Genome-Wide Identification and Expression Analysis of SOS Response Genes in Salmonella enterica Serovar Typhimurium

cells Article Genome-Wide Identification and Expression Analysis of SOS Response Genes in Salmonella enterica Serovar Typhimurium Angela Mérida-Floriano 1, Will P. M. Rowe 2and Josep Casadesús1,*   Citation: Mérida-Floriano, A.; Rowe, W.P.M.; Casadesús, J. Genome-Wide Identification and Expression Analysis of SOS Response Genes in Salmonella enterica Serovar Typhimurium. Cells 2021,10, 943. https://doi.org/ 10.3390/cells10040943 Academic Editors: Bernard S. Lopez and Ivan Matic Received: 16 February 2021 Accepted: 16 April 2021 Published: 19 April 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1Departamento de Genética, Facultad de Biología, Universidad de Sevilla, Apartado 1095, E-41080 Sevilla, Spain; [email protected] 2Institute of Microbiology and Infection, University of Birmingham, Birmingham B15 2TT, UK; w[email protected] *Correspondence: [email protected]; Tel.: +34-95-455-7105 Abstract: A bioinformatic search for LexA boxes, combined with transcriptomic detection of loci responsive to DNA damage, identified 48 members of the SOS regulon in the genome of Salmonella enterica serovar Typhimurium. Single cell analysis using fluorescent fusions revealed that heterogeneous expression is a common trait of SOS response genes, with formation of SOS OFF and SOS ON subpopulations. Phenotypic cell variants formed in the absence of external DNA damage show gene expression patterns that are mainly determined by the position and the heterology index of the LexA box. SOS induction upon DNA damage produces SOS OFF and SOS ON subpopulations that contain live and dead cells. The nature and concentration of the DNA damaging agent and the time of exposure are major factors that influence the population structure upon SOS induction. An analogy can thus be drawn between the SOS response and other bacterial stress responses that produce phenotypic cell variants. Keywords: SOS response; LexA; LexA box; heterology index; phenotypic heterogeneity 1. Introduction The SOS regulon is a bacterial gene network that facilitates survival to DNA damage [ 1 ]. Transcription of SOS genes is under the control of the LexA repressor, which binds the so-called “LexA box” or “SOS box”, a cognate DNA motif whose sequence varies among different species [ 2 , 3 ]. In the absence of DNA damage, LexA prevents transcription of SOS genes [ 4 ]. Upon DNA damage, the recombination protein RecA forms nucleofilaments on single-stranded DNA. RecA nucleoprotein filaments activate autoproteolysis of LexA, and transcription of the SOS response genes is turned on [ 4 , 5 ]. Events that produce single-stranded DNA include arrest or stalling of DNA replication, collision between transcription and replication machines, formation of R-loops, aggression by DNA-damaging agents and DNA uptake during transformation or conjugation [5]. Genes belonging to the SOS regulon encode DNA repair functions, DNA polymerases able to perform translesion DNA synthesis, proteins involved in cell division control and additional gene products with ancillary roles in genome defense [ 4 – 6 ]. SOS activation also triggers induction of prophages and transcription of genes encoding bacteriocins [ 7 ]. The composition of the SOS regulon differs from one bacterial genus to another [5,8–11]. In Salmonella enterica serovar Typhimurium, homologs of E. coli SOS genes were described in the 1980s [ 12 – 14 ]. Additional members of the SOS regulon were identified with a genetic screen [ 15 ] and by PCR-assisted RNA fingerprinting [ 16 ]. Comparison of the Salmonella and E. coli SOS responses reveal strong (but not absolute) conservation [ 15 , 17 ]. Differences have been also described among S. enterica strains: for instance, lexA null mutations are lethal in the historic strain LT2 due to induction of the Fels2 prophage [ 18 ] but are viable in other model strains. A relevant aspect of the S. enterica SOS response is the involvement of DNA repair and DNA recombination functions in the interaction with the Cells 2021,10, 943. https://doi.org/10.3390/cells10040943 https://www.mdpi.com/journal/cells Cells 2021,10, 943 2 of 17 animal host [ 19 – 23 ]. In fact, inhibition of the SOS response has been considered a potential therapeutic strategy [24]. In this study, a combination of bioinformatic and transcriptomic analysis has identified 48 genes belonging to the S. enterica SOS regulon. All genes in the list are canonical in the sense that they harbor LexA boxes and undergo transcriptional activation in response to DNA damage. Their expression patterns, however, turn out to be complex due to the formation of SOS OFF and SOS ON subpopulations, a phenomenon observed both in the absence of DNA damage and upon SOS response activation. 