Bruce S. Chertow

Cellular Mechanisms of Insulin Release: Effects of Retinoids on Rat Islet Cell-to-Cell Adhesion, Reaggregation, and Insulin Release

We studied the effects of retinoids on islet cell-to-cell adhesiveness and glucose-induced insulin release from rat islets. For adhesion studies, islets were dispersed using low concentrations of trypsin. Thirteen cis-retinoic acid (13 cis-RA) was added to a suspension of 15 × 105 islet cells and adhesion of cells was quantitated using a hemocytometer. For functional studies, we measured biphasic insulin release from collagenase-isolated perifused islets, dispersed cells, and single large aggregates (clumps) of islet cells. Thirteen cis-RA (10−4 M) stimulated insulin secretion at 9.7, 12.5, 16.7, and 27.7 mM glucose. Maximal effects of 13 cis-RA (174% of control) were evident during second phase release at 9.7 mM glucose. Thirteen cis-RA (10−7 and 10−6 M) caused cells to adhere to each other, and at higher concentrations, 13 cis-RA caused dispersed cells to reaggregate into a single clump. These retinoid-induced clumps were perifused in a Bio-Gel P-2 gel column. Secretion from the clump was twofold greater than from an equal number of perifused dispersed cells. Electron microscopic and freeze-fracture examination of the clump showed reaggregated cells to be intact and the presence of gap junctions between cells. In conclusion, 13 cis-RA has marked effects on islet cell-to-cell adhesiveness. Trypsin-dispersed cells reaggregated by 13 cis-RA have anatomical contacts and secrete more insulin as an aggregate than as dispersed cells. Thirteen cis-RA increases insulin release possibly by increasing adhesion or interactions between β-cellS.

Islet insulin release and net calcium retention in vitro in vitamin D-deficient rats

In our previous studies, perifused islets from vitamin D-deficient (D-def) rats showed marked impairment of glucose-induced biphasic release, accounted for at least in part by a decrease in food intake. In studies reported here, we test whether D-def rat islets have an impaired response to 5.6 mM glucose or tolbutamide, (T), and if so, whether this impairment is related to a decrease in food intake or a defect in islet calcium metabolism. We isolated islets of normal rats, D-def rats, and rats pair fed (PF) to D-def rats. Biphasic insulin release from perifused islets and net 45Ca retention in lot-incubated islets were measured in response to 5.6 mM glucose, 0.37 mM T, or both. Compared with secretion from normal islets, biphasic insulin release from islets of both D-def rats and PF rats was diminished by 50% in response to 5.6 mM glucose alone or 5.6 mM glucose plus T. Insulin secretion was not significantly different between islets of D-def rats and islets of PF rats. In contrast, net calcium retention in islets of D-def rats was decreased to 68% of retention in islets of PF rats. However, net calcium retention in islets of both PF and D-def rats increased in response to T. The pair-feeding experiments suggest that the decrease in insulin release from islets of D-def rats is due to the decrease in food intake associated with the D-def state. On the other hand, the defect in calcium retention in islets of D-def rats raises the possibility that vitamin D may have a specific effect on islet calcium metabolism. In this case, the mechanism of impaired insulin release in islets of D-def rats would be different from that in islets of PF rats and would involve a defect in intracellular calcium handling.

Vitamin A stimulation of insulin secretion : Effects on transglutaminase mRNA and activity using rat islets and insulin-secreting cells

Retinol or retinoic acid is required for insulin release. Retinoids increase transglutaminase activity, and transglutaminase has been implicated in islet insulin release. To examine whether transglutaminase could mediate effects of retinoids on insulin secretion, we measured (i) transglutaminase activity in islets from rats deficient in vitamin A or repleted with retinol or retinoic acid, (ii) transglutaminase activity in RINm5F and INS-1 insulin-secreting cells cultured in retinol or retinoic acid, (iii) mRNA for transglutaminase in RINm5F and INS-1 cells, and (iv) insulin secretion from INS-1 cells in response to retinoic acid. Islets from rats repleted with retinol or retinoic acid showed more than twice the transglutaminase activity of islets from vitamin A-deficient rats. Retinoic acid increased RINm5F cells and INS-1 cell transglutaminase activity. Retinol did not increase transglutaminase activity. Transglutaminase mRNA was detected in INS-1 cells but not in RINm5F cells. Retinoic acid increased insulin secretion from INS-1 cells as observed previously in RINm5F cells. In conclusion, retinoic acid increases transglutaminase activity in both rat islets and two insulin-secreting cell lines. Retinoic acid stimulates insulin secretion from INS-1 cells. Transglutaminase is a candidate for mediating retinoid-induced changes in insulin secretion.

