Veterinary Immunology and Immunopathology, 32 (1992) 195-203
195
Elsevier Science Publishers B.V., Amsterdam
A qualitative assay for beta cell antibodies. Preliminary results in dogs with diabetes mellitus* Margarethe Hoenig a and D.L. Dawe b aDepartment of Physiology and Pharmacology and Department of Small Animal Medicine, College of Veterinary Medicine, University of Georgia, Athens, GA 30602, USA ~Department of Medical Microbiology, College of Veterinary Medicine, University of Georgia, Athens, GA 30602, USA (Accepted 17 June 1991 )
ABSTRACT Hoenig, M. and Dawe, D.L., 1992. A qualitative assay for beta cell antibodies. Preliminary results in dogs with diabetes mellitus. Vet. lmmunol, lmmunopathol., 32: 195-203. Purified beta cells from a radiation-induced transplantable rat insulinoma were used to detect beta cell antibodies in serum from untreated diabetic dogs. Serum from dogs in which anti-beta cell antibodies were induced by injecting a purified beta cell suspension subcutaneously was used as positive control. Following incubation with test sera, fluorescein-labeled anti-dog immunoglobulins were used to visualize binding between the beta cells and dog gamma globulins. Nine of the 23 diabetic dogs showed a strongly positive reaction which was characterized by a ring fluorescence, three showed a weak reaction and 11 were negative, i.e. they showed diffuse fluorescence. In contrast, 14 of the 15 healthy dogs showed diffuse fluorescence and one dog showed a weakly positive reaction. Thyroid, liver and kidney cells did not elicit ring fluorescence. Although females (spayed and intact) represented the majority of the diabetic dogs, there was no correlation between sex and the occurrence of antibodies in the diabetic dogs. There was also no correlation to the age of the dogs. In conclusion, we have developed a specific test for anti-beta cell antibodies. The test is reproducible and economical to perform on a large number of samples. ABBREVIATIONS IDDM, insulin-dependent diabetes mellitus; KRB, Krebs-Ringer bicarbonate medium; KRBSA, KRB supplemented with sodium azide.
INTRODUCTION
Spontaneous diabetes mellitus has a relatively c o m m o n occurrence in the dog; however, its pathogenesis is still unknown. It is thought that autoim"Supported in part by a grant from the Veterinary Medical Experimental Station, University of Georgia, Athens, GA 30602, USA.
© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-2427/92/$05.00
196
M. H O E N I G A N D D.L. I)AWE
mune mechanisms and environmental factors play a role in the initiation of the disease process and in the precipitation of overt disease in the insulindependent diabetic (Maclaren et al., 1989). A majority of newly diagnosed human diabetic patients have antibodies directed against beta cells (Lendrum et al., 1975; Lernmark et al., 1978 ). Antibodies are also detected in the prediabetic period and are used as a screening method in siblings or offspring of diabetic patients to predict the disease (Gorsuch et al., 1981; Baekkeskov et al., 1984, 1987; Dyrberg et al., 1984). These antibodies are organ-specific but species non-specific (Lernmark et al., 1978). Taking advantage of the species non-specificity, a test was developed to detect antibodies against beta cells in the dog and it was used to screen sera from healthy and diabetic dogs for the occurrence of anti-beta cell antibodies. MATERIALSAND METHODS
Preparation of purified beta cells A suspension of purified beta cells from a glucose-responsive insulinoma grown underneath the kidney capsule of male New England Deaconess Hospital ( N E D H ) rats (Hoenig et al., 1984) was used as tissue source for the induction of anti-beta cell antibodies. The tumor tissue was digested as previously described (Hoenig and Sharp, 1986 ) and the beta cells were then separated from the other cells by centrifugation ( 7 0 0 × g , 30 min) on four identical discontinuous Ficoll gradients. The purified beta cells were harvested, consolidated into a large tube and washed twice (40 ml each) with Hanks' buffer to remove the Ficoll (Hoenig and Sharp, 1986 ).
Preparation of anti-beta cell antibodies One milliliter each of cell suspension containing l0 s purified beta cells was injected subcutaneously in the flank area of two healthy male dogs. A second injection was made 2 weeks later. Three weeks later blood was taken for the determination of anti-beta cell antibodies. It was centrifuged and serum collected and frozen at - 20 ° C.
Serum samples and preparation Serum from 15 healthy adult dogs (seven spayed females and eight intact males) and from 23 dogs with newly diagnosed diabetes mellitus, which had not yet been treated with insulin according to the owners and the veterinarians diagnosing the disease (Table 1 ), w a s used. Also examined was serum from two dogs with proven insulinomas. Serum from dogs before and after inoculation with rat insulinoma cells (see above) was used as negative con-
ASSAYFOR BETACELLANTIBODIESIN DIABETICDOGS
197
TABLE 1 Age, breed and sex and detection of anti-beta cell antibodies of 23 dogs with untreated diabetes mellitus Dog No.
Breed
Age
Sex
Antibodies
I 2 3 4 5 6 7 8 9 I0 I1 12 13 14 15 16 17 18 19 20 21 22 23
Samoyed Mixbreed Dachshund Terrier Poodle Poodle Siberian Huskey Mixbreed Basset Hound Poodle Rottweiler Min. Schnauzer Labrador Airedale Sheltie Min. Schnauzer Min. Schnauzer Doberman Poodle Mixbreed Beagle Mixbreed Mixbreed
7 years 8 years 9 years 13 years 7 years 15 years I 0 years 12 weeks 3 years 7 years 6 years 6 years 8 years 1 year 3 years 6 years 5 years 5 years 9 years 8 years 9 years 11 years I 1 years
F/S F/S F/S M/C F/S F/S F F F/S F F/S F/S F/S M F/S M M/C M F/S F/S F/S M F/S
Positive Weak positive Negative Negative Negative Negattve Positive Positive Negative Negative Negative Negative Weak positive Positive Positive Positive Weak positive Negative Negative Positive Positive Negative Positive
F/S, spayed female; M/C, male castrated; F, female; M, male.
trol and positive control, respectively. All serum samples were heat inactivated (56 °C for 20 min) and successively absorbed with red blood cells from NEDH rats ( 1 h at room temperature) and with mouse liver powder (4% w/v at 4°C for 20 min; Sigma, St. Louis, MO). Part of the positive control serum was used for the preparation of immunoglobulins as described (Van de Winkel et al., 1982 ). The IgG fraction was separated from other proteins by ion exchange chromatography using diethylaminoethyl (DEAE) cellulose. In order to ensure that the antibodies were directed against beta cells and not insulin, serum was incubated for 24 h with ~251-insulin (porcine) and separated according to the method described by Herbert et al. ( 1965 ). The serum samples examined were from the two dogs used for antibody production before and after inoculation with beta cells, from three healthy controls, from four dogs positive for anti-beta cell antibody and from one dog with proven antibodies against bovine/porcine insulin. There was no significant decrease in the bound insulin fraction in all dogs except for the one with proven insulin antibodies (non-specific binding was 11.7%; the dog with antibodies showed 79.5% binding and the binding of the other sera ranged from 8.3 to 14.5%).
198
M. tt()ENI(; A N D I).1. I)~.WE
Mernbrane fluorescence ().[beta cells A beta cell suspension was prepared as described above. Before use, the cells were washed with Hepes-buffered Krebs-Ringer bicarbonate medium without calcium (KRB), supplemented with 0.02 g 100 ml-~ sodium azide and 1 g 100 m l - i bovine serum albumin. Samples of 10 ~ cells were then suspended in 100/A KRB supplemented with 0.02 g 100 m l - t sodium azide (KRB-SA) and incubated with 100/~1 control or patient serum or they were suspended in 150/~1 KRB-SA and incubated with 50/~1 of the IgG fraction for 2 h at 4°C with constant shaking. After washing the cells twice in KRB-SA, the cell pellet was suspended in 400 1 KRB-SA and incubated for another 30 min at 4°C with fluorescein-labeled anti-dog lgG (ICN, Costa Mesa, CA) at a final dilution of 1: 100. The cells were then washed twice in KRB-SA and observed with an ultraviolet light microscope with epi-illumination (Leitz Dialux 29, Wetzlar, Germany). Each serum was tested on two different tumor cell preparations. RESULTS
Figure 1 shows the fluorescence seen in beta cells incubated with negative control serum. The cells exhibit a homogeneous staining pattern (diffuse fluorescence). In marked contrast is the pattern seen in cells incubated with positive control serum or the IgG fraction of positive control serum. The staining has a freckled pattern and is seen at the plasma membrane (ring fluorescence, Fig. 2). The ring fluorescence seems to be specific for beta cells as it was not seen when positive control serum was incubated with thyroid, liver and kidney cells (not shown ). Comparing serum from 15 healthy dogs and from 23 untreated diabetic dogs with negative or positive control serum, it was found that nine of the 23 diabetic dogs had a strong ring fluorescence, i.e. most of the cells examined showed the ring pattern and three of the 23 were weakly positive (i.e. few of the cells examined showed the ring pattern). Thus, a majority of the 23 diabetic dogs had detectable antibodies to beta cells. Figure 3 shows insulinoma cells incubated with serum from Dog No. 14. In contrast, 14 out of the 15 healthy dogs showed diffuse fluorescence and one dog showed a weakly positive reaction. Diffuse fluorescence was seen in two dogs with insulinomas. Although females (spayed and intact) represented the majority of the diabetic dogs, there was no correlation between sex and the occurrence of antibodies in the diabetic dogs. There was also no correlation to the age of the dogs. The average age of the 12 diabetic dogs with a positive antibody reaction was 6.4_ 1.0 years and 8.3 _+ 1.0 years (mean _ SEM ) for the diabetic dogs which were negative.
ASSAYFOR BETACELLANTIBODIESIN DIABETICDOGS
199
Fig. 1. Indirect fluorescence of insulinoma cells incubated with serum from a healthy dog. × 160.
Fig. 2. Indirect fluorescence of insulinoma cells incubated with serum from dogs immunized against beta cells. X 250.
200
M. ItOENI(J AND I).L. [)AWE
Fig. 3. Indirect fluorescence ofinsulinoma cells incubated with serum from a diabetic dog (No. 14, Table 1 ). × 400. DISCUSSION
Although the pathogenesis of diabetes mellitus in man and animals remains unknown, it is likely that not one but several factors play a role. In man, it is thought that insulin-dependent diabetes mellitus ( I D D M ) is an a u t o i m m u n e disease modulated by environmental factors (Maclaren et al., 1989 ). IDDM in humans does not have an acute onset but is characterized by a latent period of several months to years which precedes the onset of diabetes (Gorsuch et al., 1981; Dyrberg et al., 1984; Baekkeskov et al., 1987). IDDM is characterized by lymphocytic infiltration of the islets and formation of antibodies against beta cells (Eisenbarth, 1986). Specific loss of beta cells ensues and insulin secretion becomes inadequate to control glucose homeostasis (Ganda et al., 1984). The nature of the target for the autoantibodies in the beta cell is unclear. Several different antibodies are found in the blood of human diabetics. They are directed against cytoplasmic components (Botazzo et al., 1974; Lendrum et al., 1975), insulin (Dean et al., 1986), proinsulin (Kuglin et al., 1990) or a 64 kDa protein which has recently been identified as the GABA-synthesizing enzyme glutamic acid decarboxylase (Baekkeskov et al., 1984, 1987, 1990). It is currently thought that the 64 kDa protein might be
ASSAY FOR BETA CELL ANTIBODIES IN DIABE-iIC DOGS
201
the molecular target. Antibodies in serum from insulin-dependent diabetics specifically immunoprecipitate this protein (Baekkeskov et al., 1989). Autoantibodies to this antigen are present in a majority of diabetics at the time of onset of the disease and have been found in many prediabetics (Baekkeskov et al., 1987, 1989; Maclaren et al., 1989). Antibodies have also been found in animal models of the disease (Baekkeskov et al., 1984; Atkinson and Maclaren, 1988). The antibodies to this protein seem to be organ specific and have not been found in people with other autoimmune or endocrine disorders. Serum or immunoglobulin fractions containing autoantibodies from insulin-dependent diabetics suppress glucose induced insulin release in vitro from mouse, human and rat beta cells (Baekkeskov et al., 1982, 1987). This species non-specificity has been applied in the current study. Our test results also confirm the organ-specificity of these antibodies seen by other investigators (Lernmark et al., 1978; Van De Winkel et al., 1982; Baekkeskov et al., 1989 ) as only beta cells but neither thyroid, liver nor kidney cells exhibited ring fluorescence when exposed to antibody containing serum. Haines and Penhale ( 1985 ) demonstrated anti-islet antibodies by an indirect fluorescence method in dogs with diabetes mellitus and other endocrine diseases by incubating normal canine pancreatic tissue with serum from healthy controls and from dogs with different endocrine disorders. While no fluorescence was seen in the healthy dogs, positive fluorescence was seen in some diabetics and dogs with other endocrine disorders. However, the specificity of the positive sera for pancreatic tissue was not examined; neither was the specificity of the positive sera for beta cells tested. Our results show that about 50% of the diabetic dogs have antibodies against beta cells. This may imply that autoimmune phenomena play a role in the pathogenesis of diabetes in the dog. It is, however, possible that they are a mere consequence of beta cell destruction. It is unknown at this time what the nature of the target for the antibodies is and further studies are needed to examine this important question. The antibody is not directed against insulin. The staining pattern in the positive cells suggests that the antibody is directed against a cell membrane associated molecule (Fig. 2 ) but this needs to be confirmed in studies using purified plasma membranes from the insulinoma, a method which has been developed in our laboratory (Hoenig et al., 1989). The size of the target molecule also needs to be examined and compared with findings in the human diabetic. In conclusion, we have developed a specific test for anti-beta cell antibodies. The test is reproducible and economical to perform on a large number of serum samples. However, it is evident that many questions remain unanswered at this time. Further studies are needed to determine the target for the beta cell antibodies, the time course of antibody production and its relationship to the diabetic state. Ultimately, future studies will be designed to eluci-
202
M. H()ENI(.i ,~,NI)ILL. I)AWE
date whether our findings indicate heterogeneity of the pathogenesis of diabetes mellitus in the dog. ACKNOWLEDGMENTS
The authors thank Drs. K.M. Comer, S. Forbes, M.E. Peterson, R.W. Nelson, and K.B. Spencer for providing serum samples from diabetic dogs.
