Abstract
The human body is colonized by the microbial cells that are estimated to be as abundant as human cells, yet their genome is roughly 100 times the human genome, providing significantly more genetic diversity. The past decade has observed an explosion of interest in examining the existence of microbiota in the human body and understanding its role in various diseases including inflammatory bowel disease, neurologic diseases, cardiovascular disorders, and cancer. Many studies have demonstrated differential community composition between normal tissue and cancerous tissue, paving the way for investigations focused on deciphering the cause-and-effect relationships between specific microbes and initiation and progression of various cancers. Also, evolving are the strategies to alter tumor-associated dysbiosis and move it toward eubiosis with holistic approaches to change the entire neighborhood or to neutralize pathogenic strains. In this review, we discuss important pathogenic bacteria and the underlying mechanisms by which they affect cancer progression. We summarize key microbiota alterations observed in multiple tumor niches, their association with clinical stages, and their potential use in cancer diagnosis and management. Finally, we discuss microbiota-based therapeutic approaches.
Introduction
A balanced microbial ecosystem within human body is termed as “eubiosis” with a dominance of diverse beneficial bacteria that live in mutual harmony. On the contrary, “dysbiosis” is defined by low diversity and preponderance of pathogenic bacteria. Dysbiosis may influence carcinogenesis via microbial toxins, altered metabolites, hormonal dysregulation, sustained inflammation, immune modulation, DNA damage, and mutagenesis. Many of these effects are considered direct effects of microbiota on cancerous tissue, whereas others are indirect effects by which microbiota at one site, e.g., gut, can influence distant organ sites. Gut dysbiosis regulates not only the circulating levels of metabolites and nutrients influencing host physiology, it releases microbial toxins to influence cancer progression. Metabolites converted by microbes can enter host circulation via absorption or reabsorption and reach distant target tissues. The direct effects of microbiome on target tissues are mostly mediated by the resident microbiota such as the effect of gut, lung, skin, or vaginal microbiome on colorectal cancer, lung cancer, melanoma or cervical, and endometrial cancers, respectively. Spatiotemporal dynamics of microbial biofilm organization also plays an important role in cancer incidence, e.g., location of the colorectal cancer development (1). The cancer-microbiome field has slowly transitioned from descriptive studies of microbiota existence in various cancerous tissues to pointed questions regarding its direct/indirect involvement in cancer progression, underlying mechanisms of action, modifying clinical interventions, and developing novel therapeutic regimens based on this knowledge. Here, we discuss the current state of knowledge regarding the role and importance of microbiome in risk assessment, early detection, and response to therapy as well as insights pertaining to microbiome-based therapy ranging from altering the microbiota and targeted removal of harmful bacteria to using bacteria as drug-delivery vehicles.
Microbiome as a Telltale Partner in Cancer
The complexity of metagenome and diversity of microbiota in health and disease triggered an interest to understand the precise role of dysbiosis in cancer. Although the initial interest was mostly focused on gut microbiota, the field has expanded to microbiota niches in various organs with the identification of unique and distinct microbiota in the most “sterile organs” such as breast, lungs, pancreas, and prostate. Microbiome influences various aspects of cancer progression as well as response to therapy, including chemotherapy and immunotherapy, and better mechanistic understandings are now emerging that can help in cancer management.
