Citation
Properties of high-affinity L-Glutamate transport in glial- and neuronal-enriched fractions from rat brain

Material Information

Title:
Properties of high-affinity L-Glutamate transport in glial- and neuronal-enriched fractions from rat brain Kellye K. Daniels
Creator:
Daniels, Kellye K., 1967-
Publication Date:
Language:
English
Physical Description:
vi, 117 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Brain ( jstor )
Enzymes ( jstor )
Neuroglia ( jstor )
Neurons ( jstor )
Phorbol esters ( jstor )
Phosphatases ( jstor )
Phosphorylation ( jstor )
Protein isoforms ( jstor )
Rats ( jstor )
Synaptosomes ( jstor )
Biological Transport -- physiology ( mesh )
Brain ( mesh )
Cell Aging ( mesh )
Department of Neuroscience thesis Ph.D ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Neuroscience -- UF ( mesh )
Glutamates -- metabolism ( mesh )
Glutamates -- physiology ( mesh )
Neuroglia ( mesh )
Neurons ( mesh )
Neurotransmitter Uptake Inhibitors -- pharmacokinetics ( mesh )
Phosphorylation ( mesh )
Protein Kinase C ( mesh )
Rats ( mesh )
Research ( mesh )
Subcellular Fractions -- analysis ( mesh )
Subcellular Fractions -- isolation & purification ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1997.
Bibliography:
Bibliography: leaves 99-116.
General Note:
Typescript.
General Note:
Vita.

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:
027921167 ( ALEPH )
48925988 ( OCLC )

Downloads

This item has the following downloads:


Full Text










PROPERTIES OF HIGH-AFFINITY L-GLUTAMATE TRANSPORT IN GLIAL- AND
NEURONAL-ENRICHED FRACTIONS FROM RAT BRAIN















By

KELLYE K. DANIELS


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


1997














This work is dedicated to Maurice and Alta Bennett, Maggie Daniels, and Louise Ferrell, with great love and fond remembrance.














ACKNOWLEDGMENTS

During my time as a graduate student at the University of Florida, Dr. William G. Luttge and Dr. Mary Jo Koroly gave me exceptional guidance and support; such generosity is rare and truly appreciated.

I extend much gratitude to the Center for the Neurobiology of Aging and to the Bryan W. Robinson Neurological Foundation, Inc., for financial support.














TABLE OF CONTENTS

page

ACKN OW LED GM EN TS ...................................................................................... iii

ABSTRACT ..................................................................................................... v

CHAPTERS

1 BACKGROUND AND SIGNIFICAN CE ............................................................ 1

2 ISOLATION OF GLIAL AND NEURONAL FRACTIONS ......................... 22

Introduction ............................................................................................... 22
Experimental Procedures ........................................................................ 24
Results .................................................................................................. 32
Discussion .............................................................................................. 47

3 EFFECT OF PHOSPHORLYATION ON L-[3H]GLUTAMATE UPTAKE ....... 53

Introduction .......................................................................................... 53
Experimental Procedures ........................................................................ 57
Results .................................................................................................. 62
Discussion .............................................................................................. 64

4 EFFECT OF AGING ON L-[3H]GLUTAMATE UPTAKE .......................... 75
Introduction ............................................................................................... 75
Experim ental Procedures ........................................................................ 77
Results ................................................................................................... 82
Discussion .............................................................................................. 8?

5 CON CLU SION S ............................................................................................ 93

REFEREN CES ................................................................................................ 98

BIOGRAPHICAL SKETCH ................................................................................ 116













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

PROPERTIES OF HIGH-AFFINITY L-GLUTAMATE TRANSPORT IN GLIAL- AND
NEURONAL-ENRICHED FRACTIONS FROM RAT BRAIN By

Kellye K. Daniels

December, 1997

Chairman: Thomas W. Vickroy, Ph.D. Major Department: Neuroscience

Prior studies have established that termination of L-glutamate-mediated

neurotransmission occurs through high-affinity transporters located on glial cells and neurons. For in vitro studies of L-glutamate transport, an improved three-step density gradient centrifugation technique was developed to isolate glial- (glial plasmalemmal vesicles or GPV) and neuronal- (synaptosomes or SYN) enriched vesicles from rat brain homogenates. Morphological, biochemical, and Western blot analyses confirmed the high degree of separation between GPV and SYN tissue fractions.

GPV and SYN tissues were used to evaluate L-glutanate uptake, to address phosphorylation / dephosphorylation events as possible regulatory mechanisms of Lglutamate uptake, and to determine to what extent aging may alter these mechanisms. L[3H]Glutamate uptake by GPV and SYN revealed remarkedly similar properties. Kinetic








studies indicated a more rapid initial rate for uptake in GPV, as well as a slightly higher substrate affinity. Conversely, SYN fractions exhibited a 2- to 3-fold greater capacity for L-[3H]gutamate uptake. In studies addressing the effects of phosphorylation on L[3H]glutamate uptake, GPV and SYN tissue fractions responded differentially following incubation with phorbol- 12,13-dibutyrate (PDBu), an activator of protein kinase C (PKC), an enzyme which phosphorylates serine/threonine residues. GPV L-[3H]glutamate uptake increased significantly above control values after incubation with PDBu, whereas uptake values in SYN fractions did not. Likewise, L-[3H]glutamate uptake in GPV significantly increased above control values when fractions were incubated with okadaic acid, a phosphoserine / phosphothreonine phosphatase inhibitor, whereas L-[3H]glutamate uptake in SYN fractions did not change from control levels. In GPV and SYN fractions obtained from animals aged 5, 25, 31, and 37 months, only L-[3H]glutamate uptake in SYN fractions from 37 month old animals significantly decreased from uptake values obtained from 5 month old animals. When fractions were incubated with PDBu, L-[3H]glutamate uptake increased dose-dependently across age groups in both fractions. This response in SYN fractions, however, decreased with increasing age.

This study has demonstrated that while both GPV and SYN fractions transport Lglutamate with high-affinity and in a saturable manner, they differ in response to kinases and phosphatases, ultimately indicating that phosphorylation / dephosphorylation events may more tightly regulate GPV rather than SYN L-glutamate transport and that SYN Lglutamate transport is more vulnerable to the aging process.














CHAPTER 1
BACKGROUND AND SIGNIFICANCE Properties of EAA Transporters

Biochemical Properties and Functions

The amino acids L-aspartate and L-glutamate represent the major excitatory

neurotransmitters in the mammalian brain (Krnjevic, 1970; Fonnum, 1984). High-affinity transporters for these excitatory amino acids (EAA) are essential for recapturing the neurotransmitter L-glutamate, thereby terminating its extracellular actions and maintaining a high concentration gradient between the extracellular space and the cytosol of nerve terminals (Kanner and Schuldiner, 1987; Nicholls and Attwell, 1990). The ability of L-glutamate to cause neuronal damage and cell death, however, poses a special requirement for an efficient uptake system (Choi, 1988, 1994).

For several years it has been known, primarily through the work of Kanner and

collaborators, that L-glutamate uptake is electrogenic and dependent on external Na and internal K' ions (Kanner and Schuldiner, 1987; Danbolt et al., 1990). Detailed studies on the stoichiometry of L-glutamate uptake have been carried out on Muller cells, a specialized form of glial cells (Attwell el al., 1991; Bouvier et al., 1992). Since these cells do not express L-glutamate receptors, the current elicited by L-glutamate provides a convenient index of electrogenic uptake. Analyses based on whole cell patch clamp recordings indicate that L-glutamate (carrying one net negative charge at physiological










pH) is transported into the cell together with two sodium ions and that the return of the carrier to the outside of the cell is coupled with an outward transport of one potassium ion and one hydroxyl anion.

The maintenance of a low level of external L-glutamate is one obvious function of the L-glutamate transporter. This function is crucial as L-glutamate becomes neurotoxic when its extracellular concentration exceeds a certain level (Choi et al., 1987, Beal, 1992a, b). The concentration at which L-glutamate exerts excitotoxic actions is hard to determine, due to the efficiency of uptake and the lack of noncompetitive, irreversible blockers. Estimates range from several hundred micromolar (Nicholls and Attwell, 1990) to as low as I micromolar (Frandsen and Schousboe, 1990). Neuronal toxicity is a consequence of excessive membrane depolarization associated with swelling and an accumulation of calcium recruited from the extracellular space or from intracellular stores. The importance of glial L-glutamate transporters for reducing L-glutamate neurotoxicity is suggested by experiments showing that neurons are much more sensitive to L-glutamateinduced toxicity when grown alone than when grown in co-culture with glia (Rosenberg et al, 1992). In conditions such as hypoxia and ischemia, the driving force for L-glutamate uptake collapses due to the breakdown of the electrochemical sodium and potassium gradients, and L-glutamate transport is compromised or reversed (Szatkowski and Attwell, 1994). In fact there is increasing evidence to suggest that the extracellular overflow of L-glutamate in ischemic and anoxic conditions primarily reflects inadequate or reversed L-glutamate transport and that exocytotic release plays a minor role (Nicholls and Attwell, 1990).








3

L-Glutamate dysfunction has been implicated in other neurologic disorders such as amyotrophic lateral sclerosis (ALS) (Plaitakis, 1990; Rothstein et al., 1992; Shaw el al., 1994; Rothstein et al., 1995; Bristol and Rothstein, 1996; Leigh and Meldrum, 1996). High-affinity, sodium-dependent transport of L-glutamate was found to be markedly impaired in ALS patients in the spinal cord, motor cortex, and somatosensory cortex (Rothstein et al., 1992). These are the brain regions most affected in ALS. The changes were specific for ALS, as they were not observed in Huntington's disease, Alzheimer's disease, or nonneurologic disease controls. This finding led Rothstein and colleagues (1992) to conclude that defects in the clearance of extracellular L-glutamate, secondary to faulty transporter function, could lead to neurotoxic levels of extracellular L-glutamate and thus be pathogenic in ALS.

Regional / Cellular Localization

Three sodium-dependent transporter subtypes have been identified in rat forebrain: L-glutamate/L-aspartate transporter (GLAST) (Storck et al., 1992), glial transporter- I (GLT- 1) (Pines et al., 1992), and excitatory amino acid carrier-I (EAAC 1) (VelazFaircloth et al., 1996). Three similar transporters have been cloned from human brain, and the provisional human homologs have been termed excitatory amino acid transporters (EAAT) - 1, -2, and -3, corresponding with greater than 90% sequence homology to GLAST, GLT-1, and EAAC1, respectively (Arriza etal., 1995). A fourth subtype, EAAT4, has been identified in human cerebellum (Fairman et al., 1995), and a fifth subtype, EAAT5, has been identified in human retina (Arriza et al., 1997).










With regard to the cellular and regional locations of L-glutamate transporters,

results from immunocytochemistry (Lehre et al., 1995) and in situ hybridization (Torp et al., 1994) indicate that there are pronounced regional differences in the expression of the transporters. These studies have mapped GLAST mRNA to the Purkinje cell layer of the cerebellum, apparently associated with Bergmann glia (Storck et al., 1992), as well as to the cortex and hippocampus (Rothstein et al., 1994). Other studies, using in situ hybridization for the carboxyl-terminal domains of each transporter, initially reported a low level of GLAST immunoreactivity in both neurons and astroglia (Rothstein et al., 1994). Amino-terminal oligopeptide antibodies localized GLAST to astrocytes (Rothstein e al., 1995; Lehre et al., 1995; Schmitt et al., 1997). In addition, GLT-1 is present in astrocytes throughout the brain, with predominant expression in telencephalic structures, including the hippocampus, neocortex, and striatum (Danbolt et al., 1992; Rothstein et al., 1994). However, both radioactive as well as non-radioactive in situ hybridization techniques have shown that selected neuronal populations (i.e., thalamus, hypothalamus, and pyramidal cells of the hippocampus) express GLT- I mRNA (Torp et al., 1994; Schmitt et al., 1996). The observations that GLT-1 rmRNA is expressed in or by neurons seems to be at variance with other studies (Danbolt et al., 1992; Rothstein et al., 1994; Lehre et al., 1995) which failed to detect GLT- 1 protein in immunoelectron microscopic preparations. Schmitt and colleagues (1996) maintain that a failure to detect GLT-1 protein in neuronal membranes may be a result of the protein concentration being below the method's detection limits and that at the electron microscopic level the extremely










dense immunoreactivity of the neuropil in general makes it impossible to distinguish whether labeled membranes belong to neuronal terminations or to glial cells.

EAACI mRNA is abundantly expressed in the pyramidal layer of the hippocampus (regions CA1-CA4), the granule cell layer of the dentate gyrus, the granule cell layer of cerebellum and layers II-IV of cerebral cortex (Kanai and Hediger, 1992). In their original report, Kanai and Hediger (1992) concluded on the basis of the distribution of hybridizing mRNA that this transporter was primarily neuronal. Findings by others (Rothstein et al., 1994; Velaz-Faircloth et al., 1996) support this conclusion. The cellular locations of the more recently cloned EAAT4 (Fairman et al., 1995), which is expressed in the cerebellum, and EAAT5 (Arriza et al., 1997), which is expressed in the retina, have not been published.

In a study by Rothstein and colleagues (1995), anti-oligopeptide antibodies generated to carboxyl- and amino-terminal sequences, specific for each human Lglutamate transporter, were used to determine the nature of the L-glutamate transport defect in ALS. The conclusion of this study was that a remarkable loss of astroglial GLTI immunoreactive protein, restricted to motor cortex and spinal cord, was responsible for the defect. Notably, GLAST, also localized to glia, was unaffected. EAAC I immunoreactive protein was affected only modestly. When mRNA levels of control versus ALS tissue was examined, there were no quantitative changes in mRNA for EAATI, EAAT2, or EAAT3 in ALS motor cortex, even in patients with a large loss of










EAAT2 protein and decreased tissue L-glutamate transport (Bristol and Rothstein, 1996). These studies suggest that abnormalities in EAAT2 may be due to translational or posttranslational processes.

Regulatory Mechanisms

While many of the mechanisms that influence L-glutamate transporters in vivo are probably unknown, several factors such as arachidonic acid (AA) (Barbour et al., 1989), nitric oxide (NO) (Pogun and Kuhar, 1993), and transporter phosphorylation (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997) have been noted as possible regulators of transporter function. There is evidence that AA effects Lglutamate uptake in glial cells and nerve terminals (Barbour et al., 1989; Zerangue et al., 1995). In Xenopus oocytes injected with cRNAs encoding the human excitatory transporters EAAT 1-3, micromolar levels of AA significantly reduced L-glutamate uptake mediated by EAAT 1, while transport mediated by EAAT2 increased more than two-fold (Zerangue et al., 1995). AA had no effect on EAAT3 transport (Zerangue et al., 1995). These differential effects could constitute a regulatory mechanism, but would be nonselective as several sodium-dependent uptake systems are modulated (i.e., inhibited) by AA (Rhoads et al., 1983), including the uptake systems for y-aminobutyric acid (Chan et a., 1983) and glycine (Zafra et al, 1990). Nevertheless, AA-mediated regulation of Lglutamate transport may have physiological relevance during the induction of long term potentiation, which is associated with an increased production of AA (Bliss and Collingridge, 1993). It has previously been reported (Herrero et al, 1992) that AA may increase synaptic strength by stimulating L-glutamate release, acting presynaptically in










concert with metabotropic L-glutamate receptors. A simultaneous inhibition of Lglutamate uptake would have a synergistic effect. Interestingly, L-glutamate uptake has been found to be inhibited also by NO (Pogun and Kuhar, 1993; Pogun el al., 1994), which, like AA, has been proposed to act as a "retrograde messenger" during the induction of LTP. NO is thought to be a mediator of L-glutamate-induced neurotoxicity because inhibitors of NO synthase prevent L-glutamate neurotoxicity (Dawson et al., 1991, 1993).

Each of the cloned L-glutamate transporters contains consensus sequences for

phosphorylation by protein kinase C (PKC) (Kanai and Hediger, 1992; Pines et al., 1992; Storck et al., 1992; Fairman et al., 1995). The existence of putative phosphorylation sites indicates that L-glutamate transporters may be regulated by protein kinases and phosphatases. The finding that L-glutamate transport activity (V.,, but not K) is increased in primary astrocyte cultures, after incubation of the cells with phorbol esters (PKC activators) (Casado et al., 1991), suggests that the putative phosphorylation sites are physiologically relevant. The first direct evidence that PKC-mediated phosphorylation is involved in the regulation of L-glutamate transport was provided by Casado et al. (1993). Using antibodies directed against biochemically purified GLT-1, this group demonstrated that the amount of phosphorylation correlated with phorbol ester-induced stimulation of L-[3H]glutamate transport activity. Additionally, by site-directed mutagenesis of GLT-1 and subsequent transfection into HeLa cells, it was shown that the effect of the phorbol ester was dependent on serine 113, the likely biologically relevant phosphorylation site.










Other groups have further established that phosphorylation of brain L-glutamate transporters modulates transport function (Dowd and Robinson, 1996; Conradt and Stoffel, 1997). Preincubation of a subline of C6 glioma cells with phorbol ester, which endogenously express EAAC 1-mediated L-glutamate transport, caused a significant increase in L-glutamate transport (Dowd and Robinson, 1996). The findings that PKC activators increasing L-glutamate transport in GLT- 1 and EAAC I are in direct contrast with results reported by Conradt and Stoffel (1997) for GLAST. In this case treatment of GLAST-expressing cells with phorbol ester decreased L-glutamate transport activity with phosphorylation occurring at a non-PKC consensus site.

The fact that L-glutamate transport appears to be subject to elaborate regulatory mechanisms strongly suggests an important role for L-glutamate transporters in brain function, possibly through mechanisms that remain to be discovered.

Glial Plasmalemmal Vesicles and Synaptosomes Prior studies have established that termination of L-glutamate-mediated excitatory transmission in the mammalian central nervous system occurs through high-affinity transporters localized to both glial cells (i.e., astrocytes and oligodendrocytes) (Schrier and Thompson, 1974) and neurons (Iversen, 1973; Kuhar, 1973). While L-glutamate transporter processes of neurons and astrocytes have been studied in brain slices, cultured cells and rat brain membrane preparations, a verified detailed comparative biochemical analysis of the neuronal and astroglial L-glutamate transporter from the same brain region under identical conditions is still lacking. The three-step density gradient centrifugation method outlined in Chapter 2 will not only allow for the characterization of L-glutamate










transport in membrane-encapsulated vesicles derived from separated glial and synaptosomal populations from the same brain region at the same time, but will also circumvent many of the problems inherent to other techniques. For example, whereas in vivo situations as well as brain slices maintain circuitry, this greatly complicates the understanding of the degree to which glial versus neuronal populations contribute to Lglutamate uptake and negate any ability to ascertain the selectivities of inhibitors for glial versus neuronal transporters. The convenience of a density gradient centrifugation method is greater than that of primary cell culture procedures in that same-day animal sacrifice yields same-day results. Tissue does not have to be harvested and grown in media for varying lengths of time prior to use, plus an animal of any age may be used in this methodology. While cell culture techniques are certainly advantageous for the separation and investigation of a homogeneous cell population, it is possible that the tissue culture process may alter intracellular cascades for transporter regulation or translocation properties. Additionally, our density gradient centrifugation technique is superior to conventional synaptosomal isolation procedures, as will be shown by data to follow, because it minimizes glial contamination. In view of this, one must question the validity of published results from studies in synaptosomes which have been attributed specifically to neuronal-derived vesicles, including kinetic properties and pharmacological properties of L-glutamate transport inhibitors.

Glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) are detached, sealed glial and synaptic nerve terminals, respectively, which have been separated by differential and density-gradient centrifugation. Both maintain ionic gradients and various membrane










properties of central nervous system (CNS) cells and retain the machinery needed to accumulate neurotransmitters, as detailed in numerous reports (Dunkley et al., 1986; Ferkany and Coyle, 1986; Bridges el al., 1991; Fykse and Fonnum, 1991; Nakamura et at., 1993).

Nakamura and coworkers (1993) used electron microscopy to examine the

subcellular fractions prepared by Percoll density gradient centrifugation. It was found that the vast majority of the membrane components of the GPV fraction contained two types of vesicles, a small (0.15-0.2 microns in diameter) spherical population and a large (0.30.8 microns in diameter), agranular, and irregularly shaped population. As of yet, it has not been determined whether both populations retain functional properties. It should be noted, however, that minute amounts of disintegrated SYN, mitochondria, and membrane fragments with postsynaptic densities were observed. Likewise, SYN have been studied through electron microscopy (Dunkley et al., 1986) and found to contain an abundance of small diameter vesicles, with plasma membranes and intrasynaptosomal mitochondria intact.

Phosphorylation / Protein Kinase C

One important component in various signal transduction pathways is the

phosphorylation of a substrate, and the substrate's resultant activation or inactivation (Hardie, 1990; Barford, 1991). Protein phosphorylation is the covalent addition of inorganic phosphate to specific amino acids by protein kinase enzymes. There are two major groups of protein kinases, those that phosphorylate serine and threonine residues and those that phosphorylate tyrosine residues. The first demonstration of a direct








11

regulation of a neurotransmitter transporter by phosphorylation was published by Casado and colleagues (Casado el al., 1993). This group reported that a purified L-glutamate transporter from pig brain was phosphorylated by protein kinase C (PKC), predominantly at serine residues. When C6 glioma cells were exposed to 12-O-tetradecanoylphorbol-13acetate (a stimulator of PKC activity), an approximate 2-fold increase in L-glutamate transport was observed within 30 minutes. In view of this first published report, our studies focused on the ability of PKC to stimulate L-glutamate transport in glial-derived vesicles and neuronal-derived vesicles.

The term "protein kinase C" encompasses an eleven member family of

serine/threonine-specific protein kinases which have been identified functionally by common enzymatic properties, including phorbol ester binding, phospholipid-dependent kinase activity and common structural features (see review: Stabel and Parker, 1991). The PKC cDNA clones first isolated and their corresponding polypeptides are now called alpha, beta,, beta2, and gamma isoforms. Due to their structural organization, PKC alpha, beta,, beta2, and gamma define the "class I" PKC enzymes, those dependent upon Ca2 . This is in contrast to the "class II" enzymes, delta, epsilon, eta, eta', and theta, which are distinguished by their structure, enzymatic properties, and Ca2 -independence (Stabel and Parker, 1991). Class III or atypical PKC isozymes include zeta and lambda (Nishizuka, 1992). These isozymes are phospholipid-dependent, but Ca2- independent and do not bind phorbol esters (Nishizuka, 1992).

The ability of PKCs to interact with and be activated by membranes

(phospholipids) and diacylgylcerol (DAG) presents a functional definition of these








12

proteins. As such, binding to lipids is a critical and indeed well studied phenomenon (see review: Epand and Lester, 1990). This interaction, which has been most clearly documented for the alpha, beta, and gamma enzymes, appears to be minimally a two-step process. The first step is the formation of a ternary complex of enzyme, Ca2', and phospholipid; the second step in association entails binding of DAG (or phorbol ester), which through conformational changes leads to activation. The ability of phorbol esters to activate PKC through binding at the DAG site has aroused much interest. Phorbol esters bind to the C I domain of PKC, a domain that contains a highly conserved cysteine-rich repeated sequence found in all isoenzymes (Ono et al., 1988). In addition sustained activation with phorbol esters selectively depletes PKC that is degraded by proteolytic enzymes (Young et al., 1987). Tumor-promoting phorbol diesters, such as 12-0tetradecanoylphorbol 13-acetate (TPA, sometimes referred to as PMA) and phorbol 12,13-dibutyrate (PDBu) activate all known PKC isoenzymes in vivo, with the exception of class III PKC isoenzymes zeta and lambda (Nishizuka, 1992). To date no phorbol ester has exhibited PKC isozyme-selective activity in vivo (Kiley and Jaken, 1990; Roivainen and Messing, 1993; Kiley et al., 1994). Notably, however, in vitro work by Ryves and colleagues (1991) demonstrated that the proinflammatory, non-tumor promoting phorbol ester, 12-deoxyphorbol- I 3-O-phenylacetate-20-acetate (DOPPA), was a PKC..,-specific agonist. DOPPA effectively stimulates PKCb. kinase activity at a concentration of 20 nM (10 ng/ml), but does not activate PKC alpha, gamma, delta, or epsilon at concentrations up to 2MM.










Inhibitors at the DAG binding site have been isolated from a soil fungus (lida et at., 1989; Kobayashi et al., 1989). These compounds have a multi-tiered quinone structure and are termed calphostins. The most potent of these compounds is calphostin C, which has an IC5o of 50 nM. At a concentration of 1M, it causes complete inhibition of the binding of 50 nM [3H]phorbol dibutyrate to PKC and subsequent inhibition of its activity (Kobayashi el al., 1989). IC50 values greater than 28MM are necessary for calphostin C to inhibit cyclic AMP-dependent protein kinase, cyclic GMP-dependent protein kinase, or pp6O' protein tyrosine kinase (product of the src oncogene) (Kobayashi et al., 1989; Tamaoki et al., 1990; Bruns et al., 1991).

Dephosphorylation / Serine/Threonine Protein Phosphatases

While the role and influence of phosphorylation in the CNS is becoming more and more clear, the influence of protein dephosphorylation has yet to be investigated in detail. However, it does seem reasonable to assume that dephosphorylation is equally as important as phosphorylation in nervous system function and that protein phosphatases, enzymes which remove inorganic phosphate from a substrate, may play a key role in the regulation of such cellular processes as neurotransmitter release, ion fluxes, receptor availability, etc. (Nestler and Greengard, 1984; Cohen, 1989; Shenolikar and Nairn, 1991). This lack of defined, precise roles for protein phosphatases is primarily due to the rather ubiquitous nature of protein phosphorylation as a regulatory mechanism, such that the use of general phosphatase probes is likely to have a myriad of effects (Sim, 1991).

A large number of phosphoprotein phosphatases have been described in

mammalian tissues (Cohen, 1989). These enzymes can be divided into two broad types:










phosphoserine/phosphothreonine-specific protein phosphatases or PP and phosphotyrosine-specific protein phosphatases or PTP (Gong et al., 1993). The activity of each type of phosphoserine/phosphothreonine-specific protein phosphatase in mammalian tissues can be determined on the basis of differences in substrate specificities, dependence on divalent cations, sensitivities to specific inhibitors, and their catalytic subunit (Ingebritsen et al., 1983; Cohen, 1989). Based upon their four types of catalytic subunits, the phosphoserine/phosphothreonine-specific protein phosphatases are comprised of four main classes of enzymes: PP 1, PP2A, PP2B, and PP2C (Ingebritsen et al., 1983; Cohen, 1989).

Protein phosphatase 1 (PP 1) preferentially dephosphorylates the beta-subunit of phosphorylase kinase and is inhibited by two heat-stable inhibitor proteins, inhibitor-I (which inhibits PP I after phosphorylation by cyclicAMP-dependent protein kinase) and inhibitor-2 (also termed "modulator," which inhibits PP I by impeding the substrate binding and by inducing a conformational change of the catalytic subunit) (see review: Bollen and Stalmans, 1992). As noted in several reviews (see: Cohen, 1989; Sim, 1991; Bollen and Stalmans, 1992; Shenolikar, 1994), the substrate specificity of PP1 may be controlled by a number of different regulatory subunits both in different tissues and within the same tissue. These regulatory subunits direct PP I activity toward specific subcellular localizations, and, therefore, toward specific substrates, through modulation of the phosphorylation state of a number of different regulatory subunits. Subcellular fractionation studies have demonstrated PP I activity in cytosolic, synaptosolic, synaptic plasma membrane and synaptic junction fractions (Shields el al., 1985; Dokas et al., 1990). These findings are








15

consistent with a ubiquitous distribution of this phosphatase in brain. Alternatively, type 2 phosphatases preferentially dephosphorylate the alpha-subunit of phosphorylase kinase and are insensitive to inhibitor-I and inhibitor-2. Type 2 protein phosphatases are subdivided into three distinct classes based on their cationic requirements: PP2A, PP2B, and PP2C. PP2A is active in the absence of divalent cations. The amount of PP2A activity in brain extracts is the highest of any tissues investigated (Ingebritsen et al., 1983), and there is approximately three times as much PP2A as PP 1. Native PP2A enzymes are heterotrimers of two regulatory subunits (A and B) and a catalytic subunit (Cohen, 1989; Shenolikar and Nairn, 1991). The mechanisms responsible for regulating PP2A activity in normal cells are still poorly understood. Notably, PP2B (i.e. calcium/calmodulindependent protein phosphatase or calcineurin) and PP2C are completely dependent on calcium and magnesium, respectively. PP2B is a heterodimer composed of equal amounts of A and B subunits. The A subunit contains the catalytic and calmodulin binding domains, whereas the B subunit binds Ca2 and is highly homologous to calmodulin (Klee, 1988). Calmodulin appears to activate PP2B by neutralizing the inhibitory effect of a 4kD domain on the A subunit, distinct from the calmodulin-binding domain. This inhibitory domain is extremely susceptible to proteolysis, and its removal yields a Ca2 -dependent enzyme that cannot be further stimulated by calmodulin (see review: Bollen and Stalmans, 1992). PP2C, meanwhile, is a monomeric protein which is structurally different from all other protein phosphoserine/phosphothreonine-specific phosphatases (Shenolikar and Nairn, 1991). Beyond its dependence on Mg2 , nothing is known about the regulation of this phosphatase or its specific substrates, and functional roles have not been identified.








16

These four classes of enzymes also may be distinguished by their sensitivity to the marine toxin, okadaic acid (Cohen, 1989; Haystead et al., 1989). Okadaic acid, a polyether fatty acid, was shown to be a potent and specific inhibitor of PPI and PP2A by Takai and coworkers (1987), while not activating PKC (Suganuma et al., 1988). Okadaic acid inhibits PP2A (IC50 = 0.1-1.OnM) at concentrations ten to one hundred times less than those required to inhibit PP 1 (Suganuma et al., 1988). Due to its membrane permeability, it can be used in intact cells to identify physiological substrates of PP1 and PP2A, as well as to reveal novel intracellular processes that are controlled by phosphorylation (Hardie et al., 1991). However, in vitro sensitivity of PP2A to okadaic acid is decreased by increasing protein concentrations (see review: Shenolikar, 1994). Moreover, radiolabeled okadaic acid predominantly accumulates in membranes, where PPI is more abundant than PP2A. Thus, it may be difficult to utilize solely the okadaic acid sensitivity of a particular process in the intact vesicle to identify the phosphatase involved.

Another phosphatase inhibitor, isolated from a marine sponge (Kato et al., 1986), is calyculin A. This inhibitor, though structurally unrelated to okadaic acid, has been shown to bind to okadaic acid receptors in particulate and cytosolic fractions of mouse skin (Fujiki et al., 1989, 1991; Suganuma et al., 1989, 1990). It is a potent inhibitor of PP1 and PP2A. Distinct from okadaic acid, calyculin A was found equally effective against PP1 and PP2A, with IC50 values of 1.4nM and 2.6nM for PPI and PP2A, respectively (Ishihara et al., 1989; Suganuma et al., 1990, 1992).

Studies by Bu and colleagues (1993) and others (Turner et al., 1993) have

indicated that okadaic acid and other cell-permeable protein phosphatase inhibitors, such










as calyculin A, lead to a preferential increase in the phosphothreonine content of metabolically labeled proteins. Because of this, it has been suggested that phosphothreonines, rather than phosphoserines, are preferentially dephosphorylated in cells (Shenolikar, 1994). Therefore, phosphatase inhibitors may be most useful for visualizing threonine phosphorylation.

Aging

Human Studies

It is possible that altered protein phosphorylation is involved in the neuronal loss characteristic of Senile Dementia of the Alzheimer's Type (SDAT) (Armbrecht et al., 1993). The maintenance of neuronal structure and function is thought to require the continuous action of neurotrophic factors. One common pathway for the action of trophic factors is the activation and translocation of PKC. In this regard, it was found that PKC activity in human frontal cortex was reduced by about 50% in SDAT patients (Cole et al., 1988). Likewise, the phosphorylation of P86 protein, the major PKC substrate in the frontal cortex, was also reduced by 50%. Using antibodies against individual PKC isoforms, alterations in PKC isozymes have been shown in patients with Alzheimer's Disease (Masliah el al., 1990). Additionally, Van Huynh and his group in 1989 found that decreased PKC and decreased protein phosphorylation were also seen in fibroblasts from SDAT patients. These findings, therefore, suggest possible systemic defects in protein phosphorylation in SDAT.

The effect of normal aging on phosphatase activity has not been investigated in any detail. Studies in SDAT brains have shown aberrant protein phosphorylation, suggesting










an alteration in protein kinases and/or phosphoprotein phosphatases. The microtubuleassociated protein tau is known to be hyperphosphorylated in SDAT brains, and this abnormal hyperphosphorylation is associated with an inability of tau to promote the assembly of microtubules in the affected neurons. Previous studies by Khatoon and coworkers (1992) and others (Swaab et al., 1992; Iqbal et al., 1993) have demonstrated that abnormally phosphorylated tau could be dephosphorylated after treatment with alkaline phosphatase, thereby suggesting that the abnormal phosphorylation of tau might in part be the result of a deficiency of the phosphoprotein phosphatase system in patients with SDAT. Furthermore, Gong et al. (1993) reported that PP 1 and PP2A activity in frontal gray and white matter were significantly decreased in SDAT brains when compared to controls. PP2B, PP2C, and phosphatase activities remained unchanged. As a change in the state of phosphorylation of a substrate protein and therefore its functional activity can be achieved physiologically through increases or decreases in the activity of either a protein kinase or a phosphoprotein phosphatase, it is clear that analyses of phosphoprotein phosphatases is important in clarifying the mechanism of aberrant protein phosphorylation/dephosphorylation in SDAT brains. Rodent Studies

The activity of PKC has been investigated in the senescent rodent brain, and agerelated modifications in levels of protein phosphorylation have been reported (Barnes et al., 1988; Govoni et al., 1988; Magnoni et al., 1991; Parfitt et al., 1991; Pisano et al., 1991). Specifically, age-related changes in protein phosphorylation mediated by PKC have been described by Barnes and colleagues (1988). This group reported that in the








19

hippocampus of senescent rats, there was a 46% decrease in phosphorylated F1 (i.e. GAP43 or B-50), a 47kD protein postulated to play a role in long term potentiation, signal transduction, and neurotransmitter release. However, when the content and functional activity of PKC were measured in aged rats by Battaini and colleagues (1990), results indicated that both these parameters were enhanced in the hippocampus yet reduced in the cortex. The age findings were similar to those later obtained by Pisano and colleagues in 1991, which indicated that the phosphorylation of certain cytoskeletal/microtubular proteins by PKC declines with age. Using brain slices, they demonstrated that the phosphorylation of tubulin, microtubule-associated proteins (MAPs), and tau protein was much less in 24 month old rats as compared to 6 month old rats in response to phorbol esters. When mRNA levels of three PKC isoforms (i.e., a, P3, y) were investigated in aging rats, no change in expression was found (Battaini et al., 1994; Narang and Crews, 1995). However, Northern blot studies have indicated lower PKC,, and PKC, mRNA levels in the cortex of 28 month old rats as compared to animals aged 3 months (Battaini et al., 1993). Several groups have demonstrated that activation of calcium-dependent PKCs is defective in the brains from aged rats (Friedman and Wang, 1989; Meyer et al., 1994; Battaini el al., 1995; Undie et al., 1995). This deficit in translocation has been related to biochemical changes in membrane phospholipid composition (Undie et al., 1995) and with functional impairment in neurotransmission (Friedman and Wang, 1989; Meyer et al., 1994). A study by Pascale and colleagues (1996) suggest that alterations in receptors for activated C kinase which bind PKC contribute to the functional impairment in PKC activation observed in aged rat brain.








20

Several studies have examined EAA uptake in tissue preparations from a variety of areas of the aging rat brain. In the neocortex, uptake of both L-['4C]glutamate into a preparation of synaptosomes and D-[3H]aspartate in crude homogenates was found to decline with age, although these changes appear to reflect differences between immature (2-4 months of age) and mature (10-30 months of age) animals (Strong et al., 1984; Meldrum et al., 1992). By contrast no difference was observed in L-[3H]glutamate uptake into slices of frontal neocortex in a comparison of animals aged 6 and 24 months (Dawson et al., 1989). In the neostriatum, uptake both of L-[3H]glutamate (in crude synaptosomal fractions) and D-[3H]aspartate (in crude homogenates) was lower in mature rats (aged 10 months or more) than in immature animals, aged 6 months or less (Price et al., 1981; Wheeler and Ondo, 1986; Najlerahim el al., 1990). Additionally, Palmer and colleagues, using Fisher 344/Norwegian Brown rats aged 3, 12, 24, and 37 months, assayed D[3H]aspartate uptake in crude cortical, hippocampal, and neostriatal synaptosomes and found no significant changes with increasing age (Palmer et al., 1994).

The conflicting results in these studies may be attributable to several factors. First, because of different vulnerabilities of inbred strains of rats to the effects of aging, previous studies may have used animals that had not fully reached senescence (Coleman and Flood, 1987). Second, erroneous conclusions may be made on the basis of studies that examined animals at two ages only (Coleman and Flood, 1987). Finally, the young control animals (<12 months of age) used in many studies may not have been entirely appropriate, since there is evidence that sexual maturity (puberty) may not mark the end of what may be








21

considered maturation. Evidence indicates that this process may continue until 12 months of age in some brain regions of some rodent strains (Coleman and Flood, 1987).













CHAPTER 2
ISOLATION OF GLIAL AND NEURONAL FRACTIONS Introduction

Synaptosomes have proven to be an excellent model system for studies on the molecular mechanisms underlying presynaptic phenomena (Bradford, 1975). Since the original isolation procedures were developed (De Robertis el al., 1961; Gray and Whittaker, 1962), a number of methodological improvements have been described (Cotman, 1974; Morgan, 1976). Many researchers prefer to use crude mitochondrial or P2 fractions (Whittaker, 1972) for neurochemical investigations due to the simplicity and speed (30-40min) for isolation of this fraction as well as the avoidance of hypertonic conditions. However, standard P2 fractions have major disadvantages associated with contaminants (myelin, extrasynaptosomal mitochondria and glial cells) which can have direct and indirect effects on synaptic events (Dodd et al, 1981). Glial contamination of synaptosomal preparations is particularly problematic as evidenced by both biochemical (carbonic anhydrase activity) and morphological measures (Delaunoy et al., 1979; Henn el al., 1974).

