Thirty-eight genes highly expressed in three different growth phases were screened via RNA-seq (Figure 1). To identify the highly expressed genes in different growth phases, the growth curve of S. aureus USA300_FPR3757 in TSB medium was first drawn (Figure 1A). Bacterial cells from different time points (2, 6, or 10 h) after inoculation, corresponding to the lag phase (LP), exponential phase (EP), and stationary phase (SP), were collected for RNA extraction and subsequently submitted for RNA-seq. The transcriptome dataset, including read counts and fragments per kilobase million (FPKM) of each gene from LP, EP, and SP, is listed in Supplementary Table 1. The expression levels of all 6,030 genes during LP, EP, and SP were ranked from the most highly expressed to the least expressed according to their FPKM values (Figure 1B). Genes with FPKM values ≥ 1,500 were defined as highly expressed, and there were 152, 81, and 88 highly expressed genes in the LP, EP, and SP growth phases, respectively (Figures 1B,C). Among these genes, 38 genes were all highly expressed in all three growth phases (Figure 1C).

Since the exact -35 and -10 sequences of these promoters were unknown, the gap non-coding sequence between the highly expressed gene and its upstream gene was defined as the potential promoter sequence. Among these 38 highly expressed genes, there were eight genes with very short promoter regions or no promoter region, making them difficult or impossible to clone. Therefore, the promoter regions of the other 30 genes were selected for cloning and evaluation (Supplementary Table 3). If the promoter sequence was longer than 450 bp, then only 450 bp was chosen as a promoter sequence. In addition, the commonly used constitutive promoter of the transcriptional regulator gene sarA P_sarA (P31) was also cloned as a positive control (Bayer et al., 1996; Cheung and Manna, 2005). The length of the selected promoters ranged from 120 to 434 bp, and each promoter was named the gene locus of its corresponding gene. Information on the cloned promoters, including their length, coding products of their corresponding genes, and the transcriptional level (FPKM value) of the selected gene, is shown in Table 1. In short, among these 38 highly expressed genes, the promoter regions of 30 genes were successfully cloned (Table 1).

To check the activity of the selected promoters, the promoter-probe vector pQLV1003 carrying the lacZ reporter gene was constructed as described in the methods and materials (Table 2). All the selected promoters were amplified and cloned into pQLV1003 upstream of a lacZ reporter gene (Figure 2A). The generated plasmids were transformed into S. aureus USA300, and the activity of each cloned promoter was evaluated by measuring the beta-galactosidase activity of each transformed strain.

First, the beta-galactosidase activity of the transformed strains was tested on a TSB X-gal plate. The results showed that 20 strains exhibited blue colonies on the plate (Figure 2B), with the corresponding 20 promoters (P1, P2, P3, P4, P5, P7, P8, P10, P11, P12, P13, P14, P15, P17, P19, P20, P21, P24, and P25) and the positive control P_sarA (P31). Subsequently, the strength of these 20 verified active promoters was quantified by measuring the beta-galactosidase activity of their corresponding bacterial cell lysates (Figure 2C). Overall, the activity of most promoters in EP (6 h) and SP (10 h) was significantly stronger than that in early growing phase LP (2 h), except P3, P11, P15, P17, P19, P21, and P24, which exhibited a similar activity in all growing phases. The activity of most promoters exhibited no significant difference (p-value ≥ 0.05) in EP (6 h) and SP (10 h) with the exception of P1, P2, P4, P5, P7, and P12, among which P1, P2, P4, P7, and P12 had stronger activity in SP and P5 had stronger activity in EP. Compared with the commonly used promoter P_sarA (P31), 11 promoters displayed stronger activity. Overall, the activity of the cloned 20 promoters varied from 25 to 500% of the activity of P_sarA. Vector pQLV1003 without a promoter served as a negative control (NG), which exhibited no beta-galactosidase activity. In summary, the activity of these 30 selected promoters was evaluated in the native host, and 20 promoters showed a wide range of activity.

Using the identified 20 active promoters, 20 vectors derived from pBUS1_P_cap_HC were constructed for constitutive protein expression in S. aureus. pBUS1_P_cap_HC is a high-copy S. aureus-E. coli shuttle vector used for protein overexpression in S. aureus using the promoter of type 1 capsule gene 1A (P_cap) (Schwendener and Perreten, 2015) (Figure 3A). It contains the tetracycline marker for plasmid selection in both E. coli and S. aureus. Because some clinical S. aureus strains, such as S. aureus USA300 and S. aureus MW2, are resistant to tetracycline, an additional selection marker, the chloramphenicol cassette (cat), was amplified from pBT2 (Bruckner, 1997) and introduced into pBUS1_P_cap_HC at the BglII site, generating the vector pBUS1_P_cap_HC_cat (pQLV1002) (Figure 3B).

To generate the overexpression vectors by applying our identified constitutive promoters, the P_cap promoter sequence from the pBUS1_P_cap_HC_cat plasmid was replaced with a new DNA fragment, which contained an individual selected promoter sequence and a typical ribosomal-binding site (RBS) (Figures 3B,C). This generated a set of vectors harboring the identified strong constitutive promoter individually (Figure 3C). Downstream of the promoter, the multiple cloning site (MCS) and the RGS-6 × His coding sequence were the same as those in the original vector pBUS1_P_cap_HC_cat, which enables the expression of the target protein with a C-terminal 6 × His tag (Figure 3C). Briefly, 20 vectors carrying different constitutive promoters were developed for gene expression and the production of tagged fusion proteins.

To evaluate the effect and efficiency of the constructed expression vectors on gene expression, an egfp reporter gene was cloned into the MCS of each vector at the NdeI and XhoI sites, which allowed the egfp gene to be expressed with an RGS_6 × His tag under the control of different promoters. The verified plasmid was transformed into S. aureus USA300, and the fluorescence and OD₆₀₀ value of each resulting strain were measured in a bioreader. Promoter activity was analyzed by calculating the ratio of RLU/OD₆₀₀ (Supplementary Table 4), and a heatmap based on the RLU/OD₆₀₀ value in different growth phases was generated (Figure 4A). All tested promoters were active in LP, EP, or SP compared to the negative control, in which the egfp gene was driven by no promoter (Figure 4A).

To further quantify the strength of these vectors on gene expression, the transcriptional level of the egfp gene in each vector was quantified by RT-qPCR. The egfp gene was overexpressed in all vectors compared to the negative control (NG) (Figure 4B), which further confirmed the effect of our constructed vectors on gene expression. When compared to the commonly used promoter P_sarA (P31), nine promoters (P1, P2, P7, P8, P10, P12, P14, P19, and P25) were stronger than P_sarA (P31), while 10 promoters were weaker than P_sarA (P31) (Figure 4B). Overall, the activity of our selected promoters in the overexpression vectors varied from 18 to 650% of the activity of P_sarA (P31) (Figure 4B).

Additionally, the expression of the eGFP was confirmed by SDS-PAGE and western blotting. Overproduction of the recombinant eGFPs (28 kDa) from some strong promoters, such as P₁₀₉₃₀ (P1), P₁₀₉₃₅ (P8), and P₀₀₁₆₅ (P10), was clearly visible on the SDS-PAGE gel (Figure 4C). Owing to the eGFP being expressed with a 6 × His tag at its C-terminal, the expression of the eGFP was further confirmed by western blotting using an anti-6 × His tag antibody. The recombinant eGFP was detected in all strains at various levels (Figure 4D), most of which corresponded to their transcriptional level (Figures 4C,D). Collectively, all the overexpression vectors constructed here efficiently expressed their target genes. More importantly, this set of vectors displayed a wide range of activity in their effects on gene expression.

To evaluate the universality of the constructed expression vectors, an egfp gene was cloned into each vector and transformed into different S. aureus strains, which included S. aureus USA300, S. aureus MW2, S. aureus Newman, S. aureus RN4220, S. aureus NCTC_8325, and S. aureus ATCC 10832. As shown, the eGFP was successfully overexpressed in all vectors compared to the negative control, in which the egfp gene was regulated without a promoter (Figure 5). However, the fluorescence of eGFP expressed from the few vectors exhibited variance in different strains, such as P₁₀₉₃₅ (P8), which displayed strong promoter activity in all tested strains but not in S. aureus RN4220 (Figure 5). Importantly, most of the vectors expressed eGFPs with the same intensity in all treated S. aureus strains, which suggested that the promoters screened from S. aureus USA300 were functional in the same manner as those in other strains. These results indicated that the promoters we selected from S. aureus USA300 are conserved in S. aureus species, and the overexpression vectors constructed based on the identified active promoters were functional in different S. aureus strains with slight variance in their activity.

To further evaluate and apply our constructed overexpression system, the endogenous gene purR, which encodes a transcriptional repressor, was expressed and validated in terms of its function in S. aureus. The PurR protein has been proven to be a transcriptional repressor of purine biosynthesis that inhibits the transcription of fibronectin-binding protein-coding genes fnbA and fnbB (Goncheva et al., 2019). Three overexpression vectors carrying promoters P₀₀₁₆₅ (P10), P₀₈₆₂₀ (P25), and P₀₄₄₀₀ (P7) were randomly selected for overexpression of the purR gene. RT-qPCR analysis revealed that the purR gene was overexpressed in the three vectors compared to that in S. aureus WT (Figure 6A). The production of PurR protein (31.7 kDa) from strains overexpressing the purR gene with the P₀₀₁₆₅ (P10), P₀₈₆₂₀ (P25), or P₀₄₄₀₀ (P7) promoter was further confirmed by SDS-PAGE and western blotting (Figure 6B). The signal detected from western blotting coincided with the transcriptional level of the purR gene as quantified by RT-qPCR, with the vector harboring P₀₄₄₀₀ (P7) having the highest output, followed by P₀₈₆₂₀ (P25) and P₀₀₁₆₅ (P10). In addition, the physiological function of the overexpressed PurR protein was verified by checking the transcriptional level of its target genes fnbA and fnbB (Figure 6C). As shown, transcription of the fnbA and fnbB genes was significantly repressed in PurR overexpression strains compared to the WT (Figure 6C).

Furthermore, the endogenous catalase of S. aureus was overexpressed to further evaluate the constructed vectors. Similarly, overexpression of catalase using three vectors under the control of promoters P₀₀₁₆₅ (P10), P₀₈₆₂₀ (P25), and P₀₄₄₀₀ (P7) was confirmed by RT-qPCR and western blotting (Figures 6D,E). The results showed that the transcription of the catalase gene in the overexpression strains was 3- to 10-fold higher than that in the WT, in which P₀₀₁₆₅ (P10) possessed the highest transcription level, followed by P₀₄₄₀₀ (P7) and P₀₈₆₂₀ (P25) (Figure 6D). The same expression patterns of catalase were observed and confirmed by SDS-PAGE and western blotting (Figure 6E). In addition, the activity of catalase overproduced from different vectors was measured by H₂O₂ and ammonium molybdate-based assays. The undecomposed H₂O₂ reacts with ammonium molybdate to produce a yellowish color, i.e., more transparency of the reacted mixture represents less hydrogen peroxide left, which indicates higher catalase activity. The cell lysates from the overexpression strains exhibited significantly stronger catalase activity than those from the WT strain (Figure 6F). In line with the expression level, the vector carrying P₀₀₁₆₅ (P10) that possessed the highest expression exhibited the highest catalase activity (Figure 6F).

In conclusion, the endogenous genes purR and catalase were successfully overexpressed by using our constructed overexpression vectors, and their related functions and phenotypes were confirmed in S. aureus USA300.

