University of Minnesota

Beyond Mendel: DNA methylation in certain hybrids

--S. M. Kaeppler, G. Holland and R. L. Phillips

Heterosis, paramutation, imprinting, transposable element expression, transgene expression variation, long-range cis effects, "modifying factors"--all of these seemingly enigmatic phenomena found in maize appear to be at least partially controlled by factors beyond the bounds of Mendelian theory. In this article we will describe some preliminary results found during studies on the stability of DNA methylation through sexual generations. We will then relate these results to an allelic "cross-talk" hypothesis which suggests that these diverse observations may be controlled by a fundamental underlying mechanism, a mechanism which may have importance in understanding gene expression at the next level beyond promoters, enhancers and transcription factors.

The study involved five inbreds (B73, Mo17, P3, A679, and A682) crossed in a diallel fashion to produce 25 genotypes. An important component of the crossing strategy was that a single plant was used as a male to produce a self as well as the four respective hybrids. This provided a way to assess the methylation state of the male plant. Five seedlings from each of the 25 genotypes were pooled and DNA from these samples cut with HpaII and HhaI; both enzymes are sensitive to CG methylation. Southern blots were probed with 26 single copy DNA sequences including sucrose synthase I, and RFLP probes from Brookhaven National Laboratories, University of Missouri-Columbia, and Native Plants, Inc. Seven of the probes showed non-parental methylation patterns in at least one of the hybrids (Table 1). In all seven cases the

Table 1. Sequences detecting variant methylation patterns in maize hybrids.
 

Probe	Hybrid(s) containing variant patterns
BNL5.09	B73 x Mo17
NPI112	B73 x P3, Mo17 x P3, A682 x P3
NPI114	B73 x P3, Mo17 x P3, A682 x P3
Sucrose Synthase I	B73 x Mo17
UMC54	B73 x P3, Mo17 x P3, A679 x P3, A682 x P3
UMC85	P3 x A679
UMC175	B73 x Mo17

allele transmitted through the male was the altered allele; the hybrid made in the opposite direction contained the expected methylation pattern. An alteration occurred in all four possible hybrids only with probe UMC54. This indicates that in at least six of the seven cases the changes in methylation occurred after fertilization. B73 was involved as a female in all of the cases of variation; in four of the seven cases other female parents also were involved. These data indicate that there may be a post-fertilization modification of alleles, that the parental origin of the allele is important, and that this situation occurs frequently since random probes testing relatively few potentially methylated sites were used. A related situation is our analysis of tissue culture-derived lines. In these studies we have found methylation to be quite stable upon selfing but subject to change upon crossing back to the non-cultured source (unpublished).

This type of mechanism may have important implications regarding gene expression, but at this point we can only speculate. However, a "cross-talk" hypothesis presented by Monk (TIG 6:1120-1124, 1990) suggests that similar mechanisms may be acting in a wide range of organisms. The hypothesis states that homologous regions are "compared" after fertilization to determine the degree of genetic and epigenetic relatedness. Detection of differences may result in sequence change or epigenetic modification. Rivin, C (personal communication) has found that the copy number of tandem repeats in hybrids is often not the mean of the two inbreds as expected. Perhaps this cross-talk involves more than epigenetic modification and is ultimately manifested as chromatin or chromosome structure alteration. The finding of long-range cis effects by Schwartz, D (reports at 1990, 1991 Maize meetings) may imply that this mechanism occurs in a directional fashion along a chromosome and/or that homologous sequences at different chromosome positions are compared as frequently as homologous sequences at a given chromosome position.

As we try to rebuild genomes using DNA transfer technology, it will become increasingly important to understand the influence of position effects, chromatin and chromosome structure, and epigenetic modification on gene expression. In addition to at least partially explaining the phenomena listed at the beginning of this article we may find that these types of mechanisms may allow heritable ways for plants to respond to harsh environments (e.g. tissue culture, nutrient stress) without altering their basic genetic code. Outcrossing among plants could then minimize the effect of drastic changes in a single plant if that plant was uniquely affected by the stress.

Return to the MNL 66 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage