- Email: [email protected]
In vitro paracrine regulation of islet B-cell function by A and D cells
Life Sciences, Vol. 32, pp. 1873-1878 Printed in the U.S.A.
Pergamon Press
IN VITRO PARACRINE REGULATION OF ISLET B-CET,T, FUNCTION BY A AND D C[~Lq Wilfred Y. Fujimoto, Shoji Kawazu, Masatoshi Ikeuchi, and Yasunori Kanazawa Department of Medicine, Division of Metabolism and Endocrinology, University of Washington, Seattle, Washington 98195; 4th Department of Internal Medicine, Saitama Medical School, Saitama, Japan 350-04; and 3rd Department of Internal Medicine, University of Tokyo, Tokyo, Japan 113. (Received in final form January 26, 1983) Summary In monolayer cultures of islet cells from neonatal rats, incubation of cells for 1 hour with either anti-somatostatin serum or anti-glucagon serum enhanced insulin release. The former appears to be due to neutralization of endogenously secreted somatostatin. The latter may be due to removal of a stimulatory effect of endogenously released glucagon upon somatestatin secretion. Thus, although exogenously added glucagon stimulates insulin secretion, the effect of endogenously released glucagon upon islet B cells is a restraining one which may be mediated through an effect upon D cells and their release of endogenous somatestatin. There is increasing evidence that secreted islet hormones may locally regulate the function of islet cells. This paracrine regulatory process is thought to occur when a substance secreted by a cell into the interstitial space influences the function of another cell. We have used monolayer cultures of newborn rat pancreatic islet cells to examine this paracrine relationship among the A, B, and D cells of the islet. In these cultures, approximately 50% of the pla~na membrane external surface area of each cell is exposed to the incubation medium. If one considers the incubation medium to represent the interstitial space, then this tissue culture system becomes a reasonable in vitro model for investigation of paracrine relationships among the islet A, B and D cells in which all cells are quickly and uniformly exposed to the cc~ponents of the medium, and secretory products are released directly into the medium. Both processes occur rapidly since diffusion through intervening tissue is avoided. Thus this appears to be an ideal system for investigating the paracrine relationships among the islet A, B and D cells through the addition to the incubation medium of either islet hormones or antisera to islet hormones. Materials and Methods Monolayer cultures of neonatal (3-5 d old) Wistar rat pancreases were prepared using methods described previously (i). Cultures were established in 35 x i0 mm tissue culture dishes (Falcon 3001), and experiments were done on the fourth or fifth day after digestion of the pancreases. For experiments, cultures were preincubated for 1 hour in 1.0 ml Krebs Ringer bicarbonate (KRB) buffer containing 0.5% bovine serum albumin (BSA) and 2.7 mM glucose. BSA was treated with I~)TA and then exhaustively dialyzed against distilled water before use. Cells were then incubated for 1 hour in I. 0 ml KRB buffer with BSA and test substances. Glucagon was obtained fram 0024-3205/83/161873-06503.00/0 Copyright (c) 1983 Pergamon Press Ltd.
