Bacteria in lung tumors may fuel cancer mutations

Bacteria in lung tumors may do more than simply coexist with cancer cells. In samples of non-small cell lung cancer, researchers found an association between Mycobacterium abscessus and a distinctive pattern of mutations, then used experimental models to trace a molecular pathway from a bacterial protein to increased mutagenesis in tumor cells.

The finding does not mean that M. abscessus causes lung cancer. It points to a different possibility: once present inside an established tumor, a bacterium may activate a host defense system that can accidentally become a source of new mutations.

The researchers analyzed paired tumor and adjacent normal tissue from 43 patients with non-small cell lung cancer. They combined exome sequencing, bacterial profiling, and proteomics to examine both the tumor genome and its microbial environment.

One organism that stood out was M. abscessus, a rapidly growing non-tuberculous mycobacterium commonly found in environmental reservoirs such as soil and water and known to colonize the respiratory tract in susceptible people.

Fluorescence imaging detected intracellular mycobacterial signals in tumor cells. Because the probe recognized the Mycobacterium genus rather than M. abscessus specifically, the researchers cross-checked the finding using bacterial sequencing. In 12 tumors with strong intratumoral mycobacterial signals, M. abscessus was also highly represented in the sequencing data.

From a bacterium to a mutational signature

Tumors with a higher M. abscessus burden showed increased expression of APOBEC3A. APOBEC3A is part of the innate antiviral defense system and can convert cytosine to uracil in single-stranded DNA. That activity can help defend cells against pathogens, but persistent or misplaced APOBEC3A activity can also damage the cell's own genome.

The mutations produced by this process leave recognizable genomic patterns known as APOBEC mutational signatures. In the patient samples, greater M. abscessus abundance was associated with stronger APOBEC-related signatures.

An association in human tumors cannot by itself establish causality, so the investigators tested the proposed mechanism in experimental systems. Across cancer cells, human non-small cell lung cancer organoids, xenografts, and genetically engineered mouse lung tumors, exposure to M. abscessus increased APOBEC3A expression and APOBEC-associated mutagenesis.

In one xenograft model, the bacteria were injected directly into tumors. These experiments therefore demonstrate that the bacterium can produce the effect under controlled conditions; they do not reproduce the natural route by which M. abscessus might colonize a human tumor.

Removing APOBEC3A from cancer cells weakened both the increase in APOBEC-related mutations and part of the tumor-growth effect associated with M. abscessus. That placed APOBEC3A near the center of the pathway.

A bacterial protein hijacks an antiviral signal

The researchers then screened 23 major M. abscessus effector proteins for their ability to activate the APOBEC family. One stood out: a nucleoside diphosphate kinase called NDK.

Further experiments revealed an unusual interaction between NDK and IRF3, a human transcription factor involved in antiviral signaling. The bacterial protein modified histidine 263 of IRF3 by attaching a phosphate group to it.

This modification amplified type I interferon signaling. The stronger signal, in turn, sustained APOBEC3A expression. In simplified form, the pathway runs from bacterial NDK to IRF3, from IRF3 to interferon signaling, and from prolonged APOBEC3A activity to additional mutations in tumor DNA.

Changing histidine 263 in IRF3 disrupted NDK-dependent activation of interferon and APOBEC3A. Knocking down NDK in M. abscessus also reduced APOBEC3A induction and the associated mutational phenotype. These intervention experiments are important because they move the study beyond a correlation between microbial abundance and tumor mutations.

What the finding changes

Cancer evolution depends on genetic diversity. New mutations continually create variants from which selection can favor cells that grow more efficiently or survive immune pressure and treatment. A process that increases the supply of mutations could therefore influence how a tumor evolves.

Tumor-associated microbes have previously been studied mainly for their effects on immunity, inflammation, metabolism, and treatment response. This study describes a more direct route: a bacterial effector alters signaling inside the cancer cell and activates an endogenous enzyme capable of changing the tumor genome itself.

The authors also designed an experimental degrader targeting bacterial NDK. In mouse models exposed to M. abscessus, the compound reduced NDK and APOBEC3A levels and slowed tumor progression. This remains a preclinical proof of concept rather than a cancer treatment. Its pharmacology, long-term safety, and therapeutic potential in humans have not been established.

The human evidence also has important limits. The clinical analysis involved a single cohort of 43 patients, and the M. abscessus-positive subgroup was small. An observed association with shorter disease-free survival therefore needs independent validation in larger populations. The prevalence of this mechanism across lung cancers is still unknown.

There is also a technical caveat. Phosphohistidine is chemically unstable, which made direct mass-spectrometric mapping of the endogenous IRF3 modification difficult. The authors instead supported the proposed site through biochemical and genetic experiments.

The most defensible conclusion is therefore not that a new bacterial cause of lung cancer has been identified. Rather, the study describes a specific microbial mechanism that can increase mutation production in tumor cells under experimental conditions. The next question is how often that mechanism operates strongly enough in patients to alter the course of tumor evolution.