The Role of prostaglandin synthase II (PGS-2) in the immunopathogenesis of NOD diabetes

MISSING IMAGE

Material Information

Title:
The Role of prostaglandin synthase II (PGS-2) in the immunopathogenesis of NOD diabetes
Physical Description:
vi, 127 leaves : ill. ; 29 cm.
Language:
English
Creator:
Xie, Xianqing Tony
Publication Date:

Subjects

Subjects / Keywords:
Prostaglandins -- pharmacology   ( mesh )
Prostaglandin-Endoperoxide Synthase -- pharmacology   ( mesh )
Mice, Inbred NOD   ( mesh )
Mice, SCID   ( mesh )
Diabetes Mellitus, Type I -- immunology   ( mesh )
Diabetes Mellitus, Type I -- pathology   ( mesh )
Gene Expression   ( mesh )
Gene Expression Regulation   ( mesh )
Macrophages   ( mesh )
Sex Characteristics   ( mesh )
Dinoprostone -- biosynthesis   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 107-126).
Statement of Responsibility:
by Xianqing Tony Xie.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 49015965
ocm49015965
System ID:
AA00011188:00001

Full Text















THE ROLE OF PROSTAGLANDIN SYNTHASE II (PGS-2) IN
THE IMMUNOPATHOGENESIS OF NOD DIABETES













By

XIANQING TONY XIE


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

UNIVERSITY OF FLORIDA


1996












ACKNOWLEDGMENTS

I would especially like to thank the chairman of my committee, Dr. Michael

Clare-Salzler, for his intellectual guidance and his unflagging support and

encouragement. I am also grateful to my committee members, Drs. Ammon Peck,

Edward Wakeland, Mark Atkinson and Michael Humphreys-Beher, for their guidance

and support in various phases of my research.

I also thank, Drs. Srinivasa Reddy and Harvey Herschman for their helpful

advice and technical support. I also appreciate the help from Drs. Linder Wicker, Edward

Wakeland, Mark Atkinson and Mary Yui for providing congenic mice, NODscid/scid

mice, and many valuable discussions.

I have enjoyed my association with other members of the Dr. Clare-Salzler's lab

and members of Dr. Peck's lab, especially Andrea Hofig, Sally Litherland, Janet

Cornelius, Jeff Anderson, Chris Robinson, Joanne Johnson-Tardieu and Bertha Moreno-

Altamirano and thank them for help and encouragement. Thanks go to Christy Myrick for

confirming the PGS-2 promoter sequence and special thanks go to both Sally Litherland

and Christy Myrick for helping with thesis editing.

My special thanks go to both Michael and Rita Clare-Salzler for friendship,

support and great meals.

I want to thank my parents for their encouragement for my endeavors through

the years. Lastly, I am grateful to my wife, for her love and support throughout the years.












TABLE OF CONTENTS


ACKN OW LED GM EN TS .................................................................................................. ii

A B STR A C T .........................................................................................................................v

CHAPTERS

1 IN TR O D U CTION .......................................................................................................

Insulin Dependent (type I) Diabetes .......................................................................
The Non-Obese Diabetic (NOD) Mouse ............................................................2...
Polygenic Control of IDD Susceptibility.............................................................3...
Characterization of the Inflammatory Cellular Infiltration in NOD Insulitis .........5
T Cell Activation Defects and Tolerance in the NOD Mouse ............................7...
The Role of the Macrophage in NOD Diabetes, APC Effects on Tolerance........10
The Impact of PGE-2 on APC, T Cells and AICD............................................13
Macrophage Activation and PGS-2 Expression................................................16
Constitutive PGS-1 Expression and PGS-2 Inducibility ..................................19
Regulation of PGS-2 Expression .......................................................................22
Sex Hormones, the Immune Response and Autoimmunity ...............................24
Regulation of Lymphocyte Functions by Nitric Oxide......................................27
iNOS and PGS-2 Crosstalk................................................................................ 30
Rationale for these Studies................................................................................. 31

2. MATERIALS AND METHODS............................................................................. 34

3. R ESU LTS ................................................................................................................. 43

Effects of Macrophages on Dendritic Cell Mediated T Cell Activation in
N O D M ice...................................................................... 43
Spontaneous PGS-2 Expression in Macrophages of NOD and
NODscid/scid Female Mice ...........................................45
Aberrant PG Metabolism in NOD, NODscid/scid, and Congenic Mice ..............52
Factor(s) which Influence PGS-2 Gene Expression and Mechanism of
Spontaneous PGS-2 Expression in NOD Mice..............59
NOD PGS-2 Promoter Sequence Analysis...............................................59
Sex Hormone Influence on the PGS-2 Gene Expression in NOD Mice......63








Role of Monokines on the Expression of PGS-2 Expression in
N O D M ice...................................................................... 64
Contribution of Spontaneous PGS-2 to the Immunopathogenesis
of IDD, a Candidate Gene for Idd5...............................72
PGS and iNOS Inhibitors Affect NOD Diabetes Incidence ..............................82
Impaired AICD in NOD Mice Secondary to Enhanced PG Production ............85

4. D ISCU SSIO N ............................................................................................................. 92

5. SUMMARY AND CONCLUSION .....................................................................104

REFEREN CE LIST ....................................................................................................... 107

BIOGRAPHICAL SKETCH .........................................................................................127







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

THE ROLE OF PROSTAGLANDIN SYNTHASE II (PGS-2) IN
The IMMUNOPATHOGENESIS OF NOD DIABETES

By

Xianqing Tony Xie

December, 1996


Chairperson: Michael Clare-Salzler, M.D.
Major Department: Pathology and Laboratory Medicine


The non-obese diabetic (NOD) mouse is a well-established animal model of the

human autoimmune disease, insulin dependent diabetes (IDD). In NOD mice, both T

cells and macrophages (MPs) play an important role in the immunopathogenesis of this

disease. The cellular and molecular basis of MP involvement in NOD diabetes, however,

is largely unknown. In the present study, we have investigated the expression of

prostaglandin synthase II (PGS-2) and PGE-2 production in the NOD MPs and its role in

the immunopathogenesis of NOD diabetes.

Our results indicate that 1) MPs from NOD and NOD scid/scid estrus female

mice, unlike those of non-autoimmune mice, spontaneously express abundant PGS-2

mRNA and protein. 2) There is a sexual dimorphism in the PGS-2 mRNA and protein

expression. Whereas MPs from male NOD mice do not spontaneously express PGS-2,

spontaneous PGS-2 mRNA expression is observed at all times in MPs of NOD female

mice, while PGS-2 protein expression is observed only during estrus phase. Our in vitro

experiments demonstrate that female steroid hormones strongly influence PGS-2







expression, whereas male steroid hormones have no effect. 3) PGS-2 expression in NOD

MPs is insensitive to IL-10. In our in vitro studies, mrIL-10 completely suppressed LPS

induced PGS-2 expression in MPs of control mice at 10ng/ml. In contrast, mrIL-10 did

not suppress spontaneously expressed PGS-2 in NOD MPs at concentration as high as

500 ng/ml. 4) Different PGS-2 expression in both NOD and normal chromosome 1

congenic mice suggest that PGS-2 could be a candidate gene of diabetes susceptibility for

Idd5. 5) There is enhanced prostaglandin production by MPs through the constitutive

PGS-2 expression. Treating female NOD mice with drugs that block PGS-2 enzymatic

activity significantly delays the onset and reduces the incidence of diabetes. 6) There is a

general impaired activation and deletion of T cells in NOD mice which is predominantly

affected by the NOD MHC molecule H-27, and that PGS-2 appears to contribute to a

defect in T cell deletion. These data strongly suggest that spontaneous expression of

PGS-2 in NOD MPs plays an important role in the immunopathogenesis of diabetes in

NOD mice and is a candidate gene for Idd5.













CHAPTER 1
INTRODUCTION
Insulin Dependent (type I) Diabetes ODD)

Insulin-dependent diabetes mellitus (IDD, type I diabetes) is an autoimmune

disease caused by mononuclear cell infiltration of the pancreatic islets that leads to the

selective destruction of insulin producing P3 cells (Gepts et al., 1965; Palmer et al., 1983).

Progressive loss of islet beta cells results in insufficient insulin production, which causes

clinical diabetes, characterized by hyperglycemia, ketonuria, glycosuria, polyuria,

polydypsia, and weight loss. The inability to regulate blood glucose levels eventually

results in death unless insulin is administered. In this disease, other pancreatic endocrine

cells, such as somatostatin-producing cells, are not affected and autoantibodies to several

pancreatic islet cells/antigens are detected before the onset of disease (Bottazzo et al.,

1974).

Insulin-dependent diabetes in both humans and rodent models is a multifactorial

disease which results from both genetic and environmental factors. Barnett et al. (1981)

and Rotter et al. (1990) found the risk of disease for an identical twin of an individual

with IDD to be more than 80 times the risk for the general population which strongly

suggests a genetic contribution. However, the findings that among twins of patients with

IDD only 33% subsequently develop diabetes and that 90% percent of patients with

newly diagnosed IDD do not have an affected first-degree relative suggests that








environmental factors, such as diabetogenic virus, dietary constituents, or P cell-tropic

toxins also influence the development of IDD (reviewed by Atkinson and Maclaren,

1994).

The Non-Obese Diabetic (NOD) Mouse

The non-obese diabetic (NOD) mouse is a well-established spontaneous animal

model for IDD. Makino et al. (1980) developed the NOD mouse from a single diabetic

female that spontaneously arose from a noninbred cataract-prone mouse strain. They

established the NOD as a homozygous strain by inbreeding for over 20 generations. The

NOD mouse shares many common immunological features with human IDD, such as the

development of autoantibodies, insulitis, and ketosis-prone diabetes. NOD mice that

develop diabetes die within a few weeks of onset unless they receive insulin therapy

(Makino et al., 1980; 1981).

Insulin-dependent diabetes in the NOD mouse is influenced by many factors.

Insulitis and diabetes susceptibility are under polygenic control and include both MHC

and non-MHC related genes (Leiter, 1989; Wicker et al., 1996). Environmental factors

also influence the incidence of disease and include dietary conditions, caging conditions,

and specific pathogens such as murine hepatitis virus (Leiter et al., 1990).

The NOD, however, differs from humans in that diabetes incidence in females is

markedly higher than in males. In females, the onset of diabetes is observed at

approximately 8 weeks of age, with a cumulative incidence of 80% diabetes by 30 weeks

of age. In contrast, in male NOD mice, the age of diabetes onset is similar to females, but

the incidence by 30 weeks of age is only about 30% (Makino et al., 1981).








Polygenic Control of IDD Susceptibility

Inheritance of IDD and insulitis susceptibility are under polygenic control in the

NOD mouse. Early studies showed that all crosses between the NOD and non-IDD

susceptible control mouse strains resulted in Fl mice that did not develop insulitis and

were free of diabetes suggesting the genetic control of diabetes is recessive (Ikegami et

al., 1993).

Published studies have shown that defects in MHC molecules are associated with

several autoimmune diseases (Erlich et al., 1993). MHC molecules play a crucial role in

antigen presentation and T cell activation and may thus play a central role in T cell

tolerance. In the thymus, MHC molecules present both foreign and self antigens to

immature T cells. Through interaction via T cell receptors, T cells are chosen through

positive and negative selection (Ashton-Richardt et al., 1994; Sebzda et al., 1994). In the

periphery, the MHC also presents self antigen to self reactive T cells which causes

activation induced cell death to maintain peripheral immune tolerance (Rocha et al.,

1991; Zhang et al., 1992). The first IDD susceptibility locus (Iddl) to be identified was

the NOD MHC II, designated H-297 (Hattori et al., 1986; Prochazka et al., 1987; Ikegami

et al., 1988), a unique I-A molecule. Ikegami et al. (1988; 1993) reported that the CTS

strain (H2cs) is a natural recombinant expressing the same class II gene as NOD

(including a non-functional Ea gene) but different class I and III loci, and developed a

NOD MHC congenic mouse expressing the CTS MHC. While NOD.H-2ch congenic mice

developed diabetes, the frequency was significantly less than in the NOD strain. These

data support the hypothesis that the NOD I-A molecule is only partially responsible for








diabetes and that class I or III alleles also contribute to disease development. NOD and

CTS differ at Hsp70, Bat5, and Tnfi3 in the class III region, and therefore, in addition to

the class I loci, these loci are candidate genes for IDD. Thus, in addition to the genes

encoding the I-A molecule, other genes within the MHC of the NOD may function

together to create the effects of Iddl.

Although the NOD MHC is required for the development of diabetes, when the

MHC of the NOD strain was expressed on the B 10 or B6 genetic background, no insulitis

or diabetes was observed (Ikegami and Makino., 1993; Wicker et al., 1993). The

congenic strains, B10.H-2g and B6.H-2g7, which express the NOD MHC, demonstrate

that the MHC itself is not sufficient for developing pathology associated with diabetes

and suggest that non-MHC genes also play a role in diabetes. Indeed, many non-MHC

susceptibility loci have been localized to regions of different chromosomes. Idd2, located

on chromosome 9 near the Thyl locus was the first of the many non-MHC linked Idd loci

to be identified (41). Analysis of progeny from the (NOD x B10.H-297) Fl x NOD BCl

generation also led to the mapping of other Idd susceptibility regions including Idd3

(Ghosh et al., 1993; Todd et al., 1991; Wicker et al., 1994) and IddlO (Ghosh et al., 1993;

Wicker et al., 1994) on chromosome 3, Idd4 on chromosome 11 (Todd et al., 1991), Idd5

on chromosome 1 (Comall et al., 1991), Idd6 on chromosome 6 (Ghosh et al., 1993),

Idd7 on chromosome 7 (Todd et al., 1991), Idd8 on chromosome 14 (Todd et al., 1991),

and Idd9 on chromosome 4 (Rodrigues et al., 1994). Additionally, Iddl1 was mapped on

chromosome 4, Idd 12 on chromosome 14, Idd 13 on chromosome 2, and Iddl4 on

chromosome 13 (Wicker et al., 1995).








In addition, (NOD x B10.H-2 7) Fl x NOD BC1 progenies were examined for

pancreatic islet pathology to determine which Idd loci were linked to insulitis. It was

found that Idd3 and IddlO0 on chromosome 3 and Idd5 on chromosome 1 showed

significant linkage to insulitis (Ghosh et al., 1993; Todd et al., 1991; Wicker et al., 1994).

Two more insulitis loci were later identified on chromosome 1 by a cross between NOD

and C57BL/6 and NZW mice (Garchon et al., 1991).

With the discovery of these Idd loci, candidate genes have been suggested and

examined for functional polymorphism. Since many loci have been localized to a large

region of a particular chromosome which contains many genes, it makes candidate gene

selection difficult. IL-2 has been proposed as a candidate gene for Idd3 (Chesnut et al.,

1993), Fc receptor for IddlO (Prins et al., 1993), and both Bcl-2 and Lsh/Ity/Bcg for Idd5

(Comall et al., 1991; Garchon et al., 1991). Prostaglandin synthase 2 (PGS-2) enzyme,

the rate limiting enzyme for prostanoid metabolism, is responsible for the production of

large quantities prostaglandins by MPs. The PGS-2 is located on chromosome 1 near

Idd5. Due to the differential expression of the PGS-2 gene that I have observed in NOD

in comparison to congenic and control mouse strains, and the correlation of this PGS-2

phenotype to the development of autoimmunity, I have proposed PGS-2 as a candidate

susceptibility gene for Idd5.

Characterization of the Cellular Infiltration in NOD Insulitis

The development of insulin-dependent diabetes in NOD mice is preceded by the

progressive accumulation of lymphocytes within the pancreatic islets of Langerhans

(termed insulitis). Insulitis is first noted at 3 to 4 weeks of age, and virtually all female








and male NOD mice have insulitis by 2 months of age (Fujita et al., 1982).

Histopathological examination of the pancreas of the NOD mouse reveals that the

insulitis is initially confined to ducts and vascular areas followed by movement to the

peri-islet areas, and progresses until the islet is surrounded by lymphocytes. By eight

weeks in females, lymphocytes and monocytes penetrate into the islet itself and insulin-

producing P cells are specifically destroyed while leaving delta and alpha cells intact

(Wicker et al, 1986). Diabetes ensues after about 12 weeks when a large proportion of the

islets have been destroyed.

Immunohistological studies show that macrophages, dendritic cells, B cells, and

T cells are present in the pancreatic infiltrates with the majority of cells being Thy-1-

positive T cells. Shimizu et al. (1987) have shown that among T cells, L3T4+ T helper

cells are in greater numbers than the Lyt-2+ T cytotoxic cells. Miyazaki et al. (1985)

found a predominance of T lymphocytes in the NOD islet from 3 to 13 weeks of age.

