Functional Characterization of TetR-like Transcriptional Regulator PA3973 from Pseudomonas aeruginosa

Pseudomonas aeruginosa, a human opportunistic pathogen, is a common cause of nosocomial infections. Its ability to survive under different conditions relies on a complex regulatory network engaging transcriptional regulators controlling metabolic pathways and capabilities to efficiently use the available resources. P. aeruginosa PA3973 encodes an uncharacterized TetR family transcriptional regulator. In this study, we applied a transcriptome profiling (RNA-seq), genome-wide identification of binding sites using ChIP-seq, as well as the phenotype analyses to unravel the biological role of PA3973. Transcriptional profiling of P. aeruginosa PAO1161 overexpressing PA3973 showed changes in the mRNA level of 648 genes. Concomitantly, ChIP-seq analysis identified more than 300 PA3973 binding sites in the P. aeruginosa genome. A 13 bp sequence motif was indicated as the binding site of PA3973. The PA3973 regulon encompasses the PA3972-PA3971 genes encoding a probable acyl-CoA dehydrogenase and a thioesterase. In vitro analysis showed PA3973 binding to PA3973p. Accordingly, the lack of PA3973 triggered increased expression of PA3972 and PA3971. The ∆PA3972-71 PAO1161 strain demonstrated impaired growth in the presence of stress-inducing agents hydroxylamine or hydroxyurea, thus suggesting the role of PA3972-71 in pathogen survival upon stress. Overall our results showed that TetR-type transcriptional regulator PA3973 has multiple binding sites in the P. aeruginosa genome and influences the expression of diverse genes, including PA3972-PA3971, encoding proteins with a proposed role in stress response.


Introduction
Pseudomonas aeruginosa is a bacterium commonly found in various ecological niches and characterized by the ability to survive in very unfavorable, frequently changing environmental conditions. It is also an opportunistic human pathogen, causing infections in immunocompromised and cystic fibrosis patients, and it is often isolated from infected pulmonary and urinary tracts, eyes, or burn wounds [1]. Currently, in the pandemic era of the Coronavirus disease (COVID-19) causing severe acute respiratory syndrome, P. aeruginosa is one of the most frequently detected co-existent bacterial pathogen with COVID-19 among hospitalized patients with 12% confirmed detections, along with Haemophilus influenzae (12%) and Mycoplasma pneumoniae (42%) [2]. P. aeruginosa can increase the risk of serious complications during infection and lead to higher morbidity and mortality. It can be easily transferred, because it can survive in varying conditions, such as aquatic environments, soil, plant, and animal tissues. The success of P. aeruginosa as a pathogen is largely connected to its intrinsic resistance to antimicrobial agents and the broad spectrum of acquired resistance mechanisms making the treatment of P. aeruginosa infections a growing problem [3]. In addition to virulence and resistance factors, these bacteria modulate metabolism in response to stress factors, including antibiotics [4,5]. Thus, it is essential to understand the mechanisms responsible for the adaptation of bacteria to various environmental conditions  [40,41]. Monomers are coloured in shades of red to khaki or navy blue to olive. The helices α2 and α3 constituting the HTH DNA binding domain are marked. (D) Analysis of the oligomerization state of purified His6-PA3973 by crosslinking with glutaraldehyde. Samples were separated by SDS-PAGE (12% gel) and analyzed by dimer obtained using AlphaFold2 and HDOCK [40,41]. Monomers are coloured in shades of red to khaki or navy blue to olive. The helices α2 and α3 constituting the HTH DNA binding domain are marked. (D) Analysis of the oligomerization state of purified His 6 -PA3973 by cross-linking with glutaraldehyde. Samples were separated by SDS-PAGE (12% gel) and analyzed by Western blot using anti-His 6 antibodies. Dimeric forms are marked by two dots.
In silico analyses and database mining suggested that products of the genes encoded within the PA3973-PA3970 gene cluster could be involved in a stress response ( Figure 1A). They encode a probable acyl-CoA dehydrogenase, E. coli AidB homolog (PA3972), a putative thioesterase from PaaI family (PA3971), and an AMP nucleosidase, Amn (PA3970). According to the Pseudomonas Database, the PA3973-PA3970 genes are predicted to form an operon, but since experimental evidences are lacking, throughout this manuscript, the term gene cluster is used.

Identification of PA3973-Regulated Genes and Binding Sites for This Transcriptional Regulator in the P. aeruginosa Genome
To analyze the role of PA3973 in P. aeruginosa, a PAO1161 ∆PA3973 mutant was constructed. This mutant strain did not display obvious differences in growth in LB or M9 medium, swimming, swarming, twitching, or biofilm formation compared to the wild type (WT) parental strain PAO1161 (Figure 2A,B). In parallel, the effect of PA3973 overexpression was tested by linking the gene to an IPTG (isopropy-β-D-thiogalactopyranoside)-inducible promoter in plasmid pKKB3.11 (lacI q -tacp-PA3973). No effect of PA3973 overproduction on bacterial growth was observed in P. aeruginosa under the tested gradient of IPTG in the growth medium ( Figure 2C). RNA-seq was performed using material isolated from cultures of the strains PAO1161 pKKB3.11 (tacp-PA3973, hereafter called PA3973+) and PAO1161 pAMB9.37 (tacp, empty vector [EV]) grown in a selective LB medium supplemented with 0.05 mM IPTG ( Figure 2C; Table S3). The comparison of the PA3973+ and EV transcriptomes showed 648 loci with altered expression (fold change [FC] ≤ −2 or ≥2, adjusted p-value ≤ 0.01) ( Figure 3A; Table S3). The expression of 374 loci was downregulated, whereas 274 loci displayed increased expression. For convenience, we use the P. aeruginosa PAO1 gene names throughout the manuscript, although the corresponding PAO1161 gene names are also included. The functional classification of the identified loci, based on PseudoCAP Cultures were grown without shaking in LB medium at 37 • C for 24 h and biofilm content measurements were performed. Data represent mean ±SD from 6 biological replicates. No statistically significant changes were observed by comparing ∆PA3973 with WT (p > 0.05 in a two-sided Student t test). (C) Growth curves of P. aeruginosa PAO1161 cells carrying pKKB3.11 (tacp-PA3973) grown under selection in LB in the presence of different concentrations of IPTG. Data represent mean ± SD from 3 biological replicates.
To identify genes that display PA3973-dependent expression, we used RNA sequencing analysis (RNA-seq). In addition, we performed chromatin immunoprecipitation and sequencing analysis (ChIP-seq) to identify PA3973 binding sites in the P. aeruginosa genome. The rationale behind an analysis of cells with PA3973 in excess was based on the following: (i) the relatively low level of PA3973 expression under standard growth conditions (LB or M9 medium; Bartosik AA, personal communication); (ii) the likelihood that an excess of PA3973 might mimic the induced, activated state of the protein; and (iii) the fact that the effector, signal, or partner for this TetR-type regulator is unknown.
RNA-seq was performed using material isolated from cultures of the strains PAO1161 pKKB3.11 (tacp-PA3973, hereafter called PA3973+) and PAO1161 pAMB9.37 (tacp, empty vector [EV]) grown in a selective LB medium supplemented with 0.05 mM IPTG ( Figure 2C; Table S3). The comparison of the PA3973+ and EV transcriptomes showed 648 loci with altered expression (fold change [FC] ≤ −2 or ≥2, adjusted p-value ≤ 0.01) ( Figure 3A; Table S3). The expression of 374 loci was downregulated, whereas 274 loci displayed increased expression. For convenience, we use the P. aeruginosa PAO1 gene names throughout the manuscript, although the corresponding PAO1161 gene names are also included. The functional classification of the identified loci, based on PseudoCAP [11], showed that the upregulated genes were mostly associated with protein secretion/export systems, noncoding RNA, and cell wall functions ( Figure 3B; Table S3). Decreased expression was observed for several genes encoding proteins engaged in energy metabolism, chemotaxis, central intermediary metabolism, or adaptation and protection. The most significantly downregulated genes were norC from the PA0509-PA0527 gene cluster, nosR from the PA3391-PA3393 cluster, narK1 from the PA3874-PA3880 gene cluster, ccoP2 from the PA1555-PA1557 cluster, arcD from the PA5170-PA5173 operon, or nrdD from the PA1919-PA1920 gene cluster. Concomitantly, the most significantly upregulated loci were PA5024, PA4193 from the PA4191-PA4195 cluster, cysT from the PA0280-PA0284 cluster, PA3445-PA3446, PA2311, and PA3530 ( Figure 3B; Table S3).
The RT-qPCR (reverse transcription followed by quantitative PCR) analysis was performed to validate the changes in gene expression in the PA3973 overproducer strain in an independent set of biological replicates, collected at the same conditions as used for the RNA-seq analysis ( Figure 3C). The genes selected for the RT-qPCR analysis had a broad range of fold changes observed in PA3973+ vs. EV cells, and the selected reference gene proC (PA0393) was not altered in the tested conditions based on RNA-seq results. Essentially, most of the RT-qPCR results correlated well with the RNA-seq data, confirming a negative impact of PA3973 on the expression of PA3614, PA5208, PA5460, PA5497, and PA3972 ( Figure 3C).
To identify PA3973 binding sites in the P. aeruginosa genome, ChIP-seq analysis was performed using anti-FLAG antibodies and ∆PA3973 cells carrying plasmid pMEB255 (tacp-PA3973-flag), which were grown in a selective LB medium supplemented with 0.05 mM IPTG. As a background control for the ChIP procedure, the ∆PA3973 strain carrying plasmid pABB28.1 (tacp-flag) was grown under the same conditions, and samples were processed in parallel. The comparison of PA3973-FLAG ChIP samples with control samples using a fold enrichment (FE) cut-off value of 2 ( Figure 4A) yielded 308 PA3973-FLAG ChIPseq peaks. The summits of 139 peaks (45%) mapped to intergenic regions (Table S4), suggesting the potential of PA3973 to regulate the expression of adjacent loci. The summits of 169 peaks mapped to gene bodies (coding regions) (Table S5). A search for nucleotide motifs shared by sites bound by PA3973 using MEME [42] showed the presence of a 13 bp motif SAAGRNMTGAACG, where S-G or C; R-A or G; M-A or C ( Figure 4B) based on the analysis of 100 bp regions encompassing summits of all 308 peaks indicating PA3973 binding sites.
The functional PseudoCAP analysis of genes potentially influenced by PA3973, based on the ChIP-seq results, demonstrated enrichment in genes encoding proteins engaged in secretion/export systems, adaptation and protection, and central intermediary metabolism ( Figure 4C). Table A1 (Appendix A) lists the most enriched PA3973 bound sites (FE > 4) under the tested conditions and adjacent genes, in which expression could be modulated. This data indicates that PA3973 has multiple binding sites in the P. aeruginosa genome, which suggests that this factor may function as a crucial component of the gene expression regulatory network.  Table S3) with altered expression in response to PA3973 excess according to PseudoCAP categories [11]. The bars indicate the enrichment of genes belonging to each category relative to all genes from the category in the PAO1 genome. Numbers presented in red and blue bars correspond to the numbers of up-and down-regulated genes, respectively, in each category. (C) Validation of RNA-seq data by RT-qPCR analysis. The RT-qPCR  Table S3) with altered expression in response to PA3973 excess according to PseudoCAP categories [11]. The bars indicate the enrichment of genes belonging to each category relative to all genes from the category in the PAO1 genome. Numbers presented in red and blue bars correspond to the numbers of up-and down-regulated genes, respectively, in each category. (C) Validation of RNA-seq data by RT-qPCR analysis. The RT-qPCR was performed using RNA samples obtained for the same conditions of growth as samples used for RNA-seq analysis. The results of RNA-seq data are presented as a fold change according to data presented in Table S3, while RT-qPCR data represent the mean ratio ±SD for three biological replicates of PA3973+ cells relative to mean expression of the control EV+ strain. Reads obtained by sequencing the ChIP DNA were mapped onto the PAO1161 genome. A gene was classified as likely to be directly regulated by PA3973 if the ChIP-seq peak summit was located in the vicinity of the start codon of the gene affected in RNA-seq by PA3973 excess. (B) Sequence logo of the PA3973 binding motif and its reverse complement version obtained by MEME tool [42]. The proposed consensus sequence is presented below. (C) Functional classification of genes likely to be affected by PA3973 binding. Bars present the enrichment of loci with PA3973 binding sites identified in ChIP-seq in the vicinity of genes and coding regions (identified genes from Tables S4 and S5), belonging to each PseudoCAP category [11]. Numbers indicate the exact number of genes belonging to each category with PA3973 binding site in their vicinity. . Identification of PA3973 binding sites in P. aeruginosa (ChIP-seq). (A) Cells expressing PA3973-FLAG were subjected to chromatin immunoprecipitation using anti-FLAG antibodies. Reads obtained by sequencing the ChIP DNA were mapped onto the PAO1161 genome. A gene was classified as likely to be directly regulated by PA3973 if the ChIP-seq peak summit was located in the vicinity of the start codon of the gene affected in RNA-seq by PA3973 excess. (B) Sequence logo of the PA3973 binding motif and its reverse complement version obtained by MEME tool [42]. The proposed consensus sequence is presented below. (C) Functional classification of genes likely to be affected by PA3973 binding. Bars present the enrichment of loci with PA3973 binding sites identified in ChIP-seq in the vicinity of genes and coding regions (identified genes from Tables S4 and S5), belonging to each PseudoCAP category [11]. Numbers indicate the exact number of genes belonging to each category with PA3973 binding site in their vicinity.

Genes under the Direct Control of PA3973
Importantly, 38 of the 648 genes showing altered expression in response to a PA3973 excess possessing a binding site for this transcriptional regulator within their promoter regions or adjacent to these genes ( Figure 4A and Table A2). In addition, 26 detected in coding regions were located in the vicinity of genes that showed changes in expression level in RNA-seq analysis (Table S5), but the mechanism by which PA3973 could influence their expression requires further studies.
The ChIP-seq analysis indicated that PA3973 binds within the region preceding its own coding sequence, in agreement with its effect on the expression of the downstream located PA3972-PA3971 genes ( Table A1). The PA3973 binding site with the second highest fold enrichment, besides the PA3973 promoter region, was detected in the putative promoter of PA4156, encoding a protein highly homologous to the Vibrio cholerae vibriobactin receptor (ViuA) involved in iron acquisition [43]; however, no significant changes in PA4156 expression were detected under tested conditions in our RNA-seq analyses (Table A1).
A PA3973 binding site was also detected in the putative promoter of PA1673, the gene showing the most severe downregulation in the RNA-seq analysis (Table A2), as well as in promoter regions of PA3762 and PA3531 that were downregulated in response to PA3973 excess (Table A2). Interestingly, the PA3530, which was transcribed up-stream from the PA3531 gene, also showed changes in expression under PA3973 overproduction but with significantly increased mRNA level (Table A2).
Similarly, the PA2468, encoding a sigma factor FoxI [44], had a peak in the region preceding the gene (Table A2). Interestingly, three other genes coding for RNA polymerase sigma factors PA2426, PA3899, and PA0715 exhibited changes in expression in response to PA3973 and possess PA3973 binding sites in their promoter regions (Table A2).
To confirm the interactions of PA3973 with the selected regions, we performed electrophoretic mobility shift assays (EMSA) using purified PA3973-His 6 and DNA fragments corresponding to the putative promoter regions of the genes exhibiting the highest fold enrichment in ChIP-seq analysis: PA3973, PA4156, PA4710, PA0061, PA2722, PA0195, and PA2468 (Table A1). PA3973-His 6 was able to specifically bind to PA3973p in a concentrationdependent manner ( Figure 5A). The EMSA assay also showed PA3973-His 6 interactions with putative promoters of PA0061, PA0195, and PA2722 ( Figure 5C,D,F). No clear shifted DNA-protein complexes, but a disappearance of unbound specific DNA in the presence of the higher concentration of the protein, could be observed in the case of the PA2468p, PA4156p, or PA4710p ( Figure 5E,G,H) in tested conditions. It is possible that additional factors (e.g., a cellular partner, and/or the presence of the ligand) might be required for the efficient PA3973 binding to these sites.
All tested promoters contained sequences matching the motif SAAGRNMTGAACG predicted using MEME ( Figures 4B and 5I). In the case of PA3973p, this motif appeared twice and is a part of palindromic and pseudopalindromic sequences (TGAATCC GGATTCA and TGAAGCG TGATTCA, respectively) encompassing the predicted −10 sequence of PA3973p ( Figure 5I). To verify the importance of the identified palindromic sequence in DNA binding by PA3973, a variant of the PA3973 promoter fragment (PA3973p*) lacking the fragment encompassing the TGAATCCGGATTCA palindrome and a part of the pseudopalindrome was tested in an EMSA. PA3973-His 6 was not able to bind to this variant of PA3973p, suggesting the involvement of these sequences in the protein recruitment to this promoter ( Figure 5B,I).
To further examine the influence of PA3973 on the expression of the aforementioned genes, their promoter regions were cloned upstream of a promoter-less xylE gene in the vector pPTOI. Out of the selected genes, only PA2468 and PA4156 promoters were active in the heterologous host E. coli DH5α ( Figure 5J). The expression of PA3973 in cells carrying a plasmid with PA4156p-xylE resulted in significantly reduced XylE activity in the corresponding cell extracts, but no such regulation was observed for PA2468p-xylE under the tested conditions ( Figure 5J). Probable binding motifs are marked with the purple box. "*" indicates the site of truncation of PA3973p to obtain its shorter fragment. (J) XylE activity in E. coli DH5α carrying pMEB267 (PA2468p-xylE) or pMEB269 (PA4156p-xylE) with pKKB3.11 (tacp-PA3973) (++) or empty plasmid pAMB9.37 (−). Strains were grown in selective LB with supplementation of 0.2 mM IPTG for pKKB3.11 (++). Cells carrying the promoter-less pPTOI (-xylE) (EV) and pAMB9.37 were used as background control. The data represent the means ± SD and * indicates p < 0.05 in a Student's twotailed t-test. (K) The genes with altered expression in ΔPA3973 strain vs. WT PAO1161 as indicated by RNA-seq analysis and RT-qPCR analysis carried out using RNA samples obtained for the same conditions as samples used for RNA-seq analysis. The results of RNA-seq data are presented as a fold change according to data presented in Table S6, while RT-qPCR data are presented as a ratio. Data represent mean ratio ± SD for three biological replicates. * indicate significantly different expression (ratio > 2, p-value < 0.05 in two-sided Student's t-test assuming equal variance) comparing ΔPA3973 with WT using the proC reference gene; NS-not significant.
All tested promoters contained sequences matching the motif SAAGRNMTGAACG predicted using MEME ( Figures 4B and 5I). In the case of PA3973p, this motif appeared Strains were grown in selective LB with supplementation of 0.2 mM IPTG for pKKB3.11 (++). Cells carrying the promoter-less pPTOI (-xylE) (EV) and pAMB9.37 were used as background control. The data represent the means ± SD and * indicates p < 0.05 in a Student's two-tailed t-test. (K) The genes with altered expression in ∆PA3973 strain vs. WT PAO1161 as indicated by RNA-seq analysis and RT-qPCR analysis carried out using RNA samples obtained for the same conditions as samples used for RNA-seq analysis. The results of RNA-seq data are presented as a fold change according to data presented in Table S6, while RT-qPCR data are presented as a ratio. Data represent mean ratio ± SD for three biological replicates. * indicate significantly different expression (ratio > 2, p-value < 0.05 in two-sided Student's t-test assuming equal variance) comparing ∆PA3973 with WT using the proC reference gene; NS-not significant.
To further define the targets of PA3973 action, we performed an RNA-seq analysis of PA3973-deficient cells relative to WT cells, which showed significantly increased PA3972 and PA3971 transcript levels, confirming the role of PA3973 in PA3972-PA3971 repression ( Figure 5K; Table S6). Additionally, three genes, PA2174, PA5497, and D3C65_10195, also exhibited increased mRNA levels in the absence of PA3973, indicating a possible involvement of this regulator in their control. The RT-qPCR verification of the changes observed in the RNA-seq analysis showed the direction of changes to be consistent for both analyses, confirming the influence of PA3973 on the expression of these genes ( Figure 5K). Overall, this data confirmed that PA3973 binds to DNA fragments identified in the ChIP-seq analysis and may regulate the activity of target promoters to influence gene expression.

