E. chaffeensis Arkansas strain was propagated in THP-1 cells in RPMI 1640 medium supplemented with 2 mM _L-glutamine and 10% fetal bovine serum (FBS) (Every Green, Zhejiang, China) at 37°C in 5% CO₂ and 95% air, as described previously (Cheng et al., 2006). The THP-1 cell line present in this study was obtained from Fuheng Biotechnology (Cat. number FH0112, Shanghai, China).

Escherichia coli strains ( Supplementary Table S1 ) DH5α and BL21 (DE3) used for general cloning and protein expression, respectively, were cultured in LB broth supplemented with ampicillin (100 μg·mL^-1), chloramphenicol (34 μg·mL^-1) or kanamycin (50 µg·mL^-1), as necessary.

The DNA fragment encoding full-length CtrA was cloned into pET-41a(+) to express recombinant CtrA (GST-rCtrA) as described previously (Cheng et al., 2006). The DNA fragments encoding GshA, GshB and GST were amplified using specific primers ( Supplementary Table S1 ). The gshA fragment was cloned into pET-His-SUMO to express recombinant GshA (SUMO-rGshA) with an N-terminal His-tag. The fragment of gshB or gst was cloned into pET-33b(+) to express recombinant GshB (rGshB) or GST (rGST) with an N-terminal His-tag, respectively. The constructed plasmids were transformed into E. coli DH5α cells, extracted and confirmed by sequencing. E. coli BL21 (DE3) cells were transformed with the resulting plasmids to express SUMO-rGshA, rGshB and rGST respectively, with pET-41a(+) to express GST or pET-His-SUMO to express SUMO. E. coli BL21 (DE3) cells were induced to express recombinant proteins at an OD₆₀₀ of 0.6 with 0.1 mM isopropyl-thio-β-D-galactoside (IPTG; Solarbio, Beijing, China) at 37°C for 4 h. All proteins were purified from soluble fraction with Ni-affinity chromatography and concentrated using an Amicon Ultra-0.5 Centrifugal Filter Unit (10 kDa MWCO).

Fixed cell sample was prepared from 2.5 × 10⁷ infected THP-1 cells (> 90% infected) following the instructions for suspension cell sample preparation from IGENEBOOK Biotechnology (Wuhan, China). The ChIP-seq experiments and data analyses were performed by IGENEBOOK Biotechnology. Briefly, cell culture was crosslinked with 1% (w/v) formaldehyde and incubated at room temperature for 10 min. Crosslinking was terminated by adding 125 mM _L-Gly. Crosslinked cells were washed twice with 1 × PBS at 4°C, pelleted, frozen in liquid nitrogen and subjected to ChIP-seq. Rabbit antiserum against E. chaffeensis CtrA (Yan et al., 2022) was used for immunoprecipitation of E. chaffeensis chromosomal DNA fragments bound by CtrA. Total DNA before immunoprecipitation was used as input DNA. Fold enrichment was expressed as immunoprecipitation signal normalized against input signal.

Electrophoresis mobility shift assay (EMSA) was performed as described previously (Cheng et al., 2011). Briefly, the promoter regions of gst (305 bp) and p28 (260 bp) were amplified with PCR using specific primers ( Supplementary Table S1 ). Fifty picomoles of purified GST-rCtrA or GST (negative control) was incubated with 50 ng of DNA probes in a 20-μL reaction mixture containing 10 mM Tris-HCl (pH 7.5), 1 mM DTT and 1% (w/v) _L-Gly on ice for 30 min. Samples were loaded onto 8% native polyacrylamide gel in 1 × TBE buffer [89 mM Tris (pH 8.3), 89 mM boric acid and 2 mM EDTA], which had been prerun at 100 V for 1 h, and electrophoresed at 10 mA on ice for 1.5 h. The gel was stained in 1 × TBE containing 0.5 μg·mL^-1 ethidium bromide at room temperature for 10 min. Bands were visualized using a ChemiDoc XRS+ molecular imager (Bio-Rad).

Enhanced green fluorescent protein (EGFP) fusions were constructed as described previously (Duan et al., 2021). Briefly, the promoter region of gst or p28 was amplified and inserted upstream of the promoter-less egfp gene in a pQE60 vector. E. coli BL21 (DE3) cells containing pACYCDuet-1 harboring ctrA (pACYCDuet-1-rCtrA) or the empty pACYCDuet-1 vector were transformed with the pQE60-promoter-EGFP fusion constructs. After induction of rCtrA expression with 0.05 mM IPTG at 37°C for 3 h, bacterial samples were collected and subjected to Western blotting to measure the amounts of EGFP, RpoA, and rCtrA in each sample.

