Diabetes Research and Clinical Practice, 9 (1990)
l-5
1
Elsevier DIABET 00372
Heterogeneity of islet cell autoantibodies in terms of insulin release from rat islets and insulinoma cells Kazushige
Ejiri ‘, Hiroshi Taniguchi 1,2,5,Yasuo Morimoto ‘, Yuki Yamashiro Shigeaki Baba’, S. Srikanta2,3*4,5 and G. S. Eisenbarth2,3,4,5
‘,
I Second Department of Internal Medicine, Kobe University School of Medicine, Kobe, Japan, ’ Joslin Diabetes Center, Boston, 3 Brigham and Women’s Hospital, Boston, 4 New England Deaconess Hospital, Boston, and 5 Department of Medicine, Harvard Medical School, Boston, MA, U.S.A.
(Received 18 June 1989) (Revision received 1 September 1989) (Accepted 7 September 1989)
Summary The clinical significance of cytoplasmic islet cell autoantibodies (ICA) has been studied since their discovery by Bottazzo et al. in 1974 [ 11. Some ICAs destroy pancreatic B cells in the presence of complement [ 21, whereas others take no part in this destruction [ 31. This suggests that islet function varies with the amount of ICA produced. In the present investigation we report the heterogeneity of monoclonal islet cell antibodies produced by one of us in terms of insulin release from isolated rat islets as well as from rat insulinoma cells (RINr). Key words: Insulin; ICA; RINr;
Release; Islet of Langerhans
Materials and methods
Islet incubation Islets were isolated using collagenase [4] from adult Wistar male rats fed ad libitum. They were preincubated in Krebs-Henseleit bicarbonate buffer (KHBB), pH 7.4, containing 0.5% bovine serum albumin (BSA) and 3.3 mM glucose at 37°C for 20 min under a gas phase of 95% Address for correspondence: Ejiri, M.D., Second Department of Internal Medicine, Kobe University School of Medicine, 5-1, Kusunoki-cho 7 chome, Chuou-ku, Kobe 650, Japan. 0168-8227/90/$03.50
0 1990 Elsevier Science Publishers
O2 : 5% CO,. Each five islets were further incubated in standard medium containing various concentrations of monoclonal anti-human islet cell antibody, either A2B5, AlD2, or HISL-22 [5-91, for 60 min under the above conditions. The standard medium consisted of KHBB containing 0.5% BSA and 8.3 mM glucose. The monoclonal antibodies were derived from mouse ascites as described elsewhere [lo]. As a control, mouse ascites, P3X63, which was produced by inoculating only non-hybridized myeloma cells [lo], was used. The final medium was stored at - 20°C until the radioimmunoassay of insulin
B.V. (Biomedical
Division)
2
with rat insulin as standard [ 111. The monoclonal antibodies used did not influence the present insulin assay system. Incubation of insulinoma cells Rat insulinoma cells (RINr), which originated from the insulinoma of a rat induced by X-ray irradiation [ 121, were cultured in RPM1 1640 containing 11.1 mM glucose and 1 y0 fetal calf serum. These cultured RINr cells (1 x 106/ml) were subjected to a 30-min incubation in KHBB containing 1 y0 BSA and 3.3 mM glucose at 37°C under a gas phase of 95 y0 0, : 5 y0 CO,, followed by a 60-min incubation in medium containing 1% BSA and 16.7 mM glucose, as well as either monoclonal antibody in various concentrations or control ascites derived only from the nonsecretory murine myeloma cell line P3X63. The insulin released in the final medium was determined in a similar manner as that from isolated islets.
Results E#ect of P3X63 on insulin release from rat islets Glucose (8.3 mM)-induced insulin release was 3.0 + 0.3 ng/islet * 60 min. The addition of loand lOO-fold diluted P3X63 ascites did not change the level of insulin release significantly, i.e., 2.4 k 0.3 and 2.9 k 0.3 ng/islet * 60 min, respectively (Fig. 1).
