Host Histone Modifications Regulate the Legionella Gene Repressor RomA
Background
Legionella pneumophila is a bacterium that replicates within macrophages and causes Legionnaires' disease, a severe form of pneumonia. The bacterium injects hundreds of effector proteins into host cells during infection to rewire immune signaling and metabolism. Some effectors translocate to the nucleus and function as 'nucleomodulins' by directly altering chromatin.
The representative effector RomA and its homolog LegAS4 are SET domain lysine methyltransferases. These two proteins repress host immune gene expression and promote bacterial intracellular replication by methylating lysine 14 of histone H3 (H3K14). Previous studies assumed that this enzyme modifies target sites independently. However, it remains unclear how pre-existing post-translational modifications, such as methylation or acetylation on host chromatin, restrict the binding and catalytic activity of bacterial enzymes.
Key Findings
Researchers at the Van Andel Institute measured RomA's methylation activity on an array of 276 synthetic histone peptides. RomA preferentially used unmodified H3 tails as substrates rather than H2A, H2B, or H4. It was particularly sensitive to the methylation state of lysine 4 on H3. As the methylation level of H3K4 increased from one to three, RomA activity decreased progressively, with H3K4 trimethylation (H3K4me3) reducing activity to about 5% of that observed with unmodified H3. Asymmetric dimethylation of H3R2 and phosphorylation of H3S10 also reduced activity by approximately 50% and 20%, respectively.
AlphaScreen binding analysis revealed that H3K4me2, H3K4me3, H3R2me2a, and H3S10p all interfered with RomA's binding to the H3 tail. RomA did not bind to H3K14 acetylated substrates. When comparing 93 recombinant nucleosomes with different modifications, H3K14 methylation was completely blocked in nucleosomes with H3K4me3. Active transcription marks such as H3K4 acetylation and H4K12 monomethylation (H4K12me1) also showed strong inhibitory effects.
The mechanism operated in two ways. For RomA to modify H3K14, both H3K4 and H3K14 in the same H3 tail must remain unmodified. This is a cis cross-talk within the same histone. In contrast, H4K12me1 inhibited RomA from the opposite side of the nucleosome, acting as trans cross-talk. Cryo-electron microscopy analysis confirmed that RomA did not bind to the common acidic patch of the nucleosome but instead interacted with the flexible histone tail.
Implications and Outlook
These findings redefine bacterial effectors not as indiscriminate enzymes that leave marks on host chromatin, but as 'readers and writers' that operate only in permitted environments by reading existing histone modifications. It is also notable that bacterial effectors utilize both cis and trans cross-talk mechanisms known in eukaryotic chromatin regulators. If H3K4me3 and H4K12me1, which are abundant in active gene promoters, block RomA, Legionella may selectively repress specific chromatin regions in response to the host's transcriptional state.
However, the study primarily focused on biochemical and structural analyses using purified proteins, synthetic peptides, and recombinant nucleosomes. It remains to be determined which genomic loci are protected or targeted by RomA in infected cells, and the quantitative impact of each modification on bacterial replication and immune response. Chromatin immunoprecipitation sequencing, transcriptome analysis, and validation using RomA-deficient strains are needed to follow up on these findings.
Proceedings of the National Academy of Sciences, Volume 123, Issue 35, September 2026. SignificanceBacterial pathogens reprogram host gene expression by delivering effector proteins that modify chromatin, but it is not known how the host epigenetic environment impacts effector function. Here, we show that the Legionella effector RomA senses ...
If the mechanism by which RomA recognizes unmodified H3 tails or is inhibited by H4K12me1 can be replicated, it may be possible to design small molecules or peptide inhibitors that block Legionella's manipulation of host transcription. For example, compounds that block the H3K4 recognition surface could interfere with nucleosome binding without directly inhibiting RomA's catalytic site. Ensuring selectivity between bacterial enzymes and human SET domain enzymes could also reduce potential side effects on the host epigenome.
Additionally, analyzing histone modifications and RomA-targeted genes in patient-derived macrophages could help identify biomarkers that predict infection susceptibility or disease severity. However, it is too early to assess the efficacy of these targets based on current results. The drug-like potential, toxicity, and ability to inhibit bacterial replication must first be validated in infected cells and animal models.