The BB Diabetic Rat Profound T-Cell Lymphocytopenia R. JACKSON, N. RASSI, T. CRUMP, B. HAYNES, AND G. S. EISENBARTH

onset of diabetes was abrupt in the six animals that became SUMMARY Approximately 50% of Wistar "BB" rats spontaneously diabetic. Plasma glucose at the time of diagnosis of diabe- develop overt diabetes mellitus characterized by loss tes ranged from 284 to 649 mg/dl. Approximately every 2 of /3-cells and "insulitis." To define abnormalities of wk, animals were gently anesthetized with ether, and 1 ml of immunoregulation in these rats, we quantitated their blood was withdrawn from a tail vein into heparinized tubes major circulating subsets. Independent of for determination of white blood count, differential and the development of diabetes, we found the BB rats to lymphocyte subsets. White blood count was determined have a markedly increased percentage of circulating using a Coulter Zf counter and differential using Wright's B which is secondary to a severe T-cell stained slides. Peripheral white blood cells were then ap- lymphocytopenia, with the major circulating T-cell sub- plied to a Ficoll-Hypaque gradient (24:10 v,v; 9% Ficoll set reacting with monoclonal W3/25 markedly 400:34% Hypaque M) and the lymphocytes and monocytes decreased. This lymphocytopenia is present in every present at the interface of the gradient were harvested. This animal studied and contrasted with studies of the non- diabetic Wistar strain from which the "BB" rats were lymphocyte fraction was then incubated at 37°C for 45 min developed. DIABETES 30:887-889, October 1981. to remove cytophilic immunoglobulin, and then reacted with rhodamine conjugated F (ab')2 anti-rat antibody (Cappel) at 4°C for 45 min, followed by three washes with phosphate- he spontaneously diabetic Wistar rat (the "BB rat") buffered saline containing 1% bovine albumin. To deter- has become an important animal model of type I in- mine the percentage of B lymphocytes, the number of cells sulin-dependent diabetes mellitus.1-2 Approxi- with rim fluorescence was determined using a Leitz-diavert mately 50% of these nonobese outbred Wistar rats microscope equipped for epifluorescence. To quantitate T- subsets, the lymphocyte preparation was incubated between 60 and 120 days of age develop severe hypergly- cemia associated with /3-cell depletion, insulitis, and keto- with either monoclonal antibody supernatants W3/13 ("pan acidosis. Of major importance is the finding of Like et al. T marker") or 3/25 ("helper" T) followed by incubation with that various immunosuppressive regimens (including ad- fluorescein conjugated F(ab')2 anti-mouse antibody. To de- ministration of anti-lymphocyte globulin) can prevent the termine the percentage of T-cells, the number of cells with onset of overt diabetes in these animal.3-4 It is postulated fluorescein rim fluorescence was determined using the that the development of diabetes mellitus is secondary to in- fluorescein filter of the Leitz-diavert microscope with correc- herited abnormalities of immunologic function. In this report tion for B-cells. we describe our findings that all BB rats, independent of the development of overt diabetes, have a specific deficiency RESULTS of T-cells reacting with monoclonal antibody W3/25. Immunoglobulin molecules which function as antigen re- ceptors are integral proteins of the plasma membrane of B METHODS lymphocytes. B lymphocytes can therefore be detected by Nineteen rats of the BB strain and 20 rats of the control, non- their reaction with rhodamine-conjugated anti-rat anti- diabetic Wistar strain (male and female) were kindly pro- vided by Dr. P. Thibert. Rats were screened for glucosuria From the Divisions of Endocrinology and Rheumatology, Department of Medi- using Testape (Eli Lilly and Company, Indianapolis, Indi- cine and Physiology, Duke University Medical Center, Durham, North Caro- ana) approximately every 3 days, and plasma glucose was lina. Address reprint requests to G. S. Eisenbarth, Box 3257, Duke University Med- measured using a Beckman glucose analyzer (Beckman In- ical Center, Durham, North Carolina 27710. struments, Fullerton, California) at least once a week. The Received for publication 3 August 1981.

DIABETES, VOL 30, OCTOBER 1981 887 R. JACKSON AND ASSOCIATES

BB Wistar No D.M. BB Wistar DM. Control Wistar Strain Onset DM.

