Prostate diseases and microbiome in the prostate, gut, and urine

The microbiome in various organs involves a vast network that plays a key role in the health and wellness of the human body. With recent advances in biological technologies such as high-throughput sequencing, transcriptomics, and metabolomics, it appears that the microbial signature varies dynamically among individuals, creating various roles in metabolism, local and systemic inflammation, and host immunity. Urinary and genital organs, including the prostate, seminal vesicles, and urinary bladder, are reservoirs of several bacterial, viral, and fungal communities. Accumulating evidence has suggested profound roles for the gut, urinary, and intraprostate microbiomes in genitourinary benign and malignant diseases. This review article addresses microbiome-related evidence for three major diseases involved in prostate cancer: chronic prostatitis (CP), benign prostatic hyperplasia (BPH), and prostate cancer (PCa). Symptomatic CP is known as CP/chronic pelvic pain syndrome. CP is one of the most common prostate diseases in young men, accounting for 8% of all men visiting a urologic clinic. Although oral medication is the gold standard therapy for patients with BPH, approximately 13% of men present with clinical progression within 4 years after the initiation of treatment, with 5% requiring surgical intervention. The identification of proinflammatory cytokines and pathogens responsible for the clinical progression of BPH is still underway. Several topics regarding the association between PCa and the microbiome are discussed in this review as follows: i) intraprostatic microbiome and the risk of PCa, ii) gut microbiome and PCa, iii) gut microbiome and the risk of radiation-induced side effects, iv) isoflavone intake and equol-producing intestinal flora on PCa, and v) the inhibitory effect of daidzein and equol on tumor growth and progression of PCa. Further studies are required for a comprehensive understanding between the urogenital microbiome and prostate pathogenesis to facilitate the development of preventive and therapeutic approaches for prostate diseases.


Introduction
Although bacterial infection had been thought to be pathogenic and harmful to the human body in the past, we are currently aware that microorganisms exist in every organ system of the body in a commensal and mutualistic manner 1,2 . The microbiome in various organs involves a vast network that plays a key role in the health and wellness of the human body. With recent advances in biological technologies such as high-throughput sequencing, transcriptomics, and metabolomics, it appears that the microbial signature varies dynamically among individuals, creating various roles in metabolism, local and systemic inflammation, and host immunity 3 . Over 10 years ago, the National Institute of Health sought to explore the human microbiome to better understand roles in human health and disorders by launching the Human Microbiome Project 4 . As a great achievement, the largest microbial genome databases for various body sites, including the nares, oral cavity, skin, gastrointestinal tract, breast, and urogenital tract were generated, leading to multiple pieces of evidence on a close link between the microbiome with pathological processes such as aging, atherosclerosis, obesity, diabetes, central nervous system dysfunction, and malignancies [5][6][7][8] . The urinary and genital organs, including the prostate and seminal vesicles are a reservoir of several bacterial, viral, and fungal communities 9 . Accumulated pieces of evidence have suggested profound roles of the gut, urinary, and intra-prostate microbiomes in genitourinary (GU) diseases, such as benign prostate hyperplasia (BPH), chronic prostatitis (CP), prostate cancer (PCa), and other malignancies.
Manipulation of microbiota in GU diseases has just begun to seek and develop clinical applications. This review describes the current advancements and understanding in the field of microbiome research and discusses the close relationship between microbiome signature and prostate diseases, including BPH, CP, and PCa. A better understanding of this relationship will drive further research, which can unveil the initiation and progenitors of these diseases and provide reliable predictors of disease and chemoprevention, as well as novel therapeutic modalities.

