The bacterial SOS response promotes the expression of the transposase encoded by ISCR mobile genetic elements Authors: Claire Lallement, Thomas Jové, Cécile Pasternak, Sandra Da Re, Marie-Cécile Ploy Published: 2026 ABSTRACT ABSTRACTInsertion sequences (IS) are widely involved in bacterial genomic plasticity by disrupting, adding, moving genomic sequences, or by activating or extinguishing gene expression. A specific family of IS, ISCR (for insertion sequence of Common Region), is thought to be involved in the dissemination of antibiotic resistance genes (ARGs). While some ISCR members are commonly found in bacteria isolated in clinical settings and can contribute to downstream ARG expression, the mechanisms regulating the ISCR-encoded transposase expression have remained uncharacterized. Here, we investigated the expression of the transposase genes of ISCR1, ISCR2, and ISCR8 and its regulation in Escherichia coli. Using in silico analyses and in vitro experiments, we showed that the expression levels were extremely low, as observed for most IS transposases. We further demonstrated the direct role of DNA damages and the key SOS response repressor, LexA, in controlling the activity of the transposase promoter. These results provide evidence that the mobility of at least some ISCR elements may be promoted upon bacterial exposure to antibiotics inducing the SOS response.IMPORTANCEMobile genetic elements (MGEs) are the most prevalent cause of antibiotic resistance emergence. Among these mobile elements, insertion sequences (IS) are well known to allow the dissemination of antibiotic resistance genes (ARGs) through the action of their transposase. Here, we studied the regulation of the transposase expression in a specific family of IS, the ISCR family, some members of which are known to be involved in antibiotic resistance. Characterizing the regulation of transposase expression is an important starting point for understanding how these IS can contribute, through their movement, to the spread and expression of antibiotic resistance. INTRODUCTION Bacterial mobile genetic elements (MGEs) are key players in the dissemination of antibiotic resistance genes (ARGs) through horizontal gene transfer (1). Insertion sequences (IS) are the most widespread type of MGEs and encode all necessary material for their transposition (2). IS can differently impact the physiology of their host cells by moving, disrupting, or activating genes (3). IS elements are generally tightly regulated through diverse mechanisms acting at different levels, including transcriptional, translational, and/or post-translational levels (reviewed in reference 4). For instance, the transposition of IS10 is controlled by the SOS response in Escherichia coli (5), the transposition of IS2 by the global regulator CRP (cAMP receptor protein) (6), and the transposition of IS50 is regulated by Hfq (mediated through CRP) and the sRNA SgrS (7). IS of the Insertion Sequence Common Region (ISCR) group are of clinical concern since they are commonly adjacent to multiple ARGs or pathogenicity regions (8). ISCR elements are related to the IS91 family members with which they share several features. These include (i) a gene encoding a transposase of the HUH superfamily of single-strand nuclease (hereafter referred to as rcr, according to reference 9) and (ii) two distinct ends that include sequences forming secondary structures, namely, oriIS and terIS, where the transposition process initiates and terminates, respectively (8, 10) (Fig. 1A). These specificities suggest that ISCR elements may transpose by a rolling-circle mechanism as presumed for IS91 and can therefore transpose adjacent genes (8, 11). Structure of ISCR and alignment of their terIS regions (A) General structure of ISCR. ISCR consists of an rcr transposase gene and its promoter Prcr, bound by two pairs of inverted repeats, namely, terIS and oriIS. (B) Alignment of terIS regions. Putative −35 and −10 σ70 elements of rcr promoters of interest are shown in bold and highlighted in blue. Putative LexA box is boxed in red. Aligned sequences include the terIS region (up to the rcr start codon) limited to 150 bp. Alignment made with Clustal Omega (https://www.ebi.ac.uk/jdispatcher/msa/clustalo) with default parameters. (C) Comparison of LexA operators in ISCR1, ISCR2, and ISCR8. Weblogo was used to generate the LexA box consensus sequence (12) from a set of 28 experimentally verified Escherichia coli LexA-binding sites (listed in reference 13). Bases identical in each of the three predicted ISCR LexA-binding sites and matching the LexA box logo are given in bold. Each color represents a nucleotide (blue: cystosine, red: thymine, yellow: guanine, green: adenine) (D) Transcriptional fusion used to assess Prcr activity in this study. Reporter gene lacZ, encoding for β-galactosidase, was fused in a translational fusion with native (Prcr), LexA-box-mutated (PrcrL), or −10 region-mutated (PrcrM) promoters of ISCR1, ISCR2, and ISCR8. The prevailing ISCR element, namely, ISCR1, is exclusively found downstream of the 3′ conserved region associated with class 1 integrons and displays a large diversity of adjacent genes, especially ARGs (8, 9). Moreover, ISCR1 harbors two outwardly oriented promoters (POUT) in its oriIS end (14–16), contributing to the expression of downstream ARGs (9). Additionally, two other ISCRs commonly found in the literature and databases are (i) ISCR2, which is mainly described adjacent to ARG within integrative conjugative elements (ICEs), especially the SXT element (17), and (ii) ISCR8, which is found as multiple copies in environmental bacterial plasmids (18). Despite their potential importance in clinical settings, nothing is known about the regulation of the expression of ISCR elements. In this study, we characterized, both in silico and in vitro, the key DNA motifs involved in the regulated expression of the rcr transposase-encoding genes of three members of the ISCR group (ISCR1, -CR2, and -CR8) in E. coli. Specifically, we demonstrate that the SOS response acts as a crucial pathway to tightly control the expression of these ISCR transposases, via the LexA repressor. RESULTS AND DISCUSSION LexA binds to the putative promoter region of three ISCR transposase (rcr) genes To investigate the expression of ISCR transposase genes (rcr genes), we first analyzed in silico the terIS region sequences encompassing the promoter region of the rcr gene (Fig. 1A and B) of 17 nonredundant ISCR elements in E. coli (see Fig. S1). Interestingly, ISCR1, -CR2, and -CR8 comprised a sequence matching the CTGT-N8-ACAG β-/γ-proteobacteria LexA-binding site consensus (LexA box) (19, 20), which overlaps with the −10 element of the predicted rcr transposase gene promoters (Prcr) (Fig. 1B and C), suggesting a potential regulation of these promoters by the repressor LexA. The LexA transcriptional factor is the master regulator of the SOS response, a global bacterial response to DNA damage (reviewed in reference 20), that acts as a repressor of SOS regulon genes under normal conditions. Upon DNA damages and the formation of single-strand DNA (ssDNA), the SOS response is triggered, through the formation of a nucleoprotein filament ssDNA-RecA that stimulates LexA autoproteolysis and, therefore, the expression of SOS-regulated genes. To confirm the functionality of the predicted LexA box in ISCR1, -CR2, and -CR8, we performed electrophoretic mobility shift assay (EMSA) of their rcr transposase promoter regions with purified E. coli LexA protein. We observed a DNA shift in the presence of the LexA protein for the three promoters Prcr (Fig. 2), showing that LexA specifically binds in vitro to the putative ISCR LexA box located in the native ISCR1, -CR2, and -CR8 Prcr promoter regions. Conversely, no shift was detected when the putative ISCR LexA boxes were mutated in one of the conserved terminal