Core

Our analysis found that the large-seeded cultivar 'HZ' seeds had a higher DNA methylation level than the abortive-seeded cultivar 'NMC' seeds. 'NMC' seeds had significantly more ROS than 'HZ' seeds, and the gene LcGPX6, involved in ROS scavenging, had higher DNA methylation and lower expression than that in 'HZ' seeds, suggesting that DNA methylation-mediated ROS production plays a role in seed development, with higher DNA methylation levels of LcGPX6 suppressing its expression and leading to excessive ROS accumulation and seed abortion.

Gene and accession numbers

Sequence data from this article can be found in the litchi genome database (https://doi.org/10.1101/2022.11.25.517904) under the accession numbers: LcGPX6: LITCHI022143, LcGPX1/4/7: LITCHI016878, LcGPX2: LITCHI015847, LcGPX3: LITCHI015847, LcGPX5: LITCHI026281, LcGPX8:LITCHI022145, LcCAT: LITCHI009991, LcAPX1: LITCHI024768, LcAPX2: LITCHI018535, LcAPX3: LITCHI002021, LcAPX4: LITCHI010566, LcAPX5: LITCHI006437, LcAPX6: LITCHI016989, LctAPX/Sapx: LITCHI011263, LcCSD1: LITCHI003782, LcCSD2: LITCHI024579, LcCSD3: LITCHI003354, LcFSD1: LITCHI020212, LcFSD3: LITCHI006009, LcMSD1: LITCHI027490, LcPrxR A/B: LITCHI018945, LcPrxR F: LITCHI014732, LcPrxR Q: LITCHI017484, LcType 2-PrxR B/C/D: LITCHI016784, LcType 2-PrxR E: LITCHI018774, LcRbohA/C: LITCHI020220, LcRbohB: LITCHI030320, LcRbohD: LITCHI021682, LcRbohE: LITCHI017482, LcRbohF/I: LITCHI022022, LcRbohH/J: LITCHI024802, LcRbohG: LITCHI030338.

Introduction

Seeds serve as the primary source of nutrients for both humans and animals, while also playing a crucial role in ensuring offspring. Therefore, comprehending the intricate mechanisms underlying seed development holds paramount importance in enhancing agricultural practices and effectively managing genetic resources. In the majority of angiosperms, seeds are formed through the process of double fertilization, which culminates in the production of a mature seed comprising the embryo, endosperm, and seed coat.

Recent studies have shed light on the crucial role of epigenetic modifications in seed development. One such modification is DNA methylation, which can modulate chromatin structure and function, thereby influencing the silencing of transposable elements (TEs) and gene expression (Buitrago et al. 2021). In plants, DNA methylation occurs in different sequence contexts, including CG, CHG, and CHH (where H represents A, C, or T). The maintenance of CG methylation is regulated by ETHYLTRANSFERASE 1 (MET1), while CHG methylation is maintained by CHROMOMETHYLASE 2 (CMT2) and CHROMOMETHYLASE 3 (CMT3). Furthermore, CHH methylation is maintained by either CMT2 or DOMAINS REARRANGED METHYLTRANSFERASE 2 (DRM2) (Gallego-Bartolome 2020). Perturbations in DNA methylation, such as a global loss of CG methylation in Arabidopsis, have been associated with abnormal embryo development and impaired megaspore mother cell development (FitzGerald et al. 2008; Li et al. 2017). Additionally, disruptions in non-CG methylation in maize have been shown to cause severe defects in ovule development (Garcia-Aguilar et al. 2010). Similarly, reproductive defects have been observed in RdDM mutants in tomato and Brassica rapa, with CHH methylation playing a role in chickpea seed development (Gouil and Baulcombe 2016; Grover et al. 2018; Rajkumar et al. 2020). However, loss of non-CG methylation has no effect on seed development in Arabidopsis and soybean (Lin et al. 2017). Although it has been demonstrated that DNA methylation plays critical roles in seed development, the precise mechanisms through which it regulates this process remain poorly understood.

