Vaccination and Microbiota Manipulation Approaches for Colon Cancer Prevention in Rodent Models

Colorectal cancer represents the third most common cancer type worldwide and is a leading cause of cancer-related mortality in the United States and Western countries (1). Colorectal cancer is caused by the accumulation of genetic mutations and epigenetic alterations that drive the transformation of the colonic mucosa and the progression from early colonic lesions to adenomas and adenocarcinomas. Although the vast majority of colorectal cancers are considered sporadic, about 2% to 5% of colorectal cancers are caused by pathogenic variants in DNA mismatch-repair (MMR) genes or the adenomatous polyposis coli (APC) (2) gene resulting in the highly penetrant hereditary cancer syndromes Lynch syndrome (LS) and familial adenomatous polyposis (FAP), respectively (3, 4). Another 20% to 30% of colorectal cancer cases are considered to have a hereditary component due to an associated family history or germline mutations in other genes (5). In addition to genetic alterations, environmental and dietary factors significantly impact the development of colorectal cancer. For example, although red meat consumption and smoking increase the risk for colorectal cancer, vegetables and fruit with high fiber content decrease colorectal cancer risk (6–8). In addition, resistant starch, a component of dietary fiber in certain foods, significantly lowered the risk of extra colonic cancers in LS patients in the CAPP2 clinical trial study (9). Furthermore, chronic stress and the widespread use of antibiotics with alteration of the gut microbiota may be linked to early-onset colorectal cancer (10, 11). Although in recent years the overall incidence of colorectal cancer has declined in older individuals (>50 years), likely due to more extensive colorectal cancer population screening, the colorectal cancer incidence has significantly increased in younger adults (12) with the causes remaining largely unknown. Despite numerous advances in understanding the genetic and environmental factors affecting the etiology and progression of colorectal cancer, preclinical animal models continue to be indispensable not only for studying the mechanisms of pathogenesis but also for developing novel approaches for the prevention and treatment of colorectal cancer. In recent years, comprehensive reviews have been published that describe these models as well as their advantages and disadvantages in different uses (13–15). In prevention studies, mouse models have previously been used to study the impact of diets, such as so-called Western-style diets, on intestinal tumor development (16). In addition, nonsteroidal anti-inflammatory drugs (NSAID) such as aspirin, sulindac, and naproxen have been tested on the prevention of intestinal cancer in LS and FAP mouse models (17–19). Here we focus on the more recent use of rodent models testing novel interventions for the prevention of colorectal cancer, including immune preventive approaches and strategies involving the manipulation of intestinal microbiota.

Approaches that use immunization are currently a focus of intensive research efforts as they hold promise in the prevention and treatment of colorectal cancer. Vaccination with tumor antigens can enhance infiltration of cytotoxic T-lymphocytes (CTL) into tumors and stimulate their cytotoxic activities against tumor cells expressing specific antigens leading to their elimination and thus preventing tumor formation or recurrence (20, 21). Once the immune system is primed by vaccination, T-memory cells are present for years and can be boosted periodically by further vaccination. Tumor antigens (TAs) are presented by MHC I/II (also known as human leukocyte antigen I and II (HLA I and II)) molecules to T cells to initiate immune responses and act as crucial factors for vaccine formulation (20–22). TAs can be divided into two subtypes: tumor-associated antigens (TAAs), which are overexpressed in tumor cells compared with normal cells, and tumor-specific antigens (TSAs), which are exclusively expressed in tumor cells (20–22). Different approaches have been used to administer TA vaccines, including the use of synthetic peptides, recombinant proteins (23, 24), whole tumor cells (25, 26), tumor antigen expression from viral vectors or bacteria (27, 28), nucleic acids (21, 22), dendritic cells (29, 30), and exosomes (31). Here we will focus on studies using TA vaccines that include peptide or full-length protein vaccines and describe examples in mouse models that used colorectal cancer TAs with frameshift-derived peptides (FSP) as well as TAs specifically expressed in colorectal cancers including carcinoembryonic antigen (32), mucin 1 (MUC1), squamous cell carcinoma antigen recognized by T-cell 3 (SART3), and KRAS (Table 1).

TAA-based vaccination: The use of FSP antigens (or neoantigens) is a new promising immune-preventive approach against cancers with microsatellite instability (MSI; refs. 17, 33, 34). FSPs are TSAs that result from frameshift mutations that occur at an increased frequency at coding microsatellites associated with MMR-deficient (MMR-d) tumors including the hereditary form Lynch syndrome (LS). Recent studies identified immunogenic MSI-associated neoantigens in MMR-d colorectal cancers (23, 33) that promoted the proliferation of cytotoxic T cells to mediate anticancer immune responses (35). A recent study by Gebert and colleagues provided evidence of the tumor-preventive potential of FSP-based vaccines in the VCMsh2 mouse model of LS (17). The authors tested and validated the immunogenicity of four candidate FSPs, Nacad, Maz, Senp6, and Xirp1, that are shared by the majority of MSI-positive intestinal tumors in VCMsh2 mice. Prevaccination with these FSPs reduced the tumor load and increased survival of the mice, suggesting that vaccination with shared FSP neoantigens may be a suitable approach for cancer prevention in LS (17). These studies represent important steps toward the development of neoantigen-based vaccines for the prevention of MSI-positive cancers in humans.

KRAS oncogene mutations, and specifically KRAS (G12D) and KRAS (G12V), are found in a significant proportion of colorectal cancers (20). Recent studies showed that KRAS mutant peptides stimulate CTLs in colorectal cancers. Sakamoto and colleagues showed that administration of the KRAS (G12D)-inhibitory bicyclic peptide KS-58 in mice with CT26 colorectal cancer tumor cell transplants expressing KRASG12D inhibited tumor cell proliferation and suppressed tumor growth (36, 37).

TAA-based vaccination: In addition, TAAs overexpressed in human colorectal cancer have been tested in mouse models. One example is CEA (carcinoembryonic antigen), a nonspecific serum biomarker that is elevated in various malignancies including colorectal cancers and other cancers. CEA is used to monitor colorectal cancer metastasis in human patients and transgenic mouse models. Vaccination of C57BL/6J-CEA-A2Kb transgenic mice (expressing both CEA and HLA-A2) with a combination of CEA peptides and the anti-idiotype antibody 3H1 (mimicking a specific epitope on CEA) increased the proliferation of splenocytes and secretion of Th1 cytokines by CD4⁺ T cells and induced antitumor immunity against MC-38-CEA-A2Kb colon carcinoma cell transplants (38, 39). A CEA-based vaccine also significantly reduced tumor multiplicity in Apc^Min/CEA transgenic mice and significantly improved overall survival (OS) compared with Apc^Min mice without causing autoimmune complications (40). These data in mice suggest that immunization with CEA-based vaccines could be useful for colorectal cancer prevention in humans.

MUC-1 (CD-227) is an epithelial membrane glycoprotein that is expressed in ∼70% of colorectal cancers and correlates with poor prognosis (41). Immunization with a MUC1-specific TAA vaccine prevented the growth of MUC1 overexpressing MC38 tumor cell xenografts in MUC1 transgenic mice. In these studies, MUC1-specific vaccination stimulated interferon-γ producing CD4⁺ Th1 cells and CD8⁺ CTLs and generated strong immune responses against MUC1-expressing cancer cells without damaging normal tissues. It also induced a strong humoral memory response (42, 43). Another study in the MUC1⁺IL10^−/− mouse model of inflammatory bowel disease-associated colorectal cancer (IBD-colorectal cancer) that overexpresses abnormal hypoglycosylated MUC1 showed that vaccination against MUC1 prevented dysplasia, IBD, and development of colitis-associated colon cancer by increasing mucosal immunity (44). Increasing evidence further suggests that MUC1 participates in the maintenance, tumorigenicity, and metastasis of colorectal cancer stem cells (CCSC). The vaccination of BALB/c mice with CD133⁺ CCSCs expressing MUC1 significantly reduced the growth of CT26 colorectal cancer transplants by the targeted killing of CCSCs indicated by the decrease in cells expressing CD133 and ALDH stem cell markers in these tumors. It also increased NK cell cytotoxicity and promoted the release of IFNγ, Perforin, and Granzyme B (45). Importantly, MUC1 peptide vaccination significantly boosted the T-cell immune response in human patients with a history of advanced colon adenoma and elicited long-term memory important for cancer prevention (46).