Cancer-associated bacteria provide insights for risk assessment
Over the years, the quest for understanding the risk factors for cancer has put forth various nonmodifiable risk factors including age, family history, and genetic mutations as well as multiple modifiable risk factors such as obesity, physical inactivity, alcohol, and smoking, to name a few. Recent developments in microbiome-cancer arena have added dysbiosis as an important risk factor for various cancers. Although most biological functions of microbiota are considered “‘community effects,” a number of potent microbes have been discovered that directly interact with eukaryotic cells via small molecules/toxins inducing signaling cascades, leading to mutagenesis and carcinogenesis (Fig. 1). A foremost candidate bacterium implicated with cancer is spiral gram-negative rod, Helicobacter pylori, which colonizes the human stomach and is associated with approximately 60% of all stomach cancer (2) and its positivity rate significantly increases with the severity of disease (3). Presence of H. pylori is also associated with laryngeal carcinoma (4), hepatobiliary cancer (5), colon cancer (6), benign prostatic hyperplasia, and prostate cancer (7). Clinical strategies are available to eradicate H. pylori infection (8, 9), but genetic damage may persist even after eradication of H. pylori infection, which indicates sustained higher risk for individuals with prior exposure to a pathogenic bacterium (10). Interestingly, H. pylori infection is linked with a reduced risk of Barrett's esophagus (32%–56%) and esophageal adenocarcinoma (36%–44%; ref. 11), but eradicating H. pylori infection does not increase their risk (12). Another gram-negative bacterium associated with cancer is Chlamydia trachomatis, whose association with cervical cancer is well noted. A pooled analysis of data from different countries shows a positive association between C. trachomatis infection and a higher risk of squamous cell carcinoma though some serotype variations are noted (13). Serotypes C, F, H, and K do not show any association with squamous cell carcinoma, whereas serotypes B, D, E, G, I, and J correlate with higher risk (14). C. trachomatis also associates with serous and mucinous epithelial ovarian cancer (15). Several other candidate microbes have been shown to be associated with multiple cancers, for instance, Salmonella enterica serovar Typhi (S. typhi) with gall bladder cancer (16); Fusobacterium nucleatum and enterotoxigenic Bacteroides fragilis (ETBF) with colorectal cancer (17, 18). In contrast to procancer associations of ETBF with colon cancer, monocolonization with nontoxigenic Bacteroides fragilis reduces colon cancer burden in mice and helps maintain immunophysiologic balance (19). Indeed, identification of infections with pathogens can be utilized for risk stratification and early detection of certain cancers.
Dysbiosis induces cancer progression. A healthy gut microflora maintains a protective microenvironment by preserving mucosal integrity and normal cell proliferation. Dysbiosis and increase in opportunistic pathogens compromises mucosal integrity, resulting in leaky gut setting stage for widespread inflammation. Increase in genotoxin-producing microbes results in ulcerative colitis, genomic instability, and hyperproliferation, leading to polyp formation. The pathogenic strains then form bacterial biofilms where specific strains of bacteria communicate via quorum sensing, encouraging production of toxins, microbial metabolites, and signaling molecules, leading to chronic inflammation, recruitment of immune cells, and cytokine upsurge, finally causing carcinoma.
Several notions have been proposed to explain the involvement of candidate microbes in cancer initiation and progression. The “alpha-bug” hypothesis promotes that the microbiota at a target site can be modulated by the existence of a pathogen species like B. fragilis toxin (BFT)–producing ETBF (20). Another interesting model is the “driver-passenger model,” where the “driver” bacteria, for example ETBF, initiate aggravated inflammation and produce toxins leading to mutations and cellular proliferation and eventual development of precancerous lesions that are then colonized by “passenger” bacteria, for instance Fusobacterium spp. (21). In addition, mechanistic insights have been provided for various bacterial genotoxins (Fig. 2; refs. 22–24). Many toxins, including cytolethal distending toxin synthesized by gram-negative bacteria, colibactin synthesized by B2 phylogroup Escherichia coli, and typhoid toxin synthesized by Salmonella enterica serovars, are known to cause DNA damage and modify local immune environment in tissues (Supplementary Table S1; refs. 25–27). Mechanistic studies aimed at BFT secreted by ETBF (28, 29) show that BFT directly affects colonic epithelial cells via cleavage of E-cadherin–disrupting catenin–cadherin complexes, resulting in the induction of β-catenin-cMyc hyperproliferative pathway (30). Cytotoxin-associated gene A protein (CagA), the bacterial oncoprotein of H. pylori, recruits PKCδ via eEF1A1 and phosphorylated Stat3 to upregulate IL6 and IL11 levels to promote carcinoma (31, 32). C. trachomatis infection decreases the tumor-suppressor caveolin-1 expression and elevates c-myc, which may help promote cervical cancer (33). In addition, phosphoproteomic and transcriptomic analyses reveal several candidate kinases and transcription factors involved in C. trachomatis–induced epithelial-to-mesenchymal transition in host cells (34). A recent seminal study revealed that genotoxic pks+ E. coli induces a distinct mutational signature in human intestinal organoids via colibactin, and this unique mutational signature was also corroborated in 5,876 human cancer genomes, mainly in colorectal cancer showing that direct exposure to bacteria possessing pks pathogenicity islands can trigger a mutational process (35). Ultimately, identification of potential oncogenic bacteria and a better mechanistic understanding by which they support cancer initiation and progression will not only help in early detection and effective treatment but may also help in cancer prevention.
Microbial drivers of cancer and their underlying mechanisms. Three known bacterial drivers in carcinoma are F. nucleatum, pks+ E. coli, and B. fragilis. B. fragilis synthesizes unique toxin called BFT, which induces multiple oncogenic pathways including JAK/Stat and Wnt/β-catenin pathways. It is also known to increase cellular ROS generation via SMOX pathway. BFT induces genomic instability and mutation via NF-kB–inducing inflammation. B. fragilis polysaccharide A induces Th17 immune response by recruiting immune cells. Colibactin produced by pks+ E. coli induces genomic instability and mutation by generating stable DNA adducts. Adhesins and LPS expressed by F. nucleatum promote carcinogenesis in multiple ways. It induces Wnt pathway via E-cadherin increasing cellular proliferation; induces chemoresistance by increasing cytoprotective autophagy via TLR4; increases proinflammatory cytokine production; and promotes immune evasion.
Microbiota as an upcoming diagnostic and prognostic tool in cancer management
Multiple studies have revealed that the microbiota is predictive of disease severity and therapy outcome in cancers including, but not limited to, lung, pancreatic, colorectal, oral, breast, prostate, and liver cancers. Toward this end, a recent study examined the microbial reads from 18,116 samples over 10,481 patients encompassing 33 cancer types from The Cancer Genome Atlas consortium for whole-genome sequencing and whole transcriptome sequencing (36). Intriguingly, unique microbial signatures are identified in tissue and blood, and despite using very stringent selection and exclusion criteria, they are predictive for patients with stage Ia–IIc cancer as well as cancers with no genomic alterations. Multiple cancer types can be distinguished from healthy tissue using only plasma-derived cell-free microbial nucleic acids (36). Variations in gut, blood, and target tissue microbial diversity are being investigated in relation to early cancer detection and progression markers (Table 1).
Key microbiota signatures in various cancer types.
Extensive research shows the predictability of pancreatic cancer clinical outcomes based on microbial diversity. Fecal microbial signatures for pancreatic cancer show an enrichment of Bacteroidetes and decreased levels of Firmicutes (37). A signature distinct from pancreatic cancer comorbidities like bile-duct obstruction and liver damage includes an overrepresentation of Veillonellaceae, Akkermansia, and Odoribacter and underrepresentation of Clostridiaceae, Lachnospiraceae, and Ruminococcaceae in patients with pancreatic cancer (37). Intriguingly, tumor microbiome can predict survival of patients with pancreatic cancer where long-term survival associates with higher alpha-diversity and a distinct microbial signature (Pseudoxanthomonas, Streptomyces, Saccharopolyspora, and Bacillus clausii; ref. 38). Interestingly, pancreatic cystic fluid also harbors microbiota rich in Bacteroides spp., Escherichia/Shigella spp., Acidaminococcus spp. Staphylococcus spp., and Fusobacterium spp. (39). In addition to gut and pancreatic tumor microbiome, oral microbiome also shows association with pancreatic cancer (40) where decreased risk correlates with higher levels of Haemophilus, whereas increased risk associates with overrepresentation of Enterobacteriaceae, Lachnospiraceae G7, Bacteroidaceae, or Staphylococcaceae in oral cavity (41). Another intriguing finding is the enrichment of oral pathogens in the early cystic precursors of invasive pancreatic cancers (42).
Gut microbiome can promote hepatocellular carcinoma (HCC; ref. 43) by various mechanisms including retrograde translocation from gut to liver via hepatic portal vein, leaky gut, microbe-associated molecular patterns, and microbial metabolites (44). Patients with cirrhosis are at a higher risk of developing HCC; hence, gut dysbiosis associated with cirrhosis may have predictive value. Indeed, cirrhotic patients with HCC show increased Bacteroides/Prevotella ratio, an abundance of Erysipelotrichaceae (3-fold), and a 5-fold decrease in family Leuconostocaceae and Fusobacterium in comparison with no-HCC group. A Random Forest model trained on microbiota changes in cirrhotic patients with HCC can correctly predict patients with HCC (45). Degree of dysbiosis (Ddys) of the gut microbiota increases with the stage of HCC and can be used during treatment of primary HCC (46). In addition to gut microbiome, analysis of blood microbiome presents a model comprising of 5 HCC-associated genera (Streptococcus, Bifidobacterium, Pseudomonas, Staphylococcus, and Trabulsiella) that can accurately distinguish HCC and controls (47). Specific blood dysbiosis is correlated with nonalcoholic steatohepatitis, cirrhosis, and HCC, but it is still unclear if blood microbiome truly participates in the pathogenesis of HCC. Nonetheless, this circulating microbiota signature presents an opportunity for noninvasive early detection of HCC.
Detection of microbiota in so-called “sterile organ” lung sparked many studies investigating the connection between lung cancer and microbiome. Significant differences mark lung microbiome in squamous cell lung carcinoma and lung adenocarcinoma with and without metastasis (48, 49). Striking changes have been observed between lung microbiota of healthy individuals and patients with lung cancer that can be used as a potential tool for early detection of lung cancer by sequencing or culturing of bronchoalveolar fluid. Local microbiota of patients with lung cancer is enriched in Haemophilus influenza, Streptococcus viridans, Enterobacterspp., and Escherichia coli, whereas overall diversity is reduced in patients with lung cancer (50). Furthermore, genera Veillonella and Megasphaera were enriched in bronchoalveolar lavage of patients with lung cancer (51). Enrichment of Koribacteraceae correlates with increased disease-free survival (DFS) and recurrence-free survival (RFS), whereas abundance of Bacteroidaceae, Lachnospiraceae, and Ruminococcaceae associates with reduced RFS and DFS (52). In addition to lung microbiome, oral microbiome also associates with lung cancer as saliva samples of patients with lung cancer exhibit abundance of Veillonella, Neisseria, Capnocytophaga, and Selenomonas (53). These studies indicate that analyses of saliva or bronchoalveolar fluid can serve as a prediction tool.
Multiple lines of evidence have proved that the microbiome, both local and gut, has a definitive role in breast carcinogenesis as well a distant metastasis (54–58). Interestingly, breast tumors have higher bacterial load and richness in comparison with tumor-adjacent breast tissue and normal breast tissue (59). Breast cancer is a heterogeneous disease with multiple subtypes, and we now know their association with distinctive microbial signatures (60) indicates much deeper biological role of tumor microbiota in diseased breast. Comparative analyses of benign and malignant breast cancer show an enrichment of genus Propionicimonas and families Methylobacteriaceae, Nocardioidaceae, Rhodobacteraceae, Caulobacteraceae, and Micrococcaceae in malignant tissues. An enrichment of taxa of lower abundance including the genera Lactobacillus, Hydrogenophaga, Fusobacterium, Atopobium, and Gluconacetobacter is observed in malignant breast (61). Also, specific alterations in microbiota such as decreased relative abundance of family Bacteroidaceae and increased abundance of genus Agrococcus are noted with increasing histologic grade of malignancy (62). Various studies have shown striking differences between microbiota of normal tissue, benign disease, primary tumors, and metastatic lesions (61–63) and have set a stage for the development of prognostic tools.
