Citrate can influence the metabolome of S. aureus independent of the TCA cycle

To evaluate the effect of citrate on metabolite production in S. aureus, we utilized ultra-high-performance liquid chromatography coupled with hybrid triple quadrupole time-of-flight mass spectrometry (UHPLC-Q-TOF-MS)-based metabolome analysis. Specifically, we analyzed the wild-type (WT) Newman strain and its isogenic ^TncitB mutant (with a transposon insertion in citB), taking advantage of the fact that the ^TncitB mutant had an intracellular citrate concentration approximately 16 times higher than that of the WT S. aureus Newman (Fig 1B). Given that inactivating citB elevates intracellular citrate levels while also disrupting the TCA cycle [21], we incorporated a double knockout mutant of both citZ and citB (ΔcitZ^TncitB) into our study. In this mutant, the TCA cycle is also impaired, but intracellular citrate levels are significantly lower (~ 2 mM) compared to those in the ^TncitB mutant (Fig 1B). Furthermore, as citrate is a recognized activator of CcpE [12](Fig 1A), a double knockout mutant of both ccpE and citB (ΔccpE^TncitB), was included in the study. This mutant, with comparable intracellular citrate levels to the ^TncitB mutant (Fig 1B), was utilized to explore the involvement of CcpE in the regulatory role of citrate.

Unexpectedly, we observed that the intracellular citrate levels of the ΔcitZ^TncitB mutant were similar to those of the WT Newman strains (Fig 1B). We hypothesized that SbnG, another citrate synthase in S. aureus, might contribute to the intracellular citrate pool in the ΔcitZ^TncitB mutant. Deletion of sbnG did not significantly affect the intracellular citrate levels of the ΔcitZ^TncitB mutant cultured in TSB medium (S1A Fig). However, when cultured in tryptic soy broth (TSB) containing 1mM iron chelator 2,2′-dipyridyl (DIP), the ΔsbnGΔcitZ ^TncitB triple mutant showed an undetectable level of intracellular citrate (S1B Fig). In contrast, the complemented strain, ΔsbnGΔcitZ ^TncitB/p-sbnG, exhibited comparable intracellular citrate levels to the ΔcitZ^TncitB mutant (S1B Fig), suggesting that SbnG plays a crucial role in regulating the intracellular citrate levels of the ΔcitZ^TncitB mutant under iron deprivation conditions. The exact source of intracellular citrate in the ΔcitZ^TncitB mutant cultured in iron-replete TSB medium remains unknown, warranting further investigation to clarify this aspect.

After growth of S. aureus strains in TSB for 12 hours (Fig 1C), we identified a total of 13,314 metabolic features, comprising 6,592 features in ESI+ mode and 6,722 features in ESI- mode, in samples obtained from four testing S. aureus strains: WT Newman, ^TncitB, ΔcitZ^TncitB, and ΔccpE^TncitB (Sheets A to C of S1 Table). The orthogonal partial least squares-discriminant analysis (OPLS-DA) showed clustering of pooled quality control (QC) samples (S2A and S2B Fig), confirming the stability and reproducibility of the instrumental analysis. Analysis of the ΔcitZ^TncitB strain samples revealed 972 differential metabolic features (in ESI⁺ and/or ESI^- modes) compared to ^TncitB cells (VIP≥1 and p-value < 0.05), with 678 features up-regulated and 294 features down-regulated (Fig 1D and Sheet A of S1 Table).

Among the 678 up-regulated features, 387 were found to be up-regulated upon the deletion of ccpE in the ^TncitB mutant (Fig 1D). Meanwhile, of the 294 down-regulated features, 123 were also found to be down-regulated in the ^TncitB mutant as a result of the deletion of ccpE (Fig 1D and Sheet A of S1 Table).

Using database alignment of the measured masses, we were able to identify a total of 102 metabolites that were regulated by citrate, which corresponded to 128 metabolic features (71 in ESI+ mode and 57 in ESI- mode) (Fig 1E and Sheets D to F of S1 Table). Citrate itself was present in both ESI+ and ESI- mode (Fig 1E and Sheets D to F of S1 Table), with a production pattern similar to bacteria grown in TSB medium for 6 hours (Fig 1B). Approximately 50% of the 30 citrate-upregulated metabolites were intermediates of central carbon metabolism (such as dihydroxyacetone, dihydroxyacetone phosphate, 2-Phosphoglycerate, 3-Phospho-D-glycerate, and alpha-ketoglutarate), nicotinamide-adenine dinucleotide phosphate (NADP), and nucleotides (such as ATP, UTP, CDP, UDP, GDP, UMP) (Sheets D to F of S1 Table). In contrast, over 50% of the 72 citrate-downregulated metabolites consisted of amino acids and their derivatives (e.g., dipeptides) (Sheets D to F of S1 Table). Collectively, these findings suggest that the accumulated citrate impacts the metabolome of the ^TncitB mutant, and this effect is primarily mediated by the activation of CcpE.

Citrate can change S. aureus transcriptome independently of TCA

Through RNA-Seq experiments, we discovered that the deletion of citZ led to alterations in the expression levels of 1,219 genes in the ^TncitB mutant (≥ 2-fold change and p < 0.05) (S2 Table). Among these genes, 599 were up-regulated and 620 were down-regulated (Fig 1F). These genes accounted for around 41% of S. aureus Newman genes and were associated with various biological processes (S2 Table). Deletion of the citZ in ^TncitB mutant resulted in significant down-regulation (>16 fold) of genes encoding virulence factors (i.e., hla and scn), phosphotransferase system (PTS) (i.e., NWMN_RS01850-NWMN_RS01835 operon, NWMN_RS14005), aa3-type terminal oxidase (qoxABCD operon), and transcriptional regulators (i.e., cstR, farR and sarR) (S2C Fig and S2 Table). In contrast, transcription of clpC operon (i.e., ctsR-mcsA-mcsB-clpC) and those encoding acetoin reductase ButA and heat-inducible transcription repressor HrcA were dramatically increased (> 16-fold) (S2C Fig and S2 Table). These findings indicate that intracellular citrate accumulation contributes to significant changes in S. aureus gene expression.

Furthermore, through RNA-Seq experiments, we found that removing ccpE from the ^TncitB mutant strain led to changes in the expression of 901 genes (≥ 2-fold change and p < 0.05) (S2 Table). Intriguingly, consistent with the results of the metabolome analysis (Fig 1D), the majority of differentially expressed genes were concurrently regulated by citrate and CcpE (Fig 1F). Thus, the removal of ccpE from the ^TncitB mutant has a comparable, yet not identical, impact on the transcriptome as the deletion of citZ, providing further support and evidence for the activation of CcpE by citrate.

We conducted a differential networking analysis using Joint-Pathway Analysis [22] to further our investigation. This integrated analysis of transcriptomics and metabolomics facilitated a comprehensive visualization of the interactions between differentially expressed genes and perturbed metabolites. By integrating the data, we identified 11 and 8 enriched pathways (with p-value < 0.05 and an impact value greater than 1) in the ΔcitZ^TncitB (S3A Fig) and ΔccpE^TncitB (S3B Fig) mutants, respectively, as compared to the ^TncitB mutant (S3 Fig). Notably, all of the enriched pathways identified in the ΔccpE^TncitB mutant were in complete concordance with those found in the ΔcitZ^TncitB (S3 Fig). These pathways include pyruvate metabolism, purine metabolism, alanine, aspartate and glutamate metabolism, arginine biosynthesis, glycine, serine and threonine metabolism, pyrimidine metabolism, glycolysis or gluconeogenesis, and pentose phosphate pathway (S3 Fig).

Genome-wide identification of CcpE targets

As CcpE regulates the expression of numerous genes in S. aureus either directly or indirectly, we conducted in vitro DNA affinity purification and sequencing (IDAP-Seq) experiments [23] to investigate its direct targets. Using the peak-calling software MACS2, we identify 390 peaks (fold enrichment ≥ 2 and −log₁₀^q-value > 50) from three independent replicates and found 107 peaks were reproducible in at least two independent experiments (Sheet B of S3 Table). Out of the 390 enriched peaks, 238 (61%) had peak summits located at intergenic regions (Sheet A of S3 Table), which constitute only 16.6% of the S. aureus Newman genome [24]. The substantial enrichment of CcpE-binding sites in these intergenic regions, where regulatory sites are commonly located, supports CcpE’s role as a transcription regulator.

The enriched peaks for CcpE displayed a broad occupancy over the S. aureus Newman chromosome (Fig 2A and Sheet A of S3 Table). The promoter region of pyruvate carboxylase gene (pycA) exhibited the highest average fold-enrichment (23.7-fold) among the enriched peaks (S4A Fig and Sheet A of S3 Table). The four next most enriched peaks (ranked by average fold-enrichment) were found in the promoter region of a six-gene operon, including 5-oxoprolinase genes (i.e., pxpB, pxpC, and pxpA) and acetyl-CoA carboxylase genes (i.e., accB and accC) (S4B Fig), the promoter region of citZC operon (comprising two TCA cycle genes, citZ and citC) (S4C Fig), the coding region of NWMN_2386 (S4D Fig), and the promoter region of citB (S4E Fig). Notably, the citB promoter has been previously identified as a direct target of CcpE [12,25], supporting the effectiveness of our IDAP-seq procedure.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Identification of CcpE targets.

