Molecular basis of dual anti-CRISPR and auto-regulatory functions of AcrIF24

Abstract

CRISPR-Cas systems are adaptive immune systems in bacteria and archaea that provide resistance against phages and other mobile genetic elements. To fight against CRISPR-Cas systems, phages and archaeal viruses encode anti-CRISPR (Acr) proteins that inhibit CRISPR-Cas systems. The expression of acr genes is controlled by anti-CRISPR-associated (Aca) proteins encoded within acr-aca operons. AcrIF24 is a recently identified Acr that inhibits the type I-F CRISPR-Cas system. Interestingly, AcrIF24 was predicted to be a dual-function Acr and Aca. Here, we elucidated the crystal structure of AcrIF24 from Pseudomonas aeruginosa and identified its operator sequence within the regulated acr-aca operon promoter. The structure of AcrIF24 has a novel domain composition, with wing, head and body domains. The body domain is responsible for recognition of promoter DNA for Aca regulatory activity. We also revealed that AcrIF24 directly bound to type I-F Cascade, specifically to Cas7 via its head domain as part of its Acr mechanism. Our results provide new molecular insights into the mechanism of a dual functional Acr-Aca protein.

Figure 1.

The crystal structure of AcrIF24 reveals a novel three-domain arrangement. (A) Size-exclusion chromatography (SEC) profile of AcrIF24. An SDS-PAGE gel loaded with the peak fractions is provided to the right of the main peak. The corresponding fractions from SEC loaded onto SDS-PAGE are indicated by black arrows. (B) The overall structure of AcrIF24 has three distinct domains (wing, head and body). (C) Cartoon representation of AcrIF24. The color of the chain from the N- to the C-termini gradually moves through the spectrum from blue to red. The nine α-helices and nine β-sheets are labeled α₁–α₉ and β₁–β₉, respectively. Topology representation of AcrIF24 is provided in the inset. (D) Surface electrostatic potential of AcrIF24. The respective surface electrostatic distributions are represented by a scale ranging from −7.0 kT/e (red) to 7.0 kT/e (blue). (E) B-factor distribution in the structure of AcrIF24. The structure is presented in a putty representation. Rainbow colors from red to violet with increasing B-factor values were used for B-factor visualization. The region with the highest B-factor, corresponding to most of the head domain, is indicated by a black circle. (F) Structural superimposition of Aca1 with AcrIF24. The superposed reqion is indicated by a black circle. The magenta cartoon model represents the structure of Aca1.

Figure 2.

Dimeric structure of AcrIF24. (A) Multi-angle light scattering (MALS) profile of AcrIF24. The experimental MALS data (red line) are plotted as SEC elution volume (x-axis) versus absolute molecular mass (y-axis) distributions on the SEC chromatogram (black) at 280 nm. (B) Crystallographic packing symmetry analysis. A single molecule found in the asymmetric unit is indicated by the purple-colored ribbon structure and the two symmetry molecules found by packing analysis are indicated by green-colored (Sym1) and magenta-colored (Sym2) ribbon structures. (C) Table summarizing the interaction details of the two types of putative interfaces analyzed by the PISA server. (D) Putative dimeric structure of AcrIF24 generated and analyzed by crystal packing and PISA server. The regions of PPI magnified and presented in (E and F) are indicated by a black-dashed square for region 1 (body domain-mediated PPI) and red-dashed square for region 2 (head domain-mediated PPI). (E and F) Close-up views of two different PPIs, region 1 (E) and region 2 (F), in the dimeric structure of AcrIF24. The red-dashed and black-dashed lines indicate salt bridges and hydrogen bonds, respectively. (G) Validation of the PPI via mutagenesis. SEC-MALS profiles comparing the position of eluted peaks of various mutants with wildtype. The red line indicates the experimental molecular mass measured by MALS. (H) Table summarizing the result of SEC-MALS. Mut and MW indicate mutant and molecular weight, respectively. Fitting error indicates the MALS fitting error.

Figure 3.

AcrIF24 possesses Aca function by binding an IR and repressing acrIF23-acrIF24 promoter expression. (A) Cartoon representation of dimeric AcrIF24 colored according to the degree of amino-acid sequence conservation across different species as analyzed by the ConSurf server. (B) Sequence alignment of AcrIF24 between different species. Mostly conserved and partially conserved residues are colored in red and blue, respectively. The location of three helices α₆−α₈ containing the putative HTH motif are shown above the corresponding sequence. * indicates conserved residues that were involved in the DNA recognition. ^# indicates mutated residues in the wing domain that were predicted as DNA binding residues. (C) Genomic context of the P. aeruginosa acrIF23-acrIF24 operon, with the predicted promoter enlarged. Regulatory elements (–35 and –10 regions) are highlighted in yellow, the inverted repeat is shown by blue arrows. The IR-S and IR-L probes used for EMSAs in (E) and (F) are indicated underneath. (D) Activity of an acrIF23-acrIF24 promoter reporter in P. carotovorum in the presence and absence of AcrIF24 (wildtype and the indicated variants), determined as the median eYFP fluorescence by flow cytometry. Data are presented as the mean ± standard error, with individual replicates represented by dots; statistical significance was assessed by a Kruskal-Wallis test and Dunn's multiple comparisons test against the -AcrIF24 control (ns P≥ 0.05, * P< 0.05, *** P< 0.001). (E) Representative EMSA with AcrIF24 using IR-L and IR-S as a substrate. Purified AcrIF24 at the indicated concentrations was mixed with substrate DNA. Non-denaturing acrylamide gels stained with SYBR Gold are shown. (F) EMSA with dimer-disrupted mutant of AcrIF24 (Y128K/Y217W) using IR-L as a substrate. Non-denaturing acrylamide gels stained with SYBR Gold are shown (G) EMSA with AcrIF24 using half-site IR mutants of IR-L and IR-S as a substrate. Substituted bases are indicated in blue. Non-denaturing acrylamide gels stained with SYBR Gold are shown.

