The Impacts of SLC22A1 rs594709 and SLC47A1 rs2289669 Polymorphisms on Metformin Therapeutic Efficacy in Chinese Type 2 Diabetes Patients

Abstract

Background. We aimed to investigate the distributive characteristics of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms and their influence on metformin efficacy in Chinese T2DM patients. Methods. The distributions of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms were determined in 267 T2DM patients and 182 healthy subjects. Subsequently, 53 newly diagnosed patients who received metformin monotherapy were recruited to evaluate metformin efficacy. Results. No significant difference was found between T2DM patients and healthy subjects in SLC22A1 rs594709 and SLC47A1 rs2289669 allele frequencies and genotype frequencies. After metformin treatment, SLC22A1 rs594709 GG genotype patients showed a higher increase in FINS ( $p=\mathrm{0.015}$ ) and decrease in HOMA-IS ( $p=\mathrm{0.001}$ ) and QUICKI ( $p=\mathrm{0.002}$ ) than A allele carriers. SLC47A1 rs2289669 GG genotype patients had a higher decrease in TChol ( $p=\mathrm{0.030}$ ) and LDL-C ( $p=\mathrm{0.049}$ ) than A allele carriers. Among SLC22A1 rs594709 AA genotype, patients with SLC47A1 rs2289669 AA genotype showed a higher decrease in FBG ( $p=\mathrm{0.015}$ ), PINS ( $p=\mathrm{0.041}$ ), and HOMA-IR ( $p=\mathrm{0.014}$ ) than G allele carriers. However, among SLC22A1 rs594709 G allele carriers, SLC47A1 rs2289669 AA genotype patients showed a higher decrease in TChol ( $p=\mathrm{0.013}$ ) than G allele carriers. Conclusion. Our data suggest that SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms may influence metformin efficacy together in Chinese T2DM patients.

This work is licensed under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.

1. Introduction

Metformin is the most widely used oral antidiabetic drug, which could effectively lower blood glucose and reduce the risk of cardiovascular diseases [1]. The classical functional mechanism of metformin is to decrease the hepatic energy status by inhibiting the mitochondrial respiratory-chain complex 1, thus activating the adenosine 5′-monophosphate-activated protein kinase (AMPK), a cellular metabolic sensor, to finally lower the blood glucose level [2]. In recent years, further probable mechanisms involved in cAMP, the phosphorylation of Acc1 and Acc2, mitochondrial glycerophosphate dehydrogenase, and the activated duodenal AMPK were found to affect metformin’s effect in different ways [3–6]. With its significant glycemic benefits and neutral effects on weight, metformin remains the recommended first-line oral therapy in the guidelines of the American Diabetes Association and European Association of the Study of Diabetes [7]. However, during its clinical usage, the individual differences and drug failure widely existed [8, 9].

Metformin is not metabolized in vivo. The gastrointestinal absorption, hepatic uptake, and renal excretion of metformin are mediated greatly by organic cation transporters (OCTs) and multidrug and toxin extrusion proteins (MATEs) [10]. It was widely reported that genetic variants in coding genes of transporters mentioned above played important roles in pharmacokinetics and pharmacodynamics of metformin [9, 11–13]. Among these, SLC22A1 rs628031 (tagged to rs594709 in Asian) was reported to affect imatinib uptake in chronic myeloid leukemia (CML) cell line [14]. As we know, imatinib is also transported by OCT1 in the human organism. SLC47A1 rs2289669 was reported to be associated with metformin excretion [15] as well as its therapeutic response [16, 17]. Besides, a gene-gene interaction between genetic variants in SLC47A1 and SLC22A1, SLC47A1 and SLC22A2 had a significant influence on metformin’s efficacy [18, 19]. It was found that there was a SNP-SNP interaction between SLC47A1 rs2289669 and SLC22A1 rs622342 [18]. However, the effect of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms and SNP-SNP interaction on metformin efficacy in Chinese T2DM patients has not been widely evaluated.

In this study, we aimed to investigate the distributive characteristics of SLC22A1 rs594709 and SLC47A1 rs2289669 between T2DM patients and healthy subjects and the influences of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms on metformin’s therapeutic efficacy in glucose lowering, serum lipids, and insulin sensitivity. Moreover, we are interested in whether there is a SNP-SNP interaction between SLC22A1 rs594709 and SLC47A1 rs2289669 on metformin’s therapeutic efficacy.

2. Material and Methods

2.1. Subjects

Distributive characteristics of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms were determined in 267 unrelated patients with T2DM and 182 healthy subjects. 267 T2DM patients (143 males and 124 females) were recruited from the Second Xiangya Hospital of Central South University, the People’s Hospital of Liuyang, and the First Hospital of Changsha. 182 healthy subjects (86 males and 96 females) were recruited from Physical Examination Center of Xiangya Hospital of Central South University. T2DM was diagnosed according to DM diagnostic criteria of WHO 1999. Patients with type 1 diabetes, gestational diabetes, uncontrollable hypertension, severe liver, and kidney disease were not included in this study.

Furthermore, 53 of 267 T2DM patients who were newly diagnosed and drug naïve (31 males and 22 females; average age 49 years, range 29–73 years) were enrolled to evaluate the effect of metformin among different SLC22A1 rs594709 and SLC47A1 rs2289669 genotype individuals. Newly diagnosed type 2 diabetes was defined as being diagnosed with T2DM within 6 months before enrollment. All patients had not been treated with any insulin secretagogue and received metformin monotherapy for 3 consecutive months. Metformin doses ranged from 1000 mg/day to 2000 mg/day among individuals. The study was approved by the Ethics Committee of Xiangya School of Medicine, Central South University (Changsha, Hunan, China) (CTXY-110002-5), and was performed in accordance with the Helsinki Declaration. Written informed consent was provided by all subjects.

2.2. Clinical Laboratory Tests

Detailed clinical information was collected from each patient. Body mass index (BMI), fasting blood glucose (FBG), postprandial blood glucose (PBG), fasting insulin (FINS), postprandial insulin (PINS), glycosylated haemoglobin (HbA1c), total cholesterol (TChol), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) were evaluated before and after metformin treatment as previously described [20, 21]. Homeostasis model assessment of insulin resistance (HOMA-IR), homeostasis model assessment of insulin sensitivity (HOMA-IS), homeostasis model assessment of β-cell function (HOMA-B), and quantitative insulin sensitivity check index (QUICKI) were calculated according to previous formulas [22, 23].

2.3. DNA Sampling and Genotyping

Genomic DNA was isolated from peripheral blood leukocytes using Promega DNA purification kit (Promega, Wisconsin, USA) according to the manufacturer’s protocol and stored at −20°C until use. Genotyping was done using direct sequencing. The primers used to amplify SLC22A1 rs594709 polymorphism were 5′GGCGCTTCCCACACTCAT3′ (sense) and 5′GAGGAAAGCTCCACATGTAACC3′ (antisense), and the primers for SLC47A1 rs2289669 polymorphism were 5′CCTGACCTCTGTCCCTATT3′ (sense) and 5′GACTTCCGTGCCTGCT3′ (antisense). The PCR products were then sequenced using ABI 3700 automated sequencer (ABI, Foster City, CA, USA).

2.4. Statistical Analysis

Independent-samples t-test was used to compare baseline characteristics in T2DM patients and healthy subjects. The observed genotype frequencies in patients and controls compared with expected frequencies under Hardy-Weinberg equilibrium were evaluated by χ ² test. Comparisons of frequencies of genotypes and alleles were also analyzed by χ ² test.

We evaluated associations between each SNP and the change of clinical index before and after metformin therapy using multivariate linear regression by adjusting for sex, age, and BMI in additive model, recessive model, and dominant model, respectively. Besides, the analysis for the SLC22A1 rs594709 genotype was stratified for the SLC47A1 rs2289669 genotype and vice versa. Multivariate linear regression was used to analyze the interaction between SLC22A1 rs594709 and SLC47A1 rs2289669 genotype on metformin response in Chinese T2DM patients. All of the analyses were performed with SPSS software (version 16.0; SPSS, Chicago, Illinois, USA).

3. Results

3.1. Clinical Characteristics of Subjects

A total of 267 T2DM patients and 182 healthy subjects were enrolled to explore the allele frequencies and genotype frequencies of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms. The clinical characteristics of T2DM patients and healthy subjects were shown in Table 1. BMI ( $25.10\pm 2.74$  kg/m² versus $23.98\pm 2.94$  kg/m², $p<0.001$ ), waist circumference ( $87.40\pm 8.56$  cm versus $81.72\pm 9.52$  cm, $p<0.001$ ), waist-to-hip ratio ( $0.91\pm 0.06$ versus $0.85\pm 0.07$ , $p<0.001$ ), FBG ( $8.91\pm 3.40$  mmol/L versus $5.11\pm 0.57$  mmol/L, $p<0.001$ ), and TG ( $3.12\pm 3.71$  mmol/L versus $1.85\pm 6.82$  mmol/L, $p<0.05$ ) in T2DM patients were significantly higher than those of healthy subjects. Moreover, HDL-C level ( $1.32\pm 0.64$  mmol/L versus $1.51\pm 0.45$  mmol/L, $p<0.001$ ) in T2DM patients was significantly lower than that in healthy subjects. No significant differences were found in hip circumference, TChol, and LDL-C level between two groups.

Table 1

The clinical characteristics of T2DM patients and healthy subjects.

Parameter	Healthy controls ( $n=182$ )	T2DM patients ( $n=267$ )	$p$ value
Gender
Male	86	143
Female	96	124	0.189
Age (years)	$$49.04\pm 8.10$$	$$49.82\pm 10.11$$	0.370
BMI (kg/m2)	$$23.98\pm 2.94$$	$$25.10\pm 2.74$$	$${0.000}^{\ast \ast \ast }$$
Waist circumference (cm)	$$81.72\pm 9.52$$	$$87.40\pm 8.56$$	$${0.000}^{\ast \ast \ast }$$
Hip circumference (cm)	$$95.32\pm 6.46$$	$$96.49\pm 7.36$$	$$0.097$$
Waist-hip ratio	$$0.85\pm 0.07$$	$$0.91\pm 0.06$$	$${0.000}^{\ast \ast \ast }$$
FBG (mmol/L)	$$5.11\pm 0.57$$	$$8.91\pm 3.40$$	$${0.000}^{\ast \ast \ast }$$
TG (mmol/L)	$$1.85\pm 6.82$$	$$3.12\pm 3.71$$	$${0.025}^{\ast }$$
TChol (mmol/L)	$$4.88\pm 0.97$$	$$5.05\pm 1.81$$	$$0.255$$
LDL-C (mmol/L)	$$2.82\pm 0.78$$	$$2.90\pm 1.01$$	$$0.388$$
HDL-C (mmol/L)	$$1.51\pm 0.45$$	$$1.32\pm 0.64$$	$${0.000}^{\ast \ast \ast }$$

BMI: body mass index; FBG: fasting blood glucose; TG: triglycerides; TChol: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol. ${}^{\ast }p<0.05$ , ${}^{\ast \ast \ast }p<0.001$ .

3.2. Genotype Analysis and Allelic Frequencies

The frequencies of the SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms in T2DM patients and healthy subjects were shown in Table 2. Both SNPs were in agreement with Hardy-Weinberg equilibrium in each group ( $p>0.05$ ). Minor allele frequencies (MAFs) of SLC22A1 rs594709 in T2DM group and healthy control were 26.78% and 28.57%, respectively. MAFs of SLC47A1 rs2289669 were, respectively, 47.19% and 46.97% in T2DM patients and healthy subjects. No significant differences were found in allele frequencies ( $p=0.555$ for rs594709; $p=0.950$ for rs2289669) and genotype frequencies ( $p=0.927$ for rs594709; $p=0.669$ for rs2289669) between T2DM patients and healthy controls.

Table 2

Comparisons of allelic frequencies of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms in T2DM patients and healthy subjects.

Genotype	T2DM patients ( $$n=267$$ )	Healthy controls ( $n=182$ )	$p$ value
SLC22A1 rs594709
AA	146 (54.68%)	95 (52.20%)	0.927
GA	99 (37.08%)	70 (38.46%)
GG	22 (8.24%)	17 (9.34%)
Alleles
A	391 (73.22%)	260 (71.43%)	0.555
G	143 (26.78%)	104 (28.57%)
SLC47A1 rs2289669
AA	77 (28.84%)	49 (26.92%)	0.669
GA	128 (47.94%)	95 (52.20%)
GG	62 (23.22%)	38 (20.88%)
Alleles
A	282 (52.81%)	193 (53.02%)	0.950
G	252 (47.19%)	171 (46.97%)

3.3. Influence of SLC22A1 rs594709 and SLC47A1 rs2289669 Polymorphisms on Therapeutic Efficacy of Metformin in T2DM Patients

53 newly diagnosed T2DM patients who had taken the first prescription of metformin for 3 consecutive months were enrolled to evaluate metformin efficacy. As shown in Table 3, SLC22A1 rs594709 GG genotype patients had a higher increase in FINS ( $p=0.015$ ) and greater decrease in HOMA-IS ( $p=0.001$ ) and QUICKI ( $p=0.002$ ) than those of A allele carriers (Table 3). However, SLC47A1 rs2289669 polymorphism was associated with changes in TChol and LDL-C level; patients carrying SLC47A1 rs2289669 A allele had a relatively minor decrease in TChol ( $p=0.030$ ) and LDL-C ( $p=0.049$ ) level compared to GG genotype patients (Table 3).

Table 3

Comparisons of differential values (post-administration minus pre-administration) in T2DM patients with different SLC22A1 rs594709 and SLC47A1 rs2289669 genotypes before and after metformin treatment.

Parameter	SLC22A1 rs594709		$p$ value	SLC47A1 rs2289669		$p$ value
	AA/GA ( $n=50$ )	GG ( $n=3$ )		AA/GA ( $n=37$ )	GG ( $n=16$ )
FBG (mmol/L)	$$-2.02\pm 2.84$$	$$-0.10\pm 0.62$$	$$0.112$$	$$-2.15\pm 3.11$$	$$-1.27\pm 1.72$$	0.668
PBG (mmol/L)	$$-5.28\pm 6.03$$	$$-1.73\pm 2.66$$	$$0.171$$	$$-5.61\pm 6.17$$	$$-3.84\pm 5.32$$	0.692
FINS (mU/L)	$$3.90\pm 5.40$$	$$11.35\pm 4.46$$	$$0.015{}^{\text{a}}$$	$$4.07\pm 6.06$$	$$5.19\pm 4.47$$	0.501
PINS (mU/L)	$$-9.77\pm 25.58$$	$$-20.58\pm 22.72$$	$$0.259$$	$$-13.69\pm 29.73$$	$$-3.53\pm 6.81$$	0.457
HbA1c (%)	$$-2.83\pm 4.82$$	$$-0.60\pm 0.53$$	$$0.227$$	$$-2.89\pm 5.38$$	$$-2.21\pm 2.64$$	1.000
TG (mmol/L)	$$-0.84\pm 3.71$$	$$-0.74\pm 1.28$$	$$0.434$$	$$-0.72\pm 3.71$$	$$-1.10\pm 3.44$$	0.163
TChol (mmol/L)	$$-0.60\pm 1.46$$	$$0.05\pm 1.56$$	$$0.224$$	$$-0.39\pm 1.40$$	$$-0.96\pm 1.55$$	0.030 ${}^{\text{b}}$
LDL-C (mmol/L)	$$-0.14\pm 1.00$$	$$-0.09\pm 0.78$$	$$0.451$$	$$-0.08\pm 1.10$$	$$-0.28\pm 0.63$$	0.049 ${}^{\text{b}}$
HDL-C (mmol/L)	$$-0.03\pm 0.18$$	$$-0.05\pm 0.15$$	$$0.399$$	$$-0.02\pm 0.19$$	$$-0.08\pm 0.14$$	0.254
HOMA-IR	$$0.60\pm 2.87$$	$$3.06\pm 0.72$$	$$0.081$$	$$0.55\pm 3.27$$	$$1.31\pm 1.31$$	0.357
HOMA-IS	$$-0.003\pm 0.03$$	$$-0.08\pm 0.10$$	$$0.001{}^{\text{a}}$$	$$-0.009\pm 0.05$$	$$-0.009\pm 0.02$$	0.382
HOMA-B	$$0.90\pm 1.70$$	$$0.96\pm 0.70$$	$$0.493$$	$$0.86\pm 1.48$$	$$1.01\pm 2.06$$	0.295
QUICKI	$$-0.03\pm 0.17$$	$$-0.43\pm 0.46$$	$$0.002{}^{\text{a}}$$	$$-0.05\pm 0.25$$	$$-0.07\pm 0.10$$	0.446

BMI: body mass index; FBG: fasting blood glucose; PBG: postprandial blood glucose; FINS: fasting insulin; PINS: postprandial insulin; HbA1c: glycosylated haemoglobin; TG: triglycerides; TChol: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; HOMA-IR: homeostasis model assessment of insulin resistance; HOMA-IS: homeostasis model assessment of insulin sensitivity; HOMA-B: homeostasis model assessment of β-cell function; QUICKI: quantitative insulin sensitivity check index.

${}^{\text{a}}p<0.05$ for comparison of GG genotype with GA/AA genotype of SLC22A1 rs594709 polymorphism, adjusted for age, sex, and BMI, in recessive model.

${}^{\text{b}}p<0.05$ for comparison of GG genotype with GA/AA genotype of SLC47A1 rs2289669 polymorphism, adjusted for age, sex, and BMI, in recessive model.

Additionally, we found a significant SNP-SNP interaction between SLC47A1 rs2289669 and SLC22A1 rs594709, which affected the improvement of insulin resistance and blood lipid induced by metformin therapy. Among SLC22A1 rs594709 AA genotypes, patients with SLC47A1 rs2289669 AA genotype showed significantly higher decrease in FBG ( $p=0.015$ ), PINS ( $p=0.041$ ), and HOMA-IR ( $p=0.014$ ) than GA or GG genotypes (Table 4), while, among SLC22A1 rs594709 G allele carriers, patients with SLC47A1 rs2289669 AA genotype showed a greater decrease in TChol ( $p=0.013$ ) than GA or GG genotypes (Table 4).

Table 4

The differential values (DV) of clinical characteristics per SLC22A1 rs594709 and SLC47A1 rs2289669 genotypes.

SLC47A1 rs2289669	SLC22A1 rs594709 AA ( $n=32$ )		$p$ value	SLC22A1 rs594709 GA/GG ( $n=21$ )		$p$ value
	AA ( $n=6$ )	GA + GG ( $n=26$ )		AA ( $n=7$ )	GA + GG ( $n=14$ )
FBG (mmol/L)	$$-3.94\pm 6.25$$	$$-1.63\pm 1.95$$	0.015 ${}^{\text{a}}$	$$-1.56\pm 2.61$$	$$-1.81\pm 2.37$$	0.641
PBG (mmol/L)	$$-9.63\pm 5.83$$	$$-4.56\pm 5.84$$	0.067	$$-3.32\pm 6.46$$	$$-5.32\pm 5.79$$	0.310
FINS (mU/L)	$$5.17\pm 6.89$$	$$4.71\pm 4.28$$	0.252	$$3.25\pm 8.35$$	$$4.17\pm 6.41$$	0.865
PINS (mU/L)	$$-31.68\pm 56.56$$	$$-4.61\pm 20.91$$	0.041 ${}^{\text{a}}$	$$-13.72\pm 25.71$$	$$-14.93\pm 22.25$$	0.799
HbA1c (%)	$$-3.25\pm 3.11$$	$$-3.33\pm 6.30$$	0.487	$$-1.84\pm 1.43$$	$$-1.76\pm 2.13$$	0.778
TG (mmol/L)	$$-4.15\pm 6.97$$	$$-1.05\pm 2.70$$	0.061	$$-0.77\pm 2.20$$	$$0.69\pm 4.05$$	0.466
TChol (mmol/L)	$$-1.28\pm 1.56$$	$$-0.65\pm 1.46$$	0.197	$$-1.33\pm 1.15$$	$$0.32\pm 1.26$$	0.013 ${}^{\text{b}}$
LDL-C (mmol/L)	$$0.38\pm 1.35$$	$$-0.09\pm 0.76$$	0.148	$$-0.66\pm 0.49$$	$$-0.11\pm 1.38$$	0.228
HDL-C (mmol/L)	$$0.09\pm 0.25$$	$$-0.04\pm 0.20$$	0.134	$$-0.10\pm 0.15$$	$$-0.03\pm 0.11$$	0.205
HOMA-IR	$$-2.02\pm 7.50$$	$$1.22\pm 1.36$$	0.014 ${}^{\text{a}}$	$$0.34\pm 2.33$$	$$1.10\pm 2.77$$	0.841
HOMA-IS	$$-0.01\pm 0.02$$	$$-0.01\pm 0.02$$	0.440	$$-0.02\pm 0.08$$	$$0.003\pm 0.05$$	0.345
HOMA-B	$$1.11\pm 0.74$$	$$0.81\pm 1.57$$	0.441	$$1.62\pm 2.90$$	$$0.57\pm 0.86$$	0.149
QUICKI	$$-0.07\pm 0.13$$	$$-0.07\pm 0.12$$	0.447	$$-0.10\pm 0.42$$	$$0.0003\pm 0.25$$	0.437

BMI: body mass index; FBG: fasting blood glucose; PBG: postprandial blood glucose; FINS: fasting insulin; PINS: postprandial insulin; HbA1c: glycosylated haemoglobin; TG: triglycerides; TChol: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; HOMA-IR: homeostasis model assessment of insulin resistance; HOMA-IS: homeostasis model assessment of insulin sensitivity; HOMA-B: homeostasis model assessment of β-cell function; QUICKI: quantitative insulin sensitivity check index.

${}^{\text{a}}p<0.05$ for comparison of AA genotype with GA/GG genotype of SLC47A1 rs2289669 polymorphism in SLC22A1 rs594709 AA genotyped T2DM patients, adjusted for age, sex, and BMI, in dominant model.

${}^{\text{b}}p<0.05$ for comparison of AA genotype with GA/GG genotype of SLC47A1 rs2289669 polymorphism in SLC22A1 rs594709 GA or GG genotyped T2DM patients, adjusted for age, sex, and BMI, in dominant model.

4. Discussion

In the current study, we evaluated the distributive characteristics of SLC22A1 rs594709 and SLC47A1 rs2289669 polymorphisms and the separate as well as combined influence of their polymorphisms on metformin’s effect in Chinese T2DM patients. Our results showed that there were no significant differences in SLC22A1 rs594709 and SLC47A1 rs2289669 allele frequencies and genotype frequencies between T2DM patients and healthy subjects. SLC22A1 rs594709 A allele carriers showed a better metformin efficacy in insulin sensitivity compared with GG genotype patients. SLC47A1 rs2289669 GG genotype patients showed a better efficacy in improving blood lipid. Furthermore, we found that there was a significant interaction between SLC47A1 rs2289669 and SLC22A1 rs594709, which might function together in metformin efficacy.

OCTs mediate the interactions between cells and their environment and are involved in transporting a broad spectrum of endogenous substrates and the detoxification of xenobiotics [24]. Nutrient transporters, including OCT1 and MATE1, could affect nutrient homeostasis through different ways [25]. Polymorphisms in OCT1 or MATE1 have been correlated with the susceptibility of type 2 diabetes, diabetic nephropathy, hypertension, and primary biliary cirrhosis [26–28]. SLC47A1 rs2453583 and SLC22A1 rs651154 have been reported to be nominally associated with diabetes incidence in the DPP (Diabetes Prevention Program) [26]. Therefore, we investigated the distributive characteristics of the two SNPs that we are concerned with, namely, SLC22A1 rs594709 and SLC47A1 rs2289669, in T2DM patients and healthy subjects. However, no statistical difference was found between two groups.

Human SLC22A1 gene is highly polymorphic, and several polymorphisms have been described to affect metformin responses [29, 30], including polymorphisms located in the intron regions [31]. SLC22A1 rs594709 locates in intron 4. Pairwise linkage disequilibrium (LD) analysis indicated that there were strong LDs between rs594709 and rs2282143, rs628031, rs3798168, and rs683369. Polymorphisms mentioned above could more or less affect the metabolizing and/or efficacy of OCT1 substrates. SLC22A1 rs2282143 variation led to a decrease of transport activity, but not completely abolished [32]. SLC22A1 rs628031 (Met408Val) was identified in multiethnic groups worldwide [33] and has been characterized to have normal metformin uptake in vitro [9, 11]. However, it has been recognized as a weak positive predictor of metformin response in clinical study [8, 34, 35], though this genetic variant did not alter protein expression [34, 36]. Recent studies reported that SLC22A1 rs3798168, rs683369, and rs628031 could affect the response to imatinib therapy in chronic myeloid leukemia (CML) patients [37, 38]. As SLC22A1 rs594709 is the tagger SNP of rs3798168, rs683369, and rs628031 ( $r^2>0.8$ ), it is possible that the structural changes in SLC22A1 caused by any of these polymorphisms or even a combination effect of them may affect its transport ability of metformin.

Previously, Caucasians population-based study in diabetic patients revealed that rs2289669 G/A in SLC47A1 gene was associated with the therapeutic efficacy of metformin [16, 17]. The decrease in HbA1c was greater in AA homozygotes compared with the GA + GG genotype patients after metformin treatment. The effect of SLC47A1 rs2289669 polymorphism on pharmacodynamics ( $n=202$ ) and pharmacokinetics ( $n=28$ ) of metformin was also evaluated in Chinese patients and showed that AA genotype patients had a better glucose lowering effect after one-year follow-up. As for pharmacokinetics parameters, AA genotype patients had higher ${\mathrm{A}\mathrm{U}\mathrm{C}}_{\mathrm{12}\mathrm{h}}$ and lower renal clearance and renal clearance by secretion [15]. Thus, it is possible that the effect of the rs2289669 polymorphism on metformin glucose lowering efficacy was caused partly or completely by a reduced renal excretion and increased metformin plasma levels [39]. However, in our study, we did not find SLC47A1 rs2289669 polymorphism was correlated with metformin glucose lowering effect but was associated with the reduction of TChol and LDL-C. This result was similar to long-term metformin and sulphonylurea combination therapy study carried out in Slovenia [40]. Their results suggested that SLC47A1 rs2289669 genotype was not associated with HbA1c levels but significantly associated with cholesterol levels in metformin-treated T2DM patients even after adjustment for renal function. In contrast to the previous studies, the sample size, period of metformin treatment, and different races were all possible reasons leading to different results. Replication studies in a relatively large population in a multi-ethnic diabetic cohort are needed to confirm SLC47A1 rs2289669 polymorphism effect on metformin efficacy both in glycemic control and in lipid profiles.

Furthermore, we found that there was an interaction between SLC47A1 rs2289669 and SLC22A1 rs594709, which affected the blood glucose, insulin level, insulin resistance improvement, and blood lipid after metformin treatment. Among SLC22A1 rs594709 AA genotype patients, SLC47A1 rs2289669 AA genotype patients showed a better glycemic control compared with G allele carriers, while, among SLC22A1 rs594709 G allele carriers, patients with SLC47A1 rs2289669 AA genotype showed a better metformin efficacy in blood lipid than GA or GG genotypes. Interaction between SLC22A1 and SLC47A1 polymorphisms and metformin response had been reported previously [18]; the effect of the SLC47A1 rs2289669 polymorphism on the decrease of HbA1c by metformin was larger in incident users with the SLC22A1 rs622342 CC genotype than in incident users with the AA or AC genotype patients. Both of the rs622342 and rs594709 are located in intron regions of SLC22A1 gene. The interaction is possibly due to the genetic mutations in SLC22A1 and SLC47A1 altering the function of OCT1-mediated influx and MATE1-mediated efflux. As OCT1 transports metformin into the hepatocyte and MATE1 transports metformin out of the hepatocyte into the bile, altered OCT1 and MATE1 function may be able to influence the intracellular metformin concentrations and glucose lowering function of metformin. Further replication and mechanism studies are needed to illustrate the influences of SLC22A1 and SLC47A1 polymorphisms on metformin efficacy.

In summary, we found that SLC22A1 594709 and SLC47A1 rs2289669 polymorphisms may interact to affect the therapeutic efficacy of metformin in Chinese T2DM patients. From our data, T2DM patients with AA genotype of both SLC22A1 rs594709 and SLC47A1 rs2289669 may get maximum glucose lowering benefit from metformin monotherapy. The SLC22A1 and SLC47A1 gene-gene interaction further suggests the complexity of metformin efficacy in T2DM patients.

Acknowledgments

The authors thank the study participants and supported grants of the National High-Tech R & D Program of China (863 Program) (2009AA022704, 2012AA02A517).

