Efficient and transformation‐free genome editing in pepper enabled by RNA virus‐mediated delivery of CRISPR/Cas9

Abstract

Tomato spotted wilt virus-mediated delivery of CRISPR/Cas9 bypasses the need for stable transformation and permits efficient, DNA-free genome editing in pepper. Remarkably, up to 77.9% of regenerated pepper plants contained heritable edits. This method has been validated with two pepper varieties and is compatible with existing tissue culture protocols. Pepper (Capsicum spp.), one of the most important vegetable crops cultivated worldwide, imparts a savory depth of flavor and spiciness. Additionally, pepper fruits possess rich natural pigments, nutritional compounds, and bioactive metabolites, offering significant ornamental, industrial, and medicinal values (Kothari et al., 2010). Despite its economic importance, progress in Capsicum functional genomics and genetic improvement lags behind other major members of the Solanaceae family. While recent advancements in whole-genome sequencing and forward genetics approaches (e.g., positional cloning, quantitative trait loci mapping, and genome-wide association study) have provided insights into the genetic basis underlying important agronomic traits in pepper (Kim et al., 2014; Qin et al., 2014; Chunthawodtiporn et al., 2018; Wu et al., 2019; Cao et al., 2022), functional dissection of key regulatory genes has been severely hindered by its notorious recalcitrance to genetic transformation (Kothari et al., 2010). As an alternative, viral vector-mediated transient RNA silencing and overexpression have been employed for gene function validation in pepper (Choi et al., 2019; Zhou et al., 2021; Wang et al., 2024). However, these strategies are unable to introduce defined genetic mutations and have limited utility in trait improvement. The recently emerged clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated nuclease (Cas) genome-editing technologies offer vast opportunities for plant functional genomics research and genetic improvement. Unfortunately, the adoption of these tools in pepper studies is greatly confounded by the reliance of CRISPR/Cas delivery on stable transformation. Although poor morphogenic responses of pepper explants are commonly considered the major bottleneck in transformation, processes including efficient DNA transfer/integration and the selection of regeneration-competent recipient cells are also crucial for the successful recovery of transgenic plants (Kothari et al., 2010). Meanwhile, significant progress has been made in refining pepper regeneration protocols using various varieties, explants, and tissue culture media (Kothari et al., 2010). To circumvent the need for stable transformation, we recently developed a transient CRISPR/Cas delivery system based on engineered tomato spotted wilt virus (TSWV) vectors, enabling efficient somatic gene editing in several crop species, including pepper (Liu et al., 2023). Although this previous study has demonstrated that heritable editing could be recovered by in vitro culturing of infected tissues of tobacco and tomato plants, two crop species that are readily amenable to genetic transformation, it remains unclear whether recalcitrant crops like pepper could benefit from the viral delivery technology. Here, using the widely cultivated pepper species C. annuum as an example, we investigate the feasibility and efficiency of recovering gene-edited pepper plants and the transmission pattern of TSWV-induced mutations from regenerated plants to the next generation. Two varieties of C. annuum were tested: the whole-genome sequenced “Zunla-1” (Qin et al., 2014) and “Hangjiao-9”, a commercial variety differing in fruit shape and pungency. Zunla-1 seedlings were infected with a TSWV vector containing the Cas9 gene and a gRNA targeting the PDS-3 locus of the pepper phytoene desaturase (PDS) gene (abbreviated V-Cas9-gPDS-3), whereas a similar vector targeting the PDS-4 locus (V-Cas9-gPDS-4) was generated to infect Hangjiao-9 plants (Figure 1A). Following systemic virus infections, high-throughput sequencing (HTS) of target region amplicons revealed somatic editing frequencies of 57.65% and 75.73% at the PDS-3 and PDS-4 sites, respectively, in the upper non-inoculated leaf tissues at 12 d post-inoculation (Figure 1B). Young, symptomatic leaf tissues from infected plants were used as explants for tissue culture in non-selective medium to regenerate pepper plants with targeted mutations (Method S1). Numerous pepper shoots were regenerated from induced calli, some showing albino or mosaic/chimeric phenotypes indicative of full or partial disruption of PDS gene functions (Figure 1D). Quantitative analyses of the percentage of explants bearing induced shoots showed that TSWV