The spontaneously diabetic BB rat


Material Information

The spontaneously diabetic BB rat a model of autoimmunity and immunodeficiency
Physical Description:
ix, 183 leaves : ill. ; 29 cm.
Elder, Melissa Ellen, 1955-
Publication Date:


Subjects / Keywords:
Autoimmune Diseases   ( mesh )
Rats, Inbred Strains   ( mesh )
Pathology thesis Ph.D   ( mesh )
Dissertations, Academic -- Pathology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida.
Bibliography: leaves 168-182.
Statement of Responsibility:
by Melissa Ellen Elder.
General Note:
Photocopy of typescript.
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028968634
oclc - 08643834
System ID:

This item is only available as the following downloads:

Full Text







Dedicated to the memory of

Mr. Mellow


First and foremost, I wish to thank my parents,

Wayne and Nora Elder, for their lifelong love and support.

I would like to also especially acknowledge my sister,

Melanie Elder, for her love, understanding, and friendship.

I wish to likewise thank Phillip Ruiz for his affection and

help during my graduate education.

I wish to express my deep appreciation to Dr. Noel

Maclaren for his expert guidance and assistance during the

course of this research. The members of my committee,

Dr. Juan Scornik, Dr. Ammon Peck, Dr. Edward Wakeland,

Dr. Paul Klein and Dr. Christopher West, are also thanked

for their interaction, assistance and interest in this

project. Other members of the Pathology faculty to whom I

express special thanks for their help are Dr. William Riley,

Dr. Raul Braylan and Dr. Arthur Kimura. I am grateful to

the Department of Pathology for the financial support I

received during my graduate career.

I would like to express my gratitude to Lee Glancey,

Tom McConnell and Edith Rosenbloom for their invaluable,

technical assistance and support in the completion of these

studies. Their excellent work aided in the development and

completion of this project. My thanks also go to everyone

else associated with the Lab. The help of Crystal Grimes


and Flo Jordan in the preparation of this manuscript and

other papers is deeply appreciated. I would also like to

thank the graduate students of the Department of Pathology

for their encouragement and friendship.

Finally, these acknowledgements would not be complete

without mentioning some of the special people whose friend-

ship and support have been important in the completion of

this endeavor: Joan Appleyard, Steve Noga, Dan Cook,

and Art Alamo.


ACKNOWLEDGEMENTS ............... .................... iii

COMMONLY USED ABBREVIATIONS.......................... vi

ABSTRACT..... ............................ .............. viii

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

Insulin-dependent Diabetes (IDD)................ 2
The BB Rat........................................ 15

SPECIFIC AIMS................ ............................. 18

MATERIALS AND METHODS............................... 21

RESULTS.......... .. ...... .... .. ........... ........ 32

DISCUSSION............... .... ... ...... .............. 149

The BB rat as a model of IDD and organ-
specific autoimmunity.......................... 149
The BB rat as a model of
immunodeficiency............................. ... 156

REFERENCES................... ........... .......... .. 168

BIOGRAPHICAL SKETCH................................. 183

Commonly Used Abbreviations

BB x Lewis

Con A

BB rats obtained from Ottawa, Ontario
BB rats obtained from Worcester, Mass.
Cross of male BB rat with female WF rat
Cross of male BB rat with female Lewis rat

Degrees Centigrade
Concanavalin A
Counts per minute

HLA D region-associated antigen



IL 2







First filial generation
Second filial generation
Fetal calf serum

Major histocompatibility complex of man

Immune response-associated antigen
Islet cell antibodies (cytoplasmic)
Islet cell surface antibodies
Insulin-dependent diabetes
Immunoglobulin Class G
Immunoglobulin Class M
Interleukin 2 or T cell growth factor

Mixed lymphocyte culture
Monoclonal antibody which defines rat Ia-
positive cells
Monoclonal antibody which defines rat
cytotoxic/suppressor T cell subset

Not done

Phosphate-buffered saline
Gastric parietal cell autoantibodies
Purified phytohemagglutinin
Pokeweed motigen

Major histocompatibility complex of the rat

Standard deviation
Smooth muscle antibodies

WF Wistar Furth
W3/13 Monoclonal antibody which defines rat T
W3/25 Monoclonal antibody which defines rat
helper T cell subset

x Irradiated


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



Melissa Ellen Elder

August, 1982

Chairman: Noel K. Maclaren, M.D.
Major Department: Pathology

The BB rat is presently the best available animal

model for human insulin-dependent diabetes (IDD). IDD in

the BB rat may result from autoimmunity since it is accom-

panied by lymphocytic inflammation of the pancreatic islets

and is preventable by immunosuppression. Antibodies to pan-

creatic islets (ICA) and other endocrine tissues in BB rats

were sought as evidence for an autoimmune etiology for IDD.

ICA were not detected, while smooth muscle and gastric par-

ietal cell autoantibodies (PCA) were frequently identified.

Although functional abnormalities of gastric parietal cells

were not noted, PCA-positive rats had evidence of lympho-

cytic gastritis. PCA appeared during the age period that


the animals were developing IDD. Thus, the BB rat has an

autoimmune diathesis of which IDD may be only one result.

Immunocompetency of the BB rat was also studied.

Severe T lymphopenia was observed in all BB rats, irrespec-

tive of age or presence of IDD, while numbers of B cells

and immunoglobulin levels were normal. Both the numbers of

helper T cells and cytotoxic/suppressor T lymphocytes were

depressed, and an inversion of the ratio of helper T cells

to cytotoxic/suppressor T cells occurred in all BB rats with


Concomitantly, profound impairments of T cell-mediated

immunity were noted to mitogenic stimulation and in allo-

responses. BB lymphocytes produced IL 2 normally; however,

irradiated WF cells and Con A supernatants did not restore

BB responses, suggesting that BB lymphocytes may have

defective responses to helper factors such as IL 2. In

contrast to BB peripheral T cells, BB and WF thymocytes re-

sponded equally well to mitogens. Whereas BB thymic histol-

ogy was normal, BB spleens and lymph nodes were severely

depleted of T lymphocytes. Thymocytotoxic autoantibodies

were also detected in many BB rats. These findings suggest

that the defect in T cell immunoresponsiveness may be post-

thymic or peripherally acquired. Although the BB rat is an

intriguing model of autoimmunity and immunodeficiency, no

clear relationship between the immunoincompetence and IDD

is known to date.


Autoimmune diseases result from loss of self toler-

ance which leads to an immune response by the individual to

autologous antigens and subsequent cellular and tissue

destruction or other effects. Such diseases are separable

into two groups, i.e. organ specific disorders such as the

autoimmune endocrinopathies, and autoimmune diseases that

are systemic and not confined to any one organ such as

systemic lupus erythematosus.

A disease is generally considered to be an organ-

specific autoimmune disorder if there are mononuclear in-

filtrations of the affected organ or tissue, organ-specific

autoantibodies, and a tendency for more than one of these

diseases to occur simultaneously in individual patients (1).

Such organ-specific autoimmune diseases include chronic

lymphocytic thyroiditis, Graves' disease, Addison's disease,

acquired hypoparathyroidism and insulin-dependent diabetes


Most autoimmune endocrinopathies are associated with

disturbed frequencies of certain HLA antigens, especially

an increased frequency of the immune response gene HLA DR3

(1,2). HLA DR4 is also increased in IDD (1,2). For some

diseases of this group, such as chronic lymphocytic thyroid-

itis, however, no definitive HLA associations have been


The presence of organ-specific autoantibodies in the

patient's serum does not mean the patient has clinical dis-

ease nor that the antibodies are actually the cause of tis-

sue damage. However, because the autoantibodies are very

specific and are found in low frequencies in the general

population, they are considered to be markers or indicators

of the presence of an autoimmune disease process occurring

in the patient, regardless of whether this process leads to

overt disease (3,4).

Proof of a primary role for autoimmunity in the patho-

geneses of these putative autoimmune endocrinopathies re-

mains difficult to obtain. Current etiologic concepts for

such autoimmune diseases most probably must include discus-

sion of the relationship between environmental triggers

and immune responsiveness (as defined by HLA DR antigens)

in the genetically predisposed individual.

Insulin-Dependent Diabetes (IDD)

It has been suggested that human IDD may result from

the autoimmune destruction of the insulin-secreting beta

cells of the pancreatic islets. Evidence to support an

autoimmune etiology for IDD has included observations of

mononuclear infiltrations in pancreatic islets of patients

who have died suddenly after onset of IDD (5,6). These

infiltrative lesions were predominantly composed of

lymphocytes, with few polymorphonuclear leukocytes, eosino-

phils or plasma cells present and are referred to as

"insulitis" (7,8). The frequency of insulitis in patients

with IDD is variable, with reports ranging from extremely

rare (9,10) to greater than 50% (8,11) in patients with re-

cent onset of IDD. Lymphocytic infiltrations of pancreatic

islets have not been demonstrated in patients with IDD of

greater than one year duration, in noninsulin-dependent

diabetics or in normal controls (12).

Islet Cell Autoantibodies in IDD. The presence of anti-

bodies to pancreatic islet cells in sera of patients with

IDD and polyendocrinopathies has been extensively described

(13,14). Cytoplasmic-reactive islet cell autoantibodies

(ICA), detectable by indirect immunofluorescence on normal

pancreatic tissue, react with all cell types of the pancre-

atic islets in addition to beta cells (15-17), which is in

contrast to the specific loss of beta cells seen in IDD.

ICA do not cross react with gastrointestinal cells secreting

hormones also found in pancreatic islets, such as glucagon

and somatostatin (2). These antibodies are exclusively of

the IgG class, usually fix complement and are believed to

react with a microsomal membrane lipoprotein found normally

in all islet cells (13,14,18-20). The frequency of ICA in

patients with IDD varies according to the patient's race, to

the time elapsed after clinical diagnosis of IDD, and has

been found to precede development of the clinical disease

(13,15,21,22). Neufeld et al. determined the frequency of

ICA to be approximately 74% in Caucasian children tested

within three months of onset of IDD, whereas ICA were

detected in less than 50% of patients three years after

diagnosis (15). However, 10-15% of IDD patients have been

noted to have persistent ICA for many years and have been

shown to have higher frequencies of associated thyroid,

gastric and adrenal autoimmune diseases as well as higher

frequencies of HLA B8-bearing haplotypes than patients with

IDD who become ICA-negative over similar periods (19,23).

In contrast, black insulin-dependent diabetics have only

about half the frequency of ICA in relation to duration of

IDD as do Caucasian patients with IDD, suggesting that much

of the IDD among black patients may be different from IDD

seen in Caucasian populations (15,24). In addition,

children who are ICA-negative at onset of IDD rarely become

positive for these antibodies later (2,3). ICA have also

been demonstrated in approximately 3-5% of nondiabetic

relatives of IDD probands (15,23) and in about 0.5% of

healthy controls (15,23).

Observations that ICA-positive family members of

patients with IDD and noninsulin-dependent diabetics with

ICA tend to become insulin-requiring with time (22,25,26)

suggest that these antibodies may be of clinical value in

predicting the subsequent development of IDD (22). However,

a causal relationship between ICA and IDD remains to be

proven. A direct role for ICA in the pathogenesis of IDD in

fact seems unlikely, especially since these antibodies react

with a shared antigen found in the cytoplasm of all islet

cell types. Furthermore, autoantibody molecules are not

normally considered to be able to cross membranes of living

cells to effect damage. In addition, experimental evidence

to suggest that ICA do not cause IDD includes findings that

transfers of ICA from human patients with IDD to mice have

not resulted in IDD in these animals (19,27), nor has placen-

tal passage of ICA resulted in the development of IDD in

newborn infants or even to affect neonatal insulin secretion

(19). Observations that ICA titers decrease with time after

clinical onset of IDD is probably related to the progressive

loss of the relevant antigen, whatever the disease


Islet Cell Surface Autoantibodies in IDD. Autoantibodies

reactive to antigens on pancreatic islet cell membranes

(ICSA) have also been demonstrated in sera of patients with

IDD (21,28-30). Using cell surface immunofluorescence tech-

niques or 125I-labeled protein A assays, ICSA have been

detected either by reaction with cultured human insulinoma

cells (29), beta cells isolated from dispersed rat or mouse

pancreatic islets (21,31,32) or human fetal pancreatic islet

cell cultures (33). Analagous to ICA, ICSA have been de-

tected in about 67% of children at onset of IDD and decrease

in frequency within the first year after diagnosis (34,35).

In contrast, ICSA are reported to be more common than ICA in

controls and seem to possibly occur independently of ICA in

patients with IDD (36).

ICSA in sera from patients with IDD have been shown to

be cytotoxic to mouse pancreatic islets in vitro (37), how-

ever some nondiabetic sera without ICSA were also

cytotoxic to these cells. Sera from IDD patients have also

been shown to have complement-mediated cytotoxic effects on

hamster islets (38) and rat islets (39), correlating with

the detection of ICSA in the serum. In addition, ICSA-

positive sera in the presence of complement have been dem-

onstrated to cause increased chromium release from labeled

rat islet cells (40,41) while sera containing ICA alone did

not have such effects (40). Difficulties with this study

include defects in the method of ICA detection, and findings

that 25% of ICSA-positive sera from nondiabetic first degree

relatives were also cytotoxic to islet cells.

There has been little convincing evidence to suggest

that ICSA react specifically with pancreatic beta cells.

Indeed in one study, cytotoxicity of sera with detectable

ICSA was shown to not be restricted to pancreatic beta

cells, but also affected a somatostatin-producing tumor line

(42). However, Dobersen and Scharff has recently demon-

strated the preferential lysis of rat beta cells with

minimal killing of other types of islet cells by ICSA-

containing diabetic sera using double-label immunofluores-

cence techniques (28). In addition, eleven of twenty-one

sera from diabetic patients with ICSA were shown to be able

to suppress glucose and theophylline-stimulated insulin

release but not glucagon release in vitro by dispersed mouse

islets (43). Paradoxically, ICSA have been demonstrated to

have a stimulatory effect on basal insulin release from

cultured mouse islets (44). These results are consistent

with the primary role for ICSA in beta cell destruction seen

in IDD proposed by some investigators (28), but much more

substantial evidence is needed.

Cell-Mediated Autoimmunity in IDD. No consistent general-

ized defects in cell-mediated immunity as defined in terms

of lymphocyte responsiveness to phytohemagglutinin (PHA) (45-

48) and in the enumeration of lymphocyte subpopulations

(45,49-51) appear to be present in well-treated patients

with IDD. However, poorly controlled diabetics have been

observed to have depressed mitogenic responses to PHA in

comparison with matched adequately treated IDD patients and

healthy controls (47,52) suggesting that the metabolic

derangements of IDD have an adverse effect on T lymphocyte

responsiveness. Other studies have indicated that specific

antipancreatic cell-mediated immunity may be observed in

patients with IDD (45,47). Nerup et al. demonstrated

significant inhibition of migration of leukocytes from

insulin-dependent diabetics in the presence of porcine

pancreas (53) or fetal calf pancreas (54,55) homogenates.

Abnormal migration inhibition was especially noted in those

patients with IDD for less than one year, but was also

demonstrated in some noninsulin-dependent diabetics.

Positive leukocyte migration inhibition in IDD patients has

also been observed using human pancreas homogenates (56) or

insulinoma extracts (57) as antigens. Concomittantly, de-

layed hypersensitivity skin reactions to porcine pancreatic

suspensions were seen by Nerup in patients with IDD who

exhibited inhibition of leukocyte migration using the same

antigen (53). In order to examine the possible role of cell-

mediated immunity to insulin and thus to beta cells in the

pathogenesis of IDD, MacCuish (47) and others (58) were able

to demonstrate significantly greater blastogenesis, as

measured by 3H-thymidine incorporation, by lymphocytes

from diabetic patients when cultured with bovine or porcine

insulin than by lymphocytes from control patients.

Huang and Maclaren were able to demonstrate specifi-

cally enhanced cytoadherence and cytotoxicity of human in-

sulinoma cells in vitro by peripheral blood lymphocytes from

children with IDD (59). Specific insulinoma-tumor cell

cytotoxicity, as measured by eosin exclusion, was seen both

with and without added patient sera, but was not obtained

using either lymphocytes from patients with systemic lupus

erythematosus or autoimmune thyroid disease, or with differ-

ent tumor lines as targets (59,60). Peripheral blood lymph-

ocytes from twenty-one out of thirty-three patients with IDD

were shown to inhibit insulin release by rat islets to glu-

cose and theophyllin (61), while no inhibition of insulin

secretion was noted by lymphocytes from noninsulin-depend-

ent diabetics or controls. Significantly increased levels

of circulating killer (K) cells, classified as low affinity

E-rosette forming cells, were also found in 57% of newly

diagnosed insulin-dependent diabetics (62). These levels

returned to normal within twelve months from diagnosis of

IDD. Raised K cell numbers were accompanied by signifi-

cantly enhanced levels of antibody-dependent cell-mediated

cytotoxicity activity to chromium-labeled human erythro-

cytes sensitized with antierythrocyte antibodies in many

IDD patients (63).

Several investigators have suggested that defects in

suppressor T cell activity were present in patients with IDD

(64,65). In one study, Concanavalin A (Con A)-activated

lymphocytes from patients with IDD poorly suppressed allo-

geneic mixed lymphocyte cultures when compared to suppressor

activity demonstrated by Con A activated lymphocytes from

controls (64). In contrast, Slater et al. recently found a

statistically significant increase in Con A activated

suppressor T cell activity in thirteen patients with IDD

(66). These measurements of cell-mediated immunity,

however, are difficult to analyze even in healthy people,

and biologically significant results with diabetic patients

are even more questionable because the effects and

complications of IDD itself affect lymphocyte function and


In a provocative study, indium-labelled autologous

peripheral blood leukocytes from two of three newly diag-

nosed patients wtih IDD were observed by CAT scanning to

become distributed in the same patient's pancreas after

intravenous reinjection (67). No pancreatic localization

using the same procedure was noted in scans of patients with

other diseases. Whether the lymphocytes specifically homed

to the pancreas or were trapped there nonspecifically as a

consequence of inflammation, these results suggest the

presence of pancreatic insulitis in these patients.

