Regulation of the Myxococcus xanthus C-Signal-Dependent Ω4400 Promoter by the Essential Developmental Protein FruA

Abstract

The bacterium Myxococcus xanthus employs extracellular signals to coordinate aggregation and sporulation during multicellular development. Extracellular, contact-dependent signaling that involves the CsgA protein (called C-signaling) activates FruA, a putative response regulator that governs a branched signaling pathway inside cells. One branch regulates cell movement, leading to aggregation. The other branch regulates gene expression, leading to sporulation. C-signaling is required for full expression of most genes induced after 6 h into development, including the gene identified by Tn5 lac insertion Ω4400. To determine if FruA is a direct regulator of Ω4400 transcription, a combination of in vivo and in vitro experiments was performed. Ω4400 expression was abolished in a fruA mutant. The DNA-binding domain of FruA bound specifically to DNA upstream of the promoter −35 region in vitro. Mutations between bp −86 and −77 greatly reduced binding. One of these mutations had been shown previously to reduce Ω4400 expression in vivo and make it independent of C-signaling. For the first time, chromatin immunoprecipitation (ChIP) experiments were performed on M. xanthus. The ChIP experiments demonstrated that FruA is associated with the Ω4400 promoter region late in development, even in the absence of C-signaling. Based on these results, we propose that FruA directly activates Ω4400 transcription to a moderate level prior to C-signaling and, in response to C-signaling, binds near bp −80 and activates transcription to a higher level. Also, the highly localized effects of mutations between bp −86 and −77 on DNA binding in vitro, together with recently published footprints, allow us to predict a consensus binding site of GTCG/CGA/G for the FruA DNA-binding domain.

Understanding how cell-cell signaling coordinates gene expression during multicellular development is a fundamental problem in biology. Myxococcus xanthus is a bacterium that provides an attractive experimental system to elucidate molecular mechanisms of signal transduction leading to developmental gene expression (24, 32). When starved at a high cell density on a solid surface, rod-shaped M. xanthus cells move in a coordinated fashion to form mounds, each containing approximately 10⁵ cells. Within these nascent fruiting bodies, some of the cells differentiate into heat- and desiccation-resistant spherical myxospores. This multicellular developmental process requires several extracellular signals (9, 16). Two of these signals, called A and C, are believed to be understood at the molecular level.

A-signal appears to be amino acids and peptides that result from cleavage of outer membrane proteins by proteases secreted early in development (39, 47). A-signaling is thought to measure cell density (40) and if it is high enough, trigger expression of early developmental genes (4, 38) via a three-component system consisting of a sensor kinase (SasS), response regulator (SasR), and negative regulator (SasN) (reviewed in reference 26).

C-signal appears to be a 17-kDa protein associated with the cell surface and possibly anchored in the outer membrane (31, 42, 50). It is derived by proteolytic cleavage of the 25-kDa product of the csgA gene (31, 42). C-signaling requires cell-cell contact (28, 29, 31), but a receptor has not yet been identified.

FruA plays a key role in the C-signal transduction pathway, governing branches that control motility and gene expression leading to sporulation (10, 46, 51, 53). FruA is similar to response regulators in the FixJ subfamily (10, 46), but its putative cognate sensor kinase has not been identified. The N-terminal domain of FruA has a putative phosphorylation site, D59, and this residue is critical for function, suggesting that phosphorylation activates FruA (10). The molecular mechanism by which activated FruA would regulate motility is unknown, but somehow C-signaling increases methylation of FrzCD (51), a cytoplasmic chemoreceptor homolog that controls the frequency of reversal of the direction of gliding cell movement (3, 45). A low level of C-signaling results in rippling, the coordinated movement of cells that gives the appearance of traveling waves in movies made by time-lapse microscopy, while a higher level of C-signaling causes cell streaming, resulting in mound formation (reviewed in references 23 and 52). FruA's C-terminal domain has a putative helix-turn-helix motif (46) and has recently been shown to bind to specific DNA sequences upstream of two developmentally regulated M. xanthus promoters (55, 56). The results suggest that FruA directly activates transcription of these genes. One of these genes is expressed independently of C-signaling (55), and the other appears to exhibit partial dependence on C-signaling from one of its three promoters (56).

Other potential direct targets of FruA transcriptional activation in response to C-signaling have been identified using Tn5 lac, a transposon that fuses transcription of the Escherichia coli lacZ gene to the promoter of a gene into which it inserts (33-35). One such insertion, at site Ω4400 in the M. xanthus chromosome, fuses lacZ expression to a promoter induced at about 6 h into development (34). Expression from the Ω4400 promoter is reduced about two- to threefold in a csgA mutant that is defective in producing C-signal (34), but expression can be restored by codeveloping the csgA mutant with wild-type cells, which supply C-signal (6). Therefore, activity of the Ω4400 promoter depends partially on extracellular C-signaling. A detailed mutational analysis of the Ω4400 promoter region revealed two regulatory elements upstream of the −35 region (58). DNA between bp −63 and the promoter −35 region was found to be absolutely essential for expression. This element contains two DNA sequences that are also found upstream of other developmentally regulated M. xanthus promoters (11, 57): a C box (consensus sequence CAYYCCY, in which Y means C or T) centered at bp −49, and a 5-bp element (consensus sequence GAACA) centered at bp −61. The second important regulatory element is located slightly farther upstream and does not appear to be found in other promoter regions examined so far. Mutations between bp −86 and −81 decrease Ω4400 promoter activity two- to fourfold and eliminate or reduce dependence on C-signaling (58). How any of these regulatory elements stimulate Ω4400 promoter activity was unknown.