Another example of a TAA that has been tested in mouse models is squamous cell carcinoma recognized by T-cell-3 (SART3). SART3 is overexpressed in ∼70% of colorectal cancers and has multiple roles in gene regulation (47), cancer immunology (48, 49), and the maintenance of stem cell pluripotency (50). Vaccination with SART3-derived peptides was shown to induce cytotoxic T-cell antitumor immune responses (51) in human colorectal cancer patients by binding to various MHC haplotypes (20, 49, 52). Vaccination of BALB/c mice with a SART3 DNA expression vector in combination with CD40 L and GM-CSF expression vectors (to stimulate DC responses and enhance vaccination) on a polyplex micelle platform stimulated CTL and NK activities and inhibited the growth and metastasis of CT26 colorectal cancer xenotransplants (53).

Survivin, an inhibitor of apoptosis protein (IAP), is overexpressed in many cancers, including colorectal cancers, and its expression shows a strong correlation with TP53 mutation (54). Vaccination with a long peptide vaccine (SurVaxM) that spans amino acids 53 through 67 of the human survivin protein sequence led to survivin-specific CD8-mediated tumor cell death and increased the OS of C57BL6 mice implanted with murine GL261 glioma cells or B16f1 melanoma cells (55). Vaccination with a long synthetic peptide (LSP)-based cancer vaccine targeting amino acids 17–34, 84–110, and 122–142 of the survivin coding region also significantly upregulated tumor infiltration of CD4⁺ and CD8⁺ T cells and inhibited the growth of CT26 colorectal cancer xenotransplants in BALB/c mice and prevented tumor relapse (56). These results suggest that targeting survivin could be another approach for immunization against colorectal cancer.

Recent technological advances have made it possible to define the taxonomic profiles of the human gut bacteria in great detail and perform investigations into their roles in the etiology and development of colorectal cancer. Large-scale population studies have defined “healthy microbiota” and unhealthy colorectal cancer–associated microbiota profiles and manipulation of the gut microbiota has been evaluated in cancer prevention and treatment (57, 58). Colorectal cancer was shown to be associated with alterations in microbiota composition (dysbiosis) and specific bacteria species were found to play causal roles in the pathogenesis of colorectal cancer (59, 60). Bacteria can directly promote colorectal cancer via toxin-mediated mutagenesis, or indirectly through the opportunistic outgrowth of bacteria in the tumor microenvironment affecting the integrity of intestinal epithelia and the replication of intestinal cells. For example, Escherichia coli can secrete genotoxic DNA-damaging agents (61), whereas Fusobacterium nucleatum alters Toll-like receptor signaling and thus inhibit apoptosis (62). Other procarcinogenic effects are exerted by bacteria secreting metabolites that support the growth of cancer cells or interfere with the capability of the host immune system to attack cancer cells (63–65). In addition, recent studies have shown that the efficacy of both chemotherapy and immunotherapy depends on the gut microbiota, and approaches were developed and tested to achieve optimal treatment responses in patients with compromised gut microbiota (66).

Current interventional strategies explore the modulation of the gut bacteria to restore the altered microbiota equilibria associated with higher colorectal cancer risk. These include the use of probiotics, prebiotics, or postbiotics, the elimination of specific pathogens with antibiotics and altering pathogenic microbiota via fecal microbiota transplantation (FMT; Table 2). Each of these approaches showed encouraging results in preventing colorectal cancer in animal models, some of which are outlined below.

Probiotics are live microorganisms that can be administered in specific amounts to provide a health benefit (67). They are widely used as food supplements, and their functions span from directly affecting the microbiota to modulating inflammation and thus reducing colorectal cancer risk. Indeed, feeding the probiotic Lactobacillus salivarius reduced enterocolitis and intestinal tumor occurrence in the C57BL/10J-IL10 mouse model of colitis (68). Importantly, probiotics can help eliminate Clostridium difficile infections (CDI) that are frequently enriched after the use of antibiotics and are associated with inflammation and colon tumorigenesis (69). For example, in C57BL/6 mice with CDI, the administration of Akkermansia muciniphilia had a protective effect by limiting CDI-associated intestinal barrier damage and inflammation (70). The outgrowth of Clostridium difficile could also be prevented in C57BL/6 mice with CDI by reducing feed succinate, a metabolite that fuels the growth of Clostridium difficile. This was achieved by the colonization of Phascolarctobacterium, a probiotic bacteria consuming succinate (71). Recently, it was shown that Streptococcus thermophilus reduces intestinal cancer development by inducing β-galactosidase and by inactivating the oncogenic Hippo pathway in both Apc^Min and azoxymethane (AOM)-injected mice (72). Interestingly, a cocktail of Lactobacillus and Bifidobacteria was shown to reduce colorectal cancer incidence in dimethylhydrazine (DMH)-treated Sprague–Dawley rats by enhancing Toll-like receptor 2 signaling, which improved epithelial barrier integrity in the colonic mucosa and reduced the incidence of aberrant crypt foci (73). Other probiotic mixtures of bacteria have inhibited intestinal tumorigenesis in chemically induced or in Apc^Min mice (74, 75). In addition, genetic modification of Lactobacillus acidophilus that eliminated its ability to secrete lipoteichoic acid reduced the expression of proinflammatory molecules and the incidence of colon adenomas in TS4-Cre/Apc^lox468 mice with colon-specific Apc deletion (76).

