pause site in the leader-attenuator ... Integration host factor (IHF)' protein consists of two dis- ... binding of a single, 20,000-dalton, IHF dimer to a target site.
THE JOURNAL OF BIOLOCVXL CHEMIWRY 0 19% by The American Society for Biochemistry
Vol. 266, No. 3, Issue of January 25, pp. 1965%1996,199l and Molecular
Prmted in U.S. A.
Biology. Inc.
Integration Host Factor-mediated Operon of Escherichia coZi*
Expression
of the iZvGMEDA (Received
John M. PagelS and G. Wesley Hatfield4 From the Department of Microbiology and Molecular Genetics, Calijornia Irvine, California 92717
The structural genes of the ilvGMEDA operon of Escherichia coli are preceded by two promoters, iZvPG1 and ilvPG2, and a leader-attenuator region. Alkylation protection and hydroxyl radical footprinting techniques have been used to demonstrate that integration host factor (IHF) interacts with the nucleotides in a consensus-like DNA sequence located immediately downstream of the RNA polymerase transcriptional pause site in the leader-attenuator region. In the presence of purified IHF protein, in vitro transcriptional pausing of RNA polymerase at the leader-attenuator pause site is increased Z-fold and, concomitantly, a 2fold increase in transcriptional termination at the attenuator is observed. Strains containing chromosomal transcriptional fusions of various segments of the ilvGMEDA promoter-attenuator region to the galK gene were used to show that IHF also decreases the in vivo basal level of transcriptional readthrough at the attenuator Z-fold. The binding of IHF to another target site in the ilvPG1 promoter region represses transcription from this promoter and causes a 4-fold stimulation of transcription initiation from the downstream ilvPG2 promoter I-fold. This IHF-mediated control of transcription initiation from the upstream promoter region is independent of the regulation of transcription termination effected by IHF interaction at the attenuator site. Thus, IHF is capable of regulating the expression of the ilvGMEDA operon in opposing manners; it can activate transcription initiation of this operon from the ilvP,Z promoter 4-fold and increase the termination of this transcription at the downstream attenuator Z-fold.
Integration host factor (IHF)’ protein consists of two dissimilar subunits, cy and p, which are encoded by the Escherichia coli himA and himD genes, respectively (l-4). Based on nucleotide and amino acid comparisons, IHF is assumed to be structurally related to the bacterial histone-like protein HU, a type II DNA-binding protein for which the crystal structure has been determined (5, 6). Although the HU protein-DNA helix interactions are unknown, biochemical experiments have shown that IHF interacts primarily with the * This work was supported in part by National Institutes of Health Grant GM24330. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Trainee supported by United States Public Health Service Training Grant GM0714.
§ To whom correspondence should be addressed. 1 The abbreviations pair; MeTSOI, dimethyl
used are: IHF, sulfate; HU,
integration hydroxyurea.
host factor;
bp, base 1985
for publication,
July
19,199O)
College of Medicine, University of California,
bases and deoxyribose sugars of the DNA backbone in the minor groove of the DNA helix (7-9). Unlike HU protein, which binds to DNA in a nonspecific manner, IHF binds to a specific DNA consensus sequence, 5’-WATCAANNNNTTR3’ (where W = A or T, R = A or G, N = any base) (4). The binding of a single, 20,000-dalton, IHF dimer to a target site protects 30-40 base pairs of DNA from DNase I attack (7,9). Moreover, IHF has been shown to induce a bend, empirically estimated to be greater than 140”, in the DNA at the target binding site (4, 9-12). By analogy to the protein structure of HU, it has been proposed that an IHF-induced DNA bend may be facilitated by the interaction of two antiparallel @stranded arms, one extending from each subunit of the protein, with the minor groove of the DNA helix. It is thought that these arm interactions may help to wrap the DNA around the surface of the protein and thus accommodate the large region of DNA interaction (9). Extensive genetic and biochemical analyses have suggested a role for IHF in the regulation of several cellular functions, such as plasmid DNA replication (8, 13, 14), phage encapsidation (l&16), recombination reactions involved in the movement of transposable elements (17, 18), fimbrial phase variation (19, 20), conjugal transfer of plasmids in E. coli (21, 22), and the posttranscriptional control of bacteriophage X cI1 gene expression (23,24). Examples of IHF involvement in the regulation of transcriptional initiation at several bacterial and phage promoters also have been reported. For example, it has been demonstrated, both in uivo and in uitro, that IHF stimulates transcription initiation from the bacteriophage X pL (25) and phage Mu pE promoters (26, 27). Conversely, IHF represses transcription initiation from the &PC1 (28), Chlumydomonas chloroplast PA (29), bacteriophage X p’R (30), and phage Mu pc-2 promoters (26). The five genes of the E. coli ilvGMEDA operon encode four of the five enzymes required for the biosynthesis of the branched-chain amino acids, isoleucine and valine (31, 32). The structural genes of this operon are preceded by tandem in vitro promoters, ilvPo1 and ilvPo2, and a leader-attenuator region (33-37). In himA strains of E. cob, the ilvA gene product of the ilvGMEDA operon, threonine deaminase, is expressed at about one-half the wild type level (38). Two IHFbinding sites have been reported in the ilvGMEDA promoterattenuator region (39). The binding of IHF to the DNA in the upstream ilvPc,l promoter region completely represses transcription from this promoter (28). Winkelman and Hatfield (40) have recently shown that the IHF-DNA interactions at the ilvPc1 target site are very similar to IHF-DNA interactions at other well characterized IHF-binding sites in the X at@ region (7, 9). Tsui and Freundlich (39) have shown that IHF protects a 39-base pair region from DNase I attack at a second IHF target site located in the iluGMEDA leaderattenuator region. Gel binding assays suggested that the Kapp
1986
IHF-mediated
Expression of the ilvGMEDA
