
Editor's Introduction
RNA editing with CRISPR-Cas13
Once a disease-causing mutation has been transcribed, it can be difficult to stop. Gene editing techniques that change the DNA sequence don't affect mRNA transcripts that have been released into the cell. A new type of gene editing technique fixes this problem by targeting messenger RNA. These new tools are programmable and can be directed to edit mRNA molecules. But several issues remain: How can we make sure the editor is changing the right base? How can we avoid breaking the molecule when we edit it? And how can we make sure the editors aren't also changing other molecules and causing unintended problems?
Paper Details
Abstract
Nucleic acid editing holds promise for treating genetic disease, particularly at the RNA level, where disease-relevant sequences can be rescued to yield functional protein products. Type VI CRISPR-Cas systems contain the programmable single-effector RNA-guided ribonuclease Cas13. We profiled type VI systems in order to engineer a Cas13 ortholog capable of robust knockdown and demonstrated RNA editing by using catalytically inactive Cas13 (dCas13) to direct adenosine-to-inosine deaminase activity by ADAR2 (adenosine deaminase acting on RNA type 2) to transcripts in mammalian cells. This system, referred to as RNA Editing for Programmable A to I Replacement (REPAIR), which has no strict sequence constraints, can be used to edit full-length transcripts containing pathogenic mutations. We further engineered this system to create a high-specificity variant and minimized the system to facilitate viral delivery. REPAIR presents a promising RNA-editing platform with broad applicability for research, therapeutics, and biotechnology.
Report
Precise nucleic acid–editing technologies are valuable for studying cellular function and as novel therapeutics. Current editing tools, based on programmable nucleases such as the prokaryotic CRISPR-associated nucleases Cas9 (1–4) or Cpf1 (5), have been widely adopted for mediating targeted DNA cleavage, which in turn drives targeted gene disruption through nonhomologous end joining (NHEJ) or precise gene editing through template-dependent homology-directed repair (HDR) (6). NHEJ uses host machineries that are active in both dividing and post-mitotic cells and provides efficient gene disruption by generating a mixture of insertion or deletion (indel) mutations that can lead to frame shifts in protein-coding genes. HDR, in contrast, is mediated by host machineries whose expression is largely limited to replicating cells. Accordingly, the development of gene-editing capabilities for post-mitotic cells remains a major challenge. DNA base editors, consisting of a fusion between Cas9 nickase and cytidine deaminase, can mediate efficient cytidine-to-uridine conversions within a target window and substantially reduce the formation of double-strand break–induced indels (7, 8). However, the potential targeting sites of DNA base editors are limited by the requirement of Cas9 for a protospacer adjacent motif (PAM) at the editing site (9). Here, we describe the development of a precise and flexible RNA base editing technology using the type VI CRISPR-associated RNA-guided ribonuclease (RNase) Cas13 (10–13).
Cas13 enzymes have two higher eukaryotes and prokaryotes nucleotide-binding (HEPN) endoRNase domains that mediate precise RNA cleavage with a preference for targets with protospacer flanking sites (PFSs) observed biochemically and in bacteria (10, 11). Three Cas13 protein families have been identified to date: Cas13a (previously known as C2c2), Cas13b, and Cas13c (12, 13). We recently reported that Cas13a enzymes can be adapted as tools for nucleic acid detection (14) as well as mammalian and plant cell RNA knockdown and transcript tracking (15), and observed that the biochemical PFS was not required for RNA interference with Cas13a (15). The programmable nature of Cas13 enzymes makes them an attractive starting point to develop tools for RNA binding and perturbation applications.
The adenosine deaminase acting on RNA (ADAR) family of enzymes mediates endogenous editing of transcripts via hydrolytic deamination of adenosine to inosine, a nucleobase that is functionally equivalent to guanosine in translation and splicing (16, 17). There are two functional human ADAR orthologs, ADAR1 and ADAR2, which consist of N-terminal double-stranded RNA–binding domains and a C-terminal catalytic deamination domain. Endogenous target sites of ADAR1 and ADAR2 contain substantial double-stranded identity, and the catalytic domains require duplexed regions for efficient editing in vitro and in vivo (18, 19). The ADAR catalytic domain is capable of deaminating target adenosines without any protein cofactors in vitro (20). ADAR1 has been found to target mainly repetitive regions, whereas ADAR2 mainly targets nonrepetitive coding regions (17). Although ADAR proteins have preferred motifs for editing that could restrict the potential flexibility of targeting, hyperactive mutants, such as ADAR2(E488Q) (21), relax sequence constraints and increase adenosine-to-inosine editing rates. (Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr. In the mutants, other amino acids were substituted at certain locations; for example, E488Q indicates that glutamic acid at position 488 was replaced by glutamine.) ADARs preferentially deaminate adenosines mispaired with cytidine bases in RNA duplexes (22), providing a promising opportunity for precise base editing. Although previous approaches have engineered targeted ADAR fusions via RNA guides (23–26), the specificity of these approaches has not been reported, and their respective targeting mechanisms rely on RNA-RNA hybridization without the assistance of protein partners that may enhance target recognition and stringency.
We assayed a subset of the family of Cas13 enzymes for RNA knockdown activity in mammalian cells and identified the Cas13b ortholog from Prevotella sp. P5-125 (PspCas13b) as the most efficient and specific for mammalian cell applications. We then fused the ADAR2 deaminase domain with the E488Q mutation (ADAR2DD) to catalytically inactive PspCas13b and demonstrated RNA editing for programmable A to I (G) replacement (REPAIR) of reporter and endogenous transcripts as well as disease-relevant mutations. Last, we used a rational mutagenesis scheme to improve the specificity of dCas13b-ADAR2DD fusions in order to generate REPAIRv2 with more than 919-fold higher specificity.
Comprehensive characterization of Cas13 family members in mammalian cells
We previously developed Leptotrichia wadei Cas13a (LwaCas13a) for mammalian knockdown applications, but it required a monomeric superfolder green fluorescent protein (msfGFP) stabilization domain for efficient knockdown, and although the specificity was high, knockdown levels were not consistently below 50% (15). We sought to identify a more robust RNA-targeting CRISPR system by characterizing a genetically diverse set of Cas13 family members in order to assess their RNA knockdown activity in mammalian cells (Fig. 1A). We generated mammalian codon-optimized versions of multiple Cas13 proteins, including 21 orthologs of Cas13a, 15 of Cas13b, and seven of Cas13c, and cloned them into an expression vector with N- and C-terminal nuclear localization signal (NLS) sequences and a C-terminal msfGFP to enhance protein stability (table S1). To assay interference in mammalian cells, we designed a dual-reporter construct expressing the independent Gaussia (Gluc) and Cypridina (Cluc) luciferases under separate promoters, allowing one luciferase to function as a measure of Cas13 interference activity and the other to serve as an internal control. For each Cas13 ortholog, we designed PFS-compatible guide RNAs, using the Cas13b PFS motifs derived from an ampicillin interference assay (fig. S1, table S2, and supplementary text) and the 3′ H (not G) PFS from previous reports of Cas13a activity (10).
