Metformin is a first-line oral anti-diabetic agent that has been used clinically to treat patients with type 2 diabetes for over 60 years. Due to its efficacy in therapy and affordable price, metformin is taken by more than 150 million people each year. Metformin improves hyperglycemia mainly through the suppression of hepatic gluconeogenesis along with the improvement of insulin signaling. However, its mechanism of action remains partially understood and controversial, especially in regard to the role of AMPK in metformin's action and the mechanism of AMPK activation. In this review, we discuss recent advances in the understanding of metformin's suppression of hepatic glucose production and the mechanism related to the improvement of insulin signaling.
Metformin, a time-proven effective agent for the treatment of patients with type 2 diabetes
In the medieval Europe, the Gallega officinalis plant was used as a folk medicine to treat diabetes. In the late 1800s, the plant was found to be rich in guanidine, then a subsequent animal study in 1918 showed that guanidine had a blood glucose-lowering effect (Watanabe 1918). However, guanidine proved to be too toxic for clinical use. In 1929, the dimethylbiguanide, including phenformin and metformin, was synthesized. These compounds preserve the glucose-lowering effect of their parent compound with reduced toxicity. Phenformin is more potent than metformin in improving hyperglycemia, but the high occurrence of lactic acidosis, a serious side effect, led to the withdrawal of phenformin from the market in the 1970s.
After over 60 years of clinical use for the treatment of type 2 diabetes (T2D), metformin has proven to be safe along with affordable, now making it the most commonly prescribed oral anti-diabetic agent worldwide, taken by more than 150 million people each year. The guidelines for the treatment of T2D, published by the American Diabetes Association and the European Association for the Study of Diabetes in 2012, recommended metformin as the initial drug for T2D treatment (Inzucchi et al. 2012). Moreover, many studies have shown a reduction in cancer incidence in patients with T2D treated with metformin (Evans et al. 2005, Landman et al. 2010).
Metformin alleviates hyperglycemia in T2D mainly through the direct suppression of hepatic glucose production
T2D accounts for more than 90% of diabetes cases (Centers for Disease Control and Prevention 2011), and increased hepatic glucose production is the major cause of fasting hyperglycemia in these patients (Magnusson et al. 1992, Saltiel & Kahn 2001, Wajngot et al. 2001). It is now clear that metformin improves hyperglycemia mainly through the suppression of gluconeogenesis in the liver. In a human study, treatment of patients with T2D with metformin led to a decrease in hepatic glucose output via the suppression of gluconeogenesis by 37% (Stumvoll et al. 1995). In another human study, patients with diabetes exhibited a twofold increase in the rate of gluconeogenesis, but metformin administration led to a decrease in gluconeogenesis by 33% (Hundal et al. 2000). In animal studies, metformin reduced hepatic glucose production by over 60% in a euglycemic-clamp study (Takashima et al. 2010) and led to 50% suppression of endogenous glucose production in high-fat-diet-fed rats (Song et al. 2001). Furthermore, we have found that metformin suppressed cAMP-stimulated glucose production by 35% in primary hepatocytes (Cao et al. 2014). These in vivo and in vitro data suggest that metformin acts mainly through the suppression of hepatic gluconeogenesis.
Metformin suppresses hepatic glucose production through AMPK activation
AMPK: a master regulator of cellular energy homeostasis
AMPK was first named and reported in 1989 for its AMP-activation property and its role in regulating three key enzymes in lipid metabolism (Brown et al. 1975, Hardie et al. 1989). Over the past decades, numerous studies have confirmed the function of AMPK as a principal energy sensor of the cell and its role in regulating cellular and whole-body energy balance by sensing and responding to changes in the AMP/ADP concentration relative to ATP (Hardie et al. 2012). AMPK is activated when cellular energy levels are low, prompting a switch from ATP-consuming anabolic pathways to ATP-producing catabolic pathways by stimulating glucose uptake and utilization and fatty acid oxidation together with suppression of hepatic glucose production.
