Diabetes-Prone and Diabetes-Resistant BB Rats: Animal Models of Spontaneous and Virally Induced Diabetes Mellitus, Lymphocytic Thyroiditis, and Collagen-Induced Arthritis

Mr. Guberski is the BB/Wor project administrator at the University of Massachusetts Medical Center in Worcester. In 1974, some rats (subsequently called BB) that spontaneously developed autoimmune diabetes mellitus were found in a closed colony of outbred WI (Wistar) rats at the Bio-Breeding Laboratories, Ottawa, Ontario. Since the first report by Nakhooda et al. (1977 ), colonies of BB rats have been established worldwide. The availability of the model enables investigation into fundamental questions concerning the spontaneous etiopathogenesis ( Crisä et al., 1992 ) and virally induced mechanisms ( Guberski et al., 1991 ) of diabetes. Complications of diabetes, including peripheral neuropathy ( Greene et al., 1984 ; Yagihashi and Sima 1985a , b ; Sima et al., 1986 ; Yagihashi et al., 1986 ; McEwen and Sima, 1987 ), impotence ( Murray et al., 1985a , b ), nephropathy ( Chakrabarti et al., 1989 ; Cohen et al., 1987 ), and retinal changes ( Sima et al., 1985 ; Chakrabarti et al., 1991 ), are also widely studied in this model. This paper will provide an overview of the model, describing the development of the inbred rats, features of the diabetic syndrome, the immunopathology and mechanisms of immune suppression and immune modulation, and the interaction Of genetics and the environment, as well as a brief review of recommended husbandry practices. In 1977, Butler et al. (1983 , 1988 , 1990 ) began inbreeding BB rats at the University of Massachusetts Medical Center (laboratory code Wor) with 300 breeders purchased from the Bio-Breeding Laboratories. In 1983, the National Institutes of Health established the Wor inbred colony as a reference colony. The first priority was to develop an inbred model with good fecundity and a high incidence of diabetes. Diabetes was defined as positive glycosuria with a blood glucose concentration greater than 250 mg/dl. The incidence of diabetes was determined in breeders held through 150 days of age. Selection of siblings for breeding occurred after the onset of hyperglycemia. In 1978, between the third and fourth generation of inbreeding, an epidemic of Mycoplasma pumonis caused a substantial number of animals to die. As a result, a pathogen-free rodent barrier system was constructed for the colony. During caesarian rederivation, animals from litters with the highest incidence of diabetes were introduced into the barrier. Following the successful rederivation, genetic studies were undertaken to determine the mechanism of inheritance of diabetes. From these studies two groups of animals were produced. The first was comprised of the progeny of diabetic x diabetic matings, from which six diabetes prone (DP) strains were developed. The second was comprised of progeny selected to remain diabetes free, including the BBVB/Wor strain, a direct descendant from the original 300 breeders, and the BBDR/Wor strain derived from DP rats in the fifth generation of inbreeding. Characteristics of the DP and DR strains are listed in Table 1 . Characteristics of Diabetes Prone and Diabetes Resistant Rats Characteristics of Diabetes Prone and Diabetes Resistant Rats Genetic studies showed that a single diabetic x diabetic mating fixed the recessive diabetogenic genes. Subsequent progeny all had the same incidence of diabetes whether they were the offspring of a phenotypically diabetic x nondiabetic mating or a nondiabetic x nondiabetic mating ( Butler et al., 1983 ). This revolutionized the breeding at the Wot facility. Instead of waiting until hyperglycemia occurred and then mating chronically diabetic rats, mating was now possible before the onset of diabetes, which substantially improved production. New selection criteria were also developed, including selecting for animals that developed hyperglycemia at an early age and came from large litters with a high incidence of diabetes. Animals with an early onset of hyperglycemia are difficult to breed ( Murray et al., 1985b ; Butler et al., 1988 , 1990) due to the difficulty