Staphylococcus aureus Does Not Synthesize Arginine from Proline under Physiological Conditions

ABSTRACT

The Gram-positive pathogen Staphylococcus aureus is the only bacterium known to synthesize arginine from proline via the arginine-proline interconversion pathway despite having genes for the well-conserved glutamate pathway. Since the proline-arginine interconversion pathway is repressed by CcpA-mediated carbon catabolite repression (CCR), CCR has been attributed to the arginine auxotrophy of S. aureus. Using ribose as a secondary carbon source, here, we demonstrate that S. aureus arginine auxotrophy is not due to CCR but due to the inadequate concentration of proline degradation product. Proline is degraded by proline dehydrogenase (PutA) into pyrroline-5-carboxylate (P5C). Although the PutA expression was fully induced by ribose, the P5C concentration remained insufficient to support arginine synthesis because P5C was constantly consumed by the P5C reductase ProC. When the P5C concentration was artificially increased by either PutA overexpression or proC deletion, S. aureus could synthesize arginine from proline regardless of carbon source. In contrast, when the P5C concentration was reduced by overexpression of proC, it inhibited the growth of the ccpA deletion mutant without arginine. Intriguingly, the ectopic expression of the glutamate pathway enzymes converted S. aureus into arginine prototroph. In an animal experiment, the arginine-proline interconversion pathway was not required for the survival of S. aureus. Based on these results, we concluded that S. aureus does not synthesize arginine from proline under physiological conditions. We also propose that arginine auxotrophy of S. aureus is not due to the CcpA-mediated CCR but due to the inactivity of the conserved glutamate pathway.

IMPORTANCE Staphylococcus aureus is a versatile Gram-positive human pathogen infecting various human organs. The bacterium's versatility is partly due to efficient metabolic regulation via the carbon catabolite repression system (CCR). S. aureus is known to interconvert proline and arginine, and CCR represses the synthesis of both amino acids. However, when CCR is released by a nonpreferred carbon source, S. aureus can synthesize proline but not arginine. In this study, we show that, in S. aureus, the intracellular concentration of pyrroline-5-carboxylate (P5C), the degradation product of proline and the substrate of proline synthesis, is too low to synthesize arginine from proline. These results call into question the notion that S. aureus synthesizes arginine from proline.

INTRODUCTION

Staphylococcus aureus is a Gram-positive pathogenic bacterium causing a wide range of infections in humans (1). S. aureus genomes contain genes for the synthesis of all 20 amino acids (2). Intriguingly, however, the bacterium is phenotypically auxotrophic to several amino acids, including arginine and proline. The arginine and proline auxotrophy of S. aureus is attributed to the CcpA-mediated carbon catabolite repression (CCR) (3, 4).

CcpA is a transcription factor of the LacI/GalI family and plays a role diverse from those of CCR, nitrogen metabolism, and virulence (5, 6). In the CCR, when a preferred carbon source such as glucose is present, CcpA represses the transcription of the genes necessary for utilizing nonpreferred carbon sources, including ribose, arabinose, and sorbitol, allowing the bacterium to optimize energy use by avoiding unnecessary transcription (7). The CcpA-binding site is known as a catabolite response element (CRE) (8), and it consists of the following consensus sequence: WWTGNAARCGNWWNCAWW (W = A or T; R = A or G; N = any). Typically, the CRE binding of CcpA requires the corepressor protein HPr (histidine protein) (9).

In S. aureus, the synthesis pathways for arginine and proline are interconnected; arginine is synthesized from proline, while proline is synthesized from arginine (Fig. 1A) (3, 4). In fact, S. aureus is the only bacterium known to synthesize arginine from proline (4). In the arginine-proline interconversion pathway, CcpA represses almost all steps except for the final proline synthesis by ProC (3, 4) (Fig. 1A). Intriguingly, in addition to the arginine-proline interconversion pathway, S. aureus also contains the well-conserved glutamate pathway for arginine synthesis. This pathway converges to the arginine-proline interconversion pathway at ornithine (Fig. 1A). Intriguingly, the argDCJB operon, encoding the unique steps of the glutamate pathway, is not transcribed in S. aureus even under 44 different growth conditions (4, 10).

FIG 1.

S. aureus remains an arginine auxotroph regardless of carbon sources. (A) Metabolic pathways for arginine and proline in S. aureus. The steps repressed by CcpA (green) and arginine repressors (yellow and red) are indicated. (B) Carbon source effect on arginine and proline auxotrophy of S. aureus. S. aureus wild type (WT) and the ccpA mutant (ΔccpA) were grown at 37°C overnight in CDM containing glucose or ribose or no carbon source. Bacterial growth was measured by optical density at 600 nm (OD₆₀₀). −, absence; +, presence; R, arginine; P, proline.

At the center of the intertwined metabolic pathways is pyrroline-5-carboxylate (P5C). P5C is a degradation product of ornithine and proline and can be used for the synthesis of three amino acids: arginine, proline, and glutamate (Fig. 1A). In particular, the P5C-glutamate conversion allows S. aureus to use arginine and proline as alternative carbon sources in the absence of glucose because glutamate can be further converted into 2-oxoglutarate (i.e., α-ketoglutarate), an intermediate of the tricarboxylic acid (TCA) cycle, by GudB (Fig. 1A) (11). Not surprisingly, these P5C-to-2-oxoglutarate conversion steps (i.e., the production of P5C dehydrogenase [RocA] and [GudB]) are also repressed by CcpA (Fig. 1A).

Since the arginine-proline interconversion pathway is repressed by CcpA, S. aureus is expected to become a prototroph for both arginine and proline when ccpA is knocked out. Indeed, the ccpA knockout mutant can grow in the absence of arginine or proline (3, 4). Surprisingly, however, when the CcpA-mediated CCR is abolished by using secondary carbon sources such as ribose as a sole carbon source, although S. aureus became a prototroph for proline (3), it still remained an arginine auxotroph (4). Also, by an experiment with ¹³C-labeled amino acids, Halsey et al. clearly demonstrated that proline is not converted into arginine even in the absence of glucose (11). Therefore, it is questionable whether S. aureus synthesizes arginine from proline under physiological conditions (i.e., without gene knockout).

In this study, using ribose as a secondary carbon source, we wanted to understand the molecular mechanism behind the arginine auxotrophy in the absence of CCR and investigate whether the argDCJB operon is functional when ectopically expressed. We found that the insufficient concentration of proline degradation product P5C, not CCR, is responsible for the arginine auxotrophy of S. aureus, and the argDCJB operon is functional once transcribed.

RESULTS

The CCR of arginine synthesis is not affected by the carbon source.

Previously, in a chemically defined medium with ribose as a sole carbon source (CDMR), S. aureus became a proline prototroph but remained an arginine auxotroph (4). To confirm the results, we grew wild-type (WT) S. aureus strain Newman in CDM containing glucose (CDMG), ribose (CDMR), or no carbon source in the presence and absence of proline or arginine. As a control, the ccpA mutant was used. As shown in Fig. 1B, ΔccpA mutant grew under all conditions, whereas the WT strain failed to grow in CDMG in the absence of either proline or arginine, confirming that the proline and arginine synthesis pathways are under repression by CcpA. In the presence of ribose or without a carbon source, the WT strain grew without proline (−P in Fig. 1B), confirming that the proline auxotrophy of S. aureus is due to CCR. However, under the same conditions, the WT strain failed to grow without arginine (−R in Fig. 1B). In a previous study, the S. aureus USA300 JE2 strain could grow without arginine when either sorbitol or arabinose was used as a sole carbon source (4). However, in our hands, both Newman and USA300 JE2 strains failed to grow under these conditions (−R in Fig. S1 in the supplemental material). Based on these results, we concluded that S. aureus could not synthesize arginine regardless of carbon source.

Arginine repressors are not involved in the arginine auxotrophic phenotype of S. aureus.

In many bacteria, arginine repressors such as AhrC and ArgR repress arginine synthesis (12–16). S. aureus contains both AhrC (NWMN_1426) and ArgR (NWMN_2535). To examine whether these arginine repressors are responsible for the S. aureus arginine auxotrophic phenotype in CDMR, we grew the WT strain and the mutants of the arginine repressors in CDMG or CDMR with or without arginine. As a control for CCR, bacterial growth was also examined in the presence or absence of proline. In CDMR, all strains grew in the absence of proline; however, they failed to grow without arginine (−R in Fig. 2A), indicating that the arginine repressors are not responsible for the arginine auxotrophic phenotype of S. aureus. To identify the role of arginine repressors in arginine synthesis, we measured the transcript levels of the genes involved in arginine synthesis by qRT-PCR for WT and arginine repressor mutants. As shown, the transcription of arcB and argG was profoundly affected by the arginine repressor(s) (Fig. 2B). The transcription of arcB was dramatically increased only in the double mutant, indicating that the gene is redundantly repressed by AhrC and ArgR. Although the transcription of argG was increased in the ahrC deletion mutant, it was more increased in the argR mutant, suggesting that argG is mainly repressed by ArgR. Transcription of all other genes was not largely affected by the arginine repressor mutations. Based on these results, we concluded that the arginine repressors control the conversion of ornithine to arginine but are not responsible for the arginine auxotrophy.

FIG 2.

Arginine repressors do not control the arginine auxotrophy of S. aureus. (A) The effect of arginine repressor mutations on arginine and proline auxotrophy of S. aureus. S. aureus Newman wild type (WT) and the arginine repressor mutants were grown at 37°C overnight in CDM containing glucose or ribose in the presence or absence of arginine and proline. +, presence; −, absence; R, arginine; P, proline. Data represent means ± standard errors of the means (SEM) from three independent experiments. (B) The effect of arginine repressor mutations on the transcription of the genes involved in arginine metabolism. The test strains were grown in CDMG at 37°C until the exponential growth phase. The transcript levels of the test genes indicated under the graph were measured by qRT-PCR, where 16S rRNA was used for internal control. The results were analyzed by the comparative C_T method. The experiments were performed in triplicate and repeated independently three times with similar results. Statistical significance was assessed by unpaired, two-tailed Student's t test. *, P < 0.05.

In S. aureus, the proline degradation step is responsible for the arginine auxotrophy.

To determine which arginine synthesis steps are responsible for the arginine auxotrophy, we grew S. aureus Newman WT strain in CDMG, in which arginine was replaced by the arginine synthesis intermediate ornithine or citrulline. The WT strain grew in both growth media (Fig. 3A), indicating that staphylococcal arginine synthesis is impaired before the synthesis of ornithine (i.e., RocD or PutA).

FIG 3.

Proline degradation step is responsible for the arginine auxotrophy of S. aureus. (A) Effect of arginine synthesis intermediates on staphylococcal growth without arginine. The Newman wild-type strain was grown at 37°C overnight in CDMG-R supplemented with ornithine (Orn) or citrulline (Cit). For a control, the strain was also grown in CDMG (i.e., CDM containing glucose and all amino acids). −, no addition. Data represent means ± SEM from three independent experiments. (B) Effect of putA overexpression on arginine auxotrophy of S. aureus. S. aureus Newman containing the putA expression plasmid pYJ335-putA (pYJ-putA) was grown in CDMG-R in the absence (−) or presence of anhydrotetracycline (ATc; 100 ng/mL). The Newman strain without the plasmid was used as a negative control. Data represent means ± SEM from three independent experiments.

The transcriptions of rocD and putA are both repressed by CcpA (3, 11). Of the two, RocD is shared by both arginine and proline synthesis pathways (Fig. 1A). S. aureus also could grow without proline when ribose was used as a sole carbon source (Fig. 1B), suggesting that the CCR imposed on rocD is released by ribose. Therefore, it is likely that the arginine auxotrophic phenotype in CDMR is due to the blockage at the proline degradation step by PutA. To test this hypothesis, we fused the putA gene to the anhydrotetracycline (ATc)-inducible promoter in a multicopy plasmid, pYJ335 (17), and the resulting plasmid, pYJ335-putA, was mobilized into the WT strain. In the absence of ATc, the strain barely grew; however, the addition of ATc significantly increased the growth of the bacterium (Fig. 3B), indicating that the insufficient production of PutA is responsible for the S. aureus arginine auxotrophy in CDMR.

In the wild-type strain, ribose appears to induce the expression of putA fully.

In their previous study, Nuxoll et al. suggested that S. aureus arginine auxotrophy in CDMR is due to the inability of ribose to derepress CcpA (4). Indeed, the PutA production also seemed insufficient to support arginine synthesis in CDMR (Fig. 3B). To examine the possibility of the inadequate induction of putA in CDMR, we measured the putA transcripts levels in CDMG and CDMR for WT and the ccpA deletion mutant. The rocD transcription was used as an indicator for CCR regulation. In the WT strain, transcriptions of both putA and rocD were significantly increased in CDMR compared with CDMG (Fig. 4A), which was recapitulated in the promoter-lacZ fusion assay (Fig. 4B). These results demonstrate that, contrary to the previous postulation, putA transcription is rather normally induced by ribose. In the ccpA mutant, both putA and rocD showed significant transcription in CDMG (Fig. 4A), further confirming that both genes are repressed by CcpA. Importantly, the putA transcript level in CDMR was slightly higher than that in CDMG of ccpA mutant (Fig. 4A). Considering that the ccpA mutant grows in CDMG without arginine, these results indicate that inadequate induction of putA is not responsible for the S. aureus arginine auxotrophy. This conjecture was further supported by Western blot analysis for PutA. The PutA level of WT in CDMR was slightly higher than that of the ccpA mutant in CDMG (1.15 versus 0.9 in Fig. 4C). To further investigate the putA induction in CDMR, we generated a single-copy putA complementation plasmid, pROKA-putA. The CRE sequence or the entire promoter of putA then was replaced with those of rocD, whose transcription is fully derepressed by ribose (3). The original and variant plasmids were inserted into the chromosomes of WT, ΔputA, and ΔccpA strains, and the resulting strains were grown in CDMR with or without arginine (CDMR-R). As shown, regardless of the complementary plasmids, the WT strains failed to grow in the absence of arginine (Fig. 4D). Based on these results, we concluded that putA is fully induced in CDMR, and inadequate induction of putA is not responsible for S. aureus arginine auxotrophy.

