Characteristic dysbiosis in patients with type 2 diabetes and hyperuricemia, and the effect of empagliflozin on gut microbiota

Abstract

BACKGROUND

Gut microbiota play a crucial role in metabolic diseases, including type 2 diabetes (T2DM) and hyperuricemia (HUA). One-third of uric acid is excreted into the intestinal tract and further metabolized by gut microbiota. Thus, the gut microbiota might be a new therapeutic target for HUA. Empagliflozin significantly lowers serum uric acid levels and contributes to cardiovascular benefits which are partly attributed to altered gut microbiota. We hypothesize that gut dysbiosis in patients with diabetes and HUA, and the reduction of uric acid by empagliflozin, may be mediated by gut microbiota.

AIM

To investigate dysbiosis in patients with T2DM and HUA, and the effect of empagliflozin on gut microbiota associated with purine metabolism.

METHODS

In this age and sex-matched, case-control study, we recruited 30 patients with T2DM and HUA; 30 with T2DM; and 30 healthy controls at the Henan Provincial People’s Hospital between February 2019 and August 2023. Nine patients with T2DM and HUA were treated with empagliflozin for three months. Gut microbiota profiles were assessed using the 16S rRNA gene.

RESULTS

Patients with T2DM and HUA had the highest total triglycerides (1.09 mmol/L in heathy control vs 1.56 mmol/L in T2DM vs 2.82 mmol/L in T2DM + HUA) and uric acid levels (302.50 μmol/L in heathy control vs 288.50 μmol/L in T2DM vs 466.50 μmol/L in T2DM + HUA) among the three groups. The composition of the gut microbiota differed significantly between patients with T2DM and HUA, and those with T2DM/healthy controls (P < 0.05). Notably, patients with T2DM and HUA demonstrated a deficiency of uric acid-degrading bacteria such as Romboutsia, Blautia, Clostridium sensu stricto 1 (P < 0.05). Empagliflozin treatment was associated with significantly reduced serum uric acid levels and purine metabolism-related pathways and genes in patients with T2DM and HUA (P < 0.05).

CONCLUSION

Gut dysbiosis may contribute to the pathogenesis of HUA in T2DM, and empagliflozin may partly restore the gut microbiota related to uric acid metabolism.

Key Words: Type 2 diabetes; Hyperuricemia; Uric acid; Empagliflozin; Gut microbiota

Core Tip: Patients with type 2 diabetes (T2DM) have a significantly higher prevalence of hyperuricemia (HUA) than non-diabetic patients and are more likely to suffer from cardiovascular diseases. In recent years, the gut microbiota has been shown to play a crucial role in metabolic diseases, including T2DM and HUA. Thus, the gut microbiota may be a new therapeutic target for HUA. Empagliflozin significantly lowers serum uric acid levels and contributes to cardiovascular benefits which are partly attributed to altered gut microbiota. This study revealed that empagliflozin administered to patients with characteristic dysbiosis due to T2DM and HUA, may have gut microbiota involved in purine metabolism partially restored.

INTRODUCTION

Hyperuricemia (HUA) is a metabolic disorder characterized by abnormal purine metabolism, resulting in a sequence of tissue damage such as gout, atherosclerosis, and chronic kidney disease[1]. The global prevalence of HUA has risen to 15%-20%[2,3], making it the second most common metabolic disease after type 2 diabetes (T2DM)[4]. Emerging evidence indicates that gut microbiota play an important role in the pathogenesis of T2DM and cardiovascular diseases[5,6]. Among the commonly reported findings, the genera of Ruminococcus, Fusobacterium, and Blautia were abundant in patients with T2DM with a low diversity[7]. In patients with T2DM, the risk for atherosclerosis which is associated with elevated serum uric acid levels is partly driven by gut microbiota[8]. Radioisotope studies involving healthy individuals revealed that approximately one-third of uric acid is excreted into the intestinal tract and further metabolized by the gut microbiota, a process that is doubled in proportion in patients with kidney disease[9]. Thus, gut microbiota is a new therapeutic target for HUA[4,9].

In patients with T2DM, empagliflozin, which is a selective inhibitor of sodium-glucose cotransporter 2, mitigates hyperglycemia via reducing renal glucose reabsorption and enhancing urinary glucose excretion[10]. In patients with or without T2DM, empagliflozin significantly lowers serum uric acid levels and contributes to cardiovascular benefits which are partly attributed to altered gut microbiota[11,12]. However, the association between empagliflozin treatment, uric acid metabolism, and gut microbiota remains unclear. Therefore, the aim of the study was to investigate gut dysbiosis and the effect of empagliflozin treatment on gut microbiota associated with purine metabolism in patients with T2DM with HUA.

