The insulin-mimetic effect of vanadate is well established, and vanadate ha
s been shown to improve insulin sensitivity in diabetic rats and humans. Al
though the exact mechanism(s) remain undefined, we have previously demonstr
ated a direct relation of intracellular free magnesium (Mg) levels to gluco
se disposal, to insulinemic responses following glucose loading, and to ins
ulin-induced ionic effects. To investigate whether the insulin-mimetic effe
cts of vanadate could similarly be mediated by Mg-i, we utilized P-31-nucle
ar magnetic resonance spectroscopy to measure Mgi in erythrocytes from norm
al (NL, n= 10) and hypertensive (HTN, n= 12) subjects, before and after inc
ubation with insulin and with different doses of sodium vanadate. In NL, va
nadate elevated Mg, levels, with maximum efficacy at 50 mu mol/L (186 +/-6
to 222 +/-6 mu mol/L, P >0.01), as did physiologically maximal doses of ins
ulin, 200 muU/mL (185 +/-6 to 222 +/-8 mu mol/L, P <0.01). In HTN, only van
adate, but not insulin, increased Mg, (insulin: 173 +/-7 to 180 +/-9 mu mol
/L, P=NS; vanadate: 170 +/-7 to 208 +/- 10 mu mol/L, P <0.01). Mg-i respons
es to insulin (r=0.637, P <0.001), but not to vanadate (r=0.15, P=NS), were
closely and directly related to basal Mgi levels. We conclude that (1) bot
h vanadate and insulin stimulate erythrocyte Mg, levels; (2) cellular Mgi r
esponses to insulin, but not to vanadate, depend on basal Mgi content-the l
ower the basal Mg,, the less the Mgi response to insulin. As such, (3) Mgi
responses to vanadate were equivalent among HTN and NL, whereas HTN cells e
xhibited blunted Mg-i responses to insulin, and (4) the ability of vanadate
to improve insulin sensitivity clinically may be mediated, at least in par
t, by its ability to increase Mg-i levels, which in turn, helps to determin
e cellular insulin action.