Developing PspCas13b-based enhanced RESCUE system, eRESCUE, with efficient RNA base editing

RNA base editing is potential for cellular function research and genetic diseases treating. There are two main RNA base editors, REPAIR and RESCUE, for in vitro use. REPAIR was developed by fusing inactivated Cas13 (dCas13) with the adenine deaminase domain of ADAR2, which efficiently performs adenosine-to-inosine (A-to-I) RNA editing. RESCUE, which performs both cytidine-to-uridine (C-to-U) and A-to-I RNA editing, was developed by fusing inactivated Cas13 (dCas13) with the evolved ADAR2. However, the relatively low editing efficiency of the RESCUE system limits its broad application. Here, we constructed an enhanced RESCUE (eRESCUE) system; this dPspCas13b-RESCUE-NES system was generated by fusing inactivated PspCas13b with the evolved ADAR2. We determined the endogenous mRNA A-to-I and C-to-U editing efficiency mediated by the dPspCas13b-RESCUE-NES system in HEK-293T cells. This new RNA base editor was then used to induce 177Ser/Gly conversion of inhibitor kappa B kinase β (IKKβ) by changing the genetic code from AGU to GGU. The results showed that the eRESCUE editor mediates more efficient A-to-I and C-to-U RNA editing than the RESCUE RNA editor, as was previously reported. The 177Ser/Gly conversion of IKKβ, accomplished by converting the genetic code from AGU to GGU, resulted in a decrease in the phosphorylation of IKKβ and downregulation of downstream IKKβ-related genes. In summary, we developed a more efficient RNA base editor, eRESCUE, which may provide a useful tool for biomedical research and genetic disease treatment. Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s12964-021-00716-z.


Background
Previous studies have reported several tools that mediate adenosine-to-inosine (A-to-I) RNA editing in vivo [1][2][3]. Currently, there are two main RNA base editing tools in vitro: RNA editing for programmable A-to-I replacement (REPAIR) [4] and RNA editing for specific C-to-U exchange (RESCUE) [5]. REPAIR was the first RNA base editor developed and was constructed using a catalytically inactivated Cas13 ortholog from Prevotella sp. (dPspCas13b) fused with the adenosine deaminase acting on RNA type 2 (ADAR2) [4]. To expand the application of RNA editor systems, a RESCUE RNA base editor that performs both C-to-U and A-to-I RNA editing was successfully developed by fused with the evolved ADAR2 which served as a cytidine deaminase, then fused to an inactivated Cas13 ortholog from Riemerella anatipestifer (dRanCas13b) [5]. However, it should be noted that the efficiency of both A-to-I and C-to-U RNA editing is relatively low, especially for C-to-U base editing.
In this study, we tried to improve the RNA editing efficiency of the RESCUE system by using a Cas13 ortholog, dPspCas13b, with a nuclear export sequence (NES) to guide RNA base editors to edit mRNA in the cytoplasm. We successfully developed a PspCas13b-based enhanced RESCUE system, eRESCUE, with efficient RNA base editing abilities.

RESCUE editing in mammalian cells
Before transfection, 293T cells were seeded in 24-well plates which have been coated by D-lysine, and maintained at approximately 60-70% confluence. Then cells were transfected using EZ Trans Reagent (Shanghai Life iLab) according to the manufacturer's protocols. For transfection, dPspCas13b-RESCUE or the dRan-Cas13b-RESCUE-(800 ng) and gRNA-expressing plasmids (400 ng) were mixed and added to each well. DNA (1.2 µg) and 3.6 µL EZ Trans Reagent were diluted in 50 µL DMEM. The diluted EZ Trans Reagent was then added into the diluted DNA solution, mixed gently, and incubated for 15 min at room temperature (20-25 °C) to form DNA-EZ Trans Reagent complexes. After 15 min of incubation, the DNA-EZ Trans Reagent complexes were directly added to each well and mixed gently by rocking the plate back and forth. At 6 h post-transfection, the complexes were removed, and 0.5 mL complete growth medium was added to the cells. Post-transfection 48 h, GFP-positive cells were collected by fluorescence-activated cell sorting (FACS).

Flow cytometry
Cells were collected and subjected to FACS at 48 h after transfection. The GFP signal was detected via FACS. Cells (2 × 10 4 ) with positive GFP signals were collected and used to extract total RNA for editing efficiency analysis. More than 5 × 10 5 GFP-positive cells were harvested and used to extract total RNA for off-target analysis.

