Knocking out M. tuberculosis small RNAs MTS0997 and MTS1338 shortens mouse survival without changing lung bacterial burden

The MTS0997-deleted strain (d_0997) and the double-knockout strain (dd_0997_1338) had highly similar transcriptomes during culture growth, both clearly distinct from wild type, whereas d_1338 showed only two differentially expressed genes. Previous characterization of how these two small RNAs regulate the mycobacterial transcriptional profile was incomplete; this study directly compared unmarked single- and double-knockout strains by whole-genome RNA-seq and reported the number of differentially expressed genes per strain (2 for d_1338, 33 for d_0997, 52 for dd_0997_1338, with 30 shared between d_0997 and the double knockout). Based on triplicate RNA-seq with a threshold of adjusted p-value ≤ 0.01 and |log2FC| ≥ 1; PCA showed knockout strains forming distinct transcriptional clusters, and selected genes were validated by qRT-PCR.

MTS0997 deletion was accompanied by changes in succinate/fumarate respiration genes, up-regulation of ESX-1 secretion system genes (espA, espC, espD and others), and elevated expression of lipid metabolism genes (fad11, fadE23, fadE24, lipQ). It extends the previously known phenotype of MTS0997 promoting fatty acid utilization under acidic stress into a broader transcriptional rearrangement of respiration, virulence factor secretion and lipid metabolic pathways, and notes that only two genes, the succinate dehydrogenase components Rv0247c (>10-fold) and Rv0248c (7-fold), were up-regulated in all three knockout strains. From RNA-seq of bacteria grown to OD600 ~1.0 plus qRT-PCR validation; in the complemented strain dd_0997_1338_comp these genes returned to levels close to wild type.

After low-dose respiratory challenge (~100 CFU per mouse) of highly susceptible I/St mice, all three knockout strains produced significantly shorter survival than wild type, yet lung CFU counts at 6 weeks post-infection did not differ between groups. The same single-knockout mutants had previously given different survival outcomes in an intravenous high-dose (~10^6 CFU) model; the authors attribute the discrepancy to route and dose, and this study reports survival and bacterial burden side by side, showing the virulence phenotype is not explained by bacterial load. Kaplan-Meier multiple comparison showed significant survival differences; mice infected with the complemented strain dd_0997_1338_comp survived significantly longer than those given the double knockout (P<0.01) and did not differ statistically from wild type, although complementation appeared only partial.

Lungs of mice infected with the double-knockout strain showed lower expression of the pro-inflammatory cytokine genes il6, il12b and ifng and lower CD4+ T-cell numbers and frequency, whereas single-knockout infections reduced lung B-cell frequency and raised MHC-II H2-A surface expression about 2-fold, changes absent in double-knockout-infected mice. It links bacterial small RNA deletion to host lung immune cell composition and cytokine transcription, and interprets the discordance between single- and double-knockout phenotypes as transcriptional epistasis, a phenomenon still poorly understood in mycobacteria. Based on qRT-PCR of infected mouse lung tissue and flow cytometry (single-cell suspensions prepared individually from 9 mice), with ANOVA and t-tests and P<0.05 taken as significant.

This work provides unmarked knockout and complemented strains for studying MTS0997 and MTS1338 separately and together in the H37Rv background, along with publicly available bacterial and mouse transcriptome data (GEO GSE330228 and GSE220660; mutant strain data in NCBI SRA PRJNA701202). It applies to studies using murine bone marrow-derived macrophages as an in vitro model and low-dose aerosol infection of highly susceptible I/St mice as an in vivo model, and to researchers focused on ESX-1 secretion, succinate/fumarate respiration and lipid metabolic pathways in TB virulence. The authors note that phenotypes in non-SPF, non-sterile 'dirty mice' may be more relevant to general human populations, while SPF animals would likely give more consistent results with smaller deviations.

This is an incomplete reading: figures and supplementary tables were not included, so the full list of differentially expressed genes and the specific values and statistical details in the supplementary figures and tables cannot be checked here. The authors caution that host gene expression changes in cultured macrophages cannot be directly extrapolated to in vivo immune and inflammatory pathways; the opposite trends in MHC-II between BMDMs and lung tissue, and the fact that lower CXCL1/CXCL2 expression was not reflected in reduced neutrophil infiltration, indicate that the two systems need to be interpreted separately. A reliable target for MTS1338 has still not been identified, and the authors propose it may act on host RNA sensors such as RIG-I or MDA5, which remains a hypothesis to be tested. The mechanism of transcriptional epistasis in mycobacteria is unclear, so the explanation for the divergent single- and double-knockout phenotypes remains open. Complementation only partially restored the low-virulence phenotype, suggesting the genetic basis of that phenotype still needs further clarification.