The small non-coding RNA B11 regulates multiple facets of Mycobacterium abscessus virulence

Abstract

Mycobacterium abscessus causes severe disease in patients with cystic fibrosis. Little is known in M. abscessus about the roles of small regulatory RNAs (sRNA) in gene regulation. We show that the sRNA B11 controls gene expression and virulence-associated phenotypes in this pathogen. B11 deletion from the smooth strain ATCC_19977 produced a rough strain, increased pro-inflammatory signaling and virulence in multiple infection models, and increased resistance to antibiotics. Examination of clinical isolate cohorts identified isolates with B11 mutations or reduced expression. We used RNAseq and proteomics to investigate the effects of B11 on gene expression and test the impact of mutations found in clinical isolates. Over 200 genes were differentially expressed in the deletion mutant. Strains with the clinical B11 mutations showed expression trends similar to the deletion mutant, suggesting partial loss of function. Among genes upregulated in the B11 mutant, there was a strong enrichment for genes with B11-complementary sequences in their predicted ribosome binding sites (RBS), consistent with B11 functioning as a negative regulator that represses translation via base-pairing to RBSs. Comparing the proteomes similarly revealed that upregulated proteins were strongly enriched for B11-complementary sequences. Intriguingly, genes upregulated in the absence of B11 included components of the ESX-4 secretion system, critical for M. abscessus virulence. Many of these genes had B11-complementary sequences at their RBSs, which we show is sufficient to mediate repression by B11 through direct binding. Altogether, our data show that B11 acts as a direct negative regulator and mediates (likely indirect) positive regulation with pleiotropic effects on gene expression and clinically important phenotypes in M. abscessus. The presence of hypomorphic B11 mutations in clinical strains is consistent with the idea that lower B11 activity may be advantageous for M. abscessus in some clinical contexts. This is the first report on an sRNA role in M. abscessus.

Fig 1. Disruption of the sRNA B11 causes rough morphology with reduced GPL content in M. abscessus.

A. The indicated strains were grown in liquid media with Tween 80, diluted and plated on solid media with and without Tween 80. Colonies were imaged after 7 days of growth. See S1E Fig for exact mutations of ∆B11+B11_mutated. B. Northern blotting of duplicate RNA samples from the indicated strains confirmed that B11 expression was reduced in the transposon mutant, absent from the deletion mutant, and restored in the complemented mutant. C. Thin layer chromatography analysis of surface lipids extracted from the indicated strains grown on 7H10-ADC agar plates. Lipids extracted from equivalent amounts of bacterial cells were analyzed on silica gel 60-precoated TLC plates F254 (Merck) developed in the solvent system chloroform:methanol:water (90:10:1, v/v/v). The plate was revealed by spraying with α-naphthol and heating. D. Quantitative PCR reveals that expression of two key GPL biosynthesis genes is modestly but significantly reduced in strains lacking B11. * P < 0.05; ** P < 0.01; *** P < 0.001; ANOVA with Tukey post-test. qPCR comparisons where P > 0.05 are not shown. qPCR data represent triplicate log-phase cultures. All experiments shown in this figure were performed at least twice and representative data are shown.

Fig 2. Disruption of M. abscessus B11 increases virulence and cytokine secretion.

