INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Peptidoglycan or murein hydrolases are a family of enzymes that catalyze the cleavage of specific structural components of the bacterial cell wall. These enzymes are involved in several important physiological processes, including peptidoglycan turnover and recycling, cell wall expansion during bacterial growth, septation, and daughter cell separation (13, 32). Due to the potential of these enzymes to compromise cell wall integrity, leading to cell lysis (autolysis), murein hydrolase activity must be carefully regulated. This regulation includes transcriptional control, blocked access to the specific substrate, inhibition by choline or teichoic acid, and substrate modification (13, 32).

The Staphylococcus aureus lytS and lytR genes, whose products are members of the two-component regulator family of proteins, are involved in the control of murein hydrolase activity (3). A lytS mutant phenotypically displays altered murein hydrolase activity and an increased rate of penicillin- and Triton X-100-induced lysis (3, 11). Immediately downstream of lytS and lytR are two genes, lrgA and lrgB, whose transcription is dependent upon lytS and lytR. LrgA and LrgB show no sequence similarity to known murein hydrolases. Instead, LrgA has structural characteristics in common with the bacteriophage-encoded holin proteins involved in murein hydrolase export (4, 11). Recent data generated in our laboratory indicate that the LrgA and LrgB gene products inhibit extracellular murein hydrolase activity and increase tolerance to penicillin (11). Based on these data, it was proposed that the function of LrgA, and possibly LrgB, is analogous to that of an antiholin, i.e., blocking murein hydrolase access to the substrate peptidoglycan (1).

In another study, we have shown that the staphylococcal virulence factor regulators Agr and Sar affect the rate of penicillin-induced lysis and killing (7). Mutations within the agr and sar genes result in decreased and increased rates of penicillin-induced lysis, respectively, possibly as a result of the different effects that these mutations have on murein hydrolase activity. In contrast, both mutant strains exhibited increased sensitivity to the bactericidal effects of penicillin, including the agr mutant, which exhibited reduced penicillin-induced lysis. These data demonstrate that, in addition to controlling virulence factor expression, Agr and Sar affect murein hydrolase activity and penicillin tolerance.

In the present study, we have extended our analysis of lrgAB regulation. We show that Agr and Sar, in addition to the LytSR regulatory system, positively regulate lrgAB expression. We also identified and characterized lrgAB cis-acting elements, including a region of intrinsic DNA bending (curvature), which contribute to maximal lrgAB expression. These data suggest that the effects of Agr and Sar on penicillin tolerance involve the regulation of lrgAB expression.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and growth conditions. S. aureus strains (Table 1) were routinely cultivated in tryptic soy broth (Difco Laboratories, Detroit, Mich.) or NZY broth (3% casein enzymatic hydrolysate [Sigma Chemical Co., St. Louis, Mo.], 1% yeast extract [Fisher Scientific, Fair Lawn, N.J.] adjusted to pH 7.5). Escherichia coli DH5 (14) was grown in Luria-Bertani medium (Fisher Scientific). Liquid cultures were grown with shaking at 250 rpm and 37°C. The antibiotics used were purchased from either Sigma Chemical Co. or Fisher Scientific and used at the following concentrations: ampicillin, 50 µg/ml; chloramphenicol, 5 µg/ml; tetracycline, 3 µg/ml; kanamycin, 50 µg/ml; spectinomycin, 50 µg/ml.

DNA manipulations. S. aureus genomic DNA was isolated using the method of Dyer and Iandolo (6). Plasmid DNA purification was performed using the Wizard Plus kits from Promega, Inc. (Madison, Wis.). The restriction enzymes and T4 DNA ligase used in this study were purchased from GIBCO-BRL (Gaithersburg, Md.). Preparation and transformation of E. coli DH5 were accomplished as described by Inoue et al. (14). Electroporation of DNA into S. aureus was carried out using the procedure of Kraemer and Iandolo (17). S. aureus 11-mediated transduction was performed using the method of Shafer and Iandolo (30).

