Embryological study on apomixis in Kentucky bluegrass
Apomixis can fix any genotype across generations, and Kentucky bluegrass is a natural apomictic species closely related to major crops. The existing research on apomixis in Kentucky bluegrass has focused on various aspects, particularly cell embryology. In the late 1970s, Yong et al.[32] used the techniques of pistil cleaning and thick sectioning to observe the apomixis of Kentucky bluegrass. Subsequently, Marshall et al.[33] noted that the apomixis of Kentucky bluegrass was mainly aposporic. In addition, Kentucky bluegrass possesses a small number of parthenogenetic embryos, cocellular embryos, and antipodal embryos[11]. The occurrence of apospory in Kentucky bluegrass is caused by the expansion of one or more somatic cells in the ovules, which restores reproductive ability, and the unreduced embryo is then produced by several mitotic events[34, 35]. The number of AICs is mostly one or two, though three are occasionally observed[36]. In this process, sexual embryos and apomictic embryos can coexist in the same ovule[37] and ultimately produce seeds with two or three embryos[11]. These previous reports support our research results, which demonstrated that during the embryonic development of Kentucky bluegrass, one or more nucellar cells located near the MMC under the nucellus epidermis became enlarged and exhibited an obvious nucleus, and their meristematic ability was restored. They then further developed to form AICs. The AICs developed into unreduced and unfertilized apomictic embryos, which could coexist with the sexual embryos.
Transcriptome analysis of apomixis in Kentucky bluegrass
Plant apomixis was first discovered in Alchornea ilicifolia in 1841[48]. Since this discovery, there has been much research on the classification, genetic evolution, embryonic development, morphogenesis, physiology, biochemistry, and breeding of apomixis[38, 39]. Researchers have long tried to collect natural apomixis resources and analyze their biological characteristics and genetic mechanisms, as well as introduce key genes into crops to make use of apomixis fixed heterosis in crops. However, due to the complexity of species, occurrence processes, and formation mechanisms, as well as the diversity in apomixis types, it is extremely difficult and time-consuming to mine and clone related genes using traditional breeding methods. Thus, a reverse genetic strategy based on transcriptomics, which is used to screen and identify candidate genes and then verify them by experimental mapping and functional analysis, has become the core of apomixis studies[23]. Furthermore, transcriptome methods are applicable to all types of plants and traits[40].
In this study, a reference transcriptome of wild germplasm materials of Kentucky bluegrass with significant differences in apomictic rates in Gansu, Chain, was constructed using RNA-seq, and the gene expression changes and differences between the two genotypes were studied. However, the ovary of Kentucky bluegrass is very small, and RNA is also very easily degraded, and thus the biological sample we used included the entire spikelet of Kentucky bluegrass, which is composed of 3–4 florets, including the glumes, lemma, palea, lodicules, pistil, and stamen. Therefore, our transcriptome originates from a variety of cell types, including somatic cells and male and female germ cells from the premeiosis to anthesis stages. Based on this, we believe that this database will be very helpful in identifying any transcripts expressed in the flowers of Kentucky bluegrass that can be identified at a detectable level. However, the spatial and temporal specificity of gene expression will need to be evaluated by other experiments such as qRT-PCR[23].
Regulation of TFs in the ovule development of apomixes: During plant growth and development, the expression of genes in cells is time-specific and space-specific, which leads to differentiation and differences among different cells. One of the main reasons for this phenomenon is the regulation of TFs at the transcriptional level[41]. TFs are protein molecules that can specifically bind to cis-acting elements in the promoter region of eukaryotic genes to regulate the expression of target genes. They play an important role in plant growth and development, regulation, hormone signal transduction, and stress responses[42]. By comparing the transcripts of obligate apomixis and sexually reproductive Boechera plants, Shah et al.[43] showed that the dysregulation of TFs is one of the main components of apomixis-specific transcripts. Before this, scientists noted that no obvious developmental pathways or time changes could simply explain the shift from sexual reproduction to apomixis. The overexpression of TFs in apomixis-specific genes indicates a large-scale change in the regulation of apomixis ovules[44, 45]. Sharbel et al.[45] suggested that TFs are important motifs for the differential expression between sexual and apomictic ovules, and their mediation of the inhibition of gene expression is a characteristic of the development of sexual ovules to apomictic ovules. These studies suggest that TFs may regulate the transformation of plants from sexual reproduction to apomixis[46], and thus we further analyzed the differential expression of TFs in the two genotypes of Kentucky bluegrass. The results showed that a total of 84 TF families were differentially expressed in the two genotypes of Kentucky bluegrass, of which 52 were common across all four stages.
Present studies on plant apomixis have reported that many TFs are involved in the regulation of apomixis. Kumar et al.[40] reported a transcriptome analysis of ovule development during nucellar embryogenesis in citrus (Citrus reticulata), and the results showed that MYB and WRKY were key TFs and were differentially expressed in single and polyembryonic ovules as well as between the pre-anthesis and post-anthesis stages of polyembryonic varieties. In addition, there are many reports that TFs such as B3[47], NAC and bZIP[43], MADS[45], OFP[46], TCP and Alfin-like[40] and bHLH, ARF and AUX/IAA[44] can regulate apomixis. These TFs have diverse functions and regulate plant responses to biotic and abiotic stresses in addition to many aspects of plant growth and development. Similarly, in our results, these TFs were differentially expressed in different genotypes and were time-specific, indicating that they may play an important role in regulating the apomixis development of Kentucky bluegrass. In addition, we also found that TFs such as FAR1, C2H2, TRAF, SWI/SNF-BAF60b, and CPP were differentially expressed in different genotypes. They can regulate different genes and perform a variety of regulatory functions. For example, FAR1 can act as a Cdk inhibitor (CKI) to resist the transformation of G1/S, block cells in the G1 phase and then regulate the cell cycle[48]. It is speculated that they may be involved in the regulation of apomixis in Kentucky bluegrass. However, since our sampling included the whole spikelet, we suspect that many of the TFs we have screened may be involved in flower development, cell division, cell evolution, reproductive organ development, and so forth.
Disorder of meiosis leads to apomixes: The loss or failure of meiosis is one of the three basic events involved in apomixis[15]. Apomixis in Kentucky bluegrass is mainly aposporic, which is usually characterized by the failure of meiosis, followed by differentiation into one or more unreduced embryos from nucellus cells[49]. In Arabidopsis thaliana, apomixis also occurs through important meiotic genes, such as DYAD/SWITCH1 (SWI1), which controls meiotic chromosome tissue. Additionally, the MIME (mitosis instead of meiosis) of a triple mutant, which could be achieved in A. thaliana by combining different combinations of meiotic mutants spo11-1, osd1, and rec8, exhibits characteristics similar to diplospory[50, 51]. Although the role of these genes or potential mutations in asexual reproduction in natural apomixis, such as in Boechera, has not been documented, a comparative MMC study between A. thaliana and Boechera indicated that meiotic genes in the MMC of Boechera such, as PARTING DANCERS and SPO11-2 (meiotic recombination protein), may be downregulated compared with those in the MMC of sexual A. thaliana[52]. Interestingly, Okada et al.[53] found that meiotic genes such as DMC1 (DNA meiotic recombinase 1), RAD50 (RAD50 double strand break repair protein), and other genes required for meiosis in A. thaliana, including SWITCH1/DYAD, MULTIPOLAR SPINDLE1, SPOROCYTELESS/NOZZLE, REC8 (meiotic recombination protein) or similar transcripts, were not found in AIC, EAE, or SO cell types. It was previously reported that the meiosis control gene DMC1 was expressed in both sexual and apomixis MMCs in Hieracium spp., but could not be detected in AI cells by situ hybridization[54]. However, the results of this study indicated that these meiosis-related genes were differentially expressed between two genotypes, which may be due to the fact that our samples contained a variety of cellular tissues and originated from multiple time points. In addition, this discussion also implies the difference between diplospory and apospory. The former can be realized through loss of meiosis on the premise of avoiding fertilization, while the ability of nucellus cells to develop into mature embryo sacs through mitosis is more important for apospory.
Regulation of plant hormone signal transduction in apomixis development: Plant hormones directly or indirectly regulate the determination of cell fate, including embryogenesis and postembryonic development, and thus it is speculated that plant hormones may also affect the development of apomixis ovules[55]. In this study, many plant hormone-related genes were differentially expressed between the two genotypes, such as the auxin-related genes SAUR (SAUR-like auxin-responsive protein family), ARF (auxin response factor), IAA (auxin-responsive protein), AUX1 (AUX transcriptional regulator family protein1), cytokinin synthesis-related genes IPT (tRNA dimethylallyltransferase 2), and CKX (cytokinin dehydrogenase 7), indicating that the expression of apospory in Kentucky bluegrass may be accompanied by a response to hormone stimulation. Giulio et al.[16] found that the DEGs between sexual reproduction and apomixis in Hypericum perforatum included genes involved in cytokinin biosynthesis, several auxin response factors, and SAUR-like proteins. After further analysis, it was noted that some genes related to hormone perception and dynamic balance, including cytokinin, auxin, and brassinol, were differentially expressed in the apomictic pistils. Schmidt et al.[52] reported that compared with mature gametophytes, the upregulation of cytokinin degradation genes was detected in apomictic genotypes, while the egg cell was marked by the activity of genes that lead to cytokinin modification. At the same time, it was found that the genes related to auxin signal transduction were enriched in AIC, but not expressed in sexual MMC[13]. We also determined the endogenous hormones of two genotypes and found that ZT and IAA may have significant effects on reproductive patterns. The change in the regularity of GA3 and ABA was not obvious. We speculate that the differential expression of genes related to hormone signal transduction disrupts the balance of endogenous hormones in Kentucky bluegrass. The unbalanced hormone content affects plant morphogenesis, and so the two Kentucky bluegrass genotypes exhibit different reproductive patterns. In summary, our analysis of the results shows that the developmental pathway of apospory in Kentucky bluegrass is accompanied by the expression and regulation of multiple genes involved in hormone homeostasis.
Regulation of apomixis in Kentucky bluegrass by embryogenesis-related genes: The successful development of embryos is necessary in both sexual reproduction and apomixis. At present, it has been reported that SERK (somatic embryogenesis receptor-like kinase), LEC (laryngotracheo esophageal cleft), and WUS (WUSCHEL-related homeobox) are the main genes that regulate the development of apomixis. The expression of SERK is closely related to apomixis. Tucker et al.[56] studied the expression of AtSERK1 in sexual and apomictic Hieracium umbellatum, and the differential activity of SERK was also observed between sexual and asexual genotypes of Kentucky bluegrass[25]. In our study, the expression of SERK1 was upregulated in LN during the meiosis stage, while SERK2 was upregulated at the anthesis stage. WUS can promote the transformation from vegetative to embryogenic cells. In the absence of exogenous auxin, it binds with other genes that can induce embryo formation, such as LEC1, LEC2, and SERK1, to regulate the formation and development of embryos[57]. LEC regulates embryo formation by establishing an intercellular environment suitable for embryo development and is an important regulator for inducing embryo morphogenesis and controlling embryo development[58]. Compared with GN, we found that WUS1 and WDR23 (WD40 repeat-containing protein), as the homologous genes of LEC14B, were upregulated and time-specific. Comprehensive analysis showed that the upregulated expression of genes related to somatic and zygotic embryogenesis may regulate the occurrence of apomixis in Kentucky bluegrass.
Regulation of apomixis by the stress response pathway: One study showed that stress can induce the occurrence of somatic cell embryogenesis[59]. Many studies have observed a close correlation between stress-induced signal pathways and somatic embryogenesis induction[60-62]. One of the key steps in apospory is to confer nucellar cells embryogenesis ability. Our results showed that the stress response plays an important role in the development of apomixis in Kentucky bluegrass, and a large number of TFs related to stress responses, such as MYB, WRKY, and NAC, were differentially expressed between the two genotypes. In fact, many studies have found that genes involved in reproduction are associated with responses to stimulation and defense processes[63, 64]. Kumar et al.[40] found that many genes related to abiotic stress were significantly expressed in citrus polyembryos, including chalcone synthase, heat-shock factor protein, and chaperone protein CLPB1, indicating that the overexpression of stress-related signals was the main characteristic of the pre-anthesis ovules of citrus polyembryonic varieties. Shah et al.[43] also found that apomixis exhibited better tolerance to osmotic stress than sexual reproduction in vitro and was accompanied by the significant upregulation of a subset of NAC genes. These results indicate that stress response pathways are involved in early preparatory events before apomixis transition[23]. Previous studies hypothesized that the stress response pathway regulates apomixis due to prolonged UV exposure, which may be responsible for activating meiosis-specific proteins and initiating double-strand break formation and DNA repair, thus increasing the frequency of recombination[65, 66].
