Search

Here we describe the design and application of OSu-FMS-MAL-S-(CH(2))(15)-COOH, an agent that associates with albumin while linked to a peptide or a protein with sufficient affinity (Ka=2 to 2.6 x 10(5)M(-1)) to protract the action of short- lived peptides and proteins in vivo. Under physiological conditions this probe undergoes spontaneous hydrolysis with the concomitant reactivation of inactive conjugates. Intravenously administered (125)I-labeled-Insulin-FMS-MAL-S-(CH(2))(15)-COOH to rats shows half-life of 17+/-2h, exceeding 5.2 times that obtained with intravenously administered (125)I-labeled Insulin. In mice this derivative facilitates glucose-lowering effect over a period of 24h, yielding AUC five times greater than that obtained by a similar dose of insulin-detemir. Similarly, subcutaneous administration of Exendin-4-FMS-MAL-S-(CH(2))(15)-COOH into mice facilitated prolonged and stable reduction in glucose level, yielding a t(1/2) value of 28+/-2h, exceeding the effect of exendin-4 4.7 folds. The inactive derivative gentamicin-FMS-MAL-S-(CH(2))(15)-COOH regained its full antibacterial potency upon incubation at physiological conditions yielding a t(1/2) value of 7.1+/-0.2h. In conclusion, the albumin-binding probe we introduced enables to prolong the action of any amino containing molecule in vivo, without the drawback of inactivation that often occurs upon such derivatization., (Copyright 2009 Elsevier B.V. All rights reserved.)

We attempted to engineer a novel long-acting insulin based on the following properties: (i) action as a prodrug to preclude supraphysiological concentrations shortly after injection; (ii) maintenance of low-circulating level of biologically active insulin for prolonged period; and (iii) high solubility in aqueous solution. A spontaneously hydrolyzable prodrug was thus designed and prepared by conjugating insulin through its amino side chains to a 40kDa polyethylene glycol containing sulfhydryl moiety (PEG(40)-SH), employing recently developed hetero-bifunctional spacer 9-hydroxymethyl-7(amino-3-maleimidopropionate)-fluorene-N-hydroxysucinimide (MAL-Fmoc-0Su). A conjugate trapped in the circulatory system and capable of releasing insulin by spontaneous chemical hydrolysis has been created. PEG(40)-Fmoc-insulin is a water-soluble, reactivatable prodrug with low biological activity. Upon incubation at physiological conditions, the covalently linked insulin undergoes spontaneous hydrolysis at a slow rate and in a linear fashion, releasing the nonmodified immunologically and biologically active insulin with a t(1/2) value of 30h. A single subcutaneous administration of PEG(40)-Fmoc-insulin to healthy and diabetic rodents facilitates prolonged glucose-lowering effects 4- to 7-fold greater than similar doses of the native hormone. The beneficial pharmacological features endowed by PEGylation are thus preserved. In contrast, nonreversible, "conventional" pegylation of insulin led to inactivation of the hormone.

The covalent linkage of peptides or protein drugs to human serum albumin (HSA) greatly prolongs their lifetime in vivo but is pharmacologically irrelevant when it irreversibly inactivates them. We retain drug bioactivity by synthesizing a heterobifunctional reagent (MAL-Fmoc-OSu, 9-hydroxymethyl-2-(amino-3-maleimidopropionate)-fluorene-N-hydroxysuccinimide) that generates HSA-Fmoc-insulin on covalent conjugation to the amino group of insulin and the Cys-34 side chain of HSA. HSA-Fmoc-insulin is water-soluble and, upon incubation in aqueous buffers reflecting normal human serum conditions, slowly, spontaneously, and homogeneously hydrolyzes to release unmodified insulin with a t 1/2 of 25 +/- 2 h. A single subcutaneous or intraperitoneal administration of HSA-Fmoc-insulin to diabetic rodents lowers circulating glucose levels for about 4 times longer than an equipotent dose of Zn2+-free insulin. Following subcutaneous administration, onset of the glucose-lowering effect is delayed 0.5-1 h and persists for 12 h. Thus, we present a prototype insulin formulation possessing three desirable parameters: high aqueous solubility, delayed action following subcutaneous administration, and prolonged therapeutic effect.

FMS(3)-insulin (2-sulfo-9-fluorenylmethoxycarbonyl)(3)-insulin is a water soluble, inactive-reactivated derivative of insulin with protracted action in vivo. In this study we find that FMS(3)-insulin preserves insulin's capacity to crystallize when associated with Zn(2+) ions or with basic protamine. Zinc or protamine suspended preparations of FMS(3)-insulin manifest substantially prolonged, blood glucose-lowering pharmacokinetic profiles in STZ-treated rats (STZ-rats). A dose of up to 1mg suspended FMS(3)-insulin/STZ rat can be subcutaneously administered with no hypoglycemic episodes at any time point after administration. This dose yielded glucose-lowering profiles with t(1/2) values at the range of 50-70h, turning catabolic STZ-rats into anabolic ones over a period of 2-3 days. The obtained glucose-lowering patterns exceeded 7-8 times in duration those produced by nonhypoglycemic doses of NPH-insulin. In summary, subcutaneous administration of suspended insulin prodrugs, such as FMS(3)-insulin, can bring about prolonged, nonhypoglycemic glucose-lowering profiles, unattainable with insulin preparations, which are known to be active at the time of administration.

