MS2 RNA aptamer greatly enhances prime editing in rice

Prime editing is a universal and very promising precise genome editing technology. However, optimization of prime editor (PE) from different aspects remains vital for its use as a routine tool in plant basic research and crop molecular breeding. In this report, we tested MS2-based prime editor (MS2PE). We fused the M-MLV reverse transcriptase (RT) gene variant to the MS2 RNA binding protein gene, MCP , and allowed the MCP-RT fusion gene to co-express with the SpCas9 nickase gene, SpCas9H840A , and various engineered pegRNAs harboring MS2 RNA (MS2pegR). Compared with control PEs, MS2PEs significantly enhanced editing efficiency at four of six targets in rice protoplasts, and achieved 1.2~10.1-fold increase in editing efficiency at five of six targets in transgenic rice lines. Furthermore, we tested total 22 different MS2pegR scaffolds, 3 RT variants or genes, 2 MCP variants, and various combinations of the Cas9 nickase, RT, and MCP modules. Our results demonstrated a new strategy for more efficient prime editing and provide a platform for further directed evolution of PEs.


Introduction
CRISPR/Cas systems that induced DNA double-strand breaks (DSBs) have been mainly used as a search-and-disrupt genome editing technology in plants (Zhan et al., 2021). DSB-based tools can also induce homology-directed repair (HDR) of host cells and thus can be used as a search-andreplace genome editing (Zhan et al., 2021). However, HDR-mediated gene targeting (GT) requires donor DNA, which in turn requires efficient delivery , synergetic processing (Barone et al., 2020), and in vivo amplifications or in vitro chemical modifications when possible (Cermak et al., 2015;Lu et al., 2020). In addition, as non-homologous end joining (NHEJ) is the preferred DSB repair mechanism in somatic plant cells, inherent low efficiency of HDR has been the main obstacle to practical applications of GT in crops (Zhan et al., 2021). These aspects prevent GT from broad applications by plant researchers although major advances have been made (Barone et al., 2020;Lu et al., 2020).
CRISPR/Cas-derived base editing without requiring DSBs or donor DNA templates is a search-andconvert editing technology for base conversions (Koblan et al., 2018;Jin et al., 2020;Richter et al., 2020). However, although two types of base editors for inducing all 4 types of base transition mutations and a type of base editors for inducing C to G transversion in mammalian cells have been developed, base editors for inducing the rest 6 types of base transversion mutations remain to be developed (Anzalone et al., 2019;Kurt et al., 2021;Zhao et al., 2021). In addition, base editing has a much stricter requirement for target selection due to the restriction of editing windows and has the potential to induce off-target mutations in editing windows harboring multiple editable bases.
These aspects restrict broad applications of base editors.
Like base editing, CRISPR/Cas-derived prime editing requires no DSBs or donor DNA for precise genome modifications. However, unlike base editing, which is mainly used for inducing base conversions, prime editing is a universal search-and-replace genome modification technology (Anzalone et al., 2019). Prime editors (PEs) can induce controllable rather than random indel mutations and unlimited base conversions, and thus has a potency to substitute the HDR-based genome editors to a large extent and base editors in a complete way. In plants, the main obstacle to broad applications of prime editing is the overall low editing efficiency of PEs (Jiang et al., 2020;Lin et al., 2021). Thus, optimization of PEs for higher editing efficiency from different aspects is vital for their acceptance as routine tools by plant researchers. In this report, we developed MS2 RNA aptamer-based prime editing in rice and demonstrated that the MS2-based strategy greatly improved prime-editing efficiency.

