The bacterial strains and plasmids used in this work are listed in Table 1. The E. coli strains were used for cloning proposes and protein heterologous expression. They were transformed according to (Sambrook et al., 1989). Transformants were selected on Luria–Bertani (LB) media at 37°C supplemented with the appropriate antibiotics: 50 μg kanamycin (Km) ml⁻¹ and/ or 20 μg chloramphenicol (Cm) ml⁻¹. M. smegmatis mc²155, an electroporation proficient mutant of mc²6 (Snapper et al., 1990), was routinely grown at 37°C in Middlebrook 7H9 (BD-Difco) medium supplemented with 0.03% tyloxapol or in solid medium Middlebrook 7H10 (BD-Difco) supplemented with the appropriate antibiotics if required 15 μg kanamycin ml⁻¹, 20 μg gentamycin ml⁻¹ or 50 μg hygromycin ml⁻¹. For nitrogen analysis, bacteria were grown in modified Sauton’s minimal medium (Jenkins et al., 2013) supplemented with different nitrogen sources. Cells were cultivated in 15 mM ammonium sulfate (Sigma) to log phase and washed twice with nitrogen-free medium before inoculating fresh medium with the required nitrogen source.

Isolation of plasmid DNA, restriction enzyme digestion and agarose gel electrophoresis were carried out by conventional methods (Sambrook et al., 1989). The genomic DNA of M. smegmatis was obtained as described by Connell (1994).

Recombinant M. tuberculosis His-tagged ACCase subunits AccA3, AccD5, AccD6, and AccE5 were purified as described in Gago et al. (2006). The same protocol was applied for His-tagged M. tb PII and Amt_CTR fused to TrxA. After dialysis, TrxA was removed through digestion with TEV protease. In brief, the Amt_CTR-TrxA fraction was incubated with TEV protease (30:1) for 2 h at room temperature, followed by an additional 2 h at 4°C. The mixture was then loaded onto a Ni-NTA column, and cleaved Amt_CTR was subsequently recovered. PII proteins from E. coli, A. brasilense and M. tb were obtained without tags, according to Gerhardt et al. (2015, 2017).

Proteins were analyzed by SDS-PAGE (Laemmli, 1970). Protein contents were determined by measuring A280 nm to the solutions and by the Bradford method, using BSA as standard.

For the production of polyclonal antibodies anti PII, a rabbit was immunized biweekly with purified M. tb PII, emulsified in Freund’s complete adjuvant (Sigma) at a 1:1 (v/v) ratio. Antisera against PII were elicited in rabbit following conventional procedures (Burnette, 1981, Anal. Biochem). The anti-M. tb PII was used to evaluate the presence of M. tb or M. smegmatis PII proteins (88.4% identity), and both proteins were recognized by the antibody (Supplementary Figure S1A).

PII from mycobacteria were detected by Western blot analysis. After electrophoretic separation, proteins were transferred onto nitrocellulose membranes (Bio-Rad). Anti-M. tb PII was used at a 1:300 dilution, and anti-RpoB at 1:10,000 dilution, as a control. Antigenic polypeptides were visualized using an alkaline phosphatase-tagged secondary antibody. For biotinylated proteins western blotting procedure was modified as described by Nikolau et al. (1985). Proteins were probed alkaline phosphatase-streptavidin conjugate (AP-streptavidin diluted 1:10,000) (Bio-Rad).

M. smegmatis mc²155 was grown in 0.5 mM (NH₄)₂SO₄ to an OD 600 nm 0.6–0.8 and half of the culture was subjected to a shock of 15 mM of ammonium sulfate, for 15 min (Atkinson et al., 1994; Durand and Merrick, 2006; Moure et al., 2012; Rodrigues et al., 2014). The adenylation state of PII in cell-free extracts was determined by electrophoresis on SDS- PAGE or native PAGE and followed by Western blotting as described in (Bonatto et al., 2007).

Complex formation between M. tb His-AccA3 and PII without a tag were assessed by pull down as described by Rajendran et al. (2011), with modifications. Pull down assay was performed using 15 μL of magnetic beads (Promega) with interaction buffer (Tris–HCl 50 mM pH 8.0, NaCl 100 mM, glycerol 5%, LDAO 0,01%, imidazole 20 mM; MgCl₂ 5 mM) and saturating ATP (3.5 mM). Fifteen micrograms of purified His-AccA3 were mixed with 20 μg of PII protein from M. tb (Mt), A. brasilense (Ab) or E. coli (Ec). Proteins eluted from the Ni²⁺ were analyzed by SDS-PAGE and the gel was stained with Coomassie Blue.

The interaction between PII and Amt_CTR was analyzed in vitro. 10 μg of cell extracts expressing His-tagged M. tb PII were mixed with 10 μg of purified Amt_CTR without tag. The mixture was prepared in the presence of either 1 mM ATP or 1 mM ATP along with 1.5 mM 2-OG. After incubating the mixture for 1 h at 4°C, it was loaded into a Ni²⁺-NTA column that had been pre-equilibrated with PBS containing ATP, or ATP and 2-OG. The column was then washed with PBS, supplemented with the ATP and 2-OG, using 10 column volumes. Finally, the bound proteins were eluted from the column using PBS containing 250 mM imidazole. SDS-PAGE was performed to analyze the samples obtained from the column.

ACCase activities in cell-free extracts and in vitro reconstituted complexes were determined by following the incorporation of radioactive HCO₃⁻ into acid non-volatile material, as previously described (Diacovich et al., 2002). Substrate concentrations were 0.5 mM for acetyl-CoA or propionyl-CoA. One unit of enzyme activity catalyzed the incorporation of 1 mmol ¹⁴C into acid-stable products min⁻¹.

For the construction of the M. smegmatis mutant ∆MsPII the plasmid pDE13 was used to transform M. smegmatis cells. The mutant strain generated through a double homologous recombination event was selected according to (Crotta Asis et al., 2021).

Nitrite concentration was determined using the Griess reaction (Peter, 1879). For the uptake assay, exponentially growing cells in modified Sauton’s minimal medium supplemented with 1 mM nitrite at an OD 600 nm of 0.6–0.8 were harvested by centrifugation at 3,000 × g for 10 min at 4° C. Afterward, they were washed twice using modified Sauton’s medium that lacked a nitrogen source. The cultures were then resuspended in fresh medium at pH 7.2 or 9 containing 0.2 mM sodium nitrite and 1 mM ammonium sulfate when indicated, and adjusting to an OD of 600 nm of 0.75. Cells were cultivated under constant shaking of 180 rpm at 37° C. At regular intervals of 15 min, aliquots of the cultures were sampled and centrifuged at 5,000 × g for 10 min to remove the cells previous to nitrite quantification.

Crude extracts were obtained from cultures grown in modified Sauton’s minimal medium with 1 mM nitrate. Appropriate amounts of extracts were incubated for 30 min at 30°C in a buffer composed of 0.1 M potassium phosphate (pH 7.5), 0.4 mM NADPH, 0.025 mM FAD, 0.025 mM FMN, and 1 mM NaNO₃. To measure the concentration of accumulated nitrite, the Griess Reaction was performed (Peter, 1879).

Cell-free extracts excised from Coomassie-stained SDS-PAGE gels were subjected to digestion with trypsin. Peptide separations were carried out on a nanoHPLC Ultimate3000 (Thermo Scientific) using a nano column EASY-Spray ES903 (50 cm × 50 μm ID, PepMap RSLC C18). The mobile phase flow rate was 400 nl/min, using 0.1% formic acid in water (solvent A) and 0.1% formic acid and 100% acetonitrile (solvent B). The gradient profile was set as follows: 4–35% solvent B for 30 min, 35–90% solvent B for 1 min and 90% solvent B for 5 min. Three microliters of each sample were injected. MS analysis was performed by using a Q-Exactive HF mass spectrometer (Thermo Scientific). For ionization, 1.9 kV of liquid junction voltage and 250°C capillary temperature was used. The full scan method employed a m/z 375–1,600 mass selection, an Orbitrap resolution of 120,000 (at m/z 200), a target automatic gain control (AGC) value of 3e6, and maximum injection times of 100 ms. After the survey scan, the 7 most intense precursor ions were selected for MS/MS fragmentation. Fragmentation was performed with a normalized collision energy of 27 eV and MS/MS scans were acquired with a dynamic first mass. AGC target was 5e⁵, resolution of 30,000 (at m/z 200), intensity threshold of 4.0e4, isolation window of 1.4 m/z units and maximum IT was 200 ms. Charge state screening was enabled to reject unassigned, singly charged, and equal or more than seven protonated ions. A dynamic exclusion time of 15 s was used to discriminate against previously selected ions.