Fig. 1. Characterization of a highly active Cas13b ortholog for RNA knockdown. (A) Schematic of stereotypical Cas13 loci and corresponding CRISPR RNA (crRNA) structure. (B) Evaluation of 19 Cas13a, 15 Cas13b, and seven Cas13c orthologs for luciferase knockdown by use of two different guides. Orthologs with efficient knockdown by using both guides are labeled with their host organism name. Values are normalized to a nontargeting guide designed against the Escherichia coli LacZ transcript, with no homology to the human transcriptome. (C) PspCas13b and LwaCas13a knockdown activity (as measured by luciferase activity) by using tiling guides against Gluc. Values represent mean ± SEM. Nontargeting guide is the same as in (B). (D) PspCas13b and LwaCas13a knockdown activity (as measured by luciferase activity) by using tiling guides against Cluc. Values represent mean ± SEM. Nontargeting guide is the same as in (B). (E) Expression levels in log2[transcripts per million (TPM)+1] values of all genes detected in RNA-seq libraries of nontargeting control (x axis) compared with Gluc-targeting condition (y axis) for LwaCas13a (red) and shRNA (black). Shown is the mean of three biological replicates. The Gluc transcript data point is labeled. Nontargeting guide is the same as in (B). (F) Expression levels in log2[transcripts per million (TPM)+1] values of all genes detected in RNA-seq libraries of nontargeting control (x axis) compared with Gluc-targeting condition (yaxis) for PspCas13b (blue) and shRNA (black). Shown is the mean of three biological replicates. The Gluc transcript data point is labeled. Nontargeting guide is the same as in (B). (G) Number of significant off-targets from Gluc knockdown for LwaCas13a, PspCas13b, and shRNA from the transcriptome-wide analysis in (E) and (F).
Question
How did the authors choose the ortholog of Cas13 which was the most effective in RNA knockdown?
Panel B
Each dot represents the Gluc luciferase level in cells with different Cas13 orthologs (represented by different colors). The knockdown level is plotted for two gRNAs used with each ortholog. The luciferase signal was normalized to the signal in cells transfected with irrelevant gRNA against a bacterial gene absent from mammalian cells.
Panels C & D
These plots compare the knockdown efficiency of PspCas13b and LwaCas13a against the Gluc transcript. The x-axis shows where the first nucleotide of the gRNA target site is positioned along the Gluc transcript. Each gRNA is shifted by several nucleotides relative to the previous gRNA so that together they can cover the whole RNA molecule. Gluc is a protein of around 19 kDA, while Cluc is around 60 kDA, so note that the scale of the x-axes in C and D is different. The x-axis represent the knockdown level, as in B.
Panels E & F
These plots show expression levels of all detected cellular transcripts. The majority of them fall into a line at an angle of around 45° to the x-axis, which means that their expression is the same in the cells with either a targeting gRNA or an irrelevant control gRNA.
Only two dots corresponding with the Gluc transcript are far beyond the line, which means their expression is much lower in the samples that got the targeting gRNA (as opposed to the control gRNA).
Conclusion
The researchers accumulated results from a series of knockdown experiments targeting both the exogenous Gluc transcript and the endogenous KRAS transcript, as well as from off-target analysis. Based on these data, the PspCas13b ortholog demonstrated the best result in the knockdown of Gluc transcripts and produced minimal off-target effects. Moreover, the Cas13-mediated knockdown was much more efficient and specific than the shRNA-mediated knockdown (Panel G).
We transfected human embryonic kidney (HEK) 293FT cells with Cas13-expression, guide RNA, and reporter plasmids and then quantified levels of Cas13 expression and the targeted Gluc 48 hours later (Fig. 1B and fig. S2A). Testing two guide RNAs for each Cas13 ortholog revealed a range of activity levels, including five Cas13b orthologs with similar or increased interference across both guide RNAs relative to the recently characterized LwaCas13a (Fig. 1B), and we observed only a weak correlation between Cas13 expression and interference activity (fig. S2, B to D). We selected the top five Cas13b orthologs and the top two Cas13a orthologs for further engineering.
We next tested Cas13-mediated knockdown of Gluc without msfGFP to select orthologs that do not require stabilization domains for robust activity. We hypothesized that Cas13 activity could be affected by subcellular localization, as we previously reported for optimization of LwaCas13a (15). Therefore, we tested the interference activity of the seven selected Cas13 orthologs C-terminally fused to one of six different localization tags without msfGFP. Using the luciferase reporter assay, we identified the top three Cas13b designs with the highest level of interference activity: Cas13b from Prevotella sp. P5-125 (PspCas13b) and Cas13b from Porphyromonas gulae (PguCas13b) C-terminally fused to the HIV Rev nuclear export sequence (NES), and Cas13b from Riemerella anatipestifer (RanCas13b) C-terminally fused to the mitogen-activated protein kinase NES (fig. S3A). To further distinguish activity levels of the top orthologs, we compared the three optimized Cas13b constructs with the optimal LwaCas13a-msfGFP fusion and to short hairpin–mediated RNA (shRNA) for their ability to knock down the endogenous KRAS (V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog) transcript by using position-matched guides (fig. S3B). We observed the highest levels of interference for PspCas13b (average knockdown, 62.9%) and thus selected this for further comparison with LwaCas13a.
To more rigorously define the activity of PspCas13b and LwaCas13a, we designed position-matched guides tiling along both Gluc and Cluc transcripts and assayed their activity using our luciferase reporter assay. We tested 93 and 20 position-matched guides targeting Gluc and Cluc, respectively, and found that PspCas13b had consistently increased levels of knockdown relative to LwaCas13a (average of 92.3% for PspCas13b versus 40.1% knockdown for LwaCas13a) (Fig. 1, C and D).
