Bacillus subtilis genes for the utilization of sulfur ... - Semantic Scholar
Microbiology (1998), 144, 2555-2561
Printed in Great Britain
Bacillus subtilis genes for the utilization of sulfur from aliphatic sulfonates Jan R. van der Ploeg,’ Nicola J. Cummings,’ Thomas Leisinger’ and Ian F. Connerton’ Author for correspondence : Jan R. van der Ploeg. Tel : e-mail: [email protected]
MikrobiologischesInstitut, Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Ziirich, Switzerland Institute of Food Research, Department of Food Macromolecular Science, Reading Laboratory, Earley Gate, Whiteknights Road, Reading RG6 6BZ, UK
+ 41 1 6323332. Fax : + 41 1 6321148.
A 5 kb region upstream of katA at 82” on the Bacillus subtilis chromosome contains five ORFs organized in an operon-like structure. Based on sequence similarity, three of the ORFs are likely to encode an ABC transport system (ssuBAC) and another to encode a monooxygenase (ssuD). The deduced amino acid sequence of the last ORF (yganr)shows no similarity to any known protein. B. subtilis can utilize a range of aliphatic sulfonates such as alkanesulfonates, taurine, isethionate and sulfoacetate as a source of sulfur, but not when SSUA and ssuC are disrupted by insertion of a neomycin-resistance gene. Utilizationof aliphatic sulfonates was not affected in a strain lacking 3’phosphoadenosine 5’-phosphosulfate (PAPS) sulfotransferase, indicatingthat sulfate is not an intermediate in the assimilation of sulfonate-sulfur.Sulfate or cysteine prevented expression of p-galactosidase from a transcriptional ssuD::/acZfusion. It is proposed that ssuBACD encode a system for ATPdependent transport of alkanesulfonatesand an oxygenase required for their desulfonation.
Keywords : sulfonate, sulfate starvation, Bacillus subtilis, cysteine biosynthesis, ABC transporter
INTRODUCTION There is little information available regarding the genes and enzymes involved in the assimilation of sulfate in Bacillus subtilis. Sulfur may be assimilated from either inorganic sulfur sources like sulfate, or organic sulfur sources such as sulfate esters or sulfonates. In Escherichia coli, reduction of sulfate to sulfide and utilization of the latter in the formation of cysteine has been well characterized (Kredich, 1996). Expression of the corresponding genes is completely repressed by cysteine and fully induced when E. coli is grown with a growthrate-limiting sulfur source. It is thought that assimilation of sulfate and its regulation in B. subtilis proceeds in a similar fashion to that in E. coli, since ATP-sulfurylase, adenosine 5’-phosphosulfate kinase, 3’-phosphoadenosine 5’-phosphosulfate (PAPS) sulfotransferase and sulfite reductase activities were found in extracts of cells grown with sulfate but not (or at reduced levels) in ...............................,.......,......................................................................................................... Abbreviations: CCCP, carbonyl cyanide m-chlorophenylhydrazone; FMNH,, reduced flavin mononucleotide; PAPS, 3’-phosphoadenosine 5’phosphosulfate The EMBL accession number for the sequence reported in this paper is 293 102.
cells grown with cysteine (Pasternak et al., 1965). So far only the B. subtilis cysH gene, encoding PAPS sulfotransferase, has been identified ;its expression was found to be repressed by cysteine and sulfide (Mansilla & de Mendoza, 1997). In soil, the habitat of B. subtilis, the amount of sulfate available to support growth is likely to be low. Most of the sulfur in soil is present in the form of sulfonates (Autry & Fitzgerald, 1990), which are generally stable compounds originating from a number of sources. Methanesulfonate is a product of photo-oxidation and chemical oxidation of dimethyl sulfide in the atmosphere (Kelly et al., 1993), whereas other sulfonates, including taurine (2-aminoethanesulfonate), are products of algal and animal metabolism (Huxtable, 1992). Many sulfonates, in particular aromatic sulfonates, arise from industrial activities. Several bacterial species are able to utilize aliphatic sulfonates as a sulfur source under aerobic conditions (Seitz et al., 1995 ; Uria-Nickelsen et al., 1993). In most of these species sulfate-sulfur was found to be preferred to sulfonate-sulfur, when both were present (Seitz et al., 1995; Uria-Nickelsen et a[., 1993 ; Kertesz, 1996). The preference for utilization of sulfate over organic sulfur sources appears not to be limited to aliphatic sulfonates but is also observed for
~~
0002-2433 0 1998 SGM
2555
J. R. V A N D E R P L O E G a n d O T H E R S
aromatic sulfonates and aromatic sulfate esters (Beil et al., 1995, 1996). It has been shown that in E. coli, Pseudomonas putida and Pseudomonas aeruginosa, the phenomenon is due to repression by sulfate of the enzymes involved in the liberation of sulfur from organosulfur compounds (Beil et al., 1995 ; Kertesz, 1996; van der Ploeg et al., 1996). Recently, we have identified the tauABCD gene cluster from E. coli, whose products are required for the utilization of taurine as a sulfur source and whose expression is repressed by sulfate (van der Ploeg et al., 1996). The tauABC genes most likely encode an ABC transport system for taurine, whereas the gene product of tauD is a 2-oxoglutarate-dependenttaurine dioxygenase (Eichhorn et al., 1997). The sequences of t a u A and tauB, encoding the taurine-binding protein and the membrane component of the putative ABC transport system, respectively, are significantly similar to B. subtilis OrfK and OrfL, two ORFs of unknown function (Quirk et al., 1994). These ORFs precede OrfM, in which insertion of Tn917 was found to result in resistance to the protonophore carbonyl cyanide mchlorophenylhydrazone (CCCP; Quirk et al., 1994). As a result of the B. subtilis genome sequencing project, we now recognize these sequences as part of what appears to be an operon, which we demonstrate to be required for the utilization of aliphatic sulfonates as a sulfur source.
Spizizen, 1961) but with equimolar amounts of magnesium chloride and ammonium chloride replacing magnesium sulfate and ammonium sulfate respectively. Sulfur sources were added to a final concentration of 250 pM. When necessary, amino acids were added at the following concentrations : L-histidine (100 pg ml-l), L-threonine (100 yg ml-l) and L-tryptophan (40 yg ml-l). Chloramphenicol (5 pg ml-l) and neomycin (10 pg ml-l) were added when required. E . coli strain DH5a was grown at 37 "C in LB medium. Antibiotics were added at the following concentrations : kanamycin (50 pg ml-l), ampicillin (100 pg ml-l) and chloramphenicol (35 pg ml-l). For plates, 1.5% (w/v) agar (Serva) was added to the medium. DNA manipulation. For plasmid isolation, restriction enzyme digestion and transformation of E. coli, standard procedures were used (Ausubel et al., 1987). B. subtilis was transformed as described by Cutting & Youngman (1994).
Construction of a deletion-insertion in ssuAC. Plasmid pECl1 (Quirk et al., 1994) was digested with BglIIIEcoRV and ligated to a BamHI-SmaI fragment from pBESTSO1 (Itaya et al., 1989), containing the neomycin resistance cassette, to give plasmid pME4154. B. subtilis strain BD99 was transformed with plasmid pME4154 ; one of the resulting neomycinresistant transformants that was sensitive to chloramphenicol was selected and designated SB10. The replacement of part of ssuAC with the neomycin-resistance cassette was confirmed by using PCR analysis. p-Galactosidase assay. p-Galactosidase activities were measured by using the Miller assay (Miller, 1992).
METHODS Chemicals. All chemicals used as sulfur sources were of the highest quality available and were obtained from Fluka, Acros, Aldrich or Sigma. Bacterial strains and growth conditions. The strains used in this study are listed in Table 1. B. subtilis strains were grown at 37 "C in Spizizen salts medium (Anagnostopoulos &
DNA sequencing and analysis. Chromosome sequencing was performed by a strategy which used a combination of both random DNase I fragments shotgun-cloned into the SmaI site of pUC8 and direct sequencing of a long-range PCR product (amplified from B. subtilis 168 genomic DNA) that contained the entire region within an 11.7 kb fragment. All sequences were determined from both strands using DyeDeoxy Terminator Cycle Sequencing kits with 373A and 373A STRETCH automated sequencers (Applied Biosystems). These reactions were primed with either long versions of the M13 universal primers or custom-synthesized oligonucleo-
Table 1. Strains and plasmids used in this study
Strain or plasmid Strains E. coli DHSa B. subtilis BD99 MSll SBlO BD2620 Plasmids pBESTSO1 pECl1 pME4154
2556
Relevant phenotype/genotype
Reference or source
supE44 A lacU169 (480 lacZAM15) hsdRl7 recAl endAl gyrA96 thi-1 relAl
Life Technologies
trpC2 thr-5 his BD99, SSUD::lacZ operon fusion, Cm' BD99, AssuAC : :kan trpC2 thr-5 cysH: :Tn917
Quirk et al. (1994) Quirk et 61. (1994) This study Mansilla & de Mendoza (1997)
kan cassette, Ap' ori (ColEl), Ap", contains proximal part of ssuACD flanking Tn917 insertion from MSll BamHI-SmaI fragment from pBEST501 cloned in BglII/SmaI-digested pECl1, Ap' Nm'
Itaya et al. (1989) Quirk et al. (1994) This study
...................
katA
-
B. subtilis sulfonate utilization genes
kan (SB10) Tn977: : lacZ (MS11) ....-._____.--_.--l.
r SUB
ssuA
r ssuC
ssuD
r
........
ygaN
u 1 kb
.................................................................................................................................................
Fig. 1. Organization of ssuBACD, ygaN and adjacent genes a t 82" on the B. subtilis chromosome. The solid line indicates the fragment that has been used in this study; the dotted lines indicate fragments whose sequence is documented elsewhere (Cummings & Connerton, 1997). The direction of transcription of the ORFs is depicted by arrows. The locations of insertion of Tn917 and the kan cassette are indicated.
tides. The DNA sequences were compiled using the fragment assembly programs in the Genetics Computer Group package (Devereux et al., 1984). Nucleotide and protein sequences were analysed with the Genetics Computer Group package (Devereux et al., 1984) version 8.1. Protein sequences were compared t o those in the SWISS-PROT database (release 34) with the program FASTA (Pearson & Lipman, 1988). Multiple sequence alignments were performed with CLUSTAL w (Thompson et al., 1994).
RESULTS AND DISCUSSION Sequence dete rminat ion and anaIysis
As part of the B. subtilis genome sequencing project, the 5 kb nucleotide sequence upstream of the k a t A gene was
determined (Fig. 1).Analysis of the sequence indicated the presence of five ORFs larger than 100 codons; the corresponding genes were named ygaL, ygbA, ygaM, ygcA and ygaN (Kunst et al., 1997). We now designate the first four genes SSUB,SSUA, ssuC and ssuD based on their proposed function in sulfonate-sulfur utilization ssuC and ssuD (Fig. 1).The proteins encoded by SSUA, had also previously been designated OrfK, OrfL and OrfM, respectively (Quirk et al., 1994). The nucleotide sequence upstream of ssuB contains a potential promoter region with an inverted repeat and poly(T) sequence lying between it and the initiation codon of ssuB. The gene product of ssuB (nt 409-1176) is 255 amino acids long and has a predicted molecular mass of 28.6 kDa. Although there is another possible translation start site at nt 352, the second initiation site is preceded by a Shine-Dalgarno sequence which is partly complementary to the 3' end of the 16s rRNA of B. subtilis (McLaughlin et al., 1981). The sequence of SsuB shows up to 40% identity to ATP-binding proteins from ABC transport systems and contains an ATP-binding site motif. The protein encoded by ssuA (nt 1194-2192) consists of 332 amino acids. It has a calculated molecular mass of 36.4 kDa and contains the lipid attachment motif from the prokaryotic membrane lipoproteins. The end of ssuA has an overlap of 4 bp with the translational start site of ssuC (nt 2189-3019), which encodes a protein of 276 amino acids with a calculated molecular mass of 30.2 kDa. The product of ssuD (nt 30364172) has 378 amino acids and a calculated molecular mass of 41.4 kDa. The last ORF, ygaN (nt 42764812), encodes
Table 2. Similarity of ORF products t o other proteins Gene product" SsuB (YgaL)
SsuA (YgbA, OrfK) SsuC (YgaM, Or&) Ssib (YgcA, OrfM) YgaN
Similarity
Accession n0.t
Identity
Putative taurine transport ATP-binding protein TauB, E. coli Putative aliphatic sulfonate ATP-binding protein SsuB, E. coli Histidine transport ATP-binding protein, E. coli Putative taurine-binding protein TauA, E. coli
447538
40.3
65.4
P38053
33.3
57.5
PO7109 447537
30.8 23.5
52.6 5 1.7
Putative aliphatic-sulfonate-binding protein SsuA, E. coli Putative taurine transport membrane component TauC, E. coli Putative aliphatic sulfonate transport membrane component SsuC, E. coli Putative aliphatic sulfonate monooxygenase Ssib (sulfate-starvation-induced protein Ssib), E. coli
P75853
30-5
53.1
447539
32-7
62.0
P75851
44.1
67.9
P80645
63.6
77.3
-
-
-
Similar protein(s)
-
(O/O
I*
(O/O
)*
"-Indicated in parentheses are alternative ORF assignments: the first as they appear on the B. subtilis genome web site (http://www.pasteur.fr/Bio/Su6tilist.html),the second (where present) as according to Quirk et al. (1994). t SWISS-PROT or Trembl database All alignments were performed over the full length sequence.
+
2557
J. R. V A N D E R P L O E G a n d O T H E R S
SsuLl-Bacsu Ssi6-Ecoli SoxA-Rhoso SndA-Strpr NtaA-Chche
1
1 1 1 1
...... ...... .MTQQR HTAPRR . MGANK
SsuD-Bacsu
SsuD-Bacsu Ssi6-Ecoli SoxA-Rhoso SnaA-Strpr
..
89
..
88 118 113
SW
..
SI
...............
SsuD-Bacsu
.................
SoxA-Rhoso SnaA-Strpr NtdA-Chehe
178 DEDALVLDKAAGVFADPAKVH 171 SDGR . . . . . . . . . . . . .P... 179 EDGAYLRDKLAGRYGLSEKIH
SsuD-Bacsu Ssi6-Ecoli SoxA-Rhos0 SnaA-Strpr NtaA-Chehe
186 185 237 213 238
Fig. 2. CLUSTAL w sequence alignment between B. subtilis SsuD (SsuD-Bacsu) and FMNH,-dependent monooxygenases. Residues identical to SsuD from B. subtilis are white on a black background and similar residues are in bold. Ssi6_Ecoli, sulfate starvation protein 6 from E. coli (SWISSPROT accession no. P80645); SoxARhoso, dibenzothiophene desulfurization enzyme A from Rhodococcus sp. strain IGTS8 (P54995); SnaAStrpr, pristinamycin Ha synthase subunit A from Streptomyces pristinaespiralis (P54991) ; NtaAChehe, nitri lotriacetate monooxygenase component A from Chelatobacter heintzii (P54989).
a protein of 178 amino acids with a molecular mass of 19.9 kDa. It seems likely that SsuBAC constitute an ABC transport system, since the sequences of SsuB and SsuC are similar to ATP-binding proteins and membrane components, respectively, of members of the ABC-transporter superfamily (Higgins, 1992). Furthermore, the sequences of the proteins encoded by ssuB, ssuA and ssuC were significantly similar to those of the proteins that constitute putative ABC transporters from E. coli involved in utilization of taurine (TauB, TauA and TauC, respectively; van der Ploeg et al., 1996) and aliphatic sulfonates (SsuB, SsuA and SsuC, respectively ; J. R. van der Ploeg & T. Leisinger, unpublished) (Table 2)The sequence of SsuD shows similarity to sequences of (FMNHJ-dependent monooxygenases with diverse substrate ranges (Fig. 2). The highest similarity was obtained with Ssi6 from E. coli (SWISS-PROT accession no. P80645), an enzyme involved in desulfonation of aliphatic sulfonates (Table 2; E. Eichhorn, J. R. van der Ploeg & T. Leisinger, unpublished). The predicted sequence of YgaN is unlike any other protein in the database.
Utilization of sulfur sources by wild-type and mutant strains
The high similarity of ssuBACD with genes encoding taurine and aliphatic sulfonate utilization from E. coli prompted us to investigate whether B. subtilis is able to grow with aliphatic sulfonates as sulfur sources. Aliphatic or aromatic sulfate esters were not utilized, but all primary aliphatic sulfonates tested supported growth of strain BD99. However, mutant SB10, in which part of ssuA and ssuC was replaced by a neomycin-resistance gene (Fig. l), was unable to utilize any of the alkanesulfonates tested as sulfur sources (Table 3), whereas growth with all other sulfur sources was like the wildtype strain. Strain MS11, which has an insertion of Tn917 in the fourth before last codon of ssuD (Quirk et al., 1994; Fig. l),grew like the wild-type strain. It seems unlikely that insertion of Tn917 has an effect on the activity of SsuD, but it may prevent or decrease expression of ygaN. It is not certain at this stage whether the ygaN gene is co-transcribed with ssuBACD, but we were unable to find a sequence resembling -35 and - 10 promoter regions recognized by RNA polymerase oA.It is possible that the product of ygaN is not required for sulfonate utilization, although it cannot be excluded
B. subtilis sulfonate utilization genes Table 3. Utilization of sulfur sources by B. subtilis wildtype (BD99) and mutant (SB10) strains
+ , Growth; -, no growth; ND, not done. Sulfur source
Strain
BD99 Sulfate Cy steine Cystine Methionine Glutathione Lanthionine Hexylsulfate p-Nitrocatechol sulfate Dodecyl sulfate Thioglycolate Djenkolate Methanesulfonate Ethanesulfonate Butanesulfonate Hexanesulfonate Taurine Sulfoacetate Sulfosuccinate Isethionate MOPS Methionine sulfoxide Methylsulfamate Dimethyl sulfone Cysteate Lipoic acid
SBlO
+ + + + + +
-
+ + + + + + + + + + + -
-
-
+ +
that there is an alternative protein which can suffice. Strain MS11, originally isolated by its resistance to CCCP, showed differences in the fatty acid composition of the cytoplasmic membrane as compared to the wildtype (Quirk et al., 1994), but it is not clear how the insertion of Tn927 in SSUDshould affect the membrane composition. B. subtilis strain BD2620, which contains an insertion of Tn917 in the cysH gene encoding PAPS transferase (Mansilla & de Mendoza, 1997), grew normally with all sulfonates tested, indicating that sulfate is not an intermediate in the assimilation pathway leading from aliphatic sulfonates to cysteine. Desulfonation of n-alkane sulfonates is catalysed by monooxygenases which yield aldehydes and sulfite as products (Eichhorn et al., 1997; Higgins et al., 1996; Thysse & Wanders, 1974). For the desulfonation of sulfosuccinate, which is not a sulfur source for B. subtilis, sulfite and oxaloacetate were the products (Quick et al., 1994). In E. coli, sulfonate-sulfur could be assimilated for growth by any mutant that was deficient in the reduction from sulfate to sulfite, but not by mutants deficient in sulfite reductase (Uria-Nickelsen et al., 1994).
Several bacterial species can use aliphatic sulfonates as a carbon or sulfur source, but until now there was no information available on the genes that are involved in these processes, except for the utilization of taurinesulfur by E. coli, encoded by tauABCD. This system is specific for taurine only, since mutants containing disrupted t a d , tauC or tauD genes could still utilize a wide variety of other aliphatic sulfonates (van der Ploeg et al., 1996). The situation is different in B. subtilis, where disruption of ssuA and ssuC does not allow for growth with any of the sulfonates, ranging from methanesulfonate to MOPS and including taurine. Unlike E. coli, B. subtilis apparently does not possess a separate system for taurine utilization. Based on the sequence similarity with Ssi6 from E. coli and other FMNH,-dependent monooxygenases (Fig. 2), and the phenotype of mutant SB10, it seems likely that B. subtilis SsuD is a monooxygenase which utilizes a broad range of aliphatic sulfonates. Since we have not tested whether disruption of ssuD results in the same phenotype as disruption of ssuA and ssuC, it cannot be excluded that SsuD is the only enzyme that is able to degrade sulfonates. The proteins which are related in sequence to SsuD, SnaA from Streptomyces pristinaespiralis, NtaA from Chelatobacter heintzii and Ssi6 from E. coli, require FMNH, for activity (Blanc et al., 1995; Knobel et al., 1996; E. Eichhorn, J. R. van der Ploeg & T. Leisinger, unpublished). The FMNH, is provided by NADH : :FMN oxidoreductases, for which the corresponding genes are located in the same gene cluster as the genes encoding the monooxygenase. Apparently, the gene encoding the corresponding NADH : :FMN oxidoreductase for SsuD is located elsewhere on the B. subtilis chromosome. Expression of ssuD::/acZis regulated by the sulfur source
To determine whether expression of the ssu genes was dependent on the sulfur source used for growth, Pgalactosidase activities were measured in strain MS11, which contains a transcriptional SSUD: :lac2 fusion (Fig. 1).N o P-galactosidase activity could be detected when sulfate, methionine, cysteine or cystine was present in the growth medium (Table 4). Djenkolate, thioglycolate and sulfonates except for taurine, resulted in intermediate P-galactosidase levels, whereas lanthionine, glutathione and taurine resulted in the highest levels of b-galactosidase activity. These results indicate that the synthesis of proteins encoded by ssuBACD is repressed by sulfate and cysteine. This is clearly different from the regulation of the enzymes involved in cysteine biosynthesis, which are repressed by cystine, but not by sulfate (Pasternak et al., 1965). The situation is very similar in E. coli, where the expression of tauABCD, but not of the cys genes, is repressed by sulfate (van der Ploeg et al., 1996). Full expression of the E. coli cys genes requires the CysB protein, limitation of reduced sulfur and 0- or N2559
J. R. V A N D E R P L O E G a n d OTHERS
REFERENCES
Tabre 4. Expression of P-galactosidase from a transcriptional SSUD::lacZ fusion upon growth with different sulfur sources ........ ...............,. . ....................................................................................................... ...
........
Anagnostopoulos, C. & Spizizen, 1. (1961). Requirements for
transformation in Bacillus subtilis. J Bacteriol 81, 741-746.
, , ,, ,
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. E., Seidman, J. G., Smith, J.A. & Struhl, K. (1987). Current Protocols in
The values shown here are the mean of at least two independent experiments. Sulfur source
Molecular Biology. New York : Wiley.
Autry, A. R. & Fitzgerald, 1. W. (1990). Sulfonate S : a major form
8-Galactosidase activity (Miller units)
of forest soil organic sulfur. Biol F e d Soils 10, 50-56.
Beil, S., Kehrli, H., James, P., Staudenmann, W., Cook, A. M., Leisinger, T. & Kertesz, M. A. (1995). Purification and charac-
~~
Sulfate Cystine Methionine Cy steine D jenkolate Lanthionine Glutathione (reduced) Thioglycolate Methanesulfonate Ethanesulfonate Propanesulfonate Butanesulfonate Sulfoacetate Cy steate MOPS Taurine Isethionate Glutathione (reduced) sulfate Taurine sulfate
+
+
terization of the arylsulfatase synthesized by Pseudomonas aeruginosa P A 0 during growth in sulfate-free medium and cloning of the arylsulfatase gene (atsA). Eur J Biochem 229, 385-394. Beil, S., Kertesz, M. A., Leisinger, T. & Cook, A. M. (1996). The assimilation of sulfur from multiple sources and its correlation with expression of the sulfate-starvation-induced stimulon in Pseudomonas putida S-313. Microbiology 142, 1989-1995.
Printed in Great Britain
Bacillus subtilis genes for the utilization of sulfur from aliphatic sulfonates Jan R. van der Ploeg,’ Nicola J. Cummings,’ Thomas Leisinger’ and Ian F. Connerton’ Author for correspondence : Jan R. van der Ploeg. Tel : e-mail: [email protected]
MikrobiologischesInstitut, Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Ziirich, Switzerland Institute of Food Research, Department of Food Macromolecular Science, Reading Laboratory, Earley Gate, Whiteknights Road, Reading RG6 6BZ, UK
+ 41 1 6323332. Fax : + 41 1 6321148.
A 5 kb region upstream of katA at 82” on the Bacillus subtilis chromosome contains five ORFs organized in an operon-like structure. Based on sequence similarity, three of the ORFs are likely to encode an ABC transport system (ssuBAC) and another to encode a monooxygenase (ssuD). The deduced amino acid sequence of the last ORF (yganr)shows no similarity to any known protein. B. subtilis can utilize a range of aliphatic sulfonates such as alkanesulfonates, taurine, isethionate and sulfoacetate as a source of sulfur, but not when SSUA and ssuC are disrupted by insertion of a neomycin-resistance gene. Utilizationof aliphatic sulfonates was not affected in a strain lacking 3’phosphoadenosine 5’-phosphosulfate (PAPS) sulfotransferase, indicatingthat sulfate is not an intermediate in the assimilation of sulfonate-sulfur.Sulfate or cysteine prevented expression of p-galactosidase from a transcriptional ssuD::/acZfusion. It is proposed that ssuBACD encode a system for ATPdependent transport of alkanesulfonatesand an oxygenase required for their desulfonation.
Keywords : sulfonate, sulfate starvation, Bacillus subtilis, cysteine biosynthesis, ABC transporter
INTRODUCTION There is little information available regarding the genes and enzymes involved in the assimilation of sulfate in Bacillus subtilis. Sulfur may be assimilated from either inorganic sulfur sources like sulfate, or organic sulfur sources such as sulfate esters or sulfonates. In Escherichia coli, reduction of sulfate to sulfide and utilization of the latter in the formation of cysteine has been well characterized (Kredich, 1996). Expression of the corresponding genes is completely repressed by cysteine and fully induced when E. coli is grown with a growthrate-limiting sulfur source. It is thought that assimilation of sulfate and its regulation in B. subtilis proceeds in a similar fashion to that in E. coli, since ATP-sulfurylase, adenosine 5’-phosphosulfate kinase, 3’-phosphoadenosine 5’-phosphosulfate (PAPS) sulfotransferase and sulfite reductase activities were found in extracts of cells grown with sulfate but not (or at reduced levels) in ...............................,.......,......................................................................................................... Abbreviations: CCCP, carbonyl cyanide m-chlorophenylhydrazone; FMNH,, reduced flavin mononucleotide; PAPS, 3’-phosphoadenosine 5’phosphosulfate The EMBL accession number for the sequence reported in this paper is 293 102.
cells grown with cysteine (Pasternak et al., 1965). So far only the B. subtilis cysH gene, encoding PAPS sulfotransferase, has been identified ;its expression was found to be repressed by cysteine and sulfide (Mansilla & de Mendoza, 1997). In soil, the habitat of B. subtilis, the amount of sulfate available to support growth is likely to be low. Most of the sulfur in soil is present in the form of sulfonates (Autry & Fitzgerald, 1990), which are generally stable compounds originating from a number of sources. Methanesulfonate is a product of photo-oxidation and chemical oxidation of dimethyl sulfide in the atmosphere (Kelly et al., 1993), whereas other sulfonates, including taurine (2-aminoethanesulfonate), are products of algal and animal metabolism (Huxtable, 1992). Many sulfonates, in particular aromatic sulfonates, arise from industrial activities. Several bacterial species are able to utilize aliphatic sulfonates as a sulfur source under aerobic conditions (Seitz et al., 1995 ; Uria-Nickelsen et al., 1993). In most of these species sulfate-sulfur was found to be preferred to sulfonate-sulfur, when both were present (Seitz et al., 1995; Uria-Nickelsen et a[., 1993 ; Kertesz, 1996). The preference for utilization of sulfate over organic sulfur sources appears not to be limited to aliphatic sulfonates but is also observed for
~~
0002-2433 0 1998 SGM
2555
J. R. V A N D E R P L O E G a n d O T H E R S
aromatic sulfonates and aromatic sulfate esters (Beil et al., 1995, 1996). It has been shown that in E. coli, Pseudomonas putida and Pseudomonas aeruginosa, the phenomenon is due to repression by sulfate of the enzymes involved in the liberation of sulfur from organosulfur compounds (Beil et al., 1995 ; Kertesz, 1996; van der Ploeg et al., 1996). Recently, we have identified the tauABCD gene cluster from E. coli, whose products are required for the utilization of taurine as a sulfur source and whose expression is repressed by sulfate (van der Ploeg et al., 1996). The tauABC genes most likely encode an ABC transport system for taurine, whereas the gene product of tauD is a 2-oxoglutarate-dependenttaurine dioxygenase (Eichhorn et al., 1997). The sequences of t a u A and tauB, encoding the taurine-binding protein and the membrane component of the putative ABC transport system, respectively, are significantly similar to B. subtilis OrfK and OrfL, two ORFs of unknown function (Quirk et al., 1994). These ORFs precede OrfM, in which insertion of Tn917 was found to result in resistance to the protonophore carbonyl cyanide mchlorophenylhydrazone (CCCP; Quirk et al., 1994). As a result of the B. subtilis genome sequencing project, we now recognize these sequences as part of what appears to be an operon, which we demonstrate to be required for the utilization of aliphatic sulfonates as a sulfur source.
