The Arginine Catabolic Mobile Element Is Not Associated with Enhanced Virulence in Experimental Invasive Disease Caused by the Community-Associated Methicillin-Resistant Staphylococcus aureus USA300 Genetic Background ^▿

Isolate selection.

A convenience sample of 90 MRSA clinical isolates representing the multilocus sequence type (ST) 8 lineage (presumed to represent USA300) was selected for evaluation from our isolate library at the University of Chicago (5). These isolates caused a variety of disease syndromes associated with CA-MRSA, including skin and soft tissue infections, pneumonia, bacteremia, and sepsis. Each isolate contained the gene encoding the chromosomal arginine deiminase and the lukSF-PV genes encoding PVL. In addition, strains SF8300 (wild-type USA300) and SF8300ax (ACME deletion mutant) were a generous gift from Henry Chambers (University of California San Francisco). They were previously described (8). Strain LAC was generously provided by Michael Otto (National Institutes of Health) (19).

Colony PCR was performed to assess the presence of the ACME-arcA gene (primers 5′-GAGCCAGAAGTACGCGAG-3′ and 5′-CACGTAACTTGCTAGAACGAG-3′) (8). Conditions for PCR were as described previously (8). Eight clinical isolates, three of which contained ACME-arcA and five of which did not, were selected in addition to LAC, SF8300, and SF8300ax for pulsed-field gel electrophoresis (PFGE) (Table 1).

PCR was performed with genomic DNA prepared from the selected isolates to assess the presence of the opp-3 locus (primers 5′-GCAAATCTGTAAATGGTCTGTTC-3′ and 5′-GAAGATTGGCAGCACAAAGTG-3′) (8) and to confirm the presence or absence of ACME-arcA. The presence of both of these loci was used to assume the presence of ACME, as previously described (8). The relative virulence of strains LAC, 923, and 2145 in experimental pneumonia in rats was previously characterized (16).

PFGE.

Whole-cell-DNA-containing agarose plugs were prepared and digested with SmaI as described previously (15). Restriction fragments were resolved on a Chef DR-III (PFGE) apparatus (Bio-Rad). Strain LAC served as a standard for the USA300-0114 pulsotype.

In vitro gene expression.

Frozen stock isolates were subcultured on tryptic soy agar (TSA) and incubated overnight at 37°C in tryptic soy broth (TSB) with shaking at 250 rpm. The overnight culture was diluted 1:100 in fresh TSB (flask/volume ratio, 7:1). Bacteria were harvested during the logarithmic growth phase (3 h, optical density at 600 nm [OD₆₀₀] = 2.4) and frozen at −80°C in RNAprotect (Qiagen). The growth curves of each isolate were similar, as determined by serial OD₆₀₀ measurements and plating of appropriate serial dilutions at several time points (data not shown). Bacteria were lysed using the Fastprep FastRNA Pro Blue system (Q-BIOgene). Purification of RNA was achieved using the Qiagen RNeasy kit, including treatment with DNase. cDNA was prepared using the high-capacity cDNA Archive kit (Applied Biosystems).

Real-time reverse transcriptase PCR (RT-PCR) was performed to assess the expression of lukF-PV and hla by using an ABI Prism 7300 series RT-PCR thermocycler (Applied Biosystems) as described previously (16). LUX primers (Invitrogen) were designed for saeRS by using a forward primer labeled with 6-carboxyfluorescein (5′-CGTAAAGTCCAAGGGAACTCGTTTTA[FAMc]G-3′) and an unlabeled reverse primer (5′-TTGTTCAACTGGTTGATGATGG-3′). Data were analyzed using the ABI Prism 7300 sequence detection software version 1.2.3 (Applied Biosystems). 16S rRNA expression was used as a control for each isolate. The forward primer (5′-CGGCCTAACTACGTGCCAGCAGC[JOEc]G-3′) was labeled with 6-carboxy-4,5-dichloro-2,7-dimethoxyfluorescein (JOE), and the reverse primer (5′-GCGCTTTACGCCCAATAATTCC-3′) was left unlabeled. Quantification of expression was performed for each isolate relative to strain 451. Northern blot analysis was performed as described previously for agr (RNAIII) (16). RNA (5 μg) was added to each lane, and the equality of loading was confirmed by analysis on a 1.2% agarose-0.66 M formaldehyde gel (data not shown). Densitometry was performed using Molecular Imaging software v4.0 (Kodak). Net intensities were compared between ACME-positive and ACME-negative isolates.

Protein A and α-hemolysin abundance.

S. aureus isolates were grown as described above to stationary growth phase (6 h, OD₆₀₀ = 6). Cells were pelleted by centrifugation, and supernatants were removed. An equal volume of supernatant was added for each isolate (10 μl). Western blotting was performed on the supernatants to assess protein A abundance using a rabbit monoclonal antibody against protein A (Sigma) and anti-rabbit immunoglobulin G horseradish peroxidase-conjugated secondary antibody (Promega). For α-hemolysin, a mouse polyclonal antibody (Sigma) was used as the primary antibody and anti-mouse immunoglobulin G conjugated to horseradish peroxidase served as the secondary antibody (Promega). The membranes were developed using a chemiluminescent substrate (Pierce). Densitometry was performed as described above.