2. Materials and Methods 2.1. Bacterial Strains, Culture Media and Growth Conditions All Salmonella enterica strains used in this study belong to serovar Typhimurium. Unless indicated otherwise, the strains derive from ATCC 14028 ( Supplementary Table S2 ). Strain ST4/74 [ 25 ] was used in specific experiments. The source of the lexA (Ind − ) allele was strain DA6522 [ 18 ], kindly provided by John R. Roth, University of California, Davis, CA, USA. The recD and recF alleles were described by Cano et al. [ 20 ]. Chromosomal mutations were transduced between strains using phage P22 HT 105/1 int201 [ 26 ]. The P22 transduction protocol and the procedure for isolation of phage-free transductants were described elsewhere [27]. Liquid cultures were prepared in lysogeny broth (LB) and were grown at 37 ◦ C with shaking at 200 rpm. LB solid medium contained agar at a final concentration of 15 g/L. Chemicals were used at the following concentrations: ampicillin, 100 µ g/mL; chloramphenicol, 20 µ g/mL; kanamycin, 50 µ g/mL; nalidixic acid, 8.5 µ g/mL; hydroxyurea, 0.1 M; mitomycin C, 0.25 µ g/mL; H 2 O 2 , 0.01%. Green plates [ 28 ] contained methyl blue (Sigma-Aldrich, St Louis, MO, USA) instead of aniline blue. 2.2. RNA Isolation, cDNA Library Preparation and Illumina Sequencing Overnight cultures were diluted 1:100 in 5 mL of LB. For the ‘non-inducing’ condition the three biological replicates were incubated at 37 ◦ C, 200 rpm until early stationary phase (ESP, OD 600 ~2.00). Three ‘SOS-inducing’ condition replicates were incubated 1 h in LB, and nalidixic acid was added to a final concentration of 8.5 µ g/mL. The cultures were incubated 4 more hours before RNA extraction (ESP, OD600 ~1.60). Total RNA was isolated using the Direct-zol RNA MiniPrep Plus kit (Zymo Research, Irvine, CA, USA). For RNA-seq, cDNA libraries were prepared and sequenced by Vertis Biotechnologie AG (Freising, Germany). The indexed sequencing libraries were pooled in equimolar amounts, size-selected to 200–550 bp and sequenced on an Illumina NextSeq 500 system using 75 bp read length. 2.3. Read Processing and Alignment The quality of RNA-seq libraries was assessed using FastQC v0.11.5 (Babraham Institute, Cambridge, UK) (http://www.bioinformatics.babraham.ac.uk/projects/fastqc (accessed on 20 September 2017)). Trimmomatic v0.36 tool (The Usadel Lab, Aachen, Germany) was used to remove the Illumina TruSeq adapter sequences, leading and trailing bases. Reads with a length shorter than 40 nucleotides after trimming were discarded for further analysis. The remaining reads of each library were aligned to the sequence of the published S. Typhimurium ATCC 14028 chromosome (accession: CP001363.1) using Bowtie2 v2.2.29 [ 29 ], and alignments were filtered with Samtools v1.3.1 [ 30 ] using a MAPQ cut-off of 15. The complete RNA-seq pipeline used for this study is described in https: //github.com/will-rowe/rnaseq (accessed on 27 November 2017). 2.4. Quantification of Absolute Gene Expression Levels Absolute gene expression levels were calculated as Transcript Per Million (TPM) values [ 31 , 32 ] and were generated for protein coding genes and noncoding sRNAs in the Cells 2021,10, 943 3 of 17 chromosome of S. Typhimurium ATCC 14028. Based on the values obtained, the expression cut-off was set as TPM > 10 for both types of genes [33]. 2.5. Differential Gene Expression Analysis with Three Biological Replicates Raw read counts from the three biological replicates in ESP and the three biological replicates in ‘SOS-inducing’ conditions were uploaded into Degust (http://degust.erc. monash.edu/ (accessed on 27 November 2017)). Data were analyzed using the Voom/Limma approach [ 34 , 35 ] with an FDR of ≤ 0.001 and a Log 2 FC ≥ 1. This cut-off was set to identify genes slightly induced by nalidixic acid. To remove genes with low counts, thresholds of ≥ 10 read counts and ≥ 1 Counts Per Million (CPM) in at least the three biological replicates of one sample were used [36]. 