(A) Representative image of the CcpE IDAP-seq data illustrated by Integrated Genome Viewer (IGV) software. The black arrows depicted the top five ChIP-Seq peaks identified in the promoters of pycA, citB, pxpB operon, and citZC operon, and the coding region of NWMN_2386. (B) CcpE-binding motif derived from a MEME analysis of a set of 81-bp DNA sequences centered at the summit from 107 reproducible IDAP-seq peaks. (C) Fur-binding motif derived from 23 Fur-recognized sequence in S. aureus. (D) Visualization of promoters bounded by CcpE and Fur. Blue circles denote promoters bounded by CcpE, red circles denote promoters co-targeted by CcpE and Fur, and pink circles denote promoters targeted by Fur. The lines illustrate the connection between transcription factors and their respective target promoters. (E to G) Electropherograms showed the protection pattern of pycA, pxpB, and citZ promoter DNA after digestion with DNase I following incubation without or with CcpE in the absence or presence of citrate. Regions protected (relative to the start codon) by CcpE are underlined. (H) Alignment of the CcpE-protected DNA sequences in the promoter of pycA (E), pxpB (F), citZ (G) and citB. The potential LTTR box (T-N₁₁-A, where N is any nucleotide) are highlighted in red.

https://doi.org/10.1371/journal.ppat.1012425.g002

Upon annotating the 390 CcpE-enriched peaks based on their positions, it was revealed that 682 genes within 443 predicted transcription units (TUs) might potentially be regulated by CcpE (Sheet A of S3 Table). According to functional analyses, these potential CcpE-targeted genes are associated with a variety of biological processes, such as central carbon metabolism (e.g., pycA, zwf, accB and accC, citZC operon, citB, maeB), iron uptake and homeostasis (e.g., isdA, isd operon, isdI, fhuABG operon, tatAC operon, sufCDSUB operon), stress response (e.g., sodM, dsbA, copZ) and the synthesis or export of virulence factors (e.g., pmt operon, capABC operon, tagA, tagH, tarGBXD operon, tarIJLS operon, sasD, csa1A, and nuc) (Sheet A of S3 Table).

After comparing the genes potentially targeted by CcpE, which were obtained from IDAP-Seq experiments, with RNA-seq results (i.e., ΔccpE^TncitB/^TncitB), we discovered that CcpE regulates the expression of 218 genes (in 141 predicted transcription units) responsible for diverse cellular functions (Sheet A of S3 Table). Notably, CcpE binds to the promoters and modulates the expression of genes involved in numerous metabolic processes, such as amino sugar metabolism (e.g., nanA, nanK, nanE, glmS), gluconeogenesis (e.g., fbp), pentose phosphate pathway (e.g., zwf, tal), fatty acids synthesis (e.g., accB, accC), the TCA cycle (e.g. citB), amino acid metabolism (i.e., pxpA, pxpB, pxpC, metL), and the phosphoenolpyruvate (PEP)-pyruvate-oxaloacetate (PPO) node (e.g., pycA), an area located at the intersection of glycolysis and the TCA cycle (S5 Fig) [26]. However, the detailed mechanisms by which CcpE affects diverse metabolic processes remain to be defined.

Using the MEME suite [27], we identify an 18-bp A/T-enriched motif, TWWDWWWWRVTTATCATT (W = A or T, D = G or A or T, R = A or G, V = A or C or G), in a subset of CcpE-targeted DNAs (Fig 2B). Specifically, it was present in 94 out of the 107 reproducible peaks (Fig 2B and S4 Table). Interestingly, we observed a significant similarity between the in silico DNA motif recognized by CcpE and the ferric uptake regulator (Fur), with a statistically significant p-value of 7.42 X 10⁻³, as calculated by Motif Comparison Tool with a Pearson correlation coefficient comparison (Fig 2C). In support of our findings, 11 out of 20 known Fur-regulated promoters [28,29] were directly bound by CcpE (Fig 2D and Sheet A of S3 Table). Taken together, these results provide insights into the regulatory mechanisms of CcpE and suggest the possibility of cross-talk between CcpE and Fur in the regulation of genes.

CcpE inhibits the expression of pycA

As mentioned above, CcpE binds to promoter and modulates the expression of various metabolic gene, including pycA that encodes the pyruvate carboxylase (S4A and S5 Figs), which is a crucial phosphoenolpyruvate (PEP)-pyruvate-oxaloacetate node (PPO-node) enzyme [26] and a determinant factor of S. aureus pathogenicity [11,30]. To validate the binding of CcpE to the promoter of pycA, DNase I footprinting analyses were conducted. The results showed that CcpE indeed binds to the promoter of pycA (Fig 2E). Additionally, through comparison of the CcpE-protected region of the pycA promoter with those of pxpB (Fig 2F), citZ (Fig 2G), or citB promoter [12], we have identified a potential T-N₁₁-A motif (Fig 2H) that are recognized by LysR family proteins [31]. Through 5’-rapid amplification of cDNA ends (5’-RACE) experiments, we determined that the transcriptional start site of pycA is located 275 bp upstream of its start codon, specifically at the A site of the potential T-N₁₁-A motif. When the T-N₁₁-A motif of the pycA promoter was mutated to G-N₁₁-G, the inhibitory impact of ccpE on the pycA promoter (Fig 3A and 3B) disappeared (Fig 3C and 3D). Disrupting the T-N₁₁-A motif led to a loss of upregulation in both wild-type and ΔccpE cells (Fig 3C and 3D), indicating that the T-N₁₁-A motif is not only crucial for the regulation of pycA by CcpE but also plays a significant role in the promoter activity of pycA. Using in vitro transcription assays, we demonstrated that CcpE can suppress pycA transcription (Fig 3E) while activating transcription of citB (Fig 3F). Altogether, these results indicate that CcpE serves as a repressor for pycA.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. CcpE inhibits the expression of pycA.

(A) Effect of ccpE deletion on the expression of pycA in S. aureus Newman cultured in TSB medium for 6 h. ****p < 0.0001, by one-way ANOVA with Dunnett test. (B) Relative pycA-lacZ activity in S. aureus strains grown in TSB. ****p < 0.0001, by two-way ANOVA with Tukey test. (C) Representative image of pycA-lacZ activity in S. aureus strains grown on TSA plate containing 160 μg/mL X-gal for 24 h. Strains carry either the wild-type or a mutated G-N₁₁-G box motif of the pycA promoter. (D) pycA-lacZ activity grown in TSB for 4.5 h. The relative fluorescence intensity (RFU) was normalized to the optical density of bacteria at 600 nm (OD₆₀₀). n.s., p>0.05, ****p < 0.0001, by two-way ANOVA with Sidak test. (E and F) Effect of CcpE on the transcription of pycA (E) and citB (F) in vitro, *p < 0.05 and ***p < 0.001, by two-tailed one-sample t-test. (G to I) Bacterial growth curves in chemical defined medium (CDM) supplemented with (+ Asp) or without (- Asp) the addition of aspartate (2.4 g/L). Newman, ΔpycA, ^TncitB and ΔccpE^TncitB harbor an empty control plasmid, respectively. In (A), (B), (D), and (G to I), data represents mean ± SD from n = 3 biological replicates. In (E) and (F), data represents mean ± SD from n = 3 independent experiments.

https://doi.org/10.1371/journal.ppat.1012425.g003

In S. aureus, PycA supplies oxaloacetate, which serves as a precursor for aspartate, a central building block for many metabolic processes such as biosynthesis of other amino acids (S5 Fig)[32]. Deletion of pycA made S. aureus Newman almost unable to grow in chemically defined medium (CDM) without aspartate (Asp) (Fig 3G). To further confirm the role of CcpE in regulating the biological function of pycA, the ability of WT S. aureus Newman, ^TncitB, and ΔccpE^TncitB strains to grow in the CDM supplemented with or without Asp was tested. Our results indicated that both ^TncitB and ΔccpE^TncitB strains grew similarly to the WT Newman strain in CDM that contains an abundance of amino acids, including Asp (Fig 3H). However, when Asp was excluded from the medium, the growth of ^TncitB, but not ΔccpE^TncitB, was notably lower than that of the WT S. aureus Newman strain (Fig 3H). Complementation of the ΔccpE^TncitB mutant with a plasmid-borne ccpE made it almost unable to grow in CDM without Asp (Fig 3H), suggesting that the activation of CcpE plays a crucial role in the aspartate auxotrophy observed in the ^TncitB mutant. Moreover, the constitutive expression of pycA fully restored the growth of the ^TncitB mutant cultured in CDM without Asp to a level comparable to that of the wild-type Newman strain (Fig 3I). Collectively, these results suggest that activation of CcpE, which lead to decreased expression of pycA, may be the cause of Asp auxotrophy observed in the ^TncitB mutant.

CcpE competes with Fur for binding to the intergenic region of isdA and isdC

CcpE binds to the promoters and enhances the expression of well-known iron-regulated genes (S6A–S6D Fig and Sheet A in S3 Table) including those (i.e., isdA, isdC, isdD, and isdI) encoding the iron-regulated surface determinant (Isd) system [33]. In S. aureus, isdA is transcribed in the opposite direction from isdC, which is the first gene in the isdCDEF-srtB-isdG operon (Fig 4A) [33]. The transcription of isdA, isdCDEF-srtB-isdG, and three other isd genes (isdB, isdI, and isdH) is known to be regulated by Fur through conserved DNA sites called Fur-boxes (Fig 4A) [33]. Using in vitro transcription experiments, we demonstrated that CcpE is a direct activator of isdC (Fig 4B). Using DNase I footprinting experiments, we showed that CcpE binds to the intergenic region of isdA and isdC and that the CcpE-protected regions overlap with those of Fur (Fig 4C). This finding is consistent with our previous outcomes indicating that the in silico DNA motif recognized by CcpE is comparable to that of Fur (Fig 2B and 2C). Furthermore, through EMSA experiments, we showed that CcpE could compete with Fur for binding to the intergenic region of isdA and isdC (Fig 4D), and vice versa (S7A Fig)

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. CcpE competes with Fur for binding to the promoter region of isdC, and iron starvation activates CcpE.

(A) An illustration shows the genetic organization of the isd locus and the location of Fur-box. (B) Effect of CcpE on the in vitro transcription of isdC in the presence of sodium citrate (20 mM). Data reported as mean ± SD of n = 3 independent experiments, *p < 0.05, ****p < 0.0001 by One-way ANOVA with Dunnett test. (C) DNase I footprinting results of CcpE- and Fur-protected regions for the intergenic region between isdA and isdC. Protected regions (relative to the start codon of isdC) were framed with dashed rectangles, as indicated. Red arrows showed the hypersensitive sites. (D) EMSA experiments illustrating the impact of CcpE on the protein-DNA complexes formed by the Fur protein with the isdC promoter DNA (Fur-isdC). (E) Representative image of Western blotting for IsdC. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a loading control. Strains harbor either an empty plasmid control or p-citZ or p-ccpE, as indicated. (F and G) Effect of DIP treatment on the expression of citB, pxpB, isdC (in F), and pycA (in G) in wild-type Newman, ΔcitZ, ΔcitZΔsbnG, and ΔccpE cultured in TSB (without glucose) for about 4.5 h. Data presented with mean ± SD from n = 3 biological replicates, * p < 0.05, *** p < 0.001, ****p < 0.0001 by One-way ANOVA with Dunnett test, using the WT Newman group as the control.

https://doi.org/10.1371/journal.ppat.1012425.g004

To verify the regulatory influence of CcpE on the Isd system in S. aureus, we conducted Western blot analyses on various strains, including the WT Newman strain and its derivatives (i.e., ^TncitB, ΔcitZ^TncitB, ΔcitZ^TncitB/p-citZ, ΔccpE^TncitB, ΔccpE^TncitB/p-ccpE, ΔisdC, and ^TnsrtB). The goal was to clarify the influence of CcpE on the production of IsdC, a vital component of the Isd system, which is attached to the cell wall by sortase B (SrtB) [33]. All the strains, except for ΔisdC and ^TnsrtB mutants, generated three principal IsdC-reactive bands (two with slower migration rates and one with a faster migration rate) when cultured in the presence of 1 mM DIP for 5 hours in TSB medium (Fig 4E). The two slower bands probably represent IsdC precursors (P1 and P2), while the faster band likely corresponds to the mature form of IsdC (Fig 4E). We found that deletion of either citZ or ccpE in the ^TncitB mutant reduced the production of IsdC (Figs 4E and S7B).

Citrate accumulation and activation of CcpE are triggered by iron starvation

Since CcpE is observed to bind to the promoter region and stimulates the expression of multiple Fur-regulated genes (S6 Fig), we hypothesize that citrate could serve as a crucial mediator linking central carbon metabolism and iron homeostasis. Indeed, it was noted that treatment with the DIP (S7C Fig) resulted in a 4-fold increase in the intracellular citrate level of the WT Newman strain (S7D Fig). In contrast, the ^TncitB mutant did not exhibit any significant alteration (S7D Fig). Thus, the DIP-induced citrate accumulation is potentially due to the inhibition of the CitB.

Upon DIP treatment, an 82.2% decrease in aconitase activity was observed in the WT strain of S. aureus Newman (S7E Fig). Using RT-qPCR, we detected elevated transcription levels of the citB gene in different growth phases of S. aureus WT Newman upon DIP treatment (S7F Fig), which implies activation of CcpE [12]. In addition, without DIP treatment, the maximum transcript level of citB occurred during the post-exponential growth phase, whereas in the presence of DIP, there was at least a 6-hour delay in the appearance of maximum citB transcript level during stationary growth phase (S7F Fig). Furthermore, we observed that the effects of DIP treatment on the expression of citB, pxpB, isdC, and pycA were largely mitigated by the deletion of either citZ or ccpE (Fig 4F and 4G). These findings suggest that iron deprivation results in citrate accumulation and the subsequent activation of CcpE.

However, the relationship between iron homeostasis and citrate accumulation in S. aureus appears to be intricate and multifaceted. Fur not only binds to the promoter and triggers citB expression in vitro and in vivo (S8A–S8E Fig), but it also represses IsrR, an sRNA that in turn downregulates CitB and CcpE through post-transcriptional control mechanisms [34]. Furthermore, the revelation that Fur serves an activator for citB aligns with earlier findings in E. coli [35], reinforcing the concept that Fur links iron transport and utilization enzymes with negative-feedback loop pairs for iron homeostasis [35].

Citrate inhibits the transcriptional regulatory activity of Fur

Considering that citrate is known to sequester divalent cations [36,37], including Fe²⁺ and Mn²⁺, which are used by Fur to bind its target DNA [38,39], we performed an EMSA test to investigate whether citrate could influence the transcriptional regulatory function of Fur. In the assay, the purified recombinant Fur protein was exposed to isdC promoter DNA in the presence of Mn²⁺ as a divalent metal, substituting for Fe²⁺ as it is more stable in an aerobic environment and has the same coordination geometry [38,39]. The EMSA results demonstrated that citrate inhibits the binding of recombinant Fur to isdC promoter DNA in a dose-dependent way (Fig 5A). Furthermore, the addition of Mn²⁺ into the EMSA binding mixtures, containing 20 mM citrate, resulted in a dose-dependent increase in the binding of Fur to the isdC promoter DNA (Fig 5B). We demonstrated through in vitro transcription assays that citrate has the ability to alleviate the inhibition of isdC transcription by the recombinant Fur protein (Fig 5C). Additionally, we observed that the deletion of citZ had a greater influence on the expression of Fur-regulated genes, including sirA and sbnA (Fig 5D), and to a lesser extent on isdC and isdA (Fig 5E), in the ^TncitB mutant in comparison to the ccpE deletion. These findings suggest that citrate can modulate the transcriptional regulatory activity of Fur in S. aureus.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Citrate inhibits the regulatory activity of Fur.

(A) EMSA showing the impact of citrate on the DNA binding ability of Fur. (B) EMSA showing the impact of Mn²⁺ on the DNA binding ability of Fur in the presence of 20 mM citrate. (C) Effect of citrate on the transcriptional regulatory activity of Fur in vitro. Data presented with mean ± SD of n = 3 independent experiments. ***p < 0.001, by one sample t-test. (D and E) RT-qPCR analysis of sirA, sbnA, isdC, and isdA transcripts in S. aureus strains cultured in TSB supplemented with 1 mM DIP for 5 h. n.s., p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001 by One-way ANOVA with Tukey test. Data represents the mean ± SD from n = 3 biological replicates.

https://doi.org/10.1371/journal.ppat.1012425.g005

Citrate and CcpE can modulate the pathogenicity of S. aureus in a TCA-independent manner

To investigate the impact of citrate accumulation on the pathogenicity of S. aureus, we exposed WT S. aureus Newman and ΔcitZ, ^TncitB, and ΔcitZ ^TncitB mutant strains to a mouse model of bacteremia and analyzed bacterial survival in the host organs. Compared to the WT Newman strain, both ΔcitZ and the citrate-accumulating ^TncitB mutant exhibited statistically significant decreases in bacterial loads in the kidney or the hearts of infected mice (Fig 6A). This supports the idea that the TCA cycle regulates the pathogenicity of S. aureus [40,41]. Interestingly, we observed that the ΔcitZ^TncitB mutant displayed approximately a 0.9-log increase in CFU bacterial loads in the kidneys and hearts of the infected mice compared to the ^TncitB mutant, suggesting that intracellular citrate accumulation contributes to the decreased pathogenicity of the ^TncitB mutant (Fig 6A). Furthermore, deleting ccpE also increased the bacterial loads of the ^TncitB mutant in the kidneys (~0.9 log10 CFU/organ) in a statistically significant manner (Fig 6B). Additionally, although the increase was not statistically significant, we observed a rise (~1.5 log10 CFU/organ) in the bacterial load in the hearts for the ^TncitB mutant upon ccpE deletion (Fig 6B). These results indicate that citrate and CcpE have the ability to regulate the pathogenicity of S. aureus in the ^TncitB mutant.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Virulence of S. aureus strains and a proposed model for the regulation by citrate.

