Protein kinase B controls Mycobacterium tuberculosis growth via phosphorylation of the transcriptional regulator Lsr2 at threonine 112.

Mycobacterium tuberculosis (Mtb) is able to persist in the body through months of multi-drug therapy. Mycobacteria possess a wide range of regulatory proteins, including the protein kinase B (PknB) which controls peptidoglycan biosynthesis during growth. Here, we observed that depletion of PknB resulted in specific transcriptional changes that are likely caused by reduced phosphorylation of the H-NS-like regulator Lsr2 at threonine 112. The activity of PknB towards this phosphosite was confirmed with purified proteins, and this site was required for adaptation of Mtb to hypoxic conditions, and growth on solid media. Like H-NS, Lsr2 binds DNA in sequence-dependent and non-specific modes. PknB phosphorylation of Lsr2 reduced DNA binding, measured by fluorescence anisotropy and electrophoretic mobility shift assays, and our NMR structure of phosphomimetic T112D Lsr2 suggests that this may be due to increased dynamics of the DNA-binding domain. Conversely, the phosphoablative T112A Lsr2 had increased binding to certain DNA sites in ChIP-sequencing, and Mtb containing this variant showed transcriptional changes that correspond with the change in DNA binding. In summary, PknB controls Mtb growth and adaptations to the changing host environment by phosphorylating the global transcriptional regulator Lsr2.

Introduction

Mycobacterium tuberculosis (Mtb) is a slow‐growing bacterium that can replicate in humans and cause tuberculosis. The pathogen is able to rapidly shut down its growth to persist in non‐replicating states in infected individuals, which can be modelled in the laboratory (Wayne and Sohaskey, 2001). Mtb adaptation to non‐permissive conditions is accompanied by dramatic changes in global protein phosphorylation but the importance of these modifications is poorly defined (Prisic et al., 2010). Mtb has 11 serine/threonine protein kinases and they play significant roles in growth, virulence and metabolism (Richard‐Greenblatt and Av‐Gay, 2017). In particular, protein kinase B (PknB) is reported to be essential for Mtb growth (Fernandez et al., 2006; Forti et al., 2009) due to its critical function in the regulation of peptidoglycan biosynthesis (Gee et al., 2012; Boutte et al., 2016; Turapov et al., 2018). It is also important for Mtb survival in hypoxic conditions and resuscitation during reaeration (Ortega et al., 2014). However, the molecular mechanism for PknB‐mediated adaptation to hypoxia is unknown.

We have recently shown that PknB‐depleted Mtb can grow in osmoprotective sucrose magnesium medium (SMM) (Turapov et al., 2018). Comparative phosphoproteomic analysis of PknB‐producing against PknB‐depleted mycobacteria revealed substantial changes. Specifically, the transcriptional regulator Lsr2 showed increased phosphorylation in PknB‐producing mycobacteria, indicating that this protein may be a PknB substrate.

Lsr2 is a DNA‐binding protein that combines the properties of a nucleoid‐associated protein (Kriel et al., 2018) and a global transcriptional regulator (Bartek et al., 2014). Lsr2 has over 1000 binding sites in Mtb (Gordon et al., 2010; Minch et al., 2015). The precise role of Lsr2 in mycobacterial biology remains unclear, nevertheless parallels may be drawn with similar proteins from other bacteria. Lsr2 represents the first example of an H‐NS‐like protein identified outside Gram‐negative bacteria; moreover, lsr2 was able to complement an hns mutant in Escherichia coli (Gordon et al., 2008). Similar to the H‐NS proteins, Lsr2 has been proposed to bind to the minor groove of DNA (Gordon et al., 2011) and to possess DNA bridging properties (Chen et al., 2008). Additionally, Lsr2 has been shown to protect DNA from reactive oxygen species, and overexpression of Lsr2 improved survival of mycobacteria treated with hydrogen peroxide (Colangeli et al., 2009). Deletion of lsr2 in Mtb resulted in severe growth impairment on solid media, defects in persistence and adaptation to changing oxygen levels, all of which were accompanied by differential expression of genes involved in cell wall remodelling, respiration and lipid biosynthesis (Bartek et al., 2014).

Here, we profiled the transcriptional changes that resulted from PknB depletion, and investigated the role of Lsr2 in coordinating these changes, as suggested by reduced phosphorylation of Lsr2 at T112 during PknB depletion (Turapov et al., 2018). We probed the role of phosphorylation at this site in regulation of the structure and DNA‐binding properties of Lsr2 and in governing growth and survival of Mtb in different conditions. Based on our data, we propose that PknB‐mediated phosphorylation controls Lsr2 binding to DNA in Mtb, providing a functional link between serine/threonine protein kinase signalling in replicating bacilli and regulatory networks that enable Mtb to survive dynamic environments during infection.

Results

Transcriptome profiling of PknB‐depleted Mtb revealed an Lsr2‐regulated gene expression signature

PknB is essential for growth in standard conditions; however, we have recently developed an osmoprotective medium (SMM) that supported growth of PknB‐depleted Mtb and allowed us to identify PknB substrates (Turapov et al., 2018). Using the same system, we compared the transcriptional profile of PknB‐depleted versus PknB‐producing Mtb (Fig. 1A, Tables 1 and S1). PknB‐depletion led to specific and significant changes in gene expression: 65 genes were induced and 34 repressed (Fig. 1B, Table S1). Two functional classes were overrepresented amongst the induced genes compared to the genome as a whole: regulatory proteins and proteins involved in lipid metabolism. The induced genes annotated as transcriptional regulators were csoR, rv1129c, rv1460, rv2017, rv2250c, rv3334, sigB, whiB3 and whiB6 (Table S1). These transcription factors regulate copper homoeostasis (CsoR) (Marcus et al., 2016), iron–sulphur cluster biogenesis (Rv1460) (Willemse et al., 2018), cholesterol catabolism (Rv1129c/PrpR) (Masiewicz et al., 2012), the enduring hypoxic response (Rv3334) (Rustad et al., 2008), multiple stress responses (SigB) (Lee et al., 2008), redox stress and complex lipid biosynthesis (WhiB3) (Mehta and Singh, 2018) and virulence factor expression (WhiB6) (Bosserman et al., 2017).

Figure 1

Application of omics tools to characterise the function of pknB essential in standard growth media without affecting Mtb viability.