As an alternative to traditional methods, isolation of synaptosomes on Percoll

gradients is rapid, isotonic and leads to a relatively homogeneous isolate that is specifically depleted of damaged synaptosomes, synaptic plasma membranes, extraneous membranous material, myelin and extrasynaptosomal mitochondria (Dunkley et al., 1986). The three-










step Percoll density gradient centrifigation technique is superior to other techniques insofar as glial vesicles are separated more efficiently from synaptosomes (Booth and Clark, 1978; Fykse and Fonnum, 1988). Complete separation of neuronal- and glialderived vesicles is of critical importance since both neurons and glial cells work together in order to maintain adequate and appropriate functioning in the central nervous system (CNS).

In contrast to synaptosomes, which have been used for studies of neuronal processes, few biochemical preparations have been available for the study of glial functions. Instead, a heavy reliance has been placed upon cultured cells in order to evaluate glial functions; however, it is known that the culturing process itself may alter intrinsic properties of various cell types. Within the past several years limited success have been realized with regard to the isolation of high-quality glial-derived fractions from whole-brain homogenates. An initial report by Nakamura and coworkers (Nakamura et al., 1993) involved a three-step Percoll density gradient centrifugation technique which was used to isolate simultaneously glial-derived vesicles (termed glial plasmalemmal vesicles or GPV) and neuronal-derived vesicles from a rat brain homogenate. While the method described in this report is based loosely on this previous work, significant technical modifications have produced marked improvements in the simultaneous isolation of functionally viable glial and neuronal elements. In this report, we have used morphological, biochemical and molecular markers to validate the efficiency of this separation technique, and we describe the properties of L[3H]-glutamate uptake by the two vesicle fractions.










Experimental Procedures

Materials

L-[2,3,4-3H]-Glutamic acid (specific activity = 60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). D-[2,3-3H]-Aspartic acid (specific activity = 13.5 Ci/mmol) and [methyl-3H]-choline chloride (specific activity = 81 Ci/mmol) were purchased from DuPont/NEN (Boston, MA, USA). Glial fibrillary acidic protein (GFAP) was acquired from Biogenesis (Sandown, NH, USA), while primary antibody against GFAP was purchased from DAKO (Carpinteria, CA, USA). Neuron specific enolase (NSE) and primary antibody were obtained from Polysciences (Warrington, PA, USA). Secondary antibodies, as well as the alkaline phosphatase color reagent development kit, were bought from Biorad (Hercules, CA, USA). D-Aspartic acid and DL-threo-1-hydroxyaspartic acid were purchased from Sigma Chemical Co. (St. Louis, MO, USA), while L-trans-pyrrolidine-2,4-dicarboxylic acid was obtained from Tocris Cookson (St. Louis, MO, USA). L-a-Aminoadipic acid was bought from Calbiochem-Novabiochem International (La Jolla, CA, USA). All other chemicals were purchased from either Fisher Scientific or Sigma Chemical Co. and were of the highest quality available.

Preparation of Tissue

Young adult (3 - 4 months of age) Sprague-Dawley male rats (Zivic Miller)

weighing 250-275g were used throughout this study. Animals were housed in pairs and maintained on a 12-hr light/dark cycle with food and water available ad libitum and were transported to the laboratory approximately 15 hrs prior to use. Two rats were










decapitated quickly with a small animal guillotine, and the brains were removed rapidly and placed upon an ice-cold glass surface. Cerebellar tissue was removed and discarded, while all remaining forebrain tissue (approximately 2.4 g) was placed in 30 ml of an icecold solution containing 0.32 M sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA). The tissue was homogenized gently with a Potter-Elvehjem tissue grinder (approximately 30 rpm) and centrifuged at 1000 x g for 10 min (4*C) using a fixed-angle rotor (F28/50-DuPont). All subsequent centrifugation steps were conducted at 4C. The resultant pellet was discarded, and the supernatant was split into four equal portions, which were diluted to 30 ml with an ice-cold solution containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM dithiothreitol and 20 mM HEPES (pH 7.4 at 40C). Hereafter, this solution is referred to as SEDH. Diluted aliquots of supernatant were centrifuged at 5000 x g (15 min), and resultant supernatants were saved separately on ice. Each of the four tissue pellets were resuspended in 15 ml of ice-cold SEDH solution and centrifuged at 1000 x g for 10 min. The resultant pellets were discarded, while supernatants were combined with supernatants saved from a previous step. The four tubes, each containing approximately 45 ml of tissue homogenate, were centrifuged at 33,500 x g (20 min), and supernatants were discarded. Tissue pellets (four) were resuspended in 15 ml of ice-cold SEDH solution and gently transferred onto a three-step discontinuous Percoll gradient (10 ml each of 1.38%, 2.3%, and 4.6% Percoll in SEDH solution) with a Minipuls 2 (Gilson) peristaltic pump (flow rate = 0.88 ml/min). Tubes were centrifuged at 33,500 x g (10 min) with 15-min periods of linear acceleration to and deceleration from the top speed. The turbid layer between 0% and 1.38% Percoll was collected from all four tubes and








26

combined into two aliquots, which were diluted to a final volume of 15 ml each with icecold SEDH solution. Aliquots were centrifuged at 1000 x g (20 min), and resultant supernatants were layered onto fresh three-step Percoll gradients as described above. Tubes were centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration (see above), and the turbid layer between 0% and 1.38% Percoll was collected from both tubes and combined into a single aliquot. The tissue aliquot was diluted to a final volume of 45 nl with ice-cold SEDH solution, centrifuged at 33,500 x g (20 min), and the resultant pellet was used as the glial plasmalemmal vesicle (GPV) fraction. For the recovery of the synaptosomal (SYN) fraction, the turbid layer between 2.38% and 4.6% Percoll was collected from the initial discontinuous gradient. The four aliquots were diluted to a final volume of 15 ml (each) with ice-cold SEDH solution and centrifuged at 1000 x g (20 min). Pellets were discarded, and supernatants were layered onto a threestep Percoll gradient (see above) and centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration periods. The turbid layer between 2.38% and 4.6% Percoll was collected from each of the four tubes and combined into one aliquot. This aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution and centrifuged at 17,500 x g (20 min). The resultant pellet was designated as the SYN fraction. The total preparation time from initial animal decapitation was approximately six hours. Lipid Quantification

Tissue phospholipids were separated from protein and quantified by a modification of the method of Rodriguez-Vico and coworkers (Rodriguez-Vico et al., 1991). Five nl of hexane/isopropanol (3:2, v/v) were added to a 13-ml centrifuge tube containing 0.05 ml








27

aliquots of either GPV or SYN. Samples were mixed and centrifuged at 10,000 x g for 15 min. The liquid phase was decanted and concentrated to dryness under a stream of dry N2. Following the modified method of Bartlett (Bartlett, 1959), samples were reconstituted in 1.0 ml H20 and acidified with 0.3 ml of 10 N H2SO4. The mixture was heated to 150'-160'C for 3 hours, and following addition of two drops of 30% H202, the solution was heated for 1.5 hours in order to complete the combustion process and break down any residual peroxide. After combustion, H20 (0.65 ml), 5% ammonium heptamolybdate tetrahydrate (0.2 ml), and Fiske-SubbaRow reagent (0.05 ml) were added, and the solution was heated for 7 min at 100�C. Following a 10 min incubation at room temperature, absorbance was measured at 660nm (LKB Ultraspeed spectrophotometer, Pharmacia) and molar concentrations for unknowns were estimated by comparison with inorganic phosphate standards (5 gM - 200 pM). Electron Microscopy

Electron microscopy was carried out in the Electron Microscopy Core Facility at the University of Florida Interdisciplinary Center for Biotechnology Research. GPV and SYN fractions were fixed in 2% buffered glutaraldehyde at 40C for 1 hr. After three 10 min washes with phosphate buffered saline (pH 7.2), samples were postfixed in 1% buffered osmium tetroxide (pH 7.2) for 1 hr. Following three washes with deionized water, samples were dehydrated in 50% ethanol for 15 min. Staining occurred overnight at 4�C with 2% uranyl acetate in 75% ethanol. GPV and SYN were dehydrated in a graded series of alcohols (15 min each), using 100% acetone as a transitional solvent. Infiltration occurred in a graded series of embedding resin/acetone mixtures (1 hr each),








28

and blocks were polymerized at 60'C for two days. Thin sections were taken on a Leica Ultracut E Ultramicrotome with a diamond knife and post-stained with 2% aqueous uranyl acetate followed by Reynold's lead citrate. Micrographs were taken on a Hitachi H-7000 TEM at 75kv.

[H]Excitatory Amino Acid Uptake

The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 30�C). GPV and SYN were used immediately for uptake. Sodium-dependent [3H]L-glutamate uptake by GPV and SYN was measured by a filtration method modified from Divac and coworkers (Divac et al., 1977). For the measurement of uptake, aliquots (50 pl) of tissue fractions (approximately 1 mg protein / ml) were added to cold glass culture tubes that contained 400 M! of a buffered solution containing (in mM) NaCi (140), KCI (5), CaCl2 (1.0), MgCI2 (1.0), NaHYP04 (1.2), D-glucose (10), and HEPES (20) at pH 7.4 (30C). Sodium-independent uptake (blanks) was measured in parallel using a buffer in which NaCI was replaced by an isosmolar concentration of choline chloride. Assay tubes in triplicate containing tissue homogenate and buffer were warmed to 300C (5 min) in a shaking water bath. The uptake reaction was initiated by the addition of 50 Jul of L-[2,3,4-3H]-glutamic acid at a final concentration of 5,uM. Tubes were mixed rapidly and returned to the water bath for 90 seconds. Uptake was terminated by rapid vacuum filtration using a Brandel cell harvester and Whatman GF/B filter sheets that had been presoaked overnight at 4�C in 25 mM L-glutamate. Test tubes and filters were rinsed rapidly three times with 2 ml aliquots of ice-cold normal or sodium-deficient buffer. Tissue trapped on filters was digested with 2 ml of 0.2 M NaOH (overnight), acidified








29

with I ml of 0.5 M HCI, and assayed for tritium content in 10 ml of EcoLume scintillation fluid (ICN Biochemicals). Radioactivity was quantified in an LKB 1214 liquid scintillation counter with a counting efficiency of approximately 45% as determined by a radium-226 standard. Sodium-dependent uptake was determined as the difference between uptake in normal versus sodium-deficient buffers. For time-course and dose-response studies, tissue fractions were incubated between 5 sec and 10 min in the presence of 5 ,M [3H]Lglutamate or were incubated with concentrations of [3H]L-glutamate ranging between 0.0177,uM - 100 M.M for 90 sec, respectively. Uptake of [3H]D-aspartate (5 jM, final conc) was measured according to the same procedure used to determine [3H]L-glutamate uptake.

[3HlCholine Uptake

High-affinity [H]choline uptake by GPV and SYN fractions was measured by a method modified from Divac and coworkers (Divac et al., 1977). Aliquots (50 gl) of tissue fractions (approximately 1 mg protein / ml) were added to cold glass culture tubes containing 400 gl of a solution containing (in mM) NaCI (120), KCI (4.7), MgCl2 (1.2), D-glucose (20), CaCI2 (2.5), and HEPES (50) at pH 7.4 (30'C). Nonspecific [3H]choline uptake was measured in the presence of 1 MM hemicholinium-3. Assay tubes in triplicate were warmed to 30'C (5 min) in a shaking water bath, and the reaction was initiated by the addition of 50 ,4l [methyl-3H]choline chloride at a final concentration of I pM. Uptake was allowed to proceed for 3 min (30'C) and was stopped by the addition of I ml of icecold buffer containing 10 ,4M hemicholinium-3. Tissue fractions were trapped by vacuum filtration onto Whatman GF/B filters that had been presoaked overnight at 4�C in 100 mM








30

choline chloride. Test tubes and filters were rinsed rapidly with four 2 ml aliquots of cold buffer. Tissue was digested and tritium content was determined as described above. Specific [3H]choline uptake was determined as the difference between tissue tritium content in the absence and presence of hemicholinium-3. Western Blot Analyses

Western blot analyses were conducted in the Protein Chemistry Core Facility at the University of Florida Interdisciplinary Center for Biotechnology Research. For immunoblotting of antibodies against GFAP and NSE, aliquots (5-15 ug protein / 20 U1) of GPV and SYN samples were subjected to SDS-polyacrylamide gel electrophoresis (10% Tris-Tricine polyacrylamide gels; 50V for 15 min, followed by 120V for 40 min). The electrophoresed proteins were transferred onto polyvinylidene difluoride membrane (Immobilon P, Millipore) (90V for 1 hr). Blots were blocked (1 hr.) with 5% milk in TrisTricine blocking solution (TTBS) (10 mM Tris-HC1, 150 mM NaCl, 0.05% Tween-20), then incubated overnight at 4�C with primary antibodies against either GFAP or NSE (1:1000). Blots were washed with TTBS and then incubated with alkaline phosphataseconjugated goat anti-rabbit IgG (1:1000) for 2 hrs. The secondary antibody was washed out with TTBS, and proteins were visualized via alkaline phosphatase color reagent development kit.

Glutamine Synthetase Activity

The activity of glutamine synthetase in the GPV and SYN tissue samples was

determined according to the method of Galanopoulous and colleagues (Galanopoulos el at., 1988) with slight modification. Aliquots (100 1) of either GPV or SYN were placed










in 13-ml centrifuge tubes containing 10 g1 of 3% (w/v) sodium deoxycholate (DierksVentling et al., 1975) and sonicated. Thereafter, 450 Ail of a solution containing (in mM) L-sodium glutamate (110), disodium ATP (11), imidazole (110), MgSO4 (44.5), Lcysteine.HCl (44.5), hydroxylamine.HC1 (44.5), tricyclohexylammonium salt of phosphoenolpyruvic acid (9), and 5U of pyruvate kinase (pH 7.4 at 37�C) were added, and the mixture was incubated for 15 min at 37�C in a shaking water bath. The latter two components constituted the ATP-regenerating system (Lund, 1970). The reaction was terminated by addition of 150 g1 of a solution containing equal volumes of 10% (w/v) FeCI3.6H20 in 2 N HC1, 24% (w/v) trichloroacetic acid, and 50% concentrated HCI. After centrifugation at 50,000 x g (20 min), the optical density of the supernatant was measured within 10 min on a spectrophotometer (LKB Ultraspeed, Pharmacia) at 540nm and quantified by comparison with standard solutions of L-glutamic acid gammamonohydroxamate.

Carbonic Anhydrase Activity

Carbonic anhydrase activity in GPV and SYN samples was determined according to a technique modified from that of Brion and coworkers (Brion et al., 1988). All reactions were carried out at 00C in gas-tight glass microvials (300 pd total volume) that were sealed with teflon septa. Tissue homogenate (30 Al) was mixed with 20 /A of H20 in microvials prior to being sealed. The vessels were purged with CO2 (0.5 ml / min) for 7.5 min prior to initiating the reaction. CO2 was infused and exhausted through two hypodermic needles which pierced the septum. Following the preincubation step, enzyme activity was determined via a colorimetric reaction. Fifty microliters of a solution








32

containing (in mM) imidazole (20), TRIS (5), and p-nitrophenol (0.4) were added to each vial with the aid of a syringe and needle. The time necessary for yellow color to dissipate completely was determined and compared to standard carbonic anhydrase solutions. Protein Analyses

The amount of protein in GPV and SYN samples was determined by the Lowry protein assay (Lowry et aL., 1951) using solutions of bovine serum albumin as standards. Data Analyses

All values reported were the mean � SEM from four to twelve fractions isolated on separate experimental days, except the glutamine synthetase results, which were reported as mean + SD. Statistical significance was determined via Student's t-test and accepted at a level of p<0.05.

Results

Morphological Examination of GPV and SYN Fractions

Rat forebrain homogenates were fractionated by a multi-step procedure which entailed both differential and discontinuous density-gradient centrifugation steps. While the basis for this fractionation procedure was based loosely on the method of Nakamura and coworkers (Nakamura et al., 1993), the final method used for this investigation is substantially more complex and time-consuming than the previously published procedure. In view of the amount of time required to isolate final tissue fractions using this modified procedure (approx. 6 hr), it was necessary to verify the structural integrity of constituents in the final tissue fractions with morphological approaches. Using standard electron microscopy, both the GPV and SYN fractions were found to contain a large number of










intact membrane-encapsulated vesicles (Fig. 2-1). Within the GPV fraction, membraneencapsulated vesicles could be broadly classified as small (0.1 - 0.3 gm) spherical structures or large (0.4 - 0.9 gm) irregularly-shaped structures (panel A), though this division was somewhat arbitrary. The vast majority of vesicles in the GPV fraction contained a clear core with little or no electron-dense material evident. In addition to vesicles, the GPV fraction contained non-descript membrane fragments and electron-dense vesicles, though the prevalence of these components was quite low. The morphological characteristics of elements within the GPV fraction are consistent with observations published by Nakamura and coworkers (1993). In contrast to elements in the GPV fraction, a high percentage of membrane-encapsulated structures in SYN fractions contained numerous identifiable synaptic vesicles with identifiable synaptic vesicles, postsynaptic densities and mitochondria (Fig. 2-1, panel B). As noted previously, these morphological characteristics are consistent with a neuronal origin for elements in the SYN fraction.

In addition to morphological characterizations, GPV and SYN fractions were analyzed for total content of protein and membrane phospholipids. Normalization of protein content by total membrane phospholipid indicated that the amount of membrane phospholipid (pmol lipid per mg protein) was not statistically different between the two fractions (155 + 24, n = 10 vs. 106 + 48, n = 12 for SYN and GPV, respectively) at a level of p_ 0.05.










Glial Markers

In order to obtain supportive evidence for the efficiency with which glial and neuronal elements can be separated by this tissue fractionation protocol, several approaches were used to measure the level of marker proteins and enzymes that are selectively expressed by glia or neurons. GPV and SYN fractions separated by the threestep discontinuous density gradient method were analyzed for glial-specific enzyme activities and GFAP, a marker protein that is specific for astroglial cells. As shown in Fig. 2-2, specific activities of glutamine synthetase (GS) and carbonic anhydrase (CA) were significantly higher (p < 0.00001) in GPV fractions relative to SYN fractions. In the case of GS, GPV fractions contained more than six-fold higher enzyme activity as compared to SYN fractions isolated from the same rat forebrain homogenates. Measurements of CA activity indicated an even greater difference between GPV and SYN fractions insofar as SYN exhibited no measurable activity whereas GPV contained an average of 2360 units of activity per mg protein. Based upon the estimated lower limit for detection of CA activity by this method (50 units / assay tube), these results indicate at least 45-fold greater CA activity in GPV relative to SYN fractions. Finally, as shown in Fig. 2-2 (panel C), Western blot analyses for the astrocyte-specific protein GFAP provide additional support for the effective separation of glial elements between SYN and GPV fractions. When incubated with GFAP-specific polyclonal antibodies, GPV fractions (5 - 15 gg protein / lane) exhibited strongly positive immunostaining to a band near 47 kD. By comparison, positive immunostaining in SYN fractions was evident only in the presence of 15 gg total protein. Despite our inability to quantify the amount of immunoreactive material in each










fraction, densitometric measurements indicated the presence of substantially more immunopositive material in the 5 gg GPV sample (4104 arbitrary units) relative to the 15 pig SYN sample (1873 arbitrary units) following background subtraction. The amounts of antigen and density of reaction product lay on the linear portion of a standard scale. Neuronal Markers

While the experimental results outlined above helped confirm the presence of glial markers in GPV fractions and the diminished presence of such markers in SYN fractions, additional studies were undertaken to confirm the presence of neuronal markers in SYN fractions and to estimate contamination of such markers in GPV fractions. One marker used to address this issue was sodium-dependent high-affinity choline uptake, a process that is highly localized to cholinergic neurons. In the presence of 1 piM substrate, hemicholinium-3-sensitive uptake of [3H]choline was nearly 40,000-fold higher in SYN vs. GPV fractions (9.68 vs. 0.00025 pmol / mg protein / 3 min). This result indicates that GPV are nearly devoid of neuronal contamination. Verification of this observation was obtained in Western blots by using an antibody against NSE, a marker protein localized preferentially to neurons. As shown in Fig. 2-3 (panel B), GPV had no positive immunoreactivity (above background) with up to 15 pg of total GPV protein loaded per lane. By comparison, SYN fractions displayed strong positive immunoreactivity for NSE, as illustrated by the heavy banding at approximately 49 kD. Excitatory Amino Acid Uptake

GPV and SYN fractions from rat forebrain were tested in order to ascertain their abilities to accumulate excitatory amino acid (EAA) substrates. As shown in Fig. 2-4,










both fractions demonstrated a good capacity to transport substrates in a sodiumdependent manner. While the level of EAA transport was found to differ between GPV and SYN fractions, there was a strong correlation between L-[3H]glutamate uptake and D-[3H]aspartate uptake within individual preparations (Fig. 2-4, inset). This observation served to validate the use of L-[3H]glutamate for studies of EAA transport in GPV and SYN fractions. As shown in Fig. 2-5, L-[3H]glutamate uptake at 30'C is a rapid, highaffinity process which exhibits pseudo first-order reaction kinetics (panel B). Analyses of results from substrate saturation experiments (panel A) revealed that EAA transporters in GPV and SYN fractions exhibit high affinities for L-[3H]glutamate (Km values of 2.7 tM and 5.1 gM, respectively), but differ significantly in their maximal rate of transport with the capacity in SYN exceeding that of GPV (773 vs. 279 pmol / mg protein / 90 sec). In addition pharmacological comparisons between GPV and SYN were conducted using synthetic L-glutamate analogues which have been reported to inhibit L-glutamate transport. As shown in Fig. 2-6, L-[3H]glutamate uptake was completely inhibited by several compounds in both tissue fractions. In studies with L-trans-pyrrolidine-2,4dicarboxylic acid (PDC) (panel A) and D-aspartic acid (D-ASP) (panel D), both tissue fractions revealed similar sensitivities to these transporter inhibitors. Average fitted IC50 values were 4.7, 4.1, 4.2 and 3.7 puM for PDC and D-ASP in GPV fractions and SYN fractions, respectively. In contrast to these results, both DL-threo-p3-hydroxyaspartic acid (THA) (panel B) and L-a-aminoadipate (AAD) (panel C) revealed modest selectivity for L-glutamate transporters in GPV fractions. Average IC5o values were 1.4 and 3.2 PM (THA) and 200 and 340 gM (AAD) in GPV and SYN fractions respectively. Despite























Figure 2-1. Representative electron micrographs of tissue fractions isolated from rat forebrain homogenates. Panel A) GPV fractions contained two major types of membrane-encapsulated vesicles which apppeared as large irregular-shaped vesicles (L) and small spherical vesicles (s). Scale bar = 0.5,4M.







00


Ii


V
/
/


K!


4
























Figure 2-1 .--continued
Panel B) SYN fractions contained numerous membrane-enclosed vesicles, containing many synaptic vesicles (V), as well as mitochondria (m). Scale bar = 0.5ptM.











/ 'I


i 5

























Figure 2-2. Analysis of GPV and SYN tissue fractions for glial-specific enzyme activities and GFAP.
Panel A) Glutamine synthetase (GS) activity was determined spectrophotometrically by monitoring GS-catalyzed formation of L-glutamic acid gamma-monohydroxamate (GHA). The absorbance of the samples was measured on a spectrophotometer at 540nm. Values are mean � SEM, (n=5) for each tissue fraction. One unit of activity is defined as lpmol GHA formed / 15min / mg protein. Panel B) Carbonic anhydrase (CA) activity in GPV (n=7) and SYN (n=l 1) fractions expressed as units of enzyme activity per mg protein. Panel C) Western blot analysis for GFAP immunoreactivity in GPV (lanes 3-5) and SYN (lanes 6-8) fractions. Aliquots containing either 5kig (lanes 3 and 6), 10ug (lanes 4 and 7) or 15yg (lanes 5 and 8) of total protein were subjected to SDS-polyacrylamide gel electrophoresis prior to reaction with primary and secondary antibodies. Molecular weight markers were run in lane 1 and antigen in lane 2. Numbers to the left of lane I indicate molecular weight in kD.











Carbonic Anhydrase (units of activity)


tj
0 ( I I


I I


Glutamine Synthetase
(units of activity)


0 0 I. I I


I


LM MMM













128


4-


664630-


A)











GPV


SYN


- - l -


123 4 5 678



Figure 2-3. Measurement of neuron-specific markers in GPV and SYN. Panel A) High-affinity [3H]choline uptake by GPV and SYN fractions was determined in the presence of IMM [3H]choline. Values are mean � SEM, (n=5) for each tissue fraction. Panel B) Western blot analysis for NSE immunoreactivity in SYN (lanes 3-5) and GPV (lanes 6-8) fractions. Molecular weight markers were run in lane I and antigen in lane 2. Numbers to the left of lane I indicate molecular weight in kD.


U

E






















1200 - [HJD-A pa1UW 100 [3H]L-Giuttmate

1000 1200 -0 GPV 800 E SYN 900
800 - 90

6600
300
600 r 9
0 300 600 900 1200 400['HIASP Uptake
200

0

SYN GPV



Figure 2-4. Comparison of sodium-dependent [3H]D-aspartate and [3H]L-glutamate uptakes by GPV and SYN fractions. Tissues were incubated in the presence of 5,uM substrate for 90 seconds. Values are mean � SEM, (n=5) for each tissue fraction. Inset depicts plot of paired values from individual tissue fractions for all experiments. Degree of correlation (r) equals 0.998.












"-! 300 750
o

500 - 200 0


E 250 0-0 GPV - 100 .
OWC .... C1 SYN

0-- - i I 0

0 20 40 60 80 100 120

[GLU], (10' M)

1600 900 B)

1200
600 o

t800 o '"

0 300 0
l 400 - OGPV []E ..... El SYN

0- 0 I 1 0
0 150 300 450 600
TIME (sec)


Figure 2-5. Properties of sodium-dependent L-[3H]glutamate transport in GPV and SYN fractions isolated from rat forebrain. Panel A) Substrate-saturation curves for L-[3H]glutamate transport in GPV and SYN. Aliquots (50,.l) of tissue were incubated for 90 seconds in the presence of L[3H]glutamate (0.0 177AuM - 100PM). Data points represent the mean + SEM from 5 separate experiments. Panel B) Time-course for L-[3H]glutamate transport in GPV and SYN in the presence of 5guM substrate. Values are mean � SEM, (n=4) for each tissue fraction.

























-9 -8 -7 -6 -5
LOG [PDC] (M)


4020

0 i
-4 -8 -7 -6 -5
LOG [THA] (M)


100 80 60


-6 -5 -4 -3 -7 -6 -5 -4


LOG [AAD] (M)


LOG [D-ASP] (M)


Figure 2-6. Inhibition of sodium-dependent L-[3H]glutamate uptake in GPV and SYN fractions by selected L-glutamate analogs. Tissue fractions (n=4-12 per data point) were incubated for 5 minutes in the presence of transport inhibitors prior to measurement of L[3H]glutamate uptake. Curve-fitting procedures (see Experimental Procedures) were used to derive IC50 values (see Results) as well as estimates of the pseudo Hill slope (n) for inhibitors. Fitted values of n (for GPV and SYN, respectively) were 0.89 and 0.80 for PDC; 0.79 and 0.85 for THA, 1.23 and 0.96 for AAD; and 0.66 and 0.78 for D-ASP.


CI



0










these differences, both fractions displayed identical rank orders of potencies (THA>DASP>PDC>AAD) for these L-glutamate transport inhibitors.

Discussion

In contrast to synaptosomes, which have been used extensively to characterize neuronal processes in the CNS, few preparations derived from CNS tissues have been available for the study of glial functions. While cultured glial cell lines provide the benefits of a homogenous and stable cell population, certain limitations restrict the range of uses or applications for these preparations. In addition, since it is possible for cellular properties to be altered under culture conditions, it is prudent to determine to the greatest extent possible the degree to which cultured cells accurately reflect cell functions in living tissues. In view of these considerations, there is need for a technique which can be used to obtain nontransformed cells or functional cell derivatives from tissues of experimental animals. Such a technique provides an important adjunct to cell culture whereby the effects of aging or experimental manipulations on cellular/molecular events can be evaluated in CNS cells. The technique presented herein provides such a methodology, yielding simultaneously isolated fractions highly-enriched with either glia-derived or neuronderived vesicles. However, the point should be made that no technique is without drawbacks and, as such, important restrictions of our improved technique are discussed below.

The extensive multi-step density gradient fractionation technique described in this report is based loosely upon a previous report by Nakamura and colleagues (1993). Since a far greater number of cellular markers have been used in the present study to estimate










the extent of glial and neuronal separation, presenting an exhaustive comparison of the separation achieved with our method versus that of the Nakamura method is not possible (Nakamura et al., 1993). However, since two cellular markers (high-affinity [3H]choline uptake and GFAP immunoreactivity) were used by both groups, these markers provide a means for direct comparison of the fractions isolated by the two procedures. Using hemicholinium-3 -sensitive [3H]choline uptake as a neuronal marker, Nakamura and coworkers reported that GPV fractions exhibited uptake that was 10 percent of uptake in the SYN fractions (Nakamura et al., 1993). However, using the fractionation technique described in this report, hemicholinium-3-sensitive [3H]choline uptake in GPV was measured as less than 0.0025 percent of uptake in SYN fractions (Fig. 2-3A). Therefore, based upon this neuronal marker, it is estimated that our fractionation procedure reduces neuronal contamination in GPV fractions by a factor of 4000 compared with the previous method (Nakamura el al., 1993). Another marker used in both studies was GFAP immunoreactivity as measured by Western blot analysis. GFAP is a cell-specific intermediate filament protein found only in astrocytes (Eng, 1985; Eng et al., 1971; Goldman et al., 1978). As shown in Fig. 2-2C, GFAP was undetectable in SYN fractions that contained less than 15,ug total protein. However, Nakamura and colleagues (1993) detected GFAP in substantially smaller amounts (5/ig protein) of SYN fractions isolated with their procedure. Since neither study provided accurate quantitative determinations of GFAP levels, calculating the difference achieved by the two fractionation procedures is not possible. Nevertheless, the current method clearly offers a substantial improvement in the removal of glial elements from the SYN fraction as compared with the earlier method.








49

Since a major goal of this study was to develop and validate a method for isolating functional glia-derived elements that were devoid of neuronal contamination, results from studies with neuron-specific markers (high-affinity choline uptake and NSE immunoreactivity) provide solid evidence for the highly-enriched glial nature of the GPV fractions. High-affinity [3H]choline uptake (Fig. 2-3 A) is the rate-limiting step for acetylcholine synthesis in neurons (Kuhar and Murrin, 1978). Neuronal choline transporters are coupled to an electrochemical sodium ion gradient and are purported to be associated exclusively with cholinergic neurons (Haga and Noda, 1973; Kuhar et al., 1973; Yanamura and Snyder, 1972; Yamamura and Snyder, 1973). As shown in Fig. 2-3 (panel A), [3H]choline uptake was virtually undetectable in GPV fractions, being nearly 40,000-fold lower than uptake in SYN fractions. However, the near absence of [3H]choline uptake in GPV fractions could not be attributed to a loss of transmembrane sodium gradient or other non-specific changes, since GPV demonstrate a good capacity for sodium-dependent L-[3H]glutamate uptake (Figs. 2-5 and 2-6). The efficient removal of neuronal elements from the GPV fraction represents a striking improvement over the results obtained by Nakamura and coworkers (1993), wherein GPV fractions maintained very appreciable levels of [3H]choline accumulation. In addition to choline uptake, NSE immunostaining was used to assess neuronal contamination in GPV fractions. NSE, a unique form of the glycolytic enzyme enolase, preferentially though not exclusively localized in neurons and neuroendocrine tissues (Marangos et al., 1975; Kato et al., 1982), helped validate our observation with choline uptake. As shown in Fig. 2-3B, there was no positive immunostaining in GPV fractions, even when lanes were overloaded with








50

high levels of total protein. Taken together, these data indicate that the method described in this report provides a highly-enriched, functionally viable population of glial cell elements which are suitable for studies of glial cell functions m vitro. It should be noted, however, that "glia-derived" may well include oligodendrocyte and microglial processes as well as astrocyte processes. Indeed, the CA activity present in GPV fractions is consistent with some contribution from oligodendrocytes, since this enzyme is expressed by oligodendrocytes (Cammer, 1984; Ghandour et al., 1979) as well as astrocytes (Cammer and Tansey, 1988a, b).

Similar approaches were used in order to establish the highly-enriched neuronal nature of synaptosomes (i.e., absence of glial cell markers) that were isolated in parallel with GPV fractions. In addition to Western blot analyses for GFAP, two marker enzymes were used to further validate the relative lack of glial contamination of SYN fractions. CA, a ubiquitous enzyme localized selectively to glial cells, is associated with the control of ion and fluid movements and acid-base balance (Trachtenberg and Sapirstein, 1980). As shown in Fig. 2-2 (panel B), no carbonic anhydrase activity was evident in the SYN fractions. In contrast, a high level of CA activity was evident in GPV fractions. Finally, studies were conducted to assess activity of GS, a key enzyme which catalyzes the amidation of L-glutamate to glutamine (Palaiologos et al., 1985). Although previous studies involving inunohistochenical localization of GS have indicated its selective association with glial cells (Martinez-Hernandez et al., 1977; Norenberg and Martinez-








51

Hernandez, 1979; Norenberg, 1983) GS activity was detectable in SYN fractions (Fig. 22A). However, we do not know if this finding reflects the presence of intact glial-derived vesicles within the neuronal-derived fractions.

In view of the evidence outlined above, it was concluded that our improved

method provides an efficient means to isolate glial and neuronal elements from a rat brain homogenate. Therefore, studies were performed to characterize L-glutamate transport in these isolated fractions.

Studies of L-[3Hlglutamate uptake by GPV and SYN revealed remarkably similar properties for both fractions. Kinetic studies indicated a more rapid initial rate for uptake in GPV, as well as a slightly higher substrate affinity. Conversely, SYN fractions exhibited a 2-3-fold greater capacity for L-[3Hlglutamate transport. While there are no appropriate published values with which our findings in GPV can be compared, it should be noted that our highly enriched SYN fractions displayed somewhat lower capacities for L[3H]glutamate uptake than previously reported for crude synaptosomal preparations. The low values in our case may be a consequence of metabolic 'run-down' brought about by the lengthy (6 hrs) purification process. In this regard, it was observed that the GPV fraction undergoes a rapid loss of L-[3H]glutamate uptake capacity following brief (several minutes) exposure to elevated temperatures (data not shown). A limited study with four transport inhibitors (i.e., L-trans-pyrrolidine-2,4-dicarboxylic acid, L-a-aminoadipate, Daspartic acid, and DL-threo-3-hydroxyaspartic acid), revealed no remarkable differences between GPV and SYN fractions. This finding is in good agreement with previous studies of these compounds (Dowd et al, 1996; Rauen et al., 1992).








52

While possible pathways for regulation of L-glutamate transporters in vivo have

not been delineated, several mechanisms involving endogenous substances (Barbour et al., 1989; Pogun et at., 1994; Trotti et a., 1996) and transporter phosphorylation (Casado et al., 1991, 1993) have been noted as possible regulators of transporter function. With the advent of this improved technique which separates glia-derived and neuron-derived vesicles from fresh rat brain homogenates, it is feasible to study the regulation of Lglutamate transporter function and to evaluate the potential effects by aging on Lglutamate transporter function in the CNS.














CHAPTER 3
EFFECT OF PHOSPHORYLATION ON L-[3HIGLUTAMATE UPTAKE Introduction

The amino acids L-glutamate and L-aspartate are the predominant excitatory

neurotransmitters in the mammalian central nervous system (CNS) (see reviews: Fagg and Foster, 1983; Robinson and Coyle, 1987). Neurotransmission at glutamatergic synapses is terminated by the reuptake of neurotransmitter by sodium-dependent high-affinity transporters located in neuronal and glial cell membranes. Recently, distinct cDNAs (GLT-1, EAAC 1, GLAST, EAAT4, EAAT5) that encode subtypes of L-glutamate transporters were isolated (Kanai and Hediger, 1992; Pines et al., 1992; Storck et al., 1992; Fairman et al., 1995; Arriza et al., 1997). Each of the cloned L-glutamate transporters contain consensus sites for phosphorylation by protein kinase C (PKC).

PKC is found in high concentrations in the CNS (Nishizuka, 1992). The term "protein kinase C" encompasses an eleven member family of serine/threonine-specific protein kinases which have been identified functionally by common enzymatic properties, including phorbol ester binding, phospholipid-dependent kinase activity and common structural features (see review: Stabel and Parker, 1991). The PKC cDNA clones first isolated and their corresponding polypeptides are called alpha, beta1, beta2, and gamma isoforms. Due to their structural organization, PKC alpha, beta,, beta, and gamma define the "class " PKC (cPKC) enzymes, those dependent upon Ca2+. This is in contrast to the








54
"class II" enzymes, delta, epsilon, eta, eta', and theta, which are distinguished from class I enzymes based on structure, enzymatic properties, and Ca2*-independence (Stabel and Parker, 1991). Class III or atypical PKC isozymes include zeta and lambda (Nishizuka, 1992). These isozymes are phospholipid-dependent, but Ca2 - independent and do not bind phorbol esters (Nishizuka, 1992). The endogenous activator of PKC is diacylglycerol (DAG). Phorbol esters, which bind specifically to PKC at the DAG binding site (Castagna et al., 1982), are also potent activators of the enzyme (Nishizuka, 1984).

The alpha, beta1, beta2, gamma, epsilon, delta, and zeta isoforms and their mRNAs have been identified in the brain using Western and Northern blot analysis and in situ hybridization. Immunohistochemical analysis using isoform-specific antibodies has revealed a differential distribution of the PKC isoforms in the mammalian CNS (Tanaka and Saito, 1992). The gamma isoform is expressed solely in the brain and spinal cord (Saito et al., 1988). It is localized mainly in cortical pyramidal cells, hippocampal pyramidal and granule cells, cerebellar Purkinje cells, and thalamic neurons (Saito et al., 1988). The beta2 isoform is localized mainly in cortical and hippocampal pyramidal cells and striatal neurons. The alpha and delta isoforms are universally distributed in all tissues and cell types so far examined (Ito et al., 1990). Notably, the glutamatergic neurons in the hippocampal formation display multiple PKC isoforms. The alpha, beta2, and gamma isoforms are contained in pyramidal cells, and the alpha and gamma isoforms in granule cells (Ito et al., 1990, Kose et al., 1990). The epsilon isoform is present predominantly in nerve terminals rather than the perikarya (Saito et al., 1993). Such differences in the








55

cellular distribution patterns of individual PKC isoforms yields further insight into the role of this enzyme in nerve function, where specificity of its actions may depend on anatomical distribution and substrate specificity.

