CtaM Is Required for Menaquinol Oxidase aa3 Function in Staphylococcus aureus

ABSTRACT Staphylococcus aureus is the leading cause of skin and soft tissue infections, bacteremia, osteomyelitis, and endocarditis in the developed world. The ability of S. aureus to cause substantial disease in distinct host environments is supported by a flexible metabolism that allows this pathogen to overcome challenges unique to each host organ. One feature of staphylococcal metabolic flexibility is a branched aerobic respiratory chain composed of multiple terminal oxidases. Whereas previous biochemical and spectroscopic studies reported the presence of three different respiratory oxygen reductases (o type, bd type, and aa3 type), the genome contains genes encoding only two respiratory oxygen reductases, cydAB and qoxABCD. Previous investigation showed that cydAB and qoxABCD are required to colonize specific host organs, the murine heart and liver, respectively. This work seeks to clarify the relationship between the genetic studies showing the unique roles of the cydAB and qoxABCD in virulence and the respiratory reductases reported in the literature. We establish that QoxABCD is an aa3-type menaquinol oxidase but that this enzyme is promiscuous in that it can assemble as a bo3-type menaquinol oxidase. However, the bo3 form of QoxABCD restricts the carbon sources that can support the growth of S. aureus. In addition, QoxABCD function is supported by a previously uncharacterized protein, which we have named CtaM, that is conserved in aerobically respiring Firmicutes. In total, these studies establish the heme A biosynthesis pathway in S. aureus, determine that QoxABCD is a type aa3 menaquinol oxidase, and reveal CtaM as a new protein required for type aa3 menaquinol oxidase function in multiple bacterial genera.

the heme-copper oxygen reductase superfamily (13). Prior to the availability of the genome sequence of S. aureus, biochemical studies indicated that there were o-type and/or aa 3 -type terminal oxidases in the organism, but there is no mention of cytochrome bd (14)(15)(16)(17)(18). Typically, cytochrome bd can be identified spectroscopically due to the presence of heme D and by the resistance of the enzyme to cyanide (9). In fact, cytochrome bd cyanide resistance is a feature that distinguishes S. aureus from nonpathogenic staphylococci; S. aureus cytochrome bd is more sensitive to cyanide than the cytochrome bd expressed by the avirulent Staphylococcus carnosus (19). It was not until 2014 that the spectroscopic signature of S. aureus cytochrome bd was reported (20). Taken together, there is biochemical evidence for three types of respiratory oxygen reductases in S. aureus, a bd type, an o type, and/or an aa 3 type, despite genomic evidence for only cydAB and qoxABCD. Here, we address the origin of the three putative terminal oxidases and delineate the cyanide sensitivity of the S. aureus cytochrome bd. It is shown that the S. aureus cytochrome bd is resistant to cyanide below 1 mM and that the QoxABCD enzyme exists in two forms. One form contains two equivalents of heme A (aa 3 type), and the other incorporates hemes B and O (bo 3 type). Both forms are functional and sensitive to cyanide. We also identify a new gene in the S. aureus genome that is conserved in many Firmicutes and is required for the assembly of functional QoxABCD in any form. This gene (locus NWMN_0982), which we name ctaM, is adjacent to the gene encoding the heme O synthase, ctaB. These results define the molecular requirements for terminal oxidase function in S. aureus, provide an explanation for the previous observations describing the presence of an o-type reductase in the staphylococcal respiratory chain, and establish a conserved protein involved in terminal oxidase assembly.

RESULTS
QoxABCD is an aa 3 -type terminal oxidase and CydAB is a canonical cyanide-resistant cytochrome bd. The composition of the aerobic respiratory chain depends upon both the growth medium and the phase of growth at which the cells are harvested. Table 1 shows the comparison of the properties of membranes isolated from the wild type (WT) and different mutants of S. aureus, each grown in LB with 0.5% glucose and harvested after 8 h. The rate of O 2 consumption (per mg of membrane protein) in the presence of NADH was measured in the presence and absence of 100 M potassium cyanide (KCN). About 40% of the O 2 con-sumption is inhibited by 100 M KCN in the WT, whereas in the ⌬cydA mutant, 100% inhibition is observed. In contrast, in the ⌬qoxA mutant, 100% of the activity remains in the presence of 100 M KCN. Hence, the oxygen reductase encoded by cydAB is cyanide resistant and the oxygen reductase encoded by qoxABCD is cyanide sensitive. The reduced-minus-oxidized difference spectra of the membranes from both the WT and ⌬cydA strains (Fig. 1) clearly show the spectroscopic features of an aa 3 -type oxygen reductase (peaks at about 443 nm and 606 nm), which are absent from the spectrum of the ⌬qoxABCD strain. The spectroscopic signature typical of heme D, a peak in the reduced-minusoxidized spectrum near 630 nm, is not evident in any of the spectra in Fig. 1. This feature was previously observed with membranes from the WT strain, as shown in Fig. S4 in reference 20, so its absence is likely due to its relatively small amount and the noise in the spectra shown in Fig. 1. These results establish that qoxABCD encodes a cyanide-sensitive, a-type reductase.
QoxABCD assembles as a functional oxidase in the absence of heme A. It is expected that the aa 3 -type menaquinol oxidase encoded by qoxABCD incorporates two equivalents of heme A. The conversion of heme B to heme A is catalyzed by CtaB and CtaA, which have been characterized from other bacteria, including Bacillus subtilis (21,22). The reaction sequence is shown in Fig. 2A, showing heme B converted to heme O by CtaB and then to heme A via CtaA. Both the ctaB and ctaA genes are present in S. aureus and, in addition, there is a third gene adjacent to ctaB which we have named ctaM (locus NWMN_0982), whose function is not known (Fig. 2B).
A mutant with a deletion of ctaB was constructed, and the ⌬ctaB strain behaves like the ⌬qoxA strain insofar as the remaining respiratory activity is entirely resistant to the presence of 100 M KCN (Table 1). Figure 3A shows a liquid chromatography-mass spectrometry (LC-MS) analysis of the hemes extracted from the membranes of the WT strain, excluding heme D, which is not observed under these conditions. The major peak is due to heme B, expected to be a component of cytochrome bd, as well as succinate dehydrogenase. Two smaller peaks elute at positions indicating that they represent heme O and heme A. This is consistent with the analysis of the hemes present in the membranes of the ⌬ctaB strain, in which these two small peaks are absent (Fig. 3B). In addition, the double ⌬ctaB ⌬cydA mutant exhibits the SCV phenotype (Fig. 3C), indicating that this strain is respiration deficient. These results demonstrate that the synthesis of heme O is essential for the assembly of functional QoxABCD.
Unexpectedly, the ⌬ctaA strain, which also cannot synthesize  (Fig. 4B). Previous studies on Bacillus cereus (23,24) showed that preventing the synthesis of heme A results in the assembly of the aa 3 -type menaquinol oxidase into a bo 3 -type menaquinol oxidase. Our observations indicate that S. aureus responds in a similar manner. QoxABCD utilizes heme O at the expense of carbon restriction. The ability of B. subtilis to grow in different media is impacted by mutations that eliminate one or more of the respiratory oxygen reductases (25). To determine whether the inability to synthesize heme A leads to any growth phenotype in S. aureus, WT, ⌬qoxA, and ⌬ctaA strains were grown in a minimal medium supplemented with glucose or Casamino Acids (CAS) as the carbon source (Fig. 4C). When glucose is supplied as a carbon source, the WT, ⌬ctaA, and ⌬qoxA strains grow equally well (Fig. 4C). However, when the cells are cultured in a minimal medium supplemented with CAS, no growth of the ⌬qoxA strain is observed after 24 h and the growth of the ⌬ctaA strain exhibits a lag of at least 12 h before growth is observed (Fig. 4C).
CtaM is required for QoxABCD function. The gene adjacent to ctaB, NWMN_0982 (locus tag in S. aureus strain Newman), encodes a putative membrane protein of unknown function (Fig. 2B). The location adjacent to ctaB suggests that NWMN_0982 also facilitates QoxABCD function and that these genes are likely cotranscribed. To verify this, primers were designed to amplify between the coding regions of ctaB and NWMN_0982. Amplification between ctaB and NWMN_0982 was observed when cDNA was prepared from S. aureus grown to stationary phase and was dependent on reverse transcriptase (Fig. 5A). The finding that ctaB and NWMN_0982 are cotranscribed demonstrates that these genes reside in an operon and supports the hypothesis that these genes encode proteins that func-  tion in the same pathway. Consistent with this, we refer to NWMN_0982 as ctaM (cytochrome assembly). To test the hypothesis that ctaM supports QoxABCD function, ctaM was inactivated in the ⌬cydAB strain. The ⌬cydAB ⌬ctaM mutant displays the SCV phenotype (Fig. 5B), and a plasmid-encoded copy of ctaM complements the respiration-arrested growth of the ⌬cydAB ⌬ctaM strain (see Fig. S2B in the supplemental material). These results demonstrate that CtaM is required for the assembly of functional QoxABCD. Accordingly, the ⌬qoxB ⌬ctaM double mutant does not exhibit the SCV phenotype, indicating that CtaM is not required for CydAB activity (Fig. 5B). Consistent with this, the respiratory activity of the ⌬ctaM strain is resistant to KCN, revealing that CydAB is the only terminal oxidase active in this mutant (Table 1). Furthermore, the ⌬ctaM strain behaves like the ⌬qoxA strain in that neither strain exhibits growth over a 24-h period in liquid broth consisting of a minimal medium supplemented with CAS (Fig. 4C). The hemes extracted from membranes prepared from the ⌬ctaM strain are similar to the hemes isolated from the WT, showing a predominant LC peak due to heme B and small peaks that appear to correspond to heme O and heme A (Fig. 5C). These results suggest that CtaM is not required for the function of CtaA or CtaB but, rather, is needed either for the insertion of the hemes into the enzyme or for the activity of the final, assembled enzyme.
ctaM is conserved in respiring Firmicutes that synthesize heme O and heme A. These results establish that ctaM is required for QoxABCD terminal oxidase function and imply that ctaM is conserved in other species of bacteria that utilize QoxABCD to respire. Phylogenic investigation in the SEED database (26,27) determined that ctaM is conserved in many diverse species of bacteria, as the analysis of a set of 978 nonredundant representative prokaryotic genomes reveals a ctaM homologue in 120 organisms (12% of the genomes analyzed). The results of this analysis are illustrated in Fig. 6 and are available in full online in the SEED Subsystem "Heme O and Heme A biosynthesis (with selected terminal oxidases)" (28).
The presence of ctaM in the Archaea, Firmicutes, Aquificae, and Planctomycetes, some of the oldest evolutionary taxa, supports the conclusion that CtaM is an ancient protein. Consistent with this, ctaM is not present in Beta-or Gammaproteobacteria (the youngest taxa) but does occur in a few representatives of the Alpha-and Delta-/Epsilonproteobacteria (the older subgroups of Proteobacte-  ria) (29). Notably, within each taxon, CtaM proteins are found only in genomes encoding aa 3 -type terminal oxidases and never in nonrespiring organisms or organisms utilizing other types of terminal oxidases (28). In particular, within the Firmicutes, ctaM is conserved among those that respire but not in obligate anaerobes like the Clostridiales or aerotolerant nonrespiring bacteria like the lactic acid bacteria (LAB). In some cases, when grown in medium supplemented with the proper cofactors, LAB respire by utilizing bd-type oxidase (CydAB) (30). This finding supports the idea that CtaM specifically interacts with QoxABCD. Additional support for this notion is provided by the fact that in all of the organisms in our representative set that encode qoxABCD (31 out of 978), ctaM is also present without exception. A similar tendency for cooccurrence in the same genome is observed between ctaM and the genes encoding cytochrome c oxidase subunits CoxI, -II, -III, and -IV and the alternative cytochrome c oxidase CoxMNOP, both     Th er m op la sm a ac id op hi lu m M e th a n o p y ru s k a n d le ri  (29,45). The distributions of ctaB, the gene that encodes the heme O synthase (purple, outer ring), and of ctaA, the gene that encodes the heme A synthase (teal, outer ring), are also presented. aa 3 -type enzymes (10,13). Furthermore, this functional association inferred by co-occurrence is strengthened by colocalization of the corresponding genes in many genomes (see Subsystem spreadsheet at reference 28). On the other hand, the ctaM homologues show no tendency to co-occur or to colocalize with genes that encode bo 3 -type ubiquinol oxidase (cyoABCD), bdtype ubiquinol oxidase (cydABCD), cytochrome bc 1 complexes (mcb, ucb, or cbsAB-soxLN), or cbb 3 -type cytochrome c oxidase (ccoPONQ) (28).

