Listeria Phages Induce Cas9 Degradation to Protect Lysogenic Genomes

Abstract

Bacterial CRISPR-Cas systems employ RNA-guided nucleases to destroy phage (viral) DNA. Phages, in turn, have evolved diverse "anti-CRISPR" proteins (Acrs) to counteract acquired immunity. In Listeria monocytogenes, prophages encode two to three distinct anti-Cas9 proteins, with acrIIA1 always present. However, the significance of AcrIIA1's pervasiveness and its mechanism are unknown. Here, we report that AcrIIA1 binds with high affinity to Cas9 via the catalytic HNH domain. During lysogeny in Listeria, AcrIIA1 triggers Cas9 degradation. During lytic infection, however, AcrIIA1 fails to block Cas9 due to its multi-step inactivation mechanism. Thus, phages encode an additional Acr that rapidly binds and inactivates Cas9. AcrIIA1 also uniquely inhibits a highly diverged Cas9 found in Listeria (similar to SauCas9) and Type II-C Cas9s, likely due to Cas9 HNH domain conservation. In summary, Listeria phages inactivate Cas9 in lytic growth using variable, narrow-spectrum inhibitors, while the broad-spectrum AcrIIA1 stimulates Cas9 degradation for protection of the lysogenic genome.

Keywords: CRISPR-Cas; Cas9; Listeria; anti-CRISPR; bacteriophage; lysogen; prophage.

Figure 1.. AcrIIA1 Induces Cas9 Degradation in Listeria

(A-B) Immunoblots detecting FLAG-tagged LmoCas9 protein and a non-specific (ns) protein loading control in Listeria monocytogenes strain 10403s (Lmo10403s) lysogenized with the indicated wild-type prophages (A, top) or Lmo10403s containing Acr-expressing plasmids (B, top). Dashed lines indicate where intervening lanes were removed for clarity (B, top). Representative blots of at least three biological replicates are shown (A and B). Schematics of translational and transcriptional reporters used to measure Lmo or Spy Cas9 protein and mRNA levels in Lmo10403s (A, middle). Cas9 translational (black bars) and transcriptional (gray shaded bars) reporter measurements reflect the mean percentage mCherry relative fluorescence units (RFU/OD₆₀₀) in the indicated lysogens (A, bottom) or strains with Acr-expressing plasmids (B, bottom) relative to the control strain lacking a prophage (−prophage) (A, bottom) or containing an empty vector (B, bottom). Error bars represent the mean ± SD of at least three biological replicates. (C) SpyCas9-mCherry protein levels post Acr induction or translation inhibition. Lmo10403s expressing SpyCas9-mCherry from the constitutively active pHyper promoter and AcrIIA1 or AcrIIA4 from an inducible promoter were grown to mid-log phase and treated with 100 mM rhamnose to induce Acr expression (dashed lines) or 100 mM glycerol as a neutral carbon source control (solid lines) and 5 μg/mL gentamicin (Gent) to inhibit translation (+) or water (−) as a control. SpyCas9-mCherry protein measurements reflect the mean percentage fluorescence (RFU/OD₆₀₀) relative to the SpyCas9-mCherry levels at the time translation inhibition was initiated (0 min). Error bars (vertical lines) represent the mean ± SD of at least three biological replicates. Data were fitted by nonlinear regression to generate best-fit decay curves. See Figure S1C for additional controls and S1B for data showing tight repression of the pRha promoter under non-inducing conditions. Note: Lmo doubling time is significantly slower in LB media containing glycerol and/or rhamnose carbon sources (Fieseler et al., 2012).

Figure 2.. AcrIIA1 Selectively Binds Catalytically Active Cas9 to Trigger its Degradation

(A-B) Acr-mediated inhibition of CRISPRi (A) or self-targeting (B). Lmo10403s contains chromosomally-integrated constructs expressing dead (A) or catalytically active (B) LmoCas9 from the inducible pRha-promoter and sgRNA that targets the pHelp-promoter driving mCherry expression. For CRISPRi, mCherry expression measurements reflect the mean percentage fluorescence (RFU/OD₆₀₀) in deadCas9-induced cells relative to uninduced controls (−dCas9) of three biological replicates ± SD (error bars) (A). For self-targeting, bacterial growth was monitored after LmoCas9 induction (orange lines) or no induction (blue lines) and data are displayed as the mean OD₆₀₀ of three biological replicates ± SD (error bars) (B). See Figure S2A for CRISPRi data with Lmo10403s expressing deadSpyCas9. (C) Translational (black bars) and transcriptional (gray shaded bars) reporter levels of catalytically active (left) and dead LmoCas9 (right) in Lmo10403s lysogenized with engineered isogenic ΦA006 prophages. Cas9 reporter measurements reflect the mean percentage mCherry relative fluorescence units (RFU/OD₆₀₀) in the indicated lysogens relative to the control strain lacking a prophage (−prophage). Error bars represent the mean ± SD of at least three biological replicates. Asterisk (*) denotes genes containing the native orfA RBS (strong) in ΦA006 and unmarked genes contain their native RBS. See Figure S2B for equivalent data with Lmo10403s expressing SpyCas9. (D) Quantification of the binding affinities (K_D; boxed inset) of Acr proteins for WT, catalytically dead (dCas9), or nickase (D10A or H840A) SpyCas9-gRNA complexes using microscale thermophoresis. Data shown are representative of three independent experiments. See Figure S2D for additional controls with AcrIIA2b.3 (IIA2, dashed line). (E) Time course of SpyCas9 DNA cleavage reactions in the presence of Acr proteins that were recombinantly purified from E. coli. Dashed lines indicate where intervening lanes were removed for clarity. Solid lines indicate a separate image. Data shown are representative of three independent experiments.