Northern blot analysis. S. aureus RNA was isolated using the procedure described by Hart et al. (12). A lrgA-specific probe was PCR amplified using the primers 5'-CCAGCACACTTTTTTCACC-3' and 5'-GGTGCTTGGCTAATGACACC-3', producing a 260-bp fragment. The PCR products were gel purified and radiolabeled with [-³²P]ATP by the random priming method as described previously (28). Northern blot analysis was performed as described by Sambrook et al. (28).

lrgAB promoter fusion construction. To construct a single-copy reporter gene vector, the xylE reporter gene encoding the catechol 2,3-dioxygenase enzyme was PCR amplified from the plasmid pCR100. This was achieved using the M13 reverse primer as the 3' primer and a 5' primer designated EX (5'-CCTGAAT TCATGACTCGAGAAGAGGTGACGTCATGAACAAAGGTGTAATGCG ACC-3') containing EcoRI and XhoI restriction sites (underlined). The amplified product was digested with EcoRI and BamHI and directionally ligated into the integration vector pCL84 (18). The resulting plasmid, designated pDF16, contained a promoterless xylE gene downstream of two EcoRI and XhoI restriction sites available for cloning. To examine specific sequences involved in lrgAB expression, PCR amplification was employed to generate consecutively smaller lrgAB promoter fragments. Four different 5' oligomers that anneal to selected regions upstream of the lrgA gene (5'-CCTGTTGAAATTGAATTCAAAATTCACATGTTAAAGC-3', 5'-AATGGTGTCAGAATTCAAGTTGGACGTCCA-3', 5'-TACTTTAACAGGAATTCTTTTTTTTATGC-3', and 5'-TTGTATTCGAATTCAAATCACGCAAATCG-3') were used in separate reactions with the same 3' oligomer (5'-CCGGAAGCTTGTGCTGGTTTTGATGCGTC-3'). All four 5' oligomers generate an EcoRI restriction site (underlined), while the 3' oligomer produces a HindIII site beginning at position +60 with respect to the transcription start site. The PCRs produced products designated AB396, 147, 85 and 42, respectively, where the names refer to the 5' base present in each fragment with respect to the transcription start site of the lrgA gene (see Fig. 1). These fragments were digested with EcoRI and HindIII and directionally cloned into the polylinker of pBluescriptSK(+) (Stratagene), transformed into DH5, and sequenced using the method developed by Sanger et al. (29). Each fragment was liberated from pBluescriptSK(+) by sequential restriction digestion using EcoRI and XhoI and ligated in front of the xylE gene in pDF16, producing plasmids pDF17, pDF36, pDF35, and pDF34, respectively (Table 1). These plasmids (along with pDF16) were electroporated into S. aureus CYL316, where they specifically integrated into the geh locus encoding lipase (18). The integrated loci were transferred into RN6390 by 11-mediated transduction, producing strains designated KB700, KB707, KB706, and KB705 that were used for reporter gene assays (Table 1). Proper integration of the promoter fusion constructs was confirmed by Southern blot analysis and/or PCR analysis using the 5' primer employed for the production of the promoter fragment along with a second oligomer (5'-GTTCTGCACCTTTACGTTG-3') that is complementary to a DNA sequence downstream of the attB integration site within the geh locus (19).

Reporter gene assays. Assays of catechol 2,3-dioxygenase activity were performed by growing S. aureus strains containing the reporter gene in 20 ml of NZY medium for 12 h (stationary phase). A 10-ml sample was removed, and the cells were pelleted by centrifugation at 4,000 × g. The cell pellet was washed with 10 ml of TES buffer (50 mM Tris-HCl [pH 7.5], 10 mM EDTA, 30 mM NaCl) and pelleted a second time. The cell pellet was resuspended in 2 ml of lysis buffer (100 mM potassium acetate, 10% acetone, 20 mM EDTA), and cells were lysed by incubation at 37°C in the presence of lysostaphin (AMBI Inc., Tarrytown, N.Y.) at 50 µg/ml. The cell lysate was homogenized by repeated passage through an 18-gauge needle, and then the cellular debris was removed by centrifugation at 27,000 × g. The clarified lysate was assayed for catechol 2,3-dioxygenase activity as described by Zukowski et al. (37). Specific activity units are defined as milliunits of the product generated divided by the total cellular protein used in each reaction mixture (1 mU corresponds to the formation of 1 nmol of 2-hydroxymuconic semialdehyde per min at 37°C). Total cellular protein concentration was determined using a Bradford protein assay kit (Bio-Rad, Richmond, Calif.) with bovine serum albumin as the standard.

Two-dimensional polyacrylamide gel electrophoresis. A two-dimensional electrophoresis assay to assess the presence of intrinsic DNA bending was performed as described by Rohde et al. (26). Briefly, the far left lane(s) of a 5% nondenaturing polyacrylamide gel was loaded with 1 µg of a 1-kb DNA ladder (GIBCO-BRL) along with a mixture of EcoRI- and HindIII-digested plasmid clones that liberated different fragments originating from the lrgAB promoter region. The fragments were separated in the first dimension at 65°C for 30 min at 100 V. The gel was then rotated 90°, and the fragments were separated in the second dimension by electrophoresis at 4°C for 5 h at 80 V. The DNA fragments were stained with ethidium bromide and then visualized with UV light.

Site-directed mutagenesis. A PCR-based strategy described by Chen and Przybyla (5) was used to generate site-directed changes in the lrgAB promoter region. The 3' primers 5'-TAGGTAAAATGCATAAGCTAAAAAAAAGGATTACTGTTAAAG-3', 5'-TAGGTAAAATGCAGCTCTAGACATAAAAAAAAGGATTACTGTTAAAG-3', and 5'-GGTAAATGCATAAAAAAAATCGACGCTTTAAAATC-3' were designed to generate lrgAB promoter fragments containing 5- and 10-bp insertions and the inverted repeat deletion, respectively (the 5- and 10-bp insertions and the point of deletion in the primers are underlined). These primers were used in PCRs along with a common 5' primer, 5'-CCTGTTGAAATTGAATTCAAAATTCACATGTTAAAGC-3'. The PCR fragments generated were used in a second round of amplification along with the downstream 3' primer 5'-CCGGAAGCTTGTGCTGGTTTTGATGCGTC-3', which amplifies the entire lrgAB promoter region.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Regulatory elements involved in lrgAB expression. In previous studies, it was shown that a lytS mutation resulted in altered murein hydrolase activity, increased Triton X-100- and penicillin-induced lysis, and undetectable lrgAB transcription (3, 4, 11). Given that the lrgAB operon is located immediately downstream of the lytSR locus (Fig. 1), the possibility that the lytS mutation has a polar effect on lrgAB could not be ruled out. Thus, to examine the effects of LytSR on lrgAB expression in greater detail, an lrgAB-xylE promoter fusion construct was introduced into the lytS mutant KB300 (3) and the parental strain 8325-4, which were assayed for catechol 2,3-dioxygenase activity (Fig. 2A). The cloned lrgAB promoter region produced 3.27 ± 0.60 U of catechol 2,3-dioxygenase specific activity in 8325-4. In contrast, KB300 produced undetectable levels of catechol 2,3-dioxygenase activity. These data demonstrate that the gene products produced by the lytSR operon activate lrgAB transcription in trans, either directly or indirectly.

View larger version (35K): [in this window] [in a new window]   FIG. 1.   Nucleotide sequence of the S. aureus lrgAB promoter region. The potential ribosome binding sites and the 10 and 35 promoter elements are underlined and in bold. The asterisks mark the two transcription start sites identified previously (4). The translational stop site for the lytR gene and the translational start site for the lrgA structural gene are in bold. The four boxed sequences highlight AT tracts, and inverted repeat sequences are indicated by arrows.