FIG 4.

Ribose fully induces the expression of PutA. (A) Ribose-mediated induction of putA and rocD transcription measured by qRT-PCR analysis. The wild type (WT) and the ccpA mutant (ΔccpA) were grown in CDMG or CDMR until the exponential growth phase, and total RNA was purified. 16S rRNA was used as an internal control, and the results were analyzed by the comparative C_T method and normalized to the WT value in CDMG. (B) The ribose-mediated induction of the putA transcription measured by the promoter-lacZ fusion assay. Newman WT strain containing the putA promoter-lacZ or the rocD promoter-lacZ plasmid was grown in CDMG or CDMR until the exponential growth phase. Cells were collected and lysed, and then LacZ activity was measured. All assays were performed in triplicate and repeated independently three times. MU, Miller units. (C) Carbon source effect on the expression of the PutA protein. WT and ΔccpA strain were grown in CDMG or CDMR and harvested at the exponential growth phase. The expression of PutA was measured by Western blotting with a PutA antibody. A putA transposon mutant strain (ΔputA) was used for negative control. For loading control, SaeR, the response regulator of the SaeRS two-component system, was detected. The ratio of PutA and SaeR protein band intensity (PutA/SaeR) is shown under each column. Glc, CDMG; Rib, CDMR. (D) Effect of the CRE sequence or the promoter replacement on arginine auxotrophy of S. aureus. The CRE sequence or the entire promoter sequence of the putA promoter in pROKA-putA was replaced with that of the rocD promoter. The resulting plasmids were inserted into the wild-type (WT), the putA deletion mutant (ΔputA), and the ccpA deletion mutant (ΔccpA). The test strains were grown at 37°C overnight in CDMR and CDMR-R. −, no change; CRE, the CRE sequence replacement; promoter, the promoter replacement. Data represent means ± SEM from three independent experiments. The statistical significance was assessed by unpaired, two-tailed Student's t test. Unless indicated otherwise, the assessment was made for CDMG versus CDMR. **, P < 0.005; ***, P < 0.0005; ****, P < 0.0001.

Uncharacterized repeat sequences are involved in the ribose-mediated activation of the putA promoter.

In the previous experiment, we observed that both putA and rocD transcript levels in ΔccpA strain were further increased in CDMR compared with CDMG (see ΔccpA strain in Fig. 4A), indicating that transcription of those genes was activated by ribose in a CCR-independent manner. The putA promoter and its upstream region contain three types of cis-elements: two ARG boxes (i.e., the binding sequence for arginine repressors), uncharacterized two identical repeat sequences (RS1 and RS2), and a CRE sequence (Fig. 5A). One of the ARG boxes overlaps the −10 sequence of the putative promoter for NWMN_1657, encoding a putative lysophospholipase, whereas the CRE sequence is right next to one of the repeat sequences and overlaps the −35 sequence of the putA promoter. To investigate which of those cis-elements is responsible for the CCR-independent regulation, we generated various deletions of the cis-elements in the putA promoter-lacZ fusion plasmid (Fig. 5B). Those plasmids were mobilized into the Newman WT strain, and the LacZ activity was measured in CDMG or CDMR. The deletion of the ARG boxes did not significantly affect the LacZ activity in either growth medium, showing that the ARG boxes do not affect the putA promoter activity. On the other hand, the deletion of the uncharacterized repeat sequences additively lowered the LacZ activity in CDMR (Δ4 and Δ5 in Fig. 5B and C). As expected, when half of the CRE sequence was deleted without affecting the −35 sequence, the LacZ activity was increased in CDMG but not in CDMR (Δ6 in Fig. 5B and C). The putA core promoter (i.e., the promoter in the Δ6 mutant) showed only slightly higher LacZ activity in CDMR than in CDMG. Based on these results, we concluded that the uncharacterized repeat sequences possibly are responsible for the ribose-mediated activation of the putA promoter.

FIG 5.

Uncharacterized repeat sequences are involved in the ribose-mediated activation of the putA promoter. (A) The DNA sequence of the putA promoter region. ARG box, the putative binding site of arginine repressors; RS, repeat sequence. The promoter sequences (−10 and −35) are also indicated. The putA promoter sequence was verified by mutagenesis analysis, whereas the promoter sequence for NWMN-1657 was predicted. (B) The cis-element deletions in pOS1-PputA-lacZ. The remaining putA promoter sequences are indicated as a thick line for each deletion mutant. (C) The effect of cis-element deletion on the putA promoter activity. S. aureus Newman carrying one of the pOS1-PputA-lacZ plasmids was grown at 37°C in CDMG or CDMR, and LacZ activity was measured at the exponential growth phase. WT, no deletion in pOS1-PputA-lacZ; Δ1 to Δ6, cis-element deletion mutants of pOS1-PputA-lacZ shown in panel B. Data represent means ± SEM from three independent experiments. Statistical significance was assessed by unpaired, two-tailed Student's t test. ***, P = 0.0001.

The intracellular concentration of P5C is the determining factor for arginine synthesis.

So far, the experimental results indicate that although the transcription of putA is fully induced by ribose, S. aureus remains an arginine auxotroph in CDMR. To find the reason for the arginine auxotrophy, we decided to identify arginine prototrophic mutants. The mixture of 1,736 transposon mutants of S. aureus Newman (i.e., Phoenix library) was spread on the CDMR agar plate without arginine (CDMR-R agar) and incubated for 2 days (18). The colonies on the plate were then inoculated in CDMR-R broth to confirm their arginine prototrophy. Inverse PCR analysis of those mutants found that they have a transposon insertion in proC, the gene encoding pyrroline-5-carboxylate (P5C) reductase. This enzyme converts P5C into proline, the last step of the proline synthesis and the reverse reaction of proline degradation (Fig. 1A). The proC mutant grew in CDMR-R and CDMG-R, indicating that arginine synthesis in the strain is independent of CCR (Fig. 6A). When the transposon mutation was transduced into strain USA300, a predominant MRSA strain in the United States, the resulting mutant could also grow without arginine, indicating that the proC mutation converts S. aureus into arginine prototroph regardless of the strain background (Fig. S2).

FIG 6.