Distinct microbiome signatures have been reported for prostate cancer that correlates with different Gleason scores, grades, and stage of prostate cancer (64, 65). Presence of a prostate tissue microbiome in multiple ethnicities has also been reported (66), though no specific differences have been delineated with respect to ethnic populations. A comparison of prostatic fluid microbiome between prostate cancer and normal controls shows the association of lower diversity with the cancer group (67). Additional studies are needed to establish a prostate cancer–specific microbiome signature. Numerous studies have found associations between gut dysbiosis and colorectal carcinogenesis (68, 69). Potential colorectal carcinogenesis biomarkers are Fusobacterium nucleatum, Group B2 E. coli, Enterotoxigenic B. fragilis, Streptococcus gallolyticus, and Enterococcus faecalis (70, 71). Moving beyond the richness and diversity, Norouzi-Beirami and colleagues recently took a functional approach to interpret gut dysbiosis in relation to colorectal carcinogenesis (72). Extensive bioinformatics analysis suggests that microbial diversity does not always translate to functional diversity; hence, to fully understand dysbiosis–cancer relationship, it is important to investigate the functional changes associated with dysbiosis that might be driving the physiologic changes (72). In the past decade, great strides have been made not only to identify the presence of microbiota in so-called “sterile organs” but also to decipher microbiome signatures associated with normal, benign, different grades and stages of primary tumors as well as malignant lesions. These advances have paved the way for the development of better diagnostic and prognostic tools that take advantage of our metagenome.
Microbiome-Based Therapy: Living Treatments
Recent studies have shown the importance of host microbiome in various aspects of cancer initiation and progression as well as modulation of therapeutic responses. Guided by this knowledge, multiple strategies to modulate host microbiome are currently being investigated to maximize the benefits of cancer therapy and minimize the treatment-related toxicities and comorbidities.
Microbiota modulations as adjuvant cancer therapy
Microbiota can be modified by using dietary modulations, prebiotics, probiotics, synbiotics, or postbiotics. Although intestinal microbial community is influenced by age, ethnicity, geography, genetics, and drug usage, one of the most important determinants is the “diet.” Indeed, diet rapidly and reversibly changes gut microbiota composition (73). Higher dietary inflammatory index associates with increased risk of cancer (74). In a primate model, Mediterranean diet leads to enrichment of Lactobacillus, Clostridium, Faecalibacterium, and Oscillospira genus, and decline of Ruminococcus and Coprococcus population (75), a pattern that has also been observed in humans following a vegetarian diet versus omnivore diet (76). Western diet reduces gut microbial diversity as well as lowers the level of health-promoting Bacteroides population while increasing levels of Firmicutes and Proteobacteria (77). Distinct changes in gut microbes Prevotella and Blautia have been observed in patients with prostate cancer in response to increased red-meat diet (78). Interestingly, diet not only influences the gut microbiota, it can also alter breast microbial community (79). Mediterranean diet results in a 10-fold increase in protective Lactobacillus species in the breast tissue compared with western diet (79). It is known that diet-induced microbiota changes determine host metabolic functions and chronic inflammation, a fact that can be used to advantage in cancer therapy. In fact, several clinical studies are examining the effect of dietary interventions on gut microbiome and/or metabolome in patients with high cancer risk to achieve favorable metabolic and inflammatory state (80–82).
“Probiotics” are essentially living microorganisms used to achieve positive effects in health and disease. Multiple strains of Lactobacillus, Bifidobacterium, and Streptococcus have shown efficacy in cancer treatment both in isolation and in combination (83–85). Bifidobacterium longum administration reduces ETBF infection, suggesting a role for probiotics in eliminating infection with harmful bacteria (86). A combination regimen termed VSL#3 consisting of four strains of Lactobacillus (L. acidophilus, L. casei, L. delbrueckii, and L. plantarum), three strains of Bifidobacterium (B. infantis, B. longum, and B. breve), and Streptococcus thermophilus shows high immunomodulatory activity, inhibits tissue neutrophil influx, and reduces the level of proinflammatory cytokines IL1β, TNFα, IFNγ, IL12, and IL8 (87–89). Another promising probiotic is Akkermansia muciniphila, whose abundance increases in responders to anti–PD-1 immunotherapy in comparison to nonresponders (90–92). Strikingly, oral administration of Akkermansia muciniphila improves response to anti–PD-1 immunotherapy in nonresponders (90). Additional “good bacteria” associated with better response to anti–PD-1 immunotherapy are B. longum, Collinsella aerofaciens, and Enterococcus faecium (91) whose individual contributions are yet to be tested. Although search is ongoing for the identification of specific strains of microbes associated with modulation of specific therapies, it is evident that probiotic administration can improve response to traditional cancer therapeutics as well as reduce inflammation.