Here, we show that fruA is absolutely required for Ω4400 expression in vivo and that the C-terminal FruA DNA-binding domain binds in a sequence-specific manner to DNA upstream of the Ω4400 promoter in vitro. Mutations between bp −86 and −77 impair DNA binding in vitro, suggesting that FruA binds near bp −80 and activates Ω4400 transcription in response to C-signal during M. xanthus development. In support of this hypothesis, we devised a procedure for chromatin immunoprecipitation (ChIP) analysis of developing M. xanthus cells and used it to show that FruA associates with the Ω4400 promoter region in vivo. Moreover, we found that FruA was associated with the Ω4400 promoter region even in the absence of C-signaling. The results support a model in which FruA directly activates transcription from the Ω4400 promoter both before and after C-signaling. This model explains partial dependence of Ω4400 expression on C-signaling and may apply to other genes induced after 6 h into M. xanthus development, because many of these also exhibit partial dependence on C-signaling (30, 34, 41, 56).

MATERIALS AND METHODS

Bacterial strains and plasmids.

Strains and plasmids that were used in this study are listed in Table 1.

TABLE 1.

Bacterial strains and plasmids used in this study

Bacterial strain or plasmid	Description	Reference or source
E. coli strains
DH5α	φ80 ΔlacZΔM15 ΔlacU169 recA1 endA1 hsdR17 supE44 thi-1 gyrA relA1	17
BL21(DE3)	F−ompT hsdSB(rB− mB−) gal dcm with DE3, a λ prophage carrying the T7 RNA polymerase gene	Novagen
EDYFruA	BL21(DE3) containing pET11a/FDBD-H8	This work
M. xanthus strains
DZF1	sglA1	22
TF786	sglA1 fruA::TnV Ω786	13
MDY4400.DZF1	sglA1 attB::pJB40030	This work
MDY4400.FA	sglA1 fruA::TnV Ω786 attB::pJB40030	This work
MDY1727.DZF1	sglA1 attB::pREG1727	This work
MDY1727.FA	sglA1 fruA::TnV Ω786 attB::pREG1727	This work
DK1622	Wild type	25
DK5208	csgA::Tn5-132 Ω205	49
Plasmids
pET11a/FDBD-H8	pET11a with a gene encoding FruA-DBD-His8 under control of a T7 RNA polymerase promoter	56
pEE512	Apr vector with a gene encoding His6-FruA under control of a T7 RNA polymerase promoter	E. Ellehauge and L. Søgaard-Andersen
pREG1727	KmrlacZ transcriptional fusion vector for integration at Mx8 attB in the M. xanthus chromsome	12
pJB40029	pGEM7Zf with Ω4400 DNA from bp −101 to +155	58
pJB40030	pREG1727 with Ω4400 DNA from bp −101 to +155 fused to lacZ	58
pDY69	pJB40029 with GTC-to-TGA mutation at bp −86 to −84	58
pDY79	pJB40029 with GGGGGTG-to-TTTTTGT mutation at bp −83 to −77	58
pDY67	pJB40029 with TG-to-GT mutation at bp −76 to −75	58
pDY65	pJB40029 with GGGAGC-to-TTTCTA mutation at bp −69 to −64	58
pDY35	pJB40029 with GAAC-to-TCCA mutation at bp −63 to −60	58
pDY59	pJB40029 with GTCCC-to-TGAAA mutation at bp −58 to −54	58
pGV4400.1	pJB40029 with CATCCCT-to-ACGAAAG mutation at bp −52 to −46	58
pDY71	pJB40029 with GGCGG-to-TTATT mutation at bp −45 to −41	58

Growth and development.

E. coli strains containing plasmids were grown at 37°C in Luria-Bertani (LB) medium (48) containing 200 μg of ampicillin (Ap) per ml. M. xanthus strains were grown at 32°C in CTT broth or on agar (1.5%) plates (19) (1% Casitone, 10 mM Tris-HCl [pH 8.0], 1 mM KH₂PO₄-K₂HPO₄, 8 mM MgSO₄ [final pH, 7.6]). When necessary, 40 μg of kanamycin (Km) per ml was used for selection. Fruiting body development was performed on TPM agar (1.5%) plates (10 mM Tris-HCl [pH 8.0], 1 mM KH₂PO₄-K₂HPO₄, 8 mM MgSO₄ [final pH, 7.6]) as described previously (35).

Construction of M. xanthus strains and determination of lacZ expression during development.

Strains containing pREG1727 derivatives integrated at the Mx8 phage attachment site were constructed by electroporation (27) of M. xanthus, and transformants were selected on CTT-Km plates. To identify colonies with unusual developmental lacZ expression, at least 10 transformants were screened on TPM agar plates containing 40 μg of 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside per ml. Any colonies with unusual expression of lacZ were discarded, and of the remaining candidates, three independent isolates were chosen for analysis. In all cases, the three transformants gave similar results when developmental β-galactosidase activity was measured as described previously (35).

Preparation of FruA-DBD-His₈ and His₆-FruA proteins.

