FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W1992837017> ?p ?o ?g. }

• W1992837017 endingPage "9042" @default.
• W1992837017 startingPage "9037" @default.
• W1992837017 abstract "The Escherichia coli MarA protein mediates a response to multiple environmental stresses through the activation or repression in vivo of a large number of chromosomal genes. Transcriptional activation for a number of these genes has been shown to occur via direct interaction of MarA with a 20-bp degenerate asymmetric “marbox” sequence. It was not known whether repression by MarA was also direct. We found that purified MarA was sufficient in vitro to repress transcription of both purA and hdeA. Transcription and electrophoretic mobility shift experiments in vitro using mutant promoters suggested that the marbox involved in the repression overlapped the -35 promoter motif and was in the “backward” orientation. This organization contrasts with that of the class II promoters activated by MarA, in which the marbox also overlaps the -35 motif but is in the “forward” orientation. We conclude that MarA, a member of the AraC/XylS family, can act directly as a repressor or an activator, depending on the position and orientation of the marbox within a promoter. The Escherichia coli MarA protein mediates a response to multiple environmental stresses through the activation or repression in vivo of a large number of chromosomal genes. Transcriptional activation for a number of these genes has been shown to occur via direct interaction of MarA with a 20-bp degenerate asymmetric “marbox” sequence. It was not known whether repression by MarA was also direct. We found that purified MarA was sufficient in vitro to repress transcription of both purA and hdeA. Transcription and electrophoretic mobility shift experiments in vitro using mutant promoters suggested that the marbox involved in the repression overlapped the -35 promoter motif and was in the “backward” orientation. This organization contrasts with that of the class II promoters activated by MarA, in which the marbox also overlaps the -35 motif but is in the “forward” orientation. We conclude that MarA, a member of the AraC/XylS family, can act directly as a repressor or an activator, depending on the position and orientation of the marbox within a promoter. The Escherichia coli MarA protein, a member of the AraC/XylS family of transcriptional regulators, mediates cellular responses to stress through the differential control of a large number of chromosomal genes comprising the mar regulon (1Alekshun M.N. Levy S.B. Antimicrob. Agents Chemother. 1997; 41: 2067-2075Crossref PubMed Google Scholar, 2Gallegos M.-T. Schleif R. Bairoch A. Hofmann K. Ramos J.L. Microbiol. Mol. Biol. Rev. 1997; 61: 393-410Crossref PubMed Scopus (657) Google Scholar, 3Martin R.G. Rosner J.L. Curr. Opin. Microbiol. 2001; 4: 132-137Crossref PubMed Scopus (185) Google Scholar, 4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google Scholar, 5Pomposiello P.J. Bennik M.H. Demple B. J. Bacteriol. 2001; 183: 3890-3902Crossref PubMed Scopus (403) Google Scholar). MarA causes decreased susceptibility to structurally unrelated antibiotics, organic solvents, household disinfectants, and oxidative stress agents (1Alekshun M.N. Levy S.B. Antimicrob. Agents Chemother. 1997; 41: 2067-2075Crossref PubMed Google Scholar). The MarA homologs SoxS and Rob in E. coli are also stress-response proteins (2Gallegos M.-T. Schleif R. Bairoch A. Hofmann K. Ramos J.L. Microbiol. Mol. Biol. Rev. 1997; 61: 393-410Crossref PubMed Scopus (657) Google Scholar). Transcriptional activation of those mar regulon members that have been studied occurs when MarA binds as a monomer to the promoter region at a degenerate asymmetric 20-bp DNA sequence known as the “marbox” (6Martin R.G. Gillette W.K. Rhee S. Rosner J.L. Mol. Microbiol. 1999; 34: 431-441Crossref PubMed Scopus (143) Google Scholar, 7Martin R.G. Rosner J.L. Mol. Microbiol. 2002; 44: 1611-1624Crossref PubMed Scopus (119) Google Scholar). The mechanisms by which MarA activates members of the mar regulon and the details of MarA, SoxS, and Rob interactions with the marbox DNA have been investigated (for example see Refs. 6Martin R.G. Gillette W.K. Rhee S. Rosner J.L. Mol. Microbiol. 1999; 34: 431-441Crossref PubMed Scopus (143) Google Scholar, 8Kwon H.J. Bennik M.H. Demple B. Ellenberger T. Nat. Struct. Biol. 2000; 7: 424-430Crossref PubMed Scopus (153) Google Scholar, 9Martin R.G. Gillette W.K. Rosner J.L. Mol. Microbiol. 2000; 35: 623-634Crossref PubMed Scopus (81) Google Scholar, 10Rhee S. Martin R.G. Rosner J.L. Davies D.R. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10413-10418Crossref PubMed Scopus (232) Google Scholar, 11Dangi B. Pelupessey P. Martin R.G. Rosner J.L. Louis J.M. Gronenborn A.M. J. Mol. Biol. 2001; 314: 113-127Crossref PubMed Scopus (21) Google Scholar). MarA was originally described as an activator. Its expression directly induced transcription of all previously studied mar regulon members except ompF, whose down-regulation resulted indirectly from MarA activation of the antisense RNA micF, which then inhibited the translation of ompF (12Cohen S.P. McMurry L.M. Levy S.B. J. Bacteriol. 