Associated Autoimmune Diseases. Other organ-specific

autoimmunities are found with increased frequencies in IDD

(1,19,68,69). These diseases mainly involve the thyroid,

the adrenal and the parietal cells of the gastric mucosa.

In studies by Riley and colleagues of more than 500 chil-

dren with IDD, 17% of the patients had detectable thyroid

microsomal antibodies and 5% had overt thyroid disease in

comparison with less than 2% of the control population even

having thyroid antibodies (15). Adrenal autoantibodies and

Addison's disease were also more frequent in Caucasian chil-

dren with IDD, being found in 2% and 0.5% of insulin-depen-

dent diabetics respectively, compared to adrenal antibodies

being detected in 0.7% of matched controls (15,69). Final-

ly, gastric parietal cell autoantibodies were detected in

approximately 9% of patients with IDD and in only 1% of

matched controls (19). Children with other autoimmune

endocrinopathies in addition to IDD have even higher fre-

quencies of associated organ-specific autoantibodies (19).

For example, patients with both IDD and thyroid microsomal

antibodies have augmented incidences of adrenal antibodies

to about 6% (19). Organ-specific autoantibodies and auto-

immune disease are not found in increased frequencies in

noninsulin-dependent diabetics (19).

HLA Associations with IDD. Singal and Blajchman first

reported that IDD was associated with disturbed frequencies

of HLA antigens and noted an increase in HLA B15 in these

patients (70). Nerup and coworkers later documented

statistically significant increases in HLA B8 and HLA B15 in

patients with IDD in comparison with matched controls, while

no HLA differences were seen between controls and noninsulin

-dependent diabetics (71). Subsequent studies have both

confirmed and extended these observations (72-74). Cudworth

(75) and others (76) have postulated that there are two HLA

haplotypes associated with increased relative risks for

IDD: HLA Al (A30) B8 (B18) Cw3 DR3 and HLA A2 B15 (B40)

DR4. The primary association of IDD seems to be with the

HLA DR antigens and secondarily due to linkage disequilib-

rium with the HLA A and B antigens (2,72,73,77). HLA DR3

is found in 36-59% of IDD patients compared to between 11-

24% in normal controls (72), while HLA DR4 is seen in 32-

58% of insulin-dependent diabetics compared to 16-28% in the

general population (72). Patients with multiple autoimmune

endocrinopathies have an even higher frequency of HLA Al B8

DR3 haplotypes and especially HLA DR3, suggesting that a

gene predisposing for general organ-specific autoimmunity is

associated with HLA DR3 (2,45). The risk of developing IDD

is highest for HLA DR3/DR4 heterozygotes, implying the

possibility of the existence of at least two hereditary

susceptibility genes for IDD associated wtih HLA genes--one

associated with HLA DR3 and the other with HLA DR4 (34,45,

77,78). One haplotype has been suggested to possibly render

protection against IDD because it is decreased and virtually

absent in Caucasian patients with IDD: HLA A3 (All) B7 DR2

(2,75,77,78), although the low frequencies of HLA DR2-bear-

ing haplotypes in IDD may be due in part to the increased

frequencies of HLA DR3 and HLA DR4 (79). Black patients

with IDD as a whole do not have as significantly disturbed

frequencies of HLA DR3 and DR4 antigens as Caucasian child-

ren with IDD (24). However, those black diabetic children

who are positive for ICA seem to invariably type HLA DR3 or

HLA DR4 (24), which is probably due to the impact of Cauca-

sian IDD genes in the black genome by racial admixture (24).

Families with multiple members affected by IDD tend to show

an excess of individuals typing for HLA DR3 or HLA DR4. In

addition, several investigators (80-82) have observed

significantly disturbed frequencies of several complement

factor B alleles in insulin-dependent diabetics (2). The

mechanisms through which HLA antigens, especially the immune

response associated HLA DR antigens, affect susceptibility
to IDD remain to be elucidated.

Mechanism of Inheritance of IDD. The apparent linkage

between IDD and HLA has made it possible to analyze the

segregation of HLA haplotypes as markers for IDD in multi-

plex families with two or more siblings with IDD. Several

investigators have suggested IDD to be a recessive disease

(83,84), while others believe IDD to be dominant. In either

case, reduced penetrance would need to be invoked to explain

the segregation of IDD with HLA haplotypes (85,86). Indeed,

the penetrance or the percentage of people carrying the IDD

susceptibility gene(s) that actually have clinical IDD seems

to be quite low, approximating 15-30% in multiplex families

(86). Speilman has recently suggested a hypothesis of dif-

ferential susceptibility to IDD depending on dosage of IDD

susceptibility gene(s), rather than simple dominant or re-

cessive inheritance (87). Heterozygosity of IDD alleles

would result in significant susceptibility to IDD, but homo-

zygosity for the gene(s) would be associated with even

greater risk (penetrance) for the disease. However, because

of uncertainty as to the random frequency of IDD genes in

the general population, the crossover rates between HLA and

IDD gene loci (if not one and the same), and the probable

genetic heterogeneity of the disease, estimations of the

mode of transmission of IDD are difficult to make (19).

Environmental Factors in the Pathogenesis of IDD. As the

concordance for IDD in monozygotic twins has been shown

by Leslie and Pyke to be at most 50% (88), environmental

influences have been implicated as at least secondary

factors in the development of IDD (86). Support from animal

models for a viral role in the pathogenesis of IDD includes

evidence that encephalomyocarditis virus causes a diabetes-

like syndrome in susceptible SJL mice, associated with beta

cell loss and pancreatic insulitis (89).

There is limited direct evidence for such viral partic-

ipation in human IDD. Coxsackie B4 virus was isolated from

the pancreas of a child with fatal diabetic ketoacidosis

which proved capable of inducing insulitis and hyperglycemia

in some strains of mice (90). An epidemiological relation-

ship between annual cycles of infection with Coxsackie B4

virus in people and the seasonal incidence of IDD has also

been found (91), and a high frequency of IDD in children who

have suffered severe congenital rubella or mumps infection

has been reported (45,92). One study has suggested that the

frequency of IDD in patients after infections with these

viruses is positively associated with HLA B8 and thus by

linkage disequilibrium with HLA DR3 (93). If confirmed,

such findings would suggest a relationship between the

immune response of an individual to a virus and the

development of IDD. Several investigators suggest that

these viruses do not usually cause IDD, but instead may

trigger IDD by a stress effect in the susceptible predia-

betic individual with preexisting insulinopenia. To this

effect, it is notable that no epidemics of IDD have been

reported in several large registry studies in the United

States, London and Denmark.

The BB Rat

The spontaneously diabetic BB rat was first recognized

in 1974 in an outbred colony of Wistar rats at the BioBreed-

ing Laboratories (94). These rats have subsequently been

bred for the IDD phenotype. BB animals which have been

formally inbred for seven to twelve generations develop

spontaneous severe IDD at about 70-120 days of age, which is

characterized by insulinopenia, marked hyperglycemia,

ketoacidosis, weight loss and an absolute requirement for

exogeneous insulin (95). The BB rat most closely resembles

human IDD of all animal models known to date. IDD develops

in genetically susceptible male and female rats with equal

frequencies (95,96) and is thought to be inherited either as

a single autosomal recessive gene with reduced penetrance

(94) or as multiple genes.

Moderate insulitis is seen in the pancreatic islets of

BB rats at the time of diagnosis of IDD, resembling the pan-

creatic lesions seen in recently diagnosed human patients

with IDD (96,97). Insulitis has also been observed in non-

diabetic animals (94). Immunochemical staining of pancre-

atic islets from diabetic BB rats for insulin content re-

veals depletions of beta cells, especially marked in pan-

creases from BB rats with longstanding IDD (98,99). Such

pancreatic islets are comprised almost exclusively of glu-

cagon, somatostatin and pancreatic polypeptide-secreting

cells (94).

The IDD seen in the BB rats is probably not due to

recognized infectious agents, since animals raised in a

gnotobiotic environment, such that they did not develop

antibodies to any bacteria, viruses or parasites, did not

have decreased incidences of IDD (100). However, this study

was based on results from only one litter and, in any event,

does not rule out the possibility of a role for a vertically

transmitted virus in the etiology of IDD in these rats.

The typical insulitis lesions and the genetic predispo-

sition for IDD suggest that IDD in the BB rat may be the

result of beta cell autoimmunity. Evidence to support such

a hypothesis includes recent findings of ICSA in twelve of

fourteen diabetic BB rats using a 125-labelled protein A

assay (101). Other data include observations of the

reduction in frequency of IDD in susceptible BB rats

after administration of antilymphocyte serum (102),

neonatal thymectomy (103,104) or bone marrow reconstitu-

tion (105). However, the experimental designs were ques-

tionable since the studies were performed between groups of

litters rather than within litters. Because the incidence

of IDD varies from 0-60% in any one litter to the next, the

observed decrease in frequency of IDD, but not total IDD

prevention, may instead be due to litter assignment rather

than treatment. In addition, no research group has yet been

able to successfully transfer IDD immunologically from BB

rats to nondiabetic recipients.

Less direct evidence for an involvement of autoimmunity

in the pathogenesis of IDD in the BB rat includes recent

findings by Colle and coworkers of a genetic linkage between

IDD and the rat major histocompatibility complex RT.1 in F2

animals produced by initial matings of male BB rats with IDD

and RT.1 incompatible female Lewis rats (106). BB rats have

also been shown to have decreased numbers of circulating T

lymphocytes (107) and to be extremely susceptible to

opportunistic infections. As is the case with human IDD,

more evidence is needed in order to prove an autoimmune

etiology for the loss of pancreatic beta cells and thus IDD

in the BB rat.


From the previously mentioned findings, IDD in the BB

rat and in humans is thought to result from pancreatic beta

cell autoimmunity. However, much more evidence is required

to substantiate this hypothesis. In order to better under-

stand the etiology and genetics of IDD in the BB rat and its

similarities to human IDD, the following studies were per-


1. Identification of organ-specific autoantibodies in

BB rat sera. Several of such antibodies occur with

increased frequencies in human IDD and their pres-

ence in BB rats would support a role for autoimmu-

nity in the strain.

a. Autoantibodies to the cells of the pancreatic

islets, the thyroid, the adrenal gland, and the

gastric mucosa were sought in BB rats from an

early age before the development of IDD in or-

der to determine whether a correlation between

IDD onset and the appearance of autoantibodies


b. Clinical evidence of disease was determined in

those rats with organ-specific autoantibodies.

c. Crosses between BB and WF rats were performed

in order to indicate the inheritance of auto-

antibodies and their relationship to IDD.

2. Evaluation of the immune system of the BB rat.

These studies were done in the hopes of identifying

abnormalities which could both predispose to auto-

immune disease and explain the increased suscepti-

bility of these animals to infection. Both

diabetic and nondiabetic BB rats were studied.

a. The various circulating leukocyte populations

and major lymphocyte subsets were counted in BB

rats and compared to numbers observed in WF and

BB x WF Fl animals.

b. The ability of the immune system of BB rats to

function in vivo was determined by observing

how well these rats could reject skin grafts

involving both major and minor histocompati-

bility differences.

c. The ability of BB lymphocytes to function in

vitro was tested by the responses of lympho-

cytes to mitogens or to allogeneic cells in

mixed leukocyte cultures (MLCs).

d. The possible presence of increased suppressor

activity in BB rats and the ability of BB

lymphocytes to produce and respond to helper

factors such as interleukin 2 (IL 2) were


e. The proliferative responses of BB and WF thymo-

cytes to mitogens were compared.

f. Autoantibodies to thymocytes were sought in

sera from BB rats.

g. Histological examination of thymuses, spleens,

and lymph nodes from BB rats were made.

h. Gamma globulin levels were measured in BB rats

as a gross indication of B lymphocyte function.

3. Attempts to transfer IDD with pancreas extracts and

cells from spleens and lymph nodes from BB rats to

nondiabetic BB, WF and BB x WF Fl rats were

performed. Positive findings would thus provide

convincing proof of an autoimmune etiology for IDD

in the BB rat.

4. Evaluation of the genetics of IDD and autoanti-

bodies in the BB rat.

a. Different mating combinations were performed

between BB rats with and without IDD and/or

autoantibodies in order to observe the modes of

inheritance of these autoimmune parameters.

b. Crosses between male BB rats with IDD and both

female WF and Lewis rats were performed in

order to observe the relationships between IDD,

autoantibodies, and the rat major histocompati-

bility complex, RT.1.


Animals. One hundred and fifty BB rats of both sexes and

of varying ages in addition to Wistar Furth (WF) rats

(Charles River Laboratories, Wilmington, MA) and Lewis rats

(Charles River) were used in these studies. Initially, 40

BB rats from 15 litters were obtained from Dr. Pierre

Thibert (Animal Resources Division, Health Protection

Branch, Ottawa, Ontario) and 28 BB rats from 3 litters were

obtained from Dr. Arthur Like (University of Massachusetts,

Worcester, MA). The animals obtained from Dr. Thibert and

Dr. Like are referred to as BB/O and BB/W rats

respectively. The remaining 82 BB rats were bred in our

laboratory from the original animals obtained above. Male

BB rats with IDD were also mated with inbred WF or Lewis

females. Male and female Fl progeny were then interbred to

produce F2 rats.

The animals were given Purina Rodent Chow 5001 and

water ad libitum, with light regulation at 0730 and 1830

hours. Diabetic rats were maintained on PZI insulin given

between 1500 and 1700 hours daily by subcutaneous injections

into axillary skin folds at doses sufficient to sustain

weights, minimize polyuria and avoid clinical hypoglycemic


Detection of IDD and Sera Collection. All BB rats had

blood and urinary glucose levels determined weekly. Blood

samples were drawn from the periorbital venous sinus using

heparinized capillary tubes while the rats were under light

ether anesthesia. After clotting, serum glucose levels were

run in duplicate on a Beckman II glucose analyzer (Beckman

Instruments, Inc., Fullerton, California). Sera were

collected weekly and then stored at -200C until testing for

autoantibodies were made. Rats were considered to have IDD

if the animals had glycosuria or serum glucose levels above

250 mg/dl. For some studies, intraperitoneal glucose toler-

ance tests (1.75 g/kg) were performed after an overnight

fast on BB rats aged 6-9 months who had not developed overt


Handling and Preservation of Tissues for Autoantibody

Normal WF rat tissues were used for detection of organ-

specific autoantibodies. After removal, the tissues (thy-

roid, pancreas, adrenal and stomach) were cubed and immedi-

ately snap frozen in isopentane cooled in a mixture of dry

ice and acetone and then stored at -800C. Four micrometer

tissue sections were cut out on a SLEE HR Mark II cryostat

(Slee Medical Equipment Ltd., London, England) at -200C,

placed on slides and air-dried. The slides were either used

immediately or stored for no more than one month at -800C.

Tissue Histology. Thymuses, spleens, lymph nodes,

stomachs and pancreases were removed from BB, WF and BB x WF

Fl rats and placed in a 10% formalin solution. The final

paraffin-embedded tissue sections were stained with hematox-

ylin and eosin before histological examination.

Isolation of Peripheral Blood Mononuclear Cells. One to 3

ml blood samples were withdrawn from the periorbital venous

sinus or tail vein into 3 ml EDTA vacutainer tubes (Becton-

Dickinson, Rutherford, NJ) while the rats were under ether

anesthesia. The blood was diluted with phosphate-buffered

saline (PBS), layered over a Ficoll-Hypaque gradient (24:10

v/v; 9% Ficoll 400, Pharmacia, Piscataway, NJ: 34% Hypaque

M, Winthrop Laboratories, New York, NY), and the mononuclear

cells (PBL) isolated at the interface after centrifugation

at 2200 rpm for 13-15 minutes. Any remaining erythrocytes

were lysed by exposure to an ammonium chloride-Tris buffer

solution for 3 minutes at 370C. The PBL were then washed

twice in PBS before adjustment to the desired cell concen-

trations in either PBS for T cell subset determinations or

in complete medium: RPMI 1640 (GIBCO, Grand Island, NY)

supplemented with 50 pg/ml gentamycin (Schering, Kenilworth,

NJ), 5% heat-inactivated fetal calf serum (FCS) (GIBCO), and

5 x 10-5M 2-mercaptoethanol (2-ME) (Bio-Rad, Richmond, CA)

for mitogen and MLC assays. Viability of cells was measured

by trypan blue exclusion.

Isolation of Splenic Lymphocytes. The spleens were

removed asceptically from anesthesized rats, minced, pressed

through nylon mesh screens and resuspended in cold PBS. Any

cell clumps were dispersed by repeatedly aspirating the

suspensions through a series of varying gauge (19,21,23,25)

needles, after which the cell suspensions were centrifuged

at 1600 rpm for 10 minutes at 200C. Erythrocytes were

removed as stated previously. After washing the pellets

twice in PBS, the cells were resuspended in either PBS for

transfer experiments or in complete medium for in vitro


Preparation of Purified Splenic Lymphocytes. For some

experiments, spleen cells were purified of Fc receptor-

bearing T cells and Ig-positive B cells and monocytes by

passage through rabbit immunoglobulin (Ig)-antirat Ig-coated

glass bead columns, prepared according to the protocol of

Wigzell et al. (108). After passage through the columns,

the remaining cells were washed in PBS and resuspended in

complete medium. The purified spleen cell populations were

observed to be 98% W3/13 monoclonal antibody reactive cells

with less than 2% MRC OX8-positive cells and B lymphocytes

present. (See below under leukocyte populations and T cell

subsets for discussion of monoclonal antibodies).