The bacterial strains and plasmids used in this study are described and listed in Table 2. E. coli strains were grown in Luria-Bertani (LB) broth with constant shaking at 220 rpm or on LB agar plates at 37°C. S. aureus strains were cultured in tryptic soy broth (TSB) with shaking at 220 rpm or on TSB agar plates (TSA) at 37°C. Plasmids used for S. aureus transformation were modified by S. aureus RN4220. All S. aureus transformants were obtained through electroporation as described previously (Schenk and Laddaga, 1992). E. coli and S. aureus transformants were selected on agar plates containing 10 μg/mL tetracycline or 25 μg/mL chloramphenicol, and antibiotics were also used to maintain the plasmids in the cells.

The day culture of S. aureus USA300 was prepared by inoculating 100 μL of an overnight culture into 10 mL of TSB medium in a 100-mL flask. Two milliliters of bacterial cells were collected after 2, 6, and 10 h of cultivation by centrifugation at 5,000 g for 10 min at 4°C. Bacterial cells were resuspended in 100 μL RNase-free cell lysis buffer (20 mM Tris-Cl, pH 8.0; 2 mM sodium EDTA; 1.2% Triton X-100; 50 μg/mL lysostaphin) and incubated at 37°C for 15 min. Then, RNA extraction was performed using an RNApure Bacteria Kit (CwBIO, Jiangsu, China) following the manufacturer’s instructions. The rRNA of the RNA samples was removed with a Ribo-Zero rRNA Removal Kit (Gram-positive Bacteria, Illumina) according to the manufacturer’s instructions. cDNA library preparation and sequencing were performed by Personalbio Co. (Shanghai, China).

The isolation of genomic DNA and plasmid preparation using E. coli or S. aureus were performed with a bacterial genomic DNA kit and PurePlasmid Mini Kit, respectively (CwBIO, Jiangsu, China). Oligonucleotide primers were synthesized by Sangon Biotech (Shanghai, China). All primers used in this study are listed in Supplementary Table 2. For analytical purposes, PCRs were performed using OneTaq 2 × Master Mix (NEB, Ipswich, England). PCRs for plasmid construction were performed using Q5 High-Fidelity 2 × Master Mix from NEB according to the manufacturers’ instructions. The PCR products were purified using the DNA Clean-up Kit (CwBIO, Jiangsu, China). For cloning or plasmid construction, the plasmid was linearized by the related restriction enzymes from NEB. Cloning was performed using the ClonExpress II One Step Cloning Kit (Vazyme, Nanjing, China), which is based on homologous recombination.

An entire chloramphenicol resistance cassette was amplified from plasmid pBT2 (Bruckner, 1997) using primers QL0230/0231 (Supplementary Table 2) and cloned into S. aureus–E. coli shuttle vectors pBUS1-HC and pBUS1-Pcap-HC (Schwendener and Perreten, 2015) at the BglII site, which generated vectors pQLV1001 and pQLV1002, respectively. To generate a promoterless beta-galactosidase reporter vector, a DNA fragment containing the red fluorescence protein gene (rfp) and beta-galactosidase gene (lacZ) was amplified from the pBS1C_lacZ plasmid (Popp et al., 2017) with primers QL409/0410 (Supplementary Table 2). The rfp_lacZ fragment was inserted into pQLV1001 linearized with HindIII and BamHI to generate the lacZ promoter-probe vector pQLV1003.

The promoter regions of highly expressed genes (Supplementary Table 3) were amplified from the genomic DNA of S. aureus USA300 by PCR using the primers listed in Supplementary Table 2. Then, the promoter DNA fragment was individually cloned into the lacZ promoter-probe vector pQLV1003, which was linearized with EcoRI and PstI. After confirmation by PCR and sequencing, the generated plasmid was modified by RN4220 and subsequently transformed into S. aureus USA300 for the beta-galactosidase-based promoter activity assay.

The activity of the cloned promoters was first evaluated on the plate. Briefly, day cultures of S. aureus USA300 strains harboring the plasmid constructed above were prepared by inoculating 100 μL of the overnight culture into 10 mL of TSB medium. After the OD₆₀₀ value reached 0.6, 5 μL of the bacterial culture was pipetted out and dropped onto a TSA agar plate containing 25 μg/mL chloramphenicol and 200 μg/mL X-Gal (5-bromo-4-chloro-3-indolyl-β-D-galactoside). The plate was incubated at 30°C, and the color of the colony was monitored and photographed after 48 h.

