Biochemical and morphological evidence that the insulin receptor is internalized with insulin in hepatocytes.

There is morphological and biochemical evidence that insulin is internalized in hepatocytes. The present study was designed to investigate the fate of the insulin receptor itself, subsequently to the initial binding step of the hormone to the hepatocyte plasma membrane. The insulin receptor was labeled with a 125I-photoreactive insulin analogue (B2[2-nitro,4-azidophenylacetyl]des-PheB1-insulin). This photoprobe was covalently coupled to the receptor by UV irradiation of hepatocytes after an initial binding step of 2-4 h at 15 degrees C. At this temperature, only limited (approximately 20%) internalization of the ligand occurred. In a second step, hepatocytes were resuspended in insulin-free buffer and further incubated for 2-4 h at 37 degrees C. After h at 37 degrees C, no significant radioactivity could be detected in non-UV-irradiated cells, whereas 12-15 % of the radioactivity initially bound remained associated to UV-irradiated cells. Morphological analysis after electron microscopy revealed that approximately 70% of this radioactivity was internalized and preferentially associated with lysosomal structures. SDS PAGE analysis under reducing conditions revealed that most of the radioactivity was associated with a 130,000-dalton band, previously identified as the major subunit of the insulin receptor in a variety of tissues. Internalization of the labeled insulin-receptor complex at the end of the 37 degrees C incubation was further demonstrated by its inaccessibility to trypsin. Conversely, at the end of the association step, the receptor (also characterized as a predominant 130,000-dalton species) was localized on the cell surface since it was cleaved by trypsin. We conclude that in hepatocytes the insulin receptor is internalized with insulin.

Insulin receptors in isolated human adipocytes. Characterization by photoaffinity labeling and evidence for internalization and cellular processing.

We photolabeled and characterized insulin receptors in isolated adipocytes from normal human subjects and then studied the cellular fate of the labeled insulin-receptor complexes at physiologic temperatures. The biologically active photosensitive insulin derivative, B2(2-nitro-4-azidophenylacetyl)des-PheB1-insulin (NAPA-DP-insulin) was used to photoaffinity label the insulin receptors, and the specifically labeled cellular proteins were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and autoradiography. At saturating concentrations, the binding of 125I-NAPA-DP-insulin to the isolated adipocytes at 16 degrees C was rapid (half-maximal in approximately 1 min and maximal in approximately 10 min) and approximately 25% of the specifically bound ligand was covalently linked to the cells by a 3-min exposure to long-wave (366 nm) ultraviolet light. Analysis of the photolabeled cellular proteins by PAGE in the absence of disulfide reductants revealed the specific labeling of a major protein band of Mr 330,000 and two less intense bands of Mr 295,000 and 260,000. Upon reduction of disulfide bonds with dithiothreitol, all three unreduced forms of the insulin receptor were converted into a major labeled Mr-125,000 band and a less intensely labeled Mr-90,000 band. The labeling of the Mr-125,000 receptor subunit was saturable and native porcine insulin effectively inhibited (half-maximal inhibition at 12 ng/ml) the photolabeling of this binding subunit by NAPA-DP insulin. When intact adipocytes photolabeled at 16 degrees C (a temperature that inhibits endocytosis) were immediately trypsinized, all of the labeled receptor bands were converted into small molecular weight tryptic fragments, indicating that at 16 degrees C all of the labeled insulin-receptor complexes remained on the cell surface. However, when the photolabeled cells were further incubated at 37 degrees C and then trypsinized, a proportion of the labeled receptors became trypsin insensitive, indicating that this fraction has been translocated to the cell interior and thus was inaccessible to the trypsin in the incubation medium. The intracellular translocation of the labeled receptors was observed within 2 min, became half-maximal by 10 min, and maximal by approximately 30 min of incubation at 37 degrees C. Cellular processing of the internalized insulin-receptor complexes also occurred, since incubation at 37 degrees C (but not 16 degrees C) resulted in the generation of a Mr-115,000 component from the labeled receptors. Inclusion of chloroquine, a drug with lysosomotropic properties, in the incubation media caused a time-dependent increase (maximal increase of 50% above control by 2 h at 37 degrees C) in the intracellular pool of labeled receptors. In contrast to these findings in human adipocytes, no appreciable internalization of insulin-receptor complexes and no chloroquine effect was observed in cultures human IM-9 lymphocytes during a 1-h incubation at 37 degrees C. We concluded that in isolated human adipocytes: (a) the subunit structure of insulin receptors is the same as that reported for several other tissues, (b) insulin-receptor complexes are rapidly internalized and processed at physiologic temperatures, and (c) the cellular processing of insulin-receptor complexes occurs at one or more chloroquine-sensitive intracellular site(s).

