- Email: [email protected]
Insulin Acts Intracellularly on Proteasomes through Insulin-Degrading Enzyme
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
244, 390–394 (1998)
RC988276
Insulin Acts Intracellularly on Proteasomes through Insulin-Degrading Enzyme William C. Duckworth, Robert G. Bennett, and Frederick G. Hamel Research Service, Veterans Affairs Medical Center, Departments of Internal Medicine and Biochemistry and Molecular Biology, University of Nebraska Medical Center, Omaha, Nebraska 68198
Received February 4, 1998
Insulin decreases cellular protein degradation, but the mechanism of this action is poorly understood. We have shown that insulin can have an inhibitory effect on the action of the proteasome in vitro, which requires the presence of insulin degrading enzyme (IDE). In this study we have used an antibody which inhibits the activity of IDE to show that IDE is required for insulin inhibition of protein degradation in intact cells. The anti-IDE antibody blocked the insulin effect on cellular degradation of proteins prelabeled with radioactive amino acids. The anti-IDE antibody also decreased insulin inhibition of proteasome degradation of a specific substrate in intact cells. These data suggest that insulin works intracellularly via IDE to inhibit protein degradation by the proteasome. Thus, IDE may function as an intracellular mediator for insulin effects on protein degradation. This is a novel signal transduction mechanism for peptide hormones. q 1998 Academic Press
Key Words: insulin; insulin-degrading enzyme; proteasome; signal transduction; protease.
Major progress has been made in the understanding of insulin signal transduction pathways and in the appreciation of the complexity of the process. No single series of events explains all of the diverse effects of insulin. Insulin alters glucose, lipid, and protein metabolism and has effects on ion transport and translocation, DNA and RNA expression and metabolism. Most studies of the mechanism of insulin action have been directed at glucose metabolism or at cell growth and mitogenesis. The former is a rapid, or short term effect requiring seconds to minutes of hormone exposure while the latter is a long term effect requiring hours to days. Insulin also has intermediate-term effects on protein and lipid metabolism which require minutes to hours to occur. Little attention has been directed at the specific signal transduction system responsible for these intermediate events (1). 0006-291X/98 $25.00
Insulin’s ability to inhibit protein degradation has been suggested to be the result of suppression of lysosomal or autophagic proteolysis in liver (2, 3, 4), heart (5, 6), muscle (7), and some cultured cells (8, 9). However, many of these studies compare proteolysis in the presence and complete absence of insulin, an unphysiological condition. Nutritional factors that affect protein degradation also complicate interpretation of the data. Mortimore’s studies (2, 3, 4, 10, 11) have shown that lysosomal degradation is controlled by selected amino acids, whose metabolism is altered by insulin. Thus insulin’s effect on lysosomal degradation may be manifest only under extreme, and unphysiological, conditions. In fact, Goldberg and colleagues (12) have shown that in diabetic rats alteration of protein degradation is the result of Ca//-dependent and ATP-dependent (proteasomal) systems, and the intralysosomal proteolytic system is unaffected. This suggests that insulin’s effect in the whole animal may be primarily on these systems. The proteasome makes a particularly important target for insulin action, since it is the main cytosolic proteolytic activity, and is involved in the control of the cell cycle, antigen presentation, ubiquitinmediated degradation, and regulation of transcription factors. Evidence supporting a direct intracellular effect of insulin has been developed using a number of different approaches, including intracellular injection of insulin (13). We have shown direct effects of insulin on the proteasome, the major cytosolic proteolytic activity and have shown that the insulin effect requires the presence of the insulin degrading enzyme, IDE (14). IDE is also a regulatory factor for the androgen and glucocorticoid receptors (15). IDE associates with these receptors and increases DNA binding, and thus, activity. Insulin added to these receptor-IDE complexes decreases binding to DNA. This is analogous to IDE’s effects on proteasome activity. From these and other observations, we have hypothesized that the effect of insulin on cellular protein degradation is mediated by a direct intracellular interaction with the IDE-proteasome complex re-
390
Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
sulting in inhibition of proteolysis (1). The present studies explore this hypothesis in intact hepatocytes, using total cell protein degradation and proteasome activity as measurements. The data show that insulin inhibits overall protein degradation and specific proteasome proteolysis in intact hepatocytes, both freshly isolated and cultured. Furthermore, a monoclonal antibody which inhibits IDE activity specifically blocks the insulin effect on protein degradation and proteasome activity. MATERIALS AND METHODS Crystalline porcine insulin and 125I [A14] iodoinsulin were graciously provided by Dr. Ronald Chance and Dr. Bruce Frank, respectively (Lilly Research Laboratories). HepG2 cells were obtained from ATCC (Rockville, MD). Dulbecco’s Modified Eagle’s Medium (DMEM) was purchased from Life Technologies (Grand Island, NY). Fetal bovine serum was obtained from Intergen Co. (Purchase, NY). Succinyl-leu-leu-val-tyr-7-amido-4-methyl coumarin (LLVY) and boc-leuser-thr-arg-7-amido-4-methyl coumarin (LSTR) were from Sigma (St Louis, MO). Methoxysuccinyl-phe-leu-phe-7-amido-4-trifluoromethyl coumarin (FLF) was purchased from Enzyme Systems Products (Dublin, CA). The proteasome inhibitor N-boc-ile-glu(O-t-butyl)-alaleucinal (PSI) was from Peptide International, Inc. (Louisville, KY). The proteasome inhibitor lactacystin was from E. J. Corey (Harvard University). All other reagents were of at least reagent grade. Cellular protein degradation. Male Sprague-Dawley rats were fasted overnight and hepatocytes were isolated by a modification (16) of the method of Terris and Steiner (17). Cells were resuspended in cell buffer at an approximate density of 106 cells/ml, and incubated 30 min. at 377C. Cell viability was approximately 90%. Cellular protein degradation was measured by labeling cells with buffer containing 3 H-leucine and 5 times the normal serum concentration (18) of 19 unlabeled amino acids (excluding leucine) for 60 minutes, then washing and chasing with 2 mM unlabeled leucine. The cells were divided into 3 portions, and either 30mg/ml anti-IDE antibody C20-3.1a (14), nonspecific mouse IgG (30 mg/ml), or PBS buffer was added. To each of these systems, the following were added: 5X amino acids, 10 nM porcine insulin, or no addition. Five times the normal serum concentration of amino acids were used as a positive control to demonstrate the decrease in protein degradation due to selective inhibition as shown by Mortimore (2, 3, 4). These effects are independent of the insulin regulatory system. The cells were incubated at 377C, and 0.5 ml aliquots were taken at 0 and 120 minutes, and counted. Cellular degradation was determined by the difference in solubility (in 12.5% TCA) at 0 and 120 minutes. In vitro proteasome activity. Partially purified insulin-degrading enzyme, complexed with the multicatalytic proteinase, was obtained from rat skeletal muscle as described previously (19). IDE-proteasome complex was preincubated 5 minutes with increasing amounts of anti-IDE antibody C20-3.1A. Degradation of insulin 125I-labeled at the A14 position was measured by the generation of trichloroacetic acid soluble counts. Proteasome activity was measured with LLVY and LSTR for determination of chymotrypsin-like and trypsin-like activities, respectively, as described before (14). Cellular proteasome activity. Human hepatoma (HepG2) cells were maintained in DMEM supplemented with 10% fetal bovine serum in a 5% CO2/95% air environment. For peptide degradation assays, subconfluent cultures were serum deprived overnight (18 hours) prior to treatment. Peptide degradation was assessed with the membrane permeable substrate FLF by a modification of a previously published method (20). After hormone and/or inhibitor treatment, FLF was added to the cells to a final concentration of 13 mM, and incubated 1 hour. The DMSO concentration from the addition of
TABLE 1
Anti-IDE Antibody Eliminates Inhibition of Protein Degradation in Isolated Rat Hepatocytes
Control 5X amino acids insulin
No antibody
Anti-IDE
Nonspecific IgG
100 78.8 { 4.6* 83.5 { 6.4*
100 85.0 { 2.4* 99.7 { 2.5
100 78.4 { 6.5* 78.2 { 6.2*
Note. Cellular protein degradation as measured by the release of radiolabeled amino acids is inhibited by insulin and excess amino acids in the buffer or a non-specific IgG. In the presence of antiIDE antibody (30 mg/ml), however, insulin no longer inhibits protein degradation. The data are expressed as the percent soluble label released normalized to that in cells incubated without insulin or excess amino acids. * Significantly different from control (p õ 0.05).
FLF or inhibitors did not exceed 0.15%. The cells were disrupted by sonication, and the fluorescence due to degradation of FLF was measured at excitation 390 nm, emission 515 nm. Antibody studies. Subconfluent cultures of HepG2 cells were serum-deprived overnight, then osmotically loaded with the inhibitory anti-IDE monoclonal antibody C20-3.1A (50mg/ml) by the method of Okada and Rechsteiner (21). The cells were allowed to recover for 1 hour, then were treated with the indicated concentrations of insulin for 1 hour. The fluorogenic proteasome substrate FLF was added, and the cells were incubated for 1 hour. Fluorescence due to degradation of FLF was measured as above.
RESULTS The experiment in Table 1 shows a potential intracellular insulin-IDE effect on total cellular protein degradation. Protein degradation was decreased by incubation with insulin or an excess of amino acids. Prelabeled cells were also incubated with a monoclonal antibody which inhibits IDE activity, but does not interact with insulin (data not shown). Antibody-treated cells no longer responded to insulin by decreasing protein degradation but did respond to the addition of a five fold excess of amino acids which work independent of insulin. These data support a selective role for IDE in the cellular response to insulin for inhibition of protein degradation. The data in Table 1 do not indicate the specific site of the insulin-IDE effect. Since we have shown an insulinIDE-proteasome interaction, isolated proteasomes were incubated with varying amounts of anti-IDE antibody with and without insulin (Table 2). Two different catalytic sites of the proteasome, the trypsin-like and the chymotrypsin-like, were assayed using artificial substrates. As shown previously, insulin inhibited LLVY (chymotrypsin-like) and LSTR (trypsin-like) degradation. The anti-IDE antibody blocked the insulin effect on LLVY and LSTR degradation in a dose dependent manner. To substantiate an insulin-IDE proteasome effect, a
391
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 2
Inhibition of IDE with a Monoclonal Antibody Decreases Insulin Inhibition of the Proteasome in Vitro % inhibition by 1 mM insulin
mg/ml antibody
% insulin degradation
0 20 40 60
100 94.6 { 8.5 80.9 { 10.7 68.2 { 9.8
LLVY 64.5 45.8 45.2 37.2
{ { { {
4.9 8.7 4.6 3.9*
LSTR 69.9 52.6 38.8 47.0
{ { { {
2.9 7.3 13.7 9.4
Note. Insulin degradation was measured in the presence of increasing amounts of anti-IDE antibody. The data are expressed as the percent acid soluble counts per 15 minute incubation. Proteasome activities were measured under the same conditions in the presence or absence of 1 mM insulin. Data are expressed as the percent inhibition of proteasome activity compared with that in the absence of insulin. Data are mean { SEM for 3 independent experiments. * Significant different versus no antibody (p õ 0.05).
substrate for assessing proteasome activity in intact cells was necessary. Methoxysuccinyl-phe-leu-phe-7amido-4-trifluoromethyl coumarin (FLF) is degraded by the chymotrypsin-like catalytic activity of proteasomes and freely crosses cell membranes. We have shown that FLF degradation by intact cells is inhibited by insulin and that FLF degradation appears to be due to proteasome activity (22). The latter conclusion was based on the differential effects of calpain inhibitor I (N-acetyl-leu-leu-norleucinal, ALLN) and calpain inhibitor II (N-acetyl-leu-leu-methioninal, ALLM) on FLF degradation by cells which was consistent with that seen for the proteasome (23). Both of these agents inhibit calpains with similar potencies. Calpain inhibitors I and II also inhibit proteasome activity, but calpain inhibitor I is more effective. Calpain inhibitor I (ALLN) inhibited FLF degradation by intact cells much better than did calpain inhibitor II (ALLM), suggesting the involvement of proteasomes in this process. While suggestive, the effect of non-specific inhibitors such as calpain inhibitors I and II did not establish the proteasomes as the mechanism for FLF degradation in intact cells. This was examined further by using two specific proteasome inhibitors. Figure 1 shows the effects of PSI, a specific inhibitor of the proteasome chymotrypsin-like activity (24), and lactacystin, an inhibitor which modifies active site threonine residues in the proteasome (25), on FLF degradation. PSI and lactacystin inhibited cellular FLF degradation by 90% and 70%, respectively. In contrast, the lysosomal inhibitor methylamine had no effect. These results confirm that the majority of FLF degradation by HepG2 cells is due to the proteasome with little if any contribution from lysosomes. Insulin had no further effect on FLF degradation after inhibition of the proteasome by these agents (Table 3), indicating insulin is not acting on other proteolytic systems.
