Staphylococcal enterotoxin superantigens

MISSING IMAGE

Material Information

Title:
Staphylococcal enterotoxin superantigens structural studies and role in autoimmune disease
Alternate title:
Structural studies and role in autoimmune disease
Physical Description:
x, 142 leaves : ill. ; 29 cm.
Language:
English
Creator:
Soos, Jeanne Margaret, 1967-
Publication Date:

Subjects

Subjects / Keywords:
Enterotoxins -- immunology   ( mesh )
Staphylococcus   ( mesh )
Superantigens -- immunology   ( mesh )
Superantigens -- physiology   ( mesh )
Superantigens -- pharmacology   ( mesh )
Multiple Sclerosis -- etiology   ( mesh )
Disease Models, Animal   ( mesh )
Department of Pathology and Laboratory Medicine thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Pathology and Laboratory Medicine -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 127-140).
Statement of Responsibility:
by Jeanne Margaret Soos.
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:
aleph - 002321809
notis - ALS5339
oclc - 50697535
System ID:
AA00009051:00001

Full Text









STAPHYLOCOCCAL ENTEROTOXIN SUPERANTIGENS: STRUCTURAL STUDIES
AND ROLE IN AUTOIMMUNE DISEASE











By

JEANNE MARGARET SOOS


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


UNIVERSITY OF FLORIDA


1994























I would like to dedicate this work to my parents, Steven and Elizabeth Soos.

My father has believed in me and helped me to believe in myself. My mother's

caring, intelligence and strength and the aspirations she had for her daughter will

always be an inspiration to me.














ACKNOWLEDGEMENTS

I would like to thank my mentor, Dr. Howard M. Johnson, for his insight,

guidance and brilliant, in-depth analysis of the film The Godfather, which I will

remember fondly and for which I will always be grateful. He has taught me as

much about life as he has about science. He has also listened to ideas, mine as

well as others, with sincerity and without disdain, which is an invaluable gift to a

graduate student. I have been fortunate to have had him as my mentor. My

thanks also extends to Dr. Ammon B. Peck, Dr. Edward Hoffmann, and Dr.

Maureen Goodenow and especially Dr. Joel Schiffenbauer for his collaboration on

the EAE studies presented here. I would like to thank my labmates past and

present, Barbara, Brian, Maria, Alicia, Harold, Tim, Keisha, Amy, Sherri (Sherri-Jo),

Prem and Shaneeka for their assistance and friendship through everything. I thank

also Marilyn, Christy, Liliana, Minh-Thanh, Jeff, Mike and Karen for their friendship.

My family and friends also deserve my heartfelt thanks and gratitude for

putting up with me. All the times I've explained what I've been working on, they

listened patiently and were always supportive. My sincere thanks go to Steven

Soos, my dad, for all of his love and support over the years. I wouldn't have been

able to do this without his encouragement. I would also like to express my love

and thanks to Brian Szente, my darling husband and fellow graduate student, who

makes it all worthwhile. Our lives intermingle in so many ways.














TABLE OF CONTENTS



ACKNOW LEDGEM ENTS ....................................................................... iii

LIST O F TABLES .................................................................................. vii

LIST O F FIGURES ................................................................................ viii

ABSTRACT ................................................................................... xi

INTRODUCTION ................................................................................. 1

Overview .................................................................................. 1
Staphylococcal Enterotoxin Superantigens
Structure .................................................................................. 6
Function .................................................................................. 14
Superantigen Interaction with MHC Class II Molecules ........ 21
Superantigen Interaction with TCR and Accessory Molecules .... 24
Mis and the Viral Superantigens ................................................ 26
Experim ental Allergic Encephalom yelitis ..................................... 30
The Interferons ........................................................................ 33

MATERIALS AND M ETHODS .......................................................... 35

Synthetic Peptides ..................................................................... 35
Cell Lines and Reagents .......................................................... 35
Radioiodinations ..................................................................... 37
Peptide Com petition Studies ....................................................... 38
Induction of EAE ........................................................................ 39
Injection Schedule for Reactivation of EAE by
Staphylococcal Enterotoxins ................................................ 39
Flow Cytom etry ........................................................................ 39
Proliferation Assays ............................................................ 40
IL-2 Bioassay ..................................................................... 41
Circular Dichroism ..................................................................... 42
Statistical Analysis ........................................................................ 42









RESULTS ............................................................................................. 44

TSST-1 Peptide Binding Studies ................................................ 44
SEB Peptide Binding Studies ................................................ 55
Antagonist activity of SEB(124-154) ..................................... 70
Prevention of EAE by SEB ............................................... 73
Re-activation of EAE by Staphylococcal Enterotoxins ............... 82
Type I IFN Inhibition of Superantigen Activity .......................... 92

DISCUSSION ............................................................................... 107

Structural Studies ..................................................................... 107
Role of Superantigens in Experimental Allergic
Encephalomyelitis ........................................................................ 117
Potential Therapy for Superantigen Associated Disease ........... 123

REFERENCES ................................................................................ 127

BIOGRAPHICAL SKETCH ..................................................................... 141














LIST OF TABLES

Tables Pag

I. The superantigens ...................................................................... 2

II. VB specificities of some microbial superantigens ................ 17

III. Steps in peptide synthesis ................................................ 36

IV. Amino acid sequences of overlapping TSST-1 peptides ......... 46

V. Amino acid sequences of overlapping SEB peptides ................. 57

VI. Treatment with SEB prevents development of EAE in
PL/J m ice .......................................................................... ...... 77

VII. Re-activation of EAE with SEB ................................................ 88

VIII. Re-activation of EAE by SEA ................................................ 91

IX. Inhibition of VB specific SEB-induced T cell proliferation
by the type I IFNs as assessed by flow cytometry ................ 105

X. IFNT is less cytotoxic than other IFNs ...................................... 106













LIST OF FIGURES
Figures Pag

1. The MHC/superantigen/TCR trimolecular complex .................. 5

2. A ribbon model of the three dimensional crystal structure of
S E B ..................................................................................... .... 10

3. A ribbon model of the three dimensional crystal structure of
TSST-1 ......................................................................... ..... 13

4. Predicted composite surface profile of TSST-1 .......................... 48

5. Percent control binding of TSST-1 to Raji and A20 cells
in the presence of TSST-1 ............................................... 50

6. Dose dependent inhibition of TSST-1 binding to Raji cells
in the presence of T(39-68), T(155-194) and T(39-68)S ....... 52

7. Dose dependent inhibition of TSST-1 binding to Raji cells
in the presence of B-chain class II MHC peptides ............... 54

8. Predicted composite surface profile of SEB .......................... 59

9. Percent control binding of SEB to DR1 transfected L cells
in the presence of SEB peptides .................................... 61

10. Dose dependent inhibition of SEB binding to DR1 cells
in the presence of SEB peptides possessing activity ........ 63

11. Percent control binding of SEB to Raji cells in the
presence of SEB peptides ............................................... 65

12. Dose dependent inhibition of SEB binding to Raji cells
in the presence of SEB peptides possessing activity ....... 67

13. Percent control binding of SEB to DR1 transfected L cells
in the presence of N-terminal truncations of the
SEB peptide (179-212) ........................................................... 70








14. Lack of agonist activity of the SEB peptides ........................... 72

15. Ability of SEB peptides to inhibit the proliferation of
HPMC stimulated by SEB ............................................... 75

16. Two-color FACS analysis of spleen cells revealing
VB8 CD4 T cell depletion in SEB protected PL/J mice ........ 79

17. SEB protected PL/J mice that did not develop EAE
are unresponsive to SEB but respond to SEA in vitro ......... 81

18. Time course of reinduction of EAE in PL/J mice
given multiple injections of SEB .................................... 87

19. Induction of T cell energy by superantigen is
prevented by previous activation .................................... 90

20. Type I IFNs inhibit superantigen-induced activation
of H PM C .......................................................................... ...... 96

21. Dose dependent inhibition of superantigen-induced
activation by type I IFNs ........................................................... 98

22. Type I IFNs successfully inhibit proliferation driven
by very high concentrations of SEB ..................................... 100

23. Inhibition of IL-2 activity of HPMC cultured with
SEB in the presence of type I IFNs ..................................... 102

24. Type I IFNs individually inhibit SEB-induced
proliferation as successfully as when tested
in com bination ..................................................................... 104

25. Locations of MHC class II binding regions in
the structure of TSST-1 ........................................................... 110

26. Locations of MHC class II binding regions in
the structure of SEB ........................................................... 114













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


STAPHYLOCOCCAL ENTEROTOXIN SUPERANTIGENS: STRUCTURAL STUDIES
AND ROLE IN AUTOIMMUNE DISEASE



By

Jeanne Margaret Soos

April, 1994



Chairman: Howard M. Johnson
Major Department: Pathology and Laboratory Medicine


The staphylococcal enterotoxins have profound effects on the immune

system, causing massive T cell stimulation and cytokine release. We have studied

the interaction of toxic shock syndrome toxin-1 (TSST-1) and staphylococcal

enterotoxin B (SEB) with the major histocompatibility complex (MHC) class II

molecule, the role of superantigens in experimental allergic encephalomyelitis

(EAE) and the effect of type I interferon (IFN) on superantigen-induced stimulation.

In the structural studies, overlapping peptides of TSST-1 were synthesized and

peptides (39-68) and (155-194) were found to compete with TSST-1 for binding to

Raji and A20 cells. SEB peptides were also synthesized and competition with SEB








was examined on Raji and DR1 transfected L cells. Peptides (1-33), (31-64) and

(179-212) were effective competitors on DR1 transfected L cells while peptides (1-

33), (124-154), (150-183) and (179-212) were effective competitors on Raji cells.

These findings provide insight into superantigen interaction with MHC class II

molecules.

The observation that superantigens stimulate T cells and in some cases

anergize and delete VB specific T cell subsets has raised speculation that

superantigens may play a role in autoimmune disease. We first examined the

ability of SEB to prevent EAE, a murine autoimmune model for multiple sclerosis

(MS). PL/J mice treated with SEB prior to immunization with myelin basic protein

(MBP), the autoantigen involved in EAE and MS, were protected against

development of EAE. Study of the protected mice revealed that the T cell subset

normally responsible for disease had been depleted and anergized. In contrast,

we examined the effect of superantigen after immunization with MBP. We found

that both SEB and SEA were able to reactivate EAE in resolved and asymptomatic

mice. Thus, superantigen is able to both prevent and reactivate EAE, depending

on the time of superantigen treatment relative to immunization with MBP. In an

effort to identify potential therapeutics for superantigen associated disease, the

type I IFNs, a, .f, and r, were shown to inhibit superantigen-induced T cell

stimulation. In all, superantigens and their derivatives may offer a mechanism by

which to modulate autoimmune responses. The effects of the type I interferons on

superantigens may provide yet another level of control over such diseases.













INTRODUCTION

Overview

Superantigens are among the most potent T cell activators known (Langford

et al., 1978). They stimulate as many as 1 in 5 T cells as compared to classical

peptide antigens which stimulate as few as 1 in 10,000 T cells. These unique

molecules are produced by both bacteria and viruses and are presented in Table

I. The prototype for the bacterial superantigens is the family of staphylococcal

enterotoxins. Original studies of these superantigens have been the basis for the

characterization of more recently described superantigens. In addition to the

staphylococci, other organisms that have been shown to produce superantigens

are the group A streptococci, Mycoplasma arthritidis, Mycobacterium tuberculosis,

Yersinia pestis, and Clostridium perfringens. Like some bacteria, certain viruses

are also able to produce superantigens. The first identified virus was the mouse

mammary tumor virus (MMTV), which has served as the prototype for the viral

superantigens. Before its description as a viral product, these superantigens were

thought to be involved as self-superantigens involved in thymic education or as

coligands for higher affinity interaction of certain immune cells (Pullen et al., 1988;

Janeway et al., 1989). Other viruses that have been shown to possess

superantigenic activity are mouse leukemia virus, human spumaretrovirus, rabies,

and Epstein-Barr virus. Many of the bacterial superantigens have been shown to












Table I. The superantigens.