Western blotting was performed as described previously (Yan et al., 2022). Briefly, to detect the amount of EGFP in the reporter assay, the same numbers of E. coli cells were collected by centrifugation at 12000 × g for 1 min, resuspended in 1 × SDS sample buffer [62.5 mM Tris-HCl (pH 6.8), 2% (w/v) SDS, 0.02% (w/v) bromophenol blue, 1% (v/v) β-mercaptoethanol and 10% (v/v) glycerol] and boiled for 10 min. The samples were then subjected to 12% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a PVDF membrane (Millipore, Co. Cork, Ireland). The protein levels of EGFP, RpoA, and rCtrA were determined using a mouse monoclonal anti-GFP antibody (GeneTex Inc., North America, GTX628528), mouse monoclonal anti-E. coli RNA polymerase α antibody (BioLegend, San Diego, CA, United States, 663104), and rabbit antiserum against E. chaffeensis CtrA (Yan et al., 2022), respectively. The relative amounts of EGFP or RpoA in E. coli BL21 (DE3) cells expressing rCtrA were normalized against those in E. coli BL21 (DE3) cells containing pACYCDuet-1 vector.

E. chaffeensis was isolated from infected THP-1 cells as described previously (Yan et al., 2022). Briefly, E. chaffeensis-infected THP-1 cells (2 × 10⁷ cells, > 90% infected) were harvested at 600 × g at room temperature for 5 min. The pellet was suspended in fresh culture medium and passed through a 23-gauge needle with a syringe on ice for 20 times to rupture the host cell membrane. To remove unbroken cells and cell debris, the mixture was centrifuged at 1,000 × g at 4°C for 5 min. Bacteria in supernatant were collected by additional centrifugation at 10,000 × g at 4°C for 10 min. The bacterial pellet was suspended in fresh culture medium for synchronous culture or in 0.3 M sucrose for PNA transfection.

Total RNA was extracted from each sample and reverse transcribed to cDNA as described previously (Duan et al., 2021). The amounts of E. chaffeensis 16S rRNA, ctrA, gst, gshA, gshB, and human GAPDH were determined with qRT-PCR using specific primers ( Supplementary Table S1 ) and the ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China) on a StepOnePlus Real-Time PCR System (Applied Biosystems, MA, USA). The expression levels of bacterial genes were normalized against that of E. chaffeensis 16S rRNA. Relative bacterial number was determined as the amount of bacterial 16S rRNA normalized against that of human GAPDH mRNA.

The activity of rGST was determined as described previously with minor modifications (Habig and Jakoby, 1981). Briefly, the reaction mixture was composed of 100 mM Na₃PO₄ (pH 6.5), 1 mM 2,4-dinitrochlorobenzene (CDNB; Xiya Reagent, Shandong, China) and varying concentrations of GSH. Reactions were initiated by adding 3 μg of purified rGST or SUMO as a negative control in a final volume of 200 μL, then incubated at 37°C for 30 min. The formation of GSH-CDNB conjugate was measured by OD₃₄₀.

The activity of SUMO-rGshA was determined as described previously with minor modifications (Tu and Anders, 1998). Briefly, the reaction mixture was composed of 100 mM Tris-HCl (pH 8.0), 50 mM KCl, 20 mM MgCl₂, 2 mM EDTA, 5 mM ATP, 10 mM _L-α-aminobutyric acid and varying concentrations of monosodium _L-glutamate. Reactions were initiated by adding 1.25 μg of purified SUMO-rGshA or SUMO as a negative control in a final volume of 80 μL, then incubated at 37°C for 30 min. Twenty microliters of 5% (w/v) SDS were added into each reaction mixture to terminate the reaction and the concentration of inorganic phosphate (P_i) generated by the reaction ( Figure S1 ) was determined as described previously with minor modifications (Satoh et al., 2006). Briefly, 100 μL of reagent mixture containing 24 mg·mL^-1 ascorbic acid in (NH₄)₂MoO₄-H₂SO₄ solution [1.145 g (NH₄)₂MoO₄ and 5.95 mL H₂SO₄ in 250 mL of water] was added to each reaction mixture. Reaction mixtures were incubated at room temperature for 2 h and concentrations of P_i were measured by OD₇₅₀ using Varioskan Flash (Thermo Fisher Scientific). To calibrate the P_i concentration, 80 μL of increasing concentrations of KH₂PO₄ from 0 to 1,000 μM was mixed with 20 μL 5% (w/v) SDS and subjected to P_i determination as described above to generate a standard curve for each experiment.