Towards the Biological Function of PA3973-PA3971 Gene Cluster
In silico analyses and database mining suggested that the products of the PA3973-PA3971 gene cluster could potentially be involved in stress response. Besides the first gene coding for the TetR-type transcriptional regulator PA3973, this cluster also encodes a probable acyl-CoA dehydrogenase (PA3972) and a putative thioesterase (PA3971). The PA3972 from P. aeruginosa is the putative homolog of the AidB protein from E. coli (41% identity, 56% similarity) with possible acylo-CoA dehydrogenase activity and oxidoreductase activity, acting on the CH-CH group of donors [11]. The AidB protein is involved in the defense against methylating agents [45][46][47]. In E. coli, aidB was upregulated in response to small doses of DNA-methylating agents [48]. This gene, along with alkA and alkB, belongs to the Ada protein regulon that controls the cellular stress response induced by the presence of alkylating agents, leading to changes in DNA or RNA [49].
The PA3971 gene encodes a hypothetical protein belonging to the thioesterases family [11]. Its expression level increases when the strain is subjected to stress factors such as hydrogen peroxide [50]. A BLAST analysis showed approximately 40% similarity to a thioesterase from E. coli, which is responsible for the degradation of phenylacetic acid [51], and a high 91% similarity to a putative esterase from Acinetobacter baumanii with a role in the biosynthesis of secondary metabolites.
To assign a biological function to PA3972 or PA3971, the chromosomal double mutant of PAO1161 in PA3972-71 genes was constructed. No changes in colony morphology, bacterial motility, or biofilm formation were observed between the ∆PA3972-71 mutant and the WT strain. Lack of PA3973 or PA3972-71 did not significantly affect the growth of the cells in a rich medium, or minimal M9 or MOPS medium supplemented with glucose as the carbon source in the early stages of culture growth (Figure 6A,D,G). However, prolonged incubation over eight hours (and up to 48 h in the case of a MOPS medium) resulted in a significantly lower OD 600 of the ∆PA3972-71 mutant and to a less extent for ∆PA3973 in comparison with the WT strain ( Figure 6G). This suggests that in conditions that might involve stress possibly connected with starvation, the PA3973, PA3972, and/or PA3971 proteins might play an important role in facilitating pathogen survival. The growth in the presence of hydroxylamine and hydroxyurea, two stress-inducing agents, was also tested ( Figure 6B,C,E,F). Hydroxylamine exposure introduces mutations by acting as a DNA nucleobase amine-hydroxylating agent [52], whereas hydroxyurea is a ribonucleotide reductase inhibitor, altering DNA replication and nucleotide metabolism by the depletion of deoxyribonucleoside triphosphate pool [53].
The presence of hydroxylamine (50 µg/mL) or hydroxyurea (1-20 mM) tested in an LB and M9 minimal medium with glucose slowed down the growth of P. aeruginosa strains in comparison to growth in a medium without these compounds ( Figure 6A-F). The presence of hydroxylamine caused a prolonged lag time of bacterial culture growth with a significantly slower kinetic of the ∆PA3972-71 strain in comparison with the WT or ∆PA3973 in M9 medium ( Figure 6E), indicating that PA3972 and/or PA3971 could facilitate the growth of bacteria in such conditions. PA3971 proteins might play an important role in facilitating pathogen survival. The growth in the presence of hydroxylamine and hydroxyurea, two stress-inducing agents, was also tested ( Figure 6B,C,E,F). Hydroxylamine exposure introduces mutations by acting as a DNA nucleobase amine-hydroxylating agent [52], whereas hydroxyurea is a ribonucleotide reductase inhibitor, altering DNA replication and nucleotide metabolism by the depletion of deoxyribonucleoside triphosphate pool [53].  The significantly slower growth of ∆PA3972-71 cultures in comparison with the WT, was also observed in the presence of 20 mM (in LB medium) or 1 mM (in M9 medium) hydroxyurea ( Figure 6C,F). This data indicates that PA3972 and/or PA3971 may play a role in stress response; however, further studies are needed to discover the underlying mechanisms.
The RT-qPCR analysis was also applied to check the expression of PA3973 and selected genes from its regulon in the P. aeruginosa WT cells exposed to antibiotics and hydroxylamine. The analysis showed no significant difference in the PA3973 expression upon the addition of carbenicillin compared to control cultures (non-treated grown in LB), but significantly decreased the expression in response to kanamycin, and increased the mRNA level of PA3973 in response to tetracycline ( Figure 6H). Interestingly, the expression of PA3972-PA3970 genes was significantly downregulated in response to tetracycline (Figure 6H), which correlates with the increased expression of PA3973, confirming its repressor functions for the PA3972-71 genes. Moreover, in the presence of hydroxylamine, the expression of PA3973 was significantly down-regulated, and the PA3972, but not PA3971 or PA3970, showed a notable up-regulation, which suggests that PA3972 expression is induced in response to hydroxylamine.
This data indicates that PA3973 action might be connected with the cellular response to some stressors, but further studies are needed to discover the underlying mechanisms and the role of PA3973-dependent genes, including PA3972, in this process.

