Serine/threonine protein kinase phosphorylation of DosR alters target gene transcription mechanics and regulates Mycobacterium tuberculosis response to nitric oxide stress

TL;DR

This study shows how protein kinases regulate a key bacterial system that helps Mycobacterium tuberculosis survive in the host by adjusting gene activity in response to stress.

Contribution

The study reveals that STPK phosphorylation of DosR modulates TCS function, adding a new regulatory layer in Mtb stress response.

Findings

STPK phosphorylation of DosR reduces its DNA binding and transcriptional activation ability.

An Mtb mutant lacking STPKs shows increased DosR regulon transcription at lower NO levels.

STPKs act as a regulatory mechanism to fine-tune TCS activation in response to environmental stress.

Abstract

Successful host colonization by bacterial pathogens requires appropriate response and adaptation to environmental signals encountered during infection, with two-component systems (TCSs) and serine/threonine protein kinases (STPKs) being two important signal transduction mechanisms. Mycobacterium tuberculosis (Mtb) possesses similar numbers of STPKs (11) and TCSs (12), but if and how these two regulatory systems coordinate to enable Mtb adaptation in response to key environmental cues remains poorly understood. Here, we identify extensive interactions between STPKs and TCSs, with a subset of STPKs demonstrating interactions with multiple TCS response regulators. STPK phosphorylation of purified DosR, the response regulator of the key nitric oxide (NO)/hypoxia-responsive TCS DosRS(T), decreased its binding to target promoter DNA and its ability to activate steady-state gene transcription, in marked contrast with the opposite phenotypes observed with the activated, phospho-aspartic acid form of DosR. Strikingly, a ΔSTPK Mtb mutant exhibited increased DosR regulon transcription at lower NO levels than wild type Mtb, illustrating how STPK phosphorylation of a TCS RR may act to restrict and fine-tune conditions in which activation occurs. Together, our results support a functional relationship between STPKs and TCSs, and shed light on the mechanisms underpinning STPK-TCS interplay.

Mycobacterium tuberculosis (Mtb) is the bacterium that causes tuberculosis, which remains the largest cause of death from an infectious disease globally. Successful host colonization by Mtb requires that the bacteria appropriately sense and respond to changes encountered in its local microenvironment throughout the course of infection. Here, we provide evidence for the interplay between two key signal transduction regulatory mechanisms – two-component systems (TCSs) and serine/threonine protein kinases (STPKs). Focusing on the DosRS(T) TCS that is crucial in the response of Mtb to the critical environmental signals of nitric oxide (NO) and hypoxia, we reveal that STPK phosphorylation of the purified DosR regulator decreases target gene promoter binding and the activation of steady-state transcription. Further, an Mtb mutant that was disrupted in an STPK that phosphorylates DosR exhibited increased DosR target gene expression at lower NO concentrations than wild type Mtb. These results indicate that STPK phosphorylation serves as an additional regulatory layer for TCSs, adjusting the DosR concentration range under which full activation of the TCS occurs.

Figures

Specificity of interactions between TCS RRs and STPKs.Interactions between the TCS RRs DosR (A), PrrA (B), PhoP (C), or KdpE (D) with the various STPKs (kinase domains only for all except PknG and PknK) were tested by M-PFC assay. The positive control (“+”) was M. smegmatis expressing the S. cerevisiae GCN4 dimerization domains fused to the F1,2 or F3 domains of the murine dihydrofolate reductase gene; the negative control (“-“) was M. smegmatis expressing the respective RR fused to the F1,2 domain and a F3 domain that was not fused to any Mtb gene. Data are representative of 3 independent experiments.

Changes in DosR phosphorylation status alter binding affinity to the promoter of its target gene hspX.Electrophoretic mobility shift assays (EMSAs) using purified recombinant C-terminally 6x-His-tagged DosR and IRDye 700-labeled probes for the hspX promoter are shown. A control with no protein (“NP”) added is shown for each gel. DosR was added at indicated concentrations for all other lanes. 40 fmoles of hspX promoter DNA was used in each reaction. EMSAs shown are as follows: (A) untreated DosR, (B) DosR incubated with 50 mM acetyl phosphate (AcP), (C) DosR phosphorylated “on-bead” with 1 µM PknH, (D) DosR phosphorylated “on-bead” with 1 µM PknD, (E) DosR phosphorylated “on-bead” with 1 µM PknH, then purified and incubated with 50 mM AcP, and (F) DosR phosphorylated “on-bead” with 1 µM PknD, then purified and incubated with 50 mM AcP. Data are representative of at least 3 independent experiments. Quantification of the intensity of the non-shifted band in the 1.57 µM protein EMSA reaction versus that in the 0.2 µM reaction for each protein/treatment is shown in (G). Data are shown as means ± SEM from 3-5 experiments. p-values shown in the table on the right in panel (G) were obtained with a one-way ANOVA with Tukey’s multiple comparisons. N.S. not significant, * p < 0.05, p < 0.01, p < 0.0001. The numerical data underlying the graph shown in this figure are provided in S1 Data.

Fluorescence polarization assays demonstrate inhibition of DosR binding to its target gene promoters upon PknH or PknD phosphorylation.The change in fluorescence anisotropy (∆FP) relative to no protein was measured as a function of increasing concentrations of DosR (black curves), PknH-phosphorylated DosR (blue curves), or PknD-phosphorylated DosR (red curves) incubated with a fluorescently labeled hspX (A) or fdxA (B) promoter DNA region. The fit using a Hill equation to estimate binding parameters with 95% confidence intervals (shaded regions) are shown. The dissociation constants (Kd) and Hill coefficients (n) are indicated in the tables below the graphs.

STPK phosphorylation of purified DosR decreases the level and alters concentration-dependence of steady-state transcription rates of its target genes.(A) STPK phosphorylation of DosR inhibits the ability of DosR to increase target gene steady-state transcription. A Spinach RNA aptamer assay was run with the fdxA promoter with different concentrations of indicated DosR protein. For “PknH-DosR” and “PknD-DosR”, phosphorylation of DosR with PknH or PknD, respectively, was performed “on-bead” before final purification of the phosphorylated DosR utilized in the assay. Fluorescence (arbitrary units, “AU”) was tracked over time on a plate reader, and steady-state rate calculated. Data are shown as means ± SEM from 3-8 experiments. p-values were obtained with a 2-way ANOVA with Tukey’s multiple comparisons. p-value in blue and red correspond to those for PknH-DosR and PknD-DosR, respectively, as compared to DosR. * p < 0.05, p < 0.01, * p < 0.001, p < 0.0001. (B and C) STPK phosphorylation alters the dynamics of activated DosR. A Spinach RNA aptamer assay was run with the fdxA promoter with different concentrations of indicated DosR protein, as in (A). (D) DosR concentration-dependent transcription is not observed with a promoter lacking DosR binding motifs. A Spinach RNA aptamer assay was run with the fdxA or rrnAP3 (no DosR binding motif) promoters with different concentrations of DosR protein, as in (A). 50 mM acetyl phosphate (AcP) was used where indicated. Data are shown as means ± SEM from 3-8 experiments. WT, PknH-DosR, and PknD-DosR data where shown in panels (B)-(D) are as shown in Fig 4A. The same DosR + AcP data set for the fdxA promoter is shown in panels (B)-(D). p-values were obtained with a 2-way ANOVA with Tukey’s multiple comparisons. Significant p-values are shown for the PknH-DosR + AcP, or PknD-DosR + AcP, versus DosR + AcP proteins. p < 0.01. The numerical data underlying the graphs shown in this figure are provided in S1 Data.