A. Experimental set‐up for sample preparation and analysis. Conditional PknB mutant (pknB‐CM) was grown in sucrose‐magnesium medium with pristinamycin (SMMpri, activation of PknB expression) or without pristinamycin (SMM, PknB depletion). Phosphoproteomics analysis was previously described (Turapov et al., 2018).

B. PknB depletion results in alteration of Mtb transcriptome. Transcriptional impact of PknB depletion in Mtb. Volcano plot showing 65 genes significantly induced (red) and 34 genes repressed (blue) by PknB‐depletion in replicating Mtb. Significantly differentially expressed genes were identified using a moderated t‐test (P‐value < 0.05 with Benjamini and Hochberg multiple testing correction) and fold change >1.8 from three biological replicates. Fold change (log2) comparing Mtb pknB‐CM in SMM with and without pristinamycin is plotted on the x‐axis; corrected p‐value (−log10) on the y‐axis. Application of TFOE tool Lsr2 as potential regulator of the observed gene expression patterns Lsr2.

C. Expression of selected targets was validated using qRT‐PCR and proteomics approaches (see for detail Experimental procedures). Expression of pknB, nuoB, hsp, csoR was significantly different in SMM compared with SMMpri (P < 0.01, t‐test).

Table 1

The impact of PknB depletion on the gene and protein expression levels of serine/threonine protein kinases and Lsr2 in Mtb.

Gene	Protein	Description	Transcriptomics	Proteomics
			Fold difference SMMpri vs SMMa	Fold difference SMMpri vs SMM
Rv3597c	Lsr2	Protein Lsr2	1.01	1.20
Rv0015c	PknA	Protein kinase PknA	1.27	0.80
Rv0014c	PknB	Protein kinase PknB	5.97	6.60
Rv0931c	PknD	Protein kinase PknD	1.02	0.70
Rv1743	PknE	Protein kinase PknE	1.04	1.00
Rv1746	PknF	Protein kinase PknF	0.77	0.40
Rv0410c	PknG	Protein kinase PknG	0.93	1.30
Rv1266c	PknH	Protein kinase PknH	1.02	0.70
Rv2914c	PknI	Protein kinase PknI	0.84	N/D
Rv2088	PknJ	Protein kinase PknJ	0.89	N/D
Rv3080c	PknK	Protein kinase PknK	1.07	N/D
Rv2176	PknL	Protein kinase PknL	0.90	N/D

Fold change values, comparing Mtb pknB‐CM grown in sucrose magnesium medium with (SMMpri) or without pristinamycin (SMM) are derived from microarray and proteomics profiles.

The transcriptional signature of PknB depletion resembled features of intracellular growth (Table S1), with a significant overlap with RNA profiles from several studies of Mtb in macrophages as reflected by hypergeometric probability values: 6.7 × 10−23 (Tailleux et al., 2008), 7.34 × 10−18 (Schnappinger et al., 2003) and 3.57 × 10−17 (Rohde et al., 2007). For example, there was induction of pathways involved in mycobactin synthesis (mbtB/C/D), complex lipid phthiocerol dimycocerosate (PDIM) biosynthesis (fadD26, ppsA/B/C/D), metabolism of alternative lipid carbon sources, the glyoxylate shunt (icl), the methylcitrate cycle (prpD/C, prpR) and triacylglycerol synthase (tgs1). The isoniazid inducible genes (iniB/A/C) that respond to cell wall stress (Colangeli et al., 2007), and four of the nine genes coding for alternative ribosomal proteins, rpmB1, rpmB2, rpmG1, rpsN2 (Prisic et al., 2015) were also induced.

The 34 genes that were significantly repressed in PknB‐depleted bacteria included pknB itself (sixfold change, whereas no other protein kinases were significantly changed Table 1); nuoA/B/C, encoding subunits of NADH dehydrogenase I, which is part of the aerobic respiratory chain, and several genes involved in intermediary metabolism (Table S1). Comparison of gene expression and protein abundance of selected targets showed good agreement (Fig. 1C). Overall the number of differentially expressed genes was comparable to the number with differential expression when other regulators were similarly disrupted, for example DosR (Park et al., 2003). This is in contrast to the large‐scale changes in gene expression after treatment with an inhibitor of PknB and PknA (Carette et al., 2018), which would likely impact Mtb viability. In summary, PknB depletion in replicating bacteria resulted in co‐ordinated changes to the transcriptome with similarities to intracellular adaptations, suggesting that PknB may control the induction of alternative gene regulatory pathways.

Application of the Transcription Factor Over‐Expression (TFOE) output tool (Rustad et al., 2014) predicted Rv0081 (Galagan et al., 2013), DosR (Park et al., 2003) and Lsr2 (Bartek et al., 2014) as potential regulators of the observed gene expression patterns (Fig. 1, Table S1). We next focussed on the involvement of Lsr2 in PknB‐mediated transcriptional adaptation, since Lsr2 was identified as a putative PknB substrate in our earlier phosphoproteomic work (Turapov et al., 2018): PknB depletion decreased Lsr2 phosphorylation 2.54‐fold without impacting Lsr2 protein expression levels (Turapov et al., 2018; Table 1). By contrast, the phosphorylation of DosR was unchanged during PknB depletion, and there are no reports of regulation of Rv0081 by phosphorylation.

PknB phosphorylated Lsr2 in vitro

Kinase assays of purified PknB kinase domain with Lsr2 used anti‐phosphothreonine antibody to detect phosphorylation and demonstrated that Lsr2 was directly phosphorylated by PknB (Fig. 2A). Interestingly, phosphorylation resulted in a marked change in Lsr2 protein mobility in SDS‐PAGE (Fig. 2B) and generated several bands, indicative of multiple phosphorylated forms. Mass spectrometry confirmed the previously observed phosphosite on threonine 112 (Turapov et al., 2018) and detected additional phosphorylations at threonine 8, threonine 22 and threonine 31 (Fig. 2C and D).

Figure 2

Identification of Lsr2 as a substrate of PknB.

A. Recombinant Lsr2 was phosphorylated by recombinant PknBKD. Phosphorylated proteins were detected by western blot using a phospho‐threonine antibody. Recombinant CwlM was used as positive control. 1 – protein markers; 2 – Lsr2 incubated with PknB and ATP; 3 – Lsr2 incubated with PknB without ATP; 4 – CwlM incubated with PknB and ATP; 5 – CwlM incubated with PknB without ATP.

