Introduction

The number of microorganisms in the human intestine is approximately 1 × 1014, while the number of eukaryotic cells in the human body is around 1 × 1013. The number of microbes inhabiting the intestines has been estimated to be 1014, which encompasses approximately 10 times more bacterial cells than the number of human cells and over 100 times the amount of genomic content (microbiome) as compared to the human genome [1, 2]. Meanwhile, a more recent estimation has demonstrated that the ratio of human to bacterial cells in the human body (microbiota) is, in fact, around 1:1 [3]. As a result of co-evolution of the host mucosal immune system and the microbiota, both have developed multiple mechanisms to maintain homeostasis [3]. However, dissociations between the composition of the gut microbiota and the human host may play a crucial role in the development of type 2 diabetes (T2DM) [4,5,6,7]. Large metagenome-wide studies in China and Europe have documented gut microbial dysbiosis associated with obesity and T2DM. The common finding of these studies was a reduction in butyrate-producing microbes, together with an increase in opportunistic pathogens [4,5,6,7,8]. Data from these studies and the Danish MetaHIT project were analyzed after controlling for metformin administration [4,5,6,7,8,9]. The conclusion was that the noted reduction in butyrate-producing bacteria could imply gut microbiome shifts among patients with T2DM [4,5,6,7,8,9]. On the other hand, while recent data suggest that metformin alters gut microbiota, the classic mechanisms of action, such as inhibition of hepatic gluconeogenesis, still retain their importance in the treatment of patients with T2DM [4,5,6,7,8,9].

Metformin and metabolomics

Metformin, the most frequently administered medication to treat patients with T2DM, has only recently been suggested to alter gut microbiota composition through the increase in mucin-degrading Akkermansia muciniphila, as well as several SCFA-producing (short-chain fatty acid) microbiota [10]. A meta-analysis of 199 individuals with T2DM and 544 control participants without diabetes confirmed that metformin significantly altered gut microbiota composition: metformin patients with T2DM could be identified based on compositional changes in their gut microbiota [9].

Metabolomics, i.e., the scientific study of chemical processes involving metabolites—the small molecule intermediates and products of metabolism—are likely to shed much light on this subject [11]. Specific microbiome profiles are related to increased glucose intolerance [12]. Changes in gut microbiota may influence the gut metabolome, thus affecting the intestinal output of butyrate and acetate [12, 13]. These are major gut-derived metabolites which are implicated in the host’s insulin resistance and, thus, glycemic control. It has been suggested that an enhancement in bacteria producing the short-chain fatty acids butyrate and propionate may improve glycemia (Fig. 1) [14]. Butyrate and propionate increase intestinal gluconeogenesis. In rodents, increased intestinal gluconeogenesis results in a reduction in hepatic gluconeogenesis, appetite, and weight, leading to improved glucose homeostasis. In particular, the gut microbiota of participants on metformin exerted increased ability to produce butyrate and propionate, while in the untreated individuals, there was enhancement in microbial genes implicated in the degradation of glycine and tryptophan as well as decreases in genes involved in threonine and arginine degradation. This is of particular interest since glycine has been associated with insulin sensitivity, and glycine supplementation has been reported to improve insulin sensitivity in T2DM [12, 13].

Fig. 1
figure 1

Dysbiotic microbiota is characterized by the reductions in butyrate-producing bacteria, while metformin induces abundancies in butyrate-producing bacteria and in Lactobacillus and Escherichia spp.

Full size image

Metformin and intestinal absorption

There is a growing body of evidence in favor of a key role of metformin in the lower gut [15,16,17,18,19,20,21,22,23]. It is noteworthy that metformin levels in the gut are 100- to 300-fold higher than those in serum, making the gut the primary reservoir for metformin in humans [15,16,17,18,19,20,21,22,23]. Unlike oral administration, intravenous administration of metformin in humans does not improve glycemia [22]. Rates of systemic absorption and intestinal transit time affect the concentrations of metformin to which the microbiota are exposed: oral bioavailability is about 30–60%, and approximately 50% of an ingested dose is found in the feces, although both these factors exhibit significant person-to-person variability. Thus, the degree to which the intestinal flora is altered by the drug is affected by host factors [15,16,17,18,19,20,21,22,23].