Bacterial

Prototype: Staphylococcal enterotoxins


Organism


Protein


Staphylococcus aureus
Group A streptococci
Mycoplasma arthriditis
Mycobacterium tuberculosis
Yersinia pestis
Clostridium perfringens


Enterotoxins
Pyrogenic exotoxins
T cell mitogen
Not identified
Not identified
Exotoxin


Viral


Prototype: Superantigen from Mouse Mammary Tumor Virus (MMTV)


Virus


Protein


Mouse mammary tumor virus
- type B retrovirus
Mouse leukemia virus
- type C retrovirus
Human spumaretrovirus
- foamy virus
Rabies
Epstein-Barr virus


ORF product

Gag protein

bel 3 gene product

nucleocapsid protein
Not identified








3
play a role in disease caused by their organisms of origin. These maladies include

food poisoning, toxic shock syndrome, and Kawasaki syndrome. They may also

play a role as environmental factors in the relapsing-remitting nature of certain

autoimmune diseases. In the case of the viral superantigens, only MMTV's role

in disease is clear, as it has been shown to be important in virus life cycle. There

has been much speculation concerning the purpose of the other viral

superantigens; however, direct evidence is as yet unavailable.

Recently, the mechanisms by which the bacterial superantigens exert their

effects have been greatly clarified. It has been determined that these molecules

are presented by MHC class II antigens to T cells. However, their mode of

presentation is quite different from that of classical peptide antigen in that the

superantigens bind to the outside of the antigen binding groove, do not require

processing prior to presentation and exhibit very little MHC haplotype restriction

(Carlsson et al., 1988; Fleischer and Schrezenmeier, 1988; Mollick et al., 1989;

Dellabonna et al., 1990). The recognition of superantigen by the TCR is also

unusual. Superantigens bind to the variable region of the B chain of the TCR and

thus stimulate T cells in a VB specific manner (White et al., 1989; Gascoigne and

Ames, 1991). The MHC/superantigen/TCR trimolecular complex is depicted in

Figure 1. VB specific stimulation explains why the superantigens are able to

activate as much as 20 % of a T cell repertoire. It has been shown that the

different superantigens exhibit varying VB specificities. T cell stimulation in this

manner results in massive cytokine release, including interleukin-2 (IL-2),






























Figure 1. The MHC/superantigen/TCR trimolecular complex.







45










S. Antigen
Presenting Cell




Superantigen


r- T cell








6
interferon- (IFN-y), tumor necrosis factor-a (TNF-a) and interleukin-1 (IL-1). Under

certain circumstances, superantigens are also able to anergize and/or delete

specific VB T cell subsets subsequent to initial T cell activation in naive animals

(Kawabe and Ochi, 1990; Rellahan et al., 1990; Kawabe and Ochi, 1991).

The viral superantigens are also capable of the functions described above

for the bacterial superantigens. A stumbling block in the study of the viral

superantigens has been the difficulty of their isolation. Thus, protein

characterization, in terms of direct binding studies, for example, has remained

elusive. Clearly the superantigens, both bacterial and viral, have profound effects

on the immune systems of their victims, and the elucidation of their mechanisms

of action will enable the development of therapeutics for many of the syndromes

for which they are responsible.

Staphylococcal Enterotoxin Superantigens

Structure

The staphylococcal enterotoxins are a family of structurally related proteins

that are single chain molecules with molecular weights of approximately 24 to 30

kDa. They are acid and heat stable and rich in threonine, serine and aspartic acid

residues. Thet are also charged with pis ranging from 7.0 to 8.6. All of the

members of the staphylococcal enterotoxin family, with the exception of TSST-1,

contain a centrally located disulfide loop, the function of which remains in question.

In the cases of SEA, SEB and SEC, enzymatic cleavage in the region of the

disulfide bridge had no effect on biological activity (Spero et al., 1973; Noskova et








7
al., 1984). However, the stimulatory activities of these toxins based on the disulfide

bridge have been dissociated for T cells and monocytes (Grossman et al., 1990).

The reduction and alkylation of the staphylococcal enterotoxins affected T cell

activation but had no effect on monocyte production of interleukin-1 (IL-1) and

tumor necrosis factor-a (TNF-a). The staphylococcal enterotoxins have been

serologically classified into five groups A through E, with SEC divided further into

three subtypes on the basis of minor epitopes (Bergdoll, 1985). Another member

of this toxin family was originally designated SEF but was later renamed TSST-1

(Bergdoll, 1985). The amino acid sequences of the staphylococcal enterotoxins

exhibit appreciable homology (Betley and Melakanos, 1988). SEA, SEE and SED

are related in amino acid sequence, while SEB and SEC share greater homology

with each other. Two regions of these toxins, residues 106 to residue 119 and

residue 147 to residues 163, have been identified as highly conserved (Betley and

Melakanos, 1988; landolo, 1989).

While sequence homology of the staphylococcal enterotoxin family members

range from as little as 30 % to as much as 86 %, all of the toxins regardless of

homology exhibit similar biological activity (Betley and Melakanos, 1988). Such

common biological function suggests that their activity is based on similar

secondary and tertiary structures. Initial studies using such techniques as circular

dichroism (CD) and tryptophan quenching found that the staphylococcal

enterotoxins characteristically have low a helical content together with a high

content of B structure (Singh et al., 1988a). In general, the more closely related








8
4
toxins exhibit greater similarity in their CD spectra. It is of interest that TSST-1,

which shares only minimal sequence homology, also has the low a helix and high

B sheet content characteristic of the other staphylococcal enterotoxins (Singh et

al., 1988b). This argues strongly that the similar functions of these toxins are

maintained at the secondary and tertiary structural levels.

Recently, the three dimensional crystal structures of SEB and TSST-1 have

been determined (Swaminathan et al., 1992; Acharya et al., 1994). In the case of

SEB, the structure consists of a main chain fold containing two domains, which

suggests a general motif for the other enterotoxins (Figure 2). The first domain is

composed of residues 1 to residue 120 and contains two B sheets and three a

helices. Two of the B stands form a cylindrical barrel forming a crisscross pattern.

The disulfide loop is present in this domain in residues 99 to 105 extending into the

solvent. The second domain is composed of residue 127 to residue 239 and

contains 2 a helices and 7 B strands. The second domain is more complicated

and actually contains an infrequently observed left-handed crossover of B strands

6 and 12.

A shallow cavity, formed by the two structural domains of SEB, is

considered to be a binding site for the TCR. This is supported by a previous study

that identified residues important for TCR binding (Kappler et al., 1992), and it has

been shown that these residues lie along the sides of the cavity. Upon close

inspection of the structure, three adjacent loops at the edge of the cavity appear

to collectively form a structural unit important for TCR interaction. Of the 21































Figure 2. A ribbon model of the three dimensional crystal structure of SEB.








10








11
residues of SEB suggested to be important for TCR binding, 18 residues are

conserved between SEB and SEC1. Thus, it makes sense that these two

superantigens stimulate the same VB specific TCRs with the exception of one.

The TSST-1 structure is a monomeric, two domain, compact molecule

(Figure 3). The N-terminal domain of TSST-1 is composed of an a helix and

multiple polypeptide folds of B strands that form a continuous "roll" or what is

known as a B-barrel. An extended chain caps an end of the B-barrel. The C-

terminal domain is composed of one relatively long a-helix, which is cupped by a

highly twisted B-sheet forming an elaborate B-grasp motif (Murzin, 1992). The a-

helix present in the C-terminal domain appears to serve as the backbone of the

molecule.

The topology of TSST-1, as determined by its three dimensional structure,

is similar to the structure of SEB in that it is composed of two domains and exhibits

a high content of B-sheet relative to a-helix. However, a number of differences in

the TSST-1 and SEB structures exist. Among these differences are 1) the topology

of the TSST-1 C-terminus is unique compared to SEB as it contains a B-grasp

motif; 2) TSST-1 is extensively truncated in a number of loop regions of SEB,

including the cysteine loop, which is consistent with the fact that TSST-1 does not

contain a disulfide bridge like the other staphylococcal enterotoxins; 3) there is no

structural equivalent of the N-terminus of SEB in the TSST-1 structure; and 4) the

TSST-1 structure contains even fewer a-helices than the SEB structure. Thus, a

general motif is apparent from the SEB and TSST-1 3-dimensional structures, but






























Figure 3. A ribbon model of the three dimensional crystal structure of TSST-1.
The blue and red balls represent N and C termini, respectively.








13









414
certain structural features that appear important for SEB are not present in TSST-1

probably due to the lack of TSST-1 sequence identity with the other staphylococcal

enterotoxins.

Studies are presently underway for the high-resolution determination of the

MHC/superantigen/TCR trimolecular complex (Brown et al.,1993). It has been

suggested that the MHC and superantigen molecules involved in this study are

HLA-DR1 and SEB. If such information is obtained, it will prove extremely useful

in understanding the superantigen mechanism of pathogenicity and provide a

strong basis for the development of engineered therapeutics that can target

specific epitopes identified from the sites of interaction within the trimolecular

complex.

Function

The first observation that a staphylococcal enterotoxin, namely SEB, was

able to induce mitogenesis of human peripheral lymphocytes was made in 1970

(Peavy et al., 1970). The staphylococcal enterotoxins were subsequently shown

to be specific for the activation of T cells (Johnson and Bukovic, 1975). In addition

to their ability to induce mitogenesis in T cells, they are potent inducers of a

number of cytokines, including IFN-y, IL-2 and TNF (Langford et al., 1978; Carlsson

and Sjogren, 1985) and have been implicated in the suppression of antibody

responses and certain T cell responses (Torres et al., 1982; Papermaster et al.,

1983). In order to induce activation, the staphylococcal enterotoxins are required

in amazingly low concentrations (Langford et al., 1978). For example, SEA, the








15
most potent of the staphylococcal enterotoxins, is able to stimulate maximal DNA

synthesis and production of IFN-y at concentrations of 3.5 X 10'13 and 3.5 X 10-10

M respectively (Langford et al., 1978), making the staphylococcal enterotoxins

among the most powerful T cell activators known.

During the 1980s, further characterization of superantigen-induced T cell

activation continued. The first clue that the microbial superantigens were actually

a type of antigen came when it was discovered that antigen presenting cells and

more specifically, the MHC class II molecules on their surface were required for

presentation of these unique antigens to T cells (Carlsson et al., 1988; Fleischer

and Schrezenmeier, 1988; Mollick et al., 1989). (The term superantigen was later

given to these enterotoxins.) However, the manner in which the MHC class II

molecule presents superantigen is quite different from that of a classical peptide

antigen. First, presentation of microbial superantigens exhibits very little MHC

haplotype restriction. Second, the microbial superantigens do not require

processing prior to presentation by MHC class II molecules as evidenced by the

fact that paraformaldehyde fixed APCs are as proficient in presenting superantigen

as their untreated counterparts. Third, superantigens have been shown to bind to

the outside of the antigen binding groove of MHC class II molecules, in contrast

to nominal peptide antigens which are known to bind inside the groove

(Dellabonna et al., 1990; Russell et al., 1990). Thus, the manner in which

superantigens are presented by the MHC class II molecule proved to be quite

unique and provided insight into how they are recognized by T cells.








16
If the binding of superantigen to MHC class II molecules is so unusual, then

how are the superantigens recognized by the TCR? With the advent of

monoclonal antibodies specific for the VB elements of the TCR, much insight was

gained about how the microbial superantigens stimulate T cells. It was shown in

vitro that SEB stimulated specific T cell subsets based upon the VB region of the

TCR (White et al., 1989). For example, SEB was shown to stimulate murine T cells

bearing VBs 3, 7, 8, 11, and 17 (Callahan et al., 1989). This VB specific T cell

stimulation was also shown to hold true in the human system (Kappler et al.,1989).

In addition to in vitro experiments, one of the most interesting pieces of evidence

in the stimulation of human VBs came from studies of patients suffering from toxic

shock syndrome. A number of years earlier, it had been shown that the pathogen

responsible for this sometimes fatal disease was a newly isolated strain of

Staphylococcus aureus that produced a toxin aptly titled toxic shock syndrome

toxin-1 (TSST-1) (Bergdoll et al., 1981). Patients infected with TSST-1 producing

S. aureus, often via tampon use, exhibited as great as a 60 % expansion of their

VB 2 bearing T cells in both the CD4 and CD8 subsets during the acute phase of

infection (Abe et al., 1992). The levels of VB 2 bearing T cells remained elevated

until resolution of disease symptoms when the expanded subset exhibited a

gradual return toward normal levels. Subsequent studies have determined the Vs

specificities of the microbial superantigens in both the murine and human systems

(Table II).









4
Table II. VB specificities of some microbial superantigens.


VB specificity


Toxin Mouse Human


1, 3, 10, 11, 17

7, 8.1-8.3, 11, 17

3, 8.2, 8.3, 11, 17

3, 8.2, 10, 17

7, 8.2

3, 11, 17

11, 15, 17

15, 16

10, 11, 15

ND

ND

ND

5.1, 6, 8.1-8.3


1, 5, 6's, 7.3-7.4, 9.1

3, 12, 14, 15, 17, 20

3, 6.4, 6.9, 12, 15

12, 13.2, 14, 15, 17, 20

3, 5, 12, 13.2

5, 12

5.1, 6's, 8, 18

2

2

8, 12, 14, 15

2, 8

1, 2, 5.1, 10

3, 17


ND-not determined or undefined.


SEA

SEB

SEC1

SEC2

SEC3

SED

SEE

TSST-1

ExFT

SPE-A

SPE-B

SPE-C

MAM








4 18
For some time the only immunomodulatory activities of the microbial

superantigens known were their ability to stimulate T cells and their involvement in

certain immune suppression responses (Papermaster et al. 1983). In addition to

these, they are also capable of inducing energy (rendering T cells nonfunctional)

and depletion (reduction in the number of T cells) of VB specific T cell subsets.

Original studies of induction of specific clonal energy by SEB were shown in vitro

and in vivo and were the first examples of peripheral T cell tolerance (Kawabe and

Ochi, 1990; Rellahan et al., 1990). Studies in vitro showed that a number of

staphylococcal enterotoxins including SEA, SEB, SEC1-3, and SED were able to

induce specific clonal energy (O'Hehir and Lamb, 1990). Cloned human T cells

specific for hemagglutinin (307-319) were exposed to the various staphylococcal

enterotoxins for a 16 hour period and then challenged with peptide antigen. The

T cells failed to respond to their natural ligand when pretreated with superantigen

even in the presence of exogenous IL-2.

It was shown in vivo that adult mice administered SEB exhibited a profound

state of energy in their V88 T cell subset. Such energy was evidenced by their

inability to proliferate to subsequent in vitro stimulation with either SEB or anti-V88

antibodies. V88 T cells from SEB primed mice also failed to proliferate in response

to exogenous IL-2 indicating a defect in their IL-2 responsiveness. This same T

cell subset also exhibited reduced activity in primary cytotoxicity assays. Induction

of energy by SEB was shown to occur primarily in the VB8 CD4 T cell subset but

not in the CD8 subset. In these studies, energy induced by SEB was shown to








4 19
last at least two weeks while data presented in this dissertation suggest that

abrogation of response to SEB may last beyond 40 days after initial exposure to

superantigen. It is important to point out that superantigen-induced energy occurs

in adult naive mice. However, mice that have been primed with another antigen

prior to exposure to SEB may not be so readily anergized. Data presented in this

dissertation show that mice immunized with another antigen, in this case myelin

basic protein (MBP) in complete Freunds adjuvant (CFA) or CFA alone, are not

susceptible to the induction of energy in their VB8 T cell subset. Thus, it appears

that the circumstances and conditions of the subject can influence the induction

of energy by superantigen.

Deletion of VB8 CD4 T cells in adult naive mice primed with SEB has also

been observed (Kawabe and Ochi, 1991). Mice exposed to SEB first exhibit an

increase in both the VB8 CD4 and CD8 subsets on days one through four after

exposure. By day seven the VB8 CD8 subset returned to near normal levels. In

contrast, the levels of VB8 CD4 T cells were reduced by half of normal prior to

exposure to SEB. Deletion by superantigen was determined to occur in the

periphery as a reduction in the VB8 CD4 subsets occurred in both normal and

thymectomized mice. The mechanism of deletion was determined to occur via

programmed cell death or apoptosis. By day four after exposure to SEB, the VB8

CD4 subset, while still detectable by flow cytometry, exhibited genomic DNA

fragmentation associated with programmed death. Extremely low concentrations

of superantigens are also able to delete VB specific T cells although the events








20

prior to deletion are different compared to deletion induced by higher doses of

superantigen. When mice are chronically exposed to very low doses of

superantigen, i.e., multiple injections within a defined time period, the initial T cell

stimulation observed in previous deletion studies was not seen. Instead dramatic

deletion of the target subsets was observed (McCormack et al., 1993). Thus

dosage of superantigen may also greatly influence the course of deletion of

specific VB T cells.

The critical determinant leading to clonal deletion has been shown to be one

of the cytokines strongly induced by superantigen, namely IL-2 (Lenardo, 1991).

Experiments with the antigen specific cell line, A.E7, showed that exposure to IL-2

prior to antigen specific stimulation results in programmed cell death, while

exposure to IL-2 after antigen specific stimulation results in proliferation. In mice

injected with SEB, an anti-IL-2 antibody that blocks binding to the IL-2 receptor a

chain inhibited the marked reduction of VB8 T cells normally observed in previous

deletion studies. Therefore, superantigen-induced IL-2 is integral to VB specific T

cell deletion. It would appear that energy and deletion of VB specific T cells by

superantigen occurs concomitantly with approximately 50 % depletion of a VB

specific subset with the remainder of the subset undergoing energy. Whether the

entire remainder of the subset is anergized or if certain cells are able to escape

energy remains to be determined.

Only limited study of the energy and deletion properties of the microbial

superantigens in humans has been conducted. To date, no studies of patients








21
suffering from food poisoning have been undertaken. In the case of toxic shock

syndrome, after expansion of the VB2 subset occurs, levels of V32 bearing T cells

return to normal. No depletion of VB2 bearing T cells was observed at any time

point after disease resolution. The ability of superantigen to induce energy and/or

deletion in the human system remains to be studied in depth and much may

depend on the timing of exposure and dosage of superantigen.

Superantigen Interaction with MHC Class II Molecules

Although the staphylococcal enterotoxins had been determined to be

responsible for a large percentage of food poisoning cases, their mode of action

has remained in question for some time. This was partly resolved by studies

showing that SEA bound to murine lymphocytes (Buxser et al., 1981). SEA was

shown to bind to a single class of receptor, and via inhibition studies it was also

determined that SEA and SEE bound to the same receptor. Subsequent to this

study it was shown that paraformaldehyde fixed antigen presenting cells were as

efficient in presenting the bacterial toxins as were untreated cells suggesting that

the enterotoxins do not require processing prior to presentation (Carlsson et al.,

1988). The specific receptors for the staphylococcal enterotoxins on antigen

presenting cells were shown to be the MHC class II molecule by a number of

studies including 1) use of anti-MHC antibodies for inhibition of binding (Fleischer

and Schrezenmeier 1988), 2) direct binding to purified MHC class II molecules (Lee

and Watts, 1990), 3) genetic analysis illustrating the necessity of I-E expression

(Cole et al., 1981), and 4) use of MHC class II negative cells transfected with MHC








22

class II genes (White et al., 1989). Quantitative binding studies with a number of

the staphylococcal enterotoxins have determined their affinities for HLA and MHC

class II molecules (Buxser et al., 1981; Fraser, 1989; Mourad et al., 1989). In

general, the staphylococcal enterotoxins have a higher affinity for the HLA

molecules than for the MHC class II molecules. Of all the staphylococcal

enterotoxins, SEA exhibits the highest affinity for the class II molecules in both the

HLA and MHC systems, which may assist in explaining why SEA is the most

potent of all of the staphylococcal enterotoxins. Further characterization of

superantigen MHC interaction showed that this binding is quite unusual compared

to classical peptide antigen. In addition to the lack of processing, superantigens

bind to the outside of the antigen binding groove (Dellabona et al., 1990). The

toxins also exhibit a lack of MHC restriction. However, it has been shown that

alleles of HLA-DR differ in their ability to present the staphylococcal enterotoxins

to T cells (Herman et al., 1990). Thus, while there is lack of MHC restriction, the

enterotoxins can exhibit haplotype preference.

A number of questions have arisen regarding the requirement of MHC

binding by superantigen and the ultimate stimulation of VB specific T cell subsets.

Do MHC and TCR interactions influence superantigenic stimulation or does binding

to MHC class II induce a conformational change in the superantigen that allows for

binding to the TCR? In the cases of two superantigens, SEA and SEE, it has been

shown that zinc is required for the stabilization of the binding domain of MHC class

II molecules but is not directly involved in the interaction with the MHC class II1








4 23
molecule (Fraser et al., 1992). The experiments involving zinc as a regulatory

factor provide insight into the mode of action of these two superantigens. Binding

of SEA and SEE, but not SEB or TSST-1, was abolished by EDTA and

reconstituted in the presence of Zn2". Zinc binding appears to be an essential first

step in the formation of the MHC class II binding domain of SEA and SEE. That

zinc is able to produce an active site on SEA for MHC class II binding suggests

that SEA is not a wholly inflexible molecule as previously thought due to its total

resistance to protease and heat. Unlike SEA, TSST-1 did not require zinc for MHC

class II binding, which may be directly related to the more severe consequences

of toxic shock syndrome compared to food poisoning. The involvement of zinc

and the requirement of MHC class II for binding of superantigen to the VS region

of the TCR suggests two possible explanations. The first of these is that binding

of superantigen to MHC class II induced some type of structural modification that

reveals a cryptic site for interaction with the TCR. The second is that superantigen

binding induces a modification of the MHC class II molecule required for high

affinity interaction with the TCR Vj region.

Binding regions of the MHC class II molecule involved interaction with the

staphylococcal enterotoxins have been the subject of intense study. Competitive

binding studies have determined that TSST and SEB bind to two different sites on

HLA-DR. Both of these sites are also bound by SEA, SED, and SEE (Scholl et al.,

1989; Scholl et al., 1990; Chintagumpala et al., 1991; Pontzer et al., 1991b). It is

of interest that SEB and TSST-1 are not closely related by amino acid sequence








24
while SEA, SED and SEE, all of which bind to the same sites, share more that 70

% amino acid sequence homology. Several studies have suggested that there are

approximately three distinct, although possibly overlapping, binding sites on HLA-

DR for these bacterial toxins. Further examination of superantigen/MHC class II

interaction has revealed the specific binding sites for a number of the

staphylococcal enterotoxins. SEA, via direct involvement of its N-terminus, was

shown to bind to region 65 through 85 of the B chain of I-Ab (Russell et al., 1990).