MATERIALS AND METHODS

Recruitment of study participants

The study approved by the Ethics Committee and Committee for Clinical Investigation of Henan Provincial People’s Hospital (Henan Province, China), and registered in the Chinese Clinical Trial Registry (No. ChiCTR1800018825 and No. ChiCTR2000030049) was performed following the principles of the Declaration of Helsinki. All participants provided written informed consent.

We consecutively recruited 30 patients with T2DM and HUA; 30 patients with T2DM; and 30 healthy controls at the Henan Provincial People’s Hospital between February 2019 and August 2023. The participants in the three groups were age and sex-matched. Patients were included on the basis of the following criteria: (1) 18 to 70 years of age; (2) T2DM was defined by the 1999 World Health Organization with the typical history of hyperglycemia without requiring for immediate insulin treatment or positive islet autoantibodies[6,13], and HUA was defined as individuals with a fasting serum uric acid level was higher than 420 μmol/L in men and 360 μmol/L in women[1]; and (3) No history of type 1 diabetes mellitus, immune disorders, or other severe diabetic complication diseases. Exclusion criteria were as follows: (1) Secondary diabetes; (2) Infectious diseases; (3) Acute or chronic inflammatory disease; (4) Pregnancy; (5) Malignant tumors; (6) History of steroid or immunosuppressive drug use > 7 days; (7) History of treatment with prebiotics, probiotics, antibiotics, or any other medication that could potentially influence the gut microbiota for > 3 days in the previous three months; (8) A history of hepatic and renal malfuncting; (9) Gastrointestinal diseases; and (10) Gastrointestinal surgery in the previous year[6].

Nine individuals with treatment-naive T2DM and HUA who participated in the case-control analysis were recruited for the intervention analysis. These nine participants were treated with empagliflozin (10 mg/day, Boehringer Ingelheim Pharma, Ingelheim am Rhrin, Germany) for 3 months as previously described[12].

All patients with T2DM were educated on glycemic control. Data on health status, lifestyle, medical history, and medication use were obtained using standard questionnaires. Anthropometric and metabolic assessments were performed, and blood, urine, and fecal samples were obtained at baseline, and after empagliflozin treatment. Blood, urine, and fecal samples were frozen in dry ice immediately after collection and then stored at -80 °C.

Statistical analysis

Statistical analyses were conducted using STATA (version 15.0; STATA Corp., College Station, TX, United States). Continuous variables are presented as the means ± SDs or medians with interquartile ranges for normal distribution data or non-normal distribution data, respectively. Categorical variables are presented as numbers (proportions). One-way analysis of variance or Kruskal-Wall is test was used to compare the differences between groups. Significant difference were adjusted by Bonferroni correction. The comparison in continuous variables before and after empagliflozin treatment were conducted using paired-sample t-tests or the Wilcoxon matched-pairs signed-rank test.

DNA extraction and 16S rRNA gene sequencing

Genomic DNA was extracted from fecal samples using the QIAamp Power Fecal Pro DNA kit (51804; QIAGEN, Hilden, Germany). Gut microbiota in the fecal samples were profiled using 16S rRNA gene amplicon sequencing.

A polymerase chain reaction targeting the V3-V4 region of the 16S rRNA gene was performed using forward (5’-CCTACGGGNGGCWGCAG-3’) and reverse (5’-GACTACHVGGGTATCTAATCC-3’) primers[14]. Subsequent amplicon sequencing was performed on a MiSeq platform to generate paired-end reads of 300 base pairs each in length (Illumina, San Diego, CA, United States).

Analysis of sequencing data

Three batches of sequencing data were used in case-control analysis, and the sequences were analyzed using QIIME2 version 2023.2 respectively[15]. The adapters of the original sequences were removed using the “cutadapt” plugin of QIIME2. Sequences were then truncated with DADA2 and further filtered, denoised, cleared of chimeras, and merged to obtain the abundance and representative sequences of amplicon sequence variants (ASVs)[16]. Subsequently, the ASVs of the three batches were merged using the QIIME2. Representative sequences for ASVs were built into a phylogenetic tree using the core-metrics-phylogenetic pipeline in QIIME2 and assigned to taxonomy using SILVA database (release 138)[17]. Since the detection was carried out in different batches, the ComBatseq function of the R package was used to remove the batch effect (without adding grouping parameters) firstly. All samples were then randomly subsampled to equal depths of 9070 reads before the subsequent fecal microbiome analysis using QIIME2 diversity plugins.

