Perspectives in Diabetes What P-cell Defect Could Lead to Hyperproinsulinemia in NIDDM? Some Clues From Recent Advances Made in Understanding the Proinsulin-Processing Mechanism Christopher J. Rhodes and Cristina Alarcon

Pancreatic @-cell dysfunction is a characteristic of non- insulin-dependent diabetes mellitus (NIDDM). An aspect of this dysfunction is that an increased proportion of proinsulin is secreted, but an actual P-cell defect that leads to hyperproinsulinemia is unknown. Nevertheless, an impairment in P-cell proinsulin conversion mechanism has been suggested as the most likely cause. Insulin is produced from its precursor molecule, proinsulin, by limited proteolytic cleavage at two dibasic sequences (kg?', h82 and L ~ S ' ~ , kg6'). Two endopeptidase activ- ities catalyze this cleavage: PC2 and PC3. PC2 endopep- tidase cleaves predominately at L ~ S " ~ , Aqf6, and PC3 endopeptidase cleaves at Aqf' , M2. The recent identi- fication and characterization of these endopeptidases has enabled a better understanding of the human proinsulin- processing mechanism. In particular, experimental evi- dence suggests that the majority of human proinsulin processing is sequential. PC3 cleaves proinsulin first to generate a proinsulin conversion intermediate that is the preferred substrate of PC2. Both PC2 and PC3 activities are influenced by Ca2+ and pH, but the more stringent Ca2+ and pH requirements of PC3 suggest it as the most likely enzyme to regulate proinsulin conversion, as well as initiate it. When an increased demand is placed on the proinsulin-processing mechanism by a glucose-stimulated increase in ~roinsulin biosvnthesis. there is a coordinate increase inAPC3 biosynth\sis (bi t not in PC2). This supports PC3 as the key endopeptidase that regulates proinsulin processing. In this perspective, the current concepts of the enzymology and regulation of proinsulin conversion at a molecular level are reviewed. Then, sev- eral hypothetical points of P-cell dysfunction that might affect the proinsulin-processing mechanism and lead to hyperproinsulinernia are considered. Because the term NIDDM encompasses a wide variety of different disor- ders, any one or a combination of these possible points of P-cell dysfunction could result in hyperproinsulinemia for a certain subset of NIDDM. Diabetes 43:511-17,1994

From the E.P. Joslin Research Laboratory, Joslin Diabetes Center, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts.

Address c'orrespondence and reprint requests to Dr. Christopher J. Rhodes, E.P. Joslin Research Laboratory, Joslin Diabetes Center, Room 640D, One Joslin Place, Boston, MA 02215.

Received for publication I6 November 1993 and accepted in revised form 26 January 1994

NIDDM, non-insulin-dependent diabetes mellitus; RIA, radioin~munoassay; CP-H, carbo.xypeptidaseH.

T hat an increased proportion of proinsulin and proinsulin conversion intermediates exists in the circulation of patients with NIDDM is well estab- lished (14). Conventional radioimmunoassay

(RIA) determinations of serum insulin levels also measure proinsulin-like molecules. However, use of techniques such as high-performance liquid chromatography (5) and proinsu- lin-specific RIAs (6) have measured the proinsulin compo- nent separately from insulin. Because proinsulin and proinsulin conversion intermediates are biologically inactive compared with insulin, and an increased proportion of the insulin measured with conventional RIAs is proinsulin and conversion intermediates, patients with NIDDM could be insulin deficient (1,2,4,7). For the most part, hyperproinsu- linemia accompanies hyperglycemia (1,4,8). Even under rel- atively mild hyperglycemic circumstances (fasting plasma glucose levels of 6.1-7.8 mM), the proportion of circulating proinsulin rises from - l O O ? normally to 15-22% (8). As hyperglycemia worsens, the relative contribution of proinsu- lin in the circulation can rise to >40% (1,8). However, hyperproinsulinemia also has been observed to occur in certain individuals without hyperglycemia (1,2,9,10). Cur- rently, the cause of hyperproinsulinernia is thought to be a pancreatic p-cell defect that is augmented by the increased demand placed on the p-cell by hyperglycemia, rather than a decreased clearance of proinsulin-like molecules from the circulation (although this latter possibility has not been ruled out yet) (14). A p-cell defect could be reflected either as a primary dysfunction of the proinsulin conversion machinery itself or a malfunction in related p-cell regulatory mecha- nisms that affect insulin production secondarily. Over the last 2-3 years, great strides have been made for a better understanding of the proinsulin conversion mechanism in p-cells. After 25 years of searching, the specific enzyme activities that catalyze human proinsulin conversion have now been identified. In addition, other p-cell biochemical investigations have led to a clearer picture of the mechanics and regulation of proinsulin processing. In this perspective, we review these recent advances and hypothesize how certain defects in the human proinsulin conversion mecha- nism could contribute to hyperproinsulinemia.

DIABETES, VOL. 43, APRIL 1994

FIG. I. Cleavage sites of the two distinct proinsulin-processing endopeptidases on human proinsulin.

IDENTIFICATION OF TIIE PROINSULIN-CONVERTING ENZYMES Insulin is produced posttranslationally from its precursor molecule, proinsulin, by limited proteolysis, primarily in newly formed @-granules of the p-cell (11,12). An endopro- teolytic cleavage can be made on the carboxylic side of two dibasic sequences in the proinsulin molecule: Arg"', Arg:12 and at LysU, ~ r g " (13). This is followed by the specific removal of the newly exposed COOH-terminal basic amino acids by the rapid action of an exopeptidase, carboxypepti- dase-H (CP-H) (14,15), to yield the products insulin and C-peptide. A proinsulin-processing endopeptidase activity was first observed as a Ca2+-dependent @-granule-specific proteinase activity with an acidic pH optimum (16). This activity was subsequently found to consist of two distinct endopeptidases (17). One (type-I) cleaved exclusively at the ~ r $ ' , Arg" site, whereas the other (type-11) cleaved prefer- entially at LysU, Arg" (17-19) (Fig. 1). Around the time of the discovery of the type-I and type-I1 proinsulin-processing endopeptidases, the only unequivocal proprotein-processing endopeptidase was the Kex2 protease found in Sa,crhar.omy- res cerevisiae (20-22). The Kex2-protease was cloned (23), and its cDNA sequence provided means whereby mammalian Kex2-homologues could be isolated. Over the last 3 years, a new fanlily of mammalian Kex2-related proprotein-process- ing endopeptidases has emerged (21,24,25). Among this family of novel enzymes are two neuroendocrine specific proteinases named PC2 (26,27) and PC3 (28) (PC3 also has been named mPCl [29]). After the discovery of PC2 and PC3, it was subsequently established that PC2 and the type-I1 proinsulin-processing activity are equivalent (30) and that the PC3 is immunologically (31) and biochemically (32) indistinguishable from the type-I proinsulin-processing activ-

ity. Cotransfection experiments of PC2 and PC3 with proin- sulin have substantiated PC3 and PC2 as the proinsulin- processing type-I and type-I1 endopeptidase activities, respectively (33).

SEQUENTIAL PROCESSING OF HUMAN PROINSULIN Theoretically, proinsulin processing can take one of two routes (Fig. 2). Either PC3 will cleave intact proinsulin first at the Arg"', kg"' site t o generate the split 32, 33 proinsulin conversion intermediate or PC2 will cleave at the Lys", k g G " site to yield the split 65, 66 proinsulin conversion intennedi- ate. These split proinsulin intermediates are rarely detected in vivo because of the relatively high quantity of CP-H activity present in a p-granule (34). The more commonly seen proinsulin intermediates are the CP-H trimmed des 31, 32 proinsulin and des 64, 65 proinsulin. Another round of appropriate PC2PC3 cleavage, plus CP-H trimming of the des proinsulin intermediates, then generates the products, insulin and C-peptide.

However, the proinsulin-processing scenario outlined in Fig. 2 is probably not applicable to human proinsulin ron- version. Several lines of evidence suggest that human proin- sulin conversion has a preferred sequential route (Fig. 3). First, des 31, 32 proinsulin is present in higher quantities both in the circulation and in pancreas extracts, whereas des 64, 65 proinsulin is low (4-6). In NIDDM, hyperproinsuline- mia consists of elevated intact proinsulin and des 31, 32 proinsulin levels; des 64, 65 proinsulin levels are difficult to measure and most likely do not contribute to increased levels of proinsulin-like molecules (4,6). Second, kinetic studies of proinsulin conversion in isolated human pancre- atic islets have shown that des 31,32 proinsulin appears as a transient conversion intermediate at a greater than sevenfold higher level than des 64, 65 proinsulin (35). Third, an Arg3', Arg3"o kg", GlyZ2 proinsulin mutation (which is not processed by PC3 activity, because the Arg-Arg dibasic processing site is no longer present) was processed ineffi- ciently by PC2 activity at Lys6" Arg(j5, even though this site was present in the ~ 1 ~ " ~ ~ r o i n s u l i n variant (19,36). This finding led to the idea that some degree of secondary structure, as well as dibasic specificity, is necessary for efficient PC2 cleavage of proinsulin. Indeed, nuclear mag- netic resonance studies have shown that although most of the C-peptide moiety within the proinsulin molecule is unstructured, a degree of ordered structure occurs around the LysG4, k g B 5 site and has been termed the CA-knuckle (37). It has been proposed that the CA-knuckle might repre- sent a structural element necessary for PC2 processing of proinsulin (37). Finally, it has been demonstrated that PC2 activity has a much greater preference for des 31, 32 proin- sulin as a substrate than intact proinsulin (31,38). Further- more, intact proinsulin appeared to be inefficiently processed at LysG4, kg6 ' by PC2 compared with PC3 cleav- age at kg3', kg3' (38). This led to a proposal that intial cleavage of intact proinsulin by PC3 at Arg"', A r e is necessary to release a structural constraint around the L ~ S ' ~ , Arge6 site (perhaps represented by the CA-knuckle) that then renders easier accessibility for PC2 to cleave des 31, 32 proinsulin at L ~ s ~ ~ , (38).

Collectively, these lines of evidence indicate that the route via des 31, 32 proinsulin is the predominant pathway for proinsulin to insulin conversion in human p-cells (Fig. 3).