B. SDS‐PAGE revealed a shift in Lsr2 mobility upon phosphorylation (lanes identical to panel A).

C. Schematic presentation of phosphosites identified in phopshoproteomics studies (top) (Turapov et al., 2018), and in vitro (bottom).

D. Phosphopeptides detected by mass spectrometry; phosphorylated residues shown in bold font.

Phosphosite threonine 112 was necessary for Lsr2 function in Mtb

The functional importance of the identified phosphorylation sites in Lsr2 was further investigated by constructing a panel of phosphoablative lsr2 variants using the pMV306 plasmid that integrates at attB site of Mtb chromosome (Table S2) and measuring their ability to complement the phenotypic changes caused by lsr2 deletion in Mtb. Lsr2 deletion mutant containing the empty pMV306 plasmid (Δlsr2 pMV306) was used as a control. Lsr2 deletion significantly impaired Mtb growth on solid media, similarly to a previous study (Bartek et al., 2014). Expression of wild‐type lsr2 allele at attB site (the resultant strain designated as Δlsr2 WT) fully complemented the defect (Figs 3A and S1A). Lsr2 deletion mutant expressing phosphoablative T112A Lsr2 variant (Δlsr2 T112A) had impaired growth on solid media (Figs 3A and S1A), whereas all other phosphoablative variants (T8A, T22A, T31A) complemented growth fully (Fig. 3A). Growth of lsr2 deletion mutant expressing a T112D phosphomimetic variant of Lsr2 (Δlsr2 T112D) was indistinguishable from growth of wild‐type Mtb (Fig. S1A). Notably, Δlsr2 pMV306 and Δlsr2 T112A transformants were recovered from liquid medium, since these strains failed to produce colonies on solid media unlike transformation with plasmids carrying wild‐type lsr2 and other variants in Fig. 3.

Figure 3

Phosphoablative T112A variant does not complement growth and survival defect of lsr2 deletion mutant.

A. Lsr2 phosphoablative mutants were serially diluted and plated on 7H10 agar. Growth of lsr2 deletion mutant expressing wild‐type, T8A, T22A, T31A and T112A variants was compared with growth of the deletion mutant containing the empty vector on 7H10 agar. Experiment was repeated using two biological replicates.

B. Mtb Δlsr2, Δlsr2, Δlsr2 and Δlsr2 (~6 × 105 cells/ml of each strain) were inoculated in 7H9 liquid medium supplemented with ADC and Tween 80 and incubated at 37ºC without shaking.

C. Expression of Lsr2 and Lsr2 variants from pMV306 was verified by western blotting of Mtb lysates using an anti‐Lsr2 antibody. Expression of GroEL was used as loading control.

D. T112A mutation impairs Mtb survival in the Wayne model of non‐replicating persistence. MtbLsr2 mutants were incubated in sealed tubes with gentle mixing for up to 24 weeks. B and D Data presented as mean ± SEM (N = 6, two independent experiments done with biological triplicates). ***Statistically different in Δlsr2 or Δlsr2 compared with Δlsr2 and Δlsr2 .

Comparison of the growth rates in liquid 7H9 medium (Fig. 3B) revealed a similar pattern: the phosphomimetic T112D variant grew at a similar rate to the strain with wild‐type Lsr2 (Δlsr2 T112D 0.039 ± 0.009 h−1 and Δlsr2 WT 0.041 ± 0.004 h−1), whereas the phosphoablative T112A variant had a similar defect to the plasmid control (Δlsr2 T112A 0.017 ± 0.004 h−1 and Δlsr2 pMV306 0.016 ± 0.03 h−1). This failure of Lsr2 T112A to complement the growth defect of Δlsr2 in liquid or solid medium, despite expression at similar levels (Fig. 3C), indicates a requirement for phosphorylatable threonine at this position for the function of Lsr2 in growing Mtb, suggesting a possible mechanism by which PknB might regulate transcription via phosphorylation of Lsr2. The phosphomimetic T112D Lsr2 had reduced mobility in SDS PAGE, similar to phosphorylated Lsr2 (Figs 3C and 2B).

In separate experiments, we investigated whether expression of T112D Lsr2 from pMV306 would complement the growth defect of pknB‐CM on solid SMM agar. However, introduction of pMV306::lsr2 WT or pMV306::lsr2 T112D did not improve growth of pknB‐CM; moreover pMV306::lsr2 T112A failed to produce any transformants, suggesting that T112A Lsr2 was toxic for pknB‐CM.

Lsr2 phosphorylation at threonine 112 is crucial for Mtb adaptation to hypoxic conditions but not in prolonged stationary phase

We hypothesised that regulation of Lsr2 by PknB could account for the defects in survival of oxygen depletion reported when pknB and lsr2 function was abrogated (Bartek et al., 2014; Ortega et al., 2014). Thus, we assessed survival of Mtb carrying Lsr2 variants using the Wayne model of non‐replicating persistence in hypoxia (Wayne and Sramek, 1994). The viable counts of Δlsr2 WT and Δlsr2 T112D increased after 7 weeks of incubation, proving that Mtb grew during gradual depletion of oxygen. In opposite, Δlsr2 pMV306 showed a dramatic drop of CFU counts below the initial inoculum. Further incubation for 24 weeks resulted in near‐complete loss of Δlsr2 pMV306 bacteria but did not significantly alter the survival of Δlsr2 WT and Δlsr2 T112D. Interestingly, T112A Lsr2 was apparently unable to grow during initial 7 weeks, suggesting a requirement for T112 phosphorylation during adaptation for decreasing oxygen levels.

We also assessed the survival of Mtb expressing Lsr2 variants in late stationary phase of aerobically grown cultures by MPN and CFU counting. Δlsr2 pMV306 showed impaired survival compared to Δlsr2 WT, resulting in a 1.5 order of magnitude difference in viable counts. In this model, the survival of T112A or T112D variants (Δlsr2 T112A, Δlsr2 T112D) was not significantly different from wild‐type Lsr2 (Fig. S1B). Our results suggest that phosphorylation of Lsr2 at T112 may be specifically required during adaptation to hypoxic conditions but not for survival in prolonged stationary phase.

