Inactivation of lmo0946 (sif) induces the SOS response and MGEs mobilization and silences the general stress response and virulence program in Listeria monocytogenes

Bacteria have evolved numerous regulatory pathways to survive in changing environments. The SOS response is an inducible DNA damage repair system that plays an indispensable role in bacterial adaptation and pathogenesis. Here we report a discovery of the previously uncharacterized protein Lmo0946 as an SOS response interfering factor (Sif) in the human pathogen Listeria monocytogenes. Functional genetic studies demonstrated that sif is indespensible for normal growth of L. monocytogenes in stress-free as well as multi-stress conditions, and sif contributes to susceptibility to β-lactam antibiotics, biofilm formation and virulence. Absence of Sif promoted the SOS response and elevated expression of mobilome genes accompanied by mobilization of the A118 prophage and ICELm-1 mobile genetic elements (MGEs). These changes were found to be associated with decreased expression of general stress response genes from the σB regulon as well as virulence genes, including the PrfA regulon. Together, this study uncovers an unexpected role of a previously uncharacterized factor, Sif, as an inhibitor of the SOS response in L. monocytogenes. SUMMARY This study uncovers an unexpected role of a previously uncharacterized factor, Sif, as an inhibitor of the SOS response in L. monocytogenes.


10
Listeria monocytogenes is a Gram-positive foodborne human pathogen that is exquisitely well 11 adapted to survive exposure to severe environmental challenges including high salt concentrations, 12 a wide range of temperatures and extreme pH, and high concentration of β-lactam antibiotics used 13 in the treatment of listeriosis. The versatility of L. monocytogenes allows for its adaptation to the The study concerning identification of penicillin G-inducible genes of L. monocytogenes revealed 11 that fri is transcribed together with its downstream genes lmo0944 and lmo0945 (Krawczyk-Balska 12 et al., 2012). Further in silico analysis of the genomic region comprising the fri gene indicated that 13 lmo0946 and lhrC5 are located downstream from, and in orientation consistent with, lmo0945. This 14 led us to hypothesise that these two genes may also be a part of the fri operon ( Figure 1a). To 15 evaluate this hypothesis, RT-PCR analysis was performed that verified the co-transcription of fri, 16 lmo0944, lmo0945, lmo0946 and lhrC5 genes ( Figure 1b). Importantly, in addition to the observed 17 co-transcription event, transcription of lmo0944, lmo0945 and lhrC5 genes is expected to proceed 18 from their own promoters. To get more detailed information about individual transcripts produced 19 within the ferritin operon and their level of co-transcription, northern blot analysis was performed on 20 total RNA extracted from L. monocytogenes EGD-e grown in BHI in exponential phase, with and 21 without penicillin G exposure. The results of the northern blotting, shown in Figure 1c, demonstrate 22 signals corresponding to each gene transcript from the fri operon. According to these signals, fri, 23 lmo0944 and lhrC5 are predominantly transcribed as monocistronic transcripts with size ~ 500 nt, ~ 24 400 nt and ~ 100 nt, respectively, with the lhrC5 transcript clearly visible only in presence of penicillin 25 G. For lmo0945 and lmo0946, the signals in the northern blot analysis indicate that these two genes 26 are co-transcribed as no signal with a size corresponding to a monocistronic transcript of any of 27 5 them is visible. Each of the analyzed genes gave additional bands in the northern blot which may 1 originate from co-transcription with downstream genes and /or processing of the larger co-2 transcripts. Noteworthy is that signals corresponding to a larger co-transcript with size ~2300 nt are 3 visible for each of the analyzed genes, providing proof for co-transcription of fri with all the genes 4 analyzed. The signal intensity of this co-transcript is much lower than signals corresponding to the 5 monocistronic transcripts of fri, lmo0944 and lhrC5, and the bicistronic transcript of lmo0945 and 6 lmo0946, indicating that co-transcription of fri with downstream genes does not proceed very 7 effectively in the analyzed conditions of growth. We were wondering if co-transcription level depends 8 on the conditions of growth. To answer this, further northern blot analysis with probes specific for fri 9 and lmo0946 was performed on total RNA extracted from L. monocytogenes EGD-e in stationary 10 phase of growth. According to the signals in the northern blot analysis, shown in Figure 1d, fri is still 11 predominantly transcribed as the monocistronic transcript with size ~ 500 nt, however the co- 12 transcript with size ~2300 nt is clearly visible in stationary phase. In case of lmo0946, the intensity 13 of the signal corresponding to this co-transcript is much higher than the signal corresponding to the 14 bicistronic transcript of lmo0945-lmo0946. These observations suggest that co-transcription of fri 15 with downstream genes occurs more readily in stationary phase cells. Generally, northern blotting 16 confirmed co-transcription of fri with lmo0944, lmo0945, lmo0946 and lhrC5 genes. Thus, these 17 genes constitute an operon, named afterwards, the ferritin operon. 