Discussion
The PA3973 of P. aeruginosa based on sequence similarity is classified as the TetR-type transcriptional regulator. The N-terminal part, similar to other members of the TetR family, consists of the HTH DNA-binding domain predicted with structure modeling, in which helices α2 and α3 are involved in direct contact with cognate DNA [14,15]. The C-terminal part is presumably involved in signal perception and oligomerization. The presence of a dimeric form was confirmed with glutaraldehyde cross-linking experiments with the use of purified His 6 -PA3973. The signal to which PA3973 responds in P. aeruginosa awaits elucidation.
The ChIP-seq was applied to evaluate the binding of PA3973 across the chromosome in exponentially growing P. aeruginosa cells. The 308 PA3973-binding targets located across the genome, including intergenic and coding regions, were identified. Multiple binding sites in the genome suggest the role of PA3973 as a global regulator of gene expression in P. aeruginosa. However, the transcriptional profiling of ∆PA3973 mutant cells grown in rich medium did not support the global impact. Significantly, under this conditions, the increased expression of PA3972, PA3971, PA2174, and PA5497 genes was detected in PA3973 deficient cells in comparison with the WT ( Figure 5K; Table S6). In contrast, the RNA-seq analyses of exponentially growing cells overproducing PA3973 showed 648 genes responding to PA3973 (Table S3). Among them 38 genes possess PA3973 binding sites in their promoter regions or in the vicinity of these genes and the additional 26 genes in coding regions (Table A2, Tables S4 and S5). These genes could exhibit direct PA3973-dependent regulation in contrast to the remaining loci with altered expression in PA3973 overexpressing cells, which likely represent the indirect effect of PA3973 action on P. aeruginosa cells.
In the case of most transcriptional regulators their high-affinity binding sites are largely restricted to non-coding DNA regions to exert direct effect on the target genes' expression [54]. In the case of PA3973, more than half (169 out of 308) of the DNA targets for PA3973 are located within the coding regions of the genome and appear to have little or no effect on gene expression, at least under the conditions tested. This may suggest that additional factors (e.g., special growth conditions, presence of ligand, signal and/or cellular partner) might be involved in PA3973 regulatory activity. A similar pattern was observed; for example, in E. coli, RutR, the uracil-responsive transcription factor from the TetR family involved in the regulation of the expression of the rut operon encoding genes for the catabolism of pyrimidine. RutR showed multiple additional binding sites within genes; however, their role in gene expression control remains elusive [55]. Concomitantly, an interesting mode of action among TetR-like regulators exemplifies a LuxR protein from Vibrio alginolyticus [56]. It controls the expression of approximately 280 genes that contain either a symmetric palindrome or asymmetric binding motifs upstream, overlapping start codon, or inside open reading frames. Differences in LuxR binding to these motifs, modulated by its N-terminal extension, allow the repression or activation of cognate gene expression to impact quorum sensing and virulence [56]. Based on ChIP-seq data and the localization of identified PA3973 binding sites, we speculate that this regulator might present a similar mode of action.
The sequence analysis of PA3973 ChIP-seq peaks revealed a 13-bp binding motif with the consensus SAAGRNMTGAACG. In the case of PA3973p, this motif was found twice in the sequence encompassing the predicted −10 promoter sequence ( Figure 5I). Two closely located motifs, separated by a dozen or more base pairs, are also present in other loci identified as PA3973 targets (e.g., PA0061p, PA0195p, PA2468p, and PA4710p). One PA3973-binding motif was identified in PA2722p, PA4056p, or PA4156p. In some target regions, the binding sites for other regulatory factors could be detected, such as the Fur (ferric uptake regulation protein) binding site in PA4156p or PA4710p. This suggests that, besides PA3973, other factors might be involved in the regulation of these genes. The Fur is involved in the regulation of iron metabolism in P. aeruginosa (e.g., in the control of the production of receptors and siderophores involved in iron scavenging) [57]. Moreover, the PA3973 binding site was identified in the promoter region of the PA2468 encoding extracytoplasmic function sigma factor (ECF) FoxI [44].
Other interesting regulatory connections are the decreased expression of PA1920 in response to PA3973 excess (Table A1 and Table S3) and the up-regulation of PA5497 in ∆PA3973 mutant ( Figure 5K; Table S6). The PA1920 and PA5497 encode a class III (anaerobic) ribonucleoside-triphosphate reductase subunit, NrdD, and a class II (cobalamin-dependent) ribonucleotide-diphosphate reductase subunit, NrdJa, respectively. Importantly, it was shown that both ribonucleotide reductases (RNR) are highly expressed during infection [58]. In P. aeruginosa, three different types of RNR are encoded, allowing adaptation to different environmental conditions including anaerobiosis [59,60]. The expression of RNR genes is under the control of NrdR, a global regulator, which, in E. coli, binds to the NrdR boxes (acaCwAtATaTaTwGtg) in the promoters of target genes [60]. Such sequences could be found in PA1920 and PA5497 promoter sequences, and the ChIP-seq data showed the PA3973 binding site in the coding sequence of PA1920 and PA5497. Moreover, the PA3973 binding site with fold enrichment 10 (Table A1) was detected in the predicted PA4056-PA4057 (nrdR-ribD) operon, overlapping the end of the nrdR and the beginning of the ribD ORFs. It suggests the existence of complex regulatory interdependencies, managing the expression of these genes.
The EMSA assays for the selected targets of PA3973 showed varying affinity of PA3973 to analyzed targets, with the strongest binding observed in the case of PA3973p comprising two pseudopalindromic/palindromic sequences, which likely served as the operator sequences recognized by PA3973 ( Figure 5A-H). In the promoter region of PA3973, the identified PA3973 binding sites are part of two closely spaced pseudopalindromic/palindromic sequences of 14 bp in length separated by 11 bp (Figure 5I). Their arrangement and localization resemble the operator sequences of other TetR-type regulators (e.g., TetR with two 15 bp operator sequences (TCTATCATTGATAGG) separated by 11 bp) [14,18]. The deletion of the palindromic sequence and part of the psedopalindromic sequence prevented PA3973 binding to truncated PA3973p*, suggesting that the presence of these sequences is needed for protein binding to cognate sequence and to exert a possible effect on expression (e.g., repression). It is tempting to propose the model in which PA3973 acting as a dimer binds to these operator sequences, repressing expression from its own promoter and possibly PA3972-PA3971 genes.
Our studies on P. aeruginosa growth in the presence of stress-inducing agents like hydroxylamine or hydroxyurea showed that PA3972 could facilitate bacterial growth in such detrimental conditions. Interestingly, it was recently shown that PA3972 belongs to negative regulators of bacterial virulence as predicted with PseudomonasNet and further validated by the analysis of the PA3972 knockout mutant in the Caenorabditis elegans infection model [37]. The disruption of PA3972 resulted in elevated P. aeruginosa PAO1 virulence in C. elegans.
Taken together, these results demonstrated the participation of TetR-type transcriptional regulator PA3973 in several diverse cellular processes including the modulation of cellular response to growth conditions to keep cellular homeostasis and facilitate pathogen survival. Some of the identified PA3973 targets play a role in P. aeruginosa stress response and pathogenesis, but the exact mechanism and the signals to which PA3973 responds and regulates its targets need further investigation. Presented data brings us closer to understanding the complex P. aeruginosa system, to draw the picture of regulatory connections involved in controlling bacteria lifestyle as a free-living organism and during infection.

