Citation
Alterations of septohippocampal structure in interleukin-2 knockout mice

Material Information

Title:
Alterations of septohippocampal structure in interleukin-2 knockout mice
Creator:
Beck, Ray D
Publication Date:
Language:
English
Physical Description:
viii, 92 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Brain ( jstor )
Cholinergics ( jstor )
Cytokines ( jstor )
Hippocampus ( jstor )
Memory ( jstor )
Mice ( jstor )
Neurogenesis ( jstor )
Neurons ( jstor )
Rats ( jstor )
Receptors ( jstor )
Dissertations, Academic -- Neuroscience -- UF
Hippocampus ( mesh )
Interleukin-2 -- physiology ( mesh )
Mice Knockout -- anatomy & histology ( mesh )
Mice Knockout -- growth & development ( mesh )
Mice Knockout -- physiology ( mesh )
Neurons -- growth & development ( mesh )
Neurons -- physiology ( mesh )
Neurons -- ultrastructure ( mesh )
Neuroscience thesis, Ph. D
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references (leaves 73-91).
Additional Physical Form:
Also available online.
General Note:
Printout.
General Note:
Vita.
General Note:
Typescript.
Statement of Responsibility:
by Ray D. Beck.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022906520 ( ALEPH )
56547488 ( OCLC )
Classification:
LD1780 2004 ( lcc )

Downloads

This item has the following downloads:


Full Text










ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE IN
INTERLEUKIN-2 KNOCKOUT MICE














By

RAY D. BECK, JR.
















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004































This dissertation is dedicated to my wife Laura whose love, support, and occasional
nagging help to keep me focused on my goals.














ACKNOWLEDGMENTS

Traditionally, most acknowledgement sections begin by thanking one's advisor; in this case, such tradition is most warranted. As such, I thank Dr. John Petitto for being a wonderful mentor. Though not the stealthiest individual in the world with his penchant for crying out one's name (or impromptu nickname) exuberantly as soon as one enters anywhere within his field of vision, he is among the kindest, most supportive, enthusiastic, and knowledgeable mentors for which any graduate student could wish. Next, I would like to thank each member of my committee. I feel fortunate to be advised by such a great selection of knowledgeable and friendly people. Dr. Mike King's easygoing personality and extensive knowledge of all things stereological and cholinergic have proven invaluable in my studies. Dr. Mark Lewis' advice on statistics and experimental design, as well as his sense of humor, has been most appreciated. Dr. Jake Streit, in addition to his ability to make me laugh, always reminded me that there is more than one kind of cell in the brain. Dr. Mark Atkinson, always amiable and approachable, helped guide me in the "immunology" aspect of "neuroimmunology."

I also thank Dr. Huang Zhi. I cannot overstate how much I valued his advice on experiments and his daily conversations on topics ranging from basketball to politics to Hong Kong movies. I wish him the best of luck in his medical residency and his future as a psychiatrist. I would also like to thank Clive Wasserfall and Fletcher Schwartz for teaching me how to use the Luminex technology and Tim Vaught for teaching me the ins and outs of multiple microscopy techniques. I also thank the many technicians that


111








worked in the lab both past and present: Brent, Andrew, David, Jeannette, Dan, Jesse, and Grace. In addition to taking care of the upkeep of the lab, working with them was a pleasure. In particular, I would like to single out Andrew. He was my dearest friend here in Gainesville. I wish him the best of luck in his career as a medical doctor and hope that we will always remain friends after I leave Gainesville.

Outside of the laboratory, I thank my other friends for being my support structure. There has never been a better collection of in-the-closet geeks than Coleman, Dan, Andy, Ryan, Charles, Chris, Curtis, Nick, and Jason. They are simply the best. I would also like to thank Mozart, Michelangelo, and Sage simply because not nearly enough people thank their dogs. Obviously, they would be more likely to chew on this dissertation than read it, but I would like anyone else that does see this to know that few people are capable of matching the unconditional love that a dog has for its owner.

I also thank my family. My mother has always instilled the value of education into me. My father's love and sense of humor were crucial in the development of my personality. My sisters, Jackie and Judy, will always be among my closest friends. My Aunt Kathy and Uncle Jimmy are tremendous people who have always supported me and though they are family by marriage, our bond is stronger than blood.

Finally, I thank my lovely wife Laura. Without her, my life would be incomplete (though despite her beliefs, I could still drive effectively without her commentary from the passenger seat). She has pushed me when I needed motivation and comforted me when I need support. She is my heart and soul and I could not have achieved this dissertation without her.






iv















TABLE OF CONTENTS

page

ACKNOW LEDGM ENTS ..................................................... iii

ABSTRACT.................................................... vii

CHAPTER

I BACKGROUND AND SIGNIFICANCE.......................................................

Cytokine-Brain Interactions......................................................................................... 1
The Pleiotropic Cytokine: Interleukin-2 ....................................................................2
Interleukin-2 and the Brain ...... .................................. 4
IL-2 and the Septohippocampal System ................... ............6........... .......................6
Statem ent of the Problem ............................................................................... ........

2 ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS
RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT .......................... 11

Introduction................................................................................ .. ..... 11
Materials and Methods ..........................................13
Animals and Tissue Preparation ................... ............................ 13
Genotyping Using PCR ................................................... 14
ChA T Im m unohistochem istry ............................................................................. 14
AChE Histochemistry............................................................................ 16
Cholinergic Stereology ..................................16......... 16
Quantitative Image Analysis of AChE Staining........................ ......... 18
Cresyl Violet Staining ....................................... ...............20
Results...............................................................20
Comparison of Cholinergic Somata in the MS/vDB ........................................... 20
Density of AChE-positive Fibers in Regions of the Hippocampus .................. 20
Morphology of the Granular Cell Layer of the Lower Limb of the DG ......21
D iscussion ............................................................... ....................... 23

3 ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN
INTERLEUKIN-2 KNOCKOUT MICE ..............................................................29

Introduction..................................................................................................................... 29
M ethods .................................................................................... 32








A nim als and G enotyping .................... ... ................................ ...............32
Immunohistochemistry .......................................33
Cresyl Violet Staining .......................................34
S tereo logy ............................................................................................................ 34
Enzyme-linked Immunosorbent Assay (ELISA) Characterization of NGF and
BDN F....... .................................................................................. .......... 36
Statistical A nalysis ..................................................................................... 37
Results....................................... ...................... ........................................ 38
Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell
Number in Adult Mice............ ............................ 38
Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice........................38
A lterations in N eurotrophin Levels..................................................................38
D iscussion ................................................4..................................... ......... 40

4 INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND
NEUROGENESIS IN THE HIPPOCAMPUS .......................... ....................45

Introduction............................................................... .......................... 45
Materials and Methods .......................................... 49
A nim als and Genotyping .............................................................. ..................49
CD3' T cells and MHC II+ Microglia Immunohistochemistry ...........................50
Preparation of Serum and Brain Tissue for Cytokine Analysis .......................51
M ultiplex M icrosphere Cytokine Analysis ........................................................52
Labeling Neurogenesis with BrdU ......................................52
R esults.............................................................................................................. 54
Assessment of CD3+ T Cells and MHC II+ Activated Microglia in the
H ippocam pus ........................................................................................... 54
Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice ..............54
Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-type Mice .55 Alterations in Neurogenesis ...... ................................ 56
D iscussion.................................................................................. ...58

5 GENERAL DISCUSSION .......................................... 64

Sum m ary of the Overall Findings................................................................................ 64
Im plications .................................................................................................. 65
Caveats and Future Directions................................................................67
Concluding Remarks .............................................72

WORKS CITED ........................................73

BIOGRAPHICAL SKETCH ................... ......................... 92








vi














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE IN INTERLEUKIN-2 KNOCKOUT MICE By

Ray D. Beck, Jr.

August 2004

Chair: John M. Petitto
Major Department: Neuroscience

Interleukin-2 (IL-2) is a multifunctional cytokine involved in peripheral immune processes and may also be implicated in multiple brain functions. IL-2 gene knockout (IL-2 KO) mice exhibit deficits in several hippocampally-mediated behaviors (e.g., learning and memory) and have alterations in hippocampal structure.

In the first study, adult IL-2 KO and wild-type littermates were compared for differences in the cholinergic neurons in the media] septum and vertical limb of the diagonal band of Broca (MS/vDB; a structure associated with learning and memory). The IL-2 KO mice had significantly fewer cholinergic somata in the MS/vDB, but not in the striatum, thus indicating a selective effect of IL-2 on the MS/vDB. Cholinergic neurite density in the hippocampus was unaffected, but the length across the infrapyramidal (IP), but not the suprapyramidal (SP), granule cell layer (GCL) of the dentate gyrus was reduced.





vii








The second study assayed for variations between groups in the second largest population of neurons in the MS/vDB, the GABAergic neurons. We found no differences in these neurons in IL-2 KO animals. In 21-day-old IL-2 KO mice, we detected no changes in cholinergic neuronal number in the MS/vDB. This inconsistency with adult cholinergic neurons may be due to a failure in maintenance or might be secondary to autoimmunity. Neuronal number in the IP-GCL was also decreased, consistent with the reduction in distance detected in the first study. We also discovered that IL-2 KO correlates with a hippocampal elevation in nerve growth factor (NGF), but a reduction in the brain-derived neurotrophic factor (BDNF).

Finally, in the last study, no T cells or evidence of increased activated microglia was evident in the IL-2 KO mouse hippocampus. We noted significant elevations in several cytokines (IL-12, IL-15, IP-10, MCP-1) in the hippocampus of IL-2 KO mice. The cytokine profile of the serum was different from the hippocampus, indicating that these were not global changes throughout the bodies of the animals. We also found an alteration in hippocampal neurogenesis that appeared to be attributable to differences in male mice.

The results of these studies suggest a neuroimmune interaction that may be important in septohippocampal physiology.














viii













CHAPTER 1
BACKGROUND AND SIGNIFICANCE

Cytokine-Brain Interactions

The landmark studies of Ader and Cohen demonstrating that immune physiology could be behaviorally conditioned led to the systematic investigation of the complex interaction between the central nervous and immune systems (Ader and Cohen, 1975; Ader et al., 1982). Although the central nervous system (CNS) and peripheral immune system were once considered functionally incompatible entities separated by a nearly impermeable protective blood-brain-barrier (BBB), it is now known that there is bidirectional communication and modulation between these two systems. Cytokines have emerged as important mediators of various processes in the CNS. Their effects range from neuroinflammation in experimental autoimmune encephalomyelitis (EAE) and viral infection of the brain to neurobiological processes such as hypothalamicpituitary axis (HPA) regulation, induction of fever, sleep, analgesia, feeding behavior, and cognition (for reviews see Ader et al., 2001; Dunn, 2002; Wilson et al., 2002).

Cytokines produced both within and outside of the CNS can exert their effect on brain cells (Dunn, 2002; Streit et al., 1998). The work of Banks and others show that the BBB acts as a selective filter for peripheral cytokines (for a review see Banks et al., 2002). Multiple studies support the ability of cytokines (e.g., IL-la and -P, IL-2, IL-6, IFN-a and -y, TNF-a) to cross the BBB via different transport mechanisms (Banks et al., 1994; Banks et al., 1991; Gutierrez et al., 1993; Pan et al., 1997; Waguespack et al., 1994), and via the "leaky" circumventricular organs (CVO), four brain regions outside of


I





2


the BBB with fenestrated capillaries (Buller, 2001). Peripheral leukocytes, in particular activated T cells that enter the brain during certain conditions (e.g., EAE, facial nerve axotomy), can also release cytokines in the CNS (Hickey et al., 1991). Finally, cytokines may also interact with the brain through activation of peripheral nerves, such as IL-I stimulation of the vagus nerve, which can lead to modulation of brain functions through its afferent connections in the CNS (Maier et al., 1998). Such a cytokine-to-nerve communication pathway may not be limited to the vagus nerve, as central hyperalgesic effects are also observed by stimulating cutaneous nerves with a subcutaneous injection of IL- 10 (Fukuoka et al., 1994), TNF-a (Sorkin et al., 1997), or antibodies against TNFa (Lindenlaub et al., 2000). Thus, multiple pathways exist that allow cytokines to directly or indirectly influence the brain. The focus of this dissertation was on IL-2, which can be produced in the periphery and the CNS.

The Pleiotropic Cytokine: Interleukin-2

IL-2 was originally identified as a growth factor for bone marrow-derived T cells in 1976 (Morgan et al., 1976), and was renamed in 1979, when its pleiotropic effects between leukocytes (thus the term interleukin) became clear (Aarden et al., 1979). Further characterization of IL-2 revealed that it belongs to the four a-helix bundle family of cytokines; this family consists of cytokines with four a-helices connected by three loops in an up-up-down-down formation (Bazan, 1992). The receptor for IL-2 has a common gamma (y,) subunit shared by multiple cytokines including IL-4, IL-7, IL-9. and IL-15 (Sugamura et al., 1996); a P subunit only shared with IL-15 (Giri et al., 1995); and, in one conformation, an a subunit, which confers greater binding affinity (Leonard et al., 1984). The receptor subunits can combine in two biologically active forms: a lower








affinity heterodimer consisting of the y, and 3 subunits and a high affinity heterotrimer comprised of all three subunits (a, 1, and y,) (Ringheim et al., 1991; Takeshita et al., 1992). The 3 and Yc both possess intracellular signaling domains and in their heterodimeric form have a Kd of 10-9, whereas the addition of the a subunit forms a heterotrimer with higher affinity (Kd of 10-11) for IL-2 (Nakamura et al., 1994; Nelson et al., 1994).

In the peripheral immune system, where the physiological properties of IL-2 are most well-characterized, IL-2 has multiple biological functions, including natural killer

(NK) cell activation, T lymphocyte activation, as well as B lymphocyte differentiation (for review see Waldmann, 2002). The creation of a transgenic knockout mouse model for this cytokine suggests that the most important function of IL-2 is the maintenance of immune self-tolerance (Schmitt et al., 1994). IL-2 knockout (KO) mice develop autoimmune symptoms commonly including inflammatory bowel disease similar to ulcerative colitis in humans and advanced hemolytic anemia, although the manifestation of the phenotype is dependent on the genetic background of the knockout mice (Horak, 1995, 1996). The autoimmunity that develops when the IL-2 gene is deleted is T cell dependent (Ma et al., 1995). More recently, it has become apparent that IL-2 plays a major role in limiting T cell responses via the development of regulatory T cells (CD4'CD25* T reg cells) and other mechanisms that promote self tolerance and suppress T cell responses in vivo (Nelson, 2004). Though extensive research has characterized the effects of IL-2 in the peripheral immune system, increasing evidence indicates that IL-2 may potentially impact the central nervous system (CNS). The focus of this research project was to characterize these potential actions of IL-2 in the brain.





4


Interleukin-2 and the Brain

The effect of IL-2 on cognition and mood in humans was among the earliest findings that suggested that this cytokine might have neurobiological actions. In early clinical studies of the cognitive side effects of IL-2 therapy, 50% (i.e., 22 patients out of 44) of the subjects monitored developed cognitive changes, with 15 of them necessitating acute intervention (Denicoffet al., 1987). In addition, IL-2 therapy in patients with renal carcinoma or melanoma was found to impair spatial memory and performance in planning tasks (Capuron et al., 2001a), and induce depressive symptoms as early as two days into therapy (Capuron et al., 2000).

IL-2-brain interactions have also been investigated on an anatomical and

physiological level. In landmark studies, IL-2 was found to modulate the proliferation of oligodendrocytes (Benveniste et al., 1987; Benveniste and Merrill, 1986; Saneto et al., 1986). Exogenously administered IL-2 also has multiple effects on pituitary cells including stimulation of cortisol production and adrenal corticotropin releasing hormone release (Hanisch et al., 1994), as well as increasing pituitary cell responsiveness to corticotropin-releasing hormone (Witzke et al., 2003). IL-2 has also been shown to regulate the production and secretion of peptides from hypothalamus, in addition to pituitary cells (Karanth et al., 1993; Lapchak and Araujo, 1993; Pardy et al., 1993). Subsequent research has shown that exogenously applied IL-2 can modulate other types of central nervous system cells, such as microglia (Sakai et al., 1995). Exogenously applied IL-2 can also biphasically regulate the release of some neurotransmitters such as dopamine (Alonso et al., 1993; Petitto et al., 1997), or acetylcholine (Hanisch et al., 1993; Seto et al., 1997).








IL-2 has been shown to have neurotrophic effects on cultured neurons from several regions of the rat brain including the neocortex (Shimojo et al., 1993), cortex, striatum, medial septum, and hippocampus (Awatsuji et al., 1993). Moreover, in rat hippocampal neuronal cultures, IL-2 enhances the length and branching of hippocampal neurites and the morphology of these neurons (Sarder et al., 1996; Sarder et al., 1993). Interestingly, altered levels of IL-2 expression have been detected in schizophrenia (for reviews see Hanisch and Quirion, 1995a; Muller and Ackenheil, 1998), which is a neurological disorder where altered morphology of hippocampal neurons is well documented (for a review see Thune and Pakkenberg, 2000).

IL-2-like immunoreactivity has been localized to the hippocampal formation in rat forebrain (Lapchak et al., 1991), and detected in tissue extracts from rat and human hippocampal tissue (Araujo et al., 1989). In mouse brain, IL-2 mRNA has been found in the hippocampus (Villemain et al., 1991), and transcripts for this cytokine may be expressed in rat astrocyte cultures as well (Eizenberg et al., 1995). Our lab has cloned and sequenced the full-length mouse brain cDNAs for IL-2Rct as well as the IL2/15RP and y, subunits, and has found that the sequences of the genes expressed by lymphocytes and in brain are identical. We have also found that these genes are enriched in the hippocampus and related limbic regions. Of particular relevance to IL-2 actions in the hippocampus, in situ hybridization has shown that the IL-2/15RP and y, genes are expressed by pyramidal and granule cell neurons (Petitto and Huang, 1994, 1995, 2001; Petitto et al., 1998).





6


IL-2 and the Septohippocampal System

Hippocampal circuitry is important for encoding spatial learning and memory and some evidence supports a potential role of IL-2 in the hippocampus. IL-2, for example, alters the electrophysiological characteristics of hippocampal neurons including alterations of voltage-dependent Ca2+ currents (Plata-Salaman and ffrench-Mullen, 1993), depolarization and hyperpolarization of cultured hippocampal neurons (Hanisch and Quirion, 1995a), and changes in long-term potentiation (LTP) (Tancredi et al., 1990). IL2R subunits, as previously mentioned, are enriched in the hippocampus relative to other brain regions and exogenously applied IL-2 enhances the survival and morphological development of neurons of the hippocampus.

Knockout mice deficient in IL-2 perform significantly worse than wild-type

controls in one such test of spatial learning and memory, the Morris water maze; show an enhanced pre-pulse inhibition of the acoustic startle response (PPI; another hippocampally-mediated process); and also exhibit structural alterations in mossy fiber length (Petitto et al., 1999). Our initial studies suggest that this deficit in learning and memory is not likely due to a compromised immune system, as severe combined immunodeficient (SCID) mice perform significantly better than IL-2 KO mice in the Morris water maze. More recent studies from our lab suggest that the nature of the deficit in learning and memory seen in IL-2 knockout mice could be related to the immune status of the mother (normal heterozygote vs. autoimmune homozygote mother). In addition, the previously mentioned clinical studies of cancer patients under IL-2 treatment found alterations in spatial memory, lending some support to the potential role of IL-2 in learning and memory.





7

In the basal forebrain, the medial septum and vertical limb of the diagonal band of Broca (MS/vDB) send a large number of projections to the hippocampus, with the major neuronal phenotypes of these being cholinergic and GABAergic (Brashear et al., 1986; Kiss et al., 1990b; Kiss et al., 1990a). The septohippocampal system has been associated with learning and memory processes, with extensive data existing that link the septal cholinergic neurons that project to the hippocampus to learning and memory (Galey et al., 1994; Leanza et al., 1995), and PPI (Koch, 1996b). Moreover, variability of cholinergic fiber density in the dentate gyrus of individual mouse strains correlate with changes in spatial learning (Schwegler et al., 1996a; Schwegler et al., 1996b). Some controversy exists, however, on the relative importance of cholinergic neurons of the MS/vDB in learning and memory processes. The advent of selective toxins that target cholinergic neurons, like 192 IgG-saporin, have allowed researchers to behaviorally test animals only lacking MS/vDB cholinergic neurons, but with presumably normal distributions of GABAergic neurons. In many of these studies, animals with cholinergic septohippocampal lesions did not differ from control subjects (Baxter et al., 1996; Bizon et al., 2003; Cahill and Baxter, 2001; Chappell et al., 1998; Perry et al., 2001). Surprisingly, however, multiple other contemporary studies utilizing 192 IgG-saporin do find learning and memory deficits in the lesioned animals (Janis et al., 1998; Johnson et al., 2002; Lamprea et al., 2000; Wrenn et al., 1999). One potential explanation for this discrepancy may be that a certain threshold of cholinergic damage is necessary to elicit a deterioration in spatial learning ability (Leanza et al., 1995; Wrenn et al., 1999). Nevertheless, the negative findings are considerable and this may not explain the inconsistency well enough between groups. Another hypothesis calls into question the





8


efficacy of 192 IgG-saporin in completely removing septohippocampal cholinergic activity. In studies where 192 IgG-saporin appears to nearly completely eliminate cholinergic immunostaining of MS/vDB neurons, ACh release in the hippocampus of these lesioned animals still persists at -40% of the control values (Chang and Gold, 2004; Gold, 2003). The septohippocampal cholinergic neurons are capable of compensatory collateral sprouting (Gage et al., 1983a, 1984; Gage et al., 1983b), and functional recovery of hippocampal ACh release after complete fimbria-fornix transection (Leanza et al., 1993), which may be important as a response to damage. Thus, considering the above evidence, it is clear that cholinergic input to the hippocampus plays an important role in the complex neurobiological processes of learning and memory, though it is certainly not the only system involved.

Some evidence supports a potential trophic or regulatory role for IL-2 on the

cholinergic septohippocampal system. IL-2 enhances the survival of these septal and hippocampal neurons in culture (Awatsuji et al., 1993), and modulates the activity of choline acetyl transferase (ChAT) (Mennicken and Quirion, 1997). Also, IL-2 is a potent biphasic modulator of acetylcholine (ACh) release (Hanisch et al., 1993; Seto et al., 1997). At low concentrations (sub-pM), IL-2 stimulates the release of ACh, but at higher concentrations (nM), IL-2 inhibits ACh release. Since IL-2 is difficult to detect in the adult brain, endogenous brain IL-2 or IL-2 that crosses the BBB would bc expected to be present at a low concentration in the normal brain.

Taking into consideration 1) the neurotrophic and neuromodulatory role of IL-2 in cultured septal and hippocampal neurons, 2) the regulatory effects of IL-2 on ACh release and ChAT activity in vitro, 3) the spatial learning and memory impairments of IL-2-








deficient knockout mice and cancer patients undergoing IL-2 therapy, and 4) the enhanced concentration of IL-2R subunits in the hippocampus, I hypothesized that IL-2 may play a role in the growth and differentiation of the septohippocampal cholinergic neurons.

Statement of the Problem

Normal development of septal neurons depends on trophic factors that are

presumably secreted from the hippocampus. For example, the neurotrophins, nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) are expressed in the hippocampus and have both been shown to be important in the development, maintenance, and repair of septohippocampal neurons (Brooks et al., 1999; Conner et al., 1992; Conner and Varon, 1997; Morse et al., 1993). Perhaps, IL-2 also mediates the growth and development of septal and hippocampal neurons. This cytokine may act as a growth factor by itself, or signal the release of growth factors from neurons or glia in the hippocampal area.

In other investigations, the modulatory and growth-promoting effects of IL-2

were determined by either adding exogenous IL-2 to cultures or by injecting exogenous IL-2 into the brain (in most cases, species non-specific, e.g., human IL-2 in rats or mice). Thus, the studies utilizing exogenous IL-2 administration have several potential shortcomings: 1) the cytokine has a short half-life, and the amount of IL-2 delivered may not reflect physiologically relevant concentrations in the CNS in vivo (e.g., chronic vs. acute dosing), 2) the nature of the injections disrupts the BBB, causing potentially even more IL-2, as well as other cytokines, from the peripheral immune system to enter into the brain, 3) it is unknown during which period in neurodevelopment that IL-2 might exert these postulated effects, and 4) non-species specific IL-2 may have neurotoxic





10


effects, such as T and B cell invasion of the brain, angiogenesis, changes in the composition of the extracellular matrix, myelin damage, and neuronal cell loss seen in rats administered human IL-2 intracerebroventricularly via minipumps (Hanisch et al., 1996; Hanisch et al., 1997b). Thus, my approach was to use IL-2 knockout mice. These studies were the first to investigate the consequences of the absence of IL-2 on aspects of brain development and maintenance of the septohippocampal system in vivo. Our laboratory had found the aforementioned behavioral alterations in IL-2 knockout mice, and therefore another important goal of my research was to test the hypotheses regarding the neurobiological and neuroimmunological alterations that may underlie these behavioral abnormalities.













CHAPTER 2
ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS
RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT Introduction

One of the earliest observations suggesting that cytokines could influence brain function in humans came from cancer treatment trials in which interleukin-2 (IL-2) was found to induce cognitive dysfunction and other untoward neuropsychiatric side effects in patients (Denicoff et al., 1987). Although basic research has demonstrated that IL-2 can modulate different aspects of central nervous system (CNS) function, some of IL-2's most prominent neurobiological actions occur in the hippocampal formation and related limbic regions, where receptors for this cytokine are enriched (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto et al., 1998).

Exogenously administered IL-2 has effects on a number of parameters of septal and hippocampal neuronal function including trophic effects on cultured fetal septal and hippocampal neurons (Awatsuji et al., 1993; Sarder et al., 1996; Sarder et al., 1993). IL-2 may also modify cellular and molecular substrates of learning and memory such as longterm potentiation (Tancredi et al., 1990), and multiple parameters of cognitive behavioral performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al., 1999; Nemni et al., 1992). Moreover, the neurotrophic and neuromodulatory actions of IL-2 have been implicated in abnormal hippocampal development associated with schizophrenia (Ganguli et al., 1995; Licinio et al., 1993; McAllister et al., 1995).



11





12


The potential effects of IL-2 on cholinergic neurons are particularly relevant to this study. In addition to the aforementioned trophic effects of IL-2 on cultured septal neurons, IL-2 is among the most potent modulators of acetylcholine (ACh) release from cultured septohippocampal neurons (Araujo et al., 1989; Hanisch et al., 1993; Seto et al., 1997), and can also modulate its precursor enzyme, choline acetyltransferase (ChAT) in fetal neurons (Mennicken and Quirion, 1997). Alterations in the cytoarchitecture of cholinergic septohippocampal neurons have been shown to correlate with differences in spatial learning ability in mice (Schwegler et al., 1996a; Schwegler et al., 1996b). We found that IL-2 knockout (IL-2 KO) mice exhibited impaired learning and memory performance, sensorimotor gating, and reductions in hippocampal infrapyramidal mossy neuronal fiber length (Petitto et al., 1999), a factor shown previously to correlate positively with spatial learning ability (Schopke et al., 1991; Schwegler and Crusio, 1995; Schwegler et al., 1988).

In the present study, we therefore sought to test the hypothesis that loss of IL-2 would result in abnormal neurodevelopment of septal cholinergic neurons that project to the hippocampus. Since extensive data document that these neurons play a critical role in learning and memory performance (Galey et al., 1994; Leanza et al., 1995), and given the various in vitro neurotrophic and neuromodulatory effects of IL-2 on developing septohippocampal cholinergic neurons, we postulated that IL-2 KO mice would have fewer cholinergic neurons in the medial septum and vertical limb of the diagonal band of Broca (MS/vDB) and a reduction in the cholinergic axonal density in the hippocampus. To accomplish this goal, IL-2 KO and wild-type littermates were compared using stereological techniques to count MS/vDB cholinergic somata stained with ChAT





13

immunohistochemistry, and image analysis methods to measure the density and distribution of cholinergic neurites in several regions of the hippocampus labeled for acetylcholine esterase (AChE), a reliable marker of cholinergic axons (Hedreen et al., 1985).

Materials and Methods

Animals and Tissue Preparation

Mice used in these experiments were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2 heterozygote by IL-2 heterozygote crosses. The polymerase chain reaction (PCR) was used to genotype the offspring post-weaning (see below). The IL-2 KO mice, obtained originally from the NIH repository at Jackson Labs, were derived from ten generations of backcrossing onto the C57BL/6 background. Mice were housed under specific pathogenfree conditions. Animals used in these experiments were 8-12 weeks of age.

Each animal was anesthetized with sodium pentobarbital (50 mg/kg) and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS). The brains were removed and fixed overnight in 4% paraformaldehyde followed by overnight equilibration in 30% sucrose cryoprotective solution, and then were snap frozen in isopentane (-800C) for storage. The brains were equilibrated to -200C prior to cryostat sectioning into 40 plm slices in the coronal plane, collected into individual wells of polystyrene 24-well plates (NUNC 1147), and stored free-floating at 40C in PBS for histochemistry. Every third section was processed for ChAT immunohistochemistry, AChE histochemistry, or cresyl violet Nissl staining.





14


Genotyping Using PCR

The genotypes of all mice were determined by the PCR. PCR reactions were performed using a 25 pl total reaction volume containing I pM each of forward and reverse primers, 0.1 pig genomic DNA, 0.2 mM of each dNTP, 0.3 pl1 Taq DNA polymerase and amplified using a thermal cycler with a heated evaporation cover (Ericomp). The cycling parameters were hot start 950C (3min), denaturing 940C (30 sec), annealing 640C (30 sec), extension 720C (45 sec) with a final extension step of 4 min. Thirty cycles were used for these experiments. The 5' and 3' primers for the IL-2 KO (500 bp knockout band amplified) were 5'-TCGAATCGCCAATGACAAGACGCT3' and 5'-GTAGGTGGAAATTCTAGCATCATCC-3'. The 5' and 3'primers for the wild type (324 bp wild type band amplified) were 5'CTAGGCCACAGAATTGAAAGATCT-3' and 5'GTAGGTGGAAAATTCTAGCATCATCC-3'. ChAT Immunohistochemistry

Free-floating 40-pm sections were incubated for 20 minutes in 1% hydrogen peroxide (H202) to quench endogenous peroxidative activity. The sections were then washed twice in PBS and blocked for 1 hr in 200 gl/well 3% normal goat serum (NGS). After this incubation, the sections were incubated overnight in the primary antibody, rabbit anti-ChAT (Chemicon AB143; 1:2000 dilution in PBS with 0.3% Triton X-100 and 1% NGS, 200 jl/well). The next day, the sections were washed twice in PBS and incubated overnight in the secondary antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-100 and 1% NGS). The sections were then washed twice in PBS and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS)





15


for 2 hrs followed by two washes in PBS. The sections were developed in 0.5 mg/ml 3,3'-diaminobenzidine (DAB), 0.2 mg/ml urea H202 for approximately 5 min and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped. Figure 2-1 shows an example of ChAT immunostained section of the MS/vDB.







i I

I
















the diagonal band of Broca. The scale bar represents 200
I





16


AChE Histochemistry

AChE histochemistry was used as a marker of cholinergic innervation of the

hippocampus (Woolf et al., 1984). Brain sections were collected in individual wells of 24-well plates containing 250 pl/well 0.1 M pH 6.0 acetate buffer (AB). The sections were washed twice with AB, then placed in 200 pl preincubation solution consisting of aqueous 5 mM sodium citrate, 3 mM cupric sulfate, and 0.5 mM potassium ferricyanide. The sections were incubated for 20 minutes at room temperature on a shaker a low speed. After the preincubation period, 200 gl of the incubation solution was added to each well consisting of the same make-up as the preincubation solution supplemented with 4.84 mM acetylthiocholine iodide and 0.4 mM ethopropazine. The multi-well plate was packed on top of crushed ice and microwaved at 200 W for 2 minutes. The solution was then removed and the sections were washed twice in 0.05 M TRIS pH 7.6 buffer followed by AB. The reaction product was intensified with 0.5 mg/ml DAB, 2.5% nickel sulfate, and 0.01 % H202 in AB for 5-7 min or until definitive staining could be detected in the hippocampal subregions. Sections were then mounted on slides, dehydrated in grade ethanol washes, cleared in xylenes, and coverslipped for imaging. Cholinergic Stereology

Stained cholinergic neuronal somata of the MS/vDB were counted using the

software MCID 5.1 and the three-dimensional counting box (optical dissector) method described by Williams and Rakic (Williams and Rakic, 1988). All stereology was performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2 microscope with a motorized x-y stage made by Imaging Research, Inc. The latter is capable of making movements as fine as 0. 1 ptm. Every third section through the anterior-posterior extent of the septal region was sampled. The regions to be counted





17


were outlined at I Ox magnification and the size of the counting boxes were generated to be approximately 5% of the most rostral, and therefore, smallest, area of the MS/vDB (defined by the section where the corpus collosum first joins in the midline). The size of the outlined count regions, but not the counting box, varied depending on where the individual section was taken from the rostral to caudal extent of the MS/vDB. The defined counting box was approximately 2-2.5% of the outlined count area of the largest single section of the MS/vDB.

To assess whether the predicted septal cholinergic alterations in IL-2 KO mice might be associated with a general effect on cholinergic neurons in the brain, striatal cholinergic somata were also counted in the right hemisphere in the sections that also contained the MS/vDB. Except for a different magnification used to outline the striatum

(5X), the sampling parameters were identical to those used to generate estimates of septal cholinergic neuron number. The guard volume was set at 1 p.m for the top and bottom of the section and the counting cubes were randomly distributed throughout the user-defined count area with a total sampling frequency of 25% (-8.3% of the total area since every third section was sampled). Only somata that were clearly and distinctly stained were counted. Each counting box was examined at 40x magnification (20x for striatal neurons) and the computer-assisted focus was used to scan from the top to the bottom of the counting box. Cells were counted only if they were either completely inside of the counting box, or partially inside of the box on the top, back, or left side. They were not counted if they fell outside of the box or crossed into the box anywhere on the bottom, front, or right side. The cells that were counted were labeled on the monitor by clicking the mouse pointer on each cell and the MCID software recorded the number of marks.





18


The MCID software interpolated the total volume of the MS/vDB based on the volume of the count areas defined by the user. Cell density (Nv) was estimated by dividing the total number of cells counted by the volume of the counting boxes, which was also tracked by the software. The total cell number was estimated by multiplying Nv by the total volume.

Quantitative Image Analysis of AChE Staining

For quantitative analysis of AChE staining, we modified previously described methods used to measure intensity of staining and comparisons of normalized length across CAl, CA3b, and the suprapyramidal (SP) and infrapyramidal (IP) blades of the dentate gyrus (DG) (King et al., 1989; Schwegler et al., 1996b). Images of the hippocampus in each tissue section sampled from a light microscope (Olympus BH-2) were relayed by digital video camera (Hitachi KP-D5 81) to a computer frame grabber (Flashpoint 128, Integral Technologies) and digitized to 640x480 pixel images with 256 gray levels from black to white. Imaging software (Image Pro Plus v.4.0, Media Cybernetics) was utilized to define a broad sampling traverse across various areas of the hippocampus, approximately perpendicular to the cell body layers. For CAl and CA3b, this traverse extended from the alveus to the hippocampal fissure. For the DG, the line extended from the hilus to hippocampal fissure or to the pial surface for the SP or IP limb of the DG, respectively. AChE-containing fiber density was estimated by using the gray level of each point along the line, which was calculated by averaging the intensity value of each pixel across the width of the sampling band (e.g. a traverse with a width of 46 pixels would have an average of 46 measurements approximately parallel to each data point). Gray level intensity measurements were used to estimate the AChE reaction product density at each point along the traverse. The mean pixel intensity of a small box





19

positioned in the corpus collosum was used to represent the background staining intensity of tissue containing little cholinesterase activity. All of the average intensities along the line were then plotted by traverse position to illustrate the quantitative patterns of AChE distribution across each subregion. Conspicuous inflections marking the transition from alveus (Alv) to stratum oriens (SO, in CA1, CA3b), from stratum lacunosum-moleculare (Lmol) to dentate molecular layer (Mol) (hippocampal fissure (HiF); CAl, CA3, DG), and polymorph zone (PoDG) to granule cell layer (GCL) were used to align traverses across sections and animals (Figure 2-2). Also, length-normalized comparisons of the



















Figure 2-2. A micrograph of the sampling regions utilized for image analysis of AChEstaining. Alv=alveus, DG=dentate gyrus, GCL=granular cell layer
(IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus,
Lmol=lacunosum moleculare, Mol-molecular layer, Pi-pial surface, SO=stratum
oriens, SP=septum pellucidum, SR=stratum radiatum. The scale bar represents
200 pm.
measured regions were made by converting the sampling lengths to 100 points using the software Matlab v.5.3. Dependent variables were the absolute length of traverses, AChE intensity at anatomically identifiable inflection points (IP and SP bands, Lmol, dentate sofwae Mtlb v5.. Dpedet vrible wre heabslue engh f tavrse, 8h

inenit a a at mcalyid ntfibe nfecio p its(I ad P ads L ol d ntt





20

inner Mol) and subregions (SO, stratum radiatum (SR), dentate outer Mol), and derived values for absolute and relative positions of, and distances between, these landmarks. For each animal, three hippocampal slices were measured on both the left and right hemispheres of the brain and the measured intensities or distances were averaged together for statistical analysis.

Cresyl Violet Staining

Every third section was Nissl stained to provide a qualitative view of the

boundaries between various forebrain regions. The sections were placed on slides and allowed to air-dry. The slides were immersed in 600C cresyl violet for 45 sec, washed in running distilled water to remove the excess cresyl violet, dehydrated in graded ethanol, cleared in xylenes, and coverslipped.

Results

Comparison of Cholinergic Somata in the MS/vDB

Figure 2-3 shows the total number of ChAT-positive cells stereologically counted from the MS/vDB of IL-2 KO and wild-type mice. As seen in this figure, the IL-2 KO mice had approximately 26% fewer cholinergic somata in this region than wild-type controls. An ANOVA confirmed that this group difference was statistically significant (F(1,16)=8.6, p=.01). By contrast, the number of ChAT-positive somata in the striatum of IL-2 KO and wild-type mice were not different. There were no significant gender differences in either brain region.

Density of AChE-positive Fibers in Regions of the Hippocampus

The average AChE-staining intensity curves were generated by defining a region across CAI, CA3b, and the SP and IP blades of the DG. Figure 2-4 shows the intensity curves that were generated from imaging the AChE-histochemically stained sections for





21


Cholinergic Somata in the MS/vDB


2000


E 1500
O

o 1000
CL

500



F-- Wd-Type Groups
= IL-2 Knockou I

Fig. 2-3. ChAT-positive somata are significantly reduced in the MS/vDB of IL-2
knockout mice. Each bar represents the mean + SEM of 9 animals per group.
*p=0.01.

CAl (Figure 2-4a), CA3 (Figure 2-4b), the SP-GCL (Figure 2-4c), and IP-GCL layer of the DG (Figure 2-4d). Repeated measures ANOVA was performed on regions of the average normalized curves selected by areas that appeared to deviate between the groups. None of these areas, however, were found to differ between IL-2 KO and wild-type mice. Morphology of the Granular Cell Layer of the Lower Limb of the DG

There were no differences in the groups for the Y-axis (intensity) data. The variations in the patterns of the DG curves X-axis (distance) were also compared. Distances were compared by defining the point of lowest intensity in the regions that the curves indicated as each transition between the regions of interest. Distances are reported as a percentage of the total distance across each curve. The IP blade of the DG was broken into three regions: the Hil, the IP-GCL, and the Mol (Figure 2-4d). As depicted in





22



A) Average AChE Intensity of CA B) Average AChE ntensty of CA3


LL
9E0 t
A


MoC. S L i, Hit fi

SO SP SR Lmo I SO SP SR Lmno I
0 10 30 40 50 W4 ) 40 000 0 Ic 20 30 40 00 M 4 90 I00 Cl anoe ( Of T dal) Cmmsl e (% of Ahtal)

C) Average AC Intensty the SuprCpyAamiC Limb D) Average AChE Inentty of the Infrapyral das Lmb of the Dentate Gyrus of the Dentale Oyrus





knock m GCL
ay Socap fu H hl

CA II


Mol SP-GCL Hil b Hil I P-GCL Mol tPi
0 It 3 e0 t oa a s ut r uc0 i SO 3n 40 0 60 7R a 40 B )00 ofrr ( o T a) m sl i.e (% Tl al)
IL,2 ~ooul
0 Wo.Tmp


Fig.2-4. IL-2 knockout mice do not differ in measures of average intensity of AChEstaining across CA, CA3b, and the suprapyramidal and infrapyramidal layers of
the DG. Each curve represents the mean of 9 wild-type (open circles) or 9 IL-2
knockout mice (closed circles). Alv=alveus, GCL=granular cell layer
(IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus,
Lmol=lacunosum moleculare, Mol-=molecular layer, Pi=pial surface, SO=stratum
oriens, SP=septum pellucidum, SR=stratum radiatum. A) The CAl curve shows
the average intensity from the Alv to the Hif. The arrows delineate the transitions
between the hippocampal substructures including the SO, SP, SR, and Lmol. B) The pattern of the peaks and valleys of the CA3 region curve is similar to that of
the CAl. C) The intensity curve of SP-GCL begins at the Hif and terminates at
the dorsal border of the IP-GCL. Arrows delimit the borders of the Mol, SP-GCL, and Hil. D) The curve representing the IP-GCL begins at the ventral border of the
SP-GCL and continues to the pial surface. The second and third solid arrow
define the borders in wild-type mice between the Hil and IP-GCL and the IP-GCL
and Mol, respectively. The broken arrows define these same borders in IL-2
knockout mice.





23


Figure 2-5, the distance across the IP-GCL was significantly reduced in the IL-2 KO mice compared to wild-type mice (F(1,16)=9.2, p=0.008). The distances across the Hil and Mol, however, were not significantly different. The SP blade of the DG was separated into three regions: the Mol, the SP-GCL, and the Hil (Figure 2-4c). There were no significant differences in length between groups across any of the internal blade regions.


Distance Across GrDG of the External Blade 35
30
25
F

F 15

10
5
0

Groups
--] Wid-Type
SIL-2 Knockout

Fig. 2-5. Distance across the GrDG of the external blade was significantly decreased in
IL-2 knockout mice compared to wild-type mice. Each bar represents the mean +
SEM of nine animals per group. *p=0.008.

Discussion

These data are the first to demonstrate that loss of endogenous IL-2 results in

reduction in the number of MS/vDB cholinergic neurons and structural alterations in the morphology of the dentate gyrus. Given the role of septohippocampal cholinergic neurons in learning and memory (Galey et al., 1994; Leanza et al., 1995; Schwegler et al., 1996a; Schwegler et al., 1996b), sensorimotor gating (Caine et al., 1992; Curzon et al.,





24

1994; Koch, 1996a), and the aforementioned influence of IL-2 on parameters of cholinergic function, the present findings are consistent with alterations in these behavioral measures that we have reported previously in IL-2 KO mice (Petitto et al., 1999).

In this study, the number of cholinergic cell bodies in the MS/vDB of wild-type mice were comparable to those reported previously for C57BL/6 mice (Schwegler et al., 1996b). Cholinergic somata were reduced by 26% in IL-2 KO mice as compared to wildtype mice. This was not a general effect on cholinergic neurons in the brain, however, as striatal ChAT-positive neurons were not significantly affected. This finding is also consistent with previous research showing that exogenous IL-2 has potent effects on ACh release from septohippocampal neurons, whereas cholinergic interneurons in the striatum do not respond to IL-2 (Hanisch et al., 1993). One potential caveat of using ChAT as a marker of cholinergic neurons is that differences in cell counts may be due to a decrease in ChAT labeling intensity rather than a reduction in cholinergic neurons (Ward and Hagg, 2000). This would appear unlikely, however, as we did not note any appreciable differences in the staining intensity between groups.

Alterations in AChE-staining intensity did not differ as we hypothesized. It is not likely that the unexpected lack of difference in AChE-staining intensity would be attributable to the length normalization technique used, as the patterns of the average intensity curves seen in Figure 2-4 were remarkably similar between the groups. This unexpected result may, however, be due to compensatory sprouting of the surviving MS/vDB neurons during development. Indeed, numerous other studies have found that septohippocampal neurons can undergo compensatory sprouting in response to injury






25


(Cassel et al., 1997; Gage and Bjorklund, 1987; Gage et al., 1984; Gage et al., 1983b), and in animal disease models such as Alzheimer's transgenic mice (Bronfman et al., 2000). Future studies are needed to address this issue more directly.

Another significant finding of the current study was that IL-2 KO mice exhibited structural alterations in the distance across the IP-GCL. The neurons of the GCL have been associated with learning and memory (Collier and Routtenberg, 1984; Conrad and Roy, 1993; McLamb et al., 1988; Nanry et al., 1989; Walsh et al., 1986), and are also a target for septohippocampal cholinergic axon termination (Makuch et al., 2001). In situ hybridization studies have found the GCL to be enriched in IL-2 receptors (Petitto and Huang, 2001; Petitto et al., 1998), supporting a possible role for IL-2 in the observed structural alterations. Also, GCL development progresses from the SP layer to the IP layer (Bayer, 1980). The differences seen in the IL-2 KO mice in this study may indicate a failure of these late stage granule cells to fully develop or survive. Whether these structural changes are due to a reduced number of GCL cells or a decrease in the cell body size of these neurons requires further investigation.

The most likely mechanism whereby loss of IL-2 results in these changes in the septohippocampal cholinergic system would appear to be due to the absence of its neurotrophic actions during development. As noted earlier, IL-2 enhances neurite extension and survival of cultured fetal septal and hippocampal neurons (Awatsuji et al., 1993; Hanisch and Quirion, 1995a; Sarder et al., 1996; Sarder et al., 1993), and thus, the absence of these intrinsic effects of IL-2 could account for the observed neuroanatomical alterations. Another mechanism that may account for these findings is the possibility that the loss of endogenous IL-2 may result in lower levels of tonic ACh release during





26


critical periods of neurodevelopment. Release of ACh by developing neurons has been shown to be important for growth cone guidance (Zheng et al., 1994), neuronal growth and differentiation, synaptic plasticity (Lauder and Schambra, 1999), and survival of newly developed neurons (Knipper and Rylett, 1997). In fact, some evidence indicates that ACh released from developing neurons may engage in a positive feedback mechanism with nerve growth factor (NGF) (Knipper et al., 1994), a member of the neurotrophin family that is essential for the normal development of septal cholinergic neurons (Arimatsu and Miyamoto, 1991; Hartikka and Hefti, 1988; Mobley et al., 1986; Ruberti et al., 2000). In a series of studies, Quirion's laboratory has demonstrated that IL-2 is among the most potent modulators of ACh release from mature brain slices and fetal neurons in vitro, and can upregulate ChAT in fetal septal neurons in culture (Hanisch et al., 1993; Mennicken and Quirion, 1997; Seto et al., 1997). It is therefore possible that the loss of such potent actions of IL-2 during development could account, in part, for the cytoarchitectural alterations found in this study. Indeed, both IL-2's neurotrophic effects and action on cholinergic release may well be operative and interactive with one another. Nevertheless, the IL-2 KO mice do not exhibit complete loss of septal cholinergic neurons suggesting that the effects of IL-2 on MS/vDB neurons are likely secondary to other trophic factors like NGF.

These experiments do not enable us to differentiate between the contributions of the loss of central versus peripheral IL-2, and thus, it remains to be determined whether these septohippocampal cholinergic abnormalities are due primarily to the absence of central, peripheral, or a combination of both sources of IL-2. There is some evidence that endogenous IL-2 may be produced in neuronal areas of the mammalian hippocampal





27


formation, where its release may regulate the development and function of septal cholinergic neurons projecting to the hippocampus. IL-2-like immunoreactivity has been localized to the hippocampal formation in rat forebrain (Lapchak et al., 1991), and detected in tissue extracts from rat and human hippocampal tissue (Araujo et al., 1989). In mouse brain, IL-2 mRNA has been found in the hippocampus (Villemain et al., 1991), and transcripts for this cytokine may be expressed in rat astrocyte cultures as well (Eizenberg et al., 1995).

In the periphery, absence of endogenous IL-2 leads to an immunodysregulation that produces loss of self-tolerance and IL-2 KO mice eventually develop generalized systemic autoimmune disease (although C57BL/6-IL-2 KO mice develop clinical signs of systemic autoimmunity at a substantially slower rate than other strains such as Balb/c or C3H) (Petitto et al., 2000). Therefore, it is reasonable to speculate that the neuroanatomical alterations found in the IL-2 KO mice result from peripheral autoimmune processes. Autoimmunity could impact on brain development or induce neurodegeneration. The former seems unlikely, however, since IL-2 KO mice do not express the first signs autoimmunity (e.g., splenomegaly) until at least three to four weeks after birth (Horak, 1995); by this time, septohippocampal development should already be complete (Bender et al., 1996; Chandler and Crutcher, 1983; Super and Soriano, 1994; Yoshida and Oka, 1995). Furthermore, the likelihood that these neuroanatomical alterations may be due to autoimmune-induced degeneration of existing neurons also seems unlikely, since lymphocytes cannot be detected in the brain of adult IL-2 KO mice (Petitto et al., 1999). Nonetheless, since autoimmunity has been associated with cognitive changes in both animals and humans (Lal and Forster, 1988; Sakic et al., 1997;






28


Sakic et al., 1993), more subtle autoimmune processes may be at play in the IL-2 KO mice (e.g., autoantibodies). It would be of interest to explore the observed cholinergic changes in old versus neonatal mice to determine if the abnormalities increase with age due to neurodegeneration, or are primarily the result of abnormal development. Such knowledge will then enable us to develop a more specific model to test relevant hypotheses involving IL-2 at specific anatomical sites in the septohippocampal cholinergic system.

In summary, these data demonstrate that loss of endogenous IL-2 results in reduction in the number of cholinergic neurons in the MS/vDB and alterations in the structural morphology of dentate projection fields. These findings extend our previous experiments showing that spatial learning and hippocampal mossy fiber length are abnormal in IL-2 KO mice (Petitto et al., 1999). Further research is needed to determine whether these outcomes in IL-2 KO mice may be due to the absence of central or peripheral IL-2 during neurodevelopment (or some combination of both sources), neurodegeneration secondary to peripheral autoimmunity, or other factors associated with the absence of IL-2.













CHAPTER 3
ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN INTERLEUKIN-2 KNOCKOUT MICE

Introduction

Interleukin-2 (IL-2) has been implicated in the pathogenesis of multiple sclerosis and several major neuropsychiatric disorders such as Alzheimer's disease, schizophrenia, and Parkinson's disease (Hanisch and Quirion, 1995b). Furthermore, in case studies of humans receiving IL-2 treatment for cancer therapy, prolonged exposure to IL-2 was found to induce cognitive dysfunction and other untoward neuropsychiatric side effects (Denicoff et al., 1987). Although IL-2 has been shown to be capable of modulating different aspects of central nervous system (CNS) function, many of its known effects in the limbic system occur in the hippocampal formation, where receptors for this cytokine are enriched (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto et al., 1998). IL-2 may, for example, modify cellular and molecular substrates of learning and memory such as long-term potentiation (Tancredi et al., 1990), and can affect multiple parameters of cognitive behavioral performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al., 1999; Nemni et al., 1992). IL-2 can provide trophic support to primary cultured neurons from multiple region of the rat brain, including the hippocampus and medial septum (Awatsuji et al., 1993; Sarder et al., 1993), and positively affects the morphology of neurite branching from rat hippocampal cultures (Sarder et al., 1996; Sarder et al., 1993). Furthermore, IL-2 has been shown to be one of the most potent modulators of



29





30


acetylcholine (ACh) release from rat hippocampal slices (Hanisch et al., 1993; Seto et al., 1997), and can also increase the activity of its precursor enzyme, choline acetyltransferase (ChAT) (Mennicken and Quirion, 1997).

Previously, we found that IL-2 knockout mice (IL-2 KO) exhibited impaired

learning and memory performance, sensorimotor gating, and reductions in hippocampal infrapyramidal mossy neuronal fiber length (Petitto et al., 1999), a factor which correlates positively with spatial learning ability (Schopke et al., 1991; Schwegler and Crusio, 1995; Schwegler et al., 1988). We also found in studies of IL-2 KO mice in vivo, there was a marked reduction in cholinergic somata in medial septal/vertical limb of the diagonal band of Broca (MS/vDB) region, as well as decrease in the distance across the infrapyramidal granule cell layer (IP-GCL) of the dentate gyrus (DG) (Beck et al., 2002). Variation in the cytoarchitecture of cholinergic septohippocampal neurons correlate with differences in spatial learning ability in mice (Schwegler et al., 1996a; Schwegler et al., 1996b).

Research has shown that the neurotrophins, nerve growth factor (NGF) and brainderived neurotrophic factor (BDNF) expressed in the hippocampus, can be important in the development, maintenance, and repair of septohippocampal neurons in vitro (Arimatsu and Miyamoto, 1991; Conner and Varon, 1997; Gahwiler et al., 1987; Hartikka and Hefti, 1988; Morse et al., 1993). Similar trophic effects have been noted in studies utilizing infusion of exogenous neurotrophins in vivo (Hagg et al., 1990; Morse et al., 1993), and in studies of transgenic and knockout mice (Ruberti et al., 2000; Ward and Hagg, 2000). In the peripheral immune system of multiple animal species, both NGF and BDNF are expressed by T lymphocytes (Braun et al., 1999; Kerschensteiner et al., 1999;





31


Mizuma et al., 1999; Moalem et al., 2000). Though IL-2 regulates several aspects of T cell function, the production or release of NGF and BDNF from T lymphocytes by IL-2 has not been tested directly(Carter et al., 1998; He and Malek, 1998), and conversely, it is not known whether NGF and BDNF can modulate IL-2 in the brain.

In the present study, we sought to expand our previous findings that loss of endogenous IL-2 in knockout mice led to reductions in cholinergic neurons of the MS/vDB and a decrease in the distance across the IP-GCL of the DG. First, we compared parvalbumin (Parv)-labeled somata in 8-12 week old IL-2 KO and wild-type littermates to test the hypothesis that the loss of IL-2 is selective for cholinergic, but not GABAergic cell bodies loss in the MS/vDB (e.g., not a general effect occurring on all neurons in this region of the brain). Second, we compared cholinergic MS/vDB somata between younger wild-type and IL-2 KO mice at postnatal day 21 (P21), an age where septohippocampal development in mice is nearly complete (Armstrong et al., 1987; Gould et al., 1991; Makuch et al., 2001). This age also precedes the development of autoimmune disease in IL-2 KO mice of the C57BL/6 background, e.g., absence of splenomegaly, lymphadenopathy, and inflammatory bowel disease. Third, we sought to expand on the previous finding that there was a reduction in distance across IP-GCL by performing stereological cell counts of Nissl-stained dentate gyri of IL-2 KO and wildtype littermates to determine if the reduction in distance could be contributed to a reduction in granule cell number. Finally, we tested the hypothesis that loss of IL-2 may impact the expression and release of the neurotrophins, NGF and BDNF, which may contribute to the MS/vDB cholinergic and GCL deficits that we observed previously.





32


Methods

Animals and Genotyping

Mice used in these experiments were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2 heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the NIH repository at Jackson Laboratories, were derived from ten generations of backcrossing onto the C57BL/6 background. Mice were housed under specific pathogenfree conditions. Animals used in these experiments were either 21-days-old (for ChAT immunohistochemistry) or 8-12 weeks of age. All experiments were performed with independent groups of animals. The specific animal numbers utilized are reported at the beginning of each method descriptions below.

The genotypes of all mice were determined by the polymerase chain reaction (PCR). PCR reactions were performed using 25 p1 total reaction volume containing 1 pM each of forward and reverse primers, 0.1 tpg genomic DNA, 0.2 mM of each dNTP,

0.3 p1 Taq DNA polymerase, and amplified using a thermal cycler with a heated evaporation cover (Ericomp). The cycling parameters were hot start 950C (3min), denaturing 940C (30 sec), annealing 640C (30 sec), extension 720C (45 sec) with a final extension step of 4 min. Thirty cycles were used for these experiments. The 5' and 3' primers for the IL-2 KO (500 bp knockout band amplified) were 5'TCGAATCGCCAATGACAAGACGCT-3' and 5'GTAGGTGGAAATTCTAGCATCATCC-3'. The 5' and 3'primers for the IL-2 wild type (324 bp wild type band amplified) were 5'-





33

CTAGGCCACAGAATTGAAAGATCT-3' and 5'GTAGGTGGAAAATTCTAGCATCATCC-3'. Immunohistochemistry

For Parv stereological cell counts, six animals per group were used and for 21day-old ChAT stereological cell counts, seven animals per group were used. Each animal was anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose of 0.015 ml injection cocktail/g body weight and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS). The brains and spleens were removed and fixed overnight in 4% paraformaldehyde, followed by overnight equilibration in 30% sucrose cryoprotective solution, and then, were snap frozen in isopentane (-800C) for storage. The spleens were weighed to assay for relative splenomegaly of IL-2 KO vs. wild-type mice. The brains were equilibrated to -200C prior to cryostat sectioning into 50 p.m slices in the coronal plane, collected into individual wells of polystyrene 24-well plates (NUNC 1147), and stored free-floating at 40C in PBS for histochemistry. Every third section was processed for Parv or ChAT immunohistochemistry, or cresyl violet Nissl staining.

Free-floating 50-ptm sections were labeled for Parv and ChATimmunohistochemistry as described previously (Beck et al., 2002). Briefly, they were incubated for 20 minutes in 1% hydrogen peroxide (H202) to quench endogenous peroxidative activity. The sections were then washed and blocked for 1 hr in 200 pl/well 3% normal goat serum (NGS). After this incubation, the sections were incubated overnight in the primary antibody, rabbit anti-ChAT (Chemicon; 1:2000 in PBS with

0.3% Triton X-100 and 1% NGS, 200 pl/well) or rabbit anti-Parv (Chemicon; 1:1000 in PBS with 0.3% Triton X-100 and 1% NGS, 200 pl/well). The next day, the sections were





34

washed and incubated overnight in the secondary antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-100 and 1% NGS). The sections were then washed and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS) for 2 hrs. The sections were developed in 0.5 mg/ml 3,3'-diaminobenzidine (DAB), 0.2 mg/ml urea H202 for approximately 5 min and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped. Cresyl Violet Staining

For stereological cell count of Nissl-stained granule cells, seven animals per

group were used. The tissue for this assessment was selected, because it originated from animals utilized in a previous study of cholinergic differences in IL-2 KO mice (Beck et al., 2002). Every third section was Nissl stained to provide a qualitative view of the boundaries between various forebrain regions, as well as labeling of hippocampal granule cells layers (GCL) for stereological counting. The sections were placed on slides and allowed to air-dry. The slides were immersed in 250C cresyl violet for 10 min, washed vigorously in rapid exchanges of distilled water to remove the excess cresyl violet, dehydrated in graded ethanol, cleared in xylenes, and coverslipped. Stereology

Stained neuronal somata of the MS/vDB or IP and SP-GCL were counted using the software MCID 5.1 as previously described (Beck et al., 2002). All stereology was performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2 microscope with a motorized x-y stage made by Imaging Research, Inc. Every third section through the anterior-posterior extent of the MS/vDB or IP and SP-GCL regions were sampled. The regions to be counted were outlined at 10x (MS/vDB) or 20x (IP and SP-GCL) magnification and the size of the counting boxes were generated to be





35

approximately 2-2.5% of the outlined count area of the largest single section of the areas of interest.

The rostral border of the GCL count area was defined as the first section where the dentate granule cell layer clearly separated from the pyramidal layer of CAl. The caudal border was defined as the first section where the habenular commisure was visible in the third ventricle. Only cells that could clearly be determined to be part of either the IP or SP-GCL were counted; any cells in the area where the IP and SP-GCL connected were left uncounted. Furthermore, only Nissl-stained cells with clearly visible nucleoli were counted. For the MS/vDB, the rostral border was determined as the first section where the corpus collosum connected in the midline of the section and the caudal border was the first section where the anterior commisure joined in the midline.

The guard volume was set at 2 ptm for the top and bottom of the section and the

counting cubes were randomly distributed with a total sampling frequency of the outlined count area of 25% for MS/vDB and 33% for IP and SP-GCL (-8.3% and 11% of the total area respectively, since every third section was sampled). The outlined counting area was defined by the user and only somata that were clearly and distinctly stained were counted. Each counting box was examined at 40x magnification and the computerassisted focus was used to scan from the top to the bottom of the counting box. Cells were counted only if they were either completely inside of the counting box, or partially inside of the box on the top, back, or left side. They were not counted if they fell outside of the box or crossed into the box anywhere on the bottom, front, or right side. The cells that were counted were labeled on the monitor by clicking the mouse pointer on each cell and the MCID software recorded the number of marks.





36


The MCID software interpolated the total volume of the MS/vDB based on the volume of the count areas defined by the user. Cell density (Nv) was estimated by dividing the total number of cells counted by the volume of the counting boxes, which was also tracked by the software. The total cell number was estimated by multiplying Nv by the total volume.

Enzyme-linked Immunosorbent Assay (ELISA) Characterization of NGF and BDNF

For measurement of BDNF in hippocampus and MS/vDB, nine IL-2 KO and

seven wild-type mice were used. For measurement of NGF protein levels, seven animals per group were used. Initial test runs revealed that some, but not all, of the NGF protein levels fell below the sensitivity of detection for the kit; therefore, the homogenized NGF samples were spiked with 25 pg/ml of the known NGF standard included with the kit to bring any low levels of expression above the kit's 15.6 pg/ml lower detection limit. One column of the ELISA plate was also run with only the 25 pg/ml standard spike to provide a baseline and the data reported are corrected for this. Animals used for ELISA characterization of neurotrophin levels only received saline perfusion and were not postfixed in paraformaldehyde. The brains were removed, snap frozen, and then allowed to equilibrate to -200 C. The brains were sectioned on a cryostat at -20-220 C at 400 jm thickness and the MS/vDB and hippocampi were dissected with a 0.75 mm micropunch on a -200 C freezing platform. The dissected tissue was weighed on a microgram scale, and then transferred to 25 tl of homogenizing solution (50 mM Na/Na2 and 0.2% TX100 in H20 with Anti-protease Complete TM cocktail (Boehringer)) per mg of wet weight tissue. The tissue was sonicated in the homogenizing solution for 30-sec on ice





37


and centrifuged at 16,000 g for 15 min at 40 C. The supernatant was collected and stored at -200 C for ELISA analysis.

Levels of NGF and BDNF were analyzed in the homogenates from MS/vDB and hippocampus using a commercially available E,,a Immunoassay System according to the manufacturer's instructions (Promega). Briefly, the 96-well plates were coated with 1:6,250 anti-NGF pAb in carbonate coating buffer (0.025 M sodium bicarbonate, 0.025 M sodium carbonate, pH 9.7) and incubated overnight at 40 C. The plates were washed with TBST wash buffer (20 mM Tris-HCL pH 7.6, 150 mM NaCI, 0.05% (v/v) Tween 20) and blocked with lx Block and Sample buffer (provided with kit) for 1 hour. The plates were washed again with TBST and a set of standard curves were generated in duplicate by performing 1:2 dilutions of a known 500 pg/ml standard in a range from 500 pg/ml to 7.6 pg/ml followed by a "blank" well of 0 pg/ml. All added samples and standards were allowed to incubate at 250 C for 6 hours. The plates were washed thoroughly with TBST and 1:4000 anti-NGF mAb was added and incubated overnight at 40 C. The plates were again washed with TBST and 1:100 anti-rat IgG pAb conjugated to HRP was added for 2.5 hours at room temperature. The plates were washed and TMB One Solution was added for color development for 10 min. The reaction was stopped with the addition of equal volume of 1 N HCI and the absorbance was read at 450 nm within 30 min of the color development reaction. The data were reported as pg of protein per mg wet weight tissue.

Statistical Analysis

Results are reported as the mean SEM. Statistical differences between groups were determined using analysis of variance (ANOVA).





38


Results

Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell Number in Adult Mice

In 8-12-week-old mice, no significant differences were apparent in the relative

number of stereologically counted Parv-positive neurons between groups (F(1,1 0)=0.002, p=0.964). Thus, the GABAergic neurons appear to be unaffected by IL-2 gene deletion. We did not assay for GABAergic alterations in younger animals, since there were no differences in the adult IL-2 KO mice relative to the wild-types.

In contrast to the previously reported data from 8-12-week-old animals (Beck et al., 2002), there was no significant difference in stereologically counted cholinergic somata number in the MS/vDB of 21-day-old IL-2 KO mice relative to wild-type mice (F(1,12)=0.689, p=0.423). As expected, the 21-day-old KO mice also did not exhibit splenomegaly seen in the autoimmune 8-12-week-old group, as spleen weights did not differ between 21-day-old wild type and IL-2 KO mice (F(1,12)=0.989, p=0.340). Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice

The IP-GCL of IL-2 KO mice had significantly fewer neuronal somata than wildtype mice (Fig. 3-1; F(1,12)=10.966, p=0.006). In the SP-GCL, however, there was no significant difference in granule cell number (Fig 3-1.; F(1,12)=0.197, p=0.665). Alterations in Neurotrophin Levels

Levels of NGF protein in hippocampal tissue homogenates was significantly increased in IL-2 KO mice relative to wild-type mice (Fig. 3-2A; F(1,10)=8.261, p=0.017). The levels of BDNF protein, conversely, were significantly decreased in IL-2 KO mice relative to wild-type mice (Fig. 3-2B; F(1,12)=8.023, p=0.015).







39




Stereological Cell Count of Nissl-stained Granule Cell Layers


1200


1000




L 600


400

200


0
Infrapyramidal Suprapyramida I Wild-Type
IL-2 Knockout


Fig. 3-1. There was a significant reduction in infrapyramidal, but not suprapyramidal
granule cells in IL-2 knockout relative to wild-type mice. Each bar represents the
mean SEM of seven animals per group. *p=0.006.



A) Enhanced NGF Levels in the Hippocampus B) Reduced BDNF Levels In the Hippocampus
of IL-2 Knockout Mice of IL-2 Knockout Mice


30 80








r 0
E E


z
C, 5



Wdd-Type IL-2 Vaockout Wild-Type IL-2 Knockout Genotype Gen ope Figure 3-2. There was a significant: A) increase in NGF and B) decrease in BDNF
protein levels in the hippocampus of IL-2 knockout compared to wild-type mice.
The NGF bars represent the mean SEM of 7 animals per group. The BDNF
bars represent mean SEM of 7 IL-2 knockout and 9 wild-type mice. *p<0.05.





40


Discussion

These data are the first to demonstrate that the loss of endogenous IL-2 in knockout mice can lead to alterations in neuronal cell number in the IP-GCL and production of the neurotrophins, BDNF and NGF. Further, this study expands upon the previous finding that IL-2 gene deletion leads to a deficiency of cholinergic neurons in the MS/vDB (Beck et al., 2002), by showing a lack of significant cholinergic neuronal differences in MS/vDB of 21-day-old IL-2 KO animals or GABAergic alterations in adult animals.

The lack of a difference in GABAergic neurons in 8-12-week-old adult mice was consistent with our initial hypothesis, as there is no evidence in the literature that IL-2 has any modulatory effects on GABAergic neurons. Moreover, this is not a regional effect, but rather appears to be selective to cholinergic projection neurons. As previously mentioned, IL-2 is a potent modulator of ACh release (Hanisch et al., 1993; Seto et al., 1997), and its precursor enzyme ChAT (Mennicken and Quirion, 1997), suggesting an effect of IL-2 on cholinergic neurons. In GABAergic neurons, however, IL-2 has failed to evoke release of GABA in mesencephalic neuronal cultures (Alonso et al., 1993), or the cortex or hippocampus of mice (Bianchi et al., 1995). Since IL-2 deficiency does not affect the number of GABAergic somata in the MS/vDB of IL-2 KO mice, the neuronal loss appears to be selective for cholinergic neurons in the MS/vDB. Furthermore, we previously found no differences in the striatal cholinergic neuronal number (Beck et al., 2002), so the lack of IL-2 does not simply cause a general loss of all cholinergic neurons.

Against our initial hypothesis that 21-day-old IL-2 KO mice would have similar cholinergic deficiencies as adult 8-12-week-old mice, there was no detectable loss of cholinergic cell number in the MS/vDB. We did not examine 21-day-old mice for





41


differences in GABAergic cell number, since we did not detect any changes in adult mice using the same marker. One potential explanation for the loss of cholinergic neurons in the MS/vDB may be a failure in maintenance or survival in the late stages of, or after, development. Other studies have found decreases in cholinergic enzyme activity (i.e., ChAT and AChE) between postnatal days 30-60 in normal rats (Thal et al., 1992), and postnatal days 60-150 in C57BL/6 mice (Virgili et al., 1991). The IL-2 KO mice may potentially be more susceptible to this loss of cholinergic activity during adulthood, which could lead to the previously observed deficiencies in 8-12-week-old IL-2 KO animals.

An alternate explanation for the different cholinergic effects seen in 21-day-old vs. 8-12-week-old animals is that the loss of IL-2 may be secondary to the effects of autoimmunity present in adult IL-2 KO animals. Though we cannot completely rule out this possibility, we have previously failed to find discernable levels of infiltrating lymphocytes or clear signs of gliosis in the brains of IL-2 KO animals (Petitto et al., 1999). More research is necessary to further address this issue.

Another finding of this study was a significant decrease in neuronal cell number in the IP-GCL, but not the SP-GCL. This decrease is consistent with the in vitro studies showing a potent neurotrophic effect of IL-2 on hippocampal neurons (Awatsuji et al., 1993; Sarder et al., 1996; Sarder ct al., 1993). Furthermore, these data are supported by our previous findings that IL-2 KO mice exhibited a reduced distance across the IP-GCL (Beck et al., 2002), and that the IP mossy fiber length of IL-2 KO mice is shorter than wild-type controls (Petitto et al., 1999). The reductions in distance across the IP-GCL could also potentially be explained by increased density, but not number, of cells or





42


smaller cell body size. Qualitative assessments of random granule cells, however, do not support this hypothesis, though a more extensive study would be necessary to definitively address that issue.

Though the receptors for IL-2 are more abundant in the hippocampus, including the GCL of the DG (Petitto and Huang, 1994; Petitto et al., 1998), it is not clear whether IL-2 may act directly on these neurons, or whether it upregulates other growth factors like the neurotrophins. The observed differences in the level of the neurotrophin BDNF was consistent with our hypothesis that we would find a reduction in trophic factors important in MS/vDB and hippocampal development and maintenance. BDNF plays a role in the maintenance and repair of septal cholinergic neurons (Alderson et al., 1990; Morse et al., 1993; Ward and Hagg, 2000), can implement a positive feedback mechanism with these neurons to enhance the release of ACh (Knipper et al., 1994), and can also modulate neurogenesis (Larsson et al., 2002; Lee et al., 2002), thus potentially impacting granule cell number. Thus, the reduction of cholinergic cell number in the MS/vDB is consistent with a reduction in this trophic factor. The exact interaction between IL-2 gene deletion and the reduction of BDNF levels remains unclear. Though BDNF is expressed in the peripheral immune system by lymphocytes, IL-2 does not stimulate its production or release. IL-2 can, however, upregulate the expression of TrkB, the receptor for BDNF, in lymphocytes (Besser and Wank, 1999). Furthermore, some evidence suggests that BDNF can stimulate a positive feedback mechanism of its own production via the TrkB receptor in hippocampal neurons (Canossa et al., 1997; Saarelainen et al., 2001). In IL-2 KO mice, the absence of IL-2 may therefore potentially lead to a down-regulation of the TrkB receptor, thereby partially inhibiting the positive





43


feedback production of BDNF. Interestingly, the neurotrophin Trk receptors and IL-2 receptor share some of the same signal transduction pathways (e.g., mitogen activated protein kinase or phosphatidylinositol 3-kinase), which appear to play a role in their growth and survival promoting actions (for reviews see Gaffen, 2001; Patapoutian and Reichardt, 2001). Whether IL-2 knockout leads to disruption of one of these signal transduction pathways has not, to our knowledge, been elucidated and thus requires further study.

Against our initial hypothesis, NGF protein levels were actually increased in the IL-2 KO mice. Unlike BDNF, NGF does not appear to stimulate a positive feedback neurotrophin release from hippocampal neurons (Canossa et al., 1997). Given the reduction in cholinergic survival in the MS/vDB of IL-2 KO mice, the target neurons in the hippocampus of these animals may produce higher protein levels of NGF as a compensatory response. Similarly, moderate lesions of rat septohippocampal projections lead to increased mRNA expression of NGF, but not BDNF in hippocampal target cells (Hellweg et al., 1997).

In summary, cholinergic deficits seen in the MS/vDB of IL-2 KO mice appear to be selective for cholinergic over GABAergic neurons. In addition, the loss of cholinergic neurons in the MS/vDB may occur in the later stages of, or after, development of the septohippocampal system, as the deficits are not seen in 21-day-old IL-2 KO mice. In the hippocampus, the number of neurons in the IP-GCL is significantly reduced. A reduced production of hippocampal BDNF may contribute to many of the aforementioned changes, though NGF levels are increased in a possible compensatory response. Although overt signs of autoimmunity in the brain are not apparent (we have been unable





44


to detect significant levels of leukocyte infiltration or gliosis in IL-2 KO mice brains), further study is necessary to assess this possibility, as factors such as IL-2 induced cytokine dysregulation or autoantibodies could contribute to the hippocampal alterations in adult IL-2 KO mice.













CHAPTER 4
INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND NEUROGENESIS IN THE HIPPOCAMPUS Introduction

Receptors for interleukin-2 (IL-2) are enriched in the hippocampal formation, and many of the most prominent neurobiological functions of this cytokine occur in the hippocampus (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto et al., 1998). Previous studies from our laboratory have found that IL-2 knockout (KO) mice exhibit significantly lower numbers of medial septum and vertical limb of the diagonal band of Broca (MS/vDB) cholinergic cell bodies, a reduction in the distance across the granular cell layer (GCL) of the infrapyramidal (IP) blade of the dentate gyrus (DG), and decreased fiber length and neuronal cell number in the IP-GCL of the DG (Beck et al., 2004; Beck et al., 2002; Petitto et al., 1999). These neurobiological alterations appear to be related to abnormalities in learning and memory performance and sensory motor gating in IL-2 KO mice (Cushman et al., 2004; Petitto et al., 1999).

Because IL-2 has been shown to possess various neurotrophic and

neuromodulatory effects on hippocampal neurons in vitro (Awatsuji et al., 1993; Bianchi et al., 1995; Pauli et al., 1998; Plata-Salaman and ffrench-Mullen, 1993; Sarder et al., 1996; Sarder et al., 1993; Tancredi et al., 1990), our original working hypothesis was that the alterations exhibited by IL-2 KO mice are due to the absence of IL-2's neurotrophic actions on hippocampal neurons during development. More recent data from our



45





46

laboratory, however, suggests that these hippocampal changes may be due to neurodegenerative rather than neurodevelopmental processes. We tested mice at postnatal day 21 (P21), an age where septohippocampal cholinergic neurons are nearly fully developed, to determine if the reduction in septohippocampal cholinergic projection neurons seen in adult IL-2 KO mice was present earlier in postnatal development (e.g., at weaning) and prior to the onset of the earliest signs of autoimmune disease (e.g., splenomegaly, lymphadenopathy). Contrary to our hypothesis, we found that the number of MS/vDB cholinergic cell bodies did not differ between IL-2 KO and wild-type littermates at P21 (Beck et al., 2004). Thus, together these data indicate that the loss of cholinergic neurons that occurs between P21 and adulthood (8-12 weeks) suggests an alternate hypothesis; neurodegenerative processes may be operative in the brain of IL-2 KO mice.

Since IL-2 is an important factor in immune physiology, one possible mechanism behind these neurodegenerative processes is immune dysregulation caused by the absence of IL-2. IL-2-deficiency in mice leads to generalized systemic autoimmune disease in adult mice that may affect multiple organs in the periphery, most notably the intestines and the kidneys (Horak, 1995). The autoimmune effects in IL-2 KO mice involving peripheral organs are mediated largely by infiltrating T cells. In the colon, for example, adult IL-2 KO mice develop chronic inflammatory bowel disease with features common to inflammatory ulcerative colitis in humans, where the lamina propria is infiltrated with activated T cells responsible for the development of this inflammatory disease (Ma et al., 1995). In addition, there is a disruption of immune homeostasis that is evidenced by changes in the gene expression of several Thl, Th2, and various proinflammatory





47


cytokines in this organ (Autenrieth et al., 1997; Meijssen et al., 1998). Moreover, these cytokine changes and the onset of inflammatory bowel disease are preceded by increased gene expression of IL-15, which shares the same signal transducing receptor subunits with IL-2 (Meijssen et al., 1998). Thus, it is possible that the immune dysregulation in the brain of IL-2 KO mice may be induced by activated T cells and/or proinflammatory cytokines (e.g., IL-1, TNFa, IL-6) from the periphery crossing the blood-brain-barrier (BBB).

By contrast, IL-2 may lead to neuroimmunological changes that do not involve

peripheral immune cells. Rather than peripheral T cells and serum cytokines entering the brain, an alternative hypothesis that may account for the hippocampal differences observed in P21 versus adult IL-2 KO mice may be that the absence of IL-2 reduces the trophic support of hippocampal neurons as a result of dysregulation of other brainderived cytokines. Thus, loss of IL-2 in the brain could in turn modify the normal neuroimmunological status of the brain by modifying the normal expression of brain cytokines such as IL-15. Since IL-2 can modify the release of certain cytokines from lymphoid cells (Lauwerys et al., 2000; McDyer et al., 2002), similar actions could occur in brain cells that produce cytokines (e.g., microglia, astrocytes). Alterations in the production of brain cytokines important in normal brain physiology could alter the integrity of hippocampal neurons by decreasing levels of classic neurotrophins and/or neurotrophic cytokines on the one hand, or elicit inflammatory-like neurodegenerative processes within the brain on the other.

The present study therefore sought to test the hypothesis that IL-2 gene deletion results in neuroimmunological changes in the hippocampus by examining the possible





48

outcomes described above. We compared the hippocampi of adult IL-2 KO mice and wild-type littermates at 8-12 weeks of age, the age where differences in hippocampal cytoarchitecture and behavior have been found previously (Beck et al., 2004; Beck et al., 2002; Cushman et al., 2004; Petitto et al., 1999), for differences in several measures of neuroimmunological status. First, the groups were assessed for differences in the number of CD3 T lymphocytes and activated microglial cells (as measured by MHC-II positivity) in the hippocampus. IL-15 is also expressed in the brain (Hanisch et al., 1997a; Lee et al., 1996), and this cytokine is known to have both proinflammatory and anti-inflammatory effects, potent anti-apoptotic, and T cell chemoattractant properties (Wilkinson and Liew, 1995). Because IL-15 uses the IL-2/15R and Y subunits that are enriched in the neuronal cell layers of the hippocampus (Petitto and Huang, 2001), and may modulate microglial cell function and T cell chemoattraction, a second aim of this study was to test the hypothesis that IL-15 is elevated in the hippocampus of IL-2 KO mice. In addition, since changes in both IL-2 and IL-15 may modify levels of various cytokines in other tissues and physiological contexts, exploratory testing was performed to determine if IL-2 KO mice have increased levels of proinflammatory cytokines in the hippocampus relative to wild-type mice (also, compared to serum levels). Furthermore, as recent evidence indicates that elevation of inflammatory cytokines such as IL-6 may impair hippocampal neurogenesis (Monje et al., 2003; Vallieres et al., 2002), a third aim of this study was to test the hypothesis that the postulated changes in neuroimmunological status would be associated with reductions in neurogenesis of neurons in the dentate gyrus (DG) of IL-2 KO mice.





49


Materials and Methods

Animals and Genotyping

Mice used in these experiments were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2 heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the NIH repository at Jackson Laboratories, were derived from ten generations of backcrossing onto the C57BL/6 background. Mice were housed under specific pathogenfree conditions. Animals used in these experiments were 8-12 weeks of age. Independent animals were used for the assessment of CD3 T cells and MHC II microglial cells in the hippocampus, the determinations of hippocampal versus serum cytokine levels, and assessments of neurogenesis in the dentate gyrus. Specific numbers utilized are reported at the beginning of the description of each method.

The genotypes of all mice were determined by the polymerase chain reaction (PCR). PCR reactions were performed using 25 pl total reaction volume containing 1 CpM each of forward and reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP,

0.3 pl Taq DNA polymerase, and amplified using a thermal cycler with a heated evaporation cover (Ericomp). The cycling parameters were hot start 950C (3min), denaturing 94'C (30 sec), annealing 640C (30 sec), extension 720C (45 sec) with a final extension step of 4 min. Thirty cycles were used for these experiments. The 5' and 3' primers for the IL-2 KO (500 bp knockout band amplified) were 5'TCGAATCGCCAATGACAAGACGCT-3' and 5'GTAGGTGGAAATTCTAGCATCATCC-3'. The 5' and 3'primers for the IL-2 wild type (324 bp wild type band amplified) were 5'-





50


CTAGGCCACAGAATTGAAAGATCT-3' and 5'GTAGGTGGAAAATTCTAGCATCATCC-3'. CD3+ T cells and MHC II+ Microglia Immunohistochemistry

For qualititative assessment of autoimmunity, three IL-2 KO and three wild-type brains were processed for MHC II (an activated microglial marker) and CD3 (a pan T cell marker) immunohistochemistry. Each animal was anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose of 0.015 ml injection cocktail/g body weight and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS). The brains and spleens were removed and fixed overnight in 4% paraformaldehyde, followed by overnight equilibration in 30% sucrose cryoprotective solution, and then were snap frozen in isopentane (-800C) for storage. The spleens were weighed to assay for relative splenomegaly of IL-2 KO vs. wild-type mice. The brains were equilibrated to -200C prior to cryostat sectioning into 50 jpm slices in the coronal plane, collected into individual wells of polystyrene 24-well plates (NUNC 1147), and stored free-floating at 4oC in PBS for histochemistry. Every third section was processed for MHC II or CD3 immunohistochemistry.

Free-floating 50-ipm sections were incubated for 20 minutes in 1% hydrogen peroxide (H202) to quench endogenous peroxidative activity. The sections were then washed and blocked for 1 hr in 200 pl/well 3% normal goat serum (NGS). After this incubation, the sections were incubated overnight in the primary antibody, rat anti mouse CD3 (BD PharMingen; 1:500 in PBS with 0.3% Triton X-100 and 1% NGS) or rat anti mouse MHC II (BD PharMingen; 1:500 in PBS with 0.3% Triton X-100 and 1% NGS). The next day, the sections were washed and incubated overnight in the secondary





51


antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with

0.3% TX-100 and 1% NGS). The sections were then washed and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS) for 2 hrs. The sections were developed in

0.5 mg/ml 3,3'-diaminobenzidine (DAB), 0.2 mg/ml urea H202 for approximately 5 min and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped.

