Strains and plasmids used in this study are listed in Supplementary Table 1. Aerobic growth was achieved by shaking in air at 180 rpm at 37°C, and anaerobic growth by incubating in a sealed jar with MGC AnaeroPack pouches (Mitsubishi Gas Chemical Company, Japan). M9(gly) minimal medium contains glycerol (0.25% v/v) as an energy substrate, supplemented with M9 salts, 2 mM MgSO₄, 0.1 mM CaCl₂, and 1 mg/ml vitamin B1. Trimethylamine N oxide (TMAO, 20 mM) was added as electron acceptor during anaerobic growth, and KG was present at 20 mM. For genetic manipulations, all E. coli strains were grown routinely in lysogenic broth (LB). Selective antibiotics were added when necessary at the following concentrations: ampicillin (Amp), 100 μg ml^–1; kanamycin (Kan), 50 μg ml^–1; chloramphenicol (Chl), 25 μg ml^–1.

Polymerase chain reaction (PCR), DNA ligation, electroporation and DNA gel electrophoresis were performed according to Sambrook and Russell (2001) unless otherwise indicated. DNA sequencing services were provided by Comate Bioscience Company (China). All restriction and DNA-modifying enzymes were purchased from New England Biolabs or Thermo Fisher Scientific, and used based on the suppliers’ recommendations. Recombinant plasmids, PCR products, and restriction fragments were purified using MiniBEST DNA Fragment purification kit or MinElute gel extraction kit (Takara) as recommended by the supplier. DNA and amino acid sequence analyses were performed using CloneManager software (Scientific & Educational Software, NC). Chromosomal transcriptional lacZ fusion was constructed by homologous recombination of the suicidal plasmid pVIK112 carrying a fragment of complete 3′-region or internal fragment of the kguR gene (Cai et al., 2013). All oligonucleotides used are listed in Supplementary Table 2.

RNA-seq analysis was performed using a standard protocol with minor modifications (Sheehan et al., 2020; Li et al., 2021). Wild-type CFT073 and its ΔkguR mutant were grown anaerobically in M9(gly) in the presence of KG. After a growth period of 18 h at 37°C, bacteria were quickly spun down at 10,000 g for 2 min, followed by adding 4 mL of RNALater (Thermo Fisher Scientific) to stabilize the bacterial pellets. Total RNA of each sample was extracted using TRIzol Reagent (Invitrogen), and possible DNA contamination was removed with a TURBO DNA-free kit (Thermo Fisher Scientific). One microgram of high-quality RNA (A260/A280 ratio > 2.0 and RIN value > 7.0) was used for each NextGen sequencing library, which was constructed according to the manufacturer’s protocol (NEBNext Ultra Directional RNA Library Prep Kit for Illumina). Sequencing of the libraries was performed using a 2 × 150 paired-end (PE) configuration on an Illumina HiSeq platform according to the manufacturer’s instructions (Illumina, CA). Reads were processed by Cutadapt (v1.9.1) to remove adapter sequences and to discard reads with quality scores < 20 and reads < 75 nt after trimming. Clean reads were aligned to the CFT073 genome (GenBank accession: NC_004431.1) using bowtie2 (version v2.1.9, standard options). Reads were counted using HTseq (version V 0.6.1). Differential gene expression analysis was then performed using DESeq2 (V1.6.3) with R version 3.3.2 following a standard workflow. All genes with a |log₂(Fold-change)| > 1 and a Benjamini-Hochberg adjusted p-value (q-value) < 0.05 were considered differentially expressed (differentially expressed genes, DEGs). Supplementary Table 3 lists all DEGs in the mutant compared to the wild type, and raw data are available at the National Microbiology Data Center (NMDC40014023 to NMDC40014028).

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed by assigning KEGG pathways to each DEG. The enrichment factor was calculated as the ratio of the number of DEGs enriched in the pathway to the number of all genes in the background gene set. The top 19 enriched pathways relevant to this study are shown in the figure.

