PheP, a Putative Amino Acid Permease of Staphylococcus aureus, Contributes to Survival In Vivo and during Starvation

Abstract

PheP, a putative amino acid permease in Staphylococcus aureus, contributes to starvation survival under glucose-limiting conditions and virulence. A pheP mutation led to poor growth after microaerobic or anaerobic incubation on pig serum agar, which was recovered by phenylalanine addition. Genetic complementation of pheP restored growth and starvation survival.

The bacterial pathogen Staphylococcus aureus has the ability to cause a wide variety of infections ranging from endocarditis to pneumonia and osteomyelitis. While our knowledge of the genetics and physiology of this pathogen is increasing rapidly, we still have a poor understanding of S. aureus virulence in vivo (12).

Genes important for S. aureus survival in vivo have mostly been screened for using signature-tagged mutagenesis (STM) (4, 11). In these screens several classical virulence genes were identified; however, a large proportion of the loci found encoded proteins ascribed functions in pathways for cellular metabolism, transport, and biosynthesis (4). Furthermore, mutants with defects in the transport of amino acids formed a majority of these attenuated STM isolates (11).

One of the mutants identified by STM was extensively characterized and shown to have an insertion in a gene encoding the high-affinity proline permease, PutP (16). Large- and small-pool screening suggested that the putP mutant was attenuated in all of the animal models tested (abscess, bacteremia, endocarditis, and wound) (1, 16). The in vivo attenuation of the putP mutant was proposed to result from a subsequent deficiency in the ability to scavenge proline (16).

A screen of S. aureus Tn917 library transposants with an altered ability to survive carbon starvation identified strain ST1. Sequencing determined that the transposon insertion had produced a large deletion of the katA gene, encoding catalase, and a smaller deletion in the divergently transcribed pheP gene, which encodes a putative amino acid permease (Fig. 1A and B). Insertions of Tn917 (and transposons in general) can generate deletions adjacent to insertion junctions (24), and this is a likely explanation for the deletion observed in ST1.

FIG. 1.

Schematic representation of the pheP-katA locus. The direction of pheP and katA transcription is shown with large arrows and putative transcription terminator structures with ball and stick. (A) 8325-4 (wild-type). (B) Mutant ST1 showing the insertion of Tn917 with the corresponding 380-bp deletion of the 1,542-bp coding region of the pheP gene and 1,329-bp deletion of the 1,419-bp coding region of the katA gene. (C) MJH600 (pheP) showing allelic replacement. Through use of Northern blot and start-site mapping methods (6, 8) (data not shown), the pheP gene was found to be transcribed from a single σ^A-dependent promoter as a 1.55-kb monocistronic mRNA, which has a 3′ Rho-independent terminator sequence (ΔG = −15.5 kcal/mol) located 13 bases downstream from the stop codon.

Analysis of the PheP sequence reveals that it has 12 noncontiguous regions of hydrophobicity indicative of membrane-spanning domains. The protein has significant homology with members of the amino acid-polyamine-organocation superfamily of transport proteins (9, 13). S. aureus PheP exhibits greatest sequence identity with Escherichia coli LysP (44.6%), which functions as a lysine permease, and Bacillus subtilis RocE (34.9%), which functions as an arginine and ornithine permease.

When the virulence of ST1 (pheP katA) was tested in a murine abscess model of infection, it had significantly reduced recovery (0.15%) (P < 0.003) compared to the isogenic parent strain, 8325-4. The level of recovery for ST1 (pheP katA) was similar to that for the virulence regulator mutant, sarA (0.04%) (P < 0.003) (Fig. 2). In contrast the recovery of ST16 (katA), as described previously (7), was not significantly different from that of the 8325-4 (wild type), suggesting that the reduced virulence of ST1 (pheP katA) was due to inactivation of the pheP gene.

FIG. 2.

Pathogenicity of S. aureus strains in a murine skin abscess model of infection. Approximately 10⁸ CFU of each strain was inoculated subcutaneously into 6- to 8-week old BALB/C mice: 8325-4 (wild-type) (n = 10), ST1 (pheP katA) (n = 10), MJH600 (pheP) (n = 10), and PC1839 (sarA) (n = 10). Seven days after infection, mice were euthanized, lesions were removed and homogenized, and viable bacteria were counted after dilution and growth on brain heart infusion agar plates (3, 7). Significantly reduced recovery was observed for ST1 (pheP katA) (0.15%) (P < 0.003), MJH600 (pheP) (1.9%) (P < 0.003), and PC1839 (sarA) (0.04%) (P < 0.003). The dashed line shows the limit of recovery. Bar indicates mean value of recovery. Statistical significance was evaluated on the recovery of strains by using the Student t test with a 5% confidence limit.

An allelic replacement mutant, MJH600 (pheP) (Fig. 1C), was constructed to determine if inactivation of pheP was the contributing factor for the reduced virulence of ST1. Allelic replacement was achieved by amplifying the pheP gene in upstream and downstream fragments by using primers CCAGAATTCTGCCAATGATTAACTCTAATCG with ATGATGGTACCAGTAGCTACAAATAGACCAGTCC and AGAGGATCCGCATGTCGCAATCGTATTTGTGACC with GGACTGGTCTATTTGTAGCTACTGGTACCATCAT. The tetracycline resistance gene (tet) from pDG1513 (5) was amplified by using primer CCGGTACCCGGATTTTATGACCGATGATGAAG with CCGGTACCTTAGAAATCCCTTTGAGAATGTTT. Following purification, the three separate PCR products were digested with BamHI/KpnI, EcoRI/KpnI, and KpnI, respectively, and were simultaneously ligated to BamHI/EcoRI- digested pAZ106 (10, 14). One tetracycline-resistant clone, pMAL32, was used to transform electrocompetent S. aureus RN4220 (15) and was resolved by outcross via transduction of S. aureus 8325-4 by using φ11.