The activity of rGshB was determined as described previously with minor modifications (Oppenheimer et al., 1979). Briefly, the reaction mixture was composed of 100 mM Tris-HCl (pH 8.2), 50 mM KCl, 20 mM MgCl₂, 2 mM EDTA, 10 mM Na₃ATP, 2.5 mM DTT, 5 mM _L-γ-_L-glutamyl-_L-cysteine (γ-GC) and varying concentrations of _L-Gly. Reactions were initiated by adding 0.75 μg of purified rGshB or SUMO as a negative control in a final volume of 80 μL, then incubated at 37°C for 30 min. The concentration of P_i generated by these reactions ( Figure S1 ) was measured as described above.

PNA transfection to knock down CtrA, GST, GshA or GshB in E. chaffeensis was performed as described previously (Yan et al., 2021). Antisense PNA oligomers targeting 31-46 bp following the start codon of ctrA (CtrA PNA), 10-23 bp following the start codon of gst (GST PNA), 13-28 bp following the start codon of gshA (GshA PNA) and 19-34 bp following the start codon of gshB (GshB PNA) and a control PNA with a scrambled sequence (CTL PNA) were synthesized by KareBay Biochem, Inc. (Ningbo, China; Supplementary Table S1 ). The knockdown efficiency of CtrA PNA has been verified previously (Yan et al., 2022). To verify the specific binding of PNAs to their targets, ssRNAs derived from gst (GST ssRNA, 3-33 bp following the start codon of gst mRNA), gshA (GshA ssRNA, 6-37 bp following the start codon of gshA mRNA) and gshB (GshB ssRNA, 12-43 bp following the start codon of gshB mRNA) were synthesized by Genewiz (Suzhou, China; Supplementary Table S1 ). Hybridization of ssRNA with PNA oligomer was performed as described with minor modifications (Yan et al., 2018). Briefly, 80 pmol of ssRNA and 80 pmol of gene-specific PNA or CTL PNA were added into hybridization buffer containing 250 mM Tris-HCl (pH 7.2) in a final volume of 15 μL. Hybridization was performed on a BIO-GENER GE4852T thermocycler (Hangzhou, China) with the following program: 95°C for 3 min, then gradual cooling to 20°C over 75 cycles (20 s per cycle), 20°C for 1 min. Hybridization mixtures were then subjected to 12% native polyacrylamide gel electrophoresis and bands were visualized by GelGreen nucleic acid staining (Biomed, Beijing, China).

Three micrograms of CtrA PNA, GST PNA, GshA PNA, GshB PNA or CTL PNA dissolved in nuclease-free water were mixed with 100 μL of host cell-free E. chaffeensis in 0.3 M sucrose, then incubated on ice for 15 min. Electroporation was conducted at 2,000 V, 25 μF, and 400 Ω with a 10-ms pulse using a Gene Pulser Xcell electroporation system (Bio-Rad, Hercules, CA, United States) in a 2-mm electroporation cuvette (Bio-Rad). Then the PNA-transfected E. chaffeensis was transferred to a T25 flask to infect 5 × 10⁵ THP-1 cells and incubated at 37°C for 2 h with gentle shaking every 15 min to facilitate bacterial internalization. To detect the effect of GST PNA, GshA PNA and GshB PNA transfection, infected cells were harvested at 36 hours post infection (h p.i.). The expression levels of gst, gshA and gshB were determined using qRT-PCR.

E. chaffeensis was treated with H₂O₂ as described previously (Yen et al., 2020). THP-1 cells were synchronously infected with CtrA PNA-, GST PNA-, GshA PNA-, GshB PNA- or CTL PNA-transfected E. chaffeensis. At 36 h p.i., 20 μL of 10 mM H₂O₂ diluted in culture medium was added into 2 mL of the synchronous culture, reaching a final concentration of 100 μM. After 2 h of treatment at 37°C, the bacterial number in each sample was determined with qRT-PCR.

CtrA ChIP-seq assay was performed for a single sample and ChIP-seq data were analyzed by IGENEBOOK Biotechnology (Wuhan, China). All other experiments were performed at least three times. Statistical analyses were performed using GraphPad Prism 7.0. Statistical significance of a two-group comparison was assessed using Student’s t test (two-tailed). A value of P < 0.05 was considered significant.

