Role of gut microbiota in the pathogenesis and treatment of diabetes mullites: Advanced research-based review

Gut microbiota plays an important role in the proper functioning of human organisms, while its dysbiosis is associated with disease in various body organs. Diabetes mellitus (DM) is a set of heterogeneous metabolic diseases characterized by hyperglycemia caused by direct or indirect insulin deficiency. There is growing evidence that gut microbiota dysbiosis is closely linked to the development of DM. Gut microbiota composition changes in type 1 diabetes mullites (T1DM) and type 2 diabetes mullites (T2DM) patients, which may cause gut leakiness and uncontrolled entry of antigens into the circulation system, triggering an immune response that damages the isle β cells or metabolic disorders. This review summarizes gut microbiota composition in healthy individuals and compares it to diabetes mullites patients. The possible pathogenesis by which gut microbiota dysbiosis causes DM, particularly gut leakiness and changes in gut microbiota metabolites is also discussed. It also presents the process of microbial-based therapies of DM.


Introduction
The gut microbiota plays an essential role in the proper functioning of human organisms (Vrancken et al., 2019). It co-evolves and symbioses with humans by combating pathogenic bacteria, assisting nutrient digestion, maintaining the integrity of the intestinal epithelia, and promoting immunological development (Stecher and Hardt, 2011;Dave et al., 2012;Natividad et al., 2012;Shreiner et al., 2015). However, when the balance of the microbiota community is affected, known as gut microbiota dysbiosis, it may lead to various diseases, as summarized in Figure 1 (Hou et al., 2022a). Diabetes mullites (DM) is a set of heterogeneous metabolic disorders characterized by hyperglycemia and glucose intolerance, with high and increasing prevalence and multiple complications (Danaei et al., 2011;Saeedi et al., 2019;Classification and Diagnosis of Diabetes, 2022). The classical opinion is that type 1 diabetes mullites (T1DM) results from autoreactive T-cellmediated partially or absolute destruction of pancreatic β cells in patients (Atkinson et al., 2014). In contrast, type 2 diabetes mullites (T2DM) is the outcome of a progressive loss of sufficient pancreatic β-cell insulin secretion in the context of insulin resistance (IR) (Kahn et al., 2014). According to International Diabetes Federation, there were more than 463 million DM patients in 2019 over the world and this number is estimated to rise to 578 million by 2030 and even 700 million by 2045 (Saeedi et al., 2019).
In the past decade, studies on the gut microbiome have developed rapidly due to the advancements in sequencing technologies and data analysis. It has been observed that gut microbiota dysbiosis is presented in both T1DM and T2DM patients (Bibbò et al., 2017;Vallianou et al., 2018). Gut microbiota dysbiosis may cause gut leakiness, which leads to external antigens uncontrollably entering the circulatory system (Di Tommaso et al., 2021). These antigens could activate islet autoimmunity and directly damage pancreatic β cells, and gut microbiota metabolites may also cause hormonal effects leading to metabolic disorders (Sun et al., 2015;Zhu and Goodarzi, 2020). Immune and metabolic disorders play an important role in the pathogenesis of DM (Bachem et al., 2019;Hou et al., 2022a). In addition, many diabetic complications are proven to be linked to the gut microbiota, including diabetic retinopathy, diabetes-induced cognitive impairment, diabetic peripheral neuropathy, and diabetic nephropathy . Although the role of gut microbiota in the pathogenesis of DM is not yet understood, more and more researchers are pinning new hopes for treating DM in microbiological therapy.
Diabetes mellitus and its complications cause physical and mental injury to patients and a great economic burden to the medical system. However, there is no single treatment that can sustainably and consistently prevent the progression of β-cell failure after the onset of DM. Elucidating the role of gut microbiota in the onset and progression of DM will contribute to a better understanding of DM and the development of new treatments for it. This review summarizes the gut microbiota composition in healthy individuals and compares it with T1DM and T2DM patients. Furthermore, we review the role of gut microbiota dysbiosis in the pathogenesis of DM, particularly gut leakiness, immune disorders, and metabolite disorders. This paper will also present the process of microbial-based therapies of DM in the final. Diseases caused by gut microbiota dysbiosis.