References

[1] G. Anfossi, I. Russo, K. Bonomo, M. Trovati, "The cardiovascular effects of metformin: further reasons to consider an old drug as a cornerstone in the therapy of type 2 diabetes mellitus," Current Vascular Pharmacology, vol. 8 no. 3, pp. 327-337, DOI: 10.2174/157016110791112359, 2010.

[2] X. Stephenne, M. Foretz, N. Taleux, G. C. van der Zon, E. Sokal, L. Hue, B. Viollet, B. Guigas, "Metformin activates AMP-activated protein kinase in primary human hepatocytes by decreasing cellular energy status," Diabetologia, vol. 54 no. 12, pp. 3101-3110, DOI: 10.1007/s00125-011-2311-5, 2011.

[3] R. A. Miller, Q. Chu, J. Xie, M. Foretz, B. Viollet, M. J. Birnbaum, "Biguanides suppress hepatic glucagon signalling by decreasing production of cyclic AMP," Nature, vol. 494 no. 7436, pp. 256-260, DOI: 10.1038/nature11808, 2013.

[4] M. D. Fullerton, S. Galic, K. Marcinko, S. Sikkema, T. Pulinilkunnil, Z.-P. Chen, H. M. O'Neill, R. J. Ford, R. Palanivel, M. O'Brien, D. G. Hardie, S. L. MacAulay, J. D. Schertzer, J. R. B. Dyck, B. J. Van Denderen, B. E. Kemp, G. R. Steinberg, "Single phosphorylation sites in Acc1 and Acc2 regulate lipid homeostasis and the insulin-sensitizing effects of metformin," Nature Medicine, vol. 19 no. 12, pp. 1649-1654, DOI: 10.1038/nm.3372, 2013.

[5] A. K. Madiraju, D. M. Erion, Y. Rahimi, X.-M. Zhang, D. T. Braddock, R. A. Albright, B. J. Prigaro, J. L. Wood, S. Bhanot, M. J. MacDonald, M. J. Jurczak, J.-P. Camporez, H.-Y. Lee, G. W. Cline, V. T. Samuel, R. G. Kibbey, G. I. Shulman, "Metformin suppresses gluconeogenesis by inhibiting mitochondrial glycerophosphate dehydrogenase," Nature, vol. 510 no. 7506, pp. 542-546, DOI: 10.1038/nature13270, 2014.

[6] F. A. Duca, C. D. Côté, B. A. Rasmussen, M. Zadeh-Tahmasebi, G. A. Rutter, B. M. Filippi, T. K. T. Lam, "Metformin activates a duodenal Ampk-dependent pathway to lower hepatic glucose production in rats," Nature Medicine, vol. 21 no. 5, pp. 506-511, DOI: 10.1038/nm.3787, 2015.

[7] S. E. Inzucchi, R. M. Bergenstal, J. B. Buse, M. Diamant, E. Ferrannini, M. Nauck, A. L. Peters, A. Tsapas, R. Wender, D. R. Matthews, "Management of hyperglycemia in type 2 diabetes, 2015: a patient-centered approach: update to a position statement of the American Diabetes Association and the European Association for the Study of Diabetes," Diabetes Care, vol. 38 no. 1, pp. 140-149, DOI: 10.2337/dc14-2441, 2015.

[8] L. Tarasova, I. Kalnina, K. Geldnere, A. Bumbure, R. Ritenberga, L. Nikitina-Zake, D. Fridmanis, I. Vaivade, V. Pirags, J. Klovins, "Association of genetic variation in the organic cation transporters OCT1, OCT2 and multidrug and toxin extrusion 1 transporter protein genes with the gastrointestinal side effects and lower BMI in metformin-treated type 2 diabetes patients," Pharmacogenetics and Genomics, vol. 22 no. 9, pp. 659-666, DOI: 10.1097/FPC.0b013e3283561666, 2012.

[9] Y. Shu, S. A. Sheardown, C. Brown, R. P. Owen, S. Zhang, R. A. Castro, A. G. Ianculescu, L. Yue, J. C. Lo, E. G. Burchard, C. M. Brett, K. M. Giacomini, "Effect of genetic variation in the organic cation transporter 1 (OCT1) on metformin action," The Journal of Clinical Investigation, vol. 117 no. 5, pp. 1422-1431, DOI: 10.1172/jci30558, 2007.

[10] G. G. Graham, J. Punt, M. Arora, R. O. Day, M. P. Doogue, J. K. Duong, T. J. Furlong, J. R. Greenfield, L. C. Greenup, C. M. Kirkpatrick, J. E. Ray, P. Timmins, K. M. Williams, "Clinical pharmacokinetics of metformin," Clinical Pharmacokinetics, vol. 50 no. 2, pp. 81-98, DOI: 10.2165/11534750-000000000-00000, 2011.

[11] Y. Shu, C. Brown, R. A. Castro, R. J. Shi, E. T. Lin, R. P. Owen, S. A. Sheardown, L. Yue, E. G. Burchard, C. M. Brett, K. M. Giacomini, "Effect of genetic variation in the organic cation transporter 1, OCT1, on metformin pharmacokinetics," Clinical Pharmacology and Therapeutics, vol. 83 no. 2, pp. 273-280, DOI: 10.1038/sj.clpt.6100275, 2008.

[12] W. Hou, D. Zhang, W. Lu, T. Zheng, L. Wan, Q. Li, Y. Bao, F. Liu, W. Jia, "Polymorphism of organic cation transporter 2 improves glucose-lowering effect of metformin via influencing its pharmacokinetics in chinese type 2 diabetic patients," Molecular Diagnosis & Therapy, vol. 19 no. 1, pp. 25-33, DOI: 10.1007/s40291-014-0126-z, 2015.