Dilution
Immunohistological study of ICA Pancreata obtained from Wistar male rats were frozen in liquid nitrogen. Five-micron sections were cut on a cryostat at - 25 ‘C, laid on slides and dried in air at room temperature. The binding of the islet cell monoclonal antibodies to the sections was examined using FITC-labeled antimouse goat IgG (Medical and Biological Laboratories Co. Ltd., Nagoya, Japan) under a fluorescence microscope by an indirect fluorescence method. The monoclonal antibodies were applied at lo-, 20-, 30-, 40-, 50-, and loo-fold dilution. Insulin assay Insulin released in the incubated medium was radioimmunoassayed with a double antibody method using rat insulin as standard [ 111. Data were analyzed by Student’s t-test and expressed as mean + SEM.
I-II
X 0
32
32
x10
Xl00
Fig. 1. Insulin release from rat islets incubated in 8.3 mM glucose-containing medium supplemented with different concentrations of myeloma-cell-derived ascites P3X63. The figures at the bottom of each column indicate the number of experiments. The column expresses the mean and the bar the SEM. ns. = not significant.
A
-IIL 7
y”yy;q:boacdya’ p3x Dil’hO”
6
5
63
ND2
X10
P3X63
AID2
Xl00
Fig. 2. Insulin release from rat islets incubated in 8.3 mM glucose-containing medium supplemented with anti-islet cell monoclonal antibody AlD2. The figures at the bottom of each column show the number of experiments. The column expresses the mean and the bar the SEM. n.s. = not significant.
3
Effect of AlD2
on insulin release from rat islets
Insulin release from islets in the presence of loand lOO-fold diluted AlD2 was 2.9 + 0.9 and 2.7 + 0.7 ng/islet * 60 min, respectively (Fig. 2) which was not different from that in the presence of the same dilution of P3X63, i.e., 3.1 + 0.4 and 3.2 + 0.3 ng/islet * 60 min, respectively (Fig. 2).
Eflect of glucose on insulinrelease from RINr cells
In the absence of glucose, insulin release was 2.0 k 0.1 ng/106 cells * 60 min. In the presence of 3.3, 8.3 and 16.7 mM glucose, insulin release in the medium was 2.6 + 0.1, 3.3 2 0.1 and
150-
Eflect of A2B.5 on insulin release from rat islets
lo- and lOO-fold diluted A2B5 induced insulin release of 2.2 k 0.3 and 2.1 & 0.3 ng/islet * 60 min, respectively, which was not different from that in the presence of similarly diluted P3X63, 2.4 + 0.4 and 2.4 + 0.2 ng/islet * 60 min, Fzspectiveiy (Fig. 3).
“5
n5
100-
2 4 : 3
2
Eflect of HISL-22 on insulin release from rat islets HISL-22, diluted lOO-fold, produced 139.4 &
15.2% more insulin release than did similarly diluted P3X63. This was a significant difference (Fig. 4). Similar dilutions of A2B5 and AlD2 produced only 92.7 f 14.1 and 87.0 f 9.1% more insulin release than the P3X63 level, which was 2.4 + 0.2 ng/islet .60 min. These were not significantly different. A
Fig. 4. Effect of anti-islet cell monoclonal antibodies on 8.3 mM glucose-induced insulin release from rat islets. The insulin release induced by the monoclonal antibodies is shown against that in the presence of myeloma-cell-derived ascites P3X63 which is expressed as 100 %. The figures at the bottom of each column indicate the number of experiments. The column expresses the mean and the bar the SEM. n.s. = not significant.
T
Fig. 3. Insulin release from rat islets incubated in 8.3 mM glucose-containing medium supplemented with anti-islet cell monoclonal antibody A2B5. The figures at the bottom of each column show the number of experiments. The column expresses the mean and the bar the SEM. n.s. = not significant.
50-
P- 0.001 -A
mM
Fig. 5. release column cates
Effect of different concentrations ofglucose on insulin from RINr cells. The figures at the bottom of each show the number of experiments. The column indithe mean and the bar the SEM. n.s. = not significant.
200
I
"b
:
I 5
Mono‘b~dl nn,kmdr
Discussion
p. 0 001
p3x63
A285
T
Li 5
AiD
5
HI%-22
Fig. 6. Effect of anti-islet cell monoclonal antibodies on 16.7 mM glucose-induced insulin release from RINr cells. The insulin release induced by the monoclonal antibodies is shown against that in the presence of myeloma-cell-derived ascites P3X63 which is expressed as 100%. The figures at the bottom of each column indicate the number of experiments. The column shows the mean and the bar the SEM. n.s. = not significant.
