Microbiome in cancer progression and therapy

Summary

A myriad of microbes living together with the host constitute microbiota, which possesses very diverse functions in regulation of host physiology. Recently, it has been unequivocally demonstrated that microbiota regulates cancer initiation, progression and responses to therapy. Here we review known pro-tumorigenic and anti-tumorigenic function of microbiota and mechanisms how microbes can regulate cancer cells and immune and stromal cells within the tumor microenvironment.

Introduction

Unique complex microbial ecosystems are important and irreplaceable components of human body biology and homeostasis. Microbiota of the host appears early after birth through maternal transmission is further shaped through the lifetime by various circumstances and events including host genetics, infections, antibiotic usage, dietary and environmental factors, and presence of various diseases or use of therapies. The presence of large number of live and metabolically active microbes in close proximity to cells represents opportunity for the reciprocal interactions whereby the host modify the microbiota and microbiota derived factors. Indeed, quantitative and qualitative properties of the microbiota can regulate multitude of host processes, including tonic regulation of immune and inflammatory cells, digestion, gastrointestinal motility, cell differentiation and turnover or cognitive and behavioral functions. Likewise, changes in microbiota have been associated or found to be causative of a variety of diseases, including inflammatory conditions [1,2], metabolic [3] and cardiovascular diseases [4], neurological ailments [5] and cancer [6,7]. Since the gastrointestinal tract harbors at least 10¹⁴ bacteria [8], many aspects of the “microbiota-host-disease” triad have been studied in the context to intestinal microbiome. However, it has become apparent that other organs, previously thought to be sterile can harbor microbial organisms both in healthy and diseased states or are subject to distant regulation by microbial metabolites produced elsewhere[9,10].

In this review we summarize some of the strategies and mechanisms by which microbiota regulates cancer progression, metastasis and sensitivity to therapies and metastasis development and progression. Importantly, this relation can be both deleterious and beneficial, as the microbiome can exert anti-tumorigenic and pro-tumorigenic effects in the host through distinct molecular and cellular mechanisms.

Protective role of the microbiome in cancer

Live bacteria and their products can accumulate in many tumor types, primary or metastatic [10]. Distant regulation of tumorigenesis and tumor microenvironment is also thought to be a process that can be mediated by microbiota [9,11–15]. While many bacteria and their products were shown to be tumor-promoting [6,16,17], as discussed later, some bacteria clearly exert protective functions during the process of tumorigenesis or facilitate various forms of cancer therapies.

Diverse microbiome community which populates tumor tissue or regulates intra-tumoral processes distantly can exert protective, anti-tumor function by several mechanisms.

First, some bacterial species could initiate and sustain the formation of favorable conditions for growth of beneficial bacteria and together form the niche which suppresses the overgrowth of pathogenic bacteria [18] (Figure 1). Examples of such beneficial bacteria include Lactobacillus or various Clostridium species protecting against pathogenic E.coli or Salmonella colonization in the context of intestinal inflammation [19,20].

Figure 1. Roles of «beneficial» and «pathogenic» bacteria in primary tumor development and metastasis.

Microbiota contains multiple species with different metabolic activities and other properties, and these species co-exist with the host within the particular niche or tissue. Each relevant bacterial species plays the different role in the primary tumor growth, metastasis and anti-cancer therapy. «Beneficial» bacteria can suppress tumor growth and improve anti-cancer therapies through several direct mechanisms but also can inhibit overgrowth and outcompete «pathogenic» bacteria in the context of cancer. Pathogenic bacteria promote tumor development, metastasis and resistance to therapies. “Green” reflects anti-tumorigenic action, while “Red” is tumor-promoting function. Arrows illustrate active promotion, while blocks illustrate inhibition. Figure was produced using BioRender platform.

Second, some of the metabolites produced by bacteria may be essential tools in their interactions with epithelial and cancer cells for tumor growth suppression. Short chain fatty acids (SCFA), which are produced by different commensals from fiber fermentation, inhibit myeloid cell driven pro-tumorigenic inflammation and regulate proliferation of epithelial and stem cell compartment [21,22]. Several vitamins, including biotin, cobalamin, folate, niacin, pantothenate, pyridoxine and others have also been linked to anti-tumor activities [23]. Phytochemicals such as polyphenols, which are widely present in fruits and vegetables, are metabolized by certain gut bacterial species into the active forms. Polyphenols metabolites may prevent cancer progression through the impact on cell cycle arrest and apoptosis and also through inhibition of inflammatory cytokines production [24].

