The Staphylococcus aureus Chaperone PrsA Is a New Auxiliary Factor of Oxacillin Resistance Affecting Penicillin-Binding Protein 2A

Abstract

Expression of the methicillin-resistant S. aureus (MRSA) phenotype results from the expression of the extra penicillin-binding protein 2A (PBP2A), which is encoded by mecA and acquired horizontally on part of the SCCmec cassette. PBP2A can catalyze dd-transpeptidation of peptidoglycan (PG) because of its low affinity for β-lactam antibiotics and can functionally cooperate with the PBP2 transglycosylase in the biosynthesis of PG. Here, we focus upon the role of the membrane-bound PrsA foldase protein as a regulator of β-lactam resistance expression. Deletion of prsA altered oxacillin resistance in three different SCCmec backgrounds and, more importantly, caused a decrease in PBP2A membrane amounts without affecting mecA mRNA levels. The N- and C-terminal domains of PrsA were found to be critical features for PBP2A protein membrane levels and oxacillin resistance. We propose that PrsA has a role in posttranscriptional maturation of PBP2A, possibly in the export and/or folding of newly synthesized PBP2A. This additional level of control in the expression of the mecA-dependent MRSA phenotype constitutes an opportunity to expand the strategies to design anti-infective agents.

INTRODUCTION

Methicillin-resistant Staphylococcus aureus (MRSA) is regularly listed as one of the worldwide causes of hospital and community infections in yearly global surveillance reports (World Health Organization, Antimicrobial Resistance Global Report on Surveillance, 2014). The extreme plasticity of the S. aureus genome has allowed acquisition of new features in terms of antibiotic resistance, but also virulence and even host tropism (1, 2). Despite several antimicrobials available for the treatment of infections, MRSA is repeatedly involved in critical outcomes for infected patients because of increased length of stay and mortality (3). Cell wall-active antibiotics (CWAA), including β-lactams, glycopeptides, and daptomycin, are currently used in clinical settings to treat staphylococcal infections (4); however, resistance to these antibiotics was reported only a few years after their introduction (5, 6). Steady progresses in genome-wide analysis and high-throughput methodologies has permitted identification of cell wall remodeling (7, 8), oxidative stress modulation (9, 10), proteolysis, and chaperoning rerouting (11) as mechanisms leading to resistance against CWAA.

CWAA, especially glycopeptides and β-lactam antibiotics, target extracellular cell wall compounds such as peptidoglycan (PG) precursors or penicillin-binding proteins (PBPs). Any factor affecting the amount and/or activity of such cell wall components may represent a suitable candidate for the purpose of MRSA resensitization. In a previous study, we identified the posttranslocational protein PrsA as a factor required for both glycopeptide and oxacillin resistance in S. aureus (12). Expression of prsA is induced upon addition of CWAA and requires the activity of the cell wall stress sentinel two-component system VraSR. PrsA was first discovered in a bank of secretion defective mutants in Bacillus subtilis (13, 14). Since then, PrsA has been described as a ubiquitous protein, lipid anchored to the outer surface of the membrane, where it assists in the folding of extracellular proteins. In B. subtilis, prsA is essential for growth and secretion of alpha-amylase. In fact, PrsA-depleted cells suffer from severe morphological defects and decreased PG cross-linking, resulting from the misfolding of PBP2a, PBP2b, PBP3, and PBP4 (15). Along the same lines, cell surface properties are affected in a Streptococcus mutans prsA mutant and in Listeria monocytogenes PBPA and PBPB were identified as putative PrsA substrates on the basis of proteomics studies (16, 17).

PrsA proteins are parvulin-like peptidyl-prolyl cis-trans isomerases (PPIase) and typically consist of three domains: (i) the N-terminal domain (Nter) provides anchorage to the membrane via a cysteine residue covalently linked to a di-acyl glycerol (18); (ii) the parvulin PPIase domain catalyzes isomerization of peptide bonds preceding proline residues, which is a limiting factor for protein folding; and (iii) the C-terminal domain (Cter). Protease-coupled PPIase assays confirmed the prolyl isomerase activity of B. subtilis, S. aureus, and L. monocytogenes PrsA proteins (17, 19, 20). B. subtilis and S. aureus PPIase domains share a typical parvulin fold consisting of a four-stranded antiparallel β-sheet surrounded by four α-helices. They exhibit differences in substrate specificity, most probably due to the extended loop between sheet S1 and α-helix 1 in S. aureus PrsA (19).

In an elegant work using the signal peptidase (SPase) inhibitor arylomycin to define the S. aureus secretome, Romesberg and colleagues showed that the prsA gene and the htrA gene (encoding an extracellular protease) are induced and that PrsA is highly secreted following inhibition of SPase (21, 22). In B. subtilis, htrA and the S. aureus vraSR counterpart liaSR are among the genes induced by severe secretion stress (23). This overlap in genes activated upon secretion stress and CWAA likely implies a need for the secretion and folding of cell wall-related proteins synthesized to face cell wall stress caused by CWAA.

The role of PrsA in glycopeptide and oxacillin resistance in S. aureus is recognized, but limited information is currently available regarding PrsA substrates in S. aureus. Here, we show that the requirement of PrsA for β-lactam resistance is not dependent upon the SCCmec type, since the deletion of prsA in three different SCCmec backgrounds provokes a decreased in oxacillin resistance. More importantly, prsA deletion causes a decrease in PBP2A in the membrane. Moreover, by performing PrsA structure-function analysis, we demonstrate the role of the three PrsA domains (the N-terminal, C-terminal, and PPIase domains) in oxacillin resistance and PBP2A protein level modulation.

MATERIALS AND METHODS

Bacterial strains and plasmids.

Strains and plasmids used in this study are listed in Table 1. Escherichia coli strains were grown in Luria-Bertani broth and S. aureus strains were grown in Mueller-Hinton broth (MHB) or tryptic soy broth (TSB). Growth media were supplemented with ampicillin (100 μg/ml), kanamycin (40 μg/ml), tetracycline (3 μg/ml), and chloramphenicol (15 μg/ml), when appropriate. Recombinant lysostaphin was obtained from AMBI Products LLC (Lawrence, NY).

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Strain	Relevant genotype or characteristica	Source or reference
Strains
E. coli DH5α		Restriction-deficient DNA cloning strain	Gibco/BRL
S. aureus
RN4220		8325-4, r− m+, restriction-defective strain which accepts foreign DNA	49
COL (Tets) (cured for plasmid)	AJ717	HA-MRSA, bla− ΔmecR1 mecI− mecR2−	50
MW2		CA-MRSA, bla+ ΔmecR1 mecI− mecR2−	53
Mu3		HA-MRSA, hVISA, bla− mecR1 mecI mecR2	52
CYL316		Derivative of RN4220 containing the integrase for insertion of pLC84 into the lipase gene	51
Newman ΔprsA	AJ3	Newman prsA::kan	12
NewmanΔprsA-Cb	AJ4	Newman prsA::kan geh::pCL84_prsA	12
MW2ΔprsA	AJ724	MW2 prsA::kan, Φ80α transductant of Newman ΔprsA	This study
Mu3ΔprsA	AJ457	Mu3 ΔprsA, markerless deletion using pKOR1	This study
COLΔprsA	AJ728	COL prsA::kan, Φ80α transductant of Newman ΔprsA	This study
COLΔprsA-C	AJ760	COL prsA::kan, geh::pCL84_prsA, Φ80α transductant of Newman ΔprsA-C	This study
COL_prsA_Nterm+Cterm	AJ843	COL prsA::kan geh::pCL84_prsA_Nterm+Cterm	This study
COL_prsA_PPIase	AJ900	COL prsA::kan geh::pCL84_prsA_PPIase	This study
COL_prsA_Nterm+PPIase	AJ1016	COL prsA::kan geh::pCL84_prsA_Nterm+PPIase	This study
COL_prsA_Nter	AJ1020	COL prsA::kan geh::pCL84_prsA_Nterm	This study
COL_prsA_Cterm	AJ1018	COL prsA::kan geh::pCL84_prsA_Cterm	This study
Plasmids
pKOR1		E. coli-S. aureus thermosensitive shuttle vector, Ampr Camr, ATc counterselection	24
pAJ342		pKOR1_prsA_Mu3
pCL84		tetK; S. aureus geh locus integrating plasmid	51

^aCam^r, chloramphenicol resistance; Amp^r, ampicillin resistance.

^bprsA-C, prsA complemented.

Construction of prsA mutant strains.

The prsA::kan allele from AJ3 (12) was backcrossed to COL and MW2 genetic backgrounds by transduction using bacteriophage 80α. Transductants were verified by PCR-based assay and sequence analysis. The temperature-sensitive vector pKOR1 plasmid (24) was used to generate an in-frame markerless deletion of prsA in strain Mu3. Briefly, ∼1,000 bp of prsA upstream and downstream regions were amplified from Mu3 genomic DNA with primers 1, 2, 3, and 4 listed in Table 2. PCR products were digested with SacII (NEB) and ligated by T4 DNA ligase (TaKaRa). The ligation product was used for recombination with pKOR1 and the recombination product was used to transform E. coli DH5α. The resulting pKOR1_prsA_Mu3 (pAJ342) plasmid was used to transform RN4220, selecting for chloramphenicol resistance (Cam^r) at 30°C. From RN4220, the recombinant plasmid was introduced into strain Mu3 by electroporation at 30°C. Allelic exchange and anhydrotetracycline-inducible counterselection were performed as previously described (24). The deletion of the chromosomal prsA gene was confirmed by PCR and sequencing, generating strain AJ457.

TABLE 2.

Primers used in this study

No.	Name	Sequence (5′–3′)a	Description
1	attB2_prsA_down-R	GGGGACCACTTTGTACAAGAAGCTGGGTGCTACACAGGTGTAACAGC	pKOR1-based allelic exchange
2	attB1_prsA_up-F	GGGGACAAGTTTGTACAAAAAAGCAGGCATATGCCCTGCCATATCCAT	pKOR1-based allelic exchange
3	sacII_prsA_down-F	ATGCCCGCGGCACAAAACCGAGCGACCGTGG	pKOR1-based allelic exchange
4	sacII_prsA_up-R	ATGCCCGCGGAGTTGAAACTCCTTTGTAAG	pKOR1-based allelic exchange
5	upstream_primer_kpn_eco-F	ATGCGGTACCGAATTCTCCATATCATTTATAACAAAATAA	Flanking primer for prsA site-directed mutagenesis
6	prsA_PCR_bamHI-R	CGCGGATCCGAATTAAAAGATATCGGACAGATGAG	Flanking primer for prsA site-directed mutagenesis
7	prsA-serine-F	TTATTAGGCGCTTCTGGCGCTAGTGCCACA	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_serine)
8	prsA_serine-R	TGTGGCACTAGCGCCAGAAGCGCCTAATAA	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_serine)
9	OL_pHu_pMK4_prsA_NC_down-F	CTGATTCTGAAATTAAAGAACCAACAGACTTTAACAGTGA	Mutagenic primers for prsA site directed mutagenesis (COL_prsA_NterCter)
10	OL_pHU_pMK4_prsA_NC_up-R	TCACTGTTAAAGTCTGTTGGTTCTTTAATTTCAGAATCAG	Mutagenic primers for prsA site directed mutagenesis (COL_prsA_NterCter)
11	New_N_signal_PPIase-F	GCTTTATTATTAGGCGCTTGTGACAGCAAGAAAGCTTCACA	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_PPIase, COL_prsA_PPIase_Cter)
12	New_N_signal_PPIase-R	TGTGAAGCTTTCTTGCTGTCACAAGCGCCTAATAAAGC	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_PPIase, COL_prsA_PPIase_Cter)
13	PrsA_PPIase_UAG_Bam-R	GCATGGATCCCTATTTATCAGCTTTAATAATATGAT	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_PPIase, COL_prsA_Nter_PPIase)
14	N_UAG_Bam-R	GCATGGATCCTTATTTAATTTCAGAATCAG	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_Nter)
15	PrsA_C_N_signal-R	TCACTGTTAAAGTCTGTTGGACAAGCGCCTAATAATAAAGC	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_Cter)
16	PrsA_C_N_signal-F	GCTTTATTATTAGGCGCTTGTCCAACAGACTTTAACAGTGA	Mutagenic primers for prsA site-directed mutagenesis (COL_prsA_Cter)
17	PrsA-F	AGTTAATGATAAGAAGATTGACGAACAAA	prsA TaqMan
18	PrsA-R	GAAGGGCCTTTTCAAATTTATCTTT	prsA TaqMan
19	PrsA-P FAM/TAMRA	TGAAAAAATGCAAAAGCAATACGGCGG	prsA TaqMan
20	PPIase-F	TAA-AGT-TAA-ATC-TAA-GAA-AAG-CGA-CAA-AGA-A	prsA TaqMan
21	PPIase-R	CAA-ATT-TAC-TTG-GAT-CTT-TTG-AAA-CTT-C	prsA TaqMan
22	PPIase-P FAM/TAMRA	AGA-CGA-TAA-AGA-AGC-GAA-ACA-AAA-AGC-TGA-AGA-A	prsA TaqMan
23	mecA_MGB_F1525	TTCCACATTGTTTCGGTCTAAAATT	mecA TaqMan
24	mecA_MGB_R1603	AATGCAGAAAGACCAAAGCATACA	mecA TaqMan
25	mecA_MGB_1553 VIC	CCA CGT TCT GAT TTT AAA	mecA TaqMan

^aUnderlined regions represent restriction enzyme sites.

Construction of bacterial strains expressing truncated PrsA derivatives.

All primers used to generate PrsA derivatives are described in Table 2. Site-directed mutagenesis was performed by PCR-based overlap. Two independent PCRs were performed, carrying overlapping terminals. Both intermediate PCR products were mixed to act as the template DNA for a second PCR using primers flanking prsA. For the PrsA_NterCter construct, two intermediate PCR products were amplified with the primer pairs 5/10 and 9/6 and subsequently used as the templates to generate a final 1,120-bp PCR product using the flanking primer pair 5/6 (primer 5 encompassing the prsA promoter). For PrsA_PPIase, two intermediate PCR products, amplified with the primer pairs 5/12 and 11/13 (primers 11 and 12 encompassing the N-terminal signal sequence region, and primer 13 introducing an UAG stop codon) were used as the template to generate a final 780-bp PCR product with the flanking primer pair 5/13. A single 780-bp PCR product was amplified using the primer pair 5/14 (primer 14 introducing an UAG stop codon) for construction of the PrsA_Nter derivative. PrsA_Nter_PPIase was constructed from a single 1,110-bp PCR product, amplified using the primer pair 5/13 (primer 13 introducing an UAG stop codon). Two intermediate PCR products amplified with the primer pairs 5/12 and 11/6 were used as the template to generate a final 1,100-bp PCR product with the flanking primer pair 5/6 to construct PrsA_PPIase_Cter. For PrsA_Cter, two intermediate PCR products amplified with the primer pairs 5/15 and 16/6 were used as the template to generate a final 640-bp PCR product with the flanking primer pair 5/6. Final PCR products were cloned within EcoRI and BamHI restriction sites of the integrative plasmid pCL84 to yield pCL84_prsA*. After sequence verification, CYL316 harboring the bacteriophage L54a integrase was transformed with pCL84_prsA*, and transformants were selected on tetracycline-containing plates. The correct integration into the geh locus was verified by PCR and sequence analysis. The stably chromosomally integrated plasmid containing the complete prsA gene in CYL316 was then transduced into AJ728 prsA::kan strain using phage Φ80.

Spot test assay.

The spot population analysis profile (PAP) method was used to assess antibiotic resistance within the population, as previously described (10). Bacterial cultures were adjusted to 0.5 MacFarland standard (1.5 × 10⁸ bacteria/ml), corresponding to an optical density at 600 nm (OD₆₀₀) of 0.1. Serial dilutions (10⁻¹ to 10⁻⁵) were prepared, and then 10 μl of each dilution was spotted onto freshly prepared Mueller-Hinton agar plates containing different concentrations of antibiotics. Viable colonies were examined after 48 h of incubation at 37°C. The results reported were consistent across at least three independent assays and expressed as the CFU/ml.

Total protein extracts.