2.6. Identification of LexA Boxes Putative LexA boxes were identified using the FIMO motif scanning tool [ 37 ]. The consensus LexA box used as a motif was 5‘ TAC TGT ATA TAT ATA CAG TA 3 0 [ 38 ], scanned against the S. Typhimurium ATCC 14028 chromosome (accession number CP001363.1) using default parameters. Both DNA strands were scanned using default parameters, and only matches with p-value < 0.0001 were selected. 2.7. Construction of Transcriptional GFP Fusions A fragment containing the promoterless green fluorescent protein (gfp) gene, originally amplified from pZEP07 [ 39 ], and a kanamycin resistance cassette, originally amplified from pKD13 [40], was amplified with primers listed in Supplementary Table S3. These primers added flaking regions homologous to the 3 0 -UTR of the selected genes. The PCR products were integrated into the Salmonella chromosome using the Lambda Red recombination system [ 40 ]. Since the homologous regions are located downstream the gene stop codon, the integration creates a transcriptional gfp fusion and the upstream gene remains intact. 2.8. Flow Cytometry Flow cytometry was used to monitor gene expression at the single cell level. For both inducing and non-inducing conditions, cultures were grown in LB at 37 ◦ C. For SOS- inducing conditions, inducers were added during exponential growth. Flow cytometry assays were performed using bacterial cells diluted in PBS 1x. For discrimination of live and dead cells, propidium iodide was used [ 41 ]. Data acquisition was performed using a Cytomics FC500-MPL cytometer (Beckman Coulter, Brea, CA, USA). Data were collected for 100,000 events per sample, and were analyzed with FlowJo 8.7 software (Tree Star Inc. Ashland, OR, USA). 2.9. Statistical Analysis Student’s t-test (unpaired) was performed using GraphPad Prism version 6.0 for Mac to determine the statistical differences between two groups. 3. Results 3.1. Identification of LexA Boxes in the Chromosome of S. enterica Serovar Typhimurium Search for the consensus LexA binding sequence, 5’ TAC TGT ATA TAT ATA CAG TA 3’ [ 38 ] using the software tool FIMO [ 37 ] identified 318 potential LexA boxes with a fairly regular distribution in both DNA strands across the chromosome (Supplementary Figure S1). Unlike other studies that identified putative LexA boxes [ 2 , 42 ], no additional parameters were applied. Despite the high number of putative LexA boxes identified, FIMO analysis failed to detect at least two known LexA boxes: the second box in the lexA gene [2] and the third box in the recN gene [2]. These boxes were manually added. Cells 2021,10, 943 4 of 17 3.2. Identification of SOS Response Loci The list of putative LexA-regulated genes was narrowed down by identifying loci that altered their expression under SOS-inducing conditions. For this purpose, we used Illumina-Seq to analyze the transcriptome of S. Typhimurium ATCC 14028 in the presence of sub-lethal concentrations of nalidixic acid, a fluoroquinolone that induces the SOS response by targeting DNA gyrase and topoisomerase IV [ 43 , 44 ]. An average of 11,010,991 reads per sample were obtained for non-inducing conditions, and an average of 35,550,086 reads per sample for SOS-inducing conditions. Raw and processed data from RNA- Seq analysis have been deposited at the Gene Expression Omnibus (GEO) database ( http://www.ncbi.nlm.nih.gov/geo/ (accessed on 7 October 2020)), with accession number GSE159310. Using a Log 2 FC ≥ 1, FDR = 0.001 cut-off, 1091 upregulated loci and 1073 downregulated loci were detected (Supplementary Materials Dataset S1) Downregulated loci include metabolic, flagellar and chemotaxis genes as well as loci belonging to Salmonella pathogenicity islands 1 and 2. Downregulation was not investigated further. SOS regulon candidates were shortlisted by identifying genes that were upregulated in the presence of nalidixic acid and harbored a SOS box at maximum distance of 400 base pairs upstream of the start codon. This cross-reference identified 47 genes (Table 1), of which 25 had been identified previously as SOS response genes either S. enterica or E. coli [ 2 , 9 , 42 ]. The tisB and istR-1,2 genes, which encode the TisB-IstR toxin antitoxin system, had also been described as LexA-regulated [ 45 ]. Our analysis also identified prophage genes known to be regulated by LexA, such as the dinI homologues STM14_3210, STM14_1156 and STM14_1439 encoded by Gifsy-1, Gifsy-2 and Gifsy-3, respectively [ 46 ]. Other prophageborne candidates identified in our analysis were STM14_3214 and STM14_1432, which encode phage replication proteins of Gifsy-1 and Gifsy-3, respectively [ 47 ]. Genes that were not known to be part of the SOS regulon