Equine retinal S antigen and its relevance to equine recurrent uveitis

MISSING IMAGE

Material Information

Title:
Equine retinal S antigen and its relevance to equine recurrent uveitis
Physical Description:
Book
Language:
English
Creator:
Hines, Melissa Trogdon
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 23098608
oclc - 20674758
System ID:
AA00022808:00001

Full Text










EQUINE RETINAL S ANTIGEN AND ITS RELEVANCE
TO EQUINE RECURRENT UVEITIS









BY

MELISSA TROGDON HINES, DVM


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


1988














ACKNOWLEDGEMENTS


I would like to thank the many people who provided

assistance and support throughout my graduate training.

First, I would like to thank Dr. Richard Halliwell, who

provided not only guidance but much-needed encouragement.

Also, I would like to extend thanks to my committee members

for their valuable suggestions.

I am very grateful to everyone who helped me in the

laboratory and working with the ponies during this long

project. Special thanks go to Alyssa Jarpe, Nancy Keller,

Luis Romero, Danielle Stanek and especially Debby

Sundstrom; their assistance and friendship have been

invaluable. I would also like to thank my fellow graduate

students, especially Steve Oberste and Paul Kroger who

helped me get through the early years.

Finally, I would like to express my gratitude to Steve

Hines for his love, support and histologic descriptions.

Also I would like to recognize my entire family; they have

always been there for me.















TABLE OF CONTENTS


Page


ACKNOWLEDGEMENTS


ABSTRACT . . . . . . . . . .

CHAPTERS

I INTRODUCTION .............

II ISOLATION AND CHARACTERIZATION OF EQUINE
RETINAL S ANTIGEN

Introduction ...... ...............
Materials and Methods ... ..........
Results ....... .................
Discussion . . . . . . . .

III IMMUNOLOGIC REACTIVITY TO S ANTIGEN IN
NORMAL HORSES AND THOSE WITH UVEITIS


. . iv


. 1


Introduction ..... ............
Materials and Methods .. ........
Results ...... .............
Discussion . . . . . .

IV IMMUNIZATION OF PONIES WITH EQUINE
RETINAL S ANTIGEN

Introduction ..... ............
Materials and Methods .. ........
Results ...... ..............
Discussion . . . . . .

V SUMMARY ..... ..............

REFERENCES . . . . . . . .

BIOGRAPHICAL SKETCH ............


82
* 83
88
117


. . . 123

. . . 128

142


iii














Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
EQUINE RETINAL S ANTIGEN AND ITS RELEVANCE
TO EQUINE RECURRENT UVEITIS

BY

MELISSA TROGDON HINES, DVM

DECEMBER, 1988

Chairman: Dr. R. E. W. Halliwell
Major Department: Immunology and Medical Microbiology

S antigen, a soluble retinal antigen implicated in

experimental autoimmune uveitis and clinical uveitis of

man, was isolated from equine retinas and studies were

conducted investigating its role in the immunopathogenesis

of equine recurrent uveitis. The primary method of

purification was gel filtration followed by ion exchange

chromatography. Analysis of the equine S antigen

preparation by polyacrylamide gel electrophoresis revealed

a major band of 53 kDa corresponding closely in size to the

major band of bovine S antigen prepara- tions. There were

also several minor bands present, all of which were

recognized by both polyvalent antisera to equine, bovine

and guinea pig S antigen and by monoclonal antibodies

raised against the 53 KDa band, indicating that the bands

are antigenically related.








Studies investigating immunologic reactivity to S

antigen in the equine population demonstrated that serum

IgG antibody to S antigen was prevalent in the general

population, and that there was no correlation between the

incidence or the level of such antibody and uveitis.

However, intraocular antibody

to S antigen was found only in the animals with uveitis,

and in several cases intraocular synthesis of antibody was

docu- mented. Peripheral blood lymphocytes from both

normal animals and patients with uveitis would occasionally

respond to S antigen, but there was no evidence for

heightened cellular reactivity in those animals with

disease.

Immunization with equine S antigen in complete Freund's

adjuvant did not produce experimental autoimmune uveitis in

ponies. The immune response to the antigen was variable,

with most individuals mounting a humor response but no

cellular response.

Administration of S antigen in the anterior chamber did

consistently produce an acute uveitis which was generally

not observed after the intraocular administration of

albumin. The phenomenon appears to be a direct

inflammatory response rather than an immune mediated

response, although it suggests that release of S antigen

into the eye following ocular damage of any kind could

potentially contribute to inflammation.













CHAPTER 1
INTRODUCTION



Uveitis is characterized by inflammation of the uveal

tract or vascular tunic of the eye, which is composed of the

iris, ciliary body and choroid. All of these structures,

either singly or in combination, may be affected, and

frequently the inflammation extends to other regions of the

eye as well. Chronic or recurrent episodes often result in

irreversible ocular pathology, such as synechiae, cataracts,

retinal detachment or glaucoma. There are several different

syndromes of uveitis involving a number of etiologies, and

although many types of uveitis are believed to be immune-

mediated, the pathogenic mechanisms remain obscure. This

limited understanding of the pathophysiology of the disease

has made effective prevention and therapy difficult.

Uveitis occurs as a spontaneous disease in a number of

species, but it is of particular significance as a distinct

clinical entity in man and in the horse. In man, the

disease has an annual incidence of 15 new cases per

i00,0001. In the United States alone, approximately 45,000

new cases of uveitis and 1,650 new cases of blindness due to

uveitis were diagnosed in 1978.2 Many distinct subsets of

uveitis have been defined; among the most common is acute










anterior uveitis, which annually accounts for 8.2 new cases

per 100,0003. This syndrome primarily involves the iris and

ciliary body resulting in an iridocyclitis that generally

lasts less than three months, although it has a marked

tendency to recur (40% within 5 years). In most cases a

specific cause for the inflammation cannot be estab-

lished.4,5,6

Uveitis in man comprises a group of inflammatory

conditions of the eye, and frequently the etiology is

obscure. In the 1800's, most cases of uveitis in people

were attributed to either tuberculosis or syphilis, but

research since that time has linked the disease to a wide

variety of causative factors.7 Infectious agents are still

among the recognized causes but a specific bacterium, virus

or fungus can be identified in only 5-10% of patients.

Important parasitic agents include Toxoplasma, Onchocerca,

and Toxocara, and yet these too only account for a small

percentage of cases8,9,10,11. Uveitis has also been

associated with a number of syndromes of presumed immuno-

logic origin, including such conditions as anklyosing

spondylitis, Reiter's syndrome, ulcerative colitis and

Behcet's disease.12,13,14 Genetic factors also appear to be

involved in certain types of uveitis, the best example being

the association of acute anterior uveitis with certain HLA

antigens, especially HLA-B27. HLA-B27-associated uveitis

appears to form a distinct clinical entity accounting for










approximately 50% of acute anterior uveitis cases.5,14,15

Despite the many known etiologic agents and predisposing

factors, in many cases of ocular inflammation a specific

etiology cannot be established. In a recent retrospective

study of 600 cases, 33% were finally designated as idio-

pathic; this percentage may be deceptively low since those

cases with well-established disease patterns, such as the

condition recognized as birdshot choroidopathy, were not

considered idiopathic even though their pathophysiologic

cause remains unknown.16

Equine uveitis is of importance not only because of its

prevalence and severity, but because it shares many features

with uveitis of man, potentially allowing it to serve as a

model for human disease. It is the leading cause of

blindness in horses and mules throughout the world, with

estimates of its overall incidence reaching 12%.17,18

Although as in any species uveitis in the horse can result

from trauma or direct invasion of the eye by infectious

agents, equine recurrent uveitis is typically an isolated

ocular problem of uncertain etiology.19,20 As is the case

in acute anterior uveitis of man, the disease primarily

produces an iridocyclitis, although extension to other

ocular tissues is common with chorioretinitis and optic

neuritis often being identified. Also known as periodic

ophthalmia, the disease is characterized by recurring

episodes of inflammation often resulting in permanent damage










to the eye with time. The condition may be unilateral or

bilateral, and the clinical course is highly variable.

Most studies indicate that there is no breed, sex or age

predilection.17,20 However, there have been anecdotal

reports suggesting a genetic predisposition and one recent

study found a significantly increased incidence of the

disease in the Appaloosa breed.21

Recognition of equine recurrent uveitis dates back to

the 4th century A.D. when it was described by Vegetius, at

which time its periodic recrudescence was ascribed to the

lunar cycle and thus the familiar term "moon blindness".

Since then a number of investigations have attempted to

define the etiology and pathogenesis. The pathogen most

often reported in association with the disease is Leptospira

interrogans. Leptospirosis is associated with uveitis in

man as well as the horse, with ocular inflammation often

occurring weeks to months after the initial systemic

infection in both species.22,23,24,25 Williams experimen-

tally produced uveitis in ponies by subcutaneous inoculation

with Leptospira interrogans serovar pomona; although no

ocular damage was detected prior to 50 weeks after exposure,

18 of 36 eyes eventually developed inflammation. The

mechanism of ocular damage in leptospirosis is not under-

stood. In the horse, there appears to be no correlation

between the presence of serum antibody and disease, but

there is evidence suggesting an association with the








5

intraocular synthesis of leptospiral antibody.26 ,27 This

observation, together with the fact that uveitis generally

develops months after systemic infection when leptospires

cannot be recovered from the eye, suggests the possibility

of persistence of antigen in the eye and immune mediated

ocular damage. Still, in most cases of equine uveitis there

are no data to support a diagnosis of leptospirosis and

other possible etiologies must be investigated. Another

agent implicated is Onchocerca cervicalis. Again, oncho-

cerciasis is a well established cause of uveitis in man, and

there are several reports of ocular onchocerciasis in the

horse.28,29,30 However, the high prevalence of this

parasite in the equine population made it difficult to

establish a causal relationship in many cases.31,32 A host

of other causative factors, including viral infections,

streptococcal hypersensitivity and riboflavin deficiency

have been proposed.17,19,20,33 In most instances the

etiology of equine recurrent uveitis is undetermined making

specific prevention and therapy difficult.

Thus uveitis is a complex group of diseases with no

single etiology. However, there is evidence to suggest that

immunological mechanisms play a crucial role in the patho-

genesis whether or not a specific etiologic agent is

identified. Some characteristics which support the concept

of an immunologic basis for uveitis include its responsive-

ness to immunomodulatory agents such as corticosteroids and










cylclosporin, the lymphocytic nature of the infiltrate, and

its occurrence as a component of systemic immune disorders.

Immunologic damage to the eye could result from a primary

immunologic abnormality or secondary to damage by infectious

agents or trauma.

The eye possesses a number of distinct anatomical and

physiological features which make it immunologically unique.

Although the eye is classically considered to be an immuno-

logically privileged site, it has been shown to support a

variety of immunologic reactions. Certainly the extent of

immunologic protection within the eye is not absolute, but

dependent upon a number of factors, most importantly the

specific tissue involved.34,35 The avascularity and

relative anatomic isolation of the cornea and lens endow

them with a high degree of immunologic privilege. In

contrast, the uvea is invested with a rich vascular network,

and as a result is highly susceptible to inflammatory and

immunologic responses, which are often intimately related.

There is evidence that the uvea, along with the conjunctiva,

functions as a regional lymph node for the eye, since the

corneoscleral coat and all intraocular structures are

without lymphatic drainage.35,36 It appears that antigens

presented to the eye are processed at a distant site.

