|
|
Proc Natl Acad Sci U S A. 2007 May 8; 104(19): 7969–7974. Published online 2007 April 30. doi: 10.1073/pnas.0701569104. | PMCID: PMC1876556 |
Copyright © 2007 by The National Academy of Sciences of the USA Developmental Biology Combinatorial regulation of genes essential for Myxococcus xanthus development involves a response regulator and a LysR-type regulator Poorna Viswanathan,* Toshiyuki Ueki,† Sumiko Inouye,† and Lee Kroos*‡ *Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI 48824; and †Department of Biochemistry, Robert Wood Johnson Medical School, 675 Hoes Lane, Piscataway, NJ 08854 Received February 21, 2007. Myxococcus xanthus is a bacterium that undergoes multicellular development. C-signaling influences gene expression and movement of cells into aggregates. Expression of the dev operon, which includes genes essential for efficient sporulation, depends in part on C-signaling and reaches its highest level in cells within aggregates, ensuring that spores form within fruiting bodies. Here, an upstream DNA element was found to be essential for dev promoter activity and was bound by FruA, a response regulator in the C-signaling pathway. A second positive regulatory element, located ≈350 bp downstream of the dev transcriptional start site, was bound by LadA, a newly identified transcription factor in the LysR family. Typically, LysR-type transcription factors bind upstream of the promoter and activate transcription in response to a coinducer. LadA appears to activate transcription from an unusual location for a LysR family member and likely subjects dev transcription to a different cue than does FruA. A ladA mutant exhibited similar developmental defects as dev mutants, suggesting that LadA may be devoted to dev regulation, unlike FruA, which regulates many developmental genes. FruA and LadA act on a regulatory region spanning >400 bp to bring about proper temporal and spatial expression of the dev operon, resembling the regulation of developmental genes in multicellular eukaryotes. Keywords: C-signaling, CRISPR, dev operon, FruA, sporulation How multicellular organisms achieve proper temporal and spatial expression of genes during development is a fundamental question. In eukaryotes, the regulatory regions of developmental genes are typically large, containing enhancers that integrate signaling information via binding of multiple transcription factors ( 1). M. xanthus provides an excellent experimental system to investigate gene regulation during a simple multicellular developmental process ( 2). Starvation initiates cell movement to aggregation foci. C-signaling is necessary to complete aggregation ( 3) and changes the frequency with which cells reverse their gliding movement ( 4). C-signaling also influences the expression of genes induced after the early aggregation phase ( 5). Cell contact is required for C-signaling ( 6), which is mediated by the cell-surface-associated CsgA protein ( 7– 9). About 10 5 cells aggregate to form a mound in which cells make many contacts ( 10). This is thought to permit efficient C-signaling in the nascent fruiting body, inducing late genes and differentiation of rod-shaped cells into spherical, dormant spores ( 11, 12). To achieve proper temporal and spatial gene expression during development, M. xanthus employs eukaryotic-like signal transduction proteins and transcription factors ( 13, 14). Are the regulatory regions of M. xanthus developmental genes also eukaryotic-like (i.e., large and complex)? Several have been described, ranging from <100 bp to >1 kb in size ( 11, 15– 22), and in some cases, a single transcription factor has been shown to bind to the regulatory region ( 23– 28). Recently, the regulatory region of the dev operon was characterized ( 29). Multiple positive regulatory elements spanning from >500 bp upstream of the promoter to >580 bp downstream of the transcriptional start site (TSS) are necessary for full expression. Also, at least one negative regulatory element is located between +219 and +280. Moreover, upstream and downstream regulatory elements appear to interact functionally, suggesting contacts via DNA looping. Expression of the dev operon is induced during the aggregation phase of development, and its rise depends in part on C-signaling ( 5, 30). The C-signal dependence of dev expression may explain its higher expression in cells within the nascent fruiting body than in cells outside ( 31). Expression of the dev operon is essential for normal aggregation and spore formation. In-frame deletions in devT ( 32) or devS ( 29) delay aggregation and reduce sporulation to ≈1% of the wild-type level. DevS negatively autoregulates dev transcription ( 29). DevT positively regulates transcription of fruA ( 32), which encodes a putative response regulator that appears to directly activate transcription of several developmentally regulated genes ( 25– 27, 33, 34). The dev operon includes several CRISPR-associated ( cas) genes and at least two repeats of a series of clustered regularly interspaced short palindromic repeats (CRISPR) ( 29). CRISPR-Cas systems are found in nearly half of all sequenced bacterial and archaeal genomes ( 35). They have been proposed to form small interfering RNAs (siRNAs) that inhibit expression of plasmid and bacteriophage genes ( 36, 37). In support of this notion, the 37-bp sequence between the two repeats shown to be part of the dev transcript is complementary to the sense strand of the phage Mx8 integrase gene ( 29), which is necessary for recombination during lysogenization of M. xanthus ( 38). Expression of dev might protect developing cells from lysogen formation ( 29). Here, we identify two transcription factors that bind to the dev regulatory region. FruA binds to a site centered at about −90 bp, apparently activating transcription, possibly in response to C-signaling, and forming a regulatory loop because DevT positively regulates fruA transcription. LadA, which stands for LysR-type activator of dev, binds to a site centered at about +350. Both transcription factors are required and likely respond to different cues, subjecting dev transcription to combinatorial regulation. An Upstream DNA Element Is Essential for dev Promoter Activity, Which Depends on FruA. Recently, the dev promoter was identified, and a 17-bp sequence centered at −91 relative to the TSS was noted that is similar to a sequence centered at −74.5 in the Ω4400 promoter region ( 29). The upstream half of the latter sequence overlaps a site recognized by the DNA-binding domain of FruA ( 27). To test whether upstream DNA including the 17-bp sequence is important for dev promoter activity, we compared expression from dev promoter fragments with 5′ endpoints at −114 or −62. Both fragments had the same 3′ endpoint at +581 fused to a downstream lacZ reporter. Expression was measured during M. xanthus development after integration of the reporter constructs into the chromosome by site-specific recombination at a phage attachment site. In M. xanthus DZF1, expression from the fragment with the 5′ endpoint at −114 increased during development, but that from the fragment with its 5′ endpoint at −62 remained at a low level (), demonstrating that dev promoter activity depends on DNA upstream of the region typically bound by RNA polymerase. Developmental lacZ expression from a reporter inserted into devR ( devR::Tn 5 lac Ω4414) previously was shown to depend on FruA ( 34). To examine the dependence of dev promoter activity on FruA in the presence of a wild-type dev locus, both of the reporter constructs described above were transformed into fruA mutant M. xanthus TF786. Expression from both constructs was abolished in the fruA mutant (), indicating that dev promoter activity completely depends on FruA in a wild-type dev background. FruA Binds to the Upstream DNA Element in the dev Promoter Region. The DNA-binding domain of FruA (FruA-DBD-His 8) has been purified and shown to bind specifically to sites in three M. xanthus promoter regions ( 25– 27). Electrophoretic mobility shift assays (EMSAs) indicated that FruA-DBD-His 8 bound specifically to dev upstream DNA between −114 and −62 ( A) but not to dev DNA between −62 and +31 (data not shown). To further localize the binding, three shorter DNA segments were tested. DNA corresponding to −101 to −75 produced a single shifted complex with FruA-DBD-His 8, whereas no complex was observed for the other two segments ( A). EMSAs with full-length His 6-FruA showed similar results except that less shifted complex was formed [for example, see supporting information (SI) Fig. 7]. We conclude that FruA binds specifically in vitro to a sequence present between −101 and −75 in the dev promoter region. This sequence includes the 17-bp sequence noted above for its similarity to a sequence in the Ω4400 promoter region that is bound by FruA. Mutations in the Upstream DNA Element Impair FruA Binding in Vitro and dev Promoter Activity in Vivo. To further define the FruA-binding site and test its role in dev transcription, we made four mutations in DNA corresponding to −101 to −75 (B). In the context of DNA segments corresponding to −114 to −62, mutations M2 and M3 abolished detectable shifted complex formation by FruA-DBD-His8, and mutations M1 and M4 reduced complex formation to 65 ± 15% and 70 ± 10% of wild type, respectively (C). All four mutations abolished detectable formation of the small amount of shifted complex observed with full-length His6-FruA (data not shown). These results suggest that FruA interacts with the entire region tested and that FruA's DNA-binding domain interacts most strongly with sequences between −94 and −81. Each of the four mutations was also tested for its effect on dev promoter activity in the context of DNA spanning from −114 to +581, fused to lacZ and integrated into the M. xanthus chromosome as described above. However, in this case, the reporter constructs were transformed into a devS null mutant (DK11209) to enhance dev promoter activity by abolishing DevS-mediated negative autoregulation ( 29). All four mutations impaired expression, with M2 reaching only 4% of the wild-type maximum at 48 h into development, M3 reaching 15%, and M1 and M4 each reaching 29% ( D). These results are in qualitative agreement with the FruA-DBD-His 8 DNA-binding assays ( C). Taken together, the results strongly support a model in which FruA binds to an upstream DNA element located between −101 and −75, and activates dev transcription. Identification of a Protein that Binds to a Positive Regulatory Element Downstream of the dev Promoter. Comparison of developmental lacZ expression from fusions with different 3′ endpoints suggested that a positive regulatory element might be present between +280 and +581 relative to the dev TSS ( 29). EMSAs with a DNA fragment spanning from +282 to +456 revealed two shifted complexes using DNA-binding proteins partially purified [ammonium sulfate (AS) fraction] as described previously ( 24) from 12-h-developing M. xanthus DZF1 ( A). Each complex and the unbound probe (as a control) was excised from the gel and subjected to footprinting by treatment with 1,10-phenanthroline-copper followed by electrophoresis on a DNA sequencing gel. Complex I (CI) showed protection from +327 to +373. Complex II (CII) showed protection from +319 to +376 and two hypersensitive sites (HS1 and HS2) suggestive of DNA bending at about +340 and +380 ( B). We conclude that one or more proteins in the AS fraction binds specifically to the region from +319 to +376. The region bound by protein in the AS fraction overlaps the predicted start of cas6, the second gene in the dev operon ( 29). To test whether this region acts as a positive regulatory element, we made fusions to different 3′ endpoints and measured developmental lacZ expression in a devS null mutant (DK11209) as described above. DNA fragments spanning from −114 to +458 or +359 () exhibited a similar pattern of expression as the fragment from −114 to +581 ( D), although the maximum expression reached at 48 h of development was slightly less. In contrast, a fragment from −114 to +326 showed very little expression (). Therefore, DNA between +326 and +359, which is part of the region bound by protein in the AS fraction, acts as a positive regulatory element for dev expression. In further support of this conclusion, a 7-bp mutation of the sequence between +342 and +348 reduced expression substantially, and a 58-bp deletion from +320 to +377 abolished expression when each was tested in the context of the fragment from −114 to +581 ( SI Fig. 8). To purify the protein(s) in the AS fraction that binds to the positive regulatory element, DNA chromatography was performed with magnetic streptavidin beads to which biotinylated DNA corresponding to +282 to +456 was bound (see Materials and Methods). The affinity-purified protein showed a major species with an apparent molecular weight of ≈34 kDa (A). In EMSAs with the dev DNA fragment from +282 to +456, the affinity-purified protein formed CII much more abundantly than CI, especially as compared with the AS fraction (C). The affinity-purified protein failed to form a shifted complex with a DNA fragment having the same endpoints and a 58-bp deletion from +320 to +377 (data not shown). Although the affinity-purified protein formed too little of CI for footprint analysis, the footprint of CII was indistinguishable from that of CII produced by the AS fraction (D). We hypothesized that the 34-kDa affinity-purified protein was responsible for producing CII. To