were also identified as loci under putative LexA control. Examples include cysP, which encodes a thiosulfate-binding protein [ 48 ]; the phosphate metabolism gene yqaB [ 49 ]; and hupA, which encodes the alpha subunit of the nucleoid-associated protein HU [ 50 ]. Certain candidate genes that are divergently transcribed appeared to share an SOS box with a neighbour (e.g., ssb and uvrA, and nlhH and higB-2). Table 1. SOS response genes of S. enterica identified as loci that harbor putative LexA boxes and are upregulated in the presence of nalidixic acid. Gene ID Gene Name Fold Change Number of SOS Boxes Distance to ATG (bp) Heterology Index * STM14_4584 tisB 610.09 1 236 6.13 STM14_2287 yebG 41.34 1 17 7.03 STM14_3289 recN 39.41 3 8, 26, 48 12.97, 7.72, 8.82 STM14_3417 recA 35.71 1 64 5.92 STM14_1215 sulA 26.70 1 21 1.49 STM14_3210 dinIGifsy-1 23.20 1 19 4.55 STM14_1331 dinI 22.82 2 19, 42 4.00, 17.60 STM14_2423 umuD ** 20.10 1 15 4.38 STM14_4775 yigN 19.46 1 44 8.71 STM14_3002 cysP 15.47 1 65 18.16 STM14_4846 nlhH 15.11 1 13 7.63 STM14_1156 dinIGifsy-2 14.46 1 19 3.92 STM14_0926 uvrB 14.13 1 73 5.66 Cells 2021,10, 943 5 of 17 Table 1. Cont. Gene ID Gene Name Fold Change Number of SOS Boxes Distance to ATG (bp) Heterology Index * STM14_4752 uvrD 13.87 1 105 11.48 STM14_0369 dinP *** 11.30 1 15 6.80 STM14_5112 uvrA 11.16 1 80 6.99 STM14_953 dinG 10.61 1 14 10.08 IstR-1,2 istR-1,2 10.26 1 40 6.13 STM14_5114 ssb 9.63 1 24 6.99 STM14_4847 higB-2 8.05 1 184 7.63 STM14_4236 dinJ 7.82 1 14 8.27 STM14_3214 –7.70 1 6 12.33 STM14_5094 lexA 6.86 2 6, 27 14.48, 7.94 STM14_3568 gudD 5.43 1 0 16.97 STM14_3405 yqaB 5.32 1 12 16.26 STM14_1439 dinIGifsy-3 4.56 1 19 5.26 STM14_2752 yejK 4.54 1 102 20.89 STM14_2422 umuC ** 3.91 1 – 4.38 STM14_2753 yejL 3.74 1 57 20.89 STM14_2648 thiM 3.62 1 177 18.02 STM14_3627 mutH 3.58 1 228 12.99 STM14_2551 sbmC 3.55 1 21 6.32 STM14_1492 msgA 3.44 1 14 11.88 STM14_2650 rcnB 3.38 1 212 18.02 STM14_1589 ydjQ 3.33 1 0 6.69 STM14_0161 yacA 3.30 1 403 19.46 STM14_4344 yhjE 3.05 1 202 14.01 STM14_5011 hupA 2.74 1 190 15.31 STM14_1432 –2.71 1 6 12.33 STM14_4158 slyX 2.70 1 146 21.28 STM14_3283 corE 2.54 1 98 20.11 STM14_1385 ndh 2.37 1 157 12.04 STM14_0374 frsA 2.33 1 360 13.12 STM14_5490 deoD 2.21 1 210 18.78 STM14_5518 –2.19 1 71 17.26 STM14_1605 ydjM 2.11 2 14, 32 9.12, 12.85 * Defined in Section 3.4. ** umuD and umuC form and operon, and their SOS box is located upstream of umuD. *** Homologue of E. coli dinB. Thirty-one genes that harbored a LexA box at a maximum distance of approximately 100 bp were chosen for further analysis. Genes co-induced with adjacent canonical SOS regulon gene(s) through transcriptional read-through [ 8 , 51 ] were not included (e.g., yebF and yebE). The SOS gene ruvA was manually added to the list as it had not appeared among the loci upregulated by nalidixic acid. Genes ftsY and ybfE, identified as SOS genes by Erill et al. [ 2 ], were upregulated by nalidixic acid but their SOS boxes were not identified in our search. Cells 2021,10, 943 6 of 17 3.3. Single Cell Analysis of SOS Gene Expression in the Absence of SOS Induction To analyze at the single cell level the expression patterns of the SOS genes under study, a transcriptional fusion with the green fluorescent protein (gfp) was constructed downstream of the stop codon of each gene. After preliminary validation, certain genes were excluded from further analysis due to lack of fluorescence (dinG, ydjQ and ydjM) or because flow cytometry failed to detect induction by nalidixic acid (cysP and gudD). Failure to detect expression does not rule out that these genes may be part of the SOS regulon. In the remaining genes under study, flow cytometry analysis detected heterogeneous expression in LB (that is, in the absence of any known DNA damaging agent). Three main gene groups can be tentatively distinguished: (i). Group I, which is the most numerous (12 loci) presents bistable expression: the bacterial population contains a major subpopulation in which the gene is completely OFF and a smaller subpopulation of cells in which the gene is expressed (ON). The size of the ON subpopulation varies from one gene to another, ranging from ~8.7% in sulA to 1.3% in dinP and 0.8% and nlhH. In all loci, the ON subpopulation is absent in lexA (Ind − ) and recA backgrounds. Expression of lexA,recA,umuDC and colicin genes in non-inducing conditions has been described previously in E. coli [ 52 – 54 ], and the formation of ON subpopulations under such conditions has