Strains and plasmids used in this study are listed in Table 1. E. coli strains were cultivated either on a solid or in a liquid Luria-Bertani (LB) medium at 37°C (Sambrook et al., 1989). Streptomyces coelicolor M145 was cultivated at 30°C on R2YE agar or Mannitol Soy flour (MS) agar (Kieser et al., 2000). For growth in liquid medium, complex S-medium (Okanishi et al., 1974), and defined Evans medium (Evans et al., 1970) was used. Carbon to nitrogen ratio was set as follows: for N-limitation in Evans medium C:N of 2 and for N-excess in Evans medium C:N of 60. Media was supplemented when appropriate with: ampicillin (150 μg/ml), kanamycin (50 μg/ml), chloramphenicol (25 μg/ml), nalidixic acid (255 μg/ml) or thiostrepton (12.5 μg/ml), unless otherwise stated. Genetic manipulation of S. coelicolor M145 and E. coli was performed as described by (Kieser et al., 2000) and (Sambrook et al., 1989), respectively.

For the transcriptional analysis experiments, the S. coelicolor M145 wild type and the glnR mutant were grown in the complex S-medium for 4 days at 30°C. After 4 days, cells were harvested and washed twice with the defined Evans medium without N-source to remove traces of the S-medium. The biomass was subsequently transferred into the defined Evans medium supplemented with variable concentrations of either ammonium (low: 5 mM, high: 100 mM) or nitrate (low: 5 mM, high: 100 mM) and glucose (low: 2.5 g/l or high: 25 g/l). RT-PCR was conducted using total RNA isolated from S. coelicolor M145 and the glnR mutant after 24 h of growth in defined Evans medium. The RNA isolation was performed with an RNeasy kit (Qiagen). All RNA preparations were treated twice with DNase (Fermentas). First, an on-column digestion was carried out for 30 min at 24°C, and afterwards RNA samples were treated with DNase for 1.5 h at 37°C. RNA concentrations and quality were checked using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific). The cDNA from 3 μg RNA was generated with random nonamer primers (Sigma), reverse transcriptase and cofactors (Fermentas). The reverse transcription products (1 μl) were then used as template for PCR amplification. A standard PCR protocol using Taq DNA polymerase (GENAXXON bioscience) and primers annealing to internal parts of the various genes was used. Primers targeting hrdB were used as positive controls for RNA quality. Annealing temperatures were optimized for each primer combination. PCR reactions were performed with the primers listed in Table 2. The PCR conditions were as follows: 95°C for 5 min; 35 cycles of 95°C for 15 s, 55–60°C for 30 s and 72°C for 30 s, and 72°C for 10 min. Negative controls containing nuclease free water and total RNA were performed to exclude any DNA contamination. Positive controls containing total genomic DNA from S. coelicolor M145 were performed to ensure specific amplification of the PCR product. The PCR products were separated during electrophoresis on 2% agarose gels. All reverse transcription/PCR reactions were carried out in triplicate using RNA isolated from three independent cultivations.

For Strep-GlnR purification, S. coelicolor M145 carrying pGM-Strep-glnR overexpression strain (Table 1) was grown for 4 days in complex S-medium at 30°C. Subsequently, cells were harvested and washed twice with nitrogen-free Evans medium. Washed cell biomass was transferred into Evans medium supplemented with 5 or 100 mM NH₄Cl and 5 or 100 mM NaNO₃ as a sole nitrogen source, respectively. The expression of Strep-GlnR was induced with 12.5 μg/ml thiostrepton for 36 h. After cultivation cells were harvested and washed with a solution of 100 mM Tris and 150 mM NaCl (pH 8.0). In order to prevent phosphatase and protease activity, 5 mM sodium fluoride and 5 mM orthovanadate and the EDTA-free cOmplete protease inhibitor cocktail (Roche) were added to the buffer. Cell lysis was performed by Emulsifex (Avestin, Ottawa, Canada) with three consecutive passages. Cell debris and insoluble proteins were separated from the soluble fraction by centrifugation (60 min, 14800 g, 4°C). Soluble proteins were loaded onto a pre-equilibrated Gravity flow Strep-Tactin^®; Sepharose^®; column for one-step purification of recombinant Strep-tag^®; proteins (1-ml bed volume; IBA, Germany). Strep-GlnR was competitively eluted using elution buffer supplemented with 2.5 mM desthiobiotin. Concentrated fractions containing the pure Strep-GlnR were stored at 4°C.