Sensitized lymphocytes subsequently migrate to the antigenic

source, accumulating in the uvea or limbal conjunctiva,

where they form areas of immunologically competent cells










that can participate in cell-mediated reactions and produce

antibody locally. Although the anterior chamber is partial-

ly bordered by the reactive uvea, the immune response to

antigen within this area is significantly delayed, due to

the selective blood-aqueous barrier.34,35 Under normal

physiologic conditions, little protein is transferred across

this barrier; however, there is normally a low level of

immunoglobulin present in the aqueous.37,38,39 The blood-

retinal barrier is even more selective than the blood-

aqueous barrier, and is impervious to lymphocytes and

immunoglobulin. However, the more permeable choroidal

vessels are closely associated with the retinal pigment

epithelium and there is no barrier between the aqueous and

the vitreous or between the vitreous and the retina.34,40

Even so, only a small amount of immunoglobulin is present in

the normal retina, and this tissue normally exhibits a

considerable degree of immunologic privilege.34,3637

However, when the eye is inflammed, there is generally an

increase in vascular permeability, thus enhancing the

interaction with the immune system.

The idea that the recurrent ocular inflammation in

uveitis may be an autoimmune process has intrigued many

investigators. The factors involved in the generation of

autoimmune disease are largely obscure, but a breakdown in

self tolerance may arise through one of three major mech-

anisms: 1) a change in the distribution of a normally










sequestered self-component, allowing its exposure to

immunoreactive cells, 2) a structural change in a self-

component, allowing it to become antigenic or cross-reactive

with another antigen, 3) a basic disturbance in regulation

of the immune response, most often at the level of the

suppressor T cell.41 Autoimmune disorders can be primary or

secondary to another event, such as a viral infection or

endotoxemia. There are numerous examples of tissue-specific

autoimmune diseases. Also, the role of genetic factors in

autoimmune diseases has been clearly demonstrated both in

man and in animal models. The same is true of several

syndromes of uveitis.

The suggestion that autoimmunity might be involved in

the pathogenesis of uveitis is not new. The idea was first

proposed in 1910 by Elschnig, who demonstrated that immun-

ization with homologous uveal extracts could elicit both

complement fixing antibodies and ocular inflammation in

experimental animals, he suggested that antigens liberated

from damaged uveal tissue incited an immune response that

was involved in the production of sympathetic ophthalmia.42

Numerous studies followed, aimed at both inducing autoimmune

uveitis in experimental animals and at establishing evidence

of autoreactivity in patients with spontaneous disease. In

1925, Woods was able to demonstrate by intradermal skin

testing that there was reactivity to human uveal tissue

extracts in patients with uveitis.43 Later, in 1964,










Aronson and co-workers found precipitating antibody against

uveal tissue in the sera of patients suffering with uvei-

tis.44 However, the induction of uveitis in experimental

animals by immunization with uveal tissue proved difficult.

In general, a large amount of uveal tissue and a prolonged

series of injections were required, although Wacker iden-

tified a particulate choroidal antigen that appeared more

active than other uveal preparations.45,46,47,48 The

problems in establishing experimental disease made it

difficult to evaluate the role of the recognized immune

response in the production of disease.

Most early efforts to induce experimental uveitis

logically utilized uveal tissue since this was the primary

target, but retinal preparations have actually proved to be

far more uveitogenic. This interesting phenomenon was first

observed in 1965 by Wacker and Lipton who produced allergic

uveitis in guinea pigs by repeated immunization with

homologous retina.49 In further studies, immunization of

guinea pigs with other ocular tissues, including uvea, lens,

ciliary zonula, vitreous or optic nerve, produced little if

any pathology, while even a single injection of retina was

highly effective in the production of experimental allergic

uveitis.50,51,52 ,53 These findings were verified by a

number of studies.54,55,56'57'58

In 1977 Wacker and associates isolated from bovine

retina a soluble retinal protein, S antigen, the immuno-










pathogenicity of which was greater than that of crude

retinal preparations.59 S antigen has now been isolated

from several species and it appears to be structurally and

antigenically similar between species.60,61 S antigen is a

protein of approximately 50,000 daltons by SDS polyacryl-

amide gel electrophoresis and 42,000-44,000 daltons by

laser-light scattering and ultracentrifugation62; it appears

to have a low level of carbohydrate.63 The amino acid

sequence has been determined by standard biochemical

techniques and DNA sequence analysis, and the results reveal

local regions of homology with -transducin, a GTP-binding

protein of the retina, and other purine nucleotide binding

proteins.63,64,65 Secondary structure prediction shows that

the antigen is composed predominantly of P-pleated sheet

conformation.63

S antigen is localized primarily in the outer segment

of the photoreceptor layer. Ultrastructurally, it can be

demonstrated on the inner surface of the visual discs and on

the plasma membrane of the rod outer segment by immuno-

electron microscopy.66,67,68 In certain species, small

amounts of immunoreactive S antigen have also been iden-

tified in the vitreous, ciliary body, choroid, and the

retinal pigment epithelium. It has been postulated that

when outer segments are shed, S antigen is phagocytized by

the retinal pigment epithelium and transported to the

choriocapillaries.69,70,71 The antigen has also been










identified in pinealocytes, which are phylogenetically

related to photoreceptors.72,73 In the eye, S antigen is

synthesized in the retina and is not processed from a larger

precursor.74

S antigen is present in the photoreceptors of both

vertebrate and invertebrate species, indicating a high

degree of phylogenetic stability and suggesting an important

role in photoreceptor function.61 However, the precise

function of S antigen is still unclear. It shares certain

properties with rhodopsin kinase and the small subunit of

retinol binding protein.68,75 Broekhuyse demonstrated that

binding of S antigen to rod outer segment membranes was

induced by light, probably through an interaction with

illuminated rhodopsin, which suggests that S antigen plays a

role in phototransduction.62 Pfister has reported that S

antigen is identical to the 48K protein in rod outer

segments by biochemical, functional, immunological and

pathological tests; this protein binds to photoexited

rhodopsin and is involved in the quenching of light induced

guanosine 3'5'-monophosphate-phosphodiesterase activity.76

The sequence homologies with a-transducin and other purine

nucleotide binding proteins occur at functionally important

regions, including those sites subject to ADP-ribosylation

by toxins and those involved in phosphoryl binding. These

findings suggest that the function of S antigen is related

to that of purine nucleotide binding proteins and involves










the binding of phosphorylated rhodopsin, although a more

general function is possible as well.63,64

S antigen induced uveitis is the most extensively

studied model of autoallergic eye disease. However, other

ocular proteins such as opsin, A antigen and inter-

photoreceptor retinoid binding protein (IRBP), have more

recently been shown to be capable of producing experimental

autoimmune uveoretinitis as well.77,78,79 Immunization with

as little as 1 4g of purified S antigen in complete Freund's

adjuvant given at a site distant from the eye produces

experimental allergic uveitis in a number of species,

including guinea pigs, rabbits, rats and monkeys.59,-
80,81,82,83,84 The use of adjuvant is critical to the

production of disease. Depending on the amount given,

complete Freund's adjuvant can be highly effective.

Pertussis vaccine has also been shown to markedly facilitate

the induction of disease especially when injected concur-

rently with complete Freund's adjuvant.50,80

Early on Wacker demonstrated that the immunopathogenic

activity was related to the purity of the preparation rather

than the total amount given in a crude preparation.51,60

The reason for this has not been fully elucidated, but the

relative inactivity of crude preparations could result from

a masking of the uveitopathogenic determinants of S Ag, a

competitive effect between antigens, or the presence of

actual inhibitors. Since the elucidation of the amino acid








13

sequence of retinal S antigen, studies have been directed at

identifying the uveitogenic portion. At least one uveito-

pathogenic epitope, an 18 amino acid peptide designated as

peptide M, has been identified using synthetic peptides.

Peptide M consistently produces an experimental disease

identical to that caused by native S antigen.85,86

In most cases heterologous and homologous antigen are

both effective in producing disease, with the clinical and

histological manifestations varying, depending upon a number

of factors such as species, strain, antigen source, immun-

izing dose and amount and type of adjuvant.50,60,80,83,84

In most cases the disease resolves spontaneously, but

remission and reexacerbation has been observed in some

guinea pigs at intermediate dose levels.85 In general,

typical lesions include iridocyclitis and chorioretinitis

with focal destruction of photoreceptor cells; the infil-

trate is predominantly mononuclear, although polymorpho-

nuclear cells are also often prevalent.59,84,87 Due to the

presence of S antigen in the pineal gland, inflammatory

damages in the pineal, again typically characterized by

mononuclear infiltration, have also been recognized in some

species.88,89,90

The pathogenic mechanisms by which immunization with S

Ag produces disease are not fully understood. There are a

great deal of data suggesting that the cellular immune

response is of primary importance, although several other








14

mechanisms have also been implicated. Early experiments by

Wacker showed no correlation between the level of the

humoral response and the severity of disease, which sug-

gested that factors other than antibody were involved in

disease production.50 S antigen specific lymphocytes were

demonstrated both in vitro and in vivo in animals with

experimental allergic uveitis.91 One study indicated that

the predominant cell in the eye early in the course of the

disease was the T-helper/inducer subtype, while at a later

stage there was a relative increase of T-suppressor/

cytotoxic cells.92 Strong evidence for the pathogenicity of

these cells is provided by the fact that in inbred strains

of guinea pigs and rats experimental allergic uveitis can be

passively transferred by T cells, but generally not by

antibody.93'94'95 In particular, transfer appears to be

mediated by T lymphocytes of the helper/ inducer pheno-

type.96,97 Another observation supporting the T cell

dependence of experimental uveitis is that athymic nude rats

deficient in T cell-mediated functions do not develop

disease following injection with S antigen.98 Furthermore,

cyclosporin A, an immunosuppressant with specific anti-T

cell effects will prevent disease in susceptible spe-

cies.99,100,101

Although there is strong evidence supporting the role

of cell-mediated mechanisms, other mechanisms of hypersen-

sitivity involving antibody and complement appear to be










involved as well. Some studies have implicated an Arthus-

like reaction that can be suppressed by complement deple-

tion, while others have indicated an important role for

reaginic antibody.102,103 Interestingly, Mochizuki has

recently demonstrated a positive correlation between the

susceptibility of different rat strains to experimental

allergic uveitis and the number of choroidal mast cells;

this association also may be related to mediation by T lym-

phocytes in that mast cells facilitate the development of

delayed type hypersensitivity responses.104 It has been

suggested that a range of immunologic mechanisms come into

play, depending on a number of factors such as dose and

species, and that the differing pathogenic mechanisms

involved account for the wide variation in clinical and

histopathologic appearance.84

There is evidence that S antigen-induced experimental

uveitis is relevant to the spontaneous disease in man. Some

forms of the experimental disease clinically and histologi-

cally resemble certain syndromes of uveitis in man.84

Furthermore, autosensitization to S antigen has been

documented in many human patients. Although the results are

highly variable, depending at least in part upon the

particular subset of patients studied, evidence of both

cellular and humoral sensitization has been documented.

Several studies, using assays such as the lymphocyte

transformation test and measurement of migration inhibitory










factor, have shown a correlation between cellular immune

responsiveness and the presence of disease; one such study

demonstrated a positive response of lymphocytes to S antigen

in 53% of patients with uveitis as compared to less than 3%

of normal controls.105,106,107,108 Elevated serum anti-

bodies against S antigen have also been noted in a spectrum

of patients with uveitis, including those with toxoplasmic

uveitis, children with idiopathic chronic iridocyclitis and

children with uveitis accompanying juvenile rheumatoid

arthritis.109,110,111 Furthermore, Sainte-Laudy has

reported the presence of specific reaginic sensitization in

clinical patients.112 However, the presence of sensitiza-

tion does not always correlate with disease; in one recent

study, S antigen autoantibodies were found among a sig-

nificant number of healthy controls.107 The role of the

immune response to S antigen in clinical disease has yet to

be established, and it is still unclear whether the recog-

nized autoreactivity is the cause or the result of the

disease process.

Clearly there is a significant amount of data suggest-

ing that autoimmunity may contribute to the pathogenesis of

uveitis, but the opportunity for studying ocular immuno-

pathology in man has been restricted by the limited access

to involved tissues. Many aspects of equine recurrent

uveitis suggest that it is likewise immune-mediated: the

predominance of lymphocytes histologically, the beneficial








17

effect of immunosuppressive and anti-inflammatory drugs, the

recurrent nature, and the inability to identify a specific

causative agent. However, the mechanisms of the immune

response involved in this disease have not been extensively

studied. The horse could provide a unique opportunity to

study immunologic parameters in both experimental and

spontaneous disease in the same species. The incidence of

equine recurrent uveitis is sufficiently high to provide

ample clinical case material and the eye of the horse is

large, making it readily amenable to study and to procedures

such as anterior chamber paracentesis. It was the purpose

of this study to attempt to develop an experimental model of

autoimmune uveitis in the horse using equine retinal S

antigen. There were three major objectives: 1) to isolate

equine S antigen, 2) to develop and characterize an exper-

imental model resulting from immunization with S antigen, 3)

to evaluate humoral and cellular autoreactivity to S antigen

in horses with spontaneous equine recurrent uveitis. If an

appropriate model of allergic uveitis can be developed in

the horse, information from induced cases can be directly

compared to clinical cases, hopefully allowing for a better

assessment of the role of autoimmunity. Also, similarities

between uveitis of man and of the horse suggest that equine

uveitis could serve as a valuable model for human disease

and might thus shed light on the immunopathogenesis of

uveitis in man.













CHAPTER TWO
ISOLATION AND CHARACTERIZATION OF EQUINE RETINAL S ANTIGEN



Introduction

S antigen has been isolated from many diverse species,

including both vertebrates and invertebrates.61 While it

appears to be highly conserved, with similarities existing

between species in its composition, antigenicity and

uveitogenic potential, there are also species-specific

characteristics. For example, when comparing human and

bovine S antigen, Beneski found differences in the physical

properties, including the molecular weight, isoelectric

point and amino acid composition.113 Both unique and shared

antigenic determinants were also demonstrated. Furthermore,

analysis by monoclonal antibodies of S antigen from several

species has documented both shared and species specific

epitopes.61,I14

The differences existing in S antigen of different

species may affect the immunopathogenic activity of the

antigen. Beneski reported that the uveitis produced in

guinea pigs by human S antigen was mild in comparison with

that induced by bovine antigen.113 Early studies using

crude retinal homogenates suggested that homologous tissue

was more uveitogenic than was heterologous.55,115 Later,








19

Wacker found this to also be true for purified homologous S

antigen.83

These findings suggest that it may be advantageous to use

a homologous system when studying experimental autoimmune

uveitis and immune responsiveness in uveitis patients. It

may also be more relevant to spontaneous disease, since the

damaging immune response in naturally occurring uveitis

would be directed against autologous antigen if autoimmunity

to S antigen plays a role. Nevertheless, due to the

relatively large amounts of retinal tissue required to

obtain an adequate yield of S antigen, most studies have

employed bovine S antigen. The purpose of this study was to

isolate and characterize equine S antigen so that homologous

protein could be used in subsequent experiments.

Materials and Methods

Reference Antigens and Antisera

There were two reference preparations of bovine S

antigen, the first prepared by gel filtration and ion

exchange chromatography (kindly provided by Dr. Wacker,

Louisville, KY) and the second prepared by high performance

liquid chromatography (kindly provided by Dr. Noveen Das,

Gainesville, FL).

Reference antisera utilized as standards in these

studies included guinea pig antiserum to guinea pig S

antigen (kindly provided by Dr. Wacker) and rabbit antiserum

to bovine S antigen (kindly provided by Dr. Das).










Quantitation of Protein

Total protein determinations in retinal preparations

were made by the Coumassie blue dye-binding technique of

Bradford using bovine albumin as the standard.116

The concentration of S antigen was measured by single

radial immunodiffusion according to the method of Mancini,

Carbona and Heremans, utilizing the guinea pig antiserum to

guinea pig S antigen and bovine S antigen as the ref-

erence.-17 Since the antiserum could preferentially

recognize S antigen from different species, equal total

protein amounts from preparations of bovine and equine S

antigen were compared by immunodiffusion.

Retinal Tissue

Equine retinas were harvested from normal eyes at a

slaughterhouse and placed on ice for up to 48 hours before

freezing at -700 C.

Preparation of Equine Retinal S Antigen by Gel Filtration

and Ion Exchange Chromatography

S antigen was isolated from equine retinas according to

the procedure outlined by Wacker with some modifications.59

All steps were carried out at 40C and all buffers incor-

porated 0.04% sodium azide and 2.OmM benzamidine. A 20%

weight per volume suspension of retina was prepared in 0.01

M potassium phosphate buffer, pH 7.6. The suspension was

stirred gently for one hour and then ultracentrifuged at

30,000 rpm (80,000 x G) for one hour. Saturated ammonium








21

sulfate, pH 7.2, was added to the pooled supernatants until

50% saturation was obtained. This mixture stood for two

hours and was centrifuged at 1,200 x G for 30 minutes. The

precipitate was dissolved in deionized water and dialyzed

first against 0.15 M saline, pH 7.2, and then against 0.05 M

potassium phosphate buffer, pH 7.2, containing 0.1 M NaCl.

This solution was then ultracentrifuged at 30,000 rpm for

one hour; the supernatant was concentrated by ultrafiltra-

tion (CX-30, Millipore Corporation, Bedford, MA) and

filtered through a 0.45 Lm membrane (Acrodisc, Gelman

Sciences, Ann Arbor, MI).

The concentrate was applied to a Sephacryl S-200 column

(Pharmacia Fine Chemicals, Piscattaway, NJ) equilibrated

with the potassium phosphate buffer, and calibrated with

molecular weight standards (Bio-Rad Laboratories, Richmond,

CA). Samples were eluted with potassium phosphate buffer

and collected in 7.0 ml. fractions. The fractions were

analyzed for the presence of S antigen by radial immunodif-

fusion; those fractions containing S antigen were pooled,

concentrated and dialyzed against 0.01 M Tris buffer, pH

7.9.

This material was applied to a column of DEAE Bio-Gel A

(Bio-Rad Laboratories) equilibrated with the Tris buffer.

Columns were initially eluted with Tris buffer and then with

a linear salt gradient using the same buffer containing 0.1

M NaCl. Any residual protein was eluted with 0.5 M NaCl in










the starting buffer. The eluant was collected in 3.0 ml

fractions, and antigenic activity was again identified by

radial immunodiffusion. Fractions containing S antigen were

pooled, concentrated and dialyzed against a 0.15M saline

solution. The amount of protein in the resulting S antigen

preparation was determined by the Coumassie blue dye-binding

technique of Bradford.116 Antigen for use in immunization

studies was further dialyzed in saline without azide or

benzamidine and sterile filtered with a 0.2 iim filter

(Acrodisc, Gelman Sciences); the protein determination was

repeated after filtering. The antigen was stored in

aliquots at -700C.

Polyacrylamide Gel Electrophoresis

Polyacrylamide gel electrophoresis in sodium dodecyl

sulfate (SDS-PAGE) was carried out according to standard

methodology in slab gels composed of a 15% running gel and a

5% stacking gel. Ten to forty 4g of protein were applied to

each lane. Gels were stained with either Coumassie blue or

silver (Bio-Rad Silver Stain, Bio-Rad Laboratories).

Protein standards (Bio-Rad Laboratories) were used for

calculation of molecular weight according to the relative

mobilities.

Densitometry

A representative polyacrylamide gel stained with

Coumassie blue was evaluated by densitometry using a Bio-Rad

Model 620 densitometer (Bio-Rad Laboratories).










Production of Antisera to S Antigen

Polyvalent antisera to S antigen were produced in

rabbits. Two hundred fifty 4g of equine S antigen were

electrophoresed on a SDS-polyacrylamide gel with a single

well. The 53,000 dalton (53 kDa) molecular weight band

corresponding to the major protein in the bovine S antigen

preparation was cut from the gel and emulsified in complete

Freund's adjuvant (CFA, Difco Laboratories, Detroit, MI).

Two rabbits received three immunizations with this material

at two-week intervals. The antibody was characterized by

an enzyme linked immunosorbent assay (ELISA) and by immuno-

blotting.

The same antigen preparation was used to prepare

monoclonal antibodies to equine S antigen by standard

methods as described by Waldman and Milstein.I18 Five Balb

C mice were immunized subcutaneously three times at three-

week intervals. Four days after the last immunization the

spleens were removed and the isolated splenic lymphocytes

fused to the mouse myeloma line Sp. 2/0 using polyethylene

glycol. The fused cells were plated in 24-well polystyrene

tissue culture plates (Corning Glass Works, Corning, NY) and

grown in medium with hypoxantinie, aminopterin and thy-

midine (HAT medium, Sigma Chemical Co., St. Louis, MO) to

select for hybridomas. Hybridoma supernatants were screened

for the presence of antibody by ELISA. Positive wells were

cloned by limiting dilution, and the resultant clones were








24

again checked for the presence of antibody. The isotype was

determined by ELISA (Mouse Monoclonal Sub-isotyping Kit,

Hyclone, Logan, UT) and after it was established that a

single isotype was present, mice were innoculated for the

production of ascites. Mice were primed with 0.1 ml

Pristane (Sigma Chemical Co.) and injected intraperitoneally

with 107-108 hybridoma cells 1-3 weeks later. The resultant

ascitic fluid was harvested and tested for specific antibody

by ELISA and immunoblotting.

Detection of Antibody to S Antigen by ELISA

Antisera. The antisera employed were affinity purified

rabbit anti-guinea pig IgG, sheep anti-mouse IgG and goat

anti-rabbit IgG, all conjugated to alkaline phosphatase

(Sigma Chemical Co.). The optimal concentration for each

antiserum was determined by preliminary checkerboard

titrations.

Assay. The ELISA technique used by Gregerson was

adapted to detect antibody to S antigen in rabbits and

mice.119 The optimal concentration of antigen was again

determined by preliminary checkerboard titrations. Poly-

styrene microtiter plates (Immulon I, Dynatech, Alexandria,

VA) were coated with 0.1 4g/well of equine or bovine S

antigen in 60 i of carbonate-bicarbonate buffer with 0.02%

sodium azide, pH 9.6 and incubated overnight at 40C.

Unbound antigen was removed by washing with PBS, pH 7.2,

containing 0.2% Tween 20 (Fisher Scientific Co., Fair Lawn,










NJ). Sixty microliters of dilutions of sera, hybridoma

supernatant, or ascitic fluid in PBS with 0.04% Tween 20 and

0.5% human serum albumin (Cohn, Fraction V, Sigma Chemical

Co.) were added, incubated overnight at 40C, and the wells

were washed with PBS-Tween. The alkaline phosphatase

conjugated antisera were diluted in PBS-Tween-HSA and 60 i

were added to each well and incubated overnight at 40C.

Wells were again washed with PBS-Tween before the addition

of 60 il of substrate, which was p-nitrophenyl phosphate

(Sigma Chemical Co.) in carbonate-bicarbonate buffer, pH 9.8

with 0.1mM MgCI2. Plates were allowed to develop for 1-2

hours at room temperature and optical densities at 405 nm

were determined using a Titertek Multiskan filter photometer

(Flow Laboratories, Inc., McLean, VA).