1988; 170: 5416-5422Crossref PubMed Scopus (188) Google Scholar, 13Delihas N. Forst S. J. Mol. Biol. 2001; 313: 1-12Crossref PubMed Scopus (155) Google Scholar). Repression by MarA has been shown recently in vivo for several dozen genes by two independent macroarray studies, although the gene overlap between the two studies was not large (4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google Scholar, 5Pomposiello P.J. Bennik M.H. Demple B. J. Bacteriol. 2001; 183: 3890-3902Crossref PubMed Scopus (403) Google Scholar). To determine whether the repression by MarA was direct or occurred by an indirect action, we focused on two genes, purA and hdeA. In the intact cell, transcription of purA was decreased in cells constitutively producing MarA (4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google Scholar). Expression of hdeA was repressed by induced expression of MarA (see Ref. 5Pomposiello P.J. Bennik M.H. Demple B. J. Bacteriol. 2001; 183: 3890-3902Crossref PubMed Scopus (403) Google Scholar, Supplemental Material). purA codes for adenylosuccinate synthase, which plays an important role in the de novo pathway of purine nucleotide biosynthesis and is required for adenosine monophosphate synthesis from inosine -5′-monophosphate (14Wolfe S.A. Smith J.M. J. Biol. Chem. 1988; 263: 19147-19153Abstract Full Text PDF PubMed Google Scholar). An association between purA and virulence has also been proposed (15Dunstan S.J. Simmons C.P. Strugnell R.A. Infect. Immun. 1998; 66: 732-740Crossref PubMed Google Scholar, 16Hoffman J.A. Badger J.L. Zhang Y. Kim K.S. Microb. Pathog. 2001; 31: 69-79Crossref PubMed Scopus (11) Google Scholar). hdeA encodes a periplasmic protein involved in acid resistance (17Gajiwala K.S. Burley S.K. J. Mol. Biol. 2000; 295: 605-612Crossref PubMed Scopus (169) Google Scholar, 18Waterman S.R. Small P.L. Mol. Microbiol. 1996; 21: 925-940Crossref PubMed Scopus (144) Google Scholar). In Shigella flexneri and E. coli, both acidresistant organisms, hdeA is present, but it is absent in Salmonella, which is merely acid-tolerant (17Gajiwala K.S. Burley S.K. J. Mol. Biol. 2000; 295: 605-612Crossref PubMed Scopus (169) Google Scholar). We show here that MarA directly represses both purA and hdeA. Thus MarA is a dual regulator capable of both activation and repression, depending upon the promoter. Bacterial Strains, Plasmids, and Growth Conditions—E. coli K12 strains and plasmids used in this study are shown in Table I. Strains were routinely grown at 37 °C in LB broth (per liter: 10 g of tryptone, 5 g of yeast extract, 5 g of NaCl).Table IBacterial strains and plasmids used in this studyNameGenotype/relevant characteristicsRef./SourceE. coli K12 strainsAG100argE3 thi-1 rpsL xyl supE44 [Δ (gal-uvrB)?]42George A.M. Levy S.B. J. Bacteriol. 1983; 155: 531-540Crossref PubMed Google Scholar; for ?, see Ref. 4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google ScholarGC4468Δ (lac)U169 rpsL43Touati D. J. Bacteriol. 1983; 155: 1078-1087Crossref PubMed Google ScholarJHC1096Same as GC4468 but zdd-239::Tn9 del1738 (Δ 39 kb including marRAB locus)26Greenberg J.T. Chou J.H. Monach P.A. Demple B. J. Bacteriol. 1991; 173: 4433-4439Crossref PubMed Google ScholarPlasmidspMB102ori colE1 lacI lacZp::marA, AmpR5Pomposiello P.J. Bennik M.H. Demple B. J. Bacteriol. 2001; 183: 3890-3902Crossref PubMed Scopus (403) Google ScholarpSP-nfnB1pSP-luc+ (Promega; promoter-less luciferase gene, ori colE1, AmpR) containing a 177-bp fragment of nfnB promoter20Barbosa T.M. Levy S.B. Mol. Microbiol. 2002; 45: 191-202Crossref PubMed Scopus (30) Google Scholar Open table in a new tab General Molecular Biology Manipulations—Preparation of genomic and plasmid DNA and purification of PCR products and of DNA fragments from agarose gels were carried out with Qiagen products as follows: QIAamp Tissue kit, QIAprep Spin Miniprep kit, QIAquick PCR purification kit, and QIAquick gel extraction kit, respectively, following the manufacturer’s instructions. Restriction enzyme digestions followed standard procedures (19Sambrook J. Russell D.W. Molecular Cloning: A Laboratory Manual.3rd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY2001: 1.84-1.87Google Scholar). PCR amplification utilized either Taq polymerase (Invitrogen) or TripleMaster Polymerase Mix (proofreading) (Eppendorf) in a GeneAmp® PCR System 9700 thermocycler (PE Applied Biosystems, PerkinElmer Life Sciences). Oligonucleotide synthesis and DNA sequencing were performed by the Tufts University Core Facility. RNA Preparation and Northern Analysis—Total RNA was extracted from cells using the Qiagen RNeasy kit. Two micrograms of RNA were fractionated by electrophoresis in a denaturing formaldehyde agarose gel, which was then blotted onto a Hybond-N (Amersham Biosciences) membrane. RNA was visualized by both ethidium bromide staining of the gel and by methylene blue staining of the membrane. purA or hdeA probes were amplified by PCR using AG100 chromosomal DNA as template. For purA,an ∼1270-bp region was amplified using the primer pair F2:P-Rev (GAAAACGATTGGCTGAAC:AAGGTGGATTCAGACCAG). For hdeA,an ∼396-bp region was amplified using the primer pair HdeA1:HdeA2 (TTGATTCGTGACGGCTCT:ATGCAAGGAAGTACGATGT). Hybridization of membrane-bound RNA to the purA or hdeA probes was performed as described previously (4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google Scholar). PCR Amplification of Promoter Regions Used in Transcription in Vitro—The promoter region of the nfnB-luciferase fusion in pSP-nfnB1 was amplified with the primer pair NFN-F1:lucR (CCCGGTACCCTTCGCGATCTGTCAACG:CTTCCAGCGGATAGAATGG), producing a 261-bp product with a 105-nt 1The abbreviations used are: nt, nucleotide; EMSA, electrophoretic mobility shift assays; IPTG, isopropyl-1-thio-β-d-galactopyranoside. transcript (20Barbosa T.M. Levy S.B. Mol. Microbiol. 2002; 45: 191-202Crossref PubMed Scopus (30) Google Scholar). All other PCR products were generated using AG100 chromosomal DNA. The wild type 225-bp purA promoter-containing template was amplified using primer pair purApF:purApR (GGAAAACGATTGGCTGAAC:CGTTCAGTCAGAAGATCG). PCR products containing mutated purA marboxes were made using reverse primer purApR together with a mutated forward primer. The 5′ end of a mutated forward primer corresponded to the 5′ end of the PCR product shown in Fig. 5; its arbitrary 3′ terminus was either… AACTCTG-3′ (for D2F and D3F) or… GAAAAGC-3′ (for all others). All purA templates yielded a transcript of 105 nt. For hdeA, a 193-bp PCR product of the promoter region was amplified using the primer pair HdeA1a:HdeA2a (TCTGATGCATCTGTAACTCA:GAAGCAGACCACCAAGAATA). Mutant PCR products were made with the reverse primer HdeA2a together with primers beginning at the 5′ end shown in Fig. 6. All hdeA transcripts were 94 nt long. The 167-bp gnd promoter-containing PCR product used the primers GNDF:GNDR (TCGCAACTTTGATCGAAT:TACATACTCCTGTCAGGT); the transcript was 55 nt long. All PCR products were gel-purified before being used in the transcription reactions in vitro.Fig. 6Effect of mutations in thehdeApromoter upon repressionin vitroby MarA. See the legend of Fig. 5 for format. The arrow indicates marbox 3. The wild type template began at bp -99. All the other templates had 5′ start sites as shown. All templates ended at bp +94. Values represent an average of 4–12 experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Transcription in Vitro—N-terminal polyhistidine-tagged MarA (His6-MarA) protein was purified essentially as described by Jair et al. (21Jair K.-W. Martin R.G. Rosner J.L. Fujita N. Ishihama A. Wolf Jr., R.E. J. Bacteriol. 1995; 177: 7100-7104Crossref PubMed Google Scholar) and was a generous gift of Victoria Bartlett and Michael Alekshun (Paratek Pharmaceuticals, Boston). Single round (5 min) transcription in vitro was carried out essentially as described (22Jair K.-W. Fawcett W.P. Fujita N. Ishihama A. Wolf R.E.J. Mol. Microbiol. 1996; 19: 307-317Crossref PubMed Scopus (74) Google Scholar) in 30 μl using PCR products as DNA templates. The Reaction Buffer at pH 7.8 contained 50 mm Tris-HCl, 0.1 mm EDTA, 3 mm magnesium acetate, 0.1 mm dithiothreitol, 20 mm sodium chloride, and 250 μg/ml bovine serum albumin. Other components of the pre-incubation mixture are as follows: E. coli RNA polymerase holoenzyme (σ70; Epicentre), 40 nm; the RNase inhibitor SUPERase In™ (Ambion), 500 units/ml; test DNA (purA, hdeA, or nfnB promoter-containing DNAs); and control DNA (gnd promoter-containing DNA) each at 2 nm. In most experiments involving mutant templates with lower transcription levels, the concentration of each test DNA was raised to 4 or 6 nm. Purified His6-MarA, diluted as needed in Reaction Buffer, was added to the pre-incubation mixture on ice to give 200 nm; the buffer in which His6-MarA was prepared was similarly diluted and added to the control. Open complexes were allowed to form at 37 °C for 15 min, when a heparin/nucleotide mixture was added to initiate transcription; the final concentrations were heparin, 1.2 mg/ml, and nucleotides, 0.3 mm UTP, 0.96 mm ATP, CTP, and GTP with [α-32P]UTP at 0.5 μCi/ml (PerkinElmer Life Sciences). After 5 min, an equal volume of Ambion gel loading buffer II was added to stop the reaction; the samples were boiled for 3 min, chilled, and fractionated by electrophoresis on a 7% polyacrylamide, 8 m urea gel in Tris borate EDTA, pH 8.3. Quantification of Transcript Levels—Radioactivity was detected by exposure of membranes/dried gels to Kodak BioMax MS x-ray film; alternatively, exposure to a storage phosphor screen was followed by scanning with a Storm PhosphorImager (Amersham Biosciences). Film images for methylene blue-stained membranes showing ribosomal RNA were scanned into Adobe Photoshop, and the bands were quantified using NIH Image 1.62. Transcripts from both Northern blots and from transcription reactions in vitro were quantified after PhosphorImager scanning by ImageQuant (Amersham Biosciences). The amounts of the two major transcripts produced, in vivo by hdeA or in vitro by purA, were combined. The amount of each test transcript was normalized by the level of the ribosomal RNA (for Northern blots) or gnd transcript (for in vitro transcriptions). Identification of Putative Marboxes—Putative marboxes were identified using the “search patterns” utility of Colibri (genolist.pasteur.fr/Colibri/) with the 20-bp degenerate consensus sequence A(CAT)RGCACRWWNNRYYAAA(CAT)N, where R indicates A or G; W indicates A or T; Y indicates T or C; and N indicates A, T, G, or C (7Martin R.G. Rosner J.L. Mol. Microbiol. 