Bacterial cultures were prepared as described in the RNA-seq section. Bacterial pellets were frozen in liquid nitrogen for 15 min and stored in −80°C until further metabolites extraction. For metabolite extraction, bacterial pellets were resuspended in prechilled methanol containing 0.1% formic acid, followed by an incubation of 5 min on ice. These solutions were then centrifuged at 15,000 rpm for 15 min at 4°C. The supernatants containing extracted metabolites were harvested and filtered through a 0.22 μm filter, and subsequently subjected to liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) analysis.

LC-MS/MS analyses were performed on a Vanquish UHPLC system (Thermo Fisher Scientific) coupled with an Orbitrap Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific). For chromatographic analysis, a C18 Hyperil Gold reversed-phase column (2.1 mm × 100 mm, 1.9 μm, Thermo Scientific, United States) preheated at 40°C was selected, with a 16-min gradient at a flow rate of 0.2 mL/min. The dual-eluent for the positive polarity mode were 0.1% formic acid in H₂O (eluent A) and methanol (eluent B), respectively; on the other hand, 5 mM ammonium acetate at pH 9.0 (eluent A) and methanol (eluent B), respectively, for the negative polarity mode. The dual-solvent gradient program was set as follows: 2% B, 1.5 min; 2–100% B, 1.5–12 min; 100% B, 12–14 min; 100–2% B, 14–14.1 min; 2% B, 14.1–16 min. Q Exactive HF-X mass spectrometer was operated in positive/negative polarity mode with spray voltage of 3.2 kV, capillary temperature of 320°C, sheath gas flow rate of 35 arb and aux gas flow rate of 10 arb.

Raw data generated by UHPLC-MS/MS were analyzed by Compound Discoverer 3.0 (CD 3.0, Thermo Fisher) for peak alignment, peak picking, and quantitation of each metabolite. The main parameters were set as follows: retention time tolerance, 0.2 min; actual mass tolerance, 5 ppm; signal intensity tolerance, 30%; signal/noise ratio, 3; and minimum intensity, 100,000. Then, peak intensities were normalized to the total spectral intensity, and the normalized data were used to predict the molecular formula according to additive ions, molecular ion peaks and fragment ions. Finally, peaks were matched against the mzCloud¹ and ChemSpider² database for accurate qualitative and relative quantitative results (Li et al., 2017).

RNA isolation was performed as described above, and reverse transcription of RNA was done using a HiScript II 1st Strand cDNA Synthesis Kit (Vazyme, China). Melting curve analyses were performed after each reaction to ensure amplification specificity. Differences (n-fold) in transcripts were calculated using the relative comparison method, and amplification efficacies of each primer set were verified as described by Schmittgen et al. (2000). RNA levels were normalized using the housekeeping gene rpoB as an endogenous control (Skyberg et al., 2008). qPCR was performed with an Applied Biosystem Q5 Thermocycler using TB Green™ Premix Ex Taq™II Tli RNaseH Plus (Takara) according to the manufacture’s instructions (Li et al., 2011).

Acid resistance assay was done as previously described (Ma et al., 2003). Briefly, bacteria were grown anaerobically in M9(gly) in the absence or presence of KG to stationary phase, and then the cultures were 1:1,000 diluted into prewarmed HCl-buffered LB (pH 2.5) for 2 h acid treatment. Viable counts were measured at time 0 and 2 h after acid challenge. Survival = (CFU_{2 h}/CFU_{0 h}) × 100%.

β-galactosidase assay was performed according to Cai et al. (2013), with minor modifications. Briefly, bacteria were grown in M9(gly) with or without KG, followed by harvesting and washing with PBS, and then bacteria were diluted properly in Z buffer. These cultures were diluted 1:10 in Z buffer and assayed for β-galactosidase activity using ortho-Nitrophenyl-β-galactoside (ONPG) as a substrate.