To prepare the FruA DNA-binding domain fused to an octahistidine tag (FruA-DBD-His₈), we used the following methods, which are based on work described previously (56). E. coli BL21(DE3) (54) was transformed with pET11a/FDBD-H₈ to create strain EDYFruA. A single Ap-resistant colony was used to inoculate 5 ml of LB-Ap broth. This culture was grown overnight at 37°C with shaking. One ml of overnight culture was used to inoculate 100 ml of LB-Ap broth. The culture was grown to 80 to 90 Klett units (about 4 × 10⁸ to 4.5 × 10⁸ cells/ml) at 37°C, at which time isopropyl-1-thio-β-d-galactopyranoside (IPTG) was added to a final concentration of 1 mM. Cells were harvested after 2 h by centrifugation (5,900 × g, 25°C, 15 min). The cell pellet was resuspended in 25 ml of 10 mM Tris-HCl (pH 7.9), and cells were collected again by centrifugation. All remaining steps were performed on ice or at 4°C. The cell pellet was resuspended in 0.5 ml buffer H (10 mM Tris-HCl [pH 7.9], 500 mM NaCl, 1 mM β-mercaptoethanol, 10 mM imidazole) supplemented with protease inhibitors (Roche Mini EDTA-free tablets) and sonicated six times for 10 s to disrupt cells. After ultracentrifugation at 100,000 × g at 4°C for 30 min, Triton X-100 was added to the supernatant to a final concentration of 0.1%. Ni-nitrilotriacetic acid beads (QIAGEN) were prepared by washing three times with buffer H supplemented with protease inhibitors and added at a 1:1 ratio to the supernatant. The Ni-nitrilotriacetic acid bead-supernatant mixture was rotated for 45 min and applied to a Cell Thru 2-ml disposable column (Clontech) that had been rinsed with 1 ml of buffer H supplemented with protease inhibitors. The beads were washed with 4 ml of buffer H containing protease inhibitors and 0.1% Triton X-100, followed by 4 ml of buffer H containing protease inhibitors and 40 mM imidazole. Fru-DBD-His₈ was eluted from the column in 4 ml of buffer (10 mM Tris-HCl [pH 7.9], 50 mM NaCl, 1 mM β-mercaptoethanol, 250 mM imidazole) and dialyzed against a buffer containing 10 mM Tris-HCl (pH 7.9), 50 mM NaCl, 1 mM β-mercaptoethanol, 1 mM EDTA, and 50% (wt/vol) glycerol. Uninduced culture to which no IPTG had been added was processed in parallel with the induced culture. Both uninduced and induced protein preparations were stored at −20°C. Protein concentration was estimated using the Bradford method (5).

His₆-FruA was prepared by transforming E. coli BL21(DE3) with pEE512 (a gift from E. Ellehauge and L. Søgaard-Andersen). The plasmid encodes N-terminally hexahistidine-tagged, full-length FruA under the control of an IPTG-inducible T7 RNA polymerase promoter. The methods described above were used for culture growth, IPTG induction, and protein purification.

Gel shift assay.

DNA fragments spanning the Ω4400 promoter region from bp −101 to +25, bp −101 to −41, and bp −41 to +25 were generated by PCR using wild-type or mutant plasmid (Table 1) as the template and combinations of the following primers: 5′-CCTAAGCTTTGCACTGCGACGCGAGTC-3′ (for −101), 5′-GCGGATCCCGGTCCTTCGCGTCGCCG-3′ (for +25), 5′-CCGCCAGGGATGTGGGACTGTT-3′ (for −41 in combination with the upstream primer ending at −101), and 5′-CCGGAGGCGCGAGGCGC-3′ (for −41 in combination with the downstream primer ending at +25). The DNA fragments were purified using a PCR purification kit (QIAGEN) and labeled with [γ-³²P]ATP using T4 polynucleotide kinase (New England BioLabs). The labeled DNA fragments were purified by electrophoresis on a 15% polyacrylamide gel, excision after visualization by autoradiography, and elution by soaking overnight at 37°C in 250 μl of TE (10 mM Tris-HCl [pH 8.0], 1 mM EDTA). The concentration of the labeled DNA fragments was estimated by agarose gel electrophoresis with DNA fragments of known concentration and staining of the gel with ethidium bromide.

Binding reaction mixtures (10 μl) contained 10 mM Tris-HCl (pH 8.0), 50 mM KCl, 1 mM dithiothreitol, 10% glycerol, 10 μg/ml bovine serum albumin, 1 μg of double-stranded poly(dI-dC) (Amersham Biosciences, Inc.), and 1 ng of ³²P-labeled DNA fragment. Unless specified otherwise, 2 μg of protein prepared from cells induced to produce FruA-DBD-His₈ was added per reaction mixture. Reaction mixtures were incubated at 30°C for 10 min and loaded on a 5% polyacrylamide gel. After electrophoresis using 0.5× TBE (45 mM Tris-borate [1:1], 1 mM EDTA), gels were dried and exposed to X-ray film for autoradiography and/or bands were quantified with a Storm 820 PhosphorImager (Molecular Dynamics) using ImageQuant software (Amersham Biosciences, Inc.).

Preparation of His₆-FruA antibodies.

