MoaB2, a newly identified transcription factor, binds to σA in Mycobacterium smegmatis

Abstract

In mycobacteria, σ^A is the primary sigma factor. This essential protein binds to RNA polymerase (RNAP) and mediates transcription initiation of housekeeping genes. Our knowledge about this factor in mycobacteria is limited. Here, we performed an unbiased search for interacting partners of Mycobacterium smegmatis σ^A. The search revealed a number of proteins; prominent among them was MoaB2. The σ^A-MoaB2 interaction was validated and characterized by several approaches, revealing that it likely does not require RNAP and is specific, as alternative σ factors (e.g., closely related σ^B) do not interact with MoaB2. The structure of MoaB2 was solved by X-ray crystallography. By immunoprecipitation and nuclear magnetic resonance, the unique, unstructured N-terminal domain of σ^A was identified to play a role in the σ^A-MoaB2 interaction. Functional experiments then showed that MoaB2 inhibits σ^A-dependent (but not σ^B-dependent) transcription and may increase the stability of σ^A in the cell. We propose that MoaB2, by sequestering σ^A, has a potential to modulate gene expression. In summary, this study has uncovered a new binding partner of mycobacterial σ^A, paving the way for future investigation of this phenomenon.IMPORTANCEMycobacteria cause serious human diseases such as tuberculosis and leprosy. The mycobacterial transcription machinery is unique, containing transcription factors such as RbpA, CarD, and the RNA polymerase (RNAP) core-interacting small RNA Ms1. Here, we extend our knowledge of the mycobacterial transcription apparatus by identifying MoaB2 as an interacting partner of σ^A, the primary sigma factor, and characterize its effects on transcription and σ^A stability. This information expands our knowledge of interacting partners of subunits of mycobacterial RNAP, providing opportunities for future development of antimycobacterial compounds.

Keywords: MoaB2; RNA polymerase; mycobacteria; transcription; σA.

Fig 1

M. smegmatis MoaB2 is in the interactome of σ^A. (A) Volcano plots of proteins associating with M. smegmatis σ^A-FLAG (strain LK3207) pulled down in exponential (EXP) and stationary (STA) phases of growth. The plots show LC-MS-identified proteins enriched in IP pull downs with anti-FLAG over proteins from the control “no FLAG” strain (LK3016). Red spots indicate proteins significantly enriched (−log₁₀ P < 2, indicated with the horizontal dashed line; enrichment >log₂>2, indicated with the vertical dashed line). The spots show averages from three independent biological repeats. (B) Quantitation of relative enrichments of selected σ^A-FLAG (LK3207) associating proteins from (A) compared to the “no FLAG” strain. Data from exponential (EXP) and stationary (STA) phases are indicated. The bars show averages from three independent biological repeats. The SDs cannot be shown directly in the graph because they are calculated from the intensity values, whereas the fold change is shown in the graph. However, the variance of the replicates is one of the parameters of the P-value calculated in the t-test—the lower the variance, the lower the P-value (P-value_{MoaB2_STA} = 0.0036; P-value_{MoaB2_EXP} = 0.0037; P-value_{α_STA} < 0.001; P-value_{α_EXP} < 0.001; P-value_{β_STA} < 0.001; P-value_{β_EXP} < 0.001; P-value_{σA_STA} = 0.0015; P-value_{σA_EXP} < 0.001). (C) SDS-PAGE of IPs of FLAG-tagged σ^A (LK3207) using the anti-FLAG antibody. “No FLAG” strain is negative control (LK3016). Data from exponential (EXP) and stationary (STA) phases are shown. The dotted line indicates that this panel was electronically assembled from two parts of one gel. MoaB2 is marked with red asterisk. The identity of the bands was determined by mass spectrometry. The experiment was performed in three biological replicates [independent of experiments shown in (A)] with the same result. (D) SDS-PAGE of IP of σ^A from the stationary phase “no FLAG” strain cells (wt, LK3016) using antibody against σ⁷⁰ (anti-σ⁷⁰, clone name 2G10). IgG is a mouse nonspecific IgG used as a negative control. The band corresponding to MoaB2 is indicated with red asterisk. The identity of the bands was determined by mass spectrometry. The experiment was performed in three biological replicates with the same result.

Fig 2

M. smegmatis MoaB2 does not associate with alternative σ factors. SDS-PAGE of IPs of FLAG-tagged sigma factors [σ^A (LK2073), σ^B (LK2077), σ^E (LK2157), σ^F (LK2159), σ^H (LK2160), σ^G (LK2161), or FLAG-tagged β´ subunit of RNAP (LK1468)] using anti-FLAG antibody. The FLAG-tagged proteins were present in the genome in an additional copy under ATC inducible promoter and expressed after ATC induction. In the case of σ^B (2×), twofold amount of cells were harvested for IP to enhance the detection of MoaB2 if present. The “No FLAG” strain was used as a negative control (LK3016). Individual σ factors are marked with black asterisks. Black arrows indicate respective anti-σ factors. MoaB2 is marked with red asterisk. The identity of the bands was determined by mass spectrometry. Three independent experiments were performed with identical results. Visualization of proteins was done also by silver staining with the same result (Fig. S2).

Fig 3

M. smegmatis σ^A interacts with M. smegmatis MoaB2 in vitro. All SEC runs were done using Superdex 200 Increase 10/300 Gl column (GE Healthcare) calibrated using Blue dextran and six protein standards ranging from 12.4 to 669 kDa selected from Gel Filtration Markers Kit for Protein Molecular Weights (MW) 6,500–66,000 Da (Sigma-Aldrich, MWGF70) and Gel Filtration Markers Kit for Protein Molecular Weights (MW) 29,000–700,000 Da (Sigma-Aldrich, MWGF1000). Void volume (V₀) is marked. (A) SEC chromatogram of σ^A from M. smegmatis. Peak elution volume was 12.6 mL, which suggests MW of ~122 kDa. Fractions eluted at 12 to 13.5 mL were pooled and used for formation of the σ^A-MoaB2 complex. (B) SEC chromatogram of MoaB2 from M. smegmatis. Peak elution volume was 13.6 mL, which suggests MW of roughly 77 kDa. Fractions eluted at 13 to 14.5 mL were pooled and used for formation of the σ^A-MoaB2 complex. (C) SEC chromatogram of σ^A-MoaB2 complex formed by mixing samples from (A) and (B). σ^A and MoaB2 were mixed at a molar ratio (monomer:monomer) of 1:2. Peak elution volume corresponding to σ^A-MoaB2 complex was 10.4 mL, to unbound σ^A it was estimated to be 12.4 mL and to unbound MoaB2 13.6 mL. These elution volumes indicate MW of approximately 330 kDa,133 kDa, and 77 kDa, respectively. σ^A elutes before MoaB2 at a lower volume than expected according to its MW because it is a non-globular protein with long loops (the longest intramolecular distance in the structured part of σ^A is ~116 Å) and contains an intrinsically disordered N-terminal domain. The hexamer of MoaB2 is globular (the longest intramolecular distance ~80 Å). Individual peaks are marked with respective proteins, and corresponding fractions used for SDS-PAGE analysis (see Panel F) are indicated with red numbers. (D). SEC chromatogram of CarD from M. smegmatis. Peak elution volume was 16.3 mL that suggests MW of ~23 kDa. Fractions eluted at 16–16.5 mL were pooled and used in E to test whether MoaB2, under the used experimental conditions, does not form unphysiological complexes. The fraction used for SDS-PAGE analysis (see Panel G) is indicated with the red number. (E) SEC chromatogram of CarD and MoaB2 from M. smegmatis. Molar ratio of CarD and MoaB2 was approximately 1:1 (see Panel G). Complex between MoaB2 and CarD is not forming as there was no new peak present, and the amount of CarD eluting at the position of “free” CarD (16.3 mL) in both experiments is the same. Peak elution volume corresponding to MoaB2 was 13.5 mL which suggests MW of ~81 kDa. Individual peaks are marked with respective proteins, and corresponding fractions used for SDS-PAGE analysis (see Panel F) are indicated with red numbers. (F) SDS-PAGE analysis of the σ^A-MoaB2 complex formation. Lines 1–3 contain selected fractions from (C). MoaB2 in line 2 is present due to the tailing of σ^A-MoaB2 complex peak which therefore overlaps with the peak of free σ^A at the elution volume from which the fraction was taken for the SDS-PAGE analysis. Color Prestained Protein Standard, Broad Range (New England Biolabs) was used as a marker. The experiment was performed in three independent replicas with identical results. (G) SDS-PAGE analysis of the MoaB2-CarD. Lines 4–6 contain selected fractions from (D) and (E). SDS-PAGE was performed under same conditions as in Panel F.

Fig 4

3D structure of M. smegmatis MoaB2. (A) Top view of the MoaB2 hexamer with the subunits shown in different colors. (B) Side view of the MoaB2 hexamer colored as in Panel A. (C) Detail of the interactions between monomeric subunits in the MoaB2 trimer. The hydrogen bond contacts at the interface (residues Val79, Thr80, and Pro81 from chain C and Arg95 from chain A) are shown, and the distances are specified in Å. (D) Detail of the trimer-trimer interface of MoaB2. The hydrogen bond and salt bridge contacts at the trimeric interface between Glu34, Glu38, Arg138, and Arg142 from chain E and chain C are shown.

Fig 5

M. smegmatis σ^A_N may be involved in the σ^A-MoaB2 interaction. (A) Schematic linear representation of σ^A, σ^A_Δ60aaN, σ^A_ΔN, and σ^B. The scale bar represents 100 amino acids (aa). σ^A is 466 aa long, σ^A_Δ60aaN is 401 aa long, σ^A_ΔN is 301 aa long, and σ^B is 319 aa long. (B). SDS-PAGE gel of IP of FLAG-tagged proteins σ^A (LK2073), σ^A_Δ60aaN (LK4207), and σ^A_ΔN (LK2463) using anti-FLAG antibody. The dotted line indicates where this panel was electronically assembled from two parts of one gel. Relevant proteins are indicated on the right side of the gel; blue arrows mark different σ^A variants; red arrows indicate MoaB2. No FLAG strain (LK3016) was used as negative control. The identity of the bands was determined by mass spectrometry. The experiment was performed in three biological replicates with identical results. EXP, exponential; STA, stationary phase. (C) Relative amounts of M. smegmatis MoaB2 bound to different constructs of σ^A in EXP and STA phase of growth, calculated from signal intensities of respective bands from three independent SDS-PAGE gels from three independent experiments. QuantityOne (Bio-Rad) software was used for quantification. The relative values were normalized to molecular weight of proteins. The relative amount of σ^A immunoprecipitated from stationary phase (STA) was set as 1. The bars show the average from three biological replicates, and the error bars show ±SD. P-values that are less than 0.001 are marked as *** (t-test). The vertical arrows in the chart indicate values “zero” as the values were even below background. (D) Comparisons of proteins associating with σ^A-FLAG (LK2073 over control LK3016) vs proteins associating with σ^A_ΔN-FLAG (LK2463 over control LK3016) as determined by quantitative LC-MS. Data from stationary phase (STA) are shown. Red spots indicate proteins significantly enriched (significance: −log₁₀ P < 2; enrichment: >log₂>2). The blue spot indicates MoaB2 that was not significantly enriched in σ^A_ΔN-FLAG (in σ^A-FLAG it was significant). Each spot is the average calculated from three independent experiments. For values, see Supplementary Table (Table S3).

