Abstract
Hybridization is an important approach to the production of new varieties with exceptional traits. Although the kernel rate of wild jujube (Ziziphus jujuba Mill. var. spinosa Hu.) is generally high, that of cultivated jujube (Z. jujuba Mill.) is low, greatly hampering the jujube breeding process. However, the mechanism by which this trait changed during jujube domestication remains unclear. Here, we explored the potential regulatory network that governs jujube embryo abortion using correlation analysis of population traits, artificial pollination, sugar content measurements and multi-omics analysis. The results showed that embryo abortion was an important reason for the low kernel rate of cultivated jujube, and kernel rate was negatively correlated with edible rate. Twenty-one days after pollination was a critical period for embryo abortion. At this time, the sugar content of cultivated ‘Junzao’ kernels decreased significantly compared with that of the pulp, but sugar content remained relatively stable in kernels of wild ‘Suanzao’. A total of 1142 differentially expressed genes targeted by 93 microRNAs (miRNAs) were identified by transcriptome, miRNA and degradome sequencing, and may be involved in the regulation of embryo abortion during kernel development. Among them, DELLA protein, TCP14 and bHLH93 transcription factors have been shown to participate in the regulation of embryonic development. Our findings suggest that carbohydrate flow between different tissues of cultivated jujube exhibits a bias toward the pulp at 21 days after pollination, thereby restricting the process of kernel development. This information enhances our understanding of the embryo abortion process and reveals miRNA–target gene pairs that may be useful for molecular-assisted breeding.
Keywords: carbohydrates, degradation, embryo abortion, jujube, miRNA, transcriptome
Introduction
Jujube (Ziziphus jujuba Mill.) is a tree species in the Rhamnaceae family that is native to China and has over 7000 years of cultivation history (Qu and Wang 1993). About 6.25 million tons of dried jujube fruit is produced annually from a cultivated area of approximately two million hectares (Shi et al. 2018). Jujube has high nutritional and medicinal value, as it contains several bioactive components that are beneficial to human health, including vitamin C, phenolics, flavonoids, triterpene acids, polysaccharides and several microelements (Gao et al. 2013, Liu et al. 2015). It receives significant attention because of its unique flavor. Jujube has a wide distribution range and rich germplasm resources, which enable it to adapt to different growth environments and provide both economic and ecological benefits. Improvement of jujube varieties through the incorporation of outstanding traits can improve fruit quality and yield. Cross-breeding is an important and effective means of incorporating multiple positive traits into a single variety during fruit tree breeding, but normal seeds cannot be obtained in most cultivated jujubes.
Chinese jujube has bisexual flowers, and researchers have long focused on its self-incompatibility (SI) in breeding research. The SI of jujube was reported by Asatryan based on fluorescence microscopy observations of pollen tube development after self- and out-cross pollination (Asatryan et al. 2013). S RNase and two S-like RNase SI genes have been identified in jujube (Huang et al. 2016). Although SI is important for parent selection, embryo abortion resulting in a seedless phenotype is also a limiting factor in jujube breeding. Seeds can rarely be harvested even after successful pollination and fertilization. This phenomenon indicates that pollen germination and pollination status are not the only factors that cause a low kernel rate in jujube; embryo abortion during the middle stage of seed development is also an important factor (Li et al. 2016). This phenomenon of embryo abortion occurs in many species. In Arabidopsis, silencing of CWIN2/4 inhibits the occurrence of ovules and leads to ovule abortion (Liao et al. 2020). Knockout of GmSWEET15 genes in soybean leads to slow embryo development and persistence of the endosperm, causing significant seed abortion (Wang et al. 2019). These findings show that sugar plays a key role in ovule development as a signaling substance or a carbon source.
High-throughput omics studies have facilitated our understanding of the molecular regulation of seed abortion at the transcriptional, biochemical and metabolic levels. In total, 140 genes and 41 proteins related to embryo development were identified by transcriptomics and proteomics in chrysanthemum flowers (Zhang et al. 2014), and YG5571 was shown to be related to ovule development in seedless grapes (Zhang et al. 2013). MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression post-transcriptionally through sequence complementarity. Studies have shown that plant miRNAs and their target genes play an important role in controlling plant development, flowering time and metabolism. Target site analysis showed that miRNA167, miRNA529 and miRNA2950 directly target ZjARF4, which is associated with jujube witches’ broom disease (Ma et al. 2020). Recently, the jujube gene ZjPOD1 was reported to be associated with seed-setting rate (Guo et al. 2020, 2021), although its mechanism of action remains unclear.
Embryo abortion leading to seedlessness is common in cultivated jujube and is a significant barrier to breeding. Cultivated jujube was domesticated from its wild ancestor, the wild jujube (Huang et al. 2016), which has a small fruit size but an extremely high kernel rate. Therefore, understanding the genetic and molecular mechanisms of kernel development will be important for future breeding improvements in jujube. In this study, we explored differences in kernel development between cultivated jujube and wild jujube. The mechanism of embryo abortion was studied in detail using population multi-trait correlation analysis, artificial pollination, nutrient content measurements and multi-omics analysis, which provided new insights into the mechanism of embryo abortion.
Materials and methods
Plant materials
All experimental materials were grown under the same conditions at the Experimental Station of Jujube at Northwest A&F University in Qingjian, Shaanxi, China (N 37.13, E 110.09). The cultivated jujube ‘Junzao’, with an extremely low kernel rate, and ‘Suanzao’, a wild jujube, were selected for artificial pollination experiments. Open buds were manually removed at the flowering stage before pollination and then labeled one by one after artificial pollination. Kernel rate was calculated according to the development status of endosperm and embryo of fruits and was recorded at 7, 14, 21, 28, 35, 42 and 49 days after pollination (DAP). Kernel and pulp from the same sampling date were collected and immediately quick-frozen in liquid nitrogen after tissue separation in a low-temperature environment. The samples were stored in a −80 °C freezer for subsequent analysis. Tissues from the same cultivars and sampling dates were also collected for sectioning and were stored in FAA fixative (70% ethanol:formaldehyde:glacial acetic acid = 90:5:5) (Fei et al. 2021). The sections were observed and photographed with an Olympus BX43 front-mounted microscope (Olympus DP80, Olympus Medical Systems Co., Tokyo, Japan).