identify the affinity-purified protein, a sample was digested with protease, and the resulting peptides were subjected to mass spectrometric sequence analysis. The peptide sequences matched that predicted for M. xanthus MXAN1402, which is predicted to be a 34.2-kDa protein similar to transcription factors of the LysR family ( 14). Based on the results described below, we have named the protein LadA for LysR-type activator of dev. To test whether LadA binds to the positive regulatory element located downstream of the dev promoter, the ladA gene was PCR-amplified from M. xanthus chromosomal DNA and cloned into a plasmid that permits inducible expression from a T7 RNA polymerase promoter in Escherichia coli. The plasmid was designed to produce LadA, C-terminally tagged with six histidine residues (LadA-His6), to facilitate purification of the protein by metal-affinity chromatography. LadA-His6 purified from E. coli (B) produced CII in EMSAs with dev DNA from +282 to +456 (C). CI was not observed. The footprint of CII produced by LadA-His6 was indistinguishable from that produced by the affinity-purified protein or the AS fraction (D). We conclude that LadA binds to DNA from +319 to +376 downstream of the dev TSS. Because this region acts as a positive regulatory element, and given the similarity of LadA to transcription factors of the LysR family, we propose that LadA activates dev transcription by binding to the downstream site. A Mutation in ladA Impairs dev Expression and Spore Formation. If LadA activates dev transcription as we propose, a null mutation in ladA should reduce dev expression. An internal fragment of ladA was PCR-amplified and cloned into a plasmid unable to replicate in M. xanthus. This plasmid was transformed into wild-type M. xanthus DK1622 with selection for kanamycin resistance conferred by the plasmid. A transformant was identified in which the plasmid had integrated into the chromosome by homologous recombination, disrupting ladA. Because the gene downstream of ladA is in the opposite orientation, the plasmid insertion should not have a polar effect on downstream genes. To examine the effect of ladA disruption on dev expression, the fragment from −114 to +581 bp was fused to lacZ in a plasmid that, after transformation into M. xanthus, integrates into the chromosome by site-specific recombination at a phage attachment site and confers resistance to oxytetracycline. Developmental lacZ expression was abolished in the ladA mutant (). Likewise, the higher level of expression observed in a devS mutant (due to the absence of negative autoregulation) was abolished in a devS ladA double mutant ( SI Fig. 9). These results demonstrate that ladA is essential for developmental expression of dev and, together with our other results, support the idea that LadA activates dev transcription by binding to a site centered ≈350 bp downstream of the TSS. Mutations in the dev operon delay and/or reduce aggregation, and reduce sporulation 100-fold or more, depending on the conditions used for development ( 29, 32, 39). Under the conditions we used, wild-type M. xanthus DK1622 formed loose aggregates by 9 h, mounds with well defined edges by 15 h, and darkening, nascent fruiting bodies by 24 h. The ladA mutant was delayed by ≈6 h for aggregation ( SI Fig. 10), and the nascent fruiting bodies failed to darken even after 72 h, at which time the ladA mutant had made only ≈1% as many spores as wild type. The ladA mutant exhibits similar developmental defects as dev mutants. The aggregation and sporulation defects of the ladA mutant were complemented by a wild-type copy of the gene plus the intergenic region that precedes it, which was cloned in a plasmid that integrated into the chromosome by site-specific recombination at a phage attachment site (data not shown). This demonstrates that disruption of ladA is responsible for the aggregation and sporulation defects of the ladA mutant, and suggests that ladA is transcribed from a promoter in the intergenic region. Our results support a model in which transcription of the dev operon is activated by FruA and LadA binding to sites centered at about −90 and +350, respectively. Absence of either protein eliminates dev expression and impairs aggregation and sporulation. As discussed below, it is likely that activity of FruA and LadA is regulated by different signals, subjecting dev transcription to combinatorial regulation that ensures proper temporal and spatial expression. FruA is similar to proteins in the FixJ subfamily of response regulators and has been proposed to be activated by phosphorylation in response to C-signaling and perhaps an unidentified signal ( 24, 33, 34, 40). FruA's putative cognate histidine protein kinase(s) has not been identified. Of the three promoter regions shown previously to be bound by FruA-DBD-His 8, expression from two ( fdgA and 4400) depends partially on C-signaling, and expression from one ( dofA) is C-signal-independent ( 25– 27). Yet, developmental expression of all three, like that of dev, absolutely depends on FruA. Either FruA has some ability to activate transcription without being phosphorylated or it can be phosphorylated in response to a signal(s) other than C-signal. We favor the latter model because mutational analysis suggests that D59 must be phosphorylated in order for FruA to be active ( 34, 41). Recently a consensus sequence of GTCG/CGA/G was predicted as the binding site for FruA-DBD-His 8 ( 27). The best matches to this consensus in the dev upstream region bound by FruA-DBD-His 8 are GACGGG centered at −99.5 with one mismatch and TTTGGG centered at 86.5 with two mismatches. Both of these sequences contain GGG, a sequence observed in tandem within and adjacent to sites bound by FruA-DBD-His 8, leading us to predict a longer consensus binding site of GGGC/TA/G(N 4–6)C/TGGG ( SI Fig. 11). Sequences that match this consensus are present in the M. xanthus genome at a much higher frequency than expected based on chance (7,155 occurrences, 3,820 expected; P < 10 −6, two-tailed binomial test), suggesting that FruA may bind to many sites in the genome. In agreement with the 5′ half of our predicted consensus, a recent study of full-length FruA-His 8 binding to randomized sequences suggested GGGC/TA/G as a minimum consensus recognition sequence (C.-Y. Xu, T.U., H. Nariya, and S.I., unpublished data). The dev operon is the second direct target of FruA activation known to encode proteins essential for M. xanthus development. The first target to be discovered, fdgA, encodes a protein homologous to outer-membrane auxiliary family proteins involved in polysaccharide export ( 25). A fdgA null mutant formed larger and fewer aggregates that were not as dark, and formed ≈1% as many spores as its M. xanthus DZF1 parent. In contrast, fruA null mutants fail to form aggregates or spores ( 33, 34, 42). The more drastic defects of fruA mutants suggested that FruA regulates additional genes required for development ( 25). Can failure to express the dev operon account for the fruA mutant phenotype? It seems unlikely. A ladA mutant, which fails to express dev (), exhibits less drastic developmental defects than fruA mutants. Like previously characterized dev mutants ( 29, 32, 39), a null mutation in ladA delayed aggregation and reduced sporulation to ≈1% of the wild-type level. FruA likely activates additional genes required for aggregation and sporulation because a fruA mutant showed reduced expression of at least 50 proteins at 12 h into development ( 40). LadA is similar to LysR-type transcriptional regulators in diverse bacteria ( 14). LysR family members regulate genes whose products perform diverse functions (reviewed in ref. 43). Commonly, LysR-type regulators repress transcription from their own promoter and activate transcription from a divergent promoter; however, sometimes target genes are unlinked. When acting as an activator, the LysR-type protein typically binds to a site centered at about −65. In response to a coinducer, a second molecule of the LysR-type protein binds cooperatively to the first and interacts with additional downstream DNA near the target promoter −35 region. Coinducers include extracellular signaling molecules, metabolites, and ions, and are highly specific for a particular LysR-type regulator. Most LysR-type proteins bend the DNA, and some have been shown to interact with the C-terminal domain of the α-subunit of RNA polymerase (reviewed in refs. 43 and 44). Both of these effects may increase transcription initiation. LadA is similar to other LysR-type proteins in its domain organization. LadA's N-terminal 60 aa matches at most positions the consensus sequence for the N-terminal DNA-binding domain of LysR family proteins ( SI Fig. 12). LadA also exhibits a short sequence closer to its C terminus that matches a consensus found in other LysR-type proteins and appears to be involved in DNA binding. LysR family members typically bind to interrupted dyad sequences (reviewed in ref. 43). Near the center of the