1874
Paracrine Regulation of Islet B Cells
Vol. 32, No. 16, 1983
Novo ( ~ k ) , somatostatin from Protein Research Foundation (Japan), rabbit antisomatostatin serum fromDr. N. Yanaihara, and rabbit anti-glucagon serum (GC-5) from OtsukaAssayLaboratory (Japan). Incubation mediumwas saved for subsequent measurement of insulin and glucagon levels. Cell layers were extractedwith acid-ethanol for subsequent determination of hormone content. Insulin was assayed by double antibody radioimmunoassay, using rat insulin standards (2). Glucagon was measured by double antibody radioimmunoassay, using porcine glucagon standards. Hormone fractional release rates for both insulin and glucagon were calculated from the ratio (hormone secreted) - (hormone secreted + hormone content of cells). Guinea pig antiserum to porcine insulin, and rabbit antisera to glucagon or somatostatin (all antisera used at 1:500 dilution) were used to stain fixed (Bouin's fluid) menolayercultures of islet cells in tissue culture chamber/slides (Lab-Tek 4804) by the peroxidase-antiperoxidase cc~plex technique (3,4). In addition, control sections of rat pancreas were used to confirm staining. Immunocytochemical controls included omission of primary antiserum, c~ission of all antisera, substitution of normal guinea pig or rabbit serum for p r ~ antiserum, and using primary antiserum previously mixed (absorbed) with the appropriate antigen. Significance of differences was determined by Student's t test for isolated comparisons and by Newman Keuls test for multiple comparisons. Results The relative proportions of A, B and D cells was determined by counting the number of positively stained cells among all endocrine cells in cultures immunocytochemically stained for glucagon, insulin, and somatostatin, respectively. Results from one study are shown in Table I. In general, B cells comprise 50-60% of the endocrine type cells in these cultures, A cells about 20%, and D cells about 10%. TABLE I Relative Proportions of A, B and D Cells
Cell Type A B D a:
~aSD)
Total # Cells Counted
23.8+4.0 52.1T7.1 12.3TI.8
450 404 488
(x
Relative proportions of A, B and D cells as determined from immunocytochemical staining of newborn rat pancreatic monolayer cultures, expressed as mean +SD of 5 separate counts. Fibroblastoid cells, which constitute about 50% of the cells, have been excluded.
As shown in Figure i, glucagon 5 ng/ml significantly enhanced insulin secretion at glucose 2.7 or 16.5 hiM, whereas somatostatin i0 ng/ml reduced insulin secretion at glucose 16.5 n~4. Under conditions where glucagon release is highly stimulated (glucose 2.7 ram plus arginine 16.5 r~M), anti-glucagon serum 2% (v/v) significantly enhanced insulin secretion (Fig. 2). At lesser concentrations of anti-glucagon serum, insulin release was not significantly affected.
Vol. 32, No, 16, 1983
Paracrine Regulation of Islet B Cells
1875
~'P
.s-- 1 200
[P
I00
Glucose(mM) 2.7 2.7 2.7 16.5 16.5 16.5 Glucagon(ng/ml) 0 5 0 0 5 0 Somotostetin(ng/ml) 0 0 I0 0 0 I0 FIG. 1 Insulin secretion (% of basal secretion at glucose 2.7 nlM, mean+SE~) in the absence or presence of exogenous glucagon (5 ng/ml) or somatostatin (I0 ng/ml) and glucose (2.7 or 16.5 ~4).
150
Glucose 2.7mM+grginine 16.5mM
r - -
p
§ IOO
,
NS
c--NS
50
Rabbit Anti-cjlucagon 0 Serum (%) NormaIRabbit Serum 2 (%)
0.5
I
2
1.5
I
0
FIG. 2 Fractional insulin secretion (mean+SEM) in the presence of glucose 2.7 mM + arginine 16.5 mM and either normal rabbit s ~ , rabbit anti-glucagon serum, or a combination of both. Anti-somatostatin sertnn 2% (v/v) enhanced glucose-stia~lated but not basal insulin release (Fig. 3). Glucagon secretion was increased at glucose 16.5 mM but this was not significant.
1876
Paracrine Regulation of Islet B Cells
I
[--NS~
Vol. 32, No. 16, 1983
_,,,c _
2O a)
g c)
,o --
._O
co
0
§
~.~50 C
0 Glucose (raM) Rot?bit Anti-somatosto ti n
Serum (%)
Normol Rabbit Serum (%)
[--NSq [p
,,ml 2.7
2.7 16.5 16.5
0
2
0
2
2
0
2
0
FIG. 3 Fractional insulin and glucagon secretion (mean+_S~4) in the presence of either normal rabbit serum or rabbit anti-somatostatin serum and glucose (2.7 or 16.5 n~4). Discussion These results Lndicate that endogenously secreted sc~atostatin and glucagon restrain B cell release of insulin. The effect of endogenous somatostatin is similar to that of the exogenously added peptide. However, exogenously added glucagon stimulated insulin secretion. The difference between endogenously released and exogenously added glucagon requires explanation. The stin~latory effect of glucagon upon insulin secretion has already been well doc~nented (5) . The addition of glucagon anti-serum would be expected to bind and inactivate endogenously secreted glucagon. Such an action may lead to a number of possible outcomes. It is possible that reduction of the amount of glucagon available to B-cell receptors may occur, and insulin secretion reduced. However, we observed an increase in insulin release. Addition of glucagon anti-serum may accelerate secretion of glucagon, and some of the glucagon may remain unbound to antibody and become available to B-cell glucagon receptors. Under these conditions, insulin secretion would be increased. Alternatively, the amount of glucagon available to D-cell receptors for glucagon may be reduced, thereby lowering scmatostatin release and in turn increasing insulin secretion. Of these two possible mechanisms whereby insulin secretion may be enhanced by the addition of anti-glucagon sert~n, the latter seems more plausible.
Vol. 32, No. 16, 1983
Paracrine Regulation of Islet B Cells
1877
Islet cells examined in monolayer culture with a scanning electron microscope exhibit many round bumps on the cell surface which are thought to represent secretory granules bulging from the cell (6). It appears highly likely, therefore, that secretory products of islet cells in monolayer culture are released into the incubation medium from the cell surface facing the medium rather than the opposite cell surface which is attached to the culture vessel. When glucagon anti-serum is added to the incubation medium, it seems improbable that under these circtm~tances secreted glucagon would remain free and unbound to antibody. On the other hand, glucagon is known to stimulate somatostatin release (7-9). Addition of glucagon antiserum therefore removes an endogenous stimulus for somatostatin release. Our results with anti-somatostatin serum indicate that endogenously secreted scmatostatin is capable of restraining insulin release. Since somatostatin secretion is stimulated by glucose (8,10), the observation that anti-somatostatin serum stimulated only glucose-stimulated insulin release is appropriate. Previously, several groups of investigators have shown that either somatostatin anti-serum or anti-somatostatin gamma globulin augmented insulin secretion by isolated rat islets (11-13) or the perfused chicken pancreas (14). In some of these, glucagon secretion was also increased (12,14). However, in two studies, antisomatostatin serum had no effect upon insulin secretion by isolated organ cultured rat islets (15) or the isolated perfused rat pancreas (16). The reasons for these disparate observations remain to be clarified, but may be due to differences in amounts of anti-serum used. Nevertheless, the majority of studies of this kind suggest that endogenously secreted somatostatin modulates insulin secretion, and in some cases, glucagon release as well. We know of only one prior report examining the effect of anti-glucagon serum upon insulin release. Tanigawa et al. (13) showed in isolated rat islets that anti-glucagon serum enhanced insulin release at glucose 5.5 r~4 but not at glucose 20 hiM. This glucose dependency may be due to suppression of glucagon release by glucose; under these conditions, addition of anti-glucagon sert~m may have no further effect. Exogenously added glucagon probably also enhanced somatostatin secretion by D cells. However, since stimulation of insulin secretion was observed, the direct effect of glucagon upon the B cells apparently was enough to negate the inhibitory effect of endogenously released sc~atostatin. Kadowaki et al. (9) have shown wi N the i~. lated perfused rat pancreas a dose effect of exogenous glucagon (i0 to i0 M) upon somatostatin and insulin release. With higher glucagon, more scmatostatin was released. Glucagon at any concentration enhanced insulin release cc~pared to absence of glucagon, but glucagon-stimulated insulin release was less at the higher glucagon concentrations. In order to confirm the mechanism whereby anti-glucagon serum enhances insulin release, additional studies are required. These must be directed toward the measurement of somatostatin secretion. Although others have reported successful assays for s(m~atostatin secretion by menolayer islet cell cultures similar to the ones used by us, we have not been able to detect somatostatin secretion (W.Y. Fujimoto and J.W. Ensinck, unpublished). However, cells containing somatostatin immunoreactivity are present in the culture. In conclusion, endogenously released scmatostatin inhibits insulin secretion. In addition, endogenously released glucagon appears to have a restraining effect upon insulin secretion since neutralization of glucagon by anti-glucagon serum enhanced insulin secretion. This effect is probably mediated through glucagon stimulation of somatostatin secretion. Exogenously added glucagon has a net stimulatory effect upon insulin secretion.