[13] H. Yoon, H.-Y. Cho, H.-D. Yoo, S.-M. Kim, Y.-B. Lee, "Influences of organic cation transporter polymorphisms on the population pharmacokinetics of metformin in healthy subjects," The AAPS Journal, vol. 15 no. 2, pp. 571-580, DOI: 10.1208/s12248-013-9460-z, 2013.

[14] A. Giannoudis, L. Wang, A. L. Jorgensen, G. Xinarianos, A. Davies, S. Pushpakom, T. Liloglou, J.-E. Zhang, G. Austin, T. L. Holyoake, L. Foroni, P. D. Kottaridis, M. C. Müller, M. Pirmohamed, R. E. Clark, "The hOCT1 SNPs M420del and M408V alter imatinib uptake and M420del modifies clinical outcome in imatinib-treated chronic myeloid leukemia," Blood, vol. 121 no. 4, pp. 628-637, DOI: 10.1182/blood-2012-01-405035, 2013.

[15] R. He, D. Zhang, W. Lu, T. Zheng, L. Wan, F. Liu, W. Jia, "A variants enhance the glucose-lowering effect of metformin via delaying its excretion in Chinese type 2 diabetes patients," Diabetes Research and Clinical Practice, vol. 109 no. 1, pp. 57-63, DOI: 10.1016/j.diabres.2015.05.003, 2015.

[16] M. L. Becker, L. E. Visser, R. H. N. van Schaik, A. Hofman, A. G. Uitterlinden, B. H. C. Stricker, "Genetic variation in the multidrug and toxin extrusion 1 transporter protein influences the glucose-lowering effect of metformin in patients with diabetes: a preliminary study," Diabetes, vol. 58 no. 3, pp. 745-749, DOI: 10.2337/db08-1028, 2009.

[17] I. Tkáč, L. Klimčáková, M. Javorský, M. Fabianová, Z. Schroner, H. Hermanová, E. Babjaková, R. Tkáčová, "Pharmacogenomic association between a variant in SLC47A1 gene and therapeutic response to metformin in type 2 diabetes," Diabetes, Obesity and Metabolism, vol. 15 no. 2, pp. 189-191, DOI: 10.1111/j.1463-1326.2012.01691.x, 2013.

[18] M. L. Becker, L. E. Visser, R. H. N. van Schaik, A. Hofman, A. G. Uitterlinden, B. H. C. Stricker, "Interaction between polymorphisms in the OCT1 and MATE1 transporter and metformin response," Pharmacogenetics and Genomics, vol. 20 no. 1, pp. 38-44, DOI: 10.1097/fpc.0b013e328333bb11, 2010.

[19] M. M. H. Christensen, R. S. Pedersen, T. B. Stage, C. Brasch-Andersen, F. Nielsen, P. Damkier, H. Beck-Nielsen, K. Brøsen, "A gene-gene interaction between polymorphisms in the OCT2 and MATE1 genes influences the renal clearance of metformin," Pharmacogenetics and Genomics, vol. 23 no. 10, pp. 526-534, DOI: 10.1097/FPC.0b013e328364a57d, 2013.

[20] H.-L. Liu, Y.-G. Lin, J. Wu, H. Sun, Z.-C. Gong, P.-C. Hu, J.-Y. Yin, W. Zhang, D. Wang, H.-H. Zhou, Z.-Q. Liu, "Impact of genetic polymorphisms of leptin and TNF-alpha on rosiglitazone response in Chinese patients with type 2 diabetes," European Journal of Clinical Pharmacology, vol. 64 no. 7, pp. 663-671, DOI: 10.1007/s00228-008-0483-9, 2008.

[21] H. Sun, Z.-C. Gong, J.-Y. Yin, H.-L. Liu, Y.-Z. Liu, Z.-W. Guo, H.-H. Zhou, J. Wu, Z.-Q. Liu, "The association of adiponectin allele 45T/G and -11377C/G polymorphisms with type 2 diabetes and rosiglitazone response in Chinese patients," British Journal of Clinical Pharmacology, vol. 65 no. 6, pp. 917-926, DOI: 10.1111/j.1365-2125.2008.03145.x, 2008.

[22] A. Katz, S. S. Nambi, K. Mather, A. D. Baron, D. A. Follmann, G. Sullivan, M. J. Quon, "Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans," Journal of Clinical Endocrinology and Metabolism, vol. 85 no. 7, pp. 2402-2410, DOI: 10.1210/jc.85.7.2402, 2000.

[23] D. R. Matthews, J. P. Hosker, A. S. Rudenski, B. A. Naylor, D. F. Treacher, R. C. Turner, "Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man," Diabetologia, vol. 28 no. 7, pp. 412-419, DOI: 10.1007/bf00280883, 1985.

[24] A. Lautem, M. Heise, A. Gräsel, M. Hoppe-Lotichius, N. Weiler, D. Foltys, J. Knapstein, J. M. Schattenberg, A. Schad, A. Zimmermann, G. Otto, H. Lang, P. R. Galle, M. Schuchmann, T. Zimmermann, "Downregulation of organic cation transporter 1 (SLC22A1) is associated with tumor progression and reduced patient survival in human cholangiocellular carcinoma," International Journal of Oncology, vol. 42 no. 4, pp. 1297-1304, DOI: 10.3892/ijo.2013.1840, 2013.

[25] Y. Taketani, H. Yamanaka-Okumura, H. Yamamoto, E. Takeda, "Role of nutrient transporters in lifestyle-related diseases," The Journal of Physical Fitness and Sports Medicine, vol. 2 no. 4, pp. 409-416, DOI: 10.7600/jpfsm.2.409, 2013.

[26] K. A. Jablonski, J. B. McAteer, P. I. W. de Bakker, P. W. Franks, T. I. Pollin, R. L. Hanson, R. Saxena, S. Fowler, A. R. Shuldiner, W. C. Knowler, D. Altshuler, J. C. Florez, "Common variants in 40 genes assessed for diabetes incidence and response to metformin and lifestyle intervention in the diabetes prevention program," Diabetes, vol. 59 no. 10, pp. 2672-2681, DOI: 10.2337/db10-0543, 2010.

[27] R. Sallinen, M. A. Kaunisto, C. Forsblom, M. Thomas, J. Fagerudd, K. Pettersson-Fernholm, P.-H. Groop, M. Wessman, "Association of the SLC22A1, SLC22A2, and SLC22A3 genes encoding organic cation transporters with diabetic nephropathy and hypertension," Annals of Medicine, vol. 42 no. 4, pp. 296-304, DOI: 10.3109/07853891003777109, 2010.

[28] Y. Ohishi, M. Nakamuta, N. Ishikawa, O. Saitoh, H. Nakamura, Y. Aiba, A. Komori, K. Migita, H. Yatsuhashi, N. Fukushima, M. Kohjima, T. Yoshimoto, K. Fukuizumi, M. Ishibashi, T. Nishino, K. Shirabe, A. Taketomi, Y. Maehara, H. Ishibashi, M. Nakamura, "Genetic polymorphisms of OCT-1 confer susceptibility to severe progression of primary biliary cirrhosis in Japanese patients," Journal of Gastroenterology, vol. 49 no. 2, pp. 332-342, DOI: 10.1007/s00535-013-0795-0, 2014.

[29] H. Takane, E. Shikata, K. Otsubo, S. Higuchi, I. Ieiri, "Polymorphism in human organic cation transporters and metformin action," Pharmacogenomics, vol. 9 no. 4, pp. 415-422, DOI: 10.2217/14622416.9.4.415, 2008.

[30] M.-K. Choi, I.-S. Song, "Organic cation transporters and their pharmacokinetic and pharmacodynamic consequences," Drug Metabolism and Pharmacokinetics, vol. 23 no. 4, pp. 243-253, DOI: 10.2133/dmpk.23.243, 2008.

[31] M. L. Becker, L. E. Visser, R. H. N. van Schaik, A. Hofman, A. G. Uitterlinden, B. H. C. Stricker, "Genetic variation in the organic cation transporter 1 is associated with metformin response in patients with diabetes mellitus," Pharmacogenomics Journal, vol. 9 no. 4, pp. 242-247, DOI: 10.1038/tpj.2009.15, 2009.

[32] A. Takeuchi, H. Motohashi, M. Okuda, K. Inui, "Decreased function of genetic variants, Pro283Leu and Arg287Gly, in human organic cation transporter hOCT1," Drug Metabolism and Pharmacokinetics, vol. 18 no. 6, pp. 409-412, DOI: 10.2133/dmpk.18.409, 2003.

[33] H.-J. Kang, I.-S. Song, H. J. Shin, W.-Y. Kim, C.-H. Lee, J.-C. Shim, H.-H. Zhou, S. L. Sang, J.-G. Shin, "Identification and functional characterization of genetic variants of human organic cation transporters in a Korean population," Drug Metabolism and Disposition, vol. 35 no. 4, pp. 667-675, DOI: 10.1124/dmd.106.013581, 2007.

[34] A. T. Nies, H. Koepsell, S. Winter, O. Burk, K. Klein, R. Kerb, U. M. Zanger, D. Keppler, M. Schwab, E. Schaeffeler, "Expression of organic cation transporters OCT1 (SLC22A1) and OCT3 (SLC22A3) is affected by genetic factors and cholestasis in human liver," Hepatology, vol. 50 no. 4, pp. 1227-1240, DOI: 10.1002/hep.23103, 2009.

[35] E. Shikata, R. Yamamoto, H. Takane, C. Shigemasa, T. Ikeda, K. Otsubo, I. Ieiri, "Human organic cation transporter (OCT1 and OCT2) gene polymorphisms and therapeutic effects of metformin," Journal of Human Genetics, vol. 52 no. 2, pp. 117-122, DOI: 10.1007/s10038-006-0087-0, 2007.

[36] M. V. Tzvetkov, S. V. Vormfelde, D. Balen, I. Meineke, T. Schmidt, D. Sehrt, I. Sabolić, H. Koepsell, J. Brockmöller, "The effects of genetic polymorphisms in the organic cation transporters OCT1, OCT2, and OCT3 on the renal clearance of metformin," Clinical Pharmacology and Therapeutics, vol. 86 no. 3, pp. 299-306, DOI: 10.1038/clpt.2009.92, 2009.

[37] D. H. Kim, L. Sriharsha, W. Xu, S. Kamel-Reid, X. Liu, K. Siminovitch, H. A. Messner, J. H. Lipton, "Clinical relevance of a pharmacogenetic approach using multiple candidate genes to predict response and resistance to imatinib therapy in chronic myeloid leukemia," Clinical Cancer Research, vol. 15 no. 14, pp. 4750-4758, DOI: 10.1158/1078-0432.ccr-09-0145, 2009.

[38] O. Singh, J. Y. Chan, K. Lin, C. C. T. Heng, B. Chowbay, "SLC22A1-ABCB1 haplotype profiles predict imatinib pharmacokinetics in Asian patients with chronic myeloid leukemia," PLoS ONE, vol. 7 no. 12,DOI: 10.1371/journal.pone.0051771, 2012.

[39] M. Tsuda, T. Terada, T. Mizuno, T. Katsura, J. Shimakura, K.-I. Inui, "Targeted disruption of the multidrug and toxin extrusion 1 (Mate1) gene in mice reduces renal secretion of metformin," Molecular Pharmacology, vol. 75 no. 6, pp. 1280-1286, DOI: 10.1124/mol.109.056242, 2009.

[40] J. Klen, K. Goričar, A. Janež, V. Dolžan, "The role of genetic factors and kidney and liver function in glycemic control in type 2 diabetes patients on long-term metformin and sulphonylurea cotreatment," BioMed Research International, vol. 2014,DOI: 10.1155/2014/934729, 2014.