Bacterial Strains, Plasmids, and Growth Experiments
The bacterial strains used and constructed in this study (listed in Table S1) were grown in an L broth (LB) rich medium or on LB-agar at 37 • C, or in M9 or MOPS minimal medium [61] supplemented with glucose (0.5%) as the carbon source, and with leucine (10 mM) added in the case of P. aeruginosa PAO1161 leu − strains cultured in minimal media. Hydroxyurea (1 or 20 mM) and hydroxylamine (50 µg/mL) were used in growth experiments as indicated. For the selection of plasmids in E. coli, the media were supplemented with 10 µg/mL chloramphenicol, 50 µg/mL kanamycin, or benzyl penicillin at a final concentration of 150 µg/mL in a liquid medium or 300 µg/mL in agar plates. For P. aeruginosa strains, carbenicillin (300 µg/mL), rifampicin (300 µg/mL), and chloramphenicol (75 µg/mL in a liquid medium; 150 µg/mL in plates) were used as required. For growth experiments, cultures were grown overnight with shaking at 37 • C and diluted with fresh medium either in flasks, closed with cotton plugs, or in 96-well plates. Bacterial growth was monitored by measurements of optical density at 600 nm (OD 600 ) in a spectrophotometer Shimadzu UV-1800 (Shimadzu Corporation, Kyoto, Japan) or with the use of Varioskan Lux Multimode Microplate Reader and SkanIt RE software (Thermo Fisher Scientific, Waltham, MA, USA). Competent E. coli cells were prepared with a treatment of CaCl 2 [62]. The competent P. aeruginosa cells were prepared with a treatment of MgCl 2 as described previously [63].
All plasmids used and constructed in this study are described in Table S1. All primers used are listed in Table S2.
The P. aeruginosa PAO1161 ∆PA3973 and ∆PA3972-71 mutants were obtained with allelic exchange. The competent cells of E. coli S17-1 were transformed with plasmid pKKB60.6 or pSOB3 (derivatives of suicide vector pAKE600) to create the donor strain, and the WT P. aeruginosa PAO1161 Rif R was used as the recipient. The allele exchange procedure was performed as described previously [64]. The verification of the allele exchange was performed with PCR.