Specificity of Cas13 mammalian interference activity
To characterize the interference specificities of PspCas13b and LwaCas13a, we designed a plasmid library of luciferase targets containing single mismatches and double mismatches throughout the target sequence and the three flanking 5′ and 3′ base pairs (fig. S3C). We transfected HEK293FT cells with either LwaCas13a or PspCas13b, a fixed guide RNA targeting the unmodified target sequence, and the mismatched target library corresponding to the appropriate system. We then performed targeted RNA sequencing (RNA-seq) of uncleaved transcripts in order to quantify depletion of mismatched target sequences. We found that LwaCas13a and PspCas13b had a central region that was relatively intolerant to single mismatches, extending from base pairs 12 to 26 for the PspCas13b target and 13 to 24 for the LwaCas13a target (fig. S3D). Double mismatches were even less tolerated than single mutations, with little knockdown activity observed over a larger window, extending from base pairs 12 to 29 for PspCas13b and 8 to 27 for LwaCas13a in their respective targets (fig. S3E). Additionally, because there are mismatches included in the three nucleotides flanking the 5′ and 3′ ends of the target sequence, we could assess PFS constraints on Cas13 knockdown activity. Sequencing showed that almost all PFS combinations allowed robust knockdown, indicating that a PFS constraint for interference in mammalian cells likely does not exist for either enzyme tested. These results indicate that Cas13a and Cas13b display similar sequence constraints and sensitivities against mismatches.
We next characterized the interference specificity of PspCas13b and LwaCas13a across the mRNA fraction of the transcriptome. We performed transcriptome-wide mRNA sequencing to detect significant differentially expressed genes. LwaCas13a and PspCas13b demonstrated robust knockdown of Gluc (Fig. 1, E and F) and were highly specific compared with a position-matched shRNA, which showed hundreds of off-targets (Fig. 1G), a finding consistent with our previous characterization of LwaCas13a specificity in mammalian cells (15).
Cas13-ADAR fusions enable targeted RNA editing
Given that PspCas13b achieved consistent, robust, and specific knockdown of mRNA in mammalian cells, we envisioned that it could be adapted as an RNA-binding platform to recruit RNA-modifying domains, such as ADARDD, for programmable RNA editing. To engineer a PspCas13b lacking nuclease activity (dPspCas13b, referred to as dCas13b hereafter), we mutated conserved catalytic residues in the HEPN domains and observed loss of luciferase RNA knockdown (fig. S4A). We hypothesized that a dCas13b-ADARDD fusion could be recruited by a guide RNA to target adenosines, with the hybridized RNA creating the required duplex substrate for ADAR activity (Fig. 2A). To enhance target adenosine deamination rates, we introduced two additional modifications to our initial RNA editing design: We introduced a mismatched cytidine opposite the target adenosine, which has been previously reported to increase deamination frequency, and fused dCas13b with the deaminase domains of human ADAR1 or ADAR2 containing hyperactivating mutations in order to enhance catalytic activity [ADAR1DD(E1008Q) (27) or ADAR2DD(E488Q)] (21).
Fig. 2. Engineering dCas13b-ADAR fusions for RNA editing. (A) Schematic of RNA editing by dCas13b-ADARDD fusion proteins. dCas13b is fused to ADARDD, which naturally deaminates adenosines to insosines in dsRNA. The crRNA specifies the target site by hybridizing to the bases surrounding the target adenosine, creating a dsRNA structure for editing and recruiting the dCas13b-ADARDD fusion. A mismatched cytidine in the crRNA opposite the target adenosine enhances the editing reaction, promoting target adenosine deamination to inosine, a base that functionally mimics guanosine in many cellular reactions. (B) Schematic of Cypridina luciferase W85X target and targeting guide design. Deamination of the target adenosine restores the stop codon to the wild-type tryptophan. Spacer length is the region of the guide that contains homology to the target sequence. Mismatch distance is the number of bases between the 3′ end of the spacer and the mismatched cytidine. The cytidine mismatched base is included as part of the mismatch distance calculation. (C) Quantification of luciferase activity restoration for (left) dCas13b-ADAR1DD(E1008Q) and (right) dCas13b-ADAR2DD(E488Q) with tiling guides of length 30, 50, 70, or 84 nt. All guides with even mismatch distances are tested for each guide length. Values are background-subtracted relative to a 30-nt nontargeting guide that is randomized with no sequence homology to the human transcriptome. (D) Schematic of the sequencing window in which A-to-I edits were assessed for Cypridina luciferase W85X. (E) Sequencing quantification of A-to-I editing for 50-nt guides targeting Cypridina luciferase W85X. Blue triangle indicates the targeted adenosine. For each guide, the region of duplex RNA is outlined in red. Values represent mean ± SEM. Nontargeting guide is the same as in (C).
Problem
How to turn the nuclease Cas13b into an RNA base-editor.
Panel C
These plots show how Cluc luciferase signals are restored in the cells that underwent RNA editing. Each dot represents a signal in the cells that got a gRNA of a certain length and with a certain mismatch distance. Absence of a signal in the uppermost left plot means that short gRNAs of the length 30-nt could not direct dCas13b-ADAR1 complex to the target, so no editing occurred.
Panels D & E
The researchers used sequencing to detect off-target editing events and to calculate their frequency. The column of the brightest red squares depicts that the largest proportion of editing events occurred in the target site with almost all gRNA molecules. However, some rows which represent individual gRNA molecules contain two and even three bright squares. Notice that the majority of the edits is located inside the red outline. This corresponds to part of the RNA transcript that forms a duplex with the gRNA. As was mentioned earlier, ADAR prefers to edit bases in the duplex RNA.
To test the activity of dCas13b-ADARDD, we generated an RNA-editing reporter on Cluc by introducing a nonsense mutation [W85X (UGG→UAG)], which could functionally be repaired to the wild-type codon through A→I editing (Fig. 2B) and then be detected as restoration of Cluc luminescence. We evenly tiled guides with spacers 30, 50, 70, or 84 nucleotides (nt) long across the target adenosine so as to determine the optimal guide placement and design (Fig. 2C). We found that dCas13b-ADAR1DD(E1008Q) required longer guides to repair the Cluc reporter, whereas dCas13b-ADAR2DD(E488Q) was functional with all guide lengths tested (Fig. 2C). We also found that the hyperactive E488Q mutation improved editing efficiency as wild-type ADAR2DD displayed reduced luciferase restoration (fig. S4B). From this demonstration of activity, we chose dCas13b-ADAR2DD(E488Q) for further characterization and designated this system RNA Editing for Programmable A to I Replacement version 1 (REPAIRv1).
To validate that restoration of luciferase activity was due to bona fide editing events, we directly measured REPAIRv1-mediated editing of Cluc transcripts via reverse transcription and targeted next-generation sequencing. We tested 30- and 50-nt spacers around the target site and found that both guide lengths resulted in the expected A-to-I edit, with 50-nt spacers achieving higher editing percentages (Fig. 2, D and E, and fig. S4C). We also observed that 50-nt spacers had an increased propensity for editing at nontargeted adenosines within the sequencing window, likely because of increased regions of duplexed RNA (Fig. 2E and fig. S4C).