Through the use of phorbol esters, a wide spectrum of functional systems have been demonstrated to be sensitive to PKC activation, including sodium-dependent amino acid transport (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997). The first direct evidence that PKC-mediated phosphorylation is involved in the regulation of L-glutamate transport was provided by Casado and colleagues (1993). Using antibodies directed against biochemically purified GLT-1, this group demonstrated that the amount of phosphorylation correlated with phorbol ester-induced stimulation of L-[3H]glutamate transport activity. One consensus site for phosphorylation by PKC is shared among EAAC 1, GLT- I and GLAST L-glutamate transporters. This is a serine residue, corresponding to serine 113 in GLT- 1. Studies by Casado and colleagues (1993) revealed that a site-directed mutagenesis of GLT- 1 serine residue 113 to asparagine abolished stimulation of L-glutamate transport by phorbol esters.

Other groups have established further that phosphorylation of brain L-glutamate transporters modulates transport function (Dowd and Robinson, 1996; Conradt and Stoffel, 1997). Preincubation of C6 glioma cells with phorbol ester caused a significant increase in L-glutamate transport activity (Dowd and Robinson, 1996). Observations that GLT- 1 and EAAC 1 exhibit increased rates of L-glutamate transport following PKC activation are in direct contrast with results reported by Conradt and Stoffel (1997) for








56

GLAST. In this case, treatment of GLAST-expressing cells with phorbol ester decreased L-glutamate transport activity with phosphorylation occurring at a non-PKC consensus site.

Like PKC isoforms, L-glutamate transporters have differences in their regional and cellular locations. GLAST is present in astrocytes (Rothstein et al., 1995; Lehre et al., 1995; Schmitt et al., 1997). Immunocytochemistry (Lehre et al., 1995) and in situ hybridization (Torp et al., 1994) have localized GLAST mRNA in the Purkinje cell layer of the cerebellum, associated with Bergmann glia (Storck et al., 1992), as well as in the cortex and hippocampus (Rothstein et al., 1994). GLT-1 is present in astrocytes throughout the brain, with predominant expression in telencephalic structures, including the hippocampus, neocortex, and striatum (Danbolt et al., 1992; Rothstein et al., 1994). Radioactive and non-radioactive in situ hybridization studies have shown that select neurons in the thalamus, hypothalamus, and pyramidal cells of the hippocampus express GLT-1 mRNA (Torp et al., 1994; Schmitt et al., 1996). Expression of mRNA for EAAC1, a neuronal L-glutamate transporter, is abundant in the pyramidal layer of the hippocampus (regions CA1-CA4), the granular layer of the dentate gyrus, the granule cell layer of the cerebellum and layers II-IV of the cerebral cortex (Kanai and Hediger, 1992). The cellular locations of the more recently cloned EAAT4 (Fairman et al., 1995), which is expressed in the cerebellum, and EAAT5 (Arriza et al., 1997), which is expressed in the retina, have not been published. However, given these differences in cellular and regional localization of L-glutamate transporters, as well as their responses following activation of PKC, changes in L-glutamate transport activity may be a reflection of the PKC isoform








57

present in the same cell as the L-glutamate transporter. In essence, the presence of PKC isoforms may differentially and/or selectively modulate L-glutamate transport depending upon their cellular location.

In contrast to synaptosomes, which have been used extensively to characterize neuronal processes in the CNS, few preparations derived from CNS tissues have been available for the study of glial functions. Since it is possible for cellular properties to be altered under culture conditions, it is prudent to determine the degree to which cultured cells accurately reflect cell functions in living tissues. In view of these considerations, we developed a technique which yields functionally viable glial- (glial plasmalemmal vesicles or GPV) and neuronal- (synaptosomes or SYN) enriched fractions from rat brain homogenates. Morphological, biochemical, and Western blot analyses previously confirmed the separation of glial- and neuron-derived fractions (Daniels and Vickroy, 1998). Prior studies confirmed that GPV and SYN tissue fractions exhibit high-affinity Lglutamate transport (Daniels and Vickroy, 1998) and can be used to study phosphorylation-dependent regulation of neuronal and glial L-glutamate transporters.

Experimental Procedures

Chemicals

L-[2,3,4-3H]-Glutamic acid (specific activity = 60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). Phorbol-12,13-dibutyrate was bought from Sigma Chemical Co. (St. Louis, MO, USA). Okadaic acid (potassium salt)










and calphostin C were purchased from Calbiochem Novabiochem International (La Jolla, CA, USA). All other chemicals were purchased from commercial vendors and were of the highest quality available.

Preparation of Tissue

Glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) were prepared as previously described (Daniels and Vickroy, 1998). Briefly, young adult (3 - 4 months of age) Sprague-Dawley male rats (Zivic Miller) weighing 250-275 g were used throughout this study. Animals were housed in pairs and maintained on a 12-hr light/dark cycle with food and water available ad libitum and were transported to the laboratory approximately 15 hrs prior to use. For each experiment, two rats were decapitated quickly with a small animal guillotine, and the brains were removed rapidly and placed upon an ice-cold glass surface. Cerebellar tissue was removed and discarded, while all remaining forebrain tissue (approximately 2.4 g) was placed in 30 ml of an ice-cold solution containing 0.32 M sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA). The tissue was homogenized gently with a Potter-Elvehjem tissue grinder (approximately 30 rpm) and centrifuged at 1000 x g for 10 min (4�C) using a fixed-angle rotor (F28/50-DuPont). All subsequent centrifugation steps were conducted at 4C. The resultant pellet was discarded, and the supernatant was split into four equal portions, which were diluted to 30 ml with an icecold solution containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM dithiothreitol and 20 mM HEPES (pH 7.4 at 40C). Hereafter, this solution is referred to as SEDH. Diluted aliquots of supernatant were centrifuged at 5000 x g (15 min), and resultant supernatants were saved separately on ice. Each of the four tissue pellets were resuspended in 15 ml of










ice-cold SEDH solution and centrifuged at 1000 x g for 10 min. The resultant pellets were discarded, while supernatants were combined with supernatants saved from the previous step. The four tubes, each containing approximately 45 ml of tissue homogenate, were centrifuged at 33,500 x g (20 min), and supernatants were discarded. Tissue pellets (four) were resuspended in 15 ml of ice-cold SEDH solution and gently transferred onto a three-step discontinuous Percoll gradient (10 ml each of 1.38%, 2.3%, and 4.6% Percoll in SEDH solution) with a Minipuls 2 (Gilson) peristaltic pump (flow rate = 0.88 nl/min). Tubes were centrifuged at 33,500 x g (10 min) with 15-min periods of linear acceleration to and deceleration from the top speed. The turbid layer between 0/6 and 1.38% Percoll was collected from all four tubes and combined into two aliquots, which were diluted to a final volume of 15 ml each with ice-cold SEDH solution. Aliquots were centrifuged at 1000 x g (20 min), and resultant supernatants were layered onto fresh three-step Percoll gradients as described above. Tubes were centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration (see above), and the turbid layer between 0% and 1.38% Percoll was collected from both tubes and combined into a single aliquot. The tissue aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution, centrifuged at 33,500 x g (20 min), and the resultant pellet was used as the GPV fraction. For the recovery of the SYN fraction, the turbid layer between 2.38% and 4.6% Percoll was collected from the initial discontinuous gradient. The four aliquots were diluted to a final volume of 15 ml (each) with ice-cold SEDH solution and centrifuged at 1000 x g (20 min). Pellets were discarded, and supernatants were layered onto a three-step Percoll gradient (see above) and centrifuged at 33,500 x g (10 min) with gradual acceleration and










deceleration periods. The turbid layer between 2.38% and 4.6% Percoll was collected from each of the four tubes and combined into one aliquot. This aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution and centrifuged at 17,500 x g (20 min). The resultant pellet was designated as the SYN fraction. L-I3HlGlutamate Uptake

The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 250C). GPV and SYN were used immediately for uptake. Sodium-dependent L-[3H]glutamate uptake by GPV and SYN was measured by a filtration method modified from Divac and coworkers (1977). For the measurement of uptake, aliquots (50,ul) of tissue fractions (approximately I mg protein/ml) were added to cold glass culture tubes that contained 400 g1 of a buffered solution containing (in mM) NaCI (140), KCl (5), CaC12 (1.0), MgCl2 (1.0), NaH2PO4 (1.2), D-glucose (10), and HEPES (20) at pH 7.4 (250 C). Sodium-independent uptake (blanks) was measured in parallel using a buffer in which NaCI was replaced by an isosmolar concentration of choline chloride. The uptake reaction was initiated by the addition of 50 b1 of L-[2,3,4-3Hj-glutamic acid to duplicate assay tubes. Tubes were mixed rapidly and returned to shaking water bath (25 'C) for 90 seconds. Uptake was terminated by rapid vacuum filtration using a Brandel cell harvester and Whatman GF/B filter sheets that had been presoaked overnight at 4�C in 25 mM Lglutamate. Test tubes and filters were rinsed rapidly three times with 2 ml aliquots of icecold normal or sodium-deficient buffer. Tissue trapped on filters was digested with 2 ml of 0.2 M NaOH (overnight), acidified with 1 ml of 0.5 M HCI, and assayed for tritium content in 10 ml of EcoLume scintillation fluid (ICN Biochemicals). Radioactivity was










quantified in an LKB 1214 liquid scintillation counter with a counting efficiency of approximately 45% as determined by a radium-226 standard. Sodium-dependent uptake was determined as the difference between uptake in normal versus sodium-deficient buffers.

For concentration-response studies, tissue fractions were preincubated for 5 min in the presence of phorbol-12,13-dibutyrate (PDBu), ranging in concentration between 300 nM - 10 gM, or its vehicle (0.01% acetone v/v), prior to determination of L[3H]glutamate uptake. A similar treatment protocol was used to evaluate possible effects by okadaic acid. For studies testing the interaction between PDBu and okadaic acid, tissue fractions were exposed to both drugs simultaneously for 5 min prior to determination of L-[3H]glutamate uptake. For studies of the interaction between PDBu and calphostin C, tissues were pretreated with calphostin C for 5 min prior to addition of PDBu.

Statistical Evaluation

For the results reported below, the n indicates the number of experiments

conducted on separate days. Data were expressed as the mean + S.E.M. To determine statistical significance, One-way Analysis of Variance (ANOVA) was used, followed, when appropriate, by Bonferroni's t test. Significance was accepted at a level of p5 0.05.










Results

Concentration-Dependence of PDBu-Induced Facilitation of L-I3H]Glutamate Uptake

As shown in Figure 3-1, a 5 min pretreatment with PDBu (300 nM -10 pM)

stimulated L-[3H]glutamate uptake in a concentration-dependent manner in GPV but had no significant effect on synaptosomal L-glutamate uptake. The maximal PDBu effect in GPV occurred at 10,4M (56 � 15% above control). Significant differences (p<0.05) between GPV and vehicle were observed at both 7 ,M (45 � 12% above control) and 10 MM PDBu (56 � 15% above control).

Kinetic Analysis of PDBu-Induced Facilitation of L-[H]Glutamate Uptake

In order to determine whether the PDBu-induced increase in L-[3H]glutamate uptake was due to a change in the number of active transporters (apparent V,. or a change in transporter affinity for L-glutamate (apparent K), the concentration dependence of L-[3H]glutamate uptake was examined after preincubation of tissue fractions with 10 MM PDBu for 5 min (Figure 3-2). Under these conditions, PDBu increased the V, for L-glutamate transport in GPV but not SYN and did not change the K. of L-[3H]glutamate uptake (Figure 3-2, panels A and B). For GPV (panel A), the V., of vehicle-treated controls was 254 � 23 pmol/mg protein/90sec. Following pretreatment with PDBu, the V, increased to 365 � 21 pmol/mg protein/90sec (p<0.05). In vehicletreated GPV tissue fractions the K. was 1.0 ,M � 0.58 MiM, and in PDBu-treated GPV the K. was 1.9 pM � 0.43 gM. In vehicle-treated control SYN tissue fractions (panel B),










the V.. was 1648 � 58 pmol/mg protein/90sec with a K. of 1.8 � 0.63 4M. In the presence of 10 MM PDBu, the V, was 1989 � 83 pmol/mg protein/90sec with a Km of

3.9 � 0.94 MM.

Specificity of Protein Kinase C Pathway for PDBu-Induced Facilitation of L[3H]Glutamate Uptake

To determine whether enhanced protein kinase C activity via PDBu was causing the increase in L-[3H]glutamate uptake, the effect of an irreversible protein kinase C inhibitor, calphostin C, (Kobayashi et al, 1989) was evaluated for its effectiveness in blocking the PDBu-induced enhancement in L-[3H]glutamate uptake. As shown in Figure 3-3 (panel A), calphostin C blocked 10 gM PDBu-enhanced L-[3H]glutamate uptake in a concentration-dependent manner. Calphostin C at concentrations of either 0.1 AM or 0.5 MM alone (panels A and B) did not affect L-[3H]glutamate uptake. For GPV (panel A), when 0.1 uM calphostin C was combined with 10 MM PDBu, L-['H]glutanate uptake was reduced from 46 � 4% above control (p<0.05) to 17 � 6% above control. The combination of 0.5 gM calphostin C with 10 MM PDBu further reduced transport from 46 � 4% above control to 6 � 2% above control (p<0.05). For SYN, no significant changes were produced by any of these treatments.

Effect of Okadaic Acid, a Phosphatase Inhibitor, on L-[3H]Gutamate Transport

The concentration-response curve for OKA is shown in Figure 3-4. In GPV, a 5 min pretreatment with OKA at concentrations of either 300 nM or 1000 nM significantly enhanced (p<0.05) L-[3H]glutamate uptake (24.4 � 2.3% above control and 32.6 � 5.0% above control, respectively). Pretreatment of SYN for 5 min with increasing








64

concentrations of OKA produced no significant change in L-[3H]glutamate uptake. When 10 gM PDBu was combined with either 300 nM or 1000 nM OKA for 5 min, values of L[3H]glutamate uptake enhancement remained the same as values obtained in the presence of 10 gM PDBu or 300 nM or 1000 nM OKA alone (data not shown).

Discussion

While specific cellular mechanisms for regulation of L-glutamate transport in vivo have not been reported, several endogenous factors, such as arachidonic acid and nitric oxide (Barbour et al., 1989; Pogun and Kuhar, 1993) and transporter phosphorylation (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997) have been proposed as possible regulators of transporter function. Consensus sites for phosphorylation by PKC are present in each of the recently cloned L-glutamate transporters (Kanai and Hediger, 1992; Pines et al., 1992; Storck et al., 1992, Fairman et al., 1995; Arriza et al., 1997). The existence of these putative phosphorylation sites suggests that L-glutamate transporters may be regulated by protein kinases and phosphatases. Primary astrocyte cultures and cell lines expressing a homogenous population of L-glutamate transporters have demonstrated that L-glutamate transport activity may be modulated by PKC activation (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997). In this study we used glial- and neuronalenriched fractions isolated by three-step density gradient centrifugation to investigate phosphorylation-dependent regulation of L-glutamate transport.

The first step in addressing the issue of a change in L-[3H]glutamate transport

activity due to phosphorylation was to expose GPV and SYN tissue fractions to increasing








65









*-- GPV * 60- SYN ,


840


(c20

0



0 2 4 6 8 10 PDBu Conc. (pM)






Figure 3-1. Concentration dependence for PDBu-induced facilitation of L-[3H]glutamate uptake. GPV and SYN fractions were preincubated for 5 minutes in the presence of PDBu or vehicle (0.0 1% acetone, v/v) then assayed for uptake in the presence of 5gM L[3H]glutamate. Data points represent averaged values (mean � SEM) from five experiments. Asterisks (*) denote a significant difference (p<0.05) between PDBu-treated samples and vehicle-treated controls as determined by one-way ANOVA followed by Bonferroni's t test.












400300


200 100


::" 0
0


- 2500 a 2000 1500 1000


500


0


0 20 40 60 80 100


0 20 40 60 80 100

L-[3H]Glutamate Conc. (pM)



Figure 3-2. Substrate-saturation curves for L-[3H]glutamate uptake in PDBu-treated tissue fractions. GPV (panel A) and SYN (panel B) fractions were incubated for 5 minutes in the presence of 1OtM PDBu or vehicle prior to measurement of uptake in the presence of L-[3H]glutamate (0.02gM - 100gM). Data points represent the mean � SEM from 3-4 experiments. Fitted curves were obtained by least squares regression analysis as described under Experimental Procedures. PDBu-treatment of GPV fractions caused significant enhancement (p<0.05) of L-[3H]glutamate uptake as indicated by one-way ANOVA.


























Calphostin C


0.1 0.5


- 0.1 0.5


0.1 0.5


- 0.1 0.5


- - 10 10 10


- - 10 10 10


Figure 3-3. Effect of calphostin C on PDBu-facilitated L-[3H]glutamate uptake. Aliquots (50,1) of GPV (panel A) or SYN (panel B) were preincubated for 5 minutes in the presence of calphostin C or vehicle (0.01% ethanol, v/v) followed by 5 minute treatment with PDBu or vehicle. All drug concentrations are given in micromolar, and dashes (-) indicate the use of drug vehicles. Data are plotted as the mean � SEM, (n=3) for each tissue fraction. Significant differences (p<0.05) between PDBu and either vehicle (*) or PDBu plus calphostin C (t) were determined by one-way ANOVA followed by Bonferroni's t test.


PDBu


0



















60 -


50403020lO-


I I
100 1000


10000


OKA Conc. (nM)









Figure 3-4. Concentration dependence for OKA-induced facilitation of L-[3H]glutamate uptake. GPV and SYN fractions were preincubated for 5 minutes in the presence of OKA or buffer then assayed for uptake in the presence of L-[3H]glutamate. Data are expressed as mean � SEM for five experiments. Asterisks (*) denote significant differences (p<0.05) between OKA-treated samples and vehicle-treated controls as determined by one-way ANOVA followed by Bonferroni's t test.


-- GPV
-1- -- SYN










concentrations of the PKC activator, PDBu. Tumor-promoting phorbol esters, such as PDBu, activate all known PKC isoenzymes directly, with the exception of the atypical PKCs, zeta and lambda, by binding at the DAG site of PKC (Nishizuka, 1984). Zeta and lambda isoforms are activated by phosphotidylserine but are not affected by DAG, phorbol esters, or calcium (Ono et al., 1989; Nakanishi and Exton, 1992). L-[3H]glutamate transport activity in GPV fractions responded to PDBu in a concentration-dependent manner (Figure 3-1). This increase in L-[3H]glutamate transport appears to be due to an increase in the number of active transporters, as GPV affinity for L-[3H]glutamate did not change following exposure to PDBu (1.0 � 0.44 gM vs. 1.9 � 0.52/zM), but the V, did (254 � 23 vs. 365 � 21 pmol/mg protein/90sec; p<0.05) (Figure 3-2). These findings contrast with those of SYN in which L-[3H]glutamate uptake was not altered significantly from control levels following PDBu treatment. These findings were substantiated further by substrate-saturation studies (Figure 3-2), wherein neither the affinity for L[3H]glutamate nor the maximal rate of L-[3H]glutamate transport in SYN changed significantly from control levels. Taken together, these results indicate a potential means of differential regulation of L-glutamate transport in GPV versus SYN fractions by PKCmediated phosphorylation.

Considering these results, it was necessary to confirm that PDBu-induced

facilitation of L-[3H]glutamate transport was occurring through activation of PKC. In order to evaluate the basis for PDBu activation of L-glutamate transport, tissue fractions were pretreated with calphostin C, prior to exposure to PDBu. Calphostin C is the most potent and specific inhibitor of PKC among the family of calphostins and blocks binding at








70

the DAG / phorbol ester binding site (Kobayashi el al., 1989; Tamaoki et al., 1990). At a concentration of I juM, it causes complete inhibition of the binding of 50 nM [3H]PDBu to PKC and subsequent inhibition of its activity (Kobayashi et al., 1989). Notably, calphostin C inhibited the PDBu-induced facilitation of GPV L-[3H]glutamate transport in a concentration-dependent manner (Figure 3-3). These results confirm that PDBu is activating PKC.

There are several possibilities as to why there is a difference between the two

vesicle fractions in response to PDBu pretreatment. First, there may be a difference in the size of the PKC "pool" between GPV and SYN. It is possible that the GPV PKC "pool" is simply larger than that of SYN, and as such, responds more robustly to exposure to a PKC activator like PDBu. This, of course, assumes that the PKC isoform(s) responsible for mediation of this response is actually present in both tissue fractions but at a lower level in SYN. This may not be the case. Like L-glutamate transporters, PKC isoforms have specific regional and cellular localizations. Certainly, homogenization has disrupted regional specificity. However, cellular specificity may be conserved, and SYN fractions may not contain PKC isoforms which are activated by phorbol esters. This does not mean that SYN L-glutamate transport cannot be modulated by phosphorylation, but that this potential modulation may take place through atypical PKCs, which are not activated by phorbol esters. Second, the ability of PKCs to interact with and be activated by membranes (phospholipids) and DAG presents a functional definition of these proteins. As such, binding to lipids is a critical and indeed well-studied phenomenon (see review: Epand and Lester, 1990). Even though general levels of phospholipids were evaluated








71

previously in GPV and SYN tissue fractions and found not to differ significantly (Daniels and Vickroy, 1998), it is possible that other important PKC activators, such as Ca2 , is present to a lesser degree in SYN versus GPV. Again, this would limit the ability of PDBu to enhance SYN L-glutamate transport. Third, it is possible that SYN L-glutamate transport may be modulated by phosphorylation but simply not by PKC. In addition to having consensus PKC sites, the cloned L-glutamate transporters also contain consensus sequences for protein kinase A (Kanai and Hediger, 1992; Pines et al., 1992; Storck et al., 1992; Fairman et al., 1995). Conradt and Stoffel (1997) reported that treatment of GLAST-expressing cells with phorbol ester decreased L-glutamate transport activity with phosphorylation occurring at a non-PKC consensus site.

With the finding that phosphorylation by PKC could modulate L-[3H]glutamate

transport, it seemed prudent to investigate the possible effect(s) that dephosphorylation by phosphoserine/phosphothreonine phosphatases could have on L-[3H]glutamate transport activity in GPV and SYN. These types of phosphatases can be differentiated from one another on the basis of differences in substrate specificities, dependence on divalent cations, sensitivities to specific inhibitors, and their catalytic subunits (Ingebritsen et al., 1983; Cohen, 1989). Based upon the four types of catalytic subunits, the phosphoserine/phosphothreonine-specific protein phosphatases have been divided into four main classes of enzymes: PPI, PP2A, PP2B, and PP2C (Ingebritsen et aL, 1983; Cohen, 1989).

Protein phosphatase 1 (PP 1) preferentially dephosphorylates the beta-subunit of phosphorylase kinase and is inhibited by two heat-stable inhibitor proteins, inhibitor-i










(which inhibits PP I after phosphorylation by cyclicAMP-dependent protein kinase) and inhibitor-2 (also termed "modulator," which inhibits PP I by impeding the substrate binding and by inducing a conformational change of the catalytic subunit) (see review: Bollen and Stalmans, 1992). Subcellular fractionation studies have demonstrated PP 1 activity in cytosolic, synaptosolic, synaptic plasma membrane and synaptic junction fractions (Shields et al., 1985; Dokas et al., 1990). These findings are consistent with a ubiquitous distribution of this phosphatase in brain. Alternatively, type 2 phosphatases preferentially dephosphorylate the alpha-subunit of phosphorylase kinase and are insensitive to inhibitor1 and inhibitor-2. Type 2 protein phosphatases are subdivided into three distinct classes based on their cationic requirements: PP2A, PP2B, and PP2C. PP2A is active in the absence of divalent cations. PP2B (i.e. calcium/calmodulin-dependent protein phosphatase or calcineurin) and PP2C are completely dependent on calcium and magnesium, respectively. The amount of PP2A activity in brain extracts is the highest of any tissues investigated (Ingebritsen et al., 1983), and there is approximately three times as much PP2A as PPI. As noted in several reviews (see: Cohen, 1989; Sim, 1991; Bollen and Stalmans, 1992; Shenolikar, 1994), substrate specificity of phosphatases may be controlled by a number of different regulatory subunits. These regulatory subunits direct phosphatase activity toward specific subcellular localizations and, therefore, toward specific substrates.

OKA is a PPI and PP2A inhibitor with a higher affinity for PP2A (IC50 = 0.1 nM) than for PPI (IC50 = 10 - 15 nM) (Cohen et al., 1990). Pretreatment of GPV and SYN fractions with OKA significantly enhanced L-[3H]glutamate transport, but only in GPV










fractions (Figure 3-4). Again, these results suggest a differential modulation of L[3H]glutamate transport between GPV and SYN tissue fractions. As to why there is a difference between the two tissue fractions, it is possible that GPV and SYN differ in the size of their "pools" containing PPI and PP2A or in the distribution of PPl and PP2A within these "pools." The latter may be true considering the results from exposure of the tissue fractions to OKA. GPV fractions may contain a higher level of PP2A, thus inhibiting this phosphatase, increases L-glutamate transport due to inhibiting dephosphorylation. However, regulatory subunits direct phosphatase activity toward specific subcellular localizations and therefore toward specific substrates. It is possible that following homogenization, this process is perturbated, so that phosphatases are no longer directed toward L-glutamate transporters, or conversely, that phosphatases are directed now inappropriately towards L-glutamate transporters.

Also of interest are the results from studies in which GPV and SYN fractions

were treated at the same time with both PDBu and OKA (data not shown). The increase in L-glutamate transport activity was not additive in either fraction. In other words the increased L-glutamate uptake in GPV following PDBu treatment was not enhanced further by additional treatment with OKA. SYN L-glutamate uptake remained at control levels, and GPV L-glutamate uptake increased to levels seen with either PDBu or OKA alone. This result suggests that PDBu- and OKA-induced facilitation of L-glutamate transport in GPV fractions is occurring through a common pathway.

In summary, we have demonstrated a significant increase in Na+-dependent highaffinity L-[3H]glutamate transport in glial versus neuronal tissue fractions. This increase is










consistent with activation of PKC and subsequent phosphorylation of GPV L-glutamate transporters. Furthermore, activity of GPV L-glutamate transporters could be enhanced by blockade of PP1 and PP2A phosphatases. Together, these results suggest that phosphorylation / dephosphorylation events differentially modulate glial and neuronal Lglutamate transport, and provide some insight into these potential regulatory mechanisms in CNS tissue. It is possible that modulation of L-glutamate transport depends upon the transporters cellular colocalization with specific PKC isoforms. Future studies need to address the type and levels of PKC isoforms within the glial- and neuronal-enriched fractions in order to further access this potential means of regulation.













CHAPTER 4
EFFECT OF AGING ON L-[3H]GLUTAMATE UPTAKE Introduction

Reuptake via sodium-dependent high-affinity transport systems in neurons and glial cells is the major mechanism for termination of L-glutamate synaptic actions (Nicholls and Attwell, 1990) and maintenance of its extracellular concentration below toxic levels (McBean and Roberts, 1985). Several studies have examined EAA uptake in tissue preparations from a variety of areas of the aging rat brain. In the neocortex, uptake of both L-['4C]glutamate into a preparation of synaptosomes and D-[3H]aspartate in crude homogenates was found to decline with age, although these changes appear to reflect differences between immature (2-4 months of age) and mature (10-30 months of age) animals (Strong et al., 1984; Meldrum et al., 1992). By contrast, no difference was observed in L-[3H]glutamate uptake into slices of frontal neocortex in a comparison of animals aged 6 and 24 months (Dawson et al., 1989). In the neostriatum, uptake both of L-[3H]glutamate (in crude synaptosomal fractions) and D-[3H]aspartate (in crude homogenates) was lower in mature rats (aged 10 months or more) than in immature animals, aged 6 months or less (Price et at., 1981; Wheeler and Ondo, 1986; Najlerahim et al., 1990). Additionally, Palmer and colleagues in 1994, using Fisher 344 x Brown Norway rats aged 3, 12, 24, and 37 months, assayed D-[3H]aspartate uptake in crude cortical, hippocampal, and neostriatal synaptosomes and found no significant changes with










increasing age. Results from investigating L-glutamate transport across aging in crude synaptosomes, however, may be misleading, since these preparations are contaminated with glial cells, which also accumulate L-glutamate.

Given the necessity of proper functioning of L-glutamate transporters, it is of

extreme importance to understand their regulatory mechanisms. The isolation of cDNAs for several sodium-dependent high-affinity L-glutamate transporters (GLAST, GLT-1, EAACI, EAAT4, EAAT5) has furthered the idea first raised by Casado and colleagues (1991, 1993) that phosphorylation may provide a regulatory mechanism for L-glutamate transport since cloned transporters contain putative protein kinase C phosphorylation sites (Kanai and Hediger, 1992; Pines et al., 1992; Storck et al., 1992; Fairman et al., 1995; Arriza et al., 1997). Immunological studies have localized GLAST and GLT-1 primarily to glial cells (Rothstein et al., 1994, 1995), though GLT-1 is present in certain neurons (Torp et al., 1994; Schmitt et al., 1996). EAAC I is located in neurons (Kanai and Hediger, 1992; Rothstein et al., 1994; Velaz-Faircloth et al., 1996). The cellular localization of EAAT4 and EAAT5 expression has not been determined as of yet.

Several reports within the last few years have investigated the potential regulatory role that phosphorylation via protein kinase C (PKC) may play in modifying L-glutamate transport activity (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997). L-glutamate transport by GLT- 1 and EAAC 1 appears to be increased following incubation with phorbol esters (Casado et al., 1993; Dowd and Robinson, 1996), while GLAST-mediated transport has been reported to decline (Conradt and Stoffel, 1997). However, since these studies were conducted in cultured cells, it is










impossible to use this approach to assess possible changes in phosphorylation-mediated regulation of L-glutamate transport as a function of aging. With the advent of an improved density gradient centrifugation technique with which glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) can be isolated from a rat brain homogenate (Daniels and Vickroy, 1998), it is now possible to investigate the potential effects that aging may have on L-glutamate transport and the role that phosphorylation may play in regulating transport function.

Experimental Procedures

Chemicals

L-[2,3,4-3H]-Glutamic acid (specific activity = 60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). Phorbol-12,13-dibutyrate was bought from Sigma Chemical Co. (St. Louis, MO, USA). Primary antibody against Excitatory Amino Acid Carrier I (EAAC 1) was a gift from Dr. Michael Kilberg (University of Florida, Gainesville, FL, USA), and primary antibody against GLT-I was a gift from Dr. Jeffrey Rothstein (Johns Hopkins University, Baltimore, MD, USA). Secondary antibody (horseradish peroxidase conjugated to protein A), as well as the horseradish peroxidase color reagent kit, were bought from Biorad (Hercules, CA, USA). All other chemicals were purchased from commercial vendors and were of the highest quality available.

Preparation of Tissue

Glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) were prepared as described previously (Daniels and Vickroy, 1998). Male Fisher 344 x Brown Norway Fl










rats (NIA) aged 5, 25, 31, and 37 months, weighing 350 - 450 g, were used throughout this study. Animals were housed in pairs on a 12-hr light/dark cycle with food and water available ad libitum under "pathogen-free" conditions for 5-10 days prior to use. Animals were transported to the laboratory approximately 15 hrs prior to use. For each assay, two rats of the same age were decapitated quickly with a small animal guillotine, and the brains were removed rapidly and placed upon an ice-cold glass surface. Cerebellar tissue was removed and discarded, while all remaining forebrain tissue (approximately 2.4 g) was placed in 30 ml of an ice-cold solution containing 0.32 M sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA). The tissue was homogenized gently with a Potter-Elvehjem tissue grinder (approximately 30 rpm) and centrifuged at 1000 x g for 10 mun (4�C) using a fixed-angle rotor (F28/50-DuPont). All subsequent centrifugation steps were conducted at 4�C. The resultant pellet was discarded, and the supernatant was split into four equal portions, which were diluted to 30 ml with an ice-cold solution containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM dithiothreitol and 20 mM HEPES (pH

7.4 at 4C). Hereafter, this solution is referred to as SEDH. Diluted aliquots of supernatant were centrifuged at 5000 x g (15 min), and resultant supematants were saved separately on ice. Each of the four tissue pellets were resuspended in 15 ml of ice-cold SEDH solution and centrifuged at 1000 x g for 10 min. The resultant pellets were discarded, while supernatants were combined with supernatants saved from the previous step. The four tubes, each containing approximately 45 ml of tissue homogenate, were centrifuged at 33,500 x g (20 min), and supernatants were discarded. Tissue pellets (four) were resuspended in 15 ml of ice-cold SEDH solution and gently transferred onto a three-










step discontinuous Percoll gradient (Oml each of 1.38%, 2.3%, and 4.6% Percoll in SEDH solution) with a Minipuls 2 (Gilson) peristaltic pump (flow rate = 0.88 mI/min). Tubes were centrifuged at 33,500 x g (10 min) with 15-min periods of linear acceleration to and deceleration from the top speed. The turbid layer between 0% and 1.38% Percoll was collected from all four tubes and combined into two aliquots, which were diluted to a final volume of 15 ml each with ice-cold SEDH solution. Aliquots were centrifuged at 1000 x g (20 min), and resultant supernatants were layered onto fresh three-step Percoll gradients as described above. Tubes were centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration (see above), and the turbid layer between 0% and 1.38% Percoll was collected from both tubes and combined into a single aliquot. The tissue aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution, centrifuged at 33,500 x g (20 min), and the resultant pellet was used as the GPV fraction. For the recovery of the SYN fraction, the turbid layer between 2.38% and 4.6% Percoll was collected from the initial discontinuous gradient. The four aliquots were diluted to a final volume of 15 ml (each) with ice-cold SEDH solution and centrifuged at 1000 x g (20 min). Pellets were discarded, and supernatants were layered onto a three-step Percoli gradient (see above) and centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration periods. The turbid layer between 2.38% and 4.6% Percoll was collected from each of the four tubes and combined into one aliquot. This aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution and centrifuged at 17,500 x g (20 min). The resultant pellet was designated as the SYN fraction.










L-[3H]Glutamate Uptake

The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 25�C). GPV and SYN were used immediately for uptake. Sodium-dependent L-[3H]glutamate uptake by GPV and SYN was measured by a filtration method modified from Divac and coworkers (1977). For the measurement of uptake, aliquots (50 Al) of tissue fractions (approximately I mg protein/ml) were added to cold glass culture tubes that contained 400 4l of a buffered solution containing (in mM) NaCI (140), KCI (5), CaC12 (1.0), MgCI2 (1.0), NaH2P04 (1.2), D-glucose (10), and HEPES (20) at pH 7.4 (25 �C). Sodium-independent uptake (blanks) was measured in parallel using a buffer in which NaCl was replaced by an isosmolar concentration of choline chloride. The uptake reaction was initiated by the addition of 50 gl of L-[2,3,4-3H]-glutamic acid to duplicate assay tubes. Tubes were mixed rapidly and returned to shaking water bath (25 0 Q for 90 seconds. For concentration-response studies, tissue fractions were preincubated for 5 min in the presence of phorbol-12,13-dibutyrate (PDBu), ranging in concentration between 1 /M - 100 4M, or its vehicle (0.01% acetone v/v), prior to determination of L[3H]glutamate uptake. Uptake was terminated by rapid vacuum filtration using a Brandel cell harvester and Whatman GF/B filter sheets that had been presoaked overnight at 4�C in 25 mM L-glutamate. Test tubes and filters were rinsed rapidly three times with 2 ml aliquots of ice-cold normal or sodium-deficient buffer. Tissue trapped on filters was digested with 2 ml of 0.2 M NaOH (overnight), acidified with 1 ml of 0.5 M HCI, and assayed for tritium content in 10 ml of EcoLume scintillation fluid (ICN Biochemicals). Radioactivity was quantified in an LKB 1214 liquid scintillation counter with a counting










efficiency of approximately 45% as determined by a radium-226 standard. Sodiumdependent uptake was determined as the difference between uptake in normal versus sodium-deficient buffers.

Western Blot Analyses

The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 25�C). Aliquots (100 ,1) were placed immediately in liquid nitrogen and stored at -80�C until needed. For immunoblotting of antibodies against EAAC 1 and GLT-1, aliquots (1 jug protein / 20 kil for GLT-1 or 5 jg protein / 20 Al for EAAC1) of GPV and SYN samples from animals aged 5, 25, 31, and 37 months were subjected to SDS-polyacrylamide gel electrophoresis (200V for 2.5 hr). The electrophoresed proteins were transferred onto polyvinylidene difluoride membrane (Immobilon P, Millipore) (1 OOA for 1.25 hr). Blots were incubated (1.5 hr) with blocking solution [5% non-fat carnation instant milk in Tris-Tricine solution (TBS/T) (10 mM Tris-HCl, 200 mM NaCI,

0.05% Tween-20)], then incubated in primary antibody (1:50 for EAAC1, 1:500 for GLT1) for 1.5 hr. Blots were washed five times (5 min each) with blocking solution, then incubated with horseradish peroxidase-conjugated protein A (1:5000) for 1 hr. The secondary antibody was removed, and blots were rinsed 5 times (5 min each) with blocking solution, followed by a 10 min wash with TBS/T. Protein bands were visualized via horseradish peroxidase color reagent development kit (Biorad). Protein Analyses

The amount of protein in GPV and SYN samples was determined by the Lowry protein assay (Lowry et al., 1951), using solutions of bovine serum albumin as standards.










Data Analyses

All values reported were the mean � SEM from five fractions isolated on separate experimental days. Statistical significance, accepted at a level of p<0.05, was determined via repeated measures analysis of variance (ANOVA), followed by Tukey's multiple pairwise comparison procedure.

