Shaping rice Green Revolution traits by engineering <scp>ATG</scp> immediate upstream 5′‐<scp>UTR</scp> sequences of <scp><i>OsSBI</i></scp> and <scp><i>OsHTD1</i></scp>

Abstract

During the rice ‘Green Revolution’, the semidwarf phenotype for lodging resistance and the high-tillering trait for high yield were co-selected by pyramiding semi-dwarf 1 (SD1) and high tillering and dwarf 1/carotenoid cleavage dioxygenase 7 (HTD1/CCD7) from Dee-geo-woo-gen and Peta, two parents of the ‘miracle rice’ IR8 (Asano et al., 2011; Wang et al., 2020). Recent advances in genome editing have offered a powerful tool for crop breeding (Gao, 2021), and how to modulate endogenous protein levels has become crucial for creating new elite alleles (Song et al., 2022; Xue et al., 2023). Recent studies revealed that the ATG start codon immediate upstream 5′-UTR sequence (AUS) is an important cis-control feature in yeast, human cell, and Arabidopsis protoplast (Dvir et al., 2013; Kim et al., 2014; Sample et al., 2019). Especially, the occurrence of a trinucleotide AAA in the AUS correlates with a high translation rate in Arabidopsis protoplast, while a trinucleotide TTT is associated with decreased translation efficiency (Kim et al., 2014; Li et al., 2019). Therefore, we expected the prime editing (PE) system could be applied to design the AUS to create artificial beneficial alleles of target genes without affecting coding regions in rice. Here, we chose two genes, a strigolactone (SL) biosynthesis gene OsHTD1/OsCCD7 and a gibberellin metabolism gene shortened basal internodes (OsSBI) (Liu et al., 2018; Wang et al., 2020), to examine whether the rational design of their AUSs could create useful alleles to acquire Green Revolution traits. We first tested the protein output of the target genes' AUS in rice protoplasts via a dual-luciferase (LUC) reporter assay. To increase the protein abundance, we designed the substitution (OsSBIAUS-Sub) or insertion (OsSBIAUS-Ins) of AAA at the first three nucleotides of wide-type (WT) GCC (OsSBIAUS-WT) in OsSBI AUS. The results showed that OsSBIAUS-Sub and OsSBIAUS-Ins produced more LUC activities than OsSBIAUS-WT (Figure 1a, Figure S1A). Next, we designed OsHTD1AUS by changing first four nucleotides of OsHTD1 AUS from AAAG (OsHTD1AUS-WT) to TTTT (OsHTD1AUS-Sub) or inserting TTT (OsHTD1AUS-Ins) to decrease protein abundance. We found that the translation of LUC in OsHTD1AUS-Sub and OsHTD1AUS-Ins was substantially reduced compared to that of OsHTD1AUS-WT in transient expression assays (Figure 1b, Figure S1B). These results showed that the engineered AUSs of OsSBI and OsHTD1 could increase or decrease the translation rate as designed in rice protoplasts. We then tested this strategy in vivo by introducing the AAA insertion at OsSBI AUS via PE (Figure 1c, Figure S2). Among 46 T0 transgenic-positive plants, two homozygous lines with the 3-nt AAA insertion, OsSBIaus line 1 (L1) and L2, were obtained (Figure S3), and their transgene-free T2 plants were used for analysis (Figure 1d). We found that OsSBI mRNA levels were significantly decreased in OsSBIaus L1 and L2 compared to that of WT (Figure 1e), which may be due to the decreased mRNA stability of OsSBI in OsSBIaus lines (Figure S4). However, OsSBI protein abundance was strongly elevated in OsSBIaus L1 and L2 (Figure 1f), suggesting that the translational regulation may have a larger effect than mRNA stability after the genome editing of OsSBI AUS. Consistent with this, the plant heights of OsSBIaus L1 and L2 were significantly reduced (Figure 1g). Furthermore, we applied PE to insert the TTT sequence at OsHTD1 AUS (Figure 1c; Figures S5 and S6). Two homozygous transgene-free T2 lines with the designed 3-nt TTT insertion, OsHTD1aus L1 and L2, were obtained (Figure 1h). The OsHTD1 mRNA levels in OsHTD1aus lines were significantly increased (Figure 1i), possibly due to the increased mRNA stability of OsHTD1 (Figure S7). However, OsHTD1 protein abundance was largely decreased in OsHTD1aus lines compared with WT (Figure 1j), suggesting a larger effect of the translational regulation than mRNA stability in this condition. Moreover, the contents of 4-deoxyorobanchol, one of canonical SLs in rice, in root exudates of OsHTD1aus L1 and L2 were also markedly reduced (Figure 1k), and their tiller numbers were significantly increased compared to WT (Figure 1l). These results demonstrated that the designed AUS in OsSBIaus and OsHTD1aus could upregulate or downregulate in vivo protein abundance to acquire desired traits. In summary, we showed that the rational design of AUS in OsSBI and OsHTD1 could create beneficial alleles to obtain Green Revolution traits in rice (Figure 1m). This approach may offer some advantages in crop breeding. First, endogenous genetic elements are harnessed to manipulate gene expression levels without exogenously artificial transgene. Second, this approach has been developed to regulate protein levels by merely inserting a few nucleotides in the AUS. Third, this strategy might be applied to a large number of genes by rational design of their AUSs. Using the dual-luciferase assays, the AUSs can be rapidly screened and characterized for certain genes to obtain suitable AUSs with the desired regulatory effects. The engineered alleles of OsSBI and OsHTD1 with 3-bp insertion in AUS recapitulated Green Revolution traits, providing a convenient and practical tool in crop breeding. We thank Prof. Caixia Gao for pH-nCas9-PPE-V2 vector and insightful discussions, and Prof. Weichang Yu for ccd7c1 seeds. This work was supported by CAS Project for Young Scientists in Basic Research (YSBR-078), the National Natural Science Foundation of China (32388201, 32122064), and the Hainan Excellent Talent Team. The authors have declared no conflict of interest. H.W., H.Y., and J.L. designed the experiments and wrote the paper. H.W., M.C., D.Z., X.M., J.Y., and J.C. performed the experiments and analysed the results. H.Y. and J.L. supervised the project. All authors approved the final manuscript. The data that supports the findings of this study are available in the supplementary material of this article. Data S1 Materials and Methods. Figure S1 LUC/REN mRNA and activity in dual-luciferase assays of different AUS constructs for OsSBI and OsHTD1 in rice protoplasts. Figure S2 pegRNA design for OsSBI. The detailed sequence of peOsSBI pegRNA. Figure S3 Genotypes of the OsSBIAUS ePE-edited lines. Figure S4 mRNA stability of OsSBI in WT, OsSBIaus L1 and L2. Figure S5 pegRNA design for OsHTD1. The detailed sequence of peOsHTD1 pegRNA. Figure S6 Genotypes of the OsHTD1AUS ePE-edited lines. Figure S7 mRNA stability of OsHTD1 in WT, OsHTD1aus L1 and L2. Figure S8 ePE construct map. Figure S9 Specificity determination of OsSBI antibody by protein extracts from rice protoplasts. Figure S10 Specificity determination of OsHTD1 antibodies by protein extracts from rice plants. Table S1 Primers used in this study. Table S2 Analysis of potential off-target sites for each target. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts