Duplication of the mmoX gene in Methylosinus sporium: cloning, sequencing and mutational analysis

Methylosinus sporium strain 5 is a type II obligate methanotroph, originally isolated by Whittenbury and co-workers (Whittenbury et al., 1970), which oxidizes methane to methanol using the enzyme methane monooxygenase (MMO). Like other type II methanotrophs, such as Methylosinus trichosporium OB3b and Methylocystis sp. strain M, Ms. sporium has the ability to express two forms of MMO, a membrane-bound particulate form (pMMO) and a cytoplasmic soluble form (sMMO) (Pilkington & Dalton, 1991). One of the distinctive characteristics of Ms. sporium, compared to other species of Methylosinus, is its ability to produce a brownblack water-soluble pigment (Whittenbury et al., 1970).

The genes encoding pMMO are present in all known methanotrophs with the exception of Methylocella, a facultative genus of methanotrophs (Dedysh et al., 2000, 2004; Dunfield et al., 2003; Theisen et al., 2005). The sMMO enzyme, however, is only found in some methanotrophs, and the most extensively characterized sMMO enzymes are those from Methylococcus capsulatus (Bath) (Colby & Dalton, 1976; Green & Dalton, 1985; Woodland & Dalton, 1984) and Ms. trichosporium OB3b (Fox et al., 1989). The purified enzyme is composed of a three-component hydroxylase (αβγ)₂ encoded by mmoXYZ, a reductase encoded by mmoC, and a regulatory protein, protein B, encoded by mmoB (Cardy et al., 1991a, b; Pilkington & Dalton, 1991; Stainthorpe et al., 1989). The α subunit of the hydroxylase contains a binuclear iron centre, which is essential for catalysis (Elango et al., 1997; Rosenzweig et al., 1993). The reductase MmoC contains Fe₂S₂ sites and an FAD cofactor located in the C-terminal domain, and is responsible for electron donation (Lund et al., 1985). MmoB has no metal or prosthetic groups, but is essential for activity and electron transfer from the reductase to the hydroxylase (Green & Dalton, 1985). The genes composing the sMMO operon (mmoXYBZDC) have been cloned and sequenced from several methanotrophs, and are present as a single copy in the chromosome (Murrell et al., 2000), whereas the genes encoding pMMO, pmoCAB, can be present in multiple copies (Semrau, 1995; Stolyar et al., 1999).

The molecular regulation of the expression of MMO in methanotrophs which possess both sMMO and pMMO is not fully understood. It is known that the intracellular location of MMO activity is dependent on the copper-to-biomass ratio during growth (Stanley et al., 1983). The pMMO is expressed during growth under a high copper-to-biomass ratio, and the sMMO is expressed under a low copper-to-biomass ratio. It is also not fully resolved whether the decrease in pMMO activity during growth under a low copper-to-biomass ratio is due to repression of transcription, or whether it is due to lack of availability of copper ions, since pMMO requires copper for activity (Nguyen et al., 1998; Zahn & DiSpirito, 1996). The sMMO operon is transcribed from a σ⁵⁴ promoter located upstream of mmoX, and sMMO expression is regulated by copper ions at the level of transcription (Nielsen et al., 1996, 1997), with the exception of Methylocella silvestris BL2 (Nielsen et al., 1996, 1997; Theisen et al., 2005). Transcription driven from σ⁵⁴ promoters requires a σ⁵⁴-dependent regulator (Shingler, 1996). The σ⁵⁴-dependent transcriptional regulator designated MmoR has been shown to be essential for sMMO expression in Mc. capsulatus and Ms. trichosporium (Csaki et al., 2003; Stafford et al., 2003). A second gene essential for sMMO expression, mmoG, encoding a GroEL homologue, has also been identified (Csaki et al., 2003; Stafford et al., 2003).

sMMO from Ms. sporium has been characterized at the biochemical level (Pilkington & Dalton 1991). Until this present study, no sequence information has been available for sMMO from Ms. sporium, except that for partial mmoX sequences (GenBank accession nos. AJ458528, AJ458525, AJ458520, AJ458512 and AJ458511). In this study, we describe the cloning and sequencing of the complete sMMO operon, including mmoG and mmoR. We also report the identification of duplicate copies of mmoX which have not been found in any other methanotrophs investigated. The role of the duplicate copies of mmoX was investigated by phenotypic characterization of mmoX knock-out mutants. This gave some interesting insights into the production of the water-soluble pigment, and its possible role as a siderophore for iron acquisition in Ms. sporium. Finally, we complemented the sMMO mutant by heterologous expression of sMMO genes from Ms. trichosporium and Mc. capsulatus.

