Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
FEMS Immunology and Medical Microbiology 28 (2000) 25^35 www.fems-microbiology.org Identi¢cation and characterization of mutants with increased expression of hilA, the invasion gene transcriptional activator of Salmonella typhimurium Thomas F. Fahlen, Nitin Mathur, Bradley D. Jones * Department of Microbiology, University of Iowa School of Medicine, 3-330 Bowen Science Bldg., 51 Newton Rd., Iowa City, IA 52242-1109, USA Received 8 October 1999; accepted 8 December 1999 Abstract Induction of invasion gene transcription and expression of the invasive phenotype of Salmonella strains are regulated by environmental conditions. Experimental evidence indicates that oxygen, pH, and osmotic conditions need to closely resemble those of the host intestinal lumen for invasion gene activation. The hilA gene, encoded on Salmonella pathogenicity island 1 (SPI-1), is a transcriptional activator which is required for invasion and whose expression is modulated by oxygen, pH, and osmolarity. Additionally, hilA is regulated by genetic elements encoded on SPI-1 (hilC/sirC/sprA and hilD), as well as by elements which reside outside of SPI-1 (phoP/phoQ and sirA), although how environmental signals modulate hilA is unknown. In an effort to further characterize the Salmonella invasion gene regulon, we have created and preliminarily characterized 18 Tn5 insertions which result in upregulation of a hilA: :lacZY fusion. We have classified the mutations based on location and phenotype into three classes. Six class 1 and six class 2 mutants have insertions in SPI-1 near the invasion gene orgA or the invasion gene regulator hilD, respectively. Six class 3 mutants reside outside of SPI-1 in four different loci. The class 2 and 3 mutations induce overexpression of an episomal hilA: :lacZY fusion and significantly increase S. typhimurium invasion of HEp-2 cells in a standard invasion assay. These data implicate new regions of SPI-1 as being involved in the regulation of invasion by S. typhimurium and identify new invasion gene regulators located outside of SPI-1. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Salmonella typhimurium ; Invasion gene regulation ; hilA 1. Introduction Pathogenic Salmonella typhimurium strains cause two types of infectious disease. In healthy humans the bacteria cause a self-limiting gastroenteritis of the small intestine while infection of mice leads to a systemic enteric fever [1]. Since S. typhimurium infections in mice closely resemble human enteric fever, murine infection models are commonly used as models of typhoid fever. The pathogenic strategy of enteroinvasive Salmonella species requires that the bacteria invade host intestinal cells to initiate disease. Infection begins when organisms invade the apical surface of enterocytes in the small intestine [2^4]. The entry process occurs via actin rearrangement and ru¥ing of the cell membrane that results in cellular projections which engulf the bacteria [5]. Invasion causes destruction of the cell which allows the bacteria to proceed to submucosal sites * Corresponding author. E-mail : [email protected] where they are taken up by macrophages and, in an appropriate host, persist within the phagolysosomal compartment. Survival within circulating macrophages allows dissemination via the lymphatic system to the liver and spleen of the host causing systemic infection (enteric fever). Genetic elements responsible for the invasive phenotype of Salmonella have been identi¢ed and localized to a 40-kb chromosomal region at centisome 63 termed Salmonella pathogenicity island 1 (SPI-1) (reviewed in [6]). The SPI1 genes are predicted to encode products with similarity to the structural components of type III secretion systems found in both plant and animal pathogens including Pseudomonas, Rhizobium, Erwinia, Yersinia, Shigella, and enteropathogenic Escherichia coli [7]. Type III secretion systems are composed of 20 or more proteins which allow e¡ector proteins to be secreted across the inner and outer membranes of Gram-negative bacteria and are typically involved in the translocation of virulence proteins to eukaryotic cells [8]. In addition to virulence loci within 0928-8244 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 9 2 8 - 8 2 4 4 ( 9 9 ) 0 0 1 9 9 - 6 FEMSIM 1177 5-4-00 26 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 SPI-1, loci outside of SPI-1, which encode proteins secreted via the type III secretion apparatus of SPI-1, have been found to induce cellular changes in tissue culture cells and/or to a¡ect the ability of Salmonella to invade tissue culture cells [9^13]. The expression of the invasion-associated type III secretion apparatus and the secreted e¡ector proteins are tightly regulated in Salmonella in response to environmental conditions including osmolarity, oxygen, and pH such that all conditions must be appropriate to allow invasion gene expression [14^18]. The SPI-1-encoded hilA gene encodes a transcriptional activator which modulates expression of the type III secretion apparatus and/or the secreted e¡ector proteins [19]. The modulation of hilA expression appears to be the primary system for regulation of the invasive phenotype of Salmonella since overexpression of hilA confers a hyperinvasive phenotype and counteracts the e¡ects of invasion-repressing signals [14,19,20]. Results in our laboratory and others [21,22] show that null mutations in hilA result in dramatic attenuation of invasion demonstrating the importance of hilA in the invasion phenotype. It appears that the signi¢cance of hilA in the invasion phenotype stems from its role in the activation of expression of components of the type III secretion apparatus as well as the secreted proteins encoded on SPI-1 [14,23]. As invasion gene regulation is studied more extensively in Salmonella, it is evident that a complex regulatory scheme exists to allow expression of invasion genes only under optimal conditions. The elements involved in this scheme and their roles in regulation are not well understood. In an e¡ort to elucidate pathways involved in invasion gene regulation, we conducted a search for genetic elements involved in the modulation of hilA expression. We describe here the identi¢cation of 18 mutations created through random Tn5 mutagenesis which result in increased expression of a chromosomal hilA: :lacZ transcriptional fusion. 2. Materials and methods 2.1. Bacterial strains and growth conditions Bacterial strains and plasmids used in this study are listed in Table 1. Liquid cultures were grown in Lennox broth (LB) (Gibco BRL) with appropriate antibiotics added to the following ¢nal concentrations (Wg ml31 ): ampicillin, 100; kanamycin, 50; streptomycin, 100; and tetracycline, 20. MacConkey lactose agar (Difco) was used to assay the L-galactosidase phenotype of S. typhimurium reporter strains. High and low oxygen bacterial growth conditions were used to repress or induce invasion gene transcription. High oxygen cultures were obtained by inoculating 3 ml of LB broth with 10 Wl of a stationaryphase culture and growing vigorously at 200 rpm at 37³C to early exponential phase (V1U108 CFU ml31 ). Low oxygen cultures were initiated by inoculating 3 ml of LB broth with 10 Wl of a stationary-phase culture and growing the bacteria statically overnight at 37³C to a density of V4^5U108 CFU ml31 [21,24]. 2.2. Plasmid construction The plasmid pJK001 encodes the prgH, -I, -J, -K and orgA genes with the native prg promoter. The plasmid was constructed by combining portions of pBDJ116 and pBDJ134 [25]. Brie£y, a HindIII-BamHI fragment from pBDJ116 that contains prgH, prgI and the ¢rst 185 bp of prgJ was cloned into pACYC177 cut with HindIII and BamHI and designated pBDJ233. A second plasmid, pBDJ234, was constructed with the terminal portion of prgJ, and intact prgK and orgA genes. In addition, pBDJ234 carries a kanamycin cassette just downstream of the orgA gene. Digestion of pBDJ234 with BamHI and BclI released a fragment containing the last 118 bp of prgJ, prgK, orgA and the kanamycin gene which was cloned into pBDJ233 digested with BamHI to create pJK001. Creation of the desired plasmid was veri¢ed by restriction enzyme digestion and agarose gel electrophoresis. 2.3. Mutagenesis of S. typhimurium with Tn5 Transposon mutagenesis of the S. typhimurium hilA: : lacZY reporter strain BJ57 was performed using the plasmid pRTP1: :Tn5. This plasmid harbors the Tn5 transposon (kanR ), the rpsL gene, which confers sensitivity to streptomycin, and an ampicillin resistance marker. In BJ57, which is streptomycin-resistant due to a mutation in the chromosomal rpsL gene, pRTP1: :Tn5 imparts resistance to kanamycin and ampicillin yet the strain is streptomycin-sensitive because the rpsL allele on the pRTP1 : :Tn5 plasmid is dominant to the chromosomal rpsL gene. To create random chromosomal Tn5 insertions, BJ57 was transformed with pRTP1 : :Tn5, several hundred colonies were plated onto each of 10 plates, and the colonies were replica-plated to LB agar containing streptomycin and kanamycin. Growth on this medium results from transposition of Tn5 to the chromosome and subsequent loss of pRTP1: :Tn5. The bacteria that grew on each of the 10 selective plates were collected to form 10 separate pools. At least 3000 colonies from each of the 10 pools of streptomycin- and kanamycin-resistant bacteria were plated onto MacConkey agar to identify mutants with deviations in the `¢sheye' phenotype of BJ57. 2.4. Identi¢cation of Tn5 insertion sites The site of the Tn5 insertion in each mutant, except BJ749 and TF43, was determined by cloning the kanamycin resistance marker and £anking DNA from the chro- FEMSIM 1177 5-4-00 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 27 Table 1 Bacterial strains, plasmids, and phage used in this study Strain/plasmid Strain E. coli DH12S S. typhimurium SL1344 EE251 BJ35 BJ57 BJ606 BJ749 TF25 TF26 TF27 TF28 TF29 TF30 TF37 TF38 TF39 TF40 TF41 TF42 TF43 TF44 TF45 TF46 TF47 TF48 TF49 TF50 TF51 TF52 TF53 TF54 TF55 TF56 TF57 TF58 TF59 TF60 TF61 TF62 TF63 TF64 TF65 Plasmid pJK001 pLS31 pRTP1: :Tn5 Relevant genotype Reference/source mcrA v(mrr-hsdRMS-mcrBC) FP lacIq lacZvM15 Gibco BRL wild-type virulent strain invasive LT2 derivative, rpsL EE251 pepT: :Mud1(x) vzda88Tn10 EE251 hilA: :Tn5lacZY EE251 hilA: :Tn5lacZY prgH3 : :Tn5 EE251 prgH3: :Tn5 EE251 hilA: :Tn5lacZY prgK1 : :Tn5 EE251 hilA: :Tn5lacZY orgA1 : :Tn5 EE251 hilA: :Tn5lacZY orgA2 : :Tn5 EE251 hilA: :Tn5lacZY orgA3 : :Tn5 EE251 hilA: :Tn5lacZY orgA4 : :Tn5 EE251 hilA: :Tn5lacZY orgA5 : :Tn5 EE251 hilA: :Tn5lacZY prgH4 : :Tn5 EE251 hilA: :Tn5lacZY prgH5 : :Tn5 EE251 hilA: :Tn5lacZY prgH6 : :Tn5 EE251 hilA: :Tn5lacZY prgH7 : :Tn5 EE251 hilA: :Tn5lacZY prgH8 : :Tn5 EE251 prgH4: :Tn5 EE251 prgH5: :Tn5 EE251 prgH6: :Tn5 EE251 prgH7: :Tn5 EE251 prgH8: :Tn5 EE251 hilA: :Tn5lacZY ams1: :Tn5 EE251 hilA: :Tn5lacZY hupB1: :Tn5 EE251 hilA: :Tn5lacZY pag1 : :Tn5 EE251 hilA: :Tn5lacZY hilE3: :Tn5 EE251 hilA: :Tn5lacZY hilE2: :Tn5 EE251 hilA: :Tn5lacZY hilE1: :Tn5 EE251 ams1: :Tn5 EE251 hupB1: :Tn5 EE251 pag1: :Tn5 EE251 hilE3: :Tn5 EE251 hilE2: :Tn5 EE251 hilE1: :Tn5 EE251 pepT: :Mud1(x) EE251 pepT: :Mud1(x) ams1: :Tn5 EE251 pepT: :Mud1(x) hupB1 : :Tn5 EE251 pepT: :Mud1(x) pag1: :Tn5 EE251 pepT: :Mud1(x) hilE3: :Tn5 EE251 pepT: :Mud1(x) hilE2: :Tn5 EE251 pepT: :Mud1(x) hilE1: :Tn5 [49] [25] This [21] This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This This pACYC derivative encoding prgH, -I, -J, -K and orgA, ampR , kanR low copy plasmid with a hilA-lacZY fusion, tetR Tn5 transposon inserted onto plasmid pRTP1 which carries the rpsL gene (strS ) ampR , kanR Jones Laboratory C. Lee [50,25] mosome into the plasmid vector pBluescript (Stratagene). Approximately 0.5 Wg of chromosomal DNA was isolated using a Dneasy tissue kit (Qiagen), digested with BamHI, and ligated into the unique BamHI site of pBluescript. Ligated DNA was then used to transform E. coli DH12S (Gibco BRL) to kanamycin resistance via electroporation. Because of a unique BamHI site in Tn5, kanamycin-resistant clones carry the kanamycin resistance marker and chromosomal DNA £anking the Tn5 insertion site. These plasmids were isolated from DH12S and used work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work work as template with the primer oxregA (5P-TTCAGGACGCTACTTGTG-3P) in a sequencing reaction to determine the chromosomal sequence £anking the Tn5 insertion site. Due to di¤culty in cloning the TF43 insertion site, PCR was used to determine the location of Tn5 in this mutant. The general area of insertion was ¢rst identi¢ed using PCR with the primers oxregA (5P-TTCAGGACGCTACTTGTG-3P) and prgH3P (5P-AGATCTAAAGTGGGCTTGGGAAATAC-3P) which ampli¢ed a 1.9-kb band using TF43 chromosome as template. This band was ex- FEMSIM 1177 5-4-00 28 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 cised from an agarose gel and the DNA was eluted using a gel extraction