been proposed to be triggered by activation of the SOS response upon spontaneous DNA strand breakage [ 53 ]. In agreement with this view, our analysis of gfp fusions did not detect ON cells in a recA background (Figure 1), which seems to rule out the possibility that the subpopulation might be produced by spontaneous alleviation of LexA repression (if that were the case, an ON subpopulation should be still detected in a recA mutant). Furthermore, the size of the ON subpopulation increased in a recD background and to a lesser extent in a recF background (Supplementary Figure S3), in agreement with the major role played by the RecBCD recombination pathway in double-strand break repair [55,56]. (ii). Group II genes show “noisy”, heterogeneous expression that splits the population into OFF and ON subpopulations. The percentages of ON cells are larger than in genes of group I (from ~13% in uvrD up to ~85% in ssb). However, the main difference with group I is that formation of ON cells remains unaltered in lexA (Ind − ) and recA backgrounds (Figure 2). We thus conclude that expression of these genes under non-inducing conditions is not under LexA control. Because this group contains housekeeping genes, a tentative interpretation is that the genes may be active in a subpopulation of cells in the absence of DNA damage. It is also possible that cells in the OFF state might have expression levels below the threshold for experimental detection. (iii). Genes classified into group III show heterogeneous expression that is not bimodal, and two ON subpopulations are detected. However, these subpopulations differ in their expression level, and only the ON subpopulation with higher expression level disappears in a lexA (Ind − ) mutant (Figure 3). The latter observation has a paradoxical side as formation of the “high ON” subpopulation appears to be LexA-dependent. Interestingly, the lexA gene belongs to this group. Expression of lexA in a lexA (Ind − ) background was not tested because a lexA::gfp fusion could not be constructed. Loci that show noisy expression with a pattern opposite to that of group III are also detected (Supplementary Figure S4). In these loci, a subpopulation with higher expression levels appears in a lexA (Ind − ) background only. No gene in this group has been previously reported as a member of the SOS regulon; however, they harbor a LexA box and are activated by DNA damage. Therefore, one cannot discard that they may have SOS-related functions that remain to be identified [57]. Cells 2021,10, 943 7 of 17 Figure 1. LexA-dependent bistability of group I loci in the absence of known DNA damage. The dashed line shows the boundary between OFF and ON states, established as the expression level of a non-fluorescent sample (Supplementary Figure S2). Figure 2. LexA-independent gene expression heterogeneity in group II genes. The dashed line shows the boundary between OFF and ON states, established as the expression level of a non-fluorescent sample (Supplementary Figure S2). Figure 3. Formation of LexA dependent and LexA-independent ON subpopulations in genes of group III. The pink dashed line shows the boundary between OFF and ON states, established as the expression level of a non-fluorescent sample (Supplementary Figure S2). Cells 2021,10, 943 8 of 17 A SOS response gene with a unique expression pattern above is recA (Supplementary Figure S5). Expression in virtually all cells is consistent with the pleiotropic role of RecA in genome maintenance. 3.4. Influence of the Distance and the Nucleotide Sequence of the LexA Box on the Expression Patterns of SOS Genes The existence of various expression patterns under non-inducing conditions led us to investigate the underlying cause. For this purpose, we made the reductionist hypothesis that pattern divergence might be caused by differences in the LexA box. When we examined the distance between the LexA box and the start codon of the cognate gene, we found that bistable expression under LexA control (found in groups 1 and 3) correlates with shorter distances (21.6 ± 5.8 bp). In contrast, genes whose bistability is LexA-independent (group 2) tend to have their LexA box further away (58.8 ± 11.6 bp) (Figure 4A). In genes where the transcription start site (TSS) is known [ 33 ], the LexA boxes of genes with LexA-dependent bistability overlap with the TSS. This analysis was made omitting prophage genes, and ssb and uvrA were also excluded because their LexA box is shared. Figure 4. ( A ). Correlation between LexA-dependent and LexA-independent bistability and the distance of the SOS box. Statistically significant differences are shown (p-value: * p< 0.05; ** p< 0.01). ( B ). Correlation between LexA-dependent and LexA-independent bistability and the heterology index (HI) of the SOS box. In genes that show LexA-dependent bistability and have more than one LexA box (recN,dinI and lexA), only the closest LexA box with lower HI was considered. The sequence of LexA boxes was also examined as their relatedness to the consensus determines the binding affinity of the LexA repressor [ 58 ]. Divergence from the consensus LexA box was measured using a mathematical formula described by Berg and von Hippel [ 38 , 59 ], and the resulting score was expressed as an ‘heterology index’ (HI) as defined by Lewis et al. [ 60 ]. Genes that presented LexA-dependent bistability had lower HI values (6.29 ± 0.87), while higher HI’s (10.04 ± 1.1) were common among genes with LexA-independent bistability (Figure 4B). Altogether, these observations indicate that both the location and the sequence of the LexA box determine the expression pattern of the cognate gene under non-inducing conditions. After observing an influence of the heterology index of the LexA box on the gene expression pattern, we wondered if there was any relationship between the predicted LexA affinity and the size of the subpopulation of cells that expressed the gene under non-inducing conditions. Contrary to our expectations, genes with lower HI (and therefore with predicted higher LexA affinity) were those that presented larger ON subpopulations (Figure 5). Cells 2021,10, 943 9 of 17 Figure 5. Correlation between the HI values of the SOS boxes and the size of the ON subpopulations of genes with LexA-dependent bistability (group I). Because this observation was puzzling, we took advantage of the existence of a single nucleotide polymorphism in the LexA box of the yebG gene in S. enterica ATCC 14028 (the strain used in this work) and another model strain, ST4/74 (Figure 6A). The SNP, a G → C substitution, lowers the HI of the yebG SOS box of ST4/74 down to 4.98, compared with 7.03 in ATCC 14028. When we analyzed yebG::GFP expression by flow cytometry, the ON subpopulation was higher in ST4/74 ON, thus strengthening the unsuspected observation that genes with LexA-dependent bistability produce larger ON subpopulations in the absence of DNA damage if the LexA box has a low HI (Figure 6B). Figure 6. Comparison of YebG ON subpopulation sizes between S. Typhimurium strains ATCC 14028 and ST4/74. ( A ). Single-nucleotide polymorphism in the yebG SOS boxes of ATCC 14028 and ST4/74. HI values are indicated. ( B ). Percentages of YebG ON cells in ATCC 14028 and 4/74. ( C ). Percentages of YebG ON cells in in ATCC 14028 and 4/74 derivatives whose SOS boxes have been swapped. Statistically significant differences are indicated (p-value; * p< 0.05, ** p< 0.01, *** p< 0.001, **** p< 0.0001). To further probe this unsuspected conclusion, the SNPs were swapped in both strains. Single cell analysis revealed that a yebGATCC 14028 gene with an SNPST4/74 produced more cells in the ON state, while the introduction of SNP ATCC 14028 into the yebG ST4/74 gene lowered the percentage of ON cells (Figure 6C). Hence, the nucleotide sequence of the Cells 2021,10, 943 16 of 17 31. Wagner, G.P.; Kin, K.; Lynch, V.J. Measurement of mRNA abundance using RNA-seq data: RPKM measure is inconsistent among samples. Theory Biosci. 2012,131, 281–285. [CrossRef] 32. Wagner, G.P.; Kin, K.; Lynch, V.J. A model based criterion for gene expression calls using RNA-seq data. Theory Biosci. 2013 ,132, 159–164. [CrossRef] 33. Kroger, C.; Colgan, A.; Srikumar, S.; Handler, K.; Sivasankaran, S.K.; Hammarlof, D.L.; Canals, R.; Grissom, J.E.; Conway, T.; Hokamp, K.; et al. An infection-relevant transcriptomic compendium for Salmonella enterica serovar Typhimurium. Cell Host Microbe 2013,14, 683–695. [CrossRef] 34. Law, C.W.; Chen, Y.; Shi, W.; Smyth, G.K. voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014,15, R29. [CrossRef] 35. Ritchie, M.E.; Phipson, B.; Wu, D.; Hu, Y.; Law, C.W.; Shi, W.; Smyth, G.K. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015,43, e47. [CrossRef] 36. Robinson, M.D.; McCarthy, D.J.; Smyth, G.K. edgeR: A Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 2010,26, 139–140. [CrossRef] 37. Grant, C.E.; Bailey, T.L.; Noble, W.S. FIMO: Scanning for occurrences of a given motif. Bioinformatics 2011 ,27, 1017–1018. [CrossRef] 38. Berg, O.G. Selection of DNA binding sites by regulatory proteins: The LexA protein and the arginine repressor use different strategies for functional specificity. Nucleic Acids Res. 1988,16, 5089–5105. [CrossRef] 39. Hautefort, I.; Proenca, M.J.; Hinton, J.C. Single-copy green fluorescent protein gene fusions allow accurate measurement of Salmonella gene expression in vitro and during infection of mammalian cells. Appl. Environ. Microbiol. 2003 ,69, 7480–7491. [CrossRef] 40. Datsenko, K.A.; Wanner, B.L. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 2000,90, 6640–6645. [CrossRef] 41. Lehtinen, J.; Nuutila, J.; Lilius, E.M. Green fluorescent protein-propidium iodide (GFP-PI) based assay for flow cytometric measurement of bacterial viability. Cytom. A 2004,60, 165–172. [CrossRef] [PubMed] 42. Fernandez De Henestrosa, A.R.; Ogi, T.; Aoyagi, S.; Chafin, D.; Hayes, J.J.; Ohmori, H.; Woodgate, R. Identification of additional genes belonging to the LexA regulon in Escherichia coli.Mol. Microbiol. 2000,35, 1560–1572. [CrossRef] [PubMed] 43. Ysern, P.; Clerch, B.; Castano, M.; Gibert, I.; Barbe, J.; Llagostera, M. Induction of SOS genes in Escherichia coli and mutagenesis in Salmonella typhimurium by fluoroquinolones. Mutagenesis 1990,5, 63–66. [CrossRef] [PubMed] 44. Newmark, K.G.; O’Reilly, E.K.; Pohlhaus, J.R.; Kreuzer, K.N. Genetic analysis of the requirements for SOS induction by nalidixic acid in Escherichia coli.Gene 2005,356, 69–76. [CrossRef] 45. Vogel, J.; Argaman, L.; Wagner, E.G.; Altuvia, S. The small RNA IstR inhibits synthesis of an SOS-induced toxic peptide. Curr. Biol. 2004,14, 2271–2276. [CrossRef] 46. Lemire, S.; Figueroa-Bossi, N.; Bossi, L. Bacteriophage crosstalk: Coordination of prophage induction by trans-acting antirepressors. PLoS Genet 2011,7, e1002149. [CrossRef] 47. Figueroa-Bossi, N.; Coissac, E.; Netter, P.; Bossi, L. Unsuspected prophage-like elements in Salmonella typhimurium.Mol. Microbiol. 1997,25, 161–173. [CrossRef] 48. Hryniewicz, M.; Sirko, A.; Palucha, A.; Bock, A.; Hulanicka, D. Sulfate and thiosulfate transport in Escherichia coli K-12: Identification of a gene encoding a novel protein involved in thiosulfate binding. J. Bacteriol. 1990,172, 3358–3366. [CrossRef] 49. Zhang, Z.; Zhang, Q.; Guan, S.; Shi, H. Quantitative connection between cell size and growth rate by phospholipid metabolism. Cells 2020,9, 391. [CrossRef] 50. Hillyard, D.R.; Edlund, M.; Hughes, K.T.; Marsh, M.; Higgins, N.P. Subunit-specific phenotypes of Salmonella typhimurium HU mutants. J. Bacteriol. 1990,172, 5402–5407. [CrossRef] 51. Krin, E.; Pierle, S.A.; Sismeiro, O.; Jagla, B.; Dillies, M.A.; Varet, H.; Irazoki, O.; Campoy, S.; Rouy, Z.; Cruveiller, S.; et al. Expansion of the SOS regulon of Vibrio cholerae through extensive transcriptome analysis and experimental validation. BMC Genom. 2018,19, 373. [CrossRef] [PubMed] 52. McCool, J.D.; Long, E.; Petrosino, J.F.; Sandler, H.A.; Rosenberg, S.M.; Sandler, S.J. Measurement of SOS expression in individual Escherichia coli K-12 cells using fluorescence microscopy. Mol. Microbiol. 2004,53, 1343–1357. [CrossRef] [PubMed] 53. Pennington, J.M.; Rosenberg, S.M. Spontaneous DNA breakage in single living Escherichia coli cells. Nat. Genet 2007 ,39, 797–802. [CrossRef] [PubMed] 54. Kamensek, S.; Podlesek, Z.; Gillor, O.; Zgur-Bertok, D. Genes regulated by the Escherichia coli SOS repressor LexA exhibit heterogeneous expression. BMC Microbiol. 2010,10, 283. [CrossRef] 55. Kuzminov, A.; Stahl, F.W. Double-strand end repair via the RecBC pathway in Escherichia coli primes DNA replication. Genes Dev. 1999,13, 345–356. [CrossRef] 56. Kowalczykowski, S.C. Initiation of genetic recombination and recombination-dependent replication. Trends Biochem. Sci. 2000 ,25, 156–165. [CrossRef] 57. D’Ari, R.; Casadesus, J. Underground metabolism. Bioessays 1998,20, 181–186. [CrossRef] 58. Butala, M.; Zgur-Bertok, D.; Busby, S.J. The bacterial LexA transcriptional repressor. Cell Mol. Life Sci. 2009 ,66, 82–93. [CrossRef] 59. von Hippel, P.H.; Berg, O.G. Facilitated target location in biological systems. J. Biol. Chem. 1989,264, 675–678. [CrossRef] Cells 2021,10, 943 17 of 17 60. Lewis, L.K.; Harlow, G.R.; Gregg-Jolly, L.A.; Mount, D.W. Identification of high affinity binding sites for LexA which define new DNA damage-inducible genes in Escherichia coli.J. Mol. Biol. 1994,241, 507–523. [CrossRef] 61. Barbe, J.; Villaverde, A.; Guerrero, R. Induction of the SOS response by hydroxyurea in Escherichia coli K12. Mutat. Res. 1987 ,192, 105–108. [CrossRef] 62. Otsuji, N.; Sekiguchi, M.; Iijima, T.; Takagi, Y. Induction of phage formation in the lysogenic Escherichia coli K-12 by mitomycin C. Nature 1959,184, 1079–1080. [CrossRef] 63. Imlay, J.A.; Linn, S. Mutagenesis and stress responses induced in Escherichia coli by hydrogen peroxide. J. Bacteriol. 