Bacterial strains used in this study were S. aureus N315, N315ΔsprC (sprC deletion mutant) (Zhao et al., 2017), and S. aureus R4220. The plasmid pOS1 was a gift from Associate Professor Qian Liu. Bacteria were cultured at 37°C in tryptic soy broth medium (TSB; Oxoid), rotating at 200 revolutions per minute under aerobic conditions, with a volume to tube ratio of 1/3 according to our early study (Zhao et al., 2017).

Bacterial cultures that grew for 4 h to early logarithmic growth phase (OD₆₀₀ = 0.5) were centrifuged at 13,000×g for 10 min at 4°C. The cell pellets were resuspended in 1 ml of TE buffer (10 mM Tris HCl, 1 mM EDTA, pH 8.0) supplemented with lysostaphin (1 mg/ml, Sangon Biotech) and proteinase K (20 mg/ml, TaKaRa), and incubated for 1 h at 56°C. Total RNA extraction was performed using a MiniBEST Universal RNA Extraction Kit (TaKaRa) according to the manufacturer’s protocol, including an optional on-column DNase treatment procedure (TianGen). RNA degradation and contamination were detected using 1% agarose gel electrophoresis. The purity and concentration of RNA were analyzed by a Nanodrop spectrometer (Thermo Fisher Scientific). The integrity of RNA was precisely measured using the RNA Nano 6000 Assay Kit, and quantified by the Agilent 2100 bioanalyzer (Agilent Technologies Inc.).

Bacterial mRNA was isolated with the MICROBExpress™ kit (Invitrogen) by removing ribosomal RNA (rRNA) from total RNA according to the manufacturer’s instructions. The mRNA was fragmented and reverse transcribed with random hexamer primers (5’-d(NNNNNN)-3’ (N=G, A, T or C)) (Thermo Fisher Scientific) by using the Superscript™ Double-Stranded cDNA synthesis kit (Invitrogen). After repairing the cDNA ends using NEBNext End repair/dA-tailing module (New England Biolabs), sequencing adapter ligation was performed. Then, polymerase chain reaction (PCR) was carried out to generate the cDNA library. AMPure XP beads (Beckman Coulter Inc.) were utilized for purification after each enzymatic reaction. Following the construction of the cDNA library, preliminary quantification and the detection of the insert size were conducted by qubit 2.0 (Life Tech Invitrogen) and Agilent 2100, and the effective concentration of the library (> 2 nM) was accurately quantified by quantitative real-time PCR (qRT-PCR) to ensure the quality of the library. The cDNA library was established using NEBNext^® UltraTM II RNA Library Prep Kit (New England Biolabs).

The prepared cDNA Library was sequenced on Illumina HiSeq 4000 sequencing platform (Illumina) to generate 2 × 150 bp paired-end reads. Quality of RNA-seq data was comprehensively evaluated using RseQC package (version 2.6.3) (Wang et al., 2012). Raw data were aligned to the S. aureus NCTC 8325 genome sequence (GenBank accession number NC_007795). The quantitative expression level of each gene was represented by expected number of fragments per kilobase transcript sequence per million base pairs sequenced (FPKM), which was calculated using Cufflinks software (version 2.2.1). DESeq2 v 1.10.1 package was used to identify differentially expressed genes (DEGs) (with a |log₂(fold change)| ≥1 and a false discovery rate (FDR) ≤ 0.05). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis and Gene Ontology (GO) functional annotation on the DEGs were conducted based on the KEGG and GO databases (R package GOseq v 1.18) (Ashburner et al., 2000; Kanehisa and Goto, 2000).

In order to appraise the RNA-seq data, quantitative real-time polymerase chain reaction (qRT-PCR) was performed to quantify the level of the DEGs. Total RNA was extracted and reverse transcribed into cDNA using the Takara RNA PCR kit (AMV) ver.3.0 kit (TaKaRa), according to the manufacturer’s protocol. Then qRT-PCR was carried out in a 20 μL volume using SYBR Premix Ex Taq™ (Tli RNaseH Plus) kit (TaKaRa) as recommended by the manufacturer. Primers used are listed in Supplementary Table 1. Thermal cycling profile was as follows: 95°C for 30 sec, followed by 40 cycles of 95°C for 5 sec, 60°C for 34 sec; and 95°C for 15 sec, 60°C for 1 min, and 95°C for 15 sec. The 16S rRNA was used as the reference gene for normalization. Three independent experiments were run in triplicate. Relative gene expression was assessed by the 2^-ΔΔCT method (Yang et al., 2020). Melting curve analysis was performed for gene amplification to analyze the primer efficiency.