M e th a n o c o c c u s m a ri p a lu d is M e th a n o c o c c u s ja n n a s c h ii P y r o c o c c u s h o r io s h ii P y r o c o c c u s a b y s s i P y r o c o c c u s fu r io s u s
The fact that only a-type oxidases, such as qoxABCD and coxMNOP, are linked with ctaM homologues by strong comparative genomics evidence implies a functional association of the CtaM protein family with heme A synthesis. Consistent with this, the co-occurrence of ctaM with ctaB (heme O synthase) has no exceptions (Fig. 6). However, 4 of the 120 genomes (3%) that contain ctaM do not have a ctaA (heme A synthase) homologue. These four exceptions are all Archaea, organisms in which heme A synthesis is not well understood (31,32), and therefore do not contradict the detected functional association of the CtaM family with enzymes of heme O and A biosynthesis. In addition to a nearly absolute co-occurrence between ctaM and ctaA and/or ctaB, the ctaM genes show a strong tendency to colocalize with ctaB and/or ctaA genes within bacterial chromosomes: this is true in 79 of 120 genomes (66%) that contain ctaM within the set of genomes analyzed. In total, the phylogenic profile, colocalization on the chromosome, and functional analysis of the ctaM family clearly demonstrate that the CtaM protein functions to support the activity of a-type terminal oxidases in many species.

DISCUSSION
In order to survive and proliferate in the presence of the host immune response, bacterial pathogens fine-tune their metabolism for maximal energy production and growth. The utilization of multiple terminal oxidases is one mechanism that allows pathogens to optimize metabolism in the host environment (3,33,34). These branched aerobic respiratory chains allow for rapid adaption to the availability of oxygen and changes in available carbon sources. The availability of oxygen is likely a major environmental cue for S. aureus as it transitions from a commensal organism colonizing the skin or nasopharynx to a systemic pathogen proliferating within tissue abscesses in different host organs. Certainly, the amount of oxygen on the skin and within the nose is dramatically different than in an abscess, which has recently been demonstrated to be a hypoxic environment (35). Utilizing more than one terminal oxidase allows S. aureus to generate the maximum amount of energy to proliferate in distinct host environments, a conclusion supported by the finding that S. aureus mutants restricted to a single terminal oxidase are severely impaired for colonization of the murine heart or liver (3,4). The findings presented herein conclusively demonstrate that S. aureus utilizes two terminal oxidases to respire aerobically, QoxABCD and CydAB.
The S. aureus CydAB is resistant to micromolar concentrations (100 M) of cyanide, distinguishing it from the cyanide-sensitive cytochrome aa 3 encoded by qoxABCD. This profile of cyanide resistance is typical of canonical representatives of the cytochrome bd and cytochrome aa 3 family of oxidases. However, elegant work by Voggu et al. demonstrated that nonpathogenic S. carnosus can grow in the presence of 1.5 mM cyanide, whereas S. aureus cannot (19). This is likely due to a greater sensitivity to cyanide of the cytochrome bd from S. aureus. Given the importance of CydAB to S. aureus virulence, a particularly intriguing notion raised from these findings is that the cyanide sensitivity of S. aureus, compared to that of S. carnosus, is an expense paid to enhance fitness during colonization of the human host.
