Hypercholesterolaemia, signs of islet microangiopathy and altered angiogenesis precede onset of type 2 diabetes in the Goto-Kakizaki (GK) rat.

AIMS/HYPOTHESIS: The adult non-obese Goto-Kakizaki (GK) rat model of type 2 diabetes, particularly females, carries in addition to hyperglycaemia a genetic predisposition towards dyslipidaemia, including hypercholesterolaemia. As cholesterol-induced atherosclerosis may be programmed in utero, we looked for signs of perinatal lipid alterations and islet microangiopathy. We hypothesise that such alterations contribute towards defective pancreas/islet vascularisation that might, in turn, lead to decreased beta cell mass. Accordingly, we also evaluated islet inflammation and endothelial activation in both prediabetic and diabetic animals. METHODS: Blood, liver and pancreas were collected from embryonic day (E)21 fetuses, 7-day-old prediabetic neonates and 2.5-month-old diabetic GK rats and Wistar controls for analysis/quantification of: (1) systemic variables, particularly lipids; (2) cholesterol-linked hepatic enzyme mRNA expression and/or activity; (3) pancreas (fetuses) or collagenase-isolated islet (neonates/adults) gene expression using Oligo GEArray microarrays targeted at rat endothelium, cardiovascular disease biomarkers and angiogenesis, and/or RT-PCR; and (4) pancreas endothelial immunochemistry: nestin (fetuses) or von Willebrand factor (neonates). RESULTS: Systemic and hepatic cholesterol anomalies already exist in GK fetuses and neonates. Hyperglycaemic GK fetuses exhibit a similar percentage decrease in total pancreas and islet vascularisation and beta cell mass. Normoglycaemic GK neonates show systemic inflammation, signs of islet pre-microangiopathy, disturbed angiogenesis, collapsed vascularisation and altered pancreas development. Concomitantly, GK neonates exhibit elevated defence mechanisms. CONCLUSIONS/INTERPRETATION: These data suggest an autoinflammatory disease, triggered by in utero programming of cholesterol-induced islet microangiopathy interacting with chronic hyperglycaemia in GK rats. During the perinatal period, GK rats show also a marked deficient islet vascularisation in conjunction with decreased beta cell mass.

IL-1 antagonism reduces hyperglycemia and tissue inflammation in the type 2 diabetic GK rat.

Recent studies suggest an inflammatory process, characterized by local cytokine/chemokine production and immune cell infiltration, regulates islet dysfunction and insulin resistance in type 2 diabetes. However, the factor initiating this inflammatory response is not known. Here, we characterized tissue inflammation in the type 2 diabetic GK rat with a focus on the pancreatic islet and investigated a role for IL-1. GK rat islets, previously characterized by increased macrophage infiltration, displayed increased expression of several inflammatory markers including IL-1beta. In the periphery, increased expression of IL-1beta was observed primarily in the liver. Specific blockade of IL-1 activity by the IL-1 receptor antagonist (IL-1Ra) reduced the release of inflammatory cytokines/chemokines from GK islets in vitro and from mouse islets exposed to metabolic stress. Islets from mice deficient in IL-1beta or MyD88 challenged with glucose and palmitate in vitro also produced significantly less IL-6 and chemokines. In vivo, treatment of GK rats with IL-1Ra decreased hyperglycemia, reduced the proinsulin/insulin ratio, and improved insulin sensitivity. In addition, islet-derived proinflammatory cytokines/chemokines (IL-1beta, IL-6, TNFalpha, KC, MCP-1, and MIP-1alpha) and islet CD68(+), MHC II(+), and CD53(+) immune cell infiltration were reduced by IL-1Ra treatment. Treated GK rats also exhibited fewer markers of inflammation in the liver. We conclude that elevated islet IL-1beta activity in the GK rat promotes cytokine and chemokine expression, leading to the recruitment of innate immune cells. Rather than being directly cytotoxic, IL-1beta may drive tissue inflammation that impacts on both beta cell functional mass and insulin sensitivity in type 2 diabetes.

Trafficking/sorting and granule biogenesis in the beta-cell.

Proinsulin is packaged into nascent (immature, clathrin-coated) secretory granules in the trans-Golgi network (TGN) of the beta -cell along with other granular constituents including the proinsulin conversion enzymes. It is assumed that such packaging is dependent on an active sorting process, separating granular proteins from other secretory or membrane proteins, but the mechanism remains elusive. As granules mature, the clathrin coat is lost, the intragranular milieu is progressively acidified, and proinsulin is converted to insulin and C-peptide. Loss of clathrin is believed to arise by budding of clathrin-coated vesicles from maturing granules, carrying with them any inappropriate or unnecessary products and providing an additional means for refinement of granular content.