Gut microbiota dysbiosis in T1DM
Previous studies have identified differences in the gut microbiota between healthy individuals and T1DM patients (Bibbò et al., 2017). The stability, connectivity, abundance, and composition of the gut microbiota are probably linked to the development of T1DM (Han et al., 2018). Decreased microbiota diversity is a common gut microbiota shift associated with T1DM development (Leiva-Gea et al., 2018).
At the phyla level, it was well documented that proportions of Firmicutes phyla decreased in T1DM patients compared to the healthy individual group (Murri et al., 2013;Kostic et al., 2015;Leiva-Gea et al., 2018), while Bacteroidetes abundance increased successively (Alkanani et al., 2015;Pellegrini et al., 2017). Furthermore, previous research indicated that the F/B ratio significantly increased over time in children who eventually progressed to clinical T1DM and T1DM (Murri et al., 2013;Pellegrini et al., 2017;Leiva-Gea et al., 2018). However, research also showed no difference in the F/B ratio between T1DM patients and healthy individuals (Qi et al., 2016). Studies assessing the relationship between Proteobacteria abundance and T1DM have reported conflicting results, with some reporting a positive association (Brown et al., 2011;Cinek et al., 2018), some reporting a negative association (Leiva-Gea et al., 2018), and some reporting no difference (Murri et al., 2013).
At the genus level, T1DM patients present with a higher abundance of 12 different genera, including Bifidobacterium, Bacteroides, Escherichia, Veillonella, Clostridium, Enterobacter, Lactobacillus, Ruminococcus, Streptococcus, Sutterella, Lactococcus, and Blautia, among which Bacteroides is reported to be the dominant genus in the most research literature (Mokhtari et al., 2021). A more recent two-sample Mendelian randomization analysis revealed a close link between a higher relative abundance of the Bifidobacterium and a higher risk of T1DM (Xu et al., 2021).

Gut microbiota dysbiosis in T2DM
Environmental factors play an essential role in the onset and development of T2DM and recent evidence indicates that gut microbiota dysbiosis is one of them (Gurung et al., 2020;Hou et al., 2022b). Like T1DM, the gut microbiota in T2DM differs from those in healthy individuals (Vallianou et al., 2018;Que et al., 2021). The Integrative Human Microbiome Project found that insulin-resistant (IR) individuals had distinguishable molecular and microbial patterns at baseline from the healthy controls group (The Integrative Human Microbiome Project, 2019).
At the phyla level, Firmicutes level was reported to be lower in T2DM patients than in healthy individuals, while Bacteroidetes level was increased (Larsen et al., 2010). However, some recent research reported the opposite result; Firmicutes increased but Bacteroidetes decreased in T2DM patients (Sedighi et al., 2017;Zhao et al., 2019). Uniformly, the F/B ratio was reported to increase in some studies, but decreased in some studies (Sedighi et al., 2017;Zhao et al., 2019). In addition, the recent research also indicated an increased level of Proteobacteria in T2DM individuals (Sedighi et al., 2017;Zhao et al., 2019).
At the genus level, a large-scale metagenomic analysis in China compared the structural characteristics of the gut microbiota in healthy control and T2DM patients (Qin et al., 2012). It was recognized that conditioned pathogens (Escherichia  (Karlsson et al., 2013). Notably, the above two studies both reported a decrease in butyrate-producing microbiota in T2DM patients, particularly Roseburia (Qin et al., 2012;Karlsson et al., 2013). In addition, recent research confirmed the decrease of the Clostridium genus in T2DM patients (Allin et al., 2018). Similarly, a lower level of Akkermansia mucinphila, which is responsible for degenerating mucin in the gut, was also considered a risk factor for T2DM (Allin et al., 2018;Hasani et al., 2021).