Purified His₆-FruA was dialyzed against phosphate-buffered saline to decrease the glycerol concentration, which interferes with emulsion formation. Purified protein (100 μg) was mixed with an equal volume of TiterMax adjuvant (CytRx) and emulsified using a double-hub syringe. Rabbits were injected subcutaneously three times over the course of 3 months, and blood was collected at intervals for preparation of serum as described previously (18).

Western blot analysis.

M. xanthus DZF1 or the fruA null mutant TF786 was grown and subjected to starvation conditions to induce development. At 2 and 18 h into development, approximately 5 × 10⁸ cells were collected and boiled in 100 μl of sample buffer (0.125 M Tris-HCl [pH 6.8], 5% β-mercaptoethanol, 2% sodium dodecyl sulfate (SDS), 10% glycerol, 0.1% bromophenol blue) for 5 min. Equal volumes (8 μl) were loaded on SDS-12% polyacrylamide gels and electrophoresed. Proteins were electrotransferred to a polyvinylidene difluoride membrane (Millipore). The membrane was incubated with TBS (20 mM Tris-HCl [pH 7.5], 500 mM NaCl) containing 5% nonfat dry milk for 1 h at room temperature to block nonspecific binding of the primary antibodies to the membrane. The membrane was then probed by incubating with a 1:5,000 dilution of polyclonal His₆-FruA antibodies in 10 ml of TBS-2% nonfat dry milk. Detection of immunocomplexes was performed using goat anti-rabbit immunoglobulin G-horseradish peroxidase (Bio-Rad) and chemiluminescence (Western Lightning; Perkin-Elmer) according to the manufacturer's recommendations.

Chromatin immunoprecipitation.

M. xanthus DZF1, fruA null mutant TF786, or csgA null mutant DK5208 was grown to 100 Klett units (about 5 × 10⁸ cells/ml) in 10-ml cultures, centrifuged at 12,000 × g for 10 min, and resuspended to about 5 × 10⁹ cells/ml in TPM. The cell suspension was spotted (20 μl) on TPM agar and allowed to dry briefly at room temperature and then incubated at 32°C for either 2 or 18 h. Cells were collected by scraping and dispersed into 10 ml of SHP (10 mM sodium phosphate [pH 7.3]) containing 1% formaldehyde for 30 min at 32°C with shaking to cross-link proteins to DNA. Glycine (1.1 ml of a 1.25 M stock) was added to stop the cross-linking. The sample was centrifuged (12,000 × g at 25°C for 10 min), and the cell pellet was washed with 25 ml of SHP and then centrifuged again. The cell pellet was resuspended in 1 ml of cold (4°C) IP150 (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 0.5% Triton X-100) supplemented with 40 μM pepstatin, 1 mM EDTA, 4 mM Pefabloc, 20 μM leupeptin, 28 μM E-64, and 1.5 μM aprotinin, and then sonicated six times for 10 s on ice to disrupt cells and shear chromosomal DNA into fragments ranging from about 500 to 1,000 bp. Insoluble material was pelleted in a microcentrifuge at 14,000 rpm for 10 s. A sample (20 μl) of the supernatant, representing 2% of the total cellular extract, was removed and set aside as the input DNA sample. The remaining supernatant was rotated for 1 h at 4°C with an equal volume of a 50% slurry (in IP150) of protein A-Sepharose beads (Amersham Biosciences, Inc.) that had been washed three times in IP150. The beads were removed by microcentrifugation twice at 14,000 rpm for 1 min, and the resulting supernatant was divided into samples of equal volume. To one sample, 2 μl of preimmune sera was added. To the other, 2 μl of His₆-FruA antibodies was added. The samples were then rotated for 18 h at 4°C. Samples were microcentrifuged at 14,000 rpm for 5 min at 4°C, each supernatant was placed in a fresh tube, to each tube 50 μl of protein A-Sepharose beads was added, and the sample was rotated at 4°C for 1 h. Beads were collected by microcentrifugation at 500 rpm for 1 min and then washed (15 min with rotation) twice with IP150 and twice with IP300 (50 mM Tris-HCl [pH 8.0], 300 mM NaCl, 0.5% Triton X-100). To the washed beads and to the input DNA samples, 150 μl of elution buffer (50 mM Tris-HCl [pH 8.0], 10 mM EDTA, 1% SDS) was added, and the samples were incubated at 65°C for 4 h to reverse cross-links. Samples were microcentrifuged at 500 rpm for 1 min, supernatants were removed to fresh tubes, and proteins were digested for 30 min at 37°C by adding 150 μl of TE (pH 8.0) containing 100 μg/ml proteinase K. DNA was then purified by phenol-chloroform-isoamyl alcohol (25:24:1) extraction followed by ethanol precipitation after the addition of glycogen to a final concentration of 0.27 μg/ml. The dried DNA was resuspended in 30 μl of sterile water.

Tenfold serial dilutions were made of the input DNA samples, representing 0.2, 0.02, 0.002, or 0.0002% of the total cellular extract for each condition. PCR mixtures (25 μl) contained 10 pmol of each primer and 5 μl of template under standard conditions for Taq DNA polymerase as specified by the manufacturer (New England BioLabs). The number of cycles was optimized for each primer pair. For the Ω4400 promoter region, the −101 to +25 primers are listed above. The rpoC primer sequences were as follows: 5′-CCTTGAGCGCGATGGAGATA-3′ and 5′-CTCGGCGGCCTCATCGAC-3′. These primers allow the amplification of a 125-bp fragment starting at bp +1780 into the putative coding region of M. xanthus rpoC. The results of the PCRs were visualized after electrophoresis of 8 μl on a 1.5% agarose gel and staining with ethidium bromide.

RESULTS

A fruA null mutation abolishes Ω4400 expression.

To determine the effect of fruA on Ω4400 expression in vivo, fruA null mutant and wild-type strains of M. xanthus were transformed with a plasmid containing the Ω4400 promoter region from bp −101 to +155 transcriptionally fused to the E. coli lacZ gene. The plasmid also contained a DNA segment from phage Mx8, which allows efficient site-specific recombination into the M. xanthus chromosome. As negative controls, strains bearing the vector with promoterless lacZ were also constructed. Expression of the lacZ reporter was measured as β-galactosidase specific activity in extracts of cells harvested at different times after placement on starvation agar to induce development. The fruA mutation abolished developmental lacZ expression from the Ω4400 promoter (Fig. 1). We conclude that fruA is essential for Ω4400 expression in vivo.

FIG. 1.

Effect of a fruA null mutation on Ω4400 expression. Developmental lacZ expression was determined for wild-type M. xanthus DZF1 bearing the Ω4400 promoter region (bp −101 to +155) (filled squares) or no promoter (open squares) fused to lacZ and the corresponding fruA mutant strains with the Ω4400 promoter region (filled circles) or no promoter (open circles) fused to lacZ. Points show the average β-galactosidase specific activities of three different transformants for each plasmid in nanomoles of o-nitrophenyl phosphate per minute per milligram of protein, and error bars depict 1 standard deviation of the data.

The FruA DNA-binding domain binds to the Ω4400 promoter region.

To test whether FruA might bind directly to Ω4400 promoter region DNA, gel shift analyses were performed with a FruA DNA-binding domain purified from E. coli. The FruA DNA-binding domain (amino acids 152 to 229) fused to an octahistidine tag (FruA-DBD-His₈) was expressed from a plasmid under the control of an IPTG-inducible promoter. As a negative control, E. coli bearing the same plasmid but not induced with IPTG was subjected to the same purification protocol. A protein of the expected size for FruA-DBD-His₈ was greatly enriched in the preparation from IPTG-induced E. coli compared to the preparation from uninduced cells (Fig. 2A). In gel shift assays with Ω4400 DNA (bp −101 to +25), the preparation enriched for FruA-DBD-His₈ formed a complex with reduced mobility, whereas the preparation from uninduced cells did not produce a shifted complex (Fig. 2B). We infer from these results that FruA-DBD-His₈ binds to Ω4400 promoter region DNA. Formation of a discretely migrating shifted complex in the presence of a 1,000-fold excess of nonspecific competitor DNA [poly(dI-dC)] is suggestive of specific binding. As a control, a 125-bp DNA fragment amplified from near the center of the M. xanthus rpoC gene, which is predicted to encode the β′ subunit of RNA polymerase, was also used in gel shift experiments. No shifted complex was observed (Fig. 2B), indicating that the binding of FruA-DBD-His₈ to the Ω4400 promoter region is specific.

FIG. 2.

FruA-DBD-His₈ binds specifically to the Ω4400 promoter region. (A) Proteins were purified from IPTG-induced or uninduced cultures of E. coli EDYFruA containing pET11a/FDBD-H₈ as described in Materials and Methods. An equal volume (8 μl) of each preparation was electrophoresed on an SDS-14% polyacrylamide gel and stained with Coomassie brilliant blue. The arrow denotes FruA-DBD-His₈ in the sample from induced cells. (B) Gel shift assays were performed with ³²P-labeled DNA from the Ω4400 promoter region (bp −101 to +25) or a 125-bp fragment that starts 1,780 bp downstream of the M. xanthus rpoC putative start codon. The leftmost lane in each panel shows ³²P-labeled DNA with no addition of protein to the binding reaction mixture. Other lanes show addition of 0.02, 0.2, or 2 μl of protein preparation from IPTG-induced or uninduced cells.

FruA-DBD-His₈ binds to upstream DNA in the Ω4400 promoter region.

To determine the location of FruA-DBD-His₈ binding in the Ω4400 promoter region, gel shift assays were performed with shorter DNA segments. The region from bp −101 to +25 is sufficient for full Ω4400 expression during development (58). This region was split in two at bp −41, just upstream of the promoter −35 region. The bp −101 to −41 DNA fragment was bound by FruA-DBD-His₈ to a similar extent as the bp −101 to +25 fragment, but no shifted complex was observed with the bp −41 to +25 fragment (Fig. 3). These results further demonstrate the specificity of FruA-DBD-His₈ binding and localize it to the region upstream of the Ω4400 promoter.

FIG. 3.

FruA-DBD-His₈ binds upstream of the Ω4400 promoter. Gel shift assays were performed with ³²P-labeled DNA spanning the indicated regions relative to the start site of Ω4400 transcription. In the indicated lanes, the protein preparation enriched for FruA-DBD-His₈ was added to the binding reaction mixture. The autoradiographs represent typical results from several experiments.

FruA-DBD-His₈ binds to a region responsible for C-signal-dependent expression from the Ω4400 promoter.

To identify the sequence to which FruA-DBD-His₈ binds in the region upstream of the Ω4400 promoter, gel shift assays were performed with DNA fragments from bp −101 to +25 bearing different mutations. The mutations were all multiple-base-pair transversions (T↔ G and C↔ A) whose effects on developmental expression were measured previously (Fig. 4A). In contrast to the wild-type DNA fragment and all the other fragments with mutations, those mutated at bp −86 to −84 and bp −83 to −77 were bound very little by FruA-DBD-His₈ (Fig. 4B). Quantification of the results from four experiments indicated that these mutations reduced binding three- to fivefold (Fig. 4C). We conclude that FruA-DBD-His₈ recognizes sequences in the bp −86 to −77 region upstream of the Ω4400 promoter. Previously, the region from bp −86 to −81 was shown to mediate, at least in part, C-signal-dependent expression from the Ω4400 promoter (58). Since FruA boosts developmental expression from the Ω4400 promoter in response to C-signal (6, 34), the results taken together suggest it does so by binding to the bp −86 to −81 region and activating transcription.

FIG. 4.

FruA-DBD-His₈ binds to the bp −86 to −77 region, which is important for expression from the Ω4400 promoter. (A) Summary of mutational effects on Ω4400 developmental expression. Promoter regions with the indicated mutations (alternately underlined or boxed) were fused to a lacZ reporter, and developmental β-galactosidase production was measured as described previously (58). Downward arrows represent decreased expression, and numbers indicate the maximum β-galactosidase specific activity observed for the mutant promoter region, expressed as a percentage of the maximum observed for the wild-type promoter in the same experiment. Dashed lines above bp −86 to −81, −72 to −67, and −58 to −53 indicate potential binding sites for FruA-DBD-His₈ based on the consensus sequence presented below in Fig. 7. (B) Gel shift assays were performed with ³²P-labeled DNA fragments (bp −101 to +25) generated by PCR using wild-type or mutant template DNA. In the indicated lanes, the protein preparation enriched for FruA-DBD-His₈ was added. (C) The intensities of the shifted complex and the unbound DNA were quantified as described in Materials and Methods. The ratio for each mutant DNA fragment was normalized to that observed for the wild type in the same experiment. The graph shows the mean and 1 standard deviation of the data from four experiments.

FruA associates with the Ω4400 promoter region in vivo.

To assess whether FruA becomes associated with the Ω4400 promoter region in developing M. xanthus cells, ChIP was performed using antibodies created against full-length His₆-FruA protein. ChIP has not been reported previously for M. xanthus. Therefore, we developed a protocol to cross-link proteins to DNA in M. xanthus cells harvested at different times during development, immunoprecipitate FruA accompanied by the DNA to which it is cross-linked, and analyze the DNA by PCR after reversal of the cross-links (see Materials and Methods).

The fruA gene is expressed starting at about 6 h into development, reaching a peak at 18 to 24 h (10, 46). Expression from the Ω4400 promoter shows a similar pattern during development (Fig. 1). M. xanthus wild-type DZF1 and a fruA null mutant derivative were allowed to develop for either 2 or 18 h, and samples were collected for Western blotting and ChIP analyses. As expected, FruA was not detected in the 2-h samples and had accumulated by 18 h in the wild type but not the fruA mutant (Fig. 5A).

FIG. 5.

FruA associates with the Ω4400 promoter region in vivo. (A) Equal volumes of cellular extracts from M. xanthus wild-type DZF1 or a fruA null mutant, which had been collected at 2 or 18 h into development, were electrophoresed on an SDS-14% polyacrylamide gel and subjected to Western blot analysis using antibodies generated against His₆-FruA (see Materials and Methods). (B) ChIP was performed on cellular extracts from M. xanthus wild-type DZF1 or a fruA null mutant. Cells were collected at 2 or 18 h into development, and protein-DNA complexes were immunoprecipitated with preimmune serum or polyclonal antibodies against full-length His₆-FruA. Input samples correspond to DNA purified from 0.2%, 0.02%, 0.002%, or 0.0002% of the total cellular extract prior to immunoprecipitation. The top panels show PCR with primers designed to amplify the Ω4400 promoter region (bp −101 to +25 relative to the start site of transcription), and the bottom panels show PCR with primers designed to amplify the rpoC coding region (bp +1780 to +1905 relative to the predicted translation start).

Purified DNA resulting from the ChIP protocol was used as a template in PCRs with primers designed to amplify either the Ω4400 promoter region or the center of the rpoC coding region (as a negative control). To show that the PCR conditions were in the linear range of amplification for each primer set, a 10-fold dilution series of input DNA (i.e., DNA purified from a sample of the cellular extract prior to immunoprecipitation) was used as a template in parallel PCRs (Fig. 5B, lanes 1 to 4, 7 to 10, 13 to 16, and 19 to 22).

The top panels in Fig. 5B show the results for the region surrounding the Ω4400 promoter. ChIP analysis of wild-type cells harvested at 2 h into development showed little or no difference in signal intensity, comparing preimmune serum (lane 5) with antibodies against FruA (lane 6), as expected since FruA is absent at this time in development (Fig. 5A). In contrast, ChIP analysis of the wild type harvested at 18 h, when FruA is present (Fig. 5A), showed much greater signal intensity with FruA antibodies (Fig. 5B, lane 12) than with preimmune serum (lane 11). This difference was observed in five additional experiments, while no such difference was observed for cells harvested at 2 h. Also, no such difference was observed when comparing preimmune serum versus FruA antibodies at 2 or 18 h for the fruA null mutant (lanes 17, 18, 23, and 24). The bottom panels show the ChIP analysis of the rpoC coding region. For both the wild type and the fruA mutant, at both 2 and 18 h into development, the intensity of the signal with preimmune serum versus FruA antibodies was similar. We conclude that FruA is associated with the Ω4400 promoter region at 18 h into development but not with DNA in the vicinity of the rpoC coding region.