Storage of tissue-type plasminogen activator in Weibel-Palade bodies of human endothelial cells.

Tissue-type plasminogen activator (t-PA) is acutely released by endothelial cells. Although its endothelial storage compartment is still not well defined, t-PA release is often accompanied by release of von Willebrand factor (vWf), a protein stored in Weibel-Palade bodies. We investigated, therefore, whether t-PA is stored in these secretory organelles. Under basal culture conditions, a minority of human umbilical vein endothelial cells (HUVEC) exhibited immunofluorescent staining for t-PA, which was observed only in Weibel-Palade bodies. To increase t-PA expression, HUVEC were infected with a t-PA recombinant adenovirus (AdCMVt-PA). Overexpressed t-PA was detected in Weibel-Palade bodies and acutely released together with endogenous vWf by thrombin or calcium ionophore stimulation. In contrast, plasminogen activator inhibitor type 1 and urokinase were not detected in Weibel-Palade bodies after adenovirus-mediated overexpression. Infection of HUVEC with proinsulin recombinant adenovirus resulted in the storage of insulin in Weibel-Palade bodies, indicating that these organelles can also store nonendothelial proteins that show regulated secretion. Infection of AtT-20 pituitary cells, a cell type with regulated secretion, with AdCMVt-PA resulted in the localization of t-PA in adrenocorticotropic hormone-containing granules, indicating that t-PA can be diverted to secretory granules independently of vWf. Coinfection of AtT-20 cells with AdCMVt-PA and proinsulin recombinant adenovirus resulted in the colocalization of t-PA and insulin in the same granules. Taken together, these results suggest that HUVEC have protein sorting mechanisms similar to those of other regulated secretory cells. Although the results did not exclude an alternative storage site for t-PA in HUVEC, they established that t-PA can be stored in Weibel-Palade bodies. This finding may explain the acute coordinate secretion of t-PA and vWf.

Sorting human beta-cells consequent to targeted expression of green fluorescent protein.

Pancreatic islets of Langerhans are composed of four major endocrine cell types with a smaller number of nonendocrine cells. To study the molecular constituents and function of just one subpopulation of islet cells, it is necessary to sort them from the other cell types. While rat beta-cells can be sorted by autofluorescence-activated flow cytometry, this has not proved possible on a routine and reproducible basis for human beta-cells. In the present study, we have selectively labeled human beta-cells with green fluorescent protein (GFP), allowing for their sorting by flow cytometry. Human islet cells were infected with replication-defective (attenuated) recombinant adenovirus expressing GFP driven by the rat insulin I promoter (Ad-RIP-GFP) for targeted expression in beta-cells, or beta-galactosidase driven by the promiscuous cytomegalovirus (CMV) promoter (Ad-CMV-beta-gal) as control. Whereas the majority of islet cells can be infected by adenovirus, as shown by control infection with Ad-CMV-beta-gal, increased fluorescence after infection with Ad-RIP-GFP was limited to insulin-containing beta-cells. Infection of islet cells with Ad-RIP-GFP resulted reproducibly in the appearance of a population of intensely fluorescent cells, when analyzed by flow cytometry. These cells were sorted using a fluorescence-activated cell sorter (FACS) and shown by immunofluorescence to consist of >95% beta-cells. The targeted expression of GFP thus allows for preparation of human beta-cells purified close to homogeneity. This method should be readily applicable in any laboratory with FACS capability.

Proinsulin targeting to the regulated pathway is not impaired in carboxypeptidase E-deficient Cpefat/Cpefat mice.

Sorting of proinsulin from the trans-Golgi network to secretory granules is critical for its conversion to insulin as well as for regulated insulin secretion. The proinsulin sorting mechanism is unknown. Recently, carboxypeptidase E (CPE) was proposed as a sorting receptor for prohormones. To know whether CPE is implicated in proinsulin sorting, pancreatic islets were isolated from CPE-deficient Cpefat/Cpefat mice and Cpefat/+ controls, pulse-labeled ([3H]leucine), and then chased in basal medium (90 min) to examine constitutive secretion followed by medium with secretagogues (60 min) to stimulate regulated secretion. Secretion of labeled proinsulin via the constitutive pathway was <2% even in Cpefat/Cpefat islets. After a 150-min chase, only 13% of radioactivity remained as proinsulin in Cpefat/+ islets compared with 46% in Cpefat/Cpefat islets, reflecting slower conversion. Regulated secretion was stimulated to an equal extent from Cpefat/+ and Cpefat/Cpefat mice with 20% of the total content of labeled (pro)insulin released during the 60-min stimulatory period. It is concluded that in CPE-deficient Cpefat/Cpefat mice, proinsulin is efficiently routed to the regulated pathway and its release can be effectively stimulated by secretagogues. CPE is thus not essential for sorting proinsulin to granules.

