Molecular Characterization of a NovelStaphylococcus aureus Serine Protease Operon

The bacterial strains and plasmids used in this study are listed in Table1. S. aureus strains were routinely grown in tryptic soy broth (TSB; Difco Laboratories, Detroit, Mich.) containing erythromycin (5 μg/ml), kanamycin (50 μg/ml), or tetracycline (5 μg/ml) where indicated. Escherichia colistrains were grown in Luria-Bertani medium (Fisher Biotech) with ampicillin selection (50 μg/ml). All bacterial cultures were routinely incubated at 37°C and aerated by shaking at 250 rpm.

E. coli plasmid DNA was extracted using the Wizard Plus DNA Purification System (Promega, Madison, Wis.). Restriction endonucleases and T4 DNA ligase were purchased from GIBCO-BRL (Gaithersburg, Md.) and used as recommended by the manufacturer. Transformations of E. coli strain DH5α and S. aureus RN4220 were performed using the procedures described by Inoue et al. (17) and Kraemer and Iandolo (19), respectively. S. aureus chromosomal DNA was isolated using the method of Dyer and Iandolo (13), and plasmid DNA was isolated using the Wizard Plus DNA Purification System, except that S. aureus cells were converted to protoplasts using lysostaphin (50 μg/ml). Bacteriophage transductions in S. aureus were performed as described by Shafer and Iandolo (32) using φ11 as the transducing phage.

Several proteins with similar physiochemical properties were purified from S. aureus cultures. N-terminal sequence data was used in a BLAST search to screen the S. aureus 8325 (the parent strain of RN6390) sequencing database provided by theStaphylococcus aureus Genome Sequencing Project (University of Oklahoma Health Sciences Center). The N-terminal sequences used originally matched open reading frames found on two separate contigs. One contig contained the splA, splB, and splCgenes, while the other contig contained part of splD and all of splE and splF. Since a search of the S. aureus COL genome database (The Institute for Genome Research) indicated that these six genes make up a single operon, the sequence between the two 8325 contigs was PCR amplified and sequenced. The nucleotide sequence of the entire spl operon from 8325 was entered into the GenBank database (accession no. AF271715 ). The nucleotide sequences of the splB and splC genes were previously identified and included in the GenBank database asorf1 and orf2 (accession numbers U60589 andU63529, respectively).

S. aureusstrain RN6390 was grown with shaking at 37°C in 5 liters of Todd-Hewitt broth (Difco Laboratories) to late stationary phase. The culture was precipitated in 4 volumes of ethanol for 1 week at 4°C. The precipitate was collected by centrifugation ( 13,000 × g for 15 min at 4°C) and air dried. Proteins in the pellet were dissolved in 50 ml of water and clarified by centrifugation at 40,000 × g at 4°C for 30 min. After dialysis against multiple changes of water (4°C), the volume of the dialysate was reduced to less than 100 ml by pervaporation. The dialysate was then subjected to flatbed isoelectric focusing (IEF) in a gradient of ampholytes (pH 3 to 10) in Sephadex (Amersham Pharmacia Biotech, Piscataway, N.J.). The IEF gel was divided into 19 fractions. Proteins were eluted from each fraction and then dialyzed exhaustively at 4°C against multiple changes of water. Proteins in each fraction were resolved by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and visualized by staining with Coomassie blue. The SplA, SplB, and SplF triplet was further purified using a fast protein liquid chromatography (FPLC) sizing column (Superose 12 HR 10/30; Amersham Pharmacia Biotech). The purified Spl proteins were then separated on an SDS-PAGE gel containing 2.0 M urea and transferred to a polyvinylidene difluoride membrane (Millipore Corporation, Bedford, Mass.) according to the method of Matsudaira (20). SplB and SplC (from a separate IEF fraction) were further purified by reverse-phase high-pressure liquid chromatography (HPLC) using a Hewlett-Packard HP1090 (Hewlett-Packard, Waldbronn, Germany) apparatus equipped with a 15-cm C₁₈ column (VYDAC, Hesperia, Calif.). Protein was loaded onto the column in 0.1% trifluoroacetic acid and eluted in a gradient of 0 to 60% acetonitrile in 0.1% trifluoroacetic acid with a flow rate of 1 ml/min. The N-terminal amino acid sequences of the purified Spl proteins were determined by Edman degradation (Laboratory for Bioanalysis and Biotechnology, Pullman, Wash.).