Protein extracts from S. aureus were prepared as previously described (9). Briefly, overnight bacterial cultures were incubated in TSB at 37°C with agitation. Cells were harvested by centrifugation and then washed and resuspended in 500 μl of TE buffer (10 mM Tris [pH 7.5], 1 mM EDTA). Bacterial cells were disrupted by adding 500 μl of acid-washed glass beads (100 to 200 μm; Sigma) and using a FastPrep cell disrupter (MP Biomedicals). The cell debris was separated from soluble protein extracts by centrifugation at 14,000 rpm (10 min at 4°C). Protein concentrations were determined using the bicinchoninic acid (BCA) kit assay (Pierce) according to the manufacturer's guidelines.

Membrane extracts.

Membrane proteins were isolated as previously described (25). Strains were grown in TSB until midexponential phase and pellets were resuspended in 1/250 of the initial culture volume of buffer A (50 mM KPO₄ buffer [pH 7.4], 10 mM MgCl₂). Bacterial cells were disrupted by adding 500 μl of acid-washed glass beads (100 to 200 μm; Sigma) and using a FastPrep cell disrupter (MP Biomedicals). The resulting lysate was centrifuged at 7,000 rpm for 30 min at 4°C. The supernatant was centrifuged at 80,000 rpm in a Beckman ultracentrifuge TL100 using a TLA 100.3 rotor for 1 h at 4°C, and the membrane pellet was washed with buffer B (50 mM KPO₄ [pH 7.4], 10 mM MgCl₂, 20% glycerol) and then resuspended in the same buffer. Total membrane proteins were quantified using a BCA kit assay (Pierce), diluted to 10 mg/ml in buffer B, snap-frozen with liquid N₂, and stored at −80°C.

Bocillin labeling.

Portions (100 μg) of membrane proteins were labeled with 100 μM bocillin-FL (Molecular Probes) for 10 min at 30°C. The reaction was stopped by adding 5-fold-concentrated SDS-PAGE sample buffer (500 mM dithiothreitol, 10% SDS, 250 mM Tris-HCl [pH 6.8], 30% glycerol, 0.02% bromophenol blue). Labeled membrane proteins (20 μg) were separated on a 7.5% SDS-PAGE gel and detected using a 473-nm laser of a Fuji FLA-5100 reader. The quantification of the intensity of the fluorescent bands was performed using ImageJ software.

Western blot.

Aliquots of proteins (20 μg) were separated on 10% SDS-PAGE gels and blot transferred onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad). After blocking using 5% low-fat milk in phosphate-buffered saline (PBS), PBP2A was probed with monoclonal anti-PBP2A antibody (Slidex MRSA detection; bioMérieux) at a 1/500 dilution, followed by incubation with a secondary horseradish peroxidase (HRP)-conjugated goat anti-mouse antibody at a 1/50,000 dilution. Chemiluminescence was detected using the Western Pico Super Signal reagent (Pierce). FtsZ was probed with an anti-FtsZ antibody at a 1/5,000 dilution, followed by incubation with a secondary HRP-conjugated anti-sheep antibody at a 1/50,000 dilution. PrsA was probed with anti-PrsA antibody at a 1/10,000 dilution, followed by incubation with a secondary HRP-conjugated goat anti-rabbit antibody at a 1/50,000 dilution. PBP2 was probed with anti-PBP2 antibody at a 1/5,000 dilution, followed by incubation with a secondary HRP-conjugated anti-rabbit antibody at a 1/50,000 dilution. The PBP2A protein levels were quantified by analyzing the signal intensity of each PBP2A and FtsZ band using ImageJ software. PBP2A levels were normalized to FtsZ protein levels, used as internal controls of total protein loading.

Quantitative transcript analysis by qRT-PCR.

mRNA levels were determined by quantitative reverse transcriptase PCR (qRT-PCR) using the one-step reverse transcriptase qPCR master mix kit (Eurogentec, Seraing, Belgium), as previously described (12). Appropriate prsA and mecA primers and probes were designed using PrimerExpress software (version 1.5; Applied Biosystems) and obtained from Eurogentec (Table 2). The mRNA levels of target genes from the different strains were normalized to 16S rRNA levels, which were assayed in each round of qRT-PCR as internal controls. The statistical significance of strain-specific differences in normalized cycle threshold (C_T) values of each transcript was evaluated by Student paired t test, and data were considered significant when P was <0.05.

Determination of antibiotic susceptibility.

Determination of the MIC to an array of antibiotics was performed in TSB by microdilution in 96-well plates. Overnight cultures of COL and COLΔprsA were added at a final cell density of 5 × 10³ CFU/ml to wells containing 2-fold dilutions of each antibiotic. Plates were incubated at 37°C for 24 or 48 h, and the MIC was recorded as the lowest concentration of antibiotic that inhibited bacterial growth. The results are the average of three independent MIC determination experiments.

Labeling and imaging S. aureus.

Parental and deletion mutant strains were grown overnight, diluted 1/200 in fresh medium and incubated at 37°C. At OD_{600 nm} of 0.5 to 0.7, 2 ml of each culture was taken and incubated for 5 min with the DNA dye Hoechst 33342 (3 μg/ml; Invitrogen) and the membrane dye Nile Red (10 μg/ml; Invitrogen). The cells were then pelleted and resuspended in PBS, and 1 μl of this cell suspension was placed onto a thin film of 1.2% agarose prepared in PBS mounted on a microscopy slide.

SR-SIM.