3.2 k 0.1 ng/lO’ cells * 60 min, respectively. Insulin release rose with higher glucose concentrations in the range of O-8.3 mM. However, 16.7 mM glucose did not produce more insulin release (Fig. 5). Comparison of the eflect of monoclonal islet cell autoantibodies on insulin release from RINr insulinoma cells
lOO-fold diluted P3X63 produced 3.1 + 0.2 ng/lO’ cells - 60 min insulin release. Similarly diluted A2B5,AlD2andHISL-22produced,respectively, 83.9 + 1.8% as much release (a significant decrease), 99.6 k 4.0% as much release, and 156.1 k 7.8% as much release (a significant increase) (Fig. 6). Binding of A2B5, AlD2, creatic tissue
It is acknowledged that stimulated insulin release deteriorates more rapidly in diabetics with ICA than in those without [ 121. This suggests that ICA may impair B cell function, directly or indirectly. However, we know that some diabetics with long-standing ICA do not lose insulin release function quickly. Thus, the question is open as to whether ICA affects islet function. In the present immunohistological study, all monoclonal anti-human islet cell autoantibodies were found to react with rat islet cells. Although the human islets would have been useful for the present investigation, it was too difficult to get appropriate human pancreata and to harvest enough human islets by collagenase digestion for a complete series of studies. Therefore, we used rat islets and the antibodies reacting with them. The monoclonal antibodies had various effects on insulin release. This variety of effects also depended upon the effector cells, i.e., rat islets or insulinoma cells. The antibody HISL-22 stimulated insulin release from rat islets, while the others did not. Conversely, A2B5 suppressed insulin release from RINr cells. The pattern of insulin release induced by HISL-22 and AlD2 from RINr cells was similar to that from rat islets; HISL-22 enhanced it, while AlD2 produced no change. A2B5 and AlD2 antibodies react against cytoplasm as well as the surface of rat islet cells [ 5-91. While the binding of HISL-22 to the islet cell surface has not been studied, some of the present authors have observed its binding to islet cell cytoplasm [7,9], and this was confirmed by our present immunohistochemical work. Other workers have shown that A2B5 reacts with cell surface gangliosides and AlD2 with glycoprotein [5-91. These specific patterns of reaction may account for the different patterns of insulin release produced from islets and insulinoma cells. The divergent patterns of insulin release might also be accounted for by the different known varieties of ICA, that react to glycolipid [8,9,13] and that
and HISL-22
to rat pan-
Fluorescence was observed on rat pancreatic islets when A2B5, AlD2, and HISL-22 were applied, but not when P3X63 was used (not shown).
5 react to protein [ 131. The destruction of insulinsecreting cells by ICAs and complement is thought to increase the insulin release from incubated islets [ 21. Conversely, an antibody like HISL-22, that increases insulin secretion, might conceivably cure diabetes. In this study, insulin appeared to have been released not from damaged but from intact islet cells, since there was no complement. This heterogeneity of ICA populations may also account for some diabetics tolerating ICA without loss of insulin release. Meanwhile, the type of tissues utilized has to be taken into careful consideration when the pathophysiological significance of ICA is studied, since different responses may be seen depending upon the effector cells used, as was the case in the present investigation. These kinds of monoclonal antibodies would make good tools for the analysis of the underlying mechanism of islet cell function with special reference to insulin release.
References Bottazzo, G. F., Florin-Christensen, A. and Doniach, D. (1974) Islet cell antibodies in diabetes mellitus with autoimmune polyendocrine deficiencies. Lancet ii, 1279-1283. Rabinovitch, A., MacKay, P., Ludvigsson, J. and Lernmark, A. (1984) A prospective analysis of islet-cell cytotoxic antibodies in insulin-dependent diabetic children. Transient effects of plasmapheresis. Diabetes 33, 224-228. Madsbad, S., Bottazzo, G. F., Cudworth, A. G., Dean, B., Faber, 0. K. and Binder, C. (1980) Islet-cell antibodies and beta-cell function in insulin-dependent diabetics. Diabetologia 18, 45-47.