Results

Effect of Aging on L-[3H]glutamate Uptake

As shown in Figure 4-1 (panel A), L-[3H]glutamate uptake values for GPV

fractions did not differ among animals at 5, 25, 31 or 37 months of age. In SYN fractions (panel B), L[3H]glutamate transport at 37 months was significantly reduced F3,,6=4.20, p<0.03) from 5 month values (712 � 31 vs. 562 � 40 pmol/mg protein/90sec). Effect of Aging on PDBu-induced Facilitation of L-[3Hglutamate Uptake Within age group evaluation

Another key issue to be answered was whether or not PDBu would facilitate an increase in L-[3H]glutamate transport within each of the four age groups examined (Figures 4-2 and 4-3). Figure 4-2 shows the PDBu concentration-response curves for GPV fractions obtained from 5 (panel A), 25 (panel B), 31 (panel C), and 37 (panel D) month-old animals. Significant increases in L-[3H]glutamate transport above control values occurred in each of the four age groups (F,46=97.38, p<0.001). PDBu concentrations of 3 gM and above facilitated L-[3H]glutamate transport in all age groups, while 1 gM PDBu yielded a significant increase above control values in 31 month-old animals. Likewise, SYN tissue fractions (Figure 4-3) also showed a PDBu-induced










facilitation of L-[3H]glutamate transport above control levels in each of the four age groups examined (F5,62=61.84, p<0.001). PDBu concentrations of 10,UM and above significantly facilitated transport at 5 months (panel A) and 37 months (panel D). However, in 25 month-old and 31-month old rats, the minimally effective concentrations of PDBu were 30 gM and 3 gM, respectively. (panels B and C). Between age groups evaluation

The effect of age on the concentration-dependence for PDBu-induced facilitation of L-[3H]glutamate uptake in GPV and SYN tissue fractions among the four age groups was examined also. As shown in the PDBu concentration-response curves of Figure 4-4, GPV fractions (panel A) did not exhibit any significant differences among age groups at any of the five PDBu concentrations. However, there were significant (p<0.05) differences between age groups at each of the PDBu concentrations in SYN fractions (panel B) (F3,16=6.90, p<0.01). At each PDBu concentration examined, values for uptake at 37 months differed significantly (*) from the 5 month old group (F3,16=12.91, 3.46, 7.95, 7.44, and 4.37 for 1, 3, 10, 30, and IOOK.SM, respectively; p<0.05 in all cases). A change in SYN L-[3H]glutamate uptake leading to a significant difference between 5 and 25 month old animals occurred at IOptM PDBu only (*) (F3,16=7.95, p<.01). A significant difference in transport values between 25 and 37 month old animals occurred at I qM (F3,16=12.91, p<0.001) and 30 /sM (*) (F3,16=7.44, p<0.01), while a significant difference between the 31 and 37 month old age groups occurred at 1 M (F3,16=12.91, p









Preliminary Western Blot Analyses for EAACI and GLT-1 Immunoreactivity

With the decreasing level of L-[3H]glutamate transport observed in SYN tissue fractions from aged rats (Figure 4-3 and Figure 4-4, panel B), preliminary Western blot analysis for the neuron-specific L-glutamate transporter EAAC 1 was undertaken to determine if the amount of EAACI protein was decreased in aged animals. When 5 Ig of total protein of GPV and SYN fractions obtained from animals aged 5, 25, 31, and 37 months were incubated with EAAC I antibody, GPV fractions exhibited no immunoreactivity above background. By comparison, positive immunostaining in SYN fractions was evident in each age group (n = 1, data not shown). One microgram of total protein from both GPV and SYN tissues from each of the four age groups was screened also against an antiserum for GLT-1. While GPV fractions from all age groups exhibited positive immunoreactivity for GLT- 1, the level appeared less at 5 months when compared to 25, 31, or 37 months (n = 1, data not shown). SYN tissue fractions from animals aged 5, 25, 31, and 37 months also reacted positively following incubation with primary antibody for GLT- 1.

Discussion

This study is the first to examine the potential effects of aging on L-[3H]glutamate uptake in GPV and SYN tissue fractions isolated from rat forebrain. In these studies aged Fisher 344 x Brown Norway F1 rats were used. Several strains and lines of rats have been used for gerontologic investigations, with Fisher 344 (F344) rats being used most frequently. However, one overriding reason for choosing the F344 x BN F1 instead of the F344 for this study is that the former have a much lower incidence of several major








85


300
A) GPV

--" 225


= 150


75


0




800 -B) SYN 600 - T


400


200


0
5 25 31 37 Age (in months) Figure 4-1. Effect of age on L-[3H]glutamate uptake. GPV (panel A) and SYN (panel B) fractions from animals 5, 25, 31, and 37 months of age were incubated in the presence of 51M L-[3H]glutamate for 90 seconds. Values are mean � SEM from five separate experiments. Asterisks (*) denote a significant difference (p<0.05) between the indicated age group and the 5 month-old group, as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.











!*
400 - A) 5 Months 400 - B) 25 Months

-- 300 - * 3001 *


200 -200
E*




0 100 -/ O /41 1 10
0 1 10 100 0 1 10 100
400 - C 31 Months 400 D) 37 Months



300 - 300 [/


200 200


100 /I I 100- / I I I
0 1 10 100 0 1 10 100

PDBu Conc. (gM)


Figure 4-2. Within group evaluation of the effect of age on the concentration dependence for PDBu-induced facilitation of L-['H]glutamate uptake in GPV tissue fractions. GPV fractions from animals aged 5 months (panel A), 25 months (panel B), 31 months (panel C), and 37 months (panel D) were preincubated for 5 minutes in the presence of PDBu (I gM - I100/tM) or vehicle (0. 0 1% acetone, v/v) then assayed for uptake in the presence of 5,uM L-['Hlglutamate. Data points represent averaged values (mean � SEM) from five separate experiments. Asterisks (*) denote a significant difference (p<0.05) between PDBu-treated samples and vehicle-treated controls as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.














1000


875 750 -


625 500 1000 875 750 625


500


// 1 I I

0 1 10 100 C) 31 Months










/ I I I
0 1 10 100


I A) 5 Months ,


1000 875 750 625 500 1000 875


750 625 500


PDBu Conc. (pM)


Figure 4-3. Within group evaluation of the effect of age on the concentration dependence for PDBu-induced facilitation of L-[3H]glutamate uptake in SYN tissue fractions. SYN fractions from animals aged 5 months (panel A), 25 months (panel B), 31 months (panel C), and 37 months (panel D) were assayed for L-[3H]glutamate uptake as described in Figure 4-2. Asterisks (*) denote a significant difference (p<0.05) between PDBu-treated samples and vehicle-treated controls as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.


77 I I I
0 1 10 100


D) 37 Months











0 1 10 100


B) 25 Months













400 - A) GPV


-. 300


200
�n -- 5 Months
o [-E] 25 Months � 100 A4 31 Months A--A 37 Months
-- 0- /
0 Z/I
0
0 1 10 100


1000 B) SYN ,


800



2 600



400- /
0 1 10 100

PDBu Conc. (gM)

Figure 4-4. Between group evaluation of the effect of age on the concentration dependence for PDBu-induced facilitation of L-[3H]glutamate uptake in GPV and SYN tissue fractions. GPV (panel A) and SYN (panel B) fractions from animals aged 5, 25, 31, and 37 months were assayed for L-[3H]glutamate uptake as described in Figure 4-2. Symbols denote a significant difference (p<0.05) between indicated groups (i =5 months vs 37 months, *=25 months vs 37 months, +=31 months vs 37 months, *=5 months vs 25 months) as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.








89

pathologic processes, including glomerulonephritis, retinal atrophy, and leukemia (Maeda et al., 1985; Lipman et al., 1996). Additionally, the F344 x BN F1 hybrid attains 500/o mortality at 146 weeks of age, which is much greater than the 103 weeks for the F344 rat (Lipman et al., 1996). As a model in which to study aging, this hybrid strain provides an increased period in which changes associated with an increase in age can be examined in the relative absence of disease.

The consistent finding on age-related changes in rodent brain PKC is that

translocation is impaired in cortical and hippocampal structures of aged when compared to young and mature rats (Friedman and Wang, 1989; Meyer et al., 1994; Pisano et al., 1991; Battaini et al., 1995). In the cortex this appears not to be related to changes in either phosphorylation or isozyme levels of PKC substrates such as histone or B-50, because levels of these are similar in adult and aged rats (Pisano et al., 1991; Battaini et al., 1995). In the hippocampus from aged rats, neither mRNA nor protein levels of PKC alpha, beta, or gamma isoforms appear to be modified (Battaini et al., 1995). It may be that age-related changes in brain membrane composition (Zidovetzkd and Lester, 1992), rather than a modification in a particular PKC isoform, are responsible for the impaired PKC translocation. It should be noted, however, that observations showing unmodified PKC activity using histone as substrate are partially in contrast with previous data from the cortex of Fisher 344 rats (Friedman and Wang, 1989) and from the cortex and hippocampus of Sprague-Dawley rats (Battaini et al., 1990). It is possible that strainrelated differences are responsible for these conflicting data, since strain-related








90

differences during aging are known to occur in rats in other neurochemical (Peterson and Cotman, 1989) and electrophysiological parameters (Potier et al., 1993).

Before determining whether or not age would alter PKC activation and ultimately L-[3H]glutamate uptake, it was necessary to ascertain if L-[3H]glutamate uptake, without drug treatment, was changing with increasing age. Figure 4-1 shows that baseline L[3H]glutamate uptake did not change in the GPV fraction with increasing age (panel A). For SYN fractions (panel B), values for L-[3H]glutamate uptake declined with advancing age. However, it was not until 37 months of age that decreasing levels of L-[3H]glutamate uptake reached significance from 5 month old animals. This finding, of a decrease in SYN L-[3 H]glutamate uptake with increased aging, differs from previous reports (Dawson et al., 1989; Palmer et al., 1994). These disparate findings may be attributable to the use in prior studies of crude synaptosomal preparations, almost certainly contaminated with glial cells, which can accumulate L-glutamate. Additionally, many of these prior studies, which indicated no effect of aging on L-glutamate transport, used a broad age range of animals, sometimes only two age groups. We circumvented these issues by using four age groups of animals and tissue fractions isolated by an improved three-step density gradient centrifugation technique (Daniels and Vickroy, 1998).

In the present study the ability of PKC activation, via PDBu, to modulate L[3H]glutamate uptake by GPV and SYN fractions was tested in each age group (5, 25, 31, and 37 months) (Figures 4-2 and 4-3). Each of the four age groups examined, for both GPV and SYN, showed increased L-[3Hjglutamate uptake in response to increasing










concentrations of PDBu (p<0.001). These findings illustrate that, irrespective of animal age, PKC can respond to activation by PDBu.

Differences between age groups in their responsiveness to a particular PDBu

concentration in facilitating L-[3H]glutamate uptake were examined also in GPV and SYN tissue fractions obtained from animals aged 5, 25, 31, and 37 months (Figure 4-4). Among the six PDBu concentrations used in this study, no differences were evident among the four age groups with regard to PDBu-induced enhancement of L[3H]glutamate uptake by GPV fractions (panel A). However, differences among the SYN fractions from different age groups were noted at all PDBu concentrations (panel B). Notably, the ability of PKC at 37 months to respond to PDBu activation by increasing L[3H]glutamate uptake differed significantly from 5 month-old animals at each PDBu concentration. Taken together, these results show that GPV fractions isolated from animals of increasing age respond to PDBu. GPV fractions responded to PDBu in a concentration-dependent fashion. The four age groups exhibited no differences in their response to any particular PDBu concentration for facilitation of L-[3H]glutamate uptake. However, SYN fractions undergo a significant decrease with advancing age in their ability to respond to PDBu activation. Though SYN fractions can respond to PDBu in a concentration-dependent fashion, this response, like baseline L-[3Hjglutamate uptake, declines significantly by 37 months. Therefore, irrespective of the age of animals, PKC can respond to activation by PDBu; however, as baseline L-[3H]glutamate uptake declines with increasing age, so too does the magnitude of the PDBu-induced response. One possible explanation for this may be that all the available sites for phosphorylation on the










SYN transporters already are phosphorylated. If this is the case, PDBu exposure would be incapable of elliciting a response.

Western blot analyses were carried out to determine whether protein levels of Lglutamate transporters were changing with increasing age. Preliminary studies of EAAC 1, an L-glutamate transporter in neurons (Kanai and Hediger, 1992; Rothstein et al., 1994; Velaz-Faircloth et al., 1996), showed no remarkable changes across the four age groups (n = 1, data not shown). The EAAC I antibody used in the Western blots was characterized by expression of cDNA for EAAC I in Human Embryonic Kidney cells (Matthews et al., 1997). Specificity was determined by peptide inhibition in the aforementioned cell line as well as tissue (Matthews et al., 1997). From our preliminary studies, it appears that levels of EAAC 1 protein do not change across ages and, therefore, cannot account for the significant decrease in L-[3H]glutamate uptake. Preliminary results with GLT- I (n = 1), an L-glutamate transporter located in glial cells (Danbolt et al., 1992; Rothstein et al., 1994) and some subsets of neurons (Torp et al., 1994; Schmitt et al., 1996), indicated positive immunoreactivity in both GPV and SYN tissue fractions. Surprisingly, immunoreactivity in SYN tissue fractions was greater than GPV tissue fractions at 5, 25, and 37 months (data not shown). Again, these results offer no ready explanation for the observed differences in L-[3H]glutamate uptake by GPV and SYN tissue fractions from animals of different ages.

This study has shown that there are differential effects of aging between GPV and SYN tissue fractions. The decline in SYN response to PDBu may result from a lack of available phosphorylation sites for the PDBu-activated PKC or it may be related to an age-








93

induced impairment of PKC. As discussed previously, the consistent finding on agerelated changes in PKC is impaired translocation. Whether this decline is due to reduced levels of PKC, its translocation, or diminished level of cofactors, such as membrane phospholipids or diacylglycerol, remains to be determined.















CHAPTER 5
CONCLUSIONS

Limitations of the Technique

While the first two aims of this project could have been addressed using cultured cells, we choose not to use such techniques, due to the potential for cell culturing to transform, alter, or negate intracellular regulatory mechanisms. Instead, we developed and verified, via morphological, biochemical, and protein analyses, an improved method of cell separation by three-step Percoll density gradient centrifugation. This technique yields functionally viable glial- and neuronal-enriched fractions which can be used for direct investigation of potential L-glutamate transporter regulatory mechanisms and determinations of aging effects under identical conditions at the same time. One important advantage in using GPV and SYN is that the diffusional barriers are not as prohibitive as in other preparations such as brain slices. This feature is particularly important in experiments from Chapters 3 and 4, because the site of action of the drugs used was intracellular. Therefore, the drugs needed to permeate the GPV and SYN membranes. Caution should be used, however, when relating GPV and SYN experimental results to situations where a greater degree of anatomical connectivity exists, such as in brain slices or in vivo.




Full Text

PAGE 1

PROPERTIES OF HIGH-AFFINITY L-GLUTAMATE TRANSPORT IN GLIALAND NEURONAL-ENRICHED FRACTIONS FROM RAT BRAIN By KELLYE K. DANIELS 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 1997

PAGE 2

This work is dedicated to Maurice and Alta Bennett, Maggie Daniels, and Louise Ferrell, with great love and fond remembrance.

PAGE 3

ACKNOWLEDGMENTS During my time as a graduate student at the University of Florida, Dr. William G. Luttge and Dr. Mary Jo Koroly gave me exceptional guidance and support; such generosity is rare and truly appreciated. I extend much gratitude to the Center for the Neurobiology of Aging and to the Bryan W. Robinson Neurological Foundation, Inc., for financial support. iii

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGMENTS iii ABSTRACT v CHAPTERS 1 BACKGROUND AND SIGNIFICANCE 1 2 ISOLATION OF GLIAL AND NEURONAL FRACTIONS 22 Introduction 22 Experimental Procedures 24 Results 32 Discussion 47 3 EFFECT OF PHO SPHORL Y ATION ON L-[ 3 H]GLUTAMATE UPTAKE 53 Introduction 53 Experimental Procedures 57 Results 62 Discussion 64 4 EFFECT OF AGING ON L[ 3 H] GLUT AMATE UPTAKE 75 Introduction 75 Experimental Procedures 77 Results 82 Discussion 8? 5 CONCLUSIONS 93 REFERENCES 98 BIOGRAPHICAL SKETCH 1 16 iv

PAGE 5

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 PROPERTIES OF HIGH-AFFINITY L-GLUTAMATE TRANSPORT IN GLIALAND NEURON AL-ENRICHED FRACTIONS FROM RAT BRAIN By Kellye K. Daniels December, 1997 Chairman: Thomas W. Vickroy, Ph.D. Major Department: Neuroscience Prior studies have established that termination of L-glutamate-mediated neurotransmission occurs through high-affinity transporters located on glial cells and neurons. For in vitro studies of L-glutamate transport, an improved three-step density gradient centrifugation technique was developed to isolate glial(glial plasmalemmal vesicles or GPV) and neuronal(synaptosomes or SYN) enriched vesicles from rat brain homogenates. Morphological, biochemical, and Western blot analyses confirmed the high degree of separation between GPV and SYN tissue fractions. GPV and SYN tissues were used to evaluate L-glutamate uptake, to address phosphorylation / dephosphorylation events as possible regulatory mechanisms of Lglutamate uptake, and to determine to what extent aging may alter these mechanisms. L[ 3 H]Glutamate uptake by GPV and SYN revealed remarkedly similar properties. Kinetic v

PAGE 6

studies indicated a more rapid initial rate for uptake in GPV, as well as a slightly higher substrate affinity. Conversely, SYN fractions exhibited a 2to 3-fold greater capacity for L-[ 3 H]glutamate uptake. In studies addressing the effects of phosphorylation on L[ 3 H]glutamate uptake, GPV and SYN tissue fractions responded differentially following incubation with phorbol-12,13-dibutyrate (PDBu), an activator of protein kinase C (PKC), an enzyme which phosphorylates serine/threonine residues. GPV L-[ 3 H]glutamate uptake increased significantly above control values after incubation with PDBu, whereas uptake values in SYN fractions did not. Likewise, L-[ 3 H]glutamate uptake in GPV significantly increased above control values when fractions were incubated with okadaic acid, a phosphoserine / phosphothreonine phosphatase inhibitor, whereas L-[ 3 H]glutamate uptake in SYN fractions did not change from control levels. In GPV and SYN fractions obtained from animals aged 5, 25, 31, and 37 months, only L-[ 3 H]glutamate uptake in SYN fractions from 37 month old animals significantly decreased from uptake values obtained from 5 month old animals. When fractions were incubated with PDBu, L-[ 3 H]glutamate uptake increased dose-dependently across age groups in both fractions. This response in SYN fractions, however, decreased with increasing age. This study has demonstrated that while both GPV and SYN fractions transport Lglutamate with high-affinity and in a saturable manner, they differ in response to kinases and phosphatases, ultimately indicating that phosphorylation / dephosphorylation events may more tightly regulate GPV rather than SYN L-glutamate transport and that SYN Lglutamate transport is more vulnerable to the aging process. vi

PAGE 7

CHAPTER 1 BACKGROUND AND SIGNIFICANCE Properties of EAA Transporters Biochemical Properties and Functions The amino acids L-aspartate and L-glutamate represent the major excitatory neurotransmitters in the mammalian brain (Krnjevic, 1970;Fonnum, 1984). High-affinity transporters for these excitatory amino acids (EAA) are essential for recapturing the neurotransmitter L-glutamate, thereby terminating its extracellular actions and maintaining a high concentration gradient between the extracellular space and the cytosol of nerve terminals (Kanner and Schuldiner, 1987; Nicholls and Attwell, 1990). The ability of L-glutamate to cause neuronal damage and cell death, however, poses a special requirement for an efficient uptake system (Choi, 1988, 1994). For several years it has been known, primarily through the work of Kanner and collaborators, that L-glutamate uptake is electrogenic and dependent on external Na* and internal K + ions (Kanner and Schuldiner, 1987; Danbolt et al., 1990). Detailed studies on the stoichiometry of L-glutamate uptake have been carried out on Muller cells, a specialized form of glial cells (Attwell et al, 1991; Bouvier et al, 1992). Since these cells do not express L-glutamate receptors, the current elicited by L-glutamate provides a convenient index of electrogenic uptake. Analyses based on whole cell patch clamp recordings indicate that L-glutamate (carrying one net negative charge at physiological 1

PAGE 8

2 pH) is transported into the cell together with two sodium ions and that the return of the carrier to the outside of the cell is coupled with an outward transport of one potassium ion and one hydroxyl anion. The maintenance of a low level of external L-glutamate is one obvious function of the L-glutamate transporter. This function is crucial as L-glutamate becomes neurotoxic when its extracellular concentration exceeds a certain level (Choi et ai, 1987; Beal, 1992a, b). The concentration at which L-glutamate exerts excitotoxic actions is hard to determine, due to the efficiency of uptake and the lack of noncompetitive, irreversible blockers. Estimates range from several hundred micromolar (Nicholls and Attwell, 1990) to as low as 1 micromolar (Frandsen and Schousboe, 1990). Neuronal toxicity is a consequence of excessive membrane depolarization associated with swelling and an accumulation of calcium recruited from the extracellular space or from intracellular stores. The importance of glial L-glutamate transporters for reducing L-glutamate neurotoxicity is suggested by experiments showing that neurons are much more sensitive to L-glutamateinduced toxicity when grown alone than when grown in co-culture with glia (Rosenberg et ai, 1992). In conditions such as hypoxia and ischemia, the driving force for L-glutamate uptake collapses due to the breakdown of the electrochemical sodium and potassium gradients, and L-glutamate transport is compromised or reversed (Szatkowski and Attwell, 1994). In fact there is increasing evidence to suggest that the extracellular overflow of L-glutamate in ischemic and anoxic conditions primarily reflects inadequate or reversed L-glutamate transport and that exocytotic release plays a minor role (Nicholls and Attwell, 1990).

PAGE 9

3 L-Glutamate dysfunction has been implicated in other neurologic disorders such as amyotrophic lateral sclerosis (ALS) (Plaitakis, 1990; Rothstein et ai, 1992; Shaw et ai, 1994; Rothstein et al, 1995; Bristol and Rothstein, 1996; Leigh and Meldrum, 1996). High-affinity, sodium-dependent transport of L-glutamate was found to be markedly impaired in ALS patients in the spinal cord, motor cortex, and somatosensory cortex (Rothstein et ai, 1992). These are the brain regions most affected in ALS. The changes were specific for ALS, as they were not observed in Huntington's disease, Alzheimer's disease, or nonneurologic disease controls. This finding led Rothstein and colleagues (1992) to conclude that defects in the clearance of extracellular L-glutamate, secondary to faulty transporter function, could lead to neurotoxic levels of extracellular L-glutamate and thus be pathogenic in ALS. Regional / Cellular Localization Three sodium-dependent transporter subtypes have been identified in rat forebrain: L-glutamate/L-aspartate transporter (GLAST) (Storck etai, 1992), glial transporter1 (GLT-1) (Pines etai, 1992), and excitatory amino acid carrier1 (EAAC1) (VelazFaircloth et ai, 1996). Three similar transporters have been cloned from human brain, and the provisional human homologs have been termed excitatory amino acid transporters (EAAT) -1,-2, and -3, corresponding with greater than 90% sequence homology to GLAST, GLT-1, and EAAC1, respectively (Arriza et ai, 1995). A fourth subtype, EAAT4, has been identified in human cerebellum (Fairman et ai, 1995), and a fifth subtype, EAAT5, has been identified in human retina (Arriza et ai, 1997).

PAGE 10

4 With regard to the cellular and regional locations of L-glutamate transporters, results from immunocytochemistry (Lehre et al, 1995) and in situ hybridization (Torp et al, 1994) indicate that there are pronounced regional differences in the expression of the transporters. These studies have mapped GLAST mRNA to the Purkinje cell layer of the cerebellum, apparently associated with Bergmann glia (Storck et al, 1992), as well as to the cortex and hippocampus (Rothstein et al, 1994). Other studies, using in situ hybridization for the carboxyl-terminal domains of each transporter, initially reported a low level of GLAST immunoreactivity in both neurons and astroglia (Rothstein et al, 1994). Amino-terminal oligopeptide antibodies localized GLAST to astrocytes (Rothstein et al, 1995; Lehre et al, 1995; Schmitt et ai, 1997). In addition, GLT-1 is present in astrocytes throughout the brain, with predominant expression in telencephalic structures, including the hippocampus, neocortex, and striatum (Danbolt etal, 1992; Rothstein et al, 1994). However, both radioactive as well as non-radioactive in situ hybridization techniques have shown that selected neuronal populations (i.e., thalamus, hypothalamus, and pyramidal cells of the hippocampus) express GLT-1 mRNA (Torp et al, 1994; Schmitt et al, 1996). The observations that GLT-1 mRNA is expressed in or by neurons seems to be at variance with other studies (Danbolt et al, 1992; Rothstein et al, 1994; Lehre et al, 1995) which failed to detect GLT-1 protein in immunoelectron microscopic preparations. Schmitt and colleagues (1996) maintain that a failure to detect GLT-1 protein in neuronal membranes may be a result of the protein concentration being below the method's detection limits and that at the electron microscopic level the extremely

PAGE 11

dense immunoreactivity of the neuropil in general makes it impossible to distinguish whether labeled membranes belong to neuronal terminations or to glial cells. EAAC 1 mRNA is abundantly expressed in the pyramidal layer of the hippocampus (regions CA1-CA4), the granule cell layer of the dentate gyrus, the granule cell layer of cerebellum and layers II-IV of cerebral cortex (Kanai and Hediger, 1992). In their original report, Kanai and Hediger (1992) concluded on the basis of the distribution of hybridizing mRNA that this transporter was primarily neuronal. Findings by others (Rothstein et al, 1994; Velaz-Faircloth et al, 1996) support this conclusion. The cellular locations of the more recently cloned EAAT4 (Fairman et al, 1995), which is expressed in the cerebellum, and EAAT5 (Arriza et al., 1997), which is expressed in the retina, have not been published. In a study by Rothstein and colleagues (1995), anti-oligopeptide antibodies generated to carboxyland amino-terminal sequences, specific for each human Lglutamate transporter, were used to determine the nature of the L-glutamate transport defect in ALS. The conclusion of this study was that a remarkable loss of astroglial GLT1 immunoreactive protein, restricted to motor cortex and spinal cord, was responsible for the defect. Notably, GLAST, also localized to glia, was unaffected. EAAC1 immunoreactive protein was affected only modestly. When mRNA levels of control versus ALS tissue was examined, there were no quantitative changes in mRNA for EAAT1, EAAT2, or EAAT3 in ALS motor cortex, even in patients with a large loss of

PAGE 12

6 EAAT2 protein and decreased tissue L-glutamate transport (Bristol and Rothstein, 1996). These studies suggest that abnormalities in EAAT2 may be due to translational or posttranslational processes. Regulatory Mechanisms While many of the mechanisms that influence L-glutamate transporters in vivo are probably unknown, several factors such as arachidonic acid (AA) (Barbour etal, 1989), nitric oxide (NO) (Pogun and Kuhar, 1993), and transporter phosphorylation (Casado et al, 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997) have been noted as possible regulators of transporter function. There is evidence that AA effects Lglutamate uptake in glial cells and nerve terminals (Barbour et al, 1989; Zerangue et al, 1995). In Xenopus oocytes injected with cRNAs encoding the human excitatory transporters EAAT1-3, micromolar levels of AA significantly reduced L-glutamate uptake mediated by EAAT1, while transport mediated by EAAT2 increased more than two-fold (Zerangue et al, 1995). AA had no effect on EAAT3 transport (Zerangue et al, 1995). These differential effects could constitute a regulatory mechanism, but would be nonselective as several sodium-dependent uptake systems are modulated (i.e., inhibited) by AA (Rhoads et al., 1983), including the uptake systems for Y-aminobutyric acid (Chan et al, 1983) and glycine (Zafra et al, 1990). Nevertheless, AA-mediated regulation of Lglutamate transport may have physiological relevance during the induction of long term potentiation, which is associated with an increased production of AA (Bliss and Collingridge, 1993). It has previously been reported (Herrero et al, 1992) that AA may increase synaptic strength by stimulating L-glutamate release, acting presynaptically in

PAGE 13

concert with metabotropic L-glutamate receptors. A simultaneous inhibition of Lglutamate uptake would have a synergistic effect. Interestingly, L-glutamate uptake has been found to be inhibited also by NO (Pogun and Kuhar, 1993; Pogun et al, 1994), which, like AA, has been proposed to act as a "retrograde messenger" during the induction of LTP. NO is thought to be a mediator of L-glutamate-induced neurotoxicity because inhibitors of NO synthase prevent L-glutamate neurotoxicity (Dawson et al, 1991, 1993). Each of the cloned L-glutamate transporters contains consensus sequences for phosphorylation by protein kinase C (PKC) (Kanai and Hediger, 1992; Pines et al, 1992; Storck et al, 1992; Fairman et al, 1995). The existence of putative phosphorylation sites indicates that L-glutamate transporters may be regulated by protein kinases and phosphatases. The finding that L-glutamate transport activity (V,^, but not KJ is increased in primary astrocyte cultures, after incubation of the cells with phorbol esters (PKC activators) (Casado et al, 1991), suggests that the putative phosphorylation sites are physiologically relevant. The first direct evidence that PKC-mediated phosphorylation is involved in the regulation of L-glutamate transport was provided by Casado et al (1993). Using antibodies directed against biochemically purified GLT-1, this group demonstrated that the amount of phosphorylation correlated with phorbol ester-induced stimulation of L-[ 3 H]glutamate transport activity. Additionally, by site-directed mutagenesis of GLT-1 and subsequent transfection into HeLa cells, it was shown that the effect of the phorbol ester was dependent on serine 113, the likely biologically relevant phosphorylation site.

PAGE 14

8 Other groups have further established that phosphorylation of brain L-glutamate transporters modulates transport function (Dowd and Robinson, 1996; Conradt and Stoffel, 1997). Preincubation of a subline of C6 glioma cells with phorbol ester, which endogenously express EAAC1 -mediated L-glutamate transport, caused a significant increase in L-glutamate transport (Dowd and Robinson, 1996). The findings that PKC activators increasing L-glutamate transport in GLT-1 and EAAC1 are in direct contrast with results reported by Conradt and Stoffel (1997) for GLAST. In this case treatment of GLAST-expressing cells with phorbol ester decreased L-glutamate transport activity with phosphorylation occurring at a non-PKC consensus site. The fact that L-glutamate transport appears to be subject to elaborate regulatory mechanisms strongly suggests an important role for L-glutamate transporters in brain function, possibly through mechanisms that remain to be discovered. Glial Plasmalemmal Vesicles and Synaptosomes Prior studies have established that termination of L-glutamate-mediated excitatory transmission in the mammalian central nervous system occurs through high-affinity transporters localized to both glial cells (i.e., astrocytes and oligodendrocytes) (Sender and Thompson, 1974) and neurons (Iversen, 1973; Kuhar, 1973). While L-glutamate transporter processes of neurons and astrocytes have been studied in brain slices, cultured cells and rat brain membrane preparations, a verified detailed comparative biochemical analysis of the neuronal and astroglial L-glutamate transporter from the same brain region under identical conditions is still lacking. The three-step density gradient centrifugation method outlined in Chapter 2 will not only allow for the characterization of L-glutamate

PAGE 15

9 transport in membrane-encapsulated vesicles derived from separated glial and synaptosomal populations from the same brain region at the same time, but will also circumvent many of the problems inherent to other techniques. For example, whereas in vivo situations as well as brain slices maintain circuitry, this greatly complicates the understanding of the degree to which glial versus neuronal populations contribute to Lglutamate uptake and negate any ability to ascertain the selectivities of inhibitors for glial versus neuronal transporters. The convenience of a density gradient centrifugation method is greater than that of primary cell culture procedures in that same-day animal sacrifice yields same-day results. Tissue does not have to be harvested and grown in media for varying lengths of time prior to use, plus an animal of any age may be used in this methodology. While cell culture techniques are certainly advantageous for the separation and investigation of a homogeneous cell population, it is possible that the tissue culture process may alter intracellular cascades for transporter regulation or translocation properties. Additionally, our density gradient centrifugation technique is superior to conventional synaptosomal isolation procedures, as will be shown by data to follow, because it minimizes glial contamination. In view of this, one must question the validity of published results from studies in synaptosomes which have been attributed specifically to neuronal-derived vesicles, including kinetic properties and pharmacological properties of L-glutamate transport inhibitors. Glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) are detached, sealed glial and synaptic nerve terminals, respectively, which have been separated by differential and density-gradient centrifugation. Both maintain ionic gradients and various membrane

PAGE 16

10 properties of central nervous system (CNS) cells and retain the machinery needed to accumulate neurotransmitters, as detailed in numerous reports (Dunkley et al, 1986; Ferkany and Coyle, 1986; Bridges et al., 1991; Fykse and Fonnum, 1991; Nakamura et al., 1993). Nakamura and coworkers (1993) used electron microscopy to examine the subcellular fractions prepared by Percoll density gradient centrifugation. It was found that the vast majority of the membrane components of the GPV fraction contained two types of vesicles, a small (0. 15-0.2 microns in diameter) spherical population and a large (0.30.8 microns in diameter), agranular, and irregularly shaped population. As of yet, it has not been determined whether both populations retain functional properties. It should be noted, however, that minute amounts of disintegrated SYN, mitochondria, and membrane fragments with postsynaptic densities were observed. Likewise, SYN have been studied through electron microscopy (Dunkley et al., 1986) and found to contain an abundance of small diameter vesicles, with plasma membranes and intrasynaptosomal mitochondria intact. Phosphorylation / Protein Kinase C One important component in various signal transduction pathways is the phosphorylation of a substrate, and the substrate's resultant activation or inactivation (Hardie, 1990; Barford, 1991). Protein phosphorylation is the covalent addition of inorganic phosphate to specific amino acids by protein kinase enzymes. There are two major groups of protein kinases, those that phosphorylate serine and threonine residues and those that phosphorylate tyrosine residues. The first demonstration of a direct

PAGE 17

11 regulation of a neurotransmitter transporter by phosphorylation was published by Casado and colleagues (Casado et ai, 1993). This group reported that a purified L-glutamate transporter from pig brain was phosphorylated by protein kinase C (PKC), predominantly at serine residues. When C6 glioma cells were exposed to 12-0-tetradecanoylphorbol-13acetate (a stimulator of PKC activity), an approximate 2-fold increase in L-glutamate transport was observed within 30 minutes. In view of this first published report, our studies focused on the ability of PKC to stimulate L-glutamate transport in glial-derived vesicles and neuronal-derived vesicles. The term "protein kinase C" encompasses an eleven member family of serine/threonine-specific protein kinases which have been identified functionally by common enzymatic properties, including phorbol ester binding, phospholipid-dependent kinase activity and common structural features (see review: Stabel and Parker, 1991). The PKC cDNA clones first isolated and their corresponding polypeptides are now called alpha, beta,, betaj, and gamma isoforms. Due to their structural organization, PKC alpha, beta 1; beta 2 , and gamma define the "class I" PKC enzymes, those dependent upon Ca 2+ . This is in contrast to the "class II" enzymes, delta, epsilon, eta, eta', and theta, which are distinguished by their structure, enzymatic properties, and Ca 2+ -independence (Stabel and Parker, 1991). Class III or atypical PKC isozymes include zeta and lambda (Nishizuka, 1992). These isozymes are phospholipid-dependent, but Ca 2+ independent and do not bind phorbol esters (Nishizuka, 1992). The ability of PKCs to interact with and be activated by membranes (phospholipids) and diacylgylcerol (DAG) presents a functional definition of these

PAGE 18

12 proteins. As such, binding to lipids is a critical and indeed well studied phenomenon (see review: Epand and Lester, 1990). This interaction, which has been most clearly documented for the alpha, beta, and gamma enzymes, appears to be minimally a two-step process. The first step is the formation of a ternary complex of enzyme, Ca 2+ , and phospholipid; the second step in association entails binding of DAG (or phorbol ester), which through conformational changes leads to activation. The ability of phorbol esters to activate PKC through binding at the DAG site has aroused much interest. Phorbol esters bind to the CI domain of PKC, a domain that contains a highly conserved cysteine-rich repeated sequence found in all isoenzymes (Ono et al, 1988). In addition sustained activation with phorbol esters selectively depletes PKC that is degraded by proteolytic enzymes (Young etal, 1987). Tumor-promoting phorbol diesters, such as 12-0tetradecanoylphorbol 13 -acetate (TP A, sometimes referred to as PMA) and phorbol 12,13-dibutyrate (PDBu) activate all known PKC isoenzymes in vivo, with the exception of class III PKC isoenzymes zeta and lambda (Nishizuka, 1992). To date no phorbol ester has exhibited PKC isozyme-selective activity in vivo (Kiley and Jaken, 1990; Roivainen and Messing, 1993; Kiley et al, 1994). Notably, however, in vitro work by Ryves and colleagues (1991) demonstrated that the proinflammatory, non-tumor promoting phorbol ester, 12-deoxyphorbol-13-O-phenylacetate-20-acetate (DOPPA), was a PKC^-specific agonist. DOPPA effectively stimulates PRC^i kinase activity at a concentration of 20 nM (10 ng/ml), but does not activate PKC alpha, gamma, delta, or epsilon at concentrations up to 2^M.