Preparation of Serum and Brain Tissue for Cytokine Analysis

Hippocampal homogenates were analyzed from eight IL-2 KO and nine wild-type mice to measure cytokine levels in the hippocampus. From these subject groups, serum was collected from a smaller subset of animals (five IL-2 KO and seven wild-type mice) for comparative analysis of brain vs. peripheral cytokine levels. Animals used for characterization of endogenous cytokine levels were anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose of 0.015 ml injection cocktail/g body weight. Whole blood was collected by puncturing the right atrium of the heart and inserting heparanized micro-hematocrit capillary tubes (Fisher Scientific). The animals were then saline perfused, but were not post-fixed in paraformaldehyde. The whole blood was centrifuged in Microtainer Brand serum separator tubes (Becton Dickinson) at 5,000 rpm for 10 minutes to isolate serum and the serum was stored at -800 C until used for Luminex analysis. The brains were removed, snap frozen, and then allowed to equilibrate to -200 C. The brains were sectioned on a cryostat at -20-220 C at 400 tm thickness and the hippocampi were dissected with a 0.75 mm micropunch on a -200 C freezing platform. The dissected tissue was weighed on a microgram scale, and then transferred to 25 ul1 of homogenizing





52


solution (50 mM Na/Na2 and 0.2% TX-100 in H20 with Anti-protease Complete TM cocktail (Boehringer)) per mg of wet weight tissue. The tissue was sonicated in the homogenizing solution for 30 sec on ice and centrifuged at 16,000 g for 15 min at 40 C. The supernatant was collected and stored at -200 C for Luminex analysis. Multiplex Microsphere Cytokine Analysis

Commercial kits, Lincoplex mouse cytokine (Linco, Research, Inc) and a

Luminex 100 LabMAP system (Upstate Biotechnology), were used in attempt to measure a number of cytokines in the hippocampus and in the serum of IL-2 KO and wild-type mice. Assays were performed according to the manufacturer's instructions, and cytokine concentrations were calculated using the Softmax program and the linear range on the standard curve (3.2-10,000 pg/ml). Altogether, we attempted to detect a total of twentytwo different cytokines and chemokines from the serum and brain homogenates of these animals. In the serum, there were detectable levels of IL-6, IL-13, kerotinocyte-derived chemokines (KC), granulocyte-colony stimulating factor (G-CSF), and macrophage inflammatory protein-I alpha (MIP- la). In the brain, there were detectable levels of IL7, IL-9, IL-12, IL-15, interferon-gamma inducible protein of 10 kD (IP-10), and monocyte chemoattractant protein-i (MCP- 1). Thus, only those cytokines and chemokines detected were subjected to statistical analyses. The remainder of the cytokines and chemokincs tested could not be detected in either the serum or the brain (IFN-y; TNF-ca; IL-lcx, IL- 13; IL-2, IL-4, IL-5, IL-10, IL-17, GM-CSF, and RANTES).

Labeling Neurogenesis with BrdU

Twelve IL-2 KO (six male; six female) and eleven wild-type (five male; six female) mice were used to assay for differences in neurogenesis. The procedure for





53

labeling of neurogenesis and subsequent immunostaining in the mouse hippocampus has been adapted from (Lee et al., 2002). Briefly, the mice were given five intraperitoneal injections of BrdU (50 mg/kg of body weight) over the course of 3 days. The day following the last injection, the mice were sacrificed and perfused with 0.9% saline followed by 4% paraformaldehyde in PBS as described previously.

The BrdU-incorporated brains were equilibrated to -200 C and cryostat-sectioned at 50 pim in the coronal plane. They were collected into individual wells of polystyrene 24-well plates (NUNC 1147), and used for free-floating immunohistochemistry. The sections were then washed twice in PBS and then the DNA was denatured by a 30 minute incubation with 2 N HCI to allow binding of the antibody to the BrdU in the singlestranded DNA. The acid was neutralized with a 0.1 M borate buffer (pH 8.5) wash, followed by several washes in PBS. Afterwards, the sections were blocked for 1 hr in 3% normal goat serum (NGS). The sections were then incubated overnight in the primary antibodies, rat monoclonal anti-BrdU (Serotec; 1:400 in PBS with 0.3% Triton X-100 and 1% NGS) and either the neuronal marker mouse monoclonal anti-tubulin P III isoform (Chemicon; 1:200 in PBS with 0.3% Triton X-100 and 1% NGS), the astroglial marker rabbit anti-glial fibrillary acidic protein (GFAP; Chemicon; 1:1,000 in PBS with 0.3% Triton X-100 and 1% NGS) or the oligodendrocyte marker mouse anti-2'3'-cyclic nucleotide 3'-phosphohydrolase (CNPase; Chemicon; 1:200 in PBS with 0.3% Triton X100 and 1% NGS). The next day, the sections were washed twice in PBS and incubated for 2 hr in the dark with the secondary antibodies, goat anti-rat IgG (H+L) conjugated with Alexa Fluor-488 (green; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1% NGS) and goat anti-mouse IgG (highly cross-absorbed H+L) conjugated with Alexa





54

Fluor 568 (red; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1% NGS) or goat anti-rabbit IgG (H+L) conjugated with Alexa Fluor 350 (blue; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1% NGS). The sections were then washed twice in PBS, placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped.

The sections were imaged using a Bio-Rad 1024 ES confocal microscope and

only cells which showed colocalized staining through five consecutive 1-pm planes were considered to be double-labeled. The IP and SP-GCL area (mm2) were measured at 20x magnification using the MCID 5.1 software, a CCD High Resolution Sony camera, and a Zeiss Axioplan 2 microscope. The data were reported as a density of all double-labeled cells counted from five sections per animal divided by the total area measured Results

Assessment of CD3 T Cells and MHC II+ Activated Microglia in the Hippocampus

No CD3+ T cells were detected in the hippocampi of either 8-12-week-old IL-2 KO or wild-type mouse brains, and only an occasional MHC II+ microglial cell was detected (e.g., approximately one every other section) in both groups. By contrast, both activated MHC II microglia and CD3 T cells were readily detectable in positive control slices (sections of the axotomized FMN of wild-type C57BL/6 mice; Petitto et al., 2003) demonstrating that the immunohistochemistry procedure utilized was effective for labeling both markers. Thus, no differences in T cells or activated microglial cells were found between the subject groups.

Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice

As depicted in Figure 4-1, in the hippocampus, there were significantly increased levels of IL-12 (increased -57%; F(1,15)=9.174, p=0.008) and IL-15 (increased -38%;






55


F(1,15)=6.105, p=0.026) in the IL-2 KO compared to wild-type mice. Figure 4-1 also shows that IL-2 KO mice had increased levels of the chemokines, IP-10 (increased

-63%; F(1,15)=4.747, p=0.046) and MCP-1 (increased -46%; F(1,15)=5.218, p=0.039). Although detectable levels of IL-7 and IL-9 were found in hippocampus, they did not differ between the subject groups.


Enhanced Cytokine Concentrations in the Hippocampus




10
U.n
2 0,


06
,



02


IL-12 IL-15 IP-10 MCP-1 Cytokine
I Wild-Type
IL-2 Knockout

Figure 4-1. Protein levels of several cytokines and chemokines are elevated in the IL-2
knockout mice compared to wild-type brain. These include IL-12, IL-15, IP-10,
and MCP-1. *p<0.05; **p<0.01.

Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-type Mice

As can been seen in Figure 4-2, the IL-2 KO mice exhibited marked elevation in serum IL-6 (increased -224%; F(1,10)=8.077, p=0.017), and serum MIP-l o serum concentration (increased -327%; F(1,10)=21.538, p=0.001) compared to wild-types. Although detectable levels of IL-13, G-CSF, and KC were detectable in the serum, levels of these cytokines did not differ between the subject groups.





56


A) B)
Enhanced Cytoklne Concentrations In the Serum of Enhanced Cytokine Concentrations In the Serum of IL-2 KO Animals IL-2 KO Animals




11
2 2

4-[


IL-6 MIP-1 alpha WdsType ] 0 Wdd-Type
IL-2 Knockoul i IL 2 Knockout

Figure 4-2. Relative cytokine profile of IL-2 knockout mouse serum does not match brain
profile. A) IL-6 is increased in IL-2 knockout mice compared to wild-type; and
B) the chemokine MIP- l a is higher in IL-2 knockout mice than wild-type.
*p<0.05; **p<0.01.

Alterations in Neurogenesis

Although there was not a significant effect of group on neurogenesis in either the IP-GCL or SP-GCL, as depicted in Figure 4-3 and 4-4 respectively, there was a significant group by gender interaction in both the IP-GCL (F(1,19)=4.71, p=0.043) and SP-GCL (F(1,19)=6.43, p=0.02). Fisher's least significant difference post hoc analysis test confirmed a difference between male IL-2 KO and wild-type mice in the IP-GCL (p=0.025) and SP-GCL (p=0.014), but not between female groups in either region. There was no significant effect of group or group by gender interaction in cells around either the IP-GCL or SP-GCL labeled with the oligodendrocyte marker, CNPase (data not shown). Similarly, no significant effect of group or group by gender interaction was evident in either the IP or SP-GCL in cells labeled with the astrocyte marker, GFAP (data not shown).







57





Neurogenesis in the Infrapyramidal Granule Cell Layer






400

E
E








10

1 0 "-------.... ..
IL-2 Knockout Wild-Type C Female Male


Figure 4-3. There is no effect of group on neurogenesis in the infrapyramidal granule cell
layer, but there is a group by gender interaction between IL-2 knockout and wildtype mice (p=0.043). This effect appears to be attributable to differences in the
male mice (Fisher least significant difference test, p=0.025).



Neurogenesis in the Suprapyramidal Granule Cell Layer


50s


400

E
E







100


0
Wild-Type IL-2 Knockout CI Female Male


Figure 4-4. There is no effect of group on neurogenesis in the suprapyramidal granule
cell layer, but there is a group by gender interaction between IL-2 knockout and wild-type mice (p=0.02). This effect appears to be attributable to differences in
the male mice (Fisher least significant difference test, p=0.014).






58


Discussion

The data presented here show that IL-2 KO and wild-type littermates exhibit differences in several measures of neuroimmunological status in the hippocampus. In order to access the brain parenchyma, T cells require activation markers to cross the BBB (Hickey et al., 1991). We have previously reported leukocytes were not detectable in the cresyl violet stained hippocampal sections from IL-2 KO mice (Petitto et al., 1999), however, we recognized that it is difficult to reliably detect small numbers of peripheral leukocytes in the brain without cell-type specific stains. Although we were therefore not expecting to see substantial numbers of T cells in the IL-2 KO brain, we wanted to determine if small numbers of autoimmune T lymphocytes were present that could initiate neuroimmunological alterations in the hippocampus. The hippocampi of IL-2 KO mice were devoid of T cells, despite the fact that the majority of peripheral T cells of IL-2 KO mice express activation markers such as CD69 (Sakai et al., 1995; Schopke et al., 1991), which are thought to enhance their ability to cross the BBB. Microglia are indigenous antigen presenting cells (Hickey and Kimura, 1988; Streit et al., 1988). Contact with T cells can induce microglia to exhibit characteristics of antigen presenting cells, and microglia also have the ability to activate T cells (Aloisi et al., 2000). There was, however, no evidence of increased numbers of activated microglia in the hippocampus of IL-2 KO mice. This observation is consistent with our previous finding in C57BL/6scid-IL-2 KO (mice without mature T and B cells), which were devoid of T cells in the axotomized facial motor nucleus and had levels of axotomy-induced activated microglia that did not differ from wild-type mice (Petitto et al., 2003). Thus, at the cellular level, the hippocampus of IL-2 KO mice did not show signs of autoimmune disease.





59


Cytokines can enter brain via specific transport mechanisms and through the

circumventricular organs. In this study, we measured levels of the various cytokines of interest in the serum to determine if levels found in the hippocampus correlated with those found in the serum. If a particular inflammatory cytokine (e.g., TNFa) had been found to be significantly elevated in the serum and the hippocampus of IL-2 KO mice, it would have suggested the possibility that peripheral immune activation associated with autoimmunity (Schimpl et al., 2002) could account for the presence of the cytokine in the hippocampus (although increased gene transcription and translation in both the periphery and the brain could not be ruled out). Although there were significantly increased levels of IL-6 and MIP Ia in the circulation of IL-2 KO mice, we did not detect either of these proteins in the brain. IL-6 is a proinflammatory cytokine that was of particular interest to us because of its actions in the hippocampus, including effects on neurogenesis (Monje et al., 2003; Vallieres et al., 2002). Although levels of IL-6 were markedly increased in the serum, though not measurable in the hippocampus at the tissue homogenate concentrations used, the unlikely possibility remains that IL-6 could have entered the brain from the circulation and had functional consequences at concentrations below the limits of detection of the assay method. Nonetheless, the most parsimonious explanation is that cytokines from the peripheral circulation of IL-2 KO mice are not the source of the cytokine alterations found in the hippocampus. Thus, peripheral cytokine dysregulation associated with autoimmunity in IL-2 KO does not appear to be associated with their hippocampal pathology.

Our data indicate that loss of IL-2 in the brain results in changes in the production of several other brain cytokines. Consistent with our hypothesis, IL-15 concentrations





60

were increased in the hippocampus of IL-2 KO mice. IL-15 is structurally related to IL-2 and shares the same p and Yc signal transducing receptor subunits with the IL-2 receptor (Giri et al., 1995). IL-15 also shares and opposes several physiological functions of IL-2 in the peripheral immune system (Waldmann, 2002; Waldmann et al., 2001). IL-15 and its heterotrimeric receptor are constitutively expressed in various regions of the adult mouse brain and can be detected in microglial cultures (Hanisch et al., 1997a), astrocytes (Lee et al., 1996), and possibly neurons (Maslinska, 2001; Satoh et al., 1998). As noted earlier, increased IL-15 gene expression precedes the inflammatory cytokine changes and onset of inflammatory bowel disease in IL-2 KO mice (Meijssen et al., 1998). It also induces the onset of autoimmunity in thyroiditis (Kaiser et al., 2002). Thus, IL-15 could trigger proinflammatory cytokine-like processes in the hippocampus, including the elevations in IL-12 that were found in IL-2 KO mice in this study. IL-12-driven Thl responses are involved in inflammation (e.g., colonic) in IL-2 KO mice (Ludviksson et al., 1997), and it has been implicated as an important effector in the pathogenesis of experimental autoimmune encephalomyelitis (EAE) (Adorini, 1999; Segal et al., 1998). Moreover, IL-15 can render cells resistant to the protective effects of TGFP (Campbell et al., 2001), a Th2 cytokine that appears to play a key role in dampening processes associated with peripheral autoimmune disease in IL-2 KO mice (Ludviksson et al., 1997). Thus, these actions of IL-15 suggest that it may be involved in the hippocampal pathology seen in IL-2 KO mice. It is noteworthy, however, that IL-15 has potent antiapoptotic properties (Lauwerys et al., 2000; Waldmann, 2002; Waldmann et al., 2001) that may oppose the pro-apoptotic effects of IL-2. We have recently found that loss of brain IL-2 in C57BL/6scid-IL-2 KO mice increased neuroregeneration in the axotomized





61

facial motor nucleus (Petitto et al., 2003). It is interesting to speculate that elevated IL-15 levels could contribute to the increased motor neuronal survival in those mice. Therefore an alternative interpretation is that increased IL-15 in the hippocampus of IL-2 KO mice could be a compensatory response to counteract neuroregenerative changes in the hippocampus.

The increased levels of MCP-1 and IP-10 may possibly be induced by increased IL-15 in the hippocampus (Badolato et al., 1997). Previous studies have demonstrated that IP-10 expression can be detected in lipopolysaccharide (LPS)-treated microglial and astroglial cultures and in situ hybridization of LPS-treated rat brains (Ren et al., 1998). Similarly, MCP-1 can also be induced by the addition of the pro-inflammatory cytokine, TNF-ca, or the anti-inflammatory cytokine, TGF-P, in astrocytes (Hurwitz et al., 1995) and microglia (Meda et al., 1996). IP-10 and MCP-1 are also chemoattractant factors for T lymphocyte infiltration into the CNS (Babcock et al., 2003; Dufour et al., 2002), and may attract activated microglia in vitro (Cross and Woodroofe, 1999). In spite of the published data linking IP-10 and MCP-1 to T cell and microglial chemotaxis, we were unable to detect either T cells or increased numbers of activated microglia in the hippocampus of IL-2 KO mice. Further studies are necessary to determine the functional significance of the increased production of IP-I 0 and/or MCP- in the hippocampus of IL-2 KO mice. Finally, in addition to its immune activating effects as a Th I cytokine, IL-2 is also known to have important critical negative regulatory functions by stimulating Th2 lymphocytes to produce TGFP (Ludviksson et al., 1997), which down-regulates the ability of antigen presenting cells to produce IL-12, a powerful activator of Th 1 cell





62


development and inflammation. Thus, the increased levels of IL-12 found in the hippocampus could be the secondary to the loss IL-2.

Increased levels of inflammatory cytokines such as IL-6 may impair hippocampal neurogenesis (Monje et al., 2003; Vallieres et al., 2002); thus, we hypothesized that the alterations in neuroimmunological status would correlate with decreased neurogenesis in the DG of IL-2 KO mice. Against our hypothesis, however, no detectable levels proinflammatory cytokines like IL-6 (though IL-12 can mediate inflammatory responses) or effects of group were apparent in hippocampal neurogenesis. However, a significant group by gender interaction was detectable. Interestingly, this effect appeared to be due to variations in the male mice, but not the females. We previously noted a trend which suggested that IL-2 gene deletion may protect from experimental autoimmune encephalomyelitis (EAE), though we did not statistically analyze the data (Petitto et al., 2000). In that study, three out of the seven male mice utilized developed some symptoms of the disease, whereas none of the six females did. One hypothesis to explain this group by gender interaction is that there may be an interplay between IL-2 and the sex hormones. Sex steroids like estrogen, for example, have been linked to neurogenesis (Gould et al., 2000; Perez-Martin et al., 2003). Further, evidence that IL-2 can regulate sex hormone expression comes from studies of Leydig cell cultures where IL-2 was shown to inhibit steroidogenesis (Guo et al., 1990).

Hormonal differences influenced by IL-2 may give a possible explanation of why males differ from females, but it does not suggest a mechanism of why IL-2 KO mice appear to be protected from impaired hippocampal neurogenesis. As noted above, absence of IL-2 can enhance neuroregenerative properties in the axotomized facial motor





63

nucleus and we speculate an involvement of IL-15. Though IL-15 has not been linked to neurogenesis, its antiapoptotic properties on immune cells are well-studied (Lauwerys et al., 2000; Waldmann, 2002; Waldmann et al., 2001). If these antiapoptotic actions of IL15 can also promote survival of neurons, then this may provide a hypothesis of why the IL-2 KO male mice have higher levels of neurogenesis than the wild-types. Some cytokines are capable of promoting the development of neural stem cells (Rozental et al., 1995; Shah et al., 1996; Wong et al., 2004), and this may account for the how they influence neurogenesis. The actual mechanism whereby IL-2, sex hormones, and neurogenesis may interact is as of yet undefined and requires future study.

In summary, T cells and peripheral cytokines do not appear to enter into the hippocampi of IL-2 KO mice, which does not support the hypothesis that the CNS alterations previously seen in IL-2 KO mice are due to peripheral autoimmunity. Other potential immune indicators of autoimmunity (e.g., deposition of autoantibodies in the brain) were not addressed in this dissertation and require more research. Genetic deletion of IL-2 may, however, alter the neuroimmunological status of the mouse hippocampus through a dysregulation of cytokines produced by CNS cells (e.g., microglia, astroglia). Further studies will be required to determine how these changes impact hippocampal cytoarchitecture and function in IL-2 KO mice.













CHAPTER 5
GENERAL DISCUSSION

Summary of the Overall Findings

The overall purpose of this dissertation research was to investigate the impact of IL-2 on the septohippocampal system by observing changes in the basal physiology and structure of IL-2 KO mice brains vs. their wild-type littermates. During these experiments, we chose not to surgically, chemically, or in any other way experimentally manipulate these animals beyond the genetic knockout of IL-2, so that the effects that we detected could be attributed to loss of IL-2 and not to confounding experimental techniques (e.g., disruption of the BBB). We understand, however, that some experimental manipulations could be useful in the long term and discuss the topic more in the "Caveats and Future Directions" section of towards the end of this chapter.

In the first study in Chapter 2, we determined that IL-2 KO mice suffered a significant 26% loss ofcholinergic neurons in the MS/vDB relative to wild-type littermates. This loss ofcholinergic somata was not reflected by a similar loss of cholinergic fiber density in the hippocampal projection fields of the septal cholinergic neurons. Moreover, the deficits observed were not a general effect on all cholinergic neurons of the brain, as the cholinergic neurons of the striatum in the IL-2 KO animals studied did not appear to be affected. There was, however, a reduction in the distance across the IP, but not SP, GCL of the dentate gyrus.

In Chapter 3, we extended upon these findings by showing that the loss of

cholinergic neurons appeared to occur later in development or during adulthood as a


64





65


potential failure of these neurons to survive, since 21-day-old IL-2 KO animals did not show signs of a loss of cholinergic neurons of the MS/vDB. Furthermore, the neuronal deficiencies appeared to be selective to the cholinergic neurons of the MS/vDB, as there were no similar decreases of GABAergic neurons in IL-2 KO animals. In the hippocampus, consistent with the reductions in length across the IP-GCL of IL-2 KO compared to wild-type mice, the stereologically estimated cell count of Nissl-stained neuronal somata of IL-2 KO mice was significantly lower than wild-type littermates. Also, the hippocampal neurotrophin levels of BDNF and NGF were significantly decreased and increased, respectively.

Finally, in Chapter 4, the cytokine profile of the IL-2 KO hippocampus was altered with increases in IL-15 and IL-12 and the chemokines MCP-1 and IP-10. These experiments confirmed our hypothesis that loss of IL-2 would result in increased levels of brain IL- 15 production. Furthermore, in spite of the known roles of these cytokines and chemokines in T cell trafficking and modulation during brain insult, we were unable to label for elevated levels of the T cells or activated microglia in the hippocampus. The above cytokine profile did not match the serum cytokine levels, suggesting that the changes in cytokines in the hippocampus were likely due to changes in their production in the CNS. Finally, though there was no group alteration in adult GCL neurogenesis, there was a group by gender interaction that appcarcd to be attributable to the male mice.

Implications

This series of studies is the first to demonstrate that endogenous levels of IL-2 may be an important factor in the late development or survival of neurons in the CNS. This impact on CNS neurons involved not only cell number, but also alterations in levels of trophic factors and brain cytokines. Other studies involving IL-2 and the CNS, to date,





66

have depended on in vitro models and/or exogenous administration of IL-2 into the CNS in vivo. Whereas studies such as these can yield valuable information, in vitro models often fail to mimic the complexity of an in vivo system (e.g., lack of glia, different organization of neurons, etc.) and the manipulated in vivo models are subject to experimental design complications (e.g., disruption of the BBB, determination of physiologically meaningful doses of exogenously administered cytokines, etc.). We would be remiss, however, to claim that the IL-2 KO model was completely without complications itself (e.g., autoimmunity such as inflammatory bowel disease), but we feel that the most informative way to address a hypothesis is to examine it in multiple different ways. Thus, taken together with the previous studies in the literature, this dissertation reveals an interesting relationship between IL-2 and the septohippocampal system.

The goal of this study was to elicit whether IL-2 KO impacted the

septohippocampal system, but not to design a treatment or study a clinical disorder. Thus, at its core, this dissertation was basic research. Whereas the pursuit of knowledge is a noble goal, invariably, a simple question often finds itself in the front of our minds: how is this new knowledge clinically relevant and how can we use it?

Already, IL-2 is being used clinically to treat, or being studied as a treatment for, a number of disorders, including cancer (Atkins et al., 1999; Davis and Gillies, 2003; Fyfe et al., 1995; Guirguis et al., 2002; Parkinson et al., 1990; Rosenberg et al., 1985; Rosenberg et al., 1994), and HIV infection (Armstrong and Kazanjian, 2001; Mitsuyasu, 2001; Smith, 2001). The aforementioned treatments are not, however, completely without side effects. In a landmark study, systemic IL-2 treatment for cancer patients





67


was shown to induce dose and time-related behavioral and cognitive changes in 22 out of 44 subjects, including spatial and temporal disorientation (Denicoff et al., 1987). Others have also noted cognitive changes in patients undergoing IL-2-treatment (Capuron et al., 2000; Capuron et al., 2001b; Walker et al., 1997).

Of particular relevance to this dissertation, IL-2 immunotherapy can induce

cognitive changes, including spatial memory deficits and poor task planning (which may be partially dependent on spatial memory) as early as five days into therapy (Capuron et al., 2001a). Capuron's study utilized neuropsychological batteries to assess these cognitive dysfunctions, but did not investigate the physiological mechanism underlying them. One hypothesis to explain at least part of the cognitive changes observed is that the chronic IL-2 treatment used on these patients may affect the septohippocampal system in a similar way that loss of IL-2 causes changes in IL-2 KO animals. Deletion of IL-2 having a similar physiological effect as administration of exogenous IL-2 may seem counterintuitive; however, in at least one population of neurons (i.e., the septal neurons), IL-2 has potent biphasic modulatory actions on ACh release with low sub-pM IL-2 concentrations enhancing and higher nM concentrations inhibiting release (Hanisch et al., 1993; Seto et al., 1997). Thus, under the proposed hypothesis, a loss of IL-2 could lead to a reduction in the stimulatory action on cholinergic neurons, whereas an overabundance of exogenous IL-2 could inhibit release of ACh. Of course, more investigation would be necessary to clarify that hypothesis.

Caveats and Future Directions

In Study 1, we examined the effects of IL-2 KO on the cholinergic

septohippocampal neurons. Many of our findings were extended upon in Study 2. However, we hypothesized that the lack of any change in the cholinergic fiber density in






68


the hippocampus was likely due to compensatory sprouting, but did not overtly assay for any alterations. Thus, to address this issue further, a study utilizing a stereotaxic injection of an anterograde tracer, such as Phaseolus vulgaris-leucoagglutinin (PHAL), into the MS/vDB area would be necessary. The septohippocampal cholinergic neurites would need to be labeled with a marker for AChE and the length and branching of these double-labeled axons would be characterized. If the septohippocampal cholinergic neurons do undergo compensatory sprouting in IL-2 KO mice, we hypothesize that a detectable increase in branching should occur in those animals vs. wild-type littermates.

Though the majority of cholinergic neurons in the MS/vDB project to the

hippocampus (Schwegler et al., 1996b), a small percentage also project to other areas, including the mediodorsal nucleus of the thalamus (Gritti et al., 1998), parietal, cingulate (Gritti et al., 1997), and entorhinal cortices (Alonso and Kohler, 1984). Thus, to accurately claim that the cholinergic neuronal losses observed in the MS/vDB of the IL-2 KO mouse do indeed project to the hippocampus, a study involving a retrograde tracer like Fluoro-Gold injected in the hippocampal projection areas would be necessary. The retrogradely labeled somata of the MS/vDB would need to be double-labeled with ChAT to identify neurons of the cholinergic phenotype and the cell number would be estimated with stereology.

In Study 2, three-week-old IL-2 KO animals did not differ from wild-type

littermates in cholinergic cell number. To expand upon this finding, the brains from animals should be examined at one-week intervals from three-week-old mice to adult eight-week-old mice. This will allow us to establish a time course of cholinergic neuronal loss. Furthermore, we would be interested in investigating the nature of the cell






69

loss. A DNA fragmentation assay, like terminal transferase-mediated dUTP nick-endlabeling (TUNEL), could be used to detect apoptotic cell death and cell death by necrosis could be determined by a simple trypan blue dye exclusion method. Both assays would require immunohistochemical labeling with ChAT to identify the cholinergic somata.

In Study 3, the cytokine and chemokine profile of the IL-2 KO hippocampus was altered compared to wild-type mice. Though we did speculate on the cell types that might produce these cytokines, we did not investigate it further. To identify the potential sources of these cytokines, we could use in situ hybridization to label the mRNA and label the individual cell types with immunohistochemistry. We could expand upon this further by utilizing real time PCR techniques as a semi-quantitative assay of different levels of mRNA expression in homogenates from the hippocampi of IL-2 KO and wildtype mice. These two experiments allow us to identify the cells producing the cytokines and compare the relative levels of mRNA produced for each cytokine between IL-2 KO and wild-type mice.

Several other considerations and questions arise from the data generated from the experiments of this dissertation. First, there is possibility that the reduced levels of ChAT+ immunostaining may be due to a loss of the cholinergic phenotype (e.g., inability to produce ACh) rather than cell death. Alterations in the activity of ACh in the hippocampus (e.g., addition of receptor agonists or antagonists) are sufficient to alter learning and memory (for a review see Gold, 2003). Thus, the behavioral deficits observed in the IL-2 KO animals may not be due to cholinergic neuronal loss. Although the staining intensity was uniform (e.g., consistent between and within groups), to test this further, IL-2 KO and wild-type littermate brains should be Nissl-stained to label and






70


quantify MS/vDB neurons in this brain region. Animals with neuronal loss, should exhibit reduction in total neuronal counts, whereas the neuronal counts would not differ significantly in mice with a loss of phenotype.

Also, considering the difficulty of detection of endogenous IL-2 in the normal CNS, what is the source of the IL-2 in the wild-type brain? Taken together with the unknown non-saturable transport mechanism that allows IL-2 to cross the BBB (Waguespack et al., 1994), and the numerous studies showing cognitive effects of peripherally administered exogenous IL-2 (Capuron et al., 2001a; Denicoffet al., 1987; Lacosta et al., 1999; Walker et al., 1997), an argument can be made that some of the endogenous IL-2 in the CNS could be from the periphery, particularly during development when the BBB is not completely formed. In addition, though we did address the autoimmunity issue of IL-2 KO mice somewhat, we did not completely rule it out as a potential factor in the observed alterations. An experiment that could address both of the above issues would involve crossbreeding the IL-2 KO mice onto an immunodeficient background lacking functional lymphocytes (e.g., RAG-1 knockout mice). Next, an adoptive transfer of normal T lymphocytes from healthy IL-2 wild-type (i.e., non-RAG-i KO) mice to young (i.e., less than 3-week-old) IL-2 KO/RAG-1 KO or IL-2 wild-type/RAG-1 KO littermate animals would establish a functional immune system, and at the same time restore a major source of peripheral IL-2. Also, another set of IL-2 KO/RAG-1 KO and IL-2 wild-type/RAG-1 KO mice should receive an adoptive transfer of IL-2 KO (i.e., non-RAG-I KO) lymphocytes to mimic the autoimmune state of normal IL-2 KO mice. Finally, another group of IL-2 KO/RAG-1 KO and IL-2 wildtype/RAG-i KO mice should receive a sham reconstitution, such that they maintain any





71

normal levels (or lack thereof in IL-2 KO/RAG-1 KO mice) of endogenous brain IL-2 expression, but remain immunodeficient, thus lacking a peripheral source of IL-2. This experimental design would create a model

* that eliminates the impact of peripheral autoimmunity in the IL-2 wild-type
reconstituted IL-2 KO/RAG-I KO mice;

* in which the aforementioned animals are only lacking a CNS, but not peripheral,
source of IL-2;

* in which the IL-2 wild-type reconstituted IL-2 wild-type/RAG-i KO mice would
have a functional peripheral immune system and normal endogenous IL-2
expression in the brain;

* in which the IL-2 KO reconstituted IL-2 KO/RAG-i KO mice are similar in
phenotype to the normal IL-2 KO mouse (i.e., peripheral autoimmunity);

* in which the IL-2 KO reconstituted IL-2 wild-type/RAG-1 KO mice would have
normal brain IL-2 expression, but autoimmunity caused by the peripheral IL-2 KO
T lymphocytes; and

* in which the sham reconstituted IL-2 KO/RAG-1 KO mice lack any source of IL-2
and the sham reconstituted wild-type/RAG-I KO mice lack a peripheral source of
IL-2, with neither of the two models succumbing to autoimmunity.

Thus, if endogenous brain IL-2 is important in septohippocampal development and/or maintenance from the age of three-weeks to adulthood, then we should observe deficits in the septohippocampal system similar to those noted in this dissertation in all reconstitution and sham models above on an IL-2 KO/RAG-1 KO background. If the peripheral source of IL-2 is more important for septohippocampal physiology, then the structural and physiological alterations in this dissertation should be rescued in the above IL-2 wild-type reconstituted cases, but not IL-2 KO or sham reconstituted animals. If autoimmunity is the determining factor, then IL-2 KO reconstituted, but not sham reconstituted, animals will all exhibit deficits in the septohippocampal system regardless of background. This proposed experiment combining IL-2 KO strains, immunodeficient






72

strains, and reconstitution allows us to control for several immunological and genetic factors without disruption of the BBB.

Concluding Remarks

In the peripheral immune system, the complex interplay and interactions of the various cytokines often have redundant, supportive, or even oppositional roles. This complexity allows for a system of compensatory and regulatory control of the immune response. Similar overlaps and checks and balances are also present in the CNS, so understanding how these intricate systems interact can prove daunting. Nevertheless, elucidating the relationship between the brain and immune system molecules may have profound clinical utility. The IL-2 KO mouse is a complicated model to study a cytokine brain interaction and, in this dissertation, we have attempted to simplify the model by not experimentally manipulating it. Our goal was to lay groundwork for future studies on this topic, whereby we hope to more fully understand the precise mechanisms by which IL-2 alters brain physiology.

















WORKS CITED

LA Aarden, TK Brunner, JC Cerottini, J Dayer, AL de Weck, CA Dinarello, G Di
Sabato, JJ Farrar, I Gery, S Gillis, RE Handschumacher, CS Henney, MK
Hoffmann, WJ Koopman, SM Krane, LB Lachman, I Lefkowitz, RI Mishell, SB
Mizel, JJ Oppenheim, V Paetkau, J Plate, M Rollinghoff, D Rosenstreich, AS
Rosenthal, LJ Rosenwasser, A Schimpl, HS Shin, PL Simon, KA Smith, H
Wagner, JD Watson, E Wecker and DD Wood, 1979. Revised nomenclature for
antigen-nonspecific T cell proliferation and helper factors. J Immunol 123, 29282929.

R Ader and N Cohen, 1975. Behaviorally conditioned immunosuppression. Psychosom
Med 37, 333-340.

R Ader, N Cohen and D Bovbjerg, 1982. Conditioned suppression of humoral immunity
in the rat. J Comp Physiol Psychol 96, 517-521.

R Ader, D Felten and N Cohen (Eds), 2001. Psychoneuroimmunology, Third Edition
Edition. Academic Press, San Diego, CA.

L Adorini, 1999. Interleukin-12, a key cytokine in Thl-mediated autoimmune diseases.
Cell Mol Life Sci 55, 1610-1625.

RF Alderson, AL Alterman, YA Barde and RM Lindsay, 1990. Brain-derived
neurotrophic factor increases survival and differentiated functions of rat septal
cholinergic neurons in culture. Neuron 5, 297-306.

F Aloisi, R De Simone, S Columba-Cabezas, G Penna and L Adorini, 2000. Functional
maturation of adult mouse resting microglia into an APC is promoted by
granulocyte-macrophage colony-stimulating factor and interaction with Thl cells. J
Immunol 164, 1705-1712.

A Alonso and C Kohler, 1984. A study of the reciprocal connections between the septum
and the entorhinal area using anterograde and retrograde axonal transport methods
in the rat brain. J Comp Neurol 225, 327-343.

R Alonso, I Chaudieu, J Diorio, A Krishnamurthy, R Quirion and P Boksa, 1993.
Interleukin-2 modulates evoked release of [3H]dopamine in rat cultured
mesencephalic cells. J Neurochem 61, 1284-1290.



73






74

DM Araujo, PA Lapchak, B Collier and R Quirion, 1989. Localization of interleukin-2
immunoreactivity and interleukin-2 receptors in the rat brain: interaction with the
cholinergic system. Brain Res 498, 257-266.

Y Arimatsu and M Miyamoto, 1991. Survival-promoting effect of NGF on in vitro
septohippocampal neurons with cholinergic and GABAergic phenotypes. Brain Res
Dev Brain Res 58, 189-201.

DM Armstrong, G Bruce, LB Hersh and FH Gage, 1987. Development of cholinergic
neurons in the septal/diagonal band complex of the rat. Brain Res 433, 249-256.

WS Armstrong and P Kazanjian, 2001. Use of cytokines in human immunodeficiency
virus-infected patients: colony-stimulating factors, erythropoietin, and interleukin2. Clin Infect Dis 32, 766-773.

MB Atkins, MT Lotze, JP Dutcher, RI Fisher, G Weiss, K Margolin, J Abrams, M Sznol,
D Parkinson, M Hawkins, C Paradise, L Kunkel and SA Rosenberg, 1999. Highdose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 17, 21052116.

IB Autenrieth, N Bucheler, E Bohn, G Heinze and I Horak, 1997. Cytokine mRNA
expression in intestinal tissue of interleukin-2 deficient mice with bowel
inflammation. Gut 41, 793-800.

H Awatsuji, Y Furukawa, M Nakajima, S Furukawa and K Hayashi, 1993. Interleukin-2
as a neurotrophic factor for supporting the survival of neurons cultured from
various regions of fetal rat brain. J Neurosci Res 35, 305-311.

AA Babcock, WA Kuziel, S Rivest and T Owens, 2003. Chemokine expression by glial
cells directs leukocytes to sites of axonal injury in the CNS. J Neurosci 23, 79227930.

R Badolato, AN Ponzi, M Millesimo, LD Notarangelo and T Musso, 1997. Interleukin-15
(IL-15) induces IL-8 and monocyte chemotactic protein 1 production in human
monocytes. Blood 90, 2804-2809.

WA Banks, AJ Kastin and EG Gutierrcz, 1994. Penetration of interleukin-6 across the
murine blood-brain barrier. Neurosci Lett 179, 53-56.

WA Banks, SA Farr and JE Morley, 2002. Entry of blood-borne cytokines into the
central nervous system: effects on cognitive processes. Neuroimmunomodulation
10, 319-327.

WA Banks, L Ortiz, SR Plotkin and AJ Kastin, 1991. Human interleukin (IL) I alpha,
murine IL-1 alpha and murine IL-1 beta are transported from blood to brain in the
mouse by a shared saturable mechanism. J Pharmacol Exp Ther 259, 988-996.





75

MG Baxter, DJ Bucci, TJ Sobel, MJ Williams, LK Gorman and M Gallagher, 1996.
Intact spatial learning following lesions of basal forebrain cholinergic neurons.
Neuroreport 7, 1417-1420.

SA Bayer, 1980. Development of the hippocampal region in the rat. I. Neurogenesis
examined with 3H-thymidine autoradiography. J Comp Neurol 190, 87-114.

JF Bazan, 1992. Unraveling the structure of IL-2. Science 257, 410-413.

RD Beck, Z Huang, GK Ha and JM Petitto, 2004. Altered Hippocampal Structure and
Neurotrophin Levels in Interleukin-2 Knockout Mice. (to be submitted for
publication).

RD Beck, Jr., MA King, Z Huang and JM Petitto, 2002. Alterations in septohippocampal
cholinergic neurons resulting from interleukin-2 gene knockout. Brain Res 955, 1623.

R Bender, M Plaschke, T Naumann, P Wahle and M Frotscher, 1996. Development of
cholinergic and GABAergic neurons in the rat medial septum: different onset of
choline acetyltransferase and glutamate decarboxylase mRNA expression. J Comp
Neurol 372, 204-214.

EN Benveniste and JE Merrill, 1986. Stimulation of oligodendroglial proliferation and
maturation by interleukin-2. Nature 321, 610-613.

EN Benveniste, PK Herman and JN Whitaker, 1987. Myelin basic protein-specific RNA
levels in interleukin-2-stimulated oligodendrocytes. J Neurochem 49, 1274-1279.

M Besser and R Wank, 1999. Cutting edge: clonally restricted production of the
neurotrophins brain-derived neurotrophic factor and neurotrophin-3 mRNA by
human immune cells and Th 1/Th2-polarized expression of their receptors. J
Immunol 162, 6303-6306.

M Bianchi and AE Panerai, 1993. Interleukin-2 enhances scopolamine-induced amnesia
and hyperactivity in the mouse. Neuroreport 4, 1046-1048.

M Bianchi, P Ferrario, N Zonta and AE Panerai, 1995. Effects of interleukin-1 beta and
interleukin-2 on amino acids levels in mouse cortex and hippocampus. Neurorcport
6, 1689-1692.

JL Bizon, JS Han, C Hudon and M Gallagher, 2003. Effects of hippocampal cholinergic
deafferentation on learning strategy selection in a visible platform version of the
water maze. Hippocampus 13, 676-684.

HR Brashear, L Zaborszky and L Heimer, 1986. Distribution of GABAergic and
cholinergic neurons in the rat diagonal band. Neuroscience 17, 439-451.






76


A Braun, M Lommatzsch, A Mannsfeldt, U Neuhaus-Steinmetz, A Fischer, N Schnoy,
GR Lewin and H Renz, 1999. Cellular sources of enhanced brain-derived
neurotrophic factor production in a mouse model of allergic inflammation. Am J
Respir Cell Mol Biol 21, 537-546.

FC Bronfman, D Moechars and F Van Leuven, 2000. Acetylcholinesterase-positive fiber
deafferentation and cell shrinkage in the septohippocampal pathway of aged
amyloid precursor protein london mutant transgenic mice. Neurobiol Dis 7, 152168.

AI Brooks, MW Halterman and HJ Federoff, 1999. Focal hippocampal gain of NGF
function elicits specific septal cholinergic reorganization. Neuroreport 10, 337-344.

KM Buller, 2001. Role of circumventricular organs in pro-inflammatory cytokineinduced activation of the hypothalamic-pituitary-adrenal axis. Clin Exp Pharmacol
Physiol 28, 581-589.

JF Cahill and MG Baxter, 2001. Cholinergic and noncholinergic septal neurons modulate
strategy selection in spatial learning. Eur J Neurosci 14, 1856-1864.

SB Caine, MA Geyer and NR Swerdlow, 1992. Hippocampal modulation of acoustic
startle and prepulse inhibition in the rat. Pharmacol Biochem Behav 43, 1201-1208.

JD Campbell, G Cook, SE Robertson, A Fraser, KS Boyd, JA Gracie and IM Franklin,
2001. Suppression of IL-2-induced T cell proliferation and phosphorylation of
STAT3 and STAT5 by tumor-derived TGF beta is reversed by IL-15. J Immunol
167, 553-561.

M Canossa, O Griesbeck, B Beminger, G Campana, R Kolbeck and H Thoenen, 1997.
Neurotrophin release by neurotrophins: implications for activity-dependent
neuronal plasticity. Proc Natl Acad Sci U S A 94, 13279-13286.

L Capuron, A Ravaud and R Dantzer, 2000. Early depressive symptoms in cancer
patients receiving interleukin 2 and/or interferon alfa-2b therapy. J Clin Oncol 18,
2143-2151.

L Capuron, A Ravaud and R Dantzer, 2001a. Timing and specificity of the cognitive
changes induced by interleukin-2 and interferon-alpha treatments in cancer patients.
Psychosom Med 63, 376-386.

L Capuron, A Ravaud, N Gualde, E Bosmans, R Dantzer, M Maes and PJ Neveu, 2001b.
Association between immune activation and early depressive symptoms in cancer patients treated with interleukin-2-based therapy. Psychoneuroendocrinology 26,
797-808.

LL Carter, X Zhang, C Dubey, P Rogers, L Tsui and SL Swain, 1998. Regulation of T
cell subsets from naive to memory. J Immunother 21, 181-187.






77


JC Cassel, E Duconseille, H Jeltsch and B Will, 1997. The fimbria-fomix/cingular bundle
pathways: a review of neurochemical and behavioural approaches using lesions and
transplantation techniques. Prog Neurobiol 51, 663-716.

JP Chandler and KA Crutcher, 1983. The septohippocampal projection in the rat: an
electron microscopic horseradish peroxidase study. Neuroscience 10, 685-696.

Q Chang and PE Gold, 2004. Impaired and spared cholinergic functions in the
hippocampus after lesions of the medial septum/vertical limb of the diagonal band
with 192 IgG-saporin. Hippocampus 14, 170-179.

J Chappell, R McMahan, A Chiba and M Gallagher, 1998. A re-examination of the role
of basal forebrain cholinergic neurons in spatial working memory.
Neuropharmacology 37, 481-487.

TJ Collier and A Routtenberg, 1984. Selective impairment of declarative memory
following stimulation of dentate gyrus granule cells: a naloxone-sensitive effect.
Brain Res 310, 384-387.

JM Conner and S Varon, 1997. Developmental profile of NGF immunoreactivity in the
rat brain: a possible role of NGF in the establishment of cholinergic terminal fields
in the hippocampus and cortex. Brain Res Dev Brain Res 101, 67-79.

JM Conner, D Muir, S Varon, T Hagg and M Manthorpe, 1992. The localization of nerve
growth factor-like immunoreactivity in the adult rat basal forebrain and
hippocampal formation. J Comp Neurol 319, 454-462.

CD Conrad and EJ Roy, 1993. Selective loss of hippocampal granule cells following
adrenalectomy: implications for spatial memory. J Neurosci 13, 2582-2590.

AK Cross and MN Woodroofe, 1999. Chemokines induce migration and changes in actin
polymerization in adult rat brain microglia and a human fetal microglial cell line in
vitro. J Neurosci Res 55, 17-23.

P Curzon, DJ Kim and MW Decker, 1994. Effect of nicotine, lobeline, and
mecamylamine on sensory gating in the rat. Pharmacol Biochem Behav 49, 877882.

JD Cushman, Z Huang, G Ha, R Beck, J Lo and JM Petitto, 2004. Contextual fear
discrimination learning and memory is disrupted in IL-2 knockout mice. Society of
Neuroscience Abstracts.

CB Davis and SD Gillies, 2003. Immunocytokines: amplification of anti-cancer
immunity. Cancer Immunol Immunother 52, 297-308.





78

KD Denicoff, DR Rubinow, MZ Papa, C Simpson, CA Seipp, MT Lotze, AE Chang, D
Rosenstein and SA Rosenberg, 1987. The neuropsychiatric effects of treatment
with interleukin-2 and lymphokine-activated killer cells. Ann Intern Med 107, 293300.

JH Dufour, M Dziejman, MT Liu, JH Leung, TE Lane and AD Luster, 2002. IFNgamma-inducible protein 10 (IP-10; CXCL10)-deficient mice reveal a role for IP10 in effector T cell generation and trafficking. J Immunol 168, 3195-3204.

AJ Dunn, 2002. Mechanisms by which cytokines signal the brain. Int Rev Neurobiol 52,
43-65.

O Eizenberg, A Faber-Elman, M Lotan and M Schwartz, 1995. Interleukin-2 transcripts
in human and rodent brains: possible expression by astrocytes. J Neurochem 64,
1928-1936.

H Fukuoka, M Kawatani, T Hisamitsu and C Takeshige, 1994. Cutaneous hyperalgesia
induced by peripheral injection of interleukin-1 beta in the rat. Brain Res 657, 133140.

G Fyfe, RI Fisher, SA Rosenberg, M Sznol, DR Parkinson and AC Louie, 1995. Results
of treatment of 255 patients with metastatic renal cell carcinoma who received
high-dose recombinant interleukin-2 therapy. J Clin Oncol 13, 688-696.

SL Gaffen, 2001. Signaling domains of the interleukin 2 receptor. Cytokine 14, 63-77.

FH Gage and A Bjorklund, 1987. Trophic and growth-regulating mechanisms in the
central nervous system monitored by intracerebral neural transplants. Ciba Found
Symp 126, 143-159.

FH Gage, A Bjorklund and U Stenevi, 1983a. Reinnervation of the partially deafferented
hippocampus by compensatory collateral sprouting from spared cholinergic and
noradrenergic afferents. Brain Res 268, 27-37.

FH Gage, A Bjorklund and U Stenevi, 1984. Cells of origin of the ventral cholinergic
septohippocampal pathway undergoing compensatory collateral sprouting
following fimbria-fornix transection. Neurosci Lett 44, 211-216.

FH Gage, A Bjorklund, U Stenevi and SB Dunnett, 1983b. Functional correlates of
compensatory collateral sprouting by aminergic and cholinergic afferents in the
hippocampal formation. Brain Res 268, 39-47.

BH Gahwiler, A Enz and F Hefti, 1987. Nerve growth factor promotes development of
the rat septo-hippocampal cholinergic projection in vitro. Neurosci Lett 75, 6-10.





79

D Galey, C Destrade and R Jaffard, 1994. Relationships between septo-hippocampal
cholinergic activation and the improvement of long-term retention produced by
medial septal electrical stimulation in two inbred strains of mice. Behav Brain Res
60, 183-189.

R Ganguli, JS Brar, KR Chengappa, M DeLeo, ZW Yang, G Shurin and BS Rabin, 1995.
Mitogen-stimulated interleukin-2 production in never-medicated, first- episode
schizophrenic patients. The influence of age at onset and negative symptoms. Arch
Gen Psychiatry 52, 668-672.

JG Giri, DM Anderson, S Kumaki, LS Park, KH Grabstein and D Cosman, 1995. IL-15, a
novel T cell growth factor that shares activities and receptor components with IL-2.
J Leukoc Biol 57, 763-766.

PE Gold, 2003. Acetylcholine modulation of neural systems involved in learning and
memory. Neurobiol Learn Mem 80, 194-210.

E Gould, NJ Woolf and LL Butcher, 1991. Postnatal development of cholinergic neurons
in the rat: I. Forebrain. Brain Res Bull 27, 767-789.

E Gould, P Tanapat, T Rydel and N Hastings, 2000. Regulation of hippocampal
neurogenesis in adulthood. Biol Psychiatry 48, 715-720.

I Gritti, M Mariotti and M Mancia, 1998. GABAergic and cholinergic basal forebrain and
preoptic-anterior hypothalamic projections to the mediodorsal nucleus of the
thalamus in the cat. Neuroscience 85, 149-178.

I Gritti, L Mainville, M Mancia and BE Jones, 1997. GABAergic and other
noncholinergic basal forebrain neurons, together with cholinergic neurons, project
to the mesocortex and isocortex in the rat. J Comp Neurol 383, 163-177.

LM Guirguis, JC Yang, DE White, SM Steinberg, DJ Liewehr, SA Rosenberg and DJ
Schwartzentruber, 2002. Safety and efficacy of high-dose interleukin-2 therapy in
patients with brain metastases. J Immunother 25, 82-87.

H Guo, JH Calkins, MM Sigel and T Lin, 1990. Interleukin-2 is a potent inhibitor of
Leydig cell steroidogenesis. Endocrinology 127, 1234-1239.

EG Gutierrez, WA Banks and AJ Kastin, 1993. Murine tumor necrosis factor alpha is
transported from blood to brain in the mouse. J Neuroimmunol 47, 169-176.

T Hagg, HL Vahlsing, M Manthorpe and S Varon, 1990. Nerve growth factor infusion
into the denervated adult rat hippocampal formation promotes its cholinergic
reinnervation. J Neurosci 10, 3087-3092.

UK Hanisch and R Quirion, 1995a. Interleukin-2 as a neuroregulatory cytokine. Brain
Res Brain Res Rev 21, 246-284.






80

UK Hanisch and R Quirion, 1995b. Interleukin-2 as a neuroregulatory cytokine. Brain
Res Brain Res Rev 21, 246-284.

UK Hanisch, D Seto and R Quirion, 1993. Modulation of hippocampal acetylcholine
release: a potent central action of interleukin-2. J Neurosci 13, 3368-3374.

UK Hanisch, J Neuhaus, R Quirion and H Kettenmann, 1996. Neurotoxicity induced by
interleukin-2: involvement of infiltrating immune cells. Synapse 24, 104-114.

UK Hanisch, W Rowe, S Sharma, MJ Meaney and R Quirion, 1994. Hypothalamicpituitary-adrenal activity during chronic central administration of interleukin-2.
Endocrinology 135, 2465-2472.

UK Hanisch, SA Lyons, M Prinz, C Nolte, JR Weber, H Kettenmann and F Kirchhoff,
1997a. Mouse brain microglia express interleukin-15 and its multimeric receptor complex functionally coupled to Janus kinase activity. J Biol Chem 272, 2885328860.

UK Hanisch, J Neuhaus, W Rowe, D Van Rossum, T Moller, H Kettenmann and R
Quirion, 1997b. Neurotoxic consequences of central long-term administration of
interleukin-2 in rats. Neuroscience 79, 799-818.

J Hartikka and F Hefti, 1988. Comparison of nerve growth factor's effects on
development of septum, striatum, and nucleus basalis cholinergic neurons in vitro.
J Neurosci Res 21, 352-364.

YW He and TR Malek, 1998. The structure and function of gamma c-dependent
cytokines and receptors: regulation of T lymphocyte development and homeostasis.
Crit Rev Immunol 18, 503-524.

JC Hedreen, SJ Bacon and DL Price, 1985. A modified histochemical technique to
visualize acetylcholinesterase- containing axons. J Histochem Cytochem 33, 134140.

R Hellweg, C Humpel, A Lowe and H Hortnagl, 1997. Moderate lesion of the rat
cholinergic septohippocampal pathway increases hippocampal nerve growth factor synthesis: evidence for long-term compensatory changes? Brain Res Mol Brain Res
45, 177-181.

WF Hickey and H Kimura, 1988. Perivascular microglial cells of the CNS are bone
marrow-derived and present antigen in vivo. Science 239, 290-292.

WF Hickey, BL Hsu and H Kimura, 1991. T-lymphocyte entry into the central nervous
system. J Neurosci Res 28, 254-260.

I Horak, 1995. Immunodeficiency in IL-2-knockout mice. Clin Immunol Immunopathol
76, S172-173.





81

I Horak, 1996. Interleukin-2-knockout mice: a new model to study autoimmunity and
self-tolerance. Sb Lek 97, 25-28.

AA Hurwitz, WD Lyman and JW Berman, 1995. Tumor necrosis factor alpha and
transforming growth factor beta upregulate astrocyte expression of monocyte
chemoattractant protein-1. J Neuroimmunol 57, 193-198.

LS Janis, MM Glasier, Z Fulop and DG Stein, 1998. Intraseptal injections of 192 IgG
saporin produce deficits for strategy selection in spatial-memory tasks. Behav Brain
Res 90, 23-34.

DA Johnson, NJ Zambon and RB Gibbs, 2002. Selective lesion of cholinergic neurons in
the medial septum by 192 IgG-saporin impairs learning in a delayed matching to
position T-maze paradigm. Brain Res 943, 132-141.

P Kaiser, L Rothwell, D Vasicek and K Hala, 2002. A role for IL-15 in driving the onset
of spontaneous autoimmune thyroiditis? J Immunol 168, 4216-4220.

S Karanth, K Lyson and SM McCann, 1993. Role of nitric oxide in interleukin 2-induced
corticotropin-releasing factor release from incubated hypothalami. Proc Natl Acad
Sci U S A 90, 3383-3387.

M Kerschensteiner, E Gallmeier, L Behrens, VV Leal, T Misgeld, WE Klinkert, R
Kolbeck, E Hoppe, RL Oropeza-Wekerle, I Bartke, C Stadelmann, H Lassmann, H
Wekerle and R Hohlfeld, 1999. Activated human T cells, B cells, and monocytes
produce brain-derived neurotrophic factor in vitro and in inflammatory brain
lesions: a neuroprotective role of inflammation? J Exp Med 189, 865-870.

MA King, BE Hunter, RL Reep and DW Walker, 1989. Acetylcholinesterase stain
intensity variation in the rat dentate gyrus: a quantitative description based on
digital image analysis. Neuroscience 33, 203-221.

J Kiss, AJ Patel and TF Freund, 1990a. Distribution of septohippocampal neurons
containing parvalbumin or choline acetyltransferase in the rat brain. J Comp Neurol
298, 362-372.

J Kiss, AJ Patel, KG Baimbridge and TF Freund, 1990b. Topographical localization of
neurons containing parvalbumin and choline acetyltransferase in the medial
septum-diagonal band region of the rat. Neuroscience 36, 61-72.

M Knipper and RJ Rylett, 1997. A new twist in an old story: the role for crosstalk of
neuronal and trophic activity. Neurochem Int 31, 659-676.

M Knipper, M da Penha Berzaghi, A Blochl, H Breer, H Thoenen and D Lindholm, 1994.
Positive feedback between acetylcholine and the neurotrophins nerve growth factor
and brain-derived neurotrophic factor in the rat hippocampus. Eur J Neurosci 6,
668-671.






82


M Koch, 1996a. The septohippocampal system is involved in prepulse inhibition of the
acoustic startle response in rats. Behav Neurosci 110, 468-477.

M Koch, 1996b. The septohippocampal system is involved in prepulse inhibition of the
acoustic startle response in rats. Behav Neurosci 110, 468-477.

S Lacosta, Z Merali and H Anisman, 1999. Influence of acute and repeated interleukin-2
administration on spatial learning, locomotor activity, exploratory behaviors, and
anxiety. Behav Neurosci 113, 1030-1041.

H Lal and MJ Forster, 1988. Autoimmunity and age-associated cognitive decline.
Neurobiol Aging 9, 733-742.

MR Lamprea, FP Cardenas, R Silveira, S Morato and TJ Walsh, 2000. Dissociation of
memory and anxiety in a repeated elevated plus maze paradigm: forebrain
cholinergic mechanisms. Behav Brain Res 117, 97-105.

PA Lapchak and DM Araujo, 1993. Interleukin-2 regulates monoamine and opioid
peptide release from the hypothalamus. Neuroreport 4, 303-306.

PA Lapchak, DM Araujo, R Quirion and A Beaudet, 1991. Immunoautoradiographic
localization of interleukin 2-like immunoreactivity and interleukin 2 receptors (Tac
antigen-like immunoreactivity) in the rat brain. Neuroscience 44, 173-184.

E Larsson, RJ Mandel, RL Klein, N Muzyczka, O Lindvall and Z Kokaia, 2002.
Suppression of insult-induced neurogenesis in adult rat brain by brain-derived
neurotrophic factor. Exp Neurol 177, 1-8.

JM Lauder and UB Schambra, 1999. Morphogenetic roles of acetylcholine. Environ
Health Perspect 107 Suppl 1, 65-69.

BR Lauwerys, N Garot, JC Renauld and FA Houssiau, 2000. Cytokine production and
killer activity of NK/T-NK cells derived with IL-2, IL-15, or the combination of
IL-12 and IL-18. J Immunol 165, 1847-1853.

G Leanza, OG Nilsson and A Bjorklund, 1993. Compensatory changes of in vivo
acetylcholine and noradrenaline release in the hippocampus after partial deafferentation, as monitored by microdialysis. Brain Res 615, 147-159.

G Leanza, OG Nilsson, RG Wiley and A Bjorklund, 1995. Selective lesioning of the
basal forebrain cholinergic system by intraventricular 192 IgG-saporin:
behavioural, biochemical and stereological studies in the rat. Eur J Neurosci 7, 329343.

J Lee, W Duan and MP Mattson, 2002. Evidence that brain-derived neurotrophic factor is
required for basal neurogenesis and mediates, in part, the enhancement of
neurogenesis by dietary restriction in the hippocampus of adult mice. J Neurochem
82, 1367-1375.






83


YB Lee, J Satoh, DG Walker and SU Kim, 1996. Interleukin-15 gene expression in
human astrocytes and microglia in culture. Neuroreport 7, 1062-1066.

WJ Leonard, JM Depper, GR Crabtree, S Rudikoff, J Pumphrey, RJ Robb, M Kronke, PB
Svetlik, NJ Peffer, TA Waldmann and WC Greene, 1984. Molecular cloning and expression of cDNAs for the human interleukin-2 receptor. Nature 311, 626-631.

J Licinio, JP Seibyl, M Altemus, DS Charney and JH Krystal, 1993. Elevated CSF levels
of interleukin-2 in neuroleptic-free schizophrenic patients. Am J Psychiatry 150,
1408-1410.

T Lindenlaub, P Teuteberg, T Hartung and C Sommer, 2000. Effects of neutralizing
antibodies to TNF-alpha on pain-related behavior and nerve regeneration in mice
with chronic constriction injury. Brain Res 866, 15-22.

BR Ludviksson, RO Ehrhardt and W Strober, 1997. TGF-beta production regulates the
development of the 2,4,6-trinitrophenol-conjugated keyhole limpet hemocyanininduced colonic inflammation in IL-2-deficient mice. J Immunol 159, 3622-3628.

A Ma, M Datta, E Margosian, J Chen and I Horak, 1995. T cells, but not B cells, are
required for bowel inflammation in interleukin 2-deficient mice. J Exp Med 182,
1567-1572.

SF Maier, LE Goehler, M Fleshner and LR Watkins, 1998. The role of the vagus nerve in
cytokine-to-brain communication. Ann N Y Acad Sci 840, 289-300.

R Makuch, J Baratta, LD Karaelias, JC Lauterborn, CM Gall, J Yu and RT Robertson,
2001. Arrival of afferents and the differentiation of target neurons: studies of
developing cholinergic projections to the dentate gyrus. Neuroscience 104, 81-91.

D Maslinska, 2001. The cytokine network and interleukin-15 (IL-15) in brain
development. Folia Neuropathol 39, 43-47.

CG McAllister, DP van Kammen, TJ Rehn, AL Miller, J Gurklis, ME Kelley, J Yao and
JL Peters, 1995. Increases in CSF levels of interleukin-2 in schizophrenia: effects
of recurrence of psychosis and medication status. Am J Psychiatry 152, 1291-1297.

JF McDyer, Z Li, S John, X Yu, CY Wu and JA Ragheb, 2002. IL-2 receptor blockade
inhibits late, but not early, IFN-gamma and CD40 ligand expression in human T
cells: disruption of both IL-12-dependent and -independent pathways of IFNgamma production. J Immunol 169, 2736-2746.

RL McLamb, WR Mundy and HA Tilson, 1988. Intradentate colchicine disrupts the
acquisition and performance of a working memory task in the radial arm maze.
Neurotoxicology 9, 521-528.






84


L Meda, S Bernasconi, C Bonaiuto, S Sozzani, D Zhou, L Otvos, Jr., A Mantovani, F
Rossi and MA Cassatella, 1996. Beta-amyloid (25-35) peptide and IFN-gamma synergistically induce the production of the chemotactic cytokine MCP-1/JE in
monocytes and microglial cells. J Immunol 157, 1213-1218.

MA Meijssen, SL Brandwein, HC Reinecker, AK Bhan and DK Podolsky, 1998.
Alteration of gene expression by intestinal epithelial cells precedes colitis in
interleukin-2-deficient mice. Am J Physiol 274, G472-479.

F Mennicken and R Quirion, 1997. Interleukin-2 increases choline acetyltransferase
activity in septal- cell cultures. Synapse 26, 175-183.

RT Mitsuyasu, 2001. The potential role of interleukin-2 in HIV. Aids 15 Suppl 2, S22-27.

H Mizuma, K Takagi, K Miyake, N Takagi, K Ishida, S Takeo, A Nitta, H Nomoto, Y
Furukawa and S Furukawa, 1999. Microsphere embolism-induced elevation of
nerve growth factor level and appearance of nerve growth factor immunoreactivity
in activated T-lymphocytes in the rat brain. J Neurosci Res 55, 749-761.

G Moalem, A Gdalyahu, Y Shani, U Otten, P Lazarovici, IR Cohen and M Schwartz,
2000. Production of neurotrophins by activated T cells: implications for
neuroprotective autoimmunity. J Autoimmun 15, 331-345.

WC Mobley, JL Rutkowski, GI Tennekoon, J Gemski, K Buchanan and MV Johnston,
1986. Nerve growth factor increases choline acetyltransferase activity in
developing basal forebrain neurons. Brain Res 387, 53-62.

ML Monje, H Toda and TD Palmer, 2003. Inflammatory blockade restores adult
hippocampal neurogenesis. Science 302, 1760-1765.

DA Morgan, FW Ruscetti and R Gallo, 1976. Selective in vitro growth of T lymphocytes
from normal human bone marrows. Science 193, 1007-1008.

JK Morse, SJ Wiegand, K Anderson, Y You, N Cai, J Carnahan, J Miller, PS DiStefano,
CA Altar, RM Lindsay and et al., 1993. Brain-derived neurotrophic factor (BDNF)
prevents the degeneration of medial septal cholinergic neurons following fimbria
transection. J Neurosci 13, 4146-4156.

N Muller and M Ackenheil, 1998. Psychoneuroimmunology and the cytokine action in
the CNS: implications for psychiatric disorders. Prog Neuropsychopharmacol Biol
Psychiatry 22, 1-33.

Y Nakamura, SM Russell, SA Mess, M Friedmann, M Erdos, C Francois, Y Jacques, S
Adelstein and WJ Leonard, 1994. Heterodimerization of the IL-2 receptor beta- and gamma-chain cytoplasmic domains is required for signalling. Nature 369, 330-333.






85


KP Nanry, WR Mundy and HA Tilson, 1989. Colchicine-induced alterations of reference
memory in rats: role of spatial versus non-spatial task components. Behav Brain
Res 35, 45-53.

BH Nelson, 2004. IL-2, regulatory T cells, and tolerance. J Immunol 172, 3983-3988.

BH Nelson, JD Lord and PD Greenberg, 1994. Cytoplasmic domains of the interleukin-2
receptor beta and gamma chains mediate the signal for T-cell proliferation. Nature
369, 333-336.

R Nemni, S lannaccone, A Quattrini, S Smirne, M Sessa, M Lodi, C Erminio and N
Canal, 1992. Effect of chronic treatment with recombinant interleukin-2 on the
central nervous system of adult and old mice. Brain Res 591, 248-252.

W Pan, WA Banks and AJ Kastin, 1997. Permeability of the blood-brain and bloodspinal cord barriers to interferons. J Neuroimmunol 76, 105-111.

K Pardy, D Murphy, D Carter and KM Hui, 1993. The influence of interleukin-2 on
vasopressin and oxytocin gene expression in the rodent hypothalamus. J
Neuroimmunol 42, 131-138.

DR Parkinson, JS Abrams, PH Wiernik, AA Rayner, KA Margolin, DA Van Echo, M
Sznol, JP Dutcher, FR Aronson, JH Doroshow and et al., 1990. Interleukin-2
therapy in patients with metastatic malignant melanoma: a phase II study. J Clin
Oncol 8, 1650-1656.

A Patapoutian and LF Reichardt, 2001. Trk receptors: mediators of neurotrophin action.
Curr Opin Neurobiol 11, 272-280.

S Pauli, AC Linthorst and JM Reul, 1998. Tumour necrosis factor-alpha and interleukin-2
differentially affect hippocampal serotonergic neurotransmission, behavioural
activity, body temperature and hypothalamic-pituitary-adrenocortical axis activity
in the rat. Eur J Neurosci 10, 868-878.

M Perez-Martin, I Azcoitia, JL Trejo, A Sierra and LM Garcia-Segura, 2003. An
antagonist of estrogen receptors blocks the induction of adult neurogenesis by
insulin-like growth factor-I in the dentate gyrus of adult female rat. Eur J Neurosci
18, 923-930.

T Perry, H Hodges and JA Gray, 2001. Behavioural, histological and
immunocytochemical consequences following 192 IgG-saporin immunolesions of
the basal forebrain cholinergic system. Brain Res Bull 54, 29-48.

JM Petitto and Z Huang, 1994. Molecular cloning of a partial cDNA of the interleukin-2
receptor-beta in normal mouse brain: in situ localization in the hippocampus and
expression by neuroblastoma cells. Brain Res 650, 140-145.





86

JM Petitto and Z Huang, 1995. Molecular cloning of the coding sequence of an
interleukin-2 receptor alpha subunit cDNA in murine brain. J Neuroimmunol 59,
135-141.

JM Petitto and Z Huang, 2001. Cloning the full-length IL-2/15 receptor-beta cDNA
sequence from mouse brain: evidence of enrichment in hippocampal formation
neurons. Regul Pept 98, 77-87.

JM Petitto, Z Huang, J Lo and WJ Streit, 2003. IL-2 gene knockout affects T lymphocyte
trafficking and the microglial response to regenerating facial motor neurons. J
Neuroimmunol 134, 95-103.

JM Petitto, DB McCarthy, CM Rinker, Z Huang and T Getty, 1997. Modulation of
behavioral and neurochemical measures of forebrain dopamine function in mice by
species-specific interleukin-2. J Neuroimmunol 73, 183-190.

JM Petitto, Z Huang, MK Raizada, CM Rinker and DB McCarthy, 1998. Molecular
cloning of the cDNA coding sequence of IL-2 receptor-gamma (gammac) from
human and murine forebrain: expression in the hippocampus in situ and by brain
cells in vitro. Brain Res Mol Brain Res 53, 152-162.

JM Petitto, RK McNamara, PL Gendreau, Z Huang and AJ Jackson, 1999. Impaired
learning and memory and altered hippocampal neurodevelopment resulting from
interleukin-2 gene deletion. J Neurosci Res 56, 441-446.

JM Petitto, WJ Streit, Z Huang, E Butfiloski and J Schiffenbauer, 2000. Interleukin-2
gene deletion produces a robust reduction in susceptibility to experimental
autoimmune encephalomyelitis in C57BL/6 mice. Neurosci Lett 285, 66-70.

CR Plata-Salaman and JM ffrench-Mullen, 1993. Interleukin-2 modulates calcium
currents in dissociated hippocampal CA1 neurons. Neuroreport 4, 579-581.

LQ Ren, N Gourmala, HW Boddeke and PJ Gebicke-Haerter, 1998. Lipopolysaccharideinduced expression of IP-10 mRNA in rat brain and in cultured rat astrocytes and
microglia. Brain Res Mol Brain Res 59, 256-263.

GE Ringheim, BD Freimark and RJ Robb, 1991. Quantitative characterization of the
intrinsic ligand-binding affinity of the interleukin 2 receptor beta chain and its
modulation by the alpha chain and a second affinity-modulating element.
Lymphokine Cytokine Res 10, 219-224.

SA Rosenberg, JJ Mule, PJ Spiess, CM Reichert and SL Schwarz, 1985. Regression of
established pulmonary metastases and subcutaneous tumor mediated by the
systemic administration of high-dose recombinant interleukin 2. J Exp Med 161,
1169-1188.





87


SA Rosenberg, JC Yang, SL Topalian, DJ Schwartzentruber, JS Weber, DR Parkinson,
CA Seipp, JH Einhorn and DE White, 1994. Treatment of 283 consecutive patients with metastatic melanoma or renal cell cancer using high-dose bolus interleukin 2.
Jama 271, 907-913.

R Rozental, MF Mehler, M Morales, AF Andrade-Rozental, JA Kessler and DC Spray,
1995. Differentiation of hippocampal progenitor cells in vitro: temporal expression
of intercellular coupling and voltage- and ligand-gated responses. Dev Biol 167,
350-362.

F Ruberti, S Capsoni, A Comparini, E Di Daniel, J Franzot, S Gonfloni, G Rossi, N
Berardi and A Cattaneo, 2000. Phenotypic knockout of nerve growth factor in adult
transgenic mice reveals severe deficits in basal forebrain cholinergic neurons, cell
death in the spleen, and skeletal muscle dystrophy. J Neurosci 20, 2589-2601.

T Saarelainen, S Vaittinen and E Castren, 2001. trkB-receptor activation contributes to
the kainate-induced increase in BDNF mRNA synthesis. Cell Mol Neurobiol 21,
429-435.

N Sakai, S Kaufman and S Milstien, 1995. Parallel induction of nitric oxide and
tetrahydrobiopterin synthesis by cytokines in rat glial cells. J Neurochem 65, 895902.

B Sakic, H Szechtman and JA Denburg, 1997. Neurobehavioral alterations in
autoimmune mice. Neurosci Biobehav Rev 21, 327-340.

B Sakic, H Szechtman, S Denburg, R Carbotte and JA Denburg, 1993. Spatial learning
during the course of autoimmune disease in MRL mice. Behav Brain Res 54, 5766.

RP Saneto, A Altman, RL Knobler, HM Johnson and J de Vellis, 1986. Interleukin 2
mediates the inhibition of oligodendrocyte progenitor cell proliferation in vitro.
Proc Natl Acad Sci U S A 83, 9221-9225.

M Sarder, H Saito and K Abe, 1993. Interleukin-2 promotes survival and neurite
extension of cultured neurons from fetal rat brain. Brain Res 625, 347-350.

M Sarder, K Abe, H Saito and N Nishiyama, 1996. Comparative effect of IL-2 and IL-6
on morphology of cultured hippocampal neurons from fetal rat brain. Brain Res
715, 9-16.

J Satoh, K Kurohara, M Yukitake and Y Kuroda, 1998. Interleukin- 15, a T-cell growth
factor, is expressed in human neural cell lines and tissues. J Neurol Sci 155, 170177.

A Schimpl, I Berberich, B Kneitz, S Kramer, B Santner-Nanan, S Wagner, M Wolf and T
Hunig, 2002. IL-2 and autoimmune disease. Cytokine Growth Factor Rev 13, 369378.





88


E Schmitt, T Germann, S Goedert, P Hoehn, C Huels, S Koelsch, R Kuhn, W Muller, N
Palm and E Rude, 1994. IL-9 production of naive CD4+ T cells depends on IL-2, is
synergistically enhanced by a combination of TGF-beta and IL-4, and is inhibited
by IFN-gamma. J Immunol 153, 3989-3996.

R Schopke, DP Wolfer, HP Lipp and MC Leisinger-Trigona, 1991. Swimming
navigation and structural variations of the infrapyramidal mossy fibers in the
hippocampus of the mouse. Hippocampus 1, 315-328.

H Schwegler and WE Crusio, 1995. Correlations between radial-maze learning and
structural variations of septum and hippocampus in rodents. Behav Brain Res 67,
29-41.

H Schwegler, WE Crusio, HP Lipp and B Heimrich, 1988. Water-maze learning in the
mouse correlates with variation in hippocampal morphology. Behav Genet 18, 153165.

H Schwegler, M Boldyreva, R Linke, J Wu, K Zilles and WE Crusio, 1996a. Genetic
variation in the morphology of the septo-hippocampal cholinergic and GABAergic
systems in mice: II. Morpho-behavioral correlations. Hippocampus 6, 535-545.

H Schwegler, M Boldyreva, M Pyrlik-Gohlmann, R Linke, J Wu and K Zilles, 1996b.
Genetic variation in the morphology of the septo-hippocampal cholinergic and
GABAergic system in mice. I. Cholinergic and GABAergic markers. Hippocampus
6, 136-148.

BM Segal, BK Dwyer and EM Shevach, 1998. An interleukin (IL)-10/IL-12
immunoregulatory circuit controls susceptibility to autoimmune disease. J Exp Med
187, 537-546.

D Seto, S Kar and R Quirion, 1997. Evidence for direct and indirect mechanisms in the
potent modulatory action of interleukin-2 on the release of acetylcholine in rat
hippocampal slices. Br J Pharmacol 120, 1151-1157.

NM Shah, AK Groves and DJ Anderson, 1996. Alternative neural crest cell fates are
instructively promoted by TGFbeta superfamily members. Cell 85, 331-343.

M Shimojo, Y Imai, K Nakajima, S Mizushima, A Uemura and S Kohsaka, 1993.
Interleukin-2 enhances the viability of primary cultured rat neocortical neurons.
Neurosci Lett 151, 170-173.

KA Smith, 2001. Low-dose daily interleukin-2 immunotherapy: accelerating immune
restoration and expanding HIV-specific T-cell immunity without toxicity. Aids 15
Suppl 2, S28-35.

LS Sorkin, WH Xiao, R Wagner and RR Myers, 1997. Tumour necrosis factor-alpha
induces ectopic activity in nociceptive primary afferent fibres. Neuroscience 81,
255-262.





89

WJ Streit, MB Graeber and GW Kreutzberg, 1988. Functional plasticity of microglia: a
review. Glia 1, 301-307.

WJ Streit, SL Semple-Rowland, SD Hurley, RC Miller, PG Popovich and BT Stokes,
1998. Cytokine mRNA profiles in contused spinal cord and axotomized facial
nucleus suggest a beneficial role for inflammation and gliosis. Exp Neurol 152, 7487.

K Sugamura, H Asao, M Kondo, N Tanaka, N Ishii, K Ohbo, M Nakamura and T
Takeshita, 1996. The interleukin-2 receptor gamma chain: its role in the multiple
cytokine receptor complexes and T cell development in XSCID. Annu Rev
Immunol 14, 179-205.

H Super and E Soriano, 1994. The organization of the embryonic and early postnatal
murine hippocampus. II. Development of entorhinal, commissural, and septal
connections studied with the lipophilic tracer Dil. J Comp Neurol 344, 101-120.

T Takeshita, H Asao, K Ohtani, N Ishii, S Kumaki, N Tanaka, H Munakata, M Nakamura
and K Sugamura, 1992. Cloning of the gamma chain of the human IL-2 receptor.
Science 257, 379-382.

V Tancredi, C Zona, F Velotti, F Eusebi and A Santoni, 1990. Interleukin-2 suppresses
established long-term potentiation and inhibits its induction in the rat hippocampus.
Brain Res 525, 149-151.

LJ Thal, E Gilbertson, DM Armstrong and FH Gage, 1992. Development of the basal
forebrain cholinergic system: phenotype expression prior to target innervation.
Neurobiol Aging 13, 67-72.

JJ Thune and B Pakkenberg, 2000. Stereological studies of the schizophrenic brain. Brain
Res Brain Res Rev 31, 200-204.

L Vallieres, IL Campbell, FH Gage and PE Sawchenko, 2002. Reduced hippocampal
neurogenesis in adult transgenic mice with chronic astrocytic production of
interleukin-6. J Neurosci 22, 486-492.

F Villemain, T Owens, T Renno and A Beaudet, 1991. Localization of mRNA for
interleukin-2 (IL-2) in mouse brain by in situ hybridization (Abstract). Society of
Neuroscience Abstracts 17, 1199.

M Virgili, A Contestabile and O Bamabei, 1991. Postnatal maturation of cholinergic
markers in forebrain regions of C57BL/6 mice. Brain Res Dev Brain Res 63, 281285.

PJ Waguespack, WA Banks and AJ Kastin, 1994. Interleukin-2 does not cross the bloodbrain barrier by a saturable transport system. Brain Res Bull 34, 103-109.





90


T Waldmann, 2002. The contrasting roles of IL-2 and IL-15 in the life and death of
lymphocytes: implications for the immunotherapy of rheumatological diseases.
Arthritis Res 4 Suppl 3, S161-167.

TA Waldmann, S Dubois and Y Tagaya, 2001. Contrasting roles of IL-2 and IL-15 in the
life and death of lymphocytes: implications for immunotherapy. Immunity 14, 105110.

LG Walker, MB Walker, SD Heys, J Lolley, K Wesnes and O Eremin, 1997. The
psychological and psychiatric effects of rIL-2 therapy: a controlled clinical trial.
Psychooncology 6, 290-301.

TJ Walsh, DW Schulz, HA Tilson and DE Schmechel, 1986. Colchicine-induced granule
cell loss in rat hippocampus: selective behavioral and histological alterations. Brain
Res 398, 23-36.

NL Ward and T Hagg, 2000. BDNF is needed for postnatal maturation of basal forebrain
and neostriatum cholinergic neurons in vivo. Exp Neurol 162, 297-310.

PC Wilkinson and FY Liew, 1995. Chemoattraction of human blood T lymphocytes by
interleukin-15. J Exp Med 181, 1255-1259.

RW Williams and P Rakic, 1988. Three-dimensional counting: an accurate and direct
method to estimate numbers of cells in sectioned material. J Comp Neurol 278,
344-352.

CJ Wilson, CE Finch and HJ Cohen, 2002. Cytokines and cognition--the case for a headto-toe inflammatory paradigm. J Am Geriatr Soc 50, 2041-2056.

O Witzke, T Winterhagen, A Kribben, T Philipp, K Mann and W Reinhardt, 2003.
Interleukin-2 given to asymptomatic HIV-infected individuals leads to an
exaggerated response of the pituitary gland to the action of CRH. Clin Endocrinol
(Oxf) 59, 104-109.

G Wong, Y Goldshmit and AM Turnley, 2004. Interferon-gamma but not TNF alpha
promotes neuronal differentiation and neurite outgrowth of murine adult neural
stem cells. Exp Neurol 187, 171-177.

NJ Woolf, F Eckenstein and LL Butcher, 1984. Cholinergic systems in the rat brain: I.
projections to the limbic telencephalon. Brain Res Bull 13, 751-784.

CC Wrenn, DA Lappi and RG Wiley, 1999. Threshold relationship between lesion extent
of the cholinergic basal forebrain in the rat and working memory impairment in the
radial maze. Brain Res 847, 284-298.

K Yoshida and H Oka, 1995. Topographical projections from the medial septum-diagonal
band complex to the hippocampus: a retrograde tracing study with multiple
fluorescent dyes in rats. Neurosci Res 21, 199-209.





91


JQ Zheng, M Felder, JA Connor and MM Poo, 1994. Turning of nerve growth cones
induced by neurotransmitters. Nature 368, 140-144.














BIOGRAPHICAL SKETCH

Ray Dennis Beck, Jr. was born in York, PA, to Daisy and the appropriately named Ray Dennis Beck, Sr. Realizing the error of his ways early in life, Ray moved to Houston, TX, at the age of two, lending credence to the old Texas saying, "I wasn't born in Texas, but I got here as fast as I could." He attended Oak Ridge High School until the age of 16. Ranked 14 overall in his class, he elected to forego his senior year to attend Simon's Rock College of Bard in Great Barrington, MA. Whereas some 16-year-old adolescents are mature enough to pursue a college education while resisting distractions like geographically convenient buildings filled with members of the opposite sex, a ready supply of fermented beverages, and complete lack of parental supervision Ray was not. He returned to Houston after his first year of college with a less than stellar GPA intent on taking a year off from school. During this "year" (comprised of -1,825 days), he held various jobs ranging from perfume salesman to waiter at various restaurants. During one of the latter jobs, he met his present wife, Laura Frakey. Inspired by her enthusiasm for education, Ray enrolled in the University of Houston, while maintaining full-time employment to pay for school. He graduated cum laude with a B.S. in biology. Ray and Laura moved to Gainesville, FL, to attend University of Florida graduate programs in neuroscience and psychology, respectively. With the completion of his Ph.D., Ray is proud to be one of the most educated high school dropouts that anyone is ever likely to meet. Ray is an avid follower of movies and his other interests can be classified as "all things geeky" (e.g., computers, roleplaying, video games, etc.).