Incorporated in each plate were dilutions of a positive

control serum, either the reference guinea pig anti-guinea

pig S antigen, the reference rabbit anti-bovine S antigen,

or samples that had been shown to have significant reac-

tivity. Sera from non-immunized guinea pigs, rabbits and

mice were used as negative controls. Also, on each plate

controls consisting of coated wells with no sample and wells

with no antigen were run. All samples were run in trip-

licate at a range of dilutions.

Immunoblotting

The resolved proteins were electrophoretically trans-

ferred from gels to nitrocellulose paper and immunoblotted










essentially as described by Towbin and Tsang.120,121

Briefly, the gel to be transferred was placed next to a

nitrocellulose membrane (Bio-Rad Laboratories) and transfer

was performed at 40C in a buffer of 0.025 M Tris, 0.193 M

glycine and 20% methanol, pH 8.35, using the Trans-Blot Cell

(Bio-Rad Laboratories).

Once the antigen was bound to the nitrocellulose, the

immunoblotting was completed at room temperature. Gentle

agitation was employed in each step. The paper was washed

in 20 mm Tris with 500 mM NaCl, pH 7.5, for 10 minutes and

unbound sites were blocked by washing in 3% gelatin in the

Tris buffered saline for one hour. The membrane was then

incubated with the first antibody, which was the sample

diluted in Tris buffered saline containing 1% gelatin for 12

hours. The membrane was rinsed briefly with deionized

water, washed twice for 10 minutes with Tris buffered saline

with 0.05% Tween-20, and then incubated for two hours with

the second antibody again diluted in the Tris buffered

saline with 1% gelatin. The second antibody was a goat

antiserum to either guinea pig immunoglobulins (Accurate

Chemicals, Westbury, NY), rabbit immunoglobulins (Sigma

Chemical Co.) or mouse immunoglobulins (Sigma Chemical Co.)

conjugated with horseradish peroxidase. After incubation

with the second antibody, the washing procedure was repeated

and the nitrocellulose was immersed in the color development

solution consisting of 60 mg horseradish peroxidase reagent








27

(Bio-Rad Laboratories). in 20 ml cold methanol mixed with 60

-I cold 30% H202 in 100 ml of Tris buffered saline until

adequate color development occurred. Development was

stopped by immersing the membrane in distilled water for ten

minutes.

Production of Uveitis in Lewis Rats

Five Lewis rats were immunized subcutaneously with 50

;g of equine retinal S antigen emulsified with 500 4g of

complete Freund's adjuvant (H37Ra, Sigma Chemical Co.) in

equal volumes. Two rats were similarly immunized with

bovine S antigen and two control rats received adjuvant

alone. Daily ophthalmic examinations were done for two

months and weekly thereafter for another month.

Histopathology

Eyes were enucleated, fixed in Zenker's solution and

embedded in paraffin. Sections were stained with hema-

toxylin-eosin and periodic acid schiff.

Results

Isolation of Equine S Antigen

Several preparations of equine retinal S antigen were

made using the described procedure. The results from a

representative separation are summarized in Table 1. Since

the guinea pig antiserum to guinea pig S antigen used to

quantitate S antigen was shown by ELISA to have greater

reactivity with bovine S antigen than equine, the measure-

ment of equine S antigen might not be totally accurate.








Table 1:
retina.


Fraction


Purification of S antigen from 50 g of equine


Total protein'
mg %


S antigen2 % S antigen
mg %


Crude extract


810


50% ammonium sulfate 324
precipitate


Sephacryl S-200
filtration (fr 2)

DEAE Bio-Gel A
Chromatography


100.0

40.0


5.2


0.6


23 100.0

16 60.0


10.1 51.3


4.4 19.1


1 Bradford reaction
2 Radial immunodiffusion


2.8

4.8


24.0


88.0










However, comparison of equal total protein amounts of the

two antigens by immunodiffusion gave similar results.

Eighty equine retinas weighing 50 grams were homo-

genized and after ammonium sulfate fractionation 324mg of

protein containing 16mg of S antigen was obtained. This

material was separated on Sephacryl S-200 with the antigenic

activity eluting in the 44-158 kDa molecular weight range

(Figure 1). The most antigenic fraction was fraction 2,

which had 42mg of total protein and 10.1mg of S antigen

after concentration and dialysis. This material was applied

to a DEAE Bio-Gel A column, and S antigen eluted in the main

protein peak in the middle of the salt gradient (Figure 2).

In this experiment the final S antigen preparation had

5mg of protein and 4.4mg of S antigen; this represented a

yield of 19% of the S antigen available in the crude retinal

extract. The percent recovery in subsequent preparations

was similar, ranging from 18% to 32%. Alternative methods

of isolation, such as gel filtration and hydrophobic

interaction chromatography, were also used (data not shown).

Characterization by Polyacrylamide Gel Electrophoresis

All preparations of equine S antigen contained a major

band of approximately 53 kDa when calculated on the basis of

relative mobility; this molecular weight was just slightly

larger than that of the major band present in the reference

preparations of bovine S antigen (Figures 3 and 4). Several

minor bands were also present in all preparations of S

























Figure 1: Gel filtration chromatography on Sephacryl S-200 of an ammonium
sulfate precipitate of equine retinal extract. A sample volume of six ml
containing 324 mg of protein was applied to the column and eluted with 0.05M
potassium phosphate buffer containing O.IPI NaCl, p1 7.2. Seven ml fracLions
were collected and analyzed for S antigen by radial immunodiff Ision.
Fractions containing S antigen were pooled and designated as fracLion 2.
absorbance concentration of S antigen












-~0 C0
o 1.0




'0.8 -0.8

cO
E fr 2
-0.6
o 0.6-.
co (I)
(3)3
o L
C *.0.4 "
0 4 -

0
CO



0.2 0.2




0- - - . .-- - ----------- _

0 80 100 120 140 160 180 200 220 240 260
Fraction N umber

























Figure 2: Ion exchange chromatography of fraction 2 on DEAE Bio-Ge] A.
Forty-two mg of protein in a 5 ml volume were applied to the column and
eluted using a linear salt gradient in 0.01M Tris buffer, pH 7.9. The
eluant was collected in 5 wi. fractions, and antigenic activity was
identified by radial immunodiffusion.
absorbance concentration of S antigen












16
1.6


1.4- S Ag
-1.4


1 .2- t1.2



E 1o0 (,0
r 1 0 >
0 :
co
C'\ t- (0
(D
.8-- -.8

r- 6- ca 3
0 .6"





oL
L. .6.- ci .4inb3r




5,,
.4-. K -.
Frcio uIe

















c

Z. 2
Uj < < -

0 0
M M W W 2

r 92,500
66,200

45,000

31,000

21,500

14,400




a b c d ef






Figure 3: Analysis by SDS-PAGE in a 15% slab gel of a) 40
4g crude equine retina b) 20 4g bovine S antigen prepared
by HPLC c) 20 4g bovine S antigen prepared by gel
filtration and ion exchange chromatography d) 20 4g equine
S antigen prepared by gel filtration and hydrophobic
interaction chromatography e) 20 4g equine S antigen
prepared by gel filtration and ion exchange chromatography
f) molecular weight standards. The gel was stained with
silver stain.




















5.0

x = bovine S antigen

4.8 13 = equine S antigen



4.6



2 4.4



4.2



40
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
RELATIVE MOBILITY






Figure 4: Molecular weight determination of S antigen by
PAGE. The major protein band of bovine S antigen was
calculated to have a molecular weight of 51 kDa while tat
of equine S antigen had a molecular weight of 53 kDa.










antigen, including both standards of bovine S antigen.

These minor bands were variable between preparations and

ranged in size from 74 kDa to 18 kDa with the majority being

smaller in size than the 53 kDa major band. Although all

preparations were stored in aliquots at -700C, repeated

analysis by PAGE revealed changes with time in the band

patterns seen in both equine and bovine S antigen, usually

observed as a shift to smaller molecular weights or as a

doublet in the 50-53 kDa molecular weight range (Figure 5).

Densitometry Results

Densitometry of the equine S antigen preparation

indicated that there was a major band at 53 kDa containing

54% of the protein. (Figure 6). The remainder of the

protein was distributed between four distinct bands smaller

in size, representing approximately 35% of the protein, and a

band larger in size, representing 11%.

Characterization of the Antisera by ELISA

Reference antisera. The ELISA was a sensitive and

reproducible technique for detecting antibody to S antigen.

Titration of the reference antisera, rabbit anti-bovine S

antigen and guinea pig anti-guinea pig S antigen, yielded

standard curves with a linear portion having a correlation

coefficient(r) greater than or equal to 0.99 (Figure 7).

The slope of the line generated by the rabbit antiserum was

almost identical to that generated by the guinea pig















0 0







45,000
-45,000











ab c d



Figure 5: Analysis by SDS-PAGE in a 15% slab gel of a)
molecular weight standards b) 20 4g bovine S antigen
prepared by gel filtration and ion exchange chromatography
c) 20 4g bovine S antigen prepared by gel filtration and
ion exchange chromatography d) molecular weight standards.
Gels were stained with Coumassie blue. Samples of bovine S
antigen were of the same preparation of bovine S antigen
and gels were run in the same week.









































Figure 6: Densitometric analysis of equine S antigen.
Twenty 4g of equine S antigen were electrophoresed in a 15%
SDS polyacrylamide gel. The gel was stained with Coumassie
blue and evaluated using a Bio-Rad Model 620 densitometer;
the results demonstrated that 54% of the protein was
present in the 53 kDa band.
































Figure 7: Detection of antibody to S antigen by ELISA.
Titration of reference antisera against a) 0.1 qg bovine S
antigen b) 0.1 qg equine S antigen. Q- rabbit anti-
bovine S antigen 'A- guinea pig anti-guinea pig S antigen
0- normal rabbit serum 0- normal guinea pig serum




































16 32 64 128
Dilution (1/n X 102)


256 512


----_ _

0
4 3 16 32 64 128
Dilution (1/n X 102)


256 512


1.8
a)
1.6

1.4


= 1.2
LO
1.0

0.8


o 0.6
0.
0.4


0.2

0


1.0
E
0.8


0.6

0o.4
0


, -


< 0.


T








41

antiserum, suggesting that the two antisera are recognizing

the same antigen. Although both antisera recognized equine

and bovine S antigen, reactivity with the bovine antigen was

greater, suggesting that there is a difference in the

epitopes presented in the two preparations. The serum of

some normal rabbits had a low level of reactivity to S

antigen.

Rabbit Antisera. Antiserum was raised in two rabbits

against the 53 kDa band of the equine S antigen preparation.

The resulting polyvalent antisera were shown to recognize

both equine and bovine S antigen by ELISA, with recognition

of the equine antigen being slightly stronger (Figure 8).

Again, the slope of the linear portion of the curve equaled

that of the positive control antiserum, rabbit anti-bovine S

antigen. The reaction of the prepared antiserum with equine

S antigen was slightly stronger than that of the positive

control.

Hybridoma Antisera. Two hybridoma clones were iso-

lated, P2CI, and P4A4. Both were of the IgM isotype with

kappa light chains. Once it was established that a single

isotype was present, ascites was produced. As determined by

ELISA, P2CI and P4A4 recognized both equine and bovine S

antigen, although reactivity with bovine S antigen was weak

(Figure 9).
