Role of the gut microbiota in the development of DM
From the above discussion, we can conclude that DM patients' gut microbiota is characterized by an increased level of opportunistic pathogens and decreased level of probiotics. Furthermore, Firmicutes are negatively associated with T1DM, while Bacteroidetes is a positive factor at the phyla level. However, conflicting results have been reported on the difference between healthy individuals and T2DM patients. At the genus level, Bacteroides showed a promotive association with both T1DM and T2DM. In addition, both T1DM and T2DM patients presented a lower level of butyrate-producing microbiota. Figure 2 depicts the role of gut microbiota dysbiosis in the development of DM.

Gut leakiness: A possible origin of DM
An epithelial layer-based physical barrier and a mucosal immune cell-based functional barrier make up the majority of the intestinal barrier and separate the host from the external environment (Mu et al., 2017). Tight junctions (TJ) and mucins are key to maintaining the integrity of the intestinal barrier (Durazzo et al., 2019). Butyrate-producing gut microbiota, such as Firmicutes, has been shown to promote TJ assembly, mucin synthesis, and anti-inflammatory properties (Hague et al., 1996;Peng et al., 2007Peng et al., , 2009). On the contrary, Bacteroides can inhibit the production of TJ protein and thus promote gut leakiness (Brown et al., 2011;Tlaskalová-Hogenová et al., 2011). Besides, Bifidobacterium, Bacteroides, and Ruminococcus can degrade mucin and thus induce gut leakiness (Hooper et al., 2002). Some bacterial metabolites, such as butyrate, have also been identified to play an important role in maintaining intestinal epithelial integrity (Wang et al., 2012;Kayama et al., 2020). Correspondingly, a low abundance of the butyrateproducing gut microbiota and increased abundance of the Bacteroides are presented in DM patients. Of note, hyperglycemia can induce gut leakiness by changing the TJ and adherence junctions (Thaiss et al., 2018). Immune and metabolic disorders caused by uncontrolled exogenous substances entering the circulatory system as a result of gut leakiness may be the origin of DM.
Immune disorder: The key to developing T1DM caused by gut microbiota dysbiosis and gut leakiness Gut microbiota significantly influences the development of the immune system. Human and animal studies support a causal relationship between early microbial exposure and immune function development (Zhou et al., 2021). Defective immune response maturation is associated with T1DM progression later in life (Zhou et al., 2021). Moreover, the hygiene hypothesis posits that environmental improvements and antibiotic use lead to a reduction in microbiota diversity, possibly associated with a significant increase in allergic diseases and certain autoimmune diseases (Bach, 2018). Notably, the host's genetic set-up of can also interact with the gut microbiota, leading to alterations in microbial composition, immune system activation, and T1DM susceptibility (Alkanani et al., 2015;Bonder et al., 2016).
It is suggested that gut leakiness possibly leads to uncontrol entry of external antigens (Di Tommaso et al., 2021), which could activate islet autoimmunity and directly damage pancreatic β cells (Sun et al., 2015;Bachem et al., 2019). Lipopolysaccharides (LPS) are a possible molecular link between the gut microbiota, inflammation, and T1DM (Bachem et al., 2019), an integral part of the outer membrane of Gram-negative bacterial species. The leakage of fatty acids and LPS can activate the toll-like receptor 4 (TLR4) (Velloso et al., 2015). Activation of TLR contributes to the maturation of dendritic cells and the recognition of pathogenassociated molecular patterns (Li et al., 2013). TLR4 is a member of patten-recognition receptors, and its activation helps activate pro-inflammatory signaling pathways, expressing and secreting cytokine while bacterial pathogens present (Shi et al., 2006;Cani et al., 2007). A case-control study proved that T1DM patients have higher circulating LPS levels than healthy individuals (Devaraj et al., 2009). In addition, these antigens may be absorbed by antigen-presenting cells (APCs) in the gut mucosa, which then activates the islet-reactive T cells that would be subsequently transported to pancreatic lymph nodes and islets to induce the damage of β cells in genetically predisposed individuals (Sorini et al., 2019). Some antigens may significantly be homologous to islet autoantigen, known as molecular mimicry, so they can directly induce the activation of pathogenic CD8+ T cells to promote DM development (Tai et al., 2016). Metabolites disorder: The key to developing T2DM induced by gut microbiota dysbiosis and gut leakiness Like T1DM, gut microbiota dysbiosis in T2DM can increase serum LPS concentration to injure the intestinal barrier and change in mucosal immune response (Allcock et al., 2001). LPS and other antigens can activate TLR4 on immune cells and induce pro-inflammatory response and IR (Medzhitov, 2001;Medzhitov and Horng, 2009;Janssen and Kersten, 2017). Besides, gut leakiness leads to macrophage infiltration, which causes local inflammation by producing and activating serum IL-6, TNF-α, and other inflammatory cytokines (Ghosh et al., 2020). Gut leakiness may also introduce the gut microbiota and its metabolite into the blood systemic circulation and promote local and systemic immune responses (Sudo et al., 2002;Sumida et al., 2022). However, T2DM is traditionally characterized as a metabolic disease, so the role of gut microbiota metabolite is more remarkable in the development of T2DM. Gut microbiota produces bioactive metabolites, including short-chain fatty acids (SCFAs), ammonia, phenols, endotoxins, etc., through dietary macronutrients (Schroeder and Bäckhed, 2016). SCFAs, bile acids, indole derivatives, sulfur-containing amino acids, and vitamins are gut microbiota metabolisms that prevent DM, whereas Branched-chain amino acids (BCAAs), phenol, p-cresol, methane, amines and ammonia are gut microbiota metabolic that promote DM progression (Khan et al., 2014). Herein, we review the effect of several important metabolites in T2DM.
The short-chain fatty acids (SCFAs), such as acetate, propionate, and butyrate, are products of the metabolism of soluble fiber and amino acids by gut microbiota. Widely reported studies have demonstrated that SCFAs contribute to improved glucose homeostasis and metabolism in tissues such as the liver, adipose, and muscle (Canfora et al., 2015). Acetate can induce the browning of adipose tissue, and propionate stimulates the release of peptide YY (PYY) and glucagon-like peptide 1 (GLP-1) to reduce energy intake, whereas butyrate can reduce inflammation and reverse the decline in GLP-1 receptor expression in the liver (Sahuri-Arisoylu et al., 2005;Chambers et al., 2015;Jin et al., 2015;Zhou et al., 2017Zhou et al., , 2018. In addition, two receptors for SCFAs, free fatty acid receptor 3 (GPR41) and free fatty acid receptor 2 (GPR43), have been identified to be directly implicated in T2DM development, which are expressed in enteroendocrine cells, intestinal epithelial cells, pancreatic islet cells, and other cells (Stoddart and Smith, 2008;Priyadarshini et al., 2018). Activation of The role of gut microbiota dysbiosis in the development of diabetes mullites.