Biological potency of covalently linked insulin-receptor in rat adipocytes. Comparison with the potency of reversible complexes.

Irradiation of photoreactive insulin derivatives in the presence of isolated rat adipocytes produces a prolonged stimulation of lipogenesis in the cells even after exogenous and reversibly bound derivative has been removed by extensive washing. The quantitative nature of this response has now been studied using 125I-B2(2-nitro-4-azidophenylacetyl)des-PheB1-insulin. This derivative possesses nearly full biological potency and binding affinity prior to irradiation. After covalent linkage to adipocytes the efficacy of the derivative is reduced to 25 +/- 4% of the reversibly bound derivative, viz. 4-times as much needs to be covalently associated as reversibly bound to induce the same level of stimulation of lipogenesis. This reduced relative molar potency is due to a reduced ability of specific covalent insulin-receptor complexes to trigger a response.

Biological activity of covalently-linked insulin-receptor complexes in rat adipocytes. Effect of pH.

The relative molar potencies of covalently and reversibly-bound insulin-receptor complexes were studied as a function of pH. The insulin derivatives used were 125I-B2 (2-nitro-4-azidophenylacetyl)des-PheB1-insulin and 125I-B29(2-nitro-4-azidophenylacetyl)des-PheB1-insulin. The potencies of both types of reversible complexes were effectively identical and constant between pH 7 and 8. The relative potency of the covalent B2-complex increased from 25 to 75%, and of the covalent B29 complex from 30 to nearly 100%. This indicates that the covalently linked partners in the complex are able to flex about the cross-linkages. Variations in the potency are due to variations in the number of correctly associated, reversibly or covalently bound insulin-receptor complexes. The form of the pH dependance suggests that an ionizable group, possibly an amino group, must be deprotonated to allow effective interaction.

Evidence for an early degradative event to the insulin molecule following binding to hepatocyte receptors.

We have used photoreactive insulin analogues to investigate as related processes, early structural modification of the receptor-bound insulin molecule and internalisation of the insulin-receptor complex. In isolated rat hepatocytes an initial modification of bound insulin leads to the generation of a molecular species unchanged in molecular weight but with reduced receptor and antibody binding affinities and altered electrophoretic mobility. Using photoreactive insulin analogues and density gradient cell fractionation the insulin receptor complex has been shown to undergo internalisation from the plasma membrane to a low density vesicular fraction, the endosome. No labelled material was found in lysosomal fractions after up to 10 min incubation at 37 degrees C. The degree of labelling of the endosome fraction depended on the position of the photoreactive group within the insulin molecule. The data suggest that before or during endocytosis, a small peptide is proteolytically cleaved from the C terminus of the insulin B chain.

Demonstration that the insulin receptor undergoes an early structural modification following insulin binding.

Processing of the insulin receptor by hepatocytes was studied using a 125I-labelled photoreactive insulin derivative which could be covalently attached to the receptor and facilitate the analysis of receptor structure in isolated subcellular fractions by SDS-polyacrylamide gel electrophoresis. Following binding at the cell surface, the label was rapidly internalised and located in a low-density subcellular fraction ('endosomes'). The intact receptor (350 000 molecular weight) and binding (alpha) subunit (135 000), produced by in vitro disulphide reduction of the samples, were found in the plasma membrane fraction but not in endosomes. In endosomes, the label was concentrated in a band at 140 000 (non-reduced) which on reduction generated species of 100 000 and 68 000 predominantly. The insulin receptor therefore undergoes an early structural change during endocytosis. This modification does not involve complete disulphide reduction and may be due to a proteolytic event.

Degradation of insulin receptors in rat adipocytes.

Insulin receptors on viable rat adipocytes were affinity-labeled using a biologically active and photosensitive analogue of insulin, 125I-B2(2-nitro, 4 azidophenylacetyl)-des-PheB1-insulin (125I-NAPA-DP-insulin). The radiolabeled proteins were identified by SDS polyacrylamide gel electrophoresis and autoradiography. Binding of 125I-NAPA-DP-insulin (40 ng/ml) to rat adipocytes at 16 degrees C, followed by photolysis, resulted in the specific labeling of essentially one protein with an apparent molecular weight of 430-450,000 daltons. When this radiolabeled protein was treated with dithiothreitol prior to electrophoresis, specific labeling occurred predominantly in a 125,000-dalton protein and to a lesser extent in a 90,000-dalton protein. In addition, there was a minimal amount of specific labeling of a 115,000-dalton protein. Under certain experimental conditions, the nonreduced form of the photoaffinity-labeled receptor appeared as a heterogeneous population of proteins having apparent molecular weights of 430,000, 350,000, and 270,000 daltons. Subsequent to photoaffinity labeling of insulin receptors at 16 degrees C, adipocytes were incubated at 37 degrees C for various periods of time to allow for internalization. This resulted in an initial rapid loss of radioactivity in the 430,000- and 125,000-dalton bands. At 60 min the amount of radioactivity in each of these bands was approximately 50% of that present before incubation at 37 degrees C and stayed constant for 120 min. A first-order plot of the decline in receptor-associated radioactivity was biphasic with the initial phase having a half-life of 1.4 h.(ABSTRACT TRUNCATED AT 250 WORDS)

Brown adipose tissue in lean and obese mice. Insulin-receptor binding and tyrosine kinase activity.

Insulin-receptor binding and tyrosine kinase activity have been studied in brown adipose tissue from lean and obese mice. Brown adipose tissue carries functional insulin receptors comparable with those of conventional insulin target tissues. The alpha-subunit (Mr, 130,000) was labeled with photoreactive insulin; the beta-subunit (Mr, 95,000) was phosphorylated in a cell-free system, and its level of phosphorylation was increased in a dose-dependent manner by insulin. Two types of obese mice, mice rendered obese by gold thioglucose injection (GTG obese) and genetically obese ob/ob mice, were used. Insulin-receptor number was decreased by 60-70% in obese mice, when expressed per milligram of plasma membrane protein or per microgram of glycoprotein, whereas only a 30-40% diminution was observed in skeletal muscle, indicating that insulin receptors from brown adipose tissue are greatly affected by the downregulation process. Insulin-stimulated autophosphorylation of the insulin-receptor beta-subunit was decreased by 60-70% in preparations of obese mice compared with lean mice in direct proportion to the diminished level of insulin-receptor number. Similarly, the ability of receptors to catalyze the phosphorylation of a synthetic substrate (copolymer glutamate-tyrosine) was reduced. These results suggest that the decrease in insulin-receptor number and in associated tyrosine kinase activity could explain the insulin-resistant glucose uptake and the alteration in diet-induced thermogenesis described in obese animals.

Binding of antigen to Ia molecules on intact antigen presenting cells demonstrated by photoaffinity labeling.

We used a photoaffinity labeling technique to investigate whether a molecular interaction occurs between antigen and Ia molecules on antigen presenting cells (APC) in the absence of T lymphocytes. M.12.4.1 B lymphoma cells (Iad), which are able to present bovine insulin to Iad lymph node primed T cells, were given radioiodinated bovine insulin derivatized with the photoreactive group (2-nitro-4-azidophenylacetyl) at Lys 29 of the B chain of the insulin molecule. Processing of insulin was allowed by incubating the APC with antigen for increasing periods of time at 37 degrees C or 4 degrees C. The covalent coupling of the processed photoreactive antigen to any neighboring cellular protein was thereafter induced by u.v. irradiation. Immunoprecipitation of membrane proteins by monoclonal antibodies showed that under these conditions, the alpha and beta subunits of the Ia molecules were selectively photolabeled. Labeling was time- and temp-dependent as was the internalization of insulin. The apparent mol. wts of the antigen-Ia molecule complexes were not significantly different from that of native Ia molecules radioiodinated by surface labeling, indicating that only a small fragment of the antigen was covalently coupled to Ia molecules. Similar experiments performed with human B lymphoma cells (526 cells) gave similar results. These observations therefore indicate: (1) that Ia molecules expressed by intact APC are able to bind antigens in the absence of T lymphocyte antigen receptor; and (2) that this association, at least for insulin, requires uptake and a proteolytic fragmentation of the antigen by the APC.

Novel hepatoselective insulin analog: studies with a covalently linked thyroxyl-insulin complex in humans.

OBJECTIVE: To test whether a thyroxyl-insulin analog with restricted access to receptor sites in peripheral tissues displays relative hepatoselectivity in humans. RESEARCH DESIGN AND METHODS: Five normal human subjects received a subcutaneous bolus injection of either N(alphaBl) L-thyroxyl-insulin (Bl-T4-Ins) or NPH insulin in random order. Insulin kinetics, relative effects on hepatic glucose production, and peripheral glucose uptake were studied using euglycemic clamp and stable isotope [D-6,6-(2)H2]glucose) dilution techniques. Blood samples were taken for the determination of total immunoreactive insulin/analog concentrations and for liquid chromatography to assess the protein binding of the analog in the circulation. RESULTS: After subcutaneous administration, Bl-T4-Ins was well tolerated and rapidly absorbed. The analog had a long serum half-life and was highly protein bound (approximately 86%). Its duration of action, as judged by the duration of infusion of exogenous glucose to maintain euglycemia, was similar to that of NPH insulin. The effect of the analogs on hepatic glucose production was similar to that of NPH insulin, indicating equivalent hepatic potency. The analog demonstrated less effect on peripheral glucose uptake than NPH insulin (P = 0.025), had no effect on metabolic clearance rate of glucose, and exhibited a reduced capacity to inhibit lipolysis (P < 0.05). CONCLUSIONS: When injected subcutaneously into normal human subjects, Bl-T4-Ins is well tolerated, quickly absorbed, and highly protein bound, resulting in a long plasma halflife. This analog appears to have a hepatoselective action, and, therefore, has the potential to provide more physiological insulin action than the insulin preparations currently used.

The solution structure of a superpotent B-chain-shortened single-replacement insulin analogue.

This paper reports on an insulin analogue with 12.5-fold receptor affinity, the highest increase observed for a single replacement, and on its solution structure, determined by NMR spectroscopy. The analogue is [D-AlaB26]des-(B27-B30)-tetrapeptide-insulin-B26-amide. C-terminal truncation of the B-chain by four (or five) residues is known not to affect the functional properties of insulin, provided the new carboxylate charge is neutralized. As opposed to the dramatic increase in receptor affinity caused by the substitution of D-Ala for the wild-type residue TyrB26 in the truncated molecule, this very substitution reduces it to only 18% of that of the wild-type hormone when the B-chain is present in full length. The insulin molecule in solution is visualized as an ensemble of conformers interrelated by a dynamic equilibrium. The question is whether the "active" conformation of the hormone, sought after in innumerable structure/function studies, is or is not included in the accessible conformational space, so that it could be adopted also in the absence of the receptor. If there were any chance for the active conformation, or at least a predisposed state to be populated to a detectable extent, this chance should be best in the case of a superpotent analogue. This was the motivation for the determination of the three-dimensional structure of [D-AlaB26]des-(B27-B30)-tetrapeptide-insulin-B26-amide. However, neither the NMR data nor CD spectroscopic comparison of a number of related analogues provided a clue concerning structural features predisposing insulin to high receptor affinity. After the present study it seems more likely than before that insulin will adopt its active conformation only when exposed to the force field of the receptor surface.

Oligosaccharide heterogeneity of insulin receptors. Comparison of N-linked glycosylation of insulin receptors in adipocytes and brain.

We tested the hypothesis that the molecular weight discrepancy between insulin receptors in brain and adipocytes is due to differences in glycosylation by treating photoaffinity-labeled insulin receptors from both tissues with endo-beta-N-acetylglucosaminidase F (Endo F) and analyzing the products by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Endo F removed glycans from the adipocyte 125-kilodalton (kDa) subunit and the brain 115-kDa subunit in a manner dependent upon the concentration of enzyme and time of incubation. At a maximally effective concentration of Endo F, the adipocyte alpha-subunit was reduced from 125-kDa to 100-kDa and the brain alpha-subunit from 115-kDa to 100-kDa. We also examined the type of oligosaccharides present in both alpha-subunits by treating the proteins with endo-beta-N-acetylglucosaminidase H (Endo H), which selectively removes high mannose residues, and neuraminidase. Endo H treatment reduced the apparent molecular weight of both the adipocyte and brain alpha-subunits. In both receptors, the deglycosylated product obtained with Endo H was larger than that generated by Endo F. The adipocyte alpha-subunit demonstrated a shift in mobility on sodium dodecyl sulfate gels after neuraminidase treatment, whereas the brain alpha-subunit did not. We conclude from these studies that 1) The discrepancy in apparent molecular weight of alpha-subunits in brain and adipocytes is due to differences in N-linked glycosylation; 2) high mannose and complex type oligosaccharides are present in both receptor types; and 3) the complex oligosaccharides in the adipocyte alpha-subunit are terminated in a manner different from the complex glycans of the brain alpha-subunit.

Effects of covalently linked insulin dimers on receptor kinase activity and receptor down regulation.

Certain covalently linked insulin dimers have previously been found to have a greater ability to bind to the insulin receptor than to stimulate lipogenesis in adipocytes. The present report presents data indicating that the same insulin dimers also have a greater ability to bind to the receptor than to stimulate the kinase activity of the insulin receptor. In particular, one such covalently linked insulin dimer had less than 1% the potency of native insulin in stimulating the receptor kinase although it could bind to the solubilized receptor with 30% the potency of native insulin. In contrast, this dimer could down regulate the insulin receptor with approximately 30% the potency of native insulin. These results suggest that stimulation of the receptor kinase may require more than simple occupancy of the receptor binding site whereas down regulation of the receptor may require only the binding of ligand to the receptor.

The subunit structure of the insulin receptor and molecular interactions with major histocompatibility complex antigens.

Insulin receptors were labeled with 125I-photoreactive insulin (specifically labeling alpha-subunits) and by insulin-stimulated autophosphorylation (specifically labeling beta-subunits). The results show that the insulin receptor exists under different free and disulfide-linked combinations of alpha and beta subunits. Moreover, the insulin receptor is closely associated to class I antigens of the major histocompatibility complex to form a high molecular weight multi-molecular membrane complex.

Biological action and fate of photoaffinity-labelled insulin-receptor complexes.

Covalent linking of two photoactivatable insulin derivatives, B2-(2-nitro,4-azidophenylacetyl)-des-PheB1-insulin and B29-(2-nitro,4-azidophenylacetyl)-insulin to viable rat adipocytes gives a system, which contains a fixed stoichiometry between hormone and receptor. The biological signal of prolonged lipogenesis has been used to study several aspects of insulin binding and action: the role of the site of the crosslink between insulin and receptor, recognition of bound photoinsulin by anti-insulin antibodies, the half-life of the biologically active complex, the pH-dependence of the biological signal, and the possible role of internalization. Furthermore, the effect of trypsin on the insulin receptor, as well as the insulin-receptor complex, has been investigated and a refined model of the receptor is presented.

Quantitative dissociation of glucose transport stimulation and insulin receptor tyrosine kinase activation in isolated adipocytes with a covalent insulin dimer (B29,B29'-suberoyl-insulin).

The covalent insulin dimer B29,B29'-suberoyl-insulin was investigated for its effects on insulin receptor binding, insulin receptor tyrosine kinase activity and glucose transport in isolated adipose cells. The dimer stimulated glucose transport (initial 3-O-methylglucose uptake rate) to the same extent as insulin did (basal rate, 35 +/- 3 pmol/sec/microliter lipid; insulin, 380 +/- 27; B29,B29'-suberoyl-insulin, 369 +/- 24, means +/- S.E.), although at higher concentrations (EC50 1.94 +/- 0.64 nM versus 0.1 +/- 0.02 with insulin). In contrast, the dimer only partially (23%) mimicked insulin's effect on phosphate incorporation into insulin receptors immunoprecipitated after equilibration of cells with [32P]phosphate. Similarly, insulin receptor tyrosine kinase as assessed by receptor autophosphorylation and phosphorylation of the substrate poly-(Glu/Tyr) was not fully activated by treatment of cells with the insulin dimer (31 and 42% of the effect of insulin, respectively) in concentrations which maximally activate glucose transport and give rise to full insulin receptor occupancy (5 X 10(-7) M). Further, the dimer activated the receptor tyrosine kinase in solubilized purified insulin receptor preparations from adipose cells to only 25% of the effect of insulin (EC50 32.0 +/- 16 versus 1.9 +/- 1.0 nM with insulin) in spite of full receptor occupancy. Binding of the dimer to insulin receptors followed single site binding kinetics, indicating that the derivative is unable to induce negative cooperativity of the insulin receptor. It is concluded that a partial phosphorylation of insulin receptors and a submaximal tyrosine kinase activation are sufficient for full stimulation of glucose transport in the adipocyte. Further, it is suggested that negative cooperativity of the insulin receptor and activation of its tyrosine kinase require a similar conformational change of the receptor protein.

Peptide mapping on Northern blot analyses of insulin receptors in brain and adipocytes.

Previous studies have demonstrated differences in the size of insulin receptor subunits in brain and adipocytes that appear to involve variations in glycosylation of the proteins. In this report, we examined the degree of homology in the protein backbones of insulin receptors in both tissues by peptide mapping and compared the mRNAs encoding the receptors by Northern blot analysis. Photoaffinity-labeled insulin receptors from rat brain and adipocytes were deglycosylated and then subjected to partial proteolysis by five different enzymes with differing substrate specificities. The intact receptors and their proteolytic fragments were analyzed by electrophoresis and autoradiography. Each enzyme yielded a unique pattern of fragments ranging from 70 to 11 kDa. In all cases, there was a striking similarity in the peptide maps generated from insulin receptors in brain and adipocytes. Northern hybridization experiments were carried out using poly(A)+ RNA from rat brain, rat adipocytes, and human hepatocarcinoma (HEP G2) cells. In rat brain, two bands of 9.5 and 7.4 kb were detected and, in rat adipocytes, the same two bands were observed. The two mRNA bands observed in rat tissues represented only two of the five mRNA species seen in human HEP G2 cells. The results indicate that the protein domains and the mRNAs encoding of insulin receptors in brain and adipocytes are very similar, if not identical.

Structure-activity relationship of covalently dimerized insulin derivatives: correlation of partial agonist efficacy with cross-linkage at lysine B29.

The effects of 7 covalently dimerized insulin derivatives on glucose transport in differentiated 3T3-L1 cells were investigated. Symmetric cross-linkage at lysine B29 with a bridge of 2 (oxalyl), 8 (suberoyl) or 12 (dodecanedioyl) carbon atoms produced derivatives with essentially unaltered receptor binding affinity but largely reduced intrinsic activity. Regardless of the chain length, these derivatives inhibited the effect of submaximal insulin concentrations. Insulin derivatives cross-linked at phenylalanine B1 or asymmetrically at B1/B29 were full agonists of the insulin receptor. When lysine B29 was cross-linked with the inactive desoctapeptide(B23-B30)insulin at phenylalanine B1, the intrinsic activity of the resulting dimer was lower than that of insulin, but higher than that of the symmetric B29-dimers. It is concluded that linkage at the B29-lysines, and not at the B1-phenylalanine, leads to partial agonism of dimerized insulin derivatives, regardless of the length of the crosslinker.

In vivo metabolic activity of des-(B26-B30)-insulin-B25-amide and related analogues in the rat.

Metabolic potency of des-(B26-B30)-insulin-B25-amide, [TyrB25]des- (B26-B30)-insulin-B25-amide and [HisB25]des-(B26-B30)-insulin-B25-amide was studied in anaesthetized rats. Compared to insulin, full potency for des-(B26-B30)-insulin-B25-amide and an enhanced potency for both substituted analogues has been described previously on rat adipocytes in vitro. Hypoglycaemic effects following i.v. injection of all of these analogues were almost identical to those of native insulin with a half-maximal effective dose of approximately 3 nmol.kg-1. Stimulation of glucose metabolism during euglycaemic hyperinsulin-/analogueaemic clamp studies was indistinguishable from that of the native hormone with a maximal stimulation of approximately 19 mg.kg-1.min-1 and half-maximal effective hormone concentrations of approximately 1 pmol.ml-1. Analogue action on individual peripheral tissues estimated by the uptake of 2-deoxyglucose as well as stimulation of lipogenesis in epididymal fat was not different to that of insulin. These data demonstrate that C-terminal amidation of des-(B26-B30)-insulin results in a shortened molecule with full in vivo metabolic potency. When substituting phenylalanine in position B25 by tyrosine or histidine, the insulin-identical potency is preserved.

Crosslinked insulins: preparation, properties, and application.

Crosslinked insulins have proved to be valuable for structure-function studies and as proinsulin models. In the first part of the paper, a short review of the literature on analytical investigations, the preparation of A1-B1- and A1-B29-crosslinked derivatives, their biological activities in vivo and in vitro, and CD-spectral properties is given. The results of reduction/reoxidation studies with insulin derivatives containing irreversible and cleavable crosslinks are summarized. In the second part, new A1-B29-crosslinked monomers and 3 symmetrical dimers, linked between A1-A'1, B1-B'1 and B29-B'29, are described, as well as some results of tritium-labelling and of enzymatic degradation experiments with A1-B29-linked insulins.