Super-resolution structured illumination microscopy (SR-SIM) imaging was performed using a Plan-Apochromat 63×/1.4 oil differential interference contrast M27 objective, in an Elyra PS.1 microscope (Zeiss). Images were obtained using 5 grid rotations, with 34 μm grating period for the 561 nm laser and 23 μm period for 405 nm laser. Images were acquired using a sCMOS camera and reconstructed using ZEN software (black edition, 2012, version 8.1.0.484) based on a structured illumination algorithm as previously described (26).

RESULTS

prsA deletion decreases β-lactam resistance.

PrsA was previously reported as required for high level of oxacillin resistance in the MRSA strain COL (12). MRSA strains are resistant to virtually all β-lactam antibiotics, including cefotaxime, cefaclor, and cefoxitin, preferentially targeting PBP2, PBP3, and PBP4, respectively. As observed by spot PAPs, the deletion of prsA causes a reduction in resistance to all tested β-lactams (Fig. 1). Introduction of a chromosomal copy of the prsA gene under the expression of its own promoter restored the high level of β-lactam resistance. The recent Food and Drug Administration-approved broad-spectrum antibiotic ceftaroline has unique activity against MRSA. Ceftaroline specifically inhibits PBP2A by simulating allosteric opening of the catalytic site for further inactivation by a second β-lactam molecule (27). We tested the effect of prsA deletion on COL and Mu3 MRSA strains, which are considered ceftaroline susceptible (MIC = 0.5 μg/ml) and resistant (MIC = 2 μg/ml) strains, respectively, according to EUCAST breakpoints (see www.eucast.org). Interestingly, spot PAP assays revealed that deletion of prsA increases susceptibility to ceftaroline in COL and reduces resistance in Mu3 (Fig. 1). Testing susceptibility to other cell wall active antibiotics (d-cycloserine, phosphomycin, and bacitracin) and antibiotics targeting protein synthesis and DNA replication (chloramphenicol and nalidixic acid) did not reveal any variation in MIC values in the COLΔprsA mutant (see Table S1 in the supplemental material).

FIG 1.

Effect of prsA deletion on resistance to cefotaxime, cefaclor, cefoxitin, and ceftaroline. Spot plating population analysis profiles (spot PAP) of COL, Mu3, and their corresponding prsA deletion (ΔprsA) and prsA-complemented (ΔprsA-C) strains on MHB agar containing cefotaxime, cefaclor, cefoxitin, or ceftaroline. The upper panels correspond to MHB agar without antibiotics. The lower panels correspond to MHB agar containing antibiotic. The antibiotic concentration is indicated at the left margin. Spot serial dilutions are indicated at the right margin. The first spot, 10 μl, corresponds to 10⁵ CFU.

Impact of prsA deletion on PBP profiles.

In B. subtilis, the PrsA foldase is involved in the stability of a subset of PBPs and the formation of proper rod-shaped cells (15). Considering that the S. aureus prsA is expressly required for resistance to cell wall-active antibiotics, especially agents targeting PBPs, we evaluated the effect of prsA deletion on staphylococcal PBP profiles. The membrane abundance and activity of PBP1, PBP2, PBP3, and PBP4 were assessed by fluorescent penicillin-derivative Bocillin-FL labeling. Membrane fractions extracted from wild-type and ΔprsA strains were incubated with Bocillin-FL and visualized by SDS-PAGE. In the methicillin-sensitive S. aureus strain Newman and its ΔprsA counterpart, all four native PBPs could be distinctly identified, and equal levels of membrane-bound protein were observed. Similar PBP profiles were obtained for the MRSA strain COL and its ΔprsA derivative (Fig. 2A). Accordingly, Western blot comparison of wild-type strains with their corresponding mutants showed no variation in the levels of membrane-bound PBP2 (Fig. 2B). These data show that the quantity and substrate-binding activity of the native, membrane-bound PBPs are not affected by deletion of prsA. Moreover, the deletion of prsA in COL and Newman strains has no impact cell morphology and cell wall thickness (12), and no major change in membrane aspect and organization was observed by SR-SIM (see Fig. S1 in the supplemental material).

FIG 2.

Effect of prsA on PBP profile. (A) Detection of PBP1, PBP2, PBP3, and PBP4 in membrane preparations of wild-type and ΔprsA strains. Equal amounts (20 μg) of Bocillin-FL-labeled membrane proteins were separated on a 10% SDS-PAGE gel. Fluorescently labeled PBPs are indicated by arrows. (B) Western blot analysis of PBP2 in S. aureus protein membrane extracts. PBP2 protein (80 kDa) was detected using rabbit-polyclonal anti-PBP2 antibodies. A total of 20 μg of membrane protein extract was loaded on a 10% SDS-PAGE gel. (C) Western blot analysis of PBP2A in S. aureus membrane protein extracts. PBP2A protein (70 kDa) was detected using rabbit-polyclonal anti-PBP2A antibodies. The lower panel corresponds to FtsZ antibody probing on the same PVDF membrane, used as a loading control. A total of 20 μg of membrane protein extract was loaded on a 10% SDS-PAGE gel.

The mecA gene encodes an additional PBP, PBP2A, with a markedly reduced affinity to β-lactams. In the presence of β-lactams, PBP2A maintains transpeptidation, while PBP2 provides transglycosylase activity in order to sustain peptidoglycan synthesis. PBP2A is considered the main effector of β-lactam resistance in MRSA strains, although several accessory factors account for full expression of resistance. Accordingly, we compared the levels of membrane-bound PBP2A in the COL and COLΔprsA strains (Fig. 2C). The cell division protein FtsZ was probed on the same PVDF membrane as a loading control. Membrane-bound PBP2A levels detected by Western blotting were significantly and reproducibly decreased in COLΔprsA, whereas FtsZ membrane levels were equal in both samples. Collectively, these results indicate that prsA deletion specifically causes a decrease in PBP2A membrane levels.

mecA is not transcriptionally regulated by prsA.

The mecA gene is located on an exogenous genetic mobile fragment called the SCCmec element (Staphylococcal Chromosomal Cassette mec) containing the mecA regulatory locus itself. The regulation of mecA expression has been described so far to occur at the transcriptional level and results from the activity of a two- or three-component system consisting of a transcriptional repressor (mecI), a sensor inducer (mecR1), and an eventual antirepressor (mecR2) (28). Besides the mecA regulatory locus itself (mecR1, mecI, and mecR2), the induction of mecA by β-lactams also includes the blaR1 and blaI system, driving expression of the blaZ gene that encodes a β-lactamase. To determine whether the decrease in resistance to β-lactams observed in COLΔprsA is linked to this described transcriptional regulation, we evaluated the contribution of prsA to oxacillin resistance using various MRSA strain backgrounds (COL, MW2, and Mu3), displaying differentially functional mecR1, mecI, and mecR2 gene and/or bla gene loci (Table 1). As judged by spot PAPs, prsA deletion reduces oxacillin resistance in all MRSA strains tested, COL, MW2, and Mu3 (Fig. 3A). We conclude that prsA affects oxacillin resistance regardless of the presence of blaR1 and blaI genes or mecR12, mecI, and mecR2 functional genes. To further understand the role of prsA in mecA expression, we monitored the levels of mecA mRNA in COL and COLΔprsA strains by qRT-PCR. As shown in Fig. 3B, COL, COLΔprsA, and COLΔprsA-C strains showed similar levels of mecA mRNA levels. Together, these data suggest that prsA alters oxacillin resistance and PBP2A membrane protein levels independently of the mec regulatory elements and without affecting mecA steady-state mRNA levels.

FIG 3.

Effect of prsA deletion on oxacillin resistance in different SCCmec strain backgrounds and on mecA gene transcription. (A) Oxacillin spot plating PAPs of COL (ΔblaI ΔblaR1 ΔblaZ ΔmecR1 ΔmecI ΔmecR2), MW2 (blaI⁺ blaR1⁺ blaZ⁺ ΔmecR1 ΔmecI ΔmecR2), and Mu3 (ΔblaI ΔblaR1 ΔblaZ mecR1 mecI mecR2⁺) and their corresponding derivatives. The upper panels correspond to MHB agar without oxacillin. The lower panels correspond to MHB agar containing-oxacillin, and the oxacillin concentration is indicated at the left margin. Spot serial dilutions are indicated at the right margin. The first spot, 10 μl, corresponds to 10⁵ CFU. (B) Steady-state levels of mecA transcript of wild-type COL strain compared to COLΔprsA and COLΔprsA-C strains, determined by qRT-PCR and normalized to 16S rRNA. Values represent the mean C_T values ± the standard deviations of three independent experiments.

Expression of PrsA NC-terminal and PPIase domains.

To determine which domains of the PrsA protein are critical for β-lactam resistance and PBP2A protein levels (see below), we first constructed strains expressing only the individual PrsA domains. The canonical organization of PrsA protein family members comprises a central PPIase domain flanked by N-terminal (Nter) and C-terminal (Cter) regions. Although the PPIase activity of PrsA from B. subtilis, S. aureus, and L. monocytogenes has been confirmed in vivo, it is not an absolute requisite for essential in vivo roles (20). Furthermore, the signature motif for peptidyl-prolyl PPIases is absent in the PrsA-like proteins of several streptococci (29). We complemented the prsA deletion strain by chromosomal insertion into the geh locus of wild-type prsA (COLΔprsA-C), prsA Nter+Cter (COLΔprsA-NterCter), and/or prsA PPIase (COLΔprsA-PPIase) as depicted in Fig. 4A. S. aureus PrsA domains were previously determined by homology with other PrsA proteins (19). Of note, all constructs were expressed using the endogenous prsA promoter and include the N-terminal signal sequence required for membrane addressing and the cysteine residue necessary for membrane anchorage (Fig. 4A and Table 2).

FIG 4.

Expression of PrsA and truncated PrsA derivatives. (A) Schematic representations of PrsA domain derivatives used in this study, according to Heikkinen et al. (19). Each PrsA domain derivative was introduced into to the prsA deletion strain by chromosomal insertion in the geh locus. The white box represents the first 140 N-terminal residues of PrsA containing the cysteine membrane-anchored amino acid, the gray box represents the 105 amino acid residues of the PrsA PPIase domain, and the darker gray box corresponds to the 75 amino acid residues of the C-terminal domain. The corresponding detection (+) or absence (−) of prsA mRNA or PrsA protein of each genetic construct is shown. (B) PrsA Western blot of total protein extracts from strains COL, COLΔprsA, and COLΔprsA complemented in the geh locus with wild-type prsA (COLΔprsA-C), with N- and C-terminal domains (COLΔprsA-Nter-Cter) or with PPIase domain (COLΔprsA-PPIase). Total protein extracts were run in a 15% SDS gel, and Western blotting was done with a PrsA-specific antibody (see Materials and Methods). For the detection of the PPIase domain, a longer exposure time was used.

By RT-PCR using specific N-terminal or PPIase domain TaqMan probes, we demonstrated that all constructs express prsA mRNA (Fig. 4). The level of expression of each construct was verified by Western blotting of total protein extracts using an anti-PrsA antibody. As shown in Fig. 4B, a 35-kDa protein band, corresponding to wild-type PrsA was detected in COL and COLΔprsA-C (lanes 2 and 3), while no band was detected in in COLΔprsA (lane 1). NterCter and PPIase versions of PrsA appear as 24- and 11-kDa bands, respectively, corresponding to the expected size of each construct. Similar amounts of PrsA were observed in COL, COLΔprsA-C, and COLΔprsA-NterCter but levels of PrsA-PPIase are significantly lower compared to the wild type (the autoradiogram presented was overexposed to allow detection of PrsA-PPIase). Interestingly, all of the constructs devoid of N-terminal domains exhibit smaller amounts of PrsA. Since RT-PCR using domain-specific probes demonstrated equal amounts of prsA mRNA in all of the constructs, we propose that PrsA protein stability is compromised in the absence of the N-terminal domain.

PrsA domains involved in PBP2A regulation and β-lactam resistance in COL.

Since PrsA affects PBP2A protein levels in the membrane, we explored the function of PPIase and the flanking domains on PBP2A expression by Western blotting. Previous subcellular fractionation experiments failed to detect PBP2A in the cytosolic fraction, indicating that PBP2A is predominantly present as a membrane protein (data not shown), presumably targeted to the membrane concomitantly with synthesis. Equal amounts of total protein extracts were used for the purpose of this PBP2A Western blot. FtsZ was probed on the same PVDF membrane as a loading control (Fig. 5A). In agreement with Fig. 2C, COLΔprsA strain showed a reduction of PBP2A protein levels compared to the wild-type strain. The introduction of PrsA in the prsA deletion strain restored wild-type PBP2A protein levels. Interestingly, the reduction of PBP2A protein levels in COLΔprsA is accompanied with reduced oxacillin resistance (Fig. 5B). When introducing the PrsA-NterCter domain into COLΔprsA, nearly complete restoration of wild-type PBP2A protein levels was observed with a complete restoration of oxacillin resistance (Fig. 5). Expression of the PrsA-PPIase domain alone did not restore either PBP2A levels or oxacillin resistance (Fig. 5). Altogether, analysis of both PBP2A protein levels and oxacillin resistance suggest that PrsA-NterCter is sufficient to restore a proper amount of active PBP2A protein to help oxacillin resistance. Even considering the low expression of the PPIase construct, these observations suggest that the PPIase central domain is dispensable for oxacillin resistance.

FIG 5.

Effect of PrsA domains on PBP2A protein levels and oxacillin resistance. (A) The upper panel shows a PBP2A Western blot of the total protein extracts from strains COL, COLΔprsA, and COLΔprsA complemented in the geh locus with wild-type prsA (COLΔprsA-C), with N- and C-terminal domains (COLΔprsA_NterCter), or with PPIase domain (COLΔprsA_PPIase). Total protein extracts were run on a 10% SDS gel, and the membrane was probed with PBP2A- and FtsZ-specific antibodies (see Materials and Methods). The lower panel shows a quantification of three different Western blot membranes detecting PBP2A and FtsZ proteins, performed as described above. Relative abundance (AU) of PBP2A was measured by normalizing to FtsZ protein levels used as internal controls of total protein loading. An asterisk (*) denotes PBP2A protein levels significantly (P < 0.05) different from wild-type levels. (B) Oxacillin spot plating PAPs of COL, COLΔprsA, COLΔprsA-C, COLΔprsA_NterCter, and COLΔprsA_PPIase. Spot serial dilutions are indicated in the right margin. The first spot, 10 μl, corresponds to 10⁵ CFU.

DISCUSSION

Bacterial cell wall homeostasis relies on a finely tuned equilibrium between peptidoglycan synthesis and degradation, involving a substantial collection of partially redundant peptidases, glycosylases, and lytic enzymes. Among CWAA disturbing this fragile balance, β-lactams inhibit PG synthesis by binding to the active site of PBPs, leading to an acylated, inactive form of the enzyme. PBPs are fundamental to the maintenance of bacterial cell wall integrity. They participate in the late stages of PG biosynthesis catalyzing the transglycosylation and/or transpeptidation of precursors into the growing PG. The four native staphylococcal PBPs have been functionally and structurally scrutinized over the past 40 years (30–35). MRSA strains possess an exogenous PBP with very low affinity for β-lactam antibiotics, namely, PBP2A. In the presence of β-lactams, PBP2A functionally cooperates with the transglycosylase domain of PBP2 in order to continue PG synthesis (31). Although the MRSA phenotype is definitely multifactorial, PBP2A is the main effector of the MRSA phenotype; hence, the value of dissecting the processes, leading to optimal synthesis and secretion of an active PBP2A.

This study focused upon the role of PrsA as an auxiliary factor required for expression of β-lactam resistance. We provided evidence that S. aureus PrsA may contribute to the recruitment of correct amounts of active PBP2A at the cell membrane, by an as-yet-unknown mechanism, and that it exerts its effect on PBP2A at the posttranscriptional level. This study provides one of the first descriptions of PBP2A regulation occurring outside the mec locus.

Acquisition and expression of the mecA gene alone is not sufficient for a uniform resistance to β-lactams. The optimal expression of β-lactam resistance relies on the mecR1/mecI/mecR2 and blaR1/blaI complementary systems regulating mecA transcription. Recent large-scale studies extended the list of auxiliary factors required for β-lactam resistance (36, 37). In addition, spontaneous mutations on the chromosome referred to as chr* have been reported to contribute to “hetero-homo” conversion of β-lactam resistance (38, 39). PrsA was, curiously, never identified in these previous screenings, although its deletion has a drastic impact upon β-lactam resistance in three different SCCmec MRSA backgrounds.