1878
Paracrine Regulation of Islet B Cells
Vol. 32, No. 6, 1983
Acknowledc.~nents We are grateful to Dr. A. Kaneto for the kind donation of rabbit anti-glucagon serum; Tcmoko Chinone, MarikoFujita, and Jeanette Teague for their expert technical assistance, and Sharon Kemp for her careful preparation of this manuscript. This work was supported in part by NIH grants AM 15312 and AM 17047. Part of this work was done while Dr. Fujimoto was on sabbatical leave at the University of Tokyo and a recipient of a fellowship from the Japan Society for the Promotion of Science. References i. 2. 3.
4. 5. 6. 7. 8. 9. i0. ii.
12. 13. 14. 15. 16.
W.Y. FUJIMOTO, J.W. ENSINCK and R.H. WILLIAMS, Life Sci. 15 1999-2004 (1974). Y. KANAZAWA, T. KUZUYA, T. IDE and K. KOSAKA, Am. J. Physiol. 211 442-448 (1966). W. STRANSE, in First International Syr~posium on Immunoenzymatic Techniques (INSERM Slmlo. No. 2), G. Feldman, P. Druet, J. Bignon, and S. Avrameas, eds., p. 117 North Holland, Amsterdam (1976). D.G. BASKIN, K.C. GORRAY and W.Y. FLUIMOTO, J. Histoche~. Cytochem. 29 567-571 (1981). E. SAMOLS, G. MARRI and V. MARKS, Lancet II 415-416 (1965). G. SOMERS, B. BLONDEL, L. ORCI and W.J. MALAISSE, Endocrinology 104 255264 (1979). G.S. PATION, E. IPP, R.E. DOBBS, L. ORCI, W. VALE and R.H. UNGER, Proc. Nat. Acad. Sci. USA 74 2140-2143 (1977). G.C. WEIR, E. SAMOLS and Y.C. PATEL, Metabolism 27 (suppl i) 1223-1226 (1978). S. KADOWAKI, T. TAMINATO, T. CHIBA, K. MORI, H. ABE, Y. GOTO, Y. SEINO, S. MATSUKURA, M. NOZAk~ and T. FUJITA, Diabetes 28 600-603 (1979). P. SCHAUDER, C. McINTOSH, J. ARENDS, R. ARNOLD, H. FRERICHS and W. CREUTZFELDT, FEBS Lett. 68 225-227 (1976). H. TANI~JCHI, M. UTSUMI, M. HASEGAk%, T. KOBAYASHI, Y. WATANABE, K. MURAKAMI, M. SEKI, A. TSUTOU, H. MAKIMURA, M. SAKADA and S. BABA, Diabetes 26 700-702 (1977). M. ITOH, L. MANDARINO and J.E. GERICH, Diabetes 29 693-696 (1980). K. TANIGAWA, H. KUZUYA, H. SAKURAI, Y. SEINO, S. SEINO, K. TSUDA and H. IMURA, Horm. Metab. Res. 13 78-80 (1981). R.N. HONEY, A. ARIMURA and G.C. WEIR, Endocrinology 109 1971-1974 (1981). N. B A B D ~ , M. LAVOIE, A. DUPONT, J. COTE and J.-P. COTE, Endocrinology i01 635-638 (1977). R.L. SOR~NSON, D.V. LIND,.T. and R.P. ELDE, Diabetes 29 747-751 (1980).