Islets from 6 weeks old mice were characterized by a marked T-helper and T-cytotoxic

cell infiltrate. Shimizu et al. (1987) found that the T lymphocytes were localized to the

islets and proposed that they were responsible for beta cell destruction. B lymphocytes,

though abundant, were observed more peripherally and islet cell specific antibodies were

either not detected (Shimizu et al., 1987; Miyazaki et al., 1985) or detected late in the

disease process, at 12-18 weeks (Kanazawa et al., 1984).

Jarpe et al. (1991) observed that islet infiltration of CD8+ T cells and MHC class

II positive macrophages at 5-6 weeks of age, followed by a transient wave of CD4+

positive T cells and then by an influx of B cells and CD8+ T killer cells and proposed the








following model for the cellular time course of insulitis: antibody-dependent cellular

cytoxicity and opsonization-phagocytosis at the early stage (5-6 weeks) followed by a

transient wave of CD4+ T cells (6-8 weeks) followed by CD8+ T cell damage at the late

stage (8-10 weeks).

T cell Activation Defects and Tolerance In NOD Mouse

Many studies suggest that T cells play a central role in the pathogenesis of P3 cell

destruction in NOD mouse. A T cell role in disease has been suggested based on the

observation that neonatal thymectomy prevents diabetes in NOD mice (Ogawa et al.,

1985), and diabetes is transferred into irradiated young recipients by adoptive transfer of

splenocytes from overtly diabetic NOD mice (Wicker et al., 1986). By using purified T

cells and fractionated L3T4+ (CD4+) and Lyt-2+ (CD8+) T cells from diabetic NOD

spleens, Bendelac et al. (1988) and Miller et al. (1988) demonstrated that both CD4+ and

CD8+ T cells were required for transfer of disease while B cells were not. Shizuru et al.

(1988) and Wang et al. (1991) further showed that diabetes in NOD mice is T cell

dependent as it is prevented by injection of anti-CD4 antibodies.

T lymphocytes normally respond to a wide variety of foreign antigens and are

tolerant to self antigens. Mechanisms for regulation of T cell self-tolerance can be found

in both the thymus and in the periphery. The TCR of immature T cells differentiating

within the thymus interacts with peptides presented by MHC gene products expressed on

both hematopoietically derived APC and thymic epithelium (Sprent et al., 1987). The

interaction of immature CD3+/owCD4+CD8+ thymocytes with foreign peptides in the

context of self MHC gene products induce positive selection of T cells. The T cells which








do not undergo positive selection are eliminated. To avoid the development of T cells

bearing an autoreactive TCR, T cell precursors that recognize self-peptides presented in

the context of self-MHC gene products are negatively selected by an activation induced

cell death (AICD) process known as apoptosis (Finkel et al., 1989; Murphy et al., 1990).

When mature peripheral T cells are fully activated, they are induced to follow one of

three known paths: differentiation to effector cells (including suppressor cells), energy, or

activation induced cell death (Weber et al., 1990; Nigata et al., 1995; Dhein et al., 1995).

AICD provides an antigen specific regulatory system dependent on the qualitative and

quantitative signals provided by the interaction of antigen presenting cell (APC) and self-

reactive T cell. Any defect intrinsic to APC or T cell affecting their interaction or

subsequent T cell activation may possibly result in impairment of AICD. Impaired AICD

could contribute to a common defect in autoimmunity: the accumulation and activation

of self-reactive cells (Mountz et al., 1994; Thompson, 1995).

Many studies have demonstrated that treatment of the NOD mouse with high

doses of glutamic acid decarboxylase (GAD) induces splenic T cell tolerance to this

antigen and prevents diabetes (Tisch et al., 1993; Kaufman et al., 1993). GAD is the

target of 64 kD reactive autoantibodies in both humans and NOD mice and is one of

important islet cell autoantigens (Baekkeskov et al., 1990). Two forms of GAD (65 and

67 kD) exist (Bu et al., 1992). Both GAD 65 and 67 have been identified in islet cells and

may play a role in the inhibition of somatostatin and glucagon secretion, as well as in the

regulation of insulin secretion and proinsulin synthesis. Because treatment of the NOD

mouse with GAD induces splenic T cell tolerance to this antigen, it suggests that T cell








tolerance is not complete in NOD mice but can be effectively established by activating

autoimmune T cells with high doses of autoantigen. Studies have suggested that the

activation of T cells by powerful stimuli such as high doses of antigen, superantigen, or

by persistent encounters with self antigens on APCs leads to clone elimination in vivo

(Web et al., 1990; MacDonald et al., 1991). Antigens that are processed and presented in

an inefficient fashion can continue to stimulate T cells responses in the periphery, but

would be unable to induce tolerance (Milich et al., 1989; Mamula, 1993). Similarly, the

induction of activation-driven T cell death both in the periphery and during negative

selection in the thymus, requires quantitatively more antigenic stimulation than is needed

to induce positive selection or trigger effector T cell proliferative responses (Ucker et al.,

1992; Sebzda et al., 1994).

These findings indicate that the threshold of T cell activation required to induce

tolerance is much higher than that required to trigger an effector response. In addition, the

stimulation of immunoregulatory T cell requires a more highly activated APC than is

required to activate effector T cells (Ishikura et al., 1993). Thus, any genetic defect that

compromises the differentiation or function of APCs could preferentially diminish the

ability of these cells to present antigens in a manner quantitatively sufficient to induce

tolerance and/or activate immunoregulatory T cells, without fully abrogating their ability

to activate effector T cells.

Defects in both APCs and the T cell signaling pathway have been described to

contribute to impaired T cell activation (De et al., 1994; Yokona et al., 1989; Ransanen et

al., 1989; Serreze et al., 1993b; Rapoport et al., 1993). IL-2 production following APC -








T cell interaction appears to be of critical important for tolerance in general as mice that

do not express IL-2, or the IL-2 receptor as a result of gene disruption develop severe

generalized autoimmune disease (Sadlach et al., 1993; Willerford et al., 1995). Many

studies suggest that there is impaired IL-2 production and activation of regulatory T cells

in response to self antigen presenting cells and mitogens in both human and murine IDD

(De et al., 1994; Yokona et al., 1989; Ransanen et al., 1989; Serreze et al., 1993b). This

may be an important factor in impairing the development of tolerance in this disease.

The Role of the Macrophage in NOD Diabetes. APC Effect on Tolerance

The mechanisms) by which the (3 cell is specifically destroyed is not known, but

MPs in addition to T cells have been implicated. Jansen et al. (1994) demonstrated that

prior to infiltration of T cells (3 week of age), MPs and DCs accumulate around the islets,

with DCs remaining outside the islet, but MPs going on to penetrate the islet. Blocking

the migration of MPs into the islet by treatment of NOD mice with CR3

(myelomomocytic adhesion promoting type-3 complement receptor) prevents intra-islet

infiltration by both MPs and T cells and inhibits development of IDD (Hutchings et al.,

1990). Furthermore, depletion of MPs by treatment of NOD mice with silica injections

also blocks the development of insulitis and diabetes (Charlton et al., 1989; Lee et al.,

1988). Macrophage accumulation in the pancreas may well have a major role in the

processing and presentation of islet antigens to autoreactive T cells within the islet (Lee et

al., 1988; Chariton et al., 1988). Additionally, activated MPs produce the monokines IL-1

and TNF-a that induce islet cell and MP production of NO, a potentially toxic metabolite

to which p cells are highly sensitive (Cobet et al., 1995). Many other oxygen radical








species, produced by activated MPs, can also induce P cell death (Cobet et al., 1995;

1992).

NOD MPs differ from MPs of non-autoimmune control strains of mice in their

development and function. Serreze et al. (1993a) found that NOD bone marrow cells

proliferate poorly in response to the myeloid growth factor CSF-1 and generate fewer

phenotypically mature (Mac-3) MPs than are generated from CSF-1 stimulated marrow

from diabetes-resistant mouse strains. In addition, the finding that LPS-stimulated MPs

from NOD mice secrete relatively little IL-1 (Jacob et al., 1990; Serreze et al., 1990)

indicates that a large proportion of APCs in NOD mice may be incompletely

differentiated, as MPs do not acquire the ability to secrete IL-13 until later stages of

development. The reduced capacity to produce IL-1 may be of practicable importance as

this cytokine may preferentially activate T helper type 2 (Th2) responses associated with

disease protection in the NOD (Rabinovitch, 1994). Serreze et al. (1990) and Campbell

et al. (1991) further demonstrated that the reduced ability of NOD mice to generate

functionally mature MPs is associated with both aberrant regulation of RNA transcripts

encoding the receptors for CSF-1 and IFN-y, and a decreased ability to activate protein

kinase C (PKC) second messenger activities coupled to these receptors. PKC activity is

also reduced in NOD T cells which suggests a PKC induction defect in more than one cell

lineage.

In addition to developmental defects, MHC associated defects intrinsic to MP and

other APC may also contribute to the reduced capacity to active T cell. It has been found

that NOD mice produce significant levels of IFN-y (Campbell et al., 1991) and that MHC








class I, but not MHC class II, expression is aberrantly down-regulated in a cell-specific

manner on NOD MPs, but not on pancreatic 0 cells exposed to interferon-gamma (IFN-y)

for 6 days (Serreze et al., 1993b). In contrast, MHC class I expression is up-regulated

normally in IFN-y treated macrophages from diabetes resistant NOR mice (Serreze et al,

1993b). Serreze et al. (1993b) and Kay et al. (1991) suggest that aberrant down-

regulation of MHC class I expression by IFN-y in NOD derived APC may impair the

ability of these cells to mediate negative selection of autoreactive CD8+ T cells, which

are then efficiently targeted to pancreatic P cells that express high levels of MHC class I.

The unique H-27 MHC molecule also plays a central role in T cell activation and

murine IDD. IDD rarely develops in congenic stocks of NOD that express heterozygous

MHC haplotypes from other strains and those immunotolerogenic defects most readily

occur when the H-27 is homozygous (Wicker et al., 1992; Prochazka et al., 1989). The

effect of the H-2g7 MHC molecule maybe due to its instability and decreased efficiency

in binding and presenting self antigens as recently reported (Carrasco-Marin et al., 1996).

The finding that NON mice congenic for the H-2g7 haplotype (NOD.H-27g) are

diabetes resistant (Ucker et al., 1992) indicates that APC expression of the diabetogenic

MHC haplotype is insufficient for development of disease, and that the pathogenesis in

NOD mice is dependent on synergistic interaction between diabetes susceptibility genes

both inside and outside of the MHC.

The other defects in NOD MPs, in combination with the diabetogenic H-2g may

perturb MP processing and/or presentation of p cell autoantigens in such a way that their








ability to mediate negative selection of autoreactive T cells in thymus and periphery is

impaired while retaining their ability to active "low level" effector functions.

The Impact of Prostaglandin E-2 on APC. T cell. and AICD

Prostaglandins, especially PGE-2, exert some very important effects during an

immune response. PGE-2 produced early during the activation of MPs has

proinflammatory effects such as vasodilatation, increasing vascular permeability, and

under specific conditions stimulates MP cytoxicity. PGs also have a profound effect on

monokine expression following the activation of MPs (i.e., after LPS stimulation). They

readily suppress IL-I and TNF-a expression while up-regulating IL-10 production

through their effects mediated by cAMP on the IL-10 promoter (Meisel et al., 1996).

PGE-2 also affects T cell cytokine expression and promotes Th2-like cytokine secretion

profiles in murine and human CD4+ T cells by inhibiting Thl-associated cytokines IL-2

and IFN-y and up-regulating the production of the Th2-associated cytokines IL-4 and IL-

5 in a dose-dependent manner (Snijidewint et al., 1993).

Many studies have shown that PGE-2 modulates immune responses by markedly

inhibiting T cell activation events including IL-2 and IL-2 receptor expression (Goetzl et

al., 1995; Lee et al., 1993). The effects of PGE-2 are mediated by up-regulating the

intracellular second messenger, cAMP, which binds to its intracellular receptor protein

kinase A (PKA). As with other agents that increase cAMP (histamine, adenosine,

forskolin, and cholera toxin), the mechanism by which PGE-2 activates PKA involves the

binding of cAMP to the inactive tetrameric holoenzyme and its dissociation into two

regulatory and two catalytic subunits (Krammer, 1988). This kinase-mediated activation








leads to the inhibition of IL-2R expression and IL-2 production (both mediated by a

decrease in IL-2 nuclear transcription and IL-2 mRNA stability) (Anastassiou et al.,

1992).

Activation induced cell death is an important mechanism for peripheral T cell

tolerance (Ashton-Richardt et al., 1994; Sebzda et al., 1994; Rocha et al., 1991; Zhang et

al., 1992). After lymphocytes mature and leave the thymus, they are functionally

competent, e.g., capable of responding to antigenic stimulation by proliferating and

differentiating into effector cells. Among the progeny of antigen-stimulated lymphocytes,

only a small fraction develop into functional effector and memory cells. The majority

probably die by apoptosis (Rocha et al., 1991; Zhang et al., 1992). The process of

activation-induced cell death may be enhanced by the exposure of antigen stimulated

lymphocytes to growth factors, such as IL-2 in the case of T cells (Klas et al., 1993). In

mice, administration of large doses of anti-T cell receptor antibodies or superantigens,

such as staphylococcal endotoxins, results in the deletion of T cells that express antigen

receptors which specifically bind these antibodies or superantigens (Trauth et al., 1989;

Kawabe et al., 1991). This is due to activation induced apoptosis, and may be an

exaggerated version of the phenomenon that normally occurs in clones of antigen-specific

lymphocytes that encounter self antigen. Activation induced cell death in mature

lymphocytes, therefore is a homeostatic mechanism that functions to regulate the number

of antigen-stimulated clones and also serves a highly protective function in the case of

autoantigens. Indeed, impaired AICD, as in lprllpr mice, contributes strongly to the

development of autoimmunity (Wu et al., 1994). It has been suggested that T cell receptor








(TCR) stimulation of resting mature peripheral T cells causes them to undergo activation,

but not AICD, whereas TCR stimulation of cycling cells causes a significant number of

lymphocytes to die. Some critical events for AICD include TCR activation, lymphokine

mediated cell cycle progression (IL-2 production), and TCR re-engagement (Guery et al.,

1995; Stockinger et al., 1992). Several studies have suggested that AICD can be markedly

inhibited by anti-IL-2 antibodies and by agents that block the cell cycle, i.e., cAMP

(Boehme et al., 1993; Critchfield et al., 1995).

Yokono et al. (1989) demonstrated that aberrant PG production from NOD MPs

contributes to the suppression of T cell activation and IL-2 production in Con-A

stimulated spleen cells in NOD mice. Preliminary studies in our lab have also shown that

MPs are responsible for the suppression of T cell activation in NOD syngeneic mixed

lymphocyte response (SMLR) and that PGs mediated this defect.

Because of PG effects on T cell activation, these molecules may profoundly affect

AICD. Several studies have shown that PGE-2 markedly impairs AICD of cycling mature

T cells through activation of adenylate cyclase and generation of cAMP (Goetzl et al.,

1995; Ucker et al., 1994). The quantitative signal given by the interaction of IL-2 and IL-

2R has been shown to be decisive in cell activation, proliferation, AICD, and immune

tolerance (Sadlack et al., 1993; Willerford et al., 1995). PGE-2 is a potent inhibitor of IL-

2 and IL-2 receptor expression and also induces high levels of cAMP and blocks cell

cycle in T cells. The mechanism by which cAMP protects T cells from apoptosis may

involve its inhibition of Ras-dependent activation of the signal transmission pathway

leading from Raf to mitogen-activated protein kinases (Cook et al., 1993). This inhibitory








mechanism is postulated to require protein kinase A. In addition, PGE-2 appears to play

a regulatory role in the function of ceramide, a main AICD signal transducer (Jarvis et al.,

1994; Hannum et al., 1994). Ceramide in turn promotes the production of PGE-2 though

induction of PGS-2, thus promoting its own regulation (Hannum et al., 1994).

An established in vivo experimental model of AICD is the immunization of mice

with the bacterial superantigen, Staphylococcus aureus enterotoxin B (SEB) (MacDonald

et al., 1991; Kawabe et al., 1991). In this experimental paradigm, SEB immunization of

BALB/c mice initially leads to the expansion of SEB reactive T cells bearing the Vp8+ T

cell receptor (Vp8+) within 48 hours of immunization. By 10 days post immunization,

however, Vp8+ T cells are reduced secondary to cell death. In contrast, control T cells

that are not activated by SEB, i.e., Vp6+ TCR bearing T cells do not expand nor are they

deleted in SEB immunized mice.

