Quantitative utilization of L-cysteate (2-amino-3-sulphopropionate) as the sole source of carbon and energy for growth of the aerobic, marine bacterium Silicibacter pomeroyi DSS-3T was observed. The sulphonate moiety was recovered in the medium largely as sulphite, and the appropriate amount of the ammonium ion was also observed. Genes [suyAB (3-sulpholactate sulpho-lyase)] encoding the known desulphonation reaction in cysteate degradation were absent from the genome, but a homologue of a putative sulphate exporter gene (suyZ) was found, and its neighbour, annotated as a D-cysteine desulphhydrase, was postulated to encode pyridoxal 5′-phosphate-coupled L-cysteate sulpho-lyase (CuyA), a novel enzyme. Inducible CuyA was detected in cysteate-grown cells. The enzyme released equimolar pyruvate, sulphite and the ammonium ion from L-cysteate and was purified to homogeneity by anion-exchange, hydrophobic-interaction and gel-filtration chromatography. The N-terminal amino acid sequence of this 39-kDa subunit confirmed the identification of the cuyA gene. The native enzyme was soluble and homomultimeric. The Km-value for L-cysteate was high (11.7 mM) and the enzyme also catalysed the D-cysteine desulphhydrase reaction. The gene cuyZ, encoding the putative sulphite exporter, was co-transcribed with cuyA. Sulphite was exported despite the presence of a ferricyanide-coupled sulphite dehydrogenase. CuyA was found in many bacteria that utilize cysteate.
- cysteate dissimilation
- pyridoxal 5′-phosphate
- sequence comparisons
- sulphite exporter
L-Cysteate (2-amino-3-sulphopropionate) has been recognized as a natural product since 1946, when it was discovered as a product of the weathering of wool . In the meantime, the compound has been found in Staphylococcus aureus and Bacillus subtilis [2,3] and shown to be an educt for the synthesis of the bacterial sulpholipid capnine , of algal sulpholipids  and possibly of another bacterial sulpholipid ; extracellularly, the compound is described as a component of spiders' webs . This widespread occurrence of cysteate is mirrored by its biodiversity in microbial nutrition. The compound can serve as a source of sulphur for yeasts and bacteria; oxygenolytic desulphonation occurs in Escherichia coli [8,9]. Cysteate can serve as a sole source of nitrogen for bacterial growth (K. Denger and A. M. Cook, unpublished work); all the tested isolates degraded the compound apparently to biomass, CO2, ammonium and sulphate by the same pathway used by the organism for dissimilation of cysteate. Many aerobic and anaerobic bacteria have been found to dissimilate cysteate  (see below).
One pathway of cysteate dissimilation has been elucidated, in Paracoccus pantotrophus NKNCYSA . Cysteate is transported into the cell and transaminated to 3-sulphopyruvate, which is subsequently reduced to 3-sulpholactate. SuyAB (3-sulpholactate sulpho-lyase) [EC 4.4.1.-, the full EC number has not yet been assigned] cleaves the C-sulphonate bond to yield sulphite and pyruvate. The results indicate that the last step in the pathway is excretion of sulphate, which is generated from sulphite by a cytochrome c-coupled sulphite dehydrogenase that we assume to have some phylogenetic relationship to the characterized bacterial sulphite dehydrogenase [SorAB (sulphite oxido-reductase)]. SorA is a molybdopterin-coupled oxidoreductase, SorB a cytochrome c . The gene believed to encode the sulphate exporter, suyZ, is a paralogue of tauZ in the degradation of taurine and is co-transcribed with suyB .
Rein et al.  noted, however, that a different degradative pathway seemed to operate in most of their cysteate-dissimilating organisms. This pathway presumably involved a 35 kDa protein, which was strongly induced in cysteate-grown cells. The aim of the present study was to characterize this protein and its activity. However, each organism which Rein et al. used , had the disadvantage that its genome sequence is unavailable, and we were fortunate to obtain an alternative organism whose sequenced genome was available.