4 Taniguchi, H., Ejiri, K., Ishihara, K. and Baba, S. (1982) Method of pancreatic islet isolation and culture and functional evaluation of cultured islet. Metab. Dis. 19, 299-311. 5 Eisenbarth, G. S., Shimizu, K., Bowring, M. A. and Wells, S. (1982) Expression of receptors for tetanus toxin and monoclonal antibody A2B5 by pancreatic islet cells. Proc. Natl. Acad. Sci. USA 79, 5066-5070. 6 Crump, A., Scearce, R., Doberson, M., Kortz, W. J. and Eisenbarth, G. S. (1982) Production and characterization of a cytotoxic monoclonal antibody reacting with rat islet cells. J. Clin. Invest. 70, 659-666. 7 Srikanta, S. and Eisenbarth, G. S. (1986) Islet cell antigens. Initial studies oftheir biology and function. Mol. Biol. Med. 3, 113-127. A. C., Rabizaden, A., Akeson, R. and 8 Powers, Eisenbarth, G. S. (1984) Characterization of monoclonal antibody 3G5 and utilization of this antibody to immobilize pancreatic islet cell gangliosides in a solid phase radioassay. Endocrinology 114, 1338-1343. 9 Eisenbarth, G. S. (1987) Genes, generator of diversity, glycoconjugates, and autoimmune /?-cell insufficiency in type 1 diabetes. Diabetes 36, 355-364. 10 Srikanta, S., Krisch, K. and Eisenbarth, G. S. (1986) Islet cell proteins defined by monoclonal islet cell antibody HISL-19. Diabetes 35, 300-305. 11 Morgan, C. R. and Lazarow, A. (1963) Immunoassay of insulin: two antibody system, plasma insulin levels of normal, subdiabetic and diabetic rats. Diabetes 12, 115-126. 12 Srikanta, S., Ganda, 0. P., Gleason, R. E., Jackson, R. A., Soelner, J. S. and Eisenbarth, G. S. (1984) Pre-type I diabetes: linear loss of beta cell response to intravenous glucose. Diabetes 33, 717-720. 13 Taniguchi, H., Yamashiro, Y., Morimoto, Y., Baba, S., Kodama, S., Taniguchi, T., Mori, S. and Hara, K. (1988) Heterogeneity of islet cell antibody and its clinical implication. In: G. Mimura, C. Zhisheng and I. Fukui (Eds.), Diabetes Mellitus in East Asia, Elsevier, Amsterdam, pp. 69-71.
l-5
1
Elsevier DIABET 00372
Heterogeneity of islet cell autoantibodies in terms of insulin release from rat islets and insulinoma cells Kazushige
Ejiri ‘, Hiroshi Taniguchi 1,2,5,Yasuo Morimoto ‘, Yuki Yamashiro Shigeaki Baba’, S. Srikanta2,3*4,5 and G. S. Eisenbarth2,3,4,5
‘,
I Second Department of Internal Medicine, Kobe University School of Medicine, Kobe, Japan, ’ Joslin Diabetes Center, Boston, 3 Brigham and Women’s Hospital, Boston, 4 New England Deaconess Hospital, Boston, and 5 Department of Medicine, Harvard Medical School, Boston, MA, U.S.A.
(Received 18 June 1989) (Revision received 1 September 1989) (Accepted 7 September 1989)
Summary The clinical significance of cytoplasmic islet cell autoantibodies (ICA) has been studied since their discovery by Bottazzo et al. in 1974 [ 11. Some ICAs destroy pancreatic B cells in the presence of complement [ 21, whereas others take no part in this destruction [ 31. This suggests that islet function varies with the amount of ICA produced. In the present investigation we report the heterogeneity of monoclonal islet cell antibodies produced by one of us in terms of insulin release from isolated rat islets as well as from rat insulinoma cells (RINr). Key words: Insulin; ICA; RINr;
Release; Islet of Langerhans
Materials and methods
Islet incubation Islets were isolated using collagenase [4] from adult Wistar male rats fed ad libitum. They were preincubated in Krebs-Henseleit bicarbonate buffer (KHBB), pH 7.4, containing 0.5% bovine serum albumin (BSA) and 3.3 mM glucose at 37°C for 20 min under a gas phase of 95% Address for correspondence: Ejiri, M.D., Second Department of Internal Medicine, Kobe University School of Medicine, 5-1, Kusunoki-cho 7 chome, Chuou-ku, Kobe 650, Japan. 0168-8227/90/$03.50
0 1990 Elsevier Science Publishers