Figure 8: Detection of antibody to S antigen in rabbit serum by ELISA.
A reference rabbit anti-bovine S antigen, equine S antigen coating the
wells A- rabbit 102 pre-immunization, equine S antigen coating the wells
* rabbit 14)2 post-immunization with equine S antigen, equine S antigen
coating the wells 0- rabbit 102 post-immunization with equine S antigen,
bovine S antigen coating the wells




















1.0-




.8
E
LO
0


C,

e~ .4-




.2

0-1
0- 1 I I I I-

2 4 8 16 32 64 128 256 512 1024 2048

Dilution (1/n X 102)


























Figure 9: Detection of antibody to S antigen in mouse ascites by ELISA.
A P2C, ascites, equine S antigen coating the wells 0- PC1 ascites,
bovine S antigen coating the wells 0 unrelated hybridoma ascites, equine
S antigen coating the wells


















1.0



.8


2 4 8 16 32 64


Dilution (I/n)


128


256










Characterization of the Antisera by Immunoblotting

Both the guinea pig anti-guinea pig S antigen and the

rabbit anti-bovine S antigen demonstrated activity against

bovine and equine S antigen upon analysis by immunoblotting.

The major bands of 53 kDa and 51 kDa were recognized as well

as multiple other bands in both antigen preparations

representing both smaller and larger proteins (Figure 10).

The antisera to equine S antigen raised in rabbits displayed

a similar pattern. Notably, both monoclonal antibodies

raised against the 53 kda recognized multiple protein bands

as well; P4A4 demonstrated strong reactivity with a band

which was much smaller in size (Figure 11).

Production of Uveitis in Lewis Rats

Neither of the two negative control rats injected only

with complete Freund's adjuvant developed uveitis. Both of

the positive control rats injected with bovine S antigen in

complete Freund's adjuvant developed uveitis. In one case,

the experimental allergic uveitis observed was typical of

that described by other researchers.80,81 Briefly, on day

14 post-immunization the rat developed bilateral uveitis

characterized clinically by corneal edema, marked flare,

hypopyon and iridocyclitis. In the left eye, which was most

severely affected, the retina could not be visualized;

however, in the right eye it appeared clinically normal.

The condition progressed for five days, and then began to

gradually improve until it was clinically resolved by day 28




















0C

0
0


S

L


R anti-Bov. S Ag.


0
03
o a ca
CO wU LU


R anti-Eq. S Ag.


Figure 10: Evaluation of rabbit antisera by
immunoblotting. Twenty 4g of protein were electrophoresed
in each lane and transferred to nitrocellulose paper. The
nitrocellulose was then incubated with rabbit serum
followed by incubation with goat antiserum to rabbit
immunoglobulin conjugated with horseradish peroxidase a)
reference rabbit anti-bovine S antigen serum b) rabbit
antiserum raised against the 53 kDa band of the equine S
antigen preparation c) rabbit serum prior to immunization.


0
0
0
o "
* L.U


NRS


~. ,








48






C/)
CO
0 C)i




























mAb P4A4 GP anti-GP
S Ag.


a b



Figure ii: Evaluation of antibodies in hybridoma ascites
by immunoblotting. Twenty 4g of protein were
electrophoresed in each lane and transferred to
nitrocellulose paper. The nitrocellulose was then
incubated with a) hybridoma ascites produced against the 53
kDa band of the equine S antigen preparation followed by
goat antiserum to mouse immunoglobulin conjugated with
horseradish peroxidase b) reference guinea pig antiserum to
guinea pig S antigen followed by goat antiserum to guinea
pig immunoglobulin conjugated to horseradish peroxidase.










post-immunization. The second rat developed a similar

disease, but clinical signs were not observed until day 80

post-immunization; the condition also resolved.

Of the five rats immunized with equine S antigen in

complete Freund's adjuvant, two developed uveitis, with the

onset of inflammation being between days 90 and 100 post-

immunization in both cases. The clinical signs were similar

to those in the rats immunized with bovine S antigen.

Infiltration of the anterior chamber was so severe that the

retina could not be visualized in any of the affected eyes;

however histologic evaluation of eyes from one rat with

acute disease revealed a severe panuveitis with infiltration

of the retina by inflammatory cells (Figure 12). One rat

was maintained, and unlike those immunized with bovine S

antigen, developed a chronic uveitis resulting in bilateral

synechiae, cataracts and neovascularization of the corneas

(Figure 13).

Discussion

Equine retinal S antigen was isolated by a combination

of gel filtration and ion exchange chromatography. The

yield of this technique ranged from 18-32% of the S antigen

available in the crude homogenate, which is comparable to

that of Wacker (24%)59 and to that of other investigators

using varying techniques.122,123

In this experiment, the equine S antigen preparation

had a major band of 53 kDa as calculated by its relative
















- q~j~N~h~


Figure 12: Retinitis in the eye of a Lewis rat three
months after subcutaneous immunization with 50 4g of equine
retinal S antigen in CFA and 7 days after the onset of
clinical signs.








































Figure 13: Severe changes in the eye of a Lewis rat 18
months after subcutaneous immunization with 50 jig of equine
retinal S antigen in CFA. The uveitis is chronic, with
neovascularization and cataracts present.










preparation was consistently slightly larger than that of

the bovine standards, which had molecular weights of 51 and

52 kDa; this difference could be accounted for by a species

variation in the amino acid structure or level of glycosyla-

tion. Also, a difference in the extent of degradation could

be involved.

In addition to the 53 kDa band, which accounted for 54%

of the protein by densitometric analysis, multiple other

bands ranging in size from 18-75 kDa were present in all

preparations of equine S antigen. Several alternative

methods of purification were attempted, including those of

Dorey, Borthwick, and Kasp, but they did not alter the

composition of the preparation (data not shown).122,123,124

The presence of multiple bands in S antigen preparations is

a common occurrence, as is evidenced by the reports of

numerous other investigators and by the presence of such

bands in both standards of bovine S antigen, including that

prepared by HPLC.125 It is important to establish whether

these represent significant contamination by unrelated

proteins or whether they are related to S antigen. Although

the presence of a low level of contamination can not be

ruled out based on these experiments, it appears that most

of the material is antigenically related to S antigen. Upon

immunoblot analysis, all of the antisera, including those

raised specifically against the 53 kDa band, recognized all

bands to some degree, including those greater than 53 kDa in










size. The strongest evidence of an antigenic relationship

is that both monoclonal antibodies raised against the 53 kDa

band recognized multiple proteins; in fact, P2C4 reacted

most strongly with some of the smaller molecular weight

species. Thus, it appears that the majority of protein

present is antigenically related to S antigen, with the

minor bands resulting from degradation and aggregation.

Definite changes in the electrophoretic pattern of both

equine and bovine S antigen preparations were observed

throughout the experiments, even though the antigens were

stored in aliquots at -700C in the presence of protease

inhibitors. S antigen has been shown to be unstable over

wide pH or temperature variations and to be susceptible to

proteolytic degradation; it also has a tendency to ag-

gregate, which may be enhanced by limited proteolysis.59,-
122,125,126 Degradative changes such as those observed

could account for the heterogeneity in size, isoelectric

point and antigenicity observed between preparations of S

antigen.

Equine S antigen produced an experimental uveitis in

the Lewis rat model. Although the incidence of the disease

(2/5 rats) was lower than that generally reported for bovine

S antigen in the Lewis rat, it is impossible to draw any

firm conclusions due to the number of animals studied.80,-
81,127 A lower uveitogenicity could be attributed to

species differences or differences in the epitopes presented








54
due to variable degradation. Beneski, who reported only

mild uveitis in guinea pigs immunized with relatively high

doses of human S antigen, explained this difference by

autolytic changes occurring during the long delay in

harvesting the retinas, which could be a factor in the case

of equine S antigen.113 Despite the lower incidence of the

disease, the equine S antigen isolated in this study was

shown to be uveitogenic, producing a severe chronic uveitis.













CHAPTER THREE
IMMUNOLOGIC REACTIVITY TO S ANTIGEN IN NORMAL HORSES
AND THOSE WITH UVEITIS



Introduction

There are several investigations documenting immuno-

logic reactivity to S antigen in human patients with

uveitis, which gives support to the theory that autoimmunity

to S antigen is important in the pathogenesis of uvei-

tis.105-112,128 Both humoral and cellular responsiveness to

S antigen have been correlated with the presence of disease;

however, the actual role of the recognized autoreactivity in

the pathogenesis remains unclear. On the other hand, other

studies have failed to confirm that a correlation ex-

ists.107,110,129 Factors such as the subset of uveitis

patients studied, the stage of the disease, and whether

homologous or heterologous antigen is used may all influence

the results. In certain types of uveitis and other organ-

specific diseases, either autoimmune or infectious in

origin, it has been shown that only the local immune

response, and not the systemic response, correlates with the

presence of disease.130,131,!32 To date, the local response

to S antigen has not been investigated in man.

It is important to assess the role of autoreactivity to

S antigen in equine recurrent uveitis. Due to the avail-

55










ability of antigen and the ability to perform anterior

chamber paracentesis, local and systemic reactivity to

homologous S antigen can be investigated in the this

species. In this study, immunologic reactivity to S antigen

in normal horses and those with uveitis was assessed by

measuring relative anti-S antigen antibody levels in serum

and aqueous humor by ELISA and by detecting cellular

responsiveness by lymphocyte transformation.

Materials and Methods

Population Sampled for Antibody Studies

All animals received a thorough physical examination

and an ophthalmic examination including at least direct and

indirect ophthalmoscopy.

Controls. Serum samples were obtained from 17 healthy

horses with no history or clinical evidence of ocular

disease. Additionally, serum and aqueous humor samples were

collected from 14 horses and ponies which had no clinical

signs of ocular problems; however, complete histories were

not available.

Animals with uveitis. Forty five animals with acute or

chronic uveitis diagnosed either at the University of

Florida Veterinary Medical Teaching Hospital or by referring

veterinarians were sampled. Serum was obtained from all

cases and 23 aqueous samples were obtained from 15 of the

cases.










Population Sampled for Lymphocyte Studies

Controls. Lymphocytes from seven clinically normal

animals were assayed; again, no extended histories were

available.

Animals with uveitis. Lymphocytes from seven animals

with either acute or chronic uveitis were assayed.

Sampling Techniques for Antibody Studies

Blood was drawn by jugular venipuncture, allowed to

clot, and the serum was frozen in aliquots at -200C until

assayed.

Animals were anesthetized for the collection of aqueous

humor. In most cases, 0.2-0.4 mg/kg of xylazine (RompunR,

Haver-Lockhart Division of Cutter Laboratories, Shawnee, KS)

followed by 0.4-0.6 mg/kg of ketamine hydrochloride (Ket-

asetR, Bristol Pharmaceuticals, Syracuse, NY) was used.

Analgesia was reinforced locally with proparacaine hydro-

chloride (AlcaineR, Alcon, Fort Worth, TX). In some

clinical patients, aqueous was obtained immediately after

surgical enucleation. Anterior chamber paracentesis was

performed with a 26 gauge needle directed through the sclera

into the iridocorneal angle; a maximum of 1.0 ml of aqueous

humor was withdrawn. This has been shown to be a safe

procedure which in itself produces little change in the

aqueous even with repeated sampling.133,134 Samples were

again frozen at -200C until assayed.










Preparation of Retinal S Antigen

S antigen was isolated from equine retinas by gel

filtration and ion exchange chromatography as described in

Chapter 2. Bovine S antigen was provided by Dr. Wacker.