Increased DNA binding in the T112A Lsr2 variant of Mtb‐altered gene expression

Phosphorylation of nucleoid‐associated proteins is known to influence their interaction with DNA (Dilweg and Dame, 2018), thus providing a mechanism by which PknB might regulate transcription via Lsr2. The DNA‐binding profile of Lsr2, and the influence of phosphorylation upon its DNA‐binding profile, were investigated using ChIP‐seq and a custom anti‐Lsr2 antibody. Lsr2‐binding peaks were found at intergenic spaces and running through several open reading frames, and were well conserved across biological replicates (Fig. S2 and Table S3), whereas no DNA was precipitated from strains lacking lsr2, confirming antibody specificity. The putative regulon of Lsr2, including genes with a Lsr2‐binding site within or immediately upstream of the coding sequence, was defined as 1178 genes (Table S3), and was consistent with previously identified Lsr2‐binding patterns (Gordon et al., 2010; Minch et al., 2015) (Table S4). The regulon was significantly enriched in genes that are differentially expressed upon inactivation of Lsr2 (Bartek et al., 2014), macrophage infection (Tailleux et al., 2008), in sputum (Garton et al., 2008) and under acid‐nitrosative stress (Cossu et al., 2013; Table S4), overlapping with the transcriptional signature of PknB depletion.

To measure the effect of phosphorylation of Lsr2 on its DNA‐binding pattern, we compared the ChIP‐seq data above with parallel experiments using phosphomimetic Lsr2 (Δlsr2 T112D) and phosphoablative Lsr2 (Δlsr2 T112A). The results for phosphomimetic strain were essentially the same as those with wild‐type Lsr2 (Δlsr2 WT Mtb, no significant differences in sequence abundance), whereas DNA precipitated from the phosphoablative strain showed a significant increase in the abundance of 226 putative Lsr2‐binding sites (Fig. S2), suggesting that greater DNA binding of this variant. These binding sites may affect the expression of 94 genes (Table S5), which are involved in pathways of cell wall biosynthesis, lipid metabolism, PE/PPE protein synthesis and intermediary metabolism. Application of Motif‐based sequence analysis predicted a binding site where T122A variant was preferably binding (Fig. 4A).

Figure 4

T112A mutation alters Lsr2 binding to DNA and gene expression patterns.

A. Lsr2 T112A binding site was predicted using the MEME Suite (://meme-suite.org) and sequence of 25 DNA fragments which were >6‐fold enriched in Lsr2 T112A compared with Lsr2 WT.

B. Representative plots describe Lsr2 binding upstream of rpfA, intragenic binding in leuS, or both intergenic and intragenic binding in sodA, showing greater Lsr2 binding in three biological replicates of phosphoablative Δlsr2 (green) compared to Δlsr2 (yellow). A black asterisk marks the position of a Lsr2‐binding site in leuS. Plots adapted from Integrative Genomics Viewer IGV, (Robinson et al., 2011). Binding patterns in Δlsr2 were identical to those in Δlsr2 and not shown for clarity.

C. Expression of leuS, rpfA, rpfC, sigA and sodA relative to Δlsr2measured by quantitative RT‐PCR in Δlsr2 and Δlsr2 and normalised to 16s rRNA and Δlsr2. Data presented as mean ± SEM (N = 6). **Statistically different in Δlsr2 compared with Δlsr2 (P < 0.01).

To determine whether altered binding of Lsr2 variants resulted in corresponding changes in RNA abundance, we tested two of these genes for differential expression in the presence/absence of Lsr2 and in the presence of wild‐type versus phosphoablative Lsr2. sodA and leuS were selected for investigation on the basis of their divergent patterns of Lsr2 binding (Fig. 4B). sodA was upregulated but leuS did not change significantly in Δlsr2 WT compared with Δlsr2 (Fig. 4C). Both genes were differentially expressed in Δlsr2 versus Δlsr2 WT, but leuS expression was lower, while sodA expression was higher (Fig. 4C). These results suggest that the influence of phosphorylation on Lsr2 DNA binding can up‐ or downregulate gene expression, depending on promoter structure and other regulatory factors. Genes known to be differentially expressed upon lsr2 disruption were included alongside (rpfA and rpfC) (Bartek et al., 2014), as well as the principle sigma factor gene sigA, which has a putative Lsr2‐binding site in its promoter (Rustad et al., 2014). However, expression of these genes did not significantly change in Δlsr2 versus Δlsr2 WT in accordance with our ChIP‐sequencing data.

PknB‐mediated phosphorylation of Lsr2 or phosphomimetic mutation of Lsr2 reduced its DNA binding in vitro

To determine how phosphorylation of Lsr2 affects the affinity and sequence specificity of DNA binding, we performed electrophoretic mobility shift assays (EMSA) with a range of DNA fragments containing putative Lsr2‐binding sites. These fragments were obtained by PCR or by annealing pairs of oligonucleotides to produce double‐stranded DNA (Table S2). Lsr2 reduced the mobility of all fragments when added at concentrations above 1.9 μM and there was no apparent difference in affinity between the DNA fragments (Fig. S3). The same pattern was observed when a shorter double‐stranded DNA containing a putative Lsr2‐binding site within the leuS gene (Reddy et al., 2009) was used (Fig. 5A). We tested various DNA fragments, including a mutated leuS site (Fig. 5B), and all were shifted by Lsr2 regardless of their sequence, demonstrating that Lsr2 binds non‐specifically to DNA, as previously suggested (Colangeli et al., 2007).

Figure 5

Phosphorylatied Lsr2 or its phosphomimetic T112D variant do not bind DNA. A. Lsr2 was mixed with leuS fragment containing a putative binding site (AATTCGGCAAAATCGGTAAG), position of which is marked with an asterisk in Fig. 4A.

B. Lsr2 was mixed with the mutated leuS MUT fragment (AACTCGGCGAGGTCGGTCAG).

C. Lsr2 was phosphorylated by PknB and mixed with leuS‐binding site.

D. Lsr2 T112D variant was mixed with leuS fragment. Lsr2 was added to DNA at a range of concentrations (0.95–7.6 μM). Representative results from three independent experiments shown.

E. Quantification of Lsr2BD (DNA‐binding domain) interaction with DNA by fluorescence anisotropy. Titration of 5’ Alexa Fluor 488 double‐stranded DNA (CGCGCATATATGCG) (4 nM) by Lsr2BD WT (open circles), Lsr2BD T112A (black circles) and Lsr2BD T112D (black diamonds).