triplet (CTGT-N8-ACAG into CTGT-N8-AACT; PrcrL) (Fig. 2). These results strongly suggest that LexA plays a role in regulating the expression of the corresponding rcr transposase gene and show that the sequences identified in silico are functional LexA boxes. LexA binds to the rcr promoter region of ISCR1, ISCR2, and ISCR8. Electrophoretic mobility shift assays were performed with native rcr promoters (Prcr ; lanes 1–5) or LexA box-mutated rcr promoters (Prcr L; lanes 6–7) of rcr1 (A), rcr2 (B), or rcr8 (C), in the presence or absence of purified LexA protein (amounts in nanograms are indicated). F, free DNA; D, delayed complex. The SOS response induces ISCR1, ISCR2, and ISCR8 transposase expression To characterize the SOS regulation of the rcr transposase genes in vivo, we constructed lacZ transcriptional fusions between the lacZ gene with its RBS and the terIS regions of ISCR1, ISCR2, and ISCR8, including the putative rcr promoter (Prcr) sequence and LexA-binding site (Fig. 1D). We also constructed these fusions containing a mutation of the putative −10 box of the Prcr promoters (Prcr M) or the mutation in the LexA box (Prcr L) (Table 1). TABLE 1Bacterial strains and plasmids used in this studyStrain or plasmidGenotype or descriptionSource or referenceBacterial strains Escherichia coli MG1656lacZ- derivative of E. coli K-12 MG1655(21) MG1656ΔlexADerivative of MG1656 with deletions of sulA and lexA(22) MG1656ΔrecADerivative of MG1656 with deletion of recA(23) MG1656lexA3(Ind-)Derivative of MG1656 carrying the lexA3(Ind-) allele encoding a non-cleavable LexA proteinLaboratory collection BL21(DE3)/pLysSF– ompT gal dcm lon hsdS B(rB–mB–) (DE3), carrying the pLysS plasmid(24) BW25113/ICEPdaSpaI (ECI)Strain carrying ISCR2(25) Salmonella enterica serovar Montevideo C1Strain carrying ISCR1Laboratory collection Pseudomonas sp. ADP2 PADStrain carrying ISCR8 (26)Plasmids pSU38ΔtotlacZKmr, vector carrying the promoterless lacZ gene(27) Prcr1369-bp terIS region of ISCR1 and lacZ translation initiation signal cloned into pSU38ΔtotlacZ (oligonucleotides CP69 and CP70 and C1 strain DNA)This study Prcr1 MPrcr1 mutated in the −10 element of Prcr1 with oligonucleotides CP69 and 4This study Prcr1 LPrcr1 mutated in the LexA box of Prcr1 with oligonucleotides CP69 and CP71This study Prcr2106-bp terIS of ISCR2 and lacZ translation initiation signal cloned into pSU38ΔtotlacZ (oligonucleotides Prcr2L-EcoRI and CP72 and ECI strain DNA)This study Prcr2 MPrcr2 mutated in the −10 element of Prcr2 with oligonucleotides Prcr2-10mutL and Prcr2-10mutRThis study Prcr2 LPrcr2 mutated in the LexA box of Prcr2 with oligonucleotides Prcr2L-EcoRI and CP73This study Prcr887-bp terIS of ISCR8 and lacZ translation initiation signal cloned into pSU38ΔtotlacZ (oligonucleotides Prcr8L and Prcr8-lacZ and PAD strain DNA)This study Prcr8 MPrcr8 mutated in the −10 element of Prcr8 with oligonucleotides 15 and 16This study Prcr8 LPrcr8 mutated in the LexA box of Prcr8 with oligonucleotides Prcr8LexAmut2L and Prcr8LexAmut2RThis study pUA1107ppET15b with PT7lac::His6::lexAE.coli-termT7 pET expression system; PT7lac, N-terminal His6 tag, and Ampr(23) PsulAsulA gene promoter region and lacZ translation initiation signal cloned into pSU38ΔtotlacZ(28) Bacterial strains and plasmids used in this study The resulting recombinant plasmids were introduced into the E. coli K-12 MG1656 strain, its ΔlexA (constitutive expression of LexA-regulated promoters), and ΔrecA (constitutive repression of LexA-regulated promoters) derivatives, as well as in a lexA3(Ind-) mutant encoding for an uncleavable LexA protein (29). Measuring β-galactosidase activities showed very low activity for all three promoters in the wild-type (WT) strain (<5 Miller units), suggesting that Rcr transposases are not constitutively expressed (Fig. 3A through C). These results are consistent with those in the literature. Indeed, weak promoter activity is a feature shared by many transposase