Phylogenetically, AMPK is a highly conserved serine/threonine kinase and presents in virtually all eukaryotes. Functional AMPK is a heterotrimeric complex comprised of an α catalytic subunit and regulatory βγ non-catalytic subunits, with each subunit having several isoforms (α1, α2, β1, β2, γ1, γ2, γ3, encoded by separated genes) (Hardie et al. 2012). The catalytic α-subunit consists of an N-terminal protein kinase domain (conventional Ser/Thr kinase domain), followed by an auto-inhibitory domain, a linker domain, and a C-terminal regulatory domain (α-CTD) (Chen et al. 2009, Hardie et al. 2012). The β-subunits contain a carbohydrate-binding module (CBM) and a C-terminal domain (β-CTD). The CBM domain allows AMPK to act as a glycogen sensor, while the β-CTD works as a scaffold, interacting with α-CTD and the γ-subunit, thus forming the core of the heterotrimeric complex. The γ-subunit has a tandem of four cystathionine-β-synthase domains, functioning as the adenine nucleotide-binding region (Hardie et al. 2012). When the cellular energy status changes, an increase in the AMP/ATP or ADP/ATP ratio leads to AMP or ADP binding to the γ-subunit. This results in an allosteric change in AMPK and augments α-subunit phosphorylation at T172 by an upstream kinase or prevents dephosphorylation by a protein phosphatase (Corton et al. 1994, Xiao et al. 2011). Phosphorylation of the α-subunit at T172 leads to an increase in AMPK enzymatic activity of over several hundredfold. In comparison, AMP only causes a two- to fivefold allosteric activation of AMPK, which also depends on the nature of the isoforms present in the AMPK complex (Xiao et al. 2011).
In mammals, the upstream kinases for AMPK α-subunit phosphorylation at T172 are the LKB1–STRAD–MO25 complex (Lizcano et al. 2004), the Ca2+/calmodulin-activated protein kinases, especially CaMKKβ (also termed as CaMKK2) (Hawley et al. 2005), or the transforming growth factor-β activated protein kinase-1 (Momcilovic et al. 2006). The LKB1/AMPK pathway plays a major role in metabolic regulation.
The role of AMPK in mediating metformin action
After oral administration, metformin is absorbed into enterocytes and then delivered directly to the liver through the portal vein. Plasma concentrations in the portal vein are between 40 and 70 μM in animals after a therapeutic dose (Wilcock & Bailey 1994, He & Wondisford 2015). Over a decade ago, Zhou et al. (2001) reported that metformin at concentrations found in the portal vein (≤80 μM) activates AMPK by increasing α-subunit phosphorylation at T172. A subsequent study reported that hepatic knockout of LKB1, an upstream kinase for AMPK α-subunit phosphorylation at T172, abrogated the effect of metformin on the alleviation of hyperglycemia in mice fed a high-fat diet (Shaw et al. 2005). In agreement with this result, our study showed that depletion of AMPK catalytic α-subunits abolished the suppression of cAMP- or glucagon-stimulated glucose production by metformin at concentrations typically found in the portal vein (≤80 μM) (Cao et al. 2014). Furthermore, metformin-mediated activation of AMPK increases CBP phosphorylation at S436 via atypical protein kinase ι/λ resulting in the disassembly of the CREB co-activator complex, inhibition of gluconeogenic gene expression, and a reduction in glucose production (He et al. 2009). A growing body of evidence supports the notion that metformin suppresses mRNA or protein levels of rate-limited gluconeogenic gene expression in the liver of mice and in primary hepatocytes (Cheng et al. 2006, Heishi et al. 2008, He et al. 2009, Martin-Montalvo et al. 2013). Corroborating our findings, a P300G422S knock-in mouse model, which harbors a reconstituted metformin phosphorylation site found in CBP, exhibits exaggerated hypoglycemia in a metformin tolerance test (He et al. 2014). The previous evidence supports a mechanism for metformin suppression of hepatic glucose production via activation of the LKB–AMPK pathway.