FIG. 1. Effect of protease inhibitors on FLF degradation in HepG2 cells. Subconfluent cultures were serum-deprived overnight, then treated with the indicated concentrations of inhibitor for 2 hours, followed by addition of substrate (FLF) for 1 hour. Data for lactacystin (open circles), PSI (closed squares),and methylamine (closed circles) are expressed as mean { S.E.M. of % fluorescence with addition of inhibitor vehicle only. Results are from six replicate wells.
We therefore wanted to demonstrate that the in vivo effect of insulin also involved mediation by IDE as does the in vitro effect. Cultured hepatoma cells were osmotically loaded with IDE inhibitory monoclonal antibody or with vehicle only. After 1 hour incubation to allow recovery, insulin was added to the cells and FLF degradation assayed. In the cells loaded with vehicle only, insulin inhibited proteasome activity in a dose dependent manner as expected. Cells containing inhibitory antibody had a greatly diminished response to the hormone (Figure 2). A similar experiment (Table 4) at one concentration of insulin showed no effect of an unrelated mouse IgG on the insulin response. Thus the effect is specific for the anti-IDE antibody.
TABLE 3
Effect of Insulin on Residual FLF Degrading Activity in HepG2 Cells after Inhibition with Proteasome-Specific Inhibitors FLF degradation Inhibitor
0 Insulin
/ Insulin
% Inhibition
None Lactacystin PSI
111.7 { 2.7 63.2 { 0.8 27.2 { 1.7
98.0 { 0.4* 63.8 { 0.9 25.5 { 1.1
12.2 01.1 6.1
Note. The effect of insulin on FLF degrading activity remaining after treatment of cells with proteasome inhibitors was examined. Subconfluent HepG2 cells were serum-deprived overnight, then treated with inhibitor (1.0 mM) or vehicle for 2 hours, followed by the addition of 1.0 nM insulin and FLF as described in Methods. Data are expressed as the FLF degradation (fluorescence), mean { S.E.M. for 6 replicate wells (*p õ .001).
392
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Osmotic loading of anti-IDE antibody into cultured HepG2 cells decreases insulin inhibition of intracellular degrading activity by the proteasome. Anti-IDE antibody (50 mg/ml) was loaded into HepG2 cells, and the cells were treated with increasing amounts of insulin. The effect on proteasome activity was measured with the membrane permeable chymotrypsin-like substrate methoxysuccinylphe-leu-phe-7-amid-4-trifluoromethyl coumarin (FLF). The effect of anti-IDE is compared with that of cells loaded with vehicle only. Data are expressed as the percent FLF degrading activity compared with that in the absence of insulin (mean { SEM for 5 replicate wells). * p õ 0.05 ** p õ 0.01.
DISCUSSION The studies presented here describe a new mechanism of action for the effects of insulin on cellular protein degradation. Recent studies have suggested that under normal physiological conditions, (e.g. postprandial hormone changes) insulin inhibits protein degradation by altering cytosolic proteolysis, i.e., protein degradation by proteasomes and related complexes (26). We have shown that insulin has direct effects in vitro on proteasome activity, inhibiting at least two of the multiple catalytic sites of this complex (14, 27), and that insulin inhibits the chymotrypsin-like site in intact cells (22). These data raise the possibility that the effect of insulin on cellular protein degradation may be due to a direct intracellular interaction of the hormone with the cytosolic regulatory protein, IDE. This interaction results in inhibition of proteasome catalytic activity. This hypothesis is supported by previous work from several laboratories. Receptor bound insulin is internalized and delivered to the cytosol where it associates with IDE (28) . IDE is a regulatory factor for at least three cytosolic complexes, the proteasome (1) and the androgen and glucocorticoid receptors (15), stimulating the activity of these proteins. Insulin reverses this stimulation resulting in inhibition of proteasomes and a decrease in protein degradation. The same effect may occur with the steroid receptors, producing insulin antagonism of glucocorticoid and androgen effects. IDE also has a peroxisomal localization (29) and insulin inhibits b-oxidation of lipids when added in vitro to purified peroxisomes (30), raising the possibility that some
of the effects of insulin on lipid metabolism may be through the same mechanism. A direct role for intracellular insulin has long been postulated (31) with much evidence to support this suggestion. Miller showed that intracellular injection of insulin into oocytes resulted in insulin effects on the cells (32). Hofmann showed that insulin internalized by receptors other than the insulin receptor, has cellular effects (33). Inhibitors of intracellular insulin processing block insulin inhibition of protein degradation (34, 16). In spite of these, and many other data, general acceptance of an intracellular action of insulin has not occurred nor has this possibility been extensively investigated. The lack of a testable hypothesis for a mechanism producing intracellular effects of insulin has limited approaches to the question. The cytosolic interaction of insulin with a regulatory protein provides a testable hypothesis. IDE (insulinase, insulin protease, insulysin) was discovered and characterized by its most obvious in vitro property, a high specificity for insulin binding and degradation (35). Its unique properties resulted in controversy about the classification of this enzyme until isolation of the cDNA (36). Analysis revealed that IDE does not belong to any of the classical proteinase groups, but is instead the initial representative of a new superclass of metalloenzymes. In addition to insulin degradation, it has been implicated in growth factor binding and degradation, degradation of other cellular proteins, and in cellular growth and differentiation (37). IDE’s primary cellular role may be as a regulatory factor for cytosolic complexes and as a receptor for hormones which alter activity of these complexes. Other hormones, including IGF-I, IGF-II, TGFa, ANF, and proinsulin bind to IDE with varying affinities and susceptibility to degradation (38). In vitro studies with isolated proteasomes show inhibition of proteasome activity by these ligands (27). Potentially, some effects of these
TABLE 4
Insulin Inhibition of Proteasome Activity Is Reduced after Delivery of Anti-IDE But Not Control Antibody Loading conditions
mg/ml
No loading Vehicle Mouse antibody Anti-IDE antibody
— — 10 5.0
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
91.6 94.4 90.1 107.4
{ { { {
3.8 3.2 5.4 4.8*‡
Note. Introduction of anti-IDE antibody reduces insulin inhibition of proteasome activity. Cells were osmotically loaded with anti-IDE, isotype-matched non-specific mouse IgG, or vehicle alone. The effect of insulin (10011 M) on FLF degradation was determined. Data are mean { S.E.M. for 6 replicate wells. * p õ 0.05 compared with no loading. ‡ p õ 0.05 compared with mouse non-specific antibody.
393
AID
% activity with 10011 M insulin
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
hormones could be mediated by intracellular interaction with IDE. Regardless of potential implications for other actions of insulin or for actions of other hormones, the present data support an effect of insulin on cellular proteolysis requiring intracellular interaction of insulin and IDE with the proteasome. Short term (glucose metabolism) and long term (mitogenesis) actions of insulin have different post receptor signal transduction systems. We propose that the intermediate term effects of insulin on protein and possibly lipid metabolism also have a different signal transduction mechanism which involves post receptor interaction with the cytosolic protein, IDE. In addition, with the demonstration of a regulatory role for IDE we suggest that a new trivial name for this protein is appropriate. We propose the name ‘‘Insulin Regulatory Protein’’ (IRP) to reflect its central role in regulating insulin (and potentially insulin-related peptides) levels and activity. ACKNOWLEDGMENTS This research was supported in part by the Department of Veterans Affairs Research Service and in part by the Bly Memorial Research Fund, University of Nebraska Medical Center. The authors thank Gerri Siford and Kim Harmon for technical assistance and Janet Corr for preparing the manuscript.
REFERENCES 1. Duckworth, W. C., Bennett, R. G., and Hamel, F. G. (1997) J. Invest. Med. 45, 20–27. 2. Mortimore, G. E., and Poso, A. R. (1986) Adv. Enzyme Regul. 25, 257–276. 3. Mortimore, G. E., and Poso, A. R. (1987) Annu. Rev. Nutr. 7, 539–564. 4. Mortimore, G. E., Poso, A. R., and Lardeux, B. R. (1989) Diabetes Metab. Rev. 5, 49–70. 5. Chua, B., Kao, R., Rannels, D. E., and Morgan, H. E. (1978) Biochem. Soc. Symp. 43, 1–15. 6. Long, W. M., Chua, B. H., Munger, B. L., and Morgan, H. E. (1984) Fed. Proc. 43, 1295–1300. 7. Furuno, K., and Goldberg, A. L. (1986) Biochem. J. 237, 859– 864. 8. Gulve, E. A., and Dice, J. F. (1989) Biochem. J. 260, 377–387. 9. Gulve, E. A., Mabuchi, K., and Dice, J. F. (1991) J. Cell. Physiol. 147, 37–45. 10. Mortimore, G. E., and Poso, A. R. (1984) Fed. Proc. 43, 1289– 1294.
11. Mortimore, G. E., Poso, A. R., Kadowaki, M., and Wert, J. J., Jr. (1987) J. Biol. Chem. 262, 16322–16327. 12. Pepato, M. T., Migliorini, R. H., Goldberg, A. L., and Kettelhut, I. C. (1996) Am. J. Physiol. 271, E340–E347. 13. Miller, D. S. (1988) Science 240, 506–509. 14. Duckworth, W. C., Bennett, R. G., and Hamel, F. G. (1994) J. Biol. Chem. 269, 24575–24580. 15. Kupfer, S. R., Wilson, E. M., and French, F. S. (1994) J. Biol. Chem. 269, 20622–20628. 16. Peavy, D. E., Edmondson, J. W., and Duckworth, W. C. (1984) Endocrinology 114, 753–760. 17. Terris, S., and Steiner, D. F. (1975) J. Biol. Chem. 250, 8389– 8398. 18. Tolman, E. L., Schworer, C. M., and Jefferson, L. S. (1973) J. Biol. Chem. 248, 4552–4560. 19. Duckworth, W. C., Heineman, M. A., and Kitabchi, A. E. (1972) Proc. Natl. Acad. Sci. USA 69, 3698–3702. 20. Harding, C. V., France, J., Song, R., Farah, J. M., Chatterjee, S., Iqbal, M., and Siman, R. (1995) J. Immunol. 155, 1767–1775. 21. Okada, C. Y., and Rechsteiner, M. (1982) Cell 29, 33–41. 22. Hamel, F. G., Bennett, R. G., Harmon, K. S., and Duckworth, W. C. (1997) Biochem. Biophys. Res. Commun. 234, 671–674. 23. Rock, K. L., Gramm, C., Rothstein, L., Clark, K., Stein, R., Dick, L., Hwang, D., and Goldberg, A. L. (1994) Cell 78, 761–771. 24. Figueiredo-Pereira, M. E., Berg, K. A., and Wilk, S. (1994) J. Neurochem. 63, 1578–1581. 25. Fenteany, G., Standaert, R. F., Lane, W. S., Choi, S., Corey, E. J., and Schreiber, S. L. (1995) Science 268, 726–731. 26. Barrett, E., Jahn, L., Oliveras, E., and Fryburg, D. (1995) Am. J. Physiol. 268, E820–E824. 27. Bennett, R. G., Hamel, F. G., and Duckworth, W. C. (1997) Diabetes 46, 197–203. 28. Hari, J., Shii, K., and Roth, R. A. (1987) Endocrinology 120, 829– 831. 29. Authier, F., Bergeron, J. J. M., Ou, W.-J., Rachubinski, R. A., Posner, B. I., and Walton, P. A. (1995) Proc. Natl. Acad. Sci. USA 92, 3859–3863. 30. Bennett, R. G., Hamel, F. G., and Duckworth, W. C. (1996) Diabetes 45(Suppl. 2), 310A. 31. Steiner, D. F. (1977) Diabetes 26, 322–331. 32. Miller, D., and Sykes, D. B. (1991) J. Cell. Physiol. 147, 487– 494. 33. Hofmann, C. A., Lotan, R. M., Ku, W. W., and Oeltmann, T. N. (1983) J. Biol. Chem. 258, 11774–11779. 34. Draznin, B., and Trowbridge, M. (1982) J. Biol. Chem. 257, 11988–11993. 35. Mirsky, I. A. (1957) Recent Prog. Horm. Res. 13, 429–471. 36. Affholter, J. A., Fried, V. A., and Roth R. A. (1988) Science 242, 1415–1418. 37. Duckworth, W. C. (1988) Endocr. Rev. 9, 319–345. 38. Authier, F., Posner, B. I., and Bergeron, J. J. M. (1996) Clin. Invest. Med. 19, 149–160.
394
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
244, 390–394 (1998)
RC988276
Insulin Acts Intracellularly on Proteasomes through Insulin-Degrading Enzyme William C. Duckworth, Robert G. Bennett, and Frederick G. Hamel Research Service, Veterans Affairs Medical Center, Departments of Internal Medicine and Biochemistry and Molecular Biology, University of Nebraska Medical Center, Omaha, Nebraska 68198
Received February 4, 1998
Insulin decreases cellular protein degradation, but the mechanism of this action is poorly understood. We have shown that insulin can have an inhibitory effect on the action of the proteasome in vitro, which requires the presence of insulin degrading enzyme (IDE). In this study we have used an antibody which inhibits the activity of IDE to show that IDE is required for insulin inhibition of protein degradation in intact cells. The anti-IDE antibody blocked the insulin effect on cellular degradation of proteins prelabeled with radioactive amino acids. The anti-IDE antibody also decreased insulin inhibition of proteasome degradation of a specific substrate in intact cells. These data suggest that insulin works intracellularly via IDE to inhibit protein degradation by the proteasome. Thus, IDE may function as an intracellular mediator for insulin effects on protein degradation. This is a novel signal transduction mechanism for peptide hormones. q 1998 Academic Press
Key Words: insulin; insulin-degrading enzyme; proteasome; signal transduction; protease.
Major progress has been made in the understanding of insulin signal transduction pathways and in the appreciation of the complexity of the process. No single series of events explains all of the diverse effects of insulin. Insulin alters glucose, lipid, and protein metabolism and has effects on ion transport and translocation, DNA and RNA expression and metabolism. Most studies of the mechanism of insulin action have been directed at glucose metabolism or at cell growth and mitogenesis. The former is a rapid, or short term effect requiring seconds to minutes of hormone exposure while the latter is a long term effect requiring hours to days. Insulin also has intermediate-term effects on protein and lipid metabolism which require minutes to hours to occur. Little attention has been directed at the specific signal transduction system responsible for these intermediate events (1). 0006-291X/98 $25.00
Insulin’s ability to inhibit protein degradation has been suggested to be the result of suppression of lysosomal or autophagic proteolysis in liver (2, 3, 4), heart (5, 6), muscle (7), and some cultured cells (8, 9). However, many of these studies compare proteolysis in the presence and complete absence of insulin, an unphysiological condition. Nutritional factors that affect protein degradation also complicate interpretation of the data. Mortimore’s studies (2, 3, 4, 10, 11) have shown that lysosomal degradation is controlled by selected amino acids, whose metabolism is altered by insulin. Thus insulin’s effect on lysosomal degradation may be manifest only under extreme, and unphysiological, conditions. In fact, Goldberg and colleagues (12) have shown that in diabetic rats alteration of protein degradation is the result of Ca//-dependent and ATP-dependent (proteasomal) systems, and the intralysosomal proteolytic system is unaffected. This suggests that insulin’s effect in the whole animal may be primarily on these systems. The proteasome makes a particularly important target for insulin action, since it is the main cytosolic proteolytic activity, and is involved in the control of the cell cycle, antigen presentation, ubiquitinmediated degradation, and regulation of transcription factors. Evidence supporting a direct intracellular effect of insulin has been developed using a number of different approaches, including intracellular injection of insulin (13). We have shown direct effects of insulin on the proteasome, the major cytosolic proteolytic activity and have shown that the insulin effect requires the presence of the insulin degrading enzyme, IDE (14). IDE is also a regulatory factor for the androgen and glucocorticoid receptors (15). IDE associates with these receptors and increases DNA binding, and thus, activity. Insulin added to these receptor-IDE complexes decreases binding to DNA. This is analogous to IDE’s effects on proteasome activity. From these and other observations, we have hypothesized that the effect of insulin on cellular protein degradation is mediated by a direct intracellular interaction with the IDE-proteasome complex re-
390
Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
sulting in inhibition of proteolysis (1). The present studies explore this hypothesis in intact hepatocytes, using total cell protein degradation and proteasome activity as measurements. The data show that insulin inhibits overall protein degradation and specific proteasome proteolysis in intact hepatocytes, both freshly isolated and cultured. Furthermore, a monoclonal antibody which inhibits IDE activity specifically blocks the insulin effect on protein degradation and proteasome activity. MATERIALS AND METHODS Crystalline porcine insulin and 125I [A14] iodoinsulin were graciously provided by Dr. Ronald Chance and Dr. Bruce Frank, respectively (Lilly Research Laboratories). HepG2 cells were obtained from ATCC (Rockville, MD). Dulbecco’s Modified Eagle’s Medium (DMEM) was purchased from Life Technologies (Grand Island, NY). Fetal bovine serum was obtained from Intergen Co. (Purchase, NY). Succinyl-leu-leu-val-tyr-7-amido-4-methyl coumarin (LLVY) and boc-leuser-thr-arg-7-amido-4-methyl coumarin (LSTR) were from Sigma (St Louis, MO). Methoxysuccinyl-phe-leu-phe-7-amido-4-trifluoromethyl coumarin (FLF) was purchased from Enzyme Systems Products (Dublin, CA). The proteasome inhibitor N-boc-ile-glu(O-t-butyl)-alaleucinal (PSI) was from Peptide International, Inc. (Louisville, KY). The proteasome inhibitor lactacystin was from E. J. Corey (Harvard University). All other reagents were of at least reagent grade. Cellular protein degradation. Male Sprague-Dawley rats were fasted overnight and hepatocytes were isolated by a modification (16) of the method of Terris and Steiner (17). Cells were resuspended in cell buffer at an approximate density of 106 cells/ml, and incubated 30 min. at 377C. Cell viability was approximately 90%. Cellular protein degradation was measured by labeling cells with buffer containing 3 H-leucine and 5 times the normal serum concentration (18) of 19 unlabeled amino acids (excluding leucine) for 60 minutes, then washing and chasing with 2 mM unlabeled leucine. The cells were divided into 3 portions, and either 30mg/ml anti-IDE antibody C20-3.1a (14), nonspecific mouse IgG (30 mg/ml), or PBS buffer was added. To each of these systems, the following were added: 5X amino acids, 10 nM porcine insulin, or no addition. Five times the normal serum concentration of amino acids were used as a positive control to demonstrate the decrease in protein degradation due to selective inhibition as shown by Mortimore (2, 3, 4). These effects are independent of the insulin regulatory system. The cells were incubated at 377C, and 0.5 ml aliquots were taken at 0 and 120 minutes, and counted. Cellular degradation was determined by the difference in solubility (in 12.5% TCA) at 0 and 120 minutes. In vitro proteasome activity. Partially purified insulin-degrading enzyme, complexed with the multicatalytic proteinase, was obtained from rat skeletal muscle as described previously (19). IDE-proteasome complex was preincubated 5 minutes with increasing amounts of anti-IDE antibody C20-3.1A. Degradation of insulin 125I-labeled at the A14 position was measured by the generation of trichloroacetic acid soluble counts. Proteasome activity was measured with LLVY and LSTR for determination of chymotrypsin-like and trypsin-like activities, respectively, as described before (14). Cellular proteasome activity. Human hepatoma (HepG2) cells were maintained in DMEM supplemented with 10% fetal bovine serum in a 5% CO2/95% air environment. For peptide degradation assays, subconfluent cultures were serum deprived overnight (18 hours) prior to treatment. Peptide degradation was assessed with the membrane permeable substrate FLF by a modification of a previously published method (20). After hormone and/or inhibitor treatment, FLF was added to the cells to a final concentration of 13 mM, and incubated 1 hour. The DMSO concentration from the addition of
TABLE 1
Anti-IDE Antibody Eliminates Inhibition of Protein Degradation in Isolated Rat Hepatocytes
Control 5X amino acids insulin
No antibody
Anti-IDE
Nonspecific IgG
100 78.8 { 4.6* 83.5 { 6.4*
100 85.0 { 2.4* 99.7 { 2.5
100 78.4 { 6.5* 78.2 { 6.2*
Note. Cellular protein degradation as measured by the release of radiolabeled amino acids is inhibited by insulin and excess amino acids in the buffer or a non-specific IgG. In the presence of antiIDE antibody (30 mg/ml), however, insulin no longer inhibits protein degradation. The data are expressed as the percent soluble label released normalized to that in cells incubated without insulin or excess amino acids. * Significantly different from control (p õ 0.05).
FLF or inhibitors did not exceed 0.15%. The cells were disrupted by sonication, and the fluorescence due to degradation of FLF was measured at excitation 390 nm, emission 515 nm. Antibody studies. Subconfluent cultures of HepG2 cells were serum-deprived overnight, then osmotically loaded with the inhibitory anti-IDE monoclonal antibody C20-3.1A (50mg/ml) by the method of Okada and Rechsteiner (21). The cells were allowed to recover for 1 hour, then were treated with the indicated concentrations of insulin for 1 hour. The fluorogenic proteasome substrate FLF was added, and the cells were incubated for 1 hour. Fluorescence due to degradation of FLF was measured as above.
RESULTS The experiment in Table 1 shows a potential intracellular insulin-IDE effect on total cellular protein degradation. Protein degradation was decreased by incubation with insulin or an excess of amino acids. Prelabeled cells were also incubated with a monoclonal antibody which inhibits IDE activity, but does not interact with insulin (data not shown). Antibody-treated cells no longer responded to insulin by decreasing protein degradation but did respond to the addition of a five fold excess of amino acids which work independent of insulin. These data support a selective role for IDE in the cellular response to insulin for inhibition of protein degradation. The data in Table 1 do not indicate the specific site of the insulin-IDE effect. Since we have shown an insulinIDE-proteasome interaction, isolated proteasomes were incubated with varying amounts of anti-IDE antibody with and without insulin (Table 2). Two different catalytic sites of the proteasome, the trypsin-like and the chymotrypsin-like, were assayed using artificial substrates. As shown previously, insulin inhibited LLVY (chymotrypsin-like) and LSTR (trypsin-like) degradation. The anti-IDE antibody blocked the insulin effect on LLVY and LSTR degradation in a dose dependent manner. To substantiate an insulin-IDE proteasome effect, a
391
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 2
Inhibition of IDE with a Monoclonal Antibody Decreases Insulin Inhibition of the Proteasome in Vitro % inhibition by 1 mM insulin
mg/ml antibody
% insulin degradation
0 20 40 60
100 94.6 { 8.5 80.9 { 10.7 68.2 { 9.8
LLVY 64.5 45.8 45.2 37.2
{ { { {
4.9 8.7 4.6 3.9*
LSTR 69.9 52.6 38.8 47.0
{ { { {
2.9 7.3 13.7 9.4
Note. Insulin degradation was measured in the presence of increasing amounts of anti-IDE antibody. The data are expressed as the percent acid soluble counts per 15 minute incubation. Proteasome activities were measured under the same conditions in the presence or absence of 1 mM insulin. Data are expressed as the percent inhibition of proteasome activity compared with that in the absence of insulin. Data are mean { SEM for 3 independent experiments. * Significant different versus no antibody (p õ 0.05).
substrate for assessing proteasome activity in intact cells was necessary. Methoxysuccinyl-phe-leu-phe-7amido-4-trifluoromethyl coumarin (FLF) is degraded by the chymotrypsin-like catalytic activity of proteasomes and freely crosses cell membranes. We have shown that FLF degradation by intact cells is inhibited by insulin and that FLF degradation appears to be due to proteasome activity (22). The latter conclusion was based on the differential effects of calpain inhibitor I (N-acetyl-leu-leu-norleucinal, ALLN) and calpain inhibitor II (N-acetyl-leu-leu-methioninal, ALLM) on FLF degradation by cells which was consistent with that seen for the proteasome (23). Both of these agents inhibit calpains with similar potencies. Calpain inhibitors I and II also inhibit proteasome activity, but calpain inhibitor I is more effective. Calpain inhibitor I (ALLN) inhibited FLF degradation by intact cells much better than did calpain inhibitor II (ALLM), suggesting the involvement of proteasomes in this process. While suggestive, the effect of non-specific inhibitors such as calpain inhibitors I and II did not establish the proteasomes as the mechanism for FLF degradation in intact cells. This was examined further by using two specific proteasome inhibitors. Figure 1 shows the effects of PSI, a specific inhibitor of the proteasome chymotrypsin-like activity (24), and lactacystin, an inhibitor which modifies active site threonine residues in the proteasome (25), on FLF degradation. PSI and lactacystin inhibited cellular FLF degradation by 90% and 70%, respectively. In contrast, the lysosomal inhibitor methylamine had no effect. These results confirm that the majority of FLF degradation by HepG2 cells is due to the proteasome with little if any contribution from lysosomes. Insulin had no further effect on FLF degradation after inhibition of the proteasome by these agents (Table 3), indicating insulin is not acting on other proteolytic systems.