Kernel rates of different jujubes
A total of 99 cultivated jujubes and 56 wild jujubes grown at the Shanxi Taigu National Jujube Germplasm Resource Nursery (N 37.42 E 112.55) have been identified (Table S1 available as Supplementary data at Tree Physiology Online). We selected three fruiting trees of each jujube and collected a total of 30 mature fruits of similar size from the sun-facing and semi-sunny sides of the trees. Fruit shape indexes (FSIs—ratios of fruit length to fruit width) and kernel shape indexes (KSIs—ratios of kernel length to kernel width) were measured with vernier calipers (Liu and Wang 2009, Guo et al. 2020), and >50 fruits of each jujube were used for kernel rate calculations.
Sugar content measurements
Sugars were extracted using the method of Zhang with some modifications (Zhang et al. 2018). Three biologically replicated kernel and pulp samples (0.2 g each) were processed with a grinder and extracted in 80% ethanol. The total soluble sugar and starch content were estimated using the anthrone-sulfuric acid colorimetric method (Lu et al. 2020), and reducing sugars were measured using the 3,5-dinitrosalicylic acid method (Wang et al. 2017). Monosaccharide contents were measured with an ICS-5000+ ion chromatograph (Dionex, USA), and all monosaccharide standards were purchased from Yuanye Bio-Technology Co. (Shanghai, China).
RNA extraction, library construction and sequencing
Three biologically replicated samples of pulp and kernel tissue from ‘Junzao’ at three developmental stages (7, 21 and 35 DAP) were used for RNA-seq experiments. Total RNA was isolated and purified using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions. The amount and purity of each RNA sample were quantified using a NanoDrop ND-1000 instrument (NanoDrop, Wilmington, DE, USA). RNA integrity (RIN > 7.0) was assessed with a Bioanalyzer 2100 (Agilent, CA, USA) and confirmed by electrophoresis on a denaturing agarose gel. StringTie software (http://ccb.jhu.edu/software/stringtie/) was used for initial assembly of genes or transcripts. The initial assembly results of all samples were merged, and gffcompare (http://ccb.jhu.edu/software/stringtie/gffcompare.shtml) was used to compare the assembled transcripts with the jujube reference annotation to obtain the final assembly annotation result (Huang et al. 2016). Gene expression was quantified as Fragments Per Kilobase of exon model per Million mapped reads, and genes that were differentially expressed between samples were identified based on the criteria |fold change| ≥ 2 and false discovery rate (FDR)-adjusted P-value < 0.05. GO and KEGG enrichment analyses were performed on the differentially expressed genes (DEGs) using OmicStudio (https://www.omicstudio.cn/login).
Sequencing and identification of miRNAs
Eighteen miRNA libraries were constructed from replicated samples as described above. The raw reads were processed with the in-house program ACGT101-miR (LC Sciences, Houston, TX, USA) to remove adapter dimers, contamination, low-complexity reads, reads from common RNA families (rRNA, tRNA, snRNA snoRNA) and repeats. Unique 18–25 nucleotide (nt) sequences were mapped to specific species precursors in the miRBase 22.0 database using BLAST to identify known miRNAs and novel 3p- and 5p-derived miRNAs. Length variation at both the 3′ and 5′ ends and one mismatch inside the sequence were allowed in the alignment. Unique sequences that mapped to mature miRNAs of specific species in hairpin arms were identified as known miRNAs. Unique sequences that mapped to the other arm, opposite to the annotated mature miRNA-containing arm of specific species’ precursor hairpins, were considered to be novel 5p- or 3p-derived miRNA candidates. The remaining sequences were mapped to precursors from other selected species (with the exclusion of the specific species) in miRBase 22.0 using BLAST, and the mapped pre-miRNAs were also BLASTed against the specific species’ genomes to determine their genomic locations. The latter two miRNA types were defined as known miRNAs. Unmapped sequences were BLASTed against specific genomes, and the sequences that contained hairpin RNA structures were predicted in the flanking 120-nt sequences using RNAfold software (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi).
Identification of differentially expressed miRNAs
The expression levels of the miRNAs were quantified as transcripts per million reads (TPM = [Read count × 1,000, 000]/number of mapped reads). DESeq software was used to identify miRNAs that were differentially expressed between developmental stages based on the criteria of FDR ≤ 0.05 and |log2(fold-change) | ≥ 1.
Degradome sequencing
Nine kernel RNA samples and nine pulp RNA samples were mixed to construct two degradation libraries and were sent to Hangzhou LC-Bio Co., Ltd (Hangzhou, China) for cDNA library construction and sequencing on the Illumina GAII sequencing platform (Illumina, San Diego, CA, USA). Extracted degradome sequencing reads that had 20–21 nt were used to identify potentially cleaved targets using the CleaveLand 3.0 pipeline. The degradome reads were then mapped to the transcriptome data. The identified targets were categorized as 0, 1, 2, 3 and 4 as described in previous studies, and t-plots were constructed for high-efficiency analysis of potential miRNA targets (Ambady et al. 2012). Finally, all the identified potential target genes were used for BLASTX searches at NCBI and annotated by GO analysis (Young et al. 2010).