Activation of AMPK by low metformin concentrations (≤80 μM) via the promotion of the formation of the AMPKαβγ heterotrimeric complex
Metformin has no direct effect on AMPKα enzymatic activity (Meng et al. 2015). But, metformin is able to increase AMPK enzymatic activity by stimulating the phosphorylation of the α subunit at T172 (Zhou et al. 2001). Considering that the upstream kinase LKB1 is a constitutively active kinase (Lizcano et al. 2004), we asked whether metformin could affect the formation of functional AMPK heterotrimeric complex, consisting of an α catalytic subunit and βγ non-catalytic subunits, to increase the phosphorylation of the α subunit at T172 by upstream kinase (Davies et al. 1994, Mitchelhill et al. 1994, Stapleton et al. 1996). However, previous studies showed that AMPK α1β1γ1 or α2β1γ1 subunits exist in a 1:1:1 molar ratio (Davies et al. 1994, Stapleton et al. 1996), and it is generally thought that the heterotrimeric complex assembles immediately following translation in order to keep the subunits stable. Free subunits may not be stable and will therefore degrade rapidly (Neumann et al. 2003). Interestingly, we were able to express AMPK-α1,β1,-γ1 individually in both primary hepatocytes and Hepa1-6 cells (Cao et al. 2014, Meng et al. 2015). Purified AMPK subunits did not associate with other subunits in a coomassie stained gel, indicating that free AMPK subunits can exist and are stable. Furthermore, we examined the protein levels of the AMPK subunits in different mouse tissues and found that the protein levels of different AMPK subunits vary from tissue to tissue (Meng et al. 2015), suggesting that cellular protein levels of AMPK subunits are unlikely to exist in a 1:1:1 molar ratio (Davies et al. 1994, Stapleton et al. 1996).
To assess the cellular ratio of endogenous AMPK subunits, we generated and purified FLAG-tagged -α1, -β1, -γ1 since they account for over 90% of AMPK activity in hepatocytes (Davies et al., 1994, Stapleton et al. 1996). We first determined the amounts of purified FLAG-tagged -α1, -β1, -γ1 subunits that were equivalent to the corresponding protein levels in the same amount of cellular lysates. Subsequently, these amounts of purified FLAG-tagged -α1, -β1, and -γ1 subunits were used and immunoblotted with anti-FLAG antibody. Our data reveal that the -α1, -β1, and -γ1 subunits exist approximately in a 1:5:0.6 molar ratio in Hepa1-6 cells and in a 1:31:0.5 molar ratio in mouse liver (Meng et al. 2015). Our data suggest that the β1 subunit is much more abundant than α1 and γ1 subunits in both Hepa1-6 cells and liver.
The unequal expression of endogenous -α1, -β1, and -γ1 protein levels suggest that not all of the -α1, -β1, and -γ1 are in the heterotrimeric complex. Therefore, we reasoned that metformin might be able to affect the assembly of the AMPK heterotrimeric complex. In immunoprecipitation studies, we found that both α and γ subunits were co-immunoprecipitated when using β1/2 antibodies in metformin-treated Hepa1-6 cells. In contrast, in untreated Hepa1-6 cells, the γ1 but not the α subunit co-immunoprecipitated with β1 subunit. These data suggest that β1 and γ1 subunits are preassembled, and metformin promotes the association of the α1 subunit with β1 and γ1 complex to form the functional heterotrimeric complex (Meng et al. 2015). In agreement with these results, we found that mice fed a high-fat diet were given metformin in the drinking water exhibited an increase in α subunit phosphorylation at T172 as well as the association of the α subunit with the β subunit in the liver (Meng et al. 2015). Furthermore, low concentrations of metformin also increased the formation of AMPKαβγ heterotrimeric complex in in vitro assembly assay. Interestingly, neither AMP/ADP, nor phenformin affected the heterotrimeric assembly, suggesting that they activate AMPK through different mechanisms.
