Insulin degradation XVIII. On the regulation of glutathione-insulin transhydrogenase in the hyperglycemic obese (ob/ob) mouse

The occureence of insulin-degrading activity in the liver of the obese hyperglycemic mouse (ob/ob) and its litter mate has been studied. The trichloroacetic acid-soluble product formed from insulin upon incubation with liver homogenate was identified as the A chain of insulin. In Ouchterlony double-diffusion experiments with antibody to purified rat liver glutathione-insulin transhydrogenase, mouse liver homogenate and the microsomal fraction each gave a single precipitation band of identity with the purified rat liver enzyme. These results indicate that the insulin-degrading activity preseny in the mouse liver is, in fact, glutathione-insulin transhydrogenase. Subcellular distribution studies of glutathione-insulin transhydrogenase and marker enzymes indicate that the transhydrogenase is located primarily in the microsomal fraction of mouse liver homogenate.The ob/ob mouse, which is a genetic mutant characterized by obesity, hyper-insulinism and resistance to the hypoglycemic action of insulin, contains hepatic glutathione-insulin transhydrogenase activity (per mg microsomal protein) markedly higher (40–60%) than its lean litter mates. However, a major portion of the increased hepatic enzyme in the ob/ob mouse occurs in a latent state; the increased amount of enzyme either is unavailable or is nonfunctional, although the ob/ob mouse still contains more of the functional form than the lean mouse. Thus, the results are consistent with the suggestion that the hepatic glutathione-insulin transhydrogenase is probably under a feedback control by circulating insulin.     

Insulin degradation V. Unmasking of glutathione-insulin transhydrogenase in rat liver microsomal membrane

Glutathione-insulin transhydrogenase (glutathione:protein disulfide oxidoreductase, EC 1.8.4.2) inactivates insulin by cleaving its disulfide bonds. The distribution of GSH-insulin transhydrogenase in subcellular fractions of rat liver homogenates has been studied. From the distribution of insulin-degrading activity and marker enzymes (glucose-6-phosphatase and succinate-INT reductase) (INT, 2-p-iodophenyl-3-p-nitrophenyl-5-phenyl tetrazolium chloride) after cell fractionation by differential centrifugation, the immunological analysis of the isolated subcellular fractions with antibody to purified rat liver GSH-insulin transhydrogenase, and chromatographic analysis (on a column of Sephadex G-75 in 50% acetic acid) of the products formed from ¹²⁵I-labelled insulin after incubation with the isolated subcellular fractions, it is concluded that GSH-insulin transhydrogenase is located primarily in the microsomal fraction of rat liver homogenate. An enzyme(s) that further degrades insulin by proteolysis is located mainly in the soluble fraction; a significant amount of the protease(s) activity is also present in the mitochondrial fraction. The possibility has been discussed that the protease(s) acts upon the intermediate product of insulin degradation, A and B chains of insulin, rather than upon the intact insulin molecule itself.The GSH-insulin transhydrogenase in intact microsomes occurs in a latent state; it is readily released from the microsomal membrane and its activity is greatly increased by treatments which affect the lipoprotein membrane structure of microsomal vesicles. There include homogenization with a Polytron homogenizer, sonication, freezing and thawing, alkaline pH, the nonionic detergent Triton X-100, and phospholipases A and C.     

Thiol-protein disulphide oxidoreductases. Assay of microsomal membrane-bound glutathione-insulin transhydrogenase and comparison with protein disulphide-isomerase.

1. Inhibition of endogenous microsomal NADPH oxidase by CO enables membrane-bound glutathione-insulin transhydrogenase (EC 1.8.4.2) to be assayed conveniently by a linked assay involving NADPH and glutathione reductase (EC 1.6.4.2). 2. The specific activity of the enzyme in rat liver microsomal preparations is of the order of 1 nmol of oxidized glutathione formed/min per mg of membrane protein. 3. The specific activity of the enzyme is comparable in rough and smooth microsomal fractions, and the activity is not affected by treatment with EDTA and the removal of ribosomes from rough microsomal fractions. 4. Membrane-bound glutathione-insulin transhydrogenase is not affected by concentrations of deoxycholate up to 0.5%, whereas protein disulphide-isomerase (EC 5.3.4.1) is drastically inhibited. 5. On these grounds it is concluded that, in rat liver microsomal fractions, glutathione-insulin transhydrogenase and protein disulphide-isomerase activities are not both catalysed by a single enzyme species.     

Topology of glutathione-insulin transhydrogenase in rat liver microsomes

Glutathione-insulin transhydrogenase (EC 1.8.4.2) catalyzes the inactivation of insulin through scission of the disulfide bonds to form insulin A and B chains. In the liver, the transhydrogenase occurs primarily in the microsomal fraction where most of the enzyme is present in a latent (‘inactive’) state. We have isolated rat hepatic microsomes with latent transhydrogenase activity being an integral part of the vesicles. We have used these vesicles to study the topological location of glutathione-insulin transhydrogenase by investigating the effects of detergents (Triton X-100 and sodium deoxycholate), phospholipase A₂ and proteinases (trypsin and thermolysin) on the latent enzyme activity. Treatment of intact vesicles with variable concentrations of detergents and phospholipase A₂ resulted in the unmasking of latent transhydrogenase activity. The extent of unmasking of transhydrogenase activity is dependent upon the concentration of detergent or phospholipase used and is accompanied by a parallel release of the enzyme into the soluble fraction. Activation of the transhydrogenase by phospholipase A₂ is partially inhibited by bovine serum albumin and the extent of inhibition is inversely proportional to the phospholipase concentration. In intact vesicles, latent transhydrogenase activity is resistant to proteolytic inactivation by both trypsin and thermolysin, while in semipermeable and permeable vesicles these proteases inactivate 60 and 25% of the total transhydrogenase activity, respectively. Together these results indicate that in microsomes transhydrogenase is probably weakly bound to membrane phospholipid components and that most of the enzyme is present on the cisternal surface (i.e., the luminal surface of endoplasmic reticulum) of microsomes. Each detergent and phospholipase apparently unmasks glutathione-insulin transhydrogenase activity through disruption of the phospholipid-enzyme interaction followed by translocation of the enzyme to the soluble (cytoplasmic) fraction and not through increases in substrate availability.     

Insulin degradation. XXV. Glutathione-insulin transhydrogenase activity of rat liver and kidney during the development of streptozotocin-diabetes.

The activity of glutathione-insulin transhydrogenase (glutathione:protein-disulfide oxidoreductase, EC 1.8.4.2) in the liver and kidneys of rats during the development of streptozotocin-induced diabetes has been studied. Following a single injection of streptozotocin, the transhydrogenase activity fell rapidly for 7-8 days and then gradually with time in both organs. In contrast to the control rats where approximately 25% of the enzyme is in a 'latent' state, nearly all the transhydrogenase activity in the diabetic liver appears to be in the free or functional form. The results are consistent with the hypothesis that both hepatic and renal glutathione-insulin transhydrogenase activity are under feedback control by circulating insulin. The possibility is discussed that the latent state may represent a storage form of the enzyme, which in insulin-insufficiency states is mobilized to the free or functional form for cell function.     

Characterization and application of monoclonal antibodies directed to separate epitopes of glutathione-insulin transhydrogenase

Five monoclonal antibodies specific for glutathione-insulin transhydrogenase were characterized. None of the monoclonal antibodies cross-reacted with another insulin-degrading enzyme, neutral thiopeptidase. The isotype of four antibodies was IgG1 and of the fifth IgG2b. Affinity studies, competitive binding studies and immunoblot analysis of CNBr and trypsin cleavage products of glutathione-insulin transhydrogenase demonstrated that the four IgG1 antibodies were directed to an epitope of the enzyme which was distinct from the epitope recognized by the IgG2b antibody. Inhibition studies indicated that each monoclonal antibody, when added singly to glutathione-insulin transhydrogenase, was unable to inhibit the insulin-degrading activity of the enzyme. However, when monoclonal antibodies directed against separate epitopes of glutathione-insulin transhydrogenase were presented together (i.e., the IgG2b with any one of the four IgG1 antibodies), a loss in enzymatic activity was noted. Immunoblot analysis of rat organ extracts with the IgG1 antibodies demonstrated one immunoreactive protein band of Mr 56,000 in all tissues examined (liver, fat, pancreas and kidney) except the spleen, which demonstrated two immunoreactive protein bands of Mr 56,000 and 51,000. The same immunoblots, when probed with the IgG2b antibody, demonstrated the same immunoreactive protein banding pattern as above plus an additional immunoreactive protein band of Mr 67,000 in all tissues. Studies with spleen extracts from steptozotocin-induced diabetic rats demonstrated that there was a loss of the 51,000 immunoreactive band in diabetes. This 51,000 protein was restored upon insulin treatment of the diabetic rats and nullified upon concomitant administration of cycloheximide or actinomycin D with insulin. Immunoblots of human liver, adipose and skeletal muscle extracts indicated that each monoclonal antibody cross-reacted with the human form of the enzyme which had a molecular weight of Mr 63,000; a second minor immunoreactive band of 67,000 was detected with the IgG2b antibody. The physiological significance of additional molecular forms of the enzyme (i.e., 67,000 and 51,000) remains to be determined.     

Liver monoamine oxidase in the obese-hyperglycaemic (ob/ob) mouse.

1. The specific activity of monoamine oxidase was found to be greater in liver mitochondria from ob/ob mice than from lean mice. The activities of marker enzymes were similar in both tissues. 2. Experiments with various substrates (5-hydroxytryptamine, benzylamine and tyramine) and inhibitors (clorgyline and deprenyl) indicated that, unlike rat liver mitochondria, mouse liver mitochondria contain a predominance of the B-form of monoamine oxidase. 3. The Km values for lean and ob/ob mice were the same for any given substrate and were in the increasing order 5-hydroxytryptamine less than tyramine less than benzylamine. Vmax. was approximately 50% greater in obese than in lean mice. 4. Extraction of liver mitochondria with acetone/water or acetone/water/NH3 to remove lipids decreased the enzyme activity relatively more in obese- than in lean-mice preparations, but residual activity was the same in both preparations.     

Production of a monoclonal antibody directed against rat liver glutathione-insulin transhydrogenase

A hybridoma cell line secreting monoclonal antibody specific for glutathione-insulin transhydrogenase has been produced by fusing mouse myeloma cells with spleen cells from mice immunized to purified rat liver glutathione-insulin transhydrogenase. The secreted antibody isotypes were found to be: Ig gamma 1 heavy chains and kappa light chains. This monoclonal antibody has been used to screen glutathione-insulin transhydrogenase in various rat tissue extracts (liver, fat, heart, testis, spleen, lung and kidney) following separation on NaDodSO4/urea polyacrylamide disc-gel electrophoresis and electrophoretic transfer to nitrocellulose. Screening with the monoclonal antibody showed the presence of one immunoreactive protein band equal in molecular weight to that of purified rat liver GIT (Mr 53,000) in extracts of all tissues studied and a second immunoreactive protein band of lower molecular weight (Mr 49,000) in spleen and lung tissue extracts. Separation of these two proteins by HPLC using a TSK-DEAE column demonstrated that both proteins exhibit insulin degrading activity. These data indicate that GIT may occur in multiple forms in some tissues.     

Characterization of a cDNA for human glutathione-insulin transhydrogenase (protein-disulfide isomerase/oxidoreductase)

A human liver cDNA expression library in lambda-phage gt11 was screened with monoclonal antibodies to rat liver protein-disulfide isomerase/oxidoreductase (EC 5.3.4.1/1.8.4.2), also known as glutathione-insulin transhydrogenase (GIT). The nucleotide sequence of the largest cDNA insert (hgit-1) was determined. It contained approx. 1500 basepairs, representing an estimated 65% of the glutathione-insulin transhydrogenase message. The amino-acid sequence deduced from this cDNA insert contains a 7-amino-acid long polypeptide determined by sequencing the active-site fragment isolated from the rat GIT protein. A comparison of the nucleotide sequence of hgit-1 and a previously reported nucleotide sequence of rat glutathione-insulin transhydrogenase cDNA shows that the human hgit-1 clone corresponds to the middle of the transhydrogenase message at amino-acid residue number 275 of the rat protein, and codes for 206 amino-acid residues, including one of the two active-site regions of glutathione-insulin transhydrogenase, a stop codon (TAA), a long 3'-noncoding region of over 800 bases, a polyadenylation signal (AATAA), and a 29 base poly(A) tail. There exists high homology between the human and rat enzymes (94% in the overall amino-acid sequence, with 100% in the active site region and 81% in the nucleotide sequence within the coding portion of hgit-1). As with the rat enzyme, the human enzyme shows some identity with another dithiol-disulfide-exchange protein, Escherichia coli thioredoxin. Like rat cDNA, the human hgit-1 cDNA hybridized to rat mRNA of 2500 bases on a Northern blot. The relative quantitative abundance of GIT mRNA in nine rat tissues studied using hgit-1 as a hybridization probe was found to be in the same order as previously found with the rat cDNA. Thus, the above studies indicate that glutathione-insulin transhydrogenase is a highly conserved protein and that the human hgit-1 cDNA is suitable for use as a probe for further studies on gene regulation of this enzyme.     

Characterization of a cDNA for human glutathione-insulin transhydrogenase (protein-disulfide isomerase / oxidoreductase)

A human liver cDNA expression library in λ-phage gt11 was screened with monoclonal antibodies to rat liver protein-disulfide isomerase / oxidoreductase (EC 5.3.4.1 / 1.8.4.2), also known as glutathione-insulin transhydrogenase (GIT). The nucleotide sequence of the largest cDNA insert (hgit-1) was determined. It contained approx. 1500 basepairs, representing an estimated 65% of the glutathione-insulin transhydrogenase message. The amino-acid sequence deduced from this cDNA insert contains a 7-amino-acid long polypeptide determined by sequencing the active-site fragment isolated from the rat GIT protein. A comparison of the nucleotide sequence of hgit-1 and a previously reported nucleotide sequence of rat glutathione-insulin transhydrogenase cDNA shows that the human hgit-1 clone corresponds to the middle of the transhydrogenase message at amino-acid residue number 275 of the rat protein, and codes for 206 amino-acid residues, including one of the two active-site regions of glutathione-insulin transhydrogenase, a stop codon (TAA), a long 3′-noncoding region of over 800 bases, a polyadenylation signal (AATAA), and a 29 base poly(A) tail. There exists high homology between the human and rat enzymes (94% in the overall amino-acid sequence, with 100% in the active site region and 81% in the nucleotide sequence within the coding portion of hgit-1). As with the rat enzyme, the human enzyme shows some identity with another dithiol-disulfide-exchange protein, Escherichia coli thioredoxin. Like rat cDNA, the human hgit-1 cDNA hybridized to rat mRNA of 2500 bases on a Northern blot. The relative quantitative abundance of GIT mRNA in nine rat tissues studied using hgit-1 as a hybridization probe was found to be in the same order as previously found with the rat cDNA. Thus, the above studies indicate that glutathione-insulin transhydrogenase is a highly conserved protein and that the human hgit-1 cDNA is suitable for use as a probe for further studies on gene regulation of this enzyme.     

Resolution of protein disulphide-isomerase and glutathione-insulin transhydrogenase activities by covalent chromatography.

1. Protein disulphide-isomerase (EC 5.3.4.1) and glutathione-insulin transhydrogenase (EC 1.8.4.2) were resolved by covalent chromatography. Both activities, in a partially purified preparation from bovine liver, bind covalently as mixed disulphides to activated thiopropyl-Sepharose 6B, in a new stepwise elution procedure protein disulphide-isomerase is displaced in mildly reducing conditions whereas glutathione-insulin transhydrogenase is only displaced by more extreme reducing conditions. 2. This together with evidence for partial resolution of the two activities by ion-exchange chromatography, conclusively establishes that the two activities are not alternative activities of a single bovine liver enzyme. 3. Protein disulphide-isomerase, partially purified by a published procedure, has now been further purified by covalent chromatography and ion-exchange chromatography. The final material is 560-fold purified relative to a bovine liver homogenate; it has barely detectable glutathione-insulin transhydrogenase activity. 4. The purified protein disulphide-isomerase shows a single major band on sodium dodecyl sulphate/polyacrylamide-gel electrophoresis corresponding to a mol.wt. of 57000. 5. The purified protein disulphide-isomerase has Km values for 'scrambled' ribonuclease and dithiothreitol of 23 microgram/ml and 5.4 microM respectively and has a sharp pH optimum at 7.5. The enzyme has a broad thiol-specificity, and several monothiols, at 1mM, can replace dithiothreitol. 6. The purified protein disulphide-isomerase is completely inactivated after incubation with a 2-3 fold molar excess of iodoacetate. The enzyme is also significantly inhibited by low concentrations of Cd2+ ions. These findings strongly suggest the existence of a vicinal dithiol group essential for enzyme activity. 7. When a range of thiols were used as co-substrates for protein disulphide-isomerase activity, the activities were found to co-purify quantitatively, implying the presence of a single protein disulphide-isomerase of broad thiol-specificity. Glutathione-disulphide transhydrogenase activities, assayed with a range of disulphide compounds, did not co-purify quantitatively.     

Insulin-degrading neutral cysteine proteinase activity of adipose tissue and liver of nondiabetic, streptozotocin-diabetic, and insulin-treated diabetic rats

The activity of the insulin-degrading enzyme neutral cysteine proteinase (EC 3.4.22.11, insulinase) was studied in adipose tissue and in liver of nondiabetic, streptozotocin-diabetic, and insulin-treated diabetic rats. Proteinase activity was found to be significantly decreased during diabetes and was restored to near normal levels in both tissues following insulin treatment. The insulin-mediated increase of proteinase activity in both tissues was partially or completely blocked by actinomycin D (an inhibitor of RNA synthesis) and by cyclohexamide (an inhibitor of protein synthesis). Kinetic analysis showed that the changes in proteinase activity of both liver and adipose tissues were accompanied by a change in Vmax (i.e., maximal enzyme activity) without a change in Km (i.e., substrate affinity). These data indicate that insulin functions as an inducer for neutral cysteine proteinase in both tissues. These alterations in the proteinase activity paralleled the alterations in the activity of a second insulin-degrading enzyme, glutathione-insulin transhydrogenase in adipose tissue (this paper) and in liver (previously published papers) under the same physiological conditions.     

Production and characterization of a monoclonal antibody to rat liver thiol: protein-disulfide oxidoreductase/glutathione-insulin transhydrogenase

Rat liver thiol:protein-disulfide oxidoreductase/glutathione-insulin transhydrogenase (glutathione:protein disulfide oxidoreductase, EC 1.8.4.2) was purified and found to give two bands on sodium dodecyl sulfate polyacrylamide gel electrophoresis. A monoclonal antibody was produced against this enzyme preparation and found to remove all the insulin degrading activity of purified preparations of the enzyme. This monoclonal antibody was also found to react with the two different forms of the enzyme observed on gel electrophoresis. These results suggest that glutathione-insulin transhydrogenase can exist in more than one state.     

Effects of thyroid hormone and adrenalectomy on [Na+ + K+]ATPase in the ob/ob mouse

The absence of the thyroid stimulation of hepatic [Na+ + K+]ATPase in obese (ob/ob) mice has been investigated. A wide range of tri-iodothyronine (T3) concentrations failed to enhance the hepatic [Na+ + K+]ATPase of ob/ob mice made hypothyroid by methimazole treatment but glycerophosphate dehydrogenase activity was maximally stimulated at low doses of T3. Adrenalectomy increased the basal activity of [Na+ + K+]ATPase to levels found in lean control mice and restored the response of this enzyme to T3. Body weight gain was unaffected by the induction of hypothyroidism in either lean or ob/ob mice. It is concluded that the adrenal steroids play an important role in the expression of [Na+ + K+]ATPase activity in the ob/ob mouse.     

Lack of vasopressin receptors in liver, but not in kidney, of ob/ob mice.

The activity of phosphorylase a was measured in isolated hepatocytes from fed lean and ob/ob mice after addition of vasopressin, angiotensin, phenylephrine and glucagon. The binding of these hormones to purified liver plasma membranes was also determined. In hepatocytes of ob/ob mice, no increase in phosphorylase a was measured after addition of vasopressin, whereas the other hormones promoted an increase in the activity of the enzyme. No specific vasopressin receptors could be measured on purified liver plasma membrane of ob/ob mice. A decrease in the number of receptors for angiotensin and glucagon, without modification of the affinity, was also observed. No restoration of the number of vasopressin receptors was observed in liver of ob/ob mice starved for 3 days or in younger (5-6 weeks) animals. Vasopressin receptors and vasopressin-stimulated adenylate cyclase, measured on purified kidney medulla membranes, were similar in both lean and ob/ob mice. The data indicate a selective lack of vasopressin receptors and metabolic response in liver of the ob/ob mouse.     

Effects of cold acclimation and fasting on thyroxine 5'-deiodinase in brown adipose tissue of ob/ob mice.

Gradual acclimation to mild cold for 6 weeks increases the total activity of thyroxine 5'-deiodinase in brown adipose tissue (BAT) of genetically obese (ob/ob) mice to a level greater than that in similarly acclimated lean mice. This increase is largely due to the growth of the BAT in the ob/ob mouse, because specific activity of the enzyme is only slightly increased. In similarly cold-acclimated lean mice, the specific activity of thyroxine 5'-deiodinase was not altered. BAT mitochondrial GDP binding increased to the same high level in the gradually cold-acclimated ob/ob mouse as in cold-acclimated lean mice. We conclude that the growth and maintenance of BAT in the cold-acclimated ob/ob mouse, as in the cold-acclimated lean mouse, does not require greatly increased activity of thyroxine 5'-deiodinase. Fasting for 48 hr did not alter thyroxine 5'-deiodinase activity of BAT in either lean or ob/ob mice. The fasting-induced increase in activity seen by others in lean mice is probably due to thermoregulatory stimulation of BAT occasioned by the low environmental temperature at which the fasting occurred.     