Preparation of Antisera

A reference antiserum of guinea pig anti-guinea pig S

antigen was provided by Dr. Wacker. A standard reference

antiserum of equine anti-equine S antigen was prepared from

pooled sera from multiple bleedings of a pony immunized

subcutaneously with 1.0 mg equine S antigen in complete

Freund's adjuvant.

Goat anti-guinea pig immunoglobulin conjugated to

alkaline phosphatase and goat anti-guinea pig immunoglobulin

conjugated to horseradish peroxidase were purchased (Ac-

curate Chemicals).

Goat anti-equine IgG was raised against protein from a

33 1/3% saturated ammonium sulfate precipitate of pooled

normal equine sera that was in the pass through volume from

diethylaminoethyl cellulose (DEAE, DE52, Whatman, Maidstone,

Kent, England) in 0.025 M phosphate buffer, pH 8.0. The

antiserum detected IgG a,b and IgG (T) upon immunoelectro-

phoresis of normal equine serum. It was rendered specific

for heavy chains by passage through a cyanogen bromide

activated column of Sepharose 4B (Pharmacia Fine Chemicals)

to which had been coupled equine Fab'2 prepared by digestion

with pepsin (Sigma Chemical Co.) and subsequent filtration








59

on Sephadex GI00 (Pharmacia Fine Chemicals). This prepara-

tion demonstrated spontaneous activity against equine S

antigen which was removed by passage through a similar

affinity column which had equine S antigen coupled to it.

The resultant antiserum was conjugated to alkaline phospha-

tase (Type VII-S, Sigma Chemical Co.) by a modified glut-

araldehyde procedure.135

ELISA

Performance of the assay. The ELISA procedure outlined

in Chapter 2 was used. Briefly, polystyrene microtiter

plates (Immulon I, Dynatech) were coated with 0.1 4g/well of

equine or bovine S antigen. Sixty microliter samples of

sera or aqueous humor diluted in PBS with 0.04% Tween 20 and

0.5% human serum albumin (Cohn, Fraction V, Sigma Chemical

Co.) were assayed using the described conjugated antiserum.

Optimal concentrations of antigen and conjugated antisera

were determined for each assay by preliminary checkerboard

titrations. The substrate used was p-nitrophenyl phosphate

(Sigma Chemical Co.) in carbonate-bicarbonate buffer, pH

9.8, with 0.1mM MgCl2. Incorporated in each plate were

dilutions of the same positive standard reference equine

antiserum. As a negative control, a pool of precolostral

foal sera was used. Also included on each plate were

further negative controls consisting of two antigen coated

and two noncoated wells incubated with diluting solution

followed by conjugate and substrate. Serum samples were run










in triplicate at a range of dilutions which gave values in

the linear part of the standard curve. Due to extremely low

values, aqueous samples were run undilute and at a 1/2

dilution.

Determination of relative antibody units (RAU). An

arbitrary value of 100 antibody units/ml was assigned to the

undiluted standard reference serum. Relative antibody unit

values for samples were then calculated through the use of

linear regressions by interpolating from the absorbance of

dilutions which fell in the linear portion of the standard

curve.

Immunoblot Analysis

The proteins recognized by the antisera were examined

by immunoblotting. The procedure was performed as pre-

viously described, except that for equine serum samples

three incubation steps were employed. Following electro-

phoresis and transfer, the nitrocellulose was incubated with

the sample equine serum, then with the described goat anti-

equine IgG, and finally with rabbit anti-goat immunoglobulin

conjugated with horseradish peroxidase (Sigma Chemical Co.).

Quantitation of Immunoglobulin by Single Radial Immuno-

diffusion

IgG in serum and aqueous humor was quantitated by

single radial immunodiffusion after the method of Mancini,

Carbonara and Heremans!17 using purified antibody to equine

IgG (Miles Scientific, Naperville, IL).










Lymphocyte Transformation

Blood was collected by jugular venipuncture into

syringes with 50 units of preservative free heparin (Sher-

wood Medical Industries, St. Louis, MO) in Rosewell Park

Memorial Institute 1640 medium (RPMI 1640, Flow Labora-

tories) per ten ml of whole blood. The mononuclear cells

were separated using a Ficoll-Hypaque density gradient

(Histopaque-1077, Sigma Chemical Co.).136,137 The harvested

cells were suspended in RPMI 1640 medium supplemented with

10% heat inactivated fetal bovine serum (Flow Laboratories),

1% L-glutamine, 1% penicillin-streptomycin and 2% kanamycin.

Cultures were set up in 96 well tissue culture plates

(Corning Glass Works) with each well receiving 80,000

lymphocytes in 100 4I of the supplemented RPMI. Triplicate

wells received either 100 4i of supplemented RPMI alone, 100

41 of medium containing .25, .5, 1.0 or 2.0 Lg of phytohema-

gluttin P (PHA-P, Sigma Chemical Co.), or 100 41 of medium

containing 0.1, 1.0, 3.0 or 10 4g of equine S antigen.

Cultures were incubated for 72 hours at 370C in 5% C02,

pulsed for 16 hours of incubation with 1 4Ci per well of

tritiated thymidine (New England Nuclear, No Billerica, MA)

in 10 41 of RPMI, and harvested onto filter paper using a

cell harvester (Model M12V, Brandel, Rockville, MD). The

incorporated radioactivity was measured by a scintillation

counter (Delta 300 Liquid Scintillation System, TM Analytic,

Inc., Elk Grove Village, IL). Results were expressed as the










stimulation index (S.I.), which is derived by dividing the

mean counts per minute from wells containing antigen or

mitogen by the mean counts per minute from unstimulated

control wells.

Results

Antibodies to S Antigen

The ELISA was a reproducible and sensitive method of

detecting antibody to equine or bovine S antigen in both

serum and aqueous humor. The reference antiserum of guinea

pig anti-guinea pig S antigen, the reference positive equine

serum and the sample sera all produced curves with similar

slopes (Figure 14). That the equine sera recognized S

antigen was further confirmed by immunoblot analysis (Figure

15). The ELISA assay was consistently sensitive down to the

level of 0.08 relative antibody units. The pool of precolo-

stral foal sera used as a negative control did not demon-

strate reactivity.

A total of 62 sera were tested for the presence of

antibodies to equine S antigen: 17 sera from normal controls

and 45 sera from patients with uveitis. In both groups, the

majority of samples contained IgG antibodies to S antigen

(Figure 16). The frequency of occurrence between the two

groups, 100% (17/17) for the control group and 91% (41/45)

for the affected group, was not statistically different (2x2

contingency table; chi-square). Relative antibody unit-

levels ranged from 5 to 43 with a mean of 18+13 in the



























Figure 14: Determination of relative antibody levels to equine S antigen in
serum by ELISA. Titration of: 0- reference guinea pig anti-guinea pig S
antigen antiserum A reference positive equine serum 0- sample equine
serum l- negative control equine serum









































10240


1.4



1.2



1.0


10 40 160 64-0 2560
Dilution (1/n)











































Eq. serum
(patient)


Eq. serum
(expt.)


R anti-Bov. S Ag.




C


Figure 15: Evaluation of antibodies in equine serum by
immunoblotting. Twenty 4g of protein were electrophoresed
in each lane and transferred to nitrocellulose paper. The
nitrocellulose was then incubated with a) equine serum from
a patient diagnosed with uveitis b) reference equine
antiserum to equine S antigen c) reference rabbit
antiserum to bovine S antigen.




































= 13 19


x = 18 13


normal controls


patients with uveitis


(n=17) (n=45)






Figure 16: Antibody to equine S antigen measured by ELISA
in the serum of normal controls (n=17) and patients
diagnosed with uveitis (n=45).


LUI
5 -
C

LU


09
M~


771










control group and from 0 to 81 with a mean of 13+19 in the

affected group; again, these values did not represent a

statistically significant difference between the groups (t

test).

Ten of the control sera and ten of the sera from

animals with uveitis were also assayed for IgG antibody to

bovine S antigen. Again, antibody was prevalent in both

groups (Figure 17). However, the mean for sera from the

control group (24+13 relative antibody units) was actually

greater than that of samples from the diseased group (12+12)

(p < 0.025, t test). There was some positive correlation

between the level of anti-equine S antigen antibody and the

level of anti-bovine S antigen antibody in the individual

samples (r = 0.49, p < 0.05, t test)(Figure 18).

A total of 46 aqueous humor samples were assayed for

IgG antibody to equine S antigen: 23 samples from 15 normal

control animals and 23 samples from 15 animals with uveitis.

Aqueous was obtained from the inflamed eye of seven clinical

patients with unilateral disease and from both eyes of eight

patients with bilateral disease. In those animals with

bilateral disease, the condition was often more severe in

one eye. None of the samples from the normal controls had

detectable anti-S antibody (Figure 19); however 12/23 (52%)

of samples from animals with uveitis were positive which

represents a significantly greater frequency in this group

(p < 0.001, 2x2 table, chi-square). The level of antibody



































0 = 24 t 13


30
6~
0


I
normal controls
(n=10)


I
patients with uveltis
(n= 10)


Figure 17: Antibody to bovine S antigen measured by ELISA
in the serum of normal controls (n=10) and patients
diagnosed with uveitis (n=10).


x=12 -t 12







69


















* Controls (n=10)

o Patients with uveltis (n=10)


0 0 0 0


10 20 30 40 50
ANTIBODY TO EQUINE
(relative antibody


70 80 90


S ANTIGEN
units)


Figure 18: Correlation of antibody levels to equine and
bovine S antigen in equine sera.


50

40

30

20

10

01
0





















65-

641

63

5-


normal controls
(n=23)


I
patients with uveitis
(n=23)


Figure 19: Antibody to equine S antigen in aqueous humor
measured by ELISA in 23 samples from 15 normal controls and
23 samples from 15 patients with uveitis.








71

in aqueous from eyes in the affected group ranged from 0 to

64 relative antibody units with a mean of 3+13. This level

was greater than that of the control group (p<0.0011

Wilcoxon Ranked Sums test).

Quantitation of IgG in Aqueous Humor

The total IgG was measured in the 12 aqueous humor

samples in which specific anti-S antibodies were detected.

The IgG level ranged from 0.02 to 3.81 mg/ml with a mean of

1.70+1.32 mg/ml (Table 2). As expected, this was much

greater than the reported values of 0.032 and 0.050 mg/ml

for IgG in aqueous humor from normal eyes.133,134 It was

also greater than the mean levels of 0.576 and 0.362 mg/ml

reported for IgG in aqueous from horses with uveitis.

Evidence of Intraocular Synthesis of Anti-S Antibodies

In order to determine whether the anti-S antibody

present in the aqueous humor resulted from intraocular

synthesis or from non-specific leakage from the systemic

circulation secondary to increased capillary permeability,

the levels of IgG anti-S antibodies in aqueous and serum

were compared with the levels of total IgG. The ratio of

the anti-S IgG in relative antibody units to the total IgG








Table 2: Relative levels of total IgG and specific antibody to S
antigen in the serum and aqueous humor of horses with uveitis.