Prebiotics are defined as nondigestible food ingredients and belong mainly to a limited number of carbohydrate compounds that include fructo-oligosaccharides (FOS), inulin, galacto-oligosaccharides (GOS), and lactulose (77). They represent an alternate means to stimulate the growth of beneficial bacteria and their production of metabolites through fermentation in the colon. In addition, they can exert direct effects on the intestinal mucosa, modulate the immune system (78) and prevent infectious colonization (79). Recent studies indicate a protective effect of certain prebiotics (xylooligosaccharides/fructooligosaccharides) against colorectal cancer development in AOM-induced Sprague–Dawley rats (80). Another study showed the beneficial effects of prebiotics in reducing adenoma incidence in Apc^Min mice through the interaction between prebiotic triterpenoid saponins and Bifidobacterium (81). Studies in DMH-induced Fisher 344 or Sprague–Dawley rats reported that the addition of prebiotics, such as dietary galacto-oligosaccharides (82) and resistant starch (83), reduced the numbers of aberrant crypt foci. Inulin is a galacto-oligosaccharide for which conflicting results have been reported. For example, although the addition of inulin to the food protected Fisher 344 rats against AOM-induced colorectal cancer (84), another in vitro study in Caco-2 colorectal cancer cells showed a protumorigenic effect as inulin increased the production of genotoxic colibactin by pks+ Escherichia coli in Caco-2 colorectal cancer cells (85). Based on these and other animal studies, ongoing trials in large human cohorts are exploring the potential preventive effects of prebiotics, but conclusive results are not yet available (86, 87).

Postbiotics are products of bacterial fermentation including short-chain fatty acids (SCFA), extracellular polysaccharides, and others. In colorectal cancer, the production of SCFAs such as butyrate, acetate, and propionate appears to have protective effects (88). Butyrate has been extensively studied due to the so-called butyrate paradox, which lies in its ability to safely promote healthy colonocyte proliferation in the mucosa, but at the same time to induce apoptosis in colorectal cancer cells. This effect was initially demonstrated in vitro by the ability of butyrate to inactivate multiple oncogenic pathways through histone hyperacetylation in HT-29 and HCT-116 colorectal cancer cell lines (89, 90). Subsequent in vivo studies in Apc^Min mice fed a high-fat diet reported that butyrate, produced by Clostridium butyricum, protected against intestinal tumor formation by suppressing Wnt/β-catenin signaling, which decreased cell proliferation and increased apoptosis (91). Another study in TS4Cre/APC^lox468 conditional knockout mice showed that food enriched in dietary fiber promoted the growth of butyrate-producing bacteria and reduced polyp numbers (92). The role of butyrate in modulating pro- and anti-inflammatory cytokines and colonic Treg cell differentiation was observed in multiple murine studies (93, 94) and provided the rationale for the initiation of several clinical trials to test its ability to ameliorate ulcerative colitis, an inflammatory disease with increased risk to colorectal cancer (95, 96). Interestingly, a study using MMR-d Msh2^−/−/Apc^Min/+ mice demonstrated that microbial-derived butyrate can support rather than inhibit colorectal cancer cell proliferation (97), which raises questions as to whether somatic or germline mutations of the host affect the impact of postbiotics on the prevention of colorectal cancer.