Frontiers in Microbiology 06 frontiersin.org GPR41 was reported to stimulate leptin secretion to regulate energy expenditure and long-term food intake (Xiong et al., 2004) and peptide YY (PYY) to increase satiety (Larraufie et al., 2018). In addition, activating the GPR41 in the sympathetic nervous system can stimulate energy expenditure and reduce the risk of T2DM (Kimura et al., 2011). GLP-1 will be increasingly released when GPR43 is activated, which enhances insulin secretion, suppresses glucagon production, and increases satiety (Tolhurst et al., 2012). Of note, GLP-1 might also improve endothelial cell function by changing the composition of gut microbiota (Chen et al., 2022). The GPR43 transgenic mouse showed improved metabolic parameters, such as reduced obesity, improved homeostasis, improved lean meat quality, and higher GLP-1 secretion (Bjursell et al., 2011;Tolhurst et al., 2012;Kimura et al., 2013). Moreover, it has been demonstrated that SCFAs can decrease inflammation in mucosal and chronic systemic, possibly because of their ability to inhibit pro-inflammatory cytokines like interleukin-6 (IL-6) and tumor necrosis factor α (TNF-α) (Roelofsen et al., 2010), induct anti-inflammatory cytokines (Säemann et al., 2000) and decrease the infiltration of inflammatory cells into adipose tissue (Meijer et al., 2010). Primary bile acids are generated by the liver from cholesterol and released into the intestine by the gallbladder. The majority of these bile acids are reabsorbed in the intestinal tract, while a small amount reaches the lower intestine tract and is converted by the gut microbiota into secondary bile acid. Glucose homeostasis and intestinal fat absorption are both impacted by primary and secondary bile acids, which may significantly impact the pathophysiology of T2DM (Ahmad and Haeusler, 2019). Bile acids can lessen the amount of gluconeogenesis in the liver, enhance the production of glycogen, boost energy expenditure, stimulate the release of insulin, and reduce inflammation (Shapiro et al., 2018). These bile acids attach to and trigger nuclear hormone receptors like G-protein-coupled bile acid receptor-1 (TGR5) and nuclear hormone receptors, including the Pregnane X receptor (PXR) and Farnesoid X receptor (FXR). TGR5 is presented in digestive, immune, and adipose tissues. Intestinal TGR5 activation promotes insulin secretion and satiety by inducing the release of GLP-1 and PYY (Katsuma et al., 2005;Thomas et al., 2009;Kuhre et al., 2018). On the contrary, the knockout of TRG5 in myeloid-lineage cells and macrophages promoted IR and inflammation in adipose tissue (Perino et al., 2014). Moreover, the action of thyroid hormones may be promoted by bile acid-driven activation of TGR5 in adipose tissue, leading to an increase in energy expenditure (Watanabe et al., 2006). FXP is widely expressed in the liver, gut, kidney, and other tissues. The application of a synthetic FXR agonist was shown to significantly reduce plasma glucose, free fatty acids, triglycerides, cholesterol, and hepatic steatosis in DM mice (Zhang et al., 2006;Fang et al., 2015). Activation of PXR contributed to reducing bile acid production and increasing bile acid clearance to promote the absorption of lipid and fat-soluble vitamins (Staudinger et al., 2001). In addition, bile acids can stimulate GLP-1 secretion to regulate glucose metabolism and improve insulin sensitivity (Pathak et al., 2018;Shapiro et al., 2018).
Branched-chain amino acids (BCAAs) are essential amino acids produced by gut microbiota, including isoleucine, leucine, and valine. A case-cohort study supported that higher baseline BCAAs and their increase over 1 year were associated with a risk of T2DM (Ruiz-Canela et al., 2018). The obese mice model restored metabolic health while feeding them a reduced BCAAs diet, including improved insulin sensitivity and glucose tolerance, despite continuing to consume a high-sugar and high-fat diet (Cummings et al., 2018). However, a mendelian randomization analysis study suggested that the genetic risk score (GRS) for circulating BCAAs level was not linked to the homeostasis model assessment of insulin resistance (HOMA-IR), while the GRS for IR traits was significantly linked to increased circulating BCAAs level (Mahendran et al., 2017). This result indicates that it is not the increase of BCAAs that leads to IR, but IR that leads to the increase of BCAAs. The altered BCAAs levels are possibly due to reduced suppression of proteolysis caused by IR and reductions in BCAA catabolism in peripheral tissues caused by adiponectin signaling in T2DM (Lian et al., 2015;Giesbertz and Daniel, 2016).
In addition to the metabolites mentioned above, Trimethylamine (TMA), Indole derivatives, Imidazole propionate, etc., are associated with T2DM. TMA will be oxidized in the liver into Trimethylamine oxide (TMAO), and a case-control study reported that TMAO is positively associated with newly diagnosed T2DM . Indolepropionic acid is the product of the microbial metabolism of tryptophan, which is negatively linked to the risk of developing IR, low-grade inflammation, and T2DM (de Mello et al., 2017). Imidazole propionate is a microbial metabolite from histidine, which was demonstrated to be higher in T2DM patients compared with healthy individuals (Koh et al., 2018).