2002; 44: 1611-1624Crossref PubMed Scopus (119) Google Scholar). Only marboxes between 150-bp upstream and 100-bp downstream of the transcriptional start that had 15 or more matches and the 5th C invariant were considered. Electrophoretic Mobility Shift Assays (EMSA) for purA—Binding of purified His6-MarA to DNA from the purA promoter region was measured by EMSA. The DNA used for fragments of 107–185-bp was amplified by PCR using a chromosomal AG100 template, whereas the 20–29-mers encompassing marboxes were made by annealing equimolar concentrations of complementary DNA oligomers. The forward oligomers for the 20-bp marboxes were marbox 1 (AGTGCAAAAAGTGCTGTAAC), marbox 2 (TTGAGTGCAAAAAGTGCTGT), and marbox 3 (AGGTCATTTTTGAGTGCAAA). The forward oligomer for the wild type 29-bp fragment comprising overlapping marboxes 1 and 2 was TTTTGAGTGCAAAAAGTGCTGTAACTCTG. Two mutant oligomers were designed such that all 4-bp of RE1 of either marbox 1 or marbox 2 were mutated; the mutations are shown in Fig. 5, MB1.1F and MB2.1F, respectively. In all cases, DNA was labeled at the 3′ end with digoxigenin-11-ddUTP (DIG gel shift kit, Roche Applied Science), and the EMSA reactions were performed as described previously (20Barbosa T.M. Levy S.B. Mol. Microbiol. 2002; 45: 191-202Crossref PubMed Scopus (30) Google Scholar). Transient Induction of MarA Caused a Decrease of purA and hdeA Expression in Cells—The earlier E. coli DNA macroarray experiments showing that constitutively produced MarA reduced expression of purA (4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google Scholar) were confirmed subsequently using Northern assays (data not shown). To determine whether this repression was a result of MarA function, as opposed to a stress-related artifact associated with the constitutive overproduction of MarA, we controlled expression of marA with the IPTG-inducible lac promoter in plasmid pMB102 (5Pomposiello P.J. Bennik M.H. Demple B. J. Bacteriol. 2001; 183: 3890-3902Crossref PubMed Scopus (403) Google Scholar) in the mar-deleted strain JHC1096 (Table I). In the same experiment we also examined the expression of hdeA, which had been observed in another macroarray study to be repressed upon MarA induction in cells (5Pomposiello P.J. Bennik M.H. Demple B. J. Bacteriol. 2001; 183: 3890-3902Crossref PubMed Scopus (403) Google Scholar). After a 1-h induction of MarA by IPTG, a decline was seen in the levels of transcripts of purA (Fig. 1A) and hdeA (Fig. 1B). Typically, the decrease at 1 mm IPTG was ∼3-fold for purA and 6–25-fold for hdeA, with less IPTG required for hdeA. Each of the two bands in Fig. 1B probably includes both genes of the hdeAB operon (23Arnqvist A. Olsen A. Normark S. Mol. Microbiol. 1994; 13: 1021-1032Crossref PubMed Scopus (142) Google Scholar, 35Tucker D.L. Tucker N. Ma Z. Foster J.W. Miranda R.L. Cohen P.S. Conway T. J. Bacteriol. 2003; 185: 3190-3201Crossref PubMed Scopus (98) Google Scholar); therefore, hdeB was probably also repressed. No decrease was seen in cells bearing the vector control (pJPBH) (Fig. 1, A and B). These results show that the decrease in purA and hdeA expression was caused, directly or indirectly, by transiently synthesized MarA rather than by a stress reaction due to constitutively overabundant MarA. Purified MarA Repressed Transcription of purA and hdeA in Vitro—To determine whether repression was direct or indirect, we tested the effect of purified His6-MarA protein on transcription of purA and hdeA in vitro. This protein, engineered to facilitate purification and hereafter referred to as MarA, had been shown previously to activate nfnB transcription in vitro (20Barbosa T.M. Levy S.B. Mol. Microbiol. 2002; 45: 191-202Crossref PubMed Scopus (30) Google Scholar). The DNA templates for the single round transcription reactions were PCR products bearing a promoter, its transcriptional start, and 55–105 bp downstream. The transcriptional start sites of the purA gene (14Wolfe S.A. Smith J.M. J. Biol. Chem. 1988; 263: 19147-19153Abstract Full Text PDF PubMed Google Scholar) and the hdeAB operon (23Arnqvist A. Olsen A. Normark S. Mol. Microbiol. 1994; 13: 1021-1032Crossref PubMed Scopus (142) Google Scholar) have been defined. We mixed the promoter-containing DNA of purA or hdeA with that of the control gnd (a promoter that is unresponsive to MarA) (21Jair K.-W. Martin R.G. Rosner J.L. Fujita N. Ishihama A. Wolf Jr., R.E. J. Bacteriol. 1995; 177: 7100-7104Crossref PubMed Google Scholar, 22Jair K.-W. Fawcett W.P. Fujita N. Ishihama A. Wolf R.E.J. Mol. Microbiol. 1996; 19: 307-317Crossref PubMed Scopus (74) Google Scholar, 24Ariza R.R. Li Z. Ringstad N. Demple B. J. Bacteriol. 