To further quantify the promoter activity, analysis of beta-galactosidase activity using O-nitrophenyl-β-D-galactopyranoside (ONPG) as substrate was performed and modified as previously described (Zhu et al., 2020). A single blue colony on the TSA X-Gal plate was inoculated into chloramphenicol-containing TSB medium overnight culture preparation. The next day, the day culture was prepared as described above. After 2, 6, or 10 h of cultivation, 100 μL of the bacterial culture was collected for OD₆₀₀ value measurement in a 96-well plate by using a Synergy H1 microplate reader (Vermont, United States).

Meanwhile, bacterial cells were harvested from the 200 μL culture at each time point by centrifugation at 5,000 g for 10 min at 4°C. The cells were resuspended in 100 μL of lysis buffer (60 mM K₂HPO₄, 40 mM KH₂PO₄, 100 mM NaCl, 0.1% Triton X-100, 50 μg/mL lysostaphin) for total protein extraction. The suspension was incubated at 37°C until the bacterial cells had completely lysed. Then, 50 μL of the lysate was pipetted onto a 96-well plate for OD₄₂₀ measurement. After that, 100 μL of a reaction buffer (60 mM K₂HPO₄, 40 mM KH₂PO₄, 100 mM NaCl, 0.1% Triton X-100, 5 mg/mL ONPG) was added to each well and mixed with the cell lysate. The 96-well plate was incubated in the Bioreader at 37°C with constant shaking, and the OD₄₂₀ was measured every 5 min. The slope of the linear part of the spectrophotometric output was used to calculate the specific activity as follows: Miller units = (1,000 × slope)/(V × OD₆₀₀), where V is the volume (in milliliters) of the culture used in the assay. This assay was repeated three times for each strain.

The promoter sequences of the validated active sequences (Supplementary Table 3) were amplified by PCR from the pQLV1003-derived plasmids carrying different promoters using the primers listed in Supplementary Table 2. A fragment “AGGAGGTTTATCATATG” that contained a typical RBS (underlined), a spacer, and an NdeI site (italic) was introduced downstream of each promoter from the primer. This will lead the target gene to be cloned at the NdeI site and other restriction enzyme sites downstream of NdeI. The amplified promoter was cloned into vectors pQLV1002 linearized with KpnI and NdeI to generate a series of overexpression vectors.

To evaluate the constructed expression vectors, the egfp reporter gene was PCR amplified from plasmid pTH100 (de Jong et al., 2017) with primer pairs QL0558/0559 and inserted into each constructed expression vector that was linearized with NdeI and XhoI. The endogenous transcriptional repressor purR or catalase gene of S. aureus USA300 was PCR amplified with the primer pairs QL0609/0610 and QL0613/0614, respectively. These were cloned into the overexpression vectors at the NdeI and XhoI sites for PurR and catalase expression with an RGS_6 × His tag at their C-termini.

A day-culture was prepared as described above. Bacterial cells from the 200 μL day-culture after 2, 6, and 10 h of inoculation were harvested by centrifugation. Cells were washed twice with 500 μL of PBS buffer and suspended in PBS to an OD₆₀₀ value of approximately 0.3. The OD₆₀₀ value and relative light unit (RLU) of GFP fluorescence (excitation, 485 nm; emission, 512 nm) were measured using a Synergy H1 plate reader. Promoter activity was analyzed by calculating the ratio of RLU/OD₆₀₀. The assay for each strain was performed in three independent experiments, and the mean value of RLU/OD₆₀₀ was used to generate the heatmap.

