Gut Microbes Are Quietly Rewiring Your Liver’s DNA Switches, New Study Finds

Scientists have long known the gut and liver are in constant chemical conversation, but a new study published in Molecular Cell suggests that conversation reaches deeper than previously understood: gut microbes appear to influence which DNA regulatory switches are active in liver cells, reshaping gene expression in ways that matter for metabolic health. The findings are preliminary, but they point toward a molecular mechanism that could help explain why disruptions in gut bacteria so often track with liver disease.

How 109,386 Liver DNA Switches Were Mapped

The core achievement of the study is a systematic map of cis-regulatory elements (CREs), the stretches of non-coding DNA that act as dimmers and switches for nearby genes, built from 109,386 human liver-derived sequences. Most existing maps of these elements come from cell cultures that don't behave the way a living liver does. To get closer to reality, the researchers used massively parallel reporter assays (MPRA), a method that tests thousands of DNA sequences simultaneously for regulatory activity, run both in lab-grown HepG2 liver cells and, crucially, in living mouse hepatocytes.

The in vivo screen identified 9,178 activating and 2,570 repressing CREs, representing roughly 17.7% of all sequences tested. These functionally active CREs were enriched for a histone modification called H3K27ac (a marker of open, transcriptionally active chromatin) and tended to sit in accessible regions of the genome. More than half of the distal CREs active in living animals showed no activity in the dish-based assay, which underscores why spatial single-cell omics and similar in vivo approaches are increasingly seen as essential for understanding real liver biology.

Germ-Free Mice Reveal Microbial Influence on Liver Gene Regulation

To test whether gut bacteria affect this regulatory landscape, the team compared specific pathogen-free (SPF) mice, which carry a normal microbiome, with gnotobiotic mice raised in sterile conditions with no gut bacteria at all. The difference was substantial: 3,169 CREs showed reduced activity in germ-free animals, and 1,286 showed increased activity, for a total of 4,455 microbiota-responsive regulatory elements. About 60% of these overlapped with CREs that had been confirmed as active in vivo.

Genes linked to these responsive CREs also showed significantly lower expression in germ-free mice, and that same pattern held when the researchers cross-referenced human liver RNA-sequencing data. The transcription factor NFE2L2, a master regulator of the antioxidant response, was reduced in germ-free liver tissue, while an enzyme called Cul3, which degrades NFE2L2, was elevated. This suggests that gut bacteria help sustain the liver's defenses against oxidative stress through the KEAP1/NFE2L2 pathway. The gut-liver axis and its molecular mechanisms have attracted considerable interest, though researchers note it remains unclear in many contexts whether microbial changes drive liver dysfunction or result from it.

Specific Metabolites Target Specific Regulatory Elements

The study moved beyond the general observation that microbes matter to ask which metabolites do the work. When the team treated HepG2 cells with metabolites derived from human stool samples, 1,810 genes changed expression, with enrichment in lipid metabolic processes and cell proliferation. Three specific CREs linked to the genes Gdf15, Sept4, and Ctsz responded to metabolite treatment.

Indole-derived metabolites, including indole-3-propionic acid (IPA) and 5-methoxyindole-3-acetic acid (5-MIAA), and aromatic metabolites like phenylpropionic acid (PPA) and 4-hydroxy-phenylpropionic acid (4-OH-PPA) robustly activated the CREs controlling Gdf15 and Sept4. Short-chain fatty acids including butyrate, propionate, and acetate had minimal effects on those same elements. When Firmicutes bacteria were selectively depleted from fecal cultures using cyclic peptides, the induction of Sept4 and Gdf15 was reduced, but Ctsz was unaffected, suggesting different bacterial communities drive different regulatory programs.

This kind of metabolite-to-regulatory-element specificity matters because gut microbiome-driven epigenetic modifications in liver disease are increasingly recognized as relevant to conditions like metabolic dysfunction-associated steatotic liver disease (MASLD). Non-alcoholic fatty liver disease, a closely related condition, affects roughly 25% of the global population and can progress to cirrhosis and liver cancer, which gives this line of research considerable clinical urgency.