O2 : 5% CO,. Each five islets were further incubated in standard medium containing various concentrations of monoclonal anti-human islet cell antibody, either A2B5, AlD2, or HISL-22 [5-91, for 60 min under the above conditions. The standard medium consisted of KHBB containing 0.5% BSA and 8.3 mM glucose. The monoclonal antibodies were derived from mouse ascites as described elsewhere [lo]. As a control, mouse ascites, P3X63, which was produced by inoculating only non-hybridized myeloma cells [lo], was used. The final medium was stored at - 20°C until the radioimmunoassay of insulin
B.V. (Biomedical
Division)
2
with rat insulin as standard [ 111. The monoclonal antibodies used did not influence the present insulin assay system. Incubation of insulinoma cells Rat insulinoma cells (RINr), which originated from the insulinoma of a rat induced by X-ray irradiation [ 121, were cultured in RPM1 1640 containing 11.1 mM glucose and 1 y0 fetal calf serum. These cultured RINr cells (1 x 106/ml) were subjected to a 30-min incubation in KHBB containing 1 y0 BSA and 3.3 mM glucose at 37°C under a gas phase of 95 y0 0, : 5 y0 CO,, followed by a 60-min incubation in medium containing 1% BSA and 16.7 mM glucose, as well as either monoclonal antibody in various concentrations or control ascites derived only from the nonsecretory murine myeloma cell line P3X63. The insulin released in the final medium was determined in a similar manner as that from isolated islets.
Results E#ect of P3X63 on insulin release from rat islets Glucose (8.3 mM)-induced insulin release was 3.0 + 0.3 ng/islet * 60 min. The addition of loand lOO-fold diluted P3X63 ascites did not change the level of insulin release significantly, i.e., 2.4 k 0.3 and 2.9 k 0.3 ng/islet * 60 min, respectively (Fig. 1).
Dilution
Immunohistological study of ICA Pancreata obtained from Wistar male rats were frozen in liquid nitrogen. Five-micron sections were cut on a cryostat at - 25 ‘C, laid on slides and dried in air at room temperature. The binding of the islet cell monoclonal antibodies to the sections was examined using FITC-labeled antimouse goat IgG (Medical and Biological Laboratories Co. Ltd., Nagoya, Japan) under a fluorescence microscope by an indirect fluorescence method. The monoclonal antibodies were applied at lo-, 20-, 30-, 40-, 50-, and loo-fold dilution. Insulin assay Insulin released in the incubated medium was radioimmunoassayed with a double antibody method using rat insulin as standard [ 111. Data were analyzed by Student’s t-test and expressed as mean + SEM.
I-II
X 0
32
32
x10
Xl00
Fig. 1. Insulin release from rat islets incubated in 8.3 mM glucose-containing medium supplemented with different concentrations of myeloma-cell-derived ascites P3X63. The figures at the bottom of each column indicate the number of experiments. The column expresses the mean and the bar the SEM. ns. = not significant.
A
-IIL 7
y”yy;q:boacdya’ p3x Dil’hO”
6
5
63
ND2
X10
P3X63
AID2
Xl00
Fig. 2. Insulin release from rat islets incubated in 8.3 mM glucose-containing medium supplemented with anti-islet cell monoclonal antibody AlD2. The figures at the bottom of each column show the number of experiments. The column expresses the mean and the bar the SEM. n.s. = not significant.
3
Effect of AlD2
on insulin release from rat islets
Insulin release from islets in the presence of loand lOO-fold diluted AlD2 was 2.9 + 0.9 and 2.7 + 0.7 ng/islet * 60 min, respectively (Fig. 2) which was not different from that in the presence of the same dilution of P3X63, i.e., 3.1 + 0.4 and 3.2 + 0.3 ng/islet * 60 min, respectively (Fig. 2).
Eflect of glucose on insulinrelease from RINr cells
In the absence of glucose, insulin release was 2.0 k 0.1 ng/106 cells * 60 min. In the presence of 3.3, 8.3 and 16.7 mM glucose, insulin release in the medium was 2.6 + 0.1, 3.3 2 0.1 and
150-
Eflect of A2B.5 on insulin release from rat islets
lo- and lOO-fold diluted A2B5 induced insulin release of 2.2 k 0.3 and 2.1 & 0.3 ng/islet * 60 min, respectively, which was not different from that in the presence of similarly diluted P3X63, 2.4 + 0.4 and 2.4 + 0.2 ng/islet * 60 min, Fzspectiveiy (Fig. 3).