In the intervention analysis, the sequences were analyzed using QIIME2 version 2023.2 as described above. All samples were randomly subsampled to equal depths of 23212 reads before the subsequent fecal microbiome analysis using QIIME2 diversity plugins.

The raw Illumina sequence data in this study are available in the sequence read archive at National Center for Biotechnology Information (NCBI) under accession No. SRP513322.

For the α-diversity, microbiome diversity was evaluated using the Shannon index, and the microbiome richness was evaluated by the number of observed ASVs using QIIME2 plugin diversity (-core-metrics-phylogenetic). For the analysis of β-diversity, QIIME2 plugin diversity (-core-metrics-phylogenetic) was used to perform the principal coordinate analysis (PCoA) based on Bray-Curtis distances. Also, the diversity plugin was used to perform the permutational multivariate analysis of variance test (999 tests) to test the significance of the differences of the gut microbiota structure between groups. The unsupervised partial least squares discrimination analysis (PLS-DA) was performed using R package mixomics[18]. We used LEfSe (R package lefser) to determine the ASVs that were significantly differentiated between groups.

Microbiome co-abundance variation cluster classification

The sparse correlation for composite data algorithm was used to calculate the correlation between ASVs in the case-control study with a sample co-occurrence rate greater than 10%, and the correlation value was converted into a correlation distance (1-correlation value). Then Ward clustering algorithm was then used to cluster R-package weighted correlation network analysis, and co-abundance groups (CAGs) were obtained. The CAG network was visualized in the Cytoscape (selection criteria: 1, P value < 0.05; 2, absolute value of r value > 0.3). Differences in CAGs between groups were tested using ordinary one-way analysis of variance test combined with Tukey’s multiple comparisons in the case-control analysis, and paired t test in the intervention study. GraphPad Prism 9 was used to visualize the differences in CAGs.

Prediction of microbiome metabolic pathways

PICRUST2 was used to predict the metabolic functions of microbiota based on the representative sequences obtained by QIIME2 using the Kyoto encyclopedia of genes and genomes (KEGG) and Meta CYC databases[19]. The different metabolic pathways between the groups were selected by performing ordinary one-way analysis of variance combined with Tukey’s multiple comparisons in case-control analysis, and paired t-test in the intervention analysis. GraphPad Prism 9 was used to visualized different pathways and KEGG Orthologys.

RESULTS

Anthropometric and biochemical measurements

Thirty patients with T2DM and HUA, 30 patients with T2DM, and 30 healthy controls were age- and sex-matched (Table 1). Expectedly, the body mass index (BMI), waist-hip ratio, systolic blood pressure, fasting plasma glucose, glycated hemoglobin (HbA1c), and serum high-density lipoprotein cholesterol were significantly higher in T2DM patients with or without HUA, than in healthy controls. The patients with T2DM and HUA had significantly higher serum uric acid levels.

Table 1 Anthropometric and biochemical data of the study participates in case-control population, mean ± SD.

