Comprehensive analysis of imprinted genes in maize reveals allelic variation for imprinting and limited conservation with other species

In plants, a subset of genes exhibit imprinting in endosperm tissue such that expression is primarily from the maternal or paternal allele. Imprinting may arise as a consequence of mechanisms for silencing of transposons during reproduction, and in some cases imprinted expression of particular genes may provide a selective advantage such that it is conserved across species. Separate mechanisms for the origin of imprinted expression patterns and maintenance of these patterns may result in substantial variation in the targets of imprinting in different species. Here we present deep sequencing of RNAs isolated from reciprocal crosses of four diverse maize genotypes, providing a comprehensive analysis that allows evaluation of imprinting at more than 95% of endosperm-expressed genes. We find that over 500 genes exhibit statistically significant parent-of-origin effects in maize endosperm tissue, but focused our analyses on a subset of these genes that had >90% expression from the maternal allele (69 genes) or from the paternal allele (108 genes) in at least one reciprocal cross. Over 10% of imprinted genes show evidence of allelic variation for imprinting. A comparison of imprinting in maize and rice reveals that 13% of genes with syntenic orthologs in both species exhibit conserved imprinting. Genes that exhibit conserved imprinting between maize and rice have elevated nonsynonymous to synonymous substitution ratios compared with other imprinted genes, suggesting a history of more rapid evolution. Together, these data suggest that imprinting only has functional relevance at a subset of loci that currently exhibit imprinting in maize.Imprinting describes a biased expression of alleles that depends upon the parent of origin. Imprinting is observed in both flowering plants and mammals (1–3) but there are differences in the mechanisms and organization of imprinted genes in these organisms (1, 4). In plants, imprinting is most prevalent in the endosperm, a triploid tissue that contains two maternal genomes and a single paternal genome (5). The endosperm provides an energy source for germinating seeds and, as the majority of harvested grain consists of endosperm tissue, a major source of calories in the human diet. A better understanding of imprinting will shed further light on epigenetic gene regulation and endosperm development and could provide an avenue for altering plant reproductive processes or seed quality.Despite a widespread interest in imprinting and its potential importance, the function of most imprinted genes is not well characterized in plants, and imprinting has only recently been assayed on a genome-wide level. Imprinting is reflected in parentally biased allele-specific expression in the endosperm tissue of intraspecific reciprocal hybrids. A quantitative method for detecting the relative expression of two alleles that have nearly identical sequences is required to find such an effect, traditionally limiting analysis to a handful of imprinted genes identified based on phenotype or through targeted analyses (6–8). The implementation of deep sequencing of RNA molecules (RNA-seq) has allowed detection of additional imprinted genes (9–14). In each of these studies, allele-specific expression levels were monitored for a single cross of two parents in Arabidopsis, maize, or rice. This allowed for the analysis of imprinting in 50–58% of genes expressed in endosperm tissue. In each species there is evidence for several hundred imprinted genes with similar numbers of maternally expressed genes (MEGs) and paternally expressed genes (PEGs), but comparisons among flowering plants (8, 11, 12, 15) have revealed limited overlap in the genes that are imprinted among species.There has been considerable speculation on the mechanisms that might lead to the origin of imprinted expression as well as the evolutionary mechanisms that would lead to the maintenance of imprinting (8, 16–18). Recent studies suggest that imprinting may arise due to programmed release of silencing marks in specific nuclei of male and female gametophytes (19, 20). Plant gametophytes are multinucleate structures. The male gametophyte includes a vegetative nucleus and two sperm nuclei. The female gametophyte has multiple cells including the haploid egg cell (which is fertilized by a sperm nuclei to generate the embryo) and the diploid central cell (which is fertilized by a sperm cell to generate the endosperm) (21). The loss of DNA methylation before fertilization leads to an epigenetic asymmetry in the endosperm because the maternal genomes (from the central cell) have been demethylated, whereas the paternal genome (from a sperm nucleus) retains normal levels of methylation. Programmed DNA demethylation might result in the generation of siRNAs that could reinforce transposon silencing in adjacent cell types (egg and sperm cells) that contribute genetic material to the next generation (22). It has been hypothesized that this process, although targeted to transposons, could inadvertently influence nearby genes, resulting in imprinted expression (17). In support of this idea, several well-characterized imprinted genes contain transposon sequences in adjacent regions (6, 23–25,). The potential for transposons to contribute to the origin of imprinted expression of nearby genes may result in examples of imprinting that do not provide a selective advantage and would not be expected to persist over evolutionary time. Because imprinting at such loci would be of limited functional relevance and dependent on the presence of a transposable element, this could lead to substantial allelic variation for imprinting within a species. Indeed, several of the first characterized examples of imprinting in maize exhibit allelic variation such that certain alleles are imprinted whereas others are not (26, 27).Regardless of the mechanisms that give rise to imprinted expression, parent-of-origin expression could, in some instances, provide a selective advantage. The kinship theory (16) suggests that MEGs should restrict growth or limit the flow of resources to offspring whereas PEGs might function to promote offspring growth. There are examples of imprinted genes that appear to exhibit these functions (28), but there is no clear evidence for these predicted functions in the annotations of the full set of previously identified MEGs or PEGs (3). Genes that are subject to parental conflict might be expected to exhibit signatures of positive selection (18, 29). For some imprinted genes, such as the Arabidopsis locus MEDEA, potential evidence of positive selection has been found in some cases (30, 31) but not others (32).The presence of imprinting for a particular gene is often assumed to have functional relevance. Although this may be the case for a subset of genes, the potential for inadvertent acquisition of imprinting as a result of nearby transposon influences could result in numerous examples of imprinting that have limited functional relevance and thus show intra- or interspecific variation in imprinting. To distinguish between these possibilities and evaluate the functional importance of imprinting, we analyzed imprinting in multiple diverse genotypes of maize. Reciprocal crosses among four genotypes allowed for the surveying of imprinting at over 95% of the genes expressed in endosperm tissue. We find that only a subset of imprinted genes shows conserved imprinting in maize and rice and that these genes show evidence of distinct selective pressures. Comparison of imprinting in different haplotypes within maize reveals allelic variation for imprinting, further suggesting that imprinting may have limited functional consequence for many maize genes.