Macrophage Activation and PGS-2 Expression

Cells of the mononuclear phagocyte system originate in the bone marrow, and

after maturation and subsequent activation achieve varied morphological forms and

functions. The first cell type that enters the peripheral blood after leaving the marrow is

incompletely differentiated and is known as the monocyte. Once the monocyte migrates

into tissues, these cells undergo further maturation and become tissue specific

macrophages. In tissues, their functions are strongly down-regulated by a variety of

suppressive factors normally present. They therefore remain in a quiescent state until

stimulated or challenged with a wide variety of agents such as lympokines from

lymphocytes, tumor cells, bacteria, foreign particles, or environmental toxins. Once








activated, MPs become able to produce TNF-a, IL-1, oxygen radicals, NO and various

immunoregulatory factors that include components of the complement system and

arachidonic acid metabolites such as prostaglandins (PGs) (Meisel et al., 1996).

Following LPS activation of MPs, there is a defined order of monokine mRNA and

protein expression with TNF-a and IL-1 expression preceding the expression of PGS-2,

and IL-10 following PGS-2 expression (Meisel et al., 1996). The order of expression for

these monokines is critical as there are many complex regulatory interactions amongst

them. It has been found that TNF-a and IL-1 up regulate PGS-2 and IL-10 expression.

After PGS-2 is expressed, PGE-2 production potently suppresses TNF-a and IL-1

expression while up-regulate IL-10 production. IL-10, in turn, suppresses the expression

of TNF-a, IL-1 and PGS-2 (Mertz et al., 1994).

The activation of MPs, however, is not a singular process. Activation consists of

quantitative alterations in the expression of various gene products (proteins) that endow

the activated MPs with the capacity to perform some functions that can not be performed

by the resting MP such as tumor killing (Mackey et al., 1993; 1987). Two of the most

extensively studied marker proteins for MP activation in tumor cell killing are the

proteins p47b and p71/73 (Mackey et al., 1993; 1987). Unstimulated MPs do not express

either marker, whereas expression of p47b alone is evidence that a MP has become

primed by interferon, and simultaneous expression of both p47b and the p71/73 complex

correlate with full activation and tumor cell killing. Recently, p71/73 has been identified

as an inducible prostaglandin endoperoxide synthase (cyclooxygenase 2 or PGS-2) (Xie

et al., 1993).








Prostaglandin endoperoxide synthase is the rate limiting enzyme in prostanoid

metabolism. The biosynthetic pathways for PGs or leukotrienes include three distinct

stages: (1) release of arachidonic acid from membrane phospholipids which provides

substrate, (2) metabolization of free arachdonic acid (AA) to leukotrienes by

lipooxygenase or oxygenation by cyclooxygenase to yield PGH, a prostaglandin

endoperoxide that serves as precursor for other prostaglandins, and (3) depending on the

enzymes present in the cell, conversion of PGH to different prostanoids (including PGE-

2, PGD-2, prostacyclin, and thromboxane). The biosynthetic pathways are summarized

below in Figure 1.


Figure 1. Prostanoid Metabolism








AA release involves either the action of phospholipase A2 (PLA2) on

phosphatidylcholine or phosphatidylethanolamine, yielding arachidonate, or the action of

a phospholipase C on phosphatidylinositol, yielding diacylglycerol, which in turn

undergoes cleavage to give free arachidonate. There are two isoforms of PLA2, secretary

PLA2 (sPLA2) and cytosolic PLA2 (cPLA2). They have similar functions, but effect AA

metabolism at different locations (Glaser et al., 1995). PLA2 activation and expression

occur as a result of tissue-specific stimuli by hormones such as bradykinin, epinephrine,

or proteases such as thrombin (Balsinde et al., 1990). Release of AA can also occur if

membranes are perturbed.

Mouse and human cells express two prostaglandin synthases, PGS-1 and PGS-2.

PGS-1 is constitutively expressed, responsible for the production of low levels of

prostanoids, and serves physiological housekeeping functions. PGS-2 is an inducible

enzyme responsible for the production of large quantities of PGs (Xie et al., 1993). In

contrast to PGS-1, PGS-2 mRNA and protein are not expressed in resting MPs but can be

induced by mitogens, LPS, and cytokines such as IL-1, TNF-a (Rzymkiewicz et al.,

1994; Tordiman et al., 1995).

Constitutive PGS-1 Expression and PGS-2 Inducibility

Both PGS genes are single copy genes. They are located on different

chromosomes. PGS-1 maps to chromosome 2 in the mouse (Xie et al., 1993) and 9 in the

human (Kosaka et al., 1994). PGS-2 maps to chromosome 1 in both species (Xie et al.,

1993; Kosaka et al., 1994; Jones et al., 1993). Both human and murine PGS-1 are -22 kb

in length and contain 11 exons and 10 introns with the exon-intron structure conserved








between humans and mice. The murine PGS-1 cDNA encodes a 602-aa polypeptide,

containing a 26-aa signal peptide and its mRNA is about 2.7 kb. PGS-1 is localized to the

membrane of endoplasmic reticulum (ER). It is constitutively expressed in almost all

tissues and cells and plays an important role in maintaining normal vascular, gastric, and

renal homeostatic functions (Wang et al., 1993).

In contrast to PGS-1, the PGS-2 gene is ~8 kb in the length and contains 10 exons

and 9 introns (Xie et al., 1991; Kujubu et al., 1991; 1993). PGS-2 cDNA has been cloned

from human, rat, mouse, and avian sources. Comparison of PGS-2 sequences among

these species shows about 80% identity and about 60% homology with PGS-1. All PGS-2

cDNAs encode a 604-aa polypeptide with four potential N-glycosylation sites. The signal

peptide (17 aa) of the PGS-2 protein is shorter than the PGS-1 signal peptide. Near the C-

terminus of PGS-2, there is an 18-aa insert which is absent in PGS-1. This protein

sequence has been used to generate PGS-2 specific antibody that uniquely binds to PGS-

2.

The single best characterized distinction between PGS-1 and PGS-2 is their

differential regulation of expression. PGS-1 is expressed constitutively in almost all

tissues, whereas PGS-2 is selectively expressed, primarily in macrophages, endothelial

cells, fibroblasts, and smooth muscle cells, but only in response to activating agents such

as cytokines, growth factors, hormones, and tumor promoters. The PGS-2 mRNA

expressed in these cells is 4.0-4.5 kb, which is longer than that of PGS-1 (2.7 kb). In

contrast to PGS-1, which localizes to endoplasmic reticulum (ER), PGS-2 preferentially








localizes to the nuclear envelope (Morita et al., 1995) and suggests that prostanoids

formed via PGS-2 may directly function within the nucleus.

Constitutive PGS-1 expression suggests that cells using PGS-1 to produce

prostaglandins involves a rapid response to stimulation by circulating hormones. Because

of the time lag required for PGS-2 induction in a cell or tissue, this enzyme is available to

produce prostaglandins only after activation following specific physiological events,

such as inflammation, mitogenesis, and ovulation (Hedin et al., 1987; Wong et al., 1989).

Recent evidence suggests that the PGS-1 and -2 produce prostaglandins through separate

pathways and use different phospholipases that are coupled to different signaling

pathways (Reddy et al., 1994; Murakami et al., 1994).

The regulation of PGS-2 has been extensively investigated by examining its

mRNA and protein expression following stimulation in cultured fibroblasts, endothelial

cells, and in purified macrophages or MP cell lines. PGS-2 expression can be induced by

phorbol esters such as phorbol-12-myristate-13-acetate (PMA) (Rysecket al., 1992),

TNF-a, IL-I (O'Banion et al., 1992; Ristimaki et al., 1994; Rzymkiewicz et al., 1994),

serum (DeWitt et al., 1993), growth factors, and LPS (Habib et al., 1993; Tordiman et al.,

1995). Induction of PGS-2 by these stimulators is inhibited by cycloheximide which is in

keeping with the concept that PGS-2 is a primary response gene. The magnitude of

induction can reach 50 fold over the basal level (DeWitt et al., 1993).

The induction of PGS-2 is rapid. PGS-2 mRNA expression is induced -30 min

after stimulation, peaks at 1 hr, and returns to basal level at 4 hr (DeWitt et al., 1993).

PGS-2 mRNA level, which peaks 1 hr following the addition of serum, declines by more








than 50% by 3 hr, suggesting that half-life (ti/2) of PGS-2 mRNA is 2 h or less. PGS-2

protein levels begin to increase immediately following induction of PGS-2 mRNA and

are detectable 1 hr post induction. Protein levels peak at 2 to 4 hr after induction and

return to near baseline levels by 6 hr (DeWitt et al., 1993).

A class of agents, the non-steroid anti-inflammatory drugs (NSAIDs), inhibit the

enzymatic activity of PGS. Most PGS inhibitors such as indomethacin are non-selective

and inhibit both PGS-1 and PGS-2 (Meade et al., 1993). New NSAID drugs are being

developed, such as NS-398, which are selectively inhibits PGS-2 enzyme activity (Futaki

et al., 1994), and have an important therapeutic application in inflammatory diseases. In

addition, other physiological inhibitors of PGS-2 exist, such as glucocorticoids which

also selectively inhibit PGS-2 but not PGS-1 (DeWitt et al., 1993). Anti-inflammatory

cytokines such as IL-10, TGF-3 also inhibit PGS-2 expression (Mertz et al., 1994).

Regulation of PGS-2 Expression

Nuclear run-off experiments in murine NIH3T3 cells indicated that serum

increases the rate of PGS-2 mRNA synthesis (DeWitt et al., 1993; Evett et al., 1993).

Increases in mRNA synthesis in fibroblasts were noted 12-30 min after the addition of

fetal calf serum. The magnitude of the increase paralleled that of steady-state mRNA

levels. Induction of PGS-2 protein by mitogenic growth factors, cytokines, and hormones

has been attributed to increased transcription.

Sequence analysis of the untranslated 5'-flanking region of PGS-2 and PGS-1 has

begun to shed light on the regulation of PGS-2 transcription. There are several consensus

sequences for transcriptional activation on the 5'-flanking region of the human and








chicken PGS-2 gene (Figure 2). The PGS-2 gene has a canonic TATA box 30 bp

upstream from the transcription start site (Tazawa et al., 1994; Xie et al., 1993). It

contains several putative regulatory elements in the 280 bp of the 5'-flanking region:

cyclic AMP response element, IL-6 response element (NF-IL6), C/EBP, AP-2, nuclear

factor-Krp (NF-Kp), and SP-1 sites. Further upstream are putative PEA-3, GATA-1, NF-

Ki, and NF-IL6 binding sites. The PGS-2 gene has features of a primary response gene

and is expected to be inducible by phorbol ester, cAMP, and a number of cytokines and

growth factors.




E-BOX
SP-1
ATF/CRE
\ TATA






-966 -400 -300 -200 -100 0
NF-IL-6


NF-IL6 (C/EBP)

Figure 2. The Presumptive cis-acting regulatory sequences present in the PGS-2
promoter.



PGS-2 expression is also regulated at post-transcriptional levels (Evett et al.,

1993; Ristimaki et al., 1994). PGS-2 mRNA is unstable compared with PGS-1 mRNA, a

feature predicted from the presence of multiple RNA instability sequences (AUUUA) in








its 3'-untranslated region. PGS-2 mRNA is translated as soon as it is synthesized;

therefore, the short mRNA half-life limits PGS-2 production post-transcriptionally

(DeWitt et al., 1993).

PGS-2 expression is regulated at the protein level as well. PGS-2 protein is much

less stable than PGS-1 in fibroblasts, indicating a post-translational regulatory mechanism

that limits PGS-2 protein levels in fibroblasts (DeWitt et al., 1993). What accounts for the

different protein stability of PGS-1 and PGS-2 is not known, but increased protein

turnover of PGS-2 may be mediated via the carboxyl-terminal protein sequences that are

unique to PGS-2.

Sex Hormone. The Immune Response and Autoimmunity

Human autoimmune diseases such as systemic lupus erythematosus (SLE),

rheumatoid arthritis (RA), and multiple sclerosis are characterized by a disproportionate

effect on women with a marked increase in disease post pubertal peroid (Cutolo et al.,

1988; 1986). Animal models of human autoimmune disease similarly reflect a female

bias, i.e. in NZB/W mice, a model of SLE, and in the NOD mouse. The higher incidence

of disease in females suggests that higher concentrations of sex steroids in females, or

their hormonal cycle may strongly influence autoimmune immunopathogenesis.

Conversely, male sex steroids appear to inhibit expression of autoimmune disease (Fox,

1992; Homo-Delarch et al., 1991; Wilder, 1995).

In the NOD mouse, there is a marked difference in the incidence and the onset of

diabetes between male and female mice with female NOD mice developing diabetes at

three times the rate of their male counterparts. The influence of sex steroid hormones has








been demonstrated by exacerbation of disease in orchiectomized or estrogen treated NOD

male mice, and a reduction in insulitis and diabetes incidence following oopherectomy or

treatment with androgens in female mice (Fox, 1992; Hawkins et al., 1993; Fitzpatrick et

al., 1991). Although experimental studies both in vivo and in vitro have confirmed the

influence of sex hormones on immunoreactivity, the mechanisms of their influence have

yet to be fully elucidated.

Studies have shown that hormone receptors are present in many cells involved in

the immune response and the concentration of hormone levels play a major role in the

regulation of the immune response. Evidence has accumulated to support the concept that

female sex steroid hormones have a regulatory role in macrophage gene expression and

effector function (Zhang et al., 1988; Wang et al., 1988). In vivo regulation may be either

indirect or direct. For example, macrophages migrate into the estrogen-stimulated mouse

uterus (Zhang et al., 1988), possibly as an indirect response to chemoattractive cytokines

such as colony-stimulating factor-1 and granulocyte-macrophage colony-stimulating

factor produced by estrogen-targeted uterine cells (Wang et al., 1988). However, in vitro

studies show that estrogen and progesterone have direct effects as well. Sex hormones

may inhibit or stimulate, in a dose-dependent manner, the Fc-mediated clearance of

antibody-coated erythrocytes by guinea pig spleen macrophages (Schreiber et al., 1988)

and may regulate expression of MHC class II (la) antigen and interleukin- lp (IL- 1)

protein synthesis by mammalian monocytes, macrophages, and macrophage cell lines

(Polan et al., 1989). The effects exerted by estrogen on MP IL-1 synthesis seem to be

biphasic and dose-dependent. Polan et al. (1989) reported that a negative relationship has








been found between IL-1 mRNA levels and estrogen concentrations in cultured human

peripheral monocytes and pelvic macrophages. In particular, IL-1 mRNA levels

decreased by 80-90% as the measured estradiol concentrations increased from 10-9 to 10"

5 M. Thus, low estrogen level (physiological, 10-7 M) stimulates both IL-1 mRNA levels

and IL-1 protein secretion, whereas higher levels (pharmacological, 10-6 to 10'5 M) are

inhibitory. Levels of circulating estradiol-171 (E2) and progesterone (P) in female mice

during the estrus cycle and pregnancy has been reported to reach 60.0 pg/ml (-10-6 M)

and 39.0 pg/ml (-10"7 M), respectively. In human placentas, P levels can vary from 0.5

ug/ml to 5.1 ug/ml. Recently, Miller et al. (1996) found that concentrations greater than

0.1 ug/ml of P inhibited iNOS gene activity and NO production, while E2 had no effect

on iNOS gene activity and NO production in IFN-y/LPS-stimulated macrophages.

It has been demonstrated by many investigators that prostaglandin E2, which is

one of the major secretary products of monocytes, plays an important role in the immune

response. El Attar et al. (1982) studied the effects of sex hormones on the production of

PGE-2 using human gingival tissue and demonstrated that addition of estradiol or

progesterone significantly enhanced the synthesis of PGE-2. They also reported that

inflamed gingiva produced more PGE-2 than healthy gingiva and speculated that

inflammatory cells in the tissue were responsible for the increased level of PGE-2.

Miyaga et al. (1993) investigated the effect of sex hormones on the production of PGE-2

by LPS-stimulated human monocytes and showed that testosterone reduced the

production of PGE-2, while progesterone enhanced it. Estradiol showed a bi-directional

effect on PGE-2 production; that is, inhibitory at 0.4 ng/ml and stimulatory at 29 ng/ml.








Thus, the concentration of estradiol seems to be an important factor for this effect. They

also found that the reduced PGE-2 production by monocytes treated with low amounts of

estradiol was restored in the presence of high amounts of progesterone, while the

enhancing effects of progesterone on PGE-2 production were reduced by addition of low

amounts of estradiol. These results suggest that a balance in the combination of sex

hormones modulates PGE-2 production by monocytes and macrophages in a female

hormonal milieu.

The mechanism by which sex hormones affect MPs production of PGE-2 remains

unclear. Landers et al. (1992) proposed the following general model for steroid action:

Free steroids passively diffuse into all cells but are preferentially retained in target cells

through the formation of a high affinity complex with the steroid-specific receptor.

Binding of the steroid results in an "activation" of the receptor molecule that appears to

involve conformational and post-translational changes in the receptor itself, as well as

changes in the protein-protein associations in the receptor complex. Finally, the activated

complex (steroid-receptor) binds with high affinity to specific DNA sequences, termed

"steroid response elements" (SRE). The steroid response element acts as a transcription

factors, modulating the rate of transcription of steroid-responsive genes or altering the

post-transcriptional steps which, in turn, results in a change in the steady state levels of

specific messenger RNAs (mRNAs) (Landers et al., 1992).

Regulation of Lymphocyte Functions by Nitric Oxide

Nitric oxide synthase (NOS) catalyzes the conversion of L-arginine to L-citrulline

and nitric oxide (NO) (Marletta et al., 1993). NO possesses diverse activities and is








considered to play an important role in many physiological functions and pathological

conditions. Notably, it is a potent vascular mediator which maintains vascular

homeostasis via its actions on platelets and vascular tone. It also acts as a

neurotransmitter for central and peripheral nervous system function and as a nonspecific

immune modulator involved in controlling the invasion of microorganisms and tumors

(Moncada et al., 1991). Synthesis of this ubiquitous, diffusible gas in many tissues is

catalyzed by isoforms of NO synthase (Marietta et al., 1994). As with PGS, two major

types of NOS have been identified (Marietta et al., 1994). One is the calcium-dependent

form, which is present constitutively in a variety of tissue and produces the physiological

concentration of NO needed for 'house-keeping'. Another one is not consititutively

expressed, but can be induced in a number of cell types, including macrophages,

hepatocytes, neutrophils, muscle, and endothelium, via a variety of immunological

stimuli such as interferon-y, tumor necrosis factor, and LPS. Once induced, these cells

produce large amounts of NO, which may be cytotoxic. Thus, apart from maintaining

normal physiological function, NO is required in large amounts to combat infectious

organism and tumors. Production of excessive amounts of NO will, however, lead to a

different range of pathological outcomes and important pathologies. Therefore, the

expression of inducible NOS (iNOS) is necessarily under tight regulation.

Lipopolysaccharide and IFN-y response elements have been located in the promoter

region of iNOS gene (Marietta, 1994). A number of cytokines are able to inhibit the

expression of iNOS by murine macrophages. These include TGF-P (Ding et al., 1990),

IL-4 (Liew et al., 1991), and IL-10 (Bogdan et al., 1991). Glucocorticoids can also








suppress iNOS expression. Activation of PTK and PKC, in turn, activates iNOS, while

inhibition of these kinases blocks NO production (Marietta., 1994; Bogdan et al., 1991).

NO produced by MPs plays an important role in affecting and modulating the

immune response. NO, like PG, is also a potent inhibitor of T cell activation (Liew et al,

1991). Hoffman et al. (1990) first reported that NOS activity was detected during

phytohaemagglutinin (PHA)-stimulated proliferation of rat spleen cells and also in a

mixed lymphocyte response, and that the addition of NOS inhibitor, N-monomethy-L-

arginine (NMMA), to cultures suppressed NOS activity and allowed a robust proliferative

response to occur. These findings have since been extended to both in vitro and in vivo

marine models, indicating that NO inhibits the proliferation of T cells. Taylor-Robinson

et al. (1994) demonstrated that the inhibition of proliferation of Thl cells by NO can be

reversed by the addition of exogenous IL-2, suggesting that NO inhibits the expansion of

Thl cells by blocking the secretion of IL-2, which is an autocrine mediator of T cell

expansion. Another way NO can influence T-cell function is through its modulation of

antigen presentation by down-regulation of the expression of class II MHC molecules on

antigen-presenting cells (Sicher et al., 1994). At present, the mechanism of such down

regulation is unknown. When a panel of cloned T cells specific for malaria antigens was

examined, it appeared that only Thl cells could be induced to produce NO, whereas Th2

cells could not (Taylor-Robinson et al., 1994). This suggests low levels of NO produced

by Thl together with the NO produced by activated macrophages, may inhibit Thl

proliferation by blocking the synthesis of IL-2. It is likely that NO at a physiological

concentration (provided constitutively) is required for proliferation of T cells (and indeed








for that of other cells). At the higher concentrations, however, NO inhibits cellular

proliferation.

NO plays an important role in the immunopathogenesis of murine autoimmune

diseases including IDD, arthritis, and SLE. Treatment of SLE prone mice with the NO

inhibitors, such as the arginine analogue NMMA or aminoguandine (AG), reduces

clinical autoimmune disease (Weinberg et al., 1994). Increased NO production and iNOS

expression has been shown in MRL-lpr/lpr mouse (Wu et al., 1994). Nitric oxide also

contributes directly to the pathogenesis of autoimmune disease by causing an increase in

vascular permeability. Cell-derived NO, through interaction with superoxides forms

peroxynitrite may spontaneously produce hydroxyl radicals. These radicals are highly

reactive and readily cause cell and tissue injury and destruction. Evidence has shown that

cytokines, like IL-1, released in islets by nonendocrine cells, most probably MPs can

induce the expression of iNOS by (3 cells and inhibit their insulin secretion (Corbet et al.,

1995).

iNOS and PGS-2 Crosstalk

In several physiological and pathological conditions, NO and prostanoids work

synergistically. A recent report indicates that the inducible forms of PGS (PGS-2) and

NOS (iNOS) are concurrently induced in inflammatory tissues in experimental animals

(Vane et al., 1994). Although the inducing agents were not investigated in the

experimental model, inflammatory cytokines, most notably IL-IP, have been shown to

induce PGS-2 and iNOS in cultured cells.








PGS-2 and iNOS may also interact via their metabolites. Several reports have

shown that NO increases prostanoid synthesis (Salvemini et al., 1993; Franchi et al.,

1994). Although one study suggested that NO directly activates PGS-2 enzyme activity

(Salvemini et al., 1993), further work is needed to determine whether the effect of NO is

mediated by its direct activation of PGS-2 or via an intermediate step (Marshall et al.,

1987). Conversely, PGS-2 metabolites have been reported to influence iNOS induction.

A recent report showed that PGE-2 and iloprost, a PGI-2 stable analog, at micromolar

concentrations, suppressed iNOS induction in an murine macrophage cell line, J774,

whereas PGF-2a and lipoxygenase metabolites had no effect (Morotta et al., 1992).

In summary, there exists an intricate relationship between PGS-2 and iNOS under

physiological and pathological conditions. Synergistic interactions of these two enzymes

and their metabolites may play an important role in modulating the immune response and

inflammation.

Rationale for these Studies

In NOD mice, both T cells and macrophages (MPs) play an important role in the

immunopathogenesis of diabetes. Several studies have suggested that antigen presenting

cells from autoimmune humans and animals including NOD mice, are defective in their

capacity to active T cells (Serreze et al., 1988; Yokono et al., 1989). The cellular and

molecular basis for this defect, however, is largely unknown. Studies from our lab have

suggested a cellular basis for this defect, as NOD MPs, unlike those of control strains,

suppress T cell activation through the production of prostaglandin and nitric oxide.

Furthermore, I have established that NOD MPs display an "activation" phenotype, as








resident peritoneal MPs from NOD, unlike MPs from control MPs spontaneously and

constitutively express an early response gene, prostaglandin synthase II (PGS-2).

PGS-2 is the rate limiting enzyme in prostanoid metabolism and responsible for

the production of large quantities of PGs, therefore, I hypothesized that constitutive

expression of PGS-2 may result in enhanced production of PGE-2 which may disturb

peripheral tolerance in NOD mice through impair activation induced cell death.

Activation induced cell death (AICD) is an important mechanism for peripheral T

cell tolerance. Indeed, impaired AICD, as in the lpr/lpr mouse, contributes strongly to the

development of autoimmunity. Several studies have suggest that AICD can be markedly

inhibited by anti-IL-2 antibodies and by agents that block the cell cycle and IL-2

production/signaling, i.e., cAMP (Boehme et al., 1993; Critchfield et al., 1995). Of

interest, PGE-2 is a potent inhibitor of IL-2 and IL-2 receptor expression and can block

the cell cycle through its activation of adenylate cyclase and generation of cAMP (Goetzl

et al., 1995; Ucker et al., 1994). The potentially enhanced PGE-2 production in NOD

MPs may therefore impair AICD in NOD mice and contribute to the pathogenesis of

diabetes.

The overall objective of this proposal is to establish the role of PGS-2 and PGs in

the immunopathogenesis of NOD diabetes. The specific aims include: 1) understanding

the mechanism of the aberrant PGS-2 expression, 2) determination of the correlation

between PGS-2 phenotype and antoimmunity phenotype and evaluating the role of PGS-2

as a candidate gene for Idd5, and 3) establishing a potential mechanism and the role of

PGS-2 expression in impaired peripheral tolerance in NOD mice.








To fulfill these aims, I investigated 1) PGS-2 expression in the autoimmunity

environment, 2) prostaglandin production in NOD MPs, 3) possible sequence variations

of cis-elements of NOD PGS-2 promoter, 4) the influence of sexual hormones on PGS-2

expression, and 5) monokines regulation of NOD PGS-2 expression.

I also analyzed PGS-2 phenotype of different chromosome 1 congenic mice and

compared PGS-2 expression with the autoimmune phenotype in these congenic mice

which help me elucidate the role of PGS-2 as a candidate gene for Idd5.

To further investigate the role of PGS-2 in the pathogenesis of diabetes, we used

drugs to block PGS-2 activity and examined the effect of drug treatment on the onset and

incidence of diabetes.

Finally, I established a potential mechanism for PGS-2 expression in the

impairment of immune tolerance by using Staphyloccus aureus enterotoxin B (SEB)

immunization as a model to examine AICD in NOD mice.

From these studies, I elucidated the role of PGS-2 expression in NOD diabetes

and have derived a better understanding of the pathogenesis of this autoimmune disease.













CHAPTER 2
MATERIALS AND METHODS

Animals

Female and male NOD mice were purchased from the Taconic Laboratory

(Germantown, NY) and BALB/c and B57BL/6 mice were purchased from the Jackson

Laboratory (Bar Harbor, ME) for experiments conducted in the first and the second years.

The NOD scid/scid mice, B6.NODC1 congenic mice, and NOD.B10C1, NOD.B10C3,

NOD.B10C4 congenic mice were gifts from our collaborators: Dr. Mark Atkinson

(University of Florida, Gainesville, FL), Dr. Edward Wakeland (University of Florida,

Gainesville, FL), and Dr. Linder Wicker (Merck Laboratory, Rahway, NJ). NOD,

BALB/c, C57BL/6 and NOD.H-2b congenic mice used for experiments in this year were

bred and maintained in the animal facility of the Department of Pathology and Laboratory

Medicine at the University of Florida, Gainesville. All the mice, except B6.NODCl

congenic strain, were housed in SPF conditions. Estrus phase of female mice was

determined by vaginal cytology.

For the first set of PGS-2 inhibition experiments, NOD female mice were treated

from 8 weeks to 32 weeks of age with indomethacin (3 ug/ml) and aminoguanidine

(0.1%) in their drinking water. In a second study, NOD female mice were treated from 4

weeks to 22 weeks of age with indomethacin (20 ug/ml) in their drinking water. Water

was changed every week and the drugs freshly added at each change. Mice were housed








in SPF conditions during the entire experimental peroid. Development of diabetes was

monitored weekly by a urine glucose test and confirmed by tail vain blood glucose (250

mg/dl glucose was used as indication of diabetes).

For the SEB immunization, 6-8 week old female mice were given an

intraperitoneal injection of buffed saline or SEB (12.5 ug or 50 ug). Spleen cells were

examined at day 2 and day 10 after immunization.

Antibodies

PGS-2 specific antisera was produced in Dr. Harvey Herschman's lab. Antibodies

for the FACS analysis including FITC labeled anti-mouse VP 6 TCR and VP 8.1, 8.2

TCR, PE labeled anti-mouse CD4 and relevant isotype control antibodies were purchased

from Pharmingen Bio. Inc. (San Diego, CA).

Peritoneal Cell Preparation and Culture

Mice were sacrificed by cervical dislocation and cells were obtained by washing

the peritoneal cavity with 5 ml of cold 10% endotoxin-free FBS-supplemented RPMI

1640. For mRNA determination, adherent resident peritoneal MPs were isolated by

incubating Ix105 MPs with RPMI 1640 (GIBCO, Grand Island, NY) which containing 2

mM glutamine, 100 u/ml penicillin, and 100 ug/ml streptomycin, plus heat treated 10%

FBS (HyClone Laboratories, Logan, UT) for 2-3 hr in 96 well plates (Costar Corporation,

Cambridge, MA) at 370C in a 5% CO2 atmosphere. Cells were washed three times with

same media and cultured for 16 hr alone or with stimulation: LPS (10 ug/ml) or

hormones at physiological concentrations. Hormones used were: estradiol (10-6 M),

progesterone (10"7 M), testosterone (10"7 M), dehydroepiandrosterone (DHEA, 10"7 M),








and dehydroepiandrosterone sulfate (DHEAS, 10-7 M). MPs were released from plate by

adding 50 ul ice cold EDTA-PBS buffer (0.02% EDTA) to the MPs and incubating for 15

min on the ice. For castration experiments, NOD males were castrated at 8 weeks of age.

Four weeks after castration, MPs were collected as described above, cultured overnight,

then processed for mRNA and protein analysis by RT-PCR and immunofluorescence

assay.

Spleen T cell Preparation and Culture

Single cell suspensions of splenic leukocytes were obtained by gently pressing

freshly explanted spleens through wire mesh screens followed by a single wash in media.

Erythrocytes were lysed in a 0.84% ammonium chloride treatment. The remaining

leukocytes were washed twice with media. Purified T lymphocytes were obtained by

passing suspensions of spleen cells through a nylon wool fiber column. T cell purification

columns were pre-prepared by saturating and washing 0.6 g nylon wool fiber with 50 ml

of media in a 12 cc syringe followed by incubation with media for 1 hour at 370C in a 5%

CO2 atmosphere, and then washed with 20 ml media again. Approximately 3 ml

(40~150x106 cells) cells were added to the column, allowed to penetrate into column,

and incubated at 370C in a 5% CO2 atmosphere for 45 min. Cells were then collected in

10 ml of media. For MP and T cell co-culture, 5x105 T cells were added to xl0s5 MPs

cultured in 96 well culture plates.








Reverse Transcriptase/Polymerase Chain Reaction (RT-PCR) Analysis
of PGS-2 and other Cvtokine mRNA

Poly (A+) mRNA from Ixl05 MPs, Ixl04 DCs, 5x105 T cells, or same number of

cells from T/MP, T/DC, T/MP/DC co-culture were obtained by using the Fast Track

mRNA isolation kit (Invitrogen, San Diego, CA). Isolated mRNA was reverse transcribed

at 42C for 1 hr in 20ul reaction mixtures containing 400 units of Moloney Murine

Leukemia Virus reverse transcriptase. The resultant cDNAs were subjected to PCR (35

cycles in a Gene Machine II, programmable thermal cycler, USA/Scientific Plastics,

Ocala, FL). The oligonucleotide PCR primers specific for PGS-2 were derived from exon

9 (5'-primer, 5'-CAAGCAGTGGCAAAGGCCTCCA-3') and exon 10 (3'-primer, 5'-

GGAACTTGCATTGATGGTGGCT-3'). G3PDH primers were used as an assay control.

Each PCR cycle consisted of denaturation for 45 seconds at 94C, annealing for 45

seconds at 60C and extension at 72C for 2 minutes, followed by 7 minutes at 720C for

final extension. The 50 ul of reaction mixture contained 50 pM of each primer, 1.25 U of

AmpliTaq DNA polymerase (Boehringer Mannheim, Germany) and 2 ul of cDNA.

Fifteen microliters of each 50 ul PCR reaction were resolved by electrophoresis through a

1.4% agarose gel that was subsequently stained with ethidium bromide. PCR results were

confirmed by Southern blotting using a digoxigenin labeled internal probe.

Oligonucleotides derived from PGS-2 exon 10 (5'-GTGCTCCAAGCTCTACCA-

3') were labeled with digoxigenin-labeled uridine-triphosphate (Dig-UTP) using the

enzyme terminal transferase as per instructions from Boehringer Mannheim. Probes were

hybridized to positively charged nylon membranes to which PCR gels had been blotted.

Detection of hybridized probes revolves around the use of an anti-DIG alkaline








phosphatase conjugate (DIG-AP). After DIG-AP is reacted with any hybridized probes, a

subsequent AP-catalyzed color reaction with 5-bromo-4-chloro-3-indoyl phosphate (x-

phosphate) and nitroblue tetrazolium salt (NBT) as substrate to produces an insoluble

blue precipitate. The presence of blue bands is indicative of the presence of probe-PCR

product hybrid molecules.

The primer and probe sequences of other cytokines were designed from published

sequences (Anderson, 1991). All primers and probes were synthesized in the University

of Florida ICBR DNA Synthesis Laboratory (Gainesville, FL).

Immunofluorescence Analysis of PGS-2 Protein

MPs (Ix105) were purified in multi-chamber culture slides by the method

described above. After 16 hr of culture, the MPs were fixed in 2% paraformaldehyde/PBS

for 30 min. The fixed cultures were rinsed with PBS and further washed in PBS-GT

(0.1M glycine, 0.05% Triton X-100). MPs were incubated in normal goat serum (1:20

dilution) for 30 minutes, then overnight at 40C with anti-PGS-2 antibody (1:250 diluted

in PBS- Tween (0.2%)). After three 10-min washes with PBS-T, the cultures were

washed three times for 20 min each, dried, mounted in buffered glycerol containing 1

mg/ml paraphenylenediamine, and analyzed with a Zeiss Photomicroscope III. Exposures

were for 8 sec in all cases, using Kodak Tri-X Pan 400 film. Magnifications were 100x,

in all cases (Reddy et al., 1994).

SDS-Polyacrylamide Gel Electrophoresis and Western Analysis

Fresh peritoneal MPs were obtained by method described above. Cells were

washed three times with PBS, and harvested in SDS gel loading buffer ( 50 mM Tris-







HCL (pH 6.8), 100 mM dithiothreitol, 2% SDS, 0.1% bromphenol blue, 10% glycerol).

Samples were incubated for 10 min at 1000C, sonicated, and protein concentration was

determined. 75 gg of the protein extract was analyzed on SDS-polyacrylamide

electrophoresis (5% stacking and 8% resolving gel), using a tris-glycine buffer (pH 8.3).

The proteins were electroblotted using BioRad's Mini Protein II according to the

manufacturer's procedures. Filters were then incubated in PBS, 0.2% Tween 20, 10%

non-fat milk for 60 min, washed with PBS buffer, and incubated with primary and

secondary antibodies as suggested in the manufacturer's protocols. Dilution's of 1:6000

and 1:8000 for primary and secondary antibodies, respectively were sufficient for

immunodetection using the enhanced chemiluminescence reagents. The filters were

exposed to Kodak XAR-5 film at room temperature (Reddy et al., 1994).

PGE-2 Analysis

Peritoneal MPs (xl105) or Peritoneal MPs (1x105) plus T cells (5xl05) were

cultured for 24 hours or time specified with or without NS-398 and in the presence or

absence of LPS (10 ug/ml) in RPMI 1640 and 2% endotoxin free fetal bovine sera culture

media. Supernatants were harvested for PGE-2 analysis by using a PGE-2 specific EIA

kit (Cayman Chemical). Results are expressed as pg/ml PGE-2/105 MPs.

Spleen Cell Preparations and Flow Cytometric Analysis

Single cell suspensions of splenic leukocytes from SEB immunized NOD,

BALB/c and C57BL/6 mice were obtained as described above. The collected leukocytes

were washed twice with media, then, lxl06 cells were resuspended into FACS buffer

(lxPBS with 1% bovine serum album and 0.1 % sodium azide) and labeled with either








FITC conjugated TCR Vp6 or 8 specific antibodies (Pharmingen, San Diego, CA). After

45' min at 4*C, the cells were, and then incubated with PE conjugated anti-CD4 antibody

(Pharmingen, San Diego, CA) at 4C for another 45 min. Then, cells were washed in

FACS buffer and used for two color FACS analysis. Single color FACS using Vp6 TCR,

Vp8 TCR, CD4 antibodies were also analyzed. FITC and PE labeled isotype matched

antibodies were used as controls.

For FACS, viable cells (10,000/analysis) were gated using a combination of

forward and side scatters and analyzed on a FACScan Flow Cytometer (Becton-

Dickinson, Mountain View, CA) using logarithmic scales. TCR Vi8+/CD4+ and

VI6+/CD4+ double positive cells were determined by two color fluorescent analysis.

Data were analyzed using PC Lysis software. Results are expressed as the percentage of

VP 8+/CD4+ or VP 6+/CD4+ cells divided by total CD4+ cells.

Syngenic Mixed Lymphocyte Reaction (SMLR) Assay

Splenic cells as antigen presenting cells and purified splenic T cells were isolated

as described above. Single cell suspensions of dendritic cells were prepared by passage

axillary and inguinal lymph node through Nitex 110 mesh, then purified by metrizamide

gradient. MPs were isolated by overnight adherence. B cells were isolated by using

mouse anti-IgG coupled magnetic beads.

SMLR responses were determined in triplicate in flat-bottomed 96-well, in a final

volume of 0.2 ml/well, containing 5 x 105 nylon wool-purified T cell/well cultured in

medium alone or with DCs (1 x 104), MPs (1 x 105), or splenic cells (APCs, 1 x 105) with

or without MPs or B cells. Splenic APCs (with or without MPs and B cells) were








irradiated at 1500 Rads before used for co-culture. Microtiter plates were incubated for 5

days at 370C in a 5% C02/95% air humidified atmosphere. For the final 18 h of culture,

each well was incubated in the presence of 0.5 gCi/well of [3H] thymidine (New England

Nuclear, Boston, MA) and cultures were harvested onto glass fiber disks by a cell

harvester (Cambridge Technology, Cambridge, MA). The disks were then counted in 3

ml of Aquasol-2 (New England Nuclear). Data are presented as the mean cpm of

triplicate co-cultures of T cells and DCs, MPs, or APCs divided the mean cpm of

triplicate T cell cultured alone.

PGS-2 Promoter Sequence Analysis

Genomic DNA was extracted from 2x106 rest peritoneal macrophages of NOD

and BALB/c mice using Qiagen genomic DNA isolation kit (Qiagen Inc., Chatsworth,

CA). Two microliters genomic DNA from 20 ul total was subjected to PCR (35 cycles in

a Gene Machine II, programmable thermal cycler, USA/Scientific Plastics, Ocala, FL).

Two pairs of overlapping oligonucleotide PCR primers specific for the PGS-2 promoter

were used. The first pair of primers: 5'-primer, 5'-GGCCAACACAAACACAGTAGG-

3', 3'-primer, 5'-TCCCTCCCGGGATCTAAG-3' covers -1 to -600 bp of the PGS-2

promoter sequence; and the second pair of primers: 5'-primer, 5'-

AGTGGGGAGAGGTGAGGG-3', 3'-primer, 5'-TCTTTGCCAATACACAGCCA-3'

covers -500 to -1000 bp PGS-2 promoter sequence. Together, these primers covered the

5' promoter region from -996 nucleotide to the transcriptional start site). Each PCR cycle

consisted of denaturation for 45 seconds at 940C, annealing for 45 seconds at 60C and

extension at 72C for 2 minutes, followed by 7 minutes at 72C after all cycles were








finished. The 50ul of reaction mixture contained 50 pM of each primer, 1.25 U of

AmpliTaq DNA polymerase (Boehringer Mannheim, Germany) and 2ul of genomic

DNA. PCR products were purified by 1.4% agarose gel electrophoresis and sequenced at

the University of Florida sequence core laboratory (Gainesville, FL) by using the cycle

sequencing procedure and an ABU 373 automated sequencer.

Statistical Analyses

All measures of variance are given as SEM of the mean. Tests of significance for

difference in SMLR response and PGE-2 production in different group were performed

with the one-way analysis of variance (ANOVA). Results in which p< 0.05 were

considered significant. Test of significant for difference in incidence of diabetes between

NOD mice group treated with indomethacin/aminoguanidine or indomethacin alone and

untreated NOD mice group were performed with logrank test chi-square. p< 0.05 were

considered significant. p values are one sided.












CHAPTER 3
RESULTS

Effects of Macrophages on Dendritic Cell-Mediated T cell Activation
in NOD Mice

The syngeneic mixed lymphocyte reaction (SMLR) has been characterized as an

in vitro activation of CD4+ T lymphocytes recognizing class II molecules on the surface

of syngeneic antigen presenting cells (APC) (Glimcher et al, 1981). The marine SMLR,

provides a model system for studying immunoregulatory mechanisms in vitro. Previous

studies in our lab and in others have shown that there is a depressed syngeneic mixed

lymphocyte reaction in NOD mice (Serreze et al, 1988) and a defect in the APC

stimulation of T cell proliferation has also been suggested (Yokona et al, 1989). In those

experiments, splenic cells were used as APCs to stimulate syngeneic T cells. Spleen cells,

however, contain a mixture of APC sub-populations including dendritic cells (DCs),

macrophages (MPs), and B cells. To investigate which sub-population of APCs

specifically contributed to the decreased ability of APCs to activate T cells in NOD mice,

we purified APC populations and stimulated syngeneic T cells. We established that 5

days of culture is the peak response time for the SMLR stimulated by DC, MPs, and B

cells. As shown in figure 3, we first demonstrated that the suppression of SMLR is not

due to DC dysfunction as we found that lymph node (LN) DCs from NOD mice were

paradoxically better stimulators of syngenic T cell response than DCs from control mice.

Also NOD MPs and purified spleen B cells are similar to control strains in their capacity








to stimulate T cells. However, the mixed spleen APC populations are deficient,

suggesting the interaction of sub-populations of APCs may disturb the ability of other

APCs to activate T cells. We therefore subtracted B cells by anti-IgG magnetic bead

separation and MPs by overnight adhesion. We found that the removal of the MP

population from splenic cells markedly increased the stimulatory capacity of splenic cells

to T cells in NOD mice, but not in control strain BALB/c mice (figure 4). Furthermore,

when MPs were added back to DCs and T cells, marked suppression of the SMLR, once

again occurred (figure 5). In transwell experiments, we established that the effect of MPs

was mediated through soluble factors and have gone on to demonstrate that NO and PG

are the chief mediators of T cell suppression as it is reversed completely by the NO and

PG inhibitors NMMA and indomethacin (figure 5). These results demonstrated that MPs

suppress syngeneic T cell activation when in the environment of a strong immune

response such as that provided by DCs.

To further analyze the T cell activation and APC contribution in SMLR, IL- I3,

IL-2, IL-4, IL-6, IL-7, IL-10, IL-12, TGF-P, TNF-a, IFN-y, iNOS, and PGS-2 mRNA

expression were analyzed in MPs (xl105 cells), DC (Ixl04 cells), T cells (5xl05 cells), T

(5x105 cells)/MPs (Ixl05 cells), T (5x105 cells)/DCs (lxl04 cells), and T (5xl05

cells)/DCs (Ixl04 cells)/MPs (Ixl05 cells) following 24 hr cell culture by RT-PCR assay.

For T/DC/MP co-culture, mRNA expression was assessed in the presence or absence of

indomethacin (I) and N-monomethyl-L-arginine (NMMA).

After examining different mRNA expression in the single cell population culture

or combined cell population co-culture, it was found that cells from the BALB/c T cells








(5x10s cells)/DC ( x 104) cells)/MPs (Ix105 cells) co-culture expressed IL-2, IL-4, IL-12,

TNF-a, IFNy, TGF-p, but not PGS-2 and iNOS mRNA (Table 1). In contrast, in the same

culture system, cells from NOD mice expressed IL-103, PGS-2 and iNOS (Table 2). The

absence of IL-2/IL-4 expression in NOD is consistent with previous studies (Serreze et

al., 1993), and high levels of PGS-2, iNOS, and IL- 13 expression in NOD suggests the

MPs are activated in SMLR. Studies using NOD DC/MP/T co-cultures also showed that

addition of PG and NO inhibitors enhanced IL-2 expression and suggests that PG/NO

may suppress T cell activation through down-regulation of IL-2 or IL-2R expression.

The key finding in these experiments, however, was the differential expression of

PGS-2 in MPs of NOD and control strain mice. After examining purified resting

peritoneal MPs, I found that PGS-2 is consititutively expressed in MPs of female NOD

mice but not in BALB/c MPs. The constitutive expression of PGS-2 in NOD MPs

strongly suggested that the NOD MPs are "activated" as this enzyme is reported to be

expressed only in activated Mps.

Spontaneous PGS-2 Expression in Macrophage of NOD
and NODscid/scid Female Mice

In above mentioned experiments, I found that PGS-2 mRNA is consititutively

expressed in NOD MPs but not in MPs of control mice. However, PGS-2 mRNA

expression can be induced successfully in control MPs by overnight LPS stimulation

(figure 6). Although endotoxin contamination could upregulate PGS-2 expression in

NOD mouse, this is unlikely as BALB/c MPs cultured in the same media did not express

PGS-2. Additional proof that NOD PGS-2 expression is not due to endotoxin









50 ENOD
45 *BALBIC
40 F1(NOOBALB/c)
C57BL/U6

30
o 25
20






4wk 8wk


Figure 3. Lymph node dendritic cells (DCs) stimulation of syngeneic T cell response. LN
DCs (1xl04) from 4 wk and 8 wk old female NOD and control mice were purified by
metrizamide gradient and used to stimulate nylon wool purified splenic T cells (5x105).
DCs/T cells were co-cultured for 5 days and proliferation was assessed by [ H]thymidine
uptake in the last 18 hr of culture. Results are expressed as mean value of stimulation
index SEM (n=10 experiments, p< 0.05) (Stimulation index is the mean value of the
cpm from triplicate DC/T cultures divided by mean value of cpm from T cells alone).











4 3.5
BALBIc

0 3.5 ____3.5______________
:3
S2.5
.5 2
S1.5

S0.5
0

APC-MPs APC-B Cells


Figure 4. Syngeneic T cell response stimulated by antigen presenting cell (APC) without
macrophage (MP) or B cell. The subtraction of MPs from the spleen by overnight
adherence markedly enhances the stimulation of T cells from the 8 week old NOD
female mice, but not from spleen cells of age matched control BALB/c female mice. B
cell subtraction by mouse anti-IgG coupled magnetic beads did not affect T cell response
in either strain. APCs (1x105 cells) without MPs and B cells were irradiated at 1500 Rads
and then cultured with 5x105 nylon wool purified T cells. APCs/T cells were co-cultured
for 5 days and proliferation was assessed by [3H]thymidine uptake in the last 18 hr of
culture. Results are expressed as mean value of the fold increase of stimulation index
SEM (n=10 experiments, p< 0.05) (fold increase of stimulation index is expressed as
stimulation index of splenic APCs without MPs or B cells divided by stimulation index
of intact splenic APCs































Figure 5. Effects of iNOS and cyclooxygenase inhibitors on the MP suppression of
SMLR in 8 week old female NOD mice. Metrizamide gradient purified LN DCs (lxlO4)
and overnight adherence purified MPs (xl 0s) alone or combinations were used to
stimulate splenic T (5x105) cell activation in the presence or absence of 1 mM
indomethacin (I) and/or N-monomethyl-L-arginine (NMMA). Cells were cultured for 5
days and proliferation was assessed by [3H]thymidine uptake in the last 18 h of culture.
Results are expressed as mean value of stimulation index SEM (n=10 experiments, p<
0.05).