RT-qPCR validation of miRNAs and mRNA targets
To validate the high-throughput sequencing results, we performed Real-time quantitative polymerase chain reaction(RT-qPCR) of 10 pairs of miRNAs and their target genes using a LightCycler 96 assay system (Roche, Switzerland). The RNA samples used for RT-qPCR were the same as those used in the experiments above. The reverse transcription reactions for miRNAs and cDNA were performed using the SYBR PrimeScript miRNA RT-qPCR Kit (TaKaRa, Dalian, China) and the Superscript III First-Strand Synthesis system followed by RNase H treatment (Invitrogen). The target gene primers were designed using Primer 5 (http://frodo.wi.mit.edu/primer5/). U6 was used as an internal reference gene for miRNA qPCR. The relative expression levels of target genes were also detected by RT-qPCR, and UBQ1 was selected as an internal reference gene (Zhang et al. 2015). The gene names, sequences and primers used for RT-qPCR analysis are provided in Table S2 available as Supplementary data at Tree Physiology Online. RT-qPCR for miRNAs and their targets were performed in 96-well plates using the SYBR Premix Ex Taq Kit (TaKaRa) and the SYBR PrimeScript miRNA RT-PCR Kit (TaKaRa). The amplification procedure and data analysis were performed as described in a previous study (Shi et al. 2020). Each PCR was repeated independently three times.
Results
Kernel rates of cultivated and wild jujube
We surveyed the variation in kernel rate among 155 cultivated and wild jujubes. Kernel rate was only weakly negatively correlated with KSIs in the 99 cultivated jujubes (Figure 1a). However, when all 155 wild and cultivated jujubes were considered, kernel rate was weakly positively correlated with transverse kernel diameter and negatively correlated with other characters (Figure 1b). The strongest negative correlation was that between kernel rate and edible rate, suggesting a clear relationship between these characters. Our results reveal a wide range of embryo abortion rates among jujubes. We selected representative wild and cultivated jujubes for subsequent studies on the occurrence and mechanism of embryo abortion.
Figure 1.
Cluster analysis of correlations among multiple traits. (a) Correlation analysis of multiple characters in 99 cultivated jujubes. (b) Correlation analysis of multiple characters in 99 cultivated jujubes and 56 wild jujubes.
21 DAP: a critical time period for embryo abortion in jujube
We divided seed development into two stages based on morphological observations: immature embryo development (before 21 DAP) and kernel formation (after 21 DAP). The proportion of immature embryos with normal development was relatively high in the early developmental stage of cultivated jujube ‘Junzao’, but the kernel rate gradually decreased to ~20% during the kernel formation stage. By contrast, wild jujube ‘Suanzao’ showed a marked decrease in the proportion of normal embryos during the immature embryo developmental stage, which may have been caused by falling flowers and fruit, but the kernel rate rapidly increased and reached 100% during the kernel development period (Figure 2a). Kernel development continued throughout all stages of ‘Suanzao’ (Figure 2c and d), but stagnation of kernel development or abortion was observed in ‘Junzao’ at 28 DAPs. The period of abortion was basically consistent with variations in kernel rate.
Figure 2.
Changes in kernel rate during development and observations of embryo development through time. (a) Kernel rate in the natural state. (b) Kernel rate after artificial pollination. (c) Morphological characteristics and histological observations of multiple developmental stages in ‘Suanzao’. (d) Morphological characteristics and histological observations of multiple developmental stages in ‘Junzao’. The red arrow indicates the embryo.
To further examine the correlation between embryo abortion and kernel rate, a ‘Junzao’ (female) × ‘Suanzao’ (male) cross was performed by artificial pollination. The resulting kernel rate resembled that observed in ‘Junzao’ after spontaneous pollination (Figure 2b). These results indicated that embryo abortion is an important factor causing reductions in kernel rate during the development of jujube embryos.
Changes in carbohydrate flow during kernel development
To assess the potential relationship between embryo abortion and carbohydrate flow in jujube, we measured changes in carbohydrate content in different tissues at multiple stages during the development of cultivated jujube ‘Junzao’ and wild jujube ‘Suanzao’ (Figure 3 and Table S3 available as Supplementary data at Tree Physiology Online). Total soluble sugar, reducing sugar and starch were measured in kernel and pulp (Figure 3a–c). In general, ‘Junzao’ showed a greater difference in carbohydrate partitioning between kernel and pulp than ‘Suanzao’, and as development progressed, the total sugar and reducing sugar contents became much higher in pulp than in kernels of ‘Junzao’. By contrast, the difference in sugar contents between pulp and kernel was much smaller in ‘Suanzao’. A similar trend was observed for mono- and disaccharides (Figure 3d–f). Early in development, levels of these sugars were similar between the kernel and the pulp of ‘Junzao’, but they were significantly higher in the pulp later in development. This difference was less marked in ‘Suanzao’. We speculate that the early development of ‘Junzao’ embryos is restricted by reduced carbohydrate levels and that carbohydrates are instead directed toward the pulp, which is the more agronomically valuable tissue.
Figure 3.
The contents of soluble sugar, reducing sugar, starch and mono- and disaccharides (glucose, fructose and sucrose) in different tissues and developmental stages of ‘Junzao’ and ‘Suanzao’. DAP means days after polling. Values are expressed as the means ± standard deviation of three replicates.
Transcriptome analysis of pulp and kernel during different developmental stages
To explore genes that may be associated with embryo abortion, we performed transcriptome sequencing on pulp and kernel tissues from three developmental stages (7, 21 and 35 DAP). A total of 18 libraries were constructed, producing a total of 865,581,484 raw reads. After the removal of sequencing adapters and low-quality sequencing data, 830,609,558 clean reads remained. The quality control pipeline for the sequencing data is shown in Table S4 available as Supplementary data at Tree Physiology Online.