(A to C) S. aureus colonization in a mouse bacteremia model. Female BALB/c mice were infected retro-orbitally with 4 × 10⁶ CFU (in A), 3 × 10⁶ CFU (in B), or 1 × 10⁶ CFU (in C) of S. aureus. Each symbol represents data from one experimental animal. Horizontal bars indicate observation means, and dashed lines mark limits of detection. Statistical analysis was performed using Mann-Whitney test. (D) A proposal for how citrate regulates genes in S. aureus. Arrows indicate activation or generation, blunted lines indicate inhibition, solid lines indicate a direct influence or connection, and dotted lines indicate indirect influence. Genes expressed in blue or red indicate those whose expression is directly inhibited or activated by CcpE.

https://doi.org/10.1371/journal.ppat.1012425.g006

To further investigate how ccpE modulates the pathogenicity of S. aureus, we conducted mouse infection experiments again using WT S. aureus Newman and ΔpycA, ΔccpE, and ΔpycAΔccpE mutant strains. We observed a significant decrease in virulence of the S. aureus Newman strain upon deletion of pycA, which is consistent with previous studies [11,30]. The removal of ccpE resulted in a statistically significant increase in bacterial loads of the ΔpycA mutant in the kidneys and hearts (Fig 6C), indicating that CcpE can modulate the pathogenicity of S. aureus Newman in a manner independent of pycA. However, the role of citrate in the pathogenicity of S. aureus may be underestimated in the current infection model, as S. aureus has developed its Isd system with a heightened capacity to bind hemoglobin derived from humans compared to other mammals [42] and citrate influences the expression of Isd genes (S6 Fig). The regulatory effect of citrate or CcpE on the pathogenicity of S. aureus seems to be multifaceted, and additional studies are warranted.

Ethics statement

Animal experiments were carried out strictly in accordance with the regulations of the Administration of Affairs Concerning Experimental Animals approved by State Council of People’s Republic of China (11-14-1988). Animal studies were approved and supervised by the Institutional Animal Care and Use Committee (IACUC) of Shanghai Public Health Clinical Centre. Laboratory animal usage license was certified by Shanghai Committee of Science and Technology, the certification number is SYXK-HU-2010-0098.

Bacterial strains, plasmids, and culture conditions

Bacterial strains and plasmids used in this study were listed in S5 Table. Unless otherwise described, Escherichia coli was grown in Luria Bertani (LB) broth or on LA (LB agar) plates. S. aureus strains were cultivated in Tryptone Soya Broth (TSB) (Oxoid, Catalog no. CM0129) or on TSA plate (Oxoid, Catalog no. CM0131). When necessary, antibiotics were used at the following concentrations for maintaining plasmids: for E. coli, 100 μg/mL carbenicillin or 50 μg/mL kanamycin; for S. aureus, 10 μg/mL erythromycin or 10 μg/mL chloramphenicol. For routine cultivating, all of the strains were grown aerobically at 37°C with shaking at 250 rpm.

Construction of S. aureus null or transposon insertion mutants

For creating S. aureus gene null mutants, an allelic replacement strategy was employed [59]. Generally, polymerase chain reactions (PCRs) were performed to amplify upstream sequences (~1 kb) as well as downstream sequences (~1 kb) of the targeted deletion gene. About 10 ng of both the upstream and downstream products were used as templates for the second round of overlapping PCR, resulting in the upstream + downstream fragment. This fragment was used for recombination with pKOR1, a plasmid which permits rapid cloning via lambda recombination and ccdB selection. Recombination reactions were performed using Gateway cloning kit (Invitrogen, Catalog no. 11789–020). The resulting products were subsequently transformed into E. coli DH5a competent cells (Yeasen Biotech. Co. Ltd., Catalog no. 11802ES80). Correct plasmids were electroporated transformed into S. aureus 4220 strain (restriction minus and modification plus) to modify plasmids. The modified plasmids were subsequently transformed into S. aureus Newman strain. Allelic replacement was performed and the null mutant alleles were verified by PCR and DNA sequencing.

For the deletion of citZ, the upstream fragment and downstream fragment of intended deletion was amplified from S. aureus Newman genomic DNA using primers citZ-up-F/citZ-up-R and citZ-down-F/citZ-down-R (S6 Table.), respectively. For the deletion of ccpE, four primer, ccpE-up-F/ccpE-up-R and ccpE-down-F/ccpE-down-R, were used. For the deletion of sbnG, primers sbnG-up-F/sbnG-up-R and sbnG-down-F/sbnG-down-R were used. For the deletion of pycA, primers pycA-up-F/pycA-up-R and pycA-down-F/pycA-down-R were used. For the deletion of isdC, primers isdC-up-F/isdC-up-R and isdC-down-F/isdC-down-R were used. For the deletion of fur, primers fur-up-F/fur-up-R and fur-down-F/fur-down-R were used.

JE2 ^TncitB (NE861, Catalog no. NR-47404) and JE2 ^TnsrtB (NE1363, Catalog no. NR-47905) were acquired from the Nebraska transposon mutant library via the Network on Antimicrobial Resistance in S. aureus (http://www.narsa.net). The transposon was inserted 591 bp downstream the translational start site of citB (locus tag: SAUSA300_1246). For JE2 ^TnsrtB, the transposon was inserted 260 bp downstream the translational start site of srtB (locus tag: SAUSA300_1034). Newman ^TncitB and Newman ^TnsrtB mutant were generated by phage transduction [41] with Φ85 phage lysate from JE2 ^TncitB or JE2 ^TnsrtB, respectively. To obtain ΔcitZ^TncitB or ΔccpE^TncitB double mutants, ΔcitZ or ΔccpE (either of each was generated by allelic replacement as described above) was used as the recipient, and transduction was performed similarly as described above. All of the mutants were confirmed by PCR and DNA sequencing.

Construction of plasmids for gene complementation

To construct plasmids for gene complementation, DNA fragments covering the ribosome binding site and the whole coding regions were amplified from the genomic DNA of Newman, and subsequently cloned into pYJ335 pre-digested with EcoRV. To create plasmid p-citB, a ~2.7 kb DNA fragment containing citB was amplified from S. aureus Newman genomic DNA with primers citB-F and citB-R. To create plasmid p-citZ, a~1.2 kb DNA fragment containing citZ was amplified from S. aureus Newman genomic DNA with primers citZ-F and citZ-R. To create plasmid p-sbnG, a~0.8 kb DNA fragment containing sbnG was amplified from S. aureus Newman genomic DNA with primers sbnG-F and sbnG-R. To create plasmid p-pycA, a~3.5 kb DNA fragment containing pycA was amplified from S. aureus Newman genomic DNA with primers pycA-F and pycA-R. To create plasmid p-fur, a~0.5 kb DNA fragment containing fur was amplified from S. aureus Newman genomic DNA with primers fur-F and fur-R. Clones with target gene downstream of the tetracycline-inducible xyl/tetO promoter were screened by PCR and confirmed by DNA sequencing.

Construction of plasmids for protein expression

The coding regions of CcpE, Sigma A, and Fur were amplified from the genomic DNA of S. aureus Newman with primers pET28a::ccpE-F/pET28a::ccpE-R, pET28a::sigma A-F/pET28a::sigma A-R, and pET28a-sumo::fur-F/pET28a-sumo::fur-R, respectively. Amplified fragments were digested with BamHI and XhoI (for pET28a::ccpE and pET28a::sigma A) or KpnI and HindIII (for pET28a-sumo::fur), and inserted into pET28a or pET28a-sumo digested with the same pair of restriction enzymes to generate protein expression plasmids. All of the constructs were confirmed by PCR and DNA sequencing.

Construction of plasmids for pycA-lacZ transcriptional fusions

For constructing pycA-lacZ-reporter plasmid, the pycA promoter region (-554 to +21 relative to the translational start codon) was PCR amplified from the genomic DNA of Newman with primers pycA-lacZ-F/pycA-lacZ-R and digested with EcoRI and KpnI. The resulting fragment was cloned into pCL-lacZ, which carrying a promoterless lacZ reporter gene, generating a promoter-lacZ reporter fusion (pCL-pycA-lacZ). For site-directed mutagenesis of pycA-lacZ-reporter plasmid, PCR was performed with primers pycA (G-N₁₁-G)-lacZ-F/pycA (G-N₁₁-G)-lacZ-R, using pCL-pycA-lacZ as the template. After amplification, the template was digested with Dpn I at 37°C for 3 h. The digested mixtures were then transformed into chemically-competent E. coli DH5α cells. The presence of the desired mutations was confirmed through DNA sequencing.

Metabolomic analysis

Overnight cultures of S. aureus Newman, ^TncitB, ΔcitZ^TncitB, and ΔccpE^TncitB strains were adjusted to OD₆₀₀ = 5.0 and transferred 1:100 into 10 mL fresh TSB medium, with a tube volume to medium volume = 5:1. Six biological replicates were included in each group and all of the sub-cultures were cultivated at 37°C with shaking at 250 rpm for 12 h. Bacterial cultures were flash frozen in liquid nitrogen and stored at -80°C until use. For LC-MS/MS analysis, samples were thawed at 4°C and washed twice with pre-cold PBS buffer. Subsequently, 1 mL methanol/acetonitrile/H₂O (2:2:1, v/v/v) was added to pellets, followed by vortex for 60s. The mixtures were further sonicated at 4°C with a total time of 60 min, and vortex was performed between two intervals. Sonicated samples were stored at -20°C for 1 h to precipitate proteins and proteins were removed by centrifugated at 14,000 g, 4°C for 20 min. Supernatant was then dried by lyophilization and re-dissolved in 100 μl acetonitrile/water (1:1, v/v) solvent. To monitor the repeatability and stability of instrument, quality control (QC) sample was prepared by pooling 10 μL of each sample and analyzed together with other samples.