Pyrroline-5-carboxylate (P5C) intracellular concentration is the determining factor for arginine synthesis. (A) Effect of the proC disruption on the arginine auxotrophy of S. aureus. S. aureus Newman wild type (WT) and proC deletion mutant (ΔproC) were grown at 37°C overnight in CDMG or CDMR with or without arginine. + R, presence of arginine; − R, no arginine. Data represent means ± SEM from three independent experiments. (B) The intracellular concentration of pyrroline-5-carboxylate (P5C) in WT and ΔproC cells. Cells were grown at 37°C overnight in CDMG and CDMR, and then the concentration of P5C was measured as described in Materials and Methods. AU, arbitrary units. (C) Effect of proC overexpression on the growth of the ccpA deletion mutant (ΔccpA) in CDMR-R (i.e., CDMR without arginine). The ΔccpA strain carrying either pYJ335 (pYJ) or the proC expression plasmid pYJ335-proC (pYJ-proC) was grown at 37°C for 18 h with (+ATc; 100 ng/mL) or without anhydrotetracycline. For clarity, the error bars are not shown. The growth curve of ΔccpA mutant in CDMG-R is provided in Fig. S4. (D) Effect of proC overexpression on the intracellular P5C concentration in ΔccpA strain. The ΔccpA strain carrying either pYJ335 or the proC expression plasmid pYJ335-proC was grown at 37°C for 18 h. ATc, anhydrotetracycline; −, absence; +, presence. Data represent the means ± SEM from three independent experiments. The statistical significance was assessed by unpaired, two-tailed Student's t test. Unless indicated otherwise, the assessment was made for pYJ335 versus pYJ-proC. *, P < 0.05; ***, P < 0.0005; ns, not significant.

A straightforward explanation for the arginine prototrophy of the proC mutant is that the disruption of proC increases the intracellular concentration of P5C, mimicking the overexpression of PutA (Fig. 3B). To examine this hypothesis, we measured the intracellular levels of P5C in WT and △proC strains using the ninhydrin derivatization method (19, 20). Indeed, the concentration of P5C was significantly increased in the proC transposon mutant regardless of the carbon source (Fig. 6B). When the transcription of putA was induced from the multicopy plasmid pYJ335-putA, which converted S. aureus into arginine prototroph (Fig. 3B), the intracellular concentration of P5C was also increased (Fig. S3).

To further test whether the intracellular concentration of P5C is a critical determinant for the arginine prototrophy, we generated pYJ335-proC, where proC is transcribed from an ATc-inducible promoter. The plasmid was mobilized into ΔccpA strain, which can grow without arginine, and the resulting strain was grown in CDMG-R and CDMR-R. As shown, the inducer alone did not significantly affect the growth of ΔccpA strain (pYJ versus pYJ+ATc in Fig. 6C and Fig. S4). Intriguingly, the growth of the ΔccpA mutant carrying pYJ335-proC was severely inhibited even without the addition of ATc (pYJ-proC in Fig. 6C and Fig. S4). qRT-PCR analysis showed that the ATc-inducible promoter in pYJ335 is leaky and produces a significant level of proC transcripts without ATc (proC in Fig. S5). The addition of ATc further increased the proC transcript level by approximately twofold. Importantly, under both conditions, the P5C concentration was reduced by 3- to 3.5-fold (Fig. 6D and Fig. S4B), resulting in severe growth inhibition of ΔccpA (Fig. 6C and Fig. S4A). As expected, the transcription of putA was increased in CDMR-R compared with that in CDMG-R (putA in Fig. S5); however, this did not block the decrease of P5C concentration (Fig. 6D). Based on these results, we concluded that the intracellular concentration of P5C is the determining factor for arginine prototrophy.

The argDCJB operon is functional if transcribed.

The argDCJB operon encodes enzymes for arginine synthesis from glutamate (Fig. 1A). However, the operon is known not to be transcribed (4, 10). To examine whether the operon is functional if transcribed, we fused the entire argDCJB operon to the ATc-inducible promoter in pYJ335 and mobilized the resulting plasmid into WT S. aureus Newman. In CDMG-R, the strain grew when the inducer ATc was added (Fig. 7), demonstrating that, once transcribed, those genes can produce functional enzymes that confer arginine prototrophy to S. aureus.

FIG 7.

argDCJB operon is functional once transcribed. S. aureus Newman containing pYJ335-argDCJB was grown in CDMG-R without (−) or with (+) anhydrotetracycline (100 ng/mL). For negative control, wild-type (WT) Newman strain without the plasmid was used. pargDCJB, pYJ335-argDCJB. Data represent means ± SEM from three independent experiments.

The arginine/proline role as an alternative carbon source does not contribute to the in vivo survival of S. aureus.

In S. aureus, arginine and proline can be fed into the TCA cycle through glutamate in the absence of glucose (Fig. 1A) (11). Next, we decided to examine whether the role of these amino acids as an alternative carbon source is critical for staphylococcal virulence and survival during infection. To block the synthesis of P5C, we disrupted both rocD and putA. We also deleted rocA to stop the flow from P5C to glutamate. These two mutant strains were administered to mice via a retro-orbital route with the WT strain. At 3 days postinfection, the bacterial loads in the kidney, liver, and heart were measured. As shown in Fig. 8, none of the mutations reduced the bacterial loads in the organs. On the contrary, the number of CFU of the ΔrocD::putA double mutant was slightly but significantly higher than that of the WT in the kidney. These results demonstrate that those amino acids’ role as an alternative carbon source is, at the least, not essential for the virulence and survival of S. aureus.

FIG 8.

Arginine-proline interconversion pathway does not contribute to the virulence and survival of S. aureus. The test strains (10⁷ CFU) were injected into 6 to 12 C57BL/6 mice via the retro-orbital route. At 3 days postinfection, the bacterial loads in the organs were measured by CFU counting. The CFU in the kidney and liver were counted in 12 mice, whereas the heart CFU counting was in six mice. Statistical significance was assessed by unpaired, two-tailed Student's t test. **, P < 0.005.

DISCUSSION

The genome of S. aureus harbors genes for the synthesis of all 20 amino acids (21). However, due to CCR by CcpA, S. aureus is phenotypically auxotrophic to proline (3). S. aureus is also known to be auxotrophic to arginine due to the CcpA-mediated CCR (4). Indeed, the ccpA mutant does not need arginine for growth (Fig. 1B) (4, 11). However, the disruption of CCR with nonpreferred carbon sources failed to convert S. aureus into arginine prototroph (Fig. 1B) (11). In this study, we demonstrated that the intracellular concentration of P5C, not CCR, is the determining factor for arginine auxotrophy, and the reduction of P5C concentration by the P5C reductase ProC is responsible for the S. aureus arginine auxotrophy. Therefore, it is unlikely that S. aureus can synthesize arginine from proline under physiological conditions without artificial gene knockouts.

To explain S. aureus arginine auxotrophy in CDMR, Nuxoll et al. suggested that ribose is unable to derepress CcpA (4). Initially, we also hypothesized that the inadequate induction of the putA gene by ribose is responsible for the arginine auxotrophy in CDMR. However, the transcription of putA seems to be fully induced by ribose. The putA transcription was increased eightfold in CDMR compared with CDMG, while the rocD transcription was increased 3- to 6-fold (Fig. 4A and B). The transcription from the putA promoter also was not significantly affected by the deletion of the CRE sequence, the CcpA binding site, when the bacterium was grown in CDMR (Δ5 versus Δ6 in Fig. 5C). Therefore, it is likely that ribose fully derepresses CcpA, and CCR is not responsible for the arginine auxotrophy.