González et al.  detected that Silicibacter pomeroyi DSS-3T dissimilated cysteate, but it is clear from the genome sequence , that the bacterium does not encode orthologues of SuyAB. The organism, however, encodes a potential orthologue (SPOA0157) of SuyZ , which is located on the megaplasmid. As both the suyZ and tauZ genes are found immediately downstream of the gene whose product releases sulphite [10,14], we examined the gene upstream of SPOA0157. SPOA0158 was annotated as a 36.5 kDa protein in COG2515, which includes D-cysteine desulphhydrase [EC 18.104.22.168] and ACC (1-aminocyclopropane-1-carboxylate) deaminase [EC 22.214.171.124]. D-Cysteine desulphhydrase is a pyridoxal 5′-phosphate-coupled enzyme, which converts D-cysteine into pyruvate, and the sulphide and ammonium ions (e.g. [15,16]). We could thus hypothesize that the gene product from SPOA0158 catalyses the conversion of L-cysteate into pyruvate and the sulphite and ammonium ions (Figure 1).
We now report the presence of inducible CuyA (L-cysteate sulpho-lyase) in S. pomeroyi DSS-3. The enzyme was purified and characterized as a pyridoxal 5′-phosphate-coupled enzyme, and its corresponding gene cuyA identified. CuyA activity was found in some of the organisms used by Rein et al.  that did not contain SuyAB.
L-Cysteic acid (>99% purity), D-cysteine (99% purity) and L-cysteine hydrochloride (>99% purity), L-cysteine sulphinate (99% purity) and 1-aminocyclopropane-1-carboxylic acid (>99%) were purchased from Sigma–Aldrich. DL-3-Sulpholactate was synthesized by R. Gueta  (Department of Biology, The University, D-78457 Konstanz, Germany) as described in . Oligonucleotides were synthesized by Hermann GbR Synthetische Biomoleküle (Denzlingen, Germany). Taq DNA polymerase and M-MuLV reverse transcriptase were obtained from Fermentas GmbH (St Leon-Rot, Germany) and they were used as specified by the supplier. RNase-free DNase was purchased from Qiagen. Chromosomal DNA was isolated from bacteria as described by Desomer in . Total RNA was isolated using the E.Z.N.A. bacterial RNA kit (Peqlab Biotechnologie GmbH, Erlangen, Germany). A 50 bp DNA ladder (Fermentas) was used.
Organisms, growth, harvesting of cells and preparation of cell-free extracts
S. pomeroyi DSS-3T (DSM 15171)  and Desulfomicrobium norvegicumT (DSM 1741) were obtained from the German Culture Collection (DSMZ, Braunschweig, Germany). Bilophila wadsworthia RZATAU (DSM 11045) and Desulfovibrio sp. strain RZACYSA, which respire using L-cysteate as electron acceptor to yield sulphide and acetate, and Desulfovibrio sp. strain GRZCYSA (DSM 11493), which ferments cysteate to acetate, sulphate and sulphide, were from our own laboratory [19,20].
Cells of S. pomeroyi DSS-3 were grown aerobically at 30 °C in modified SBM-M (Silicibacter basal medium) [Dr J. R. Henriksen, personal communication (Department of Microbiology, University of Georgia, Athens, GA, U.S.A.)] which contained 200 mM NaCl, 50 mM MgCl2, 10 mM KCl, 10 mM CaCl2, 50 mM Pipes, 1 mM K2HPO4, 10 mM NH4Cl, 0.07 mM FeEDTA and trace amounts of vitamins and minerals cited in : the sole added source of carbon and energy was 10 mM L-cysteate, unless otherwise stated. Precultures (3 ml) were grown in 30 ml screw-cap tubes in a roller. Growth experiments were done on the 100 ml scale in 500 ml Erlenmeyer flasks shaken in a water bath. Samples were taken at intervals to measure attenuance (at 580 nm), to assay protein and to determine the concentrations of sulphate, sulphite, carbon source and the ammonium ion. Similar cultures were used to generate small amounts of cells for enzyme assays or for molecular analyses. Cultures (1 litre) for protein purification were grown in 5 litre Erlenmeyer flasks on a shaker. Cells were harvested at D580 0.5 (about 150 mg protein/litre) by centrifugation (15000 g, 20 min, 4 °C), washed with 50 mM potassium phosphate buffer, pH 7.5 (containing 5 mM MgCl2) and stored frozen (−20 °C). The same buffer served as extraction buffer. Cell-free extracts free of nucleic acids were generated after disruption by at least five passages through a French pressure cell set at 140 MPa . The membrane/particulate fraction was pelleted by ultracentrifugation (170000 g, 30 min, 4 °C) and the supernatant fluid was referred to as the soluble fraction.