Spizizen, 1961) but with equimolar amounts of magnesium chloride and ammonium chloride replacing magnesium sulfate and ammonium sulfate respectively. Sulfur sources were added to a final concentration of 250 pM. When necessary, amino acids were added at the following concentrations : L-histidine (100 pg ml-l), L-threonine (100 yg ml-l) and L-tryptophan (40 yg ml-l). Chloramphenicol (5 pg ml-l) and neomycin (10 pg ml-l) were added when required. E . coli strain DH5a was grown at 37 "C in LB medium. Antibiotics were added at the following concentrations : kanamycin (50 pg ml-l), ampicillin (100 pg ml-l) and chloramphenicol (35 pg ml-l). For plates, 1.5% (w/v) agar (Serva) was added to the medium. DNA manipulation. For plasmid isolation, restriction enzyme digestion and transformation of E. coli, standard procedures were used (Ausubel et al., 1987). B. subtilis was transformed as described by Cutting & Youngman (1994).
Construction of a deletion-insertion in ssuAC. Plasmid pECl1 (Quirk et al., 1994) was digested with BglIIIEcoRV and ligated to a BamHI-SmaI fragment from pBESTSO1 (Itaya et al., 1989), containing the neomycin resistance cassette, to give plasmid pME4154. B. subtilis strain BD99 was transformed with plasmid pME4154 ; one of the resulting neomycinresistant transformants that was sensitive to chloramphenicol was selected and designated SB10. The replacement of part of ssuAC with the neomycin-resistance cassette was confirmed by using PCR analysis. p-Galactosidase assay. p-Galactosidase activities were measured by using the Miller assay (Miller, 1992).
METHODS Chemicals. All chemicals used as sulfur sources were of the highest quality available and were obtained from Fluka, Acros, Aldrich or Sigma. Bacterial strains and growth conditions. The strains used in this study are listed in Table 1. B. subtilis strains were grown at 37 "C in Spizizen salts medium (Anagnostopoulos &
DNA sequencing and analysis. Chromosome sequencing was performed by a strategy which used a combination of both random DNase I fragments shotgun-cloned into the SmaI site of pUC8 and direct sequencing of a long-range PCR product (amplified from B. subtilis 168 genomic DNA) that contained the entire region within an 11.7 kb fragment. All sequences were determined from both strands using DyeDeoxy Terminator Cycle Sequencing kits with 373A and 373A STRETCH automated sequencers (Applied Biosystems). These reactions were primed with either long versions of the M13 universal primers or custom-synthesized oligonucleo-
Table 1. Strains and plasmids used in this study
Strain or plasmid Strains E. coli DHSa B. subtilis BD99 MSll SBlO BD2620 Plasmids pBESTSO1 pECl1 pME4154
2556
Relevant phenotype/genotype
Reference or source
supE44 A lacU169 (480 lacZAM15) hsdRl7 recAl endAl gyrA96 thi-1 relAl
Life Technologies
trpC2 thr-5 his BD99, SSUD::lacZ operon fusion, Cm' BD99, AssuAC : :kan trpC2 thr-5 cysH: :Tn917
Quirk et al. (1994) Quirk et 61. (1994) This study Mansilla & de Mendoza (1997)
kan cassette, Ap' ori (ColEl), Ap", contains proximal part of ssuACD flanking Tn917 insertion from MSll BamHI-SmaI fragment from pBEST501 cloned in BglII/SmaI-digested pECl1, Ap' Nm'
Itaya et al. (1989) Quirk et al. (1994) This study
...................
katA
-
B. subtilis sulfonate utilization genes
kan (SB10) Tn977: : lacZ (MS11) ....-._____.--_.--l.
r SUB
ssuA
r ssuC
ssuD
r
........
ygaN
u 1 kb
.................................................................................................................................................
Fig. 1. Organization of ssuBACD, ygaN and adjacent genes a t 82" on the B. subtilis chromosome. The solid line indicates the fragment that has been used in this study; the dotted lines indicate fragments whose sequence is documented elsewhere (Cummings & Connerton, 1997). The direction of transcription of the ORFs is depicted by arrows. The locations of insertion of Tn917 and the kan cassette are indicated.
tides. The DNA sequences were compiled using the fragment assembly programs in the Genetics Computer Group package (Devereux et al., 1984). Nucleotide and protein sequences were analysed with the Genetics Computer Group package (Devereux et al., 1984) version 8.1. Protein sequences were compared t o those in the SWISS-PROT database (release 34) with the program FASTA (Pearson & Lipman, 1988). Multiple sequence alignments were performed with CLUSTAL w (Thompson et al., 1994).
RESULTS AND DISCUSSION Sequence dete rminat ion and anaIysis
As part of the B. subtilis genome sequencing project, the 5 kb nucleotide sequence upstream of the k a t A gene was
determined (Fig. 1).Analysis of the sequence indicated the presence of five ORFs larger than 100 codons; the corresponding genes were named ygaL, ygbA, ygaM, ygcA and ygaN (Kunst et al., 1997). We now designate the first four genes SSUB,SSUA, ssuC and ssuD based on their proposed function in sulfonate-sulfur utilization ssuC and ssuD (Fig. 1).The proteins encoded by SSUA, had also previously been designated OrfK, OrfL and OrfM, respectively (Quirk et al., 1994). The nucleotide sequence upstream of ssuB contains a potential promoter region with an inverted repeat and poly(T) sequence lying between it and the initiation codon of ssuB. The gene product of ssuB (nt 409-1176) is 255 amino acids long and has a predicted molecular mass of 28.6 kDa. Although there is another possible translation start site at nt 352, the second initiation site is preceded by a Shine-Dalgarno sequence which is partly complementary to the 3' end of the 16s rRNA of B. subtilis (McLaughlin et al., 1981). The sequence of SsuB shows up to 40% identity to ATP-binding proteins from ABC transport systems and contains an ATP-binding site motif. The protein encoded by ssuA (nt 1194-2192) consists of 332 amino acids. It has a calculated molecular mass of 36.4 kDa and contains the lipid attachment motif from the prokaryotic membrane lipoproteins. The end of ssuA has an overlap of 4 bp with the translational start site of ssuC (nt 2189-3019), which encodes a protein of 276 amino acids with a calculated molecular mass of 30.2 kDa. The product of ssuD (nt 30364172) has 378 amino acids and a calculated molecular mass of 41.4 kDa. The last ORF, ygaN (nt 42764812), encodes
Table 2. Similarity of ORF products t o other proteins Gene product" SsuB (YgaL)
SsuA (YgbA, OrfK) SsuC (YgaM, Or&) Ssib (YgcA, OrfM) YgaN
Similarity
Accession n0.t
Identity
Putative taurine transport ATP-binding protein TauB, E. coli Putative aliphatic sulfonate ATP-binding protein SsuB, E. coli Histidine transport ATP-binding protein, E. coli Putative taurine-binding protein TauA, E. coli
447538
40.3
65.4
P38053
33.3
57.5
PO7109 447537
30.8 23.5
52.6 5 1.7
Putative aliphatic-sulfonate-binding protein SsuA, E. coli Putative taurine transport membrane component TauC, E. coli Putative aliphatic sulfonate transport membrane component SsuC, E. coli Putative aliphatic sulfonate monooxygenase Ssib (sulfate-starvation-induced protein Ssib), E. coli
P75853
30-5
53.1
447539
32-7
62.0
P75851
44.1
67.9
P80645
63.6
77.3
-
-
-
Similar protein(s)
-
(O/O
I*
(O/O
)*
"-Indicated in parentheses are alternative ORF assignments: the first as they appear on the B. subtilis genome web site (http://www.pasteur.fr/Bio/Su6tilist.html),the second (where present) as according to Quirk et al. (1994). t SWISS-PROT or Trembl database All alignments were performed over the full length sequence.