Regulation of apomixis by epigenetic pathways and other factors: The genetics of apomixis is quite complex, often with large scale segregation aberrations[67]. Increasing studies have shown that mutations in the apparent pathway lead to phenotypes similar to apomixis phenotypes[68]. Studies on Paspalum notatum and Paspalumsimplex showed that the methylation level of genomic regions controlling apomixis was very high, and it was suggested that the relevant reproductive control factors were inactivated by methylation in Paspalum[69]. APOLLO (DNA cross-link repair 1B protein) is both an apomictic allele and a sexual allele[70], and so this gene has the potential to produce apomixis or sexual reproduction by regulation. In this study, APOLLO was up-regulated compared with GN during the meiosis and postmeiosis stages, indicating that this gene regulates the development of apomixis in Kentucky bluegrass. In European rape (Brassica napus), the existence, deletion, and distribution of BnMET1a-like DNA methyltransferase genes are related to methylation, while methylation genes can transform pollen growth and development to embryogenesis, and the resulting embryo and zygotic embryo exhibit similar DNA methylation and MET1a-like expression patterns[71]. Our results also demonstrated that MET1 (methyltransferase 1) was differentially expressed between the two genotypes. Thus it can be seen that the regulation of epigenetics may be a key focus for the continued study of natural apomixis[68]. AGO9 can combine with different types of small RNAs to play a role in the regulation pf apomixis[72], and other genes involved in RNA-guided DNA methylation pathways also play important roles[73]. In parallel, according to a study that reported that small RNAs can promote DNA methylation in plants[74], including in maize, mutations in the effectors of the small RNA pathway have been shown to be important for the reproduction of events similar to diplospory[75]. In addition, some studies have shown that the mechanisms dependent on small RNAs play an important role in the cell specification within the ovule[76, 77]. Therefore, the regulatory mechanisms of epigenetics on apomixis in Kentucky bluegrass from a small RNA perspective should be explored in the future.
In addition, SPL (squamosa-promoter binding protein-like protein)/NOZZLE, which is involved in the initiation of primary spore differentiation, also plays an important role in the regulation of apomixis[78, 79]. In this study, SPL11 and SPL15 were differentially expressed in the two genotypes, indicating that the expression of SPL/NOZZLE regulates the apomixis development of Kentucky bluegrass.
Molecular regulation model of the apomixis mechanism of Kentucky bluegrass
Based on the observation of the histological process of apomixis embryos and the study of gene regulation at the transcriptome level of wild germplasm materials of Kentucky bluegrass, we predicted the molecular regulatory model of apomixis embryogenesis in Kentucky bluegrass (Fig. 10). Compared with sexual reproduction, the apospory of Kentucky bluegrass avoids meiosis and double fertilization, which may be caused by temporal and spatial changes in related gene expression. Therefore, we speculate that the temporal and spatial changes in gene expression lead to the inhibition or imbalance of normal sexual reproduction in Kentucky bluegrass. As the plant needs to ensure the survival of offspring and produce seeds, it promotes the occurrence of apomixis. In essence, in facultative apomixis species such as Kentucky bluegrass, the occurrence of apomixis is mainly due to the disorder of the sexual reproductive pathway. The differential expression of stress response-related genes is extremely important for the acquisition of embryogenic cells and the change in hormone content. We speculate that the occurrence of events related to the stress response may be a preliminary preparation for apomixis in Kentucky bluegrass,which promotes the embryonic ability of the nucellus cells, causes a hormone imbalance, and leads to a disordered sexual reproductive pathway, which promotes the occurrence and development of apomixis. In summary, the regulatory mechanism of apomixis in Kentucky bluegrass is extremely complex, and its occurrence and development involve the coexpression of genes related to many aspects, such as the imbalance of meiosis, hormone signal transduction, embryonic development, stress response pathway, and epigenetic regulation pathway. In addition, we also found a number of new DEGs that have not been previously reported to be related to apomixis regulation, and so their apomixis regulatory mechanisms require further research.