DIABETES, VOL. 43, APRII, 1444

FIG.

SPLIT 32, 33 PROlNSULlN

I

PC2 (Type-ll) Endopeptidase 0 6 -

PC3 (Type-I) Endopeptidase

+ Carboxypeptidase-H + Carboxypeptidase-H

C-PEPTIDE

PROlNSULlN

2. Tlie two possible routes of proinsulin conversion in human p-cells.

The strong preference of PC2 for processing des 31, 32 proinsulin ensures that the conversion mechanism strongly favors insulin and C-peptide production in p-granules, which compliments the observations made in vivo where >95% of proinsulin is fully converted to insulin (35,39). At most, < 10% of human proinsulin is converted to insulin via des 64, 65 proinsulin (35). The vast majority of proinsulin processing occurs in newly formed p-granules (11,40), where the in- traorganellar environment of ca2+ and pH ideally suits maximal. activity of the PC2 and PC3 endopeptidases (see below) 1(17,38). The more stringent ca2' and pH require- ments of the PC3 activity ensure that it is only active in p-granules (17,38). In contrast, the relatively broad biochem- ical requirements for PC2 activity suggest that it might also be partially active in the trans-Golgi network (the preceding organelle to the p-granule of the p-cell's secretory pathway) (2,17,38). Cleavage of proinsulin by partially active PC2 in the tra.ns-Golgi network (17,38) could account for a small amount of des 64, 65 proinsulin produced in human p-cells. However, note that PC2 is relatively inefficient at cleaving intact proinsulin, as well as not being fully active in the trans-Golgi compartment (17,31,38). Thus, the path via des 64, 65 proinsulin would be a relatively minor route of proinsulin conversion. Once in the p-granule compartment, >90% of proinsulin is converted by the preferred sequential route via des 31, 32 proinsulin (Fig. 3). In humans this hypothesis predicts that des 64, 65 proinsulin would be produced in negligible amounts, and that des 31, 32 proinsu- lin would be the major proinsulin conversion intermediate. This is actually what has been observed in both the human circulation and islets (4,6,35).

SPLIT 65, 66 PROlNSULlN

DES 64,65 PROlNSULlN

REGULATION OF PROINSULIN CONVERSION By examining the biochemistry and cell biology of the PC2 and PC3 endopeptidases, an understanding of the regulation of proinsulin processing is now becoming apparent. Ca2+ and pH regulation of endopeptidase activity. Both PC2 and PC3 are Ca"- and pH-dependent endopeptidases; however, they differ from each other in their pH and Ca2' requirements (17,31,38). PC2 is half-maximally activated between 50 and 100 pM Ca2+, whereas PC3 is only active above 0.5 mM ca2+ (17,38). Both endopeptidases have an acidic pH 5.5 optimum, but PC2 is more active than PC3 at neutral pH. The major intracellular site of proinsulin process- ing in the pancreatic p-cell is the p-granule ( l l ) , which contains an acidic environment of pH 5.5 (41) and an estimated free ca2' concentration of 1-10 mM (42). This ideally suits the optimal requirements for full activity of the proinsulin-processing endopeptidases (17,38). The strict bio- chemical requirements for PC3 activity of high mM Ca" and narrow acidic pH range (17,38) ensures that it is only active in the inner environment of a p-granule. This then provides means whereby insulin can only be produced in the organelle in which it is stored (17,38). Sequential proinsulin processing in human p-cells (Fig. 3) implies that PC3 makes the initial cut and is the enzyme that is likely t o control the process. Thus, it follows that PC3 is more stringently regulated by CaZ+ and pH. To render PC3 and PC2 fully active in a p-granule, the key events for initiating proinsulin to insulin conversion will be activation of the proton-pumping ATPase (34,40,41) and ca2+-transporting proteins (43). However, the means of triggering these particular p-granule Ca2' and H' translocation events in a newly formed p-granule have yet to be elucidated.

ULABETES, VOL. 43, APRIL 1994

PROINSULIN CONVERSION AND HYPERPROINSULINEMIA I

PROINSULIN

w PC3 (type-I) endopeptidase

+ carboxypeptidase-H

PC2 (type-l I) endopeptidase + carboxypeptidase-H

INSULIN

C-PEPTIDE

FIG. 3 . Sequential processing o f proinsulin.

