Loss of multidrug and toxin extrusion 1 (MATE1) is associated with metformin-induced lactic acidosis

Abstract

Backgrounds and purpose: Lactic acidosis is a fatal adverse effect of metformin, but the risk factor remains unclear. Multidrug and toxin extrusion 1 (MATE1) is expressed in the luminal membrane of the kidney and liver. MATE1 was revealed to be responsible for the tubular and biliary secretion of metformin. Therefore, some MATE polymorphisms, that cause it to function abnormally, are hypothesized to induce lactic acidosis. The purpose of this study is to clarify the association between MATE dysfunction and metformin-induced lactic acidosis.

Experimental approach: Blood lactate, pH and bicarbonate ion (HCO(3) (-) ) levels were evaluated during continuous administration of 3 mg·mL(-1) metformin in drinking water using Mate1 knockout (-/-), heterozygous (+/-) and wild-type (+/+) mice. To determine the tissue accumulation of metformin, mice were given 400 mg·kg(-1) metformin orally. Furthermore, blood lactate data were obtained from diabetic patients given metformin.

Key results: Seven days after metformin administration in drinking water, significantly higher blood lactate, lower pH and HCO(3) (-) levels were observed in Mate1(-/-) mice, but not in Mate1(+/-) mice. The blood lactate levels were not affected in patients with the heterozygous MATE variant (MATE1-L125F, MATE1-G64D, MATE2-K-G211V). Sixty minutes after metformin administration (400 mg·kg(-1) , p.o.) the hepatic concentration of metformin was markedly higher in Mate1(-/-) mice than in Mate1(+/+) mice.

Conclusion and implications: MATE1 dysfunction caused a marked elevation in the metformin concentration in the liver and led to lactic acidosis, suggesting that the homozygous MATE1 variant could be one of the risk factors for metformin-induced lactic acidosis.

Figure 1

Long-term toxicity of metformin in Mate1^+/+, Mate1^+/− and Mate1^−/− mice. Mate1^+/+ (n= 9), Mate1^+/− (n= 9) and Mate1^−/− (n= 11) mice were given 3 mg·mL⁻¹ metformin in drinking water for 21 days. The mean daily dose of each 7 days (A), body weight change from baseline (B), blood lactate level (C), pH (D) and HCO₃^- levels (E) were determined at 0, 7, 14 and 21 days after metformin treatment. Blood samples were collected under anaesthesia in 4 h-fasted mice. Blood lactate, pH and HCO₃^- levels were measured by i-STAT. Each point represents the mean ± SEM. *P < 0.05, significantly different from Mate1^+/+ mice at each day.

Figure 2

Lactate concentration–time profile in diabetic patients. Metformin was administered to patients in the MATE-reference group (n= 25) and heterozygous MATE-variant group (n= 4). Blood lactate levels were measured by Lactate Pro at 0, 4 and 9 h after the oral administration of metformin. One MATE1-L125F variant carrier, two MATE1-G64D variant carriers and one MATE2-K-G211V variant carriers were found in this study. Each point represents the mean ± SD.

Figure 3

Blood lactate level in vehicle- or metformin-treated Mate1^+/+ and Mate1^−/− mice after a single dose of metformin. All mice were given 400 mg·kg⁻¹ metformin via oral gavage. Blood lactate levels were determined before and 24 h after the oral administration of metformin in vehicle-treated Mate1^+/+ mice (n= 6), vehicle-treated Mate1^−/− mice (n= 7), metformin-treated Mate1^+/+ mice (n= 7) and metformin-treated Mate1^−/− mice (n= 10). Each point represents the mean ± SEM. *P < 0.05, significantly different from vehicle-treated mice with the same genotype. ^#P < 0.05, significantly different from metformin-treated Mate1^+/+ mice.

Figure 4

Pharmacokinetics of metformin in Mate1^+/+ and Mate1^−/− mice after a single dose of metformin. In the same mice as shown in Figure 3, metformin concentrations in plasma (A), liver (B), kidney (D) and skeletal muscle (F) were determined by HPLC. K_p values in the liver (C), kidney (E) and skeletal muscle (G) were calculated by dividing the tissue concentration by the plasma concentration of metformin. Data represent mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, significantly different from Mate1^+/+ mice.

Figure 5

Lactate and metformin concentration–time profiles in Mate1^−/− mice. Overnight-fasted mice were given 150 mg·kg⁻¹ metformin (n= 8). Plasma concentration of metformin and blood lactate levels ere measured by HPLC and Lactate Pro respectively. Data represent mean ± SEM.

Figure 6

Concentration dependence of [¹⁴C]-metformin uptake by mouse Mate1-, mouse Oct1- or mouse Oct2-expressing cells. The cells were incubated with [¹⁴C]-metformin in the presence or absence of 5 mM MPP at 37°C for 1 min in mouse Mate1-expressing cells (A) and for 2 min in mouse Oct1- (B) or mouse Oct2-expressing cells (C). Data represent mean ± SEM.