Bio-Layer Interferometry Analysis of the Target Binding Activity of CRISPR-Cas Effector Complexes

Abstract

CRISPR-Cas systems employ ribonucleoprotein complexes to identify nucleic acid targets with complementarity to bound CRISPR RNAs. Analyses of the high diversification of these effector complexes suggest that they can exhibit a wide spectrum of target requirements and binding affinities. Therefore, streamlined analysis techniques to study the interactions between nucleic acids and proteins are necessary to facilitate the characterization and comparison of CRISPR-Cas effector activities. Bio-layer Interferometry (BLI) is a technique that measures the interference pattern of white light that is reflected from a layer of biomolecules immobilized on the surface of a sensor tip (bio-layers) in real time and in solution. As streptavidin-coated sensors and biotinylated oligonucleotides are commercially available, this method enables straightforward measurements of the interaction of CRISPR-Cas complexes with different targets in a qualitative and quantitative fashion. Here, we present a general method to carry out binding assays with the Type I-Fv complex from Shewanella putrefaciens and the Type I-F complex from Shewanella baltica as model effectors. We report target specificities, dissociation constants and interactions with the Anti-CRISPR protein AcrF7 to highlight possible applications of this technique.

Keywords: CRISPR-Cas; DNA-binding; affinity; bio-layer interferometry; cascade.

FIGURE 1

Schematic representation of Bio-layer Interferometry (BLI). BLI measures the shifts in reflected white light upon changes in the thickness of the bio-layer. The measured shift depends both on the size and the affinity of the interactors. Bio-layers are coated with molecules (e.g., streptavidin) that allow the immobilization of one of the interactors. For this study, 5′-biotin-tagged oligonucleotides were used. The long dsDNA oligonucleotides have complementary arms separated by a 5-thymidine loop to allow for dsDNA hybridization. In addition, they contain a sequence complementary to the crRNA carried by the effector complex (protospacer) and a protospacer-adjacent motif (PAM). After baselining the light signal on buffer (blue), the oligonucleotide is added, binding to the streptavidin and generating a shift (green) that is also baselined by incubating the bio-layer back in buffer (black straight line). After signal stabilization, the label-free ribonucleoprotein complex is added, and the generated shift is recorded until equilibrium is reached (orange). Dissociation of the complex is then measured by transferring the bio-layer back into buffer (black curve).

FIGURE 2

Determination of optimal oligonucleotide concentrations and saturating protein concentrations. (A) The wavelength shift (nm) generated by the addition of 1 μM Type I-Fv Cascade to different concentrations of complementary dsDNA oligonucleotide was recorded for 5 min. At an initial concentration of 100 nM of oligonucleotide, the reaction reached equilibrium (plateau) and produced the strongest signal. (B) Different concentrations of Type I-Fv Cascade were tested for binding to complementary dsDNA (immobilized at a concentration of 100 nM) for 5 min. The saturating concentration was established as the lowest concentration at which the binding curve reached a stable plateau (blue line). Higher concentrations did not further increase the wavelength shift at equilibrium or caused deviations from the expected binding behavior (as depicted above for 3.5 μM by the dotted line).

FIGURE 3

Determination of the equilibrium dissociation constants (K_D) of the interactions of Type I-Fv and I-F CRISPR-Cas complexes with target dsDNA. Wavelength shift (nm) generated by the addition of Type I-Fv Cascade from S. putrefaciens CN-32 (A) or Type I-F from S. baltica OS195 (B) to complementary dsDNA oligonucleotide immobilized on a streptavidin biosensor at a concentration of 100 nM. Binding was followed for 7 min in order to reach equilibrium. Afterward, the protein-bound biosensor was incubated for 3 min in buffer to measure the dissociation reaction. The interaction of the complexes with the dsDNA-free biosensor is shown as control (Complex alone). The wavelength shifts recorded at 420 s after the start of binding were plotted against the corresponding complex concentration in order to calculate the respective K_D values. Data were fitted to the non-linear equation: Y = B_max*X/(K_D + X) + NS*X, where B_max is the maximum wavelength shift and NS the slope of the non-linear component, representing non-specific binding. The coefficients of determination (R²) and K_D values obtained are shown in the graphs for each complex. Binding assays were performed in duplicate. Error bars indicate the Standard Error of the Mean (SEM).

FIGURE 4

Qualitative analysis of PAM preference by Type I-Fv Cascade. Wavelength shift (nm) generated by the addition of 1.5 μM of Type I-Fv Cascade from S. putrefaciens CN-32 to dsDNA oligonucleotides containing a spacer matching to the crRNA and either a GG or a TT PAM. The GG PAM is recognized by the complex, while the TT PAM is not recognized, giving shifts similar to that procured by a non-complementary target. Assays were performed in duplicate. The lighter outlines represent the SEM.

FIGURE 5

Influence of dsDNA mismatches on Type I-Fv Cascade binding. (A) Schematic structures of dsDNA oligonucleotides used in the binding assay. Targets are either fully hybridized or contain 5 nt mismatches at positions in the protospacer close to the PAM, in the centre or at the PAM-distal end. (B) Wavelength shift (nm) generated by the addition of 1.5 μM Type I-Fv Cascade from S. putrefaciens CN-32 to the indicated dsDNA oligonucleotides. Assays were performed in duplicate. The lighter outlines represent the SEM.

FIGURE 6

Effect of Anti-CRISPR (Acr) proteins on the binding of CRISPR-Cas complexes to dsDNA. (A) Efficiency of Transformation Assays (EOTs) of CRISPR-Cas systems I-Fv from S. putrefaciens CN-32 (left) and I-F from S. baltica OS195 (right) expressed in E. coli BL21-AI. The activity of the effector complexes was tested when expressed alone (WT Cascade) or co-expressed with AcrF7 from Pseudomonas aeruginosa (Pawluk et al., 2016). EOT equals to the colony ratio between the strain of interest and its corresponding Cas2/3 HD mutant strain, presented as percentages. Error bars represent the SEM. Three replicates were quantified. (B) Wavelength shift (nm) generated by the binding of 500 nM of either Type I-Fv (left) or Type I-F (right) Cascades to complementary oligonucleotide. In order to test the effect of AcrF7, the protein was mixed with Cascades at a 1:1 molar ratio or at a 5-fold molear excess (1:5). Samples were incubated at room temperature for 10 min before analysis. The wavelength shift elicited by the binding of AcrF7 to the dsDNA in the absence of Casade complexees is shown as a control (AcrF7 to DNA). Assays were performed in duplicate. The lighter outlines represent the SEM. (C) Wavelength shift (nm) induced by the binding of 500 nM of S. baltica I-F Cascade to dsDNA oligonucleotide. The effect of AcrF7 on complexes already bound to dsDNA was tested by incubation of the biosensor in BLI buffer containing 2.5 μM AcrF7 (400 s). Assays were performed in duplicate. The lighter outline represents the SEM.