Proteolysis of PC2 and PC3. The Kex2-related family of mammalian endopeptidases have homology to bacterial sub- tilisin, which requires a proregion so that the molecule can become a correctly folded, active enzyme. Once folded, the proregion is autocatalytically cleaved (44). Kex2 and another member of the Kex2-related endopeptidase family, furin, undergo autocatalytic cleavage activation (45,46). PC2 and PC3 are initially synthesized as proproteins that are post- translationally cleaved by limited proteolysis (47,48). In pancreatic islets, PC2 is initially synthesized as a 76 kD glycoprotein that is cleaved via 71/68 kD intermediates to a mature 64-66 kD p-granule form (47,48). PC3 is cleaved from a 92/94 kD glycoprotein precursor to a 66 kD mature 6-granule form (48). However, it has yet to be demonstrated whether proteolysis of PC2 and PC3 is an autocatalytic and/or activation process (24). Apparently, the precursor form of PC3 has proteolytic activity (32,49). However, it is unclear whether processing of PC3 to its mature 66 kD form affects its processing activity. It has been hypothesized that furin (a putative housekeeping proprotein-processing endo- peptidase present in the majority of mammalian cells [21,24]), rather than autocatalytic cleavage, processes PC2 and PC3. For the moment, proteolysis of PC2 and PC3 as a regulatory activation process remains to be demonstrated, but until proven otherwise, it should be considered as a possible regulation of proinsulin processing. Glucose regulation of PC2 and PC3. Proinsulin biosynthe- sis can be stimulated by glucose at both translational and transcriptional levels (50-52). An increase in proinsulin biosynthesis places an obvious increased demand on the proinsulin conversion mechanism in p-cells. For short-term glucose stimulation ((2 h), the biosynthesis of PC3 is stimulated in parallel with that of proinsulin at a translational level (48). In contrast, the biosynthesis of PC2 (48) and CP-H (54) are not glucose regulated. However, in the short term, proinsulin processing proceeds just as efficiently at stimula- tory or basal glucose concentrations (39), thus implying that enough PC2 and CP-H is already present in p-cells to cater for increased conversion demands. Given that PC3 initiates the majority of sequential processing of proinsulin conver- sion in human 6-cells (Fig. 3), the coordinate glucose regu- lation of PC3 with its proinsulin substrate at the biosynthetic level (48) further documents PC3 as the key endopeptidase that regulates proinsulin processing in p-cells. Such regula- tion ensures that proinsulin conversion proceeds efficiently and is adaptable to varying glucose concentrations in the short term (39,53). For longer term situations, it remains to be shown whether PC2, PC3, and CP-H genes are coordi- nately transcriptionally regulated with the preproinsulin gene by glucose (55-57). This is a distinct possibility, espe- cially because it has been shown that PC2 and PC3 mRNAs are coordinately regulated with that of pro-opiomelanocortin mRNA in pituitary cells (58).

WHAT DEFECT IN THE P-CELL PROINSULIN CONVERSION MECHANISM COULD CAUSE HYPERPROINSULINEMIA? Recent advances in understanding the proinsulin-processing mechanism can now lead to speculation of what might be a p-cell defect that may lead to hyperproinsulinemia in NIDDM. Several possibilities of p-cell dysfunction can be considered. Genetic variants of proinsulin conversion endopepti- dases. A mutation in either the PC2 and/or PC3 genes could

DIABETES, VOL. 43, APRIL 1994

adversely affect the activity of these endopeptidases. Such PC2/PC3 mutants in p-cells would result in inefficient pro- cessing of proinsulin leading to hyperproinsulinemia, insulin deficiency, and NIDDM. This is likely to be a very rare occurrence. Nevertheless, the possibility is potentially anal- ogous to the recent discovery of glucokinase mutations being a primary cause of maturity onset diabetes of the young (59). PC2PC3 mutations might also account for the hyperproinsulinemia found in some relatives of insulin- dependent diabetes mellitus patients (10). A general defective glucose stimulus-coupling mecha- nism in the p-cell. In some animal models of NIDDM, a selective loss of the glucose-regulated insulin release mech- anism occurs in p-cells (3). The glucose stimulus-coupling mechanism for insulin release shares some similarities with other glucose-regulated p-cell functions (60,61). A general dysfunction in the glucose-sensing (e.g., GLZTT2 downregu- lation (31) andlor signal transduction mechanisms could adversely influence general p-cell function that then second- arily affects proinsulin processing. Defective transcriptional regulation of the PC3 andlor PC2 genes. Preproinsulin gene transcription is regulated by glucose (55,57,62), and an extended stability of preproinsulin mRNA is observed with increased glucose concentrations (51). A rise in preproinsulin gene expression induced by hyperglycemia should increase proinsulin biosynthesis that, in t u n , would place an increased demand on the proinsulin conversion mechanism. However, it has not yet been deter- mined whether the PC3 and PC2 genes are coordinately regulated in p-cells. If the transcriptional machinery for PC2 and/or PC3 gene expression is not coregulated by glucose in parallel to the preproinsulin gene, or an inadequacy of PC2/PC3 gene expression during prolonged states of glucose stimulation (e.g., hyperglycemia in NIDDM) develops, defi- cient proinsulin conversion and hyperproinsulinemia could result. Defective translational regulation of glucose-induced PC3 biosynthesis. Glucose specifically regulates PC3 bio- synthesis in parallel with that of proinsulin (48). A deficiency in the translational machinery for regulating PC3 biosynthe- sis versus that of proinsulin could lead to less of the key endopeptidase thatinitiates the proinsulin conversion mech- anism in p8-cells. This could result in less efficient processing of proinsulin and hence, hyperproinsulinemia. A disproportionate ratio between PC3 and PC2 in NIDDM. In contrast to PC3 biosynthesis, PC2 biosynthesis is not translationally regulated by glucose in pancreatic islets (48). Thus, it is possible that an adverse discrepancy might arise in the ratio between PC2 and PC3 present in p-cells. Under conditions of prolonged glucose stimulation of the p-cell, PC3 biosynthesis should be able to keep pace with glucose-stimulated proinsulin biosynthesis (48). However, because PC2 biosynthesis is not glucose regulated, the ratio between PC2 and PC3 would grow ever wider. Eventually, under hyperglycemic conditions, the amount of PC2 present compared with that of PC3 and proinsulin would become rate limiting for proinsulin processing. The second PC2 catalyzed step in sequential processing of human proinsulin would be hindered (Fig. 3). As a result, des 31, 32 proinsulin would accumulate, which is what has been observed in hyperproinsulinemia in NIDDM (4,6). Defective processing of PC2 and PC3. PC2 and PC3 are posttranslationally cleaved by limited proteolysis in p-cells