Animal Antibody to .S Antigen IgG Aqueous/serum
Number- Sample Relative Antibody Units -mg/ml ratio*


I Serum 1 9.4
Aqueous 1.82 0.91 18.8

2 Serum 8 5.4
Aqueous 0.25 3.41 0.05

3 Serum 81 16.7
Aqueous 0.17 0.02 1.75

4 Serum 73 12.9
Aqueous 64 2.44 4.64

6 Serum 4 13.6
Aqueous 0.28 3.24 0.25
Aqueous 0.18 2.48 0.29

7 Serum 3 9.0
Aqueous 0.09 0.03 9

10 Serum 10 9.6
Aqueous 0.08 0.80 0.10

Ii Serum 1 8.8
Aqueous 0.17 3.81 0.39

12 Serum 16 9.1
Aqueous 0.44 0.80 0.31

14 Serum 2 9.2
Aqueous 0.08 1.30 0.28

15 Serum 10 16.5
Aqueous 3.80 1.14 5.5

The ratio of the aqueous IgG anti-S antibody level to
total IgG is compared to the same ratio derived for serum.










in mg/ml in the aqueous was divided by the same ratio for

serum, i.e.; aqueous anti-S IgG
aqueous IgG
serum anti-S IgG
serum IgG

When this value exceeded 1.0, intraocular antibody synthesis

was suggested.

In five of 12 cases where anti-S antibodies were

detected in the aqueous humor, there was a higher level of

specific anti-S in the aqueous than in the serum relative to

the total IgG (Table 2). The ratio as defined above ranged

from 0.05-18.8 overall and from 1.75 to 18.8 in the five

cases with presumed intraocular antibody synthesis.

Cellular Immune Responsiveness to S Antigen

Lymphocytes from seven normal controls and seven

animals with uveitis were assayed for reactivity to equine S

antigen by lymphocyte blastogenesis. The stimulation index

ranged from 0.1 to 4.1 in the controls and from 0.4 to 2.5

in the diseased animals (Figure 20). Two animals from each

group had a stimulation index above 2.0.

DISCUSSION

Serum antibody to equine S antigen is prevalent in the

equine population, with 100% (17/17) of normal controls

having IgG to S antigen detectable by ELISA. Similarly,

serum antibody to S antigen has been identified in normal

people, with studies reporting a frequency of occurence

ranging from 5.1 to 64%.107,108,111 The presence of such


















5
a) 4
3
2
1


2t--------------------------- I-------------


4 3


normal controls
(n=7)


patients with ureltis
(n=7)


Figure 20: Transformation responses of peripheral blood
lymphocytes from normal controls (n=7) and patients with
uveitis (n=7) to a)0.l ig b) 1.0 4g c) 3.0 ig of equine S
antigen.










antibodies is not unexpected, since low levels of specific

antibody to other self-components have been previously

observed without accompanying pathology.138,139 Since S

antigen is localized in the rod outer segments, which are

rapidly turned over, sensitization of the immune system

could occur during the normal turnover process. The rod

outer segments that are shed are phagocytosed and degraded

in the retinal pigment epithelium.140,141 The close

association of this tissue with the highly permeable

choroidal capillaries may facilitate exposure of the antigen

to the immune system. Also degradation may be incomplete,

leaving antigenic determinants intact.

There are conflicting reports in the literature

concerning the significance of serum anti-S antigen anti-

body. Several investigators have reported either an

increased frequency or amount of such antibody in the serum

of patients with various syndromes of uveitis.108,109,111,-
112,119 However, in all of these studies there were also

patients with either no or low levels of anti-S antibody.

Petty proposed a genetic predisposition to the development

of the autoantibody, since among children with uveitis

accompanying juvenile rheumatoid arthritis, those who also

had the HLA-B35 phenotype had the highest frequency of

antibody to S antigen.108 Although Uusitalo and Abrahams

both found an increased level of anti-S antibody in patients

with uveitis, neither was able to document a correlation










between the severity of the disease and the titer, sug-

gesting that the antibody may not be pathogenic.109,110'111

Chan and co-workers were able to detect serum antibody to S

antigen in patients with Vogt-Koyangi-Harada syndrome,

Behcet's disease and sympathetic ophthalmia by indirect

immunoperoxidase staining, but were unable to do so by

ELISA.142 Furthermore, Doekes reported that serum antibody

was not associated with uveitis; no difference in the

frequency or level of IgG, IgM or IgA antibody to S antigen

was found between healthy controls and patients with either

anterior, intermediate, posterior, or panuveitis.I07 Thus,

it is unclear whether any correlation between disease and

serum antibody exists in human patients with uveitis.

In this study, there was no difference in either the

frequency of occurence or the level of serum IgG to equine S

antigen in horses with or without uveitis. Unless another

class of immunoglobulin is primarily involved in the

pathogenesis of the disease, it appears that in the horse

there is no association with serum antibody. Perhaps IgE

should be investigated, since Sainte-Laudy found evidence

for immediate hypersensitivity in human disease using the

human basophil degranulation test. Also, the humoral

response could still be of importance at the local level.

There was a correlation, although not exact, between

the serum levels of anti-equine and anti-bovine S antigen in

individuals. This suggests that equine and bovine S antigen








77

are antigenically related and yet have some species specific

epitopes. Thus, the results obtained using the heterologous

and homologous antigens were generally comparable. The

major difference was that the level of antibody to bovine S

antigen was significantly greater in the control group

(24+13 relative antibody units) than in the patients with

uveitis (12+12), whereas there was no difference in the

levels of antibody to equine S antigen in the two groups.

The reasons for this are obscure. However, antibody levels

to the homologous protein would clearly be expected to be

more relevant to the pathogenesis.

Both the incidence and the level of IgG to equine S

antigen were significantly greater in the aqueous humor of

horses with uveitis than in the aqueous humor of normal

controls. However, the role of this antibody in the

pathogenesis of the disease has not been determined. The

humoral response may be part of the primary pathogenic

mechanism or it may occur secondary to ocular damage from

some other cause but contribute to continued inflammation.

It may also simply be an epiphenomenon which has no patho-

logic significance. The fact that in five of twelve cases

there was evidence for local production of specific anti-S

antibodies within the eye suggests that there was an

autoimmune response to the release of antigen at that site.

The presence of antibody in the aqueous humor is likely

to represent the presence of antibody in the vitreous body










as well, based on the anatomic and physiologic features of

the eye. Except for the presence of collagen and hyaluronic

acid, the composition of the vitreous humor is similar to

that of the aqueous. Also, a small percentage of the

aqueous normally leaves by diffusion through the iris,

ciliary body, and vitreous.40 In one study investigating

experimental ocular toxoplasmosis in rabbits, antitoxo-

plasmal IgG antibodies were in fact detected in both the

aqueous and vitreous humors of infected eyes.143 Assuming

then that since anti-S antibody was detected in the aqueous,

there was also some present in the vitreous, this antibody

should have access to the presumed target tissue, since

there is no barrier between vitreous body and the retina.40

Generally equine recurrent uveitis is a disease of the

anterior segment, although extension to the posterior

segment is common. Clinically, chorioretinitis, vitreous

opacities, vitreous traction bands, retinal edema and

retinal detachments are some of the posterior lesions

observed.18 Due to the severity of the intraocular inflam-

mation, the posterior segment could not be visualized in

seven of the fifteen cases in this study in which aqueous

humor samples were available. Also, of the five cases with

evidence of intraocular synthesis of anti-S antibody, one

had chorioretinitis but the posterior segment was obscured

in the other four. Thus, although posterior involvement

cannot be ruled out, it is difficult to draw any conclusions










about the presence of intraocular anti-S antibody and

retinal damage. However, autoreactivity to S antigen could

potentially contribute to the immunopathogenesis regardless

of the distribution of the lesions. Not only do some human

patients with anterior uveitis have antibodies to S antigen

even though the antigen is localized in the retina, but in

certain laboratory animal species, experimental autoimmune

uveitis induced by S antigen primarily affects the anterior

segment.53,81,102,110 The same is true in uveitis induced

in primates using the retinal antigen interphotorecptor

retinoid-binding protein.144 The reasons for this are

obscure, although the distribution of lesions seems to be

related in part to dosage of the antigen and to species

differences in the vascular pattern of the retina. The

horse, like the guinea pig which commonly exhibits acute

anterior uveitis in response to S antigen, has a pauran-

giotic retina with small vessels extending only a short

distance from the disc.40 Therefore an autoimmune response

to S antigen might be expected to produce primarily anterior

uveitis in the horse.

Cell mediated immunity appears to be an important

mechanism in many autoimmune diseases. In experimental

allergic uveitis induced by S antigen there is considerable

data indicating that cellular immunity is involved in the

pathogenesis.91-101 As was the case for studies inves-

tigating the humoral response to S antigen in human patients








80

with uveitis, the results of studies investigating cellular

responsiveness in patients are variable. While several

investigators found heightened lymphocyte responsiveness in

patients, others were unable to detect any such reac-

tivity.I05-108,128,129 In this study, in which peripheral

blood lymphocytes from seven normal controls and seven

animals with uveitis were assayed, there was no increased

lymphocyte proliferation in response to S antigen in the

diseased animals. However, assessment of local reactivity

may be more relevant, as has been demonstrated in other

diseases and in humoral studies.130,131,145 Unfortunately,

insufficient numbers of lymphocytes were harvested from the

aqueous humor for performance of the lymphocyte prolifera-

tion assay as described.

In summary, these data indicated that in the equine

species there is no association between serum antibody or

reactivity of peripheral blood lymphocytes to equine S

antigen and uveitis. Serum IgG to S antigen was prevalent

in the general population and there was no correlation

between the level of this antibody and equine uveitis.

Also, there was no correlation between reactivity of

peripheral blood lymphocytes to equine S antigen uveitis.

There was, however, a statistically significant increase in

both the frequency and level of antibody to S antigen in the

aqueous humor of animals with uveitis as compared with

normal controls. In several cases there was evidence that








81

this antibody was locally produced, suggesting that it may

be relevant to the disease process. However, its role in

the pathogenesis, if any, has yet to be elucidated.













CHAPTER FOUR
IMMUNIZATION OF PONIES WITH EQUINE RETINAL S ANTIGEN



Introduction

S antigen induced experimental autoimmune uveitis is

the most extensively studied model of autoallergic eye

disease. Although the antigen has generally demonstrated

considerable uveitogenicity, the results of immunization are

variable depending on several factors, particularly the

species involved. A distinct difference in overall suscep-

tibility to experimental uveitis has been demonstrated

between species and even between strains.59,80,81,82,-
83,84,104 The clinicopathologic features of the disease

also vary between species. The variability may in part be

explained by differences in anatomic and physiologic

characteristics of the eye, such as the differences in the

vascular supply and the number of choroidal mast

cells.59,81,83,104 Differences in the genetic ability to

mount an autoimmune response and in the precise immunologic

mechanism evoked may also affect the response to S antigen.

While certain syndromes of S antigen induced experimental

uveitis resemble some forms of naturally occurring uveitis

of man, the variable manner in which different species

respond to S antigen makes a direct correlation between










experimental disease and the spontaneous human disease

difficult.

Considerable evidence suggests that immunologic

mechanisms are of importance in equine recurrent uveitis,

but the actual immunopathogenesis is not understood. The

development of a reproducible experimental model of autoim-

mune uveitis in the horse using S antigen would provide an

opportunity to directly compare and contrast experimental

and spontaneous disease in the same species. Hopefully,

this would lead to a better assessment of the relevance of

experimental autoimmune uveitis to clinical disease and an

improved understanding of the pathogenesis of equine

recurrent uveitis. Therefore, in this study, ponies were

immunized with equine S antigen, first at a site remote to

the eye and then intraocularly. Their clinical condition,

humoral response and cellular reactivity were monitored.

Materials and Methods

Experimental Animals

Sixteen ponies of varying sex and age were utilized.

All were normal on physical and ophthalmoscopic examination,

but no histories were available.

Sampling Techniques

Blood was drawn by jugular venipuncture, allowed to

clot and the serum was frozen in aliquots at -200C, until

assayed.








84

Aqueous humor samples were obtained under anesthesia as

described in Chapter Three.