40
35
30
S2s
o 20
" 15
E
U 10
5
0


So i
a z
04 0Z


0z
SZ








Table 1. Summary data of cytokine mRNA expression in cell cultures of 6-8 week old
female BALB/c mice. Lymph node dendritic cells (DC, lx104), peritoneal macrophages
(MP, 1x10s) and splenic T cells (T, 5xl05) by themselves or in combination were
cultured for 24 hours. DC/MP/T were cultured in the presence or absence of
indomethacin (I) (13 ug/ml) and N-monomethyl-L-arginine (NMMA) (250 ug/ml). After
24 hr culture, cells were harvested and mRNA were extracted by using mRNA isolation
kit and amplified by RT-PCR followed by Southern blotting using a digoxigenin labeled
cytokine specific internal probe. (N=2 set of experiments, "++" : PCR product visible on
the gel (further verified by blot), "+" : PCR product visible on the blot, "-" : PCR product
undetectable on the blot)

T DC MP T/DC T/MP DC/MP/T DC/MP/T DC/MP/T DC/MP/T

+1 +NMMA +I+NMMA
3-actin ++ ++ ++ + ++ ++

IL-2 + + + + + +

IL-4 + + + +

IL-12 + + + +

IFN-y + + + ++ + + +

TNF-a + + + + +

TGF-3 + + + + + + +

IL-lp + + +

iNOS

PGS-2 -





50

Table 2. Summary data of cytokine mRNA expression in cell cultures of 6-8 week old
female NOD mice. Lymph node dendritic cells (DC, Ix104), peritoneal macrophages
(MP, lxl0s) and splenic T cells (T, 5x105) by themselves or in combination were
cultured for 24 hours. DC/MP/T were cultured in the presence or absence of
indomethacin (I) (13 ug/ml) and N-monomethyl-L-arginine (NMMA) (250 ug/ml). After
24 hr culture, cells were harvested and mRNA were extracted by using mRNA isolation
kit and amplified by RT-PCR followed by Southern blotting using a digoxigenin labeled
cytokine specific internal probe. (N=3 set of experiments, "++" : PCR product visible on
the gel (further verified by blot), "+" : PCR product visible on the blot, "-" : PCR product
undetectable on the blot)








contamination comes from experiment where polymixin B, a specific endotoxin inhibitor,

was added to NOD MPs cell cultures and it did not block PGS-2 expression. In order to

established that the gene is truly expressed constitutively, I evaluated the kinetics of PGS-

2 mRNA expression in MPs of NOD and control strains. I harvested mRNA from cells

immediately after peritoneal lavage, as well as after 1, 2, 6, 12, 24, and 48 hr of cell

culture. I found that PGS-2 is expressed at all time points in the MPs of NOD mice, but is

not expressed at any time points in the MPs of control strains (BALB/c and C57BL/6

mice).

PGS-2 protein expression was also examined in the resident peritoneal MPs of 6-

8 weeks old female NOD, BALB/c and C57BL/6 mice. Since previous reports suggested

that PGE-2 production from MPs can be upregulated by exposure to female sex steroid

hormones (El Attar et al., 1982; Smith et al., 1986; Yagel et al., 1987), I examined PGS-2

protein expression at various points during the estrus cycle as determined by vaginal

smear. I found that PGS-2 protein expression in MPs of female NOD mice is related to

the phase of the estrus cycle and can be detected only in the MPs of estrus phase mice. In

contrast, I did not detect any PGS-2 protein expression at any point of the estrus cycle of

control mouse strains (BALB/c, C57BL/6) (figure 7). The kinetics of PGS-2 protein

expression in MPs of estrus phase NOD female mice demonstrated that the PGS-2 protein

is constitutively expressed at all time points; i.e., immediately after peritoneal lavage, as

well as after 2, 6, 12, 24, and 48 hours of cell culture. In contrast, PGS-2 protein

expression was not detected at any time points in the MPs of control mice.








The spontaneous expression of PGS-2 in NOD MPs could caused by either a

primary defect in the regulation of PGS-2 gene itself or secondary to MP activation by

the autoimmune environment. To investigate whether the autoimmune milieu in NOD

mice contributes to PGS-2 expression, we examined MP PGS-2 mRNA and protein

expression in female NODscid/scid mice. The NOD scid/scid mouse, is a congenic strain

genetically identical to NOD but lacking functional T and B cells due to a homologous

mutation at the severe combined immunodeficiency (SCID) locus. NODscid/scid mice do

not develop autoantibodies, insulitis or diabetes. We examined PGS-2 mRNA (figure 8)

and protein (figure 9) expression in unstimulated peritoneal macrophages of 6-8 wk old

female NODscid/scid mice and found that PGS-2 mRNA and protein expression were

readily detectable, as in NOD MPs. The spontaneous expression of PGS-2 in the

NODscid/scid mice suggests that functional lymphocytes and active autoimmunity are

not required for MP expression of PGS-2 in NOD mice. This data supports the idea that

the PGS-2 enzyme is spontaneously expressed in estrus female NOD MPs and suggests

that the NOD PGS-2 gene regulation differs substantially from control strains.

Aberrant PG Metabolism in NOD. NODscid/scid. and Congenic Mice

PGS-2 is the rate limiting enzyme in PG metabolism and is responsible for the

production of large quantities of PGs. Since PGS-2 is consititutively expressed in NOD

mice, I hypothesized that there is enhanced PG production by NOD MPs. Examination of

PGE-2 production from resting peritoneal MPs of NOD and control mice in cell culture

revealed that PGE-2 production, as measured in the supernatant of Ixl05 unstimulated











470 bp, PGS-2








1050 bp, G3PDH







+ + + LPS
NOD BALB/c C57BL/6

Figure 6. PGS-2 mRNA expression in estrus NOD, BALB/c and C57BL/6 mice. mRNA
expression was assessed in 1 x 105 resident peritoneal MPs from 8 weeks old female mice
by RT-PCR after overnight culture with and without LPS stimulation. Unstimulated 8
weeks old BALB/c and C57BL/6 mice do not spontaneously express PGS-2 mRNA, but
do so when stimulated in vitro by LPS (10 ug/ml). PGS-2 and G3PDH PCR product
were detected on the 1.4% agarose gel (verified by blot). This is a representative PCR
result. (NOD, n=16 mice examined; BALB/c, n=16; C57BL/6, n=8). G3PDH mRNA was
used as an internal positive control and was equally expressed in all samples analyzed.









































Figure 7. Expression of PGS-2 protein in the resident peritoneal MP of 8 week old estrus
NOD, BALB/c and C57BL/6 mice. PGS-2 was detected by indirect immunofluorescence
using a specific PGS-2 specific anti-sera (Herschman's laboratory, UCLA; Reddy et al.,
1994) after MPs were cultured for 16 hours with or without LPS stimulation (10 ug/ml).
NOD MPs spontaneously expressed PGS-2 protein without LPS stimulation. In contrast,
BALB/c and C57BL/6 MPs did not express PGS-2 protein, but it was readily induced by
LPS stimulation. There is no apparent increase in PGS-2 expression in NOD MPs after
LPS stimulation.











S470 bp, PGS-2






S1050 bp, G3PDH


- + + LPS
NOD NODscid/scid


Figure 8. PGS-2 mRNA expression in estrus NOD, NODscid/scid mice. mRNA
expression was assessed in 1 x 105 resident peritoneal MPs from 8 weeks old female mice
by RT-PCR after overnight culture with and without LPS (10 ug/ml) stimulation. The
NODscid/scid mouse, which does not contain functional T and B lymphocyte, and does
not develop active autoimmune disease, expresses PGS-2 mRNA in a manner identical to
NOD. LPS stimulation does not markedly upregulate mRNA expression in NOD and
NOD scid/scid mice. PGS-2 and G3PDH PCR product were detected on the 1.4% agarose
gel (verified by blot). This is a representative PCR result. (NOD, n=16 mice examined;
NODscid/scid, n=10).


































Figure 9. Expression of PGS-2 protein in the resident peritoneal MPs of 8 week old estrus
NOD and NODscid/scid mice. PGS-2 was detected by indirect immunofluorescence
using a specific PGS-2 anti-sera (Reddy et al., 1994) after MPs were cultured for 16
hours with or without LPS stimulation (10 ug/ml). NOD and NOD scid/scid MPs
spontaneously express PGS-2 protein without LPS (10 ug/ml) stimulation and there is no
demonstrable change of PGS-2 expression in NOD or NODscid/scid MPs after LPS
stimulation.








resident peritoneal NOD MPs cultured overnight, was consistently higher than MPs from

control mice (figure 10).

To evaluate the influence of the autoimmune environment on PGE-2 production

in NOD mice, I also examined PGE-2 production from MPs of NODscid/scid mice (PGS-

2 positive). I found that PGE-2 production from NODscid/scid MPs was similar to NOD

mice (figure 10). I also examined PGE-2 production by MPs of congenic strain,

NOD.B10Cl mice. This mouse has an NOD background with a large interval of

chromosome 1 containing the PGS-2 gene from C57BL/10snj mice (detailed information

about this mouse and its PGS-2 expression will be discussed later in this chapter). This

mouse is negative for PGS-2 mRNA and protein expression, and its PGE-2 production is

lower than NOD (figure 10). We further demonstrated that enhanced PGE-2 production

by NOD MPs is mediated by the PGS-2 enzyme, as the PGS-2 specific inhibitor, NS-398,

completely blocks its production (figure 10). These data demonstrated that enhanced PG

production in NOD MPs is mediated through PGS-2 and not due to the autoimmune

environment. Although there is no statistically significant difference (p= 0.08) in PGE-2

production between MPs of NOD and control strains mice, the consistently higher PGE-

2 production from NOD MPs may contribute to the biologically functional differences

between NOD and control strains of mice. In control MPs, NS-398 reduced PGE-2

production to a small degree. This is most likely due to the transient expression of PGS-2

that occurs following the adherence of control MPs, as previously published studies

suggest (DeWitt et al., 1993).








Since PGE-2 production potently affects the activation of T cells in DC/T/MP co-

culture system as previous discussed, I hypothesized that a T cell or perhaps a DC factor

may contribute to the enhanced production of PGE-2 by MPs. Analysis of PGE-2

production from the co-cultures of MP/T or DC/MP/T in NOD and control strains mice

was under taken to test this theory.

I found that T cells enhanced PGE-2 production by MPs in all mice strains tested,

but PGE-2 production from NOD MP/T cells co-culture was consistently higher than

MP/T cell cultures from control strains. Furthermore, enhanced PGE-2 production was

mediated by PGS-2, since NS-398 totally blocked PGE-2 production (figure 11). The

production of PGE-2 by NOD.B 10C 1 MPs which are negative for PGS-2 expression was

also examined in co-culture. NOD.Bl0Cl MPs were cultured in the presence of both

NOD and NOD.B10Cl T cells and found to have PGE-2 production similar to controls.

These data suggest that NOD T cells by themselves are not capable of stimulating the

higher PGE-2 production seen in NOD, and that PGS-2 phenotype in NOD is essential

for enhanced PGs production. PGE-2 production in the DC/ MP/T co-culture were found

to be even higher in all mice strains examined, again, NOD DC/MP/T co-cultures are

higher in PGE-2 production than control mice strains and the difference is significant (P<

0.04) (figure 12).

These data suggest that signals from T cells, which may be enhanced by the

presence of DC, contribute to a general upregulation of PGE-2 production. The

mechanism involved is still largely unknown but Wang et al (Wang et al., 1996) found

that leukoregulin, a 50-kD cytokine product of mitogen-activated T lymphocytes, can








dramatically increase PGE-2 synthesis in cultured human orbital fibroblasts. This up-

regulation is mediated through a substantial increase in steady-state PGS-2 mRNA and

protein levels. This in turn results markedly increased generation of cAMP plus an

alteration in orbital fibroblast morphology. In the murine macrophage, leukoregulin could

play a similar role in the transient activation of PGS-2 and upregulation of PGE-2

production in both the NOD and control strains.

Factors) which Influence PGS-2 Gene Expression and the Mechanism
of Spontaneous PGS-2 Expression in NOD Mice

NOD PGS-2 Promoter Sequence Analysis

The PGS-2 regulatory region has several consensus sequences for transcriptional

activation including two potential NF-IL6 response elements, a Spl consensus sequence,

two cyclic AMP (cAMP) response elements (ATF/CRE), and an E box. Defects in any

regulatory element may potentially affect PGS-2 gene expression.

To investigate the contribution of cis-regulatory elements on PGS-2 expression in

the NOD mice, 5' promoter region was sequenced from -966 nucleotides to the

transcriptional start site. Genomic DNA from MPs of female NOD mice was extracted

from MPs and the PGS-2 promoter sequence was amplified the by PCR. PCR amplified

products were sequenced in the University of Florida sequence core laboratory

(Gainesville, FL). Sequence analysis showed that there were no mutations in the NOD

PGS-2 promoter compared with PGS-2 promoter from C57BL/6 or the published

sequence of3T3 fibroblast cells (Fletcher et al., 1992). Christy Myrick in Dr. Wakeland's

laboratory (University of Florida, Gainesville, FL) further confirmed our results by

cloning the NOD and C57BL/6 PGS-2 promoter gene into a TA vector system










OC57BU6
SBALB/c
*NOCB10C1
*NODBIOC
INctIs)
dNI:)scid/sd


9Z 250


' 200


S 150


100


B: MPs+NS398


Figure 10. PGE-2 production by murine macrophages. PGE-2 production from cell
culture (MEM media without phenol red and with 2% fetal bovine serum) of estrus phase
resident peritoneal MPs (Ixl05) of 8 week old female NOD, NODscid/scid, congenic,
and control mice was detected by ELISA ( ELISA kit, Cayman Chemicals) after MPs
were cultured 16 hours with and without a PGS-2 specific inhibitor, NS-395 (5 uM).
Values are given as the mean SEM. NOD, n=6 experiments; BALB/c, n=6; C57BL/6,
n=5; NODscid/scid, n=4; NOD.B10C1, n=4. One-way ANOVA was used to compare the
means of PGE-2 production in different MPs, p= 0.0795.


A: MPs


















800

700


600

500

2400


300 -

200.-

100. -


OC57BL/6
*BALB/c
BNODBIOCI
S NOD






: I4


A: MP/T


B: MP/T + NS398


Figure 11. PGE-2 production by murine macrophage and T cell co-culture. PGE-2
production from MP (Ixl05) /T cell (5x105) co-culture (MEM media without phenol red
and with 2% fetal bovine serum) of resident peritoneal MPs and purified splenic T cells
from 8 week old NOD, congenic, and control mice was detected by ELISA co-culture for
16 hours with and without a PGS-2 specific inhibitor, NS-395 (5 uM). For the
NOD.B10C1 MP/T co-culture, T cells are derived from NOD mice and MPs are from
NOD.B10C1 congenic mice. Values are given as the mean SEM. NOD, n=4
experiments; BALB/c, n=4; C57BL/6, n=4; NOD.B10C1, n=3. One-way ANOVA was
used to compare the means of PGE-2 production in different MPs, p= 0.06.


00


90D



700

600



400

300


200.-------------


I












3000


2500


2000


1500


1000.


500.


0.


OC57BU6
HBANLBc
*NOD


1000

900.

800.



600.

500.

400

300

200.


A: DC/MP/T


B: DC/ MP/T + NS398


Figure 12. PGE-2 production by dendritic cell, murine macrophage and T cell co-culture.
PGE-2 production from DC (xl104)/ MP (xl105) /T cell (5x105) co-culture (MEM media
without phenol red and with 2% fetal bovine serum) of lymph node dendritic cells,
resident peritoneal MPs and purified splenic T cells of 8 week old NOD, and control mice
was detected by ELISA after co-culture for 16 hours with and without a PGS-2 specific
inhibitor, NS-395 (5 uM). Values are given as the mean SEM. NOD, n=4 experiments;
BALB/c, n=4; C57BL/6, n=4. One-way ANOVA was used to compare the means of
PGE-2 production, p< 0.04.


11








(Invitrogen) and sequencing. She found no sequence differences between NOD,

C57BL/6, and published data.

Sex Hormones Influence on PGS-2 Gene Expression in the NOD Mice

The sexual dimorphism in the incidence of diabetes suggests that sex hormones

may play a role in the autoimmunity of the NOD mouse. The influence of sex steroid

hormones has been demonstrated by exacerbation of disease in orchiectomized or

estrogen treated NOD male mice, and a reduction in insulitis and diabetes incidence

following oopherectomy or treatment with androgens in female mice (Fox, 1992;

Hawkins et al., 1993; Fitzpatrick et al., 1991). Since estrogen and progesterone receptors

are expressed on macrophage and female sex steroid hormones have been described to

stimulate PGE-2 production (Smith et al., 1986; Yagel et al., 1987), these hormones may

influence prostanoid metabolism through the induction of PGS-2. The murine female

estrus cycle lasts about 4-5 days and the estrus phase about 12 hours; with estrogen and

progesterone reaching their highest level during estrus phase (Allen, 1922). I therefore

hypothesized that estrus related changes in sex steroids may play a role in the regulation

of PGS-2 mRNA and protein expression. My data demonstrate a sexual dimorphism in

the expression of PGS-2 in female NOD mice and dependence on the estrus phase. I

found first, that the level of peritoneal MP PGS-2 mRNA expression was lower in female

mice during non-estrus phase compared with that of estrus phase (table 3). Second, PGS-

2 protein expression also varies during the estrus cycle as the protein was only detected

during estrus phase (table 3). Third, PGS-2 protein expression in NOD female mice was

only detected in sexually mature mice as it was expressed at 4 week old but not in








sexually immature 2 week old female NOD mice (table 5). Finally, PGS-2 mRNA

(figure 13) and protein (figure 14) expression were not found in male NOD MPs.

