Diabetes mellitus (DM) is a common, serious, and costly disease, which is a major public health issue (1). In recent decades, the global prevalence of DM has increased from 4.7% in 1980 to 8.5% in 2014, while the number of people with diabetes has risen from 108 million in 1980 to 422 million in 2014 (1). In 2012, 1.5 million deaths were directly caused by DM, which has been projected as the seventh leading cause of death in 2030 (1). A previous study has reported that type 2 diabetes mellitus (T2DM) comprises the majority of the people with DM around the world, approximately 90% of DM patients are T2DM patients (2).

Patients with DM are prone to various infections, with the urinary tract as the most common infection site (3, 4). Urinary tract infection (UTI) in hospitalized DM patients was nearly two times higher than that caused by other factors (5). The probability of UTIs in T2DM patients was 60% higher than that in non-DM individuals. A survey showed that the incidence of UTIs in T2DM patients was 46.9 per 1,000 person-years, compared with 29.9 for non-T2DM subjects (6).

Cytokines are small, soluble proteins produced by various cells in response to infections and inflammation (7). Interleukin (IL)-8, a potent proinflammatory chemokine and activator of neutrophils, can be stimulated by lipopolysaccharide, IL-1, and tumor necrosis factor alpha (TNF-α). Previous studies have found that interleukin-8 (IL-8) causes migration of neutrophils to the place of inflammation, leading to pyuria in patients with UTIs (8, 9). Elevated urine levels of IL-8 were detected in febrile children with UTIs compared to children with asymptomatic bacteriuria (7). It is reported that urinary IL-8 can be identified as a potential novel biomarker for the diagnosing of UTIs with 93% sensitivity and 90% specificity (10, 11). Ko et al. found that confirmation of the presence of bioactive IL-8 in urine suggests the participation of IL-8 in UTI, providing additional evidence of the role of IL-8 in inflammation (12). Consequently, the measurement of urinary IL-8 is thought to be a potential bioindicator of the localization and severity of inflammation within the urinary tract (9). Epithelial cell lines secrete interleukin in response to stimulation with bacteria (13). Therefore, it is possible that urinary microbiota modulate the presence and levels of IL-8 in the urinary tract.

Previous studies show that the production of cytokines is related to human microbiota. For instance, Schueller et al. found that Veillonellaceae and Neisseriaceae are the most abundant indicators for high IL-8 status, whereas Erysipelotrichaceae is the most abundant bacteria in low IL-8 subjects in an oral microbiota study (14). One study that explores the cytokine secretion of vaginal epithelial cells induced by commensal bacteria shows that colonization of parallel multilayer cultures with Staphylococcus epidermidis resulted in a significant increase in IL-8 relative to non-colonized cultures (15). Cervicovagina those dominated by Gardnerella and Prevotella induced higher levels of IL-8, whereas Lactobacillus iners induced moderate IL-8 secretion and Lactobacillus crispatus did not elicit IL-8 secretion (16). On the other hand, IL-8 in combination with other cytokines plays crucial roles in regulating the microbiota. Previous studies have shown that IL-6 and IL-10 play essential roles for the maintenance of intestinal homeostasis and the prevention of colitis (17, 18), while proinflammatory cytokines, such as IL-17 and IL-15 promote intestinal dysbiosis associated with increased susceptibility to colitis (19, 20). In addition, levels of IL-α, IL-β, and IL-6 are able to stimulate the expression of antimicrobial peptides, which help to shape the skin microbiota and have a protective role against potential pathogen attack (21).

Seventy female T2DM patients were recruited in our previous study, since female patients are known to have higher prevalence of UTI than males (22). We explored whether dysbiosis of urinary microbiota was related to T2DM females. We found that both diversity and richness of urinary microbiota declined in T2DM patients. Actinobacteria, Flavobacteriales, and Flavobacteria were recognized as potential distinguishing biomarkers for T2DM patients. Fourteen bacterial genera, including Lactobacillus, Prevotella, and Streptococcus, were enriched in the T2DM patients, while 19 bacterial genera, such as Pseudomonas, Klebsiella, and Akkermansia, decreased. Meanwhile, the relative abundance of Actinobacteria, Lactobacillus, and Akkermansia muciniphila correlated to the levels of fasting blood and urine glucose (23). The correlations between cytokines and microbiota from intestine, oral cavity, vagina, and skin have been extensively investigated, while the relationship between IL-8 and urinary microbiota remains poorly studied. Here, we investigated whether dysbiosis of urinary microbiota was associated with the presence of urinary IL-8, which might be useful to explore the interactions between the urinary microbiota and immune system and shed light on potential novel diagnosis and therapy for UTIs in T2DM patients.

Type 2 diabetes mellitus patients are prone to a higher occurrence of certain infections compared with the healthy population (27). The hospitalization rate for UTIs caused by diabetes is over twice those caused by other factors (5). Defects in maintaining the integrity of mucosal barriers can result in systemic endotoxemia that contributes to chronic low-grade inflammation (28). Recent advances in understanding the pathophysiology of T2DM have established the involvement of low-grade inflammation due to an increased production of proinflammatory cytokines (29, 30). This study focused on the correlations between urinary microbiota and the proinflammatory chemokine IL-8 for the first time, with the aim of providing new insights on host–urinary microbiota interactions in T2DM patients.

Urine specimens from 46 T2DM patients had detectable levels of IL-8 (WIL8 group), while IL-8 was not detected in 24 T2DM patients (NIL8 group). The alpha diversity indices, such as Shannon and Simpson, did not show significant differences between the NIL8 and WIL8 groups, indicating that bacterial diversity was not affected by the presence of IL-8 in urine. The beta diversity index such as PCoA analysis indicated that most of the patients with urinary IL-8 formed their own cluster, which reflected a contribution from IL-8 was prominent in differentiating urinary microbiota in groups.

Interestingly, the distinguishing biomarker, Bifidobacteriaceae, which is associated with diabetic patients with higher BMIs (31), had a higher abundance in WIL8 subjects, and those patients also had higher BMIs than NIL8 participants in our study. Elfeky et al. reported that exosomes increased IL-8 release from endothelial cells, and the effect was even higher when exosomes were isolated from obese women compared to lean subjects (32). Thus, BMI might play a role in regulating IL-8 levels, and subsequently IL-8 modulates the abundance of urinary Bifidobacteriaceae.

A higher abundance of Proteobacteria was found in the WIL8 group in this study. Fricke et al. and Rani et al. have demonstrated that Proteobacteria in urine decreased in patients that had renal transplantation and were exposed to high dose immunosuppressant (33, 34), during which cytokine production might be inhibited by the suppressing T cells induced by administrating the immunosuppressant. In addition, a recent study on intestinal microbiota reported that Proteobacteria was detected in feces of rats at the peak of experimental autoimmune encephalomyelitis (35). Demmer et al. demonstrated that inflammation explained 30–98% of the observed associations between levels of microbiota in a subgingival microbiome study, and the percentages of the overall phyla associations with inflammation were 27% for Proteobacteria (36). The above findings illustrated that Proteobacteria might be correlated to the onset or development of the inflammatory process. It is demonstrated that a rise in species belonging to the phylum Proteobacteria may have a larger impact on host autoimmunity which may make a protein molecule non-functional and thereby may be involved in the onset of inflammatory disorders, including diabetes (37).