Sites in this region of the f3-chain, specifically amino acid residues 72, 80 and 81,

were shown to be important for SEA binding by the synthetic peptide approach

and site directed mutagenesis of DR transfected L929 cells (Russell et al., 1991;

Herman et al., 1991). The a chain of the I-Ab molecule was also shown to be

involved in SEA binding (Russell et al., 1991). Studies of SEE binding also point

to the importance of histidine 81 of the B chain of HLA-DR (Karp and Long, 1992).

In the case of TSST-1, both the a and B chains of both HLA and MHC class II

molecules have been shown to contribute to binding of this toxin (Braunstein et al.,

1992). Interpretation of these studies suggests that residues of the two helices

together define the binding site for TSST-1 and that only one binding site for TSST-

1 is present on a single class II molecule. This is supported by the crystal

structure of TSST-1 (Acharya et al., 1994). Thus, multiple superantigen binding

regions are present on the MHC class II molecules.

Superantigen Interaction with TCR and Accessory Molecules

The manner in which superantigens stimulate T cells is unique in that they








25

cause the activation of many or all T cells bearing specific TCR B chain variable

region elements (White et al., 1989). Superantigen interaction with the TCR has

been of interest, however, such studies have proved more difficult due to the

requirement of superantigen binding to MHC class II prior to recognition by the

TCR. Initial studies involved the production of a chimeric TCR in which products

of human VB genes were introduced into a murine TCR expressed on the surface

of a mouse T cell hybridoma (Choi et al., 1990a). Residues 67 through 77 of

human VB 13.2 were determined to be important for the function of SEC2. Direct

binding of superantigen to the TCR was demonstrated by the use of a secreted

form of the TCR B chain of VB 3 specificity (Gascoigne and Ames, 1991). SEA, in

the presence of Raji cells, was shown to bind directly to the secreted B chain and

was dependent on the presence of the MHC class 11 molecule as SEA did not bind

the B chain in the absence of Raji cells. This was the first report of direct TCR

binding with a natural ligand and provided clear evidence that presentation of

superantigen by MHC is a requirement for interaction with the VW region of the

TCR. Studies to define the region of the variable portion of the B chain have

identified amino acid residues 57 through 77 of VB3 of the TCR as being

responsible for SEA/TCR interaction (Pontzer et al., 1992). These residues lie in

the CDR4 region of the TCR. The CDR4 region forms part of the side of the B

chain, exposed to the aqueous phase and well away from the classical peptide

antigen/MHC class II binding site or any site of interaction between the Va and VB

chains. Thus, the trimolecular complex appears to involve MHC and TCR








26
interaction with the superantigen on the external sides of the MHC 8 chain a helix

with amino acids between residues 57 and 77 of the TCR.

Interestingly, the observation has been made that there is only a limited

requirement for accessory molecules on T cells for staphylococcal enterotoxin

stimulation. Clearly, CD4 and CD8 molecules do not play a role, as T cell clones

and hybridomas lacking these two accessory molecules usually respond to the

bacterial superantigens (Fleischer and Schrezenmeier, 1988; Fleischer and

Schrezenmeier, 1989). The one set of accessory molecules that has been shown

to be directly involved in superantigen stimulation is ICAM-1/LFA-1 (van Seventer

et al., 1991). This was evidenced by the use of a solid matrix assay in which

soluble MHC class II, SEA and ICAM-1 were present and T cell stimulation was

measured. In the absence of soluble ICAM-1, SEA-induced stimulation did not

occur. When soluble ICAM-1 was added to the solid matrix assay, SEA-induced

stimulation of T cells was close to that of previously observed levels under normal

conditions. The role of another set of accessory molecules, B7/CD28, remains

uncertain as a number of conflicting reports have been made concerning their

effect on superantigen-induced activation (Fischer et al., 1992; Green et al., 1992;

Damle et al., 1993).

MIs and the Viral Superantigens

The endogenous superantigens originally termed minor lymphocyte

stimulatory (Mis) determinants were first detected based on their ability to stimulate

a strong, primary mixed lymphocyte reaction (MLR) between cells from mice








27

bearing the same MHC haplotype (Festenstein, 1973). In the MLR, naive T cells

were shown to respond to MIs determinants presented by MHC class II molecules

on the surface of various cell types (Peck et al., 1977). This stimulation by MIs

was shown to be greater than that induced by the products of MHC in a primary

MLR (Wilson et al., 1968; Janeway et al., 1980; Lutz et al., 1981; Miller and

Stutman, 1982). Original studies of MIs murine strain distribution suggested that

the Mis locus expressed four alleles, a, b, c, and d, and that these alleles encoded

polymorphic determinants with variable stimulatory activity (Festenstein, 1974).

Other non-MHC related determinants were also designated as MIs determinants

based upon their MLR stimulatory activity such as Misx in PL/J and MIs" in C3H/Tif

(Coutinho et al., 1977; Janeway and Katz, 1985).

With the advent of monoclonal antibodies specific for the VS elements of the

TCR, important information was gained about the mechanism by which the Mis

determinants induce T cell stimulation. It was shown in vitro that naive T cells

bearing particular VB elements respond to Mis determinants and that different MIs

determinants were specific for particular V3s. In contrast, in vivo studies showed

that T cells bearing specific Vj3s were clonally deleted during development in the

thymus of mice expressing particular MIs determinants (Blackman et al., 1990).

The observation that MIs determinants could clonally delete thymic T cells led to

much speculation and raised the question of MIs determinants as self

superantigens. The self superantigen hypothesis suggested that mice expressing

self superantigens would remove responsive T cells from the periphery; this








28

elimination would occur via clonal deletion during thymic education resulting in self

tolerance. It had been shown that EAE and induced rheumatoid arthritis were

chiefly caused by T cells expressing VB8.2 and VB6, respectively (Acha-Orbea et

al., 1988). Thus, the deletion of VB specific T cell subsets by self superantigens

could confer protection against autoimmunity as well as the ill effects of bacterial

toxins. Another hypothesis on the function of MIs was that the as yet unidentified

MIs protein served as a coligand between the TCR and MHC, its purpose being

to potentiate T cell responses, allowing such responses to occur more rapidly

(Janeway, 1990). The nagging question that remained was why had structures

similar to MIs not been detected in humans. Ultimately, the answer to the true

identity of the MIs determinants proved somewhat surprising.

The Mis determinants are actually a collection of products of the 3' open

reading frame encoded in the long terminal repeat of the murine retrovirus, mouse

mammary tumor virus (MMTV) (Dyson et al., 1991; Frankel et al., 1991; Marrack

et al., 1991; Woodland et al., 1991). MMTVs are known to be type B retroviruses

responsible for the induction and transmission of mammary carcinoma in mice

(Heston et al., 1945). Endogenous MMTV proviruses are present in the germ line

of all inbred mice. Different strains of mice harbor distinct proviruses at multiple

locations in their genome. Many are defective and unable to produce infectious

viruses while some can be activated in the mammary gland, resulting in the

shedding of viral particles into milk (Kozak et al., 1987). Thus, the transmission of

MMTV superantigen can occur by both inheritance and milk-borne infection.








29

The structural characterization of the MMTV superantigen, MMTV-7,

suggests that it is probably a 45K type II integral membrane protein with an

intracellular N terminus and an extracellular, glycosylated C terminus (Choi et al.,

1992; Korman et al., 1992). Gene truncation experiments showed that the N

terminus was important to the activity of MMTV-7 but its function was unclear (Choi

et al., 1992). Subsequently, it was shown that the MMTV-7 superantigen may be

synthesized as a precursor protein, undergo proteolytic cleavage and be

expressed as an 18.5K surface protein (Winslow et al., 1992). Such proteolytic

cleavage would involve the removal of the N-terminal transmembrane domain

resulting in the expression of the C-terminal residues that are somehow "tethered"

to the membrane. Such tethering to the cell surface may occur to MHC class II

or the N terminal portion of MMTV-7 via noncovalent associations or via covalent

lipid addition. However, it has been shown that intracellular transfer of MMTV-7

does take place, but how the transfer of a type II integral membrane protein or

tethered protein occurs is difficult to explain. One potential explanation is that the

mature, functional form of the viral superantigen protein is a soluble protein

although this has yet to be proven. While MMTV-7 superantigen has been

detected on the surface of B cells, the only studies that have shown direct binding

of MMTV-7 have employed the synthetic peptide approach. It was shown that

residues 76 through 119 of MMTV-7 are responsible for binding to MHC class II

and that MMTV-7 shares an MHC class II binding site with the bacterial

superantigen, SEA (Torres et al., 1993). In this study, the labelled MMTV-7 peptide








4 30
ORF(76-119) was shown to bind directly to murine B cells and an MHC class II

peptide encompassing the alpha helical residues 60 through 90 of the beta chain.

Using antipeptide antibodies, the immediate C-terminal residues 310 through 322

of MMTV-7 were found to be important for interaction with the TCR (Winslow et al.,

1992). While recent studies have provided information regarding regions important

for binding and function, direct evidence as to the mature form of the MMTV-7

protein remains elusive.

Experimental Allergic Encephalomyelitis

The animal model experimental allergic encephalomyelitis (EAE) is a

prototype for antigen specific T cell mediated autoimmune disease (Gonatas and

Howard, 1974; Ortiz-Ortiz and Weigle, 1976). The primary autoantigen involved in

the induction of EAE is myelin basic protein (MBP). MBP is a predominant protein

present in myelin in the central nervous system (CNS). The mediators of

pathogenesis in EAE are MHC class II restricted, CD4 T lymphocytes that are MBP

specific. EAE can be induced in a number of species, and certain forms are

characterized by relapsing paralysis. Histopathology studies have demonstrated

the infiltration of perivascular lymphocytes and demyelination in the CNS. The

characteristics of EAE suggest it as the primary model for the human

demyelinating disease, multiple sclerosis (MS) (Alvord, 1984; Raine, 1983).

The first description of EAE was made when adverse allergic responses to

the original rabies vaccine developed by Louis Pasteur were observed (Remlinger,

1905). The original vaccine consisted of fixed rabies virus that has been grown in








31

rabbit CNS tissue. An extremely small percentage of those who received the

vaccine developed a monophasic paralysis termed acute disseminated

encephalomyelitis (Stuart and Krikorian, 1928). Considering that paralysis is not

a symptom associated with rabies infection and that individuals exposed to the

vaccine but not infected with rabies virus also developed paralysis suggested that

tissue from the CNS which contaminated the vaccine was responsible for the

paralytic illness (Einstein et al., 1962). Initial studies showed that immunization with

CNS tissue alone could induce demyelinating encephalomyelitis and that EAE

could be induced in a wide range of species including mice, rats, guinea pigs,

monkeys, sheep, dogs and chickens (Stuart and Krikorian, 1928; Paterson, 1976;

Martonson, 1984).

Lymphocytes, as mediators of EAE, were first implicated by experiments in

which anti-lymphocyte antibodies inhibited induction of EAE (Waksman et al.,

1961). Further evidence that T cells were involved stemmed from the observation

that thymocytes are required for EAE induction (Arnason et al., 1962). More

specifically, the T helper subset appeared to be the primary candidate as shown

by the elimination of Lyl T cells preventing the transfer of EAE to naive recipients

(Brostoff and Mason, 1984). Conformation of the involvement of CD4 T cells is

their abundant presence in inflammatory EAE lesions in the CNS (Traugott et al.,

1986).