gene promoters, including those from IS21 (30), IS30 (31), IS911 (32), and IS200 (33), to limit their impact on bacterial fitness. We observed a significant increase in β-galactosidase activity in the lexA-deleted strain, with varying strength for the three native promoters Prcr1, Prcr2, and Prcr8 (18.2-, 1.6-, and 4.1-fold, respectively; Fig. 3A through C and Table S1). However, we did not observe a strong effect in conditions where the SOS response could not be induced (strains ΔrecA and lexA3(Ind-)), with basal Prcr activities being very low (<5 Miller units) in the WT strain (Fig. 3D). Overall, the results are consistent with the potential regulation of Prcr1, Prcr2, and Prcr8 by LexA. To further validate these results, we measured the β-galactosidase activities from PrcrL-mutated promoters (no binding of LexA). As expected, the mutation of the LexA-binding site led to a 48.5-, 14.5-, and 10.6- fold increase in β-galactosidase activities for Prcr1, Prcr2, and Prcr8, respectively (Fig. 3A through C and Table S1). These results demonstrate the involvement of LexA protein in the regulation of ISCR1, CR2, and CR8 rcr transposase gene promoters. The SOS response controls the expression of ISCR1, ISCR2, and ISCR8 transposase rcr genes. (A–D) LexA represses the expression of the transposase genes. β-galactosidase activities (in Miller units mU) were measured in the E. coli MG1656 reference strain (gray) and its ΔlexA (red), ΔrecA (light blue), and lexAind3- (non-autoproteolytic LexA; turquoise) derivatives. These strains carry a plasmid containing a lacZ transcriptional fusion with the rcr promoter of ISCR1 (A, D), ISCR2 (B, D), or ISCR8 (C, D) in either the native (Prcr), LexA-box-mutated (PrcrL), or −10 region-mutated (Prcr M) promoter configuration, as indicated. Experiments were performed three times in technical triplicates. Statistical significance was determined by unpaired two-tailed t tests (*** P-value < 0.01). (E) The ISCR transposase gene expression is induced by SOS response inducers. β-galactosidase activities were measured in the E. coli MG1656 reference strain carrying a plasmid containing a lacZ transcriptional fusion with Prcr1 (orange), Prcr2 (red), Prcr8 (green), or PsulA (light gray; E. coli SOS-induced promoter, after treatment (+) or not (−) with mitomycin C (MMC), an SOS-inducing agent. All experiments were performed three times in technical triplicates. Statistical significance was determined by unpaired two-tailed t tests (*** P-value < 0.01, * P-value < 0.05). Graphs in panels A and E display a logarithmic y-axis. To confirm the SOS induction, we measured Prcr1, Prcr2, and Prcr8 activity after treatment with mitomycin C (MMC), a DNA-damaging molecule known to induce the SOS response in E. coli (34, 35). As shown in Fig. 3E, treatment with MMC induced a significant and strong increase in β-galactosidase activity for Prcr1 (7-fold) and milder ones for Prcr2 and Prcr8 (2- and 3.2-fold, respectively; Table S1). The intrinsic activity of the three Prcr promoters was then analyzed to validate the σ70 putative promoter sequences identified in silico (Fig. 1B) by introducing mutations into the −10 region of the Prcr promoters (Prcr M ; see Materials and Methods). As anticipated from the very weak activity of the native promoters, Prcr M mutation almost completely abolished the activity of Prcr1 and Prcr8 in the ΔlexA background only, demonstrating the existence of an active promoter tightly repressed by LexA (Fig. 3A through C and Table S1). Surprisingly, for Prcr2 mutated in the −10 region, we obtained a 3.1-fold higher β-galactosidase activity compared to Prcr2, in the wild-type strain. We did not observe such an increase in the ΔlexA strain (Fig. 3B). These results suggest either that the Prcr2 promoter has not been correctly defined by the in silico approach (BPROM analysis [36]) or that the mutations performed to inactivate this promoter indeed disrupted its activity