SEARCH

SEARCH BY CITATION

Keywords:

  • Purple sulfur bacterium;
  • Hydrogenase;
  • Sulfite reductase;
  • Heterodisulfide reductase;
  • Heterologous hybridization;
  • Sulfur oxidation;
  • Allochromatium vinosum;
  • Thiocapsa roseopersicina

Abstract

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgments
  7. References

The dsr genes and the hydSL operon are present as separate entities in phototrophic sulfur oxidizers of the genera Allochromatium, Marichromatium, Thiocapsa and Thiocystis and are organized similarly as in Allochromatium vinosum and Thiocapsa roseopersicina, respectively. The dsrA gene, encoding the α subunit of ‘reverse’ siroheme sulfite reductase, is also present in two species of green sulfur bacteria pointing to an important and universal role of this enzyme and probably other proteins encoded in the dsr locus in the oxidation of stored sulfur by phototrophic bacteria. The hupSL genes are uniformly present in the members of the Chromatiaceae family tested. The two genes between hydS and hydL encode a membrane-bound b-type cytochrome and a soluble iron-sulfur protein, respectively, resembling subunits of heterodisulfide reductase from methanogenic archaea. These genes are similar but not identical to dsrM and dsrK, indicating that the derived proteins have distinct functions, the former in hydrogen metabolism and the latter in oxidative sulfur metabolism.


1Introduction

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgments
  7. References

Purple and green sulfur bacteria are able to utilize reduced sulfur compounds, such as sulfide, as photosynthetic electron donors. Usually, ‘elemental’ sulfur is formed as an intermediate en route to the end product sulfate. Recently, it has been shown that siroheme sulfite reductase and other proteins encoded on the dsr gene locus are essential for the further oxidation of ‘elemental’ sulfur in Allochromatium vinosum (formerly Chromatium vinosum[1]) [2]. However, the general relevance of sulfite reductase for oxidative sulfur metabolism in phototrophic sulfur bacteria remained dubious because Ach. vinosum was the only representative of this physiological group in which the enzyme had been found [3,4]. Hydrogen serves as an alternative photosynthetic electron donor in most phototrophic sulfur bacteria and several green and purple photosynthetic bacteria including Ach. vinosum and Thiocapsa roseopersicina, which have been reported to contain [NiFe]hydrogenase(s) [5]. Genes for a stable (hydSL) and an unstable (hupSL) [NiFe]hydrogenase have been sequenced from the latter [6,7]. The hydS and hydL genes are separated by two open reading frames (isp1 and isp2). The molecular biology of hydrogenases has been thoroughly studied in purple non-sulfur phototrophic bacteria, and this rather unusual gene arrangement was not found. The intriguing question remains if the gene arrangement is common among Chromatiaceae.

The isp1- and isp2-encoded proteins from Tca. roseopersicina show significant similarity to two of the proteins (DsrM and DsrK) encoded in the dsr gene cluster of Ach. vinosum. Both the DsrM and Isp1 proteins contain five putative membrane-spanning helices and are homologous to the heme b-containing subunit of heterodisulfide reductase from Methanosarcina barkeri, the γ subunits of nitrate reductases, and predicted b-type cytochromes from the sulfate reducers Desulfovibrio vulgaris and Archaeoglobus fulgidus. DsrK and Isp2 both resemble the catalytic iron-sulfur cluster-carrying subunits of heterodisulfide reductases of methanogenic archaea, the hmc6 gene product from D. vulgaris and several hypothetical proteins from A. fulgidus[2,7]. The occurrence of homologous genes in the dsr locus of Ach. vinosum on the one hand, and in the hyd operon of Tca. roseopersicina on the other hand, suggests a possible link between hydrogen and sulfur metabolism in phototrophic sulfur bacteria. We examined this potential link with comparative Southern hybridization analyses of the dsr and hyd loci in various purple and green sulfur bacteria.

2Materials and methods

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgments
  7. References

2.1Bacterial strains and growth conditions

Ach. vinosum strains DSMZ 180T and DSMZ 185 were grown as described elsewhere [8]. Thiocystis violacea DSMZ 208, Allochromatium minutissimum (formerly Chromatium minutissimum[1]) DSMZ 1376T, Marichromatium gracile (formerly Chromatium gracile[1]) DSMZ 203T, Marichromatium purpuratum (formerly Chromatium purpuratum[1]) DSMZ 1591T, Tca. roseopersicina strains BBS and 6311, Chlorobium limicola DSMZ 257 and Chlorobium vibrioforme DSMZ 263 were grown as described in [9]. Cells were harvested by centrifugation and stored as a cell paste at −20°C.

2.2Genetic methods

DNA was isolated either by sarcosyl lysis after Bazaral and Helinski [10] followed by phenol/chloroform extraction and dialysis against water or using Nucleobond AXG 100 cartridges with buffer Set III (Macherey-Nagel, Düren, Germany). 5 μg DNA was used for each digest. Southern hybridizations were performed overnight at 60°C following the protocol given in [11]. In some cases, membranes were used twice and probes were stripped off before the second hybridization by washing with distilled water at 68°C for 5 min, followed by six incubations in 0.2 M NaOH, 1% SDS. The latter solution was boiling when applied and washing was continued at 68°C for 10 min. Finally the membranes were equilibrated in 2×SSC.

2.3Primers and PCR for probe generation

The following primers were used for probe generation: dsrA probe: dsoA1 (5′-GAATTCCACACCGTCCG-3′) and dsoA3 (5′-AGCGGGTGATGACGTT-3′); dsrE-C probe: 370F (5′-CAGGACGATCGTCACAT-3′) and 3r (5′-TCGGTGGAGCTTGATGGA-3′); dsrM: 6f (5′-GGCTCGCAAGATCATTCA-3′) and 7r (5′-CCACCAGGAGCAGATGCA-3′); dsrK: 10f (5′-GGTGCCTGCACCGACAAAT-3′) and 9r (5′-TGACCTTGAGCTTGATCG-3′). The dsr gene probes were amplified via PCR with chromosomal Ach. vinosum DNA as the template. The hupSL structural genes were amplified from Tca. roseopersicina genomic DNA using the PCR primers OHUP1 (5′-ATGCCAACCACCGAAACCTATTAC-3′) and OHUP2 (5′-TCAGCGGACTTGATGCGAGTG-3′). These primers cover the region from the start codon of hupS to the stop codon of hupL. The PCR fragment was isolated from agarose gel, 3′ end polished and 5′ end phosphorylated before cloning into the HincII site of pBluescript vector. The insert was isolated from the vector for DIG labeling. The hydS probe was amplified from plasmid pTSUP2 using primers TRSA/N and ACX15/R [7]. The isp1 probe was generated by PCR amplification from plasmid pTSH20/1 (783-bp BglII/XhoI fragment covering isp1 and the first 222 bp of isp2 in the XhoI/BamHI sites of pBluescript) using M13 forward and reverse universal primers. The isp2 probe was amplified from plasmid pTSH14/3 (1038-bp XhoI/EcoRI fragment covering most of isp2 in pBluescript SK+) using M13 forward and reverse universal primers. The hydL probe was generated by PCR from plasmid pACRG4 using primers ACX12/N and ACX13/R [7]. PCR with Taq DNA polymerase was done essentially as described in [12]. 49°C was chosen as the annealing temperature for amplification of the isp1 and isp2 probes, annealing for production of the hydS and hydL probes was done at 60°C. Extension at 72°C was for 1.5 min. Annealing temperature and extension time were 50°C and 1 min for the dsrM, dsrK, and dsrA probes, and 48°C and 2 min for the dsrE-C probe. Probes were labelled with digoxigenin via PCR with final concentrations of 20 μM DIG-dUTP and 180 μM dTTP in the reaction mixture. For digoxigenin labelling of the probes, 10% of the dTTP in the PCR reactions was replaced by DIG-dUTP.