LC-MS/MS analysis was performed using an Ultra-High-Performance Liquid Chromatography (UHPLC, 1290 Infinity LC, Agilent Technologies) coupled to a quadrupole time of flight mass spectrometry (AB Sciex TripleTOF 6600). For Hydrophilic Interaction Liquid Chromatography (HILIC) separation, samples were analyzed by a 100 mm×2.1 mm ACQUIY UPLC BEH 1.7 μm column (Waters, Ireland) with a flow rate of 0.3 ml/min and column temperature of 25°C. In both Electrospray Ionization (ESI) negative and positive modes, the mobile phase was A = 25 mM ammonium hydroxide and 25 mM ammonium acetate in water and B = acetonitrile. The gradient was 95% B for 1 min and was linearly reduced to 65% in 13 min, followed by reducing to 40% in 2 min and kept for 2 min, and then increased to 95% in 0.1 min. 5 min of re-equilibration period was employed before analysis for the next sample. ESI source conditions were set as the following: Ion Souce Gas1 (Gas1) as 60, Ion Source Gas2 (Gas2) as 60, curtain gas (CUR) as 30, source temperature: 600°C, Ion Spray Voltage Floating (ISVF) ±5500V. Instrument was set to acquire over the m/z range of 60–1000 Da in MS only acquisition, and TOF MS accumulation time was set at 0.20 s/spectra. In auto MS/MS acquisition, a m/z range of 25–1000 Da and 0.05 s/spectra for accumulation time of product ion scan were employed. The product ion scan is acquired using information dependent acquisition (IDA) with high sensitivity mode selected. Collision energy (CE) was fixed at 35 V±15 eV and declustering potential (DP) was set as ± 60 V.

For data processing, raw MS data (wiff. scan files) was converted to MzXML files via ProteoWizard MSConvert and processed using XCMS for retention time correction, feature detection and alignment, with default parameter settings. Identification of metabolites was performed by comparing the accuracy mass (<25 ppm), and MS/MS spectra with the laboratory’s self-built commercial database (Shanghai Applied Protein Technology Co., Ltd.). In the extracted ion features, only the variables having more than 50% of the non-zero measurement values in at least one group were kept. For the multivariate statistical analysis, the MetaboAnalyst (www.metaboanalyst.ca) web-based system was used. After the data was preprocessed by Pareto-scaling, multi-dimensional statistical analysis, including unsupervised principal component analysis (PCA) and orthogonal partial least-squares discriminant analysis (OPLS-DA) were performed. The first principal component of the variable importance in the projection (VIP) was obtained from OPLS-DA to refine the analysis. The metabolites with VIP values ≥1 and the variables assessed by Student’s two-tailed t-test with p-value < 0.05 were considered as significantly differential metabolites. Differential metabolites were further identified and validated by databases including Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.kegg.jp/) and Human Metabolome Database. Fold-change values of metabolites were calculated by comparing the mean value between each group. Metabolomic analysis was performed with the assistance of Shanghai Applied Protein Technology Co. Ltd. All metabolomic data (six biological replicates for each sample) has been submitted to the database of MetaboLights (https://www.ebi.ac.uk/metabolights/) with the accession numbers MTBLS4276.

RNA-seq and data analysis

Overnight cultures of S. aureus Newman, ^TncitB, ΔcitZ^TncitB, and ΔccpE^TncitB strains were adjusted to OD₆₀₀ = 5.0 and transferred to 10 mL of fresh TSB medium at a ratio of 1:100. The tube volume to medium volume ratio was of 5:1. Three biological replicates were included in each group and all of the sub-cultures were cultivated at 37°C with shaking at 250 rpm for 12 h. For RNA extraction, cultures were harvested by centrifugation at 4°C. After resuspended in 1 mL pre-cold sterile 1 × TE buffer, pellets were homogenized by a homogenizer (Next Advance Bullet Blunder Storm 24). Supernatant containing RNA were collected after centrifugation at 14,000 g, 4°C for 4 min. Total RNA was extracted by a QIAGEN RNeasy extraction kit (Catalog no. 74104) following the manufacture’s recommendations. Ribosomal RNA removal, cDNA library construction and paired-end sequencing with Illumina HiSeq 2000 were performed by Guangdong Magigene Biotechnology Co. Ltd. Differential expressed genes (DEGs) were detected by the edgeR software package with a fold change ≥ 2 and an p-value < 0.05 (edgeR, exact test). All RNA-seq data (triplicate biological replicates for each sample) has been submitted to the NCBI Sequence Read Archive (SRA, https://ncbi.nlm.nih.gov/sra/) under the BioProject accession number PRJNA797550, with the BioSample accession numbers SAMN25010747 to SAMN25010758.

Joint pathway analysis

The transcriptomics-metabolomics coupled analysis was conducted utilizing the Joint Pathway Analysis module of MetaboAnalyst 6.0 [22], with the S. aureus N315 strain serving as the model organism. Genes and metabolites exhibiting significant regulation (gene with ≥2-fold change and p<0.05, and metabolites with VIP≥1 and p<0.05) were selected for Joint Pathway Analysis. This analysis was based on all pathways, including both metabolic pathways and gene-only pathways (i.e. regulatory pathways), with the enrichment analysis grounded in the hypergeometric test. The calculation of p-value was achieved through an integrated approach, wherein genes and metabolites are pooled into a single query and used to perform enrichment analysis within their "pooled universe". For pathways with hits from only one input type, p values calculated from their individual universe will be used. Pathway impact values were determined based on pathway topology measures, which evaluates the potential importance of a particular molecule based on its position within a pathway. Degree centrality was employed as the default topology measurement method for this analysis.

Overexpression and purification of recombinant CcpE, SigA, and Fur proteins

The proteins were expressed in E. coli strain BL21(DE3) and purified by Ni-NTA affinity chromatography. Generally, overnight cultures were sub-cultured 1:100 into 1L of fresh LB medium supplemented with 50 μg/mL kanamycin and grown to an OD₆₀₀ of 0.6–0.8. Expression of CcpE with a N-terminal fused His tag (6×His-CcpE) or Sigma A with a N-terminal fused His tag (6×His-SigA) was induced with 0.5 mM isopropyl-β-D-thiogalactoside (IPTG) overnight at 16°C or 25°C, respectively, with shaking at 220 rpm. Expression of Fur with a N-terminal fused Sumo tag (Sumo-Fur) was induced with 0.5 mM IPTG at 30°C for 4 h. Cells were harvested and bacteria pellets were resuspended in 50 mL buffer A (for 6×His-CcpE: 20 mM Tris-HCl, pH 8.0; 300 mM NaCl; 10% glycerol; 5 mM 2-mercaptoethanol; for 6×His-SigA: 50 mM Tris-HCl, pH 8.0; 300 mM NaCl; 10% glycerol; for Sumo-Fur: 50 mM Tris-HCl, pH 7.5; 300 mM NaCl; 10% glycerol; 5 mM 2-mercaptoethanol). Bacterial suspensions were lysed at 4°C by high pressure homogenization for about 5 min. Lysates were centrifugated at 15,000 g for 25 min, the resulting supernatants were loaded onto a His Trap HP 5-ml column (GE Healthcare, Catalog no. 17-5248-01) pre-equilibrated with buffer A. His-tagged proteins were purified with an AKTA pure system. Briefly, 40 mL 10% buffer B (buffer A plus 400 mM imidazole) was used to elute most of the non-specific binding proteins, followed by step elution with 10–100% buffer B at a flow rate of 2 mL/min for 15 column volumes (CVs). Purified proteins were analyzed by SDS-PAGE with Coomassie blue staining. Fractions containing the targeted proteins were pooled.

For 6×His-SigA and Sumo-Fur, pooled proteins were dialyzed at 4°C with 2 L buffer C (for His-SigA: 50 mM Tris-HCl, pH 8.0; 150 mM NaCl; 10% glycerol; for Sumo-Fur: 50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 10% glycerol; 5 mM 2-mercaptoethanol;) for 6 h~8 h to remove most of the imidazole. After dialysis was complete, purified proteins were sub-packaged and stored at -80°C until use.

For 6×His-CcpE, proteins were concentrated and filtered, followed by removing imidazole with a Hi Trap Desalting column (GE healthcare, Catalog no. 17-1408-01) using buffer A. The resulting 6×His-CcpE was stored and used for electrophoretic mobility shift and DNase I footprinting assays. 6×His-CcpE used for in vitro transcription were further purified by Superdex 200 (10/300) gel-filtration chromatography (GE Healthcare, Catalog no. 28-9909-44) with buffer A, after Ni-column affinity purification.