MS data were analyzed with MaxQuant (V: 2.1.4.00) using standardized workflows. Mass spectra *.raw files were searched against a database from Mycobacterium smegmatis mc2155 (reviewed and unreviewed proteins), entry UP00000075 from Uniprot. The “MSMEG_4206” sequence was added to the database. Precursor and fragment mass tolerance were set to 10 ppm and 0.02 Da, respectively, allowing 2 missed cleavages. Fixed modification: carbamidomethylation of cysteines. Variable modifications: protein N-terminal acetylation and methionine oxidation. The statistical analysis was performed with Perseus v1.6.15.0 software (MaxQuant).

The culture was grown in a modified Sautons’ medium supplemented with 0.5 mM (NH₄)₂SO₄ until reaching an OD of 0.8 at 600 nm. The culture was then split into four fractions. In three of the cultures, (NH₄)₂SO₄ was added to a final concentration of 15 mM, inducing an ammonium shock. After 15 min, the cultures were harvested by centrifugation at 3,000 x g for 10 min at 4°C. The cells were washed twice with PBS and resuspended in 100 μl of PBS containing 1 mM PMSF and either 0.6 mM ADP, or 4.5 mM ATP, and 1.5 mM 2-OG. Then, cultures were lysed using a water bath sonicator (Diagenode), and the lysates were clarified by centrifugation at 11,180 × g for 30 min at 4°C. The resulting cell-free extracts were further centrifuged at 16,000 x g for 1 h at 4°C, and the supernatant, representing the cytosolic fraction, was collected. The pellet, containing the membrane fraction, was washed once with PBS, containing the corresponding additives, and centrifuged again for 30 min at 16,000 g. The membranes were then resuspended in 50 μl of PBS. The samples were analyzed by SDS-PAGE followed by western blotting, performed to detect PII and RpoB, serving as a control.

The genomes of M. tb and M. smegmatis contain one type of PII protein-encoding gene. The PII protein expression in mycobacteria is induced in response to nitrogen limitation (Amon et al., 2008; Williams et al., 2013) and PII is rapidly adenylated by GlnD (Williams et al., 2013). To understand the dynamics of PII protein adenylation in response to the nutritional status, we analyzed the state of PII adenylation under different nitrogen regimen. M. smegmatis mc²155 was grown in modified Sauton’s minimal medium supplemented with 0.5 mM ammonium sulfate (Williams et al., 2013, 2015) to an OD 600 nm of 0.7; at this stage, PII expression can be detected by Western-blot (Supplementary Figure S1A). At this particular OD, the culture was divided into two fractions and one was subjected to a concentration of 30 mM of ammonium, which is considered an ammonium shock (Coutts and Thomas, 2002) After 15 min, cells were collected and protein extracts were analyzed by SDS-PAGE or native gels, and detected by Western blotting. As can be observed in Figures 1A,B, the migration pattern of PII changed after the ammonium shock, with a pattern consistent with the loss of the nucleoside (Atkinson et al., 1994; Durand and Merrick, 2006; Moure et al., 2012; Rodrigues et al., 2014). In native gels, this modification provokes PII to run faster than unmodified PII, due to PII-AMP’s higher negative charge imparted by the AMP residues. On the contrary, in SDS-PAGE, as the gel in Figure 1A, the adenylation of PII produces a delay in protein migration. The impact of adenylation on protein electrophoretic mobility, under denaturing conditions, was validated by LC-MS/MS (data not shown). These results suggest that the adenylation of PII is readily reversible by increasing the availability of ammonium.