B. wadsworthia RZATAU, Desulfovibrio sp. strain RZACYSA, Desulfovibrio sp. strain GRZCYSA, and D. norvegicum DSM 1741 were grown anoxically in carbonate-buffered, fresh-water medium (pH 7.2) with titanium(III)nitriloacetate as the reducing agent as described in . The medium for B. wadsworthia RZATAU contained 50 mM formate as the electron donor, that for Desulfovibrio sp. strain RZACYSA and D. norvegicum DSM 1741 contained 20 mM lactate as the electron donor.
L-Cysteate, L-cysteine sulphinate, taurine, and alanine were determined by HPLC after derivatization with fluoro-2,4-dinitrobenzene . The ammonium ion was routinely determined by the improved Berthelot reaction (salicylate in place of phenol)  or by the specific reaction with glutamate dehydrogenase . Sulphite was routinely determined as the fuchsin adduct  and occasionally measured by ion chromatography . Sulphate was measured as the attenuance of a suspension of insoluble BaSO4 . Pyruvate was routinely determined by the specific reaction of lactate dehydrogenase  and confirmed by MALDI-TOF-MS (matrix-assisted laser-desorption ionization–time-of-flight mass spectrometry) used in the positive-ion mode, after reaction with 1,2-phenylendiamine, by Dr K. Hollemeyer (University of the Saarland, Saarbrücken, Germany) .
Protein in whole cells was quantified by a Lowry-type reaction . Protein in extracts was assayed by protein-dye binding . Denatured proteins were separated in SDS/12% PAGE gels and stained with Coomassie-Brillant-Blue R250 . The N-terminal sequence of a blotted protein was determined after Edman-degradation and HPLC separation at TopLab (Martinsried, Germany).
L-Cysteate sulpho-lyase was assayed routinely as the formation of sulphite after 2 min incubation at 30 °C. The reaction mixture (1 ml) contained 50 μmol Tris/HCl buffer (pH 9.0), 25 μmol L-cysteate and 0.1–0.8 mg protein, with which the reaction was started. Addition of up to 1 μmol pyridoxal 5′-phosphate to crude extracts or purified protein did not increase the activity and was thus omitted. On occasion, the reaction mixture was used to determine the formation of pyruvate or of the ammonium ion. Alanine dehydrogenase was assayed as reduction of NAD+ in Caps [3-(cyclohexylamino)propane-1-sulphonic acid] buffer, pH 10.0 . 3-Sulpholactate sulpho-lyase was assayed discontinuously at 30 °C and pH 7.5 as the formation of sulphite (as described in ). Sulphite dehydrogenase [EC 126.96.36.199] was assayed photometrically with ferricyanide as electron acceptor at pH 7.5 . Beef-heart cytochrome c could not replace ferricyanide.
Purification of CuyA
A three-step purification protocol was used. The soluble fraction was diluted with water (1:2.5) and loaded on to an anion-exchange column (Mono Q 10/10; GE Healthcare) equilibrated with 20 mM potassium phosphate buffer (pH 7.5); the flow rate was 2 ml/min, and proteins eluted with a gradient to 0.5 M Na2SO4 in buffer as described in . CuyA eluted at about 100 mM Na2SO4. Active fractions were pooled, concentrated using Vivaspin concentrators (10 kDa cut-off; Sartorius), rebuffered in 20 mM potassium phosphate buffer (pH 7.5), containing 1.7 M ammonium sulphate, and subjected to hydrophobic interaction chromatography on phenyl Superose HR (5/5 column; GE Healthcare). A linear decreasing gradient of ammonium sulphate in 20 mM potassium phosphate buffer (pH 7.5) was applied, and CuyA eluted at 0 mM ammonium sulphate. Concentrated active fractions were loaded on to a gel filtration column (Superose 12 HR 10/30; GE Healthcare) and eluted at a flow rate of 0.4 ml/min with 50 mM potassium phosphate buffer (pH 7.5) containing 150 mM Na2SO4.
The reverse PCR primers listed below were used for the RT-PCR reactions. PCR was performed as described previously in . The PCR primers were as follows: for cuyA, primer combination cuyAF (5′-atctggatcaagcgcgacgactg-3′) and cuyAR (5′-gcatctcgaaagcgcagttgaca-3′), and for cuyZ, primer combination cuyZF (5′-gacactgttgcggttcggggtc-3′) and cuyZR (5′-ccgcagatagccaccgacccac-3′). Positive controls for RNA integrity after isolation were done using the 16S rRNA-specific primers 27F and 533R . For the combination cuyAZ, primer combination cuyAF and cuyZR was used. RT-PCR products were visualized on 1.5–2% agarose gels.