92




Full Text
82
M Koch, 1996a. The septohippocampal system is involved in prepulse inhibition of the
acoustic startle response in rats. Behav Neurosci 110, 468-477.
M Koch, 1996b. The septohippocampal system is involved in prepulse inhibition of the
acoustic startle response in rats. Behav Neurosci 110, 468-477.
S Lacosta, Z Merali and H Anisman, 1999. Influence of acute and repeated interleukin-2
administration on spatial learning, locomotor activity, exploratory behaviors, and
anxiety. Behav Neurosci 113, 1030-1041.
H Lai and MJ Forster, 1988. Autoimmunity and age-associated cognitive decline.
Neurobiol Aging 9, 733-742.
MR Lamprea, FP Cardenas, R Silveira, S Morato and TJ Walsh, 2000. Dissociation of
memory and anxiety in a repeated elevated plus maze paradigm: forebrain
cholinergic mechanisms. Behav Brain Res 117, 97-105.
PA Lapchak and DM Araujo, 1993. Interleukin-2 regulates monoamine and opioid
peptide release from the hypothalamus. Neuroreport 4, 303-306.
PA Lapchak, DM Araujo, R Quirion and A Beaudet, 1991. Immunoautoradiographic
localization of interleukin 2-like immunoreactivity and interleukin 2 receptors (Tac
antigen-like immunoreactivity) in the rat brain. Neuroscience 44, 173-184.
E Larsson, RJ Mandel, RL Klein, N Muzyczka, O Lindvall and Z Kokaia, 2002.
Suppression of insult-induced neurogenesis in adult rat brain by brain-derived
neurotrophic factor. Exp Neurol 177, 1-8.
JM Lauder and UB Schambra, 1999. Morphogenetic roles of acetylcholine. Environ
Health Perspect 107 Suppl 1, 65-69.
BR Lauwerys, N Garot, JC Renauld and FA Houssiau, 2000. Cytokine production and
killer activity of NK/T-NK cells derived with IL-2, IL-15, or the combination of
IL-12 and IL-18. J Immunol 165, 1847-1853.
G Leanza, OG Nilsson and A Bjorklund, 1993. Compensatory changes of in vivo
acetylcholine and noradrenaline release in the hippocampus after partial
deafferentation, as monitored by microdialysis. Brain Res 615, 147-159.
G Leanza, OG Nilsson, RG Wiley and A Bjorklund, 1995. Selective lesioning of the
basal forebrain cholinergic system by intraventricular 192 IgG-saporin:
behavioural, biochemical and stereological studies in the rat. Eur J Neurosci 7, 329-
343.
J Lee, W Duan and MP Mattson, 2002. Evidence that brain-derived neurotrophic factor is
required for basal neurogenesis and mediates, in part, the enhancement of
neurogenesis by dietary restriction in the hippocampus of adult mice. J Neurochem
82, 1367-1375.


34
washed and incubated overnight in the secondary antibody, biotinylated goat anti-rabbit
IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-100 and 1% NGS). The
sections were then washed and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS)
for 2 hrs. The sections were developed in 0.5 mg/ml 3,3-diaminobenzidine (DAB). 0.2
mg/ml urea H2O2 for approximately 5 min and were placed on slides, dehydrated in
graded ethanol washes, cleared in two changes of xylenes, and coverslipped.
Cresyl Violet Staining
For stereological cell count of Nissl-stained granule cells, seven animals per
group were used. The tissue for this assessment was selected, because it originated from
animals utilized in a previous study of cholinergic differences in IL-2 KO mice (Beck et
al., 2002). Every third section was Nissl stained to provide a qualitative view of the
boundaries between various forebrain regions, as well as labeling of hippocampal granule
cells layers (GCL) for stereological counting. The sections were placed on slides and
allowed to air-dry. The slides were immersed in 25C cresyl violet for 10 min, washed
vigorously in rapid exchanges of distilled water to remove the excess cresyl violet,
dehydrated in graded ethanol, cleared in xylenes, and coverslipped.
Stereology
Stained neuronal somata of the MS/vDB or IP and SP-GCL were counted using
the software MCID 5.1 as previously described (Beck et al., 2002). All stereology was
performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2
microscope with a motorized x-y stage made by Imaging Research, Inc. Every third
section through the anterior-posterior extent of the MS/vDB or IP and SP-GCL regions
were sampled. The regions to be counted were outlined at lOx (MS/vDB) or 20x (IP and
SP-GCL) magnification and the size of the counting boxes were generated to be


The second study assayed for variations between groups in the second largest
population of neurons in the MS/vDB, the GABAergic neurons. We found no
differences in these neurons in IL-2 KO animals. In 21 -day-old IL-2 KO mice, we
detected no changes in cholinergic neuronal number in the MS/vDB. This inconsistency
with adult cholinergic neurons may be due to a failure in maintenance or might be
secondary to autoimmunity. Neuronal number in the IP-GCL was also decreased,
consistent with the reduction in distance detected in the first study. We also discovered
that IL-2 KO correlates with a hippocampal elevation in nerve growth factor (NGF), but a
reduction in the brain-derived neurotrophic factor (BDNF).
Finally, in the last study, no T cells or evidence of increased activated microglia
was evident in the IL-2 KO mouse hippocampus. We noted significant elevations in
several cytokines (IL-12, IL-15, IP-10, MCP-1) in the hippocampus of IL-2 KO mice.
The cytokine profile of the serum was different from the hippocampus, indicating that
these were not global changes throughout the bodies of the animals. We also found an
alteration in hippocampal neurogenesis that appeared to be attributable to differences in
male mice.
The results of these studies suggest a neuroimmune interaction that may be
important in septohippocampal physiology.
viii


37
and centrifuged at 16,000 g for 15 min at 4 C. The supernatant was collected and stored
at -20 C for ELISA analysis.
Levels of NGF and BDNF were analyzed in the homogenates from MS/vDB and
hippocampus using a commercially available Emax Immunoassay System according to the
manufacturers instructions (Promega). Briefly, the 96-well plates were coated with
1:6,250 anti-NGF pAb in carbonate coating buffer (0.025 M sodium bicarbonate, 0.025
M sodium carbonate, pH 9.7) and incubated overnight at 4 C. The plates were washed
with TBST wash buffer (20 mM Tris-HCL pH 7.6, 150 mM NaCl, 0.05% (v/v) Tween
20) and blocked with lx Block and Sample buffer (provided with kit) for 1 hour. The
plates were washed again with TBST and a set of standard curves were generated in
duplicate by performing 1:2 dilutions of a known 500 pg/ml standard in a range from 500
pg/ml to 7.6 pg/ml followed by a blank well of 0 pg/ml. All added samples and
standards were allowed to incubate at 25 C for 6 hours. The plates were washed
thoroughly with TBST and 1:4000 anti-NGF mAb was added and incubated overnight at
4 C. The plates were again washed with TBST and 1:100 anti-rat IgG pAb conjugated
to HRP was added for 2.5 hours at room temperature. The plates were washed and TMB
One Solution was added for color development for 10 min. The reaction was stopped
with the addition of equal volume of 1 N HC1 and the absorbance was read at 450 nm
within 30 min of the color development reaction. The data were reported as pg of protein
per mg wet weight tissue.
Statistical Analysis
Results are reported as the mean SEM. Statistical differences between groups
were determined using analysis of variance (ANOVA).


84
L Meda, S Bemasconi, C Bonaiuto, S Sozzani, D Zhou, L Otvos, Jr., A Mantovani, F
Rossi and MA Cassatella, 1996. Beta-amyloid (25-35) peptide and IFN-gamma
synergistically induce the production of the chemotactic cytokine MCP-l/JE in
monocytes and microglial cells. J Immunol 157, 1213-1218.
MA Meijssen, SL Brandwein, HC Reinecker, AK Bhan and DK Podolsky, 1998.
Alteration of gene expression by intestinal epithelial cells precedes colitis in
interleukin-2-defcient mice. Am J Physiol 274, G472-479.
F Mennicken and R Quirion, 1997. Interleukin-2 increases choline acetyltransferase
activity in septal- cell cultures. Synapse 26, 175-183.
RT Mitsuyasu, 2001. The potential role of interleukin-2 in HIV. Aids 15 Suppl 2, S22-27.
H Mizuma, K Takagi, K Miyake, N Takagi, K Ishida, S Takeo, A Nitta, H Nomoto, Y
Furukawa and S Furukawa, 1999. Microsphere embolism-induced elevation of
nerve growth factor level and appearance of nerve growth factor immunoreactivity
in activated T-lymphocytes in the rat brain. J Neurosci Res 55, 749-761.
G Moalem, A Gdalyahu, Y Shani, U Otten, P Lazarovici, IR Cohen and M Schwartz,
2000. Production of neurotrophins by activated T cells: implications for
neuroprotective autoimmunity. J Autoimmun 15, 331-345.
WC Mobley, JL Rutkowski, GI Tennekoon, J Gemski, K Buchanan and MV Johnston,
1986. Nerve growth factor increases choline acetyltransferase activity in
developing basal forebrain neurons. Brain Res 387, 53-62.
ML Monje, H Toda and TD Palmer, 2003. Inflammatory blockade restores adult
hippocampal neurogenesis. Science 302, 1760-1765.
DA Morgan, FW Ruscetti and R Gallo, 1976. Selective in vitro growth of T lymphocytes
from normal human bone marrows. Science 193, 1007-1008.
JK Morse, SJ Wiegand, K Anderson, Y You, N Cai, J Carnahan, J Miller, PS DiStefano,
CA Altar, RM Lindsay and et al., 1993. Brain-derived neurotrophic factor (BDNF)
prevents the degeneration of medial septal cholinergic neurons following fimbria
transection. J Neurosci 13, 4146-4156.
N Muller and M Ackenheil, 1998. Psychoneuroimmunology and the cytokine action in
the CNS: implications for psychiatric disorders. Prog Neuropsychopharmacol Biol
Psychiatry 22, 1-33.
Y Nakamura, SM Russell, SA Mess, M Friedmann, M Erdos, C Francois, Y Jacques, S
Adelstein and WJ Leonard, 1994. Heterodimerization of the IL-2 receptor beta- and
gamma-chain cytoplasmic domains is required for signalling. Nature 369, 330-333.


62
development and inflammation. Thus, the increased levels of IL-12 found in the
hippocampus could be the secondary to the loss IL-2.
Increased levels of inflammatory cytokines such as IL-6 may impair hippocampal
neurogenesis (Monje et al., 2003; Vallieres et al., 2002); thus, we hypothesized that the
alterations in neuroimmunological status would correlate with decreased neurogenesis in
the DG of IL-2 KO mice. Against our hypothesis, however, no detectable levels pro-
inflammatory cytokines like IL-6 (though IL-12 can mediate inflammatory responses) or
effects of group were apparent in hippocampal neurogenesis. However, a significant
group by gender interaction was detectable. Interestingly, this effect appeared to be due
to variations in the male mice, but not the females. We previously noted a trend which
suggested that IL-2 gene deletion may protect from experimental autoimmune
encephalomyelitis (EAE), though we did not statistically analyze the data (Petitto et al.,
2000). In that study, three out of the seven male mice utilized developed some symptoms
of the disease, whereas none of the six females did. One hypothesis to explain this group
by gender interaction is that there may be an interplay between IL-2 and the sex
hormones. Sex steroids like estrogen, for example, have been linked to neurogenesis
(Gould et al., 2000; Perez-Martin et al., 2003). Further, evidence that IL-2 can regulate
sex hormone expression comes from studies of Leydig cell cultures where IL-2 was
shown to inhibit steroidogenesis (Guo et al., 1990).
Hormonal differences influenced by IL-2 may give a possible explanation of why
males differ from females, but it does not suggest a mechanism of why IL-2 KO mice
appear to be protected from impaired hippocampal neurogenesis. As noted above,
absence of IL-2 can enhance neuroregenerative properties in the axotomized facial motor


6
IL-2 and the Septohippocampal System
Hippocampal circuitry is important for encoding spatial learning and memory and
some evidence supports a potential role of IL-2 in the hippocampus. IL-2, for example,
alters the electrophysiological characteristics of hippocampal neurons including
alterations of voltage-dependent Ca:+ currents (Plata-Salaman and ffrench-Mullen, 1993),
depolarization and hyperpolarization of cultured hippocampal neurons (Hanisch and
Quirion, 1995a), and changes in long-term potentiation (LTP) (Tancredi et ah, 1990). IL-
2R subunits, as previously mentioned, are enriched in the hippocampus relative to other
brain regions and exogenously applied IL-2 enhances the survival and morphological
development of neurons of the hippocampus.
Knockout mice deficient in IL-2 perform significantly worse than wild-type
controls in one such test of spatial learning and memory, the Morris water maze; show an
enhanced pre-pulse inhibition of the acoustic startle response (PPI; another
hippocampally-mediated process); and also exhibit structural alterations in mossy fiber
length (Petitto et ah, 1999). Our initial studies suggest that this deficit in learning and
memory is not likely due to a compromised immune system, as severe combined
immunodeficient (SCID) mice perform significantly better than IL-2 KO mice in the
Morris water maze. More recent studies from our lab suggest that the nature of the
deficit in learning and memory seen in IL-2 knockout mice could be related to the
immune status of the mother (normal heterozygote vs. autoimmune homozygote mother).
In addition, the previously mentioned clinical studies of cancer patients under IL-2
treatment found alterations in spatial memory, lending some support to the potential role
of IL-2 in learning and memory.


68
the hippocampus was likely due to compensatory sprouting, but did not overtly assay for
any alterations. Thus, to address this issue further, a study utilizing a stereotaxic
injection of an anterograde tracer, such as Phaseolus vulgaris-leucoagglutinin (PHAL),
into the MS/vDB area would be necessary. The septohippocampal cholinergic neurites
would need to be labeled with a marker for AChE and the length and branching of these
double-labeled axons would be characterized. If the septohippocampal cholinergic
neurons do undergo compensatory sprouting in IL-2 KO mice, we hypothesize that a
detectable increase in branching should occur in those animals vs. wild-type littermates.
Though the majority of cholinergic neurons in the MS/vDB project to the
hippocampus (Schwegler et al., 1996b), a small percentage also project to other areas,
including the mediodorsal nucleus of the thalamus (Gritti et al., 1998), parietal, cingulate
(Gritti et al., 1997), and entorhinal cortices (Alonso and Kohler, 1984). Thus, to
accurately claim that the cholinergic neuronal losses observed in the MS/vDB of the IL-2
KO mouse do indeed project to the hippocampus, a study involving a retrograde tracer
like Fluoro-Gold injected in the hippocampal projection areas would be necessary. The
retrogradely labeled somata of the MS/vDB would need to be double-labeled with ChAT
to identify neurons of the cholinergic phenotype and the cell number would be estimated
with stereology.
In Study 2, three-week-old IL-2 KO animals did not differ from wild-type
littermates in cholinergic cell number. To expand upon this finding, the brains from
animals should be examined at one-week intervals from three-week-old mice to adult
eight-week-old mice. This will allow us to establish a time course of cholinergic
neuronal loss. Furthermore, we would be interested in investigating the nature of the cell


71
normal levels (or lack thereof in IL-2 KO/RAG-1 KO mice) of endogenous brain IL-2
expression, but remain immunodeficient, thus lacking a peripheral source of IL-2. This
experimental design would create a model
that eliminates the impact of peripheral autoimmunity in the IL-2 wild-type
reconstituted IL-2 KO/RAG-1 KO mice;
in which the aforementioned animals are only lacking a CNS, but not peripheral,
source of IL-2;
in which the IL-2 wild-type reconstituted IL-2 wild-type/RAG-1 KO mice would
have a functional peripheral immune system and normal endogenous IL-2
expression in the brain;
in which the IL-2 KO reconstituted IL-2 KO/RAG-1 KO mice are similar in
phenotype to the normal IL-2 KO mouse (i.e., peripheral autoimmunity);
in which the IL-2 KO reconstituted IL-2 wild-type/RAG-1 KO mice would have
normal brain IL-2 expression, but autoimmunity caused by the peripheral IL-2 KO
T lymphocytes; and
in which the sham reconstituted IL-2 KO/RAG-1 KO mice lack any source of IL-2
and the sham reconstituted wild-type/RAG-1 KO mice lack a peripheral source of
IL-2, with neither of the two models succumbing to autoimmunity.
Thus, if endogenous brain IL-2 is important in septohippocampal development and/or
maintenance from the age of three-weeks to adulthood, then we should observe deficits in
the septohippocampal system similar to those noted in this dissertation in all
reconstitution and sham models above on an IL-2 KO/RAG-1 KO background. If the
peripheral source of IL-2 is more important for septohippocampal physiology, then the
structural and physiological alterations in this dissertation should be rescued in the above
IL-2 wild-type reconstituted cases, but not IL-2 KO or sham reconstituted animals. If
autoimmunity is the determining factor, then IL-2 KO reconstituted, but not sham
reconstituted, animals will all exhibit deficits in the septohippocampal system regardless
of background. This proposed experiment combining IL-2 KO strains, immunodeficient


61
facial motor nucleus (Petitto et al., 2003). It is interesting to speculate that elevated IL-15
levels could contribute to the increased motor neuronal survival in those mice. Therefore
an alternative interpretation is that increased IL-15 in the hippocampus of IL-2 KO mice
could be a compensatory response to counteract neuroregenerative changes in the
hippocampus.
The increased levels of MCP-1 and IP-10 may possibly be induced by increased
IL-15 in the hippocampus (Badolato et al., 1997). Previous studies have demonstrated
that IP-10 expression can be detected in lipopolysaccharide (LPS)-treated microglial and
astroglial cultures and in situ hybridization of LPS-treated rat brains (Ren et al., 1998).
Similarly, MCP-1 can also be induced by the addition of the pro-inflammatory cytokine,
TNF-a, or the anti-inflammatory cytokine, TGF-P, in astrocytes (Hurwitz et al., 1995)
and microglia (Meda et al., 1996). IP-10 and MCP-1 are also chemoattractant factors for
T lymphocyte infiltration into the CNS (Babcock et al., 2003; Dufour et al., 2002), and
may attract activated microglia in vitro (Cross and Woodroofe, 1999). In spite of the
published data linking IP-10 and MCP-1 to T cell and microglial chemotaxis, we were
unable to detect either T cells or increased numbers of activated microglia in the
hippocampus of IL-2 KO mice. Further studies are necessary to determine the functional
significance of the increased production of IP-10 and/or MCP-1 in the hippocampus of
IL-2 KO mice. Finally, in addition to its immune activating effects as a Thl cytokine,
IL-2 is also known to have important critical negative regulatory functions by stimulating
Th2 lymphocytes to produce TGFp (Ludviksson et al., 1997), which down-regulates the
ability of antigen presenting cells to produce IL-12, a powerful activator of Thl cell


86
JM Petitto and Z Huang, 1995. Molecular cloning of the coding sequence of an
interleukin-2 receptor alpha subunit cDNA in murine brain. J Neuroimmunol 59,
135-141.
JM Petitto and Z Huang, 2001. Cloning the full-length IL-2/15 receptor-beta cDNA
sequence from mouse brain: evidence of enrichment in hippocampal formation
neurons. Regul Pept 98, 77-87.
JM Petitto, Z Huang, J Lo and WJ Streit, 2003. IL-2 gene knockout affects T lymphocyte
trafficking and the microglial response to regenerating facial motor neurons. J
Neuroimmunol 134, 95-103.
JM Petitto, DB McCarthy, CM Rinker, Z Huang and T Getty, 1997. Modulation of
behavioral and neurochemical measures of forebrain dopamine function in mice by
species-specific interleukin-2. J Neuroimmunol 73, 183-190.
JM Petitto, Z Huang, MK Raizada, CM Rinker and DB McCarthy, 1998. Molecular
cloning of the cDNA coding sequence of IL-2 receptor-gamma (gammac) from
human and murine forebrain: expression in the hippocampus in situ and by brain
cells in vitro. Brain Res Mol Brain Res 53, 152-162.
JM Petitto, RK McNamara, PL Gendreau, Z Huang and AJ Jackson, 1999. Impaired
learning and memory and altered hippocampal neurodevelopment resulting from
interleukin-2 gene deletion. J Neurosci Res 56, 441-446.
JM Petitto, WJ Streit, Z Huang, E Butfiloski and J Schiffenbauer, 2000. Interleukin-2
gene deletion produces a robust reduction in susceptibility to experimental
autoimmune encephalomyelitis in C57BL/6 mice. Neurosci Lett 285, 66-70.
CR Plata-Salaman and JM ffrench-Mullen, 1993. Interleukin-2 modulates calcium
currents in dissociated hippocampal CA1 neurons. Neuroreport 4, 579-581.
LQ Ren, N Gourmala, HW Boddeke and PJ Gebicke-Haerter, 1998. Lipopolysaccharide-
induced expression of IP-10 mRNA in rat brain and in cultured rat astrocytes and
microglia. Brain Res Mol Brain Res 59, 256-263.
GE Ringheim, BD Freimark and RJ Robb, 1991. Quantitative characterization of the
intrinsic ligand-binding affinity of the interleukin 2 receptor beta chain and its
modulation by the alpha chain and a second affinity-modulating element.
Lymphokine Cytokine Res 10, 219-224.
SA Rosenberg, JJ Mule, PJ Spiess, CM Reichert and SL Schwarz, 1985. Regression of
established pulmonary metastases and subcutaneous tumor mediated by the
systemic administration of high-dose recombinant interleukin 2. J Exp Med 161,
1169-1188.


CHAPTER 2
ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS
RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT
Introduction
One of the earliest observations suggesting that cytokines could influence brain
function in humans came from cancer treatment trials in which interleukin-2 (IL-2) was
found to induce cognitive dysfunction and other untoward neuropsychiatric side effects in
patients (Denicoff et al., 1987). Although basic research has demonstrated that IL-2 can
modulate different aspects of central nervous system (CNS) function, some of IL-2s
most prominent neurobiological actions occur in the hippocampal formation and related
limbic regions, where receptors for this cytokine are enriched (Araujo et al., 1989;
Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto
et al., 1998).
Exogenously administered IL-2 has effects on a number of parameters of septal
and hippocampal neuronal function including trophic effects on cultured fetal septal and
hippocampal neurons (Awatsuji et al., 1993; Sarder et al., 1996; Sarder et al., 1993). IL-2
may also modify cellular and molecular substrates of learning and memory such as long
term potentiation (Tancredi et al., 1990), and multiple parameters of cognitive behavioral
performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al.,
1999; Nemni et al., 1992). Moreover, the neurotrophic and neuromodulatory actions of
IL-2 have been implicated in abnormal hippocampal development associated with
schizophrenia (Ganguli et al., 1995; Licinio et al., 1993; McAllister et al., 1995).
11


30
acetylcholine (ACh) release from rat hippocampal slices (Hanisch et al., 1993; Setoet al.,
1997), and can also increase the activity of its precursor enzyme, choline
acetyltransferase (ChAT) (Mennicken and Quirion, 1997).
Previously, we found that IL-2 knockout mice (IL-2 KO) exhibited impaired
learning and memory performance, sensorimotor gating, and reductions in hippocampal
infrapyramidal mossy neuronal fiber length (Petitto et al., 1999), a factor which correlates
positively with spatial learning ability (Schopke et al., 1991; Schwegler and Crusio,
1995; Schwegler et al., 1988). We also found in studies of IL-2 KO mice in vivo, there
was a marked reduction in cholinergic somata in medial septal/vertical limb of the
diagonal band of Broca (MS/vDB) region, as well as decrease in the distance across the
infrapyramidal granule cell layer (IP-GCL) of the dentate gyrus (DG) (Beck et al., 2002).
Variation in the cytoarchitecture of cholinergic septohippocampal neurons correlate with
differences in spatial learning ability in mice (Schwegler et al., 1996a; Schwegler et al.,
1996b).
Research has shown that the neurotrophins, nerve growth factor (NGF) and brain-
derived neurotrophic factor (BDNF) expressed in the hippocampus, can be important in
the development, maintenance, and repair of septohippocampal neurons in vitro
(Arimatsu and Miyamoto, 1991; Conner and Varn, 1997; Gahwiler et al., 1987;
Hartikka and Hefti, 1988; Morse et al., 1993). Similar trophic effects have been noted in
studies utilizing infusion of exogenous neurotrophins in vivo (Hagg et al., 1990; Morse et
al., 1993), and in studies of transgenic and knockout mice (Ruberti et al., 2000; Ward and
Hagg, 2000). In the peripheral immune system of multiple animal species, both NGF and
BDNF are expressed by T lymphocytes (Braun et al., 1999; Kerschensteiner et al., 1 999;



PAGE 1

ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE IN INTERLEUKIN-2 KNOCKOUT MICE By RAY D. BECK, JR. A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

This dissertation is dedicated to my wife Laura whose love, support, and occasional nagging help to keep me focused on my goals.

PAGE 3

ACKNOWLEDGMENTS Traditionally, most acknowledgement sections begin by thanking one's advisor; in this case, such tradition is most warranted. As such, I thank Dr. John Petitto for being a wonderful mentor. Though not the stealthiest individual in the world with his penchant for crying out one's name (or impromptu nickname) exuberantly as soon as one enters anywhere within his field of vision, he is among the kindest, most supportive, enthusiastic, and knowledgeable mentors for which any graduate student could wish. Next, I would like to thank each member of my committee. I feel fortunate to be advised by such a great selection of knowledgeable and friendly people. Dr. Mike King's easygoing personality and extensive knowledge of all things stereological and cholinergic have proven invaluable in my studies. Dr. Mark Lewis' advice on statistics and experimental design, as well as his sense of humor, has been most appreciated. Dr. Jake Streit, in addition to his ability to make me laugh, always reminded me that there is more than one kind of cell in the brain. Dr. Mark Atkinson, always amiable and approachable, helped guide me in the "immunology" aspect of "neuroimmunology." I also thank Dr. Huang Zhi. I cannot overstate how much I valued his advice on experiments and his daily conversations on topics ranging from basketball to politics to Hong Kong movies. I wish him the best of luck in his medical residency and his future as a psychiatrist. I would also like to thank Clive Wasserfall and Fletcher Schwartz for teaching me how to use the Luminex technology and Tim Vaught for teaching me the ins and outs of multiple microscopy techniques. 1 also thank the many technicians that iii

PAGE 4

worked in the lab both past and present: Brent, Andrew, David, Jeannette, Dan, Jesse, and Grace. In addition to taking care of the upkeep of the lab, working with them was a pleasure. In particular, I would like to single out Andrew. He was my dearest friend here in Gainesville. I wish him the best of luck in his career as a medical doctor and hope that we will always remain friends after I leave Gainesville. Outside of the laboratory, I thank my other friends for being my support structure. There has never been a better collection of in-the-closet geeks than Coleman, Dan, Andy, Ryan, Charles, Chris, Curtis, Nick, and Jason. They are simply the best. I would also like to thank Mozart, Michelangelo, and Sage simply because not nearly enough people thank their dogs. Obviously, they would be more likely to chew on this dissertation than read it, but I would like anyone else that does see this to know that few people are capable of matching the unconditional love that a dog has for its owner. I also thank my family. My mother has always instilled the value of education into me. My father's love and sense of humor were crucial in the development of my personality. My sisters, Jackie and Judy, will always be among my closest friends. My Aunt Kathy and Uncle Jimmy are fremendous people who have always supported me and though they are family by marriage, our bond is stronger than blood. Finally, I thank my lovely wife Laura. Without her, my life would be incomplete (though despite her beliefs, I could still drive effectively without her commentary from the passenger seat). She has pushed me when I needed motivation and comforted me when I need support. She is my heart and soul and I could not have achieved this dissertation without her. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS hi ABSTRACT vii CHAPTER 1 BACKGROUND AND SIGNIFICANCE 1 Cytokine-Brain Interactions 1 The Pleiotropic Cytokine: Interleukin-2 2 Interleukin-2 and the Brain 4 IL-2 and the Septohippocampal System 6 Statement of the Problem 9 2 ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT 11 Introduction 1 1 Materials and Methods 13 Animals and Tissue Preparation 13 Genotyping Using PCR 14 ChAT Immunohistochemistry 14 AChE Histochemistry 16 Cholinergic Stereology 16 Quantitative Image Analysis of AChE Staining 18 Cresyl Violet Staining 20 Results 20 Comparison of Cholinergic Somata in the MS/vDB 20 Density of AChE-positive Fibers in Regions of the Hippocampus 20 Morphology of the Granular Cell Layer of the Lower Limb of the DG 21 Discussion 23 3 ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN INTERLEUKIN-2 KNOCKOUT MICE 29 Introduction 29 Methods 32 V

PAGE 6

Animals and Genotyping 32 Immunohistochemistry 33 Cresyl Violet Staining 34 Stereology 34 Enzyme-linked Immunosorbent Assay (ELISA) Characterization of NGF and BDNF 36 Statistical Analysis 37 Results 38 Cholinergic MS/vDB Cell Number in 21 -day-old Mice and GABAergic Cell Number in Adult Mice 38 Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice 38 Alterations in Neurotrophin Levels 38 Discussion 40 4 INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND NEUROGENESIS IN THE HIPPOCAMPUS 45 Introduction 45 Materials and Methods 49 Animals and Genotyping 49 CD3^ T cells and MHC iV Microglia Immunohistochemistry 50 Preparation of Serum and Brain Tissue for Cytokine Analysis 51 Multiplex Microsphere Cytokine Analysis 52 Labeling Neurogenesis with BrdU 52 Results 54 Assessment of CD3+ T Cells and MHC 11+ Activated Microglia in the Hippocampus 54 Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice 54 Comparison of Serum C>1okine Levels in lL-2 Knockout vs. Wild-type Mice .55 Alterations in Neurogenesis 56 Discussion 53 5 GENERAL DISCUSSION 64 Summary of the Overall Findings 64 Implications 65 Caveats and Future Directions 67 Concluding Remarks 72 WORKS CITED 73 BIOGRAPHICAL SKETCH 92 vi

PAGE 7

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ALTERATIONS OF SEPTOHPPOCAMPAL STRUCTURE IN INTERLEUKIN-2 KNOCKOUT MICE By Ray D. Beck, Jr. August 2004 Chair: John M. Petitto Major Department: Neuroscience Interleukin-2 (IL-2) is a multifunctional cytokine involved in peripheral immune processes and may also be implicated in multiple brain functions. IL-2 gene knockout (IL-2 KO) mice exhibit deficits in several hippocampally-mediated behaviors (e.g., learning and memory) and have alterations in hippocampal structure. In the first study, adult IL-2 KO and wild-t>pe littermates were compared for differences in the cholinergic neurons in the medial septum and vertical limb of the diagonal band of Broca (MS/vDB; a structure associated with learning and memory). The IL-2 KO mice had significantly fewer cholinergic somata in the MS/vDB, but not in the striatum, thus indicating a selective effect of IL-2 on the MS/vDB. Cholinergic neurite density in the hippocampus was unaffected, but the length across the infi-apyramidal (IP), but not the suprapyramidal (SP), granule cell layer (GCL) of the dentate gyrus was reduced. vii

PAGE 8

The second study assayed for variations between groups in the second largest population of neurons in the MS/vDB, the GABAergic neurons. We found no differences in these neurons in IL-2 KO animals, hi 2 1 -day-old IL-2 KO mice, we detected no changes in cholinergic neuronal number in the MS/vDB. This inconsistency with adult cholinergic neurons may be due to a failure in maintenance or might be secondary to autoimmunity. Neuronal number in the IP-GCL was also decreased, consistent with the reduction in distance detected in the first study. We also discovered that IL-2 KO correlates with a hippocampal elevation in nerve growth factor (NGF), but a reduction in the brain-derived neurotrophic factor (BDNF). Finally, in the last study, no T cells or evidence of increased activated microglia was evident in the IL-2 KO mouse hippocampus. We noted significant elevations in several cytokines (IL-12, IL-15, IP10, MCP-1) in the hippocampus of IL-2 KO mice. The cytokine profile of the serum was different from the hippocampus, indicating that these were not global changes throughout the bodies of the animals. We also found an alteration in hippocampal neurogenesis that appeared to be attributable to differences in male mice. The results of these studies suggest a neuroimmune interaction that may be important in septohippocampal physiology. viii

PAGE 9

CHAPTER 1 BACKGROUND AND SIGNIFICANCE Cytokine-Brain Interactions The landmark studies of Ader and Cohen demonstrating that immune physiology could be behaviorally conditioned led to the systematic investigation of the complex interaction between the central nervous and immune systems (Ader and Cohen, 1975; Ader et al., 1982). Although the central nervous system (CNS) and peripheral immune system were once considered functionally incompatible entities separated by a nearly impermeable protective blood-brain-barrier (BBB), it is now known that there is bidirectional communication and modulation between these two systems. Cytokines have emerged as important mediators of various processes in the CNS. Their effects range from neuroinflammation in experimental autoimmune encephalomyelitis (EAE) and viral infection of the brain to neurobiological processes such as hypothalamicpituitary axis (HP A) regulation, induction of fever, sleep, analgesia, feeding behavior, and cognition (for reviews see Ader et al., 2001 ; Dunn, 2002; Wilson et al., 2002). Cytokines produced both within and outside of the CNS can exert their effect on brain cells (Dunn, 2002; Streit et al., 1998). The work of Banks and others show that the BBB acts as a selective filter for peripheral cytokines (for a review see Banks et al., 2002). Multiple studies support the ability of cytokines (e.g., IL-la and -p, IL-2, IL-6, IFN-a and -y, TNF-a) to cross the BBB via different transport mechanisms (Banks et al., 1994; Banks et al., 1991; Gutierrez et al., 1993; Pan et al., 1997; Waguespack et al., 1994), and via the "leaky" circumventricular organs (CVO), four brain regions outside of 1

PAGE 10

2 the BBB with fenestrated capillaries (Duller, 2001). Peripheral leukocytes, in particular activated T cells that enter the brain during certain conditions (e.g., EAE, facial nerve axotomy), can also release cytokines in the CNS (Hickey et al., 1991). Finally, cytokines may also interact with the brain through activation of peripheral nerves, such as IL-1 stimulation of the vagus nerve, which can lead to modulation of brain functions through its afferent connections in the CNS (Maier et al., 1998). Such a cytokine-to-nerve communication pathway may not be limited to the vagus nerve, as central hj'peralgesic effects are also observed by stimulating cutaneous nerves with a subcutaneous injection of IL-1 p (Fukuoka et al, 1994), TNF-a (Sorkin et al., 1997), or antibodies against TNFa (Lindenlaub et al., 2000). Thus, multiple pathways exist that allow cytokines to directly or indirectly influence the brain. The focus of this dissertation was on IL-2, which can be produced in the periphery and the CNS. The Pleiotropic Cytokine: InterIeukin-2 IL-2 was originally identified as a growth factor for bone marrow-derived T cells in 1976 (Morgan et al., 1976), and was renamed in 1979, when its pleiotropic effects between leukocytes (thus the term interleukin) became clear (Aarden et al., 1979). Further characterization of IL-2 revealed that it belongs to the four a-helix bundle family of cytokines; this family consists of cytokines with four a-helices connected by three loops in an up-up-down-down formation (Bazan, 1992). The receptor for IL-2 has a common gamma (Yc) subunit shared by multiple cytokines including IL-4, IL-7, IL-9, and IL-15 (Sugamura et al., 1996); a p subunit only shared with IL-15 (Giri et al., 1995); and, in one conformation, an a subunit, which confers greater binding affinity (Leonard et al., 1984). The receptor subunits can combine in two biologically active forms: a lower

PAGE 11

3 affinity heterodimer consisting of the yc and (3 subunits and a high affinity heterotrimer comprised of all three subunits (a, p, and Yc) (Ringheim et al., 1991; Takeshita et al., 1992). The p and Yc both possess intracellular signaling domains and in their heterodimeric form have a of 10"'', whereas the addition of the a subunit forms a heterotrimer with higher affinity (K
PAGE 12

4 Interleukin-2 and the Brain The effect of IL-2 on cognition and mood in humans was among the earliest findings that suggested that this cytokine might have neurobiological actions. In early clinical studies of the cognitive side effects of IL-2 therapy, 50% (i.e., 22 patients out of 44) of the subjects monitored developed cognitive changes, with 15 of them necessitating acute intervention (Denicoff et al., 1987). In addition, IL-2 therapy in patients with renal carcinoma or melanoma was found to impair spatial memory and performance in planning tasks (Capuron et al., 2001a), and induce depressive symptoms as early as two days into therapy (Capuron et al., 2000). IL-2-brain interactions have also been investigated on an anatomical and physiological level. In landmark studies, IL-2 was found to modulate the proliferation of oligodendrocytes (Benveniste et al., 1987; Benveniste and Merrill, 1986; Saneto et al., 1 986). Exogenously administered IL-2 also has multiple effects on pituitary cells including stimulation of Cortisol production and adrenal corticotropin releasing hormone release (Hanisch et al., 1994), as well as increasing pituitary cell responsiveness to corticotropin-releasing hormone (Witzke et al., 2003). IL-2 has also been shown to regulate the production and secretion of peptides from hypothalamus, in addition to pituitary cells (Karanth et al., 1993; Lapchak and Araujo, 1993; Pardy et al., 1993). Subsequent research has shown that exogenously applied IL-2 can modulate other types of central nervous system cells, such as microglia (Sakai et al., 1995). Exogenously applied IL-2 can also biphasically regulate the release of some neurotransmitters such as dopamine (Alonso et al., 1993; Petitto et al., 1997), or acetylcholine (Hanisch et al., 1993; Seto et al., 1997).

PAGE 13

5 IL-2 has been shown to have neurotrophic effects on cultured neurons from several regions of the rat brain including the neocortex (Shimojo et al., 1993), cortex, striatum, medial septum, and hippocampus (Awatsuji et al., 1993). Moreover, in rat hippocampal neuronal cultures, IL-2 enhances the length and branching of hippocampal neurites and the morphology of these neurons (Sarder et al., 1996; Sarder et al., 1993). Interestingly, altered levels of IL-2 expression have been detected in schizophrenia (for reviews see Hanisch and Quirion, 1995a; MuUer and Ackenheil, 1998), which is a neurological disorder where altered morphology of hippocampal neurons is well documented (for a review see Thune and Pakkenberg, 2000). IL-2-like immunoreactivity has been localized to the hippocampal formation in rat forebrain (Lapchak et al., 1991), and detected in tissue extracts from rat and human hippocampal tissue (Araujo et al., 1989). In mouse brain, IL-2 mRNA has been found in the hippocampus (Villemain et al., 1991), and transcripts for this cytokine may be expressed in rat astrocyte cultures as well (Eizenberg et al., 1995). Our lab has cloned and sequenced the full-length mouse brain cDNAs for IL-2Ra as well as the IL2/1 5RP and Yc subunits, and has found that the sequences of the genes expressed by lymphocytes and in brain are identical. We have also found that these genes are enriched in the hippocampus and related limbic regions. Of particular relevance to IL-2 actions in the hippocampus, in situ hybridization has shown that the IL-2/15Rp and yc genes are expressed by pyramidal and granule cell neurons (Petitto and Huang, 1994, 1995, 2001; Petitto et al., 1998).

PAGE 14

6 IL-2 and the Septohippocampal System Hippocampal circuitry is important for encoding spatial learning and memory and some evidence supports a potential role of IL-2 in the hippocampus. IL-2, for example, alters the electrophysiological characteristics of hippocampal neurons including alterations of voltage-dependent Ca^^ currents (Plata-Salaman and ffrench-Mullen, 1993), depolarization and hyperpolarization of cultured hippocampal neurons (Hanisch and Quirion, 1995a), and changes in long-term potentiation (LTP) (Tancredi et al., 1990). IL2R subunits, as previously mentioned, are enriched in the hippocampus relative to other brain regions and exogenously applied IL-2 enhances the survival and morphological development of neurons of the hippocampus. Knockout mice deficient in IL-2 perform significantly worse than wild-type controls in one such test of spatial learning and memory, the Morris water maze; show an enhanced pre-pulse inhibition of the acoustic startle response (PPI; another hippocampally-mediated process); and also exhibit structural alterations in mossy fiber length (Petitto et al., 1999). Our initial studies suggest that this deficit in learning and memory is not likely due to a compromised immune system, as severe combined immunodeficient (SCID) mice perform significantly better than IL-2 KO mice in the Morris water maze. More recent studies fi-om our lab suggest that the nature of the deficit in learning and memory seen in IL-2 knockout mice could be related to the immune status of the mother (normal heterozygote vs. autoimmune homozygote mother). In addition, the previously mentioned clinical studies of cancer patients under IL-2 treatment found alterations in spatial memory, lending some support to the potential role of IL-2 in learning and memory.

PAGE 15

7 In the basal forebrain, the medial septum and vertical limb of the diagonal band of Broca (MS/vDB) send a large number of projections to the hippocampus, with the major neuronal phenotypes of these being cholinergic and GABAergic (Brashear et al., 1986; Kiss et al., 1990b; Kiss et al., 1990a). The septohippocampal system has been associated with learning and memory processes, with extensive data existing that link the septal cholinergic neurons that project to the hippocampus to learning and memory (Galey et al., 1994; Leanza et al., 1995), and PPI (Koch, 1996b). Moreover, variability of cholinergic fiber density in the dentate gyrus of individual mouse strains correlate with changes in spatial learning (Schwegler et al., 1996a; Schwegler et al., 1996b). Some controversy exists, however, on the relative importance of cholinergic neurons of the MS/vDB in learning and memory processes. The advent of selective toxins that target cholinergic neurons, like 192 IgG-saporin, have allowed researchers to behaviorally test animals only lacking MS/vDB cholinergic neurons, but with presumably normal distributions of GABAergic neurons. In many of these studies, animals with cholinergic septohippocampal lesions did not differ from control subjects (Baxter et al., 1996; Bizon et al., 2003; Cahill and Baxter, 2001; Chappell et al., 1998; Perry et al., 2001). Surprisingly, however, multiple other contemporary studies utilizing 192 IgG-saporin do find learning and memory deficits in the lesioned animals (Janis et al., 1998; Johnson et al., 2002; Lamprea et a!., 2000; Wrenn et a!., 1999). One potential explanation for this discrepancy may be that a certain threshold of cholinergic damage is necessary to elicit a deterioration in spatial learning ability (Leanza et al., 1995; Wrenn et al., 1999). Nevertheless, the negative findings are considerable and this may not explain the inconsistency well enough between groups. Another hypothesis calls into quesfion the

PAGE 16

8 efficacy of 192 IgG-saporin in completely removing septohippocampal cholinergic activity. In studies where 192 IgG-saporin appears to nearly completely eliminate cholinergic immunostaining of MS/vDB neurons, ACh release in the hippocampus of these lesioned animals still persists at -40% of the control values (Chang and Gold, 2004; Gold, 2003). The septohippocampal cholinergic neurons are capable of compensatory collateral sprouting (Gage et al., 1983a, 1984; Gage et al., 1983b), and functional recovery of hippocampal ACh release after complete fimbria-fornix transection (Leanza et al., 1993), which may be important as a response to damage. Thus, considering the above evidence, it is clear that cholinergic input to the hippocampus plays an important role in the complex neurobiological processes of learning and memory, though it is certainly not the only system involved. Some evidence supports a potential trophic or regulatory role for IL-2 on the cholinergic septohippocampal system. IL-2 enhances the survival of these septal and hippocampal neurons in culture (Awatsuji et al., 1993), and modulates the activity of choline acetyl transferase (ChAT) (Mennicken and Quirion, 1997). Also, IL-2 is a potent biphasic modulator of acetylcholine (ACh) release (Hanisch et al., 1993; Seto et al., 1997). At low concentrations (sub-pM), IL-2 stimulates the release of ACh, but at higher concentrations (nM), IL-2 inhibits ACh release. Since IL-2 is difficult to detect in the adult brain, endogenous brain IL-2 or IL-2 that crosses the BBB would be expected to be present at a low concentration in the normal brain. Taking into consideration 1 ) the neurotrophic and neuromodulatory role of IL-2 in cultured septal and hippocampal neurons, 2) the regulatory effects of IL-2 on ACh release and ChAT activity in vitro, 3) the spatial learning and memory impairments of IL-2-

PAGE 17

9 deficient knockout mice and cancer patients undergoing IL-2 therapy, and 4) the enhanced concentration of IL-2R subunits in the hippocampus, I hypothesized that IL-2 may play a role in the growth and differentiation of the septohippocampal cholinergic neurons. Statement of the Problem Normal development of septal neurons depends on trophic factors that are presumably secreted from the hippocampus. For example, the neurotrophins, nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) are expressed in the hippocampus and have both been shown to be important in the development, maintenance, and repair of septohippocampal neurons (Brooks et al., 1999; Conner et al, 1992; Conner and Varon, 1997; Morse et al., 1993). Perhaps, IL-2 also mediates the growth and development of septal and hippocampal neurons. This cytokine may act as a growth factor by itself, or signal the release of growth factors from neurons or glia in the hippocampal area. In other investigations, the modulatory and growth-promoting effects of IL-2 were determined by either adding exogenous IL-2 to cultures or by injecting exogenous IL-2 into the brain (in most cases, species non-specific, e.g., human IL-2 in rats or mice). Thus, the studies utilizing exogenous IL-2 administration have several potential shortcomings: 1) the cytokine has a short half-life, and the amount of IL-2 delivered may not reflect physiologically relevant concentrations in the CNS in vivo (e.g., chronic vs. acute dosing), 2) the nature of the injections disrupts the BBB, causing potentially even more IL-2, as well as other cytokines, from the peripheral immune system to enter into the brain, 3) it is unknown during which period in neurodevelopment that IL-2 might exert these postulated effects, and 4) non-species specific IL-2 may have neurotoxic

PAGE 18

10 effects, such as T and B cell invasion of the brain, angiogenesis, changes in the composition of the extracellular matrix, myelin damage, and neuronal cell loss seen in rats administered human IL-2 intracerebroventricularly via minipumps (Hanisch et al., 1996; Hanisch et al., 1997b). Thus, my approach was to use IL-2 knockout mice. These studies were the first to investigate the consequences of the absence of IL-2 on aspects of brain development and maintenance of the septohippocampal system in vivo. Our laboratory had found the aforementioned behavioral alterations in IL-2 knockout mice, and therefore another important goal of my research was to test the hypotheses regarding the neurobiological and neuroimmunological alterations that may underlie these behavioral abnormalities.

PAGE 19

CHAPTER 2 ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT Introduction One of the earliest observations suggesting that cytokines could influence brain function in humans came from cancer treatment trials in which interleukin-2 (IL-2) was found to induce cognitive dysfunction and other untoward neuropsychiatric side effects in patients (Denicoff et al., 1987). Although basic research has demonstrated that IL-2 can modulate different aspects of central nervous system (CNS) function, some of IL-2's most prominent neurobiological actions occur in the hippocampal formation and related limbic regions, where receptors for this cytokine are enriched (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto etal., 1998). Exogenously administered IL-2 has effects on a number of parameters of septal and hippocampal neuronal function including trophic effects on cultured fetal septal and hippocampal neurons (Awatsuji et al., 1993; Sarder et al., 1996; Sarder et al., 1993). IL-2 may also modify cellular and molecular substrates of learning and memory such as longterm potentiation (Tancredi et al., 1990), and multiple parameters of cognitive behavioral performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al., 1999; Nemni et al., 1992). Moreover, the neurotrophic and neuromodulatory actions of IL-2 have been implicated in abnormal hippocampal development associated with schizophrenia (Ganguli et al., 1995; Licinio et al., 1993; McAllister et al., 1995). 11

PAGE 20

12 The potential effects of IL-2 on cholinergic neurons are particularly relevant to this study. In addition to the aforementioned trophic effects of IL-2 on cultured septal neurons, IL-2 is among the most potent modulators of acetylcholine (ACh) release from cultured septohippocampal neurons (Araujo et al., 1989; Hanisch et al., 1993; Seto et al., 1997), and can also modulate its precursor enzyme, choline acetyltransferase (ChAT) in fetal neurons (Mennicken and Quirion, 1997). Alterations in the cytoarchitecture of cholinergic septohippocampal neurons have been shown to correlate with differences in spatial learning ability in mice (Schwegler et al., 1996a; Schwegler et al., 1996b). We found that IL-2 knockout (IL-2 KO) mice exhibited impaired learning and memory performance, sensorimotor gating, and reductions in hippocampal infrapyramidal mossy neuronal fiber length (Petitto et al., 1999), a factor shown previously to correlate positively with spatial learning ability (Schopke et al., 1991; Schwegler and Crusio, 1995; Schwegler et al., 1988). In the present study, we therefore sought to test the hypothesis that loss of IL-2 would result in abnormal neurodevelopment of septal cholinergic neurons that project to the hippocampus. Since extensive data document that these neurons play a critical role in learning and memory performance (Galey et al., 1994; Leanza et al., 1995), and given the various in vitro neurotrophic and neuromodulator/ effects of IL-2 on developing septohippocampal cholinergic neurons, we postulated that IL-2 KO mice would have fewer cholinergic neurons in the medial septum and vertical limb of the diagonal band of Broca (MS/vDB) and a reduction in the cholinergic axonal density in the hippocampus. To accomplish this goal, IL-2 KO and wild-type littermates were compared using stereological techniques to count MS/vDB cholinergic somata stained with ChAT

PAGE 21

13 immunohistochemistry, and image analysis methods to measure the density and distribution of cholinergic neurites in several regions of the hippocampus labeled for acetylcholine esterase (AChE), a reliable marker of cholinergic axons (Hedreen et al., 1985). Materials and Methods Animals and Tissue Preparation Mice used in these experiments were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2 heterozygote by lL-2 heterozygote crosses. The polymerase chain reaction (PCR) was used to genotype the offspring post-weaning (see below). The IL-2 KO mice, obtained originally from the NIH repository at Jackson Labs, were derived from ten generations of backcrossing onto the C57BL/6 background. Mice were housed under specific pathogenfree conditions. Animals used in these experiments were 8-12 weeks of age. Each animal was anesthetized with sodium pentobarbital (50 mg/kg) and perftised with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS). The brains were removed and fixed overnight in 4% paraformaldehyde followed by overnight equilibration in 30% sucrose cryoprotective solution, and then were snap frozen in isopentane (-80C) for storage. The brains were equilibrated to -20C prior to cryostat sectioning into 40 ^im slices in the coronal plane, collected into individual wells of polystyrene 24-well plates (NUNC 1 147), and stored free-floating at 4C in PBS for histochemistry. Every third section was processed for ChAT immunohistochemistry, AChE histochemistry, or cresyl violet Nissl staining.

PAGE 22

14 Genotyping Using PCR The genotypes of all mice were determined by the PCR. PCR reactions were performed using a 25 |al total reaction volume containing 1 [iM each of forward and reverse primers, 0.1 ^g genomic DNA, 0.2 mM of each dNTP, 0.3 i^l Taq DNA polymerase and amplified using a thermal cycler with a heated evaporation cover (Ericomp). The cycling parameters were hot start 95 C (3min), denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final extension step of 4 min. Thirty cycles were used for these experiments. The 5' and 3' primers for the IL-2 KO (500 bp knockout band amplified) were 5'-TCGAATCGCCAATGACAAGACGCT3' and 5'-GTAGGTGGAAATTCTAGCATCATCC-3'. The 5' and 3 'primers for the wild type (324 bp wild type band amplified) were 5'CTAGGCCACAGAATTGAAAGATCT-3' and 5'GTAGGTGGAAAATTCTAGCATC ATCC-3 ChAT Immunohistochemistry Free-floating 40-^m sections were incubated for 20 minutes in 1% hydrogen peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then washed twice in PBS and blocked for 1 hr in 200 ^l/well 3% normal goat serum (NGS). After this incubation, the sections were incubated overnight in the primary antibody, rabbit anti-ChAT (Chemicon AB143; 1:2000 dilution in PBS with 0.3% Triton X-100 and 1% NGS, 200 nl/well). The next day, the sections were washed twice in PBS and incubated overnight in the secondary antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-lOO and 1% NGS). The sections were then washed twice in PBS and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS)

PAGE 23

15 for 2 hrs followed by two washes in PBS. The sections were developed in 0.5 mg/ml 3,3'-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped. Figure 2-1 shows an example of ChAT immunostained section of the MS/vDB. 2Q0 tim Figure 2-1. ChAT immunohistochemical staining of the medial septum/vertical limb of the diagonal band of Broca. The scale bar represents 200 ^m.

PAGE 24

16 AChE Histochemistry AChE histochemistry was used as a marker of chohnergic innervation of the hippocampus (Woolf et al., 1984). Brain sections were collected in individual wells of 24-well plates containing 250 ^1/well 0.1 M pH 6.0 acetate buffer (AB). The sections were washed twice with AB, then placed in 200 ^1 preincubation solution consisting of aqueous 5 mM sodium citrate, 3 mM cupric sulfate, and 0.5 mM potassium ferricyanide. The sections were incubated for 20 minutes at room temperature on a shaker a low speed. After the preincubation period, 200 ^il of the incubation solution was added to each well consisting of the same make-up as the preincubation solution supplemented with 4.84 mM acetylthiocholine iodide and 0.4 mM ethopropazine. The multi-well plate was packed on top of crushed ice and microwaved at 200 W for 2 minutes. The solution was then removed and the sections were washed twice in 0.05 M TRIS pH 7.6 buffer followed by AB. The reaction product was intensified with 0.5 mg/ml DAB, 2.5% nickel sulfate, and 0.01 % H2O2 in AB for 5-7 min or until definitive staining could be detected in the hippocampal subregions. Sections were then mounted on slides, dehydrated in grade ethanol washes, cleared in xylenes, and coverslipped for imaging. Cholinergic Stereology Stained cholinergic neuronal somata of the MS/vDB were counted using the software MCID 5.1 and the three-dimensional counting box (optical dissector) method described by Williams and Rakic (Williams and Rakic, 1988). All stereology was performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2 microscope with a motorized x-y stage made by Imaging Research, Inc. The latter is capable of making movements as fine as 0.1 ^im. Every third section through the anterior-posterior extent of the septal region was sampled. The regions to be counted

PAGE 25

17 were outlined at 1 Ox magnification and the size of the counting boxes were generated to be approximately 5% of the most rostral, and therefore, smallest, area of the MS/vDB (defined by the section where the corpus collosum first joins in the midline). The size of the outlined count regions, but not the counting box, varied depending on where the individual section was taken fi-om the rostral to caudal extent of the MS/vDB. The defined counting box was approximately 2-2.5% of the outlined count area of the largest single section of the MS/vDB. To assess whether the predicted septal cholinergic alterations in IL-2 KO mice might be associated with a general effect on cholinergic neurons in the brain, striatal cholinergic somata were also counted in the right hemisphere in the sections that also contained the MS/vDB. Except for a different magnification used to outline the striatum (5X), the sampling parameters were identical to those used to generate estimates of septal cholinergic neuron number. The guard volume was set at 1 |am for the top and bottom of the section and the counting cubes were randomly distributed throughout the user-defined count area with a total sampling frequency of 25% (-8.3% of the total area since every third section was sampled). Only somata that were clearly and distinctly stained were counted. Each counting box was examined at 40x magnification (20x for striatal neurons) and the computer-assisted focus was used to scan fi-om the top to the bottom of the counting box. Cells were counted only if they were either completely inside of the counting box, or partially inside of the box on the top, back, or left side. They were not counted if they fell outside of the box or crossed into the box anywhere on the bottom, fi-ont, or right side. The cells that were counted were labeled on the monitor by clicking the mouse pointer on each cell and the MCID software recorded the number of marks.

PAGE 26

18 The MCID software interpolated the total volume of the MS/vDB based on the volume of the count areas defined by the user. Cell density (Ny) was estimated by dividing the total number of cells counted by the volume of the counting boxes, which was also tracked by the software. The total cell number was estimated by multiplying Ny by the total volume. Quantitative Image Analysis of AChE Staining For quantitative analysis of AChE staining, we modified previously described methods used to measure intensity of staining and comparisons of normalized length across CAl, CA3b, and the suprapyramidal (SP) and infi-apyramidal (IP) blades of the dentate gyrus (DG) (King et al., 1989; Schwegler et al., 1996b). Images of the hippocampus in each tissue section sampled from a light microscope (Olympus BH-2) were relayed by digital video camera (Hitachi KP-D581) to a computer frame grabber (Flashpoint 128, Integral Technologies) and digitized to 640x480 pixel images with 256 gray levels from black to white, hnaging software (Image Pro Plus v.4.0. Media Cybernetics) was utilized to define a broad sampling traverse across various areas of the hippocampus, approximately perpendicular to the cell body layers. For CAl and CA3b, this traverse extended from the alveus to the hippocampal fissure. For the DG, the line extended from the hilus to hippocampal fissure or to the pial surface for the SP or IP limb of the DG, respectively. AChE-containing fiber density was estimated by using the gray level of each point along the line, which was calculated by averaging the intensity value of each pixel across the width of the sampling band (e.g. a traverse with a width of 46 pixels would have an average of 46 measurements approximately parallel to each data point). Gray level intensity measurements were used to estimate the AChE reaction product density at each point along the traverse. The mean pixel intensity of a small box

PAGE 27

19 positioned in the corpus collosum was used to represent the background staining intensity of tissue containing little cholinesterase activity. All of the average intensities along the line were then plotted by traverse position to illustrate the quantitative patterns of AChE distribution across each subregion. Conspicuous inflections marking the transition from alveus (Alv) to stratum oriens (SO, in CAl, CA3b), from stratum lacunosum-moleculare (Lmol) to dentate molecular layer (Mol) (hippocampal fissure (HiF); CAl, CA3, DG), and polymorph zone (PoDG) to granule cell layer (GCL) were used to align traverses across sections and animals (Figure 2-2). Also, length-normalized comparisons of the j\\y ..i.iii' Figure 2-2. A micrograph of the sampling regions utilized for image analysis of AChEstaining. Alv=alveus, DG=dentate gyrus, GCL=granular cell layer (IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus, Lmol=lacunosum moleculare, Mol=Tnolecular layer, Pi=pial surface, SO=stratum oriens, SP^septum pellucidum, SR=stratum radiatum. The scale bar represents 200 )am. measured regions were made by converting the sampling lengths to 100 points using the software Matlab v.5.3. Dependent variables were the absolute length of traverses, AChE intensity at anatomically identifiable inflection points (IP and SP bands, Lmol, dentate

PAGE 28

20 inner Mol) and subregions (SO, stratum radiatum (SR), dentate outer Mol), and derived values for absolute and relative positions of, and distances between, these landmarks. For each animal, three hippocampal slices were measured on both the left and right hemispheres of the brain and the measured intensities or distances were averaged together for statistical analysis. Cresyl Violet Staining Every third section was Nissl stained to provide a qualitative view of the boundaries between various forebrain regions. The sections were placed on slides and allowed to air-dry. The slides were immersed in 60C cresyl violet for 45 sec, washed in running distilled water to remove the excess cresyl violet, dehydrated in graded ethanol, cleared in xylenes, and coverslipped. Results Comparison of Cholinergic Somata in the MS/vDB Figure 2-3 shows the total number of ChAT-positive cells stereologically counted from the MS/vDB of IL-2 KO and wild-type mice. As seen in this figure, the IL-2 KO mice had approximately 26% fewer cholinergic somata in this region than wild-type controls. An ANOVA confirmed that this group difference was statistically significant (F(l,16)=8.6, p=.01). By contrast, the number of ChAT-positive somata in the striatum of IL-2 KO and wild-type mice were not different. There were no significant gender differences in either brain region. Density of AChE-positive Fibers in Regions of the Hippocampus The average AChE-staining intensity curves were generated by defining a region across CAl, CA3b, and the SP and IP blades of the DG. Figure 2-4 shows the intensity curves that were generated from imaging the AChE-histochemically stained sections for

PAGE 29

21 Cholinergic Somata in the MSA/DB I 500 I 1500 0 I 1 Wid-Type Ib2 Knockout Groups Fig. 2-3. ChAT-positive somata are significantly reduced in the MS/vDB of IL-2 knockout mice. Each bar represents the mean SEM of 9 animals per group. *p=0.01. CAl (Figure 2-4a), CA3 (Figure 2-4b), the SP-GCL (Figure 2-4c), and IP-GCL layer of the DG (Figure 2-4d). Repeated measures ANOVA was performed on regions of the average normalized curves selected by areas that appeared to deviate between the groups. None of these areas, however, were found to differ between IL-2 KO and wild-type mice. Morphology of the Granular Cell Layer of the Lower Limb of the DG There were no differences in the groups for the Y-axis (intensity) data. The variations in the patterns of the DG curves X-axis (distance) were also compared. Distances were compared by defining the point of lowest intensity in the regions that the curves indicated as each transition between the regions of interest. Distances are reported as a percentage of the total distance across each curve. The IP blade of the DG was broken into three regions: the Hil, the IP-GCL, and the Mol (Figure 2-4d). As depicted in

PAGE 30

22 A) Average AChE InUnsily of CA1 0 10 30 50 60 ?0 aO W 100 Qsance (% olToial) C) Average AChE Intensity of ttie Suprapyramldal Limb of ine Dentate Gyrus 10 30 30 10 50 60 703:90 100 Daarce (% ol Tota) B) Average AChE Intensity of CA3 D) Average AChE Intensity of the Infrapyramldal Limb of the Dentate Gyrus 30 30
PAGE 31

23 Figure 2-5, the distance across the EP-GCL was significantly reduced in the IL-2 KO mice compared to wild-type mice (F(l,16)=9.2, p=0.008). The distances across theHil and Mol, however, were not significantly different. The SP blade of the DG was separated into three regions: the Mol, the SP-GCL, and the Hil (Figure 2-4c). There were no significant differences in length between groups across any of the internal blade regions. Distance Across GrDG of the External Blade Will-Type IL-2 Knockout Groups Fig. 2-5. Distance across the GrDG of the external blade was significantly decreased in IL-2 knockout mice compared to wild-type mice. Each bar represents the mean SEM of nine animals per group. *p=0.008. Discussion These data are the first to demonstrate that loss of endogenous IL-2 results in reduction in the number of MS/vDB cholinergic neurons and structural alterations in the morphology of the dentate gyrus. Given the role of septohippocampal cholinergic neurons in learning and memory (Galey et al., 1994; Leanza et al., 1995; Schwegler et al., 1996a; Schwegler et al., 1996b), sensorimotor gating (Caine et al., 1992; Curzon et al..

PAGE 32

24 1994; Koch, 1996a), and the aforementioned influence of IL-2 on parameters of cholinergic function, the present findings are consistent with alterations in these behavioral measures that we have reported previously in IL-2 KO mice (Petitto et al., 1999). In this study, the number of cholinergic cell bodies in the MS/vDB of wild-type mice were comparable to those reported previously for C57BL/6 mice (Schwegler et al., 1 996b). Cholinergic somata were reduced by 26% in IL-2 KO mice as compared to wildtype mice. This was not a general effect on cholinergic neurons in the brain, however, as striatal ChAT-positive neurons were not significantly affected. This finding is also consistent with previous research showing that exogenous IL-2 has potent effects on ACh release fi-om septohippocampal neurons, whereas cholinergic intemeurons in the striatum do not respond to IL-2 (Hanisch et al., 1993). One potential caveat of using ChAT as a marker of cholinergic neurons is that differences in cell counts may be due to a decrease in ChAT labeling intensity rather than a reduction in cholinergic neurons (Ward and Hagg, 2000). This would appear unlikely, however, as we did not note any appreciable differences in the staining intensity between groups. Alterations in AChE-staining intensity did not differ as we hypothesized. It is not likely that the unexpected lack of difference in AChE-staining intensity would be attributable to the length normalization technique used, as the patterns of the average intensity curves seen in Figure 2-4 were remarkably similar between the groups. This unexpected result may, however, be due to compensatory sprouting of the surviving MS/vDB neurons during development. Indeed, numerous other studies have found that septohippocampal neurons can undergo compensatory sprouting in response to injury

PAGE 33

25 (Cassel et al., 1997; Gage and Bjorklund, 1987; Gage et al., 1984; Gage et al., 1983b), and in animal disease models such as Alzheimer's transgenic mice (Bronfman et al., 2000). Future studies are needed to address this issue more directly. Another significant finding of the current study was that IL-2 KO mice exhibited structural alterations in the distance across the IP-GCL. The neurons of the GCL have been associated with learning and memory (Collier and Routtenberg, 1984; Conrad and Roy, 1993; McLamb et al., 1988; Nanry et al., 1989; Walsh et al., 1986), and are also a target for septohippocampal cholinergic axon termination (Makuch et al., 2001). In situ hybridization studies have found the GCL to be enriched in IL-2 receptors (Petitto and Huang, 2001 ; Petitto et al., 1998), supporting a possible role for IL-2 in the observed structural alterations. Also, GCL development progresses from the SP layer to the IP layer (Bayer, 1980). The differences seen in the IL-2 KO mice in this study may indicate a failure of these late stage granule cells to fully develop or survive. Whether these structural changes are due to a reduced number of GCL cells or a decrease in the cell body size of these neurons requires further investigation. The most likely mechanism whereby loss of IL-2 results in these changes in the septohippocampal cholinergic system would appear to be due to the absence of its neurotrophic actions during development. As noted earlier, IL-2 enhances neurite extension and survival of cultured fetal septal and hippocampal neurons (Awatsuji et al., 1993; Hanisch and Quirion, 1995a; Sarder et al., 1996; Sarder et al., 1993), and thus, the absence of these intrinsic effects of IL-2 could account for the observed neuroanatomical alterations. Another mechanism that may account for these findings is the possibility that the loss of endogenous IL-2 may result in lower levels of tonic ACh release during

PAGE 34

26 critical periods of neurodevelopment. Release of ACh by developing neurons has been shown to be important for growth cone guidance (Zheng et al., 1994), neuronal growth and differentiation, synaptic plasticity (Lauder and Schambra, 1999), and survival of newly developed neurons (Knipper and Rylett, 1997). In fact, some evidence indicates that ACh released from developing neurons may engage in a positive feedback mechanism with nerve growth factor (NGF) (Knipper et al., 1994), a member of the neurotrophin family that is essential for the normal development of septal cholinergic neurons (Arimatsu and Miyamoto, 1991; Hartikka and Hefti, 1988; Mobley et al., 1986; Ruberti et al., 2000). In a series of studies, Quirion's laboratory has demonstrated that IL-2 is among the most potent modulators of ACh release from mature brain slices and fetal neurons in vitro, and can upregulate ChAT in fetal septal neurons in culture (Hanisch et al., 1993; Mennicken and Quirion, 1997; Seto et al., 1997). It is therefore possible that the loss of such potent actions of IL-2 during development could account, in part, for the cytoarchitectural alterations found in this study. Indeed, both IL-2's neurotrophic effects and action on cholinergic release may well be operative and interactive with one another. Nevertheless, the IL-2 KO mice do not exhibit complete loss of septal cholinergic neurons suggesting that the effects of IL-2 on MS/vDB neurons are likely secondary to other trophic factors like NGF. These experiments do not enable us to differentiate between the contributions of the loss of central versus peripheral IL-2, and thus, it remains to be determined whether these septohippocampal cholinergic abnormalities are due primarily to the absence of central, peripheral, or a combination of both sources of IL-2. There is some evidence that endogenous IL-2 may be produced in neuronal areas of the mammalian hippocampal

PAGE 35

27 formation, where its release may regulate the development and function of septal cholinergic neurons projecting to the hippocampus. IL-2-like immunoreactivity has been localized to the hippocampal formation in rat forebrain (Lapchak et al., 1991), and detected in tissue extracts from rat and human hippocampal tissue (Araujo et al., 1989). In mouse brain, IL-2 mRNA has been found in the hippocampus (Villemain et al., 1991), and transcripts for this cytokine may be expressed in rat astrocyte cultures as well (Eizenberg et al., 1995). In the periphery, absence of endogenous IL-2 leads to an immunodysregulation that produces loss of self-tolerance and IL-2 KO mice eventually develop generalized systemic autoimmune disease (although C57BL/6-1L-2 KO mice develop clinical signs of systemic autoimmunity at a substantially slower rate than other strains such as Balb/c or C3H) (Petitto et al., 2000). Therefore, it is reasonable to speculate that the neuroanatomical alterations found in the IL-2 KO mice result from peripheral autoimmune processes. Autoimmunity could impact on brain development or induce neurodegeneration. The former seems unlikely, however, since IL-2 KO mice do not express the first signs autoimmunity (e.g., splenomegaly) until at least three to four weeks after birth (Horak, 1995); by this time, septohippocampal development should already be complete (Bender et al., 1996; Chandler and Crutcher, 1983; Super and Soriano, 1994; Yoshida and Oka, 1995). Furthermore, the likelihood that these neuroanatomical alterations may be due to autoimmune-induced degeneration of existing neurons also seems unlikely, since lymphocytes cannot be detected in the brain of adult IL-2 KO mice (Petitto et al., 1999). Nonetheless, since autoimmunity has been associated with cognitive changes in both animals and humans (Lai and Forster, 1988; Sakic et al., 1997;

PAGE 36

28 Sakic et al., 1993), more subtle autoimmune processes may be at play in the IL-2 KO mice (e.g., autoantibodies). It would be of interest to explore the observed cholinergic changes in old versus neonatal mice to determine if the abnormalities increase with age due to neurodegeneration, or are primarily the result of abnormal development. Such knowledge will then enable us to develop a more specific model to test relevant hypotheses involving IL-2 at specific anatomical sites in the septohippocampal cholinergic system. In summary, these data demonstrate that loss of endogenous IL-2 results in reduction in the number of cholinergic neurons in the MS/vDB and alterations in the structural morphology of dentate projection fields. These findings extend our previous experiments showing that spatial learning and hippocampal mossy fiber length are abnormal in IL-2 KO mice (Petitto et al., 1999). Further research is needed to determine whether these outcomes in IL-2 KO mice may be due to the absence of central or peripheral IL-2 during neurodevelopment (or some combination of both sources), neurodegeneration secondary to peripheral autoimmunity, or other factors associated with the absence of IL-2.

PAGE 37

CHAPTER 3 ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN INTERLEUKIN-2 KNOCKOUT MICE Introduction Interleukin-2 (IL-2) has been implicated in the pathogenesis of multiple sclerosis and several major neuropsychiatric disorders such as Alzheimer's disease, schizophrenia, and Parkinson's disease (Hanisch and Quirion, 1995b). Furthermore, in case studies of humans receiving IL-2 treatment for cancer therapy, prolonged exposure to IL-2 was found to induce cognitive dysfunction and other untoward neuropsychiatric side effects (Denicoff et al., 1987). Although IL-2 has been shown to be capable of modulating different aspects of central nervous system (CNS) function, many of its known effects in the limbic system occur in the hippocampal formation, where receptors for this cytokine are enriched (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto et al., 1998). IL-2 may, for example, modify cellular and molecular substrates of learning and memory such as long-term potentiation (Tancredi et al., 1990), and can affect multiple parameters of cognitive behavioral performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al., 1999; Nemni et al., 1992). IL-2 can provide trophic support to primary cultured neurons from multiple region of the rat brain, including the hippocampus and medial septum (Awatsuji et al., 1993; Sarder et al., 1993), and positively affects the morphology of neurite branching from rat hippocampal cultures (Sarder et al., 1996; Sarder et al., 1993). Furthermore, lL-2 has been shown to be one of the most potent modulators of 29

PAGE 38

30 acetylcholine (ACh) release from rat hippocampal slices (Hanisch et al., 1993; Setoet al., 1997), and can also increase the activity of its precursor enzyme, choline acetyltransferase (ChAT) (Mennicken and Quirion, 1997). Previously, we found that IL-2 knockout mice (IL-2 KO) exhibited impaired learning and memory performance, sensorimotor gating, and reductions in hippocampal infrapyramidal mossy neuronal fiber length (Petitto et al., 1999), a factor which correlates positively with spatial learning ability (Schopke et al., 1991; Schwegler and Crusio, 1995; Schwegler et al., 1988). We also found in studies of IL-2 KO mice in vivo, there was a marked reduction in cholinergic somata in medial septal/vertical limb of the diagonal band of Broca (MS/vDB) region, as well as decrease in the distance across the infrapyramidal granule cell layer (IP-GCL) of the dentate gyrus (DG) (Beck et al., 2002). Variation in the cytoarchitecture of cholinergic septohippocampal neurons correlate with differences in spatial learning ability in mice (Schwegler et al., 1996a; Schwegler et al., 1996b). Research has shown that the neurotrophins, nerve growth factor (NGF) and brainderived neurotrophic factor (BDNF) expressed in the hippocampus, can be important in the development, maintenance, and repair of septohippocampal neurons in vitro (Arimatsu and Miyamoto, 1991; Conner and Varon, 1997; Gahwiler et al., 1987; Hartikka and Hefti, 1988; Morse et al., 1993). Similar trophic effects have been noted in studies utilizing infusion of exogenous neurotrophins in vivo (Hagg et al., 1990; Morse et al., 1993), and in studies of transgenic and knockout mice (Ruberti et al., 2000; Ward and Hagg, 2000). In the peripheral immune system of multiple animal species, both NGF and BDNF are expressed by T lymphocytes (Braun et al., 1999; Kerschensteiner et al., 1 999;

PAGE 39

31 Mizuma et al., 1999; Moalem et al., 2000). Though IL-2 regulates several aspects of T cell function, the production or release of NGF and BDNF from T lymphocytes by IL-2 has not been tested directly(Carter et al., 1998; He and Maiek, 1998), and conversely, it is not known whether NGF and BDNF can modulate IL-2 in the brain. In the present study, we sought to expand our previous findings that loss of endogenous IL-2 in knockout mice led to reductions in cholinergic neurons of the MS/vDB and a decrease in the distance across the IP-GCL of the DG. First, we compared parvalbumin (Parv)-labeled somata in 8-12 week old IL-2 KO and wild-type littermates to test the hypothesis that the loss of IL-2 is selective for cholinergic, but not GABAergic cell bodies loss in the MS/vDB (e.g., not a general effect occurring on all neurons in this region of the brain). Second, we compared cholinergic MS/vDB somata between younger wild-type and IL-2 KO mice at postnatal day 21 (P21), an age where septohippocampal development in mice is nearly complete (Armstrong et al., 1987; Gould et al., 1991 ; Makuch et al., 2001). This age also precedes the development of autoimmune disease in IL-2 KO mice of the C57BL/6 background, e.g., absence of splenomegaly, lymphadenopathy, and inflammatory bowel disease. Third, we sought to expand on the previous finding that there was a reduction in distance across IP-GCL by performing stereological cell counts of Nissl-stained dentate gyri of IL-2 KO and wildtjTDe littermates to determine if the reduction in distance could be contributed to a reduction in granule cell number. Finally, we tested the hypothesis that loss of IL-2 may impact the expression and release of the neurotrophins, NGF and BDNF, which may contribute to the MS/vDB cholinergic and GCL deficits that we observed previously.

PAGE 40

32 Methods Animals and Genotyping Mice used in these experiments were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2 heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the NIH repository at Jackson Laboratories, were derived from ten generations of backcrossing onto the C57BL/6 background. Mice were housed under specific pathogenfree conditions. Animals used in these experiments were either 2 1 -days-old (for ChAT immunohistochemistry) or 8-12 weeks of age. All experiments were performed with independent groups of animals. The specific animal numbers utilized are reported at the beginning of each method descriptions below. The genotypes of all mice were determined by the polymerase chain reaction (PCR). PCR reactions were performed using 25 |il total reaction volume containing 1 }iM each of forward and reverse primers, 0.1 fag genomic DNA, 0.2 mM of each dNTP, 0.3 10.1 Taq DNA polymerase, and amplified using a thermal cycler with a heated evaporation cover (Ericomp). The cycling parameters were hot start 95C (3min), denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final extension step of 4 min. Thirty cycles were used for these experiments. The 5' and 3' primers for the IL-2 KO (500 bp knockout band amplified) were 5'TCGAATCGCCAATGACAAGACGCT-3' and 5'GTAGGTGGAAATTCTAGCATCATCC-3'. The 5' and 3 'primers for the IL-2 wild type (324 bp wild type band amplified) were 5'-

PAGE 41

33 CTAGGCCACAGAATTGAAAGATCT-3' and 5'GTAGGTGGAAAATTCTAGCATCATCC-3 Immunohistochemistry For Parv stereological cell counts, six animals per group were used and for 21day-old ChAT stereological cell counts, seven animals per group were used. Each animal was anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine ( 1 0 mg/ml) at a dose of 0.01 5 ml injection cocktail/g body weight and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS). The brains and spleens were removed and fixed overnight in 4% paraformaldehyde, followed by overnight equilibration in 30% sucrose cryoprotective solution, and then, were snap frozen in isopentane (-80C) for storage. The spleens were weighed to assay for relative splenomegaly of IL-2 KO vs. wild-type mice. The brains were equilibrated to -20C prior to cryostat sectioning into 50 \im slices in the coronal plane, collected into individual wells of polystyrene 24-well plates (NUNC 1 147), and stored free-floating at 4C in PBS for histochemistry. Every third section was processed for Parv or ChAT immunohistochemistry, or cresyl violet Nissl staining. Free-floating 50-|am sections were labeled for Parv and ChATimmunohistochemistry as described previously (Beck et al., 2002). Briefly, they were incubated for 20 minutes in 1% hydrogen peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then washed and blocked for 1 hr in 200 )il/well 3% normal goat serum (NGS). After this incubation, the sections were incubated overnight in the primary antibody, rabbit anti-ChAT (Chemicon; 1 :2000 in PBS with 0.3% Triton X-100 and 1% NGS, 200 ^1/well) or rabbit anti-Parv (Chemicon; 1:1000 in PBS with 0.3% Triton X-100 and 1% NGS, 200 ^il/well). The next day, the sections were

PAGE 42

34 washed and incubated overnight in the secondary antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-lOO and 1% NGS). The sections were then washed and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS) for 2 hrs. The sections were developed in 0.5 mg/ml 3,3'-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped. Cresyl Violet Staining For stereological cell count of Nissl-stained granule cells, seven animals per group were used. The tissue for this assessment was selected, because it originated from animals utilized in a previous study of cholinergic differences in IL-2 KO mice (Beck et al., 2002). Every third section was Nissl stained to provide a qualitative view of the boundaries between various forebrain regions, as well as labeling of hippocampal granule cells layers (GCL) for stereological counting. The sections were placed on slides and allowed to air-dry. The slides were immersed in 25C cresyl violet for 10 min, washed vigorously in rapid exchanges of distilled water to remove the excess cresyl violet, dehydrated in graded ethanol, cleared in xylenes, and coverslipped. Stereology Stained neuronal somata of the MS/vDB or IP and SP-GCL were counted using the software MCID 5.1 as previously described (Beck et al., 2002). All stereology was performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2 microscope with a motorized x-y stage made by Imaging Research, Inc. Every third section through the anterior-posterior extent of the MS/vDB or IP and SP-GCL regions were sampled. The regions to be counted were outlined at lOx (MS/vDB) or 20x (IP and SP-GCL) magnification and the size of the counting boxes were generated to be

PAGE 43

35 approximately 2-2.5% of the outlined count area of the largest single section of the areas of interest. The rostral border of the GCL count area was defined as the first section where the dentate granule cell layer clearly separated from the pyramidal layer of CAl. The caudal border was defined as the first section where the habenular commisure was visible in the third ventricle. Only cells that could clearly be determined to be part of either the IP or SP-GCL were counted; any cells in the area where the IP and SP-GCL connected were left uncounted. Furthermore, only Nissl-stained cells with clearly visible nucleoli were counted. For the MS/vDB, the rostral border was determined as the first section where the corpus collosum connected in the midline of the section and the caudal border was the first section where the anterior commisure joined in the midline. The guard volume was set at 2 |im for the top and bottom of the section and the counting cubes were randomly distributed with a total sampling frequency of the outlined count area of 25% for MS/vDB and 33% for IP and SP-GCL (-8.3% and 1 1% of the total area respectively, since every third section was sampled). The outlined counting area was defined by the user and only somata that were clearly and distinctly stained were counted. Each counting box was examined at 40x magnification and the computerassisted focus was used to scan from the top to the bottom of the counting box. Cells were counted only if they were cither completely inside of the counting box, or partially inside of the box on the top, back, or left side. They were not counted if they fell outside of the box or crossed into the box anywhere on the bottom, front, or right side. The cells that were counted were labeled on the monitor by clicking the mouse pointer on each cell and the MCID software recorded the number of marks.

PAGE 44

36 The MCID software interpolated the total volume of the MS/vDB based on the volume of the count areas defined by the user. Cell density (Nv) was estimated by dividing the total number of cells counted by the volume of the counting boxes, which was also tracked by the software. The total cell number was estimated by multiplying Nv by the total volume. Enzyme-linked Immunosorbent Assay (ELISA) Characterization of NGF and BDNF For measurement of BDNF in hippocampus and MS/vDB, nine IL-2 KO and seven wild-type mice were used. For measurement of NGF protein levels, seven animals per group were used. Initial test runs revealed that some, but not all, of the NGF protein levels fell below the sensitivity of detection for the kit; therefore, the homogenized NGF samples were spiked with 25 pg/ml of the known NGF standard included with the kit to bring any low levels of expression above the kit's 15.6 pg/ml lower detection limit. One column of the ELISA plate was also run with only the 25 pg/ml standard spike to provide a baseline and the data reported are corrected for this. Animals used for ELISA characterization of neurotrophin levels only received saline perfusion and were not postfixed in parafonnaldehyde. The brains were removed, snap frozen, and then allowed to equilibrate to -20 C. The brains were sectioned on a cryostat at -20-22 C at 400 [im thickness and the MS/vDB and hippocampi were dissected with a 0.75 mm micropunch on a -20 C freezing platform. The dissected tissue was weighed on a microgram scale, and then transferred to 25 ^il of homogenizing solution (50 mM Na/Na2 and 0.2% TX100 in H2O with Anti-protease Complete TM cocktail (Boehringer)) per mg of wet weight tissue. The tissue was sonicated in the homogenizing solution for 30-sec on ice

PAGE 45

37 and centrifuged at 16,000 g for 15 min at 4" C. The supernatant was collected and stored at -20 C for ELISA analysis. Levels of NGF and BDNF were analyzed in the homogenates from MS/vDB and hippocampus using a commercially available Emax Immunoassay System according to the manufacturer's instructions (Promega). Briefly, the 96-well plates were coated with 1 :6,250 anti-NGF pAb in carbonate coating buffer (0.025 M sodium bicarbonate, 0.025 M sodium carbonate, pH 9.7) and incubated overnight at 4 C. The plates were washed with TBST wash buffer (20 mM Tris-HCL pH 7.6, 1 50 mM NaCl, 0.05% (v/v) Tween 20) and blocked with Ix Block and Sample buffer (provided with kit) for 1 hour. The plates were washed again with TBST and a set of standard curves were generated in duplicate by performing 1 :2 dilutions of a known 500 pg/ml standard in a range from 500 pg/ml to 7.6 pg/ml followed by a "blank" well of 0 pg/ml. All added samples and standards were allowed to incubate at 25 C for 6 hours. The plates were washed thoroughly with TBST and 1 :4000 anti-NGF mAb was added and incubated overnight at 4 C. The plates were again washed with TBST and 1 : 100 anti-rat IgG pAb conjugated to HRP was added for 2.5 hours at room temperature. The plates were washed and TMB One Solution was added for color development for 10 min. The reaction was stopped with the addition of equal volume of 1 N HCl and the absorbance was read at 450 nm within 30 min of the color development reaction. The data were reported as pg of protein per mg wet weight tissue. Statistical Analysis Results are reported as the mean SEM. Statistical differences between groups were determined using analysis of variance (ANOVA).

PAGE 46

38 Results Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell Number in Adult Mice In 8-12-week-old mice, no significant differences were apparent in the relative number of stereologically counted Parv-positive neurons between groups (F(l,10)=0.002, p=0.964). Thus, the GABAergic neurons appear to be unaffected by IL-2 gene deletion. We did not assay for GABAergic alterations in younger animals, since there were no differences in the adult IL-2 KO mice relative to the wild-types. In contrast to the previously reported data from 8-12-week-old animals (Beck et al., 2002), there was no significant difference in stereologically counted cholinergic somata number in the MS/vDB of 21-day-old IL-2 KO mice relative to wild-type mice (F(l,12)=0.689, p=0.423). As expected, the 21-day-old KO mice also did not exhibit splenomegaly seen in the autoimmune 8-12-week-old group, as spleen weights did not differ between 21-day-old wild type and IL-2 KO mice (F(l,12)=0.989, p=0.340). Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice The IP-GCL of IL-2 KO mice had significantly fewer neuronal somata than wildtype mice (Fig. 3-1; F(l,12)=l 0.966, p=0.006). In the SP-GCL, however, there was no significant difference in granule cell number (Fig 3-1.; F(l,12)=0.197, p=0.665). Alterations in Neurotrophin Levels Levels of NGF protein in hippocampal tissue homogenates was significantly increased in IL-2 KO mice relative to wild-type mice (Fig. 3-2A; F(l,10)=8.261, p=0.017). The levels of BDNF protein, conversely, were significantly decreased in IL-2 KO mice relative to wild-type mice (Fig. 3-2B; F(l,12)=8.023, p=0.015).

PAGE 47

39 Stereologicd Cell Count of Nissl-steuned Granule Cell Layers 1200 Infrapvramidal Suprapyramidal I Wild-Type IL-2 Knockout Fig. 3-1. There was a significant reduction in infrapyramidal, but not suprapyramidal granule cells in IL-2 knockout relative to wild-type mice. Each bar represents the mean SEM of seven animals per group. *p=0.006. A) Enhanced NGF Levels in the Hippocampus oriL-2 Knocl(out Mice B) Reduced BDNF Levels in the Hippocampus of IL-2 Knockout Mice 30 25 5" 20 Z 15LL 2 10 W*d-Type IL-2 Knoctioul Genotype Wild-Type IL-2 Knockout Genaype Figure 3-2. There was a significant: A) increase in NGF and B) decrease in BDNF protein levels in the hippocampus of IL-2 knockout compared to wild-t>'pe mice. The NGF bars represent the mean SEM of 7 animals per group. The BDNF bars represent mean SEM of 7 IL-2 knockout and 9 wild-t>pe mice. *p<0.05.

PAGE 48

40 Discussion These data are the first to demonstrate that the loss of endogenous IL-2 in knockout mice can lead to alterations in neuronal cell number in the IP-GCL and production of the neurotrophins, BDNF and NGF. Further, this study expands upon the previous finding that IL-2 gene deletion leads to a deficiency of cholinergic neurons in the MS/vDB (Beck et al., 2002), by showing a lack of significant cholinergic neuronal differences in MS/vDB of 21-day-old IL-2 KO animals or GABAergic alterations in adult animals. The lack of a difference in GABAergic neurons in 8-12-week-old adult mice was consistent with our initial hypothesis, as there is no evidence in the literature that IL-2 has any modulatory effects on GABAergic neurons. Moreover, this is not a regional effect, but rather appears to be selective to cholinergic projection neurons. As previously mentioned, IL-2 is a potent modulator of ACh release (Hanisch et al., 1993; Seto et al., 1997), and its precursor enzyme ChAT (Mennicken and Quirion, 1997), suggesting an effect of IL-2 on cholinergic neurons. In GABAergic neurons, however, IL-2 has failed to evoke release of GAB A in mesencephalic neuronal cultures (Alonso et al., 1993), or the cortex or hippocampus of mice (Bianchi et al., 1995). Since IL-2 deficiency does not affect the number of GABAergic somata in the MS/vDB of IL-2 KO mice, the neuronal loss appears to be selective for cholinergic neurons in the MS/vDB. Furthermore, we previously found no differences in the striatal cholinergic neuronal number (Beck et al., 2002), so the lack of IL-2 does not simply cause a general loss of all cholinergic neurons. Against our initial hypothesis that 21-day-old IL-2 KO mice would have similar cholinergic deficiencies as adult 8-12-week-old mice, there was no detectable loss of cholinergic cell number in the MS/vDB. We did not examine 21-day-old mice for

PAGE 49

41 differences in GABAergic cell number, since we did not detect any changes in adult mice using the same marker. One potential explanation for the loss of cholinergic neurons in the MS/vDB may be a failure in maintenance or survival in the late stages of, or after, development. Other studies have found decreases in cholinergic enzyme activity (i.e., ChAT and AChE) between postnatal days 30-60 in normal rats (Thai et al., 1992), and postnatal days 60-150 in C57BL/6 mice (Virgili et al., 1991). The IL-2 KO mice may potentially be more susceptible to this loss of cholinergic activity during adulthood, which could lead to the previously observed deficiencies in 8-12-week-old IL-2 KO animals. An alternate explanation for the different cholinergic effects seen in 21 -day-old vs. 8-12-week-old animals is that the loss of IL-2 may be secondary to the effects of autoimmunity present in adult IL-2 KO animals. Though we cannot completely rule out this possibility, we have previously failed to find discemable levels of infiltrating lymphocytes or clear signs of gliosis in the brains of IL-2 KO animals (Petitto et al., 1 999). More research is necessary to further address this issue. Another finding of this study was a significant decrease in neuronal cell number in the DP-GCL, but not the SP-GCL. This decrease is consistent with the in vitro studies showing a potent neurotrophic effect of IL-2 on hippocampal neurons (Awatsuji et al., 1993; Sarder et al., 1996; Sarder et al., 1993). Furthermore, these data are supported by our previous findings that IL-2 KO mice exhibited a reduced distance across the IP-GCL (Beck et al., 2002), and that the IP mossy fiber length of IL-2 KO mice is shorter than wild-type controls (Petitto et al., 1999). The reductions in distance across the IP-GCL could also potentially be explained by increased density, but not number, of cells or

PAGE 50

42 smaller cell body size. Qualitative assessments of random granule cells, however, do not support this hypothesis, though a more extensive study would be necessary to definitively address that issue. Though the receptors for IL-2 are more abundant in the hippocampus, including the GCL of the DG (Petitto and Huang, 1994; Petitto et al., 1998), it is not clear whether IL-2 may act directly on these neurons, or whether it upregulates other growth factors like the neurotrophins. The observed differences in the level of the neurotrophin BDNF was consistent with our hypothesis that we would find a reduction in trophic factors important in MS/vDB and hippocampal development and maintenance. BDNF plays a role in the maintenance and repair of septal cholinergic neurons (Alderson et al., 1990; Morse et al., 1993; Ward and Hagg, 2000), can implement a positive feedback mechanism with these neurons to enhance the release of ACh (Knipper et al., 1994), and can also modulate neurogenesis (Larsson et al., 2002; Lee et al., 2002), thus potentially impacting granule cell number. Thus, the reduction of cholinergic cell number in the MS/vDB is consistent with a reduction in this trophic factor. The exact interaction between IL-2 gene deletion and the reduction of BDNF levels remains unclear. Though BDNF is expressed in the peripheral immune system by lymphocytes, IL-2 does not stimulate its production or release. IL-2 can, however, upregulate the expression of TrkB, the receptor for BDNF, in lymphocytes (Besser and Wank, 1999). Furthermore, some evidence suggests that BDNF can stimulate a positive feedback mechanism of its own production via the TrkB receptor in hippocampal neurons (Canossa et al., 1997; Saarelainen et al., 2001). In IL-2 KO mice, the absence of IL-2 may therefore potentially lead to a down-regulation of the TrkB receptor, thereby partially inhibiting the positive

PAGE 51

43 feedback production of BDNF. Interestingly, the neurotrophin Trk receptors and IL-2 receptor share some of the same signal transduction pathways (e.g., mitogen activated protein kinase or phosphatidylinositol 3-kinase), which appear to play a role in their growth and survival promoting actions (for reviews see Gaffen, 2001 ; Patapoutian and Reichardt, 2001). Whether IL-2 knockout leads to disruption of one of these signal transduction pathways has not, to our knowledge, been elucidated and thus requires further study. Against our initial hypothesis, NGF protein levels were actually increased in the IL-2 KO mice. Unlike BDNF, NGF does not appear to stimulate a positive feedback neurotrophin release from hippocampal neurons (Canossa et al., 1997). Given the reduction in cholinergic survival in the MS/vDB of IL-2 KO mice, the target neurons in the hippocampus of these animals may produce higher protein levels of NGF as a compensatory response. Similarly, moderate lesions of rat septohippocampal projections lead to increased mRNA expression of NGF, but not BDNF in hippocampal target cells (Hellweg et al., 1997). In summary, cholinergic deficits seen in the MS/vDB of IL-2 KO mice appear to be selective for cholinergic over GABAergic neurons. In addition, the loss of cholinergic neurons in the MS/vDB may occur in the later stages of, or after, development of the septohippocampal system, as the deficits are not seen in 21 -day-old IL-2 KO mice. In the hippocampus, the number of neurons in the EP-GCL is significantly reduced. A reduced production of hippocampal BDNF may contribute to many of the aforementioned changes, though NGF levels are increased in a possible compensatory response. Although overt signs of autoimmunity in the brain are not apparent (we have been unable

PAGE 52

44 to detect significant levels of leukocyte infiltration or gliosis in IL-2 KO mice brains), further study is necessary to assess this possibility, as factors such as IL-2 induced cytokine dysregulation or autoantibodies could contribute to the hippocampal alterations in adult IL-2 KO mice.

PAGE 53

CHAPTER 4 INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND NEUROGENESIS IN THE HIPPOCAMPUS Introduction Receptors for interleukin-2 (IL-2) are enriched in the hippocampal formation, and many of the most prominent neurobiological functions of this cytokine occur in the hippocampus (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto et al., 1998). Previous studies from our laboratory have found that IL-2 knockout (KO) mice exhibit significantly lower numbers of medial septum and vertical limb of the diagonal band of Broca (MS/vDB) cholinergic cell bodies, a reduction in the distance across the granular cell layer (GCL) of the infrapyramidal (IP) blade of the dentate gyrus (DG), and decreased fiber length and neuronal cell number in the IP-GCL of the DG (Beck et al., 2004; Beck et al., 2002; Petitto et al., 1999). These neurobiological alterations appear to be related to abnormalities in learning and memory performance and sensory motor gating in IL-2 KO mice (Cushman et al., 2004; Petitto et al., 1999). Because IL-2 has been shown to possess various neurotrophic and neuromodulatory effects on hippocampal neurons in vitro (Awatsuji et al., 1993; Bianchi et al., 1995; Pauli et al., 1998; Plata-Salaman and ffrench-Mullen, 1993; Sarder et al., 1996; Sarder et al., 1993; Tancredi et al., 1990), our original working hypothesis was that the alterations exhibited by IL-2 KO mice are due to the absence of IL-2's neurotrophic actions on hippocampal neurons during development. More recent data from our 45

PAGE 54

46 laboratory, however, suggests that these hippocampal changes may be due to neurodegenerative rather than neurodevelopmental processes. We tested mice at postnatal day 21 (P21), an age where septohippocampal cholinergic neurons are nearly fully developed, to determine if the reduction in septohippocampal cholinergic projection neurons seen in adult IL-2 KO mice was present earlier in postnatal development (e.g., at weaning) and prior to the onset of the earliest signs of autoimmune disease (e.g., splenomegaly, lymphadenopathy). Contrary to our hypothesis, we found that the number of MS/vDB cholinergic cell bodies did not differ between IL-2 KO and wild-type littermates at P21 (Beck et al., 2004). Thus, together these data indicate that the loss of cholinergic neurons that occurs between P21 and adulthood (8-12 weeks) suggests an alternate hypothesis; neurodegenerative processes may be operative in the brain of IL-2 KO mice. Since IL-2 is an important factor in immune physiology, one possible mechanism behind these neurodegenerative processes is immune dysregulation caused by the absence of IL-2. IL-2-deficiency in mice leads to generalized systemic autoimmune disease in adult mice that may affect multiple organs in the periphery, most notably the intestines and the kidneys (Horak, 1995). The autoimmune effects in IL-2 KO mice involving peripheral organs are mediated largely by infiltrating T cells. In the colon, for example, adult IL-2 KO mice develop chronic inflammatory bowel disease with features common to inflammatory ulcerative colitis in humans, where the lamina propria is infiltrated with activated T cells responsible for the development of this inflammatory disease (Ma et al., 1995). In addition, there is a disruption of immune homeostasis that is evidenced by changes in the gene expression of several Thl, Th2, and various proinflammatory

PAGE 55

47 cytokines in this organ (Autenrieth et al., 1997; Meijssen et al., 1998). Moreover, these cytokine changes and the onset of inflammatory bowel disease are preceded by increased gene expression of IL-15, which shares the same signal transducing receptor subunits with IL-2 (Meijssen et al., 1998). Thus, it is possible that the immune dysregulation in the brain of IL-2 KO mice may be induced by activated T cells and/or proinflammatory cytokines (e.g., IL-1, TNFa, IL-6) from the periphery crossing the blood-brain-barrier (BBB). By contrast, IL-2 may lead to neuroimmunological changes that do not involve peripheral immune cells. Rather than peripheral T cells and serum cytokines entering the brain, an alternative hypothesis that may account for the hippocampal differences observed in P21 versus adult IL-2 KO mice may be that the absence of IL-2 reduces the trophic support of hippocampal neurons as a result of dysregulation of other brainderived cytokines. Thus, loss of IL-2 in the brain could in turn modify the normal neuroimmunological status of the brain by modifying the normal expression of brain cytokines such as IL-15. Since IL-2 can modify the release of certain cytokines from lymphoid cells (Lauwerys et al., 2000; McDyer et al., 2002), similar actions could occur in brain cells that produce cytokines (e.g., microglia, astrocytes). Alterations in the production of brain cytokines important in normal brain physiology could alter the integrity of hippocampal neurons by decreasing levels of classic neurotrophins and/or neurotrophic cytokines on the one hand, or elicit inflammatory-like neurodegenerative processes within the brain on the other. The present study therefore sought to test the hypothesis that IL-2 gene deletion results in neuroimmunological changes in the hippocampus by examining the possible

PAGE 56

48 outcomes described above. We compared the hippocampi of adult IL-2 KO mice and wild-type littermates at 8-12 weeks of age, the age where differences in hippocampal cytoarchitecture and behavior have been found previously (Beck et al., 2004; Beck et al., 2002; Cushman et al., 2004; Petitto et al., 1999), for differences in several measures of neuroimmunological status. First, the groups were assessed for differences in the number of CD3* T lymphocytes and activated microglial cells (as measured by MHC-II positivity) in the hippocampus. IL-15 is also expressed in the brain (Hanisch et al., 1997a; Lee et al., 1996), and this cytokine is known to have both proinflammatory and anti-inflammatory effects, potent anti-apoptotic, and T cell chemoattractant properties (Wilkinson and Liew, 1995). Because IL-1 5 uses the IL-2/15Rp and Yc subunits that are enriched in the neuronal cell layers of the hippocampus (Petitto and Huang, 2001), and may modulate microglial cell function and T cell chemoattraction, a second aim of this study was to test the hypothesis that IL-15 is elevated in the hippocampus of IL-2 KO mice. In addition, since changes in both IL-2 and IL-15 may modify levels of various c>1okines in other tissues and physiological contexts, exploratory testing was performed to determine if IL-2 KO mice have increased levels of proinflammatory cytokines in the hippocampus relative to wild-type mice (also, compared to serum levels). Furthermore, as recent evidence indicates that elevation of inflammatory cytokines such as IL-6 may impair hippocampal neurogenesis (Monje et al., 2003; Vallieres et al., 2002), a third aim of this study was to test the hypothesis that the postulated changes in neuroimmunological status would be associated with reductions in neurogenesis of neurons in the dentate gyrus (DG) of IL-2 KO mice.

PAGE 57

49 Materials and Methods Animals and Genotyping Mice used in these experiments were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2 heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the NIH repository at Jackson Laboratories, were derived from ten generations of backcrossing onto the C57BL/6 background. Mice were housed under specific pathogenfree conditions. Animals used in these experiments were 8-12 weeks of age. Independent animals were used for the assessment of CD3^ T cells and MHC 11"^ microglial cells in the hippocampus, the determinations of hippocampal versus serum cytokine levels, and assessments of neurogenesis in the dentate gyrus. Specific numbers utilized are reported at the beginning of the description of each method. The genotypes of all mice were determined by the polymerase chain reaction (PCR). PCR reactions were performed using 25 \i\ total reaction volume containing 1 \jlM each of forward and reverse primers, 0.1 \ig genomic DNA, 0.2 mM of each dNTP, 0.3 |j,l Taq DNA polymerase, and amplified using a thermal cycler with a heated evaporation cover (Ericomp). The cycling parameters were hot start 95 C (3min), denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final extension step of 4 min. Thirty cycles were used for these experiments. The 5' and 3' primers for the IL-2 KO (500 bp knockout band amplified) were 5'TCGAATCGCCAATGACAAGACGCT-3' and 5'GTAGGTGGAAATTCTAGCATCATCC-3'. The 5' and 3 'primers for the IL-2 wild type (324 bp wild type band amplified) were 5'-

PAGE 58

50 CTAGGCCACAGAATTGAAAGATCT-3' and 5'GTAGGTGGAAAATTCTAGCATCATCC-3'. CD3^ T cells and MHC 11^ Microglia Immunohistochemistry For quantitative assessment of autoimmunity, three IL-2 KO and three wild-type brains were processed for MHC II (an activated microglial marker) and CD3 (a pan T cell marker) immunohistochemistry. Each animal was anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose of 0.015 ml injection cocktail/g body weight and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS). The brains and spleens were removed and fixed overnight in 4% paraformaldehyde, followed by overnight equilibration in 30% sucrose cryoprotective solution, and then were snap frozen in isopentane (-80C) for storage. The spleens were weighed to assay for relative splenomegaly of IL-2 KO vs. wild-type mice The brains were equilibrated to -20C prior to cryostat sectioning into 50 ^m slices in the coronal plane, collected into individual wells of polystyrene 24well plates (NUNC 1 147), and stored free-floating at 4C in PBS for histochemistry. Every third section was processed for MHC II or CDS immunohistochemistry. Free-floating 50-|im sections were incubated for 20 minutes in 1% hydrogen peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then washed and blocked for 1 hr in 200 ^1/well 3% normal goat serum (NGS). After this incubation, the sections were incubated overnight in the primary antibody, rat anti mouse CD3 (BD PharMingen; 1 :500 in PBS with 0.3% Triton X-100 and 1% NGS) or rat anti mouse MHC II (BD PharMingen; 1 :500 in PBS with 0.3% Triton X-100 and 1% NGS). The next day, the sections were washed and incubated overnight in the secondary

PAGE 59

51 antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-lOO and 1% NGS). The sections were then washed and incubated in ExtrAvidin (Sigma E-2886; 1 : 1000 in PBS) for 2 hrs. The sections were developed in 0.5 mg/ml 3,3'-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped. Preparation of Serum and Brain Tissue for Cytokine Analysis Hippocampal homogenates were analyzed from eight IL-2 KO and nine wildtype mice to measure cytokine levels in the hippocampus. From these subject groups, serum was collected from a smaller subset of animals (five IL-2 KO and seven wild-type mice) for comparative analysis of brain vs. peripheral cytokine levels. Animals used for characterization of endogenous cytokine levels were anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose of 0.01 5 ml injection cocktail/g body weight. Whole blood was collected by puncturing the right atrium of the heart and inserting heparanized micro-hematocrit capillary tubes (Fisher Scientific). The animals were then saline perfused, but were not post-fixed in paraformaldehyde. The whole blood was centrifuged in Microtainer Brand serum separator tubes (Becton Dickinson) at 5,000 rpm for 10 minutes to isolate serum and the serum was stored at -80 C until used for Luminex analysis. The brains were removed, snap frozen, and then allowed to equilibrate to -20 C. The brains were sectioned on a cryostat at -20-22^ C at 400 [im thickness and the hippocampi were dissected with a 0.75 mm micropunch on a -20 C freezing platform. The dissected tissue was weighed on a microgram scale, and then transferred to 25 yd of homogenizing

PAGE 60

52 solution (50 mM Na/Na2 and 0.2% TX-lOO in H2O with Anti-protease Complete TM cocktail (Boehringer)) per mg of wet weight tissue. The tissue was sonicated in the homogenizing solution for 30 sec on ice and centrifuged at 16,000 g for 15 min at 4" C. The supernatant was collected and stored at -20 C for Luminex analysis. Multiplex Microsphere Cytokine Analysis Commercial kits, Lincoplex mouse cytokine (Linco, Research, Inc) and a Luminex 100 LabMAP system (Upstate Biotechnology), were used in attempt to measure a number of cytokines in the hippocampus and in the serum of IL-2 KO and wild-type mice. Assays were performed according to the manufacturer's instructions, and cytokine concentrations were calculated using the Softmax program and the linear range on the standard curve (3.2-10,000 pg/ml). Altogether, we attempted to detect a total of twentytwo different cytokines and chemokines from the serum and brain homogenates of these animals. In the serum, there were detectable levels of IL-6, IL-13, kerotinocyte-derived chemokines (KC), granulocyte-colony stimulating factor (G-CSF), and macrophage inflammatory protein1 alpha (MlP-la). In the brain, there were detectable levels of IL7, IL-9, IL-12, IL-15, interferon-gamma inducible protein of 10 kD (IP10), and monocyte chemoattractant protein1 (MCP-1 ). Thus, only those cytokines and chemokines detected were subjected to statistical analyses. The remainder of the cytokines and chemokines tested could not be detected in either the semm or the brain (IFN-y; TNF-a; IL-la, IL-IP; IL-2, IL-4, IL-5, IL-10, IL-17, GM-CSF, and RANTES). Labeling Neurogenesis with BrdU Twelve IL-2 KO (six male; six female) and eleven wild-type (five male; six female) mice were used to assay for differences in neurogenesis. The procedure for

PAGE 61

53 labeling of neurogenesis and subsequent immunostaining in the mouse hippocampus has been adapted from (Lee et al., 2002). Briefly, the mice were given five intraperitoneal injections of BrdU (50 mg/kg of body weight) over the course of 3 days. The day following the last injection, the mice were sacrificed and perfused with 0.9% saline followed by 4% paraformaldehyde in PBS as described previously. The BrdU-incorporated brains were equilibrated to -20 C and cryostat-sectioned at 50 |im in the coronal plane. They were collected into individual wells of polystyrene 24-well plates (NUNC 1 147), and used for free-floating immunohistochemistry. The sections were then washed twice in PBS and then the DNA was denatured by a 30 minute incubation with 2 N HCl to allow binding of the antibody to the BrdU in the singlestranded DNA. The acid was neutralized with a 0. 1 M borate buffer (pH 8.5) wash, followed by several washes in PBS. Afterwards, the sections were blocked for 1 hr in 3% normal goat serum (NGS). The sections were then incubated overnight in the primary antibodies, rat monoclonal anti-BrdU (Serotec; 1:400 in PBS with 0.3% Triton X-100 and 1% NGS) and either the neuronal marker mouse monoclonal anti-tubulin p III isoform (Chemicon; 1 :200 in PBS with 0.3% Triton X-100 and 1% NGS), the astroglial marker rabbit anti-glial fibrillary acidic protein (GFAP; Chemicon; 1 : 1,000 in PBS with 0.3% Triton X-100 and 1% NGS) or the oligodendrocyte marker mouse anti-2'3 '-cyclic nucleotide 3'-phosphohydrolase (CNPase; Chemicon; 1 :200 in PBS with 0.3% Triton X100 and 1% NGS). The next day, the sections were washed twice in PBS and incubated for 2 hr in the dark with the secondary antibodies, goat anti-rat IgG (H+L) conjugated with Alexa Fluor-488 (green; Molecular Probes; 1 :400 in PBS with 0.3% TX-lOO and 1% NGS) and goat anti-mouse IgG (highly cross-absorbed H+L) conjugated with Alexa

PAGE 62

54 Fluor 568 (red; Molecular Probes; 1 :400 in PBS with 0.3% TX-1 00 and 1% NGS) or goat anti -rabbit IgG (H+L) conjugated with Alexa Fluor 350 (blue; Molecular Probes; 1 :400 in PBS with 0.3% TX-lOO and 1% NGS). The sections were then washed twice in PBS, placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes, and coverslipped. The sections were imaged using a Bio-Rad 1024 ES confocal microscope and only cells which showed colocalized staining through five consecutive l-|a.m planes were considered to be double-labeled. The IP and SP-GCL area (mm') were measured at 20x magnification using the MCID 5.1 software, a CCD High Resolution Sony camera, and a Zeiss Axioplan 2 microscope. The data were reported as a density of all double-labeled cells counted from five sections per animal divided by the total area measured Results Assessment of CD3^ T Cells and MHC 11^ Activated Microglia in the Hippocampus No CD3"' T cells were detected in the hippocampi of either 8-12-week-old IL-2 KO or wild-type mouse brains, and only an occasional MHC if microglial cell was detected (e.g., approximately one every other section) in both groups. By contrast, both activated MHC if microglia and CD3"^ T cells were readily detectable in positive control slices (sections of the axotomized FMN of wild-type C57BL/6 mice; Petitto et al., 2003) demonstrating that the immunohistochemistry procedure utilized was effective for labeling both markers. Thus, no differences in T cells or activated microglial cells were found between the subject groups. Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice As depicted in Figure 4-1, in the hippocampus, there were significantly increased levels of IL-12 (increased -57%; F(l,15)=9.174, p=0.008) and IL-15 (increased -38%;

PAGE 63

55 F(l,15)=6.105, p=0.026) in the IL-2 KO compared to wild-type mice. Figure 4-1 also shows that IL-2 KO mice had increased levels of the chemokines, IP10 (increased -63%; F(l,15)=4.747, p=0.046) and MCP-1 (increased -46%; F(l,15)=5.218, p=0.039). Although detectable levels of IL-7 and IL-9 were found in hippocampus, they did not differ between the subject groups. Enhanced Cytokine Concentrations in the Hippocampus IL-12 WJd-Type IL-2 Khockout IL-16 IP-10 Oyftokine MCP-1 Fi gure 4-1. Protein levels of several cytokines and chemokines are elevated in the IL-2 knockout mice compared to wild-type brain. These include IL-12, IL-15, IP-10, and MCP-1. *p<0.05; **p<0.01. Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-t>'pe Mice As can been seen in Figure 4-2, the IL-2 KO mice exhibited marked elevation in serum IL-6 (increased -224%; F(l,10)=8.077, p=0.017), and serum MlP-la serum concentration (increased -327%; F(l,10)=21.538, p=0.001) compared to wild-types. Although detectable levels of IL-13, G-CSF, and KC were detectable in the serum, levels of these cytokines did not differ between the subject groups.

PAGE 64

56 A) Enhanced Cytokine Concentrations In the Serum of IL-2 KO Animals B) Enhanced Cytokine Concentrations in the Semmot IL-2 KO Animals E 5 9i 1 M O s ICC 140 E 120 5 100 E g 80 60 40 20 • 0 MIP-1 alpha 3 WilS-Type I IL 2 Knockout ] W*i)-Tie I IL 2 Knockout Figure 4-2. Relative cytokine profile of IL-2 knockout mouse serum does not match brain profile. A) IL-6 is increased in IL-2 knockout mice compared to wild-type; and B) the chemokine MIP-1 a is higher in IL-2 knockout mice than wild-type. *p<0.05; **p<0.01. Alterations in Neurogenesis Although there was not a significant effect of group on neurogenesis in either the IP-GCL or SP-GCL, as depicted in Figure 4-3 and 4-4 respectively, there was a significant group by gender interaction in both the IP-GCL (F(1,19)=4.7I, p=0.043) and SP-GCL (F(l,19)=6.43, p=0.02). Fisher's least significant difference post hoc analysis test confirmed a difference between male IL-2 KO and wild-type mice in the IP-GCL (p=0.025) and SP-GCL (p=0.014), but not between female groups in either region. There was no significant effect of group or group by gender interaction in cells around either the IP-GCL or SP-GCL labeled with the oligodendrocyte marker, CNPase (data not shown). Similarly, no significant effect of group or group by gender interaction was evident in either the IP or SP-GCL in cells labeled with the astrocyte marker, GFAP (data not shown).

PAGE 65

57 Neurogenesis in the Infrapyramidal Granule Cell Layer 500 Figure 4-3. There is no effect of group on neurogenesis in the infrapyramidal granule cell layer, but there is a group by gender interaction between IL-2 knockout and wildtype mice (p=0.043). This effect appears to be attributable to differences in the male mice (Fisher least significant difference test, p=0.025). Neurogenesis in the Suprapyramidal Granule Cell Layer 500 4C0 E E 2 300 s 200 1X1 WildType IL-2 Knockout I Female Male Figure 4-4. There is no effect of group on neurogenesis in the suprapyramidal granule cell layer, but there is a group by gender interaction betv-'een IL-2 knockout and wild-type mice (p=0.02). This effect appears to be attributable to differences in the male mice (Fisher least significant difference test, p=0.014).

PAGE 66

58 Discussion The data presented here show that IL-2 KO and wild-type httermates exhibit differences in several measures of neuroimmunological status in the hippocampus. In order to access the brain parenchyma, T cells require activation markers to cross the BBB (Hickey et al., 1991). We have previously reported leukocytes were not detectable in the cresyl violet stained hippocampal sections from IL-2 KO mice (Petitto et al., 1999), however, we recognized that it is difficult to reliably detect small numbers of peripheral leukocytes in the brain without cell-type specific stains. Although we were therefore not expecting to see substantial numbers of T cells in the IL-2 KO brain, we wanted to determine if small numbers of autoimmune T lymphocytes were present that could initiate neuroimmunological alterations in the hippocampus. The hippocampi of IL-2 KO mice were devoid of T cells, despite the fact that the majority of peripheral T cells of IL-2 KO mice express activation markers such as CD69 (Sakai et al., 1995; Schopke et al., 1991), which are thought to enhance their ability to cross the BBB. Microglia are indigenous antigen presenting cells (Hickey and Kimura, 1988; Streit et al., 1988). Contact with T cells can induce microglia to exhibit characteristics of antigen presenting cells, and microglia also have the ability to activate T cells (Aloisi et al., 2000). There was, however, no evidence of increased numbers of activated microglia in the hippocampus of IL-2 KO mice. This observation is consistent with our previous finding in C57BL/65c/i/-IL-2 KO (mice without mature T and B cells), which were devoid of T cells in the axotomized facial motor nucleus and had levels of axotomy-induced activated microglia that did not differ from wild-type mice (Petitto et al., 2003). Thus, at the cellular level, the hippocampus of IL-2 KO mice did not show signs of autoimmune disease.

PAGE 67

59 Cytokines can enter brain via specific transport mechanisms and through the circumventricular organs. In this study, we measured levels of the various cytokines of interest in the serum to determine if levels found in the hippocampus correlated with those found in the serum. If a particular inflammatory cytokine (e.g., TNFa) had been found to be significantly elevated in the serum and the hippocampus of IL-2 KO mice, it would have suggested the possibility that peripheral immune activation associated with autoimmunity (Schimpl et al., 2002) could account for the presence of the cytokine in the hippocampus (although increased gene transcription and translation in both the periphery and the brain could not be ruled out). Although there were significantly increased levels of IL-6 and MlPla in the circulation of lL-2 KO mice, we did not detect either of these proteins in the brain. IL-6 is a proinflammatory cytokine that was of particular interest to us because of its actions in the hippocampus, including effects on neurogenesis (Monje et al., 2003; Vallieres et al., 2002). Although levels of IL-6 were markedly increased in the serum, though not measurable in the hippocampus at the tissue homogenate concentrations used, the unlikely possibility remains that IL-6 could have entered the brain from the circulation and had functional consequences at concentrations below the limits of detection of the assay method. Nonetheless, the most parsimonious explanation is that cytokines from the peripheral circulation of IL-2 KO mice are not the source of the cytokine alterations found in the hippocampus. Thus, peripheral cytokine dysregulation associated with autoimmunity in IL-2 KO does not appear to be associated with their hippocampal pathology. Our data indicate that loss of IL-2 in the brain results in changes in the production of several other brain cytokines. Consistent with our hypothesis, IL-15 concentrations

PAGE 68

60 were increased in the hippocampus of IL-2 KO mice. IL-15 is structurally related to IL-2 and shares the same (5 and yc signal transducing receptor subunits with the IL-2 receptor (Giri et al., 1995). IL-15 also shares and opposes several physiological functions of IL-2 in the peripheral immune system (Waldmann, 2002; Waldmann et al., 2001). IL-15 and its heterotrimeric receptor are constitutively expressed in various regions of the adult mouse brain and can be detected in microglial cultures (Hanisch et al., 1997a), astrocytes (Lee et al., 1996), and possibly neurons (Maslinska, 2001; Satoh et al., 1998). As noted earlier, increased IL-15 gene expression precedes the inflammatory cytokine changes and onset of inflammatory bowel disease in IL-2 KO mice (Meijssen et al., 1998). It also induces the onset of autoimmunity in thyroiditis (Kaiser et al., 2002). Thus, IL-15 could trigger proinflammatory cytokine-like processes in the hippocampus, including the elevations in IL-12 that were found in IL-2 KO mice in this study. IL-12-driven Thl responses are involved in inflammation (e.g., colonic) in IL-2 KO mice (Ludviksson et al., 1997), and it has been implicated as an important effector in the pathogenesis of experimental autoimmune encephalomyelitis (EAE) (Adorini, 1999; Segal et al., 1998). Moreover, IL-15 can render cells resistant to the protective effects of TGpp (Campbell et al., 2001), a Th2 cytokine that appears to play a key role in dampening processes associated with peripheral autoimmune disease in IL-2 KO mice (Ludviksson et al., 1997). Thus, these actions of IL-15 suggest that it may be involved in the hippocampal pathology seen in IL-2 KO mice. It is noteworthy, however, that IL-15 has potent antiapoptotic properties (Lauwerys et al., 2000; Waldmann, 2002; Waldmann et al., 2001) that may oppose the pro-apoptotic effects of IL-2. We have recently found that loss of brain IL-2 in C57BL/65c?i/-IL-2 KOmice increased neuroregeneration in the axotomized

PAGE 69

61 facial motor nucleus (Petitto et al., 2003). It is interesting to speculate that elevated IL-15 levels could contribute to the increased motor neuronal survival in those mice. Therefore an alternative interpretation is that increased IL-15 in the hippocampus of IL-2 KO mice could be a compensatory response to counteract neuroregenerative changes in the hippocampus. The increased levels of MCP-1 and IP10 may possibly be induced by increased IL-15 in the hippocampus (Badolato et al., 1997). Previous studies have demonstrated that IP10 expression can be detected in lipopolysaccharide (LPS)-treated microglial and astroglial cultures and in situ hybridization of LPS-treated rat brains (Ren et al., 1998). Similarly, MCP-1 can also be induced by the addition of the pro-inflammatory cytokine, TNF-a, or the anti-inflammatory cytokine, TGF-P, in astrocytes (Hurwitz et al., 1995) and microglia (Meda et al., 1996). IP10 and MCP-1 are also chemoattractant factors for T lymphocyte infiltration into the CNS (Babcock et al., 2003; Dufour et al., 2002), and may attract activated microglia in vitro (Cross and Woodroofe, 1999). In spite of the published data linking IP10 and MCP-1 to T cell and microglial chemotaxis, we were unable to detect either T cells or increased numbers of activated microglia in the hippocampus of IL-2 KO mice. Further studies are necessary to determine the functional significance of the increased production of IP-10 and/or MCP-1 in the hippocampus of IL-2 KO mice. Finally, in addition to its immune activating effects as a Thl cytokine, IL-2 is also known to have important critical negative regulatory functions by stimulating Th2 lymphocytes to produce TGFp (Ludviksson et al., 1997), which down-regulates the ability of antigen presenting cells to produce IL-12, a powerful activator of Thl cell

PAGE 70

62 development and inflammation. Thus, the increased levels of IL-12 found in the hippocampus could be the secondary to the loss IL-2. Increased levels of inflammatory cytokines such as IL-6 may impair hippocampal neurogenesis (Monje et al., 2003; Vallieres et al., 2002); thus, we hypothesized that the alterations in neuroimmunological status would correlate with decreased neurogenesis in the DG of IL-2 KO mice. Against our hypothesis, however, no detectable levels proinflammatory cytokines like IL-6 (though IL-12 can mediate inflammatory responses) or effects of group were apparent in hippocampal neurogenesis. However, a significant group by gender interaction was detectable. Interestingly, this effect appeared to be due to variations in the male mice, but not the females. We previously noted a trend which suggested that IL-2 gene deletion may protect from experimental autoimmune encephalomyelitis (EAE), though we did not statistically analyze the data (Petitto et al., 2000). In that study, three out of the seven male mice utilized developed some symptoms of the disease, whereas none of the six females did. One hypothesis to explain this group by gender interaction is that there may be an interplay between IL-2 and the sex hormones. Sex steroids like estrogen, for example, have been linked to neurogenesis (Gould et al., 2000; Perez-Martin et al., 2003). Further, evidence that IL-2 can regulate sex hormone expression comes from studies of Leydig cell cultures where IL-2 was shown to inliibit steroidogenesis (Guo et al., 1990). Hormonal differences influenced by IL-2 may give a possible explanation of why males differ from females, but it does not suggest a mechanism of why IL-2 KO mice appear to be protected from impaired hippocampal neurogenesis. As noted above, absence of IL-2 can enhance neuroregenerative properties in the axotomized facial motor

PAGE 71

63 nucleus and we speculate an involvement of IL-15. Though IL-15 has not been linked to neurogenesis, its antiapoptotic properties on immune cells are well-studied (Lauwerys et al., 2000; Waldmann, 2002; Waldmann et al., 2001). If these antiapoptotic actions of IL15 can also promote survival of neurons, then this may provide a hypothesis of why the IL-2 KO male mice have higher levels of neurogenesis than the wild-types. Some cytokines are capable of promoting the development of neural stem cells (Rozental et al., 1995; Shah et al., 1996; Wong et al., 2004), and this may account for the how they influence neurogenesis. The actual mechanism whereby IL-2, sex hormones, and neurogenesis may interact is as of yet undefined and requires fiature study. In summary, T cells and peripheral cytokines do not appear to enter into the hippocampi of IL-2 KO mice, which does not support the hypothesis that the CNS alterations previously seen in IL-2 KO mice are due to peripheral autoimmunity. Other potential immune indicators of autoimmunity (e.g., deposition of autoantibodies in the brain) were not addressed in this dissertation and require more research. Genetic deletion of IL-2 may, however, alter the neuroimmunological status of the mouse hippocampus through a dysregulation of cytokines produced by CNS cells (e.g., microglia, astroglia). Further studies will be required to determine how these changes impact hippocampal cytoarchitecture and function in IL-2 KO mice.

PAGE 72

CHAPTER 5 GENERAL DISCUSSION Summary of the Overall Findings The overall purpose of this dissertation research was to investigate the impact of IL-2 on the septohippocampal system by observing changes in the basal physiology and structure of IL-2 KO mice brains vs. their wild-type littermates. During these experiments, we chose not to surgically, chemically, or in any other way experimentally manipulate these animals beyond the genetic knockout of IL-2, so that the effects that we detected could be attributed to loss of IL-2 and not to confounding experimental techniques (e.g., disruption of the BBB). We understand, however, that some experimental manipulations could be useful in the long term and discuss the topic more in the "Caveats and Future Directions" section of towards the end of this chapter. In the first study in Chapter 2, we determined that IL-2 KO mice suffered a significant 26% loss of cholinergic neurons in the MS/vDB relative to wild-type littermates. This loss of cholinergic somata was not reflected by a similar loss of cholinergic fiber density in the hippocampal projection fields of the septal cholinergic neurons. Moreover, the deficits observed were not a general effect on all cholinergic neurons of the brain, as the cholinergic neurons of the striatum in the IL-2 KO animals studied did not appear to be affected. There was, however, a reduction in the distance across the IP, but not SP, GCL of the dentate gyrus. In Chapter 3, we extended upon these findings by showing that the loss of cholinergic neurons appeared to occur later in development or during adulthood as a 64

PAGE 73

65 potential failure of these neurons to survive, since 21 -day-old IL-2 KO animals did not show signs of a loss of cholinergic neurons of the MS/vDB. Furthermore, the neuronal deficiencies appeared to be selective to the cholinergic neurons of the MS/vDB, as there were no similar decreases of GABAergic neurons in IL-2 KO animals. In the hippocampus, consistent with the reductions in length across the IP-GCL of IL-2 KO compared to wild-type mice, the stereologically estimated cell count of Nissl-stained neuronal somata of IL-2 KO mice was significantly lower than wild-type littermates. Also, the hippocampal neurotrophin levels of BDNF and NGF were significantly decreased and increased, respectively. Finally, in Chapter 4, the cytokine profile of the IL-2 KO hippocampus was altered with increases in IL-15 and IL-12 and the chemokines MCP-1 and LP10. These experiments confirmed our hypothesis that loss of IL-2 would result in increased levels of brain IL-15 production. Furthermore, in spite of the known roles of these cytokines and chemokines in T cell trafficking and modulation during brain insult, we were unable to label for elevated levels of the T cells or activated microglia in the hippocampus. The above cytokine profile did not match the serum cytokine levels, suggesting that the changes in cytokines in the hippocampus were likely due to changes in their production in the CNS. Finally, though there was no group alteration in adult GCL neurogenesis, there was a group by gender interaction that appeared to be attributable to the male mice. Implications This series of studies is the first to demonstrate that endogenous levels of IL-2 may be an important factor in the late development or survival of neurons in the CNS. This impact on CNS neurons involved not only cell number, but also alterations in levels of trophic factors and brain cytokines. Other studies involving IL-2 and the CNS, to date,

PAGE 74

66 have depended on in vitro models and/or exogenous administration of IL-2 into theCNS in vivo. Whereas studies such as these can yield valuable information, in vitro models often fail to mimic the complexity of an in vivo system (e.g., lack of glia, different organization of neurons, etc.) and the manipulated in vivo models are subject to experimental design complications (e.g., disruption of the BBB, determination of physiologically meaningful doses of exogenously administered cytokines, etc.). We would be remiss, however, to claim that the IL-2 KO model was completely without complications itself (e.g., autoimmunity such as inflammatory bowel disease), but we feel that the most informative way to address a hypothesis is to examine it in multiple different ways. Thus, taken together with the previous studies in the literature, this dissertation reveals an interesting relationship between IL-2 and the septohippocampal system. The goal of this study was to elicit whether IL-2 KO impacted the septohippocampal system, but not to design a treatment or study a clinical disorder. Thus, at its core, this dissertation was basic research. Whereas the pursuit of knowledge is a noble goal, invariably, a simple question often finds itself in the front of our minds: how is this new knowledge clinically relevant and how can we use it? Already, IL-2 is being used clinically to treat, or being studied as a treatment for, a number of disorders, including cancer (Atkins et al., 1999; Davis and Gillies, 2003; Fyfe et al., 1995; Guirguis et al., 2002; Parkinson et al., 1990; Rosenberg et al., 1985; Rosenberg et al., 1994), and HIV infection (Armstrong and Kazanjian, 2001; Mitsuyasu, 2001; Smith, 2001). The aforementioned treatments are not, however, completely without side effects. In a landmark study, systemic IL-2 treatment for cancer patients

PAGE 75

67 was shown to induce dose and time-related behavioral and cognitive changes in 22 out of 44 subjects, including spatial and temporal disorientation (Denicoff et al., 1987). Others have also noted cognitive changes in patients undergoing IL-2-treatment (Capuron et al., 2000; Capuron et al., 2001b; Walker et al., 1997). Of particular relevance to this dissertation, IL-2 immunotherapy can induce cognitive changes, including spatial memory deficits and poor task planning (which may be partially dependent on spatial memory) as early as five days into therapy (Capuron et al., 2001a). Capuron's study utilized neuropsychological batteries to assess these cognitive dysfunctions, but did not investigate the physiological mechanism underlying them. One hypothesis to explain at least part of the cognitive changes observed is that the chronic IL-2 treatment used on these patients may affect the septohippocampal system in a similar way that loss of IL-2 causes changes in IL-2 KO animals. Deletion of IL-2 having a similar physiological effect as administration of exogenous IL-2 may seem counterintuitive; however, in at least one population of neurons (i.e., the septal neurons), IL-2 has potent biphasic modulatory actions on ACh release with low sub-pM IL-2 concentrations enhancing and higher nM concentrations inhibiting release (Hanisch et al., 1993; Seto et al., 1997). Thus, under the proposed hypothesis, a loss of IL-2 could lead to a reduction in the stimulatory action on cholinergic neurons, whereas an overabundance of exogenous IL-2 could inhibit release of ACh. Of course, more investigation would be necessary to clarify that hypothesis. Caveats and Future Directions In Study 1, we examined the effects of IL-2 KO on the cholinergic septohippocampal neurons. Many of our findings were extended upon in Study 2. However, we hypothesized that the lack of any change in the cholinergic fiber density in

PAGE 76

68 the hippocampus was hkely due to compensatory sprouting, but did not overtly assay for any alterations. Thus, to address this issue further, a study utilizing a stereotaxic injection of an anterograde tracer, such as Phaseolus vulgaris-leucoagglutinin (PHAL), into the MS/vDB area would be necessary. The septohippocampal cholinergic neurites would need to be labeled with a marker for AChE and the length and branching of these double-labeled axons would be characterized. If the septohippocampal cholinergic neurons do undergo compensatory sprouting in IL-2 KO mice, we hypothesize that a detectable increase in branching should occur in those animals vs. wild-type littermates. Though the majority of cholinergic neurons in the MS/vDB project to the hippocampus (Schwegler et al., 1996b), a small percentage also project to other areas, including the mediodorsal nucleus of the thalamus (Gritti et al., 1998), parietal, cingulate (Gritti et al., 1997), and entorhinal cortices (Alonso and Kohler, 1984). Thus, to accurately claim that the cholinergic neuronal losses observed in the MS/vDB of the IL-2 KO mouse do indeed project to the hippocampus, a study involving a retrograde tracer like Fluoro-Gold injected in the hippocampal projection areas would be necessary. The relrogradely labeled somata of the MS/vDB would need to be double-labeled with ChAT to identify neurons of the cholinergic phenotype and the cell number would be estimated with stereology. In Study 2, three-week-old IL-2 KO animals did not differ from wild-type littermates in cholinergic cell number. To expand upon this finding, the brains from animals should be examined at one-week intervals from three-week-old mice to adult eight-week-old mice. This will allow us to establish a time course of cholinergic neuronal loss. Furthermore, we would be interested in investigating the nature of the cell

PAGE 77

69 loss. A DNA fragmentation assay, like terminal transferase-mediated dUTP nick-endlabeling (TUNEL), could be used to detect apoptotic cell death and cell death by necrosis could be determined by a simple trypan blue dye exclusion method. Both assays would require immunohistochemical labeling with ChAT to identify the cholinergic somata. In Study 3, the cytokine and chemokine profile of the IL-2 KO hippocampus was altered compared to wild-type mice. Though we did speculate on the cell types that might produce these cytokines, we did not investigate it further. To identify the potential sources of these cytokines, we could use in situ hybridization to label the mRNA and label the individual cell types with immunohistochemistry. We could expand upon this further by utilizing real time PCR techniques as a semi-quantitative assay of different levels of mRNA expression in homogenates from the hippocampi of IL-2 KO and wildtype mice. These two experiments allow us to identify the cells producing the cytokines and compare the relative levels of mRJMA produced for each cytokine between IL-2 KO and wild-type mice. Several other considerations and questions arise from the data generated from the experiments of this dissertation. First, there is possibility that the reduced levels of ChAT"^ immunostaining may be due to a loss of the cholinergic phenotype (e.g., inability to produce ACh) rather than cell death. Alterations in the activity of ACh in the hippocampus (e.g., addition of receptor agonists or antagonists) are sufficient to alter leaming and memory (for a review see Gold, 2003). Thus, the behavioral deficits observed in the IL-2 KO animals may not be due to cholinergic neuronal loss. Although the staining intensity was uniform (e.g., consistent between and within groups), to test this further, IL-2 KO and wild-type littermate brains should be Nissl-stained to label and

PAGE 78

70 quantify MS/vDB neurons in this brain region. Animals with neuronal loss, should exhibit reduction in total neuronal counts, whereas the neuronal counts would not differ significantly in mice with a loss of phenotype. Also, considering the difficulty of detection of endogenous IL-2 in the normal CNS, what is the source of the IL-2 in the wild-type brain? Taken together with the unknown non-saturable transport mechanism that allows IL-2 to cross the BBB (Waguespack et al., 1994), and the numerous studies showing cognitive effects of peripherally administered exogenous IL-2 (Capuron et al., 2001a; Denicoff et al., 1987; Lacosta et al., 1999; Walker et al., 1997), an argument can be made that some of the endogenous IL-2 in the CNS could be from the periphery, particularly during development when the BBB is not completely formed. In addition, though we did address the autoimmunity issue of IL-2 KO mice somewhat, we did not completely rule it out as a potential factor in the observed alterations. An experiment that could address both of the above issues would involve crossbreeding the IL-2 KO mice onto an immunodeficient background lacking functional lymphocytes (e.g., RAG-1 knockout mice). Next, an adoptive transfer of normal T lymphocytes from healthy IL-2 wild-type (i.e., non-RAG-1 KO) mice to young (i.e., less than 3-week-old) IL-2 KO/RAG-1 KO or IL-2 wild-type/RAG-1 KO littermate animals would establish a functional immune system, and at the same time restore a major source of peripheral IL-2. Also, another set of IL-2 KO/RAG-1 KO and IL-2 wild-type/RAG-1 KO mice should receive an adoptive transfer of IL-2 KO (i.e., non-RAG-1 KO) lymphocytes to mimic the autoimmune state of normal IL-2 KO mice. Finally, another group of IL-2 KO/RAG-1 KO and IL-2 wildtype/RAG-1 KO mice should receive a sham reconstitution, such that they maintain any

PAGE 79

71 normal levels (or lack thereof in IL-2 KO/RAG-1 KO mice) of endogenous brain IL-2 expression, but remain immunodeficient, thus lacking a peripheral source of IL-2. This experimental design would create a model • that eliminates the impact of peripheral autoimmunity in the IL-2 wild-type reconstituted IL-2 KO/RAG-1 KO mice; • in which the aforementioned animals are only lacking a CNS, but not peripheral, source of IL-2; • in which the IL-2 wild-type reconstituted IL-2 wild-type/RAG-1 KO mice would have a functional peripheral immune system and normal endogenous IL-2 expression in the brain; • in which the IL-2 KO reconstituted IL-2 KO/RAG-1 KO mice are similar in phenotype to the normal IL-2 KO mouse (i.e., peripheral autoimmunity); • in which the IL-2 KO reconstituted IL-2 wild-type/RAG-1 KO mice would have normal brain IL-2 expression, but autoimmunity caused by the peripheral IL-2 KO T lymphocytes; and • in which the sham reconstituted IL-2 KO/RAG-1 KO mice lack any source of IL-2 and the sham reconstituted wild-type/RAG-1 KO mice lack a peripheral source of IL-2, with neither of the two models succumbing to autoimmunity. Thus, if endogenous brain IL-2 is important in septohippocampal development and/or maintenance from the age of three-weeks to adulthood, then we should observe deficits in the septohippocampal system similar to those noted in this dissertation in all reconstitution and sham models above on an IL-2 KO/RAG-1 KO background. If the peripheral source of IL-2 is more important for septohippocampal physiology, then the structural and physiological alterations in this dissertation should be rescued in the above IL-2 wild-type reconstituted cases, but not IL-2 KO or sham reconstituted animals. If autoimmunity is the determining factor, then IL-2 KO reconstituted, but not sham reconstituted, animals will all exhibit deficits in the septohippocampal system regardless of background. This proposed experiment combining IL-2 KO strains, immunodeficient

PAGE 80

72 strains, and reconstitution allows us to control for several immunological and genetic factors without disruption of the BBB. Concluding Remarks In the peripheral immune system, the complex interplay and interactions of the various cytokines often have redundant, supportive, or even oppositional roles. This complexity allows for a system of compensatory and regulatory control of the immune response. Similar overlaps and checks and balances are also present in the CNS, so understanding how these intricate systems interact can prove daunting. Nevertheless, elucidating the relationship between the brain and immune system molecules may have profound clinical utility. The IL-2 KO mouse is a complicated model to study a cytokine brain interaction and, in this dissertation, we have attempted to simplify the model by not experimentally manipulating it. Our goal was to lay groundwork for future studies on this topic, whereby we hope to more fully understand the precise mechanisms by which IL-2 alters brain physiology.

PAGE 81

WORKS CITED LA Aarden, TK Brunner, JC Cerottini, J Dayer, AL de Week, CA Dinarello, G Di Sabato, JJ Farrar, I Gery, S Gillis, RE Handschumacher, CS Henney, MK Hoffmann, WJ Koopman, SM Krane, LB Lachman, I Lefkowitz, RI Mishell, SB Mizel, JJ Oppenheim, V Paetkau, J Plate, M Rollinghoff, D Rosenstreich, AS Rosenthal, LJ Rosenwasser, A Schimpl, HS Shin, PL Simon, KA Smith, H Wagner, JD Watson, E Wecker and DD Wood, 1 979. Revised nomenclature for antigen-nonspecific T cell proliferation and helper factors. J Immunol 123, 29282929. R Ader and N Cohen, 1975. Behaviorally conditioned immunosuppression. Psychosom Med 37, 333-340. R Ader, N Cohen and D Bovbjerg, 1982. Conditioned suppression of humoral immunity in the rat. J Comp Physiol Psychol 96, 517-521. R Ader, D Felten and N Cohen (Eds), 2001. Psychoneuroimmunology, Third Edition Edition. Academic Press, San Diego, CA. L Adorini, 1999. Interleukin-12, a key cytokine in Thl-mediated autoimmune diseases. Cell Mol Life Sci 55, 1610-1625. RF Alderson, AL Alterman, YA Barde and RM Lindsay, 1 990. Brain-derived neurotrophic factor increases survival and differentiated functions of rat septal cholinergic neurons in culture. Neuron 5, 297-306. F Aloisi, R De Simone, S Columba-Cabezas, G Penna and L Adorini, 2000. Functional maturation of adult mouse resting microglia into an APC is promoted by granulocyte-macrophage colony-stimulating factor and interaction with Thl cells. J Immunol 164, 1705-1712. A Alonso and C Kohler, 1 984. A study of the reciprocal connections between the septum and the entorhinal area using anterograde and retrograde axonal transport methods in the rat brain. J Comp Neurol 225, 327-343. R Alonso, I Chaudieu, J Diorio, A Krishnamurthy, R Quirion and P Boksa, 1993. Interleukin-2 modulates evoked release of [3H]dopamine in rat cultured mesencephalic cells. J Neurochem 61, 1284-1290. 73

PAGE 82

74 DM Araujo, PA Lapchak, B Collier and R Quirion, 1989. Localization of interleukin-2 immunoreactivity and interleukin-2 receptors in the rat brain: interaction with the cholinergic system. Brain Res 498, 257-266. Y Arimatsu and M Miyamoto, 1991 Survival-promoting effect of NGF on in vitro septohippocampal neurons with cholinergic and GABAergic phenotypes. Brain Res Dev Brain Res 58, 189-201. DM Armstrong, G Bruce, LB Hersh and FH Gage, 1987. Development of cholinergic neurons in the septal/diagonal band complex of the rat. Brain Res 433, 249-256. WS Armstrong and P Kazanjian, 2001 Use of cytokines in human immunodeficiency virus-infected patients: colony-stimulating factors, erythropoietin, and interleukin2. Clin Infect Dis 32, 766-773. MB Atkins, MT Lotze, JP Dutcher, RI Fisher, G Weiss, K Margolin, J Abrams, M Sznol, D Parkinson, M Hawkins, C Paradise, L Kunkel and SA Rosenberg, 1999. Highdose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 17, 21052116. IB Autenrieth, N Bucheler, E Bohn, G Heinze and 1 Horak, 1997. Cytokine mRNA expression in intestinal tissue of interleukin-2 deficient mice with bowel inflammation. Gut 41, 793-800. H Awatsuji, Y Furukawa, M Nakajima, S Furukawa and K Hayashi, 1993. Interleukin-2 as a neurotrophic factor for supporting the survival of neurons cultured from various regions of fetal rat brain. J Neurosci Res 35, 305-31 1. AA Babcock, WA Kuziel, S Rivest and T Owens, 2003. Chemokine expression by glial cells directs leukocytes to sites of axonal injury in the CNS. J Neurosci 23, 79227930. R Badolato, AN Ponzi, M Millesimo, LD Notarangelo and T Musso, 1997. hiterleukin-15 (IL-15) induces IL-8 and monocyte chemotactic protein 1 production in human monocytes. Blood 90, 2804-2809. WA Banks, AJ Kastin and EG Gutierrez, 1994. Penetration of iiiterleukin-6 across the murine blood-brain barrier. Neurosci Lett 179, 53-56. WA Banks, SA Farr and JE Morley, 2002. Entry of blood-borne cytokines into the central nervous system: effects on cognitive processes. Neuroimmunomodulation 10, 319-327. WA Banks, L Ortiz, SR Plotkin and AJ Kastin, 1991 Human interleukin (IL) 1 alpha, murine IL-1 alpha and murine IL-1 beta are transported from blood to brain in the mouse by a shared saturable mechanism. J Pharmacol Exp Ther 259, 988-996.

PAGE 83

75 MG Baxter, DJ Bucci, TJ Sobel, MJ Williams, LK Gorman and M Gallagher, 1996. Intact spatial leaming following lesions of basal forebrain cholinergic neurons. Neuroreport 7, 1417-1420. SA Bayer, 1 980. Development of the hippocampal region in the rat. I. Neurogenesis examined with 3H-thymidine autoradiography. J Comp Neurol 190, 8711 4. JF Bazan, 1992. Unraveling the structure of IL-2. Science 257, 410-413. RD Beck, Z Huang, GK Ha and JM Petitto, 2004. Altered Hippocampal Structure and Neurotrophin Levels in Interleukin-2 Knockout Mice, (to be submitted for publication). RD Beck, Jr., MA King, Z Huang and JM Petitto, 2002. Alterations in septohippocampal cholinergic neurons resulting from interleukin-2 gene knockout. Brain Res 955, 1623. R Bender, M Plaschke, T Naumann, P Wahle and M Frotscher, 1996. Development of cholinergic and GABAergic neurons in the rat medial septum: different onset of choline acetyltransferase and glutamate decarboxylase mRNA expression. J Comp Neurol 3 72, 204-214. EN Benveniste and JE Merrill, 1986. Stimulation of oligodendroglial proliferation and maturation by interleukin-2. Nature 321, 610-613. EN Benveniste, PK Herman and JN Whitaker, 1987. Myelin basic protein-specific RNA levels in interleukin-2-stimulated oligodendrocytes. J Neurochem 49, 1274-1279. M Besser and R Wank, 1999. Cutting edge: clonally restricted production of the neurotrophins brain-derived neurotrophic factor and neurotrophin-3 mRNA by human immune cells and Thl/Th2-polarized expression of their receptors. J Immunol 162, 6303-6306. M Bianchi and AE Panerai, 1993. Interleukin-2 enhances scopolamine-induced amnesia and hyperactivity in the mouse. Neuroreport 4, 1046-1048. M Bianchi, P Ferrario, N Zonta and AE Panerai, 1995. Effects of interleukin-1 beta and interleukin-2 on amino acids levels in m.ouse cortex and hippocampus. Neuroreport 6, 1689-1692. JL Bizon, JS Han, C Hudon and M Gallagher, 2003. Effects of hippocampal cholinergic deafferentation on leaming strategy selection in a visible platform version of the water maze. Hippocampus 13, 676-684. HR Brashear, L Zaborszky and L Heimer, 1986. Distribution of GABAergic and cholinergic neurons in the rat diagonal band. Neuroscience 17, 439-451.

PAGE 84

76 A Braun, M Lommatzsch, A Mannsfeldt, U Neuhaus-Steinmetz, A Fischer, N Schnoy, GR Lewin and H Renz, 1999. Cellular sources of enhanced brain-derived neurotrophic factor production in a mouse model of allergic inflammation. Am J Respir Cell Mol Biol 21, 537-546. FC Bronfman, D Moechars and F Van Leuven, 2000. Acetylcholinesterase-positive fiber deafferentation and cell shrinkage in the septohippocampal pathway of aged amyloid precursor protein london mutant transgenic mice. Neurobiol Dis 7, 152168. AI Brooks, MW Halterman and HJ Federoff, 1999. Focal hippocampal gain of NGF function elicits specific septal cholinergic reorganization. Neuroreport 10, 337-344. KM BuUer, 2001. Role of circumventricular organs in pro-inflammatory cytokineinduced activation of the hypothalamic-pituitary-adrenal axis. Clin Exp Pharmacol Physiol 28, 581-589. JF Cahill and MG Baxter, 2001 Cholinergic and noncholinergic septal neurons modulate strategy selection in spatial learning. Eur J Neurosci 14, 1856-1864. SB Caine, MA Geyer and NR Swerdlow, 1992. Hippocampal modulation of acoustic startle and prepulse inhibition in the rat. Pharmacol Biochem Behav 43, 1201-1208. JD Campbell, G Cook, SE Robertson, A Eraser, KS Boyd, JA Gracie and IM Franklin, 2001 Suppression of IL-2-induced T cell proliferation and phosphorylation of STAT3 and STAT5 by tumor-derived TGF beta is reversed by lL-15. J Immunol 167, 553-561. M Canossa, O Griesbeck, B Beminger, G Campana, R Kolbeck and H Thoenen, 1997. Neurotrophin release by neurotrophins: implications for activity-dependent neuronal plasticity. Proc Natl Acad Sci USA 94, 13279-13286. L Capuron, A Ravaud and R Dantzer, 2000. Early depressive symptoms in cancer patients receiving interleukin 2 and/or interferon alfa-2b therapy. J Clin Oncol 18, 2143-2151. L Capuron, A Ravaud and R Dantzer, 2001a. Timing and specificity of the cognitive changes induced by interleukin-2 and interferon-alpha treatments in cancer patients. Psychosom Med 63, 376-386. L Capuron, A Ravaud, N Gualde, E Bosmans, R Dantzer, M Maes and PJ Neveu, 2001b. Association between immune activation and early depressive symptoms in cancer patients treated with interleukin-2-based therapy. Psychoneuroendocrinology 26, 797-808. LL Carter, X Zhang, C Dubey, P Rogers, L Tsui and SL Swain, 1998. Regulation of T cell subsets from naive to memory. J Immunother 2 1 1 81-187.

PAGE 85

77 JC Cassel, E Duconseille, H Jeltsch and B Will, 1997. The fimbriafomix/cingular bundle pathways: a review of neurochemical and behavioural approaches using lesions and transplantation techniques. Prog Neurobiol 51, 663-716. JP Chandler and KA Crutcher, 1983. The septohippocampal projection in the rat: an electron microscopic horseradish peroxidase study. Neuroscience 10, 685-696. Q Chang and PE Gold, 2004. Impaired and spared cholinergic functions in the hippocampus after lesions of the medial septum/vertical limb of the diagonal band with 192 IgG-saporin. Hippocampus 14, 170-179. J Chappell, R McMahan, A Chiba and M Gallagher, 1998. A re-examination of the role of basal forebrain cholinergic neurons in spatial working memory. Neuropharmacology 37, 481-487. TJ Collier and A Routtenberg, 1984. Selective impairment of declarative memory following stimulation of dentate gyrus granule cells: a naloxone-sensitive effect. Brain Res 310, 384-387. JM Conner and S Varon, 1997. Developmental profile of NGF immunoreactivity in the rat brain: a possible role of NGF in the establishment of cholinergic terminal fields in the hippocampus and cortex. Brain Res Dev Brain Res 101, 67-79. JM Coimer, D Muir, S Varon, T Hagg and M Manthorpe, 1992. The localization of nerve growth factor-like immunoreactivity in the adult rat basal forebrain and hippocampal formation. J Comp Neurol 319, 454-462. CD Conrad and EJ Roy, 1993. Selective loss of hippocampal granule cells following adrenalectomy: implications for spatial memory. J Neurosci 13, 2582-2590. AK Cross and MN Woodroofe, 1 999. Chemokines induce migration and changes in actin polymerization in adult rat brain microglia and a human fetal microglial cell line in vitro. J Neurosci Res 55, 17-23. P Curzon, DJ Kim and MW Decker, 1994. Effect of nicotine, lobeline, and mecamylamine on sensory gating in the rat. Pharmacol Biochem Behav 49, 877882. JD Cushman, Z Huang, G Ha, R Beck, J Lo and JM Petitto, 2004. Contextual fear discrimination learning and memory is disrupted in IL-2 knockout mice. Society of Neuroscience Abstracts. CB Davis and SD Gillies, 2003. humunocytokines: amplification of anti-cancer immunity. Cancer Immunol Immunother 52, 297-308.

PAGE 86

78 KD Denicoff, DR Rubinow, MZ Papa, C Simpson, CA Seipp, MT Lotze, AE Chang, D Rosenstein and SA Rosenberg, 1987. The neuropsychiatric effects of treatment with interleukin-2 and lymphokine-activated killer cells. Ann Intern Med 1 07, 293300. JH Dufour, M Dziejman, MT Liu, JH Leung, TE Lane and AD Luster, 2002. IFNgamma-inducible protein 10 (IP10; CXCL10)-deficient mice reveal a role for IP10 in effector T cell generation and trafficking. J Immunol 168, 3195-3204. AJ Dunn, 2002. Mechanisms by which cytokines signal the brain. Int Rev Neurobiol 52, 43-65. O Eizenberg, A Faber-Elman, M Lotan and M Schwartz, 1995. Interleukin-2 transcripts in human and rodent brains: possible expression by astrocytes. J Neurochem 64, 1928-1936. H Fukuoka, M Kawatani, T Hisamitsu and C Takeshige, 1 994. Cutaneous hyperalgesia induced by peripheral injection of interleukin-1 beta in the rat. Brain Res 657, 133140. G Fyfe, RI Fisher, SA Rosenberg, M Sznol, DR Parkinson and AC Louie, 1995. Results of treatment of 255 patients with metastatic renal cell carcinoma who received high-dose recombinant interleukin-2 therapy. J Clin Oncol 13, 688-696. SL Gaffen, 2001. Signaling domains of the interleukin 2 receptor. Cytokine 14, 63-77. FH Gage and A Bjorklund, 1987. Trophic and growth-regulating mechanisms in the central ner\'ous system monitored by intracerebral neural transplants. Ciba Found Symp 126, 143-159. FH Gage, A Bjorklund and U Stenevi, 1983a. Reinnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and noradrenergic afferents. Brain Res 268, 27-37. FH Gage, A Bjorklund and U Stenevi, 1 984. Cells of origin of the ventral cholinergic septohippocampal pathway undergoing compensatory collateral sprouting following fimbria-fornix transection. Neurosci Lett 44, 211-216. FH Gage, A Bjorklund, U Stenevi and SB Dunnett, 1983b. Functional correlates of compensatory collateral sprouting by aminergic and cholinergic afferents in the hippocampal formation. Brain Res 268, 39-47. BH Gahwiler, A Enz and F Hefti, 1987. Nerve growth factor promotes development of the rat septo-hippocampal cholinergic projection in vitro. Neurosci Lett 75, 6-10.

PAGE 87

79 D Galey, C Destrade and R Jaffard, 1994. Relationships between septo-hippocampal cholinergic activation and the improvement of long-term retention produced by medial septal electrical stimulation in two inbred strains of mice. Behav Brain Res 60, 183-189. R Ganguli, JS Brar, KR Chengappa, M DeLeo, ZW Yang, G Shurin and BS Rabin, 1 995. Mitogen-stimulated interleukin-2 production in never-medicated, firstepisode schizophrenic patients. The influence of age at onset and negative symptoms. Arch Gen Psychiatry 52, 668-672. JG Giri, DM Anderson, S Kumaki, LS Park, KH Grabstein and D Cosman, 1995. IL-15, a novel T cell growth factor that shares activities and receptor components with IL-2. J Leukoc Biol 57, 763-766. PE Gold, 2003. Acetylcholine modulation of neural systems involved in learning and memory. Neurobiol Learn Mem 80, 194-210. E Gould, NJ Woolf and LL Butcher, 1991. Postnatal development of cholinergic neurons in the rat: I. Forebrain. Brain Res Bull 27, 767-789. E Gould, P Tanapat, T Rydel and N Hastings, 2000. Regulation of hippocampal neurogenesis in adulthood. Biol Psychiatry 48, 715-720. I Gritti, M Mariotti and M Mancia, 1998. GABAergic and cholinergic basal forebrain and preoptic-anterior hypothalamic projections to the mediodorsal nucleus of the thalamus in the cat. Neuroscience 85, 149-178. I Gritti, L Mainville, M Mancia and BE Jones, 1997. GABAergic and other noncholinergic basal forebrain neurons, together with cholinergic neurons, project to the mesocortex and isocortex in the rat. J Comp Neurol 383, 163-177. LM Guirguis, JC Yang, DE White, SM Steinberg, DJ Liewehr, SA Rosenberg and DJ Schwartzentruber, 2002. Safety and efficacy of high-dose interleukin-2 therapy in patients with brain metastases. J Immunother 25, 82-87. H Guo, JH Calkins, MM Sigel and T Lin, 1990. Interleukin-2 is a potent inhibitor of Leydigcell steroidogenesis. Endocrinology 127, 1234-1239. EG Gutierrez, WA Banks and AJ Kastin, 1993. Murine tumor necrosis factor alpha is transported from blood to brain in the mouse. J Neuroimmunol 47, 169-176. T Hagg, HL Vahlsing, M Manthorpe and S Varon, 1990. Nerve growth factor infusion into the denervated adult rat hippocampal formation promotes its cholinergic reinnervation. J Neurosci 10, 3087-3092. UK Hanisch and R Quirion, 1995a. Interleukin-2 as a neuroregulatory cytokine. Brain Res Brain Res Rev 21, 246-284.

PAGE 88

80 UK Hanisch and R Quirion, 1995b. Interleukin-2 as a neuroregulatory cytokine. Brain Res Brain Res Rev 21, 246-284. UK Hanisch, D Seto and R Quirion, 1993. Modulation of hippocampal acetylcholine release: a potent central action of interleukin-2. J Neurosci 13, 3368-3374. UK Hanisch, J Neuhaus, R Quirion and H Kettenmann, 1996. Neurotoxicity induced by interleukin-2: involvement of infiltrating immune cells. Synapse 24, 104-1 14. UK Hanisch, W Rowe, S Sharma, MJ Meaney and R Quirion, 1994. Hypothalamicpituitary-adrenal activity during chronic central administration of interleukin-2. Endocrinology 135, 2465-2472. UK Hanisch, SA Lyons, M Prinz, C Nolte, JR Weber, H Kettenmann and F Kirchhoff, 1997a. Mouse brain microglia express interleukin-15 and its multimeric receptor complex functionally coupled to Janus kinase activity. J Biol Chem 272, 2885328860. UK Hanisch, J Neuhaus, W Rowe, D Van Rossum, T Moller, H Kettenmann and R Quirion, 1997b. Neurotoxic consequences of central long-term administration of interleukin-2 in rats. Neuroscience 79, 799-818. J Hartikka and F Hefti, 1988. Comparison of nerve growth factor's effects on development of septum, striatum, and nucleus basalis cholinergic neurons in vitro. J Neurosci Res 21, 352-364. YW He and TR Malek, 1998. The structure and function of gamma c-dependent cytokines and receptors: regulation of T lymphocyte development and homeostasis. Crit Rev hnmunol 18, 503-524. JC Hedreen, SJ Bacon and DL Price, 1985. A modified histochemical technique to visualize acetylcholinesterasecontaining axons. J Histochem Cytochem 33 134140. R Hellweg, C Humpel, A Lowe and H Hortnagl, 1997. Moderate lesion of the rat cholinergic septohippocampal pathway increases hippocampal nerve growth factor synthesis: evidence for long-term compensatory changes? Brain Res Mol Brain Res 45, 177-181. WF Hickey and H Kimura, 1988. Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in vivo. Science 239, 290-292. WF Hickey, BL Hsu and H Kimura, 1991. T-lymphocyte entry into the central nervous system. J Neurosci Res 28, 254-260. I Horak, 1995. hnmunodeficiency in IL-2-knockout mice. Clin Immunol Immunopathol 76, S172-173.

PAGE 89

81 I Horak, 1996. Interleukin-2-knockout mice: a new model to study autoimmunity and self-tolerance. Sb Lek 97, 25-28. AA Hurwitz, WD Lyman and JW Berman, 1995. Tumor necrosis factor alpha and transforming growth factor beta upregulate astrocyte expression of monocyte chemoattractant protein1. J Neuroimmunol 57, 193-198. LS Janis, MM Glasier, Z Fulop and DG Stein, 1998. hitraseptal injections of 192 IgG saporin produce deficits for strategy selection in spatial-memory tasks. Behav Brain Res 90, 23-34. DA Johnson, NJ Zambon and RB Gibbs, 2002. Selective lesion of cholinergic neurons in the medial septum by 192 IgG-saporin impairs learning in a delayed matching to position T-maze paradigm. Brain Res 943, 132-141 P Kaiser, L Rothwell, D Vasicek and K Hala, 2002. A role for IL-15 in driving the onset of spontaneous autoimmune thyroiditis? J Immunol 168, 4216-4220. S Karanth, K Lyson and SM McCann, 1993. Role of nitric oxide in interleukin 2-induced corticotropin-releasing factor release from incubated hypothalami. Proc Natl Acad Sci U S A 90, 3383-3387. M Kerschensteiner, E Gallmeier, L Behrens, VV Leal, T Misgeld, WE Klinkert, R Kolbeck, E Hoppe, RL Oropeza-Wekerle, I Bartke, C Stadelmann, H Lassmann, H Wekerle and R Hohlfeld, 1999. Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor in vitro and in inflammatory brain lesions: a neuroprotective role of inflammation? J Exp Med 189, 865-870. MA King, BE Hunter, RL Reep and DW Walker, 1 989. Acetylcholinesterase stain intensity variation in the rat dentate gyrus: a quantitative description based on digital image analysis. Neuroscience 33, 203-221. J Kiss, AJ Patel and TF Freund, 1990a. Distribution of septohippocampal neurons containing parvalbumin or choline acetyltransferase in the rat brain. J Comp Neurol 298, 362-372. J Kiss, AJ Patel, KG Baimbridge and TF Freund, 1990b. Topographical localization of neurons containing parvalbumin and choline acetyltransferase in the medial septum-diagonal band region of the rat. Neuroscience 36, 61-72. M Knipper and RJ Rylett, 1997. A new twist in an old story: the role for crosstalk of neuronal and trophic activity. Neurochem Int 31, 659-676. M Knipper, M da Penha Berzaghi, A Blochl, H Breer, H Thoenen and D Lindholm, 1994. Positive feedback between acetylcholine and the neurotrophins nerve growth factor and brain-derived neurotrophic factor in the rat hippocampus. Eur J Neurosci 6, 668-671.

PAGE 90

82 M Koch, 1 996a. The septohippocampal system is involved in prepulse inhibition of the acoustic startle response in rats. Behav Neurosci 110, 468-477. M Koch, 1996b. The septohippocampal system is involved in prepulse inhibition of the acoustic startle response in rats. Behav Neurosci 1 10, 468-477. S Lacosta, Z Merali and H Anisman, 1999. Influence of acute and repeated interleukin-2 administration on spatial learning, locomotor activity, exploratory behaviors, and anxiety. Behav Neurosci 113, 1030-1041. H Lai and MJ Forster, 1988. Autoimmunity and age-associated cognitive decline. Neurobiol Aging 9, 733-742. MR Lamprea, FP Cardenas, R Silveira, S Morato and TJ Walsh, 2000. Dissociation of memory and anxiety in a repeated elevated plus maze paradigm: forebrain cholinergic mechanisms. Behav Brain Res 117, 97-105. PA Lapchak and DM Araujo, 1993. Interleukin-2 regulates monoamine and opioid peptide release from the hypothalamus. Neuroreport 4, 303-306. PA Lapchak, DM Araujo, R Quirion and A Beaudet, 1991. Immunoautoradiographic localization of interleukin 2-like immunoreactivity and interleukin 2 receptors (Tac antigen-like immunoreactivity) in the rat brain. Neuroscience 44, 173-184. E Larsson, RJ Mandel, RL Klein, N Muzyczka, O Lindvall and Z Kokaia, 2002. Suppression of insult-induced neurogenesis in adult rat brain by brain-derived neurotrophic factor. Exp Neurol 177, 1-8. JM Lauder and UB Schambra, 1999. Morphogenetic roles of acetylcholine. Environ Health Perspect 107 Suppl 1, 65-69. BR Lauwerys, N Garot, JC Renauld and FA Houssiau. 2000. Cytokine production and killer activity of NK/T-NK cells derived with IL-2, IL-15, or the combination of IL-12 and IL-18. J hnmunol 165, 1847-1853. G Leanza, OG Nilsson and A Bjorklund, 1993. Compensatory changes of in vivo acetylcholine and noradrenaline release in the hippocampus after partial deafferentation, as monitored by microdialysis. Brain Res 615, 147-159. G Leanza, OG Nilsson, RG Wiley and A Bjorklund, 1995. Selective lesioning of the basal forebrain cholinergic system by intraventricular 192 IgG-saporin: behavioural, biochemical and stereological studies in the rat. Eur J Neurosci 7 329343. J Lee, W Duan and MP Mattson, 2002. Evidence that brain-derived neurotrophic factor is required for basal neurogenesis and mediates, in part, the enhancement of neurogenesis by dietary restriction in the hippocampus of adult mice J Neurochem 82, 1367-1375.

PAGE 91

83 YB Lee, J Satoh, DG Walker and SU Kim, 1996. Interleukin-15 gene expression in human astrocytes and microglia in culture. Neuroreport 7, 1062-1066. WJ Leonard, JM Depper, GR Crabtree, S Rudikoff, J Pumphrey, RJ Robb, M Kronke, PB Svetlik, NJ Peffer, TA Waldmann and WC Greene, 1984. Molecular cloning and expression of cDNAs for the human interleukin-2 receptor. Nature 311, 626-63 L J Licinio, JP Seibyl, M Altemus, DS Chamey and JH Krystal, 1993. Elevated CSF levels of interleukin-2 in neuroleptic-free schizophrenic patients. Am J Psychiatry 150 1408-1410. T Lindenlaub, P Teuteberg, T Hartung and C Sommer, 2000. Effects of neutralizing antibodies to TNF-alpha on pain-related behavior and nerve regeneration in mice with chronic constriction injury. Brain Res 866, 15-22. BR Ludviksson, RO Ehrhardt and W Strober, 1997. TGF-beta production regulates the development of the 2,4,6-trinitrophenol-conjugated keyhole limpet hemocyanininduced colonic inflammation in IL-2-deficient mice. J hnmunol 159, 3622-3628. A Ma, M Datta, E Margosian, J Chen and I Horak, 1995. T cells, but not B cells, are required for bowel inflammation in interleukin 2-deficient mice. J Exp Med 182, 1567-1572. SF Maier, LE Goehler, M Fleshner and LR Watkins, 1998. The role of the vagus nerve in cytokine-to-brain communication. Ann N Y Acad Sci 840, 289-300. R Makuch, J Baratta, LD Karaelias, JC Lauterbom, CM Gall, J Yu and RT Robertson, 2001 Arrival of afferents and the differentiation of target neurons: studies of developing cholinergic projections to the dentate gyrus. Neuroscience 104, 81-91. D Maslinska, 2001. The cytokine network and interleukin-15 (IL-15) in brain development. Folia Neuropathoi 39, 43-47. CG McAllister, DP van Kammen, TJ Rehn, AL Miller, J Gurklis, ME Kelley, J Yao and JL Peters, 1995. Increases in CSF levels of interleukin-2 in schizophrenia: effects of recurrence of psychosis and medication status. Am J Psychiatry 152, 1291-1297. JF McDyer, Z Li, S John, X Yu, CY Wu and JA Ragheb, 2002. IL-2 receptor blockade inhibits late, but not early, IFN-gamma and CD40 ligand expression in human T cells: disruption of both IL-12-dependent and -independent pathways of IFNgamma production. J hnmunol 169, 2736-2746. RL McLamb, WR Mundy and HA Tilson, 1988. Intradentate colchicine disrupts the acquisition and performance of a working memory task in the radial arm maze. Neurotoxicology 9, 521-528.