Pergamon Press
IN VITRO PARACRINE REGULATION OF ISLET B-CET,T, FUNCTION BY A AND D C[~Lq Wilfred Y. Fujimoto, Shoji Kawazu, Masatoshi Ikeuchi, and Yasunori Kanazawa Department of Medicine, Division of Metabolism and Endocrinology, University of Washington, Seattle, Washington 98195; 4th Department of Internal Medicine, Saitama Medical School, Saitama, Japan 350-04; and 3rd Department of Internal Medicine, University of Tokyo, Tokyo, Japan 113. (Received in final form January 26, 1983) Summary In monolayer cultures of islet cells from neonatal rats, incubation of cells for 1 hour with either anti-somatostatin serum or anti-glucagon serum enhanced insulin release. The former appears to be due to neutralization of endogenously secreted somatostatin. The latter may be due to removal of a stimulatory effect of endogenously released glucagon upon somatestatin secretion. Thus, although exogenously added glucagon stimulates insulin secretion, the effect of endogenously released glucagon upon islet B cells is a restraining one which may be mediated through an effect upon D cells and their release of endogenous somatestatin. There is increasing evidence that secreted islet hormones may locally regulate the function of islet cells. This paracrine regulatory process is thought to occur when a substance secreted by a cell into the interstitial space influences the function of another cell. We have used monolayer cultures of newborn rat pancreatic islet cells to examine this paracrine relationship among the A, B, and D cells of the islet. In these cultures, approximately 50% of the pla~na membrane external surface area of each cell is exposed to the incubation medium. If one considers the incubation medium to represent the interstitial space, then this tissue culture system becomes a reasonable in vitro model for investigation of paracrine relationships among the islet A, B and D cells in which all cells are quickly and uniformly exposed to the cc~ponents of the medium, and secretory products are released directly into the medium. Both processes occur rapidly since diffusion through intervening tissue is avoided. Thus this appears to be an ideal system for investigating the paracrine relationships among the islet A, B and D cells through the addition to the incubation medium of either islet hormones or antisera to islet hormones. Materials and Methods Monolayer cultures of neonatal (3-5 d old) Wistar rat pancreases were prepared using methods described previously (i). Cultures were established in 35 x i0 mm tissue culture dishes (Falcon 3001), and experiments were done on the fourth or fifth day after digestion of the pancreases. For experiments, cultures were preincubated for 1 hour in 1.0 ml Krebs Ringer bicarbonate (KRB) buffer containing 0.5% bovine serum albumin (BSA) and 2.7 mM glucose. BSA was treated with I~)TA and then exhaustively dialyzed against distilled water before use. Cells were then incubated for 1 hour in I. 0 ml KRB buffer with BSA and test substances. Glucagon was obtained fram 0024-3205/83/161873-06503.00/0 Copyright (c) 1983 Pergamon Press Ltd.
