Identification of a gene involved in polysaccharide export as a transcription target of FruA, an essential factor for Myxococcus xanthus development.

Fruiting body development in Myxococcus xanthus is a multicellular event that is coordinated by exchanging intercellular signals. FruA is a transcription factor essential for fruiting body development and is thought to play a key role in the C-signal pathway. Here we present the first identification of a gene regulated by FruA. The gene was isolated from a genomic library via in vitro selection in a DNA binding assay by using the DNA-binding domain of FruA tagged with His(8) at the C-terminal end (FruA-DBD-H(8)). The gene, named fdgA (FruA-dependent gene A), encodes a protein homologous to the outer-membrane auxiliary family protein involved in the polysaccharide export system. FruA-DBD-H(8) bound the upstream promoter region of the fdgA gene from nucleotide -89 to nucleotide -64 with respect to the transcription initiation site, which was required for the induction of fdgA expression during development. fdgA mRNA induced during development was absent in a fruA deletion strain. The deletion of fdgA resulted in defective fruiting body formation and reduced sporulation efficiency (1% that of the parent strain). Moreover, FruA was required for the developmental expression of sasA, which is also involved in the biosynthesis of the lipopolysaccharide O-antigen and is required for fruiting body development. Furthermore, the expression of both fdgA and sasA was partially dependent on the C-signal. These findings expand our understanding of the signal transduction pathway mediated by FruA during development in M. xanthus.


Fruiting body development in Myxococcus xanthus is a multicellular event that is coordinated by exchanging intercellular signals.
FruA is a transcription factor essential for fruiting body development and is thought to play a key role in the C-signal pathway. Here we present the first identification of a gene regulated by FruA. The gene was isolated from a genomic library via in vitro selection in a DNA binding assay by using the DNA-binding domain of FruA tagged with His 8 at the C-terminal end (FruA-DBD-H 8 ). The gene, named fdgA (FruA-dependent gene A), encodes a protein homologous to the outer-membrane auxiliary family protein involved in the polysaccharide export system. FruA-DBD-H 8 bound the upstream promoter region of the fdgA gene from nucleotide ؊89 to nucleotide ؊64 with respect to the transcription initiation site, which was required for the induction of fdgA expression during development. fdgA mRNA induced during development was absent in a fruA deletion strain. The deletion of fdgA resulted in defective fruiting body formation and reduced sporulation efficiency (1% that of the parent strain). Moreover, FruA was required for the developmental expression of sasA, which is also involved in the biosynthesis of the lipopolysaccharide O-antigen and is required for fruiting body development. Furthermore, the expression of both fdgA and sasA was partially dependent on the C-signal. These findings expand our understanding of the signal transduction pathway mediated by FruA during development in M. xanthus.
Developmental programs in living organisms are guided by a number of different signal transduction systems involved in the integration of environmental and cellular signals. These systems regulate the temporal and spatial expression of genes specific to distinctive stages of development. The developmental progression of a soil bacterium, Myxococcus xanthus, is also controlled by a number of signal transduction systems (1,2). Upon nutrient limitation, M. xanthus cells begin to coordinately migrate toward aggregation centers by gliding on a solid surface to form a fruiting body that holds 10 5 cells. Inside fruiting bodies only 10% of the starved cells differentiate to spores. The developmental process is achieved by a series of sophisticated intercellular signaling pathways that regulate the expression of a specific set of genes. Five intercellular signals, the A-, B-, C-, D-, and E-signals, are known to be involved and each is essential for fruiting body development. The C-signal is a cell-surface associated morphogen required for rippling, aggregation, and sporulation, and its transmission occurs by a contact-dependent mechanism that involves end-to-end contact between cells (3)(4)(5). Recently, it has been reported that the C-signal (p17) is a proteolytic product of CsgA (p25) and corresponds to the C-terminal part of p25 (6).