Third, distinct bacterial species or “aggregate microbiota” are essential for maturation and tonic stimulation of the immune system, which then can exert immunosurveillance functions at different stages of tumor development [25,26]. Classical Coley experiments and anti-tumor effect of Coley toxins where bacterial infections led to regression of established tumors served as a foundation for studies in this direction [27]. On the other hand, germ free mice lack fully matured immune system and also depletion of microbiota with broad spectrum antibiotics can reduce the efficacy of immunesurveillance in mouse models [28]. This function is further extended to the ability of microbes to stimulate activation of various immune cells subsets; for example, members of Lactobacilla phylum activate antitumor immune responses by dendritic cell (DC) maturation [29] and subsequently for acquisition of cytotoxic properties by T cells, NK cells, NK T cells and anti-tumorigenic myeloid cells [30].

Fourth mechanism stems from the aforementioned ability of bacteria to drive activation of immune system and is related to the ability of some bacteria to improve anti-cancer therapies [6]. Germ-free or antibiotic-treated tumor bearing mice do not properly respond to standard chemotherapeutic treatment with oxaliplatin [31] and gut bacteria are needed for the effect of cyclophosphamide, an anti-cancer immunomodulatory agent, which acts in particular through an increase in the intestinal permeability and translocation of immunostimulatory bacteria into secondary lymphoid organs, such as lymph nodes, for proper stimulation of anticancer immune responses [32]. Microbiota is also essential for efficiency of and sensitivity to various immunotherapies, such as combination of CpG-oligodeoxynucleotides (ODN), a ligand of Toll-like receptor 9 (TLR9), and inhibitory interleukin-10 (IL-10) receptor antibodies (anti-IL-10R) therapy [31]. It was shown that some commensal bacteria such as Alistipes and Ruminococcus are directly responsible for production of anti-tumorigenic TNF and regulation of reactive oxygen species essential for tumor restriction in this model [31]. Another facet of microbiota action is its ability to induce and heighten the responses to immunotherapies based on the usage of immune checkpoint blockade approaches. Responses to anti-CTLA4 or anti-PD-1/PD-L1 or combinational therapies positively correlated with overall diversity of microbiota as well as with the presence of particular species, such as Bacteroides and others [25,33–35] . These findings have led to trials of probiotics or fecal transplants from responders to non-responders in order to understand the mechanisms which underline these effects of microbiota. Mechanistically, the microbiota can regulate the production of cytokines with anti-tumorogenic and immunostimulatory properties, such as IL-12, TNF and others [31]. As the microbial diversity associated with improved responses, it is tempting to hypothesize that microbial genes provide an excellent basis for molecular mimicry of cancer neoantigens, i.e. when the same peptides present as cancer neoantigen in the tumor are present within one of the bacterial proteins. In this scenario, the presence of bacteria provides both a signal (so called “signal 1”) for stimulation of innate immunity, and an additional signal (“signal 2”) as the MHC- peptide complex, which is identical to that on cancer cells; and the presence of checkpoint blocker alleviates the exhaustion or repression of neoantigen/tumor specific T cells. This has been demonstrated in particular for Bacteroides fragilis which encodes for the peptide found in melanoma [33] or for a bacteriophage infecting Enterococcus species associated with enhanced response to anti-PD-1 immunotherapy [36].

Pro-tumorigenic effects of microbiota in tumorigenesis and tumor progression

A concept of “infection-driven” cancers is highly plausible, as HPV, HCV, HBV, Helicobacter pylori and other infectious organisms have been shown to aid to directly promote transformation and tumor growth and can be considered as bona fide “infectious carcinogens” [37]. However, recent evidence also suggests that pathobionts, commensal organisms that can sometimes become pathogenic, can also drive tumorigenesis. The underlying mechanisms for microbiota driving tumorigenesis are somehow overlapping with the case of anti-tumorigenic mechanisms discussed above, and include the induction of pro-tumorigenic inflammation and immunosuppression, aberrant tissue repair program, direct genotoxic effects, as well as metabolic effects on cancer cells and cells of the tumor microenvironment (Figure 2).

