Dexamethasone Induces Neuropeptide Y (NPY) Expression and Impairs Insulin Release in the Insulin-producing Cell Line RINm5F RELEASE OF NPY AND INSULIN THROUGH DIFFERENT PATHWAYS*

Neuropeptide Y (NPY) occurs in adrenergic as well as in non-adrenergic nerves innervating the islets of Langerhans and inhibits glucose-stimulated insulin secretion. Recently we demonstrated that NPY is expressed within islet beta cells of the rat pancreas following treatment with dexamethasone in vivo . In this study we examined the cellular expression of NPY following dexamethasone treatment of the insulin-producing cell line RINm5F, which under control conditions does not express or release NPY. The cells were cultured with or without dexamethasone (100 n M ) for 5 days. Over the 5-day culture period, dexamethasone time dependently induced an increased release of NPY with a concomitant decrease in the release of insulin. Northern blot and in situ hybridization revealed a corresponding time-dependent increase in the amount of NPY transcripts and in the number of cells labeled for NPY mRNA, whereas immunocytochemistry for NPY revealed only a few immunoreactive cells, indicating a rapid release of the formed peptide. Following 5 days of culture with dexamethasone, acute stimulation with D -glyceralde-hyde (10 m M ) or KCl (20 m M ) Ca 2 (cid:49) dependently stimulated the release of insulin. In contrast neither stimulation with D -glyceraldehyde or KCl nor removal of extracellular Ca 2 (cid:49) affected the release of NPY. yielding a specific 9 carried out overnight at 37 using a of first in (cid:51) SSC (0.15 M at then in 0.5 (cid:51) (four times and finally in SSC for at followed by were in °C They were then counterstained hematoxylin. purposes, hybridization was also performed after or in the presence of a excess the no


Neuropeptide Y (NPY) occurs in adrenergic as well as in non-adrenergic nerves innervating the islets of Langerhans and inhibits glucose-stimulated insulin secretion. Recently we demonstrated that NPY is expressed within islet beta cells of the rat pancreas following treatment with dexamethasone in vivo.
In this study we examined the cellular expression of NPY following dexamethasone treatment of the insulin-producing cell line RINm5F, which under control conditions does not express or release NPY. The cells were cultured with or without dexamethasone (100 nM) for 5 days. Over the 5-day culture period, dexamethasone time dependently induced an increased release of NPY with a concomitant decrease in the release of insulin. Northern blot and in situ hybridization revealed a corresponding timedependent increase in the amount of NPY transcripts and in the number of cells labeled for NPY mRNA, whereas immunocytochemistry for NPY revealed only a few immunoreactive cells, indicating a rapid release of the formed peptide. Following 5 days of culture with dexamethasone, acute stimulation with D-glyceraldehyde (10 mM) or KCl (20 mM) Ca 2؉ dependently stimulated the release of insulin. In contrast neither stimulation with D-glyceraldehyde or KCl nor removal of extracellular Ca 2؉ affected the release of NPY. Furthermore the D-glyceraldehyde-and KCl-induced increase in cytosolic Ca 2؉ , evident in control RINm5F cells, was impaired after dexamethasone treatment. We conclude that RINm5F cells show steroid-sensitive plasticity and express NPY after dexamethasone treatment concomitantly with a decreased insulin secretion and impaired increase in cytosolic Ca 2؉ upon depolarization with KCl or stimulation with D-glyceraldehyde. We also conclude that NPY and insulin secretion are regulated differently and suggest that the inability of the removal of extracellular Ca 2؉ to inhibit NPY secretion and the failure of D-glyceraldehyde and KCl to stimulate NPY secretion reflect a constitutive release of this peptide from the cells in contrast to the regulated release of insulin.
Neuropeptide Y (NPY) 1 is a widely distributed neurotransmitter in the mammalian central and peripheral nervous systems (1,2) and belongs to the NPY family of peptides also comprising pancreatic polypeptide (PP) and peptide YY (PYY) (3) (for review, see Ref. 2). In the pancreas of several species, including the rat, NPY has been demonstrated in adrenergic as well as in non-adrenergic nerve fibers around blood vessels, ducts, and acini as well as within the islets (4,5). In the pancreas NPY inhibits glucose-stimulated insulin secretion as demonstrated both in vitro and in vivo (6,7). Recent studies have revealed expression of NPY also in islet cells, e.g. in hamster islet somatostatin cells (8), in mouse islet glucagon cells (9), and in rat islet cells following in vivo treatment or treatment of isolated islets with dexamethasone (10 -12). The dexamethasone-induced expression of NPY in rat islet cells has been localized predominantly to the insulin-producing beta cells (11,12). However, the mechanisms of dexamethasoneinduced NPY expression and the regulation of NPY secretion from insulin-producing cells after dexamethasone treatment are not yet established.
To study whether insulin-producing cells show steroid plasticity and display induced expression and secretion of NPY after dexamethasone treatment in a defined environment without influence of peptides derived from other islet cells, we examined the RINm5F cells by using Northern blot, in situ hybridization (ISH), and immunocytochemistry. Furthermore, we examined the release of insulin and NPY by radioimmunoassay from the RINm5F cells after dexamethasone treatment during basal conditions and after stimulation with D-glyceraldehyde and KCl in the presence or absence of extracellular Ca 2ϩ . The levels of cytosolic Ca 2ϩ in control-and dexamethasone-treated cells during basal and stimulated conditions were also measured.

EXPERIMENTAL PROCEDURES
Cell Culture-Stock cultures of RINm5F cells were grown continuously at 37°C in humified air equilibrated with 5% CO 2 in RPMI 1640 medium supplemented with 2 mM L-glutamine, 10% fetal calf serum, 100 units/ml penicillin G, 100 g/ml streptomycin, and 2.5 g/ml amphothericin B. Cells were subcultured once weekly after trypsination and the medium was changed every 3-4 days. All experimental work was performed with cells in passages number 76 -80.
For ISH, slides were rinsed twice with PBS and then fixed for 10 min in buffered 4% paraformaldehyde (pH 7.2), permeabilized in 0.25% Triton X-100, rinsed in PBS for 5 min, and dehydrated. For detection of NPY mRNA, a 36-mer probe complementary to the nucleotide sequence 266 -301 of rat NPY cDNA was used. The probe has been checked for complementarity to other genes present in the GenBank TM data base, using the NCBI BLAST E-mail server, but no such complementarity was found (11,12). For detection of insulin mRNA, a probe mixture consisting of six different 30-mer oligodeoxyribonucleotides was used (BPR 236; R & D Systems, Abingdon, Oxfordshire, UK). ISH was also performed for glucagon mRNA using a probe complementary to nucleotides 153-182 of rat glucagon cDNA, for somatostatin mRNA using a probe complementary to nucleotides 310 -339 (code NEP-503, DuPont NEN, Solna, Sweden) of rat somatostatin cDNA, for PYY and PP mRNAs using probes complementary to nucleotides 266 -295 of rat PYY cDNA, and nucleotides 265-294 of rat PP cDNA. The probes were 3Ј-end-labeled with 35 S-dATP using a terminal transferase (both supplied by DuPont NEN, Stockholm, Sweden) yielding a specific activity of approximately 2 ϫ 10 9 cpm/g. Hybridization was carried out overnight at 37°C in sealed moisturizing chambers using a probe concentration of 1.3 pmol/ml. After hybridization, the slides were washed first in 1 ϫ SSC (0.15 M NaCl, 0.015 M sodium citrate) at room temperature, then in 0.5 ϫ SSC (four times for 15 min at 55°C), and finally in 1 ϫ SSC (once for 30 min at room temperature), followed by dehydration. The slides were coated with Ilford K-5 emulsion and stored in light-sealed boxes at 4°C for 3-4 days. They were then developed and counterstained with hematoxylin. For control purposes, hybridization was also performed after incubation in RNase A (Sigma) (45 g/ml for 30 min at 37°C) or in the presence of a 100-fold molar excess of unlabeled probe. In the control experiments no autoradiographic labeling was obtained.
The percentage of NPY mRNA-labeled cells of the total number of cells counted was calculated for each day during the 5-day culture period with and without dexamethasone. At least 2 slides from three separate experiments for each day (control and dexamethasone-treated cells) were examined. In each slide approximately 1000 cells, in randomly selected visual fields in the microscope, were counted and checked for NPY mRNA labeling.
NPY and Insulin Release during Five Days of Culture/Long Term Incubation-Cells were seeded in 24-well plates at a density of 15 ϫ 10 4 cells per well (n ϭ 6) and cultured for 5 days with or without dexamethasone (100 nM). The medium was replaced, and 800 l were sampled from each well daily and then centrifuged at 200 rpm for 5 min. Aliquots for assay of insulin (50 l) and NPY (100 l) were stored at Ϫ20°C until assayed.
The number of cells per well was counted each day during the culture period. After removal of the medium, the cells were rinsed with 2 ϫ 500 l of PBS. Washed cells were then trypsinated (200 l, Life Technologies, Inc., Paisley, Scotland) and collected in vials. To inactivate the trypsin, 200 l of 10% fetal calf serum were added to each vial. The vials were then centrifuged at 1000 rpm for 5 min. The supernatant was discarded, and the remaining cell pellet was resuspended with 2 ml of RPMI 1640. The cells were then counted in a Bü rker chamber.
Acute Stimulation of NPY and Insulin Release/Short Term Incubation-After 3 days of culture in flasks with or without dexamethasone (100 nM) cells were detached with trypsin, then seeded in 24-well plates at a concentration of 23 ϫ 10 4 cells per well (n ϭ 6), and cultured for 2 more days (reaching about 80% confluence) with or without dexamethasone. The cells were then washed twice in a HEPES medium containing in mM: 125 NaCl, 5.9 KCl, 1.28 CaCl 2 , 1.2 MgCl 2 , 25 HEPES, and 0.1% human serum albumin (pH 7.36). Thereafter, the cells were incubated in the HEPES medium in a volume of 200 l supplemented with 3.3 mM glucose, 10 mM D-glyceraldehyde, or 20 mM KCl. The influence of Ca 2ϩ deficiency was studied by excluding CaCl 2 in the incubation medium at incubations with 3.3 mM glucose and 20 mM KCl. After a 60-min incubation at 37°C in 95% O 2 , 5% CO 2 , 150 l of the medium were collected and centrifuged at 200 rpm for 5 min. Aliquots for insulin (50 l) and NPY (50 l) were then frozen and stored at Ϫ20°C until assayed.
NPY and Insulin Assays-For the radioimmunoassay of NPY, a rabbit antiserum to synthetic porcine NPY was used (gift from Dr. P. C. Emson, Cambridge, UK). Porcine 125 I-NPY was used as a tracer. The antiserum used has been characterized previously (13). For the radioimmunoassay of insulin, we used a guinea pig anti-rat insulin antibody, 125 I-labeled human insulin, as tracer and rat insulin as standard (Linco Research, St. Louis, MO). The separation of free and bound radioactivity was performed by the double antibody technique.
RNA Extraction and Northern Blot Analysis-Total RNA was isolated from cells cultured without or with dexamethasone (100 nM) for 1-5 days by extraction in guanidium thiocyanate using the single step method described by Chomczynski and Sacchi (14). RNA was quantified spectrophotometrically. In each lane, 20 g of total cellular RNA were separated electrophoretically on a denaturing 1.2% agarose-formaldehyde gel and transferred to a nylon membrane (Hybond-N, Amersham International plc, Bucks, UK) by capillary transfer in 20 ϫ SSC. RNA was cross-linked to the filter by exposure to UV light (120,000 J). Prehybridization was performed at 63°C for 60 min. Hybridization with the [␥-32 P]ATP 5Ј-end-labeled (Amersham International plc) NPY oligoprobe (same sequence as used for ISH) (1.6 ϫ 10 6 cpm/ml hybridization solution) was carried out at 63°C overnight. The filter was then washed as follows: twice for 15 min, 2 ϫ SSC, 0.1% SDS, at room temperature; once for 15 min, 0.1 ϫ SSC, 0.1% SDS, at room temperature; and once for 15 min, 0.1 ϫ SSC, 0.1% SDS at 60°C. As control for the amount of total RNA application, the filter was hybridized with a randomly [␣-32 P]dCTP-labeled actin probe (15). Autoradiography was performed at Ϫ70°C for 5 days for both actin and NPY by exposing the filter to imaging film (CEA AB, Strä ngnä s, Sweden) and developed according to standard procedures. Densitometric laser scanning of each autoradiograph was performed. Quantitation of NPY mRNA was obtained by expressing the autoradiographic signal after hybridization with the NPY probe relative to the autoradiographic signal after hybridization with the actin probe.
Measurement of Intracellular Ca 2ϩ -The free cytoplasmic calcium concentration [Ca 2ϩ ] i was analyzed with the fluorescent probe FURA-2AM (Sigma) as described elsewhere (16). Cells cultured in control medium or dexamethasone-supplemented medium for 1, 3, and 5 days were trypsinated and thereafter allowed to recover at a concentration of 0.5 ϫ 10 6 cells/ml for 2 h in 10 ml of RPMI 1640 medium supplemented with 10% fetal calf serum at 37°C. During this period the cells were continuously shaken to avoid attachment. The cells were then loaded with 1 M indicator for 45 min, rinsed three times in HEPES buffer, and allowed to equilibrate for 20 min in the HEPES medium at room temperature. The cells were again rinsed, counted in a Bü rker chamber, resuspended in the HEPES medium at a concentration of 0.5 ϫ 10 6 cells/ml, and transferred to a cuvette for measurement of [Ca 2ϩ ] i . [Ca 2ϩ ] i was measured by dual wavelength spectrophotofluorimetry in a Perkin-Elmer LS-50 spectrophotofluorometer. The excitation wavelength was continuously altered between 340 and 380 nm, and the emission wavelength was 510 nm. The cell suspension was continuously stirred with a stir bar mounted at the side of the cuvette. D-Glyceraldehyde (10 mM) and KCl (20 mM) were added to the cuvette in microliter volumes from stock solutions of high concentrations at specific time points as indicated in the figures. When experimental and control series were compared, the substances under study were always added at identical time points from the start of the experiment: at 200 s for D-glyceraldehyde and at 900 s for KCl. At the end of each experiment, fluorescence maximum was obtained by adding 0.03% Triton X-100, and fluorescence minimum was obtained by adding EGTA in excess to the cell suspensions. The [Ca 2ϩ ] i was calculated according to the previously described formula (17). The K d was assumed to be 224 nM.
Statistical Analysis-The data are presented as the mean Ϯ S.E. Statistical comparisons of differences between mean values were performed by the use of the Student's t test or with the Mann-Whitney U test. Bonferroni correction was used when multiple comparisons were performed.

Dexamethasone Induces NPY Expression in RINm5F
Cells-ISH performed on control RINm5F cells displayed no labeling for NPY mRNA at any time (Fig. 1), and no signal for NPY mRNA could be detected with Northern blot (Fig. 2). In contrast, after dexamethasone treatment of RINm5F cells, a timedependent increase in the amount of NPY mRNA transcripts (Fig. 2) and in the number of NPY mRNA-labeled cells (Fig. 1) was observed. Thus, after 1 day of dexamethasone treatment, Northern blot displayed a weak signal (Fig. 2), and only 5 Ϯ 1% of the cells displayed labeling ( Fig. 1 and Fig. 3). However, after 5 days of culture with dexamethasone, there was a 13-fold increase in the amount of NPY mRNA transcripts compared with 1 day of dexamethasone treatment (Fig. 2), and 91 Ϯ 1% of the cells displayed intense labeling (Figs. 1 and 3). ISH for insulin mRNA revealed labeling of all cells, and there was no obvious difference in labeling intensity between dexamethasone-treated and control cells (data not shown). ISH for glucagon, somatostatin, PP, or PYY revealed no labeling for respective mRNA in dexamethasone-treated or control cells during the 5-day culture period. NPY immunoreactivity, located in the cell cytoplasm, was detected in a minor population of cells treated with dexamethasone for 2-5 days (Fig. 4A). In contrast, control cells (Fig. 4B) and cells cultured with dexamethasone for only 1 day did not stain for NPY. In all groups studied glucagon, somatostatin, PP, or PYY immunoreactivities could not be detected, whereas a subpopulation of cells displayed insulin immunoreactivity.
Dexamethasone Increases Release of NPY and Decreases Release of Insulin from RINm5F Cells-The release of NPY from dexamethasone-treated RINm5F cells increased in a time-dependent manner concomitantly with a decreased insulin release (Table I). Thus, the content of NPY in the medium after a 5-day culture period with dexamethasone was increased 58fold compared with 1 day of culture with dexamethasone (p Ͻ 0.001). In contrast NPY could not be detected in the medium of control cells (Table I). Insulin was detected in the medium of both control and dexamethasone-treated cells. In the medium of control cells, the insulin content increased gradually during the first 3 days of culture and then leveled off, whereas the insulin content in the medium of dexamethasone-treated cells was significantly lower compared with controls (p Ͻ 0.001) ( Table I) at all time points throughout the 5-day culture period. During the 5-day culture period, the cells displayed a steady increase in number. However, treatment with dexamethasone decreased cell recovery by approximately 30% at day 5 compared with control cells (data not shown). When calculating the release of insulin and NPY per cell for 24 h, insulin release from control and dexamethasone-treated cells decreased with time, and by day 5 had decreased to approximately 35% from control cells (p Ͻ 0.001) and approximately 5% from dexamethasone-treated cells (p Ͻ 0.001) compared with cells cultured for 1 day (Fig. 5). In contrast, the release of NPY displayed a time-dependent increase from cells cultured with dexamethasone by a 22-fold increase (p Ͻ 0.001) at day 5 compared with cells cultured with dexamethasone for 1 day (Fig. 5).
Dexamethasone Impairs D-Glyceraldehyde-and KCl-stimulated Increase of [Ca 2ϩ ] i -Base-line [Ca 2ϩ ] i levels in RINm5F cells was 89 Ϯ 3 nM. D-Glyceraldehyde (10 mM) induced a gradual increase in [Ca 2ϩ ] i . At 1 day after initiation of culture, [Ca 2ϩ ] i increased to 212 Ϯ 4 nM at 300 s after addition of D-glyceraldehyde (p Ͻ 0.001). A similar pattern of influence was observed at days 3 and 5 after initiation of culture (Fig. 6). When the cells were cultured in the presence of dexamethasone for 1 day, the D-glyceraldehyde-induced increase in [Ca 2ϩ ] i was the same as in control cells, both in pattern and magnitude (Fig. 6). However, the D-glyceraldehyde-induced increase in [Ca 2ϩ ] i was markedly impaired at day 3 when cultured with dexamethasone and more pronounced at day 5. In fact after 5 days of culture with dexamethasone, D-glyceraldehyde failed to increase [Ca 2ϩ ] i . In contrast, base-line [Ca 2ϩ ] i was not affected by dexamethasone. Depolarization by KCl (20 mM) markedly increased [Ca 2ϩ ] i due to opening of voltage-sensitive Ca 2ϩ channels (Fig. 6). After 5 days of culture in the presence of dexamethasone, this depolarization-induced increase in [Ca 2ϩ ] i was reduced by approximately 50% (p Ͻ 0.001).
Different Secretory Pathways for Insulin and NPY-Acute stimulation with 10 mM D-glyceraldehyde or 20 mM KCl for 60 min significantly increased insulin release from both control and dexamethasone-treated RINm5F cells (p Ͻ 0.001). However, the insulin secretory response to D-glyceraldehyde and KCl from cells treated with dexamethasone was much lower than the response from control cells (p Ͻ 0.001) ( Table II). Removal of extracellular Ca 2ϩ abolished the KCl-stimulated release of insulin from both control and dexamethasone-treated cells (Table II). In contrast neither stimulation with D-glyceraldehyde or KCl nor removal of Ca 2ϩ from the incubation medium affected the release of NPY from cells cultured with dexamethasone as determined with radioimmunoassay (Table  II). NPY could not be detected in the incubation medium from cells cultured without dexamethasone. DISCUSSION Although NPY is widely distributed in the nervous system, it has been demonstrated also in pancreatic endocrine cells in the hamster (8), mouse (9), human (18), and rat when treated with glucocorticoids (10 -12) and during fetal development (19). In the dexamethasone-treated rat, we recently showed that NPY is expressed in insulin-producing beta cells (11,12). Glucocorticoids administered in vivo are known to induce peripheral insulin insensitivity with elevated plasma insulin and glucose levels associated with hypertrophy and proliferation of the beta cells (20; for review see Ref. 21). Since NPY has been shown to inhibit glucose-stimulated insulin secretion (6, 7), it is possible that the expression of NPY in islet cells after dexamethasone  has a modulatory role in intraislet hormonal regulation during glucocorticoid-induced islet adaptation in vivo. However, the mechanisms involved in the regulation of NPY expression in beta cells and its potential importance for islet function are not yet established.