FruA associates with the Ω4400 promoter region in the absence of C-signaling.

Expression from the Ω4400 promoter is absolutely dependent on fruA (Fig. 1) but only partially dependent on C-signaling (6, 34). In a csgA null mutant unable to produce C-signal, Ω4400 expression is reduced two- to threefold. The absolute dependence of Ω4400 expression on fruA might reflect a direct or indirect role of FruA in the absence of C-signaling. To distinguish between these possibilities, we performed ChIP analysis on a csgA null mutant at 18 h into development. Under the conditions of development we used (starvation on TPM agar), C-signaling first affects the pattern of gene expression at about 6 h into development (34), and under submerged culture conditions of development, which like TPM agar causes abrupt starvation and similar timing of developmental gene expression, the level of CsgA protein that mediates C-signaling reaches a maximum at 12 to 21 h into development (15, 36). Greater signal intensity was observed with FruA antibodies than with preimmune serum (Fig. 6). This difference was observed in two additional experiments. Similar results were observed for wild-type DK1622 (data not shown), the strain from which the csgA null mutant was derived. We conclude that FruA is associated with the Ω4400 promoter region at 18 h into development in the absence of C-signaling. This suggests that FruA directly activates transcription from the Ω4400 promoter in csgA null mutant cells, accounting for moderate Ω4400 expression.

FIG. 6.

FruA associates with the Ω4400 promoter region in the absence of C-signaling. ChIP was performed on cellular extracts from M. xanthus strain DK5208, a csgA null mutant unable to produce C-signal. Cells were collected at 18 h into development, and protein-DNA complexes were immunoprecipitated with preimmune serum or polyclonal antibodies against full-length His₆-FruA. Input samples correspond to DNA purified from 0.2%, 0.02%, 0.002%, or 0.0002% of the total cellular extract prior to immunoprecipitation. PCR was performed with primers designed to amplify the Ω4400 promoter region (bp −101 to +25 relative to the start site of transcription).

DISCUSSION

Our results provide new insight into the regulation of developmental gene expression by FruA. We have shown that expression of Ω4400, which depends partially on C-signaling, absolutely requires fruA in vivo. The FruA DNA-binding domain bound specifically to a region upstream of the Ω4400 promoter, and binding was impaired by mutations between bp −86 and −77. These mutations were shown previously to reduce Ω4400 expression two- to fourfold, and one of these mutations rendered expression independent of C-signaling (the other was not tested) (58). Taken together, these results suggest that (i) FruA binds near bp −80 and activates Ω4400 transcription in response to C-signal during M. xanthus development and (ii) the portion of Ω4400 expression that does not depend on C-signaling nevertheless depends on FruA. We developed antibodies and a protocol for ChIP experiments and found that FruA is associated with the Ω4400 promoter region during M. xanthus development, even in the absence of C-signaling. Therefore, FruA appears to directly activate Ω4400 transcription moderately prior to C-signaling and more vigorously in response to C-signal.

Regulation of Ω4400 expression shares several features with regulation of fdgA, a gene identified recently by virtue of its ability to be bound by FruA-DBD-His₈ (56). FdgA is predicted to be involved in polysaccharide export. The fdgA gene appears to be transcribed from three promoters, one active during vegetative growth (P_V) and two active during development (P_D1 and P_D2). P_V activity was fruA independent, but like the Ω4400 promoter, P_D1 and P_D2 appeared to be inactive in a fruA mutant. FruA-DBD-His₈ bound from bp −89 to −64 relative to the inferred transcriptional start site for P_D1, which is 81 bp upstream of that for P_D2. Hence, FruA likely binds at a similar position relative to P_D1 and the Ω4400 promoter, but farther upstream of P_D2. Interestingly, however, P_D1 activity was C-signal independent, whereas P_D2 activity appeared to be reduced in a csgA mutant defective in C-signaling. Therefore, P_D2 behaved more like the Ω4400 promoter with respect to C-signal dependence, suggesting that FruA can activate transcription in response to C-signal from positions ranging at least from about 80 to 160 bp upstream of the transcriptional start site.

Like the fdgA P_D1 promoter, other fruA-dependent but csgA-independent promoters have been described or inferred in M. xanthus (20, 21, 34, 46). Among these, only the dofA promoter region has been shown to be bound by FruA-DBD-His₈ (55). In this case, two regions, from bp −82 to −67 and from bp −57 to −42, were protected from DNase I digestion. A 5′ deletion to bp −62 decreased promoter activity to 13%, suggesting that FruA binding to the upstream region is important for dofA transcription, independent of C-signaling. Likewise, a 5′ deletion to bp −71 (relative to the inferred transcriptional start site for P_D1) decreased fdgA transcription during development markedly, suggesting that FruA binding to the upstream region is important for both C-signal-independent transcription from P_D1 and C-signal-dependent transcription from P_D2. Horuichi et al. (21) have proposed that a low level of FruA phosphorylation might be sufficient to activate C-signal-independent promoters, whereas a higher level of FruA phosphorylation, resulting from C-signaling, might be required to activate other promoters. While it is attractive to think, based on this model, that binding of phosphorylated FruA near bp −80 could account for C-signal-independent, as well as C-signal-dependent, Ω4400 promoter activity, this is not the case, because a 5′ deletion to bp −73 decreased promoter activity less than twofold (58), while promoter activity was abolished in a fruA mutant (Fig. 1). Taking these observations together with our finding that FruA is associated with the Ω4400 promoter region in the absence of C-signaling (Fig. 6), it seems likely that FruA (possibly phosphorylated at a low level or not phosphorylated at all) binds to one or more sites downstream of bp −73, in addition to binding near bp −80.

