The bacterial strains and plasmids used in this work are listed in Supplementary Table S1. Salmonella Typhi IMSS-1 (Puente et al., 1987) and Escherichia coli strains were grown aerobically at 37°C in LB (10 g tryptone, 5 g yeast extract, and 10 g NaCl per liter), MA (7 g nutrient broth, 1 g yeast extract, 2 ml glycerol, 3.75 g K₂HPO₄, and 1.3 g KH₂PO₄ per liter; Kawaji et al., 1979) or N-MM media [0.37 g KCl, 0.99 g (NH₄)₂SO₄, 0.087 g K₂SO₄, 0.14 g KH₂PO₄, 0.019 g MgCl₂, 1 g casamino acids, 5 ml glycerol, and 100 mM of Tris-HCl (pH 7.5) per liter] (Deiwick et al., 1999). When required, the following antibiotics were added: kanamycin (Km), 30 μg ml⁻¹; tetracycline (Tc), 12 μg ml⁻¹, and ampicillin (Ap), 200 μg ml⁻¹.

Plasmid and genomic DNA isolations were carried out according to published protocols (Sambrook et al., 1989). Primers for PCR amplifications were provided by the Oligonucleotide Synthesis Facility at our Institute (Supplementary Table S2). Restriction enzymes, ligase, nucleotides, and polymerases were acquired from New England Biolabs, Invitrogen, or Thermo Scientific. For sequencing, double-stranded DNA was purified with the High Pure Plasmid Isolation Kit (Roche) and sequenced with an automatic Perkin Elmer/Applied Biosystems 377-18 system.

The Salmonella mutants were obtained by the one-step non-polar mutagenesis procedure (Datsenko and Wanner, 2000). The target gene was replaced with selectable antibiotic resistance gene markers. The resistance cassette was removed using the pCP20 plasmid. Each mutation was further characterized by sequencing to verify the authenticity of the deletion.

For transcriptional cat constructs, oligonucleotides (see Supplementary Table S2) were designed to amplify DNA fragments of different lengths from the ompC, ompF, ompS2, and ompR regulatory regions. PCR products were double-digested with BamHI-KpnI and ligated into pKK232-8 or pKK232-9 (Supplementary Table S1), which contain the promoterless cat gene. All constructs were sequenced to verify the correct DNA sequence of the PCR fragments.

To determine the expression of the cat reporter gene mediated by the S. Typhi promoters, chloramphenicol acetyltransferase (CAT) assays were performed according to a previously published protocol (Martínez-Laguna et al., 1999). Briefly, S. Typhi strains harboring the reporters were grown in N-MM or MA to different optical densities (OD), and the latter medium was supplemented when required with Ap and Km, with or without IPTG (isopropyl-β-d-thiogalactopyranoside; 50 μM). Cells were harvested, centrifuged, washed with 0.8 ml of TDTT buffer (50 mM Tris-HCl, 30 μM DL-dithiothreitol, and pH 7.8), resuspended in 0.5 ml of TDTT, and sonicated on ice for 10-s intervals with 10-s rest periods until the extract was clear. The homogenate was centrifuged at 12,000 g/15 min, and the supernatant used for activity measurement. For CAT assays, 5 μl of each extract were added in duplicate to a 96-well enzyme-linked immunosorbent assay (ELISA) plate, followed by the addition of 0.2 ml of a reaction mixture containing 1 mM DTNB [5,5'-dithiobis (2-nitrobenzoic acid)], 0.1 mM acetyl-coenzyme A (acetyl-CoA), and 0.1 mM chloramphenicol in 0.1 M Tris-HCl, pH 7.8. The absorbance at 412 nm was measured every 5 s for 5 min using a Ceres 900 scanning auto reader and microplate workstation. The protein concentration of the cell extracts was obtained using the bicinchoninic acid (BCA) protein assay reagent (Pierce). Protein values and the mean rate of product formation by CAT were used to determine CAT-specific activity as micromoles per minute per milligram of protein.