Antibiotics are helpful in preventing colorectal cancer when used selectively, such as in the elimination of Fusobacterium nucleatum, a pathogen highly enriched at the adenoma stage in the colon (98, 99). Metronidazole treatment of mice transplanted with human Fusobacterium-positive colorectal cancer xenografts lowered the load of Fusobacterium and reduced cancer cell proliferation and tumor growth (100). In another example, the administration of cefoxitin (a broad-spectrum cephalosporin) to Apc^Δ716 mice led to a complete clearance of previously inoculated enterotoxigenic Bacteroides fragilis, known to be associated with IBD-colorectal cancers (101) and reduced formation of intestinal microadenomas (102). Other studies used antibiotic cocktails rather than specific antibiotics, to assess their ability to prevent IBD-colorectal cancer tumorigenesis. Studies in AOM/dextran sodium sulfate (DSS)-treated BALB/c (103) and C57BL/6 mice (104, 105) showed that different combinations of antibiotics could suppress colitis-associated colorectal cancer.

Fecal microbiota transplantation (FMT) uses the transplantation of fecal matter from healthy donors to patients with intestinal pathologies (106). Several studies explored the effect of FMT from human donors on colorectal cancer development in mouse models. For example, one study showed that FMT of human stool from colorectal cancer survivors that consumed rice bran daily protected C57BL/6 mice from AOM/DSS-induced colon carcinogenesis (107). In another example of FMT in colorectal cancer prevention, FMT from volunteers who regularly used the YYFZBJS herbal medicine protected Apc^Min mice from intestinal tumorigenesis (108). Similarly, restoration of a “healthy microbiota composition” in C57BL/6 mice by FMT with microbiota from wild mice (Mus musculus domesticus) reduced inflammation and improved resistance against AOM/DSS-induced colorectal tumorigenesis (109).

These and many other studies in rodent models suggest that approaches involving prevaccination with tumor antigens or manipulation of intestinal microbiota may hold promise for the prevention of colorectal cancer in patients with familial syndromes or in patients at increased risk, such as those with inflammatory bowel disease. However, each of these approaches still need to be carefully tested in human patients to alleviate safety concerns and assess efficacy. For vaccination approaches, initial clinical trials found that three MMR-d FSP neoantigens (AIM2, HT001, and TAF1B, exhibiting one-base pair deletions at their respective coding microsatellites) are well tolerated and induced T-cell and humoral immune responses (110). The FSP-based vaccine Nous209 (encoding 209 different FSP neoantigens shared by both sporadic and hereditary MSI-positive tumors) showed broad induction of FSP-specific CD8 and CD4 T-cell responses in C56F1 mice (111). Nous209 is currently being tested in early clinical trials for the treatment of MSI-H solid tumors in combination with pembrolizumab and cancer prevention in Lynch syndrome carriers. Vaccination against TSAs such as KRAS (G12D) or mutations in other tumor driver genes that commonly occur might also be useful in colorectal cancer prevention. For example, substitution at R248Q residue is the most frequent TP53 alteration in human colorectal cancer. A recent study in humanized TP53R248Q^floxmice showed that conditional inactivation of the mutp53R248Q allele strongly reduced AOM/DSS-induced colon carcinogenesis, which suggests that vaccination against the TP53R248Q may protect against colorectal cancers expressing this TSA (112), which could be tested in mouse models.

Although studies in rodents yielded promising results suggesting the beneficial effects of probiotics, prebiotics, and postbiotics, concerns remain about their safety in the clinical setting, especially in patients with underlying medical conditions such as compromised immunity (113). Issues regarding optimal strain usage, dosage, and formulation will need to be addressed. In this regard, the beneficial use of antibiotics in colorectal cancer prevention has to be balanced with the potential risks it poses as meta-analyses in human patients found conflicting associations with both delayed and accelerated colon tumorigenesis (114). Although these studies do not directly prove a causal role for antibiotics in colon carcinogenesis, they highlight important questions regarding the safe use of these medications, as they can also induce the colonization of colorectal cancer–associated pathogens (115). Not surprisingly, the use of FMT also raises concerns as it may have the potential of transmitting multidrug-resistant bacteria and viruses or causing other host infections (116). Therefore, the manipulation of microbiota in the clinical setting is still under scientific scrutiny, and providing clinical benefit might be highly dependent on several factors, including the microbiota diversity and susceptibility to specific gut microbiota, preexisting genetic alterations, and individual immune responses in patients. It is likely that colorectal cancer prevention approaches involving the gut microbiota need to be tailored around individual patient needs, and animal models will continue to provide valuable tools to test these parameters in a controlled manner.

No disclosures were reported.

This work was supported by NIH grants (CA248536 and CA222358 and Cancer Center grant CA13330) and a Feinberg Family Donation to W. Edelmann.