To identify the genes regulated by CtrA in E. chaffeensis, we first performed ChIP-seq assay. Using rabbit antiserum against E. chaffeensis CtrA, we identified that the promoter regions of 211 genes were enriched by more than 1.5-fold (P-value < 0.01) compared to input DNA ( Supplementary Table S2 ). Among these genes, 69 were annotated and 64 were successfully classified into Cluster of Orthologous Group (COG) functional categories (E-value < 0.01). COG analysis revealed that these genes were involved in coenzyme transport and metabolism, energy production and conversion, replication, recombination, and repair, and various other functions ( Figure 1 ). Consistent with the previous studies (Cheng et al., 2011; Yan et al., 2022), the promoter regions of ctrA, bolA, surE and gshB were enriched by 1.64-, 1.72-, 1.61- and 1.71-fold, respectively. The promoter regions of gshA and ompA were also enriched by 1.45- and 1.30-fold respectively ( Supplementary table S2 ).

Among these enriched promoter regions, notably, the gst promoter region was enriched 1.81-fold compared to input DNA ( Supplementary Table S2 ). GST is an enzyme that is able to catalyze the reaction of GSH with ROS, then eliminate ROS (Skopelitou et al., 2012a). The amino acid sequence of E. chaffeensis GST showed 34.35% identity with that of Agrobacterium tumefaciens GST (GenBank ID: AAK86108.1) ( Figure S2A ) (Skopelitou et al., 2012b). We found a 9-mer CtrA binding motif (TTAA-N₇-TTAAC) in the gst promoter region (-181 to -166 calculated from the translational start codon) using pDRAW32 DNA analysis software (http://acaclone.com/) ( Figure S2B ). These results suggest that CtrA regulates the gst expression in E. chaffeensis.

We then investigated whether the gst expression was regulated by CtrA. CtrA is expressed at the late stage of E. chaffeensis intracellular growth (Cheng et al., 2011; Yan et al., 2022). Using qRT-PCR, we examined the gst expression pattern in synchronously cultured E. chaffeensis in THP-1 cells. We found that the gst expression was also upregulated at the late stage of E. chaffeensis intracellular growth ( Figure 2A ), indicating that the gst expression may be regulated by CtrA.

GshA and GshB are glutamate-cysteine ligase and glutathione synthetase respectively that catalyze GSH synthesis from _L-Glu, _L-Gly, _L-Cys and ATP ( Figure S1 ). Previously we found that CtrA activates the expression of gshA and gshB in E. chaffeensis upon oxidative stress (Yan et al., 2022). We then investigated whether the gst expression was activated by CtrA upon oxidative stress. PNA is a DNA mimic that has been shown to bind single- and double-stranded DNA and RNA with high affinity and specificity (Pelc et al., 2015), and can be used to knock down the levels of E. chaffeensis proteins (Sharma et al., 2017; Yan et al., 2018; Yan et al., 2021; Yan et al., 2022). Previously we used PNA to specifically knock down CtrA level in E. chaffeensis (Yan et al., 2022). Host cell-free E. chaffeensis was transfected with CtrA PNA or CTL PNA, then used to infect THP-1 cells. At 36 h p.i. the infected THP-1 cells were treated with 100 μM H₂O₂ at 37°C for 2 h. We found that the gst expression in E. chaffeensis transfected with CTL PNA was significantly upregulated after H₂O₂ treatment, suggesting that GST may be required for conferring oxidative stress resistance to E. chaffeensis ( Figure 2B ). After CtrA level was knocked down by CtrA PNA, the gst expression showed no changes after H₂O₂ treatment ( Figure 2B ), indicating that the gst expression upon oxidative stress is activated by CtrA.

We then performed EMSA using GST-rCtrA to investigate whether CtrA directly binds to the gst promoter. Purified GST-rCtrA showed a single band on the SDS-PAGE gel ( Figure S3 ). The DNA probe derived from the gst promoter region shifted upon incubation with GST-rCtrA ( Figure 2C ). The binding specificity was confirmed by the result that no shifted band was detected when the DNA probe was incubated with purified GST protein or when a DNA probe derived from the p28 promoter region, which is not bound by CtrA, was incubated with GST-rCtrA ( Figure 2C ). These results indicate that CtrA directly binds to the gst promoter in E. chaffeensis.