S. aureus cultures were grown in 10 ml of TSB to late stationary phase. Cultures then were centrifuged at 6,000 × g for 5 min (4°C), and the supernatant was filter sterilized through a 0.2-μm-pore-size filter (Pall Corporation, Ann Arbor, Mich.). The exoproteins were concentrated by passage through a Centricon 3 centrifugal filter device (Millipore) at 4°C by following the manufacturer's instructions. Protease activity associated with Spl proteins was detected using zymographic analysis essentially as described by Arvidson et al. (2). A 52.5-μl volume of each sample was mixed with 17.5 μl of 4× sample buffer (without β-mercaptoethanol) and denatured for 15 min at room temperature. The samples were loaded on an SDS–12% polyacrylamide gel containing 0.1% β-casein (Sigma, St. Louis, Mo.), and electrophoresis was performed with a constant voltage of 120 V using buffer precooled to 4°C. Renaturation of proteases was obtained by incubating the gels in 2.5% (vol/vol) Triton X-100 for 30 min with gentle agitation at room temperature. Gels were then placed in development buffer (50 mM Tris, 200 mM NaCl, 5 mM CaCl₂, 0.02% Brij 35) for 10 to 12 h with gentle agitation at 37°C and stained with 0.5% Coomassie blue. Zones of hydrolysis were visualized after destaining in a 40% methanol–10% acetic acid solution.

The S. aureus spl operon was deleted in strain RN6390 by allele replacement as follows. Initially, a 480-bp DNA fragment spanning a region 3′ tosplF was PCR amplified using Taq DNA polymerase (GIBCO-BRL) and the splD3 and splD4 primers (5′-GGTAATAAGCCATCAGGTGAAAGCTCTAGAGG-3′ and 5′-GCGAACGTTGTTCTGCAGTAATAGAC-3′, respectively; see Fig. 2). The DNA product was then digested with XbaI andPstI (underlined sequences) and ligated into theXbaI and PstI sites of pDG647 downstream of the plasmid-encoded erythromycin resistance (Em^r) cassette. This recombinant plasmid was designated pSR1.

Similarly, a 450-bp DNA fragment spanning a region 5′ tosplA was PCR amplified using the primers splA1 and splA2 (5′-GTTTCATTGATGAATTCATATTGGC-3′ and 5′-GTTAAAATAGTTAAAGCAGTGGTACCTTTAACC-3′, respectively; see Fig. 2). This fragment was then digested withEcoRI and KpnI (underlined sequences) and ligated into the EcoRI and KpnI restriction sites of the plasmid pSR1 upstream of the (Em^r) cassette. This recombinant plasmid, designated pSR2, was then digested withPstI and EcoRI, and the resulting fragment containing the Em^r cassette and the two flankingspl sequences was ligated into pCL52.2, generating pSR3. This plasmid was transformed into S. aureus RN4220 and subsequently transferred into RN6390 by bacteriophage-mediated transduction. This strain was grown at the nonpermissive temperature (43°C) on Trypticase soy agar (TSA) plates containing tetracycline to select for cells in which the plasmid had been integrated into the RN6390 chromosome by homologous recombination. To allow a second recombination event to occur, a single colony was inoculated into antibiotic-free TSB and grown at 30°C for 5 days with 1:1,000 dilution into fresh medium each day. After day 5, the bacteria were transferred to TSA containing erythromycin and isolated colonies were screened for the loss of tetracycline resistance. PCR amplification and Southern blot analysis (data not shown) confirmed the replacement of the spl operon. This spl mutant strain was designated KB600.

