RNA Targeting by Functionally Orthogonal Type VI-A CRISPR-Cas Enzymes

Abstract

CRISPR adaptive immunity pathways protect prokaryotic cells against foreign nucleic acids using CRISPR RNA (crRNA)-guided nucleases. In type VI-A CRISPR-Cas systems, the signature protein Cas13a (formerly C2c2) contains two separate ribonuclease activities that catalyze crRNA maturation and ssRNA degradation. The Cas13a protein family occurs across different bacterial phyla and varies widely in both protein sequence and corresponding crRNA sequence conservation. Although grouped phylogenetically together, we show that the Cas13a enzyme family comprises two distinct functional groups that recognize orthogonal sets of crRNAs and possess different ssRNA cleavage specificities. These functional distinctions could not be bioinformatically predicted, suggesting more subtle co-evolution of Cas13a enzymes. Additionally, we find that Cas13a pre-crRNA processing is not essential for ssRNA cleavage, although it enhances ssRNA targeting for crRNAs encoded internally within the CRISPR array. We define two Cas13a protein subfamilies that can operate in parallel for RNA detection both in bacteria and for diagnostic applications.

Fig. 1. Pre-crRNA processing is broadly conserved within the Cas13a protein family

(A) Schematic of crRNA processing reaction catalyzed by Cas13a. pre-crRNA substrates are cleaved by Cas13a to generate mature crRNAs. Below, a schematic of a pre-crRNA highlighting important functional features. (B) Alignment of the 5′ portion of CRISPR repeat sequences from the studied type VI CRISPR systems highlighting the pre-crRNA cleavage site. Mapped cut cleavage sites are shown as red bars. Deviations from the Lbu crRNA-repeat sequence are noted in black text. Lowercase g’s were required for transcription purposes and are not part of the native crRNA repeat sequences. Full CRISPR repeat sequence is diagrammed in Fig. S1B. Cleavage sites mapped in previous studies for LbuCas13a, LshCas13a and LseCas13a (East-Seletsky et al., 2016). (C) Representative gel of Cas13a- mediated pre-crRNA cleavage by nine Cas13a homologs after 60 min incubation with 5′-radiolabelled pre-crRNA substrates. Cleavage by LbuCas13a has been previously demonstrated (East-Seletsky et al., 2016) and can be observed in Fig 2.

Fig. 2. Identification of residues important for pre-crRNA cleavage by LbuCas13a

(A) General Cas13a domain organization schematic with annotated based off a LshCas13a crystal structure (Liu et al., 2017). Multiple-sequence amino acid alignment of (B) the local region in Cas13a’s helical 1 domain implicated in pre-crRNA processing by studies on LshCas13a, and (C) the region within in Cas13a’s HEPN2 domain implicated in pre-crRNA processing by studies on LbuCas13a. See Fig. S2 for full family tree alignment of these regions and Data File S1 for a complete protein alignment. Residues whose mutation severely affect pre-crRNA processing are marked by red triangles, minimal impacts on processing by yellow squares and residues whose mutation did not affect pre-crRNA processing marked with teal diamonds. Symbols above the LbuCas13a sequences correspond to mutations made to LbuCas13a, and symbols below the LshCas13a sequence correspond to mutations made to LshCas13a by Liu et al, 2017. Coloration of the matrix alignment denotes residue conservation using the ClustalX scheme, with darker hues indicating the strength of the conservation. pre-crRNA processing under single turnover conditions measured for mutants in (D) the helical 1 domain, and (E) the HEPN2 domain. Quantified data were fitted with single-exponential decays with calculated pseudo-first-order rates constants (k_obs) (mean ± s.d., n =3) as follows: Lbu WT 0.074 ± 0.003 min⁻¹, E299A 0.071 ± 0.005 min⁻¹, K310A 0.071 ± 0.003 min⁻¹, R311A 0.054 ± 0.007 min⁻¹, N314A 0.029 ± 0.008 min⁻¹, R1079A 0.009 ± 0.007 min⁻¹, D1078A 0.023 ± 0.002 min⁻¹, K1080A 0.016 ± 0.004 min⁻¹, and K1087A 0.076 ± 0.007 min⁻¹, while R1072A and K1082A could not be fitted. (F) Representative gel of pre-crRNA processing of LshCas13a, LwaCas13a, and LbuCas13a pre-crRNAs by LbuCas13a, LshCas13a, LwaCas13a and PprCas13a proteins using standard conditions. Hydrolysis ladders for each pre-cRNA substrate demonstrate subtle differences in the migration of these fragments from differing sequences.