Preparation of Thymocytes. Thymocyte suspensions were

prepared using the above method for splenic lymphocytes, but

were resuspended in RPMI 1640 without FCS or 2-ME when used

in microcytotoxicity assays.

Preparation of Pancreas and Peritoneal Cavity Suspensions.

Pancreases were removed aseptically, rinsed in PBS, minced

and pressed through nylon mesh screens. The suspension was

then centrifuged at 1600 rpm for 10 minutes at 200C and re-

suspended in 1 ml of PBS. Peritoneal cells were obtained by

rinsing the open cavity with PBS and withdrawing the fluid

by pipette. This procedure was repeated several times until

the fluid removed from the peritoneal cavity was clear. The

suspension was then spun and resuspended in 1 ml of PBS.

Production of Con A Supernatants. Spleen cells from both

BB and WF rats at concentrations of 4-5 x 106 cells/ml

were incubated with 5 Vg/ml Con A in complete medium for

16-20 hours at 370C in 25 cm2 tissue culture flasks

(Corning, Medfield, MA). The contents of the flasks were

then centrifuged at 2000 rpm for 10 minutes, the cell-free

supernatants (Con A sups) removed and filtered through

0.45 pm filters (Sybron/Nalge, Rochester, NY) and stored at

-700 C for future use.

Detection of Organ-Specific Autoantibodies. BB and WF

rat sera were applied undiluted to slides of air-dried,

unfixed pancreatic or adrenal tissue and at 1:4 dilution and

at 1:10 dilution to thyroid and stomach sections respect-

ively. Optimal results (lowest background and highest sen-

sitivity) were obtained when these specified dilutions were

used on the various tissues. After incubation with rat sera

for 30 minutes in the dark, the slides were washed three

times in PBS. Rabbit antirat IgG-fluorescein isothiocya-

nate conjugate (Cappel Laboratories, Cochranville, PA) was

then added at 1:60 final dilution to the slides and

incubated for 60 minutes in the dark. After washing in PBS,

the slides were dried, covered with glycerol and glass slips

and read under a Leitz Dialux 20 ultraviolet glass micro-

scope fitted with an HB-100 mercury lamp and KP490, TK510

and K515 filters (E. Leitz, Inc., Rockleigh, NJ). All

results were read double blind with control negative and

antibody-positive sera in each batch. Sere were considered

to have autoantibodies if positive immunofluoresence of the

tissue was observed. Sera were tested for the presence of

autoantibodies at least monthly after the rats were 40 days

of age.

Measurement of Gastric Acidity and Serum Iron and Vitamin
B12 Levels.

Serum vitamin B12 levels were measured by radioimmunoassay

(Diagnostic Products, Los Angeles, CA). Serum iron levels

and iron binding capacities were measured by the ACA-III

autoanalyzer (Dupont Inc., Wilmington, DL). Gastric acidity

after pentagastrin administration (0.06 mg/kg body weight)

was determined by gastric aspirates during the fasting state

and was quantitated by color changes of pH indicator paper.

Transfer Studies. Spleen, mesenteric lymph nodes and

peritoneal suspensions from BB rats 1 to 10 days after onset

of IDD were injected into the tail veins of anesthesized

recipient nondiabetic BB, WF or BB x WF Fl rats using 25

gauge butterflys (Deseret, Sandy, UT). Pancreas suspensions

were given intraperitoneally. In some experiments, the

recipients were given 300-350 rads of irradiation 24 hours

before the transfer. Blood glucose levels were measured on

sera from recipient rats on day 0, day 3 and every week

thereafter for at least 90 days. Autoantibodies to the

cells of the pancreatic islets, the thyroid gland, and the

gastric mucosa were also sought monthly in sera recipient

rats. After approximately 90-120 days, the recipients were

sacrificed and the pancreases removed for histological


Determination of Leukocyte Populations and Rat T Cell

White blood cell counts were determined on whole blood after

1:200 dilution in acetic acid using a hemacytometer, and

differentials were made from slides of Wright-Giemsa stained


Monoclonal antisera to the various T cell subpopula-

tions were kindly provided by Dr. Alan Williams (Medical

Research Council Cellular Immunology Unit, Oxford, England).

The specificities of the antibodies were known to be as

follows: W3/13 all T lymphocytes, W3/25 helper T cells,

MRC OX8 cytotoxic/suppressor T cells, and MRC OX6 Ia-

positive cells (109). Aliquots of 1-2 x 106 PBL were

incubated at 40C for 30 minutes with 1:20 dilutions of these

antisera in addition to a 1:15 dilution of rabbit antirat Ig

(Accurate Chemicals, Westbury, NY) for determination of the

numbers of B cells and monocytes, and PBS or normal rat

serum alone as a control. After 2 washings with cold PBS,

all PBL aliquots were incubated with a 1:20 dilution of a

fluorescein isothiocynate conjugated goat antimouse IgG

(Cappel Laboratories) for 30 minutes at 40C, washed in cold

PBS, and resuspended in 30 Ils of PBS-glycerol. Slides were

then prepared and were observed for positive immunofluore-

scence under the ultraviolet microscope.

Detection of Thymocytotoxic Autoantibodies. Two p1

samples of sera from both diabetic and nondiabetic BB

animals of varying ages, WF rats and Lewis rats were applied

undiluted or at 1:2 dilution to 72 well microcytotoxicity

trays (Falcon, Oxnard, CA). A modification of the microcy-

totoxicity method of Amos was used for determination of

thymocytotoxic autoantibodies (110). One pl samples of

either a BB or WF thymocyte suspension at a concentration of

2 x 106 cells/ml were added to each well and incubated

with the various sera for 30 minutes at room temperature.

The cells were then washed in PBS and incubated with 4 pls

of a 1:10 dilution of guinea pig serum (Dutchland Labora-

tories, Denver, PA) for 60 minutes at room temperature.

After washing and staining with a 1% trypan blue solution,

the plates were read for determination of positive

cytotoxicity (greater than 50% cell death in any well).

Skin Graftings. Sections of ear pinnae were removed from

anesthesized nondiabetic BB, WF or Lewis rats, split in

half, and washed in sterile PBS. Sections of skin conform-

ing to the shapes of the ear grafts were removed from the

sides of shaved, anesthesized BB or WF rats. After washing

the wounds with sterile PBS, the ear grafts were placed onto

the prepared areas with the hairless sides down and lightly

sprayed with a plastic dressing (Aeroplast, Parke-Davis,

Greenwood, SC). Gauze bandages were wrapped around the rats

and left in place for one week and then removed. The grafts

were considered to have taken if no macroscopic evidence of

necrosis was apparent at this time. Skin grafted rats were

then followed for evidence of rejection by daily inspection.

Mitogen Assays. In most mitogen or MLC experiments, only

nondiabetic BB rats were used due to the possible effects of

hyperglycemia on lymphocyte responsiveness. However, a few

experiments were performed using well controlled diabetic BB

rats in order to see if similar results would be obtained.

Unseparated or purified (rabbit Ig antirat Ig treated)

splenic lymphocytes or PBL from nondiabetic or well con-

trolled diabetic BB and WF rats at various cell concentra-

tions ranging from 0.3-2 x 105 cells/well were cultured in

round-bottom microtiter plates (Costar, Cambridge, MA) with

several mitogen concentrations. Pokeweed mitogen (PWM;

1-25 yg/ml) (Sigma, St. Louis, MO), PHA (0.025-1%) (Difco,

Detroit, MI), Con A (0.5-10 yg/ml) (Miles-Yeda, Rehovoth,

Israel) and WF Con A sup (0.1/ml/well) were used for a total

volume of 0.2 ml of complete medium in each well. After 48

hours incubation at 370C in 5% C02, the cultures were

pulsed with 1.0 pCi/well of 3H-thymidine (Schwarz/Mann,

Spring Valley, NY, specific activity of 6 Ci/mM) and

harvested 18 hours later onto filter paper with a 24-line

cell harvester (Otto Hiller, Madison, WI). The filters were

then air-dried, placed into vials containing scintillation

fluid, and counted in a LKB Model 8100 liquid scintillation

counter (LKB Instruments, Rockville, MD). In several

experiments, 0.5 x 105 spleen cells irradiated with 3000

rads from a 137Cs source (Gammator Model M) were also

added to some wells. Some 0.5 x 105 or 1 x 105 BB and

WF thymocytes/well were also incubated with 0.1 pg/ml and 1

pg/ml Con A and 0.1 ml/well WF Con A sup for 48 hours at 370

C, pulsed with 1.0 pCi/well of 3H-thymidine and harvested

18 hours later.