PAGE 19

13 Inhibitors at the DAG binding site have been isolated from a soil fungus (lida et al, 1989; Kobayashi et al, 1989). These compounds have a multi-tiered quinone structure and are termed calphostins. The most potent of these compounds is calphostin C, which has an IC 50 of 50 nM. At a concentration of 1//M, it causes complete inhibition of the binding of 50 nM [ 3 H]phorbol dibutyrate to PKC and subsequent inhibition of its activity (Kobayashi et al, 1989). IC 50 values greater than 28//M are necessary for calphostin C to inhibit cyclic AMP-dependent protein kinase, cyclic GMP-dependent protein kinase, or pp60 v " src protein tyrosine kinase (product of the src oncogene) (Kobayashi et al, 1989; Tamaoki efa/., 1990; Brunse/a/., 1991). Dephosphorylation / Serine/Threonine Protein Phosphatases While the role and influence of phosphorylation in the CNS is becoming more and more clear, the influence of protein dephosphorylation has yet to be investigated in detail. However, it does seem reasonable to assume that dephosphorylation is equally as important as phosphorylation in nervous system function and that protein phosphatases, enzymes which remove inorganic phosphate from a substrate, may play a key role in the regulation of such cellular processes as neurotransmitter release, ion fluxes, receptor availability, etc. (Nestler and Greengard, 1984; Cohen, 1989; Shenolikar and Nairn, 1991). This lack of defined, precise roles for protein phosphatases is primarily due to the rather ubiquitous nature of protein phosphorylation as a regulatory mechanism, such that the use of general phosphatase probes is likely to have a myriad of effects (Sim, 1991). A large number of phosphoprotein phosphatases have been described in mammalian tissues (Cohen, 1989). These enzymes can be divided into two broad types:

PAGE 20

14 phosphoserine/phosphothreonine-specific protein phosphatases or PP and phosphotyrosine-specific protein phosphatases or PTP (Gong et ai, 1993). The activity of each type of phosphoserine/phosphothreonine-specific protein phosphatase in mammalian tissues can be determined on the basis of differences in substrate specificities, dependence on divalent cations, sensitivities to specific inhibitors, and their catalytic subunit (Ingebritsen et al., 1983; Cohen, 1989). Based upon their four types of catalytic subunits, the phosphoserine/phosphothreonine-specific protein phosphatases are comprised of four main classes of enzymes: PP1, PP2A, PP2B, and PP2C (Ingebritsen et ai, 1983; Cohen, 1989). Protein phosphatase 1 (PP1) preferentially dephosphorylates the beta-subunit of phosphorylase kinase and is inhibited by two heat-stable inhibitor proteins, inhibitor1 (which inhibits PP1 after phosphorylation by cyclicAMP-dependent protein kinase) and inhibitor-2 (also termed "modulator," which inhibits PP1 by impeding the substrate binding and by inducing a conformational change of the catalytic subunit) (see review: Bollen and Stalmans, 1992). As noted in several reviews (see: Cohen, 1989; Sim, 1991; Bollen and Stalmans, 1992; Shenolikar, 1994), the substrate specificity of PP1 may be controlled by a number of different regulatory subunits both in different tissues and within the same tissue. These regulatory subunits direct PP1 activity toward specific subcellular localizations, and, therefore, toward specific substrates, through modulation of the phosphorylation state of a number of different regulatory subunits. Subcellular fractionation studies have demonstrated PP1 activity in cytosolic, synaptosolic, synaptic plasma membrane and synaptic junction fractions (Shields et ai, 1985; Dokas et ai, 1990). These findings are

PAGE 21

15 consistent with a ubiquitous distribution of this phosphatase in brain. Alternatively, type 2 phosphatases preferentially dephosphorylate the alpha-subunit of phosphorylase kinase and are insensitive to inhibitor1 and inhibitor-2. Type 2 protein phosphatases are subdivided into three distinct classes based on their cationic requirements: PP2A, PP2B, and PP2C. PP2A is active in the absence of divalent cations. The amount of PP2A activity in brain extracts is the highest of any tissues investigated (Ingebritsen et ai, 1983), and there is approximately three times as much PP2 A as PP 1 . Native PP2 A enzymes are heterotrimers of two regulatory subunits (A and B) and a catalytic subunit (Cohen, 1989; Shenolikar and Nairn, 1991 ). The mechanisms responsible for regulating PP2A activity in normal cells are still poorly understood. Notably, PP2B (i.e. calcium/calmodulindependent protein phosphatase or calcineurin) and PP2C are completely dependent on calcium and magnesium, respectively. PP2B is a heterodimer composed of equal amounts of A and B subunits. The A subunit contains the catalytic and calmodulin binding domains, whereas the B subunit binds Ca 2+ and is highly homologous to calmodulin (Klee, 1988). Calmodulin appears to activate PP2B by neutralizing the inhibitory effect of a 4kD domain on the A subunit, distinct from the calmodulin-binding domain. This inhibitory domain is extremely susceptible to proteolysis, and its removal yields a Ca 2+ -dependent enzyme that cannot be further stimulated by calmodulin (see review: Bollen and Stalmans, 1992). PP2C, meanwhile, is a monomelic protein which is structurally different from all other protein phosphoserine/phosphothreonine-specific phosphatases (Shenolikar and Nairn, 1991). Beyond its dependence on Mg 2+ , nothing is known about the regulation of this phosphatase or its specific substrates, and functional roles have not been identified.

PAGE 22

16 These four classes of enzymes also may be distinguished by their sensitivity to the marine toxin, okadaic acid (Cohen, 1989; Haystead et al, 1989). Okadaic acid, a polyether fatty acid, was shown to be a potent and specific inhibitor of PP1 and PP2A by Takai and coworkers (1987), while not activating PKC (Suganuma et al, 1988). Okadaic acid inhibits PP2A (IC 50 = 0. 1-1 OnM) at concentrations ten to one hundred times less than those required to inhibit PP1 (Suganuma et al, 1988). Due to its membrane permeability, it can be used in intact cells to identify physiological substrates of PP1 and PP2A, as well as to reveal novel intracellular processes that are controlled by phosphorylation (Hardie et al, 1991). However, in vitro sensitivity of PP2A to okadaic acid is decreased by increasing protein concentrations (see review: Shenolikar, 1994). Moreover, radiolabeled okadaic acid predominantly accumulates in membranes, where PP1 is more abundant than PP2A. Thus, it may be difficult to utilize solely the okadaic acid sensitivity of a particular process in the intact vesicle to identify the phosphatase involved. Another phosphatase inhibitor, isolated from a marine sponge (Kato et al., 1986), is calyculin A. This inhibitor, though structurally unrelated to okadaic acid, has been shown to bind to okadaic acid receptors in particulate and cytosolic fractions of mouse skin (Fujiki et al., 1989, 1991; Suganuma et al, 1989, 1990). It is a potent inhibitor of PP1 and PP2A. Distinct from okadaic acid, calyculin A was found equally effective against PP1 and PP2A with IC 50 values of 1.4nM and 2.6nM for PP1 and PP2A respectively (Ishihara et al., 1989; Suganuma et al, 1990, 1992). Studies by Bu and colleagues (1993) and others (Turner et al., 1993) have indicated that okadaic acid and other cell-permeable protein phosphatase inhibitors, such

PAGE 23

17 as calyculin A, lead to a preferential increase in the phosphothreonine content of metabolically labeled proteins. Because of this, it has been suggested that phosphothreonines, rather than phosphoserines, are preferentially dephosphorylated in cells (Shenolikar, 1994). Therefore, phosphatase inhibitors may be most useful for visualizing threonine phosphorylation. Aging Human Studies It is possible that altered protein phosphorylation is involved in the neuronal loss characteristic of Senile Dementia of the Alzheimer's Type (SDAT) (Armbrecht et al, 1993). The maintenance of neuronal structure and function is thought to require the continuous action of neurotrophic factors. One common pathway for the action of trophic factors is the activation and translocation of PKC. In this regard, it was found that PKC activity in human frontal cortex was reduced by about 50% in SDAT patients (Cole et al, 1988). Likewise, the phosphorylation of P86 protein, the major PKC substrate in the frontal cortex, was also reduced by 50%. Using antibodies against individual PKC isoforms, alterations in PKC isozymes have been shown in patients with Alzheimer's Disease (Masliah et al, 1990). Additionally, Van Huynh and his group in 1989 found that decreased PKC and decreased protein phosphorylation were also seen in fibroblasts from SDAT patients. These findings, therefore, suggest possible systemic defects in protein phosphorylation in SDAT. The effect of normal aging on phosphatase activity has not been investigated in any detail. Studies in SDAT brains have shown aberrant protein phosphorylation, suggesting

PAGE 24

18 an alteration in protein kinases and/or phosphoprotein phosphatases. The microtubuleassociated protein tau is known to be hyperphosphorylated in SDAT brains, and this abnormal hyperphosphorylation is associated with an inability of tau to promote the assembly of microtubules in the affected neurons. Previous studies by Khatoon and coworkers (1992) and others (Swaab et al., 1992; Iqbal et al., 1993) have demonstrated that abnormally phosphorylated tau could be dephosphorylated after treatment with alkaline phosphatase, thereby suggesting that the abnormal phosphorylation of tau might in part be the result of a deficiency of the phosphoprotein phosphatase system in patients with SDAT. Furthermore, Gong et al. (1993) reported that PP1 and PP2A activity in frontal gray and white matter were significantly decreased in SDAT brains when compared to controls. PP2B, PP2C, and phosphatase activities remained unchanged. As a change in the state of phosphorylation of a substrate protein and therefore its functional activity can be achieved physiologically through increases or decreases in the activity of either a protein kinase or a phosphoprotein phosphatase, it is clear that analyses of phosphoprotein phosphatases is important in clarifying the mechanism of aberrant protein phosphorylation/dephosphorylation in SDAT brains. Rodent Studies The activity of PKC has been investigated in the senescent rodent brain, and agerelated modifications in levels of protein phosphorylation have been reported (Barnes et al., 1988; Govometal., 1988; Magnoni etal., 1991; Parfitt etal., 1991; Pisano etal., 1991). Specifically, age-related changes in protein phosphorylation mediated by PKC have been described by Barnes and colleagues (1988). This group reported that in the

PAGE 25

19 hippocampus of senescent rats, there was a 46% decrease in phosphorylated Fl (i.e. GAP43 or B-50), a 47kD protein postulated to play a role in long term potentiation, signal transduction, and neurotransmitter release. However, when the content and functional activity of PKC were measured in aged rats by Battaini and colleagues (1990), results indicated that both these parameters were enhanced in the hippocampus yet reduced in the cortex. The age findings were similar to those later obtained by Pisano and colleagues in 1991, which indicated that the phosphorylation of certain cytoskeletal/microtubular proteins by PKC declines with age. Using brain slices, they demonstrated that the phosphorylation of tubulin, microtubule-associated proteins (MAPs), and tau protein was much less in 24 month old rats as compared to 6 month old rats in response to phorbol esters. When mRNA levels of three PKC isoforms (i.e., a, P, y) were investigated in aging rats, no change in expression was found (Battaini et al, 1994; Narang and Crews, 1995). However, Northern blot studies have indicated lower PKC H and PKC p mRNA levels in the cortex of 28 month old rats as compared to animals aged 3 months (Battaini et al, 1993). Several groups have demonstrated that activation of calcium-dependent PKCs is defective in the brains from aged rats (Friedman and Wang, 1989; Meyer et al, 1994; Battaini et al, 1995; Undie et al, 1995). This deficit in translocation has been related to biochemical changes in membrane phospholipid composition (Undie et al, 1995) and with functional impairment in neurotransmission (Friedman and Wang, 1989; Meyer et al, 1994). A study by Pascale and colleagues (1996) suggest that alterations in receptors for activated C kinase which bind PKC contribute to the functional impairment in PKC activation observed in aged rat brain.

PAGE 26

20 Several studies have examined EAA uptake in tissue preparations from a variety of areas of the aging rat brain. In the neocortex, uptake of both L-[ 14 C]glutamate into a preparation of synaptosomes and D-[ 3 H]aspartate in crude homogenates was found to decline with age, although these changes appear to reflect differences between immature (2-4 months of age) and mature (10-30 months of age) animals (Strong et al, 1984; Meldrum et al., 1992). By contrast no difference was observed in L-[ 3 H]glutamate uptake into slices of frontal neocortex in a comparison of animals aged 6 and 24 months (Dawson et al., 1989). In the neostriatum, uptake both of L-[ 3 H]glutamate (in crude synaptosomal fractions) and D-[ 3 H]aspartate (in crude homogenates) was lower in mature rats (aged 10 months or more) than in immature animals, aged 6 months or less (Price et al., 1981; Wheeler and Ondo, 1986; Najlerahim etal., 1990). Additionally, Palmer and colleagues, using Fisher 344/Norwegian Brown rats aged 3, 12, 24, and 37 months, assayed D[ 3 H]aspartate uptake in crude cortical, hippocampal, and neostriatal synaptosomes and found no significant changes with increasing age (Palmer et al., 1994). The conflicting results in these studies may be attributable to several factors. First, because of different vulnerabilities of inbred strains of rats to the effects of aging, previous studies may have used animals that had not fully reached senescence (Coleman and Flood, 1987). Second, erroneous conclusions may be made on the basis of studies that examined animals at two ages only (Coleman and Flood, 1987). Finally, the young control animals (<12 months of age) used in many studies may not have been entirely appropriate, since there is evidence that sexual maturity (puberty) may not mark the end of what may be

PAGE 27

considered maturation. Evidence indicates that this process may continue until 12 months of age in some brain regions of some rodent strains (Coleman and Flood, 1987).

PAGE 28

CHAPTER 2 ISOLATION OF GLIAL AND NEURONAL FRACTIONS Introduction Synaptosomes have proven to be an excellent model system for studies on the molecular mechanisms underlying presynaptic phenomena (Bradford, 1975). Since the original isolation procedures were developed (De Robertis et al, 1961; Gray and Whittaker, 1962), a number of methodological improvements have been described (Cotman, 1974; Morgan, 1976). Many researchers prefer to use crude mitochondrial or P 2 fractions (Whittaker, 1972) for neurochemical investigations due to the simplicity and speed (30-40min) for isolation of this fraction as well as the avoidance of hypertonic conditions. However, standard P 2 fractions have major disadvantages associated with contaminants (myelin, extrasynaptosomal mitochondria and glial cells) which can have direct and indirect effects on synaptic events (Dodd et al, 1981). Glial contamination of synaptosomal preparations is particularly problematic as evidenced by both biochemical (carbonic anhydrase activity) and morphological measures (Delaunoy et al, 1979; Henn et al, 1974). As an alternative to traditional methods, isolation of synaptosomes on Percoll gradients is rapid, isotonic and leads to a relatively homogeneous isolate that is specifically depleted of damaged synaptosomes, synaptic plasma membranes, extraneous membranous material, myelin and extrasynaptosomal mitochondria (Dunkley et al, 1986). The three22

PAGE 29

23 step Percoll density gradient centrifiigation technique is superior to other techniques insofar as glial vesicles are separated more efficiently from synaptosomes (Booth and Clark, 1978; Fykse and Fonnum, 1988). Complete separation of neuronaland glialderived vesicles is of critical importance since both neurons and glial cells work together in order to maintain adequate and appropriate functioning in the central nervous system (CNS). In contrast to synaptosomes, which have been used for studies of neuronal processes, few biochemical preparations have been available for the study of glial functions. Instead, a heavy reliance has been placed upon cultured cells in order to evaluate glial functions; however, it is known that the culturing process itself may alter intrinsic properties of various cell types. Within the past several years limited success have been realized with regard to the isolation of high-quality glial-derived fractions from whole-brain homogenates. An initial report by Nakamura and coworkers (Nakamura et al., 1993) involved a three-step Percoll density gradient centrifiigation technique which was used to isolate simultaneously glial-derived vesicles (termed glial plasmalemmal vesicles or GPV) and neuronal-derived vesicles from a rat brain homogenate. While the method described in this report is based loosely on this previous work, significant technical modifications have produced marked improvements in the simultaneous isolation of functionally viable glial and neuronal elements. In this report, we have used morphological, biochemical and molecular markers to validate the efficiency of this separation technique, and we describe the properties of L[ 3 H]-glutamate uptake by the two vesicle fractions.

PAGE 30

24 Experimental Procedures Materials L-[2,3,43 H]-Glutamic acid (specific activity = 60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). D-[2,33 H]-Aspartic acid (specific activity =13.5 Ci/mmol) and [methyl3 H]-choline chloride (specific activity = 81 Ci/mmol) were purchased from DuPont/NEN (Boston, MA, USA). Glial fibrillary acidic protein (GFAP) was acquired from Biogenesis (Sandown, NH, USA), while primary antibody against GFAP was purchased from DAKO (Carpinteria, CA, USA). Neuron specific enolase (NSE) and primary antibody were obtained from Polysciences (Warrington, PA, USA). Secondary antibodies, as well as the alkaline phosphatase color reagent development kit, were bought from Biorad (Hercules, CA, USA). D-Aspartic acid and DL-f/veo-P-hydroxyaspartic acid were purchased from Sigma Chemical Co. (St. Louis, MO, USA), while L-/ra/w-pyrrolidine-2,4-dicarboxylic acid was obtained from Tocris Cookson (St. Louis, MO, USA). L-a-Aminoadipic acid was bought from Calbiochem-Novabiochem International (La Jolla, CA, USA). All other chemicals were purchased from either Fisher Scientific or Sigma Chemical Co. and were of the highest quality available. Preparation of Tissue Young adult (3 4 months of age) Sprague-Dawley male rats (Zivic Miller) weighing 250-275g were used throughout this study. Animals were housed in pairs and maintained on a 12-hr light/dark cycle with food and water available ad libitum and were transported to the laboratory approximately 15 hrs prior to use. Two rats were

PAGE 31

25 decapitated quickly with a small animal guillotine, and the brains were removed rapidly and placed upon an ice-cold glass surface. Cerebellar tissue was removed and discarded, while all remaining forebrain tissue (approximately 2.4 g) was placed in 30 ml of an icecold solution containing 0.32 M sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA). The tissue was homogenized gently with a Potter-El vehj em tissue grinder (approximately 30 rpm) and centrifuged at 1000 x g for 10 min (4°C) using a fixed-angle rotor (F28/50-DuPont). All subsequent centrifugation steps were conducted at 4°C. The resultant pellet was discarded, and the supernatant was split into four equal portions, which were diluted to 30 ml with an ice-cold solution containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM dithiothreitol and 20 mM HEPES (pH 7.4 at 4°C). Hereafter, this solution is referred to as SEDH. Diluted aliquots of supernatant were centrifuged at 5000 x g (15 min), and resultant supernatants were saved separately on ice. Each of the four tissue pellets were resuspended in 15 ml of ice-cold SEDH solution and centrifuged at 1000 x g for 10 min. The resultant pellets were discarded, while supernatants were combined with supernatants saved from a previous step. The four tubes, each containing approximately 45 ml of tissue homogenate, were centrifuged at 33,500 x g (20 min), and supernatants were discarded. Tissue pellets (four) were resuspended in 15 ml of ice-cold SEDH solution and gently transferred onto a three-step discontinuous Percoll gradient (10 ml each of 1.38%, 2.3%, and 4.6% Percoll in SEDH solution) with a Minipuls 2 (Gilson) peristaltic pump (flow rate = 0.88 ml/min). Tubes were centrifuged at 33,500 x g (10 min) with 15-min periods of linear acceleration to and deceleration from the top speed. The turbid layer between 0% and 1 .38% Percoll was collected from all four tubes and

PAGE 32

26 combined into two aliquots, which were diluted to a final volume of 15 ml each with icecold SEDH solution. Aliquots were centrifuged at 1000 x g (20 min), and resultant supernatants were layered onto fresh three-step Percoll gradients as described above. Tubes were centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration (see above), and the turbid layer between 0% and 1.38% Percoll was collected from both tubes and combined into a single aliquot. The tissue aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution, centrifuged at 33,500 x g (20 min), and the resultant pellet was used as the glial plasmalemmal vesicle (GPV) fraction. For the recovery of the synaptosomal (SYN) fraction, the turbid layer between 2 .38% and 4.6% Percoll was collected from the initial discontinuous gradient. The four aliquots were diluted to a final volume of 15 ml (each) with ice-cold SEDH solution and centrifuged at 1000 x g (20 min). Pellets were discarded, and supernatants were layered onto a threestep Percoll gradient (see above) and centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration periods. The turbid layer between 2.38% and 4.6% Percoll was collected from each of the four tubes and combined into one aliquot. This aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution and centrifuged at 17,500 x g (20 min). The resultant pellet was designated as the SYN fraction. The total preparation time from initial animal decapitation was approximately six hours. Lipid Quantification Tissue phospholipids were separated from protein and quantified by a modification of the method of RodriguezVico and coworkers (RodriguezVico et al., 1991). Five ml of hexane/isopropanol (3:2, v/v) were added to a 13-ml centrifuge tube containing 0.05 ml

PAGE 33

27 aliquots of either GPV or SYN. Samples were mixed and centrifuged at 10,000 x g for 15 min. The liquid phase was decanted and concentrated to dryness under a stream of dry N 2 . Following the modified method of Bartlett (Bartlett, 1959), samples were reconstituted in 1 .0 ml H 2 0 and acidified with 0.3 ml of 10 N H 2 S0 4 . The mixture was heated to 150°-160°C for 3 hours, and following addition of two drops of 30% H 2 0 2 , the solution was heated for 1.5 hours in order to complete the combustion process and break down any residual peroxide. After combustion, H 2 0 (0.65 ml), 5% ammonium heptamolybdate tetrahydrate (0.2 ml), and Fiske-SubbaRow reagent (0.05 ml) were added, and the solution was heated for 7 min at 100°C. Following a 10 min incubation at room temperature, absorbance was measured at 660nm (LKB Ultraspeed spectrophotometer, Pharmacia) and molar concentrations for unknowns were estimated by comparison with inorganic phosphate standards (5 //M 200 /jM). Electron Microscopy Electron microscopy was carried out in the Electron Microscopy Core Facility at the University of Florida Interdisciplinary Center for Biotechnology Research. GPV and SYN fractions were fixed in 2% buffered glutaraldehyde at 4 °C for 1 hr. After three 10 min washes with phosphate buffered saline (pH 7.2), samples were postfixed in 1% buffered osmium tetroxide (pH 7.2) for 1 hr. Following three washes with deionized water, samples were dehydrated in 50% ethanol for 15 min. Staining occurred overnight at 4°C with 2% uranyl acetate in 75% ethanol. GPV and SYN were dehydrated in a graded series of alcohols (15 min each), using 100% acetone as a transitional solvent. Infiltration occurred in a graded series of embedding resin/acetone mixtures (1 hr each),

PAGE 34

28 and blocks were polymerized at 60 °C for two days. Thin sections were taken on a Leica Ultracut E Ultramicrotome with a diamond knife and post-stained with 2% aqueous uranyl acetate followed by Reynold's lead citrate. Micrographs were taken on a Hitachi H-7000 TEM at 75kv. [ 3 H] Excitatory Amino Acid Uptake The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 30°C). GPV and SYN were used immediately for uptake. Sodium-dependent [ 3 H]L-glutamate uptake by GPV and SYN was measured by a filtration method modified from Divac and coworkers (Divac et al, 1977). For the measurement of uptake, aliquots (50 //l) of tissue fractions (approximately 1 mg protein / ml) were added to cold glass culture tubes that contained 400 ful of a buffered solution containing (in mM) NaCl (140), KC1 (5), CaCl 2 (1.0), MgCl 2 (1.0), NaH 2 P0 4 (1.2), D-glucose (10), and HEPES (20) at pH 7.4 (30°C). Sodium-independent uptake (blanks) was measured in parallel using a buffer in which NaCl was replaced by an isosmolar concentration of choline chloride. Assay tubes in triplicate containing tissue homogenate and buffer were warmed to 30°C (5 min) in a shaking water bath. The uptake reaction was initiated by the addition of 50 (A of L-[2,3,43 H]-glutamic acid at a final concentration of 5 ^M. Tubes were mixed rapidly and returned to the water bath for 90 seconds. Uptake was terminated by rapid vacuum filtration using a Brandel cell harvester and Whatman GF/B filter sheets that had been presoaked overnight at 4°C in 25 mM L-glutamate. Test tubes and filters were rinsed rapidly three times with 2 ml aliquots of ice-cold normal or sodium-deficient buffer. Tissue trapped on filters was digested with 2 ml of 0.2 M NaOH (overnight), acidified

PAGE 35

29 with 1 ml of 0.5 M HC1, and assayed for tritium content in 10 ml of EcoLume scintillation fluid (ICN Biochemicals). Radioactivity was quantified in an LKB 1214 liquid scintillation counter with a counting efficiency of approximately 45% as determined by a radium-226 standard. Sodium-dependent uptake was determined as the difference between uptake in normal versus sodium-deficient buffers. For time-course and dose-response studies, tissue fractions were incubated between 5 sec and 10 min in the presence of 5 fuM [ 3 H]Lglutamate or were incubated with concentrations of [ 3 H]L-glutamate ranging between 0.0177 {OA 100 juM for 90 sec, respectively. Uptake of [ 3 H]D-aspartate (5 juM, final cone) was measured according to the same procedure used to determine [ 3 H]L-glutamate uptake. [ 3 H]Choline Uptake High-affinity [ 3 H]choline uptake by GPV and SYN fractions was measured by a method modified from Divac and coworkers (Divac et al, 1977). Aliquots (50 (A) of tissue fractions (approximately lmg protein / ml) were added to cold glass culture tubes containing 400 \A of a solution containing (in mM) NaCl (120), KC1 (4.7), MgCl 2 (1 .2), D-glucose (20), CaCl 2 (2.5), and HEPES (50) at pH 7.4 (30°C). Nonspecific [ 3 H]choline uptake was measured in the presence of 1 (jM hemicholinium-3 . Assay tubes in triplicate were warmed to 30°C (5 min) in a shaking water bath, and the reaction was initiated by the addition of 50 \A [methyl3 H]choline chloride at a final concentration of 1 //M. Uptake was allowed to proceed for 3 min (30°C) and was stopped by the addition of 1 ml of icecold buffer containing 10 /uM hemicholinium-3. Tissue fractions were trapped by vacuum filtration onto Whatman GF/B filters that had been presoaked overnight at 4°C in 100 mM

PAGE 36

30 choline chloride. Test tubes and filters were rinsed rapidly with four 2 ml aliquots of cold buffer. Tissue was digested and tritium content was determined as described above. Specific [ 3 H]choline uptake was determined as the difference between tissue tritium content in the absence and presence of hemicholinium-3. Western Blot Analyses Western blot analyses were conducted in the Protein Chemistry Core Facility at the University of Florida Interdisciplinary Center for Biotechnology Research. For immunoblotting of antibodies against GFAP and NSE, aliquots (5-15 //g protein / 20 fj.1) of GPV and SYN samples were subjected to SDS-polyacrylamide gel electrophoresis (10% Tris-Tricine polyacrylamide gels; 50V for 15 min, followed by 120V for 40 min). The electrophoresed proteins were transferred onto polyvinylidene difluoride membrane (Immobilon P, Millipore) (90V for 1 hr). Blots were blocked (1 hr.) with 5% milk in TrisTricine blocking solution (TTBS) (10 mM Tris-HCl, 150 mM NaCl, 0.05% Tween-20), then incubated overnight at 4°C with primary antibodies against either GFAP or NSE (1 : 1000). Blots were washed with TTBS and then incubated with alkaline phosphataseconjugated goat anti-rabbit IgG (1 : 1000) for 2 hrs. The secondary antibody was washed out with TTBS, and proteins were visualized via alkaline phosphatase color reagent development kit. Giutamine Synthetase Activity The activity of giutamine synthetase in the GPV and SYN tissue samples was determined according to the method of Galanopoulous and colleagues (Galanopoulos et al., 1988) with slight modification. Aliquots (100 fxl) of either GPV or SYN were placed

PAGE 37

31 in 13 -ml centrifuge tubes containing 10 /A of 3% (w/v) sodium deoxycholate (DierksVentling et al, 1975) and sonicated. Thereafter, 450 /A of a solution containing (in mM) L-sodium glutamate (110), disodium ATP (11), imidazole (110), MgS0 4 (44.5), Lcysteine.HCl (44.5), hydroxylamine.HCl (44.5), tricyclohexylammonium salt of phosphoenolpyruvic acid (9), and 5IU of pyruvate kinase (pH 7.4 at 37°C) were added, and the mixture was incubated for 15 min at 37°C in a shaking water bath. The latter two components constituted the ATP -regenerating system (Lund, 1970). The reaction was terminated by addition of 150 /A of a solution containing equal volumes of 10% (w/v) FeCl 3 .6H 2 0 in 2 N HC1, 24% (w/v) trichloroacetic acid, and 50% concentrated HC1. After centrifugation at 50,000 x g (20 min), the optical density of the supernatant was measured within 10 min on a spectrophotometer (LKB Ultraspeed, Pharmacia) at 540nm and quantified by comparison with standard solutions of L-glutamic acid gammamonohydroxamate. Carbonic Anhydrase Activity Carbonic anhydrase activity in GPV and SYN samples was determined according to a technique modified from that of Brion and coworkers (Brion et al, 1988). All reactions were carried out at 0°C in gas-tight glass microvials (300 (A total volume) that were sealed with teflon septa. Tissue homogenate (30 /A) was mixed with 20 [A of H 2 0 in microvials prior to being sealed. The vessels were purged with C0 2 (0.5 ml / min) for 7.5 min prior to initiating the reaction. C0 2 was infused and exhausted through two hypodermic needles which pierced the septum. Following the preincubation step, enzyme activity was determined via a colorimetric reaction. Fifty microliters of a solution

PAGE 38

32 containing (in mM) imidazole (20), TRIS (5), and />-nitrophenol (0.4) were added to each vial with the aid of a syringe and needle. The time necessary for yellow color to dissipate completely was determined and compared to standard carbonic anhydrase solutions. Protein Analyses The amount of protein in GPV and SYN samples was determined by the Lowry protein assay (Lowry et ai, 1951) using solutions of bovine serum albumin as standards. Data Analyses All values reported were the mean ± SEM from four to twelve fractions isolated on separate experimental days, except the glutamine synthetase results, which were reported as mean ± SD. Statistical significance was determined via Student's /-test and accepted at a level of p<0.05. Results Morphological Examination of GPV and SYN Fractions Rat forebrain homogenates were fractionated by a multi-step procedure which entailed both differential and discontinuous density-gradient centrifugation steps. While the basis for this fractionation procedure was based loosely on the method of Nakamura and coworkers (Nakamura et ai, 1993), the final method used for this investigation is substantially more complex and time-consuming than the previously published procedure. In view of the amount of time required to isolate final tissue fractions using this modified procedure (approx. 6 hr), it was necessary to verify the structural integrity of constituents in the final tissue fractions with morphological approaches. Using standard electron microscopy, both the GPV and SYN fractions were found to contain a large number of

PAGE 39

33 intact membrane-encapsulated vesicles (Fig. 2-1). Within the GPV fraction, membraneencapsulated vesicles could be broadly classified as small (0. 1 0.3 //m) spherical structures or large (0.4 0.9 jum) irregularly-shaped structures (panel A), though this division was somewhat arbitrary. The vast majority of vesicles in the GPV fraction contained a clear core with little or no electron-dense material evident. In addition to vesicles, the GPV fraction contained non-descript membrane fragments and electron-dense vesicles, though the prevalence of these components was quite low. The morphological characteristics of elements within the GPV fraction are consistent with observations published by Nakamura and coworkers (1993). In contrast to elements in the GPV fraction, a high percentage of membrane-encapsulated structures in SYN fractions contained numerous identifiable synaptic vesicles with identifiable synaptic vesicles, postsynaptic densities and mitochondria (Fig. 2-1, panel B). As noted previously, these morphological characteristics are consistent with a neuronal origin for elements in the SYN fraction. In addition to morphological characterizations, GPV and SYN fractions were analyzed for total content of protein and membrane phospholipids. Normalization of protein content by total membrane phospholipid indicated that the amount of membrane phospholipid (/umol lipid per mg protein) was not statistically different between the two fractions (155 ± 24, n = 10 vs. 106 ± 48, n = 12 for SYN and GPV, respectively) at a level of p< 0.05.

PAGE 40

34 Glial Markers In order to obtain supportive evidence for the efficiency with which glial and neuronal elements can be separated by this tissue fractionation protocol, several approaches were used to measure the level of marker proteins and enzymes that are selectively expressed by glia or neurons. GPV and SYN fractions separated by the threestep discontinuous density gradient method were analyzed for glial-specific enzyme activities and GFAP, a marker protein that is specific for astroglial cells. As shown in Fig. 2-2, specific activities of glutamine synthetase (GS) and carbonic anhydrase (CA) were significantly higher (p < 0.00001) in GPV fractions relative to SYN fractions. In the case of GS, GPV fractions contained more than six-fold higher enzyme activity as compared to SYN fractions isolated from the same rat forebrain homogenates. Measurements of CA activity indicated an even greater difference between GPV and SYN fractions insofar as SYN exhibited no measurable activity whereas GPV contained an average of 2360 units of activity per mg protein. Based upon the estimated lower limit for detection of CA activity by this method (50 units / assay tube), these results indicate at least 45-fold greater CA activity in GPV relative to SYN fractions. Finally, as shown in Fig. 2-2 (panel C), Western blot analyses for the astrocyte-specific protein GFAP provide additional support for the effective separation of glial elements between SYN and GPV fractions. When incubated with GFAP-specific polyclonal antibodies, GPV fractions (5-15 //g protein / lane) exhibited strongly positive immunostaining to a band near 47 kD. By comparison, positive immunostaining in SYN fractions was evident only in the presence of 15 //g total protein. Despite our inability to quantify the amount of immunoreactive material in each

PAGE 41

35 fraction, densitometric measurements indicated the presence of substantially more immunopositive material in the 5 //g GPV sample (4104 arbitrary units) relative to the 15 //g SYN sample (1873 arbitrary units) following background subtraction. The amounts of antigen and density of reaction product lay on the linear portion of a standard scale. Neuronal Markers While the experimental results outlined above helped confirm the presence of glial markers in GPV fractions and the diminished presence of such markers in SYN fractions, additional studies were undertaken to confirm the presence of neuronal markers in SYN fractions and to estimate contamination of such markers in GPV fractions. One marker used to address this issue was sodium-dependent high-affinity choline uptake, a process that is highly localized to cholinergic neurons. In the presence of 1 //M substrate, hemicholinium-3 -sensitive uptake of [ 3 H]choline was nearly 40,000-fold higher in SYN vs. GPV fractions (9.68 vs. 0.00025 pmol / mg protein / 3 min). This result indicates that GPV are nearly devoid of neuronal contamination. Verification of this observation was obtained in Western blots by using an antibody against NSE, a marker protein localized preferentially to neurons. As shown in Fig. 2-3 (panel B), GPV had no positive immunoreactivity (above background) with up to 15 //g of total GPV protein loaded per lane. By comparison, SYN fractions displayed strong positive immunoreactivity for NSE, as illustrated by the heavy banding at approximately 49 kD. Excitatory Amino Acid Uptake GPV and SYN fractions from rat forebrain were tested in order to ascertain their abilities to accumulate excitatory amino acid (EAA) substrates. As shown in Fig. 2-4,

PAGE 42

both fractions demonstrated a good capacity to transport substrates in a sodiumdependent manner. While the level of EAA transport was found to differ between GPV and SYN fractions, there was a strong correlation between L-[ 3 H]glutamate uptake and D-[ 3 H] aspartate uptake within individual preparations (Fig. 2-4, inset). This observation served to validate the use of L-[ 3 H]glutamate for studies of EAA transport in GPV and SYN fractions. As shown in Fig. 2-5, L-[ 3 H]glutamate uptake at 30°C is a rapid, highaffinity process which exhibits pseudo first-order reaction kinetics (panel B). Analyses of results from substrate saturation experiments (panel A) revealed that EAA transporters in GPV and SYN fractions exhibit high affinities for L-[ 3 H]glutamate (K,,, values of 2.7 (jM and 5. 1 fxM, respectively), but differ significantly in their maximal rate of transport with the capacity in SYN exceeding that of GPV (773 vs. 279 pmol / mg protein / 90 sec). In addition pharmacological comparisons between GPV and SYN were conducted using synthetic L-glutamate analogues which have been reported to inhibit L-glutamate transport. As shown in Fig. 2-6, L-[ 3 H]glutamate uptake was completely inhibited by several compounds in both tissue fractions. In studies with L-frarcs-pyrrolidine-2,4dicarboxylic acid (PDC) (panel A) and D-aspartic acid (D-ASP) (panel D), both tissue fractions revealed similar sensitivities to these transporter inhibitors. Average fitted IC 50 values were 4.7, 4.1, 4.2 and 3.7 //M for PDC and D-ASP in GPV fractions and SYN fractions, respectively. In contrast to these results, both DL-//»ra>-P-hydroxyaspartic acid (THA) (panel B) and L-a-aminoadipate (AAD) (panel C) revealed modest selectivity for L-glutamate transporters in GPV fractions. Average IC 50 values were 1.4 and 3.2 //M (THA) and 200 and 340 //M (AAD) in GPV and SYN fractions respectively. Despite

PAGE 43

o £ o .a . S c/l — CO td * C T3 n. Q. a +~J 1—1 E '55 Ill ^ .2 d CO 3 II 9 S" !3 .2 §^ a § » I J s CO 1_ * £ ^ is| ° 55 O «t> all h 3 &> O C t/5 'Spa a Bl a ^> 5 S£ S c .£ 8 w a 52 «J t; C 7. P O .SJ a> •« c« n ti IS u S > "2 * £ I 00

PAGE 45

% 3 > gfl U > 2 -o o

PAGE 47

Figure 2-2. Analysis of GPV and SYN tissue fractions for glial-specific enzyme activities and GFAP. Panel A) Glutamine synthetase (GS) activity was determined spectrophotometrically by monitoring GS-catalyzed formation of L-glutamic acid gamma-monohydroxamate (GHA). The absorbance of the samples was measured on a spectrophotometer at 540nm. Values are mean ± SEM, (n=5) for each tissue fraction. One unit of activity is defined as 1 pmol GHA formed / 15min / mg protein. Panel B) Carbonic anhydrase (CA) activity in GPV (n=7) and SYN (n=l 1) fractions expressed as units of enzyme activity per mg protein. Panel C) Western blot analysis for GFAP immunoreactivity in GPV (lanes 3-5) and SYN (lanes 6-8) fractions. Aliquots containing either 5/ug (lanes 3 and 6), 10/yg (lanes 4 and 7) or 15yL/g (lanes 5 and 8) of total protein were subjected to SDS-polyacrylamide gel electrophoresis prior to reaction with primary and secondary antibodies. Molecular weight markers were run in lane 1 and antigen in lane 2. Numbers to the left of lane 1 indicate molecular weight in kD.

PAGE 48

GPV SYN 2000 1000 GPV SYN C) 97.446• 301 2 3 4 5 6 7 8

PAGE 49

43 C .5 2 u | " — 12 8 A) GPV SYN B) 6646301 2 3 5 6 7 8 Figure 2-3. Measurement of neuron-specific markers in GPV and SYN. Panel A) High-affinity [ 3 H]choline uptake by GPV and SYN fractions was determined in the presence of l^M [ 3 H]choline. Values are mean ± SEM, (n=5) for each tissue fraction. Panel B) Western blot analysis for NSE immunoreactivity in SYN (lanes 3-5) and GPV (lanes 6-8) fractions. Molecular weight markers were run in lane 1 and antigen in lane 2. Numbers to the left of lane 1 indicate molecular weight in kD.

PAGE 50

44 II 3 o o. bo E 1 — 1 5, 1200 1000 800 600 o 400 200 T J [ 3 H]D-Aspartate | 1 I 3 H]L-Glutamate Jj 1200 — 3 5" 900 — 3 600 o ^ 300 — 0 SYN OPV SYN ^=0.998 I I I I 0 300 600 900 1200 [ 3 H]ASP Uptake GPV Figure 2-4. Comparison of sodium-dependent [ 3 H]D-aspartate and [ 3 H]L-glutamate uptakes by GPV and SYN fractions. Tissues were incubated in the presence of 5//M substrate for 90 seconds. Values are mean ± SEM, (n=5) for each tissue fraction. Inset depicts plot of paired values from individual tissue fractions for all experiments. Degree of correlation (r) equals 0.998.

PAGE 51

45 0 20 40 60 80 100 120 [GLU],(10" 6 M) 0 150 300 450 600 TIME (sec) Figure 2-5. Properties of sodium-dependent L-[ 3 H]glutamate transport in GPV and SYN fractions isolated from rat forebrain. Panel A) Substrate-saturation curves for L-[ 3 H]glutamate transport in GPV and SYN. Aliquots (50/^1) of tissue were incubated for 90 seconds in the presence of L[ 3 H]glutamate (0.0177^M 100^M). Data points represent the mean ± SEM from 5 separate experiments. Panel B) Time-course for L-[ 3 H]glutamate transport in GPV and SYN in the presence of 5fjM substrate. Values are mean ± SEM, (n=4) for each tissue fraction.