In mice, PGE-2 production from macrophages has been reported to be upregulated

by exposure to 10"7 to 108 M concentrations of progesterone (Smith et al., 1986; Yagel et

al., 1987). The mechanism for this enhancement is still unknown. I hypothesized that sex

hormones may affect PGE-2 production through regulation of PGS-2 expression. Sex

hormone regulation of PGS-2 expression was examined in both in vitro and in vivo

experiments to test this hypothesis.

I demonstrated that physiological concentrations of estradiol (10'6 M) and

progesterone (10"7) upregulate PGS-2 protein expression in NOD non-estrus female and

male NOD MPs (figure 15). In contrast, the male hormones testosterone (107 M), has no

effect on the PGS-2 mRNA and protein expression in NOD female MPs (figure 15). I

also castrated 8 week old male NOD mice and examined PGS-2 expression four weeks

after orchiectomy and found that PGS-2 mRNA and protein were not affected by removal

of the testicles (table 4). PGS-2 mRNA and protein expression in female and male

resident peritoneal MPs are summarized in table 3 and table 4. Table 5 summarizes PGS-

2 expression in NOD female mice of different ages.

Role of Monokines on the Regulation of PGS-2 Expression in NOD Mice

PGS-2 expression in MPs can be induced by the monokines TNF-a and IL-1

(Ristimaki et al., 1994; Lee et al., 1992). Once expressed, PGE-2 production can suppress

TNF-a and IL-1 gene expression while up-regulating IL-10 expression. IL-10 production,

in turn, suppresses TNF-a, IL-1, and PGS-2 expression (Mertz et al., 1994). In addition












470 bp, PGS-2








S1050 bp, G3PDH





+ LPS
NOD (male)

Figure 13. PGS-2 mRNA expression in male NOD mice. mRNA expression was assessed
in 1 x 105 resident peritoneal MPs from 8 week old male mice by RT-PCR after overnight
culture with and without LPS stimulation. The male NOD mice do not express, or express
limited amounts of PGS-2 mRNA. PGS-2 is readily induced by LPS (10ug/ml).








































Figure 14. Expression of PGS-2 protein in resident peritoneal MPs of 8 week old NOD
male mice was detected by indirect immunofluorescence using PGS-2 specific anti-sera
(Reddy et al., 1994) after MPs were cultured for 16 hours with or without LPS
stimulation (10 ug/ml). NOD male MPs did not spontaneously express PGS-2 protein, but
it was readily induced by LPS stimulation.


NOD Nlale










.........


N0D,'\Iale+LPS




































Figure 15. Effect of sex steroid hormones on the expression of PGS-2 protein in resident
peritoneal MPs of 8 week old NOD (male and female). PGS-2 was detected by indirect
immunofluorescence using PGS-2 specific anti-sera after MPs were cultured for 16 hours
with or without stimulation. NOD male PGS-2 expression (A) is negative as is NOD non-
estrus female MP expression, but can be induced by progesterone (P, 10" M) (B). NOD
estrus female MPs spontaneously express PGS-2 protein (C) and testosterone (T, 108 M)
had no effect on PGS-2 expression in estrus female NOD MPs (D).








Table 3. Summary of PGS-2 mRNA and protein expression in female resident peritoneal
MPs. mRNA and protein expression were assessed in resident peritoneal MPs from 8
week old mice (estrus and non-estrus phase) by RT-PCR and immunocytochemistry after
overnight culture with or without LPS (10 ug/ml), and with or without various sex steroid
hormones at physiological concentration (NOD only). Estradiol (106 M), progesterone
(10'7 M), and testosterone (10'7 M).

PGS-2 mRNA PGS-2 mRNA PGS-2 Protein PGS-2 Protein

(Estrus) (Non-Estrus) (Estrus) (Non-Estrus)

C57BL/6- -

BALB/c-

NOD + + + -/low

NOD scid/scid + + + -/low

C57BL/6+LPS + + + +

BALB/c + LPS + + + +

NOD + LPS + + + +

NODscid/scid+LPS + + + +

NOD + Estradiol + + + +

NOD + Progesterone + + + +

NOD + Testosterone + + + -/low








Table 4. Summary of PGS-2 mRNA and protein expression in male resident peritoneal
MPs. mRNA and protein expression were assessed in resident peritoneal MPs from 8
weeks old mice by RT-PCR and immunocytochemistry assay after overnight culture with
or without LPS, and without or with estradiol (10 6 M) and progesterone (10-7 M). NOD
males were also castrated at 8 weeks of age. Four weeks after castration, mRNA and
protein expression was assessed in peritoneal MPs after overnight culture.

PGS-2 mRNA PGS-2 Protein

BALB/c

BALB/c + LPS + +

NOD

NOD + LPS + +

Castrated NOD-

NOD + Estradiol + +

NOD + Progesterone + +


Table 5. Summary of PGS-2 protein expression in female resident peritoneal MPs. PGS-2
protein expression was assessed in resident peritoneal MPs from mice by
immunocytochemistry after overnight culture.








to monokines, cytokines produced by T cells, i.e., IFN-y, upregulate PGS-2 expression,

and both T cell and MP derived TGF-P can suppress PGS-2 expression (Reddy et al.,

1993).

The spontaneous expression of PGS-2 in NOD could be due to: 1) a primary

defect in regulatory elements of the PGS-2 gene, or 2) a defect in the monokines that

affect PGS-2 expression. Studies have shown that TNF-a and IL-1 can upregulate PGS-2

expression. Defects leading to enhanced production of TNF-a or IL-1 may ultimately

contribute to constitutive expression of PGS-2 in the NOD. On the other hand, IL-10 or

TGF-P can suppress PGS-2 expression, so defects in the down regulation of PGS-2 by

IL-10 or TGF-P could also contribute to the differential expression of PGS-2 in NOD

MPs. Therefore, perturbing the autocrine regulation of PGS-2 may contribute to the

spontaneous expression of PGS-2 in NOD MPs.

To investigate the influence of monokines on PGS-2 expression, TNF-a, and IL-

1 P production were quantitated from the supernatant of 1x105 resident peritoneal MPs of

NOD, BALB/c and C57BL/6 mice cultured overnight. The production of monokines

were found to be undetectable by ELISA (assay detection limit 30 pg/ml and 70 pg/ml

respectively) in all three mouse strains examined; and were produced at similar levels

following LPS stimulation ( about 70 pg/ml for TNF-a, and 130 pg/ml for IL-10). These

data suggest that spontaneous PGS-2 expression is not due to the enhanced production of

the monokines TNF-a and IL- 1 by NOD MPs. Since the TNF-a and IL-1 levels in MPs

cell culture are below ELISA detection limit, further in vitro studies to investigate the

effect of quantitative TNF-a and IL-1 neutralizing antibodies on the NOD MPs PGS-2








expression can more directly address the issue of the sensitivity of PGS-2 expression to

the autocrine TNF-ac and IL-1 in NOD MPs.

Previous studies have shown that PGS-2 mRNA and protein expression is readily

suppressed by IL-10 (Mertz et al., 1994). If IL-10 production in NOD MPs is absent or

low, or if PGS-2 becomes insensitive to IL-10 regulation in NOD MPs, then, PGS-2

expression may be enhanced. IL-10 production from the supernatant of lxlO5 resident

peritoneal MPs of both BALB/c and NOD mice cultured overnight was examined. IL-10

in these samples was undetectable (below 30 pg/ml detection limit) but can be induced by

LPS to similar levels (NOD:70 pg/ml Vs BALB/c: 75 pg/ml). These data suggest that

following LPS stimulation, NOD mice produce IL-10 at the levels which normally would

suppress PGS-2 expression. To examine the regulation of PGS-2 by IL-10, I examined

the suppressive effect of this cytokine on PGS-2 protein expression in both NOD MPs

and LPS stimulated BALB/c and C57BL/6 MPs. I found that 10 and 100 ng/ml rmlL-10

totally blocked the LPS induced PGS-2 expression in BALB/c and C57BL/6 mice (figure

16). In contrast, up to 500 ng/ml rmIL-10 did not block spontaneous PGS-2 expression in

NOD MPs. These data suggest that there may be a general desensitization in NOD MP

responses to IL-10 which could be due to a defect in the IL-10 receptor or post-IL-10

receptor signal transduction cascade. This data may also provide a mechanism for

spontaneous expression of PGS-2 in NOD MP: the loss of the suppression function of IL-

10. Alternatively, there may be a differential regulation regarding IL-10 mediated

suppression of PGS-2 expression. i.e., LPS induced PGS-2 expression is sensitive to IL-

10 suppression; whereas, spontaneous PGS-2 expression in NOD mice which is








upregulated by sex steroid hormones may be IL-10 insensitive. Further experiments are

needed to elucidate the mechanisms responsible for the defect in the IL-10 suppression of

PGS-2 expression in NOD MPs. Table 6 summarizes the rmIL-10 inhibitory effect on

PGS-2 protein expression.

Contribution of Spontaneous PGS-2 to the Immunopathogenesis of IDD
a Candidate Gene for Idd5

A particularly powerful approach to the study of polygenetic disease is the

development of congenic mouse strains with genetically defined segments of

chromosomes from non-autoimmune control mice on the background of the autoimmune

mouse, or visa versa. Once established, these strains can be used to assess the effect of

these chromosomal segments on disease or on an immunophenotype, i.e., PGS-2

expression in our case. The immunophenotype may be particularly helpful in suggesting

candidate genes in the congenic interval if sufficiently well characterized.

My studies suggest the constitutive expression of PGS-2 is a primary defect based

on data from NODscid/scid mice. I hypothesize that aberrant PGS-2 expression may be

genetically determined. I, hence, examined PGS-2 expression in congenic mouse strains

that contain segments of chromosome 1 which include the PGS-2 gene from either NOD

or C57BL/10snj. The first congenic strain examined was NOD.B10C1. This strain was

derived by Dr. Linda Wicker's laboratory at Merck Laboratory, Rahway, NJ. These mice

contain an interval of C57BL/10snj chromosome 1 that includes the B10 PGS-2 gene on

the NOD background (figure 17). In female homozygous NOD.B 10C1 mice, replacement

of the NOD chromosome 1 segment with the C57BL/10snj segment delayed the onset of

diabetes to 14 weeks of age (compared with onset in NOD which is 10 weeks of age) and
















B: BALB/c+LPS C: BALB/c+LPS+
IL-10(1Ong/ml)


E: NOD F: NOD+
IL-10(10ng/ml)


D: BALB/c+LPS
+DEX(2uM)


G: NOD+ H: NOD+
IL-10(500ng/ml) DEX(2uM)


Figure 16. mrIL-10 suppresses LPS induced PGS-2 expression in BALB/c but not
spontaneously expressed PGS-2 in estrus NOD mice. Peritoneal MPs (xl105 ) were
cultured with or without LPS (10 ug/ml) and/or different doses of mrIL-10 overnight.
NOD PGS-2 protein expression is completely suppressed by dexamethasone (DEX, 2
uM). PGS-2 protein was examined by immufluorescence detection using a PGS-2
specific anti-sera (Reddy et al., 1994).


A: BALB/c








Table 6. Summary of rmIL-10 mediated suppression of PGS-2 protein expression in
resident peritoneal MPs of female mice. Protein expression was assessed in resident
peritoneal MPs from 8 week old female mice (estrus phase) by immunocytochemistry
after overnight culture with or without LPS (10 ug/ml), in combination with
dexamethasone (DEX, 2 uM) and mrIL-10 (10, 100, and 500 ng/ml).

BALB/c C57BL/6 NOD

MPs ++

MPs+IL-10 (100 ng/ml) -+

MPs+IL-10 (500 ng/ml) ++

MPs+DEX

MPs+LPS ++ ++ ++

MPs+LPS+DEX (2 uM)

MPs+LPS+IL-10 (10 ng/ml)

MPs+LPS+IL-10 (500 ng/ml) ++








the cumulative incidence of diabetes by the age of 30 weeks was decreased to

approximately 40% (compared with NOD which is 80%) (figure 18). I examined PGS-2

expression in the resident peritoneal MPs of estrus female NOD.B10C1 mice and found

that MPs from NOD.B10C1, unlike those from parental NOD mice, do not express PGS-

2 mRNA or protein (figure 19). As a control, PGS-2 protein expression in NOD.B10C3

(C57BL/10snj chromosome 3) and NOD.B10C4 (C57BL/10snj chromosome 4) congenic

mice were also examined. Both of these congenic strains have the NOD PGS-2 gene and

express the PGS-2 protein to the same level as the NOD MPs (figure 20)

I also examined the B6.NODCl congenic mouse. The B6.NODC1 mouse was

derived in Dr. Edward Wakeland's laboratory (University of Florida, Gainesville, FL)

and contains an interval of NOD chromosome 1 (Idd 5) including NOD PGS-2 on the

C57BL/6 background (figure 17). B6.NODC1 develop peri-insulitis by 6 months of age. I

found that MPs from estrus B6.NODC1 mice constitutively express PGS-2 protein

(figure 19). PGS-2 expression in these mice was further confirmed by Western blotting

(figure 20).

My studies demonstrated that the PGS-2 phenotype is correlated with the

autoimmune phenotype in these congenic animals. The positive expression of PGS-2 in

B6.NODC1 mice is correlated with the development of peri-insulitis in this congenic

strain; whereas, the lack of expression of PGS-2 in NOD.B10C1 is correlated with a 50%

drop of diabetes incidence by 30 weeks of age. These data suggest that PGS-2 is a

candidate susceptibility gene for Idd5 in NOD mice.







Mouse Chromosome 1


NOD.BO1C1
(Wicker/Peterson)


D1MIT4


D1MIT5
D1MIT8


PGS-2
(76.2cM)
D1MIT15

- D1MIT17


B6.NODC1
(Wakeland/Yui)


Idd 5


Figure 17. Location of the PGS-2 gene on chromosome 1 and the intervals included in the
congenic mouse strains evaluated.









Diabetes Incidence (%)

100


80


60


40


20


0
60 80 100 120 140 160 180 200 220

AGE (days)


NOD

-NOD.B10C1-



NOD.B10C1


NOD.B10C1: NOD background with an interval from C57BL/10snJ Chromosome 1

Figure 18. Cumulative incidence of diabetes in NOD.B10C1 congenic mice (from Dr.
Linder Wicker, Merck Laboratory, Rahway, NJ).



































Figure 19. Expression of PGS-2 protein in resident peritoneal MPs of 8 week old female
estrus congenic mice NOD.B10CI and B6.NODC1. PGS-2 was detected by indirect
immunofluorescence using a specific PGS-2 specific anti-sera (Reddy et al., 1994) after
MPs were cultured for 16 hours without stimulation. Both homozygous and heterozygous
(not shown) B6.NODC1 spontaneously express PGS-2, while NOD.B10C1 homozygous
mice do not.


B6.NODCI honio/N ,(ms estrus female











NOD.1110(A holliozN-olls esti-tis felliale








Table 7. Summary of background information and MP PGS-2 expression in the congenic
strains NOD.B 1 OC 1 and B6.NODC 1.

NOD.BIOC1 B6.NODC1

Background NOD C57BL/10snj

Chrom.1/PGS-2 C57BL/10snj NOD

Insulitis/ Insulitis/ Peri-Insulitis/

Diabetes Incidence Diabetes Incidence: 45% No Diabetes

PGS-2 Expression +-










U,

i"sI

MQ U Z z


70 Kd -


-- -


Figure 20. Detection of PGS-2 protein in resident peritoneal MPs from 8 week old female
estrus NOD, non estrous NOD, NOD.Bl0C1, NOD.B10C3, NOD.B10C4, and control
mouse strains. PGS-2 was detected by SDS-polyacrylamide gel electrophoresis (5%
stacking and 8% resolving gel) and Western blotting. Fresh resident peritoneal MPs
(2x106) were harvested in SDS gel loading buffer. Extracts were prepared and subjected
to electrophoresis and western analysis with PGS-2 specific antiserum (Reddy et al.,
1994). Seventy five ug of protein was loaded per lane.


0

z^


0
0


-PGS-2








Table 8. Summary of PGS-2 expression in NOD, NODscid/scid, congenic, and parental
mouse strains.

Background Chrom. 1 PGS-2 Insulitis/Diabetes Mellitus

NOD NOD NOD + +/80%

NODscid/scid NOD NOD + -/0%

B6.NODC1 C57BL/6 NOD + Peri-insulitis/0%

C57BL/6 C57BL/6 C57BL/6 -/0%

NOD.B10C1 NOD C57BL/10snj +/45%

C57BL/10snj C57BL/10snj C57BL/10snj -/0%








PGS and iNOS Inhibitors Affect NOD Diabetes Incidence

Our data suggest that aberrantly expressed PGS-2 contributes to the enhanced

PGE-2 production in NOD MPs. Since PGE-2 has multiple important effects on the

immune response and peripheral tolerance, we investigated whether blocking PGS-2

enzyme activity pharmacologically prevents diabetes in NOD. We treated NOD mice

with a combination of indomethacin, a PGS-2 inhibitor, and aminoguanidine, an

inducible nitric oxide synthase (iNOS) inhibitor. We chose this combination because our

data suggested that NO also contributes to MP mediated suppression of T cell activation,

and because NO markedly enhances the enzymatic activity of PGS-2 (Salvemini et al.,

1993; Franchi et al., 1994). In two separate experiments, we either treated NOD female

mice from 8 weeks of age with a combination of indomethacin (3 ug/ml) and

aminoguanidine (0.1%) or from 4 weeks of age with high doses of indomethacin alone

(20 ug/ml) in the drinking water. In an initial study, we found that treatment of NOD

female mice from 8 weeks of age with a combination of indomethacin/aminoguanidine

(I/A) significantly delayed the onset (onset began at 20 weeks of age in I/A treated group

compared with 8 weeks in the untreated group) and reduced the diabetes incidence by

40% compared to untreated NOD of the same age. (P< 0.02, figure 21). In a truncated

experiment, we observed that treatment of NOD female mice from 4 weeks of age with

high dose of indomethacin alone (20 ug/ml) provided an identical effect in delaying the

onset of diabetes and reducing the diabetes incidence and difference is also significant

(p< 0.04, figure 22). These data further support a critical role for PGE-2 in the

pathogenesis of diabetes.










Incidence of diabetes


100


80
60
40
20
0


-u- Control
-*-AG
- IN+AG


Weeks

Figure 21. Cumulative incidence of diabetes in NOD mice treated with indomethacin (IN,
3 ug/ml) and aminoguanidine (AG, 0.1%). NOD female mice were treated from 8 to 32
weeks of age with indomethacin and aminoguanidine in their drinking water. Fresh drugs
were added to the water during weekly water change. Diabetes was monitored weekly by
a urine glucose test and confirmed by tail vain blood glucose level (250 mg/dl glucose
was used as indication of diabetes). The logrank chi-square statistic (Mantel, 1988) was
used to compare incidence of diabetes between the control and the IN/AG treated group.
p< 0.02, p values are one sided. n=20 animals per group.









Incidence of Diabetes
100-
80 Control
60 -
40-
20 -
0 *--- -
4 6 8 10 12 14 16 18 20 22

Weeks

Figure 22. Cumulative incidence of diabetes in NOD mice treated with indomethacin (IN,
20 ug/ml). NOD female mice were treated from 4 to 22 weeks of age with indomethacin
in their drinking water. Fresh drugs were added to the water during weekly water change.
Diabetes was monitored weekly by a urine glucose test and confirmed by tail vain blood
glucose level (250 mg/dl glucose was used as indication of diabetes). The logrank chi-
square statistic (Mantel, 1988) was used to compare incidence of diabetes between the
control and the IN/AG treated group. p< 0.02, p values are one sided. n=9 animals per
group.







Impaired AICD in NOD Mice Secondary to Enhanced PG Production

Activation induced cell death (AICD), the apoptosis of T cells following the

activation with antigen (Rocha et al., 1991; Zhang et al., 1992), is an important

mechanism for peripheral T cell tolerance. Several studies have suggested that AICD can

be markedly inhibited by anti-IL-2 antibodies and by agents that block the cell cycle, i.e.,

cAMP. Of interest, PGE-2 is a potent inhibitor of IL-2 and IL-2 receptor expression and

blocks the cell cycle through its activation of adenylate cyclase and generation of cAMP

(Goetzl et al., 1995; Ucker et al., 1994). Based on my data and other published studies

(Lety et al., 1992) there is enhanced PG production in NOD mice. I therefore

hypothesized that enhanced PGs production may result in an impaired AICD and

contribute to NOD diabetes. To test my hypothesis, I followed an established in vivo

experimental model of ACID, the immunization of mice with the bacterial superantigen,

Staphylococcus aureus enterotoxin B (SEB), as described by MacDonald et al. (1991) and

Kawabe et al. (1991). Following MacDonald's protocol, I examined ACID in the groups

of untreated NOD mice, indomethacin (20ug/ml)/aminoguanidine (0.1%) (IN/AG) treated

NOD and BALB/c mice (IN/AG were added to the drinking water 2 weeks prior to SEB

immunization and continued on these drugs throughout the remainder of the experiment)

and untreated control strains of mice (C57BL/6 and BALB/c). On day zero, animals were

given an intraperitoneal injection of buffered saline or SEB. Splenic cells from three

mice in each group were examined on day 2 and day 10 after immunization for the

percentage of Vp8+/CD4+ T cells and Vp6+/CD4+ T cells by fluorescence activated cell

sorter (FACS) analysis (figure 23, 24, 25).








In addition to the effects of PGS-2, MHC molecules may also affect AICD in the

NOD mouse. MHC molecules play a crucial role in immune response. Any defect in

MHC molecules leading to down regulation of antigen presentation or activation of T cell

would directly impair immune tolerance. Changes or variations in MHC molecules are

associated with several autoimmune diseases (Erlich et al., 1993). Studies also

demonstrated that the unique H-27 MHC plays a central role in NOD autoimmunity

(Hattori et al., 1986; Prochazka et al., 1987; Ikegami et al., 1988). Recent studies suggest

that the NOD MHC molecule is inherently unstable and as it does not readily dimerize,

that decreases its efficiency in binding and presenting antigens (Carrasco-Marin et al.,

1996). I therefore hypothesized that unique properties of H-2g7 in combination with

aberrant PGS-2 expression may disturb its ability to present self antigen to self reactive T

cells and impair AICD in NOD mice.

AICD in NOD.H-2b congenic mice were examined as control for the effect of

NOD MHC molecule (figure 26). These mice express the C57BL/6 H-2b MHC gene on a

NOD background.

Figure 23 shows the expansion and deletion of V38+/CD4+ T cells in response to

SEB (12.5 ug) immunization in C57BL/6 mice. BALB/c mice were also immunized with

SEB (50 ug) and a similar pattern of expansion and deletion as that of C57BL/6 was seen

but at higher levels (figure 24). Figure 25 illustrates the markedly impaired expansion of

the Vp8+ T cell population in NOD. However, if NOD mice are pre-treated with drugs

that block PGS-2 activity, the early expansion is not affected, but deletion on day 10

increased two fold. Similar indomethacin and aminoguanidine treatment of BALB/c mice








did not shown effect on the early expansion and later deletion of VP8+/CD4+ T cells

(figure 24). The potent effect of the NOD MHC H-297 on the expansion and deletion of

Vp8+ T cells is illustrated in the data shown in figure 26.

These experiments demonstrated that AICD in NOD is predominantly affected by

the MHC molecule and that PGS-2 contributes, but to a much lesser degree than MHC to

the impairment of deletion. Our study suggests that the NOD MHC molecule

incombination with PGS-2 expression plays a dramatic role in the impairment of AICD in

superantigen driven expansion and deletion of T cells in NOD mouse.












~i~I
'
-


- Control
---SEB: 12.5


20
15
10
5
0
-5
-10
-15
-20
-25
-30
0


Figure 23. Percent change in Vp8+/CD4+ and VP6+/CD4+ splenic T cells in SEB (12.5
ug) immunized C57BL/6 mice. Mice were immunized with 12.5 ug of SEB by
intraperitoneal injection and the number of splenic Vp8+/CD4+ (a) and VP6+/CD4+ (b)
were measured by dual color FACS analysis on day 2 and day 10 following
immunization. Three animals at each time point were tested. Results are expressed as
mean value of percentage change SEM (Percent change is the percent of Vp8+/CD4+
or VP6+/CD4+ T cells in each individual animal subtract the mean value of percent of
Vp8+/CD4+ or V36+/CD4+ T cells in control group, and then divided by the mean value
of percent of Vp8+/CD4+ or VP6+/CD4+ T cells in control group).


2 10
Days


20
- 15 Control
i 10 SEB: 12.5
+ 5

-5
4 -10
S-15
-20
4 -25
-30
0 2 10
Days





89


100 100
Control Control

..60 A -...SEB:50 0 o_ _.. _B:50
+ 60 ,j V. + 60
40 ... .. IN/AG 40 ... ..IN/AG
20 l. ... IN/AC Q 20 _. .I / A
..N/AQ
0 0 B:50 0 | B:50
-20 -20
-40 -40
0 2 10 0 2 10
Days Days


a. b.


Figure 24. Percent change in V38+/CD4+ and V36+/CD4+ splenic T cells in SEB (50
ug) immunized BALB/c mice. Untreated BALB/c mice and BALB/c mice treated for 2
weeks with indomethacin (IN, 20 ug/ml)/aminoguanidine (AG, 0.1%) were immunized
with 50 ug of SEB by intraperitoneal injection and the number of splenic Vp8+/CD4+ (a)
and V06+/CD4+ (b) were measured by dual color FACS analysis on day 2 and day 10
following immunization. Three animals at each time point were tested. Results are
expressed as mean value of percentage change SEM (Percent change is the percent of
VJ8+/CD4+ or VD6+/CD4+ T cells in each individual animal subtract the mean value of
percent of Vp8+/CD4+ or V06+/CD4+ T cells in control group, and then divided by the
mean value of percent of Vp8+/CD4+ or V06+/CD4+ T cells in control group).











.---. Control

--- SEB:12.5

......IN/AG

.... IN/AG,
SEB:12.5


0 2
Days


15
10
5
o .... ... .. -
.5
-10
-15
-20
-25
-30
0 2 10


-... Control

__..SEB: 12.5

......IN/AG

....IN IN/AG
SB:12.5


Figure 25. Percent change in Vi8+/CD4+ and Vp6+/CD4+ splenic T cells in SEB (12.5
ug) immunized NOD mice. Untreated NOD mice and NOD mice treated for 2 weeks with
indomethacin (IN, 20 ug/ml)\aminoguanidine (AG, 0.1%) were immunized with 12.5 ug
of SEB by intraperitoneal injection and the number of splenic Vp8+/CD4+ (a) and
Vp6+/CD4+ (b) were measured by dual color FACS analysis on day 2 and day 10
following immunization. Three animals at each time point were tested. Results are
expressed as mean value of percentage change SEM (Percent change is the percent of
Vp8+/CD4+ or V36+/CD4+ T cells in each individual animal subtract the mean value of
percent of Vp8+/CD4+ or Vp6+/CD4+ T cells in control group, and then divided by the
mean value of percent of Vp8+/CD4+ or Vp6+/CD4+ T cells in control group).











..........Control
Z : SEB: 12.5

#
\


0 2 10
Days


Figure 26. Percent change in Vp8+/CD4+ and V36+/CD4+ splenic T cells in SEB (12.5
ug) immunized NOD.H-2b congenic mice. Mice were immunized with 12.5 ug of SEB by
intraperitoneal injection and the number of splenic Vp8+/CD4+ (a) and Vp6+/CD4+ (b)
were measured by dual color FACS analysis on day 2 and day 10 following
immunization. Three animals at each time point were tested. Results are expressed as
mean value of percentage change SEM (Percent change is the percent of Vp8+/CD4+
or Vp6+/CD4+ T cells in each individual animal subtract the mean value of percent of
Vp8+/CD4+ or Vp6+/CD4+ T cells in control group, and then divided by the mean value
of percent of Vp8+/CD4+ or Vp6+/CD4+ T cells in control group).


I


20 ...__
_a 15 Control
U 10 ....SEB: 12.5
+ 5
U 0 L--
S-5
> -10
-15
-20
-25
-30
0 2 10
Days













CHAPTER 4
DISCUSSION

Insulin-dependent diabetes (IDD) is an autoimmune disease. The non-obese diabetic

mouse is a well established, spontaneous animal model of the human disease (Makino et al.,

1980). Both T cells and antigen presenting cells, especially macrophages, appear to play

important roles in the immunopathogenesis of this disease. Several studies have suggested

that antigen presenting cells are defective in their capacity to active T cells in IDD (Yokona

et al., 1989; Ransanen et al., 1989; Rapoport et al., 1993; Serrze et al., 1990). The cellular

and molecular basis of this defect, however, is largely unknown. The studies presented here

demonstrate that resident NOD peritoneal MPs, unlike control MPs, spontaneously express

mRNA transcripts and protein for the early response gene prostaglandin synthase II (PGS-

2). My initial studies suggest that PGS-2 insensitivity to the suppression by murine IL-10

may play a role in its constitutive expression. I also found that constitutive expression of

PGS-2 in NOD MPs is responsible for enhanced prostanoid production, and that treating

NOD mice with drugs that block PGS-2 enzymatic activity significantly delay the onset and

reduce the incidence of diabetes. I further demonstrate that enhanced prostaglandin

production may interfere with tolerance mechanisms by impairing AICD. Finally, my data

from chromosome 1 congenic mice suggest a genetic basis for the NOD PGS-2 phenotype

and its candidacy as a diabetes susceptibility gene.








My data show that unstimulated peritoneal NOD MPs from 8 week old estrus

female NOD mice housed in a SPF facility constitutively express high levels of PGS-2

mRNA and protein (figure 6, 7). In contrast, unstimulated resident MPs from control

strains (BALB/c, C57BL/6, C57BL/10) do not express this enzyme without induction by

agents such as LPS. To exclude the possibility that the autoimmune environment and

autoimmune T and B cells in NOD mice contribute to the PGS-2 expression, I also

examined its expression in MPs of female NODscid/scid mouse. The NODscid/scid

mouse which lacks functional T and B cells and consequently does not develop

autoimmunity, also expresses PGS-2 constitutively. These data suggest that the

autoimmune environment is not necessary for its expression and may be a primary defect

in the regulation of PGS-2 in MP of NOD mice.

Because PGS-2 was differentially expressed in NOD mice, its expression may

result in enhanced prostaglandin production by NOD MP. PGE-2 production was

analyzed in MPs cell culture supernatant of NOD, NODscid/scid, and control mice. PGE-

2 production of resident peritoneal MPs was found to be consistently higher than control

MPs. Also PGE-2 levels produced by NODscid/scid MPs are similar to those of NOD

mice. Furthermore, PGE-2 production by NOD and NODscid/scid MPs is mediated by

the PGS-2 enzyme since a PGS-2 specific inhibitor, NS-398, completely blocks its

production. These data demonstrated that there is an enhanced PG production in NOD

MPs and that this production is secondary to the constitutive expression of PGS-2.








Since the finding indicate NOD PGS-2 mRNA is constitutively expressed and

protein expression is up regulated during estrus phase, both transcriptional and

translational regulation may be involved in the aberrant PGS-2 expression.

The expression of PGS-2 expression following induction is regulated mainly by

transcriptional activation, but post-transcriptional regulation also occurs (DeWitt et al.,

1993; Evett et al., 1993; Ristimaki et al., 1994). PGS-2 mRNA is unstable compared with

PGS-1 mRNA; a feature predicated from the presence of multiple RNA instability

sequence (AUUUA) in its 3'-untranslated region. PGS-2 mRNA is translated as soon as it

is synthesized; therefore, the short mRNA half-life (less than 2 hrs) limits PGS-2

production post-transcriptionally (DeWitt et al., 1993). Evett et al. (1993) suggest that

instability of PGS-2 protein, a relatively short half-life of 22 min in chicken embryo

fibroblasts, also plays a role in the regulation of PGS-2 enzyme expression. Due to the

complex regulation and the instability of PGS-2 mRNA and protein, PGS-2 expression is

transiently expressed with mRNA levels peaking at 1 hr and protein levels peaking at 2 hr

following induction and returning to basal levels by 4 hr and 6 hr respectively (DeWitt et

al., 1993).

My data show that PGS-2 expression in NOD MPs differs from PGS-2 expression

in the MPs of control strain mice. The defect that contributes to aberrant PGS-2

expression may occur at any stage of gene expression; i.e., transcription level, post-

transcription level (mRNA stability), translation level, and post translation level (protein

stability).