A specific disruption of splB was achieved by PCR amplifying a 320-bp DNA fragment internal to splB using the primers splB1 and splB2 (5′-GATACTAATATTTTTCCATATACTGG-3′ and 5′-CGCTCACAAGCTTTAGCCCCTGCCGC-3′, respectively). This amplified fragment was ligated into the EcoRV site of pBluescript II KS (Stratagene, La Jolla, Calif.) and subsequently ligated into the EcoRI/KpnI sites of pER924 (Table 1). This recombinant plasmid was then electroporated intoS. aureus strain RN4220, which was spread onto TSA containing erythromycin, and incubated at 30°C overnight. The plasmid was then transferred into RN6390 by phage-mediated transduction. This strain was grown at the nonpermissive temperature (43°C) in the presence of erythromycin to select for cells in which the plasmid had been integrated into the RN6390 chromosome via homologous recombination. PCR amplification and Southern blot analysis confirmed that splB had been disrupted (data not shown). This strain was designated KB601.

For use in complementation experiments, the spl operon (excluding the promoter region) was PCR amplified using primers splA3 (5′-CATTAACTATAAAAATAAGCTTGGAAGGAGG-3′) and splD2 (5′-GTATATTTTGTCAGGATCCGGTGAATGTCTAAG-3′) and then cloned into the BamHI andHindIII sites (underlined sequences) of pBluescript II KS. This 5.0-kb DNA fragment was then liberated by digestion withBamHI and SalI, ligated to the corresponding sites within pRN5548, and transformed into S. aureus RN4220. Plasmid DNA from one positive clone was then transferred to variousS. aureus strains by electroporation. This plasmid construct, designated pSR7, placed the expression of the sploperon under the control of the constitutively expressed β-lactamase promoter.

To examine growth phase-dependent expression, cultures were diluted 1:100 in TSB and grown with aeration at 37°C. At 4, 6, 8, 10, and 12 h after inoculation, 10-ml culture aliquots were mixed with an equal volume of ethanol-acetone solution (1:1, vol/vol) and the cells were stored at −20°C. After sampling was complete, the cell suspensions were centrifuged at 6,000 × g (4°C) for 5 min and the pellets were resuspended in 10 ml of TEN buffer (29). The cells were centrifuged again, and the pellets were resuspended in 1 ml of TEN buffer (containing 2.5 M NaCl). The cells were converted to protoplasts by incubation (37°C, 25 min) in the presence of recombinant lysostaphin (50 μg/ml). RNA was isolated using 4 ml of Trizol (GIBCO-BRL) in accordance with the manufacturer's instructions and resuspended in sterile, dimethylpyrocarbonate-treated water containing 0.5% SDS.

Twenty micrograms of RNA (determined by spectrophotometric analysis) from each S. aureus strain was incubated at 55°C in denaturing buffer (5% 10× morpholinepropanesulfonic acid [MOPS], 5.5% formaldehyde [37%, wt/vol], 50% formamide) for 15 min and resolved by electrophoresis through a 0.8% agarose gel containing formaldehyde (3). The RNA was transferred to a charged nylon membrane (MSI, Westboro, Mass.) using downward capillary action (29). For dot blot analysis, 20- and 40-μg portions of RNA were denatured as described above and applied to a charged nylon membrane using a dot blot manifold. Immobilized RNA was cross-linked twice on each side of the membrane using a UV Stratalinker 1800 (Stratagene). Next, prehybridization was carried out at 65°C for 1 h in 50 ml of hybridization buffer (5× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate]; 0.02% SDS, 0.1%N-laurylsarcosine, 1% blocking stock [10% blocking reagent from Boehringer Mannheim, Indianapolis, Ind.] in maleic acid buffer). Digoxigenin (DIG)-labeled gene-specific DNA probes were generated using a PCR DIG Probe Synthesis Kit (Boehringer Mannheim) with primers specific for each spl gene. Hybridization withspl-specific probes (see above) was carried out at 65°C in hybridization buffer containing heat-denatured DIG-dUTP-labeled probe for 12 to 16 h. Membranes were washed twice (for 15 min each time) in 20× SSC–0.01% SDS and then twice (for 30 min each time) in 0.5× SSC–0.01% SDS. All washes were carried out at room temperature (with the exception of the second wash in 0.5× SSC–0.1% SDS, which was performed in a water bath at 68°C) with gentle agitation. The remainder of the detection procedure followed the protocol in theDIG System User's Guide for Filter Hybridization.

In an initial study to examine if thespl operon contributes to virulence, 29-day-old male Sprague-Dawley rats were given intraperitoneal injections of S. aureus RN6390, KB600, and KB600(pSR7). The dosages used were 10⁷, 10⁸, and 10⁹ CFU in a 500-μl volume. The overall health and viability of the rats were assessed over a 3-day period. After this time, the rats were sacrificed and their organs were examined by visual inspection and bacterial culturing.

Fractionation of proteins in S. aureusRN6390 culture supernatants revealed a comigrating group of three extremely basic and similarly sized proteins (SplA, SplB, and SplF) ranging from 26 to 28 kDa (Fig. 1A). Upon preparative IEF, these proteins migrated to the same position in the pH gradients representing pIs of approximately 9.0. Further purification using FPLC resulted in an apparently pure protein preparation containing only these three Spl proteins (Fig. 1A, lane 4). To identify the proteins, they were separated by SDS-PAGE, transferred to polyvinylidene difluoride membrane, and subjected to N-terminal sequencing. The results of this analysis revealed that the N-terminal sequences of the proteins from the upper, middle, and lower bands were ENNVTKVKDTNIFPYTGVVAF, EKNVKEITDATKEPY, and ENTVKQITNTNVAPYSG, respectively.

Comparison of the N-terminal sequences of the purified proteins to the S. aureus 8325 genome database indicated that these proteins were encoded by genes within a six-gene operon (Fig. 2). The N-terminal sequences of the proteins from the upper, middle, and lower bands (Fig. 1A, lane 4) were identical to the deduced amino acid sequences of the first, second, and sixth gene products, respectively. Interestingly, products of all six genes share 43.9 to 94.6% sequence identity with each other and, as a group, contain significant sequence similarity to known serine proteases (Fig. 3 and4) such as S. aureus V8 protease and the epidermolytic toxins. Notably, each gene product contains the conserved amino acids His-74, Asp-113, and Ser-189 (splA gene product numbering), which make up the classic catalytic triad of trypsin-like serine proteases (Fig. 3). Unlike V8 protease, the Spl proteins appear to be synthesized without propeptides. Because of their sequence similarities to serine proteases, the genes encoding these proteins were designatedsplA, splB, splC, splD, splE, and splF for serine protease like.

Comparison of the SplA, SplB, and SplF N-terminal sequences with the deduced amino acid sequences of the splA, splB, andsplF gene products revealed matches starting 35 or 36 residues downstream of the putative N-terminal methionines, respectively. The product of the splC gene was also detected in IEF gels, purified (Fig. 1A, lane 5), and shown to have an N-terminal sequence (EKNVTQVKDT) that was identical to the deduced amino acid sequence 37 residues downstream of the N-terminal methionine. These data indicate that at least four of the proteins encoded by this operon contain 35-or 36-residue-long signal peptides (Table 2) consistent with secretion in a Sec-dependent manner. The predicted mature protein products ranged in size from 21.9 to 22.4 kDa and, with the exception of the third gene product (SplC), are predicted to be basic proteins having isoelectric points of approximately 9.0 (Table 2). The predicted neutral isoelectric point of the third gene product is consistent with its absence in our initial purified protein preparation of SplA, SplB, and SplF (Fig. 1A, lane 4).

To determine whether the spl operon encodes protease activity, we tested purified preparations of SplB and SplC for the ability to degrade casein. As shown in Fig. 1B, a zymographic analysis using casein as the substrate revealed clear zones of hydrolysis corresponding to the purified proteins, indicating that SplB and SplC exhibit protease activity. Although the other Spl proteins were not tested, it is likely that they also exhibit protease activity based on the high degree of sequence conservation exhibited by all of thespl gene products (including the catalytic triads).

Since the expression of most staphylococcal exoprotein genes is activated during the transition into stationary phase and is dependent on the virulence factor regulator Agr, we examined the expression of the spl operon during various stages of growth. RNA was isolated from S. aureus RN6390 cultures in 2-h increments beginning in exponential phase (4 h) and extending to stationary phase (12 h). RNA was analyzed by dot blot hybridization using an splB-specific DNA probe. As shown in Fig. 5, maximal expression ofsplB-specific transcripts was detected at the 8-h time point, which corresponded to early stationary phase (unpublished results).

The six spl genes span a total of 4,916 nucleotides and are separated by 57- to 157-bp spacer regions. Immediately downstream ofsplF is an inverted repeat sequence that is a potential factor-independent transcription terminator. The absence of inverted repeat sequences downstream of any other spl gene suggests that the six spl genes are cotranscribed, with transcription terminating downstream of splF. As determined by Northern blot analysis (Fig. 6), several transcripts in RN6390 hybridized to splB-, splC-, and splF-specific probes, while only the largest (5.5 kb) of these hybridized to the splA-specific probe. The observation that the splF-specific probe hybridized to all of the transcripts suggests that all of these transcripts have the same 3′ terminus. Assuming that these transcripts terminate just downstream of the splF gene (near the putative transcription terminator), the start of transcription for the 5.5-kb transcript is estimated to be located approximately 400 bp upstream of the splA start codon. Experiments using any of the spl-specific sequences to probe RNA from KB601 (a strain that contained an integrated plasmid in the splB gene) did not detect hybridization to the 5.5-kb RNA. Instead, a 2.3-kb transcript, possibly representing a truncated version resulting from the plasmid integration in splB, was detected only with the splA- or splB-specific probes. Since any promoters downstream of splB would likely be unaffected by the integrated plasmid, these results suggest that the largest transcript spans the entire operon and that the smaller transcripts are nonrandom degradation products.

To examine the roles of Agr and Sar in spl expression, a Northern blot analysis was performed on RNA isolated from RN6911 (agr), ALC135 (agr sar), and ALC136 (sar) when the spl operon was maximally expressed in the parental strain (during the transition from exponential to stationary phase). As shown in Fig. 7, there was no hybridization to RN6911 (agr) and ALC136 (agr sar) RNA using an splB-specific probe (otherspl-specific probes gave similar [unpublished] results), demonstrating regulation by Agr. Hybridization of each probe to ALC135 (sar) RNA occurred at levels equivalent to hybridization to RN6390.

To examine if thespl genes are present in other strains, 11 randomly selectedS. aureus isolates (Table 1) were analyzed by PCR amplification using splA- and splD-specific primers. PCR amplification of DNA from these isolates usingspl operon-specific primers successfully amplified a 2.8-kb DNA fragment (presumably containing splA, splB,splC, and part of splD) in seven (64%) of the isolates tested (data not shown). With the exception of the Brittingham strain, the results of the PCR survey were confirmed by Southern blot analysis. Although the Brittingham strain produced no DNA fragment in PCR amplification using primers specific to the spl operon, Southern blot analysis revealed the presence of sequences that hybridized to splC- and splD/F-specific probes, but not to probes specific for splA and splB(data not shown).

The role of virulence factors during the course of infection is most easily determined by generating mutant strains and then assessing their ability to cause disease in an animal model of infection. Accordingly, the entire spl operon was replaced in RN6390 with an Em^r determinant (see Materials and Methods), resulting in strain KB600. SDS-PAGE analysis of the exoproteins produced by this strain confirmed that it failed to produce the Spl proteins (unpublished data). Furthermore, transformation of KB600 with the spl operon expression plasmid pSR7 resulted in restoration of spl expression. To assess the effects of the spl operon on staphylococcal pathogenesis, three dosages (10⁷, 10⁸, and 10⁹ CFU) of RN6390, KB600, and KB600(pSR7) were intraperitoneally injected into 29-day-old rats. Examination of these animals revealed no differences in the 50% lethal dose or the time required to cause death. Furthermore, no differences in the lesions produced by these strains were observed. Thus, the spl operon plays no obvious role in virulence as determined by this animal infection model.