The virulence regulator CovR boosts CRISPR-Cas9 immunity in Group B Streptococcus

Abstract

CRISPR-Cas9 immune systems protect bacteria from foreign DNA. However, immune efficiency is constrained by Cas9 off-target cleavages and toxicity. How bacteria regulate Cas9 to maximize protection while preventing autoimmunity is not understood. Here, we show that the master regulator of virulence, CovR, regulates CRISPR-Cas9 immunity against mobile genetic elements in Streptococcus agalactiae, a pathobiont responsible for invasive neonatal infections. We show that CovR binds to and represses a distal promoter of the cas operon, integrating immunity within the virulence regulatory network. The CovR-regulated promoter provides a controlled increase in off-target cleavages to counteract mutations in the target DNA, restores the potency of old immune memory, and stimulates the acquisition of new memory in response to recent infections. Regulation of Cas9 by CovR is conserved at the species level, with lineage specificities suggesting different adaptive trajectories. Altogether, we describe the coordinated regulation of immunity and virulence that enhances the bacterial immune repertoire during host-pathogen interaction.

Fig. 1. CovR represses cas transcription via the P2_cas distal promoter.

A Organization of the CRISPR-Cas9 chromosomal locus. Upper: the cas operon and the crRNA encoded by the CRISPR array are organized in two transcriptional units. Lower: zoom-in on the upstream region of cas9 containing the tracrRNA with its short and long promoters (P_tracr-S and P_tracr-L) and the two cas promoters (P1_cas and P2_cas). The tracr-L long transcript represses cas transcription by binding to the P1_cas promoter. The absence of the P1_cas promoter following a chromosomal deletion in CC-17 strains is highlighted, as well the region corresponding to CovR binding identified by ChIP-seq. B CovR binds to the P2_cas promoter in vivo. ChIP-seq profiles of FLAG-tagged CovR at the P2_cas loci in two WT strains (BM110 and NEM316) in non-inducing (-) and inducing condition (+). C CovR binds to the P2_cas promoter in vitro. Upper panel: Electrophoretic mobility shift assay with recombinant rCovR showing delayed migration of the rCovR-P2_cas complex compared to the free P2_cas radiolabelled probe. Lower panel: DNase I protection assay with increasing rCovR concentrations. The in vitro rCovR-protected sequence is highlighted, along with the P2_cas transcriptional start site (+1) and the 3’ end of the tracrRNA. Source data are provided as a Source Data file. (D) CovR represses cas transcription. Gene expression by RT-qPCR relative to the BM110 strain in the ΔcovR mutant (orange), the ΔcovR/covR complemented strain (light violet), and the double ΔcovR ΔP2_cas mutant (light yellow) in which the −10 box and +1 TSS of P2_cas were mutated. In addition to CRISPR-cas genes (cas9, cas1, csn2, and crRNA), expression of the cylI gene directly repressed by CovR and required for toxin synthesis is provided as control. Bars represent the mean ± SD of biological replicates (n = 3). Significant P-value from multiple two-tailed unpaired t-tests (FDR 1%, two-stage method of Benjamini- Krieger and Yekutieli) against the WT strain are highlighted with stars (****P < 0.0001; *** <0.001; ** <0.01; * <0.05). Source data, including exact P-value, are provided as a Source Data file. (E) CovR regulates P2_cas activity. The P2_cas and P_tracr promoters were cloned upstream of a ß-galactosidase reporter in the pTCV-lac vector. Promoter activity was quantified in BM110 (WT: dark violet) and mutants using colorimetric assays. Bars represent the mean ± SD of biological replicates (n = 3). Significance was determined by unpaired, parametric, two-tailed t-test (****P < 0.0001; ns not significant). Source data, including exact P-value, are provided as a Source Data file. F CovR represses Cas9 expression. The FLAG epitope coding sequence was introduced in-frame at the 5’ end of cas9 (cas9::FLAG) in BM110 (WT) and in ∆covR, ΔP2_cas, and ΔcovR ΔP_cas mutants. Upper: representative Western blot from a biological duplicate (n = 2) using total protein extract and anti-FLAG antibody. Lower: corresponding Coomassie staining of total proteins used as loading control. Source data are provided as a Source Data file.

Fig. 2. CovR inactivation increases the immune repertoire.

A CRISPR immunity assay. Protospacers (X, Y, or Z) corresponding to spacer sequences in the genomic CRISPR array are cloned into the shuttle vector pTCV and transformed into GBS strains. The number of kanamycin resistant transformants by µg of empty vector (N_empty) is used as reference to infer the relative transformation efficiency for each protospacer (R_T). Immunity results in cleavage of the vector and the absence of transformants (R_T = 0), while ineffective binding or cleavage of the protospacer results in a number of transformants similar to that obtained with the empty vector (R_T = 1). The increase in immune repertoire refers to a difference in R_T between the WT and the ∆covR mutant for a given protospacer. B Immunity depends on spacer position and CovR inactivation. Immunity assays (Log2 R_T) in BM110 strain (WT, dark violet) and ∆covR mutant (orange) with protospacers corresponding to the first (P1), internal, (P4, P8, P12), and last (P13) spacers in the CRISPR array. Controls without a protospacer adjacent motif (-PAM) are included. Full immunity is defined as no transformants (below the limit of detection of 10³ transformants per µg) and is represented by a dashed line (Log2 = −10). C CovR inactivation enhances immunity against single-point mutation. Similar immunity assays using single mutations (transversion) introduced into the P8 protospacer sequence at the PAM-proximal position 1 to 14. D CovR inactivation enhances immunity against mutated protospacers. Similar immunity assays using consecutive mutations in the P4 protospacer. E CovR acts through the P2_cas promoter. Immunity assays with non-optimal spacers in the ΔcovR/covR complemented strain (light violet) and the ΔcovR ΔP_cas double mutant (light yellow). For individual immunity assays (B–E), bars represent the mean and error bars the SD of at least three biological triplicates (n = 3). Significant P-values from multiple two-tailed unpaired t-tests (FDR 1%, two-stage method of Benjamini, Krieger, and Yekutieli) are highlighted with stars (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant). Source data, including exact P-value, are provided as a Source Data file.

Fig. 3. CovR inactivation confers tolerance to mutations in the seed region.

A Bulk CRISPR immunity assay. Synthetic protospacers were designed and cloned en masse to obtain a pool of plasmids with all possible single mutations at each position. The plasmid pool (INPUT) was transformed into the wild-type strain BM110 (WT) and the ∆covR mutant. Kanamycin-resistant GBS transformants were then pooled and plasmid extracted (OUTPUT pools). Vector inserts were amplified by PCR, and the amplicons sequenced by Illumina to quantify the proportion of each protospacer sequence in each condition. B Bulk CRISPR immunity assay. Normalized counts of all possible single mutations at each position of the P4 and P8 protospacer in the input pool (white dots) and after transformation in the WT (blue dots) or ∆covR mutant (orange dots). Each dot represents the mean +/− SD of biological replicate for the WT and ∆covR mutant (n = 2), and of technical replicate (n = 3 plasmid purification) for the input pool. C CovR inactivation increases immunity to mutated seed regions. Gain in immunity (Log₂ fold change: green to brown) for the ∆covR mutant compared to the WT strain for all possible single mutations in the P4 and P8 protospacers. The wild-type spacer sequences are depicted with a dot at each position. Statistical analysis with DESeq2 uses the Wald test to compute p-values, followed by Benjamini-Hochberg correction to adjust for multiple comparisons. Significances (p-adj <0.05) are highlighted with black square borders. Raw count, p-values, and p-adj are provided in Supplementary Data S1.

Fig. 4. CovR inactivation promotes adaptative immunity.

A Spacer acquisition assay. Acquisition of a new spacer at the leader end of the CRISPR array is tested by PCR with specific forward (Fw) and reverse (Rv) primers. Spacer acquisition was monitored in the population over 9 serial subcultures without selective antibiotic pressure to maintain the pZN123 vector. B Representative PCRs showing products corresponding to N and N + 1 spacer (+66 bp) over time in the WT, ∆covR mutant, and ∆covR/covR complemented strain. Source data are provided as a Source Data file. C Quantification of new spacer acquisition rate. Proportion of N (light violet) and N + 1 (light yellow) spacers in the bacterial population are quantified by densitometry of PCR products. Data represent the mean ± SD of biological replicates (n = 3). Significance against the WT for each subculture was determined by unpaired, parametric, two-tailed t-tests (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant). Source data, including PCR gels and exact P-value, are provided as a Source Data file.