CP and microbiome
CP is one of the most common prostate diseases in young men, with an estimated prevalence of 1.8e8.2% 10 . CP is reported to be responsible for approximately 2 million outpatient clinic visits per year, including 8% of all men who visit a urologic clinic 11 . As chronic pelvic and/or GU pain is a primary component of the CP-caused condition, symptomatic CP is called CP/chronic pelvic pain syndrome (CP/CPPS) 12 . A questionnaire-based cross-sectional study enrolling 20-to 59-year-old men in Estonia demonstrated that a third of men suffering from CP/CPPS considered their own health poor, with more frequently diagnosed renal diseases, BPH, sexually transmitted infections, chronic nervous system diseases, and depression as compared with controls without CP/CPPS 13 . In addition, men with CP reported more frequent cystitis and gynecological inflammations in their partners and more occurrences of prostatitis in their relatives. CP/CPPS can have a strong negative impact on the quality of life, with a remarkable complex of comorbidities, habits, and attitudes.
One of the biggest clinical issues is the challenge faced by many clinicians in diagnosing and treating CP/CPPS effectively owing to poor comprehension. Besides genetic factors and central nervous system imbalance, the host-microbiota of the seminal vesicles, prostate, and gut and the partner's genital microbiota could potentially contribute to the onset of CP/CPPS. Researchers have focused on microorganisms in the seminal fluid, urine, and stool of patients with CP/CPPS. Comparison of the microbiota in the specimens between patients with CP/CPPS and controls differed in various urologic disorders, including CP/CPPS. Table 1 lists the previous studies addressing the possible association between CP/ CPPS and the composition of the microbiota 14-24 . Choi et al. investigated the different microbiological etiology of CP in Korean patients between a general hospital and a primary care clinic using bacterial culture and PCR detection of expressed prostatic secretion (EPS) and post-massage urine (VB3). They reported culture-positive patients in the primary care clinic were significantly higher than in the general hospital, but the number of PCR positive patients in the primary care clinic was similar to that in the general hospital 16 . Biological technologies such as high-throughput sequencing, transcriptomics, and metabolomics, have been advancing over time, and the influence of infectious factors in the onset of CP/CPPS continues to be studied. Enterobacteriaceae dominance is the classical concept of the development of CP/CPPS, supplemented with recent data, including the etiological role for some species, as well as microorganisms that are associated with sexually transmitted infections (in particular, Chlamydia trachomatis) 25,26 . New concepts related to pathological conditions are in the context of microbiota diversity in the affected organs. M€ andar et al. investigated profiles of semen microbiomes using the HiSeq2000 sequencing of the ribosomal ribonucleic acid V6 region, demonstrating that the semen of patients with CP/CPPS contained fewer health-supporting lactobacilli and had higher species diversity than that of healthy men 20 . Overall, CP/CPPS has several etiological and potentiating contributions not only from infection and inflammation, but also from central nervous system changes and mental stress, and so on. No singular cause of CP/CPPS has been identified and it is most likely a syndrome with multifactorial factors. All factors play important roles in the crosstalk between the human body and microbiome. Recent research found a significant association between disease severity and the degree of dysbiosis in microbiomes of urine and gut in patients with CP/CPPS as compared to unaffected controls 12 . Stapleton et al. conducted randomized, placebo-controlled phase 2 trial of a Lactobacillus crispatus intravaginal suppository probiotic (Lactin-V) for prevention of female recurrent urinary tract infection 27 . A significant reduction in recurrent urinary tract infection for the Lactin-V group suggested that modification of urogenital microbiome by probiotic supplementation potentially prevents symptomatic UTI in males. In the future, the successful work-up of patients with CP/CPPS requires assessment of the microbiota composition on aspects of the pathophysiology of prostate inflammation to develop multidimensional treatment modalities that take into consideration major factors of the syndrome.

