The results of this study have important implications for the role of OCT3 in response to the anti-diabetic drug, metformin. Metformin is widely used as a first-line therapy for the treatment of type 2 diabetes (33
). The action of metformin appears to be related to its activation (phosphorylation) of the so-called energy sensor, AMP-activated protein kinase (AMPK), which results in suppression of glucagon-stimulated glucose production and enhancement of glucose uptake in muscle and in hepatic cells (21
). In hepatic cells, we previously demonstrated that the OCT3 paralog, OCT1, is a determinant of metformin activity in the liver, a major site of action of the drug (13
). However, though metformin is known to have action in skeletal muscle (34
), the transporter(s) responsible for metformin uptake and action in skeletal muscle is largely not known. Data in the current study suggests that OCT3, which unlike OCT1/2, has a broader expression pattern, is a one of the determinants of metformin action in skeletal muscle. First, we confirmed that metformin is a substrate of OCT3, which was also recently shown by Anne et al.(18
). Further, our quantitative RT-PCR studies revealed that OCT3 is expressed at higher levels than OCT1/2 in muscle type tissues like skeletal muscle and heart. Importantly, previous studies had demonstrated that in skeletal muscle, metformin significantly increased AMPK α2 activity by increasing phosphorylation of AMPK at Thr172 (34
). In this study we showed that OCT3 is an important determinant of the effects of metformin in skeletal muscle. That is, the effect of metformin on phosphorylation of AMPK in cultured skeletal muscle cells was not only greatly inhibited by cimetidine (38
), but also was substantially inhibited by OCT3 shRNA suggesting that OCT3 plays a major role in the therapeutic action of metformin. Both the RT-PCR and immuostaining showed that OCT3 was expressed at high levels in cardiac myocytes. OCT3 might play an important role for the metformin uptake in heart. The MPP+ accumulation in heart of Orct3/Slc22a3-deficient mice decrease more than 70% (3
) and congestive heart failure has been included as a contraindication to metformin therapy (39
). Collectively, our data suggest that OCT3 is an important determinant of the peripheral effects of metformin.
Previously, we demonstrated that coding region variants of OCT1 play a role response to metformin by controlling access of the drug to AMPK in the liver (13
). To determine whether genetic variants of OCT3 may also play a role in the action of metformin, we identified and functionally characterized coding region variants of OCT3. Using heterologous expression of amino acid-altering variants of OCT3, we discovered that 3 of the 6 variants had significantly altered function with respect to metformin uptake as well as other selected endogenous amines, i.e., T44M, T400I and V423F. Studies of protein expression and subcellular localization revealed that the amino-acid substitutions, T400I and V423F, did not appreciably affect expression or subcellular localization of these variants, suggesting that the impairment of transport function may result from a disruption in the structure of OCT3. T400 is highly conserved across all the species and V423 is partially conserved in mammalian OCT1 (). The T400I and V423F variants showed obvious substrate selectivity with respect to monoamines. In particular, these two variants exhibited a substantial reduction in the transport of metformin and catecholamines in comparison to the other monoamines. The main structural difference among the various monoamines is the hydroxyl group in the phenyl ring, which is only present in the catecholamines (). Changing from a threonine to an isoleucine by T400I may disrupt the hydrogen bonding of the catecholamines to the hydroxyl group of threonine. A larger hydrophobic replacement by phenylalanine at V423 could interfere with the hydrophobic interactions between the phenol ring of the catecholamines and V400F. We created two comparative models for OCT3 based on the crystal structures of the glycerol-3-phosphate transporter and the lactose permease (LacY) from E. coli
(Fig. 3S, Supplemental Digital Content 1
). The templates have similar fold and belong to the Major Facilitator Superfamily. Both models place T400 in the 8th
transmembrane helix while V423 is plaed in the 9th
(LacY, 2CFQ) and 10th
(GlpT, 1PW4) transmembrane helix, respectively, in close proximity to the extracellular loop between the 9th and the 10th transmembrane helices. This is in agreement with a previous study describing a comparative model of rOCT1 based on a structure of the lactose permease LacY from E. coli
). Helix 8th
is predicted to be one of the helixes (H2, H5, H8 and H11) lining the substrate-transporting pore (26
), suggesting that it may interact with substrates of OCT3. However, the possibility that the amino acid residue substitutions may alter the structure of OCT3 and indirectly affect the interaction of substrates with the transporter cannot be excluded. The effect of V423F on the substrate selectivity, which was not among the helixes lining the pore, might be indirect and remain elusive (Fig. 3S, Supplemental Digital Content 1
). Further information including a high resolution crystal structure of a mammalian organic cation transporter is clearly needed to identify residues involved in the interaction of various substrates with OCTs.
The endogenous role of OCT3 may be versatile due to its role in the uptake of multiple monoamines (8
). The uptake data showed OCT3 especially favors histamine over other monoamines. The discovery of genetic variants with functional changes may have some implications for the regulation of tissue levels of endogenous substrates such as histamine as well as to the pharmacologic action of the important anti-diabetic drug, metformin. The study suggests that in addition to OCT1, OCT2 and MATE1 (6
), OCT3 should be considered as an important candidate gene for the uptake of metformin in muscle type cells and its variation may modulate the response to metformin.