Rab11 and its effector Rip11 participate in regulation of insulin granule exocytosis.

Rab GTPases and their effectors play important roles in membrane trafficking between cellular compartments in eukaryotic cells. In the present study, we examined the roles of Rab11B and its effectors in insulin secretion in pancreatic beta-cells. In the mouse insulin-secreting cell line MIN6, Rab11 was co-localized with insulin-containing granules, and over-expression of the GTP- or the GDP-bound form of Rab11B significantly inhibited regulated secretion, indicating involvement of Rab11B in regulated insulin secretion. To determine the downstream signal of Rab11-mediated insulin secretion, we examined the effects of various Rab11-interacting proteins on insulin secretion, and found that Rip11 is involved in cAMP-potentiated insulin secretion but not in glucose-induced insulin secretion. Analyses by immunocytochemistry and subcellular fractionation revealed Rip11 to be co-localized with insulin granules. The inhibitory effect of the Rip11 mutant was not altered in MIN6 cells lacking Epac2, which mediates protein kinase A (PKA)-independent potentiation of insulin secretion, compared with wild-type MIN6 cells. In addition, Rip11 was found to be phosphorylated by PKA in MIN6 cells. The present study shows that both Rab11 and its effector Rip11 participate in insulin granule exocytosis and that Rip11, as a substrate of PKA, regulates the potentiation of exocytosis by cAMP in pancreatic beta-cells.

Islet amyloid polypeptide in human insulinomas. Evidence for intracellular amyloidogenesis.

Amyloid deposits that characteristically form in the pancreatic islets of patients with non-insulin-dependent diabetes mellitus (NIDDM) and in insulinomas are both derived from islet amyloid polypeptide (IAPP). Evidence from previous studies has suggested that deposition of IAPP-derived amyloid is related to inherent amyloidogenic sequences present within normal human IAPP, together with an increased production and local concentration of IAPP. However, whether the aggregation of IAPP to form amyloid fibrils is primarily an intra- or extracellular event is not clear. To address this question, we studied 20 human insulinomas by light and electron microscopy. By light microscopy, amyloid deposits were demonstrated in 13 of 20 (65%) human insulinomas. Furthermore, evaluation of Congo red-stained tumor sections showed small, globular or irregular, congophilic amyloid deposits within the cytoplasm of many tumor cells in 10 of 13 (77%) amyloid-containing insulinomas. Dense, punctate areas of IAPP immunoreactivity within tumor cells corresponded with the congophilic intracellular deposits. Ubiquitin immunoreactivity also was observed as punctate intracellular labeling and within large extracellular amyloid deposits. Among the 10 insulinomas available for electron microscopic evaluation, pathological IAPP-immunoreactive (immunogold) deposits were found in 3 of 5 insulinomas in which amyloid was demonstrated by light microscopy and in none of 5 tumors found negative for amyloid by light microscopy. Morphology of IAPP-immunoreactive deposits varied from those with the classical distinct 7- to 10-nm diameter nonbranching fibrils to those with distinct but faint fibrillarity to those without discernable fibrils.(ABSTRACT TRUNCATED AT 250 WORDS)

Clonal insulinoma cell line that stably maintains correct glucose responsiveness.

A number of pancreatic beta-tumor cell (beta TC) lines have been derived from insulinomas arising in transgenic mice expressing the SV40 T antigen gene under control of the insulin promoter. Some of these lines secrete insulin in response to physiological glucose concentrations. However, this phenotype is unstable. After propagation in culture, these nonclonal lines become responsive to subphysiological glucose levels and/or manifest reduced insulin release. Here we report the use of soft-agar cloning to isolate single-cell clones from a beta TC line, which give rise to sublines that maintain correct glucose responsiveness and high insulin production and secretion for > 55 passages (over a year) in culture. One of these clonal lines, denoted beta TC6-F7, was characterized in detail. beta TC6-F7 cells expressed high glucokinase and low hexokinase activity, similarly to normal islets. In addition, they expressed mRNA for the GLUT2 glucose transporter isotype and no detectable GLUT1 mRNA, as is characteristic of normal beta-cells. These results demonstrate that transformed beta-cells can maintain a highly differentiated phenotype during prolonged propagation in culture, which has implications for the development of continuous beta-cell lines for transplantation therapy of diabetes.

Site-specific phosphorylation of synapsin I by Ca2+/calmodulin-dependent protein kinase II in pancreatic betaTC3 cells: synapsin I is not associated with insulin secretory granules.

Increasing evidence supports a physiological role of Ca2+/calmodulin-dependent protein kinase II (CaM kinase II) in the secretion of insulin from the pancreatic beta-cell, but the precise sites of action are not known. A role of this enzyme in neuroexocytosis is implicated by its phosphorylation of a vesicle-associated protein, synapsin I. Because of emerging similarities to the neuron with respect to exocytotic mechanisms, the expression and phosphorylation of synapsin I in the beta-cell have been studied. Synapsin I expression in clonal mouse beta-cells (betaTC3) and primary rat islet beta-cells was initially confirmed by immunoblot analysis. By immunoprecipitation, in situ phosphorylation of synapsin I was induced in permeabilized betaTC3 cells within a Ca2+ concentration range shown to activate endogenous CaM kinase II under identical conditions. Proteolytic digests of these immunoprecipitates revealed that calcium primarily induced the increased phosphorylation of sites identified as CaM kinase II-specific and distinct from protein kinase A-specific sites. Immunofluorescence and immunogold electron microscopy verified synapsin I expression in betaTC3 cells and pancreatic slices but demonstrated little if any colocalization of synapsin I with insulin-containing dense core granules. Thus, although this study establishes that synapsin I is a substrate for CaM kinase II in the pancreatic beta-cell, this event appears not to be important for the mobilization of insulin granules.

A novel insulin secretagogue is a phosphodiesterase inhibitor.

The arylpiperazine L-686,398 was described as an oral hypoglycemic agent and is shown to be an insulin secretagogue in vitro. The characteristics of its activity were similar to those of the incretin glucagon-like peptide I (GLP-I). We demonstrate that both the peptide and L-686,398 increase the accumulation of cAMP in isolated ob/ob mouse pancreatic islet cells, but by different mechanisms. Although GLP-I activates adenylate cyclase, the arylpiperazine has no effect on this enzyme or on the binding of 125I-labeled GLP-I to its receptor on RINm5F rat insulinoma cell membranes. However, L-686,398 inhibits the total cAMP phosphodiesterase (PDE) activity in homogenates of ob/ob mouse pancreatic islets with an EC50 of approximately 50 mumol/l. To determine the mechanism of PDE inhibition by the arylpiperazine and to examine its specificity, we studied the kinetics of arylpiperazine inhibition of two recombinant PDEs. The arylpiperazine is a competitive inhibitor of both a human heart type III PDE and a rat type IV-D PDE. Inhibition of the type III and IV isozymes are characterized by Ki values of 27 and 5 mumol/l, respectively. Although not extremely potent, the arylpiperazine does exhibit modest selectivity between these PDEs. The observation that L-686,398 acts as a PDE inhibitor suggests that exploration for beta-cell-specific PDE isoforms may reveal novel PDEs as targets for the development of therapeutically useful glucose-dependent insulin secretagogues.

Structure, function, and immunogenicity of human insulinoma cells.

Dissociated human insulinoma cells were plated onto plastic multiwell dishes. Cells were maintained for 1 mo on plastic with three passages. Cultures consisted of small colonies with some areas of stratification and few intercellular spaces. Ultrastructural studies indicated that cultured cells had epithelial features with desmosomes at cell-to-cell contacts and intermediate filaments in addition to secretory granules in the cytoplasm. Insulin and C-peptide were released in equimolar amounts in culture media. When challenged for 30 min with 16.7 mM glucose, 1 mM 3-isobutyl-1-methylxanthine, 4 mM tolbutamide, or 10(-6) M glucagon, insulinoma cells responded by a 1.5-, 1.5-, 2-, or 3-fold increase, respectively, in insulin release above baseline levels. A 15-min challenge with 10(-5) M isoproterenol increased insulin secretion by 1.85-fold. By indirect immunofluorescence, an anti-insulin antibody reacted positively with cell cytoplasm, whereas anti-somatostatin and anti-glucagon antibodies did not. Insulinoma cell surface expressed class I MHC molecules but not class II molecules. Immediately after isolation, crude insulinoma cells were contaminated by 2% of DR+ cells from nonislet components that disappeared after several weeks in culture. The ability of insulinoma cells to stimulate allogenic T-lymphocyte proliferation was assessed by [3H]thymidine incorporation in mixed culture combinations. Crude insulinoma cells elicited a strong lymphoproliferative response with a stimulation index ranging between 3.5 and 7, whereas no stimulation was found after 1 mo in culture. It is postulated that absence of class II-positive cells in the stimulatory cell preparation conditioned this immune tolerance across the major histocompatibility barrier.(ABSTRACT TRUNCATED AT 250 WORDS)

Centrosome abnormalities in pancreatic ductal carcinoma.

The centrosome plays an important role in microtubule nucleation and organization, ensuring the establishment of cell polarity and balanced chromosome segregation. Recent studies have suggested that the loss of cell polarity and/or chromosome missegregation (aneuploidy) in human malignant tumors could result from defects in centrosome function. Using immunofluorescence analysis with an antibody to gamma-tubulin (a well-characterized centrosomal component), we examined surgically resected human pancreatic tissues for centrosome abnormalities. The tissues included ductal carcinomas (n = 13), adenomas (n = 3), endocrine tumors (n = 3), chronic pancreatitis (n = 5), and normal pancreatic tissues (n = 12). We found that most (85%) carcinomas and some adenomas displayed abnormal centrosome profiles, characterized by an increase in size and number of centrosomes, and by their irregular distribution. In contrast, none of normal ductal and stromal tissues showed these abnormalities. These findings suggest that centrosome abnormalities may develop at a relatively early stage of pancreatic ductal carcinogenesis.

Truncated SNAP-25 (1-197), like botulinum neurotoxin A, can inhibit insulin secretion from HIT-T15 insulinoma cells.

We and others have previously shown that insulin-secreting cells of the pancreas express high levels of SNAP-25 (synaptosomal-associated protein of 25 kDa), a 206-amino acid t-SNARE (target soluble N-ethylmaleimide-sensitive factor attachment protein receptors) implicated in synaptic vesicle exocytosis. In the present study, we show that SNAP-25 is required for insulin secretion by transient transfection of Botulinum Neurotoxin A (BoNT/A) into insulin-secreting HIT-T15 cells. Transient expression of BoNT/A cleaved the endogenous as well as overexpressed SNAP-25 proteins and caused significant reductions in K+ and glucose-evoked secretion of insulin. To determine whether the inhibition of release was due to the depletion of functional SNAP-25 or the accumulation of proteolytic by-products, we transfected cells with SNAP-25 proteins from which the C-terminal nine amino acids had been deleted to mimic the effects of the toxin. This modified SNAP-25 (amino acids 1-197) remained bound to the plasma membrane but was as effective as the toxin at inhibiting insulin secretion. Microfluorimetry revealed that the inhibition of secretion was due neither to changes in basal cytosolic Ca2+ levels nor in Ca2+ influx evoked by K(+)-mediated plasma membrane depolarization. Electron microscopy revealed that cells transfected with either BoNT/A or truncated SNAP-25 contained significantly higher numbers of insulin granules, many of which clustered close to the plasma membrane. Together, these results demonstrate that functional SNAP-25 proteins are required for insulin secretion and suggest that the inhibitory action of BoNT/A toxin on insulin secretion is in part caused by the production of the plasma membrane-bound cleavage product, which itself interferes with insulin granule docking and fusion.

XIHbox 8, an endoderm-specific Xenopus homeodomain protein, is closely related to a mammalian insulin gene transcription factor.