The STPK PknH regulates response of the hspX’::GFP reporter to NO.WT, ∆pknH, and pknH* (complemented mutant) Mtb each carrying the hspX’::GFP reporter were exposed for 1 day to the indicated DETA NONOate concentrations, before fixation and reporter signal analysis via flow cytometry. Fold induction is in comparison to the untreated condition for each strain. Data are shown as means ± SEM from 3 experiments. p-values were obtained with a 2-way ANOVA with Tukey’s multiple comparisons. Only significant comparisons are indicated. * p < 0.05, p < 0.01, * p < 0.001, p < 0.0001. The numerical data underlying the graph shown in this figure are provided in S1 Data.

The STPK PknH regulates the DosR-dependent transcriptional response of Mtb to NO.(A-C) ∆pknH Mtb exhibits increased induction of DosR-dependent NO-responsive genes at intermediate DETA NONOate levels. Log-phase WT, ∆pknH, and pknH* Mtb were exposed for 4 hours to 7H9, pH 7 media ± indicated DETA NONOate concentrations, before RNA was extracted for qRT-PCR analysis. (D-F) pknH deletion does not affect induction of DosR-independent NO-responsive genes. qRT-PCR analysis on DosR-independent NO-responsive genes were performed on samples obtained as in (A-C). In all cases, fold change compares each DETA NONOate condition to the control untreated condition for each strain. Data are shown as means ± SEM from 3 experiments, and p-values were obtained with unpaired t-tests with Holm-Sidak multiple comparisons. N.S. not significant, * p < 0.05, * p < 0.001. The numerical data underlying the graphs shown in this figure are provided in S1 Data.

Full text

Introduction

The ability of Mycobacterium tuberculosis (Mtb) to sense and respond to dynamic changes in environmental signals encountered throughout the course of infection is critical for successful host colonization. This includes cues such as acidic pH, chloride (Cl^-^), potassium (K^+^), nitric oxide (NO), and hypoxia, which are associated with the host immune response [1–7]. Mtb is the causative agent of tuberculosis and is the leading global cause of death from an infectious disease [8]. The heterogeneity of the environment Mtb experiences during host infection further affects the bacterial physiological state and contributes importantly to the difficulty of tuberculosis treatment [9–13].

Phosphorylation-based signal transduction enables environmental adaptation by linking extracellular signals to intracellular regulatory mechanisms. In bacteria, the best-studied mechanism of phosphorylation-based transmembrane signaling are two-component systems (TCSs), where ligand binding by the transmembrane histidine kinase (HK) sensor protein initiates a phosphorelay to the cognate intracellular response regulator (RR) protein, canonically a transcription factor that controls expression of specific genes [14,15]. The number of TCSs per bacterial genome strongly correlates with ecological and environmental niche [16,17] – bacteria living in more constant environments usually encode fewer TCSs, while bacteria that inhabit rapidly changing or diverse environments typically encode larger numbers of these signaling proteins that are critical for cellular processes [18–20]. Mtb encodes 12 TCSs that play a key role in virulence, environmental adaptation, and infection [14]. For example, inactivation of PhoPR, a key TCS involved in Mtb response to acidic pH and Cl^-^ [3,5], results in significant attenuation in the ability of Mtb to colonize its host [21]. An essential TCS in Mtb, PrrAB, was reported to be involved in early adaptation to intracellular infection [22,23], and we have since shown that PrrA is a global regulator of Mtb response to acidic pH, high [Cl^-^], hypoxia and NO [24]. The TCS DosRS(T) regulates Mtb response to hypoxia and NO, and upon sensing of these signals, mediates induction of a “dormant” state through the control of 48 genes known as the “dormancy regulon” [4,6,7,25]. The TCS KdpDE is responsible for regulation of the Kdp K^+^ uptake system and is pivotal for Mtb adaptation to low environmental potassium levels ([K^+^]) [1,26]. These examples illustrate the importance of TCSs for Mtb adaptation to its local environment.

Importantly, in addition to TCSs, signal transduction and environmental response in Mtb is also mediated by serine/threonine protein kinases (STPKs) [27–29]. In contrast to TCSs, STPKs have a larger set of phosphorylation targets, and while STPKs are often less numerous in the genome of a given bacterium, their widespread presence and impact on bacterial biology have become increasingly appreciated [27–32]. For example, phosphorylation of glutamyl tRNA reductase by the STPK Stk1 plays an important role in the regulation of heme biosynthesis in Staphylococcus aureus [33]. Another example is with Listeria monocytogenes, where phosphorylation of the protein ReoM by the STPK PrkA is essential for viability, due to its role in peptidoglycan synthesis [34]. The Mtb genome markedly contains a comparatively large number of STPKs compared to other bacterial species, with 11 STPKs encoded [27,30,35]. Notably, global phosphoproteome studies have identified TCS RRs as potential substrates of STPK phosphorylation [31,32], raising the concept of STPK-TCS interplay in the regulation of TCS function. Indeed, there is increasing support for this interplay, with studies in bacterial species ranging from Mtb and Bordetella, to Streptococcus pneumoniae, S. aureus, and Bacillus subtilis [31,32,36–42]. Specifically for Mtb, all 12 TCS RRs have been identified from global phosphoproteome studies as potential substrates of STPK phosphorylation [31,32]. In the case of DosR and PrrA, both RRs from TCSs that respond to NO stress among other signals, biochemical assays with purified proteins have further verified their phosphorylation by STPKs [36,43], with the STPK PknH shown to phosphorylate DosR [36]. Strikingly, we have shown that a STPK phosphoablative PrrA-T6A Mtb mutant was significantly altered in its transcriptional response to acidic pH and high [Cl^-^], and dampened in environmental response to both NO and hypoxia [24]. Consequently, this mutant was incapable of entry into an adaptive state of growth arrest upon extended exposure to NO and attenuated for host colonization in vivo [24]. While there is clear evidence for interplay between STPKs and TCSs, much remains unknown regarding how the two systems interact in the coordination of Mtb environmental adaptation.

Here, we examined STPK interactions with the RRs DosR, PrrA, PhoP, and KdpE, all RRs known to play critical roles in Mtb transcriptional response to environmental cues encountered throughout host infection [1,3–7], revealing both specificity and overlap in interactions. Focusing on DosR as an exemplar system, we find that PknH and PknD phosphorylation of purified DosR decreased its binding to the promoter of its target genes, with strong binding restored by combined treatment with acetyl phosphate (AcP), which mimics HK phosphotransfer [44,45]. Further, PknH and PknD phosphorylation of purified DosR decreased the steady-state transcription rate of DosR target genes, in contrast to the increase observed upon AcP treatment of DosR. Additionally, STPK phosphorylation shifted the AcP-treated DosR concentration-dependence of target gene activation. Finally, we found that a ∆pknH Mtb mutant exhibited increased DosR regulon transcription at lower NO levels than wild type (WT) Mtb. Combined, our work sheds light on the mechanisms underpinning STPK-TCS interplay, illustrating how STPK phosphorylation of a TCS RR can act to restrain its activation to ensure response initiation only when appropriate.