18 19 Inactivation of lmo0946 impairs the Growth under Different Conditions 20 To investigate the relevance of the ferritin operon in the physiology of L. monocytogenes, each gene 21 of the operon was subjected to inactivation. In case of lmo0944 and lmo0945, in frame deletion 22 mutants were constructed. Due to the presence of genes coding for antisense RNAs Anti0943, 23 Anti0945 and Anti0946 within the operon (Figure 1a), single point mutations were designed for 24 inactivation of fri, lmo0946 and lhrC5 to minimize the risk of undesirable changes in the expression 25 of these asRNAs. In case of fri and lmo0946 the nucleotide substitutions were introduced to create 26 nonsense mutations (named fri* and lmo0946*), and in case of lhrC5, nucleotide substitutions in the 27 6 -10 promoter region were introduced to prevent initiation of transcription (named lhrC5*). Moreover, 1 due to high sequence similarity between LhrC5 and six other sRNAs from the LhrC family, the 2 functional analysis LhrC5 may be difficult or impossible when the other LhrCs remain functional. To 3 resolve this problem, the genes encoding for these six LhrCs were deleted in the background of the 4 wild-type and lhrC5 mutant strains. First, the growth rates of the mutants and the parent strain in 5 BHI broth at 37°C were compared. Interestingly, in these stress-free conditions the inactivation of 6 all the studied genes but lmo0946 had no effect on rate of growth ( Figure 2). 7 The observed growth impairment of the lmo0946* mutant strain prompted us to focus on assessing 8 the effects of inactivation of this gene. Prior to detailed analysis, complementation of lmo0946 was 9 performed in the background of the mutant strain. The wild-type, lmo0946* mutant and 10 complementated strain (lmo0946*-lmo0946) were compared with respect to their abilities to grow in 11 stress-free conditions. The growth of the lmo0946* mutant was clearly impaired relative to the wild  Table S1). 17 18 Inactivation of lmo0946 reduces the ability of L. monocytogenes to form biofilm 19 In order to study the effect of lmo0946 inactivation on biofilm formed by L. monocytogenes, the 20 sessile biomass of the studied strains was stained by crystal violet. The biofilm formation assay 21 revealed significant differences between the lmo0946* mutant, and wild-type or complemented  Lmo0946 contributes to the resistance of L. monocytogenes to cephalosporins and tolerance 1 to penicillin G 2 To investigate whether inactivation of lmo0946 affects the susceptibility of L. monocytogenes to 3 antibiotics, the parent and mutant strains were subjected to antibiotic disk assays. This assay 4 revealed that the wild-type and lmo0946* strains displayed similar levels of resistance to various 5 antibiotics (ampicillin, penicillin G, vancomycin, aztreonam, meropenem, tetracycline, rifampicin, 6 gentamicin, trimethoprim, and ciprofloxacin), however significantly greater zones of growth inhibition 7 were observed for the mutant with cefuroxime and cefoxitin. The MICs of these specific 8 cephalosporin antibiotics were then determined for L. monocytogenes EGD-e and the 9 lmo0946* mutant. In confirmation of the antibiotic disk assay result, the MIC of cefuroxime for wild-10 type and lmo0946* was 8 μg/ml and 4 μg/ml, respectively, whereas the MIC of cefoxitin for wild-type 11 and lmo0946* was 32 μg/ml and 16 μg/ml, respectively. Thus, inactivation of the lmo0946 gene 12 caused a 2-fold increase in the sensitivity of L. monocytogenes to these cephalosporins. Next, 13 tolerance was examined by testing the ability of the strains to survive in concentrations of penicillin 14 G over 100 fold higher than the MIC value which is 0.12 μg/ml. The tolerance assay revealed that 15 the survival of lmo0946* was significantly impaired since reduced numbers of viable cells were 16 recovered for the mutant relative to the wild-type strain after exposure to high concentration of 17 penicillin G ( Figure 5). In summary, our results indicate that Lmo0946 contributes to the resistance 18 to cephalosporins and tolerance to penicillin G of L. monocytogenes. 19 20 Inactivation of lmo0946 impairs the virulence of L. monocytogenes. To determine whether 21 Lmo0946 is important for the virulence of L. monocytogenes, in vitro infection studies were 22 performed using the lmo0946* mutant and wild-type L. monocytogenes EGD-e. The infection assay 23 showed that the lmo0946* mutant was significantly impaired in the ability to proliferate intracellularly 24 in the murine macrophage cell line P388D1 relative to the wild-type strain (Figure 6a). Subsequently, 25 the mouse infection model was used to assess the virulence properties of the lmo0946* mutant 26 relative to that of wild-type. For the mutant, survival and/or growth in the liver and spleen was 27 8 significantly reduced at day 3 post-infection when compared with wild-type ( Figure 6b). These 1 results indicate that Lmo0946 contributes to the virulence of L. monocytogenes. In silico analysis of Lmo0946 indicates that it is a small (75 aa) conserved bacterial protein of 5 unknown function, and the function of its homologs, found mainly in Bacilli and Clostridia classes of 6 the Firmicutes phylum (Supplementary Figure S1), has not been established yet. To further 7 elucidate why inactivation of lmo0946 results in severe phenotypic changes of L. monocytogenes, 8 a transcriptomic analysis of L. monocytogenes lmo0946* mutant and the parental strain EGD-e was 9 performed using RNA-seq during exponential phase of growth at 37°C in BHI. The RNA-seq analysis 10 revealed that inactivation of lmo0946 caused 589 genes to be differentially expressed when 11 compared to the wild-type strain using an adjusted p-value < 0.01, corresponding to approximately 12 20 % of protein encoding genes and 10 % of the sRNAs (Figure 7a and Supplementary Table S2). 13 The relative abundance of each COG category in the set of differentially transcribed genes was 14 further determined. A set of 320 genes showed higher transcript levels while 269 showed lower 15 transcript levels in the lmo0946* mutant relative to the wild-type strain (Figure 7b and Supplementary   16   Table S3). COGs analysis revealed that genes involved in carbohydrate transport and metabolism, 17 energy production and conversion, and lipid transport and metabolism were highly represented in 18 the downregulated set of genes, indicating that inactivation of lmo0946 results in a general 19 slowdown of metabolism. The analysis also showed that genes related to nucleotide transport and 20 metabolism, replication, recombination and repair as well as the mobilome were highly represented 21 among the upregulated genes, suggesting that inactivation of lmo0946 leads to disorders of 22 processes related to DNA metabolism and replication.

23
To better understand the causes of global changes in gene expression resulting from lmo0946 24 inactivation we further analyzed the genes that were at least fourfold up-or downregulated in the 25 background of lmo0946* (corresponding to 83 and 45 genes, respectively) (Supplementary Table   26 S4). The most up-regulated genes with more than 20-fold change were lmo1097 (encoding an 27 9 integrase), hfq (encoding an RNA chaperone), genes involved in cadmium resistance, and few 1 genes coding for hypothetical proteins (Supplementary Table S4 and Figure 7a). Going through the 2 most numerous parts of the up-regulated genes, it was noticeable that most of them are phage-3 related genes. These genes accounted for over 50% of the upregulated genes and most of them, 4 though not all, were genes of the A118 prophage which is one of the four mobile genetic elements  Given the high frequency of prophage genes in the set of up-regulated genes, the expression of all 9 MGE genes in the lmo0946* strain was further analyzed. This analysis showed that genes of three 10 of the four MGEs, i.e., prophage A118, monocin locus and ICELm-1, are highly expressed in the 11 lmo0946* mutant (Supplementary Table S5). 12 The monocin locus, a cryptic prophage region, is conserved in all L. monocytogenes lineages and ). The expression of all genes from the monocin locus is up-regulated in the lmo0946* mutant, 15 and for most of them more than 10-fold up-regulation is observed. ICELm1 is an MGE specific for 16 L. monocytogenes EGD-e. It consists of 19 genes, including an integrase-encoding gene (lmo1097), 17 cadmium resistance genes lmo1100 (cadA), lmo1101 (lspB) and lmo1102 (cadC), a gene coding 18 for a fibrinogen-binding protein with an LPXTG domain (lmo1115) and Tn916-like genes of unknown  Table S5). Interestingly, genes belonging to ICELm1 showed over 22 30-fold up-regulation of expression in the lmo0946* mutant strain, and this was the highest increase 23 in expression recorded in these studies. An analysis of the A118 prophage showed that the 24 expression of 50 of 62 prophage genes is up-regulated in the lmo0946* strain with more than 4-fold 25 up-regulation observed for 45 of these genes (Supplementary Table S5). Interestingly, the 26 upregulated genes belong to virtually all functional modules of prophage A118 including genes 27 encoding phage lysis proteins holin and lysin. In summary, these data indicate that inactivation of 1 lmo0946 led to upregulation of a large portion of the mobilome of L. monocytogenes. 2 The most highly down-regulated gene in the lmo0946* mutant was lmo1634. It encodes the Listeria 3 adhesion protein (LAP), a putative bifunctional acetaldehyde-CoA/alcohol dehydrogenase involved 4 in pyruvate metabolism. Three pyruvate-formate lyase encoding genes were found to be 5 downregulated as well (pflB, pflA and pflC). Additionally, genes involved in arginine biosynthesis 6 (argC, argG, argH), ferrous iron transport (feoA and feoB), and amino acid transport and metabolism 7 (arpJ and gadD) were found to be highly down-regulated in the lmo0946* mutant relative to the wild-8 type strain (Supplementary Table S4 and Figure 7a). Collectively, these findings suggest that 9 inactivation of lmo0946 led to a decrease in energy production and conversion as well as 10 descreased transport and metabolism of amino-acids and other cell compounds. Interestingly, high 11 level of downregulation was also noticed for lmo2686 encoding the RNA binding protein Zea which 12 was shown to specifically interact with phage RNAs of L. monocytogenes (Pagliuso et al., 2019). 13 This observation further supports a putative link between lmo0946 and the mobilome of L. 14 monocytogenes. 15 Inactivation of lmo0946 Triggers a Variety of Stress Responses in L. monocytogenes, 16 Including the Sos Response 17 Out of 128 genes with at least a 4-fold change in expression level, 68 were found to be allocated 18 into known regulons, and some of them are under control of more than one regulator 19 (Supplementary Table S4). For these genes the largest overlap was observed with the CodY 20 regulon. The transcriptional regulator CodY responds to different nutritional and environmental 21 stresses and generally controls adaptive responses in conditions that limit bacterial growth (Bennett 22 et al., 2007). We noticed that 30 of the highly upregulated genes, coding for prophage A118 proteins, 23 were under positive control of CodY. In addition, seven of the most downregulated genes were under 24 negative control of CodY. Among them were arg genes coding for proteins engaged in the 25 biosynthesis of arginine, and genes involved in the transport and metabolism of amino acids, 26 carbohydrates, and inorganic ion. 27 In addition to the CodY regulon, five genes belonging to LexA/RecA regulon were upregulated at 1 least fourfold in lmo0946* mutant strain (Supplementary Table S4). The LexA/RecA regulon covers  Table S6). Notably, the expression levels of genes from the SOS response noticed 8 for the lmo0946* mutant strain is comparable to the levels expressed in wild-type L. monocytogenes  Table S6). These findings clearly demonstrate that inactivation of lmo0946 triggers 13 the expression of SOS response genes known to promote the DNA repair in DNA-damaged cells.