Bacterial strains and growth conditions.
Bacterial strains used in this study are shown in Table 1. Methanotrophs were routinely grown on nitrate mineral salts (NMS) medium (Whittenbury et al., 1970) at 30 °C, either in batch cultures with a headspace of methane and air (1 : 5), or on NMS agar plates using the same conditions. Large amounts of biomass were obtained by growing methanotrophs in a 5 l fermenter (Inceltech LH Series 210) with a continuous flow of air (1 l min¹) and methane (140 ml min¹). pH of the culture was maintained between 6.9 and 7.1 with the automatic addition of 0.5 M HCl or 0.5 M NaOH. Growth was monitored by measuring OD₅₄₀ using a Beckman DU-70 spectrophotometer, and by monitoring the dissolved oxygen tension in the fermenter vessel using an oxygen electrode. To prevent limitation of oxygen, and to maintain optimal growth conditions, the dissolved oxygen levels were maintained above 5 % by increasing the agitation speed and the airflow rate. Culture purity was routinely monitored by microscopic observation and streaking cultures onto nutrient agar plates. pMMO- and sMMO-expressing cells were obtained by cultivation of methanotrophs on NMS medium containing copper (1 µM CuSO₄), and NMS medium without added copper, respectively. Under normal growth conditions, Ms. sporium was cultivated with 1 µM FeEDTA, and with 5 µM FeEDTA for growth under excess iron conditions. All Escherichia coli strains were maintained in LuriaBertani medium. Antibiotics were used, as required, for E. coli, at the following final concentrations: ampicillin (50 µg ml¹), kanamycin (25 µg ml¹) and gentamicin (15 µg ml¹).

Table 1. Bacterial strains, plasmids and primers used in this study

DNA manipulations.
All DNA manipulations were carried out according to Sambrook et al. (1989), unless otherwise stated. Small-scale plasmid preparation from E. coli TOP10 was performed using the Qiaprep Spin Miniprep kit (Qiagen). DNA from methanotrophs was extracted using the method described by Marmur (1961). All plasmids used in this study are listed in Table 1.

PCR.
PCR amplifications were performed in total volumes of 50 µl. Following the initial denaturation step at 94 °C for 5 min, 2.5 U Taq DNA polymerase (MBI Fermentas) were added. All PCR amplifications were performed using 30 cycles of 94 °C for 1 min, 5060 °C for 1 min (adjusted according to the annealing temperature of the primers) and 72 °C for 1 min per kilobase of DNA amplified, and a final extension step at 72 °C for 10 min. Typically, 100 ng of either plasmid or genomic DNA was used as a DNA template. PCR products were routinely gel purified using the QIAquick Gel Extraction Kit (Qiagen), and used directly in cloning reactions or via the pCR2.1-TOPO vector using the TA-TOPO Cloning kit (Invitrogen).

Random priming, Southern blotting and colony hybridization.
Aliquots of genomic DNA of ∼35 µg were digested with various restriction enzymes, and the DNA fragments were resolved on a 0.9 % (w/v) agarose gel in Tris/acetate/EDTA (TAE) buffer. The DNA was transferred by capillary blotting onto Hybond N+ membranes (Amersham), and fixed by UV crosslinking using a Stratalinker (Stratagene). Approximately 25 ng purified PCR products were labelled with 50 µCi (1.9 MBq) [α-³²P]dGTP by random priming with 10 U Klenow fragment and random hexanucleotides (Roche), according to the manufacturer's instructions. Typically, the membranes were hybridized overnight at 50 °C with the labelled probes with 1 ml hybridization buffer (0.5 M sodium phosphate buffer, pH 7.2, 5 mM EDTA, pH 8.0, 7 % (w/v) SDS) per cm² of Hybond N+ membrane. Following hybridization, the membranes were washed with 2x saline sodium citrate (SSC; 17.3 g NaCl l¹, 8.8 g trisodium citrate l¹, pH 7.0) as described by Sambrook et al. (1989), and DNA fragments hybridizing to the labelled probes were identified by autoradiography. These fragments were targeted for cloning by ligating size-selected restriction enzyme digests of Ms. sporium DNA into pUC19. Libraries of clones were generated by transforming E. coli TOP10 cells, which were screened by colony hybridization. Clones identified by colony hybridization were fully sequenced by primer walking.

Inverse PCR.
Inverse PCR was used to complete the sequencing of the sMMO operon. Briefly, 1 µg genomic DNA was digested with a restriction enzyme that cut frequently (PstI). DNA was then self-ligated in 50 µl reaction volumes using T4 ligase (MBI Fermentas) to generate small circular molecules. The ligation mix was then used directly as the PCR template, and outward facing primers were designed with positions near the end of the known DNA sequence. Primers used in this study are listed in Table 1. The PCR products were purified and sequenced directly using the outward facing primers.

Isolation of total RNA.
A hot-phenol method (Gilbert et al., 2000) was used to isolate total RNA from 4 ml exponential-phase fermenter cultures of Ms. sporium, expressing either pMMO or sMMO. DNA was removed from the total RNA samples using 24 U DNase I (Promega) per microgram of total nucleic acid, according to the manufacturer's instructions. Using the DNase I-treated RNA as a PCR template, the removal of all traces of DNA was confirmed by the absence of a 16S rRNA PCR product.

RNA dot blots and RT-PCR.
Various amounts of RNA (0.53 µg), extracted from pMMO- and sMMO-expressing cultures, were blotted onto Hybond N+ membranes, according to the dot hybridization of purified RNA method outlined in Sambrook et al. (1989). mmoX DNA amplified by PCR from Ms. sporium was used as a probe (∼25 ng) for hybridization and as a positive control (∼6 ng). The first strand cDNA synthesis for RT-PCR was carried out using SuperScript II reverse transcriptase (RT) (Invitrogen). RNA (0.51 µg) was added to 50 pmol gene-specific reverse primer and 1 µl dNTP mix (10 mM each) in a final volume of 12 µl. The mixture was heated to 65 °C for 5 min and then chilled on ice. To the reaction mixture, 4 µl 5xfirst strand buffer, 2 µl 0.1 M DTT and 1 µl (200 U) SuperScript II RT were added to give a 20 µl final volume before incubation at 42 °C for 50 min, followed by 15 min at 70 °C to inactivate the RT enzyme. The cDNA (2 µl) was used as a template for PCR amplification without further purification.

Construction and confirmation of mutant mmoX strains.
Two regions flanking mmoX1 and mmoX2 were amplified by PCR and were designated products AD (Fig. 1a, b). In order to facilitate cloning of these products into the cloning vector pK18mob, XbaI and SphI sites were introduced into the forward and reverse primers of products A and C, respectively, and SphI and HindIII sites into the forward and reverse primers of products B and D, respectively. Products A and B, and C and D, were cloned into pK18mob via XbaI and HindIII sites to give constructs pMHA500.1 and pMHA501.1, respectively. A 913 bp SphI fragment from p34S-Gm, containing the gentamicin resistance (Gm^R) gene, was cloned via the SphI site between products A and B, and C and D, to give the final constructs pMHA500 and pMHA501, respectively (Fig. 1c). pMHA500 and pMHA501 were electroporated into E. coli S17.1 λpir strain (Herrero et al., 1990), and this was used as a donor strain for conjugation into Ms. sporium. Conjugations were carried out using the method of Martin & Murrell (1995). Transconjugants were initially selected on NMS agar plates containing gentamicin (5 µg ml¹), which selected for both single and double recombinants. These transconjugants were further subcultured onto replica plates containing kanamycin (15 µg ml¹) and gentamicin (5 µg ml¹), and double recombinants were identified by screening for kanamycin sensitivity. The genotypes of the mutants were confirmed by PCR using primers designed outside the knock-out region.

(13K): Fig. 1. Strategy for constructing Ms. sporium ΔmmoX1 and ΔmmoX2 strains. (a, b) mmoX1 and mmoX2, and flanking DNA regions. Primers used to amplify the flanking regions of mmoX, yielding products AD, are indicated. The region between the SphI sites is the knock-out (KO) region and is indicated by the dotted lines. (c) Suicide plasmid constructs pMHA500 and pMHA501 used to knock out mmoX1 and mmoX2, respectively. (d) Physical maps of the chromosomal region containing the sMMO operons. Clones used for sequencing are indicated by thick lines. Inverse PCR products are indicated by horizontal dotted lines between the PstI sites. The unique PvuII and MscI sites within the mmoX genes are indicated (not all restriction sites are shown).

Analysis of polypeptides by SDS-PAGE and liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS).
Cells were harvested during the exponential-growth phase either from a 5 l fermenter culture (OD₅₄₀ 46) or from small batch cultures grown in 250 ml flasks (OD₅₄₀ 0.40.6). The cell pellets were resuspended in 25 mM MOPS buffer (pH 7) containing 1 mM benzamidine, 5 mM DTT and a few crystals of DNase. The cells were broken by two passages through a pre-cooled French pressure cell at 110 MPa. The cell-free extracts were then separated into soluble and particulate fractions by centrifugation (38 000 g for 60 min at 4 °C). The protein content in the soluble fraction was quantified using Bio-Rad reagent, and BSA was used as the protein standard. Protein samples were separated by 12.5 % (w/v) SDS-PAGE using an X-cell II Mini-Cell apparatus (Novex), and stained with Coomassie brilliant blue R250. The MmoX polypeptides corresponding to mmoX1 and mmoX2 were excised from SDS-PAGE gels, subjected to in-gel digestion, and analysed by LC-ESI MS/MS, using methods outlined by Schäfer et al. (2005).

Naphthalene assay.
A naphthalene oxidation assay (Brusseau et al., 1990) was routinely used to assay the activity of sMMO on NMS agar plates or in liquid cultures. Briefly, methanotrophs were grown on NMS medium containing no added copper. Once grown (810 days), 1 ml culture or NMS agar plates were incubated in the presence of a few crystals of naphthalene at 30 °C for 30 min. A few drops of tetrazotized o-dianisidine (10 mg ml¹) solution were then added, and the formation of a deep purple colour indicated sMMO expression and activity.

Cloning and sequencing of the sMMO operon, mmoG and mmoR
In order to clone and sequence the complete sMMO operon from Ms. sporium, probes for mmoX were initially generated by PCR from the same organism using the mmoX primers described by Hutchens et al. (2004), and a Southern blot containing Ms. sporium and Ms. trichosporium DNA was probed. As expected, Ms. trichosporium DNA probed with mmoX confirmed only a single copy of mmoX. Interestingly, Ms. sporium DNA probed with mmoX suggested that there may have been duplicate copies (Fig. 2).

(41K): Fig. 2. Southern blot showing duplication of mmoX genes in Ms. sporium. The Southern blot contained Ms. trichosporium digested DNA (lanes 14) and Ms. sporium digested DNA (lanes 58), which was probed with a 32P-labelled mmoX gene probe. The 4.3 and 5 kb EcoRI fragments used for the subsequent cloning are indicated.

The duplication of mmoX in Ms. sporium was subsequently verified by cloning and sequencing the two copies of the mmoX gene. An EcoRI library of ∼45 kb genomic DNA fragments was constructed. By probing with the mmoX gene, two clones, designated pHA006 and pHA007, containing a 4.3 and a 5.0 kb EcoRI insert, respectively, were fully sequenced (Fig. 1d). Clone pMHA007 contained mmoG, mmoX and a partial sequence of mmoY. Clone pHA006 only contained mmoX, with no additional ORFs identified 5' or 3' of the mmoX gene. In order to target the remainder of the sMMO structural genes, and to investigate duplication of other sMMO structural genes, mmoY was PCR-amplified from pHA007 using primers mmoY_F and mmoY_R (Table 1), and another Southern blot was probed. The Southern blot indicated that mmoY was most likely present as a single copy in the chromosome. Following the screening of a ClaI library with mmoY, a 5.7 kb DNA fragment was identified (Clone pHA010), which was cloned and sequenced (Fig. 1d). Clone pHA010 contained a partial sequence of mmoC and all the other structural genes for sMMO (mmoXYBZDC). Inverse PCR using mmoC_F and mmoC_R, and mmoG_F and mmoG_R primers, was used to complete the sequence of mmoC and to obtain a partial DNA sequence of mmoR, located 5' of mmoG. The mmoX in the full sMMO operon was designated mmoX1 and the lone mmoX was designated mmoX2 (Fig. 1d).

Sequence analysis of the sMMO operon, mmoG and mmoR
The arrangement of the structural genes encoding sMMO in Ms. sporium is similar to those sequenced previously (Cardy et al., 1991b; McDonald et al., 1997; Shigematsu et al., 1999; Stainthorpe et al., 1990). Alignment of the sMMO gene cluster with other sMMO gene clusters clearly demonstrated that Ms. sporium is most closely related to Methylocystis sp. strain M, with nucleotide sequences of 8197 % identity. The nucleotide sequence identity of mmoX1 and mmoX2 with Methylocystis sp. strain M was 96 and 88 %, respectively, and there was 89 % identity between mmoX1 and mmoX2. The evolutionary relatedness of the duplicate copies of mmoX was further investigated by phylogenetic analysis using reference mmoX sequences from GenBank (Fig. 3). Both copies of mmoX from Ms. sporium branched within mmoX sequences from other type II methanotrophs. The mmoX1 sequence was closest to other Ms. sporium mmoX sequences, with ∼97 % identity, whereas the mmoX2 sequence was closest to the mmoX sequence of Methylocystis species, with ∼90 % identity.

(11K): Fig. 3. Maximum-likelihood tree showing the relationship of the duplicate mmoX sequences (bold type) obtained in this study to the mmoX sequences available at the National Center for Biotechnology Information (NCBI) as analysed with the ARB software package (). The alignment was based on mmoX sequences longer than 1200 bp. The two main clades separating the Type I and Type II methanotrophs are indicated. The related butane monooxygenase (bmo) from Pseudomonas butanovora (Sluis et al., 2002), belonging to the γ-subdivision of the Proteobacteria, was used as an outgroup. The accession numbers of the reference mmoX sequences are shown in parentheses.

The two ORFs located upstream of mmoX1 were identified as mmoG and mmoR, encoding a GroEL homologue and a σ⁵⁴-dependent transcriptional regulator, respectively. Both of these genes have been identified and characterized in other methanotrophs, and are essential for sMMO expression (Csaki et al., 2003; Stafford et al., 2003). mmoG and mmoR are found in close proximity to the sMMO operon; however, the exact arrangement of these genes varies from one strain to another. The arrangement of mmoG and mmoR with respect to the sMMO operon in Ms. sporium is identical to that found in Ms. trichosporium OB3b (Stafford et al., 2003).

Transcription of the sMMO operon in Mc. capsulatus (Bath) and Ms. trichosporium (OB3b) is initiated from a σ⁵⁴ promoter located at the 5' end of mmoX (Csaki et al., 2003; Nielsen et al., 1996, 1997). A closer examination of the sequences upstream of mmoX in Ms. sporium revealed a highly conserved sequence resembling σ⁵⁴ recognition sites (Barrios et al., 1999), and putative transcriptional start sites were also identified. The σ⁵⁴ recognition site (TGGCAC-N₅-TTGCW) has been identified 5' of mmoX in all methanotrophic strains analysed (Table 2). The ShineDalgarno (SD) sequence located upstream of the initiation codon (AUG) for mmoX was also identified and aligned with other SD sequences for mmoX. The SD sequence (GAGGA), with the exception of mmoX2 of Ms. sporium, was found to be highly conserved amongst the sequences analysed. Upstream of the AUG of mmoX2, two overlapping SD sequences were identified. The first SD sequence had a mismatch of 1 bp to that of the consensus SD sequence, and was located 6 bp from the start codon. The second SD sequence was located 3 bp from the start codon, and was identical to the consensus SD sequence (Table 2). It is not known which of these potential SD sequences is involved with ribosome binding and thus initiation of translation from the AUG of mmoX2.

Table 2. Alignment of the mmoX σ54 promoters and the SD sequences of various methanotrophs The 24 and 12 σ54 recognition sites are highlighted in bold type, and the putative transcriptional start sites are highlighted in bold italic type. The distances between the transcriptional start sites and the initiation codon (AUG) are in parentheses. The conserved SD sequences are highlighted in bold type, and the distances between the SD sequence and AUG are indicated in parentheses. W=A or T.

Transcription and expression of mmoX in Ms. sporium
The strict transcriptional regulation of sMMO operons from mmoX σ⁵⁴ promoters by copper ions is a characteristic feature of methanotrophs containing both pMMO and sMMO (Murrell et al., 2000; Nielsen et al., 1996). RNA dot hybridization with the mmoX gene probe was used to investigate transcriptional regulation of sMMO in Ms. sporium. The mmoX DNA probe only hybridized with RNA extracted from sMMO-expressing cultures, indicating that transcription from the sMMO σ⁵⁴ promoter was totally repressed during growth under a high copper-to-biomass ratio (Fig. 4a). It is noteworthy that the mmoX gene probe used for probing the dot blot would have hybridized with both copies of mmoX. Therefore, in order to investigate whether both copies of mmoX were transcribed, RT-PCR was performed on RNA extracted from sMMO-expressing cultures. The cDNA was synthesized and amplified using the degenerate mmoX primers described by Hutchens et al. (2004), which yielded a 719 bp product. The PCR products containing the duplicate copies of the mmoX gene were distinguished from each other by digesting with the restriction enzymes PvuII or MscI. mmoX1 contained a unique PvuII site, and yielded two fragments of 429 and 291 bp when the PCR product was digested with PvuII. Conversely, the PCR product containing mmoX2 yielded two DNA fragments of 555 and 163 bp when digested with MscI. In order to clarify that the degenerate primers used were not biased towards one copy of mmoX, mmoX was amplified from genomic DNA, and digested individually with PvuII and MscI. Equivalent amounts of DNA fragments were obtained from both PvuII and MscI digestions, indicating that there was no inherent PCR bias towards a particular copy of mmoX (Fig. 4b). This also ruled out the possibility of inefficient digestion by either of the restriction enzymes. A similar experiment was performed in duplicate on cDNA generated using the mmoX primers. Judging by the amounts of the DNA fragments that were obtained following the selective digestion of the mmoX genes, the data clearly suggest that there were more transcripts corresponding to mmoX1 than mmoX2 (Fig. 4b). Based on these data, it can be concluded that both copies of mmoX were transcribed; however, transcription of mmoX1 was more efficient than that of mmoX2. It is noteworthy that, although RT-PCR cannot be used as an absolute measure of the relative quantification of mmoX mRNA in situ, it can be used as a crude method for the measurement of the relative abundance of mRNA transcripts in situ, with the support of appropriate controls.

(71K): Fig. 4. RNA dot blot and RT-PCR showing transcription of mmoX. (a) Dot blot containing RNA from pMMO- (top row) and sMMO-expressing cells (bottom row) hybridized with the mmoX gene probe. Columns 1, 2 and 3 contained 0.5, 1.5 and 3 µg total RNA, respectively. Approximately 6 ng mmoX gene probe was used as a positive control for hybridization. (b) mmoX amplified by PCR from genomic DNA using mmoX206F and mmoX886R primers, and digested with MscI (lane 1) and PvuII (lane 2); mmoX amplified by RT-PCR from cDNA using mmoX206F and mmoX886R primers, and digested with MscI (lane 3) and PvuII (lane 4). Lanes 5 and 6 are replicas of lanes 3 and 4. (Note: PvuII restriction site is unique to mmoX1, and MscI is unique to mmoX2).

The question of whether both copies of mmoX were translated was also addressed by analysing polypeptides of the α subunit of the hydroxylase by LC-ESI-MS/MS using three alternative methods. The first method analysed the most abundant peptides for which MS/MS experiments were performed. The second analysis was performed on peptides unique to mmoX2, and the third analysis was performed by instructing the software to ignore all peptides common to both copies of mmoX. The resulting MS/MS spectra were searched against an in-house database that included the duplicate copies of MmoX. The most abundant peptides identified were either unique to MmoX1 or common to both MmoXs, and therefore could not be differentiated. No peptides unique to MmoX2 could be identified.

Phenotypic characterization of ΔmmoX1 and ΔmmoX2 mutations
In order to definitively conclude whether the duplicate copies of mmoX were functional, and thus played a role in the formation of an active sMMO, knock-out mutants were constructed individually on both copies of mmoX by marker-exchange mutagenesis. This is believed to be the first report of successful construction of knock-out mutants in Ms. sporium by marker-exchange mutagenesis.

The genotypes of the Ms. sporium ΔmmoX1 and ΔmmoX2 strains were confirmed by PCR amplification of the region flanking the knock-out regions, using genomic DNA extracted from wild-type and mutant strains. The differences in length of the PCR products obtained from the wild-type and mutant strains confirmed the chromosomal deletion of the knock-out region of mmoX, and the insertion of the gentamicin cassette (data not shown). The ability of the mutant strains to oxidize methane using the sMMO enzyme was assessed using the naphthalene plate assay (Fig. 5a). The Ms. sporium ΔmmoX1 strain did not oxidize naphthalene, indicating loss of sMMO activity, whereas the Ms. sporium ΔmmoX2 strain retained wild-type sMMO activity and could oxidize naphthalene.

(73K): Fig. 5. sMMO expression and pigmentation profile of Ms. sporium ΔmmoX1 and ΔmmoX2 strains. (a) The sMMO activity of Ms. sporium ΔmmoX1 and ΔmmoX2 strains was analysed using the naphthalene plate assay. For each mutant strain, three transconjugants resulting from double homologous recombination events were tested, and Ms. sporium wild-type strain was included as a positive control. sMMO-positive colonies turned purple. (b) Pigmentation profiles of Ms. sporium wild-type and ΔmmoX1 and ΔmmoX2 strains. Ms. sporium strains grown on NMS medium containing excess FeEDTA (5 µM), with no added copper (sMMO-expressing condition) or excess copper (5 µM CuSO4; pMMO-expressing condition) are shown. The pigment-producing colonies appear brown. (c) Production of the water-soluble pigment in Ms. sporium : ΔmmoX1 strain following complementation of the sMMO phenotype by heterologous expression of sMMO operons from Ms. trichosporium (pVK100Sc) and Mc. capsulatus (pVK104) (compare with Fig. 5b). Six transconjugants from each conjugation were selected on NMS agar plates. Secretion of the water-soluble pigment caused discoloration of the agar plates, indicating complementation of the sMMO phenotype by heterologous expression of sMMO operons.

Ms. sporium forms a brownblack water-soluble pigment, and this is one of the main differences between the Methylosinus subgroups (Whittenbury et al., 1970; Anthony, 1982). Production of the pigment was notably apparent after growth on NMS agar plates for 1014 days at 30 °C, when the agar plate turned brown. The ΔmmoX1 strain, which had lost sMMO activity, also lost the ability to form the water-soluble pigment, whereas the ΔmmoX2 strain still produced the pigment (Fig. 5b). The production of the pigment seemed to be related to the expression of sMMO. However, it was noted by Whittenbury et al. (1970) that the pigmentation was only apparent on an iron-deficient medium. This was confirmed by growing Ms. sporium and the mutant strains on NMS plates containing excess or no added iron. Under excess iron conditions, none of the strains produced the water-soluble pigment, whereas under iron-limiting conditions, only the wild-type and ΔmmoX2 strains produced the water-soluble pigment (Fig. 5b).

Complementation and heterologous expression of sMMO in Ms. sporium : ΔmmoX1
The broad-host-range plasmids pVK100Sc and pVK104, containing the sMMO structural genes, native promoter and the regulatory genes from Ms. trichosporium and Mc. capsulatus, respectively (Martin, 1994; Lloyd et al., 1999), were conjugated into the Ms. sporium : ΔmmoX1 strain to demonstrate the complementation and expression of heterologous sMMO genes in Ms. sporium : ΔmmoX1. Six transconjugants from each conjugation were selected on replica NMS agar plates without added copper, and containing gentamicin (5 µg ml¹) and kanamycin (15 µg ml¹). Ms. sporium : ΔmmoX1 strains containing pVK100Sc and pVK104 all expressed active sMMO, as judged by the naphthalene plate assay and the ability to secrete the water-soluble pigment, causing discoloration of the agar plates. The intensity of the purple colour formed from the naphthalene plate assay, and the discoloration of the agar plates from the production of the water-soluble pigment, were low compared to that of the wild-type strain, thus indicating a low level of expression of the recombinant sMMO (Fig. 5c).

The duplication of pMMO genes in methanotrophs and ammonium monooxygenase genes in nitrifiers is well documented (Arp et al., 2002; Stolyar et al., 1999). A comprehensive phylogenetic analysis of sMMO and other soluble di-iron monooxygenases (reviewed by Leahy et al., 2003) suggests that these enzymes have been largely spread through horizontal gene transfer, and are not maintained permanently in a bacterial lineage. The identification of a second lone copy of mmoX in Ms. sporium, and the isolation of plasmids from this organism in previous studies (Lidstrom & Wopat, 1984), suggest that sMMO genes in Ms. sporium may have been acquired through horizontal gene transfer. Functional redundancy is common among bacterial genes, and in this case, the identification of a duplicate copy of mmoX may indicate a competitive advantage for Ms. sporium in adapting to local environments, such as peat-bogs, where copper is likely to be limiting for growth.

