Elevated levels of palmitate accentuate glucose-stimulated insulin secretion (GSIS) after short-term and cause beta-cell dysfunction after prolonged exposure. We investigated whether metformin, the first-line oral drug for treatment of T2DM, has beneficial effects on FFA-treated human islets and the potential mechanisms behind the effects. Insulin secretion, oxygen consumption rate (OCR), AMPK activation, endoplasmic reticulum (ER) stress and apoptosis were examined in isolated human islets after exposure to elevated levels of palmitate in the absence or presence of metformin. Palmitate exposure doubled GSIS after 2 days but halved after 7 days compared with control. Inclusion of metformin during palmitate exposure normalized insulin secretion both after 2 and 7 days. After 2-day exposure to palmitate, OCR and the marker of the adaptive arm of ER stress response (sorcin) were significantly raised, whereas AMPK phosphorylation, markers of pro-apoptotic arm of ER stress response (p-EIF2α and CHOP) and apoptosis (cleaved caspase 3) were not affected. Presence of metformin during 2-day palmitate exposure normalized OCR and sorcin levels. After 7-day exposure to palmitate, OCR and sorcin were not significantly different from control level, p-AMPK was reduced and p-EIF2α, CHOP and cleaved caspase 3 were strongly upregulated. Presence of metformin during 7-day culture with palmitate normalized the level of p-AMPK, p-EIF2α, CHOP and cleaved caspase 3 but significantly increased the level of sorcin. Our study demonstrates that metformin prevents early insulin hypersecretion and later decrease in insulin secretion from palmitate-treated human islets by utilizing different mechanisms.
Supplementary Fig 1. Dose-dependent effect of metformin on GSIS, insulin content, and phosphorylation of AMPK from human islets. Human islets were incubated at 5.5 mM glucose in the absence (C) or presence of indicated concentrations of metformin for 7 days. After culture, islets were perifused with 2 mM glucose (G2) for 1 hour followed by 20 minutes of perifusion with 20 mM glucose (G20). Representative graphs are shown (Panel A). Total insulin secretion at 20 mM glucose during 20 min was calculated, normalized to protein and expressed as fold control (Panel B). After culture, islets were lysed and used for measurements of insulin content and AMPK phosphorylation. Insulin content is normalized to total protein and expressed as fold control (Panel C). p-AMPK/AMPK ratio was calculated and expressed as fold control (Panel D). Results are means ± SEM of 4-7 donors. * P<0.05 vs C.
Supplementary Fig 2. Effect of metformin on human islets in the absence of palmitate. Human islets were cultured in the absence (C) or presence of 25 μM metformin (M) for 2 days (M_2d) and 7 days (M_7d). After culture, human islets were perifused for 1 hour with 2 mM glucose (G2) followed by 20 minutes of perifusion with 20 mM glucose (G20). Representative graphs are shown (Panel A). Total insulin secretion at 20 mM glucose during 20 min was calculated, normalized to protein, and expressed as fold control (Panel B). After perifusion, islets were lysed, and insulin content was normalized to total protein content and expressed as fold control (Panel C). Oxygen consumption rate (OCR) was measured. OCR from a representative experiment, where each point is a mean ± SEM of 5 replicates from one donor, is shown (Panel D). The injection of 5 μM oligomycin, 5 μM FCCP, and combination of 5 μM of rotenone and 5 μM antimycin are shown as indicated (Panel D). Mitochondrial OCR was calculated by subtracting non-mitochondrial OCR from the total OCR (Panel E). ATP-coupled (Panel F) and proton leak (Panel G) portions of mitochondrial OCR were calculated based on the effect of oligomycin. Maximal OCR (Panel H) was determined by analyzing the effect of FCCP. ECAR measurements (Panel I) were taken in parallel. After treatments, human islets were lysed, and expression of p-AMPK and AMPK (Panel J), p-eIF2α (Panel K), CHOP (Panel L), sorcin (Panel M) and cleaved caspase-3 (Panel N) were determined by western blotting. p-AMPK/AMPK ratio was calculated and expressed as fold control. Other protein levels were normalized to β-actin and expression as fold control. Results are means ± SEM of 4-6 donors. * P<0.05 vs C.