The cis-acting sequences that mediate insulin gene expression exclusively in pancreatic islet beta-cells are localized within the 5'-flanking region between nucleotides -340 and -91. We have identified an evolutionarily conserved, A+T-rich element at -201/-196 basepairs in the rat insulin II gene that is essential for efficient expression in beta-cells. Affinity-purified antibody to the XIHbox 8 protein super-shifted the major beta-cell-activator factor complex binding to the -201/-196 element. XIHbox 8 is a Xenopus endoderm-specific homeodomain protein whose expression is restricted to the nucleus of endodermal cells of the duodenum and developing pancreas. Antibody to XIHbox 8 specifically interacts with a 47-kilodalton protein present in this DNA complex. Immunohistochemical studies revealed XIHbox 8-like proteins within the nucleus of almost all mouse islet beta-cells and a subset of islet alpha- and beta-cells. These results are consistent with the proposal that an XIHbox 8-related homeoprotein of 47 kilodalton is required for expression of the mammalian insulin gene in beta-cells. Experiments conducted with antiserum raised to somatostatin transcription factor-1 (STF-1), a recently isolated mammalian XIHbox 8-related homeoprotein, indicate that the STF-1 protein is the mammalian homolog of Xenopus XIHbox 8.

Characterization and significance of sulfonylurea receptors.

This study describes and characterizes a putative sulfonylurea receptor. The radioligand used was [3H]glipizide (9 Ci/mmol). The beta-cell plasma membranes were derived from a transplantable rat insulinoma generated by subcutaneous injection of RINm5F cells and purified by ultracentrifugation on a 15-55% sucrose gradient. Specific binding of [3H]glipizide to purified beta-cell plasma membranes was determined to be maximal at temperatures of 4-23 degrees C, pH 7.3, and an incubation of 2 h. Scatchard analysis indicated a single binding site with Kd = 7 nM and sulfonylurea binding of 0.93 pmol/mg membrane protein. Displacement of [3H]glipizide from the purified beta-cell plasma membranes by various sulfonylureas and their analogues correlated well with their known hypoglycemic and insulin-releasing activities. Various agents, including nutrients, agents affecting Ca2+ flux, gastrointestinal hormones, and pancreatic hormones, had no effect on [3H]glipizide binding to the beta-cell plasma membranes. Putative sulfonylurea receptors on beta-cell and brain cell plasma membranes have been reported by several groups of investigators. Sulfonylurea binding to the beta-cell is hypothesized to close an ATP-sensitive K+ channel, which leads to depolarization of the membrane and activation of a voltage-dependent Ca2+ channel.

Loss of polarization of plasma membrane domains in transformed pancreatic endocrine cell lines.

Using enveloped RNA viruses that bud selectively from either the apical or basolateral surface in polarized epithelial cells, we have recently provided evidence for polarization of plasma membrane domains in cultured pancreatic islet cells. In this study, we have followed the same experimental strategy to establish whether these polarized properties are maintained in transformed pancreatic endocrine cells. We find that influenza virus and vesicular stomatitis virus emerge from both the attached and free surfaces of cultured insulinoma cells (RIN cells) and SV40-transformed beta-cells (HIT cells). This demonstrates loss of polarization in transformed pancreatic endocrine cells.

Molecular heterogeneity and cellular localization of carboxypeptidase H in the islets of Langerhans.

The intracellular distribution and molecular heterogeneity of carboxypeptidase H was studied in rat insulinoma tissue and isolated islets of Langerhans by a combination of immunohistochemical, ultrastructural, subcellular fractionation, and immunoblotting analyses. Immunofluorescence microscopy of islets demonstrated the presence of carboxypeptidase H in both insulin-containing B cells and glucagon-containing A cells. Quantitative ultrastructural analyses of islet B cells indicated that the enzyme was concentrated in mature insulin secretory granules, clathrin-coated condensing granules, and to a lesser extent the Golgi apparatus. Carboxypeptidase H activity was localized principally to secretory granule subfractions of insulinoma tissue, where it was present for the major part (70%) as a form which is readily solubilizable at pH values prevailing in the granule interior (5.5). This species migrated as a diffuse band of 53-57 kilodaltons (kDa) on immunoblot analysis using antisera raised against the purified native enzyme. In contrast, the insoluble form which was associated with the granule membrane at pH 5.5, migrated as a relatively compact band of 55-57 kDa. Carboxypeptidase H activity was also present in subcellular fractions which contained Golgi membranes together with elements of the endoplasmic reticulum, and in a low density secretory granule fraction which may represent immature granules. The enzyme in these compartments, like the granule membrane species, migrated as a compact 55-57 kDa band on immunoblots. Two-dimensional electrophoretic immunoblot analysis of secretory granules suggested that both membrane and soluble forms of the enzyme were glycoproteins and that the terminal glycosylation was similar in both instances. Antiserum raised against the deduced C-terminal 11 amino acids of the cloned carboxypeptidase H sequence recognized the 55-57 kDa membrane component in granules but did not react with the 53-57 kDa soluble species. A major difference between the soluble and membrane forms therefore appears to be a structural modification or proteolytic removal of the C-terminal domain in the trans-Golgi or early secretory granule compartment. The concept that proteolysis is involved is further supported by the observation that the relative proportion of the high and low mol wt forms of the enzyme in different subcellular fractions correlated with that of proinsulin and insulin, respectively. The membrane association of the 55-57 kDa form of carboxypeptidase H is disrupted at pH values of 9 and is dependent on ionic strength. This further suggests that the C-terminus of the protein may have an important role in the sorting or concentration of the enzyme in vesicular elements of the regulated pathway of secretion.

Establishment of 2-mercaptoethanol-dependent differentiated insulin-secreting cell lines.

New insulin-secreting cell lines (INS-1 and INS-2) were established from cells isolated from an x-ray-induced rat transplantable insulinoma. The continuous growth of these cells was found to be dependent on the reducing agent 2-mercaptoethanol. Removal of this thiol compound caused a 15-fold drop in total cellular glutathione levels. These cells proliferated slowly (population doubling time about 100 h) and, in general, showed morphological characteristics typical of native beta-cells. Most cells stained positive for insulin and did not react with antibodies against the other islet hormones. The content of immunoreactive insulin was about 8 micrograms/10(6) cells, corresponding to 20% of the native beta-cell content. These cells synthesized both proinsulin I and II and displayed conversion rates of the two precursor hormones similar to those observed in rat islets. However, glucose failed to stimulate the rate of proinsulin biosynthesis. In static incubations, glucose stimulated insulin secretion from floating cell clusters or from attached cells. Under perifusion conditions, 10 mM but not 1 mM glucose enhanced secretion 2.2-fold. In the presence of forskolin and 3-isobutyl-1-methylxanthine, increase of glucose concentration from 2.8-20 mM caused a 4-fold enhancement of the rate of secretion. Glucose also depolarized INS-1 cells and raised the concentration of cytosolic Ca2+. This suggests that glucose is still capable of eliciting part of the ionic events at the plasma membrane, which leads to insulin secretion. The structural and functional characteristics of INS-1 cells remained unchanged over a period of 2 yr (about 80 passages). Although INS-2 cells have not been fully characterized, their insulin content was similar to that of INS-1 cells and they also remain partially sensitive to glucose as a secretagogue. INS-1 cells retain beta-cell surface antigens, as revealed by reactivity with the antigangloside monoclonal antibodies R2D6 and A2B5. These findings indicate that INS-1 cells have remained stable and retain a high degree of differentiation which should make them a suitable model for studying various aspects of beta-cell function.

Glucose sensitivity and metabolism-secretion coupling studied during two-year continuous culture in INS-1E insulinoma cells.

Rat insulinoma-derived INS-1 cells constitute a widely used beta-cell surrogate. However, due to their nonclonal nature, INS-1 cells are heterogeneous and are not stable over extended culture periods. We have isolated clonal INS-1E cells from parental INS-1 based on both their insulin content and their secretory responses to glucose. Here we describe the stable differentiated INS-1E beta-cell phenotype over 116 passages (no. 27-142) representing a 2.2-yr continuous follow-up. INS-1E cells can be safely cultured and used within passages 40-100 with average insulin contents of 2.30 +/- 0.11 microg/million cells. Glucose-induced insulin secretion was dose-related and similar to rat islet responses. Secretion saturated with a 6.2-fold increase at 15 mm glucose, showing a 50% effective concentration of 10.4 mm. Secretory responses to amino acids and sulfonylurea were similar to those of islets. Moreover, INS-1E cells retained the amplifying pathway, as judged by glucose-evoked augmentation of insulin release in a depolarized state. Regarding metabolic parameters, INS-1E cells exhibited glucose dose-dependent elevations of NAD(P)H, cytosolic Ca(2+), and mitochondrial Ca(2+) levels. In contrast, mitochondrial membrane potential, ATP levels, and cell membrane potential were all fully activated by 7.5 mm glucose. Using the perforated patch clamp technique, 7.5 and 15 mm glucose elicited electrical activity to a similar degree. A K(ATP) current was identified in whole cell voltage clamp using diazoxide and tolbutamide. As in native beta-cells, tolbutamide induced electrical activity, indicating that the K(ATP)conductance is important in setting the resting potential. Therefore, INS-1E cells represent a stable and valuable beta-cell model.

In vitro study of cultured human insulinoma cells: evidence of abnormal sensitivity to glucose.

Human insulinoma cells were isolated and cultured in vitro, and their functional and morphological characteristics were determined. The cells, isolated as single cells or small cell clusters, reaggregated to almost the size of islets by the fifth culture day and were maintained in vitro for more than 1 month. Morphologically (light and electron microscopies) they were intact throughout the culture period. Immunohistochemically more than 50% of the cells in each reaggregate contained insulin. Incubation experiments revealed that a low glucose concentration (15 mg/dL) was sufficient to produce maximal insulin release. In the absence of glucose, 1 microgram/mL glibenclamide increased insulin release. On the other hand, 5 mM theophylline and 10 mM arginine did not alter insulin release significantly. Theophylline, arginine, and glibenclamide did not have any stimulatory effect on insulin release in the presence of 50 mg/dL glucose. Perifusion experiments with 50 mg/dL glucose disclosed a biphasic pattern of insulin release, and no significant change in insulin release occurred when the glucose concentration in the perifusate was switched from 50 to 150 and then back to 50 mg/dL. These findings demonstrate that human insulinoma cells can be isolated and maintained in vitro and that the cells have abnormal sensitivity to glucose.

Characterization of the receptor for glucagon-like peptide-1(7-36)amide on plasma membranes from rat insulinoma-derived cells by covalent cross-linking.

125I-Labelled glucagon-like peptide-1(7-36)amide was cross-linked to a specific binding protein in plasma membranes prepared from RINm5F rat insulinoma-derived cells using disuccinimidyl suberate. Consistent with the presence of a single class of binding site on the surface of intact cells, only a single radiolabelled band at Mr63,000 was identified by SDS-PAGE after solubilization of the ligand-binding protein complex. The band was not observed when 10nM glucagon-like peptide-1(7-36)amide was included in the binding assay, but 1 microM concentrations of glucagon-like peptide-1(1-36)amide, glucagon-like peptide-2 and glucagon did not decrease the intensity of labelling. No change in the mobility of the band was observed under reducing conditions, suggesting that the binding protein in the receptor is not attached to other subunits via disulphide bonds. In control incubations using plasma membranes from pig intestinal epithelial cells, which do not contain specific binding sites for glucagon-like peptide-1(7-36)amide, no cross-linked ligand-binding protein complex was observed.

An antigen retrieval method using an alkaline solution allows immunoelectron microscopic identification of secretory granules in conventional epoxy-embedded tissue sections.