LadA footprint is the interrupted dyad GATTT-N 13-AAATC (+342 to +364), which may be a recognition sequence for LadA. Although less well defined, the central domain of LysR-type regulators is involved in coinducer recognition and response (reviewed in ref. 43). LadA exhibits high similarity to other LysR-type proteins across its central domain, but it is impossible to predict the nature of the coinducer. The putative LadA coinducer would be intracellular and therefore distinct from at least one of the signals believed to lead to phosphorylation of FruA, the extracellular C-signal. Hence, dev transcription appears to be subject to combinatorial regulation with at least two signal inputs. This is reminiscent of regulation of xpsR, which acts as a signal integrator in the virulence gene regulatory network of the plant pathogen Ralstonia solanacearum ( 45). Transcriptional activation of xpsR involves the LysR-type protein PhcA binding primarily to a site centered at −77 and the response regulator VsrD binding to a site centered at −315. LadA is unusual among LysR-type regulators in that it binds far (350 bp) downstream of the TSS of its target promoter. To our knowledge, only two LysR family members have been shown previously to bind downstream of a target promoter. CatR binds to a site that extends from +162 to +193 in the Pseudomonas putida catBCA operon ( 46). CatR bound to the downstream site mediates 3- to 4-fold repression of catBCA, and this has been proposed to involve interaction with CatR bound upstream of the promoter via DNA looping. MetR binds to a site spanning from +62 to +85 in the Salmonella typhimurium metF gene, in addition to binding at −95 to −50 ( 47). Binding of MetR to both sites is necessary to antagonize MetJ-mediated repression of metF transcription. We propose that LadA, like MetR, functions as an antirepressor. This model is based on results that suggest a negative regulatory element lies between +219 and +280 in the dev operon ( 29). This element reduces dev expression 4.6-fold. Part of LadA's role might be to relieve this negative effect. The LadA-binding site near +350 is dispensable in the absence of DNA downstream of +71 because a dev promoter-containing fragment spanning from −114 to +71 exhibited similar developmental lacZ expression in a devS mutant ( 29) as the fragment from −114 to +581 ( D). Whether LadA itself is dispensable for dev transcription under these conditions and whether LadA binds to other sites in the dev promoter region are important questions for further study. In any case, the dev promoter region is unusual for a bacterium. Only a few examples of promoter regions involving a downstream positive regulatory element have been described in M. xanthus ( 22, 48) and other bacteria ( 47, 49– 54). The dev promoter region is also unusual in that upstream and downstream regulatory elements interact functionally ( 29). These interactions have been proposed to involve DNA looping. A second possible role of LadA would be to influence DNA-loop formation, changing the architecture of a nucleoprotein complex that regulates dev promoter activity. In this role, the function of LadA would be analogous to that of DNA-bending architectural transcription factors like integration host factor (IHF) and HU in bacteria ( 55) and HMG-domain proteins in eukaryotes ( 56). It is likely that LadA bends the DNA upon binding, because LadA binding caused hypersensitivity to cleavage by 1,10-phenanthroline-copper at two sites (near +340 and +380; B), and hypersensitivity to DNA-cleaving agents is correlated with DNA bending for many LysR-type proteins (reviewed in ref. 43). LadA may be devoted specifically to direct regulation of the dev operon, because the ladA null mutant did not exhibit more drastic developmental defects than dev mutants. However, LadA may indirectly affect expression of other genes, because DevT positively regulates fruA ( 32). Because FruA appears to activate dev transcription, FruA and DevT form a positive regulatory loop. Such loops can form bistable switches (reviewed in ref. 57), and developing cells exhibit bimodal expression of dev ( 58), with expression predominantly in cells within fruiting bodies and the cells outside failing to activate dev ( 31). It has been proposed that C-signaling accounts for spatial control of dev expression ( 31) by activating the positive FruA/DevT feedback loop ( 32). Alternatively or in addition, signal input via LadA might control the switch. Failure to establish the FruA/DevT