A. Galleria mellonella infections. 36 larvae per group were infected by ~3000 CFU of WT M. abscessus, the ΔB11 mutant, or the ΔB11+B11 complemented strain. 20 larvae were injected with saline. Larvae were followed for two weeks. The survival in each group is showed in a Kaplan-Mayer curve. The survival of infection with WT M. abscessus was compared to each other group by log-rank (Mantel-Cox) test. *, P < the Bonferroni-corrected threshold of 0.0167. B. 20 larvae were infected by WT, and 20 by the ΔB11 mutant, ~1000 cfu/larva. 10 larvae from each group were sacrificed on day 0, and the remaining on day 3, and CFU were enumerated by plating. Y axis represents proliferation at day 3 over day 0, showing no significant advantage of either strain. C. RAW macrophage-like cells were infected with the indicated M. abscessus strains and bacteria were enumerated three days later. Three separate experiments were performed (expt. 1-expt. 3) with 2–4 replicates per experiment. The mean fold-increase for each experiment is shown. D. RAW cells were infected with M. abscessus at an MOI of 5 and TNF secretion was measured after 6 hours. E. RAW cells were exposed to heat-killed M. abscessus at the equivalent to an MOI of 20, washed, and TNF secretion measured after 24 hours. F. Lung bacterial burden for C57BL/6NCrl mice four days after infection with the indicated M. abscessus strains. G. Weight change in mice following infection with the indicated M. abscessus strains or mock infection with PBS-infused agar beads. Points are error bars represent mean and SD. n = 9 mice per group for infections and 3 mice for the mock infection. H. The indicated cytokines were measured in lung homogenates four days after infection with the indicated M. abscessus strains. Bar charts display means and SD. Black dots indicate individual samples. ** P < 0.01; *** P < 0.001; **** P < 0.0001, ANOVA with Tukey post-test. All experiments shown in this figure except for the murine infections were done at least twice and representative data are shown.

Fig 3. Disruption of the sRNA B11 in M. abscessus causes gene expression changes at the mRNA and protein levels particularly in genes with B11-complementary ribosome-binding sites.

A-C. DESeq2 (93) was used to identify genes differentially expressed in a ∆B11 strain compared to a WT strain and in a ∆B11 strain compared to a∆B11 complemented with B11 expressed from an ectopic location (∆B11+B11). mRNA abundance was measured by RNAseq. Genes that met the thresholds of > = 2-fold mRNA abundance change and adjusted P < 0.05 were classified as differentially expressed. A-B, Venn diagrams show genes that met the criteria for overexpression or underexpression in the ∆B11 strain compared to the WT or complemented strains. Only genes with data in both datasets are shown. C. The log₂ fold changes of genes that met the differential expression criteria for either comparison were plotted to assess their correlation. Pink dots indicate genes that met the differential expression criteria for the ∆B11/WT comparison only, blue dots indicate genes that met the differential expression criteria for the ∆B11/∆B11+B11 comparison only, and purple dots indicate genes that met the criteria for both comparisons. D-F. LC-MS/MS was used to compare protein abundance in the WT and ∆B11 strains. Proteins with > = 2-fold abundance differences and between-replicate CVs < 0.5 were classified as differentially expressed. D-E, Venn diagram show the overlap between genes that met the criteria for overexpression or underexpression by RNAseq (as in A-B) and the proteins that met the criteria for overexpression or underexpression by LC-MS/MS. Only genes/proteins with data in both datasets are shown. F. The log₂ fold changes of genes that met the differential expression criteria for RNAseq or LC-MS/MS were plotted to assess their correlation. Pink dots indicate genes that met the differential expression criteria by RNAseq only, teal dots indicate genes that met the differential expression criteria by LC-MS/MS only, and gray dots indicate genes that met the criteria for both methods. G. Differentially expressed genes are enriched for the presence of B11-complementary sequences in their ribosome bind sites. Genes/proteins were categorized according to whether or not they met the differential expression criteria used in D-F, and the proportion of genes/proteins in each category that had > = 6 nt of consecutive sequence in the 25 nt upstream of their start codons that was complementary to one or both loops of B11 was determined. Only genes (and their encoded proteins) with defined transcription start sites were included in the analysis. * P < 0.05; **** P < 0.0001, Fisher’s exact test. All Venn diagrams were made by BioVenn (101).

Fig 4. Disruption of M. abscessus B11 causes increased expression of several ESX-related genes with B11-complementary sequences.

A. Disruption or deletion of B11 causes increased expression of genes encoding the EsxU, EsxT, and the ESX-4 structural protein EccB. B. Deletion of B11 causes increased expression of sigM and MAB_0628, which is predicted to be transcribed upstream of espI on a polycistronic transcript (see S4 Fig). Bars display means and SD. Black dots show the values for each sample. * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001, ANOVA with Tukey post-test. qPCR comparisons where P > 0.05 are not shown. qPCR data represent triplicate log-phase cultures. Data are representative of at least two independent experiments.