F. control experiments with GC‐rich double‐stranded DNA fragment (AACTCGGCGAGGTCGGTCAG).

We next assessed the effect of Lsr2 phosphorylation on binding to these DNA fragments. As shown in Fig. 5C, phosphorylation of Lsr2 by PknB completely abolished binding (Figs 5C and S3), whereas Lsr2 T112D variant showed reduced DNA binding (Fig. 5D). This recalls the ChIP‐sequencing results where Lsr2 T112D and wild‐type Lsr2 (that is phosphorylated in cells), precipitated less DNA than phosphoablative T112A Lsr2.

In vitro phosphorylation of Lsr2 resulted in phosphorylation of both domains. Like Lsr2, HN‐S proteins also consist of an oligomerisation domain and DNA‐binding domain. In the HN‐S family, the two domains have distinct roles: shaping the nucleoid and regulating gene expression respectively (Winardhi et al., 2015). In the case of Lsr2, threonine 112 in the DNA‐binding domain was required for function in Mtb, whereas other putative phosphorylation sites in the oligomerisation domain were mutated without loss of function (Fig. 3). To discriminate between the effects of phosphorylation on these two domains, we purified the DNA‐binding domain (Lsr2BD) to compare DNA binding of phosphorylated and unphosphorylated protein. A fluorescence anisotropy approach was used, since truncated Lsr2BD was not suitable for EMSA.

Previous studies demonstrated that Lsr2BD mainly recognised AT‐rich DNA sequences that formed a hook‐like structure (Gordon et al, 2011). We measured to ability of Lsr2BD and T112 mutants to bind an Alexa 488N labelled AT‐rich double‐stranded DNA fragment (5′‐CGCGCATATATGCG‐3′) (Fig. 5E). At pH 7.5 the Kd value for Lsr2BD was 0.21 ± 0.06 µM, that for Lsr2BDT112A was 0.02 ± 0.008 µM and Lsr2BD T112D showed no significant DNA binding (Kd could not be determined). In control experiments, Lsr2BD or Lsr2BDT112D showed no binding to a GC‐rich DNA sequence (Fig. 5F), confirming Lsr2BD preferential binding to AT‐rich DNA. These results show that the T112D mutation, mimicking the Lsr2 phosphorylated state, reduced Lsr2BDT112D binding to DNA in vitro. In summary, phosphorylation controls both Lsr2 and Lsr2BD binding to DNA.

Phosphomimetic T112D variant changed the conformation entropy of the Lsr2 DNA‐binding domain

To elucidate a molecular basis for the reduced binding of phosphomimetic Lsr2 to DNA we compared nuclear magnetic resonance (NMR) structures of Lsr2BDWT and Lsr2BDT112D. In accordance with the previously published structure (Gordon et al., 2010), our data confirmed that Lsr2 BD consists of two perpendicular α‐helices (α1, residues 78–89; α2, residues 102–112) linked by a long loop (residues 90–101). The two major components involved in Lsr2 to DNA binding are residues Arg97–Gly98–Arg99 that are inserted into the minor grove of DNA and Arg77, Ser 80, Arg84 and Ser95 that interact with the phosphate‐sugar backbone of the minor grove (Gordon et al., 2010). This organisation was preserved in Lsr2BDT112D (Table S6). However, T112D mutation resulted in a shorter α2 helix (Fig. S4), which ended with alanine 111 in Lsr2BDT112D compared with threonine 112 in Lsr2BDWT (Fig. 6A and B). Both variants were monomeric and the N‐terminal segment (residues 66‐75) upstream of the DNA‐binding domain was disordered. We also found that in Lsr2BD the methyl group of threonine 112 interacted with tyrosine 108, while the hydroxyl group of threonine 112 interacted with tryptophan 86 (Fig. 6C). These interactions did not form in Lsr2BDT112D, likely accounting for the shorter helix. Lsr2BD T112A, which must lack both interactions, bound DNA in our anisotropy experiments (Fig. 5E), demonstrating that the interactions themselves are not required for DNA binding, while suggesting that changes in conformation or dynamics related to the shorter helix could account for changes in DNA binding.

Figure 6

Solution structure of Lsr2BD (A) and Lsr2BDT112D (B). Superimposition of the 20 best calculated structures in cartoon representation with the last residues represented as sticks (A and B). α1 helix and linker region are involved in DNA binding; the structure of these regions is not affected by T112D mutation. In Lsr2BD, the threonine T112 side chain interacts with both tyrosine Y108 (3.7 Å) and tryptophan W86 (3.0 Å), all represented in sticks (C).

15N heteronuclear NMR relaxation analysis was performed to assess the dynamic behaviour of the two proteins (Fig. S5). In both isoforms, two α helices showed similar amplitudes for internal motion. However, in Lsr2BD mainly the C‐terminal helix was affected, while in Lsr2BDT112D the internal motion was extended to the N‐terminal helix. A previous study on the catabolite activator protein (CAP) demonstrated that different protein mutants with the same structure of interaction interface displayed very different affinity for their target DNA (Tzeng and Kalodimos, 2012). The authors showed that changes of the binding affinity were linked to fast internal dynamics (conformation entropy). Similarly, the T112D mutation and, presumably, phosphorylation of threonine 112, resulted in a shorter α2 helix and a more mobile loop that increased the Lsr2BD dynamics and impaired DNA binding.

Discussion

Mtb can subvert the immune system to survive in the host for many years. This remarkable ability is determined by mechanisms that allow Mtb to respond to multiple environments and adjust metabolic activity and cell division. One of these regulatory systems is protein phosphorylation involving 11 serine/threonine kinases, including protein kinase B. PknB is expressed during active replication (Kang et al., 2005) and is essential for growth in standard media (Fernandez et al., 2006; Forti et al., 2009). PknB has been implicated in the regulation of peptidoglycan biosynthesis (Bellinzoni et al., 2019); its main substrate CwlM stimulates biosynthesis of peptidoglycan precursors (Boutte et al., 2016) and may facilitate their transport to the cell surface (Turapov et al., 2018). This regulation is crucial for adjusting bacterial growth and synthesis of the cell wall. The external PASTA domain of PknB (Barthe et al., 2010; Prigozhin et al., 2016) is essential for PknB‐mediated signalling and its disruption results in bacterial death and alteration of antimicrobial susceptibility (Chawla et al., 2014; Turapov et al., 2015; Turapov et al., 2018). Furthermore, PknB can phosphorylate sigma factor SigH and its cognate anti‐sigma factor RshA; phosphorylation disrupts interaction of these two proteins and result in increased expression of the SigH regulon (Park et al., 2008).