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Catechol 2,3-dioxygenase assays. (A) Reporter gene assays of 8325-4 (wild-type [wt]) and KB300 (lytSR mutant) strains containing the plasmid reporter gene pDF3 (Table 1). (B) Enzyme activity detected in the virulence regulator mutants using the integrated reporter gene system with the full-length promoter (AB396). RN6390, wild type; RN6390agr-1, agr mutant; ALC488, sar mutant. Control strain RN6390, containing promoterless reporter gene plasmid pDF16, did not produce detectable enzyme activity. The specific activities shown are averages of three independent experiments reported in milliunits (1 mU equals 1 nmol of 2-hydroxymuconic semialdehyde min1 mg of total protein1).

View larger version (26K): [in this window] [in a new window]   FIG. 3.   Northern blot analysis. Total RNA was isolated from S. aureus strains RN6390 (wild type [wt]), RN6911 (agr), ALC136 (sar), and ALC135 (agr sar), and 10-µg samples were separated in a 1% formaldehyde gel. Hybridization was performed using an lrgA-specific probe.

Identification of the lrgAB promoter cis-acting elements. Inspection of the lrgAB promoter region (Fig. 1) revealed the presence of two inverted repeat sequences, centered at nucleotides (nt) 192 and 375, that are potential binding sites for regulatory proteins. Furthermore, four homopolymeric AT tracts (Fig. 1) are located immediately upstream of the 35 and 10 hexamers. Koo et al. (15) determined that homopolymeric AT tracts at least 4 bp long positioned in phase with the helical screw contribute to overall DNA bending. As shown in Fig. 1, AT tracts I and II are in phase with the helical turn and are thus a potential site containing intrinsic DNA curvature. Tracts III and IV are also nearly in phase and may also contain intrinsic curvature. In support of this, the Bend-It software created by Gabrielian et al. (8) to predict DNA curvature predicts that both of these sites contain intrinsic DNA curvature (Fig. 4). To examine the potential role of all these sequences in lrgAB regulation, three consecutively smaller deletions (designated 147, 85, and 42; Fig. 5) of the lrgAB promoter region were made, cloned in front of the xylE gene in pDF16, and inserted into the RN6390 chromosome. The 147 construct, which excluded the inverted repeat sequence centered at nt 192, produced 32.80 ± 6.24 U of enzyme specific activity. This level of activity is similar to that produced by the fusion strain containing AB396 (36.12 ± 12.16 U of specific activity), suggesting that the sequences upstream of nt 324 (including the inverted repeat centered at nt 192) do not affect lrgAB expression under these conditions. However, the 85 construct, which excludes the inverted repeat sequence centered on nt 375, produced 15-fold less enzyme activity (2.3 ± 1.8 U of specific activity) compared to AB396, indicating that the sequence between nt 324 and 386 (including the inverted repeat sequence) is required for optimal lrgAB expression. The 42 construct, which eliminates the potential intrinsic bend sites, produced undetectable levels of reporter gene activity. Since the removal of the sequences spanning the inverted repeat sequence centered on nt 375 significantly decreased reporter gene activity, a full-length promoter fusion construct in which this inverted repeat sequence was specifically deleted was generated and designated AB396IR. The AB396IR fragment was ligated into pDF16 and integrated into the RN6390 chromosome, producing strain KB701. Reporter gene activity was undetectable in this strain. These experiments show that cis-acting regulatory elements controlling lrgAB expression, including the inverted repeat centered at position 375, are contained within approximately 134 nt upstream of the transcription start site.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   DNA curvature prediction. Predicted bending or curvature of the lrgAB promoter region as determined by the Bend-It program constructed by Gabrielian et al. (8) (http://www2.icgeb.trieste.it/~dna/cgi-bin/BendIt2) using a 20-bp window.

View larger version (10K): [in this window] [in a new window]   FIG. 5.   Deletion analysis. Promoter fusion analysis of four consecutively smaller fragments of the lrgAB promoter and the internal deletion fragment of the inverted repeat centered on nt 96 analyzed using the integrated reporter gene construct. The specific activity values shown are averages of at least three independent experiments and are in milliunits (1 mU equals 1 nmol of 2-hydroxymuconic semialdehyde min1 mg of total protein1).

