Engineered gut bacteria fused to T cells shrink colorectal tumors in mice

A newly published mouse study connects two things gut-health readers already think about separately: engineering gut bacteria for a purpose, and the immune system's T cells doing the work of fighting disease. Researchers in China physically bonded an engineered probiotic strain onto the surface of T cells, creating a living hybrid cell that shrank colorectal tumors far more effectively than either the bacteria or the T cells could on their own. The work, published in Science Bulletin, is a mouse and cell-culture study only. It has not been tested in a single human patient, but it offers a genuinely new mechanical idea for a cancer treatment approach that has struggled outside of blood cancers for years.

Key takeaways

• Researchers chemically bonded an engineered probiotic (Lactococcus lactis) directly onto the surface of T cells, creating a hybrid called T-FOLactis, and used it to treat colorectal tumors in mice.
• T-FOLactis cut tumor burden by 78.1% compared with untreated controls and by 65.0% compared with standard T-cell therapy, while roughly doubling median survival from 22 to 46 days.
• The effect depended heavily on a signaling molecule called IL-18; blocking it reversed most of the benefit, pointing to the biological mechanism at work.
• This is a small, mouse-only study (groups of 3 to 6 animals) with no human trial data yet; it has not been reported by any mainstream outlet.

Probiotic-T Cell Chimera 'T-FOLactis' Cuts Tumor Burden by 78% in Mice

The core finding is that fusing an engineered probiotic onto the surface of T cells produced a hybrid cell, which the study authors named T-FOLactis, that was substantially better at shrinking tumors than either component delivered separately. T cells are the immune system's cytotoxic (cell-killing) white blood cells, and adoptive cell transfer (ACT) is the general strategy of collecting, boosting, and reinfusing a patient's own T cells to fight cancer. The bacterial half of the hybrid is FOLactis, an engineered strain of Lactococcus lactis (the bacterium used to ferment cheese and yogurt) modified to produce a fusion of two immune-signaling proteins, Flt3L and OX40L, that recruit and activate the dendritic cells (immune cells that "present" tumor targets to T cells) needed to sustain an attack. Rather than mixing the bacteria and T cells together in a syringe, the researchers used a chemistry technique called bio-orthogonal click chemistry to covalently weld FOLactis directly onto the T-cell surface, so the two travel and act as a single unit. In mice with colorectal tumors, this combined cell reduced final tumor weight by 78.1% compared with saline-treated animals and by 65.0% compared with mice given unmodified T cells, while extending median survival from 22 to 46 days, without causing detectable weight loss or organ damage.

Why Solid Tumors Are the Hard Case for Cell Therapy

Adoptive T-cell therapies, including the CAR-T treatments now approved for leukemia and lymphoma, have achieved durable remissions in blood cancers but have repeatedly underperformed against solid tumors such as colorectal, lung, and pancreatic cancer. The problem is not that the engineered T cells cannot kill cancer cells in a dish; it is that solid tumors are hard for T cells to reach, hard to survive inside once they arrive, and actively suppress the immune attack once T cells do infiltrate. This gap matters most to anyone following colorectal cancer research, since solid-tumor immunotherapy has lagged years behind the blood-cancer successes that get most of the headlines. A related engineered-probiotic approach, in which tumor-colonizing bacteria "tag" solid tumors so CAR-T cells can find them, was reported in Science in 2023, reflecting a broader research push toward pairing bacteria with T cells to solve exactly this delivery problem. Reviews of adoptive cell therapy in solid tumors consistently identify poor tumor infiltration, limited T-cell persistence, and an immune-suppressing tumor microenvironment as the three central obstacles the new study is aimed at.

Mouse and Human-Cell Experiments Behind the Numbers

The study's central experiments were done in two mouse models of colorectal cancer (strains called CT26 and MC38), using small groups of five mice per treatment arm, with additional cell-based experiments run in triplicate or in groups of three to six samples. Researchers also confirmed, in the lab dish rather than in a living animal, that the same conjugation trick worked with primary human T cells: human T-FOLactis showed the same pattern of increased activation markers and improved killing of human colorectal and gastric cancer cell lines. That human-cell result strengthens the case that the biology is not a mouse-only quirk, but it is still a test-tube experiment, not evidence the approach works in an actual human body with an intact immune system and a real tumor.

What This Could Mean, and What It Does Not Mean Yet

For anyone managing colorectal cancer or following immunotherapy research because a family member has a solid tumor, this study is a mechanistic advance worth knowing about, not a treatment to ask a doctor about yet. Nothing here is an approved therapy, an available clinical trial, or a probiotic supplement; there is no over-the-counter equivalent of eating more Lactococcus lactis that recreates this effect, since the bacteria used were genetically engineered and chemically bonded onto lab-grown immune cells under sterile conditions. What is genuinely useful to track going forward is the underlying idea: that combining a tumor-homing microbe with a cytotoxic immune cell, rather than choosing one strategy over the other, might address the infiltration and persistence problems that have stalled solid-tumor cell therapy. That is a research direction to watch, not a claim about outcomes in people.

Small Sample Sizes and No Human Testing Are the Main Caveats

The most important limitation is scale: the in vivo therapy experiments used groups of five mice, and most supporting lab experiments used three to six biological replicates, which is standard for early preclinical work but far too small to estimate how consistent or generalizable the effect is. The entire therapeutic result comes from two mouse tumor models; no human patient has received T-FOLactis, and the human-cell work was confined to cells in culture. The authors are also candid that manufacturing this kind of living cell-bacteria hybrid at the quality and safety standard required for human use, sometimes called GMP-compatible manufacturing, remains a substantial unsolved problem, since it involves controlling bacterial load per cell, sterility, and product stability in a way that ordinary drug manufacturing does not. The study reports no conflicts of interest and was funded by Chinese government science foundations, not corporate sponsors, which reduces one common source of bias.

What Comes Next for This Cell-Bacteria Platform

The researchers frame this as a proof-of-concept for a broader engineering strategy, and the immediate next steps they describe are scientific rather than clinical: mapping exactly which cells produce the IL-18 signal that drives the effect, and testing the platform in more complex model systems, including patient-derived tumor models, that better mimic real human solid tumors before any thought of a clinical trial. No timeline for human testing has been announced, and the paper itself does not propose one; the honest read is that this is early-stage academic research, not a therapy in the pipeline.

Common Questions

Is T-FOLactis a probiotic I could take or buy? No. It is a genetically engineered bacterial strain chemically welded onto lab-grown T cells under sterile conditions, not a supplement, food, or anything a person could consume to replicate the effect.

Does this study show the treatment works in humans? No. All of the tumor-shrinking and survival data come from mice; the only human-related experiments were done on isolated human cells in a lab dish, not in a person with cancer.

What is IL-18 and why does it matter here? IL-18 is a signaling protein (cytokine) that helps immune cells called dendritic cells and T cells coordinate an attack on a target. When researchers blocked IL-18, T-FOLactis lost most of its cancer-fighting benefit, which is how they identified it as central to the mechanism.

Why does this matter if it is only in mice? Solid-tumor cell therapy has a persistent infiltration and persistence problem, and this study offers a specific engineering fix, worth tracking as later, larger studies test whether it holds up.

References

1. Jingwen Shen, Yanhong Chu, Yan Wang, et al. (2026). A bio-orthogonally engineered probiotic-T cell chimera amplifies adoptive cell therapy through spatial DC-T cell immune licensing. Science Bulletin. https://doi.org/10.1016/j.scib.2026.09.009
2. Rosa L. Vincent, Candice R. Gurbatri, Fangda Li, et al. (2023). Probiotic-guided CAR-T cells for solid tumor targeting. Science. https://doi.org/10.1126/science.add7034
3. Víctor Albarrán, María San Román, Javier Pozas, et al. (2024). Adoptive T cell therapy for solid tumors: current landscape and future challenges. Frontiers in Immunology. https://doi.org/10.3389/fimmu.2024.1352805