Mixed Leukocyte Cultures (MLCs). MLCs were performed in

round-bottom microtiter plates, with each well containing

0.25-1 x 105 unseparated or purified responder spleen

cells and 0.5-3 x 105 irradiated (3000 rads) unseparated

spleen cells as stimulators. In addition, third-party

irradiated or nonirradiated unseparated spleen cells at a

concentration of either 0.25 x 105 or 0.5 x 105 cells

were added to some wells with a final volume in each well

always of 0.2 ml of complete medium. The plates were incu-

batd at 370C in 5% CO2 for 5 days, pulsed with 1.0 pCi/

well of 3H-thymidine as previously described, and

harvested 18 hours later. Also, 0.1 ml of WF Con A sup was

also added in some cases to wells containing either

responders alone or both responder and stimulator cells.

Interleukin 2 (IL 2) Assay. Levels of IL 2 in Con A sups

from both BB and WF rats were determined by the stimulatory

activity of these samples on a murine IL 2-dependent cyto-

toxic T cell line (anti-EL4, generously provided by Dr.

Shiro Shimuzu, University of Florida). Some 20 x 103 cyto-

toxic T cells in 0.1 ml of complete medium were incubated at

370C for 22 hours with 0.1 ml of various Con A sups and 6

dilutions (50%-1.5%) of a reference murine Con A sup. The

cultures were then pulsed with 0.5 PCi/well of 3H-thymi-

dine and harvested 6 hours later.

Measurement of Gamma Globulin Levels. Gamma globulin

levels were determined on 10 pl serum samples from BB rats

with and without IDD, and WF rats using the Beckman micro-

zone serum protein electrophoresis system.


Age at Onset and Frequency of IDD. The frequency of IDD

was determined in the original 20 BB/O and 28 BB/W rats.

The incidence of IDD was found to be 80% (16/20) in BB/O

rats, while IDD developed in 75% (21/28) of BB/W animals

(Table 1). However, two female BB/O rats greater than 285

days of age and one male BB/W rat who had reached 180 days

of age without developing overt IDD, had one hour peak

glucose levels of 513 mg/dl, 671 mg/dl and 298 mg/dl

respectively, after glucose loading. Peak glucose levels

seen in three control rats did not exceed 220 mg/dl after

glucose loading (data not shown). Thus, these three BB rats

were considered to have noninsulin-dependent diabetes.

No significant sex difference was seen in the frequency

of IDD in either BB/O or BB/W animals. The time period for

onset of IDD in the BB/O rats was from 90 to 120 days of age

(Figure 1), while the BB/W rats had a broader age range for

onset of IDD of between 70 and 161 days (Figure 2). All BB

rats were studied beyond 180 days of age, which exceeded the

critical age span for development of the disease in these

rats. Mild to moderate insulitis was seen in pancreases

from BB rats at onset of IDD (Figure 3). None of the










BB x WF Fl



aOne rat had equivocal PCA. All rats ascertained for the
above were studied up to and beyond 6 months of age.

Thyroid colloid autoantibodies.


FIGURE 1. Frequencies of insulin-dependent diabetes (IDD),
gastric parietal cell antibodies (PCA) and smooth muscle
antibodies (SMA) in 20 BB/O rats with age.

o---o PCA
1001 o--0 SMA
90 IDD

m 60



60 80 0 120 140 160 180 220 4
AGE (days)






5 ~
o ^3

o---o PCA
--a SMA

AGE (days)



FIGURE 3. Section of pancreas from a BB rat at onset of
IDD. The islet in the center has disorganized architecture
and lymphocytic infiltration (black arrow). A normal beta
cell is indicated by a clear arrow.


control WF animals or BB x WF Fl rats, which were followed

for at least 180 days after birth, developed IDD or

pancreatic insulitis (Table 1), indicating that IDD in the

BB rat is not due to a single dominant gene.

Presence of Organ-Specific Autoantibodies. ICA and other

organ-specific autoantibodies were sought in BB/O and BB/W

rats because these antibodies are characteristically found

in human IDD, are evidence for an autoimmune etiology for

IDD in the BB rat, and can be used as markers for animals

with autoimmune tendencies. ICA were never detected in any

BB rats, regardless of age or duration of IDD. To rule out

the possibility that ICA were present but were directed

against diabetic antigens, ICA were also sought on pancre-

atic sections from both nondiabetic BB rats and BB rats at

onset of IDD. However, no such autoantibodies were identi-

fied. In addition, ICA were not demonstrated in BB rat sera

when using fluorescein conjugated antirat Ig (all classes)

instead of antirat IgG in case ICA of the IgM class were

present. Thyroid microsomal autoantibodies and adrenal

autoantibodies were never found in any BB animals.

However, autoantibodies to the parietal cells of the

gastric mucosa (PCA) (Figure 4), thyroid colloid antigens

and smooth muscle (SMA) (Figure 5) were demonstrated in the

sera of a considerable number of BB/O and BB/W rats (Table

1). PCA were detected in the sera of 35% (7/20) of the BB/O

rats and in the sera of of 68% (19/28) of the BB/W animals.

FIGURE 4. Positive indirect immunofluorescence staining of
the parietal cells of rat gastric funds.


FIGURE 5. Positive indirect immunofluorescence staining of
smooth muscle in rat gastric funds.


The intensity of immunofluorescence of the PCA-positive sera

tended to increase with duration of PCA. Except for an

equivocal result from the serum of one WF rat, none of the

WF or BB x WF Fl animals had demonstrable PCA in their

serum. As was the case with IDD, PCA did not seem to have a

dominant mode of inheritance. Thyroid colloid autoanti-

bodies were of low immunofluorescence intensity and were

less frequent than PCA, being found in only 5% (1/20) and

18% (5/28) of the sera from BB/O rats and BB/W rats,

respectively. None of the control animals had autoanti-

bodies to thyroid colloid. SMA were demonstrated in 55%

(11/20) of BB/O rats, in 61% (17/28) of BB/W rats, and in 9%

(7/80) of WF and BB x WF Fl animals. The presence of SMA

was unrelated to the presence of IDD or PCA in the BB rats

(Table 2).

Some 71% (5/7) of the BB/O rats with PCA and 79%

(15/19) of the BB/W rats with PCA had IDD (Table 2).

However, two PCA-positive BB animals (one BB/O and one BB/W)

without clinical evidence of IDD had abnormal glucose

tolerance tests as previously mentioned. Thus, only 15%

(4/26) of the BB rats with detectable PCA had no discernible

IDD confirmed by normal glucose tolerance tests.

The appearance of PCA in the serum usually shortly

preceded the onset of IDD in BB rats that developed the

disease. The frequency of PCA in the BB rats was seen to

increase at an age coincident with the development of IDD

and did not further rise after the critical age span for



Rats Studied Total PCA % SMA %

I. Total BB rats with IDD 37 54 57

Total BB rats without IDD 11 54 64

BB/O rats with IDD 16 31 50

BB/O rats without IDD 4 50 75

BB/W rats with IDD 21 71 62

BB/W rats without IDD 7 57 57

Rats Studied




II. Total


BB rats with PCA

BB rats without PCA

BB/O rats with PCA

BB/O rats without PCA

BB/W rats with PCA

BB/W rats without PCA



onset of IDD in these rats had passed (Figures 1 and 2).

Although BB/W rats were susceptible to IDD for a longer age

range than BB/O animals, the marked increase in the

frequency of PCA in BB/W rats between 70 and 110 days of age

mirrored the increase in PCA frequency observed in BB/O rats

between 90 and 113 days of age. The frequency of SMA in BB

rats increased with age and, unlike PCA, did not closely

parallel the development of IDD in BB/O rats. The frequency

of SMA, however, increased in parallel with PCA in the BB/W


Functional abnormalities of the gastric parietal cells

were sought in BB rats with PCA as evidence that gastric

autoimmunity may result in clinical disease. Achlorhydria

was not demonstrated in the gastric aspirates of BB rats

with PCA, the pH values of which ranged from 2 to 3.

Vitamin B12 levels were measured in sera from 14 BB rats

with PCA, 14 PCA-negative BB rats, and 7 control WF rats.

No statistically significant differences were seen in serum

vitamin B12 levels between the three groups as evaluated by

Student's t-test (Table 3). Comparison of serum iron levels

and total iron binding capacities between 14 BB rats with

PCA, 11 BB rats without PCA and 7 control WF animals also

revealed no significant differences (Table 3).

Histological examinations of stomach sections from 12

BB rats with PCA were performed in order to see if lympho-

cytic infiltration suggestive of autoimmunity were present.

All sections revealed mild to moderate lymphocytic



BB rats with PCA (n=14)a BB rats without PCAa (n=14) WF rats (n=7)

x = 472 + 38 (244-710) x = 466 + 44 (109-759) x = 473 + 49 (275-666)

Vitamin B12

BB rats with PCA (n=14) BB rats without PCA (n=ll) WF rats (n=7)

Iron x = 223 + 13 ( 84-352) x = 210 + 21 (131-326) x = 234 + 33 ( 84-372)

Total iron binding capacity x = 492 + 17 (328-620) x = 474 + 36 (355-565) x = 519 + 78 (285-742)

% iron binding x = 45 + 2 ( 33- 78) x = 48 + 5 ( 35- 60) x = 51 + 6 ( 26- 70)

aThe BB rats used are from the group of 48 animals designated as either BB/O (20 rats) or as BB/W (28 rats).
All results are expressed as the mean + 1 standard error of the mean.
bSera from 11 BB rats without PCA were used to determine serum iron levels, however only sera from 9 BB rats
without PCA had measurements for iron binding capacity and percent iron binding performed.

infiltrations of the gastric mucosa with some loss of normal

mucosal cells and increased fibrosis (Figure 7), in

comparison with gastric funds obtained from WF and PCA-

negative BB rats (Figure 6). However, severe atrophy of the

gastric mucosa was not found in any of the tissues studied.