REFERENCES Atkinson, M. and Maclaren, N.K., 1988. Autoantibodies in nonobese diabetic mice immunoprecipitate 64,000-Mr islet antigen. Diabetes, 37:1587-1590. Baekkeskov. S., Nielsen, J.H., Marner, B., Bilde, T., Ludvigsson, J. and Lernmark, A.. 1982. Autoantibodies in newly diagnosed diabetic children immunoprecipitate human pancreatic islet proteins. Nature, 298:167-169. Baekkeskov, S., Dyrberg, T. and Lernmark, A., 1984. Autoantibodies to a 64-kilodalton islet cell protein precede the onset of spontaneous diabetes in the BB rat. Science, 224: 13481350. Baekkeskov. S., Landin, M., Kristenscn, J.K., Srikanta, S., Briuning, G.J., Mandrup-Poulsen. T., de Beaufort, C., Soeldner, J.S., Eisenbarth, G., Lindgren, F., Sundquist. G. and Lcrnmark, A.. 1987. Antibodies to a 64,000 M human islet cell antigen precede the clinical onset of insulin-dependent diabetes. J. Clin, Invest., 79: 926-934. Baekkcskov, S., Warnock, G., Christie, M., Rajotte, R.V., Larsen, P.M. and Frey, S., 1989. Revelation of specificity of 64K autoantibodies in IDDM serums by high-resolution 2-D gel electrophoresis. Unambiguous identification of 64 K target antigen. Diabetes, 38: 1133-1141. Baekkeskov, S., Aanstoot, H.-J., Christgau, S., Reetz, A., Solimcna, M., Cascalho, M., Folli. F.. Richter-Olesen, H. and Camilli, P.-D., 1990. Identification of the 64 K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylasc. Nature, 347:151 - 156. Botazzo, G.F., Florin-Christensen, A. and Doniach, D., 1974. Islet cell antibodies in diabetes mellitus with autoimmune polycndocrinc deficiencies. Lancet, 2:1279-1282. Dean, B.M., Becker, J.M., McNally, A.C., Tarn, C., Schwartz, G., Gale, A.M. and Botazzo, G.F., 1986. Insulin autoantibodies in the prediabetic period: Correlation with islet cell antibodies and development of diabetes. Diabetologia, 29: 339-342. Dyrberg, T., Poussier, P., Nakhooda, F., Marliss, E,B. and Lcrnmark, A.. 1984. Islet cell surface and lymphocyte antibodies often precede the spontaneous diabetes in the BB rat. DiabetoIogia, 26: 159-165. Eiscnbarth, G., 1986. Insulin-dependent diabetes mellitus: a chronic autoimmunc disease. N. Engl. J. Med., 314: 1360-1368. Ganda, O.P., Srikanta, S., Brink, S.J., Morris, M.A., Gleason, R.E., Soeldncr, J.S. and Eiscnbarth, G.S., 1984. Differential sensitivity to beta cell secretagogues in "early", type I diabetes mellitus. Diabetes, 33: 516-521. Gorsuch, A.N., Lister, J., Dean, B.M., Spencer, K.M., McNally, J.M., Botazzo, G.F. and Cudsworth, A.G., 1981. Evidence for a long prediabetic period in type I (insulin-dependent) diabetes mellitus. Lancet, 2: 1363-1365. Haines. D.M. and Penhale, W.J., 1985. Autoantibodies to pancreatic islet cells in canine diabetes mellitus. Vet. Immunol. lmmunopathol., 8:149-156.
ASSAY FOR BETA CELL ANTIBODIES IN DIABETIC DOGS
203
Herbert, V., Lau, K. and Gottlieb, C.W., 1965. Coated charcoal immunoassay of insulin. J. Clin. Endocrinol. Metab., 25:1375-1384. Hoenig, M. and Sharp, G.W.G., 1986. Glucose induces insulin release and a rise in cytosolic calcium concentration in a transplantable rat insulinoma. Endocrinology, 119: 2502-2507. Hoenig, M., Ferguson, D.C. and Matschinsky, F.M., 1984. Fuel-induced insulin release in vitro from insulinomas transplanted into the rat kidney. Diabetes, 33: 1-7. Hocnig, M., Russo, L.L. and Ferguson, D.C., 1989. Characterization of calcium channels of a glucose-responsive rat insulinoma. Am. J. Physiol., 256: E488-E493. Kuglin, B., Halder, B., Benrams, J., Griinekle¢, D., Gries, F.A. and Kolb, H., 1990. Proinsulin autoantibodies: Association with type I diabetes but not with islet cell antibodies, insulin autoantibodies or HLA-DR type. J. Autoimmunol., 3: 573-577. Lendrum, R., Walker, G. and Gamble, D.R., 1975. Islet-cell antibodies in juvenile diabetes mellitus of recent onset. Lancet, 1: 880-883. Lernmark, A., Freedman, Z.R., Hofman, C., Rubenstein, A.H., Steiner, D.F., Jackson, R.L., Winter, R.J. and Traisman, H.S., 1978. Islet-cell-surface antibodies in juvenile diabetes mellitus. N. Engl. J. Med., 299: 375-380. Maclaren, N., Schatz, D., Drash, A. and Grave, G., 1989. Initial pathogenic events in IDDM. Diabetes, 38: 534-538. Van De Winkel, M., Smets, G., Gepts, W. and Pipeleers, D., 1982. Islet cell surface antibodies from insulin-dependent diabetics bind specifically to pancreatic B cells. J. Clin. Invest., 70: 41-49.
195
Elsevier Science Publishers B.V., Amsterdam
A qualitative assay for beta cell antibodies. Preliminary results in dogs with diabetes mellitus* Margarethe Hoenig a and D.L. Dawe b aDepartment of Physiology and Pharmacology and Department of Small Animal Medicine, College of Veterinary Medicine, University of Georgia, Athens, GA 30602, USA ~Department of Medical Microbiology, College of Veterinary Medicine, University of Georgia, Athens, GA 30602, USA (Accepted 17 June 1991 )
ABSTRACT Hoenig, M. and Dawe, D.L., 1992. A qualitative assay for beta cell antibodies. Preliminary results in dogs with diabetes mellitus. Vet. lmmunol, lmmunopathol., 32: 195-203. Purified beta cells from a radiation-induced transplantable rat insulinoma were used to detect beta cell antibodies in serum from untreated diabetic dogs. Serum from dogs in which anti-beta cell antibodies were induced by injecting a purified beta cell suspension subcutaneously was used as positive control. Following incubation with test sera, fluorescein-labeled anti-dog immunoglobulins were used to visualize binding between the beta cells and dog gamma globulins. Nine of the 23 diabetic dogs showed a strongly positive reaction which was characterized by a ring fluorescence, three showed a weak reaction and 11 were negative, i.e. they showed diffuse fluorescence. In contrast, 14 of the 15 healthy dogs showed diffuse fluorescence and one dog showed a weakly positive reaction. Thyroid, liver and kidney cells did not elicit ring fluorescence. Although females (spayed and intact) represented the majority of the diabetic dogs, there was no correlation between sex and the occurrence of antibodies in the diabetic dogs. There was also no correlation to the age of the dogs. In conclusion, we have developed a specific test for anti-beta cell antibodies. The test is reproducible and economical to perform on a large number of samples. ABBREVIATIONS IDDM, insulin-dependent diabetes mellitus; KRB, Krebs-Ringer bicarbonate medium; KRBSA, KRB supplemented with sodium azide.
INTRODUCTION
Spontaneous diabetes mellitus has a relatively c o m m o n occurrence in the dog; however, its pathogenesis is still unknown. It is thought that autoim"Supported in part by a grant from the Veterinary Medical Experimental Station, University of Georgia, Athens, GA 30602, USA.
© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-2427/92/$05.00
196
M. H O E N I G A N D D.L. I)AWE
mune mechanisms and environmental factors play a role in the initiation of the disease process and in the precipitation of overt disease in the insulindependent diabetic (Maclaren et al., 1989). A majority of newly diagnosed human diabetic patients have antibodies directed against beta cells (Lendrum et al., 1975; Lernmark et al., 1978 ). Antibodies are also detected in the prediabetic period and are used as a screening method in siblings or offspring of diabetic patients to predict the disease (Gorsuch et al., 1981; Baekkeskov et al., 1984, 1987; Dyrberg et al., 1984). These antibodies are organ-specific but species non-specific (Lernmark et al., 1978). Taking advantage of the species non-specificity, a test was developed to detect antibodies against beta cells in the dog and it was used to screen sera from healthy and diabetic dogs for the occurrence of anti-beta cell antibodies. MATERIALSAND METHODS
Preparation of purified beta cells A suspension of purified beta cells from a glucose-responsive insulinoma grown underneath the kidney capsule of male New England Deaconess Hospital ( N E D H ) rats (Hoenig et al., 1984) was used as tissue source for the induction of anti-beta cell antibodies. The tumor tissue was digested as previously described (Hoenig and Sharp, 1986 ) and the beta cells were then separated from the other cells by centrifugation ( 7 0 0 × g , 30 min) on four identical discontinuous Ficoll gradients. The purified beta cells were harvested, consolidated into a large tube and washed twice (40 ml each) with Hanks' buffer to remove the Ficoll (Hoenig and Sharp, 1986 ).
Preparation of anti-beta cell antibodies One milliliter each of cell suspension containing l0 s purified beta cells was injected subcutaneously in the flank area of two healthy male dogs. A second injection was made 2 weeks later. Three weeks later blood was taken for the determination of anti-beta cell antibodies. It was centrifuged and serum collected and frozen at - 20 ° C.
Serum samples and preparation Serum from 15 healthy adult dogs (seven spayed females and eight intact males) and from 23 dogs with newly diagnosed diabetes mellitus, which had not yet been treated with insulin according to the owners and the veterinarians diagnosing the disease (Table 1 ), w a s used. Also examined was serum from two dogs with proven insulinomas. Serum from dogs before and after inoculation with rat insulinoma cells (see above) was used as negative con-
ASSAYFOR BETACELLANTIBODIESIN DIABETICDOGS
197
TABLE 1 Age, breed and sex and detection of anti-beta cell antibodies of 23 dogs with untreated diabetes mellitus Dog No.