Preparation of Equine Retinal S Antigen

S antigen was isolated from equine retinas by gel

filtration and ion exchange chromatography as described in

Chapter Two. The antigen was analyzed by PAGE throughout

the study to ensure that significant degradative changes did

not occur.

Preparation of Equine Albumin

Purified equine albumin (Sigma Chemical Co.) was

utilized as a control protein. It was handled in precisely

the same manner as S antigen; that is, it was applied to

columns of Sephacryl S-200 and DEAE Biogel A, dialyzed and

sterile filtered according to the procedure described for S

antigen in Chapter Two.

Experimental Design

Experiment one. A single pony was immunized sub-

cutaneously in the neck with 1.0 mg of equine S antigen

emulsified with complete Freund's adjuvant (H37Ra, Sigma

Chemical Co.) in a 1:1 volume ratio. Three months later the

animal was immunized in the anterior chamber of the left eye

with 0.5 mg of equine S antigen in 0.5 ml sterile saline;

the right eye received a control injection of 0.5 ml sterile

saline. This pony was followed clinically for two years.










Experiment two. Twelve ponies were randomly assigned

to three groups. The immunization schedule for all groups

was as follows:

Week 0: Subcutaneous immunization with 200ig of

protein emulsified with complete Freund's

adjuvant.

Week 3: Subcutaneous immunization with 2004g of

protein emulsified with complete Freund's

adjuvant.

Week 6: Left eye immunization in the anterior

chamber with 1004g of protein in saline.

Right eye injection of saline.

Week 9: Left eye immunization in the anterior

chamber with 1004g of protein in saline.

Right eye injection of saline.

The four ponies in group I received equine S antigen

throughout the study, while group II received equine

albumin. In Group III, two ponies received systemic albumin

followed by intraocular S antigen (IIIa) while two received

systemic S antigen followed by intraocular albumin (IIIb).

There was an additional control pony which received no

immunizations or ocular manipulations.

All ponies were examined daily. Serum was sampled

prior to immunization and at weekly intervals, and addi-

tionally when active ocular inflammation was present.










Peripheral blood lymphocytes were assayed by lymphocyte

transformation prior to immunization and bi-weekly.

Experiment Three. Two neonatal pony foals deprived of

colostrum were immunized in the anterior chamber of the left

eye with 100 4g of equine S antigen in 0.5 ml saline. In

the right eye"they received a control injection of 0.5 ml

saline.

Clinical Assessment

All ponies received a daily physical examination and an

ophthalmic examination using direct and indirect ophthal-

moscopy. Six parameters were evaluated and scored on a

scale of 0-3 according to their severity (0 = not present, 1

= mild, 2 = moderate, 3 = severe). The six parameters

monitored were pain as evidenced by blepharospasm, lacrima-

tion, conjunctivitis or chemosis, corneal abnormalities,

anterior chamber infiltrates, and pupillary changes; each

animal was then given an average score for the day. Since

in all cases posterior segment changes were either not

present or could not be evaluated due to the severity of

anterior segment disease, they were not included in the

scoring system.

Cell Counts and Differentials in Aqueous Humor

Cell counts were performed on aqueous humor samples

using a hemacytometer (Bright-Line Hemacytometer Counting

Chamber, American Optical, Buffalo, NY). If more than 2000

cells/ml were present, differential cell counts were done on










cytocentrifuge preparations (Cytospin 2, Shandon Inc.,

Pittsburgh, PA).

Detection of Antibody to Equine S Antigen by ELISA

Antibody to equine S antigen was measured in serum and

aqueous humor samples using the ELISA technique outlined in

Chapter Three. Again, relative antibody unit values for

samples were calculated by interpolating from the absorbance

of dilutions which fell in the linear portion of the

standard curve through the use of linear regressions.

Quantitation of Immunoglobulin by Single Radial Immunodif-

fusion

IgG in serum and aqueous humor was quantitated by

single radial immunodiffusion according to the method of

Mancini, Carbonara, and Heremans using antibody to equine

IgG (Miles Scientific).117

Lymphocyte Transformation for Assessment of Reactivity to

Equine S Antigen

Peripheral blood lymphocytes were assayed for reac-

tivity to S antigen by the lymphocyte transformation

procedure described in Chapter Three. As an additional

control, lymphocytes from a pony receiving no ocular

manipulations were also assayed each time.

Intradermal Skin Testing

To further assess immunologic reactivity to equine S

antigen, animals in experiment two were skin tested in-

tradermally at week ten. All ponies were injected intrader-










mally with 0.05 ml containing 50 4g equine S antigen, 54g

equine S antigen, 504g equine albumin, and 54g equine

albumin. Histamine and saline were injected as controls.

The skin tests were read at 15 minutes, 4 hours, 24 hours

and 48 hours. Biopsies of the skin test sites were taken at

4 and 48 hours.

Histopathology

Animals in experiment two were euthanized 48 hours

after intradermal skin testing. The eyes were enucleated

immediately, fixed in Zenker's solution and embedded in

paraffin. Sections were stained with hematoxylin-eosin and

periodic acid schiff. Sections of normal eyes were addi-

tionally stained with Giemsa and toluidine blue.

Results

Clinical Assessment

No ocular inflammation was detected in any ponies after

immunization with equine S antigen at a site distant to the

eye. However, ocular changes were observed after the

intraocular administration of antigens. In experiment one,

the pony developed chronic relapsing uveitis in the S

antigen injected eye while the control eye injected with

saline remained normal. At 24 hours following injection

there was mild inflammation (grade 1) which progressed to

severe inflammation (grade 3) at 72 hours post injection.

The condition was characterized by marked hypopyon and

corneal edema; there was also significant swelling of the








89

iris and the pupil was miotic. At two weeks post-injection,

the clinical signs began to gradually improve, but at three

months there was still evidence of a chronic active uveitis

with neovascularization of the cornea, keratic precipitates

and pupillary occlusion (Figure 21). Throughout the

following two years the pony had periodic recurrences of

active inflammation resulting in permanent ocular pathology.

The results of clinical evaluation of ponies in

experiment two are presented in Table 3. No ponies ex-

hibited ocular inflammation prior to the intraocular

injection of antigen and the saline injected control eyes

remained clinically normal in all cases. There was a

significant difference in the incidence of uveitis between

the groups after two intraocular injections (p < 0.05, 2x4

contingency table, x2). Specifically, those ponies injected

with equine S antigen throughout the study (Group I), had a

higher incidence (4/4) than those injected with albumin

throughout the study (Group II, 1/4)(p < 0.05, 2x2 contin-

gency table, x2). Both the two ponies which received in-

traocular S antigen after systemic albumin (Group IIIa), and

the two ponies which received intraocular albumin after

systemic albumin (Group IlIb), manifested some degree of

uveitis, but did not represent a statistically significant

difference in incidence when compared with group I or II.

The clinical signs and the course of the disease were

similar in all animals which developed uveitis, although










































Figure 21a: Acute uveitis
Clinical signs one week
administration of S antigE
lacrimation, hypopyon and cornea


in the pony experiment one.
after the intraocular
!n included blephorospasm,
Ll edema.









































Figure 21b: Chronic uveitis in the pony in experiment one.
Clinical signs six months after the intraocular
administration of S antigen included excessive lacrimation
and neovascularization in the ventral region of the cornea.
Synechia were also present but could not be visualized in
the photograph.








Table 3a: Clinical assessment of ocular inflammation
scored on a scale of 0-3 (0 = none present, 1 = mild, 2 =
moderate, 3 = severe). Ponies received intraocular
injections at weeks 6 and 9 and were skin tested on day 72.



Group I S antigen group


Week Day 1 2 3 4 mean


6 42 0 0 0 0 0

43 0 2.2 1.5 1.8 1.4

45 0 2.0 1.0 1.8 1.2

47 0 1.7 0.5 1.3 0.9

49 0 2.0 0.2 1.2 0.9

51 0 2.0 0 1.7 0.9

53 0 1.0 0 0.5 0.4

55 0 0.7 0 0.2 0.2

57 0 0.5 0 0.2 0.2

59 0 0.3 0 0.2 0.1

61 0 0.3 0 0.2 0.1

9 63 0 0.3 0 0.2 0.1

64 1.2 2.0 1.2 1.0 1.35

66 0.5 1.8 0.5 0.8 0.9

68 0.5 1.8 0.5 0.2 0.8

70 0 1.5 0 0.2 0.42

72 0 1.3 0 0.2 0.4

74 0 0.7 0 0 0.2







93
Table 3b: Clinical assessment of ocular inflammation
scored on a scale of 0-3 (0 = none present, 1 = mild, 2 =
moderate, 3 = severe).


Group II albumin group


Week Day 5 6 7 8 mean


6 42 0 0 0 0 0

43 0 0 0.8 0 0.2

45 0 0 0.3 0 0.i

47 0 0 0 0 0

49 0 0 0 0 0

51 0 0 0 0 0

53 0 0 0 0 0

55 0 0 0 0 0

57 0 0 0 0 0

59 0 0 0 0 0

61 0 0 0 0 0

9 63 0 0 0 0 0

64 0 0 0.5 0 0.1

66 0 0 0.2 0 0.1

68 0 0 0.2 0 0.1

70 0 0 0 0 0

72 0 0 0 0 0

74 0 0 0 0 0







94
Table 3c: Clinical assessment of ocular inflammation
scored on a scale of 0-3 (0 = none present, 1 = mild, 2 =
moderate, 3 = severe).


Group III -
antigen b)
albumin



Week Day


6 42

43

45

47

49

51

53

55

57

59

61

9 63

64

66

68

70

72

74


a) systemic albumin
systemic S antigen


10 mean


0

1.8

1.3

0.5

0

0

0

0

0

0

0

0

0.5

0.2

0

0

0

0


0

1.5

1.2

0.8

0.7

0.3

0.3

0.3

0.2

0

0

0

1.2

0.8

0.2

0

0

0


0

1.7

1.3

0.7

0.4

0.2

0.2

0.2

0.1

0

0

0

0.9

0.2

0.2

0

0

0


followed by intraocular S
followed by intraocular


b)

11 12 mean


0

0.8

0.6

0

0

0

0

0

0

0

0

0

0.3

0

0

0

0

0


0

0.3

0.2

0

0

0

0

0

0

0

0

0

0.2

0

0

0

0

0


0

0.6

0.4

0

0

0

0

0

0

0

0

0

0.3

0

0

0

0

0










there was a difference in severity. After the intraocular

injection of antigen, the onset of inflammation was rapid in

all cases, with clinical signs occurring within 24 hours.

The disease was primarily characterized by hypopyon and

corneal edema which was most severe in the early stages

(Figure 22). As the condition improved, a more localized

area of corneal edema developed.

The initial onset of inflammation after the intraocular

injection of antigen was rapid, with clinical signs occur-

ring within 24 hours in all cases. Also, the clinical signs

were most severe initially, with gradual improvement

thereafter.

There was a difference in the severity of the uveitis

between the groups. Following the first intraocular

injection, ponies in group I scored from 0 to 2.2 when the

clinical signs were most severe, with an average score for

the group of 1.4. Ponies in group II scored from 0 to 0.8,

with an average of 0.2; those in group IIIa scored from 1.5

to 1.8 with an average of 1.7 and those in group IlIb scored

from 0.3 to 0.8 with an average of 0.6. When these averages

were compared, the disease in groups I and IIIa, which

included animals receiving intraocular S antigen, was

significantly worse than that in group II (p < 0.05, Duncans

Multiple Range Comparison); although these groups also

appeared clinically worse that group IIIb, the difference

was not statistically significant. There was clearly no