The relative abundance of Bacteroidetes was lowered in the WIL8 group. A similar alteration was found in a study that showed that patients with urgency urinary incontinence had decreased abundance of Bacteroidetes compared to HCs (38). The patients suffering from renal transplantation and successive immunosuppressing therapy also had lowered levels of Bacteroidetes (34). Bacteroidetes can suppress enteric inflammation, suggesting that members of the phylum play a similar role in regulating the level of IL-8 in urinary tract system. Interestingly, there was a noticeable decrease in the phylum Bacteroidetes in newly diagnosed diabetics (39), and in our present study, we observed that patients with detectable IL-8 had T2DM for a longer duration than the NIL8 patients, suggesting that in the early state of diabetes, Bacteroidetes play a minor role in regulating the inflammatory process in diabetes.

We found that the genus of Corynebacterium in the urinary microbiota was enriched in WIL8 patients. A recent study reported that the prevalence of Corynebacterium correlated with concentrations of IL-6 and C-reaction protein in cancer patients (40). In our present study, the WIL8 group had a higher level of urine white blood cells, more leukocyte esterase and urine nitrite-positive cases, which indicated an inflammatory reaction was present in the urinary tract system, although no UTI was diagnosed currently. This might suggest that an inflammatory reaction and its correlation with urinary microbiota is established before a patient is diagnosed with UTI and the presence of obvious clinical manifestation, thus early detection of the interaction of the inflammatory response with urinary microbiota is valuable for clinicians to diagnose and treat UTIs in the early stages.

A higher abundance of A. muciniphila was correlated with a higher concentration of IL-8. Moreover, Akkermansia in the “≥HCs” group had a higher level of IL-8 than the “<HC” group. An in vitro model demonstrated that A. muciniphila could induce IL-8 production by enterocytes at cell concentrations 100-fold higher than those for Escherichia coli (41). Another study demonstrated that mice with low levels of inflammation were enriched for Akkermansia (42). Surprisingly, the patients from the WIL8 group had higher levels of urine leukocyte esterase and nitrite than the NIL8 patients, suggesting Akkermansia might play important roles in protecting patients away from UTIs, since patients in either WIL8 or NIL8 groups were not currently diagnosed with UTIs while they were recruited to the present study.

Lin et al. reported that the overgrowth of Enterococcus in diabetic mice was accompanied with increased IL-1β and TNF-α expression from Kupffer cells in intestine (43). Interestingly, our data indicated that Enterococcus was dramatically increased in the WIL8 group. Therefore, the abundance of Enterococcus might be regulating the level of cytokines in diabetic populations. Nienhouse et al. reported that Pseudomonas was enriched in positive urine culture specimens comparing to negative specimens. Also, Pseudomonas might cause the most severe inflammation which was accompanied by an increase in the number of inflammatory cells and IL-6 (44). This trend was similar to our study, in which patients with Pseudomonas in the “≥HCs” group had a higher level of IL-8. Interestingly, Pseudomonas aeruginosa were the predominant isolates of non-healing ulcers in diabetic foot patients (45), and diabetic foot infection has demonstrated higher concentrations of IL-6 and IL-1β than controls (46). Thus, Pseudomonas might be correlated to the inflammatory process in diabetic patients and might regulate the levels of cytokines. Klebsiella, being a recognized uropathogen (47), was enriched in female participants who developed UTIs after pelvic floor surgery (48). Interestingly, patients with Klebsiella in the “≥HCs” group had a higher concentration of IL-8. Huang et al. reported that diabetic patients was associated with relapse of recurrent bacteremia caused by Klebsiella pneumonia (49), suggesting that this bacteria is responsible for the inflammation process in diabetes. S. anginosus was also increased in the WIL8 group compared to the NIL8 group. A similar result was reported in the study conducted by Price et al., in which S. anginosus increased in subjects with UTIs compared to the non-UTI participants (50).