MS2-based prime editors enhance prime editing in rice protoplasts and transgenic lines
The MS2 RNA aptamer and its binding protein MCP in combination with CRISPR/Cas9 system have been used for efficient gene activation and base editing (Konermann et al., 2015;Zalatan et al., 2015;Hess et al., 2016;Li et al., 2020), however, it remains unknown whether the MS2 system can be used for efficient prime editing. To test MS2-based prime editors (MS2PEs), as an initial step we tested five engineered pegRNA scaffolds, named MS2pegRs. These five MS2pegRs harbor two copies of MS2 RNA aptamers, MS2 and/or f6, which are located at different regions of the pegRNA scaffold (Figure 1a, b; Figure S1). We fused the M-MLV reverse transcriptase (RT) gene variant to MCP and allowed the MCP-RT fusion gene to co-express with a Cas9 nickase gene, SpCas9H840A, and MS2pegR genes (Figure 1a, c). We tested two MCP variants, MCP1 and MCP2 (Supplemental information), and the five MS2pegRNA scaffolds at six rice genomic targets.
We first tested editing efficiency of MS2PEs in rice protoplasts. Compared with control PEs, out of the five MS2pegR scaffolds, pegR2.3b and pegR2f.3b ( Figure S1) significantly enhanced editing efficiency at four of six targets, and displayed similar editing efficiency at the rest two targets ( Figure   1d). In general, there was no significant difference between pegR2.3b and pegR2f.3b. There was also no significant difference between MCP1 and MCP2 ( Figure 1d). We further tested editing efficiency of MS2PEs in rice transgenic lines ( Figure 1e). The results indicated that MS2PEs, each of which harbors two similar MS2pegRs, pegR2.3b and pegR2f.3b, for enhancing expression, increased editing efficiencies by a factor of 1.2~ 10.1 times at five of six targets compared with control PEs, each of which harbors two same pegRNAs for enhancing expression (Figure 1c, e). These results demonstrated that MS2PEs greatly enhanced prime editing in rice.
In protoplasts, the most efficient three MS2PEs were those inducing mutations in OsACC-I1879V, OsACC-D2176G, and OsALS-S627I.a, whereas in transgenic lines, the most efficient three MS2PEs were those inducing mutations in OsACC-D2176G, OsACC-I1879V, and OsALS-S627I.a. This inconsistency may reflect on differences of edited cell types, i.e., protoplast cells and callus cells. In theory, editing processes last for much longer time in callus cells than in protoplast cells and promoters may have different activity in these two types of cells.

Comparisons of various MS2PEs with different MS2pegR scaffolds and RT modules
To attempt to optimize MS2PEs, we tested more MS2PEs with different MS2pegR scaffolds and RT modules. To more extensively evaluate effects of MS2pegRNA scaffolds on editing efficiency, we designed additional 17 MS2pegR scaffolds (Figure 2a; Figure S2-S6). The results indicated that only pegR1.3b and pegR1f.3b achieved editing efficiency similar to pegR2.3b or pegR2f.3b (Figure 2a

Methods
All primers used in this study are listed in Table S1, and the sequences of the MCP-RT and pegRNA expression cassettes are listed in Supplemental information. Vectors described in this study together with their annotated sequences are available from Addgene and/or MolecularCloud (GenScript).

Rice protoplast transfection and analysis of prime editing
We used the Japonica rice (Oryza sativa) variety Zhonghua11 to prepare protoplasts. We transferred the plasmids into protoplasts by PEG-mediated methods and incubated the transfected protoplasts at 26 °C for 48 hours. To analyze efficiency of prime editing, we extracted the genomic DNA and used the primers listed in Table S1 to amplify target fragments for deep sequencing. For each target site, sequencing fragment was repeated three times using genomic DNA extracted from three independent protoplast samples.

Rice transformation and analysis of prime editing
We separately transformed the pGreen3 binary vectors into the engineered Agrobacterium strain LBA4404/pVS1-VIR2 to generate strains harboring the ternary vector system (Zhang et al., 2019).
We used these strains to transform callus cells of Zhonghua11 separately. To analyze the mutations induced by prime editing, we amplified fragments spanning the target sites in genes from genomic DNA of the transgenic lines using PCR with primers listed in Table S1. We then submitted the purified PCR products to direct sequencing with primers listed in Table S1.    original PEs served as positive controls named CK(+). Efficiency (mean ± s.e.m.) was calculated from three independent experiments (n = 3). The P values were obtained by using the two-tailed Student's t-test, comparing MS2PEs with positive controls. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. d, Comparisons of various combinations of MCP and RT by using a dual vectorbased strategy for protoplast transfection. In the dual vector system, one vector harbors the Cas9n cassette and two same MS2pegRNAs whereas another vector harbors a fusion gene of MCP and RT.