3.1Cloning of hypF, hypC, hypD, hypE, hypA and hypB– analysis of the deduced protein sequences
The hyp genes were cloned while searching for genes upstream and downstream of hupSL in Nostoc PCC 73102 . The 5′-end of hypF was located at one end of an EcoRI 4.7-kb subfragment of a hupSL-containing cosmid, the other end of which encoded the 5′-end of hupS on the opposite strand. The distance between hupS and hypF was determined to be 3.8 kb by PCR using primers in hupS and hypF, respectively (not shown). The five hyp genes (hypF, hypB, hypA, hypE, hypD; originally published as hup genes) previously identified in Anabaena PCC 7120 were also located upstream of hupS and the deduced proteins are 84–90% similar to the corresponding Nostoc PCC 73102 gene products . Lower similarities (43–63%) of the Nostoc PCC 73102 hyp gene products are observed to the putative Synechocystis PCC 6803 Hyp proteins . The highest similarity of the deduced Nostoc PCC 73102 proteins with identified Hyp proteins can be found with R. capsulatus, being 60, 58, 57, 69, 66, and 55% for HypABCDEF, respectively.
The Nostoc PCC 73102 hypF product (782 amino acids) is characterised by an N-terminal acylphosphatase domain and a double zinc finger motif [C-x2-C-x18-C-x2-C]-x24-[C-x2-C-x18-C-x2-C]. These two signatures are found in all HypF sequences reported so far, except HypF1 of R. eutropha. The zinc fingers in the N-terminal part of HypF are believed to be involved in sensing Ni. Mutants of R. eutropha with a deletion in the recently cloned HypF2, however, exhibit hydrogenase activity, due to the still present HypF1, a truncated HypF lacking the N-terminal half and thus the acylphosphatase and the zinc finger domain. Since a ΔhypF1/hypF2 double mutant does not show any hydrogenase activity , the two N-terminal motifs do not seem to be essential for the Ni sensing and insertion function. Instead, this function of HypF should be located in the C-terminal half. His residues are notorious for their ability to bind Ni, and many of the conserved His residues are located in the C-terminal half. It also contains a short pentapeptide consisting of the residues 492–496 with the sequence HHHAH. This motif is also found in the Klebsiella aerogenes UreE protein, an accessory protein known to be essential for the insertion of Ni into the urease (UreE) of this bacterium. Mutants with a deleted ureE gene contain an inactive, Ni-free urease . The conserved His residues might therefore be more important for the postulated function of HypF, the incorporation of Ni into the hydrogenase active centre, than the zinc finger motifs.
Nostoc PCC 73102 HypC (89 amino acids) shows a PROSITE signature motif typical for the family of the small hydrophobic HupF/HypC proteins (75–110 residues). It contains all amino acid residues of the motif M-C-[LIV]-[GA]-[LIV]-P-x-[QKR]-[LIV]. This signature motif is not found in any other protein sequence in the databases and thus identifies HypC. In E. coli it could be demonstrated that HypC forms a complex with the hydrogenase 3 large subunit precursor pre-HycE. This complex formation is essential for the maturation of hydrogenase 3 .
hypD encodes a 392-amino acid protein, which carries the potential metal binding motif CPVC. No precise function of HypD has been identified yet.
The Nostoc PCC 73102 hypE encoded a 367-amino acid protein and has a conserved AIR synthetase (phosphoribosyl-aminoimidazole synthetase; EC 184.108.40.206) motif. No information about the function of HypE in the bacterial hydrogenase maturation process is available yet, although an interaction between HypF and HypE could be demonstrated in Helicobacter pylori.
The 113-amino acid gene product of hypA has an [FeS] cluster or zinc finger motif. The function of HypA is not known yet.
HypB has been proposed to be a major contributor to insertion of the Ni atom into the large hydrogenase subunit. As in the case of hypC and hypF, ΔhypB mutants produce Ni-free hydrogenase precursors [8,31]. Like all HypB proteins, the deduced Nostoc PCC 73102 275-amino acid protein has a conserved GTP binding motif at the C-terminus, which was shown to be essential for Ni insertion [6,32,33]. The N-terminal part of the protein is rich in His residues, which are probably involved in Ni storage and/or transport .
The cloned ORF upstream of hypF shows no homology to any known gene/protein in the databases.
3.2Expression analysis of the cloned genes
RT-PCR experiments using RNA isolated from Nostoc PCC 73102 grown under non-N2-fixing and N2-fixing conditions with hypF- and hupL-specific primers (Table 1) showed that hypF and hupL transcripts were not present in non-N2-fixing cultures, but can be detected under N2-fixing conditions (Fig. 1). Previously, hypB of Anabaena PCC 7120 was also shown to be expressed exclusively during N2 fixation . This may be considered unexpected, since this strain contains both a bidirectional and an uptake hydrogenase .