As MBP is the primary autoantigen in EAE, the various epitopes of the

protein have been the subject of intense study. Using peptic fragmentation and








32

the synthetic peptide approach, epitopes of MBP that are encephalitogenic have

been identified. Different species appear to respond to different regions of MBP.

In the H-2" mice, PL/J and B10.PL, the amino terminal region 1-37 can induce EAE

while the H-2' mice, SJL/J and A.SW, recognize MBP 89-169 as an

encephalitogenic epitope (Fritz et al., 1983; Fritz et al., 1985). The

encephalitogenic determinant in the Lewis rat strain has been shown to be MBP

68-88. Further characterization of these regions has been conducted and specific

amino acid residues have been identified as being responsible for conferring

encephalitogenicity (Zamvil et al., 1986). An intriguing discovery in the T cell

subsets responsive to some epitopes of MBP demonstrated restricted TCR usage.

The CD4 T cells in PL/J mice and Lewis rats that mediate EAE are VB8.2 specific.

Experiments employing anti-VB8.2 antibodies have successfully prevented the

induction of EAE (Acha-Orbea et al., 1988). Also, peptides of the V88.2 TCR were

shown to be able to block induction of EAE in Lewis rats. Thus, insight into T cell

responses to MBP has provided clues to the mechanism of such pathogenesis.

The EAE model has also served as a vital means for testing of novel forms

of immunotherapy. Among the most recently tested and successful treatments are

anti-TCR antibodies, anti-MHC antibodies, anti-CD4 antibodies, T cell vaccination

and peptide therapy (Steinman et al., 1983; Brostoff and Mason, 1984; Howell et

al., 1989; Vandenbark et al., 1989). Development of effective and safe forms of

therapy in EAE is the first step toward testing potential therapies in human

autoimmune disease.









4
The Interferons

The family of secretary proteins known as interferons (IFNs) were first

described in 1957 by Issacs and Lindenmann (Issacs and Lindenmann, 1957).

Their name is derived from their ability to interfere with virus replication. In addition

to their antiviral effects, these molecules possess many varied activities including

those which are antimicrobial, antitumor, and immunomodulatory (Pestka et al.,

1987). Advances in the understanding of the basic mechanism of IFN actions and

a number of clinical trials have led to Food and Drug Administration approval of

IFN as treatment for several diseases including hairy cell leukemia, condyloma

acuminatum, acquired immune deficiency syndrome (AIDS) related Kaposi's

sarcoma, type C hepatitis, and most recently multiple sclerosis.

The IFN family of molecules are divided into three main species: IFNa, IFNB,

and IFN-. Two other types of IFNs, IFNr and IFNw, are close relatives of IFNa.

These molecules can be further classified into two distinct groups. The type I IFNs

include IFNs a, B, T, and w, and are induced primarily by viruses and tumor cells

while type II (immune) IFN or IFN-y is induced by antigens and mitogens that

stimulate T and NK cells. There are as many as 25 species of structurally similar

forms of IFNa, encoded by a family of as many IFNa genes (Allen, 1982). A

number of IFNa classes exist and recently the IFNa2 family was renamed IFNo.

Unlike IFNa, IFNB is encoded by a single IFNB gene and the mature protein is

composed of 166 amino acid residues. It was thought that an additional IFNB

subtype had been identified. However, the protein possessed no antiviral activity










and is presently referred to as interleukin-6 (Van Damme et al., 1987).

While the first described IFNs have been well studied, the type I IFN, IFNT,

has only been recently characterized. IFNT was first identified as a major

concepts secretary protein in sheep, and in its reproductive function, it is

important for its antileuteolytic properties (Godkin et al., 1982). IFNT shares 45-55

% amino acid sequence homology with various IFNas and is as potent as any

IFNa in its antiviral activity (Imakawa et al., 1987; Pontzer et al., 1988).

IFN-y is a pleiotropic lymphokine with numerous unique biologic effects.

IFN- is predominantly made by TH1 cells and CD8 cytotoxic T lymphocytes

(Salgame et al., 1991). IFN7-, while able to exert antiviral activity, is not as potent

as the type I IFNs. Other biological activities of IFN-y include, macrophage

activation, upregulation of MHC class I and II on macrophage and B cells, and B

cell maturation (Pestka et al., 1987).

The suppressive and antiproliferative effects of the IFNs as applied to

cellular growth and viral replication, coupled with their potent immunomodulatory

effects render them as attractive agents for use in combating autoimmune disease.

This is evidenced by the amelioration of the relapsing-remitting nature of multiple

sclerosis by IFNB (IFNB Multiple Sclerosis Study Group, 1993). Furthermore, if

superantigens do, as speculated, play a role in the initiation or persistence of

autoimmune disease, then the suppressive effect of IFNs on superantigen-driven

activation, presented in this dissertation, may form the basis of future therapies for

autoimmune disease.













MATERIALS AND METHODS

Synthetic Peptides

Overlapping peptides corresponding to the entire sequence of both TSST-1

and SEB were synthesized on a Biosearch 9500AT automated peptide synthesizer

(Milligen/Biosearch, Burlington, MA) using 9-fluoroenylmethyl oxycarbonyl (FMOC)

chemistry (Chang and Meienhofer, 1978). N-Terminal truncations of SEB(179-212)

were generated by removal of the peptidyl resin from the reaction vessel when the

desired sequence length was obtained. Peptides were cleaved from the resins

usingtrifluoroacetic acid/ethanedithiol/thioanisole/crystalline phenol/distilledwater

at a ratio of 80/3/5/7/5. The cleaved peptides were then extracted in ether and

subsequently dissolved in water and lyophilized. Reverse phase high performance

liquid chromatography (HPLC) (Perkin Elmer, Norwalk, CT) analysis of the crude

peptides revealed one major peak in each profile. Amino acid analysis of the

peptides, performed by the University of Florida Protein Core Facility, showed that

the amino acid composition corresponded closely to theoretical values.

Cell Lines and Reagents

Three cell lines were employed for the structural binding studies. The Raji

is an EBV-transformed human B cell line that bears DR3, DRw10, DOwl and DQw2

(Merryman et al., 1989). The A20 line is a murine B cell line that bears I-Ad and I-Ed

(Russell et al., 1990). Both the Raji and A20 lines were obtained from ATCC











Table III. Steps in peptide synthesis


Step


Effect


Step 1: Deprotection


Step 2: Neutralization
and activation



Step 3: Coupling


Step 4: Capping


Removal of Fmoc group


Conversion of a-amino
group from protonated to
deprotonated form


Coupling of amino acid to
nascent peptide chain

Multi-coupling if poor
coupling efficiency is
observed


Permanent acylation of
unreacted a-amino groups,
elimination from further
participation in synthesis








37
(Rockville, MD). DRi-transfected L cells were kindly provided by Dr. Eric 0. Long

(National Institutes of Health), and are described elsewhere (Long et al., 1991).

The IL-2 dependent cell line, HT-2, was kindly provided by Dr. Janet Yamamoto

(University of Florida, Gainesville, FL). SEB, TSST-1 and SEA were obtained from

Toxin Technology (Sarasota, FL). Anti-VB antibodies were obtained from T Cell

Sciences (Cambridge, MA), AMAC (Westbrook, ME) and Pharmingen (San Diego,

CA). Human IFNa (HulFNa) and human IFNB (HulFNf) were purchased from Lee

Biomolecular (San Diego, CA). Bovine IFNT (BolFNT) was kindly provided by Drs.

Fuller Bazer and Troy Ott (Texas A&M University, College Station, TX). Human

IFN-i (HulFN-y) was purchased from Intergen (Purchase, NY).

Radioiodinations

SEB and TSST-1 (2.5 jig) were radioiodinated with 500 iCi of Na125l (15

mCi/Ag, Amersham Corp., Arlington Heights, IL) in 25 ul of 0.5 M potassium

phosphate buffer, pH 7.4, and 10 pl of chloramine-T (5 mg/ml) for 2 min. After

neutralization of the reaction with 10 pl vol of sodium bisulfite (10 mg/ml),

potassium iodide (70 mg/ml), bovine serum albumin (BSA) (20 mg/ml) and 15 p,

of NaCI (4 M), the preparation was sieved on a 5 ml Sepharose G-10 column. The

fraction containing the highest radioactivity in the first eluted peak was used in the

radioligand binding assays. The specific activity of the iodinated proteins ranged

from 40 to 100 gCi/lg.

Peptide Competition Studies

For analysis of TSST-1 peptides, 2 x 105 cells of Raji or A20 line were








38
washed with phosphate buffered saline (PBS), and incubated with 50 p1 of

competitor, unlabeled toxin or PBS at room temperature for 1 hr. The cells were

then incubated with 121I-TSST-1 in PBS/1 % BSA for 1 hr. 75 p1 of the mixture was

placed on 0.65 p filter units (Millipore, Bedford, MA), spun, washed twice with

PBS/1 % BSA and counted on a gamma counter. For analysis of class II MHC

peptides, 50 p1 of various concentrations of peptide were preincubated with 50 p1

of 1251-TSST-1 at room temperature for 1 hr. Peptide and labeled TSST-1 were

then incubated with 2 x 105 of either Raji or A20 cells at room temperature for 1 hr.

Samples were then harvested and counted as described above.

For analysis of SEB peptides, DR1-transfected L cells were plated into a 96

well microtiter plate at a concentration of 6 X 104 cells per well and allowed to

adhere and grow to confluency over 24 hrs at 37'C. Media was removed and 50

Al of competitor were added and incubated with the cells at room temperature for

1 hr. The cells were then incubated with 1251I-SEB in PBS/1 % BSA for 1 hr. The

cells were washed twice with PBS/1 % BSA and solubilized by 1 % SDS. The

liquid was then absorbed by cotton tip applicators and assayed on a gamma

counter. For analysis of Raji cells, 2 x 10s cells of the Raji line were washed with

PBS, and incubated with 50 pl of competitor, unlabeled toxin or PBS at room

temperature for 1 hr. The cells were then incubated with 1251-SEB in PBS / 1 %

BSA for 1 hr. 75 pl of the mixture was placed on 0.65 p filter units (Millipore,

Bedford, MA), spun, washed twice with PBS / 1 % BSA and assayed on a gamma

counter.









4
Induction of EAE

Six to eight week old female PL/J mice (The Jackson Laboratory, Bar

Harbor, Maine) were immunized subcutaneously at the base of the tail with 300 lg

rat MBP in an emulsion of equal volumes of complete Freund's adjuvant (CFA)

containing H37Ra (4 mg/ml) and PBS. On the day of injection with antigen and 48

hrs later, Bordetella pertussis toxin (List Biological Laboratories, Campbell, CA)

(400 ng) was injected i.p.. Mice were examined daily for signs of EAE. Mice that

were treated with SEB to prevent EAE received 40 jg SEB in 0.2 ml PBS i.p. while

untreated mice received 0.2 ml PBS and five days later were injected with rat MBP

as described above.

Injection Schedule for Re-activation of EAE by Staphylococcal Enterotoxins

For injection of staphylococcal enterotoxin for initial re-activation of disease,

SEB (40 jig) (Sigma Chemicals, St. Louis, MO), SEA (40 /ig) (Toxin Technology,

Sarasota, FL) in 0.2 ml PBS and pertussis toxin (500 ng) both were administered

i.p. on the same day one month after resolution of clinical symptoms. For

subsequent re-activations of EAE by SEB, 40 gg of SEB in 0.2 ml PBS was

administered i.p. with or without pertussis toxin seven to nine days after resolution

of clinical symptoms.