1995; 177: 1655-1661Crossref PubMed Google Scholar, 25Jair K.W. Yu X. Skarstad K. Thony B. Fujita N. Ishihama A. Wolf Jr., R.E. J. Bacteriol. 1996; 178: 2507-2513Crossref PubMed Google Scholar). The promoter for the MarA-activated gene nfnB (20Barbosa T.M. Levy S.B. Mol. Microbiol. 2002; 45: 191-202Crossref PubMed Scopus (30) Google Scholar), also mixed with gnd DNA, was tested in parallel as a positive control for MarA activity. In the presence of 200 nm MarA, there was a 5–6-fold activation in nfnB transcription, as expected (Fig. 2). As anticipated, MarA did not affect the expression from the control gnd promoter. The same concentration and batch of MarA protein, on the other hand, repressed both purA and hdeA (Fig. 2). The average repression ratio was 0.44 for purA (see Fig. 5, WT) and 0.60 for hdeA (see Fig. 6, WT) relative to controls without MarA. These results showed that MarA by itself could directly repress purA and hdeA transcription. DNA Sequences Involved in Binding by MarA, EMSA—Activation of transcription by MarA involves binding of this protein as a monomer to a degenerate 20-bp “marbox” located upstream from or overlapping the -35 motif within the promoter (6Martin R.G. Gillette W.K. Rhee S. Rosner J.L. Mol. Microbiol. 1999; 34: 431-441Crossref PubMed Scopus (143) Google Scholar, 10Rhee S. Martin R.G. Rosner J.L. Davies D.R. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10413-10418Crossref PubMed Scopus (232) Google Scholar). It seemed likely that for repression, a marbox in the promoter region would also be involved, although perhaps with a different location or orientation. The purA promoter region contained five consensus marboxes (see “Experimental Procedures” for criteria); all were upstream of the translational start site. By EMSA using PCR products covering various regions of the purA promoter, we showed that MarA bound with similar affinity to fragments spanning bp -140 to +45, -121 to -14, and -96 to +45, relative to the transcriptional start (data not shown). Therefore, binding occurred between bp -96 and -14, eliminating three of the marboxes from consideration. The remaining were marbox 1 (forward, bp -49 to -30, 15/20 consensus bp) and the overlapping marbox 2 (backward, bp -33 to -52, 15/20 consensus bp) (see Fig. 5). Further mobility shift assays with 20-bp DNA fragments comprising the individual marboxes showed that MarA did not bind to a “control” poor consensus backward “marbox 3” (bp -42 to -61; only 12/20 consensus bp) but did bind to marboxes 1 and 2. The affinity for marbox 2 (Kd < 100 nm) was greater than that for marbox 1 (Kd > 400 nm) (Fig. 3A), suggesting that marbox 2 was preferred. It was not straightforward to further distinguish the relative importance of marboxes 1 and 2 because they overlapped at 17 of their 20 bp (see Fig. 5). We therefore took advantage of the two highly conserved marbox “recognition elements,” RE1 and RE2, of 4 bp each which are particularly critical for activation by SoxS (27Griffith K.L. Wolf Jr., R.E. Mol. Microbiol. 2001; 40: 1141-1154Crossref PubMed Scopus (52) Google Scholar) and presumably by MarA. The crystal structure of MarA bound to marbox DNA shows the two recognition helices of the protein in contact with the recognition elements of the marbox (10Rhee S. Martin R.G. Rosner J.L. Davies D.R. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10413-10418Crossref PubMed Scopus (232) Google Scholar). We mutated all 4 bp of RE1 of either marbox 1 or marbox 2 (see “Experimental Procedures” and Fig. 5) in 29-bp duplexes containing both purA marboxes 1 and 2. The mutations in the RE1 of one marbox did not affect the RE1 of the other. We found that MarA still bound to mutated purA marbox 1 but did not bind to mutated marbox 2 (Fig. 3B). This result confirmed the importance of marbox 2. In the case of hdeA, there were six consensus marboxes (Fig. 4; see “Experimental Procedures” for criteria). However, EMSA experiments were not conclusive because a shift in mobility was seen for only a very small fraction of hdeA promoter DNA, even with MarA at 750 nm. DNA Sequences Involved in Repression by MarA, Transcription Assays in Vitro—For purA, the EMSA analyses had identified marbox 2 as more important for MarA binding but did not prove its involvement in MarA-mediated transcriptional repression. Therefore, we repeated the in vitro transcription reactions using promoters with mutations in marboxes 1 and 2 to look for loss of repression. Both deletions and point mutations were used. Some of these mutations caused a reduction in transcription itself (see “Discussion”), so a doubling or tripling in the amount of template DNA (including with wild type controls) was used to enhance the amount of transcript (see “Experimental Procedures”). For purA, changing 3 bp in RE2 of marbox 2 did not prevent repression (Fig. 5, MB2.2F) nor did deletions of 7 or 14 bp in marboxes 1 and 2 (Fig. 5, purD2F, purD3F, and DelF). Truncation of the final 6 bp of marbox 2 decreased but did not eliminate repression (Fig. 5, purApF2). The unexpected repression still seen despite marbox deletions and truncation is discussed below. Nevertheless, repression was clearly prevented by the alteration of all 4 bp of RE1 in marbox 2 (Fig. 5, MB2.1F), whereas no effect was seen for mutations in RE1 of marbox 1 (Fig. 5, MB1.1F and mut1F). These latter results again pointed to marbox 2, and not marbox 1, as critical for repression of purA. For hdeA, a template missing marbox 1, the last 13 bp of marbox 6 and the last 6 bp of marbox 5 still showed repression (Fig. 6, MB5.5), as did a template missing all of marbox 5 (and consequently the last 2 bp of marbox 3) (Fig. 6, ΔMB5). Therefore, it appeared that marboxes 1, 5, and 6 were not needed for repression. Not surprisingly, mutations of the last 2 bp of marbox 3 did not block repression (Fig. 6, MB3a). However, alteration of all 4 bp of RE2 of marbox 3 did block repression (Fig. 6, MB3b). Therefore, marbox 3 was necessary for repression of hdeA. To investigate the role of RE1 of marbox 3, two of its 4 bp (within the -35 hexamer) were changed, but this reduced basal levels of transcription such that quantification was not possible. When the -35 hexamer (atgaca) was mutated to the consensus hexamer (ttgaca), transcription increased notably; repression by MarA still occurred (Fig. 6, MB3e), even though the mutation also affected 1 bp of RE1 of marbox 3. When this single mutation in the -35 hexamer was then combined with the RE2 mutations of MB3b, transcription remained high, and repression was again lost (Fig. 6, MB3k). These results together indicated that marbox 3, and particularly its RE2, was critical for repression of hdeA by MarA. No further mutations were attempted in RE1 as it completely overlapped the -35 hexamer. By using transcription analysis in vitro, we have demonstrated that purified MarA protein was sufficient to down-regulate expression of purA and hdeA via the promoter region of these genes. How the same protein is able both to activate and to repress transcription probably relates to the identity and location of DNA sequence motifs that are recognized by MarA. For our studies here, we presumed that marboxes within the promoter region would be involved. By a combination of EMSA and transcription experiments in vitro using wild type and mutant marboxes, we found that the marbox most likely to contribute to repression of both purA and hdeA overlapped the -35 promoter motif and was oriented in the “backward” direction. “Class II” promoters activated by MarA also have marboxes overlapping the -35 motif, but these marboxes are in the forward direction (6Martin R.G. Gillette W.K. Rhee S. Rosner J.L. Mol. Microbiol. 1999; 34: 431-441Crossref PubMed Scopus (143) Google Scholar). Presumably the two different marbox orientations result in two types of interactions between MarA, RNA polymerase, and/or DNA, one leading to activation and the other to repression. In the case of purA, marbox 2 was defined as critical by its affinity for MarA in EMSA and by the fact that point mutations in RE1 of marbox 2 prevented both the mobility shift and the repression by MarA in vitro. Although deletions within this marbox did not prevent repression, this result could readily be explained (for the 7-bp deletions) by the formation of new backward marboxes from the resultant joined sequences, with consensus bp of 15/20 (D2F) or 16/20 (D3F) (see Fig. 5). In the case of the 14-bp deletion (DelF), the newly created marbox was 1 bp too short, but it had 14 consensus bp. However, mutant template purApF2, lacking the terminal 6 bp of the marbox, was still somewhat repressed (Fig. 5). That finding suggested that only the first part of this marbox was needed for repression. Transcription using the templates with deletions was considerably reduced, possibly due to destruction of a potential “UP element” (28Estrem S.T. Gaal T. Ross W. Gourse R.L. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 9761-9766Crossref PubMed Scopus (237) Google Scholar) comprising bp -41 to -57 in the purA promoter. For hdeA, even though repression of transcription was seen in vitro, EMSA experiments showed minimal binding of MarA to the hdeA promoter region. Discrepancies between DNA binding as measured by EMSA and by other means have also been observed for other promoters, for example ybjC (29Martin R.G. Gillette W.K. Martin N.I. Rosner J.L. Mol. Microbiol. 2002; 43: 355-370Crossref PubMed Scopus (71) Google Scholar) and inaA (6Martin R.G. Gillette W.K. Rhee S. Rosner J.L. Mol. Microbiol. 