Despite the large number of publish studies addressing the function, structure, and localization of PBPs, the mechanisms of secretion of these enzymes remain elusive. By virtue of specific N-terminal signal peptides, proteins directed to the membrane or to extracellular compartment are predominantly translocated by the general secretory (Sec) system or the twin-arginine translocation (Tat) pathway. The Sec pathway targets N-terminal signal peptide containing sequences and exports proteins in an unfolded conformation, while folded proteins with a characteristic motif, including two consecutive arginine residues, are exported via the Tat pathway. The prominent Sec translocation pathway is based on a three step process. The nascent polypeptide emerging from the ribosome interacts with the signal recognition particle. The signal recognition complex directs the nascent polypeptide to the Sec membrane machinery consisting of SecA, SecYEG, and accessory proteins SecDF. Shortly after translocation, secretory preproteins are processed by a signal peptidase. In this scenario, newly secreted proteins need to rapidly fold into an active conformation, since the unfolded state is rather prone to degradation, especially for Gram-positive bacteria that by nature lack a protective outer membrane. In B. subtilis, the capacity of the secretion apparatus is directly dependent on PrsA (40), with α-amylase, protective agent, and PBPs specifically identified as PrsA substrates (41). Although numerous studies have attempted to define the S. aureus secretome, the subset of Sec exported proteins is a missing piece (42–44). PBP profiles are not affected by deletion of the secDF accessory factors of the Sec machinery (45); nevertheless, PBP3 is 1 of the 46 proteins affected by the inhibition of type 1 signal peptidase (21), and the SOSUI algorithm predicts N-terminal signal sequences for PBP4 and PBP2A (46). As supported by qRT-PCR experiments, the decreased level of membrane-bound PBP2A cannot be attributed to a defect in mecA transcription and/or mecA mRNA destabilization. Since PrsA and PBP2A share a membrane localization in S. aureus, it is tempting to speculate that PrsA helps with the folding of PBP2A. In order to elucidate the precise mechanisms by which PrsA affects the levels of PBP2A in the membrane more thorough biochemical investigations are required, including the testing for a direct interaction between PBP2A and PrsA.

In an effort to appreciate the structural characteristics of PrsA that are decisive for a PBP2A membrane protein level decrease, our structure function study revealed the essential role of the N- and C-terminal domains upon PBP2A regulation and oxacillin resistance. In spite of the relative conservation of PrsA structural organization among Gram-positive bacteria, mechanistic aspects and substrate specificities are often not conserved. The peptidyl-prolyl isomerase in vitro activity lies in the central domain of the PrsA like proteins, but this activity is not required for the tested cellular functions in B. subtilis (20) and apparently not in S. aureus since the truncated construct lacking the central domain is sufficient for correct protein level of PBP2A and subsequent oxacillin resistance. Conversely, in vivo L. monocytogenes infection in a mouse model is strictly dependent on the PPIase domain of PrsA, even if secretion of various virulence factors is restored by the NterCter domain alone (17). The PPIase domain seems critical for a defined set of substrates, but the NterCter domain is responsible for mediating the most essential functions of PrsA.

There are two reports identifying discrepancies between mecA mRNA and PBP2A protein levels in the literature to date. First, it was reported that stringent response induced by mupirocin triggers an increase of PBP2A protein levels without affecting mecA transcription (39). Second, insertional inactivation of the vraS gene in a CA-MRSA strain induced mecA mRNA, but the PBP2A protein levels do not reflect this change. In this model, PBP2 overexpression was not enough to restore β-lactam resistance (47). PrsA has no recognized role in the stringent response to date; however, we have previously shown that PrsA is a member of the vraRS regulon (12). In a vraS-inactivated strain, the low levels of PrsA could account for the reduced levels of PBP2A observed, despite the proper induction of mecA transcription by mec regulatory elements. Here, we propose a model where PrsA is involved in the posttranscriptional regulation of mecA after β-lactam exposure, where PrsA-assisted folding of PBP2A is one hypothesis among others (Fig. 6). Upon prsA deletion, PBP2A active membrane levels decrease, correlating with a further decrease in β-lactam resistance. PrsA is also involved in glycopeptide resistance, which does not absolutely rely on PBP2A expression (12). This implies that PBP2A is not the only cell wall-related substrate of PrsA.

FIG 6.

Model of proposed PrsA posttranscriptional regulation of PBP2A following β-lactams exposure. (Step 1) β-Lactam-induced mecA and prsA transcription. The dashed arrows denote the presumed pathways leading to PrsA effects on PBP2A. PrsA protein may affect directly or indirectly PBP2A (step 2) at the translational level, and/or PBP2A protein intracellular stability (step 3), and/or (step 4) PBP2A cell membrane export/folding. The resulting PBP2A and PrsA proteins will ultimately be found active in the cell membrane compartment (49–53).

The expression of the MRSA phenotype key player mecA proves to be meticulously controlled even after transcription. The PrsA proteins expressed by B. subtilis and S. aureus cannot substitute for all the functions of L. monocytogenes PrsA2, but PBPs are overlapping PrsA substrates in these three Gram-positive organisms (48). Unraveling the connection between PBPs and secretion could reshuffle our approaches to examine cell wall synthesis and reveal unexpected targets for innovative anti-infective strategies.