A large number of DEGs were identified between different developmental stages: DAP7 vs DAP21 (3278 DEGs, 2066 upregulated), DAP21 vs DAP35 (6176 DEGs, 3755 upregulated) and DAP7 vs DAP35 (6624 DEGs, 4332 upregulated). A total of 13,192 genes were differentially expressed across the three time points, and 8417 genes were differentially expressed in at least one comparison (Figure 4a and b). KEGG enrichment analysis was performed on the DEGs from each comparison, and highly enriched pathways included ko04075 (plant hormone signal transduction), ko04626 (plant–pathogen interaction), ko04016 (MAPK signaling pathway-plant), ko00940 (phenylpropanoid biosynthesis), ko00500 (starch and sucrose metabolism), ko00040 (pentose and glucuronate interconversions) and ko00520 (amino sugar and nucleotide sugar metabolism) (Figure 4c). These results suggest that hormones and carbohydrate metabolism may have key regulatory roles in the early stages of kernel development until abortion.
Figure 4.
Analysis of DEGs during early development of ‘Junzao’ embryos. (a) Venn diagram of the number of DEGs between different comparison groups. (b) Up- and downregulated genes in pairwise comparisons between three developmental stages. (c) KEGG enrichment analysis of the three groups of DEGs. The size of the circle represents the number of genes enriched in the pathway, the P-value indicates the significance of the enrichment and red indicates a higher degree of enrichment.
Differences in gene expression were also examined in kernel vs pulp at the same stage of development. There were 2647, 4550 and 5685 DEGs between kernel and pulp at 7, 21 and 35 DAP: 1290, 2401 and 2440 genes upregulated in kernels at the three stages and the remaining 1357, 2149 and 3245 genes downregulated (Table S3 available as Supplementary data at Tree Physiology Online). A total of 9304 DEGs between kernel and pulp across the three successive developmental stages (Figure 5a). KEGG enrichment results for these genes were consistent with the results of the kernel development analysis, above. The DEGs were significantly enriched in ko04075 (plant hormone signal transduction), ko04626 (plant–pathogen interaction), ko04016 (MAPK signaling pathway-plant), ko00940 (phenylpropanoid biosynthesis) and ko00500 (starch and sucrose metabolism). The enrichment analysis results suggested that there was active hormone regulation and energy metabolism flow between the two tissues in the early stages of fruit development (Figure 5b). A heat map was drawn to illustrate the expression of a subset of genes (Figure 5c).
Figure 5.
Analysis of DEGs during early development of ‘Junzao’ embryos. (a) Venn diagram of DEGs between different comparison groups. (b) KEGG enrichment analysis of the two groups of DEGs. The size of the circle represents the number of genes enriched in the pathway, the P-value indicates the significance of the enrichment and red indicates a higher degree of enrichment. (c) Expression of DEGs in pairwise comparisons between the three developmental stages. Red indicates upregulation, and blue indicates downregulation.
Identification of known and novel miRNAs
To investigate the regulatory role of miRNAs in embryo abortion, 18 miRNA libraries (corresponding to the transcriptome samples) were constructed to identify miRNAs expressed at three developmental stages of kernel and pulp. After quality control of the raw reads and removal of the 3′ linker, the lengths were 18–25 nt; these sequences were compared with various RNA databases (excluding miRNAs), such as the mRNA database, the RFam database (which includes rRNA, tRNA, snRNA and snoRNA sequences) and the Repbase repeated sequence database. Hits to these databases were removed (Table S5 available as Supplementary data at Tree Physiology Online), and the remaining reads were compared with the ‘Junzao’ genome.
A total of 927 miRNAs were identified, and their length distribution characteristics are shown in Figure S1 available as Supplementary data at Tree Physiology Online. Due to the high degree of conservation of miRNA sequences among species, all miRNA sequences obtained by miRNA sequencing were compared with all mature miRNAs of plant species in the miRBase database (Version 22.1) (http://www.mirbase.org/) to identify known miRNAs expressed during the early development of jujube. Forty-three known miRNAs from 13 miRNA families were identified in the 18 miRNA libraries. The MIRNA156 family had the most members (eight miRNAs), followed by the MIRNA171 family (seven miRNAs). There were four miRNAs from the MIRNA160, MIRNA319 and MIRNA8005 families, and the remaining families were represented by one to two miRNAs (Figure S2 available as Supplementary data at Tree Physiology Online). In total, 507 miRNA sequences were not matched to pre-miRNAs of plant species in miRbase. However, they did match genomic sequences and were predicted to form hairpin structures and were thus identified as novel miRNAs. The regulation of miRNAs in organisms is microregulated, and we therefore gave priority to a number of miRNAs with high expression levels. The expression levels of some miRNAs were relatively low, and it was unclear whether their presence simply reflected transcriptional noise; we therefore filtered out miRNAs with expression levels <10 during analysis. miRNAs whose precursor miRNAs could not be mapped to the genome were not analyzed further because they lacked genomic support.
miRNAs and their target genes related to embryo abortion
miRNAs participate in plant growth and development and have important roles in regulating development and responding to environmental stress. Their expression patterns can provide insight into kernel development, as they help to regulate this process (Chen et al. 2019, Siddiqui et al. 2019). miRNAs have been identified in Arabidopsis, corn, grape and other species (Wu et al. 2017, Chen et al. 2019). In this study, 332 miRNAs were differentially expressed at three time points during kernel development, and 248 miRNAs were differentially expressed during pulp development. In total, the expression patterns of 425 miRNAs changed during kernel and pulp development in jujube. In total, 123 miRNAs were classified as known and novel miRNAs, and their corresponding target genes were identified by degradome group sequencing. Real-time fluorescent quantitative PCR was used to verify the expression of the predicted target genes and confirm the RNA-seq results (Figure S3 available as Supplementary data at Tree Physiology Online). Because one miRNA may regulate multiple target genes and one target gene may be controlled by multiple miRNAs, 2716 target genes were identified for the 123 miRNAs. These 2716 genes were compared with the list of 9304 DEGs identified in different tissues, and 1142 of the predicted target genes were differentially expressed in the different tissues. These mRNAs were predicted to be the targets of 96 miRNAs. To gain new insights into possible mechanisms of jujube embryo abortion, we focused our attention on these 1142 target genes and performed enrichment analysis to identify their functions and possible regulatory pathways. Thirty-three of the 1142 target genes were found to be transcription factors that were regulated by 24 miRNAs (7 known miRNAs and 17 novel miRNAs), forming 43 miRNA–mRNA target gene pairs (Table 1). Their expression changes at each developmental stage are shown in Figure 6.
