Dysbiotic microbes and how to find them: a review of microbiome profiling in prostate cancer

The human body comprises of trillions of microorganisms with the estimated bacterial population in the order of 10¹³ cells, resulting in an approximate 1:1 ratio between bacterial and human cells in an average human [1]. It is thus fathomable that the human microbial ecosystem (microbiota) can influence aspects of human health and disease through direct or indirect effects [2], for example by manipulating nutrient uptake and drug metabolism or by inducing systemic inflammatory responses [2‐5]. While the resident microbiota is typically associated with beneficial effects to its host, changes to the microbial composition, known as microbial dysbiosis, could be associated with diseases such as inflammatory bowel disease, diabetes mellitus, and obesity (reviewed by [6]). Since inflammation is a signature of many pre-neoplastic and malignant lesions, chronic inflammation has also been implicated in carcinogenesis likely mediated by bacterial toxins as in the case of Helicobacter pylori and gastric carcinoma [7]. In fact, a growing body of evidence now suggests a crucial role for microbial dysbiosis in cancer development and progression, including significant associations with both bacterial and viral species [8‐12].

Prostate cancer (PCa) is the second most frequent male malignancy worldwide with over 358,000 estimated deaths in 2018 [13]. PCa has a highly heterogeneous clinical course. Most organ confined (localized) PCa have an indolent course with a 5 year overall survival of ~100% even without any treatment. In such cases, active surveillance is recommended. In aggressive forms of PCa where the tumor is still organ confined, complete removal of the prostate through radical prostatectomy and/or radiation therapy is necessary to prevent further spread of the disease. However, the decision whether or not to treat localized PCa is a major clinical challenge as currently there are no accurate tests to distinguish between indolent and aggressive PCa at the localized (early) stages. This leads to an over-treatment of indolent cases and an under-treatment of aggressive PCa, resulting in patient morbidity and mortality. Thus, there is an urgent need for better risk stratification tools, that incorporate other risk factors such as bacterial infections or inflammatory markers [14].

The advent of next generation sequencing (NGS) technologies has opened a new area of PCa research enabling unparalleled access to the genomic and transcriptomic underpinnings in PCa. Utilizing the potential for these technologies, several molecular markers have been proposed to stratify PCa [15], although complete success in this regard is yet to be achieved. A major reason could be the interaction between the neoplastic cells and the tumor microenvironment, which remains dynamic. Prostate microbiota could also be hypothesized to be a major driver enabling differential clinical course in localized PCa. In fact, among the well known risk factors for PCa such as age and ethnicity, factors such as microbial composition has also found a potentially essential place in recent years due to the increased scientific scrutiny of its role in mediating inflammation and thereby driving prostate carcinogenesis and progression [16]. While earlier studies relied on culturing bacteria from the prostate, NGS based methods have enabled genotyping the microbial ecosystem within a prostate for hundreds to thousands of patients in parallel, providing a better overview of the landscape of the PCa associated microbiome.

The aim of this review is twofold, 1) provide the researcher with the necessary technical know-how to perform microbiome analysis, and 2) inform the reader of the advances that have been made in the field of prostate cancer microbiome research. We start out by describing in general how to analyse microbiome data and note several automated pipelines that are available to the researcher. For the sake of simplicity, we restricted our review to the three most frequently used NGS based analysis methodologies (amplicon sequencing, shotgun DNA and total RNA sequencing) that have been widely adopted by the microbiome research community. We next provide a detailed overview of the current knowledge of the role of the human microbiome in PCa development, progression and treatment response that was made possible by some of the aforementioned methodologies. Finally, we describe the proposed mechanisms of host-microbe interactions that could lead to PCa development, progression, or treatment resistance. A clear distinction between microbial association with PCa development (carcinogenesis) and its association with PCa progression (e.g. metastatic dissemination) is difficult to make due to the lack of healthy non-cancer control samples in most studies, and consequently this remains an outstanding question in the field.

Most of the current research enumerating the microbial species present in the prostate and various other body sites utilize NGS based methodologies as opposed to the culture based techniques employed during the last century, which could detect only species that could be cultured. Three main methodologies are most commonly used now (Fig. 1A). These include amplicon sequencing, shotgun DNA sequencing, and RNA sequencing based methodologies. An in depth explanation of the analysis methodologies and best practises for microbiome research is beyond the scope of this review, but we direct readers to other published reviews [17, 18]. A graphical summary of the general steps involved in microbiome sequence analysis is shown in Fig. 1B.

Since 2015 there has been a steady rise in the number of publications looking at the association between the human microbiome and prostate cancer development, progression, and treatment outcome. While most research has focussed on the so called direct effect on PCa of the microbiome in the prostate tissue, others have also investigated associations between PCa and the core microbiota from different body sites, the so called indirect effects [2], as depicted in Fig. 2. These have mainly focussed on the effect of the gastrointestinal microbiota and the urinary microbiome on neoplastic transformation of prostatic epithelia [46, 47], but also include studies evaluating associations between PCa and prostatic and seminal fluid microbiomes [48, 49].

