Probiotic E. coli engineered to manufacture cancer drugs inside tumors

What if instead of flooding the entire body with toxic chemotherapy drugs, you could send a living organism straight into a tumor and have it manufacture the drug right there?

That is the premise behind a study from Shandong University in Qingdao, China, published March 17 in PLOS Biology. The researchers took Escherichia coli Nissle 1917 (EcN) - a probiotic strain that has been used safely in humans for over a century - and genetically engineered it to produce romidepsin (FK228), an FDA-approved anticancer drug. They then injected these modified bacteria into tumor-bearing mice.

The bacteria did what E. coli does best: they colonized. But instead of the gut, they homed in on tumors, establishing themselves in the oxygen-poor, immune-suppressed environment that solid tumors create. Once there, they began producing romidepsin on site.

Building a drug factory from scratch

Romidepsin is not a simple molecule. It is a bicyclic depsipeptide - a complex chemical structure that would normally require sophisticated industrial synthesis. The research team, led by Tianyu Jiang, used genetic and genomic engineering techniques to reconstruct the romidepsin biosynthetic pathway inside the EcN bacteria. They essentially gave a probiotic the genetic instructions to build a pharmaceutical.

The engineered bacteria produced romidepsin both in laboratory cultures (in vitro) and inside living animals (in vivo), under varying conditions. The dual-action mechanism combines two effects: the natural tendency of certain bacteria to colonize tumors, and the localized release of a proven anticancer agent.

Why bacteria target tumors

The concept of bacteria-mediated cancer therapy is not new, but the execution has lagged behind the theory. Certain bacterial strains, including EcN, preferentially accumulate in tumors because solid tumors create conditions that favor bacterial colonization: low oxygen levels, suppressed local immune responses, and abundant nutrients from rapidly dying cells. This natural targeting means the bacteria concentrate where the drug is needed most.

Previous attempts at bacterial cancer therapy have struggled with efficacy. The bacteria reach the tumor, but the therapeutic payloads they carry have often been insufficient. By engineering the bacteria to synthesize a potent, FDA-approved drug rather than carrying a modest therapeutic protein, this study attempts to close that gap.

Mouse model, human distance

The study used a mouse model with breast cancer tumors. The results showed that the engineered EcN colonized tumors and delivered romidepsin effectively in this controlled setting. But the distance between a mouse tumor model and a human cancer patient is vast.

Several critical questions remain unanswered. How would the human immune system respond to injected bacteria, even a probiotic strain, over time? Would the bacteria maintain drug production long enough to be therapeutically useful? What happens to the engineered bacteria after treatment - can they be reliably cleared from the body? The researchers themselves note that future studies identifying adverse outcomes and methods for eliminating the bacteria post-treatment are necessary.

There is also the regulatory challenge. A living, genetically modified organism that produces a drug inside the body is a fundamentally different product from a pill or an injection. The safety and manufacturing standards for such a therapy would need to be developed largely from scratch.

A foundation, not a finish line

The researchers describe their work as establishing "a solid foundation" for engineering bacteria capable of producing small-molecule anticancer drugs for tumor-targeted therapy. That framing is appropriate. This is proof-of-concept work in an animal model, not a clinical therapy.

But the underlying idea - using bacteria as programmable drug factories that self-deliver to tumors - has appeal precisely because it addresses one of oncology's core problems: getting enough drug to the tumor without poisoning the rest of the body. If the approach can be made safe and reliable in humans, it could change how localized cancer treatment works.

For now, the engineered E. coli exists in the laboratory and in mouse models. The path to a patient's bedside, if it gets there, will require years of safety testing, optimization, and regulatory navigation. But the demonstration that a common gut bacterium can be turned into a tumor-colonizing drug manufacturer is a technical achievement that opens a distinct line of research.