The formation of endosperm in angiosperms begins with synchronous divisions of the triploid endosperm nucleus unaccompanied by cell wall formation. Earlier studies have shown that multinucleate endosperm cell stage is followed by cellularization, its pattern being different in different groups of angiosperms. The identification and cloning of a number of genes concerning endosperm development in Arabidopsis have generated renewed interest in endosperm development, creating the need for updating research findings in this field.
The publication of a review article in the February, ‘03 issue of Curr Opin Plant Biol (vol. 6:42-50), by F. Berger at ENS-Lyon, France, has filled up that void.
The author begins the article by describing the importance of endosperm in world’s 60 per cent food supply and how its in-depth developmental studies may help in engineering endosperm development for enhancing food production.
Effect of certain mutations on endosperm development: The author describes the formation of a multinucleate endosperm chamber with well defined sectors along the anterior-posterior axis. Some recent studies have shown that the nutritional status of the cell decides the pattern of endosperm development. For instance, reduction in the levels of hexose supply, affects the endosperm development of the maize mutant miniature1 by slowing down the rate of mitosis, seemingly without affecting endoduplication process. On the other hand, in wee1 mutants, cdc (cell division cycle) activity is suppressed and as a result entry of the cell into mitosis is blocked.
Role of ttn class of genes in the developmental process: In the endosperm of the maize mutant miniature1, mitotic activity is reduced but the endoduplication process does not appear to be affected by this mutation. Recent study has also revealed that the TOR (Target of Rapamycin) gene plays a vital role in controlling G1/S phase and the mutation of this gene arrests free nuclear division of the endosperm during the initial developmental phase. Next, the author describes the titan (ttn) class of embryo-lethal mutants, in which endosperm cells show enlarged nuclei. In some yeast mutants, namely, in ttn3, ttn7 and ttn8, structural chromosome proteins become defective, affecting chromosome condensation and chromatid separation.
PILZ, MCM7, Prolifera, Keule and Spätzle genes in endosperm development: The author also describes a subclass of titan mutants called pilz. Its wild type allele PILZ encodes components of tubulin-folding complex that are conserved in eukaryotes. Another member of the titan class is MINICHROMOSOME MAINTENANCE7 (MCM7) gene PROLIFERA. The mutants of this gene have enlarged nuclei in the endosperm, reflected in a complex variable phenotype. PROLIFERA is expressed in the endosperm and appears to control the onset of DNA replication. In Keule mutants, cellularization is normal in spite of defective cytokinesis, while
in spätzle mutants, it is reversed i.e., cytokinesis is normal but cellulari- zation does not occur. Some recent studies have indicated that the eighth cycle of synchronous mitosis in the peripheral endosperm is linked with cellularization in the Arabidopsis endosperm. Thus , as in the Drosophila embryo where a similar cellularization process occurs, the onset of endosperm cellularization might be control by a threshold nucleocytoplasmic ratio
An important class of Arabidopsis mutants showing seed development without fertilization: Fertilization-independent seed (fis) mutants influence the patterning of the posterior zone of Arabidopsis endosperm. Endosperm development in the posterior domain of these mutants is abnormal and their endosperm is marked by the presence of nodules and cysts. Interestingly, the proteins encoded by FIS genes are similar to that encoded by Polycomb Group (PcG) class of genes. The PcG proteins, isolated in Drosophila, were found to repress the expression of zygotic gene, thereby ensuring maintenance of patterns established by maternal controls. FIS genes are expressed in the mature embryo sac. The activity of MEA and FIS2 continues till the endosperm is at the middle of the multinucleate stage; thereafter, such an activity is only confined to the posterior pole. FIE (Fertilization Independent Endosperm), on the other hand, is expressed in all endosperm domains tapering of its activity during the late free nuclear phase. Homologues of the PcG genes have been reported in maize. Such a discovery opens up the possibility that FIS homologues may be involved in patterning the more complex cereal endosperm. Such a piece of information may be helpful to compare the different endosperm domains in cereals and in Arabidopsis.