View larger version (24K): [in this window] [in a new window]   FIG. 6.   Operator titration experiment using strain KB700 containing high-copy-number plasmids pMK4, pDF28, and pDF29. The specific activities shown are in milliunits (1 mU equals 1 nmol of 2-hydroxymuconic semialdehyde min1 mg of total protein1) and are averages of three independent experiments.

Two-dimensional gel electrophoresis. As described above, the 42 construct, which lacks the potential bend sites upstream of the 10 and 35 elements, produced no detectable reporter gene activity, indicating that these sequences affect lrgAB expression. To test if this promoter region contains intrinsic DNA bending, a two-dimensional gel electrophoresis assay was performed (Fig. 7A). First, the AB396, 147, and 85 promoter fragments were mixed with a 1-kb ladder and separated in a 5% polyacrylamide gel at 65°C. In this dimension, the rate of DNA migration was a function of size alone. Next, the gel was rotated 90° and the DNA fragments were separated at 4°C. During electrophoresis in this dimension, the migration rate of the DNA fragments is a function of size and secondary structure (intrinsic bending). As shown in Fig. 7A, significant retardation of the migration rate in the second dimension of the AB396 and 147 fragments and, to a lesser extent, the 85 fragment was observed. These data indicate that the lrgAB promoter region does contain intrinsic DNA bending, as predicted by DNA sequencing and computer analysis. Furthermore, the increased rate of migration of the 85 fragment compared to the AB396 and 147 fragments suggests that the bend is centered near the 5' end of 85, which encompasses the AT tract sites.

View larger version (35K): [in this window] [in a new window]   FIG. 7.   Two-dimensional gel electrophoresis assays. (A) lrgAB promoter region fragments AB396, 147, and 85. (B) The 147 fragment and 147 fragments containing the 5- and 10-bp insertions between AT tracts I and II. Fragments are labeled 147, 147+5, and 147+10, respectively.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In a previous study, the global virulence regulator loci agr and sar were shown to affect murein hydrolase activity and tolerance to penicillin-induced lysis (7). Both the agr and sar mutant strains were found to be more sensitive to penicillin-induced lysis and killing compared to the parental strain (7). A more recent study demonstrated that the lrgAB gene products function to decrease extracellular murein hydrolase activity and increase tolerance to penicillin (11). Based on these findings, we hypothesized that the virulence regulators Agr and Sar may affect the expression of the lrgAB operon. Indeed, the results of promoter fusion (Fig. 2B) and Northern analyses (Fig. 3) demonstrate that this is the case. The agr mutation decreased the level of lrgAB expression detected approximately two- to sixfold, while the sar mutation reduced lrgAB expression to nearly undetectable levels. These data suggest that SarA controls lrgAB expression through an Agr-independent pathway. Furthermore, results from our laboratory also indicate that the Agr and Sar effects on lrgAB expression are mediated through a LytSR-independent pathway (unpublished data).

Inspection of the lrgAB promoter region revealed several potential cis-acting regulatory sequences, including two inverted repeat sequences and several AT tracts phased with the DNA helical turn positioned upstream of the 35 and 10 promoter elements. To determine if these sequences are involved in lrgAB expression, we constructed a set of nested deletions spanning the lrgAB promoter region and fused each fragment to the xylE reporter gene. These constructs were integrated into the S. aureus chromosome, and reporter gene activity was assayed. This analysis revealed that the inverted repeat sequence centered on nt 375 and the AT tracts present upstream of the 35 hexamer are important for maximal lrgAB expression. To examine the inverted repeat sequence in greater detail, we produced a promoter fragment (AB396IR) that was identical to AB396 except for a specific deletion of this inverted repeat sequence. This fragment failed to produce detectable enzyme activity using the integrated reporter gene system.