PAGE 52

46 'O "9 -8 -7 -6 -5 -4 -8 -7 -6 -5 -4
PAGE 53

these differences, both fractions displayed identical rank orders of potencies (THA>DASP>PDOAAD) for these L-glutamate transport inhibitors. Discussion In contrast to synaptosomes, which have been used extensively to characterize neuronal processes in the CNS, few preparations derived from CNS tissues have been available for the study of glial functions. While cultured glial cell lines provide the benefits of a homogenous and stable cell population, certain limitations restrict the range of uses or applications for these preparations. In addition, since it is possible for cellular properties to be altered under culture conditions, it is prudent to determine to the greatest extent possible the degree to which cultured cells accurately reflect cell functions in living tissues. In view of these considerations, there is need for a technique which can be used to obtain nontransformed cells or functional cell derivatives from tissues of experimental animals. Such a technique provides an important adjunct to cell culture whereby the effects of aging or experimental manipulations on cellular/molecular events can be evaluated in CNS cells. The technique presented herein provides such a methodology, yielding simultaneously isolated fractions highly-enriched with either glia-derived or neuronderived vesicles. However, the point should be made that no technique is without drawbacks and, as such, important restrictions of our improved technique are discussed below. The extensive multi-step density gradient fractionation technique described in this report is based loosely upon a previous report by Nakamura and colleagues (1993). Since a far greater number of cellular markers have been used in the present study to estimate

PAGE 54

48 the extent of glial and neuronal separation, presenting an exhaustive comparison of the separation achieved with our method versus that of the Nakamura method is not possible (Nakamura et al, 1993). However, since two cellular markers (high-affinity [ 3 H]choline uptake and GFAP immunoreactivity) were used by both groups, these markers provide a means for direct comparison of the fractions isolated by the two procedures. Using hemicholinium-3 -sensitive [ 3 H]choline uptake as a neuronal marker, Nakamura and coworkers reported that GPV fractions exhibited uptake that was 10 percent of uptake in the SYN fractions (Nakamura et al., 1993). However, using the fractionation technique described in this report, hemicholinium-3 -sensitive [ 3 H]choline uptake in GPV was measured as less than 0.0025 percent of uptake in SYN fractions (Fig. 2-3 A). Therefore, based upon this neuronal marker, it is estimated that our fractionation procedure reduces neuronal contamination in GPV fractions by a factor of 4000 compared with the previous method (Nakamura et al, 1993). Another marker used in both studies was GFAP immunoreactivity as measured by Western blot analysis. GFAP is a cell-specific intermediate filament protein found only in astrocytes (Eng, 1985; Eng et al., 1971; Goldman et al, 1978). As shown in Fig. 2-2C, GFAP was undetectable in SYN fractions that contained less than 15//g total protein. However, Nakamura and colleagues (1993) detected GFAP in substantially smaller amounts (5/ig protein) of SYN fractions isolated with their procedure. Since neither study provided accurate quantitative determinations of GFAP levels, calculating the difference achieved by the two fractionation procedures is not possible. Nevertheless, the current method clearly offers a substantial improvement in the removal of glial elements from the SYN fraction as compared with the earlier method.

PAGE 55

49 Since a major goal of this study was to develop and validate a method for isolating functional glia-derived elements that were devoid of neuronal contamination, results from studies with neuron-specific markers (high-affinity choline uptake and NSE immunoreactivity) provide solid evidence for the highly-enriched glial nature of the GPV fractions. High-affinity [ 3 H]choline uptake (Fig. 2-3 A) is the rate-limiting step for acetylcholine synthesis in neurons (Kuhar and Murrin, 1978). Neuronal choline transporters are coupled to an electrochemical sodium ion gradient and are purported to be associated exclusively with cholinergic neurons (Haga and Noda, 1973; Kuhar et al, 1973; Yamamura and Snyder, 1972; Yamamura and Snyder, 1973). As shown in Fig. 2-3 (panel A), [ 3 H]choline uptake was virtually undetectable in GPV fractions, being nearly 40,000-fold lower than uptake in SYN fractions. However, the near absence of [ 3 H]choline uptake in GPV fractions could not be attributed to a loss of transmembrane sodium gradient or other non-specific changes, since GPV demonstrate a good capacity for sodium-dependent L-[ 3 H]glutamate uptake (Figs. 2-5 and 2-6). The efficient removal of neuronal elements from the GPV fraction represents a striking improvement over the results obtained by Nakamura and coworkers (1993), wherein GPV fractions maintained very appreciable levels of [ 3 H]choline accumulation. In addition to choline uptake, NSE immunostaining was used to assess neuronal contamination in GPV fractions. NSE, a unique form of the glycolytic enzyme enolase, preferentially though not exclusively localized in neurons and neuroendocrine tissues (Marangos et al, 1975; Kato et al, 1982), helped validate our observation with choline uptake. As shown in Fig. 2-3B, there was no positive immunostaining in GPV fractions, even when lanes were overloaded with

PAGE 56

50 high levels of total protein. Taken together, these data indicate that the method described in this report provides a highly-enriched, functionally viable population of glial cell elements which are suitable for studies of glial cell functions in vitro. It should be noted, however, that "glia-derived" may well include oligodendrocyte and microglial processes as well as astrocyte processes. Indeed, the CA activity present in GPV fractions is consistent with some contribution from oligodendrocytes, since this enzyme is expressed by oligodendrocytes (Cammer, 1984; Ghandour et ai, 1979) as well as astrocytes (Cammer and Tansey, 1988a, b). Similar approaches were used in order to establish the highly-enriched neuronal nature of synaptosomes (i.e., absence of glial cell markers) that were isolated in parallel with GPV fractions. In addition to Western blot analyses for GFAP, two marker enzymes were used to further validate the relative lack of glial contamination of SYN fractions. CA, a ubiquitous enzyme localized selectively to glial cells, is associated with the control of ion and fluid movements and acid-base balance (Trachtenberg and Sapirstein, 1980). As shown in Fig. 2-2 (panel B), no carbonic anhydrase activity was evident in the SYN fractions. In contrast, a high level of C A activity was evident in GPV fractions. Finally, studies were conducted to assess activity of GS, a key enzyme which catalyzes the amidation of L-glutamate to glutamine (Palaiologos et ai, 1985). Although previous studies involving immunohistochemical localization of GS have indicated its selective association with glial cells (Martinez-Hernandez et ai, 1977; Norenberg and Martinez-

PAGE 57

51 Hernandez, 1979; Norenberg, 1983) GS activity was detectable in SYN fractions (Fig. 22A). However, we do not know if this finding reflects the presence of intact glial-derived vesicles within the neuronal-derived fractions. In view of the evidence outlined above, it was concluded that our improved method provides an efficient means to isolate glial and neuronal elements from a rat brain homogenate. Therefore, studies were performed to characterize L-glutamate transport in these isolated fractions. Studies of L-[ 3 H]glutamate uptake by GPV and SYN revealed remarkably similar properties for both fractions. Kinetic studies indicated a more rapid initial rate for uptake in GPV, as well as a slightly higher substrate affinity. Conversely, SYN fractions exhibited a 2-3-fold greater capacity for L-[ 3 H]glutamate transport. While there are no appropriate published values with which our findings in GPV can be compared, it should be noted that our highly enriched SYN fractions displayed somewhat lower capacities for L[ 3 H]glutamate uptake than previously reported for crude synaptosomal preparations. The low values in our case may be a consequence of metabolic 'run-down' brought about by the lengthy (6 hrs) purification process. In this regard, it was observed that the GPV fraction undergoes a rapid loss of L-[ 3 H]glutamate uptake capacity following brief (several minutes) exposure to elevated temperatures (data not shown). A limited study with four transport inhibitors (i.e., L-tows-pyrrolidine-2,4-dicarboxylic acid, L-a-aminoadipate, Daspartic acid, and DL-tfrao-p-hydroxyaspartic acid), revealed no remarkable differences between GPV and SYN fractions. This finding is in good agreement with previous studies of these compounds (Dowd et al., 1996; Rauen et al., 1992).

PAGE 58

52 While possible pathways for regulation of L-glutamate transporters in vivo have not been delineated, several mechanisms involving endogenous substances (Barbour et al., 1989; Pogun et al, 1994; Trotti et al, 1996) and transporter phosphorylation (Casado et al, 1991, 1993) have been noted as possible regulators of transporter function. With the advent of this improved technique which separates glia-derived and neuron-derived vesicles from fresh rat brain homogenates, it is feasible to study the regulation of Lglutamate transporter function and to evaluate the potential effects by aging on Lglutamate transporter function in the CNS.

PAGE 59

CHAPTER 3 EFFECT OF PHOSPHORYLATION ON L-[ 3 H]GLUTAMATE UPTAKE Introduction The amino acids L-glutamate and L-aspartate are the predominant excitatory neurotransmitters in the mammalian central nervous system (CNS) (see reviews: Fagg and Foster, 1983; Robinson and Coyle, 1987). Neurotransmission at glutamatergic synapses is terminated by the reuptake of neurotransmitter by sodium-dependent high-affinity transporters located in neuronal and glial cell membranes. Recently, distinct cDNAs (GLT-1, EAAC1, GLAST, EAAT4, EAAT5) that encode subtypes of L-glutamate transporters were isolated (Kanai and Hediger, 1992; Pines etal., 1992; Storck etal., 1992; Fairman et ai, 1995; Arriza et al., 1997). Each of the cloned L-glutamate transporters contain consensus sites for phosphorylation by protein kinase C (PKC). PKC is found in high concentrations in the CNS (Nishizuka, 1992). The term "protein kinase C" encompasses an eleven member family of serine/threonine-specific protein kinases which have been identified functionally by common enzymatic properties, including phorbol ester binding, phospholipid-dependent kinase activity and common structural features (see review: Stabel and Parker, 1991). The PKC cDNA clones first isolated and their corresponding polypeptides are called alpha, beta,, beta 2 , and gamma isoforms. Due to their structural organization, PKC alpha, beta,, beta 2 , and gamma define the "class I" PKC (cPKC) enzymes, those dependent upon Ca 2+ . This is in contrast to the 53

PAGE 60

"class II" enzymes, delta, epsilon, eta, eta', and theta, which are distinguished from class I enzymes based on structure, enzymatic properties, and Ca 2+ -independence (Stabel and Parker, 1991). Class III or atypical PKC isozymes include zeta and lambda (Nishizuka, 1992). These isozymes are phospholipid-dependent, but Ca 2+ independent and do not bind phorbol esters (Nishizuka, 1992). The endogenous activator of PKC is diacylglycerol (DAG). Phorbol esters, which bind specifically to PKC at the DAG binding site (Castagna et aL, 1982), are also potent activators of the enzyme (Nishizuka, 1984). The alpha, beta,, beta 2 , gamma, epsilon, delta, and zeta isoforms and their mRNAs have been identified in the brain using Western and Northern blot analysis and in situ hybridization. Immunohistochemical analysis using isoform-specific antibodies has revealed a differential distribution of the PKC isoforms in the mammalian CNS (Tanaka and Saito, 1992). The gamma isoform is expressed solely in the brain and spinal cord (Saito et aL, 1988). It is localized mainly in cortical pyramidal cells, hippocampal pyramidal and granule cells, cerebellar Purkinje cells, and thalamic neurons (Saito etal., 1988). The beta, isoform is localized mainly in cortical and hippocampal pyramidal cells and striatal neurons. The alpha and delta isoforms are universally distributed in all tissues and cell types so far examined (Ito et aL, 1990). Notably, the glutamatergic neurons in the hippocampal formation display multiple PKC isoforms. The alpha, betaj, and gamma isoforms are contained in pyramidal cells, and the alpha and gamma isoforms in granule cells (Ito et aL, 1990, Kose et aL, 1990). The epsilon isoform is present predominantly in nerve terminals rather than the perikarya (Saito et aL, 1993). Such differences in the

PAGE 61

55 cellular distribution patterns of individual PKC isoforms yields further insight into the role of this enzyme in nerve function, where specificity of its actions may depend on anatomical distribution and substrate specificity. Through the use of phorbol esters, a wide spectrum of functional systems have been demonstrated to be sensitive to PKC activation, including sodium-dependent amino acid transport (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997). The first direct evidence that PKC -mediated phosphorylation is involved in the regulation of L-glutamate transport was provided by Casado and colleagues (1993). Using antibodies directed against biochemically purified GLT-1, this group demonstrated that the amount of phosphorylation correlated with phorbol ester-induced stimulation of L-[ 3 H]glutamate transport activity. One consensus site for phosphorylation by PKC is shared among EAAC1, GLT-1 and GLAST L-glutamate transporters. This is a serine residue, corresponding to serine 1 13 in GLT-1. Studies by Casado and colleagues (1993) revealed that a site-directed mutagenesis of GLT-1 serine residue 1 13 to asparagine abolished stimulation of L-glutamate transport by phorbol esters. Other groups have established further that phosphorylation of brain L-glutamate transporters modulates transport function (Dowd and Robinson, 1996; Conradt and Stoffel, 1997). Preincubation of C6 glioma cells with phorbol ester caused a significant increase in L-glutamate transport activity (Dowd and Robinson, 1996). Observations that GLT-1 and EAAC1 exhibit increased rates of L-glutamate transport following PKC activation are in direct contrast with results reported by Conradt and Stoffel (1997) for

PAGE 62

56 GLAST. In this case, treatment of GL AST-expressing cells with phorbol ester decreased L-glutamate transport activity with phosphorylation occurring at a non-PKC consensus site. Like PKC isoforms, L-glutamate transporters have differences in their regional and cellular locations. GLAST is present in astrocytes (Rothstein et ai, 1995; Lehre et ai, 1995; Schmitt et ai, 1997). Immunocytochemistry (Lehre et al, 1995) and in situ hybridization (Torp et ai, 1994) have localized GLAST mRNA in the Purkinje cell layer of the cerebellum, associated with Bergmann glia (Storck et ai, 1992), as well as in the cortex and hippocampus (Rothstein et ai, 1994). GLT-1 is present in astrocytes throughout the brain, with predominant expression in telencephalic structures, including the hippocampus, neocortex, and striatum (Danbolt et ai, 1992; Rothstein et ai, 1994). Radioactive and non-radioactive in situ hybridization studies have shown that select neurons in the thalamus, hypothalamus, and pyramidal cells of the hippocampus express GLT-1 mRNA (Torp et ai, 1994; Schmitt et ai, 1996). Expression of mRNA for EAAC1, a neuronal L-glutamate transporter, is abundant in the pyramidal layer of the hippocampus (regions CA1-CA4), the granular layer of the dentate gyrus, the granule cell layer of the cerebellum and layers II-IV of the cerebral cortex (Kanai and Hediger, 1992). The cellular locations of the more recently cloned EAAT4 (Fairman et ai, 1995), which is expressed in the cerebellum, and EAAT5 (Arriza et ai, 1997), which is expressed in the retina, have not been published. However, given these differences in cellular and regional localization of L-glutamate transporters, as well as their responses following activation of PKC, changes in L-glutamate transport activity may be a reflection of the PKC isoform

PAGE 63

57 present in the same cell as the L-glutamate transporter. In essence, the presence of PKC isoforms may differentially and/or selectively modulate L-glutamate transport depending upon their cellular location. In contrast to synaptosomes, which have been used extensively to characterize neuronal processes in the CNS, few preparations derived from CNS tissues have been available for the study of glial functions. Since it is possible for cellular properties to be altered under culture conditions, it is prudent to determine the degree to which cultured cells accurately reflect cell functions in living tissues. In view of these considerations, we developed a technique which yields functionally viable glial(glial plasmalemmal vesicles or GPV) and neuronal(synaptosomes or SYN) enriched fractions from rat brain homogenates. Morphological, biochemical, and Western blot analyses previously confirmed the separation of glialand neuron-derived fractions (Daniels and Vickroy, 1998). Prior studies confirmed that GPV and SYN tissue fractions exhibit high-affinity Lglutamate transport (Daniels and Vickroy, 1998) and can be used to study phosphorylation-dependent regulation of neuronal and glial L-glutamate transporters. Experimental Procedures Chemicals L-[2,3,43 H]-Glutamic acid (specific activity = 60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). Phorbol-12,13-dibutyrate was bought from Sigma Chemical Co. (St. Louis, MO, USA). Okadaic acid (potassium salt)

PAGE 64

58 and calphostin C were purchased from Calbiochem Novabiochem International (La Jolla, CA, USA). All other chemicals were purchased from commercial vendors and were of the highest quality available. Preparation of Tissue Glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) were prepared as previously described (Daniels and Vickroy, 1998). Briefly, young adult (3-4 months of age) Sprague-Dawley male rats (Zivic Miller) weighing 250-275 g were used throughout this study. Animals were housed in pairs and maintained on a 12-hr light/dark cycle with food and water available ad libitum and were transported to the laboratory approximately 15 hrs prior to use. For each experiment, two rats were decapitated quickly with a small animal guillotine, and the brains were removed rapidly and placed upon an ice-cold glass surface. Cerebellar tissue was removed and discarded, while all remaining forebrain tissue (approximately 2.4 g) was placed in 30 ml of an ice-cold solution containing 0.32 M sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA). The tissue was homogenized gently with a Potter-Elvehjem tissue grinder (approximately 30 rpm) and centrifuged at 1000 x g for 10 min (4°C) using a fixed-angle rotor (F28/50-DuPont). All subsequent centrifugation steps were conducted at 4°C. The resultant pellet was discarded, and the supernatant was split into four equal portions, which were diluted to 30 ml with an icecold solution containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM dithiothreitol and 20 mM HEPES (pH 7.4 at 4°C). Hereafter, this solution is referred to as SEDH. Diluted aliquots of supernatant were centrifuged at 5000 x g (15 min), and resultant supernatants were saved separately on ice. Each of the four tissue pellets were resuspended in 15 ml of

PAGE 65

ice-cold SEDH solution and centrifiiged at 1000 x g for 10 min. The resultant pellets were discarded, while supernatants were combined with supernatants saved from the previous step. The four tubes, each containing approximately 45 ml of tissue homogenate, were centrifiiged at 33,500 x g (20 min), and supernatants were discarded. Tissue pellets (four) were resuspended in 1 5 ml of ice-cold SEDH solution and gently transferred onto a three-step discontinuous Percoll gradient (10 ml each of 1 .38%, 2.3%, and 4.6% Percoll in SEDH solution) with a Minipuls 2 (Gilson) peristaltic pump (flow rate = 0.88 ml/min). Tubes were centrifiiged at 33,500 x g (10 min) with 15-min periods of linear acceleration to and deceleration from the top speed. The turbid layer between 0% and 1 .38% Percoll was collected from all four tubes and combined into two aliquots, which were diluted to a final volume of 15 ml each with ice-cold SEDH solution. Aliquots were centrifiiged at 1000 x g (20 min), and resultant supernatants were layered onto fresh three-step Percoll gradients as described above. Tubes were centrifiiged at 33,500 x g (10 min) with gradual acceleration and deceleration (see above), and the turbid layer between 0% and 1.38% Percoll was collected from both tubes and combined into a single aliquot. The tissue aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution, centrifiiged at 33,500 x g (20 min), and the resultant pellet was used as the GPV fraction. For the recovery of the SYN fraction, the turbid layer between 2.38% and 4.6% Percoll was collected from the initial discontinuous gradient. The four aliquots were diluted to a final volume of 15 ml (each) with ice-cold SEDH solution and centrifiiged at 1000 x g (20 min). Pellets were discarded, and supernatants were layered onto a three-step Percoll gradient (see above) and centrifiiged at 33,500 x g (10 min) with gradual acceleration and

PAGE 66

deceleration periods. The turbid layer between 2.38% and 4.6% Percoll was collected from each of the four tubes and combined into one aliquot. This aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution and centrifuged at 17,500 x g (20 min). The resultant pellet was designated as the SYN fraction. L-| 3 H JGIutamate Uptake The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 25 °C). GPV and SYN were used immediately for uptake. Sodium-dependent L-[ 3 H]glutamate uptake by GPV and SYN was measured by a filtration method modified from Divac and coworkers (1977). For the measurement of uptake, aliquots (50 /A) of tissue fractions (approximately 1 mg protein/ml) were added to cold glass culture tubes that contained 400 (A of a buffered solution containing (in mM) NaCl (140), KC1 (5), CaCl 2 (1.0), MgCl 2 (1.0), NaH 2 P0 4 (1.2), D-glucose (10), and HEPES (20) at pH 7.4 (25 °C). Sodium-independent uptake (blanks) was measured in parallel using a buffer in which NaCl was replaced by an isosmolar concentration of choline chloride. The uptake reaction was initiated by the addition of 50 (A of L-[2,3,43 H]-glutamic acid to duplicate assay tubes. Tubes were mixed rapidly and returned to shaking water bath (25 °C) for 90 seconds. Uptake was terminated by rapid vacuum filtration using a Brandel cell harvester and Whatman GF/B filter sheets that had been presoaked overnight at 4°C in 25 mM Lglutamate. Test tubes and filters were rinsed rapidly three times with 2 ml aliquots of icecold normal or sodium-deficient buffer. Tissue trapped on filters was digested with 2 ml of 0.2 M NaOH (overnight), acidified with 1 ml of 0.5 M HC1, and assayed for tritium content in 10 ml of EcoLume scintillation fluid (ICN Biochemicals). Radioactivity was

PAGE 67

61 quantified in an LKB 1214 liquid scintillation counter with a counting efficiency of approximately 45% as determined by a radium-226 standard. Sodium-dependent uptake was determined as the difference between uptake in normal versus sodium-deficient buffers. For concentration-response studies, tissue fractions were preincubated for 5 min in the presence of phorbol-12,13-dibutyrate (PDBu), ranging in concentration between 300 nM 10 juM, or its vehicle (0.01% acetone v/v), prior to determination of L[ 3 H]glutamate uptake. A similar treatment protocol was used to evaluate possible effects by okadaic acid. For studies testing the interaction between PDBu and okadaic acid, tissue fractions were exposed to both drugs simultaneously for 5 min prior to determination of L-[ 3 H]glutamate uptake. For studies of the interaction between PDBu and calphostin C, tissues were pretreated with calphostin C for 5 min prior to addition of PDBu. Statistical Evaluation For the results reported below, the n indicates the number of experiments conducted on separate days. Data were expressed as the mean ± S.E.M. To determine statistical significance, One-way Analysis of Variance (ANOVA) was used, followed, when appropriate, by Bonferroni's t test. Significance was accepted at a level of p< 0.05.

PAGE 68

62 Results Concentration-Dependence of PDBu-Induced Facilitation of L-[ 3 H]Glutamate Uptake As shown in Figure 3-1, a 5 min pretreatment with PDBu (300 nM -10 (jM) stimulated L-[ 3 H]glutamate uptake in a concentration-dependent manner in GPV but had no significant effect on synaptosomal L-glutamate uptake. The maximal PDBu effect in GPV occurred at 10 //M (56 ± 15% above control). Significant differences (p<0.05) between GPV and vehicle were observed at both 7 fjM (45 ± 12% above control) and 10 IjM PDBu (56 ± 15% above control). Kinetic Analysis of PDBu-Induced Facilitation of L-[ 3 H]Glutamate Uptake In order to determine whether the PDBu-induced increase in L-[ 3 H]glutamate uptake was due to a change in the number of active transporters (apparent V^J or a change in transporter affinity for L-glutamate (apparent KJ, the concentration dependence of L-[ 3 H]glutamate uptake was examined after preincubation of tissue fractions with 10 fuM PDBu for 5 min (Figure 3-2). Under these conditions, PDBu increased the V,^ for L-glutamate transport in GPV but not SYN and did not change the K,,, of L-[ 3 H]glutamate uptake (Figure 3-2, panels A and B). For GPV (panel A), the V,^ of vehicle-treated controls was 254 ± 23 pmol/mg protein/90sec. Following pretreatment with PDBu, the V^ increased to 365 ± 21 pmol/mg protein/90sec (p<0.05). In vehicletreated GPV tissue fractions the K„ was 1 .0 /jM ± 0.58 //M, and in PDBu-treated GPV the K,,, was 1 .9 //M ± 0.43 fj,M. In vehicle-treated control SYN tissue fractions (panel B),

PAGE 69

63 the was 1648 ± 58 pmol/mg protein/90sec with a K™ of 1 .8 ± 0.63 //M. In the presence of 10 fjM PDBu, the was 1989 ± 83 pmol/mg protein/90sec with a K„, of 3.9 ± 0.94 mM. Specificity of Protein Kinase C Pathway for PDBu-Induced Facilitation of L[ 3 H]Glutamate Uptake To determine whether enhanced protein kinase C activity via PDBu was causing the increase in L-[ 3 H]glutamate uptake, the effect of an irreversible protein kinase C inhibitor, calphostin C, (Kobayashi et al, 1989) was evaluated for its effectiveness in blocking the PDBu-induced enhancement in L-[ 3 H]glutamate uptake. As shown in Figure 3-3 (panel A), calphostin C blocked 10 fjM PDBu-enhanced L-[ 3 H]glutamate uptake in a concentration-dependent manner. Calphostin C at concentrations of either 0. 1 /jM or 0.5 A*M alone (panels A and B) did not affect L-[ 3 H]glutamate uptake. For GPV (panel A), when 0. 1 calphostin C was combined with 10 fjM PDBu, L-[ 3 H]glutamate uptake was reduced from 46 ± 4% above control (p<0.05) to 17 ± 6% above control. The combination of 0.5 fiM calphostin C with 10 yuM PDBu further reduced transport from 46 ± 4% above control to 6 ± 2% above control (p<0.05). For SYN, no significant changes were produced by any of these treatments. Effect of Okadaic Acid, a Phosphatase Inhibitor, on L~[ 3 H]Glutamate Transport The concentration-response curve for OKA is shown in Figure 3-4. In GPV, a 5 min pretreatment with OKA at concentrations of either 300 nM or 1000 nM significantly enhanced (p<0.05) L-[ 3 H]glutamate uptake (24.4 ± 2.3% above control and 32.6 ± 5.0% above control, respectively). Pretreatment of SYN for 5 min with increasing

PAGE 70

64 concentrations of OKA produced no significant change in L-[ 3 H]glutamate uptake. When 10 (jM PDBu was combined with either 300 nM or 1000 nM OKA for 5 min, values of L[ 3 H]glutamate uptake enhancement remained the same as values obtained in the presence of 10 v-M PDBu or 300 nM or 1000 nM OKA alone (data not shown). Discussion While specific cellular mechanisms for regulation of L-glutamate transport in vivo have not been reported, several endogenous factors, such as arachidonic acid and nitric oxide (Barbour etai, 1989; Pogun and Kuhar, 1993) and transporter phosphorylation (Casado etal., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997) have been proposed as possible regulators of transporter function. Consensus sites for phosphorylation by PKC are present in each of the recently cloned L-glutamate transporters (Kanai and Hediger, 1992; Pines et al., 1992; Storck etal., 1992, Fairman et al, 1995; Arriza et al., 1997). The existence of these putative phosphorylation sites suggests that L-glutamate transporters may be regulated by protein kinases and phosphatases. Primary astrocyte cultures and cell lines expressing a homogenous population of L-glutamate transporters have demonstrated that L-glutamate transport activity may be modulated by PKC activation (Casado et al., 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997). In this study we used glialand neuronalenriched fractions isolated by three-step density gradient centrifugation to investigate phosphorylation-dependent regulation of L-glutamate transport. The first step in addressing the issue of a change in L-[ 3 H]glutamate transport activity due to phosphorylation was to expose GPV and SYN tissue fractions to increasing

PAGE 71

65 0 2 4 6 8 PDBu Cone. (|uM) Figure 3-1 Concentration dependence for PDBu-induced facilitation of L-[ 3 H]glutamate uptake. GPV and S YN fractions were preincubated for 5 minutes in the presence of PDBu or vehicle (0.01% acetone, v/v) then assayed for uptake in the presence of 5//M L[ 3 H]glutamate. Data points represent averaged values (mean ± SEM) from five experiments. Asterisks (*) denote a significant difference (p<0.05) between PDBu-treated samples and vehicle-treated controls as determined by one-way ANOVA followed by Bonferroni's / test.

PAGE 72

66 u O c — 93 O — a. M E 400 300 200 100 o 0 £ 0 20 40 60 80 100 a. 2500 2000 "S 1500 J3 a 5c i 1000 0—0 Vehicle PDBu i i i r~ 0 20 40 60 80 100 L-[ 3 H]Glutamate Cone. (jiM) Figure 3-2. Substrate-saturation curves for L-[ 3 H]glutamate uptake in PDBu-treated tissue fractions. GPV (panel A) and SYN (panel B) fractions were incubated for 5 minutes in the presence of 10//M PDBu or vehicle prior to measurement of uptake in the presence of L-[ 3 H]glutamate (0.02//M lOO^M). Data points represent the mean ± SEM from 3-4 experiments. Fitted curves were obtained by least squares regression analysis as described under Experimental Procedures. PDBu-treatment of GPV fractions caused significant enhancement (p<0.05) of L-[ 3 H]glutamate uptake as indicated by one-way ANOVA

PAGE 73

o — o — o o o — i 2 in d ([OIJUOO 3AOCJTJ %) e •a o. 3 Q — 5 € *8 S a c ftm z I 2 6 CO T3 ft. g CD c S T3 M CD T3 CD o O •E 3 m Q ft. B O u .s e 99 • E 1) *J *-> CO 3 O B •5 "3. s ^ o t2 o £ il •S I $ e .£Pn > M Z '5b u u « 09 E _o cc H E s o E o U E _o '-*-» cj 03 cfc U 3 55 C/l CO I CD E O >, X) -a E CD +-» -H « . O C, 3 3 £ OQ SO U Q tU £ ft. E cs S O o JS o. •a o

PAGE 74

68 10 100 1000 10000 OKA Cone. (nM) Figure 3-4. Concentration dependence for OKA-induced facilitation of L-[ 3 H]glutamate uptake. GPV and SYN fractions were preincubated for 5 minutes in the presence of OKA or buffer then assayed for uptake in the presence of L-[ 3 H]glutamate. Data are expressed as mean ± SEM for five experiments. Asterisks (*) denote significant differences (p<0.05) between OKA-treated samples and vehicle-treated controls as determined by one-way ANOVA followed by Bonferroni's / test.

PAGE 75

69 concentrations of the PKC activator, PDBu. Tumor-promoting phorbol esters, such as PDBu, activate all known PKC isoenzymes directly, with the exception of the atypical PKCs, zeta and lambda, by binding at the DAG site of PKC (Nishizuka, 1984). Zeta and lambda isoforms are activated by phosphotidylserine but are not affected by DAG, phorbol esters, or calcium (Ono et ai, 1989; Nakanishi and Exton, 1992). L-[ 3 H]glutamate transport activity in GPV fractions responded to PDBu in a concentration-dependent manner (Figure 3-1). This increase in L-[ 3 H]glutamate transport appears to be due to an increase in the number of active transporters, as GPV affinity for L-[ 3 H]glutamate did not change following exposure to PDBu (1 .0 ± 0.44 (jM vs. 1 .9 ± 0.52 //M), but the V ra did (254 ± 23 vs. 365 ± 21 pmol/mg protein/90sec; p<0.05) (Figure 3-2). These findings contrast with those of SYN in which L-[ 3 H]glutamate uptake was not altered significantly from control levels following PDBu treatment. These findings were substantiated further by substrate-saturation studies (Figure 3-2), wherein neither the affinity for L[ 3 H]glutamate nor the maximal rate of L-[ 3 H]glutamate transport in SYN changed significantly from control levels. Taken together, these results indicate a potential means of differential regulation of L-glutamate transport in GPV versus SYN fractions by PKCmediated phosphorylation. Considering these results, it was necessary to confirm that PDBu-induced facilitation of L-[ 3 H]glutamate transport was occurring through activation of PKC. In order to evaluate the basis for PDBu activation of L-glutamate transport, tissue fractions were pretreated with calphostin C, prior to exposure to PDBu. Calphostin C is the most potent and specific inhibitor of PKC among the family of calphostins and blocks binding at

PAGE 76

70 the DAG / phorbol ester binding site (Kobayashi et al., 1989; Tamaoki et al., 1990). At a concentration of 1 ^M, it causes complete inhibition of the binding of 50 nM [ 3 H]PDBu to PKC and subsequent inhibition of its activity (Kobayashi et al., 1989). Notably, calphostin C inhibited the PDBu-induced facilitation of GPV L-[ 3 H]glutamate transport in a concentration-dependent manner (Figure 3-3). These results confirm that PDBu is activating PKC. There are several possibilities as to why there is a difference between the two vesicle fractions in response to PDBu pretreatment. First, there may be a difference in the size of the PKC "pool" between GPV and SYN. It is possible that the GPV PKC "pool" is simply larger than that of SYN, and as such, responds more robustly to exposure to a PKC activator like PDBu. This, of course, assumes that the PKC isoform(s) responsible for mediation of this response is actually present in both tissue fractions but at a lower level in SYN. This may not be the case. Like L-glutamate transporters, PKC isoforms have specific regional and cellular localizations. Certainly, homogenization has disrupted regional specificity. However, cellular specificity may be conserved, and SYN fractions may not contain PKC isoforms which are activated by phorbol esters. This does not mean that SYN L-glutamate transport cannot be modulated by phosphorylation, but that this potential modulation may take place through atypical PKCs, which are not activated by phorbol esters. Second, the ability of PKCs to interact with and be activated by membranes (phospholipids) and DAG presents a functional definition of these proteins. As such, binding to lipids is a critical and indeed well-studied phenomenon (see review: Epand and Lester, 1990). Even though general levels of phospholipids were evaluated

PAGE 77

71 previously in GPV and SYN tissue fractions and found not to differ significantly (Daniels and Vickroy, 1998), it is possible that other important PKC activators, such as Ca 2+ , is present to a lesser degree in SYN versus GPV. Again, this would limit the ability of PDBu to enhance SYN L-glutamate transport. Third, it is possible that SYN L-glutamate transport may be modulated by phosphorylation but simply not by PKC. In addition to having consensus PKC sites, the cloned L-glutamate transporters also contain consensus sequences for protein kinase A (Kanai and Hediger, 1992; Pines etai, 1992; Storck etal, 1992; Fairman et al., 1995). Conradt and Stoffel (1997) reported that treatment of GLAST-expressing cells with phorbol ester decreased L-glutamate transport activity with phosphorylation occurring at a non-PKC consensus site. With the finding that phosphorylation by PKC could modulate L-[ 3 H]glutamate transport, it seemed prudent to investigate the possible effect(s) that dephosphorylation by phosphoserine/phosphothreonine phosphatases could have on L-[ 3 H]glutamate transport activity in GPV and SYN. These types of phosphatases can be differentiated from one another on the basis of differences in substrate specificities, dependence on divalent cations, sensitivities to specific inhibitors, and their catalytic subunits (Ingebritsen et al, 1983; Cohen, 1989). Based upon the four types of catalytic subunits, the phosphoserine/phosphothreonine-specific protein phosphatases have been divided into four main classes of enzymes: PP1, PP2A, PP2B, and PP2C (Ingebritsen et al, 1983; Cohen, 1989). Protein phosphatase 1 (PP1) preferentially dephosphorylates the beta-subunit of phosphorylase kinase and is inhibited by two heat-stable inhibitor proteins, inhibitor1

PAGE 78

72 (which inhibits PP1 after phosphorylation by cyclicAMP-dependent protein kinase) and inhibitor-2 (also termed "modulator," which inhibits PP1 by impeding the substrate binding and by inducing a conformational change of the catalytic subunit) (see review: Bollen and Stalmans, 1992). Subcellular fractionation studies have demonstrated PP1 activity in cytosolic, synaptosolic, synaptic plasma membrane and synaptic junction fractions (Shields et ai, 1985; Dokas et ai, 1990). These findings are consistent with a ubiquitous distribution of this phosphatase in brain. Alternatively, type 2 phosphatases preferentially dephosphorylate the alpha-subunit of phosphorylase kinase and are insensitive to inhibitor1 and inhibitor-2. Type 2 protein phosphatases are subdivided into three distinct classes based on their cationic requirements: PP2A, PP2B, and PP2C. PP2A is active in the absence of divalent cations. PP2B (i.e. calcium/calmodulin-dependent protein phosphatase or calcineurin) and PP2C are completely dependent on calcium and magnesium, respectively. The amount of PP2A activity in brain extracts is the highest of any tissues investigated (Ingebritsen et ai, 1983), and there is approximately three times as much PP2A as PP1. As noted in several reviews (see: Cohen, 1989; Sim, 1991; Bollen and Stalmans, 1992; Shenolikar, 1994), substrate specificity of phosphatases may be controlled by a number of different regulatory subunits. These regulatory subunits direct phosphatase activity toward specific subcellular localizations and, therefore, toward specific substrates. OKA is a PP1 and PP2A inhibitor with a higher affinity for PP2A (IC 50 = 0. 1 nM) than for PP1 (IC 50 = 10 1 5 nM) (Cohen et ai, 1990). Pretreatment of GPV and SYN fractions with OKA significantly enhanced L-[ 3 H]glutamate transport, but only in GPV

PAGE 79

73 fractions (Figure 3-4). Again, these results suggest a differential modulation of L[ 3 H]glutamate transport between GPV and SYN tissue fractions. As to why there is a difference between the two tissue fractions, it is possible that GPV and SYN differ in the size of their "pools" containing PP1 and PP2A or in the distribution of PP1 and PP2A within these "pools." The latter may be true considering the results from exposure of the tissue fractions to OKA. GPV fractions may contain a higher level of PP2A, thus inhibiting this phosphatase, increases L-glutamate transport due to inhibiting dephosphorylation. However, regulatory subunits direct phosphatase activity toward specific subcellular localizations and therefore toward specific substrates. It is possible that following homogenization, this process is perturbated, so that phosphatases are no longer directed toward L-glutamate transporters, or conversely, that phosphatases are directed now inappropriately towards L-glutamate transporters. Also of interest are the results from studies in which GPV and SYN fractions were treated at the same time with both PDBu and OKA (data not shown). The increase in L-glutamate transport activity was not additive in either fraction. In other words the increased L-glutamate uptake in GPV following PDBu treatment was not enhanced further by additional treatment with OKA. SYN L-glutamate uptake remained at control levels, and GPV L-glutamate uptake increased to levels seen with either PDBu or OKA alone. This result suggests that PDBuand OKA-induced facilitation of L-glutamate transport in GPV fractions is occurring through a common pathway. In summary, we have demonstrated a significant increase in Na + -dependent highaffinity L-[ 3 H]glutamate transport in glial versus neuronal tissue fractions. This increase is

PAGE 80

consistent with activation of PKC and subsequent phosphorylation of GPV L-glutamate transporters. Furthermore, activity of GPV L-glutamate transporters could be enhanced by blockade of PP1 and PP2A phosphatases. Together, these results suggest that phosphorylation / dephosphorylation events differentially modulate glial and neuronal Lglutamate transport, and provide some insight into these potential regulatory mechanisms in CNS tissue. It is possible that modulation of L-glutamate transport depends upon the transporters cellular colocalization with specific PKC isoforms. Future studies need to address the type and levels of PKC isoforms within the glialand neuronal-enriched fractions in order to further access this potential means of regulation.