In E. coli two PII proteins were described, namely GlnB and GlnK. Both proteins can interact with the BCCP subunit of the acetyl-CoA carboxylase (ACC) complex, but only GlnB is able to interact with the complex inhibiting the ACC holoenzyme activity (Gerhardt et al., 2015). Given that the regulation of ACC by PII is conserved in distantly related organisms, ranging from plants to bacteria (Forchhammer et al., 2022), we hypothesized that such PII function could be also conserved in mycobacteria. In M. tb the acetyl-CoA carboxylase activity is present in two enzyme complexes, named acyl-CoA carboxylases (ACCases) based on their relaxed substrate specificity. These two complexes, ACCase 5 and ACCase 6, consist of two subunits: an α-subunit named AccA3, which is shared by both ACCases, containing the biotin carboxylase (BC) and the biotin carboxyl carrier protein (BCCP) domains, and a specific β-subunit carrying the carboxyltransferase (CT) domain, named AccD5 or AccD6, respectively (Gago et al., 2006; Daniel et al., 2007; Kurth et al., 2009; Pawelczyk et al., 2011; Bazet Lyonnet et al., 2014). A third subunit, named ε, is also required for maximal activity of the ACCase 5 complex. These two enzymes are able to carboxylate both acetyl-CoA (ACC activity) and propionyl-CoA (PCC activity), but with different substrate preferences (Gago et al., 2006; Daniel et al., 2007). On the other hand, in E. coli there is only one type of ACC enzyme of which the BC and BCCP components are encoded by separated polypeptides.

To investigate the interaction between PII and the biotinylated subunit of the ACCase, purified AccA3-His₆ (α-subunit) was immobilized into nickel beads and used as bait which was challenged with purified M. tb PII without tags (Gerhardt et al., 2015). Orthologue PII proteins, GlnB from Azospirillum brasiliense (Ab) and GlnB from E. coli (Ec), which share 60.7 and 61.6% similarity with M. tb PII (Mt), respectively, were used as controls. The pull-down assay data revealed that AccA3 interacted strongly with Ab-GlnB but weakly with Ec-GlnB or Mt-PII (Figure 2A). We also observed that the specificity of the AccA3 PII proteins interaction is dependent on the PII effectors. The interaction between AccA3 and PII is only effective in the presence of ATP, but it does not occur in the presence of ADP (Figure 2A and Supplementary Figure S2A). On the other hand, we also used lysozyme as a non-specific protein and showed that this protein does not interact with AccA3, in the same conditions used to detect AccA3-PII interaction (Supplementary Figure S2B).

In order to analyze if these interactions could affect M. tb ACCase activities, we assayed the propionyl-CoA carboxylase (PCC) activity in vitro using the reconstituted ACCase 5 complex which is composed of AccA3 (BC-BCCP α-subunit), AccD5 (CT), and AccE5 (ε subunit). When Ab-GlnB PII protein was incorporated into the reaction mix, an inhibition of the PCC activity was detected (Supplementary Figure S3), which is consistent with the strong interaction observed in the pull-down assay (Figure 2A). This data support that the interaction between ACCase and PII acts to reduce ACC/PCC activity as described previously in bacteria such as E. coli and Streptomyces hygroscopicus var. ascomyceticus (Wang et al., 2021). However, in the presence of M. tb PII no significant modulation of PCC activity was detected (Figure 2B), which is consistent with the weak interaction between M. tb PII and AccA3 (Figure 2A). Likewise, the ACC activity of the M. tb ACCase 6 complex, which shares the BC-BCCP subunit with ACCase 5, was not affected by the presence of PII either (Figure 2C).

In E. coli the formation of the ACC-GlnB complex is inhibited when GlnB is uridylated (Gerhardt et al., 2015). Hence, we analyzed ACCase activities in M. smegmatis cell-free extracts, grown in limiting nitrogen conditions (1 mM of ammonium) or after the ammonium shock (30 mM of ammonium), conditions were PII is adenylated or de-adenylated, respectively (Figure 1). Under these conditions, no changes were observed in the PCC or the ACC activities of the corresponding ACCases (Figure 3).

Previous high-throughput studies suggested that PII is not essential for the viability of M. tb (Sassetti et al., 2003). Assuming that this would also be applied to M. smegmatis PII coding gene (MSMEG_2426), we constructed a knockout (KO) mutant in M. smegmatis to gain information about the physiological relevance of PII. We disrupted MSMEG_2426 by using an in-frame deletion strategy, which would not have a polar effect on the expression of glnD. For this, we replaced MSMEG_2426, the PII coding gene, with a mutant allele, using a classic two-step homologous recombination approach (Crotta Asis et al., 2021; Figure 4A). The strain obtained was named ∆MsPII, confirming that the MSMEG_2426 gene is not essential for the viability of M. smegmatis. PII was not detected by Western blotting in cell-free extracts from two individual colonies of the putative mutants, while PII was perfectly distinguished in cell-free extracts deriving from wild-type strain, which confirmed that the expected allelic exchange at the chromosomal MSMEG_2426 locus had been successful (Figure 4B). The PII expression levels were partially complemented by the plasmid pDE35, expressing MSMEG_2426 under the control of the constitutive promoter P_hsp60 (Figure 4B); this strain was named ∆MsPIIc. In all the cases, the anti-M. tb PII antibody was used to detect PII proteins from M. tb or M. smegmatis; this antibody was able to well recognize both proteins (Supplementary Figure S1).