kit (Qiagen). The puri¢ed PCR product was then sequenced as above using the oxregA primer to direct sequencing o¡ the end of the Tn5 segment into the ampli¢ed £anking DNA. Inverse PCR [26] was used to identify the location of the Tn5 insertion in S. typhimurium BJ749. Chromosomal DNA from BJ749 was isolated using the technique of Marmur [27] and V1 Wg was digested to completion with BamHI, which cuts within the Tn5 transposon. The BamHI-digested chromosomal DNA was diluted 1:100 in dH2 0 and ligated overnight at 16³C to form closed circular DNA. The ligated DNA was used as template in a 50 Wl Expand Long Template PCR reaction (Promega). The primers Tn5B21L (5P-GTTCAGGACGCTACTTGTG-3P) and Mathur (5P-AGGACGACGAGGCTTGCA-3P) anneal to sequences adjacent to the BamHI site of Tn5 and direct DNA synthesis away from each other. PCR ampli¢cation with these primers resulted in ampli¢cation of a 4.5-kb product which was ligated into pGEM-T (Promega) and electroporated into E. coli DH12S. Plasmid DNA was isolated from white colonies on X-gal plates and the presence of an insert in the vector was veri¢ed by restriction enzyme digestion and agarose gel electrophoresis. Plasmid DNA was used as template with Tn5B21L primer (5P-GTTCAGGACGCTACTTGTG-3P) to direct sequencing into DNA £anking the Tn5 insertion site. DNA sequencing was conducted by the DNA core facility at the University of Iowa. Sequences obtained were used to search DNA databases available at the National Center for Biotechnology Information (www.ncbi.nlm. nih.gov), the Sanger Centre (www.sanger.ac.uk), and the Genome Sequencing Center at the Washington University School of Medicine (genome.wustl.edu/gsc/bacterial/salmonella.shtml). 2.5. P22-mediated transductions Mutations marked by antibiotic resistance genes were moved by transduction with P22 HT int3 as previously described [28]. Transductants were selected on LB plates with appropriate antibiotics and EGTA to prevent secondary infections by P22. 2.6. L-Galactosidase assays L-Galactosidase production from lacZ reporter constructs was quantitated using the method of Miller [29] after growth of the bacteria under high or low oxygen conditions. 2.7. Tissue culture conditions and invasion assays HEp-2 tissue culture cells [30] were maintained in RPMI 1640 (Gibco BRL) containing 10% (v/v) fetal bovine serum (FBS) and passaged every 2^3 days. Invasion assays were done as previously described [24]. Brie£y, wells of a 24-well dish containing 105 HEp-2 cells per well seeded the previous day were inoculated with 107 bacteria. Bacteria were allowed to invade for 1 h before washing each well with PBS three times. One ml of RPMI 1640 medium containing 50 Wg ml31 of gentamicin was added to each well and the plates were incubated at 37³C for 90 min. Wells were then washed with PBS again before lysing the cells by the addition of 100 Wl of 1% Triton X-100 to each well. After incubation at 37³C for 10 min, 400 Wl of LB was added to each well and dilutions from each well were plated to determine the number of intracellular bacteria. The percent invasion is de¢ned as the number of gentamicin-resistant bacteria recovered from the HEp-2 cells divided by the number of bacteria in the inoculum multiplied by 100. 3. Results 3.1. Identi¢cation of S. typhimurium mutants with altered hilA expression To identify mutants with aberrant hilA expression, we took advantage of the MacConkey agar phenotype of S. typhimurium strain BJ57 which carries an oxygen-regulated hilA: :Tn5lacZY transcriptional fusion. Colonies of this strain have a `¢sheye' phenotype on MacConkey agar (red center and a white periphery) due to the expression of the lacZ fusion in the low oxygen environment at the center of the colony [25,31]. We were interested in isolating solid red mutants on MacConkey agar since these may carry disruptions in negative regulators of hilA. S. typhimurium BJ57 was mutagenized with a plasmid-mediated transposon delivery system that allowed us to create Tn5 insertions randomly throughout the S. typhimurium chromosome. Ten separate pools of random Tn5 chromosomal insertions were created and screened by plating for isolated colonies on MacConkey agar. More than 15 000 colonies were screened which yielded 30 mutants (three from each pool) that exhibited a darker red phenotype on MacConkey agar compared to wild-type. To con¢rm a single Tn5 insertion as the cause of the altered phenotype, the Tn5 was moved from each mutant to S. typhimurium EE251 (wild-type) and then back into BJ57 via P22 phage transduction. Strains isolated in this manner contained single Tn5 insertions which a¡ected hilA: :lacZ expression and were characterized further. 3.2. The class 1 mutants cluster near the orgA invasion gene The insertion site of Tn5 in each of the mutants was determined by cloning kanamycin resistance and £anking DNA (see Section 2). Sequence analysis of the mutants revealed that several of the transposons had inserted near orgA, a gene encoded on SPI-1 which is required FEMSIM 1177 5-4-00 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 29 for invasion and the secretion of virulence proteins [21,25]. This group of mutants has been designated class 1. Although these mutants were isolated as colonies which appeared darker red than wild-type on MacConkey agar, the hilA: :lacZ expression of these strains as measured in a standard L-galactosidase assay was not signi¢cantly di¡erent from wild-type (data not shown). Therefore, these mutations were not characterized further. 3.3. The class 2 mutants cluster upstream of the hilA derepressor hilD and cause increased expression of chromosomal and episomal hilA : :lacZ Six Tn5 insertions termed class 2 were found to cluster in the region of SPI-1 encoding the prgH operon and the newly described hilD gene [32,33] (Fig. 1B). Five of these insertions lie within the prgH, -I, -J, -K and orgA operon ([32] and J. Klein, manuscript in preparation) while the sixth insertion in this region lies between the prgH and hilD genes and as such does not disrupt either of these protein coding regions. The class 2 mutations alter the normal expression of the chromosomal hilA: :lacZ fusion which is typically induced approximately 2^3-fold by growth in low versus high oxygen conditions. Increases caused by these mutations range from 0- to 3-fold high oxygen and 0.5- to 14-fold in low oxygen (Fig. 1A). To further characterize the class 2 mutants, strains containing the mutations in an otherwise wild-type back- Fig. 1. Location of the class 2 mutations and their e¡ect on chromosomal hilA: :lacZ expression. A: Chromosomal hilA : :lacZ expression in class 2 mutant backgrounds. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and assays were conducted to quantitate L-galactosidase activity (see Section 2). BJ57 is the hilA: :lacZ parent strain. Data presented represent the averages of at least three independent experiments. Error bars depict the standard deviation from the mean. B: Partial map of SPI-1 showing the location of the class 2 mutant Tn5 insertions in the S. typhimurium chromosome. The distances in nucleotides of the insertions to the ¢rst nucleotide of the HilD coding sequence are as follows : TF37, 936; TF38, 811 ; TF39, 664; BJ606, 652, TF40, 349 ; TF41, 99. Fig. 2. E¡ect of the class 2 mutations on expression of an episomal hilA : :lacZ fusion. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and assays were conducted to quantitate L-galactosidase activity (see Section 2). EE251 is the wild-type parent strain. Plasmid pLS31 encodes a hilA: :lacZ fusion in the low copy vector pRW50 [33]. Data presented represent the averages of at least three independent experiments. Error bars depict the standard deviation from the mean. ground were transformed with the low copy hilA: :lacZ reporter plasmid, pLS31 [33]. As shown in Fig. 2, these mutations cause a signi¢cant increase in hilA: :lacZ expression from pLS31 demonstrating that the mutations can a¡ect hilA expression in trans. 3.4. Invasion by class 2 mutants is not repressed by high oxygen conditions Overexpression of hilA has been shown to result in a hyperinvasive phenotype for tissue culture cells [20]. Since the class 2 mutations cause a signi¢cant increase in hilA expression, we were interested in examining the e¡ect of these mutations on the invasion phenotype. As ¢ve of the class 2 mutations disrupt the prgH, -I, -J, -K and orgA genes, which are required for invasion, strains containing these mutations are attenuated for invasion in a standard invasion assay of HEp-2 cell monolayers (data not shown). However, when these strains are transformed with plasmid pJK001, which encodes the prgH, -I, -J, -K, and orgA, genes, low oxygen invasion is restored to levels near wild-type and high oxygen invasion is 10^100fold above wild-type, with the exception of TF43 and TF45 (Fig. 3). The increase in invasion (high or low oxygen) by strains TF42, TF44, BJ749, and TF46 is not due to the presence of pJK001 since invasion by the wild-type strain containing this plasmid is still repressed by growth in high oxygen (Fig. 3). One of the class 2 mutants, TF46, has a Tn5 insertion between prgH and hilD that does not disrupt genes required for invasion (Fig. 1B). Similar to the other class 2 mutants with pJK001 present, TF46 is nearly 100-fold more invasive under high oxygen conditions than the wild-type strain (Fig. 3). Taken together, these data indicate that four of the class 2 mutations result in a decreased ability to respond to the invasion-repressing growth condition of high oxygen and that the decreased ability is not due to the disruption of prgH -I, -J, -K or orgA. FEMSIM 1177 5-4-00 30 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 Fig. 3. E¡ect of the class 2 mutations on invasion of HEp-2 cell monolayers. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and invasion assays were conducted as described in Section 2. EE251 is the wild-type parent strain. Data presented represent the average of at least three independent experiments with duplicate wells. Error bars depict the standard deviation from the mean. Fig. 4. E¡ect of the class 3 mutations on expression of chromosomal hilA : :lacZ. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and assays were conducted to quantitate L-galactosidase activity (see Section 2). BJ57 is the hilA : :lacZ parent strain. Data presented represent the average of at least three independent experiments. Error bars depict the standard deviation from the mean. 3.5. The class 3 mutations map outside of SPI-1 ducted on strains harboring chromosomal or episomal hilA: :lacZ fusions. The class 3 mutations cause small effects on chromosomal hilA: :lacZ expression under noninducing conditions but up to 3-fold increases of hilA : : lacZ under inducing conditions as measured by L-galactosidase assay (Fig. 4). The ams mutation and one of the mutations in hilE appear to express hilA: :lacZ only slightly higher than wild-type. It should be noted that the hilE3: :Tn5 mutation in TF50 is at the extreme 3P end of the putative ORF and thus the mutation may only partially disrupt activity of the protein. The other strains with Tn5 insertions in hilE, TF51 and TF52, cause disruptions in the 5P end and result in a 1.5-fold increase in chromosomal hilA: :lacZ expression under our low oxygen assay conditions (Fig. 4). To further characterize the e¡ect of the class 3 mutations on hilA expression, we transformed mutant strains with the hilA: :lacZ reporter plasmid pLS31 and measured L-galactosidase production. Interestingly, the patterns of hilA: :lacZ expression in high and low oxygen di¡er in Six Tn5 insertions a¡ecting hilA: :lacZ expression were determined to reside outside of SPI-1 and were designated class 3 mutations. The S. typhimurium genes disrupted by these insertions were identi¢ed by using the DNA sequence £anking the transposon to search databases at the National Center for Biotechnology Information (NCBI) or the Sanger Centre S. typhi sequencing project (www.sanger.ac.uk). Sequences identi¢ed in this manner were examined for open reading frames (ORFs) disrupted by the Tn5 using a DNA sequence analysis program. Homologs of each of the genes identi¢ed in S. typhi were found in S. typhimurium by using the S. typhi sequence to search the S. typhimurium genome database available from the Washington University School of Medicine (genome.wustl.edu/gsc/bacterial/salmonella.shtml). One of the strains (TF47) was found to have Tn5 inserted at amino acid 194 of the S. typhimurium ams gene, which, in E. coli, encodes the endoribonuclease RNase E [34]. Another mutant (TF48) was found to have a Tn5 insertion 188 bp upstream of the putative translational start of the hupB gene which encodes the L subunit of the heterodimeric histone-like protein HU [35]. A third insertion site (in TF49) was determined to be in a small ORF of unknown function between the pagK and pagM genes which we refer to here as pag [36]. Interestingly, the remaining three class 3 mutations (in TF50, TF51, and TF52) reside in a previously uncharacterized ORF that we propose be termed hilE. Computer analysis of the hilE gene predicts that it encodes a soluble protein but a secondary structure analysis of the putative protein showed no helix turn helix or coiled coil motifs. 