Here, we present data demonstrating that PknB also controls global gene expression via another substrate, the DNA‐binding protein Lsr2. Transcriptomic analysis of PknB‐depleted Mtb suggest that PknB phosphorylation may silence alternative pathways, which are not important for logarithmic growth but which may be critical for mediating stress responses and virulence. These include enzymes involved in alternative metabolic pathways, synthesis of complex lipids, regulators of stress responses and antimicrobial tolerance (Fig. 1, Table S1). PknB depletion resulted in the altered expression of several transcriptional regulators and the regulons of Rv0081, DosR and Lsr2. However only Lsr2 was more phosphorylated in PknB‐producing Mtb compared with PknB‐depleted Mtb. According to our transcriptomic and proteomic signatures, none of the other serine/threonine protein kinases were upregulated in these conditions (Table 1), suggesting that PknB was responsible for Lsr2 phosphorylation in growing Mtb.

We demonstrate that Lsr2 could be phosphorylated by PknB at several threonines but only the T112 was essential for growth and survival of Mtb. The phosphoablative T112A mutant could not complement the growth defect of the Mtb lsr2 deletion mutant and was impaired in the Wayne hypoxia model. To investigate the molecular mechanism responsible for the growth defect in Δlsr2 T112A we conducted ChIP‐Seq analysis and compared DNA‐binding patterns in Δlsr2 T112A, Δlsr2 T112D and Δlsr2 WT. In agreement with previously published data (Gordon et al., 2010; Minch et al., 2015) we detected multiple occurrences of DNA binding in all three Lsr2 backgrounds supporting both nucleoid shaping and gene regulatory functions of Lsr2. However, phosphoablative Lsr2 T112A variant had increased DNA binding, potentially directly affecting expression of 94 genes. Interestingly, most of these genes were not essential for growth in vitro (DeJesus et al., 2017). Moreover, deletion of some of these genes has been previously shown to be advantageous for growth (DeJesus et al., 2017), including genes of unknown function (Rv0888, Rv1958, Rv1957), PE/PPE genes (Rv0878, Rv1983), aprA and genes controlling transport of PDIM (drrB and drrC). While products of these genes might be disadvantageous for growth in vitro, they likely play a critical role in adaptations to stress and virulence (Camacho et al., 2001; Sassetti and Rubin, 2003; Rohde et al., 2007).

Our data demonstrate that PknB phosphorylation of Lsr2 in vitro completely abolished DNA binding, while the phosphomimetic mutation reduced Lsr2–DNA interactions (Fig. 5). We have not investigated the effect of T8, T22 and T31 phosphorylations on DNA binding. Based on previously published data we hypothesise that phosphorylation of these sites in the oligomerisation domain might be important for controlling nucleoid shape and DNA bridging properties. The results of our study suggest that phosphorylation of T112 in the DNA‐binding domain controls interaction of Lsr2 with AT‐rich DNA sequences which modifies gene expression (Fig. 4C). Our structural studies further confirm that phosphomimetic T112D Lsr2 variant had a shorter C‐terminal helix and increased dynamics of the DNA‐binding domain, leading to impaired Lsr2–DNA binding (Figs 6 and S5).

Post‐translational modifications are common mechanisms for the regulation of DNA binding both in eukaryotes (Bannister and Kouzarides, 2011) and prokaryotes (Dilweg and Dame, 2018). Phosphorylation or nitrosylation of transcriptional regulators abolish DNA binding (Leiba et al., 2014; Smith et al., 2017). H‐NS protein, a homologue of Lsr2 in E. coli, has been shown to be acetylated and phosphorylated; however, the precise function of these modifications remains to be characterised (Dilweg and Dame, 2018). Our data show that phosphorylation of Lsr2 is important for Mtb growth and that this may be a key mechanism for controlling mycobacterial adaptations to permissive and non‐permissive environments. Thus, PknB mediates two critical components of mycobacterial growth, peptidoglycan biosynthesis and gene expression of alternative pathways.

Based on our data, we propose that PknB controls Mtb growth by phosphorylating Lsr2. Like other H‐NS‐like proteins Lsr2 plays a dual role in mycobacterial biology, it shapes and protects the nucleoid and it controls gene expression (Bartek et al., 2014; Kriel et al., 2018). However, unlike H‐NS proteins in Gram‐negative bacteria that mainly silence the expression of foreign DNA (Lucchini et al., 2006), Lsr2 regulates expression of genes that are essential for growth, virulence and adaptation (Bartek et al., 2014). Our study suggests that phosphorylation of T112 might be important for tuning gene expression during growth, and the dynamic change between phosphorylated and non‐phosphorylated Lsr2 may help to adjust transcriptional patterns according to growth conditions. Reduced T112 phosphorylation, for example during starvation, may increase Lsr2 binding and upregulate pathways that are critical for Mtb survival under these conditions. Our data suggest that PknB is the main serine/threonine kinase responsible for phosphorylation of Lsr2 at T112 during growth, however we cannot exclude that other kinases can phosphorylate Lsr2 at this or other sites under different conditions as it has previously been shown for other substrates (Baer et al., 2014).

While there are many outstanding questions on the precise mechanisms of PknB‐mediated regulation of gene expression and Lsr2 binding to DNA, our findings provide a functional link between serine/threonine protein kinase signalling and transcriptional regulatory pathways that enable Mtb to survive the varied environments encountered during infection.

