Oligonucleotides were obtained from Bonsaitech (Madrid, Spain) and Taq polymerase from 5PRIME (Gaithersburg, MD, USA). Escherichia coli cells (M15 strain), plasmids, and plasmidic DNA purification systems were obtained from Qiagen (Hilden, Germany). Genomic DNA purification system was purchased from Macherey Nagel (Düren, Germany), Ni²⁺-Chelating Sepharose™ Fast-Flow, and isopropyl-thio-β-D-galacto-pyra- noside (IPTG) were obtained from GE Healthcare (Uppsala, Sweden). Anti-His tag antibody and anti-rabbit and anti-mouse IgG were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other chemicals were of analytical or molecular biology grade.

The thermophilic, unicellular, non-diazotrophic cyanobacterium Thermosynechococcus elongatus BP-1 (NIES- 2133) was originally obtained from Prof. M. Rögner (University of Bochum, Germany). Cells were grown photoautotrophically at 48˚C in 250 ml Erlenmeyer flasks containing 100 ml of sterile BG11 [22] mineral medium. Flasks were shaken at 140 - 180 rpm and continuously illuminated by fluorescent lamps providing a light intensity of about 20 - 50 µE∙m²∙s⁻¹. Air atmosphere inside the orbital incubator (Kuhner Shaker X, Basel, Switzerland) was enriched with 1% (v/v) CO₂.

Sterile culture medium was used either without nitrogen source added (medium BG11₀) or supplemented in each case with the indicated inorganic nitrogen salt. Nitrate or nitrite was added as sodium salt and sterilized in autoclave. Ammonium was supplemented as sulphate, using a 100 mM stock solution sterilized by filtration (0.2 μm Millex^®-FG filter, Millipore Co., Bedford, MA, USA) before addition to sterile BG11₀ medium. The final concentration of nitrogen in the culture media was 4 mM in all cases.

Protein concentration was determined by the Bradford [23] method using crystalline bovine serum albumin (Sigma-Aldrich) as standard T. elongatus BP-1 growth was assessed by measuring the turbidity of cultures at 730 nm in a DU^®800 UV/Vis spectrophotometer (Beckman Coulter Inc., Fullerton, CA, USA).

T. elongatus BP-1 genomic DNA was extracted and purified using Nucleo Spin Tissue Kit (Machery Nagel). PCR was performed using the oligonucleotides 5’-GTGAGCAACAAAATTGAAGC-3’ and 5’-TTAGGAGACCATCATGGCCA-3’, designed using the T. elongatus BP-1 nirA sequence obtained from genome.kazusa.or.jp/cyanobase. A DNA fragment of 1536 bp carrying the NirA coding region was amplified. DNA was purified from the agarose band using the DNA Gel Extraction Kit (Millipore Co.) and cloned into the vector pQE-30UA using a Ligation Mixture Mix (Qiagen). The resulting plasmid (pQE-30UA-nirA) carried a chimeric gene, which encoded a fusion protein consisting of an N-terminal amino acid tag with six consecutive histidine residues. The final expression plasmid was verified by DNA sequencing Cells of E. coli strain M15 [pREP4] were transformed with the expression plasmid and grown overnight at 37˚C in Luria-Bertani medium (2 × 100 ml) supplemented with ampicillin (100 mg∙l⁻¹) and kanamycin (25 mg∙l⁻¹). Then, this culture was used to inoculate 1 l of medium containing per liter: 10 g tryptone, 5 g yeast extract, 10 g NaCl, 100 mg ampicillin and 25 mg kanamycin. Cells were grown at 37˚C under vigorous orbital shaking. When the A₆₀₀ of cultures reached a value of 0.8, 0.3 mM IPTG was added, and temperature lowered to 25˚C. After 90 min of induction, cells were collected by centrifugation (10 min, at 8000´g, 4˚C) and stored at –20˚C until needed. To isolate His₆NirA, the pelleted biomass from a 1 l of culture was suspended in 25 ml of lysis Buffer A (50 mM Tris-HCl, 300 mM NaCl, 10 mM imidazole, pH 7.3). The cell suspension was incubated for 1 h at 4˚C with 0.1% (w/v) lysozyme, and finally disrupted for 15 min by sonication (Branson Sonifier^® 250, Danbury, CT, USA) at 14% amplitude (60 s on/30 s off).