BPH, male lower urinary tract symptoms (LUTS), and the microbiome
Oral medications, such as alpha-1 blockers and/or 5-alphareductase inhibitors (finasteride or dutasteride), are gold standard therapies and decrease the risk of clinical progression 28 . Unfortunately, approximately 13% of males present with clinical progression within 4 years after the initiation of treatment, with 5% requiring surgical intervention, such as transurethral resection of the prostate 29 . The fact that some patients show a lack of response or durability to oral medication drives researchers to fully understand the underlying pathophysiology of BPH and benign prostatic enlargement. Although a previous study revealed that an individual's urinary microbiome/microbiota could provide information about bacteria related to urinary urgency incontinence status and treatment response to anticholinergic drugs, there is a lack of evidence regarding predictive tools for treatment response to drugs in patients with BPH 30 . As intraprostatic inflammation may play a vital role in the pathogenesis of BPH and clinical progression, 31,32 this physiological phenomenon is a potential target for diagnosis and treatment. Profiling the urinary microbiome in men with positive vs. negative biopsies for PCa led to the hypothesis of a link between the prostate and proinflammatory bacterial species 33 .
An investigation into the association between the microbiota and CP/CPPS paved the way to studying similar associations between the microbiota in BPH and male LUTS (Table 2) 34À38 . A previous study published in 2013 demonstrated that aging affects the microbiota in male midstream voided urine and showed an interesting trend. The total number of bacteria in the distal urethra significantly decreased with age, while the number of genera increased (increased 'diversity') 39 . This finding suggested that the change in the urethra and bladder microbiota with age might be associated with increasing LUTS in older males, which is typically due to BPH. Midstream urine, catheterized urine, seminal fluid, expressed prostatic secretion, stool, and resected prostatic tissue have been used for profiling the microbiome. However, Bajic et al. pointed out that voided urine cannot adequately characterize the male bladder microbiome due to the existence of distinct microbiota in the anterior urethra of the male lower urinary tract 35 . In the analysis of urine microbiome from males with and without prostatic enlargement, the International Prostate Symptom Score severity was associated with detectable bacteria in catheterized Table 1 Representative studies for the association between chronic prostatitis/chronic pelvic pain syndrome and microbiome. In patients with CP/CPPS, a predominance of anaerobes or a combination of aerobes and anaerobes in a titer of ! 10 3 colony-forming units per mL in post-massage urine is associated with worse clinical status.
The main components that determine the severity of symptoms are an increase in microbial load and qualitative differences in the composition of the microbiota.

Table 2
Representative studies for the association between benign prostatic hyperplasia/lower urinary tract symptoms and microbiome.  34 . Given this evidence, materials used for microbiota profiling surely affect the study findings. Infection with certain bacteria may induce a chronic inflammatory state in the prostate via the enhanced production of proinflammatory cytokines. Further investigation into the urogenital microbiota may reveal potential associations between BPH, male LUTS, and the causative pathogens.

Intraprostatic microbiome and the risk of PCa
Carcinogenesis is associated with multiple factors, including age, race, diet, heredity, the environment, and inflammation. Approximately one-fifth of malignancies can be attributed to infection by bacteria, viruses, and parasites 40 . Ravich et al. first advocated the infection hypothesis for PCa and the potential of prophylaxis in 1951 41 . Then, a meta-analysis of thousands of patients with PCa and controls from 29 case-control studies showed a significant association between carcinogenesis risk and infection history of any sexually transmitted infection, including Mycoplasma genitalium (MG) and human papillomavirus (HPV) 9,40 . There has been great interest in understanding the key role of inflammation in the initiation and progression of PCa. Proliferative inflammatory atrophy is a putative pathological lesion wherein inflammation stress drives prostate carcinogenesis through excessive reactive oxygen species, epigenetic alterations, and subsequent mutagenesis 42 . Intense exposure of the numerous microorganisms in the prostate through the urethra can generate chronic inflammation, in which stress drives prostate carcinogenesis through excessive reactive oxygen species, epigenetic alterations, and subsequent mutagenesis. Cavarretta et al. investigated the microbiome of the prostate microenvironment using 16s rRNA gene amplification followed by massive sequencing to identify the specific microbiota or bacteria associated with prostate carcinogenesis 43 . The evidence from this study was limited because there was a lack of evaluation of intraprostatic inflammation such as proliferative inflammatory atrophy and the detection of non-bacterial microorganisms such as mycoplasmas and viruses.
We previously examined robot-assisted prostatectomy specimens from 45 patients with PCa and transurethral resection specimens from 33 patients with BPH 9 . Using these specimens and DNA-based detection analysis, we screened tissue infection of a panel of seven sexually transmitted infection-related organisms and compared the pathological severity of prostatic inflammation to seek the potential association of infection, inflammatory environment, and carcinogenesis. Among the tested microorganisms, Neisseria gonorrhoeae, Chlamydia trachomatis, Ureaplasma urealyticum, and Mycoplasma hyorhinis were not detected in any specimen. The positive rate of MG was significantly different between the PCa cohort (18/45, 40%) and the BPH cohort (6/33, 18%). HPV18 DNA was detected in five (11%) patients in the PCa cohort and two (6%) patients in the BPH cohort, while only one patient in the PCa cohort was positive for HPV16. Of the six patients with PCa positive for HPV16/18, four (67%) were positive for MG, indicating a high rate of mixed infection. The positive status of MG in surgical specimens was associated with lower age in the PCa cohort (N ¼ 45, Fig. 1A) and the PCa/BPH cohort (N ¼ 78, Fig. 1B). MG is one of the most significant sexually transmitted pathogens, which can cause several human inflammatory diseases such as urethritis and CP in men and cervicitis, endometritis, salpingitis, tubal factor infertility, and pelvic inflammatory disease in women 44 . The higher rate of MG infection in the younger population could be explained by higher sexual activity. In the analysis of MG infection, patient background, and intraprostatic inflammation in the PCa cohort, the rate of extensive disease (pT2ce3b) was higher in MG-positive patients than in MG-negative patients (77% vs. 44%, P ¼ 0.027). However, no significant correlation was observed between the MG infection status and the grade of pathological intraprostatic inflammation.
The possible clinical application of testing prostatic microorganisms would be an indication or recommendation for re-biopsy or close follow-up after negative results of the initial prostate biopsy. We examined the feasibility of detecting infectious agents from the corresponding prostate needle biopsy cores. The detection sensitivity of biopsy specimens was 61% for MG and 60% for HPV18 9 . Although the detection accuracy would be high enough, non-invasiveness of material sampling is important for routine   34 . The investigators successfully detected the presence of diverse bacteria from all three materials and found a significant difference in the microbial population between PCa and BPH. Given that data obtained from non-invasive sampling accurately reflects the microbial environment in the prostate, another feasible and possible intervention would be the targeting of infectious agents by antibiotics to reduce the risk of PCa.