Table 1.
In total, 43 miRNA–mRNA target gene pairs and their gene annotations.
| miRNA | Targets | Target annotation | Abbreviations |
|---|---|---|---|
| zju-MIR319a-p5 | evm.TU.Contig111.156 | Trihelix transcription factor GT-2-like, partial | GT-2-1 |
| zju-MIR2916-p3_1ss8TC | evm.TU.Contig101.72 | DELLA transcription factor | DELLA |
| zju-novel-8972_790-5p | evm.TU.Contig82.147 | Nuclear transcription factor Y subunit C-1 | Subunit C-1 |
| zju-novel-50463_154-5p | evm.TU.Contig111.156 | Trihelix transcription factor GT-2-like, partial | GT-2-1 |
| zju-novel-42882_184-5p | evm.TU.Contig28.0.99 | Transcription factor MYB1R1 | MYB1R1 |
| zju-novel-39360_202-5p | evm.TU.Contig36.127 | myb family transcription factor EFM | EFM |
| zju-novel-7303_940-3p | evm.TU.Contig12.2.11 | GATA transcription factor 4 | GATA |
| evm.TU.Contig16.2.14 | Trihelix transcription factor GT-2-like | GT-2 | |
| evm.TU.Contig2.462 | Transcription factor DIVARICATA | DIVARICATA | |
| evm.TU.Contig27.125 | Transcription factor MYB46 | MYB46 | |
| evm.TU.Contig38.46 | myb family transcription factor APL isoform X1 | APL | |
| evm.TU.Contig45.38 | Probable WRKY transcription factor 69 | WRKY69 | |
| evm.TU.Contig70.102 | Transcription factor bHLH143-like isoform X2 | bHLH143 | |
| evm.TU.Contig80.34 | Transcription factor ILR3 | ILR3 | |
| zju-novel-6811_994-3p | evm.TU.Contig0.2.23 | Ethylene-responsive transcription factor 4 | ERF4 |
| zju-novel-64849_114-3p | evm.TU.Contig116.279 | Ethylene-responsive transcription factor ERF106-like | ERF106 |
| zju-novel-48348_162-3p | evm.TU.Contig84.447 | Trihelix transcription factor ASR3 | ASR3 |
| zju-novel-4257_1458-3p | evm.TU.Contig4.44 | Trihelix transcription factor ASIL2 | ASIL2 |
| evm.TU.Contig47.104_ evm.TU.Contig47.105 | Transcription factor BIM1 isoform X1 | BIM1 | |
| zju-novel-3975_1540-3p | evm.TU.Contig111.433 | Transcription factor bHLH91-like | bHLH91 |
| evm.TU.Contig3.23 | Transcription factor ICE1-like | ICE1 | |
| evm.TU.Contig4.44 | Trihelix transcription factor ASIL2 | ASIL2 | |
| evm.TU.Contig45.52 | Heat stress transcription factor A-4c-like | A-4c-like | |
| evm.TU.Contig5.172 | Nuclear transcription factor Y subunit A-3-like | A-3-like | |
| evm.TU.Contig62.374 | Ethylene-responsive transcription factor RAP2–13-like | RAP2 | |
| evm.TU.Contig73.14 | Ethylene-responsive transcription factor ERF109-like | ERF109 | |
| zju-novel-39447_201-3p | evm.TU.Contig19.281 | Basic transcription factor 3-like isoform X2 | Basic3 |
| zju-novel-38951_204-3p | evm.TU.Contig13.274 | Transcription factor PIF3 isoform X1 | PIF3 |
| evm.TU.Contig4.44 | Trihelix transcription factor ASIL2 | ASIL2 | |
| evm.TU.Contig47.104_ evm.TU.Contig47.105 | Transcription factor BIM1 isoform X1 | BIM1 | |
| zju-novel-35927_222-3p | evm.TU.Contig70.239 | Transcription factor MYB3R-1-like | MYB3R |
| zju-novel-24202_327-3p | evm.TU.Contig116.279 | Ethylene-responsive transcription factor ERF106-like | ERF106 |
| evm.TU.Contig69.109 | Transcription factor TCP14 | TCP14 | |
| zju-novel-22253_354-3p | evm.TU.Contig0.1.42 | Transcription factor MYB41 | MYB41 |
| evm.TU.Contig69.109 | Transcription factor TCP14 | TCP14 | |
| zju-novel-21677_363-3p | evm.TU.Contig0.2.23 | Ethylene-responsive transcription factor 4 | ERF4 |
| zju-novel-16578_463-3p | evm.TU.Contig124.33 | Transcription factor MYB124-like isoform X1 | MYB124 |
| zju-miR156e | evm.TU.Contig34.338 | Nuclear transcription factor Y subunit gamma isoform X2 | Nuclea |
| zju-MIR5368p3_1ss16CT | evm.TU.Contig116.64 | Ethylene-responsive transcription factor ERF027 | ERF027 |
| zju-miR171o-5p | evm.TU.Contig70.102 | Transcription factor bHLH143-like isoform X2 | bHLH143 |
| zju-miR159b-5p_1ss1AG | evm.TU.Contig101.72 | DELLA transcription factor | DELLA |
| evm.TU.Contig84.447 | Trihelix transcription factor ASR3 | ASR3 | |
| zju-MIR160d-p3_1ss10GA | evm.TU.Contig57.409 | Transcription factor bHLH93-like | bHLH93 |
Figure 6.