Due to the lack of effective genetic manipulation methods in E. chaffeensis, in order to investigate whether CtrA activates the gst expression, we used an E. coli reporter assay system (Cheng et al., 2011; Duan et al., 2021; Yan et al., 2022). The gst promoter region was inserted upstream of the promoter-less egfp gene in the pQE60 plasmid to generate gst promoter-EGFP fusion construct. The p28 promoter region was used to generate p28 promoter-EGFP fusion construct as a negative control. E. coli BL21 (DE3) cells containing pACYCDuet-1 vector harboring E. chaffeensis ctrA gene (pACYCDuet-1-rCtrA) or pACYCDuet-1 vector only (negative control) were transformed with the EGFP fusion constructs, respectively. The rCtrA expression induced by IPTG resulted in a significantly higher expression of EGFP in bacteria harboring gst promoter-EGFP construct compared to the vector control, while it had no effect on bacteria harboring p28 promoter-EGFP construct ( Figure 2D ; Figure S4 ). These results indicate that CtrA activates the gst expression in E. chaffeensis.

We next determined the GST function in vitro. The purified rGST and the negative control SUMO, which was purified using the same method, showed single bands on the SDS-PAGE gel ( Figure S5 ). The enzymatic activity of rGST was measured by GSH-CDNB conjugate formation ( Figure 3A ). The K _m value of rGST was determined for GSH. The K _m and V _max values of the enzymes are listed in Table 1 . The results showed that E. chaffeensis rGST was functional and its K _m value for GSH was found to be in a similar range with that of A. tumefaciens GST reported previously (Skopelitou et al., 2012a), indicating that GST is a functional enzyme in E. chaffeensis that catalyzes the conjugation of GSH to its typical substrate CDNB.

In our previous study, we verified the functions of E. chaffeensis GshA and GshB by complementing Pseudomonas aeruginosa respective mutants (Yan et al., 2022). Here, we also determined the functions of E. chaffeensis GshA and GshB in vitro. Since rGshA was unable to be expressed efficiently in E. coli BL21 (DE3) cells, we constructed a plasmid to express SUMO-fused recombinant GshA (SUMO-rGshA). The purified SUMO-rGshA and rGshB showed single bands on the SDS-PAGE gel ( Figure S5 ). The enzymatic activities of SUMO-rGshA and rGshB were measured based on P_i formation ( Figures 3B, C ). The K _m value of SUMO-rGshA or rGshB was determined for _L-Glu or _L-Gly respectively. The K _m and V _max of the enzymes are listed in Table 1 . The results showed that SUMO-rGshA was functional and the K _m value for _L-Glu was found to be in a similar range with that of A. tumefaciens GshA (Gromes et al., 2008). Also, rGshB was functional and the K _m value for _L-Gly was found to be in a similar range with that of E. coli GshB (Tanaka et al., 1997). These results indicate that GshA and GshB are functional enzymes in E. chaffeensis, which catalyze the synthesis of GSH.

Since E. chaffeensis GST was functional in vitro, we then investigated its function in vivo. We designed GST PNA that specifically binds to the region following the translation start codon of gst and confirmed the specificity by showing that GST PNA bound to the GST ssRNA (synthesized gst mRNA fragment, 3-33 bp following the start codon) ( Figure 4 ). Transfection of host cell-free E. chaffeensis with GST PNA significantly reduced the gst mRNA level at 36 h p.i. ( Figure 4 ). We then investigated whether the reduction of the gst expression influenced the bacterial survival upon oxidative stress. At 36 h p.i. the infected THP-1 cells were treated with 100 μM H₂O₂ at 37 °C for 2 h. The survival ability of E. chaffeensis transfected with GST PNA was significantly lower than that of E. chaffeensis transfected CTL PNA with H₂O₂ treatment ( Figure 5A ), indicating that GST is critical for E. chaffeensis to cope with oxidative stress.

We also determined the survival ability of E. chaffeensis transfected with GshA PNA or GshB PNA upon oxidative stress. We designed GshA PNA and GshB PNA and confirmed the specificity by showing that both GshA PNA and GshB PNA bound to their corresponding ssRNA (synthesized gshA or gshB mRNA fragment, 6-37 bp or 12-43 bp following the start codon, respectively) ( Figure 4 ). Transfection of host cell-free E. chaffeensis with GshA PNA or GshB PNA significantly reduced the mRNA level of gshA or gshB at 36 h p.i. ( Figure 4 ). At 36 h p.i. the infected THP-1 cells were treated with H₂O₂ as described above. The survival ability of E. chaffeensis transfected with GshA PNA or GshB PNA was also significantly lower than that of E. chaffeensis transfected CTL PNA ( Figures 5B, C ). All these results together indicate that GSH synthesis and utilization are critical for E. chaffeensis to cope with oxidative stress.