We next targeted an endogenous gene, PPIB. We designed 50-nt spacers tiling PPIB and found that we could edit the PPIB transcript with up to 28% editing efficiency (fig. S4D). To test whether REPAIR could be further optimized, we modified the linker between dCas13b and ADAR2DD(E488Q) (fig. S4E and table S3) and found that linker choice modestly affected luciferase activity restoration. Additionally, we tested the ability of dCas13b and guide alone to mediate editing events, finding that the ADARDD is required for editing (fig. S5, A to D).
Defining the sequence parameters for RNA editing
Given that we could achieve precise RNA editing at a test site, we wanted to characterize the sequence constraints for programming the system against any RNA target in the transcriptome. Sequence constraints could arise from dCas13b-targeting limitations, such as the PFS, or from ADAR sequence preferences (28). To investigate PFS constraints on REPAIRv1, we designed a plasmid library that carryies a series of four randomized nucleotides at the 5′ end of a target site on the Cluc transcript (Fig. 3A). We targeted the center adenosine within either a UAG or AAC motif and found that for both motifs, all PFSs demonstrated detectable levels of RNA editing, with a majority of the PFSs having >50% editing at the target site (Fig. 3B). Next, we sought to determine whether the ADAR2DD in REPAIRv1 had any sequence constraints immediately flanking the targeted base, as has been reported previously for ADAR2DD (28). We tested every possible combination of 5′ and 3′ flanking nucleotides directly surrounding the target adenosine (Fig. 3C) and found that REPAIRv1 was capable of editing all motifs (Fig. 3D). Last, we analyzed whether the identity of the base opposite the target A in the spacer sequence affected editing efficiency and found that an A-C mismatch had the highest luciferase restoration, in agreement with previous reports of ADAR2 activity, with A-G, A-U, and A-A having drastically reduced REPAIRv1 activity (fig. S5E).
Fig. 3. Measuring sequence flexibility for RNA editing by REPAIRv1. (A) Schematic of screen for determining PFS preferences of RNA editing by REPAIRv1. A randomized PFS sequence is cloned 5′ to a target site for REPAIR editing. After exposure to REPAIR, deep sequencing of reverse-transcribed RNA from the target site and PFS is used to associate edited reads with PFS sequences. (B) Distributions of RNA editing efficiencies for all 4-N PFS combinations at two different editing sites. (C) Quantification of the percent editing of REPAIRv1 at Cluc W85 across all possible three-base motifs. Values represent mean ± SEM. Nontargeting guide is the same as in Fig. 2C. (D) Heatmap of 5′ and 3′ base preferences of RNA editing at Cluc W85 for all possible three-base motifs.
Question
What sequences are most effectively targeted by the REPAIRv1 system?
Panel B
The efficiency of the REPAIRv1 editing at UAG and AAC motifs is demonstrated by two violin plots. The y-axis shows the proportion of molecules edited, whereas the x-axis represents how common a particular editing rate is. The wider the shape, the more common a particular result is. Both target motifs mediate editing at a rate of more than 50% (the widest parts of the plot are above 50%). However, targeting at UAG is more efficient, since the majority of the data is between 75% and 100%, whereas the AAC motif allows only around 65% rate.
This is a cumulative result for all randomized PFS sequences for UAG or AAC sites. From these data we cannot conclude what individual sequences are the most suitable for targeting. But we can compare the two site targets for their ability to promote editing.
Panel C
The figure shows how the editing rate of the target adenosine is influenced by the two flanking nucleotides. Note the difference in the editing rate between this histogram and the plot in B. Here the authors picked just one targeting RNA molecule for each triplet combination.
We can see that the majority of combinations mediate targeted editing, while the UAC triplet also shows a high level of untargeted editing.
Panel D
This heatmap is a practical summary of the data from C. The y-axis represents the 5' (left) flanking nucleotide, and the x-axis represents the 3' (right) flanking nucleotide.
As can be seen by the brightest red color, the nucleotide U at both positions allows for the most efficient editing. It is followed by combinations of UAC and AAC, which corresponds to the bars in C. However, as we saw in C, the UAC triplet is associated with increased untargeted editing.
Conclusion
The authors tested how the activity of REPAIRv1 is influenced by the PFS sequence and two flanking nucleotides around the target site. These differences can affect either dCas13b targeting or ADAR2 modification, respectively. By comparing all possible nucleotide combinations at these positions, the authors concluded that neither PFS nor the flanking nucleotides have a major impact on REPAIRv1. However, individual combinations vary the editing efficiency by around 20%.
Correction of disease-relevant human mutations using REPAIRv1
To demonstrate the broad applicability of the REPAIRv1 system for RNA editing in mammalian cells, we designed REPAIRv1 guides against two disease-relevant mutations: 878G→A (AVPR2W293X) in X-linked nephrogenic diabetes insipidus and 1517G→A (FANCC W506X) in Fanconi anemia. We transfected expression constructs for cDNA of genes carrying these mutations into HEK293FT cells and tested whether REPAIRv1 could correct the mutations. Using guide RNAs containing 50-nt spacers, we were able to achieve 35% correction of AVPR2 and 23% correction of FANCC (Fig. 4, A to D). We then tested the ability of REPAIRv1 to correct 34 different disease-relevant G→A mutations (table S4) and found that we were able to achieve substantial editing at 33 sites with up to 28% editing efficiency (Fig. 4E). The mutations we chose are only a fraction of the pathogenic G-to-A mutations (5739) in the ClinVar database, which also includes an additional 11,943 G-to-A variants (Fig. 4F and fig. S6). Because there are no strict sequence constraints (Fig. 3), REPAIRv1 is capable of potentially editing all of these disease-relevant mutations, especially given that we observed editing regardless of the target motif (Figs. 3C and 4G).
Fig. 4. Correction of disease-relevant mutations with REPAIRv1. (A) Schematic of target and guide design for targeting AVPR2 878G→A. (B) The 878G→A mutation (indicated by blue triangle) in AVPR2 is corrected to varying levels by using REPAIRv1 with three different guide designs. For each guide, the region of duplex RNA is outlined in red. Values represent mean ± SEM. Nontargeting guide is the same as in Fig. 2C. (C) Schematic of target and guide design for targeting FANCC 1517G→A. (D) The 1517G→A mutation (indicated by blue triangle) in FANCC is corrected to varying levels by using REPAIRv1 with three different guide designs. For each guide, the region of duplex RNA is outlined in red. The heatmap scale bar is the same as in (B). Values represent mean ± SEM. Nontargeting guide is the same as in Fig. 2C. (E) Quantification of the percent editing of 34 different disease-relevant G→A mutations selected from ClinVar by using REPAIRv1. Nontargeting guide is the same as in Fig. 2C. (F) Analysis of all the possible G→A mutations that could be corrected by using REPAIR as annotated in the ClinVar database. (G) The distribution of editing motifs for all G→A mutations in ClinVar is shown versus the editing efficiency by REPAIRv1 per motif as quantified on the Gluc transcript. Values represent mean ± SEM.
Question
Is REPAIRv1 able to correct endogenous transcripts, including disease-relevant mutant variants?
Panels A, B, C, & D
The figures A and C show the design of three targeting spaces for either AVPR2 or FANCC. As we can see from B and D, both transcripts can be edited with an efficiency level higher than 20%. One spacer targeting the AVPR2 transcript is particularly advantageous (the middle row in B). All spacers for the FANCC yield similar results.
Panel E
The histogram shows the editing rate for 34 transcripts. It varies greatly from almost 0% to 30% and in some cases is accompanied by significant untargeted editing. Different tendency to form duplexes and different nucleotides around the target site contribute to this difference.
Panel F
There are around 80,000 records for single nucleotide polymorphisms (SNPs) in the ClinVar database. 25% of them are G to A substitutions. However, only 25,000 SNPs have been lined to pathological conditions and round 5,000 of them are caused by G to A mutations.
Conclusion
REPAIRv1 can potentially edit all G>A substitutions in disease-causing SNPs. However, these SNPs comprise only 20% of all pathogenic SNPs.
Delivering the REPAIRv1 system to diseased cells is a prerequisite for therapeutic use, and we therefore sought to design REPAIRv1 constructs that could be packaged into therapeutically relevant viral vectors, such as adeno-associated viral (AAV) vectors. AAV vectors have a packaging limit of 4.7 kb, which cannot accommodate the large size of dCas13b-ADARDD [4473 base pairs (bp)] along with promoter and expression regulatory elements. To reduce the size, we tested a variety of N-terminal and C-terminal truncations of dCas13 fused to ADAR2DD(E488Q) for RNA-editing activity. We found that all C-terminal truncations tested were still functional and able to restore luciferase signal (fig. S7), and the largest truncation, C-terminal Δ984–1090 (total size of the fusion protein, 4152 bp) was small enough to fit within the packaging limit of AAV vectors.
Transcriptome-wide specificity of REPAIRv1
Although RNA knockdown with PspCas13b was highly specific in our luciferase tiling experiments, we observed off-target adenosine editing within the guide:target duplex (Fig. 2E). To see whether this was a widespread phenomenon, we tiled an endogenous transcript, KRAS, and measured the degree of off-target editing near the target adenosine (Fig. 5A). We found that for KRAS, although the on-target editing rate was 23%, there were many sites around the target site that also had detectable A-to-I edits (Fig. 5B).
Fig. 5. Characterizing specificity of REPAIRv1. (A) Schematic of KRAS target site and guide design. (B) Quantification of percent A-to-I editing for tiled KRAS-targeting guides. Editing percentages are shown for the on-target (blue triangle) and neighboring adenosine sites. For each guide, the region of duplex RNA is outlined in red. Values represent mean ± SEM. (C) Transcriptome-wide sites of significant RNA editing by REPAIRv1 (150 ng REPAIR vector transfected) with Cluc targeting guide. The on-target site Cluc site (254 A→I) is highlighted in orange. (D) Transcriptome-wide sites of significant RNA editing by REPAIRv1 (150 ng REPAIR vector transfected) with nontargeting guide. Nontargeting guide is the same as in Fig. 2C.
Question
How specific is the REPAIRv1 system at adenosine editing?
Panel B
The figure illustrates the editing rate at different adenosine positions for seven gRNAs with spacers shifted along the transcript (see Panel A for the design of the spacers). The different spacers are shown in different rows, and adenosines that may undergo editing are shown the columns. Darker red squares mean a particular spacer had a higher editing rate than the corresponding adenosine.
The spacer that mediates the highest rate of editing for the target (26) has noticeable off-target sites. Moreover, the next most active spacer predominantly targets irrelevant adenosine bases.
Panels C & D
Each dot in these plots represents a transcript. These transcripts are ranked along their y-axis corresponding to the proportion of molecules of the transcript that have been edited by REPAIRv1. Along the x-axis, the transcripts are placed by their genomic location.
Conclusion
Whole-transcriptome sequencing demonstrated that REPAIRv1 has a substantial level of off-target editing.
Because of the observed off-target editing within the guide:target duplex, we initially evaluated transcriptome-wide off-targets by performing RNA-seq on all mRNAs with 12.5x coverage. Of all the editing sites across the transcriptome, the on-target editing site had the highest editing rate, with 89% A-to-I conversion. We also found that there was a substantial number of A-to-I off-target events, with 1732 off-targets in the targeting guide condition and 925 off-targets in the nontargeting guide condition, with 828 off-targets shared between the targeting and nontargeting guide conditions (Fig. 5, C and D). Given the high number of overlapping off-targets between the targeting and nontargeting guide conditions, we reasoned that the off-targets may arise from ADARDD. To test this hypothesis, we repeated the Cluc-targeting experiment, this time comparing transcriptome changes for REPAIRv1 with a targeting guide, REPAIRv1 with a nontargeting guide, REPAIRv1 alone, or ADARDD(E488Q) alone (fig. S8). We found differentially expressed genes and off-target editing events in each condition (fig. S8, B and C). There was a high degree of overlap in the off-target editing events between ADARDD(E488Q) and all REPAIRv1 off-target edits, supporting the hypothesis that REPAIR off-target edits are driven by dCas13b-independent ADARDD(E488Q) editing events (fig. S8D).
