A Clumping Protein Lets Common Oral Bacteria Shift From Harmless to Dangerous

June 29, 2021

The oral bacterium Fusobacterium nucleatum is widespread and, though usually harmless, has been linked to a variety of health problems, including cancer, periodontal disease, and pregnancy complications. New research from Columbia University suggests that an amyloid-like form of the protein FadA is a crucial player in F. nucleatum’s transformation from innocuous to pathogenic.

The study, published June 29 in EMBO Reports, builds on more than a decade of research into FadA, a small, filamentous protein produced by oral Fusobacteria to help them bind to their host’s cells. FadA comes in two basic flavors: a longer version called pre-FadA and a shorter version called mFadA. The two FadA configurations also form a complex, FadAc, that was found in earlier studies to advance colorectal cancer.

“We were so puzzled by this protein,” says Yiping Han, PhD, the paper’s lead author and a professor at Columbia University’s College of Dental Medicine and Vagelos College of Physicians and Surgeons. Han’s team was able to determine the structure of mFadA, but pre-FadA and the complex proved difficult to solve because of their heterogeneous fibers.

Around that time, says Han, a new faculty member moved into same floor as her research lab. That faculty member was Luke Berchowitz, PhD, an assistant professor of genetics and development, whose research centers on fibrous protein clumps called amyloids. “I showed him our FadA pictures, and he said, ‘That looks like amyloid.’”

The insight sparked a collaboration between their labs and led to the current work characterizing FadA and its role in F. nucleatum’s pathogenicity.

FadA shares properties of amyloid proteins

Fusobacteria producing pre-FadA were coated with amyloid fibers (noted with arrows) in the stationary phase (top), whereas Fusobacteria lacking pre-FadA had a smooth surface (bottom). Image adapted from Meng et al. (2021).

Pursuing this lead, Han and her colleagues tested other properties of FadA and found that pre-FadA and FadAc share several characteristics of amyloids. First, they bind to dyes known to stain amyloid protein. Second, their structure is unaffected by detergents that damage most proteins. Finally, they are recognized by anti-amyloid antibodies, which react to the amyloid fibril structure.

F. nucleatum strains that lacked the gene for FadA or secreted only mFadA did not have these amyloid-like properties and, when examined under the microscope, did not have the fibrous coating seen with the standard strain. These findings allowed the researchers to establish FadA, when pre-FadA was present, as “amyloid-like.”

A bacterial stress response that makes some diseases worse

The team examined F. nucleatum cultures to see when FadA aggregates develop over the bacterial life cycle. During the growth phase, amyloid-like FadA was largely absent, but in the stationary phase that followed, the researchers detected substantial amounts of amyloid-like FadA.

In patients with periodontal disease, immunohistochemical staining detected amyloid aggregates in plaque but not in healthy tissue. Image adapted from Meng et al. (2021).

“That means it responds to stress conditions, because in the stationary phase you have nutrient deprivation or other stress signals,” explains Han.

In the body, such signals might be triggered by disease or the resulting interaction with immune cells. When the researchers looked at human samples, they found FadA aggregates in colorectal tumors but not in healthy tissues from the same patients. Paired diseased and healthy tissues from patients with periodontal disease—when infection damages the gums and bone supporting the teeth—told the same story.

Experiments in mouse models offered more evidence of amyloid-like FadA’s role in periodontal disease and colorectal cancer. In mice with periodontal disease, adding F. nucleatum strains that secreted amyloid-like FadA resulted in more bone loss; in mice with colorectal cancer xenographs, these strains led to more aggressive tumor growth.

“This has big significance,” says Han, “because Fusobacteria nucleatum exists in everybody’s oral cavity, but not everybody is having [health issues]. So we wondered, how does this common commensal turn into a rampant pathogen? We think we are just starting to find the answer, which could be that this amyloid-like FadA serves as a molecular switch: When F. nucleatum receives certain signals, it secretes amyloid FadA and these bacteria become pathogenic.”

F. nucleatum gears up for growth and migration

Han and her colleagues observed several ways in which F. nucleatum received a boost when FadA aggregates were present. For one, amyloid-like FadA shielded bacteria from acidic conditions that are generally fatal to F. nucleatum.

Amyloid-like FadA also supported the growth of biofilm, a community of microbes and substances they release. Biofilm protects bacteria, helps them stick to the tooth surface and cells, and allows them to communicate with each other and share nutrients. The protein aggregates could enhance biofilm formation by providing a scaffold, says Han.

 Illustrated summary showing how F. nucleatum (Fn) turns pathogenic and contributes to disease. Image adapted from Meng et al. (2021).

The acid tolerance and biofilm support help explain how Fusobacteria in the mouth can survive passage through the GI tract to take up residence in the colon, where they can exacerbate cancer. Likewise, the FadA-mediated changes provide a mechanistic link between periodontal disease and colorectal cancer, which may co-occur.

There is still much to learn about FadA, and Han’s lab will focus next on determining the structure of FadA fibers and investigating the relationship between FadA and the protein that secretes it. A better understanding of FadA and how it leads F. nucleatum to advance disease will open the door to effective, targeted interventions.


Cancer, Colon Cancer, Research


microbiology, periodontal disease


The study is titled “Fusobacterium nucleatum secretes amyloid-like FadA to enhance pathogenicity.”

Yiping Han is a professor of microbial sciences at the Columbia University College of Dental Medicine and of microbiology and immunology at Columbia University Vagelos College of Physicians and Surgeons.

Luke Berchowitz is an assistant professor of genetics and development at Columbia University Vagelos College of Physicians and Surgeons.

All authors (from Columbia University unless otherwise noted): Qing Meng, Qiuqiang Gao, Shebli Mehrazarin, Kamonchanok Tangwanichgapong (now at University, Khon Kaen, Thailand), Yu Wang (now at University of Pennsylvania, Philadelphia, USA), Yiming Huang, Yutong Pan, Samuel Robinson, Ziwen Liu, Amirali Zangiabadi, Renate Lux (University of California, Los Angeles, Los Angeles, USA), Panos N. Papapanou, X. Edward Guo, Harris Wang, Luke E. Berchowitz, and Yiping W. Han.

The research was supported by the U.S. National Institutes of Health (grants R01CA192111, R01DE029532, R35GM124633-01, R01AI132403, R01AR065564, and 1S10OD025102), Burroughs Wellcome Fund (PATH1016691), Irma T. Hirschl Career Scientist Award, and China Scholarship Council.

Disclosures: Yiping Han is a scientific advisor to Emanate Biomedical. Harris Wang is a scientific advisor to and equity holder of SNIPR Biome and Imvela.