To analyze the physiological consequences of altered PII protein levels, the wild-type, ∆MsPII, strains were grown in 7H9 liquid medium and modified Sauton’s minimal medium supplemented with 15 mM or 0.5 mM of (NH₄)₂SO₄ (Williams et al., 2013, 2015). The growth dynamic of the ∆MsPII strain was very similar to that corresponding to the isogenic wild-type strain (Figures 4C–E). Hence, under these growth conditions, the absence of PII does not modify the cell growth rate.

It is well described that PII proteins are involved in the regulation of nitrogen metabolism. To further understand the metabolic role of mycobacterial PII, we grew the three strains, wild-type, ∆MsPII and ∆MsPIIc in Sauton’s minimal media where the unique source of nitrogen was nitrate or nitrite. When the medium was supplemented with 1 mM of nitrite (Akhtar et al., 2013) a clear growth delay was observed in the ∆MsPII strain, with a marked and prolonged lag phase, when compared to the wild-type; however, both strains reached similar final ODs. The complemented strain shows an intermediate behavior, suggesting that the expression of the M. smegmatis from the plasmid pDE35 partially compensates for the loss of M. smegmatis PII expression from the chromosome (Figure 5A). Simultaneously, the concentration of the nitrite remaining in the medium was measured through the Griess assay. The nitrite was not detectable after 33 h in the wild-type cultures, while it took over 41 and 52 h to become undetectable in the ∆MsPIIc and ∆MsPII strains, respectively (Figure 5B).

Similar but strongest differences were observed when the strains were grown in Sauton’s media supplemented with 1 mM of nitrate (Figure 5C). The ∆MsPII strain had a reduced growth rate and lower final ODs when compared to the wild type. To determine whether the nitrite assimilation of ∆MsPII cells was the cause for the lower growth rate in nitrate, we measured the nitrite concentration in extracts of cells that were growing in 1 mM nitrate. In this assay, the three cultures resulted in comparable levels of nitrite, indicating the mutant cells are not accumulating or excreting nitrite (Figure 5D).

These experiments suggest that M. smegmatis PII protein could be involved in the modulation of nitrate and nitrite assimilation. We conclude that PII is not completely required for cell growth but it increases cell fitness when nitrate or nitrate are used as sole nitrogen sources.

The ability to take up nitrite at different pH was analyzed in wild-type, ΔMsPII and ΔMsPIIc cells, previously adapted to nitrite. The three strains presented a similar uptake activity when nitrite was the sole nitrogen source at pH 7.2 (Figure 6A). The addition of ammonium resulted in a diminution of the nitrite uptake in all the strains; however, this effect was retarded in the ΔMsPII strain (Figure 6B). At pH 9, the wild-type strain presented a similar ammonium-dependent inhibition, but this effect was not observed in the ΔMsPII strain. The ammonium inhibition resulted slower in ΔMsPIIc strain (Figures 6C,D). At pH 7.2, nitrite predominantly exists in the form of nitrous acid which could permeate the cell via diffusion. Conversely, at pH 9, the nitrite ions must be transported through the cell envelope by specific permeases (Lee et al., 1998). Therefore, the lack of ammonium-induced inhibition of nitrite uptake at pH 9 in the ΔMsPII strain would imply a potential role of M. smegmatis PII in orchestrating the hierarchical utilization of ammonium over nitrite as a nitrogen source.