Preparation of bacteria for inoculation.

To prepare an animal inoculum, a frozen stock of S. aureus was subcultured onto TSA and incubated overnight in TSB at 37°C. The overnight culture was diluted 1:100 in fresh TSB and grown to early stationary phase (OD₆₀₀ = 3.0). The cells were then centrifuged (3,200 × g for 10 min), washed in phosphate-buffered saline (PBS), and resuspended in PBS. All bacteria in the inocula were quantified by plating serial dilutions on TSA and enumerating the colonies. All animal experiments were conducted in accordance with the regulations of the University of Chicago Institutional Animal Care and Use Committee.

Rat model of CA-MRSA pneumonia.

A model of necrotizing pneumonia in rats has been described (16). Briefly, male Sprague-Dawley rats (Harlan) weighing approximately 300 g each were housed for 7 days prior to inoculation. Animals were anesthetized with ketamine and xylazine. The trachea of each animal was intubated with an 18-gauge angiocatheter (Becton-Dickinson), and S. aureus (9 × 10⁸ CFU) or PBS was instilled (100 μl). The catheter was removed, and the animal was held upright for 15 s. Animals were observed for signs of illness. All surviving animals were euthanized at 24 h with pentobarbital.

Quantitative culture and histopathology in the pneumonia model.

After sacrifice, the left lung was removed and placed in PBS. The lung was washed twice in PBS and homogenized (Tissuemiser; Fisher); serial dilutions of the homogenate were plated on mannitol salt agar for quantification of bacteria. The right lung was placed in formalin, sequentially infiltrated with increasing concentrations of ethanol and xylene, and embedded in paraffin. The tissues were sectioned and stained with hematoxylin and eosin. Lung sections were graded by an observer in a blind manner according to a previously described scoring scheme, with a scale of 0 to 11 (16). The severity of the pulmonary pathology was based on the distribution and extent of inflammation, the presence of necrosis, and the presence of bacterial colonies.

Mouse skin infection model.

Based on the method of Voyich et al. (19), Crl:SKH1-hrBR hairless immunocompetent mice (Charles River) were inoculated by subcutaneous injection in the right flank with 50 μl of S. aureus (1 × 10⁷ CFU) or PBS (bacterial growth described above). The area of dermonecrosis was the product of the length and the width of the lesion.

Statistical analysis.

Mortality rates, the prevalences of necrotizing pneumonia, and the prevalences of dermonecrosis were compared by Fisher's exact test. Quantitative cultures, expression data, and dermonecrosis areas were compared by Student's t test. Pathology scores were compared by the Mann-Whitney U test (SPSS-W v15). Statistical significance was considered present when the P value was <0.05.

Isolate selection.

Of the 90 ST 8 MRSA isolates screened by colony PCR, 8 lacked ACME-arcA. Three ST 8 isolates containing ACME-arcA were selected for further analysis, as were ACME-arcA-positive strains LAC and SF8300. Five ST 8 isolates lacking ACME-arcA and SF8300ax, an isogenic ACME deletion mutant used as a comparator with SF8300, were also chosen (Table 1). All ACME-arcA-positive isolates also contained the opp-3 operon, as determined by PCR, and thus were considered to contain ACME (8). All ACME-arcA-negative isolates also lacked the opp-3 operon and were considered ACME negative.

PFGE.

The pulsotype of all five selected ACME-positive isolates was consistent with USA300-0114 (Fig. 1). The six ACME-negative isolates (including SF8300ax) were also USA300. Each ACME-negative isolate had a single band shift compared with the ACME-positive isolates (Fig. 1). Diep et al. have demonstrated that this shift is due to the absence of ACME (8). Additionally, strains 451, 1185, and 1358 had a single, identical, additional band difference; these strains were also missing a single low-molecular-weight band present in the other strains (Fig. 1).

In vitro virulence factor expression and production.

In vitro expression of lukF-PV by RT-PCR varied slightly among the tested isolates, but the magnitude of the variation was small (less than approximately twofold relative to that of strain 451) (Fig. 2B). There was no significant difference between the mean lukF-PV expression in ACME-positive isolates and that among the ACME-negative isolates (P = 0.68). Expression of lukF-PV was unchanged in SF8300ax compared with SF8300. There was also no difference in agr (RNAIII) expression as determined by Northern blotting between ACME-positive isolates and ACME-negative isolates (P = 0.19) (Fig. 2D). Similarly, there was no difference in mean saeRS expression as determined by RT-PCR between ACME-positive isolates and ACME-negative isolates (P = 0.24) (Fig. 2C).

The strains we tested had considerable variation in hla expression by RT-PCR; for example, expression was increased >11-fold in strains LAC and 1135 compared with strain 451 (Fig. 2A). Nevertheless, there was no significant difference between the mean hla expression among ACME-positive isolates and the mean among the ACME-negative isolates (P = 0.85). Expression of hla was 1.8-fold higher in SF8300ax than in SF8300.