Deletion mutagenesis showed that the putA promoter is regulated differently depending on the carbon source. In CDMG, the activity of putA promoter was solely controlled by CcpA (Δ5 versus Δ6 in Fig. 5C), whereas in CDMR it was regulated by two repeat sequences (RS1 and RS2 in Fig. 5A and Δ3 to Δ5 in Fig. 5C). Those repeat sequences might be binding sites for a hitherto unknown transcription factor. Since the repeat sequences are right next to CRE, the DNA binding events of the putative transcription factor and CcpA are likely mutually exclusive. Indeed, in CDMG, the deletion of the repeat sequence did not affect the activity of the putA promoter, whereas, in CDMR, the deletion of CRE showed little effect on the putA promoter activity (Fig. 5C). It should also be noted that the distance between RS2 and the −35 sequence of the putA promoter is 10 bp, the number of nucleotides in one helix turn. Therefore, once bound to DNA, the putative transcription activator will be on the same side as RNA polymerase, a prerequisite for direct interaction. Intriguingly, each repeat sequence contributed to the putA promoter activity in an additive manner (Δ4 and Δ5 in Fig. 5C). Therefore, it is possible that the putative transcription factor binds to the sequence as a monomer. However, these are just speculations at this stage, and more studies are required to confirm the existence of such a transcription factor.

In S. aureus, the P5C concentration appears to be the key determinant for arginine synthesis from proline. The increased P5C concentration in ΔproC mutant allowed the strain to grow without arginine regardless of the carbon sources, whereas the decreased P5C concentration inhibited the arginine prototrophy of the ccpA mutant (Fig. 6; see also Fig. S4 in the supplemental material). Since RocA also uses P5C as a substrate to synthesize glutamate, the ΔrocA mutant is expected to have an increased concentration of P5C. Indeed, in CDMR, the P5C concentration in ΔrocA mutant was higher than that in WT but lower than that in either ΔccpA or ΔproC mutant (Fig. S6A). However, the ΔrocA mutant barely grew in CDMR without arginine (Fig. S5B), suggesting that the P5C concentration in ΔrocA mutant is not sufficient to support arginine synthesis fully (Fig. 1A). Intriguingly, when S. aureus was supplemented with ¹³C-labeled proline in CDM without carbon source, it produced ¹³C-labeled ornithine but not citrulline (11), indicating that, even at a concentration insufficient to support arginine synthesis, P5C still can be converted to ornithine by RocD to a certain extent. Nonetheless, the concentration of the produced ornithine is likely too low to be further processed by ArcB/ArgF into citrulline (Fig. 1A).

Because of the arginine prototrophy of the ccpA mutant, S. aureus has been known to synthesize arginine from proline via CCR control (4). However, when the CcpA-mediated CCR is abolished by secondary carbon sources, S. aureus still remains an arginine auxotroph (Fig. 1B and Fig. S1). Like the ccpA deletion, the knockout mutation of proC also converts S. aureus into arginine prototroph (Fig. 6A and B). Therefore, the arginine prototrophic phenotype of the ccpA mutant is more an artifact of the gene disruption than a regulatory outcome of the derepression of CCR. In CDMR, the putA promoter activity was higher in the ccpA mutant than in the WT (Fig. 4A), with a concomitant increase of P5C (Fig. S6). Therefore, we hypothesize that CcpA can repress, directly or indirectly, the transcription of putA in a CCR-independent manner. Alternatively, although unlikely, CcpA might repress another gene(s), which can contribute to the synthesis of P5C, in a CCR-independent manner. Both possibilities are currently being investigated in our laboratory. Regardless of the molecular mechanism of the arginine prototrophy of the ccpA mutant, the proline catabolite pathway seems to function to provide an alternative carbon source in the absence of glucose, not to synthesize arginine, as Halsey et al. clearly showed (11).

Commonly, bacteria synthesize arginine from glutamate using the argDCJB operon. Although the S. aureus genome contains the argDCJB operon, the operon appears to be dormant. Mader et al. reported that the argDCJB operon was not transcribed in S. aureus HG001, a derivative of NCTC 8325, under 44 different growth conditions (10). However, when the operon was transcribed from an ATc-inducible promoter, S. aureus became an arginine prototroph (Fig. 7), indicating that the genes in the operon are intact. Without significant contribution to bacterial growth and survival, it is unlikely that the entire operon survives natural selection and remains intact. Therefore, it is reasonable to predict that S. aureus requires the argDCJB operon to synthesize arginine under certain environmental conditions. In their study, Mader et al. identified 312 genes never expressed under 44 different growth conditions (10). Intriguingly, those genes include arcB, whose transcription is repressed by ArgR/AhrC (Fig. 2B), and vraDE, whose transcription is induced by bacitracin and nisin via the BraSR two-component system (22). Therefore, those 44 different growth conditions are not comprehensive and did not include the growth condition where the transcription of the argDCJB operon is induced. Considering that, without mutation in ccpA or proC, the P5C concentration is too low for arginine synthesis (Fig. 6B), ArgDCJB likely constitutes the genuine pathway for arginine synthesis in S. aureus.

Finally, we propose the following model for the arginine-proline interconversion pathway (Fig. 9). In CDMG, almost all steps are repressed either by the CcpA-mediated CCR or by the arginine repressors. Although the conversion of P5C to proline by ProC is not repressed, proline is not synthesized due to the low concentration of P5C. Therefore, S. aureus is auxotrophic to arginine and proline under this condition. In CDMR, CCR is abolished, but due to the presence of arginine, the arginine repressors are functional and repress the conversion of ornithine to arginine. Under this condition, both arginine and proline are converted into P5C and fuel the TCA cycle through glutamate synthesis by RocA. Due to the constitutive activity of ProC, P5C is also converted to proline, leading to proline prototrophy. In CDMR-R (i.e., CDMR without arginine), neither CcpA nor arginine repressor is functional, and all steps are open. Although proline can be converted to P5C by PutA and might be further processed to glutamate, due to the constant consumption of P5C by ProC, the intracellular P5C concentration is too low to produce a sufficient amount of ornithine for arginine synthesis, and S. aureus fails to grow. In summary, S. aureus cannot synthesize arginine from proline under any of these three conditions.

FIG 9.

Models for the regulation of arginine and proline metabolism by carbon catabolite repression and arginine repressors. Red arrow, repressed; blue arrow, induced; black arrow, constitutive. The dashed lines indicate nonfunctional steps. The gradient of gray colors represents relative concentrations.

MATERIALS AND METHODS

Ethics statement.

All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (23). The animal protocol was approved by IUSM-NW IACUC (protocol no. NW-48). Every effort was made to minimize the suffering of the animals.

Bacterial strains and growth conditions.

Bacterial strains used in this study are listed in Table 1. For bacterial culture in chemically defined medium (CDM) (24, 25), single colonies on tryptic soy agar (TSA) were inoculated in tryptic soy broth (TSB), and the culture was incubated at 37°C with shaking overnight. Cells were collected by centrifugation, washed with sterile phosphate-buffered saline (PBS), and suspended in sterile water. The cell suspension (1% of the culture volume) was inoculated into CDM and incubated at 37°C with shaking. For transduction of transposon mutation, heart infusion broth (HIB; BD Difco) supplemented with 5 mM CaCl₂ (Sigma) was used. Escherichia coli was used for genetic manipulation of plasmids and grown in lysogeny broth (LB; BD Difco). When necessary, antibiotics were added to the cultures at the following concentrations: ampicillin, 100 μg/mL; chloramphenicol, 15 μg/mL; erythromycin, 10 μg/mL.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)	Source or reference
Strain
E. coli DH5α	Restriction deficient, plasmid free	NEB
S. aureus RN4220	Restriction deficient, prophage cured	32
S. aureus Newman	Clinical isolate	33
NMΔccpA	Deletion mutant of ccpA	3
NMΔproC	Deletion mutant of proC	3
NMΔrocA	Deletion mutant of rocA	28
NM::ahrC	Transposon mutant of ahrC	18
NM::argR	Transposon mutant of argR	18
NMΔargR	Deletion mutant of argR	This study
NMΔargR::ahrC	NM::ahrC with argR deletion	This study
NM::proC	Transposon mutant of proC	18
USA300-P23	USA300-0114 without plasmid 2 and 3	34
USA300::proC	Transposon mutant of proC	This study
Plasmid
pKOR1	Allele replacement plasmid	28
pKOR1-mcs	pKOR1 with multicloning site	26
pKOR1-ΔargR	pKOR1-mcs with argR-deletion module	This study
pROKA	Chromosome integration plasmid with the integration site of ϕNM2	This study
pROKA-putA	pROKA containing the putA gene and its promoter sequence	This study
pROKA-putA_CRE	pROKA-putA, where CRE is replaced by the rocD CRE sequence	This study
pROKA-putA_ProcD	pROKA-putA, where the promoter is replaced by the rocD promoter	This study
pYJ335	Multicopy plasmid with an anhydrotetracycline-inducible promoter	17
pYJ335-putA	putA expression plasmid	This study
pYJ335-proC	proC expression plasmid	This study
pYJ335-argCDJB	argCDJB expression plasmid	This study
pOS1-P2640-lacZ	Promoter-lacZ fusion plasmid used as a template for other lacZ reporter plasmids	35
pOS1-PputA-lacZ	lacZ reporter plasmid for the putA promoter (PputA)	This study
pOS1-ProcD-lacZ	lacZ reporter plasmid for the rocD promoter (ProcD)	This study
pOS1-PputA-lacZ-Δ1~Δ6	pOS1-PputA-lacZ with deletion of cis-elements in PputA	This study
pET28a-putA	PutA expression vector	This study

Generation of an ahrC argR double mutant.

To delete argR (NWMN_2535), we used the plasmid pKOR1-mcs (26). Flanking sequences (1 kb) of argR were PCR amplified with the primer sets P1742/P1743 and P1744/P1755 (Table 2), whereas pKOR1-mcs was amplified with the primer set P1740/P1741 (Table 2). The PCR products were purified and joined with ligation-independent cloning (27) and mobilized into E. coli DH5α by transformation. After confirmation of the DNA sequences of the cloned DNA by DNA sequencing, the resulting plasmid, pKOR1-ΔargR, was mobilized into S. aureus RN4220 by electroporation and subsequently into S. aureus Newman by transduction with ϕ85. Deletion of argR was carried out as described previously (28). The transposon in NM::ahrC then was transduced into the argR deletion mutant, resulting in NMΔargR::ahrC mutant (Table 1). The transduction was confirmed by PCR with the primer pair P3183/P3184 (Table 2).

TABLE 2.