Software for sequence analyses and accession numbers
Sequence analyses of the genome (accession number NC_003911) and the megaplasmid (accession number NC_006569) of S. pomeroyi DSS-3 were performed using the BLAST algorithm on the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov/). Sequence data were manipulated with different subroutines from the LASERGENE program package (DNASTAR). Transmembrane regions were predicted using the program TMHMM at the Center for Biological Sequence Analysis (CBS; http://www.cbs.dtu.dk/services/).
Growth of S. pomeroyi DSS-3 with sulfonates
Strain DSS-3 grew exponentially in L-cysteate-SBM-M with a μ (specific growth rate) of 0.07 h−1 (Figure 2A). The utilization of cysteate was concomitant with growth and was quantitative (Figure 2B), and approx. 80% of the sulphonate moiety was recovered as sulphite, the remainder as sulphate (results not shown). Some 80% of the cysteate-nitrogen, as the ammonium ion, was excreted into the growth medium concomitantly with growth (Figure 2B); the balance of the cysteate-nitrogen was calculated to be in cell material. The molar growth yield was 5.0 g of protein per mol C (a value which indicates quantitative utilization of the carbon source ), so, together with the growth rate, a specific degradation rate for cysteate of 1.3 mkat per kg of protein could be calculated.
Strain DSS-3 grew quantitatively with taurine or isothionate and excreted solely sulphate (and the ammonium ion, if appropriate) (results not shown). Cysteine sulphinate was a good growth substrate (with excretion of 20% sulphite and 80% sulphate); D-cysteine, L-cysteine and 1-aminocyclopropane-1-carboxylate were not utilized. The organism grew slowly with a poor yield (in 21 days compared with 3–5 days for other substrates) in DL-sulpholactate-SBM-M, and excreted sulphate (10% of the theoretical yield), but no sulphite. The growth medium interfered with the ion chromatographic analysis of sulpholactate, so we were unable to determine its fate.
CuyA in crude extracts of S. pomeroyi DSS-3
Soluble extracts of cysteate-grown cells of strain DSS-3 catalysed the hypothesized CuyA reaction, whereas extracts of acetate- or taurine-grown cells did not. The specific activity of CuyA in the crude extract (7.7 mkat per kg of protein; Table 1) was higher than the specific degradation rate (1.3 mkat (kg protein)−1) in growing cells and thus was sufficient to support the growth rate of the organism. The reaction products pyruvate (three independent methods), ammonium (two independent methods) and sulphite (two independent methods) were thoroughly identified. The pH optimum for the enzyme in crude cell extract was determined to be 8.8–9.0 when using a 50 mM Tris buffer in the pH range 7.5–9.0, whereas in sodium carbonate buffer pH 9.0, in Caps buffer at pH 9.5 or in 50 mM potassium phosphate buffer at pH 7.5, a lower activity was obtained.
The proteins in soluble extracts from cysteate-grown and acetate-grown cells were separated by SDS/PAGE (Figure 3). The major difference was the presence in cysteate-grown cells of two strongly-induced bands at 39 kDa and 42 kDa. The 39 kDa protein was shown to represent the cysteate sulpho-lyase (see below). The 42 kDa protein was isolated and determined to be alanine dehydrogenase, whose high activity was observed in crude cell extracts (Table 2). SuyAB activity was not detected (Table 2), which excluded the possibility that the 42 kDa protein represented SuyB (a 42 kDa protein ).
Purification and some properties of CuyA from S. pomeroyi DSS-3
Proteins in the soluble fraction of cysteate-grown cells were separated on an anion-exchange column. Two peaks of activity, measured as production of sulphite from L-cysteate, were observed. The minor peak eluted non-reproducibly at low salt concentrations often in the apparent absence of protein, whereas the major peak eluted reproducibly at about 100 mM Na2SO4 in the presence of protein. We worked with the major peak only, which catalysed the CuyA reaction and contained the 39 kDa protein (Figure 3). Further separation on hydrophobic-interaction and gel-filtration columns yielded an effectively homogeneous protein with an apparent molecular mass of approx. 39 kDa (Figure 3). Unit stoichiometry of all three products from the reaction, pyruvate, ammonium and sulphite ions was observed (Figure 4). The Km value was determined to be 11.7±2.1 mM. Determination of the stoichiometry of substrate disappearance and product formation was impractical, because 25 mM cysteate had to be added in the assay and during the linear phase of the assay [maximally 8 min (Figure 4)], and a turnover of only 0.3 mM substrate was obtained based on the quantification of the end products.
The specific activity of CuyA from S. pomeroyi did not apparently rise during purification, although the enzyme was effectively homogeneous after the third step of the purification. This might be due to inhibition by sulphate ions, which were present in all separated fractions, although desalting improved the specific activity only slightly (results not shown). The cofactor pyridoxal 5′-phosphate was tightly bound to the enzyme and its addition did not enhance the specific activity. CuyA from S. pomeroyi DSS-3 could be stored for at least 2 months at 4 °C without significant loss of activity. Frozen samples lost about 90% of their activity on thawing.
A UV–visible spectrum (pH 9.0) of the purified enzyme, which was yellow, gave maxima at 278 and 412 nm and a minimum at 260 nm, which supported the presence of bound pyridoxal 5′-phosphate, whose spectrum in the same buffer had maxima at 278 and 411 nm and a minimum at 260 nm.
D-Cysteine was converted into sulphide and pyruvate (and presumably the ammonium ion) by CuyA from S. pomeroyi, but at 15% of the rate with L-cysteate; L-cysteine was not a substrate. Neither L-cysteine sulphinate nor ACC was a substrate for the enzyme.
The N-terminal amino-acid sequence of CuyA was unambiguously MHLARYP, which was identical with the deduced N-terminal sequence of SPOA0158. This was the gene we predicted (see the Introduction). The deduced molecular mass of this protein, however, was 36.5 kDa, whereas we routinely observed 39 kDa for the denatured protein (Figure 3). Altered preparation procedures for SDS/PAGE did not significantly alter this value. The molecular mass of the native enzyme was estimated by gel filtration chromatography to be 100±6 kDa, indicating a homomultimeric enzyme.
Sulphite dehydrogenase in crude extracts of S. pomeroyi DSS-3
The almost quantitative excretion of sulphite during growth with cysteate (Figure 2) was initially attributed to the absence of a sulphite dehydrogenase, because no convincing homologue of SorA  was found on the genome of S. pomeroyi DSS-3. However, the excretion of sulphate during growth with taurine caused us to assay extracts for sulphite dehydrogenases. No cytochrome c-coupled activity was detected, but inducible ferricyanide-coupled oxidation of sulphite was detected in crude cell extracts of both cysteate-grown (3.2 mkat per kg of protein) and taurine-grown (3.8 mkat per kg of protein) cells of S. pomeroyi DSS-3 (Table 2). The level in crude cell extracts of acetate-grown cells was below the detection limit (<0.03 mkat (kg protein)−1). The dehydrogenase was found in the soluble fraction of the extract after ultracentrifugation (7.0 mkat (kg protein)−1) and an apparent Km value of 10.7±3.7 mM was observed.
Genes involved in cysteate degradation and their transcription in S. pomeroyi DSS-3
CuyA, encoded by cuyA, was identified as SPOA0158, which is located on the megaplasmid present in S. pomeroyi DSS-3. The locus is annotated as a 36.5 kDa protein in COG2515, which includes D-cysteine desulphhydrase and ACC deaminase (Supplementary data, Figure S1 http://www.BiochemJ.org/bj/394/bj3940657add.htm). Directly downstream of cuyA, a potential orthologue (SPOA0157) of SuyZ  is found. A potential LysR-type regulator (SPOA0159), which we termed cuyR, is located upstream of cuyA, on the complementary strand (Figure 1B). A chromosomal paralogue of cuyA (SPO2657) was obviously not involved in cysteate metabolism.
The gene product of cuyZ was predicted to be a sulphate exporter (see Introduction section), whereas excretion of sulphite was observed (Figure 2). RT-PCR-data showed that cuyZ was transcribed in cysteate-grown cells (Figure 5), but not in either taurine-grown (Figure 5) or in acetate-grown cells. A 1358 bp amplicon spanning cuyAZ was also observed (results not shown), so it was evident that the two genes were co-transcribed during growth with cysteate.