Electrophoretic Mobility Shift Assay (EMSA) was performed essentially as the reference (Cai et al., 2013). MBP-KguR-His₆ fusion protein was expressed on the pMal-c2x vector (NEB) and induced by 1 mM IPTG at 16°C. Proteins were purified to homogeneity using Ni-NTA Spin Columns (Qiagen) and dialyzed against the binding buffer. P_kguS and P_c5038 probes were PCR amplified and gel purified using OMEGA MicroElute Gel Extracion Kit; and P_{kguSΔ BS} and P_{c5038Δ BS} probes were chemically synthesized by Genewiz (China). A probe amplified from the coding sequence of c5036 was used as a negative control (Cai et al., 2013). EMSAs were performed by adding increasing amounts of purified MBP-KguR-His₆ fusion protein (0–4 μM) to the DNA probe in the binding buffer (10 mM Tris (pH 7.5), 1 mM EDTA, 1 mM dithiothreitol, 50 mM KCl, 50 mM MgCl₂, 10 mM acetyl phosphate, 1 μg/mL bovine serum albumin, 10% glycerol) for a 30 min incubation at room temperature. The reaction mixtures were then subjected to electrophoresis on a 6% polyacrylamide gel in 0.5 × TBE buffer (44.5 mM Tris, 44.5 mM boric acid, 1 mM EDTA, pH 8.0) at 100 V for 120 min. The gel was stained in 0.5 × TBE buffer containing the SYBR Gold nucleic acid stain for 15 min, before an image was taken using a ChampGel7000 imager (SAGE, China).

All binary comparisons were analyzed by a Student’s t-test (GraphPad 9.0, Prism). A P-value < 0.05 was considered statistically significant.

To identify genes whose expression are affected by kguR deletion, we first constructed a kguR deletion mutant, ΔkguR, and assayed the growth kinetics of ΔkguR and the wild type (WT) in M9(gly) + KG medium (M9 minimal media containing glycerol as a carbon source, as well as 20 mM KG as an inducing signal) under anaerobiosis. The results showed that ΔkguR grew slightly slower than the WT in M9(gly) + KG (Figure 1A), but not in LB rich medium (data not shown), suggesting that KguR is induced and plays a role in the utilization of KG (likely as a carbon and energy source) under these conditions.

Then, we cultured the WT and the ΔkguR mutant anaerobically in M9(gly) + KG, and compared their transcriptomes using RNA-seq. With a false discovery rate ≤ 0.05 and fold-change ≥ 2, 620 genes (Figure 1B, volcano plot) were differentially expressed in the ΔkguR mutant compared to the WT; of these genes, 513 genes were downregulated and 107 genes were upregulated (A full list of differentially expressed genes is available in Supplementary Table 3).

To validate the RNA-seq data, we performed qPCR on a few representative genes that were differentially expressed in the ΔkguR mutant. These genes were selected based on the direction of differential expression (upregulation or downregulation), the extent of the change (2–16-fold), and the relevance to UPEC virulence. The qPCR results are largely consistent with the RNA-seq data (Figure 1C). As expected, target genes encoded on the KG island (except c5038) were dramatically downregulated in this experiment, indicating the credibility of this experiment.

All DEGs were then subjected to KEGG pathway enrichment analysis. As shown in Figure 1D, the majority of enriched pathways are related to amino acid metabolism, for example the valine, leucine and isoleucine (branch-chained amino acids, BCAA) biosynthesis. These results indicate that amino acid biosynthesis is likely severely affected due to the loss of kguR.

Sixty-four DEGs showed |log2fold−change| ≥ 3, and these were all downregulated in response to the loss of kguR. We categorized them into several different functional groups (Table 1), including KG utilization, acid resistance, iron uptake system, amino acid metabolism, capsule biosynthesis, ribosome synthesis, and sulfur metabolism. Each group contains ≥ 4 DEGs, with an average of log₂fold-change < −3. These results suggest that the abovementioned functions were among the most affected functions of all in the ΔkguR mutant.