“Prebiotics” refer to dietary components that encourage the growth of beneficial bacteria in the gut. Recent years have seen multiple preclinical and clinical studies showing the benefits of prebiotics, but it is important to remember that “one size does not fit all.” Oral prebiotic amylase-resistant starch does not prevent acute radiation proctitis in patients with cervical cancer when given in conjunction with chemotherapy (93). But a prebiotic regimen containing fructo-oligosaccharides, xylooligosaccharides, polydextrose, and resistant dextrin improves serum immunologic indicators and reverses surgical stress-induced increase in opportunistic pathogens in patients with colorectal cancer when given 7 days prior to surgery (85). Similarly, prebiotics (inulin and fructo-oligosaccharides) administered prior to therapy encourage growth of Lactobacillus spp. and Bifidobacterium spp. and offer relief from radiotherapy-induced side effects in patients with gynecological cancer (94, 95). These clinical trials indicate that timing of administration of prebiotics is important and show that more benefits are reaped when prebiotics are given prior to the therapeutic intervention. In addition to reducing side effects associated with cancer therapy, studies have been examined whether prebiotics can modulate gut bacteria to prevent cancer (96, 97).
Some studies have examined the positive effects of “synbiotics” referring to the combined administration of probiotics and prebiotics (98, 99). Significant reductions in inflammatory cytokines are observed in patients with colorectal cancer upon synbiotics (Simbioflora) administration 7 days prior to surgery (100). In contrast to prebiotics that seem to impart beneficial effects when started prior to therapy/surgery, synbiotics show reduction in chemotherapy-induced side effects even when it is administered in conjunction with chemotherapy (101). Additional tools gaining traction to modulate the gut microbiota are “postbiotics” that include substances produced or released by metabolic activity of microbes that typically benefit the host. Postbiotics include exopolysaccharides, cell-free supernatants from beneficial bacteria, enzymes, bacterial cell wall fragments, short chain fatty acids, bacterial lysates, and bacterial metabolites (102). Additional research is still needed to answer critical questions related to dose, timing and specificity of dietary modulations, prebiotics, probiotics, synbiotics, and postbiotics in cancer prevention and treatment.
Fecal microbiota transplantation as adjuvant cancer therapy
Fecal microbiota transplant (FMT) has been successfully used for recurrent Clostridium difficile infections, inflammatory bowel diseases, and intractable functional constipation (103). In recent years, it has gained immense interest in cancer treatment owing to the recognition of gut microbiota as an important player. Proposed routes of FMT administration include capsule, nasogastric tube, nasoduodenal tube, enema, or colonoscopy (104). Preclinical studies have shown that FMT from healthy controls protects against chemotherapy-induced toxicity (105) and induces anti-inflammatory function in colon cancer (106). In contrast, FMT from patients with colorectal cancer to Apcmin/+ mice or carcinogen-induced colon cancer models leads to significantly increased tumor burden via activation of Wnt/β-catenin pathway, increased tumor proliferation, and higher levels of proinflammatory cytokines (107, 108). Highlighting the difference in gut microbiome of patients that respond to therapy versus nonresponders, FMT from patients with cancer that responded to PD-1/PD-L1 therapy elicits retardation of tumor growth and better response to immunotherapy in recipient mice (90, 92). Another interesting observation is that altered gut microbiota after FMT can effectively modulate tumor microbiota, tumor growth, and immune infiltration (38). Immune checkpoint inhibitors (ICI) have shown efficacy against multiple cancer types but are marred by the immune-related adverse effects (irAE). In this regard, FMT may prove useful against irAEs as FMT from a healthy donor alleviates ICI treatment–related colitis (109). FMT is currently being evaluated in multiple clinical studies to decipher its impact on restoration of gut microbial diversity to alleviate treatment-related side effects (Supplementary Table S2). Few limitations associated with FMT studies are the identification of optimum healthy donors, reproducibility of donor-microbiome, recruitment of large number of patients, inclusion of control arms, and careful assessment of timing and frequency of FMT. Few adverse effects of FMT including the accidental introduction of a pathogenic species have also been noted (110, 111) that need careful consideration. The Food and Drug Administration has issued guidelines for FMT that are continuously revised to address new situations, for instance, the risk of SARS-CoV-2 exposure in 2020.