but allowed, at the same time, the recognition of a slightly more active promoter. We performed another in silico analysis confirming that introducing mutations in the initial putative −10 region of Prcr2 allowed the identification of a new Prcr2 promoter upstream of the initial one Fig. S2. This could explain the increased activity for Prcr2 M in the wild-type background. This emphasizes two major facts: (i) introducing mutations to study regulation or expression can lead to a polarized effect; (ii) insertion sequence promoters can be plastic and display a range of regulation yet to be explored, such as recognition by alternative sigma factors, making them difficult to predict. It should be noted that we searched only for σ70 promoter sequences to identify potential transposase rcr gene promoters in silico. The low levels of Prcr2 activity observed under our experimental conditions could also reflect direct or indirect regulation of Prcr2 by alternative sigma factors and/or other specific stress conditions beyond DNA damage, as reported for certain IS. A recent study showed that several ISs in Deinococcus geothermalis are induced and transpose during the exponential phase upon oxidative stress, depending on a noncharacterized alternative sigma factor (37). Geobacillus kaustophilus enhances IS transposition via an extracytoplasmic function sigma factor sigX-dependent stress responses at elevated temperatures when proliferative cells were prevented from active propagation (38). Interestingly, the Tn4652 transposase gene promoter is controlled by the stationary-phase-specific sigma factor σS in Pseudomonas putida (39), and it is silent in E. coli, indicating that its expression requires a specific factor absent in E. coli (40). Implications for ISCR mobility Although it is well known that the expression of genes involved in the mobility in many diverse MGEs, like prophages, plasmids, pathogenic islands, gene cassettes from integrons (reviewed in reference 41), and ICE (34), is controlled by the SOS response through LexA repression, only a few articles in the literature report the contribution of the host SOS system in the regulation of IS transposable elements. The SOS response has been involved in the UV-induced transposition of IS10 (5) and possibly in the transposition of Tn5 (IS50) in E. coli (42), but controversy persists regarding the role of RecA and LexA in regulating Tn5 (43, 44). Here, we demonstrate that the expression of the transposases of ISCR1 and ISCR8 is directly and tightly regulated by the repressor of the SOS response, LexA. Although LexA can specifically bind to the rcr transposase promoter region of ISCR2 in vitro, Prcr2 activity remains extremely low (below 5 Miller units) even when induced with SOS-inducing agents or in the ΔlexA mutant (less than 2-fold induction, Fig. 3B through E). By performing an in silico blastP search of the Rcr2 protein and analyzing the different Rcr2-encoding sequences, we found that 51% of the rcr2 sequences were identified in E. coli, highlighting the relevance of using E. coli for studying the regulation of ISCR2. Furthermore, 84% of all the available rcr2 sequences display a β-/γ-proteobacteria canonical LexA-binding sequence (Supplementary methods and Table S3). This strong conservation of the LexA site in the rcr2 promoter region supports the existence of an SOS-dependent regulatory mechanism that is, however, likely associated with another regulatory mechanism, as discussed above. Intriguingly, ISCR1 also appears to be restricted to β-/γ-proteobacteria (9), i.e., hosts mostly displaying a canonical SOS response. Furthermore, the transposases encoded by ISCR elements are members of the HUH family of endonucleases that use as a substrate an ssDNA known to trigger the SOS response. Thereby, the availability of the ssDNA substrates for ISCR transposase may be sensed as a signal to stimulate the transposase expression through the induction of