Results

There is both overlap and specificity in interactions between STPKs and TCS RRs

To systematically identify possible interactions between STPKs and TCSs RRs, we utilized the mycobacterial protein fragment complementation (M-PFC) assay [46]. Comparable to the bacterial two-hybrid assay, this method is based on functional reconstitution of the murine dihydrofolate reductase (mDHFR) driven by interactions between two test proteins, thereby conferring resistance to trimethoprim (TRIM) [46]. A previous study applying this method uncovered an interaction between DosR and the STPK PknH, providing precedence for its use in this context [36]. In particular, we examined the RRs DosR, PrrA, PhoP, and KdpE, fusing the open reading frame of each to the mDHFR fragment F1,2, with each of 9 STPKs fused to the mDHFR fragment F3. Some combinations of RRs and specific STPKs showed clear growth at 50 µg/mL TRIM, indicating a strong interaction (Figs 1 and S1). As controls, Mycobacterium smegmatis containing empty vectors were unable to grow on 7H10 TRIM plates, whereas all strains grew well on 7H10 plates lacking TRIM (Fig 1). We compared interactions across four RRs to explore possible relationships between environmental signals responded to by a TCS and interactions with STPKs, as shown in S1 Fig. Interestingly, we observed almost complete overlap in the STPKs that interact with PrrA and DosR, with smaller subsets of those same STPKs interacting with PhoP and KdpE. These results indicate that there is both overlap and specificity in the interactions between STPKs and RRs.

Specificity of interactions between TCS RRs and STPKs.Interactions between the TCS RRs DosR (A), PrrA (B), PhoP (C), or KdpE (D) with the various STPKs (kinase domains only for all except PknG and PknK) were tested by M-PFC assay. The positive control (“+”) was M. smegmatis expressing the S. cerevisiae GCN4 dimerization domains fused to the F1,2 or F3 domains of the murine dihydrofolate reductase gene; the negative control (“-“) was M. smegmatis expressing the respective RR fused to the F1,2 domain and a F3 domain that was not fused to any Mtb gene. Data are representative of 3 independent experiments.

In accord with previous results [36], mass spectrometry (MS) analysis of recombinantly expressed and purified DosR that was subsequently phosphorylated in vitro with recombinant PknH identified phosphorylation at the Thr198 and Thr205 residues (S2–S4 Figs). Analysis of DosR samples also directly confirmed that AcP treatment, mimicking HK phosphotransfer [44,45], led to phosphorylation of the Asp54 residue (S5 Fig) [47,48]. Notably, analyzing the results from a previous global study of the Mtb O-phosphoproteome, where phosphosites in individual STPK overexpression and deletion mutants were compared to WT Mtb [32], provided support for the likely physiological interactions of several STPK-TCS RRs indicated by our M-PFC assay results. In particular, significant differences in phosphopeptides within DosR (i.e., increased presence upon STPK overexpression and/or decreased presence with the STPK deletion mutant, as compared to WT Mtb) were identified for all of the STPKs identified as DosR interactors in our M-PFC assay [32]. This was similarly the case for PhoP, while significant differences in phosphopeptides within PrrA were identified for PknD and PknE in the global phosphoproteome study [32]. Together, our results reinforce the concept of extensive interplay between STPKs and TCS RRs in Mtb.

Changes in DosR phosphorylation status alter its binding affinity to target promoters

TCS RRs are canonically transcription factors that mediate their activity through changes in gene transcription [14,15]. To examine how STPK phosphorylation of TCS RRs mechanistically affect their function, we thus first analyzed effects on target promoter binding, focusing our studies on DosR as an exemplar RR. As expected, AcP treatment to mimic HK phosphotransfer enhanced DosR binding to the promoter of hspX, a member of the DosR regulon [7,25,45], as indicated by electrophoretic mobility shifts observed at lower DosR concentrations with AcP-treated DosR (Fig 2A, 2B, quantified in 2G). Intriguingly, in vitro phosphorylation of DosR with the STPK PknH or PknD resulted in decreased DNA binding affinity to a similar degree (compare Fig 2C and 2D to 2A, quantified in 2G). In a live bacterium, both HK phosphotransfer and STPK phosphorylation might be expected to co-occur at times, therefore we next examined the influence of both types of phosphorylation on DosR target promoter binding affinity. AcP treatment of PknH or PknD phosphorylated DosR resulted in a DNA binding affinity that resembled that of AcP-treated DosR alone (compare Fig 2E and 2F to 2B, quantified in 2G). Similar changes in promoter binding affinity depending on DosR phosphorylation status were observed with DosR binding to the promoter of fdxA, another member of the DosR regulon (S6 Fig) [7,25,49]. For both the hspX and fdxA promoters, competition experiments with unlabeled probes of each respective promoter reversed the gel shift in a concentration-dependent manner (S7A and S7B Fig). Conversely, competition experiments with an unlabeled probe of a promoter not bound by DosR (rv2390c promoter [1]) did not affect the gel shift (S7C and S7D Fig). These results demonstrate the specificity of the binding results.

Changes in DosR phosphorylation status alter binding affinity to the promoter of its target gene hspX.Electrophoretic mobility shift assays (EMSAs) using purified recombinant C-terminally 6x-His-tagged DosR and IRDye 700-labeled probes for the hspX promoter are shown. A control with no protein (“NP”) added is shown for each gel. DosR was added at indicated concentrations for all other lanes. 40 fmoles of hspX promoter DNA was used in each reaction. EMSAs shown are as follows: (A) untreated DosR, (B) DosR incubated with 50 mM acetyl phosphate (AcP), (C) DosR phosphorylated “on-bead” with 1 µM PknH, (D) DosR phosphorylated “on-bead” with 1 µM PknD, (E) DosR phosphorylated “on-bead” with 1 µM PknH, then purified and incubated with 50 mM AcP, and (F) DosR phosphorylated “on-bead” with 1 µM PknD, then purified and incubated with 50 mM AcP. Data are representative of at least 3 independent experiments. Quantification of the intensity of the non-shifted band in the 1.57 µM protein EMSA reaction versus that in the 0.2 µM reaction for each protein/treatment is shown in (G). Data are shown as means ± SEM from 3-5 experiments. p-values shown in the table on the right in panel (G) were obtained with a one-way ANOVA with Tukey’s multiple comparisons. N.S. not significant, * p < 0.05, ** p < 0.01, **** p < 0.0001. The numerical data underlying the graph shown in this figure are provided in S1 Data.

Complementary to the electrophoretic mobility shift assays (EMSAs), we utilized fluorescence polarization to further quantitatively measure DosR binding to target gene promoters. Consistent with the EMSA results, the dissociation constant (K_d_) of DosR binding to its target hspX and fdxA promoters was significantly increased upon PknH or PknD phosphorylation of DosR, demonstrating a decrease in binding affinity (Fig 3). Together, these results show that STPK phosphorylation of DosR alone, in the absence of HK phosphotransfer, decreases the affinity of DosR for its target gene promoters, indicating that STPK phosphorylation can serve as a modulatory mechanism providing tighter control of DosR activation.

Fluorescence polarization assays demonstrate inhibition of DosR binding to its target gene promoters upon PknH or PknD phosphorylation.The change in fluorescence anisotropy (∆FP) relative to no protein was measured as a function of increasing concentrations of DosR (black curves), PknH-phosphorylated DosR (blue curves), or PknD-phosphorylated DosR (red curves) incubated with a fluorescently labeled hspX (A) or fdxA (B) promoter DNA region. The fit using a Hill equation to estimate binding parameters with 95% confidence intervals (shaded regions) are shown. The dissociation constants (Kd) and Hill coefficients (n) are indicated in the tables below the graphs.

STPK phosphorylation of purified DosR decreases the level and alters concentration-dependence of steady-state transcription rates of its target genes

While EMSAs and fluorescence polarization assays provide insight into target promoter DNA binding, RR promoter binding affinity is just one factor determining its effect on gene transcription. For example, a transcription factor with high DNA-binding affinity but with reduced ability to interact productively with RNA polymerase [50,51], or to modulate the kinetics of transcription initiation [52], would not be expected to be a strong transcriptional activator. To investigate the effects of STPK phosphorylation on RNA production directly, we utilized a fluorescent RNA aptamer-based method to quantify the effect of STPK phosphorylation of purified DosR on the steady-state rate of target gene transcription. This assay exploits the use of a Spinach-mini aptamer that produces a fluorescence-based enhancement when binding a small molecule fluorophore, such that each transcription event results in a consequent increase in fluorescence, allowing for transcript production to be monitored in real time [53,54]. Examining fdxA as the target gene, phosphorylation of purified DosR with PknH or PknD inhibited the ability of DosR to increase steady-state transcription (Fig 4A), corresponding with the decrease in binding to the fdxA promoter observed with PknH or PknD-phosphorylated DosR (Figs 3B, S6C and S6D).

STPK phosphorylation of purified DosR decreases the level and alters concentration-dependence of steady-state transcription rates of its target genes.(A) STPK phosphorylation of DosR inhibits the ability of DosR to increase target gene steady-state transcription. A Spinach RNA aptamer assay was run with the fdxA promoter with different concentrations of indicated DosR protein. For “PknH-DosR” and “PknD-DosR”, phosphorylation of DosR with PknH or PknD, respectively, was performed “on-bead” before final purification of the phosphorylated DosR utilized in the assay. Fluorescence (arbitrary units, “AU”) was tracked over time on a plate reader, and steady-state rate calculated. Data are shown as means ± SEM from 3-8 experiments. p-values were obtained with a 2-way ANOVA with Tukey’s multiple comparisons. p-value in blue and red correspond to those for PknH-DosR and PknD-DosR, respectively, as compared to DosR. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. (B and C) STPK phosphorylation alters the dynamics of activated DosR. A Spinach RNA aptamer assay was run with the fdxA promoter with different concentrations of indicated DosR protein, as in (A). (D) DosR concentration-dependent transcription is not observed with a promoter lacking DosR binding motifs. A Spinach RNA aptamer assay was run with the fdxA or rrnAP3 (no DosR binding motif) promoters with different concentrations of DosR protein, as in (A). 50 mM acetyl phosphate (AcP) was used where indicated. Data are shown as means ± SEM from 3-8 experiments. WT, PknH-DosR, and PknD-DosR data where shown in panels (B)-(D) are as shown in Fig 4A. The same DosR + AcP data set for the fdxA promoter is shown in panels (B)-(D). p-values were obtained with a 2-way ANOVA with Tukey’s multiple comparisons. Significant p-values are shown for the PknH-DosR + AcP, or PknD-DosR + AcP, versus DosR + AcP proteins. ** p < 0.01. The numerical data underlying the graphs shown in this figure are provided in S1 Data.

dosR expression is itself upregulated by the very signals that the DosRS(T) system responds to [4,6,7]. We found that concentrations as low as 0.25 µM AcP-treated DosR resulted in an increase in target gene transcription rate, with maximal rates obtained at 0.5-1 µM DosR, before levels again decreased at concentrations ≥1.5 µM (Fig 4B, black dashed line). Interestingly, when purified DosR was phosphorylated by PknH or PknD, in addition to being treated with AcP, an increase in transcription rate was also observed, but with a narrower activation window, as the “de-activation” response occurred with a steeper decline (for example, 4.76 ± 1.67 AU/s and 6.10 ± 1.56 AU/s for 3 µM PknH and PknD-phosphorylated, AcP-treated DosR, respectively, versus 12.14 ± 1.26 AU/s for only AcP-treated DosR, p < 0.01 in each case; Fig 4B and 4C, compare blue and red dashed lines to black dashed line). Importantly, this de-activation response is dependent on the presence of DosR binding motifs (“DosR boxes”) and not simply due to non-specific DNA coating, as no transcription decrease was observed for the ribosomal rrnAP3 promoter that is not controlled by DosR and lacks DosR boxes (Fig 4D).

These biochemical results further demonstrate how STPK phosphorylation can alter DosR activity output, and also intriguingly show a concentration-dependent effect of AcP treatment (HK phosphotransfer) on DosR activity that is modulated by STPK phosphorylation. Together, these results illuminate how STPK phosphorylation of DosR can affect its function, and suggests a mechanism by which the response of Mtb to key environmental signals can be tightly regulated.

The STPK PknH regulates the DosR-dependent transcriptional response of Mtb to NO

Given our working hypothesis, supported by the observations above, that STPK activity is functionally relevant to regulation of the Mtb stress response, we finally sought to examine how pknH deletion might alter the sensitivity of Mtb to NO. To this end, we first transformed WT, ∆pknH, and pknH* (complemented mutant) with our NO/hypoxia-responsive hspX’::GFP reporter [5], and tested reporter response in different DETA NONOate concentrations. Markedly, hspX’::GFP reporter induction was higher at an intermediate DETA NONOate concentration (50 µM) with ∆pknH Mtb as compared to WT or pknH* Mtb (Fig 5). In contrast, in the presence of 100 µM DETA NONOate, reporter induction was similar across all strains (Fig 5).