	HC (n = 30)	T2DM (n = 30)	T2DM + HUA (n = 30)	P value (HC vs T2DM)	P value (HC vs T2DM + HUA)	P value (T2DM vs T2DM + HUA)
Age (year), medians IQR	36.5 (30.0, 48.0)	36.0 (30.0, 48.0)	35.5 (29.0, 48.0)	1.000	1.000	1.000
Male, n (%)	21 (70)	21 (70)	21 (70)
BMI (kg/m2)	23.33 ± 3.01	26.08 ± 3.41	27.96 ± 4.26	0.012	< 0.001	0.139
WHR	0.87 ± 0.07	0.93 ± 0.08	0.95 ± 0.07	0.007	< 0.001	1.000
SBP (mmHg), medians IQR	114.00 (110.00, 127.00)	125.00 (115.00, 135.00)	129.50 (120.00, 136.00)	0.032	< 0.001	0.203
DBP (mmHg)	73.43 ± 8.84	77.50 ± 9.96	81.73 ± 9.54	0.299	0.003	0.260
FPG (mmol/L)	5.01 ± 0.57	8.06 ± 3.33	8.20 ± 2.75	< 0.001	< 0.001	1.000
HbA1c (%), medians IQR	4.92 (4.60, 5.40)	6.90 (6.20, 9.25)	7.79 (6.32, 8.90)	< 0.001	< 0.001	1.000
TG (mmol/L), medians IQR	1.09 (0.93, 1.51)	1.56 (1.00, 2.77)	2.82 (1.49, 4.72)	0.085	< 0.001	0.004
TC (mmol/L)	4.47 ± 0.91	4.65 ± 0.86	5.05 ± 0.97	1.000	0.047	0.281
LDL-C (mmol/L)	2.75 ± 0.59	2.60 ± 0.78	2.87 ± 0.69	1.000	1.000	0.415
HDL-C (mmol/L), medians IQR	1.26 (1.13, 1.52)	0.97 (0.85, 1.16)	0.89 (0.79, 1.07)	< 0.001	< 0.001	0.575
Creatine (μmol/L)	58.93 ± 10.45	55.24 ± 13.44	57.95 ± 13.09	0.767	1.000	1.000
UA (μmol/L), medians IQR	302.50 (263.00, 351.00)	288.50 (257.00, 325.00)	466.50 (422.00, 513.00)	0.543	< 0.001	< 0.001

IQR: Interquartile ranges; HC: Healthy control; T2DM: Type 2 diabetes mellitus; HUA: Hyperuricemia; BMI: Body mass index; WHR: Waist–hip circumference ratio; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; FPG: Fasting plasma glucose; HbA1c: Glycated hemoglobin; TG: Total triglycerides; TC: Total cholesterol; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol; UA: Uric acid.

Altered gut microbiota in patients with T2DM and HUA

No differences in the richness (observed ASVs), evenness (Pielou index) and diversity (Shanon index) were found among the healthy subjects group; T2DM group; T2DM and HUA group (Supplementary Figure 1). The PCoA and score plots of PLS-DA suggested that composition of the gut microbiota differed significantly among the three groups (Figure 1A and B, Supplementary Figure 2). And we identified 322 ASVs and constructed ten CAGs based on sparse correlation analysis (Figure 1). The abundance of CAG3 (i.e., genus Bacteroides) was significantly higher in patients with T2DM and HUA group; whereas the abundance of CAG7 (i.e., family Lachnospiraceae) was significantly lower the group of T2DM patients with or without HUA group (Figure 1C). Additionally, serum uric acid was positively related to CAG1 (i.e., families Enterobacteriaceae and Bacteroidaceae) (Figure 1D). LEfSe analysis showed that the gut microbes in patients with T2DM and HUA significantly differed from those in the other two groups (Supplementary Figure 3). Patients with T2DM and HUA demonstrated a deficiency of uric acid-degrading bacteria such as Romboutsia, Blautia, and Clostridium sensu stricto 1 (Supplementary Figure 3). LEfSe analysis was used to further explore microbial function and identify differentially abundant KEGG pathways and genes among three groups. Notably, purine-related metabolic pathways and genes were significantly more abundant in patients with T2DM and HUA group (Figure 1E and F).

Figure 1 Characteristics of the microbiome in the cross-sectional study. A: Partial least squares discrimination analysis was used to show the difference in microbiome structures, and the permutational multivariate analysis of variance test results based on the Bray Curtis distance were labeled in the figure; B: The Bray Curtis distance between the microbiome of the two patient groups and the healthy control group. The difference between groups was tested by the student’s t-test; C: Differences of microbiome co-abundance between groups were tested by ordinary one-way analysis of variance test, combined with Tukey’s multiple comparisons and the threshold of significance was P < 0.05; D: Network diagrams of the co-abundance groups. Lines between nodes represent correlations; only correlations with magnitudes > 0.3 are drawn. Red lines mean positive correlations, and blue lines mean negative correlations. The significant linear correlations between two co-abundance groups and uric acid are shown on the right; E: Metabolic pathways related to the uric acid metabolism which showed differences among the three groups; F: Kos related to the uric acid metabolism which showed differences among the three groups. Differences were tested by ordinary one-way analysis of variance test combined with Tukey’s multiple comparisons. ¹P < 0.05 vs type 2 diabetes mellitus group. ²P < 0.05 vs healthy control group. ³P < 0.01 vs healthy control group. PLS-DA: Partial least squares discrimination analysis; PERMANOVA: Permutational multivariate analysis of variance; HC: Healthy control; T2DM: Type 2 diabetes mellitus; HUA: Hyperuricemia; CAG: Co-abundance group; ASV: Amplicon sequence variant.