In order to further confirm the influence of SprC on the expressions of DEGs, a complementation strain of sprC was constructed. Briefly, using DNA from strain N315 as a template, sprC gene was amplified by PCR with primer sprC-up-Nhel (CCGGCTAGCAAGTATTGAAAAATAAAATATTT) and sprC-down-BamHI (CGCGGATCCAACATATATATATTTACTATGAAC). SprC complementation plasmid was generated by cloning the sprC gene into the vector pOS1. The recombinant plasmid pOS1-sprC was first transferred into S. aureus RN4220 and then transformed into strain N315ΔsprC via electroporation (electrotransformation conditions: 1.8 kv, 2.5 ms). According to the methods described above, qRT-PCR was performed to check whether the expression levels of the DEGs (10 randomly selected genes) in the complementation strain (named N315ΔsprC-C) were backfilled to those in the wild-type N315 strain.

To validate that the mRNAs transcribed by DEGs could act as the targets of SprC, we calculated the thermodynamic stability of SprC:mRNA duplex using a bioinformatics tool “standalone algorithm RNAhybrid” on Bielefeld Bioinformatics Service website (https://bibiserv.cebitec.uni-bielefeld.de/). On this website, the molecular free energy (Mfe, kcal/mol) between the two molecules can be obtained. According to the description of Akhtar (Akhtar et al., 2019), a value of Mfe less than 0 indicated that SprC and mRNA could bind spontaneously with a good affinity.

Strain N315 and N315ΔsprC were collected at logarithmic growth phase and centrifuged at 14,000×g for 10 minutes at 4°C. The cell pellets were washed with phosphate-buffered saline (PBS) twice, then flash-frozen in liquid nitrogen, and stored at -80°C. The collected samples were thawed on ice, and metabolites were extracted by vortexing with 50% methanol buffer for 1 min, and incubating at room temperature for 10 min. The extraction mixtures were stored overnight at -20°C, then centrifugated for 20 min at 4,000×g. The supernatants were transferred into new 96-well plates and stored at -80°C prior to metabolomic analysis. The data of non-targeted metabolomic profiling were assembled in both positive ion (pos) and negative ion (neg) modes using LC-MS/MS technique (high-resolution mass spectrometer (Q Exactive), Thermo Fisher Scientific) to probe into the metabolomic compositions and their biofunctions (data from both groups, 6 biological repeats for each). Based on partial least squares method-discriminant analysis (PLS-DA) and variable importance in projection (VIP) value, the fold changes were used to identify differential metabolites. The differential ion defined needed to meet 3 conditions: (1) ratio ≥ 2 or ≤ 1/2; (2) P-value ≤ 0.05; (3) VIP ≥ 1. Compound Discoverer 3.1.0 software (Thermo Fisher Scientific) was used to identify statistically different metabolites. KEGG pathway database was utilized to annotate the different metabolites and exhibit the pathways with differential metabolite enrichment.

All statistical analyses were performed with SAS 9.3 for Windows software (SAS Institute Inc.) and R language. Data comparisons were performed using Student’s t-test. A P-value < 0.05 was considered a statistically significant difference.

Six cDNA libraries constructed based on samples of S. aureus N315 and N315ΔSprC (3 for each) were sequenced. The gene expression results of the three biological replicates obtained from the RNA-seq were highly consistent, which was more conducive to the subsequent transcriptome analysis. A total of 50,637,714 raw reads were generated. After quality control, 47,575,359 reads (24,279,637 and 23,295,722 for wild-type and knockout groups, respectively) were produced. The quality indicators of Q20 and Q30 were > 98% and > 95%, respectively, suggesting successful sequencing of the S. aureus transcriptome. Then the qualified reads were utilized for mapping to the reference genome of S. aureus NCTC 8325 for subsequent analysis. A detailed analysis of the RNA-seq on the six samples is represented in Table 1.

To identify DEGs in S. aureus N315 following sprC knockout, we utilized DESeq2 program to yield S. aureus gene expression profiles. The program identified a total of 2497 genes expressed, among which 60 (2.4%) were differentially regulated, including 37 (1.5%) significantly up-regulated and 23 (0.9%) distinctly down-regulated DEGs (Figure 1A). The scatter plot revealed the expression discrepancies of genes between the wild-type strain and the knockout strain (Figure 1B). In addition, a volcano plot was created to picture the DEGs intuitively between both groups (Figure 1C). The heat map visualized the global expression changes of 60 DEGs between N315 and N315ΔsprC (Figure 1D). From the heat map, we apparently recognized that the expressions of the genes encoding leucocidin ED (SAOUHSC_01955 and SAOUHSC_01954) were markedly up-regulated in N315ΔsprC, which were associated with impetigo, antibiotic-associated diarrhea, and bloodstream infections (Yang et al., 2020). Detailed information about each DEG is available in Supplementary Table 2. The representatives of the significant DEGs involved in metabolism and virulence are listed in Table 2.