CuyA activity in other bacteria
Rein et al.  postulated that the inducible, 35 kDa protein in extracts of many cysteate-utilizing bacteria represented a new desulphonative enzyme. The results from the present study indicate that the enzyme could be CuyA. Cell extracts from B. wadsworthia RZATAU and from Desulfovibrio spp. strains GRZCYSA and RZACYSA were examined. Each organism contained inducible CuyA activity. CuyA from B. wadsworthia RZATAU was purified 10-fold to near homogeneity on the anion-exchange column (results not shown): this was the 35 kDa protein (Table 1).
cuyA-like genes in genome projects
We found cuyA-like genes in the sequences from several genome projects (see Supplementary data, Figure S1 http://www.BiochemJ.org/bj/394/bj3940657add.htm). These genes are often designated as COG2515, whose most prominent characterized protein is ACC deaminase. The deduced protein sequences had <62% identity with CuyA. The E. coli cysteine desulphhydrase (YedO, also known as DcyD) had 45.5% sequence identity with CuyA, the characterized ACC deaminases <40% identity. One complete CuyA homologue from the Sargasso Sea metagenome project  had 64% identity with CuyA (see Supplementary data, Figure S1), while several fragments (single reads) had similar sequence identities (results not shown).
The relatively high level of identity of some orthologues led us to the hypothesis that these proteins could also encode CuyAs. However, neither Desulfotalea psychrophila LSv54 (two paralogues sharing 51.2% and 46.4% sequence identity with CuyA, respectively) nor Caulobacter crescentus CB15 (59.7% sequence identity to CuyA) grew with cysteate [Dr R. Rabus, personal communication (Max Planck Institute for Marine Microbiology, Bremen, Germany); Dr A. Sporman, personal communication (Stanford University, Palo Alto, CA, U.S.A.), so we presume that the products of these cuyA-like genes are other enzymes in COG2515, such as D-cysteine desulphhydrase or ACC deaminase.
The data, which indicated that S. pomeroyi grew with L-cysteate as a sole source of carbon and energy , have been confirmed in medium in which the mass balances for carbon, nitrogen and sulphur can be calculated (Figure 2). A key hypothesis, for the cysteate degradative pathway involved in growth, was a novel desulphonative enzyme, CuyA (see the Introduction and Figure 1). This hypothesis has been confirmed experimentally (Figure 4) with a pure enzyme (Figure 3), as was predicted based on the identity of the corresponding gene, SPOA0158. The presumed reaction mechanism of β-elimination of sulphite followed by deamination , though a novel desulphonation reaction, falls within the standard reaction range of enzymic transformations involving pyridoxal 5′phosphate . The enzyme thus carries out a lytic reaction that is not a hydrolysis, and can be attributed to the lyases [EC 4], specifically the ‘carbon-sulphur lyases’, EC 4.4, which has only one subgroup, EC 4.4.1. We propose that the enzyme (EC number has not yet been assigned) could have the systematic name ‘L-cysteate sulpho-lyase (deaminating)’ and the common name ‘L-cysteate sulpho-lyase’. The fact that the enzyme can catalyse the D-cysteine desulphhydrase reaction does not alter the nomenclature, because the organism does not dissimilate D-cysteine, and the biological function is L-cysteate sulpho-lyase.
Native CuyA may be trimeric, based on the facts that the N-terminal sequence analysis gave a reliable start codon, on a reliable subunit molecular mass of 36.5 kDa (derived) to 39 kDa (observed), and on a value of 100 kDa for the native enzyme derived from gel filtration, which, unsupported, is not a robust method . The first bacterial ACC deaminase also appeared to be trimeric, based on gel filtration data , whereas the enzyme of the yeast Williopsis (Hansenula) saturnus is a dimer, based on both chromatographic and crystal structure data . The L-cysteine desulphhydrase from E. coli is described as being a dimer (gel filtration and ultracentrifugation) . A conservative conclusion is that CuyA is homomultimeric.
CuyA is induced to about 10% of soluble cell protein (e.g. Figure 3). This high level of expression is typical of all the organisms in our studies (results not shown) and elsewhere . The enzyme has a poor Km value, 11.7 mM, which might contribute to the requirement for a large amount of enzyme. We thus speculate that a powerful pump is required to supply cysteate at adequate concentrations in the cell. It has been noted that ACC deaminases have high Km values . The high level of induction may involve the hypothetical LysR-type transcriptional regulator, cuyR, encoded upstream of cuyA in the opposite orientation (SPOA0159; Figure 1B). The spacing between cuyA and cuyR would allow this type of regulation of cuyAZ, and three possible LysR-boxes [48,49] are located in the promoter region between cuyA and cuyR, so there is circumstantial evidence for transcriptional regulation by CuyR.
