Does the combined action of methylation and a maternally imprinted factor repress endosperm expression of paternal specific alleles of the zein multigene family?
--Castelli, S; Ciceri, P; Genga, A; Lazzari, B and Viotti, A

We are interested in elucidating the molecular mechanisms underlying the specific expression of those zein genes that undergo parental imprinting in maize endosperm. Previous data on zein gene modification and transcription (Bianchi and Viotti, Plant Mol. Biol. 11:203-214, 1988; Lund et al., Plant J. 8:571-581, 1995) suggested that endosperm-specific expression of some zein alleles occurs via parental imprinting. This could be mediated by the differential methylation of the maternal and paternal zein gene sequences, the hypomethylation state of the maternal copies correlating with their expression (imprinted state).

It is reported and generally accepted that mutations at the Opaque2 regulatory locus severely reduce the synthesis and the accumulation of the heavy chain zeins (H1 and H2 bands in SDS-PAGE). In analyzing many maize lines carrying different mutations at the O2 locus (Bernard et al., Plant Mol. Biol. 24:949-959, 1994) we confirmed the previous observation only for some of them. We noticed, however, that several lines showed the presence of the H1 band or only a moderate reduction of both H1 and H2 bands. Two dimensional analysis, carried out first by charge and then by size fractionation, evidenced that those o2 lines showing the presence of the H1 band in fact expressed three to four polypeptides with different charge.

These particular patterns allowed us to further and more accurately investigate the imprinting phenomenon by proper crosses between H1-plus (H1p) lines and H1-null (H1n) lines. A preliminary analysis using the H1p line, NYRo2-It, in reciprocal crosses with three different H1n lines (Rossmano2-R, W64Ao2-T or Mo17o2-R) indicates among the six possible crosses the absence of the H1 band only in the Rossmano2-R /NYRo2-It cross. This suggests the presence of a maternally imprinted factor (mif) that specifically represses the expression of those zein genes contributing to the H1 band. This was confirmed by the analysis of the reciprocal crosses between three other H1p lines (W22o2-It, 3316o2-It, Bianchio2-It) and two of the three H1n lines used in the previous experiments (Rossmano2-R and W64Ao2-T). Within the twelve resulting crosses only the ones that have as maternal contribution -to the endosperm complements- the Rossmano2-R genotype show the absence of the H1 polypeptides.

We should remark that the link between a specific modification state of certain zein alleles and a MIF is not an exclusiveness of the crosses between Rossmano2-R and the H1p-lines. In fact, the imprinting phenomenon has been observed also for the other size class zein polypeptides in reciprocal crosses involving other lines, with maintenance of the specific uniparental behaviour.

At present, the working hypothesis not only favours the occurrence of the trans-acting factor(s) MIF (line specific?) but also predicts the possibility that the zein modified genes mediate the action of the MIF by remaining in the default methylation state when they participate in the crosses as paternal contribution to the endosperm complements (Lund et al., Plant J. 8:571-581, 1995). 

Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors

Return to the MNL 70 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page