As we found that metformin was able to promote the formation of the AMPKαβγ heterotrimeric complex, we further tested whether the formation of the AMPKαβγ heterotrimeric complex had any effect on the phosphorylation of the α1 subunit at T172 by an upstream kinase. In in vitro phosphorylation assays, the phosphorylation of the α1 subunit at T172 by LKB1 is tenfold greater in the presence of β1 and γ1 subunits than α1 subunit alone. The presence of the γ1 subunit is essential for metformin's effect on the phosphorylation of the α1 subunit at T172 by LKB1, as metformin's effect was lost when the γ1 subunit was absent in the reaction. A previous study showed that cells expressing a mutant AMPKγ subunit (R531G) were resistant to metformin activation of AMPK (Hawley et al. 2010), suggesting that R531 in the γ subunit is critical for metformin binding. Furthermore, low metformin concentrations (<100 μM) prevented the dephosphorylation of the catalytic α1 subunit by PP2C in the presence of β1 and γ1 subunits, and this effect was lost in the absence of β1 and γ1 subunits. Taken together, our data support that metformin binds directly to AMPK subunits, resulting in increased assembly of the AMPK αβγ heterotrimeric complex as well as allosteric changes, making it more accessible for upstream kinases, such as LKB1, to phosphorylate the α1 subunit at T172 (Fig. 1). The formation of the AMPK αβγ heterotrimeric complex also hinders the dephosphorylation of the α1 subunit at T172 by protein phosphatase (Meng et al. 2015). Overall, the collective effects of metformin action lead to the activation of AMPK.
Metformin suppresses hepatic glucose production through the inhibition of mitochondrial respiratory chain complex 1 and AMP deaminase
Mitochondria respiratory-chain complex 1, also termed NADH:ubiquinone oxidoreductase, is the first membrane-embedded supramolecular complex in the electron transport chain. Complex 1 catalyzes the transfer of a hydride ion from NADH and a proton from the matrix to ubiquinone. Ubiquinol (QH2), a lipid-soluble carrier ubiquinone, can diffuse quickly in the inner mitochondrial membrane from complex 1 to complex 3. Complex 1 also transfers four protons from the matrix to the intermembrane space to generate a high electron gradient, which is the energy source in ATP synthesis. In 1955, it was documented that guanidine can inhibit the oxygen uptake of mitochondria. This was the first study that showed that guanidine may have an effect on the mitochondrial respiratory chain (Hollunger 1955). In 2000, two studies showed that metformin's action is through the disruption of mitochondrial complex I (El-Mir et al. 2000, Owen et al. 2000), leading to a decrease in ATP production. More recently, Foretz et al. (2010) suggested that the change in the AMP/ATP or ADP/ATP ratio after metformin's inhibition of the mitochondrial complex is primarily responsible for metformin's effect and that this occurs through an AMPK-independent pathway. It was also shown that metformin increases AMP levels via the inhibition of AMP deaminase (Fig. 2, Ouyang et al. 2011). In addition, elevated AMP levels after metformin treatment led to the inhibition of adenylate cyclase and a decrease in cAMP levels, resulting in the suppression of the cAMP-PKA pathway, and the inhibition of gluconeogenesis was proposed (Miller et al. 2013). We also found that metformin is able to decrease cellular ATP levels and increase the ratios of AMP/ATP or ADP/ATP (Cao et al. 2014). Since gluconeogenesis is an energy-demanding process in which synthesis of one molecule of glucose from lactate or pyruvate requires four molecules of ATP and two molecules of GTP, reduction of cellular ATP levels by metformin will lead to the suppression of hepatic gluconeogenesis (Fig. 2).
However, metformin's inhibition of mitochondrial complex I or AMP deaminase activity occurred at concentrations ≥5 mM in hepatocytes (El-Mir et al. 2000, Owen et al. 2000, Ouyang et al. 2011). The concentrations of metformin required to affect the ratios of AMP/ATP or ADP/ATP are ≥250 μM (Foretz et al. 2010, Cao et al. 2014). These metformin concentrations are unreachable in the portal vein after a therapeutic dose (Wilcock & Bailey 1994, He & Wondisford 2015). Interestingly, using typical pharmacologic dosing in mice, metformin administration did not decrease, but significantly increased mitochondrial complex I activity in the liver (Martin-Montalvo et al. 2013). We also found that a low metformin concentration (80 μM) significantly increased ATP levels in primary hepatocytes treated with cAMP. This may be due to the inhibition of gluconeogenesis and activation of fatty acid oxidation (Cao et al. 2014), but metformin still significantly activated AMPK (Zhou et al. 2001, Cao et al. 2014) and the suppression of glucose production relative to cAMP treatment alone in primary hepatocytes (Cao et al. 2014). These data indicate that inhibition of mitochondrial complex I activity is unlikely at pharmacologic metformin concentrations found in the portal vein, and that the inhibition of mitochondrial complex I is unnecessary for metformin's suppression of gluconeogenesis. Since very high metformin concentrations are required to inhibit mitochondrial complex I and perhaps, the AMP deaminase activity in hepatocytes as well (El-Mir et al. 2000, Ouyang et al. 2011), we suggest the effects of metformin seen at high concentrations might be either an off-target phenomena or signals of cell toxicity and are therefore clinically irrelevant.