Figure 4.

The HTH body domain of AcrIF24 is involved in DNA binding. (A) Predicted DNA binding residues on AcrIF24. (B) Surface electrostatic features of dimeric AcrIF24. The scale bar ranges from −7.0 kT/e (red) to 7.0 kT/e (blue). (C) EMSA with the indicated AcrIF24 mutants using IR-L and IR-S as a substrate. Purified proteins were added at concentration of 1 μM. BSA was used as a negative control. Non-denaturing acrylamide gels stained with SYBR Gold are shown. (D) Native-PAGE with the wildtype and indicated mutants using IR-L as a substrate. DNA binding ability of wildtype and each mutant of AcrIF24 were analyzed by comparison of band-shift by adding DNA substrate (IR-L). (E) EMSA with the indicated AcrIF24 mutants using IR-L in a various different concentration of protein. Non-denaturing acrylamide gels stained with SYBR Gold are shown.

Figure 5.

AcrIF24 directly binds type I-F Cascade via Cas7f1. (A) Schematic showing the composition and interference process by the type I-F Cascade complex. The crRNA processed and bound by Cas6f acts as a scaffold for type I-F Cascade assembly. (B) Interaction analysis between AcrIF24 and Cascade by SEC. SEC profiles produced by AcrIF24 (blue line), Cascade (red line), and the mixture of AcrIF24 and Cascade (black line) are shown. (C) SDS-PAGE gels produced by loading one of the main fractions from Cascade sample with (+AcrIF24) or without (–AcrIF24) providing AcrIF24. The position of each Cascade subunit on the gel is indicated. AcrIF24 co-migrated with Cascade is indicated by a red arrow. (D) Position of mutated residues marked on the dimeric structure of AcrIF24. Mutated residues G22Y (representative wing domain disruption mutant), D105K and W110K (representative head domain disruption mutants), and G189K (representative body domain disruption mutant) were indicated by red color. (E) Interaction analysis of Cascade with wildtype AcrIF24 and with various mutants of AcrIF24 by SEC followed by SDS-PAGE. SDS-PAGE gels produced by loading one of the main fractions from Cascade sample with various mutants of AcrIF24. The position of each Cascade subunit and AcrIF24 on the gel is indicated. The reduced co-purification of AcrIF24 (W110K) with Cascade is denoted on the gel by red dot-box. The tentative position of the ΔHead mutant that did not co-migrate with Cascade is also denoted in the gel by red dot-box. (F) Bar-chart showing the quantified intensity of the co-eluted AcrIF24 and various mutants on SDS-PAGE provided in (E). Data are presented as the mean ± standard deviation of three independent experiments. (G–I) Interaction analysis between AcrIF24 and each Cascade subunit by SEC. SEC profiles produced by AcrIF24 (blue line), each Cascade subunit (red line) of Cas5f1 (G), Cas6f (H) and Cas7f1 (I), and the mixture of AcrIF24 and each Cascade subunit (black line) are shown. SDS-PAGE gel produced by the mixture of AcrIF24 and each Cascade subunit is provided under the SEC profile. Loaded fractions for SDS-PAGE were indicated by a black bar.

Figure 6.

The head domain and dimerization of AcrIF24 are important for its anti-CRISPR activity. (A) EMSA performed with Cascade using target DNA as substrate. Purified wildtype AcrIF24 at the indicated concentrations was pre-mixed with Cascade and Cas2/3 before adding DNA substrate for analysing the anti-CRISPR function of AcrIF24. (B) EMSA for analysing the anti-CRISPR activity of two Cascade-binding disrupted mutants, W110K and ΔHead. (C) EMSA for analysing the anti-CRISPR activity of two promoter-binding disrupted mutants, K197Y and R207W. (D) EMSA for analysing the anti-CRISPR activity of dimer-disrupted mutant, Y128K/Y217W. Non-denaturing acrylamide gels stained with SYBR Gold are shown. Cas2/3 has been added in these assays but no particular role was detected. (E) Phage ZF40 infecting a sensitive (–CRISPR, PCF425) or immune (+CRISPR, PCF835) P. carotovorum host carrying either an empty vector (–AcrIF24, pBAD30) or a derived plasmid (pPF2964, pPF3482, pPF3483, pPF3484 or pPF3487) for production of the indicated AcrIF24 variants. ZF40 was added as spots from a 10-fold serial dilution, indicated by the back triangle.