Oligonucleotides used in this study

Name	Sequence (5′–3′)	Target
P1740	ATTGGAAGTGGATAACGGTACCGGTTCCGAGGCTC	pKOR1-mcs
P1741	ATTGGATTGGAAGTACGGGCCCGAGCTTAAGACTGG	pKOR1-mcs
P1742	ATGTCTATTTAAAACTTTTCTTCATATTGTAAACTCC	argR upstream
P1743	TACTTCCAATCCAATGCGTGAATCATATGACAACCAAGG	argR upstream
P1744	TATGAAGAAAAGTTTTAAATAGACATGTATCAAATGAATA	argR downstream
P1745	TTATCCACTTCCAATGAAACAAGTTCATCGCCGCCTTC	argR downstream
P3356	TTGATATCGAATTCGGAGGCCACAAGGGTTTGGAGGAGTAGCAA	putA
P3357	TTATTATCCGGAGGTGTAGCCCAGACTCCTAAATTATTGTTTATGTCTTGCT	putA
P4135	GATTACGCCGAGCAATAACAACGACGCCTT
P4136	CGGCCAGGTTTCAGAACTGCCAATATAACATTTAGGAGC
P4137	TATTGCTCGGCGTAATCATGGTCATAGCTGTT
P4138	TCTGAAACCTGGCCGTCGTTTTACAACGT
CRE-F	ACCCTTACATCAAAAACAATCAAAAATAATTATGTATAATTACACTA
CRE-R	TTTCAAGAATATTTTTTTCATGCTAAACTTATTGTAAACACAAGGG
PrD-F	GCTATGACCATGATTACGCCGAAGGGTTTTATCCATTACCACG
PrD-R	AAATTCTTTAATAGTGCCATCATATCTCCCCTTTTCATCA
Vec-F	GGTAATGGATAAAACCCTTCGGCGTAATCATGGTCATAG
Vec-R	TGATGAAAAGGGGAGATATGATGGCACTATTAAAGAATTTTTTTATCG
P3358	TACTCCTCCAAACCCTTGTGGCCTCCGAATTCGATATCAAGCT	pYJ335
P3359	ATAATTTAGGAGTCTGGGCTACACCTCCGGATAATAAATATATATAAACGTATATAGATT	pYJ335
P3418	TTGATATCGAATTCGGAGGCGTATAAATATAAAGGAGGGAGACCGATGAATTCAAT	argDCJB
P3419	TTATTATCCGGAGGTGTAGCGCGCAATGGATAATCGAACTGGAAT	argDCJB
P3416	GATTATCCATTGCGCGCTACACCTCCGGATAATAAATATATATAAACGTATATAGATTTC	pYJ335
P3417	TCCCTCCTTTATATTTATACGCCTCCGAATTCGATATCAAGCTTATT	pYJ335
P5173	AAATAAGCTTGATCAAAAACTAAAGGGATGTGACGTTAATG	proC
P5174	CATTATCTCATATTATTGGTCTTCTATATTAGAAAGTTCAATACTACGGTCG	proC
P5175	CTTTAGTTTTTGATCAAGCTTATTTTAATTATACTCTATCAATGATAGAGTGTCAAT	pYJ335
P5176	AGAAGACCAATAATATGAGATAATGCCGACTGTACTTTTTACAGT	pYJ335
P3348	AGCCTTAAAGACGATCCGGGCGACGCCTTTCGCATCATTTTCA	putA promoter
P3349	GTTCACCACCTTTTCCCTATGTGCCATTGCTACTCCTCCAAAC	putA promoter
P3350	TGGAGGAGTAGCAATGGCACATAGGGAAAAGGTGGTGAACTACT	pOS1-P2640-lacZ
P3351	AAATGATGCGAAAGGCGTCGCCCGGATCGTCTTTAAGGCT	pOS1-P2640-lacZ
P3352	AGCCTTAAAGACGATCCGGGGGCATCAATTTAATGAAGGGTTTTATCCATTACC	rocD promoter
P3353	GTTCACCACCTTTTCCCTATGTGCTCCGTAATGATTTGTTAACTCAATAATTTTTTCAG	rocD promoter
P3354	AACAAATCATTACGGAGCACATAGGGAAAAGGTGGTGAACTACT	pOS1-P2640-lacZ
P3355	CCCTTCATTAAATTGATGCCCCCGGATCGTCTTTAAGGCT	pOS1-P2640-lacZ
P3551	AAATGATGCGAAAGGCGTCGCCCGGATCGTCTTTAAGGCT	pOS1-PputA-lacZ
Δ1	TGTAGATTTCTATTTATAGTATTATTGTTGTCCAT	Cis-element deletion
Δ2	GTCCATATTATTATATATAAATGAAATCAACATCAATAATAGTG	Cis-element deletion
Δ3	CAACATCAATAATAGTGTAATTATACATAATTATTTTTGATTG	Cis-element deletion
Δ4	CATAATTATTTTTGATTGTTTTTGATGAAAACGC	Cis-element deletion
Δ5	TTGATGAAAACGCTTTCTCGAATATTTTTTTC	Cis-element deletion
Δ6	GCTTTCTCGAATATTTTTTTCATGCTAAAC	Cis-element deletion
P4105	CATGGTATATCTCCTTCTTAAAGTTAAAC	pET28a
P4106	CACCACCACCACCACCACTGAGATC	pET28a
P4910	CTTTAAGAAGGAGATATACCATGGCACTATTAAAGAATTTTTTTATCGGATTATC	putA
P4911	GTGGTGGTGGTGGTGTTTACGGCATAATTTTTTAATTGTACTTAAACCTAAC	putA
P142	AAACTGATTTTTAGTAAACAGTTGACGATATTC	Inverse PCR
P143	TTTATGGTACCATTTCATTTTCCTGCTTTTTC	Inverse PCR
16S-F	TATAGATGGATCCGCGCT	qRT-PCR
16S-R	CTTTCGCCCATTGCGGAAG	qRT-PCR
putA-F	GTTAGCAGAAGAAATCGCAAATG	qRT-PCR
putA-R	GTGGGCGTTCTGCTAATCTT	qRT-PCR
arcB-F	GGTGGCAACATATTAATCACAG	qRT-PCR
arcB-R	CCATTGATACCCAAACGTCAG	qRT-PCR
argG-F	GATCGAACGCATCCTCTGGT	qRT-PCR
argG-R	GGGCATGGAGTCGTGAAGAA	qRT-PCR
argD-F	TGGGAGCAAGTCGTTCCAGA	qRT-PCR
argD-R	CAAGCGCTGCCGTCGATATT	qRT-PCR
argF-F	TTGCAGCGCATGATCAAGGT	qRT-PCR
argF-R	TTCCACACTGGTACGCCTGAA	qRT-PCR
rocF-F	CTAAATCTCTCATACCAATTAGTAC	qRT-PCR
rocF-R	TAAGGATCCGGGGGACGCTTATGACAAAGAC	qRT-PCR
rocD-F	TAAGGATCCGATATGATGACTAAATCTGAAAAAATT	qRT-PCR
rocD-R	ATCTCCAAAATCAACTTTTCTAAATC	qRT-PCR
proC-F	TGGCACAAGCTATATTTACAGG	qRT-PCR
proC-R	TTTGATGCGTGTTGCTAGAG	qRT-PCR
P3183	CGTCCCAATTTTTCTAAAGGATGG	Transduction
P3184	CACTCAAGCAACTGTTTCTCGTG	Transduction

Construction of gene expression plasmids.

pROKA is a chromosome integration plasmid containing p15A origin for the replication in E. coli and the integrase gene and phage attachment sequence (attP) of ϕNM2 in S. aureus Newman. The plasmid integrates into the intergenic region between rpmF and isdB (29).

To generate pROKA-putA, we PCR amplified the putA gene with primer pair P4135/P4136 (Table 2). pROKA was also PCR amplified with primer pair P4137/P4138 (Table 2). The PCR products were treated with DpnI (NEB) at 37°C for 30 min, purified with a PCR purification kit (Qiagen), and assembled by the Gibson method (30). The assembled plasmid was mobilized into E. coli DH5α by transformation.

To replace the CRE sequence in pROKA-putA with that of rocD, we PCR amplified pROKA-putA with primer pair CRE-F/CRE-R (Table 2). The PCR product was treated by DpnI, purified, phosphorylated with PNK (NEB), and subsequently self-ligated with T4 ligase at 16°C overnight. The resulting plasmid, pROKA-putA__CRE, was mobilized into E. coli DH5α (Table 1).

To replace the putA promoter sequence in pROKA-putA with that of rocD, we PCR amplified pROKA-putA with primer pair promoter Vec-F/Vec-R (Table 2). In contrast, the rocD promoter was PCR amplified with primer pair PrD-F/PrD-R (Table 2). PCR products were assembled by the Gibson method (30), resulting in pROKA-putA__ProcD (Table 1). The plasmid was mobilized into E. coli DH5α by transformation.

After verification of the correct sequence by DNA sequencing, the plasmids were electroporated into S. aureus RN4220 and subsequently transduced into S. aureus strain Newman with ϕ85.

To construct the putA expression plasmid pYJ335-putA, the putA gene, including 300 bp of upstream sequence, was PCR amplified with the primer pair P3356/P3357 (Table 2). The multicopy plasmid pYJ335 was also PCR amplified with the primer pair P3358/P3359 (17) (Table 2). To make the pYJ335-argDCJB plasmid, the entire argDCJB operon, including 100 bp of upstream sequence, was PCR amplified with the primer pair P3418/P3419 (Table 2). The vector pYJ335 was PCR amplified with the primer pair P3416/P3417 (17) (Table 2). pYJ335-proC was constructed similarly. The proC gene was PCR amplified with the primer pair P5173/P5174, whereas the vector pYJ335 was PCR amplified with primer pair P5175/P5176 (Table 2). The PCR products were treated with DpnI (NEB, R0176) at 37°C for 30 min, purified with a PCR purification kit (Qiagen), and assembled by the Gibson method (30). The assembled DNA was mobilized into E. coli DH5α by transformation. The insert DNAs of all plasmids were verified by DNA sequencing. Finally, the plasmids were electroporated into S. aureus RN4220 and subsequently transduced into S. aureus strain Newman with ϕ85. The transduction was confirmed by plasmid purification.

Construction of LacZ reporter plasmids.