Heat map of miRNA and target gene expression. (a) Expression heatmap of 24 miRNAs in kernel and pulp at three consecutive developmental periods, and (b) expression heatmap of 33 transcription factors that may be associated with embryo abortion.
The most highly enriched KEGG pathway in the 1142 target genes was ko03010 (ribosome) with 33 genes, followed by ko04075 (plant hormone signal transduction) with 28 genes (Table S6 available as Supplementary data at Tree Physiology Online). Other significantly enriched pathways included ko04626 (plant–pathogen interaction), ko04141 (protein processing in endoplasmic reticulum) and ko00500 (starch and sucrose metabolism). The hormonal regulation of seed development has been studied in many species, and apricot has been reported to undergo embryo abortion caused by uneven energy distribution between tissues. We therefore performed a more in-depth analysis of the genes in the enriched ko04075 and ko00500 pathways (Figure 7). Four transcription factors were found to be enriched in the hormone signal transduction pathway. A DELLA transcription factor was co-regulated by two miRNAs, constituting different miRNA–target gene pairs, and the other three transcription factors were all regulated by a single miRNA.
Figure 7.
miRNAs and their target genes. (a) miRNA–target gene pairs in the hormone signal transduction pathway. (b) miRNA–target gene pairs in the starch and sucrose metabolism pathway.
Discussion
The key period of embryo abortion and nutritional competition in jujube
The timing and mechanism of embryo abortion is the key to understanding jujube kernel development. By observing the changes in kernel rate and paraffin sections of cultivated ‘Junzao’ fruit after pollination, we found that the initial stage of embryo abortion occurred at 21 DAPs; the jujube kernel then atrophied and browned, and embryo abortion occurred. Only microglobular embryos were observed in ‘Junzao’, whereas the entire developmental process from microglobular embryos to cotyledons was observed in wild jujube ‘Suanzao’ (Figure 1). Li et al. (2016) used different pollination combinations to investigate the phenomenon of embryo abortion during kernel development in cultivated jujube, and this study identified the key time point for embryo abortion in cultivated jujube ‘Junzao’.
A correlation analysis of fruit traits in 99 cultivated and 56 wild jujubes showed that kernel rate was negatively correlated with FSIs and edible rate. Changes in carbohydrate levels of pulp and kernel tissues at different developmental stages suggested that there was competition for carbohydrates between these two sink tissues at the immature embryo development stage (Ruan et al. 2012). From wild jujube to cultivated jujube, the kernel rate decreased from 100% to 20% (or 0% in some cases), but the edible rate increased from 20% to 96%. We therefore speculate that more carbohydrates flowed to the pulp in cultivated jujube, limiting carbohydrate supply to the developing kernel and leading to embryo abortion. Wang et al. (2004) also proposed an effect of carbohydrate competition on kernel development in jujube. Similarly, knockout of two GmSWEET15 genes (GmSWEET15a and GmSWEET15b) in soybean resulted in delayed embryonic development and persistence of the endosperm and led to severe seed abortion (Wang et al. 2019). However, some studies on assimilate restriction and seed abortion found that a lack of carbohydrates was the result of embryo abortion rather than the cause, and genes that affected embryo growth played their role earlier than those that affected glucose metabolism (Oury et al. 2016, Shen et al. 2018).
Identification of genes related to embryo abortion in jujube
Identification of the genes that regulate embryo abortion is the key to improving the kernel rate of jujube and speeding up the process of cross breeding. Transcriptome, miRNA and degradome sequencing of samples collected 7, 21 and 35 DAP after pollination identified 1142 DEGs targeted by 93 miRNAs. The DEGs were enriched in hormone signal transduction and starch and sucrose metabolism pathways, suggesting that hormones may also play an important role in embryo abortion (Singh et al. 2010, Zu et al. 2021). Among the 33 miRNA-targeted transcription factors, DELLA protein, TCP14 and bHLH93 are known to be involved in the regulation of kernel development through hormone metabolism and synthesis. DELLA protein, a negative regulator of gibberellin (GA) signaling, belongs to a subfamily of plant-specific GRAS transcriptional regulators that interact with the transcription factor ABERRANT TESTA SHAPE (ATS). The inhibition of GA biosynthesis by ATS leads to DELLA homeostasis. This provides evidence for the role of GAs in ovule and seed development (Gomez et al. 2016). TCP14, generally reported to be involved in plant growth and development as well as hormone regulation, modulates GA-dependent stamen elongation and is expressed primarily in the vascular tissue of the embryo (Resentini et al. 2021). Transcriptomic analysis of rubber ovules at different developmental stages showed that bHLH93 was highly expressed in the cotyledon embryo stage, indicating that it plays a key role in late embryo development (Wang et al. 2021). In this study, the expression level of bHLH93 showed a general decrease with increasing degree of embryo abortion. In addition, grape and other fruits regulate the occurrence of seedless traits through gibberellin (Li et al. 2019, Gomez et al. 2020). The EMBRYO-DEFECTIVE (EMB) genes of Arabidopsis have roles in seed development (Meinke 2020); the expression level of an EMB homolog was significantly reduced as embryo abortion occurred in ‘Junzao’, and its potential regulatory role in embryo abortion requires further study.
Conclusions
miRNA sequencing and transcriptome profiling were performed to gain insight into the miRNA-mediated regulation of embryo abortion. Forty-three known and 108 potentially novel miRNAs from 13 miRNA families were identified; 96 miRNAs were differentially expressed during early kernel development and were predicted to target 1142 genes. These differentially expressed miRNAs (zju-mir2916-p3_1ss8tc zju-novel-24202_327-3p and zju-mir160d-p3_1ss10ga) and their corresponding target genes (transcription factors DELLA protein, TCP14, bHLH93 and an embryo-sensitive genes) may play key roles in the regulation of embryo abortion. These findings provide a foundation for further study of miRNA-mediated embryonic development.