The STPK PknH regulates response of the hspX’::GFP reporter to NO.WT, ∆pknH, and pknH (complemented mutant) Mtb each carrying the hspX’::GFP reporter were exposed for 1 day to the indicated DETA NONOate concentrations, before fixation and reporter signal analysis via flow cytometry. Fold induction is in comparison to the untreated condition for each strain. Data are shown as means ± SEM from 3 experiments. p-values were obtained with a 2-way ANOVA with Tukey’s multiple comparisons. Only significant comparisons are indicated. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. The numerical data underlying the graph shown in this figure are provided in S1 Data.*

While the DosR regulon encompasses the subset of genes most highly induced upon Mtb exposure to NO, there is also a significant number of Mtb genes that are responsive to NO in a non-DosR-dependent manner [4,24]. To more specifically probe the relationship of PknH phosphorylation with DosR-mediated response of Mtb to NO, we thus next examined induction of DosR-dependent and -independent NO-responsive genes by qRT-PCR, examining induction across a range of DETA NONOate exposures as with the hspX’::GFP reporter assay. Analyses of three representative genes in the DosR regulon show markedly increased induction for all three genes in ∆pknH Mtb at the intermediate 50 µM DETA NONOate concentration that was restored to WT Mtb levels by complementation (Fig 6A), with induction levels at the higher 75 µM and 100 µM DETA NONOate concentrations similar across the strains (Fig 6B and 6C), supporting the hspX’::GFP reporter data (Fig 5). Strikingly, no change in induction profile between WT and ∆pknH was observed at all DETA NONOate concentrations for the three NO-responsive, non-DosR-regulated genes tested (Fig 6D–6F).

The STPK PknH regulates the DosR-dependent transcriptional response of Mtb to NO.(A-C) ∆pknH Mtb exhibits increased induction of DosR-dependent NO-responsive genes at intermediate DETA NONOate levels. Log-phase WT, ∆pknH, and pknH Mtb were exposed for 4 hours to 7H9, pH 7 media ± indicated DETA NONOate concentrations, before RNA was extracted for qRT-PCR analysis. (D-F) pknH deletion does not affect induction of DosR-independent NO-responsive genes. qRT-PCR analysis on DosR-independent NO-responsive genes were performed on samples obtained as in (A-C). In all cases, fold change compares each DETA NONOate condition to the control untreated condition for each strain. Data are shown as means ± SEM from 3 experiments, and p-values were obtained with unpaired t-tests with Holm-Sidak multiple comparisons. N.S. not significant, * p < 0.05, *** p < 0.001. The numerical data underlying the graphs shown in this figure are provided in S1 Data.*

Together, these results support a role of PknH as a fine-tuning regulatory mechanism for Mtb environmental response, acting to restrain DosR activation to ensure response initiation only when appropriate.

Discussion

The ability of Mtb to sense and respond to environmental cues is dependent on signal transduction regulatory mechanisms such as STPKs and TCSs, which are critical for bacterial adaptation and hence survival within a host [14,21,22,24,27,55–57]. While the interplay between STPKs and TCSs has become increasingly appreciated, how such interplay may alter TCS function and thus the response of Mtb to environmental cues has remained understudied. Here, we uncover a role of STPK phosphorylation of TCSs as a fine-tuning regulatory mechanism that acts to provide an additional layer of regulation of TCS RR activity.

In particular, our findings that PknH phosphorylation of purified DosR decreases DNA binding affinity to target promoters and steady-state transcription of target genes, and that deletion of pknH in Mtb results in increased induction of DosR-dependent, but not DosR-independent, NO-responsive genes at intermediate levels of NO stress, strongly support that regulation of DosR by STPKs serve as a second axis of regulation, in addition to regulation by its cognate histidine kinases. A previous study had conversely reported increased binding of DosR to the hspX promoter when DosR and PknH were co-expressed in Escherichia coli, with DosR subsequently purified for EMSAs, as well as decreased induction of several DosR regulon genes in ∆pknH Mtb upon NO exposure [36]. It is difficult, however, to separate out what other modifications may also occur in E. coli, and there were differences in the EMSA assays, with the previous study testing binding to separate DosR boxes present in the hspX promoter (38 bp or 54 bp segments), versus the longer DNA segments encompassing all DosR boxes upstream of the hspX promoter utilized in our study. Complementation was not shown for the ∆pknH transcriptional response phenotypes in that study [36], and differences in Mtb strain (H37Rv in the previous study versus CDC1551 here) and growth conditions between our studies may also account for the different results.

Notably, both the DosR T198 and T205 sites phosphorylated by PknH map within the α10 helix of the DosR crystal structure [58,59]. DosR is thought to adopt two dimeric structures, one active and one inactive, that exist in equilibrium with each other as well as a monomeric form [60]. Phosphotransfer to the D54 site in the receiver domain favors the active, DNA-binding DosR, with the α10 helix from each monomer mediating the dimerization of that species, making this helix critical for protein activation [58,60]. The T198 and T205 PknH phosphorylation sites are closely positioned in the DosR dimer interface in the DNA-binding conformation, and we thus posit that PknH phosphorylation of these sites shifts the equilibrium of DosR in such a way as to disfavor the active DNA binding-competent dimer form, resulting in the decreased DNA binding affinity and steady-state transcription observed. Examples of opposing effects of STPK and HK phosphorylation on activity of the corresponding RR have been previously reported in S. aureus and Streptococcus agalactiae [41,42], and our data further support this concept of STPKs serving to restrain TCS activity to ensure appropriate response.

Our attempt to generate a STPK phosphoablative DosR was unsuccessful, as mutation of the T198 and T025 sites to alanine residues resulted in a non-functional protein with poor DNA binding and minimal fdxA steady-state transcription (S8 Fig). Interestingly, positioning of known or putative STPK phosphorylation sites close to or within the DNA binding region is also found for other Mtb TCS RRs such as MtrA and PhoP [32,57,61,62], as well as in TCS RRs of other bacterial species [37,38,42]. In the case of the Mtb TCS RR PrrA, STPK phosphorylation occurs within the receiver domain, affecting Mtb response to various environmental signals [24]. Future studies seeking to uncover how STPK phosphorylation alters TCS structure and hence function will be important for continued insight into the role of STPKs as regulators of TCSs.

A key outstanding question is how STPK activity is regulated. Expression levels of STPKs have not generally been found to be affected by key environmental signals for Mtb, and Mtb encodes only a single Ser/Thr phosphatase, with broad activity across phosphorylated Ser/Thr residues [27]. The large number of targets for a given STPK would however strongly suggest a need for regulation of their activity, and indeed the basal activity level for many STPKs appears low, as evidenced by deletion mutants of the STPKs exhibiting little reduction in phosphosites, compared to WT Mtb, in standard rich media [32]. STPKs autophosphorylate and dimerization is required for their activation [27,63,64]; ligand triggering of such events thus represent a route for control of STPK activity separate from expression differences. Indeed in B. subtilis, the STPK PrkC senses cell wall fragments, with binding of peptidoglycan fragments to its extracellular domain leading to phosphorylation and activation of an essential ribosomal GTPase involved in initiating vegetative growth [65]. In Mtb, a ∆pknG mutant is defective for growth when glutamate or asparagine is used as the sole nitrogen source, and phosphorylation of GarA, a key target of PknG, was greatest in the presence of these same amino acids, suggesting that these amino acids may act as triggers for PknG activity [66]. Intriguingly, we found almost complete overlap in the STPKs that interact with the PrrA and DosR RRs, both known to mediate Mtb response to NO and hypoxia [24,25,67]. A smaller subset of those same STPKs also interacted with PhoP, which like PrrA, functions as a global regulator of pH and Cl^-^ [5,24]. Future studies analyzing the extent of phosphorylation of each RR by a given interacting STPK identified here, and testing whether the same environmental signals that drive TCS activity also affect STPK activity, will provide important insight into the coordination of TCS and STPK activity.