Salmonella Typhi IMSS-1 and S. Typhi Δcas-CRISPR harboring plasmid pFMTrcleuO-50 were grown in MA medium supplemented with Ap and IPTG (50 μM) to an optical density of 0.6 at 595 nm (OD₅₉₅). Salmonella cultures (100 ml) were pelleted and washed with 1X phosphate-buffered saline (PBS). Cellular proteins were obtained by sonication at 24 kHz for 1 min in the on position and 1 min in the off position, for five cycles at 4°C using a Vibra Cell (Sonics, United States), in the presence of a protease inhibitor (Complete tablets; Roche Diagnostics GmbH, Mannheim, Germany). To further limit proteolysis, protein isolation was performed using phenol extraction (Hurkman and Tanaka, 1986). To solubilize proteins and to obtain completely denatured and reduced proteins, pellets were dried and resuspended as previously reported (Encarnación et al., 2005). Prior to electrophoresis, samples were mixed with 7 M urea, 2 M thiourea, 4% 3-[(3-choloamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS; Roche Diagnostics GmbH, Germany), 2 mM tributylphosphine, 2% ampholytes, and 60 mM dithiothreitol.

Methods used for sample preparation, analytical two-dimensional gel electrophoresis (2-DGE), image analysis, and preparative 2-DGE were described previously (Encarnación et al., 2003). pH gradients were determined using a two-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis standard (Sigma, United States). For isoelectric focusing, 500 μg of total proteins were loaded. All gel experiments were repeated at least two times.

Selected spots from Coomassie blue-stained preparative one- or two-dimensional gels were excised manually and frozen at −70°C until use. Samples were prepared for mass spectrum analysis using a slight modification of a previously described procedure (Encarnación et al., 2005). Protein spots were destained, reduced, alkylated, and digested with trypsin (Promega, Madison, WI). Before the mass spectra of the peptide mixtures were obtained, the mixtures were desalted using a C₁₈ Zip Tip (Millipore, Bedford, MA) according to the manufacturer’s recommendations. Mass spectra were determined using a Bruker Daltonics Autoflex (Bruker Daltonics, Billerica, MA) operated in the delayed extraction and reflectron mode. Spectra were externally calibrated using a peptide calibration standard (Bruker Daltonics 206095). Peptide mixtures were analyzed using a saturated solution of alpha-cyano-4-hydroxycinnamic acid in 50% acetonitrile-0.1% trifluoroacetic acid. Peak lists of the tryptic peptide masses were generated and searched against the NCBInr databases using the Mascot search program (Matrix Science, London, United Kingdom).¹

Outer Membrane Proteins (OMPs) were isolated from S. Typhi IMSS-1 strains grown in N-MM to an OD₅₉₅ of 0.6 and 1.3 according to previous protocols (Puente et al., 1995). Fifteen milliliter of each culture was harvested and centrifuged at 5,000 g for 10 min at 4°C. Cells were resuspended in 500 μl of 10 mM Na₂HPO₄ buffer (pH 7.2) and sonicated on ice until the suspensions were clear. Intact cells and debris were eliminated by centrifugation (15,000 g) for 2 min, and the supernatants were transferred to clean microcentrifuge tubes and membrane fractions were pelleted by centrifugation at 12,000 g for 1 h at 4°C. Inner membrane proteins were solubilized by resuspension in 500 μl of 10 mM Na₂HPO₄ buffer, pH 7.2, containing 2% Triton X-100 for 30 min at 37°C. After incubation, the samples were centrifuged at 12,000 g for 1 h at 4°C. The remaining outer membrane insoluble fraction was washed with 500 μl of 10 mM Na₂HPO₄, pH 7.2, centrifuged at 12,000 g for 1 h at 4°C, and finally resuspended in 50 μl 1X PBS, pH 7.4. OMP concentrations were determined by BCA assay (Thermo), and 15 μg of each sample was analyzed by SDS-12% polyacrylamide gel electrophoresis. One-dimensional OMP gels were visualized by staining with Coomassie brilliant blue.