Empagliflozin partly restored altered gut microbiota in T2DM with HUA

Clinical characteristics and biochemical parameters of individuals with treatment-naive T2DM and HUA before and after empagliflozin treatment are shown in Table 2. After three months’ treatment of empagliflozin treatment, HbA1c and uric acid were significantly reduced (Table 2). No differences in the richness (observed ASVs), evenness (Pielou index), and diversity (Shanon index) were found between patients before and after empagliflozin treatment (Supplementary Figure 4). Based on the Bray-Curtis distances, composition of gut microbiota after empagliflozin treatment differed significantly from that of before empagliflozin use (Figure 2A). Moreover, uric acid related to CAG1 and CAG3, which were elevated in patients in the T2DM and HUA group decreased after three months’ of empagliflozin treatment (Figure 2B). After adjusting for BMI, CAG1 was significantly associated with uric acid (Figure 2C). Empagliflozin treatment increased number of species from (Ruminococcus gauvreauii) group (Supplementary Figure 5). After empagliflozin treatment, the participants showed a significantly lower relative abundance of purine related metabolic pathways and genes than that observed before empagliflozin treatment (Figure 2D and E).

Figure 2 The characteristics of the microbiomes in the empagliflozin intervention study. A: Partial least squares discrimination analysis was used to show the differences of microbiome structures, and the permutational multivariate analysis of variance test results based on Bray Curtis distance were labeled in the figure; B: Differences of microbiome co-abundance between groups were tested by the paired t test, and the threshold of significance was P < 0.05; C: The linear correlation between co-abundance group 1 and uric acid; D: Metabolic pathways related to the uric acid metabolism among the three groups; E: Kos related to the uric acid metabolism among the three groups. Differences were tested by the paired t test. ^aP < 0.05. ^bP < 0.01. PLS-DA: Partial least squares discrimination analysis; PERMANOVA: Permutational multivariate analysis of variance; BMI: Body mass index; CAG: Co-abundance group.

Table 2 Anthropometric and biochemical data of patients received treatment of empagliflozin, mean ± SD.

	Pre-empagliflozin (n = 9)	Post-empagliflozin (n = 9)	P value
Age (year), medians IQR	38.00 (34.00, 46.00)
Male, n (%)	8 (88.9)
BMI (kg/m2)	29.47 ± 3.34	28.14 ± 3.52	0.051
WHR	0.98 ± 0.04	0.96 ± 0.03	0.140
SBP (mmHg), medians IQR	122.00 (120.00, 124.00)	120.00 (119.00, 126.00)	0.171
DBP (mmHg)	77.67 ± 6.40	74.33 ± 7.45	0.224
FPG (mmol/L)	7.34 ± 1.44	7.13 ± 1.17	0.607
HbA1c (%), medians IQR	8.40 (7.90, 8.90)	6.60 (6.40, 6.90)	0.008
TG (mmol/L), medians IQR	2.81 (1.89, 3.20)	1.86 (1.62, 2.97)	0.051
TC (mmol/L)	5.21 ± 1.08	5.20 ± 0.97	0.981
LDL-C (mmol/L)	2.82 ± 0.89	3.17 ± 0.76	0.231
HDL-C (mmol/L), medians IQR	0.87 (0.83, 1.18)	1.08 (0.87, 1.15)	0.342
Creatine (μmol/L)	62.56 ± 14.98	57.63 ± 10.23	0.271
UA (μmol/L), medians IQR	503.00 (477.00, 513.00)	370.50 (328.00, 423.00)	0.012

IQR: Interquartile ranges; HC: Healthy control; T2DM: Type 2 diabetes mellitus; HUA: Hyperuricemia; BMI: Body mass index; WHR: Waist–hip circumference ratio; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; FPG: Fasting plasma glucose; HbA1c: Glycated hemoglobin; TG: Total triglycerides; TC: Total cholesterol; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol; UA: Uric acid.

DISCUSSION

In this study, we found that gut microbiota differed significantly between patients with T2DM and HUA and those of healthy subjects and patients with T2DM alone. In patients with T2DM and HUA, empagliflozin not only ameliorated glycemic metabolism, but also yielded significant uric acid reduction. Additionally, empagliflozin treatment may have partially restored gut microbiota involving purine metabolism in patients with T2DM and HUA.