Figure 3.. AcrIIA1 Inhibits Cas9 to Protect Prophages During Lysogeny

(A) Left: Representative image of plaquing assays where isogenic ΦA006 phages are titrated in ten-fold serial dilutions (black spots) on a lawn of Lmo10403s (gray background). Dashed lines indicate where intervening rows were removed for clarity. Right: Efficiency of plaquing (EOP) of isogenic ΦA006 phages expressing the indicated Acrs on Lmo10403s. Plaque forming units (PFUs) were quantified on Lmo10403s overexpressing the first spacer in the native CRISPR array that targets ΦA006 (cas9;pHyper-spacer#1) and normalized to the number of PFUs measured on a non-targeting Lmo10403s-derived strain (Δcas9). Data are displayed as the mean EOP of at least three biological replicates ± SD (error bars). See Figure S3A for EOP measurements of additional ΦA006 phages. (B) Bacterial growth curves of self-targeting Lmo10403s::ΦA006 isogenic lysogens expressing the indicated Acrs and rhamnose-inducible WT or dead LmoCas9. WT LmoCas9 (blue lines) is lethal in an Acr-deficient (Δacr) strain because the Lmo10403s CRISPR array contains a spacer targeting the ΦA006 prophage integrated in the bacterial genome. Data are displayed as the mean OD₆₀₀ of at least three biological replicates ± SD (error bars) as a function of time (min). Asterisk (*) indicates the native orfA RBS (strong) in ΦA006 was used for Acr expression.

Figure 4:. AcrIIA1 Uses its C-terminal Domain to Lock Cas9 in an Inhibited State

(A) Left: Alignment of AcrIIA1 homolog protein sequences denoting key residues. Right: Phylogenetic tree of the protein sequences of AcrIIA1 homologs. See companion manuscript for a complete alignment of the AcrIIA1 homolog protein sequences (Osuna et al., 2020b). (B-C) Fold reduction in phage titer in response to SpyCas9 targeting of a P. aeruginosa DMS3m-like phage in the presence of AcrIIA1 homologs (B) or mutants (C). The percent protein sequence identities of each homolog to the full-length (FL) or domains (NTD or CTD) of AcrIIA1_ΦA006 are listed in (B). The displayed fold reductions in phage titer were qualitatively determined by examining three biological replicates of each phage-plaquing experiment. See Figure S4B for representative pictures of the corresponding phage-plaquing experiments. (D) Immunoblots detecting GST-tagged anti-CRISPR proteins that co-immunoprecipitated with Myc-tagged SpyCas9 in a P. aeruginosa strain heterologously expressing Type II-A SpyCas9-gRNA and the indicated Acrs. For input samples, one-hundredth lysate volume was analyzed to verify tagged protein expression and RNA-polymerase was used as a loading control. Representative blots of three biological replicates are shown. See Figure S4E for the reciprocal GST-Acr pulldown. (E) Time course of SpyCas9 DNA cleavage reactions conducted with SpyCas9-gRNA-Acr (or no Acr, −) complexes immunoprecipitated from P. aeruginosa. Dashed lines indicate where intervening lanes were removed for clarity. Data shown are representative of three independent experiments. See Figure S4F for reactions with AcrIIA1 mutants.

Figure 5.. AcrIIA1 is a Broad-Spectrum Cas9 Inhibitor

(A) Phylogenetic tree of the protein sequences of Cas9 orthologues. The percent query coverage and percent protein sequence identities relative to LmoCas9 are listed in parentheses. Cas9 orthologue names: Francisella novicida (Fno), Listeria monocytogenes (Lmo), Streptococcus pyogenes (Spy), Staphylococcus aureus (Sau), Listeria ivanovii (Liv), Neisseria meningitidis (Nme), Haemophilus parainfluenzae (Hpa), Brackiella oedipodis (Boe), Geobacillus stearothermophilus (Geo), Campylobacter jejuni (Cje), Corynebacterium diphtheriae (Cdi). (B) Plaquing assays where the Listeria phage ΦP35 is titrated in ten-fold serial dilutions (black spots) on lawns of L. monocytogenes Mack (gray background) strains that express chromosomally-integrated LivCas9/tracrRNA and contain pLEB plasmids expressing two components: an Acr or no Acr (−) and a crRNA that targets phage DNA or a scrambled crRNA (non-targeting control). (C) Plaquing assays where the E. coli phage Mu is titrated in ten-fold serial dilutions (black spots) on lawns of E. coli (gray background) expressing the indicated anti-CRISPR proteins and Type II-A, II-B and II-C Cas9-sgRNA programmed to target phage DNA. Representative pictures of at least 3 biological replicates are shown. (D) Gene editing activities of SpyCas9 and CjeCas9 in human cells in the presence of AcrIIA1 variants and orthologues. Control inhibitors (references in Methods): AcrIIA4 selective inhibitor of SpyCas9; AcrIIA5 broad-spectrum Cas9 inhibitor; AcrVA1 Cas12 inhibitor (negative control for Cas9 orthologues). Editing assessed by targeted sequencing. NT indicates a no-sgRNA control condition. Error bars indicate SEM for three independent biological replicates. See Figure S5C for editing experiments with additional Cas9 orthologues.