Breed
Age
Sex
Antibodies
I 2 3 4 5 6 7 8 9 I0 I1 12 13 14 15 16 17 18 19 20 21 22 23
Samoyed Mixbreed Dachshund Terrier Poodle Poodle Siberian Huskey Mixbreed Basset Hound Poodle Rottweiler Min. Schnauzer Labrador Airedale Sheltie Min. Schnauzer Min. Schnauzer Doberman Poodle Mixbreed Beagle Mixbreed Mixbreed
7 years 8 years 9 years 13 years 7 years 15 years I 0 years 12 weeks 3 years 7 years 6 years 6 years 8 years 1 year 3 years 6 years 5 years 5 years 9 years 8 years 9 years 11 years I 1 years
F/S F/S F/S M/C F/S F/S F F F/S F F/S F/S F/S M F/S M M/C M F/S F/S F/S M F/S
Positive Weak positive Negative Negative Negative Negattve Positive Positive Negative Negative Negative Negative Weak positive Positive Positive Positive Weak positive Negative Negative Positive Positive Negative Positive
F/S, spayed female; M/C, male castrated; F, female; M, male.
trol and positive control, respectively. All serum samples were heat inactivated (56 °C for 20 min) and successively absorbed with red blood cells from NEDH rats ( 1 h at room temperature) and with mouse liver powder (4% w/v at 4°C for 20 min; Sigma, St. Louis, MO). Part of the positive control serum was used for the preparation of immunoglobulins as described (Van de Winkel et al., 1982 ). The IgG fraction was separated from other proteins by ion exchange chromatography using diethylaminoethyl (DEAE) cellulose. In order to ensure that the antibodies were directed against beta cells and not insulin, serum was incubated for 24 h with ~251-insulin (porcine) and separated according to the method described by Herbert et al. ( 1965 ). The serum samples examined were from the two dogs used for antibody production before and after inoculation with beta cells, from three healthy controls, from four dogs positive for anti-beta cell antibody and from one dog with proven antibodies against bovine/porcine insulin. There was no significant decrease in the bound insulin fraction in all dogs except for the one with proven insulin antibodies (non-specific binding was 11.7%; the dog with antibodies showed 79.5% binding and the binding of the other sera ranged from 8.3 to 14.5%).
198
M. tt()ENI(; A N D I).1. I)~.WE
Mernbrane fluorescence ().[beta cells A beta cell suspension was prepared as described above. Before use, the cells were washed with Hepes-buffered Krebs-Ringer bicarbonate medium without calcium (KRB), supplemented with 0.02 g 100 ml-~ sodium azide and 1 g 100 m l - i bovine serum albumin. Samples of 10 ~ cells were then suspended in 100/A KRB supplemented with 0.02 g 100 m l - t sodium azide (KRB-SA) and incubated with 100/~1 control or patient serum or they were suspended in 150/~1 KRB-SA and incubated with 50/~1 of the IgG fraction for 2 h at 4°C with constant shaking. After washing the cells twice in KRB-SA, the cell pellet was suspended in 400 1 KRB-SA and incubated for another 30 min at 4°C with fluorescein-labeled anti-dog lgG (ICN, Costa Mesa, CA) at a final dilution of 1: 100. The cells were then washed twice in KRB-SA and observed with an ultraviolet light microscope with epi-illumination (Leitz Dialux 29, Wetzlar, Germany). Each serum was tested on two different tumor cell preparations. RESULTS
Figure 1 shows the fluorescence seen in beta cells incubated with negative control serum. The cells exhibit a homogeneous staining pattern (diffuse fluorescence). In marked contrast is the pattern seen in cells incubated with positive control serum or the IgG fraction of positive control serum. The staining has a freckled pattern and is seen at the plasma membrane (ring fluorescence, Fig. 2). The ring fluorescence seems to be specific for beta cells as it was not seen when positive control serum was incubated with thyroid, liver and kidney cells (not shown ). Comparing serum from 15 healthy dogs and from 23 untreated diabetic dogs with negative or positive control serum, it was found that nine of the 23 diabetic dogs had a strong ring fluorescence, i.e. most of the cells examined showed the ring pattern and three of the 23 were weakly positive (i.e. few of the cells examined showed the ring pattern). Thus, a majority of the 23 diabetic dogs had detectable antibodies to beta cells. Figure 3 shows insulinoma cells incubated with serum from Dog No. 14. In contrast, 14 out of the 15 healthy dogs showed diffuse fluorescence and one dog showed a weakly positive reaction. Diffuse fluorescence was seen in two dogs with insulinomas. Although females (spayed and intact) represented the majority of the diabetic dogs, there was no correlation between sex and the occurrence of antibodies in the diabetic dogs. There was also no correlation to the age of the dogs. The average age of the 12 diabetic dogs with a positive antibody reaction was 6.4_ 1.0 years and 8.3 _+ 1.0 years (mean _ SEM ) for the diabetic dogs which were negative.
ASSAYFOR BETACELLANTIBODIESIN DIABETICDOGS
199
Fig. 1. Indirect fluorescence of insulinoma cells incubated with serum from a healthy dog. × 160.
Fig. 2. Indirect fluorescence of insulinoma cells incubated with serum from dogs immunized against beta cells. X 250.
