Growing evidence suggests that certain types of bacteria are capable of causing colorectal cancers, indicating that a sub-set of these cancers could be the result of infectious disease.
But understanding how bacteria interact in the human gut – our microbiome – has been challenging because of the complex microbial mixture of “good” and “bad” bacteria.
Over a decade ago, French scientists discovered a pathway in certain strains of E. coli, a bacterium normally found in 90% of humans, that is “genotoxic” – toxic to DNA – causing tumor formation and colorectal cancer in mice.
While the “colibactin” pathway is not found in all strains of E. coli, colibactin-producing strains are supported as epidemiological risk factors for colorectal cancer in humans.
Until recently, important questions remained about how this pathway actually damages the DNA to cause colorectal cancer, ultimately barring progress towards the development of future treatments.
“We found that certain E. coli produce specific molecules that ‘cross-link’ our DNA, effectively locking it together,” said Jason Crawford, Associate Professor of Chemistry and of Microbial Pathogenesis, and co-corresponding author of the findings, published today in Science.
Through close collaboration between Crawford’s lab at the Yale Chemical Biology Institute and Seth Herzon, Professor of Chemistry, for the first time the scientists were able to crack the identity of the structure of the cross link, establishing how the molecules work – their so called “mode of action” a- nd how they are made.
“Establishing what is driving colorectal cancer at an atomic resolution expands on previous cellular studies over the past decade, ultimately bringing us closer to strategies to effectively treat and eliminate this risk factor,” said Herzon.
The paper was written by co-first authors Mengzhao (Lucy) Xue, graduate student in the Herzon lab, and Chung Sub Kim, a post-doc in the Crawford lab.
There are an estimated 142,820 new cases of colorectal cancer (CRC) annually in the United States, with over 50,000 deaths .
Researchers have established the major risk factors of CRC. American males have 25% higher risk of CRC than females, and African Americans have 20% higher incidence than Caucasians.
Age is also a risk factor, with less than 10% of cases occurring under the age of 50 .
Colorectal carcinomas are classified by etiology as inherited (e.g., hereditary nonpolyposis colorectal cancer due to genetic instability and familial adenomatous polyposis (FAP) coli due to a mutation in the adenomatous polyposis coli gene, APC), inflammatory (e.g., Crohn’s disease and ulcerative colitis), or sporadic .
Over 80% of CRCs are classified as sporadic CRC and these have a poorly defined etiology.
Sporadic tumorigenesis is thought to involve mutations in the APC (5q), DNA hypomethylation, and multiple epigenetic changes, notably in KRAS2 (12p), DCC (18q), and p53 (17p) .
As well as genetic factors, there are environmental factors that increase the risk of CRC. In particular, diet has been associated with CRC.
Previous studies have established that consumption of red and processed meats, highly refined carbohydrates, and alcohol carry an increased risk of CRC .
The gut microbiota has emerged as an environmental promoter of CRC in both animal models and human studies, which is the focus of discussion from here onward.
It is now generally accepted that bacteria are ubiquitous colonizers of the human body, including along the gastrointestinal tract .
These bacterial communities are established at birth, and a lifelong symbiotic relationship forms, with the human providing nutrients.
In return, bacteria are involved in many processes, including tempering immune responses, metabolizing food and byproducts, and preventing pathogenic bacterial diseases.
The concept that normal bacterial microbiota plays a role in the development of inflammation-induced cancer has gained prominence from the considerable colonic microbiota literature.
Microbiome and Colorectal Carcinogenesis
Despite large interpersonal variability, it is known that an average colorectal microbiota includes anaerobic bacteria, including Bacteroides, Eubacterium, Bifidobacterium, Fusobacterium, Peptostreptococcus, and Atopobium [6,7]. Facultative anaerobes include Lactobacilli, Enterococci, Streptococci, and Enterobacteriaceae are also usually present at about 1000-fold lower abundance . However, an individual’s microbiota is influenced by diet, age, gender, and ethnicity, and its dynamic nature makes it difficult to characterize .