The putA promoter (PputA) and the rocD promoter (ProcD) were PCR amplified with the primer pairs P3348/P3349 and P3352/P3353, respectively (Table 2). The plasmid vector sequence was PCR amplified from pOS1-P2640-lacZ with the primer pairs P3350/P3351 and P3354/P3355 (Table 2). The PCR products were assembled by the Gibson method (30). The assembled plasmids were used to transform E. coli and purified from the transformants, and the cloned DNA sequence was confirmed by DNA sequencing. To generate cis-element deletion mutants of PputA, we PCR amplified pOS1-PputA-lacZ plasmid with the reverse primer 3551 and one of six forward primers, Δ1 to Δ6 (Table 2). The PCR products were treated with DpnI at 37°C for 30 min and purified with a PCR purification kit (Qiagen). The purified DNA was phosphorylated with PNK (NEB), self-ligated with T4 ligase (NEB), and used to transform E. coli DH5a. The fidelity of insert DNA was verified by DNA sequencing for all resulting plasmids. The sequence-verified plasmids were electroporated into S. aureus RN4220 and subsequently transduced into S. aureus strain Newman with ϕ85. The transduction was verified by plasmid purification.

RNA extraction and qRT-PCR.

Cells were grown in CDMG and harvested by centrifugation from bacterial cultures at 3 h postinoculation for the mid-exponential growth phase and at 18 h postinoculation for stationary-growth-phase cells. Total RNA was isolated from the harvested cultures using an RNeasy mini prep kit (Qiagen). Genomic DNA was eliminated with Turbo DNA-free kit (Invitrogen). The cDNA was synthesized from 2 μg of total RNA with SuperiorScript III master mix (Enzynomics). cDNA reaction mixture (8 μL) was used for PCR amplification with SYBR green PCR reagent (Enzynomics) and the cognate primer pairs (Table 2). Reactions were performed in a MicroAmp optical 96-well reaction plate (Applied Biosystems) and monitored with a 7300 sequence detector (Applied Biosystems). DNA was amplified 40 cycles under the following conditions: 95°C for 10 s, 60°C for 15 s, and 72°C for 30 s. All qRT-PCR experiments were performed with 50 ng of cDNA in triplicate and repeated independently three times. 16S rRNA was used as an internal control, and the results were analyzed by the comparative threshold cycle (C_T) method (31).

β-Galactosidase assay.

The test strains were grown in CDM containing 15 μg/mL chloramphenicol at 37°C for 16 h, and the optical density at 600 nm (OD₆₀₀) was measured for the cultures. The bacterial cells were collected by centrifugation and suspended in Z-buffer (60 mM Na₂HPO₄, 40 mM NaH₂PO₄, 10 mM KCl, 1 mM MgSO₄, 50 mM 2-mercaptoethanol, pH 7.0). The suspension was treated with lysostaphin (50 μg/mL) at 37°C for 30 min. After the addition of 4 volumes of Z-buffer to the lysed cells, 500 μL of the cell lysate was mixed with 10 μL of O-nitrophenyl-beta-galactopyranoside (ONPG; 12 mg/mL) and incubated at room temperature for 15 min to 1 h. When yellow color developed, 500 μL of Na₂CO₃ (1 M) was added, and the samples were centrifuged; the optical density of the supernatant then was measured at 420 nm and 550 nm. The LacZ activity was determined by Miller units as U = 1,000 × [(OD₄₂₀) − (1.75 × OD₅₅₀)]/[(incubation time) × (volume) × (OD₆₀₀)]. This assay was performed in triplicate and independently repeated at least three times.

Isolation of proC transposon mutant.

The mix of Phoenix transposon mutant library was grown in TSB containing 10 μg/mL erythromycin at 37°C overnight. Cells were collected from 1 mL culture by centrifugation, washed twice with 1 mL sterile water, and suspended in 1 mL sterile water. The cell suspension (100 μL) was spread on a CDMR agar medium without arginine (CDMR-R) and incubated at 37°C for 2 days. The resulting colonies were examined for their growth in CDMR-R broth. The transposon insertion site was determined by inverse PCR with the primers P142 and P143 (Table 2), as described previously (18).

Quantification of P5C.

P5C was quantified by the ninhydrin derivatization method (19, 20). In the method, the test strains were grown in CDM broth at 37°C for 18 h. The bacterial cells were collected by centrifugation and suspended in 500 μL of TSM (50 mM Tris, 0.5 M sucrose, 10 mM MgCl₂, pH 7.5). The suspension was treated with lysostaphin (50 μg/mL) at 37°C for 15 min and centrifuged. The protoplast pellet was suspended in 1 mL of 50 mM Tris (pH 8.0) and mixed with 0.25 mL of perchloric acid (3.6 N) and 0.25 mL of ninhydrin (2%, wt/vol). The mixture was boiled for 15 min, and the insoluble ninhydrin derivatives were collected by centrifugation. The collected derivatives were dissolved in 0.5 mL of ethanol, and 0.5 mL of 100 mM Tris (pH 8.0) was added. After centrifugation, the supernatant was collected, and its optical density was measured at 620 nm. The sample’s optical density was normalized by the culture optical density at 600 nm. Finally, the P5C concentration was calculated with the molar extinction coefficient (1.96 × 10⁵) and normalized by OD₆₀₀ of the culture: P5C concentration = OD₆₂₀/[ε/(volume) × (OD₆₀₀)]. This assay was performed in triplicate and independently repeated at least three times.

Generation of PutA antibody.

To produce the PutA protein, we generated pET28a-putA. pET28a was amplified with primers P4105/P4106, and putA gene was amplified with primers 4910/4911. The purified PCR products were assembled by the Gibson method (30) and used to transform E. coli DH5α. After confirmation of the putA sequence by DNA sequencing, the plasmid was mobilized into E. coli BL21(DE3).

To express His-tagged PutA (His-PutA), we grew E. coli BL21(DE3) carrying pET28a-putA in LB at 37°C and treated it with 1 mM IPTG (isopropyl β-d-1-thiogalactopyranoside). After 3 h of incubation, the cells were collected and His-PutA was purified with Ni column chromatography (Qiagen). The purified His-PutA protein (50 μg) was 1:1 (vol/vol) mixed with either complete (1st injection) or incomplete (2nd and 3rd injection) Freund’s adjuvant (Sigma) and injected into C57BL/6 mice (6 to 8 week old) intramuscularly in 2-week intervals. One week after the 3rd injection, all mice were anesthetized and blood was collected via cardiac puncture.

Western blot analysis of PutA.

The test strains were grown in CDM broth at 37°C, harvested at exponential growth phase, and normalized to an OD₆₀₀ of 4. The cells were collected by centrifugation, suspended in 100 μL of TSM, and treated with lysostaphin (50 μg/mL) at 37°C for 15 min. The protoplasts cells were collected by centrifugation and suspended in 80 μL of lysis buffer (50 mM Tris, 150 mM NaCl, pH 7.4). The lysed cells were added to 1 μL of DNase I and incubated at 37°C for 30 min. Twenty microliters of 5× SDS sample buffer was added to the cell lysates, and samples were boiled for 10 min. After a brief centrifugation, 10 μL of samples was subjected to 12% SDS-PAGE. Western blotting was carried out with the PutA antibody (1:1,000 dilution). Signals were detected by SuperSignal West atto chemiluminescent substrate (Thermo Scientific).