Next, we sought to compare two RNA-guided ADAR systems that have been described previously (fig. S9A). The first uses a fusion of ADAR2DD to the small viral protein lambda N (ƛN), which binds to the BoxB-ƛ RNA hairpin (24). A guide RNA with double BoxB-ƛ hairpins guides ADAR2DD(E488Q) to edit sites encoded in the guide RNA (25). The second design uses full-length ADAR2 (ADAR2) and a guide RNA with a hairpin that the double-strand RNA (dsRNA)–binding domains (dsRBDs) of ADAR2 recognize (23, 26). We analyzed the editing efficiency of these two systems compared with REPAIRv1 and found that the BoxB-ADAR2 and full-length ADAR2 systems demonstrated 50 and 34.5% editing rates, respectively, compared with the 89% editing rate achieved by REPAIRv1 (fig. S9, B to E). Additionally, the BoxB and full-length ADAR2 systems created 1814 and 66 observed off-targets, respectively, in the targeting guide conditions, compared with the 2111 off-targets in the REPAIRv1 targeting guide condition. All the conditions with the two ADAR2DD-based systems (REPAIRv1 and BoxB) showed a high percentage of overlap in their off-targets, whereas the full-length ADAR2 system had a largely distinct set of off-targets (fig. S9F). The overlap in off-targets between the targeting and nontargeting conditions and between REPAIRv1 and BoxB conditions suggests that ADAR2DD drives off-targets independent of dCas13 targeting (fig. S9F).
Improving specificity of REPAIR through rational protein engineering
To improve the specificity of REPAIRv1, we used structure-guided protein engineering of ADAR2DD(E488Q). Because of the guide-independent nature of the off-targets, we hypothesized that destabilizing ADAR2DD(E488Q)–RNA binding would selectively decrease off-target editing, but maintain on-target editing because of increased local concentration from dCas13b tethering of ADAR2DD(E488Q) to the target site. We mutated residues in ADAR2DD(E488Q) previously determined to contact the duplex region of the target RNA (Fig. 6A) (19). To assess efficiency and specificity, we tested 17 single mutants with both targeting and nontargeting guides, under the assumption that background luciferase restoration in the nontargeting condition would be indicative of broader off-target activity. We found that mutations at the selected residues had substantial effects on the luciferase activity for targeting and nontargeting guides (Fig. 6, A and B, and fig. S10A). A majority of mutants either significantly improved the luciferase activity for the targeting guide or increased the ratio of targeting to nontargeting guide activity, which we termed the specificity score (Fig. 6, A and B).
Fig. 6. Rational mutagenesis of ADAR2DD to improve the specificity of REPAIRv1. (A) Quantification of luciferase signal restoration (on-target score, red boxes) by various dCas13-ADAR2DD mutants as well as their specificity score (blue boxes) plotted along a schematic of the contacts between key ADAR2 deaminase residues and the dsRNA target (the target strand is shown in gray; the nontarget strand is shown in red). All deaminase mutations were made on the dCas13-ADAR2DD(E488Q) background. The specificity score is defined as the ratio of the luciferase signal between targeting guide and nontargeting guide conditions. [Schematic of ADAR2 deaminase domain contacts with dsRNA is adapted from (20)] (B) Quantification of luciferase signal restoration by various dCas13-ADAR2 mutants versus their specificity score. Nontargeting guide is the same as in Fig. 2C. (C) Quantification of on-target editing and the number of significant off-targets for each dCas13-ADAR2DD(E488Q) mutant by transcriptome-wide sequencing of mRNAs. Values represent mean ± SEM. Nontargeting guide is the same as in Fig. 2C. (D) Transcriptome-wide sites of significant RNA editing with (top) REPAIRv1 and (bottom) REPAIRv2, with a guide targeting a pretermination site in Cluc. The on-target Cluc site (254 A→I) is highlighted in orange. Ten nanograms of REPAIR vector was transfected for each condition. (E) Representative RNA-seq reads surrounding the on-target Cluc-editing site (254 A→I; blue triangle) highlighting the differences in off-target editing between (top) REPAIRv1 and (bottom) REPAIRv2. A-to-I edits are highlighted in red; sequencing errors are highlighted in blue. Gaps reflect spaces between aligned reads. Nontargeting guide is the same as in Fig. 2C. (F) RNA editing with REPAIRv1 and REPAIRv2, with guides targeting an out-of-frame UAG site in the endogenous KRAS and PPIB transcripts. The on-target editing fraction is shown as a sideways bar chart on the right for each condition row. For each guide, the region of duplex RNA is outlined in red. Values represent mean ± SEM. Nontargeting guide is the same as in Fig. 2C.
Question
How can we select the best REPAIR system from many mutant versions?
Panels A & B
The mutants with an increased on-target editing ability are depicted by brighter red boxes in A or fall on the right side of the plot (to the right of E488Q) in B. The mutants with an improved specificity are represented by brighter blue boxes in A or are shifted upwards in B.
We can see that some dots shifted to the right lay almost on the same line as the E488Q protein. These mutants have enhanced editing activity (higher on-target score), but demonstrate the same specificity as the starting protein (almost the same position on the y-axis). This means that their editing activity is roughly equally enhanced toward the target and non-target sites and may result in even greater off-target numbers.
Panel C
Several mutant variants demonstrate an editing level of above 75%, similar to the starting protein. The bars on the right show the majority of the selected variants have reduced off-target numbers both in targeting and in non-targeting conditions.
You may notice that the variant R455E neither produced off-targets nor was able to edit the desired adenosine, although it had a high specificity score in B. To understand the reason, see Figure S10A (in the paper's supplementary materials) and see that this mutant had a low on-target signal but division by an even smaller non-target signal resulted in a substantial specificity score. In fact, both signals were at the background level.
Panels D & E
Two systems, REPAIRv1 and REPAIRv2, are compared in their off-target activity at the level of the whole transcriptome (D) or at the level of the target transcript (E). Each dot in D represents an off-target transcript (except the orange dot for the Cluc) plotted on the y-axis according to its editing rate, i.e., the percentage of molecules of this transcript with the modified adenosine. Each fragment in a row in E is an individual read obtained from the sequencing experiment. You see that some reads have G instead of A at other positions than the target site. It is the proportion of these sequences with the off-target A>G conversion that represents the editing rate (in D). The more coverage, the larger the number of reads is generated. Therefore, the higher the change to detect additional irrelevant edits that occur in a low number of molecules.
Conclusion
The versions of the REAPIR with mutant ADAR2 were compared by two characteristics, the level of the target adenosine editing and off-target activity. Because off-target activity has to be minimized, the authors chose the variant with the lowest off-target level which still demonstrated reasonable on-target editing.
We selected a subset of these mutants (Fig. 6B) for transcriptome-wide specificity profiling by next-generation sequencing. As expected, off-targets measured from transcriptome-wide sequencing correlated with our specificity score (fig. S10B) for mutants. We found that with the exception of ADAR2DD(E488Q/R455E), all sequenced REPAIRv1 mutants could effectively edit the reporter transcript (Fig. 6C), with many mutants showing reduction in the number of off-targets (Fig. 6C and figs. S10C and S11). We further explored motifs surrounding off-targets for the various specificity mutants and found that REPAIRv1 and most of the engineered variants exhibited a strong 3′ G preference for their off-target edits, which is in agreement with the characterized ADAR2 motif (fig. S12A) (28).
We focused on the mutant ADAR2DD(E488Q/T375G)—because it had the highest percent editing of the four mutants with the lowest numbers of transcriptome-wide off-targets—and termed it REPAIRv2. Compared with REPAIRv1, REPAIRv2 exhibited increased specificity, with a reduction from 18,385 to 20 transcriptome-wide off-targets with high-coverage sequencing (125x coverage, 10 ng of REPAIR vector transfected) (Fig. 6D). In the region surrounding the targeted adenosine in Cluc, REPAIRv2 also had reduced off-target editing, visible in sequencing traces (Fig. 6E). In motifs derived from the off-target sites, REPAIRv1 presented a strong preference toward 3′ G but showed off-target edits for all motifs (fig. S12B); by contrast, REPAIRv2 only edited the strongest off-target motifs (fig. S12C). The distribution of edits on transcripts was heavily skewed for REPAIRv1, with highly edited genes having more than 60 edits (fig. S13A), whereas REPAIRv2 only edited one transcript (EEF1A1) multiple times (fig. S13B). REPAIRv1 off-target edits were predicted to result in numerous variants, including 1000 missense base changes (fig. S13C), with 93 events in genes related to cancer processes (fig. S13D). In contrast, REPAIRv2 only had six predicted missense changes (fig. S13E), none of which were in cancer-related genes (fig. S13F). Analysis of the sequence surrounding off-target edits for REPAIRv1 or -v2 did not reveal homology to guide sequences, suggesting that off-targets are likely dCas13b-independent (fig. S14), which is consistent with the high overlap of off-targets between REPAIRv1 and the ADAR2 deaminase domain (fig. S8D). To directly compare REPAIRv2 with other programmable ADAR systems, we repeated our Cluc-targeting experiments with all systems at two different dosages of ADAR vector, finding that REPAIRv2 had comparable on-target editing with that of BoxB and ADAR2 but with substantially fewer off-target editing events at both dosages (fig S15). REPAIRv2 had enhanced specificity compared with REPAIRv1 at both dosages (fig. S15B), a finding that also extended to two guides targeting distinct sites on PPIB (fig. S16, A to D). It is also worth noting that in general, the lower-dosage condition (10 ng REPAIR vector) had fewer off-targets than that of the higher dosage condition (150 ng REPAIR vector) (fig. S5).
To assess editing specificity with greater sensitivity, we sequenced the low-dosage condition (10 ng of transfected DNA) of REPAIRv1 and v2 at much higher sequencing depth (125x coverage of the transcriptome). Increased numbers of off-targets were found at higher sequencing depths corresponding to detection of rarer off-target events (fig. S17). Furthermore, we speculated that different transcriptome states could also potentially alter the number of off-targeting events. Therefore, we tested REPAIRv2 activity in the osteosarcoma U2OS cell line, observing six and seven off-targets for the targeting and nontargeting guide, respectively (fig. S18).
We targeted REPAIRv2 to endogenous genes to test whether the specificity-enhancing mutations reduced nearby edits in target transcripts while maintaining high-efficiency on-target editing. For guides targeting either KRAS or PPIB, we found that REPAIRv2 had no detectable off-target edits, unlike REPAIRv1, and could effectively edit the on-target adenosine at efficiencies of 27.1% (KRAS) or 13% (PPIB) (Fig. 6F). This specificity extended to additional target sites, including regions that demonstrate high levels of background in nontargeting conditions for REPAIRv1, such as other KRAS or PPIB target sites (fig. S19). Overall, REPAIRv2 eliminated off-targets in duplexed regions around the edited adenosine and showed dramatically enhanced transcriptome-wide specificity.
Discussion
We show here that the RNA-guided RNA-targeting type VI-B CRISPR effector Cas13b is capable of highly efficient and specific RNA knockdown, providing the basis for improved tools for interrogating essential genes and noncoding RNA as well as controlling cellular processes at the transcript level. Catalytically inactive Cas13b (dCas13b) retains programmable RNA-binding capability, which we leveraged here by fusing dCas13b to the adenosine deaminase domain of ADAR2 to achieve precise A-to-I edits, a system we term REPAIRv1. Further engineering of the system produced REPAIRv2, which has dramatically higher specificity than previously described RNA-editing platforms (25, 29) while maintaining high levels of on-target efficacy.
Although Cas13b exhibits high fidelity, our initial results with dCas13b-ADAR2DD(E488Q) fusions revealed a substantial number of off-target RNA editing events. To address this, we used a rational mutagenesis strategy to vary the ADAR2DD residues that contact the RNA duplex, identifying a variant, ADAR2DD(E488Q/T375G), that is capable of precise, efficient, and highly specific editing when fused to dCas13b. Editing efficiency with this variant was comparable with or better than that achieved with two currently available systems, BoxB-ADAR2DD(E488Q) or ADAR2 editing. Moreover, the REPAIRv2 system created only 20 observable off-targets in the whole transcriptome, which is at least an order of magnitude better than both alternative editing technologies. Although it is possible that ADAR could deaminate adenosine bases on the DNA strand in RNA-DNA heteroduplexes (20), it is unlikely to do so in this case because Cas13b does not bind DNA efficiently and because REPAIR is cytoplasmically localized. Additionally, the lack of homology of off-target sites to the guide sequence and the strong overlap of off-targets with the ADARDD(E488Q)–only condition suggest that off-targets are not mediated by off-target guide binding. Deeper sequencing and novel inosine enrichment methods could further refine our understanding of REPAIR specificity in the future.
