ABSTRACT
Gene clusters coding for the chaperone/usher (CU) pathway are widely distributed in many important environmental and pathogenic microbes; however, information about the regulatory machineries controlling CU gene expression during multicellular morphogenesis is missing. The Myxococcus xanthus Mcu system, encoded by the mcuABCD gene cluster, represents a prototype of the archaic CU family that functions in spore coat formation. Using genome-wide transposon mutagenesis, we identified MXAN2872 to be a potential regulator of the mcuABC operon and demonstrated the necessity of MXAN2872 for mcuABC expression and fruiting body morphogenesis in early development. In silico, biochemical, and genetic analyses suggest that MXAN2872 encodes a Baeyer-Villiger monooxygenase (BVMO) of flavoproteins, and the potential cofactor-binding site as well as the BVMO fingerprint sequence is important for the regulatory role of the MXAN2872 protein. The expression profile of mcuA in strains with an MXAN2872 deletion and point mutation agrees well with the timing of cell aggregation of these mutants. Furthermore, McuA could not be detected either in a fruA-null mutant, where starvation-induced aggregation was completely blocked, or in the glycerol-induced spores, where sporulation was uncoupled from cell aggregation. In sum, the present work uncovers a positive role for MXAN2872, a metabolic enzyme-encoding gene, in controlling M. xanthus development. MXAN2872 functions by affecting the onset of cell aggregation, thereby leading to a secondary effect on the timing of mcuABC expression of this model organism.
IMPORTANCE Identification of the players that drive Myxococcus xanthus fruiting body formation is necessary for studying the mechanism of multicellular morphogenesis in this model organism. This study identifies MXAN2872, a gene encoding a putative flavin adenine dinucleotide-binding monooxygenase, to be a new interesting regulator regulating the timing of developmental aggregation. In addition, MXAN2872 seems to affect the expression of the chaperone/usher gene cluster mcu in a manner that is aggregation dependent. Thus, in organisms characterized by a developmental cycle, expression of the chaperone/usher pathway can be controlled by morphological checkpoints, demonstrating another layer of complexity in the regulation of this conserved protein secretion pathway.
INTRODUCTION
To survive in hostile environments, bacterial cells possess a fundamental ability to sense and respond to changes in their environment. The model organism Myxococcus xanthus, a Gram-negative bacterium characterized by its social behavior, displays two different cellular patterns depending on its nutritional status. In the presence of nutrients, cells grow and organize into spreading colonies; in the absence of nutrients, cells aggregate to form multicellular fruiting bodies inside which the rod-shaped cells differentiate into dormant spherical spores that are stress resistant (1, 2). For the purpose of initiating fruiting body development, M. xanthus needs to sense its nutritional status. It has been proposed that this process is accomplished by measuring the intracellular level of the guanosine nucleotide (p)ppGpp, a general starvation signal that is both necessary and sufficient to activate the developmental program (3). Synthesis of (p)ppGpp is catalyzed by the enzyme named RelA, and other M. xanthus genes, such as nsd and socE, can influence the accumulation of this starvation signal (4, 5). In addition, like the nsd or socE mutant, cells carrying a mutation in the bcsA gene can develop on medium containing nutrient levels high enough to block development in wild-type cells (6), suggesting that BcsA, which is homologous to flavin-binding monooxygenases, might also play a role in sensing or responding to nutritional deprivation, although whether this protein influences the RelA-dependent stringent response remains to be explored. It is likely that these genes that have been identified are only a few of many genes implicated in nutrient sensing and the regulation of entry into development. Identification of other genes that modulate the timing of entry into development is important for understanding how M. xanthus responds to environmental stimuli.
Starvation-induced sporulation normally occurs when the aggregation process is complete, suggesting that M. xanthus is capable of monitoring progress toward aggregation prior to the initiation of sporulation. The emerging regulatory machinery that coordinates aggregation and sporulation appears to be the C-signaling pathway. Distinct thresholds of the C-signal protein, combined with an ordered increase in the level of C signaling, allow the C signal to act as a timer that induces, first, aggregation and early C-signal-dependent gene expression and then sporulation and late C-signal-dependent gene expression (7).
M. xanthus uses a large number of proteins to construct a spore (8–14). In this bacterium, the Mcu system, encoded by the mcuABCD gene cluster (i.e., MXAN3885 to MXAN3882), represents a prototype of the archaic chaperone/usher (CU) pathway that functions in spore coat formation (15, 16). CU pathways are used for the assembly and secretion of a large family of adhesive protein polymers termed pili or fimbriae at the cell surfaces of Gram-negative bacteria (17, 18), and genes involved in the biosynthesis of CU pili are often organized into operons. These operons encode, at a minimum, one major structural pilus subunit and two nonstructural assembly components: a specialized periplasmic chaperone and an outer membrane protein called the usher. The mcu locus is the only CU gene cluster in M. xanthus, with mcuABC being expressed as a single transcriptional unit (16). Both McuA and McuD contain a spore coat protein U (SCPU) domain, determined on the basis of in silico analysis, and the presence of McuA on spore surfaces has been experimentally demonstrated (16). While mcuD and mcuA are not cotranscribed, McuD is necessary for the surface localization of McuA, implying that these two proteins may interact with each other and are likely to be exported together through the putative outer membrane usher McuC (19). McuB acts as a molecular chaperone to stabilize McuA by forming a complex with it (19). Compared with classical CU systems, although the M. xanthus Mcu system exhibits structural variation in mediating the chaperone-subunit interaction, it is a bona fide CU pathway involved in spore coat formation (19, 20). This is a newly recognized function for CU pathways, which in most cases are responsible for pilus biogenesis.