PAGE 92

84 L Meda, S Bemasconi, C Bonaiuto, S Sozzani, D Zhou, L Otvos, Jr., A Mantovani, F Rossi and MA Cassatella, 1996. Beta-amyloid (25-35) peptide and IFN-gamma synergistically induce the production of the chemotactic cytokine MCP-l/JE in monocytes and microglial cells. J Immunol 157, 1213-1218. MA Meijssen, SL Brandwein, HC Reinecker, AK Bhan and DK Podolsky, 1998. Alteration of gene expression by intestinal epithelial cells precedes colitis in interIeukin-2-deficient mice. Am J Physiol 274, G472-479. F Mennicken and R Quirion, 1997. hiterleukin-2 increases choline acetyltransferase activity in septalcell cultures. Synapse 26, 175-183. RT Mitsuyasu, 2001. The potential role of interleukin-2 in HIV. Aids 15 Suppl 2, S22-27. H Mizuma, K Takagi, K Miyake, N Takagi, K Ishida, S Takeo, A Nitta, H Nomoto, Y Furukawa and S Furukawa, 1 999. Microsphere embolism-induced elevation of nerve growth factor level and appearance of nerve growth factor immunoreactivity in activated T-lymphocytes in the rat brain. J Neurosci Res 55, 749-761. G Moalem, A Gdalyahu, Y Shani, U Otten, P Lazarovici, IR Cohen and M Schwartz, 2000. Production of neurotrophins by activated T cells: implications for neuroprotective autoimmunity. J Autoimmun 15, 331-345. WC Mobley, JL Rutkowski, GI Tennekoon, J Gemski, K Buchanan and MV Johnston, 1986. Nerve growth factor increases choline acetyltransferase activity in developing basal forebrain neurons. Brain Res 387, 53-62. ML Monje, H Toda and TD Palmer, 2003. Inflammatory blockade restores adult hippocampal neurogenesis. Science 302, 1760-1765. DA Morgan, FW Ruscetti and R Gallo, 1976. Selective in vitro growth of T lymphocytes from normal human bone marrows. Science 193, 1007-1008. JK Morse, SJ Wiegand, K Anderson, Y You, N Cai, J Camahan, J Miller, PS DiStefano, CA Altar, RM Lindsay and et al., 1993. Brain-derived neurotrophic factor (BDNF) prevents the degeneration of medial septal cholinergic neurons following fimbria transection. J Neurosci 13, 4146-4156. N Muller and M Ackenheil, 1998. Psychoneuroimmunology and the cytokine action in the CNS: implications for psychiatric disorders. Prog Neuropsychopharmacol Biol Psychiatry 22, 1-33. Y Nakamura, SM Russell, SA Mess, M Friedmann, M Erdos, C Francois, Y Jacques, S Adelstein and WJ Leonard, 1994. Heterodimerization of the IL-2 receptor betaand gamma-chain cytoplasmic domains is required for signalling. Nature 369, 330-333.

PAGE 93

85 KP Nanry, WR Mundy and HA Tilson, 1989. Colchicine-induced alterations of reference memory in rats: role of spatial versus non-spatial task components. Behav Brain Res 35, 45-53. BH Nelson, 2004. IL-2, regulatory T cells, and tolerance. J Immunol 172, 3983-3988. BH Nelson, JD Lord and PD Greenberg, 1994. Cytoplasmic domains of the interleukin-2 receptor beta and gamma chains mediate the signal for T-cell proliferation. Nature 369, 333-336. R Nemni, S lannaccone, A Quattrini, S Smime, M Sessa, M Lodi, C Erminio and N Canal, 1992. Effect of chronic treatment with recombinant interleukin-2 on the central nervous system of adult and old mice. Brain Res 591, 248-252. W Pan, WA Banks and AJ Kastin, 1997. Permeability of the blood-brain and bloodspinal cord barriers to interferons. J Neuroimmunol 76, 105-1 11 K Pardy, D Murphy, D Carter and KM Hui, 1993. The influence of interleukin-2 on vasopressin and oxytocin gene expression in the rodent hypothalamus. J Neuroimmunol 42, 131-138. DR Parkinson, JS Abrams, PH Wiemik, AA Rayner, KA Margolin, DA Van Echo, M Sznol, JP Dutcher, PR Aronson, JH Doroshow and et al., 1990. Interleukin-2 therapy in patients with metastatic malignant melanoma: a phase II study. J Clin Oncol 8, 1650-1656. A Patapoutian and LP Reichardt, 2001. Trk receptors: mediators of neurotrophin action. Curr Opin Neurobiol 11, 272-280. S Pauli, AC Linthorst and JM Reul, 1998. Tumour necrosis factor-alpha and interleukin-2 differentially affect hippocampal serotonergic neurotransmission, behavioural activity, body temperature and hypothalamic-pituitary-adrenocortical axis activity in the rat. Eur J Neurosci 10, 868-878. M Perez-Martin, I Azcoitia, JL Trejo, A Sierra and LM Garcia-Segura, 2003. An antagonist of estrogen receptors blocks the induction of adult neurogenesis by insulin-like growth factor-I in the dentate gyrus of adult female rat. Eur J Neurosci 18, 923-930. T Perry, H Hodges and JA Gray, 2001. Behavioural, histological and immunocytochemical consequences following 192 IgG-saporin immunolesions of the basal forebrain cholinergic system. Brain Res Bull 54, 29-48. JM Petitto and Z Huang, 1994. Molecular cloning of a partial cDNA of the interleukin-2 receptor-beta in normal mouse brain: in situ localization in the hippocampus and expression by neuroblastoma cells. Brain Res 650, 140-145.

PAGE 94

86 JM Petitto and Z Huang, 1995. Molecular cloning of the coding sequence of an interleukin-2 receptor alpha subunit cDNA in murine brain. J Neuroimmunol 59, 135-141. JM Petitto and Z Huang, 2001. Cloning the full-length IL-2/1 5 receptor-beta cDNA sequence from mouse brain: evidence of enrichment in hippocampal formation neurons. Regul Pept 98, 77-87. JM Petitto, Z Huang, J Lo and WJ Streit, 2003. lL-2 gene knockout affects T lymphocyte trafficking and the microglial response to regenerating facial motor neurons. J Neuroimmunol 134, 95-103. JM Petitto, DB McCarthy, CM Rinker, Z Huang and T Getty, 1997. Modulation of behavioral and neurochemical measures of forebrain dopamine function in mice by species-specific interleukin-2. J Neuroimmunol 73, 183-190. JM Petitto, Z Huang, MK Raizada, CM Rinker and DB McCarthy, 1998. Molecular cloning of the cDNA coding sequence of IL-2 receptor-gamma (gammac) from human and murine forebrain: expression in the hippocampus in situ and by brain cells in vitro. Brain Res Mol Brain Res 53, 152-162. JM Petitto, RK McNamara, PL Gendreau, Z Huang and AJ Jackson, 1999. Impaired learning and memory and altered hippocampal neurodevelopment resulting from interleukin-2 gene deletion. J Neurosci Res 56, 441-446. JM Petitto, WJ Streit, Z Huang, E Butfiloski and J Schiffenbauer, 2000. Interleukin-2 gene deletion produces a robust reduction in susceptibility to experimental autoimmune encephalomyelitis in C57BL/6 mice. Neurosci Lett 285, 66-70. CR Plata-Salaman and JM ffrench-MuUen, 1993. Interleukin-2 modulates calcium currents in dissociated hippocampal CAl neurons. Neuroreport 4, 579-581. LQ Ren, N Gourmala, HW Boddeke and PJ Gebicke-Haerter, 1998. Lipopolysaccharideinduced expression of IP-10 mRNA in rat brain and in cultured rat astrocytes and microglia. Brain Res Mol Brain Res 59, 256-263. GE Ringheim, BD Freimark and RJ Robb, 1991. Quantitative characterization of the intrinsic ligand-binding affinity of the interleukin 2 receptor beta chain and its modulation by the alpha chain and a second affinity-modulating element. Lymphokine Cytokine Res 10, 219-224. SA Rosenberg, JJ Mule, PJ Spiess, CM Reichert and SL Schwarz, 1985. Regression of established pulmonary metastases and subcutaneous tumor mediated by the systemic administration of high-dose recombinant interleukin 2. J Exp Med 161, 1169-1188.

PAGE 95

87 SA Rosenberg, JC Yang, SL Topalian, DJ Schwartzentruber, JS Weber, DR Parkinson, CA Seipp, JH Einhom and DE White, 1994. Treatment of 283 consecutive patients with metastatic melanoma or renal cell cancer using high-dose bolus interleukin 2. Jama 271, 907-913. R Rozental, MF Mehler, M Morales, AF Andrade-Rozental, JA Kessler and DC Spray, 1995. Differentiation of hippocampal progenitor cells in vitro: temporal expression of intercellular coupling and voltageand ligand-gated responses. Dev Biol 167, 350-362. F Ruberti, S Capsoni, A Comparini, E Di Daniel, J Franzot, S Gonfloni, G Rossi, N Berardi and A Cattaneo, 2000. Phenotypic knockout of nerve growth factor in adult transgenic mice reveals severe deficits in basal forebrain cholinergic neurons, cell death in the spleen, and skeletal muscle dystrophy. J Neurosci 20, 2589-2601. T Saarelainen, S Vaittinen and E Castren, 2001 trkB-receptor activation contributes to the kainate-induced increase in BDNF mRNA synthesis. Cell Mol Neurobiol 21, 429-435. N Sakai, S Kaufman and S Milstien, 1995. Parallel induction of nitric oxide and tetrahydrobiopterin synthesis by cytokines in rat glial cells. J Neurochem 65, 895902. B Sakic, H Szechtman and JA Denburg, 1997. Neurobehavioral alterations in autoimmune mice. Neurosci Biobehav Rev 21, 327-340. B Sakic, H Szechtman, S Denburg, R Carbotte and JA Denburg, 1993. Spatial learning during the course of autoimmune disease in MRL mice. Behav Brain Res 54 5766. RP Saneto, A Altman, RL Knobler, HM Johnson and J de Vellis, 1986. hiterleukin 2 mediates the inhibition of oligodendrocyte progenitor cell proliferation in vitro. Proc Natl Acad Sci U S A 83, 9221-9225. M Sarder, H Saito and K Abe, 1993. Interleukin-2 promotes survival and neurite extension of cultured neurons from fetal rat brain. Brain Res 625, 347-350. M Sarder, K Abe, H Saito and N Nishiyama, 1996. Comparative effect of lL-2 and IL-6 on morphology of cultured hippocampal neurons from fetal rat brain. Brain Res 715,9-16. J Satoh, K Kurohara, M Yukitake and Y Kuroda, 1998. Interleukin15, a T-cell growth factor, is expressed in human neural cell lines and tissues. J Neurol Sci 155 170177. A Schimpl, I Berberich, B Kneitz, S Kramer, B Santner-Nanan, S Wagner, M Wolf and T Hunig, 2002. IL-2 and autoimmune disease. Cytokine Growth Factor Rev 13 369378.

PAGE 96

88 E Schmitt, T Germann, S Goedert, P Hoehn, C Huels, S Koelsch, R Kuhn, W Muller, N Palm and E Rude, 1994. IL-9 production of naive CD4+ T cells depends on IL-2, is synergistically enhanced by a combination of TGF-beta and IL-4, and is inhibited by IFN-gamma. J Immunol 153, 3989-3996. R Schopke, DP Wolfer, HP Lipp and MC Leisinger-Trigona, 1991. Swimming navigation and structural variations of the infrapyramidal mossy fibers in the hippocampus of the mouse. Hippocampus 1, 315-328. H Schwegler and WE Crusio, 1995. Correlations between radial-maze learning and structural variations of septum and hippocampus in rodents. Behav Brain Res 67, 29-41. H Schwegler, WE Crusio, HP Lipp and B Heimrich, 1988. Water-maze learning in the mouse correlates with variation in hippocampal morphology. Behav Genet 18, 153165. H Schwegler, M Boldyreva, R Linke, J Wu, K Zilles and WE Crusio, 1996a. Genetic variation in the morphology of the septo-hippocampal cholinergic and GABAergic systems in mice: II. Morpho-behavioral correlations. Hippocampus 6, 535-545. H Schwegler, M Boldyreva, M Pyrlik-Gohlmann, R Linke, J Wu and K Zilles, 1996b. Genetic variation in the morphology of the septo-hippocampal cholinergic and GABAergic system in mice. I. Cholinergic and GABAergic markers. Hippocampus 6, 136-148. BM Segal, BK Dwyer and EM Shevach, 1998. An interleukin (IL)-10/IL-12 immunoregulatory circuit controls susceptibility to autoimmune disease. J Exp Med 187, 537-546. D Seto, S Kar and R Quirion, 1997. Evidence for direct and indirect mechanisms in the potent modulatory action of interleukin-2 on the release of acetylcholine in rat hippocampal slices. Br J Pharmacol 120, 1151-1157. NM Shah, AK Groves and DJ Anderson, 1996. Alternative neural crest cell fates are instructively promoted by TGFbeta superfamily members. Cell 85, 331-343. M Shimojo, Y Imai, K Nakajima, S Mizushima, A Uemura and S Kohsaka, 1993. Interleukin-2 enhances the viability of primary cultured rat neocortical neurons. Neurosci Lett 151, 170-173. KA Smith, 2001. Low-dose daily interleukin-2 immunotherapy: accelerating immune restoration and expanding HlV-specific T-cell immunity without toxicity. Aids 15 Suppl 2, S28-35. LS Sorkin, WH Xiao, R Wagner and RR Myers, 1997. Tumour necrosis factor-alpha induces ectopic activity in nociceptive primary afferent fibres. Neuroscience 8 1 255-262.

PAGE 97

89 WJ Streit, MB Graeber and GW Kreutzberg, 1988. Functional plasticity of microglia: a review. Glia 1, 301-307. WJ Streit, SL Semple-Ro\Yland, SD Hurley, RC Miller, PG Popovich and BT Stokes, 1998. Cytokine mRNA profiles in contused spinal cord and axotomized facial nucleus suggest a beneficial role for inflammation and gliosis. Exp Neurol 152 7487. K Sugamura, H Asao, M Kondo, N Tanaka, N Ishii, K Ohbo, M Nakamura and T Takeshita, 1996. The interleukin-2 receptor gamma chain: its role in the multiple cytokine receptor complexes and T cell development in XSCID. Annu Rev Immunol 14, 179-205. H Super and E Soriano, 1994. The organization of the embryonic and early postnatal murine hippocampus. II. Development of entorhinal, commissural, and septal connections studied with the lipophilic tracer Dil. J Comp Neurol 344, 101-120. T Takeshita, H Asao, K Ohtani, N Ishii, S Kumaki, N Tanaka, H Munakata, M Nakamura and K Sugamura, 1992. Cloning of the gamma chain of the human IL-2 receptor. Science 257, 379-382. V Tancredi, C Zona, F Velotti, F Eusebi and A Santoni, 1990. Interleukin-2 suppresses established long-term potentiation and inhibits its induction in the rat hippocampus. Brain Res 525, 149-151. LJ Thai, E Gilbertson, DM Armstrong and FH Gage, 1992. Development of the basal forebrain cholinergic system: phenotype expression prior to target innervation. Neurobiol Aging 13, 67-72. JJ Thune and B Pakkenberg, 2000. Stereological studies of the schizophrenic brain. Brain Res Brain Res Rev 31, 200-204. L Vallieres, IL Campbell, FH Gage and PE Sawchenko, 2002. Reduced hippocampal neurogenesis in adult transgenic mice with chronic astrocytic production of interleukin-6. J Neurosci 22, 486-492. F Villemain, T Owens, T Renno and A Beaudet, 1991. Localization of mRNA for interleukin-2 (IL-2) in mouse brain by in situ hybridization (Abstract). Society of Neuroscience Abstracts 17, 1199. M Virgili, A Contestabile and O Bamabei, 1991 Postnatal maturation of cholinergic markers in forebrain regions of C57BL/6 mice. Brain Res Dev Brain Res 63 281285. PJ Waguespack, WA Banks and AJ Kastin, 1994. hiterleukin-2 does not cross the bloodbrain barrier by a saturable transport system. Brain Res Bull 34, 103-109.

PAGE 98

90 T Waldmann, 2002. The contrasting roles of IL-2 and IL-15 in the life and death of lymphocytes: implications for the immunotherapy of rheumatological diseases. Arthritis Res 4 Suppl 3, S161-167. TA Waldmann, S Dubois and Y Tagaya, 2001. Contrasting roles of IL-2 and IL-15 in the life and death of lymphocytes: implications for immunotherapy. Immunity 14, 105110. LG Walker, MB Walker, SD Heys, J Lolley, K Wesnes and O Eremin, 1997. The psychological and psychiatric effects of rIL-2 therapy: a controlled clinical trial. Psychooncology 6, 290-301. TJ Walsh, DW Schulz, HA Tilson and DE Schmechel, 1986. Colchicine-induced granule cell loss in rat hippocampus: selective behavioral and histological alterations. Brain Res 398, 23-36. NL Ward and T Hagg, 2000. BDNF is needed for postnatal maturation of basal forebrain and neostriatum cholinergic neurons in vivo. Exp Neurol 162, 297-310. PC Wilkinson and FY Liew, 1995. Chemoattraction of human blood T lymphocytes by interleukin-15. J Exp Med 181, 1255-1259. RW Williams and P Rakic, 1988. Three-dimensional counting: an accurate and direct method to estimate numbers of cells in sectioned material. J Comp Neurol 278, 344-352. CJ Wilson, CE Finch and HJ Cohen, 2002. Cytokines and cognition-the case for a headto-toe inflammatory paradigm. J Am Geriatr Soc 50, 2041-2056. O Witzke, T Winterhagen, A Kribben, T Philipp, K Mann and W Reinhardt, 2003. Interleukin-2 given to asymptomatic HIV-infected individuals leads to an exaggerated response of the pituitary gland to the action of CRH. Clin Endocrinol (0x1) 59, 104-109. G Wong, Y Goldshmit and AM Tumley, 2004. Interferon-gamma but not TNF alpha promotes neuronal differentiation and neurite outgrowth of murine adult neural stem cells. Exp Neurol 187, 171-177. NJ Woolf, F Eckenstein and LL Butcher, 1984. Cholinergic systems in the rat brain: I. projections to the limbic telencephalon. Brain Res Bull 13, 751-784. CC Wrenn, DA Lappi and RG Wiley, 1999. Threshold relationship between lesion extent of the cholinergic basal forebrain in the rat and working memory impairment in the radial maze. Brain Res 847, 284-298. K Yoshida and H Oka, 1995. Topographical projections from the medial septum-diagonal band complex to the hippocampus: a retrograde tracing study with multiple fluorescent dyes in rats. Neurosci Res 21, 199-209.

PAGE 99

91 JQ Zheng, M Felder, JA Connor and MM Poo, 1994. Turning of nerve growth cones induced by neurotransmitters. Nature 368, 140-144.

PAGE 100

BIOGRAPHICAL SKETCH Ray Dennis Beck, Jr. was bom in York, PA, to Daisy and the appropriately named Ray Dennis Beck, Sr. Realizing the error of his ways early in life, Ray moved to Houston, TX, at the age of two, lending credence to the old Texas saying, "I wasn't bom in Texas, but I got here as fast as I could." He attended Oak Ridge High School until the age of 16. Ranked 14 overall in his class, he elected to forego his senior year to attend Simon's Rock College of Bard in Great Barrington, MA. Whereas some 16-year-old adolescents are mature enough to pursue a college education while resisting distractions like geographically convenient buildings filled with members of the opposite sex, a ready supply of fermented beverages, and complete lack of parental supervision Ray was not. He returned to Houston after his first year of college with a less than stellar GPA intent on taking a year off from school. During this "year" (comprised of -1,825 days), he held various jobs ranging from perfume salesman to waiter at various restaurants. During one of the latter jobs, he met his present wife, Laura Frakey. Inspired by her enthusiasm for education, Ray enrolled in the University of Houston, while maintaining full-time employment to pay for school. He graduated cum laude with a B.S. in biology. Ray and Laura moved to Gainesville, FL, to attend University of Florida graduate programs in neuroscience and psychology, respectively. With the completion of his Ph.D., Ray is proud to be one of the most educated high school dropouts that anyone is ever likely to meet. Ray is an avid follower of movies and his other interests can be classified as "all things geeky" (e.g., computers, roleplaying, video games, etc.). 92

PAGE 101

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. John IS^. Petitto, Chair Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Michael A. King, Cochair Associate Scientist of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. -"^ Mark H. Lewis Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.' Wolfgang J. Streit Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fiiUy adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ,^ ^"^•V*'""'" Mark A. Atkinson Professor of Pathology, Immunology and Laboratory Medicine

PAGE 102

This dissertation was submitted to the Graduate Faculty of the College of Medicine and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. / 7 August 2004 X' ^Dean, College of Medicine Dean, Graduate School


28
Sakic et al., 1993), more subtle autoimmune processes may be at play in the IL-2 KO
mice (e.g., autoantibodies). It would be of interest to explore the observed cholinergic
changes in old versus neonatal mice to determine if the abnormalities increase with age
due to neurodegeneration, or are primarily the result of abnormal development. Such
knowledge will then enable us to develop a more specific model to test relevant
hypotheses involving IL-2 at specific anatomical sites in the septohippocampal
cholinergic system.
In summary, these data demonstrate that loss of endogenous IL-2 results in
reduction in the number of cholinergic neurons in the MS/vDB and alterations in the
structural morphology of dentate projection fields. These findings extend our previous
experiments showing that spatial learning and hippocampal mossy fiber length are
abnormal in IL-2 KO mice (Petitto et al., 1999). Further research is needed to determine
whether these outcomes in IL-2 KO mice may be due to the absence of central or
peripheral IL-2 during neurodevelopment (or some combination of both sources),
neurodegeneration secondary to peripheral autoimmunity, or other factors associated with
the absence of IL-2.


19
positioned in the corpus collosum was used to represent the background staining intensity
of tissue containing little cholinesterase activity. All of the average intensities along the
line were then plotted by traverse position to illustrate the quantitative patterns of AChE
distribution across each subregion. Conspicuous inflections marking the transition from
alveus (Alv) to stratum oriens (SO, in CA1, CA3b), from stratum lacunosum-moleculare
(Lmol) to dentate molecular layer (Mol) (hippocampal fissure (HiF); CAI, CA3, DG),
and polymorph zone (PoDG) to granule cell layer (GCL) were used to align traverses
across sections and animals (Figure 2-2). Also, length-normalized comparisons of the
Figure 2-2. A micrograph of the sampling regions utilized for image analysis of AChE-
staining. Alv=alveus, DG=dentate gyrus, GCL=granular cell layer
(IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus,
Lmol=lacunosum moleculare, Mol=molecular layer, Pi=pial surface, SO=stratum
oriens, SP=septum pellucidum, SR=stratum radiatum. The scale bar represents
200 pm.
measured regions were made by converting the sampling lengths to 100 points using the
software Matlab v.5.3. Dependent variables were the absolute length of traverses, AChE
intensity at anatomically identifiable inflection points (IP and SP bands, Lmol, dentate


38
Results
Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell Number
in Adult Mice
In 8-12-week-old mice, no significant differences were apparent in the relative
number of stereologically counted Parv-positive neurons between groups (F(l,10)-0.002,
p=0.964). Thus, the GABAergic neurons appear to be unaffected by IL-2 gene deletion.
We did not assay for GABAergic alterations in younger animals, since there were no
differences in the adult IL-2 KO mice relative to the wild-types.
In contrast to the previously reported data from 8-12-week-old animals (Beck et
al., 2002), there was no significant difference in stereologically counted cholinergic
somata number in the MS/vDB of 21-day-old IL-2 KO mice relative to wild-type mice
(F(l,12)=0.689, p=0.423). As expected, the 21-day-old KO mice also did not exhibit
splenomegaly seen in the autoimmune 8-12-week-old group, as spleen weights did not
differ between 21-day-old wild type and IL-2 KO mice (F(l,12)=0.989, p=0.340).
Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice
The IP-GCL of IL-2 KO mice had significantly fewer neuronal somata than wild-
type mice (Fig. 3-1; F( 1,12)=10.966, p=0.006). In the SP-GCL, however, there was no
significant difference in granule cell number (Fig 3-1.; F( 1,12)=0.197, p=0.665).
Alterations in Neurotrophin Levels
Levels of NGF protein in hippocampal tissue homogenates was significantly
increased in IL-2 KO mice relative to wild-type mice (Fig. 3-2A; F( 1,10)8.261,
p=0.017). The levels of BDNF protein, conversely, were significantly decreased in IL-2
KO mice relative to wild-type mice (Fig. 3-2B; F( 1,12)=8.023, p=0.015).


69
loss. A DNA fragmentation assay, like terminal transferase-mediated dUTP nick-end
labeling (TUNEL), could be used to detect apoptotic cell death and cell death by necrosis
could be determined by a simple trypan blue dye exclusion method. Both assays would
require immunohistochemical labeling with ChAT to identify the cholinergic somata.
In Study 3, the cytokine and chemokine profile of the IL-2 KO hippocampus was
altered compared to wild-type mice. Though we did speculate on the cell types that
might produce these cytokines, we did not investigate it further. To identify the potential
sources of these cytokines, we could use in situ hybridization to label the mRNA and
label the individual cell types with immunohistochemistry. We could expand upon this
further by utilizing real time PCR techniques as a semi-quantitative assay of different
levels of mRNA expression in homogenates from the hippocampi of IL-2 KO and wild-
type mice. These two experiments allow us to identify the cells producing the cytokines
and compare the relative levels of mRNA produced for each cytokine between IL-2 KO
and wild-type mice.
Several other considerations and questions arise from the data generated from the
experiments of this dissertation. First, there is possibility that the reduced levels of
ChAT* immunostaining may be due to a loss of the cholinergic phenotype (e.g., inability
to produce ACh) rather than cell death. Alterations in the activity of ACh in the
hippocampus (e.g., addition of receptor agonists or antagonists) are sufficient to alter
learning and memory (for a review see Gold, 2003). Thus, the behavioral deficits
observed in the IL-2 KO animals may not be due to cholinergic neuronal loss. Although
the staining intensity was uniform (e.g., consistent between and within groups), to test
this further, IL-2 KO and wild-type littermate brains should be Nissl-stained to label and


79
D Galey, C Destrade and R Jaffard, 1994. Relationships between septo-hippocampal
cholinergic activation and the improvement of long-term retention produced by
medial septal electrical stimulation in two inbred strains of mice. Behav Brain Res
60, 183-189.
R Ganguli, JS Brar, KR Chengappa, M DeLeo, ZW Yang, G Shurin and BS Rabin, 1995.
Mitogen-stimulated interleukin-2 production in never-medicated, first- episode
schizophrenic patients. The influence of age at onset and negative symptoms. Arch
Gen Psychiatry 52, 668-672.
JG Giri, DM Anderson, S Kumaki, LS Park, KH Grabstein and D Cosman, 1995. IL-15, a
novel T cell growth factor that shares activities and receptor components with IL-2.
J Leukoc Biol 57, 763-766.
PE Gold, 2003. Acetylcholine modulation of neural systems involved in learning and
memory. Neurobiol Leam Mem 80, 194-210.
E Gould, NJ Woolf and LL Butcher, 1991. Postnatal development of cholinergic neurons
in the rat: I. Forebrain. Brain Res Bull 27, 767-789.
E Gould, P Tanapat, T Rydel and N Hastings, 2000. Regulation of hippocampal
neurogenesis in adulthood. Biol Psychiatry 48, 715-720.
I Gritti, M Mariotti and M Mancia, 1998. GABAergic and cholinergic basal forebrain and
preoptic-anterior hypothalamic projections to the mediodorsal nucleus of the
thalamus in the cat. Neuroscience 85, 149-178.
I Gritti, L Mainville, M Mancia and BE Jones, 1997. GABAergic and other
noncholinergic basal forebrain neurons, together with cholinergic neurons, project
to the mesocortex and isocortex in the rat. J Comp Neurol 383, 163-177.
LM Guirguis, JC Yang, DE White, SM Steinberg, DJ Liewehr, SA Rosenberg and DJ
Schwartzentruber, 2002. Safety and efficacy of high-dose interleukin-2 therapy in
patients with brain metastases. J Immunother 25, 82-87.
H Guo, JH Calkins, MM Sigel and T Lin, 1990. Interleukin-2 is a potent inhibitor of
Leydig cell steroidogenesis. Endocrinology 127, 1234-1239.
EG Gutierrez, WA Banks and AJ Kastin, 1993. Murine tumor necrosis factor alpha is
transported from blood to brain in the mouse. J Neuroimmunol 47, 169-176.
T Hagg, HL Vahlsing, M Manthorpe and S Varn, 1990. Nerve growth factor infusion
into the denervated adult rat hippocampal formation promotes its cholinergic
reinnervation. J Neurosci 10, 3087-3092.
UK Hanisch and R Quirion, 1995a. Interleukin-2 as a neuroregulatory cytokine. Brain
Res Brain Res Rev 21, 246-284.


21
Cholinergic Somata in the MSA/DB
2000
| 1500
O
>
o 1000
4
I-
2
o
500
0
Fig. 2-3. ChAT-positive somata are significantly reduced in the MS/vDB of IL-2
knockout mice. Each bar represents the mean SEM of 9 animals per group.
*p=0.01.
CA1 (Figure 2-4a), CA3 (Figure 2-4b), the SP-GCL (Figure 2-4c), and IP-GCL layer of
the DG (Figure 2-4d). Repeated measures ANOVA was performed on regions of the
average normalized curves selected by areas that appeared to deviate between the groups.
None of these areas, however, were found to differ between IL-2 KO and wild-type mice.
Morphology of the Granular Cell Layer of the Lower Limb of the DG
There were no differences in the groups for the Y-axis (intensity) data. The
variations in the patterns of the DG curves X-axis (distance) were also compared.
Distances were compared by defining the point of lowest intensity in the regions that the
curves indicated as each transition between the regions of interest. Distances are reported
as a percentage of the total distance across each curve. The IP blade of the DG was
broken into three regions: the Hil, the IP-GCL, and the Mol (Figure 2-4d). As depicted in


10
effects, such as T and B cell invasion of the brain, angiogenesis, changes in the
composition of the extracellular matrix, myelin damage, and neuronal cell loss seen in
rats administered human IL-2 intracerebroventricularly via minipumps (Hanisch et al.,
1996; Hanisch et al., 1997b). Thus, my approach was to use IL-2 knockout mice. These
studies were the first to investigate the consequences of the absence of IL-2 on aspects of
brain development and maintenance of the septohippocampal system in vivo. Our
laboratory had found the aforementioned behavioral alterations in IL-2 knockout mice,
and therefore another important goal of my research was to test the hypotheses regarding
the neurobiological and neuroimmunological alterations that may underlie these
behavioral abnormalities.


27
formation, where its release may regulate the development and function of septal
cholinergic neurons projecting to the hippocampus. IL-2-like immunoreactivity has been
localized to the hippocampal formation in rat forebrain (Lapchak et al., 1991), and
detected in tissue extracts from rat and human hippocampal tissue (Araujo et al., 1989).
In mouse brain, IL-2 mRNA has been found in the hippocampus (Villemain et al., 1991),
and transcripts for this cytokine may be expressed in rat astrocyte cultures as well
(Eizenberg et al., 1995).
In the periphery, absence of endogenous IL-2 leads to an immunodysregulation
that produces loss of self-tolerance and IL-2 KO mice eventually develop generalized
systemic autoimmune disease (although C57BL/6-IL-2 KO mice develop clinical signs of
systemic autoimmunity at a substantially slower rate than other strains such as Balb/c or
C3H) (Petitto et al., 2000). Therefore, it is reasonable to speculate that the
neuroanatomical alterations found in the IL-2 KO mice result from peripheral
autoimmune processes. Autoimmunity could impact on brain development or induce
neurodegeneration. The former seems unlikely, however, since IL-2 KO mice do not
express the first signs autoimmunity (e.g., splenomegaly) until at least three to four weeks
afterbirth (Horak, 1995); by this time, septohippocampal development should already be
complete (Bender et al., 1996; Chandler and Crutcher, 1983; Super and Soriano, 1994;
Yoshida and Oka, 1995). Furthermore, the likelihood that these neuroanatomical
alterations may be due to autoimmune-induced degeneration of existing neurons also
seems unlikely, since lymphocytes cannot be detected in the brain of adult IL-2 KO mice
(Petitto et al., 1999). Nonetheless, since autoimmunity has been associated with
cognitive changes in both animals and humans (Lai and Forster, 1988; Sakic et al., 1997;


55
F(l,15)=6.105, p=0.026) in the IL-2 KO compared to wild-type mice. Figure 4-1 also
shows that IL-2 KO mice had increased levels of the chemokines, IP-10 (increased
-63%; F(l,15)=4.747, p=0.046) and MCP-1 (increased -46%; F(l,15)=5.218, p=0.039).
Although detectable levels of IL-7 and IL-9 were found in hippocampus, they did not
differ between the subject groups.
Enhanced Cytokine Concentrations in the Hippocampus
CD
Z3
I
a>
S
oo
5
o>
E
ai
c
O
S'
o
O)
Q.
12
1 0
0.8
06
04
02
0.0
*
IL-12 IL-15 IP-10 MCP-1
Cytokine
Figure 4-1. Protein levels of several cytokines and chemokines are elevated in the IL-2
knockout mice compared to wild-type brain. These include IL-12, IL-15, IP-10,
and MCP-1. *p<0.05; **p<0.01.
Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-type Mice
As can been seen in Figure 4-2, the IL-2 KO mice exhibited marked elevation in
serum IL-6 (increased -224%; F(l,10)=8.077, p=0.017), and serum MIP-la serum
concentration (increased -327%; F(l,10)=21.538, p=0.001) compared to wild-types.
Although detectable levels of IL-13, G-CSF, and KC were detectable in the serum, levels
of these cytokines did not differ between the subject groups.


52
solution (50 mM Na/Na2 and 0.2% TX-100 in H2O with Anti-protease Complete TM
cocktail (Boehringer)) per mg of wet weight tissue. The tissue was sonicated in the
homogenizing solution for 30 sec on ice and centrifuged at 16,000 g for 15 min at 4 C.
The supernatant was collected and stored at -20 C for Luminex analysis.
Multiplex Microsphere Cytokine Analysis
Commercial kits, Lincoplex mouse cytokine (Lineo, Research, Inc) and a
Luminex 100 LabMAP system (Upstate Biotechnology), were used in attempt to measure
a number of cytokines in the hippocampus and in the serum of IL-2 KO and wild-type
mice. Assays were performed according to the manufacturers instructions, and cytokine
concentrations were calculated using the Softmax program and the linear range on the
standard curve (3.2-10,000 pg/ml). Altogether, we attempted to detect a total of twenty-
two different cytokines and chemokines from the serum and brain homogenates of these
animals. In the serum, there were detectable levels of IL-6. IL-13, kerotinocyte-derived
chemokines (KC), granulocyte-colony stimulating factor (G-CSF), and macrophage
inflammatory protein-1 alpha (MIP-la). In the brain, there were detectable levels of IL-
7, IL-9, IL-12, IL-15, interferon-gamma inducible protein of 10 kD (IP-10), and
monocyte chemoattractant protein-1 (MCP-1). Thus, only those cytokines and
chemokines detected were subjected to statistical analyses. The remainder of the
cytokines and chemokines tested could not be detected in either the serum or the brain
(IFN-y; TNF-oc; IL-la, IL-lf); IL-2, IL-4, IL-5, IL-10, IL-17, GM-CSF, and RANTES).
Labeling Neurogenesis with BrdU
Twelve IL-2 KO (six male; six female) and eleven wild-type (five male; six
female) mice were used to assay for differences in neurogenesis. The procedure for


80
UK Hanisch and R Quirion, 1995b. Interleukin-2 as a neuroregulatory cytokine. Brain
Res Brain Res Rev 21, 246-284.
UK Hanisch, D Seto and R Quirion, 1993. Modulation of hippocampal acetylcholine
release: a potent central action of interleukin-2. J Neurosci 13, 3368-3374.
UK Hanisch, J Neuhaus, R Quirion and H Kettenmann, 1996. Neurotoxicity induced by
interleukin-2: involvement of infiltrating immune cells. Synapse 24, 104-114.
UK Hanisch, W Rowe, S Sharma, MJ Meaney and R Quirion, 1994. Hypothalamic-
pituitary-adrenal activity during chronic central administration of interleukin-2.
Endocrinology 135, 2465-2472.
UK Hanisch, SA Lyons, M Prinz, C Nolte, JR Weber, H Kettenmann and F Kirchhoff,
1997a. Mouse brain microglia express interleukin-15 and its multimeric receptor
complex functionally coupled to Janus kinase activity. J Biol Chem 272, 28853-
28860.
UK Hanisch, J Neuhaus, W Rowe, D Van Rossum, T Moller, H Kettenmann and R
Quirion, 1997b. Neurotoxic consequences of central long-term administration of
interleukin-2 in rats. Neuroscience 79, 799-818.
J Hartikka and F Hefti, 1988. Comparison of nerve growth factor's effects on
development of septum, striatum, and nucleus basalis cholinergic neurons in vitro.
J Neurosci Res 21, 352-364.
YW He and TR Malek, 1998. The structure and function of gamma c-dependent
cytokines and receptors: regulation of T lymphocyte development and homeostasis.
Crit Rev Immunol 18, 503-524.
JC Hedreen, SJ Bacon and DL Price, 1985. A modified histochemical technique to
visualize acetylcholinesterase- containing axons. J Histochem Cytochem 33, 134-
MO.
R Hellweg, C Humpel, A Lowe and H Hortnagl, 1997. Moderate lesion of the rat
cholinergic septohippocampal pathway increases hippocampal nerve growth factor
synthesis: evidence for long-term compensatory changes? Brain Res Mol Brain Res
45, 177-181.
WF Hickey and H Kimura, 1988. Perivascular microglial cells of the CNS are bone
marrow-derived and present antigen in vivo. Science 239, 290-292.
WF Hickey, BL Hsu and H Kimura, 1991. T-lymphocyte entry into the central nervous
system. J Neurosci Res 28, 254-260.
I Horak, 1995. Immunodeficiency in IL-2-knockout mice. Clin Immunol Immunopathol
76, S172-173.


59
Cytokines can enter brain via specific transport mechanisms and through the
circumventricular organs. In this study, we measured levels of the various cytokines of
interest in the serum to determine if levels found in the hippocampus correlated with
those found in the serum. If a particular inflammatory cytokine (e.g., TNFa) had been
found to be significantly elevated in the serum and the hippocampus of IL-2 KO mice, it
would have suggested the possibility that peripheral immune activation associated with
autoimmunity (Schimpl et al., 2002) could account for the presence of the cytokine in the
hippocampus (although increased gene transcription and translation in both the periphery
and the brain could not be ruled out). Although there were significantly increased levels
of IL-6 and MIPla in the circulation of IL-2 KO mice, we did not detect either of these
proteins in the brain. IL-6 is a proinflammatory cytokine that was of particular interest to
us because of its actions in the hippocampus, including effects on neurogenesis (Monje et
al., 2003; Vallieres et al., 2002). Although levels of IL-6 were markedly increased in the
serum, though not measurable in the hippocampus at the tissue homogenate
concentrations used, the unlikely possibility remains that IL-6 could have entered the
brain from the circulation and had functional consequences at concentrations below the
limits of detection of the assay method. Nonetheless, the most parsimonious explanation
is that cytokines from the peripheral circulation of IL-2 KO mice are not the source of the
cytokine alterations found in the hippocampus. Thus, peripheral cytokine dysregulation
associated with autoimmunity in IL-2 KO does not appear to be associated with their
hippocampal pathology.
Our data indicate that loss of IL-2 in the brain results in changes in the production
of several other brain cytokines. Consistent with our hypothesis, IL-15 concentrations


67
was shown to induce dose and time-related behavioral and cognitive changes in 22 out of
44 subjects, including spatial and temporal disorientation (Denicoff et al., 1987). Others
have also noted cognitive changes in patients undergoing IL-2-treatment (Capuron et ah,
2000; Capuron et ah, 2001b; Walker et ah, 1997).
Of particular relevance to this dissertation, IL-2 immunotherapy can induce
cognitive changes, including spatial memory deficits and poor task planning (which may
be partially dependent on spatial memory) as early as five days into therapy (Capuron et
ah, 2001a). Capurons study utilized neuropsychological batteries to assess these
cognitive dysfunctions, but did not investigate the physiological mechanism underlying
them. One hypothesis to explain at least part of the cognitive changes observed is that
the chronic IL-2 treatment used on these patients may affect the septohippocampal
system in a similar way that loss of IL-2 causes changes in IL-2 KO animals. Deletion of
IL-2 having a similar physiological effect as administration of exogenous IL-2 may seem
counterintuitive; however, in at least one population of neurons (i.e., the septal neurons),
IL-2 has potent biphasic modulatory actions on ACh release with low sub-pM IL-2
concentrations enhancing and higher nM concentrations inhibiting release (Hanisch et al.,
1993; Seto et al., 1997). Thus, under the proposed hypothesis, a loss of IL-2 could lead
to a reduction in the stimulatory action on cholinergic neurons, whereas an
overabundance of exogenous IL-2 could inhibit release of ACh. Of course, more
investigation would be necessary to clarify that hypothesis.
Caveats and Future Directions
In Study 1, we examined the effects of IL-2 KO on the cholinergic
septohippocampal neurons. Many of our findings were extended upon in Study 2.
However, we hypothesized that the lack of any change in the cholinergic fiber density in


CHAPTER 4
INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND
NEUROGENESIS IN THE HIPPOCAMPUS
Introduction
Receptors for interleukin-2 (IL-2) are enriched in the hippocampal formation, and
many of the most prominent neurobiological functions of this cytokine occur in the
hippocampus (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991;
Petitto and Huang, 1994, 2001; Petitto et al., 1998). Previous studies from our laboratory
have found that IL-2 knockout (KO) mice exhibit significantly lower numbers of medial
septum and vertical limb of the diagonal band of Broca (MS/vDB) cholinergic cell
bodies, a reduction in the distance across the granular cell layer (GCL) of the
infrapyramidal (IP) blade of the dentate gyrus (DG), and decreased fiber length and
neuronal cell number in the IP-GCL of the DG (Beck et al., 2004; Beck et al., 2002;
Petitto et al., 1999). These neurobiological alterations appear to be related to
abnormalities in learning and memory performance and sensory motor gating in IL-2 KO
mice (Cushman et al., 2004; Petitto et al., 1999).
Because IL-2 has been shown to possess various neurotrophic and
neuromodulatory effects on hippocampal neurons in vitro (Awatsuji et al., 1993; Bianchi
et al., 1995; Pauli et al., 1998; Plata-Salaman and ffrench-Mullen, 1993; Sarder et al.,
1996; Sarder et al., 1993; Tancredi et al., 1990), our original working hypothesis was that
the alterations exhibited by IL-2 KO mice are due to the absence of IL-2s neurotrophic
actions on hippocampal neurons during development. More recent data from our
45


33
CTAGGCCACAGAATTGAAAGATCT-3 and 5-
GTAGGTGG AAAATTCTAGC ATC ATCC-3 .
Immunohistochemistrv
For Parv stereological cell counts, six animals per group were used and for 21-
day-old ChAT stereological cell counts, seven animals per group were used. Each animal
was anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20
mg/ml): acepromazine (10 mg/ml) at a dose of 0.015 ml injection cocktail/g body weight
and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered
saline (PBS). The brains and spleens were removed and fixed overnight in 4%
paraformaldehyde, followed by overnight equilibration in 30% sucrose cryoprotective
solution, and then, were snap frozen in isopentane (-80C) for storage. The spleens were
weighed to assay for relative splenomegaly of IL-2 KO vs. wild-type mice. The brains
were equilibrated to -20C prior to cryostat sectioning into 50 pm slices in the coronal
plane, collected into individual wells of polystyrene 24-well plates (NUNC 1147), and
stored free-floating at 4C in PBS for histochemistry. Every third section was processed
for Parv or ChAT immunohistochemistry, or cresyl violet Nissl staining.
Free-floating 50-pm sections were labeled for Parv and ChAT-
immunohistochemistry as described previously (Beck et al., 2002). Briefly, they were
incubated for 20 minutes in 1% hydrogen peroxide (H2O2) to quench endogenous
peroxidative activity. The sections were then washed and blocked for 1 hr in 200 pl/well
3% normal goat serum (NGS). After this incubation, the sections were incubated
overnight in the primary antibody, rabbit anti-ChAT (Chemicon; 1:2000 in PBS with
0.3% Triton X-100 and 1% NGS, 200 pl/well) or rabbit anti-Parv (Chemicon; 1:1000 in
PBS with 0.3% Triton X-100 and 1% NGS, 200 pl/well). The next day, the sections were


89
WJ Streit, MB Graeber and GW Kreutzberg, 1988. Functional plasticity of microglia: a
review. Glia 1, 301-307.
WJ Streit, SL Semple-Roxyland, SD Hurley, RC Miller, PG Popovich and BT Stokes,
1998. Cytokine mRNA profiles in contused spinal cord and axotomized facial
nucleus suggest a beneficial role for inflammation and gliosis. Exp Neurol 152, 74-
87.
K Sugamura, H Asao, M Kondo, N Tanaka, N Ishii, K Ohbo, M Nakamura and T
Takeshita, 1996. The interleukin-2 receptor gamma chain: its role in the multiple
cytokine receptor complexes and T cell development in XSCID. Annu Rev
Immunol 14, 179-205.
H Super and E Soriano, 1994. The organization of the embryonic and early postnatal
murine hippocampus. II. Development of entorhinal, commissural, and septal
connections studied with the lipophilic tracer Dil. J Comp Neurol 344, 101-120.
T Takeshita, H Asao, K Ohtani, N Ishii, S Kumaki, N Tanaka, H Munakata, M Nakamura
and K Sugamura, 1992. Cloning of the gamma chain of the human IL-2 receptor.
Science 257, 379-382.
V Tancredi, C Zona, F Velotti, F Eusebi and A Santoni, 1990. Interleukin-2 suppresses
established long-term potentiation and inhibits its induction in the rat hippocampus.
Brain Res 525, 149-151.
LJ Thai, E Gilbertson, DM Armstrong and FH Gage, 1992. Development of the basal
forebrain cholinergic system: phenotype expression prior to target innervation.
Neurobiol Aging 13, 67-72.
JJ Thune and B Pakkenberg, 2000. Stereological studies of the schizophrenic brain. Brain
Res Brain Res Rev 31, 200-204.
L Vallieres, IL Campbell, FH Gage and PE Sawchenko, 2002. Reduced hippocampal
neurogenesis in adult transgenic mice with chronic astrocytic production of
interleukin-6. J Neurosci 22, 486-492.
F Villemain, T Owens, T Renno and A Beaudet, 1991. Localization of mRNA for
interleukin-2 (IL-2) in mouse brain by in situ hybridization (Abstract). Society of
Neuroscience Abstracts 17, 1199.
M Virgili, A Contestabile and O Bamabei, 1991. Postnatal maturation of cholinergic
markers in forebrain regions of C57BL/6 mice. Brain Res Dev Brain Res 63, 281-
285.
PJ Waguespack, WA Banks and AJ Kastin, 1994. Interleukin-2 does not cross the blood-
brain barrier by a saturable transport system. Brain Res Bull 34, 103-109.


This dissertation was submitted to the Graduate Faculty of the College of Medicine
and to the Graduate School and was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
August 2004
Dean, College of Medicine
Dean, Graduate School


CHAPTER 1
BACKGROUND AND SIGNIFICANCE
Cytokine-Brain Interactions
The landmark studies of Ader and Cohen demonstrating that immune physiology
could be behaviorally conditioned led to the systematic investigation of the complex
interaction between the central nervous and immune systems (Ader and Cohen, 1975;
Ader et al., 1982). Although the central nervous system (CNS) and peripheral immune
system were once considered functionally incompatible entities separated by a nearly
impermeable protective blood-brain-barrier (BBB), it is now known that there is
bidirectional communication and modulation between these two systems. Cytokines
have emerged as important mediators of various processes in the CNS. Their effects
range from neuroinflammation in experimental autoimmune encephalomyelitis (EAE)
and viral infection of the brain to neurobiological processes such as hypothalamic-
pituitary axis (HPA) regulation, induction of fever, sleep, analgesia, feeding behavior,
and cognition (for reviews see Ader et ah, 2001; Dunn, 2002; Wilson et ah, 2002).
Cytokines produced both within and outside of the CNS can exert their effect on
brain cells (Dunn, 2002; Streit et ah, 1998). The work of Banks and others show that the
BBB acts as a selective filter for peripheral cytokines (for a review see Banks et ah,
2002). Multiple studies support the ability of cytokines (e.g., IL-la and -(3, IL-2, IL-6,
IFN-a and -y, TNF-a) to cross the BBB via different transport mechanisms (Banks et ah,
1994; Banks et ah, 1991; Gutierrez et ah, 1993; Pan et ah, 1997; Waguespack et ah,
1994), and via the leaky circumventricular organs (CVO), four brain regions outside of
1


This dissertation is dedicated to my wife Laura whose love, support, and occasional
nagging help to keep me focused on my goals.


worked in the lab both past and present: Brent, Andrew, David, Jeannette, Dan, Jesse, and
Grace. In addition to taking care of the upkeep of the lab, working with them was a
pleasure. In particular, I would like to single out Andrew. He was my dearest friend here
in Gainesville. I wish him the best of luck in his career as a medical doctor and hope that
we will always remain friends after I leave Gainesville.
Outside of the laboratory, I thank my other friends for being my support structure.
There has never been a better collection of in-the-closet geeks than Coleman, Dan, Andy,
Ryan, Charles, Chris, Curtis, Nick, and Jason. They are simply the best. I would also
like to thank Mozart, Michelangelo, and Sage simply because not nearly enough people
thank their dogs. Obviously, they would be more likely to chew on this dissertation than
read it, but I would like anyone else that does see this to know that few people are
capable of matching the unconditional love that a dog has for its owner.
I also thank my family. My mother has always instilled the value of education into
me. My fathers love and sense of humor were crucial in the development of my
personality. My sisters, Jackie and Judy, will always be among my closest friends. My
Aunt Kathy and Uncle Jimmy are tremendous people who have always supported me and
though they are family by marriage, our bond is stronger than blood.
Finally, I thank my lovely wife Laura. Without her, my life would be incomplete
(though despite her beliefs, I could still drive effectively without her commentary from
the passenger seat). She has pushed me when I needed motivation and comforted me
when I need support. She is my heart and soul and I could not have achieved this
dissertation without her.
IV


3
affinity heterodimer consisting of the yc and P subunits and a high affinity heterotrimer
comprised of all three subunits (a, P, and yc) (Ringheim et al., 1991; Takeshita et al.,
1992). The P and yc both possess intracellular signaling domains and in their
heterodimeric form have a IQ of 109, whereas the addition of the a subunit forms a
heterotrimer with higher affinity (IQ of 10n) for IL-2 (Nakamura et al., 1994; Nelson et
al., 1994).
In the peripheral immune system, where the physiological properties of IL-2 are
most well-characterized, IL-2 has multiple biological functions, including natural killer
(NK) cell activation, T lymphocyte activation, as well as B lymphocyte differentiation
(for review see Waldmann, 2002). The creation of a transgenic knockout mouse model
for this cytokine suggests that the most important function of IL-2 is the maintenance of
immune self-tolerance (Schmitt et al., 1994). IL-2 knockout (KO) mice develop
autoimmune symptoms commonly including inflammatory bowel disease similar to
ulcerative colitis in humans and advanced hemolytic anemia, although the manifestation
of the phenotype is dependent on the genetic background of the knockout mice (Horak,
1995, 1996). The autoimmunity that develops when the IL-2 gene is deleted is T cell
dependent (Ma et al., 1995). More recently, it has become apparent that IL-2 plays a
major role in limiting T cell responses via the development of regulatory T cells
(CD4+CD25* T reg cells) and other mechanisms that promote self tolerance and suppress
T cell responses in vivo (Nelson, 2004). Though extensive research has characterized the
effects of IL-2 in the peripheral immune system, increasing evidence indicates that IL-2
may potentially impact the central nervous system (CNS). The focus of this research
project was to characterize these potential actions of IL-2 in the brain.


41
differences in GABAergic cell number, since we did not detect any changes in adult mice
using the same marker. One potential explanation for the loss of cholinergic neurons in
the MS/vDB may be a failure in maintenance or survival in the late stages of, or after,
development. Other studies have found decreases in cholinergic enzyme activity (i.e.,
ChAT and AChE) between postnatal days 30-60 in normal rats (Thai et al., 1992), and
postnatal days 60-150 in C57BL/6 mice (Virgili et al., 1991). The IL-2 KO mice may
potentially be more susceptible to this loss of cholinergic activity during adulthood,
which could lead to the previously observed deficiencies in 8-12-week-old IL-2 KO
animals.
An alternate explanation for the different cholinergic effects seen in 21-day-old
vs. 8-12-week-old animals is that the loss of IL-2 may be secondary to the effects of
autoimmunity present in adult IL-2 KO animals. Though we cannot completely rule out
this possibility, we have previously failed to find discemable levels of infiltrating
lymphocytes or clear signs of gliosis in the brains of IL-2 KO animals (Petitto et al.,
1999). More research is necessary to further address this issue.
Another finding of this study was a significant decrease in neuronal cell number
in the DP-GCL, but not the SP-GCL. This decrease is consistent with the in vitro studies
showing a potent neurotrophic effect of IL-2 on hippocampal neurons (Awatsuji et al.,
1993; Sardcr et al., 1996; Sarder et al., 1993). Furthermore, these data are supported by
our previous findings that IL-2 KO mice exhibited a reduced distance across the IP-GCL
(Beck et al., 2002), and that the IP mossy fiber length of IL-2 KO mice is shorter than
wild-type controls (Petitto et al., 1999). The reductions in distance across the IP-GCL
could also potentially be explained by increased density, but not number, of cells or


22
A)
Average AChE Intensity of CA1
B)
Average AChE Intensity of CA3
stance (% of Total) Dtaence(% of Total)
C)
Average AChE Intensity of the Suprapyramldal Limb
of the Dentate Gyrus
D)
Average AChE Intensity of the Infrapyramidal Limb
of the Dentate Gyrus
IL-? ¡Miockout
O WHd*Typ
Fig.2-4. IL-2 knockout mice do not differ in measures of average intensity of AChE-
staining across CA1, CA3b, and the suprapyramidal and infrapyramidal layers of
the DG. Each curve represents the mean of 9 wild-type (open circles) or 9 IL-2
knockout mice (closed circles). Alv=alveus, GCL=granular cell layer
(IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus,
Lmol=lacunosum moleculare, Mol=molecular layer, Pi=pial surface, SO=stratum
oriens, SP=septum pellucidum, SR=stratum radiatum. A) The CA1 curve shows
the average intensity from the Alv to the Hif. The arrows delineate the transitions
between the hippocampal substructures including the SO, SP, SR, and Lmol. B)
The pattern of the peaks and valleys of the CA3 region curve is similar to that of
the CAL C) The intensity curve of SP-GCL begins at the Hif and terminates at
the dorsal border of the IP-GCL. Arrows delimit the borders of the Mol, SP-GCL,
and Hil. D) The curve representing the IP-GCL begins at the ventral border of the
SP-GCL and continues to the pial surface. The second and third solid arrow
define the borders in wild-type mice between the Hil and IP-GCL and the IP-GCL
and Mol, respectively. The broken arrows define these same borders in IL-2
knockout mice.


26
critical periods of neurodevelopment. Release of ACh by developing neurons has been
shown to be important for growth cone guidance (Zheng et al., 1994), neuronal growth
and differentiation, synaptic plasticity (Lauder and Schambra, 1999), and survival of
newly developed neurons (Knipper and Rylett, 1997). In fact, some evidence indicates
that ACh released from developing neurons may engage in a positive feedback
mechanism with nerve growth factor (NGF) (Knipper et al., 1994), a member of the
neurotrophin family that is essential for the normal development of septal cholinergic
neurons (Arimatsu and Miyamoto, 1991; Hartikka and Hefti, 1988; Mobley et al., 1986;
Ruberti et al., 2000). In a series of studies, Quirions laboratory has demonstrated that
IL-2 is among the most potent modulators of ACh release from mature brain slices and
fetal neurons in vitro, and can upregulate ChAT in fetal septal neurons in culture
(Hanisch et al., 1993; Mennicken and Quirion, 1997; Seto et al., 1997). It is therefore
possible that the loss of such potent actions of IL-2 during development could account, in
part, for the cytoarchitectural alterations found in this study. Indeed, both IL-2s
neurotrophic effects and action on cholinergic release may well be operative and
interactive with one another. Nevertheless, the IL-2 KO mice do not exhibit complete
loss of septal cholinergic neurons suggesting that the effects of IL-2 on MS/vDB neurons
are likely secondary to other trophic factors like NGF.
These experiments do not enable us to differentiate between the contributions of
the loss of central versus peripheral IL-2, and thus, it remains to be determined whether
these septohippocampal cholinergic abnormalities are due primarily to the absence of
central, peripheral, or a combination of both sources of IL-2. There is some evidence that
endogenous IL-2 may be produced in neuronal areas of the mammalian hippocampal


87
SA Rosenberg, JC Yang, SL Topaban, DJ Schwartzentruber, JS Weber, DR Parkinson,
CA Seipp, JH Einhom and DE White, 1994. Treatment of 283 consecutive patients
with metastatic melanoma or renal cell cancer using high-dose bolus interleukin 2.
Jama 271, 907-913.
R Rozental, MF Mehler, M Morales, AF Andrade-Rozental, JA Kessler and DC Spray,
1995. Differentiation of hippocampal progenitor cells in vitro: temporal expression
of intercellular coupling and voltage- and ligand-gated responses. Dev Biol 167,
350-362.
F Ruberti, S Capsoni, A Comparini, E Di Daniel, J Franzot, S Gonfloni, G Rossi, N
Berardi and A Cattaneo, 2000. Phenotypic knockout of nerve growth factor in adult
transgenic mice reveals severe deficits in basal forebrain cholinergic neurons, cell
death in the spleen, and skeletal muscle dystrophy. J Neurosci 20, 2589-2601.
T Saarelainen, S Vaittinen and E Castren, 2001. trkB-receptor activation contributes to
the kainate-induced increase in BDNF mRNA synthesis. Cell Mol Neurobiol 21,
429-435.
N Sakai, S Kaufman and S Milstien, 1995. Parallel induction of nitric oxide and
tetrahydrobiopterin synthesis by cytokines in rat glial cells. J Neurochem 65, 895-
902.
B Sakic, H Szechtman and JA Denburg, 1997. Neurobehavioral alterations in
autoimmune mice. Neurosci Biobehav Rev 21, 327-340.
B Sakic, H Szechtman, S Denburg, R Carbotte and JA Denburg, 1993. Spatial learning
during the course of autoimmune disease in MRL mice. Behav Brain Res 54, 57-
66.
RP Saneto, A Altman, RL Knobler, HM Johnson and J de Vellis, 1986. Interleukin 2
mediates the inhibition of oligodendrocyte progenitor cell proliferation in vitro.
Proc Natl Acad Sci U S A 83, 9221-9225.
M Sarder, H Saito and K Abe, 1993. Interleukin-2 promotes survival and neurite
extension of cultured neurons from fetal rat brain. Brain Res 625, 347-350.
M Sarder, K Abe, H Saito and N Nishiyama, 1996. Comparative effect of IL-2 and IL-6
on morphology of cultured hippocampal neurons from fetal rat brain. Brain Res
715,9-16.
J Satoh, K Kurohara, M Yukitake and Y Kuroda, 1998. Interleukin-15, a T-cell growth
factor, is expressed in human neural cell lines and tissues. J Neurol Sci 155, 170-
177.
A Schimpl, I Berberich, B Kneitz, S Kramer, B Santner-Nanan, S Wagner, M Wolf and T
Hunig, 2002. IL-2 and autoimmune disease. Cytokine Growth Factor Rev 13, 369-
378.


44
to detect significant levels of leukocyte infiltration or gliosis in IL-2 KO mice brains),
further study is necessary to assess this possibility, as factors such as IL-2 induced
cytokine dysregulation or autoantibodies could contribute to the hippocampal alterations
in adult IL-2 KO mice.


66
have depended on in vitro models and/or exogenous administration of IL-2 into theCNS
in vivo. Whereas studies such as these can yield valuable information, in vitro models
often fail to mimic the complexity of an in vivo system (e.g., lack of glia, different
organization of neurons, etc.) and the manipulated in vivo models are subject to
experimental design complications (e.g., disruption of the BBB, determination of
physiologically meaningful doses of exogenously administered cytokines, etc.). We
would be remiss, however, to claim that the IL-2 KO model was completely without
complications itself (e.g., autoimmunity such as inflammatory bowel disease), but we feel
that the most informative way to address a hypothesis is to examine it in multiple
different ways. Thus, taken together with the previous studies in the literature, this
dissertation reveals an interesting relationship between IL-2 and the septohippocampal
system.
The goal of this study was to elicit whether IL-2 KO impacted the
septohippocampal system, but not to design a treatment or study a clinical disorder.
Thus, at its core, this dissertation was basic research. Whereas the pursuit of knowledge
is a noble goal, invariably, a simple question often finds itself in the front of our minds:
how is this new knowledge clinically relevant and how can we use it?
Already, IL-2 is being used clinically to treat, or being studied as a treatment for, a
number of disorders, including cancer (Atkins et al., 1999; Davis and Gillies, 2003; Fyfe
et al., 1995; Guirguis et al., 2002; Parkinson et al., 1990; Rosenberg et al., 1985;
Rosenberg et al., 1994), and HIV infection (Armstrong and Kazanjian, 2001; Mitsuyasu,
2001; Smith, 2001). The aforementioned treatments are not, however, completely
without side effects. In a landmark study, systemic IL-2 treatment for cancer patients


14
Genotyping Using PCR
The genotypes of all mice were determined by the PCR. PCR reactions were
performed using a 25 pi total reaction volume containing 1 pM each of forward and
reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP, 0.3 pi Taq DNA
polymerase and amplified using a thermal cycler with a heated evaporation cover
(Ericomp). The cycling parameters were hot start 95C (3min), denaturing 94C (30
sec), annealing 64C (30 sec), extension 72C (45 sec) with a final extension step of 4
min. Thirty cycles were used for these experiments. The 5 and 3 primers for the IL-2
KO (500 bp knockout band amplified) were 5-TCGAATCGCCAATGACAAGACGCT-
3 and 5-GTAGGTGGAAATTCTAGCATCATCC-3\ The 5 and 3primers for the wild
type (324 bp wild type band amplified) were 5-
CTAGGCCACAGAATTGAAAGATCT-3 and 5-
GTAGGTGGAAAATTCTAGC ATC ATCC-3 .
ChAT Immunohistochemistry
Free-floating 40-pm sections were incubated for 20 minutes in 1% hydrogen
peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then
washed twice in PBS and blocked for 1 hr in 200 pl/well 3% normal goat serum (NGS).
After this incubation, the sections were incubated overnight in the primary antibody,
rabbit anti-ChAT (Chemicon AB143; 1:2000 dilution in PBS with 0.3% Triton X-100
and 1% NGS, 200 pl/well). The next day, the sections were washed twice in PBS and
incubated overnight in the secondary antibody, biotinylated goat anti-rabbit IgG (Sigma
B-7389; 1:1000 dilution in PBS with 0.3% TX-100 and 1% NGS). The sections were
then washed twice in PBS and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS)


25
(Cassel et al., 1997; Gage and Bjorklund, 1987; Gage et al., 1984; Gage et al., 1983b),
and in animal disease models such as Alzheimers transgenic mice (Bronfman et al.,
2000). Future studies are needed to address this issue more directly.
Another significant finding of the current study was that IL-2 KO mice exhibited
structural alterations in the distance across the IP-GCL. The neurons of the GCL have
been associated with learning and memory (Collier and Routtenberg, 1984; Conrad and
Roy, 1993; McLamb et al., 1988; Nanry et al., 1989; Walsh et al., 1986), and are also a
target for septohippocampal cholinergic axon termination (Makuch et al., 2001). In situ
hybridization studies have found the GCL to be enriched in IL-2 receptors (Petitto and
Huang, 2001; Petitto et al., 1998), supporting a possible role for IL-2 in the observed
structural alterations. Also, GCL development progresses from the SP layer to the IP
layer (Bayer, 1980). The differences seen in the IL-2 KO mice in this study may indicate
a failure of these late stage granule cells to fully develop or survive. Whether these
structural changes are due to a reduced number of GCL cells or a decrease in the cell
body size of these neurons requires further investigation.
The most likely mechanism whereby loss of IL-2 results in these changes in the
septohippocampal cholinergic system would appear to be due to the absence of its
neurotrophic actions during development. As noted earlier, IL-2 enhances neurite
extension and survival of cultured fetal septal and hippocampal neurons (Awatsuji et al.,
1993; Hanisch and Quirion, 1995a; Sarder et al., 1996; Sarder et al., 1993), and thus, the
absence of these intrinsic effects of IL-2 could account for the observed neuroanatomical
alterations. Another mechanism that may account for these findings is the possibility that
the loss of endogenous IL-2 may result in lower levels of tonic ACh release during


88
E Schmitt, T Germann, S Goedert, P Hoehn, C Huels, S Koelsch, R Kuhn, W Muller, N
Palm and E Rude, 1994. IL-9 production of naive CD4+ T cells depends on IL-2, is
synergistically enhanced by a combination of TGF-beta and IL-4, and is inhibited
by IFN-gamma. J Immunol 153, 3989-3996.
R Schopke, DP Wolfer, HP Lipp and MC Leisinger-Trigona, 1991. Swimming
navigation and structural variations of the infrapyramidal mossy fibers in the
hippocampus of the mouse. Hippocampus 1, 315-328.
H Schwegler and WE Crusio, 1995. Correlations between radial-maze learning and
structural variations of septum and hippocampus in rodents. Behav Brain Res 67,
29-41.
H Schwegler, WE Crusio, HP Lipp and B Heimrich, 1988. Water-maze learning in the
mouse correlates with variation in hippocampal morphology. Behav Genet 18, 153-
165.
H Schwegler, M Boldyreva, R Linke, J Wu, K Zilles and WE Crusio, 1996a. Genetic
variation in the morphology of the septo-hippocampal cholinergic and GABAergic
systems in mice: II. Morpho-behavioral correlations. Hippocampus 6, 535-545.
H Schwegler, M Boldyreva, M Pyrlik-Gohlmann, R Linke, J Wu and K Zilles, 1996b.
Genetic variation in the morphology of the septo-hippocampal cholinergic and
GABAergic system in mice. I. Cholinergic and GABAergic markers. Hippocampus
6, 136-148.
BM Segal, BK Dwyer and EM Shevach, 1998. An interleukin (IL)-10/IL-12
immunoregulatory circuit controls susceptibility to autoimmune disease. J Exp Med
187,537-546.
D Seto, S Kar and R Quirion, 1997. Evidence for direct and indirect mechanisms in the
potent modulatory action of interleukin-2 on the release of acetylcholine in rat
hippocampal slices. Br J Pharmacol 120, 1151-1157.
NM Shah, AK Groves and DJ Anderson, 1996. Alternative neural crest cell fates are
instructively promoted by TGFbeta superfamily members. Cell 85, 331-343.
M Shimojo, Y Imai, K Nakajima, S Mizushima, A Uemura and S Kohsaka, 1993.
Interleukin-2 enhances the viability of primary cultured rat neocortical neurons.
Neurosci Lett 151, 170-173.
KA Smith, 2001. Low-dose daily interleukin-2 immunotherapy: accelerating immune
restoration and expanding HIV-specific T-cell immunity without toxicity. Aids 15
Suppl 2, S28-35.
LS Sorkin, WH Xiao, R Wagner and RR Myers, 1997. Tumour necrosis factor-alpha
induces ectopic activity in nociceptive primary afferent fibres. Neuroscience 81,
255-262.


16
AChE Histochemistry
AChE histochemistry was used as a marker of cholinergic innervation of the
hippocampus (Woolf et al., 1984). Brain sections were collected in individual wells of
24-well plates containing 250 plAvell 0.1 M pH 6.0 acetate buffer (AB). The sections
were washed twice with AB, then placed in 200 pi preincubation solution consisting of
aqueous 5 mM sodium citrate, 3 mM cupric sulfate, and 0.5 mM potassium ferricyanide.
The sections were incubated for 20 minutes at room temperature on a shaker a low speed.
After the preincubation period, 200 pi of the incubation solution was added to each well
consisting of the same make-up as the preincubation solution supplemented with 4.84
mM acetylthiocholine iodide and 0.4 mM ethopropazine. The multi-well plate was
packed on top of crushed ice and microwaved at 200 W for 2 minutes. The solution was
then removed and the sections were washed twice in 0.05 M TRIS pH 7.6 buffer
followed by AB. The reaction product was intensified with 0.5 mg/ml DAB, 2.5% nickel
sulfate, and 0.01 % H2O2 in AB for 5-7 min or until definitive staining could be detected
in the hippocampal subregions. Sections were then mounted on slides, dehydrated in
grade ethanol washes, cleared in xylenes, and coverslipped for imaging.
Cholinergic Stereologv
Stained cholinergic neuronal somata of the MS/vDB were counted using the
software MCID 5.1 and the three-dimensional counting box (optical dissector) method
described by Williams and Rakic (Williams and Rakic, 1988). All stereology was
performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2
microscope with a motorized x-y stage made by Imaging Research, Inc. The latter is
capable of making movements as fine as 0.1 pm. Every third section through the
anterior-posterior extent of the septal region was sampled. The regions to be counted


ACKNOWLEDGMENTS
Traditionally, most acknowledgement sections begin by thanking ones advisor; in
this case, such tradition is most warranted. As such, I thank Dr. John Petitto for being a
wonderful mentor. Though not the stealthiest individual in the world with his penchant
for crying out ones name (or impromptu nickname) exuberantly as soon as one enters
anywhere within his field of vision, he is among the kindest, most supportive,
enthusiastic, and knowledgeable mentors for which any graduate student could wish.
Next, I would like to thank each member of my committee. I feel fortunate to be advised
by such a great selection of knowledgeable and friendly people. Dr. Mike Kings
easygoing personality and extensive knowledge of all things stereological and cholinergic
have proven invaluable in my studies. Dr. Mark Lewis advice on statistics and
experimental design, as well as his sense of humor, has been most appreciated. Dr. Jake
Streit, in addition to his ability to make me laugh, always reminded me that there is more
than one kind of cell in the brain. Dr. Mark Atkinson, always amiable and approachable,
helped guide me in the immunology aspect of neuroimmunology.
I also thank Dr. Huang Zhi. I cannot overstate how much I valued his advice on
experiments and his daily conversations on topics ranging from basketball to politics to
Hong Kong movies. I wish him the best of luck in his medical residency and his future
as a psychiatrist. I would also like to thank Clive Wasserfall and Fletcher Schwartz for
teaching me how to use the Luminex technology and Tim Vaught for teaching me the ins
and outs of multiple microscopy techniques. I also thank the many technicians that
iii


CHAPTER 5
GENERAL DISCUSSION
Summary of the Overall Findings
The overall purpose of this dissertation research was to investigate the impact of
IL-2 on the septohippocampal system by observing changes in the basal physiology and
structure of IL-2 KO mice brains vs. their wild-type littermates. During these
experiments, we chose not to surgically, chemically, or in any other way experimentally
manipulate these animals beyond the genetic knockout of IL-2, so that the effects that we
detected could be attributed to loss of IL-2 and not to confounding experimental
techniques (e.g., disruption of the BBB). We understand, however, that some
experimental manipulations could be useful in the long term and discuss the topic more in
the Caveats and Future Directions section of towards the end of this chapter.
In the first study in Chapter 2, we determined that IL-2 KO mice suffered a
significant 26% loss of cholinergic neurons in the MS/vDB relative to wild-type
littermates. This loss of cholinergic somata was not reflected by a similar loss of
cholinergic fiber density in the hippocampal projection fields of the septal cholinergic
neurons. Moreover, the deficits observed were not a general effect on all cholinergic
neurons of the brain, as the cholinergic neurons of the striatum in the IL-2 KO animals
studied did not appear to be affected. There was, however, a reduction in the distance
across the IP, but not SP, GCL of the dentate gyrus.
In Chapter 3, we extended upon these findings by showing that the loss of
cholinergic neurons appeared to occur later in development or during adulthood as a
64


47
cytokines in this organ (Autenrieth et al., 1997; Meijssen et al., 1998). Moreover, these
cytokine changes and the onset of inflammatory bowel disease are preceded by increased
gene expression of IL-15, which shares the same signal transducing receptor subunits
with IL-2 (Meijssen et al., 1998). Thus, it is possible that the immune dysregulation in
the brain of IL-2 KO mice may be induced by activated T cells and/or proinflammatory
cytokines (e.g., IL-1, TNFa, IL-6) from the periphery crossing the blood-brain-barrier
(BBB).
By contrast, IL-2 may lead to neuroimmunological changes that do not involve
peripheral immune cells. Rather than peripheral T cells and serum cytokines entering the
brain, an alternative hypothesis that may account for the hippocampal differences
observed in P21 versus adult IL-2 KO mice may be that the absence of IL-2 reduces the
trophic support of hippocampal neurons as a result of dysregulation of other brain-
derived cytokines. Thus, loss of IL-2 in the brain could in turn modify the normal
neuroimmunological status of the brain by modifying the normal expression of brain
cytokines such as IL-15. Since IL-2 can modify the release of certain cytokines from
lymphoid cells (Lauwerys et al., 2000; McDyer et al., 2002), similar actions could occur
in brain cells that produce cytokines (e.g., microglia, astrocytes). Alterations in the
production of brain cytokines important in normal brain physiology could alter the
integrity of hippocampal neurons by decreasing levels of classic neurotrophins and/or
neurotrophic cytokines on the one hand, or elicit inflammatory-like neurodegenerative
processes within the brain on the other.
The present study therefore sought to test the hypothesis that IL-2 gene deletion
results in neuroimmunological changes in the hippocampus by examining the possible


BIOGRAPHICAL SKETCH
Ray Dennis Beck, Jr. was bom in York, PA, to Daisy and the appropriately named
Ray Dennis Beck, Sr. Realizing the error of his ways early in life, Ray moved to
Houston, TX, at the age of two, lending credence to the old Texas saying, I wasnt bom
in Texas, but I got here as fast as I could. He attended Oak Ridge High School until the
age of 16. Ranked 14 overall in his class, he elected to forego his senior year to attend
Simons Rock College of Bard in Great Barrington, MA. Whereas some 16-year-old
adolescents are mature enough to pursue a college education while resisting distractions
like geographically convenient buildings filled with members of the opposite sex, a ready
supply of fermented beverages, and complete lack of parental supervision Ray was not.
He returned to Houston after his first year of college with a less than stellar GPA intent
on taking a year off from school. During this year (comprised of-1,825 days), he held
various jobs ranging from perfume salesman to waiter at various restaurants. During one
of the latter jobs, he met his present wife, Laura Frakey. Inspired by her enthusiasm for
education, Ray enrolled in the University of Houston, while maintaining full-time
employment to pay for school. He graduated cum laude with a B.S. in biology. Ray and
Laura moved to Gainesville, FL, to attend University of Florida graduate programs in
neuroscience and psychology, respectively. With the completion of his Ph.D., Ray is
proud to be one of the most educated high school dropouts that anyone is ever likely to
meet. Ray is an avid follower of movies and his other interests can be classified as all
things geeky (e.g., computers, roleplaying, video games, etc.).
92


24
1994; Koch, 1996a), and the aforementioned influence of IL-2 on parameters of
cholinergic function, the present findings are consistent with alterations in these
behavioral measures that we have reported previously in IL-2 KO mice (Petitto et al.,
1999).
In this study, the number of cholinergic cell bodies in the MS/vDB of wild-type
mice were comparable to those reported previously for C57BL/6 mice (Schwegler et al.,
1996b). Cholinergic somata were reduced by 26% in IL-2 KO mice as compared to wild-
type mice. This was not a general effect on cholinergic neurons in the brain, however, as
striatal ChAT-positive neurons were not significantly affected. This finding is also
consistent with previous research showing that exogenous IL-2 has potent effects on ACh
release from septohippocampal neurons, whereas cholinergic intemeurons in the striatum
do not respond to IL-2 (Hanisch et al., 1993). One potential caveat of using ChAT as a
marker of cholinergic neurons is that differences in cell counts may be due to a decrease
in ChAT labeling intensity rather than a reduction in cholinergic neurons (Ward and
Hagg, 2000). This would appear unlikely, however, as we did not note any appreciable
differences in the staining intensity between groups.
Alterations in AChE-staining intensity did not differ as we hypothesized. It is not
likely that the unexpected lack of difference in AChE-staining intensity would be
attributable to the length normalization technique used, as the patterns of the average
intensity curves seen in Figure 2-4 were remarkably similar between the groups. This
unexpected result may, however, be due to compensatory sprouting of the surviving
MS/vDB neurons during development. Indeed, numerous other studies have found that
septohippocampal neurons can undergo compensatory sprouting in response to injury


90
T Waldmann, 2002. The contrasting roles of IL-2 and IL-15 in the life and death of
lymphocytes: implications for the immunotherapy of rheumatological diseases.
Arthritis Res 4 Suppl 3, S161-167.
TA Waldmann, S Dubois and Y Tagaya, 2001. Contrasting roles of IL-2 and IL-15 in the
life and death of lymphocytes: implications for immunotherapy. Immunity 14. 105-
110.
LG Walker, MB Walker, SD Heys, J Lolley, K Wesnes and O Eremin, 1997. The
psychological and psychiatric effects of rIL-2 therapy: a controlled clinical trial.
Psychooncology 6, 290-301.
TJ Walsh, DW Schulz, HA Tilson and DE Schmechel, 1986. Colchicine-induced granule
cell loss in rat hippocampus: selective behavioral and histological alterations. Brain
Res 398, 23-36.
NL Ward and T Hagg, 2000. BDNF is needed for postnatal maturation of basal forebrain
and neostriatum cholinergic neurons in vivo. Exp Neurol 162, 297-310.
PC Wilkinson and FY Liew, 1995. Chemoattraction of human blood T lymphocytes by
interleukin-15. J Exp Med 181, 1255-1259.
RW Williams and P Rakic, 1988. Three-dimensional counting: an accurate and direct
method to estimate numbers of cells in sectioned material. J Comp Neurol 278,
344-352.
CJ Wilson, CE Finch and HJ Cohen, 2002. Cytokines and cognitionthe case for a head-
to-toe inflammatory paradigm. J Am Geriatr Soc 50, 2041-2056.
O Witzke, T Winterhagen, A Kribben, T Philipp, K Mann and W Reinhardt, 2003.
Interleukin-2 given to asymptomatic HIV-infected individuals leads to an
exaggerated response of the pituitary gland to the action of CRH. Clin Endocrinol
(0x0 59, 104-109.
G Wong, Y Goldshmit and AM Tumley, 2004. Interferon-gamma but not TNF alpha
promotes neuronal differentiation and neurite outgrowth of murine adult neural
stem cells. Exp Neurol 187, 171-177.
NJ Woolf, F Eckenstein and LL Butcher, 1984. Cholinergic systems in the rat brain: I.
projections to the limbic telencephalon. Brain Res Bull 13, 751-784.
CC Wrenn, DA Lappi and RG Wiley, 1999. Threshold relationship between lesion extent
of the cholinergic basal forebrain in the rat and working memory impairment in the
radial maze. Brain Res 847, 284-298.
K Yoshida and H Oka, 1995. Topographical projections from the medial septum-diagonal
band complex to the hippocampus: a retrograde tracing study with multiple
fluorescent dyes in rats. Neurosci Res 21, 199-209.


I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
John Petitto,
Profssor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
j y l t j..,'
Michael A. King, Cochair
Associate Scientist of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Mark H. Lewis
Professor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.' ft ,
.M-- *
Wolfgang J. Streit
Professor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.. v
K a \ v'
"
Mark A. Atkinson
Professor of Pathology, Immunology and
Laboratory Medicine


35
approximately 2-2.5% of the outlined count area of the largest single section of the areas
of interest.
The rostral border of the GCL count area was defined as the first section where
the dentate granule cell layer clearly separated from the pyramidal layer of CA1. The
caudal border was defined as the first section where the habenular commisure was visible
in the third ventricle. Only cells that could clearly be determined to be part of either the
IP or SP-GCL were counted; any cells in the area where the IP and SP-GCL connected
were left uncounted. Furthermore, only Nissl-stained cells with clearly visible nucleoli
were counted. For the MS/vDB, the rostral border was determined as the first section
where the corpus collosum connected in the midline of the section and the caudal border
was the first section where the anterior commisure joined in the midline.
The guard volume was set at 2 pm for the top and bottom of the section and the
counting cubes were randomly distributed with a total sampling frequency of the outlined
count area of 25% for MS/vDB and 33% for IP and SP-GCL (-8.3% and 11% of the total
area respectively, since every third section was sampled). The outlined counting area
was defined by the user and only somata that were clearly and distinctly stained were
counted. Each counting box was examined at 40x magnification and the computer-
assisted focus was used to scan from the top to the bottom of the counting box. Cells
were counted only if they were either completely inside of the counting box, or partially
inside of the box on the top, back, or left side. They were not counted if they fell outside
of the box or crossed into the box anywhere on the bottom, front, or right side. The cells
that were counted were labeled on the monitor by clicking the mouse pointer on each cell
and the MCID software recorded the number of marks.


85
KP Nanry, WR Mundy and HA Tilson, 1989. Colchicine-induced alterations of reference
memory in rats: role of spatial versus non-spatial task components. Behav Brain
Res 35, 45-53.
BH Nelson, 2004. IL-2, regulatory T cells, and tolerance. J Immunol 172, 3983-3988.
BH Nelson, JD Lord and PD Greenberg, 1994. Cytoplasmic domains of the interleukin-2
receptor beta and gamma chains mediate the signal for T-cell proliferation. Nature
369, 333-336.
R Nemni, S Iannaccone, A Quattrini, S Smime, M Sessa, M Lodi, C Erminio and N
Canal, 1992. Effect of chronic treatment with recombinant interleukin-2 on the
central nervous system of adult and old mice. Brain Res 591, 248-252.
W Pan, WA Banks and AJ Kastin, 1997. Permeability of the blood-brain and blood-
spinal cord barriers to interferons. J Neuroimmunol 76, 105-111.
K Pardy, D Murphy, D Carter and KM Hui, 1993. The influence of interleukin-2 on
vasopressin and oxytocin gene expression in the rodent hypothalamus. J
Neuroimmunol 42, 131-138.
DR Parkinson, JS Abrams, PH Wiemik, AA Rayner, KA Margolin, DA Van Echo, M
Sznol, JP Dutcher, FR Aronson, JH Doroshow and et al., 1990. Interleukin-2
therapy in patients with metastatic malignant melanoma: a phase II study. J Clin
Oncol 8, 1650-1656.
A Patapoutian and LF Reichardt, 2001. Trk receptors: mediators of neurotrophin action.
Curr Opin Neurobiol 11, 272-280.
S Pauli, AC Linthorst and JM Reul, 1998. Tumour necrosis factor-alpha and interleukin-2
differentially affect hippocampal serotonergic neurotransmission, behavioural
activity, body temperature and hypothalamic-pituitary-adrenocortical axis activity
in the rat. Eur J Neurosci 10, 868-878.
M Perez-Martin, I Azcoitia, JL Trejo, A Sierra and LM Garcia-Segura, 2003. An
antagonist of estrogen receptors blocks the induction of adult neurogenesis by
insulin-like growth factor-I in the dentate gyrus of adult female rat. Eur J Neurosci
18,923-930.
T Perry, H Hodges and JA Gray, 2001. Behavioural, histological and
immunocytochemical consequences following 192 IgG-saporin immunolesions of
the basal forebrain cholinergic system. Brain Res Bull 54, 29-48.
JM Petitto and Z Huang, 1994. Molecular cloning of a partial cDNA of the interleukin-2
receptor-beta in normal mouse brain: in situ localization in the hippocampus and
expression by neuroblastoma cells. Brain Res 650, 140-145.


