Highly efficient, genotype‐independent transformation and gene editing in watermelon (Citrullus lanatus) using a chimeric ClGRF4‐GIF1 gene

INTRODUCTION

In the past few decades, genetic transformation has proven to be an essential tool for characterizing gene function in plants and even more so for improving agronomic traits in crops (Gelvin, 2003). Watermelon (Citrullus lanatus) is the fifth most consumed fresh fruit in the world, favored by consumers for its fresh taste and nutritional content (Guo et al., 20132019). However, low regeneration rates have hindered watermelon transformation for years. Many different explant sources have been used for watermelon transformation, including the stem tip, shoot meristem, immature embryo, cotyledon, cotyledon node, hypocotyl, and protoplast, but these approaches have achieved only relatively low genetic transformation efficiencies (Choi et al., 1994; Ellul et al., 2003Tian et al., 2017). Moreover, as in many other plant species, watermelon genetic transformation is highly genotype dependent, with only a few cultivars or inbred lines having been successfully transformed (Tian et al., 20172018Ren et al., 2018).

Recent work showed that expression of a fusion protein made up of two interacting transcription factors, GROWTH-REGULATING FACTOR4 (GRF4) and GRF-INTERACTING FACTOR1 (GIF1), greatly improved transformation efficiency in wheat (Triticum aestivum), citrus (Citrus sp.), grape (Vitis vinifera), and hemp (Cannabis sativa) (Debernardi et al., 2020Zhang et al., 2021) thanks to its ability to remodel chromatin and activate genes (Kim, 2019). In this study, we established a genotype-independent watermelon transformation method with high efficiency by overexpressing a chimeric watermelon ClGRF4-GIF gene during transformation. Our method will greatly accelerate watermelon research and molecular breeding in the future.

RESULTS AND DISCUSSION

To develop a transformation system for watermelon, first we screened different genotypes for regeneration ability. Among those tested, the wild watermelon variety “TC” displayed the highest regeneration rate. Following a conventional Agrobacterium tumefaciens-mediated transformation protocol (Tian et al., 2017), we successfully transformed “TC” using cotyledons as explant material, achieving a 6.5% (n = 465) transformation efficiency.

Based on recent work showing that the chimeric fusion protein GRF4-GIF1 improved regeneration in multiple plant species (Debernardi et al., 2020Zhang et al., 2021), we next tested whether a watermelon GRF4-GIF1 could improve transformation in watermelon. To this end, the endogenous watermelon genes ClGRF4 (Cla97C02G034420) and ClGIF1 (Cla97C02G042620), linked by a 4x Ala linker (ClGRF4-4xAla-ClGIF1) and driven by a 35S promoter, were cloned into a pCambia1300 backbone vector along with green fluorescent protein (GFP) and a hygromycin selectable marker to make the final construct termed pGH-GRF-GIF (Figures 1AS1). We transformed “TC” watermelon with pGH-GRF-GIF, and positive transgenic shoots were regenerated with an efficiency of 47.02% (positive shoots/explants). This represents an almost 9-fold improvement compared to the 5.23% efficiency achieved with the control vector pGH, containing only the GFP and hygromycin cassettes driven by 35S promoters (Figure 1C, D; Table S1).

image

Efficient genotype-independent genetic transformation in watermelon by overexpressing a chimeric ClGRF4-GIF1 gene

(A) Schematic diagram of the vectors used in this study. (B) The sequence of the miR396 targeting site inside ClGRF4 before and after mutation. (C) Transformation efficiencies achieved using different constructs in “TC” watermelon. (D) Fluorescent and bright-field images of regenerated transformed shoots during “TC” transformation. Bar = 25 mm. (E) Non-transformable watermelon “97103” can be transformed in the presence of ClrGRF-GIF. Fluorescent and bright-field images indicated a positive shoot from a “97103” cotyledon explant. Bar = 3 mm. (F) Average transformation efficiencies achieved using pGH-rGRF-GIF and pGH vectors during transformation. All tested materials were transformable with the chimeric ClrGRF-GIF gene. (G) Genome editing by combining rGRF-GIF with the clustered regularly interspaced palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) system. Top panel, gene editing efficiencies achieved when LEC1 and LEC2 were mutated either individually or simultaneously. Middle panel, several sample gene edits of the T1 plants at two target sites in the LEC1 gene; Bottom panel, results from Sanger sequencing of one target site in the wild type and one editing line. (H) Fluorescent and bright-field images of a transgenic plant and diploid seedless watermelon obtained after mutating LEC1 and LEC2 simultaneously. Bars = 5 cm.

The previous study showed that mutating the potential mi396 microRNA targeting site inside GRF4 further improved transformation efficiency in wheat (Debernardi et al., 2020). Therefore, we employed the same strategy using our system, designing a new vector called pGH-rGRF-GIF (Figure 1AB). We transformed “TC” watermelon with the pGH-GRF-GIF, pGH-rGRF-GIF, and pGH vectors to compare efficiencies (Figure 1A). Transforming with the vector pGH-rGRF-GIF further improved regeneration efficiency to 67.27%, which is about 12-fold higher than the transformation efficiency achieved without addition of the chimeric gene (Figure 1C, D; Table S1).

Extremely high regeneration efficiency achieved using rGRF-GIF in “TC” inspired us to test its efficacy in additional watermelon genotypes, including non-transformable watermelon varieties. We chose eight watermelon genotypes for additional testing, namely, “YL”, “M08”, “148”, “M57”, “M20”, “97103”, “WY”, and “BJ”. Strikingly, all tested genotypes were transformable using rGRF-GIF, with regeneration rates ranging from 19.64% (M57) to 61.84% (148) (Figures 1E, F, S2; Table S2). Notably, non-transformable or nearly non-transformable genotypes, including “M57”, “M20”, “97103”, “WY”, and “BJ”, achieved regeneration efficiencies of 19.64%, 40.94%, 44.44%, 21.9%, and 48.55%, respectively (Figures 1E, F, S2; Table S2). These results indicate that use of the chimeric rGRF-GIF transgene may allow for truly genotype-independent transformation in watermelon. Healthy plants were recovered from regenerated shoots, and all of them were confirmed to be true transgenic plants by the presence of both the GFP fluorescent marker and polymerase chain reaction-verified transgenic fragments (Figures S3S4). Importantly, we only used GFP fluorescence to inspect positive transformants and did not used the antibiotic hygromycin for selection to avoid external selective pressure on newly transformed cells. We suspected that it would be much easier to obtain positive transgenic events if no antibiotic selection was applied in the ClrGRF-GIF transformation system, especially with the non-transformable materials. No obvious defects were observed during plant vegetative growth (Figure S5). Considering the reference genome for watermelon is from the variety “97103” and the cultivars “M08”, “148”, “M57”, and “M20” are all elite inbred lines, our successful transformation of these genotypes significantly advances watermelon research and breeding (Figures 1E, FS2–S4; Table S2).

Genome editing is a powerful tool for research and molecular breeding in plants (Xing et al., 2014). We combined rGRF-GIF with the clustered regularly interspaced palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) system to generate the vector pGH-rGRF-GIF-CR (Figure 1A). We chose two genes, ClLEC1 (Cla97C10G188900) and ClLEC2 (Cla97C01G022960), which are critical for watermelon reproductive development to test editing efficiency. Two guide RNAs were designed for each gene, and “YL” was used as the explant genotype. We achieved 68.75% and 50.00% editing efficiency for LEC1 and LEC2, respectively (Figure 1G). All edited plants were confirmed by Sanger sequencing and next-generation deep sequencing (Figure 1G). We also tested the feasibility of editing both genes at the same time. Four guide RNAs, two for each gene, were assembled into the pGH-rGRF-GIF-CR vector via the tRNA flanking system (Xie et al., 2015). We showed that 54.55% of transgenic plants showed edits in both genes (Figure 1G). Because these target genes are involved in plant reproduction, double mutant plants showed severe reproductive defects and could not develop normal seeds after pollination. However, because watermelon fruits can enlarge normally after pollination stimulation, we successfully created seedless watermelon at the diploid level by disrupting essential reproductive genes using our newly established watermelon transformation system (Figure 1H).

Although many efforts have been made in pursuit of successful genetic transformation in different plant species, to date only a limited number of species can be genetically transformed. Watermelon, an important summer fruit crop, could not be efficiently transformed due to a low regeneration rate. Here, we have established a highly efficient and genotype-independent transformation method by overexpressing the chimeric watermelon ClGRF-GIF gene (Figures 1C‒FS2‒S4; Tables S1S2). Using our system, the highest efficiency achieved in a single transformation was 86.08% in the “TC” genotype after mutating the mi396 microRNA target site inside ClGRF (Figure 1C; Table S1). By combining rGRF-GIF with the CRISPR/Cas9 system, we were able to efficiently edit both tested genes individually and simultaneously in watermelon. We also used this efficient genome editing tool to create seedless watermelon at the diploid level by disrupting two important reproductive genes, ClLEC1 and ClLEC2 (Figure 1G, H).

Recently, Zhang's group from Peking University showed that the Arabidopsis thaliana gene AtGRF improves watermelon transformation efficiency to about 28%, while the chimeric TaGRF4-OsGIF1 increases efficiency to 13.78% (Pan et al., 2021). However, in our study we fused endogenous genes to generate the chimeric ClGRF-GIF, and we mutated the mi396 target site inside the chimeric gene to further boost transformation efficiency. The transformation efficiencies achieved using our method are more than sufficient for research and molecular breeding purposes, and we hope the powerful genetic transformation tool presented in this work will benefit future watermelon research and breeding.

ACKNOWLEDGEMENTS

This work was supported by the National Youth Talent Program (A279021801), the Fundamental Research Fund from Northwest A&F University (Z1090221008) and the Key R&D Project from Yangling Seed Industry Innovation Center (2021).

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

AUTHOR CONTRIBUTIONS

L.Y. conceived the study and designed the overall experimental scheme. Q.F., L.X. and Y.Z.H. executed the experiments. M.L., J.F.W., S.J.T. and X.Z. provided technical suggestions and supervised the study. L.Y. wrote the manuscript. All authors read and approved the final manuscript.

留言 (0)

沒有登入
gif