Immunoelectron microscopy using chromogranin A-specific antibodies has been proposed as an efficient technique for identification of secretory granules (SGs) in tumor cells with evidence of apparent neuroendocrine differentiation. Using an antigen retrieval (AR) method, we succeeded in immunolabeling SGs with antibodies in ultrathin sections of routinely processed epoxy-embedded blocks of tissue. Samples of an insulinoma were fixed in 2% glutaraldehyde, postfixed in 1% OsO(4), and embedded in epoxy resin. Ultrathin sections were immunostained with chromogranin A-specific antibodies and gold-conjugated second antibodies. There was no significant labeling in the absence of AR. Neither etching with sodium metaperiodate nor microwave irradiation of ultrathin sections in citrate buffer (pH 6.0) or in EDTA buffer (pH 8.0) was effective in improving the efficiency of immunolabeling. However, ultrathin epoxy-embedded sections that were microwaved in alkaline solution (pH 10) were adequately labeled (5.2 +/- 0.34 particles per SG). Moreover, considerably improved efficiency of immunostaining was achieved by microwaving sections in alkaline solution (pH 10) with subsequent immunostaining at 60C (12.2 +/- 0.51 particles per SG). This method can also be applied to epoxy-embedded sections obtained from formalin-fixed, paraffin-embedded blocks of tissue and was even valid for an old epoxy-embedded block of tissue prepared 15 years previously.

Culture of human insulinoma cells: development of a neuroendocrine tumor cell- and human pancreatic islet cell-specific monoclonal antibody.

We report on the culture of human insulinoma cells derived from a 32-year-old male patient with hyperinsulinism due to an insulinoma of the pancreas. A single-cell suspension was made by passing insulinoma fragments through a fine-gauge stainless-steel mesh. Cluster-forming insulinoma cells resembling pancreatic islets grew in the presence of fibroblasts. The insulinoma cell clusters could be differentiated from fibroblasts by using in situ pan optic staining and specific immunocytochemical staining (anti-human insulin and anti-human insulinoma monoclonal antibody (mAb) D24). mAb D24 was generated using insulinoma cells as antigen for immunization of a Balb/C mouse and cell fusion by the hybridoma cell technique. The anti-insulinoma cell mAb recognized a 32 kDa protein on immunoblot analysis of neuroendocrine tumor cells. D24 mAb also reacted immunohistochemically with normal pancreatic beta-cells and tumors such as vipoma, gastrinoma and carcinoid. Insulinoma cell clusters separated from fibroblasts by micromanipulation and plated into multiwell culture dishes exhibited an insulin-secretion rate of approximately 30 U/100 cells per 24 h with no insulin-secretory response to elevated glucose concentration. Purified insulinoma cells incubated with 1 ng/ml human nerve growth factor expressed neurofilament and neurite extension. These findings together with earlier observations in animal models suggest that human pancreatic beta-cells share some properties with neurons and are related to other neuroendocrine cells in the gastrointestinal tract.

The post-translational processing and intracellular sorting of carboxypeptidase H in the islets of Langerhans.

The post-translational processing and intracellular sorting of the proinsulin-converting enzyme carboxypeptidase H (CPH) was studied in isolated rat islets of Langerhans. Pulse-chase-radiolabelling experiments using sequence-specific antisera showed that CPH was synthesized initially as a 57-kDa glycoprotein which was processed to a 54-kDa mature form by proteolytic processing at the N-terminus. Processing of the CPH precursor occurred rapidly (t(1/2) = 30) after an initial delay of 15-30 min and the enzyme was secreted in parallel with the insulin-related peptides in response to glucose-stimulation within 1 h after radiolabelling. This indicated that the proteins were packaged into nascent secretory granules at approximately the same rate following synthesis. Conversion of proinsulin and the 57-kDa form was inhibited markedly by chase incubation of islets at 20 degrees C, indicating that maturation of both proteins occurs in a post-Golgi compartment. Affinity purification of the enzyme from insulinoma subcellular fractions showed that the 57-kDa form was associated with endoplasmic reticulum or Golgi elements, and the 54-kDa form was present in secretory granules. Structural analysis showed that the granule form of the enzyme had an N-terminal amino acid sequence beginning at residue 42 of rat CPH, thereby implicating cleavage of the precursor after the fourth Arg in a site containing five consecutive Arg residues. These findings indicate that post-translational processing of CPH is mediated by an endoprotease which cleaves at sites containing multiple basic amino acid residues upon segregation of the enzyme to the secretory granules.