Role of the prohormone convertase PC2 in the processing of proglucagon to glucagon.

Proglucagon is alternatively processed to glucagon in pancreatic alpha-cells, or to glucagon-like peptide-1 in intestinal L cells. Here, the specificity of PC2, the major prohormone convertase of alpha-cells, was examined both in vivo and in vitro. Adenovirus-mediated co-expression of proglucagon and PC2 in GH4C1 cells resulted in a pattern of processing products very similar to that observed in alpha-cells. Oxyntomodulin, an intermediate in the processing of proglucagon, was quantitatively converted to glucagon in vitro by purified recombinant PC2, in combination with carboxypeptidase E. It is concluded that PC2 is able to act alone in the pancreatic pathway of proglucagon processing.

Proinsulin conversion in GH3 cells after coexpression of human proinsulin with the endoproteases PC2 and/or PC3.

Proinsulin conversion to insulin occurs in secretory granules of pancreatic beta-cells. This processing has been suggested to require both the endoproteases PC2 and PC3 with each cleaving at only one of the two sites linking the insulin A- and B-chains with C-peptide. To evaluate this in an appropriate cellular setting, conversion of human proinsulin was followed in GH3 (rat pituitary) cells normally unable to convert this prohormone but equipped with the regulated secretory pathway. For this purpose, human proinsulin was expressed in GH3 cells, alone or in combination with PC2 and/or PC3, using recombinant adenoviruses. Cells were infected with the given adenoviruses and 24 h later were pulse-chased. Kinetics of proinsulin conversion were monitored by reverse-phase high-performance liquid chromatography. It was observed that while the two endoproteases do display a preference for a single site of cleavage (PC2 at the A-chain/C-peptide junction; PC3 at the B-chain/C-peptide junction) and act in a synergistic manner to promote proinsulin conversion, either PC2 or PC3 alone can cleave at both sites to fully convert proinsulin to insulin. These results also show that a cell can be successfully infected by three different recombinant adenoviruses.

Role of the prohormone convertase PC3 in the processing of proglucagon to glucagon-like peptide 1.

Proglucagon is processed differentially in pancreatic alpha-cells and intestinal endocrine L cells to release either glucagon or glucagon-like peptide-1-(7-36amide) (tGLP-1), two peptide hormones with opposing biological actions. Previous studies have demonstrated that the prohormone convertase PC2 is responsible for the processing of proglucagon to glucagon, and have suggested that the related endoprotease PC3 is involved in the formation of tGLP-1. To understand better the biosynthetic pathway of tGLP-1, proglucagon processing was studied in the mouse pituitary cell line AtT-20, a cell line that mimics the intestinal pathway of proglucagon processing and in the rat insulinoma cell line INS-1. In both of these cell lines, proglucagon was initially cleaved to glicentin and the major proglucagon fragment (MPGF) at the interdomain site Lys70-Arg71. In both cell lines, MPGF was cleaved successively at the monobasic site Arg77 and then at the dibasic site Arg109-Arg110, thus releasing tGLP-1, the cleavages being less extensive in INS-1 cells. Glicentin was completely processed to glucagon in INS-1 cells, but was partially converted to oxyntomodulin and very low levels of glucagon in AtT-20 cells in the face of generation of tGLP-1. Adenovirus-mediated co-expression of PC3 and proglucagon in GH4C1 cells (normally expressing no PC2 or PC3) resulted in the formation of tGLP-1, glicentin, and oxyntomodulin, but no glucagon. When expressed in alphaTC1-6 (transformed pancreatic alpha-cells) or in rat primary pancreatic alpha-cells in culture, PC3 converted MPGF to tGLP-1. Finally, GLP-1-(1-37) was cleaved to tGLP-1 in vitro by purified recombinant PC3. Taken together, these results indicate that PC3 has the same specificity as the convertase that is responsible for the processing of proglucagon to tGLP-1, glicentin and oxyntomodulin in the intestinal L cell, and it is concluded that this enzyme is thus able to act alone in this processing pathway.

