The CRISPR-Cas System Differentially Regulates Surface-Attached and Pellicle Biofilm in Salmonella enterica Serovar Typhimurium

ABSTRACT The CRISPR-Cas mediated regulation of biofilm by Salmonella enterica serovar Typhimurium was investigated by deleting CRISPR-Cas components ΔcrisprI, ΔcrisprII, ΔΔcrisprI crisprII, and Δcas op. We determined that the system positively regulates surface biofilm while inhibiting pellicle biofilm formation. Results of real-time PCR suggest that the flagellar (fliC, flgK) and curli (csgA) genes were repressed in knockout strains, causing reduced surface biofilm. The mutants displayed altered pellicle biofilm architecture. They exhibited bacterial multilayers and a denser extracellular matrix with enhanced cellulose and less curli, ergo weaker pellicles than those of the wild type. The cellulose secretion was more in the knockout strains due to the upregulation of bcsC, which is necessary for cellulose export. We hypothesized that the secreted cellulose quickly integrates into the pellicle, leading to enhanced pellicular cellulose in the knockout strains. We determined that crp is upregulated in the knockout strains, thereby inhibiting the expression of csgD and, hence, also of csgA and bcsA. The conflicting upregulation of bcsC, the last gene of the bcsABZC operon, could be caused by independent regulation by the CRISPR-Cas system owing to a partial match between the CRISPR spacers and bcsC gene. The cAMP-regulated protein (CRP)-mediated regulation of the flagellar genes in the knockout strains was probably circumvented through the regulation of yddx governing the availability of the sigma factor σ28 that further regulates class 3 flagellar genes (fliC, fljB, and flgK). Additionally, the variations in the lipopolysaccharide (LPS) profile and expression of LPS-related genes (rfaC, rfbG, and rfbI) in knockout strains could also contribute to the altered pellicle architecture. Collectively, we establish that the CRISPR-Cas system differentially regulates the formation of surface-attached and pellicle biofilm. IMPORTANCE In addition to being implicated in bacterial immunity and genome editing, the CRISPR-Cas system has recently been demonstrated to regulate endogenous gene expression and biofilm formation. While the function of individual cas genes in controlling Salmonella biofilm has been explored, the regulatory role of CRISPR arrays in biofilm is less studied. Moreover, studies have focused on the effects of the CRISPR-Cas system on surface-associated biofilms, and comprehensive studies on the impact of the system on pellicle biofilm remain an unexplored niche. We demonstrate that the CRISPR array and cas genes modulate the expression of various biofilm genes in Salmonella, whereby surface and pellicle biofilm formation is distinctively regulated.

Supplementary Figure S1: Schematic representation for generating and confirming the knockout strains. The successful generation of knockout strains (ΔcrisprI, ΔcrisprII , Δcas op, and ΔΔcrisprI crisprII) would require homologous recombination between the gene of interest (GOI) and the antibiotic resistance cassette. For ΔcrisprI, ΔcrisprII , and Δcas op the genes were replaced with chloramphenicol resistance cassette, whereas for generation of ΔΔcrisprI crisprII, the crisprI gene was replaced with kanamycin resistance cassette in the ΔcrisprII strain.
Supplementary Figure S2: The deletion of the CRISPR-Cas components was confirmed through PCR using expression primers (A), and confirmatory primers (B). The colony PCR of potential knockout strains was done using respective primers mentioned in Supplementary Figure S1, and the amplicons were visualized using agarose electrophoresis. A. The presence of CRISPR-Cas genes was checked in WT and knockout strains (ΔcrisprI, ΔcrisprII , Δcas op, and ΔΔcrisprI crisprII), while 16s rRNA was used as a positive control for each strain. B. Amplicons of appropriate sizes were obtained for each knockout strain, whereas WT did not yield any bands.
Supplementary Figure S4: The CRISPR-Cas system knockout strains of S. enterica subsp. enterica serovar Typhimurium 14028s showed a similar growth trend to wild-type in LB without NaCl media. The S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media for 12 h, at 37°C, shaking condition. The graph represents OD620nm for each strain.
Supplementary Figure S5: Morphology of air-exposed side of surface-attached (glass) biofilm at early (24 h) time point. A. The knockout (ΔcrisprI, ΔcrisprII, Δcas op, and ΔΔcrisprI crisprII) strains formed patchy bacterial aggregates, in comparison to wild-type (WT), which had tightly packed bacterial aggregates covering larger area, with a few dome-like structure (arrow-head in the WT micrograph). Few elongated cells (arrow-head in the micrographs) were also observed in the biofilms of the knockout strains. The strains were grown in LB without NaCl media for 24 h, at 25°C, static conditions. The pellicle biofilms formed was fixed using 2.5% glutaraldehyde were dehydrated with increasing concentrations of ethanol. The images were captured at 5000x magnification and scaled to bar. B. The graph represents the average size (in µm) of WT and Knockout strains. Unpaired t-test was used to determine significant differences between the WT and knockout strains. Error bars indicate SD. Statistical significance: *≤ 0.05, **≤ 0.01, ***≤ 0.001, ****<0.0001, ns = not significant. A.U., arbitrary units.