The abundance of α-hemolysin and protein A (Spa) in 6-h culture supernatants, as assessed by Western blotting, also varied among the isolates (Fig. 2D). ACME-positive isolates had a higher mean abundance of α-hemolysin than did ACME-negative isolates, as assessed by densitometry (P = 0.04). SF8300ax had a 1.4-fold-higher α-hemolysin abundance than SF8300. Although there was some interstrain variability, there was no significant difference between the protein A abundance in culture supernatants of ACME-positive isolates and that in ACME-negative isolates (P = 0.96). In addition, there was a 2.9-fold-higher protein A abundance in SF8300ax than in SF8300.

Comparison of hla expression at 3 h with α-hemolysin protein abundance in 6-h culture supernatants revealed a weak correlation (r = 0.42; data not shown). Expression of hla was more strongly correlated with saeRS expression (r = 0.65, data not shown). This was not surprising, because expression of hla is controlled, at least in part, by the expression of saeRS (21).

Rat model of pneumonia. (i) Clinical isolates.

Two ACME-positive (LAC and 923) and two ACME-negative (1135 and 1866) USA300 isolates were selected for evaluation in the rat model of pneumonia (seven to eight animals/group). Animals inoculated with 9 × 10⁸ CFU of either isolate became ill, characterized by decreased mobility and labored breathing. All animals inoculated with LAC or 1135 died by 24 h; 50% of animals inoculated with 923 and 62% of animals inoculated with 1866 died (Fig. 3A). Interestingly, strains LAC and 1135 had higher expressions of hla and abundances of α-hemolysin than strains 923 and 1866, although the differences were <3-fold (Fig. 2A and D). There was a trend toward higher in vivo bacterial recovery from the lung after inoculation with LAC or 1135 than with 923 and 1866, but these differences were not significant (Fig. 3C). When data from animals inoculated with strains LAC and 923 (ACME positive) and 1135 and 1866 (ACME negative) were pooled, there was no significant difference in survival (P = 0.7) or in vivo bacterial recovery (P = 0.78), regardless of the presence of ACME (Fig. 3B, D). Pulmonary necrosis (pathology score, ≥7) was common after inoculation with LAC (75%), 923 (57%), 1135 (100%), or 1866 (71%). The median pathology scores were nearly the same, regardless of the strain inoculated (Fig. 3E) or the presence of ACME (Fig. 3F).

(ii) Isogenic wild-type/ACME deletion mutant pair.

There was no significant difference in the clinical features of illness after inoculation with 1 × 10⁹ CFU of either SF8300 or SF8300ax (21 animals/group). Among infected animals, 73% infected with SF8300 and 76% infected with SF8300ax died within 24 h of inoculation (P = 0.77) (Fig. 4A). There was no significant difference in bacterial recovery from the lung, regardless of the presence of ACME (P = 0.42) (Fig. 4B). Necrotizing pneumonia was present in all animals inoculated with either isolate. The median pathology score was slightly higher in animals inoculated with SF8300ax, but the difference was not significant (P = 0.13) (Fig. 4C).

Mouse model of skin infection. (i) Clinical isolates.

Mice were inoculated with 1 × 10⁷ CFU of 1 of 10 S. aureus isolates (5 ACME positive, 5 ACME negative; 5 to 20 animals in each group). A local dermonecrotic lesion occurred in 40 of 55 animals (72%) inoculated with an ACME-positive isolate and in 35 of 45 animals (78%) inoculated with an ACME-negative isolate (P = 0.37). The mean maximum areas of dermonecrosis ranged from 19.8 mm² (strain 2609) to 91.4 mm² (strain SF8300) in the ACME-positive isolates (Fig. 5A) and from 38.4 mm² (strain 451) to 90.6 mm² (strain 1185) in the ACME-negative isolates (Fig. 5B). Although there were strain-specific differences, when data were pooled according to the presence or absence of ACME in the inoculated strain, there was no significant difference in the mean maximum areas of dermonecrosis (ACME positive, 58.6 ± 7.6 mm²; ACME negative, 62.1 ± 11.0 mm²; P = 0.79) (Fig. 5C). The mean maximum area of dermonecrosis was not significantly associated with either expression of hla (r = 0.44) or α-hemolysin abundance (r = 0.2).

(ii) Isogenic wild-type/ACME deletion mutant pair.

To further define the role of ACME in experimental skin infection, mice were inoculated with 1 × 10⁷ CFU of SF8300 or SF8300ax (24 or 25 animals/group). Dermonecrosis was present in 22/25 (88%) animals inoculated with SF8300 and 21/24 (88%) animals inoculated with SF8300ax (P = 0.65). The mean maximal area of dermonecrosis was slightly greater after inoculation with SF8300 (80.4 ± 10.2 mm²) than with SF8300ax (71.2 ± 10.8 mm²), but the difference was not significant (P = 0.53) (Fig. 6A). There was no difference in the appearances of lesions after inoculation with SF8300 or SF8300ax (Fig. 6B).