Development of endosperm without fertilization; paternal silencing of many loci: Mutations in FIS genes produce characteristic maternal effects on the embryo sac (female gametophyte) in that the endosperm develops without fertilization i.e., autonomously. Whether the heterozygotes (fis/+) are self-fertilized or cross-fertilized with wild-type pollen, the number of seeds does not vary. Moreover, transmission of the fis allele through the female parent, occurs at a very low frequency. In the fis/fis genotype, endosperm development continues unabated without fertilization. This may be due to silencing of the paternal copy in the endosperm. With the help of reporter constructs, silencing for 20 other loci has been demonstrated. Although not proven conclusively, the development of endosperm at the early stage is attributed to paternal silencing of many loci.
Relationship between paternal silencing and DNA methylation: It has been shown that paternal silencing may be abolished in fis mutants,, if the latter are pollinated by individuals with a reduced level of DNA methylation. Such a treatment restores seed viability in fis mutants,, which in a normal situation do not form viable seeds. The restoration of viability has been shown to be due to demethylation caused by the presence of FIE (fertilization independent endosperm) gene carried by the pollen of the male parent. Demethylation seems to act by derepressing silenced genes. Another striking observation in this connection is the role of DDM1 (decrease in DNA methylation) gene. The mutated allele ddm1 reduces DNA levels. The proof of this comes from the observation that ddm1 pollen with reduced levels of DNA methylation rescues the maternal effect of mea by functionally reactivating paternally inherited MEA alleles during seed development.
The author describes the role of a recently identified gene, DEMETER, which encodes a DNA glycosylase and transcriptionally activate MEA. According to the results of a recent study, the absence of DME expression in the pollen leads to paternal imprinting of MEA. About the two recently identified mutants called capulet, the author states that in both the mutants, the endosperm is arrested at a very early stage. The author is of the opinion that with more rigorous search applying more direct screens, many more mutants with gametophytic maternal effects will come to light.
Evolutionary origin of endosperm: The author describes the development of the 7-nucleate angiosperm embryo sac consisting of six haploid cells, distributed in two sets of three at the anterior and posterior end, and the diploid central cell. The set located at the anterior end consists of an egg cell flanked by two synergids and the other set of three forms what is called antipodals at the opposite i.e., chalazal end. Of the two male gametes, released by the pollen tube, one fertilizes the egg and the other fertilizes the diploid central cell. After fertilization, the egg is turned into the diploid zygote and the central cell into the triploid endosperm.
The author puts forward his arguments whether to consider the endosperm as gametophytic or sporophytic. In Gnetales, the two male nuclei released by the pollen tube fertilize two nuclei in a multinucleate female gametophyte, giving rise initially to two embryos of which only one survives. The occurrence of multiple embryos in Gnetales leads one school of thought to propose that the ancestor of the central cell in flowering plants was probably haploid and it gave rise to a diploid zygotic endosperm. In fact, there are some primitive angiosperm families such as Nymphaeales in which the central cell is haploid. However, because of some basic differences between the pattern of endosperm development in angiosperms and Gnetales, and the fact that the latter are now considered to be more distantly related to angiosperms than what was hitherto believed, this hypothesis might be no longer tenable. The author also points out that in angiosperms ploidy of the central cell varies from one to five and does not give a clue whether the endosperm tissue is to be regarded sporophytic or gametophytic.
The author points out that a gametophytic female control of endosperm development is established through the silencing of the paternal alleles of many loci. The author believes that heterochronic mutations (a kind of mutation which alters the development pattern through regulation of timing and duration of phases of development), may have changed the development of the female gametophyte in such a way that gametogenesis occurs very early and that gametophytic development is prolonged after fertilization into the endosperm. Such type of mutations may have been responsible as well for certain types of apomixis (development of seeds without fertilization) in flowering plants as illustrated by Arabis holboelii (a member of the Brassicaceae family). In this species side by side with sexual reproduction, aneuploid and triploid apomictic seeds develop without fertilization. Such natural events seem to show that usual and unusual types of evolutionary forces may be operative in determining the development and fate of endosperm.
The author concludes on a note of confidence that further insight into various aspects of endosperm development, including epigenetic controls, will be accomplished when the results of earlier studies on endosperm development in cereals will be combined with those of future genetic studies of the model plant Arabidopsis, where a number of genes related to endosperm development have already been isolated. The author foresees that in not too distant future, common grounds for the epigenetic control of both plant and animal development will be unraveled.