Fig. 3. Members of the Cas13a protein family cleave ssRNA with a range of efficiencies

(A) Schematic of ssRNA-targeting by Cas13a. For simplicity, trans-ssRNA cleavage was the focus of study. (B) Representative gel of Cas13a mediated trans-ssRNA cleavage by all ten homologs after 60 min incubation. Cas13a:crRNA complexes were formed as described in the methods using mature crRNA products with a final RNP complex concentration of 50 nM. <1 nM radiolabeled trans-ssRNA target was added to initiate reaction in the presence and absence of 50 nM unlabeled, crRNA-complementary ssRNA activator. Weak trans-ssRNA cleavage activity was observed by LshCas13a with product bands noted by double arrows to the right. (C) Heat map reporting Cas13a-catalyzed trans-ssRNA cleavage percentages for each 5-mer homopolymer ssRNA substrate, for six different Cas13a. Assay conditions were identical to part (B), except LbuCas13a and LwaCas13a which were incubated for 5 min instead of 60 min. (n=3, values with associated errors presented in Table S3). (D) Apparent cleavage rates of a fluorescent ssRNA reporter by five homologs across a range of ssRNA activator concentrations. Cas13a:crRNA complexes were pre-incubated at a 2:1 ratio respectively with a final active complex concentration of 50 nM. Complementary ssRNA activator and fluorescent ssRNA cleavage reporter were added to initiate reactions. Normalized reporter signal curves timecourses were fitted with single-exponential decays and the apparent rates are plotted (n =3). Some conditions plateaued before first measured time-point therefore their rates are minimally assumed to be 0.5 min⁻¹ and are labeled with a * in the chart.

Fig. 4. crRNA exchangeability within the Cas13a family

(A) Maximum-likelihood phylogenetic tree of all Cas13a family members. Homologs used in this study are bolded and clades are highlighted. Bootstrapped values are located in previous work (East-Seletsky et al. 2016). (B) Symmetrical similarity score matrix for CRISPR repeats from homologs used in this study. Rows and columns are ordered by CRISPR repeat clustering. (C) Asymmetrical similarity score matrix for CRISPR repeats from homologs used in this study. The same pairwise scores are presented here as in (B), except the rows are reordered to correspond to the Cas13a phylogenetic tree of the subset of homologs used in this study. Bootstrap values for this smaller tree are in Fig S1A. (D–F) Functional activity matrix for (D) pre-crRNA processing by non-cognate proteins, (E) trans-ssRNA cleavage directed by pre-crRNAs, and (F) trans-ssRNA cleavage directed by mature crRNAs. Processing assays were performed using standard conditions and 60 min reaction endpoints were analyzed. trans-ssRNA cleavage assays were performed using the fluorescent ssRNA reporter assay with fitted initial rates. ssRNA activator concentrations were as follows: LbuCas13a:100 pM, LwaCas13a:100 pM, PprCas13a:100 nM, LbaCas13a:10 nM, and HheCas13a:100 nM. Initial rates were fit across three replicates to account for differences in fluorescence plateau values and normalized to each Cas13a:crRNA cognate pair. See Tables S4–6 for numerical values and associated errors (n=3).

Fig. 5. Functional validation of orthogonal Cas13a subfamilies for RNA detection

(A) Schematic of the RNA detection assay modified to use fluorescent homopolymer ssRNA reporter substrates to assay trans-ssRNA cleavage activation by either LbuCas13a or LbaCas13a. (B) Timecourse of raw fluorescence measurements generated by homopolymer reporters incubated with either LbuCas13a: Lbu-crRNA: 10 pM ssRNA activator or LbaCas13a: Lba-crRNA: 1 nM ssRNA activator (mean ± s.d., n=3). (C) Raw fluorescence measurements generated by the fluorescent homopolymer ssRNA reporters across a panel of crRNA, ssRNA activator, and Cas13a protein combinations (mean ± s.d., n=3).

Fig. 6. Deciphering the role of crRNA array processing for LbuCas13a

(A) Quantified timecourse data of pre-crRNA processing assays for R1079A/K1080A mutant compared to wildtype LbuCas13a. Quantified data was fitted to single- exponential decays and pseudo-first-order rate constant (k_obs) (mean ± s.d., n =3) for LbuCas13a WT of 0.074 ± 0.003 min⁻¹, while the R1079A/K1080A mutant could not be fit with sufficient confidence to yield a rate constant. (B) Apparent rate of fluorescent reporter by LbuCas13a wildtype and R1079A/K1080A processing inactive mutant as directed by pre-crRNA and mature crRNAs. Cas13a:RNA complexes were pre-incubated for 60 min at a 1:1 ratio, and then 10 pM of activator and 150 nM reporter were added to initiate reaction. (mean ± s.d., n =3) (C) Apparent rates of fluorescent ssRNA reporter cleavage by 300 nM wildtype LbuCas13a or R1079A/K1080A pre-crRNA processing inactive mutant as directed by 50 nM of a CRISPR array containing six crRNA repeat-spacers. Each bar group represents the addition of 100 pM of a distinct ssRNA activator sequence complementary to schematized positions within the CRISPR array indicated below each bar group. Each rate is fitted from data from three biological replicates and the standard deviation of the rate is depicted. Mutant protein rate is statistically different from the wildtype LbuCas13a for all spacer positions (one-sided t-test: p<0.001). (D) Data from (C) depicted as a percentage of wildtype LbuCAs13a activity demonstrating the positional effect of the decreased trans-ssRNA targeting efficiencies by the pre-crRNA processing inactive mutant. Processing of the array is shown in Fig. S5.

Fig. 7. A revised model for Type VI-A CRISPR-Cas system

(A) Graphical summary of key findings in this study. Homologs used in this study are indicated with abbreviations in bold, with trans-ssRNA cleavage inactive homologs depicted in grey. Colored circles highlight the two orthogonal Cas13a enzyme groups, as defined by their generalized crRNA exchangeability and trans-ssRNA cleavage substrate nucleotide preference. (B) Revised model of Type VI-A CRISPR-Cas systems incorporating the interplay between crRNA maturation and RNA targeting efficiency.