1999; 34: 431-441Crossref PubMed Scopus (143) Google Scholar). We therefore depended upon the transcription assays in vitro to define the particular marbox critical for repression of hdeA. Specifically, upstream marboxes 1, 5, and 6 were not required. Marbox 3, overlapping the -35 promoter motif, was critical for repression, because there was no repression if mutations were placed in all 4 bp of its RE2 (Fig. 6). The role of RE1 of marbox 3 remained unclear because RE1 completely overlapped the -35 hexamer and could not be extensively mutagenized. The last 2 bp of marbox 3 could be deleted without eliminating repression in vitro (ΔMB5), suggesting that, as with purA, an entire 20-bp marbox may not be needed for repression (Fig. 6). Of note, RE2 was important for repression of hdeA but not purA. Two relatives of MarA were reported to directly repress transcription of other promoters, but the marbox involved was not determined. IscR, a newly recognized member of the AraC family of transcriptional regulators and a “distant” MarA relative (20% identity, 40% amino acid similarity), was able to function in vitro as a transcriptional repressor of the iscRSUA operon (30Schwartz C.J. Giel J.L. Patschkowski T. Luther C. Ruzicka F.J. Beinert H. Kiley P.J. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14895-14900Crossref PubMed Scopus (332) Google Scholar). Intriguingly, we had observed down-regulation of the same promoter by constitutive expression of MarA (4Barbosa T.M. Levy S.B. J. Bacteriol. 2000; 182: 3467-3474Crossref PubMed Scopus (293) Google Scholar) (iscS = b2530). Also, the MarA homolog SoxS repressed its own expression in whole cells and bound its own promoter in vitro (31Nunoshiba T. Hidalgo E. Li Z. Demple B. J. Bacteriol. 1993; 175: 7492-7494Crossref PubMed Google Scholar); no further studies of this observation have been reported. The physiological significance of the repression of purA and hdeA by MarA is not known. Other proteins also repress purA and hdeA. The purA promoter is repressed by PurR in the presence of purines (32He B. Zalkin H. J. Bacteriol. 1994; 176: 1009-1013Crossref PubMed Google Scholar), and the promoter of hdeA is repressed by HN-S (33Hommais F. Krin E. Laurent-Winter C. Soutourina O. Malpertuy A. Le Caer J.P. Danchin A. Bertin P. Mol. Microbiol. 2001; 40: 20-36Crossref PubMed Scopus (343) Google Scholar). The hdeA promoter can be recognized by both σ70 and RpoS (23Arnqvist A. Olsen A. Normark S. Mol. Microbiol. 1994; 13: 1021-1032Crossref PubMed Scopus (142) Google Scholar), and MarA can activate both σ70- and RpoS-dependent promoters (34Van Dyk T.K. Ayers B.L. Morgan R.W. Larossa R.A. J. Bacteriol. 1998; 180: 785-792Crossref PubMed Google Scholar). Interestingly, GadX binds to a region in the hdeA promoter (35Tucker D.L. Tucker N. Ma Z. Foster J.W. Miranda R.L. Cohen P.S. Conway T. J. Bacteriol. 2003; 185: 3190-3201Crossref PubMed Scopus (98) Google Scholar) which overlaps the MarA-binding site, marbox 3. The repression of purA by MarA was quantitatively similar in vivo and in vitro, whereas that of hdeA was greater in vivo (∼6–25-fold) than in vitro (∼1.7-fold). Possibly repression in vivo of hdeA by MarA is aided by a cellular product that is constitutive or is induced by MarA. From our results we postulate that MarA represses by binding to a region of the promoter that overlaps the RNA polymerase -35 binding motif. Certain other repressors also have binding sites that overlap or lie between the RNA polymerase-binding motifs; they may provide models for MarA action (36Choy H. Adhya S. Neidhardt F.C. Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society for Microbiology, Washington, D. C.1996: 1287-1299Google Scholar, 37Gralla J.D. Collado-Vides J. Neidhardt F.C. Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society for Microbiology, Washington, D. C.1996: 1232-1245Google Scholar). LexA sterically blocks the binding of RNA polymerase to the uvrA promoter (38Bertrand-Burggraf E. Hurstel S. Daune M. Schnarr M. J. Mol. Biol. 1987; 193: 293-302Crossref PubMed Scopus (59) Google Scholar). MerR bends the -10 motif of the merT promoter away from RNA polymerase, causing a reduction in the rate of open complex formation (39Ansari A.Z. Bradner J.E. O'Halloran T.V. Nature. 1995; 374: 371-375Crossref PubMed Scopus (170) Google Scholar, 40Caslake L.F. Ashraf S.I. Summers A.O. J. Bacteriol. 1997; 179: 1787-1795Crossref PubMed Google Scholar). Arc protein at the Pant promoter of bacteriophage P22 slows the rate of open complex formation (41Smith T.L. Sauer R.T. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 8868-8872Crossref PubMed Scopus (35) Google Scholar). Although our sample of genes directly repressed by MarA is small, the location and orientation of the marboxes in the two repressed genes studied here are unlike that in either of the two classes of the MarA-activated genes. We have postulated that it is the backward marbox overlapping the -35 motif that uniquely gives rise to repression. Other cases in which proteins activate some promoters and repress others have been described (37Gralla J.D. Collado-Vides J. Neidhardt F.C. Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society for Microbiology, Washington, D. C.1996: 1232-1245Google Scholar). However, we have found no previous example in which the change between activation and repression relies on the orientation of the binding site. We thank Michael N. Alekshun and Mark Silby for helpful comments. We are also grateful to Victoria Bartlett and Michael N. Alekshun for providing the purified MarA protein, Bruce Demple (Harvard School of Public Health) for plasmid pMB102, and Xiaowen Bina for pJPBH." @default.
• W1992837017 created "2016-06-24" @default.