Flow Cytometry

For analysis of superantigen treated PL/J mice, spleens were removed from

untreated and SEB treated PL/J mice, prepared as single cell suspensions and

treated with Tris-buffered 0.16 M ammonium chloride. 1 x 106 cells were washed









40
with FACS buffer (PBS containing 0.5 % bovine serum albumin and 10 mM sodium

azide) and then incubated with biotinylated anti-V88 or anti-V.B6 antibodies

(Pharmingen, San Diego, CA) for 30 min at 37'C. Cells were washed and

incubated with streptavidin-phycoerythrin for 15 min at room temperature. Cells

were then washed twice and incubated with FITC labeled anti-CD4 or anti-CD8

antibodies (Pharmingen, San Diego, CA) for 30 min at 37'C. Cells were washed

again and analyzed on a FACScan (Becton-Dickenson, Mountain View, CA) as

10,000 events per sample.

For VB analysis of type I IFN effects on superantigen activity, HPMC treated

with SEB in the presence or absence of the type I IFNs for 72 hr were washed with

FACS buffer and then incubated with FITC conjugated anti-VB antibodies for 1 hr

at 37'C. Cells were washed, resuspended in 1 ml FACS buffer and analyzed on

a FACScan (Becton-Dickinson, Mountain View, CA) in duplicate as 10,000 events

per sample.

Proliferation Assays

For SEB peptide studies, human peripheral mononuclear cells (HPMC) were

separated from heparinized blood donated by healthy volunteers using Histopaque

density centrifugation. HPMC were collected, washed and resuspended in RPMI

1640 medium supplemented with 5 % fetal bovine serum, 0.1 mM 2-

mercaptoethanol, and antibiotics (100 units of penicillin, 100 ug of streptomycin per

ml). Cells (5 x 105) were added in 96-well microtiter plates and preincubated with

the various SEB peptides in triplicate at 37'C for 3 to 5 hr. SEB was added at a








41
final concentration of 10 pg/ml and the cells were incubated for 3 days. The cells

were pulsed with 1 'iCi of [3H]thymidine for 18 hr before harvesting. Cells were

harvested on a model M12 Brandel cell harvester (Gaithersburg, MD) and

[3H]thymidine incorporation was determined in a 8-scintillation counter.

For EAE prevention studies, spleen cells from either SEB treated or

untreated EAE resolved PL/J mice were ammonium chloride treated and

stimulated in vitro with SEB or SEA in 96-multiwell tissue culture plates at 5 x 104

cells per well. Incubation and harvest were performed as stated above.

For EAE re-activation studies, spleen cells were obtained from PL/J mice

seven days after last injection or immunization and the red cells lysed with 0.84 %

ammonium chloride. Spleen cells (3 x 105/well) were incubated for 3 days in

round bottom microtiter wells that had been coated with an anti-Vf38 antibody. The

purified anti-V/38 antibody, F23.1, was diluted to 10 pg/ml with PBS and 30 il

added per microtiter well. Plates were incubated at 37'C for 2 hours and washed

with PBS before adding lymphocytes. Incubation and harvest was performed as

stated above.

For IFN and superantigen studies, HPMC (2.5 x 105/well) were added to 96-

well microtiter plates and incubated with the various superantigens and IFNs in

quadruplicate at 370C. Incubation and harvest were performed as stated above.

IL-2 Bioassay

IL-2 activity of 48 hr supernatants was determined using the IL-2 dependent

cell line, HT-2 (Ho et al., 1987). HT-2 cells (10 /well) were added in a 96-well








42

microtiter plate and incubated with supernatants and serial dilutions of recombinant

human IL-2 (rHulL-2) for the generation of a standard curve. Plates were

incubated for 72 hr and pulsed with 1 ACi of [3H]thymidine for 6 hr before

harvesting. Cells were harvested on a model M12 Brandel cell harvester

(Gaithersburg, MD) and [3H]thymidine incorporation was determined in a B-

scintillation counter.

Circular Dichroism (CD)

CD for peptides was determined at room temperature using a JASCO 500C

spectropolarimeter. Scans were done with a 0.1 mm path length cell at a

sensitivity of 2.0 and a time constant of 8 sec. The wavelength range measured

from 250 nm to 188 nm at a scan rate of 10 nm/min. Scans were carried out on

peptides in water at concentrations of 0.1-0.5 mg/ml. The CD spectra were the

average of three scans and expressed in terms of e (Yang et al 1986):

e = [9] /3298 ; [9] = [9]observed / c x I

where [9] and [9]observed are expressed in degrees, c equals the mean residue

concentration in mol/liter, and I is the path length of the cell in cm. Secondary

structure was estimated with the SSE program provided by JASCO Inc.

Statistical Analysis

Statistical analysis of experiments involving mouse treatment groups was

performed using a nested design with mice nested with in treatment groups. Mice

were considered a random effect. Statistical significance of certain data presented

in this dissertation (at alpha level 0.05) was assessed by analysis of variance









43
followed by Student's T-test and ANOVA. All statistical analysis was done under

consultation with the University of Florida Statistics Consulting Division.













RESULTS

TSST-1 Peptide Binding Studies

Overlapping peptides of the entire TSST-1 molecule were synthesized and

tested for their ability to block binding of TSST-1 to the human Burkitt's lymphoma

line, Raji and the murine B cell line, A20. Peptide sequences and secondary

structure as predicted by CD are presented in Table IV. Peptides were designed

based upon the composite surface profile and the primary sequence of TSST-1.

The predicted composite surface profile of TSST-1 is presented in Figure 4. The

CD of the TSST-1 peptides revealed a predominance of B-structure relative to a-

helix. It has been shown that TSST-1 binds to HLA-DR, HLA-DQ but with low

affinity to HLA-DP (Scholl et al., 1990). The Raji line bears DR3 and DRw10 as well

as DQwl and DQw2 and the A20 line bears I-Ad and I-Ed (Merryman et al., 1989).

Competitive binding studies between TSST-1 peptides and 12I-TSST-1 molecule

for MHC class II antigens were performed to determine the region on TSST-1

involved in interaction with the class II MHC molecule. Of the seven peptides

tested, T(39-68) and T(155-194) were able to displace binding of 'lI-TSST-1 to Raji

and A20 cells, suggesting that these regions on TSST-1 are involved in binding to

the class II molecule (Figure 5). The peptide, interferon -y(108-133), was tested as

a negative control, and did not compete with TSST-1. Dose response studies were

performed on Raji cells to determine the relative blocking ability of the peptides








45
with activity in Figure 5. 50 % inhibition of binding was observed for T(39-68) at

80 AM and for T(155-194) at 30 AM (Figure 6). In addition, a scrambled peptide

of T(39-68) (Table IV), the sequence of which was produced by the sequence edit

program (Devereux et al., 1984), was tested to determine if the binding of T(39-68)

was sequence specific. T(39-68)S did not block 12I-TSST-1 binding at 500 pM

indicating that binding of T(39-68) was sequence specific. A scrambled peptide

of T(155-194) was also unable to inhibit binding of 1251-TSST-1 (data not shown).

Similar dose dependent patterns of inhibition for the peptides with inhibitive activity

and their nonactive scrambled counterparts was observed on murine A20 cells

(data not shown). Thus, the N-terminal peptide T(39-68) and the C-terminal

peptide T(155-194) represent regions of TSST-1 that are involved in binding to both

HLA and MHC class II molecules on Raji and A20 cells.

In order to determine the region of the class II molecule that binds TSST-1,

class II MHC, I-Ab, a-chain peptides I-Ab(30-60), I-Asb(60-90), I-Asb(65-85), and

IAGb(80-100) were examined for their ability to inhibit binding of TSST-1 to Raji and

A20 cells (Russell et al., 1990). Peptides I-Ab(30-60) and l-As(60-90)

preferentially inhibited binding of 1lI-TSST-1 in a dose dependent manner to both

Raji cells (Figure 7) and A20 cells (data not shown) The peptide I-A1b(60-90)

encompasses the entire 8-chain a-helix while I-Ab(30-60) is a beta-turn underlying

the N-terminal region of the (60-90) a-helix (Bjorkman et al., 1987; Brown et al.,

1988). Thus, TSST-1 interacts with residues of the a-helix and 8-turn of the class

II MHC molecule B-chain.















0) T- 0c) 0 C\j Cj I'-
C\j Cj C\M I- CMj (D CMJ


.Q















o



4-'
aO
















a)
-C










C.









e


-4--
-- _S- S


6~c d) c) t) 0) LO)
- a v cla ,M ,-- T- a i-
a to a r-c co uS a a

pr'~fz ^ jr r -jr r r


cO 0) N- CO) Nr- i-


d

0






co
00
a -
CLi
E
-o
XD X

a. _
up
a)
0




C G)

a)
E C
So>






a) U


io
E c
:p 0







S0)








CO
a) 0
Sa)
U -?







Ec
0
n


0 U)
o 5

SIi






























Figure 4. Predicted composite surface profile for TSST-1 using the three
parameters of HPLC hydrophilicity, accessibility, and segmental mobility (Parker
et al., 1986). Residues with composite profile values greater than 50 % are
predicted to be more likely to be accessible to the surface of the molecule.













S
U
R
F
A
C
E 5 -
U
A
L
U
E

9


I 190
due NupMer


49 60












mg-a



C
o0 0

o

(D

o c


-C0L

0 0



0 (D
S0 E
C"







oa


VC
0 (




.a U )
62


















C C
a)a






0- 0e







c c







u -: 3 C


































HI


0 a
o


50


4Q







C,,
0
z



a





I-





A5

D.


*
C,
IC
co





4-









0%.
CJ





1OAUO %


























Figure 6. Dose dependent inhibition of '1I-TSST-1 binding to Raji cells in the
presence of T(39-68), T(155-194) and T(39-68)S. T(39-68)S is a scrambled peptide
whose sequence was generated via the sequence edit program (Deveraux et al.,
1984)). Labeled TSST-1 was used at a final concentration of 2.5 nM. In the
absence of competitor, binding of 1Il-TSST-1 to Raji cells was 6407 21 cpm.
Each point represents the mean percent reduction SE of TSST-1 control binding
in the presence of TSST-1 peptides.


























100




80


1 10 100 1000


Concentration (pM)


10000





























Figure 7. Dose dependent inhibition of TSST-1 binding to Raji cells in the presence
of a-chain class II MHC peptides. 12I-TSST-1 was used at a final concentration
of 2.5 nM. Each point represents the mean percent reduction SE of TSST-1
control binding in the presence of class II MHC peptides.



















































100 1000

Concentration (pM)


100






S8o

0



S40






40


10000








4
SEB Peptide Binding Studies

Overlapping peptides of the entire SEB molecule were synthesized and

examined for their ability to inhibit binding of 12S-SEB to DR1 transfected L cells.

The peptides were designed and synthesized to correspond to discrete secondary

structures of SEB and the sequences and corresponding regions they encompass

in the crystal structure of SEB (Swaminathan et al., 1992) are presented in Table

V. The predicted composite surface profile of SEB was also considered in the

designation of peptide sequences (Figure 8). Competitive binding studies between

SEB peptides and 12I-SEB for MHC class II antigens were performed to determine

the regions) of SEB involved in interaction with the MHC class II molecule. Of the

overlapping peptides tested, peptides corresponding to amino acid residues 1-33,

31-64 and 179-212 were able to displace binding of 1Is-SEB to DR1 transfected L

cells (Figure 9). No appreciable binding of 1lI-SEB to L cells transfected with the

DR a-chain, but which did not express surface DR1, was observed (data not

shown). Dose response studies were performed on DR1 transfected L cells to

determine the relative blocking abilities of the peptides possessing inhibitory activity

in Figure 9. Each of the three peptides, (1-33), (31-64), and (179-212), showed

inhibition of 1Il-SEB binding at concentrations as low as 20 pM, and caused 50

% inhibition of binding in the 100 /M range (Figure 10). In addition, IFNT- (108-

133), an irrelevant peptide, was employed as a negative control and was shown

not to inhibit 1251-SEB binding. Thus, the amino terminal regions 1-33 and 31-64

and the carboxy terminal region 179-212 of SEB are involved in binding of SEB to









4
the HLA-DR1 class II molecule.