In vitro DNA affinity purification and sequencing (IDAP-Seq)

The genomic DNA of S. aureus Newman was extracted using a DNA purification kit (Tiangen, Catalog no. DP302-02) after lysis with lysostaphin. DNase free RNase A was added to a final concentration of 20 μg/mL to digest RNA at 37°C for 1h. After digestion, genomic DNA was sheared by sonication (SCIENTZ, Ultrasonic homogenizer, JY92-IIDN) to a fragment size range of 300 to 600 bp with a total working time of 8 min (one second on and one second off on ice, amplitude transformer 2, 10% power). Sheared DNA was recovered with a Cycle-Pure kit (Omega, Catalog no. D6492-02). A total volume of 300 μl, including 2.4 μg recovered DNA, 140 nM purified 6×His-CcpE, and 1×binding buffer (20 mM Tris-HCl, pH 7.4; 50 mM KCl; 20 mM MgCl₂; 1 mM EDTA; 5% NP-40; 5% glycerol; 1 mM DTT; 20 mM citrate), was used to perform in vitro DNA affinity purification. The binding mixtures were incubated at room temperature for 15 min. During this time, Ni-NTA agarose resin (Yeasen Biotech. Co.Ltd., Catalog no. 20502ES10) was washed twice with ddH₂O and pre-equilibrated twice with 300 μL 1×binding buffer. Subsequently, the binding mixtures were added to the pre-equilibrated Ni-NTA agarose resin and the whole mixtures were incubated at 4°C with gentle rotation for 30 min. After that, the resin was washed four times with 300 μL 1×binding buffer and followed by eluted with 300 μL elution buffer (20 mM Tris-HCl, pH 7.4; 50 mM KCl; 20 mM MgCl₂; 1 mM EDTA; 5% NP-40; 5% glycerol; 1 mM DTT; 20 mM citrate; 500 mM imidazole) at room temperature for 30 min. Finally, the eluent was purified with a Cycle-Pure Kit (Omega, Catalog no. D6492-02).

Purified DNA samples were quantified by a Qubit 2.0 Fluorometer (Invitrogen) and qualified by Agilent Bioanalyszer 2100 (Agilent Technologies). Next generation sequencing library preparations were constructed according to the manufacturer’s protocol (NEBNext II DNA library Prep kit for Illumina). For each sample, at least 10 ng in vitro affinity purified DNA or input DNA (sheared genomic DNA before in vitro DNA affinity purification) was used for library preparation. Final libraries were validated with Agilent 2100 Bioanalyzer and then sequenced with the HiSeq X system (Illumina). Each sample generated about 20 million reads and technical sequences, including adapters, PCR primers, and quality of bases lower than 20 bp were removed by Cutadapt (version 1.9.1) to obtain high quality clean data. Clean reads were mapped to the S. aureus Newman genomes using Bowtie (version 2.2.6) [60]. Enriched peaks were identified by Model-based Analysis of Chip-seq 2 (MACS2) software [61]. Peaks identified by MACS2 with an adjusted score of False Discovery Rate (FDR) corrected p-value (i.e., -log₁₀ q-value) of at least 50 was set to ensure that a high-quality peak annotation was obtained. The original IDAP-Seq data files were deposited to the NCBI Sequence Read Archive (SRA, https://ncbi.nlm.nih.gov/sra/) under the BioProject accession number PRJNA796052, with the BioSample accession numbers SAMN24812581 to SAMN24812584.

Multiple EM for Motif Elicitation (MEME) analysis

To identify the conserved DNA binding sites of 6×His-CcpE, all 107 peaks (fold enrichment ≥ 2 and -log₁₀ q value of >50) occurred at least twice in the three independent replicates were selected and used to find the conserved motif. For motif detection, sequences of ± 40 bp from the peak summit (81bp in total) were submitted for motif analysis by MEME (https://meme-suite.org/meme/tools/meme) with the minimum motif width set at thirteen, and other parameters set as default [62].

To find the conserved DNA binding sites of Fur, twenty-three potential Fur boxes, each comprising a 19 bp fragment, were download from the RegPrecise website (https://regprecise.lbl.gov/sites.jsp?regulog_id=670), which used S. aureus N315 as a model organism. These 23 Fur boxes distributed in the promoter regions of twenty genes or operons in total. Among them, three Fur boxes were located in the promoter region of ftnA (SA1709), two Fur boxes were located in the promoter region of the arlRS operon (SA1248-SA1246), and the remaining eighteen Fur boxes were located in the promoter regions of icuB (SA1983), isdH (SA1552), isdB (SA0976), isdCDEF-srtB-isdG (SA0978-SA0983), isdA (SA0977), fhuD2 (SA2079), fhuABG (SA0602-SA0604), feoAB (SA2338-SA2337), efeOBU (SA0331-SA0333), citB (SA1184), yfmCDE (SA1979-SA1977), trxB (SA2162), tatAC (SA0335-SA0334), sstABCD (SA0688-SA0691), sirABC (SA0111-SA0109), sbnABCDEFGHI (SA0112-SA0120), sfnaABC (SA1982-SA1980), ntrA (SA0757), respectively. These Fur boxes were aligned with the homologs from S. aureus Newman, and only one base pair of the tatA putative Fux box (ATTGATAATAATTATCGTT in Newman) is different from that of S. aureus N315 (ATTGATAATCATTATCGTT). Finally, the Fur boxes from S. aureus Newman were submitted for motif analysis by MEME with default settings.

Electrophoretic Mobility Shift Assay (EMSA)

EMSA was performed with minor modifications following the previously described protocol [63,64]. Fragments used for EMSA were generated by PCR using the genomic DNA of Newman as a template. For the promoter of isdC, a 359 bp fragment covering nucleotide -252 to nucleotide +107 relative to the start codon of isdC was amplified with the primer isdC-EMSA-F/isdC-EMSA-R. For the promoter of citB, a 697 bp fragment covering nucleotide -624 to nucleotide +73 relative to the start codon of citB was amplified with the primer citB-EMSA-F/citB-EMSA-R. For the coding sequence of murA, a 1260 bp fragment covering nucleotide +3 to nucleotide +1262 relative to the start codon of murA was amplified with the primer murA-EMSA-F/murA-EMSA-R. This fragment was used as a control DNA fragment to detect the binding ability of Fur with the promoter of citB. All PCR products were purified with a Cycle-Pure kit (Omega, Catalog no. D6492-02).

To perform EMSA with Sumo-Fur, the following binding buffer was used: 10 mM Tris-borate, pH 8.0; 1 mM MgCl₂; 25 mM KCl; 100 μM MnCl₂; and 50 μg/mL BSA. Reaction mixtures were incubated at room temperature for 20 min and 2 μL 10×EMSA loading buffer (Beyotime, Catalog no. GS007) was added to the mixtures before electrophoresis. Electrophoresis was carried out with 5% non-denaturing polyacrylamide gel in 0.5×TB buffer (45 mM Tris, 45 mM borate, pH 8.0) plus 100 μM MnCl₂. To evaluate the impact of citrate on DNA binding affinity of Sumo-Fur, citrate was added to the binding buffer with indicated concentrations.

To test the competition of CcpE and Fur for their binding to the promoter of isdC, one protein was first added at a fixed concentration into the reaction buffer (10 mM Tris-borate, pH 8.0; 1 mM MgCl₂; 25 mM KCl; 100 μM MnCl₂; and 50 μg/mL BSA, 20 mM citrate) and incubation for 10 minutes at 25°C, the other protein was then added at a gradually increased concentration in order to compete with the first one for the binding, after incubation at 25°C for another 25 min, reaction products were loaded onto 4.5% non-denaturing polyacrylamide gel and followed the procedure as described above. A 200-bp DAN fragment covering nucleotide -127 to nucleotide +73 relative to the start codon of citB was amplified with the primer citB-EMSA-F2/citB-EMSA-R, and was used as the control DNA. The Fur proteins were obtained by treating Sumo-Fur with ULP1 (invitrogen) at 4°C overnight following the manufacturing instructions. Fur instead of Sumo-Fur was used in competition experiment to better distinguish the shift bands between CcpE and Fur.

Dye primer-based DNase I footprinting assay

Dye primer-based DNase I footprinting assay was performed as described previously [12,64]. Generally, DNA fragments were amplified using S. aureus Newman genomic DNA as a template. For amplification the promoter of isdC, a 359 bp fragment covering nucleotide -252 to nucleotide +107 relative to the start codon of isdC was amplified with the primer isdC-footprinting-F/isdC-EMSA-R. For the promoter of citB, a 697 bp fragment covering nucleotide -624 to nucleotide +73 relative to the start codon of citB was amplified with the primer citB-footprinting-F/citB-EMSA-R. For the promoter of pycA, a 369 bp fragment covering nucleotide -469 to nucleotide -101 relative to the start codon of citB was amplified with primers pycA-footprinting-F/pycA-footprinting-R. For the promoter of pxpB, a 265 bp fragment covering nucleotide -226 to nucleotide +39 relative to the start codon of citB was amplified with primers pxpB-footprinting-F/pxpB-footprinting-R. For the promoter of citZ, a 370 bp fragment covering nucleotide -359 to nucleotide +11 relative to the start codon of citZ was amplified with primers citZ-footprinting-F/citZ-footprinting-R. All the DNA fragments were purified with a Cycle-Pure Kit (Omega, Catalog no. D6492-02).

For foot-printing assay, 300 ng purified DNA fragment, indicated amount of proteins (6×His-CcpE or Sumo-Fur) and the corresponding binding buffer was added to a total volume of 50 μL. The composition of the binding buffer for Sumo-Fur is as follows: 10 mM Tris-borate (pH 8.0), 1 mM MgCl₂, 25 mM KCl, 100 μM MnCl₂, and 50 μg/mL BSA. For 6×His-CcpE, the binding buffer includes: 20 mM Tris-HCl (pH 7.4), 50 mM KCl, 20 mM MgCl₂, 1 mM EDTA, 5% NP-40, 5% glycerol, and 1 mM DTT. To study the effect of citrate on the protection region of 6×His-CcpE for the pycA, pxpB or citZ promoter, 20 mM citrate was added to the buffer as needed.