Supplementary Material
Contributor Information
Jiangtao Du, College of Forestry, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Key Comprehensive Laboratory of Forestry of Shaanxi Province, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China.
Qianqian Shi, College of Forestry, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Key Comprehensive Laboratory of Forestry of Shaanxi Province, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China.
Yu Liu, College of Forestry, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Key Comprehensive Laboratory of Forestry of Shaanxi Province, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China.
Guozhao Shi, College of Forestry, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Key Comprehensive Laboratory of Forestry of Shaanxi Province, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China.
Xi Li, College of Forestry, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Key Comprehensive Laboratory of Forestry of Shaanxi Province, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China.
Xingang Li, College of Forestry, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Key Comprehensive Laboratory of Forestry of Shaanxi Province, Northwest Agriculture and Forestry University, Yangling 712100, Shaanxi, China; Research Center for Jujube Engineering and Technology of State Forestry Administration, Northwest A&F University, Yangling 712100, Shaanxi, China.
Conflicts of interest
The authors declare that they have no competing financial interests.
Funding
This work was supported by The National Key Research and Development Program of China (Grant No. 2018YFD1000607), the National Natural Science Foundation of China (Grant No. 32101564), the earmarked fund of Xinjiang Jujube Industrial Technology System (Grant No. XJCYTX-01) and Shaanxi Province Science and Technology Project (NYKJ-2021-YL(XN)21).
Acknowledgments
The authors are grateful to Dr. Xitong Fei and Chunmei Zhang for the suggestion on the experiment.
Authors’ contributions
X.-G.L. conceived the project. J.D. and Q.S. designed the experiments and performed the experiment. J.D. wrote the paper. Y.L., G.S. and X.L. conducted the sample collection and data statistics. All authors (J.D., Q.S., Y.L, G.S, X.L. and X.-G.L.) contributed to the article and approved the final manuscript.
References
- Ambady S, Wu Z, Dominko T (2012) Identification of novel microRNAs in Xenopus laevis metaphase II arrested eggs. Genesis 50:286–299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Asatryan A, Tel-Zur N (2013) Pollen tube growth and self-incompatibility in three Ziziphus species (Rhamna-ceae). Flora 208:390–399. [Google Scholar]
- Chen H, Yang Q, Chen K et al. (2019) Integrated microRNA and transcriptome profiling reveals a miRNA-mediated regulatory network of embryo abortion under calcium deficiency in peanut (Arachis hypogaea L.). BMC Genomics 20:392. 10.1186/s12864-019-5770-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fei XT, Lei Y, Qi YC, Wang SJ, Hu HC, Wei AZ (2021) Small RNA sequencing provides candidate miRNA-target pairs for revealing the mechanism of apomixis in Zanthoxylum bungeanum. BMC Plant Biol 21:178. 10.1186/s12870-021-02935-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gao QH, Wu CS, Wang M (2013) The jujube (Ziziphus jujuba Mill.) fruit: a review of current knowledge of fruit composition and health benefits. J Agric Food Chem 61:3351–3363. [DOI] [PubMed] [Google Scholar]
- Gomez MD, Ventimilla D, Sacristan R, Perez-Amador MA (2016) Gibberellins regulate ovule integument development by interfering with the transcription factor ATS. Plant Physiol 172:2403–2415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gomez MD, Barro-Trastoy D, Fuster-Almunia C, Tornero P, Alonso JM, Perez-Amador MA (2020) Gibberellin-mediated RGA-LIKE1 degradation regulates embryo sac development in Arabidopsis. J Exp Bot 71:7059–7072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Guo MX, Zhang ZR, Cheng YW et al. (2020) Comparative population genomics dissects the genetic basis of seven domestication traits in jujube. Hortic Res 7:89. 10.1038/s41438-020-0312-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Guo MX, Zhang ZR, Li SP et al. (2021) Genomic analyses of diverse wild and cultivated accessions provide insights into the evolutionary history of jujube. Plant Biotechnol J 19:517–531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang J, Zhang CM, Zhao X et al. (2016) The jujube genome provides insights into genome evolution and the domestication of sweetness/acidity taste in fruit trees. PLoS Genet 12:e1006433. 10.1371/journal.pgen.1006433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li DK, Xue XF, Wang YK, Zhao AL, Ren HY, Sui CL, Shi MJ, Du XH, Liang Q (2016) Observation of embryo development an abortion in Chinese jujube (Ziziphus jujuba Mill.). Acta Agric Bor Sin 25:1379–1385. [Google Scholar]
- Li YD, Zhang SL, Dong RZ, Wang L, Yao J, van Nocker S, Wang XP (2019) The grapevine homeobox gene VvHB58 influences seed and fruit development through multiple hormonal signaling pathways. BMC Plant Biol 19:523. 10.1186/s12870-019-2144-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liao SJ, Wang L, Li J, Ruan YL (2020) Cell wall invertase is essential for ovule development through sugar signaling rather than provision of carbon nutrients. Plant Physiol 183:1126–1144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu MJ, Wang M (2009) Germplasm resources of Chinese jujube. China Forestry Publishing House, Beijing, China, (in Chinese). [Google Scholar]