Fig 6

CD and ¹H-¹⁵N HSQC NMR spectra of σ^A and in of σ^A with MoaB2. (A) CD spectra of σ_1.1 from B. subtilis and σ^A_N from M. smegmatis. The spectrum of B. subtilis σ_1.1 domain (residues 1–72 of σ_1.1 and additional eight residues of a His-tag) manifests a shape that is typical for well-ordered protein with high α-helical content, whereas the shape of M. smegmatis σ^A_N (residues 1–165 of σ^A_N preceded by a glycine) spectrum reveals a mostly disordered protein. (B) ¹H-¹⁵N HSQC NMR spectra of M. smegmatis σ^A_N (black contours) and full-length σ^A (light blue contours). The narrow dispersion of proton chemical shifts indicates a disordered protein since most of the peaks are clustered around the center of the measured spectrum and is not dispersed as is typical for proteins with well-defined structure. (C–F) Selected regions of ¹H-¹⁵N HSQC NMR spectra of M. smegmatis full-length σ^A. Spectra of free σ^A and σ^A with MoaB2 are shown in light blue and red, respectively. Peaks are labeled with single-letter symbols and residue numbers. (G) Plot of the combined chemical shift changes upon MoaB2 addition. (H) Plot of the relative heights of peaks of free σ^A (blue) and σ^A with MoaB2 (red).

Fig 7

M. smegmatis MoaB2 is not essential. (A) The essentiality of the M. smegmatis MoaB2-encoding gene was tested by a CRISPR-based depletion approach. Discs soaked with ATC were placed on agar media. ATC induced expression of sgRNA and dCas9 to deplete σ^A (mysA, MSMEG_2758, and LK2203) mRNA, moaB2 (MSMEG_5485 and LK2263) mRNA, and no-target negative control (LK2261). The genes that were targeted by sgRNAs are indicated above the dishes. The experiment was performed in three biological replicates with the same result. (B) RT-qPCR relative quantitation of three mRNAs (moaB2, mysA and rpoC) in the moaB2 CRISPR depletion strain (LK2263; gray bars) used in (A) compared to the control strain (LK2261; black bars, set as 1). The mRNA levels were also normalized to an external spike (RNA control introduced during the RNA extraction protocol). mysA: codes for σ^A; rpoC: codes for the β′ subunit of RNAP. The graph shows averages from three independent experiments, and error bars show ±SD. (C) Growth curves in the 7H9 rich medium of the control strain (control oligo, LK2261) and the strain with depleted moaB2 (LK2263). The graph shows averages from three independent experiments, and the error bars indicate ±SD. CRISPR-based depletion of MoaB2 was induced with ATC (100 ng/mL), which was added at the beginning of growth.

Fig 8

M. smegmatis MoaB2 modulates σ^A-dependent but not σ^B-dependent transcription in vitro. Multiple-round transcriptions were performed with RNAP (LK1853) reconstituted with σ^A (LK2832) at the 1:5 and 1:20 ratios or σ^B (LK1248) at the 1:5 ratio in the absence or presence of increasing amounts of MoaB2 (LK2936) and at the presence of CarD (LK3209) and RbpA (LK3210; indicated below the graph). Representative primary data (full gels are shown in Fig. S13B) with lane numbers are shown above the graphs (M, MoaB2). All samples were run on 7% polyacrylamide gel (for more details see Materials and Methods). The stochiometric amounts of MoaB2 refer to monomers. As promoter, the M. smegmatis rRNA promoter PrrnAPCL1 was used in all panels (LK1548; full sequence is shown in Fig. S13C). All graphs show averages from three independent experiments, and the error bars indicate ±SD. Representative gel with transcription from the vector containing PrrnAPCL1 and the “empty” vector (LK2385) is shown in Fig. S13A, demonstrating the identity of the transcript.

Fig 9

Model of interplay between MoaB2, σ^A, and RNAP in the mycobacterial cell. A model of functional interactions between MoaB2, σ^A, and RNAP is shown. Binding of MoaB2 to σ^A occurs at a ratio of 1:1 (one chain of each) and has the potential to decrease the available pool of σ^A, likely modulating transcription by competing with the RNAP core for σ^A. σ^A bound to MoaB2 is not able to bind to RNAP. MoaB2 by interacting with σ^A may positively affect its stability.