To assimilate nitrate into the organic nitrogen pool, it undergoes two consecutive reductions. Initially, nitrate is mainly intracellularly reduced to nitrite by the action of a nitrate reductase activity, and subsequently, nitrite is further reduced to ammonia by the reaction catalyzed by the nitrite reductase (Morozkina and Zvyagilskaya, 2007; Gouzy et al., 2014). In mycobacteria, a unique nitrite reductase enzyme was identified (NirBD) (Malm et al., 2009), while several genes encode for proteins with nitrate reductase activity, including NarGHJI, and NarB. A novel type of nitrate reductase was recently described in M. smegmatis, which was named NasN. This NADPH-dependent diflavin enzyme is uniquely responsible for nitrate assimilation (Tan et al., 2020; Cardoso et al., 2021). We measured the cytoplasmic nitrate reductase activities in cell–free extracts of M. smegmatis wild-type and ∆MsPII strains and found that the protein extracts of the mutant strain had higher nitrate reductase activity than the wild-type (Figure 7). This variation could be the result of protein levels or a direct or indirect regulation of the nitrate reductase activity by PII. NasN is part of the GlnR regulon, the major nitrogen-regulator in M. smegmatis (Jeßberger et al., 2013). Even though there is no evidence that PII and GlnR may interact, PII could be modulating protein expression as seen in other systems (Huergo et al., 2013). Therefore, we compared the protein profiles of both strains when grown in 1 mM nitrate by LC–MS/MS. The proteins with significant differences were screened according to the criteria of │FC│ > 1.5 and p < 0.05 (Figure 8). Though, the amount of a few proteins involved in nitrogen metabolism was altered (Table 2), no differences were observed in the GlnR or NasN levels between these strains, suggesting that the modulation of nitrate reductase activity by PII occurs by posttranslational mechanism. The subunits of the NirBD complex, which is under the regulation of GlnR and NasR, were equally represented in both samples as well. The comprehensive roster of differentially expressed proteins is detailed in Supplementary Tables S1, S2.

The enrichment analysis using a Fisher’s exact test¹ (Ge et al., 2018) of the significantly changed proteins screened from the LC–MS/MS showed that proteins involved in ammonium assimilation pathways were increased, including GlnA and GlnD, as well as proline and arginine metabolism (Supplementary Figure S4). On the other hand, many transmembrane transporters, including putative amino acid ABC transporters, were reduced. Surprisingly, one of these was the Mce4 transport system involved in cholesterol uptake (García-Fernández et al., 2017; Rank et al., 2021), and which is considered a virulence factor in M. tb. In addition, the expression of proteins related to the stress response was also altered.

In many bacteria and Archaea, the PII protein named GlnK, interacts with the ammonia channel AmtB to modulate its function (Thomas et al., 2000; Coutts and Thomas, 2002; Sant'Anna et al., 2009). When the cellular nitrogen status increases, the GlnK-AmtB binding promotes a block of ammonia transport into the cell (Coutts and Thomas, 2002; Zheng et al., 2004; Conroy et al., 2007). Regulation of AmtB activity seems to be the archetypical function of the GlnK protein, also these genes form a conserved glnK-amtB operon in a range of prokaryotes.

Based on both synteny and protein crystal structure analyses, it has been previously proposed that M. tb GlnB may in fact share an analogous function to E. coli GlnK and not GlnB (Shetty et al., 2010). However, the precise function of PII in mycobacteria has not been studied at a biochemical level.

To verify if M. tb PII function as GlnK interacting with Amt, we analyzed the presence of PII on the membrane and cytoplasmic fractions by SDS-PAGE and Western blotting. Cells were grown in ammonium–limiting conditions for 28 h. An aliquot of the culture was incubated in 30 mM of ammonium for 15 min. The membranes were isolated by centrifugation and washed in the presence of ATP, ADP, or ATP and 2-OG. PII was only present in the cytoplasmic fraction in the ammonium-limiting condition. In contrast, after the ammonium shock, PII was found in both the cytoplasmic and membrane fractions, indicating that PII associates with the membrane in an ammonium-dependent manner. This association was weakened by the presence of the PII protein allosteric effectors ATP and 2-OG (Figure 9A).

To determine if the association with the membrane was through Amt transporter, we performed a pull-down assay in which purified M. tb His₆-PII was mixed with a Ni-NTA resin (Novagen) and incubated with purified C-terminal Amt (Amt_CTR), without tags, for 30 min. After three washes with buffer containing different combination of the PII protein allosteric effectors Mg-ATP and 2-OG, the elutions were analyzed by SDS-PAGE. Figure 9B shows that PII is able to interact with Amt_CTR being negatively modulated by 2-OG, in agreement with the PII cellular localization analysis (Figure 9A). These data confirm that M. tb GlnB plays an ancestral role as a GlnK protein, in addition to its function as a modulator of nitrite assimilation.