One of the mechanisms that can produce “clones” of mother plants is apomixis, which is a mode of asexual reproduction through seeds that have been identified in over four hundred plant species
[7][8]. In apomicts, meiosis and fertilization are modified or completely absent. Consequently, the embryo is formed without prior meiosis (by apomeiosis) and fertilization (i.e., by parthenogenesis), while endosperm development occurs either autonomously, i.e., without fertilization, or pseudogamously (by fertilization of the central cell)
[6][8][9][10][11][12][13]. During double fertilization, intrinsic to all Angiosperms, the pollen tube transports two sperm cells (male gametes) to the embryo sac, one of which fertilizes the egg cell (female gamete) and the second one fertilizes the central cell, giving rise to a 2n embryo and a 3n endosperm, respectively. In sexual species, departures from the 2 maternal:1paternal (2m:1p) genome ratio in endosperm nuclei result in seed abortion. While in pseudogamous apomicts, endosperm ploidy varies according to the ploidy of the sperm and central cell. Deviations from the 2m:1p genome ratio sometimes occur, demonstrating that the apomictic system is more resilient compared to sexual species
[14].
It was shown that
APOLLO has several polymorphic sex- and apo-alleles
[18][19]. Apo-alleles of this gene is missing in sexual ovules and are up-regulated in apomeiotic ovules in
Boechera plants. It was reported that genomes of apomictic plants are always heterozygous carrying at least one of the apo-alleles, while sexual genotypes were always homozygous for sex-alleles
[18][19]. While sex-alleles were upregulated during meiosis in sexual plants and downregulated at the same stage in apomictic
Boechera plants. Evolutionary
APOLLO gene analysis presented in this study shows that sexual and apomictic species of
Boechera are clustered in different clades in the phylogenetic tree based on the multiple protein sequence alignment. This could be partly because the
APOLLO apo-alleles present in the genomes of apomictic species could acquire a new function
[19]. Expression analysis of the
APOLLO presented here showed that the expression levels of the gene dramatically decreased during meiosis in gynoecia of both apomictic and sexual species; however, after meiosis and fertilization, the expression of the
APOLLO was upregulated in apomictic siliques compared to sexual ones.
2. Phylogenetic and Expression Analysis of CENH3 and APOLLO Genes in Sexual and Apomictic Boechera Species
Apomixis via clonal seeds produces offspring, that are genetically identical to the maternal plant. Understanding of the genetic components that regulate apomixis is very important for studying plant development and evolution, in addition, the introduction of apomixis in agricultural plants would allow a long-term fixation of complex genotypes, including F1 hybrids, often used in agriculture. However, the molecular mechanisms underlying apomixis are poorly understood. Namely, the factors inducing avoidance or modification of meiosis (apomeiosis) and parthenogenesis. To study the genes regulating meiosis and embryogenesis in comparison of sexual vs. apomictic plants we use a convenient model plant from the
Boechera genus, that comprises species naturally reproducing both by sexual and apomictic ways and show features of hybrid origin
[20].
The
CENH3 gene plays an important role in cell divisions and genome elimination when mutated. Mutations in
CENH3 of
Arabidopsis thaliana cause disturbed meiotic chromosome segregation
[21][22] that was also used to induce genome elimination in
A. thaliana and rice. In hybrids of
cenh3 mutant lines with diploid wild-type plants, the
cenh3 line genome was eliminated
[23].
CENH3 is a single-copy gene and its structure is almost identical among the seven studied
Boechera species irrespective of the reproduction (sexual or apomictic) mode. Polymorphic sites were mostly found at the N-tail protein regions, although
B. retrofracta and
B. arcuata had one site at 91 a.a. and two polymorphic sites at 67 a.a. and at 96 a.a. respectively, in conservative Histone H3/CENP-A domain. Variability within the N-terminal tail might lead to apomeiosis, since it influences the chromosome segregation in meiosis
[21][24], although this assumption still needs to be tested on mutant lines with the replacement of the corresponding polymorphic sites. Still, the similarity index between all studied
Boechera CENH3 was ≥97% (both at nucleotide and protein levels). Furthermore, in the
CENH3 phylogenetic tree, performed by multiple alignments, all studied
Boechera species were clustered into the same clade, although being very close to other Brassicaceae species from
Arabidopsis,
Eutrema,
Cardamine genera. The
CENH3 expression profile analysis showed that during meiosis expression levels of
BsCENH3 in gynoecium of sexual
B. stricta was more than twice as high compared with
BdCENH3 in gynoecium of apomictic
B. divaricarpa. By the meiosis time, expression levels of both these genes dramatically dropped in gynoecia. After pollination, the expression
CENH3 significantly increased in
B. stricta siliques by 4th DAP still remaining low in siliques of
B. divaricarpa. The lower expression levels before and after pollination in
B. divaricarpa (ES9) could indicate a feasible
CENH3 role in apomeiosis and initiation of parthenogenesis. In the meiotic anthers of
B. divaricarpa, the expression of
CENH3 was 1.5 times higher than that in
B. stricta, which may be associated with apomeiosis during pollen maturation. After meiosis, the level of gene expression dropped to zero in the anthers of both species.
