It is estimated that about 20% of cancer incidence are linked to infectious agents (zur Hausen, 2009; Plummer et al., 2016). In contrast to the numerous known onco-viruses, Helicobacter pylori, the causative agent of gastric cancer is, to date, the only bacteria classified as a carcinogen (Plummer et al., 2016). Recently, the oral bacteria, Fusobacterium nucleatum was shown to accelerate the progression of colon cancer and to confer the growing tumor with protection against attacking immune cell (Kostic et al., 2013; Rubinstein et al., 2013; Gur et al., 2015a,b; Yang et al., 2017).

F. nucleatum is a gram negative oral anaerobe that plays a key role in the development of the dental plaque by physically bridging between early and late oral bacterial colonizers (Kolenbrander and London, 1993). F. nucleatum numbers rise 10,000-fold in the gingival inflammation that precedes periodontal disease (Moore and Moore, 1994; Socransky et al., 1998). F. nucleatum is also frequently isolated (often as pure cultures) from samples collected in preterm births (Hill, 1998; Han, 2011). Most recently, F. nucleatum was found to be enriched in colorectal cancer (Castellarin et al., 2012; Kostic et al., 2012).

Colorectal cancer (CRC) is the second most commonly occurring cancer, and the fourth most common cause of cancer death (Siegel et al., 2012). Thus, new approaches for CRC diagnosis and treatment are required.

It is assumed that transient bacteremia (frequent during periodontal disease) enables the trafficking of oral fusobacteria to CRC. Tumor-induced angiogenesis, increased blood-vessel permeability, hypoxia, and local immunosuppression, are non-specific factors that aid CRC colonization by blood-borne oral fusobacteria (Abed et al., 2016). However, CRC-specific recognition by fusobacteria, is mediated by the fusobacterial Fap2 lectin, that specifically recognizes and binds tumor-displayed D-galactose-β(1-3)-N-acetyl-D-galactosamine (Gal-GalNAc) (Yang and Shamsuddin, 1996; Abed et al., 2016). High Gal-GalNAc levels were also detected in CRC metastases and were correlated with fusobacterial gDNA occurrence in these metastases (Abed et al., 2016), demonstrating the ability of fusobacteria to colonize CRC metastases.

Once in the tumor, fusobacteria can accelerate cancer development by enhancing cellular proliferation (Rubinstein et al., 2013; Chen et al., 2017; Yang et al., 2017), creating a tumor-favorable inflammatory environment (Kostic et al., 2013) and by protecting tumors from killing by NK cells and tumor infiltrating T cells. The latter is mediated through activation of the TIGIT inhibitory receptor, by the fusobacterial Fap2 protein (in a Gal-GalNAc-independent manner) (Gur et al., 2015a). Not surprisingly, high fusobacterial abundance in CRC was correlated with poor disease outcome (Flanagan et al., 2014), suggesting that therapeutic elimination of CRC-fusobacteria should be considered.

Interestingly, due to their Gal-GalNAc—specific homing, F. nucleatum could potentially be used as a platform for specific targeting and elimination of Gal-GalNAc displaying tumors and metastases. In this regards, besides CRC, additional tumors were previously found to display Gal-GalNAc (Springer, 1984; Lin et al., 2011). Here, we therefore re-screened for tumors that display high levels of Gal-GalNAc.

The results above support our hypothesis that in addition to CRC, fusobacteria home-to and colonize additional tumors that display high levels of Gal-GalNAc. As F. nucleatum was shown to accelerate tumor progression (Kostic et al., 2013; Rubinstein et al., 2013; Gur et al., 2015a; Yang et al., 2017), fusobacterial elimination in these tumors might improve treatment outcome.

Numbers of periodontal bacteria, including fusobacteria, greatly increase during periodontal inflammation (Socransky et al., 1998; Hajishengallis et al., 2011). This raise in bacterial numbers together with the frequent bleeding of the gums during periodontitis, increases the probability of hematogenous translocation of oral bacteria to distant tumors. Indeed, as part of the growing interest in the effect of oral health on general health (Pihlstrom et al., 2005; Rautemaa et al., 2007), the relationship between periodontitis and cancer development is of growing interest (Hiraki et al., 2008; Zeng et al., 2016).

As fusobacteria appear to specifically home-to tumors and metastases displaying high amounts of Gal-GalNAc, fusobacteria might be used in the future as a platform for directing treatment (immunological or chemically based) to such cancers. It should be noted that not all of the samples of each type of adenocarcinoma displayed high Gal-GalNAc levels. This implies that a potential fusobacterial-based cancer therapy will have to be personalized to high Gal-GaNAc displaying tumors.

Bacterioncology, tumor-bacterial interactions, is a rapidly developing field. The tumor microbiome was found recently to play an important role in the effectiveness of cancer treatment (Bashiardes et al., 2017). Bladder cancer is routinely treated with the live bacterial tuberculosis vaccine Bacillus Calmette-Guerin (BCG) (Babjuk et al., 2011), and additional bacterial species are being explored for future cancer treatment (Quispe-Tintaya et al., 2013; Zheng et al., 2017).

It is interesting to note that the fusobacterial Fap2 surface protein, that mediates fusobacterial attachment to tumor-displayed Gal-GalNAc, also endows tumor protection by fusobacteria by activating the immune cells - suppressing TIGIT receptor. Both lectin and immunosuppression functions appear to be on different Fap2 epitopes. This is deduced from the fact that while tumor binding is inhibited by GalNAc (Abed et al., 2016), immunosuppression by TIGIT activation, is not (Gur et al., 2015a). It seems to make evolutionary sense to couple both tumor-associated traits on the same virulence factor.

As immunosuppression is undesired in a future fusobacterial-based tumor BacterioImmunotherapy, the challenging task of identifying and inactivation of the Fap2 TIGIT-activating domain remains.

JA designed and carried out experiments, participated in writing the ms; LP, NM, and SC carried out experiments and participated in writing the ms; GB and OM designed experiments, and participated in writing ms.