Our bodies harbor countless microbes—and so do our tumors, it turns out. Over the past 5 years, researchers have shown cancer tissue contains entire communities of bacteria and fungi. Now, it appears some of the bacteria may be cancer’s accomplices. In a paper in Nature this week, a team led by Susan Bullman of the Fred Hutchinson Cancer Center reports that in oral and colorectal tumors, bacteria live inside cancer cells and boost their production of proteins known to suppress immune responses. The microbial interlopers may set off a chain reaction that prevents the immune system from killing cancerous cells, and they may also help cancer metastasize to other parts of the body.
The study doesn’t entirely clinch the case for a bacterial role in cancer, but it is very suggestive, says Laurence Zitvogel, a tumor immunologist at the Gustave Roussy Institute. “It shows that bacteria in colorectal and oral tumors can actively disturb the immune equilibrium,” she says.
Confirmation that microbes can cause tumors to grow or spread could open up new ways to make cancer treatment more effective, for instance by killing bacteria with antibiotics. And because each type of cancer appears to come with a unique microbiome, researchers are exploring whether microbes could be used as a diagnostic tool to detect cancer early in a blood sample.
Until recently, most cancer researchers believed tumors were sterile, says Ravid Straussman, a cancer researcher at the Weizmann Institute of Science. But about a decade ago, as a postdoc at the Broad Institute, Straussman accidentally discovered that human pancreatic and colorectal cancer cells grown in the lab stopped responding to a cancer drug named gemcitabine when Mycoplasma bacteria were present in the culture. The bacteria, he discovered, “protected” the cells by producing an enzyme that breaks down gemcitabine.
Straussman found he could render gemcitabine ineffective in mice with colon cancer by injecting the animals with other types of bacteria, including an Escherichia coli strain, and that treating them with antibiotics restored the drug’s effectiveness. When he studied 113 human pancreatic cancer samples, he found bacteria that produced the drug-chewing enzymes in 76% of them—raising the question of whether they contributed to drug resistance in human cancers. Straussman and his colleagues are now planning a clinical trial to test whether antibiotics can improve pancreatic cancer treatment.
Soon afterward, Gregory Sepich-Poore, a doctoral student in the lab of microbiome researcher Rob Knight at the University of California, San Diego (UCSD), was hunting for ways to diagnose pancreatic cancers early. He was motivated by his grandmother’s death from the cancer, which is often diagnosed too late for treatment to be effective. Inspired by Straussman’s 2017 paper, Sepich-Poore began to scour the Cancer Genome Atlas, a large DNA database of human cancers, for snippets of genetic material from microbes.
In March 2020, he, Knight, and colleagues reported that microbial RNA and DNA was present in each of the 33 types of cancers they studied, and that each cancer type had a unique microbiome. The team also found those distinct microbial signatures in blood samples from cancer patients. Based on their findings, Sepich-Poore and Knight co-founded San Diego–based Micronoma, a startup that aims to identify early-stage cancer in blood samples—a so-called liquid biopsy.
Later in 2020, Straussman and his colleagues confirmed that many tumors have distinctive populations of microbes and found they mostly reside inside cancer and immune cells, rather than between those cells.
This paper has taken the field a big step forward.
- Ravid Straussman
- Weizmann Institute of Science
Fungi often take up residence in tumors as well. In a study of 17,000 tumors, published in Cell in September, the UCSD and Weizmann groups found fungal species residing in each of 35 cancer types. Again, each cancer type was associated with a distinct combination of species, which could help refine Micronoma’s diagnostic tools. (Straussman now sits on the company’s scientific advisory board.)
The paper reported another striking finding: Certain combinations of fungal species correlated with lower odds of survival in several types of cancers, most strongly in ovarian and breast cancer. In October, another group reported something similar in Cancer Cell: The presence of a particular bacterial signature seemed to hasten death in pancreatic cancer. The probability of surviving 2 years after treatment doubled in patients that did not have the signature. “That’s an eyebrow-raising finding,” says co-author Martin Blaser, a cancer microbiome researcher at Rutgers University, Piscataway, who also sits on Micronoma’s scientific advisory board.
But none of these findings showed just how fungi or bacteria might lead to a worse outcome. Now, Bullman and her colleagues have addressed the question by studying eight tumors removed from patients with oral cancer and 19 others from colorectal cancer patients. Mapping the distribution of the microbes showed they only colonized specific areas of the tumors. These infected regions had high levels of proteins known to suppress cancer-fighting T cells or fuel cancer growth. T cells amassed outside these regions, the researchers found, but few were found inside. (Instead, the regions contained neutrophils—a type of immune cell that fights infections, among other jobs.) “It’s conceivable that the bacteria are somehow causing the T cells to move away from the tumor,” Blaser says.
Using a technique called single cell sequencing, the researchers found bacteria preferentially infect cancer epithelial cells—which line the inside surface of organs—and that only cells in which Fusobacterium and Treponema bacteria were dominant tended to show both immunosuppressive and cancer promoting characteristics.
“This paper fills a critical gap” by showing that bacteria inside cancer cells may alter the cells’ behavior, says George Miller, a cancer doctor and researcher at Trinity Health of New England.
Bullman and her colleagues also co-cultured Fusobacterium species with colon cancer spheroids—small models of human cancers—embedded in a matrix that contained neutrophils, and compared them with bacteria-free spheroids. With the bacteria present, neutrophils tended to move toward the cancer cells, just as they did in the patient tumor samples. And the researchers saw infected cancer cells breaking off the spheroids and migrating, which Bullman thinks may be a sign that they are metastasizing.
Zitvogel says the paper paints a plausible picture of how microbes could hamper the body’s defenses against cancer. Still, the spheroid model “is a reductionist approach,” she cautions; the human body, which has a varied arsenal of immune cells and a diverse and largely beneficial microbiome, may have other mechanisms that keep cancers from metastasizing.
The study was small and only included two types of cancers, Straussman adds, which leaves plenty of work to do. But, “Bullman’s research has shown us how we should be exploring the tumor microbiome,” he says. “This paper has taken the field a big step forward.”