Experimental procedures

Strains and media

Mtb H37Rv was grown in Middlebrook 7H9 (Becton, Dickinson and Company) liquid medium supplemented with 10% (v/v) Albumin–Dextrose Complex (ADC), 0.2% (v/v) glycerol and 0.1% (w/v) Tween 80 at 37°C with shaking or in SMM containing hygromycin with or without pristinamycin. SMM comprised of 0.3 M sucrose, 20 mM MgSO4, 0.1% Tween 80 (w/v), 10% (v/v) ADC in standard 7H9 broth. Bacterial growth was measured by absorbance at 580nm, or by colony‐forming unit (CFU) counting on 7H10 agar (Becton, Dickinson and Company), or by most probable number (MPN) counting using established protocols (Loraine et al., 2016) and the MPN calculator program (Jarvis et al., 2010). Escherichia coli OverExpress™ C41(DE3) and DH5α were grown in Lysogeny broth. For protein expression, E. coli was grown to mid‐log phase (OD600 0.6‐0.8) at 37°C with shaking at 200 rpm before adding 0.5 mM isopropyl β‐D‐1‐thiogalactopyranoside followed by incubation at 18°C overnight. Antibiotics were used at the following concentrations (μg/ml): pristinamycin, 0.5; kanamycin, 50; hygromycin, 50; ampicillin, 50. Wayne model of non‐replicating persistence was set up as previously described (Wayne and Sramek, 1994). CFU and MPN counts were determined at 0‐, 7‐ and 12‐week time points.

Genetic manipulations

Previously described pknB‐CM (Forti et al., 2009) and Δlsr2 mutants (Bartek et al., 2014) were used in this study. In Δlsr2 mutant, a gene fragment encoding the C‐terminal DNA‐binding domain (corresponding to 174–268 bp region) was deleted. The Rv3597 (lsr2) coding sequence with additional 200 bp upstream region containing the putative promoter was amplified from the MtbH37Rv genome using Platinum Taq‐HF polymerase (ThermoFisher) and cloned into the pMV306 plasmid that integrates at attB site of Mtb chromosome (primers in Table S2). Lsr2 Mtb variants were obtained using the GeneArt™ Site‐Directed Mutagenesis System (Thermo Fisher Scientific) according to the manufacturer's instructions. All constructs were sequenced before transformation into an Mtb lsr2 deletion mutant. Transformants were selected on 7H10 agar or in supplemented 7H9 containing kanamycin. For expression of Lsr2 proteins Mtb H37Rv lsr2‐coding sequence or shorter fragments were PCR amplified and cloned into pET15‐TEV and the resultant constructs were transformed into E. coli OverExpress™ C41 (DE3). Bacterial strains generated in this study are shown in Table S2.

Preparation of samples for proteomics and transcriptomics

Mtb pknB‐CM was grown to OD580nm ~0.7 in SMM with or without pristinamycin. For proteomics studies, washed pknB‐CM bacteria were resuspended in buffer containing 20 mM TrisCl, pH 7.5, 1 M NaCl, 8 M urea and proteinase/phosphatase inhibitors. After bead beating, lysates were cleared by centrifugation and filtration and treated using the FASP protocol, as described previously (Turapov et al., 2018; Iswahyudi et al., 2019). For transcriptomics analysis, bacterial cultures were incubated with four volumes of guanidine thiocynate (GTC) for 30 minutes prior centrifugation.

Quantitative label‐free proteomics analysis

Analysis was performed as previously described (Turapov et al., 2018; Iswahyudi et al., 2019). For quantification, all peptides of an identified protein were included and the total cumulative abundance was calculated by summing the abundance of all peptides allocated to the respective protein. Additionally, Scaffold Q+ (version Scaffold_4.3.4, Proteome Software Inc) was used for peptide and protein identifications as previously described (Turapov et al., 2015). Peptide probabilities from Mascot were assigned by the Scaffold Local FDR algorithm.

Transcriptomic analyses

RNA was extracted from three biological replicates using the GTC/Trizol method (Waddell and Butcher, 2010). Mtb RNA (2 μg) was enzymatically labelled with Cy3 fluorophore and hybridised to a Mtb complex microarray (ArrayExpress accession number A‐BUGS‐41) as previously described (Salina et al., 2014). For quantitative RT‐PCR, total RNA was isolated from triplicate Mtb cultures; cDNA was generated using Superscript Reverse Transcriptase II and mycobacterial genome‐directed primers (Rachman et al., 2006). qPCR was performed in a Corbett Rotor Gene 6000 real‐time thermocycler using Absolute qPCR SYBR Green mix and gene expression values were normalised to 16S rRNA expression. For qPCR three biological and two replicates were assessed.

ChIP‐Seq analysis

DNA–Lsr2 interactions in Δlsr2 T112A, Δlsr2 T112D, Δlsr2 WT were assayed using ChIP‐seq methods as previously described (Minch et al., 2015). Briefly, mid log‐phase Mtb H37Rv cultures (OD600 0.4–0.6) were crosslinked with 1% formaldehyde, followed by incubation with 125 mM glycine for 5 minutes at 37°C. The cells were mechanically lysed and then sonicated to produce 200–500 bp fragments. Input control samples were taken for each genotype before antibody was added to assess antibody specificity. Samples were immunoprecipitated using a polyclonal anti‐rabbit anti‐Lsr2 antibody and protein‐G agarose beads. The Lsr2 complexes were de‐crosslinked by heating at 65°C overnight and proteins removed by treatment with proteinase K (10 mg/ml) for 2 h at 55°C. The DNA samples were column‐purified (Qiagen) and the quality of purified IP‐Lsr2 DNA verified using the Qubit DNA HS quantification assay and Nanodrop spectrophotometer. Libraries were prepared and sequenced using Illumina HiSeq SE50, 20 million reads (Novogene, Hong Kong). Raw fastq files were aligned to the Mtb H37Rv (NC_000962.3) reference genome using bwa samse (Li and Durbin, 2009). MACS2 version 2.1.1.20160309 (Zhang et al., 2008) was used to compare each of the input controls to the immunoprecipitated samples, identifying Lsr2‐binding sites (callpeak) using default parameters but including ‘‐g 4.41e+06 –nomodel –extsize 147’ (Table S3). Differential peaks comparing Δlsr2 T112A to Δlsr2 WT were then identified using ‘macs2 bdgdiff’.

Purification of recombinant proteins and phosphorylation in vitro

Recombinant Lsr2 proteins were purified using immobilised metal affinity chromatography (Ni‐NTA agarose, Qiagen) and size exclusion chromatography. The recombinant catalytic domain of PknB was purified using Glutathione Sepharose 4B GST‐tagged protein purification resin (GE Healthcare). For phosphorylation recombinant Lsr2 (10 μM) was mixed with the recombinant catalytic domain of PknB (5 μM) in a kinase buffer (20 mM Tris–HCl, pH 8.0; 0.5 mM DTT; 10 mM MgCl2; 0.2 mM ATP) and incubated at 37°C for 1h. Phosphorylation was confirmed by western blot analysis. Phosphorylated residues were identified in trypsin‐digested proteins using LTQ‐Orbitrap‐Velos mass spectrometer.