1874
Paracrine Regulation of Islet B Cells
Vol. 32, No. 16, 1983
Novo ( ~ k ) , somatostatin from Protein Research Foundation (Japan), rabbit antisomatostatin serum fromDr. N. Yanaihara, and rabbit anti-glucagon serum (GC-5) from OtsukaAssayLaboratory (Japan). Incubation mediumwas saved for subsequent measurement of insulin and glucagon levels. Cell layers were extractedwith acid-ethanol for subsequent determination of hormone content. Insulin was assayed by double antibody radioimmunoassay, using rat insulin standards (2). Glucagon was measured by double antibody radioimmunoassay, using porcine glucagon standards. Hormone fractional release rates for both insulin and glucagon were calculated from the ratio (hormone secreted) - (hormone secreted + hormone content of cells). Guinea pig antiserum to porcine insulin, and rabbit antisera to glucagon or somatostatin (all antisera used at 1:500 dilution) were used to stain fixed (Bouin's fluid) menolayercultures of islet cells in tissue culture chamber/slides (Lab-Tek 4804) by the peroxidase-antiperoxidase cc~plex technique (3,4). In addition, control sections of rat pancreas were used to confirm staining. Immunocytochemical controls included omission of primary antiserum, c~ission of all antisera, substitution of normal guinea pig or rabbit serum for p r ~ antiserum, and using primary antiserum previously mixed (absorbed) with the appropriate antigen. Significance of differences was determined by Student's t test for isolated comparisons and by Newman Keuls test for multiple comparisons. Results The relative proportions of A, B and D cells was determined by counting the number of positively stained cells among all endocrine cells in cultures immunocytochemically stained for glucagon, insulin, and somatostatin, respectively. Results from one study are shown in Table I. In general, B cells comprise 50-60% of the endocrine type cells in these cultures, A cells about 20%, and D cells about 10%. TABLE I Relative Proportions of A, B and D Cells
Cell Type A B D a:
~aSD)
Total # Cells Counted
23.8+4.0 52.1T7.1 12.3TI.8
450 404 488
(x
Relative proportions of A, B and D cells as determined from immunocytochemical staining of newborn rat pancreatic monolayer cultures, expressed as mean +SD of 5 separate counts. Fibroblastoid cells, which constitute about 50% of the cells, have been excluded.
As shown in Figure i, glucagon 5 ng/ml significantly enhanced insulin secretion at glucose 2.7 or 16.5 hiM, whereas somatostatin i0 ng/ml reduced insulin secretion at glucose 16.5 n~4. Under conditions where glucagon release is highly stimulated (glucose 2.7 ram plus arginine 16.5 r~M), anti-glucagon serum 2% (v/v) significantly enhanced insulin secretion (Fig. 2). At lesser concentrations of anti-glucagon serum, insulin release was not significantly affected.
Vol. 32, No, 16, 1983
Paracrine Regulation of Islet B Cells
1875
~'P
.s-- 1 200
[P
I00
Glucose(mM) 2.7 2.7 2.7 16.5 16.5 16.5 Glucagon(ng/ml) 0 5 0 0 5 0 Somotostetin(ng/ml) 0 0 I0 0 0 I0 FIG. 1 Insulin secretion (% of basal secretion at glucose 2.7 nlM, mean+SE~) in the absence or presence of exogenous glucagon (5 ng/ml) or somatostatin (I0 ng/ml) and glucose (2.7 or 16.5 ~4).
150
Glucose 2.7mM+grginine 16.5mM
r - -
p
§ IOO
,
NS
c--NS
50
Rabbit Anti-cjlucagon 0 Serum (%) NormaIRabbit Serum 2 (%)
0.5
I
2
1.5
I
0
FIG. 2 Fractional insulin secretion (mean+SEM) in the presence of glucose 2.7 mM + arginine 16.5 mM and either normal rabbit s ~ , rabbit anti-glucagon serum, or a combination of both. Anti-somatostatin sertnn 2% (v/v) enhanced glucose-stia~lated but not basal insulin release (Fig. 3). Glucagon secretion was increased at glucose 16.5 mM but this was not significant.
1876
Paracrine Regulation of Islet B Cells
I
[--NS~
Vol. 32, No. 16, 1983
_,,,c _
2O a)
g c)
,o --
._O
co
0
§
~.~50 C
0 Glucose (raM) Rot?bit Anti-somatosto ti n
Serum (%)
Normol Rabbit Serum (%)
[--NSq [p
,,ml 2.7
2.7 16.5 16.5
0
2
0
2
2
0
2
0
FIG. 3 Fractional insulin and glucagon secretion (mean+_S~4) in the presence of either normal rabbit serum or rabbit anti-somatostatin serum and glucose (2.7 or 16.5 n~4). Discussion These results Lndicate that endogenously secreted sc~atostatin and glucagon restrain B cell release of insulin. The effect of endogenous somatostatin is similar to that of the exogenously added peptide. However, exogenously added glucagon stimulated insulin secretion. The difference between endogenously released and exogenously added glucagon requires explanation. The stin~latory effect of glucagon upon insulin secretion has already been well doc~nented (5) . The addition of glucagon anti-serum would be expected to bind and inactivate endogenously secreted glucagon. Such an action may lead to a number of possible outcomes. It is possible that reduction of the amount of glucagon available to B-cell receptors may occur, and insulin secretion reduced. However, we observed an increase in insulin release. Addition of glucagon anti-serum may accelerate secretion of glucagon, and some of the glucagon may remain unbound to antibody and become available to B-cell glucagon receptors. Under these conditions, insulin secretion would be increased. Alternatively, the amount of glucagon available to D-cell receptors for glucagon may be reduced, thereby lowering scmatostatin release and in turn increasing insulin secretion. Of these two possible mechanisms whereby insulin secretion may be enhanced by the addition of anti-glucagon sert~n, the latter seems more plausible.
Vol. 32, No. 16, 1983
Paracrine Regulation of Islet B Cells
1877
Islet cells examined in monolayer culture with a scanning electron microscope exhibit many round bumps on the cell surface which are thought to represent secretory granules bulging from the cell (6). It appears highly likely, therefore, that secretory products of islet cells in monolayer culture are released into the incubation medium from the cell surface facing the medium rather than the opposite cell surface which is attached to the culture vessel. When glucagon anti-serum is added to the incubation medium, it seems improbable that under these circtm~tances secreted glucagon would remain free and unbound to antibody. On the other hand, glucagon is known to stimulate somatostatin release (7-9). Addition of glucagon antiserum therefore removes an endogenous stimulus for somatostatin release. Our results with anti-somatostatin serum indicate that endogenously secreted scmatostatin is capable of restraining insulin release. Since somatostatin secretion is stimulated by glucose (8,10), the observation that anti-somatostatin serum stimulated only glucose-stimulated insulin release is appropriate. Previously, several groups of investigators have shown that either somatostatin anti-serum or anti-somatostatin gamma globulin augmented insulin secretion by isolated rat islets (11-13) or the perfused chicken pancreas (14). In some of these, glucagon secretion was also increased (12,14). However, in two studies, antisomatostatin serum had no effect upon insulin secretion by isolated organ cultured rat islets (15) or the isolated perfused rat pancreas (16). The reasons for these disparate observations remain to be clarified, but may be due to differences in amounts of anti-serum used. Nevertheless, the majority of studies of this kind suggest that endogenously secreted somatostatin modulates insulin secretion, and in some cases, glucagon release as well. We know of only one prior report examining the effect of anti-glucagon serum upon insulin release. Tanigawa et al. (13) showed in isolated rat islets that anti-glucagon serum enhanced insulin release at glucose 5.5 r~4 but not at glucose 20 hiM. This glucose dependency may be due to suppression of glucagon release by glucose; under these conditions, addition of anti-glucagon sert~m may have no further effect. Exogenously added glucagon probably also enhanced somatostatin secretion by D cells. However, since stimulation of insulin secretion was observed, the direct effect of glucagon upon the B cells apparently was enough to negate the inhibitory effect of endogenously released sc~atostatin. Kadowaki et al. (9) have shown wi N the i~. lated perfused rat pancreas a dose effect of exogenous glucagon (i0 to i0 M) upon somatostatin and insulin release. With higher glucagon, more scmatostatin was released. Glucagon at any concentration enhanced insulin release cc~pared to absence of glucagon, but glucagon-stimulated insulin release was less at the higher glucagon concentrations. In order to confirm the mechanism whereby anti-glucagon serum enhances insulin release, additional studies are required. These must be directed toward the measurement of somatostatin secretion. Although others have reported successful assays for s(m~atostatin secretion by menolayer islet cell cultures similar to the ones used by us, we have not been able to detect somatostatin secretion (W.Y. Fujimoto and J.W. Ensinck, unpublished). However, cells containing somatostatin immunoreactivity are present in the culture. In conclusion, endogenously released scmatostatin inhibits insulin secretion. In addition, endogenously released glucagon appears to have a restraining effect upon insulin secretion since neutralization of glucagon by anti-glucagon serum enhanced insulin secretion. This effect is probably mediated through glucagon stimulation of somatostatin secretion. Exogenously added glucagon has a net stimulatory effect upon insulin secretion.
1878
Paracrine Regulation of Islet B Cells
Vol. 32, No. 6, 1983
Acknowledc.~nents We are grateful to Dr. A. Kaneto for the kind donation of rabbit anti-glucagon serum; Tcmoko Chinone, MarikoFujita, and Jeanette Teague for their expert technical assistance, and Sharon Kemp for her careful preparation of this manuscript. This work was supported in part by NIH grants AM 15312 and AM 17047. Part of this work was done while Dr. Fujimoto was on sabbatical leave at the University of Tokyo and a recipient of a fellowship from the Japan Society for the Promotion of Science. References i. 2. 3.
4. 5. 6. 7. 8. 9. i0. ii.
12. 13. 14. 15. 16.
W.Y. FUJIMOTO, J.W. ENSINCK and R.H. WILLIAMS, Life Sci. 15 1999-2004 (1974). Y. KANAZAWA, T. KUZUYA, T. IDE and K. KOSAKA, Am. J. Physiol. 211 442-448 (1966). W. STRANSE, in First International Syr~posium on Immunoenzymatic Techniques (INSERM Slmlo. No. 2), G. Feldman, P. Druet, J. Bignon, and S. Avrameas, eds., p. 117 North Holland, Amsterdam (1976). D.G. BASKIN, K.C. GORRAY and W.Y. FLUIMOTO, J. Histoche~. Cytochem. 29 567-571 (1981). E. SAMOLS, G. MARRI and V. MARKS, Lancet II 415-416 (1965). G. SOMERS, B. BLONDEL, L. ORCI and W.J. MALAISSE, Endocrinology 104 255264 (1979). G.S. PATION, E. IPP, R.E. DOBBS, L. ORCI, W. VALE and R.H. UNGER, Proc. Nat. Acad. Sci. USA 74 2140-2143 (1977). G.C. WEIR, E. SAMOLS and Y.C. PATEL, Metabolism 27 (suppl i) 1223-1226 (1978). S. KADOWAKI, T. TAMINATO, T. CHIBA, K. MORI, H. ABE, Y. GOTO, Y. SEINO, S. MATSUKURA, M. NOZAk~ and T. FUJITA, Diabetes 28 600-603 (1979). P. SCHAUDER, C. McINTOSH, J. ARENDS, R. ARNOLD, H. FRERICHS and W. CREUTZFELDT, FEBS Lett. 68 225-227 (1976). H. TANI~JCHI, M. UTSUMI, M. HASEGAk%, T. KOBAYASHI, Y. WATANABE, K. MURAKAMI, M. SEKI, A. TSUTOU, H. MAKIMURA, M. SAKADA and S. BABA, Diabetes 26 700-702 (1977). M. ITOH, L. MANDARINO and J.E. GERICH, Diabetes 29 693-696 (1980). K. TANIGAWA, H. KUZUYA, H. SAKURAI, Y. SEINO, S. SEINO, K. TSUDA and H. IMURA, Horm. Metab. Res. 13 78-80 (1981). R.N. HONEY, A. ARIMURA and G.C. WEIR, Endocrinology 109 1971-1974 (1981). N. B A B D ~ , M. LAVOIE, A. DUPONT, J. COTE and J.-P. COTE, Endocrinology i01 635-638 (1977). R.L. SOR~NSON, D.V. LIND,.T. and R.P. ELDE, Diabetes 29 747-751 (1980).