Finally, the RNA aptamer-based transcriptional assay utilized here provides a real time, quantitative, and high-throughput method with broad utility for understanding basal and regulated transcription dynamics. This encompasses the ability to test effects of different intrabacterial signals and regulatory factors on gene transcription kinetics, through to analysis of antibiotic-dependent inhibition of RNA polymerase on transcription steady-state rates [53,54]. Interestingly, this transcriptional assay revealed a non-monotonic relationship between AcP-activated DosR concentration and target gene steady-state transcription rates. More specifically, after the expected increase in transcription with initial increases in DosR concentration, we observed an unexpected decrease in transcription with further DosR concentration increases, a phenomenon that was dependent on the presence of DosR binding motifs on the target promoter. Strikingly, the slope of this “de-activation” was modulated by STPK phosphorylation. In Mtb, expression of the DosR regulon is markedly elevated upon initial exposure to NO or hypoxia [4,6,7], reflecting its role in mediating the early transcriptional response to these stresses. However, this induction is transient, and expression levels decline over time even in the continued presence of the inducing signal [68]. Our results suggest a possible negative feedback mechanism incorporating DosR phosphorylation state, whereby active DosR concentrations above a threshold result in decreased target gene expression relative to the maximum activation at lower concentrations. The HK DosS has been reported to also possess phosphatase activity [69,70]; defining how such phosphatase activity may work in concert with both the observed RR concentration dependence and STPK phosphorylation to enable a negative feedback loop in DosR regulon expression will be vital for understanding physiological adaptation of Mtb to environmental cues that are maintained over the course of a chronic infection.

TCSs and STPKs are the two major regulatory systems through which environmental signals are transduced into adaptive outputs in bacteria. Our findings here illustrate how STPK-mediated phosphorylation of TCS RRs can act to fine tune transcriptional outputs, serving to ensure response initiation only when appropriate. Our work establishes a framework for dissecting STPK-TCS interplay, and we propose that further studies probing the interactions of these two regulatory systems will continue to yield important insight into molecular pathways critical for Mtb environmental adaptation and host colonization.

Materials and methods

Mycobacterial protein fragment complementation assays

Mycobacterial protein fragment complementation (M-PFC) assays were performed essentially as described previously [46]. The open reading frames of dosR, prrA, phoP, and kdpE were each cloned into pUAB100, generating C-terminal translational fusions to the murine dihydrofolate reductase (mDHFR) fragment F1,2 domain. All Mtb STPK genes were cloned into pUAB400, generating N-terminal translational fusions to the mDHFR fragment F3 domain. For STPKs with transmembrane domains (all except PknG and PknK), only the kinase domain was cloned. M. smegmatis transformed with each respective plasmid pair (each STPK with each RR) were plated on 7H11 plates supplemented with 25 µg/ml kanamycin and 50 µg/ml hygromycin, with TRIM added at 10, 20, 30, or 50 µg/ml. pUAB100 and pUAB400 plasmids containing the Saccharomyces cerevisiae GCN4 homodimerization domain served as the positive control [46]. As a negative control, M. smegmatis expressing the RR-mDHFR F1,2 fusion in pUAB100 with an empty pUAB400 plasmid was used. M. smegmatis transformants carrying mDHFR F3-PknB or mDHFR F3-PknI could not be obtained, and M-PFC assays with these STPKs were thus not pursued.

Recombinant protein expression and purification

The open reading frames of DosR and DosR-T198A/T205A were individually cloned into the isopropyl-β-D-1-thiogalactopyranoside (IPTG)-inducible pET-23a vector to generate a C-terminally 6xHis-tagged DosR and DosR-T198A/T205A, respectively. Expression constructs for the kinase domains of PknH and PknD were previously described [64]. All constructs were transformed into Escherichia coli BL21(DE3) for recombinant expression and purification. For expression, 2 ml of overnight E. coli cultures started from single colonies were grown in 5 ml LB + 50 μg/ml ampicillin at 37°C and were used to inoculate 1 L LB media + 50 μg/ml ampicillin. Cultures were grown at 37°C, 160 rpm, until the culture reached an OD_600_ of ~0.6. Induction of constructs was initiated by adding 1 mM IPTG, and the cultures grown for an additional 16 hours at 16°C, 160 rpm. Afterwards, supernatants were removed, and pellets were stored at -80°C prior to further processing.

Recombinant purification of DosR and its variants and STPKs followed previously described protocols [1,24]. The template for the DosR T198A/T205A mutant was generated using QuikChange mutagenesis of WT DosR (Agilent). To remove phosphorylated residues accrued from expression in E. coli, the protein was treated with alkaline phosphatase (Sigma #P0114) according to previously described protocols [71], when the protein was still bound to the nickel beads. Dephosphorylated DosR was then washed three times with a minimal low imidazole buffer (500 mM NaCl, 50 mM Tris, pH 7.5, 15 mM imidazole, 10% glycerol) to remove residual alkaline phosphatase. For in vitro phosphorylation of DosR, DosR still bound to nickel beads was treated with 1 µM PknH or PknD and incubated at room temperature in kinase buffer (40 mM Tris-HCl, pH 7.5, 2 mM MnCl_2_, 20 mM MgCl_2_, 2 mM DTT, 0.5 mg/mL BSA) for one hour. The nickel bead-bound DosR was then washed to remove the purified STPKs, prior to continuation of the protein purification protocol to obtain STPK-phosphorylated DosR. DosR protein was dialyzed into electrophoretic mobility shift assay (EMSA) buffer as described [45]. Protein concentrations were quantified by using a Bradford assay (Bio-Rad).

Mtb RNAP σ^A^ holoenzyme complex was purified by a 10X N-terminal His-tag on the alpha subunit, using pET-Duet-rpoB-rpoC, pAcYc-HisrpoA-rpoZ, and pAC27-sigA plasmids, expressed and purified as previously described [54]. The final holoenzyme fractions were dialyzed into storage buffer (10 mM Tris-Cl, pH 7.0, 200 mM NaCl, 0.1 mM EDTA, 1 mM MgCl_2_, 20 μM ZnCl_2_, 2 mM DTT, 50% glycerol), concentrated to 4.5 μM (determined using an extinction coefficient of 280,425 M^-1^ cm^-1^), aliquoted, flash frozen in liquid nitrogen, and stored at -80°C.

Mtb CarD and RbpA, in pET-SUMO plasmid vectors, were expressed, purified, and the His-SUMO tag removed as previously described [54,72]. Eluted fractions were dialyzed overnight in 20 mM Tris, pH 8.0, 150 mM NaCl, 1 mM β-mercaptoethanol, then concentrated to 200 μM determined using extinction coefficients of 16,900 M^-1^ cm^-1^ for Mtb CarD and 13,980 M^-1^ cm^-1^ for Mtb RbpA.

Mass spectrometry

Protein samples underwent in-gel trypsin digestion at the Mass Spectrometry Technology Access Center at the McDonnell Genome Institute (MTAC@MGI) at Washington University School of Medicine. The resulting peptides were analyzed by LC-MS/MS using data-dependent acquisition on an Orbitrap Eclipse Tribad mass spectrometer (ThermoFisher Scientific). Data were searched against a custom E. coli BL21 (DE3) database supplemented with the DosR sequence, using Mascot software to identify phosphorylation on aspartic acid, serine, threonine, and tyrosine residues. In all samples, DosR accounted for at least 80% of the identified peptides and 99% sequence coverage of DosR was obtained.