“5
n5
100-
2 4 : 3
2
Eflect of HISL-22 on insulin release from rat islets HISL-22, diluted lOO-fold, produced 139.4 &
15.2% more insulin release than did similarly diluted P3X63. This was a significant difference (Fig. 4). Similar dilutions of A2B5 and AlD2 produced only 92.7 f 14.1 and 87.0 f 9.1% more insulin release than the P3X63 level, which was 2.4 + 0.2 ng/islet .60 min. These were not significantly different. A
Fig. 4. Effect of anti-islet cell monoclonal antibodies on 8.3 mM glucose-induced insulin release from rat islets. The insulin release induced by the monoclonal antibodies is shown against that in the presence of myeloma-cell-derived ascites P3X63 which is expressed as 100 %. The figures at the bottom of each column indicate the number of experiments. The column expresses the mean and the bar the SEM. n.s. = not significant.
T
Fig. 3. Insulin release from rat islets incubated in 8.3 mM glucose-containing medium supplemented with anti-islet cell monoclonal antibody A2B5. The figures at the bottom of each column show the number of experiments. The column expresses the mean and the bar the SEM. n.s. = not significant.
50-
P- 0.001 -A
mM
Fig. 5. release column cates
Effect of different concentrations ofglucose on insulin from RINr cells. The figures at the bottom of each show the number of experiments. The column indithe mean and the bar the SEM. n.s. = not significant.
200
I
"b
:
I 5
Mono‘b~dl nn,kmdr
Discussion
p. 0 001
p3x63
A285
T
Li 5
AiD
5
HI%-22
Fig. 6. Effect of anti-islet cell monoclonal antibodies on 16.7 mM glucose-induced insulin release from RINr cells. The insulin release induced by the monoclonal antibodies is shown against that in the presence of myeloma-cell-derived ascites P3X63 which is expressed as 100%. The figures at the bottom of each column indicate the number of experiments. The column shows the mean and the bar the SEM. n.s. = not significant.
3.2 k 0.1 ng/lO’ cells * 60 min, respectively. Insulin release rose with higher glucose concentrations in the range of O-8.3 mM. However, 16.7 mM glucose did not produce more insulin release (Fig. 5). Comparison of the eflect of monoclonal islet cell autoantibodies on insulin release from RINr insulinoma cells
lOO-fold diluted P3X63 produced 3.1 + 0.2 ng/lO’ cells - 60 min insulin release. Similarly diluted A2B5,AlD2andHISL-22produced,respectively, 83.9 + 1.8% as much release (a significant decrease), 99.6 k 4.0% as much release, and 156.1 k 7.8% as much release (a significant increase) (Fig. 6). Binding of A2B5, AlD2, creatic tissue
It is acknowledged that stimulated insulin release deteriorates more rapidly in diabetics with ICA than in those without [ 121. This suggests that ICA may impair B cell function, directly or indirectly. However, we know that some diabetics with long-standing ICA do not lose insulin release function quickly. Thus, the question is open as to whether ICA affects islet function. In the present immunohistological study, all monoclonal anti-human islet cell autoantibodies were found to react with rat islet cells. Although the human islets would have been useful for the present investigation, it was too difficult to get appropriate human pancreata and to harvest enough human islets by collagenase digestion for a complete series of studies. Therefore, we used rat islets and the antibodies reacting with them. The monoclonal antibodies had various effects on insulin release. This variety of effects also depended upon the effector cells, i.e., rat islets or insulinoma cells. The antibody HISL-22 stimulated insulin release from rat islets, while the others did not. Conversely, A2B5 suppressed insulin release from RINr cells. The pattern of insulin release induced by HISL-22 and AlD2 from RINr cells was similar to that from rat islets; HISL-22 enhanced it, while AlD2 produced no change. A2B5 and AlD2 antibodies react against cytoplasm as well as the surface of rat islet cells [ 5-91. While the binding of HISL-22 to the islet cell surface has not been studied, some of the present authors have observed its binding to islet cell cytoplasm [7,9], and this was confirmed by our present immunohistochemical work. Other workers have shown that A2B5 reacts with cell surface gangliosides and AlD2 with glycoprotein [5-91. These specific patterns of reaction may account for the different patterns of insulin release produced from islets and insulinoma cells. The divergent patterns of insulin release might also be accounted for by the different known varieties of ICA, that react to glycolipid [8,9,13] and that
and HISL-22
to rat pan-
Fluorescence was observed on rat pancreatic islets when A2B5, AlD2, and HISL-22 were applied, but not when P3X63 was used (not shown).