QoxABCD is promiscuous regarding its heme cofactors and can be assembled either as an aa 3 -type oxidase using two equivalents of heme A or as a bo 3 -type oxidase using one equivalent each of heme B and heme O. In the current work, QoxABCD was constrained to assemble as the bo 3 -type enzyme by genetically eliminating heme A biosynthesis (⌬ctaA mutation), but a number of previous studies identified an o-type oxidase in the WT strain (14)(15)(16)(17)(18). Such promiscuity could possibly be an advantage, since the CtaA enzyme requires O 2 to synthesize heme A (36). This might provide an advantage to S. aureus when shifting from an anoxic to oxic environment.
However, although QoxABCD is assembled as a functional enzyme in the ⌬ctaA strain, the mutation is not benign, indicating some significant differences either in the properties of the assembled bo 3 and aa 3 variants of QoxABCD or in the efficiency of assembly or stability of the different forms of the enzyme. It is also possible that the inability to convert heme O to heme A reflects a different function for heme A, though none are known, or toxicity of heme O. The inability of the ⌬ctaA strain (no heme A) to grow rapidly when CAS is substituted for glucose suggests that there is no functional respiratory oxidase present, and in the absence of a fermentable substrate (glucose), aerobic respiration is essential. Presumably, the CydAB oxygen reductase is not expressed under the conditions employed or it would be sufficient to support aerobic growth by respiration. It appears to take at least 12 h for the ⌬ctaA strain to assemble a sufficient amount of QoxABCD to permit growth. Understanding the drastic difference in the growth phenotype will take more study, including the control of gene expression of the various respiratory components.
The elimination of the synthesis of heme A in the ⌬ctaA strain also has consequences during infection, most prominently that this strain cannot colonize the murine liver (4). The reason for this is not clear but may reflect different concentrations of O 2 required for the synthesis of heme A and for aerobic respiration. The growth nutrients present, as indicated by the contrast between glucose and CAS in vitro, may also play a role. This study validates further investigation of the requirements for QoxABCD and CydAB function as an approach to define the environmental conditions encountered by S. aureus during infection.
A key finding of the current work is that the assembly of a functional QoxABCD oxygen reductase requires CtaM. Our data suggest that CtaM may act as a chaperone for the insertion of heme O or heme A into QoxABCD. In support of this prediction, ctaM is highly conserved in respiring Firmicutes that also encode qoxABCD (Fig. 6). Additionally, heme O and heme A are still synthesized in the ctaM mutant (Fig. 5C), but QoxABCD is not functional (Fig. 5). This finding suggests that the hemes are not trafficked to the QoxABCD complex. However, it is currently unknown whether the hemes in the ctaM mutant reside within an inactive form of the QoxABCD complex or are restrained in CtaB or CtaA. The latter possibility would provide additional support for the hypothesis that CtaM acts as a heme chaperone and is required to traffic heme A or heme O to QoxABCD. The fact that B. subtilis ctaB and ctaA can be ectopically expressed and are able to synthesize heme O and heme A in Escherichia coli, an organism that does not express a QoxABCD oxidase, supports the possibil-ity that the heme O and heme A are stable in the absence of an active QoxABCD complex (22). Alternatively, the presence of heme O and heme A may indicate that the function of CtaM is to assemble QoxABCD independent of heme. Further study is needed to determine the precise mechanism by which CtaM supports the assembly of QoxABCD.

MATERIALS AND METHODS
Bacterial strains and growth conditions. The strains used in this study are described in Table 2. All strains are derivatives of the human clinical isolate S. aureus Newman or the human clinical isolate USA300 LAC JE2 adapted for laboratory use (37,38). The ⌬cydAB and ⌬ctaB isogenic inframe deletion mutants were constructed using previously described methods (39). The cydA, qoxA, ctaA, and ctaM transposon mutants were obtained from the Nebraska transposon library (NE117, NE92, NE769, and NE1084, respectively) (38). PCR confirmed that the transposon insertion mapped to the respective gene. Backcross strains of the transposon-disrupted cydA, qoxA, ctaA, and ctaM alleles into WT S. aureus Newman or USA300 LAC JE2 were produced via phage 85mediated transduction (38). The backcrossed strains were used in the studies described here. Phage 85-mediated transduction of the cydA and qoxA transposon-disrupted alleles was used to create the ⌬ctaB ⌬cydA and ⌬ctaB ⌬qoxA double mutants. Likewise, phage 85-mediated transduction of the ctaA and ctaM transposon-disrupted allele was also used to construct the ⌬cydAB ⌬ctaA, ⌬cydAB ⌬ctaM, ⌬qoxB ⌬ctaA, and ⌬qoxB ⌬ctaM double mutants. For complementation studies, the ctaB gene was PCR amplified and cloned into the pOS1 P lgt plasmid using primers 5= GCGCTCGAGATGAGCAAAGAGCATACTTTGTC 3= and 5= GCGGGA TCCCTAGATCAAAGTAAGTAATGAAAC 3= and the restriction enzymes XhoI and BamHI. BamHI and XhoI were also used to clone the ctaM gene into pOS1 P lgt . ctaM was PCR amplified using primers 5= GC GCTCGAGATGGGCGTTCCAATTTTACCA 3= and 5= GCGGGATCCT TAATGACCAAATGTTGCTTTAAT 3=. Antibiotic selection of erythromycin cassette-containing resistant recipient cells was achieved with 10 g ml Ϫ1 erythromycin. Antibiotic selection of strains containing the pOS P lgt plasmid was achieved by using 10 g ml Ϫ1 of chloramphenicol. Bacteria were routinely grown in tryptic soy broth (TSB) at 37°C. All strains were diluted 1:100 from overnight cultures into fresh TSBcontaining 96-well round-bottom plates, and growth was measured by optical density at 600 nm. The PN minimal medium was prepared as previously described (38). The cells were grown overnight in TSB, washed thrice, and resuspended in PN minimal medium lacking a carbon source. Glucose and CAS were supplemented at the indicated concentrations.
Membrane preparations. Cells were grown in LB plus 0.5% glucose for 8 h. Membranes were obtained by incubating cells with lysostaphin for 2 h at 37°C and then passing them more than five times through a microfluidizer at a pressure of 80,000 lb/in 2 . The cell extract was centrifuged at 14,000 ϫ g for 10 min to remove the unbroken cells. Membranes were obtained by centrifugation at 230,000 ϫ g for 4 h. Pellets, containing 30 to 40 mg/ml of protein, were resuspended in sodium phosphate buffer, pH 7.5, 50 mM NaCl, 10% glycerol, and the suspension was stored at Ϫ80°C. The NADH-reduced minus air-oxidized difference spectrum of S. aureus membranes was obtained at room temperature using an Agilent Technologies spectrophotometer (model 8453). Reduction was achieved by the addition of 5 mM NADH.
Heme extraction and identification. The hemes from S. aureus membranes were extracted and analyzed using high-performance liquid chromatography (HPLC) elution profiles according to established protocols (40,41). One milliliter of the membrane suspension was solubilized by the addition of a stock solution of 20% DDM (dodecyl-␤-D-maltoside) dropwise to a final concentration of 1%. The solution was incubated at 4°C for 2 h with mild agitation. The suspension was cleared by centrifugation at 230,000 ϫ g for 1 h. The supernatant was mixed with 9 ml of acetone-HCl (19:1, vol/vol) and shaken on a rotary shaker for 20 min at room temperature. The mixture was centrifuged at 14,000 ϫ g for 2 min, followed by the addition of 20 ml of ice-cold water and 6 ml of ethyl acetate to the supernatant. The water-ethyl acetate mixture was vortexed and centrifuged again for 10 min at 4°C. The ethyl acetate phase was recovered, and the solvent evaporated by flushing it with nitrogen gas. The residues were dissolved in 0.05 to 0.10 ml of acetonitrile and stored at Ϫ20°C. The extracted hemes were analyzed by HPLC using a Waters 2795 separation module and Waters 2996 photodiode array (PDA) detector. For heme separation, a 1-mm-inner-diameter reverse-phase C 18 column was used with an acetonitrile (0.05% trifluoroacetic acid [TFA])-water (0.05% TFA) gradient from 50 to 100% (41). The hemes were confirmed by their molecular weights using a Waters quadrupole time of flight (Q-TOF) Ultima electrospray ionization (ESI) mass spectrometer.
Oxygen consumption. The oxygen concentration was monitored using a dual-channel respirometer system, model 782 from Strathkelvin Instruments, equipped with a temperature-controlled 0.5-ml electrode chamber at 37°C. The reaction mixture for this assay consisted of sodium phosphate buffer, pH 7.5, 50 mM NaCl, and 200 to 1,000 g/ml membranes. The concentration of oxygen in the air-saturated buffer at this temperature was assumed to be 237.5 M, and the reaction was initiated by injecting 500 M NADH. One hundred micromoles of potassium cyanide (KCN) was added to distinguish between the cyanide-sensitive aa 3 menaquinol oxidase and the cyanide-resistant bd menaquinol oxidase present in S. aureus membranes. The percentage of inhibition after the addition of KCN was calculated for each mutant.
RNA purification and ctaB-ctaM intergenic amplification. Bacterial RNA was harvested as previously described (42). Briefly, overnight cultures of S. aureus were subcultured (1:100) into TSB and cultures were harvested at late exponential phase (6 h) or stationary phase (12 h). The cells were pelleted, resuspended in LETS buffer (1 M LiCl, 0.5 M EDTA, 1 M Tris HCl, pH 7.4, 10% SDS), and lysed using mechanical bead beating. RNA was extracted using TRIzol (TRI Reagent; Sigma), chloroform precipitated with isopropyl alcohol, washed with 70% ethanol, and dissolved in distilled water. Contaminating DNA was removed with DNase I (Amersham Biosciences) treatment. RNA then was purified using the RNeasy minikit according to the manufacturer's protocol (Qiagen). The RNA concentration and purity were measured by the optical density at 260 nm and 280 nm, respectively. For cDNA synthesis, 2 g of RNA was left untreated (without reverse transcriptase [RT]) or was treated with Moloney murine leukemia virus reverse transcriptase according to the manufacturer's recommendations (Promega). The region between ctaB and ctaM was PCR amplified using the following standard cycling param- Bioinformatic analysis of ctaM. With over 60,000 prokaryotic genomes currently available in public databases and over 15,000 more in the pipelines (www.genomesonline.org), it is not practical or possible to perform meaningful comparative analysis on all of them simultaneously. Thus, a set of diverse representative prokaryotic genomes has been developed in the SEED database as follows. The algorithm for computing molecular operational taxonomic units (OTUs) based on DNA bar code data (43,44) was used to group~12,600 prokaryotic genomes available in the SEED database in October 2013 into about 1,000 taxon groups. A representative genome for each OTU was selected based on the largest amount of published experimental data and the highest level of research interest within the scientific community for different microorganisms within each OTU. The resultant collection of 978 diverse eubacterial (924) and archaeal (54) genomes creates a manageable set that accurately represents the immense diversity of the over 12,000 prokaryotic organisms with sequenced genomes. Importantly, it is not skewed by an overabundance of genomes for a few microbial genera (medically or industrially important), such as Enterobacteriaceae, staphylococci, mycobacteria, etc.
The hypothetical DUF420 family (CtaM) has been exhaustively annotated for this set of 978 representative microbial genomes (54 archaeal and 928 eubacterial) in the SEED database (26). Functional associations for this gene family were predicted based on the patterns of co-occurrence and/or colocalization of its members with other protein families using the set of tools for comparative genome analysis available in SEED (27) within the functional and genomic contexts provided by the subsystem "Heme O and Heme A biosynthesis (with selected terminal oxidases)" (28).