Modulation of Gut Microbiota to Enhance Effect of Checkpoint Inhibitor Immunotherapy

The innate immune system has physiological functions protecting our body. The innate immune system in intestinal consists of epithelial cells, myeloid cells, innate lymphoid cells, and other types of cells (21). Gut microbiota has broader effects on innate immunity (22). The immune system inside the intestinal plays a vital role in preventing external bacterial invasion and infection (22). Gut mucus is the first barrier providing the underlying epithelium from bacteria, to some extent, gut mucus is shaped by microbiota. The penetrability of gut mucus of germ-free mice and conventionally raised mice is different, the small intestine mucus become normally detached and colonic inner mucus become impenetrable after gavaged with microbiota from conventionally raised mice (23). In addition, immunoglobulin A (IgA) is an antibody isotype existing in intestinal lumen (24), mucosal IgA is binding to the polymeric immunoglobulin receptor and secreted across the epithelium (25), and it has a close relationship with Bacteroides fragilis. A research found that Bacteroides fragilis can take advantage of the immune system to settle down in the intestine of mice (Figure 1), however, it is difficult for Bacteroides fragilis to settle on the surface of the intestine and maintain long-term stability once the mice lack IgA, indicating that the immune system has a close relationship with commensal bacteria (26).

Gut microbiota has a close relation with adaptive immunity, especially T cells. A research found that in mice lamina propria, Th17 cells can be induced by Segment filamentous bacteria (SFB). Transplantation of SFB in mice induces the appearance of CD4+ T helper cells, increases level of IL-17 and IL-22 in lamina propria (Figure 1) (27). This research reveals that certain bacteria can cause changes in the number of specific cells in the immune system and expression of immune-related genes in vitro. However, how SFB affects Th17 cells remains unclear. Another research found that Indigenous Clostridium species promote the accumulation of CD4+ T regulatory cells (Tregs) in colonic mucosa. They found clusters IV and XIVa of the genus Clostridium is associated with Tregs accumulation in the colon (Figure 1) (28). Tregs plays an important role in maintenance of immune homeostasis, so this finding might offer a new therapeutic method to autoimmunity disease and allergy (28, 29). Gut microbiota influences immune system mainly through metabolites such as short-chain fatty acids (SCFAs). SCFAs are considered as the most abundant metabolites produced by gut microbiota in gut lumen, and they are produced by various types of bacteria including Faecalibacterium prausnitzii, Roseburia intestinalis, and Anaerostipes butyraticus (30, 31). SCFAs consist of acetate, propionate, and butyrate, they have an extensive influence on immune system. Studies have shown that butyrate can induce Treg cell differentiation in vivo or in vitro (32). Further research found that butyrate regulates Foxp3, which demonstrates that butyrate plays a role in inhibiting the activation and proliferation of T cells (32, 33). Butyrate also has an impact on Th17 cells through controlling the cytokine production by dentritic cells (DCs) (Figure 1). Butyrate produced by gut microbiota increases the level of GPR109a and induces expressing of IL-10 and Aldh1a1 in colonic macrophage and DCs, result in suppressing of development of Th17 cells (34). Feeding mice with Bifidobacterium infantis increased the production of retinoic acid in dendritic cells (DCs), result in increasing of TGF-β production and Treg expansion (35). Studies with gnotobiotic mice reveals an immunomodulatory effect of secreted microbial metabolites. In mice, Bacteroides fragilis secreted polysaccharide A binding to Toll-like receptor 2 (TLR2) on Treg cells, which can enhance the expansion of Tregs while suppressing the proinflammatory Th17 response (Figure 1), this indicated that polysaccharide A secreted by Bacteroides fragilis is able to prevent and cure experimental colitis in animals (36, 37). Through reading all of these researches, we can conclude that gut microbiota is indispensable for maintaining the immune homeostasis in animals and in human.

Gut microbiota has a great influence on gastrointestinal cancer (38, 39). The colonization of Helicobacter pylori (H. pylori) causes a persistent inflammatory response, which can lead to cancers of the stomach (38). Infected individuals with low abundance of H. pylori have a lower risk of gastric cancer, proving that H. pylori is associated with gastric carcinogenesis (40). Further studies revealed that H. pylori might promote cancer through β-catenin signal pathway (38). Colorectal cancer (CRC) is associated with specific bacteria including Bacteroides fragilis, Fusobacterium nucleatum, Porphyromonas asaccharolytica, Parvimonas micra, Prevotella intermedia, Alistipes finegoldii, and Thermanaerovibrio acidaminovorans, which can serve as potential diagnostic markers across population (41). Certain types of gut microbes might promote CRC. Sunny H. Wong et al. feed conventional and germ-free mice with stool samples from CRC patients and healthy individuals, then they found mice feeding with stool samples from CRC patients developed high-grade dysplasia (P <.05) and macroscopic polyps (P <.01) than mice feeding with stool samples from healthy individuals (42). Shaoguang W et al. found that Enterotoxigenic Bacteroides fragilis (ETBF) promotes colon tumorigenesis by stimulating immune response via Th17 cells in mouse CRC model, ETBF stimulates rapid colitis and colon tumors in multiple intestinal neoplasia (Min) mice and induce activation of transcription-3 (Stat3) in the colon of Min mice, Stat3 signaling is necessary for the generation of TH17 cells, and IL-17 produced by CD4+ Th17 is sufficient to induce tumorigenesis in the absence of γδ T cells (43). Gut microbiota also has a direct role in the occurrence of oncogenic mutations in colorectal cancer, a research reveals that colibactin—a compound generated by Escherichia coli—is believed to alkylate DNA on adenine residues and cause DNA double-strand breaks, which is related to the generation of colorectal cancer (44). Tumor promoting effects are mostly related to depletion or dysbiosis of gut-microbiota. After killing some bacteria with antibiotics (ATBs) in mice, the development of the tumor in the liver and the colon can be reduced (45, 46). Dianne H. Dapito et al. feed mice with ATBs which can eliminates commensal bacteria and reduce systemic lipopolysaccharide levels, then they can be found in DEN plus CCl4 HCC model, the tumor size was reduced (45).

Liver does not have direct contact with microbiota, but it has tight anatomic links to the gut (47, 48). In most of the studies, liver cirrhosis is a major potential risk of liver cancer (49). In 2016, people noticed that gut microbiota might associate with HCC progression in cirrhotic patients (50). Until now, there is no clinical evidence showing that a specific bacteria is associated with HCC (51). But experiments on gut-sterilized mice or germ-free mice showed that the development of HCC might associate with microbiota or microbially activated pathways, microbial metabolites, or microbiota-associated molecular patterns (MAMPs) also associate with the development of HCC (51). Gut microbiota and Toll-like receptors (TLRs) are required for HCC promotion but not required for HCC initiation (45). One research revealed that gut microbiota is associated with cholangiocarcinoma. According to the study which included 60 patients with cholangiocarcinoma, there existed a distinct, tissue-specific microbiome in bile duct tissues. Compared with normal samples (non-neoplastic liver), the patients with cholangiocarcinoma (CCA) tumors had a significant increase in Stenotrophomonas species in bile duct tissues (52).

There are also several studies focused on the gut microbiota and breast cancer. In 2015, Bard and his colleagues observed that gut microbiota was associated with clinical stages of breast cancer. They collected fecal samples from different stages of breast cancer patients, and performed 16S rRNA sequencing, they found that stage I patients had a lower absolute number of Blautia sp. than stage III, several other bacterial species (Bifidobacterium, Blautia, F. Prausnitzii, and Blautia) were different according to different clinical stages of breast cancer (53). Another research also found that gut microbiota was associated with different stages of breast cancer, compared to patients in stage 0/I, the number of Bacteroidetes, Clostridium coccoides cluster, Clostridium leptum cluster, Faecalibacterium prausnitzii, and Blautia sp. are increased in stage II/III patients. In addition, the absolute numbers of total bacteria are different according to the patient’s body mass index, which illustrated that obesity is associated with gut microbiota (54). Gut microbiota and lung has an essential cross-talk called “gut-lung axis” (55). In lung cancer patients, the level of Enterococcus sp. was decreased and the levels of Actinobacteria sp. and Bifidobacterium sp. were increased (56), indicating that lung cancer had an impact on the composition of gut microbiota.

Metabolites of gut microbiota are known to induce proinflammatory cytokines and mediate tumor-associated inflammation in colon cancer (57, 58). The short chain fatty acids (SCFAs) secreted by bacteria (especially propionic acid and butyric acid) can mediate anti-inflammatory response and reduce the occurrence of colorectal cancer (59). Besides, SCFAs mediate p21 gene expression through modulation of microRNA, which will in turn influence colonic carcinogenesis (60). Bacteroides fragilis in the intestines of patients with colorectal cancer produce a variety of toxic metabolites such as β-glucuronidase, spermine oxidase, reactive oxygen species, reactive nitrogen species, and nitroso compounds, which can induce DNA damage and promote colorectal cancer (61, 62). Bile acid is an important component synthesized by the liver to regulate fat metabolism. Some anaerobic bacteria in intestine such as Bacteroides metabolize bile acids into secondary bile acids, which can cause DNA damage, stimulate EGFR or wnt/β-cantenin pathway, inducing colorectal cancer or liver cancer (46, 63, 64). Dietary or genetic obesity induces alterations of gut microbiota. Fed mice with high-fat diet resulting in the increase of Gram-positive bacteria (46). High-fat diet also increases the levels of bacterially generated deoxycholic acid (DCA). DCA is an intestinal bacterial metabolite which can cause DNA damage (65). Accumulating of DCA can facilitate obesity-associated HCC development in mice (46). In liver cancer, bile acid metabolism changes caused by gut microbiota regulates the expression of CXCL16 and recruits CXCL16-mediated natural killer T (NKT) cells, which could control the liver tumor growth (66). NKT cells can kill tumor cells in a CD1d-dependent manner (51). Recent study revealed that inosine—a purine nucleoside molecule produced by gut microbiota—can enhance cancer immunotherapy response in mice (67).

