Endoreduplication can be affected by a number of physiological and environmental factors. In a cross between two inbred lines, the endoreduplication pattern of the endosperm in the F1 generation generally resembles that of the maternal parent (Kowles et al. 1997; Dilkes et al. 2002). This "maternal effect" could be due to gene dosage, in that two copies of the maternal and one copy of the paternal genome contribute to the genetic makeup of the endosperm. Alterations to this 2 : 1 ratio cause deleterious changes during development of the endosperm (LeBlanc et al. 2002). A maternal excess results in precocious endoreduplication and a shortened mitotic phase, although it does not alter the degree of endoreduplication. Excess of the paternal genome, on the other hand, delays entry into endoreduplication, not by prolonging the mitosis phase, but by causing an extended mitotic hiatus prior to the switch to endoreduplication. Also, DNA amplification is severely reduced in all cells that are committed to endoreduplication (LeBlanc et al. 2002).
A possible explanation for the maternal effect on endosperm development in general, and endoreduplication in particular, could be due to imprinting, in which specific paternal genes are silenced and only their maternal counterparts are expressed. Genomic imprinting is a phenomenon observed in plants and animals, often during the reproductive phase (Guitton and Berger 2005; chapter by Penterman, in this volume). In Arabidopsis, a group of transcrip-tional regulators belonging to the Polycomb group complex, including FIS, FIE, and MEA, are imprinted during early seed development (S0rensen et al. 2001). After fertilization, only the maternal alleles in the Polycomb group genes are expressed in the endosperm, while the paternal alleles are silenced, at least during the early stages of development. A homolog to Arabidopsis FIE gene has been found in maize, where it occurs as two copies, FIE1 and FIE2 (Danilevskaya et al. 2003). Presumably, the function of these genes in maize endosperm is similar to that in Arabidopsis, in that the paternal FIE1 allele is not expressed, while that of FIE2 is initially silenced but is expressed 5 DAP. Undoubtedly, there are genes involved in endoreduplication that are imprinted, thereby contributing to the described maternal effect.
Hormones appear to play an important role in the induction of endoredu-plication in the endosperm. The levels of several different phytohormones change during its development. It is known that cytokinins stimulate cell division, and indeed during the mitotic stage of maize endosperm development, the levels of cytokinins, such as zeatin and zeatin-riboside, increase sharply and peak at 9 DAP, and then abruptly decline between 9 and 15 DAP (Lur and Setter 1993). Conversely, auxin levels are low when cytokinins peak, and then start to increase at 10 DAP, when the cytokinins begin to decline. Thus, a high auxin/cytokinin ratio coincides with the onset of endoreduplication (Lur and Setter 1993; Cheikh and Jones 1994). Apparently, auxin is involved in the induction of endoreduplication, but not its continuation. Application of exogenous 2,4-dichlorophenoxyacetic acid (2,4-D), a synthetic auxin, during the mitotic phase caused precocious DNA amplification, but when applied during the peak of endoreduplication (around 12-15 DAP), it did not increase DNA content compared to the control (Lur and Setter 1993). When the anti-auxin, p-chlorophenoxyisobutyric acid (PCIB), was applied to the endosperm during the period of endogenous auxin increase, i.e., the onset of endoreduplication, there was no difference in endosperm DNA content compared to the control. But when PCIB was applied during the peak of endoreduplication, DNA accumulation was reduced by 20-25% (Lur and Setter 1993). PCIB has been shown to inhibit auxin-stimulated events rather than auxin production itself, indicating that it is auxin-induced events that cause endoreduplication and not auxin directly.
Endoreduplication in maize endosperm is inhibited by environmental stresses, such as water deficit and heat. However, it appears that the reduction in endoreduplication is more a result of perturbation during early events of development, rather than a direct consequence of the inhibition of the en-doreduplication cell cycle per se. When the kernel is treated with high heat (Engelen-Eigles et al. 2000, 2001) or deprived of water (Artlip et al. 1995; Setter and Flannigan 2001) during the mitotic phase, there is a marked reduction of both cell number and DNA content; however, when either form of stress is applied during active endoreduplication, the reduction in DNA amplification is not as dramatic.
The reduced level of endoreduplication as a consequence of physiological and genetic perturbations during the mitotic phase shows that there is a relationship between the mitotic and endoreduplication phases of endosperm development. Of 35 maize mutants categorized as defective kernel (dek) mutants, all but one had endosperm with both a reduced cell number and a lower level of endoreduplication (Kowles et al. 1992). This suggests a regulatory mechanism that coordinates both the mitotic and endoreduplication phases of development. However, the mitotic cell cycle is not necessarily coupled with that of endoreduplication, as one dek mutant showed a 70% reduction in cell number, but little or no reduction in DNA content when compared to the wild type. Another mutation that shows uncoupling of the mitotic and endoreduplication phases of endosperm development is miniaturel (mnl) (Vilhar et al. 2002). Maize plants with this mutation produce small kernels with endosperm that is only 20% the volume of wild type. Mnl encodes a cell wall invertase, INCW2, that provides hexose sugars to the developing endosperm. The mnl mutation limits the amount of energy available for endosperm development, causing a reduction in mitotic activity and inhibition of cell expansion, but it does not severely affect endoreduplication. mnl starchy endosperm cells go through at least six rounds of endoreduplication by 16 DAP and have the same ploidy distribution as wild-type endosperm.
Was this article helpful?