Disruption of mmoX1 resulted in an sMMO phenotype, whereas disruption of mmoX2 had no noticeable effect on sMMO expression. RT-PCR data suggested that the majority of the mmoX transcripts corresponded to mmoX1. In comparison, mmoX2 transcripts could be detected at relatively low levels. Nevertheless, both copies were transcribed. However, following analysis of trypsin- digested MmoX polypeptides by LC-ESI-MS/MS, no unique peptides corresponding to MmoX2 could be identified. It is possible that translation is not initiated from the SD sequence of mmoX2 due to the presence of an overlapping SD sequence, which may affect the efficiency of translation of the mRNA. Based on translational studies carried out on model organisms such as E. coli, it is known that the efficiency of translation of mRNA is affected by the SD sequence and the extent of complementarity between the SD sequence and 16S rRNA. The efficiency of translation is also affected by the distance between the SD sequence and the start codon (Schottel et al., 1984). We conclude that mmoX1 is crucial for sMMO activity, whereas mmoX2, which is transcribed only at low levels, is probably not translated; however, if it is, it does not play a significant role in sMMO activity.

Mutational analysis of mmoX gave insights into the expression of the water-soluble pigment produced by Ms. sporium. The pigment was only produced when sMMO was expressed or when iron bioavailability was low. Since sMMO is an iron-containing enzyme (Elango et al., 1997; Rosenzweig et al., 1993), and under these conditions, iron is rapidly used up, the pigmentation profile might reflect an iron-scavenging mechanism, which can be correlated with iron-acquisition systems mediated by siderophores (Schalk et al., 2004). In such systems, siderophores are secreted into the extracellular medium, where they bind iron, and are then subsequently transported back into the bacteria. In addition, an analogous system for copper acquisition in Mc. capsulatus has been reported, in which the copper chelator methanobactin is found to accumulate in the growth medium of Ms. trichosporium and Mc. capsulatus when grown under copper-limited conditions. However, it has been found to rapidly internalize when copper is provided (Kim et al., 2004). Therefore, it is possible that the water-soluble pigment in Ms. sporium is operating in a similar fashion, as a strategy to acquire iron under iron-limiting conditions. This conclusion is further supported by the observation that, under excess iron conditions, no pigment was produced. However, at this stage, no definitive conclusions can be drawn, and further work is required to correlate the water-soluble pigment with the siderophore-mediated iron-acquisition system in Ms. sporium.

The amenability of Ms. sporium to genetic manipulations was further demonstrated through the complementation of the sMMO mutant strain Ms. sporium : ΔmmoX1 with the broad-host-range plasmids pVK104 and pVK100Sc, which contain heterologous sMMO operons. All the respective transconjugants selected showed restoration of sMMO activity, although at much-reduced activity compared to that of the wild-type strain. Similar reduced activities have been observed with identical experiments on Methylomicrobium album BG8, and Methylocystis parvus OBBP and pVK100Sc containing the Ms. trichosporium OB3b sMMO gene cluster (Lloyd et al., 1999). We complemented the sMMO mutant with sMMO from both Ms. trichosporium OB3b and Mc. capsulatus (Bath). The relatively low-level heterologous expression of sMMO is probably due to the reasons discussed in detail by Lloyd et al. (1999). However, since the construction of the expression plasmids pVK104 and pVK100Sc, a number of broad-host-range plasmids have been used in methanotrophs for the heterologous expression of high-activity reporter genes (Csaki et al., 2003; Theisen et al., 2005). To further improve the heterologous expression of sMMO activity in methanotrophs, as well as in non-methanotrophic strains, new broad-host-range plasmids containing sMMO fragments need to be constructed and analysed, so that the full potential of the wide substrate specificity of sMMO can be exploited in biotransformation and bioremediation processes (reviewed by Smith & Dalton, 2004).

We gratefully acknowledge Hendrik Schaefer for assistance with the ARB software package (), Andreas Theisen for helpful comments and critical review of the manuscript, Sue Slade for MS analyses, and the Natural Environment Research Council for funding a studentship to H. A. and grant NER/A/S/2002/00876 to J. C. M.