Oral Bacteria as Drivers for Colorectal Cancer

Colorectal cancer (CRC) is the second leading cause of cancer death in males and females combined in the United States, affecting one in every 20 individuals.¹ Although it is considered a “Western” disease, the incidence is increasing worldwide. CRC has long been recognized as a genetic disease and follows an “adenoma-carcinoma” model, developing as mutations accumulate.² Recent publications have provided evidence that Fusobacterium nucleatum, one of the predominant subgingival microbial species found in chronic periodontitis, is causally implicated in CRC.^3,4 This commentary provides some perspective on the biologic mechanisms that appear to be involved in the role that F. nucleatum may play, together with the genetic mutations that lead to CRC.

The genes most commonly mutated in CRC include adenomatous polyposis coli (APC), β-catenin, P53, Kirsten rat sarcoma oncogene, and myelocytomatosis oncogene (MYC).⁵ These are driver mutations associated with several cancer hallmarks:1) uncontrolled cell growth and replication; 2) angiogenesis, tissue invasion, metastasis and resistance to apoptosis; 3) reprogramming of energy metabolism; 4) evading immune destruction; and 5) inflammation.⁶

The advancements in microbial identification and human microbiome studies have revolutionized our view of the microorganisms associated with disease, including CRC. Using various “omics” approaches, a number of studies consistently identified F. nucleatum, a Gram-negative anaerobe, to be highly enriched in colorectal carcinomas.^7–10 A subsequent study reported that F. nucleatum was not only enriched in CRC but also in benign precancerous polyps.¹¹

F. nucleatum is found in high numbers subgingivally implicated in periodontal disease.^12,13 In creasing evidence suggests that oral bacteria are not confined to the oral cavity and can migrate to extraoral sites, causing infection and inflammation.¹⁴ F. nucleatum has been isolated from a wide range of organ abscesses and infections, although it is never or rarely detected in those floras under normal conditions. In particular, this organism is capable of crossing the placental barrier, causing pregnancy complications such as preterm birth, stillbirth, and neonatal sepsis.¹⁴

Detection of F. nucleatum in the colorectal adenomas and carcinomas does not prove causality. The oral cavity is at the beginning of the digestive tract; it is not surprising that microbial species find their way down this path. Two recent studies by the author’s group⁴ and others³ aimed to address this issue and demonstrate that F. nucleatum is indeed a driver of CRC. The author’s group showed that the unique FadA (Fusobacterium adhesin A) adhesion of F. nucleatum stimulated human CRC cell growth.⁴ FadA mediates F. nucleatum binding to endothelial and epithelial cells,^15–19 and, perhaps most importantly to this discussion, it is only present in limited spaces including F. nucleatum and is absent in non-oral Fusobacterium species.¹⁵ FadA binds vascular endothelial-cadherin on endothelial cells, causing increased endothelial cell permeability thus allowing bacteria to penetrate, a likely mechanism used by F. nucleatum for systemic dissemination.¹⁹ Rubinstein et al.⁴ demonstrated that F. nucleatum binds to and invades both normal and cancerous epithelial cells via FadA binding to epithelial (E)-cadherin. This binding leads to growth stimulation of human CRC cells but not the non-cancerous cells. FadA binding to E-cadherin on CRC cells activates β-catenin-regulated transcription, resulting in increased expression of oncogenes cyclin D1 and c-Myc, Wnt (wingless-related integration site) signaling genes Wnt7a, Wnt7b, and Wnt9a, and inflammatory genes nuclear factor-κ B, interleukin-6 (IL-6), IL-8, and IL-18, all of which are hallmarks of carcinogenesis. The FadA binding site on E-cadherin has been mapped to an 11–amino acid domain. A synthetic peptide corresponding to this domain prevents F. nucleatum from binding and invasion of CRC cells, thus blocking the oncogenic, Wnt, and inflammatory responses/gene expression. It also inhibits F. nucleatum–driven CRC growth both in vitro and in xenograft mice.⁴

Our evidence further indicates that F. nucleatum has increased fadA expression in the carcinomas that correlate with increased tumorigenesis responses.⁴ The fadA gene levels are significantly increased in the adenoma and carcinoma tissues compared with the normal controls.⁴ The increase is stepwise, from normal controls to the precancerous state (including benign polyps and tissues surrounding the benign and malignant polyps), and from the precancerous state to carcinoma, with an average 10-fold increase between each step. FadA transcription in F. nucleatum in the carcinoma tissues is significantly increased compared with that in the normal controls and the precancerous tissues, indicating an increased virulence activity of F. nucleatum in CRC.

In the study by Kostic et al.,³ F. nucleatum was shown to induce tumor multiplicity and selectively recruit tumor-infiltrating myeloid cells to promote tumorigenesis in APC^+/− mice. Unlike other bacteria, F. nucleatum does not induce colitis, enteritis, or inflammation-associated intestinal carcinogenesis; instead, it induces sporadic colorectal tumorigenesis, which is the most common form of CRC in humans. Fusobacterium spp. were found to be enriched in the stools from patients with adenoma and carcinoma.³

Together, these studies demonstrate that F. nucleatum stimulates colorectal carcinogenesis. The discovery of a microbial driver of CRC provides a brand new perspective on the etiology, mechanism, diagnosis, treatment, and prevention of this debilitating disease. The mechanistic studies identified novel diagnostic and therapeutic targets. Because it is unique to F. nucleatum, fadA may be an ideal diagnostic marker for early detection of CRC. Diagnostic criteria may be developed to define healthy, precancerous, and cancerous states according to the fadA gene levels. The inhibitory peptide and/or its derivatives may be used in precision medicine to specifically eradicate F. nucleatum to treat CRC or reduce CRC risk, similar to eradicating Helicobacter pylori to reduce gastric cancer risk.²⁰ Compared with antibiotic therapies, the precision elimination avoids disturbance of the flora. The potential use of FadA in disease diagnosis, treatment, and prevention warrants additional testing.

It will be challenging to prove causation by periodontal disease, especially because it would require impractical numbers of patients to do a periodontitis intervention study to assess the prevention of CRC. Despite these challenges, the work by the author’s group⁴ and others³ has started to draw direct causal links between a specific subgingival microorganism found in high numbers in periodontitis and the cancerous changes that lead to CRC. Elucidation of the role of F. nucleatum in CRC has broadened the scope of oral-systemic connections. F. nucleatum is abundantly present in the oral cavity and increases in the presence of periodontal disease. Incidentally, both periodontal disease and CRC are considered “old people’s diseases,” with their risks increasing with age. Several questions thus arise. Is periodontal disease a risk factor for CRC? Given the connectivity of the digestive tract, could F. nucleatum or other oral bacteria be involved in additional gastrointestinal disorders? Furthermore, based on the “mobility” of F. nucleatum and the omnipresence of cadherins, could this organism be involved in cancers beyond the gastrointestinal tract? Answers to these questions will shed new lights on the role of the oral microbiome in human health.