For western blot experiments, S. Typhi wild-type strain and its derivatives were grown in N-MM to OD₅₉₅ of 1.0 or MA medium to an OD₅₉₅ of 0.6. The cultures were supplemented, when required, with Ap and IPTG (50 μM). Fifteen milliliter of each culture was harvested and centrifuged at 5,000 g for 8 min. The pellets were resuspended in 600 μl of 1X PBS and sonicated on ice for 12 min at intervals of 10-s with 5-s rest. Total protein concentration was determined by BCA assay (Thermo), and 80 μg of each sample was loaded on a 10% SDS polyacrylamide gel. Following electrophoresis, proteins were transferred to 0.45-μm-pore-size polyvinylidene difluoride membranes (Immobilon; Millipore) using the Trans-Blot SD system (Bio-Rad) according to a previously described procedure (Guadarrama et al., 2014). Membranes were blocked with 10% non-fat milk and incubated with anti-OmpR or anti-GroEL (StressGen) polyclonal antibodies. Then, they were washed with 1X PBS, 0.1% Tween 20. Immunodetection was performed with a 1:10,000 dilution of horseradish peroxidase-conjugated Anti-Rabbit antibody (Pierce) for polyclonal antibodies, and the Western Lightning Plus-ECL Chemiluminescence Reagent Kit (PerkinElmer). The membranes containing the proteins were exposed to Carestream X-OMAT LS films.

Salmonella Typhi wild-type and the different mutant strains were grown 24 h in LB plates at 37°C. A bacterial colony was inoculated in liquid LB broth (5 ml) and grown for 16 h at 37°C/200 rpm. Then, 50 ml of LB broth supplemented with 5% sodium deoxycholate (Sigma Chemical, St. Louis, MO) were inoculated with the pre-inoculum to give an initial OD at 595 nm of 0.02. The cultures were incubated at 37°C/200 rpm during 15 h with OD₅₉₅ measurements being done every 2 h.

The quantification of biofilm formation was performed following a previous established protocol (O’Toole, 2011). Briefly, bacterial cells were grown overnight in LB broth (5 ml) at 37°C/200 rpm. Cells were diluted 1:100 in fresh LB without NaCl for stimulates biofilm production. One hundred microliter of this dilution was added per well in a 96-well polystyrene microtitre plate (Costar Cat. No. 3599, flat bottom with lid). Six replicate wells were prepared for each strain. Microtitre plates were incubated at 30°C for 24 h. Total bacterial growth was measured at OD₆₀₀, using a GloMax®-Muti Detection System (Promega). The planktonic cells were then discarded, and the plate was washed three times with water. The remaining biofilm was fixed with 200 μl per well of methanol (100%) and stained with a 0.2% solution of crystal violet in water. After incubation at room temperature for 10 min, the plates were rinsed three times with water. The dye was solubilized by adding 125 μl of 33% acetic acid to each well and incubated the microtiter plate at room temperature for 15 min. Finally, the OD₅₆₀ was determined with the microplate reader. The amount of formed biofilm is reported as the ratio of the OD₅₆₀/OD₆₀₀ values (Oropeza et al., 2015).

Studies on the regulation and the signals that induce the CRISPR-cas locus in S. Typhi are available (Hernández-Lucas et al., 2008; Medina-Aparicio et al., 2011, 2017). However, its biological function in this human pathogen remains to be determined. In this regard, previous results in Francisella novicida demonstrated that the CRISPR-Cas system is involved in the synthesis of outer membrane proteins (Sampson et al., 2013, 2014). Therefore, we obtained a strain devoid of cas3, the cas3-cse1 intergenic region, cse1, cse2, cas7, cas5, cas6e, cas1, cas2, and the CRISPR locus (the entire Type I-E CRISPR-Cas system, Figure 1), which was named as ΔCRISPR-cas (Supplementary Table S1). By electrophoretic profiles, the presence of the major outer membrane proteins OmpC, OmpF, and OmpA was detected in the wild-type strain; whereas in the isogenic S. Typhi strain devoid of CRISPR-cas locus, OmpC, and OmpF were not visualized (Figure 2A). To confirm these results, the transcriptional expression of ompC and ompF promoter regions was evaluated. Thus, the reporter plasmids pKK9/ompC-772 + 27 and pKK8/ompF-782 + 184 (Supplementary Table S1) were transformed into S. Typhi IMSS-1 wild type and, in the isogenic ΔCRISPR-cas strain, to perform CAT assays. The experiments showed that the transcriptional activity of the ompC and ompF regulatory regions were of 4,328 and 5,512 CAT units, respectively, in the wild-type strain. However, in the ∆CRISPR-cas strain the ompC and ompF activity decreased by 99 and 73%, respectively (Figure 2B). These data demonstrated that the CRISPR-Cas system is relevant for the expression of the major OmpC and OmpF porins in S. Typhi.

To determine the specific CRISPR-Cas genetic element involved in OmpC and OmpF regulation, a collection of individual cas mutants was generated, and porin profiles of these strains showed that Δcse1, Δcas7, and wild-type S. Typhi present a similar outer membrane protein profile. Nevertheless, in the absence of cas5 and cas2, the amount of OmpC decreased substantially; whereas in the individual cse2, cas6e, cas1 and cas3 mutants the OmpF porin was not observed (Figure 2C). These data support the fundamental role of specific Cas proteins in the regulation of OmpC and OmpF major outer membrane proteins and also are in agreement with the results obtained from the deletion of the entire CRISPR-Cas locus, since this strain lacks cas5, cas2, cse2, cas6e, cas1, and cas3, which resulted in the absence of the two main porins in S. Typhi (Figure 2A).

To continue with the identification of more CRISPR-Cas dependent outer membrane proteins, and since the overexpression of LeuO induces quiescent porins, such as OmpS2 (Fernández-Mora et al., 2004), the induction of this protein was evaluated in the absence of CRISPR-Cas. Salmonella Typhi IMSS-1 harboring plasmid pFMTrcleuO-50 and S. Typhi ∆CRISPR-cas containing pFMTrcleuO-50 were grown to an OD₅₉₅ of 0.6 in MA medium supplemented with IPTG (50 μM), and 2-DGE profiles were obtained with these cultures. The results showed the presence of OmpS2 in the wild-type strain. However, in the absence of the CRISPR-cas locus, OmpS2 decreased its expression by 99% (Figure 3A). Even more, the expression of a transcriptional fusion of the 5' intergenic region of ompS2 (pKK9/ompS2-482 + 77, Supplementary Table S1), upon induction of the LeuO regulator at various points in the growth curve, was essentially abolished in the ∆CRISPR-cas as compared with the wild-type strain (Figure 3B). Therefore, CRISPR-Cas is also fundamental for OmpS2 expression mediated by LeuO.

To determine the specific cas genetic element involved in OmpS2 regulation, the individual cas mutants were transformed with the transcriptional CAT fusion containing the 5' regulatory region of ompS2 and plasmid pFMTrcleuO-50 for overexpressing LeuO. The expression results showed that ompS2 activity mediated by LeuO depends on cse2, cas5, cas6e, cas1, cas2, and cas3, since in the absence of each of these genetic elements ompS2 was not transcribed (Figure 3C). These results indicated that the majority of Cas proteins, with exception of Cse1 and Cas7, are essential for the synthesis of the quiescent porin OmpS2. Thus, the presence of CRISPR-Cas cluster is essential for the synthesis of major and quiescent porins in S. Typhi.

The results mentioned above showed that CRISPR-Cas is involved in the synthesis of outer membrane proteins in S. Typhi. Interestingly, in the 2-DGE image shown in Figure 3A, a small spot of 27.3 kDa was absent in the ∆CRISPR-cas, and the mass spectrometry (MS) results of the same spot from S. Typhi IMSS-1, demonstrated that it corresponded to OmpR. To define whether CRISPR-Cas is involved in the control of the gene for this two-component system regulator, we evaluated its transcriptional expression in the S. Typhi wild type and in a ∆CRISPR-cas. The results showed that ompR displayed 941 CAT units in the wild-type strain, and the activity decreased by 60% in the CRISPR-Cas deficient S. Typhi strain (Figure 4A). Previously, it was demonstrated that ompR contains two promoters (Villarreal et al., 2014). To define whether the ompRP1 or ompRP2 promoters are under CRISPR-Cas control, the transcriptional activity of each promoter in the wild-type strain and in the ∆CRISPR-cas was evaluated. The transcriptional results showed ompRP2 activity values of 248 and 279 CAT units in the wild type and in the ∆CRISPR-cas isogenic strain, respectively (Figure 4A). Thus, CRISPR-Cas is not involved in ompRP2 promoter control. However, the activities obtained with ompRP1 were 120 and 37 CAT units in the wild type and in the ∆CRISPR-cas, respectively (Figure 4A). Therefore, CRISPR-Cas is involved in the regulation of the ompRP1 promoter to induce ompR expression.

To validate the results obtained and to determine the Cas proteins involved in ompRP1 genetic control, individual cas mutants were transformed with CAT fusions containing either the ompRP2 (pKK8/ompRP2-383-133) or the ompRP1 (pKK8/ompRP1-134-1) promoters. The activity results showed that in the S. Typhi wild type as well as in individual cse1, cse2, cas7, cas5, cas6e, cas1, cas2, and cas3 null mutants, the ompRP2 promoter expression was similar, supporting the notion that the Cas proteins are not implicated in its regulation (Figure 4B). In the case of the ompRP1 promoter, its genetic activity in the individual cse2, cas5, cas6e, cas1, cas2, and cas3 deficient strains was considerably reduced, compared with the CAT units obtained in the wild-type strain and in the cse1 and cas7 mutants (Figure 4C). The data support the proposal that the Cas proteins involved in ompRP1 promoter regulation correspond to Cse2, Cas5, Cas6e, Cas1, Cas2, and Cas3; whereas Cse1 and Cas7 are not implicated in ompRP1 induction.

To determine whether the reduction of ompRP1 promoter activity in the ∆CRISPR-cas, as well as in each cas individual mutant, has an effect on the synthesis of OmpR, western blot experiments were performed. The wild-type S. Typhi, the ∆CRISPR-cas, as well as the individual cas3, cse1, cse2, cas7, cas5, cas6e, cas1, and cas2 deletion mutants were grown in N-MM to an OD₅₉₅ of 1.0. Total crude cell protein extracts were transferred to membranes and probed using anti-OmpR polyclonal antibody. The western blot results showed a prominent OmpR band of 27.3 KDa in the wild-type strain and in the cse1 and cas7 individual mutants; whereas in the ∆CRISPR-cas strain, and in the individual cse2, cas5, cas6e, cas1, cas2, and cas3 mutants the OmpR protein was absent (Figure 4D).

These results explain the lack of OmpC and OmpF in the corresponding cas deficient strains (Figure 2C), since it is well-known that OmpR binds to their regulatory regions to promote their expression (Yoshida et al., 2006). Therefore, cse2, cas5, cas6e, cas1, cas2, and cas3 genes are fundamental for OmpR expression, whereas cse1 and cas7 are not involved in OmpR regulation, demonstrating that specific cas genes are necessary for OmpR production to control porin synthesis. Moreover, complementation of the S. Typhi ΔCRISPR-cas with the ompR gene on a plasmid restored the presence of OmpC and OmpF porins in this strain (Figure 4E), further supporting the notion that the deletion of the entire CRISPR-cas loci results in the lowering of the expression of the OmpR regulator and thus porin expression.

With respect to the OmpS2 quiescent porin, it is well accepted that LeuO counteracts the negative effect of H-NS on the ompS2 promoter, upon which OmpR binds to its regulatory region promoting ompS2 expression (Fernández-Mora et al., 2004). Thus, it was determined whether the OmpR protein was produced in the S. Typhi strains that overexpress LeuO. Western blot experiments demonstrated the presence of OmpR in S. Typhi IMSS-1 wild-type strain harboring the pFMTrcleuO-50 plasmid. However, OmpR was not detected in the S. Typhi ∆cas-CRISPR mutant overexpressing LeuO (Figure 4F). Thus, OmpS2 was not visualized in the 2-DGE of this strain (Figure 3A) because of the lack of the two-component system regulator OmpR.

In conclusion, the results obtained here showed that the CRISPR-Cas system acts hierarchically on the ompRP1 promoter to induce OmpC, OmpF, or OmpS2 synthesis in S. Typhi.

In this report, we have shown that the S. Typhi cse2, cas5, cas6e, cas1, cas2, and cas3 genes are involved in porin synthesis through the regulation of the ompR gene which codes for the OmpR transcriptional regulator. In previous studies, it has been demonstrated that ompR is involved in virulence, sodium deoxycholate resistance, biofilm formation, the production of flagella, and curli (Pickard et al., 1994; Shin and Park, 1995; Vidal et al., 1998; Cameron and Dorman, 2012; Villarreal et al., 2014). Therefore, we evaluated whether the cas genes are involved in some of these biological processes. Growth rate experiments of S. Typhi IMSS-1 and the cas individual deleted strains were performed in LB broth supplemented with 5% of the human bile salt sodium deoxycholate. The results showed that the wild-type S. Typhi strain grew in this condition, reaching an OD₅₉₅ of 0.86 after 15 h. However, growth of the ∆cas5, ∆cas2, and ∆cas-CRISPR mutant strains was impaired in the presence of this bile salt (Figure 5A) since their OD₅₉₅ were of 0.43, 0.54, and 0.18, respectively, after 15 h of incubation. Remarkably, these strains did not express the OmpC porin, which was previously shown to be determinant for allowing S. Typhi to proliferate in the presence of sodium deoxycholate (Villarreal et al., 2014). The growth rate of ∆cas6e, ∆cas1, and ∆cas3 was similar to that observed with the wild-type strain (Figure 5A), consistent with the presence of the OmpC porin in these mutants. The same experiment was performed with ∆ompR, ∆ompC, ∆ompF, and ∆ompS2 strains. As expected, the ompR and ompC mutants were also impaired in their growth in 5% sodium deoxycholate (OD₅₉₅ = 0.5 and 0.33, respectively; Figure 5B), as previously reported (Villarreal et al., 2014); whereas the ∆ompF and ∆ompS2 mutant strains grew like the S. Typhi IMSS-1 wild type.

Additionally, we also evaluated the biofilm formation ability of the S. Typhi IMSS-1 wild type, and of the ∆cse2, ∆cas5, ∆cas6e, ∆cas1, ∆cas2, ∆cas3, ∆cas-CRISPR, ∆ompR, ∆ompC, ∆ompF, and ∆ompS2 mutant strains. The experiments showed that the wild type produced moderate biofilm (0.26 OD₅₆₀/OD₆₀₀ ratio). However, the cse2, cas5, cas6e, cas1, cas2, cas3, and cas-CRISPR mutants displayed an increased biofilm formation (OD₅₆₀/OD₆₀₀ ratio of 1.06, 1.04, 0.95, 1.04, 1.17, 0.90, and 0.50, respectively; Figure 6). These results suggest that the CRISPR-Cas system negatively regulates genes involved in biofilm production, i.e., that the absence of cas genes allows the expression of factors that increase the ability of S. Typhi to form biofilm.

The ∆ompR mutant showed an increased biofilm formation (0.77), as compared to the values obtained with the wild type (0.26; Figure 6). However, strains carrying deletions either in the ompC, ompF, or the ompS2 genes presented a slightly decreased biofilm formation (OD₅₆₀/OD₆₀₀ ratio of 0.15, 0.14, and 0.18, respectively; Figure 6). Therefore, the biofilm production was independent of the individual absence of the OmpC, OmpF, or OmpS2 porins. Interestingly, it has been demonstrated that ompR mutants in Salmonella enteritidis, Salmonella pullorum, E. coli, and Yersinia enterocolitica presented a decreased biofilm formation ability (Dong et al., 2011; Lu et al., 2012; Samanta et al., 2013; Meng et al., 2019), suggesting that the pathway toward regulating biofilm synthesis is different in S. Typhi.

The data shown are consistent with the notion that the CRISPR-Cas system is relevant for S. Typhi virulence, since this pathogen needs to survive the presence of bile salts in the gut and gallbladder, as well as to persist inside the gallbladder, where the biofilm formation is relevant (Crawford et al., 2010; Gonzalez-Escobedo et al., 2011; Spector and Kenyon, 2012).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.