Hypothesis / aims of study
Overactive bladder (OAB) reduces quality of life physically and mentally due to symptoms such as urinary urgency and urinary frequency. OAB patients are mainly treated with antimuscarinic agents or β3 agonists. However, these treatments cannot be curative and are often discontinued in a year because of lack of efficacy. For better treatment, the pathophysiology of OAB needs to be elucidated. Next-generation sequencing has made it possible to study microbiota without culture and has revealed the existence of microbiota in urine and on the bladder wall1). Clinically, the presence of bacteria in the urine is not considered a problem and is not treated unless the urine is pyuria. However, we previously revealed that the instillation of lipopolysaccharide from Escherichia coli in mouse bladder increases urothelial ATP release and induces urinary frequency2). We hypothesized that urinary microbiota causes OAB by the similar mechanism even if it does not develop cystitis. Based on this hypothesis, we aimed to clarify the characteristics of urinary microbiota associated with OAB.
Study design, materials and methods
First, we examined the association between urinary bacterial count and the prevalence of OAB. We collected catheterized urine from non-neurogenic OAB patients and asymptomatic subjects between November 2019 to October 2020. Patients with a malignant tumor, patients with pyuria, and patients receiving antibiotics within 6 months were excluded from this study. We immediately centrifuged catheterized urine, and decanted the supernatants, leaving 1/25th volume. We used the centrifuged urine to prepare urine smears and Gram-stained them. We evaluated bacterial counts of urine smears by observing 5 fields of view under 400x magnification and if the count was 100 or more, it was classified into a bacteria-rich urinary microbiota group (BR group), and if the count was less than 100, it was classified into a non-bacteria-rich urinary microbiota group (non-BR group). The prevalence of the BR group was compared between OAB patients and healthy subjects.
Second, we investigated the characteristics of the urinary microbiota of the BR group. We analyzed the microbiota of the above-mentioned urine samples by 16Smetagenomic analysis if the remaining urine sample was ≧1 mL. The microbiota was compared between the BR and non-BR groups. Alpha diversity was evaluated by rarefaction analysis of the Chao1 index, Faith’s phylogenetic diversity (PD) index, and Shannon index. Beta diversity was assessed via principal coordinate analysis (PCoA) based on weighted UniFrac distance and analysis of similarities. Linear discriminant analysis effect size (LEfSe) was performed to identify significant differences in operational taxonomic units (OTUs) between the BR and non-BR groups. The threshold on the logarithmic LDA score for discriminative features was 5.0.
Results
We recruited 48 participants, 35 were non-neurogenic OAB patients and 13 were healthy subjects. There were 7 participants classified into the BR group and 41 participants classified into the non-BR group. The median age (range) of the BR and non-BR groups was 75 years (51-91) and 75 years (72-84). The prevalence of the BR group was 20% among the OAB patients (7/35) and 0% among the healthy subjects (0/13). The prevalence of bacteria-rich urinary microbiota tended to be higher in the OAB patients than in the healthy subjects.
We analyzed the microbiota of 39 urine samples. Of the 39 cases, 29 were OAB patients and 10 were healthy subjects. Also, 5 cases were in the BR group and 34 were in the non-BR group. There was no difference between the BR and non-BR groups in Chao1 index (median (range), 130 (11-520) vs 485 (32-882), p=0.08) and Faith’s PD index (22 (5.8-45) vs 44 (11-56), p=0.07). The microbiota of the BR group had significantly lower Shannon index than that of the non-BR group (2.6 (2.1-5.9) vs 7.4 (2.5-8.8), p=0.008). PCoA showed a significant difference between the microbiota of the BR and non-BR groups (p<0.01) (Fig. 1). LEfSe analysis revealed that 6 operational taxonomic units (OTUs) had significantly higher abundance in the BR group than in the non-BR group (Fig. 2). Of the 6 OTUs, 2 OTUs were the genus Escherichia and an unannotated genus classified in the same family as the genus Escherichia. The other 4 OTUs were taxonomic names of their upper hierarchies. The proportion of the genus Escherichia and related genus together in the BR and non-BR groups was 56% (8.8-94) and 3.6% (0.0-95), respectively. LEfSe analysis also revealed that 3 OTUs had significantly lower abundance in the BR group than in the non-BR group. One of the 3 OTUs was the order Clostridiales and the other 2 OTUs were taxonomic names of its upper hierarchies. The proportion of the order Clostridiales in the BR and non-BR groups was 3.2% (0-22) and 43% (2.1-59), respectively.
Interpretation of results
We demonstrated that some OAB patients have bacteria-rich urinary microbiota, which was characterized by a high composition of the genus Escherichia and a related genus.
The participants of the BR group occupied 20% of OAB patients and did not occupy healthy subjects. It is suggested that urinary microbiota is the cause of some OAB patients.
According to the comparison of the microbiota, the microbiota of the BR group consisted of more biased bacteria than that of the non-BR group and had different bacterial composition from that of the non-BR group. These findings are mainly due to the dominance of the genus Escherichia and a related genus in the microbiota of the BR group with a median value of 56%. As the microbiota on the bladder wall was reported to be similar to urinary microbiota1), the bladder wall of some OAB patients could have higher numbers of closely related species of Escherichia coli. Lipopolysaccharide from Escherichia coli in the mouse bladder is reported to increase urothelial ATP release and induces urinary frequency2). Therefore, in humans, the increase of the genus Escherichia and a related genus may increase urothelial ATP release by the similar mechanism and cause OAB symptoms.