Progress in microbial-based therapies in DM
With the growing understanding of the role of gut microbiota in DM, more and more researchers are trying to use microbial-based therapies to treat DM. These therapies directly or indirectly alter the composition of gut microbiota to function. Herein, we review the progress in microbial-based therapies to treat DM.

Probiotics and prebiotics
Probiotics are referred to living microorganisms that are beneficial to the host's health when administered in adequate amounts (Hill et al., 2014). Research on probiotics has pointed Frontiers in Microbiology 07 frontiersin.org out their ability to enhance gut barrier function, regulate immunity, and competitively adhere to mucus and epithelial cells to benefit DM (Lomax and Calder, 2009;Ohland and Macnaughton, 2010). In animal models, oral administration of the probiotics increased the anti-inflammatory cytokines, such as TGF-β and IL-10 (Calcinaro et al., 2005;Sturm et al., 2005), decreased the pro-inflammatory cytokines like TNF-α, IL-6, and IL-1β (Mariño et al., 2017), and regulated the immune balance between Th1/Th2/ Th17/Treg cells (Lau et al., 2011;Jia et al., 2017). In addition, bone mineral density was positively correlated with the number of lactic acid bacteria and Bifidobacterium in DM patients (Hou et al., 2017).
Human and animal trials have demonstrated the potential of probiotics in preventing and treating DM (Que et al., 2021). It has been reported that the onset of DM in non-obese diabetics (NOD) was reduced by the early supplement of complexprobiotic-preparation VSL#3 (Calcinaro et al., 2005). Another research identified that early supplemental probiotic use reduced the risk of islet autoimmunity in HLA genotype T1DM high-risk children (Uusitalo et al., 2016), while another demonstrated a negative result (Savilahti et al., 2018). Early oral probiotic medication Clostridium butyricum was also observed to prevent NOD mice from developing DM (Jia et al., 2017). Combination of galactooligosaccharides, Bifidobacterium breve strain Yakult, and Lacticaseibacillus paracasei strain Shirota improved the amounts of total lactobacilli and Bifidobacterium, the proportions of Bifidobacterium, and the concentrations of acetate and butyrate in excrement in T2DM patients (Kanazawa et al., 2021).
Colonizing the host gut is an important process by which probiotics function. However, according to a systematic review, probiotics had no impact on the fecal microbiome composition in six of the seven studies that have been examined (Kristensen et al., 2016). Therefore, using prebiotics to promote the propagation of probiotics is another idea to regulate gut microbiota composition. Prebiotics are non-digestible dietary components that promote the host's health by regulating the gut microbiota composition, especially by augmenting the abundance of Bifidobacteria and/ or Lactobacillus (Jakaitis and Denning, 2014). Prebiotics improved glycemic control and decreased intestinal permeability in T1DM patients in a randomized placebocontrolled study, which enhanced insulin sensitivity (Ho et al., 2019). Human milk oligosaccharides (HMOs), the most common prebiotics that comprises multiple prebiotic oligosaccharides, are one of the most prevalent constituents of human milk (Knip and Honkanen, 2017). HMOs are believed to protect against autoimmune DM, possibly through selectively stimulating the growth of Bifidobacteria (Brown et al., 2011;de Goffau et al., 2013). In addition, HMOs also have microbially independent immunomodulatory properties, such as the induction of Treg cells (Lehmann et al., 2015) and the ability to maintain intestinal integrity (Aakko et al., 2017). Other prebiotics such as long-chain inulin-type fructosan and β-glucan-rich products have also been demonstrated to directly inhibit the progression of insulitis in NOD mice and reduce the incidence of DM Gudi et al., 2019).