14 Finally, four genes belonging to the VirR regulon were upregulated at least fourfold in the lmo0946*  Table S4). 18 Inactivation of lmo0946 Leads to Downregulation of Virulence Genes and Modulation of 19 Table S7). Furthermore, decreased level of expression was 1 observed for genes preceded by a PrfA box and belonging to the core set of the PrfA regulon, like 2 hly, inlA and inlB (Milohanic et al., 2003). Additionally, the most highly downregulated gene,  Table S4). The alternative sigma  Table S4). The only gene positively regulated by σB, and highly 15 upregulated in the lmo0946* strain, is hfq coding for an RNA chaperone involved in post- 16 transcriptional control of gene expression in L. monocytogenes. Notably, transcription of hfq was 17 shown to proceed from a σB-dependent promoter (Christiansen et al., 2004). These data suggest 18 that modulation of expression of the σB regulon occurs in the lmo0946* strain. This assumption is 19 supported by a 2-fold decrease in expression of the rsbV gene coding for the anti-anti-σB factor 20 (Supplementary Table S2). As rsbV plays an important role in a signaling pathway controlling σB  Table S8). This observation suggests that the σB 27 13 signaling cascade is affected by the lmo0946 mutation. Among the highly downregulated genes 1 were also three genes downregulated by Fur, which is a Fe 2+ dependent repressor of genes involved 2 in iron/heme uptake and utilization. These genes encode a ferrous iron transport system (feoA and 3 feoB) and an element of a transport system involved in cytochrome biosynthesis (cydC) (Lechowicz 4 & Krawczyk-Balska, 2015). 5 It should be noted that as many as 16 out of 25 highly downregulated genes belonging to known 6 regulons are regulated in lmo0946* in a manner opposite to regulation by the appropriate regulators 7 (Supplementary Table S4). Besides the genes from the PrfA and σB regulons described above,   Subsequently, to verify whether the transcriptomic changes observed in the lmo0946* mutant strain 19 are the result of inactivation of lmo0946, the level of expression of selected genes in wild-type L. 20 monocytogenes EGD-e, the lmo0946* mutant and the complemented strain was examined by RT-21 qPCR. The analysis of genes from the SOS regulon and mobile genetic elements revealed 22 increased level of expression in the lmo0946* mutant (Figure 8a). Upon complementation, the 23 expression of these genes was restored to that observed in the wild-type strain, except for the 24 lmo1097 gene belonging to ICELm1 MGE ( Figure 8a). Next, the expression of virulence genes was 25 examined. Increased expression of prfA and decreased expression of hly was observed in the 26 lmo0946* mutant, and upon complementation the expression of these genes was restored to that 27 14 observed in the wild-type strain ( Figure 8b). Finally, the expression of genes with potent regulatory 1 roles was studied. Increased expression of hfq and decreased expression of zea was observed in 2 the lmo0946* mutant, but upon complementation the expression of these genes was not restored to 3 that observed in the wild-type strain. In case of rsbV and sigB, expression of these genes is 4 decreased in the lmo0946* mutant strain, however only in case of rsbV the difference in expresion 5 between wild-type and mutant strain was significantly important (P=0.04), while for sigB the 6 difference was just above the cut-off of statistical significance (P=0.08). Furthermore, the expression 7 of rsbV was not restored upon compementation to the level observed in wild-type EGD-e strain 8 ( Figure 8c). In conclusion, the results of the RT-qPCR analysis confirmed the changes in gene 9 expression in the mutant strain observed in the RNA-seq analysis. Furthermore, gene expression 10 was restored in the complemented strain for genes involved in the SOS response and virulence as 11 well as two MGEs i.e., prophage A118 and monocin locus. However, the expression of genes 12 involved in RNA binding and the σB response as well as ICELm1 MGE was not restored upon 13 complementation.