2 3 4 10 15 20 BB Control B Non B Determination Age (Weeks) Wistar Lymphocytes Lymphocytes FIGURE 2. Percentage of B lymphocytes and number of B and non-B FIGURE 1. Percentage of B-cells of BB Wistar and control Wistar rats. 3 Panel A: Mean ± SEM percentage of Sig+ cells of Ficoll-Hypaque lymphocytes/mm of control Wistar rats and BB Wistar rats, plotted purified lymphocytes of BB Wistar rats, which did and did not develop as the mean ± SEM of four control strain Wistar rats and five BB rats. 14 Panel A: Percentage of B cells determined as described In METHODS. overt diabetes, and control Wistar rats. Determinations are grouped 3 In 3-wk intervals starting at 10 wk of age. Panel B: Percentage of Sig+ Panel B: Number of B and non-B lymphocytes/mm determined as cells of three representative individual rats, one BB Wistar that described in METHODS. developed diabetes, one BB Wistar that did not develop diabetes, and one animal of the control rat strain. tion of B lymphocytes to the latest (even after the develop- ment of overt diabetes), the percentage of B lymphocytes of bodies which bind to cell surface immunoglobulin (slg). BB rats for each animal was elevated. When we initially determined the percentage of slg-positive This could be either because of an increase in B lympho- lymphocytes of BB rats, we were surprised to find it to be cytes, or because of a decrease in non-B lymphocytes, pri- greater than 25% in every animal studied. We expected the marily T lymphocytes. That the latter explanation is correct percentage of B lymphocytes to be less than 15%, as is the is shown by the data illustrated in Figure 2 and Table 1, ex- case with a number of other rat strains and with the control, periment 1. The percentage of B lymphocytes, as well as the nondiabetic Wistar strain provided by Dr. P. Thibert, the number of B lymphocytes/mm3 and non-B lympho- strain from which the BB rat was derived. As is apparent cytes/mm3, were determined for both BB rats and control from Figure 1, the percentage of B lymphocytes is markedly Wistar rats. The percentage of B lymphocytes, as previously elevated not only in the 6 BB rats that become diabetic but shown, was markedly increased (34.4% vs. 13%, P < 0.01) also in the 13 animals that did not become diabetic. Panel A as compared with control Wistar rats (Figure 2). Despite the of Figure 1 plots the mean percentage of B lymphocytes for dramatic increased percentage of B lymphocytes, their ab- all the animals, while panel B illustrates studies of three rep- solute numbers (1581 vs. 1338) were not significantly differ- resentative individual animals. From the earliest determina- ent. The non-B lymphocytes were markedly decreased

TABLE 1 Comparison of circulating white blood cells

Control strain BB Wistar Difference

Experiment 1 White blood count 12,612* ± 1677 7,422 ± 805 5,190f Polymorphonuclear leucocytes 1,982 ±419 2600 ± 560 -618 Monocytes 418 ± 64 283 ± 68 135 Lymphocytes 10,290 ± 1220 4,549 ± 409 5,741f B lymphocytes 1,338 ± 171 1,581 ± 224 -243 Non-B lymphocytes 8,953 ± 1058 2,968 ± 246 5,985f Experiment 2 B lymphocytes 1,569 ±242 1,495 ±595 74 Non-B lymphocytes 6,245 ± 765 2,412 ± 451 3,833f W3/13 + non-B 5,929 ± 727 2,111 ± 396 3,818f W3/25 + non-B 4,996 ±612 169 ±32 4,827f W3/25 - non-B 1,249 ± 153 2,243 ±419 -994

* In experiment 1, mean ± SEM of four control rats and five BB Wistar rats; experiment 2, mean ± SEM of five control rats and three BB Wistar rats. t P < 0.01 by Student's t test.