Vitamin A status affects the development of diabetes and insulitis in BB rats

BB/Wor rats develop autoimmune diabetes mellitus with many features in common with human insulin-dependent diabetes mellitus. Since retinoids are known to have effects on insulin secretion and immune function, these studies were designed to investigate the effects of retinoid deficiency on diabetes in BB/Wor rats and to identify a role for retinoid status in the pathogenesis of autoimmune diabetes mellitus. Litters of diabetes-prone (DP) and diabetes-resistant (DR) BB/Wor rats were divided at weaning and fed a diet either (1) devoid of retinoids and leading to clinical deficiency at approximately 60 days of age (A-def diet)-following 10 days of clinical deficiency, rats on the A-def diet were changed to a diet containing 2 microg/g retinoic (A-def/RA diet); (2) containing 2 microg/g retinoic acid but deficient in retinol (RA diet); or (3) replete in retinol with 4 microg/g retinyl palmitate (RP diet). Rats receiving RP or RA diets were pair-fed to rats on the A-def/RA diet. Diabetes by 120 days of age was greatly reduced (P < .01) in DP rats that received the A-def/RA diet (four of 27) or RA diet (four of 29) versus the RP diet (13 of 31). Insulitis progressed with age in nondiabetic DP rats receiving the RP diet (P < .02) or RA diet (P < .05), but not the A-def/RA diet (P > .22). Insulin secretion was measured in perfused pancreas of nondiabetic rats after age 120 days and correlated negatively with insulitis (P < .05). DP rats receiving the RP diet had reduced insulin secretion as compared with other DP and DR rats (P < .05). In DR rats, retinoid status had no effects on insulitis through 120 days of age or on insulin secretion after 120 days of age. In conclusion, retinol deficiency reduces diabetes and insulitis in DP BB/Wor rats, and retinoic acid can at least partly substitute for retinol in the development of insulitis.

Increasing incidence and characteristics of differentiated thyroid cancer in Huntington, West Virginia.

Since 1985, we have observed an increasing number of differentiated thyroid cancer cases in Huntington, West Virginia. We describe tumor incidence, patient and tumor characteristics, treatment modalities, and tumor recurrence and death. One hundred seventeen patients with differentiated thyroid cancer were identified between 1976 and 1999. Data were collected from patient records in our practice and the tumor registries at the three hospitals serving our community. The annual incidence of differentiated thyroid cancer increased significantly from fewer than 3 cases per 100,000 prior to 1996 to 9.4 cases per 100,000 in 1999. The median age at diagnosis was 49 years (range, 16-80). The median tumor size was 2.5 cm (range, 1.2-10). Forty-seven percent of the patients had bilateral disease, 28% had three or more tumors, 44% had thyroid capsular invasion, and 16% had gross extrathyroid invasion at surgery. Twenty-two percent had cervical lymph node involvement and 9% had distant metastases at diagnosis. During 1-month to 23-year follow-up, 11% had recurrence, and 5% died of thyroid cancer. In summary, differentiated thyroid cancer has increased dramatically in our community. The tumors appear to be aggressive at diagnosis as reflected by the high percentage of tumors with bilateral, multicentric, and locally invasive disease.

Effects of all-trans-retinoic acid (ATRA) and retinoic acid receptor (RAR) expression on secretion, growth, and apoptosis of insulin-secreting RINm5F cells

To define the functions of retinoids and their receptors in insulin secretion, we tested the effects of all-trans-retinoic acid (ATRA) and retinoic acid receptor (RAR) expression on cell growth, differentiation, and secretion using insulin-secreting RINm5F cells. Wild-type cells with a low abundance of mRNA for RARβ were transfected with RARβ or chloramphenicol acetyltransferase (CAT control). Cells were cultured for 2–7 days in media without (A-def) or with ATRA, 1, 10, 100, and 1,000 nM. At day 2 of culture, ATRA stimulated insulin release in wild-type and transfected cells, and this effect was dose dependent. At 7 days, ATRA stimulated insulin secretion from wild-type cells twofold at glucose concentrations of 0.5 mM (A-def, 5.1 ± 0.27; ATRA, 1,000 nM, 10.5 ± 1.43 ng/106 cells) and at 11.0 mM (A-def, 6.9 ± 0.24; ATRA, 1,000 nM, 13.6 ± 1.86 ng/106 cells). The cellular insulin content was increased about threefold (A-def, 39.2 ± 2.95; ATRA, 1,000 nM, 118 ± 8.54 ng/106 cells). ATRA inhibited growth of wild-type cells as early as 3 days, and this effect was dose dependent. Whereas in the absence of ATRA, the cell number increased over fivefold between day 3 and day 5, ATRA, 1,000 nM, inhibited cell growth completely. ATRA, 1,000 nM, increased apoptotic RINm5F cells (day 3 A-def, 0.53 ± 0.27% of total cells, and ATRA, 2.30 ± 1.44; day 5 A-def, 0.38 ± 0.23, and ATRA, 2.14 ± 0.59; day 7 A-def, 0.90 ± 0.29, and ATRA, 6.02 ± 1.64). RARβ-transfected cells showed overexpression of mRNA to RARβ and dose-dependent inhibition of growth, with almost-complete inhibition at ATRA concentrations as low as 100 nM. Overexpression of RARβ increased insulin secretion at ATRA, 100–1,000 nM. In summary, ATRA increased the insulin secretion and content of RINm5F cells, while inhibiting growth and increasing apoptosis. Increased expression of RARβ facilitated these effects on growth and secretion. These findings may reflect the known effect of ATRA on differentiation of cells and mediation through RARβ.

Effects of vitamin A deficiency and repletion on rat glucagon secretion

To determine whether vitamin A is involved in pancreatic a cell function, we tested for (a) effects of vitamin A deficiency on glucagon release from perifused islets and perfused pancreases, and (b) the presence of cytosolic retinol-binding proteins (CRBP) and retinoic acid-binding proteins (CRABP), in the glucagon-secreting a cell line, In-Rl-G9. Arginine 19 mM plus glucose 2.8 M-stimulated glucagon secretion was markedly impaired in islets and pancreases of vitamin A-deficient rats or rats that had at some time been cycled through vitamin A deficiency (ever A-def) despite repletion with retinoids for 2–4 weeks. Insulin secretion was impaired likewise. Repletion starting early in the development of vitamin A deficiency and for a longer period of time (18 or 60 days) did not restore glucagon secretion, but did normalize insulin secretion. CRBP and CRABP were present in In-RIG9 cells. We conclude that (a) vitamin A deficiency is associated with a defect in glucagon secretion; (b) The defect in secretion occurs early in the course of vitamin A deficiency; (c) The defect persists despite repletion; and (d) The requirement of vitamin A for secretion and the presence of CRBP and CRABP in glucagon-secreting cells support a physiologic role for vitamin A at the a cell level.

Retinoic Acid Receptor, Cytosolic Retinol-Binding and Retinoic Acid-Binding Protein mRNA Transcripts and Proteins in Rat Insulin-Secreting Cells

To define the mechanism of vitamin A action at the β-cell level, we tested for the presence of messenger RNA for retinoic acid receptors α, β, and γ; cytosolic retinol-binding protein; and cytosolic retinoic acid-binding protein in RINm5F cells, an insulin-secreting cell line, and determined whether cytosolic retinol-binding protein and cytosolic retinoic acid-binding protein are present in isolated purified normal rat β-cells. Northern blot analyses showed two transcripts of retinoic acid receptor α messenger RNA (3.8 and 2.4 kb), one transcript of retinoic acid receptor messenger RNA (3.8 kb), and one transcript of cytosolic retinol-binding protein (0.9 kb) in RINm5F cells. Ribonuclease protection assays also showed the presence of cytosolic retinol-binding protein and cytosolic retinoic acid-binding protein in RINm5F cells. Quantitatively, cytosolic retinol-binding protein levels were 0.10 ± 0.02 pg/micrograms total RNA. Using specific radioimmunoassays, normal isolated purified rat β-cells contained CRBP (19.2 ± 2.38) and cytosolic retinoic acid-binding protein (16 ± 0.53 ng/10(6) cells). The presence of message for retinoic acid receptors α and γ, cytosolic retinol-binding protein, cytosolic retinoic acid-binding protein, and the gene products of cytosolic retinol-binding protein and cytosolic retinoic acid-binding protein in insulin-secreting cells support a mechanism of vitamin A action and role for cytosolic and nuclear receptors at the β-cell level similar to that suggested in nonendocrine cells. The presence of nuclear retinoic acid receptor a. and 7 suggests that vitamin A may affect insulin secretion through gene expression in the β-cell.