15
Inactivation of lmo0946 Results in excision of prophage A118 without inducing lysogeny 16 The prophage A118 is integrated within the comK gene of L. monocytogenes EGD-e and while it 17 was not extensively studied, it exhibits high similarity to Φ10403S prophage of L. monocytogenes that in the lmo0946* mutant strain precise excision of prophage A118 is induced and it is reversed 7 upon complementation. 8 We further wondered whether the observed phage excision in lmo0946* mutant strain is followed 9 by induction of the lytic cycle. To answer this question, infective virion production was monitored in 10 the wild-type, lmo0946* mutant and complemented strains by using a plaque forming assay. The 11 infective phages production was not detected in any of the strains tested (data not shown). 12 Therefore, prophage A118 does not go into the lytic cycle, even though excision of the prophage 13 occurs in the lmo0946* mutant strain. Likewise, PCR analysis was performed to detect possible 14 excision of cryptic prophage monocin locus as a high level of expression of this MGE in lmo0946* 15 mutant was also observed. However, no PCR products were obtained in this analysis indicating that 16 excision of monocin locus in lmo0946* mutant does not occur (data not shown). 17 Inactivation of lmo0946 Results in replication of ICELm1 in circular form and increased 18 resistance to cadmium 19 The integrative and conjugative element ICELm1 is specific for the EGD-e strain and accordingly, it  (Figure 10a). This analysis revealed the presence of a PCR product 1 corresponding to the fragment containing the ICELm1 integration site in the lmo0946* mutant and 2 complemented strains, however this product was not detected in the wild-type strain. Surprisingly, 3 the fragment covering distal guaA and proximal lmo1116 genes was not amplified in any of the 4 tested strains suggesting the presence of ICELm1 in the genome of the studied strains. This ICELm1 is induced in the lmo0946* mutant strain, and it is not reversed upon complementation. 13 As the genes involved in cadmium resistance are encoded within the ICELm1 (Kuenne et al., 2013) 14 we further wondered whether the observed replication of the circular form of ICELm1 in the lmo0946* 15 mutant and complemented strain led to an increase in cadmium resistance of these strains. To 16 answer this question, the susceptibility of the wild-type, lmo0946* mutant and complemented strains 17 to cadmium was tested using an agar and broth dilution assay. This assay revealed that the MIC of 18 cadmium for EGD-e was 200 μg/ml while the lmo0946* and complemented strains were extremely 19 resistant to cadmium. The precise determination of the MIC values for these strains was impossible 20 due to their ability to grow in the presence of cadmium to the limit of its solubility, which was 1000 21 μg/ml. These observations indicate that replication of the circular form of ICELm1 in the lmo0946* 22 mutant and complemented strain is accompanied by high increase of cadmium resistance.

23
Lmo0946 protein does not bind to DNA of promoters of selected genes 24 Frequently, pronunced changes in bacterial transcriptomics and physiology are observed upon 25 inactivation of regulatory genes. Considering that inactivation of Lmo0946 resulted in major changes 26 in gene expression, we wondered whether Lmo0946 could be a DNA-binding transcriptional 27 regulator. To investigate this idea, we performed electrophoretic mobility shift assays (EMSAs) using 1 native polyacrylamide gels to detect binding of the purified Lmo0946-His6 protein to promoter showed that the fri gene is co-transcribed together with its downstream genes lmo0944 and lmo0945 12 (Krawczyk-Balska et al., 2012). As genes co-transcribed with fri could play an important role in stress 13 adaptation and/or virulence, this observation prompted us to extend the co-transcription analysis 14 and functional analyses of genes transcribed together with fri. The RT-qPCR revealed that fri is co- 15 transcribed with four downstream genes, namely lmo0944, lmo0945, lmo0946 and lhrC5, the last of 16 which encodes for an sRNA from the LhrC family known for its role in the stress response of L. 17 monocytogenes (Sievers et al., 2014(Sievers et al., , 2015. As the result from the co-transcription analysis 18 contradicts previous studies showing that Fri is expressed from a monocistronic mRNA both in 19 stress-free conditions and under heat-or cold-shock (Hébraud & Guzzo, 2000), we further analyzed 20 the transcripts of the fri operon by northern blot. The analysis showed that the larger co-transcript, 21 corresponding to the mRNA of fri and downstream genes, is clearly visible in the stationary phase 22 of growth, however it is barely detectable during growth under stress-free conditions and under 23 penicilin G pressure. These results indicate that the transcriptional terminator of fri is regulated in a 24 condition-specific manner that results in generation of full-length mRNA at higher levels in the 25 stationary phase than in other conditions. Therefore, the co-transcription of fri with downstream 26 genes could be easily missed in earlier studies. Subsequently, all the genes from the ferritin operon 27 were subjected to functional analysis. The inactivation of fri, lmo0944, lmo0946 and lhrC5 had no 1 effect on the growth of L. monocytogenes in stress-free conditions which in the case of fri and lhrC5 factors, it must be related to the changes in cellular processes that occur because of the mutation. 5 The induction of the SOS response observed in the lmo0946* mutant is not likely caused by induced 6 production of ROS since the transcriptome analysis did not reveal an induction of kat (encoding 7 catalase) and sod (encoding superoxide dismutase) known to work together in detoxification of ROS 8 (Dallmier & Martin, 1988). Furthermore, these genes were recently reported to be highly co-induced 9 with the SOS response in L. monocytogenes exposed to heme stress causing oxidative damage  Table S4). Worth of note is that different stimuli can indirectly generate the SOS-14 inducing signal by activation of endogenous DNA damage mechanisms (Aertsen & Michiels, 2006). 15 Given this, we speculated that Lmo0946 may be a transcriptional regulator of genes with important 16 roles in the physiology or stress response of L. monocytogenes and its regulatory activity in turn 17 could lead to activation of the SOS response. However, the results of EMSA analyses led us to 18 reject this hypothesis. Therefore, the specific mechanism underlying the link between Lmo0946 19 inactiation and activation of the SOS response remains unclear at the current stage of research. 20 Although the present study does not explain how the SOS system is activated in the lmo0946*  bacterial viruses (Zink & Loessner, 1992). This bacteriophage was shown to reproduce by both lytic 1 and lysogenic cycles, in the latter the phage's genome is integrated at a specific attachment site 2 located within the comK gene (Loessner et al., 2000). Surprisingly, we observed that phage A118 3 mobilization is not accompanied by switching from lysogeny to the lytic pathway in a sif mutant strain 4 of L. monocytogenes EGD-e. This result is consistent with previous reports on hardly induction of 5 the lytic cycle of prophage A118 in the EGD-e strain exposed to UV irradiation or mitomycin C; well 6 known factors inducing the SOS response (Pasechnek et al., 2020). Notably, a similar prophage to 7 A118, named Φ10403S, is integrated like A118 in the comK gene of another widely used L.  In the light of these data, we assume that the interplay between comK-phages and their L.  Our study also revealed that L. monocytogenes responds to lmo0946 inactivation by upregulation 13 of the monocin genetic locus. As the monocin element in L. monocytogenes is reported to be 14 activated under UV irradiation (Argov et al., 2017), we assume that high level of expression of this 15 MGE in the sif imutant strain, similarly to phage A118, results from activation of the SOS response. 16 It is worth noting that the monocin locus resembles a phage tail module of A118-like phages but it 17 does not contain the DNA region responsible for DNA replication and recombination (Lee et al.,  stage of research, the reason for the lack of reversion of the expression level of these genes and 10 the mobilization of ICELM1 upon complementation remains unclear. However, we suppose that the 11 lack of complementation may be due to a lower level of Sif protein in the complemented strain as 12 compared to the wild-type strain. In the complemented strain, the sif gene is expressed in trans from 13 the native promoter of lmo0945. As a result, Sif is not translated from the larger co-transcripts 14 corresponding to sif mRNA arising in L. monocytogenes EGD-e, which in turn leads to an overall 15 lower level of Sif protein in the complemented strain than in the wild-type strain. 16 Activation of the SOS response and the mobilization of MGEs, is accompanied by a decrease in 17 energy production and conversion as well as transport and metabolism of different cell compounds 18 in the sif mutant strain. These changes indicate general metabolic slowdown in response to sif 19 inactivation. Our analysis also revealed that L. monocytogenes responds to sif inactivation by   Table S2). This suggests that 1 putative modulation of CodY activity in the sif mutant strain does not result from BCAAs insufficiency 2 and could be rather GTP-dependent. This assumption is in line with the observed general 3 downregulation of genes involved in energy production and conversion in the sif mutant strain. 4 However, it should be noted that only some of the genes overlapping with the CodY regulon are 5 involved in transport and metabolism, while most of them correspond to A118 prophage genes. 6 Therefore, the involvement of CodY in the downregulation of energy production and conversion in 7 the sif mutant strain is uncertain. Similarly, the engagement of CodY in the regulation of A118  Among the highly down-regulated genes in the sif mutant strain, we found genes positively regulated 12 by the alternative sigma factor σB. The σB factor controls the general stress response in L. which is an important factor in the signaling pathway controlling σB activity. In stress-free conditions, 20 σB is sequestered by the anti-sigma factor RsbW that prevents interaction of σB with core RNA 21 polymerase, and therefore the σB regulon is not transcribed (Ferreira et al., 2001). When stress 22 signals are sensed, the anti-anti-sigma factor RsbV undergoes dephosphorylation and interacts 23 directly with RsbW. The RsbV-RsbW interaction results in release of σB from RsbW-σB complexes. 24 Subsequently, σB binds to the core RNA polymerase and directs transcription of the σB regulon 25 (Hecker et al., 2007). Therefore, RsbV is required to activate σB in response to stress signals in L.

26
monocytogenes (Chaturongakul & Boor, 2004. It seems likely that the reduced level of rsbV 27 24 expression in the sif mutant strain may result in decreased liberation of σB from the RsbW-σB 1 complexes, and thus in inhibition of transcription of the σB regulon despite of stress signals sensed. 2 While the impairment of the σB signaling pathway does not explain the downregulation of numerous 3 genes which are under positive control of σB, it clearly indicates that the σB protective response is 4 prevented in sif mutant strain. Interestingly, while rsbV is the first gene of the four-gene operon which 5 includes besides rsbV also rsbW, sigB and rsbX, we observed that expression of rsbV but not sigB 6 is downregulated. This result suggests that the expression of this operon is regulated post-7 transcriptionally in the sif mutant strain. 8 Finally, we found that expression of a large portion of genes regulated by the master virulence 9 regulator PrfA is downregulated in the sif mutant strain despite increased expression of prfA itself.

10
It should be noted that σB-dependent promoters preceede multiple PrfA-dependent genes which 11 are downregulated in the sif mutant strain (Supplementary Table S7, Milohanic et al., 2003). This 12 suggests that downregulation of PrfA-dependent genes is related to the downregulation of σB-

25
To summarize, we found that fri is transcribed together with four downstream genes, namely 26 lmo0944, lmo0945, lmo0946 and lhrC5. Functional analysis of genes from the ferritin operon 27 25 revealed that inactivation of lmo0946, coding for a protein with unknown function, results in severe  The wild-type strain L. monocytogenes EGD-e, its derivative mutant strains in ferritin operon genes 18 (described in detail below) and strain with complemented mutation of lmo0946 were used in this 19 study. L. monocytogenes was routinely grown in brain heart infusion broth (BHI, Oxoid) at 37 •C with 20 aeration. When appropriate, cultures were supplemented with chloramphenicol (7.5 μg/ml), 21 erythromycin (5 µg/ml) or X-gal (50 mg/ml). 22 For growth experiments, overnight cultures grown at 37°C with aeration in BHI medium were diluted 23 1:1000 in fresh BHI broth in titration plates. Depending on the experiment, BHI medium was 24 suplemented with 5 % ethanol, 0.09 µg/ml penicillin G, acidified with HCl to pH 5 or alkalized with 25 NaOH to pH 9. The cultures were grown at 37°C with aeration and bacterial growth was monitored 26 by measuring the optical density at 600 nm (OD600). For thermal stress, the cultures were grown at 27 26 37°C to an OD600 of 0.03 and at this point the incubation temperature was shifted to 43°C. 1 Escherichia coli strain Top10 and BL21(DE3) (Invitrogen) were used in cloning experiments and 2 purification of Lmo0946-His6, respectively. E. coli strains were grown on Luria-Bertani medium. 3 When required, the LB medium was supplemented with ampicillin (100 μg/ml), chloramphenicol (25 4 μg/ml) or kanamycin (30 µg/ml). 5 Construction of ferritin operon mutant strains and complementation of lmo0946 mutation. 6 For the construction of in-frame mutants with deletions of lmo0944, lmo0945, lhrC1-4, lhrC6 (rli22) 7 and lhrC7 (rli33.1) L. monocytogenes EGD-e chromosomal DNA was used as the template for the 8 PCR amplification of DNA fragments representing either the 5' end and upstream sequences or the   In case of lhrC5, the introduced nucleotide substitutions changed the -10 promoter consensus 21 sequence TATTAT to CAGTGC. Primers used for constructing the in-frame deletions as well as site- 22 directed mutagenesis are listed in Supplementary Table S9. The obtained pMAD derivatives were 23 used for genes replacement which was performed via double-crossover homologous recombination 24 as described previously (Arnaud et al., 2004). Erythromycin-sensitive clones were screened for the 25 presence of the mutations by PCR. Accuracy of the desired DNA modifications in the obtained 26 mutant strains was confirmed by the sequencing of obtained PCR products. 27 27 For the construction of the strain with complemented lmo0946 mutation L. monocytogenes∆lmo0945 1 chromosomal DNA was used as the template for the PCR amplification of DNA fragment covering 2 ORFs of lmo0945 and lmo0946. The obtained PCR product was cloned into the pPL2 vector and 3 subsequently introduced into lmo0946* mutant strain as described previously (Lauer et al., 2002). 4 Biofilm Formation analysis 5 Biofilm formation was carried out essentially as described previously (Djordjevic et al., 2002). Briefly, 6 12-well polystyrene plates containing 1 ml of BHI broth were inoculated with the overnight culture of 7 each strain to OD600 = 0.01. The plates were then incubated at 37°C under static condition. At water. After fixation with methanol, the wells were air dried and stained with crystal violet solution 10 for 5 min. After washing, wells were air dried and the bound dye was solubilized with an 11 solution of acetic acid. Following 10-fold dilution in destiled wated, the absorbance of the solubilized 12 dye was measured at 570 nm. 13

14