Protein electrophoresis and western blotting

Proteins were separated on 4–20% gradient SERVA gels and transferred onto a nitrocellulose membrane using a Trans‐Blot® Turbo™ Transfer System (Bio‐Rad). SIGMAFAST™ BCIP®/NBT or SignalFire™ Elite ECL Reagent were used to visualise proteins on C‐DiGit Chemiluminescent Blot Scanner (LI‐COR Biosciences). The following antibodies were used: custom polyclonal antibody raised against Lsr2 in rabbit (Gemini Biosciences); monoclonal murine anti‐polyhistidine antibody (Sigma‐Aldrich); phospho‐threonine antibody (Cell Signaling Technology); monoclonal anti‐MtbGroEL2 (Rv0440), clone IT‐70 (BEIResources); mouse anti‐rabbit IgG antibody:alkaline phosphatase (Sigma‐Aldrich; anti‐mouse IgG (whole molecule:alkaline phosphatase antibody produced in rabbit (Sigma‐Aldrich), and anti‐rabbit IgG, HRP‐linked antibody (Cell Signaling Technology).

Electrophoretic mobility shift assay

Electrophoretic mobility shift assays were carried out with DNA fragments amplified by PCR or annealed oligonucleotides (see Table S2 for primer details). DNA (1.2 nM) was mixed with indicated amounts of Lsr2 in a total volume of 20µl reaction buffer containing (10 mM Tris–HCl, pH 7.5, 50 mM KCl, 1 mM DTT, 5 mM MgCl2, and 2.5% glycerol). The mixture was incubated for 30 min at room temperature followed by native polyacrylamide gel electrophoresis using 8% gels in 0.5 x Tris‐Borate‐EDTA buffer, pH 7.5 for 24 min at 120 V. The gels were stained with SYBR Safe DNA stain (Thermo Fisher Scientific) and visualised using a ChemiDoc system (Bio‐Rad).

Fluorescence anisotropy

Custom made 5′ Alexa Fluor 488 succinimidyl ester labelled oligonucleotide probe (sequence 5′‐CGCATATATGCGCG‐3) was purchased from Integrated DNA Technologies. Steady‐state fluorescence anisotropy‐binding titrations were performed on a Tecan Saphire II microplate reader, using a 470 nm LED for excitation and a monochromator set at 530 nm (bandwidth 20 nm) for emission.

Determination of solution structures and dynamics of Lsr2BD and Lsr2BDT112D

All 1H‐15N double‐resonance NMR experiments were performed at 20°C on Bruker Avance III spectrometers (700 or 800 MHz) using previously described methods (Barthe et al., 1999; Gordon et al., 2010). NMR samples of 0.5mM 15N‐labelled protein dissolved in 25mM sodium phosphate buffer (pH 6.8), 150mM NaCl with 10% D2O for the lock. 1H chemical shifts were directly referenced to the methyl resonance of DSS, while 15N chemical shifts were referenced indirectly to the absolute 15N/1H frequency ratio. All NMR spectra were processed with GIFA (Pons et al, 1996). Chemical shift assignments were made using standard NOESY, TOCSY experiments performed on the 15N‐labelled protein sample. NOE cross‐peaks identified on 3D [1H, 15N] NOESY‐HSQC (mixing time 160 ms) were assigned through automated NMR structure calculations with CYANA 2.1 (Güntert, 2004). Backbone ϕ and φ torsion angle constraints were obtained from a database search procedure on the basis of backbone (15N, HN, Hα) chemical shifts using TALOS+ (Shen et al., 2009). For each protein, a total of 200 three‐dimensional structures were generated using the torsion angle dynamics protocol of CYANA 2.1. The 20 best structures of each protein (based on the final target penalty function values) were minimised with CNS 1.2. All statistical parameters are summarised in (Table S6). Relaxation rate constant measurements were performed on a 0.5mM protein sample, at 18.8 T (800 MHz). The pulse sequences used to determine 15N RN(Nz) (R1), RN(Nxy) (R2), and 15N{1H} NOE values were similar to those described (Barthe et al., 1999). The 15N longitudinal relaxation rates (RN(Nz)) were obtained from 10 standard inversion recovery experiments, with relaxation delays ranging from 18 to 1206 ms. The 15N transverse relaxation experiments (RN(Nxy)) were obtained from 10 standard CPMG experiments, with relaxation delays ranging from 16 to 160 ms. Both series of experiments were acquired in two single interleaved matrices to ensure uniformity of the experimental conditions. Heteronuclear 15N{1H} NOE were determined from the ratio of two experiments, with and without saturation.

Statistical analysis

Calculation of the protein p‐values was performed on the sum of the normalised abundance across all runs using one‐way ANOVA. Significantly differentially expressed genes (Table S1) were identified using a moderated t‐test (P‐value < 0.05 with Benjamini and Hochberg multiple testing correction), and fold change> 1.8 in GeneSpring 14.5 (Agilent Technologies). Hypergeometric probability and TFOE analysis (Rustad et al., 2014) were used to identify significantly enriched signatures. An unpaired t‐test was performed to compare gene expression in Δlsr2 WT and Δlsr2 T112A. For evaluation of growth parameters and survival in the Wayne model, one‐way ANOVA (GraphPad Prism) was used, comparing Δlsr2 pMV or Δlsr2 T112A to Δlsr2 WT and Δlsr2 T112D.

Author contributions

Conceptualisation, OT, MCG, SJW and GVM; Methodology, ARB, PA, HJ, CR, AAW, MW; Investigation KA, OT, PB, HJ, ADV, GVM; Analysis, OT, AAW, KA, HJ, MCG, SJW, GVM; Resources, MW, ILB and MIV; Writing – Original Draft, GVM, MCG, SWJ; Writing – Review and Editing, GVM, SJW, MIV, HMO. Funding Acquisition, KA, MCG, SJW, GVM. Supervision, GVM, OT, HMO, MCG and SJW.

Conflict of interest

Paul Ajuh is the director and shareholder in Gemini Biosciences Limited, Liverpool, UK. Other authors declare no competing interests.

Click here for additional data file.

Click here for additional data file.

Click here for additional data file.

Click here for additional data file.

Click here for additional data file.