Non-proliferative E. coli therapeutic platform that inhibits in vivo proliferation through genome fragmentation and secretes AI-designed peptides

Background
Live Biotherapeutic Products (LBPs), which use living microorganisms as therapeutic agents, have emerged as a new pillar of precision medicine through integration with synthetic biology. Probiotic Escherichia coli Nissle 1917 (EcN), a representative delivery vehicle, has been a key player due to its proven intestinal colonization ability and biocompatibility.
However, safety barriers have hindered clinical application. The risk of viable bacteria proliferating beyond control or causing abnormal horizontal gene transfer with the gut microbiota has consistently been cited as a critical limitation. Although auxotrophic strains (which die without specific nutrients) and cell suicide circuits have been engineered, they remain insufficient to completely prevent strain leakage due to metabolic bypasses or mutations.
Technical difficulties also persisted in the design and production of therapeutic agents. The therapeutic molecules bacteria secrete are often structurally simple or extremely limited in variety. There was also a lack of proper means to rapidly search for and verify functional peptides that can selectively penetrate cancer cell membranes to kill tumors. Ultimately, constructing a next-generation microbial chassis that satisfies both safe biological isolation and high-efficiency drug production has been a long-standing goal in academia.
Key Findings
The researchers removed the native plasmids pM1 and pM2 from the EcN strain and redesigned them to separate biological safety functions from drug production capabilities. This approach involved reconstructing the endogenous plasmids to maintain strain-specific stability while accommodating multi-functional circuits.
The first plasmid was equipped with a CRISPR-Cas12-based chromosome-shredding device. It induces the Cas12 enzyme to cut the host genome at multiple sites, permanently halting bacterial replication and proliferation. While the strain with excised genome cannot divide on its own, it synthesizes predetermined drugs while intracellular transcription and translation machinery is temporarily maintained, before gradually dying off. This marks the completion of a sterile therapeutic production system that eliminates concerns about in vivo persistence.
The second plasmid was engineered with a programmable circuit that precisely controls the expression of target therapeutic agents. To broaden the range of therapeutic candidates, an AI-based bioinformatics prediction pipeline was also established. This structure derives Anticancer Cell-Penetrating Peptides (ACCPPs) that combine tumor cell penetration and cancer cell-killing capabilities by learning amino acid sequences and physicochemical properties.
Researchers expressed promising ACCPP candidates, selected via the AI pipeline, in the redesigned EcN chassis. Subsequent human cancer cell culture experiments confirmed that the non-proliferative bacteria stably secrete the target peptides and induce tumor cell death. This is the fruit of integrating computer-based drug discovery and genetic engineering-based microbial production into a single platform.
Significance and Outlook
This achievement has raised the safety standards for synthetic biology-based therapeutics by separating proliferation control and drug production into independent modules. Thanks to the plasmid replacement architecture, it is also relatively easy to swap genetic circuits for various target molecules in the future, such as those associated with other cancers or inflammatory bowel diseases. Bacteria with proliferation fundamentally blocked are expected to have an advantage in overcoming regulatory hurdles for clinical entry, as the risk of systemic sepsis or extracorporeal leakage is reduced.
Clear practical challenges remain. Since this experiment has completed proof-of-concept in human cancer cell culture models, it is necessary to determine whether the same therapeutic response is reproduced in actual animal tumor models and complex gut microbiota environments. It is essential to obtain long-term safety data to determine whether the Cas12 enzyme accurately cleaves the genome at the precise time in vivo without malfunction, and whether a sufficient therapeutic concentration can be stably maintained before strain death. Establishing a standardized production process at the Good Manufacturing Practice (GMP) level is also a key factor for commercialization.
Synthetic biology enables the installation of programmable genetic functions in microorganisms, positioning them as versatile platforms for therapeutic applications. However, clinical translation requires chassis organisms that integrate safety, genetic stability, and precise control over the activity. Here, we present an innovative engineering framework for the probiotic Escherichia coli Nissle (EcN) that transforms EcN into a safe, efficient, and non-proliferative chassis for therapeutic production. The native EcN plasmids pM1 and pM2 were cured and repurposed to encode two functional modules: a CRISPR-Cas12 chromosome-shredding device that prevents bacterial propagation and a programmable circuit for payload expression, thereby generating a non-proliferative therapeutic-producing system. To broaden the range of therapeutic payloads, we developed an artificial-intelligence-guided bioinformatic pipeline to predict anticancer cell-penetrating peptides (ACCPPs). As a proof of concept, a candidate identified through this workflow was produced in the engineered EcN chassis and evaluated in human cell culture for anticancer activity. Collectively, this strategy established a modular platform that integrates computational therapeutic discovery with engineered probiotic production, paving the way for the future development of safer microbial therapies and their clinical applications.
This technology can be immediately applied as a patient-customized biological agent acting locally in the tumor microenvironment. A representative scenario involves rapidly designing optimal anticancer peptides via the AI pipeline according to the characteristics of an individual's cancer tissue and loading them into a standardized non-proliferative EcN chassis for administration via oral delivery or local intratumoral injection.
A bacterial platform that has permanently lost its ability to proliferate significantly lowers the biological isolation regulatory barriers faced by existing live biotherapeutics. Because it does not persist in the body after drug delivery and self-degrades, it is assessed as more likely to clear the stringent approval hurdles set by regulatory authorities such as the US FDA, which are concerned with environmental release risks and the spread of antibiotic resistance genes. The structure is also well-suited for expansion beyond anticancer agents to include substances that regulate autoimmune diseases or secretomes for the treatment of metabolic disorders.