For the reaction, the mixtures were incubated for 25 min on ice (for 6×His-CcpE) or 20 min at room temperature (for Sumo-Fur), followed by adding of 0.01 Unit RQ1 RNase-Free DNase I (Promega, Catalog no. M610A). DNase I digestion was carried out at 30°C for 3 min and quenched with 90 μL stopping buffer (200 mM NaCl; 30 mM EDTA; 1% SDS). The mixtures were further extracted with 200 μL phenol: chloroform: isoamylalcohol (v:v:v, 25:24:1), and aqueous phase containing DNA fragment was precipitated with equal volume of isopropanol, washed twice with ice-cold 70% ethanol (v/v), and dried under vacuum. Finally, 10 μL ddH₂O was used to dissolve the digested DNA fragment. Next, 5 μL digested DNA was mixed with 4.9 μL HiDi formamide and 0.1 μL Genescan-500 LIZ size standards (Applied Biosystems). A 3730XL DNA analyzer was used to detect the sample, and the result was analyzed with GeneMapper software (Applied Biosystems). A dye primer based sequencing kit (Thermofisher, Catalog no. 79260) was applied to accurately determine the sequences of protection region, and the corresponding label-free promoter DNA fragment was used as a template in the sequencing reaction. Electropherograms were viewed with GeneMarker v1.91 (Applied Biosystems).

Rapid amplification of sequences from the 5’ ends of mRNAs (5’-RACE)

S. aureus Newman strain was grown to mid-log phase and harvested by centrifugation. Total RNA was isolated and the transcription start site identification of pycA was identified using the 5′-RACE kit (TAKARA). 5′-RACE experiment was performed according to the instructions of the manufacturer. Generally, reaction without reverse transcriptase was used as the negative control. The synthesis of first-strand cDNA initiated from gene-specific reverse primer pycA-race-1. Using terminal deoxynucleotide transferase, an oligo-dC tail was then added to the 3′-end of the cDNA. Subsequently, direct amplification of tailed cDNA using nested primer pycA-race-2 and an abridged anchor primer. PCR products were cloned into Stellar Competent Cells and sequenced using the universal primer M13F of the T vector. The first nucleotide following the oligo-dC sequence was taken as the transcriptional start.

In vitro transcription assay

For in vitro transcription of citB, a 662 bp fragment covering nucleotide -318 to nucleotide +344 relative to the start codon of citB was amplified with the primers citB-trans-F/citB-trans-R. For in vitro transcription of pycA, a 757 bp fragment covering nucleotide -469 to nucleotide +288 relative to the start codon of pycA was amplified with the primers pycA-trans-F/pycA-trans-R. For in vitro transcription of isdC, a 800 bp fragment covering nucleotide -480 to nucleotide +320 relative to the start codon of isdC was amplified with the primers isdC-trans-F/isdC-trans-R.

All the DNA fragments was amplified form S. aureus Newman genomic DNA, purified using a Cycle-Pure kit (Omega, Catalog no. D6492-02), dissolved in RNase-free ddH₂O, and stored at -20°C until used as template for in vitro transcription assay. In general, RNA polymerase holoenzyme (RNAP) was reconstituted by incubating the E. coli RNAP^core (1U/μl, NEB, Catalog no.M0550S) with same volume of S. aureus SigA (28 μM as stock) on ice for 15 min prior to use. For in vitro transcription, DNA template (10 nM) was pre-incubated with reconstituted RNAP (2 μL) in transcription buffer (final volume 20 μL) containing 40 mM Tris-HCl, pH 7.5; 150 mM KCl; 10 mM MgCl₂; 1 mM DTT; 0.01% Triton X-100; 0.1 mg/mL bovine serum albumin and 10 U RNasin (RNase inhibitor, Solarbio, Catalog no. R8060) for 10 min on ice. To evaluate the impact of 6×His-CcpE or Sumo-Fur on transcription, the indicated concentrations of 6×His-CcpE or Sumo-Fur was pre-incubated with the DNA for 20 min on ice before the addition of RNAP. When necessary, MnCl₂ or citrate was included in transcription buffer at a final concentration of 100 μM and 20 mM, respectively.

Transcription was initiated by the addition of 1 μL nucleotide mixtures (NEB, Catalog no. N0466S), and the reaction mixtures were incubated at 37°C for 15 min. Subsequently, 1U RNase free DNase RQ1 (Promega, Catalog no. M6101) was used to digest the template DNA according to the manufacture’s instruction. After which, transcripts were recovered by RNA Extraction Reagent (Solarbio, Catalog no. P1011), and precipitated with isopropanol. The precipitates were washed twice with ice cold 70% ethanol and dried by vacuum centrifugal concentration. Then, the dried transcripts were dissolved in 200 μL RNase free ddH₂O and reversely transcribed to cDNA with the PrimeScript RT reagent kit (Takara, Catalog no. RR037A) with random primers. Finally, cDNA was used as a template to quantitate the transcripts. For quantitation transcripts of isdC, citB and pycA, primer pairs of isdC-RT-F/isdC-RT-R, citB-RT-F/citB-RT-R and pycA-RT-trans-F/pycA-RT-trans-R, respectively, were used. Groups without RNAP were included as a negative control, and the results were normalized to groups with RNAP but no 6His-CcpE or Sumo-Fur.

Preparation of anti-IsdC and Western blot analysis

The DNA fragment coding the 29–192 amino acid residues of IsdC was PCR amplified using primers pET28a-sumo::isdC_29-192-F/pET28a-sumo::isdC_29-192-R, and ligated to pET28a-Sumo to generate pET28a-sumo::isdC_29-192. The Proteins were expressed in E. coli Rosetta (DE3) after induction with 0.8 mM IPTG at 37°C for 4 h and then were purified by a His Trap HP column. Purified Sumo-IsdC_29-192 was submitted to immune Japanese White Rabbit at day 1, 12, 26, 40 with 0.3 mg, 0.15 mg, 0.15 mg and 0.15 mg proteins, respectively. At day 52, serum of rabbit was used to enrich anti-IsdC antibody by affinity purification. Concentrated antibodies were obtained with a concentration of 2 mg/mL and were stored in 50% glycerol in PBS buffer. Anti-IsdC antibodies were prepared by ABclonal Biotechnology Co., Ltd.

To detect the protein abundance of IsdC, overnight cultures were sub-cultured with an initial OD₆₀₀ of 0.1 into 10 mL fresh TSB medium supplemented with 1 mM 2,2’-bipyridyl (DIP), with a tube volume to medium volume = 5:1. After 5 h growth at 37°C with shaking at 250 rpm, cells were harvested and the pellets were washed once with 1×TE buffer (10 mM Tris, pH 8.5; 1 mM EDTA). The pellets were further suspended in 1×TE buffer to an OD₆₀₀ of 6.0, ten units lysostaphin (Sigma, Catalog no. 9011-93-2) were added to 100 μL suspensions and the mixtures were incubated at 37°C with gentle shaking for about 30 min to obtain cell lysate. 25 μL 5×SDS loading buffer (250 mM Tris-HCl, pH 6.8; 10% SDS; 0.5% bromophenol blue; 50% glycerol; and 100 mM DTT) was added to the 100 μL cell lysate and the mixtures were heated at 100°C for 10 min. SDS-PAGE was performed with a 10% polyacrylamide gel at 150 V for about 60 min. Proteins on the gel was transferred to PVDF membrane (Bio-Rad, Catalog no. 1620177) by semi-dry transfer for 20 min at room temperature. Proteins on the membrane was blocked with 5% (w/v) skimmed milk (prepared in TTBS buffer:10 mM Tris-HCl, pH 7.5; 150 mM NaCl; 0.04% Tween 20) at room temperature for 2.5 h with gentle shaking and then incubated with the primary anti-IsdC with a 1:1000 dilution at 4°C overnight. Following wash out the non-specific binding of primary antibody, a 1:5000 diluted anti-rabbit IgG conjugated to horseradish peroxidase (Cwbiotech, Catalog no. CW0103S) was added and incubated at room temperature for 2 h with gentle shaking. Chemiluminescence was detected by Tanon-5200 multi according to the manufacturing instructions and band intensity was analyzed by using Image J software.

S. aureus growth in TSB with or without DIP

Overnight cultures of S. aureus Newman or its isogenic mutants was transferred into 10 mL fresh TSB medium supplemented with or without 1 mM DIP in a 50 mL tube. All the strains were adjusted with an initial OD₆₀₀ of 0.1 and sub-cultured at 37°C with shaking at 250 rpm. Cultures were sampled at indicated time points, and the optical density at 600 nm (OD₆₀₀) was monitored using a NanoDrop 2000c (Thermo Scientific) spectrophotometer.