888 DIABETES, VOL. 30, OCTOBER 1981 THE BB DIABETIC RAT: T-CELL LYMPHOCYTOPENIA (8953 vs. 2968, P < 0.01) in the BB rats. As shown in Table eases, erythemaosus in particular.9 We have recently 1, experiment 1, which lists the number of cells per mm3, ab- found that some patients with lupus erythematosus and normalities in total white blood cells, and in total lympho- other diseases characterized by polyclonal B lymphocyte cytes results from marked non-B-cell lymphocytopenia. activation have anti-islet reacting with sections Since the number of monocytes are unchanged, this proba- of Bouin's fixed pancreas.10 Similarly, autoimmune animal bly represents a severe T-cell lymphocytopenia. models which exhibit insulitis have been discovered.11 It is To confirm the T-cell lymphocytopenia and to evaluate T- possible that the BB rat is similar to these autoimmune mice, cell subsets of the BB rat, we studied the reaction of murine but with quantitatively greater /3-celI destruction. Studies to monoclonal antibody W3/13 (a rat "pan T" cell marker) and ascertain the mechanism of the lymphocytopenia of the BB W3/25 (reacts with a subset of T-cells which mediates rat, including characterization of minor T-cell subsets using "help" for antibody reaponse and graft vs. host reactions). other monoclonal antibodies, should help in defining the As shown in Table 1, experiment 2, the non-B lymphocyte pathogenic significance of their severe lymphocytopenia. fraction was approximately 90% T-cells as indicated by reaction with antibody W3/13 for both control rat and BB ACKNOWLEDGMENTS Wistar rats. Contrasting with the similar percentage of cells The expert technical assistance of Hersh Chopra and the positive for antibody W3/13, 84% of control rat T-cells secretarial assistance of Rena Wethington are greatly ap- (W3/13 positive) were positive for W3/25 while only 8% of preciated. Dr. Eisenbarth is a recipient of a Basil O'Connor BB Wistar T-cells were W3/25 positive. Starter Grant from the National Foundation and a Career De- velopment Award from the Juvenile Diabetes Foundation. DISCUSSION Supported in part by grant Am 25778-01 from the National Our findings indicate that in addition to their development of Institutes of Health. We thank Dr. P. Thibert, the Animal Re- diabetes mellitus, BB rats exhibit severe non-B-cell lympho- sources Division, the Health Protection Branch of Health cytopenia. This severe non-B-cell lymphocytopenia is ac- and Welfare, and Canada for provision of BB and control counted for by a decrease in the W3/25 T-cell subset. This Wistar rats. major T-cell subset,5 made up of roughly 80% of control Wistar T-cells, comprises only approximately 8% of BB T- REFERENCES cells. W3/25 T-cell lymphocytopenia can account for all of 1 Nakhooda, A. F., Like, A. A., Chappel, C. I., Murray, F. T., and Marliss, the lymphocytopenia of the BB rat. The function of W3/25 E. B.: The spontaneously diabetic Wistar rat: metabolic and morphologic studies. Diabetes 26:100-12, 1977. cells has not been completely defined, but these cells are 2 Nakhooda, A. F., Like A. A., Chappel, C. I., Wei, C. N., and Marliss, E. known to mediate "help" in the production of anti-hapten an- B.: The spontaneously diabetic Wistar rat (the "BB" rat). Diabetologia 14: tibodies and graft vs. host reactions. Furthermore, antibody 199-207, 1978. 3 Like, A. A., Rossini, A. A., Guberski, D. L, Appel, M. C, and Williams, W3/25 is able to block the in vitro mixed lymphocyte reac- R. M.: Spontaneous diabetes mellitus: reversal and prevention in the BB/W rat tion.6 with antiserum to rat lymphocytes. Science 206:1421-23, 1979. 4 Like, A. A., Williams, R. M., Kislauskis, E., and Rossini, A. A.: Neo- All BB rats we have studied express this severe T-cell natal thymectomy prevents spontaneous diabetes in biobreeding/Worcester lymphocytopenia, which is independent of the development (BB/w) rat. Clin. Res. 29:542, 1981. Abstract. 5 of overt diabetes mellitus. It has been reported that some BB Williams, A. F., Galfre, G., and Milstein, C: Analysis of cell surfaces by xenogeneic myeloma-hybrid antibodies: differentiation antigens of rat rats develop both abnormal glucose tolerance and insulitis lymphocytes. Cell 72:663-73, 1977. in the absence of overt diabetes mellitus.7 If all BB rats do 6 Webb, M., Mason, D. W., and Williams, A. F.: Inhibition of mixed lymphocyte response by monoclonal antibody specific for a rat T lymphocyte not have some degree of islet cell pathology, then the ab- subset. Nature 282:841-43, 1979. normality of lymphocyte subsets we have discovered is ei- 7 Rossini, A. A., Williams, R. M., Appel, M. C, and Like, A. A.: Animal ther independent of the development of diabetes or, more models of type I diabetes. In of Diabetes. Irvine, W. J., Ed. Edin- burgh, Teviot Scientific Publications, 1980. likely, predisposes animals to the development of overt dis- ' Jackson, R., Bowrings, M., Morris, M., Haynes, B., and Eisenbarth, G. ease, with other factors (genetic or environmental) trigger- S.: Increased circulating la positive T cells in recent onset Graves' disease ing severe /3-cell destruction. and insulin dependent diabetes. 63rd Endocrine Society, Abstr. 450, 1981. 9 Fauci, A. S., Steinberg, A. D., Haynes, B. F., and Whalen, G.: Immu- The abnormalities of lymphocytes we find in the BB rat are noregulatory aberrations in systemic lupus erythematosus. J. Immunol. 727:1473-81, 1977. not analogous to our studies to date in type I diabetes mel- 10 8 Eisenbarth, G. S., and Crump, M. A.: Anti-islet antibodies in patients litus. In this illness, a specific T-cell subset bearing the la with polyclonal B-cell activation: mononucleosis, lupus erythematosus, and antigen is markedly increased at the onset of diabetes with- . Diabetes 30(Suppl.):260, 1981. 11 out a T-cell lymphocytopenia.8 T-cell lymphocytopenia is, Kolb, H., Kiesel, U., and Freytag, G.: Spontaneous autoimmune reactions against pancreatic islets in autoimmune mouse strains. 4th Interna- however, a concomitant of other human autoimmune dis- tional Congress of Immunology, Abstr. 323, 1980.

DIABETES, VOL. 30, OCTOBER 1981 889