Members of the genus Lactobacillus exhibit probiotic effects in epithelial attachment, pathogen inhibition, and intestinal immunomodulation (51–53). However, their relative abundance was increased in the WIL8 group compared to the NIL8 group. Furthermore, Lactobacillus was one of the predictors of the presence of IL-8 in urine in our study. It was reported that Lactobacillus showed a rise in interstitial cystitis patients than HCs (54). However, a recent study reported that there were no associations between the presence of Lactobacillus and urinary cytokine levels in patients with interstitial cystitis (55). Therefore, Lactobacillus might play distinctive roles in regulating cytokine production in urine in different health statuses. L. iners, which may induce moderate IL-8 secretion and has moderate proinflammatory activity in the cervicovaginal bacterial community (16), was also increased in the WIL8 group. L. mucosae which has been demonstrated to possess IL-6 induction ability in macrophages (56), was found enriched in the WIL8 group. L. reuteri can suppress intestinal inflammation in a trinitrobenzene sulfonic acid-induced mouse colitis model via downregulation of gene expression of mucosal cytokine IL-6 and IL-1β in the colon (57). In another study, L. reuteri could produce molecules that had potential anti-TNF-α activity in vitro and antimicrobial compounds in diabetes (58, 59). Furthermore, intake of L. reuteri can increase insulin secretion, which might be due to augmented incretin release (60). Low-grade chronic inflammation is accepted as an internal metabolic adaptation pathway in T2DM, so the increased levels of members of the genus Lactobacillus might be believed to be players in controlling the development of inflammation in diabetes. Moreover, patients with the Lactobacillus “≥HCs” group had lower levels of IL-8 than the “<HCs” group, which also demonstrated that Lactobacillus contributed to inhibit the inflammatory process in the bladder.

In total, 18 bacterial genera contributed to the presence of IL-8 in the urine of T2DM patients. Most of these genera have not been reported by human urinary microbiota studies, except for Actinomyces (61) and Lactobacillus (47, 50, 61–65). Furthermore, very few of these genera have been reported in terms of their correlations with inflammatory cytokines. Interestingly, the abundance of Ruminococcus was positively correlated with the concentration of urinary IL-8, and can be considered as an important contributor to the presence of IL-8 as well. Ruminococcus spp. could induce dominant effector ex vivo mesenteric lymph node T-helper 17 responses (66), which is correlated to inducing the expression of IL-8 (67). In addition, Ruminococcus was more abundant in diabetic mice, and correlated negatively with delayed diabetes onset age (68). Thus, Ruminococcus might play a role in diabetes development by regulating the level of cytokines. Actinomyces spp. has been demonstrated to induce inflammatory cytokines by researchers (69), which were linked to the presence of urinary IL-8 in the present study suggesting therapy for inflammatory development in diabetes involving urinary microbiota should take a considerable to the 18 bacteria included in multiple analysis model.

There were several limitations in our study. First, the sample size in the WIL8 and NIL8 groups was not equal, which might affect the reliability of the results. Second, although all participants were not currently diagnosed with UTIs, we could not completely rule out the influence caused by previous occurrence of UTIs since it takes 6 months for diabetic patients to revert to normal glomerular filtration rate trends after an infection is cured (70); this might affect the growth environment of urinary microbiota and urine IL-8.

To our knowledge, this is the first study focused on the associations between urinary microbiota and concentrations of urine IL-8 in T2DM patients. The bacterial community showed differences in the WIL8 and the NIL8 groups, and the concentrations of IL-8 were different in the “≥HCs” group and “<HCs” group. Findings from this study indicated that urine IL-8 is interplayed with urinary microbiota among T2DM patients. Future studies should focus on how the urinary microbiota affects the inflammatory cytokine excretion in the urinary tract, which might be conducive to explore novel therapies to regulate inflammation in T2DM patients.

Ethics Committee of the First Affiliated Hospital, School of Medicine, Zhejiang University approved the study (Reference Number: 295).

LL and FL conceived and designed the study. ZL generated the sequencing data. FL and SL collected the samples. FL and YC conducted urine cultures and the urinalysis. FL extracted the bacterial DNA. ZL and FL analyzed the data, carried out the computational analysis, interpreted the data, and drafted the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.