RNA editing efficiency analysis
GFP-positive cells were sorted by FACS. Total RNA of the collected cells was immediately extracted by using the TRIzol reagent (Invitrogen) according to the manufacturer's instructions. cDNA was generated by using the HiScript II Q RT SuperMix (Vazyme). Phanta ® Max Super-Fidelity DNA Polymerase (Vazyme) was used for PCR amplifying. The PCR amplification primers are listed in Additional file 1: Table S2. The online software EditR (https:// moria rityl ab. shiny apps. io/ editr_ v10/) was used to calculate editing efficiency by analysing the Sanger sequencing results of the PCR-amplified fragments.

Differential gene expression analysis
Total RNA from GFP-positive cells was extracted using TRIzol reagent. cDNA was synthesised using oligo d (T) primers and used as the templates for real-time PCR. SYBR green-based real-time PCR was performed to evaluate the mRNA expression level. PCR primers were showed in Additional file 1: Table S3. The reaction was performed at 95 °C for 2 min, followed by 40 cycles of 95 °C for 15 s and 61 °C for 1 min using an ABI 7300 detection system. The standard curve method was used for quantification, and the cDNA of the detected mRNA was tenfold serially diluted to generate the standard curve. The mRNA quantities of the samples were determined by linear extrapolation of the Ct values plotted against the standard curve. All assays were repeated at least three times, and each experiment was performed in triplicate. One-way or two-way ANOVA with multiple comparison corrections was used to assess the statistical significance of transcript changes using Prism 7.

Whole-transcriptome sequencing
To examine RNA off-target effects across the transcriptome, each selected sample was comprised of two biological repeats, and total RNA was extracted using TRIzol. A total amount of 1 µg RNA per sample was used as input material for the RNA sample preparations. Sequencing libraries were generated using NEBNext ® UltraTM RNA Library Prep Kit for Illumina ® (NEB, USA) following manufacturer's recommendations and index codes were added to attribute sequences to each sample. All the cDNA samples were sequenced by Novogene Co., Ltd.
GATK2 (v3.7) software was used to perform SNP calling. Raw vcf files were filtered with GATK standard filter method and other parameters (cluster:3; WindowSize:35; QD < 2.0 o; FS > 30.0; DP < 10 and Snp Eff software was used to annotation for the variable site.

Statistical analysis
All data are presented as mean ± SEM, as indicated. Statistical analysis of the results was performed using GraphPad Prism (GraphPad PRISM, Version 8.0). The statistical difference between the two groups was evaluated using one-way ANOVA. All data are expressed as arithmetic mean ± SEM. The level of significance was set at p < 0.05, whereas asterisks show differences at the following levels: *p < 0.05, **p < 0.01, ***p < 0.001.

dPspCas13b-RESCUE-NES performed improved RNA base editing at exogenous sites relative to dRanCas13b-RESCUE-NES
In eukaryotic cells, the primary transcript (pre-mRNA) is synthesised from a DNA template in the cell nucleus by transcription, and the pre-mRNA is completely processed to mature messenger RNA (mRNA) in the cytoplasm [6]. The dRanCas13b protein is guided by the NES for mRNA A-to-I and C-to-U editing in the RES-CUE system. Given that PspCas13b mediated the highest efficiency in knocking down endogenous KRAS compared to other optimised Cas13b systems or hairpin-mediated RNA (shRNA) [4], we generated mammalian codon-optimized dPspCas13b and constructed the dPspCas13b-based RESCUE system by replacing dRanCas13b with dPspCas13b to explore whether the dPspCas13b-RESCUE system could be more efficient. Two RESCUE RNA base editors, dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES, were constructed ( Fig. 1a). We selected four targets, KRAS site 1 and 2, CTNNB1, and NF2, which showed relatively high A-to-I or C-to-U editing efficiency in the dRanCas13b-RESCUE system [5]. We compared the A-to-I and C-to-U RNA editing efficiency in 293T cells by co-transfection with the dPspCas13b-RESCUE or dRanCas13b-RESCUE systems with the sgRNA expression constructs for target sites (i.e., KRAS site 1 and site 2, CTNNB1, and NF2). At 48 h post-transfection, we collected 2 × 10 4 GFP-positive cells by FACS. We extracted the total RNA of the 2 × 10 4 GFP-positive cells and obtained cDNA by reverse transcription. The targeted sequences were amplified by PCR and subjected to Sanger sequencing. We used the general analytic tool, EditR (https:// moria rityl ab. shiny apps. io/ editr_ v10/) [7] to analyse the Sanger sequencing results for calculating the editing efficiency. Notably, the results showed that the dPspCas13b-RESCUE-NES system mediated A-to-I RNA editing that is more efficient than the dRanCas13b-RESCUE-NES system. In particular, the A-to-I RNA editing efficiency of KRAS site 1 mediated by dPspCas13b-RESCUE-NES was up to 78% (Fig. 1b).
Similarly, the dPspCas13b-RESCUE-NES system showed more efficient C-to-U RNA editing than the dRan-RESCUE-NES system at two sites, KRAS site 1 and 2 (Fig. 1c). As shown in Fig. 1c, the C-to-U RNA editing efficiency of KRAS site 2 mediated by dPspCas13b-RES-CUE-NES was up to 58%. The results demonstrated that the NES guided dPspCas13b-RESCUE system might be a more efficient RNA editor for RNA A-to-I and C-to-U editing.

IKKβ 177Ser/Gly substitution via dPspCas13b-RESCUE-NES system resulted in AGU/GGU genetic code change
Given the success of the dPspCas13b-RESCUE-NES RNA base editor in vitro, it was next examined in biological studies. Diverse cellular stresses, such as inflammatory cytokines, bacterial or viral products and DNA damage could activate the NF-κB signalling pathway [8]. Genes involved in immune response, growth control and protection against apoptosis are activated by unbound NF-κB, which located in the nucleus. NF-κB activation depends on the IκB kinase (IKK). The phosphorylation of various IκB and NF-κB proteins are catalysed by IKK complex integrates signals which are from NF-κB activating stimuli [9]. Phosphorylation of Ser 177 and Ser 181 which locate in the activation loop of IKKβ determine the activation of IKK [10]. Taking advantage of dPsp-Cas13b-RESCUE-NES, we set out to study the effects of phosphorylation of 177 Ser of IKKβ. We designed three dPspCas13b-RESCUE-NES guides for IKKβ 177Ser The comparison of C-to-U RNA editing efficiency between dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems. Bars represent the mean ± SEM. Different asterisks indicate significant differences as follows: *p < 0.05; **p < 0.01; ***p < 0.001; ns, no significance (AGU)/Gly (GGU) substitution by mediating IKKβ mRNA A-to-I editing. We constructed three sgRNA expression vectors in which the target base A is positioned at location 24, 25, and 26 for IKKβ 177Ser (AGU)/ Gly (GGU) substitution. We then detected the A-to-I efficiency of the three sgRNAs that were mediated by the dPspCas13b-RESCUE-NES system in 293T cells. The results showed that the A-to-I editing efficiency of three replicates were 23%, 26%, and 19% for position 24 of sgRNA1; 36%, 32%, and 37% for position 25 of sgRNA2; and 29%, 30%, 34%, for position 26 of sgRNA3, respectively (Fig. 3a). Then, we used sgRNA2, which performed The effective editing of targeted IKKβ mRNA contributed to IKKβ-related gene expression detection. Bars represent the mean ± SEM. Different asterisks indicate significant differences as follows: *p < 0.05; **p < 0.01; ***p < 0.001; ns, no significance. c The red optical density statistics of p65 proteins in the cell nucleus. Statistical method: The analysis tool Image J was used to determine the grey scale of the mCherry protein located in the nucleus of cells in the IKKβ-177 Ser gRNA group and the non-target control group for semiquantitative analysis. Three representative pictures from each group were analysed. Bars represent mean ± SEM. "**"p < 0.01 best among the three sgRNAs for IKKβ mRNA A-to-I editing.
The sgRNA2 expression construct and dPspCas13b-RESCUE-NES editor were co-transfected into 293T cells. We then performed immunostaining of the p65 subunit, which is a major component of NF-κB complexes and is responsible for trans-activation [11]. The results showed that more p65 proteins shifted into the cytoplasm in the sgRNA2 and dPspCas13b-RESCUE-NES co-transfection groups compared to that with the control group or the non-targeting sgRNA and dPspCas13b-RESCUE-NES co-transfection groups (Fig. 3c). Consistently, detection of the relative expression of the downstream regulated genes by qPCR showed that the expression of p53 and GADD45, which function in cell cycle regulation, c-Myc and Bcl2 for cell proliferation, and TRAF1 and TRAF2 for apoptosis were significantly decreased (Fig. 3b). These results demonstrated that the dPspCas13b-RES-CUE-NES system mediated efficient A-to-I base editing, which resulted in the dephosphorylation of IKKβ 177Ser by changing the genetic code from AGU to GGU. The dephosphorylation of IKKβ 177Ser further downregulated IKKβ-related genes. These results indicate that the dPspCas13b-RESCUE-NES RNA editor is a versatile tool for regulation of protein phosphorylation and functional amino acid mutations.