In two BB rats with PCA for the longest periods of approxi-

mately 7 months, degrees of squamous metaplasia of the

gastric mucosa were seen in sections taken well below the

junction between the proximal stomach and the funds (Figure

8). Only one of these rats had IDD. No stomach sections

from spleen BB rats without PCA or 5 control WF rats

revealed inflammatory lesions of the gastric mucosa.

Characterization of Peripheral Leukocyte Populations. Due

to observations of increased susceptibility to opportunistic

infections (especially of the respiratory tract) in both

diabetic and nondiabetic BB rats, and indications of both

pancreatic and gastric autoimmunities in this strain, immuno-

logical studies of these animals were performed.

Increased percentages and absolute numbers of

peripheral blood polymorphonuclear leukocytes (PMNs) were

observed in all BB rats regardless of the presence of IDD,

in comparison with WF and Fl hybrid rats (p <0.0025, Table

4), perhaps reflective of the increased rate of infections

among these animals. In contrast, all 16 BB rats with IDD

and all 32 nondiabetic rats, ranging in age from 25 to 400

days, were observed to have significantly decreased absolute

FIGURE 6. Section of normal BB rat gastric funds stained
with hematoxylin and eosin. Organized rows of parietal
cells (black arrow) are present.

'" r -


FIGURE 7. Hematoxylin and eosin stained section of gastric
funds from a BB rat with PCA showing lymphocytic
infiltration of the mucosa (black arrow).

53 iu

FIGURE 8. Hematoxylin and eosin stained section of gastric
funds from a BB rat with PCA for 7 months. Lymphocytic
infiltration, fibrosis and squamous metaplasia (black arrow)
of the mucosa are present.






% Lymphn


% Monoc


% Eosinc




BB Rats With NonDiabetic BB
IDD (n=16) Rats (n=32)

tes/mn3 5673 + 1840a'b 5655 + 1761b

3 2770 + 1132c 2225 + 1015c

57 + 11c 34 + 11C

tes/mn3 2805 + 1071c'e 3403 + 1041b

cytes 47 + 15c'd 60 + 12b

es/mn3 228 + 140 187 + 113

ytes 4+ 4 4+ 2

lils/mn3 46 + 45 136 + 309

)phils 2+ 4 2+ 4



WF Rats (n=28) Fl

9680 + 2858 108

1334 + 923 12'

14 + 7

7740 + 2317 911

82 + 7

277 + 201 4

3+ 1

31 + 58

1+ 3


43 +

75 +

12 +

05 +

34 +

32 +














Table 4-extended.

a Each value is stated as mean + S.D.
p <0.0005 by Student's t test when compared to WF and Fl rats.

Sp <0.0025 when compared to WF and Fl rats.
p <0.0025 when aorpared to BB rats without IDD.

p <0.01 when ccapared to BB rats without IDD.

numbers of both total peripheral blood leukocytes and

lymphocytes, when compared to 28 control WF rats and 8 Fl

hybrid animals (p <0.0005). The lymphopenia was more

striking than the leukopenia, reflecting significant

complementary decreases in the percentage of lymphocytes

seen in BB rats from that observed in WF and BB x WF Fl rats

(p <0.0005). Significant differences were also seen in both

the percentages and absolute numbers of lymphocytes in

diabetic BB rats when compared to BB rats without IDD (p


Due to the severe lymphopenia observed in all BB rats,

analyses of lymphocyte subpopulations were next performed.

Irrespective of age or the presence of IDD, increased

percentages (p <0.0005) but similar absolute numbers of Ia-

positive cells (MRC OX6 monoclonal antibody reactive) and Ig-

positive cells were found in BB rats in comparison with WF

rats and BB x WF Fl rats (Table 5). In contrast, the

absolute numbers of peripheral T lymphocytes (W3/13 mono-

clonal antibody reactive) were significantly lower in all BB

rats when compared to WF and Fl animals (p <0.001). Corre-

spondingly significant depressions of both absolute numbers

(p <0.005) and percentages (p <0.025) of W3/25-positive

cells (helper T cells) were observed in all BB rats,

independent of age or IDD, in comparison with WF and Fl

animals. Increased percentages (p <0.005) but decreased

absolute numbers (p <0.005) of cytotoxic/ suppressor T

lymphocytes (MRC OX8 monoclonal antibody reactive) were also




Animals Studied

BB Rats With BB Rats Without
IDD (n=8) IDD (n=15) WF Rats (n=12) Fl Rats (n=4)



W3/13+ Cells/mn3

% W3/13+ Cells

W3/25+ Cells/mmn3

% W3/25' Cells

MRC OX8+ Cells/rm3

% MRC OXB+ Cells

MRC X6m+ Cells/mm3

% MRC OX6+ Cells















































































Table 5-extended.

Ig Cells/mn3

% Ig+ Cells

1158 + 446

36.5 + 10b

1384 +

47 +



1324 + 509

16 + 4

1370 + 899

17 + 8

aEach value is stated as mean + S.D.

p <0.0005 by Student's t test when compared to WF rats.

p <0.005 by Student's t test when compared to WF rats.

p <0.025 by Student's t test when compared to WF rats.
ep 01 by Student's t test when pared to WF rats.
p <0.01 by Student's t test when compared to WF rats.

observed in all BB rats when compared to WF control and BB x

WF Fl rats. Numbers of T lymphocytes in both lymphocyte

subsets of Fl hybrid animals tended to range between those

of WF and BB rats. However, there were no significant

differences between WF and BB x WF Fl values.

An inversion of the W3/25-positive subset to MRC OX8-

positive subset ratio to less than 1.0 (mean 0.7 + 0.2) also

occurred in BB rats between 75 to 115 days of age, which was

not influenced by the presence of IDD (p <0.001, Figure 9).

In younger BB rats with and without IDD, the mean W3/25-

positive subset to MRC OX8-positive subset ratio was similar

to the mean ratio seen in WF rats at all ages studied (1.2 +

0.2 versus 1.3 + 0.1).

Presence of Thymocytotoxic Autoantibodies. As a possible

explanation for the extremely decreased numbers of T lympho-

cytes present in BB rats, autoantibodies to BB or WF thymo-

cytes were sought in these animals. As shown in Table 6,

many unabsorbed sera from both diabetic and nondiabetic BB

rats had demonstrable thymocytotoxic autoantibodies,

especially when the sera were tested at 1:2 dilution.

Although a few WF sera were also antibody positive, signifi-

cantly more BB sera had antibodies to thymocytes (p <0.05).

Only sera giving reactions of greater than 50% cytotoxicity

were considered to be positive for these autoantibodies.

Depressed Ability to Reject Allografts. Because BB rats

had both severely decreased numbers of peripheral T lympho-

FIGURE 9. A plot of the ratio of the W3/25-positive subset
(helper T lymphocytes) to the MRC OX8-positive subset
(cytotoxic/ suppressor T cells) versus age of the BB rats.
N designates nondiabetic BB rats and D designates BB rats
with IDD.

I- 1.6
S1.5 N
1.2 NDN
o1.l N ND
S.9 NN N
.8 N D
m *7 N ND
.6 D ND
.5 D N
.4 N

58 188 150 288 258 388 35 400 88

AGE (days)




BB rats with IDD

BB rats without IDD


1:2 dilution




1:2 dilution







WF rats undiluted

1:2 dilution

ap <0.05 by Chi-square

bp 0.01 when compared
p <0.01 when compared

6% (1/18)

9% (2/23)

analysis when compared to WF rats.

to WF rats.



cytes and circulating thymocytotoxic autoantibodies, the

ability of T lymphocytes from BB rats to function normally

in vivo was studied. Both BB and WF rats are thought to

share the rat major histocompatibility complex

genotype, while Lewis rats have the RT.11 haplotype (106-

111). Lewis skin grafts would thus be expected to be

rejected by both BB and WF rats in less than 14 days.

However, as seen in Table 7, nondiabetic BB rats rejected

Lewis allografts significantly more slowly than expected and

in comparison with WF controls (p <0.0005). Furthermore, WF

skin grafts were not rejected by nondiabetic BB rats, while

rejections of BB allografts by WF rats occurred in 17 + 3

days (p <0.0005), suggesting the presence of multiple minor

histocompatibility differences between BB and WF rats. BB

rats were thus extremely deficient in their ability to

reject grafts across both major and minor histocompatibility

barriers, and this defect was not dependent on the presence

of IDD.