A Genetic Variant That Changes How Your Liver Hears Microbial Signals

Perhaps the most striking finding involves a rare non-coding genetic variant, rs553586668, located within a microbiota-responsive CRE that regulates two genes: Inf2, which has oncogenic associations, and Adss1. The alternative allele (A) of this variant showed significantly higher baseline regulatory activity than the reference allele (G) in dual-luciferase reporter assays, and it was also more responsive to microbial metabolites. In silico mutagenesis predicted that the alternative allele increases chromatin accessibility in hepatocytes, meaning the DNA in that region is physically more open and available for transcription factor binding.

This is a proof-of-concept for a broader idea: genetic variation in non-coding regulatory regions could alter how sensitive a person's liver is to microbial signals, potentially explaining some of the individual variation in liver disease risk that standard risk factors don't account for.

What the Study Can't Yet Prove, and Where It Points Next

The researchers are careful to note real limitations. The in vivo MPRA uses an episomal, or free-floating, reporter construct that doesn't sit inside the normal chromatin architecture of the genome, which could affect how faithfully it reports on endogenous regulatory activity. The system also runs human CRE sequences through mouse hepatocytes, and species differences in metabolism may introduce noise. The authors flag CRISPR-based editing of endogenous CREs as the necessary next step for establishing direct causal links between specific regulatory elements and gene expression changes.

On the therapeutic side, microbiome-targeted strategies for liver disease, including probiotics, prebiotics, and fecal microbiota transplantation, are already being explored in early clinical trials. This study adds a mechanistic layer to those efforts: if specific metabolites from specific bacterial groups activate specific liver regulatory elements, it may eventually be possible to engineer microbial interventions that target precise gene expression programs rather than simply shifting the microbiome in a general direction. That kind of precision is still years away, but the map is now considerably more detailed.

Evidence rating

EMERGING — 84/100

Caveats:

• The in vivo reporter assay used an episomal (non-integrated) construct, which does not fully replicate native chromatin architecture and may over- or under-estimate the regulatory effects observed.
• Key microbiota-depletion experiments were conducted in germ-free mice; direct causal validation in humans has not yet been performed.
• Funding source was not disclosed in the available materials, so independent replication by other groups has not yet occurred.
• CRISPR-based endogenous editing of the identified CREs was not performed, leaving the causal role of specific variants partially inferred.

This study used massively parallel reporter assays (MPRA) to profile 109,386 human liver-derived cis-regulatory elements under matched in vitro and in vivo conditions, including comparisons between germ-free and specific-pathogen-free mice. It demonstrates association-to-mechanism links rather than pure association, with named metabolites and pathways, though the episomal reporter design does not fully capture native chromatin context and causal endogenous validation via genome editing was not performed. Findings have not yet been independently replicated in a separate cohort or laboratory.

References

1. Chadmirah Zaratiana, Yiamunaa M, Yong-An Lee, et al. (2026). Gut microbiota modulation of regulatory DNA elements revealed by massively parallel functional characterization. Molecular Cell. https://doi.org/10.1016/j.molcel.2026.03.036
2. Yuan Suo, Robert Thimme, Bertram Bengsch (2025). Spatial single-cell omics: new insights into liver diseases. Gut. https://doi.org/10.1136/gutjnl-2024-332105
3. Liang Ma, He Wang, Qiuyu Jin, Zhiwen Sun, Shuang Yu, Yang Zhang (2025). The Gut-Liver Axis: Molecular Mechanisms and Therapeutic Targeting in Liver Disease. International Journal of General Medicine. https://doi.org/10.2147/ijgm.s551494
4. Suki Ha, Vincent Wai-Sun Wong, Xiang Zhang, Jun Yu (2024). Interplay between gut microbiome, host genetic and epigenetic modifications in MASLD and MASLD-related hepatocellular carcinoma. Gut. https://doi.org/10.1136/gutjnl-2024-332398
5. The role of gut microbiota in the occurrence and progression of non-alcoholic fatty liver disease – Frontiers. frontiersin.org. https://www.frontiersin.org/articles/10.3389/fmicb.2023.1259275/full
6. Dharma Ayer, Alexa Trovato, Micheal Tadros (2023). Utilizing the Gut Microbiome as a Therapeutic Target for Liver Disease – Narrative Review. Journal of Translational Gastroenterology. https://doi.org/10.14218/jtg.2023.00027