The cell lysate was centrifuged (30 min, at 9500´g, 4˚C) and the clear supernatant was loaded onto a Ni²⁺- Chelating Sepharose Fast-Flow column equilibrated in buffer A. The column (1.5 cm × 4 cm) was washed with 30 column volumes of 50 mM Tris-HCl, 300 mM NaCl, 20 mM imidazole, pH 7.3. Then, His₆NirA was eluted with a linear gradient of imidazole (from 20 mM to 250 mM) in 50 mM Tris-HCl, 300 mM NaCl, and pH 7.3. The same buffer containing 250 mM imidazole was supplied until the complete elution of enzyme. A flow-rate of 1 ml min^-1 was used in all steps. The elution was monitored continuously at 280 nm, and 1.5-ml fractions were collected and analyzed by SDS-PAGE followed by Western blotting.

SDS-PAGE was performed in a homogeneous 12% (w/v) acrylamide running gel and an upper 4% (w/v) acrylamide stacking gel. Gels were run in a Mini Protean III apparatus (Bio-Rad, Hercules, CA, USA). When proteins were not transferred for immunodetection, bands were visualized after staining with Coomassie brilliant blue R-250. In order to estimate protein molecular sizes, Unstained Precision Plus Protein Standards (Bio-Rad, denoted as M1 in Figure 1(a)) were used. When proteins separated in gels were transferred to nitrocellulose membranes, Prestained SDS-PAGE Standards Broad Range (Bio-Rad, denoted as M2 in Figure 1(b)) markers were used. In this case, prior to transference, membranes and gels were pre-equilibrated for 20 min in transference buffer (0.30% w/v Tris, 1.44% w/v glycine in 20% v/v methanol).

Western blotting analysis was performed after transferring proteins from gels to nitrocellulose membranes for 1 h at 150 mA. The membranes were then incubated for 16 h at 4˚C in blocking buffer containing Tris-buffered saline (TBS, 50 mM Tris, 150 mM NaCl, pH 7.5) supplemented with 1% (w/v) bovine serum albumin and 0.3% (v/v) tween 20. Therefore, corresponding antibodies (anti-His₆ tag or anti-His₆NirA) were added and membranes were incubated for 1 h with gentle shaking. After several washes with TBS, membranes were incubated for 1 h with anti-rabbit or anti-mouse IgG (depending on the assay) conjugated to alkaline phosphatase. Once washed with TBS, membranes were developed using nitro-blue tetrazolium chloride and 5-bromo-4-chloro-3’-indolyphosphate p-toluidine salt (NBT-BCIP^®) (Sigma-Aldrich).

Protein bands of interest were cut from polyacrylamide gels and subjected to tryptic digestion. Once dehydrated with acetonitrile the withdrawn proteins were in situ digested with trypsin (12.5 ng∙µl⁻¹) for 16 h at 37˚C. In tandem mass spectrum were obtained using a SYNAPT HDMS mass spectrometer (Waters Corporation, Milford, MA, USA) connected to a NanoAcquity UPLC (Waters) chromatography system. A sample (8 µl) obtained after tryptic digestion was loaded onto a Symmetry 300 C18, 180 µm × 20 mm (Waters) column. The column was washed with 0.1% (v/v) formaldehyde for 3 min using a 5 µl∙min⁻¹ flow-rate, and connected to a 75 µm × 200 mm, 1.7 µm BEH130 C18 column (Waters) equilibrated with 3% (v/v) acetonitrile containing 0.1% (v/v) formic acid. The flow-rate was adjusted to 0.3 µl∙min⁻¹ and peptides were eluted with a linear gradient (30 min) of acetonitrile from 3% to 60% (v/v). Voltage was adjusted from 3000 to 3500 V.

Mass spectrometric (MS/MS) analysis was performed with the mass spectrometer programmed in mode DDA (Data Dependent Acquisition) in order to fragment the three most intense precursors of each scan within an interval m/z ranging from 350 to 1990, presenting 2, 3 or 4 positive charges and intensities starting from 20 counts/scan. Mass spectra were processed with VEMS program [24] and analyzed using the NCBI database and the on-line form of the MASCOT program (http://www.matrixscience.com).

Samples of His₆NirA purified by IMAC, containing both the complete (58 kDa) and truncated (44 kDa) forms of the recombinant protein, were provided to Abyntek Biopharma, S.L. to immunize a set of rabbits according to standard protocols in order to obtain polyclonal antibodies raised against His₆NirA.

Prior to the expression studies, cells of T. elongatus BP-1 were acclimated to grow with different nitrogen sources through several subcultures in media containing nitrate, nitrite or ammonium as the sole nitrogen source. When cells were acclimated to the corresponding nitrogen source, they were grown in their respective medium until the cultures reached the exponential phase of growth. In this moment, cells were harvested and washed (three times with BG11₀ medium) by centrifugation at room temperature for 5 min, at 1500´g (Allegra^TM X-22R centrifuge, Beckman Coulter Inc.). Finally, aliquots of each type of cells were used to inoculate flaks containing 100 ml of medium BG11₀, either without nitrogen source or supplemented with nitrate, nitrite or ammonium at a final nitrogen concentration of 4 mM. Then, inoculated flasks were incubated in the orbital shaker under the same conditions used to grow cells. Duplicate samples (1 ml) of each culture were withdrawn at 0, 1, 3, 6, 9, 24, 48 and 96 h and deposited in 1.5 ml Eppendorf tubes. One of the collected samples was used to estimate cell density. The biomass from the other one was harvested by centrifugation at room temperature for 5 min at 2000´g (Microfuge^® 18, Beckman Coulter Inc.). The liquid supernatant was discarded and the pelleted cells were immediately frozen and stored at −80˚C. Once all samples were collected, SDS-PAGE loading buffer was added to each tube (proportionally to cell density previously determined) in order to disrupt cells and completely solubilized proteins. Samples were incubated for 5 min at 95˚C, centrifuged for 10 min at 12,000´g, and, finally 10 µl of each supernatant was analysed by SDS-PAGE. After electrophoretical separation, protein detection was carried out by Western blotting using anti-His₆NirA (diluted 1/1000 v/v in TBS) as a primary antibody. The expression results shown in this study have been obtained in triplicate cultures and duplicate Western blotting analyses with consistent results, although only one Western blot membrane is shown. To quantify the proteins revealed in the Western blot membranes, digital images were taken using a Pentax *istD camera. Then, images were analyzed using the TLC Analyzer software [25] available free at http://www.sciencebuddies.org/science-research-papers/tlc_analyzer.shtml, and results were exported and integrated numerically in a spreadsheet.

In this work nitrite reductase from T. elongatus BP-1 has been successfully cloned and expressed as a hexahistidine-tagged protein in host cells of E. coli M15, as detailed in the Materials and Methods section. The recombinant protein was purified by a procedure which is fast, easy to accomplish and yields about 6 mg of a pure preparation of soluble His₆NirA per liter of culture. The protein was strongly retained by the IMAC resin and eluted at the end of the imidazole gradient (about 250 mM imidazole). Analysis of fractions by SDS-PAGE followed by Western blotting (Figure 1) revealed that those fractions containing His₆NirA (with a theoretical molecular mass of 58,487 kDa), contained another protein of about 44 kDa. Moreover, all fractions containing the 58 kDa recombinant protein showed NirA activity (measured with reduced dithionite-methyl viologen as reported by Arizmendi and Serra [11] , indicating that the His₆NirA expressed in E. coli cells is active (data not shown). Both the 58- and 44-kDa proteins were maintained soluble in diluted preparations stored at 4˚C for at least 7 days without apparent loss of their NirA activity. However, they showed a strong tendency to form aggregates and precipitate at concentrations above 1 mg∙ml⁻¹. Thus, 2 mM nitrite was added to the samples in order to enhance the stability of the soluble proteins.

In order to ascertain that the 44 kDa-protein was not the result of intact His₆NirA proteolysis occurring during the purification procedure, all purification steps were carried out at 4˚C and a protease inhibitor cocktail was added. Even under these conditions, the 44 kDa-protein always coeluted with the complete 58 kDa form (results not shown). Western blotting analysis of samples of E. coli cells during the induction of protein expression by IPTG confirmed that the 44 kDa-protein also carries the extension of hexahistidine residues, and that can be overexpressed in E. coli cells after the addition of IPTG (Figure 2). Mass spectrometry analyses of the protein bands sliced out from the SDS-polyacrylamide gel corresponding to the 44 kDa and 58 kDa forms of His₆NirA (Figure 3) are in agreement with former results, indicating that the 44 kDa-protein is a truncated form of the 58 kDa His₆NirA lacking a polypeptide located at the carboxyl domain.

The specificity of the obtained rabbit polyclonal antibodies (rose against the purified complete and truncated forms of His₆NirA) to recognize NirA in extracts of wild-type cells of T. elongatus BP-1 was evaluated. Several dilutions (1:10 and 1:20 v/v) of cell extracts (initially containing 100 µg protein ml⁻¹) were analyzed by SDS- PAGE followed by Western blotting using different antibodies dilutions (1/500, 1/1000 and 1/2000 v/v). Western blotting analysis revealed the presence of two bands of NirA whose molecular masses matched well with those previously calculated for the complete (58,487 kDa) and truncated (approx. 44 kDa) forms of the recombinant His₆NirA.

Anti-His₆NirA antibody was used to assess the expression of NirA in wild-type cells of T. elongatus BP-1 growing in media containing different inorganic nitrogen sources. Due to the fact that these polyclonal antibodies recognize both the complete and the truncated form of NirA, the expression of both proteins was monitored simultaneously with the aim of detecting variations due to the nitrogen source used and/or the time-course of their expression in cells. For this purpose, cells were initially grown in BG11₀ medium supplemented with ammonium, nitrite or nitrate as the sole nitrogen source. At the exponential phase of growth, cells of each culture were subsequently used to inoculate (using similar initial cell density, A₇₃₀ = 0.14 ± 0.04, in all cases), fresh BG11₀ medium supplemented with ammonium, nitrite, nitrate or without nitrogen source (N-free medium). In all cases, growth was followed by determining, at the indicated times, cell density of cultures growing in the four different nitrogen sources indicated above (Figure 4, upper panels) and which were inoculated with cells grown in ammonium (Figure 4(a)), nitrite (Figure 4(b)) or nitrate (Figure 4(c)), respectively. Simultaneously, the time-course appearance of NirA forms was analyzed in the twelve cultures by Western blotting from cell samples withdrawn in the initial 96 h of culture (Figure 4, lower panels).

To investigate NirA expression, cells acclimated to grow with a specific nitrogen source were transferred to media containing another nitrogen source or to a N-free medium. In all cases, cells grew better in media containing ammonium or nitrate than nitrite as the sole nitrogen source (Figure 4, upper panels), no matter which nitrogen source was used to grow cells used as inocula. At the initial stages, growth with nitrite is even slower than that observed in a N-free medium. These results demonstrate once more that nitrite concentration is critical for cyanobacterial growth and that this anion can be toxic if it is accumulated inside cells [14] [26] .

The expression profiles of NirA depended on the inorganic nitrogen source used to grow cells. Thus, both the complete (58 kDa) and truncated (44 kDa) forms of NirA are present in some cases, whereas only one of them can be clearly detected in other situations. Furthermore, expression profiles change not only depending on the

nitrogen source present in the media, but also with the age of culture (Figure 4, lower panels).

Thus, when cultures were growing with ammonium, independently whether cells used as inocula were grown with ammonium, nitrite or nitrate, the truncated NirA is the major form detected in most cases. Complete NirA can also be detected, particularly when cells used as inocula were grown with ammonium, nitrite or nitrate, the truncated NirA is the major form detected in most cases. Complete NirA can also be detected, particularly when cells used as inocula were grown with nitrate. These results indicate that, in the case of T. elongatus BP-1, even when ammonium is the sole nitrogen source available, a negligible amount of NirA could be observed, although the protein is detected mainly as an inactive truncated 44 kDa protein. When cultures were growing with nitrite, both complete and truncated forms of NirA were detected no matter the nitrogen source used to grow inocula.

When cultures were growing with nitrate, independently whether the cells used as inocula were grown with ammonium, nitrite or nitrate, a strong expression of complete NirA was observed, especially from 3 h to 24 h of culture. In most cases the complete protein coexists with the truncated NirA, even though the expression levels of the latest one are always lower. Finally, when cells grown with ammonium, nitrite or nitrate were used to inoculate flasks with fresh N-free media, the complete NirA is the protein mainly expressed in most cases, although cell growth is scarce. These results indicate that NirA protein is expressed in cells of T. elongatus BP-1, even in conditions of nitrogen depletion.

The effect of light during the expression of NirA was assessed in T. elongatus BP-1 cultures either continuously illuminated or subjected to darkness by covering flasks with aluminum foil and incubated under the same conditions. The same regulatory effect was observed in light and dark cultures, although the expression of the complete and truncated forms of NirA appeared enhanced in cultures grow under continuous illumination (data not shown). To confirm the differences observed in the images (Figure 4, lower panels), Western blot membranes were analyzed quantitatively and the obtained results appear in Supplementary Table 1.