Figure 2. Mechanisms of microbiota-host interactions promoting tumorigenesis.

Non-modifiable (genetic) and environmental factors promote intestinal epithelial barrier disruption and dysbiosis. Barrier dysfunction causes bacterial penetration through the epithelium and their interaction with host cells by several mechanisms: (1) immune cell activation and triggering of pro-tumorigenic inflammation; (2) Epithelial cell DNA damage by bacterial toxins or bacteria-induced genotoxic products (e.g. ROS) ; (3) modulation of immune, stromal and cancer cell function through the effect of bacterial metabolites acting locally or distantly. Figure was produced using BioRender platform.

Loss of epithelial barrier function and increased bacterial colonization of transformed tissues

Epithelial and mucosal surfaces of the body as well as tumors are colonized by bacteria at different degrees ranging from modest (lung, pancreas) to high bacterial loads (colon). Indeed, colonic epithelium and lumen are rich in bacterial content, which requires the presence of a strong dedicated barrier including fortified cell-to-cell contacts, mucus coverage and continuous secretion of anti-bacterial peptides and immunoglobulins A (IgA) [38]. Because of this barrier, in healthy state bacterial transfer into systemic circulation and transport into other tissue is a rare and inefficient process. Even local interaction of bacteria with epithelial cells via bacteria-derived metabolites, such as SCFA, secondary bile acids, vitamins and polyamines may be resctricted by a functional barrier [39]. Barrier function however can be deteriorated by autoimmunity and inflammation, diets characterized by high fat or sugar content, the presence of invasive species or dysbiosis of different etiologies, including antibiotics overuse [40–44]. However, neoplastic transformation itself can lead to dysregulation of the epithelial barrier, directly linking early transformation with the loss of epithelial integrity and barrier against microbiota [45,46]. For example, in colorectal cancer, early transformation caused by dysregulation of tumor suppressor APC leads to the loss of proper cell junction organization and tight junctions as well as to cellular hyperproliferation and a loss of terminally differentiated subset of mucus-producing Goblet cells [45,46]. Altogether, this causes an increase in direct contact of commensal with neoplastic epithelial cells and other cells of tumor microenvironment and facilitate the access of microbially-derived products to the neoplastic tissue[45,46]. In addition, as inflamed or normal tissue progresses to adenoma and to carcinoma, microbiota is disrupted, probably due to changes in the barrier and epithelial compartment itself [47,48]. Among the bacteria enriched in CRC patients is Fusobacterium nucleatum [17,49,50] that encodes surface receptor FadA which binds to E-cadherin. Such binding and accumulation in transformed tissues is likely aided by the loss of barrier function and by direct physical FadA-E-cadherin interaction. Loss or alteration in barrier function and changes in metabolic state of epithelial cells can also lead to accumulation of other species, such as biofilm-forming Escherichia coli [51] and enterotoxigenic Bacteroides fragilis [52]. Such biofilms initially found in CRC would be probably detected in other epithelial cancers, such as lung and pancreatic cancers. Overgrowth of bacterial biofilms in the direct proximity of transformed tissue is likely to enforce stronger and persistent interactions between microbes and transformed tissue surface [53] and could facilitate reciprocal loss of epithelial barrier function at the tumor site. Furthermore, some bacteria like B. fragilis are especially efficient at establishment of biofilms and their presence initiates a cascade of local colonization with other bacteria such as E.coli [54]. The presence of biofilms is not obligatory for the bacteria to be associated with tumors or to drive tumor promoting functions though, as pro-tumorigenic bacteria and components of microbiota have been detected in lung [55–57], pancreatic [58,59] and other tumors without formal demonstration that they are organized in biofilms.

Once increased interaction of bacteria and epithelial-tumor microenvironment compartment is established due to the loss of barrier function, increased colonization or formation of biofilms, bacteria can drive tumor progression by several mechanisms as it is described next.