To examine the function of the inverted repeat sequence further, we performed a titration experiment by subcloning the AB396 and AB396IR fragments into a high-copy-number plasmid and introducing them into an S. aureus strain containing the integrated AB396 reporter gene fusion (KB700). Multiple copies of the full-length promoter sequence significantly decreased the level of chromosomally encoded reporter gene activity. These data suggest that the inverted repeat is a specific binding site for an activator protein, although plasmid copy number differences could also impact these results. However, this construct did not completely abolish detectable reporter gene activity. One explanation for this observation is that the promoter topology within the plasmid may not accurately reflect the DNA structure of the promoter region as it exists within the chromosome. Thus, the promoter sequence located within the chromosome may display higher affinity and outcompete the plasmid-encoded promoter sequence for trans-acting factors required for lrgAB expression. The 10-fold increase in promoter activity obtained using the integrated reporter gene construct (Fig. 2) supports this idea.

The results of two-dimensional gel electrophoresis analysis show that the lrgAB promoter also contains a region of intrinsic DNA curvature. Using the deletion fragments generated for the promoter fusion analysis, intrinsic DNA curvature was localized between the inverted repeat sequence centered on nt 375 and the 35 hexamer. This is based on the observation that the mobility of the 85 fragment was not affected by low temperature as much as the other promoter fragments were. It has been shown that when a bend occurs at the center of a DNA molecule it will affect the shape of that molecule to a greater degree than if the bend occurs near an end (9). The results of the two-dimensional gel assay (Fig. 7A) indicate that the intrinsic bend site is located near the end of the 85 fragment, near AT tracts I and II. To confirm this, we created 5- and 10-bp insertions between AT tracts I and II in the 147 fragment. The 5-bp insertion (which inserts half of a helical turn) would be predicted to disrupt the overall curvature of this region, producing a zigzag structure and increasing the migration rate in the second dimension compared to the wild-type fragment. As shown in Fig. 7B, this is precisely what was observed. On the other hand, the 10-bp insertion (which inserts a full helical turn) was predicted to create a deeper curve, reducing the migration rate compared to the wild-type fragment. Again, the results shown in Fig. 7B show this to be the case. When these 5- and 10-bp insertions were introduced into full-length promoter fragment AB396 and analyzed using the chromosomal reporter gene construct, no detectable enzyme activity was produced (data not shown). These data argue against an enhancer loop model of lrgAB activation in which the placement of the operator and the promoter on the same face of the helix is more important than the spacing between these cis-acting elements. On the other hand, the data suggest that the geometry and/or the position of the inverted repeat relative to the promoter is critical for lrgAB expression to occur. Alternatively, the insertions may disrupt primary structure elements necessary for binding of an activator protein.

Intrinsic DNA curvature has been implicated in the activation of transcription by promoting the juxtaposition of DNA sequences near the terminal loop of a superhelical domain (35). Yang et al. (36) suggested that the introduction of a curved DNA sequence brings about changes in the overall shape of a supercoiled polymer. These examples raise the possibility that intrinsic DNA bending influences the supercoiled topology within promoter regions. Furthermore, there are several cases in which intrinsic DNA bending and/or protein-induced bending upstream of promoters affect transcription initiation (24). Thus, it is clear that DNA bending, supercoiling, and transcription factor binding can work either in concert or independently to control gene expression.

In S. aureus, the expression of the eta gene has been shown to be enhanced by DNA relaxation (31). More recently, Morfeldt et al. (20) hypothesized that SarA binding to the agr promoter region affects the localized superhelicity of this promoter. Interestingly, SarA exhibits moderate sequence similarity to Shigella flexneri VirF (21), which has been shown to activate virB transcription by binding upstream in a DNA topology-dependent manner (34). Based on this observation, along with the results of the current study, it is possible that SarA recognizes and binds distinct topological features in the lrgAB promoter region that may lead to the generation of an active transcription complex. Recently, it has also been shown that SarA specifically binds upstream of the collagen adhesion gene (cna), controlling its expression independently of Agr (2, 10). When applying the Bend-It software to examine whether this and the agr P3 promoter also contained DNA curvature, we found that both analyses predicted DNA curvature within these promoter regions (data not shown). Thus, similar topological features may be required for the activation of other SarA-regulated promoters.