• W1992837017 creator A5030266583 @default.
• W1992837017 creator A5066070289 @default.
• W1992837017 creator A5084429683 @default.
• W1992837017 creator A5084730147 @default.
• W1992837017 date "2004-03-01" @default.
• W1992837017 modified "2023-09-26" @default.
• W1992837017 title "The Escherichia coli Transcriptional Regulator MarA Directly Represses Transcription of purA and hdeA" @default.
• W1992837017 cites W1485773442 @default.
• W1992837017 cites W1488562396 @default.
• W1992837017 cites W1513616826 @default.
• W1992837017 cites W1515974038 @default.
• W1992837017 cites W1517253990 @default.
• W1992837017 cites W1530210993 @default.
• W1992837017 cites W1546369838 @default.
• W1992837017 cites W1556556200 @default.
• W1992837017 cites W1570559908 @default.
• W1992837017 cites W1572693909 @default.
• W1992837017 cites W1574886161 @default.
• W1992837017 cites W1656818934 @default.
• W1992837017 cites W1892603358 @default.
• W1992837017 cites W1917998177 @default.
• W1992837017 cites W1936237469 @default.
• W1992837017 cites W1967638733 @default.
• W1992837017 cites W1968366930 @default.
• W1992837017 cites W1971945312 @default.
• W1992837017 cites W1974830640 @default.
• W1992837017 cites W2002254680 @default.
• W1992837017 cites W2008782260 @default.
• W1992837017 cites W2008795497 @default.
• W1992837017 cites W2009646503 @default.
• W1992837017 cites W2017495709 @default.
• W1992837017 cites W2022289292 @default.
• W1992837017 cites W2040046706 @default.
• W1992837017 cites W2043154296 @default.
• W1992837017 cites W2069224592 @default.
• W1992837017 cites W2080782737 @default.
• W1992837017 cites W2094191029 @default.
• W1992837017 cites W2125021980 @default.
• W1992837017 cites W2130379757 @default.
• W1992837017 cites W2131410730 @default.
• W1992837017 cites W2131884256 @default.
• W1992837017 cites W2140841500 @default.
• W1992837017 cites W2150825662 @default.
• W1992837017 cites W2152094323 @default.
• W1992837017 cites W2162198167 @default.
• W1992837017 cites W2166884322 @default.
• W1992837017 doi "https://doi.org/10.1074/jbc.m313602200" @default.
• W1992837017 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/14701822" @default.
• W1992837017 hasPublicationYear "2004" @default.
• W1992837017 type Work @default.
• W1992837017 sameAs 1992837017 @default.
• W1992837017 citedByCount "36" @default.
• W1992837017 countsByYear W19928370172012 @default.
• W1992837017 countsByYear W19928370172013 @default.
• W1992837017 countsByYear W19928370172014 @default.
• W1992837017 countsByYear W19928370172015 @default.
• W1992837017 countsByYear W19928370172016 @default.
• W1992837017 countsByYear W19928370172019 @default.
• W1992837017 countsByYear W19928370172020 @default.
• W1992837017 countsByYear W19928370172022 @default.
• W1992837017 countsByYear W19928370172023 @default.
• W1992837017 crossrefType "journal-article" @default.
• W1992837017 hasAuthorship W1992837017A5030266583 @default.
• W1992837017 hasAuthorship W1992837017A5066070289 @default.
• W1992837017 hasAuthorship W1992837017A5084429683 @default.
• W1992837017 hasAuthorship W1992837017A5084730147 @default.
• W1992837017 hasBestOaLocation W19928370171 @default.
• W1992837017 hasConcept C104317684 @default.
• W1992837017 hasConcept C138885662 @default.
• W1992837017 hasConcept C179926584 @default.
• W1992837017 hasConcept C185592680 @default.
• W1992837017 hasConcept C27153228 @default.
• W1992837017 hasConcept C2908821069 @default.
• W1992837017 hasConcept C41895202 @default.
• W1992837017 hasConcept C54355233 @default.
• W1992837017 hasConcept C547475151 @default.
• W1992837017 hasConcept C6929976 @default.
• W1992837017 hasConcept C86339819 @default.
• W1992837017 hasConcept C86803240 @default.
• W1992837017 hasConcept C95444343 @default.
• W1992837017 hasConceptScore W1992837017C104317684 @default.
• W1992837017 hasConceptScore W1992837017C138885662 @default.
• W1992837017 hasConceptScore W1992837017C179926584 @default.
• W1992837017 hasConceptScore W1992837017C185592680 @default.
• W1992837017 hasConceptScore W1992837017C27153228 @default.
• W1992837017 hasConceptScore W1992837017C2908821069 @default.
• W1992837017 hasConceptScore W1992837017C41895202 @default.
• W1992837017 hasConceptScore W1992837017C54355233 @default.
• W1992837017 hasConceptScore W1992837017C547475151 @default.
• W1992837017 hasConceptScore W1992837017C6929976 @default.
• W1992837017 hasConceptScore W1992837017C86339819 @default.
• W1992837017 hasConceptScore W1992837017C86803240 @default.
• W1992837017 hasConceptScore W1992837017C95444343 @default.
• W1992837017 hasIssue "10" @default.
• W1992837017 hasLocation W19928370171 @default.
• W1992837017 hasOpenAccess W1992837017 @default.

• next