Supplementary Figure S6: CRISPR-Cas system knockout strains show reduced swarming motility. Swarm plates (0.5% agar, 20g/L of LB and 0.5% glucose) were point inoculated with overnight cultures and incubated at 37°C for 9 h. The complement strains (ΔcrisprI + pcrisprI, ΔcrisprII+ pcrisprII) showed reversal of swarming ability confirming the mutation process was not polar.
Supplementary Figure S7: Silver-stained Lipopolysaccharide (LPS) profiling of wild-type (WT), and CRISPR-Cas system knockout strains. The variation in O-antigen was analyzed by LPS profiling. The strains were grown in LB without NaCl media for 96 h, at 25°C, static conditions. pellicle biofilm was homogenized, and heated, followed by DNase, RNases and Proteinase-K treatment to extract crude LPS. The processed samples were loaded on 15% SDS-PAGE MIDI gel, which was later stained using a silver staining kit. Variations in banding pattern and intensity between knockout (ΔcrisprI, ΔcrisprII, Δcas op, and ΔΔcrisprI crisprII) strains and WT were observed in long O-Ag, low molecular weight O-Ag and core glycoforms regions. #Ratio indicates the intensity of the bands observed on the gel for all strains normalized by the intensity of the band for wildtype. The graph represents dry pellicle biofilm weight (in gms) of each strain normalized by the dry pellicle biofilm weight (in gms) of WT at respective time points. B. The metabolic activity was assessed by resazurin assay. S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media for 96 h, at 25°C, static condition. The pellicle biofilm formed after 96 h incubation was stained with resazurin dye and fluorescence was measured using a fluorimeter at excitation (λEx) 550 nm and emission (λEm) of 600 nm. The graph represents fluorescence intensity observed for each strain normalized by fluorescence intensity of WT. C. The S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media for different time point (24 h and 96 h), at 25°C, static condition. The pellicle biofilm formed was stained with SYTO 9, for 30 mins in the dark, at RT. The graph represents Mean intensity of SYTO9 observed for each strain. D. Qualitative analysis of the amount of cellulose present in the pellicle biofilm was done by measuring the calcofluor bound, at excitation of 350 nm and emission 475 nm. The S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media 96 h, at 25°C, static condition. E-G. S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media for 96 h, at 25°C, static condition. E. The exopolysaccharides were quantified by the phenol-sulfuric acid method, by measuring absorbance at 490nm. The graph represents absorbance observed at 490nm for each strain normalized by absorbance observed at 490nm for WT. F & G. The protein and DNA concentrations in the supernatants of each sample was estimated spectrophotometrically and was further normalized by absorbance for WT in each case. Unpaired t-test was used to determine significant differences between the WT and knockout strains. Error bars indicate SD. Statistical significance: *≤ 0.05, **≤ 0.01, ***≤ 0.001, ****<0.0001, ns = not significant. A.U., arbitrary units.
Supplementary Figure  Though thicker than wild-type, pellicle biofilms formed by CRISPR-Cas knockout strains were found to be more delicate (C). Curli production in the pellicle biofilms and planktonic culture of wild-type, CRISPR, and cas operon knockout strains was assessed with the help of Congo red depletion (A), and Thioflavin (ThT) Fluorescence intensity (B). The S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media 48 h, at 25°C, static condition. A. Congo red depletion was determined by measuring absorbance of the supernatant of cultures stained with Congo-red at 500nm. The graph represents absorbance at 500nm for each strain, normalized by absorbance at 500nm for WT. B. Thioflavin (ThT) Fluorescence intensity was determined by measuring absorbance at excitation 440 nm and emission 482 of nm. ΔcsgD was used as a negative control. The graph represents intensity readings of each strain, normalized by intensity readings of WT. C. The S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔΔcrisprI crisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media for 96 h, at 25°C, static condition. The pellicle biofilm strength was determined by addition of glass beads (1 mm, HiMedia) using a tweezer until disruption (collapse of pellicle biofilm to the bottom). The glass bead weight tolerated by pellicle biofilm of each strain was normalized to that of WT. Unpaired t-test was used to determine significant differences between the WT and knockout strains. Error bar indicates SD. Statistical significance: *≤ 0.05, **≤ 0.01, ***≤ 0.001, ****<0.0001, ns = not significant. A.U., arbitrary units.
Supplementary Figure S12: CRISPR-Cas system knockout strains showed differences in the expressions of genes associated with flagellar protein flgJ (A), fljB (B), rfbG (C), rfbI (D),and bcsA(E) when compared to WT. The S. Typhimurium strain 14028s wild-type (WT), CRISPR (ΔcrisprI, ΔcrisprII and ΔcrisprI ΔcrisprII) and cas operon (Δcas op) knockout strains were cultured in LB without NaCl media for different time periods (24 h and 96 h), at 25°C, static condition. Total RNA was isolated from bacteria (24 h) and pellicle biofilm (96 h). 1 µg of RNA was used for cDNA synthesis, followed by qRT-PCR. Relative expression of the gene was calculated using the 2 -ΔΔCt method, and normalized to reference gene rpoD.