Fecal microbiota transplantation
FMT refers to the transfer of healthy microbiomes to dysregulated receptors in the gut microbiota to restore the normal bacterial community (Vindigni and Surawicz, 2017). Notably, these healthy microbiomes can be not only allogeneic (allo-FMT) but also autologous (auto-FMT). Auto-FMT is often implemented by collecting someone's microbiota when they are healthy and transplanting it back when their gut microbiota is out of whack. Participants with abdominal obesity or dyslipidemia can maintain losing weight and controlling blood glucose after accepting auto-FMT whose gut microbiota was collected during the weight-loss intervention phase (Rinott et al., 2021). Another recent randomized controlled trial discovered that auto-FMT prevented the decline in endogenous insulin secretion for 12 months after onset in patients with a recent diagnosis of T1DM, suggesting FMT can possibly prevent the ongoing β cells damage in T1DM patients (de Groot et al., 2021). Compared to auto-FMT, allo-FMT transplants healthy donors' gut microbiota to receivers, which has been an important therapy in treating chronic diarrhea caused by Clostridium difficile infections (van Nood et al., 2013;Saha et al., 2021). Early-life FMT from MyD88-deficient mice has been shown to significantly delay the onset of T1DM in NOD mice (Peng et al., 2014). In addition, the development of T1DM was also significantly delayed by the non-selective transplantation of human gut microbiota into GF NOD mice (Neuman et al., 2019). On the contrary, the incidence of T1DM significantly increased when mice received antibiotics (Livanos et al., 2016). Of note, in newly diagnosed T1DM patients, auto-FMT delivered through duodenal tubes is more effective than allo-FMT in protecting β-cell function (de Groot et al., 2021).
In recent years, the use of FMT in the treatment of T2DM has also made great progress. A study with T2DM mice discovered that FMT can reduce hyperglycemia, improve IR, inhibit the level of chronic inflammation in pancreatic, and reduce β-cell apoptosis (Wang et al., 2019). Similarly, it has been demonstrated that the insulin sensitivity of patients with metabolic syndrome increased after receiving the transfer of gut microbiota from lean donors (Vrieze et al., 2012;Kootte et al., 2017). In addition, a randomized clinical trial indicated that butyrate-producing microbiota increased while transferring the microbiota from healthy lean donors to T2DM patients (Ng et al., 2022). Moreover, recent research began to try to combine FMT with dietary or lifestyle intervention to treat T2DM and got better results than single FMT (Ng et al., 2022;Su et al., 2022). It has been observed that FMT can increase the Frontiers in Microbiology 08 frontiersin.org abundance of beneficial microbiota such as Bifidobacterium and decrease the level of Sulfate-reducing bacteria, Desulfovibrio, and Bilophila (Ng et al., 2022;Su et al., 2022). Similar to probiotics, FMT also has difficulty in colonizing the gut, but repeated FMTs can significantly increase the engraftment of lean-associated microbiota (Ng et al., 2022). Notably, FMT transfers not only the healthy microbiota to recipients but also compounds of potentially dangerous microbes (Walker, 2017;Hanssen et al., 2021).

Dietary intervention
Diet and nutrition play the most important influence on how the gut microbiota and the host interact over a lifetime. The ingestion of nutrients influences the composition of microbial metabolism and serves as a substrate for it (Albenberg and Wu, 2014). Therefore, diet interventions are effective ways to alter the gut microbiota composition and influence the host's health. The earliest human diet in life is breastmilk, and a meta-analysis elucidate that breastfed infant presents a stable and Bifidobacteriadominating gut microbiota community, which is conducive to immune maturation (Uusitalo et al., 2016). A case-control study that found a lengthy breastfeeding time to be linked to a lower risk of T1DM suggests that exclusive and long-term breastfeeding is an independent protective factor for T1DM (Rosenbauer et al., 2008).