In 2015, Thomas and his colleagues first noticed that there were correlations between gut microbiota and ICI immunotherapy (18). They used mice which were harbored with different commensal microbiota, then compared the melanoma growth of these mice. They also found that different microbiota might relate to different spontaneous antitumor immunity. Of which, they found that Bifidobacterium could facilitate antitumor effect of PD-L1 blockade (Table 1). Commensal Bifidobacterium controlled the growth of melanoma tumor in mice, oral administration of Bifidobacterium was an effective way to block tumor growth. Bifidobacterium treated mice showed a better antitumor effect compared to non-Bifidobacterium treated mice and this tumor control effect was related to tumor-specific T cells in periphery and accumulation of antigen-specific CD8+ T cells within the tumor (18). This study demonstrated that commensal Bifidobacterium can enhance antitumor immunity in vivo with an antigen-independent manner and it has a synergistic effect with PD-L1 blockade (18). In 2020, another team found that Bifidobacterium enhanced antitumor immunity through production of metabolite inosine (68). In this research, they found three bacterial species—Bifidobacterium pseudolongum, Lactobacillus johnsonii, and Olsenella species, these three bacterial species can significantly enhance the efficacy of ICIs in mouse models of CRC (Table 1) (68). They found that Bifidobacterium pseudolongum can enhance the antitumor effect through metabolite inosine, and this effect was dependent on adenosine 2A receptor (A_2AR) signaling specifically in T cells. They also found that the antitumor effect of Akkermansia muciniphia in human also relies on inosine-A_2AR signal (Figure 2). Recent research found that Bifidobacterium can enhance local anti-CD47 immunotherapy in tumor (71). The researchers used wild type (WT) mice, ATB-fed mice, and sterile mice to test the antitumor effects of CD47-based immunotherapy. The result showed that intestinal microbes outside the gastrointestinal tract are effective in promoting CD47 tumor immunotherapy. The results of Bifidobacterium injection experiments show that Bifidobacterium has antitumor effects in mice that do not respond to CD47 inhibition (71). This study reveals that the tumor-targeting ability of Bifidobacterium might be a possible mechanism that gut microbiota affect the antitumor response.

In 2015, another group published a research article focused on the gut microbiota and the treatment of CTLA-4 blockade (19). They find that the effect of CTLA-4 blockade treatment is related to distinct Bacteroides species (Bacteroides fragilis and/or Bacteroides thetaiotaomicron and Burkholderiales) (Table 1). If gut microbiota was removed or destroyed in mice, then CTLA-4 blockade has no effect on tumors. Transplant with B. fragilis can benefit for antitumor effect of CTLA-4 blockade. They found that Bacteroides thetaiotaomicron or Bacteroides fragilis cause T cell response, which is associated with antitumor efficacy of CTLA-4 blockade (19). Treatment of anti-CTLA-4 blockade also influence the abundance of immunogenic Bacteroides spp. in the gut, which would in turn affect the antitumor efficacy of ICIs. Further research revealed that CTLA-4 blockade influences the composition of microbiota, oral feeding of Bacteroides fragilis induce Th1 immune response, promoting the maturation of dendritic cell, which facilitate tumor control in mice and patients (Figure 2) (19). However, a research focused on the PD-1 immunotherapy of melanoma patients revealed that responders have a lower abundance of Bacteroidales than the non-responders (17). They collected 112 melanoma patients who were undergoing anti-PD-1 immunotherapy, then they examined the gut microbiome of these patients, and they found the diversity and composition of gut microbiota are different between the responders and non-responders: Clostridiales order and Ruminococcaceae family are enriched in responders while Bacteroidales order are enriched in non-responders. The abundance of Bacteroidales were positively related to the frequency of Treg cells and myeloid-derived suppressor cells (17). This study revealed that the microbiome composition might have an impact on anti-PD-1 immunotherapy, patients with favorable gut microbiome like Ruminococcaceae or Faecalibacterium might have an enhancing antitumor immune response, and this antitumor immune response was related to more antigen presentation in the periphery and improved effector T cell function in the tumor microenvironment. On the contrary, patients with a high abundance of Bacteroidales might have a limited response to anti-PD-1 immunotherapy (17).