- Liu MJ, Wang JR, Liu P, Zhao J, Zhao ZH, Dai L, Li XS, Liu ZG (2015) Historical achievements and frontier advances in the production and research of Chinese jujube (Ziziphus jujuba) in China. Acta Hortic Sin 31:1683–1702. [Google Scholar]
- Lu JZ, Cheng L, Huang YY, et al. (2020) Resumptive Streptococcus mutans persisters induced from dimethylaminododecyl methacrylate elevated the cariogenic virulence by up-regulating the quorum-sensing and VicRK pathway genes. Front Microbiol 10:3102. 10.3389/fmicb.2019.03102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meinke DW (2020) Genome-wide identification of EMBRYO-DEFECTIVE (EMB) genes required for growth and development in Arabidopsis. New Phytol 226:306–325. [DOI] [PubMed] [Google Scholar]
- Ma FL, Huang JQ, Yang J, Zhou JY, Sun QB, Sun J (2020) Identification, expression and miRNA targeting of auxin response factor genes related to phyllody in the witches’ broom disease of jujube. Gene 746:144656. [DOI] [PubMed] [Google Scholar]
- Oury V, Caldeira CF, Prodhomme D, Pichon JP, Gibon Y, Tardieu F, Turc O (2016) Is change in ovary carbon status a cause or a consequence of maize ovary abortion in water deficit during flowering? Plant Physiol 171:997–1008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Qu ZZ, Wang Y (1993) Chinese fruit trees record-Chinese jujube. China Forestry Publishing House, Beijing, China. [Google Scholar]
- Resentini F, Felipo-Benavent A, Colombo L, Blázquez MA, Alabadí D, Masiero S (2021) TCP14 and TCP15 mediate the promotion of seed germination by gibberellins in Arabidopsis thaliana. Mol Plant 8:482–485. [DOI] [PubMed] [Google Scholar]
- Ruan YL, Patrick JW, Bouzayen M, Osorio S, Fernie AR (2012) Molecular regulation of seed and fruit set. Trends Plant Sci 17:656–665. [DOI] [PubMed] [Google Scholar]
- Shen S, Zhang L, Liang XG, Zhao X, Lin S, Qu LH, Liu YP, Gao Z, Ruan YL, Zhou SL (2018) Delayed pollination and low availability of assimilates are major factors causing maize kernel abortion. J Exp Bot 69:1599–1613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shi QQ, Zhang Z, Su JJ, Zhou J, Li XG (2018) Comparative analysis of pigments, phenolics, and antioxidant activity of Chinese jujube (Ziziphus jujuba Mill.) during fruit development. Molecules 23:1917. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shi QQ, Du JT, Zhu DJ, Li X, Li XG (2020) Metabolomic and transcriptomic analyses of anthocyanin biosynthesis mechanisms in the color mutant Ziziphus jujuba cv. Tailihong J Agric Food Chem 68:15186–15198. [DOI] [PubMed] [Google Scholar]
- Siddiqui ZH, Abbas ZK, Ansari MW, Khan MN (2019) The role of miRNA in somatic embryogenesis. Genomics 111:1026–1033. [DOI] [PubMed] [Google Scholar]
- Singh DP, Filardo FF, Storey R, Jermakow AM, Yamaguchi S, Swain SM (2010) Overexpression of a gibberellin inactivation gene alters seed development, KNOX gene expression, and plant development in Arabidopsis. Physiol Plant 138:74–90. [DOI] [PubMed] [Google Scholar]
- Wang JR (2004) Research on male sterility and embryo abortion of jujube trees. Hebei Agricultural University. [Google Scholar]
- Wang P, Wu Q, Jiang XJ, Wang ZQ, Tang JL, Xu Y (2017) Bacillus licheniformis affects the microbial community and metabolic profile in the spontaneous fermentation of Daqu starter for Chinese liquor making. Int J Food Microbiol 250:59–67. [DOI] [PubMed] [Google Scholar]
- Wang S, Yokosho K, Guo R, Whelan J, Ruan YL, Ma JF, Shou H (2019) The soybean sugar transporter GmSWEET15 mediates sucrose export from endosperm to early embryo. Plant Physiol 180:2133–2141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang Y, Li HL, Zhou YK, Guo D, Zhu JH, Peng SQ (2021) Transcriptomes analysis reveals novel insight into the molecular mechanisms of somatic embryogenesis in Hevea brasiliensis. BMC Genomics 22:183. 10.1186/s12864-021-07501-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu Y, Yang LY, Yu ML, Wang JB (2017) Identification and expression analysis of microRNAs during ovule development in rice (Oryza sativa) by deep sequencing. Plant Cell Rep 36:1815–1827. [DOI] [PubMed] [Google Scholar]
- Young MD, Wakefield MJ, Smyth GK, Oshlack A (2010) Gene ontology analysis for RNA-seq: accounting for selection bias. Genome Biol 11:R14. 10.1186/gb-2010-11-2-r14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang CH, Gong PJ, Wei R, Li SX, Zhang XT, Yu YH, Wang YJ (2013) The metacaspase gene family of Vitis vinifera L.: characterization and differential expression during ovule abortion in stenospermocarpic seedless grapes. Gene 528:267–276. [DOI] [PubMed] [Google Scholar]
- Zhang CM, Bian Y, Hou SH, Li XG (2018) Sugar transport played a more important role than sugar biosynthesis in fruit sugar accumulation during Chinese jujube domestication. Planta 248:1187–1199. [DOI] [PubMed] [Google Scholar]
- Zhang CM, Huang J, Li XG (2015) Identification of appropriate reference genes for RT-qPCR analysis in Ziziphus jujuba Mill. Sci Hortic Amst 197:166–169. [Google Scholar]
- Zhang FJ, Wang ZQ, Dong W, et al. (2014) Transcriptomic and proteomic analysis reveals mechanisms of embryo abortion during chrysanthemum cross breeding. Sci Rep 4:6536. 10.1038/srep06536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zu SH, Jiang YT, Chang JH, Zhang YJ, Xue HW, Lin WH (2022) Interaction of brassinosteroid and cytokinin promotes ovule initiation and increases seed number per silique in Arabidopsis. Journal of integrative plant biology 64(3):702–716. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.