The susceptibility to antibiotics and cadmium as well as the tolerance to penicillin G were examined  The tolerance to penicillin G was examined by innoculation of BHI broth supplemented with 32 μg/ml 24 penicillin G with 5 × 10 7 CFU/ml of mid-exponential phase cultures. The cultures were subsequently 25 incubated with aeration at 37°C and at indicated points of time viable cell counts were performed by 26 standard plate counting. 27 28 In vitro infection experiment 1 Infection experiments with L. monocytogenes strains were performed using P388.D1 murine 2 macrophage (ECACC Collection) essentially as described previously (Chatterjee et al., 2006). 3 Briefly, macrophages were grown in 12-well plates until reached at least 80% confluence. Bacteria  For northern blot analysis in stationary phase, L. monocytogenes EGD-e was innoculated from 23 single colony into BHI broth and cultured for 18h. For northern blot analysis in exponential phase, 24 L. monocytogenes EGD-e was grown to OD600 ~ 0.35, split, and half stressed with 0.09 µg/ml 25 penicyllin G for 60 min. 10 ml samples from stationary phase cultures and 20 ml from exponential 26 phase cultures were immedietely cooled down in liquid nitrogen and centrifuged at 12 000 x g for 3 27 min at 4•C. The cell pellets were frozen in liquid nitrogen and stored at -80 °C until further 1 processing. The cells were disrupted by the FastPrep instrument and total RNA was extracted using 2 1 ml of TRItidy G™ Reagent (Applichem) as described previously (Nielsen et al., 2010). The 3 concentration and purity of RNA were determined with a NanoDrop ND-1000 spectrophotometer 4 and integrity of RNA was confirmed by agarose gel electrophoresis. 5 For transcriptomic and RT-qPCR analysis, L. monocytogenes cultures were grown to OD600 ~ 0.35. NanoDrop ND-1000 spectrophotometer. The absence of DNA from RNA preparations was verified 16 by the failure to amplify a rpoB gene fragment in a 30-cycle PCR using 100 ng of each RNA isolation 17 as the template. A 2100 Bioanalyzer (Agilent Technologies) was used to measure the RNA quality. 18 The prepared RNA was stored at −80°C before further analysis.  For co-transcription analysis, 100 ng of total RNA was used for synthesis of first strand cDNA. Reverse transcription was performed with RevertAid First Strand cDNA Synthesis Kit (Thermo 2 Fisher Scientific) and primer specific for the lhrC5 gene. The obtained cDNA was then used as the 3 template for PCR performed with primers specific for internal fragments of the fri and lmo0944 4 genes. Primers used for the co-transcription analysis are listed in Supplementary Table S9. 5 For northern blotting, 20 µg of total RNA in loading buffer containing 50% formamide and 20% 6 formaldehyde was separated on a formaldehyde agarose gel and subsequently transferred to a Zeta  Table S9). 21 Purification of Sif protein and Electrophoretic mobility shift assay (EMSA) 22 The His-tag system (Novagen) was used for Sif purification. The fragment representing the entire 23 lmo0946 coding sequence was PCR-amplified on the template of L. monocytogenes EGD-e 24 chromosomal DNA using primers listed in Supplementary Table S9. PCR product was cloned into 25 vector pET28a (Novagen) and the obtained construct was used to transform E. coli BL21(DE3). The                      Northern blot analysis of fri, lmo0944, lmo0945, lmo0946, and lhrC5 in exponential-phase of L. 15 monocytogenes EGD-e growth. Samples were taken from EGD-e wild type mid-exponential cultures 16 exposed to 1 hour of penicillin G stress (+) as well as from non-stressed cultures (-). Northern blots 17 were probed for fri mRNA, lmo0944 mRNA, lmo0945 mRNA, lmo0946 mRNA, and LhrC5. Numbers 18 on the right side of each panel indicate the estimated lengths of the transcripts in nucleotides. (d) 19 Northern blot analysis of fri and lmo0946 in stationary-phase of L. monocytogenes EGD-e growth. 20 Samples were taken from EGD-e wild type overnight culture. Northern blots were probed for fri 21 mRNA and lmo0946 mRNA. Numbers on the right side of each panel indicate the estimated lengths 22 of the transcripts in nucleotides.     Figure 5. Inactivation of lmo0946 impairs tolerance of L. monocytogenes to penicillin G. BHI 3 broth supplemented with 32 μg/ml penicillin G was inoculated with a mid-exponential culture of wild-4 type L. monocytogenes EGD-e (wt) and lmo0946* mutant strains and incubated with shaking at 5 37°C. Viable cell counts were measured on agar plates following serial dilutions of culture samples. 6 The mean values from three independent experiments are plotted and the error bars indicate 7 standard deviations. The asterisks indicate a significant differences (*P < 0.05, **P < 0.01).    Figure 8. Impact of lmo0946 inactivation and complementation on expression of selected 2 SOS regulon, MGE, virulence and regulatory genes. The expression was determined in 3 exponential growth phase in BHI for wild-type EGD-e, lmo0946* mutant strain and complemented 4 strain lmo0946*-lmo0946 by RT-qPCR. Relative transcripts levels of (a) recA, umuC, lmo2279, 5 lmo2308, lmaA and lmo1097, (b) prfA and hly, and (c) hfq, zea, rsbV and sigB were normalized to 6 the amount of rpoB gene. The results shown are the average of three biological replicates, error 7 bars indicate standard deviations. The asterisks indicate a significant differences (*P < 0.05, **P < 8 0.01, ***P < 0.001, ****P < 0.0001).