3.6. The class 3 mutations upregulate chromosomal and episomal hilA: :lacZ fusions To quantitate the e¡ects of the class 3 mutations on hilA : :lacZ expression, L-galactosidase assays were con- Fig. 5. E¡ect of the class 3 mutations on expression of an episomal hilA : :lacZ fusion. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and assays were conducted to quantitate L-galactosidase activity (see Section 2). EE251 is the wild-type parent strain. Plasmid pLS31 encodes a hilA : :lacZ fusion in the low copy vector pRW50 [33]. Data presented represent the average of at least three independent experiments. Error bars depict the standard deviation from the mean. FEMSIM 1177 5-4-00 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 these mutants when the hilA: :lacZ reporter fusion is encoded on the plasmid rather than on the chromosome (Fig. 5). Among these di¡erences is the loss of oxygen regulation of hilA: :lacZ on the plasmid in the ams1 : :Tn5 background of TF53. Additionally, plasmid encoded hilA : :lacZ in the hupB1: :Tn5 background is expressed at levels above wild-type in high oxygen growth conditions while chromosomal hilA: :lacZ expression in the hupB1 : :Tn5 background is repressed to levels similar to that of wild-type (Fig. 4). Although pLS31 is a low copy plasmid, the di¡erences seen in plasmid and chromosomal hilA : :lacZ expression may be due to di¡erences in copy number. Alternatively, the 500 bp of native DNA upstream of hilA present on pLS31 may not be su¤cient to mimic the chromosomal context of this gene thus resulting in the observed di¡erences between chromosomal and episomal hilA: :lacZ expression. It is clear from these data, however, that the class 3 mutations exert a trans e¡ect on hilA : :lacZ expression. 3.7. The class 3 mutants are hyperinvasive for HEp-2 cell monolayers To determine the e¡ect of the class 3 mutations on the invasion phenotype, standard invasion assays of HEp-2 monolayers were conducted with strains grown under high oxygen (non-inducing) or low oxygen (inducing) conditions. As shown in Fig. 6, members of this class show moderate to dramatic increases in invasion of HEp-2 cells. Most notable is the increase in invasion of TF53 (ams1: :Tn5) which is 33-fold more invasive under high oxygen and 4.3-fold more invasive under low oxygen conditions compared to wild-type. This strain appears to be unable to repress invasion under high oxygen conditions. TF57 (hilE2 : :Tn5) and TF58 (hilE1 : :Tn5) show a 2^3fold increase in invasion under low oxygen while TF56 (hilE3 : :Tn5) does not show increased invasion compared to wild-type. This phenotype coincides with our hilA : :lacZ fusion data which show that hilA: :lacZ expression in the hilE3 : :Tn5 mutant background is not a¡ected to the degree of the two other mutations in the same ORF (Fig. 4). The mutations in TF54 (hupB1: :Tn5) and TF55 (pag1 : :Tn5) each cause moderate (1.5-fold) increases in invasion. 31 Fig. 6. E¡ect of the class 3 mutations on invasion of HEp-2 cell monolayers. Strains were grown under high oxygen (white bars) or low oxygen (dark bars) and invasion assays were conducted as described in Section 2. EE251 is the wild-type parent strain. Data presented represent at least three independent experiments with duplicate wells. Error bars depict the standard deviation from the mean. 3.8. The class 3 mutants speci¢cally upregulate hilA To test the hypothesis that the e¡ect of the class 3 mutations on hilA: :lacZ transcription is speci¢c to hilA : : lacZ, we transduced each of these mutations into TF59, an EE251 derivative with a pepT: :lacZ transcriptional fusion, and measured L-galactosidase production under high and low oxygen conditions. The pepT gene has been characterized in Salmonella and encodes an endopeptidase which is regulated at the transcriptional level by oxygen [31] and is not involved in the invasion phenotype of Salmonella [25]. Although we observed that the ams1: :Tn5 mutation slightly reduced pepT: :lacZ expression in TF60, there is no signi¢cant e¡ect of the class 3 mutations on pepT: :lacZ expression (Table 2) and certainly no increase in pepT: :lacZ expression as observed hilA: :lacZ reporter fusions (Fig. 4). This control experiment suggests that the upregulation of hilA by the class 3 mutations is speci¢c to hilA. 4. Discussion As the regulation of Salmonella invasion gene expression is studied more intensively, it is apparent that a complex regulatory network involving a multitude of regulators is present to ensure that invasion occurs under Table 2 E¡ect of class 3 mutations on expression of pepT: :lacZ Strain TF59 (wild-type) TF60 TF61 TF62 TF63 TF64 TF65 L-Galactosidase activity Induction ratio High oxygen Low oxygen 388.3 þ 29.5 100.1 þ 70.4 426.4 þ 21.4 378.6 þ 28.9 282.4 þ 16.3 366.1 þ 33.4 268.4 þ 9.87 959.8 þ 87.4 641.9 þ 22.3 1057.5 þ 124 979.2 þ 79.4 790.8 þ 20.5 895.6 þ 140.1 740.5 þ 24 FEMSIM 1177 5-4-00 2.47 6.41 2.48 2.59 2.80 2.45 2.76 32 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 optimal conditions, those which mimic the intestinal lumen of the host. At the center of invasion gene regulation in S. typhimurium is hilA which encodes a transcriptional activator homologous to the ompR/toxR family of transcriptional activators [19]. In an e¡ort to identify factors involved in the complex regulation of hilA and Salmonella invasion, we conducted a search for negative regulators of hilA. We have isolated and characterized 18 Tn5 insertions in the S. typhimurium chromosome that result in increased expression of hilA on MacConkey agar. Twelve of these insertions, which we have divided into classes 1 and 2, de¢ne two loci within SPI-1 ; one which has a minor e¡ect on hilA (class 1) and one which has a major e¡ect on hilA (class 2). A third class of mutants, comprised of six isolates, identi¢es four loci outside of SPI-1 involved in hilA regulation: ams, hupB, pag, and hilE. Six of these mutations cluster in the prgH-hilD region of SPI-1 and have been termed class 2 mutations (Fig. 1). Each of these mutations upregulates chromosomal hilA: : lacZ under low oxygen 2^6-fold, with the prgH7: :Tn5 mutation causing the greatest increase. Interestingly, we noticed that two of the class 2 mutants, BJ606 and TF40, increase expression of hilA: :lacZ after growth in high and low oxygen conditions. This observation suggests the presence of two regulatory regions near hilD, one involved in low oxygen expression and one involved in both high and low oxygen expression of hilA. Most of the class 2 insertions interrupt the invasion gene prgH, which results in polar disruption of the invasion genes prgI, -J, -K and orgA ([32], and J. Klein, manuscript in preparation). In Yersinia, virulence proteins (Yops) are secreted via a type III secretion system similar to that of Salmonella and blockage of secretion of the Yop proteins causes a feedback repression of Yop synthesis [37]. By comparison, it is possible that the trans e¡ect of class 2 mutations on hilA transcription is due to a lack of production of the PrgH, -I, -J, -K, and OrgA proteins which are required for secretion of invasion e¡ector proteins [21,32]. We tested invasion in these strains by complementation of the disrupted genes with plasmid pJK001 and observed a hyperinvasive phenotype. The plasmid pJK001 restored high oxygen invasion in the mutants to levels above wild-type yet did not alter the oxygen regulation of the wild-type strain. These experiments, along with the isolation of a mutation which does not disrupt prgH (prgH8: :Tn5), indicate that disruption of the prgH, -I, -J, -K and orgA genes by the class 2 mutations is not responsible for the upregulation of hilA. The class 2 mutations are in close proximity to hilD, a recently identi¢ed derepressor of hilA that is 300 nucleotides upstream and transcribed in the opposite orientation as prgH (Fig. 1B) [33]. Therefore, a potential e¡ect of these mutations is to increase hilD transcription which in turn would increase hilA transcription. Consistent with this are data showing that class 2 mutations increase hilA transcription in trans (Fig. 2). With the exception of TF43 and TF45, the e¡ect of the class 2 mutants on the invasion phenotype appears to be to reduce the ability of high oxygen conditions to repress invasion (Fig. 3). Although low oxygen invasion was increased in these strains, the mutants displayed 10^100-fold higher invasion compared to wild-type after growth in high oxygen conditions. This `oxygen blind' invasion phenotype suggests that the class 2 mutations are impaired in their ability to sense oxygen. However, hilA expression in these strains, as measured with the lacZ reporter, is oxygen-regulated (Fig. 1). This may be explained as a phenomenon of invasion saturation. That is, e¡ects causing increases above 0.01% invasion are detectable in these assays whereas those causing increases above 1% invasion are not. Such a phenomenon would make strains overexpressing hilA under both high and low oxygen conditions appear oxygen-insensitive in invasion assays. It seems clear from these assays, however, that the class 2 mutations allow increased invasion after high oxygen growth. During the course of this work we noticed that although the class 2 mutants showed upregulation of chromosomal hilA: :lacZ only under low oxygen conditions, the invasion phenotype of these mutants showed increases under high oxygen conditions as well, suggesting high oxygen overexpression of hilA (Figs. 1 and 3). Also, strains containing the plasmid reporter hilA: :lacZ showed increases in hilA: :lacZ expression under both high and low oxygen conditions (Fig. 2), an observation consistent with the invasion phenotype and inconsistent with the chromosomal hilA: :lacZ reporter data. Perhaps a critical di¡erence between these strains is the fact that the chromosomal hilA: :lacZ creates a hilA null mutation. Since the class 2 mutations in the chromosomal hilA: :lacZ reporter strains do not cause overexpression of hilA whereas the mutations in hilA background apparently do, it is possible that overexpression of hilA under high oxygen conditions requires hilA. This suggests autoregulation by HilA. This is in contrast to work done by Bajaj et al. [14] who examined autoregulation by hilA and found that hilA does not regulate its own expression. Therefore, the reason for the di¡erence between high oxygen expression of hilA in hilA3 and hilA backgrounds remains unclear. Two of the class 2 mutants did not appear hyperinvasive in our assays. One of these, TF43, exhibited wild-type levels of invasion under both high and low oxygen conditions (Fig. 3). Although dissimilar to our other class 2 mutant invasion phenotypes, this phenotype is in agreement with data in Fig. 1 which show that this mutation only slightly increases (less than 2-fold) hilA: :lacZ expression. The other strain, TF45, is less invasive than wildtype cells. This is inconsistent with data in Fig. 1 which show the mutation in this strain causes a large increase in hilA expression which should result in a hyperinvasive phenotype [20]. Observations in our laboratory, and others, suggest that strains that express hilA at very high levels have growth defects. This situation selects for muta- FEMSIM 1177 5-4-00 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 tions that downregulate the expression of the invasion genes. Six of the Tn5 insertions map outside of SPI-1 and are designated class 3 mutations. One of these mutations disrupts amino acid 194 of the S. typhimurium ams gene and is the ¢rst reported ams mutation in Salmonella. Although this mutation caused a slight increase in expression of chromosomal hilA: :lacZ (Fig. 4), both plasmid-encoded hilA : :lacZ expression and invasiveness were signi¢cantly increased in high oxygen growth conditions (Figs. 5 and 6). This suggests that the ams gene is involved in the repression of hilA under high oxygen conditions or, as discussed above, a functional hilA gene is required for high oxygen over expression of hilA in an ams null mutant background. In E. coli, ams encodes RNase E, an endoribonuclease with an N-terminal catalytic domain [38] that cleaves A/U-rich RNA species [39]. The mutation that we have identi¢ed disrupts this catalytic domain in the ams gene of S. typhimurium. Interestingly, Schechter et al. [33] have recently determined the transcriptional start site of hilA and found that the message for the HilA protein has a long 5P untranslated region which is A/U-rich. One possible role of RNase E in invasion is to degrade hilA mRNA to modulate translation of the transcript. Another class 3 mutation