The REPAIR system offers many advantages compared with other nucleic acid–editing tools. First, the exact target site can be encoded in the guide by placing a cytidine within the guide across from the desired adenosine to create a favorable A-C mismatch ideal for ADAR-editing activity. Second, Cas13 has no targeting sequence constraints, such as a PFS or PAM, and no motif preference surrounding the target adenosine, allowing any adenosine in the transcriptome to be potentially targeted with the REPAIR system. The lack of motif for ADAR editing, in contrast with previous literature, is likely due to the increased local concentration of REPAIR at the target site owing to dCas13b binding. DNA base editors can target either the sense or antisense strand, whereas the REPAIR system is limited to transcribed sequences, constraining the total number of possible editing sites. However, because of the less constrained nature of targeting with REPAIR, this system can effect more edits within ClinVar (Fig. 4C) than Cas9-DNA base editors. Third, the REPAIR system directly deaminates target adenosines to inosines and does not rely on endogenous repair pathways to generate desired editing outcomes. Therefore, REPAIR should be able to mediate efficient RNA editing even in post-mitotic cells such as neurons. Fourth, in contrast to DNA editing, RNA editing is transient and can be more easily reversed, allowing the potential for temporal control over editing outcomes. The transient nature of REPAIR-mediated edits will likely be useful for treating diseases caused by temporary changes in cell state, such as local inflammation, and could also be used to treat disease by modifying the function of proteins involved in disease-related signal transduction. For instance, REPAIR editing would allow the recoding of some serine, threonine, and tyrosine residues that are the targets of kinases (fig. S20). Phosphorylation of these residues in disease-relevant proteins affects disease progression for many disorders, including Alzheimer’s disease and multiple neurodegenerative conditions (30). REPAIR might also be used to transiently or even chronically change the sequence of expressed, risk-modifying G-to-A variants so as to decrease the chance of entering a disease state for patients. For instance, REPAIR could be used to functionally mimic A-to-G alleles of IFIH1 that protect against autoimmune disorders such as type I diabetes, immunoglobulin A deficiency, psoriasis, and systemic lupus erythematosus (31, 32).
The REPAIR system provides multiple opportunities for additional engineering. Cas13b possesses pre–CRISPR-RNA (crRNA) processing activity (13), allowing for multiplex editing of multiple variants—any one of which alone may not affect disease, but together might have additive effects and disease-modifying potential. Extension of our rational design approach, such as combining promising mutations and directed evolution, could further increase the specificity and efficiency of the system, while unbiased screening approaches could identify additional residues for improving REPAIR activity and specificity.
Currently, the base conversions achievable by REPAIR are limited to generating inosine from adenosine; additional fusions of dCas13 with other catalytic RNA editing domains, such as APOBEC (apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like), could enable cytidine-to-uridine editing. Additionally, mutagenesis of ADAR could relax the substrate preference to target cytidine, allowing for the enhanced specificity conferred by the duplexed RNA substrate requirement to be exploited by C-to-U editors. Adenosine-to-inosine editing on DNA substrates may also be possible with catalytically inactive DNA-targeting CRISPR effectors, such as dCas9 or dCpf1, either through formation of DNA-RNA heteroduplex targets (20) or mutagenesis of the ADAR domain.
We have demonstrated the use of the PspCas13b enzyme as both an RNA-knockdown and RNA-editing tool. The dCas13b platform for programmable RNA binding has many applications, including live transcript imaging, splicing modification, targeted localization of transcripts, pulldown of RNA-binding proteins, and epitranscriptomic modifications.We used dCas13 to create REPAIR, adding to the existing suite of nucleic acid–editing technologies. REPAIR provides a new approach for treating genetic disease or mimicking protective alleles and establishes RNA editing as a useful tool for modifying genetic function.
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/358/6366/1019/suppl/DC1
Materials and Methods
Supplementary Text
Figs. S1 to S20
Tables S1 to S9
http://www.sciencemag.org/about/science-licenses-journal-article-reuse
This is an article distributed under the terms of the Science Journals Default License.
REFERENCES AND NOTES
1. P. D. Hsu, E. S. Lander, F. Zhang, Cell 157, 1262–1278 (2014).
2. A. C. Komor, A. H. Badran, D. R. Liu, Cell 168, 20–36 (2017).
3. L. Cong et al., Science 339, 819–823 (2013).
4. P. Mali et al., Science 339, 823–826 (2013).
5. B. Zetsche et al., Cell 163, 759–771 (2015).
6. H. Kim, J. S. Kim, Nat. Rev. Genet. 15, 321–334 (2014).
7. A. C. Komor, Y. B. Kim, M. S. Packer, J. A. Zuris, D. R. Liu, Nature 533, 420–424 (2016).
8. K. Nishida et al., Science 353, aaf8729 (2016).
9. Y. B. Kim et al., Nat. Biotechnol. 35, 371–376 (2017).
10. O. O. Abudayyeh et al., Science 353, aaf5573 (2016).
11. S. Shmakov et al., Mol. Cell 60, 385–397 (2015).
12. S. Shmakov et al., Nat. Rev. Microbiol. 15, 169–182 (2017).
13. A. A. Smargon et al., Mol. Cell 65, 618–630.e7 (2017).
14. J. S. Gootenberg et al., Science 356, 438–442 (2017).
15. O. O. Abudayyeh et al., Nature 550, 280–284 (2017).
16. K. Nishikura, Annu. Rev. Biochem. 79, 321–349 (2010).
17. M. H. Tan et al., Nature 550, 249–254 (2017).
18. B. L. Bass, H. Weintraub, Cell 55, 1089–1098 (1988).
19. M. M. Matthews et al., Nat. Struct. Mol. Biol. 23, 426–433 (2016).
20. Y. Zheng, C. Lorenzo, P. A. Beal, Nucleic Acids Res. 45, 3369–3377 (2017).
21. A. Kuttan, B. L. Bass, Proc. Natl. Acad. Sci. U.S.A. 109, E3295–E3304 (2012).
22. S. K. Wong, S. Sato, D. W. Lazinski, RNA 7, 846–858 (2001).
23. M. Fukuda et al., Sci. Rep. 7, 41478 (2017).
24. M. F. Montiel-Gonzalez, I. Vallecillo-Viejo, G. A. Yudowski, J. J. Rosenthal, Proc. Natl. Acad. Sci. U.S.A. 110, 18285–18290 (2013).
25. M. F. Montiel-González, I. C. Vallecillo-Viejo, J. J. Rosenthal, Nucleic Acids Res. 44, e157 (2016).
26. J. Wettengel, P. Reautschnig, S. Geisler, P. J. Kahle, T. Stafforst, Nucleic Acids Res. 45, 2797–2808 (2017).
27. Y. Wang, J. Havel, P. A. Beal, ACS Chem. Biol. 10, 2512–2519 (2015).
28. K. A. Lehmann, B. L. Bass, Biochemistry 39, 12875–12884 (2000).
29. T. Stafforst, M. F. Schneider, Angew. Chem. Int. Ed. Engl. 51, 11166–11169 (2012).
30. C. Ballatore, V. M. Lee, J. Q. Trojanowski, Nat. Rev. Neurosci. 8, 663–672 (2007).
31. Y. Li et al., J. Invest. Dermatol. 130, 2768–2772 (2010).
32. R. C. Ferreira et al., Nat. Genet. 42, 777–780 (2010).
For references 33-38, please see the Supplementary Materials section.