5 react to protein [ 131. The destruction of insulinsecreting cells by ICAs and complement is thought to increase the insulin release from incubated islets [ 21. Conversely, an antibody like HISL-22, that increases insulin secretion, might conceivably cure diabetes. In this study, insulin appeared to have been released not from damaged but from intact islet cells, since there was no complement. This heterogeneity of ICA populations may also account for some diabetics tolerating ICA without loss of insulin release. Meanwhile, the type of tissues utilized has to be taken into careful consideration when the pathophysiological significance of ICA is studied, since different responses may be seen depending upon the effector cells used, as was the case in the present investigation. These kinds of monoclonal antibodies would make good tools for the analysis of the underlying mechanism of islet cell function with special reference to insulin release.
References Bottazzo, G. F., Florin-Christensen, A. and Doniach, D. (1974) Islet cell antibodies in diabetes mellitus with autoimmune polyendocrine deficiencies. Lancet ii, 1279-1283. Rabinovitch, A., MacKay, P., Ludvigsson, J. and Lernmark, A. (1984) A prospective analysis of islet-cell cytotoxic antibodies in insulin-dependent diabetic children. Transient effects of plasmapheresis. Diabetes 33, 224-228. Madsbad, S., Bottazzo, G. F., Cudworth, A. G., Dean, B., Faber, 0. K. and Binder, C. (1980) Islet-cell antibodies and beta-cell function in insulin-dependent diabetics. Diabetologia 18, 45-47.
4 Taniguchi, H., Ejiri, K., Ishihara, K. and Baba, S. (1982) Method of pancreatic islet isolation and culture and functional evaluation of cultured islet. Metab. Dis. 19, 299-311. 5 Eisenbarth, G. S., Shimizu, K., Bowring, M. A. and Wells, S. (1982) Expression of receptors for tetanus toxin and monoclonal antibody A2B5 by pancreatic islet cells. Proc. Natl. Acad. Sci. USA 79, 5066-5070. 6 Crump, A., Scearce, R., Doberson, M., Kortz, W. J. and Eisenbarth, G. S. (1982) Production and characterization of a cytotoxic monoclonal antibody reacting with rat islet cells. J. Clin. Invest. 70, 659-666. 7 Srikanta, S. and Eisenbarth, G. S. (1986) Islet cell antigens. Initial studies oftheir biology and function. Mol. Biol. Med. 3, 113-127. A. C., Rabizaden, A., Akeson, R. and 8 Powers, Eisenbarth, G. S. (1984) Characterization of monoclonal antibody 3G5 and utilization of this antibody to immobilize pancreatic islet cell gangliosides in a solid phase radioassay. Endocrinology 114, 1338-1343. 9 Eisenbarth, G. S. (1987) Genes, generator of diversity, glycoconjugates, and autoimmune /?-cell insufficiency in type 1 diabetes. Diabetes 36, 355-364. 10 Srikanta, S., Krisch, K. and Eisenbarth, G. S. (1986) Islet cell proteins defined by monoclonal islet cell antibody HISL-19. Diabetes 35, 300-305. 11 Morgan, C. R. and Lazarow, A. (1963) Immunoassay of insulin: two antibody system, plasma insulin levels of normal, subdiabetic and diabetic rats. Diabetes 12, 115-126. 12 Srikanta, S., Ganda, 0. P., Gleason, R. E., Jackson, R. A., Soelner, J. S. and Eisenbarth, G. S. (1984) Pre-type I diabetes: linear loss of beta cell response to intravenous glucose. Diabetes 33, 717-720. 13 Taniguchi, H., Yamashiro, Y., Morimoto, Y., Baba, S., Kodama, S., Taniguchi, T., Mori, S. and Hara, K. (1988) Heterogeneity of islet cell antibody and its clinical implication. In: G. Mimura, C. Zhisheng and I. Fukui (Eds.), Diabetes Mellitus in East Asia, Elsevier, Amsterdam, pp. 69-71.