48
outcomes described above. We compared the hippocampi of adult IL-2 KO mice and
wild-type littermates at 8-12 weeks of age, the age where differences in hippocampal
cytoarchitecture and behavior have been found previously (Beck et al., 2004; Beck et al.,
2002; Cushman et al., 2004; Petitto et al., 1999), for differences in several measures of
neuroimmunological status. First, the groups were assessed for differences in the number
of CD3+ T lymphocytes and activated microglial cells (as measured by MHC-II
positivity) in the hippocampus. IL-15 is also expressed in the brain (Hanisch et al.,
1997a; Lee et al., 1996), and this cytokine is known to have both proinflammatory and
anti-inflammatory effects, potent anti-apoptotic, and T cell chemoattractant properties
(Wilkinson and Liew, 1995). Because IL-15 uses the IL-2/15RP and yc subunits that are
enriched in the neuronal cell layers of the hippocampus (Petitto and Huang, 2001), and
may modulate microglial cell function and T cell chemoattraction, a second aim of this
study was to test the hypothesis that IL-15 is elevated in the hippocampus of IL-2 KO
mice. In addition, since changes in both IL-2 and IL-15 may modify levels of various
cytokines in other tissues and physiological contexts, exploratory testing was performed
to determine if IL-2 KO mice have increased levels of proinflammatory cytokines in the
hippocampus relative to wild-type mice (also, compared to serum levels). Furthermore,
as recent evidence indicates that elevation of inflammatory cytokines such as IL-6 may
impair hippocampal neurogenesis (Monje el al., 2003; Vallieres et al., 2002), a third aim
of this study was to test the hypothesis that the postulated changes in
neuroimmunological status would be associated with reductions in neurogenesis of
neurons in the dentate gyrus (DG) of IL-2 KO mice.


8
efficacy of 192 IgG-saporin in completely removing septohippocampal cholinergic
activity. In studies where 192 IgG-saporin appears to nearly completely eliminate
cholinergic immunostaining of MS/vDB neurons, ACh release in the hippocampus of
these lesioned animals still persists at -40% of the control values (Chang and Gold, 2004;
Gold, 2003). The septohippocampal cholinergic neurons are capable of compensatory
collateral sprouting (Gage et al., 1983a, 1984; Gage et al., 1983b), and functional
recovery of hippocampal ACh release after complete fimbria-fomix transection (Leanza
et al., 1993), which may be important as a response to damage. Thus, considering the
above evidence, it is clear that cholinergic input to the hippocampus plays an important
role in the complex neurobiological processes of learning and memory, though it is
certainly not the only system involved.
Some evidence supports a potential trophic or regulatory role for IL-2 on the
cholinergic septohippocampal system. IL-2 enhances the survival of these septal and
hippocampal neurons in culture (Awatsuji et al., 1993), and modulates the activity of
choline acetyl transferase (ChAT) (Mennicken and Quirion, 1997). Also, IL-2 is a potent
biphasic modulator of acetylcholine (ACh) release (Hanisch et al., 1993; Seto et al.,
1997). At low concentrations (sub-pM), IL-2 stimulates the release of ACh, but at higher
concentrations (nM), IL-2 inhibits ACh release. Since IL-2 is difficult to detect in the
adult brain, endogenous brain IL-2 or IL-2 that crosses the BBB would be expected to be
present at a low concentration in the normal brain.
Taking into consideration 1) the neurotrophic and neuromodulatory role of IL-2 in
cultured septal and hippocampal neurons, 2) the regulatory effects of IL-2 on ACh release
and ChAT activity in vitro, 3) the spatial learning and memory impairments of IL-2-


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
ABSTRACT vii
CHAPTER
1 BACKGROUND AND SIGNIFICANCE 1
Cytokine-Brain Interactions 1
The Pleiotropic Cytokine: Interleukin-2 2
Interleukin-2 and the Brain 4
IL-2 and the Septohippocampal System 6
Statement of the Problem 9
2 ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS
RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT 11
Introduction 11
Materials and Methods 13
Animals and Tissue Preparation 13
Genotyping Using PCR 14
ChAT Immunohistochemistry 14
AChE Histochemistry 16
Cholinergic Stereology 16
Quantitative Image Analysis of AChE Staining 18
Cresyl Violet Staining 20
Results 20
Comparison of Cholinergic Somata in the MS/vDB 20
Density of AChE-positive Fibers in Regions of the Hippocampus 20
Morphology of the Granular Cell Layer of the Lower Limb of the DG 21
Discussion 23
3 ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN
INTERLEUKIN-2 KNOCKOUT MICE 29
Introduction 29
Methods 32
v


32
Methods
Animals and Genotyping
Mice used in these experiments were cared for in accordance with the NIH Guide
for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2
heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the
NIH repository at Jackson Laboratories, were derived from ten generations of
backcrossing onto the C57BL/6 background. Mice were housed under specific pathogen-
free conditions. Animals used in these experiments were either 21-days-old (for ChAT
immunohistochemistry) or 8-12 weeks of age. All experiments were performed with
independent groups of animals. The specific animal numbers utilized are reported at the
beginning of each method descriptions below.
The genotypes of all mice were determined by the polymerase chain reaction
(PCR). PCR reactions were performed using 25 pi total reaction volume containing 1
pM each of forward and reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP,
0.3 pi Taq DNA polymerase, and amplified using a thermal cycler with a heated
evaporation cover (Ericomp). The cycling parameters were hot start 95C (3min),
denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final
extension step of 4 min. Thirty cycles were used for these experiments. The 5 and 3
primers for the IL-2 KO (500 bp knockout band amplified) were 5-
TCGAATCGCCAATGACAAGACGCT-3 and 5-
GTAGGTGGAAATTCTAGCATCATCC-3. The 5 and 3primers for the IL-2 wild
type (324 bp wild type band amplified) were 5-


53
labeling of neurogenesis and subsequent immunostaining in the mouse hippocampus has
been adapted from (Lee et al., 2002). Briefly, the mice were given five intraperitoneal
injections of BrdU (50 mg/kg of body weight) over the course of 3 days. The day
following the last injection, the mice were sacrificed and perfused with 0.9% saline
followed by 4% paraformaldehyde in PBS as described previously.
The BrdU-incorporated brains were equilibrated to -20 C and cryostat-sectioned
at 50 pm in the coronal plane. They were collected into individual wells of polystyrene
24-well plates (NUNC 1147), and used for free-floating immunohistochemistry. The
sections were then washed twice in PBS and then the DNA was denatured by a 30 minute
incubation with 2 N HC1 to allow binding of the antibody to the BrdU in the single-
stranded DNA. The acid was neutralized with a 0.1 M borate buffer (pH 8.5) wash,
followed by several washes in PBS. Afterwards, the sections were blocked for 1 hr in 3%
normal goat serum (NGS). The sections were then incubated overnight in the primary
antibodies, rat monoclonal anti-BrdU (Serotec; 1:400 in PBS with 0.3% Triton X-100 and
1% NGS) and either the neuronal marker mouse monoclonal anti-tubulin p III isoform
(Chemicon; 1:200 in PBS with 0.3% Triton X-100 and 1% NGS), the astroglial marker
rabbit anti-glial fibrillary acidic protein (GFAP; Chemicon; 1:1,000 in PBS with 0.3%
Triton X-100 and 1% NGS) or the oligodendrocyte marker mouse anti-23-cyclic
nucleotide 3-phosphohydrolase (CNPase; Chemicon; 1:200 in PBS with 0.3% Triton X-
100 and 1% NGS). The next day, the sections were washed twice in PBS and incubated
for 2 hr in the dark with the secondary antibodies, goat anti-rat IgG (H+L) conjugated
with Alexa Fluor-488 (green; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1%
NGS) and goat anti-mouse IgG (highly cross-absorbed H+L) conjugated with Alexa


49
Materials and Methods
Animals and Genotyping
Mice used in these experiments were cared for in accordance with the NIH Guide
for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2
heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the
NIH repository at Jackson Laboratories, were derived from ten generations of
backcrossing onto the C57BL/6 background. Mice were housed under specific pathogen-
free conditions. Animals used in these experiments were 8-12 weeks of age.
Independent animals were used for the assessment of CD3^ T cells and MHC II+
microglial cells in the hippocampus, the determinations of hippocampal versus serum
cytokine levels, and assessments of neurogenesis in the dentate gyrus. Specific numbers
utilized are reported at the beginning of the description of each method.
The genotypes of all mice were determined by the polymerase chain reaction
(PCR). PCR reactions were performed using 25 pi total reaction volume containing 1
pM each of forward and reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP,
0.3 pi Taq DNA polymerase, and amplified using a thermal cycler with a heated
evaporation cover (Ericomp). The cycling parameters were hot start 95C (3min),
denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final
extension step of 4 min. Thirty cycles were used for these experiments. The 5 and 3
primers for the IL-2 KO (500 bp knockout band amplified) were 5-
TCGAATCGCCAATGACAAGACGCT-3 and 5-
GTAGGTGGAAATTCTAGCATCATCC-3. The 5 and 3primers for the IL-2 wild
type (324 bp wild type band amplified) were 5-


Animals and Genotyping 32
Immunohistochemistry 33
Cresyl Violet Staining 34
Stereology 34
Enzyme-linked Immunosorbent Assay (ELISA) Characterization of NGF and
BDNF 36
Statistical Analysis 37
Results 38
Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell
Number in Adult Mice 38
Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice 38
Alterations in Neurotrophin Levels 38
Discussion 40
4 INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND
NEUROGENESIS IN THE HIPPOCAMPUS 45
Introduction 45
Materials and Methods 49
Animals and Genotyping 49
CD3+ T cells and MHC II+ Microglia Immunohistochemistry 50
Preparation of Serum and Brain Tissue for Cytokine Analysis 51
Multiplex Microsphere Cytokine Analysis 52
Labeling Neurogenesis with BrdU 52
Results 54
Assessment of CD3+ T Cells and MHC 11+ Activated Microglia in the
Hippocampus 54
Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice 54
Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-type Mice .55
Alterations in Neurogenesis 56
Discussion 58
5 GENERAL DISCUSSION 64
Summary of the Overall Findings 64
Implications 65
Caveats and Future Directions 67
Concluding Remarks 72
WORKS CITED 73
BIOGRAPHICAL SKETCH 92
vi


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EKXDISQA8_I2M3K0 INGEST_TIME 2014-12-01T20:46:17Z PACKAGE AA00026443_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


78
KD Denicoff, DR Rubinow, MZ Papa, C Simpson, CA Seipp, MT Lotze, AE Chang, D
Rosenstein and SA Rosenberg, 1987. The neuropsychiatric effects of treatment
with interleukin-2 and lymphokine-activated killer cells. Ann Intern Med 107, 293-
300.
JH Dufour, M Dziejman, MT Liu, JH Leung, TE Lane and AD Luster, 2002. IFN-
gamma-inducible protein 10 (IP-10; CXCL10)-deficient mice reveal a role for IP-
10 in effector T cell generation and trafficking. J Immunol 168, 3195-3204.
AJ Dunn, 2002. Mechanisms by which cytokines signal the brain. Int Rev Neurobiol 52,
43-65.
O Eizenberg, A Faber-Elman, M Lotan and M Schwartz, 1995. Interleukin-2 transcripts
in human and rodent brains: possible expression by astrocytes. J Neurochem 64,
1928-1936.
H Fukuoka, M Kawatani, T Hisamitsu and C Takeshige, 1994. Cutaneous hyperalgesia
induced by peripheral injection of interleukin-1 beta in the rat. Brain Res 657, 133-
MO.
G Fyfe, RI Fisher, SA Rosenberg, M Sznol, DR Parkinson and AC Louie, 1995. Results
of treatment of 255 patients with metastatic renal cell carcinoma who received
high-dose recombinant interleukin-2 therapy. J Clin Oncol 13, 688-696.
SL Gaffen, 2001. Signaling domains of the interleukin 2 receptor. Cytokine 14, 63-77.
FH Gage and A Bjorklund, 1987. Trophic and growth-regulating mechanisms in the
central nervous system monitored by intracerebral neural transplants. Ciba Found
Symp 126, 143-159.
FH Gage, A Bjorklund and U Stenevi, 1983a. Reinnervation of the partially deafferented
hippocampus by compensatory collateral sprouting from spared cholinergic and
noradrenergic afferents. Brain Res 268, 27-37.
FH Gage, A Bjorklund and U Stenevi, 1984. Cells of origin of the ventral cholinergic
septohippocampal pathway undergoing compensatory collateral sprouting
following fimbria-fornix transection. Neurosci Lett 44, 211-216.
FH Gage, A Bjorklund, U Stenevi and SB Dunnett, 1983b. Functional correlates of
compensatory collateral sprouting by aminergic and cholinergic afferents in the
hippocampal formation. Brain Res 268, 39-47.
BH Gahwiler, A Enz and F Hefti, 1987. Nerve growth factor promotes development of
the rat septo-hippocampal cholinergic projection in vitro. Neurosci Lett 75, 6-10.


83
YB Lee, J Satoh, DG Walker and SU Kim, 1996. Interleukin-15 gene expression in
human astrocytes and microglia in culture. Neuroreport 7, 1062-1066.
WJ Leonard, JM Depper, GR Crabtree, S Rudikoff, J Pumphrey, RJ Robb, M Kronke, PB
Svetlik, NJ Peffer, TA Waldmann and WC Greene, 1984. Molecular cloning and
expression of cDNAs for the human interleukin-2 receptor. Nature 311, 626-631.
J Licinio, JP Seibyl, M Altemus, DS Chamey and JH Krystal, 1993. Elevated CSF levels
of interleukin-2 in neuroleptic-free schizophrenic patients. Am J Psychiatry 150,
1408-1410.
T Lindenlaub, P Teuteberg, T Hartung and C Sommer, 2000. Effects of neutralizing
antibodies to TNF-alpha on pain-related behavior and nerve regeneration in mice
with chronic constriction injury. Brain Res 866, 15-22.
BR Ludviksson, RO Ehrhardt and W Strober, 1997. TGF-beta production regulates the
development of the 2,4,6-trinitrophenol-conjugated keyhole limpet hemocyanin-
induced colonic inflammation in IL-2-defcient mice. J Immunol 159, 3622-3628.
A Ma, M Datta, E Margosian, J Chen and I Horak, 1995. T cells, but not B cells, are
required for bowel inflammation in interleukin 2-deficient mice. J Exp Med 182,
1567-1572.
SF Maier, LE Goehler, M Fleshner and LR Watkins, 1998. The role of the vagus nerve in
cytokine-to-brain communication. Ann N Y Acad Sci 840, 289-300.
R Makuch, J Baratta, LD Karaelias, JC Lauterbom, CM Gall, J Yu and RT Robertson,
2001. Arrival of afferents and the differentiation of target neurons: studies of
developing cholinergic projections to the dentate gyrus. Neuroscience 104, 81-91.
D Maslinska, 2001. The cytokine network and interleukin-15 (IL-15) in brain
development. Folia Neuropathol 39, 43-47.
CG McAllister, DP van Kammen, TJ Rehn, AL Miller, J Gurklis, ME Kelley, J Yao and
JL Peters, 1995. Increases in CSF levels of interleukin-2 in schizophrenia: effects
of recurrence of psychosis and medication status. Am J Psychiatry 152, 1291-1297.
JF McDyer, Z Li, S John, X Yu, CY Wu and JA Ragheb, 2002. IL-2 receptor blockade
inhibits late, but not early, IFN-gamma and CD40 ligand expression in human T
cells: disruption ofboth IL-12-dependent and -independent pathways of IFN-
gamma production. J Immunol 169, 2736-2746.
RL McLamb, WR Mundy and HA Tilson, 1988. Intradentate colchicine disrupts the
acquisition and performance of a working memory task in the radial arm maze.
Neurotoxicology 9, 521-528.


70
quantify MS/vDB neurons in this brain region. Animals with neuronal loss, should
exhibit reduction in total neuronal counts, whereas the neuronal counts would not differ
significantly in mice with a loss of phenotype.
Also, considering the difficulty of detection of endogenous IL-2 in the normal
CNS, what is the source of the IL-2 in the wild-type brain? Taken together with the
unknown non-saturable transport mechanism that allows IL-2 to cross the BBB
(Waguespack et al., 1994), and the numerous studies showing cognitive effects of
peripherally administered exogenous IL-2 (Capuron et al., 2001a; Denicoff et al., 1987;
Lacosta et al., 1999; Walker et al., 1997), an argument can be made that some of the
endogenous IL-2 in the CNS could be from the periphery, particularly during
development when the BBB is not completely formed. In addition, though we did
address the autoimmunity issue of IL-2 KO mice somewhat, we did not completely rule it
out as a potential factor in the observed alterations. An experiment that could address
both of the above issues would involve crossbreeding the IL-2 KO mice onto an
immunodeficient background lacking functional lymphocytes (e.g., RAG-1 knockout
mice). Next, an adoptive transfer of normal T lymphocytes from healthy IL-2 wild-type
(i.e., non-RAG-1 KO) mice to young (i.e., less than 3-week-old) IL-2 KO/RAG-1 KO or
IL-2 wild-type/RAG-1 KO littermate animals would establish a functional immune
system, and at the same time restore a major source of peripheral IL-2. Also, another set
of IL-2 KO/RAG-1 KO and IL-2 wild-type/RAG-1 KO mice should receive an adoptive
transfer of IL-2 KO (i.e., non-RAG-1 KO) lymphocytes to mimic the autoimmune state
of normal IL-2 KO mice. Finally, another group of IL-2 KO/RAG-1 KO and IL-2 wild-
type/RAG-1 KO mice should receive a sham reconstitution, such that they maintain any


31
Mizuma et al., 1999; Moalem et al., 2000). Though IL-2 regulates several aspects of T
cell function, the production or release of NGF and BDNF from T lymphocytes by IL-2
has not been tested directly(Carter et al., 1998; He and Malek, 1998), and conversely, it is
not known whether NGF and BDNF can modulate IL-2 in the brain.
In the present study, we sought to expand our previous findings that loss of
endogenous IL-2 in knockout mice led to reductions in cholinergic neurons of the
MS/vDB and a decrease in the distance across the IP-GCL of the DG. First, we
compared parvalbumin (Parv)-labeled somata in 8-12 week old IL-2 KO and wild-type
littermates to test the hypothesis that the loss of IL-2 is selective for cholinergic, but not
GABAergic cell bodies loss in the MS/vDB (e.g., not a general effect occurring on all
neurons in this region of the brain). Second, we compared cholinergic MS/vDB somata
between younger wild-type and IL-2 KO mice at postnatal day 21 (P21), an age where
septohippocampal development in mice is nearly complete (Armstrong et al., 1987;
Gould et al., 1991; Makuch et al., 2001). This age also precedes the development of
autoimmune disease in IL-2 KO mice of the C57BL/6 background, e.g., absence of
splenomegaly, lymphadenopathy, and inflammatory bowel disease. Third, we sought to
expand on the previous finding that there was a reduction in distance across IP-GCL by
performing stereological cell counts of Nissl-stained dentate gyri of IL-2 KO and wild-
type littermates to determine if the reduction in distance could be contributed to a
reduction in granule cell number. Finally, we tested the hypothesis that loss of IL-2 may
impact the expression and release of the neurotrophins, NGF and BDNF, which may
contribute to the MS/vDB cholinergic and GCL deficits that we observed previously.


60
were increased in the hippocampus of IL-2 KO mice. IL-15 is structurally related to IL-2
and shares the same p and yc signal transducing receptor subunits with the IL-2 receptor
(Giri et al., 1995). IL-15 also shares and opposes several physiological functions of IL-2
in the peripheral immune system (Waldmann, 2002; Waldmann et al., 2001). IL-15 and
its heterotrimeric receptor are constitutively expressed in various regions of the adult
mouse brain and can be detected in microglial cultures (Hanisch et al., 1997a), astrocytes
(Lee et al., 1996), and possibly neurons (Maslinska, 2001; Satoh et al., 1998). As noted
earlier, increased IL-15 gene expression precedes the inflammatory cytokine changes and
onset of inflammatory bowel disease in IL-2 KO mice (Meijssen et al., 1998). It also
induces the onset of autoimmunity in thyroiditis (Kaiser et al., 2002). Thus, IL-15 could
trigger proinflammatory cytokine-like processes in the hippocampus, including the
elevations in IL-12 that were found in IL-2 KO mice in this study. IL-12-driven Thl
responses are involved in inflammation (e.g., colonic) in IL-2 KO mice (Ludviksson et
al., 1997), and it has been implicated as an important effector in the pathogenesis of
experimental autoimmune encephalomyelitis (EAE) (Adorini, 1999; Segal et al., 1998).
Moreover, IL-15 can render cells resistant to the protective effects of TGFP (Campbell et
al., 2001), a Th2 cytokine that appears to play a key role in dampening processes
associated with peripheral autoimmune disease in IL-2 KO mice (Ludviksson et al.,
1997). Thus, these actions of IL-15 suggest that it may be involved in the hippocampal
pathology seen in IL-2 KO mice. It is noteworthy, however, that IL-15 has potent anti-
apoptotic properties (Lauwerys et al., 2000; Waldmann, 2002; Waldmann et al., 2001)
that may oppose the pro-apoptotic effects of IL-2. We have recently found that loss of
brain IL-2 in C57BL/6sc/d-IL-2 KOmice increased neuroregeneration in the axotomized


9
deficient knockout mice and cancer patients undergoing IL-2 therapy, and 4) the
enhanced concentration of IL-2R subunits in the hippocampus, I hypothesized that IL-2
may play a role in the growth and differentiation of the septohippocampal cholinergic
neurons.
Statement of the Problem
Normal development of septal neurons depends on trophic factors that are
presumably secreted from the hippocampus. For example, the neurotrophins, nerve
growth factor (NGF) and brain-derived neurotrophic factor (BDNF) are expressed in the
hippocampus and have both been shown to be important in the development,
maintenance, and repair of septohippocampal neurons (Brooks et al., 1999; Conner et al.,
1992; Conner and Varn, 1997; Morse et al., 1993). Perhaps, IL-2 also mediates the
growth and development of septal and hippocampal neurons. This cytokine may act as a
growth factor by itself, or signal the release of growth factors from neurons or glia in the
hippocampal area.
In other investigations, the modulatory and growth-promoting effects of IL-2
were determined by either adding exogenous IL-2 to cultures or by injecting exogenous
IL-2 into the brain (in most cases, species non-specific, e.g., human IL-2 in rats or mice).
Thus, the studies utilizing exogenous IL-2 administration have several potential
shortcomings: 1) the cytokine has a short half-life, and the amount of IL-2 delivered may
not reflect physiologically relevant concentrations in the CNS in vivo (e.g., chronic vs.
acute dosing), 2) the nature of the injections disrupts the BBB, causing potentially even
more IL-2, as well as other cytokines, from the peripheral immune system to enter into
the brain, 3) it is unknown during which period in neurodevelopment that IL-2 might
exert these postulated effects, and 4) non-species specific IL-2 may have neurotoxic


46
laboratory, however, suggests that these hippocampal changes may be due to
neurodegenerative rather than neurodevelopmental processes. We tested mice at
postnatal day 21 (P21), an age where septohippocampal cholinergic neurons are nearly
fully developed, to determine if the reduction in septohippocampal cholinergic projection
neurons seen in adult IL-2 KO mice was present earlier in postnatal development (e.g., at
weaning) and prior to the onset of the earliest signs of autoimmune disease (e.g.,
splenomegaly, lymphadenopathy). Contrary to our hypothesis, we found that the number
of MS/vDB cholinergic cell bodies did not differ between IL-2 KO and wild-type
littermates at P21 (Beck et al., 2004). Thus, together these data indicate that the loss of
cholinergic neurons that occurs between P21 and adulthood (8-12 weeks) suggests an
alternate hypothesis; neurodegenerative processes may be operative in the brain of IL-2
KO mice.
Since IL-2 is an important factor in immune physiology, one possible mechanism
behind these neurodegenerative processes is immune dysregulation caused by the absence
of IL-2. IL-2-defciency in mice leads to generalized systemic autoimmune disease in
adult mice that may affect multiple organs in the periphery, most notably the intestines
and the kidneys (Horak, 1995). The autoimmune effects in IL-2 KO mice involving
peripheral organs are mediated largely by infiltrating T cells. In the colon, for example,
adult IL-2 KO mice develop chronic inflammatory bowel disease with features common
to inflammatory ulcerative colitis in humans, where the lamina propria is infiltrated with
activated T cells responsible for the development of this inflammatory disease (Ma et al.,
1995). In addition, there is a disruption of immune homeostasis that is evidenced by
changes in the gene expression of several Thl, Th2, and various proinflammatory


15
for 2 hrs followed by two washes in PBS. The sections were developed in 0.5 mg/ml
3,3-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min and were
placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes,
and coverslipped. Figure 2-1 shows an example of ChAT immunostained section of the
MS/vDB.
Figure 2-1. ChAT immunohistochemical staining of the medial septum/vertical limb of
the diagonal band of Broca. The scale bar represents 200 pm.


58
Discussion
The data presented here show that IL-2 KO and wild-type littermates exhibit
differences in several measures of neuroimmunological status in the hippocampus. In
order to access the brain parenchyma, T cells require activation markers to cross the BBB
(Hickey et al., 1991). We have previously reported leukocytes were not detectable in the
cresyl violet stained hippocampal sections from IL-2 KO mice (Petitto et al., 1999),
however, we recognized that it is difficult to reliably detect small numbers of peripheral
leukocytes in the brain without cell-type specific stains. Although we were therefore not
expecting to see substantial numbers of T cells in the IL-2 KO brain, we wanted to
determine if small numbers of autoimmune T lymphocytes were present that could
initiate neuroimmunological alterations in the hippocampus. The hippocampi of IL-2 KO
mice were devoid of T cells, despite the fact that the majority of peripheral T cells of IL-2
KO mice express activation markers such as CD69 (Sakai et al., 1995; Schopke et al.,
1991), which are thought to enhance their ability to cross the BBB. Microglia are
indigenous antigen presenting cells (Hickey and Kimura, 1988; Streit et al., 1988).
Contact with T cells can induce microglia to exhibit characteristics of antigen presenting
cells, and microglia also have the ability to activate T cells (Aloisi et al., 2000). There
was, however, no evidence of increased numbers of activated microglia in the
hippocampus of IL-2 KO mice. This observation is consistent with our previous finding
in C57BL/65C/W-IL-2 KO (mice without mature T and B cells), which were devoid of T
cells in the axotomized facial motor nucleus and had levels of axotomy-induced activated
microglia that did not differ from wild-type mice (Petitto et al., 2003). Thus, at the
cellular level, the hippocampus of IL-2 KO mice did not show signs of autoimmune
disease.


54
Fluor 568 (red; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1% NGS) or goat
anti-rabbit IgG (H+L) conjugated with Alexa Fluor 350 (blue; Molecular Probes; 1:400
in PBS with 0.3% TX-100 and 1% NGS). The sections were then washed twice in PBS,
placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes,
and coverslipped.
The sections were imaged using a Bio-Rad 1024 ES confocal microscope and
only cells which showed colocalized staining through five consecutive 1-pm planes were
considered to be double-labeled. The IP and SP-GCL area (mm2) were measured at 20x
magnification using the MCID 5.1 software, a CCD High Resolution Sony camera, and a
Zeiss Axioplan 2 microscope. The data were reported as a density of all double-labeled
cells counted from five sections per animal divided by the total area measured
Results
Assessment of CD3+ T Cells and MHC II+ Activated Microglia in the Hippocampus
No CD3+ T cells were detected in the hippocampi of either 8-12-week-old IL-2
KO or wild-type mouse brains, and only an occasional MHC II+ microglial cell was
detected (e.g., approximately one every other section) in both groups. By contrast, both
activated MHC IF microglia and CD3" T cells were readily detectable in positive control
slices (sections of the axotomized FMN of wild-type C57BL/6 mice; Petitto et al., 2003)
demonstrating that the immunohistochemistry procedure utilized was effective for
labeling both markers. Thus, no differences in T cells or activated microglial cells were
found between the subject groups.
Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice
As depicted in Figure 4-1, in the hippocampus, there were significantly increased
levels of IL-12 (increased -57%; F(l,15)=9.174, p=0.008) and IL-15 (increased -38%;


ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE IN
INTERLEUKIN-2 KNOCKOUT MICE
By
RAY D. BECK, JR.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2004

This dissertation is dedicated to my wife Laura whose love, support, and occasional
nagging help to keep me focused on my goals.

ACKNOWLEDGMENTS
Traditionally, most acknowledgement sections begin by thanking ones advisor; in
this case, such tradition is most warranted. As such, I thank Dr. John Petitto for being a
wonderful mentor. Though not the stealthiest individual in the world with his penchant
for crying out ones name (or impromptu nickname) exuberantly as soon as one enters
anywhere within his field of vision, he is among the kindest, most supportive,
enthusiastic, and knowledgeable mentors for which any graduate student could wish.
Next, I would like to thank each member of my committee. I feel fortunate to be advised
by such a great selection of knowledgeable and friendly people. Dr. Mike Kings
easygoing personality and extensive knowledge of all things stereological and cholinergic
have proven invaluable in my studies. Dr. Mark Lewis advice on statistics and
experimental design, as well as his sense of humor, has been most appreciated. Dr. Jake
Streit, in addition to his ability to make me laugh, always reminded me that there is more
than one kind of cell in the brain. Dr. Mark Atkinson, always amiable and approachable,
helped guide me in the immunology aspect of neuroimmunology.
I also thank Dr. Huang Zhi. I cannot overstate how much I valued his advice on
experiments and his daily conversations on topics ranging from basketball to politics to
Hong Kong movies. I wish him the best of luck in his medical residency and his future
as a psychiatrist. I would also like to thank Clive Wasserfall and Fletcher Schwartz for
teaching me how to use the Luminex technology and Tim Vaught for teaching me the ins
and outs of multiple microscopy techniques. I also thank the many technicians that
iii

worked in the lab both past and present: Brent, Andrew, David, Jeannette, Dan, Jesse, and
Grace. In addition to taking care of the upkeep of the lab, working with them was a
pleasure. In particular, I would like to single out Andrew. He was my dearest friend here
in Gainesville. I wish him the best of luck in his career as a medical doctor and hope that
we will always remain friends after I leave Gainesville.
Outside of the laboratory, I thank my other friends for being my support structure.
There has never been a better collection of in-the-closet geeks than Coleman, Dan, Andy,
Ryan, Charles, Chris, Curtis, Nick, and Jason. They are simply the best. I would also
like to thank Mozart, Michelangelo, and Sage simply because not nearly enough people
thank their dogs. Obviously, they would be more likely to chew on this dissertation than
read it, but I would like anyone else that does see this to know that few people are
capable of matching the unconditional love that a dog has for its owner.
I also thank my family. My mother has always instilled the value of education into
me. My fathers love and sense of humor were crucial in the development of my
personality. My sisters, Jackie and Judy, will always be among my closest friends. My
Aunt Kathy and Uncle Jimmy are tremendous people who have always supported me and
though they are family by marriage, our bond is stronger than blood.
Finally, I thank my lovely wife Laura. Without her, my life would be incomplete
(though despite her beliefs, I could still drive effectively without her commentary from
the passenger seat). She has pushed me when I needed motivation and comforted me
when I need support. She is my heart and soul and I could not have achieved this
dissertation without her.
IV

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
ABSTRACT vii
CHAPTER
1 BACKGROUND AND SIGNIFICANCE 1
Cytokine-Brain Interactions 1
The Pleiotropic Cytokine: Interleukin-2 2
Interleukin-2 and the Brain 4
IL-2 and the Septohippocampal System 6
Statement of the Problem 9
2 ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS
RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT 11
Introduction 11
Materials and Methods 13
Animals and Tissue Preparation 13
Genotyping Using PCR 14
ChAT Immunohistochemistry 14
AChE Histochemistry 16
Cholinergic Stereology 16
Quantitative Image Analysis of AChE Staining 18
Cresyl Violet Staining 20
Results 20
Comparison of Cholinergic Somata in the MS/vDB 20
Density of AChE-positive Fibers in Regions of the Hippocampus 20
Morphology of the Granular Cell Layer of the Lower Limb of the DG 21
Discussion 23
3 ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN
INTERLEUKIN-2 KNOCKOUT MICE 29
Introduction 29
Methods 32
v

Animals and Genotyping 32
Immunohistochemistry 33
Cresyl Violet Staining 34
Stereology 34
Enzyme-linked Immunosorbent Assay (ELISA) Characterization of NGF and
BDNF 36
Statistical Analysis 37
Results 38
Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell
Number in Adult Mice 38
Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice 38
Alterations in Neurotrophin Levels 38
Discussion 40
4 INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND
NEUROGENESIS IN THE HIPPOCAMPUS 45
Introduction 45
Materials and Methods 49
Animals and Genotyping 49
CD3+ T cells and MHC II+ Microglia Immunohistochemistry 50
Preparation of Serum and Brain Tissue for Cytokine Analysis 51
Multiplex Microsphere Cytokine Analysis 52
Labeling Neurogenesis with BrdU 52
Results 54
Assessment of CD3+ T Cells and MHC 11+ Activated Microglia in the
Hippocampus 54
Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice 54
Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-type Mice .55
Alterations in Neurogenesis 56
Discussion 58
5 GENERAL DISCUSSION 64
Summary of the Overall Findings 64
Implications 65
Caveats and Future Directions 67
Concluding Remarks 72
WORKS CITED 73
BIOGRAPHICAL SKETCH 92
vi

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE
IN INTERLEUKIN-2 KNOCKOUT MICE
By
Ray D. Beck, Jr.
August 2004
Chair: John M. Petitto
Major Department: Neuroscience
Interleukin-2 (IL-2) is a multifunctional cytokine involved in peripheral immune
processes and may also be implicated in multiple brain functions. IL-2 gene knockout
(IL-2 KO) mice exhibit deficits in several hippocampally-mediated behaviors (e.g.,
learning and memory) and have alterations in hippocampal structure.
In the first study, adult IL-2 KO and wild-type littermates were compared for
differences in the cholinergic neurons in the medial septum and vertical limb of the
diagonal band of Broca (MS/vDB; a structure associated with learning and memory).
The IL-2 KO mice had significantly fewer cholinergic somata in the MS/vDB, but not in
the striatum, thus indicating a selective effect of IL-2 on the MS/vDB. Cholinergic
neurite density in the hippocampus was unaffected, but the length across the
infrapyramidal (IP), but not the suprapyramidal (SP), granule cell layer (GCL) of the
dentate gyrus was reduced.
Vll

The second study assayed for variations between groups in the second largest
population of neurons in the MS/vDB, the GABAergic neurons. We found no
differences in these neurons in IL-2 KO animals. In 21 -day-old IL-2 KO mice, we
detected no changes in cholinergic neuronal number in the MS/vDB. This inconsistency
with adult cholinergic neurons may be due to a failure in maintenance or might be
secondary to autoimmunity. Neuronal number in the IP-GCL was also decreased,
consistent with the reduction in distance detected in the first study. We also discovered
that IL-2 KO correlates with a hippocampal elevation in nerve growth factor (NGF), but a
reduction in the brain-derived neurotrophic factor (BDNF).
Finally, in the last study, no T cells or evidence of increased activated microglia
was evident in the IL-2 KO mouse hippocampus. We noted significant elevations in
several cytokines (IL-12, IL-15, IP-10, MCP-1) in the hippocampus of IL-2 KO mice.
The cytokine profile of the serum was different from the hippocampus, indicating that
these were not global changes throughout the bodies of the animals. We also found an
alteration in hippocampal neurogenesis that appeared to be attributable to differences in
male mice.
The results of these studies suggest a neuroimmune interaction that may be
important in septohippocampal physiology.
viii

CHAPTER 1
BACKGROUND AND SIGNIFICANCE
Cytokine-Brain Interactions
The landmark studies of Ader and Cohen demonstrating that immune physiology
could be behaviorally conditioned led to the systematic investigation of the complex
interaction between the central nervous and immune systems (Ader and Cohen, 1975;
Ader et al., 1982). Although the central nervous system (CNS) and peripheral immune
system were once considered functionally incompatible entities separated by a nearly
impermeable protective blood-brain-barrier (BBB), it is now known that there is
bidirectional communication and modulation between these two systems. Cytokines
have emerged as important mediators of various processes in the CNS. Their effects
range from neuroinflammation in experimental autoimmune encephalomyelitis (EAE)
and viral infection of the brain to neurobiological processes such as hypothalamic-
pituitary axis (HPA) regulation, induction of fever, sleep, analgesia, feeding behavior,
and cognition (for reviews see Ader et ah, 2001; Dunn, 2002; Wilson et ah, 2002).
Cytokines produced both within and outside of the CNS can exert their effect on
brain cells (Dunn, 2002; Streit et ah, 1998). The work of Banks and others show that the
BBB acts as a selective filter for peripheral cytokines (for a review see Banks et ah,
2002). Multiple studies support the ability of cytokines (e.g., IL-la and -(3, IL-2, IL-6,
IFN-a and -y, TNF-a) to cross the BBB via different transport mechanisms (Banks et ah,
1994; Banks et ah, 1991; Gutierrez et ah, 1993; Pan et ah, 1997; Waguespack et ah,
1994), and via the leaky circumventricular organs (CVO), four brain regions outside of
1

2
the BBB with fenestrated capillaries (Buller, 2001). Peripheral leukocytes, in particular
activated T cells that enter the brain during certain conditions (e.g., EAE, facial nerve
axotomy), can also release cytokines in the CNS (Hickey et al., 1991). Finally, cytokines
may also interact with the brain through activation of peripheral nerves, such as IL-1
stimulation of the vagus nerve, which can lead to modulation of brain functions through
its afferent connections in the CNS (Maier et ah, 1998). Such a cytokine-to-nerve
communication pathway may not be limited to the vagus nerve, as central hyperalgesic
effects are also observed by stimulating cutaneous nerves with a subcutaneous injection
of IL-1 P (Fukuoka et ah, 1994), TNF-a (Sorkin et ah, 1997), or antibodies against TNF-
a (Lindenlaub et ah, 2000). Thus, multiple pathways exist that allow cytokines to
directly or indirectly influence the brain. The focus of this dissertation was on IL-2,
which can be produced in the periphery and the CNS.
The Pleiotropic Cytokine: Interleukin-2
IL-2 was originally identified as a growth factor for bone marrow-derived T cells
in 1976 (Morgan et ah, 1976), and was renamed in 1979, when its pleiotropic effects
between leukocytes (thus the term interleukin) became clear (Aarden et ah, 1979).
Further characterization of IL-2 revealed that it belongs to the four a-helix bundle family
of cytokines; this family consists of cytokines with four a-helices connected by three
loops in an up-up-down-down formation (Bazan, 1992). The receptor for IL-2 has a
common gamma (yc) subunit shared by multiple cytokines including IL-4, IL-7, IL-9. and
IL-15 (Sugamura et ah, 1996); a p subunit only shared with IL-15 (Giri et ah, 1995); and,
in one conformation, an a subunit, which confers greater binding affinity (Leonard et ah,
1984). The receptor subunits can combine in two biologically active forms: a lower

3
affinity heterodimer consisting of the yc and P subunits and a high affinity heterotrimer
comprised of all three subunits (a, P, and yc) (Ringheim et al., 1991; Takeshita et al.,
1992). The P and yc both possess intracellular signaling domains and in their
heterodimeric form have a IQ of 109, whereas the addition of the a subunit forms a
heterotrimer with higher affinity (IQ of 10n) for IL-2 (Nakamura et al., 1994; Nelson et
al., 1994).
In the peripheral immune system, where the physiological properties of IL-2 are
most well-characterized, IL-2 has multiple biological functions, including natural killer
(NK) cell activation, T lymphocyte activation, as well as B lymphocyte differentiation
(for review see Waldmann, 2002). The creation of a transgenic knockout mouse model
for this cytokine suggests that the most important function of IL-2 is the maintenance of
immune self-tolerance (Schmitt et al., 1994). IL-2 knockout (KO) mice develop
autoimmune symptoms commonly including inflammatory bowel disease similar to
ulcerative colitis in humans and advanced hemolytic anemia, although the manifestation
of the phenotype is dependent on the genetic background of the knockout mice (Horak,
1995, 1996). The autoimmunity that develops when the IL-2 gene is deleted is T cell
dependent (Ma et al., 1995). More recently, it has become apparent that IL-2 plays a
major role in limiting T cell responses via the development of regulatory T cells
(CD4+CD25* T reg cells) and other mechanisms that promote self tolerance and suppress
T cell responses in vivo (Nelson, 2004). Though extensive research has characterized the
effects of IL-2 in the peripheral immune system, increasing evidence indicates that IL-2
may potentially impact the central nervous system (CNS). The focus of this research
project was to characterize these potential actions of IL-2 in the brain.

4
Interleukin-2 and the Brain
The effect of IL-2 on cognition and mood in humans was among the earliest
findings that suggested that this cytokine might have neurobiological actions. In early
clinical studies of the cognitive side effects of IL-2 therapy, 50% (i.e., 22 patients out of
44) of the subjects monitored developed cognitive changes, with 15 of them necessitating
acute intervention (Denicoff et al., 1987). In addition, IL-2 therapy in patients with renal
carcinoma or melanoma was found to impair spatial memory and performance in
planning tasks (Capuron et ah, 2001a), and induce depressive symptoms as early as two
days into therapy (Capuron et ah, 2000).
IL-2-brain interactions have also been investigated on an anatomical and
physiological level. In landmark studies, IL-2 was found to modulate the proliferation of
oligodendrocytes (Benveniste et ah, 1987; Benveniste and Merrill, 1986; Saneto et ah,
1986). Exogenously administered IL-2 also has multiple effects on pituitary cells
including stimulation of cortisol production and adrenal corticotropin releasing hormone
release (Hanisch et ah, 1994), as well as increasing pituitary cell responsiveness to
corticotropin-releasing hormone (Witzke et ah, 2003). IL-2 has also been shown to
regulate the production and secretion of peptides from hypothalamus, in addition to
pituitary cells (Karanth et ah, 1993; Lapchak and Araujo, 1993; Pardy et ah, 1993).
Subsequent research has shown that exogenously applied IL-2 can modulate other types
of central nervous system cells, such as microglia (Sakai et ah, 1995). Exogenously
applied IL-2 can also biphasically regulate the release of some neurotransmitters such as
dopamine (Alonso et ah, 1993; Petitto et ah, 1997), or acetylcholine (Hanisch et ah,
1993; Seto et ah, 1997).

5
IL-2 has been shown to have neurotrophic effects on cultured neurons from
several regions of the rat brain including the neocortex (Shimojo et al., 1993), cortex,
striatum, medial septum, and hippocampus (Awatsuji et ah, 1993). Moreover, in rat
hippocampal neuronal cultures, IL-2 enhances the length and branching of hippocampal
neurites and the morphology of these neurons (Sarder et ah, 1996; Sarder et ah, 1993).
Interestingly, altered levels of IL-2 expression have been detected in schizophrenia (for
reviews see Hanisch and Quirion, 1995a; Muller and Ackenheil, 1998), which is a
neurological disorder where altered morphology of hippocampal neurons is well
documented (for a review see Thune and Pakkenberg, 2000).
IL-2-like immunoreactivity has been localized to the hippocampal formation in rat
forebrain (Lapchak et ah, 1991), and detected in tissue extracts from rat and human
hippocampal tissue (Araujo et ah, 1989). In mouse brain, IL-2 mRNA has been found in
the hippocampus (Villemain et ah, 1991), and transcripts for this cytokine may be
expressed in rat astrocyte cultures as well (Eizenberg et ah, 1995). Our lab has cloned
and sequenced the full-length mouse brain cDNAs for IL-2Ra as well as the IL-
2/15RP and yc subunits, and has found that the sequences of the genes expressed by
lymphocytes and in brain are identical. We have also found that these genes are enriched
in the hippocampus and related limbic regions. Of particular relevance to IL-2 actions in
the hippocampus, in situ hybridization has shown that the IL-2/15RP and yc genes are
expressed by pyramidal and granule cell neurons (Petitto and Huang, 1994, 1995, 2001;
Petitto et ah, 1998).

6
IL-2 and the Septohippocampal System
Hippocampal circuitry is important for encoding spatial learning and memory and
some evidence supports a potential role of IL-2 in the hippocampus. IL-2, for example,
alters the electrophysiological characteristics of hippocampal neurons including
alterations of voltage-dependent Ca:+ currents (Plata-Salaman and ffrench-Mullen, 1993),
depolarization and hyperpolarization of cultured hippocampal neurons (Hanisch and
Quirion, 1995a), and changes in long-term potentiation (LTP) (Tancredi et ah, 1990). IL-
2R subunits, as previously mentioned, are enriched in the hippocampus relative to other
brain regions and exogenously applied IL-2 enhances the survival and morphological
development of neurons of the hippocampus.
Knockout mice deficient in IL-2 perform significantly worse than wild-type
controls in one such test of spatial learning and memory, the Morris water maze; show an
enhanced pre-pulse inhibition of the acoustic startle response (PPI; another
hippocampally-mediated process); and also exhibit structural alterations in mossy fiber
length (Petitto et ah, 1999). Our initial studies suggest that this deficit in learning and
memory is not likely due to a compromised immune system, as severe combined
immunodeficient (SCID) mice perform significantly better than IL-2 KO mice in the
Morris water maze. More recent studies from our lab suggest that the nature of the
deficit in learning and memory seen in IL-2 knockout mice could be related to the
immune status of the mother (normal heterozygote vs. autoimmune homozygote mother).
In addition, the previously mentioned clinical studies of cancer patients under IL-2
treatment found alterations in spatial memory, lending some support to the potential role
of IL-2 in learning and memory.

7
In the basal forebrain, the medial septum and vertical limb of the diagonal band of
Broca (MS/vDB) send a large number of projections to the hippocampus, with the major
neuronal phenotypes of these being cholinergic and GABAergic (Brashear et al., 1986;
Kiss et al., 1990b; Kiss et al., 1990a). The septohippocampal system has been associated
with learning and memory processes, with extensive data existing that link the septal
cholinergic neurons that project to the hippocampus to learning and memory (Galey et
al., 1994; Leanza et al., 1995), and PPI (Koch, 1996b). Moreover, variability of
cholinergic fiber density in the dentate gyrus of individual mouse strains correlate with
changes in spatial learning (Schwegler et al., 1996a; Schwegler et al., 1996b). Some
controversy exists, however, on the relative importance of cholinergic neurons of the
MS/vDB in learning and memory processes. The advent of selective toxins that target
cholinergic neurons, like 192 IgG-saporin, have allowed researchers to behaviorally test
animals only lacking MS/vDB cholinergic neurons, but with presumably normal
distributions of GABAergic neurons. In many of these studies, animals with cholinergic
septohippocampal lesions did not differ from control subjects (Baxter et al., 1996; Bizon
et al., 2003; Cahill and Baxter, 2001; Chappell et al., 1998; Perry et al., 2001).
Surprisingly, however, multiple other contemporary studies utilizing 192 IgG-saporin do
find learning and memory deficits in the lesioned animals (Janis et al., 1998; Johnson et
al., 2002; Lamprea et al., 2000; Wrenn et a!., 1999). One potential explanation for this
discrepancy may be that a certain threshold of cholinergic damage is necessary to elicit a
deterioration in spatial learning ability (Leanza et al., 1995; Wrenn et al., 1999).
Nevertheless, the negative findings are considerable and this may not explain the
inconsistency well enough between groups. Another hypothesis calls into question the

8
efficacy of 192 IgG-saporin in completely removing septohippocampal cholinergic
activity. In studies where 192 IgG-saporin appears to nearly completely eliminate
cholinergic immunostaining of MS/vDB neurons, ACh release in the hippocampus of
these lesioned animals still persists at -40% of the control values (Chang and Gold, 2004;
Gold, 2003). The septohippocampal cholinergic neurons are capable of compensatory
collateral sprouting (Gage et al., 1983a, 1984; Gage et al., 1983b), and functional
recovery of hippocampal ACh release after complete fimbria-fomix transection (Leanza
et al., 1993), which may be important as a response to damage. Thus, considering the
above evidence, it is clear that cholinergic input to the hippocampus plays an important
role in the complex neurobiological processes of learning and memory, though it is
certainly not the only system involved.
Some evidence supports a potential trophic or regulatory role for IL-2 on the
cholinergic septohippocampal system. IL-2 enhances the survival of these septal and
hippocampal neurons in culture (Awatsuji et al., 1993), and modulates the activity of
choline acetyl transferase (ChAT) (Mennicken and Quirion, 1997). Also, IL-2 is a potent
biphasic modulator of acetylcholine (ACh) release (Hanisch et al., 1993; Seto et al.,
1997). At low concentrations (sub-pM), IL-2 stimulates the release of ACh, but at higher
concentrations (nM), IL-2 inhibits ACh release. Since IL-2 is difficult to detect in the
adult brain, endogenous brain IL-2 or IL-2 that crosses the BBB would be expected to be
present at a low concentration in the normal brain.
Taking into consideration 1) the neurotrophic and neuromodulatory role of IL-2 in
cultured septal and hippocampal neurons, 2) the regulatory effects of IL-2 on ACh release
and ChAT activity in vitro, 3) the spatial learning and memory impairments of IL-2-

9
deficient knockout mice and cancer patients undergoing IL-2 therapy, and 4) the
enhanced concentration of IL-2R subunits in the hippocampus, I hypothesized that IL-2
may play a role in the growth and differentiation of the septohippocampal cholinergic
neurons.
Statement of the Problem
Normal development of septal neurons depends on trophic factors that are
presumably secreted from the hippocampus. For example, the neurotrophins, nerve
growth factor (NGF) and brain-derived neurotrophic factor (BDNF) are expressed in the
hippocampus and have both been shown to be important in the development,
maintenance, and repair of septohippocampal neurons (Brooks et al., 1999; Conner et al.,
1992; Conner and Varn, 1997; Morse et al., 1993). Perhaps, IL-2 also mediates the
growth and development of septal and hippocampal neurons. This cytokine may act as a
growth factor by itself, or signal the release of growth factors from neurons or glia in the
hippocampal area.
In other investigations, the modulatory and growth-promoting effects of IL-2
were determined by either adding exogenous IL-2 to cultures or by injecting exogenous
IL-2 into the brain (in most cases, species non-specific, e.g., human IL-2 in rats or mice).
Thus, the studies utilizing exogenous IL-2 administration have several potential
shortcomings: 1) the cytokine has a short half-life, and the amount of IL-2 delivered may
not reflect physiologically relevant concentrations in the CNS in vivo (e.g., chronic vs.
acute dosing), 2) the nature of the injections disrupts the BBB, causing potentially even
more IL-2, as well as other cytokines, from the peripheral immune system to enter into
the brain, 3) it is unknown during which period in neurodevelopment that IL-2 might
exert these postulated effects, and 4) non-species specific IL-2 may have neurotoxic

10
effects, such as T and B cell invasion of the brain, angiogenesis, changes in the
composition of the extracellular matrix, myelin damage, and neuronal cell loss seen in
rats administered human IL-2 intracerebroventricularly via minipumps (Hanisch et al.,
1996; Hanisch et al., 1997b). Thus, my approach was to use IL-2 knockout mice. These
studies were the first to investigate the consequences of the absence of IL-2 on aspects of
brain development and maintenance of the septohippocampal system in vivo. Our
laboratory had found the aforementioned behavioral alterations in IL-2 knockout mice,
and therefore another important goal of my research was to test the hypotheses regarding
the neurobiological and neuroimmunological alterations that may underlie these
behavioral abnormalities.

CHAPTER 2
ALTERATIONS IN SEPTOHIPPOCAMPAL CHOLINERGIC NEURONS
RESULTING FROM INTERLEUKIN-2 GENE KNOCKOUT
Introduction
One of the earliest observations suggesting that cytokines could influence brain
function in humans came from cancer treatment trials in which interleukin-2 (IL-2) was
found to induce cognitive dysfunction and other untoward neuropsychiatric side effects in
patients (Denicoff et al., 1987). Although basic research has demonstrated that IL-2 can
modulate different aspects of central nervous system (CNS) function, some of IL-2s
most prominent neurobiological actions occur in the hippocampal formation and related
limbic regions, where receptors for this cytokine are enriched (Araujo et al., 1989;
Hanisch and Quirion, 1995a; Lapchak et al., 1991; Petitto and Huang, 1994, 2001; Petitto
et al., 1998).
Exogenously administered IL-2 has effects on a number of parameters of septal
and hippocampal neuronal function including trophic effects on cultured fetal septal and
hippocampal neurons (Awatsuji et al., 1993; Sarder et al., 1996; Sarder et al., 1993). IL-2
may also modify cellular and molecular substrates of learning and memory such as long
term potentiation (Tancredi et al., 1990), and multiple parameters of cognitive behavioral
performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al.,
1999; Nemni et al., 1992). Moreover, the neurotrophic and neuromodulatory actions of
IL-2 have been implicated in abnormal hippocampal development associated with
schizophrenia (Ganguli et al., 1995; Licinio et al., 1993; McAllister et al., 1995).
11

12
The potential effects of IL-2 on cholinergic neurons are particularly relevant to
this study. In addition to the aforementioned trophic effects of IL-2 on cultured septal
neurons, IL-2 is among the most potent modulators of acetylcholine (ACh) release from
cultured septohippocampal neurons (Araujo et al., 1989; Hanisch et al., 1993; Seto et ah,
1997), and can also modulate its precursor enzyme, choline acetyltransferase (ChAT) in
fetal neurons (Mennicken and Quirion, 1997). Alterations in the cytoarchitecture of
cholinergic septohippocampal neurons have been shown to correlate with differences in
spatial learning ability in mice (Schwegler et ah, 1996a; Schwegler et ah, 1996b). We
found that IL-2 knockout (IL-2 KO) mice exhibited impaired learning and memory
performance, sensorimotor gating, and reductions in hippocampal infrapyramidal mossy
neuronal fiber length (Petitto et ah, 1999), a factor shown previously to correlate
positively with spatial learning ability (Schopke et ah, 1991; Schwegler and Crusio,
1995; Schwegler et ah, 1988).
In the present study, we therefore sought to test the hypothesis that loss of IL-2
would result in abnormal neurodevelopment of septal cholinergic neurons that project to
the hippocampus. Since extensive data document that these neurons play a critical role in
learning and memory performance (Galey et ah, 1994; Leanza et ah, 1995), and given the
various in vitro neurotrophic and neuromodulatory effects of IL-2 on developing
septohippocampal cholinergic neurons, we postulated that IL-2 KO mice would have
fewer cholinergic neurons in the medial septum and vertical limb of the diagonal band of
Broca (MS/vDB) and a reduction in the cholinergic axonal density in the hippocampus.
To accomplish this goal, IL-2 KO and wild-type littermates were compared using
stereological techniques to count MS/vDB cholinergic somata stained with ChAT

13
immunohistochemistry, and image analysis methods to measure the density and
distribution of cholinergic neurites in several regions of the hippocampus labeled for
acetylcholine esterase (AChE), a reliable marker of cholinergic axons (Hedreen et al.,
1985).
Materials and Methods
Animals and Tissue Preparation
Mice used in these experiments were cared for in accordance with the NIH Guide
for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2
heterozygote by IL-2 heterozygote crosses. The polymerase chain reaction (PCR) was
used to genotype the offspring post-weaning (see below). The IL-2 KO mice, obtained
originally from the NIH repository at Jackson Labs, were derived from ten generations of
backcrossing onto the C57BL/6 background. Mice were housed under specific pathogen-
free conditions. Animals used in these experiments were 8-12 weeks of age.
Each animal was anesthetized with sodium pentobarbital (50 mg/kg) and perfused
with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS).
The brains were removed and fixed overnight in 4% paraformaldehyde followed by
overnight equilibration in 30% sucrose cryoprotective solution, and then were snap
frozen in isopentane (-80C) for storage. The brains were equilibrated to -20C prior to
cryostat sectioning into 40 pm slices in the coronal plane, collected into individual wells
of polystyrene 24-well plates (NUNC 1147), and stored free-floating at 4C in PBS for
histochemistry. Every third section was processed for ChAT immunohistochemistry,
AChE histochemistry, or cresyl violet Nissl staining.

14
Genotyping Using PCR
The genotypes of all mice were determined by the PCR. PCR reactions were
performed using a 25 pi total reaction volume containing 1 pM each of forward and
reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP, 0.3 pi Taq DNA
polymerase and amplified using a thermal cycler with a heated evaporation cover
(Ericomp). The cycling parameters were hot start 95C (3min), denaturing 94C (30
sec), annealing 64C (30 sec), extension 72C (45 sec) with a final extension step of 4
min. Thirty cycles were used for these experiments. The 5 and 3 primers for the IL-2
KO (500 bp knockout band amplified) were 5-TCGAATCGCCAATGACAAGACGCT-
3 and 5-GTAGGTGGAAATTCTAGCATCATCC-3\ The 5 and 3primers for the wild
type (324 bp wild type band amplified) were 5-
CTAGGCCACAGAATTGAAAGATCT-3 and 5-
GTAGGTGGAAAATTCTAGC ATC ATCC-3 .
ChAT Immunohistochemistry
Free-floating 40-pm sections were incubated for 20 minutes in 1% hydrogen
peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then
washed twice in PBS and blocked for 1 hr in 200 pl/well 3% normal goat serum (NGS).
After this incubation, the sections were incubated overnight in the primary antibody,
rabbit anti-ChAT (Chemicon AB143; 1:2000 dilution in PBS with 0.3% Triton X-100
and 1% NGS, 200 pl/well). The next day, the sections were washed twice in PBS and
incubated overnight in the secondary antibody, biotinylated goat anti-rabbit IgG (Sigma
B-7389; 1:1000 dilution in PBS with 0.3% TX-100 and 1% NGS). The sections were
then washed twice in PBS and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS)

15
for 2 hrs followed by two washes in PBS. The sections were developed in 0.5 mg/ml
3,3-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min and were
placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes,
and coverslipped. Figure 2-1 shows an example of ChAT immunostained section of the
MS/vDB.
Figure 2-1. ChAT immunohistochemical staining of the medial septum/vertical limb of
the diagonal band of Broca. The scale bar represents 200 pm.

16
AChE Histochemistry
AChE histochemistry was used as a marker of cholinergic innervation of the
hippocampus (Woolf et al., 1984). Brain sections were collected in individual wells of
24-well plates containing 250 plAvell 0.1 M pH 6.0 acetate buffer (AB). The sections
were washed twice with AB, then placed in 200 pi preincubation solution consisting of
aqueous 5 mM sodium citrate, 3 mM cupric sulfate, and 0.5 mM potassium ferricyanide.
The sections were incubated for 20 minutes at room temperature on a shaker a low speed.
After the preincubation period, 200 pi of the incubation solution was added to each well
consisting of the same make-up as the preincubation solution supplemented with 4.84
mM acetylthiocholine iodide and 0.4 mM ethopropazine. The multi-well plate was
packed on top of crushed ice and microwaved at 200 W for 2 minutes. The solution was
then removed and the sections were washed twice in 0.05 M TRIS pH 7.6 buffer
followed by AB. The reaction product was intensified with 0.5 mg/ml DAB, 2.5% nickel
sulfate, and 0.01 % H2O2 in AB for 5-7 min or until definitive staining could be detected
in the hippocampal subregions. Sections were then mounted on slides, dehydrated in
grade ethanol washes, cleared in xylenes, and coverslipped for imaging.
Cholinergic Stereologv
Stained cholinergic neuronal somata of the MS/vDB were counted using the
software MCID 5.1 and the three-dimensional counting box (optical dissector) method
described by Williams and Rakic (Williams and Rakic, 1988). All stereology was
performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2
microscope with a motorized x-y stage made by Imaging Research, Inc. The latter is
capable of making movements as fine as 0.1 pm. Every third section through the
anterior-posterior extent of the septal region was sampled. The regions to be counted

17
were outlined at lOx magnification and the size of the counting boxes were generated to
be approximately 5% of the most rostral, and therefore, smallest, area of the MS/vDB
(defined by the section where the corpus collosum first joins in the midline). The size of
the outlined count regions, but not the counting box, varied depending on where the
individual section was taken from the rostral to caudal extent of the MS/vDB. The
defined counting box was approximately 2-2.5% of the outlined count area of the largest
single section of the MS/vDB.
To assess whether the predicted septal cholinergic alterations in IL-2 KO mice
might be associated with a general effect on cholinergic neurons in the brain, striatal
cholinergic somata were also counted in the right hemisphere in the sections that also
contained the MS/vDB. Except for a different magnification used to outline the striatum
(5X), the sampling parameters were identical to those used to generate estimates of septal
cholinergic neuron number. The guard volume was set at 1 pm for the top and bottom of
the section and the counting cubes were randomly distributed throughout the user-defined
count area with a total sampling frequency of 25% (-8.3% of the total area since every
third section was sampled). Only somata that were clearly and distinctly stained were
counted. Each counting box was examined at 40x magnification (20x for striatal
neurons) and the computer-assisted focus was used to scan from the top to the bottom of
the counting box. Cells were counted only if they were either completely inside of the
counting box, or partially inside of the box on the top, back, or left side. They were not
counted if they fell outside of the box or crossed into the box anywhere on the bottom,
front, or right side. The cells that were counted were labeled on the monitor by clicking
the mouse pointer on each cell and the MCID software recorded the number of marks.

18
The MCID software interpolated the total volume of the MS/vDB based on the
volume of the count areas defined by the user. Cell density (Ny) was estimated by
dividing the total number of cells counted by the volume of the counting boxes, which
was also tracked by the software. The total cell number was estimated by multiplying Ny
by the total volume.
Quantitative Image Analysis of AChE Staining
For quantitative analysis of AChE staining, we modified previously described
methods used to measure intensity of staining and comparisons of normalized length
across CA1, CA3b, and the suprapyramidal (SP) and infrapyramidal (IP) blades of the
dentate gyrus (DG) (King et al., 1989; Schwegler et al., 1996b). Images of the
hippocampus in each tissue section sampled from a light microscope (Olympus BH-2)
were relayed by digital video camera (Hitachi KP-D581) to a computer frame grabber
(Flashpoint 128, Integral Technologies) and digitized to 640x480 pixel images with 256
gray levels from black to white. Imaging software (Image Pro Plus v.4.0, Media
Cybernetics) was utilized to define a broad sampling traverse across various areas of the
hippocampus, approximately perpendicular to the cell body layers. For CA1 and CA3b,
this traverse extended from the alveus to the hippocampal fissure. For the DG, the line
extended from the hilus to hippocampal fissure or to the pial surface for the SP or IP limb
of the DG, respectively. AChE-containing fiber density was estimated by using the gray
level of each point along the line, which was calculated by averaging the intensity value
of each pixel across the width of the sampling band (e.g. a traverse with a width of 46
pixels would have an average of 46 measurements approximately parallel to each data
point). Gray level intensity measurements were used to estimate the AChE reaction
product density at each point along the traverse. The mean pixel intensity of a small box

19
positioned in the corpus collosum was used to represent the background staining intensity
of tissue containing little cholinesterase activity. All of the average intensities along the
line were then plotted by traverse position to illustrate the quantitative patterns of AChE
distribution across each subregion. Conspicuous inflections marking the transition from
alveus (Alv) to stratum oriens (SO, in CA1, CA3b), from stratum lacunosum-moleculare
(Lmol) to dentate molecular layer (Mol) (hippocampal fissure (HiF); CAI, CA3, DG),
and polymorph zone (PoDG) to granule cell layer (GCL) were used to align traverses
across sections and animals (Figure 2-2). Also, length-normalized comparisons of the
Figure 2-2. A micrograph of the sampling regions utilized for image analysis of AChE-
staining. Alv=alveus, DG=dentate gyrus, GCL=granular cell layer
(IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus,
Lmol=lacunosum moleculare, Mol=molecular layer, Pi=pial surface, SO=stratum
oriens, SP=septum pellucidum, SR=stratum radiatum. The scale bar represents
200 pm.
measured regions were made by converting the sampling lengths to 100 points using the
software Matlab v.5.3. Dependent variables were the absolute length of traverses, AChE
intensity at anatomically identifiable inflection points (IP and SP bands, Lmol, dentate

20
inner Mol) and subregions (SO, stratum radiatum (SR), dentate outer Mol), and derived
values for absolute and relative positions of, and distances between, these landmarks. For
each animal, three hippocampal slices were measured on both the left and right
hemispheres of the brain and the measured intensities or distances were averaged together
for statistical analysis.
Cresyl Violet Staining
Every third section was Nissl stained to provide a qualitative view of the
boundaries between various forebrain regions. The sections were placed on slides and
allowed to air-dry. The slides were immersed in 60C cresyl violet for 45 sec, washed in
running distilled water to remove the excess cresyl violet, dehydrated in graded ethanol,
cleared in xylenes, and coverslipped.
Results
Comparison of Cholinergic Somata in the MS/vDB
Figure 2-3 shows the total number of ChAT-positive cells stereologically counted
from the MS/vDB of IL-2 KO and wild-type mice. As seen in this figure, the IL-2 KO
mice had approximately 26% fewer cholinergic somata in this region than wild-type
controls. An ANOVA confirmed that this group difference was statistically significant
(F(l,16)=8.6, p=.01). By contrast, the number of ChAT-positive somata in the striatum
of IL-2 KO and wild-type mice were not different. There were no significant gender
differences in either brain region.
Density of AChE-positive Fibers in Regions of the Hippocampus
The average AChE-staining intensity curves were generated by defining a region
across CA1, CA3b, and the SP and IP blades of the DG. Figure 2-4 shows the intensity
curves that were generated from imaging the AChE-histochemically stained sections for

21
Cholinergic Somata in the MSA/DB
2000
| 1500
O
>
o 1000
4
I-
2
o
500
0
Fig. 2-3. ChAT-positive somata are significantly reduced in the MS/vDB of IL-2
knockout mice. Each bar represents the mean SEM of 9 animals per group.
*p=0.01.
CA1 (Figure 2-4a), CA3 (Figure 2-4b), the SP-GCL (Figure 2-4c), and IP-GCL layer of
the DG (Figure 2-4d). Repeated measures ANOVA was performed on regions of the
average normalized curves selected by areas that appeared to deviate between the groups.
None of these areas, however, were found to differ between IL-2 KO and wild-type mice.
Morphology of the Granular Cell Layer of the Lower Limb of the DG
There were no differences in the groups for the Y-axis (intensity) data. The
variations in the patterns of the DG curves X-axis (distance) were also compared.
Distances were compared by defining the point of lowest intensity in the regions that the
curves indicated as each transition between the regions of interest. Distances are reported
as a percentage of the total distance across each curve. The IP blade of the DG was
broken into three regions: the Hil, the IP-GCL, and the Mol (Figure 2-4d). As depicted in

22
A)
Average AChE Intensity of CA1
B)
Average AChE Intensity of CA3
stance (% of Total) Dtaence(% of Total)
C)
Average AChE Intensity of the Suprapyramldal Limb
of the Dentate Gyrus
D)
Average AChE Intensity of the Infrapyramidal Limb
of the Dentate Gyrus
IL-? ¡Miockout
O WHd*Typ
Fig.2-4. IL-2 knockout mice do not differ in measures of average intensity of AChE-
staining across CA1, CA3b, and the suprapyramidal and infrapyramidal layers of
the DG. Each curve represents the mean of 9 wild-type (open circles) or 9 IL-2
knockout mice (closed circles). Alv=alveus, GCL=granular cell layer
(IP=infrapyramidal, SP=suprapyramidal), Hif=hippocampal fissure, Hil=hilus,
Lmol=lacunosum moleculare, Mol=molecular layer, Pi=pial surface, SO=stratum
oriens, SP=septum pellucidum, SR=stratum radiatum. A) The CA1 curve shows
the average intensity from the Alv to the Hif. The arrows delineate the transitions
between the hippocampal substructures including the SO, SP, SR, and Lmol. B)
The pattern of the peaks and valleys of the CA3 region curve is similar to that of
the CAL C) The intensity curve of SP-GCL begins at the Hif and terminates at
the dorsal border of the IP-GCL. Arrows delimit the borders of the Mol, SP-GCL,
and Hil. D) The curve representing the IP-GCL begins at the ventral border of the
SP-GCL and continues to the pial surface. The second and third solid arrow
define the borders in wild-type mice between the Hil and IP-GCL and the IP-GCL
and Mol, respectively. The broken arrows define these same borders in IL-2
knockout mice.

23
Figure 2-5, the distance across the IP-GCL was significantly reduced in the IL-2 KO
mice compared to wild-type mice (F( 1,16)=9.2, p=0.008). The distances across the Hil
and Mol, however, were not significantly different. The SP blade of the DG was
separated into three regions: the Mol, the SP-GCL, and the Hil (Figure 2-4c). There were
no significant differences in length between groups across any of the internal blade
regions.
Distance Across GrDG of the External Blade
35 -
Fig. 2-5. Distance across the GrDG of the external blade was significantly decreased in
IL-2 knockout mice compared to wild-type mice. Each bar represents the mean
SEM of nine animals per group. *p=0.008.
Discussion
These data are the first to demonstrate that loss of endogenous IL-2 results in
reduction in the number of MS/vDB cholinergic neurons and structural alterations in the
morphology of the dentate gyrus. Given the role of septohippocampal cholinergic
neurons in learning and memory (Galey et al., 1994; Leanza et al., 1995; Schwegler et ah,
1996a; Schwegler et ah, 1996b), sensorimotor gating (Caine et ah, 1992; Curzon et ah,

24
1994; Koch, 1996a), and the aforementioned influence of IL-2 on parameters of
cholinergic function, the present findings are consistent with alterations in these
behavioral measures that we have reported previously in IL-2 KO mice (Petitto et al.,
1999).
In this study, the number of cholinergic cell bodies in the MS/vDB of wild-type
mice were comparable to those reported previously for C57BL/6 mice (Schwegler et al.,
1996b). Cholinergic somata were reduced by 26% in IL-2 KO mice as compared to wild-
type mice. This was not a general effect on cholinergic neurons in the brain, however, as
striatal ChAT-positive neurons were not significantly affected. This finding is also
consistent with previous research showing that exogenous IL-2 has potent effects on ACh
release from septohippocampal neurons, whereas cholinergic intemeurons in the striatum
do not respond to IL-2 (Hanisch et al., 1993). One potential caveat of using ChAT as a
marker of cholinergic neurons is that differences in cell counts may be due to a decrease
in ChAT labeling intensity rather than a reduction in cholinergic neurons (Ward and
Hagg, 2000). This would appear unlikely, however, as we did not note any appreciable
differences in the staining intensity between groups.
Alterations in AChE-staining intensity did not differ as we hypothesized. It is not
likely that the unexpected lack of difference in AChE-staining intensity would be
attributable to the length normalization technique used, as the patterns of the average
intensity curves seen in Figure 2-4 were remarkably similar between the groups. This
unexpected result may, however, be due to compensatory sprouting of the surviving
MS/vDB neurons during development. Indeed, numerous other studies have found that
septohippocampal neurons can undergo compensatory sprouting in response to injury

25
(Cassel et al., 1997; Gage and Bjorklund, 1987; Gage et al., 1984; Gage et al., 1983b),
and in animal disease models such as Alzheimers transgenic mice (Bronfman et al.,
2000). Future studies are needed to address this issue more directly.
Another significant finding of the current study was that IL-2 KO mice exhibited
structural alterations in the distance across the IP-GCL. The neurons of the GCL have
been associated with learning and memory (Collier and Routtenberg, 1984; Conrad and
Roy, 1993; McLamb et al., 1988; Nanry et al., 1989; Walsh et al., 1986), and are also a
target for septohippocampal cholinergic axon termination (Makuch et al., 2001). In situ
hybridization studies have found the GCL to be enriched in IL-2 receptors (Petitto and
Huang, 2001; Petitto et al., 1998), supporting a possible role for IL-2 in the observed
structural alterations. Also, GCL development progresses from the SP layer to the IP
layer (Bayer, 1980). The differences seen in the IL-2 KO mice in this study may indicate
a failure of these late stage granule cells to fully develop or survive. Whether these
structural changes are due to a reduced number of GCL cells or a decrease in the cell
body size of these neurons requires further investigation.
The most likely mechanism whereby loss of IL-2 results in these changes in the
septohippocampal cholinergic system would appear to be due to the absence of its
neurotrophic actions during development. As noted earlier, IL-2 enhances neurite
extension and survival of cultured fetal septal and hippocampal neurons (Awatsuji et al.,
1993; Hanisch and Quirion, 1995a; Sarder et al., 1996; Sarder et al., 1993), and thus, the
absence of these intrinsic effects of IL-2 could account for the observed neuroanatomical
alterations. Another mechanism that may account for these findings is the possibility that
the loss of endogenous IL-2 may result in lower levels of tonic ACh release during

26
critical periods of neurodevelopment. Release of ACh by developing neurons has been
shown to be important for growth cone guidance (Zheng et al., 1994), neuronal growth
and differentiation, synaptic plasticity (Lauder and Schambra, 1999), and survival of
newly developed neurons (Knipper and Rylett, 1997). In fact, some evidence indicates
that ACh released from developing neurons may engage in a positive feedback
mechanism with nerve growth factor (NGF) (Knipper et al., 1994), a member of the
neurotrophin family that is essential for the normal development of septal cholinergic
neurons (Arimatsu and Miyamoto, 1991; Hartikka and Hefti, 1988; Mobley et al., 1986;
Ruberti et al., 2000). In a series of studies, Quirions laboratory has demonstrated that
IL-2 is among the most potent modulators of ACh release from mature brain slices and
fetal neurons in vitro, and can upregulate ChAT in fetal septal neurons in culture
(Hanisch et al., 1993; Mennicken and Quirion, 1997; Seto et al., 1997). It is therefore
possible that the loss of such potent actions of IL-2 during development could account, in
part, for the cytoarchitectural alterations found in this study. Indeed, both IL-2s
neurotrophic effects and action on cholinergic release may well be operative and
interactive with one another. Nevertheless, the IL-2 KO mice do not exhibit complete
loss of septal cholinergic neurons suggesting that the effects of IL-2 on MS/vDB neurons
are likely secondary to other trophic factors like NGF.
These experiments do not enable us to differentiate between the contributions of
the loss of central versus peripheral IL-2, and thus, it remains to be determined whether
these septohippocampal cholinergic abnormalities are due primarily to the absence of
central, peripheral, or a combination of both sources of IL-2. There is some evidence that
endogenous IL-2 may be produced in neuronal areas of the mammalian hippocampal

27
formation, where its release may regulate the development and function of septal
cholinergic neurons projecting to the hippocampus. IL-2-like immunoreactivity has been
localized to the hippocampal formation in rat forebrain (Lapchak et al., 1991), and
detected in tissue extracts from rat and human hippocampal tissue (Araujo et al., 1989).
In mouse brain, IL-2 mRNA has been found in the hippocampus (Villemain et al., 1991),
and transcripts for this cytokine may be expressed in rat astrocyte cultures as well
(Eizenberg et al., 1995).
In the periphery, absence of endogenous IL-2 leads to an immunodysregulation
that produces loss of self-tolerance and IL-2 KO mice eventually develop generalized
systemic autoimmune disease (although C57BL/6-IL-2 KO mice develop clinical signs of
systemic autoimmunity at a substantially slower rate than other strains such as Balb/c or
C3H) (Petitto et al., 2000). Therefore, it is reasonable to speculate that the
neuroanatomical alterations found in the IL-2 KO mice result from peripheral
autoimmune processes. Autoimmunity could impact on brain development or induce
neurodegeneration. The former seems unlikely, however, since IL-2 KO mice do not
express the first signs autoimmunity (e.g., splenomegaly) until at least three to four weeks
afterbirth (Horak, 1995); by this time, septohippocampal development should already be
complete (Bender et al., 1996; Chandler and Crutcher, 1983; Super and Soriano, 1994;
Yoshida and Oka, 1995). Furthermore, the likelihood that these neuroanatomical
alterations may be due to autoimmune-induced degeneration of existing neurons also
seems unlikely, since lymphocytes cannot be detected in the brain of adult IL-2 KO mice
(Petitto et al., 1999). Nonetheless, since autoimmunity has been associated with
cognitive changes in both animals and humans (Lai and Forster, 1988; Sakic et al., 1997;

28
Sakic et al., 1993), more subtle autoimmune processes may be at play in the IL-2 KO
mice (e.g., autoantibodies). It would be of interest to explore the observed cholinergic
changes in old versus neonatal mice to determine if the abnormalities increase with age
due to neurodegeneration, or are primarily the result of abnormal development. Such
knowledge will then enable us to develop a more specific model to test relevant
hypotheses involving IL-2 at specific anatomical sites in the septohippocampal
cholinergic system.
In summary, these data demonstrate that loss of endogenous IL-2 results in
reduction in the number of cholinergic neurons in the MS/vDB and alterations in the
structural morphology of dentate projection fields. These findings extend our previous
experiments showing that spatial learning and hippocampal mossy fiber length are
abnormal in IL-2 KO mice (Petitto et al., 1999). Further research is needed to determine
whether these outcomes in IL-2 KO mice may be due to the absence of central or
peripheral IL-2 during neurodevelopment (or some combination of both sources),
neurodegeneration secondary to peripheral autoimmunity, or other factors associated with
the absence of IL-2.

CHAPTER 3
ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN
INTERLEUKIN-2 KNOCKOUT MICE
Introduction
Interleukin-2 (IL-2) has been implicated in the pathogenesis of multiple sclerosis
and several major neuropsychiatric disorders such as Alzheimer's disease, schizophrenia,
and Parkinson's disease (Hanisch and Quirion, 1995b). Furthermore, in case studies of
humans receiving IL-2 treatment for cancer therapy, prolonged exposure to IL-2 was
found to induce cognitive dysfunction and other untoward neuropsychiatric side effects
(Denicoff et al., 1987). Although IL-2 has been shown to be capable of modulating
different aspects of central nervous system (CNS) function, many of its known effects in
the limbic system occur in the hippocampal formation, where receptors for this cytokine
are enriched (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991;
Petitto and Huang, 1994, 2001; Petitto et al., 1998). IL-2 may, for example, modify
cellular and molecular substrates of learning and memory such as long-term potentiation
(Tancredi et al., 1990), and can affect multiple parameters of cognitive behavioral
performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al.,
1999; Nemni et al., 1992). IL-2 can provide trophic support to primary cultured neurons
from multiple region of the rat brain, including the hippocampus and medial septum
(Awatsuji et al., 1993; Sarder et al., 1993), and positively affects the morphology of
neurite branching from rat hippocampal cultures (Sarder et al., 1996; Sarder et al., 1993).
Furthermore, IL-2 has been shown to be one of the most potent modulators of
29

30
acetylcholine (ACh) release from rat hippocampal slices (Hanisch et al., 1993; Setoet al.,
1997), and can also increase the activity of its precursor enzyme, choline
acetyltransferase (ChAT) (Mennicken and Quirion, 1997).
Previously, we found that IL-2 knockout mice (IL-2 KO) exhibited impaired
learning and memory performance, sensorimotor gating, and reductions in hippocampal
infrapyramidal mossy neuronal fiber length (Petitto et al., 1999), a factor which correlates
positively with spatial learning ability (Schopke et al., 1991; Schwegler and Crusio,
1995; Schwegler et al., 1988). We also found in studies of IL-2 KO mice in vivo, there
was a marked reduction in cholinergic somata in medial septal/vertical limb of the
diagonal band of Broca (MS/vDB) region, as well as decrease in the distance across the
infrapyramidal granule cell layer (IP-GCL) of the dentate gyrus (DG) (Beck et al., 2002).
Variation in the cytoarchitecture of cholinergic septohippocampal neurons correlate with
differences in spatial learning ability in mice (Schwegler et al., 1996a; Schwegler et al.,
1996b).
Research has shown that the neurotrophins, nerve growth factor (NGF) and brain-
derived neurotrophic factor (BDNF) expressed in the hippocampus, can be important in
the development, maintenance, and repair of septohippocampal neurons in vitro
(Arimatsu and Miyamoto, 1991; Conner and Varn, 1997; Gahwiler et al., 1987;
Hartikka and Hefti, 1988; Morse et al., 1993). Similar trophic effects have been noted in
studies utilizing infusion of exogenous neurotrophins in vivo (Hagg et al., 1990; Morse et
al., 1993), and in studies of transgenic and knockout mice (Ruberti et al., 2000; Ward and
Hagg, 2000). In the peripheral immune system of multiple animal species, both NGF and
BDNF are expressed by T lymphocytes (Braun et al., 1999; Kerschensteiner et al., 1 999;

31
Mizuma et al., 1999; Moalem et al., 2000). Though IL-2 regulates several aspects of T
cell function, the production or release of NGF and BDNF from T lymphocytes by IL-2
has not been tested directly(Carter et al., 1998; He and Malek, 1998), and conversely, it is
not known whether NGF and BDNF can modulate IL-2 in the brain.
In the present study, we sought to expand our previous findings that loss of
endogenous IL-2 in knockout mice led to reductions in cholinergic neurons of the
MS/vDB and a decrease in the distance across the IP-GCL of the DG. First, we
compared parvalbumin (Parv)-labeled somata in 8-12 week old IL-2 KO and wild-type
littermates to test the hypothesis that the loss of IL-2 is selective for cholinergic, but not
GABAergic cell bodies loss in the MS/vDB (e.g., not a general effect occurring on all
neurons in this region of the brain). Second, we compared cholinergic MS/vDB somata
between younger wild-type and IL-2 KO mice at postnatal day 21 (P21), an age where
septohippocampal development in mice is nearly complete (Armstrong et al., 1987;
Gould et al., 1991; Makuch et al., 2001). This age also precedes the development of
autoimmune disease in IL-2 KO mice of the C57BL/6 background, e.g., absence of
splenomegaly, lymphadenopathy, and inflammatory bowel disease. Third, we sought to
expand on the previous finding that there was a reduction in distance across IP-GCL by
performing stereological cell counts of Nissl-stained dentate gyri of IL-2 KO and wild-
type littermates to determine if the reduction in distance could be contributed to a
reduction in granule cell number. Finally, we tested the hypothesis that loss of IL-2 may
impact the expression and release of the neurotrophins, NGF and BDNF, which may
contribute to the MS/vDB cholinergic and GCL deficits that we observed previously.