In sexually reproduced plants, two sperms enter the embryo sac, while one of them fuses with the central cell nucleus that further forms endosperm, the second fertilizes the egg, in contrast, during pseudogamous apomixis, the embryo develops without fertilization, which is the cause of the “spare sperm problem”
[25][26]. In apomicts fertilization of the central cell with haploid sperm generally leads to a 4m:1p genomes ratio in endosperm cells, which causes seed abortion
[27]. Apomictic species can tolerate such deviations in endosperm via changing their imprinting systems; however, preventing the ‘spare sperm’ from fusing with the central cell nucleus might also be important
[28]. The fusion of reduced or unreduced ‘spare sperm’ to the central cell can potentially affect the parental genomic ratio. Therefore, it is not known how the central cell in apomicts can avoid fertilization by enhancing a polyspermy barrier for that ‘spare sperm problem’
[26]. Thus, different expression behavior of
CENH3 before and after pollination in
B. divaricarpa vs.
B. stricta gynoecium/silique might suggest its involvement in apomictic development such as elimination of the chromosomes from male gamete during endosperm development. However, this assumption requires further proof.
Promoter analysis revealed the presence of several MYB transcription factor binding and recognition motifs within the promoter regions of
BsCENH3 and
BdCENH3. These motifs in
Boechera CENH3 promoters might suggest the regulation of these genes by an MYB family protein. A study on the
Arabidopsis CENH3 promoter region and its regulation revealed two E2F binding regions at −163 and −115 sites
[27]. Further functional studies on the
Boechera CENH3 gene and its epigenetic and transcription regulation could elucidate a functional difference of
CENH3 between apomictic and sexual
Boechera.
Concerning
APOLLO, it is the only so far found gene comprising the alleles with apomixis-associated polymorphism in
Boechera species
[18]. Thus, the identification of
APOLLO apo-alleles might be used as molecular markers to spot apomictic individuals among the
Boechera species. Earlier it was shown that
APOLLO encodes the exonuclease NEN3 and suggested an evolutionary scenario where after a series of duplications one of the NEN3 protein copies of the
Boechera ancestors acquired an altered function leading to apomictic development from the sexual state
[19]. Moreover, it was demonstrated that apo-alleles are under a positive selection
[19]. In the current study, we retrieved five apo- and five sex-alleles from
[18] to screen the genomes of ten
Boechera species with different reproduction modes as well as other Brassicaceae species from the
Arabidopsis,
Eutrema,
Cardamine, and
Capsella genera to perform
APOLLO gene phylogenetic analysis. The results showed that sexual and apomictic
Boechera species are clustered into the separate sub-clades, while being very similar to each other and to other Brassicaceae species. As well as for the
CENH3 gene, functional studies of the
APOLLO alleles and their putative epigenetic and transcriptional regulation in
Boechera species is required in order to find out if there are functional differences between sexual and apomictic accessions. However, the latter will only be possible when good quality diploid genome assemblies of the studied species have been implemented. Investigation of
APOLLO expression in pre-and after meiotic gynoecium and anthers and after meiosis in siliques of sexual diploid
B. stricta and apomictic diploid
B. divaricarpa 1–5 DAP have been performed. The upregulation of the
APOLLO apo-allele in
B. divaricarpa apomicts after meiosis in the anthers and gynoecium and the decrease in the level of its expression with the onset of embryo sac formation may indicate a certain relationship between the expression of the
APOLLO apo-allele in apomicts and apomeiosis. This is consistent with the previously published data that the
APOLLO apo-allele is exclusively expressed in ovules of apomictic
Boechera species around the stage of meiosis
[18]. The upregulation of the
APOLLO sex-allele in anthers and gynoecium of sexual
B. stricta during and after meiosis and very low levels of its expression in apomicts is possibly associated with a connection of the sexual allele of the Aspartate Glutamate Aspartate Aspartate histidine exonuclease gene (
APOLLO) with meiosis and further postmeiotic processes. However, the increase in the expression of apo- and sex-allele in siliques on the 5th DAP in apomicts and sexual species respectively, requires further explanation. Upregulation of
APOLLO apo- and sex-alleles in meiotic anthers in apomictic
B. divaricarpa two folds higher than in sexual
B. stricta may indicate to a role of
APOLLO in apomeiosis or meiosis during pollen formation; moreover, the male gametophyte development in apomictic
B. divaricarpa was reported to produce both reduced and nonreduced gametes
[29][30].
In conclusion, detailed knowledge of the structure, phylogeny of genes related to apomixis and the dynamics of their expression can presumably help to better understand the nature and regulation of apomixis vs sexual reproduction and facilitate further study of the evolutionary, ecological, and population role of apomixis. However, for more accurate studies on the phylogeny and evolution of the Boechera species, it is necessary to have a good quality diploid level whole genome assembly for these species, which we are now actively working on.