We next determined the ability of the SEB peptides to inhibit binding of 125-

SEB to the human Burkitt's lymphoma line, Raji. As stated previously, the Raji line

bears HLA-DR3, DRw10, DQwl and DQw2 (Merryman et al., 1989). The purpose

of examining the activity of the SEB peptides against a second cell line bearing a

different haplotype was to determine if the same or different regions of SEB were

involved in binding to MHC class II molecules of different haplotypes. Of the eight

SEB peptides tested, peptides corresponding to amino acid residues 1-33, 124-

154, 150-183 and 179-212 were able to displace binding of 12I-SEB to Raji cells

(Figure 11). Dose response studies showed inhibition of binding at concentrations

as low as 5 to 10 jiM (Figure 12). 50 % inhibition of binding was observed for all

four peptides in the range of 20 to 40 pM (Figure 12). IFN-y (108-133), an

unrelated peptide, was also examined and did not compete with 1'I-SEB for

binding. Thus, the regions of SEB encompassing residues 1-33, 124-154, 150-183

and 179-212 represent sites that are involved in binding to the class II molecules,

DR3, DRw10, DQwl and DQw2 on Raji cells. Further, while some SEB peptides

inhibited SEB binding to both L cells transfected with DR 1 and Raji cells

expressing different DR haplotypes, binding of SEB to the two cell lines was also

differentially inhibited by distinct peptides. This suggests that different regions of

SEB bind to different MHC class II molecule haplotypes.

Peptide (179-212) was selected to determine if peptide truncations could

further define discrete regions or residues of peptides that are involved in binding









0 57
C-





c C C r 4
8 a o a a 0a
ci


(0



S-0 0

-1o )
cL z > C y0
a Z z z


CO y z 4 a


QV S 0i o -J < w a
z0 z E U
a >> C

SZ 0
~ O2 2 o
L. > o



E i g >
Sw ) w H Z w *)
0) 0 z 0 0 a


o M > > s
a) Q. C/ m 0 a) 2 a)OC
ca > o a -J 0 )- a

mj O
0) 0 (0


LE 0 U3
0 a a
(o y > > ) a o1.e
QC C') '2 r If) CD C- (0


C a CO > 0
H1)l C C' Cw i ) Cl C ( c






























Figure 8. Predicted composite surface profile for SEB using the three parameters
of HPLC hydrophilicity, accessibility, and segmental mobility (Parker et al., 1986).
Residues with composite profile values greater than 50 % are predicted to be more
likely to be accessible to the surface of the molecule.







59

CoM osite Graph file: SEB





1i




B 1


30 60 99 120 150 189 219
Residue Nu ier













me

We
*.'_ .
(D 0
S0

)0 C


sw-






owo
(14
C U

tc O


its








Bo .5







000
CO
b S



0- U

S0..
Oc0




i. e E










61
















0
0
C4








0
Cl










ox
C4











cm **
.. CL
C4






0


0
0










O1O
I
1 1 1 1 11 1 0
secoo1-
N sj00a)
Yo- o4




























Figure 10. Dose dependent inhibition of 12I-SEB binding to DR1 cells in the
presence of SEB peptides possessing inhibitory activity. I-SEB was used at a
final concentration of 4 nM. Binding of '2l-SEB in the absence of competitor was
2435 187 cpm. Closed square (1-33), closed circle (31-64), closed triangle (179-
212), and open triangle IFN-y (108-133). Each point represents the mean percent
control SE.































----- SEB (1-33)
----- SEB (31-64)
A SEB (179-212)
- IFN Y (108-133)


I ,,, ... / ., ,


.1 1 1f


Concentration (pM)


120


100-

80-


1000


10000


0










" c
0D (0










CU
0


c r



-a
a-c







o0

Sc
0
-Q.









00


3 c
St











00


'4--
o e

C'-
0- 0
8 -0.









ma
o5







o .c



0:5









c ac
U. e








65
4










CJ

C1
8T-





CS,
0)



co
i-.


6
Cu









a a a '- a.
VS
cm
CL
co




(0










o CI C C 0 0 0
lCuo

10AU03 %




























Figure 12. Dose dependent inhibition of 1251-SEB binding to Raji cells in the
presence of SEB peptides possessing inhibitory activity. 1 I-SEB was used at a
final concentration of 4 nM. Binding of '2l-SEB in the absence of competitor was
4768 355. Closed square (1-33), closed circle (124-154), open square (150-183),
closed triangle (179-212), and open triangle IFNy (108-133). Each point represents
the mean percent control SE.

























L H


-- SEB (1-33)
-- SEB (124-154)
--- SEB (150-183)
-a- SEB (179-212)
--- IFN Y (108-133)


........ ...... .......-


1 10 100 1000


Concentration (pM)


120


100-


80-


601


10000





























Figure 13. Percent control binding of 12I-SEB to DR 1 transfected L cells in the
presence of N-terminal truncations of the SEB peptide (179-212). SEB peptide
truncations were assayed at a final concentration of 200 pM. 2lI-SEB was used
at a final concentration of 5 nM. Each bar represents the mean percent of SEB
control binding in the presence of SEB (179-212) truncations SE.





















































(186-212) (195-212)

SEB (179-212) Truncations


120-



100-



80-



60-



40-



20-



0-


(179-212)


(202-212)








70
of SEB to HLA-DR1 molecules. Accordingly, we synthesized three amino-terminal

truncations of SEB (179-212) encompassing regions 186-212, 195-212 and 202-212

as shown in Table VI. Competitive binding studies between the SEB (179-212)

truncations and 1lI-SEB for DR 1 present on transfected L cells were performed.

A gradual decline of competition was observed corresponding to a decrease in the

size of the peptide (Figure 13). Similar results were observed with competitive

binding studies conducted with Raji cells (data not shown). Such results suggest

that the binding of the SEB peptide encompassing residues 179 thru 212 may not

be localized to a specific region or residue or the truncations may have resulted

in a corresponding reduction in the overall affinity and/or avidity of the peptide for

binding to the MHC class II molecule.

Antagonist Activity of SEB(124-154)

Recent studies have suggested that peptides of the bacterial superantigen,

SEA, have actual superantigenic properties themselves. The SEA peptide (121-

149) is able to stimulate proliferation and cytokine induction (Pontzer et al., 1993),

while SEA peptide (39-68) is able to stimulate proliferation (N. Griggs, personal

communication). We tested the eight overlapping SEB peptides for their ability to

induce T cell proliferation and found none of them capable of T cell stimulation

(Figure 14). Another member of Dr. Johnson's laboratory, Dr. Carol H. Pontzer,

tested the peptides for induction of TNF-a and B, however the studies were

inconclusive.

We next attempted to determine if any of the SEB peptides were able to











x 7E 0
xlo

- 0

0
cE



._ 8 -

4-e
U) c








xO=
co
(D 2 e


CA
) cE

0o "


0c.0
_o "o





CD









m O
Sa2. a-


C 0



Mo r5 c o
0 UN 0









4














co





04
CL
1 II












Tcn
w
Q)
di-d


V.












l)





xopu 1 o
V.
*
cl.
I I I
O O I)
xepu uo~'elnu!:I








4 73
inhibit the mitogenic function of SEB, the eight peptides were tested for direct

functional competition with SEB for inhibition of SEB-induced proliferation of HPMC

(Figure 15). It was observed that at a concentration of 300 pM the peptide

corresponding to amino acid residues 124-154 was able to significantly (p < 0.02)

inhibit SEB-induced proliferation. The addition of an irrelevant peptide, IFN-y(108-

133), had no effect on SEB-induced proliferation. These data suggest that

residues 124 through 154 are part of a functional site directly responsible for

activation of some lymphocytes by SEB. The other SEB receptor-binding sites

may facilitate the direct activation of cells by residues encompassed in the 124-154

peptide.

Prevention of EAE by SEB

SEB has been shown to induce clonal energy and deletion in peripheral

V88+ CD4+ T cells after in vivo administration (Kawabe and Ochi, 1990; Rellahan

et al., 1990; Kawabe and Ochi, 1991). Interestingly, the culprit T cells in the PL/J

mouse strain responsible for induction of EAE are V88 specific (Acha-Orbea et al.,

1988). In order to determine if mice treated with SEB could be protected from

development of EAE, PL/J mice were injected i.p. with SEB (40 pg in 0.2 ml of

PBS) or 0.2 ml PBS alone and allowed to rest for five days. SEB injected mice did

not exhibit signs of toxicity at any time after administration.. After a five day period,

treated mice and control PBS injected mice were immunized with rat MBP as

described in Materials and Methods and observed for signs of disease. In Table

VI, Experiment 1, five out of six mice in the group that received rat MBP, but no









2 C-4





c --- o-
.o I-

S0"














E2
co








~C0 "
L. 0a,-







ICl









c,.c
3c ci




- CD 4)






m cno 3
*F" c: 0



0 c x































1-I






















.cq
C\
























I
0\








0







m
-0



.0


00 0 0 0 0 0 0 0
Go p- \0 il IT fn


UorUiZp2a %








76
SEB showed signs of EAE at an mean day of onset of 16.2 (t 2.5) days (Table VI)

with a mean severity of 1.8 ( 0.88) (mild paraplegia). By contrast, four out of five

mice treated with SEB prior to immunization with rat MBP showed no signs of EAE

after 35 days. Only one mouse exhibited signs of EAE at an extended time point

of 24 days with a severity of onset of 1 (loss of tail tone). Likewise, five out of six

mice in experiment 2 that received rat MBP, but no SEB showed signs of EAE at

a mean day of onset of 15.4 ( 2.1) with a mean severity of 1.4 ( 0.49). Three

of five mice injected with SEB did not develop signs of EAE while two mice

exhibited onset of EAE at days 15 and 19 with a severity of 1 and 5 respectively.

The specific VB populations of mice that had been protected by SEB

treatment were studied next. It has been shown that V88.2+ CD4+ T cells are

associated with induction of EAE in PL/J mice (Acha-Orbea et al., 1988). Study

of spleen cells of SEB treated PL/J mice (no signs of EAE 35 days post

immunization) by two-color FACS analysis revealed a reduction of V88+ CD4 T

cells (Figure 16). In comparison, analysis of spleen cells from untreated EAE

resolved PL/J mice showed the presence of VB8+ CD4* T cells. A total of four

mice/group (SEB protected and EAE resolved) were studied. Inspection of VB6*

T cells showed no significant difference between SEB treated and untreated EAE

resolved mice (data not shown). The VB profiles of SEB treated and untreated

EAE resolved mice correspond to the pattern expected for EAE and the known VB

specificity of SEB.

Presented in Figure 17 is the comparison of the ability of spleen cells from










Table VI. Treatment with SEB prevents development of EAE in PL/J mice*



Mean Mean day of
SEB Incidence severity onset


Expt. 1


+ 1/5 1 24

5/6 1.8 (0.88) 16.2 (2.4)

Expt. 2


+ 2/5 1,5 15,19

5/6 1.4 (0.49) 15.4 (2.1)



* PL/J mice were injected with 300 Ag of rat MBP in a 1:1 emulsion of complete
Freund's adjuvant, H37Ra (4 mg/ml). Mice received Bordetella pertussis toxin (400
ng) on day of injection and 48 hrs later. SEB treated mice received 40 Ag SEB in
0.2 ml PBS i.p. 5 days prior to immunization with rat MBP. Untreated mice received
0.2 ml PBS i.p. 5 days prior to immunization with rat MBP. Mean and standard
deviation are given only for groups of mice that contained statistically significant
populations. Severity of EAE is represented as a graded scale in which: 0, no
signs of EAE; 1, loss of tail tone only; 2, mild paraparesis; 3, severe paraparesis;
4, paraplegia; 5, moribund. Chi-square test of the significance of the difference in
incidence between SEB injected and control PL/J mice from the two experiments;
p < 0.012.
