Proinsulin processing in the rat insulinoma cell line INS after overexpression of the endoproteases PC2 or PC3 by recombinant adenovirus.

Proinsulin is converted to insulin by the two endoproteases PC2 and PC3. For complete conversion to insulin, cleavage must occur at both the B-chain/C-peptide and C-peptide/A-chain junctions of proinsulin. Studies in vitro have shown the specificity of PC3 for the B-chain/C-peptide junction and that of PC2 for the C-peptide/A-chain junction. In contrast, studies in vivo have suggested that the proinsulin cleavage substrate specificity of these two endoproteases might be more complex. We have now used recombinant adenovirus to overexpress PC2 or PC3 in the rat insulinoma cell line INS. These cells convert proinsulin more slowly than primary pancreatic beta-cells, possibly reflecting their lower levels of PC3. Infection of INS cells with the corresponding recombinant adenovirus led to 5-10-fold and 20-40-fold increases in PC2 and PC3 expression respectively. Recombinant adenovirus is thus an effective tool for expressing proteins at high levels in slow-growing INS cells. Overexpression of either PC2 or PC3 in INS cells led to a striking acceleration of conversion of proinsulin to insulin and to a decreased accumulation of the conversion intermediate des-64.65-split proinsulin (cleaved only at the A-chain/C-peptide junction). There was no detectable accumulation of des-31.32-split proinsulin (cleaved only at the B-chain /C-peptide junction) after overexpression of either enzyme. Taken together, the data indicate that when expressed at high levels, both PC2 and PC3 seem to be able to cleave proinsulin at both its junctions in vivo.

Des-(27-31)C-peptide. A novel secretory product of the rat pancreatic beta cell produced by truncation of proinsulin connecting peptide in secretory granules.

Insulin and connecting peptide (C-peptide) are produced in equimolar amounts during proinsulin conversion in the pancreatic beta cell secretory granule. To determine whether insulin and C-peptide are equally stable in beta cell granules (and thus secreted in equimolar amounts), neonatal and adult rat beta cells were pulse-chased, and radiolabeled insulin and C-peptide analyzed by high performance liquid chromatography. A novel truncated C-peptide was identified and shown by mass spectrometry to be des-(27-31)C-peptide (loss of 5 C-terminal amino acids). Des-(27-31)C-peptide is a major beta cell secretory product, accounting for 37.4 +/- 1.6% (neonatal) and 8.5 +/- 0.6% (adult) of total labeled C-peptide in secretory granules after 10 h of chase. Des-(27-31)C-peptide is also secreted in a glucose-sensitive manner from the perfused adult rat pancreas, accounting for approximately 10% of total C-peptide immunoreactivity secreted. Human C-peptide is also a substrate for truncation in granules. Thus, when human proinsulin was expressed (infection with recombinant adenovirus) in transformed (INS) rat beta cells, human des-(27-31)C-peptide was secreted along with the intact human peptide and both intact and truncated rat C-peptide. In addition to truncation, 33.1 +/- 1.2% of C-peptide in neonatal but not adult rat beta cell granules was further degraded. Such degradation was completely inhibited by ammonium chloride (known to neutralize intra-granular pH), whereas truncation was only partially inhibited by approximately 50%. In conclusion, a novel beta cell secretory product, des-(27-31)C-peptide, has been identified and should be considered as a potential bioactive peptide. Both truncation and degradation of C-peptide are responsible for non-equimolar secretion of insulin and C-peptide in rat beta cells.

Sequence requirements for proinsulin processing at the B-chain/C-peptide junction.

Proinsulin is converted into insulin by the action of two endoproteases. Type I (PC1/PC3) is thought to cleave between the B-chain and the connecting peptide (C-peptide) and type II (PC2) between the C-peptide and the A-chain. An acidic region immediately C-terminal to the point of cleavage at the B-chain/C-peptide junction is well conserved throughout evolution and has been suggested to be important for proinsulin conversion [Gross, Villa-Komaroff, Kahn, Weir and Halban (1989) J. Biol. Chem. 264, 21486-21490]. We have here compared the precise role of this region as a whole and just the first acidic residue C-terminal to the point of cleavage in processing of proinsulin by PC3. To this end, several mutations were introduced in this region of human proinsulin (native sequence, B-chain RREAEDL C-peptide): RRPAEDL (C1Pro mutant); RRLAEDL (C1Leu mutant); RRL (C1-C4del mutant); RRE (del-C1Glu mutant). Mutant and native cDNAs were stably transfected into AtT20 (pituitary corticotroph) cells, in which PC3 is known to be the major conversion endoprotease, and kinetics of proinsulin conversion were studied (pulse-chase/HPLC analysis of proinsulin-related peptides). The results show that the acidic region following the B-chain/C-peptide junction is indeed important for PC3 cleavage at this site, and that the reduced cleavage observed for the C1-C4del mutant proinsulin can be partially overcome by replacing the acidic region with a single acidic residue (del-C1Glu mutant). Replacing only the first residue of the acidic region with leucine (C1Leu mutant) has no impact on conversion, whereas its replacement with proline (C1Pro mutant) almost completely abolishes cleavage at the B-chain/C-peptide junction without affecting that at the C-peptide/A-chain junction.