To determine metabolic changes caused by kguR deletion, such as amino acids, a metabolic profiling was performed on the WT and ΔkguR strains using LC-MS/MS (Mitsuwan et al., 2017). A Principal Component Analysis (PCA) was carried out on each group that contains 6 replicates, and the results showed a clear distinction between the WT group and ΔkguR group (Figure 2A). In total, 185 metabolites were unambiguously identified. Metabolites in the ΔkguR mutant with a p-value < 0.05 and an absolute log₂fold change ≥ 1 relative to the WT were considered significantly differentially produced. A total of 36 metabolites were differentially produced, with 20 being more abundant and 16 less produced in the ΔkguR mutant (Figure 2B); and information about these metabolites was listed in Table 2. Several amino acids were downregulated in the ΔkguR mutant, including glutamate, threonine, proline, and lysine. Collectively, these data demonstrate that ΔkguR mutant has a different metabolic profile than the WT and that amino acid production is severely affected because of kguR deletion, which agrees with our KEGG enrichment analysis of DEGs.

To test whether the loss of kguR render UPEC more sensitive to acid treatment, the WT and ΔkguR strains were cultured in M9(gly) and M9(gly) + KG, respectively, followed by a treatment with pH 2.5 acidic LB. Figure 3 showed that there was no difference in acid sensitivity between the WT and ΔkguR strains when they were both cultured in M9(gly); by contrast, ΔkguR was significantly more sensitive to the WT when they were both cultured in M9(gly) + KG. Additionally, wild-type CFT073 grown in the presence of KG was more resistant to acid than those grown in the absence of KG. Together, these data indicate that KguR promotes acid resistance of UPEC, and this may be attributed to the import and conversion of KG into glutamate.

We have shown that kguS and kguR co-transcribe, thus forming a transcriptional unit (Cai et al., 2013). RNA-seq analysis revealed that deletion of kguR reduced kguS expression (Supplementary Table 3), implying that kguSR might be autoregulated. To test this, a 5′ chromosomal fusion, in which lacZ was placed immediately downstream kguR, and an internal fusion, in which kguR was fused to lacZ but was disrupted by the suicide vector pVIK112, were constructed (Figure 4A and Supplementary Figure 1). As shown in Figure 4B, disruption of kguR abolished the kguSR-lacZ expression in the presence of KG; and this aligns with the RNA-seq data. Furthermore, expression of plasmid-borne kguR in the kguSR′-lacZ fusion strain dramatically increased the kguSR′-lacZ expression. Therefore, these results indicate that KguR autoregulates its own expression.

We then hypothesized that KguR autoregulates itself through directly binding to the kguSR promoter. By alignment of the promoter regions of c5038 and kguS, we identified a relatively conserved motif (Figure 4C). Using EMSA, we showed that KguR could indeed bind to the kguSR promoter region; whereas it could not bind to a promoter variant lacking the putative binding motif. Similar results were also obtained with the c5038 promoter region (Figures 4D,E). Altogether, our data demonstrate that KguR directly autoregulates its own expression.

To identify more KguR binding sites and potentially discover additional direct targets of KguR, we carried out a genome-wide search of KguR binding regions in the CFT073 genome using the Pattern Locator program (Mrazek and Xie, 2006) with a TGTG(T/C)G-N_5–15-C(G/A)C(G/A)CA consensus. Table 3 lists an output of 22 entries, including the binding regions of c5038 and kguS. Among these entries, only 6 are localized in intergenic regions, while others are within coding sequences. Aside from c5038 and kguS, 3 genes containing putative binding sites exhibited gene expression changes in the RNA-seq analysis; and these three genes code for iron transport protein SitA (∼20-fold downregulation), acid phosphatase AppA (∼2.3-fold downregulation), and an aromatic amino acid transaminase (∼2.2-fold downregulation), respectively. Therefore, these results suggest that KguR may associate with additional sites within the genome and directly regulate more target genes, such as sitA and appA.