Gut microbiome and PCa
Another topic to be discussed is the close association between gut microbiota and the risk of PCa. Emerging data have proven that the gut microbiota of men with PCa is different from that of men without PCa 45 . Due to racial differences and heterogeneity in the microbiome among patients, its role in prostate carcinogenesis remains unclear, especially in Asian men. Several comparative analyses using stool samples or rectal swabs proved the compositional differences in bacteria between patients with PCa and healthy controls (Table 3) 47 Sfanos et al. evaluated rectal swabs from 21 patients with PCa and demonstrated that Ruminococcaceae and Akkermansia muciniphila were more abundant in men taking oral androgen receptor axistargeted therapies such as abiraterone acetate and enzalutamide 48 . A Japanese report found that Rikenellaceae, Alistipes, and Lachnospira (all short-chain fatty acid-producing bacteria) were significantly higher in those at a high risk (grade group !2) of PCa 50 . The same group previously reported that a high-fat diet and obesity promoted local prostate inflammation and cancer proliferation and that short-chain fatty acids and major metabolites of intestinal bacteria promoted PCa growth via the insulin growth factor-1 signaling pathway in a preclinical model. 52 Interestingly, compositional differences of the gut microbiome in matched hormone-sensitive and castration-resistant PCa were reported by Liu et al 51 . The abundance of several bacterial florae, including genus Phascolarctobacterium and Ruminococcus, increased in castration-resistant PCa. Functional analyses revealed that bacterial gene pathways involved in terpenoid/polyketide metabolism and ether lipid metabolism are activated significantly in the transition from hormone-sensitive PCa to castration-resistant PCa.

Gut microbiome and the risk of radiation-induced side effects
Radiation therapy is one of the gold standard treatments for the broad spectrum of PCa. The biggest concern is radiation-induced side effects such as gastrointestinal toxicity and GU toxicity, which can limit the radiation dose 53À55 . The Microbiota-and Radiotherapy-Induced Gastrointestinal Side-Effects study hypothesized that the microbiome differs between patients with and without radiation enteropathy 56 . The results clearly showed that microbiota diversity dynamically decreased over time in patients with increasing radiation enteropathy. Similarly, a consistent relationship between low diversity and late radiation enteropathy has been noted. Metataxonomic (16S rRNA gene) and imputed metataxonomic (Piphillin) analyses demonstrated that a higher abundance of Clostridium IV, Roseburia, and Phascolarctobacterium were associated with radiation enteropathy. Homeostatic intestinal mucosal cytokines linked with microbiota regulation and maintenance of the intestinal wall were decreased in radiation enteropathy. Moreover, interleukin-15 levels were inversely correlated with the counts of Roseburia and Propionibacterium. Microbiome testing before and during pelvic radiation would exhibit the potential for risk assessment, prevention, or treatment of radiation-induced enteropathy.