Another strategy that has gained attention is the “lab-grown microbiota consortium,” which can potentially circumvent the accidental introduction of a pathogenic species, a caveat associated with a traditional FMT. Though most of the research has been conducted on recurrent C. difficile infection and IBD, lab-grown microbial cultures have shown promise and proved to be safer over traditional FMT. A phase II clinical trial shows treatment as well as prevention of recurrent infections with C. difficile using oral administration of spores from nontoxigenic C difficile strain M3 (112). In another strategy, 33 bacterial isolates purified from a healthy donor show efficacy against C. difficile infection (113). These proof-of-principle studies show the efficacy of lab-grown microbiota consortium, but their efficacy in cancer therapy has yet to be tested. A limiting factor associated with this approach is that a large proportion of the gut microbiota is known to be viable but nonculturable; hence, it might be difficult to replicate all the benefits of a traditional FMT.
Bacteria as targeted cancer therapy
The idea of utilizing bacteria for cancer treatment was introduced more than a century ago when Dr. William Coley used a mixture of inactivated Streptococcus pyogenes and Serratia marcescens, termed Coley's toxin, against cancer (114). Renewed interest in this treatment strategy has emerged with better understanding of host–microbiome interactions. Multiple strains of obligate as well as facultative anaerobes, including Salmonella, E. coli, Clostridia, Corynebacterium, and Bifidobacterium, have proven their efficacy in cancer treatment in translational studies. This approach holds promise for most solid tumors as they naturally develop a hypoxic core attracting anaerobic bacterial strains for colonization. By virtue of this specificity, apathogenic Clostridium can function as a plausible vector delivering specific gene products or liposome-encapsulated drugs in hypoxic core of solid tumors (115, 116). Intratumoral injection of attenuated Clostridium novyi spores elicits a precise, localized antitumor response in rodent and canine models, an observation also reproduced in a human patient with an advanced leiomyosarcoma (117). Systemic treatment with Bifidobacterium species has also shown efficacy as targeted drug carriers in liver and lung cancer models (118–120). In addition, many facultative anaerobes including multiple strains of E. coli, Salmonella, Vibrio cholerae, and Shigella flexneri show favorable response in several cancer models (121–124). Of note, Salmonella enterica serovar Typhimurium presents a unique advantage as it can grow in both aerobic and anaerobic conditions allowing its colonization in small nonhypoxic tumors, metastatic lesions, and multiple cell types including epithelial cells, macrophages, and dendritic cells making it most suitable for cancer therapeutics (125, 126). Intriguingly, S. typhimurium administration inhibits metastatic progression in pancreatic (127) and prostate cancer (128). Mechanistically, tumor cells secrete unique chemoattractant (129) to attract Salmonella, which express surface chemoreceptors, facilitating accumulation of the bacteria in specific regions of the tumor. Tumor-targeting bacteria owing to their vast gene packaging capacity and tumor selectivity present a unique opportunity to not only deliver the drug into the tumor but also activate them at site to enhance efficiency and reduce systemic toxicity.
Future Directions and Challenges
Initial observational studies showing the presence of gut and tumor tissue microbiota have paved the way for queries designed to explore the causal relationship between dysbiosis and cancer. Deciphering key microbiota changes between normal tissue and cancer tissue at various steps of cancer progression (grade, stage, and metastases) can help delineate dysbiosis signatures. Understanding dysbiosis may also help us answer the questions regarding discrepancies in cancer progression and response to therapy in individuals with otherwise similar clinical profile. Various candidate microbes have been identified that modulate tumor cells directly or indirectly via production of toxins. Some “alpha bugs” are known to dominate the microbiota or pave the way for a “passenger” pathogenic bacteria. The obvious next steps are to decipher their underlying mechanisms and develop better strategies for early detection and effective interventions. Multiple strategies are being developed to alter microbiota through dietary modulations, prebiotics, probiotics, synbiotics, or postbiotics with the underlying idea to push dysbiosis toward eubiosis. FMT, lab-grown microbiota consortium, and engineered bacteria are some of the upcoming strategies to improve therapeutic responses in patients with cancer (Fig. 3).