Fig 5. B11 represses expression of genes harboring the 5’ UTR of M. abscessus eccB4 via direct binding.

A. Schematic of the predicted secondary structure of B11 demonstrating the possibility of base-pairing to the ribosome binding site of the eccB4 5’ UTR. The eccB4 start codon is underlined and bolded. Note that the 6-G sequence in the UTR could also base-pair with loop 1 of B11. B. Reporter constructs were built containing different versions of B11 expressed from its native promoter (left) and different versions of the eccB4 5’ UTR linked to the mCherry gene and expressed from the B11-insensitive MOP promoter. Mutated nts are indicated with bolded blue. The eccB4 5’ UTR sequence complementary to B11 is indicated in bolded red. Start codons are bolded and underlined. C. Reporter constructs containing the WT eccB4 5’ UTR (left) or the mutated version (right) along with the indicated version of B11 were transformed into WT or ΔB11 M. abscessus as indicated. mCherry fluorescence was measured by flow cytometry of biological replicate cultures. In each chart, the ΔB11 strain complemented with WT B11 was compared to the other strains by ANOVA followed by Dunnett’s multiple comparisons test. Bar charts display means and SD. Black dots represent individual samples. * P < 0.05; ** P < 0.01; **** P < 0.0001. Data are representative of two independent experiments.

Fig 6. B11 mutations found in clinical M. abscessus strains cause partial loss of function.

A. Deletion of one C from loop 1 was found in one clinical isolate, while insertion of one C into loop 2 was found in 3 clinical isolates. B. Smooth morphology was largely restored upon complementation of the ΔB11 strain with B11 harboring each of the two clinical loop mutations. C. RNAseq revealed that genes involved in GPL synthesis and transport had reduced expression in the ΔB11 strain and this was largely complemented by each of the clinical loop mutations. Each dot represents a gene associated with GPL metabolism (62). mps1 and mps2 are indicated with green dots. ** P < 0.01, **** P < 0.0001, one way RM ANOVA. D-E. Intersections of genes differentially expressed (log₂ fold change > 1 or < -1, P < 0.05) in ΔB11/ΔB11+B11, ΔB11+B11_del1/ΔB11+B11, and ΔB11+B11_ins1/ΔB11+B11. Venn diagrams were created with BioVenn (101). P < 0.0001 for the overlaps between between all pairs of comparisons, hypergeometric test. F. There is a correlation between log₂ fold change in genes differentially expressed in ΔB11/ΔB11+B11 and/or in ΔB11+B11_del1/ΔB11+B11. Dark green dots represent genes differentially expressed in both comparisons. Pink dots represent genes that met the criteria for differential expression in ΔB11/ΔB11+B11 but not in ΔB11+B11_del1/ΔB11+B11. Light green dots represent genes that met the criteria for differential expression in ΔB11+B11_del1/ΔB11+B11 but not in ΔB11/ΔB11+B11. G. There is a correlation between log₂ fold change in genes differentially expressed in ΔB11/ΔB11+B11 and/or in ΔB11+B11_ins1/ΔB11+B11. Magenta green dots represent genes differentially expressed in both comparisons. Pink dots represent genes that met the criteria for differential expression in ΔB11/ΔB11+B11 but not in ΔB11+B11_ins1/ΔB11+B11. Periwinkle dots represent genes that met the criteria for differential expression in ΔB11+B11_ins1/ΔB11+B11 but not in ΔB11/ΔB11+B11.

Fig 7. A clinical M. abscessus strain does not express B11 and had increased antibiotic sensitivity upon ectopic B11 expression.

A. Northern blot for B11 in log phase cultures of WT M. abscessus, a transposon-mutant of mps2 (MAB_4098c), and twelve clinical isolates from an Israeli cohort. B. The clinical isolate LAH was electroporated with a single copy plasmid encoding B11 under the MOP promoter, and sensitivity to linezolid and rifampin was examined. Data are representative of two independent experiments.