+
2557
J. R. V A N D E R P L O E G a n d O T H E R S
SsuLl-Bacsu Ssi6-Ecoli SoxA-Rhoso SndA-Strpr NtaA-Chche
1
1 1 1 1
...... ...... .MTQQR HTAPRR . MGANK
SsuD-Bacsu
SsuD-Bacsu Ssi6-Ecoli SoxA-Rhoso SnaA-Strpr
..
89
..
88 118 113
SW
..
SI
...............
SsuD-Bacsu
.................
SoxA-Rhoso SnaA-Strpr NtdA-Chehe
178 DEDALVLDKAAGVFADPAKVH 171 SDGR . . . . . . . . . . . . .P... 179 EDGAYLRDKLAGRYGLSEKIH
SsuD-Bacsu Ssi6-Ecoli SoxA-Rhos0 SnaA-Strpr NtaA-Chehe
186 185 237 213 238
Fig. 2. CLUSTAL w sequence alignment between B. subtilis SsuD (SsuD-Bacsu) and FMNH,-dependent monooxygenases. Residues identical to SsuD from B. subtilis are white on a black background and similar residues are in bold. Ssi6_Ecoli, sulfate starvation protein 6 from E. coli (SWISSPROT accession no. P80645); SoxARhoso, dibenzothiophene desulfurization enzyme A from Rhodococcus sp. strain IGTS8 (P54995); SnaAStrpr, pristinamycin Ha synthase subunit A from Streptomyces pristinaespiralis (P54991) ; NtaAChehe, nitri lotriacetate monooxygenase component A from Chelatobacter heintzii (P54989).
a protein of 178 amino acids with a molecular mass of 19.9 kDa. It seems likely that SsuBAC constitute an ABC transport system, since the sequences of SsuB and SsuC are similar to ATP-binding proteins and membrane components, respectively, of members of the ABC-transporter superfamily (Higgins, 1992). Furthermore, the sequences of the proteins encoded by ssuB, ssuA and ssuC were significantly similar to those of the proteins that constitute putative ABC transporters from E. coli involved in utilization of taurine (TauB, TauA and TauC, respectively; van der Ploeg et al., 1996) and aliphatic sulfonates (SsuB, SsuA and SsuC, respectively ; J. R. van der Ploeg & T. Leisinger, unpublished) (Table 2)The sequence of SsuD shows similarity to sequences of (FMNHJ-dependent monooxygenases with diverse substrate ranges (Fig. 2). The highest similarity was obtained with Ssi6 from E. coli (SWISS-PROT accession no. P80645), an enzyme involved in desulfonation of aliphatic sulfonates (Table 2; E. Eichhorn, J. R. van der Ploeg & T. Leisinger, unpublished). The predicted sequence of YgaN is unlike any other protein in the database.
Utilization of sulfur sources by wild-type and mutant strains
The high similarity of ssuBACD with genes encoding taurine and aliphatic sulfonate utilization from E. coli prompted us to investigate whether B. subtilis is able to grow with aliphatic sulfonates as sulfur sources. Aliphatic or aromatic sulfate esters were not utilized, but all primary aliphatic sulfonates tested supported growth of strain BD99. However, mutant SB10, in which part of ssuA and ssuC was replaced by a neomycin-resistance gene (Fig. l), was unable to utilize any of the alkanesulfonates tested as sulfur sources (Table 3), whereas growth with all other sulfur sources was like the wildtype strain. Strain MS11, which has an insertion of Tn917 in the fourth before last codon of ssuD (Quirk et al., 1994; Fig. l),grew like the wild-type strain. It seems unlikely that insertion of Tn917 has an effect on the activity of SsuD, but it may prevent or decrease expression of ygaN. It is not certain at this stage whether the ygaN gene is co-transcribed with ssuBACD, but we were unable to find a sequence resembling -35 and - 10 promoter regions recognized by RNA polymerase oA.It is possible that the product of ygaN is not required for sulfonate utilization, although it cannot be excluded
B. subtilis sulfonate utilization genes Table 3. Utilization of sulfur sources by B. subtilis wildtype (BD99) and mutant (SB10) strains
+ , Growth; -, no growth; ND, not done. Sulfur source
Strain
BD99 Sulfate Cy steine Cystine Methionine Glutathione Lanthionine Hexylsulfate p-Nitrocatechol sulfate Dodecyl sulfate Thioglycolate Djenkolate Methanesulfonate Ethanesulfonate Butanesulfonate Hexanesulfonate Taurine Sulfoacetate Sulfosuccinate Isethionate MOPS Methionine sulfoxide Methylsulfamate Dimethyl sulfone Cysteate Lipoic acid
SBlO
+ + + + + +
-
+ + + + + + + + + + + -
-
-
+ +
that there is an alternative protein which can suffice. Strain MS11, originally isolated by its resistance to CCCP, showed differences in the fatty acid composition of the cytoplasmic membrane as compared to the wildtype (Quirk et al., 1994), but it is not clear how the insertion of Tn927 in SSUDshould affect the membrane composition. B. subtilis strain BD2620, which contains an insertion of Tn917 in the cysH gene encoding PAPS transferase (Mansilla & de Mendoza, 1997), grew normally with all sulfonates tested, indicating that sulfate is not an intermediate in the assimilation pathway leading from aliphatic sulfonates to cysteine. Desulfonation of n-alkane sulfonates is catalysed by monooxygenases which yield aldehydes and sulfite as products (Eichhorn et al., 1997; Higgins et al., 1996; Thysse & Wanders, 1974). For the desulfonation of sulfosuccinate, which is not a sulfur source for B. subtilis, sulfite and oxaloacetate were the products (Quick et al., 1994). In E. coli, sulfonate-sulfur could be assimilated for growth by any mutant that was deficient in the reduction from sulfate to sulfite, but not by mutants deficient in sulfite reductase (Uria-Nickelsen et al., 1994).