(47,48), but it is not yet clear whether this affects their activities (24). Nevertheless, it is possibile that PC2 and PC3 processing might also be affected in NIDDM. A defect in proteolytic processing of PC2 andlor PC3 could limit the activity of these endopeptidases, which results in further inefficient processing of proinsulin and hyperproinsulinenda. An adverse change in the inner environment of the P-granule. The major intracellular site of proinsulin pro- cessing in the p-cell is a newly formed p-granule (11,40). Under normal circumstances, the PC2 and PC3 endopepti- dases are fully activated in this compartment because their CaZ+ and pH requirements are ideally suited by the intra- granular environment (17,38). Glucose can affect Ca'" and pH fluxes in the p-cell (61), and furthermore, secretory granule Ca" and pH are responsive under certain physio- logical conditions (21,63). Therefore, under hyperglycemic conditions, the possibility arises that the inner environment of the p-granule could be altered, which then adversely affects PC2 and PC3 activity. This would affect proinsulin processing, and hyperproinsulinemia might ensue. Mistargeting of conversion enzymes to the P-granule. In normal p-cells, >99% of newly synthesized proinsulin is targeted to the p-granule compartment (39,64), i.e., the regulated secretory pathway of p-cells (65). Presumably, PC2 and PC3 are as efficiently targeted to the p-granule, as indicated in that they are both cosecreted with a peptide hormone under regulatory circumstances (47,66). However, a defect in efficient targeting of PC2, PC3, and proinsulin to p-granules, perhaps as a result of overstirr~ulation of p-cell secretory pathway by hyperglycemia, would result in these proteins being constitutively secreted. Proinsulin processing is compromised severely when mistargeted to the constitu- tive secretory pathway, and hyperproinsulinemia results (64). This is analogous to insulinoma cells, where an in- creased proportion of proinsulin-like molecules also is se- creted (67). Increased secretory demand of the p-cell. Overstimula- tion of the p-cell's regulated secretory pathway, arising from hyperglycemia, could result in defective proinsulin conver- sion. Pulse-chase radiolabeling experiments have deter- mined that -3 h are required for 900/0 of newly synthesized proinsulin to be converted to insulin within the p-cell(35,39). Similar experiments have shown preferential release of newly synthesized proinsulin from p-cells and have shown that under stimulated conditions of elevated glucose, -50% of the newly synthesized @ro)insulin can be release within 3 h with an elevated proportion as proinsulin (39). If hypergly- cemia persists, so does glucose stimulation of the P-cell. Although PC2, PC3, and CP-H are cosecreted with Cpro)in- sulin (47,66,68), the environment and dilution of enzyme: substrate ratio on secretion would not be amenable for proinsulin processing to occur outside the p-cell. Therefore, even with the correct inner environment and compliment of conversion enzymes in a p-granule, it is possible that chronic increased demand of the p-cell's secretory pathway by hyperglycemia results in newly synthesized proinsulin not being retained within the P-granule compartment long enough for complete processing to insulin. As a result, an increased proportion of proinsulin secreted from overstimu- lated p-cells would gradually accun~ulate in the circulation and worsen so that chronic hyperproinsulinemia developed.

Note that NIDDM encompasses a family of diabetes disor- ders, and in all likelihood, hyperproinsulinemia in a given

DIABETES, VOL. 43. APRIL 1994

subset of NIDDM could result from any one or a combination of the above possible P-cell dysfunctions. However, the clearer understanding of the human proinsulin conversion mechanism has now provided experimental tools, such as specific antisera and DNA probes for PC2, PC3, and CP-H, so it is likely that pcell defects that cause hyperproinsulinemia in certain types of NIDDM will be identified soon. Flnally, note that proinsulin, although the major p-cell product, is not the only secretory proprotein to traverse the secretory pathway that might have its conversion affected in NIDDM. Other p-cell peptides, such as (pro)-islet amyloid polypep- tide (2), also might have their processing aEected adversely by abnormalities occurring in PC2 and PC3, which might further the deterioration of P-cell function in NIDDM. The next few years in P-cell research should be rather intriguing. Identification of p-cell defects in NIDDM might point to possibilities of novel therapies andlor design of pharmaceu- tical agents that alleviate hyperproinsulinemia in this disease and correct the apparent insulin deficeincy.

ACKNOWLEDGMENTS This work was supported in part by National Institutes of Health Grant DK-36836 and the Juvenile Diabetes Founda- tion International.

We thank Drs. Jack Leahy and Brian C. Horsburgh for helpful criticism and discussion.

REFERENCES 1. Leahy JL: Nat,ural history of B-cell dysfunction in NIDDM. Diabetes Core

13992-1010, 1990 2. Porte D, Kahn SE: Hyperproinsulinemia and amyloid in NIDDM: clues to

etiology of islet pcell dysfunction. Diabetes 38:133%36, 1989 3. Leahy JL, Bonner-Weir S, Weir GC: P-ccll dysfunction by chronic hyper-

glycemia: current ideas on mechanism of glucose-induced insulin secre- tion. Diabetes Care 15:442-54, 1992

4. Temple RC, Carrington CA, Luzio SD, Owens DR, Schneider AE, Sobey WJ, Hales CN: Insulin deficiency in non-insulin-dependent diabetes. Lancet 2DS95, 1989

5 . Given BD, Cohen RM, Shoelson SE, Frank BH, Rubenstein AH, Tagcr HS: Biochentical and clinical implications of proinsulin conversion intcrme- diates. J Clin Invest 76:1398-1405, 1985

6. Sobey WJ, Beer SF, Carrington CA, Clark PMS, Frank BH, Gray P, Luzio SD, Owens DR, Schneider AE, Siddle K, Temple RC, Hales CN: Sensitive and specific two-site inimunoradiometric assays for human insulin, proinsulin, 6 5 4 6 split, and 3 2 3 3 split proinsulins. Biorhem. .I 260:53-1, 1989

7. Levy JC, Clark PM, Hales CN, Turner RC: Normal proinsulin responses to glucose in mild type I1 subjects with subnormal insulin response. Diabetes 42:16'2-69, 1993

8. Yoshioka N, Kuzuya T, Taniguchi M, Iwamob Y: Serum proinsulin levels at fasting and after oral glucose load in patients with lype I1 (non-insulin- dependent) diabetes mellitus. Diabetologia 31355-60, 1988

9. Hartling SG, Lindgren F, Dahlqvist G, Persson B, Binder C: Elevated proinsulin in healthy siblings of IDDM pat,ients independent of HLA identity. Diabetes 38:1271-74, 1989

10. Heaton DA, Millward BA, Gray IP, Tun Y, Halcs CN, Pyke DA, Leslie RDG: Increased proinsulin levels as an early indicator of B-cell dysfunction in nondiabetic twins of type I (insulin-dependent) patients. Diabetologia 31:182-'34, 1988

11. Orci L, Ravazzola M, Storch M J , Anderson RGW, Vassalli J-D, Perrelet A: Protcolytic maturation of insulin is a postGolgi event which occurs in acidyfyng clathrin-coated secretory vesicles. Cell 49:86.>68, 1987

12. Orci L: The insulin factory: a tour of the plant surroundings and a visit to the assembly line. Dinbetologio 28:52846, 1985

13. Steiner DF, Kemntler W, Clark JL, Oyer PE, Rubcnstcin AH: The biosynttlesis of insulin. In Handbook of Physiology (Section 7). Greep RO, Astwood EB, Eds. Washington, D.C., American Physiological Soci- ety, 1972, p. 175-98

14. Docherty KD, Hutton JC:: Carboxypeptidase activity in the insulin secre- Lory granule. FEBS Lett 16'2: 13741, 1983

15. Davidson HW, Hutton JC: The insulin secretory granule carboxypepti- daseH: Purification and demonstration of involvement in proinsulin proressing. Biochem, J 245:57542, 1987

16. Davidson HW, Peshavaria M, Hutton JC: Proteolytic conversion of proinsulin into insulin: identification of a Ca"-dependent acidic endo-

peptidase in isolaled insulin secretory granules. Biochm .I 246279-86, 1987

17. Davidson HW, Rhodes CJ, Hutton JC: Intraorgenellar Ca and pH control proinsulin cleavage in the pancreatic p-cell via two site-specific endopep- tidases. Nature 333:93-96, 1988

18. Rhodcs CJ, Zumbrunn A, Bailyes EM, Shaw E, Hutton JC: The inhibition of proinsulin-processing endopeptidase activities by active-site-directed peptides. Biochem J 258:305-308, 1989

19. Docherty KD, Rhodes CJ, Taylor NA, Sheman KIJ, Hutton JC: Proinsulin endopeptidase substrate specificities defined by site-directed mulagene- sis of proinsulin. J Biol Chem 264:1833549, 1989

20. Fuller RS, Sterne RE, Thorner J: Enzymes required for yeast prohormone processing. Anr~ Rcu Physiol 50:345-62, 1988

21. Hutton JC: Subtilisin-like proteinases involved in the activation of proproteins of the secretory pathway of eukaryotic cells. C u m Opin CcU Biol 2:113142, 1991

22. Julius D, Brake A, Blair L, Kunisawa R. Thorner J: Isolation of the putative structural gene for the lysine-arginine cleaving endopeptidase for processing yeast prepro-a-factor. Cell 371075-89, 1984

23. Mizuno K, Nakanura T, Osh'ilna T, Tanaka S, Matsuo H: Yeast KEX2 genes encodes an endopeptidase homologous to subtilisin-like proteases. B i o c h m Biopl~ys Res Commun 15624654, 1988