Isoflavone intake and equol-producing intestinal flora on the risk of PCa
The age-adjusted incidence of PCa is lower in Asian men than that in Western populations. According to the 2020 worldwide cancer statistics, the incidence rates of PCa vary from 6.3 to 83.4 per 100,000 men across regions, with the highest rates found in Northern and Western Europe, the Caribbean, Australia/New Zealand, Northern America, and Southern Africa and the lowest rates in Asian countries 57 . Moreover, the statistic of region-specific mortality age-standardized rates for PCa in 2020 shows lower mortality rates in Asian countries 57 . The A substantial number of studies have assessed dietary habits in association with PCa risk. One of the most attractive foods has been products from soybean, a traditional Japanese meal, or isoflavones, which may be associated with a decreased risk of incidence and progression of PCa 58 . A large epidemiological survey of multiethnic men in Hawaii and Los Angeles demonstrated that men with the highest intake of legumes lowered the risk of PCa by 11% and the risk of nonlocalized or highgrade PCa by 26% compared to men with the lowest intake of legumes 59 . Separate analyses showed that similar risk reductions were noted for soy products and legumes; however, the isoflavones in soy products were probably not responsible for this effect. The main mechanism of the inhibitory effect against PCa is that isoflavones function in the human body as strong phytoestrogens.
The three most abundant isoflavones are daidzein, genistein, and glycitein. The former two together comprise 0.1e0.3 mg/g soybean 60 . Moreover, daidzein is converted by a specific human intestinal bacterium, Slackia sp. NATTS, to equol, which is biologically more active than any other isoflavones 60,61 . Equol exerts a strong antiandrogen function by binding to dihydrotestosterone (DHT) 62 . A person who can produce equol in response to consuming soy isoflavones is classified as an equol producer. It is estimated that health benefits from soy-based diets may be greater in equol producers than in equol non-producers 58,63 . We previously evaluated the effect of supplementing healthy men with soy isoflavones on the serum levels of sex hormones, which were involved in PCa development 64 . A total of 28 healthy Japanese volunteers (age, 30e59 years), comprising 18 equol producers and 10 equol nonproducers, were given soy isoflavone (60 mg daily) supplements for 3 months, and their sex hormone levels were at the baseline and after 3 months were compared. No changes in the serum levels of estradiol and total testosterone were observed, while a significant increase in the serum levels of sex hormone-binding globulin and a significant decrease in the serum levels of free testosterone and DHT after 3 months of supplementation were detected. Interestingly, in two out of 10 equol non-producers, equol in the serum was detectable after 3 months of supplementation. Our findings demonstrated that short-term supplementation with soy isoflavones stimulated the production of serum equol and decreased the serum DHT level in healthy Japanese volunteers, leading to a possible PCa risk reduction. Another by-product of our study was the possibility of converting equol non-producers to equol producers by prolonged and consistent intake of soy isoflavones.
Epidemiological evidence of serum isoflavone levels and equol production by a specific human intestinal bacterium has long been evaluated with regard to the risk of PCa. A Japanese casecontrol study collecting data on serum isoflavone levels, NATTS identification in feces, and semi-quantitative frequency of food intake from patients with PCa and controls was completed and reported in 2016 65 . Higher serum levels of genistein, daidzein, and glycitein were significantly associated with a decreased risk of PCa, while serum levels of equol and NATTS in feces were not associated. A meta-analysis of two studies from Japan and five studies from Europe summarized the data of prediagnostic circulating levels of isoflavones (genistein, daidzein, and equol) and lignans (enterolactone and enterodiol) and the risk of PCa from 2828 patients with PCa and 5593 controls. 66 Overall, there is no strong evidence regarding the association between prediagnostic circulating levels of isoflavones or lignans and the risk of PCa. However, the authors emphasized the need for further research in populations where isoflavone intake is high, as in Japan, because of the positive link between higher circulating equol and a lower risk of PCa. Figure 2. Association between epithelial-mesenchymal transition, high Gleason' score, and pretreatment serum daidzein level in patients with localized prostate cancer. A) Immunohistochemical staining analysis using primary antibodies against E-cadherin, vimentin, Slug, and Snail. The association between each marker and the Gleason grade are evaluated. B) Pretreatment serum daidzein level among tumors with different Gleason scores is compared using the ManneWhitney U test. C) Correlation between the pretreatment serum daidzein level and the rate of vimentin-positive cancer cells are tested using Spearman's rank correlation coefficient.

Inhibitory effect of daidzein and equol on tumor growth and progression of PCa
According to an epidemiological study, isoflavone consumption was associated with a lower incidence of metastatic PCa, 67 supporting experimental data showing the potential inhibitory effect of isoflavones against the metastatic capability of tumor cells 68 . The molecular mechanism underlying this inhibitory effect is still unclear. Therefore, we focused on the possible inhibitory effect of isoflavones on epithelial-mesenchymal transition (EMT) in PCa. EMT has been implicated in the progression and metastasis of various malignant tumors, including PCa 69,70 .
Our analysis included 76 patients diagnosed with organconfined PCa between April 2008 and June 2011 (ethical approval ID: NMU-899). The materials and methods used are described in Supplementary File 1. In vitro experiments demonstrated that both daidzein and equol suppressed EMT and the migratory ability of PC-3 cells (Supplementary File 2). Immunostaining analysis using EMT markers, E-cadherin, vimentin, Slug, and Snail, revealed that higher Gleason scores were characterized by increased EMT potential ( Fig. 2A). Pretreatment serum daidzein levels were associated with higher Gleason scores and higher rates of vimentin-positive cancer cells (Fig. 2B). Next, we performed an animal experiment using the orthotopic carcinogenesis mouse model, the transgenic adenocarcinoma of the mouse prostate model in the FVB background (transgenic adenocarcinoma of the mouse prostate/FVB hybrid mouse model). 71 The mean serum level of equol was 1638 ng/mL in the equol-containing food group and 73 ng/mL in the standard food group (P < 0.001, Fig. 3A). The testosterone and DHT levels were not significantly different between the two groups (Supplementary File 3). Cancer-specific survival was significantly longer in the equolcontaining food group than in the standard food group (31 weeks vs. 23 weeks, Fig. 3B). At 20 weeks of age, most mice developed high-grade prostatic intraepithelial neoplasia precancerous lesions, whereas cancer lesions were detected in some mice; the cancer occupancy rate was only 8% in the equol-containing food group and 42% in the standard food group (Fig. 3C). The tumor expression of Ecadherin was higher, and that of Slug was lower in the equol group . Daily intake of equol prolongs survival and inhibits development of prostate cancer. A) Serum level of equol in transgenic adenocarcinoma of the mouse prostate (TRAMP)/FVB mice comparing the equol-containing food group and the standard food group. A P-value is obtained using the ManneWhitney U test. B) Cancer-specific survival curves of TRAMP/FVB mice comparing the equol-containing food group and the standard food group. A P-value is obtained using the log-rank test. C) Hematoxylin and eosin staining images of 20 week-old mice prostate sections comparing the equol-containing food group and the standard food group. Dashed black lines indicate borders between the cancer lesion and the normal area. Cancer occupation rate in the prostate section was calculated based on hematoxylin and eosin-stained sections and was compared between the equol-containing food group and the standard food group. PCa, prostate cancer; HG-PIN, high-grade prostatic intraepithelial neoplasia. than in the standard food group (data not shown). Moreover, the incidence of lung metastases was significantly lower compared with 27% in the equol group and 87% in the control group (P < 0.01). Our preclinical study suggests that treatment with daidzein and equol may be a potential chemopreventive approach to decrease the risk of PCa development and progression via the suppression of EMT. Our findings also provide evidence that isoflavones are potential agents for treating metastatic PCa.

Conclusion and future perspectives
A comprehensive understanding of the correlation between the urogenital microbiome and chronic prostatic inflammation could facilitate the development of novel strategies for PCa prevention. Further investigation would help to better understand the roles of gut microbiota dysbiosis in progression of prostate diseases and develop new therapeutic modalities.

Conflicts of interest
The authors disclose no potential conflicts of interest.