32
Methods
Animals and Genotyping
Mice used in these experiments were cared for in accordance with the NIH Guide
for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2
heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the
NIH repository at Jackson Laboratories, were derived from ten generations of
backcrossing onto the C57BL/6 background. Mice were housed under specific pathogen-
free conditions. Animals used in these experiments were either 21-days-old (for ChAT
immunohistochemistry) or 8-12 weeks of age. All experiments were performed with
independent groups of animals. The specific animal numbers utilized are reported at the
beginning of each method descriptions below.
The genotypes of all mice were determined by the polymerase chain reaction
(PCR). PCR reactions were performed using 25 pi total reaction volume containing 1
pM each of forward and reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP,
0.3 pi Taq DNA polymerase, and amplified using a thermal cycler with a heated
evaporation cover (Ericomp). The cycling parameters were hot start 95C (3min),
denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final
extension step of 4 min. Thirty cycles were used for these experiments. The 5 and 3
primers for the IL-2 KO (500 bp knockout band amplified) were 5-
TCGAATCGCCAATGACAAGACGCT-3 and 5-
GTAGGTGGAAATTCTAGCATCATCC-3. The 5 and 3primers for the IL-2 wild
type (324 bp wild type band amplified) were 5-

33
CTAGGCCACAGAATTGAAAGATCT-3 and 5-
GTAGGTGG AAAATTCTAGC ATC ATCC-3 .
Immunohistochemistrv
For Parv stereological cell counts, six animals per group were used and for 21-
day-old ChAT stereological cell counts, seven animals per group were used. Each animal
was anesthetized with an injection cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20
mg/ml): acepromazine (10 mg/ml) at a dose of 0.015 ml injection cocktail/g body weight
and perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered
saline (PBS). The brains and spleens were removed and fixed overnight in 4%
paraformaldehyde, followed by overnight equilibration in 30% sucrose cryoprotective
solution, and then, were snap frozen in isopentane (-80C) for storage. The spleens were
weighed to assay for relative splenomegaly of IL-2 KO vs. wild-type mice. The brains
were equilibrated to -20C prior to cryostat sectioning into 50 pm slices in the coronal
plane, collected into individual wells of polystyrene 24-well plates (NUNC 1147), and
stored free-floating at 4C in PBS for histochemistry. Every third section was processed
for Parv or ChAT immunohistochemistry, or cresyl violet Nissl staining.
Free-floating 50-pm sections were labeled for Parv and ChAT-
immunohistochemistry as described previously (Beck et al., 2002). Briefly, they were
incubated for 20 minutes in 1% hydrogen peroxide (H2O2) to quench endogenous
peroxidative activity. The sections were then washed and blocked for 1 hr in 200 pl/well
3% normal goat serum (NGS). After this incubation, the sections were incubated
overnight in the primary antibody, rabbit anti-ChAT (Chemicon; 1:2000 in PBS with
0.3% Triton X-100 and 1% NGS, 200 pl/well) or rabbit anti-Parv (Chemicon; 1:1000 in
PBS with 0.3% Triton X-100 and 1% NGS, 200 pl/well). The next day, the sections were

34
washed and incubated overnight in the secondary antibody, biotinylated goat anti-rabbit
IgG (Sigma B-7389; 1:1000 dilution in PBS with 0.3% TX-100 and 1% NGS). The
sections were then washed and incubated in ExtrAvidin (Sigma E-2886; 1:1000 in PBS)
for 2 hrs. The sections were developed in 0.5 mg/ml 3,3-diaminobenzidine (DAB). 0.2
mg/ml urea H2O2 for approximately 5 min and were placed on slides, dehydrated in
graded ethanol washes, cleared in two changes of xylenes, and coverslipped.
Cresyl Violet Staining
For stereological cell count of Nissl-stained granule cells, seven animals per
group were used. The tissue for this assessment was selected, because it originated from
animals utilized in a previous study of cholinergic differences in IL-2 KO mice (Beck et
al., 2002). Every third section was Nissl stained to provide a qualitative view of the
boundaries between various forebrain regions, as well as labeling of hippocampal granule
cells layers (GCL) for stereological counting. The sections were placed on slides and
allowed to air-dry. The slides were immersed in 25C cresyl violet for 10 min, washed
vigorously in rapid exchanges of distilled water to remove the excess cresyl violet,
dehydrated in graded ethanol, cleared in xylenes, and coverslipped.
Stereology
Stained neuronal somata of the MS/vDB or IP and SP-GCL were counted using
the software MCID 5.1 as previously described (Beck et al., 2002). All stereology was
performed using a CCD High Resolution Sony camera and a Zeiss Axioplan 2
microscope with a motorized x-y stage made by Imaging Research, Inc. Every third
section through the anterior-posterior extent of the MS/vDB or IP and SP-GCL regions
were sampled. The regions to be counted were outlined at lOx (MS/vDB) or 20x (IP and
SP-GCL) magnification and the size of the counting boxes were generated to be

35
approximately 2-2.5% of the outlined count area of the largest single section of the areas
of interest.
The rostral border of the GCL count area was defined as the first section where
the dentate granule cell layer clearly separated from the pyramidal layer of CA1. The
caudal border was defined as the first section where the habenular commisure was visible
in the third ventricle. Only cells that could clearly be determined to be part of either the
IP or SP-GCL were counted; any cells in the area where the IP and SP-GCL connected
were left uncounted. Furthermore, only Nissl-stained cells with clearly visible nucleoli
were counted. For the MS/vDB, the rostral border was determined as the first section
where the corpus collosum connected in the midline of the section and the caudal border
was the first section where the anterior commisure joined in the midline.
The guard volume was set at 2 pm for the top and bottom of the section and the
counting cubes were randomly distributed with a total sampling frequency of the outlined
count area of 25% for MS/vDB and 33% for IP and SP-GCL (-8.3% and 11% of the total
area respectively, since every third section was sampled). The outlined counting area
was defined by the user and only somata that were clearly and distinctly stained were
counted. Each counting box was examined at 40x magnification and the computer-
assisted focus was used to scan from the top to the bottom of the counting box. Cells
were counted only if they were either completely inside of the counting box, or partially
inside of the box on the top, back, or left side. They were not counted if they fell outside
of the box or crossed into the box anywhere on the bottom, front, or right side. The cells
that were counted were labeled on the monitor by clicking the mouse pointer on each cell
and the MCID software recorded the number of marks.

36
The MCID software interpolated the total volume of the MS/vDB based on the
volume of the count areas defined by the user. Cell density (Nv) was estimated by
dividing the total number of cells counted by the volume of the counting boxes, which
was also tracked by the software. The total cell number was estimated by multiplying Nv
by the total volume.
Enzvme-Iinked Immunosorbent Assay (ELISA) Characterization of NGF and
BDNF
For measurement of BDNF in hippocampus and MS/vDB, nine IL-2 KO and
seven wild-type mice were used. For measurement of NGF protein levels, seven animals
per group were used. Initial test runs revealed that some, but not all, of the NGF protein
levels fell below the sensitivity of detection for the kit; therefore, the homogenized NGF
samples were spiked with 25 pg/ml of the known NGF standard included with the kit to
bring any low levels of expression above the kits 15.6 pg/ml lower detection limit. One
column of the ELISA plate was also run with only the 25 pg/ml standard spike to provide
a baseline and the data reported are corrected for this. Animals used for ELISA
characterization of neurotrophin levels only received saline perfusion and were not post-
fixed in parafonnaldehyde. The brains were removed, snap frozen, and then allowed to
equilibrate to -20 C. The brains were sectioned on a cryostat at -20-22 C at 400 pm
thickness and the MS/vDB and hippocampi were dissected with a 0.75 mm micropunch
on a -20 C freezing platform. The dissected tissue was weighed on a microgram scale,
and then transferred to 25 pi of homogenizing solution (50 mM Na/Na2 and 0.2% TX-
100 in H20 with Anti-protease Complete TM cocktail (Boehringer)) per mg of wet
weight tissue. The tissue was sonicated in the homogenizing solution for 30-sec on ice

37
and centrifuged at 16,000 g for 15 min at 4 C. The supernatant was collected and stored
at -20 C for ELISA analysis.
Levels of NGF and BDNF were analyzed in the homogenates from MS/vDB and
hippocampus using a commercially available Emax Immunoassay System according to the
manufacturers instructions (Promega). Briefly, the 96-well plates were coated with
1:6,250 anti-NGF pAb in carbonate coating buffer (0.025 M sodium bicarbonate, 0.025
M sodium carbonate, pH 9.7) and incubated overnight at 4 C. The plates were washed
with TBST wash buffer (20 mM Tris-HCL pH 7.6, 150 mM NaCl, 0.05% (v/v) Tween
20) and blocked with lx Block and Sample buffer (provided with kit) for 1 hour. The
plates were washed again with TBST and a set of standard curves were generated in
duplicate by performing 1:2 dilutions of a known 500 pg/ml standard in a range from 500
pg/ml to 7.6 pg/ml followed by a blank well of 0 pg/ml. All added samples and
standards were allowed to incubate at 25 C for 6 hours. The plates were washed
thoroughly with TBST and 1:4000 anti-NGF mAb was added and incubated overnight at
4 C. The plates were again washed with TBST and 1:100 anti-rat IgG pAb conjugated
to HRP was added for 2.5 hours at room temperature. The plates were washed and TMB
One Solution was added for color development for 10 min. The reaction was stopped
with the addition of equal volume of 1 N HC1 and the absorbance was read at 450 nm
within 30 min of the color development reaction. The data were reported as pg of protein
per mg wet weight tissue.
Statistical Analysis
Results are reported as the mean SEM. Statistical differences between groups
were determined using analysis of variance (ANOVA).

38
Results
Cholinergic MS/vDB Cell Number in 21-day-old Mice and GABAergic Cell Number
in Adult Mice
In 8-12-week-old mice, no significant differences were apparent in the relative
number of stereologically counted Parv-positive neurons between groups (F(l,10)-0.002,
p=0.964). Thus, the GABAergic neurons appear to be unaffected by IL-2 gene deletion.
We did not assay for GABAergic alterations in younger animals, since there were no
differences in the adult IL-2 KO mice relative to the wild-types.
In contrast to the previously reported data from 8-12-week-old animals (Beck et
al., 2002), there was no significant difference in stereologically counted cholinergic
somata number in the MS/vDB of 21-day-old IL-2 KO mice relative to wild-type mice
(F(l,12)=0.689, p=0.423). As expected, the 21-day-old KO mice also did not exhibit
splenomegaly seen in the autoimmune 8-12-week-old group, as spleen weights did not
differ between 21-day-old wild type and IL-2 KO mice (F(l,12)=0.989, p=0.340).
Reduction in the IP-GCL Neuronal Number in IL-2 KO Mice
The IP-GCL of IL-2 KO mice had significantly fewer neuronal somata than wild-
type mice (Fig. 3-1; F( 1,12)=10.966, p=0.006). In the SP-GCL, however, there was no
significant difference in granule cell number (Fig 3-1.; F( 1,12)=0.197, p=0.665).
Alterations in Neurotrophin Levels
Levels of NGF protein in hippocampal tissue homogenates was significantly
increased in IL-2 KO mice relative to wild-type mice (Fig. 3-2A; F( 1,10)8.261,
p=0.017). The levels of BDNF protein, conversely, were significantly decreased in IL-2
KO mice relative to wild-type mice (Fig. 3-2B; F( 1,12)=8.023, p=0.015).

39
Stereological Cell Count of Nissl-stained
Granule Cell Layers
1200
r 1 1 1
Infrapyramidal Suprapyramidal
I 1 Wild-Type
I IL-2 Knockout
Fig. 3-1. There was a significant reduction in infrapyramidal, but not suprapyramidal
granule cells in IL-2 knockout relative to wild-type mice. Each bar represents the
mean SEM of seven animals per group. *p=0.006.
A)
Enhanced NGF Levels In the Hippocampus
of IL-2 Knockout Mice
B)
Reduced BDNF Levels in the Hippocampus
of IL-2 Knockout Mice
Genotype
Genotype
Figure 3-2. There was a significant: A) increase in NGF and B) decrease in BDNF
protein levels in the hippocampus of IL-2 knockout compared to wild-type mice.
The NGF bars represent the mean SEM of 7 animals per group. The BDNF
bars represent mean SEM of 7 IL-2 knockout and 9 wild-type mice. *p<0.05.

40
Discussion
These data are the first to demonstrate that the loss of endogenous IL-2 in
knockout mice can lead to alterations in neuronal cell number in the IP-GCL and
production of the neurotrophins, BDNF and NGF. Further, this study expands upon the
previous finding that IL-2 gene deletion leads to a deficiency of cholinergic neurons in
the MS/vDB (Beck et al., 2002), by showing a lack of significant cholinergic neuronal
differences in MS/vDB of 21-day-old IL-2 KO animals or GABAergic alterations in adult
animals.
The lack of a difference in GABAergic neurons in 8-12-week-old adult mice was
consistent with our initial hypothesis, as there is no evidence in the literature that IL-2 has
any modulatory effects on GABAergic neurons. Moreover, this is not a regional effect,
but rather appears to be selective to cholinergic projection neurons. As previously
mentioned, IL-2 is a potent modulator of ACh release (Hanisch et al., 1993; Seto et al.,
1997), and its precursor enzyme ChAT (Mennicken and Quirion, 1997), suggesting an
effect of IL-2 on cholinergic neurons. In GABAergic neurons, however, IL-2 has failed
to evoke release of GABA in mesencephalic neuronal cultures (Alonso et al., 1993), or
the cortex or hippocampus of mice (Bianchi et al., 1995). Since IL-2 deficiency does not
affect the number of GABAergic somata in the MS/vDB of IL-2 KO mice, the neuronal
loss appears to be selective for cholinergic neurons in the MS/vDB. Furthermore, we
previously found no differences in the striatal cholinergic neuronal number (Beck et al.,
2002), so the lack of IL-2 does not simply cause a general loss of all cholinergic neurons.
Against our initial hypothesis that 21-day-old IL-2 KO mice would have similar
cholinergic deficiencies as adult 8-12-week-old mice, there was no detectable loss of
cholinergic cell number in the MS/vDB. We did not examine 21-day-old mice for

41
differences in GABAergic cell number, since we did not detect any changes in adult mice
using the same marker. One potential explanation for the loss of cholinergic neurons in
the MS/vDB may be a failure in maintenance or survival in the late stages of, or after,
development. Other studies have found decreases in cholinergic enzyme activity (i.e.,
ChAT and AChE) between postnatal days 30-60 in normal rats (Thai et al., 1992), and
postnatal days 60-150 in C57BL/6 mice (Virgili et al., 1991). The IL-2 KO mice may
potentially be more susceptible to this loss of cholinergic activity during adulthood,
which could lead to the previously observed deficiencies in 8-12-week-old IL-2 KO
animals.
An alternate explanation for the different cholinergic effects seen in 21-day-old
vs. 8-12-week-old animals is that the loss of IL-2 may be secondary to the effects of
autoimmunity present in adult IL-2 KO animals. Though we cannot completely rule out
this possibility, we have previously failed to find discemable levels of infiltrating
lymphocytes or clear signs of gliosis in the brains of IL-2 KO animals (Petitto et al.,
1999). More research is necessary to further address this issue.
Another finding of this study was a significant decrease in neuronal cell number
in the DP-GCL, but not the SP-GCL. This decrease is consistent with the in vitro studies
showing a potent neurotrophic effect of IL-2 on hippocampal neurons (Awatsuji et al.,
1993; Sardcr et al., 1996; Sarder et al., 1993). Furthermore, these data are supported by
our previous findings that IL-2 KO mice exhibited a reduced distance across the IP-GCL
(Beck et al., 2002), and that the IP mossy fiber length of IL-2 KO mice is shorter than
wild-type controls (Petitto et al., 1999). The reductions in distance across the IP-GCL
could also potentially be explained by increased density, but not number, of cells or

42
smaller cell body size. Qualitative assessments of random granule cells, however, do not
support this hypothesis, though a more extensive study would be necessary to definitively
address that issue.
Though the receptors for IL-2 are more abundant in the hippocampus, including
the GCL of the DG (Petitto and Huang, 1994; Petitto et al., 1998), it is not clear whether
IL-2 may act directly on these neurons, or whether it upregulates other growth factors
like the neurotrophins. The observed differences in the level of the neurotrophin BDNF
was consistent with our hypothesis that we would find a reduction in trophic factors
important in MS/vDB and hippocampal development and maintenance. BDNF plays a
role in the maintenance and repair of septal cholinergic neurons (Alderson et al., 1990;
Morse et al., 1993; Ward and Hagg, 2000), can implement a positive feedback
mechanism with these neurons to enhance the release of ACh (Knipper et al., 1994), and
can also modulate neurogenesis (Larsson et al., 2002; Lee et al., 2002), thus potentially
impacting granule cell number. Thus, the reduction of cholinergic cell number in the
MS/vDB is consistent with a reduction in this trophic factor. The exact interaction
between IL-2 gene deletion and the reduction of BDNF levels remains unclear. Though
BDNF is expressed in the peripheral immune system by lymphocytes, IL-2 does not
stimulate its production or release. IL-2 can, however, upregulate the expression of
TrkB, the receptor for BDNF, in lymphocytes (Besser and Wank, 1999). Furthermore,
some evidence suggests that BDNF can stimulate a positive feedback mechanism of its
own production via the TrkB receptor in hippocampal neurons (Canossa et al., 1997;
Saarelainen et al., 2001). In IL-2 KO mice, the absence of IL-2 may therefore potentially
lead to a down-regulation of the TrkB receptor, thereby partially inhibiting the positive

43
feedback production of BDNF. Interestingly, the neurotrophin Trk receptors and IL-2
receptor share some of the same signal transduction pathways (e.g., mitogen activated
protein kinase or phosphatidylinositol 3-kinase), which appear to play a role in their
growth and survival promoting actions (for reviews see Gaffen, 2001; Patapoutian and
Reichardt, 2001). Whether IL-2 knockout leads to disruption of one of these signal
transduction pathways has not, to our knowledge, been elucidated and thus requires
further study.
Against our initial hypothesis, NGF protein levels were actually increased in the
IL-2 KO mice. Unlike BDNF, NGF does not appear to stimulate a positive feedback
neurotrophin release from hippocampal neurons (Canossa et al., 1997). Given the
reduction in cholinergic survival in the MS/vDB of IL-2 KO mice, the target neurons in
the hippocampus of these animals may produce higher protein levels of NGF as a
compensatory response. Similarly, moderate lesions of rat septohippocampal projections
lead to increased mRNA expression of NGF, but not BDNF in hippocampal target cells
(Hellweg et al., 1997).
In summary, cholinergic deficits seen in the MS/vDB of IL-2 KO mice appear to
be selective for cholinergic over GABAergic neurons. In addition, the loss of cholinergic
neurons in the MS/vDB may occur in the later stages of, or after, development of the
septohippocampal system, as the deficits are not seen in 21-day-old IL-2 KO mice. In the
hippocampus, the number of neurons in the IP-GCL is significantly reduced. A reduced
production of hippocampal BDNF may contribute to many of the aforementioned
changes, though NGF levels are increased in a possible compensatory response.
Although overt signs of autoimmunity in the brain are not apparent (we have been unable

44
to detect significant levels of leukocyte infiltration or gliosis in IL-2 KO mice brains),
further study is necessary to assess this possibility, as factors such as IL-2 induced
cytokine dysregulation or autoantibodies could contribute to the hippocampal alterations
in adult IL-2 KO mice.

CHAPTER 4
INTERLEUKIN-2 DEFICIENCY: NEUROIMMUNOLOGICAL STATUS AND
NEUROGENESIS IN THE HIPPOCAMPUS
Introduction
Receptors for interleukin-2 (IL-2) are enriched in the hippocampal formation, and
many of the most prominent neurobiological functions of this cytokine occur in the
hippocampus (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991;
Petitto and Huang, 1994, 2001; Petitto et al., 1998). Previous studies from our laboratory
have found that IL-2 knockout (KO) mice exhibit significantly lower numbers of medial
septum and vertical limb of the diagonal band of Broca (MS/vDB) cholinergic cell
bodies, a reduction in the distance across the granular cell layer (GCL) of the
infrapyramidal (IP) blade of the dentate gyrus (DG), and decreased fiber length and
neuronal cell number in the IP-GCL of the DG (Beck et al., 2004; Beck et al., 2002;
Petitto et al., 1999). These neurobiological alterations appear to be related to
abnormalities in learning and memory performance and sensory motor gating in IL-2 KO
mice (Cushman et al., 2004; Petitto et al., 1999).
Because IL-2 has been shown to possess various neurotrophic and
neuromodulatory effects on hippocampal neurons in vitro (Awatsuji et al., 1993; Bianchi
et al., 1995; Pauli et al., 1998; Plata-Salaman and ffrench-Mullen, 1993; Sarder et al.,
1996; Sarder et al., 1993; Tancredi et al., 1990), our original working hypothesis was that
the alterations exhibited by IL-2 KO mice are due to the absence of IL-2s neurotrophic
actions on hippocampal neurons during development. More recent data from our
45

46
laboratory, however, suggests that these hippocampal changes may be due to
neurodegenerative rather than neurodevelopmental processes. We tested mice at
postnatal day 21 (P21), an age where septohippocampal cholinergic neurons are nearly
fully developed, to determine if the reduction in septohippocampal cholinergic projection
neurons seen in adult IL-2 KO mice was present earlier in postnatal development (e.g., at
weaning) and prior to the onset of the earliest signs of autoimmune disease (e.g.,
splenomegaly, lymphadenopathy). Contrary to our hypothesis, we found that the number
of MS/vDB cholinergic cell bodies did not differ between IL-2 KO and wild-type
littermates at P21 (Beck et al., 2004). Thus, together these data indicate that the loss of
cholinergic neurons that occurs between P21 and adulthood (8-12 weeks) suggests an
alternate hypothesis; neurodegenerative processes may be operative in the brain of IL-2
KO mice.
Since IL-2 is an important factor in immune physiology, one possible mechanism
behind these neurodegenerative processes is immune dysregulation caused by the absence
of IL-2. IL-2-defciency in mice leads to generalized systemic autoimmune disease in
adult mice that may affect multiple organs in the periphery, most notably the intestines
and the kidneys (Horak, 1995). The autoimmune effects in IL-2 KO mice involving
peripheral organs are mediated largely by infiltrating T cells. In the colon, for example,
adult IL-2 KO mice develop chronic inflammatory bowel disease with features common
to inflammatory ulcerative colitis in humans, where the lamina propria is infiltrated with
activated T cells responsible for the development of this inflammatory disease (Ma et al.,
1995). In addition, there is a disruption of immune homeostasis that is evidenced by
changes in the gene expression of several Thl, Th2, and various proinflammatory

47
cytokines in this organ (Autenrieth et al., 1997; Meijssen et al., 1998). Moreover, these
cytokine changes and the onset of inflammatory bowel disease are preceded by increased
gene expression of IL-15, which shares the same signal transducing receptor subunits
with IL-2 (Meijssen et al., 1998). Thus, it is possible that the immune dysregulation in
the brain of IL-2 KO mice may be induced by activated T cells and/or proinflammatory
cytokines (e.g., IL-1, TNFa, IL-6) from the periphery crossing the blood-brain-barrier
(BBB).
By contrast, IL-2 may lead to neuroimmunological changes that do not involve
peripheral immune cells. Rather than peripheral T cells and serum cytokines entering the
brain, an alternative hypothesis that may account for the hippocampal differences
observed in P21 versus adult IL-2 KO mice may be that the absence of IL-2 reduces the
trophic support of hippocampal neurons as a result of dysregulation of other brain-
derived cytokines. Thus, loss of IL-2 in the brain could in turn modify the normal
neuroimmunological status of the brain by modifying the normal expression of brain
cytokines such as IL-15. Since IL-2 can modify the release of certain cytokines from
lymphoid cells (Lauwerys et al., 2000; McDyer et al., 2002), similar actions could occur
in brain cells that produce cytokines (e.g., microglia, astrocytes). Alterations in the
production of brain cytokines important in normal brain physiology could alter the
integrity of hippocampal neurons by decreasing levels of classic neurotrophins and/or
neurotrophic cytokines on the one hand, or elicit inflammatory-like neurodegenerative
processes within the brain on the other.
The present study therefore sought to test the hypothesis that IL-2 gene deletion
results in neuroimmunological changes in the hippocampus by examining the possible

48
outcomes described above. We compared the hippocampi of adult IL-2 KO mice and
wild-type littermates at 8-12 weeks of age, the age where differences in hippocampal
cytoarchitecture and behavior have been found previously (Beck et al., 2004; Beck et al.,
2002; Cushman et al., 2004; Petitto et al., 1999), for differences in several measures of
neuroimmunological status. First, the groups were assessed for differences in the number
of CD3+ T lymphocytes and activated microglial cells (as measured by MHC-II
positivity) in the hippocampus. IL-15 is also expressed in the brain (Hanisch et al.,
1997a; Lee et al., 1996), and this cytokine is known to have both proinflammatory and
anti-inflammatory effects, potent anti-apoptotic, and T cell chemoattractant properties
(Wilkinson and Liew, 1995). Because IL-15 uses the IL-2/15RP and yc subunits that are
enriched in the neuronal cell layers of the hippocampus (Petitto and Huang, 2001), and
may modulate microglial cell function and T cell chemoattraction, a second aim of this
study was to test the hypothesis that IL-15 is elevated in the hippocampus of IL-2 KO
mice. In addition, since changes in both IL-2 and IL-15 may modify levels of various
cytokines in other tissues and physiological contexts, exploratory testing was performed
to determine if IL-2 KO mice have increased levels of proinflammatory cytokines in the
hippocampus relative to wild-type mice (also, compared to serum levels). Furthermore,
as recent evidence indicates that elevation of inflammatory cytokines such as IL-6 may
impair hippocampal neurogenesis (Monje el al., 2003; Vallieres et al., 2002), a third aim
of this study was to test the hypothesis that the postulated changes in
neuroimmunological status would be associated with reductions in neurogenesis of
neurons in the dentate gyrus (DG) of IL-2 KO mice.

49
Materials and Methods
Animals and Genotyping
Mice used in these experiments were cared for in accordance with the NIH Guide
for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2
heterozygote by heterozygote crosses. The IL-2 KO mice, obtained originally from the
NIH repository at Jackson Laboratories, were derived from ten generations of
backcrossing onto the C57BL/6 background. Mice were housed under specific pathogen-
free conditions. Animals used in these experiments were 8-12 weeks of age.
Independent animals were used for the assessment of CD3^ T cells and MHC II+
microglial cells in the hippocampus, the determinations of hippocampal versus serum
cytokine levels, and assessments of neurogenesis in the dentate gyrus. Specific numbers
utilized are reported at the beginning of the description of each method.
The genotypes of all mice were determined by the polymerase chain reaction
(PCR). PCR reactions were performed using 25 pi total reaction volume containing 1
pM each of forward and reverse primers, 0.1 pg genomic DNA, 0.2 mM of each dNTP,
0.3 pi Taq DNA polymerase, and amplified using a thermal cycler with a heated
evaporation cover (Ericomp). The cycling parameters were hot start 95C (3min),
denaturing 94C (30 sec), annealing 64C (30 sec), extension 72C (45 sec) with a final
extension step of 4 min. Thirty cycles were used for these experiments. The 5 and 3
primers for the IL-2 KO (500 bp knockout band amplified) were 5-
TCGAATCGCCAATGACAAGACGCT-3 and 5-
GTAGGTGGAAATTCTAGCATCATCC-3. The 5 and 3primers for the IL-2 wild
type (324 bp wild type band amplified) were 5-

50
CTAGGCCACAGAATTGAAAGATCT-3 and 5-
GTAGGTGGAAAATTCTAGCATC ATCC-3 .
CD3+ T cells and MHC II+ Microglia Immunohistochemistry
For quantitative assessment of autoimmunity, three IL-2 KO and three wild-type
brains were processed for MHC II (an activated microglial marker) and CD3 (a pan T cell
marker) immunohistochemistry. Each animal was anesthetized with an injection cocktail
of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose
of 0.015 ml injection cocktail/g body weight and perfused with 0.9% saline followed by
4% paraformaldehyde in phosphate buffered saline (PBS). The brains and spleens were
removed and fixed overnight in 4% paraformaldehyde, followed by overnight
equilibration in 30% sucrose cryoprotective solution, and then were snap frozen in
isopentane (-80C) for storage. The spleens were weighed to assay for relative
splenomegaly of IL-2 KO vs. wild-type mice The brains were equilibrated to -20C
prior to cryostat sectioning into 50 pm slices in the coronal plane, collected into
individual wells of polystyrene 24-well plates (NUNC 1147), and stored free-floating at
4C in PBS for histochemistry. Every third section was processed for MHC II or CD3
immunohistochemistry.
Free-floating 50-pm sections were incubated for 20 minutes in 1% hydrogen
peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then
washed and blocked for 1 hr in 200 pl/well 3% normal goat serum (NGS). After this
incubation, the sections were incubated overnight in the primary antibody, rat anti mouse
CD3 (BD PharMingen; 1:500 in PBS with 0.3% Triton X-100 and 1% NGS) or rat anti
mouse MHC II (BD PharMingen; 1:500 in PBS with 0.3% Triton X-100 and 1% NGS).
The next day, the sections were washed and incubated overnight in the secondary

51
antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with
0.3% TX-100 and 1% NGS). The sections were then washed and incubated in
ExtrAvidin (Sigma E-2886; 1:1000 in PBS) for 2 hrs. The sections were developed in
0.5 mg/ml 3,3-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min
and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes
of xylenes, and coverslipped.
Preparation of Serum and Brain Tissue for Cytokine Analysis
Hippocampal homogenates were analyzed from eight IL-2 KO and nine wild-type
mice to measure cytokine levels in the hippocampus. From these subject groups, serum
was collected from a smaller subset of animals (five IL-2 KO and seven wild-type mice)
for comparative analysis of brain vs. peripheral cytokine levels. Animals used for
characterization of endogenous cytokine levels were anesthetized with an injection
cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml)
at a dose of 0.015 ml injection cocktail/g body weight. Whole blood was collected by
puncturing the right atrium of the heart and inserting heparanized micro-hematocrit
capillary tubes (Fisher Scientific). The animals were then saline perfused, but were not
post-fixed in paraformaldehyde. The whole blood was centrifuged in Microtainer Brand
serum separator tubes (Becton Dickinson) at 5,000 rpm for 10 minutes to isolate serum
and the serum was stored at -80 C until used for Luminex analysis. The brains were
removed, snap frozen, and then allowed to equilibrate to -20 C. The brains were
sectioned on a cryostat at -20-22 C at 400 pm thickness and the hippocampi were
dissected with a 0.75 mm micropunch on a -20 C freezing platform. The dissected
tissue was weighed on a microgram scale, and then transferred to 25 pi of homogenizing

52
solution (50 mM Na/Na2 and 0.2% TX-100 in H2O with Anti-protease Complete TM
cocktail (Boehringer)) per mg of wet weight tissue. The tissue was sonicated in the
homogenizing solution for 30 sec on ice and centrifuged at 16,000 g for 15 min at 4 C.
The supernatant was collected and stored at -20 C for Luminex analysis.
Multiplex Microsphere Cytokine Analysis
Commercial kits, Lincoplex mouse cytokine (Lineo, Research, Inc) and a
Luminex 100 LabMAP system (Upstate Biotechnology), were used in attempt to measure
a number of cytokines in the hippocampus and in the serum of IL-2 KO and wild-type
mice. Assays were performed according to the manufacturers instructions, and cytokine
concentrations were calculated using the Softmax program and the linear range on the
standard curve (3.2-10,000 pg/ml). Altogether, we attempted to detect a total of twenty-
two different cytokines and chemokines from the serum and brain homogenates of these
animals. In the serum, there were detectable levels of IL-6. IL-13, kerotinocyte-derived
chemokines (KC), granulocyte-colony stimulating factor (G-CSF), and macrophage
inflammatory protein-1 alpha (MIP-la). In the brain, there were detectable levels of IL-
7, IL-9, IL-12, IL-15, interferon-gamma inducible protein of 10 kD (IP-10), and
monocyte chemoattractant protein-1 (MCP-1). Thus, only those cytokines and
chemokines detected were subjected to statistical analyses. The remainder of the
cytokines and chemokines tested could not be detected in either the serum or the brain
(IFN-y; TNF-oc; IL-la, IL-lf); IL-2, IL-4, IL-5, IL-10, IL-17, GM-CSF, and RANTES).
Labeling Neurogenesis with BrdU
Twelve IL-2 KO (six male; six female) and eleven wild-type (five male; six
female) mice were used to assay for differences in neurogenesis. The procedure for

53
labeling of neurogenesis and subsequent immunostaining in the mouse hippocampus has
been adapted from (Lee et al., 2002). Briefly, the mice were given five intraperitoneal
injections of BrdU (50 mg/kg of body weight) over the course of 3 days. The day
following the last injection, the mice were sacrificed and perfused with 0.9% saline
followed by 4% paraformaldehyde in PBS as described previously.
The BrdU-incorporated brains were equilibrated to -20 C and cryostat-sectioned
at 50 pm in the coronal plane. They were collected into individual wells of polystyrene
24-well plates (NUNC 1147), and used for free-floating immunohistochemistry. The
sections were then washed twice in PBS and then the DNA was denatured by a 30 minute
incubation with 2 N HC1 to allow binding of the antibody to the BrdU in the single-
stranded DNA. The acid was neutralized with a 0.1 M borate buffer (pH 8.5) wash,
followed by several washes in PBS. Afterwards, the sections were blocked for 1 hr in 3%
normal goat serum (NGS). The sections were then incubated overnight in the primary
antibodies, rat monoclonal anti-BrdU (Serotec; 1:400 in PBS with 0.3% Triton X-100 and
1% NGS) and either the neuronal marker mouse monoclonal anti-tubulin p III isoform
(Chemicon; 1:200 in PBS with 0.3% Triton X-100 and 1% NGS), the astroglial marker
rabbit anti-glial fibrillary acidic protein (GFAP; Chemicon; 1:1,000 in PBS with 0.3%
Triton X-100 and 1% NGS) or the oligodendrocyte marker mouse anti-23-cyclic
nucleotide 3-phosphohydrolase (CNPase; Chemicon; 1:200 in PBS with 0.3% Triton X-
100 and 1% NGS). The next day, the sections were washed twice in PBS and incubated
for 2 hr in the dark with the secondary antibodies, goat anti-rat IgG (H+L) conjugated
with Alexa Fluor-488 (green; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1%
NGS) and goat anti-mouse IgG (highly cross-absorbed H+L) conjugated with Alexa

54
Fluor 568 (red; Molecular Probes; 1:400 in PBS with 0.3% TX-100 and 1% NGS) or goat
anti-rabbit IgG (H+L) conjugated with Alexa Fluor 350 (blue; Molecular Probes; 1:400
in PBS with 0.3% TX-100 and 1% NGS). The sections were then washed twice in PBS,
placed on slides, dehydrated in graded ethanol washes, cleared in two changes of xylenes,
and coverslipped.
The sections were imaged using a Bio-Rad 1024 ES confocal microscope and
only cells which showed colocalized staining through five consecutive 1-pm planes were
considered to be double-labeled. The IP and SP-GCL area (mm2) were measured at 20x
magnification using the MCID 5.1 software, a CCD High Resolution Sony camera, and a
Zeiss Axioplan 2 microscope. The data were reported as a density of all double-labeled
cells counted from five sections per animal divided by the total area measured
Results
Assessment of CD3+ T Cells and MHC II+ Activated Microglia in the Hippocampus
No CD3+ T cells were detected in the hippocampi of either 8-12-week-old IL-2
KO or wild-type mouse brains, and only an occasional MHC II+ microglial cell was
detected (e.g., approximately one every other section) in both groups. By contrast, both
activated MHC IF microglia and CD3" T cells were readily detectable in positive control
slices (sections of the axotomized FMN of wild-type C57BL/6 mice; Petitto et al., 2003)
demonstrating that the immunohistochemistry procedure utilized was effective for
labeling both markers. Thus, no differences in T cells or activated microglial cells were
found between the subject groups.
Hippocampal Cytokine Levels in IL-2 Knockout vs. Wild-type Mice
As depicted in Figure 4-1, in the hippocampus, there were significantly increased
levels of IL-12 (increased -57%; F(l,15)=9.174, p=0.008) and IL-15 (increased -38%;

55
F(l,15)=6.105, p=0.026) in the IL-2 KO compared to wild-type mice. Figure 4-1 also
shows that IL-2 KO mice had increased levels of the chemokines, IP-10 (increased
-63%; F(l,15)=4.747, p=0.046) and MCP-1 (increased -46%; F(l,15)=5.218, p=0.039).
Although detectable levels of IL-7 and IL-9 were found in hippocampus, they did not
differ between the subject groups.
Enhanced Cytokine Concentrations in the Hippocampus
CD
Z3
I
a>
S
oo
5
o>
E
ai
c
O
S'
o
O)
Q.
12
1 0
0.8
06
04
02
0.0
*
IL-12 IL-15 IP-10 MCP-1
Cytokine
Figure 4-1. Protein levels of several cytokines and chemokines are elevated in the IL-2
knockout mice compared to wild-type brain. These include IL-12, IL-15, IP-10,
and MCP-1. *p<0.05; **p<0.01.
Comparison of Serum Cytokine Levels in IL-2 Knockout vs. Wild-type Mice
As can been seen in Figure 4-2, the IL-2 KO mice exhibited marked elevation in
serum IL-6 (increased -224%; F(l,10)=8.077, p=0.017), and serum MIP-la serum
concentration (increased -327%; F(l,10)=21.538, p=0.001) compared to wild-types.
Although detectable levels of IL-13, G-CSF, and KC were detectable in the serum, levels
of these cytokines did not differ between the subject groups.

56
A) B)
Enhanced Cytokine Concentrations In the Serum of Enhanced Cytokine Concentrations In the Serum of
IL-2 KO Animals IL-2 KO Animals
MIP-1 alpha
1 1 Wild-Type
IL-6
1 1 Wild-Tyoe
wau IL-2 knockout
1 IL-2 knockout
Figure 4-2. Relative cytokine profile of IL-2 knockout mouse serum does not match brain
profile. A) IL-6 is increased in IL-2 knockout mice compared to wild-type; and
B) the chemokine MIP-1 a is higher in IL-2 knockout mice than wild-type.
*p<0.05; **p<0.01.
Alterations in Neurogenesis
Although there was not a significant effect of group on neurogenesis in either the
IP-GCL or SP-GCL, as depicted in Figure 4-3 and 4-4 respectively, there was a
significant group by gender interaction in both the IP-GCL (F(l,19)=4.71, p=0.043) and
SP-GCL (F(l,19)=6.43, p=0.02). Fishers least significant difference post hoc analysis
test confirmed a difference between male IL-2 KO and wild-type mice in the IP-GCL
(p=0.025) and SP-GCL (p=0.014), but not between female groups in either region. There
was no significant effect of group or group by gender interaction in cells around either
the IP-GCL or SP-GCL labeled with the oligodendrocyte marker, CNPase (data not
shown). Similarly, no significant effect of group or group by gender interaction was
evident in either the IP or SP-GCL in cells labeled with the astrocyte marker, GFAP (data
not shown).

57
Neurogenesis in the Infrapyramidal Granule Cell Layer
500
4X1
IL-2 Knockout Wild-Type
i i Female
I Male
Figure 4-3. There is no effect of group on neurogenesis in the infrapyramidal granule cell
layer, but there is a group by gender interaction between IL-2 knockout and wild-
type mice (p=0.043). This effect appears to be attributable to differences in the
male mice (Fisher least significant difference test, p=0.025).
Neurogenesis in the Suprapyramidal Granule Cell Layer
500
4X -
1 1 '
Wild-Type IL-2 Knockout
1 I Female
I Male
Figure 4-4. There is no effect of group on neurogenesis in the suprapyramidal granule
cell layer, but there is a group by gender interaction between IL-2 knockout and
wild-type mice (p=0.02). This effect appears to be attributable to differences in
the male mice (Fisher least significant difference test, p=0.014).

58
Discussion
The data presented here show that IL-2 KO and wild-type littermates exhibit
differences in several measures of neuroimmunological status in the hippocampus. In
order to access the brain parenchyma, T cells require activation markers to cross the BBB
(Hickey et al., 1991). We have previously reported leukocytes were not detectable in the
cresyl violet stained hippocampal sections from IL-2 KO mice (Petitto et al., 1999),
however, we recognized that it is difficult to reliably detect small numbers of peripheral
leukocytes in the brain without cell-type specific stains. Although we were therefore not
expecting to see substantial numbers of T cells in the IL-2 KO brain, we wanted to
determine if small numbers of autoimmune T lymphocytes were present that could
initiate neuroimmunological alterations in the hippocampus. The hippocampi of IL-2 KO
mice were devoid of T cells, despite the fact that the majority of peripheral T cells of IL-2
KO mice express activation markers such as CD69 (Sakai et al., 1995; Schopke et al.,
1991), which are thought to enhance their ability to cross the BBB. Microglia are
indigenous antigen presenting cells (Hickey and Kimura, 1988; Streit et al., 1988).
Contact with T cells can induce microglia to exhibit characteristics of antigen presenting
cells, and microglia also have the ability to activate T cells (Aloisi et al., 2000). There
was, however, no evidence of increased numbers of activated microglia in the
hippocampus of IL-2 KO mice. This observation is consistent with our previous finding
in C57BL/65C/W-IL-2 KO (mice without mature T and B cells), which were devoid of T
cells in the axotomized facial motor nucleus and had levels of axotomy-induced activated
microglia that did not differ from wild-type mice (Petitto et al., 2003). Thus, at the
cellular level, the hippocampus of IL-2 KO mice did not show signs of autoimmune
disease.

59
Cytokines can enter brain via specific transport mechanisms and through the
circumventricular organs. In this study, we measured levels of the various cytokines of
interest in the serum to determine if levels found in the hippocampus correlated with
those found in the serum. If a particular inflammatory cytokine (e.g., TNFa) had been
found to be significantly elevated in the serum and the hippocampus of IL-2 KO mice, it
would have suggested the possibility that peripheral immune activation associated with
autoimmunity (Schimpl et al., 2002) could account for the presence of the cytokine in the
hippocampus (although increased gene transcription and translation in both the periphery
and the brain could not be ruled out). Although there were significantly increased levels
of IL-6 and MIPla in the circulation of IL-2 KO mice, we did not detect either of these
proteins in the brain. IL-6 is a proinflammatory cytokine that was of particular interest to
us because of its actions in the hippocampus, including effects on neurogenesis (Monje et
al., 2003; Vallieres et al., 2002). Although levels of IL-6 were markedly increased in the
serum, though not measurable in the hippocampus at the tissue homogenate
concentrations used, the unlikely possibility remains that IL-6 could have entered the
brain from the circulation and had functional consequences at concentrations below the
limits of detection of the assay method. Nonetheless, the most parsimonious explanation
is that cytokines from the peripheral circulation of IL-2 KO mice are not the source of the
cytokine alterations found in the hippocampus. Thus, peripheral cytokine dysregulation
associated with autoimmunity in IL-2 KO does not appear to be associated with their
hippocampal pathology.
Our data indicate that loss of IL-2 in the brain results in changes in the production
of several other brain cytokines. Consistent with our hypothesis, IL-15 concentrations

60
were increased in the hippocampus of IL-2 KO mice. IL-15 is structurally related to IL-2
and shares the same p and yc signal transducing receptor subunits with the IL-2 receptor
(Giri et al., 1995). IL-15 also shares and opposes several physiological functions of IL-2
in the peripheral immune system (Waldmann, 2002; Waldmann et al., 2001). IL-15 and
its heterotrimeric receptor are constitutively expressed in various regions of the adult
mouse brain and can be detected in microglial cultures (Hanisch et al., 1997a), astrocytes
(Lee et al., 1996), and possibly neurons (Maslinska, 2001; Satoh et al., 1998). As noted
earlier, increased IL-15 gene expression precedes the inflammatory cytokine changes and
onset of inflammatory bowel disease in IL-2 KO mice (Meijssen et al., 1998). It also
induces the onset of autoimmunity in thyroiditis (Kaiser et al., 2002). Thus, IL-15 could
trigger proinflammatory cytokine-like processes in the hippocampus, including the
elevations in IL-12 that were found in IL-2 KO mice in this study. IL-12-driven Thl
responses are involved in inflammation (e.g., colonic) in IL-2 KO mice (Ludviksson et
al., 1997), and it has been implicated as an important effector in the pathogenesis of
experimental autoimmune encephalomyelitis (EAE) (Adorini, 1999; Segal et al., 1998).
Moreover, IL-15 can render cells resistant to the protective effects of TGFP (Campbell et
al., 2001), a Th2 cytokine that appears to play a key role in dampening processes
associated with peripheral autoimmune disease in IL-2 KO mice (Ludviksson et al.,
1997). Thus, these actions of IL-15 suggest that it may be involved in the hippocampal
pathology seen in IL-2 KO mice. It is noteworthy, however, that IL-15 has potent anti-
apoptotic properties (Lauwerys et al., 2000; Waldmann, 2002; Waldmann et al., 2001)
that may oppose the pro-apoptotic effects of IL-2. We have recently found that loss of
brain IL-2 in C57BL/6sc/d-IL-2 KOmice increased neuroregeneration in the axotomized

61
facial motor nucleus (Petitto et al., 2003). It is interesting to speculate that elevated IL-15
levels could contribute to the increased motor neuronal survival in those mice. Therefore
an alternative interpretation is that increased IL-15 in the hippocampus of IL-2 KO mice
could be a compensatory response to counteract neuroregenerative changes in the
hippocampus.
The increased levels of MCP-1 and IP-10 may possibly be induced by increased
IL-15 in the hippocampus (Badolato et al., 1997). Previous studies have demonstrated
that IP-10 expression can be detected in lipopolysaccharide (LPS)-treated microglial and
astroglial cultures and in situ hybridization of LPS-treated rat brains (Ren et al., 1998).
Similarly, MCP-1 can also be induced by the addition of the pro-inflammatory cytokine,
TNF-a, or the anti-inflammatory cytokine, TGF-P, in astrocytes (Hurwitz et al., 1995)
and microglia (Meda et al., 1996). IP-10 and MCP-1 are also chemoattractant factors for
T lymphocyte infiltration into the CNS (Babcock et al., 2003; Dufour et al., 2002), and
may attract activated microglia in vitro (Cross and Woodroofe, 1999). In spite of the
published data linking IP-10 and MCP-1 to T cell and microglial chemotaxis, we were
unable to detect either T cells or increased numbers of activated microglia in the
hippocampus of IL-2 KO mice. Further studies are necessary to determine the functional
significance of the increased production of IP-10 and/or MCP-1 in the hippocampus of
IL-2 KO mice. Finally, in addition to its immune activating effects as a Thl cytokine,
IL-2 is also known to have important critical negative regulatory functions by stimulating
Th2 lymphocytes to produce TGFp (Ludviksson et al., 1997), which down-regulates the
ability of antigen presenting cells to produce IL-12, a powerful activator of Thl cell

62
development and inflammation. Thus, the increased levels of IL-12 found in the
hippocampus could be the secondary to the loss IL-2.
Increased levels of inflammatory cytokines such as IL-6 may impair hippocampal
neurogenesis (Monje et al., 2003; Vallieres et al., 2002); thus, we hypothesized that the
alterations in neuroimmunological status would correlate with decreased neurogenesis in
the DG of IL-2 KO mice. Against our hypothesis, however, no detectable levels pro-
inflammatory cytokines like IL-6 (though IL-12 can mediate inflammatory responses) or
effects of group were apparent in hippocampal neurogenesis. However, a significant
group by gender interaction was detectable. Interestingly, this effect appeared to be due
to variations in the male mice, but not the females. We previously noted a trend which
suggested that IL-2 gene deletion may protect from experimental autoimmune
encephalomyelitis (EAE), though we did not statistically analyze the data (Petitto et al.,
2000). In that study, three out of the seven male mice utilized developed some symptoms
of the disease, whereas none of the six females did. One hypothesis to explain this group
by gender interaction is that there may be an interplay between IL-2 and the sex
hormones. Sex steroids like estrogen, for example, have been linked to neurogenesis
(Gould et al., 2000; Perez-Martin et al., 2003). Further, evidence that IL-2 can regulate
sex hormone expression comes from studies of Leydig cell cultures where IL-2 was
shown to inhibit steroidogenesis (Guo et al., 1990).
Hormonal differences influenced by IL-2 may give a possible explanation of why
males differ from females, but it does not suggest a mechanism of why IL-2 KO mice
appear to be protected from impaired hippocampal neurogenesis. As noted above,
absence of IL-2 can enhance neuroregenerative properties in the axotomized facial motor

63
nucleus and we speculate an involvement of IL-15. Though IL-15 has not been linked to
neurogenesis, its antiapoptotic properties on immune cells are well-studied (Lauwerys et
al., 2000; Waldmann, 2002; Waldmann et al., 2001). If these antiapoptotic actions of IL-
15 can also promote survival of neurons, then this may provide a hypothesis of why the
IL-2 KO male mice have higher levels of neurogenesis than the wild-types. Some
cytokines are capable of promoting the development of neural stem cells (Rozental et al.,
1995; Shah et al., 1996; Wong et al., 2004), and this may account for the how they
influence neurogenesis. The actual mechanism whereby IL-2, sex hormones, and
neurogenesis may interact is as of yet undefined and requires future study.
In summary, T cells and peripheral cytokines do not appear to enter into the
hippocampi of IL-2 KO mice, which does not support the hypothesis that the CNS
alterations previously seen in IL-2 KO mice are due to peripheral autoimmunity. Other
potential immune indicators of autoimmunity (e.g., deposition of autoantibodies in the
brain) were not addressed in this dissertation and require more research. Genetic deletion
of IL-2 may, however, alter the neuroimmunological status of the mouse hippocampus
through a dysregulation of cytokines produced by CNS cells (e.g., microglia, astroglia).
Further studies will be required to determine how these changes impact hippocampal
cytoarchitecture and function in IL-2 KO mice.

CHAPTER 5
GENERAL DISCUSSION
Summary of the Overall Findings
The overall purpose of this dissertation research was to investigate the impact of
IL-2 on the septohippocampal system by observing changes in the basal physiology and
structure of IL-2 KO mice brains vs. their wild-type littermates. During these
experiments, we chose not to surgically, chemically, or in any other way experimentally
manipulate these animals beyond the genetic knockout of IL-2, so that the effects that we
detected could be attributed to loss of IL-2 and not to confounding experimental
techniques (e.g., disruption of the BBB). We understand, however, that some
experimental manipulations could be useful in the long term and discuss the topic more in
the Caveats and Future Directions section of towards the end of this chapter.
In the first study in Chapter 2, we determined that IL-2 KO mice suffered a
significant 26% loss of cholinergic neurons in the MS/vDB relative to wild-type
littermates. This loss of cholinergic somata was not reflected by a similar loss of
cholinergic fiber density in the hippocampal projection fields of the septal cholinergic
neurons. Moreover, the deficits observed were not a general effect on all cholinergic
neurons of the brain, as the cholinergic neurons of the striatum in the IL-2 KO animals
studied did not appear to be affected. There was, however, a reduction in the distance
across the IP, but not SP, GCL of the dentate gyrus.
In Chapter 3, we extended upon these findings by showing that the loss of
cholinergic neurons appeared to occur later in development or during adulthood as a
64

65
potential failure of these neurons to survive, since 21-day-old IL-2 KO animals did not
show signs of a loss of cholinergic neurons of the MS/vDB. Furthermore, the neuronal
deficiencies appeared to be selective to the cholinergic neurons of the MS/vDB, as there
were no similar decreases of GABAergic neurons in IL-2 KO animals. In the
hippocampus, consistent with the reductions in length across the IP-GCL of IL-2 KO
compared to wild-type mice, the stereologically estimated cell count of Nissl-stained
neuronal somata of IL-2 KO mice was significantly lower than wild-type littermates.
Also, the hippocampal neurotrophin levels of BDNF and NGF were significantly
decreased and increased, respectively.
Finally, in Chapter 4, the cytokine profile of the IL-2 KO hippocampus was altered
with increases in IL-15 and IL-12 and the chemokines MCP-1 and IP-10. These
experiments confirmed our hypothesis that loss of IL-2 would result in increased levels of
brain IL-15 production. Furthermore, in spite of the known roles of these cytokines and
chemokines in T cell trafficking and modulation during brain insult, we were unable to
label for elevated levels of the T cells or activated microglia in the hippocampus. The
above cytokine profile did not match the serum cytokine levels, suggesting that the
changes in cytokines in the hippocampus were likely due to changes in their production
in the CNS. Finally, though there was no group alteration in adult GCL neurogenesis,
there was a group by gender interaction that appeared to be attributable to the male mice.
Implications
This series of studies is the first to demonstrate that endogenous levels of IL-2 may
be an important factor in the late development or survival of neurons in the CNS. This
impact on CNS neurons involved not only cell number, but also alterations in levels of
trophic factors and brain cytokines. Other studies involving IL-2 and the CNS, to date,

66
have depended on in vitro models and/or exogenous administration of IL-2 into theCNS
in vivo. Whereas studies such as these can yield valuable information, in vitro models
often fail to mimic the complexity of an in vivo system (e.g., lack of glia, different
organization of neurons, etc.) and the manipulated in vivo models are subject to
experimental design complications (e.g., disruption of the BBB, determination of
physiologically meaningful doses of exogenously administered cytokines, etc.). We
would be remiss, however, to claim that the IL-2 KO model was completely without
complications itself (e.g., autoimmunity such as inflammatory bowel disease), but we feel
that the most informative way to address a hypothesis is to examine it in multiple
different ways. Thus, taken together with the previous studies in the literature, this
dissertation reveals an interesting relationship between IL-2 and the septohippocampal
system.
The goal of this study was to elicit whether IL-2 KO impacted the
septohippocampal system, but not to design a treatment or study a clinical disorder.
Thus, at its core, this dissertation was basic research. Whereas the pursuit of knowledge
is a noble goal, invariably, a simple question often finds itself in the front of our minds:
how is this new knowledge clinically relevant and how can we use it?
Already, IL-2 is being used clinically to treat, or being studied as a treatment for, a
number of disorders, including cancer (Atkins et al., 1999; Davis and Gillies, 2003; Fyfe
et al., 1995; Guirguis et al., 2002; Parkinson et al., 1990; Rosenberg et al., 1985;
Rosenberg et al., 1994), and HIV infection (Armstrong and Kazanjian, 2001; Mitsuyasu,
2001; Smith, 2001). The aforementioned treatments are not, however, completely
without side effects. In a landmark study, systemic IL-2 treatment for cancer patients

67
was shown to induce dose and time-related behavioral and cognitive changes in 22 out of
44 subjects, including spatial and temporal disorientation (Denicoff et al., 1987). Others
have also noted cognitive changes in patients undergoing IL-2-treatment (Capuron et ah,
2000; Capuron et ah, 2001b; Walker et ah, 1997).
Of particular relevance to this dissertation, IL-2 immunotherapy can induce
cognitive changes, including spatial memory deficits and poor task planning (which may
be partially dependent on spatial memory) as early as five days into therapy (Capuron et
ah, 2001a). Capurons study utilized neuropsychological batteries to assess these
cognitive dysfunctions, but did not investigate the physiological mechanism underlying
them. One hypothesis to explain at least part of the cognitive changes observed is that
the chronic IL-2 treatment used on these patients may affect the septohippocampal
system in a similar way that loss of IL-2 causes changes in IL-2 KO animals. Deletion of
IL-2 having a similar physiological effect as administration of exogenous IL-2 may seem
counterintuitive; however, in at least one population of neurons (i.e., the septal neurons),
IL-2 has potent biphasic modulatory actions on ACh release with low sub-pM IL-2
concentrations enhancing and higher nM concentrations inhibiting release (Hanisch et al.,
1993; Seto et al., 1997). Thus, under the proposed hypothesis, a loss of IL-2 could lead
to a reduction in the stimulatory action on cholinergic neurons, whereas an
overabundance of exogenous IL-2 could inhibit release of ACh. Of course, more
investigation would be necessary to clarify that hypothesis.
Caveats and Future Directions
In Study 1, we examined the effects of IL-2 KO on the cholinergic
septohippocampal neurons. Many of our findings were extended upon in Study 2.
However, we hypothesized that the lack of any change in the cholinergic fiber density in

68
the hippocampus was likely due to compensatory sprouting, but did not overtly assay for
any alterations. Thus, to address this issue further, a study utilizing a stereotaxic
injection of an anterograde tracer, such as Phaseolus vulgaris-leucoagglutinin (PHAL),
into the MS/vDB area would be necessary. The septohippocampal cholinergic neurites
would need to be labeled with a marker for AChE and the length and branching of these
double-labeled axons would be characterized. If the septohippocampal cholinergic
neurons do undergo compensatory sprouting in IL-2 KO mice, we hypothesize that a
detectable increase in branching should occur in those animals vs. wild-type littermates.
Though the majority of cholinergic neurons in the MS/vDB project to the
hippocampus (Schwegler et al., 1996b), a small percentage also project to other areas,
including the mediodorsal nucleus of the thalamus (Gritti et al., 1998), parietal, cingulate
(Gritti et al., 1997), and entorhinal cortices (Alonso and Kohler, 1984). Thus, to
accurately claim that the cholinergic neuronal losses observed in the MS/vDB of the IL-2
KO mouse do indeed project to the hippocampus, a study involving a retrograde tracer
like Fluoro-Gold injected in the hippocampal projection areas would be necessary. The
retrogradely labeled somata of the MS/vDB would need to be double-labeled with ChAT
to identify neurons of the cholinergic phenotype and the cell number would be estimated
with stereology.
In Study 2, three-week-old IL-2 KO animals did not differ from wild-type
littermates in cholinergic cell number. To expand upon this finding, the brains from
animals should be examined at one-week intervals from three-week-old mice to adult
eight-week-old mice. This will allow us to establish a time course of cholinergic
neuronal loss. Furthermore, we would be interested in investigating the nature of the cell

69
loss. A DNA fragmentation assay, like terminal transferase-mediated dUTP nick-end
labeling (TUNEL), could be used to detect apoptotic cell death and cell death by necrosis
could be determined by a simple trypan blue dye exclusion method. Both assays would
require immunohistochemical labeling with ChAT to identify the cholinergic somata.
In Study 3, the cytokine and chemokine profile of the IL-2 KO hippocampus was
altered compared to wild-type mice. Though we did speculate on the cell types that
might produce these cytokines, we did not investigate it further. To identify the potential
sources of these cytokines, we could use in situ hybridization to label the mRNA and
label the individual cell types with immunohistochemistry. We could expand upon this
further by utilizing real time PCR techniques as a semi-quantitative assay of different
levels of mRNA expression in homogenates from the hippocampi of IL-2 KO and wild-
type mice. These two experiments allow us to identify the cells producing the cytokines
and compare the relative levels of mRNA produced for each cytokine between IL-2 KO
and wild-type mice.
Several other considerations and questions arise from the data generated from the
experiments of this dissertation. First, there is possibility that the reduced levels of
ChAT* immunostaining may be due to a loss of the cholinergic phenotype (e.g., inability
to produce ACh) rather than cell death. Alterations in the activity of ACh in the
hippocampus (e.g., addition of receptor agonists or antagonists) are sufficient to alter
learning and memory (for a review see Gold, 2003). Thus, the behavioral deficits
observed in the IL-2 KO animals may not be due to cholinergic neuronal loss. Although
the staining intensity was uniform (e.g., consistent between and within groups), to test
this further, IL-2 KO and wild-type littermate brains should be Nissl-stained to label and

70
quantify MS/vDB neurons in this brain region. Animals with neuronal loss, should
exhibit reduction in total neuronal counts, whereas the neuronal counts would not differ
significantly in mice with a loss of phenotype.
Also, considering the difficulty of detection of endogenous IL-2 in the normal
CNS, what is the source of the IL-2 in the wild-type brain? Taken together with the
unknown non-saturable transport mechanism that allows IL-2 to cross the BBB
(Waguespack et al., 1994), and the numerous studies showing cognitive effects of
peripherally administered exogenous IL-2 (Capuron et al., 2001a; Denicoff et al., 1987;
Lacosta et al., 1999; Walker et al., 1997), an argument can be made that some of the
endogenous IL-2 in the CNS could be from the periphery, particularly during
development when the BBB is not completely formed. In addition, though we did
address the autoimmunity issue of IL-2 KO mice somewhat, we did not completely rule it
out as a potential factor in the observed alterations. An experiment that could address
both of the above issues would involve crossbreeding the IL-2 KO mice onto an
immunodeficient background lacking functional lymphocytes (e.g., RAG-1 knockout
mice). Next, an adoptive transfer of normal T lymphocytes from healthy IL-2 wild-type
(i.e., non-RAG-1 KO) mice to young (i.e., less than 3-week-old) IL-2 KO/RAG-1 KO or
IL-2 wild-type/RAG-1 KO littermate animals would establish a functional immune
system, and at the same time restore a major source of peripheral IL-2. Also, another set
of IL-2 KO/RAG-1 KO and IL-2 wild-type/RAG-1 KO mice should receive an adoptive
transfer of IL-2 KO (i.e., non-RAG-1 KO) lymphocytes to mimic the autoimmune state
of normal IL-2 KO mice. Finally, another group of IL-2 KO/RAG-1 KO and IL-2 wild-
type/RAG-1 KO mice should receive a sham reconstitution, such that they maintain any

71
normal levels (or lack thereof in IL-2 KO/RAG-1 KO mice) of endogenous brain IL-2
expression, but remain immunodeficient, thus lacking a peripheral source of IL-2. This
experimental design would create a model
that eliminates the impact of peripheral autoimmunity in the IL-2 wild-type
reconstituted IL-2 KO/RAG-1 KO mice;
in which the aforementioned animals are only lacking a CNS, but not peripheral,
source of IL-2;
in which the IL-2 wild-type reconstituted IL-2 wild-type/RAG-1 KO mice would
have a functional peripheral immune system and normal endogenous IL-2
expression in the brain;
in which the IL-2 KO reconstituted IL-2 KO/RAG-1 KO mice are similar in
phenotype to the normal IL-2 KO mouse (i.e., peripheral autoimmunity);
in which the IL-2 KO reconstituted IL-2 wild-type/RAG-1 KO mice would have
normal brain IL-2 expression, but autoimmunity caused by the peripheral IL-2 KO
T lymphocytes; and
in which the sham reconstituted IL-2 KO/RAG-1 KO mice lack any source of IL-2
and the sham reconstituted wild-type/RAG-1 KO mice lack a peripheral source of
IL-2, with neither of the two models succumbing to autoimmunity.
Thus, if endogenous brain IL-2 is important in septohippocampal development and/or
maintenance from the age of three-weeks to adulthood, then we should observe deficits in
the septohippocampal system similar to those noted in this dissertation in all
reconstitution and sham models above on an IL-2 KO/RAG-1 KO background. If the
peripheral source of IL-2 is more important for septohippocampal physiology, then the
structural and physiological alterations in this dissertation should be rescued in the above
IL-2 wild-type reconstituted cases, but not IL-2 KO or sham reconstituted animals. If
autoimmunity is the determining factor, then IL-2 KO reconstituted, but not sham
reconstituted, animals will all exhibit deficits in the septohippocampal system regardless
of background. This proposed experiment combining IL-2 KO strains, immunodeficient

72
strains, and reconstitution allows us to control for several immunological and genetic
factors without disruption of the BBB.
Concluding Remarks
In the peripheral immune system, the complex interplay and interactions of the
various cytokines often have redundant, supportive, or even oppositional roles. This
complexity allows for a system of compensatory and regulatory control of the immune
response. Similar overlaps and checks and balances are also present in the CNS, so
understanding how these intricate systems interact can prove daunting. Nevertheless,
elucidating the relationship between the brain and immune system molecules may have
profound clinical utility. The IL-2 KO mouse is a complicated model to study a cytokine
brain interaction and, in this dissertation, we have attempted to simplify the model by not
experimentally manipulating it. Our goal was to lay groundwork for future studies on
this topic, whereby we hope to more fully understand the precise mechanisms by which
IL-2 alters brain physiology.

WORKS CITED
LA Aarden, TK Brunner, JC Cerottini, J Dayer, AL de Week, CA Dinarello, G Di
Sabato, JJ Farrar, I Gery, S Gillis, RE Handschumacher, CS Henney, MK
Hoffmann, WJ Koopman, SM Krane, LB Lachman, I Lefkowitz, RI Mishell, SB
Mizel, JJ Oppenheim, V Paetkau, J Plate, M Rollinghoff, D Rosenstreich, AS
Rosenthal, LJ Rosenwasser, A Schimpl, HS Shin, PL Simon, KA Smith, H
Wagner, JD Watson, E Wecker and DD Wood, 1979. Revised nomenclature for
antigen-nonspecific T cell proliferation and helper factors. J Immunol 123, 2928-
2929.
R Ader and N Cohen, 1975. Behaviorally conditioned immunosuppression. Psychosom
Med 37, 333-340.
R Ader, N Cohen and D Bovbjerg, 1982. Conditioned suppression of humoral immunity
in the rat. J Comp Physiol Psychol 96, 517-521.
R Ader, D Felten and N Cohen (Eds), 2001. Psychoneuroimmunology, Third Edition
Edition. Academic Press, San Diego, CA.
L Adorini, 1999. Interleukin-12, a key cytokine in Thl-mediated autoimmune diseases.
Cell Mol Life Sci 55, 1610-1625.
RF Alderson, AL Alterman, YA Barde and RM Lindsay, 1990. Brain-derived
neurotrophic factor increases survival and differentiated functions of rat septal
cholinergic neurons in culture. Neuron 5, 297-306.
F Aloisi, R De Simone, S Columba-Cabezas, G Penna and L Adorini, 2000. Functional
maturation of adult mouse resting microglia into an APC is promoted by
granulocyte-macrophage colony-stimulating factor and interaction with Thl cells. J
Immunol 164, 1705-1712.
A Alonso and C Kohler, 1984. A study of the reciprocal connections between the septum
and the entorhinal area using anterograde and retrograde axonal transport methods
in the rat brain. J Comp Neurol 225, 327-343.
R Alonso, I Chaudieu, J Diorio, A Krishnamurthy, R Quirion and P Boksa, 1993.
Interleukin-2 modulates evoked release of [3H]dopamine in rat cultured
mesencephalic cells. J Neurochem 61, 1284-1290.
73

74
DM Araujo, PA Lapchak, B Collier and R Quirion, 1989. Localization of interleukin-2
immunoreactivity and interleukin-2 receptors in the rat brain: interaction with the
cholinergic system. Brain Res 498, 257-266.
Y Arimatsu and M Miyamoto, 1991. Survival-promoting effect ofNGF on in vitro
septohippocampal neurons with cholinergic and GABAergic phenotypes. Brain Res
Dev Brain Res 58, 189-201.
DM Armstrong, G Bruce, LB Hersh and FH Gage, 1987. Development of cholinergic
neurons in the septal/diagonal band complex of the rat. Brain Res 433, 249-256.
WS Armstrong and P Kazanjian, 2001. Use of cytokines in human immunodeficiency
virus-infected patients: colony-stimulating factors, erythropoietin, and interleukin-
2. Clin Infect Dis 32, 766-773.
MB Atkins, MT Lotze, JP Dutcher, RI Fisher, G Weiss, K Margolin, J Abrams, M Sznol,
D Parkinson, M Hawkins, C Paradise, L Kunkel and SA Rosenberg, 1999. High-
dose recombinant interleukin 2 therapy for patients with metastatic melanoma:
analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 17, 2105-
2116.
IB Autenrieth, N Bucheler, E Bohn, G Heinze and I Horak, 1997. Cytokine mRNA
expression in intestinal tissue of interleukin-2 deficient mice with bowel
inflammation. Gut 41, 793-800.
H Awatsuji, Y Furukawa, M Nakajima, S Furukawa and K Hayashi, 1993. Interleukin-2
as a neurotrophic factor for supporting the survival of neurons cultured from
various regions of fetal rat brain. J Neurosci Res 35, 305-31 1.
AA Babcock, WA Kuziel, S Rivest and T Owens, 2003. Chemokine expression by glial
cells directs leukocytes to sites of axonal injury in the CNS. J Neurosci 23, 7922-
7930.
R Badolato, AN Ponzi, M Millesimo, LD Notarangelo and T Musso, 1997. Interleukin-15
(IL-15) induces IL-8 and monocyte chemotactic protein 1 production in human
monocytes. Blood 90, 2804-2809.
WA Banks, AJ Kastin and EG Gutierrez, 1994. Penetration of interleukin-6 across the
murine blood-brain barrier. Neurosci Lett 179, 53-56.
WA Banks, SA Farr and JE Morley, 2002. Entry of blood-borne cytokines into the
central nervous system: effects on cognitive processes. Neuroimmunomodulation
10,319-327.
WA Banks, L Ortiz, SR Plotkin and AJ Kastin, 1991. Human interleukin (IL) 1 alpha,
murine IL-1 alpha and murine IL-1 beta are transported from blood to brain in the
mouse by a shared saturable mechanism. J Pharmacol Exp Ther 259, 988-996.

75
MG Baxter, DJ Bucci, TJ Sobel, MJ Williams, LK Gorman and M Gallagher, 1996.
Intact spatial learning following lesions of basal forebrain cholinergic neurons.
Neuroreport 7, 1417-1420.
SA Bayer, 1980. Development of the hippocampal region in the rat. I. Neurogenesis
examined with 3H-thymidine autoradiography. J Comp Neurol 190, 87-114.
JF Bazan, 1992. Unraveling the structure of IL-2. Science 257, 410-413.
RD Beck, Z Huang, GK Ha and JM Petitto, 2004. Altered Hippocampal Structure and
Neurotrophin Levels in Interleukin-2 Knockout Mice, (to be submitted for
publication).
RD Beck, Jr., MA King, Z Huang and JM Petitto, 2002. Alterations in septohippocampal
cholinergic neurons resulting from interleukin-2 gene knockout. Brain Res 955, 16-
23.
R Bender, M Plaschke, T Naumann, P Wahle and M Frotscher, 1996. Development of
cholinergic and GABAergic neurons in the rat medial septum: different onset of
choline acetyltransferase and glutamate decarboxylase mRNA expression. J Comp
Neurol 372, 204-214.
EN Benveniste and JE Merrill, 1986. Stimulation of oligodendroglial proliferation and
maturation by interleukin-2. Nature 321,610-613.
EN Benveniste, PK Herman and JN Whitaker, 1987. Myelin basic protein-specific RNA
levels in interleukin-2-stimulated oligodendrocytes. J Neurochem 49, 1274-1279.
M Besser and R Wank, 1999. Cutting edge: clonally restricted production of the
neurotrophins brain-derived neurotrophic factor and neurotrophin-3 mRNA by
human immune cells and Thl/Th2-polarized expression of their receptors. J
Immunol 162, 6303-6306.
M Bianchi and AE Panerai, 1993. Interleukin-2 enhances scopolamine-induced amnesia
and hyperactivity in the mouse. Neuroreport 4, 1046-1048.
M Bianchi, P Ferrario, N Zonta and AE Panerai, 1995. Effects of interleukin-1 beta and
interleukin-2 on amino acids levels in mouse cortex and hippocampus. Neuroreport
6, 1689-1692.
JL Bizon, JS Han, C Hudon and M Gallagher, 2003. Effects of hippocampal cholinergic
deafferentation on learning strategy selection in a visible platform version of the
water maze. Hippocampus 13, 676-684.
HR Brashear, L Zaborszky and L Heimer, 1986. Distribution of GABAergic and
cholinergic neurons in the rat diagonal band. Neuroscience 17, 439-451.

76
A Braun, M Lommatzsch, A Mannsfeldt, U Neuhaus-Steinmetz, A Fischer, N Schnoy,
GR Lewin and H Renz, 1999. Cellular sources of enhanced brain-derived
neurotrophic factor production in a mouse model of allergic inflammation. Am J
Respir Cell Mol Biol 21, 537-546.
FC Bronfman, D Moechars and F Van Leuven, 2000. Acetylcholinesterase-positive fiber
deafferentation and cell shrinkage in the septohippocampal pathway of aged
amyloid precursor protein london mutant transgenic mice. Neurobiol Dis 7, 152-
168.
AI Brooks, MW Halterman and HJ Federoff, 1999. Focal hippocampal gain of NGF
function elicits specific septal cholinergic reorganization. Neuroreport 10, 337-344.
KM Buller, 2001. Role of circumventricular organs in pro-inflammatory cytokine-
induced activation of the hypothalamic-pituitary-adrenal axis. Clin Exp Pharmacol
Physiol 28, 581-589.
JF Cahill and MG Baxter, 2001. Cholinergic and noncholinergic septal neurons modulate
strategy selection in spatial learning. Eur J Neurosci 14, 1856-1864.
SB Caine, MA Geyer and NR Swerdlow, 1992. Hippocampal modulation of acoustic
startle and prepulse inhibition in the rat. Pharmacol Biochem Behav 43, 1201-1208.
JD Campbell, G Cook, SE Robertson, A Fraser, KS Boyd, JA Gracie and IM Franklin,
2001. Suppression of IL-2-induced T cell proliferation and phosphorylation of
STAT3 and STAT5 by tumor-derived TGF beta is reversed by IL-15. J Immunol
167,553-561.
M Canossa, O Griesbeck, B Beminger, G Campana, R Kolbeck and H Thoenen, 1997.
Neurotrophin release by neurotrophins: implications for activity-dependent
neuronal plasticity. Proc Natl Acad Sci USA 94, 13279-13286.
L Capuron, A Ravaud and R Dantzer, 2000. Early depressive symptoms in cancer
patients receiving interleukin 2 and/or interferon alfa-2b therapy. J Clin Oncol 18,
2143-2151.
L Capuron, A Ravaud and R Dantzer, 2001a. Timing and specificity of the cognitive
changes induced by interleukin-2 and interferon-alpha treatments in cancer patients.
Psychosom Med 63, 376-386.
L Capuron, A Ravaud, N Gualde, E Bosmans, R Dantzer, M Maes and PJ Neveu, 2001b.
Association between immune activation and early depressive symptoms in cancer
patients treated with interleukin-2-based therapy. Psychoneuroendocrinology 26,
797-808.
LL Carter, X Zhang, C Dubey, P Rogers, L Tsui and SL Swain, 1998. Regulation of T
cell subsets from naive to memory. J Immunother 21, 181-187.

77
JC Cassel, E Duconseille, H Jeltsch and B Will, 1997. The fimbria-fomix/cingular bundle
pathways: a review of neurochemical and behavioural approaches using lesions and
transplantation techniques. Prog Neurobiol 51, 663-716.
JP Chandler and KA Crutcher, 1983. The septohippocampal projection in the rat: an
electron microscopic horseradish peroxidase study. Neuroscience 10, 685-696.
Q Chang and PE Gold, 2004. Impaired and spared cholinergic functions in the
hippocampus after lesions of the medial septum/vertical limb of the diagonal band
with 192 IgG-saporin. Hippocampus 14, 170-179.
J Chappell, R McMahan, A Chiba and M Gallagher, 1998. A re-examination of the role
of basal forebrain cholinergic neurons in spatial working memory.
Neuropharmacology 37, 481-487.
TJ Collier and A Routtenberg, 1984. Selective impairment of declarative memory
following stimulation of dentate gyrus granule cells: a naloxone-sensitive effect.
Brain Res 310, 384-387.
JM Conner and S Varn, 1997. Developmental profile of NGF immunoreactivity in the
rat brain: a possible role of NGF in the establishment of cholinergic terminal fields
in the hippocampus and cortex. Brain Res Dev Brain Res 101,67-79.
JM Conner, D Muir, S Varn, T Hagg and M Manthorpe, 1992. The localization of nerve
growth factor-like immunoreactivity in the adult rat basal forebrain and
hippocampal formation. J Comp Neurol 319, 454-462.
CD Conrad and EJ Roy, 1993. Selective loss of hippocampal granule cells following
adrenalectomy: implications for spatial memory. J Neurosci 13, 2582-2590.
AK Cross and MN Woodroofe, 1999. Chemokines induce migration and changes in actin
polymerization in adult rat brain microglia and a human fetal microglial cell line in
vitro. J Neurosci Res 55, 17-23.
P Curzon, DJ Kim and MW Decker, 1994. Effect of nicotine, lobeline, and
mecamylamine on sensory gating in the rat. Pharmacol Biochem Behav 49, 877-
882.
JD Cushman, Z Huang, G Ha, R Beck, J Lo and JM Petitto, 2004. Contextual fear
discrimination learning and memory is disrupted in IL-2 knockout mice. Society of
Neuroscience Abstracts.
CB Davis and SD Gillies, 2003. Immunocytokines: amplification of anti-cancer
immunity. Cancer Immunol Immunother 52, 297-308.

78
KD Denicoff, DR Rubinow, MZ Papa, C Simpson, CA Seipp, MT Lotze, AE Chang, D
Rosenstein and SA Rosenberg, 1987. The neuropsychiatric effects of treatment
with interleukin-2 and lymphokine-activated killer cells. Ann Intern Med 107, 293-
300.
JH Dufour, M Dziejman, MT Liu, JH Leung, TE Lane and AD Luster, 2002. IFN-
gamma-inducible protein 10 (IP-10; CXCL10)-deficient mice reveal a role for IP-
10 in effector T cell generation and trafficking. J Immunol 168, 3195-3204.
AJ Dunn, 2002. Mechanisms by which cytokines signal the brain. Int Rev Neurobiol 52,
43-65.
O Eizenberg, A Faber-Elman, M Lotan and M Schwartz, 1995. Interleukin-2 transcripts
in human and rodent brains: possible expression by astrocytes. J Neurochem 64,
1928-1936.
H Fukuoka, M Kawatani, T Hisamitsu and C Takeshige, 1994. Cutaneous hyperalgesia
induced by peripheral injection of interleukin-1 beta in the rat. Brain Res 657, 133-
MO.
G Fyfe, RI Fisher, SA Rosenberg, M Sznol, DR Parkinson and AC Louie, 1995. Results
of treatment of 255 patients with metastatic renal cell carcinoma who received
high-dose recombinant interleukin-2 therapy. J Clin Oncol 13, 688-696.
SL Gaffen, 2001. Signaling domains of the interleukin 2 receptor. Cytokine 14, 63-77.
FH Gage and A Bjorklund, 1987. Trophic and growth-regulating mechanisms in the
central nervous system monitored by intracerebral neural transplants. Ciba Found
Symp 126, 143-159.
FH Gage, A Bjorklund and U Stenevi, 1983a. Reinnervation of the partially deafferented
hippocampus by compensatory collateral sprouting from spared cholinergic and
noradrenergic afferents. Brain Res 268, 27-37.
FH Gage, A Bjorklund and U Stenevi, 1984. Cells of origin of the ventral cholinergic
septohippocampal pathway undergoing compensatory collateral sprouting
following fimbria-fornix transection. Neurosci Lett 44, 211-216.
FH Gage, A Bjorklund, U Stenevi and SB Dunnett, 1983b. Functional correlates of
compensatory collateral sprouting by aminergic and cholinergic afferents in the
hippocampal formation. Brain Res 268, 39-47.
BH Gahwiler, A Enz and F Hefti, 1987. Nerve growth factor promotes development of
the rat septo-hippocampal cholinergic projection in vitro. Neurosci Lett 75, 6-10.

79
D Galey, C Destrade and R Jaffard, 1994. Relationships between septo-hippocampal
cholinergic activation and the improvement of long-term retention produced by
medial septal electrical stimulation in two inbred strains of mice. Behav Brain Res
60, 183-189.
R Ganguli, JS Brar, KR Chengappa, M DeLeo, ZW Yang, G Shurin and BS Rabin, 1995.
Mitogen-stimulated interleukin-2 production in never-medicated, first- episode
schizophrenic patients. The influence of age at onset and negative symptoms. Arch
Gen Psychiatry 52, 668-672.
JG Giri, DM Anderson, S Kumaki, LS Park, KH Grabstein and D Cosman, 1995. IL-15, a
novel T cell growth factor that shares activities and receptor components with IL-2.
J Leukoc Biol 57, 763-766.
PE Gold, 2003. Acetylcholine modulation of neural systems involved in learning and
memory. Neurobiol Leam Mem 80, 194-210.
E Gould, NJ Woolf and LL Butcher, 1991. Postnatal development of cholinergic neurons
in the rat: I. Forebrain. Brain Res Bull 27, 767-789.
E Gould, P Tanapat, T Rydel and N Hastings, 2000. Regulation of hippocampal
neurogenesis in adulthood. Biol Psychiatry 48, 715-720.
I Gritti, M Mariotti and M Mancia, 1998. GABAergic and cholinergic basal forebrain and
preoptic-anterior hypothalamic projections to the mediodorsal nucleus of the
thalamus in the cat. Neuroscience 85, 149-178.
I Gritti, L Mainville, M Mancia and BE Jones, 1997. GABAergic and other
noncholinergic basal forebrain neurons, together with cholinergic neurons, project
to the mesocortex and isocortex in the rat. J Comp Neurol 383, 163-177.
LM Guirguis, JC Yang, DE White, SM Steinberg, DJ Liewehr, SA Rosenberg and DJ
Schwartzentruber, 2002. Safety and efficacy of high-dose interleukin-2 therapy in
patients with brain metastases. J Immunother 25, 82-87.
H Guo, JH Calkins, MM Sigel and T Lin, 1990. Interleukin-2 is a potent inhibitor of
Leydig cell steroidogenesis. Endocrinology 127, 1234-1239.
EG Gutierrez, WA Banks and AJ Kastin, 1993. Murine tumor necrosis factor alpha is
transported from blood to brain in the mouse. J Neuroimmunol 47, 169-176.
T Hagg, HL Vahlsing, M Manthorpe and S Varn, 1990. Nerve growth factor infusion
into the denervated adult rat hippocampal formation promotes its cholinergic
reinnervation. J Neurosci 10, 3087-3092.
UK Hanisch and R Quirion, 1995a. Interleukin-2 as a neuroregulatory cytokine. Brain
Res Brain Res Rev 21, 246-284.