3Results and discussion

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgments
  7. References

Southern hybridizations were performed to reveal the occurrence and arrangement of the dsr and hyd genes in six species of purple sulfur bacteria and two species of green sulfur bacteria. The linkage of the dsr and hyd genes was deduced not only from the fact that hybridizing fragments of similar sizes were detected in independent hybridizations but also by rehybridization of membranes with various probes. Thereby, it was ensured that the very same fragments were detected by distinct probes. The organization of the dsr and hyd gene clusters in Ach. vinosum and Tca. roseopersicina, respectively, and the gene probes employed are given in Figs. 1B and 2B.

image

Figure 1. A: Southern hybridization of EcoRI-digested DNA from Tcs. violacea with four different gene probes derived from the dsr gene locus of Ach. vinosum (indicated in B). B: Restriction map of the Tcs. violacea dsr region, derived from the results presented in A as well as from hybridizations of BamHI-restricted and EcoRI/BamHI double-digested DNA (not shown). The arrows indicate the positions of hybridizing bands.

Download figure to PowerPoint

image

Figure 2. A: Southern hybridization of EcoRI-digested DNA from Tcs. violacea with four different gene probes derived from the hyd gene locus of Tca. roseopersicina (indicated in B). B: Restriction map of the Tcs. violacea hyd region, derived from the results presented in A as well as from hybridizations of BamHI-restricted and EcoRI/BamHI double-digested DNA (not shown). The arrangement of the genes was furthermore confirmed by Southern hybridization analysis of BglII- and BglII/EcoRI-digested DNA (not shown). The arrows indicate the positions of hybridizing bands.

Download figure to PowerPoint

3.1dsr gene hybridizations

All purple and green sulfur bacteria tested contain DNA reacting with gene probes comprising dsrE-C, dsrM or dsrK. For Tcs. violacea (Fig. 1) and Ach. minutissimum (data not shown), the results indicate that the dsr genes have a very similar, if not the same, arrangement as in Ach. vinosum[2]. In Tca. roseopersicina BBS and 6311 strains and in Mch. gracile the dsr genes are located close together on the same restriction fragments, but the fragment size of double digestions prevented a more precise determination of the organization of the genes. For Mch. purpuratum, the presence of all tested dsr genes was detected, although the extremely low DNA yields did not allow enough hybridizations to map the gene arrangement. The genes dsrE-C, dsrK and dsrM could not be unambiguously recognized in the two Chlorobium species tested. This may indicate either the absence of these genes in Chlorobiaceae or the negative result is due to the large phylogenetic distance between purple and green sulfur bacteria [13], leading to decreased homology between the corresponding genes which prevented positive hybridization at the applied stringency.

3.2hyd and hup gene hybridizations

The structural genes encoding a stable (hydSL) and an unstable (hupSL) hydrogenase from Tca. roseopersicina have been analyzed earlier [6,7]. Unlike most known hydrogenase sequences, hydS and hydL are separated by two ORFs, isp1 and isp2. An intriguing question is whether or not this gene organization is common among Chromatiaceae and if other Chromatiaceae species contain more than one hydrogenase. If they do, the phenomenon may have a physiological importance. Molecular biology of hydrogenase(s) has not been investigated in Ach. vinosum or in other members of the family, except for Tca. roseopersicina. Southern hybridizations were therefore carried out using Tca. roseopersicina hupSL, hydSL and isp12 probes. Genomic DNA from the following organisms was probed: Ach. vinosum, Ach. minutissimum, Mch. purpuratum, Mch. gracile, Tcs. violacea.

