Society: AGA

Background and Aims: Peptostreptococcus stomatis (P. stomatis) is an enriched bacterium in colorectal cancer (CRC) patients as compared to healthy subjects in multiple cohorts. However, the potential role of P. stomatis in colorectal tumorigenesis remains obscure. In this study, we investigate the function of P. stomatis in CRC, and its impact on CRC therapy response.

Methods: Pro-tumorigenic effect of P. stomatis on colorectal carcinogenesis was determined in Apc^Min/+ mice and azoxymethane/Dextran sodium sulfate (AOM-DSS)-treated mice models. In vitro function of P. stomatis was analysed by cell proliferation, apoptosis, and cell cycle assays. Interplay between P. stomatis and CRC cells was evaluated by fluorescence in situ hybridization (FISH), scanning electron microscopy (SEM), transmission electron microscopy (TEM), adhesion and invasion assays. The interaction between P. stomatis surface protein and the receptor of CRC cells was investigated by far-western, biotin pulldown, and glutathione S-transferase (GST) pulldown assays.

Results: Oral gavage of P. stomatis significantly promoted CRC tumorigenesis in Apc^Min/+ mice with increased colon tumor number and tumor load, which accompanied with enhanced cell proliferation, reduced apoptosis and impaired gut barrier function. The pro-tumorigenic effect of P. stomatis was validated in another CRC tumorigenesis model induced by AOM-DSS and in CRC cell line Caco2 formed xenograft models. Consistently, P. stomatis co-culture with CRC cells accelerated cell growth, induced G_1-S cell cycle progression, and suppressed apoptosis. FISH, TEM and in vitro adhesion/invasion assays showed that P. stomatis colonized on CRC cells and colonic epithelium cells of mice, with preferential adherence in tumor cells compared to normal colon cells. Mechanistically, P. stomatis surface protein fructose-1,6-bisphosphate aldolase (FBA) binds to integrin α6/β4 receptor of CRC cells and activates ERBB2 phosphorylation, leading to the consequential activation of MEK-ERK-p90 oncogenic signaling. Integrin α6/β4 knockdown abolished P. stomatis adherence and P. stomatis-induced ERBB2-MEK-ERK-p90 activation in vitro, implying that the FBA-integrin α6/β4 interaction is essential for P. stomatis-mediated tumorigenesis. Finally, we demonstrated that P. stomatis co-culture with colon cancer cell lines Caco-2 and HT-29 significantly abolished the therapeutic effects of Cetuximab and Vemurafenib, implying P. stomatis as a novel factor driving tyrosine kinase (RTK) inhibitor resistance through alternative activation of ERBB2 in CRC.

Conclusion: P. stomatis is a novel oncogenic bacterium in promoting colorectal carcinogenesis. P. stomatis surface protein FBA engages an integrin α6/β4-ERBB2 complex leading to aberrant activation of host oncogenic MEK-ERK-p90 signaling to drive CRC development and cause resistance to RTK inhibitors.

BACKGROUND & AIMS: Peptostreptococcus anaerobius is a pathogenic bacterium enriched in colorectal cancer (CRC) and it promotes colorectal tumorigenesis in mice. Here, we investigated the functional role of P. anaerobius in the tumor immune microenvironment (TIME) of CRC, the molecular basis of its action, and the consequential impact on the efficacy of immune checkpoint blockade (ICB) therapy.
METHODS: Immunomodulatory effects of P. anaerobius in Apc^Min/+, azoxymethane (AOM)- or AOM/ Dextran Sodium Sulfate (DSS)-treated CRC mouse models were investigated by single cell RNA-sequencing (scRNA-seq), flow cytometry and immunofluorescence staining. The effect of P. anaerobius on anti-PD-1 efficacy was evaluated in syngeneic mouse CRC cell line MC38 allograft and in AOM/DSS-induced CRC tumorigenesis in mice. Cytokine array was performed to identify chemokines produced by CRC cells upon P. anaerobius treatment. Immunomodulatory protein secreted by P. anaerobius was identified by mass spectrometry. Microscale thermophoresis and GST pull-down assay were used to identify corresponding receptor on MDSCs.
RESULTS: P. anaerobius significantly induced MDSCs accumulation, whilst suppressed IFNγ⁺CD8⁺T cells in colonic tumors of Apc^Min/+, which was confirmed in AOM-induced CRC and AOM/DSS-induced CRC mouse models by scRNA-seq, flow cytometry and immunofluorescence staining analyses. By fostering an immunosuppressive TIME, P. anaerobius administration abolished the efficacy of anti-PD-1 therapy both in MC38 allografts and in AOM/DSS-induced CRC tumorigenesis in mice. Mechanistically, P. anaerobius simultaneously promotes MDSCs recruitment and activation through a two-step pathway. Firstly, P. anaerobius co-culture promoted CXCL1 secretion by CRC cells via the integrin α₂/β₁-NF-κB axis. P. anaerobius-induced CXCL1 up-regulation in turn stimulated the migration of CXCR2⁺ MDSCs. Knockdown of integrin α₂/β₁or NF-κB inhibitor reversed P. anaerobius-induced CXCL1 secretion by CRC cells, accompanied by attenuated MDSCs chemotaxis. Secondly, P. anaerobius culture supernatant directly activated immunosuppressive function of intratumoral MDSCs. Mass spectrometry revealed a P. anaerobius secreted protein, LytC_22, which binds to SLAMF4 receptor on MDSCs to promote ARG1 and iNOS expression, thus enhancing the ability of MDSCs to inhibit cytotoxic T cell proliferation and function. Blockade of integrin α₂/β₁or Slamf4 receptor abrogated P. anaerobius-induced MDSCs trafficking and activation, and restored anti-PD-1 efficacy in CRC tumor models.
CONCLUSION: P. anaerobius provokes resistance to anti-PD-1 treatment in CRC by promoting MDSCs chemotaxis and activation, which in turn suppresses T cell activation. Targeting integrin α2/β1 or SLAMF4 receptor is a promising strategy to overcome P. anaerobius-induced anti-PD-1 resistance in CRC patients.

Background: Inflammatory Bowel Disease (IBD) is characterized by chronic, relapsing episodes of tissue-damaging inflammation in the gastrointestinal tract. Compelling evidence suggests that patients with IBD have higher chances of developing psychiatric comorbidities (e.g., anxiety and depression), further impairing their disease course and quality of life. The intestinal microbiota has been linked to the modulation of mood and anxiety. Moreover, intestinal bacterial composition and function are altered in patients with IBD. We hypothesized that the microbiota of patients with IBD contributes to the onset of psychiatric comorbidities.

Methods: We used the microbiota-humanized mouse model to test our hypothesis. Fecal samples from two patients with Crohn’s disease (CD) and one age-matched healthy control (HC) were used as microbiota sources to colonize adult germ-free NIH Swiss mice of both sexes. Three weeks after colonization, mice were submitted to a battery of behavioural tests to assess anxiety- and depression-like behaviours; light preference, step-down, open field, digging, marble burying, and tail suspension tests.

Results: Colonization with CD microbiota did not induce overt colonic inflammation as body weight and colon length were similar in HC and CD mice. Male CD mice spent less time in light and displayed increased immobility in the tail suspension test when compared with HC mice, suggestive of anxiety- and depression-like behaviour, respectively. Importantly, no alterations in overall locomotion were observed among the two groups of mice, ruling out sickness-like behaviour. Female CD mice showed increased latency to step down, indicative of anxiety-like behaviour. No changes were observed in the open field, digging, and marble burying tests for both sexes.

Conclusion: Our data show that human-derived microbiota from patients with CD induces changes in emotionality-related tests in a gnotobiotic mouse model, and this occurs in a sex-dependent manner. Future studies are needed to investigate the molecular mechanisms underlying these behavioural changes, to guide the development of novel microbiota-based therapies.

Background: The worldwide disease burden caused by rotavirus (RV) has been reduced by broad deployment of live attenuated RV vaccines. Such orally administered RV vaccines infect enterocytes resulting in elicitation of mucosal antibodies that lastingly protect against virulent RV strains. However, such RV vaccines frequently fail to elicit protective immunity, particularly in children in low-income countries. Such ineffectiveness associates with lack of fecal RV antigens and subsequently RV-specific antibodies following vaccination indicating that RV vaccine failure results from the vaccine virus failing to infect its host. We hypothesize that such RV vaccine failure may, in part, reflect that select microbes harbored in some hosts impede infection of RV vaccine viruses.
Aim: Examine the extent to which microbiomes of children not responding to RV vaccination might contribute to their RV vaccine failure phenotype.
Approach: Groups of germ-free mice were administered fecal microbial transplants (FMT) from Mexican children who had exhibited robust or minimal RV-specific antibodies following RV vaccination. Two weeks later, mice were subjected to RV vaccination. The extent of infection of the vaccine virus was assayed by measuring fecal RV antigen shedding days 0-15. Vaccine efficacy was determined by measuring fecal and serum RV-specific antibodies on days 14-28. Microbes associated with RV vaccine failure were assessed by 16S sequencing.
Results: Microbiome transplant recapitulated donor RV vaccine response phenotype. Specifically, mice receiving FMT from successfully vaccinated children exhibited copious fecal RV antigen shedding and high RV antibody titers in response to RV vaccination. Concomitantly, such mice were impervious to subsequent RV challenges. In contrast, mice receiving FMT from children who had not responded to RV vaccination exhibited only modest levels of RV antigen shedding and RV antibody generation in response to RV vaccination. Moreover, such mice remained prone to RV infection when later challenged with this virus. Microbiota analysis revealed that one microbe associating with RV vaccine failure in some FMT recipients was Clostridium perfringens. Oral administration of cultured C. perfringens to gnotobiotic mice partially recapitulated the RV vaccine non-responder phenotype.
Conclusion: Intestinal microbiota composition can dictate the extent to which RV vaccine viruses infect and thus elicit protective immunity in their host. C. perfringens may be one, perhaps of many, bacterial species harbored in the intestine of RV-vaccine non-responders that influences RV vaccine outcomes.

Robert Britton

Baylor College of Medicine

Appearances