Table 1 provides a list of NGS based studies since 2015 investigating microbial dysbiosis associated with PCa. These studies were selected based on a PubMed search for ‘prostate cancer microbiome’ resulting in 159 records since 2015 (as of July 2021). Of these, review articles (n=58) were excluded, and original studies (n=20) profiling the prostate tissue, gut, urinary, seminal fluid and prostatic fluid microbiomes in relation to prostate cancer were selected. A recent study from our group is also discussed in this context.

A number of studies have investigated the possible functional role of specific microbes in relation to PCa due to their inflammatory potential and frequent detection in prostatic tissue (Table 2). Cutibacterium acnes (formerly, Propionibacterium acnes) is a skin-associated commensal that has been detected in the prostate of men with PCa in several studies [53, 65] with few studies reporting a higher prevalence in patients with PCa compared to men without [66, 67]. C. acnes has also been associated with chronic inflammation in the prostate of men with PCa [68] and shown to induce acute and chronic inflammation in mice inoculated with human prostatectomy-derived C. acnes isolates [69]. Evidence from cell-based experiments suggests that C. acnes can induce cell proliferation [66] and the secretion of cytokines and chemokines such as IL-6 and IL-8 [70, 71], which are crucial for maintaining active inflammation. However, a later study failed to observe any statistical difference in IL-6 secretion between men with vs. without C. acnes infection [72]. In another study, prostatic epithelial cell lines infected with C. acnes responded via activation of transcription factors such as NF-κB and STAT3 [70], which are associated with cellular proliferation and tumor growth in various cancers, such as PCa and colon cancer [73, 74]. Others have also provided evidence for C. acnes-induced production of reactive oxygen species by keratinocytes in the skin, thereby inducing oxidative stress response and inflammation [75], an event that could perhaps be replicated in the prostate leading to pre-cancerous transformation of the prostatic epithelia [70]. Taken together, all of this evidence points to a possible role for C.acnes-induced inflammation in PCa development or progression.

Many studies have also investigated the role of E. coli in prostate carcinogenesis. Uro-pathogenic strains of E. coli are known to induce prostate tissue damage in rat models of prostatitis [81], mediated by cytotoxic necrotizing factor 1 (CNF1), a virulence factor that has also been shown to promote PCa progression [85]. In a mouse model [52], all mice experimentally infected with E. coli for 12 weeks developed chronic inflammation in the prostate, and with prolonged infection showed cytological changes typical for prostatic intraepithelial neoplasia and high-grade dysplasia. Increased epithelial cell proliferation, and oxidative DNA damage was observed in the prostate glands exhibiting dysplasia, together with decreased androgen receptor and PTEN gene expression, as compared to the control glands [52]. This could indicate a mechanistic link between E. coli-induced inflammation and the onset of PCa or pre-neoplastic lesions.

Others such as Faecalibacterium prausnitzii have also been associated with PCa, with higher fecal abundances of F. prausnitzii observed in benign as compared to malignant patient samples [59]. F. prausnitzii is generally considered to have anti-inflammatory properties with its ability to produce butyrate and induce secretion of anti-inflammatory cytokines such as IL-10, TGF-β2 and IL-1Ra (reviewed by [83]). Furthermore, it has been reported that F. prausnitzii also down-regulated the expression of pro-inflammatory cytokines such as TNF-α, TNF-β and IL-6 in lung cancer cell line [82] and could inhibit the phosphorylation of JAK2/STAT3 in breast cancer cells, potentially leading to growth inhibition of cancer cells [86].

Through this review we aimed to explore the extent of microbial dysbiosis that is associated with PCa by providing an overview of the current knowledge in the field. Moving away from the notion of cancer as solely being a disease of the genome, we believe a more holistic approach towards cancer treatment, informed by genetic, epigenetic, and host-microbiome interactions could benefit treatment decisions in the future. PCa, like other cancers, is a dynamic and heterogeneous disease that has several layers of molecular and cellular complexity associated with it. Microbiome analyses have revealed a bacteria rich environment in the prostate that might be altered during disease onset, progression or treatment. Several species of the microbial community have been associated with PCa aggressiveness and response to therapy, a finding that has also been observed in many other cancer types. A better understanding of the role microbes play in these processes will help us develop novel treatment strategies as well as better risk stratification tools. For instance, it is possible that removal of certain gut microbial species prior to androgen deprivation/hormonal therapy could delay disease progression to CRPC. Several clinical trials are underway looking at e.g. the efficacy of fecal microbiome transplant in combination with established treatment strategies as a means to control tumor progression in different cancers, including in PCa. If successful, these could give us an upper hand in the battle against cancer.