24. Steiner DF, Smeekens SP, Ohagi S, Chan U The new enzymology of precursor processing endopeptidases. J Biol C h a 26732343548, 1992

25. Lindberg I: The new eukaryot,ic precursor processing proteinases. Mol Endoc~inol 5: 1361-65, 1991

26. Sn~eekens SP, Steiner DF: Identification of a human ins~llinoma cDNA encoding a novel mammalian protein structurally related to the yeast dibasic processing K E B . J Biol Chcm 265:2997-3000, 1090

27. Seidah NG, Gaspar L, Mion P, Marcinkiewicz M, Mbikay M, Chrktien M: cDNA sequence of two distinct pituitary proteins homologous to Kex2 and furin gene products: tissue-specific mRNAs encoding candidates for pro-hormone processing proteinases. DNA Cell Biol9:415-24, 1990

28. Slneekens SP, Avruch AS, LaMendola J, Chan ST, Steiner DF: Identifica- tion of a cDNA encoding a second putative prohormone convertase related to PC2 in AtT20 cells a ~ d islets of Langerhans. Proc Natl Acad Sci USA 88:30-W, 1991

29. Seidah NG, Marcinkiewicz M, Beqjannet S, Gaspar L, Beaubien G, Mattei MG, Lazure C. Mbikay M, Chetien M: Cloning and primaly sequence of a mouse candidate prohornlone convertase PC1 homologous to PC2, Furin, and KexZ: distinct chromosomal localisation and messenger RNA distribution in brain and pituitary compared to PC2. Mol Endocrinol 5:111-22, 1991

30. Bennett DL, Bailyes EM, Nielson E, Guest PC, Rutherford NG, Arden SD, Hutton JC: Identification of the type-11 proinsulin processing endopcpti- dase as PC2, a member of the eukaryotic subtilisin family. J Biol Chern 267:15229-36, 1992

31. Bailyes EM, Bennett DL, Hutton JC: Proprotein-processing endopepti- dases of the insulin secretory granule. Enzyme 45301-13, 1991

32. Bailyes EM, Shema1 IUJ. Seal AJ, Smeckens SP, Steiner DF, Hutton JC, Docherty K: A member of the eukaryotic subtilisin family (PC3) has the enzymic properties of the type-I proinsulin-converting endopeptidase. B i o c h m J 285:391-94, 1992

33. Smeekens SP, Montag AG, Thomas G. Albiges-Hizo C, Carroll R, Benig M, Phillips LA, Martin S, Ohagi S, Gardner P, Swift HH, Steiner DF: Proinsulin processing by the subtilisin-related proprotein covertases furin, PC2 and PC3. h c Natl Acad Sci USA 89:8822-2% 1982

34. Hutton JC: The insulin secretory granule. Diabetologia 32:271-81, 1989 35. Sizonenko S , Irminger J-C, Buhler L, Deng S, Morel P, Halban P: Kinetics

of proinsulin conversion in hun~an isltes. Diabetes 42:93%36, 1993 36. Taylor NA, Docherty K: Sequence requirements for processing of proin-

sulin in transfected mouse pituitary AtT20 cells. Biochem J 286:619-622, 1992

37. wkcs MA, Frank BH, Khait I, Beckage M, Shoelson SE, Neuringer LG: NMR and photo-CIDNP of human proinsulin and prohornone processing intermediates with application to endoproteas& recognition. ~ i o c h & 29:8389-8401, 1990

38. Rhodes CJ, Lincoln B, Shoelson SE: Preferential cleavage of des 31, 32 proinsulin over intact proinsulin by the insulin secretory granule type-I1 endopeptidase: implications for a favoured route for prohormone pro- cessing. J BioL Chmn 26722719-27, 1RY2

39. Rhodes CJ, Halban PA: Newly synthesized proinsulinlinsulin and stored insulin are released from pancreatic B-cells via a regulated, rather than a constitutive pathway. J Cell Biol 105:145-53, 1987

40. Rhodes CJ, Lucas CA, Mutkoski RL, Orci L, IIalban PA: Stimulation by ATP of oroinstllin to insulin conversion in isolated rat ~ancreat ic islet secretoly granules assoclatlon ul th ATP-dependent proton pump J Bzol C h n 262.10712-17 1987

41 Hutton JC The internal pH and membrane potential of the ~nsulin secretory granule. ~ i o c h & J 20-1:171-78, 1982

42. Hutton JC, Penn J2J, Peshavaria M: Low moIecular weight constituents of isolated insulin secretory granules: bivalent cations, adenine nucleotides, and inorganic phosphate. Biochem J 210:297-305. 1983

43. Fornlby B, Capito K, Egeberg J, Hedeskov CJ: Ca-activated ATPase activity in snbcellular fractions of mouse pancreatic islets. Am .I Physiol 230:44148, 1976

DIABETES, VOL. 43, APRIL 1994

44. Takagi H, Matsuzawa H, Ohta T, Yamsaki M, Inouye M: Studies on the 55. Giddings SI, Chirgwln J, Permutt MA: Effects of glucose on proinsulin strucl.ure and function of subtilisin E by protein engineering. Ann NY mRNA in rats in vivo. Diabetes 31524-29, 1982 Acad Sci G72:52-59, 1992 56. Permutt MA, Chirgwin J, Giddings S, Kakita K. Rotwein P: Insulin