Mitogen Responsiveness. Since the skin graft results

suggested that BB rats have defective in vivo T cell-

mediated immune responses, the ability of T lymphocytes from

BB rats to respond to mitogens in vitro was next studied.

PWM, PHA and Con A, which are primarily T cell mitogens in

the rat (112-114), were used. Dose-response curves were

initially prepared for each mitogen using varying concentra-

tions of each mitogen and of WF spleen cells (0.3-2 x 105



Recipients Donor Graft Survival (Days)a

8 BB rats without IDD Lewis 30 + 4

6 WF rats Lewis 12 + 2

6 BB rats without IDD WF > 90

6 WF rats BB 17 + 3

aThe period of graft survival was measured from day of
graft placement to day of complete graft rejection. Each
value is expressed as mean + S.D.

cells/well) as responders (Figures 10-12). Optimal stimu-

lation indices were obtained from these curves and

subsequent experiments using 1.0 pg/ml and 5.0 pg/ml PWM,

0.5% and 1.0% PHA, and 1.0 Vg/ml and 5.0 pg/ml Con A in each

well. WF cells responded well to these mitogens at all cell

concentrations tested, but subsequent response comparisons

with BB cells were made using spleen cell or PBL concentra-

tions at 0.3 x 105, 1 x 105 and 2 x 105 cells/well.

The results were not dependent on the BB cell concentrations

used or on the presence of IDD in the BB rats.

Utilizing spleen cell concentrations of 0.3 x 105

cells/well (Figure 13), 1 x 105 cells/well (Figure 14),

and 2 x 105 cells/well (p <0.0005, Figure 15), BB splenic

lymphocytes showed markedly diminutive responses to PWM in

comparison to WF splenic lymphocytes at the same mitogen

concentration. Similar relative results were obtained when

using 0.3 x 105 PBL/well (Figure 16) or 1 x 105 PBL/well

(p <0.01, Figure 17). However, the level of proliferation

of WF lymphocytes to PWM using 0.3 x 105 PBL/well was

quite low, suggesting that insufficient numbers and/or types

of cells were present for optimal proliferation. Both BB

and WF proliferative responses to PWM (also PHA and Con A)

were higher using -spleen cells rather than PBL as responders

but background counts also tended to be higher. PWM

responses of BB lymphocytes did increase with increasing

cell numbers per well as expected. However, BB results

comparable to those by WF cells were only observed when

FIGURE 10. Proliferative responses (cpm) of increasing
concentrations of WF splenic lymphocytes to 3 concentrations
of PWM measured on day 3. The following spleen cell
concentrations were used: 0.2 x 105 cells/well (open
circle), 0.4 x 105 cells/well (open triangle), 0.6 x 105
cells/well (open square), 0.8 x 105 cells/well (open
hexagon), 1.0 x 105 cells/well (closed circlet, 1.2 x
105 cells/well (closed triangle) and 1.6 x 105 cells/
well (closed square).






8 5 10 25

PHM Concentration (pg/ml)

FIGURE 11. Proliferative responses (cpm) of increasing
concentrations of WF splenic lymphocytes to 5 concentrations
of PHA measured after 3 days. These spleen cell concentra-
tions were used: 0.2 x 105 cells/well (open circle), 0.4
x 105 cells/well (closed circle), 0.6 x 105 cells/well
(open triangle), 0.8 x 105 cells/well (closed triangle),
1.0 x 105 cells/well (open square 1.2 x 105 cells/
well (closed square) and 1.6 x 10 cells/well (open



c se


PHR Concentration (%)

a .025 .05 .1 .2 .4

FIGURE 12. Proliferative responses (cpm) of increasing
concentrations of WF splenic lymphocytes to 3 concentrations
of Con A measured on day 3. The following spleen cell con-
centrations were used: 0.2 x 105 cells/well (open
circle), 0.4 x 105 cells/well (open triangle), 0.6 x 105
cells/well (open square), 0.8 x 105 cells/well (open
hexagon), 1.0 x 100 cells/well (closed circle), 1.2 x
105 cells/well (closed triangle) and 1.6 x 105
cells/well (closed square).

0 1.0 5.0 10.8

Con R Concentrations (ug/ml)

FIGURE 13. Responses (cpm) of splenic lymphocytes at 0.3 x
105 cells/well from 2 nondiabetic BB rats (striped bars)
and 2 WF rats (open bars) to 1 pg/ml PWM and 4 pg/ml PWM.
Each value is mean of triplicate cultures + S.D.



m 36


Q_ 24


1 ug/ml PWM

0 ug/ml PWM

5 mg/ml PWM

FIGURE 14. Proliferative responses (cpm) of splenic lympho-
cytes at 1 x 105 cells/well from 2 nondiabetic BB rats
(striped bars) and 3 WF rats (open bars) to 2 concentrations
of PWM. Each value is mean of triplicate cultures + S.D.



8 ug/ml PWM

5 ag/ml PNM

1 pg/ml PWM

FIGURE 15. Responses (cpm) of splenic lymphocytes at 2 x
105 cells/well from 4 nondiabetic BB rats (striped bars)
and 4 WF rats (open bars) to 10 pg/ml PWM. Each value is
mean of triplicate cultures + S.D.



18 Mg/ml PHM

80 0g/ml PWM

FIGURE 16. Responses (cpm) of PBL at 0.3 x 105 cells/well
from 1 diabetic BB rat (first striped bar), 2 nondiabetic
BB rats (remaining striped bars) and 3 WF rats (open bars)
to 1 g/ml PWM and 5 pg/ml PWM. Each value is mean of
triplicate cultures + S.D.

9 jg/ml PWM

1 pg/ml PM

5 Mg/ml PHM

FIGURE 17. PWM responses (cpm) of PBL at 1.0 x 105
cells/well from 3 nondiabetic BB rats (striped bars) and 5
WF rats (open bars). Each value is mean of triplicate
cultures + S.D.

0~ug/mI PWM 1.0 pg / ml PWM 5.0 ~g/ml PWM






? ..^r~ 0,riir~ 7

Og/ml PWM

1.0 Ajg / mi PWM

5.0 ug/mi PWM


using 2 x 105 BB spleen cells/well and 0.3 x 105 WF

spleen cells/well.

Lymphocytes from BB rats also responded poorly to

various concentrations of PHA at 0.3 x 105 spleen cells/

well (Figure 18), 1 x 105 spleen cells/well (Figure 19),

and 2 x 105 spleen cells/well (p <0.05, Figure 20), in

comparison with WF lymphocytes. Although similar results

were obtained when using 1 x 105 PBL/well (p <0.01, Figure

21), low levels of proliferation with no significant

differences between WF and BB PHA responses at 0.3 x 105

PBL/well (Figure 22) were noted, most likely due to

insufficient numbers of cells as stated previously.

Although lymphocytes from BB rats were most responsive

to Con A of the mitogens used, WF cells still responded

significantly better at all of the following cell concen-

trations: 0.3 x 105 spleen cells/well (Figure 23),

1 x 105 spleen cells/well (Figure 24), 1 x 105 PBL/well

(p <0.01, Figure 25) and 2 x 105 spleen cells/well (p

<0.005, Figure 26). As was true for PWM and PHA, WF

lymphocytes at 0.3 x 105 PBL/well responded poorly to Con

A and thus no significant differences were observed at this

cell concentration between BB and WF proliferative responses

(Figure 27). However, in one experiment, splenic

lymphocytes from a single BB rat at 2 x 105 cells/well

were able to mount a significant response to Con A

comparable to responses by WF spleen cells. However, BB

lymphocytes from the same animal minimally responded to Con

FIGURE 18. PHA responses (cpm) of splenic lymphocytes at
0.3 x 105 cells/well from 2 nondiabetic BB rats (striped
bars) and 2 WF rats (open bars). Each value is mean of
triplicate cultures + S.D.



8I H 8.5 PHA -

ex PHR

8.5% PHA

FIGURE 19. Proliferative responses (cpm) of splenic lympho-
cytes at 1 x 105 cells/well from 2 nondiabetic BB rats
(striped bars) and 2 WF rats (open bars) to 0.5% and 1.0%
PHA. Each value is mean of triplicate cultures + S.D.



' 158




8.5% PHR


1.8% PHR

FIGURE 20. Comparison of proliferative responses (cpm) of
splenic lymphocytes at 2 x 105 cells/well from 2 non-
diabetic BB rats (striped bars) and 2 WF rats (open bars)
to 5 concentrations of PHA. Each value is mean of
triplicate cultures + S.D.



8 PH .825% PHA .85% PHR

.1% PHR .2% PHR .4% PHR

FIGURE 21. Proliferative responses (cpm) of PBL at 1 x
105 cells/ well from 1 diabetic BB rat (first striped
bar), 2 nondiabetic BB rats (remaining striped bars) and 5
WF rats (open bars) to 2 PHA concentrations. Each value is
mean of triplicate cultures + S.D.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E9EMTPBE6_MNBAA0 INGEST_TIME 2013-01-23T14:01:12Z PACKAGE AA00012914_00001