The results clearly indicate that all phototrophic sulfur bacteria tested contain homologous sequences to both Tca. roseopersicina hydrogenase structural genes. For Tcs. violacea (Fig. 2) and Ach. vinosum (data not shown), the hybridization results showed the same hydSisp1isp2hydL arrangement as in Tca. roseopersicina BBS [7]. In Ach. minutissimum, the hyd genes are located close together, i.e., they reside on the same restriction fragment. Further characterization was not possible because of the arrangement of the restriction sites on the fragment. Similarly, the hyd and isp genes were present in Mch. purpuratum and Mch. gracile, but the data did not allow unambiguous mapping.

In all tested purple sulfur bacteria, the probes for the related dsrM and isp1, as well as those for the likewise related genes dsrK and isp2, bound to distinct DNA fragments, indicating that each organism contained one copy of each gene. The results also strongly suggest that the dsrM- and dsrK-derived proteins play a specific role in sulfur metabolism, while Isp1 and Isp2 possibly function in hydrogen metabolism, albeit with a pronounced homology at the amino acid level to DsrM and DsrK, respectively.

3.3Sequence comparisons

The latter finding was confirmed by phylogenetic analysis of DsrM/Isp1 and DsrK/Isp2 proteins (Fig. 3). For orthologous proteins [18] that retained their ancestral physiological role, a polypeptide-based tree would have been expected to yield, at least roughly, the same ordering of taxa as a 16S rRNA-based tree. However, this is not true for the proteins compared here. In the case of the DsrM/Isp1-related proteins, nitrate reductase subunits form one major group, and Tca. roseopersicina Isp1, D. vulgaris Hmc5 and a protein encoded between the hydS and hydL genes of Aquifex aeolicus constitute a second major group. A third group consists of Ach. vinosum DsrM and three, relatively closely related hypothetical proteins from the sulfate reducing archaeon Archaeoglobus fulgidus. The most conspicuous difference between the DsrM and Isp1 related proteins is the spacing between the putative heme b-binding histidine residues in helix B, which is 9 amino acid long in the DsrM group and 12 amino acids long in the Isp1 related polypeptides.

image

Figure 3. Bootstrap parsimony analyses of amino acid sequence alignments of DsrM and ISP1 (A) and DsrK and Isp2 (B) with related characterized and hypothetical proteins. In the analyses only those proteins were included that produced significant alignments with DsrK/Isp2 or DsrM/Isp1 over the entire lengths of the polypeptides in PSI-BLAST [14] searches. Amino acid sequence alignments were produced with the program ClustalW [15] using the PAM 001 Mutation Data Matrix [16], gaps in the data sets were removed and the unrooted trees shown were constructed with the program PROTPARS [17] and compared with trees obtained by the distance matrix method (programs PROTDIST and FITCH [17]). Protein parsimony and distance matrix bootstrap analyses were based on 1000 and 100 resamplings, respectively. Bootstrap values are reported at the forks in the order parsimony/distance matrix. Asterisks indicate that the fork in question was not recovered in the majority of bootstrap replicates. GenBank accession numbers are given below the sequence names. Af, Archaeoglobus fulgidus; Aq, Aquifex aeolicus; Av, Allochromatium vinosum; Bs, Bacillus subtilis; Dv, Desulfovibrio vulgaris; Ec, Escherichia coli; Mj, Methanococcus jannaschii; Msb, Methanosarcina barkeri; Mth, Methanobacterium thermoautotrophicum; Myt, Mycobacterium tuberculosis; Pd, Paracoccus denitrificans; Psa, Pseudomonas aeruginosa; Tc, Thiocapsa roseopersicina; Tt, Thermus thermophilus.