Electrophoretic mobility shift assays

Electrophoretic mobility shift assays (EMSAs) were performed essentially as previously described [1,73]. In brief, promoter regions for hspX (558 bp) and fdxA (221 bp) were amplified using IRDye 700 labeled primers (Integrated DNA technologies) and the PCR products purified using a QIAquick PCR purification kit (Qiagen). For acetyl phosphate-treated reactions, purified DosR-6xHis protein was treated as described previously [45]. Indicated amounts of unphosphorylated and phosphorylated DosR were mixed with 40 fmoles of DNA in EMSA buffer (25 mM Tris-HCl, pH 8, 20 mM KCl, 6 mM MgCl_2_, 5% glycerol, 0.5 mM EDTA, 25 µg/ml salmon sperm DNA [45]) in a 10 µl final reaction volume. For competition EMSAs with unlabeled probes (hspX, fdxA, or the non-DosR target rv2390c [1]), 1 fmole of labeled probe was used, with indicated fold-molar excess ratios of unlabeled probe and protein in the reaction mixture. For all EMSAs, reactions were incubated at room temperature for 30 minutes and then run on non-denaturing 7.5% Tris-glycine gels in 0.5X Tris-borate-EDTA buffer at 4°C for ~4-4.5 hours. Gels were imaged using the 700 nm channel of an Odyssey CLx imaging system (LI-COR).

Fluorescence polarization assays

Fluorescence polarization (FP) experiments were performed using linear double-stranded DNA probes containing either the hspX (85 bp) or fdxA (55 bp) promoter labeled with Alexa Fluor 488 on the downstream 5’ end (Integrated DNA Technologies). These were titrated with increasing concentrations of purified DosR, or PknH/PknD-treated DosR. 10 µl binding reactions in 384-well black, low volume, round bottom assay plates (Corning) were sealed with optical adhesive film (Applied Biosystems) and measured on a CLARIOstar Plus plate reader (BMG LABTECH). DosR and its phosphorylation variants were serially diluted (1:1.5) from a 25 µM stock solution into binding buffer (25 mM Tris-HCl 8.0, 20 mM KCl, 6 mM MgCl_2_, 10% glycerol), generating a final concentration range of 0–25 µM. Labeled DNA substrates were added to each well at a final concentration of 25 nM (5 µl of 50 nM stock). Plates were incubated at 37°C for 15 minutes before polarization measurements were taken using 485/520 nm excitation/emission FP filters.

For analysis, binding curves were fit to a Hill-type saturation equation using maximum likelihood estimation (MLE) implemented in R. Raw data were plotted, and fits were generated using the following Hill equation: ΔFP = (ΔFP_max_⋅[P]^n^)/ (K_d_^n^ + [P]^n^); Where ΔFP is the change in fluorescence polarization, [P] is the protein concentration (µM), ΔFP_max_ is the maximal signal, K_d_ is the apparent dissociation constant, and n is the Hill coefficient. For the DosR condition, where the data reach saturation, all three parameters were allowed to float. For the PknH-DosR and PknD-DosR conditions, which did not reach saturation, ΔFPmax was fixed to the value obtained from DosR alone, and MLE was used to estimate K_d_ and n, and the residual variance error parameter, σ. 95% confidence intervals (CIs) for parameter estimates were obtained by bootstrapping. For each condition, the Hill model was refit to 10000 synthetic datasets generated by data resampling. The 2.5^th^ and 97.5^th^ percentiles of the resulting distributions of parameter estimates defined the CI limits for each parameter. These confidence intervals were propagated to the fitted curves, producing shaded ribbons in the plots that visualize uncertainty in the predicted fluorescence polarization across the protein concentration range. On the hspX promoter the residual variances (σ) were as follows: 16.3 ± 0.98 (DosR), 5.52 ± 0.47 (PknH-DosR), and 13.2 ± 0.91 (PknD-DosR). On the fdxA promoter, σ values were: 12.1 ± 0.89 (DosR), 12.9 ± 1.27 (PknH-DosR), and 10.0 ± 0.54 (PknD-DosR).

hspX probe:/5Alex488N/ACAACAGGGTCAATGGTCCCCAAGTGGATCACCGACGGGCGCGGACAAATGGCCCGCGCTTCGGGGACTTCTGTCCCTAGCCCTG

fdxA probe:

/5Alex488N/TGACGAATAAGGCCTTTGGTCCTTTCCGGTAGGGGTCTTTG GATAGGCGCGATCC

RNA aptamer-based transcriptional assay

Aptamer-based transcription data was collected using a CLARIOstar Plus Microplate reader (BMG LABTECH) in a 384 well, low volume, round-bottom, non-binding polystyrene assay plate (Corning) with the corresponding Voyager analysis software and following a previously published protocol [54]. To measure multi-round, steady-state transcription kinetics in real-time, we monitored the change in 3,5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI) fluorescence upon binding to a transcribed, full-length RNA sequence containing the fdxA or rrnAP3 promoter [54] and the iSpinach D5 aptamer. DFHBI fluorescence was measured using an excitation wavelength of 480 ± 15 nm (monochromator) while monitoring the emission signal at 530 ± 20 nm (filter). All reactions were conducted at 37°C in 10 μl final volume in 20 mM Tris (pH 8.0 at 37°C), 40 mM NaCl, 75 mM potassium glutamate, 10 mM MgCl_2_, 5 μM ZnCl_2_, 20 μM EDTA, 5% glycerol with 1 mM DTT and 0.1 mg/ml BSA. Reactions contained 100 nM RNAP holoenzyme, 20 μM DFHBI dye (Sigma Aldrich), 0.4 U/μl RiboLock RNase inhibitors (Thermo Scientific), CarD and RbpA at 1 μM and 2 μM, respectively, and dilutions of DosR from 1 μM to 50 μM. 2.5 μl stock rNTPs (Thermo Scientific) were injected in situ using the reader’s automated reagent injector to a final concentration of 1 mM NTP. Data were acquired in 10–20 second intervals for up to 40 minutes total. A minimum of 3 technical replicates of the negative control (i.e., no rNTPs) were collected and measured concurrently with the experimental data. Using the average of this negative control, the experimental data was corrected as previously described [53], bringing all starting fluorescence values to zero and correcting for any time-dependent drift in fluorescence. Between 2 and 8 independent experiments were collected for each condition with 3 technical replicates each. Standard deviations were used as a statistical weight during the linear regression analyses as previously described to obtain the steady-state rate [53].