200
M. ItOENI(J AND I).L. [)AWE
Fig. 3. Indirect fluorescence ofinsulinoma cells incubated with serum from a diabetic dog (No. 14, Table 1 ). × 400. DISCUSSION
Although the pathogenesis of diabetes mellitus in man and animals remains unknown, it is likely that not one but several factors play a role. In man, it is thought that insulin-dependent diabetes mellitus ( I D D M ) is an a u t o i m m u n e disease modulated by environmental factors (Maclaren et al., 1989 ). IDDM in humans does not have an acute onset but is characterized by a latent period of several months to years which precedes the onset of diabetes (Gorsuch et al., 1981; Dyrberg et al., 1984; Baekkeskov et al., 1987). IDDM is characterized by lymphocytic infiltration of the islets and formation of antibodies against beta cells (Eisenbarth, 1986). Specific loss of beta cells ensues and insulin secretion becomes inadequate to control glucose homeostasis (Ganda et al., 1984). The nature of the target for the autoantibodies in the beta cell is unclear. Several different antibodies are found in the blood of human diabetics. They are directed against cytoplasmic components (Botazzo et al., 1974; Lendrum et al., 1975), insulin (Dean et al., 1986), proinsulin (Kuglin et al., 1990) or a 64 kDa protein which has recently been identified as the GABA-synthesizing enzyme glutamic acid decarboxylase (Baekkeskov et al., 1984, 1987, 1990). It is currently thought that the 64 kDa protein might be
ASSAY FOR BETA CELL ANTIBODIES IN DIABE-iIC DOGS
201
the molecular target. Antibodies in serum from insulin-dependent diabetics specifically immunoprecipitate this protein (Baekkeskov et al., 1989). Autoantibodies to this antigen are present in a majority of diabetics at the time of onset of the disease and have been found in many prediabetics (Baekkeskov et al., 1987, 1989; Maclaren et al., 1989). Antibodies have also been found in animal models of the disease (Baekkeskov et al., 1984; Atkinson and Maclaren, 1988). The antibodies to this protein seem to be organ specific and have not been found in people with other autoimmune or endocrine disorders. Serum or immunoglobulin fractions containing autoantibodies from insulin-dependent diabetics suppress glucose induced insulin release in vitro from mouse, human and rat beta cells (Baekkeskov et al., 1982, 1987). This species non-specificity has been applied in the current study. Our test results also confirm the organ-specificity of these antibodies seen by other investigators (Lernmark et al., 1978; Van De Winkel et al., 1982; Baekkeskov et al., 1989 ) as only beta cells but neither thyroid, liver nor kidney cells exhibited ring fluorescence when exposed to antibody containing serum. Haines and Penhale ( 1985 ) demonstrated anti-islet antibodies by an indirect fluorescence method in dogs with diabetes mellitus and other endocrine diseases by incubating normal canine pancreatic tissue with serum from healthy controls and from dogs with different endocrine disorders. While no fluorescence was seen in the healthy dogs, positive fluorescence was seen in some diabetics and dogs with other endocrine disorders. However, the specificity of the positive sera for pancreatic tissue was not examined; neither was the specificity of the positive sera for beta cells tested. Our results show that about 50% of the diabetic dogs have antibodies against beta cells. This may imply that autoimmune phenomena play a role in the pathogenesis of diabetes in the dog. It is, however, possible that they are a mere consequence of beta cell destruction. It is unknown at this time what the nature of the target for the antibodies is and further studies are needed to examine this important question. The antibody is not directed against insulin. The staining pattern in the positive cells suggests that the antibody is directed against a cell membrane associated molecule (Fig. 2 ) but this needs to be confirmed in studies using purified plasma membranes from the insulinoma, a method which has been developed in our laboratory (Hoenig et al., 1989). The size of the target molecule also needs to be examined and compared with findings in the human diabetic. In conclusion, we have developed a specific test for anti-beta cell antibodies. The test is reproducible and economical to perform on a large number of serum samples. However, it is evident that many questions remain unanswered at this time. Further studies are needed to determine the target for the beta cell antibodies, the time course of antibody production and its relationship to the diabetic state. Ultimately, future studies will be designed to eluci-
202
M. H()ENI(.i ,~,NI)ILL. I)AWE
date whether our findings indicate heterogeneity of the pathogenesis of diabetes mellitus in the dog. ACKNOWLEDGMENTS
The authors thank Drs. K.M. Comer, S. Forbes, M.E. Peterson, R.W. Nelson, and K.B. Spencer for providing serum samples from diabetic dogs.
REFERENCES Atkinson, M. and Maclaren, N.K., 1988. Autoantibodies in nonobese diabetic mice immunoprecipitate 64,000-Mr islet antigen. Diabetes, 37:1587-1590. Baekkeskov. S., Nielsen, J.H., Marner, B., Bilde, T., Ludvigsson, J. and Lernmark, A.. 1982. Autoantibodies in newly diagnosed diabetic children immunoprecipitate human pancreatic islet proteins. Nature, 298:167-169. Baekkeskov, S., Dyrberg, T. and Lernmark, A., 1984. Autoantibodies to a 64-kilodalton islet cell protein precede the onset of spontaneous diabetes in the BB rat. Science, 224: 13481350. Baekkeskov. S., Landin, M., Kristenscn, J.K., Srikanta, S., Briuning, G.J., Mandrup-Poulsen. T., de Beaufort, C., Soeldner, J.S., Eisenbarth, G., Lindgren, F., Sundquist. G. and Lcrnmark, A.. 1987. Antibodies to a 64,000 M human islet cell antigen precede the clinical onset of insulin-dependent diabetes. J. Clin, Invest., 79: 926-934. Baekkcskov, S., Warnock, G., Christie, M., Rajotte, R.V., Larsen, P.M. and Frey, S., 1989. Revelation of specificity of 64K autoantibodies in IDDM serums by high-resolution 2-D gel electrophoresis. Unambiguous identification of 64 K target antigen. Diabetes, 38: 1133-1141. Baekkeskov, S., Aanstoot, H.-J., Christgau, S., Reetz, A., Solimcna, M., Cascalho, M., Folli. F.. Richter-Olesen, H. and Camilli, P.-D., 1990. Identification of the 64 K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylasc. Nature, 347:151 - 156. Botazzo, G.F., Florin-Christensen, A. and Doniach, D., 1974. Islet cell antibodies in diabetes mellitus with autoimmune polycndocrinc deficiencies. Lancet, 2:1279-1282. Dean, B.M., Becker, J.M., McNally, A.C., Tarn, C., Schwartz, G., Gale, A.M. and Botazzo, G.F., 1986. Insulin autoantibodies in the prediabetic period: Correlation with islet cell antibodies and development of diabetes. Diabetologia, 29: 339-342. Dyrberg, T., Poussier, P., Nakhooda, F., Marliss, E,B. and Lcrnmark, A.. 1984. Islet cell surface and lymphocyte antibodies often precede the spontaneous diabetes in the BB rat. DiabetoIogia, 26: 159-165. Eiscnbarth, G., 1986. Insulin-dependent diabetes mellitus: a chronic autoimmunc disease. N. Engl. J. Med., 314: 1360-1368. Ganda, O.P., Srikanta, S., Brink, S.J., Morris, M.A., Gleason, R.E., Soeldncr, J.S. and Eiscnbarth, G.S., 1984. Differential sensitivity to beta cell secretagogues in "early", type I diabetes mellitus. Diabetes, 33: 516-521. Gorsuch, A.N., Lister, J., Dean, B.M., Spencer, K.M., McNally, J.M., Botazzo, G.F. and Cudsworth, A.G., 1981. Evidence for a long prediabetic period in type I (insulin-dependent) diabetes mellitus. Lancet, 2: 1363-1365. Haines. D.M. and Penhale, W.J., 1985. Autoantibodies to pancreatic islet cells in canine diabetes mellitus. Vet. Immunol. lmmunopathol., 8:149-156.
ASSAY FOR BETA CELL ANTIBODIES IN DIABETIC DOGS
203
Herbert, V., Lau, K. and Gottlieb, C.W., 1965. Coated charcoal immunoassay of insulin. J. Clin. Endocrinol. Metab., 25:1375-1384. Hoenig, M. and Sharp, G.W.G., 1986. Glucose induces insulin release and a rise in cytosolic calcium concentration in a transplantable rat insulinoma. Endocrinology, 119: 2502-2507. Hoenig, M., Ferguson, D.C. and Matschinsky, F.M., 1984. Fuel-induced insulin release in vitro from insulinomas transplanted into the rat kidney. Diabetes, 33: 1-7. Hocnig, M., Russo, L.L. and Ferguson, D.C., 1989. Characterization of calcium channels of a glucose-responsive rat insulinoma. Am. J. Physiol., 256: E488-E493. Kuglin, B., Halder, B., Benrams, J., Griinekle¢, D., Gries, F.A. and Kolb, H., 1990. Proinsulin autoantibodies: Association with type I diabetes but not with islet cell antibodies, insulin autoantibodies or HLA-DR type. J. Autoimmunol., 3: 573-577. Lendrum, R., Walker, G. and Gamble, D.R., 1975. Islet-cell antibodies in juvenile diabetes mellitus of recent onset. Lancet, 1: 880-883. Lernmark, A., Freedman, Z.R., Hofman, C., Rubenstein, A.H., Steiner, D.F., Jackson, R.L., Winter, R.J. and Traisman, H.S., 1978. Islet-cell-surface antibodies in juvenile diabetes mellitus. N. Engl. J. Med., 299: 375-380. Maclaren, N., Schatz, D., Drash, A. and Grave, G., 1989. Initial pathogenic events in IDDM. Diabetes, 38: 534-538. Van De Winkel, M., Smets, G., Gepts, W. and Pipeleers, D., 1982. Islet cell surface antibodies from insulin-dependent diabetics bind specifically to pancreatic B cells. J. Clin. Invest., 70: 41-49.