Figure 16. Two-color FACS analysis of PL/J spleen cells showing V88" CD4 T
cell depletion in SEB protected PL/J mice. Percentage of VB88 CD4+ T cells
standard error for SEB protected PL/J mice, 2.36 1.32 (A) and EAE resolved
PL/J mice, 5.72 0.30 (B). Values given are the mean of three individual
experiments. Chi-square test of the significance of reduction between EAE
resolved and SEB protected mice; p < 0.05. Either SEB treated (A) or untreated
EAE resolved (B) were killed and the spleens were removed 40 days after
treatment with SEB. Ammonium chloride treated single cell suspensions were
analyzed with biotinylated anti-V88 and anti-VB6 antibodies and FITC labeled anti-
CD4 and -CD8 antibodies. Results for staining with anti-VB8 antibodies are
presented as individual contour graphs. Staining with anti-VB6 antibodies exhibited
no significant difference between SEB treated and untreated EAE resolved mice
(data not shown).












79
4




















A -* S



c o




t ., o s'-'
I -





is' top gas is' -00 rLt
















r%.
VB '.



























Figure 17. SEB treated PL/J mice that did not develop EAE are unresponsive to
SEB but respond to SEA in vitro (A). Untreated EAE resolved PL/J mice respond
to both SEB and SEA in vitro in a dose-dependent manner (B); (closed bars) SEB;
(hatched bars) SEA. Proliferation assays were performed 40 days after SEB
treatment or PBS treatment alone. Data are indicated as arithmetic means of
quadruplicate samples. 5 x 104 cells/well were used.





























A

* SEB
0 SEA


10 1 0.1 0.01
Concentra ion (pg/ml)


1 0.1
ConcenLration (pag/ml)


70000 -

60000 -

50000 -

40000 -

30000 -

20000 -

10000 -

0



80000

70000 -

60000 -

50000 -

40000 -

30000 -

20000 -

10000

0 -


B

* SEB
Q SEA












0.01








82
4
SEB treated mice that did not develop EAE and untreated mice that had resolved

all signs of EAE to proliferate in response to SEB and SEA. Spleen cells were

stimulated in vitro with 10.0, 1.0, 0.1, 0.01 Ag/ml of either SEB or SEA. Cells from

mice treated with SEB that did not develop EAE were unresponsive to in vitro

stimulation by SEB while stimulation by SEA produced a vigorous response. Such

a response to SEA is understandable in that SEA has a VB specificity different from

SEB, thus it would not be affected by previous exposure of the animal to SEB. In

contrast, cells from EAE mice that were not treated with SEB were able to respond

to both SEB and SEA in a dose-dependent manner. The greater response to SEA

is consistent with the previously reported greater reactivity of this superantigen

(Smith and Johnson, 1975). Therefore it would appear that the Vs specificity of

SEB plays an important role in its ability to protect PL/J mice from induction of

EAE.

Reactivation of EAE by Staphylococcal Enterotoxins

Six PL/J mice were initially immunized with rat MBP in complete Freunds

adjuvant (CFA) followed by pertussis toxin. As can be seen in Figure 18, 5/6 mice

developed signs of EAE with a mean severity index of 1.8, which resolved within

a week. One month after the development of EAE, when the mice were clinically

normal and no other relapse of EAE was noted, four mice were injected with SEB

and pertussis toxin only (rat MBP was not repeated). Three weeks following this

injection, 2/4 mice developed a second episode of clinical EAE. This episode

resolved and the mice were again clinically normal. Two weeks following resolution








83
4
of all clinical signs the same 4 mice were again injected with SEB and pertussis,

with a relapse of the clinical signs of EAE seen in 3/4 mice (onset 1-2 weeks post

injection) (Figure 18 and Table VII, Experiment 1). A similar result was seen in an

additional experiment (Table VII, Experiment 2). No relapses occurred when mice

were injected with pertussis only. Another group of 15 PL/J mice were

administered SEB (40 pg i.p.) (mice had not been previously immunized with rat

MBP), and none of them developed evidence of encephalitis (data not shown).

Next, we wished to determine whether SEB could re-induce EAE in the

absence of pertussis toxin. In a total of four mice that had previously been

immunized with MBP and received SEB plus pertussis toxin, two developed clinical

signs of EAE (Table VIII, Experiment 1) with SEB only. However, the severity and

duration of clinical signs was less than previous episodes in mice re-induced with

SEB only (compared to SEB and pertussis). Nevertheless, SEB alone was capable

of re-inducing EAE in some mice that had previously received SEB and pertussis

toxin.

Finally, we evaluated whether EAE could be induced in those mice that were

immunized with rat MBP in CFA and pertussis toxin, but who never developed

clinical evidence of EAE. A total of 3/7 mice developed EAE after injection with

SEB and pertussis toxin (Table VII, Experiment 3), and of those 3, one animal

developed EAE after re-injection with SEB only. Thus, SEB is able to induce EAE

in mice immunized with rat MBP but who did not develop clinical signs of EAE.

The finding that mice injected repeatedly with SEB developed clinical EAE








84
following each injection was unexpected. Although SEB can activate V#8+ cells,

it has been demonstrated that following this period of activation the V#8+ T cells

become unresponsive to further stimulation with SEB (or to stimulation with anti

V#8* antibodies) (Kawabe and Ochi, 1990; Rellahan et al., 1990; Kawabe and

Ochi, 1991). We reasoned that either V8'+ T cells that were previously activated

must be resistant to the development of energy upon stimulation with SEB, or that

the pertussis toxin injected along with the SEB prevented the induction of energy

by SEB. Figure 19, Panels A and B demonstrate that T cells activated in vivo with

rat MBP in CFA are indeed resistant to the induction of energy. In Panel A, mice

were administered either SEB one week before immunization with rat MBP in CFA

or administered SEB one week after immunization with rat MBP in CFA. V#8+

proliferation was evaluated by stimulation of T cells for three days in microtiter wells

coated with an anti-V#8 antibody followed by 3H-thymidine incorporation. V,88 T

cells obtained from animals administered SEB one week before immunization with

rat MBP in CFA exhibited a reduced response, while VP8+ T cells from mice

administered SEB one week after immunization with rat MBP in CFA were not

anergized. Interestingly, mice that were immunized with only CFA and received

SEB one week later were also not anergized (data not shown). Controls consisted

of PL/J mice immunized with only rat MBP in CFA. T cells from all three groups

proliferated equally well when stimulated with an anti-VP9+ antibody in vitro (data

not shown), suggesting that V88 T cells had been specifically anergized. In

addition, in Panel B we evaluated V#8+ T cell proliferation in one group of mice








85
that had previously developed a clinical episode of EAE and who were

administered SEB one month after the episode of EAE. V#8* T cells from these

mice also failed to be anergized after in vivo exposure to SEB. These results show

that the timing of administration of SEB determines whether cells are anergized or

activated. Thus, it appears that VP8* T cells stimulated in vivo by rat MBP and

CFA or CFA alone are resistant to the induction of energy by SEB.

Figure 19 demonstrates that pertussis toxin cannot overcome the anergizing

effects of SEB when injected simultaneously with SEB in previously unimmunized

animals although pertussis toxin itself is not a superantigen (Kamradt et al., 1991).

It had previously been demonstrated that pertussis toxin could prevent the

induction of T cell energy to an encephalitogenic peptide of MBP injected I.V. and

that pertussis toxin has strong mitogenic effects (Kamradt et al., 1991). However,

pertussis toxin, when injected simultaneously with SEB, was not able to overcome

SEB induced T cell energy.

While SEB can activate VP8* T cells, SEA cannot. In order to determine

whether reactivation of EAE was specific for SEB, we immunized a separate group

of PL/J mice with rat MBP and pertussis toxin. Six out of 10 mice developed EAE.

Following remission of clinical EAE, mice were then injected with SEA and pertussis

toxin. The 6 EAE mice developed a clinical relapse of EAE (Table VIII). The 4

mice that were immunized with MBP but that did not develop clinical symptoms

were also injected with SEA (40/g i.p.) and pertussis toxin one month after

resolution of clinical symptoms. A total of 2/4 of these mice exhibited clinical





























Figure 18. Time course of reinduction of EAE in PL/J mice initially immunized with
rat MBP in CFA and pertussis and given multiple injections of SEB. A relapse of
EAE occurred each time following immunization with either SEB and pertussis or
SEB alone.











































1 5/6](1.8)


Pertussis EAE
MBP/CFA begins

Pertussis I


0 2 14


12/41(3.0)


SEB -
Pert.
Clinical EAE
Resolution i begins
I // I I


22 1 55 75


(3/41(2.6)


SES *
Dert.
I EAE
Resolves begins
I I I


82 91 98


11/31(1.0)


SEB
Resolving EAE
begins
I I 1I


124 130 137 143


Days


Resolving
I





















c
0
s >,
-2.E
. 0




oEn


.c')
(0
C U)



.c C
0

0 a




Un
U)

c't
0)
C
0

L W




.0


p :* 8 w
E 4 c

- C -0 8.





:3 > '0 0






con EC
S) E E
-0. to a.

"0 oa C o o0
105 )0 6 .E
.- E
c oC


So 5 W I
C QC EW



70 c .- c



C 0 >
E O -



Zn Z: ow



a d 0 > o-u
Na w. w .
E >. 0

, 0 0 OW

Si c m .|
a m -a .9-

I.5 5 L-
00 >.
.a) 0 > 0,


a^.o I-






>ci( .u <


r-





















Figure 19. Induction of T cell energy by superantigen is prevented by previous
activation. (A) Bars: 1, group was immunized with MBP only; 2, group received
SEB (40 Ag i.p.) and was immunized 1 week later with MBP; 3, group was
immunized with MBP and received SEB (40pg i.p.) 1 week later. All groups were
sacrificed 1 week after the last immunization or injection. Procedures were timed
for sacrifice of all groups on the same day. Proliferation was induced with anti VB8
antibodies and measured by [3H]thymidine incorporation (Rellahan et al., 1990).
Group means differed significantly by an ANOVA (P>0.0031). (B) Bars: 1, group
received SEB (40 pg i.p.) only; 2, group received both toxins (SEB at 40 pg and
pertussis toxin at 500 ng i.p.) simultaneously; 3, group was immunized with MBP
and i week later received SEB (40 pg i.p.); 4, group was immunized for induction
of EAE and developed acute symptoms, and symptoms were resolved. One week
after resolution mice received SEB (i.p.). All groups were sacrificed 1 week after
the last immunization or injection. Again, procedures were timed for sacrifice of
all groups on the same day. Two to three mice were used per group per
experiment and proliferation was induced with anti-V88 antibodies. Significance
between groups 1 and 2 versus groups 3 and 4 was determined by an ANOVA
(P< 0.0001). Controls were performed for all experiments by measuring
proliferation induced by anti-VB9 antibodies or in the absence of antibody. Cells
from all groups proliferated equally well when stimulated with anti-V89 antibodies.





























A
120,000


80,000
E

40,000



0

B


120.000


| 80.000


40,000


1


T


2


2