Examination of RNA off-target effects resulting from the dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems
Given that RNA editors usually induce off-target effects, we checked the possible off-targets resulting from the dPspCas13b-RESCUE-NES and dRanCas13b-RES-CUE-NES systems at three endogenous sites, including KRAS site 1, CTNNB1, and RAF1. The RNA base editing efficiency of KRAS site 1, CTNNB1, and RAF1, which are mediated by dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems were detected. The results showed that the A-to-I on-target efficiency of the dPspCas13b-RESCUE-NES system were 68% and 72% for KRAS, 53% and 51% for CTNNB1, and 30% and 31% for RAF1, respectively; additionally, the dRanCas13b-RESCUE-NES system results were 35% and 40% for KRAS, 44% and 46% for CTNNB1, and 28% and 29% for RAF1, respectively (Fig. 4a). The C-to-U on-target editing efficiency of the dPspCas13b-RES-CUE-NES system was 46% and 42% for KRAS, 45% and 53% for CTNNB1, 29% and 31% for RAF1, respectively, whereas the dRanCas13b-RESCUE-NES system were 20% and 21% for KRAS, 20% and 22% for CTNNB1, and 19% and 12% for RAF1, respectively (Fig. 4b). For A-to-I and C-to-U RNA off-target detection, three experimental groups were examined with two replicates per transfection group: dPspCas13b-RESCUE-NES and target sgRNAs group; dRanCas13b-RESCUE-NES group; and target sgRNAs and EGFP expression plasmid transfection groups. Approximately 5 × 10 6 GFP + positive cells were collected by FACS, and total RNA was extracted from these cells for RNA-seq analysis. We evaluated transcriptome-wide off-targets by RNA-seq over all mRNAs with 50× coverage. Data from the dPspCas13b-RESCUE and target sgRNA transfection groups, as well as the dRanCas13b-RESCUE-NES and target sgRNA transfection groups were filtered by the data of the EGFP expression plasmid transfection groups. We found that there were substantial A-to-I off-target events (Fig. 4c, e) and C-to-U off-target events (Fig. 4d, f ). A greater number of A-to-I and C-to-U off-target events appeared with the dPspCas13b-RESCUE-NES RNA editor, suggesting that further optimisation is necessary to decrease the offtargets of the dPspCas13b-RESCUE-NES system.

Discussion
RESCUE is a potential RNA base editing technology which can mediate both A-to-I and C-to-U conversion in RNA [5]. However, the relatively low editing efficiency of the RESCUE system limits its applications. To improve this system, we comprehensively compared the A-to-I and C-to-U editing efficiency between the two editors, dPspCas13b-RESCUE-NES, and dRanCas13b-RESCUE-NES. Our results demonstrated that dPsp-Cas13b-RESCUE-NES system was more efficient of the two. The dPspCas13b-RESCUE-NES system mediated up to 78% A-to-I editing efficiency (Fig. 2a) and 58% C-to-U Fig. 4 Transcriptome-wide off-target determination with the dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems. a Heatmap of the A-to-I editing rate of adenines covered by dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems targeting KRAS, CTNNB1, and RAF1 genes. Each target contains two experimental repeats. b Heatmap of the C-to-U editing rate on cytosine covered by dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES system targeting at KRAS, CTNNB1, and NFKB1 gene. Each target contains two experiment repeats. c Manhattan plots of dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES A-to-I off-targets. n, total number of A-to-I off-target SNPs. d Manhattan plots of dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES C-to-U off-targets. n, total number of C-to-U off-target SNPs. e Comparison of A-to-I RNA off-target numbers between dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems. Bars represent the mean ± SEM. Different asterisks indicate significant differences as follows: **p < 0.01; ***p < 0.001. f The comparison of C-to-U RNA off-target numbers between dPspCas13b-RESCUE-NES and dRanCas13b-RESCUE-NES systems. Bars represent the mean ± SEM. Different asterisks indicate significant differences as follows: **p < 0.01; ***p < 0.001 (See figure on next page.) Rep

Learn more biomedcentral.com/submissions
Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: