![[Previous Articles]](pubtitle.gif){kind=small}
Phosphoinositol Turnover Pathway
ACKNOWLEDGEMENTS: We would like to thank undergraduate researchers Brant Gillam, Melanie Overley and Pat Ayers for their help in the lab. This work was supported by grants to B.J.S. from LIPHA Laboratories, the Undergraduate Research Opportunities Program at UCD, and the National Science Foundation (MCB-9220108).
Metformin stimulated the in vitro activity of three tyrosine kinases by 20-40% (cAMP-dependent kinase A was not stimulated). These data are supported by in vivo studies showing metformin stimulation of different tyrosine kinases including the insulin [4,5,6], IGF-1 [5, 10] and EGF [10] receptors. The stimulation of tyrosine kinases by cationic metformin may be similar to the well-characterized ability of polycations (e.g., polylysine) to stimulate IPßIRK and the intact insulin receptor [e.g., 11, 12].
The inability of others to note the direct effect of metformin on the insulin receptor may be due to the metformin concentration used, the use of detergent-solubilized or partially activated receptors (high basal levels of phosphorylation may obscure metformin stimulation), or the fact that the assay was not sufficiently sensitive to record the small stimulation noted here.
Direct activation of the insulin receptor tyrosine kinase by metformin may be important in diabetes therapy since the relative clinical efficacy of metformin and 3 derivatives is equivalent to the relative efficacy of these drugs to stimulate IPßIRK. In addition, the concentration of metformin required for activation of tyrosine kinases in the plasma membrane-cortex or the intact cell is similar to that required for in vitro metformin stimulation of IPßIRK. For example, with in vivo Xenopus preparations, maximal metformin stimulation occurred at 1-10 µg/ml [6]. For rat adipocytes [2], 1-10 µg/ml metformin was optimal, whereas for human adipose tissue [13], 2-4 µg/ml was best. Further evidence is needed to conclude whether direct metformin activation or inhibition of kinases is fully responsible for the multiple, in vivo actions of metformin (e.g., enhanced fibrinolysis, glycogen synthesis and glucose uptake, inhibition of PAI-1 production, decreased cell proliferation, blood triglyceride and platelet aggregation, and various effects on microcirculation).
There is a difference between these results and those we previously reported: metformin stimulation of IPßIRK tyrosine kinase activity shown here was less than the 2- to 3-fold increase in tyrosine kinase activity in the plasma membrane-cortex preparation [6]. The rather small 20-30% stimulation of IPßIRK by metformin is not entirely unexpected since metformin is unable to fully mimic insulin action in the intact cell [e.g., 6] and since the plasma membrane-cortex preparation should contain numerous tyrosine kinases that may show different levels of activation by metformin. Furthermore, an activating calcium feedback loop found in the plasma membrane-cortex preparation [6] is not present with the IPßIRK preparation.
Higher concentrations of metformin or MGBG inhibited IPßIRK and tyrosine kinase activity in a Xenopus plasma membrane-cortex preparation (50-200 µg/ml metformin; unpublished results, BJS). This kinase inhibition by higher levels of metformin may be reflected in the fact that higher metformin concentrations inhibit insulin action in human Hep G2 cells (at 13-130 µg/ml metformin)[14], human adipose tissue (at >10 µg/ml)[13] and rat adipocytes (at 100 µg/ml)[2].
Metformin also inhibits the proliferation of HeLa or KB cells [15, 16] and the drug may reduce atherosclerosis through its ability to inhibit proliferation of smooth muscle cells [17, 18]. The biguanide MGBG has been used clinically as an anticancer agent and it inhibits cell division in leukemia cells [19]. We suggest that inhibition of cell division by metformin or MGBG is due to inhibition of tyrosine kinases associated with proliferation.