Oligonucleotide primers (Table 2) were designed to incorporate an NdeI and BamHI restriction sites into PCR-fragments containing cobB1 (SCO0452) and cobB2 (SCO6464) genes. The PCR products were cloned into pET15b (Novagen, UK) to generate plasmids pET15b-CobB1 and pET15b-CobB2 used to transform E. coli BL21 (DE3). The over-night cultures of E.coli BL21 pET15b-CobB1 and E.coli BL21 pET15b-CobB2 were used to inoculate 500 ml fresh LB containing 12.5 μg/ml chloramphenicol and 50 μg/ml ampicillin. At an OD₅₇₈ of 0.3, the cells were induced for the expression of His-CobB1 and His-CobB2 with of 0.1 mM IPTG and further incubated at 30°C for 5–10 h. All subsequent procedures were performed at 4°C. Cells were harvested by centrifugation at 5000 rpm for 30 min and resuspended in lysis buffer [50 mM Tris-HCl buffer; pH 7.5; supplemented with the EDTA-free cOmplete protease inhibitor cocktail (Roche)]. Cell lysis was performed by Emulsifex (Avestin, Ottawa, Canada). Cell debris and insoluble proteins were separated from the soluble fraction by centrifugation (30 min, 14,800 g, 4°C). Soluble proteins were loaded onto a preequilibrated Ni²⁺-nitrilotriacetic acid (NTA)-agarose column (1-ml bed volume; IBA, Germany), and His-CobB1 and CobB2 were competitively eluted using elution buffer supplemented with 200 mM imidazole. Fractions containing the pure proteins were pooled and desalted using a HiTrap desalting column (GE Healthcare) with 20 mM Tris-HCl buffer. Concentrated fractions His-CobB1 and His-CobB2 were stored at 4°C.

S. coelicolor M145 was grown, harvested and lysed as described above. The protein concentration was determined using a Nanodrop (PEQLAB, Erlangen, Germany). Proteins were separated by SDS-PAGE (Laemmli, 1970) and transferred to a nitrocellulose membrane (Roth, Karlsruhe, Germany) by semidry electroblotting (PEQLAB) using transfer buffer (25 mM Tris, 150 mM glycine, 20% methanol, pH 9.2) for 30 min at 400 mA. The membrane was blocked with TBST buffer with 5% BSA at room temperature for 1 h. Subsequently the membrane was incubated at room temperature for 1 h with rabbit anti-GlnR polyclonal antibodies (SEQLAB, Göttingen, Germany) or overnight at 4°C with rabbit anti-N-acetyl lysine polyclonal antibodies (Cell Signaling TECHNOLOGY) in TBST buffer for GlnR detection (10 mM Tris pH 8, 150 mM NaCl, 0.05% Tween 20 supplemented with 2.5% of milk powder) or TBST buffer for N-acetyl-lysine detection (25 mM Tris pH 8.0, 125 mM NaCl, 0.1% Tween 20 supplemented with 3% of BSA), respectively. Membranes were washed with TBST (four times) for 5 min and the binding of the primary antibodies against GlnR protein or N-acetylated lysine was detected using anti-rabbit IgG horse-radish-peroxidase-conjugated antiserum (BIORAD, München, Germany) solved in TBST buffer. After 2 h of incubation at ambient temperature the binding of secondary antibodies against rabbit antibodies was detected using the ECL Western blotting detection system (GE Healthcare, München, Germany).

DNA fragments containing promoter regions of GlnR target genes were amplified with Taq polymerase (GENAXXON bioscience, Germany) using genomic DNA from S. coelicolor M145. For this, genomic DNA was isolated with the NucleoSpin® Tissue Kit (Macherey-Nagel, Düren, Germany). Primer sequences, PCR and labeling conditions were carried out as reported in Tiffert et al. (2008). All EMSA reactions were carried out in EMSA buffer (100 mM Tris, 150 mM NaCl, 10 mM β-mercaptoethanol, pH 8) containing an excess of unlabeled nonspecific salmon sperm DNA. The Cy5-labeled target DNA (2 ng) and 4 μg of Strep-GlnR protein or 50 μg of the cleared cell lysate from glnR mutant were dissolved in EMSA buffer and incubated for 10 min at 24°C. After incubation, loading buffer (0.25 × TBE buffer and 60% glycerol) was added and the fragments were separated using gel electrophoresis on 2% TAE agarose gels. DNA bands were visualized by the fluorescence imaging using a Typhoon Trio+ Variable Mode Imager (GE Healthcare).

Deacetylation of Strep-GlnR was performed in the presence of NAD⁺ as a substrate for the deacetylase. For this reaction, 100 μg of Strep-GlnR in 50 mM Tris-HCl pH 8.3, were incubated with 1 mM NAD⁺ for 6 h at 30°C in the presence and absence of 20 μg of His-CobB1 or His-CobB2. Acetylated and deacetylated Strep-GlnR samples were analyzed by immunoblotting using rabbit polyclonal anti-N-acetyl lysine antibodies (Cell Signaling TECHNOLOGY).

Purified Strep-GlnR was—either in solution or in gel—digested with trypsin (1:100 w/w) as described previously (Borchert et al., 2010). Ten percentage of the peptide mixtures in the resulting in-solution digests were directly analyzed by LC-MS/MS. Additionally, the tryptic peptides were subjected to titanium dioxide chromatography to enrich detection of the phosphorylated peptides. For phosphopeptide enrichment acetonitrile was added to the peptide mixture to a final concentration of 30% and the pH was adjusted to 2–3. Enrichment of phosphopeptides by titanium dioxide chromatography was done as described previously (Olsen and Macek, 2009) with the following modifications: phosphopeptide elution from the beads was performed three times with 100 ml of 40% ammonia hydroxide solution in 60% acetonitrile at a pH > 10.5.

For peptide analysis a Proxeon Easy-LC system (Proxeon Biosystems, Odense, Denmark) coupled to a LTQ-Orbitrap-XL (Thermo Fisher Scientific, Bremen, Germany) equipped with a nanoelectrospray ion source (Proxeon Biosystems) as described previously (Koch et al., 2011) was used. The five most intense precursor ions were fragmented by activation of neutral loss ions at −98, −49, and −32.6 relative to the precursor ion (multistage activation). Mass spectra were analyzed using the software suite MaxQuant, version 1.0.14.3 (Cox et al., 2009). The data were searched against a target-decoy Streptomyces coelicolor database containing 8154 forward protein sequences and 262 frequently observed contaminants. Trypsin was set as protease in which two missed cleavage sites were allowed. Beside acetylation at the N-terminus of lysine and oxidation of methionine, phosphorylation of serine, threonine, and tyrosine were set as variable modifications. Carbamidomethylation of cysteine was set as fixed modification. Initial precursor mass tolerance was set to 7 parts per million (ppm) at the precursor ion and 0.5 Da at the fragment ion level. Identified peptides were parsed using the identify module of MaxQuant and further processed for statistical validation of identified peptides, modified sites and protein groups. False discovery rates were set to 1% at peptide, modified site, and the protein group level. To assign a phosphorylation and acetylation site, respectively, to a specific residue a minimal reported localization probability of 0.75 was set as a threshold. The fragmentation spectra of potential modified peptides were manually validated for presence of phosphorylation and acetylation sites.

The GlnR regulator controls genes related to N-assimilation including genes involved in the ammonium uptake (operon amtB-glnK-glnD), nitrate and nitrite assimilation (nnaR, nasA, nirB, narK) and synthesis of the central metabolic nitrogen donors glutamine and glutamate (glnA, glnII, and gdhA; Tiffert et al., 2008; Amin et al., 2012). Although many genetic studies on the regulation of the N-assimilatory genes in S. coelicolor and other actinobacteria have been performed, the regulation of the GlnR activity itself is still enigmatic. In a first step, we aimed to show how glnR itself and selected GlnR target genes are regulated at the transcriptional level upon growth in four defined N-conditions as well as in a complex medium. For this purpose, RT-PCR was performed using total RNA isolated from S. coelicolor M145 and the glnR mutant. The strains were grown for 4 days in the complex S-medium to obtain high biomass. Cells were harvested by centrifugation and washed with Evans medium to remove traces of S-medium. Subsequently the biomass was transferred into defined Evans medium containing different N-sources (ammonium chloride or sodium nitrate) at 100 mM (N-excess) or 5 mM (N-limitation) or into complex S-medium. All cultures were further cultivated for 24 h, at 30°C. Total RNA isolated from S. coelicolor M145 and the glnR mutant was used to generate cDNA. Subsequently, RT-PCR analysis using internal primers for glnR and selected GlnR target genes was performed. The hrdB (encoding the essential principal sigma factor of RNA polymerase) was used as an internal standard due to its relatively constant levels of expression throughout the growth (Buttner et al., 1990). All reverse transcription/PCR reactions were carried out in triplicate using RNA isolated from three independent cultures (for details see Section RT-PCR). For the transcriptional analysis the following GlnR target genes encoding proteins involved in ammonium and nitrate assimilation were selected: glnA, glnII (encoding glutamine synthetase GSI and GSII, respectively), amtB (encoding an ammonium transporter), and nirB (encoding a nitrate reductase). The transcriptional analysis confirmed that GlnR enhanced the expression of glnA, glnII, amtB, and nirB under low concentration of ammonium chloride, whereas high ammonium chloride concentration inhibited expression of these genes (Figure 1A). To verify whether this regulatory effect was ammonium chloride dependent, transcript levels for the selected GlnR target genes were also analyzed in the presence of sodium nitrate. Our results showed that expression of all tested genes glnA, glnII, amtB, and nirB was strongly induced in condition of low nitrate concentration in S. coelicolor M145 (Figure 1B). At last, transcriptional analysis revealed that the glnA, glnII, amtB, and nirB were not expressed in S. coelicolor M145 grown in S-medium (Figure 1C). Expression of glnA, glnII, amtB, and nirB was totally abolished in the glnR mutant under all tested conditions, indicating that GlnR was necessary for their expression. These results indicate that expression of the selected GlnR target genes was strictly regulated by GlnR whose activity seemed to be modulated according to the N-status of the cell. Transcriptional analysis showed that the glnR transcript was present under all tested conditions, demonstrating that glnR is not regulated at the transcriptional level in the response to changing N-concentrations. These analyses led us to assume that GlnR regulatory activity might be modulated by post-translational modifications.

In order to verify whether the GlnR protein was present (as its transcript) in all tested conditions, Western blot analysis was performed. As a control cell lysates from the glnR mutant as well as purified Strep-GlnR protein were used. For this, S. coelicolor M145 was grown in S-medium for 4 days at 30°C; cells were washed twice with Evans medium to remove traces of the complex S-medium. The biomass was transferred into defined Evans medium and further cultivated for 36 h at 30°C. S. coelicolor M145 cells from S-medium and Evans medium were harvested and disrupted. As a negative control cell lysate generated from the glnR mutant grown in S-medium was used. As a positive control Strep-GlnR overexpressed and purified from the S. coelicolor M145 grown in S-medium was used. Clarified cell lysates (200 μg of total protein) as well as isolated Strep-GlnR protein (20 μg) were run on a 12.5% SDS polyacrylamide gel and subsequently transferred onto a nitrocellulose membrane. Signals for the native GlnR and Strep-GlnR were detected by using anti-GlnR antibodies. Western blot analysis revealed the presence of native GlnR in S. coelicolor M145 under all tested conditions, whereas no signal was detected in the cell lysate from the glnR mutant. No sign of GlnR proteolysis could be detected under N-limitation or proficiency. The predicted size of the GlnR protein calculated from the amino acid sequence is 29.8 kDa. Interestingly, GlnR appeared as a double band with the estimated size of ~35 and 38 kDa (Figure 2). This indicated that this regulator may undergo post-translational modification.

In order to determine whether other regulators besides GlnR other regulators are able to recognize and interact with selected promoters of the GlnR-target genes, comparative EMSAs were performed using cell lysates from the glnR mutant and Cy5-labeled promoter regions of P_glnA, P_amtB, and P_glnII. To do so, the glnR mutant was grown in S-medium for 4 days at 30°C; cells were washed twice with Evans medium, transferred into Evans medium or directly transferred in S-medium and further cultivated for 36 h at 30°C. Cells were harvested, disrupted and the cell lysates (50 μg of total protein) were used for EMSAs. No shifts were observed with cell lysates generated from the glnR mutant (Figure S1), indicating that solely GlnR was able to recognize tested promoters under studied conditions.

Post-translational modifications of GlnR were identified by LC-MS/MS using Strep-GlnR overexpressed and isolated from S. coelicolor M145 grown in the complex and N-rich S-medium. The purified Strep-GlnR samples were run on a 12.5% SDS polyacrylamide gel and subsequently stained with Coomassie blue. Finally, the band corresponding to Strep-GlnR was cut out from the gel and subjected to proteolytic digestion with trypsin. In addition to a direct measurement of tryptic Strep-GlnR peptides, phosphopeptides were selectively enriched using titanium dioxide affinity chromatography prior to LC-MS/MS analysis. To specifically detect peptides carrying phosphorylated and acetylated residues, the LC-MS/MS spectra of the modified peptides were compared to the intensities of spectra of the corresponding non-modified peptides. Phosphorylation and acetylation sites were considered as resulting from high confidence phosphorylation and acetylation events only if the LP was higher to 0.75 (or equal), the PEP score was lower than 0.01 (or equal) and the Mascot score higher than 39 (or equal; for details of this evaluation see Materials and Methods, in the Section Nano LC-MS/MS Analysis of the Purified Strep-GlnR). Mapping of the post-translational modifications on the Strep-GlnR purified from S. coelicolor M145 grown in this medium revealed peptides carrying phosphorylated serine and threonine residues: Ser 133, Thr 138, Ser 207, Thr 211, Thr 256, Ser 264/265, and only one acetylated residue, Lys 142 (Table 3 and Supplementary Material Data Sheets 1–4).

Post-translational modifications of GlnR under defined N-conditions were identified by LC-MS/MS using Strep-GlnR isolated from S. coelicolor M145 grown in Evans media supplemented with either 100 mM NaNO₃ (nitrate proficiency) or 5 mM NaNO₃ (nitrate limitation) as a sole N-source. Preparation of the Strep-GlnR samples, LC-MS/MS analysis and data processing were performed as stated in the Section Different GlnR Post-translational Modification Patterns Were Detected under Complex N-Rich Conditions and N-Defined Conditions. Strep-GlnR isolated from cells grown under nitrate proficient conditions revealed only two phosphorylated residues: Ser 140 and Thr 256, whereas no serine/threonine phosphorylation was detected in Strep-GlnR isolated from cells grown under nitrate limitation. Interestingly, Strep-GlnR isolated from S. coelicolor M145 cultivated in condition of nitrate proficiency or limitation exhibited identical acetylation pattern (Table 3 and Supplementary Material Data Sheets 1–4). The following lysine residues were acetylated in Strep-GlnR independent of the nitrate concentration in the Evans medium: Lys 142, Lys 153, Lys 159, and Lys 200, whereas only Lys142 was acetylated when Strep-GlnR originated from culture in S-medium. The Strep-GlnR isolated from cells grown in defined Evans medium supplemented with nitrate as a sole N-source revealed higher acetylation level than the Strep-GlnR from complex S-medium. Most of the acetylated and phosphorylated GlnR residues were localized within the helix turn helix motif involved in the DNA recognition and binding. These modifications are thus likely to have an impact on the GlnR DNA-binding activity.

GlnR was previously shown to specifically bind numerous promoters of genes that encode proteins involved in N-assimilation in S. coelicolor M145. GlnR binding boxes were defined and localized in the promoter regions of glnA, glnII, amtB, nirB, and other genes encoding proteins involved in N-metabolism (Tiffert et al., 2008; Amin et al., 2012). In order to determine whether the differently modified GlnR protein isolated from S. coelicolor M145 could recognize and interact with the selected target promoters comparative electrophoretic mobility shift assays (EMSAs) were performed using the Cy5-labeled promoter regions P_glnA, P_amtB, P_glnII, P_nirB. Three different modified forms of Strep-GlnR isolated from S. coelicolor M145 grown either under complex nitrogen rich conditions (Strep-GlnR_N++) or nitrogen defined conditions (nitrate limited—Strep-GlnR_N−; nitrate excess—Strep-GlnR_N+) were used. Post-translational modifications of Strep-GlnR_N++, Strep-GlnR_N−, and Strep-GlnR_N+ were confirmed by LC-MS/MS prior to EMSAs analysis. Multiply phosphorylated Strep-GlnR_N++ was not able to interact with the promoter regions P_glnA, P_amtB, P_glnII, and P_nirB indicating that phosphorylation inhibits the binding of Strep-GlnR_N++ to the target DNA. Interestingly, multiply acetylated GlnR (Strep-GlnR_N−) was able to interact with and to shift all tested promoter regions. Finally, the multiply acetylated Strep-GlnR_N+ also phosphorylated on the Ser 140 and Thr 256 residues, generated diffuse shifts with P_glnA, P_amtB, P_glnII, and P_nirB, suggesting different binding of the GlnR or instability of the GlnR-DNA complex (Figure 3). The Ser/Thr phosphorylations altered the in vitro DNA-binding activity of GlnR purified from cells grown under N-excess conditions while the multiple acetylations did not inhibit the formation of GlnR-DNA complexes under N-limiting conditions. These results are consistent with the transcriptional analysis of the GlnR-target genes under N-limited and excess conditions reported in Results Revisiting the Transcriptional Regulation of glnR and GlnR-Target Genes under Defined and Complex Nitrogen Conditions. High induction of the expression of GlnR target genes under nitrogen limitation is thus achieved by the unphosphorylated GlnR as shown by the LC-MS/MS analysis.

Interestingly, the Ser/Thr phosphorylation of GlnR occurs in condition of N-proficiency whereas acetylation is independent of the N-status of the cell. To study the impact of the acetylation on the GlnR DNA-binding activity, we attempted to remove the acetyl groups from GlnR by an enzymatic deacetylation. S. coelicolor M145 possess two genes (SCO0452 and SCO6464) encoding enzymes annotated as a sirtuin-like (deacetylase-like). SCO0452 named CobB1 is most similar to human sirtuin SIRT4 and was functionally characterized as NAD⁺-dependent deacetylase from S. coelicolor (Mikulik et al., 2012). The SCO6464 designated as CobB2, shares significant homology with human SIRT5 (Moore et al., 2012). The CobB2 homolog in S. erythraea (with 68% similarity on the protein level) was shown to catalyze deacetylation of acetyl-CoA synthetase AcsA in vitro (You et al., 2014). Bacterial sirtuins are able to deacetylate a large number of target proteins (Castaño-Cerezo et al., 2014) and shows no preference for enzymatic and nonenzymatic lysine acetylation substrate sites (AbouElfetouh et al., 2014). We therefore assumed that the CobB1 and CobB2 deacetylases from S. coelicolor might be potentially able to deacetylate GlnR. Overexpression of N-terminally His₆-tagged CobB1 and CobB2 (His-CobB1 and His-CobB2) was achieved with the IPTG-inducible system in E.coli BL21. Acetylated Strep-GlnR, isolated from S. coelicolor M145 grown under N-limiting conditions was used as a substrate for the in vitro deacetylation assays. A successful GlnR deacetylation was proved by Western blot analysis. Signals for acetylated Strep-GlnR (Ac+) protein and deacetylated Strep-GlnR (Ac−) were detected by anti-GlnR antibodies and anti-N-acetyl lysine antibodies. Western blot analysis revealed that the deacetylase CobB2 was able to remove the acetyl groups from Strep-GlnR in the in vitro assay (Figure S2). EMSAs performed with double-stranded Cy5-labeled selected promoter regions of P_glnA, P_amtB, P_glnII, and P_nirB showed that both the Strep-GlnR (Ac−) as well as the Strep-GlnR (Ac+) were able to interact with the target DNA (Figure 4). However, the Strep-GlnR (Ac+) generated more retarded shift than did the deacetylated Strep-GlnR (Ac−), suggesting that acetylation changes the DNA-binding affinity of GlnR.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.