Scientists engineer probiotic bacteria to infiltrate tumors and produce cancer-fighting drugs on-site

Researchers have successfully engineered probiotic bacteria to seek out tumors, penetrate the tumor microenvironment, and produce cancer-fighting compounds directly at the site of disease. In mouse models, the approach reduced tumor growth while causing significantly fewer systemic side effects than standard chemotherapy. The bacteria used are derived from strains already considered safe for human consumption, which gives the research a more direct path toward clinical testing than most experimental cancer therapies.

The core problem with most cancer drugs is that they do not stay where they are needed. A chemotherapy drug delivered intravenously circulates through the entire body. It hits the tumor, but it also hits healthy tissue in the gut, bone marrow, and hair follicles, which is why nausea, immunosuppression, and hair loss are so common. If you could put a drug factory inside the tumor itself and have it produce the therapeutic compound only there, you eliminate much of that collateral damage. That is the logic behind this approach, and the mouse data suggests it works in practice, not just in theory.

Why bacteria can find tumors that drugs cannot

Solid tumors create a distinctive internal environment. They are often hypoxic, meaning they have low oxygen levels due to poor blood vessel organization. They tend to be acidic, with a lower pH than surrounding tissue. And they suppress normal immune activity in their immediate vicinity, which is part of how they evade the body's defenses. These same properties that make tumors hard to treat also make them attractive environments for certain anaerobic bacteria, which thrive in low-oxygen conditions and are not easily cleared by a suppressed immune system.

The research team engineered their bacterial strain to exploit exactly this biology. Once administered, the bacteria preferentially colonize tumor tissue rather than healthy organs. The engineering adds a second layer: genetic circuits that cause the bacteria to produce therapeutic proteins only after they have reached the tumor interior, using the local chemical conditions as a trigger. The bacteria do not start producing the drug during transit through the bloodstream. They activate on arrival.

What the bacteria actually produce inside the tumor

In the published mouse experiments, the engineered bacteria were programmed to produce two types of therapeutic compounds inside the tumor. The first was a nanobody, a small antibody fragment, designed to block a protein that tumors use to suppress T-cell activity. The second was a cytokine, specifically interleukin-2, which stimulates immune cells to attack the tumor. Delivering both of these within the tumor microenvironment produced a localized immune activation response that reduced tumor size in treated mice by an average of 62% compared to untreated controls, according to the published data.

The choice of these two compounds is deliberate and reflects a broader strategy in cancer immunotherapy. Checkpoint inhibitors, the class of drugs that block tumor immune suppression, have been approved for multiple cancer types but cause serious autoimmune side effects in a significant share of patients when given systemically. Interleukin-2, which was one of the first approved cancer immunotherapies in the 1990s, was largely abandoned as a treatment because intravenous doses high enough to be effective also caused severe vascular toxicity. Delivering both compounds locally, in concentrations calibrated to the tumor rather than the whole body, addresses the toxicity problem directly.

The safety case for using probiotic-derived strains

One of the persistent concerns about using bacteria as therapeutic agents is safety. The strain used in this research is derived from Lactobacillus, a genus of bacteria widely used in fermented foods and probiotic supplements, with an established safety record in humans. The researchers also built in a containment mechanism: the engineered bacteria are auxotrophic for a specific amino acid that is not naturally abundant in the human body. Without a supplemental dose of that amino acid, the bacteria cannot replicate indefinitely. This limits their ability to persist or spread beyond the intended site.

The FDA classifies a number of Lactobacillus strains as Generally Recognized as Safe, which gives engineered versions of these bacteria a regulatory starting point that more exotic bacterial candidates would not have. That does not mean the path to human trials is straightforward. An engineered organism carrying therapeutic gene circuits requires a more detailed safety review than an unmodified probiotic strain, and the agency will want extensive data on off-target colonization, genetic stability, and immune response before approving any human study.

What comes next and what this field still needs to prove

The current results are from mouse tumor models, which replicate some but not all features of human cancer. Mouse tumors are typically generated by implanting cancer cell lines under the skin, which produces a different microenvironment than the spontaneous tumors that develop in human patients over years. The next stage of research will need to test the approach in more complex animal models and eventually in non-human primates before a human trial application is realistic.

The research team, based at the University of California San Diego, has stated their intention to file an Investigational New Drug application with the FDA within the next 18 to 24 months. If that application is approved, a Phase I human trial focused primarily on safety and dosing would likely begin with patients who have solid tumors that have not responded to existing treatments. Phase I trials of this type typically enroll between 20 and 40 patients and take 18 months to three years to complete.