The fruA gene encodes a putative transcription factor essential for development and is a member of the FixJ response regulator subfamily of the two-component His-Asp phosphorelay system (7,8). FruA is required for the execution of aggregation, fruiting body formation, and subsequent sporulation. The expression of fruA is developmentally regulated, dependent on A-and E-signals, and initiates at 6 h after the onset of development. Recently, a fruA promoter-binding protein was isolated from developmental cell extracts by using a DNA binding assay with the DNA fragment containing the upstream region of the fruA promoter (9). Based upon its partial sequence, this fruA promoter-binding protein was found to be identical to MrpC, which belongs to the cyclic AMP receptor protein family of transcriptional regulators (9). MrpC activates the expression of fruA by binding cis-acting elements in the upstream region of the fruA promoter. mrpC and mrpAB had previously been identified via transposon insertion mutagenesis as essential loci for cellular aggregation and sporulation (10). The mrpA and mrpB genes encode a histidine kinase and a NtrC-like response regulator, respectively, of the two-component His-Asp phosphorelay system. It has been proposed that the expression of mrpC depends on the mrpAB operon and that MrpC autoregulates its own gene expression (11).
Based on the analysis of developmental markers in a fruA::Tc strain, the expression of various developmental genes induced after 5 h of development was severely hampered (7). In addition, genes known to be expressed before 4 h of development were found to be normally expressed during early development but were unable to be repressed later in fruA::Tc (7). Furthermore, among six proteins identified by the analysis of protein expression patterns in the wild-type, fruA::Tc, and ⌬csgA strains during development, the expression of five proteins was found to be C-signal-independent and one to be C-signal-dependent (12). These results indicate that developmental genes under the control of FruA can be classified into two groups, C-signal-independent and C-signal-dependent (12). In addition, transposon mutagenesis identified mutants involved in the C-signal pathway that were phenotypically divided into two classes (13). Class I mutants are deficient in aggregation but sporulate and produce C-factor at wild-type levels. Class II mutants have deficiencies in all C-factor responses. Because fruA was also identified genetically as the only class II mutant gene (14), FruA has been proposed to play a key role and to be activated by phosphorylation with a C-signal receptor kinase in the C-signal transduction system (15). However, such a kinase has not been identified to date.
Given that the expression of a number of genes was altered between the wild-type and fruA::Tc strains during development (7,8,12), it was important to identify which genes are directly activated or repressed by FruA. To identify genes directly regulated by FruA, we performed in vitro selection for a genomic DNA library by using the DNA-binding domain of FruA tagged with eight histidine residues at the C-terminal end (FruA-DBD-H 8 ). 2 We isolated a DNA fragment containing the promoter region and the N-terminal part of a protein homologous to the outer membrane auxiliary (OMA) family protein, which is involved in the polysaccharide export system. This gene was named fdgA (FruAdependent gene A). cis-Acting elements in the fdgA promoter were identified, and developmental expression of fdgA was not detectable in a ⌬fruA strain newly constructed in this study. Moreover, the deletion of fdgA resulted in defective fruiting body formation and reduced sporulation efficiency. In addition, the induction of the sasA locus that is required for the biosynthesis of the lipopolysaccharide O-antigen and development (16,17) was not detected in the ⌬fruA strain during development.

MATERIALS AND METHODS
Bacterial Strains and Growth Conditions-M. xanthus DZF1 (18) was used as a parent strain and was grown in CYE medium (19) supplemented with kanamycin or streptomycin when necessary. For fruiting body formation, M. xanthus was spotted on clone-fruiting (CF) agar plates (20). Escherichia coli JM83 (21) was used as a recipient strain for transformation unless otherwise mentioned and was grown in Luria-Bertaini medium (22) supplemented with antibiotics when necessary.
Preparation of FruA-DBD-H 8 -The DNA fragment encoding FruA-DBD-H 8 from Pro at 152 to Leu at 229 tagged with histidine residues at the C-terminal end was amplified by PCR with pMF05 (7) and the oligonucleotide primers 5Ј-TCCATATGCCGGTGACGTCACCCAC-C-3Ј and 5Ј-TCGGATCCCTA(GTG) 8 GAGGTCCGGCGGCGGCC-GGA-3Ј (NdeI and BamHI sites are underlined). PCR products were digested with NdeI and BamHI and ligated into pET11a (Novagen). The resultant plasmid, pET11a/FDBD-H 8 , was transformed into E. coli BL21 (DE3) (23). FruA-DBD-H 8 was induced by the addition of isopropyl 1-thio-␤-D-galactopyranoside at a final concentration of 1 mM in Luria-Bertani medium supplemented with ampicillin at 37°C for 2 h. Cells were harvested by centrifugation and washed with 10 mM Tris-HCl (pH 7.9). Cells were resuspended in the buffer H (10 mM Tris-HCl (pH 7.9), 500 mM NaCl, 1 mM ␤-mercaptoethanol, and 10 mM imidazole) supplemented with protease inhibitors (Roche Applied Science) and disrupted by sonication. After centrifugation at 100,000 ϫ g for 30 min, the supernatant was mixed with Triton X-100 at a final concentration of 0.1% and was applied to nickel-nitrilotriacetic acid (Qiagen). After incubation for 30 min, the column was washed with the buffer H containing Triton X-100 and then with the buffer H containing 40 mM imidazole. FruA-DBD-H 8 was eluted with a buffer containing 10 mM Tris-HCl (pH 7.9), 50 mM NaCl, 1 mM ␤-mercaptoethanol, and 250 mM imidazole. The eluate was dialyzed against a buffer composed of 10 mM Tris-HCl (pH 7.9), 50 mM NaCl, 1 mM ␤-mercaptoethanol, 1 mM EDTA, and 50% (w/v) glycerol. All procedures for the protein purification were performed either on ice or in a cold room.
In Vitro Selection for a Genomic DNA Library-A genomic DNA library was constructed by digesting partially with MspI chromosomal DNA prepared from M. xanthus DZF1. DNA fragments of ϳ 400 base pairs were isolated by PAGE. For in vitro selection a DNA binding assay was performed as described below with the library (200 ng) as a probe and purified FruA-DBD-H 8 (50 ng) in a total volume of 50 l. After the DNA binding reactions, the mixtures were subjected to PAGE. FruA-DBD-H 8 ⅐DNA complexes were isolated from the gel after PAGE. To isolate the complexes from the gel, the 32 P-labeled DNA fragment of 350 bp containing FruA-binding sites from the dofA promoter was used as a probe for reference. 3 The gel pieces containing FruA-DBD-H 8 ⅐DNA complexes with migration similar to that of FruA-DBD-H 8 ⅐ 32 P-labeled dofA promoter complexes were excised. The complexes were electroeluted from the gel pieces and extracted with phenol and chloroform. The DNA fragments were precipitated with ethanol and resuspended in water. A portion of the recovered DNA fragments was ligated at the ClaI site in modified pBluescript SKϪ (Stratagene), which contains an additional HindIII site between the ClaI site and the SalI site (total of two HindIII sites next to the ClaI site). The ligation mixtures were transformed, and plasmids were prepared from transformed cells growing on the plates. The plasmids were digested with HindIII, and the DNA fragments were isolated with PAGE. The DNA fragments were used for the next in vitro selection as a probe. The in vitro selection was repeated three times as described above. After the fourth selection, plasmids containing the selected fragments were sequenced.
DNA Binding Assay and Footprint Analysis-The fdgA promoter region from nt Ϫ165 to ϩ9 was amplified by PCR using the plasmid containing the 388-bp MspI fragment shown in Fig. 2 and the oligonucleotide primers (a) (5Ј-TCAAGCTTCCGGGAAATGGGAAGCGG-GA-3Ј) and (b) (5Ј-TCGGATCCGGAAAGCCCTTGGATCGCA-3Ј) (see Fig. 2 for the location; the HindIII and BamHI sites are underlined). An amplified product was digested with HindIII and BamHI and cloned in pBluescript SKϪ (Stratagene). A DNA fragment for a probe was prepared by digesting the plasmid with HindIII and BamHI and isolating the fragment by PAGE. The isolated DNA fragment was then labeled with [␣-32 P]dCTP by a Klenow fragment of DNA polymerase I.
For footprint analysis, a probe corresponding to the region from nt Ϫ165 to ϩ41 was prepared as described above with oligonucleotide primers (a) and (c) (5Ј-TCGGATCCTTAGCAGCGCGTGTGCCA-3Ј. A DNA fragment for a probe was prepared by digesting the plasmid with HincII and BamHI and isolating the fragment by PAGE. The isolated fragment was labeled with [␣-32 P]dCTP by a Klenow fragment of DNA polymerase I. Therefore, only the leading strand shown in Fig. 2 was labeled. DNA binding reactions were performed at 25°C for 10 min in 10 l of the reaction mixtures containing 20 mM Tris acetate (pH 7.9), 50 mM potassium acetate, 10 mM magnesium acetate, 1 mM dithiothreitol, 10 g/ml bovine serum albumin, 10% glycerol, and 1 g of poly(dI-dC)⅐poly(dI-dC) (Amersham Biosciences). The binding patterns were analyzed by 5% PAGE followed by autoradiography. For footprint analysis, after DNA binding reactions DNase I was added to the reaction mixtures, and the mixtures were further incubated at 25°C for 3 min. The reactions were stopped by adding the solution containing 25 mM EDTA and 0.5 M ammonium acetate. After extraction of the reaction mixtures with phenol/chloroform/isoamylalcohol (25:24:1), the DNA probes were precipitated with ethanol in the presence of glycogen. The DNA probes were resuspended in sequence loading buffer and analyzed by sequencing gel followed by autoradiography.
lacZ Fusion Analysis-The fdgA promoter regions from nt Ϫ165, Ϫ93, Ϫ82, Ϫ71, and Ϫ62 to ϩ162 were fused to lacZ in the transcriptional fusion vector pZKAT (9). ␤-Galactosidase activity was measured at 0 and 20 h after the initiation of development on CF agar plates as described (24).
Construction of Mutant Strains-A ⌬fruA strain was constructed by replacing the ClaI-PvuII fragment with a streptomycin resistance gene. The ClaI and the PvuII sites are located at the positions 11 and 213 from the N-terminal end of FruA, respectively (7).
A ⌬fdgA strain was constructed by replacing the BsaAI-SmaI fragment with a kanamycin resistance gene. The BsaAI and SmaI sites are located in the promoter region and the open reading frame (ORF) of the fdgA gene, respectively. Thus, amino acid residues from 1 to 120 of FdgA are deleted. A ⌬fdgA/fdgA ϩ strain was constructed by integrating the plasmid containing the promoter region (ϳ3 kbp) and the entire ORF of the fdgA gene into the chromosome of the ⌬fdgA strain.

Identification of a Gene Regulated by FruA from a Genomic DNA
Library-To identify a gene directly regulated by FruA, we performed in vitro selection for a genomic DNA library in a DNA binding assay. We used the DNA-binding domain of FruA tagged with eight histidine residues at the C-terminal end. The DNA fragments isolated after each selection were labeled with 32 P and were used for a DNA binding assay with FruA-DBD-H 8 . As the selections were repeated, more fragments were bound by FruA-DBD-H 8 (Fig. 1). A DNA fragment in the length of 388 bp isolated after the fourth cycle of the selection was found to contain a putative promoter region and an N-terminal part of an ORF based on a search in the Institute for Genomic Research web site at www.tigr.org (Fig. 2). The DNA fragment encompassing an entire gene was cloned from the genome. The isolated gene was termed FruA-dependent gene A or fdgA.
Expression of fdgA during Vegetative Growth and Development-fdgA expression was examined by primer extension analysis with total RNA prepared from DZF1 and ⌬fruA during vegetative growth (0 h) and fruiting body development (12 and 20 h). Three 5Ј-ends of fdgA mRNA were identified. One (P V ) of them was detected in both vegetative and developmental cells, and its expression was independent of fruA ( Fig.  3A). In contrast, the other two were detected only in developmental cells. One (P D1 ) of them was observed in early (12 h) and late (20 h) development, and the other (P D2 ) was in late development. These development-specific mRNAs were not detectable in the ⌬fruA strain. The nucleotides of these 5Ј-ends were determined by DNA sequence ladders (data not shown) and are shown in Fig. 2. The DNA fragment containing the promoter region from nt Ϫ165 to ϩ7 with respect to the transcription initiation site (P D1 ) was found to be bound by FruA-DBD-H 8 (Fig.  3B). Footprint analysis showed that the two regions from nt Ϫ89 to Ϫ74 (region b) and from nt Ϫ72 to at least Ϫ64 (region a) were protected by FruA-DBD-H 8 from DNase I and that the band (C at nt Ϫ73) was hypersensitive to DNase I between the two regions (Fig. 3C). As shown in Fig. 3B, because three complexes were observed when an amount of FruA-DBD-H 8 was increased, it appears to have three binding sites.
To examine the roles of these regions in fdgA expression in vivo, various regions of the upstream fdgA promoter were transcriptionally fused to the lacZ gene and introduced into the attB site. The activity of ␤-galactosidase was measured at 0 and 20 h of development (Fig. 3D). The deletion of the region from nt Ϫ165 to Ϫ94 (promoter Ϫ93) did not affect fdgA expression during vegetative growth and development, and the deletion of the region from nt Ϫ165 to Ϫ83 (promoter Ϫ82) slightly reduced fdgA expression during development. In contrast, the deletion of 11 bases between nt Ϫ82 and Ϫ72 (promoter Ϫ71) from promoter Ϫ82 resulted in drastic decrease in fdgA expression, and region a alone (promoter Ϫ62) failed in the activation of the fdgA promoter. The increase of ␤-galactosidase activity by promoter Ϫ62 during development was most likely due to the increase in basal level expression of fdgA (P V ) during development (Fig. 3A).
Function of fdgA-The fdgA gene encodes a protein homologous to GumB of Xanthomonas campestris that belongs to the OMA family involved in polysaccharide export (25). FdgA has a typical lipoprotein consensus sequence that includes the leader sequence followed by the invariant Cys, circled in Fig. 2, and the signal peptidase II cleavage site (26).
Analysis of DNA sequences revealed that another ORF (named Orf1) is located downstream of fdgA. However, reverse transcription PCR analysis showed that orf1 was transcribed independently of the fgdA promoter (data not shown), suggesting that fdgA is mono-cistronic.
To elucidate the function of the fdgA gene, a ⌬fdgA strain was constructed. Although the fdgA gene was expressed during vegetative growth, the ⌬fdgA strain grew in CYE liquid medium (19) similarly as the parent strain (data not shown). In contrast, during development the ⌬fdgA strain showed defective fruiting body formation on CF agar plates (Fig. 4). The ⌬fdgA strain formed bigger and fewer fruiting bodies, which were not as dark as those of the parent strain. In contrast, under this condition the ⌬fruA strain formed no apparent fruiting bodies (Fig.  4). A number of viable spores were then examined after the fruiting bodies were harvested and treated with sonication. The efficiency of sporulation of the ⌬fdgA strain was ϳ1% that of the parent strain after 5 days of development on CF agar plates.
To confirm that the defective development of the ⌬fdgA strain was due to the absence of the fdgA gene, the DNA fragment containing the promoter region and the coding region of the fdgA gene was introduced into the ⌬fdgA strain. The ⌬fdgA/fdgA ϩ strain formed fruiting bodies similar to those of the parent strain (Fig. 4). Furthermore, the ⌬fdgA/ fdgA ϩ strain formed viable spores in efficiency similar to that of the parent strain. Therefore, the fdgA gene was essential for fruiting body formation and sporulation during development.
Expression of the sasA Locus-Genes at the sasA locus that are involved in lipopolysaccharide O-antigen biosynthesis have been shown  to be essential for development (see "Discussion" for details) (16,17). Therefore, we next examined whether the sasA locus is regulated by FruA. Primer extension analysis showed that sasA mRNA induced during development (P D ) was undetectable in the ⌬fruA strain, but vegetative mRNA (P V ) was expressed in both the parent and the ⌬fruA strains (Fig. 5). Interestingly, the DNA binding assay revealed that FruA-DBD-H 8 did not bind the sasA promoter region under conditions in which FruA-DBD-H 8 could bind the fdgA promoter region (data not shown). Therefore, it appears likely that the sasA locus is indirectly regulated by fruA.
Effect of C-signal on Expression of fdgA and sasA-Because FruA is proposed to function in the C-signal transduction pathway (14,15), the expression of fdgA and sasA was examined in a ⌬csgA strain (27). Primer extension analysis revealed that fdgA mRNA induced during development (P D2 ) decreased in the ⌬csgA strain (Fig. 6A). In addition, sasA mRNA induced during development also decreased in the ⌬csgA strain (Fig. 6B). Therefore, it appears likely that the expression of both fdgA and sasA is partially dependent on the C-signal.

DISCUSSION
We have identified a gene, named fdgA, encoding an OMA family protein involved in the polysaccharide export system as a transcription target of FruA from the genomic library of M. xanthus. The expression of fdgA was differentially regulated during vegetative growth and development. The developmental expression of fdgA was dependent on FruA. The deletion of fdgA resulted in defective fruiting body formation and reduced sporulation efficiency during development. fdgA is the first gene identified as a transcription target of FruA.
To identify genes directly regulated by FruA, we performed in vitro selection for the M. xanthus genomic DNA library by using a DNA binding assay with FruA-DBD-H 8 . In the in vitro selection described here, a DNA fragment containing a binding site for a specific transcription factor can be isolated from a genomic DNA (28). The recent progress of the genome projects makes it easy to determine whether the binding site identified by in vitro selection is located in the promoter region of a target gene. In contrast, other methods for global analysis of gene expression such as microarray and two-dimensional gels may result in the identification of genes or proteins indirectly regulated by a transcription factor as well as those that are directly regulated.
It appears that FruA activates fdgA expression during development by binding regions a and b identified by the footprint analysis. The foot-  print analysis indicates that these regions have similar affinity to FruA-DBD-H 8 in vitro. However, lacZ fusion analysis shows that region b has the most effect on the activation of the fdgA promoter. Because the region hypersensitive to DNase I is located between regions a and b, it is possible that binding of FruA-DBD-H 8 to the fdgA promoter results in structural modification in the fdgA promoter and that this modification is important for the activation of the fdgA promoter. At the early stage of development, fdgA expression is initiated from the P D1 promoter by binding of FruA to regions a and b. When the developmental program is progressed, it is possible that other developmental factors such as a different transcription factor and a factor may activate fdgA expression at P D2 . The induction of fdgA at P D2 is partially dependent on the C-signal, suggesting that the C-signal may be also involved in the activation of FruA-dependent gene expression during late development. M. xanthus produces large amounts of extracellular polysaccharide, often referred to as slime, and excreted slime plays a role in fruiting body development (29). Although similar studies have not been done for sporulation during fruiting body development, the increase in polysaccharide is ϳ200%, and polysaccharide accounts for 20% of the dry weight of the cell during sporulation induced by glycerol (30). The major monosaccharide components of the hydrolyzed extracellular polysaccharide are glucose, mannose, and rhamnose, whereas the nature of unhydrolyzed polysaccharide remains to be examined. It is not known which polysaccharides are excreted by vegetative cells and which are involved in fruiting body development. Thus, characterization of polysaccharide by using the ⌬fdgA strain may answer those questions, because the fdgA gene encodes an OMA family protein involved in the polysaccharide export system, and fdgA expression is differentially regulated during vegetative growth and development. In addition, at least two homologues of FdgA were identified from the preliminary M. xanthus genome sequence (data not shown). The analysis of these homologues may also allow us to understand the roles of polysaccharide in M. xanthus.
It has been shown that both OMA proteins and ABC transporter proteins are required for the polysaccharide export system in Gramnegative bacteria (25). The M. xanthus sasA locus, required for lipopolysaccharide O-antigen biosynthesis, contains rfbA, rfbB, and rfbC (16,17). The rfbA and rfbB genes encode an ABC transporter (17). Furthermore, the fdgA and the sasA mutants show similar phenotypes in that fruiting body formation is impaired and sporulation efficiency is reduced during development (16). Thus, it is tempting to speculate that fdgA and rfbAB gene products may function in the same pathway during development. In addition, the expression of both genes is dependent on fruA during development. However, it appears that sasA expression is differentially regulated from fdgA expression. FruA-DBD-H 8 alone could bind the fdgA promoter but not the sasA promoter in vitro. Because some response regulator proteins are known to bind a target sequence with higher affinity when phosphorylated (31), phosphorylation of FruA may be required to bind the sasA promoter. It is known that the expression of ompF and ompC, encoding outer membrane porins, is differentially regulated by the phosphorylation level of OmpR response regulator in the E. coli osmoregulation system (32). It is also possible that the sasA promoter requires an additional factor for FruA binding that may be C-signal-dependent, because sasA expression is partially dependent on the C-signal. The conditions for FruA to bind both promoters remain to be examined.
The developmental phenotypes of the fruA mutant are more drastic than those of the fdgA mutant. The fruA mutant exhibits little fruiting body formation and yields few spores (8,9). It is likely that FruA regulates genes other than fdgA as shown by two-dimensional gel analysis (14). This is supported by the identification of genes using the same method described here (data not shown). We are currently characterizing these genes. Analysis of genes regulated by FruA should elucidate the signal transduction pathways it controls.