FIG. 1. Effect of protease inhibitors on FLF degradation in HepG2 cells. Subconfluent cultures were serum-deprived overnight, then treated with the indicated concentrations of inhibitor for 2 hours, followed by addition of substrate (FLF) for 1 hour. Data for lactacystin (open circles), PSI (closed squares),and methylamine (closed circles) are expressed as mean { S.E.M. of % fluorescence with addition of inhibitor vehicle only. Results are from six replicate wells.
We therefore wanted to demonstrate that the in vivo effect of insulin also involved mediation by IDE as does the in vitro effect. Cultured hepatoma cells were osmotically loaded with IDE inhibitory monoclonal antibody or with vehicle only. After 1 hour incubation to allow recovery, insulin was added to the cells and FLF degradation assayed. In the cells loaded with vehicle only, insulin inhibited proteasome activity in a dose dependent manner as expected. Cells containing inhibitory antibody had a greatly diminished response to the hormone (Figure 2). A similar experiment (Table 4) at one concentration of insulin showed no effect of an unrelated mouse IgG on the insulin response. Thus the effect is specific for the anti-IDE antibody.
TABLE 3
Effect of Insulin on Residual FLF Degrading Activity in HepG2 Cells after Inhibition with Proteasome-Specific Inhibitors FLF degradation Inhibitor
0 Insulin
/ Insulin
% Inhibition
None Lactacystin PSI
111.7 { 2.7 63.2 { 0.8 27.2 { 1.7
98.0 { 0.4* 63.8 { 0.9 25.5 { 1.1
12.2 01.1 6.1
Note. The effect of insulin on FLF degrading activity remaining after treatment of cells with proteasome inhibitors was examined. Subconfluent HepG2 cells were serum-deprived overnight, then treated with inhibitor (1.0 mM) or vehicle for 2 hours, followed by the addition of 1.0 nM insulin and FLF as described in Methods. Data are expressed as the FLF degradation (fluorescence), mean { S.E.M. for 6 replicate wells (*p õ .001).
392
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Osmotic loading of anti-IDE antibody into cultured HepG2 cells decreases insulin inhibition of intracellular degrading activity by the proteasome. Anti-IDE antibody (50 mg/ml) was loaded into HepG2 cells, and the cells were treated with increasing amounts of insulin. The effect on proteasome activity was measured with the membrane permeable chymotrypsin-like substrate methoxysuccinylphe-leu-phe-7-amid-4-trifluoromethyl coumarin (FLF). The effect of anti-IDE is compared with that of cells loaded with vehicle only. Data are expressed as the percent FLF degrading activity compared with that in the absence of insulin (mean { SEM for 5 replicate wells). * p õ 0.05 ** p õ 0.01.
DISCUSSION The studies presented here describe a new mechanism of action for the effects of insulin on cellular protein degradation. Recent studies have suggested that under normal physiological conditions, (e.g. postprandial hormone changes) insulin inhibits protein degradation by altering cytosolic proteolysis, i.e., protein degradation by proteasomes and related complexes (26). We have shown that insulin has direct effects in vitro on proteasome activity, inhibiting at least two of the multiple catalytic sites of this complex (14, 27), and that insulin inhibits the chymotrypsin-like site in intact cells (22). These data raise the possibility that the effect of insulin on cellular protein degradation may be due to a direct intracellular interaction of the hormone with the cytosolic regulatory protein, IDE. This interaction results in inhibition of proteasome catalytic activity. This hypothesis is supported by previous work from several laboratories. Receptor bound insulin is internalized and delivered to the cytosol where it associates with IDE (28) . IDE is a regulatory factor for at least three cytosolic complexes, the proteasome (1) and the androgen and glucocorticoid receptors (15), stimulating the activity of these proteins. Insulin reverses this stimulation resulting in inhibition of proteasomes and a decrease in protein degradation. The same effect may occur with the steroid receptors, producing insulin antagonism of glucocorticoid and androgen effects. IDE also has a peroxisomal localization (29) and insulin inhibits b-oxidation of lipids when added in vitro to purified peroxisomes (30), raising the possibility that some
of the effects of insulin on lipid metabolism may be through the same mechanism. A direct role for intracellular insulin has long been postulated (31) with much evidence to support this suggestion. Miller showed that intracellular injection of insulin into oocytes resulted in insulin effects on the cells (32). Hofmann showed that insulin internalized by receptors other than the insulin receptor, has cellular effects (33). Inhibitors of intracellular insulin processing block insulin inhibition of protein degradation (34, 16). In spite of these, and many other data, general acceptance of an intracellular action of insulin has not occurred nor has this possibility been extensively investigated. The lack of a testable hypothesis for a mechanism producing intracellular effects of insulin has limited approaches to the question. The cytosolic interaction of insulin with a regulatory protein provides a testable hypothesis. IDE (insulinase, insulin protease, insulysin) was discovered and characterized by its most obvious in vitro property, a high specificity for insulin binding and degradation (35). Its unique properties resulted in controversy about the classification of this enzyme until isolation of the cDNA (36). Analysis revealed that IDE does not belong to any of the classical proteinase groups, but is instead the initial representative of a new superclass of metalloenzymes. In addition to insulin degradation, it has been implicated in growth factor binding and degradation, degradation of other cellular proteins, and in cellular growth and differentiation (37). IDE’s primary cellular role may be as a regulatory factor for cytosolic complexes and as a receptor for hormones which alter activity of these complexes. Other hormones, including IGF-I, IGF-II, TGFa, ANF, and proinsulin bind to IDE with varying affinities and susceptibility to degradation (38). In vitro studies with isolated proteasomes show inhibition of proteasome activity by these ligands (27). Potentially, some effects of these
TABLE 4
Insulin Inhibition of Proteasome Activity Is Reduced after Delivery of Anti-IDE But Not Control Antibody Loading conditions
mg/ml
No loading Vehicle Mouse antibody Anti-IDE antibody
— — 10 5.0
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
91.6 94.4 90.1 107.4
{ { { {
3.8 3.2 5.4 4.8*‡
Note. Introduction of anti-IDE antibody reduces insulin inhibition of proteasome activity. Cells were osmotically loaded with anti-IDE, isotype-matched non-specific mouse IgG, or vehicle alone. The effect of insulin (10011 M) on FLF degradation was determined. Data are mean { S.E.M. for 6 replicate wells. * p õ 0.05 compared with no loading. ‡ p õ 0.05 compared with mouse non-specific antibody.
393
AID
% activity with 10011 M insulin
bbrcg
AP: BBRC
Vol. 244, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
hormones could be mediated by intracellular interaction with IDE. Regardless of potential implications for other actions of insulin or for actions of other hormones, the present data support an effect of insulin on cellular proteolysis requiring intracellular interaction of insulin and IDE with the proteasome. Short term (glucose metabolism) and long term (mitogenesis) actions of insulin have different post receptor signal transduction systems. We propose that the intermediate term effects of insulin on protein and possibly lipid metabolism also have a different signal transduction mechanism which involves post receptor interaction with the cytosolic protein, IDE. In addition, with the demonstration of a regulatory role for IDE we suggest that a new trivial name for this protein is appropriate. We propose the name ‘‘Insulin Regulatory Protein’’ (IRP) to reflect its central role in regulating insulin (and potentially insulin-related peptides) levels and activity. ACKNOWLEDGMENTS This research was supported in part by the Department of Veterans Affairs Research Service and in part by the Bly Memorial Research Fund, University of Nebraska Medical Center. The authors thank Gerri Siford and Kim Harmon for technical assistance and Janet Corr for preparing the manuscript.
REFERENCES 1. Duckworth, W. C., Bennett, R. G., and Hamel, F. G. (1997) J. Invest. Med. 45, 20–27. 2. Mortimore, G. E., and Poso, A. R. (1986) Adv. Enzyme Regul. 25, 257–276. 3. Mortimore, G. E., and Poso, A. R. (1987) Annu. Rev. Nutr. 7, 539–564. 4. Mortimore, G. E., Poso, A. R., and Lardeux, B. R. (1989) Diabetes Metab. Rev. 5, 49–70. 5. Chua, B., Kao, R., Rannels, D. E., and Morgan, H. E. (1978) Biochem. Soc. Symp. 43, 1–15. 6. Long, W. M., Chua, B. H., Munger, B. L., and Morgan, H. E. (1984) Fed. Proc. 43, 1295–1300. 7. Furuno, K., and Goldberg, A. L. (1986) Biochem. J. 237, 859– 864. 8. Gulve, E. A., and Dice, J. F. (1989) Biochem. J. 260, 377–387. 9. Gulve, E. A., Mabuchi, K., and Dice, J. F. (1991) J. Cell. Physiol. 147, 37–45. 10. Mortimore, G. E., and Poso, A. R. (1984) Fed. Proc. 43, 1289– 1294.
11. Mortimore, G. E., Poso, A. R., Kadowaki, M., and Wert, J. J., Jr. (1987) J. Biol. Chem. 262, 16322–16327. 12. Pepato, M. T., Migliorini, R. H., Goldberg, A. L., and Kettelhut, I. C. (1996) Am. J. Physiol. 271, E340–E347. 13. Miller, D. S. (1988) Science 240, 506–509. 14. Duckworth, W. C., Bennett, R. G., and Hamel, F. G. (1994) J. Biol. Chem. 269, 24575–24580. 15. Kupfer, S. R., Wilson, E. M., and French, F. S. (1994) J. Biol. Chem. 269, 20622–20628. 16. Peavy, D. E., Edmondson, J. W., and Duckworth, W. C. (1984) Endocrinology 114, 753–760. 17. Terris, S., and Steiner, D. F. (1975) J. Biol. Chem. 250, 8389– 8398. 18. Tolman, E. L., Schworer, C. M., and Jefferson, L. S. (1973) J. Biol. Chem. 248, 4552–4560. 19. Duckworth, W. C., Heineman, M. A., and Kitabchi, A. E. (1972) Proc. Natl. Acad. Sci. USA 69, 3698–3702. 20. Harding, C. V., France, J., Song, R., Farah, J. M., Chatterjee, S., Iqbal, M., and Siman, R. (1995) J. Immunol. 155, 1767–1775. 21. Okada, C. Y., and Rechsteiner, M. (1982) Cell 29, 33–41. 22. Hamel, F. G., Bennett, R. G., Harmon, K. S., and Duckworth, W. C. (1997) Biochem. Biophys. Res. Commun. 234, 671–674. 23. Rock, K. L., Gramm, C., Rothstein, L., Clark, K., Stein, R., Dick, L., Hwang, D., and Goldberg, A. L. (1994) Cell 78, 761–771. 24. Figueiredo-Pereira, M. E., Berg, K. A., and Wilk, S. (1994) J. Neurochem. 63, 1578–1581. 25. Fenteany, G., Standaert, R. F., Lane, W. S., Choi, S., Corey, E. J., and Schreiber, S. L. (1995) Science 268, 726–731. 26. Barrett, E., Jahn, L., Oliveras, E., and Fryburg, D. (1995) Am. J. Physiol. 268, E820–E824. 27. Bennett, R. G., Hamel, F. G., and Duckworth, W. C. (1997) Diabetes 46, 197–203. 28. Hari, J., Shii, K., and Roth, R. A. (1987) Endocrinology 120, 829– 831. 29. Authier, F., Bergeron, J. J. M., Ou, W.-J., Rachubinski, R. A., Posner, B. I., and Walton, P. A. (1995) Proc. Natl. Acad. Sci. USA 92, 3859–3863. 30. Bennett, R. G., Hamel, F. G., and Duckworth, W. C. (1996) Diabetes 45(Suppl. 2), 310A. 31. Steiner, D. F. (1977) Diabetes 26, 322–331. 32. Miller, D., and Sykes, D. B. (1991) J. Cell. Physiol. 147, 487– 494. 33. Hofmann, C. A., Lotan, R. M., Ku, W. W., and Oeltmann, T. N. (1983) J. Biol. Chem. 258, 11774–11779. 34. Draznin, B., and Trowbridge, M. (1982) J. Biol. Chem. 257, 11988–11993. 35. Mirsky, I. A. (1957) Recent Prog. Horm. Res. 13, 429–471. 36. Affholter, J. A., Fried, V. A., and Roth R. A. (1988) Science 242, 1415–1418. 37. Duckworth, W. C. (1988) Endocr. Rev. 9, 319–345. 38. Authier, F., Posner, B. I., and Bergeron, J. J. M. (1996) Clin. Invest. Med. 19, 149–160.
394
AID
BBRC 8276
/
694c$$$301
02-19-98 16:05:44
bbrcg
AP: BBRC