45. Wilcox CA, N l e r RS: Posttranslational processing of the prohormone- biosynthesis and diabetes mellitus. Clin B i o c h a 14:230-36, 1981 cleaving Kex2 protease in the Sacchammuces cereuisiae secretory 57. German MS, Moss LG, Rutter WJ: Regulation of gene expression by pathway. J Cell Biol 115:297-307, 1991 glucose and calcium in transfected primary islet cultures. J Biol Chem

46. Rehemtulla A, Dorner A, Kaufman RJ: Regulation of PACEIfurin process- 265:22063-66. 1990 ing: requirement for a post-endoplasmic reticulum compartment and autoproteolytic activation. Proc Natl Acad Sci USA 89:823.539. 1992

47. Guest PC, Arden SD, Bennett DL, Clark A, Rutherford NG, Hutton JC: The post-translational processing and intracellular sorting of PC2 in the islets of Langerhans. J Biol C h a 267:22401-406, 1992

48. Alarcc~n C, Lincoln B, Rhodes CJ: The biosynthesis of the subtilisin- related proprotein covertase PC3, but not that of the PC2 convertase, is regulated by glucose in parallel to proinsulin biosynthesis in rat pancre- atic islets. J Biol C h m 268:427@0, 1993

49. Zhou Y, Lindberg 1: Purification and characterization of the prohonnone conver-tase PC1 (PC3). J Biol Chem 268:561&23, 1993

50. Nielsen DA, U'elsh M, Casadaban MJ, Steiner DF: Control of insulin gene expression in pancreatic p-cells and in an insulin-producing cell line, RIN-SF' cells. 1. Effects of glucose and CAMP on the transcription of insulin mRNA. J Biol Chem 260:1358&89, 1985

51. Welsh M, Nielsen DA, MacKrell AJ, Steiner DF: Control of insulin gene emression in oancreatic 0-cells and in an insulin-~roducine cell tine. RIN-SF' cells, i1. ~ e ~ u l a t i b n of insulin mRNA stability. J i i o l ~ h c & 260:13690-94, 1985

52. Welsh M, Scherberg N, Gilmore R, Steiner Dl? Translational control of insulin biosynthesis evidence for regulation of elongation, initiation and signal-recognition particle-mediated translational arrest by glucose. Bio- c h a J 235459-67, 1986

53. Nagamatsu S, B o l a JL, Grodsky GM; Direct effects of glucose on proins~~lin synthesis and processing during desensitization. Endocrinol- ogy 120:122531, 1987

54. Guest PG. Rhodes CJ, Hutton JC: Regulation of the biosynthesis of insulin secretory granule proteins: coordinate translational control is exerted on some but not all granule matrix constituents. Biochem J 257:431-37, 1989

Birch NP, racer HL, Hakes DJ, Loh YP: Coordinate regulation of mRNA levels of pro-opiomelanocortin and the candidate processing enzymes PC2 and PC3, but not furin, in rat pituitary intermediate lobe. B i o c h a Biophys Res Commun 179:1311-19, 1991 Froguel P, Zouali H, Vionnet N, Velho G, Vaxillaire M, Sun F, Lesage S, Stoffel M, Takeda J, Passa P, Permutt MA, Beckrnann JS, Bell GI, Cohen D: Familial hyperglycemia due to mutations in glucokinase: definition of a subtype of diabetes mellitus. N Engl J Med 328:697-702, 1993 Campbell IC, Hellqvist LNB. Taylor KW: Insulin biosynthesis and its regulation. Clin Sci 62:1982 Prentki M, Matchinsky FM: Ca2+, CAMP, and phospholipid-derived mes- sengers in coupling mechanisms of insulin secretion. Physiol Rev 67: 11%-1248, 1987 Goodison S, Kenna S, Ashcroft SJH: Control of insulin gene expression by glucose. B i o c h a J 285:563-68, 1992 Njus D, Kelley PM, Hamadek GJ: Bioenergetics of secretory vesicles. Biochim Biophys Acta 853:237-65, 1986 Halban PA: Proinsulin t racking and processing in the pancreatic pcell. TEM 261-65, 1990 Burgess TL, Kelly RB: C,onstitutive and regulated secretion of proteins. Ann Rev Cell Biol 3:243-93, 1987 Vindrola 0, Lindberg I: Biosynthesis of the prohormone convertasc mPCl in AtT-20 cells. Mol Endomino1 6:1088-1094, 1992 Gold G, Gishizky ML, Chick WL, Grodsky GM: Contrasting patterns of insulin biosynthesis, compartmental storage, and secretion: rat tumor verses islet cells. Diabetes 33:55641, 1984 Guest PG, Pipeleers D, Rossier J, Rhodes CJ, Hutton JC: Co-secretion of carboxgpeptidase-H and insulin from isolated rat islets of Langerhans. Biochem J 264503-508, 1989

DIABETES, VCJL. 43, APRIL 1994