The qPCR was performed in a CFX-96 Touch Real-Time PCR system (Bio-Rad, Hercules, CA, United States) using SYBR green master mix TB Green Premix Ex Ta II (Takara, Beijing, China). Total RNA from S. aureus was isolated as described above. Approximately 1 μg of total RNA was used for reverse transcription using a PrimeScript™ RT reagent kit with gDNA Eraser (Takara, Beijing, China) according to the manufacturer’s instructions. After the transcribed cDNAs were diluted fivefold, 2 μL of the cDNA was used as DNA template in 15-μL amplification volumes with 400 nM of each primer and 7.5 μL of SYBR green master mix using the following cycling parameters: 95°C 30 s; followed by 40 cycles of 5 s at 95°C, 30 s at 55°C, and 30 s at 72°C. The primer pairs QL0615/QL0616, QL0619/QL0620, QL0633/QL0634, QL0635/QL0636, QL0637/QL0638, and QL0645/QL0646 were used to amplify the egfp, purR, catalase, fnbA, fnbB, and sigA genes, respectively, by qPCR. The endogenous gene sigA, which encodes the housekeeping RNA polymerase sigma factor, was used as an internal control for promoter characterization. The expression levels of the egfp gene under different promoters were normalized to the expression of the internal control.

Total protein extracts of S. aureus were prepared as described in the beta-galactosidase assay. The protein concentration was measured using a Bradford assay kit (Thermo Scientific, Waltham, MA, United States). Twenty micrograms of the total protein extract was separated by 12% SDS-PAGE and transferred onto a nitrocellulose membrane (Bio-Rad). The membrane was blocked with TBS-T buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% [vol/vol] Tween-20) containing 5% non-fat dry milk at room temperature (RT) for 1 h. Detection of the target proteins eGFP_6 × His, PurR_6 × His, and catalase_6 × His overexpressed in S. aureus was performed by incubating the membrane with the horseradish peroxidase (HRP)-labeled His-Tag (27E8) mouse mAb (CST, Danvers, MA, United States) (diluted 1:2,000 in TBS-T buffer) at RT for 1 h. After four washing steps (10 min per step) with TBS-T buffer, the signals were detected with a ChemiDoc MP (Bio-Rad) imaging system using Pierce ECL (Thermo Scientific, Waltham, MA, United States) as a chemiluminescence substrate.

Bacterial cells from the 200 μL day-culture were collected in the middle of the lag growth phase (6 h after inoculation) by centrifugation. The precipitated cells were resuspended and lysed in 800 μL of phosphate-buffered saline (PBS, pH 7.4) by sonication on a SCIENTZ-IID system (SCIENTZ, Ningbo, China) according to the user guide. The protein concentration of the cell lysate was measured as described above. To measure the activity of catalase in the lysate, 20 μL of the cell lysate was mixed with 100 μL of 200 mM H₂O₂ (diluted with PBS) and incubated at 25°C for 10 min. Then, the reaction was stopped by adding 180 μL of ammonium molybdate (50 mM in H₂O) into the mixture. After incubation at room temperature for 10 min, the absorbance of the yellow complex of molybdate and undecomposed hydrogen peroxide was measured at a wavelength of 405 nm in a 96-well plate (A_Sample) using the plate reader Synergy H1 (BioTek, Vermont, VT, United States). One unit of catalase activity was defined as 1 μM of H₂O₂ catalyzed and hydrolyzed by 1 mg lysate in 1 min at 25°C. Thereby, the catalase activity of the lysate (CAT) was calculated as follows: CAT (U/mg) = [(ΔA_Standard − ΔA_Sample)/ΔA_Standard] × N/m/T × F, where ΔA_Standard = A_Standard − A_Blank and ΔA_Sample = A_Sample − A_Control. A_Control is the absorbance of the control group, in which H₂O₂ was replaced by PBS. A_Standard is the absorbance of the standard group, in which the cell lysate was replaced by PBS, while A_Blank is the absorbance of the blank group, in which cell lysate and H₂O₂ were both replaced by PBS. N represents the molar mass of H₂O₂ (micromole) used in the assay, and m is the mass of the cell lysate (mg) used in the assay. T is the reaction time in minutes, and F is the dilution factor of the cell lysate. The assay for each strain was performed in three independent experiments.