the SOS response, ultimately promoting their transposition. Although the transposition mechanism of the ISCR elements has not been experimentally proven, the present study demonstrates that under DNA-damaging conditions, the transposase genes of ISCR1, -CR2, and -CR8 elements can be expressed. While their level of expression remains low, even when derepressed, they could contribute to the spread of ISCR and adjacent AR genes. A recent study indeed highlighted that the very low promoter activity of the transposase gene of IS1 (around 15 Miller units) from a single copy on the chromosome can lead to high levels of its cis-transposition. The authors also show that high levels of transposase allow trans-activity on other normally nonmobile IS1 elements and that the trans-transposition frequency is proportional to the levels of transposase expressed (45). In a previous study, we analyzed in silico the diversity of ISCR1 oriS downstream genes and found that 84% of the analyzed sequences (946 out of 1,127) carried a gene downstream of the transposase gene rcr1, encoding for known or putative ARGs conferring resistance to five antibiotic families: trimethoprim (dfrA19, dfrA3, etc.), β-lactams (blaCTX-M-9, blaCMY-9, blaPER-1, etc.), quinolones (qnrA1, qnrA3, and qnrVC6), chloramphenicol (catA2, putative catA, and floR), and aminoglycosides (aphA6 and rmtB) (9). Here, we analyzed in silico 325 sequences of ISCR2 (Table S3). Contrary to the diversity of genes found downstream of ISCR1, several identical combinations of genes are present upstream and downstream of ISCR2, with four combinations being over-represented, containing floR or sul2 resistance genes. i) floR, DUF3363, ISCR2, putative/LysR-like, and recombinase/resolvase; ii) floR, LysR-like, ISCR2, putative, and sul2; iii) sul2, glmM, ISCR2, and putative; and iv) sul2, putative, ISCR2, and putative/LysR-like (Table S3). These two genes confer resistance to florfenicol and sulfonamides, respectively. While the floR gene is primarily found in animals (46), sul2 is found in both clinical settings (47) and in animals (48). For ISCR8, we did not find any ARG downstream of the rcr8 gene, but the small number of full sequences of the rcr8 transposase gene in the database (n = 9, Table S3) does not allow us to draw any conclusions on its ability to mobilize DNA fragments. The rcr2 gene has been frequently associated with the SXT ICE (8), and our in silico analysis also showed that ISCR2 is often found on a plasmid or combined with a recombinase/resolvase gene (Table S3). Given the extremely low level of expression of ISCR1, -CR2, and -CR8 rcr transposases, even when derepressed, these ISCR elements may also rely on the MGE carrying them, such as plasmids or ICE (for ISCR2), to spread. This study demonstrates, in E. coli, a direct regulation of the expression of the transposase gene by the SOS response and its repressor LexA, in three members of the ISCR/IS91 family of bacterial IS elements. As we witness an ever-increasing spread of antibiotic resistance globally, understanding the mechanisms that govern the emergence of ARGs is critical. In this context, our findings reveal the importance of the host SOS response in regulating ISCR expression and offer evidence that, by relying on host-regulatory mechanisms, these elements exhibit a β-/γ-proteobacteria host specificity. Such SOS-dependent repression limits the production of a likely costly transposase, while ensuring that the transposase gene required for mobility is expressed only under specific conditions, and potentially allows the host bacteria to adapt to antibiotic pressure. A further study will be conducted to determine whether ISCR1 mobility increases in response to stress caused by DNA damage. MATERIALS AND METHODS Bioinformatic analysis of ISCR elements Seventeen ISCR members considered in this study were recovered from the initial list published by Toleman et al., completed by the current