Reactive oxygen species (ROS) have been identified as both toxic byproducts of aerobic metabolism and crucial regulators of development, including seed development, in plants. ROS can be generated through various enzymatic activities, with NADPH oxidases (Rboh) being extensively studied in this context. To maintain ROS homeostasis, plants possess ROS-scavenging enzymes including superoxide dismutase (SOD), ascorbate peroxidase (APX), catalase (CAT), glutathione peroxidase (GPX), and peroxiredoxin (PrxR) (Apel and Hirt 2004; Mittler et al. 2004). In Arabidopsis, it has been demonstrated that maintaining ROS homeostasis during female gametophyte development is crucial for proper embryo sac patterning and fertilization. ROS are detected in the nucellus during megasporogenesis and the central cell of the embryo sac during megagametogenesis. Pollination leads to an oxidative burst, after which ROS are cleared from the embryo sac (Martin et al. 2013). MnSOD (MSD) has been identified as a pivotal protein that regulates ROS levels during female gametogenesis. Mutations in MSD, such as the oiwa mutant, disrupt ROS homeostasis, resulting in high ROS levels in the embryo sac and leading to sterility or arrested embryogenesis (Victoria Martin et al. 2013). Recently, there have been reports highlighting the crucial roles of DNA methylation-mediated ROS homeostasis in various aspects of plant development and stress response, including salt stress (Chen et al. 2015; Hu et al. 2021), heat stress(Ma et al. 2018; Sakai et al. 2022; Zhu et al. 2021), chilling/freezing stress(Guo et al. 2019; Zheng et al. 2022), and fruit ripening (He et al. 5). Therefore, we propose that DNA methylation-mediated excessive ROS accumulation in 'NMC' disrupts intracellular ROS homeostasis during early embryo development, ultimately leading to seed abortion. Recent reports have highlighted the critical roles of DNA methylation-mediated ROS homeostasis in various aspects of plant development and stress response, including salt stress (Chen et al. 2015; Hu et al. 2021), heat stress (Ma et al. 2018; Sakai et al. 2022; Zhu et al. 2021), chilling/freezing stress (Guo et al. 2019; Zheng et al. 2022), and fruit ripening (He et al. 2009). To ensure high quality, reads containing more than 10% unknown nucleotides (N) and 40% low-quality bases (Q-value ≤ 20) were removed. Methylation levels were calculated by determining the percentage of cytosine methylation across the genome, in each chromosome, and in different genomic regions in the context of CG, CHG, and CHH sequences. To identify differentially methylated regions (DMRs) between two samples, a minimum read coverage of 4 was used to determine the methylation status of a base. WGBS was performed on two biological replicates for each stage of seed development. The sequencing depth of the samples ranged from 19 to 23 X (Table S4) and the C-to-T conversion rate was evaluated for all samples (Table S5).

RNA-seq analysis

Total RNA was extracted from seeds using Trizol (Invitrogen, Carlsbad, CA, USA). Gene Denovo Biotechnology Co. performed sequencing using the Illumina Novaseq6000 platform (Guangzhou, China). To obtain high-quality clean reads, sequences containing adapters, more than 10% unknown nucleotides (N), or 50% low-quality bases (Q-value ≤ 20) were filtered using fastp (version 0.18.0) (Chen et al. 2018). The filtered reads were then aligned to the litchi genome using HISAT2 (Kim et al. 2015). The expression level of each transcription region was quantified using the FPKM (fragment per kilobase of transcript per million mapped reads) value, calculated with RSEM software (Li and Dewey 2011). Differentially expressed genes (DEGs) between two groups were determined using DESeq2 software (Love et al. 2014). Genes/transcripts with a false discovery rate (FDR) < 0.05 and an absolute fold change ≥ 2 were considered as DEGs. Gene Ontology (GO) enrichment analysis was performed using TBtools software (Chen et al. 2020).

Detection of H2O2 levels

The litchi seeds/Arabidopsis plants were subjected to vacuum treatment for a duration of 2 h/30 min, respectively. Subsequently, the litchi seeds/Arabidopsis plants were incubated in a 0.1% NBT solution at room temperature for a period of 20 h/2 h to visualize endogenous H2O2 and O2−. During this incubation, the samples were subjected to slow vibration. Following the incubation, the samples were immersed in 95% ethanol that had been heated by boiling water for 20 min to eliminate the green background. For each staining, a total of ten litchi seeds and three Arabidopsis plants were utilized and images were captured using a stereomicroscope (ZEISS). To detect intracellular H2O2, the litchi seeds and Arabidopsis plants were incubated in a solution of 100 μM DCFH-DA at a temperature of 37 °C for 1 h [32]. Concurrently, the samples were subjected to slow vibration. The samples were then washed three times with distilled water to remove any residue. The fluorescence of DCF was visualized using confocal laser scanning microscopy (LSM) with excitation at 488 nm (ZEISS LCM-800). Image J software was employed to quantify the intensity of fluorescence (Yang et al. 2018).

Analysis and treatment of LcGPX6 transgenic Arabidopsis

The coding sequence of LcGPX6 was fused into the vector pCAMBIA1302 followed by transformation of contructs into A. tumefaciens. Later on, LcGPX6 was introduced into Arabidopsis using the floral dip technique (Clough and Bent 1998). For seed development analysis, T3 homozygous plants were employed. A quantitative assay was conducted to determine the seed number per silique in the wild type Col and LcGPX6 transgenic lines. Statistical significance was determined via employing Independent-Sample t-test (***P < 0.001). The primers for generation of LcGPX6 transgenic Arabidopsis are listed in Table S6.