Motility and Biofilm Formation Assays
Motility assays were performed as described previously using cultures in an LB medium [65]. To standardize the assay, all plates contained the same volume of the medium. The biofilm amount was measured with the crystal violet staining method [66].

Protein Purification
The E. coli BL21(DE3) strains transformed with pKKB28.3 or pMEB265 encoding a His 6 -PA3973 or PA3973-His 6 fusion, respectively, were grown to exponential phase in an LB medium with the supplementation of 0.5 mM IPTG. The cells were harvested with centrifugation, resuspended in a LEW buffer [50 mM sodium phosphate buffer, pH 8.0; 300 mM NaCl] supplemented with lysozyme (1 mg/mL), PMSF (1 mM), and benzonase nuclease (250 U, Sigma), and sonicated. His 6 -tagged proteins were purified from the cell lysate using chromatography on Ni-agarose columns (Protino Ni-TED 1000, Macherey-Nagel, Düren, Germany) with 300 mM imidazole in LEW buffer used for elution. The purification procedure was monitored by SDS-PAGE using a Pharmacia PHAST gel system. The fractions containing the purified protein were dialyzed overnight in 50 mM Tris-HCl buffer (pH 8.0) containing 5% (v/v) glycerol and stored in aliquots at −80 • C.

Cross-Linking of Purified Protein
The purified His 6 -PA3973 was cross-linked with increasing concentrations of glutaraldehyde as previously described [67]. Samples were then suspended in the loading buffer [50 mM Tris-HCl (pH 8.0), 0.1 M DTT, 2% SDS, 0.1% bromophenol blue, 10% glycerol], boiled, and separated on an SDS-PAGE gel, transferred to nitrocellulose membrane (Amersham Protran, Cytiva, Germany), and used in a Western blot analysis with the use of anti-His 6 antibodies.