disrupts the regulatory region of the hupB gene which leads to increased expression of chromosomal and plasmid hilA: :lacZ reporter fusions (Figs. 4 and 5) but does not signi¢cantly increase invasion (Fig. 6). In E. coli, hupB encodes the L subunit of the histone-like protein HU which binds to and condenses DNA [40^42]. The promoter region of hupB has been characterized extensively in E. coli [43,44] and by comparison to these studies, our hupB1: :Tn5 mutation disrupts the crpcAMP binding site within the hupB promoter region. Since a Tn5 insertion in this region disrupts the positive regulation of hupB by crp-cAMP as well as the negative regulation by FIS we currently do not know if the transposon insertion increases or decreases normal expression of hupB. The K subunit of the HU protein is encoded by the hupA gene. Since HU protein can exist as a homoor heterodimer of K and L subunits, alterations in hupA or hupB expression a¡ect the ratio of KK, KL, or LL dimers present in the bacterium. Thus, it is possible that our hupB1 : :Tn5 mutation alters the composition of the HU dimer and thus a¡ects hilA expression. A third locus identi¢ed by our screen for negative regulators of hilA is a small ORF between the phoP/Q-activated genes pagK and pagM [45] and downstream of the macrophage-induced (mig) gene mig-3 [46] which we refer to as pag. This mutation increases hilA: :lacZ expression (Figs. 4 and 5) but does not cause a signi¢cant increase in invasion (Fig. 6). Since phoP/Q is integral in regulating invasion genes [32,47] as well as genes important for interacting with macrophages [36,48], this mutant is of great interest. Future work will be aimed at analyzing this mutation in a phoQ constitutive background, which represses 33 hilA, to determine whether this ORF is involved in phoP/ Q-mediated repression of hilA. Three of the class 3 Tn5 insertions disrupt a previously unidenti¢ed ORF which we named hilE. This putative gene is predicted to encode a protein which lacks homology to any known protein. Two of the insertions, hilE1: :Tn5 and hilE2 : :Tn5, disrupt the 5P end of hilE and presumably abolish the production of a functional protein from this locus. Strains containing these mutations have signi¢cantly increased hilA: :lacZ expression (Figs. 4 and 5) and invasion 2^3-fold higher under low oxygen conditions (Fig. 6). This suggests that this ORF is involved in the modulation of hilA expression and the invasion phenotype. A third mutation in this locus, hilE3: :Tn5, is at the 3P end of the ORF and has a smaller e¡ect on hilA: : lacZ expression and invasion (Figs. 4^6). This suggests that in this mutant a protein is made that retains some ability to repress hilA. These data show that we have identi¢ed a novel ORF involved in hilA regulation and expression of the invasive phenotype. By isolating the class 3 mutations, we have identi¢ed four new genes which appear to be involved in the regulation of hilA. However, these mutations are due to the insertional inactivation of the genes by Tn5 which can exert polar e¡ects on downstream genes. Future work aimed at characterizing these genes will include determining whether the e¡ect of the transposon insertions on hilA transcription is due to a polar e¡ect on downstream genes or to disruption of the gene at the insertion site. In this study we have searched for repressors of hilA by transposon mutagenesis, identi¢ed several loci involved in hilA regulation both on SPI-1 and elsewhere in the chromosome, and reported a preliminary characterization of the identi¢ed loci. Future work to characterize these loci will help to elucidate the complex regulatory scheme that serves to ensure expression of invasion genes only under conditions optimal for invasion. Acknowledgements We thank Lisa Schechter and Cathy Lee for the gift of plasmid pLS31 and helpful discussions in the preparation of the manuscript. T.F.F. is supported by a predoctoral fellowship on NIH Training Grant 5T32 AI 07511. This work was supported by National Institutes of Health Grant AI 38268 and a grant from the Roy J. Carver Charitable Trust (B.D.J.). References [1] Jones, B.D. and Falkow, S. (1996) Typhoid fever: Host immune response and Salmonella virulence determinants. Annu. Rev. Immunol. 14, 533^561. [2] Jones, B.D., Ghori, N. and Falkow, S. (1994) Salmonella typhimuri- FEMSIM 1177 5-4-00 34 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 um initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer's patches. J. Exp. Med. 180, 15^23. Clark, M.A., Jepson, M.A., Simmons, N.L. and Hirst, B.H. (1994) Preferential interaction of Salmonella typhimurium with mouse Peyer's patch M cells. Res. Microbiol. 145, 543^552. Jones, B., Pascopella, L. and Falkow, S. (1995) Entry of microbes into the host: Using M cells to break the mucosal barrier. Curr. Opin. Immunol. 7, 474^478. Finlay, B.B. and Falkow, S. (1997) Common themes in microbial pathogenicity revisited. Microbiol. Mol. Biol. Rev. 61, 136^169. Darwin, K.H. and Miller, V.L. (1999) Molecular basis of the interaction of Salmonella with the intestinal mucosa. Clin. Microbiol. Rev. 12, 405^428. Groisman, E.A. and Ochman, H. (1993) Cognate gene clusters govern invasion of host epithelial cells by Salmonella typhimurium and Shigella £exneri. EMBO J. 12, 3779^3787. Lee, C.A. (1997) Type III secretion systems: machines to deliver bacterial proteins into eukaryotic cells? Trends Microbiol. 5, 148^ 156. Wood, M.W., Rosqvist, R., Mullan, P.B., Edwards, M.H. and Galyov, E.E. (1996) SopE, a secreted protein of Salmonella dublin, is translocated into the target eukaryotic cell via a sip-dependent mechanism and promotes bacterial entry. Mol. Microbiol. 22, 327^ 338. Kaniga, K., Uralil, J., Bliska, J.B. and Galan, J.E. (1996) A secreted protein tyrosine phosphatase with modular e¡ector domains in the bacterial pathogen Salmonella typhimurium. Mol. Microbiol. 21, 633^ 641. Fu, Y. and Galan, J.E. (1998) The Salmonella typhimurium tyrosine phosphatase SptP is translocated into host cells and disrupts the actin cytoskeleton. Mol. Microbiol. 27, 359^368. Hardt, W.D., Chen, L.M., Schuebel, K.E., Bustelo, X.R. and Galan, J.E. (1998) S. typhimurium encodes an activator of Rho GTPases that induces membrane ru¥ing and nuclear responses in host cells. Cell 93, 815^826. Hong, K.H. and Miller, V.L. (1998) Identi¢cation of a novel Salmonella invasion locus homologous to Shigella ipgDE. J. Bacteriol. 