Download figure to PowerPoint

In the case of the DsrK related sequences, the alignment of the 22 related sequences was straightforward because all the [4Fe-4S] cluster binding cysteines as well as the cysteines in the C-terminal part of heterodisulfide reductases [CCG(G/A)GGGV] were perfectly conserved. It is to be noted that in Methanococcus jannaschii and Methanobacterium thermoautotrophicum the sequence corresponding to dsrK is split into two genes (hdrC and hdrB) [19]. For the alignment, the respective amino acid sequences were put together as if they formed a single polypeptide. The phylogenetic tree based on the DsrK/Isp2 sequences showed that the related sequences also fell into three major groups: first, the hdrD-encoded subunit of heterodisulfide reductase from Methanosarcina barkeri[20] and related putative proteins from other methanogens, A. fulgidus and Bacillus subtilis. Second is the group that comprises HdrCB from M. thermoautotrophicum and closely related proteins from other species. Ach. vinosum DsrK and Tca. roseopersicina Isp2 are both found in the third major group. In spite of the very close phylogenetic relationship of the two purple sulfur bacteria [1] the closest relative of the Allochromatium protein is a hypothetical polypeptide from A. fulgidus while Thiocapsa Isp2 groups with a polypeptide encoded in the hmc operon of D. vulgaris and a protein encoded between the hydS and hydL genes of A. aeolicus. The close similarity between the A. fulgidus and the Ach. vinosum proteins on the one hand, and the close similarity between the Tca. roseopersicina and the A. aeolicus proteins on the other, can be explained when it is assumed that the proteins compared here evolved under dissimilar functional constraints and possibly developed into independent lineages prior to the divergence of the organisms in which they are found. In order to appreciate this apparent relatedness, one has to consider the huge phylogenetic distance between Chromatium/Archaeoglobus and Thiocapsa/Aquifex. Although the proteins still show remarkable resemblance in their amino acid sequences, they are most likely optimized for and perform dissimilar metabolic functions.

Interestingly, the proteins most closely related to the Dsr proteins from Ach. vinosum are found in the sulfate reducer A. fulgidus. Possibly, these proteins are essential for the function of sulfite reductase and/or electron flow to or from sulfite reductase, a protein that is important for sulfur oxidation in Ach. vinosum and – acting in the reverse direction – for sulfide production from sulfite in A. fulgidus.

3.4Conclusions

The hupSL genes are uniformly present in the members of the Chromatiaceae family tested. The hydSL genes show strong homology in all Chromatiaceae tested. The organization of the hup and hyd operons is similar in the six species investigated.

The dsr operon is present in all species investigated and is organized as in Ach. vinosum. The dsrA gene, encoding the α subunit of a reverse siroheme-sulfite reductase, is also present in two species of green sulfur bacteria. Therefore we propose that this enzyme, and probably other proteins encoded in the dsr locus, play an important and universal role in the oxidation of stored sulfur by phototrophic bacteria.

The dsr genes are localized in a locus different from the one containing the hyd genes. The two open reading frames between hydS and hydL are similar, but not identical to the dsrM and dsrK genes. The putative physiological significance of the homologous sequences needs further study.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgments
  7. References

The excellent technical assistance of Dorit Glass is gratefully acknowledged. The work in C.D.'s lab was supported by the Deutsche Forschungsgemeinschaft. K.L.K. and coworkers acknowledge financial support from UNDP-HUN/95/002-0102/-0103, OMFB-00017/99, TEMPUS S_JEP-12011-97, and COST Action 818.