From the 1990s onward, studies emerged demonstrating associations between colon cancer and specific bacterial species found in fecal or mucosal tissue samples, including Streptococcus bovis, Bacteroides, and Clostridia [6,8,9].
Chen et al. undertook 16S metagenomic profiling to characterize the microbiota present in the intestinal lumen and mucosa of patients with CRC compared to healthy controls. Bifidobacterium, Faecalibacterium, and Blautia were attenuated in CRC patients, whereas Porphyromonas and Mogibacterium were increased . In the lumen, Erysipelotrichaceae, Prevotellaceae, and Coriobacteriaceae were increased in CRC patients. This suggests that intestinal lumen microbiota may interact with the host to increase CRC risk .
In a study by Shen et al., sequencing 21 adenoma and 23 nonadenoma subjects showed enriched Proteobacteria and reduced Bacteroidetes in cancer tissue . Sobhani et al. analyzed stool bacterial DNA using PCA and found Bacteroides/Prevotella species to be more abundant in cancer patients than in control subjects .
These sequencing studies have demonstrated the occurrence of gut microbiota alterations in CRC.
Fusobacteria, anaerobic gram-negative rods, are rare agents of severe human diseases  that have recently been the center of academic debate after researchers repeatedly noted their link to CRC. F. nucleatum (Fn) and F. necrophorum are the commonly encountered members of the Fusobacterium species.
They commonly inhabit the oral cavity, occasionally causing periodontal and gingival infections . The rest of the chapter explores the evidence for the role of Fn in CRC.
Clinical Significance of Fn in CRC
This has led to fecal Fn measurement being touted as a useful predictive marker in the clinical management of CRC .
Several metagenomics-sequencing studies have shown that increased Fn abundance was positively associated with CRC mortality. For example, one study found CRC patients with high fusobacterial levels had significantly lower overall survival than patients with average levels of Fn (p = 0.008) .
This would suggest that the enrichment of Fn in CRC tissue could serve as a prognostic biomarker [18,20,21]. Studies have also demonstrated a positive association between increased amounts of Fn and CRC metastases, such as hepatic metastatic disease [15,20,22].
These findings indicate that Fn-high CRC maybe be a clinically relevant subtype of CRC that promotes tumor progression. However, several studies did not identify an association between Fn and CRC prognosis [23,24,25]. There are still many gaps in our understanding of the role of Fn in CRC and, hence, Fn as a prognostic marker.
Translocation of Fn from the Oral Cavity
How Fn colonizes the colon from the mouth is not yet known.
Several studies have reported that microbial colonizers of the oral cavity predict the microbial composition of the gut and vice versa [26,27]. This suggests the oral cavity may act as a repository for Fn, which is then swallowed to the gut.
Fn is a major cause of periodontal disease, but epidemiological studies linking periodontal disease and CRC have been inconclusive. Momen-Heravi et al. (2017) described the increased risk of CRC in a large cohort study , while a separate study found no association between the amount of Fn in the oral cavity and CRC .
Meyerson et al. (2017) showed that Fn colonization of CRC is conserved in nonadjacent liver metastases, indicating microbiome similarities between paired primary–metastatic tumors .
This suggests that Fn might form a symbiotic association with the tumor during metastasis via lymphatic or hematogenous pathways.
There is also uncertainty about where Fn is found within the colon. Notably, Fn has been observed principally in proximal colon cancer, with moderately increasing proportions of Fn-high CRC detected on the right side, from rectum to cecum [24,31].
However, this effect was not seen in other experiments .
These observations are complicated by the fact that right colon cancer exhibits mucus-invasive bacterial biofilms, which can display blooms of Fn .
Overall, there are no conclusive data to support the hypothesis that the origin of the tumor-associated fusobacteria is as passengers from the oral cavity rather than residential bacteria in the colon.
More information: Mengzhao Xue et al. Structure elucidation of colibactin and its DNA cross-links, Science (2019) DOI: 10.1126/science.aax2685
Journal information: Science
Provided by Yale University