loop in a ladA mutant presumably results in a reduced level of FruA. This might impair expression of FruA-dependent genes, but as noted above, it does not cause developmental defects as dramatic as those caused by a null mutation in fruA. Little is known about how dev operon products regulate events leading to sporulation. A devRS mutant fails to express lacZ inserted at the Ω7536 locus, and the products of this locus are required for sporulation ( 59). The dev operon encodes Cas proteins with similarity to helicases and nucleases ( 35, 37, 60), and at least two repeats of the downstream CRISPR are transcribed as part of the operon ( 29). The 37-bp unique insert between the two repeats matches perfectly a sequence in the phage Mx8 integrase gene ( 29). Because some unique inserts of other CRISPR are similar to segments of phage and plasmid genes, it has been proposed that CRISPR-Cas systems are defense mechanisms that function analogously to eukaryotic RNA interference (RNAi) systems ( 36, 37). In addition to regulating genes required for sporulation, dev expression might protect the developing cell from Mx8 lysogenization ( 29). Multicellular eukaryotes employ large DNA regions with multiple signaling inputs to ensure proper temporal and spatial expression of crucial developmental genes. Our results provide evidence that M. xanthus multicellular development is accomplished by using a similar strategy. Bacterial Strains and Plasmids. The strains and plasmids used in this study are listed in SI Table 1, as is a description of their construction. Bacterial Growth and Development, and Assays for lacZ Expression and Sporulation. Growth of E. coli and M. xanthus was as described in ref. 20. Oxytetracycline was used at 12.5 μg/ml. Development of M. xanthus and determination of lacZ expression was as described in ref. 20. Production of heat- and sonication-resistant spores was assayed as described in ref. 61. EMSAs and Footprint Analysis. 32P-labeled DNA probes were synthesized by PCR after labeling one of the primers with [γ- 32P]ATP and T4 polynucleotide kinase, and probes were purified after 5% PAGE ( 62). FruA-DBD-His 8 and His 6-FruA were purified as described in ref. 27, except that cells were lysed by passage through a French pressure cell (SLM Aminco) at 14,000 lb/in 2. Likewise, E. coli BL21(DE3) transformed with pPVlysRflHis was induced with 1 mM IPTG and lysed, and Lad-His 6 was purified as described for FruA DBD-His 8 ( 27). EMSAs were performed as described in ref. 24. DNA-binding proteins were partially purified (AS fraction) as described in ref. 24 from 12-h-developing M. xanthus DZF1. Footprint analysis of gel slices using 1,10-phenanthroline-copper ion was as described in ref. 63. DNA-Affinity Chromatography. DNA corresponding to +282 to +456 relative to the dev TSS was synthesized by PCR with a 5′-biotin label, bound to streptavidin magnetic beads, and used for DNA-affinity chromatography with AS fraction as described in SI Materials and Methods. We thank K. Carr and G. Velicer for help with analysis of the number of predicted FruA-binding sites in the M. xanthus genome, and one of the manuscript reviewer's for suggesting that the FruA/DevT loop might form a bistable switch with signal input via LadA. This work was supported by National Science Foundation Grant MCB-0416456, the Michigan Agricultural Experiment Station, and the Foundation of Medicine and Dentistry of New Jersey. | | AS | ammonium sulfate | | | CI | complex I | | | CII | complex II | | | CRISPR | clustered regularly interspaced short palindromic repeats | | | FruA-DBD-His8 | DNA-binding domain of FruA | | | TSS | transcriptional start site. |
1. Stathopoulos A, Levine M. Dev Cell. 2005;9:449–462. [PubMed] 2. Dworkin M, Kaiser D. Myxobacteria II. Washington, DC: Am Soc Microbiol; 1993. 3. Shimkets LJ, Gill RE, Kaiser D. Proc Natl Acad Sci USA. 1983;80:1406–1410. [PubMed] 4. Sogaard-Andersen L, Kaiser D. Proc Natl Acad Sci USA. 1996;93:2675–2679. [PubMed] 5. Kroos L, Kaiser D. Genes Dev. 1987;1:840–854. [PubMed] 6. Kim SK, Kaiser D. Science. 1990;249:926–928. [PubMed] 7. Kim SK, Kaiser D. Cell. 1990;61:19–26. [PubMed] 8. Lobedanz S, Sogaard-Andersen L. Genes Dev. 2003;17:2151–2161. [PubMed] 9. Shimkets LJ, Rafiee H. J Bacteriol. 1990;172:5299–5306. [PubMed] 10. Sager B, Kaiser D. Proc Natl Acad Sci USA. 1993;90:3690–3694. [PubMed] 11. Li S-F, Lee B, Shimkets LJ. Genes Dev. 1992;6:401–410. [PubMed] 12. Kim SK, Kaiser D. J Bacteriol. 