updated list from the Galileo database, and curated to discard ISCR elements sharing more than 98% nucleotide identity (8, 49). When available, we aligned the sequences upstream of the rcr transposase START codon up to 150 bp. Putative σ70 promoters were predicted by BPROM (36). Alignments were performed using Clustal Omega (https://www.ebi.ac.uk/jdispatcher/msa/clustalo) (50) with default parameters. Sequences and accession numbers of all ISCR used for this alignment are given in Fig. S1. Bacterial strains and plasmids Bacterial strains and plasmids are listed in Table 1. Bacterial strains were grown in lysogeny broth (LB) at 37°C. The medium was supplemented with kanamycin (25 µg/mL) when required. The expression of E. coli His6::LexA protein in E. coli BL21(DE3)/pLysS was induced by adding 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) to the medium. DNA manipulation Plasmid DNA was extracted from E. coli using the Wizard Plus SV Minipreps DNA Purification Kit (Promega, Madison, WI, United States), and genomic DNA was extracted using the SaMag Bacterial DNA Extraction Kit (Sacace, Biotechnologies). PCRs were carried out with Phusion DNA Polymerase (Thermo Fisher Scientific) to amplify fragments subsequently used for cloning or with the Taq’Ozyme Purple Mix 2 (Ozyme) to screen transformants. Oligonucleotides used in this study are listed in Table S2. PCR products were loaded and visualized on 1.6% agarose gel, extracted, and purified with the Wizard SV Gel (Promega, Madison, WI, United States) or PCR Clean-Up kit (Macherey-Nagel). Plasmid constructions Transcriptional lacZ fusions were constructed in the pSU38ΔtotlacZ reporter plasmid (27). The rcr promoter region amplified from bacterial strains known to carry the corresponding ISCR was cloned into the EcoRI/BamHI unique restriction sites of pSU38ΔtotlacZ (see Table 1). Mutations in LexA boxes were performed by PCR assembly to convert the LexA box essential triplet “CAG” into “ACT,” as previously done (22). The Prcr promoters were mutated by PCR assembly in their putative −10 elements: from TACTGT to CGCTGT (ISCR1; Prcr1 M), TAGACT to CGGACT (ISCR2; Prcr2 M), and TACTGT to GCCTGT (ISCR8; Prcr8 M). β-Galactosidase assays β-Galactosidase assays were performed in the E. coli MG1656 strain from at least three independent assays and three technical replicates for each construct, as previously described, at 37°C (27). The SOS response was induced with the addition of mitomycin C (MMC) 1.6 µg/mL in the culture 1 hour before reaching 0.6–0.8 OD600 (22). Electrophoretic mobility shift assay (EMSA) Over-expression and purification of the E. coli His6::LexA protein was performed as described previously (51). EMSA probes, with natural (Prcr) or mutated LexA-binding site (Prcr L), were amplified by PCR from the C1 (ISCR1) or ECI (ISCR2) bacterial strains (Table 1) using oligonucleotide pairs ISCR1-GS-L/ISCR1-GS-R, ISCR1-lexAmutL/ISCR1-lexAmutL, Prcr2L-EcoRI/ISCR2_GS_R, and Prcr2-LexAmutL/Prcr2-LexAmutR for probes Prcr1, Prcr1 L (218 bp), Prcr2, and Prcr1 L (209 bp), respectively (Table S2). EMSA probes were purified on agarose gel 1.6%. For Prcr8, the WT and mutated probes were ordered from Merck as bottom and top strands (C129+CP130 and CP131+CP132, respectively) (Table S2) that were assembled prior to the EMSA experiment. Increasing quantities of recombinant His6::LexA protein (0 to 600 ng) were incubated on ice for 20 min with 40 ng of DNA probes in a final volume of 10 µL of binding buffer (10 mM HEPES pH 8, 10 mM Tris-HCl pH 8, 50 mM KCl, and 1 mM EDTA). Then, 2 µL of 6X EMSA gel-loading solution (EMSA Kit, Molecular Probes BioRad) was added to the reaction mixture. Samples were separated in 6% nondenaturing Tris-glycine polyacrylamide gels and visualized with SYBR Green following the manufacturer’s protocol (EMSA Kit, Molecular Probes BioRad). Statistical analysis Statistical analysis was performed using the t-test with two unpaired groups.