180, 1793^1802. Bajaj, V., Lucas, R.L., Hwang, C. and Lee, C.A. (1996) Co-ordinate regulation of Salmonella typhimurium invasion genes by environmental and regulatory factors is mediated by control of hilA expression. Mol. Microbiol. 22, 703^714. Ernst, R.K., Dombroski, D.M. and Merrick, J.M. (1990) Anaerobiosis, type 1 ¢mbriae, and growth phase are factors that a¡ect invasion of HEp-2 cells by Salmonella typhimurium. Infect. Immun. 58, 2014^ 2016. Galän, J.E. and Curtiss III, R. (1990) Expression of Salmonella typhimurium genes required for invasion is regulated by changes in DNA supercoiling. Infect. Immun. 58, 1879^1885. Lee, C.A. and Falkow, S. (1990) The ability of Salmonella to enter mammalian cells is a¡ected by bacterial growth state. Proc. Natl. Acad. Sci. USA 87, 4304^4308. Schiemann, D.A. and Shope, S.R. (1991) Anaerobic growth of Salmonella typhimurium results in increased uptake by Henle 407 epithelial and mouse peritoneal cells in vitro and repression of a major outer membrane protein. Infect. Immun. 59, 437^440. Bajaj, V., Hwang, C. and Lee, C.A. (1995) hilA is a novel ompR/toxR family member that activates the expression of Salmonella typhimurium invasion genes. Mol. Microbiol. 18, 715^727. Lee, C.A., Jones, B.D. and Falkow, S. (1992) Identi¢cation of a Salmonella typhimurium invasion locus by selection for hyperinvasive mutants. Proc. Natl. Acad. Sci. USA 89, 1847^1851. Penheiter, K.L., Mathur, N., Giles, D., Fahlen, T. and Jones, B.D. (1997) Non-invasive Salmonella typhimurium mutants are avirulent because of an inability to enter and destroy M cells of ileal Peyer's patches. Mol. Microbiol. 24, 697^709. [22] Bajaj, V., Hwang, C. and Lee, C.A. (1995) hilA is a novel ompR/toxR family member that activates the expression of Salmonella typhimurium invasion genes. Mol. Microbiol. 18, 715^727. [23] Eichelberg, K., Hardt, W.D. and Galan, J.E. (1999) Characterization of SprA, an AraC-like transcriptional regulator encoded within the Salmonella typhimurium pathogenicity island 1. Mol. Microbiol. 33, 139^152. [24] Jones, B.D., Lee, C.A. and Falkow, S. (1992) Invasion by Salmonella typhimurium is a¡ected by the direction of £agellar rotation. Infect. Immun. 60, 2475^2480. [25] Jones, B.D. and Falkow, S. (1994) Identi¢cation and characterization of a Salmonella typhimurium oxygen-regulated gene required for bacterial internalization. Infect. Immun. 62, 3745^3752. [26] Ling, M.M. and Robinson, B.H. (1997) Approaches to DNA mutagenesis: an overview. Anal. Biochem. 254, 157^178. [27] Marmur, J. (1961) A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3, 208^218. [28] Davis, R.W., Botstein, D. and Roth, J.R. (1980) Advanced Bacterial Genetics : A Manual for Genetic Engineering. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [29] Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [30] Moore, A.E., Sabachewsky, L. and Toolan, H.W. (1955) Culture characteristics of four permanent lines of human cancer cells. Cancer Res. 15, 598. [31] Strauch, K.L., Lenk, J.B., Gamble, B.L. and Miller, C.G. (1985) Oxygen regulation in Salmonella typhimurium. J. Bacteriol. 161, 673^680. [32] Pegues, D.A., Hantman, M.J., Behlau, I. and Miller, S.I. (1995) PhoP/PhoQ transcriptional repression of Salmonella typhimurium invasion genes: evidence for a role in protein secretion. Mol. Microbiol. 17, 169^181. [33] Schechter, L.M., Damrauer, S.M. and Lee, C.A. (1999) Two AraC/ XylS family members can independently counteract the e¡ect of repressing sequences upstream of the hilA promoter. Mol. Microbiol. 32, 629^642. [34] McDowall, K.J. and Cohen, S.N. (1996) The N-terminal domain of the rne gene product has RNase E activity and is non-overlapping with the arginine-rich RNA-binding site. J. Mol. Biol. 255, 349^355. [35] Marsh, M. and Hillyard, D.R. (1988) Nucleotide sequence of the HU-1 gene of Salmonella typhimurium. Nucleic Acids Res. 16, 7196. [36] Belden, W.J. and Miller, S.I. (1994) Further characterization of the PhoP regulon: identi¢cation of new PhoP-activated virulence loci. Infect. Immun. 62, 5095^6101. [37] Allaoui, A., Schulte, R. and Cornelis, G.R. (1995) Mutational analysis of the Yersinia enterocolitica virC operon: characterization of yscE, F, G, I, J, K required for Yop secretion and yscH encoding YopR. Mol. Microbiol. 18, 343^355. [38] Taraseviciene, L., Bjork, G.R. and Uhlin, B.E. (1995) Evidence for an RNA binding region in the Escherichia coli processing endoribonuclease RNase E. J. Biol. Chem. 270, 26391^26398. [39] McDowall, K.J., Lin, C.S. and Cohen, S.N. (1994) A+U content rather than a particular nucleotide order determines the speci¢city of RNase E cleavage. J. Biol. Chem. 269, 10790^10796. [40] Rouviere, Y.J. and Gros, F. (1975) Characterization of a novel, lowmolecular-weight DNA-binding protein from Escherichia coli. Proc. Natl. Acad. Sci. USA 72, 3428^3432. [41] Rouviere, Y.J., Yaniv, M. and Germond, J.E. (1979) E. coli DNA binding protein HU forms nucleosomelike structure with circular double-stranded DNA. Cell 17, 265^274. [42] Kano, Y., Wada, M., Nagase, T. and Imamoto, F. (1986) Genetic characterization of the gene hupB encoding the HU-1 protein of Escherichia coli. Gene 45, 37^44. [43] Kohno, K., Wada, M., Kano, Y. and Imamoto, F. (1990) Promoters and autogenous control of the Escherichia coli hupA and hupB genes. J. Mol. Biol. 213, 27^36. [44] Claret, L. and Rouviere, Y.J. (1996) Regulation of HU alpha and FEMSIM 1177 5-4-00 T.F. Fahlen et al. / FEMS Immunology and Medical Microbiology 28 (2000) 25^35 HU beta by CRP and FIS in Escherichia coli. J. Mol. Biol. 263, 126^ 139. [45] Gunn, J.S., Belden, W.J. and Miller, S.I. (1998) Identi¢cation of PhoP-PhoQ activated genes within a duplicated region of the Salmonella typhimurium chromosome. Microb. Pathog. 25, 77^90. [46] Valdivia, R.H. and Falkow, S. (1997) Fluorescence-based isolation of bacterial genes expressed within host cells. Science 277, 2007^2011. [47] Behlau, I. and Miller, S.I. (1993) A PhoP-repressed gene promotes Salmonella typhimurium invasion of epithelial cells. J. Bacteriol. 175, 4475^4484. 35 [48] Miller, S.I., Kukral, A.M. and Mekalanos, J.J. (1989) A two-component regulatory system (phoP/phoQ) controls Salmonella typhimurium virulence. Proc. Natl. Acad. Sci. USA 86, 5054^5058. [49] Wray, C. and Sojka, W.J. (1978) Experimental Salmonella typhimurium in calves. Res. Vet. Sci. 25, 139^143. [50] Stibitz, S., Black, W. and Falkow, S. (1986) The construction of a cloning vector designed for gene replacement in Bordetella pertussis. Gene 50, 133^140. FEMSIM 1177 5-4-00