Processing of proinsulin by furin, PC2, and PC3 in (co) transfected COS (monkey kidney) cells.

The enzymology of proinsulin conversion was studied in COS cells by cotransfection of three species of proinsulin and each of three conversion endoproteases (furin, PC2, and PC3). In addition to the parts of basic residues linking the B-chain to C-peptide (Arg31-Arg32) and C-peptide to the A-chain (Lys64-Arg65), which were present in all three proinsulins studied, human proinsulin presents a P4 basic residue (four residues NH2-terminal to the point of cleavage) only at the former junction (Lys29) and rat proinsulin II only at the latter (Arg62). Human proinsulin Arg62 (prepared by site-directed mutagenesis of human proinsulin) contains a P4 basic residue at both junctions. Transfected cells were incubated for four successive 2-h periods. The media were pooled, and pro-insulin, conversion intermediates, and insulin were separated by reverse-phase high-performance liquid chromatography to monitor conversion activity. There was little conversion of any proinsulin in COS cells without cotransfection of an exogenous endoprotease. When furin or PC3 was cotransfected with any of the three proinsulins, there was extensive processing, with insulin as the major conversion product. PC2, by contrast, failed to cleave human proinsulin but was able to cleave both human proinsulin Arg62 and rat proinsulin II. Cleavage by PC2 of these proinsulins was predominantly at the C-peptide-A-chain junction, generating the conversion intermediate des-64,65-split proinsulin as the major product and only very small amounts of insulin itself.

Human proinsulin conversion in the regulated and the constitutive pathways of transfected AtT20 cells.

AtT20 (mouse pituitary corticotroph) cells were stably transfected with human proinsulin cDNA. Clone H12 displayed low basal release of insulin-like immunoreactivity (< 1% cell content/30 min) with 17-fold stimulation by isobutylmethylxanthine/forskolin. Clone H23, by contrast, showed higher basal release (2.8% cell content/30 min) and only 6-fold stimulation. To follow the kinetics of conversion and release of only newly synthesized proinsulin, cells were pulse-chased, and labeled proinsulin-related products were analyzed by high pressure liquid chromatography. In the cells of both clones, [3H]proinsulin was converted to insulin with des-31,32-split proinsulin as the only detectable intermediate. While basal release of labeled products from H12 cells was low (3%/60 min), it was rapid and elevated from H23 cells (12.5% by 30 min and 24.8% by 60 min of chase) with [3H]des-31,32-split proinsulin the predominant molecular form. Stimulation of [3H] insulin release increased with time, reaching 3.8-fold by 90 min of chase, whereas that of [3H]des-31,32-split proinsulin was approximately 1.5-fold regardless of the chase time. Rapid secretion of newly synthesized products that is insensitive to secretagogues is the hallmark of the constitutive pathway. Thus in H23 cells an unusually large amount of proinsulin is diverted to the constitutive pathway, where it is partially converted to des-31,32-split proinsulin before release.

Substrate specificity of proinsulin conversion in the constitutive pathway of transfected FAO (hepatoma) cells.

Proinsulin is usually targetted to the regulated secretory pathway of beta cells, and converted to insulin in beta granules. Under certain pathological situations, a significant amount of proinsulin becomes diverted to the constitutive pathway. To study the kinetics of proinsulin conversion in the constitutive pathway, FAO (hepatoma) cells, which secrete proteins uniquely via this pathway and not the regulated pathway, were stably transfected with cDNA encoding human, rat I or rat II proinsulin. Products released to the medium of transfected cells were analysed by reversed phase HPLC and radioimmunoassay. For human proinsulin, des 31,32 split proinsulin (the conversion intermediate resulting from cleavage only at the B-chain/C-peptide junction followed by trimming of C-terminal basic residues by carboxypeptidase) was the only detectable conversion intermediate; for rat proinsulin II it was des 64,65 split proinsulin (cleaved and trimmed only at the C-peptide/A-chain junction); for rat proinsulin I, both intermediates were seen. Complete processing to insulin occurred for all three, but was most extensive for rat proinsulin I. When considered with the corresponding proinsulin sequences, these data show that a -4 basic residue (i.e. 4 residues N-terminal to the site of cleavage) facilitates proinsulin conversion in the constitutive pathway, and that arginine is preferred over lysine.

Kinetics of proinsulin conversion in human islets.

Islets isolated from human cadaver pancreas were pulse-labeled (10 min with [3H]leucine) and then incubated for a 180-min chase. Islets and chase medium were collected every 15 min and analyzed by reversed-phase HPLC to quantify the percentage of radioactively labeled proinsulin, conversion intermediates, and fully processed insulin. Release of proinsulin-related labeled products into the chase medium was < 10% of total. Whereas 50% of labeled proinsulin had been lost by conversion within 45 min, fully processed insulin only appeared with a half-time of 100 min. This discrepancy is attributable to accumulation of radioactive conversion intermediates. Des 64.65 split proinsulin was a minor component, reaching a maximum of 5.2 +/- 1.7% (n = 4) at 60 min of chase. By contrast, des 31.32 split proinsulin--and a truncated form lacking the first three residues of C-peptide--rose progressively to 29.3 +/- 1.4% by 75 min, and declined thereafter. The accumulation of des 31.32 split proinsulin rather than the des 64.65 split form during the conversion of human proinsulin reflects slower conversion at the C-peptide/A-chain than at the B-chain/C-peptide junction, and is consistent with the appearance of this particular conversion intermediate in the circulation.

A novel 7-nucleotide motif located in 3' untranslated sequences of the immediate-early gene set mediates platelet-derived growth factor induction of the JE gene.

A cohort of the serum and growth factor regulated immediate-early gene set is induced with slower kinetics than c-fos. Two of the first immediate-early genes characterized as such, c-myc and JE, are contained within this subset. cis-acting genomic elements mediating induction of the slower responding subset of immediate-early genes have never been characterized. Herein we characterize two widely separated genomic elements which are together essential for induction of the murine JE gene by platelet-derived growth factor, serum, interleukin-1, and double-stranded RNA. One of these elements is novel in several regards. It is a 7-mer, TTTTGTA, found in the proximal 3' sequences downstream of the JE stop codon. The 3' element is position dependent and orientation independent. It does not function in polyadenylation, splicing, or destabilization of the JE transcript. Copies of the 7-mer or its inverse are found at comparable 3' sites in 25 immediate-early genes that encode transcription factors or cytokines. Given its general occurrence, the 7-mer may be a required cis-acting control element mediating induction of the immediate-early gene set.

Processing of proinsulin by transfected hepatoma (FAO) cells.

Rat hepatoma (FAO) cells were stably transfected with the gene encoding either rat proinsulin II (using the DOL retroviral vector) or human proinsulin (using the RSV retroviral vector). Using the DOL vector, production of insulin immunoreactive material was stimulated up to 30-fold by dexamethasone (5 x 10(-7) M). For both proinsulins, fractional release of immunoreactive material relative to cellular content was high, in keeping with the absence of any storage compartment for secretory proteins in these cells. Pulse-chase experiments showed kinetics of release of newly synthesized products in keeping with release via the constitutive pathway. High performance liquid chromatography analysis showed immunoreactivity in the medium distributed between three peaks. For rat proinsulin II, the first coeluted with intact proinsulin; the second coeluted with des-64,65 split proinsulin (the product of endoproteolytic attack between the insulin A-chain and C-peptide followed by trimming of C-terminal basic residues by carboxypeptidase); the third (and minor peak) coeluted with native (fully processed) insulin. For human proinsulin, by contrast, the second peak coeluted with des-31,32 split proinsulin (split and trimmed at the B-chain/C-peptide junction). Analysis of cellular extracts showed intact proinsulin as the major product. The generation of the putative conversion intermediates and insulin was not due to proteolysis of proinsulin after its release but rather to an intracellular event. The data suggest that proinsulin, normally processed in secretory granules and released via the regulated pathway, may also be processed, albeit less efficiently, by the constitutive pathway conversion machinery. The comparison of the sites preferentially cleaved in rat II or human proinsulin suggests cleavage by endoprotease(s) with a preference for R/KXR/KR as substrate.