of controlling blood glucose in females during pregnancy and lactation and impotence in chronically diabetic males ( Murray et al., 1985a , b ). Selection criteria that reduces fertility can cause residual heterozygosity and delay the development of fully inbred lines ( Hedrich, 1993 ). Because diabetes reduces productivity in DP rats and the onset of the disease typically occurs during pregnancy or lactation ( Figure 1 ), residual heterozygosity may still be present in the Wor reference colony. Sound management practice requires continuation of sibling matings. Age at onset of diabetes n diabetes-prone BB/Wor rats was determined by glycosuric testing animals 3 times per week. Diabetes was defined as 4+ glycosuria with subsequent sera glucose measurement greater than 250 mg%/(13.8mmol/L). Animals studied represent viral antibody-free breeders from all diabetes-prone lines from generations 37-48. DP strains are designated BBBA/Wor, BBDP/Wor, BBBE/Wor, BBNA/Wor, BBNB/Wor, and BBPA/Wor, and DR strains are designated BBDR/Wor and BBVB/Wor ( Table 1 ). All strains have been inbred for more than 48 generations. DP rats develop a cell-mediated autoimmune destruction of the pancreaticβ cells, resulting in an abrupt onset of insulin-dependent, ketosis-prone diabetes mellitus. The development of diabetes in DP strains occurs between 60 and 120 days of age in both males and females. The abrupt onset of disease is characterized by weight loss, polyuria, polydipsia, glycosuria, hyperglycemia, hypoinsulinemia, and ketosis. In the absence of exogenous insulin therapy, animals usually die in 4-7 days ( Figure 2 ). The overall incidence of diabetes in DP strains is 86 percent, and the average age of onset is 80 days ( Like et al., 1991 ). This peripubertal onset of diabetes is comparable to the peak onset observed in 12- to 14-year-old humans. Clinical course of diabetes in 23 lean diabetic male rats after cessation of insulin therapy. Twenty out of twenty-three rats died in diabetic ketoacidosis within 4 days after cessation of insulin therapy. Insulitis is a term coined by Gepts (1965 ), who described mononuclear infiltration of the islets of Langerhans in the pancreas of a child who died with acute type I diabetes mellitus. DP rats develop an intense mononuclear insulitis that precedes the onset of disease by 2-3 weeks. Insulitis destroys only the pancreatic β cells (( Figures 3 –( 4 ), resulting in an islet devoid of insulin-positive cells ( Figure 5 ); pancreatic alpha, delta, and polypeptide cells are not destroyed. Studies to determine which cells make up the insulitis lesion suggest that different lymphoid populations are present, depending on the stage of insulitis. The initial infiltrate is comprised of ED1+ cells (macrophages/dendritic cells) ( Kiesel et al., 1986 ; Lee et al., 1988 ). Later, CD4 +, CD8 +, and natural killer (NK) cells are found. Recently Ellerman et al. (1993 ) and Shachner et al. (1992 ) demonstrated that NK cells are not the final effector cells for β-cell destruction. In situ studies have demonstrated the presence of TNF, IL- 1, and IL-6 mRNA in islets of acutely diabetic DP rats ( Jiang and Woda, 1991 ). Pancreatic lymphocytic infiltration (PLI) has also been reported to precede the insulitis lesion ( Guttmann et al., 1983 ). However, PLI can also be found in animals that never become diabetic or demonstrate insulitis (Dr. Arthur A. Like, Personal Communication, University of Massachusetts Medical School, Worcester, Massachusetts). Normal islet from a BBDR/Wor rat. The centrally located β cells stain intensely for insulin (3A). Alpha cells located at the islets periphery stain positively for glucagon (3B). Islet from an acutely diabetic BB rat. Insulitis is widespread and is responsible for the destruction of the insulin producing β cells (4A). Glucagon positive cells (4B) are unaffected. End stage islet from a chronically diabetic BB rat. Insulin positive β cells are absent (5A) due to autoimmune destruction of pancreatic β cells. Peripherally located β cells (5B) have clustered following the collapse of the centrally located β cell core. DP strains are severely lymphopenic ( Elder and Maclaren, 1983 ), as manifested by a reduction in T lymphocytes in peripheral blood, spleen, and lymph nodes. Peripheral blood samples obtained from venous sinus punctures, after being sequentially incubated with anti-CD8 (OX8) and anti-CD5 (OX19), show a significantly lower than normal percentage of T-helper cells (CD5+/CD8-) and virtually no cytotoxic and suppressor (CD5+/CD8 +) T cells. DP strains also have a slightly greater than normal percentage of NK cells (CD8+/ CD5-). DR strains are not lymphopenic and have a normal number of T cells ( Like et al., 1991 ). RT6 is a T-cell differentiation alloantigen controlled by two alleles RT6a and RT6h . Those alleles make two products, RT6.1 and RT6.2, respectively, which are co-expressed in F1 progeny derived from mating a strain that is RT6.1 + with a strain that is RT6.2 + ( Crisä et al., 1992 ). Although normally absent on thymocytes and bone marrow cells, these cell-surface markers can be demonstrated on approximately 50 percent of CD4 + T-helper cells and 70 percent of CD8 + T-cytotoxic/suppressor cells. Overall, 60 percent of peripheral T cells express an RT6 marker. Intraepithelial lymphocytes (IELs) also express RT6 cell markers ( Fangman et al., 1991 ). T cells that express RT6 markers are believed to be immunoregulatory ( Greiner et al., 1988 ). Although DP strains have a functional RT6 a allele ( Crisä et al., 1993 ), they do not express RT6.1 antigen on peripheral T cells, on splenocytes, or in lymph nodes. These rats do, however, express RT6.1 markers on their IELs ( Fangman et al., 1991 ). DR strains express RT6.1 antigen in peripheral T cells, splenocytes, and in lymph nodes ( Like et al., 1991 ). DP rats develop autoantibodies to smooth muscle, thyroid colloid, gastric parietal cells, and islet cell-surface markers ( Elder et al., 1982 ; Like et al., 1982a ; Pipeleers et al., 1987 ; Dean et al., 1986 ). The significance of the production of autoantibodies and its pathogenic consequences to diabetes remains unclear. However, humans with organ-specific autoimmune disease typically produce similar autoantibodies. More than 90 percent of BBBA/Wor, BBNA/Wor, and BBNB/Wor rats develop lymphocytic thyroiditis (LT) by 150 days of age ( Sternthal et al., 1981 ; Rajatanavin et al., 1991 ), and the severity of the lymphocytic infiltration is more pronounced in these strains. In addition, the administration of iodine to the drinking water induces LT at an increased frequency by 90 days of age ( Allen et al., 1986 ). The peripheral T-cell receptor β-chain variable region (V β ) repertoire fails to develop in DP rats ( Gold and Bellgrau, 1991 ). That limited V β repertoire could be due to a differential T-cell expansion or a selective deletion of T-cell receptors. The importance of the defect and its contribution to the pathogenesis of diabetes requires further study. DR rats do not spontaneously develop autoantibodies, LT, or insulitis. DR rats are not lymphopenic, but they do harbor an effector-cell population that can be induced to become autoreactive T cells, resulting in diabetes or thyroiditis ( Like, 1990 ). The DR rat is also an excellent animal model for studying collagen-induced arthritis. When immunized with human type II collagen suspended in incomplete Freunds adjuvant, virtually all DR rats develop arthritis within 10 days ( Watson et al., 1990 ). It is a good model for collagen-induced arthritis because the third hypervariable region of the RT1.D β gene in DR rats encodes a nucleotide sequence identical to the human disease response genes (called HLA DR) associated with rheumatoid arthritis Watson et al., 1990 ). In DP rats , susceptibility to diabetes is complex and requires the interaction of several genes. First, there is a close association of genes present in the major histocompatibility complex (MHC) encoding for the RTl' haplotype on chromosome 20 ( Jacob et al., 1992 ). The development of diabetes is thought to be independent of class I expression but requires the presence of at least one class II RTl' allele ( Fuks et al., 1990 ). Additional support for that hypothesis is derived from in vivo studies, which show