Growth of S. aureus in chemical defined medium (CDM)

Chemical defined medium (CDM) was prepared as previously described [65], except that L-aspartic acid (Asp) was omitted. In general, 1 L of this medium contained the following nutrients: L-cystine 0.24 g, L-glutamic acid 2.4 g, L-proline 2.4 g, L-arginine 0.36 g, L-glycine 2.4 g, L-histidine 0.48 g, L-lysine HCl 0.6 g, L-serine 2.4 g, L-valine 0.48 g, L-tyrosine 0.18 g, L-threonine 2.4 g, L-alanine 2.4 g, L-isoleucine 0.6 g, L-leucine 0.6 g, L-phenylalanine 0.2 g, L-tryptophan 0.06 g, L-methionine 0.18 g, D-glucose 4.0 g, (NH4)₂SO₄ 6.84 g, nicotinic acid 11.94×10⁻³ g, thiamine HCl 6.0×10⁻⁴ g, KH₂PO4 1.34 g, NaHPO₄ 5.7 g, CaCl₂.2H₂O 5.0×10⁻³ g, CoCl₂.6H₂O 2.4×10⁻³ g, CuSO₄.5H₂O 5.0×10⁻⁴ g, MnSO₄.4H₂O 5.6×10⁻³ g, NaCl 5.8×10⁻³ g, ZnCl₂ 3.4×10⁻³ g, FeCl₃.6H₂O 2.7×10⁻² g, MgSO₄. 7H₂O 2.47×10⁻¹ g. The resulting CDM was defined as CDM (-Asp). L-Aspartic acid (Asp, CAS no. 56-84-8) was dissolved with NaOH as a 100× stock solution, and adjusted with HCl to pH 7.5 for use. To evaluate the influence of Asp on the growth ability of strains, overnight cultures of S. aureus (in TSB) strains were washed once with CDM (-Asp) and transferred to fresh CDM supplemented without or with Asp (final concentration: 2.4 g/L) with an initial OD₆₀₀ of 0.05. 150 μL bacteria samples were triplicately added to 96 well plate with a transparent bottom, and 60 μL filter-sterilized mineral oil was added to prevent evaporation during the growth at 37°C. The A₆₀₀ absorption was detected by a Synergy 2 Multi-mode Microplate Reader with an interval of 1 h.

Intracellular citrate concentration determination

Liquid-state nuclear magnetic resonance (NMR) was used to detect the intracellular citrate concentration of bacteria. Generally, overnight cultures were transferred to 10 mL fresh TSB medium (tube volume to medium volume = 5:1) with an initial OD₆₀₀ of 0.1. After 6 h sub-cultivation at 37°C with shaking at 250 rpm, cultures were centrifugated at 4°C to discard the supernatant. Cell pellets were flash frozen in liquid nitrogen and stored in -80°C until NMR analysis. To quantify the intracellular citrate concentration, bacteria cell pellets were washed twice with ice cold PBS buffer (137 mM NaCl, 2.7 mM KCl, 8 mM Na₂HPO₄, and 2 mM KH₂PO₄), resuspended in 1 mL D₂O dissolved PB buffer (410 mM Na₂HPO₄, 50 mM NaH₂PO₄, in 100 mL D₂O, pH = 7.4) and homogenized with a FastPrep-24 sample preparation system (MP Biomedicals, 6.5 M/s, 40 s one time, three times in total with 3 min interval on ice between each homogenization). Subsequently, the homogenized mixtures were centrifugated at 4°C, 16,000 g for 5 min to remove the cell debris. 800 μL supernatant was used to quantify the citrate concentration at an analytical platform based on a liquid high-resolution Bruker Avance-III NMR spectrometer equipped with a high-sensitivity cryogenic probe operating at a frequency of 600.17 MHz for 1H observation at 298 K. Different concentrations of citrate (Shanghai Urchem Limited, CAS no. 6132-04-3), including 5 μM, 10 μM, 20 μM, 40 μM, 80 μM, 160 μM, 320 μM, 640 μM, 1280 μM were prepared with D₂O, and used for the construction of standard curves. Intracellular citrate concentration was calculate based on the assumptions that S. aureus cell volume is 5×10⁻¹³ mL [66] and one OD₆₀₀ corresponds to 2×10⁸ cells/mL [67].

Citrate assay kit (Abnova, Catalog no. KA0864) was also used for the determination of intracellular citrate concentration. In this circumstance, overnight cultures were sub-cultured with an initial OD₆₀₀ of 0.1 into 10 mL fresh TSB medium supplemented with or without 1 mM DIP, and with a tube volume to medium volume = 5:1. All Cultures were sampled after 2 h cultivation at 37°C with shaking at 250 rpm and washed twice with ice cold PBS buffer, followed by resuspended in citrate assay buffer (provided in the assay kit) with an OD₆₀₀ of 10. Thirty Units lysostaphin (Sigma, Catalog no. 9011-93-2) were added to 300 μL suspensions, and the mixtures were incubated at 37°C with gentle shaking for 15 min to lyse bacterial cells. Clear lysate was submitted to citrate concentration quantification according to the manufacturing instructions.

Aconitase activity assays

Overnight cultures of S. aureus Newman and ^TncitB were transferred to 50 mL fresh TSB (flask volume to medium volume = 5:1) without or with 1 mM DIP (for Newman) with an initial OD₆₀₀ of 0.1. All of the sub-cultures were cultivated at 37°C with shaking at 250 rpm for 2 h. Cultures were centrifugated at 4°C to discard the supernatant. Cell pellets were washed twice with ice-cold PBS buffer and resuspended with 1 mL ice-cold PBS buffer. Suspensions were submitted to homogenization by a Bullet Blender Storm tissue homogenizer (Next Advance). Supernatant was collected after centrifugation at 4°C, 16,000 g for 5 min. Aconitase activities were assayed with an Aconitase Assay kit (Sigma, MAK337) according to the manufacturing instructions. Bicinchoninic acid (BCA) protein assay kit (meilunbio, MA0082) was used to quantify the protein concentration of the supernatant, and aconitase activities were normalized to total proteins of samples.

Real-time quantitative PCR (RT-qPCT) analysis

Total RNA was extracted by a QIAGEN RNeasy extraction kit (Catalog no. 74104) following the manufacture’s recommendations. Subsequently, total DNase-treated RNA (~400 ng) was reversely transcribed to cDNA with the PrimeScript RT reagent kit (Takara, Catalog no. RR037A) with random primers. For RT-qPCR analysis, 10 μL 2×Hieff qPCR SYBR Green Master Mix (Yeasen Biotech. Co.Ltd.), 300 nM primers and 2 μL diluted cDNA (as template) was used, ddH₂O was added to a final volume of 20 μL. RT-qPCR was performed using a CFX Connect Real-Time PCR Detection System (Bio-Rad). Dissociation curve was analyzed to check the product homogeneity. The primer pairs used for RT-qPCR of citB, pycA, pxpB, isdC, isdA, sirA, sbnA, and 16S rRNA are citB-RT-F/citB-RT-R, pycA-RT-F/pycA-RT-R, pxpB-RT-F/pxpB-RT-R, isdC-RT-F/isdC-RT-R, isdA-RT-F/isdA-RT-R, sirA-RT-F/sirA-RT-R, sbnA-RT-F/sbnA-RT-R and 16S-RT-F/16S-RT-R, respectively. Amplicon of 16S rRNA was used as the internal control. Expression levels of targeted genes were calculated by relative quantification method (ΔΔCT) as previously described [12].

β-galactosidase assays

Overnight cultures of the promoter-lacZ reporter strains were sub-cultured into fresh TSB medium with a final optical density 600 nm (OD₆₀₀) = 0.1. Liquid cultures were grown in a 50-mL tube with a volume to medium ration of 5:1, shaking at 37°C with 250 rpm. Cultures were sampled at the indicated time points and bacteria pellets were washed once with Z-buffer (60 mM Na₃PO₄.7H₂O, 40 mM NaH₂PO₄.H₂O, 10 mM KCl, 1 mM MgSO₄.7H₂O, pH7.0). Subsequently, β-galactosidase activity assays were performed according to the method described previously [12]. Product (7-hydroxy-4-methylcoumarin, 4 MU) was detected with an excitation/emission wavelength of 360 ± 20 nm/460 ± 20 nm respectively, using a Synergy 2 (Biotek) plate reader following the manufacturer’s instructions. Samples were detected in triplicate and relative LacZ activity was normalized based on OD₆₀₀.

Mouse infection models

Overnight cultures of S. aureus strains were diluted 100-fold in 50 mL fresh TSB medium in a 250 mL flask. Sub-cultures were performed at 37°C with shaking at 250 rpm for 3 h. Bacteria pellets were collected by centrifugation, washed twice and resuspended with sterile ice-cold phosphate-buffered saline (PBS). Colony formation units (CFUs) were enumerated before mice were infected.

Mouse infections were performed as described previously [12,68,69]. Briefly, 6 to 8-week-old female BALB/c mice purchased from Shanghai SLAC Laboratory Animal Co. Ltd. were housed under specified pathogen-free conditions. Before infection, mice were anaesthetized with pentobarbital sodium (80 mg/kg) by intraperitoneal injection. Anaesthetized mice were infected retro-orbitally with S. aureus. Four days post-infection, the animals were euthanized and, their kidneys and hearts were aseptically removed and homogenized in 1 mL sterile PBS. Serial dilutions of each organ were then plated on TSA plates to enumerate the colony-forming units (CFUs). Statistical significance was determined using the two-tailed Mann-Whitney test.

Statistical analysis

Statistical analysis was conducted using GraphPad Prism version 7.00 software. Two-tailed one-sample t-test, Student’s two-tailed unpaired t-test, One-way ANOVA, Two-way ANOVA or Mann-Whitney two-tailed test were used to compare datasets where appropriate.