PD-1-based immunotherapy is also associated with gut microbiota. Analyzing of stool samples from renal cell carcinoma (RCC) and non-small cell lung cancer (NSCLC) patients after treating with PD-1 blockade revealed that clinical outcomes were associated with the abundance of Akkermansia muciniphia, oral supplementation with Akkermansia muciniphia to non-responders restored the efficacy of PD-1 blockade, which demonstrated that the Akkermansia muciniphia can enhance the effect of PD-1 inhibitor (Table 1) (20). They also found that treating with ATB in mice inhibited the antitumor effect of PD-1 blockade, which demonstrated that dysbiosis of gut microbiome system is another factor affecting results of ICIs (20). Fecal microbiota transplantation (FMT) from patients who respond to ICIs to mice can enhance the antitumor effect of PD-1 blockades in mice, while FMT from non-responders failed to do so. Oral supplementation with Akkermansia muciniphia to non-responders can enhance the efficacy of PD-1 blockade, besides, a higher incidence of E. hirae was found in responders of NSCLC patients than in non-responders (20). Another research consisting of 31 metastatic renal cell carcinoma (mRCC) patients also found that the abundance of Akkermansia muciniphila is positively correlated with the outcomes of checkpoint inhibitors. After treating with nivolumab or nivolumab plus ipilimumab, stool samples were collected and analyzed from mRCC patients, the result revealed that several bacteria are related to clinical benefit. They also found that a better treatment outcome is associated with high microbial diversity (72). The mechanism of Akkermansia muciniphila which promotes the effect of ICIs might be related to inosine (Figure 2), Akkermansia muciniphila can produce inosine in vitro in ICIs-treated tumors. Anti-CTLA-4 immunotherapy combined with monocolonization of Akkermansia muciniphila increased antitumor immunity, and this was associated with T cell expression of A_2AR (68). In HCC patients, Akkermansia muciniphila increased in responders after PD-1 immunotherapy. Besides, Ruminococcaceae spp. and other 18 species of bacteria were also increased in responders. In non-responders, the level of Proteobacteria increased (73).

Gopalakrishnan V et al. examined gut microbiome of melanoma patients who were treating with anti-PD-1 immunotherapy, they found the composition and diversity of responders are different from those of non-responders. Clostridiales order and Ruminococcaceae family and Faecalibacterium genus were enriched in responders and Bacteroidales order was enriched in non-responders (17). The existence of Ruminococcaceae family, Clostridiales order, and Faecalibacterium genus can promote the infiltration of CD8+ T cells in tumors (Table 1). Patients rich in Clostridiales, Faecalibacterium, or Ruminococcaceae in gut have a higher frequency of CD4+ and CD8+ T cells. Besides, patients with high abundance of Faecalibacterium had better antigen processing and presenting ability and a relatively high density of immune cells compared to patients with high abundance of Bacteroidales in the gut (17).

Since gut microbiota is a whole system, a small change in one species of bacteria may have an impact on others. A recent research reveals that after immunotherapy, the ratio of Prevotella and Bacteroides will be elevated (Table 1) (69). In a research which included 74 patients with advanced-stage gastrointestinal cancer, after anti-PD-1/PD-L1 treatment, they observed an elevation of Prevotella/Bacteroides ratio in patients with a preferred response to anti-PD-1/PD-L1 treatment and a particular responder harboring a significantly higher abundance of Prevotella, Ruminococcaceae, and Lachnospiraceae. Ruminococcaceae is also associated with a favorable objective response rate in NSCLC after treatment of immune checkpoint inhibitors (74). Besides, gut bacteria which are capable of SCFAs, were positively associated with anti-PD-1/PD-L1 response (69).

As we discussed before, gut microbiota is a complex system. The alteration of whole system not just specific bacteria also plays an important role which can influence the effect of ICIs. Lisa Derosa et al. analyzed fecal samples from RCC patients treated with nivolumab, they found the patients who had used ATBs recently reduced objective response by 19% compared to the patients who had not used ATBs, and the composition of gut microbiota was significantly changed in these patients (75). This research indicates that ATBs shift the composition of gut microbiota, resulting in decreased efficacy of ICIs. Bertrand Routy et al. studied the impact of ATBs on NSCLC, RCC, and urothelial patients who were received anti PD-1/PD-L1 blockade, and they found that progression-free survival (PFS) and overall survival (OS) were significantly shorter in the ATB-treated group, demonstrating that dysbiosis has an impact on therapeutic efficacy of ICIs (20). Besides, alpha diversity is also important factor in the composition of gut microbiota. Gopalakrishnan V et al. analyzed fecal microbiome samples from 43 melanoma patients who were undergoing PD-1 blockade immunotherapy, and they found responding patients have a higher alpha diversity of bacteria than the non-responders, they also found that patients with higher alpha diversity in their fecal samples had significantly prolonged PFS than the patients in low alpha diversity (17).