MJH600 (pheP) had reduced virulence (P < 0.003) compared with 8325-4 (wild type) when tested in a murine abscess model (Fig. 2). This demonstrated that reduced virulence was associated with mutation of pheP and confirmed that the inactivated permease gene was the determinant responsible for the reduced virulence of ST1. The virulence of ST1 (katA pheP) was significantly reduced (P < 0.02) relative to that of MJH600 (pheP), suggesting that katA might contribute to survival in a pheP mutant; the reason for this is unclear, since mutation of katA alone had no effect on virulence in this model, as described previously (7). MJH600 (pheP) showed an exoprotein profile similar to that of 8325-4 (wild type), ruling out major effects of the mutation on expression of known extracellular virulence factors (data not shown). Reduced virulence of MJH600 (pheP) was similarly observed in a Drosophila melanogaster model of infection (A. Needham and S. J. Foster, unpublished data).

ST16 (katA) has reduced capacity to survive glucose starvation (7, 22). Since ST1 (katA pheP) had reduced survival compared to ST16 (katA) during prolonged aerobic incubation in glucose-limiting chemically defined medium (CDM), the starvation survival of MJH600 (pheP) was tested. MJH600 (pheP) lost viability more rapidly than the parental strain, such that after 5 days viability was 10 times lower than that of 8325-4 (wild type) (Fig. 3). Survival was restored by the presence of pMAL43R, containing a full-length copy of the pheP gene on the complementation vector pMK4 (20) (Fig. 3). Plasmid pMAL43R was constructed by PCR amplifying pheP by using primer GAGAGGATCCTAGATGGGAGACTAAATATGG with CACAGAATTCGAATGGTAACATGGTAATAAT; the product was digested with BamHI/EcoRI, ligated to pMK4, and cloned directly in S. aureus RN4220.

FIG. 3.

Starvation survival capabilities of 8325-4 (wild-type) (▪), ST1 (•), ST16 (katA) (○), MJH600 (pheP) (▵), and MJH621 (pheP pMAL43R) (□) after prolonged aerobic incubation in glucose-limiting CDM (22). Samples were aseptically removed at the times indicated, and viability was assessed by dilution and counting colonies after 14 h of incubation on brain heart infusion agar. The experiment was repeated three times, giving very similar results; results from a representative experiment are shown.

MJH600 (pheP) grew in a manner similar to that of the parent strain, 8325-4, on nutrient and CDM under all conditions tested (data not shown). In contrast, a strong growth defect was observed on pig serum agar (23) for MJH600 (pheP) under microaerobic (5% CO₂, 87% N₂, and 8% O₂) (Fig. 4A) and anaerobic conditions (data not shown); normal growth was observed for aerobic growth on pig serum agar. Addition of micromolar concentrations of components from CDM identified that phenylalanine completely restored normal growth of MJH600 (pheP) (Fig. 4B), indicating that the permease mutant was likely to have a defect in phenylalanine uptake. The growth defect was also rescued by complementation using pMAL43R, indicating that only the pheP mutation was responsible for the observed phenotype (Fig. 4C).

FIG. 4.

Growth phenotype and complementation of MJH600 (pheP). Strains were incubated overnight in a microaerobic environment on pig serum agar without (A) or with (B) added phenylalanine (1 mM final concentration). Also shown is growth of MJH620 (8325-4 pMAL43R) and MJH621 (pheP pMAL43R) on pig serum agar (C). For complementation studies, pMAL43 and pMAL43R were electroporated into 8325-4 (wild-type) and MJH600 (pheP). The pheP gene was cloned directly in S. aureus due to toxicity of the gene in E. coli, which prevented complementation of transport mutants of E. coli.

The ability of S. aureus to scavenge amino acids from the host during infection would appear to be an important adaptation to growth in vivo (1, 4, 11, 16). S. aureus 8325-4 is not auxotrophic for phenylalanine (21). Biosynthesis must not be sufficient, however, to fulfill cellular growth requirements in an environment such as the abscess, thereby creating a requirement for phenylalanine uptake during infection.

The defective growth of MJH600 (pheP) on pig serum agar under microaerobic and anaerobic conditions might be due to active transport of amino acids by S. aureus requiring prior oxidation of the electron donor for transport (17, 18, 19). The absolute dependence of these respiration-linked transport systems on coupling to dehydrogenases has been shown to result in amino acid uptake having a dependence on the presence of separate electron donors under aerobic and anaerobic conditions (19). Under conditions of reduced oxygen availability, therefore, the transport of phenylalanine via an unidentified active transport system would result in growth limitation in the absence of the cognate electron donor for this system or would be dependent on a second transport system. We hypothesize that pheP is important for phenylalanine transport under conditions of reduced oxygen due to reduced function of a respiration-linked transport system. The attenuation of MJH600 (pheP) in the murine abscess model might therefore be indicative of reduced growth of this strain under conditions of reduced oxygen availability (2) where there might also be limiting free phenylalanine.