vector-infected leaf explants exhibited a moderate reduction in regeneration efficiency compared to uninfected ones, although this reduction was statistically insignificant (Figure 1C). Nearly all green and some mosaic shoots were well-rooted and fertile, whereas albino shoots failed to develop roots (Figures 1D, S1A). For both varieties, the tissue culture process takes approximately 2.5 to 4.0 months. Pepper genome editing with tomato spotted wilt virus (TSWV)-mediated delivery of clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated nuclease 9 (Cas9) (A) Diagrams of TSWV CRISPR/Cas vectors, phytoene desaturase (PDS) gene structure, and target site sequences. Blue bars, horizontal lines, vertical red lines denote PDS exon, intron, and target sites, respectively. Protospacer adjacent motifs (PAMs) are shown in red. (B) Somatic editing frequency at two target sites in pepper leaf tissues systemically infected with the TSWV vectors. (C) Regeneration efficiencies of uninfected versus the TSWV vector-infected pepper leaf explants. NS, not statistically significant (two-tailed Student's t-test). Data are shown as individual data points and means ± SD. (D) Images showing various stages of plant regeneration from leaf tissues of two pepper varieties infected with the TSWV vectors. From left to right are leaf explants at d 0, callus induced at d 30, elongating shoots at d 60, rooting plantlet at d 80, and plants transplanted in soil at d 200 (upper) and 141 (bottom). White and red scale bars are 1 cm and 10 cm. (E–F) Percentages of the regenerated M0 plants with different phenotypes (E) and genotypes (F). Mo, mono-allelic; Bi/Ho, bi-allelic/homozygous; Ch, chimeric; WT, wild-type. (G) Target site mutations in representative pepper regenerants. PAM sequences are highlighted in red, and numbers in brackets are the percentages of reads for each allele. d# and i# denote deletions and insertion of # nucleotides, respectively. The “d3” deletion type in the PDS-3 target site, which fails to disrupt PDS function, is labeled in green letters. (H) Phenotypes of M1 progeny derived from representative pepper regenerants. Scale bars, 1 cm. (I) Statistical analyses of the phenotype segregation ratios of M1 progeny. P-values were calculated using a Chi-squared test of goodness-of-fit to a segregation ratio of 3 (green): 1 (albino). 0.1 < P < 0.5, in good agreement with theoretical segregation ratios; P > 0.5, in very good agreement with theoretical segregation ratios. Mo, mono-allelic; Bi/Ho, bi-allelic/homozygous; Chi, chimeric; WT, wild-type. Regenerated plantlets did not show discernible virus symptoms during tissue culture stages (Figures 1D, S1A), although most tested positive for TSWV by reverse transcription – polymerase chain reaction (RT-PCR) (Figure S1B; Tables S1, S2). However, after transplantation into soil, many plants initially exhibited transient viral symptoms but later spontaneously recovered from the infection and showed normal fecundity (Figure S1C). This differs from Nicotiana benthamiana plants regenerated from TSWV-infected tissues, which displayed severe symptoms persistently and produced few seeds (Liu et al., 2023). Among the regenerated plantlets (M0 generation) analyzed, 53 (61.6%), 11 (12.8%), and 22 (25.6%) out of 86 Hangjiao-9 plantlets, and 5 (6.3%), 2 (2.5%), and 73 (91.2%) out of 80 Zunla-1 plantlets, exhibited albino, mosaic, and normal phenotypes, respectively (Figure 1E; Table S1, S2). Genotyping of these regenerants by HTS revealed that among the 22 green Hangjiao-9 regenerants, three carried mono-allelic mutations at the PDS-4 site, while the remaining 19 were wild-type. Among the 73 green Zunla-1 regenerants, seven contained bi-allelic mutations, including a non-frameshifting allele with a three-nucleotide deletion (d3), and 18 and six plants were mono-allelic and chimeric at the PDS-3 site, respectively (Figure 1F; Table S1, S2; Dataset S1). As expected, all albino plants of both varieties harbored exclusively mutated PDS alleles, while all mosaic lines indeed carried chimeric mutations at their target sites (Figure 1G; Table S1, S2; Dataset S1). In total, 77.9% of Hangjiao-9 and 47.5% of Zunla-1 regenerants carried targeted mutations (Figure 1F), which are proportional to the somatic editing frequencies observed in virus-infected plants (Figure 1B). Interestingly, although most albino plantlets contained bi-allelic or homozygous mutations at the PDS target sites, some had more than two types of mutated alleles (e.g., ZL-#12, and HJ-#1/-#2/-#3; Table S1, S2). In the latter cases, the targeted editing might have occurred after the first division of the founder cell during tissue culture, or it is possible that the regenerants originated from multiple edited cells. Eighteen pepper regenerants with different mutation types, nine for each variety, were randomly selected for analyzing their DNA content by flow cytometry. The data revealed that all the regenerants were diploid plants (Figure S2), ruling out the possibility of genome doubling after tissue culture. Next, we examined the transmission of mutations from M0 to the next generation (M1). Initially, we focused on analyzing representative Zunla-1 green regenerants due to their viability and fertility. Two M0 lines (ZL-#30/-#59), each harboring bi-allelic mutations including a d3-type allele, and three lines with mono-allelic mutations (ZL-#32/-#58/-#69), were selected for self-pollination, and the phenotypes of their M1 progeny were assessed. All M0 lines yielded albino offspring in a ratio approaching 3:1, consistent with Mendel's law of segregation (Figure 1H, I). Genotyping of randomly selected M1 progeny by HTS revealed the faithful transmission of the mutations (Table S3). Additionally, we also examined several mosaic M0 lines (HJ-#50/-#74, and ZL-#73) to determine the inheritance of the mutations. These regenerants exhibited extensive albino tissues when cultured in heterotrophic culture media; however, upon transplantation into soil for photoautotrophic growth, the plants gradually transitioned to nearly normal green, with the albino tissues fading out over time (Figure S3). This highlights the remarkable developmental plasticity of plants. Despite being phenotypically normal at late stages of growth, these mosaic M0 lines also produced a substantial number of albino offspring, albeit with segregation ratios potentially deviating from the classic Mendelian laws (Figure 1H, I). Further genotyping of M1 offspring revealed that certain types of mutated alleles in the mosaic lines, while not necessarily the predominant alleles (with the highest percentage of reads as shown in Figure 1G), were germline transmissible (Tables S3, S4). All analyzed M1 offspring tested negative for the viral vector (Figure S4). Taken together, these findings demonstrate that the bi-allelic/homozygous, mono-allelic, and at least some chimeric mutations in the regenerated plants are germline transmissible. In summary, we have successfully developed a highly efficient method for obtaining gene-edited, transgene-free pepper plants. This method circumvents the need for stable transformation and is compatible with established pepper tissue culture protocols, yielding up to 77.9% of regenerated plants with heritable mutations. These results suggest that circumventing the bottleneck of stable transformation mitigates inherent challenges associated with selecting totipotent recipient cells and subsequent in vitro regeneration of these cells into fertile plants. Our work not only paves the way for pepper functional genomics studies and precise trait improvement but also illuminates a potential pathway for genome editing in other crop species and varieties that are recalcitrant to genetic transformation. We thank Dr. Xiaolin Yu (Zhejiang University, China) for providing the Hangjiao-9 pepper seeds and suggestions for pepper tissue cultures and Dr. Changshui Yu (Zunyi Agricultural Sciences and Technology Research Institute, Guizhou, China) for sharing the Zunla-1 pepper seeds. This work was supported by the National Key R&D Program of China grant numbers 2023YFD1400300 and 2022YFC2601000 (to Z.L.) and the Postdoctoral Fellowship Program of CPSF under grant number GZC20241527 (to C.Z.). The authors declare no conflict of interest. Z.L. and C.Z. designed the research. C.Z., H.L., Q.L. and S.P. performed the experiments. Z.L., C.Z. and Q.L. analyzed the data. Z.L. and C.Z. wrote the manuscript. All authors read and approved the contents of this paper. All the data are included in this published article and its supporting information. Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.13741/suppinfo Figure S1. Viral symptoms in regenerated pepper plants Figure S2. Ploidy analysis of representative M0 regenerants by flow cytometry Figure S3. Phenotypes of two mosaic Hangjiao-9 M0 lines during growth Figure S4. Detection of tomato spotted wilt virus in M1 offspring by reverse transcription – polymerase chain reaction Method S1. Materials and methods Table S1. Genotype and phenotype of M0 plants regenerated from Zunla-1 leaf tissues infected with V-Cas9-gPDS-3 Table S2. Genotype and phenotype of M0 plants regenerated from Hangjiao-9 leaf tissues infected with V-Cas9-gPDS-4 Table S3. Genotype and phenotype of M1 progeny derived from representative Zunla-1 M0 lines Table S4. Genotype and phenotype of M1 progeny derived from two mosaic Hangjiao-9 M0 lines Table S5. Oligos used in this study 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