RNA Isolation, RNA-Seq
The total RNA was isolated from three independent replicate samples of P. aeruginosa PAO1161 overexpressing the PA3973 gene, samples of control strain carrying the empty vector, as well as P. aeruginosa PAO1161 WT and ∆PA3973 strain. RNA isolation, sequencing, as well as data analysis were performed essentially as described previously [68]. The raw data is available in the NCBI's Gene Expression Omnibus (GEO) database (http://www. ncbi.nlm.nih.gov/geo/) under accession number GSE211771 (released 22 November 2022).

RT-qPCR Analyses
For the qRT-PCR analyses, cells from the PAO1161 pKKB3.11 (lacI q -tacp-PA3973) and PAO1161 pAMB9.37 (lacI q -tacp) strains grown in a selective LB medium supplemented with 0.05 mM IPTG, or from PAO1161 WT and ∆PA3973 mutant cultures grown in the rich medium were collected from 2 mL of cultures at an optical density at 600 nm 0.5. For the testing influence of stress conditions on the expression of selected genes, the RNA was isolated from the WT PAO1161 cultures grown in a rich medium supplemented with subinhibitory concentrations of carbenicillin (32 µg/mL), kanamycin (10 µg/mL), tetracycline (4 µg/mL), or hydroxylamine at concentration 50 µg/mL. The RNA was isolated from 2 mL of cultures collected at an optical density at 600 nm 0.5.
Three biological replicates of total RNA (2 µg per reaction) from each strain served as a template for cDNA synthesis with TranScriba Kit (A&A Biotechnology, Gdansk, Poland). The cDNA was used as a template in qPCR performed with 5xHOT FIREPol EvaGreen qPCR Mix Plus (Solis Biodyne, Tartu, Estonia). Three technical replicates per each gene/sample were used. The sequences of primers used for RT-qPCR analysis are listed in Table S2. The efficiency of the quantitative PCR reaction with each primer pair was calculated and used to calculate the ratio of each studied gene to the reference gene. All primer pairs used showed the efficiency of amplification between 0.95-1.05. Changes in individual gene expression between the mutant and WT strain, between PA3973 overproducer and control strain, or between treated and untreated cells were calculated with the normalization of Cp values to mean Cp value for a proC (PA0393) reference gene using the Pfaffl method [69]. The qPCR was performed using the Light Cycler 480 (Roche). PCR products were detected with SYBR green fluorescent dye and amplified according to the following protocol: one cycle at 95 • C for 15 min followed by 40 cycles at 95 • C for 15 s and 60 • C for 20 s. Each 18 µL reaction contained 3.6 µL 5× reaction mix, 1 µL of five times diluted cDNA, and 1.5 µL of mixed 6 µM primers. In each run, negative controls (no cDNA) for each primer set were included.

In Vitro Protein-DNA Interactions
The electrophoretic mobility shift assay with purified PCR products (~100 ng in reaction) and PA3973-His 6 was conducted in a binding buffer [10 mM Tris-HCl (pH 8.5), 10 mM MgCl 2 , 100 mM KCl, 0.1 mg/mL BSA]. The tested DNA fragments were amplified using the appropriate pair of primers listed in  (Table S2), was used as an internal control in each reaction. The DNA fragments were incubated in a binding buffer with the increasing concentration of purified PA3973-His 6 at 23 • C for 30 min. All reactions were prepared in a total volume of 20 µL. The complexes were separated on a 1.5% agarose gel in 0.5× Tris-borate-EDTA (TBE) buffer at 4 • C. The DNA was stained with ethidium bromide and visualized under UV light.

Regulatory Experiments with Promoter-xylE Fusions in E. coli
The E. coli DH5α double transformants carrying pPT01 derivatives with the promoter regions of selected P. aeruginosa genes fused to the xylE reporter gene plus pAMB9.37 (tacp) or pKKB3.11 (tacp-PA3973) were assayed for catechol 2,3-oxygenase activity (the activity of the product of xylE). The experiment was performed as previously described [70] using cell extracts prepared from exponentially growing cultures in LB. Measurements of the amount of the protein in extracts were conducted using the Bradford assay [71].

Acknowledgments: We thank Szymon
Owczarek, a graduate student at the Warsaw University of Technology, Faculty of Chemistry, Warsaw, Poland for the construction of vectors for PA3972-71 mutagenesis and preliminary studies on the characterization of PA3972-71 from P. aeruginosa. This work was supported by the National Science Centre in Poland (grant 2015/18/E/NZ2/00675). We acknowledge financial and technical support from the Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Warsaw, Poland.

Conflicts of Interest:
The authors declare no conflict of interest.
Appendix A Table A1. List of the most enriched PA3973 bound sites identified in the P. aeruginosa PAO1161 genome (FE > 4). Abbreviations: P-promoter; T-terminator; gb-gene body; +/− strand in PAO1161 genome; FE-fold enrichment for ChIP-seq data; FC-fold change for RNA-seq data; NA-not annotated, genes identified only in PAO1161 strain but not in PAO1; PseudoCAP categories as in Figure 3B [11].  Table A2. List of 38 genes altered in response to PA3973 excess with PA3973 binding site in their promoter regions or adjacent to these genes. Full data for each peak/gene could be found in Tables S3-S6. Abbreviations: P-promoter; T-terminator; gb-gene body; +/− strand in PAO1161 genome; FE-fold enrichment for ChIP-seq data; FC-fold change for RNA-seq data; NA-not annotated, genes identified only in PAO1161 strain but not in PAO1; PseudoCAP categories as in Figure 3B [11].