To further understand the possible effects of SprC on DEGs, GO functional enrichment analysis was conducted. Our results displayed that the significantly enriched annotations were related to three GO domains: biological process [15 terms, such as oxidation-reduction process (8 genes), translation (3 genes), regulation of transcription (2 genes) and pathogenesis (2 genes)], molecular function [24 terms, most involved in catalytic activity (13 genes) and binding (15 genes)], and cellular component [4 terms, most belonging to intracellular (4 genes) and ribosomal components (3 genes)] (Figure 2). Discrepant transcription profiles acquired from the DEGs in wild-type and knockout strains indicated a potential effect of SprC on S. aureus metabolism and pathogenicity.

In order to further identify the functions of DEGs, all DEGs were mapped to KEGG database for the KEGG pathway enrichment analysis. Totally, 32 pathways were enriched, and the detailed information was presented in Supplementary Table 2. From Supplementary Table 2, we observed that some determinants were involved in several biological activities, and some were limited to a single pathway. The top four clearly enriched pathways, namely metabolic pathways, biosynthesis of antibiotics, biosynthesis of secondary metabolites and purine metabolism were shown in Figure 3A. In addition, data represented in Figure 3A reflected that there were 15 genes mapped to metabolic pathways, 11 genes involved in the pathways of biosynthesis of antibiotics, 11 genes mapped to biosynthesis of secondary metabolites, and 10 genes related to purine metabolism pathway. Besides, the pathways involving fructose and mannose metabolism (5 genes), Staphylococcus aureus infection and ribosome (3 genes each) were also remarkedly enriched (Figure 3A). A bubble chart (Figure 3B) generated with factors of enrichment factor, P value and numbers of DEGs revealed the 20 most significant pathways, among which the above mainly pathways were included.

To confirm the accuracy of gene expression detected by RNA-seq, the DEGs with a defined function including 32 up-regulated genes and 11 down-regulated genes were reanalyzed using qRT-PCR. As indicated in Figure 4, the mRNA levels of 35 of 43 DEGs (81.4%) assessed by qRT-PCR were congruent with the results of RNA-seq (a consistent trend for two genes with no significant difference). However, opposite results were shown for the remaining 8 genes (dltc, purC, purL, SA0231, modA, glcK, SA1360 and pfk) (Figure 4E). Although there were several contradictory results obtained by both methods, the data from RNA-seq were generally reliable.

In order to further verify the influence of SprC on gene expression, the complementation strain N315ΔsprC-C was constructed, screened and verified (Supplementary Figure 1 and Supplementary Sequencing Data). The result of qRT-PCR showed that the expression level of sprC in the complementation strain was returned to that in the wild-type N315 strain (Figure 5). Further analysis discovered that the mRNA levels of the 10 randomly selected DEGs in the complementation strain were also restored to those in the wild-type strain (Supplementary Figure 2).

We predicted the binding of SprC with the mRNA transcribed from the DEGs, except for genes encoding hypothetical proteins. The Mfes of SprC:target mRNA are listed in Table 3. According to Akhtar et al. (2019), low level of Mfe means that an RNA-RNA duplex is thermodynamically stable and it is more difficult to break up the base-paring between RNA strands. A tight RNA-RNA combination is more likely to be viewed as a real target (Akhtar et al., 2019). As shown in Table 3, the predicted Mfes of SprC:mRNA were exceptionally low (-95.5 to-54.4), signaling that SprC and the mRNAs predicted could bind spontaneously with a good affinity. The predicted diagram of the secondary structure of the binding between SprC and mRNA is given in Supplementary Figure 3.

With the raw mass spectrometry data imported from the LC-MS/MS acquisition, 7123 metabolites were identified. Compared with the wild-type strain N315, 256 differential ions (3.6%) were screened based on the conditions of ratio ≥ 2 or ≤ 1/2, P value ≤ 0.05 and VIP ≥ 1 in sprC deletion mutant strain (N315ΔsprC), including 187 (2.6%) up-regulated ions and 69 (1.0%) down-regulated ions. All the differentially regulated metabolites are available in Supplementary Table 3. KEGG pathways analysis indicated that these differential metabolites were enriched in various KEGG pathways of level 1, including cellular processes, environmental information processing, genetic information processing, human diseases, metabolism and organismal system, and the pathways of metabolism enriched the maximum amount of differential metabolites (Figure 6). Among the level 2 pathways of metabolism, we observed that global and overview maps enriched the most differential metabolites (25 metabolites), and amino acid metabolism and metabolism of other amino acid each enriched 10 metabolites (Figure 6). Specific pathway enrichment derived from level 2 pathways using the compounds in the KEGG database was presented as a bubble chart in Supplementary Figure 4, in which 77 specific pathways were identified and many, such as metabolic pathways (the most significantly enriched pathways, including 28 metabolites), biosynthesis of secondary metabolites and purine metabolism, were consistent with those enriched in the RNA-seq data.