Fig. 5. The P1_cas promoter dampens CovR-dependent P2_cas regulation.

A Presence (blue: P1+) or absence (red: P1∆) of the constitutive P1_cas promoter in the GBS population. The number of genomes analysed in each clonal groups (CG) is given in brackets. B Diversity of the CRISPR-Cas9 locus in the population. Left: core-gene phylogeny of representative isolates of the major sequence lineages (SL). Right: co-phylogeny of Cas9 variants. The percentage of identities (%) for P1+ (blue), P1∆ (red), and Cas9 (grey) in the population is indicated by the colour gradient. C Quantification of gene expression in ∆covR mutants by RT-qPCR relative to WT strains. Transcription of the cas9 and csn2 genes were quantified in the P1∆ strains BM110 (red) and COH1 (SL-17; light red), and in the P1+ strains NEM316 (SL-23; blue), 515 (SL-23; dark blue), 2603 V/R (SL-110; grey), and A909 (SL-7; light grey). Transcription levels of the CovR-regulated virulence gene cylE are provided. Data represents the mean +/− SD of biological replicates (n = 3). Significant P-value from multiple two-tailed unpaired t-tests (FDR 1%, two-stage method of Benjamini-Krieger and Yekutieli) against the corresponding WT strain are highlighted with stars (****P < 0.0001; ***<0.001; **<0.01; * <0.05; ns, not significant). Source data, including exact P-value, is provided as a Source Data file. D Quantification of cas transcription in ∆covR mutants relative to WT strains by RNA-seq. Dots represent the mean fold change of biological triplicate (n = 3) for the four genes of the cas operon in the ∆covR mutants of strains NEM316 (blue), 2603 V/R (grey), and A909 (light grey). Statistical analysis with DESeq2 uses the Wald test to compute p-values, followed by Benjamini-Hochberg correction to adjust for multiple comparisons. Raw count, normalization, and adjust P-value are provided in Supplementary Data S2.

Fig. 6. Conservation of CovR-regulated immunity in the GBS population.

A CovR inactivation increases NEM316 immunity to sub-optimal spacers. Immunity assays in NEM316 (blue) and ∆covR (light yellow) with protospacers corresponding to spacers at positions 1, 3, 5, 6, 9, 11, 13, and 14 in the NEM316 CRISPR array. The mutation (C- > T) in the repeated sequence of the CRISPR array between spacers 5 to 11 is highlighted in red below the graphic. B CovR inactivation increases NEM316 immunity to mutated protospacers. Similar immunity assays with the n3 protospacer containing single mismatches at position 5 to 10. C CovR inactivation increases NEM316 immune memory. Spacer acquisition assay with representative PCRs (left) and the percentage of the population (right) having N (light violet) or N + 1 (light yellow) spacers in NEM316 and ∆covR. D Quantification of gene expression by RT-qPCR in the NEM316 strain deleted for the P1_cas promoter (dark violet) and in the ∆covR ∆P1_cas double mutant (yellow). Transcription of the cas and cyl genes was quantified relative to the WT strain. E Immunity assays in NEM316 (blue), ∆P1_cas mutant (dark violet), and ∆covR ∆P1_cas double mutant (yellow) with protospacers corresponding to spacers at position 1, 3, 9, 11, and 14 in the CRISPR array. Deletion of P1_cas decreases immune efficiency, while an additional deletion of covR restores immunity to a WT level. For all panels (A–E), bars represent the mean and error bars the standard deviation (SD) for biological triplicate (n = 3). Statistical significance in all panels (A–E) is determined by multiple two-tailed unpaired t-tests (FDR 1%, two-stage method of Benjamini- Krieger and Yekutieli) and highlighted with stars (****P < 0.0001; *** <0.001; ** <0.01; * <0.05; ns, not significant). Source data, including PCR gels and exact P-value, are provided as a Source Data file.