PAGE 81

CHAPTER 4 EFFECT OF AGING ON L-[ 3 H]GLUTAMATE UPTAKE Introduction Reuptake via sodium-dependent high-affinity transport systems in neurons and glial cells is the major mechanism for termination of L-glutamate synaptic actions (Nicholls and Attwell, 1990) and maintenance of its extracellular concentration below toxic levels (McBean and Roberts, 1985). Several studies have examined EAA uptake in tissue preparations from a variety of areas of the aging rat brain. In the neocortex, uptake of both L-[ 14 C]glutamate into a preparation of synaptosomes and D-[ 3 H]aspartate in crude homogenates was found to decline with age, although these changes appear to reflect differences between immature (2-4 months of age) and mature (10-30 months of age) animals (Strong et al, 1984; Meldrum et al, 1992). By contrast, no difference was observed in L-[ 3 H]glutamate uptake into slices of frontal neocortex in a comparison of animals aged 6 and 24 months (Dawson et al, 1989). In the neostriatum, uptake both of L-[ 3 H]glutamate (in crude synaptosomal fractions) and D-[ 3 H]aspartate (in crude homogenates) was lower in mature rats (aged 10 months or more) than in immature animals, aged 6 months or less (Price et al, 1981; Wheeler and Ondo, 1986; Najlerahim et al., 1990). Additionally, Palmer and colleagues in 1994, using Fisher 344 x Brown Norway rats aged 3, 12, 24, and 37 months, assayed D-[ 3 H] aspartate uptake in crude cortical, hippocampal, and neostriatal synaptosomes and found no significant changes with 75

PAGE 82

76 increasing age. Results from investigating L-glutamate transport across aging in crude synaptosomes, however, may be misleading, since these preparations are contaminated with glial cells, which also accumulate L-glutamate. Given the necessity of proper functioning of L-glutamate transporters, it is of extreme importance to understand their regulatory mechanisms. The isolation of cDNAs for several sodium-dependent high-affinity L-glutamate transporters (GLAST, GLT-1, EAAC1, EAAT4, EAAT5) has furthered the idea first raised by Casado and colleagues (1991, 1993) that phosphorylation may provide a regulatory mechanism for L-glutamate transport since cloned transporters contain putative protein kinase C phosphorylation sites (Kanai and Hediger, 1992; Pines et al, 1992; Storck et al, 1992; Fairman et al, 1995; Arriza et al, 1997). Immunological studies have localized GLAST and GLT-1 primarily to glial cells (Rothstein et al, 1994, 1995), though GLT-1 is present in certain neurons (Torp et al, 1994; Schmitt et al, 1996). EAAC1 is located in neurons (Kanai and Hediger, 1992; Rothstein et al, 1994; Velaz-Faircloth etal, 1996). The cellular localization of EAAT4 and EAAT5 expression has not been determined as of yet. Several reports within the last few years have investigated the potential regulatory role that phosphorylation via protein kinase C (PKC) may play in modifying L-glutamate transport activity (Casado etal, 1991, 1993; Dowd and Robinson, 1996; Conradt and Stoffel, 1997). L-glutamate transport by GLT-1 and EAAC1 appears to be increased following incubation with phorbol esters (Casado et al, 1993; Dowd and Robinson, 1996), while GLAST-mediated transport has been reported to decline (Conradt and Stoffel, 1997). However, since these studies were conducted in cultured cells, it is

PAGE 83

77 impossible to use this approach to assess possible changes in phosphorylation-mediated regulation of L-glutamate transport as a function of aging. With the advent of an improved density gradient centrifugation technique with which glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) can be isolated from a rat brain homogenate (Daniels and Vickroy, 1998), it is now possible to investigate the potential effects that aging may have on L-glutamate transport and the role that phosphorylation may play in regulating transport function. Experimental Procedures Chemicals L-[2,3,43 H]-Glutamic acid (specific activity = 60 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). Phorbol-12,13-dibutyrate was bought from Sigma Chemical Co. (St. Louis, MO, USA). Primary antibody against Excitatory Amino Acid Carrier 1 (EAAC1) was a gift from Dr. Michael Kilberg (University of Florida, Gainesville, FL, USA), and primary antibody against GLT-1 was a gift from Dr. Jeffrey Rothstein (Johns Hopkins University, Baltimore, MD, USA). Secondary antibody (horseradish peroxidase conjugated to protein A), as well as the horseradish peroxidase color reagent kit, were bought from Biorad (Hercules, CA, USA). All other chemicals were purchased from commercial vendors and were of the highest quality available. Preparation of Tissue Glial plasmalemmal vesicles (GPV) and synaptosomes (SYN) were prepared as described previously (Daniels and Vickroy, 1998). Male Fisher 344 x Brown Norway Fl

PAGE 84

78 rats (NIA) aged 5, 25, 3 1, and 37 months, weighing 350 450 g, were used throughout this study. Animals were housed in pairs on a 12-hr light/dark cycle with food and water available ad libitum under "pathogen-free" conditions for 5-10 days prior to use. Animals were transported to the laboratory approximately 15 hrs prior to use. For each assay, two rats of the same age were decapitated quickly with a small animal guillotine, and the brains were removed rapidly and placed upon an ice-cold glass surface. Cerebellar tissue was removed and discarded, while all remaining forebrain tissue (approximately 2.4 g) was placed in 30 ml of an ice-cold solution containing 0.32 M sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA). The tissue was homogenized gently with a Potter-Elvehjem tissue grinder (approximately 30 rpm) and centrifuged at 1000 x g for 10 min (4°C) using a fixed-angle rotor (F28/50-DuPont). All subsequent centrifugation steps were conducted at 4°C. The resultant pellet was discarded, and the supernatant was split into four equal portions, which were diluted to 30 ml with an ice-cold solution containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM dithiothreitol and 20 mM HEPES (pH 7.4 at 4°C). Hereafter, this solution is referred to as SEDH. Diluted aliquots of supernatant were centrifuged at 5000 x g (15 min), and resultant supernatants were saved separately on ice. Each of the four tissue pellets were resuspended in 15 ml of ice-cold SEDH solution and centrifuged at 1000 x g for 10 min. The resultant pellets were discarded, while supernatants were combined with supernatants saved from the previous step. The four tubes, each containing approximately 45 ml of tissue homogenate, were centrifuged at 33,500 x g (20 min), and supernatants were discarded. Tissue pellets (four) were resuspended in 15 ml of ice-cold SEDH solution and gently transferred onto a three-

PAGE 85

79 step discontinuous Percoll gradient (10ml each of 1 .38%, 2.3%, and 4.6% Percoll in SEDH solution) with a Minipuls 2 (Gilson) peristaltic pump (flow rate = 0.88 ml/min). Tubes were centrifuged at 33,500 x g (10 min) with 15-min periods of linear acceleration to and deceleration from the top speed. The turbid layer between 0% and 1.38% Percoll was collected from all four tubes and combined into two aliquots, which were diluted to a final volume of 1 5 ml each with ice-cold SEDH solution. Aliquots were centrifuged at 1000 x g (20 min), and resultant supernatants were layered onto fresh three-step Percoll gradients as described above. Tubes were centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration (see above), and the turbid layer between 0% and 1.38% Percoll was collected from both tubes and combined into a single aliquot. The tissue aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution, centrifuged at 33,500 x g (20 min), and the resultant pellet was used as the GPV fraction. For the recovery of the SYN fraction, the turbid layer between 2.38% and 4.6% Percoll was collected from the initial discontinuous gradient. The four aliquots were diluted to a final volume of 15 ml (each) with ice-cold SEDH solution and centrifuged at 1000 x g (20 min). Pellets were discarded, and supernatants were layered onto a three-step Percoll gradient (see above) and centrifuged at 33,500 x g (10 min) with gradual acceleration and deceleration periods. The turbid layer between 2.38% and 4 6% Percoll was collected from each of the four tubes and combined into one aliquot. This aliquot was diluted to a final volume of 45 ml with ice-cold SEDH solution and centrifuged at 17,500 x g (20 min). The resultant pellet was designated as the SYN fraction.

PAGE 86

80 L-[ J H]Glutamate Uptake The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 25 °C). GPV and SYN were used immediately for uptake. Sodium-dependent L-[ 3 H]glutamate uptake by GPV and SYN was measured by a filtration method modified from Divac and coworkers (1977). For the measurement of uptake, aliquots (50 /A) of tissue fractions (approximately 1 mg protein/ml) were added to cold glass culture tubes that contained 400 fi\ of a buffered solution containing (in mM) NaCl (140), KC1 (5), CaCl 2 (1.0), MgCl 2 (1.0), NaH 2 P0 4 (1.2), D-glucose (10), and HEPES (20) at pH 7.4 (25 °C). Sodium-independent uptake (blanks) was measured in parallel using a buffer in which NaCl was replaced by an isosmolar concentration of choline chloride. The uptake reaction was initiated by the addition of 50 /A of L-[2,3,43 H]-glutamic acid to duplicate assay tubes. Tubes were mixed rapidly and returned to shaking water bath (25 °C) for 90 seconds. For concentration-response studies, tissue fractions were preincubated for 5 min in the presence of phorbol-12,13-dibutyrate (PDBu), ranging in concentration between 1 //M 100 /iM, or its vehicle (0.01% acetone v/v), prior to determination of L[ 3 H]glutamate uptake. Uptake was terminated by rapid vacuum filtration using a Brandel cell harvester and Whatman GF/B filter sheets that had been presoaked overnight at 4°C in 25 mM L-glutamate. Test tubes and filters were rinsed rapidly three times with 2 ml aliquots of ice-cold normal or sodium-deficient buffer. Tissue trapped on filters was digested with 2 ml of 0.2 M NaOH (overnight), acidified with 1 ml of 0.5 M HC1, and assayed for tritium content in 10 ml of EcoLume scintillation fluid (ICN Biochemicals). Radioactivity was quantified in an LKB 1214 liquid scintillation counter with a counting

PAGE 87

81 efficiency of approximately 45% as determined by a radium-226 standard. Sodiumdependent uptake was determined as the difference between uptake in normal versus sodium-deficient buffers. Western Blot Analyses The GPV and SYN pellets were collected and then resuspended in SEDH solution (pH 7.4 at 25°C). Aliquots (100 (A) were placed immediately in liquid nitrogen and stored at -80 °C until needed. For immunoblotting of antibodies against EAAC1 and GLT-1, aliquots (1 /zg protein / 20 iA for GLT-1 or 5 /zg protein / 20 fA for EAAC1) of GPV and SYN samples from animals aged 5, 25, 31, and 37 months were subjected to SDS-polyacrylamide gel electrophoresis (200V for 2.5 hr). The electrophoresed proteins were transferred onto polyvinylidene difluoride membrane (Immobilon P, Millipore) (100A for 1.25 hr). Blots were incubated (1.5 hr) with blocking solution [5% non-fat carnation instant milk in Tris-Tricine solution (TBS/T) (10 raM Tris-HCl, 200 mM NaCl, 0.05% Tween-20)], then incubated in primary antibody (1:50 for EAAC1, 1 :500 for GLT1) for 1.5 hr. Blots were washed five times (5 min each) with blocking solution, then incubated with horseradish peroxidase-conjugated protein A (1 :5000) for 1 nr. The secondary antibody was removed, and blots were rinsed 5 times (5 min each) with blocking solution, followed by a 10 min wash with TBS/T. Protein bands were visualized via horseradish peroxidase color reagent development kit (Biorad). Protein Analyses The amount of protein in GPV and SYN samples was determined by the Lowry protein assay (Lowry et al, 1951), using solutions of bovine serum albumin as standards.

PAGE 88

82 Data Analyses All values reported were the mean ± SEM from five fractions isolated on separate experimental days. Statistical significance, accepted at a level of p<0.05, was determined via repeated measures analysis of variance (ANOVA), followed by Tukey's multiple pairwise comparison procedure. Results Effect of Aging on L-[ 3 H]gIutamate Uptake As shown in Figure 4-1 (panel A), L-[ 3 H]glutamate uptake values for GPV fractions did not differ among animals at 5, 25, 31 or 37 months of age. In SYN fractions (panel B), L[ 3 H]glutamate transport at 37 months was significantly reduced (F 3 16 =4.20, p<0.03) from 5 month values (712 ± 31 vs. 562 ± 40 pmol/mg protein/90sec). Effect of Aging on PDBu-induced Facilitation of L-[ 3 H]glutamate Uptake Within age group evaluation Another key issue to be answered was whether or not PDBu would facilitate an increase in L-[ 3 FT]glutamate transport within each of the four age groups examined (Figures 4-2 and 4-3). Figure 4-2 shows the PDBu concentration-response curves for GPV fractions obtained from 5 (panel A), 25 (panel B), 3 1 (panel C), and 37 (panel D) month-old animals. Significant increases in L-[ 3 H]glutamate transport above control values occurred in each of the four age groups (F 5 ^=97.38, p<0.001). PDBu concentrations of 3 /jM and above facilitated L-[ 3 H]glutamate transport in all age groups, while 1 ijM PDBu yielded a significant increase above control values in 3 1 month-old animals. Likewise, SYN tissue fractions (Figure 4-3) also showed a PDBu-induced

PAGE 89

facilitation of L-[ 3 H]glutamate transport above control levels in each of the four age groups examined (F J62 =61.84, p<0.001). PDBu concentrations of 10 /uM and above significantly facilitated transport at 5 months (panel A) and 37 months (panel D). However, in 25 month-old and 3 1 -month old rats, the minimally effective concentrations of PDBu were 30 f/M and 3 /iM, respectively, (panels B and C). Between age groups evaluation The effect of age on the concentration-dependence for PDBu-induced facilitation of L-[ 3 H]glutamate uptake in GPV and SYN tissue fractions among the four age groups was examined also. As shown in the PDBu concentration-response curves of Figure 4-4, GPV fractions (panel A) did not exhibit any significant differences among age groups at any of the five PDBu concentrations. However, there were significant (p<0.05) differences between age groups at each of the PDBu concentrations in SYN fractions (panel B) (F 3 16 =6.90, p<0.01). At each PDBu concentration examined, values for uptake at 37 months differed significantly (it) from the 5 month old group (F 3 16 =12.91, 3.46, 7.95, 7.44, and 4.37 for 1, 3, 10, 30, and 100a/M, respectively; p<0.05 in all cases). A change in SYN L-[ 3 H]glutamate uptake leading to a significant difference between 5 and 25 month old animals occurred at 10/iM PDBu only (*) (F 3 ,i 6 =7.95, p<0.01). A significant difference in transport values between 25 and 37 month old animals occurred at 1 (jM (F 3>16 =12.91, p<0.001) and 30 /jM () (F 3 I6 =7.44, p<0.01), while a significant difference between the 3 1 and 37 month old age groups occurred at 1 /uM (F 3 16 =12.91, p<0.001) and 100 (jM PDBu (+) (F 3 16 =4.37, p<0.05).

PAGE 90

84 Preliminary Western Blot Analyses for EAAC1 and GLT-1 Immunoreactivity With the decreasing level of L-[ 3 H]glutamate transport observed in SYN tissue fractions from aged rats (Figure 4-3 and Figure 4-4, panel B), preliminary Western blot analysis for the neuron-specific L-glutamate transporter EAAC1 was undertaken to determine if the amount of EAAC1 protein was decreased in aged animals. When 5 //g of total protein of GPV and SYN fractions obtained from animals aged 5, 25, 3 1, and 37 months were incubated with EAAC1 antibody, GPV fractions exhibited no immunoreactivity above background. By comparison, positive immunostaining in SYN fractions was evident in each age group (n = 1, data not shown). One microgram of total protein from both GPV and SYN tissues from each of the four age groups was screened also against an antiserum for GLT-1 . While GPV fractions from all age groups exhibited positive immunoreactivity for GLT-1, the level appeared less at 5 months when compared to 25, 3 1, or 37 months (n = 1, data not shown). SYN tissue fractions from animals aged 5, 25, 3 1, and 37 months also reacted positively following incubation with primary antibody for GLT1 . Discussion This study is the first to examine the potential effects of aging on L-[ 3 H]glutamate uptake in GPV and SYN tissue fractions isolated from rat forebrain. In these studies aged Fisher 344 x Brown Norway Fl rats were used. Several strains and lines of rats have been used for gerontologic investigations, with Fisher 344 (F344) rats being used most frequently. However, one overriding reason for choosing the F344 x BN Fl instead of the F344 for this study is that the former have a much lower incidence of several major

PAGE 91

85 o o ON • — 2 a E o £ a. w I a 5 25 31 37 Age (in months) Figure 4-1 . Effect of age on L-[ 3 H]glutamate uptake. GPV (panel A) and SYN (panel B) fractions from animals 5, 25, 3 1, and 37 months of age were incubated in the presence of 5//M L-[ 3 H]glutamate for 90 seconds. Values are mean ± SEM from five separate experiments. Asterisks (*) denote a significant difference (p<0.05) between the indicated age group and the 5 month-old group, as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.

PAGE 92

86 Figure 4-2. Within group evaluation of the effect of age on the concentration dependence for PDBu-induced facilitation of L-[ 3 H]glutamate uptake in GPV tissue fractions. GPV fractions from animals aged 5 months (panel A), 25 months (panel B), 3 1 months (panel C), and 37 months (panel D) were preincubated for 5 minutes in the presence of PDBu (\(uM 100//M) or vehicle (0.01% acetone, v/v) then assayed for uptake in the presence of 5//M L-[ 3 H]glutamate. Data points represent averaged values (mean ± SEM) from five separate experiments. Asterisks (*) denote a significant difference (p<0.05) between PDBu-treated samples and vehicle-treated controls as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.

PAGE 93

87 1000 PDBu Cone. (nM) Figure 4-3. Within group evaluation of the effect of age on the concentration dependence for PDBu-induced facilitation of L-[ 3 H]glutamate uptake in SYN tissue fractions. SYN fractions from animals aged 5 months (panel A), 25 months (panel B), 3 1 months (panel C), and 37 months (panel D) were assayed for L-[ 3 H]glutamate uptake as described in Figure 4-2. Asterisks (*) denote a significant difference (p<0.05) between PDBu-treated samples and vehicle-treated controls as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.

PAGE 94

88 o O c u o c (50 £ o £ 1 Gu 3 o i -J 400 300 200 100 A)GPV — 5 Months 3— fl 25 Months A^-A 31 Months A— A 37 Months // 0 1 10 100 1000 800 600 t 400 0 1 10 100 PDBu Cone. (pM) Figure 4-4. Between group evaluation of the effect of age on the concentration dependence for PDBu-induced facilitation of L-[ 3 H]glutamate uptake in GPV and SYN tissue fractions. GPV (panel A) and SYN (panel B) fractions from animals aged 5, 25, 31, and 37 months were assayed for L-[ 3 H]glutamate uptake as described in Figure 4-2. Symbols denote a significant difference (p<0.05) between indicated groups (-£=5 months vs 37 months, *=25 months vs 37 months, +=3 1 months vs 37 months, *=5 months vs 25 months) as determined by repeated measures ANOVA followed by Tukey's multiple pairwise comparison.

PAGE 95

pathologic processes, including glomerulonephritis, retinal atrophy, and leukemia (Maeda et ai, 1985; Lipman et aL, 1996). Additionally, the F344 x BN Fl hybrid attains 50% mortality at 146 weeks of age, which is much greater than the 103 weeks for the F344 rat (Lipman et ai, 1996). As a model in which to study aging, this hybrid strain provides an increased period in which changes associated with an increase in age can be examined in the relative absence of disease. The consistent finding on age-related changes in rodent brain PKC is that translocation is impaired in cortical and hippocampal structures of aged when compared to young and mature rats (Friedman and Wang, 1989; Meyer et aL, 1994; Pisano et ai, 1991; Battaini et ai, 1995). In the cortex this appears not to be related to changes in either phosphorylation or isozyme levels of PKC substrates such as histone or B-50, because levels of these are similar in adult and aged rats (Pisano et ai, 1991; Battaini et ai, 1995). In the hippocampus from aged rats, neither mRNA nor protein levels of PKC alpha, beta, or gamma isoforms appear to be modified (Battaini et aL, 1995). It may be that age-related changes in brain membrane composition (Zidovetzki and Lester, 1 992) , rather than a modification in a particular PKC isoform, are responsible for the impaired PKC translocation. It should be noted, however, that observations showing unmodified PKC activity using histone as substrate are partially in contrast with previous data from the cortex of Fisher 344 rats (Friedman and Wang, 1989) and from the cortex and hippocampus of Sprague-Dawley rats (Battaini et ai, 1990). It is possible that strainrelated differences are responsible for these conflicting data, since strain-related

PAGE 96

90 differences during aging are known to occur in rats in other neurochemical (Peterson and Cotman, 1989) and electrophysiological parameters (Potier et al, 1993). Before determining whether or not age would alter PKC activation and ultimately L-[ 3 H]glutamate uptake, it was necessary to ascertain if L-[ 3 H]glutamate uptake, without drug treatment, was changing with increasing age. Figure 4-1 shows that baseline L[ 3 H]glutamate uptake did not change in the GPV fraction with increasing age (panel A). For SYN fractions (panel B), values for L-[ 3 H]glutamate uptake declined with advancing age. However, it was not until 37 months of age that decreasing levels of L-[ 3 H]glutamate uptake reached significance from 5 month old animals. This finding, of a decrease in SYN L-[ 3 H]glutamate uptake with increased aging, differs from previous reports (Dawson et al, 1989; Palmer et al, 1994). These disparate findings may be attributable to the use in prior studies of crude synaptosomal preparations, almost certainly contaminated with glial cells, which can accumulate L-glutamate. Additionally, many of these prior studies, which indicated no effect of aging on L-glutamate transport, used a broad age range of animals, sometimes only two age groups. We circumvented these issues by using four age groups of animals and tissue fractions isolated by an improved three-step density gradient centrifugation technique (Daniels and Vickroy, 1998). In the present study the ability of PKC activation, via PDBu, to modulate L[ 3 H]glutamate uptake by GPV and SYN fractions was tested in each age group (5, 25, 31, and 37 months) (Figures 4-2 and 4-3). Each of the four age groups examined, for both GPV and SYN, showed increased L-[ 3 H]glutamate uptake in response to increasing

PAGE 97

91 concentrations of PDBu (p<0.001). These findings illustrate that, irrespective of animal age, PKC can respond to activation by PDBu. Differences between age groups in their responsiveness to a particular PDBu concentration in facilitating L-[ 3 H]glutamate uptake were examined also in GPV and SYN tissue fractions obtained from animals aged 5, 25, 3 1, and 37 months (Figure 4-4). Among the six PDBu concentrations used in this study, no differences were evident among the four age groups with regard to PDBu-induced enhancement of L[ 3 H]glutamate uptake by GPV fractions (panel A). However, differences among the SYN fractions from different age groups were noted at all PDBu concentrations (panel B). Notably, the ability of PKC at 37 months to respond to PDBu activation by increasing L[ 3 H]glutamate uptake differed significantly from 5 month-old animals at each PDBu concentration. Taken together, these results show that GPV fractions isolated from animals of increasing age respond to PDBu. GPV fractions responded to PDBu in a concentration-dependent fashion. The four age groups exhibited no differences in their response to any particular PDBu concentration for facilitation of L-[ 3 H]glutamate uptake. However, SYN fractions undergo a significant decrease with advancing age in their ability to respond to PDBu activation. Though SYN fractions can respond to PDBu in a concentration-dependent fashion, this response, like baseline L-[ 3 H]glutamate uptake, declines significantly by 37 months. Therefore, irrespective of the age of animals, PKC can respond to activation by PDBu; however, as baseline L-[ 3 H]glutamate uptake declines with increasing age, so too does the magnitude of the PDBu-induced response. One possible explanation for this may be that all the available sites for phosphorylation on the

PAGE 98

92 SYN transporters already are phosphorylated. If this is the case, PDBu exposure would be incapable of elliciting a response. Western blot analyses were carried out to determine whether protein levels of Lglutamate transporters were changing with increasing age. Preliminary studies of EAAC1, an L-glutamate transporter in neurons (Kanai and Hediger, 1992; Rothstein et al., 1994; Velaz-Faircloth et al., 1996), showed no remarkable changes across the four age groups (n = 1, data not shown). The EAAC1 antibody used in the Western blots was characterized by expression of cDNA for EAAC1 in Human Embryonic Kidney cells (Matthews et al. , 1 997). Specificity was determined by peptide inhibition in the aforementioned cell line as well as tissue (Matthews et al., 1997). From our preliminary studies, it appears that levels of EAAC1 protein do not change across ages and, therefore, cannot account for the significant decrease in L-[ 3 H]glutamate uptake. Preliminary results with GLT-1 (n = 1), an L-glutamate transporter located in glial cells (Danbolt et al., 1992; Rothstein et al., 1994) and some subsets of neurons (Torp et al., 1994; Schmitt et al., 1996), indicated positive immunoreactivity in both GPV and SYN tissue fractions. Surprisingly, immunoreactivity in SYN tissue fractions was greater than GPV tissue fractions at 5, 25, and 37 months (data not shown). Again, these results offer no ready explanation for the observed differences in L-[ 3 H]glutamate uptake by GPV and SYN tissue fractions from animals of different ages. This study has shown that there are differential effects of aging between GPV and SYN tissue fractions. The decline in SYN response to PDBu may result from a lack of available phosphorylation sites for the PDBu-activated PKC or it may be related to an age-

PAGE 99

93 induced impairment of PKC. As discussed previously, the consistent finding on agerelated changes in PKC is impaired translocation. Whether this decline is due to reduced levels of PKC, its translocation, or diminished level of cofactors, such as membrane phospholipids or diacylglycerol, remains to be determined.

PAGE 100

CHAPTER 5 CONCLUSIONS Limitations of the Technique While the first two aims of this project could have been addressed using cultured cells, we choose not to use such techniques, due to the potential for cell culturing to transform, alter, or negate intracellular regulatory mechanisms. Instead, we developed and verified, via morphological, biochemical, and protein analyses, an improved method of cell separation by three-step Percoll density gradient centrifugation. This technique yields functionally viable glialand neuronal-enriched fractions which can be used for direct investigation of potential L-glutamate transporter regulatory mechanisms and determinations of aging effects under identical conditions at the same time. One important advantage in using GPV and SYN is that the diffusional barriers are not as prohibitive as in other preparations such as brain slices. This feature is particularly important in experiments from Chapters 3 and 4, because the site of action of the drugs used was intracellular. Therefore, the drugs needed to permeate the GPV and SYN membranes. Caution should be used, however, when relating GPV and SYN experimental results to situations where a greater degree of anatomical connectivity exists, such as in brain slices or in vivo. 94

PAGE 101

95 It is obvious that the GPV fractions are cleaner (i.e., contain less detectable neuronal elements) than the SYN fractions. This is most likely due to the procedure used to remove the SYN fractions from the Percoll gradient following centrifugation. Unfortunately, at present, resolution of this problem seems difficult and would require extremely creative modifications to circumvent. Additional problems include that a large amount of starting brain tissue (approximately 2.4 g) is necessary to obtain a sufficient, workable volume of GPV tissue (1 .4 ml at approximately 1 mg/mg). This means that two animals must be killed per experiment, and the justification of the need for so many animals is often difficult. Again, due to the necessary quantity of brain tissue initially needed, discrete brain regions, such as the hippocampus, cannot be studied unless many animals are killed for only one experiment. Furthermore, by their nature, GPV fractions limit experimental protocol. This tissue fraction is particularly vulnerable to pH changes, has a very limited lifespan (<30 min), and is exceedingly temperature-sensitive. Obviously, working within the confines of these limitations is possible, but one must keep them in mind whenever establishing a new experimental protocol. Contributions from this Project It is clear that altered excitatory amino acid (EAA) function, due to altered phosphorylation or dephosphorylation, may play a role in age-related phenomena ranging from cell death within discrete brain regions, to the global effects of Alzheimer's disease, to the selective degeneration of upper and lower motor neurons in amyotrophic lateral sclerosis. In terms of basic research, much more needs to be known about the mechanisms responsible for any age-related changes that may occur in the above, and the impact that

PAGE 102

96 those changes will have on EAA transporter function. For example, if the number of transporters phosphorylated in response to increasing extracellular EAA levels declines with age, is this because there are fewer transporters present? Is the conformation of the transporter protein somehow different in the aged brain so that it is less accessible to kinases and/or phosphatases? Are there age-related changes in phosphatases specific for EAA transporters? The regulatory roles of phosphoprotein phosphatases are becoming more apparent, yet there have been few studies of what effects aging may have on protein phosphatase function(s). Regarding clinical application, protein phosphorylation may be potentially useful as a diagnostic tool, as well as a possible site of intervention. Using immunological techniques, developing antibodies specific for the phosphorylated or non-phosphorylated form of a protein is possible (Green and Cotton, 1990; Czernik et ah, 1991), though extensive testings would have to be performed in order to insure specificity. Thereafter, however, developing clinical tests that measure the phosphorylation state of certain key proteins may be possible. Additionally, performing these tests using blood samples may be feasible, since some documented phosphorylation defects also occur in peripheral tissue. For example, fibroblasts from patients with Senile Dementia of the Alzheimer's Type have altered PKC activity (Van Huynh et ah, 1989). Regarding intervention, the most promising strategy may be to alter the kinases and phosphatases responsible for the phosphorylation state of a protein. There are many compounds that either stimulate or block protein kinase and phosphatase activity in vitro (Nishizuka, 1984; Kobayashi et ah, 1989; Cohen et ah, 1990). One problem is that these

PAGE 103

97 compounds may act nonspecifically in a variety of cell types. It is becoming clear that each type of protein kinase has multiple isoforms and that, in many diseases, only one form is altered. Thus, it may be possible to design drugs that are specific only for the affected isoform. In addition, more general compounds, such as antioxidants, may modulate protein kinase activity. For example, the antioxidant 2-mercaptoethanol enhances the PKC activity of old T cells in response to interleukin-2 (Fong and Makinodan, 1989). Therefore, the three-step density gradient centrifugation technique described herein is very useful, not only for answering the many questions concerning the role of phosphorylation / dephosphorylation on EAA transporter function in general, but also in aging for screening compounds that may be potentially useful in correcting defects in transporter function, thereby negating the neurotoxic effects of high extracellular levels of L-glutamate. Future Directions As the project currently stands, phosphorylation / dephosphorylation events, as well as aging, appear to differentially modulate L-glutamate uptake in glia and neurons. This may be attributable to the potentially different functions or roles that glial and neuronal transporters may serve. Some hint of this has come from experiments in which molecular biological methods were used to prevent different transporters from being expressed (Rothstein et al, 1996; Tanaka et ai, 1997). Deleting the glial transporters led to a general rise of extracellular L-glutamate concentration, which caused cell death. By contrast, preventing EAAC1 function did not raise the extracellular glutamate concentration, but caused epileptic-like seizures. These results suggest that glial transporters may function mainly to keep the glutamate concentration low in the

PAGE 104

98 extracellular space, whereas neuronal transporters may play a more specific role in synaptic transmission. It is possible that phosphorylation-mediated changes in EAA transporters may play a role in regulating these functions. Questions remain as to why there are differences in L-glutamate uptake between GPV and SYN tissue fractions. As a first approach to answering this question, the composition and levels of PKC isozymes, as well as PKC translocation, should be determined, as there may be differences in these parameters between the two fractions, and this is affecting (whether increasing or decreasing) the ability of PKC to catalyze phosphorylation of L-glutamate transporters. The presence and levels of various PKC cofactors, such as diacylglycerol, and for some isozymes, calcium, should be determined. Furthermore, the composition of phosphatases within GPV and SYN fractions needs to be ascertained, because L-glutamate transporters may be a targeted substrate for one or more phosphatases. This will be difficult, however, because there are no commercial antibodies available for immunoblotting. Once the aforementioned are determined in young adult animals, aged animals should be screened for the same. Alterations in any of the above may be responsible for the aging changes in L-glutamate uptake seen in very old animals.

PAGE 105

REFERENCES Armbrecht, H.J., Nemani, R.K. and Wongsurawat, N. 1993. Protein phosphorylation: Changes with age and age-related diseases. J. Am. Geriatr. Soc. 41:873-879. Arriza, J.L., Eliasof, S., Kavanaugh, M P. and Amara, S.G. 1997. Excitatory amino acid transporter 5, A retinal glutamate transporter coupled to a chloride conductance. Proc. Natl. Acad. Sci. USA 94:4155-4160. Arriza, J.L., Fairman, W.A., Wadiche, J.I., Murdoch, G.H., Kavanaugh, M P. and Amara, S.G. 1995. An excitatory amino-acid transporter with properties of a ligand-gated chloride channel . Nature 375:5 99-603. Attwell, D., Sarantis, M., Szatkowski, M., Barbour, B. And Brew, H. 1991. Patch-clamp studies of electrogenic glutamate uptake: Ionic dependence, modulation and failure in anoxia. Pages 223-237, in H. Wheal and A. Thomson (Eds.), Excitatory Amino Acids and Synaptic Transmission, Academic Press, London. Balcar, V.J., Johnston, GAR., and Twitchin, B. 1977. Stereospecificity of the inhibition of L-glutamate and L-aspartate high affinity uptake in rat brain slices by threo-3hydroxyaspartate. J. Neurochem. 28:1145-1146. Bartlett, G.R. 1959. Phosphorus assay in column chromatography. J. Biol. Chem. 234: 466-469. Barbour, B., Szatkowski, M., Ingledew, N., and Attwell, D. 1989. Arachidonic acid induces a prolonged inhibition of glutamate uptake into glial cells. Nature 342:918-920. Barford, D. 1991 . Molecular mechanisms for the control of enzymic activity by protein phosphorylation. Biochim. Biophys. Acta 1133:55-62. Barnes, C.A., Mizumori, S.J., Lovinger, D M, Sheu, F.S., Murakami, K, Chan, S.Y., Linden, D.J., Nelson, R.B. and Routtenberg, A. 1988. Selective decline in protein Fl phosphorylation in hippocampus of senescent rats. Neurobiol. Aging 9:393398. 99

PAGE 106

100 Battaini, F., Del-Vesco, R, Govoni, S. and Trabucci, M. 1990. Regulation of phorbol ester binding and protein kinase C activity in aged rat rain. Neurobiol. Aging 11:563-566. Battaini, F., Elkabes, S., Bergamaschi, S., Ladisa, V., Lucchi, L., DeGraan, P.N.E. , Schuurman, T., Wetsel, W.C., Trabucchi, M. and Govoni, S. 1995. Protein kinase C activity, translocation and conventional isoforms in aging rat brain. Neurobiol. Aging 16:137-148. Battaini, F., Govoni, S., Lucchi, L., Ladisa, V., Bergamaschi, S. and Trabucchi, M. 1993. Age-related changes in brain protein kinase C expression, activity, and translocation. Drugs in Dev. 2:275-282. Battaini, F., Lucchi, L., Berganaschi, S., Ladisa, V., Trabucchi, M. and Govoni, S. 1994. Intracellular signalling in the aging bain. The role of protein kinase C and its calcium-dependent isoforms. Ann. NY. Acad. Sci. 719: 271-184. Beal, M.F. 1992a. Mechanisms of excitotoxicity in neurologic sideases. FASEB J. 6:3338-3344. Beal, M.F. 1992b. Does impairment of energy metabolism result in excitotoxic neuronal death in neurodegenerative illnesses? Ann. Neurol. 3 1 : 1 19-130. Bliss, TV. and Collingridge, GL. 1993. A synaptic model of memory: long-term potentiation in the hippocampus. Nature 361 :3 1-39. Bollen, M. and Stalmans, W. 1992. The structure, role, and regulation of type 1 protein phosphatases. Crit. Rev. Biochem. Mol. Biol. 27:227-281. Booth, R.F. and Clark, J.B. 1978. A rapid method for the preparation of relatively pure metabolically competent synaptosomes from rat brain. Biochem. J. 176:365-370. Bouvier, M., Szatkowdki, M., Amato, A. and Attwell, D. 1992. The glial cell glutamate uptake carrier countertransports pH-changing anions. Nature 360:471-474. Bradford, J.F. 1975. Isolated nerve terminals as an in vitro preparation for the study of dynamic aspects of transmitter metabolism and release. Pages 96-1 17, in L.L. Iversen, S.D. Iversen, and S.M. Snyder (Eds ), Handbook of Psychopharmacology, Vol. I, Plenum Press, London.

PAGE 107

101 Bridges, R.J., Stanley, M.S., Anderson, M.W., Cotman, C.W. and Chamberlin, A.R. 1991 . Conformationally defined neurotransmitter analogues. Selective inhibition of glutamate uptake by one pyrrolidine-2,4-dicarboxylate diastereomer. J. Med. Chem. 34:717-725. Brion, L P., Schwartz, J.H., Zavilowitz, B.J., and Schwartz, G.J. 1988. Micro-method for the measurement of carbonic anhydrase activity in cellular homogenates. Analytical Biochem. 175:289-297. Bristol, L A. and Rothstein, J.D. 1996. Glutamate transporter gene expression in amyotrophic lateral sclerosis motor cortex. Ann. Neurol. 39:676-679. Bruns, R.F., Miller, F.D., Merriman, R.L., Howbert, J.J., Heath, W.F., Kobayashi, E., Takahashi, I., Tamaoki, T. And Nakano, H. 1991. Inhibition of protein kinase C by calphostin C is light-dependent. Biochem. Biophys. Res. Commun. 176:288293. Bu, X , Haas, D.W. and Hagedorn, C.H. 1993. Novel phosphorylation sites of eukaryotic initiation factor-4F and evidence that phosphorylation stabilizes interactions of the p25 and p220 subunits. J. Biol. Chem. 268:4975-4978. Cammer, W. 1984. Carbonic anhydrase in oligodendrocytes and myelin in the central nervous system. Ann. N Y. Acad. Sci. 429:494-497. Cammer, W. and Tansey, F.A. 1988a. The astrocyte as a locus of carbonic anhydrase in the brains of normal and desmyelinating mutant mice. J. Comp. Neurol. 275:6567. Cammer, W. and Tansey, F.A. 1988b. Carbonic anhydrase immunostaining in astrocytes in the rat cerebral cortex. J. Neurochem. 50:3 19-322. Campbell, G.LeM. and Shank, R.P. 1978. Glutamic acid uptake by cerebellar granule and glial enriched populations. Brain Res. 153:618-622. Casado, M., Zafra, F., Aragon, C, and Gimenez, C. 1991. Activation of high-affinity uptake of glutamate by phorbol esters in primary glial cell cultures. J. Neurochem. 57:1185-1190. Casado, M., Bendahan, A., Zafra, F., Danbolt, N.C., Aragon, C, Gimenez, C, and Kanner, B.I. 1993. Phosphorylation and modulation of brain glutamate transporters by protein kinase C. J. Biol. Chem. 268:27313-27317.