value for IHF binding to its target site in the attenuator is 45-fold higher than the Kapp value obtained for IHF interaction at the upstream iluP~1 promoter region (39). In this report, we have used alkylation protection and hydroxyl radical “footprinting” techniques to show that IHF protein interacts with the bases of a consensus IHF target sequence, previously suggested by Tsui and Freundlich (39), located immediately downstream of the RNA polymerase transcriptional pause site in the ilvGMEDA leader-attenuator region. We also demonstrate that the binding of purified IHF protein to the leader-attenuator target site increases in vitro pausing of RNA polymerase at the transcriptional pause site. We have used both in vitro transcriptional studies and in vivo transcriptional fusions of various segments of the ilvGMEDA promoter-attenuator region with the galK gene to show: (i) IHF increases the basal level of transcriptional termination at the attenuator 2-fold, (ii) IHF interaction at the iluPGl target site activates transcription from the ituPo2 promoter 4-fold, and (iii) these IHF-mediated transcriptional effects are due to IHF protein interacting independently at the &PC1 promoter and leader-encoded target sites. Thus, IHF is capable, at least under certain conditions, of regulating the expression of the ilvGMEDA operon in opposing independent manners; IHF activates transcription of this operon 4-fold and can increase the termination of this transcription at the downstream attenuator site 2-fold. EXPERIMENTAL
PROCEDURES
Enzymes and Reagents-RNA polymerase was purchased from Boehringer Mannheim; restriction endonucleases and other enzymes were purchased from either Boehringer Mannheim or New England Biolabs, Inc. Deoxynucleoside triphosphates and were from Sigma unless otherwise noted. [w~‘P]GTP
[y-32P]dATP
(5000 Ci/mmol),
and [“Clgalactose
all other reagents (400 Ci/mmol),
(61 mCi/mmol)
were obtained from Amersham Corp. Purified integration host factor was prepared by the method of Nash et al. (41) or was a gift of H. Nash (National Institutes of Health). IHF preparations were estimated to be >95% pure as determined by sodium dodecyl sulfate gel electrophoresis (data not shown). Bacterial Strains and Plasmid Constructions-The E. coli strains and plasmids used in these studies are described in Table I. To isolate himA, single copy, ilu::galK fusion strains, the following bacteriophage Pl transduction experiments were performed. Strain CAG18518 (42), which contains a TnlOkan marker inserted into the host chromosome near the wild type himA locus at 37.5 min on the E. coli genetic linkage map, was infected with Pl bacteriophage, and a lysate was obtained (43). These phage were used to transduce the TnlOkan element into the him.4 strain K5242 (44). To ensure that the resulting kanamycin-resistant transductants retained the himA phenotype and that the wild type himA locus was not co-transduced with the TnlOkan element, several transductants were transformed with plasmid pSClO1 (45,46). This tetracycline-resistant plasmid can replicate only when functional IHF molecules are present in the cell (8, 44, 47). A potential himA strain containing the TnlOkan element was selected for growth on LB plates (48) supplemented with 35 pg/ml kanamycin. These cells were screened for the absence of growth on LB plates supplemented with I.0 fig/ml tetracycline. A himA TnZOkan transductant was used as the donor strain in a second generalized transduction experiment. Pl bacteriophage were used to infect the donor strain and the resulting transducing particles were collected after cell lysis. The recombinant transducing phage were used to infect the ampicillinand tetracycline-resistant, single copy, ilv::gaZK transcriptional fusion strains, IHl-10 and IH1384, which previously were transformed with plasmid pSC101. Co-transductants were isolated by: (i) positively selecting for growth in Luria medium (48) containing both ampicillin (50 pg/ml) and kanamycin (35 pg/ml) and (ii) screening for the failure of cells to replicate in Luria medium containing tetracycline (10 pg/ml). DNA Manipulations-Recombinant DNA techniques were performed by standard methods (43). Radiolabeled linear DNA fragments used in the Me,SO, and hydroxyl radical protection experiments were separated on nondenaturing 5% (w/v) polyacrylamide gels (19~1 ratio of acrylamide to N,N’-bisacrylcystamine) buffered
Operon
with TBE (50 mM Tris the gel containing the dissolved in 20% (v/v)
columns. Preparation
borate (pH 8.3), DNA fragments fi-mercaptoethanol,
1 mM EDTA). Portions of of interest were isolated, and purified on DE52
of chromosomal DNA (49), Southern blot anal-
yses, and hybridization reactions (43) were performed as described. End Labeling of DNA Fragments-uniquely 5’ end-labeled DNA restriction endonuclease fragments used in the Me*SO, and hydroxyl radical footprinting experiments were isolated as described above. Plasmid pJP1 was digested with Hind111 and EcoRI, and the 355base pair restriction endonuclease DNA fragment was isolated by standard methods (43). Following treatment with calf intestinal alkaline phosphatase, this DNA fragment was 5’ end-labeled with [y32P]dATP using T4 polynucleotide kinase (43). For analysis of the transcribed DNA strand, the 355-base pair 5’ end-labeled DNA fragment was digested with TaqI (ilvGMEDA base pair +7) (50) to yield a 288-base pair DNA fragment uniquely labeled at the EcoRI site from the pUC18 polylinker cloning region (51). For analysis of the nontranscribed (Fig. 1) DNA strand, the 5’ end-labeled 355-base pair HindIII-EcoRI restriction endonuclease fragment was digested with KpnI at the pUCl8 polylinker-encoded KpnI site to generate a 280-base pair DNA fragment uniquely 5’ end-labeled at the HindI site. MezS04 Protection Experiments-Me&Or protection experiments were performed as described by Siebenlist and Gilbert (52). Purified IHF (250 nM final concentration) was added to a loo-p1 binding solution (50 mM sodium cacodylate (pH 8.0), 40 mM Tris-HCl (pH 7.6), 10 mM MgQ, 100 mM KCl, 0.1 mM EDTA, 0.1 mh# dithiothreitol) containing 25 pg/ml sonicated herring sperm DNA, 25 rg/ml bovine serum albumin, and the ilvPo2 promoter-attenuator DNA fragment (containing ilvGMEDA sequences from base pairs -50 to +280). The ilvPo2 promoter-attenuator DNA fragment was endlabeled with Y-~‘P at a unique 5’ end as described above and included in the binding mixture at a concentration of approximately 1 X lo-* M. IHF protein was incubated with the DNA for 10 min at 37 ‘C, and the mixture was cooled to 0 “C. 1 pl of Me&O, (Aldrich) was added to the mixture and, after 15 min, the reactions were terminated by the addition of a 30-~1 solution of 1.5 M sodium acetate (pH 7.0) and 1.0 mM @-mercaptoethanol containing 100 pg/ml E. coli tRNA (Sigma). The DNA was precipitated with ethanol and treated according to the G > A chemical cleavage reaction methods of Maxam and Gilbert (53). Samples of the reaction mixture were fractionated in a 6% denaturing polyacrylamide gel (5.7% acrylamide, 0.3% N,N’methylenebisacrylamide) containing 8 M urea in TBE buffer and
FIG.