Mtb strains and culture

Mtb CDC1551 was used as the parental strain for all assays here, and Mtb cultures were cultured and maintained as described previously, with 7H9 broth supplemented with 10% OADC, 0.2% glycerol, 0.05% Tween-80, and 100 mM MOPS used for buffering to pH 7.0 [74]. Generation of ∆dosR, ∆pknH, and their complements were constructed with methods as described previously [5]. The dosR deletion consisted of a region from the beginning of the open reading frame through nucleotide 650, while the pknH deletion encompassed the entire pknH open reading frame. Complementation in both cases utilized the respective endogenous promoters and open reading frames in the integrating plasmid pMV306. The hspX’::GFP reporter introduced into indicated strains was previously reported [5]. Antibiotics were added as needed at the following concentrations: 100 μg/ml streptomycin, 50 μg/ml hygromycin, 50 μg/ml apramycin, and 25 μg/ml kanamycin.

qRT-PCR analyses

Mtb grown to log-phase (OD_600_ ~ 0.6) in aerated conditions was used to inoculate filter-capped T75 flasks laid flat, containing 12 ml 7H9, pH 7.0 ± indicated concentrations of DETA NONOate (Cayman Chemicals) at OD_600_ = 0.3. Bacteria were incubated for 4 hours, before RNA extracted as previously described [3]. qRT-PCR experiments were conducted and analyzed according to previously established protocols [75].

hspX’::GFP reporter assay

Indicated Mtb strains carrying the hspX’::GFP reporter were propagated to log phase (OD_600_ ~ 0.6) and subcultured to an OD_600_ = 0.05 in flat T75 flasks with filter caps containing 4 ml 7H9, pH 7. After 4 passages, Mtb was subcultured at OD_600_ = 0.05 in 7H9, pH 7.0 media with 0, 15, 30, 75, or 100 μM DETA NONOate. 1-day post-exposure, culture aliquots were taken and fixed in 4% paraformaldehyde. Reporter signal was analyzed via flow cytometry as previously described [75].

Statistical analyses

GraphPad Prism software was used for all statistical analyses, with p < 0.05 considered significant. The statistical test used for a given assay is described in the figure legends.

Supporting information

S1 FigThere is both overlap and specificity in interactions between STPKs and TCS RRs.Summary table of interactions between the RRs PrrA, DosR, PhoP, and KdpE with various STPKs (kinase domains only for all except PknG and PknK) as determined by M-PFC assays. Results are representative of 2–3 independent experiments.(TIF)

S2 FigAnnotated MS/MS spectrum confirming phosphorylation of DosR T198 after PknH treatment.(A) MS/MS spectrum of the peptide tQAAVFATELK (m/z 629.81, z = 2). Observed b and y ions, along with neutral-loss fragments (-98 Da), are indicated. (B) Fragmentation map displaying detected ions (red = b; blue = y; green = neutral-loss/derived fragments) confirming site localization at T198. Spectra were searched using Mascot v2.8.3 and MSFragger in Scaffold 5.3.3 (precursor tolerance = 10 ppm; fragment tolerance = 0.05 Da; fixed Cys + 57.02; variable phospho +79.97 [STY]; enzyme = trypsin, ≤ 3 missed cleavages; peptide/protein FDR < 1%; peptide probability > 90%).(TIF)

S3 FigAnnotated MS/MS spectrum confirming phosphorylation of DosR T205 after PknH treatment.(A) MS/MS spectrum of the peptide RTQAAVFAtELKR (m/z 524.27, z = 3). Observed b and y ions, along with neutral-loss fragments (-98 Da), are indicated. (B) Fragmentation map displaying detected ions (red = b; blue = y; green = neutral-loss/derived fragments) confirming site localization at T205.(TIF)

S4 FigAnnotated MS/MS spectrum confirming phosphorylation of DosR T198 and T205 after PknH treatment.(A) MS/MS spectrum of the peptide RtQAAVFAtELKR (m/z 550.93, z = 3). Observed b and y ions, along with neutral-loss fragments (-98 Da), are indicated. (B) Fragmentation map displaying detected ions (red = b; blue = y; green = neutral-loss/derived fragments) confirming site localization at T198 and T205.(TIF)

S5 FigAnnotated MS/MS spectrum confirming phosphorylation of DosR D54 after PknH and acetyl phosphate treatment.(A) MS/MS spectrum of the peptide VPAARPDVAVLdVRLPDGnGIELcR of (m/z 928.5, z = 3). Observed b and y ions, along with neutral-loss fragments (-98 Da), are indicated. (B) Fragmentation map displaying detected ions (red = b; blue = y; green = neutral-loss/derived fragments) confirming site localization at D54.(TIF)

S6 FigChanges in DosR phosphorylation status alter DNA binding affinity to the promoter of its target gene *fdxA.*Electrophoretic mobility shift assays (EMSAs) using purified recombinant C-terminally 6x-His-tagged DosR and IRDye 700-labeled probes for the fdxA promoter are shown. A control with no protein (“NP”) added is shown for each gel. DosR was added at indicated concentrations for all other lanes. 40 fmoles of fdxA promoter DNA was used in each reaction. EMSAs shown are as follows: (A) untreated DosR, (B) DosR incubated with 50 mM acetyl phosphate (AcP), (C) DosR phosphorylated “on-bead” with 1 µM PknH, (D) DosR phosphorylated “on-bead” with 1 µM PknD, (E) DosR phosphorylated “on-bead” with 1 µM PknH, then purified and incubated with 50 mM AcP, and (F) DosR phosphorylated “on-bead” with 1 µM PknD, then purified and incubated with 50 mM AcP. Data are representative of 3 independent experiments.(TIF)

S7 FigDosR binding to the hspX and fdxA promoters is specific.Electrophoretic mobility shift assays (EMSAs) using purified recombinant C-terminally 6x-His-tagged DosR and IRDye 700-labeled probes for the hspX promoter (A and C) and the fdxA promoter (B and D) are shown. A control with no protein (“NP”) added is shown for each gel. Purified DosR was added at 1.57 µM for all other reactions in the hspX promoter EMSAs, and at 6.25 µM for the fdxA promoter EMSAs. 1 fmole of each labeled promoter DNA was used in each reaction. “+” are reactions with DosR and the indicated labeled probe, with no competitor unlabeled probe. Where noted, unlabeled specific competitive hspX (A) or fdxA (B) probes were added at the indicated fold molar excess. In (C) and (D), unlabeled non-specific rv2390c promoter probes were added at 200-fold molar excess for reactions in the “rv2390c comp” lane.(TIF)

S8 FigMutation of DosR T198/T205 sites render the protein non-functional.(A) shows an EMSA using purified recombinant C-terminally 6x-His-tagged DosR-T198A/T205A and IRDye 700-labeled probes for the hspX promoter. A control with no protein (“NP”) added is also shown. DosR-T198A/T205A was added at indicated concentrations for all other lanes. 40 fmoles of hspX promoter DNA was used in each reaction. Data are representative of 3 independent experiments. (B) shows a Spinach RNA aptamer assay run with the fdxA promoter with different concentrations of indicated DosR protein. The WT DosR and PknH-phosphorylated DosR (“PknH-DosR) data are as shown in Fig 4A. Fluorescence (arbitrary units, “AU”) was tracked over time on a plate reader, and steady-state rate calculated. Data are shown as means ± SEM from 2-8 experiments. The numerical data underlying the graph shown in this figure are provided in S1 Data.(TIF)

S1 DataNumerical data underlying the presented graphs.Excel file with numerical data underlying graphed average data presented.(XLSX)