Genotoxic effect of bacteria and induction of tumor-driving mutations

Distinct components of microbiota can release or induce synthesis of genotoxic compounds to affect host DNA by inducing mutations. Arguably, these processes are critical for tumor initiation and progression. Toxins, genotoxic compounds and other bacteria-induced metabolites have been directly implicated tumorigenesis by mutations or otherwise altering the expression levels of key oncogenes and tumor suppressors [60]. Helicobacter pylori gene product CagA inhibit the function of p53 tumor suppressor protein in gastric epithelial cells [61], while some Staphylococcus strains secrete toxic hemolysins and enterotoxins, which may contribute to the accumulation of mutations in malignant T cells [62]. Salmonella typhi produces a protein toxin virulence factor A that directly increases the expression of oncogenic β-catenin expression [63]. Genotoxic strains of E. coli harbor PKS “pathogenicity island” which encodes biosynthetic machinery for production of colibactin- a toxin which alkylates DNA at adenine residues [51] and causes mutations in the intestinal epithelial cells [64,65]. By genomic sequencing, a similar mutational signature was found in a subgroup of over 5800 human cancer genomes, mainly in colorectal cancers and these tumors were specifically associated with E.coli expressing colibactin [65]. Monocolonization with PKS+ but not with colibactin/PKS negative E. coli NC101 increase cancer development in Il10^−/− mice treated with azoxymethane [64] and other models of colon cancer, and mutation and PKS-specific mutational signature can be directly recapitulated in vitro using intestinal organoids colonized with E. coli [65]. Enterotoxic strains of B. Fragilis produce BFT toxin, previously implicated into exacerbation of colitis and colon cancer tumorigenesis. BFT binds to a specific colon epithelial receptor, activates the Wnt and NF-κB pathways, promotes proliferation and accelerates tumorigenesis [66]. Secondary bile acids, which are converted from primary bile acids by various members of Firmicutes family, possess DNA–damaging capabilities and therefore can aid cancer induction [67,68]. Another mechanism by which microbiota can induce DNA damage is the induction of oxidative stress and nitric oxide production through the activation of innate immune cells. Such oxidative DNA damage is characterized by presence of 8-oxoguanine and double-stranded DNA breaks and can affect key tumor suppressors, such as p53 [69]. During chronic inflammation, such genetic lesions can be present in visibly normal areas of epithelium and subsequently give rise to cancer through the process of «field cancerization» [70]. Prolonged IBD also triggers ROS-dependent, mutation-induced colitis associated cancer in animal models without further addition of exogenous carcinogens, suggesting an instrumental role for inflammation in induction of cancer-driving mutations [71]. As discussed above, changes in barrier function may be again facilitate of closer and persistent contact between bacteria and target population of cells acquiring mutations.

2. Metabolic effect and regulation of epithelial and stem cell compartment

Collectively, microbiota harbors at least 100 times more genes than the host, with substantial numbers of these genes encoding for the production of various metabolites, some of them identical to the host metabolites and some very unique. Diets modulate cancer risk and some do so through the metabolic activity of microbiota [72–74] involved in the regulation of lipid or cholesterol metabolism [75,76] or fructose content [77,78]. Short chain fatty acids (SCFA), such as butyrate, acetate and propionate, are produced during the fermentation of fibers by various bacteria including Bacteroides and Firmicutes [79]. SCFA has anti-inflammatory effects [7,79] and can act on both epithelial and immune cells [78] within the tumor microenvironment, with its effects on CRC still being studied. Microbial-driven metabolism of bile acids is important in the development of cancer [80]. Deoxycholic acid (DCA) produced by intestinal microbiota promotes CRC development by modulating signaling through muscarinic 3 and Wnt receptors pathways [81,82]. Bacterial metabolism can further affect and promote tumorigenesis through additional mechanisms including metabolic activation of pro-carcinogenic compounds, regulation of hormone metabolism; and modification of inflammatory pathways. Moreover, during therapy, microbiota can participate in metabolic dismantling of active therapeutic compounds, including in metastatic cancers [83–85], resulting in toxic secondary byproducts, lowering the tolerable dose and effectively weakening the therapy [86,87]. Since such metabolites or their secondary metabolites are absorbable and go into systemic circulation, they can affect many other cancers beyond the GI tract. For instance, the intestinal microbiome secretes bioactive metabolites (reactivated estrogens, short chain fatty acids, amino acid metabolites, or secondary bile acids) that modulate breast cancer [88].

Induction of pro-tumorigenic inflammation and bacteria-driven tissue repair.