1. The DNA sequence of the leader region of the iluoperon. The nucleotide sequence of the nontranscribed DNA strand of the promoter-leader region of the ilvGMEDA operon is shown. The DNA sequence begins at a HhaI restriction site at nucleotide position -347. The nucleotides are numbered corresponding to the in vivo transcriptional start site from the ilvPo2 promoter. The transcriptional initiation site of the upstream in vitro promoter, ilvPo1, is indicated at nucleotide position -72. The conserved -10 and -35 hexamer nucleotide regions (75) of the ilvPo1 and ilvPG2 promoters are underlined. Underlined sections IA, lB, 2, 3, and 4 represent potential base-pairing regions of the leader RNA in the absence of translation (36). The sequence of the encoded leader polypeptide is shown using the one-letter amino acid symbols. Arrows indicate the approximate 3’ ends of pause and attenuated transcripts as determined from their mobility on denaturing polyacrylamide gels (50, 55).
GMEDA
IHF-mediated
Expression of the ilvGMEDA
examined by autoradiography following exposure of the gels to Kodak XAR-5 film at -70 “C in the presence of a Cronex Quanta III intensifying screen (Du Pant). Hvdrorvl Radical Protection ExDeriments-Interactions of IHF with”tbe bickbone of the DNA helid were determined as described by Tullius and Dombroski (54). 5’ end-labeled DNA fragments containing the leader-attenuator region (approximately 1 X 10m8 M) were incubated in a 167-~1 final volume of binding buffer. IHF (250 nM final concentration in the 167-~1 reaction mixture) was incubated with the DNA reaction mixture for 10 min at 37 “C. Following incubation, 10 ~1 each of fresh 20 mM sodium ascorbate, 0.6% hydrogen peroxide, and 0.2 mM iron-EDTA were added and the reaction mixtures were incubated for 2 min at 37 “C. The reaction was terminated with the addition of 20 ~1 of 0.1 M thiourea followed by precipitation with ethanol. Samples were analyzed by electrophoresis bn a 6% denaturingpolyacryla&ide gel (5.7% acrylamide, 0.3% N,N’methvlenebisacrvlamide) containing 8 M urea in TBE buffer. and the DNAfragments were examined b;I autoradiography as described above. In Vitro Transcription Assays-Zn vitro transcription assays (55) were performed with plasmids pJP29 or pJP30 as closed, circular, supercoiled DNA templates. For synchronous single round transcription reactions, with plasmid pJP1 as the DNA template, reaction volumes were increased to 150 ~1, and at each time point, a 10.~1 aliquot was transferred to a tube containing 10 ~1 of an 8 M urea, 0.1% sodium dodecyl sulfate, 0.025% bromphenol blue, 0.025% xylene cyan01 solution. RNA polymerase-DNA complexes were formed by preincubating 0.5 units (1.2 pmol) of RNA polymerase and 200 ng (0.1 pmol) of DNA in a 9-Fi solution (150 &IM KCl, 45 mM TrisHOAc (pH 7.9). 4.5 mM MirOAc. 0.1 mM dithiothreitol. 0.1 mM EDTA. 200 fiM‘ATP, 2b PM GTPyconiaining 6.4 units of Rkasin (Promegaj and 5-10 PCi of [a-“*P]GTP. Transcription reactions were initiated by the addition of 1 ~1 of a 2 mM UTP, 2 mM CTP solution. For single round transcription experiments, 100 rg/ml rifampicin was also included in the solution containing UTP and CTP. Reactions were terminated by the addition of an equal volume of an 8 M urea, 0.1% sodium dodecyl sulfate, 0.025% bromphenol blue, 0.025% xylene cyan01 solution. When present, the final concentration of IHF in the assay was 50 nM unless otherwise noted. After denaturation at 65 “C for 2 min, the transcription products (5-10 ~1) were analyzed by electrophoresis in a 6% polyacrylamidedenaturing gel (5.7% acrylamide, 0.3% N,N’-methylenebisacrylamide) prepared in 8 M urea buffered with TBE. The transcription reaction products were examined by autoradiography following exposure of the gels to Kodak XAR-5 film at -70 “C. Peak areas of transcript bands obtained over a linear range of multiple film exposure times were quantified with an LKB Ultrascan II Laser densitometer (model 2222-010). Construction of Chromosomal ilvPGlPG2 Attenuatorand ilvPG2 Attenuator-g&K Transcription Fusion Strains-The pK09-based g&K transcription fusion vectors, containing either the ilvPclPc2 promoters and the leader-attenuator region (i1vGMEDA sequences from base pairs -347 to +280), or just the ilvPo2 promoter and the leader-attenuator region (iluGMEDA seauences from base nairs -50 to +280), were integrated into the E. toll chromosome in single copy (56, 57). MluI-7’thlllI restriction endonuclease digestion. followed
by an “end fill in” reaction using the Klenow
frigmend
of DNA
polymerase I and subsequent religation of the plasmids, resulted in the formation of a 3’-truncated nonfunctional galK gene product. These galK truncated plasmid constructs were transformed into the polA mutant strain NO3434 and plated onto LB plates (48) containing 50 pg/ml ampicillin. The polA strain is unable to support the autonomous replication of ColEl-derived plasmids such as pK09 (58).
Thus, growth in the presence of ampicillin
results in the selection of
transformed cells that have integrated the plasmid into the chromosome by homologous recombination. This integration event results in the transcription of the nonfunctional g&K gene from the chromosomal gal promoter, whereas the functional g&K gene is transcribed from the inserted promoter contained on the plasmid. Southern blot hybridization was used to confirm single copy integration of the plasmid into the gal operon on the chromosome (57). Gafnctokinase Assays-For the galactokinase assays, strains were grown overnight to saturation and diluted 1:50 in 50 ml of M63 minimal medium (48) supplemented with 0.5% glucose, 1 mM MgSOd, and 5 rg/ml thiamine. The auxotrophies of individual strains were satisfied by the addition of 50 pg/ml of the appropriate amino acid. Ampicillin selection was performed at a concentration of 50 fig/ml.
Bacterial cultures were grown to an optical density (600 nm) of 0.6
1987
Operon
at 37 “C. Culture samples (10 ml) were collected and chilled on ice, and the cells were collected by centrifugation. The bacterial pellets were resuspended in a l-ml solution of 20 mM Tris-HCl (pH 7.5) and 2 mM dithiothreitol. Cell lysates were prepared by sonication with four 5-s bursts at 50 watts using a Tekmar sonic disrupter (model Tm250B). The extracts were clarified by centrifugation, and galactokinase activity (nmol of galactose l-phosphate/min) was assayed at 30 “C for 15 min as described (57). Total protein in the clarified extract was measured by the method of Bradford (59) using the Bio-
Rad protein assay kit according to the manufacturer’s and galactokinase min/mg of protein)
specific activity was determined.
(nmol
of galactose
instructions, l-phosphate/
RESULTS
Characterization
of the IHF-binding Site in the ilvGMEDA Leader-Attenuator Region
Me2S04 Protection Experiments-DNase I protection experiments have revealed the existence of an IHF-binding site in the ilvGMEDA leader-attenuator region located between base pair positions +lll and +150 (39). To define specific IHF protein-DNA interactions at this binding site, the Me2S04 methylation patterns of the DNA in this region were examined. An autoradiogram of the methylation pattern from both strands of DNA fragments containing the leader-attenuator region in the presence or in the absence of purified IHF protein is shown in Fig. 2. Fig. 3 summarizes the positions of purine residues that were either protected from methylation or showed an enhanced methylation pattern in the presence of IHF protein. As seen in Figs. 2A and 3 (bottom DNA strand), the transcribed DNA strand contains 3 guanine residues, at base pair positions +127, +135, and +146, and 5 adenine residues, at base pair positions +128, +129, +135, +142, and +143, that are protected from methylation by IHF. The methylation of a guanine residue at base pair position +150 is enhanced in the presence of IHF. The guanine residues at base pair positions +135, +146, and +150 are separated from the guanine residue at base pair position +127 by 9,19, and 23 base pairs, respectively. Therefore, since Me2S04 methylates the N7 position of guanine residues in the major groove of the DNA helix and the N3 position of adenine residues in the minor groove (52), these results show that IHF spans two adjacent major grooves on the same face of the DNA helix. However, most of the IHF contacts on this DNA strand are with adenine residues. Thus, as at other IHF target sites, most of the IHF protein-DNA contacts are located in the minor groove of the DNA helix. The Me,S04-mediated methylation pattern of the nontranscribed DNA strand, in the presence and absence of IHF protein, is shown in Fig. 2B and summarized in Fig. 3 (top DNA strand). Three adenine residues at base pair positions +134, +137, and +138 are protected from methylation in the presence of IHF. The N7 major groove-specific methylation of 2 guanine residues, at base pair positions +131 and +133, and the N3 minor groove-specific methylation of a single adenine residue, at base pair position +132, are enhanced by IHF addition. The pattern of methylation of this strand of the DNA also suggests that IHF interacts with adjacent major and minor grooves on one face of the DNA helix. It is of interest to note the occurrence of a DNA cleavage site at or adjacent to a thymine residue at base pair position +141 and a cytidine residue at base pair position +158 on the nontranscribed DNA strand in the presence of IHF (Fig. 2B). A similar pattern of DNA cleavage at a thymine residue was observed in the Me&O4 protection pattern for IHF interaction at the upstream ilvPo1 promoter region target site (40). While the reason for these unusual cleavages is unclear, it has been supposed that the altered DNA conformation wrought by the
IHF-mediated Expression of the ilvGMEDA Operon
1988
IHF
-A+ -B+
- t 155
,+’I85
- + I45
+115-
. 175 125-
-+
C
I
I35
. 165
1352
A
-IH:
B
G>A
IHC
-
170. 100-
160
110-
120150
7
3
1
1989
10 in Table 11). To determine if this decreased transcription into the galK gene mightbe due to the interaction of IHF at the leader-attenuator IHF-binding site, in vivo levels of galK expression from the ilvPc2-attenuat0r::galK transcriptional fusion strains, whichdo not contain the upstream iluPC1 promoter region IHF-binding site, were compared in the presence and in the absence of functional IHF protein expression. The data presented in Table I1 show that the interaction of IHF atthe leader-attenuator target sitedoes, in fact, cause a 2-fold decrease in galactokinase specific activity (compare the galactokinase specific activities in strains IH1020 and IH109 in Table 11). This implies that, under these conditions, IHF binding to the leader region reduces the level of transcriptional readthrough at the attenuator 2-fold.
I n Vitro Effect of IHF on E. coli RNA Polymerase Transcriptional Pause Time in the ilvGMEDA Leader Region The wild type ilvGMEDA leader RNA is capable of base140 pairing with itselfto form mutually exclusive alternate struc140tures (stem loops1:2,2:3, and 3:4)(34-36).One structure, stem loop 3:4, results in the formation of a p-independent 7 transcription terminator at the 3‘ end of the leader RNA. The formation of the antiterminatorstructure, stem loop 2:3, 130 precludes the formation of the terminator (36). Similar to 150results obtained from studies on the ilvBN (55), trp (60), his (61), pyrBI (62), and thr (63) attenuators, a strong in vitro transcription pause site in the ilvGMEDA leader has been documented at a site immediately followingthe DNA region encoding the 1:2 stem loop structure of the leader RNA (55). 160Mutations which affect the duration of RNA polymerase 120 pausing in the leader region of the trp operon alter the basal level of transcriptional readthrough at the trp attenuator (64). Mutations that decrease the RNA polymerase pause time increase transcriptional readthrough, at the attenuator, into the downstream structural genes. The intriguing position of the IHF consensus binding site immediately following the FIG.4. IHF protection of deoxyribose residues in the ilu- transcriptional pause site at nucleotide +117 in the leader GMEDA leader. Autoradiograms displayingthe positions of deox- region of the ilvGMEDA operon (Fig. 1) suggests that the yribose sugar residues protected from hydroxyl radical cleavage on binding of an IHF protomer to this site might affect RNA the nontranscribed ( A ) and transcribed ( B ) DNA strands, displayed polymerase pausing and, in turn, the level of transcriptional by brackets, in the presence and absence of IHF protein are shown. readthrough at the attenuator. Toascertain if IHF affects the A G > A sequence ladder is also displayedfor each DNA strand (53). rate of pausing of RNA polymerase in the ilvGMEDA leader Base pair positions, relevant to the in vivo transcriptional start site region, synchronized in vitro single-round transcription asfrom the iloP~2promoter, are shownnext to each panel. says, in the presence or absence of purified IHF, were performed using plasmid pJP1 asa closed, circular, supercoiled may alter the DNA conformation in and around the leader- DNA template. Under the conditions outlined under “Experattenuator target site by inducing a symmetrical bend in the imental Procedures,” this DNA template, which contains the DNA helix (9, 40). downstream ilvPc2 promoter but not the upstream ilvPC1 promoter, results in the production of a 117-nucleotide paused In Vivo Analysis of i1v::galK Transcriptional Fusion Strains RNA transcript and a 186-nucleotide full length attenuated To determine the potential regulatory effects of IHF on the RNA transcript. The addition of IHF protein (50 nM final expression of the ilvGMEDA operon, single copy chromosomalconcentration) tothe in vitro transcription reactions inilvPclPc2:galK, iluP~lPc2-attenuator::galK,and ilvPc2-at- creased the paused RNA transcript half-life 2-fold (from 0.5 tenuator::galK transcriptional fusions were constructed in to 1.0 min), yet the position of RNA polymerase pausing in wild type IHF+ (strains IH1384, IH1-10, and IH109 respec- the leader region remained unaltered (Figs. 5 and 6). Moretively (Table I)) and h i d IHF- (strains IH1-5, IH1-6, and over, the apparent rate of transcription elongation, as deterIH1020,respectively (Table I)) strains as described under mined by comparing the rates of appearance and disappearance of leader and paused RNA transcripts at each time point, “Experimental Procedures.” Thedata presented in Table I1show that the in vivo is unchanged by the addition of IHF protein. The appearance consequenceof IHF interaction at the upstream iluPG1 target of additional bands of unknown origin in the autoradiogram site is a 4-fold activation of galactokinaseexpression (compare (Fig. 5) are unaffected by the presence of IHF in these galactokinase specificactivities in strains IH1384 and IH1-5 experiments. in Table 11).However,in strains containing both the upstream In Vitro Characterizationof IHF-mediated Transcriptional ilvPcl promoter region IHF target site and the downstream Regulation of the ilvGMEDA Operon IHF target site in the leader-attenuator region, a less than 4fold increasein galactokinase expression is observed(compare To determine if the IHF-mediated increase in RNA polymthe galactokinase specificactivities in strains IH1-6 and IH1- erase pausing in the leader region of the ilvGMEDA operon
1
1
130-
IHF-mediated
Expression of the ilvGMEDA TABLE
I
List of bacterial strains and plasmid constructions Desianation Bacterial JMIOS C600K NO3434
Operon
used in this studv
Descriution
Ref.
strains
CAG18518 K5242 IHOS IH1384
IHl-10
IHlOS
IH2422
IHl-5
recA1, endAl, gyrA96, thi-I, hsdRl7, supE44, relA1, X-, Aclac-proAB), F-, thr-1, thi-1, leuB6, lacYI, galK2 tonA21, supE44, XSpontaneous strR mutant strs MG 1655” zdi-3123::TnZOkan
galK, rpsL, himAA81 galK- derivative of NO3434
of NO2383
Hfr
(same
origin
of chromosome
traD36, proAB, la&ZzhMIS]
[F’,
transfer
as HfrH),
lysA, polA1,
51 57 71
42 created
by the insertion of plasmid pKOSA3 into the chromosome of NO3434 by homologous recombination into the galK gene; galK-, lysA, polA1, stra An iluPolPo2::galK derivative of NO3434 created by the insertion of plasmid pCAPlP2 into the chromosome of NO3434 by homologous recombination into the galK gene; iluPoIPoP::galK, lysA,polAl, strR An iluPolPo2-attenuator::galK derivative of NO3434 created by the insertion of plasmid pJP639A3 into the chromosome of NO3434 by homologous recombination into the galK gene; iluPolPc2-attenuator::galK, lysA, polA1, strR An iluPo2-attenuator::galK derivative of NO3434 created by the insertion of plasmid pJPlOSA3 into the chromosome of NO3434 by homologous recombination into the galK gene; iluPo2-attenuator::galK, lysA, polA1, strR Strain created by a Pl-mediated transduction of a TnlOkan element from the donor strain, CAG18518, into the recipient strain K5242: strain was selected for Kans and screened for Tets after transformation with plasmid pSC101; zdi-3123::TnlOkan, galK, rpsL, himAA81 Strain created by a Pl-mediated co-transduction of TnlOkan element and himA locus from the donor strain, IH2422, into the plasmid pSClOl-transformed recipient strain IH1384; strain was selected for KanR and AmpR and screened for Tets; iluPolPoB::galK, lysA, polA1, stra, zdi-3123::TnIOkan, rpsL,
44 72 67
b
*
b
*
himAA81 IHl-6
IH1020
Strain created by a Pl-mediated co-transduction of TnlOkan element and himA locus from the donor strain, IH2422, into the plasmid pSClOl-transformed recipient strain IHl-10; strain was selected for and screened for Tets; KanR and Amps iluPolPc2-attenuator::galK, lysA, polA1, strR, zdi3123::TniOkan, rpsL, himAA81 Strain created bv a Pl-mediated co-transduction of TnlOkan element and himA locus from the donor strain, IH2422, into the plasmid pSClOl-transformed recipient strain IH109, strain was selected for KanR and AmpR and screened for Tets; iluPo2-attenuator::galK, lysA, polA1, strR, zdi-3123::TnlOknn,
b
b
rpsL, himAA81 Plasmids pCAP1P2 pCAPlP2A3 pJPllO0
pJP16
pJPl-10 pK09 pKOl1 pKOSA3 pJP629A3 pJP1 pJPlOSA3 psc101 pDD3 pJP29 pJP30
Plasmid contains a 520.bp SauDAl restriction fragment (ilu base pair positions, -384 to +136)’ ligated into the unique BamHI site of ~K04~ Plasmid constructed by digesting pCAPlP2 with Mu1 and TthlllI, “end-filling” at both restriction site 3’-recessed ends, and ligating. This results in the deletion of the 3’ end of the galK gene Plasmid constructed by ligating a l.l-kilobase AluI restriction fragment (ilu base pair positions, -459 to +643),’ containing the iluGMEDA upstream promoter and attenuator regions, into the SmaI polylinker site of pUC19. This results in ilu transcription in the same direction as transcription from the lccZ gene Plasmid consists of a 627-bp HhaI restriction fragment (ilu base pair positions -347 to +280)’ from pJP1100, containing the iluP~lPc2 promoters and attenuator, Klenow end-filled to blunt and ligated into the SmaI site of pUC18 Plasmid contains a 654-bp EcoRI-Sal1 restriction fragment of pJP16 ligated to EcoRI-Sal1 sites of M13mp8 A derivative of pKOld galK transcriptional fusion vector containing unique SalI, BamHI, and EcoRI sites upstream from the galK gene A derivative of pKO-Id galK transcriptional fusion vector with unique SmaI, BamHI, SalI, PstI, EcoRI, and Hind111 sites upstream of the galK gene Plasmid constructed by digesting pK09 with Mu1 and TthlllI, end-filling at both restriction site 3’recessed ends, and ligating. This results in deletion of the 3’ end of galK gene Plasmid consists of a 654-bn EcoRI-Sal1 restriction fragment from pJPl-10 inserted into the unique EcoRI and Sal1 sites of pKOSA3 Plasmid has a 344-bp HaeIII-EcoRI restriction fragment, containing the iluPo2 promoter and attenuator, from pJPl-10 ligated to end-filled SalI-EcoRI sites of plasmid pUC18 Plasmid consists of the 344-bp HaeIII-EcoRI fragment of pJPl-10 inserted into end-filled SaZI-EcoRI sites of DKGSA~ TetR, low copy number cloning vector Plasmid consists of a uniaue BamHI site flanked bv divergina rrnBTIT2 terminator sequences replacing pBR322 sequences extending from EcoRI to the Sal1 site (pBR322 base pair positions 4361-651) Plasmid contains a 344-bp HaeIII-EcoRI restriction fragment, containing the iluPo2 promoter and attenuator, from pJPl-10 ligated to the end-tilled BamHI site of plasmid pDD3 The 654-bp EcoRI-Sal1 restriction. fragment of pJPl-10 was inserted into the end-filled BamHI site of . . . plasmid pDD3 to construct this plasmid LIE. coli Genetic Stock Center (CGSC) b This work. ‘All ilu base pair positions are relative ’ Transcription fusion vectors pKO-1
strain
6300
to the in and pKO-9
as designated
uiuo initiation are described
by Bachman of transcription by McKenney
(74). from
the
et al. (58).
iluPo2 promoter
(34).