that treatment of DP rats with a monoclonal antibody directed against class II RT1.D but not anti-RTI.B prevented diabetes ( Boitard et al., 1985 ). Second, a recessive gene for lymphopenia, lyp, is permissive for diabetes and is tightly linked to the neuropeptide Y Nyp gene on chromosome 4 ( Jacob et al., 1992 ). Lymphopenia segregates independently of the RTl' haplotype, and studies show that lyp is an autosomal recessive gene ( Guberski et al., 1989 ; Herold et al., 1989 ). The relationship between lymphopenia and diabetes is also complex. While lymphopenia is permissive for diabetes, it is not an absolute requirement, because nonlympho-penic DR rats (BBDR/Wor and BBVB/Wor) can be induced to develop hyperglycemia by various means ( Like et al., 1986a ; Like, 1990 ). Other factors contributing to the diabetic phenotype include a dominant gene for PLI (called Pli) ( Guttmann et al., 1983 ) and a proposed genetic factor that controls age at onset ( Guberski et al., 1989 ). To date the use of biochemical and molecular markers have not identified a specific diabetogenic genotype ( Kastern et al., 1984 ; Kryspin- Sorensen et al., 1986 ; Bjorck et al., 1986 ). DR rats share the same MHC haplotype as DP rats but are not lymphopenic ( Like et al., 1986a ). If maintained in a pathogen-free environment, DR rats do not become diabetic. Since the rederivation of the strains in July 1989, more than ten thousand DR rats have been studied, and no diabetes has been detected. DR rats harbor effector cells capable of causing diabetes, however, and hyperglycemia can be induced by injection of polyinosinicpolycytidylic acid (usually called POLY-IC) ( Sobel et al., 1992 ), x irradiation ( Handler et al., 1989 ; Rossini et al., 1984a ), and exposure to or injection of Kilham rat virus (KRV) ( Guberski et al., 1991 ). RT6 depletion of conventionally housed DR rats can also induce diabetes ( Greiner et al., 1987 ). The diabetic phenotype is a combination of an animal's genetic predisposition and environmental factors that determine the penetrance of the diabetes genes. In inbred colonies of DP rats the genetic predisposition towards development of diabetes is uniform, but not all of the rats develop diabetes. Further, the incidence of diabetes varies among research institutions. One must therefore consider environmental factors, including viruses, bacteria, and diets. When the BB/Wor colony was rederived in 1989 by caesarian derivation, environmental pathogens were eliminated, causing significant changes in how many animals developed diabetes and at what age. Prior to rederivation, Sendai virus and sialodacryoadenitis virus were ubiquitous in the DP and DR colonies. The animals were seronegative for Mycoplasma pumonis and intermittently seropositive for KRV. The incidence of diabetes in DP rats was 70 percent, and the average age at onset was 91 days. Seven percent developed diabetes before 72 days of age, 74 percent between 72 and 102 days, and 19 percent after 103 days. After the viruses were eliminated, the incidence of diabetes increased from 70 percent to 86 percent, and the average age at onset was now 80 days. In addition, the distribution of ages at onset of diabetes was significantly earlier: 32 percent developed diabetes before 72 days of age, 62 percent between 72 and 102 days of age, and only 6 percent developed diabetes after 103 days. These studies ( Like et al., 1991 ) suggest that viruses may modulate the immune system of DP rats by interfering with the genetically programmed process of [3-cell destruction. Additional support for the hypothesis of viral immune modulation is the observation that injecting lymphocytic choriomeningitis virus into young DP rats prevents diabetes ( Dyrberg et al., 1988 ). Finally, Sadelain et al. (1990 ) has demonstrated that complete Freund's adjuvent (CFA) administered to young DP rats prevents diabetes, which suggests that CFA-induced cytokine release regulates immune responses. A review on cytokine regulation has recently been published ( Rabinovitch, 1993 ). the incidence of spontaneous diabetes in DR rats was than percent before the of environmental spontaneous of