In particular, the oral absorption and hepatic uptake of metformin is performed by organic cation transporters (OCTs), mainly OCT1 and OCT3, whereas renal excretion of metformin is mediated by OCT2 [24,25,26]. A study performed in rodents has demonstrated that metformin could be a superior substrate for renal OCT2 rather than for hepatic OCT1 [27]. Besides, OCT2 accounts for gender differences in renal organic cation transporting activity [28,29,30]. Furthermore, OCT1 is also highly polymorphic: 34 OCT1 polymorphisms have been identified among 10 ethnic groups. Met408Val (rs628031) variant has been the most frequently explored as regards metformin effects. Although some alleles have been linked to negative effects on metformin, others have shown positive effects [31]. These polymorphisms provide a plausible explanation for the interindividual variations in metformin response [31]. However, is there any relationship between polymorphisms as concerns OCTs, metformin, and microbiota? Only recently, a study demonstrated that differences in abundance of Bifidobacterium could be explained by opioids acting as organic cation transporter 1 (OCT1) inhibitors [32]. This study was the first to document that abundance of Bifidobacterium and Prevotella was influenced by interactions between T2DM and the use of metformin and opioid (acting as OCT inhibitors) [32]. Therefore, the issue of metformin, OCT, and gut microbiota remains to be further investigated in the near future. The pharmacological actions of metformin include changes in bile acid recirculation and gut microbiota, resulting in enhanced levels of glucagon-like polypeptide-1 (GLP-1). These observations have been confirmed by clinical studies testing the response to delayed-release metformin (MetDR) in healthy participants and in patients with T2DM [33,34,35]. The delayed release of metformin relies on pH-dependent dissolution of the enteric coat of the tablet; this mainly occurs at pH 6.5 in the distal intestine, where there is abundancy of L cells, which accounts for the gut-based mechanisms of metformin. Studies have documented that MetDR has a glucose-lowering effect, which is comparable to that of extended release metformin, but with a 40% increase in its potential, thus allowing for lower doses of MetDR to be used [33,34,35]. Moreover, two randomized trials have shown that MetDR resulted in glucose-lowering effects comparable to immediate-release metformin, with lower systemic exposure [33,34,35]. Nevertheless, once-daily dosing with MetDR 1500 mg results in improved glucose-lowering effect, with immediate-release metformin 1000 mg twice daily and better gastrointestinal tolerance [33,34,35,36]. Whether MetDR affects the gut microbiota in a manner similar to that of immediate-release metformin remains unknown and needs to be elucidated [33,34,35,36].

Metformin and its adverse gastrointestinal effects

The work by Forslund et al. postulates that microbiota-mediated mechanisms lie behind both the therapeutic and adverse effects of metformin [9]. Alterations in the microbiome by metformin may account for its gastrointestinal intolerance. Burton et al. carried out a study using a gastrointestinal microbiome modulator (GIMM) or placebo together with metformin in metformin-intolerant patients with T2DM [33]. They found that treatment with metformin in conjunction with GIMM resulted in lower fasting glucose levels, suggesting better tolerance of metformin therapy for longer or at a higher dose. Thus, metformin appears to affect the microbiome, and an individual’s metformin tolerance or intolerance may be influenced by their microbiome. Furthermore, increases in the genus Escherichia in metformin-treated individuals have been functionally associated with gas metabolism indicating involvement of the genus Escherichia in the adverse gastrointestinal effects of metformin [37].

Metformin’s main adverse side effect, i.e., lactic acidosis, is suggested as being due to the accumulation of lactate via the inhibition of hepatic glucose production from lactate molecules. Moderate to severe renal impairment as well as drugs inhibiting the metformin transporters (OCTs) could lead to the development of lactic acidosis, with potentially fatal results, due to an increase in its plasma concentrations [38, 39]. Recently, the 808G>T variance in the SLC22A2 gene can affect the plasma lactate level and the incidence of hyperlactacidemia in T2DM patients on metformin treatment [40] (Fig. 2). In addition, the effect of metformin on gut microbiota among patients with GFR < 60 ml/min/1.73 m2 remains unknown, as studies dealing with this particular issue are lacking.

Fig. 2
figure 2

Metformin-associated lactic acidosis (MALA) may be associated with polymorphisms in OCTs, renal impairment, or drug interactions

Full size image

Conclusions

It must be recognized that metformin may have an impact on systemic carbohydrate metabolism via gastroenteric mechanisms that do not directly involve gut microbiota. These include changes in bile acid physiology, enteric hormones, and effects on duodenal AMPK signaling, which contribute to suppression of hepatic gluconeogenesis [34,35,36]. While recent data suggest that metformin alters gut microbiota, the classic mechanisms of action, such as inhibition of hepatic gluconeogenesis, still retain their importance in the treatment of patients with T2DM. Further research is thus mandatory to evaluate the effects of metformin, which may be attributed to changes in gut microbiota. It is anticipated that these effects will vary considerably among metformin-treated patients.