Metformin suppresses hepatic glucose production through the inhibition of mitochondrial glycerol 3-phosphate dehydrogenase and through affecting gut–brain communication
It has been reported that metformin suppresses hepatic glucose production by inhibiting the enzymatic activity of mitochondrial glycerol 3-phosphate dehydrogenase (mG3PDH), which blocks the transport of NADH from the cytoplasm into mitochondria (Madiraju et al. 2014). This mechanism of metformin action may be important for the suppression of hepatic glucose production in patients with diabetes who have high levels of serum lactate, as the conversion of lactate to pyruvate leads to the generation of NADH. The inhibition of mG3PDH will result in an increase in NADH levels and a decrease in NAD+ levels, resulting in no NAD+ available for converting lactate to pyruvate. Subsequently, this action will halt glucose production from lactate (Fig. 2, right panel). Interestingly, AMPK negatively regulates the activity of yeast glycerol 3-phosphate dehydrogenase 2 (an analog of mammalian mG3PDH) by phosphorylating S72 (Lee et al. 2012). However, it remains to be determined whether AMPK can affect mG3PDH enzymatic activity by phosphorylation.
Moreover, intestinal nutrients can regulate glucose homeostasis in a manner involving the gut–brain–liver axis (Wang et al. 2008). Affecting gut–brain communication will lead to a change in hepatic glucose production. Duca et al. (2015) tested whether metformin has any effect on the gut–brain–liver axis and found that metformin suppression of hepatic glucose production occurs in part through the gut–brain–liver axis. Gut–brain communication is critical for acute metformin action, as metformin suppression of hepatic glucose production was negated in rats with blockade of the gut–brain–liver axis. The acute metformin effect is AMPK-dependent, involving the activation of PKA by GLP-1 in duodenal enterocytes (Duca et al. 2015). However, the importance of the gut–brain–liver axis in chronic metformin action remains to be determined.
Metformin alleviates hyperglycemia in T2D through the improvement of insulin signaling: an indirect pathway of metformin action
Even though metformin improves hyperglycemia mainly through the suppression of gluconeogenesis in the liver, it has also been found to be able to increase insulin sensitivity. This effect would improve insulin-mediated suppression of hepatic glucose production and enhance insulin-stimulated glucose disposal in skeletal muscle (Stumvoll et al. 1995, Inzucchi et al. 1998, Hundal et al. 2000). The potential mechanism of a metformin-mediated increase in insulin sensitivity will be discussed.
Metformin alters gut microbiota and maintains the integrity of the intestinal barrier
Obese individuals have a different composition of gut microbiota, and in a rodent model, obesity is associated with altered gut microbiota (Ley et al. 2005, 2006, Ridaura et al. 2013). High-fat-diet feeding leads to a similar change in gut microbiota composition and to a two- to threefold increase in blood lipopolysaccharide (LPS) (endotoxin) levels in mice (Cani et al. 2007, 2008). In addition, a single oral administration of lipids in human or mice acutely increased serum LPS concentration (Cani et al. 2007, Laugerette et al. 2011). Germ-free mice resisted high-fat-diet induced insulin resistance (Backhed et al. 2007). A low-dose LPS infusion mimicked high-fat-diet feeding and induced fasting hyperglycemia and hepatic insulin resistance in a mouse model (Ley et al. 2005). This line of evidence suggests that LPS from the gut is a critical etiologic factor for the development of insulin resistance. Increased LPS leakage from the gut has emerged as one of the most appealing mechanisms to explain the connection between changes in the intestinal microbiota and insulin resistance (Cani et al. 2007).
