Zoysiagrass exhibits high morphological, physiological, and genetic diversity, and its ability for interspecific hybridization has led to the development of numerous varieties with valuable traits (Chandra et al. 2017; Patton et al. 2017). However, due to the absence of the evergreen trait within its diverse genetic resources, conventional breeding methods could not facilitate the development of stay-green varieties. Our research group had previously selected Z. matrella ‘Wakaba’ as the superior ecotype from 250 eco-collections for turf quality in Japan and subsequently developed it as a cultivar (Hashiguchi et al. 2006; Tanaka et al. 2016b). We also generated a high-quality draft genome sequence of Z. japonica, and draft genome sequences of Z. matrella and Z. pacifica, with 'Wakaba' serving as the representative material for Z. matrella (Tanaka et al. 2016a). Additionally, by leveraging the experience of our research group in establishing tissue culture systems for various warm-season grasses (Muguerza et al. 2022), we have established an efficient transformation and tissue culture system using this cultivar (Ng et al. 2022). These collective achievements have laid the foundation for molecular breeding within the Zoysia genus, leading to our success in this study by conducting genome editing for the first time in zoysiagrass, targeting the NYC1 gene responsible for chlorophyll degradation in Z. matrella 'Wakaba'.
Achieving a complete knockout of ZmNYC1 proved challenging due to the highly heterozygous nature of the allotetraploid Z. matrella ‘Wakaba’. The ZmNYC1 loci consist of two sub-genomes, suggested to be derived from Z. japonica (A sub-genome) and Z. pacifica (B sub-genome), respectively (Tanaka et al. 2016b, a). We designed a gRNA targeting both A and B sub-genomes and employed the all-in-one vector containing MMCas9 and OsU6-gRNA expression constructs (pZH_OsU6gRNA_MMCas9), given its demonstrated effectiveness in inducing high-frequency mutagenesis in target genes as confirmed in rice and wheat (Mikami et al. 2015; Abe et al. 2019). Despite the relatively low transformation efficiency, we found it to be adequately sufficient for generating genome-edited T0 plants with multiple genotypes in zoysiagrass. Through this approach, we successfully achieved efficient targeted mutagenesis and accomplished the knockout of both sub-genomes of the ZmNYC1 gene in the T0 generation.
Sequencing analysis of some transformed calli revealed a wide range of mutations, mostly consisting of 1 or 2 bp indels. T0 plants from the same callus lines exhibited a variety of mutations, with some plants carrying mutations not initially detected during sequencing analysis of the initial callus line. This phenomenon of enhanced mutation frequency with prolonged culture duration, as seen in rice (Mikami et al. 2015), likely resulted from the continuous subculture of transformed calli prior to plant regeneration, which extended the exposure to CRISPR/Cas9 and contributed to additional mutations. Furthermore, several T0 plants displayed chimeric mutations in both A and B sub-genomes. New mutations might have occurred during the transition from somatic embryo to plant regeneration, possibly due to the rapid cell division of somatic embryo development, which could have promoted the active expression of Cas9/gRNA. Other than that, the complexity of the Z. matrella genome, with multiple gene copies, might have contributed to unequal editing by the CRISPR/Cas9 system, as observed in other CRISPR/Cas9-edited polyploidy crops (Ahmad et al. 2023). Consequently, it is necessary to identify the mutation patterns to confirm the genotypes of each regenerated plant.
In this study, the genome-edited ZmNYC1 in Z. matrella displayed stay-green phenotypes following dark-induced senescence. Moreover, the T0 plants with the aabb mutant genotype showed a visually darker green color than the WT even before treatments showcasing a superior stay-green trait. These results are consistent with previous studies that demonstrated mutants of the NYC1 gene in Arabidopsis (Jibran et al. 2015) and rice (Kusaba et al. 2007) retained greenness in leaves after induced senescence by dark treatment. The ZmNYC1 T0 lines also showed higher levels of total chlorophyll, Chl a, and particularly Chl b, after dark treatment, leading to a lower Chl a/Chl b ratio, consistent with the rice NYC1 mutant (Kusaba et al. 2007). These results suggest that the targeted mutations effectively disrupted the functionality of the ZmNYC1 gene. The Fv/Fm of ZmNYC1 mutant lines declined similarly to the WT during dark treatment. Higher chlorophyll content in NYC1 mutant does not necessarily imply a higher photosynthesis ability, as NYC1 is involved in chlorophyll degradation, not leaf longevity (Kusaba et al. 2007). Hence, the ZmNYC1 mutant is classified as a cosmetic stay-green mutant, retaining green leaf color while exhibiting a loss in photosynthetic capacity (Thomas and Howarth 2000).
In addition to exhibiting a stay-green phenotype, the ZmNYC1 mutation also inhibited plant growth in homozygous mutant genotypes, leading to reduced tiller numbers (Fig. S3). This contrasts with findings in other plant species with NYC1 mutants, as no compromised yield or fitness was observed except for reduced seed germination in the nyc1/nol Arabidopsis mutant (Nakajima et al. 2012; Jibran et al. 2015). This result might be associated with the complex regulation of the chlorophyll cycle, responsible for the interconversion between Chl a and Chl b, which is crucial for maintaining the balance of the Chl a/Chl b ratio under different physiological conditions (Ito et al. 1996). In this study, the Chl a/Chl b ratio for WT was generally above 2.5 before treatment. However, a lower Chl a/Chl b ratio became more pronounced with an increased number of mutated alleles, correlating with the observed reduction in tiller growth. This indicates the importance of the chlorophyll cycle in regulating the Chl a/Chl b ratio, an important determinant of light absorption efficiency of photosynthesis and its overall impact on plant growth. Additionally, mutants with suppressed Chl b degradation may hinder the degradation of LHCII, which is the most abundant membrane protein, potentially disrupting the efficient allocation of nitrogen and carbon to sink organs (Kusaba et al. 2007; Horie et al. 2009). Moreover, accumulated LHCII in these mutant genotypes may result in increased light capture but cannot fully utilize the absorbed light energy for photosynthesis because of the reduced amount of the PSII core complexes (Kusaba et al. 2007). As a result, the Fv/Fm ratio, which serves as an indicator of maximum photochemical efficiency of PSII, was significantly lower in the aabb mutant genotype compared to the WT, indicating reduced photosynthetic efficiency. Furthermore, leaves of this mutant genotype may have been causing cell death as a result of reactive oxygen species (ROS) generated by excess chlorophyll (Hu et al. 2021). Among the four mutant genotypes observed, the complete homozygous mutant exhibited the lowest growth, whereas the AAbb and Aabb mutant genotypes with the homozygous sub-genome B showed moderate greenness and growth between the complete knockout mutant and the WT. Consequently, a moderately ZmNYC1-edited mutant that retains a stay-green phenotype without adversely affecting growth is likely a more suitable candidate for practical applications.
Under simulated winter conditions, the aabb mutant genotype exhibited better preservation of greenness, with significantly higher chlorophyll content, particularly Chl b, compared to the WT. This highlights the potential of the ZmNYC1 knockout approach to mitigate the reduction in greenness during winter dormancy in zoysiagrass. Nevertheless, the aabb mutant genotype displayed gradual chlorophyll decline, indicating the potential upregulation of other genes involved in photosynthesis, carbon assimilation, ROS production, and hormone synthesis under cold conditions and reduced photoperiods (Teng et al. 2016, 2021). This observation aligns with a recent RNA-seq analysis of zoysiagrass under cold stress, which revealed the down-regulation of genes associated with chlorophyll biosynthesis and chlorophyll a/b binding, alongside the upregulation of six genes related to chlorophyll catabolism (Wei et al. 2015; Long et al. 2020). The combination of low temperature with high light has the potential to induce chronic photoinhibition of PSII in warm-season plants (Allen and Ort 2001). This inhibition was observed in both WT and line 19 − 1 mutant under simulated winter conditions, as the Fv/Fm values were zero after one week. While the ZmNYC1 mutant shows promise in maintaining a green color during winter, we consider it is also essential to identify the key genes directly involved in winter dormancy. This first established genome editing technology for Z. matrella can be effectively applied to the functional analysis of novel genes and the development of new varieties of zoysiagrass in the future.
In conclusion, we successfully applied the CRISPR/Cas9 system in the highly heterozygous allotetraploid Z. matrella ‘Wakaba’ using our previously established stable Agrobacterium transformation system (Ng et al. 2022). Specifically, we targeted the ZmNYC1 loci and obtained a total knockout of all four ZmNYC1 alleles in the T0 generation. These mutants exhibited a longer stay-green phenotype than the WT under dark-induced and winter simulation treatments. We believe that the mutants developed in this study hold the promise of being utilized as novel breeding materials capable of maintaining greenness for extended periods during winter. In the future, off-target screening and the removal of transgenes by self-pollination will enable breeding development with a view to practical application.