Our gel shift analyses with FruA-DBD-His₈ did not detect binding to multiple sites in the Ω4400 promoter region (i.e., a single shifted complex was observed). Perhaps the full-length, phosphorylated form of FruA is necessary for binding to low-affinity sites downstream of bp −73. Phosphorylation enhances DNA binding by other response regulators similar in sequence to FruA (see below). On the other hand, multiple shifted complexes were observed when FruA-DBD-His₈ bound to either the fdgA or dofA promoter region (55, 56). In both cases, a large region upstream of the promoter was protected by FruA-DBD-His₈ from digestion by DNase I. Within these regions, three sequences in the case of fdgA and four in the case of dofA are similar to the GTCGGG sequence from bp −86 to −81 in the Ω4400 upstream region (Fig. 7), which is important for FruA-DBD-His₈ binding (Fig. 4). All these sequences are in the same orientation with respect to the promoter. Based on the sequences shown in Fig. 7, we predict a consensus binding site for Fru-DBD-His₈ of GTCG/CGA/G.

FIG. 7.

Alignment of sequences in FruA-DBD-His₈-binding sites. Sequences upstream of the fdgA and dofA promoters that lie within regions protected by Fru-DBD-His₈ from digestion by DNase I are aligned with the sequence upstream of the Ω4400 promoter, which is important for Fru-DBD-His₈ binding. Numbers indicate the position of each sequence relative to the putative transcriptional start site (for P_D1 in the case of fdgA). A consensus sequence for Fru-DBD-His₈ binding is shown at the bottom. For each position in the consensus sequence, at least six of the eight sequences match the consensus at that position.

Interestingly, two other sequences matching the predicted Fru-DBD-His₈ binding site consensus in five of six positions are found in the same orientation as the GTCGGG sequence in the Ω4400 upstream region (Fig. 4A). One of these, GTGGGG located at bp −72 to −67, could be mutated in either half of its sequence with little or no effect on Ω4400 expression in vivo (58). Also, changing the sequence to GTGTTT did not impair FruA-DBD-His₈ binding in vitro (Fig. 4B and C). The other sequence matching the consensus, GTCCCA located at bp −58 to −53, was critical for expression in vivo (58), but there was little effect on FruA-DBD-His₈ binding in vitro when the sequence was changed to TGAAAA (Fig. 4B and C). The lack of effects on FruA-DBD-His₈ binding in vitro to mutated sequences is consistent with our observation of a single shifted complex which appears to reflect binding of FruA-DBD-His₈ only to the GTCGGG sequence from bp −86 to −81 in vitro. However, as noted above, full-length FruA (possibly phosphorylated at a low level or not phosphorylated at all) might contact a larger region of DNA in vivo that extends farther downstream, and so it remains possible that the loss of Ω4400 promoter activity caused by mutations in the GTCCCA sequence from bp −58 to −53 is due to loss of FruA binding.

FruA is similar to response regulators in the FixJ subfamily (10, 46). Phosphorylation of FixJ by FixL induces dimerization and enhances DNA binding and transcriptional activation (2, 7, 8, 14). FruA's putative cognate sensor kinase has not yet been identified. The C-terminal DNA-binding domain of FruA is similar to that of FixJ, which is structurally similar to that of NarL and Spo0A (37). For each of these well-studied response regulators, phosphorylation of the N-terminal receiver domain enhances DNA binding by the C-terminal helix-turn-helix-containing domain, although the effects on dimerization (i.e., in solution versus on the DNA) can be different and the sequence and arrangement of recognition sites in target promoters are different (1, 37, 43, 44, 59). Identification of a histidine protein kinase capable of producing phosphorylated FruA in vitro is an important goal of future studies in order to better understand the mechanism of transcriptional activation by FruA.

An important contribution of this study is the development of a protocol for ChIP experiments on developing M. xanthus cells. Together with the availability of DNA microarrays, genome-wide location analysis (i.e., ChIP on chip) of FruA-binding sites and other transcription factor-binding sites can now be performed. By applying the ChIP protocol to the Ω4400 promoter region, we discovered that FruA is present even in the absence of C-signaling. Our results imply a direct role of FruA in activating transcription from the Ω4400 promoter both before and after C-signaling during M. xanthus development. This is the first example of a single promoter implicated to be directly activated by FruA via both C-signal-independent and C-signal-dependent mechanisms. Expression of many other M. xanthus genes, like that of Ω4400, depends partially on C-signaling (30, 34, 41, 56). It will be interesting to determine whether FruA directly activates transcription of these genes prior to C-signaling and does so more vigorously after cells engage in C-signaling.