Oral insulin-mimetic compounds that act independently of insulin.

The hallmarks of insulin action are the stimulation and suppression of anabolic and catabolic responses, respectively. These responses are orchestrated by the insulin pathway and are initiated by the binding of insulin to the insulin receptor, which leads to activation of the receptor's intrinsic tyrosine kinase. Severe defects in the insulin pathway, such as in types A and B and advanced type 1 and 2 diabetes lead to severe insulin resistance, resulting in a partial or complete absence of response to exogenous insulin and other known classes of antidiabetes therapies. We have characterized a novel class of arylalkylamine vanadium salts that exert potent insulin-mimetic effects downstream of the insulin receptor in adipocytes. These compounds trigger insulin signaling, which is characterized by rapid activation of insulin receptor substrate-1, Akt, and glycogen synthase kinase-3 independent of insulin receptor phosphorylation. Administration of these compounds to animal models of diabetes lowered glycemia and normalized the plasma lipid profile. Arylalkylamine vanadium compounds also showed antidiabetic effects in severely diabetic rats with undetectable circulating insulin. These results demonstrate the feasibility of insulin-like regulation in the complete absence of insulin and downstream of the insulin receptor. This represents a novel therapeutic approach for diabetic patients with severe insulin resistance.

The imidazoline I2-site ligands BU 224 and 2-BFI inhibit MAO-A and MAO-B activities, hydrogen peroxide production, and lipolysis in rodent and human adipocytes.

Numerous imidazolinic agents exhibit antihyperglycaemic properties and have been described to promote insulin secretion, however their effects on adipose tissue development have been poorly investigated. Since white adipose tissue (WAT) plays an important role in glucose homeostasis and expresses imidazoline (I(2)) binding sites abundantly, this work aimed at studying extrapancreatic actions of two I(2)-site ligands, BU 224 and 2-BFI in adipocytes. Interaction with monoamine oxidase (MAO) was investigated by measuring the ability to modulate [(14)C]tyramine oxidation and hydrogen peroxide production. Direct influence on glucose uptake or on lipolytic activity was tested on mouse, rat, rabbit and human adipocytes. BU 224 and 2-BFI behaved as reversible inhibitors of both MAO-A and -B, as demonstrated by total inhibition of tyramine oxidation in human adipocytes and platelets or in liver from rats previously treated with selective MAO-inhibitors. Moreover, they weakly inhibited semicarbazide-sensitive amine oxidase. Like classical MAO-inhibitors, they were unable to produce hydrogen peroxide and to activate glucose uptake but prevented tyramine to do so in rodent or human adipocytes. BU 224 and 2-BFI also differed from MAO-inhibitors since they inhibited lipolysis at millimolar concentrations via a still undefined pathway independent of alpha(2)-adrenoceptor stimulation, beta-adrenergic antagonism and MAO activation. However, chronic treatment of obese Zucker rats with 2-BFI did not modify the maximal lipolytic capacity or the mild insulin resistance status of their adipocytes. Taken together, our observations demonstrate on WAT novel effects of BU 224 and 2-BFI different from their already reported actions on brain or endocrine pancreas.

Role of mu-opioid receptors in insulin release in the presence of inhibitory and excitatory secretagogues.

In mouse pancreatic islets incubated under static conditions, the inhibitory effects on glucose-evoked insulin release induced by adrenaline (1 microM), clonidine (2 microM) and UK 14,304 (brimonidine, 0.001-1 microM) were abolished by naloxone (30 nM). Only CTOP (D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Phe-Thr-NH(2), 0.1 microM), a very selective mu-opioid receptor antagonist, blocked the response to UK 14,304. Glucose-induced insulin secretion was attenuated by both beta-endorphin (0.01 microM) and endomorphin-1 (0.1 microM). Naloxone and CTOP prevented these inhibitory responses. The stimulatory effect of glibenclamide (1 microM) was also reduced by endomorphin-1. However, when islets were incubated in the presence of K(+) (30 mM), carbachol (100 microM) or forskolin (0.1 microM), neither the inhibitory effect induced by UK 14,304 was reversed by naloxone, nor endomorphin-1 altered the responses promoted by the excitatory agents. Thus, alpha(2)-adrenoceptor stimulation might inhibit glucose-induced insulin secretion by releasing endogenous opioids. Mu-Opioid receptor activation and opening of K(ATP) channels could be involved in the response.

Differential sensitivity to adrenergic stimulation underlies the sexual dimorphism in the development of diabetes caused by Irs-2 deficiency.

The diabetic phenotype caused by the deletion of insulin receptor substrate-2 (Irs-2) in mice displays a sexual dimorphism. Whereas the majority of male Irs-2(-/-) mice are overtly diabetic by 12 weeks of age, female Irs-2(-/-) animals develop mild obesity and progress less rapidly to diabetes. Here we investigated ß-cell function and lipolysis as potential explanations for the gender-related differences in this model. Glucose-stimulated insulin secretion was enhanced in islets from male null mice as compared to male WT whereas this response in female Irs-2(-/-) islets was identical to that of female controls. The ability of α(2)-adrenoceptor (α(2)-AR) agonists to inhibit insulin secretion was attenuated in male Irs2 null mice. Consistent with this, the expression of the α(2A)-AR was reduced in male Irs-2(-/-) islets. The response of male Irs-2(-/-) islets to forskolin was enhanced, owing to increased production of cAMP. Basal lipolysis was increased in male Irs-2(-/-) but decreased in female Irs-2(-/-) mice, concordant with the observation that adipose tissue is sparse in males whereas female Irs2 null mice are mildly obese. Adipocytes from both male and female Irs-2(-/-) were resistant to the anti-lipolytic effects of insulin but female Irs-2(-/-) fat cells were additionally resistant to the catabolic effects of beta-adrenergic agonists. This catecholamine resistance was associated with impaired generation of cAMP. Consequently, targets of cAMP-dependent protein kinase (PKA) which mediate lipolysis were not phosphorylated in adipose tissue of female Irs-2(-/-) mice. Our findings suggest that IRS-2 deficiency in mice alters the expression and/or sensitivity of components of adrenergic signaling.

Methylamine but not mafenide mimics insulin-like activity of the semicarbazide-sensitive amine oxidase-substrate benzylamine on glucose tolerance and on human adipocyte metabolism.

It has been reported that benzylamine reduces blood glucose in rabbits, stimulates hexose uptake, and inhibits lipolysis in mouse, rabbit, and human adipocytes. In the presence of vanadate, benzylamine is also able to improve glucose disposal in normoglycaemic and diabetic rats. Such insulin-mimicking properties are the consequence of hydrogen peroxide production during benzylamine oxidation by semicarbazide-sensitive amine oxidase (SSAO). The aim of the study was to determine whether other SSAO-substrates could share such potential antidiabetic properties. Thus, mafenide, a synthetic antimicrobial sulfonamide structurally related to benzylamine, and which has been recently reported to interact with SSAO, was tested in the above mentioned models, in parallel with methylamine, a proposed endogenous SSAO-substrate. All tested amines stimulated glucose uptake and inhibited lipolysis in rat and mouse fat cells. Methylamine and benzylamine, but not mafenide, reduced the hyperglycaemic response during a glucose tolerance test in rabbits while the three amines tested were devoid of insulin-releasing activity under both in vivo and in vitro conditions. In human adipocytes, mafenide did not stimulate glucose transport since it was not a high-affinity substrate for SSAO and generated less hydrogen peroxide than benzylamine or methylamine. Therefore, mafenide could not be considered as an antidiabetic drug despite being oxidized and exhibiting insulin-mimicking effects in rat and mouse adipocytes. By contrast, the endogenous substrate methylamine improved glucose utilization in all in vitro and in vivo models, leading to consider novel SSAO substrates as drugs with potential anti-hyperglycaemic properties.

Benzylamine exhibits insulin-like effects on glucose disposal, glucose transport, and fat cell lipolysis in rabbits and diabetic mice.

Benzylamine, a substrate of semicarbazide-sensitive amine oxidase (SSAO), stimulates glucose transport in rat adipocytes and improves glucose disposal in diabetic rats only in the presence of vanadate. These effects have been described to result from a synergism between the hydrogen peroxide formed during amine oxidation and vanadate, via the generation of pervanadate, a powerful insulin mimicker. However, it has also been reported that benzylamine alone can stimulate glucose uptake and inhibit lipolysis in human fat cells. In this work, we therefore investigated whether benzylamine on its own was able to induce both in vivo and in vitro insulin-like responses in animal models other than rat. In rabbits, the i.v. infusion of 7 micromol/kg benzylamine before a glucose tolerance test resulted in a net reduction of the hyperglycemic response without a change in insulin secretion. Benzylamine also improved glucose tolerance and reduced lipid mobilization in hyperglycemic/obese mice. In vitro, 0.1 mM benzylamine stimulated glucose transport and inhibited lipolysis in mouse and rabbit adipocytes. These effects were blocked by previous treatments with semicarbazide, a SSAO inhibitor. Levels of benzylamine oxidation were more elevated in mouse than in rabbit adipose tissues, whereas the reverse was observed for skeletal muscles. Finally, benzylamine was unable to stimulate insulin secretion by isolated pancreatic islets from both species and SSAO activity was hardly detectable in pancreas. Together, our results bring evidence that benzylamine on its own can improve glucose tolerance in rabbit and mouse, likely by stimulating glucose uptake via amine oxidase activation in insulin-sensitive tissues.

Ras-GRF1 signaling is required for normal beta-cell development and glucose homeostasis.

Development of diabetes generally reflects an inadequate mass of insulin-producing beta-cells. beta-cell proliferation and differentiation are regulated by a variety of growth factors and hormones, including insulin-like growth factor I (IGF-I). GRF1 is a Ras-guanine nucleotide exchange factor known previously for its restricted expression in brain and its role in learning and memory. Here we demonstrate that GRF1 is also expressed in pancreatic islets. Interestingly, our GRF1-deficient mice exhibit reduced body weight, hypoinsulinemia and glucose intolerance owing to a reduction of beta-cells. Whereas insulin resistance is not detected in peripheral tissues, GRF1 knockout mice are leaner due to increased lipid catabolism. The reduction in circulating insulin does not reflect defective glucose sensing or insulin production but results from impaired beta-cell proliferation and reduced neogenesis. IGF-I treatment of isolated islets from GRF1 knockouts fails to activate critical downstream signals such as Akt and Erk. The observed phenotype is similar to manifestations of preclinical type 2 diabetes. Thus, our observations demonstrate a novel and specific role for Ras-GRF1 pathways in the development and maintenance of normal beta-cell number and function.