Cyanobacterial sulfide-quinone reductase: cloning and heterologous expression

Abstract

The gene encoding sulfide-quinone reductase (SQR; E.C.1.8.5.'), the enzyme catalyzing the first step of anoxygenic photosynthesis in the filamentous cyanobacterium Oscillatoria limnetica, was cloned by use of amino acid sequences of tryptic peptides as well as sequences conserved in the Rhodobacter capsulatus SQR and in an open reading frame found in the genome of Aquifex aeolicus. SQR activity was also detected in the unicellular cyanobacterium Aphanothece halophytica following sulfide induction, with a V(max) of 180 micromol of plastoquinone-1 (PQ-1) reduced/mg of chlorophyll/h and apparent K(m) values of 20 and 40 microM for sulfide and quinone, respectively. Based on the conserved sequences, the gene encoding A. halophytica SQR was also cloned. The SQR polypeptides deduced from the two cyanobacterial genes consist of 436 amino acids for O. limnetica SQR and 437 amino acids for A. halophytica SQR and show 58% identity and 74% similarity. The calculated molecular mass is about 48 kDa for both proteins; the theoretical isoelectric points are 7.7 and 5.6 and the net charges at a neutral pH are 0 and -14 for O. limnetica SQR and A. halophytica SQR, respectively. A search of databases showed SQR homologs in the genomes of the cyanobacterium Anabaena PCC7120 as well as the chemolithotrophic bacteria Shewanella putrefaciens and Thiobacillus ferrooxidans. All SQR enzymes contain characteristic flavin adenine dinucleotide binding fingerprints. The cyanobacterial proteins were expressed in Escherichia coli under the control of the T7 promoter. Membranes isolated from E. coli cells expressing A. halophytica SQR performed sulfide-dependent PQ-1 reduction that was sensitive to the quinone analog inhibitor 2n-nonyl-4-hydroxyquinoline-N-oxide. The wide distribution of SQR genes emphasizes the important role of SQR in the sulfur cycle in nature.

FIG. 1

Sequence alignment of the known SQR enzymes and putative homologs. Multiple amino acid sequence alignment of the known SQR enzymes of A. halophytica, O. limnetica, and R. capsulatus (33) as well as the putative close homologs from A. aeolicus (13), S. putrefaciens (sequence number >4287 from the genome of S. putrefaciens in the TIGR bank), T. ferrooxidans (sequence number >62 from the genome of T. ferrooxidans in the TIGR bank), Anabaena PCC7120 (sequence number C279 from the genome of Anabaena in CyanoBase), and the distant homologs from S. pombe as well as the flavoprotein subunit of A. vinosum flavocytochrome c (FCC). Residues that are identical among at least four of the sequences are indicated by black shading. The conserved FAD binding domains are indicated by overlining. The two fully conserved cysteines are marked by asterisks.

FIG. 1

Sequence alignment of the known SQR enzymes and putative homologs. Multiple amino acid sequence alignment of the known SQR enzymes of A. halophytica, O. limnetica, and R. capsulatus (33) as well as the putative close homologs from A. aeolicus (13), S. putrefaciens (sequence number >4287 from the genome of S. putrefaciens in the TIGR bank), T. ferrooxidans (sequence number >62 from the genome of T. ferrooxidans in the TIGR bank), Anabaena PCC7120 (sequence number C279 from the genome of Anabaena in CyanoBase), and the distant homologs from S. pombe as well as the flavoprotein subunit of A. vinosum flavocytochrome c (FCC). Residues that are identical among at least four of the sequences are indicated by black shading. The conserved FAD binding domains are indicated by overlining. The two fully conserved cysteines are marked by asterisks.

FIG. 2

SQR activity in A. halophytica. The assay mixture contained 10 mM potassium HEPES (pH 7.4), 10 mM MgCl₂, 10 mM KCl, 30 μM PQ-1, and cells containing 3 μg of chlorophyll per ml. The reduction of PQ-1 was detected at A₂₉₂ minus A₂₆₆. Where indicated, the reaction was started by the injection of 60 μM Na₂S. PQH₂-1, reduced PQ-1.

FIG. 3

Identity and similarity of SQR sequences. ah, A. halophytica. The number in the upper left corner of each box is the percent identity (from a Blast search), and the number in the lower right corner is the percent similarity (from a Blast search).

FIG. 4

Overexpression of cyanobacterial SQR in E. coli (A) and solubilization (B) and antibody recognition (C) of overexpressed O. limnetica SQR. (A) Cells of E. coli BL21(DE3) transformed with either pTOLsqr or pTAHsqr were grown at 37°C in LB medium. At an optical density at 600 nm of 0.6, IPTG was added to 0.5 mM. After 2 h, 0.5-ml cultures were centrifuged, resuspended in 2% SDS loading buffer, and boiled for 5 min. Samples of 15 μl were loaded on the gels. Lane 1, noninduced BL21(DE3)/pTOLsqr; lane 2, induced BL21(DE3)/pTOLsqr; lane 3, induced BL21(DE3)/pTAHsqr. (B and C) Coomassie blue staining (B) and Western blotting (C) with polyclonal antibody raised against denatured O. limnetica SQR (2). Lane 1, O. limnetica membranes (containing 2 μg of chlorophyll); lane 2, noninduced BL21(DE3)/pTOLsqr membranes; lane 3, as in lane 2 but IPTG induced; lane 4, DM sup; lane 5, Triton sup; lane 6, Urea sup; lane 7, Urea pellet (see Materials and Methods). All samples loaded in lanes 2 to 7 were derived from membranes containing 40 μg of protein.

FIG. 5

SQR activity in membranes of E. coli expressing A. halophytica SQR (A) and kinetics of PQ-1 reduction and reoxidation in the presence of E. coli membranes (B). (A) Membranes of induced BL21(DE3)/pTAHsqr (60 μg of protein) were incubated under anaerobic conditions in an assay mixture containing 10 mM potassium HEPES (pH 7.4), 10 mM MgCl₂, 10 mM KCl, and 50 μM PQ-1. The reaction was started by the addition of 60 μM Na₂S. NQNO (1 μM) was added where indicated. (B) An assay mixture containing 60 μM H₂S was incubated under anaerobic conditions in either the presence (———) or the absence (–––) of noninduced E. coli membranes (60 μg of protein). Partially purified SQR obtained from O. limnetica thylakoids (40 μg of chlorophyll) was added where indicated. PQH₂-1, reduced PQ-1.

FIG. 6

Distribution of SQRs. Schematic demonstration of the phylogenetic distribution of the SQRs known so far. Species that are already known to possess SQR are underlined. ∗, SQR detected only by membrane ∗∗,  both membrane activity and cloned SQR genes; ∗∗∗, SQR-like ORFs.