diabetes in DR rats were the first from ( Like et al., 1986a ) and the second from ( et al., 1991 ; Sadelain et al., 1990 ). Following the first of diabetes in the DR breeding studies out a genetic for diabetes. First, mating diabetic rats with normal DR rats not the incidence of diabetes in the Second, in which both were and diabetic to produce diabetic offspring ( Guberski et al., 1989 ). the spontaneous was linked to the presence of a rodent ( Guberski et al., 1991 ). During the second of diabetes in the DR a determined to be by of et al., was from the of a diabetic DR rat. That into to DR rats, induced diabetes and insulitis. viruses that induce diabetes, be demonstrated in the pancreatic islets by or in situ ( Guberski et al., 1991 ; et al., When is present, it is also permissive for of diabetes in DR rats following depletion of RT6.1 + T cells ( Like, 1990 ). into young DP rats fails to induce diabetes their immune have been with from DR immune of DP rats with by injection fails to induce diabetes ( Guberski et al., 1991 ). These studies suggest that the immune system is being by the by viral molecular molecular could induce the pancreatic cells to express a viral on the to the is that has a direct on the immune the between autoreactive effector and T cells. This mechanism could by a virally induced of effector cells or a virally induced suppression of T cells. The hypothesis that the DR immune system is by is derived from the following into to DP rats fails to induce DP rats immune are by of from DR or rats do not become DP rats immune are with from DR rats and are with become and DP rats immune are with from rats and with do not become diabetic. These that diabetes requires a DR immune system the DP cells were during these ( Guberski et al., 1991 ; Ellerman et al., 1992 ). The that environmental pathogens or their may was recently by Ellerman and Like (1992 ). In these were to in DR Those cells were in diabetes, which that are capable of autoreactive T cells, at least in Since approximately of of produce this ubiquitous may an autoimmune process in genetically The of on development of diabetes have been studied ( et al., 1985a , b ; et al., 1986 ; et al., , b ; and 1991 ). that products, as and are diabetes in and the incidence and delay the onset of diabetes ( et al., ), which may be caused by or by or associated with the of ( and 1991 ). for diabetes is by the presence of insulitis in the pancreatic islets at the onset of hyperglycemia and by the to disease by injecting cells from diabetic rats into normal rats ( et al., 1983 ). T-cell lines from diabetic rats are also capable of ( et al., 1993 ). Diabetes can be prevented by ( Like et al., ), injection of sera ( Like et al., ), and a of immune ( Like et al., ; Like et al., 1984 ; Rossini et al., ). Finally, studies have that diabetes and insulitis can be prevented by islet cells from DP or DR rats into the of DP rats, which suggests that a of prevents autoimmune destruction of pancreatic cells ( and 1992 ; et al., 1992 ). DP and DR rats be maintained as in or barrier to the of environmental pathogens ( Like et al., 1991 ). for glycosuria is the to rats for diabetes. is from the by and with A blood which can be obtained by the of the with a be within 2 of a positive for to blood Animals blood glucose 250 are diabetic and insulin therapy. for glycosuria before the onset of diabetes and be at least times at the of the treatment of diabetic rats with insulin is for their and on the that glycosuria is found and diabetes is The of insulin will be a of age, presence of ketoacidosis and and whether the animal is or The of insulin be for with the various are published in in in the rat be well of and maintained in a of to If with a the of insulin be and with be The successful treatment of requires a in insulin with are in in for in diabetes-prone rats insulin of females not not insulin during of 1 of 1 for in diabetes-prone rats insulin of females not not insulin during of 1 of 1 for in diabetes-prone (DP) rats for in diabetes-prone (DP) rats in diabetes-prone (DP) rats before and after from rats in diabetes-prone (DP) rats before and after from rats