Chronic inflammation promotes cancer progression and metastasis through a variety of mechanisms and microbes and their metabolites are one of the most prominent inducers of inflammation. If barrier function of transformed epithelium is weakened or if tumor or metastatic seed is permanently colonized with bacteria, chronic exposure to bacterial metabolites leads to chronic inflammation. Inflammation can then drive tumor initiation, growth, progression and metastasis via mechanisms related to regulation of cancer stem cells compartments, immunosuppression, induction of mutations, modulation of tumor microenvironment and stromal cells [89,90]. One of the most prominent mechanisms is the ability of bacteria to activate innate and adaptive immune cells within the TME via recognition through Toll-like receptor, NOD- like receptor and other PRR to stimulate the production of pro-inflammatory cytokines, which both regulate immunity and microenvironment , but also serve as a tissue protective and repair responses in epithelial and cancer cells [45,46]. Indeed, the presence of some of the cytokines (IL-17A, IL-6, IL-11) correlate with poor prognosis in CRC and other types of cancer [91]. Expression of these cytokines can be induced by intratumoral or barrier-seeding microbes. Alternatively, the microbiota can also induce specific types of inflammation, for example through the activation of Th17 cell differentiation in response to bacterial infected apoptotic innate immune cells [92].

Microbiota and inflammation can be also linked to the activation of stem cell/cancer stem cell compartments, which is essential for cancer growth , progression, metastasis and resistance to therapy [93,94]. The plasticity of cells in response to microbe- driven inflammation allows the transition of cells from one stage of differentiation to another, which in turn promote the epithelial–mesenchymal transition and metastasis [95,96]. Survival of metastasis can be regulated by microbiota and microbial products locally (e.g. in the liver, which is generally rich in microbe-derived signals) [97] or systemically, for example in the case of metastasis to the lung. Lipopolysaccharide (LPS) was shown to induce survival and outgrowth of cancer cells metastasizing to the lung by activation of inflammation-driven NF-kB signaling [98,99]. LPS also promoted metastatic spread by increasing integrin-mediated adhesion of cancer cells to the endothelial cells in the vessels [100].

Recent studies of microbiome, in particular associated with tumor development and therapy, only started to uncover the modalities and the mechanisms of host-microbe interactions during the disease progression and therapy. It is likely that further insights will come from cataloguing new and unique microbial metabolic functions followed by mechanistic experiments in animal models of cancer. These mechanisms will be further validated in patients to establish common functions and physiological denominators for the effects of microbes on cancer, particularly using existing wealthy collections of normal and cancer tissues Next, additional diagnostic, preventive or therapeutic approaches in cancer can be developed based on the presence of specific intra-tumoral or surface microbes, links between microbiota and diet, pro- and pre-biotic approaches, ability to introduce or deplete specific microbial species or their mutant versions, or approaches to bypass live microbiota function(s) with distinct set of bacteria-derived metabolites.

Table 1.

Summary: roles of microbiota in tumorigenesis and anti-cancer therapy

Anti-tumorigenic	Pro-tumorigenic
«Beneficial» bacteria > «pathogenic» bacteria competition	Activation of Tumor elicit inflammation (TEI)
Tumor suppressive metabolites	Genotoxic effect on epithelial and stromal cells
Immune system maturation and tonic stimulation for effective cancer immunosurveillance	Metabolic effect via the metabolic activation of carcinogenic compounds
Immunostimulation and improvement of responses to chemo-, radio- and immunotherapies of cancer	Åctivation of cancer stem cell compartments
	Activation of cell plasticity in the tumor microenvironment
	Inhibition of cytotoxic immune responses (immunosuppression)

Highlights.

1. Microbiota is important in various aspects of host physiology including regulation of cancer development and progression

2. Microbiota can play negative or positive effects on tumorigenesis and efficacy of therapies

3. Microbiota regulates tumorigenesis via few distinct mechanisms including interactions with cancer cells and tumor microenvironment

4. Pro-tumorigenic microbiota stimulates tumor elicited inflammation, alters metabolism, hinder therapies and can promote genotoxicity

5. Anti-tumorigenic microbiota outcompetes «pathogenic» microbes, metabolically suppresses cancer, facilitates immunity and therapies.