In the present study we demonstrate that dexamethasone induces expression of NPY also in the insulin-producing cell line RINm5F. This implies that expression of NPY in insulinproducing cells following dexamethasone treatment occurs in a defined environment without influences from peptides produced in islet non-beta cells. Although some RINm5F cell cultures have been reported to contain minor amounts of glucagon and somatostatin (22, 23), we were not able to detect these islet peptides, nor PP or PYY in these cells. The previous findings of glucocorticoid-responsive elements in the promoter region of the rat NPY gene (24) and glucocorticoid receptors on the beta cells (25) strengthen the evidence that the expression of NPY is induced by a specific effect of dexamethasone on the cells.
The expression of NPY in RINm5F cells increased in a timedependent manner when treated with dexamethasone. After 5 days of culture with dexamethasone the vast majority of cells displayed labeling for NPY mRNA, and Northern blot revealed an intense signal for NPY mRNA. Also measurement of immunoreactive NPY in the culture medium from dexamethasonetreated cells revealed gradually increased levels of the peptide. However, immunocytochemistry for NPY performed on dexamethasone-treated cells grown on glass slides revealed only a few immunoreactive cells even after 5 days of culture. These findings indicate a rapid release and poor intracellular storage of NPY. The release rate and pattern of NPY seem therefore to be different from that of insulin from dexamethasone-treated RINm5F cells prompting a comparison of the regulation of the release of NPY and insulin.
D-Glyceraldehyde and KCl stimulate insulin secretion from RINm5F cells (22). The effect of D-glyceraldehyde has been shown to be mediated by a raise in cytosolic Ca 2ϩ in RINm5F cells by opening voltage-sensitive Ca 2ϩ channels of the -conotoxin-sensitive type after depolarization induced by closure of ATP-sensitive K ϩ channels (26,27). KCl on the other hand depolarizes the cells causing the opening of the L-type voltagesensitive Ca 2ϩ channels and a raise in cytosolic Ca 2ϩ (28). The raise in cytosolic Ca 2ϩ activates exocytosis through a series of mechanisms (29).
In the present study we found that following culture of RINm5F cells in dexamethasone-containing medium for 5 days, the release of insulin upon stimulation with D-glyceraldehyde and KCl was markedly reduced compared with control cells. Still, however, insulin secretion was Ca 2ϩ -dependent and abolished by removal of extracellular Ca 2ϩ . The D-glyceraldehyde-induced raise in [Ca 2ϩ ] i was impaired after dexamethasone, suggesting impairment either of the closure of ATP-dependent K ϩ channels or of the opening of the -conotoxinsensitive Ca 2ϩ channels. Also the increase in cytosolic [Ca 2ϩ ] i when stimulated with KCl was lowered by dexamethasone, suggesting an inhibitory influence of dexamethasone also on the L-type Ca 2ϩ channels. Hence, dexamethasone markedly impaired the raising of [Ca 2ϩ ] i in RINm5F cells. Considering the importance of cytosolic Ca 2ϩ for insulin secretion (29), this might explain the impairment of D-glyceraldehyde-and KClstimulated insulin secretion after dexamethasone, although the exact nature of the inhibited insulin secretion after dexamethasone treatment of RINm5F cells remains to be established.
In other insulin-producing cell lines, no similar study on the influence of long term dexamethasone treatment on insulin secretion has been performed. However, several short term studies exist with divergent results on the influence of dexamethasone on insulin secretion using insulin-producing cell lines and isolated rodent islets. For example, in the hamster insulinoma cell line HIT T-15 insulin biosynthesis and secretion are reduced by glucocorticoids (30), whereas RINm5F cells and isolated rodent islets treated with dexamethasone for 48 -72 h respond with increased insulin mRNA levels but decreased insulin secretion (31,32). Recently Goodman et al. (33) demonstrated that the human insulin promoter is also a negative glucocorticoid-responsive element, and the beta cells have previously been demonstrated to express glucocorticoid receptors (25). Altogether this strongly indicates a functional role for glucocorticoids in the physiology of beta cells with regard to both insulin formation and secretion although the effects might be different in different types of insulin-producing cells.