1991;173:1722–1728. [PubMed] 13. Kroos L. Proc Natl Acad Sci USA. 2005;102:2681–2682. [PubMed] 14. Goldman BS, Nierman WC, Kaiser D, Slater SC, Durkin AS, Eisen J, Ronning CM, Barbazuk WB, Blanchard M, Field C, et al. Proc Natl Acad Sci USA. 2006;103:15200–15205. [PubMed] 15. Downard JS, Kim S-H, Kil K-S. J Bacteriol. 1988;170:4931–4938. [PubMed] 16. Kil K-S, Brown G, Downard J. J Bacteriol. 1990;172:3081–3088. [PubMed] 17. Romeo JM, Zusman DR. J Bacteriol. 1991;173:2969–2976. [PubMed] 18. Gulati P, Xu D, Kaplan H. J Bacteriol. 1995;177:4645–4651. [PubMed] 19. Hao T, Biran D, Velicer GJ, Kroos L. J Bacteriol. 2002;184:3348–3359. [PubMed] 20. Viswanathan P, Kroos L. J Bacteriol. 2003;185:1405–1414. [PubMed] 21. Yoder D, Kroos L. J Bacteriol. 2004;186:3766–3776. [PubMed] 22. Viswanathan K, Viswanathan P, Kroos L. J Bacteriol. 2006;188:515–524. [PubMed] 23. Ueki T, Inouye S. J Biol Chem. 2002;277:6170–6177. [PubMed] 24. Ueki T, Inouye S. Proc Natl Acad Sci USA. 2003;100:8782–8787. [PubMed] 25. Ueki T, Inouye S. J Biol Chem. 2005;280:32279–32284. [PubMed] 26. Ueki T, Inouye S. J Bacteriol. 2005;187:8504–8506. [PubMed] 27. Yoder-Himes D, Kroos L. J Bacteriol. 2006;188:5167–5176. [PubMed] 28. Nariya H, Inouye S. Mol Microbiol. 2006;60:1205–1217. [PubMed] 29. Viswanathan P, Murphy K, Julien B, Garza AG, Kroos L. J Bacteriol. 2007 in press. 30. Kruse T, Lobedanz S, Berthelsen NM, Sogaard-Andersen L. Mol Microbiol. 2001;40:156–168. [PubMed] 31. Julien B, Kaiser AD, Garza A. Proc Natl Acad Sci USA. 2000;97:9098–9103. [PubMed] 32. Boysen A, Ellehauge E, Julien B, Sogaard-Andersen L. J Bacteriol. 2002;184:1540–1546. [PubMed] 33. Ogawa M, Fujitani S, Mao X, Inouye S, Komano T. Mol Microbiol. 1996;22:757–767. [PubMed] 34. Ellehauge E, Norregaard-Madsen M, Sogaard-Andersen L. Mol Microbiol. 1998;30:807–817. [PubMed] 35. Haft DH, Selengut J, Mongodin EF, Nelson KE. PLoS Comput Biol. 2005;1:e60. [PubMed] 36. Mojica FJ, Diez-Villasenor C, Garcia-Martinez J, Soria E. J Mol Evol. 2005;60:174–182. [PubMed] 37. Makarova KS, Grishin NV, Shabalina SA, Wolf YI, Koonin EV. Biol Direct. 2006;1:7. [PubMed] 38. Tojo N, Sanmiya K, Sugawara H, Inouye S, Komano T. J Bacteriol. 1996;178:4004–4011. [PubMed] 39. Thony-Meyer L, Kaiser D. J Bacteriol. 1993;175:7450–7462. [PubMed] 40. Horiuchi T, Taoka M, Isobe T, Komano T, Inouye S. J Biol Chem. 2002;277:26753–26760. [PubMed] 41. Kroos L, Inouye S. In: Multicellularity and Differentiation Among the Myxobacteria and Their Neighbors. Kaplan HB, Whitworth DE, editors. Washington, DC: Am Soc Microbiol; 2007. in press. 42. Sogaard-Andersen L, Slack F, Kimsey H, Kaiser D. Genes Dev. 1996;10:740–754. [PubMed] 43. Schell MA. Annu Rev Microbiol. 1993;47:597–626. [PubMed] 44. McFall SM, Chugani SA, Chakrabarty AM. Gene. 1998;223:257–267. [PubMed] 45. Huang J, Yindeeyoungyeon W, Garg RP, Denny TP, Schell MA. J Bacteriol. 1998;180:2736–2743. [PubMed] 46. Chugani SA, Parsek MR, Chakrabarty AM. J Bacteriol. 1998;180:2367–2372. [PubMed] 47. Cowan JM, Urbanowski ML, Talmi M, Stauffer GV. J Bacteriol. 1993;175:5862–5866. [PubMed] 48. Martinez-Argudo I, Ruiz-Vazquez RM, Murillo FJ. Mol Microbiol. 1998;30:883–893. [PubMed] 49. Slominska M, Konopa G, Ostrowska J, Kedzierska B, Wegrzyn G, Wegrzyn A. Mol Microbiol. 2003;47:1669–1679. [PubMed] 50. Szalewska-Palasz A, Wegrzyn A, Blaszczak A, Taylor K, Wegrzyn G. Proc Natl Acad Sci USA. 1998;95:4241–4246. [PubMed] 51. Versteeg S, Escher A, Wende A, Wiegert T, Schumann W. J Bacteriol. 2003;185:466–474. [PubMed] 52. Munson GP, Scott JR. Mol Microbiol. 2000;36:1391–1402. [PubMed] 53. Qi Y, Hulett FM. Mol Microbiol. 1998;28:1187–1197. [PubMed] 54. Gerstel U, Park C, Romling U. Mol Microbiol. 2003;49:639–654. [PubMed] 55. Johnson R, Johnson L, Schmidt J, Garder J. In: Bacterial Chromosomes. Higgins N, editor. Washington, DC: Am Soc Microbiol; 2005. pp. 65–132. 56. Merika M, Thanos D. Curr Opin Genet Dev. 2001;11:205–208. [PubMed] 57. Dubnau D, Losick R. Mol Microbiol. 2006;61:564–572. [PubMed] 58. Russo-Marie F, Roederer M, Sager B, Herzenberg LA, Kaiser D. Proc Natl Acad Sci USA. 1993;90:8194–8198. [PubMed] 59. Licking E, Gorski L, Kaiser D. J Bacteriol. 2000;182:3553–3558. [PubMed] 60. Jansen R, Embden JD, Gaastra W, Schouls LM. Mol Microbiol. 2002;43:1565–1575. [PubMed] 61. Viswanathan P, Singer M, Kroos L. J Bacteriol. 2006;188:3246–3256. [PubMed] 62. Sambrook J, Fritsch E, Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor, NY: Cold Spring Harbor Lab Press; 1989. 63. Kuwabara MD, Sigman DS. Biochemistry. 1987;26:7234–7238. [PubMed]
|
This article has been 