Retinoid-X receptors and the effects of 9-cis-retinoic acid on insulin secretion from RINm5F cells☆

Retinoid-X receptors (RXRs) are 9-cis-retinoic acid (9CRA)-dependent gene transcription factors, which modulate the action of all-trans-retinoic acid (ATRA), fatty acids, thyroid hormone (TH), and vitamin D (VD) by forming dimers with themselves or ATRA, TH, peroxisome proliferator activator receptors (PPARs), or VD receptors (VDRs). To determine if 9CRA and RXRs have a role in secretion, RINm5F cells were assayed for RXR transcripts and effects of 9CRA and ATRA on secretion. A single RXR alpha transcript and two RXR beta transcripts, but not RXR gamma, were evident by Northern blot. Cells were cultured for 48 hours without and with 9CRA 1 to 1,000 nmol/L and then stimulated with glucose 0, 0.5, 2.8, 7, and 11 mmol/L 9CRA increased secretion at each glucose concentration, 9CRA increased secretion by 50% to 100% (ANOVA, P < .001) with consistent concentration-dependent responses (eg. at glucose 2.8 mmol/L 9CRA: 0 nmol/L, 5.02 +/- .20 ng/(10(6) cells.h); 1 nmol/L, 6.97 +/- .30; 10 nmol/L, 8.36 +/- .18; 100 nmol/L, 9.15 +/- .28; 1,000 nmol/L, 10.24 +/- .24; n = 6). Although RINm5F cells respond slightly if at all to glucose, 9CRA facilitated glucose-induced insulin release (eg, at 9CRA 100 nmol/L, glucose: 0.5 mmol/L, 7.47 +/- .22 ng/(10(6) cells.h); 2.8 mmol/L, 9.15 +/- .27; 7 mmol/L, 9.81 +/- .19; 11 mmol/L, 11.16 +/- .23; n = 6). ATRA increased secretion by 28% to 57% (ANOVA, P < .001: at glucose 2.8 mmol/L, ATRA: 0 nmol/L, 6.17 +/- .32 ng/(10(6) cells.h); 1 nmol/L, 7.91 +/- .29; 10 nmol/L, 9.75 +/- .14; 100 nmol/L, 9.66 +/- .33; n = 6). 9CRA was more potent than ATRA (eg, at 2.8 mmol/L; baseline, 8.17 +/- .32 ng/(10(8) cells.h); ATRA 100 nmol/L, 9.66 +/- .33; 9CRA 100 nmol/L, 10.81 +/- .15; P < .05, n = 6). When 9CRA was combined with ATRA, the combination was not additive or synergistic (eg, at 2.8 mmol/L: ATRA 100 nmol/L, 9.66 +/- .33 ng/(10(6) cells.h); 9CRA 100 nmol/L, 10.81 +/- .15; ATRA 100 nmol/L + 9CRA 100 nmol/L, 10.79 +/- .28; P < .05, n = 6). These studies show that (1) 9CRA stimulates insulin secretion from RINm5F cells. This effect appears to be at least equal to if not greater than that observed with ATRA, but additive or synergistic effects with ATRA were not evident; (2) 9CRA may facilitate glucose-induced release; and (3) multiple RXR transcripts are present in insulin-secreting cells, implying specific functions. Our findings support the idea that the effects of 9CRA on insulin secretion are mediated through RXR homodimers or heterodimers with retinoic acid receptors (RARs) or possibly other nuclear receptors. Retinoid deficiency or alterations in retinoid receptor function could lead to abnormalities of cell growth or secretion.

Cytoplasmic Retinoid-Binding Proteins and Retinoid Effects on Insulin Release in RINm5F β-Cells

Vitamin A (retinol) is required for insulin secretion, and retinoic acid substitutes for retinol in this function. To determine if retinol acts at the β-cell level, we assayed β-cells of the rat insulinoma (RINm5F) line for cytosolic retinol- and retinoic acid-binding proteins (CRBP and CRABP) by radioimmunoassay (RIA) and [3H]retinol and [3H]retinoic acid binding to cytosol extracts. Furthermore, we tested whether insulin release from cells was affected by addition of retinol or retinoic acid to culture medium. RINm5F cells were grown to near confluence before assay of CRBP and CRABP. Scatchard analysis showed the Kd for retinol to be ∼6 nM at a level of 4.5 pmol/mg protein or 300,000 sites/cell. Sucrose density-gradient assay showed single discrete peaks migrating at 2S for both retinol and retinoic acid. RIA of whole-cell extracts showed CRBP and CRABP levels of 5.27 ± 0.41 and 2.95 ± 0.75 pmol/mg protein, respectively. Retinol (1.75 μM) and retinoic acid (0.175 and 1.75 μM) increased KCI-induced insulin release. Considered together, the presence of CRBP and CRABP in a β-cell line and the increase in KCI-induced insulin release by retinol and retinoic acid are consistent with the idea that retinol has a functional role in insulin secretion and suggest a potential mechanism of action at the β-cell level similar to that observed in other retinoid-responsive cells.