1987 ,169, 2967–2976. [CrossRef] 64. Vila, J.; Fabrega, A.; Roca, I.; Hernandez, A.; Martinez, J.L. Efflux pumps as an important mechanism for quinolone resistance. Adv. Enzymol. Relat. Areas Mol. Biol. 2011,77, 167–235. 65. Yang, L.; Mih, N.; Anand, A.; Park, J.H.; Tan, J.; Yurkovich, J.T.; Monk, J.M.; Lloyd, C.J.; Sandberg, T.E.; Seo, S.W.; et al. Cellular responses to reactive oxygen species are predicted from molecular mechanisms. Proc. Natl. Acad. Sci. USA 2019 ,116, 14368–14373. [CrossRef] 66. Xia, J.; Chiu, L.Y.; Nehring, R.B.; Bravo Nunez, M.A.; Mei, Q.; Perez, M.; Zhai, Y.; Fitzgerald, D.M.; Pribis, J.P.; Wang, Y.; et al. Bacteria-to-human protein networks reveal origins of endogenous DNA damage. Cell 2019,176, 127–143 e24. [CrossRef] 67. Sigal, N.; Delius, H.; Kornberg, T.; Gefter, M.L.; Alberts, B. A DNA-unwinding protein isolated from Escherichia coli: Its interaction with DNA and with DNA polymerases. Proc. Natl. Acad. Sci. USA 1972,69, 3537–3541. [CrossRef] 68. Shinagawa, H.; Iwasaki, H. Processing the holliday junction in homologous recombination. Trends Biochem. Sci. 1996 ,21, 107–111. [CrossRef] 69. Seeberg, E.; Nissen-Meyer, J.; Strike, P. Incision of ultraviolet-irradiated DNA by extracts of E. coli requires three different gene products. Nature 1976,263, 524–526. [CrossRef] 70. Culyba, M.J.; Kubiak, J.M.; Mo, C.Y.; Goulian, M.; Kohli, R.M. Non-equilibrium repressor binding kinetics link DNA damage dose to transcriptional timing within the SOS gene network. PLoS Genet 2018,14, e1007405. [CrossRef] 71. Davidson, C.J.; Surette, M.G. Individuality in bacteria. Annu. Rev. Genet 2008,42, 253–268. [CrossRef] [PubMed] 72. Silva-Rocha, R.; de Lorenzo, V. Noise and robustness in prokaryotic regulatory networks. Annu. Rev. Microbiol. 2010 ,64, 257–275. [CrossRef] [PubMed] 73. Goni-Moreno, A.; Benedetti, I.; Kim, J.; de Lorenzo, V. Deconvolution of gene expression noise into spatial dynamics of transcription factor-promoter Interplay. ACS Synth. Biol. 2017,6, 1359–1369. [CrossRef] [PubMed] 74. Kaern, M.; Elston, T.C.; Blake, W.J.; Collins, J.J. Stochasticity in gene expression: From theories to phenotypes. Nat. Rev. Genet 2005,6, 451–464. [CrossRef] [PubMed] 75. Silander, O.K.; Nikolic, N.; Zaslaver, A.; Bren, A.; Kikoin, I.; Alon, U.; Ackermann, M. A genome-wide analysis of promotermediated phenotypic noise in Escherichia coli.PLoS Genet 2012,8, e1002443. [CrossRef] 76. Woo, A.C.; Faure, L.; Dapa, T.; Matic, I. Heterogeneity of spontaneous DNA replication errors in single isogenic Escherichia coli cells. Sci. Adv. 2018,4, eaat1608. [CrossRef] 77. Pribis, J.P.; Garcia-Villada, L.; Zhai, Y.; Lewin-Epstein, O.; Wang, A.Z.; Liu, J.; Xia, J.; Mei, Q.; Fitzgerald, D.M.; Bos, J.; et al. Gamblers: An antibiotic-induced evolvable cell subpopulation differentiated by reactive-oxygen-induced general stress response. Mol. Cell 2019,74, 785–800 e7. [CrossRef] 78. Uphoff, S.; Lord, N.D.; Okumus, B.; Potvin-Trottier, L.; Sherratt, D.J.; Paulsson, J. Stochastic activation of a DNA damage response causes cell-to-cell mutation rate variation. Science 2016,351, 1094–1097. [CrossRef] 79. Vincent, M.S.; Uphoff, S. Bacterial phenotypic heterogeneity in DNA repair and mutagenesis. Biochem. Soc. Trans. 2020 ,48, 451–462. [CrossRef] 80. Ackermann, M. A functional perspective on phenotypic heterogeneity in microorganisms. Nat. Rev. Microbiol. 2015 ,13, 497–508. [CrossRef] 81. van Boxtel, C.; van Heerden, J.H.; Nordholt, N.; Schmidt, P.; Bruggeman, F.J. Taking chances and making mistakes: Non-genetic phenotypic heterogeneity and its consequences for surviving in dynamic environments. J. R. Soc. Interface 2017 ,14, 20170141. [CrossRef] [PubMed] 82. Schroter, L.; Dersch, P. Phenotypic diversification of microbial p athogens—cooperating and preparing for the future. J. Mol. Biol. 2019,431, 4645–4655. [CrossRef] [PubMed] 83. Zgur-Bertok, D. Phenotypic heterogeneity in bacterial populations. Acta Agric. Slov. 2007,90, 17–24. 84. Hernandez, S.B.; Cota, I.; Ducret, A.; Aussel, L.; Casadesus, J. Adaptation and preadaptation of Salmonella enterica to bile. PLoS Genet 2012,8, e1002459. [CrossRef]