Human uricase is a pseudogene, which has been inactivated early in hominid evolution[9]. Almost 1/3 of uric acid was excreted into the intestinal tract and metabolized by the gut microbiota, such as Clostridium, Anaerostipes, and Blautia[4,9]. The Clostridiaceae family has been reported to be able to use purines such as uric acid as sole carbon, nitrogen, and energy sources[20]. Besides, Romboutsia lituseburensis maintains microbial butyrate production in the presence of uric acid in humans[21]. In this study, we found a reduced abundance of uric acid-degrading bacteria, such as species from Romboutsia, Blautia, and Clostridium sensu stricto 1 in patients with T2DM and HUA, compared to that in healthy controls. Zhang et al[22] found that individuals with elected serum uric acid levels had significantly more Escherichia Shigella and fewer Faecalibacterium and Oscillospiraceae based on the tertiles of uric acid levels[22]. And the composition of the fecal microbiota in patients with gout has also been shown to be different from that in healthy individuals. In patients with gout, Fecalibacterium, Lachnospiraceae Clostridium, Roseburia, Cytophaga, Ruminococcaceae Clostridium and Alistipes are enriched, whereas Millisia, Bifidobacterium and Enterococcus were decreased[23].

In addition to ameliorating hyperglycemia, empagliflozin has potential cardiovascular and renal benefits. What’s more, lots of studies have suggested that empagliflozin induces a rapid and sustained reduction of serum uric acid levels and clinical events related to HUA[11,24]. Empagliflozin’s decreasing serum uric acid may be due to accentuating urinary acid excretion, which is related to activation of the tubular transporter urate transporter 1 and urinary excretion of glucose and sodium[25]. However, the underlying mechanisms remain unclear. We previously found that the effect empagliflozin reducing uric acid associated with altered gut microbiota[12]. A multi-omics correlation analysis revealed that plasma uric acid was related to Lachnoclostridium and Ruminococcus gnavus in patients with T2DM treated with empagliflozin[12]. In the current study, empagliflozin significantly decreased uric acid related to CAG1, including Escherichia Shigella and Klebsiella, and affected purine metabolism-related pathways/genes in patients with T2DM and HUA. Additionally, empagliflozin treatment increased the number of species from the Ruminococcus gauvreauii group, which has been indicated to be a protective factor against chronic renal failure and diabetic nephropathy[26,27]. Empagliflozin treatment may partially restore the gut microbiome in patients with T2DM and HUA. However, this still needs to be confirmed in future studies.

Studies on the mechanisms by which empagliflozin might influence the gut microbiome were limited. Previous study suggested that empagliflozin treatment could affect the composition of gut microbiota in T2DM: Increased short-chain fatty acid-producing bacteria, such as Eubacterium, Roseburia, and Faecalibacterium, and lowered the levels of damaging bacteria, such as Escherichia Shigella, Bilophila, and Hungatella in patients with T2DM[12]. Additionally, empagliflozin also increased the levels of plasma metabolites such as sphingomyelin, but reduced glycochenodeoxycholate, cis-aconitate, and uric acid levels[12]. In obesity-related glomerulopathy C57BL/6J mice, empagliflozin reduced abundances of Firmicutes and Desulfovibrio and increased abundance of Akkermansia, and disrupted lipid metabolism which was closely associated with gut microbiota alterations[28]. Thus, the potential mechanisms by which empagliflozin might influence the gut microbiome associated with altered composition of the gut bacterium and metabolites.

The strength of our study lies in the fact that it is the first study to evaluate the effect of empagliflozin on gut microbiota in patients with T2DM and HUA. However, this study had some limitations. First, as the study was primarily observational, causality could not be established. Second, the sample size of our cohort was relatively small, although we used a strict screening process for recruitment to avoid the potential confounders. Large-scale studies are required to validate these results. Third, we investigated the effects of empagliflozin on uric acid and gut microbiota; however, no control group was used in this interventional analysis. A well-controlled trial is required to clarify that the effect of empagliflozin on uric acid reduction is associated with altered gut microbiota. Finally, batch adjustment in sequencing data analysis may have attenuated the differences among groups in the case-control analysis.

CONCLUSION

In conclusion, our study emphasizes the association between HUA and gut microbiota dysbiosis. The gut microbiota play an important role in purine metabolism in humans. Empagliflozin may partly restore gut microbiota involved in purine metabolism in patients with T2DM and HUA, suggesting that gut microbiota may be a promising target for preventive intervention against HUA in the future.