Nonlatent and latent hepatic glutathione-insulin transhydrogenase activity during perinatal development and liver regeneration in rats and in rat placenta

We have studied glutathione-insulin transhydrogenase (GIT) activity in differentiating rat liver during parturition and neonatal growth and during compensatory liver growth. Parturition is characterized by a rapid but transient increase in total (i.e., nonlatent plus latent) hepatic GIT activity resulting from changes in the quantity (Vm) of the enzyme while neonatal growth is characterized by an increase in the nonlatent (active) form which persists until just prior to weaning. During liver regeneration following partial hepatectomy, GIT activity/mg protein is lowest after 12 h of regeneration and then progressively increases exceeding the control levels after 72 h of regeneration. Placenta from near-term rats contain a significant concentration of GIT which is immunologically similar to hepatic GIT.     

The role of insulin in the regulation of stearic acid desaturase activity in liver and adipose tissue from obese--hyperglycaemic (ob/ob) and lean mice.

The relationship between the hyperinsulinaemia of obese--hyperglycaemic (ob/ob) mice and their high activity of stearic acid delta 9-desaturase compared with lean mice has been investigated. The concentrations of plasma insulin in obese mice were decreased by 71, 88 and 96% after treatment either with alloxan or food restriction to maintain the same weight as lean mice, or treatment of the weight restricted mice with alloxan followed by feeding ad libitum. The concentration of plasma insulin produced by the latter treatment was the same as in normal lean mice. After treatment the hepatic desaturase activities were 24, 68 and 19% less respectively on a cell basis than in livers from untreated obese mice, and the total epididymal fat-pad activities were lower by 16, 62 and 57%. These results suggest that hyperinsulinaemia is not essential for the increased hepatic desaturase, controlling the hepatic desaturase activity, but even this may be subject to overriding regulation by the concentration of esterified linoleic acid in the liver lipids, which was negatively correlated (r = 0.91, P less than 0.001) with desaturase activity.     

Studies on the subcellular localization and role of glutathione-insulin transhydrogenase in rat liver

₁₂₅I-insulin was shown to be internalized in vivo to a discrete population of low-density membranes (ligandosomes), distinct from the Golgi, endoplasmic reticulum, plasma membrane, and lysosomes. However, analytical subcellular fractionation shows that glutathione-insulin transhydrogenase is localized to the endoplasmic reticulum. Measurement of the specific enzyme activity of glutathione-insulin transhydrogenase showed no differences between normal, diabetic, and hyperinsulinaemic rats. These results suggest that glutathione-insulin transhydrogenase is not directly involved in the subceltular processing of receptor-bound internalized insulin.     

Stabilization of insulin receptor subunit structure by glutathione-insulin transhydrogenase

A partially purified insulin receptor preparation from rat liver was incubated at 37 degrees C with and without the protein-disulfide interchange enzyme, glutathione-insulin transhydrogenase (thiol: protein-disulfide oxidoreductase/isomerase, EC 1.8.4.2/5.3.4.1). Insulin-binding activity was then assessed by crosslinking receptor-125I-insulin complexes and subjecting them to electrophoresis on SDS-polyacrylamide gels in the absence and presence of reductant followed by autoradiography. Prior incubation of the receptor at 37 degrees C in the absence of the enzyme markedly decreased the subsequent binding of 125I-insulin to the holoreceptor (Mr 350 000) and to its subunits (Mr 180 000 and 130 000), while addition of the enzyme to the preincubation medium served to substantially prevent this decrease. The loss in binding at 37 degrees C was not restored by subsequent addition of the enzyme, nor was the loss prevented by any of the several known inhibitors of proteolysis. The apparent stabilization of receptor by transhydrogenase, as evidenced by the increase in binding above control levels, was proportional to both the enzyme concentration and the duration of incubation. These effects seem to be specific for transhydrogenase, since several other disulfide-containing proteins were found to be ineffective. These data suggest that the stabilization of the subunit structure of the insulin receptor at physiological temperatures may take place via a disulfide interchange reaction catalyzed by glutathione-insulin transhydrogenase.