Several bacterial species can use aliphatic sulfonates as a carbon or sulfur source, but until now there was no information available on the genes that are involved in these processes, except for the utilization of taurinesulfur by E. coli, encoded by tauABCD. This system is specific for taurine only, since mutants containing disrupted t a d , tauC or tauD genes could still utilize a wide variety of other aliphatic sulfonates (van der Ploeg et al., 1996). The situation is different in B. subtilis, where disruption of ssuA and ssuC does not allow for growth with any of the sulfonates, ranging from methanesulfonate to MOPS and including taurine. Unlike E. coli, B. subtilis apparently does not possess a separate system for taurine utilization. Based on the sequence similarity with Ssi6 from E. coli and other FMNH,-dependent monooxygenases (Fig. 2), and the phenotype of mutant SB10, it seems likely that B. subtilis SsuD is a monooxygenase which utilizes a broad range of aliphatic sulfonates. Since we have not tested whether disruption of ssuD results in the same phenotype as disruption of ssuA and ssuC, it cannot be excluded that SsuD is the only enzyme that is able to degrade sulfonates. The proteins which are related in sequence to SsuD, SnaA from Streptomyces pristinaespiralis, NtaA from Chelatobacter heintzii and Ssi6 from E. coli, require FMNH, for activity (Blanc et al., 1995; Knobel et al., 1996; E. Eichhorn, J. R. van der Ploeg & T. Leisinger, unpublished). The FMNH, is provided by NADH : :FMN oxidoreductases, for which the corresponding genes are located in the same gene cluster as the genes encoding the monooxygenase. Apparently, the gene encoding the corresponding NADH : :FMN oxidoreductase for SsuD is located elsewhere on the B. subtilis chromosome. Expression of ssuD::/acZis regulated by the sulfur source
To determine whether expression of the ssu genes was dependent on the sulfur source used for growth, Pgalactosidase activities were measured in strain MS11, which contains a transcriptional SSUD: :lac2 fusion (Fig. 1).N o P-galactosidase activity could be detected when sulfate, methionine, cysteine or cystine was present in the growth medium (Table 4). Djenkolate, thioglycolate and sulfonates except for taurine, resulted in intermediate P-galactosidase levels, whereas lanthionine, glutathione and taurine resulted in the highest levels of b-galactosidase activity. These results indicate that the synthesis of proteins encoded by ssuBACD is repressed by sulfate and cysteine. This is clearly different from the regulation of the enzymes involved in cysteine biosynthesis, which are repressed by cystine, but not by sulfate (Pasternak et al., 1965). The situation is very similar in E. coli, where the expression of tauABCD, but not of the cys genes, is repressed by sulfate (van der Ploeg et al., 1996). Full expression of the E. coli cys genes requires the CysB protein, limitation of reduced sulfur and 0- or N2559
J. R. V A N D E R P L O E G a n d OTHERS
REFERENCES
Tabre 4. Expression of P-galactosidase from a transcriptional SSUD::lacZ fusion upon growth with different sulfur sources ........ ...............,. . ....................................................................................................... ...
........
Anagnostopoulos, C. & Spizizen, 1. (1961). Requirements for
transformation in Bacillus subtilis. J Bacteriol 81, 741-746.
, , ,, ,
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. E., Seidman, J. G., Smith, J.A. & Struhl, K. (1987). Current Protocols in
The values shown here are the mean of at least two independent experiments. Sulfur source
Molecular Biology. New York : Wiley.
Autry, A. R. & Fitzgerald, 1. W. (1990). Sulfonate S : a major form
8-Galactosidase activity (Miller units)
of forest soil organic sulfur. Biol F e d Soils 10, 50-56.
Beil, S., Kehrli, H., James, P., Staudenmann, W., Cook, A. M., Leisinger, T. & Kertesz, M. A. (1995). Purification and charac-
~~
Sulfate Cystine Methionine Cy steine D jenkolate Lanthionine Glutathione (reduced) Thioglycolate Methanesulfonate Ethanesulfonate Propanesulfonate Butanesulfonate Sulfoacetate Cy steate MOPS Taurine Isethionate Glutathione (reduced) sulfate Taurine sulfate
+
+
terization of the arylsulfatase synthesized by Pseudomonas aeruginosa P A 0 during growth in sulfate-free medium and cloning of the arylsulfatase gene (atsA). Eur J Biochem 229, 385-394. Beil, S., Kertesz, M. A., Leisinger, T. & Cook, A. M. (1996). The assimilation of sulfur from multiple sources and its correlation with expression of the sulfate-starvation-induced stimulon in Pseudomonas putida S-313. Microbiology 142, 1989-1995.