PAGE 108

102 Castagna, M, Takai, Y., Kaibuchi, K., Sano, K., Kikkawa, U. and Nishizuka, Y. 1982. Direct activation of calcium-activated, phospholipid-dependent protein kinase by tumor-promoting phorbol esters. J. Biol. Chem. 257:7847-7851. Chan, P.H., Kerlan, R. and Fishman, R.A. 1983. Reductions of Y-aminobutyric acid and glutamate uptake and (Na + + K + )-ATPase activity in brain slices and synaptosomes by arachidonic acid. J. Neurochem. 40:309-316. Choi, D.W. 1987. Ionic dependence of glutamate neurotoxicity and cortical cell culture. J. Neurosci. 7:369-379. Choi, D.W. 1988. Glutamate neurotoxicity and diseases of the nervous system. Neuron 1:623-634. Choi, D.W. 1994. Glutamate receptors and the induction of excitotoxic neuronal death. Prog. Brain Res. 100:47-51. Cohen, P. 1989. The structure and regulation of protein phosphatases. Annu. Rev. Biochem. 58:453-508. Cohen, P., Brewis, N.D., Hughes, V. and Mann, D. 1990. Protein serine/threonine phosphatases: An expanding family. FEBS Lett. 268:355-359. Cole, G, Dobkins, K.R., Hansen, L.A, Terry, R.D. and Saitoh, T. 1988. Decreased levels of protein kinase C in Alzheimer brain. Brain Res. 452: 165-174. Coleman, P C. and Flood, D.G 1987. Neuron numbers and dendritic extent in normal aging and Alzheimer's disease. Neurobiol. Aging 8:521-545. Conradt, M. and Stoffel, W. 1997. Inhibition of the high-affinity brain glutamate transporter GLAST-1 via direct phosphorylation. J. Neurochem. 68:1244-1251. Cotman, C.W. 1974. Isolation of synaptosomal and synaptic plasma membrane fractions. Methods Enzymol. 31:445-452. Czernik, AJ., Girault, J.A., Nairn, A.C., Chen, J., Snyder, G., Kebabian, J. and Greengard, P. 1991. Production of phosphorylation state-specific antibodies. Methods Enzymol. 201:264-283. Danbolt, N.C., Pines, G. and Kanner, B.I. 1990. Purification and reconstitution of the sodiumand potassium-coupled glutamate transport glycoprotein from rat brain. Biochemistry 29:6734-6740.

PAGE 109

103 Danbolt, N.C., Storm-Mathisen, J., and Kanner, B.I. 1992. An [Na + + K + ] coupled Lglutamate transporter purified from rat brain is located in glial cell processes. Neuroscience 5 1 :295-3 1 0. Daniels, K.K. and Vickroy, T.W. 1998. Simultaneous isolation of glial and neuronal fractions from rat brain homogenates: Comparison of high-affinity L-glutamate transport properties. Neurochem. Res. 21:1053-1060. Dawson, V.L., Dawson, T.M., Bartley, D.A., Uhl, G.R. and Snyder, S.H. 1993. Mechanisms of nitric oxide mediated neurotoxicity in primary brain cultures. J. Neurosci. 13:2651-2661. Dawson, V.L., Dawson, T.M., London, E.D., Bredt, D.S. and Snyder, S.H. 1991. Nitric oxide mediates glutamate neurotoxicity in primary cortical cultures. Proc. Natl. Acad. Sci. USA 88:6368-6371. Dawson, R., Wallace, D R. and Meldrum, M.J. 1989. Endogenous glutamate release from frontal cortex of adult and aged rats. Neurobiol. Aging 10:665-668. Delaunoy, J.P., Hog, F., DeFeudis, F.V., and Mandel, P. 1979. Estimation of glial contamination of synaptosomal-mitochondrial fractions of rat brain by radioimmunoassay of carbonic anhydrase (CA II). J. Neurochem. 33:61 1-612. De Robertis, E., De Iraldi, P., Rodriquez, G. and Gomez, C.J. 1961. On the isolation of nerve endings and synaptic vesicles. J. Biophys. Biochem. Cytol. 9:229-235. DierksVending, C, Cone, A.L., and Bressman, S.P. 1975. The attachment of glutamine synthetase to brain membranes. Biochem. Med. 13:213-223. Divac, I., Fonnum, F., and Storm-Mathisen, J. 1977. High affinity uptake of glutamate in terminals of corticostriatal axons. Nature 266:377-378. Dokas, L.A., Pisano, M R, Schrama, L.H., Zwiers, H. and Gispen, W.H. 1990. Dephosphorylation of B-50 in synaptic plasma membranes. Brain Res. Bull. 24:321-329. Dodd, P R, Hardy, J.H., Oakley, A.E., Edwardson, J.A., Perry, E.K., and Delaunoy, J.P. 1981. A rapid method for preparing synaptosomes: comparison with alternative procedures. Brain Res. 226: 107-1 1 8.

PAGE 110

104 Dowd, LA, Coyle, A.J., Rothstein, J.D., Pritchett, D.B., and Robinson, M B. 1996. Comparison of Na + -dependent glutamate transport activity in synaptosomes, C6 glioma, and Xenopus oocytes expressing excitatory amino acid carrier 1 (EAAC1). Molec. Pharmacol. 49:465-473. Dowd, LA. and Robinson, M B. 1996. Rapid stimulation of EAAC1 -mediated Na + dependent L-glutamate transport activity in C6 glioma cells by phorbol ester. J. Neurochem. 67:508-516. Dunkley, P R, Jarvie, P.E., Heath, J.W., Kidd, G.J., and Rostas, J A P. 1986. A rapid method for isolation of synaptosomes on percoll gradients. Brain Res. 372: 115129. Eng, L.F. 1985. Glial fibrillary acidic protein (GFAP): The major protein of glial intermediate filaments in differentiated astrocytes. J. Neuroimmunol. 8:203-214. Eng, L.F., Vanderhaegen, J.J., Bignami, A., and Gersh, B. 1971. An acidic protein isolated from fibrous astrocytes. Brain Res. 28:351-354. Epand, R.M. and Lester, D.S. 1990. The role of membrane biophysical properties in the regulation of protein kinase C activity. Trends Pharmacol. Sci. 1 1 :3 17-320. Fagg, G.E. and Foster, A C. 1983. Amino acid neurotransmitters and their pathways in the mammalian central nervous system. Neurosci. 9:701-719. Fairman, W.A, Vandenberg, R.J., Arriza, J.L., Kavanaugh, M P., and Amara, S.G. 1995. An excitatory amino-acid transporter with properties of a ligand-gated chloride channel. Nature 375:599-603. Faivre-Bauman, A., Rossier, J., and Benda, P. 1974. Glutamate accumulation by a clone of glial cells. Brain Res. 76:371-375. Ferkany, J. and Coyle, J.T. 1986. Heterogeneity of sodium-dependent excitatory amino acid uptake mechanisms in rat brain. J. Neurosci. Res. 16:491-503. Fong, T.C. and Makinodan, T. 1989. Preferential enhancement by 2-mercaptoethanol of IL-2 responsiveness of T blast cells from old over young mice is associated with potentiated protein kinase C translocation. Immunol. Lett. 20:149-154. Fonnum, F. 1984. Glutamate: A neurotransmitter in mammalian brain. J. Neurochem. 42:1-11.

PAGE 111

105 Frandsen, A. and Schousboe, A. 1990. Development of excitatory amino acid induced cytotoxicity in cultured neurons. Intl. J. Devi. Neurosci. 8:209-216. Friedman, E. and Wang, H.-Y. 1989. Effect of age on brain cortical protein kinase C and its mediation of 5-hydroxytryptamine release. J. Neurochem. 52:187-192. Fujiki, H., Suganuma, M, Yoshizawa, S., Kanazawa, H., Sugimura, T., Manam, S., Kahn, S.M., Jiang, W., Hoshina, S. and Weinstein, I B. 1989. Codon 61 mutations in the c-Harvey-ras gene in mouse skin tumors induced by 7,12dimethylbenz[a]anthracene plus okadaic acid class tumor promoters. Mol. Carcinog. 2:184-187. Fujiki, H., Suganuma, M, Yoshizawa, S., Nishiwaki, S., Winyar, B. and Sugimura, T. 1991 . Mechanisms of action of okadaic acid class tumor pormoters on mouse skin. Environ. Health Perspect. 93:211-214. Fykse, E M. and Fonnum, F. 1988. Uptake of gamma-aminobutyric acid by a synaptic vesicle fraction isolated from rat brain. J. Neurochem. 50:1237-1242. Fykse, E M. and Fonnum, F. 1991 . Transport of gamma-aminobutyrate and L-glutamate into synaptic vesicles. Effect of different inhibitors on the vesicular uptake of neurotransmitters and on the Mg2(+)-ATPase. Biochem. J. 276:363-367. Galanopoulos, E., Lellos, V., Papadakis, M., Philippidis, H., and Palaiologos, G. 1988. Effects of fasting and diabetes on some enzymes and transport of glutamate in cortex slices or synaptosomes from rat brain. Neurochem. Res. 13:243-248. Ghandour, M.C., Langley, O.K., Vincendon, G, and Gombos, G. 1979. Double labeling immunohistochemical technique provides evidence of the specificity of glial cell markers. J. Histochem. Cytochem. 27:1634-1637. Goldman, J.E., Schaumburg, H.H., and Norton, W.T. 1978. Isolation and characterization of glial filaments from human brain. J. Cell. Biol. 78:426-440. Gong, C.X., Singh, T.J., Grundke-Iqbal, I. and Iqbal, K. 1993. Phosphoprotein phosphatase activities in Alzheimer disease brain. J. Neurochem. 61 :92 1-927. Govoni, S., Rius, R.A., Battaini, F. and Trabucchi, M. 1988. Reduced cAMP-dependent phosphorylation in striatum and nucleus accumbens of aged rats: Evidence of an altered functioning of Dl dopaminoceptive neurons. J. Gerontol. 43:B93-97.

PAGE 112

106 Gray, E.G. and Whittaker, V P. 1962. The isolation of nerve endings from brain: an electron microscopic study of cell fragments derived by homogenization and centrifugation. J. Anat. 96:79-88. Green, A.K. and Cotton, R.G. 1990. Experimental determination of the phosphorylation state of phenylalanine hydroxylase. Biochem. J. 265:563-568. Haga, T. and Noda, H. 1973. Choline uptake systems of rat brain synaptosomes. Biochim. Biophys. Acta (Amst.) 291:564-575. Hardie, D.G. 1990. Roles of protein kinases and phosphatases in signal transduction. Symp. Soc. Exp. Biol. 44:241-255. Hardie, D.G., Haystead, T.A. and Sim, A.T. 1991. Use of okadaic acid to inhibit protein phosphatases in intact cells. Methods Enzymol. 201 :469-476. Haystead, T.A, Sim, A T., Carling, D., Honnor, R.C., Tsukitani, Y., Cohen, P. and Hardie, D.G. 1989. Effects of the tumour promoter okadaic acid on intracellular protein phosphorylation and metabolism. Nature 337:78-81. Henn, FA, Goldstein, M.N., and Hamberger, A. 1974. Uptake of the neurotransmitter candidate glutamate by glia. Nature 249:663-664. Herrero, I., Miras-Portugal, M.T. and Sanchez-Prieto, J. 1992. Positive feedback of glutamate exocytosis by metabotropic presynaptic receptor stimulation. Nature 360:163-166. Hertz, L., Schousboe, A, Boechler, N, Mukerji, S., and Fedoroff, S. 1978. Kinetic characteristic of the glutamate uptake into normal astrocytes in cultures. Neurochem. Res. 3:1-14. Iida, T., Kobayashi, E., Yochida, M. and Sano, H. 1989. Calphostins, novel and specific inhibitors of protein kinase C. II. Chemical structures. J. Antibiot. Tokyo 42:1475-1481. Ingebritsen, T.S., Stewart, A. A. and Cohen. P. 1983. The protein phosphatases involved in cellular regulation. 6. Measurement of assessment of their physiological roles. Eur. J. Biochem. 132:297-307. Iqbal, K, Alonso, A., Gong, C, Khatoon, S., Kudo, T., Singh, T. and Grundke-Iqbal, I. 1993. Molecular pathology of Alzheimer neurofibrillary degeneration. Acta Neurobiol. Exp. Warsz. 53:325-335.

PAGE 113

107 Ishihara, H., Martin, B.L., Brautigan, D.L., Karaki, H., Ozaki, R, Kato, Y., Fusetani, N, Watabe, S , Hashimoto, K. and Uemura, D. 1989. Calyculin A and okadaic acid: Inhibitors of protein phosphatase activity. Biochem. Biophys. Res. Commun. 159:871-877. Ito, A., Saito, N, and Hirata, M. 1990. Immunocytochemical localization of the a subspecies of protein kinase C in rat brain. Proc. Natl. Acad. Sci. USA 87:31953199. Iversen, L.L. 1973. Catecholamine uptake processes. Br. Med. Bull. 29:130-135. Kanai, Y. and Hediger, M A. 1992. Primary structure and functional characterization of a high-affinity glutamate transporter. Nature 360:467-471 . Kanner, B.I. and Schuldiner, S. 1987. Mechanism of transport and storage of neurotransmitters. CRC Crit. Rev. Biochem. 22:1-38. Kato, K., Ishiguro, Y , Suzuki, F., Ito, A., and Siemba, R. 1982. Distribution of nervous system-specific forms of enolase in peripheral tissues. Brain Res. 237:441-448. Kato, T., Hara, A., Nakayama, T., Sawada, R, Hamatake, M. And Matsumoto, Y. 1986. Purification and characterization of purple acid phosphatase from rat bone. Comp. Biochem. Physiol. B. 83:813-817. Khatoon, S., Grundke-Iqbal, I. and Iqbal, K. 1992. Brain levels of microtubuleassociated protein tau are elevated in Alzheimer's disease: a radioimmuno-slot-blot assay for nanograms of the protein. J. Neurochem. 59:750-753. Kiley, SC. and Jaken, S. 1987. Purification and characterization of three types of protein kinase C from rabbit brain cytosol. Proc. Natl. Acad. Sci. USA 84:4418-4422. Kiley, S C. and Jaken, S. 1990. Activation of alpha-protein kinase C leads to association with detergent-insoluble components of GH4C1 cells. Mol. Endocrinol. 4:59-68. Kiley, S.C., Olivier, A.R., Gordge, PC, Ryves, W.J., Evans, F.J., Ways, D.K. and Parker, P.J. 1994. 12-Deoxyphorbol-13-O-phenylacetate-20-acetate is not protein kinase C-beta isozyme-selective in vivo. Carcinogenesis 15:31 9-324. Klee, C.B. 1988. Calcineurin. Adv. Enzymol. Relat. Areas Mol. Biol. 61:149-200. Kobayashi, E , Nakano, H., Morimoto, M. and Tamaoki, T. 1989. Calphostin C (UCN1028C), a novel microbial compound, is a highly potent and specific inhibitor of protein kinase C. Biochem. Biophys. Res. Comm. 159:548-553.

PAGE 114

108 Kose, A., Ito, A., Saito, N. and Tanaka, C. 1990. Electron microscopic localization of yand pil-subspecies of protein kinase C in rat hippocampus. Brain Res. 518:209217. Kose, A., Saito, N., and Ito, H. 1988. Electron microscopic localization of type I protein kinase C in rat Purkinje cells. J. Neurosci. 8:4262-4268. Krnjevic, K. 1970. Glutamate and gamma-aminobutyric acid in brain. Nature 228:1 19124. Kuhar, M.J. 1973. Neurotransmitter uptake: A tool in identifying neurotransmitterspecific pathways. Life Sci. 13:1623-1634. Kuhar, M.J. and Murrin, L.C. 1978. Sodium-dependent high affinity choline uptake. J. Neurochem. 30:15-21. Kuhar, M.J., Sethy, V.H., Roth, R.H., and Aghajanian, G.K. 1973. Choline: Selective accumulation by central cholinergic neurons. J. Neurochem. 20:581-593. Lehre, K.P., Levy, L.M., Ottersen, O.P., Storm-Mathisen, J., and Danbolt, N.C. 1995. Differential expression of two glial glutamate transporters in the rat brain: Quantitative and immunocytochemical observations. J. Neurosci. 15:1835-1853. Leigh, P.N. and Meldrurn, B.S. 1996. Excitotoxicity in ALS. Neurology 47 (Suppl. 4):S221-S227. Lipman, R.D., Chrisp, C.E., Hazzard, D.G. and Branson, RT. 1996. Pathologic characterization of brown Norway, brown Norway x Fischer 344, and Fischer 344 x brown Norway rats with relation to age. J. Gerontol. A Biol. Sci. Med. Sci. 5LB54-B59. Logan, W.J. and Snyder, S.H. 1972. High affinity uptake systems for glycine, glutamic and aspartic acids in synaptosomes of rat central nervous tissues. Brain Res. 42:413-431. Lowry, O.W., Rosenbrough, N.J., Farr, A.L., and Randal, R.J. 1951. Protein measurement with folin phenol reagent. J. Biol. Chem. 193:265-275. Lund, P. 1970. A radiochemical assay for glutamine synthetase and activity of the enzyme in rat tissues. Biochem. J. 118:35-39.

PAGE 115

109 Maeda, H, Gleiser, C.A., Masoro, E.J., Murata, I., McMahan, C.A. and Yu, B P. 1985. Nutritional influences on aging of Fischer 344 rats. II. Pathology. J.Gerontol. 40:671-688. Magnoni, M.S., Govoni, S., Battaini, F. and Trabucchi, M. 1991. The aging brain: Protein phosphorylation as a target of changes in neuronal function. Life Sci. 48:373-385. Marangos, P.J., Zomzely-Neurath, C, Luk, D.C.M., and York, C. 1975. Isolation and characterization of the nervous system-specific protein 14-3-2 from rat brain. Purification, subunit composition, and comparison to the beef brain protein. J. Biol. Chem. 250:1884-1891. Martinez-Hernandez, A., Bell, K.P., and Norenberg, M.D. 1977. Glutamine synthetase. Glial localization in brain. Science 195: 1356-1385. Masliah, E., Cole, G., Shimohama, S., Hansen, L., DeTeresa, R, Terry, R.D. and Saitoh, T. 1990. Differential involvement of protein kinase C isozymes in Alzheimer's disease. J. Neurosci. 10:2113-2124. Matthews, J.C., Beveridge, M.J., Malandro, M.S., Rothstein, J.D., Campbell-Thompson, M., Verlander, J.W., Kilberg, M.S. and Novak, D A. 1997. Activity and protein localization of multiple glutamate transporters in gestation Day 14 versus Day 20 rat placenta. Am. J. Physiol. X:XX. McBean , G.J. and Roberts, P.J. 1985. Neurotoxicity of L-glutamate and DL-threo-3hydroxyaspartate in the rat striatum. J. Neurochem. 44:247-254. Meldrum, M.J., Glenton, P. and Dawson, R. 1992. [ 3 H]D-aspartic acid release in brain slices of adult and aged Fischer 344 rats. Neurochem. Res. 17: 1 5 1-1 56. Meyer, M., Judkins, J.H., Momol, A.E. and Hardwik, E.O. 1994. Effects of peroxidation and aging on rat neocortical Ach release and protein kinase C. Neurobiol. Aging 15:63-67. Morgan, I. G. 1976. Synaptosomes and cell separation. Neuroscience 1:159-165. Najlerahim, A., Francis, P.T. and Bowen, D M. 1990. Age-related alteration in excitatory amino acid neurotransmission in rat brain. Neurobiol. Aging 1 1 : 1 55-1 58. Nakamura, Y, Iga, K., Shibata, T.L., Shudo, M., and Kataoka, K. 1993. Glial plasmalemmal vesicles: A subcellular fraction from rat hippocampal homogenate distinct from synaptosomes. Glia 9:48-56.

PAGE 116

110 Nakanishi, J. and Exton, J.H 1992. Purification and characterization of the C-isoform of protein kinase C from bovine kidney. J. Biol. Chem. 267:16347-16354. Narang, N. and Crews, F T. 1995. Age does not alter protein kinase C isozymes mRNA expression in rat brain. Neurochem. Res. 20: 1 1 19-1 126. Nestler, E.J. and Greengard, P. 1984. Neuron-specific phosphoproteins in mammalian brain. Adv. Cyclic Nucleotide Protein Phosphorylation Res. 17:483-488. Nicholls, D. And Attwell, D. 1990. The release and uptake of excitatory amino acids. Trends Pharmacol. Sci. 11:462-468. Nishizuka, Y. 1984. The role of protein kinase C in cell surface signal transduction and tumor promotion. Nature 308:693-695. Nishizuka, Y. 1992. Intracellular signaling by hydrolysis of phospholipids and activation of protein kinase C. Science 258:607-614. Norenberg, M.D. 1983. Immunohistochemistry of glutamine synthetase. Pages 95-1 11, in L. Hertz, E. Kvamme, E. McGeer, and A. Schousboe (Eds ), Glutamine, Glutamate, and GAB A in the Central Nervous System Alan R. Liss, Inc., New York. Norenberg, M.D. and Martinez-Hernandez, A. 1979. Fine structural localization of glutamine synthetase in astrocytes of rat brain. Brain Res. 161 :303-3 10. Ono, Y , Fujii, T. and Ogita, K. 1988. The structure, expression, and properties of additional members of the protein kinase C family. J. Biol. Chem. 263:6927-6932. Ono, M., Akiyama, K., Tsutsui, K. and Kuroda, S. 1994. Differential changes in the activities of multiple protein kinase C subspecies in the hippocampal-kindled rat. Brain Res. 660:27-33. Palaiologos, G, Galanopoulos, E., Lellos, V., Papadakis, M., and Philippidis, H. 1985. Glutamate metabolizing enzymes and its transport in brain cortex from normal and diabetic rats. 13th International Congress Biochem. Abs. 287:98. Palmer, A.M., Robichaud, P.J. and Reiter, C.T. 1994. The release and uptake of excitatory amino acids in rat brain: Effect of aging and oxidative stress. Neurobiol. Aging 15:103-111.

PAGE 117

Ill Parfitt, K.D., Hoffer, B.J. and Browning, M.D. 1991. Norepinephrine and isoproterenol increase the phosphorylation of synapsin I and synapsin II in dentate slices of young but not aged Fisher 344 rats. Proc. Natl. Acad. Sci. USA 88:2361-2365. Pascale, A, Fortino, I., Govoni, S., Trabucchi, M, Wetsel, W. and Battaini, F. 1996. Functional impairment in protein kinase C by RACK1 (receptor for activated C kinase 1) deficiency in aged rat brain cortex. J. Neurochem. 67:2471-2477. Peterson, C. and Cotman, C.W. 1989. Strain-dependent decrease in glutamate binding to the N-methyl-aspartic acid receptor during aging. Neurosci. Lett. 104:309-313. Pines, G, Danbolt, N.C., Bjoras, M, Zhang, Y., Bendahan, A, Eide, L., Koepsell, H., Storm-Mathisen, J., Seeberg, E., and Kanner, B.I. 1992. Cloning and expression of a rat brain L-glutamate transporter. Nature 360:464-467. Pisano, M R, Wang, H.Y. and Friedman, E. 1991 . Protein kinase activity changes in the aging brain. Biomed. Environ. Sci. 4:173-181. Plaitakis, A. 1990. Glutamate dysfunction and selective motor neuron degeneration in amyotrophic lateral sclerosis: A hypothesis. Ann. Neurol. 28:3-8. Pogun, S., Dawson, V., and Kuhar, M.J. 1994. Nitric oxide inhibits 3 H-glutamate transport in synaptosomes. Synapse 18:21-26. Pogun, S. and Kuhar, M.J. 1993. Glutamic acid uptake inhibition by nitric oxide. Soc. Neurosci. Abstr. 19:1351. Potier, B., Lamour, Y. and Dutar, P. 1993. Age-related alterations in the properties of hippocampal pyramidal neurons among rat strains. Neurobiol. Aging 14:17-25. Price, M.T., Olney, J.W., and Haft, R. 1981 . Age-related changes in glutamate concentration and synaptosomal glutamate uptake in adult rat striatum. Life Sci. 28:1365-1370. Ramaharobandro, N, Borg, J., Mandel, P., and Mark, J. 1982. Glutamine and glutamate transport in cultured neuronal and glial cells. Brain Res. 244: 113-121. Rauen, T., Jeserich, G., Danbolt, N.C., and Kanner, B.I. 1992. Comparative analysis of sodium-dependent L-glutamate transport of synaptosomal and astroglial membrane vesicles from mouse cortex. FEBS 312:1 5-20.

PAGE 118

112 Rhoads, D.E., Ockner, R.K., Peterson, N.A. and Raghupathy, E. 1983. Modulation of membrane transport by free fatty acids: Inhibition of synaptosomal sodiumdependent amino acid uptake. Biochemistry 22:1965-1970. Robinson, M B. and Coyle, J.T. 1987. Glutamate and related acidic excitatory neurotransmitters: From basic science to clinical application. FASEB J. 1 :446455. RodriguezVico, F.L., Martinez-Cayuela, M , Zafra, M.F., Garcia-Peregrin, E., and Ramirez, H. 1991. A procedure for the simultaneous determination of lipid and protein in biomembranes and other biological samples. Lipids 26:77-80. Roivainen, R. and Messing, R.O. 1993. The phorbol derivatives thymeleatoxin and 12deoxyphorbol-13-O-phenylacetate-10-acetate cause translocation and downregulation of multiple protein kinase C isozymes. FEBS Lett. 3 19:3 1-34. Rosenburg, P. A., Amin, S. and Leitner, M. 1992. Glutamate uptake disguises neurotoxic potency of glutamate agonists in cerebral cortex in dissociated cell culture. J. Neurosci. 12:56-61. Rothstein, J.D., Dykes-Hoberg, M., Pardo, C.A, Bristol, L.A., Jin, L., Kuncl, R.W., Kanai, Y., Hediger, MA, Wang, Y , Schielki, J.P. and Welty, D.F. 1996. Knockout of glutamate transporters reveals a major role for astroglial transport in excitotoxicity and clearance of glutamate. Neuron 16:675-686. Rothstein, J.D., Martin, L.J. and Kuncl, R.W. 1992. Decreased glutamate transport by the brain and spinal cord in amyotrophic lateral sclerosis. New Engl. J. Med. 326:1464-1468. Rothstein, J.D., Martin, L., Levey, A.I., Dykes-Hoberg, M., Jin, L., Wu, D., Nash, N, and Kuncl, R. W. 1994. Localization of neuronal and glial glutamate transporters. Neuron 13:713-725. Rothstein, J.D., Van Kammen, M., Levey, A.I., Martin, L.J., and Kuncl, R.W. 1995. Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann. Neurol. 38:73-84. Ryves, W.J., Evans, AT., Olivier, A.R., Parker, P.J. and Evans, F.J. 1991 . Activation of the PKC-isotypes alpha, beta 1, gamma, delta and epsilon by phorbol esters of different biological activities. FEBS Lett. 288:5-9. Saito, N., Kikkawa, U., Nishizuka, Y. and Tanaka, C. 1988. Distribution of protein kinase C-like immunoreactive neurons in rat brain. J. Neurosci. 8:369-382.

PAGE 119

113 Saito, N, Itouji, A. and Totani, Y. 1993. Cellular and intracellular localization of esubspecies of protein kinase C in the rat brain: Presynaptic localization of esubspecies. Brain Res. 6099:241-248. Schmitt, A., Asan, E., Puschel, B., Jons, T., and Kugler, P. 1996. Expression of the glutamate transporter GLT1 in neural cells of the rat central nervous system: Nonradioactive in situ hybridization and comparative immunocytochemistry. Neurosci. 71:989-1004. Schmitt, A., Asan, E., Puschel, B. and Kugler, P. 1997. Cellular and regional distribution of the glutamate transporter GLAST in the CNS of rats: Nonradioactive in situ hybridization and comparative immunocytochemistry. J. Neurosci. 17:1-10. Schousboe, A., Svenneby, G, and Hertz, L. 1977. Uptake and metabolism of glutamate in astrocytes cultured from dissociated mouse brain hemispheres. J. Neurochem. 29:999-1005. Schrier, B.K. and Thompson, E.J. 1974. On the role of glial cells in the mammalian nervous system. Uptake, excretion, and metabolism of putative neurotransmitters by cultured glial tumor cells. J. Biol. Chem. 249: 1769-1780. Shaw, P. J., Chinnery, R.M. and Ince, P.G 1994. [ 3 H]D-aspartate binding sites in the normal human spinal cord and changes in motor neuron disease: A quantitative autoradiographic study. Brain Res. 655:195-201. Shenolikar, S. 1994. Protein serine/threonine phosphatases~new avenues for cell regulation. Annu. Rev. Cell Biol. 10:55-86. Shenolikar, S. and Nairn, A.C. 1991. Protein phosphatases: Recent progress. Adv. Second Messenger Phosphoprotein Res. 23: 1-121. Shearman, M.S., Shinomura, T., Oda, T. and Nishizuka, Y. 1991. Dynamic changes in the hippocampus and cerebellar cortex concomitant with synaptogenesis. J. Neurochem. 56:1255-1262. Shields, S.M., Ingebritsen, T.S. and Kelly, P.T. 1985. Identification of protein phosphatase 1 in synaptic junctions: Dephosphorylation of endogenous calmodulin-dependent kinase II and synapse-enriched phosphoproteins. J. Neurosci. 5:3414-3422. Sim, A.T. 1991 . The regulation and function of protein phosphatases in the brain. Mol. Neurobiol. 5:229-246.

PAGE 120

114 Stabel, S. and Parker, P.J. 1991. Protein kinase C. Pharmacol. Ther. 51:71-95. Stewart, R.M., Martuza, R.L., Baldessarini, R.J., and Kornblith, P L. 1976. Glutamate accumulation by human gliomas and meningiomas in tissue culture. Brain Res. 118:441-452. Storck, T., Schulte, S., Hoffman, K. and Stoffel, W. 1992. Structure, expression and functional analysis of a Na"-dependent glutamate/aspartate transporter from rat brain. Proc. Natl. Acad. Sci. USA 89:10955-10959. Strong, R., Samorajski, T., and Gottesfeld, A. 1984. High-affinity uptake of neurotransmitters in rat neostriatum: Effects of aging. J. Neurochem. 43: 17661768. Suganuma, M., Fujiki, H, Furuya-Suguri, H., Yoshizawa, S., Yasumoto, S., Kato, Y., Fusetani, N. and Sugimura, T. 1990. Calyculin A an inhibitor of protein phosphatases, a potent tumor promoter on CD-I mouse skin. Cancer Res. 50:3521-3525. Suganuma, M, Fujiki, H., Okabe, S., Nishiwaki, S., Brautigan, D., Ingebritsen, T.S. and Tosner, M R. 1992. Structurally different members of the odadaic acid class selectively inhibit protein serine/threonine but not tyrosine phosphatase activity. Toxicon. 30:873-878. Suganuma, M., Fujiki, H, Suguri, H, Yoshizawa, S., Hirota, M., Nakayasu, M, Ojika, M , Wakamatsu, K , Yamada, K. and Sugimura, T. 1988. Okadaic acid: An additional non-phorbol1 2-tetradecanoate1 3 -acetate-type tumor promoter. Proc. Natl. Acad. Sci. USA 85:1768-1771. Suganuma, M, Suttajit, M., Suguri, H., Ojika, M, Yamada, K. and Fujiki, H. 1989. Specific binding of okadaic acid, A new tumor promoter in mouse skin. FEBS Lett. 250:615-618. Swaab, D.F., Grundke-Iqbal, I., Iqbal, K., Kremer, H P., Ravid, R. and Van de Nes, J.A 1992. Tau and ubiquitin in the human hypothalamus in aging and Alzheimer's disease. Brain Res. 590:239-249. Szatkowski, M. and Attwell, D. 1994. Triggering and execution of neuronal death in brain ischaemia: Two phases of glutamate release by differenct mechanisms. Trends Neurosci. 17:359-365.

PAGE 121

115 Takai, A., Bialojan, C, Troschka, M. and Ruegg, J.C. 1987. Smooth muscle myosin phosphatase inhibition and force enhancement by black sponge toxin. FEBS Lett. 217:81-84. Tamaoki, T., Takahashi, I., Kobayashi, E., Nakano, H., Akinaga, S. and Suzuki, K. 1990. Calphostin (UCN1028) and calphostin related compounds, A new class of specific and potent inhibitors of protein kinase C. Adv. Second Messenger Phosphoprotein Res. 24:497-501. Tanaka, C. and Saito, N. 1992. Localization of subspecies of protein kinase C in the mammalizan central nervous system. Neurochem. Int. 21:499-512. Tanaka, K., Watase, K., Manabe, T., Yamada, K., Watanabe, M., Takahashi, K., Iwama, H., Nishikawa, T., Ichihara, N, Kikuchi, T., Okuyama, S., Kawashima, N., Hon, S., Takimoto, M. and Wada, K. 1997. Epilepsy and exacerbation of brain injury in mice lacking the glutamate transporter GLT1 . Science 276 : 1 6991 702. Torp, R, Danbolt, N.C., Babaie, E., Bjoras, M., Seeberg, E., Storm-Mathisen, J. and Ottersen, OP. 1994. Differential expression of two glial glutamate transporters in the rat brain: An in situ hybridization study. Eur. J. Neurosci. 6:936-942. Trachtenberg, M.C. and Sapirstein, V S. 1980. Carbonic anhydrase distributions in central and peripheral nervous system of the rat. Neurochem. Res. 5:573-581. Trotti, D., Rossi, D., Gjesdal, O., Levy, L.M., Racagni, G., Danbolt, N.C., and Volterra, A. 1996. Peroxynitrite inhibits glutamate transporter subtypes. J. Biol. Chem. 271:5976-5979. Turner, B.C., Tonks, N.K., Rapp, U.R and Reed, J.C. 1993. Interleukin 2 regulates Raf1 kinase activity through a tyrosine phosphorylation-dependent mechanism in a Tcell line. Proc. Natl. Acad. Sci. USA 90:5544-5548. Undie, AS., Wang, H.-Y. and Friedman, E. 1995. Decreased phospholipase C-P immunoreactivity, phosphoinositide metabolism and protein kinase C activation in senescent F-344 rat brain. Neurobiol. Aging 16:19-28. Van Huynh, T., Cole, G., Katzman, R, Huang, K.P. and Saitoh, T. 1989. Reduced protein kinase C immunoreactivity and altered protein phosphorylation in Alzheimer' s disease fibroblasts. Arch. Neurol. 46. 1 1 951 1 99. Velaz-Faircloth, M, McGraw, T.S., Alandro, M.S., Fremeau, R.T., Kilberg, M.S. and Anderson, K.J. 1996. Characterization and distribution of the neuronal glutamate transporter EAAC1 in rat brain. Am. J. Physiol. 270:C67-75.

PAGE 122

116 Weiler, C.T., Nystrom, B., and Hamberger, A. 1979. Characteristics of glutamine vs glutamate transport in isolated glia and synaptosomes. J. Neurochem. 32:559-565. Wheeler, D.D. 1979. A model of high affinity glutamic acid transport by rat cortical synaptosomes: A refinement of the originally proposed model. J. Neurochem. 33:883-894. Wheeler, D.D. and Ondo, J.G. 1986. Time course of the aging of the high affinity Lglutamate transporter in rat cortical synaptosomes. Exp. Gerontol. 21:159-168. Whittaker, V. 1972. The synaptosome. Pages 327-364, in S. Lajtha (Ed ), Handbook of Neurochemistry, Vol. 2, Plenum Press, New York. Yamamura, H.I. and Snyder, S.H. 1972. Choline high-affinity uptake by rat brain synaptosomes. Science 178:626-628. Yamamura, H.I. and Snyder, S.H. 1973. High affinity transport of choline into synaptosomes of rat brain. J. Neurochem. 21 : 1355-1374. Yoshihara, C, Saito, N, Taniyama, K. and Tanaka, C. 1991. Differential localization of four subspecies of protein kinase C in the rat striatum and substantia nigra. J. Neurosci. 11:690-700. Young, S., Parker, P.J. Ullrich, A. and Stabel, S. 1987. Down-regulation of protein kinase C is due to an increased rate of degradation. Biochem. J. 244:775-779. Young, W.S. 1988. Expression of three (and a putative four) protein kinase C genes in brains of rat and rabbit. J. Chem. Neuroanat. 1 : 177-194. Zafra, F., Alcantara, R., Gomeza, J., Aragon, C. and Gimenez, C. 1990. Arachidonic acid inhibits glycine transport in cultured glial cells. Biochem. J. 271 : 23 7-242. Zerangue, N., Arriza, J.L., Amara, S.G. and Kavanaugh, M P. 1995. Differential modulation of human glutamate transporter subtypes by arachidonic acid. J. Biol. Chem. 270:6433-6435. Zidovetzki, R. and Lester, D.S. 1992. The mechanism of activation of protein kinase C: a biophysical perspective. Biochim. Biophys. Acta 1 134:261-272.

PAGE 123

BIOGRAPHICAL SKETCH My name is Kellye K. Daniels. I was born March 21, 1967, in Valdosta, Georgia. My hometown is Jasper, Florida. I received my B.S. degree in biology from the Florida State University in 1990. After completing a Master of Science degree in zoology from the University of South Florida, I began my doctoral work in the Department of Neuroscience at the University of Florida in 1992. 117

PAGE 124

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. Thomas W. Vickroy, Chair Associate Professor of Neuros 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^atyef quality, as a dissertation for the degree of Doctor of Philosophy. levin J. Anc Associate 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. 4 Gerard P. J. Shaw 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. es W. Simpkins rofessor of Pharmacodynamics 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. Don Walker Professor of Neuroscience

PAGE 125

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. December, 1997 )ean, College of Medicine Dean, Graduate School