CuyA was first hypothesized from data in the genome of the α-proteobacterium S. pomeroyi DSS-3, but it is also found in several δ-proteobacteria (Table 1). We thus presume the enzyme to be widespread, although it is not represented in currently accessible genome sequences (Supplementary data, Figure S1 http://www.BiochemJ.org/bj/394/bj3940657add.htm). We also presume that CuyA takes part in a degradative pathway for 3-sulpholactate, because all tested organisms which encode CuyA degrade 3-sulpholactate (Table 1). We hypothesize that sulpholactate is converted into cysteate prior to desulphonation , a pathway implicit in the synthesis de novo of the bacterial sulpholipid, capnine [4,50].
The cuyAZ-genes are co-transcribed (Figure 5), analogous to the suy(A)BZ genes in the degradation of L-cysteate in P. pantotrophus NKNCYSA , so they presumably represent an operon; the up- and downstream genes are on the complementary strand. CuyZ is deduced to have a leader peptide and nine membrane-spanning helices. We again  presume, that the only metabolic function, which is equally important to the cell as desulphonation (CuyA), is the excretion of the excess inorganic sulphur oxyanion (CuyZ) to maintain homoeostasis in the cell.
The relatively simple desulphonation reaction, to cleave the stable C-sulphonate bond and release pyruvate into the amphibolic pathways to allow energy conservation and growth (Figure 1), thus represents only a small portion of the solution to the metabolic problem of sulphonate dissimilation. The nature of cysteate uptake is still unknown. The sulphite (and ammonium) ions must be safely disposed of [10,14]. Sulphite is usually converted into and excreted by aerobes as sulphate, sometimes with a significant portion of sulphite [10,34,51–53]. The excretion of sulphite, with some sulphate, is novel, though we recently detected it in Paracoccus versutus N-MT utilizing N-methyltaurine [53a]: a dendrogram of sequences of putative exporters does not show an especially close similarity of CuyZ and TauZ (from strain N-MT) (Supplementary material, Figure S2 http://www.BiochemJ.org/bj/394/bj3940657add.htm). We also detected an induced sulphite dehydrogenase in strain N-MT [53a]. We speculate that the presumed sulphite-exporter, CuyZ (TauZ in strain N-MT), has a higher affinity for sulphite than the sulphite dehydrogenase (Km=0.7 mM). Bacterial sulphite dehydrogenases, especially the ferricyanide-coupled enzymes , are poorly characterized, as is their regulation, while the sulphate and sulphite exporters have only just been recognized [10,14].
Given the wide range of amino acids degraded by S. pomeroyi DSS-3 [12,13], we presume that excretion of ammonium occurs under many nutritional conditions and independent of any specific amino acid. This general process, though recognized for many decades, does not seem to have been explored in this context, though the structure of an AmtB (ammonia/methylammonia transport) protein was recently established , and in S. pomeroyi DSS-3, amt-1 (SPO 2093; closest orthologue to E. coli amtB) is expressed under all conditions tested (A. K. Gorzynska, K. Denger, A. M. Cook and T. H. M. Smits, unpublished work).
We are grateful to Ms Andzelika K. Gorzynska for the RT-PCR experiments, to Professor M. A. Moran (Department of Marine Sciences, University of Georgia, Athens, GA, U.S.A.) for advice on growing S. pomeroyi, to Dr R. Rabus (MPI Bremen) and to Dr A. Spormann (Stanford University, Palo Alto) for testing the growth of D. psychrophila and C. crescentus with cysteate, and to Dr H. B. F. Dixon (Department of Biochemistry, Cambridge University, Cambridge U.K.) for valuable discussions. This work was supported by the University of Konstanz. The International Association for the Exchange of Students for Technical Experience (IAESTE) funded Ms Gorzynska's visit.
Abbreviations: ACC, 1-aminocyclopropane-1-carboxylate; Caps, 3-(cyclohexylamino)propane-1-sulphonic acid; CuyA, L-cysteate sulpho-lyase; L-cysteate, 2-amino-3-sulphopropionate; RT, reverse transcriptase; SBM-M, modified Silicibacter basal medium; SorAB, sulphite oxido-reductase; SuyAB, 3-sulpholactate sulpho-lyase
- The Biochemical Society, London