In this study we designed, prepared, and analyzed a water-soluble, long-acting insulin derivative whose protracted action in vivo is based on a new principle rather than on slower absorption rates of suspended insulin formulations. To this end, we have prepared (9-fluorenylmethoxycarbonyl-SO(3)H)(3)-insulin ((FMS)(3)-insulin), a derivative having three 9-fluorenylmethoxycarbonyl-SO(3)H (FMS) moieties covalently linked to the three amino side chains of insulin. (FMS)(3)-insulin is soluble in aqueous buffers at neutral pH, at a concentration range of 0.15-0.60 mM, and has about 1% of both the biological potency and the receptor-binding affinity of the native hormone. Upon incubation at pH 7.4 and 37 degrees C, it undergoes a slow hydrolysis with linear regeneration of insulin possessing full biological potency. A single subcutaneous administration of (FMS)(3)-insulin to streptozocin-treated rats lowered circulating glucose levels for a prolonged period (t(1/2) = 30 h). Similarly, intraperitoneal administration of (FMS)(3)-insulin to healthy rats had a prolonged glucose-lowering effect. In this experimental system, recovery from hypoglycemia proceeded with a t(1/2) value of 14 +/- 1 h, as compared with t(1/2) = 8.0 +/- 1 h for native insulin and t(1/2) = 10.0 +/- 1 h for NPH-insulin. (FMS)(3)-insulin is more resistant to proteolysis and appears to be nonimmunogenic. On the whole, (FMS)(3)(-)insulin represents a prototype version of a water-soluble, long-acting preparation of insulin. It is nearly inactive at the time of administration, and therefore can be administered, at high dose, with no concern for hypoglycemia. Because of its decreased receptor-binding affinity, (FMS)(3)-insulin evades receptor-mediated endocytosis and degradation and, hence, persists for a long period in the circulation. The insulin constituent of the (FMS)(3)-insulin conjugate undergoes a slow, spontaneous activation in the circulatory system, manifesting a prolonged glucose-lowering action in vivo. According to the data presented here, (FMS)(3)-insulin represents a typical prodrug: a compound which by itself shows only marginal activity but over time it is chemically hydrolyzed to the fully active hormone.

Most mammalian cells contain vanadium at a concentration of about 20 nM, the bulk of which is probably in the reduced vanadyl (+4) form. Although this trace element is essential and should be present in the diet in minute quantities, no known physiological role for vanadium has been found thus far. In the late 1970s the vanadate ion was shown to act as an efficient inhibitor of Na+,K+-ATPase as well as of other related phosphohydrolases. In 1980 vanadium was reported to mimic the metabolic effects of insulin in rat adipocytes. During the last decade, vanadium has been found to act in an insulin-like manner in all three main target tissues of the hormone, namely skeletal muscles, adipose, and liver. Subsequent studies revealed that the action of vanadium salts is mediated through insulin-receptor independent alternative pathway(s). The investigation of the antidiabetic potency of vanadium soon ensued. Vanadium therapy was shown to normalize blood glucose levels in STZ-rats and to cure many hyperglycemia-related deficiencies. Therapeutic effects of vanadium were then demonstrated in type II diabetic rodents, which do not respond to exogenously administered insulin. Finally, clinical studies indicated encouraging beneficial effects. A major obstacle, however, is overcoming vanadium toxicity. Recently, several organically chelated vanadium compounds were found more potent and less toxic than vanadium salts in vivo. Such a newly discovered organic chelator of vanadium is described in this review.

Insulin is a short-lived species in the circulatory system. After binding to its receptor sites and transmission of its biological signals, bound insulin undergoes receptor-mediated endocytosis and consequent degradation. An inactive insulin derivative that is not recognized by the receptor has a longer circulation life, but obviously is biologically impotent. (Fmoc)2 insulin is an insulin derivative purified through high-performance liquid chromatography in which two 9-fluorenylmethoxycarbonyl (Fmoc) moieties are covalently linked to the (alpha-amino group of phenylalanine B1 and the epsilon-amino group of lysine B29. It has 1-2% of the biological potency and receptor binding capacity of the native hormone. After incubation, (Fmoc)2 insulin undergoes a time-dependent spontaneous conversion to fully active insulin in aqueous solution at 37 degrees C and a pH range of 7-8.5. At pH 7.4, the conversion proceeds slowly (t1/2 = 12 +/- 1 days) and biological activity is generated gradually. A single subcutaneous administration of (Fmoc)2 insulin to streptozocin-treated diabetic rats normalized their blood glucose levels and maintained the animals in an anabolic state over 2-3 days. A broad shallow peak of immunoreactive insulin was found to persist in circulation over this period. To confirm further that the long-acting effect of (Fmoc)2 insulin proceeds via slow release in the blood circulation itself, we administered native insulin, NPH insulin, or the (Fmoc)2 derivative intraperitoneally. The rats recovered from hypoglycemia at t1/2 = 8.0 +/- 0.3 and 10 +/- 0.4 h after administration of native and NPH insulin, respectively. In contrast, (Fmoc)2 insulin was active for a significantly longer time, with an extended onset of t1/2 = 26 +/- 1h, and a glucose-lowering effect even 40 h after administration. (Fmoc)2 insulin was also found to be more resistant to proteolysis. Finally, we found that (Fmoc)2 insulin does not induce antigenic effects. In summary, we present here a new concept for prolonging the half-life of insulin in the circulatory system, in which receptor-mediated endocytosis and degradation is delayed and accompanied by a time-dependent generation of basal insulin.

The activating effect of vanadate on glycogenesis and on glycogen synthase (uridine diphosphate-glucose-glycogen glucosyl transferase) activity was studied in rat adipocytes and compared with that of insulin. Using several approaches and specific blockers, we found that vanadate and insulin resemble each other, in the activation of glycogen synthase, in several aspects: both require nonarrested protein phosphatase 1 activity; they are equally suppressed by conditions that elevate cAMP-levels; and both depend on the activation of phosphatidylinositol-3 kinase. The basic differences between them are as follows: 1) vanadate promotes glycogenesis through the activation of a cytosolic protein tyrosine kinase, in an insulin-receptor-independent manner; 2) vanadate elevates glucose-6-phosphate (G-6-P) to a higher level than insulin; 3) vanadate-activated glycogenesis is accompanied by an increase in the cellular content of immunoreactive glycogen synthase, an effect less noticeable with insulin; 4) adipose glucose-6-phosphatase is inhibited by vanadate (dose for 50% inhibition, IC50 = 7 +/- 0.7 microM) but not by insulin. We have concluded that insulin and vanadate activate glycogenesis through a phosphatidylinositol-3 kinase and dephosphorylation-dependent mechanism. Vanadate, however, uses a receptor-independent pathway and is superior to insulin in elevating the level of G-6-P, a key metabolite for activating glycogen synthase. This is attributed to the combined effect of vanadate in enhancing glucose entry and in inhibiting dephosphorylation of endogenously formed G-6-P. The latter effect is not exerted by insulin.

We report here on a novel procedure for measuring glycogenolysis in rat adipocytes. In this procedure, cells are incubated for 30 min at 37 degrees C with insulin or vanadate, and with [U-14C]glucose to label the glycogen pool with radioactive glucose. The cells are washed and preincubated for an additional 1 h, before being assayed. The extent of glycogenolysis is determined by the decrease in radioactivity in precipitated glycogen, which was quite substantial under experimental conditions facilitating glycogenolysis. From the assay, we determined the following. (a) Glycogenolysis is activated in rat adipocytes in response to lipolytic hormones (i.e. catecholamines and adrenocorticotropic hormone). (b) Other agents and conditions elevating intracellular adenosine 3',5'-monophosphate levels (i.e. cholera toxin, dibutyryladenosine 3',5'-monophosphate, and isobutylmethylxanthine) also activate glycogenolysis. (c) Glycogenolysis (as opposed to lipolysis) is activated at concentrations of adrenocorticotropic hormone or isoproterenol 7-11-fold lower and at adenosine 3',5'-monophosphate concentrations 7-fold lower. (d) Calyculin A, a specific inhibitor of protein phosphatase 1, activates glycogenolysis as well. Calyculin A also activates lipolysis at an equimolar potency. (e) Insulin does not antagonize glycogenolysis in rat adipocytes. In conclusion, the assay allowed us to compare glycogenolysis to lipolysis within the same cell, and to find that the sensitivity to hormones and adenosine 3',5'-monophosphate was about 1 order of magnitude higher for glycogenolysis than for lipolysis. A more striking finding was the inability of insulin to antagonize glycogenolysis in the rat adipose cell, an effect which occurs readily in liver and muscle cells via protein phosphatase 1-activating machinery. This rules out a role for adipose protein phosphatase 1 activation in the mechanism by which insulin antagonizes lipolysis and supports the contention that the insulin effect in lowering adenosine 3',5'-monophosphate levels is the central mechanism by which insulin antagonizes lipolysis.

Vanadate, which mimics the biological effects of insulin, also inhibits lipolysis in rat adipocytes. Here we demonstrate that the antilipolytic effect of vanadate differs from that of insulin at least by the five following criteria: 1) vanadate inhibits lipolysis mediated by high (supraphysiological) concentrations of catecholamines; 2) vanadate antagonizes (Bu)2cAMP-mediated lipolysis; 3) vanadate antagonizes isobutylmethylxanthine-dependent lipolysis, 4) vanadate inhibits lipolysis mediated by okadaic acid; and 5) wortmannin, which blocks the antilipolytic effect of insulin, fails to block vanadate-mediated antilipolysis. Vanadate does activate phosphoinositol 3-kinase, and wortmannin blocks this activation. Our working hypothesis assumes that all of the insulin-like effects of vanadate, including antilipolysis, are initiated by the inhibition of protein phosphotyrosine phosphatases (PTPases). Among documented PTPase inhibitors we found that VOSO4 (oxidation state +4), several organic vanadyl compounds (+4), zinc (Zn2+), tungstate (W), and molybdate (Mo) also had antilipolytic activity. The order of potency was vanadyl acetylacetonate > or = VOSO4 > or = NaVO3 > or = vanadyl-dipicolinate > Zn2+ >> W > Mo, and it correlated better with the inhibition of adipose membranal-PTPases in cell-free experiments. We have concluded that the antilipolytic effect of vanadate is 1) mechanistically distinct from that of insulin, 2) independent of phosphoinositol 3-kinase activation, and 3) independent of the lipolytic cascade. We also strongly suggest that the antilipolytic effect of vanadate emanates from inhibiting adipose membranal, rather than cytosolic PTPases, and present preliminary data showing distinct differences in catalysis between these two PTPase categories. Overall, the study indicates that antilipolysis can be manifested via alternative, insulin-independent, signal-transducing pathways.

Vanadium and its compounds exhibit a wide variety of insulin-like effects. In this review, these effects are discussed with respect to the treatment of type I and type II diabetes in animal models, in vitro actions, antineoplastic role, treatment of IDDM and NIDDM patients, toxicity, and the possible mechanism(s) involved. Newly established CytPTK plays a major role in the bioresponses of vanadium. It has a molecular weight of approximately 53 kDa and is active in the presence of Co2+ rather than Mn2+. Among the protein-tyrosine kinase blockers, staurosporine is found to be a potent inhibitor of CytPTK but a poor inhibitor of InsRTK. Vanadium inhibits PTPase activity, and this in turn enhances the activity of protein tyrosine kinases. Our data show that inhibition of PTPase and protein tyrosine kinase activation has a major role in the therapeutic efficacy of vanadium in treating diabetes mellitus.

Both exogenously added vanadate (oxidation state +5) and vanadyl (oxidation state +4) mimic the rapid responses of insulin through alternative signaling pathways, not involving insulin receptor activation [reviewed in Shechter et al. (1995) Mol. Cell. Biochem. 153, 39-47]. Vanadium exhibits complex chemistry, fluctuating between vanadate(+5) and vanadyl(+4), according to the prevailing conditions. Using several experimental approaches, we report here on a distinct vanadate(+5)-independent, vanadyl(+4)-dependent activating pathway. The key components of this pathway are membrane protein phosphotyrosine phosphatases (PTPases) and a cytosolic (nonreceptor) protein-tyrosine kinase (CytPTK). We further suggest that vanadate(+5) is not reduced rapidly to vanadyl(+4) inside the cell, and entered vanadyl sulfate(+4) is capable of undergoing spontaneous oxidation to vanadate(+5) in vivo. Finally, we show that the promotion and full expression of a downstream bioeffect such as lipogenesis requires both activation of CytPTK and prolonged stability of vanadyl(+4) against oxidation.

In previous studies, tungstate and molybdate were found to mimic the biological actions of insulin. It was suggested that these metallooxides initially inhibit vanadate-sensitive protein phosphotyrosine phosphatase (PTPase). This, in turn, stimulates a staurosporine-sensitive cytosolic protein tyrosine kinase (cytPTK), which activates several insulin bioeffects via insulin-independent pathways (Shisheva & Shechter, 1991, 1993; Elberg et al., 1994). Tungstate and molybdate, however, facilitate bioeffects in rat adipocytes only at high (millimolar) concentrations (Goto et al., 1992). We report here that incubations of tungstate or molybdate with hydrogen peroxide (H2O2) result in the formation of pertungstate (pW, peroxide of tungstate) or permolybdate (pMo, peroxide of molybdate). Pertungstate and permolybdate were found to stimulate all or most of the insulin bioeffects in rat adipocytes. Moreover, these permetallooxides are 80-180-fold more potent stimulators than the corresponding metallooxides. This shift in potency resembles that of pervanadate relative to vanadate in stimulating the same effect in rat adipocytes (Fantus et al., 1989). pW and pMo are also active in normalizing blood glucose levels in streptozotocin-induced diabetic rats. Further studies aimed at understanding the higher efficacy of this permetallooxide revealed the following: (a) All three permetallooxides (pV, pW, pMo) are oxidizing agents relative to reduced glutathione (GSH). They oxidize stoichiometric amounts of GSH to GSSG. (b) All three metallooxides do not oxidize GSH to GSSG. (c) Both metallooxides and permetallooxides inhibit rat adipocytic PTPase at micromolar quantities (IC50 = 3-10 microM). Permetallooxides, however, inhibited a larger PTPase fraction (80-100%) compared to metallooxides (40-70% of the total).(ABSTRACT TRUNCATED AT 250 WORDS)

In previous studies, we demonstrated that while okadaic acid stimulates glucose metabolism, it suppresses the bioresponses of insulin itself in rat adipocytes (Shisheva and Shechter, Endocrinology 129: 2279-2288, 1991). Both stimulation and suppression were attributed to okadaic acid-dependent inhibition of protein phosphatases 1 and 2A. We report here that exposure of adipocytes to staurosporine prior to okadaic acid restored insulin-stimulated actions on glucose metabolism. The effect was half-maximal at staurosporine concentrations as low as 70 nM and was fully expressed (80-87% of the control) at 400-500 nM. Similarly, the insulin-like effect of pervanadate, which was also suppressed by okadaic acid, was restored completely with staurosporine pretreatment. Staurosporine was less effective in restoring cell responses inhibited by high concentrations of okadaic acid, or when added to the cells after okadaic acid. Cell resensitization was unique to staurosporine and could not be produced by various agents that reduce cellular protein kinase A- or protein kinase C-dependent phosphorylation, such as phenylisopropyl adenosine (PIA), K-252a and GF 109203X. Staurosporine (400 nM) partially reversed lipolysis induced by okadaic acid but not that induced by beta-adrenergic stimulation. PIA, which antagonized okadaic acid-induced lipolysis to the same extent as staurosporine, was not capable of restoring insulin responses. Further studies aimed at elucidating this reversing effect revealed that staurosporine did not reactivate okadaic acid-inhibited protein phosphatases 1 and 2A in both cellular and cell-free systems. In summary, we report here a unique dynamic system in which insulin and pervanadate bioeffects can be fully suppressed and again re-expressed without reactivation of protein phosphatase 1 or 2A. The precise site for both effects, although still obscure, appears to be downstream from autophosphorylated insulin receptor.

Previous studies have shown that the combination of vanadate and H2O2 generates peroxide(s) of vanadate (pervanadate) that is able to mimic insulin in stimulating lipogenesis or protein synthesis and inhibiting lipolysis in rat adipocytes. Here we report that pervanadate is a potent trigger of 3-O-methylglucose transport in rat adipocytes, with an effective concentration of 5 microM and a maximum at 20 microM. Moreover, pervanadate produced an additional activation of approximately 60% on glucose influx in cells treated with maximally activating concentrations of insulin. Vanadate was ineffective in potentiating insulin-stimulated glucose uptake. Quercetin, a bioflavonoid that inhibits insulin receptor tyrosine kinase, blunted this effect of pervanadate. Treatment of adipocytes with pervanadate inhibited protein phosphotyrosyl phosphatase activity of cell extracts in a dose-dependent manner, with an ID50 of 5 microM and complete inhibition at 80 microM. In contrast, vanadate (1-800 microM) did not appreciably inhibit cell phosphotyrosyl phosphatases. The inhibitory effect of pervanadate correlated with the increase in protein phosphotyrosine accumulation, as determined by Western blotting with antiphosphotyrosine antibodies. The most prominent phosphotyrosine-containing band detected in pervanadate-treated adipocytes was that of autophosphorylated insulin receptor, identified by immunoblotting or immunoprecipitation with antiinsulin receptor antibodies. The addition of insulin to pervanadate-treated adipocytes (20 microM) caused a further increase (approximately 70%) in receptor autophosphorylation. In a cell-free system using partially purified insulin receptor devoid of tyrosine phosphatase activity, pervanadate did not stimulate the receptor autophosphorylation or interfere with the stimulating effect of insulin. These results suggest that 1) pervanadate triggers glucose uptake by increasing autophosphorylation of insulin receptor, preventing its dephosphorylation; 2) under physiological conditions, cellular protein phosphotyrosyl phosphatase activity is high, thereby significantly opposing insulin-mediated hexose transport; and 3) pervanadate has the unique ability to markedly increase maximal cell responsiveness in stimulating glucose transport achieved at a saturating insulin concentration. These findings suggest a possible clinical application in the management of glucose uptake in pathological conditions of insulin resistance and hyperinsulinemia.

We report here that quercetin, a naturally occurring bioflavonoid, is an effective blocker of insulin receptor tyrosine kinase-catalyzed phosphorylation of exogenous substrate. The ID50 was estimated to be 2 +/- 0.2 microM in cell-free experiments, using a partially purified insulin receptor and a random copolymer of glutamic acid and tyrosine as a substrate. Insulin-stimulated autophosphorylation of the receptor itself was not blocked by quercetin (up to 500 microM). In intact rat adipocytes, quercetin inhibited insulin-stimulating effects on glucose transport, oxidation, and its incorporation into lipids. Inhibition of lipogenesis (50%) occurred at 47 +/- 4 microM, whereas full inhibition was evident at 110 +/- 10 microM quercetin. In contrast, the effect of insulin in inhibiting lipolysis remained unaltered in quercetin-treated adipocytes. The inhibitor was devoid of general adverse cell affects. Basal activities and the ability of lipolytic agents to stimulate lipolysis were not affected. Inhibition by quercetin enabled us to evaluate which insulinomimetic agents are dependent on tyrosine phosphorylation of endogenous substrates for stimulating glucose metabolism. Quercetin blocked lipogenesis mediated by insulin, wheat germ agglutinin, and concanavalin A. The lipogenic effect of Zn2+ and Mn2+ was partially blocked, whereas that of vanadate was not affected at all.(ABSTRACT TRUNCATED AT 250 WORDS)

1. Bovine insulin dissolved in 0.05 M deoxycholic acid was absorbed through oral intubation in fish weighing 200-300 g. 2. The peak of absorption appeared in all fish, 30-45 min after intubation and was followed by a gradual decrease. The extent of absorption was extremely variable, with up to a 20-fold difference between individuals. No detectable insulin was found in fish intubated with vehicle. 3. The absorbed hormone retained 17-88% of its lipogenic bioactivity in vitro. The absorbed insulin lowered the levels of several amino acids in the intubated fish, indirectly indicating that its bioactivity was also retained in vivo.

Vanadate ions, low-molecular-weight phosphate analogues, mimic most of the rapid actions of insulin in various cell types. When administered orally to diabetic hyperglycemic rats, vanadate reaches the circulation, mimics insulin stimulation of glucose uptake and metabolism, and leads to normoglycemic and partial anabolic states. In addition, vanadate restores tissue responsiveness to insulin and hepatic glycogen levels and activates new synthesis of key enzymes for carbohydrate metabolism. This suggests that correcting hyperglycemia is sufficient to correct the typical metabolic alterations found in streptozocin-induced diabetic rats. Several weeks of oral administration of vanadate to diabetic rats has not produced detectable liver or kidney toxicity. The mechanism by which vanadate mimics the actions of insulin is still obscure. Unlike insulin, vanadate does not seem to stimulate the autophosphorylation and endogenous tyrosine phosphorylation of insulin-receptor kinase or other intracellular proteins either directly or by virtue of its known inhibitory effect on protein phosphotyrosine phosphatase. Results from many studies support a model in which vanadate activates glucose metabolism by either utilizing an alternative (insulin-independent) cascade or bypassing the early events of the insulin-dependent cascade. Either of these possibilities is of clinical importance, because early insulin events may become defective, as a result of severe hyperinsulinemia, and may contribute to insulin resistance. Alternative pathways by which vanadate may stimulate glucose metabolism, e.g., by increasing intracellular Ca2+ levels and/or regulating intracellular and intravesicular pH, are discussed. From a clinical perspective, studies should be continued in evaluating the level of vanadate toxicity after prolonged treatment and searching for agents that potentiate its insulin mimetic actions in vitro and in vivo.

We have studied the involvement of central nervous system (CNS) insulin receptors in mediating the central hypoglycemic effect of insulin by using insulin derivatives modified at regions of the hormone necessary for receptor reactivity and peripheral bioactivity. Acetylation or succinylation of the 3 free amino groups of insulin at positions A1, B1 and B29 resulted in a corresponding decrease in lipogenic activity in isolated rat adipocytes, with concentrations of hormone required to produce half the maximal effect (ED50) being 0.15 ng/ml, 3 ng/ml and 50 ng/ml for native insulin, acetyl3 insulin and succinyl3 insulin, respectively. Moreover, the modified insulins exhibited diminished hypoglycemic effect following central administration in mice, with the doses needed for suppression of plasma glucose to 50% of basal levels being 1 microgram, 10 micrograms and 25 micrograms for native insulin, acetyl3 insulin and succinyl3 insulin, respectively. Because binding of insulin derivatives to CNS receptors can be predicted from their peripheral bioactivity, the present finding of parallel decrements in lipogenic activity in vitro and central hypoglycemic effect in vivo, following modification of insulin at regions implicated in receptor activation, is consistent with the view that insulin exerts its central effect on plasma glucose by interacting with specific CNS receptor sites which are closely related to the peripheral insulin receptors.

We have studied in detail the binding of fluorescent derivatives of insulin and epidermal growth factor to 3T3 fibroblasts. We have used two types of fluorescent analogues of insulin and epidermal growth factor: highly fluorescent derivatives which have seven to eight rhodamine molecules or fluorescent derivatives which have a single rhodamine molecule per one molecule of insulin or epidermal growth factor. Both types of analogue retained substantial binding affinity as determined by radioreceptor assays and biological activity. The cells labeled with the fluorescent analogues were visualized with a sensitive video intensification microscopic system that enabled us to directly observe the location of the fluorescent hormone on the surface and within the living fibroblasts. We found that both insulin and epidermal growth factor initially bound diffusely to the cell surface and, at 4 degrees , remained dispersed. Within a few minutes at 23 degrees or 37 degrees the hormone-receptor complexes aggregated into patches that could be readily removed by trypsin but not by excess native hormone. The hormone-receptor complexes, which were initially mobile in the plane of the membrane, become immobilized later as the consequence of the receptor aggregation or internalization. Within approximately 30 min at 37 degrees , much of the labeled hormone was found within the cell in endocytic vesicles that moved about in the cytoplasm in a saltatory manner. The aggregation and immobilization of the hormone-receptor complexes could be due to either hormone-hormone interactions on the cell membrane or a hormone-induced conformational change in the hormone-receptor complex. Aggregation and internalization of hormone-receptor complexes could be associated with certain aspects of hormone action, hormone degradation, down regulation of receptors, or negative cooperativity of hormone binding.

We have studied the possible involvement of the calcium- and phospholipid/diacylglycerol-dependent enzyme, protein kinase C (PKC) in mediating insulin action in the central nervous system (CNS) by testing the effect of direct activation or blockade of the CNS PKC system on the plasma glucose responses to central insulin injection in mice. Insulin (0.1-1 microgram), injected into the CNS, produced rapid transient hypoglycemia. This effect appeared to involve interaction of insulin with specific receptors, since insulin analogs exhibiting diminished receptor binding affinity and peripheral bioactivity compared to the native hormone were much less active (i.e., insulin much greater than acetyl 3 insulin greater than proinsulin greater than IGF-I) or not active at all (i.e., insulin chain A and chain B). Central injection of the specific PKC activator, 12-O-tetradecanoylphorbol-13-acetate (TPA) (0.01-0.5 microgram), but not the inactive TPA analog, 4-alpha-phorbol or the unstable synthetic diacylglycerol analog, 1-oleoyl-2-acetyl-sn-glycerol (OAG), significantly enhanced the hypoglycemic response to co-administered insulin (0.5 microgram) or the insulin derivative, acetyl 3 insulin (2.5 micrograms). Central TPA had no effect on basal glucose levels. Furthermore, central administration of the selective PKC blockers, polymyxin B (PMB, 1-25 micrograms) or 1-beta-galactosylsphingosine (psychosine, 0.5-10 micrograms) but not their respective inactive analogs, polymyxin E and sphingomyelin, strongly inhibited the hypoglycemic response to insulin (1 microgram) or acetyl 3 insulin (5 micrograms). PMB and psychosine, injected alone had no effect on basal glucose levels.(ABSTRACT TRUNCATED AT 250 WORDS)

Incubation of physiological concentrations of 125I-labeled insulin with liver membranes in the presence of anti-insulin IgG results in a 7- to 15-fold increase in the specific binding of the hormone. The low-affinity/high-capacity binding sites are replaced by an apparently homogeneous class of high-affinity sites, and the nonlinear Scatchard plots are converted to linear plots without a change in the maximum number of binding sites. Similarly, the binding of insulin to receptors in 3T3 fibroblasts is increased substantially in the presence of anti-insulin antibody, and the biological activity of subactive concentrations of insulin is enhanced by antibody in these cells. However, the affinity of 125I-labeled epidermal growth factors (EGF) in fibroblasts is not affected by anti-EGF IgG. In adipocytes anti-insulin IgG in the same concentration range only inhibits the binding of insulin and suppresses insulin-mediated glucose oxidation. Monovalent Fab' fragments from anti-insulin IgG inhibit the binding of the hormone, indicating that the enhancement of binding in liver membranes and fibroblasts requires the bivalency of the antibody.

Short preincubation of rat adipocytes with wheat germ agglutinin, followed by removal of unbound lectin, resulted in persistent activation of lipogenesis, which lasted at least 3 h. The bound lectin also inhibited lipolysis initiated by isoproterenol 1 or 2 h after the removal of the free lectin. This property was also shared by other insulinomimetic lectins, such as Concanavalin A and wax bean agglutinin. Persistent bioactivation is the consequence of lectin adsorbed to external cell surface determinants in a permanent fashion. This fraction is not internalized, processed, or appreciably dissociated from the cells, since the addition of N-acetyl-D-glucosamine at any time after the onset of wheat germ agglutinin-induced persistent lipogenesis leads to termination. The property of producing persistent bioactivation is not shared by insulin itself, since removal of the unbound hormone results in termination of bioactivation. This study indicates the existence of externally located fat cell surface determinants, which, upon being occupied continuously, produce persistent insulin-like activities. The study also strongly supports the notion that the initial perturbation is sufficient to activate the insulin machinery system, and that internalization and processing of hormone-receptor complexes are the sole pathway for termination.

Mice immunized with insulin developed antibodies to both insulin and the insulin receptor. The antibodies to insulin receptor displaced labeled insulin from insulin receptors and mimicked the actions of insulin in stimulating the oxidation of glucose and its incorporation into lipids, and in inhibiting lipolysis. The antibodies to insulin receptor could be blocked by or bound to the antibodies to insulin, and therefore were identified as anti-idiotypes. Thus, immunization against a hormone may activate spontaneously an idiotype-anti-idiotype network resulting in antibodies to the hormone receptor.

This review includes data related to two substances that modulate insulin mechanisms, both in vitro and in the whole animal model. It seems to us that these agents (vanadate and PMXB) will be of potential use in the next decade for basic and applied research. They may assist in characterizing the essential post-binding events involved in insulin action, which cannot presently be identified. As vanadate and PMXB modulate the effects of insulin both in vivo and in vitro, they may be of use in clinical and pathophysiological research as well. VO3- is a low molecular weight substance which permeates the intestinal tract and mimics the actions of insulin in target tissues. Studies that were summarized here may even suggest that VO3- is superior to insulin in stimulating its effects in tissues that are down-regulated or desensitized to the hormone itself. Both VO3- and PMXB may be useful in the treatment of diabetes in the future if long-range toxicity studies prove these agents to be clinically safe. PMXB has already been in use in medicine for several decades now.

Protein tyrosine kinase (PTK) blockers which competitively inhibit the kinase activity of insulin receptors were synthesized and their properties examined. The best insulin receptor kinase (IRK) inhibitors possess either one hydroxyphenyl ring and two carboxyl groups or two phenyl rings and one carboxyl group. All the inhibitors, except tBoc-tyrosine aminomalonate, effectively block the IRK-catalyzed phosphorylation of exogenous substrate, but only partially block receptor autophosphorylation. These PTK blockers inhibit the insulin induced [14C]glucose assimilation into lipids (lipogenesis), but fail to inhibit the anti-lipolytic effect of the hormone. Only tBocTyr-aminomalonate was found to inhibit all the effects of insulin measured: insulin-stimulated phosphorylation of exogenous substrate, IRK autophosphorylation, insulin-dependent lipogenesis and the insulin-dependent anti-lipolytic effect. This inhibitor is the first blocker which is reported to block insulin-dependent anti-lipolysis. The inhibitors examined are devoid of general adverse effects since they have no effect on insulin-independent lipolysis, on [U14C]fructose assimilation or on (-)isoproterenol-stimulated lipolysis. These studies suggest that insulin-dependent lipogenesis and anti-lipolysis may be mediated by two distinguishable signalling pathways. This study also suggests that PTK inhibitors may become useful tools in the investigation of the signalling pathways of PTKs.

The effect of insulin on rat adipocytes was studied in isotonic buffers (pH 7.4) containing NaCl, CaCl2, MgSO4, KCl, and bovine serum albumin but no phosphate or bicarbonate anions. In phosphate- and bicarbonate-free buffers the dose-response curve to insulin is shifted to the right, the effects of the hormone on hexose uptake, glucose metabolism, and inhibition of lipolysis being observed at much higher (nearly 2 orders of magnitude) concentrations of insulin. The insulin binding capacity of the cells is only slightly changed. The dose-response curve for isoproterenol which stimulates lipolysis in the same cell type is almost the same in both Krebs-Ringer bicarbonate buffer and phosphate- and bicarbonate-free buffers. The dose-response curves for agents that mimic the action of insulin such as wheat germ agglutinin or vanadate ions are also shifted to the right. The dose-response curve to insulin can be returned to "normal" by readdition of either bicarbonate or phosphate. Almost complete recovery is obtained at either 10 mM bicarbonate or 24 mM phosphate, respectively. External Ca2+ ions which are not required for the proper action of insulin in fat cells maintained in Krebs-Ringer bicarbonate buffer, become essential for insulin action in bicarbonate-free buffer. The study indicates that depletion of bicarbonate and, to a lesser extent, phosphate anions, interferes with an essential insulin-dependent post-binding event. Also, in bicarbonate-free medium, external Ca2+ ions are essential for insulin-mediated processes. The implications of this study to the mode of action of insulin, and to physiological and clinical states of insulin desensitization are discussed.

We immunized mice with insulin and found that those strains that develop insulin antibodies subsequently produce insulin-like activity in amount equivalent to 300-400 ng insulin per ml serum. The activity was due exclusively to IgG2 antibodies. Bioactivity could be blocked efficiently by insulin antibodies from guinea pigs and from mice. The active IgG2 also displaced labeled insulin from fat cells. Preliminary in vivo studies have indicated that the appearance of insulin-like antibodies in the mouse resulted in abnormal glucose homeostasis and "down regulation" of insulin receptors. These results indicate that immunization to insulin can initiate an idiotype-anti-idiotype network resulting in antibodies to the hormone receptor.

In the preceding paper (Shechter, Y., and Ron, A. (1986) J. Biol. Chem. 261, 14945-14950) we have shown that in fat cells, prepared and maintained in an isotonic buffer (pH 7.4) containing neither phosphate nor bicarbonate anions (Buffer A), the dose-response curve to insulin shifted to the right by about 2 logarithms and insulin binding affinity or capacity was only slightly decreased. In the current paper we demonstrate that progressive loss of insulin binding, either by treatment with trypsin or preincubating the cells with isoproterenol, correlates well with the reduced ability of the cells to elicit maximal lipogenesis in response to insulin. We further demonstrate in the "new" system that: the dissociation of labeled insulin from fat cells is not accelerated by the inclusion of unlabeled insulin in the medium; termination of lipogenesis in Buffer A occurs immediately; ligand-induced receptor internalization is grossly defective; and insulin is unable to stimulate lipogenesis at 15 degrees C. The data support the hypothesis that in the new experimental system all measurable binding sites are linked to a coupling mechanism. Each site behaves as an independent, separate entity and there are no site to site interactions. This leads to a linear relationship between binding and bioactivation, lack of negative or positive cooperatively, accelerated rate of termination, defective internalization, a shift to the right in the dose-response curve to insulin, and a lack of insulin response at a lower temperature. In more general terms, the study indicates that all measurable insulin receptors are chemically homogeneous in their potential capability to be coupled to an insulin effector (biologically relevant) system, and they do so under particular experimental conditions.

We have used video intensification microscopy to observe fluorescent derivatives of insulin, epidermal growth factor and alpha2-macroglobulin added to Swiss 3T3-4 cells. At 4 degrees C, each of these polypeptides binds diffusely to specific receptors on the cell surface. When the cells are warmed to 23 or 37 degrees C, the bound insulin epidermal growth factor or alpha2-macroglobulin rapidly forms patches on the cell surface and is internalized. Using fluorescein-labeled alpha2-macroglobulin and rhodamine-labeled derivatives of insulin and epidermal growth factor, we show that all three polypeptides are internalized within the same vesicles by a common pathway. The mechanism for the internalization of these molecules is discussed.

The cyclic decapeptide, polymyxin B (PMXB), was found to inhibit hypoglycemia in mice receiving exogenous insulin (Amir, S., and Shechter, Y. (1985) Eur. J. Pharmacol. 110, 283-285). In this study, we have extended this observation to rats. Insulin-dependent hypoglycemia in rats is efficiently blocked at a 12:1 molar ratio of PMXB to insulin. This effect is highly specific, as it could not be mimicked by a variety of antibiotics or positively charged substances. Chemical modifications of PMXB have revealed that the ring structure, rather than the tail structure, is important for anti-insulin-like activity. Colistin A, which differs from PMXB by one conservative amino acid substitution in the ring structure, is devoid of this activity. Polymyxin B does not interact with insulin, nor does it alter the rate of insulin absorption and/or degradation, or the ability of insulin to bind to target tissues. This peptide inhibits hypoglycemia by blocking insulin-dependent activation of the hexose transport mechanism, as deduced by in vitro studies. The effect of insulin in stimulating hexose uptake (and subsequent glucose metabolism) in both isolated muscle tissue and adipocytes is blocked with little or no effect on the basal activities of these processes. Colistin A has no significant inhibiting effect. Other insulin-dependent activities, such as inhibition of lipolysis in adipocytes or synthesis of DNA in muscle cells, are not inhibited. It is concluded that PMXB inhibits, in a highly specific manner, the action of insulin in stimulating hexose transport and subsequent glucose metabolism, both in vitro and in the whole animal model.

Fluorescent derivatives of insulin and epidermal growth factor bound to 3T3 mouse fibroblasts are mobile on the cell surface, with similar diffusion coefficients, D approximately (3--5) x 10(-10) cm2/sec at 23 degrees C. Increasing the temperature to 37 degrees C results in rapid receptor immobilization. The immobilization is attributed to aggregation of hormone-receptor complexes, their internalization, or a combination of both processes.