Recently, metformin has been shown to modulate the composition of gut microbiota in high-fat diet-fed and diabetic mice and decreases serum LPS levels (Shin et al. 2014). In addition, metformin can activate mucosal AMPK in the intestine (Duca et al. 2015). An earlier study showed that high-fat-diet feeding increases the intestinal permeability of LPS (Cani et al. 2007, 2008, Creely et al. 2007, Oliveira et al. 2011), and AMPK plays a critical role in maintaining intestinal barrier integrity (Peng et al. 2009, Elamin et al. 2013). Therefore, metformin-mediated activation of AMPK should decrease LPS leakage from the gut (Fig. 3).
Metformin blocks the activation of the NF-κB pathway and affects the expression of PTEN
The liver is the major organ responsible for LPS clearance. After being delivered to the liver, LPS binds to the multireceptor complex constituted of CD14, TLR4 and MD2 within the lipid rafts of hepatocytes and triggers the signal activation of IRAK, and subsequently, its downstream target, the NF-κB pathway (Scott & Billiar 2008). NF-κB first drew attention in the development of insulin resistance and T2D after the landmark finding that the anti-inflammatory agent aspirin inhibits NF-κB. Activation of NF-κB leads to insulin resistance (Arkan et al. 2005), and inactivation of this pathway protects against the development of insulin resistance, as evident in mice with NF-κB (p50) knockout (Gao et al. 2009). This mouse model exhibited increased insulin sensitivity in the liver and produced significantly less glucose in a hyperinsulinemic–euglycemic clamp. Furthermore, inhibition of the NF-κB pathway improved insulin resistance in db/db mice (Kim et al. 2013). Of particular interest is the finding that the activation of AMPK by AICAR inhibited the NF-κB pathway (Cacicedo et al. 2004). Aligning with this, metformin-mediated AMPK activation attenuates the activation of the NF-κB pathway (Hattori et al. 2006, Huang et al. 2009). Therefore, the inhibition of the NF-κB pathway by metformin-mediated AMPK activation would lead to an improvement in hepatic insulin signaling (Fig. 3).
Phosphatase and tensin homolog (PTEN), a tumor suppressor, can reverse PI3K (Phosphatidylinositol-4, 5-bisphosphate 3-kinase) function by dephosphorylating the PI(3,4,5)P3 to PI(4,5)P2, therefore, suppressing the PI3K-PKB/AKT pathway (Myers et al. 1998, Stiles et al. 2004). Intriguingly, LPS can induce the expression of PTEN (Okamura et al. 2007), and metformin can suppress PTEN expression in pre-adipocyte 3T3 cells (Okamura et al. 2007, Lee et al. 2011). This metformin action is AMPK dependent, as the metformin effect is lost in cells treated with Compound C (an AMPK inhibitor) or with AMPK depletion by shRNA. This report showed that PTEN is a downstream regulator of AMPK and that the AMPK–PTEN pathway plays a critical role in regulating inflammatory response (Fig. 3). However, further studies will be needed to demonstrate conclusively how metformin's effect on PTEN occurs in the liver as well as muscle and determines how activated AMPK suppresses PTEN expression.
Since the maximum metformin dose prescribed to patients with diabetes is ∼2.5 g/day, this high therapeutic dose might affect multiple targets. As an oral agent, metformin can change the composition of gut microbiota (Shin et al. 2014) and activate mucosal AMPK (Duca et al. 2015) that will maintain intestinal barrier integrity (Peng et al. 2009, Elamin et al. 2013). Together, these metformin effects will decrease LPS levels in the circulation as well as hepatic LPS levels, since LPS is delivered directly to the liver through the portal vein. In addition, the activation of duodenal AMPK by acute metformin treatment will stimulate the gut–brain–liver axis to regulate hepatic glucose production (Duca et al. 2015). After being delivered to the liver from the intestines, metformin can inhibit gluconeogenesis through different mechanisms. First, it can activate hepatic AMPK, which will then inhibit gluconeogenesis by phosphorylating the critical co-activators CBP and CRTC2 (He et al. 2009, Fig. 1). Second, high metformin concentrations can suppress hepatic gluconeogenesis by inhibiting mitochondrial respiratory chain complex 1 to reduce cellular ATP levels and increase the AMP/ATP ratio (El-Mir et al. 2000, Owen et al. 2000, Foretz et al. 2010) due to the high-energy demand of gluconeogenesis. We found that depletion of AMPKα1/2 decreased the suppression of glucose production in primary hepatocytes at high metformin concentrations (Cao et al. 2014). Thus, the increased ratio of AMP/ATP or ADP/ATP should activate AMPK (Oakhill et al. 2011, Xiao et al. 2011, Gowans et al. 2013) to augment the inhibition of gluconeogenesis. Third, inhibition of mG3PDH enzymatic activity by metformin will affect transport of NADH from the cytoplasm into mitochondria (Madiraju et al. 2014), suppressing gluconeogenesis process from lactate.