References

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgments
  7. References
  • [1]
    Imhoff, J.F., Süling, J., Petri, R. (1998) Phylogenetic relationships and taxonomic reclassification of Chromatium species and related purple sulfur bacteria. Int. J. Syst. Bacteriol. 48, 11291143.
  • [2]
    Pott, A.S., Dahl, C. (1998) Sirohaem-sulfite reductase and other proteins encoded in the dsr locus of Chromatium vinosum are involved in the oxidation of intracellular sulfur. Microbiology 144, 18811894.
  • [3]
    Schedel, M., Vanselow, M., Trüper, H.G. (1979) Siroheme sulfite reductase from Chromatium vinosum. Purification and investigation of some of its molecular and catalytic properties. Arch. Microbiol. 121, 2936.
  • [4]
    Brune, D.C. (1995) Sulfur compounds as photosynthetic electron donors. In: Anoxygenic Photosynthetic Bacteria (Blankenship, R.E., Madigan, M.T. and Bauer, C.E., Eds.), pp. 847–870. Kluwer Academic, Dordrecht.
  • [5]
    Vignais, P., Toussaint, B. and Colbeau, A. (1995) Regulation of hydrogenase gene expression. In: Anoxygenic Photosynthetic Bacteria (Blankenship, R.E., Madigan, M.T. and Bauer, C.E., Eds.), pp. 1175–1190. Kluwer Academic, Dordrecht.
  • [6]
    Colbeau, A., Kovacs, K.L., Chabert, J., Vignais, P.M. (1994) Cloning and sequences of the structural (hupSLC) and accessory (hupDHL) genes for hydrogenase biosynthesis in Thiocapsa roseopersicina. Gene 140, 2531.
  • [7]
    Rakhely, G., Colbeau, A., Garin, J., Vignais, P.M., Kovacs, K.L. (1998) Unusual organization of the genes coding for HydSL, the stable [NiFe] hydrogenase in the photosynthetic bacterium Thiocapsa roseopersicina BBS. J. Bacteriol. 180, 14601465.
  • [8]
    Dahl, C., Trüper, H.G. (1994) Enzymes of dissimilatory sulfide oxidation in phototrophic bacteria. Methods Enzymol. 243, 400421.
  • [9]
    Pfennig, N. and Trüper, H.G. (1992) The family Chromatiaceae. In: The Prokaryotes. A Handbook on the Biology of Bacteria: Ecophysiology, Isolation, Identification, Applications (Balows, A., Trüper, H.G., Dworkin, M., Harder, W. and Schleifer, K.-H., Eds.), pp. 3200–3221. Springer-Verlag, New York.
  • [10]
    Bazaral, M., Helinski, D.R. (1968) Circular DNA forms of colicinogenic factors E1, E2 and E3 from Escherichia coli. J. Mol. Biol. 36, 185194.
  • [11]
    Dahl, C., Speich, N., Trüper, H.G. (1994) Enzymology and molecular biology of sulfate reduction in the extremely thermophilic archaeon Archaeoglobus fulgidus. Methods Enzymol. 243, 331349.
  • [12]
    Dahl, C. (1996) Insertional gene inactivation in a phototrophic sulphur bacterium: APS-reductase-deficient mutants of Chromatium vinosum. Microbiology 142, 33633372.
  • [13]
    Olsen, J., Woese, C.R., Overbeek, R. (1994) The winds of (evolutionary) change: breathing new life into microbiology. J. Bacteriol. 176, 16.
  • [14]
    Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 33893402.
  • [15]
    Higgins, D.G., Sharp, P.M. (1989) Clustal: a package for performing multiple sequence alignment on a microcomputer. Gene 73, 237244.
  • [16]
    Dayhoff, M.O. (1978) Atlas of Protein Sequence and Structure. National Biomedical Research Foundation, Washington, DC.
  • [17]
    Felsenstein, J. (1993) PHYLIP (Phylogeny Inference Package) version 3.5c. ftp.bio.indiana.edu/molbio/evolve.
  • [18]
    Gogarten, J.P. (1994) Which is the most conserved group of proteins? Homology-orthology, paralogy, xenology, and the fusion of independent lineages. J. Mol. Evol. 39, 541543.
  • [19]
    Hedderich, R., Koch, J., Linder, D., Thauer, R.K. (1994) The heterodisulfide reductase from Methanobacterium thermoautotrophicum contains sequence motifs characteristic of pyridine-nucleotide-dependent thioredoxin reductases. Eur. J. Biochem. 225, 253261.
  • [20]
    Künkel, A., Vaupel, M., Heim, S., Thauer, R.K., Hedderich, R. (1997) Heterodisulfide reductase from methanol-grown cells of Methanosarcina barkeri is not a flavoenzyme. Eur. J. Biochem. 244, 226234.