80
UK Hanisch and R Quirion, 1995b. Interleukin-2 as a neuroregulatory cytokine. Brain
Res Brain Res Rev 21, 246-284.
UK Hanisch, D Seto and R Quirion, 1993. Modulation of hippocampal acetylcholine
release: a potent central action of interleukin-2. J Neurosci 13, 3368-3374.
UK Hanisch, J Neuhaus, R Quirion and H Kettenmann, 1996. Neurotoxicity induced by
interleukin-2: involvement of infiltrating immune cells. Synapse 24, 104-114.
UK Hanisch, W Rowe, S Sharma, MJ Meaney and R Quirion, 1994. Hypothalamic-
pituitary-adrenal activity during chronic central administration of interleukin-2.
Endocrinology 135, 2465-2472.
UK Hanisch, SA Lyons, M Prinz, C Nolte, JR Weber, H Kettenmann and F Kirchhoff,
1997a. Mouse brain microglia express interleukin-15 and its multimeric receptor
complex functionally coupled to Janus kinase activity. J Biol Chem 272, 28853-
28860.
UK Hanisch, J Neuhaus, W Rowe, D Van Rossum, T Moller, H Kettenmann and R
Quirion, 1997b. Neurotoxic consequences of central long-term administration of
interleukin-2 in rats. Neuroscience 79, 799-818.
J Hartikka and F Hefti, 1988. Comparison of nerve growth factor's effects on
development of septum, striatum, and nucleus basalis cholinergic neurons in vitro.
J Neurosci Res 21, 352-364.
YW He and TR Malek, 1998. The structure and function of gamma c-dependent
cytokines and receptors: regulation of T lymphocyte development and homeostasis.
Crit Rev Immunol 18, 503-524.
JC Hedreen, SJ Bacon and DL Price, 1985. A modified histochemical technique to
visualize acetylcholinesterase- containing axons. J Histochem Cytochem 33, 134-
MO.
R Hellweg, C Humpel, A Lowe and H Hortnagl, 1997. Moderate lesion of the rat
cholinergic septohippocampal pathway increases hippocampal nerve growth factor
synthesis: evidence for long-term compensatory changes? Brain Res Mol Brain Res
45, 177-181.
WF Hickey and H Kimura, 1988. Perivascular microglial cells of the CNS are bone
marrow-derived and present antigen in vivo. Science 239, 290-292.
WF Hickey, BL Hsu and H Kimura, 1991. T-lymphocyte entry into the central nervous
system. J Neurosci Res 28, 254-260.
I Horak, 1995. Immunodeficiency in IL-2-knockout mice. Clin Immunol Immunopathol
76, S172-173.

81
I Horak, 1996. Interleukin-2-knockout mice: a new model to study autoimmunity and
self-tolerance. Sb Lek 97,25-28.
AA Hurwitz, WD Lyman and JW Berman, 1995. Tumor necrosis factor alpha and
transforming growth factor beta upregulate astrocyte expression of monocyte
chemoattractant protein-1. J Neuroimmunol 57, 193-198.
LS Janis, MM Glasier, Z Fulop and DG Stein, 1998. Intraseptal injections of 192 IgG
saporin produce deficits for strategy selection in spatial-memory tasks. Behav Brain
Res 90, 23-34.
DA Johnson, NJ Zambn and RB Gibbs, 2002. Selective lesion of cholinergic neurons in
the medial septum by 192 IgG-saporin impairs learning in a delayed matching to
position T-maze paradigm. Brain Res 943, 132-141.
P Kaiser, L Rothwell, D Vasicek and K Hala, 2002. A role for IL-15 in driving the onset
of spontaneous autoimmune thyroiditis? J Immunol 168, 4216-4220.
S Karanth, K Lyson and SM McCann, 1993. Role of nitric oxide in interleukin 2-induced
corticotropin-releasing factor release from incubated hypothalami. Proc Natl Acad
Sci U S A 90, 3383-3387.
M Kerschensteiner, E Gallmeier, L Behrens, VV Leal, T Misgeld, WE Klinkert, R
Kolbeck, E Hoppe, RL Oropeza-Wekerle, I Bartke, C Stadelmann, H Lassmann, H
Wekerle and R Hohlfeld, 1999. Activated human T cells, B cells, and monocytes
produce brain-derived neurotrophic factor in vitro and in inflammatory brain
lesions: a neuroprotective role of inflammation? J Exp Med 189, 865-870.
MA King, BE Hunter, RL Reep and DW Walker, 1989. Acetylcholinesterase stain
intensity variation in the rat dentate gyrus: a quantitative description based on
digital image analysis. Neuroscience 33, 203-221.
J Kiss, AJ Patel and TF Freund, 1990a. Distribution of septohippocampal neurons
containing parvalbumin or choline acetyltransferase in the rat brain. J Comp Neurol
298, 362-372.
J Kiss, AJ Patel, KG Baimbridge and TF Freund, 1990b. Topographical localization of
neurons containing parvalbumin and choline acetyltransferase in the medial
septum-diagonal band region of the rat. Neuroscience 36, 61-72.
M Knipper and RJ Rylett, 1997. A new twist in an old story: the role for crosstalk of
neuronal and trophic activity. Neurochem Int 31, 659-676.
M Knipper, M da Penlia Berzaghi, A Blochl, H Breer, H Thoenen and D Lindholm, 1994.
Positive feedback between acetylcholine and the neurotrophins nerve growth factor
and brain-derived neurotrophic factor in the rat hippocampus. Eur J Neurosci 6,
668-671.

82
M Koch, 1996a. The septohippocampal system is involved in prepulse inhibition of the
acoustic startle response in rats. Behav Neurosci 110, 468-477.
M Koch, 1996b. The septohippocampal system is involved in prepulse inhibition of the
acoustic startle response in rats. Behav Neurosci 110, 468-477.
S Lacosta, Z Merali and H Anisman, 1999. Influence of acute and repeated interleukin-2
administration on spatial learning, locomotor activity, exploratory behaviors, and
anxiety. Behav Neurosci 113, 1030-1041.
H Lai and MJ Forster, 1988. Autoimmunity and age-associated cognitive decline.
Neurobiol Aging 9, 733-742.
MR Lamprea, FP Cardenas, R Silveira, S Morato and TJ Walsh, 2000. Dissociation of
memory and anxiety in a repeated elevated plus maze paradigm: forebrain
cholinergic mechanisms. Behav Brain Res 117, 97-105.
PA Lapchak and DM Araujo, 1993. Interleukin-2 regulates monoamine and opioid
peptide release from the hypothalamus. Neuroreport 4, 303-306.
PA Lapchak, DM Araujo, R Quirion and A Beaudet, 1991. Immunoautoradiographic
localization of interleukin 2-like immunoreactivity and interleukin 2 receptors (Tac
antigen-like immunoreactivity) in the rat brain. Neuroscience 44, 173-184.
E Larsson, RJ Mandel, RL Klein, N Muzyczka, O Lindvall and Z Kokaia, 2002.
Suppression of insult-induced neurogenesis in adult rat brain by brain-derived
neurotrophic factor. Exp Neurol 177, 1-8.
JM Lauder and UB Schambra, 1999. Morphogenetic roles of acetylcholine. Environ
Health Perspect 107 Suppl 1, 65-69.
BR Lauwerys, N Garot, JC Renauld and FA Houssiau, 2000. Cytokine production and
killer activity of NK/T-NK cells derived with IL-2, IL-15, or the combination of
IL-12 and IL-18. J Immunol 165, 1847-1853.
G Leanza, OG Nilsson and A Bjorklund, 1993. Compensatory changes of in vivo
acetylcholine and noradrenaline release in the hippocampus after partial
deafferentation, as monitored by microdialysis. Brain Res 615, 147-159.
G Leanza, OG Nilsson, RG Wiley and A Bjorklund, 1995. Selective lesioning of the
basal forebrain cholinergic system by intraventricular 192 IgG-saporin:
behavioural, biochemical and stereological studies in the rat. Eur J Neurosci 7, 329-
343.
J Lee, W Duan and MP Mattson, 2002. Evidence that brain-derived neurotrophic factor is
required for basal neurogenesis and mediates, in part, the enhancement of
neurogenesis by dietary restriction in the hippocampus of adult mice. J Neurochem
82, 1367-1375.

83
YB Lee, J Satoh, DG Walker and SU Kim, 1996. Interleukin-15 gene expression in
human astrocytes and microglia in culture. Neuroreport 7, 1062-1066.
WJ Leonard, JM Depper, GR Crabtree, S Rudikoff, J Pumphrey, RJ Robb, M Kronke, PB
Svetlik, NJ Peffer, TA Waldmann and WC Greene, 1984. Molecular cloning and
expression of cDNAs for the human interleukin-2 receptor. Nature 311, 626-631.
J Licinio, JP Seibyl, M Altemus, DS Chamey and JH Krystal, 1993. Elevated CSF levels
of interleukin-2 in neuroleptic-free schizophrenic patients. Am J Psychiatry 150,
1408-1410.
T Lindenlaub, P Teuteberg, T Hartung and C Sommer, 2000. Effects of neutralizing
antibodies to TNF-alpha on pain-related behavior and nerve regeneration in mice
with chronic constriction injury. Brain Res 866, 15-22.
BR Ludviksson, RO Ehrhardt and W Strober, 1997. TGF-beta production regulates the
development of the 2,4,6-trinitrophenol-conjugated keyhole limpet hemocyanin-
induced colonic inflammation in IL-2-defcient mice. J Immunol 159, 3622-3628.
A Ma, M Datta, E Margosian, J Chen and I Horak, 1995. T cells, but not B cells, are
required for bowel inflammation in interleukin 2-deficient mice. J Exp Med 182,
1567-1572.
SF Maier, LE Goehler, M Fleshner and LR Watkins, 1998. The role of the vagus nerve in
cytokine-to-brain communication. Ann N Y Acad Sci 840, 289-300.
R Makuch, J Baratta, LD Karaelias, JC Lauterbom, CM Gall, J Yu and RT Robertson,
2001. Arrival of afferents and the differentiation of target neurons: studies of
developing cholinergic projections to the dentate gyrus. Neuroscience 104, 81-91.
D Maslinska, 2001. The cytokine network and interleukin-15 (IL-15) in brain
development. Folia Neuropathol 39, 43-47.
CG McAllister, DP van Kammen, TJ Rehn, AL Miller, J Gurklis, ME Kelley, J Yao and
JL Peters, 1995. Increases in CSF levels of interleukin-2 in schizophrenia: effects
of recurrence of psychosis and medication status. Am J Psychiatry 152, 1291-1297.
JF McDyer, Z Li, S John, X Yu, CY Wu and JA Ragheb, 2002. IL-2 receptor blockade
inhibits late, but not early, IFN-gamma and CD40 ligand expression in human T
cells: disruption ofboth IL-12-dependent and -independent pathways of IFN-
gamma production. J Immunol 169, 2736-2746.
RL McLamb, WR Mundy and HA Tilson, 1988. Intradentate colchicine disrupts the
acquisition and performance of a working memory task in the radial arm maze.
Neurotoxicology 9, 521-528.

84
L Meda, S Bemasconi, C Bonaiuto, S Sozzani, D Zhou, L Otvos, Jr., A Mantovani, F
Rossi and MA Cassatella, 1996. Beta-amyloid (25-35) peptide and IFN-gamma
synergistically induce the production of the chemotactic cytokine MCP-l/JE in
monocytes and microglial cells. J Immunol 157, 1213-1218.
MA Meijssen, SL Brandwein, HC Reinecker, AK Bhan and DK Podolsky, 1998.
Alteration of gene expression by intestinal epithelial cells precedes colitis in
interleukin-2-defcient mice. Am J Physiol 274, G472-479.
F Mennicken and R Quirion, 1997. Interleukin-2 increases choline acetyltransferase
activity in septal- cell cultures. Synapse 26, 175-183.
RT Mitsuyasu, 2001. The potential role of interleukin-2 in HIV. Aids 15 Suppl 2, S22-27.
H Mizuma, K Takagi, K Miyake, N Takagi, K Ishida, S Takeo, A Nitta, H Nomoto, Y
Furukawa and S Furukawa, 1999. Microsphere embolism-induced elevation of
nerve growth factor level and appearance of nerve growth factor immunoreactivity
in activated T-lymphocytes in the rat brain. J Neurosci Res 55, 749-761.
G Moalem, A Gdalyahu, Y Shani, U Otten, P Lazarovici, IR Cohen and M Schwartz,
2000. Production of neurotrophins by activated T cells: implications for
neuroprotective autoimmunity. J Autoimmun 15, 331-345.
WC Mobley, JL Rutkowski, GI Tennekoon, J Gemski, K Buchanan and MV Johnston,
1986. Nerve growth factor increases choline acetyltransferase activity in
developing basal forebrain neurons. Brain Res 387, 53-62.
ML Monje, H Toda and TD Palmer, 2003. Inflammatory blockade restores adult
hippocampal neurogenesis. Science 302, 1760-1765.
DA Morgan, FW Ruscetti and R Gallo, 1976. Selective in vitro growth of T lymphocytes
from normal human bone marrows. Science 193, 1007-1008.
JK Morse, SJ Wiegand, K Anderson, Y You, N Cai, J Carnahan, J Miller, PS DiStefano,
CA Altar, RM Lindsay and et al., 1993. Brain-derived neurotrophic factor (BDNF)
prevents the degeneration of medial septal cholinergic neurons following fimbria
transection. J Neurosci 13, 4146-4156.
N Muller and M Ackenheil, 1998. Psychoneuroimmunology and the cytokine action in
the CNS: implications for psychiatric disorders. Prog Neuropsychopharmacol Biol
Psychiatry 22, 1-33.
Y Nakamura, SM Russell, SA Mess, M Friedmann, M Erdos, C Francois, Y Jacques, S
Adelstein and WJ Leonard, 1994. Heterodimerization of the IL-2 receptor beta- and
gamma-chain cytoplasmic domains is required for signalling. Nature 369, 330-333.

85
KP Nanry, WR Mundy and HA Tilson, 1989. Colchicine-induced alterations of reference
memory in rats: role of spatial versus non-spatial task components. Behav Brain
Res 35, 45-53.
BH Nelson, 2004. IL-2, regulatory T cells, and tolerance. J Immunol 172, 3983-3988.
BH Nelson, JD Lord and PD Greenberg, 1994. Cytoplasmic domains of the interleukin-2
receptor beta and gamma chains mediate the signal for T-cell proliferation. Nature
369, 333-336.
R Nemni, S Iannaccone, A Quattrini, S Smime, M Sessa, M Lodi, C Erminio and N
Canal, 1992. Effect of chronic treatment with recombinant interleukin-2 on the
central nervous system of adult and old mice. Brain Res 591, 248-252.
W Pan, WA Banks and AJ Kastin, 1997. Permeability of the blood-brain and blood-
spinal cord barriers to interferons. J Neuroimmunol 76, 105-111.
K Pardy, D Murphy, D Carter and KM Hui, 1993. The influence of interleukin-2 on
vasopressin and oxytocin gene expression in the rodent hypothalamus. J
Neuroimmunol 42, 131-138.
DR Parkinson, JS Abrams, PH Wiemik, AA Rayner, KA Margolin, DA Van Echo, M
Sznol, JP Dutcher, FR Aronson, JH Doroshow and et al., 1990. Interleukin-2
therapy in patients with metastatic malignant melanoma: a phase II study. J Clin
Oncol 8, 1650-1656.
A Patapoutian and LF Reichardt, 2001. Trk receptors: mediators of neurotrophin action.
Curr Opin Neurobiol 11, 272-280.
S Pauli, AC Linthorst and JM Reul, 1998. Tumour necrosis factor-alpha and interleukin-2
differentially affect hippocampal serotonergic neurotransmission, behavioural
activity, body temperature and hypothalamic-pituitary-adrenocortical axis activity
in the rat. Eur J Neurosci 10, 868-878.
M Perez-Martin, I Azcoitia, JL Trejo, A Sierra and LM Garcia-Segura, 2003. An
antagonist of estrogen receptors blocks the induction of adult neurogenesis by
insulin-like growth factor-I in the dentate gyrus of adult female rat. Eur J Neurosci
18,923-930.
T Perry, H Hodges and JA Gray, 2001. Behavioural, histological and
immunocytochemical consequences following 192 IgG-saporin immunolesions of
the basal forebrain cholinergic system. Brain Res Bull 54, 29-48.
JM Petitto and Z Huang, 1994. Molecular cloning of a partial cDNA of the interleukin-2
receptor-beta in normal mouse brain: in situ localization in the hippocampus and
expression by neuroblastoma cells. Brain Res 650, 140-145.

86
JM Petitto and Z Huang, 1995. Molecular cloning of the coding sequence of an
interleukin-2 receptor alpha subunit cDNA in murine brain. J Neuroimmunol 59,
135-141.
JM Petitto and Z Huang, 2001. Cloning the full-length IL-2/15 receptor-beta cDNA
sequence from mouse brain: evidence of enrichment in hippocampal formation
neurons. Regul Pept 98, 77-87.
JM Petitto, Z Huang, J Lo and WJ Streit, 2003. IL-2 gene knockout affects T lymphocyte
trafficking and the microglial response to regenerating facial motor neurons. J
Neuroimmunol 134, 95-103.
JM Petitto, DB McCarthy, CM Rinker, Z Huang and T Getty, 1997. Modulation of
behavioral and neurochemical measures of forebrain dopamine function in mice by
species-specific interleukin-2. J Neuroimmunol 73, 183-190.
JM Petitto, Z Huang, MK Raizada, CM Rinker and DB McCarthy, 1998. Molecular
cloning of the cDNA coding sequence of IL-2 receptor-gamma (gammac) from
human and murine forebrain: expression in the hippocampus in situ and by brain
cells in vitro. Brain Res Mol Brain Res 53, 152-162.
JM Petitto, RK McNamara, PL Gendreau, Z Huang and AJ Jackson, 1999. Impaired
learning and memory and altered hippocampal neurodevelopment resulting from
interleukin-2 gene deletion. J Neurosci Res 56, 441-446.
JM Petitto, WJ Streit, Z Huang, E Butfiloski and J Schiffenbauer, 2000. Interleukin-2
gene deletion produces a robust reduction in susceptibility to experimental
autoimmune encephalomyelitis in C57BL/6 mice. Neurosci Lett 285, 66-70.
CR Plata-Salaman and JM ffrench-Mullen, 1993. Interleukin-2 modulates calcium
currents in dissociated hippocampal CA1 neurons. Neuroreport 4, 579-581.
LQ Ren, N Gourmala, HW Boddeke and PJ Gebicke-Haerter, 1998. Lipopolysaccharide-
induced expression of IP-10 mRNA in rat brain and in cultured rat astrocytes and
microglia. Brain Res Mol Brain Res 59, 256-263.
GE Ringheim, BD Freimark and RJ Robb, 1991. Quantitative characterization of the
intrinsic ligand-binding affinity of the interleukin 2 receptor beta chain and its
modulation by the alpha chain and a second affinity-modulating element.
Lymphokine Cytokine Res 10, 219-224.
SA Rosenberg, JJ Mule, PJ Spiess, CM Reichert and SL Schwarz, 1985. Regression of
established pulmonary metastases and subcutaneous tumor mediated by the
systemic administration of high-dose recombinant interleukin 2. J Exp Med 161,
1169-1188.

87
SA Rosenberg, JC Yang, SL Topaban, DJ Schwartzentruber, JS Weber, DR Parkinson,
CA Seipp, JH Einhom and DE White, 1994. Treatment of 283 consecutive patients
with metastatic melanoma or renal cell cancer using high-dose bolus interleukin 2.
Jama 271, 907-913.
R Rozental, MF Mehler, M Morales, AF Andrade-Rozental, JA Kessler and DC Spray,
1995. Differentiation of hippocampal progenitor cells in vitro: temporal expression
of intercellular coupling and voltage- and ligand-gated responses. Dev Biol 167,
350-362.
F Ruberti, S Capsoni, A Comparini, E Di Daniel, J Franzot, S Gonfloni, G Rossi, N
Berardi and A Cattaneo, 2000. Phenotypic knockout of nerve growth factor in adult
transgenic mice reveals severe deficits in basal forebrain cholinergic neurons, cell
death in the spleen, and skeletal muscle dystrophy. J Neurosci 20, 2589-2601.
T Saarelainen, S Vaittinen and E Castren, 2001. trkB-receptor activation contributes to
the kainate-induced increase in BDNF mRNA synthesis. Cell Mol Neurobiol 21,
429-435.
N Sakai, S Kaufman and S Milstien, 1995. Parallel induction of nitric oxide and
tetrahydrobiopterin synthesis by cytokines in rat glial cells. J Neurochem 65, 895-
902.
B Sakic, H Szechtman and JA Denburg, 1997. Neurobehavioral alterations in
autoimmune mice. Neurosci Biobehav Rev 21, 327-340.
B Sakic, H Szechtman, S Denburg, R Carbotte and JA Denburg, 1993. Spatial learning
during the course of autoimmune disease in MRL mice. Behav Brain Res 54, 57-
66.
RP Saneto, A Altman, RL Knobler, HM Johnson and J de Vellis, 1986. Interleukin 2
mediates the inhibition of oligodendrocyte progenitor cell proliferation in vitro.
Proc Natl Acad Sci U S A 83, 9221-9225.
M Sarder, H Saito and K Abe, 1993. Interleukin-2 promotes survival and neurite
extension of cultured neurons from fetal rat brain. Brain Res 625, 347-350.
M Sarder, K Abe, H Saito and N Nishiyama, 1996. Comparative effect of IL-2 and IL-6
on morphology of cultured hippocampal neurons from fetal rat brain. Brain Res
715,9-16.
J Satoh, K Kurohara, M Yukitake and Y Kuroda, 1998. Interleukin-15, a T-cell growth
factor, is expressed in human neural cell lines and tissues. J Neurol Sci 155, 170-
177.
A Schimpl, I Berberich, B Kneitz, S Kramer, B Santner-Nanan, S Wagner, M Wolf and T
Hunig, 2002. IL-2 and autoimmune disease. Cytokine Growth Factor Rev 13, 369-
378.

88
E Schmitt, T Germann, S Goedert, P Hoehn, C Huels, S Koelsch, R Kuhn, W Muller, N
Palm and E Rude, 1994. IL-9 production of naive CD4+ T cells depends on IL-2, is
synergistically enhanced by a combination of TGF-beta and IL-4, and is inhibited
by IFN-gamma. J Immunol 153, 3989-3996.
R Schopke, DP Wolfer, HP Lipp and MC Leisinger-Trigona, 1991. Swimming
navigation and structural variations of the infrapyramidal mossy fibers in the
hippocampus of the mouse. Hippocampus 1, 315-328.
H Schwegler and WE Crusio, 1995. Correlations between radial-maze learning and
structural variations of septum and hippocampus in rodents. Behav Brain Res 67,
29-41.
H Schwegler, WE Crusio, HP Lipp and B Heimrich, 1988. Water-maze learning in the
mouse correlates with variation in hippocampal morphology. Behav Genet 18, 153-
165.
H Schwegler, M Boldyreva, R Linke, J Wu, K Zilles and WE Crusio, 1996a. Genetic
variation in the morphology of the septo-hippocampal cholinergic and GABAergic
systems in mice: II. Morpho-behavioral correlations. Hippocampus 6, 535-545.
H Schwegler, M Boldyreva, M Pyrlik-Gohlmann, R Linke, J Wu and K Zilles, 1996b.
Genetic variation in the morphology of the septo-hippocampal cholinergic and
GABAergic system in mice. I. Cholinergic and GABAergic markers. Hippocampus
6, 136-148.
BM Segal, BK Dwyer and EM Shevach, 1998. An interleukin (IL)-10/IL-12
immunoregulatory circuit controls susceptibility to autoimmune disease. J Exp Med
187,537-546.
D Seto, S Kar and R Quirion, 1997. Evidence for direct and indirect mechanisms in the
potent modulatory action of interleukin-2 on the release of acetylcholine in rat
hippocampal slices. Br J Pharmacol 120, 1151-1157.
NM Shah, AK Groves and DJ Anderson, 1996. Alternative neural crest cell fates are
instructively promoted by TGFbeta superfamily members. Cell 85, 331-343.
M Shimojo, Y Imai, K Nakajima, S Mizushima, A Uemura and S Kohsaka, 1993.
Interleukin-2 enhances the viability of primary cultured rat neocortical neurons.
Neurosci Lett 151, 170-173.
KA Smith, 2001. Low-dose daily interleukin-2 immunotherapy: accelerating immune
restoration and expanding HIV-specific T-cell immunity without toxicity. Aids 15
Suppl 2, S28-35.
LS Sorkin, WH Xiao, R Wagner and RR Myers, 1997. Tumour necrosis factor-alpha
induces ectopic activity in nociceptive primary afferent fibres. Neuroscience 81,
255-262.

89
WJ Streit, MB Graeber and GW Kreutzberg, 1988. Functional plasticity of microglia: a
review. Glia 1, 301-307.
WJ Streit, SL Semple-Roxyland, SD Hurley, RC Miller, PG Popovich and BT Stokes,
1998. Cytokine mRNA profiles in contused spinal cord and axotomized facial
nucleus suggest a beneficial role for inflammation and gliosis. Exp Neurol 152, 74-
87.
K Sugamura, H Asao, M Kondo, N Tanaka, N Ishii, K Ohbo, M Nakamura and T
Takeshita, 1996. The interleukin-2 receptor gamma chain: its role in the multiple
cytokine receptor complexes and T cell development in XSCID. Annu Rev
Immunol 14, 179-205.
H Super and E Soriano, 1994. The organization of the embryonic and early postnatal
murine hippocampus. II. Development of entorhinal, commissural, and septal
connections studied with the lipophilic tracer Dil. J Comp Neurol 344, 101-120.
T Takeshita, H Asao, K Ohtani, N Ishii, S Kumaki, N Tanaka, H Munakata, M Nakamura
and K Sugamura, 1992. Cloning of the gamma chain of the human IL-2 receptor.
Science 257, 379-382.
V Tancredi, C Zona, F Velotti, F Eusebi and A Santoni, 1990. Interleukin-2 suppresses
established long-term potentiation and inhibits its induction in the rat hippocampus.
Brain Res 525, 149-151.
LJ Thai, E Gilbertson, DM Armstrong and FH Gage, 1992. Development of the basal
forebrain cholinergic system: phenotype expression prior to target innervation.
Neurobiol Aging 13, 67-72.
JJ Thune and B Pakkenberg, 2000. Stereological studies of the schizophrenic brain. Brain
Res Brain Res Rev 31, 200-204.
L Vallieres, IL Campbell, FH Gage and PE Sawchenko, 2002. Reduced hippocampal
neurogenesis in adult transgenic mice with chronic astrocytic production of
interleukin-6. J Neurosci 22, 486-492.
F Villemain, T Owens, T Renno and A Beaudet, 1991. Localization of mRNA for
interleukin-2 (IL-2) in mouse brain by in situ hybridization (Abstract). Society of
Neuroscience Abstracts 17, 1199.
M Virgili, A Contestabile and O Bamabei, 1991. Postnatal maturation of cholinergic
markers in forebrain regions of C57BL/6 mice. Brain Res Dev Brain Res 63, 281-
285.
PJ Waguespack, WA Banks and AJ Kastin, 1994. Interleukin-2 does not cross the blood-
brain barrier by a saturable transport system. Brain Res Bull 34, 103-109.

90
T Waldmann, 2002. The contrasting roles of IL-2 and IL-15 in the life and death of
lymphocytes: implications for the immunotherapy of rheumatological diseases.
Arthritis Res 4 Suppl 3, S161-167.
TA Waldmann, S Dubois and Y Tagaya, 2001. Contrasting roles of IL-2 and IL-15 in the
life and death of lymphocytes: implications for immunotherapy. Immunity 14. 105-
110.
LG Walker, MB Walker, SD Heys, J Lolley, K Wesnes and O Eremin, 1997. The
psychological and psychiatric effects of rIL-2 therapy: a controlled clinical trial.
Psychooncology 6, 290-301.
TJ Walsh, DW Schulz, HA Tilson and DE Schmechel, 1986. Colchicine-induced granule
cell loss in rat hippocampus: selective behavioral and histological alterations. Brain
Res 398, 23-36.
NL Ward and T Hagg, 2000. BDNF is needed for postnatal maturation of basal forebrain
and neostriatum cholinergic neurons in vivo. Exp Neurol 162, 297-310.
PC Wilkinson and FY Liew, 1995. Chemoattraction of human blood T lymphocytes by
interleukin-15. J Exp Med 181, 1255-1259.
RW Williams and P Rakic, 1988. Three-dimensional counting: an accurate and direct
method to estimate numbers of cells in sectioned material. J Comp Neurol 278,
344-352.
CJ Wilson, CE Finch and HJ Cohen, 2002. Cytokines and cognitionthe case for a head-
to-toe inflammatory paradigm. J Am Geriatr Soc 50, 2041-2056.
O Witzke, T Winterhagen, A Kribben, T Philipp, K Mann and W Reinhardt, 2003.
Interleukin-2 given to asymptomatic HIV-infected individuals leads to an
exaggerated response of the pituitary gland to the action of CRH. Clin Endocrinol
(0x0 59, 104-109.
G Wong, Y Goldshmit and AM Tumley, 2004. Interferon-gamma but not TNF alpha
promotes neuronal differentiation and neurite outgrowth of murine adult neural
stem cells. Exp Neurol 187, 171-177.
NJ Woolf, F Eckenstein and LL Butcher, 1984. Cholinergic systems in the rat brain: I.
projections to the limbic telencephalon. Brain Res Bull 13, 751-784.
CC Wrenn, DA Lappi and RG Wiley, 1999. Threshold relationship between lesion extent
of the cholinergic basal forebrain in the rat and working memory impairment in the
radial maze. Brain Res 847, 284-298.
K Yoshida and H Oka, 1995. Topographical projections from the medial septum-diagonal
band complex to the hippocampus: a retrograde tracing study with multiple
fluorescent dyes in rats. Neurosci Res 21, 199-209.

91
JQ Zheng, M Felder, JA Connor and MM Poo, 1994. Turning of nerve growth cones
induced by neurotransmitters. Nature 368, 140-144.

BIOGRAPHICAL SKETCH
Ray Dennis Beck, Jr. was bom in York, PA, to Daisy and the appropriately named
Ray Dennis Beck, Sr. Realizing the error of his ways early in life, Ray moved to
Houston, TX, at the age of two, lending credence to the old Texas saying, I wasnt bom
in Texas, but I got here as fast as I could. He attended Oak Ridge High School until the
age of 16. Ranked 14 overall in his class, he elected to forego his senior year to attend
Simons Rock College of Bard in Great Barrington, MA. Whereas some 16-year-old
adolescents are mature enough to pursue a college education while resisting distractions
like geographically convenient buildings filled with members of the opposite sex, a ready
supply of fermented beverages, and complete lack of parental supervision Ray was not.
He returned to Houston after his first year of college with a less than stellar GPA intent
on taking a year off from school. During this year (comprised of-1,825 days), he held
various jobs ranging from perfume salesman to waiter at various restaurants. During one
of the latter jobs, he met his present wife, Laura Frakey. Inspired by her enthusiasm for
education, Ray enrolled in the University of Houston, while maintaining full-time
employment to pay for school. He graduated cum laude with a B.S. in biology. Ray and
Laura moved to Gainesville, FL, to attend University of Florida graduate programs in
neuroscience and psychology, respectively. With the completion of his Ph.D., Ray is
proud to be one of the most educated high school dropouts that anyone is ever likely to
meet. Ray is an avid follower of movies and his other interests can be classified as all
things geeky (e.g., computers, roleplaying, video games, etc.).
92

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
John Petitto,
Profssor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
j y l t j..,'
Michael A. King, Cochair
Associate Scientist of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Mark H. Lewis
Professor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.' ft ,
.M-- *
Wolfgang J. Streit
Professor of Neuroscience
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.. v
K a \ v'
"
Mark A. Atkinson
Professor of Pathology, Immunology and
Laboratory Medicine

This dissertation was submitted to the Graduate Faculty of the College of Medicine
and to the Graduate School and was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
August 2004
Dean, College of Medicine
Dean, Graduate School



18
The MCID software interpolated the total volume of the MS/vDB based on the
volume of the count areas defined by the user. Cell density (Ny) was estimated by
dividing the total number of cells counted by the volume of the counting boxes, which
was also tracked by the software. The total cell number was estimated by multiplying Ny
by the total volume.
Quantitative Image Analysis of AChE Staining
For quantitative analysis of AChE staining, we modified previously described
methods used to measure intensity of staining and comparisons of normalized length
across CA1, CA3b, and the suprapyramidal (SP) and infrapyramidal (IP) blades of the
dentate gyrus (DG) (King et al., 1989; Schwegler et al., 1996b). Images of the
hippocampus in each tissue section sampled from a light microscope (Olympus BH-2)
were relayed by digital video camera (Hitachi KP-D581) to a computer frame grabber
(Flashpoint 128, Integral Technologies) and digitized to 640x480 pixel images with 256
gray levels from black to white. Imaging software (Image Pro Plus v.4.0, Media
Cybernetics) was utilized to define a broad sampling traverse across various areas of the
hippocampus, approximately perpendicular to the cell body layers. For CA1 and CA3b,
this traverse extended from the alveus to the hippocampal fissure. For the DG, the line
extended from the hilus to hippocampal fissure or to the pial surface for the SP or IP limb
of the DG, respectively. AChE-containing fiber density was estimated by using the gray
level of each point along the line, which was calculated by averaging the intensity value
of each pixel across the width of the sampling band (e.g. a traverse with a width of 46
pixels would have an average of 46 measurements approximately parallel to each data
point). Gray level intensity measurements were used to estimate the AChE reaction
product density at each point along the traverse. The mean pixel intensity of a small box


12
The potential effects of IL-2 on cholinergic neurons are particularly relevant to
this study. In addition to the aforementioned trophic effects of IL-2 on cultured septal
neurons, IL-2 is among the most potent modulators of acetylcholine (ACh) release from
cultured septohippocampal neurons (Araujo et al., 1989; Hanisch et al., 1993; Seto et ah,
1997), and can also modulate its precursor enzyme, choline acetyltransferase (ChAT) in
fetal neurons (Mennicken and Quirion, 1997). Alterations in the cytoarchitecture of
cholinergic septohippocampal neurons have been shown to correlate with differences in
spatial learning ability in mice (Schwegler et ah, 1996a; Schwegler et ah, 1996b). We
found that IL-2 knockout (IL-2 KO) mice exhibited impaired learning and memory
performance, sensorimotor gating, and reductions in hippocampal infrapyramidal mossy
neuronal fiber length (Petitto et ah, 1999), a factor shown previously to correlate
positively with spatial learning ability (Schopke et ah, 1991; Schwegler and Crusio,
1995; Schwegler et ah, 1988).
In the present study, we therefore sought to test the hypothesis that loss of IL-2
would result in abnormal neurodevelopment of septal cholinergic neurons that project to
the hippocampus. Since extensive data document that these neurons play a critical role in
learning and memory performance (Galey et ah, 1994; Leanza et ah, 1995), and given the
various in vitro neurotrophic and neuromodulatory effects of IL-2 on developing
septohippocampal cholinergic neurons, we postulated that IL-2 KO mice would have
fewer cholinergic neurons in the medial septum and vertical limb of the diagonal band of
Broca (MS/vDB) and a reduction in the cholinergic axonal density in the hippocampus.
To accomplish this goal, IL-2 KO and wild-type littermates were compared using
stereological techniques to count MS/vDB cholinergic somata stained with ChAT


77
JC Cassel, E Duconseille, H Jeltsch and B Will, 1997. The fimbria-fomix/cingular bundle
pathways: a review of neurochemical and behavioural approaches using lesions and
transplantation techniques. Prog Neurobiol 51, 663-716.
JP Chandler and KA Crutcher, 1983. The septohippocampal projection in the rat: an
electron microscopic horseradish peroxidase study. Neuroscience 10, 685-696.
Q Chang and PE Gold, 2004. Impaired and spared cholinergic functions in the
hippocampus after lesions of the medial septum/vertical limb of the diagonal band
with 192 IgG-saporin. Hippocampus 14, 170-179.
J Chappell, R McMahan, A Chiba and M Gallagher, 1998. A re-examination of the role
of basal forebrain cholinergic neurons in spatial working memory.
Neuropharmacology 37, 481-487.
TJ Collier and A Routtenberg, 1984. Selective impairment of declarative memory
following stimulation of dentate gyrus granule cells: a naloxone-sensitive effect.
Brain Res 310, 384-387.
JM Conner and S Varn, 1997. Developmental profile of NGF immunoreactivity in the
rat brain: a possible role of NGF in the establishment of cholinergic terminal fields
in the hippocampus and cortex. Brain Res Dev Brain Res 101,67-79.
JM Conner, D Muir, S Varn, T Hagg and M Manthorpe, 1992. The localization of nerve
growth factor-like immunoreactivity in the adult rat basal forebrain and
hippocampal formation. J Comp Neurol 319, 454-462.
CD Conrad and EJ Roy, 1993. Selective loss of hippocampal granule cells following
adrenalectomy: implications for spatial memory. J Neurosci 13, 2582-2590.
AK Cross and MN Woodroofe, 1999. Chemokines induce migration and changes in actin
polymerization in adult rat brain microglia and a human fetal microglial cell line in
vitro. J Neurosci Res 55, 17-23.
P Curzon, DJ Kim and MW Decker, 1994. Effect of nicotine, lobeline, and
mecamylamine on sensory gating in the rat. Pharmacol Biochem Behav 49, 877-
882.
JD Cushman, Z Huang, G Ha, R Beck, J Lo and JM Petitto, 2004. Contextual fear
discrimination learning and memory is disrupted in IL-2 knockout mice. Society of
Neuroscience Abstracts.
CB Davis and SD Gillies, 2003. Immunocytokines: amplification of anti-cancer
immunity. Cancer Immunol Immunother 52, 297-308.


23
Figure 2-5, the distance across the IP-GCL was significantly reduced in the IL-2 KO
mice compared to wild-type mice (F( 1,16)=9.2, p=0.008). The distances across the Hil
and Mol, however, were not significantly different. The SP blade of the DG was
separated into three regions: the Mol, the SP-GCL, and the Hil (Figure 2-4c). There were
no significant differences in length between groups across any of the internal blade
regions.
Distance Across GrDG of the External Blade
35 -
Fig. 2-5. Distance across the GrDG of the external blade was significantly decreased in
IL-2 knockout mice compared to wild-type mice. Each bar represents the mean
SEM of nine animals per group. *p=0.008.
Discussion
These data are the first to demonstrate that loss of endogenous IL-2 results in
reduction in the number of MS/vDB cholinergic neurons and structural alterations in the
morphology of the dentate gyrus. Given the role of septohippocampal cholinergic
neurons in learning and memory (Galey et al., 1994; Leanza et al., 1995; Schwegler et ah,
1996a; Schwegler et ah, 1996b), sensorimotor gating (Caine et ah, 1992; Curzon et ah,


17
were outlined at lOx magnification and the size of the counting boxes were generated to
be approximately 5% of the most rostral, and therefore, smallest, area of the MS/vDB
(defined by the section where the corpus collosum first joins in the midline). The size of
the outlined count regions, but not the counting box, varied depending on where the
individual section was taken from the rostral to caudal extent of the MS/vDB. The
defined counting box was approximately 2-2.5% of the outlined count area of the largest
single section of the MS/vDB.
To assess whether the predicted septal cholinergic alterations in IL-2 KO mice
might be associated with a general effect on cholinergic neurons in the brain, striatal
cholinergic somata were also counted in the right hemisphere in the sections that also
contained the MS/vDB. Except for a different magnification used to outline the striatum
(5X), the sampling parameters were identical to those used to generate estimates of septal
cholinergic neuron number. The guard volume was set at 1 pm for the top and bottom of
the section and the counting cubes were randomly distributed throughout the user-defined
count area with a total sampling frequency of 25% (-8.3% of the total area since every
third section was sampled). Only somata that were clearly and distinctly stained were
counted. Each counting box was examined at 40x magnification (20x for striatal
neurons) and the computer-assisted focus was used to scan from the top to the bottom of
the counting box. Cells were counted only if they were either completely inside of the
counting box, or partially inside of the box on the top, back, or left side. They were not
counted if they fell outside of the box or crossed into the box anywhere on the bottom,
front, or right side. The cells that were counted were labeled on the monitor by clicking
the mouse pointer on each cell and the MCID software recorded the number of marks.


56
A) B)
Enhanced Cytokine Concentrations In the Serum of Enhanced Cytokine Concentrations In the Serum of
IL-2 KO Animals IL-2 KO Animals
MIP-1 alpha
1 1 Wild-Type
IL-6
1 1 Wild-Tyoe
wau IL-2 knockout
1 IL-2 knockout
Figure 4-2. Relative cytokine profile of IL-2 knockout mouse serum does not match brain
profile. A) IL-6 is increased in IL-2 knockout mice compared to wild-type; and
B) the chemokine MIP-1 a is higher in IL-2 knockout mice than wild-type.
*p<0.05; **p<0.01.
Alterations in Neurogenesis
Although there was not a significant effect of group on neurogenesis in either the
IP-GCL or SP-GCL, as depicted in Figure 4-3 and 4-4 respectively, there was a
significant group by gender interaction in both the IP-GCL (F(l,19)=4.71, p=0.043) and
SP-GCL (F(l,19)=6.43, p=0.02). Fishers least significant difference post hoc analysis
test confirmed a difference between male IL-2 KO and wild-type mice in the IP-GCL
(p=0.025) and SP-GCL (p=0.014), but not between female groups in either region. There
was no significant effect of group or group by gender interaction in cells around either
the IP-GCL or SP-GCL labeled with the oligodendrocyte marker, CNPase (data not
shown). Similarly, no significant effect of group or group by gender interaction was
evident in either the IP or SP-GCL in cells labeled with the astrocyte marker, GFAP (data
not shown).


63
nucleus and we speculate an involvement of IL-15. Though IL-15 has not been linked to
neurogenesis, its antiapoptotic properties on immune cells are well-studied (Lauwerys et
al., 2000; Waldmann, 2002; Waldmann et al., 2001). If these antiapoptotic actions of IL-
15 can also promote survival of neurons, then this may provide a hypothesis of why the
IL-2 KO male mice have higher levels of neurogenesis than the wild-types. Some
cytokines are capable of promoting the development of neural stem cells (Rozental et al.,
1995; Shah et al., 1996; Wong et al., 2004), and this may account for the how they
influence neurogenesis. The actual mechanism whereby IL-2, sex hormones, and
neurogenesis may interact is as of yet undefined and requires future study.
In summary, T cells and peripheral cytokines do not appear to enter into the
hippocampi of IL-2 KO mice, which does not support the hypothesis that the CNS
alterations previously seen in IL-2 KO mice are due to peripheral autoimmunity. Other
potential immune indicators of autoimmunity (e.g., deposition of autoantibodies in the
brain) were not addressed in this dissertation and require more research. Genetic deletion
of IL-2 may, however, alter the neuroimmunological status of the mouse hippocampus
through a dysregulation of cytokines produced by CNS cells (e.g., microglia, astroglia).
Further studies will be required to determine how these changes impact hippocampal
cytoarchitecture and function in IL-2 KO mice.


50
CTAGGCCACAGAATTGAAAGATCT-3 and 5-
GTAGGTGGAAAATTCTAGCATC ATCC-3 .
CD3+ T cells and MHC II+ Microglia Immunohistochemistry
For quantitative assessment of autoimmunity, three IL-2 KO and three wild-type
brains were processed for MHC II (an activated microglial marker) and CD3 (a pan T cell
marker) immunohistochemistry. Each animal was anesthetized with an injection cocktail
of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml) at a dose
of 0.015 ml injection cocktail/g body weight and perfused with 0.9% saline followed by
4% paraformaldehyde in phosphate buffered saline (PBS). The brains and spleens were
removed and fixed overnight in 4% paraformaldehyde, followed by overnight
equilibration in 30% sucrose cryoprotective solution, and then were snap frozen in
isopentane (-80C) for storage. The spleens were weighed to assay for relative
splenomegaly of IL-2 KO vs. wild-type mice The brains were equilibrated to -20C
prior to cryostat sectioning into 50 pm slices in the coronal plane, collected into
individual wells of polystyrene 24-well plates (NUNC 1147), and stored free-floating at
4C in PBS for histochemistry. Every third section was processed for MHC II or CD3
immunohistochemistry.
Free-floating 50-pm sections were incubated for 20 minutes in 1% hydrogen
peroxide (H2O2) to quench endogenous peroxidative activity. The sections were then
washed and blocked for 1 hr in 200 pl/well 3% normal goat serum (NGS). After this
incubation, the sections were incubated overnight in the primary antibody, rat anti mouse
CD3 (BD PharMingen; 1:500 in PBS with 0.3% Triton X-100 and 1% NGS) or rat anti
mouse MHC II (BD PharMingen; 1:500 in PBS with 0.3% Triton X-100 and 1% NGS).
The next day, the sections were washed and incubated overnight in the secondary


ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE IN
INTERLEUKIN-2 KNOCKOUT MICE
By
RAY D. BECK, JR.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2004


7
In the basal forebrain, the medial septum and vertical limb of the diagonal band of
Broca (MS/vDB) send a large number of projections to the hippocampus, with the major
neuronal phenotypes of these being cholinergic and GABAergic (Brashear et al., 1986;
Kiss et al., 1990b; Kiss et al., 1990a). The septohippocampal system has been associated
with learning and memory processes, with extensive data existing that link the septal
cholinergic neurons that project to the hippocampus to learning and memory (Galey et
al., 1994; Leanza et al., 1995), and PPI (Koch, 1996b). Moreover, variability of
cholinergic fiber density in the dentate gyrus of individual mouse strains correlate with
changes in spatial learning (Schwegler et al., 1996a; Schwegler et al., 1996b). Some
controversy exists, however, on the relative importance of cholinergic neurons of the
MS/vDB in learning and memory processes. The advent of selective toxins that target
cholinergic neurons, like 192 IgG-saporin, have allowed researchers to behaviorally test
animals only lacking MS/vDB cholinergic neurons, but with presumably normal
distributions of GABAergic neurons. In many of these studies, animals with cholinergic
septohippocampal lesions did not differ from control subjects (Baxter et al., 1996; Bizon
et al., 2003; Cahill and Baxter, 2001; Chappell et al., 1998; Perry et al., 2001).
Surprisingly, however, multiple other contemporary studies utilizing 192 IgG-saporin do
find learning and memory deficits in the lesioned animals (Janis et al., 1998; Johnson et
al., 2002; Lamprea et al., 2000; Wrenn et a!., 1999). One potential explanation for this
discrepancy may be that a certain threshold of cholinergic damage is necessary to elicit a
deterioration in spatial learning ability (Leanza et al., 1995; Wrenn et al., 1999).
Nevertheless, the negative findings are considerable and this may not explain the
inconsistency well enough between groups. Another hypothesis calls into question the


81
I Horak, 1996. Interleukin-2-knockout mice: a new model to study autoimmunity and
self-tolerance. Sb Lek 97,25-28.
AA Hurwitz, WD Lyman and JW Berman, 1995. Tumor necrosis factor alpha and
transforming growth factor beta upregulate astrocyte expression of monocyte
chemoattractant protein-1. J Neuroimmunol 57, 193-198.
LS Janis, MM Glasier, Z Fulop and DG Stein, 1998. Intraseptal injections of 192 IgG
saporin produce deficits for strategy selection in spatial-memory tasks. Behav Brain
Res 90, 23-34.
DA Johnson, NJ Zambn and RB Gibbs, 2002. Selective lesion of cholinergic neurons in
the medial septum by 192 IgG-saporin impairs learning in a delayed matching to
position T-maze paradigm. Brain Res 943, 132-141.
P Kaiser, L Rothwell, D Vasicek and K Hala, 2002. A role for IL-15 in driving the onset
of spontaneous autoimmune thyroiditis? J Immunol 168, 4216-4220.
S Karanth, K Lyson and SM McCann, 1993. Role of nitric oxide in interleukin 2-induced
corticotropin-releasing factor release from incubated hypothalami. Proc Natl Acad
Sci U S A 90, 3383-3387.
M Kerschensteiner, E Gallmeier, L Behrens, VV Leal, T Misgeld, WE Klinkert, R
Kolbeck, E Hoppe, RL Oropeza-Wekerle, I Bartke, C Stadelmann, H Lassmann, H
Wekerle and R Hohlfeld, 1999. Activated human T cells, B cells, and monocytes
produce brain-derived neurotrophic factor in vitro and in inflammatory brain
lesions: a neuroprotective role of inflammation? J Exp Med 189, 865-870.
MA King, BE Hunter, RL Reep and DW Walker, 1989. Acetylcholinesterase stain
intensity variation in the rat dentate gyrus: a quantitative description based on
digital image analysis. Neuroscience 33, 203-221.
J Kiss, AJ Patel and TF Freund, 1990a. Distribution of septohippocampal neurons
containing parvalbumin or choline acetyltransferase in the rat brain. J Comp Neurol
298, 362-372.
J Kiss, AJ Patel, KG Baimbridge and TF Freund, 1990b. Topographical localization of
neurons containing parvalbumin and choline acetyltransferase in the medial
septum-diagonal band region of the rat. Neuroscience 36, 61-72.
M Knipper and RJ Rylett, 1997. A new twist in an old story: the role for crosstalk of
neuronal and trophic activity. Neurochem Int 31, 659-676.
M Knipper, M da Penlia Berzaghi, A Blochl, H Breer, H Thoenen and D Lindholm, 1994.
Positive feedback between acetylcholine and the neurotrophins nerve growth factor
and brain-derived neurotrophic factor in the rat hippocampus. Eur J Neurosci 6,
668-671.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
ALTERATIONS OF SEPTOHIPPOCAMPAL STRUCTURE
IN INTERLEUKIN-2 KNOCKOUT MICE
By
Ray D. Beck, Jr.
August 2004
Chair: John M. Petitto
Major Department: Neuroscience
Interleukin-2 (IL-2) is a multifunctional cytokine involved in peripheral immune
processes and may also be implicated in multiple brain functions. IL-2 gene knockout
(IL-2 KO) mice exhibit deficits in several hippocampally-mediated behaviors (e.g.,
learning and memory) and have alterations in hippocampal structure.
In the first study, adult IL-2 KO and wild-type littermates were compared for
differences in the cholinergic neurons in the medial septum and vertical limb of the
diagonal band of Broca (MS/vDB; a structure associated with learning and memory).
The IL-2 KO mice had significantly fewer cholinergic somata in the MS/vDB, but not in
the striatum, thus indicating a selective effect of IL-2 on the MS/vDB. Cholinergic
neurite density in the hippocampus was unaffected, but the length across the
infrapyramidal (IP), but not the suprapyramidal (SP), granule cell layer (GCL) of the
dentate gyrus was reduced.
Vll


57
Neurogenesis in the Infrapyramidal Granule Cell Layer
500
4X1
IL-2 Knockout Wild-Type
i i Female
I Male
Figure 4-3. There is no effect of group on neurogenesis in the infrapyramidal granule cell
layer, but there is a group by gender interaction between IL-2 knockout and wild-
type mice (p=0.043). This effect appears to be attributable to differences in the
male mice (Fisher least significant difference test, p=0.025).
Neurogenesis in the Suprapyramidal Granule Cell Layer
500
4X -
1 1 '
Wild-Type IL-2 Knockout
1 I Female
I Male
Figure 4-4. There is no effect of group on neurogenesis in the suprapyramidal granule
cell layer, but there is a group by gender interaction between IL-2 knockout and
wild-type mice (p=0.02). This effect appears to be attributable to differences in
the male mice (Fisher least significant difference test, p=0.014).


74
DM Araujo, PA Lapchak, B Collier and R Quirion, 1989. Localization of interleukin-2
immunoreactivity and interleukin-2 receptors in the rat brain: interaction with the
cholinergic system. Brain Res 498, 257-266.
Y Arimatsu and M Miyamoto, 1991. Survival-promoting effect ofNGF on in vitro
septohippocampal neurons with cholinergic and GABAergic phenotypes. Brain Res
Dev Brain Res 58, 189-201.
DM Armstrong, G Bruce, LB Hersh and FH Gage, 1987. Development of cholinergic
neurons in the septal/diagonal band complex of the rat. Brain Res 433, 249-256.
WS Armstrong and P Kazanjian, 2001. Use of cytokines in human immunodeficiency
virus-infected patients: colony-stimulating factors, erythropoietin, and interleukin-
2. Clin Infect Dis 32, 766-773.
MB Atkins, MT Lotze, JP Dutcher, RI Fisher, G Weiss, K Margolin, J Abrams, M Sznol,
D Parkinson, M Hawkins, C Paradise, L Kunkel and SA Rosenberg, 1999. High-
dose recombinant interleukin 2 therapy for patients with metastatic melanoma:
analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 17, 2105-
2116.
IB Autenrieth, N Bucheler, E Bohn, G Heinze and I Horak, 1997. Cytokine mRNA
expression in intestinal tissue of interleukin-2 deficient mice with bowel
inflammation. Gut 41, 793-800.
H Awatsuji, Y Furukawa, M Nakajima, S Furukawa and K Hayashi, 1993. Interleukin-2
as a neurotrophic factor for supporting the survival of neurons cultured from
various regions of fetal rat brain. J Neurosci Res 35, 305-31 1.
AA Babcock, WA Kuziel, S Rivest and T Owens, 2003. Chemokine expression by glial
cells directs leukocytes to sites of axonal injury in the CNS. J Neurosci 23, 7922-
7930.
R Badolato, AN Ponzi, M Millesimo, LD Notarangelo and T Musso, 1997. Interleukin-15
(IL-15) induces IL-8 and monocyte chemotactic protein 1 production in human
monocytes. Blood 90, 2804-2809.
WA Banks, AJ Kastin and EG Gutierrez, 1994. Penetration of interleukin-6 across the
murine blood-brain barrier. Neurosci Lett 179, 53-56.
WA Banks, SA Farr and JE Morley, 2002. Entry of blood-borne cytokines into the
central nervous system: effects on cognitive processes. Neuroimmunomodulation
10,319-327.
WA Banks, L Ortiz, SR Plotkin and AJ Kastin, 1991. Human interleukin (IL) 1 alpha,
murine IL-1 alpha and murine IL-1 beta are transported from blood to brain in the
mouse by a shared saturable mechanism. J Pharmacol Exp Ther 259, 988-996.


76
A Braun, M Lommatzsch, A Mannsfeldt, U Neuhaus-Steinmetz, A Fischer, N Schnoy,
GR Lewin and H Renz, 1999. Cellular sources of enhanced brain-derived
neurotrophic factor production in a mouse model of allergic inflammation. Am J
Respir Cell Mol Biol 21, 537-546.
FC Bronfman, D Moechars and F Van Leuven, 2000. Acetylcholinesterase-positive fiber
deafferentation and cell shrinkage in the septohippocampal pathway of aged
amyloid precursor protein london mutant transgenic mice. Neurobiol Dis 7, 152-
168.
AI Brooks, MW Halterman and HJ Federoff, 1999. Focal hippocampal gain of NGF
function elicits specific septal cholinergic reorganization. Neuroreport 10, 337-344.
KM Buller, 2001. Role of circumventricular organs in pro-inflammatory cytokine-
induced activation of the hypothalamic-pituitary-adrenal axis. Clin Exp Pharmacol
Physiol 28, 581-589.
JF Cahill and MG Baxter, 2001. Cholinergic and noncholinergic septal neurons modulate
strategy selection in spatial learning. Eur J Neurosci 14, 1856-1864.
SB Caine, MA Geyer and NR Swerdlow, 1992. Hippocampal modulation of acoustic
startle and prepulse inhibition in the rat. Pharmacol Biochem Behav 43, 1201-1208.
JD Campbell, G Cook, SE Robertson, A Fraser, KS Boyd, JA Gracie and IM Franklin,
2001. Suppression of IL-2-induced T cell proliferation and phosphorylation of
STAT3 and STAT5 by tumor-derived TGF beta is reversed by IL-15. J Immunol
167,553-561.
M Canossa, O Griesbeck, B Beminger, G Campana, R Kolbeck and H Thoenen, 1997.
Neurotrophin release by neurotrophins: implications for activity-dependent
neuronal plasticity. Proc Natl Acad Sci USA 94, 13279-13286.
L Capuron, A Ravaud and R Dantzer, 2000. Early depressive symptoms in cancer
patients receiving interleukin 2 and/or interferon alfa-2b therapy. J Clin Oncol 18,
2143-2151.
L Capuron, A Ravaud and R Dantzer, 2001a. Timing and specificity of the cognitive
changes induced by interleukin-2 and interferon-alpha treatments in cancer patients.
Psychosom Med 63, 376-386.
L Capuron, A Ravaud, N Gualde, E Bosmans, R Dantzer, M Maes and PJ Neveu, 2001b.
Association between immune activation and early depressive symptoms in cancer
patients treated with interleukin-2-based therapy. Psychoneuroendocrinology 26,
797-808.
LL Carter, X Zhang, C Dubey, P Rogers, L Tsui and SL Swain, 1998. Regulation of T
cell subsets from naive to memory. J Immunother 21, 181-187.


2
the BBB with fenestrated capillaries (Buller, 2001). Peripheral leukocytes, in particular
activated T cells that enter the brain during certain conditions (e.g., EAE, facial nerve
axotomy), can also release cytokines in the CNS (Hickey et al., 1991). Finally, cytokines
may also interact with the brain through activation of peripheral nerves, such as IL-1
stimulation of the vagus nerve, which can lead to modulation of brain functions through
its afferent connections in the CNS (Maier et ah, 1998). Such a cytokine-to-nerve
communication pathway may not be limited to the vagus nerve, as central hyperalgesic
effects are also observed by stimulating cutaneous nerves with a subcutaneous injection
of IL-1 P (Fukuoka et ah, 1994), TNF-a (Sorkin et ah, 1997), or antibodies against TNF-
a (Lindenlaub et ah, 2000). Thus, multiple pathways exist that allow cytokines to
directly or indirectly influence the brain. The focus of this dissertation was on IL-2,
which can be produced in the periphery and the CNS.
The Pleiotropic Cytokine: Interleukin-2
IL-2 was originally identified as a growth factor for bone marrow-derived T cells
in 1976 (Morgan et ah, 1976), and was renamed in 1979, when its pleiotropic effects
between leukocytes (thus the term interleukin) became clear (Aarden et ah, 1979).
Further characterization of IL-2 revealed that it belongs to the four a-helix bundle family
of cytokines; this family consists of cytokines with four a-helices connected by three
loops in an up-up-down-down formation (Bazan, 1992). The receptor for IL-2 has a
common gamma (yc) subunit shared by multiple cytokines including IL-4, IL-7, IL-9. and
IL-15 (Sugamura et ah, 1996); a p subunit only shared with IL-15 (Giri et ah, 1995); and,
in one conformation, an a subunit, which confers greater binding affinity (Leonard et ah,
1984). The receptor subunits can combine in two biologically active forms: a lower


20
inner Mol) and subregions (SO, stratum radiatum (SR), dentate outer Mol), and derived
values for absolute and relative positions of, and distances between, these landmarks. For
each animal, three hippocampal slices were measured on both the left and right
hemispheres of the brain and the measured intensities or distances were averaged together
for statistical analysis.
Cresyl Violet Staining
Every third section was Nissl stained to provide a qualitative view of the
boundaries between various forebrain regions. The sections were placed on slides and
allowed to air-dry. The slides were immersed in 60C cresyl violet for 45 sec, washed in
running distilled water to remove the excess cresyl violet, dehydrated in graded ethanol,
cleared in xylenes, and coverslipped.
Results
Comparison of Cholinergic Somata in the MS/vDB
Figure 2-3 shows the total number of ChAT-positive cells stereologically counted
from the MS/vDB of IL-2 KO and wild-type mice. As seen in this figure, the IL-2 KO
mice had approximately 26% fewer cholinergic somata in this region than wild-type
controls. An ANOVA confirmed that this group difference was statistically significant
(F(l,16)=8.6, p=.01). By contrast, the number of ChAT-positive somata in the striatum
of IL-2 KO and wild-type mice were not different. There were no significant gender
differences in either brain region.
Density of AChE-positive Fibers in Regions of the Hippocampus
The average AChE-staining intensity curves were generated by defining a region
across CA1, CA3b, and the SP and IP blades of the DG. Figure 2-4 shows the intensity
curves that were generated from imaging the AChE-histochemically stained sections for


36
The MCID software interpolated the total volume of the MS/vDB based on the
volume of the count areas defined by the user. Cell density (Nv) was estimated by
dividing the total number of cells counted by the volume of the counting boxes, which
was also tracked by the software. The total cell number was estimated by multiplying Nv
by the total volume.
Enzvme-Iinked Immunosorbent Assay (ELISA) Characterization of NGF and
BDNF
For measurement of BDNF in hippocampus and MS/vDB, nine IL-2 KO and
seven wild-type mice were used. For measurement of NGF protein levels, seven animals
per group were used. Initial test runs revealed that some, but not all, of the NGF protein
levels fell below the sensitivity of detection for the kit; therefore, the homogenized NGF
samples were spiked with 25 pg/ml of the known NGF standard included with the kit to
bring any low levels of expression above the kits 15.6 pg/ml lower detection limit. One
column of the ELISA plate was also run with only the 25 pg/ml standard spike to provide
a baseline and the data reported are corrected for this. Animals used for ELISA
characterization of neurotrophin levels only received saline perfusion and were not post-
fixed in parafonnaldehyde. The brains were removed, snap frozen, and then allowed to
equilibrate to -20 C. The brains were sectioned on a cryostat at -20-22 C at 400 pm
thickness and the MS/vDB and hippocampi were dissected with a 0.75 mm micropunch
on a -20 C freezing platform. The dissected tissue was weighed on a microgram scale,
and then transferred to 25 pi of homogenizing solution (50 mM Na/Na2 and 0.2% TX-
100 in H20 with Anti-protease Complete TM cocktail (Boehringer)) per mg of wet
weight tissue. The tissue was sonicated in the homogenizing solution for 30-sec on ice


39
Stereological Cell Count of Nissl-stained
Granule Cell Layers
1200
r 1 1 1
Infrapyramidal Suprapyramidal
I 1 Wild-Type
I IL-2 Knockout
Fig. 3-1. There was a significant reduction in infrapyramidal, but not suprapyramidal
granule cells in IL-2 knockout relative to wild-type mice. Each bar represents the
mean SEM of seven animals per group. *p=0.006.
A)
Enhanced NGF Levels In the Hippocampus
of IL-2 Knockout Mice
B)
Reduced BDNF Levels in the Hippocampus
of IL-2 Knockout Mice
Genotype
Genotype
Figure 3-2. There was a significant: A) increase in NGF and B) decrease in BDNF
protein levels in the hippocampus of IL-2 knockout compared to wild-type mice.
The NGF bars represent the mean SEM of 7 animals per group. The BDNF
bars represent mean SEM of 7 IL-2 knockout and 9 wild-type mice. *p<0.05.


5
IL-2 has been shown to have neurotrophic effects on cultured neurons from
several regions of the rat brain including the neocortex (Shimojo et al., 1993), cortex,
striatum, medial septum, and hippocampus (Awatsuji et ah, 1993). Moreover, in rat
hippocampal neuronal cultures, IL-2 enhances the length and branching of hippocampal
neurites and the morphology of these neurons (Sarder et ah, 1996; Sarder et ah, 1993).
Interestingly, altered levels of IL-2 expression have been detected in schizophrenia (for
reviews see Hanisch and Quirion, 1995a; Muller and Ackenheil, 1998), which is a
neurological disorder where altered morphology of hippocampal neurons is well
documented (for a review see Thune and Pakkenberg, 2000).
IL-2-like immunoreactivity has been localized to the hippocampal formation in rat
forebrain (Lapchak et ah, 1991), and detected in tissue extracts from rat and human
hippocampal tissue (Araujo et ah, 1989). In mouse brain, IL-2 mRNA has been found in
the hippocampus (Villemain et ah, 1991), and transcripts for this cytokine may be
expressed in rat astrocyte cultures as well (Eizenberg et ah, 1995). Our lab has cloned
and sequenced the full-length mouse brain cDNAs for IL-2Ra as well as the IL-
2/15RP and yc subunits, and has found that the sequences of the genes expressed by
lymphocytes and in brain are identical. We have also found that these genes are enriched
in the hippocampus and related limbic regions. Of particular relevance to IL-2 actions in
the hippocampus, in situ hybridization has shown that the IL-2/15RP and yc genes are
expressed by pyramidal and granule cell neurons (Petitto and Huang, 1994, 1995, 2001;
Petitto et ah, 1998).


75
MG Baxter, DJ Bucci, TJ Sobel, MJ Williams, LK Gorman and M Gallagher, 1996.
Intact spatial learning following lesions of basal forebrain cholinergic neurons.
Neuroreport 7, 1417-1420.
SA Bayer, 1980. Development of the hippocampal region in the rat. I. Neurogenesis
examined with 3H-thymidine autoradiography. J Comp Neurol 190, 87-114.
JF Bazan, 1992. Unraveling the structure of IL-2. Science 257, 410-413.
RD Beck, Z Huang, GK Ha and JM Petitto, 2004. Altered Hippocampal Structure and
Neurotrophin Levels in Interleukin-2 Knockout Mice, (to be submitted for
publication).
RD Beck, Jr., MA King, Z Huang and JM Petitto, 2002. Alterations in septohippocampal
cholinergic neurons resulting from interleukin-2 gene knockout. Brain Res 955, 16-
23.
R Bender, M Plaschke, T Naumann, P Wahle and M Frotscher, 1996. Development of
cholinergic and GABAergic neurons in the rat medial septum: different onset of
choline acetyltransferase and glutamate decarboxylase mRNA expression. J Comp
Neurol 372, 204-214.
EN Benveniste and JE Merrill, 1986. Stimulation of oligodendroglial proliferation and
maturation by interleukin-2. Nature 321,610-613.
EN Benveniste, PK Herman and JN Whitaker, 1987. Myelin basic protein-specific RNA
levels in interleukin-2-stimulated oligodendrocytes. J Neurochem 49, 1274-1279.
M Besser and R Wank, 1999. Cutting edge: clonally restricted production of the
neurotrophins brain-derived neurotrophic factor and neurotrophin-3 mRNA by
human immune cells and Thl/Th2-polarized expression of their receptors. J
Immunol 162, 6303-6306.
M Bianchi and AE Panerai, 1993. Interleukin-2 enhances scopolamine-induced amnesia
and hyperactivity in the mouse. Neuroreport 4, 1046-1048.
M Bianchi, P Ferrario, N Zonta and AE Panerai, 1995. Effects of interleukin-1 beta and
interleukin-2 on amino acids levels in mouse cortex and hippocampus. Neuroreport
6, 1689-1692.
JL Bizon, JS Han, C Hudon and M Gallagher, 2003. Effects of hippocampal cholinergic
deafferentation on learning strategy selection in a visible platform version of the
water maze. Hippocampus 13, 676-684.
HR Brashear, L Zaborszky and L Heimer, 1986. Distribution of GABAergic and
cholinergic neurons in the rat diagonal band. Neuroscience 17, 439-451.


CHAPTER 3
ALTERED HIPPOCAMPAL STRUCTURE AND NEUROTROPHIN LEVELS IN
INTERLEUKIN-2 KNOCKOUT MICE
Introduction
Interleukin-2 (IL-2) has been implicated in the pathogenesis of multiple sclerosis
and several major neuropsychiatric disorders such as Alzheimer's disease, schizophrenia,
and Parkinson's disease (Hanisch and Quirion, 1995b). Furthermore, in case studies of
humans receiving IL-2 treatment for cancer therapy, prolonged exposure to IL-2 was
found to induce cognitive dysfunction and other untoward neuropsychiatric side effects
(Denicoff et al., 1987). Although IL-2 has been shown to be capable of modulating
different aspects of central nervous system (CNS) function, many of its known effects in
the limbic system occur in the hippocampal formation, where receptors for this cytokine
are enriched (Araujo et al., 1989; Hanisch and Quirion, 1995a; Lapchak et al., 1991;
Petitto and Huang, 1994, 2001; Petitto et al., 1998). IL-2 may, for example, modify
cellular and molecular substrates of learning and memory such as long-term potentiation
(Tancredi et al., 1990), and can affect multiple parameters of cognitive behavioral
performance in animals (Bianchi and Panerai, 1993; Hanisch et al., 1997a; Lacosta et al.,
1999; Nemni et al., 1992). IL-2 can provide trophic support to primary cultured neurons
from multiple region of the rat brain, including the hippocampus and medial septum
(Awatsuji et al., 1993; Sarder et al., 1993), and positively affects the morphology of
neurite branching from rat hippocampal cultures (Sarder et al., 1996; Sarder et al., 1993).
Furthermore, IL-2 has been shown to be one of the most potent modulators of
29


51
antibody, biotinylated goat anti-rabbit IgG (Sigma B-7389; 1:1000 dilution in PBS with
0.3% TX-100 and 1% NGS). The sections were then washed and incubated in
ExtrAvidin (Sigma E-2886; 1:1000 in PBS) for 2 hrs. The sections were developed in
0.5 mg/ml 3,3-diaminobenzidine (DAB), 0.2 mg/ml urea H2O2 for approximately 5 min
and were placed on slides, dehydrated in graded ethanol washes, cleared in two changes
of xylenes, and coverslipped.
Preparation of Serum and Brain Tissue for Cytokine Analysis
Hippocampal homogenates were analyzed from eight IL-2 KO and nine wild-type
mice to measure cytokine levels in the hippocampus. From these subject groups, serum
was collected from a smaller subset of animals (five IL-2 KO and seven wild-type mice)
for comparative analysis of brain vs. peripheral cytokine levels. Animals used for
characterization of endogenous cytokine levels were anesthetized with an injection
cocktail of 3:3:1 ketamine (100 mg/ml): xylazine (20 mg/ml): acepromazine (10 mg/ml)
at a dose of 0.015 ml injection cocktail/g body weight. Whole blood was collected by
puncturing the right atrium of the heart and inserting heparanized micro-hematocrit
capillary tubes (Fisher Scientific). The animals were then saline perfused, but were not
post-fixed in paraformaldehyde. The whole blood was centrifuged in Microtainer Brand
serum separator tubes (Becton Dickinson) at 5,000 rpm for 10 minutes to isolate serum
and the serum was stored at -80 C until used for Luminex analysis. The brains were
removed, snap frozen, and then allowed to equilibrate to -20 C. The brains were
sectioned on a cryostat at -20-22 C at 400 pm thickness and the hippocampi were
dissected with a 0.75 mm micropunch on a -20 C freezing platform. The dissected
tissue was weighed on a microgram scale, and then transferred to 25 pi of homogenizing


WORKS CITED
LA Aarden, TK Brunner, JC Cerottini, J Dayer, AL de Week, CA Dinarello, G Di
Sabato, JJ Farrar, I Gery, S Gillis, RE Handschumacher, CS Henney, MK
Hoffmann, WJ Koopman, SM Krane, LB Lachman, I Lefkowitz, RI Mishell, SB
Mizel, JJ Oppenheim, V Paetkau, J Plate, M Rollinghoff, D Rosenstreich, AS
Rosenthal, LJ Rosenwasser, A Schimpl, HS Shin, PL Simon, KA Smith, H
Wagner, JD Watson, E Wecker and DD Wood, 1979. Revised nomenclature for
antigen-nonspecific T cell proliferation and helper factors. J Immunol 123, 2928-
2929.
R Ader and N Cohen, 1975. Behaviorally conditioned immunosuppression. Psychosom
Med 37, 333-340.
R Ader, N Cohen and D Bovbjerg, 1982. Conditioned suppression of humoral immunity
in the rat. J Comp Physiol Psychol 96, 517-521.
R Ader, D Felten and N Cohen (Eds), 2001. Psychoneuroimmunology, Third Edition
Edition. Academic Press, San Diego, CA.
L Adorini, 1999. Interleukin-12, a key cytokine in Thl-mediated autoimmune diseases.
Cell Mol Life Sci 55, 1610-1625.
RF Alderson, AL Alterman, YA Barde and RM Lindsay, 1990. Brain-derived
neurotrophic factor increases survival and differentiated functions of rat septal
cholinergic neurons in culture. Neuron 5, 297-306.
F Aloisi, R De Simone, S Columba-Cabezas, G Penna and L Adorini, 2000. Functional
maturation of adult mouse resting microglia into an APC is promoted by
granulocyte-macrophage colony-stimulating factor and interaction with Thl cells. J
Immunol 164, 1705-1712.
A Alonso and C Kohler, 1984. A study of the reciprocal connections between the septum
and the entorhinal area using anterograde and retrograde axonal transport methods
in the rat brain. J Comp Neurol 225, 327-343.
R Alonso, I Chaudieu, J Diorio, A Krishnamurthy, R Quirion and P Boksa, 1993.
Interleukin-2 modulates evoked release of [3H]dopamine in rat cultured
mesencephalic cells. J Neurochem 61, 1284-1290.
73


91
JQ Zheng, M Felder, JA Connor and MM Poo, 1994. Turning of nerve growth cones
induced by neurotransmitters. Nature 368, 140-144.


72
strains, and reconstitution allows us to control for several immunological and genetic
factors without disruption of the BBB.
Concluding Remarks
In the peripheral immune system, the complex interplay and interactions of the
various cytokines often have redundant, supportive, or even oppositional roles. This
complexity allows for a system of compensatory and regulatory control of the immune
response. Similar overlaps and checks and balances are also present in the CNS, so
understanding how these intricate systems interact can prove daunting. Nevertheless,
elucidating the relationship between the brain and immune system molecules may have
profound clinical utility. The IL-2 KO mouse is a complicated model to study a cytokine
brain interaction and, in this dissertation, we have attempted to simplify the model by not
experimentally manipulating it. Our goal was to lay groundwork for future studies on
this topic, whereby we hope to more fully understand the precise mechanisms by which
IL-2 alters brain physiology.


4
Interleukin-2 and the Brain
The effect of IL-2 on cognition and mood in humans was among the earliest
findings that suggested that this cytokine might have neurobiological actions. In early
clinical studies of the cognitive side effects of IL-2 therapy, 50% (i.e., 22 patients out of
44) of the subjects monitored developed cognitive changes, with 15 of them necessitating
acute intervention (Denicoff et al., 1987). In addition, IL-2 therapy in patients with renal
carcinoma or melanoma was found to impair spatial memory and performance in
planning tasks (Capuron et ah, 2001a), and induce depressive symptoms as early as two
days into therapy (Capuron et ah, 2000).
IL-2-brain interactions have also been investigated on an anatomical and
physiological level. In landmark studies, IL-2 was found to modulate the proliferation of
oligodendrocytes (Benveniste et ah, 1987; Benveniste and Merrill, 1986; Saneto et ah,
1986). Exogenously administered IL-2 also has multiple effects on pituitary cells
including stimulation of cortisol production and adrenal corticotropin releasing hormone
release (Hanisch et ah, 1994), as well as increasing pituitary cell responsiveness to
corticotropin-releasing hormone (Witzke et ah, 2003). IL-2 has also been shown to
regulate the production and secretion of peptides from hypothalamus, in addition to
pituitary cells (Karanth et ah, 1993; Lapchak and Araujo, 1993; Pardy et ah, 1993).
Subsequent research has shown that exogenously applied IL-2 can modulate other types
of central nervous system cells, such as microglia (Sakai et ah, 1995). Exogenously
applied IL-2 can also biphasically regulate the release of some neurotransmitters such as
dopamine (Alonso et ah, 1993; Petitto et ah, 1997), or acetylcholine (Hanisch et ah,
1993; Seto et ah, 1997).


13
immunohistochemistry, and image analysis methods to measure the density and
distribution of cholinergic neurites in several regions of the hippocampus labeled for
acetylcholine esterase (AChE), a reliable marker of cholinergic axons (Hedreen et al.,
1985).
Materials and Methods
Animals and Tissue Preparation
Mice used in these experiments were cared for in accordance with the NIH Guide
for the Care and Use of Laboratory Animals. Mice were bred in our colony using IL-2
heterozygote by IL-2 heterozygote crosses. The polymerase chain reaction (PCR) was
used to genotype the offspring post-weaning (see below). The IL-2 KO mice, obtained
originally from the NIH repository at Jackson Labs, were derived from ten generations of
backcrossing onto the C57BL/6 background. Mice were housed under specific pathogen-
free conditions. Animals used in these experiments were 8-12 weeks of age.
Each animal was anesthetized with sodium pentobarbital (50 mg/kg) and perfused
with 0.9% saline followed by 4% paraformaldehyde in phosphate buffered saline (PBS).
The brains were removed and fixed overnight in 4% paraformaldehyde followed by
overnight equilibration in 30% sucrose cryoprotective solution, and then were snap
frozen in isopentane (-80C) for storage. The brains were equilibrated to -20C prior to
cryostat sectioning into 40 pm slices in the coronal plane, collected into individual wells
of polystyrene 24-well plates (NUNC 1147), and stored free-floating at 4C in PBS for
histochemistry. Every third section was processed for ChAT immunohistochemistry,
AChE histochemistry, or cresyl violet Nissl staining.


42
smaller cell body size. Qualitative assessments of random granule cells, however, do not
support this hypothesis, though a more extensive study would be necessary to definitively
address that issue.
Though the receptors for IL-2 are more abundant in the hippocampus, including
the GCL of the DG (Petitto and Huang, 1994; Petitto et al., 1998), it is not clear whether
IL-2 may act directly on these neurons, or whether it upregulates other growth factors
like the neurotrophins. The observed differences in the level of the neurotrophin BDNF
was consistent with our hypothesis that we would find a reduction in trophic factors
important in MS/vDB and hippocampal development and maintenance. BDNF plays a
role in the maintenance and repair of septal cholinergic neurons (Alderson et al., 1990;
Morse et al., 1993; Ward and Hagg, 2000), can implement a positive feedback
mechanism with these neurons to enhance the release of ACh (Knipper et al., 1994), and
can also modulate neurogenesis (Larsson et al., 2002; Lee et al., 2002), thus potentially
impacting granule cell number. Thus, the reduction of cholinergic cell number in the
MS/vDB is consistent with a reduction in this trophic factor. The exact interaction
between IL-2 gene deletion and the reduction of BDNF levels remains unclear. Though
BDNF is expressed in the peripheral immune system by lymphocytes, IL-2 does not
stimulate its production or release. IL-2 can, however, upregulate the expression of
TrkB, the receptor for BDNF, in lymphocytes (Besser and Wank, 1999). Furthermore,
some evidence suggests that BDNF can stimulate a positive feedback mechanism of its
own production via the TrkB receptor in hippocampal neurons (Canossa et al., 1997;
Saarelainen et al., 2001). In IL-2 KO mice, the absence of IL-2 may therefore potentially
lead to a down-regulation of the TrkB receptor, thereby partially inhibiting the positive


43
feedback production of BDNF. Interestingly, the neurotrophin Trk receptors and IL-2
receptor share some of the same signal transduction pathways (e.g., mitogen activated
protein kinase or phosphatidylinositol 3-kinase), which appear to play a role in their
growth and survival promoting actions (for reviews see Gaffen, 2001; Patapoutian and
Reichardt, 2001). Whether IL-2 knockout leads to disruption of one of these signal
transduction pathways has not, to our knowledge, been elucidated and thus requires
further study.
Against our initial hypothesis, NGF protein levels were actually increased in the
IL-2 KO mice. Unlike BDNF, NGF does not appear to stimulate a positive feedback
neurotrophin release from hippocampal neurons (Canossa et al., 1997). Given the
reduction in cholinergic survival in the MS/vDB of IL-2 KO mice, the target neurons in
the hippocampus of these animals may produce higher protein levels of NGF as a
compensatory response. Similarly, moderate lesions of rat septohippocampal projections
lead to increased mRNA expression of NGF, but not BDNF in hippocampal target cells
(Hellweg et al., 1997).
In summary, cholinergic deficits seen in the MS/vDB of IL-2 KO mice appear to
be selective for cholinergic over GABAergic neurons. In addition, the loss of cholinergic
neurons in the MS/vDB may occur in the later stages of, or after, development of the
septohippocampal system, as the deficits are not seen in 21-day-old IL-2 KO mice. In the
hippocampus, the number of neurons in the IP-GCL is significantly reduced. A reduced
production of hippocampal BDNF may contribute to many of the aforementioned
changes, though NGF levels are increased in a possible compensatory response.
Although overt signs of autoimmunity in the brain are not apparent (we have been unable


65
potential failure of these neurons to survive, since 21-day-old IL-2 KO animals did not
show signs of a loss of cholinergic neurons of the MS/vDB. Furthermore, the neuronal
deficiencies appeared to be selective to the cholinergic neurons of the MS/vDB, as there
were no similar decreases of GABAergic neurons in IL-2 KO animals. In the
hippocampus, consistent with the reductions in length across the IP-GCL of IL-2 KO
compared to wild-type mice, the stereologically estimated cell count of Nissl-stained
neuronal somata of IL-2 KO mice was significantly lower than wild-type littermates.
Also, the hippocampal neurotrophin levels of BDNF and NGF were significantly
decreased and increased, respectively.
Finally, in Chapter 4, the cytokine profile of the IL-2 KO hippocampus was altered
with increases in IL-15 and IL-12 and the chemokines MCP-1 and IP-10. These
experiments confirmed our hypothesis that loss of IL-2 would result in increased levels of
brain IL-15 production. Furthermore, in spite of the known roles of these cytokines and
chemokines in T cell trafficking and modulation during brain insult, we were unable to
label for elevated levels of the T cells or activated microglia in the hippocampus. The
above cytokine profile did not match the serum cytokine levels, suggesting that the
changes in cytokines in the hippocampus were likely due to changes in their production
in the CNS. Finally, though there was no group alteration in adult GCL neurogenesis,
there was a group by gender interaction that appeared to be attributable to the male mice.
Implications
This series of studies is the first to demonstrate that endogenous levels of IL-2 may
be an important factor in the late development or survival of neurons in the CNS. This
impact on CNS neurons involved not only cell number, but also alterations in levels of
trophic factors and brain cytokines. Other studies involving IL-2 and the CNS, to date,


40
Discussion
These data are the first to demonstrate that the loss of endogenous IL-2 in
knockout mice can lead to alterations in neuronal cell number in the IP-GCL and
production of the neurotrophins, BDNF and NGF. Further, this study expands upon the
previous finding that IL-2 gene deletion leads to a deficiency of cholinergic neurons in
the MS/vDB (Beck et al., 2002), by showing a lack of significant cholinergic neuronal
differences in MS/vDB of 21-day-old IL-2 KO animals or GABAergic alterations in adult
animals.
The lack of a difference in GABAergic neurons in 8-12-week-old adult mice was
consistent with our initial hypothesis, as there is no evidence in the literature that IL-2 has
any modulatory effects on GABAergic neurons. Moreover, this is not a regional effect,
but rather appears to be selective to cholinergic projection neurons. As previously
mentioned, IL-2 is a potent modulator of ACh release (Hanisch et al., 1993; Seto et al.,
1997), and its precursor enzyme ChAT (Mennicken and Quirion, 1997), suggesting an
effect of IL-2 on cholinergic neurons. In GABAergic neurons, however, IL-2 has failed
to evoke release of GABA in mesencephalic neuronal cultures (Alonso et al., 1993), or
the cortex or hippocampus of mice (Bianchi et al., 1995). Since IL-2 deficiency does not
affect the number of GABAergic somata in the MS/vDB of IL-2 KO mice, the neuronal
loss appears to be selective for cholinergic neurons in the MS/vDB. Furthermore, we
previously found no differences in the striatal cholinergic neuronal number (Beck et al.,
2002), so the lack of IL-2 does not simply cause a general loss of all cholinergic neurons.
Against our initial hypothesis that 21-day-old IL-2 KO mice would have similar
cholinergic deficiencies as adult 8-12-week-old mice, there was no detectable loss of
cholinergic cell number in the MS/vDB. We did not examine 21-day-old mice for