In contrast to insulin, RINm5F cells cultured in control medium and subsequently stimulated with D-glyceraldehyde or KCl did not release NPY. This is in accordance with the finding that control RINm5F cells do not express the peptide or its mRNA. However, following 5 days of dexamethasone treatment NPY was abundantly released from the cells during the subsequent 60-min incubation in accordance with the abundant NPY mRNA expression. The NPY release from the cells was of similar magnitude as that of insulin, illustrating the marked induction of NPY expression by dexamethasone. In contrast to the release of insulin from dexamethasone-treated cells the release of NPY was not augmented by D-glyceraldehyde or KCl nor inhibited by removal of extracellular Ca 2ϩ . This indicates that the mechanisms of insulin and NPY secretion differ. Insulin secretion exhibits several characteristics of a regulated secretion (stimulation by nutrient secretagogue and depolarization and inhibition by removal of extracellular Ca 2ϩ ), whereas NPY secretion does not seem to be at all regulated. Rather, NPY secretion from dexamethasone-treated RINm5F cells displays more characteristics of constitutive secretion. In a constitutive release process the formed protein bypasses the sorting in the Golgi network and packaging into secretory granules, and therefore escapes the regulation of exocytosis, e.g. in response to raised intracellular Ca 2ϩ (34); Ca 2ϩ is required for the sorting of proteins for the regulated pathway (35). Our results therefore indicate that NPY is released by a constitutive rather than a regulated mechanism. It is worth noticing in this context that RINm5F cells are known to possess both constitutive and regulated pathways. Previous studies have demonstrated that RINm5F cells express the prohormone convertase PC2 (36,37), which is specific for the regulated pathway (38,39). Also the processing enzyme furin, which is specific for the constitutive pathway (40), has been detected in RINm5F cells (41). Analogous with our present results the insulin-producing cell line INS-1 has been found to express and secrete NPY in a constitutive manner since the cells did not respond with increased release of NPY to stimulation with glucose (42). Furthermore, of the islet regulatory peptides, islet amyloid polypeptide (known to be co-localized with insulin in beta cell secretory granules and co-released with insulin (43)) has been found to be released from rat neonatal beta cells during inhibition of the regulated pathway by removal of extracellular Ca 2ϩ , suggesting release of islet amyloid polypeptide also through a constitutive pathway (44). Thus, different types of insulin-producing cells exhibit characteristics for both types of secretion.
In conclusion, our findings demonstrate a steroid-sensitive plasticity of the RINm5F cells causing induction of NPY expression after long term treatment with dexamethasone in conjunction with impaired insulin secretion and inhibited in-

TABLE II
Effects of D-glyceraldehyde, KCl, and depletion of extracellular Ca 2ϩ on insulin and NPY release from RINm5F cells after 5 days of culture in the absence or presence of dexamethasone Cells cultured without (control) or with dexamethasone (100 nM) for 5 days and then incubated for 60 min in D-glyceraldehyde (GA) or KCl in the presence (ϩ) or absence (Ϫ) of extracellular Ca 2ϩ are as indicated. All incubations were performed in HEPES medium containing 3.3 mM glucose. The release of insulin and NPY into the medium was determined by radioimmunoassay. Values represent the mean Ϯ S.E., n ϭ 6. crease in cytosolic Ca 2ϩ . In contrast to the secretion of insulin, NPY secretion from RINm5F cells is not affected by D-glyceraldehyde or KCl nor inhibited by removal of extracellular Ca 2ϩ . Thus, whereas insulin is released by the regulated pathway, NPY is released by a mechanism showing characteristics of the constitutive pathway. The plasticity of RINm5F cells is evident not only by its sensitivity for dexamethasone to markedly induce NPY expression and secretion but also by its ability to use both the regulated and constitutive secretory pathways for the release of bioactive peptides.