It has been reported that an animal-based diet reduced the abundance of metabolized plant polysaccharides Firmicutes, leading to a reduction in beneficial SCFAs (David et al., 2014), while high dietary fiber intake selectively promoted the growth of a group of SCFA-producing gut microbiota and increased HbA1c levels possibly by increasing the production of GLP-1 . After 1 month of a vegan diet, obese patients with T2DM and/or hypertension presented a reduced F/B ratio and increased Clostridium and Bacteroidetes fragile, and they experienced significant reductions in HbA1c and triglyceride levels, weight loss, and improved fasting and postprandial glucose levels (Kim et al., 2013).In addition, high-fat diets regulate the composition of gut microbiota, mainly reducing the number of Bifidobacterium (Marietta et al., 2013). F/B ratio renewed while obese patients accepted carbohydrate-restricted or fat-restricted low-calorie diets (Cotillard et al., 2013;Bouter et al., 2017). Fiber-rich diet contributes to increasing Prevotella, whereas a protein-rich diet is associated with an increased abundance of Bacteroides. (Hartstra et al., 2015).
Concretely, green tea, caffeine, and omega-3 polyunsaturated fatty acids are beneficial for restoring the changed gut microbiota composition (Lau and Wong, 2018;Pascale et al., 2018). Consuming guar gum promoted T2DM patients to experience lipid-lowering effects (Li et al., 2021). Gluten intake promotes T1DM development by altering gut microbiota composition and immune response resulting in β -cell damage (Marietta et al., 2013). Zinc deficiency influences the inflammatory response and metabolic control (Xia et al., 2017), whereas Vitamin A deficiency increases the F/B ratio and decreases butyrate-producing gut microbiota level (Tian et al., 2018).

Antidiabetic drugs
The gut microbiota interacts with various popular hypoglycemic drugs, such as metformin, liraglutide, acarbose, and thiazolidinedione (Gurung et al., 2020;Lee et al., 2021;Smits et al., 2021;Takewaki et al., 2021). The antihyperglycemic effect of metformin has been traditionally attributed to its direct action on the signaling process in liver cells, leading to lower hepatic gluconeogenesis. However, it has been reported that metformin can alleviate the reduction of butyrate-producing microbiota, and gut microbiota supports the therapeutic effects of metformin through SCFAs production (Forslund et al., 2015). Treating the newly diagnosed T2DM naively with metformin can increase the amount of the bile acid GUDCA, thus improving IR (Sun et al., 2018). A randomized controlled trial showed that metformin strongly affected the gut microbiota and germ-free mice that received a transfer of fecal samples from metformintreated donors presented improved glucose tolerance . In a rodent model, the level of Lactobacillus in the upper small intestine was increased, and sodium-glucose cotransporter-1 (SGLT1) expression was restored, thus increasing glucose sensitivity after using metformin (Bauer et al., 2018).

Conclusion and perspectives
In conclusion, the gut microbiota plays an important role in the occurrence and development of DM, and its composition in both T1DM and T2DM patients differs from that of healthy individuals. In general, opportunistic pathogens are usually reported increasing in DM patients, while probiotics are decreased. This review discusses the role of immune responses, inflammatory responses, metabolic disorders, and other aspects caused by gut microbiota dysbiosis in the pathogenesis of DM, enriching the content of intervention of DM through gut microbiota. This article also summarizes the progress of microbiological therapy in the prevention and treatment of DM, especially the current situation and application prospect of FMT in DM. Nowadays, there are a variety of drugs available to intervene in DM, but these drugs cannot sustainably and consistently prevent the progression of β-cell failure after the onset of DM. Therefore, it is of profound significance to intervene in DM through gut microbiota. In the future, we will verify the effectiveness of gut microbiota intervention in DM through clinical trials, and further, explore the advantages of gut microbiota intervention in DM.
Frontiers in Microbiology 09 frontiersin.org Author contributions JY and ZW were responsible for the design and manuscript writing. JY, ZW, and YZ were responsible for editing the structure of the article, obtaining the documents, and further arranging the manuscripts. YS and WL were responsible for the supervision, review, and final editing of the manuscript. All authors contributed to the article and approved the submitted version.