Reducing LPS leakage from the gut and blocking the activation of the NF-κB pathway, in concert with down-regulating the expression of PTEN, will result in an increase in insulin sensitivity. Moreover, metformin administration will activate AMPK to increase fatty acid oxidation by phosphorylating acetyl-CoA carboxylase and consequently improve insulin sensitivity in the liver (Fullerton et al. 2013). Collectively, suppression of gluconeogenesis by metformin's direct effects and the indirect improvement of insulin signaling in the liver will lead to the amelioration of hyperglycemia in patients with T2D. Even though our studies demonstrate that suppression of gluconeogenesis by low metformin concentrations is through the activation of AMPK by promoting the formation of the AMPKαβγ heterotrimeric complex in both in vivo and in vitro assays (Cao et al. 2014, Meng et al. 2015), and the γ subunit plays a critical role in the phosphorylation of the α subunit at T172 by upstream kinase LKB1. However, the exact metformin binding site(s) on AMPK subunits and the detailed mechanism leading to the improvement of insulin signaling by metformin still remain unknown and need to be unraveled in future studies.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this review.
This work was supported in part by grants from the National Institute of Diabetes and Digestive and Kidney Diseases R00DK085142 and R01DK107641.
CacicedoJMYagihashiNKeaneyJFRudermannNBIdoY2004AMPK inhibits fatty acid-induced increases in NF-κ B transactivation in cultured human umbilical vein endothelial cells. Biochemical and Biophysical Research Communications3241204–1209. (doi:10.1016/j.bbrc.2004.09.177)
CaoJMengSMChangEBeckwith-FickasKXiongLSColeRNRadovickSWondisfordFEHeL2014Low concentrations of metformin suppress glucose production in hepatocytes through AMP-activated protein kinase (AMPK). Journal of Biological Chemistry28920435–20446. (doi:10.1074/jbc.M114.567271)
Centers for Disease Control and Prevention 2011 In Diabetes: Successes and Opportunities for Population-Based Prevention and Control At A Glance 2011. Atlanta GA USA: Centers for Disease Control and Prevention. (available at: http://www.cdc.gov/nccdphp/publications/aag/pdf/diabetes.pdf)
CreelySJMcTernanPGKusminskiCMFisherfMDa SilvaNFKhanolkarMEvansMHarteALKumarS2007Lipopolysaccharide activates an innate immune system response in human adipose tissue in obesity and type 2 diabetes. American Journal of Physiology. Endocrinology and Metabolism292E740–E747. (doi:10.1152/ajpendo.00302.2006)
ForetzMHebrardSLeclercJZarrinpashnehESotyMMithieuxGSakamotoKAndreelliFViolletB2010Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state. Journal of Clinical Investigation1202355–2369. (doi:10.1172/JCI40671)
HuangNLChiangSHHsuehCHLiangYJChenYJLaiLP2009Metformin inhibits TNF-α-induced I κ B kinase phosphorylation, I κ B-α degradation and IL-6 production in endothelial cells through PI3K-dependent AMPK phosphorylation. International Journal of Cardiology134169–175. (doi:10.1016/j.ijcard.2008.04.010)
InzucchiSEBergenstalRMBuseJBDiamantMFerranniniENauckMPetersALTsapasAWenderRMatthewsDR2012Management of hyperglycaemia in type 2 diabetes: a patient-centered approach. Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetologia551577–1596. (doi:10.1007/s00125-012-2534-0)
LaugeretteFVorsCGeloenAChauvinMASoulageCLambert-PorcheronSPerettiNAlligierMBurcelinRLavilleM2011Emulsified lipids increase endotoxemia: possible role in early postprandial low-grade inflammation. Journal of Nutritional Biochemistry2253–59. (doi:10.1016/j.jnutbio.2009.11.011)