Role of gut microbiota in cancer progression and strategies to exploit it to aid cancer therapy. A number of factors determine the community composition of intestinal microbiota including diet, physical activity, age, ethnicity, geography, host genetics, and consumption of pro-, pre-, and postbiotics. Dysbiosis can result from poor lifestyle, unhealthy diet, surgical procedures, antibiotic and drug usage, altered circadian rhythm, and alcohol and tobacco consumption. Dysbiosis aids in tumor progression in multiple ways like inducing genomic instability and mutations by genotoxins, such as BFT and colibactin; sustained proliferative signaling via F. nucleatum FadA, colibactin, and BFT; inducing tumor angiogenesis via TLR ligands; and inducing a tumor-promoting inflammatory state rich in cytokines and chemokines. A healthy microbiome, on the other hand, promotes an anti-inflammatory state, regulates hormone levels, activates drugs by metabolizing them, and encourages immune infiltration into the tumors. A healthy microbiota can be restored by various strategies including FMT, administration of probiotics, prebiotics, and specific microbial strains. Some newer strategies include usage of bacteriophages to eliminate selective pathogens and capsules filled with spores from healthy donors. Direct targeting of tumors has also been shown to be a promising approach using oncolytic microbes like S. typhimurium, V. Cholerae, S. flexneri, etc. Bioengineered strains like recombinant B. longum and S. typhimurium VPN20009 and spores from attenuated strains like C. novyi have also been shown to be effective.
Although exciting, there are inherent challenges associated with microbiome research. Microbiota composition can change with geography, age, dietary patterns, body mass index, prescription drugs, antibiotics, and pet ownership; hence, utmost care is required in designing the study cohorts and incorporating all the important variables in statistical analyses. Two additional issues that are particularly important for microbiome studies are preserving the original microbiota and avoiding any contaminants during sample collection and analyses. Several groups have addressed the impact of sample storage conditions including preservatives and temperature on microbiota (111, 130–134) and shown that alterations owing to storage conditions are not significant in comparison with variations between individual participants. However, it is extremely important to select storage conditions that minimize changes in the original microbiota and consistently follow them for all the study samples. Contamination of samples during processing is another challenge especially for samples with low microbial content, which can be overpowered by contaminating DNA from lab reagents. Recent reports of the presence of bacteria in DNA extraction kits—“kitome or contamin-ome” that varies between kits and even different lots of same kit (135, 136)—strongly suggest using the same batch of kits for processing microbiome samples or consider “different kits” as a variable. Inclusion of negative control samples to assess the contamination background during all processing steps and well-vetted positive control samples is highly recommended during sequencing. Several microbiome studies have low number of participants owing to the complexity of research questions or unavailability of sufficient resources, but it is extremely important to validate the results of discovery cohorts in independent validation cohorts. Efforts are underway to develop guidelines and standards for microbiome research (137). The Human Microbiome Project conducted by the NIH and METAgenomics of the Human Intestinal Tract (MetaHIT) project orchestrated by European Commission involving 15 institutes from 8 countries rigorously developed and presented good clinical practice standards for microbiome research (138, 139). In addition, the Microbiome Quality Control, another collaborative project, has been focusing on evaluating experimental designs to facilitate the development of guidelines for best practices in the microbiome field (140). With rapid technological advances and development of standard guidelines, human microbiome studies are paving their way to facilitate cancer risk assessment and developing novel prevention and cancer treatment strategies.
Authors' Disclosures
No disclosures were reported.
Footnotes
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Cancer Res 2021;81:790–800
- Received August 6, 2020.
- Revision received October 1, 2020.
- Accepted October 28, 2020.
- Published first November 4, 2020.
- ©2020 American Association for Cancer Research.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
- 116.↵
- 117.↵
- 118.↵
- 119.↵
- 120.↵
- 121.↵
- 122.↵
- 123.↵
- 124.↵
- 125.↵
- 126.↵
- 127.↵
- 128.↵
- 129.↵
- 130.↵
- 131.↵
- 132.↵
- 133.↵
- 134.↵
- 135.↵
- 136.↵
- 137.↵
- 138.↵
- 139.↵
- 140.↵
- 141.
- 142.
- 143.
- 144.
- 145.
Volume 81, Issue 4