Figure 1. Expression of hypF in comparison to hupL in Nostoc PCC 73102. Total RNA isolated from Nostoc PCC 73102 cells was used in RT reactions using primers specific for hypF (F4, position see Fig. 2) and hupL. cDNA generated with F4 was used in a PCR with primer F5 for hypF (A), cDNA generated with L1 in a PCR with L2 for hupL (B). For sequences of the primers, see Table 1. Lane 1: 100 bp ladder; lane 2: RT-PCR with RNA from cells grown under non-N2-fixing conditions; lane 3: RT-PCR with RNA from cells grown under N2-fixing conditions; lane 4: control without reverse transcriptase in the control without reverse transcriptase in the RT-PCr with RNA from cells grown under N2-fixing conditions; lane 6: control with dH2O in the PCR; lane 7: positive controls with genomic DNA in the RT-PCR.
Download figure to PowerPoint
In 5′-RACE experiments, using primers of the 5′ region of both the hypF and hypC transcripts, no transcription start site could be identified upstream of either hypF or hypC. However, using a primer binding to the 5′ region of the ORF resulted in a 300-bp product (result not shown, but see Table 1 and Fig. 2). By cloning and sequencing the PCR product the transcriptional start site could be identified 17 bp upstream of the potential ORF (Fig. 2). Because of the short distance from the putative translation start of the ORF we analysed the identified ORF sequence in more detail. In addition to the ATG representing the start of the complete ORF, several additional start codons could be identified. Three putative start codons are indicated as a, b and c in Fig. 2. Being a unique ORF with no similarities to any known sequences in the databases, it is premature to exactly determine the size of this putative ORF.
Figure 2. Physical map of hypFCDEAB (grey shaded) in Nostoc PCCC 73102 including the upstream ORF and the downstream putative hlyA. Conserved regions in the deduced Hyp proteins are marked at the corresponding positions in the respective genes. The positions of the primers used in 5′-RACE experiments (grey box), RT reactions and PCR are shown (for primer sequences see Table 1). The primer O1 was used in the RT reaction for 5′-RACE experiments (grey box). The transcriptional start site (+1), as well as putative −10 and −35 sequence elements (italic bold) and putative NtcA binding sites (underlined) are indicated. Furthermore, three possible start codons (a, b, c) of the ORF are presented (see text). The primers B1 and C1 were used for cDNA generation in transcript analyses. Primer pairs used in the subsequent PCR, the resulting PCR products and the corresponding agarose gels are shown. M=100-bp ladder, (+)=PCR fragment, (−)=negative control without RT, H2O=negative control with water, DNA=positive control with DNA.
Download figure to PowerPoint
Using the primer B1, binding within hypB, in an RT reaction led to the synthesis of a cDNA, which generated PCR products with primer pairs covering hypFCD (F4/D1) and hypEAB (E1/B1) (Fig. 2). However, no PCR products with primer pairs covering the ORF and hypF were obtained. This is explainable by the limited capacity of the reverse transcriptase in the generation of long cDNAs. A primer pair covering the ORF and hypF (O2/F1) in a PCR following an RT reaction with a hypC-specific primer (C1) led to the amplification of a DNA fragment with the expected size (Fig. 2) and sequence. These results demonstrate that hypFCDEAB are transcribed, together with the ORF upstream from hypF, as a single mRNA (Fig. 2). Downstream of hypB an ORF in the opposite direction encoding a putative haemolysin-type calcium binding protein (hlyA) was identified. The transcription of hyp gene clusters as operons was previously reported for several bacteria such as B. japonicum or R. eutropha. However, in Synechococcus PCC 6301 a transcript consisting of hoxUYHWhypAB was recently demonstrated by RT-PCR, while hypF, which is located directly downstream of hypAB, seems to be part of another transcript .
Upstream of the transcriptional start site, putative −10 and −35 elements and putative NtcA binding sites , one of them overlapping the −35 element, can be identified (Fig. 2). NtcA is a global nitrogen regulator in cyanobacteria . In Anabaena PCC 7120 an NtcA binding site upstream of hetC, an ATP binding cassette transporter, has been identified, which appears to substitute for the −35 element . A putative NtcA binding site has also been identified upstream of the hupSL transcriptional start site , and might be responsible for the regulation of both ORFhypFCDEAB and hupSL.