33 b 51*
51’
51b 33 58 * b 51’ b 45,46 72,73 * *
IHF-mediated Expression of the ilvGMEDA Operon TABLE I1 iluPclPc2::galK, ilvP~lPc2-attenwtor::galK, and ilvPc2attenuator::galK transcriptional fusions on the chromosome Strain”
Relevant genotype
Specific activity nmol Gal-I -PO,/ minjmg protein
1991
I\
Pause Transcript Disappearance
4.0
0.038 IH09 pKO9:galK IH1384 iluPclPc2:galK 61.8 f 3.5 12.8 f 1.5 IH1-5 iluPGlPGZ::galK, h i d IH1-10 iluP~lP~2-attenuator:~alK 17.0 f 0.3 iluPclPc2-attenuator::galK,h i d 6.2 f 0.9 IH1-6 IH109 iluPc2-attenuator::galK 2.3 f 0.4 4.9 f 0.7 IH1020 iluPc2-attenuator::galK,h i d a Galactokinase was assayed as described under “Experimental Procedures.” The galactokinase specific activity values are themean k S.D.of three separate experiments. The background value, measured by terminating reactions on dry ice a t time 0, was subtracted from the experimental value obtained in each experiment. Under these conditions, wild type strains grew with an average doubling time of 60 min, and h i d strains had an average doubling time of 78 min.
A
-IHF
180.
+IHF
B
0 2 0 4 0 00 80 100120140 160
0
2 0 40 80 80 100 120140 100
20
60140 100
180
Seconds
FIG. 6. Time course of paused RNA transcript disappearance. A semi-log plot of the disappearance of the 117-nucleotide paused RNA transcript during synchronized single round transcription reactions of a DNA template containing the ilvGMEDA leader and presence (0)of purified IHF protein region, in the absence (0) is shown. Peak areas of paused RNA transcript bands, obtained over
181
a linear range of multiple film exposure times, were quantified as described under “Experimental Procedures.” Each experiment was performed three times, and themean values of the relative peak area were plotted.
.
tide RNA transcript originating a t the iluP~2promoter (base pair position +l),which terminates at the attenuator, and a less abundant 336-nucleotide readthrough RNA transcript that also initiates at the iluP~2promoter but terminates 10& downstream of the attenuator site at the rrnB terminator. When purified IHF protein is present in the transcription FIG. 5. IHF-mediated transcriptional pausing in the leader reaction, the 336-nucleotide readthrough RNA transcript is region of the ilvGMEDA operon. The autoradiogram of a urea- reduced approximately 2.5-fold (Fig. 7,A and B ) . Thus, both polyacrylamide gel shown displays the kinetics of transcriptional pausing in the iluGMEDA leader region. In uitro transcriptions were in uiuo and in uitro, IHF decreases transcriptional readperformed at 23 “C. Transcription reaction time (seconds) is shown through at the attenuator of the ilvGMEDA operon approxia t the top of each lane. The positions of the 117-nucleotide paused mately 2.5-fold (Table I1 and Fig. 7). and the 186-nucleotide attenuated leader RNA transcripts are indiAn autoradiogram displaying the RNA transcription prodcated. The 108-nucleotide RNA transcript band originates from the ucts of plasmid pJP30 ( i ~ v P ~ ~ P ~ 2 - a t t e n u a t o r - r ~ n ~ ~ ~ ColEl origin of replication (ori) present on the plasmid. The origin which contains both the upstream and the downstream IHF of other RNA transcripts is unknown. target sites, is shown in Fig. 8A. In the absence of purified IHF protein, two attenuated RNA transcription products are is functionally related to the basal level of transcriptional observed a 186-nucleotide RNAtranscript originating at the readthrough at the downstream attenuator site, a series of in ilvP~2promoter (base pair position +1)and a 258-nucleotide uitro transcriptional experiments were performed.Additranscript originating at the iluPGl promoter (base pair positionally, in order to examine the functional relationship of tion -72). When purified IHF protein is present in the tranIHF interactions at the upstream iluPGl promoter region and scription reaction, the production of the ihPG1-initiated 258the downstream leader-encoded sites, DNA fragments from nucleotide attenuated RNA transcript is repressed, and the the iluGMEDA promoter-attenuator region containing both production of the ihP~2-initiated186-nucleotide attenuated IHF-binding sites, or just the downstream leader-attenuator RNA transcript is enhanced 4-5-fold. It is of functional IHF site, were inserted upstream of the strong rrnB termi- significance to note that IHF half-maximally represses trannator sequence contained in the plasmid pDD3 to yield the scription from the iluPGl promoter at a concentration of plasmid constructs pJP29 and pJP30 (Table I). Theseplasmid approximately 5 nM. This concentration is close to the KaPp constructs allow forthe detection of RNA transcription prod- value obtained from gel shift assays for IHF interaction at ucts that initiate at the iluGMEDA promoter(s) and terminate the iluPcl target site using the same purified IHF protein either at the attenuatoror at the downstream rrnB terminator preparation? The Kappdetermined from the results of these site. experiments compares favorably with the value reported by An autoradiogram displayingthe RNA transcription prod- Tsui and Freundlich (39). is shown ucts of plasmid pJ29 (iluP~2-attenuator-rmnBT~T~) A long exposure of an autoradiogram displaying the RNA in Fig. 7A. In the absence of purified IHF protein, two RNA transcription products are observed a prominent 186-nucleo* J. M. Pagel and G . W. Hatfield, unpublished results. ..
1992
IHF-mediated Expression of the ilvGMEDA Operon A
IHF o
25 5 0
nM
FIG. I. In vitro transcriptional analysis of the ilvGMEDA leaderattenuator region. In vitro transcriptions were performed a t 37 “C with plasmid pJP29 as a DNA template. IHF was allowed to prebind to theDNA template for 10 min, RNA polymerase was added, and the reaction was incubated for 10 min. A, autoradiogram of a urea-polyacrylamide gel in the absence of IHF protein or in the presence of 25 or 50 nM IHF protein. P ~ 2 a t tdenotes the full length 186-nucleotide attenuated RNA transcript originating from the iluPc2 promoter. &2rt represents the 336-nucleotide iluPc2 readthrough RNA transcript that terminates at the rrnBTITp terminators in the vector. B, schematic representation of the transcripts obtained either in the absence or presence of 50 nM IHF protein. Numbers a t the ends of the arrows denote the length, in nucleotides, of the RNA products. The black box depicts the IHF target site. Relative peak areas of RNA transcript bands, obtained over a linear range of multiple film exposure times, were quantified by laser densitometry. Relative peak area values were determined from mean values of relative peak areas obtained from three separate experiments normalized for the guanine content of each transcript and therelative intensity of the RNA transcript band from the ColEl origin of replication. The percentages of transcriptional readthrough a t the attenuator, obtained from experiments performed in the presence or absence of IHF, are mean values f S.D. calculated from these corrected data.
==2rt
336,
P&?att
e
4186
B
pJP29
-
rrnB T1T2
attenuator
Perce&ge of Ttanscrbtional Readthrough atthe Attenuator
-+le6
-IHF
5.3% % 1.0 4
+
3
3
6
+
3
3
6
-+la6
+IHF
M
2.2% % 0.4
site in the leader region and examined the IHF-mediated regulation of this operon both in vivo and in vitro. The results of the alkylation protection experiments, depicted in Fig. 2 and summarized in Fig. 3, showthat the IHF protein interacts specifically with the nucleotides (5”GATCAAGCCTTAA-3’)located between basepair positions +133 and +145 in the iluGMEDA leader region. This IHF-binding site sharessequence similarity with the IHF consensus DNA sequence (5’-WATCAANNNNTTR-3’) proposed by Friedman (4).Moreover, these experiments demonstrate that IHF interacts with the DNAhelixin this regionin a manner similar to theway that IHFinteracts with the well characterized IHF-binding sites in the bacteriophage X attP (7,9)and E. coli iluPcl regions (40).The data summarized in Fig. 3 further show that, aswith other IHF-binding sites, the interDISCUSSION action of IHF at the target site in the leader region occurs Previous work has shown that IHF interacts with two primarily with chemical groups exposed inthe minor groove separate target sites in the ilvGMEDA promoter-attenuator of the DNA helix (7-9). Yang and Nash (9) detected hypersensitive cleavage sites region: an upstream site in the iluPcl promoter region (28, of DNA 39) and a downstream site in the leader region (39). Recent at deoxyribosesugar residues located between regions results from our laboratory showed that IHFbinds to a near protected from hydroxyl radical attack. These protected reconsensus DNA sequence in the upstream iluPGl promoter gions define the interactions of the backbone of the DNA region and that the IHF-DNA interactions at thissite are the helix with the IHFprotein or distinguish extruding bases due same as at other well characterized IHF-binding sites involved to alterations in the DNA helix, or both. Yang and Nash in replication and recombination (4,7-9, 40). In this study, proposed that the intervening regions are hypersensitive to we have characterized the interaction of IHF at the target hydroxyl radical attack because the backbone of the DNA
transcription products of plasmid pJP30 shows that three prominent RNA transcription products are detectable (Fig. 8B). In addition to the 186- and 258-nucleotide attenuated RNA transcripts, a 336-nucleotide readthrough RNA transcript initiating from the iluPG2 promoter is also detectable in this longer exposure.The data from densitometric scanning (Fig. 8C) of the 336-nucleotide RNAtranscript indicate that the amount of readthrough transcription terminating at the rrnB terminator site increases almost 2-fold with the addition of purified IHF protein. This IHF-mediated increase in overall readthrough transcription is the combined result of a 4-5-fold activation of transcription initiation from the iluPG2 promoter and an independent 2-2.5-fold increase in transcription termination at the attenuator.
IHF-mediated Expression 'A
ilvGMEDA of the
1993
IHF..O...5_50.nM
258,
FIG. 8. IHF-mediated activation, repression, and attenuation in the ilvGMEDA promoter-leader regulatory region. In vitro transcriptions were performed at 37 "C with plasmid pJP30 as a DNA template. IHF was allowed to prebind to the template for 10 min, RNA polymerase was added, and the reaction was incubated for 10 min. A, autoradiogram of a urea-polyacrylamide gel in the absence of IHF protein or in the presence of 5 or 50 nM IHF protein. Pc2att denotes the position of the full length, 186-nucleotide attenuated RNA transcript that originates from the iluPc2 promoter. Pclatt denotes the positions of the 258-nucleotide attenuated RNA transcript that origi- B nates from the iluPclpromoter. B, a long exposure of an autoradiogram displaying the RNA transcription products of plasmid pJP30. Transcription reactions depicted in the four lanes contained 0, l, 10, and 50 nM concentrations of IHF protein, respectively. The bands marked P ~ 2 r t represent the 336-nucleotide iluP~2 originating readthrough RNA transcripts that terminate at the rrnBTIT*terminators in the vector. C, schematic representation of the RNA transcription products from reactions performed in the absence or presence of purified IHF protein (50 nM). Numbers a t theends of the arrows denote the length, in nucleotides, of the RNA products. The black boxes depict theIHF target sites. Relative peak areas of RNA transcript bands, obtained over a linear range of multiple film exposure times, were quantified by laser densitometry. The relative peak area values presented arethe mean values f S.D. obtained from threeseparate experiments normalized for the guanine content of each transcript and the relative intensity of the RNA transcript band from the ColEl origin of replication. N.D., not detected.
Operon
PGlatt
-
IHF o 1 1050 nM
k2rt
436
258,
pJP30
C "It
Po'
Po2
-lo5
Js -I0
+l
-
attenuator
T l T2
rel. Deak area
I+ 24.9 % 1.4 17.9 % 1.7 1.8 % 0.5
N.D.
116.1 +/- 6.2