The mcuABC operon is not expressed until 12 to 15 h after the initiation of M. xanthus development. The mechanism underlying the developmentally regulated expression of mcuABC is not clear. Studies of different CU gene clusters by numerous groups have demonstrated that the regulation of CU pilus gene expression involves complex regulatory circuits with multiple components required for optimal activity (21). However, current knowledge of the control of CU fimbrial gene expression is derived solely from studies on the CU systems of bacteria that do not display a complex developmental cycle, such as the CupABCDE systems of Pseudomonas aeruginosa (22), the Csu system of Acinetobacter baumannii (23), and the Coo system of enterotoxigenic Escherichia coli (24). To better understand the molecular processes controlling CU pathways during multicellular morphogenesis, we used random mutagenesis techniques to identify potential regulators of the mcuABC locus and demonstrate here that MXAN2872, a gene encoding a putative flavin adenine dinucleotide (FAD)-binding monooxygenase, positively modulates the onset of M. xanthus development, which might in turn serve as a gate for the proper timing of mcuABC expression in early development.
MATERIALS AND METHODS
Cell growth and development and measurement of β-galactosidase activity.Escherichia coli strains were grown in LB medium in the presence of relevant antibiotics. M. xanthus strains were grown in CTT liquid medium (1% Casitone, 8 mM MgSO4, 10 mM Tris-HCl, 1 mM potassium phosphate, pH 7.6) and CTT agar (1.5%) plates. Kanamycin or oxytetracycline was used for selection at concentrations of 40 μg ml−1 and 12.5 μg ml−1, respectively. M. xanthus fruiting body development was induced on TPM (10 mM Tris-HCl, 1 mM KH2PO4, 8 mM MgSO4, pH 7.6) with 1.5% Difco Bacto agar. Specific β-galactosidase activities were determined by the protocol of Kroos et al. (25). Protein concentrations were measured by a bicinchoninic acid protein assay (Pierce).
M. xanthus strains.The M. xanthus strains used in this study are listed in Table 1. DK1622 (26) was used as the parent wild-type (wt) strain for all M. xanthus strains constructed in this study. All strains constructed were confirmed by PCR.
Plasmids and strains
Transposon mutagenesis.The plasmid p15A-MycoMar-oriT-R6K-hyg-neo-lacZ, which carries the MycoMar transposable element of the mariner family (27), was digested with MluI to remove a 1.2-kb central region of the lacZ open reading frame (ORF) and then self-ligated to generate p15A-MycoMar-oriT-R6K-hyg-neo. This plasmid was transformed by electroporation into mcuABC-lacZ fusion strain DK1622/pMP-mcuABC as described previously (28). Electroporation mixes were plated on CTT agar containing kanamycin and incubated at 30°C for 5 to 7 days. The resultant colonies were transferred to TPM starvation agar containing 20 μg ml−1 X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) and screened visually after 2 days of incubation to identify colonies displaying no blue color. Colonies thus identified were checked again on TPM agar containing X-Gal to confirm the lack of a blue phenotype.
Transposon insertion location.To determine the target disrupted by the MycoMar transposon, chromosomal DNA was isolated from M. xanthus mutants carrying the MycoMar insertion and digested with ApaLI. The digest was heat inactivated and then treated with T4 DNA ligase at 15°C overnight. Subsequently, the ligation mixture was incubated at 65°C for 15 min and dialyzed on a 0.025-μm-pore-size filter (Millipore) against distilled water for 30 min (drop dialysis) before electroporation into E. coli CC118 λpir. Plasmids were prepared from kanamycin-resistant colonies, and subcloned ApaLI fragments were sequenced with primers corresponding to the sequences of the ends of magellan4 (28).
Construction of the MXAN2872-lacZ fusion.The MXAN2872-lacZ transcriptional fusion was constructed using the vector pMP220 (29). A 0.9-kb DNA fragment containing part of the MXAN2872 5′ coding sequence and its putative promoter region was generated by PCR using DK1622 genomic DNA as the template. The PCR primer pair was MXAN2872-5 and MXAN2872-3, spanning nucleotides between 828 bp upstream and 59 bp downstream from the MXAN2872 start codon. The PCR product obtained was digested with XbaI and PstI, cloned into the corresponding sites in the pBlueScript vector, and verified by DNA sequencing. Then, the plasmid-derived XbaI-PstI fragment, together with a 2.9-kb EcoRI-XbaI myxophage Mx8 attP fragment recovered from a pUC18 derivative carrying Mx8 attP, was ligated with pMP220 that had been digested with EcoRI and PstI, yielding a transcriptional lacZ fusion vector, pMP-MXAN2872. The plasmid was then introduced into M. xanthus through electroporation (30). Transformants were selected by plating the cells onto CTT agar plates containing oxytetracycline. The plasmid used in this study could not independently replicate out of the M. xanthus chromosome. Thus, all antibiotic-resistant transformants resulted from integration of the plasmid into the chromosome either by homologous recombination or by site-specific recombination between attP on the plasmid and the attB site on the chromosome. PCR was performed to screen antibiotic-resistant colonies for proper integration of each plasmid at attB using the following primer pair: forward primer 5′-GAAGGGCCCGGAACCTTGCGATTCCGGG-3′ and reverse primer 5′-CCATGAGCGGGAAGCGGTCGGGGAGCGT-3′.
Immunoblot analysis.The accumulation of McuA in M. xanthus strains was analyzed by immunoblotting. Starvation-induced developmental cells were harvested at the times indicated below, and protein samples were prepared as previously described (19). For each sample, a 5-μl aliquot representative of an equal number of cells at the beginning of development was applied to each lane. The blots were probed with rabbit anti-McuA serum, followed by anti-rabbit IgG conjugated to alkaline phosphatase with chromogenic substrates.
Analysis of glycerol-induced spores.M. xanthus DK1622 cells were cultivated in CTT medium and induced at an optical density at 600 nm (OD600) of 0.8 with glycerol to a final concentration of 0.5 M. At 2, 4, and 8 h after addition of the inducer, 5 ml of cells was harvested, and the cells were pelleted by centrifugation at 5,000 × g for 5 min, resuspended in 100 μl of 1% SDS, and then boiled for 5 min to give soluble (S100) and pellet (P100) fractions. Note that McuA can be released from starvation-induced mature spores by boiling in 1% SDS. To test whether McuA accumulates inside cells of the P100 fraction, unbroken cells were disintegrated further by sonication in the presence of glass beads. The resulting cell lysate was mixed with an equal amount of 2× SDS loading buffer and analyzed by immunoblotting as described above.
Construction of MXAN2872, MXAN2873, and fruA deletion mutants.The Plasmid pBJ113-ΔMXAN2872 is a pBJ113 derivative (31) generated to create a deletion mutation of the MXAN2872 gene. To construct pBJ113-ΔMXAN2872, a 1.2-kb upstream fragment and a 1-kb downstream fragment flanking MXAN2872 were PCR amplified using the primer pairs MXAN2872 H1HindIII/MXAN2872 H1BamHI and MXAN2872 H2BamHI/MXAN2872 H2EcoRI, respectively (Table 2). The two amplified products were digested, ligated, and cloned into pBJ113 to obtain the plasmid pBJ113-ΔMXAN2872. After being verified by DNA sequencing, the plasmid was introduced into M. xanthus DK1622 by electroporation, and transformants were selected on CTT agar plates containing 40 μg ml−1 kanamycin. Individual Kmr transformants were then grown in CTT broth in the absence of kanamycin and plated onto CTT agar plates supplemented with 1% galactose for negative selection. PCRs were used to screen Galr and Kms colonies for proper excision of the wt copy, as described previously (32). The resulting ΔMXAN2872 strain contained a 1,809-bp in-frame deletion corresponding to codons 2 to 604 in the 604-amino-acid MXAN2872 protein. MXAN2873 and fruA deletion mutants were constructed in a similar manner using their respective primers. The ΔMXAN2873 strain contains a 1,648-bp deletion within the 1,662-bp MXAN2873 ORF, while the ΔfruA strain contains a 662-bp deletion within the 690-bp fruA ORF.
Primers used in this study
Construction of plasmids expressing wt and mutant MXAN2872 genes in the ΔMXAN2872 strain.To construct the plasmid expressing the wt MXAN2872 protein, a 2.6-kb fragment encompassing the full-length MXAN2872 ORF and its putative promoter sequence was amplified using the primer pair MXAN2872FL-5/FL-3 and the DK1622 chromosome as a template. Site-directed mutagenesis of MXAN2872 was carried out by the two-step PCR method. All final PCR products were gel purified, digested with XbaI and PstI, and ligated to the pBluescript vector cut with the same enzymes, followed by transformation into E. coli JM83. The fragments representing the wt and mutant MXAN2872 genes were confirmed by DNA sequencing. Subsequently, the 2.6-kb XbaI-PstI fragment of wt MXAN2872 or its mutant version, together with a 2.9-kb EcoRI-XbaI myxophage Mx8 attP fragment, was cloned into the EcoRI-PstI sites of pMP220 to create pMP-X, where X denotes either a full-length wt gene (MXAN2872FL) or a mutant gene (G71A or H219A). For the expression of wt and mutant MXAN2872 proteins, the resulting plasmids were separately introduced into the ΔMXAN2872 strain by electroporation. Proper integration of each plasmid at attB was verified by PCR.
Expression and purification of N-terminal histidine-tagged MXAN2872 wt and mutant proteins in E. coli.The MXAN2872 wt gene was amplified from the DK1622 chromosome using primers MXAN2872His6-5 and MXAN2872His6-3 (Table 2) and cloned into the NdeI/BamHI sites of the vector pET16b(+), generating pET-MXAN2872. After confirmation of the DNA sequence, the resulting plasmid was transformed into E. coli BL21(DE3). pET-MXAN2872-G71A was constructed in a similar way, except that pMP-MXAN2872-G71A was used as the PCR template.
To induce the expression of MXAN2872 wt or mutant proteins, E. coli BL21(DE3) cells harboring pET-MXAN2872 or pET-MXAN2872-G71A were grown to an OD600 of 0.4 in 80 ml of LA (LB broth with 50 μg ml−1 ampicillin) at 37°C. Then, IPTG (isopropyl-β-d-thiogalactopyranoside) was added at a final concentration of 0.4 mM, and the culture was incubated for another 6 h at 27°C. The following procedure was carried out at 4°C. Cells were harvested by centrifugation at 10,000 × g for 10 min, resuspended in 8 ml of lysis buffer (50 mM NaH2PO4, 1 M NaCl, 10 mM imidazole, pH 8.0), sonicated, and centrifuged at 10,000 × g for 10 min to pellet the cellular debris and inclusion bodies. After adding Triton X-100 to a final concentration of 0.5%, the resulting supernatant, representing the soluble fraction of proteins, was incubated with 3 ml of Ni-nitrilotriacetic acid (NTA)–agarose (Qiagen) for 2 h on a rotating platform. Subsequently, the protein-Ni-NTA mixture was loaded into a Poly-Prep chromatography column (Bio-Rad). The column was then washed sequentially with lysis buffer containing 50 and 80 mM imidazole and finally eluted with lysis buffer containing 200 mM imidazole. Samples were resolved by SDS-PAGE to detect the 70-kDa His-tagged MXAN2872 wt or mutant proteins. Fractions containing the target protein were pooled and dialyzed overnight against phosphate-buffered saline (pH 7.6).
Spectroscopy.Purified His-tagged MXAN2872 protein or its mutant version was spectrophotometrically scanned for the FAD- or NADPH-binding absorption spectrum from 250 to 700 nm with a Shimadzu UV-2450 spectrophotometer. To determine if the cofactor was lost during purification, MXAN2872 wt or mutant proteins were also mixed with FAD (Sigma) or NADPH (Roche) at a 1:5 molar ratio of protein to cofactor at room temperature for 2 h. Then, unbound cofactor was removed by dialysis overnight at 4°C and the resulting solutions were subjected to UV-visible scanning.
RESULTS
Genetic screen for genes controlling mcuABC expression.Youderian and colleagues described the use of the MycoMar transposable element of the mariner family to identify genes required for adventurous and social gliding motility in M. xanthus (28, 33). Their findings indicated that the MycoMar transposon is active in M. xanthus and the spectrum of insertions generated by this transposon is broader than that generated by the commonly used eubacterial transposon Tn5. In this study, we chose DK1622/pMP-mcuABC, a strain carrying an mcuABC-lacZ transcriptional fusion (i.e., MXAN3883-lacZ) (16), as our starting strain. We electroporated this reporter strain with a plasmid carrying MycoMar and screened Kmr electroporants for the lack of a blue color after 24 h of development on X-Gal-containing TPM plates. Among approximately 15,000 colonies screened, 3 colonies displayed a yellow or white color, and diminished McuA accumulation in these transposon mutants was confirmed by Western blotting analyses. Figure 1 shows the result for one mutant, which is the focus of the present study. McuA accumulation in this mutant strain was nearly eliminated, and when assessed for developmental competency, the mutant was found to be defective for fruiting up to 6 days on TPM agar (data not shown). To map the position of the transposon insertion, chromosomal DNA from this mutant was purified, cleaved, ligated, and used to transform an E. coli pir+ recipient. Plasmids were prepared from E. coli kanamycin-resistant colonies, and the junction sequences of the transposon insertion were determined. It turned out that the transposon was located in MXAN2872, a gene predicted to code for a flavin-binding monooxygenase which is not identical to BcsA (34).
Expression of mcuA is abolished in a mutant generated by MycoMar insertion. Total cell lysates were prepared from M. xanthus cells after exposure to starvation on TPM agar for the indicated periods of time and analyzed by Western blotting using McuA antiserum. Protein from an equal number of input cells (5 × 107 cells) was loaded in each lane. The smaller protein band in DK1622 cells presumably corresponds to the McuA degradation product. Lane M, molecular size markers; arrowhead, McuA protein.
MXAN2872 is required for McuA accumulation during early development.Examination of the MXAN2872 locus revealed that there is a 4-base overlap between the ORFs of MXAN2872 and its downstream gene, MXAN2873, suggesting that MXAN2872 and MXAN2873 may be cotranscribed. The proposed transcriptional coupling of these two genes was confirmed by reverse transcription-PCR (RT-PCR) (Fig. 2A). Because MXAN2872 and MXAN2873 are flanked by two genes that are predicted to be transcribed in the opposite direction, interpretation of the precise effect of the transposon insertion on mcu expression is thus complicated by a potential polar effect on the downstream gene, MXAN2873, which, according to the NCBI annotation, encodes an NHL (named after NCL-1, HT2A, and Lin-41) repeat-containing protein. The NHL repeat can be found in a variety of enzymes. To know whether the effect of the insertion mutation on mcu expression was strictly due to the loss of MXAN2872, a mutant carrying an in-frame deletion in the MXAN2872 gene (ΔMXAN2872) was constructed in a DK1622 (wt) background. The isogenic strains (DK1622 and DK1622 ΔMXAN2872) were exposed to starvation on TPM starvation agar, and then cells were collected at different time points and boiled in 1× SDS-PAGE loading buffer and soluble cell lysates were resolved by SDS-PAGE followed by Western blotting using McuA antiserum. Accumulation of McuA was observed in the wt strain at all developmental time points examined, whereas in the ΔMXAN2872 strain, even though McuA was detectable at 48 and 72 h, it was completely missing at 24 h (Fig. 2B). Note that with respect to the level of McuA, strain ΔMXAN2872 was clearly complemented by integrating at attB a plasmid (pMP-MXAN2872FL) expressing MXAN2872 under the control of its native promoter (Fig. 2C), indicating that a lack of McuA in the ΔMXAN2872 strain was indeed caused by the loss of MXAN2872 function. Meanwhile, we also constructed an MXAN2873 deletion mutant (the ΔMXAN2873 mutant). The amount of McuA in ΔMXAN2873 cells that had been starved for 24, 48, and 72 h was similar to that in wt cells, although at 18 h, the McuA in the ΔMXAN2873 cells was detected at a level slightly lower than that in the wt cells (Fig. 2D). From these observations, we conclude that the effect of the transposon insertion on mcu expression was mainly due to the disruption of MXAN2872 and to a lesser extent was probably due to a polar effect on MXAN2873 expression. Taken together, genome-wide transposon mutagenesis identified MXAN2872 to be a candidate necessary for mcu expression during the early development of M. xanthus.
MXAN2872 and MXAN2873 are transcriptionally coupled, yet they play distinct roles in regulating mcuABC expression. (A) Detection of the MXAN2872 and MXAN2873 transcripts by RT-PCR. Two primers spanning the MXAN2872 and MXAN2873 ORFs were used for PCR amplification with cDNA (lane 3), total RNA that was used for cDNA synthesis (lane 2), and chromosomal DNA (lane 4) as the templates. Lane 1, DNA markers. (B) Accumulation of McuA was eliminated in strain ΔMXAN2872 during early development (up to 24 h poststarvation). (C) The plasmid (pMP-MXAN2872FL) expressing wild-type MXAN2872 complemented the McuA accumulation defect of strain ΔMXAN2872 (ΔMXAN2872/MXAN2872) during early development. (D) MXAN2873 deletion only moderately inhibited McuA accumulation at 18 h poststarvation. For panels B to D, cells were treated and samples were analyzed as described in the legend to Fig. 1. wt, wild-type DK1622. Lanes M, molecular size markers.
MXAN2872 and MXAN2873 regulate the timing of fruiting body morphogenesis.When assessed on TPM agar for developmental competency, the ΔMXAN2872 strain did not form aggregates and translucent mounds until after 24 h, whereas in the wild-type DK1622 strain, aggregation occurred as early as 6 h and translucent mounds were evident at 12 h (Fig. 3); thus, the MXAN2872 deletion mutation delayed aggregation by more than 12 h. To determine whether the developmental defect in the ΔMXAN2872 strain was caused by the loss of MXAN2872 function, the phenotype of the ΔMXAN2872 strain complemented with pMP-MXAN2872 was also examined. As shown in Fig. 3, the developmental defect in the ΔMXAN2872 strain was corrected by the plasmid expressing MXAN2872, providing evidence that the defect in the ΔMXAN2872 strain was caused by the loss of MXAN2872 function. In comparison, the aggregation defect of the ΔMXAN2873 mutation was less severe. ΔMXAN2873 cells started to aggregate at 12 h, and mound-shaped structures were evident at 18 h poststarvation. Therefore, in the DK1622 background, an MXAN2873 deletion mutation delayed aggregation by less than 6 h. These observations suggest that developmental defects in the transposon mutant may be ascribed primarily to the inactivation of MXAN2872 and secondarily to a polar effect on its downstream gene, MXAN2873. These observations, together with the data shown in Fig. 2, also indicate that in MXAN2872 and MXAN2873 mutants there is a correlation between aggregation delay and corresponding delay in McuA production. In other words, developmental expression of mcuABC was induced at a time when cells were beginning to aggregate.
Developmental phenotype of the wt (DK1622) and various mutant strains containing different MXAN2872 or MXAN2873 alleles. Cells were exposed to starvation on TPM agar for the indicated period of times and examined under a Nikon SMZ1500 stereomicroscope.
MXAN2872 protein binds FAD and NADPH.Since MXAN2872 is more important than MXAN2873 in regulating mcu expression and M. xanthus development, our subsequent studies focused on MXAN2872. An analysis of the MXAN2872 protein with the TMHMM2 server (35) did not identify transmembrane-spanning regions, suggesting that this protein is localized in the cytoplasm. According to the NCBI annotation, MXAN2872 encodes a flavin-binding monooxygenase. Sequence analysis of the MXAN2872 protein revealed three key characteristic sequence motifs of Baeyer-Villiger monooxygenases (BVMOs): two dinucleotide binding motifs (Rossmann folds) predicted to bind FAD and NADPH, respectively [GXGXX(G/A)], a BVMO fingerprint motif [FXGXXXHXXXW(P/D)], and a recently recognized BVMO-typifying motif [(A/G)GXWXXXX(F/Y)P(G/M)XXXD] (Fig. 4) (36–38). To test if the MXAN2872 protein binds flavin, the MXAN2872 gene was cloned into the vector pET-16b (pET-MXAN2872) and expressed in E. coli. A band of 70 kDa, which was in agreement with the size of the His-tagged MXAN2872 protein, was specifically detected in the strain carrying pET-MXAN2872 (data not shown). The recombinant MXAN2872 protein from the soluble fraction was purified to near homogeneity by affinity chromatography (Fig. 5A). On observation with the naked eye, the purified product did not show a yellow color, which is characteristic of flavoproteins; in addition, further investigation of this protein by a UV-visible scanning experiment also showed that it had not bound the flavin prosthetic group. To determine if flavin was lost during purification, the purified protein was mixed with FAD, dialyzed, and then spectrophotometrically scanned for the FAD-binding absorption spectrum. As shown in Fig. 5B, the MXAN2872 protein displayed discernible absorption maxima near 370 and 450 nm, characteristic of the spectrum of FAD, suggesting that this protein does bind FAD. Similarly, binding of NADPH to MXAN2872 was also spectrophotometrically demonstrated, with an absorption maximum being found at 340 nm (Fig. 5D).
Alignment of the MXAN2872 protein sequence with the sequences of two prototypes of the Baeyer-Villiger monooxygenase (BVMO) subfamily belonging to class B flavoprotein monooxygenases. HAPMO, 4-hydroxyacetophenone monooxygenase from Pseudomonas fluorescens ACB (GenBank accession number AF355751); CHMO, cyclohexanone monooxygenase from Acinetobacter sp. strain NCIB 9871 (GenBank accession number M19029). Dinucleotide binding motifs and BVMO signature sequence motifs are indicated in italics on the top line. Numbers indicate the coordinates of the amino acid sequence of each protein.
UV-visible absorbance spectra of MXAN2872 protein and its G71A mutant form. Purified proteins were separately incubated with FAD or NADPH at room temperature for 2 h. Then, unbound free FAD or NADPH was removed by dialysis. Solutions inside and outside the tubing were spectrophotometrically scanned. MXAN2872 represents the protein before being incubated with FAD or NADPH. (A) SDS-polyacrylamide gels showing the His-tagged MXAN2872 wt and mutant proteins purified from E. coli. The gels were stained with Coomassie blue. (B) Spectra of the MXAN2872 protein after incubation with FAD. (C) Spectra of the MXAN2872-G71A protein after incubation with FAD. (D) Spectra of the MXAN2872 protein after incubation with NADPH.
The putative FAD-binding site in the MXAN2872 protein may be required for activity.In the MXAN2872 protein sequence there are two GXGXX(G/A) motifs involved in the binding of either FAD or NADP(H) (Fig. 4). Given the location of the NADP and FAD sites of phenylacetone monooxygenase, the first BVMO for which the crystal structure has been determined (37), the MXAN2872 protein probably binds FAD at its N-terminal Rossmann fold, while NADP is bound to the C-terminal one. The importance of the first glycine in the GXGXX(G/A) motifs is to allow a tight turn of the main chain from a β strand into a ligand-binding loop between the β strand and α helix of a Rossmann fold (39). To examine whether the potential FAD-binding site is important for protein activity, we constructed an MXAN2872 mutant allele in which the first glycine residue in the N-terminal Rossmann fold was replaced by alanine (G71A). Note that the mutated protein lost its ability to bind FAD (Fig. 5C). A plasmid carrying this mutant allele (pMP-MXAN2872-G71A) was introduced into the ΔMXAN2872 strain and integrated at the Mx8 attB site. As shown in Fig. 3, although the mutant allele ameliorated the aggregation defect of the ΔMXAN2872 strain to a certain extent, it was unable to fully complement the ΔMXAN2872 mutation like the wt allele (DK1622 ΔMXAN2872/pMP-MXAN2872FL). Meanwhile, we carried out immunoblot analyses of total proteins isolated from developing ΔMXAN2872 cells containing either wt MXAN2872 or its mutant version encoding the G71A substitution. Compared with the findings for cells into which the wt allele was introduced, McuA was not detected at 18 h in cells into which a plasmid expressing MXAN2872-G71A was introduced (Fig. 6A). Therefore, on the basis of morphogenetic and mcu expression studies, we conclude that the G71A mutant allele is unable to fully correct the defects in the ΔMXAN2872 strain, suggesting that the putative FAD-binding motif may be important for MXAN2872 protein activity. Notably, in the ΔMXAN2872 strain complemented with the G71A allele, slightly delayed McuA accumulation paralleled a slightly delayed cell aggregation phenotype.
The putative FAD-binding motif and a BVMO fingerprint motif in the MXAN2872 protein may be important for MXAN2872 protein activity. M. xanthus cells were treated and samples were analyzed as described in the legend to Fig. 1. (A) Accumulation of McuA in total cell extracts of the ΔMXAN2872 strain complemented with a plasmid encoding the MXAN2872 wt or G71A mutant. (B) Accumulation of McuA in total cell extracts of the ΔMXAN2872 strain complemented with a plasmid encoding the MXAN2872 wt or H219A mutant. Lanes M, molecular size markers.
The BVMO fingerprint sequence of the MXAN2872 protein may be required for activity.BVMOs exhibit a three-dimensional architecture consisting of two domains for binding FAD and NADP, respectively. The fingerprint sequence FXGXXXHXXXW(P/D) connects the FAD-binding domain to the NADP-binding domain and is involved in coordinating domain movements that are predicted to occur during the catalytic cycle (37, 40). It has been demonstrated that the strictly conserved central histidine in the signature sequence is crucial for catalysis (41). Therefore, we attempted to investigate the importance of H219 for MXAN2872 protein activity. For this purpose, a plasmid expressing MXAN2872-H219A (pMP-MXAN2872-H219A) was integrated at the Mx8 attB of the ΔMXAN2872 strain. Similar to the findings for the MXAN2872-G71A allele, the H219A allele was unable to fully complement the aggregation and mcuA expression defects of the ΔMXAN2872 strain (Fig. 6B and data not shown), suggesting the necessity of the fingerprint sequence for the full activity of the MXAN2872 protein.
MXAN2872 is developmentally regulated.The observations that MXAN2872 plays a role in mcu expression and M. xanthus development led us to examine the timing of MXAN2872 expression. For this purpose, we constructed the plasmid pMP-MXAN2872, which carried a MXAN2872-lacZ transcriptional fusion and was integrated at the attB site of the DK1622 chromosome. This fusion construct contained the N-terminal 20-amino-acid region of MXAN2872 together with its upstream regulatory sequence. As shown in Fig. 7, β-galactosidase activity was not detectable within 15 h after starvation-initiated development, began to increase by 18 h, reached a maximum at about 48 h, and then slightly decreased over the next 24 h. Therefore, like mcuABC expression, expression of MXAN2872 is specifically induced during development. However, the activity of β-galactosidase expressed from MXAN2872-lacZ did not follow the same pattern observed for the mcuABC-lacZ fusion (i.e., MXAN3883-lacZ) with respect to both the timing and the level of expression (16). For instance, developmental expression of MXAN2872 started about 3 to 6 h later than that of mcuABC and remained at a relatively high level by 72 h, suggesting that there might be regulatory machineries other than MXAN2872 that modulate mcuABC expression, either directly or indirectly. On the other hand, the phenotypes of the mcu mutant and the MXAN2872-null mutant provide a clear indication that MXAN2872 is doing much more than just regulating the mcuABC operon (16).
Expression of MXAN2872 is developmentally regulated. The specific activity of β-galactosidase expressed from a MXAN2872-lacZ transcription fusion during development on TPM starvation agar was measured in a wt (DK1622) background. Samples were harvested in triplicate at various time points in each experiment. Error bars represent standard deviations of the means. The results shown here are representative of those from five independent experiments. ONP, o-nitrophenol.
Expression of mcuA and MXAN2872 depends on FruA.FruA is a key response regulator in the C-signaling pathway which induces aggregation, sporulation, and developmental gene expression at specific thresholds after 6 h of starvation (7, 42). To resolve the effect of FruA on mcuABC expression, we analyzed McuA accumulation in DK7873 (43) and ΔfruA strains, both of which are fruA-null mutant strains. As shown in Fig. 8, the fruA mutants failed to synthesize the McuA protein at all time points examined, indicating that FruA is necessary for mcuABC expression throughout development.
mcuABC expression depends on FruA. McuA accumulation patterns in wt cells (DK1622), fruA disruption mutant cells (DK7873), and fruA deletion mutant cells (ΔfruA) were analyzed by Western blotting. Cells were treated and samples were analyzed as described in the legend to Fig. 1. Lanes M, molecular size markers.
Clearly, both MXAN2872 and FruA exert control over the mcuABC operon. This prompted us to examine the relationship between FruA and MXAN2872 in regulating mcuABC expression. As mentioned above, although the MXAN2872 deletion mutation delayed cell aggregation and diminished mcuA expression at early developmental stages, no difference between wt and mutant strains was detectable at 24 to 48 h after starvation because normal fruiting bodies as well as McuA accumulation were observed in both cases. In view of this, it seems unlikely that MXAN2872 is involved in regulating FruA-dependent activities which are essential for fruiting body morphogenesis and McuA production throughout development.
To determine if MXAN2872 could be a target regulated by FruA, we constructed a plasmid, pMP-MXAN2872, that carries an MXAN2872-lacZ transcriptional fusion. The plasmid was separately introduced into DK1622 (wt) and a ΔfruA mutant, and expression of MXAN2872-lacZ in each strain was assayed. The data presented in Fig. 7 demonstrate that FruA is absolutely required for MXAN2872 expression, as β-galactosidase activity in the ΔfruA background was hardly detectable.
MXAN2872 is not directly involved in regulating mcuABC expression.Because the extent to which MXAN2872 deletion and point mutations affect mcu expression agrees well with the extent to which they affect the efficiency of cell aggregation, we reasoned that mcuABC expression may be tied to the aggregation stage of the M. xanthus multicellular developmental program. It is well-known that myxospore formation can be induced in an aggregation-independent way simply by adding certain chemicals to vegetatively growing liquid cultures. In particular, glycerol is capable of inducing the differentiation of virtually all of the cells in the culture into spores within 8 h (44). Therefore, we examined whether McuA is produced during glycerol-induced spore formation. Our results revealed that McuA was not produced either inside or on the surface of the cells harvested at 2, 4, and 8 h after addition of glycerol (data not shown). On the contrary, protein U, another spore coat protein of M. xanthus which is highly homologous to McuA, can be detected in glycerol-induced spores (45). These observations indicate that mcuA is expressed only in starvation-induced sporulating cells that have undergone the aggregation process and suggest that the effect of various MXAN2872 mutations on mcuA expression may be secondary to that on cell aggregation.
DISCUSSION
Using a powerful mariner mutagenesis approach, we were able to identify a MXAN2872-dependent mechanism for controlling mcuABC expression, which was subsequently confirmed by in-frame deletion, genetic complementation, and site-directed mutagenesis. However, MXAN2872 was indispensable for mcuABC expression only during early development (<24 to 48 h). Considering that aggregation of the ΔMXAN2872 strain was not complete until approximately 36 h, the timing of the developmental defects in the MXAN2872 mutant seems to coincide with the timing of the diminished expression of mcuABC. In other words, MXAN2872 is needed to regulate the efficiency of cell aggregation at early developmental stages besides mcuABC expression.
M. xanthus is capable of decoding multicellular morphological checkpoints and of transforming this information into appropriate changes in gene expression (7), and any mutation that blocks an early stage can affect the expression of genes at later stages without being directly involved in their regulation (46). In view of a putative enzyme-like nature of the MXAN2872 protein and the good agreement between the timing of McuA accumulation and that of the developmental progression of several mutants carrying MXAN2872 deletion or point mutations, it is reasonable to propose that the loss of MXAN2872 delays the onset of aggregation, which may in turn result in the delay of mcuABC expression. This hypothesis is supported by the absence of McuA in glycerol-induced spores, where sporulation is uncoupled from aggregation. Similarly, the complete block of aggregation by the loss of FruA results in the complete block of McuA production. Therefore, MXAN2872 indirectly participates in modulating the mcuABC operon. In addition, several lines of evidence indicate that factors other than MXAN2872 are directly or indirectly involved in regulating the mcuABC operon: (i) the expression profile of MXAN2872 did not perfectly match that of mcuABC, with MXAN2872 being expressed slightly later in development than mcuABC; (ii) deletion of MXAN2872 did not cause a defect in mcuABC expression after >24 h of development, indicative of the functional redundancy of mcuABC regulators; and (iii) we have identified at other loci transposon insertions that affected mcuABC expression (unpublished data).
At present we do not know how the MXAN2872-encoded putative BVMO, an enzyme catalyzing the insertion of an oxygen atom into a carbon-carbon bond of a carbonylic substrate, modulates M. xanthus development and, consequently, affects mcuABC expression. However, it has been well documented that bacteria are capable of using metabolic intermediates to sense the metabolome and, accordingly, modulate gene expression and cell behavior. In filamentous cyanobacteria, for instance, the Krebs cycle metabolite 2-oxoglutarate serves as a nitrogen limitation signal which is sensed by NtcA, a transcription factor responsible for the initiation of heterocyst differentiation (47). Therefore, we speculate that in M. xanthus, besides the well-known intracellular starvation signal (p)ppGpp, which is necessary to induce the developmental cycle (3), other critical metabolites could also function as nutrient limitation signals. The MXAN2872-encoded BVMO may be involved in generating such a metabolic signal which activates a potential signaling pathway necessary for cell aggregation. The formation of cell aggregates where cells are exposed to a high level of C signaling triggers the activation of transcriptional regulators, resulting in the induction of sporulation genes, including mcuABC. The function of MXAN2872 is apparently different from that of another putative flavin-binding monooxygenase, BcsA, which normally acts to inhibit development (34). It is noteworthy that in M. xanthus, expression of C-signal-dependent genes is usually morphogenesis dependent. Consistent with this, mcuA, whose product failed to accumulate in a C-signal-null mutant (data not shown), depends on cell aggregation for timely expression, supporting the existence of one or more morphological checkpoints important for the expression of C-signal-dependent genes (7).
Our finding that a putative metabolic enzyme participates in regulating archaic CU gene expression, albeit most likely in an indirect manner, is not original, because Qaisar et al. recently reported that paerucumarin, a secondary metabolite produced by pvcABCD-encoded enzymes, influences P. aeruginosa biofilm development by regulating the expression of two classical CU pathways in this bacterium, CupB and CupC (48). A metabolic enzyme can sometimes act directly as a transcriptional factor (49). Because mutations in two potential sites involved in catalysis compromised the ability of the MXAN2872 protein to regulate mcuABC expression, the present data support a function of the MXAN2872 protein as an enzyme; however, we cannot completely rule out the possibility of a second function of this protein as a transcription regulator.
ACKNOWLEDGMENTS
We thank Zhaomin Yang, Yuezhong Li, Javier Abellón-Ruiz, and Montserrat Elías-Arnanz for providing pBJ113 and/or their advice on constructing in-frame deletions based on this plasmid, Youming Zhang for the gift of a plasmid harboring the MycoMar transposon, and Mitchell Singer for strains DK1622 and DK7873. We also thank the reviewers for their insightful comments.
This work was supported by a grant from the Natural Science Foundation of China (30571008) and by a research fund from the Key Laboratory of Ministry of Education for Developmental Genes and Human Diseases.
FOOTNOTES
- Received 9 December 2014.
- Accepted 12 January 2015.
- Accepted manuscript posted online 20 January 2015.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
