Paternal imprinting of dosage-effect defectivel contributes to seed weight xenia in maize

The maize kernel's weight primarily depends on the endosperm, which develops from a diploid maternal central cell fertilized by a haploid sperm cell. Historically, the influence of pollen genotype on seed phenotype was described as 'xenia', implying a unique paternal contribution beyond simple Mendelian inheritance. Before the rediscovery of Mendelian genetics, xenia was considered a unique interaction between parental genomes. However, most observed xenia effects are explained by dominant Mendelian alleles. Genomic imprinting, where parent-of-origin effects determine gene expression, allows for phenotypic variation independent of simple dominance. The maize *R*' allele, a maternally expressed gene (MEG), is a classic example, causing reduced kernel anthocyanin when inherited paternally. Over 100 imprinted genes are known in maize, but only a few MEGs have demonstrable effects on seed phenotypes. Notably, until this study, no paternally expressed genes (PEGs) had been shown to directly regulate kernel development, although three paternal-effect seed mutants were previously identified. This research investigates the *ded1* locus, a quantitative PEG, to understand its role in maize kernel development and whether it represents a true xenia effect.

Previous research has established the significance of genomic imprinting in maize seed development, highlighting the roles of maternally expressed genes (MEGs) such as *R*, *meg1*, and *floury3* in determining kernel phenotypes. However, the functional contribution of paternally expressed genes (PEGs) in this process remained largely unexplored. While some paternal-effect seed mutants have been documented, none had been demonstrated to directly regulate kernel development. This research sought to address this gap by investigating the *ded1* locus, a quantitative PEG with potential implications for understanding xenia effects in maize.

The study started by identifying the *ded1-ref* allele through a quantitative screen of 1068 self-pollinated ears from the UniformMu population, focusing on defective kernel mutants. Kernel weights were measured and analyzed using a near-infrared grain analyzer. A cumulative distribution plot revealed a distinct weight difference between homozygous normal and *ded1* heterozygotes, indicating a dosage effect. Subsequent mapping and sequencing located the *ded1* locus to a 470 kbp region on chromosome 1, pinpointing a copia-like retrotransposon insertion in the *ZmMyb73* gene, which was further validated through CRISPR-Cas9-induced mutations. Allele-specific expression was confirmed using a cleaved amplified polymorphic sequence (CAPS) RT-PCR assay. The subcellular localization of DED1 was determined by transient expression of a DED1-GFP fusion in maize protoplasts. Transcriptional activation capabilities were tested using a yeast two-hybrid system. To identify direct targets of DED1, Differential Expression analysis (RNA-seq on 12 DAP endosperm tissue from homozygous *ded1-ref* mutants and normal siblings) and DAP-seq (DNA affinity purification sequencing) were performed, identifying differentially expressed genes (DEGs) and DED1 DNA binding sites, respectively. Electrophoretic mobility shift assays (EMSAs) validated DED1 binding to specific promoter regions of target genes. Lastly, to assess the impact of *ded1* dosage on gene expression, quantitative RT-PCR was conducted on a dosage series of 12 DAP endosperm RNA samples generated from controlled crosses. Histological analysis complemented the molecular data by examining the endosperm and embryo development of *ded1* mutants and normal siblings.

The research identified *ded1*, encoding an R2R3-MYB transcription factor, as a quantitative PEG significantly influencing seed weight. Hypomorphic *ded1* alleles showed a 5-10% seed weight reduction when paternally inherited, whereas homozygous mutants exhibited a 70-90% reduction. DED1 was found to be specifically expressed during early endosperm development with a clear paternal allele bias, directly activating early endosperm genes and repressing late grain-fill genes. Allele-specific expression analysis confirmed the paternal allele's dominant contribution to *Ded1* expression. The study identified approximately 43,225 DED1-enriched peaks via DAP-seq, with a substantial portion located near the transcriptional start sites of genes involved in endosperm development. Overlapping DEGs and DED1 binding sites revealed 258 genes directly activated and 180 genes directly repressed by DED1. Several direct targets, including *fl3*, *sus1*, *c1*, and *vp1*, were confirmed via EMSA. Analysis of temporal gene expression patterns across endosperm development revealed that *Ded1* and its activated targets exhibited peak expression during early endosperm development and in the endosperm adjacent to the scutellum (EAS) cell layer, critical for nutrient transfer to the embryo. In contrast, DED1-repressed genes displayed later expression patterns, primarily in the embryo, pericarp, and scutellar aleurone. The *ded1* mutants displayed severe developmental defects, including arrested embryo development and incompletely developed basal endosperm transfer layers (BETL). Dosage-dependent analysis showed that the paternal allele's expression level strongly correlates with seed weight. Specifically, a single paternally inherited *Ded1* allele is sufficient to complement a defective maternal allele, leading to near-normal seed development.

The findings strongly support the hypothesis that *ded1* acts as a major paternal regulator of seed size in maize, directly addressing the research question. The observed dosage effects and paternal imprinting of *Ded1* provide a clear example of xenia, where the paternal genome plays an obligate role in setting the pace of endosperm development. This work offers a crucial piece of evidence supporting the parental conflict theory for the evolution of imprinted gene expression, demonstrating that a PEG directly regulates seed development, unlike previously characterized MEGs. The detailed analysis of DED1's molecular function, through identification of its direct targets and their temporal expression patterns, provides insights into the developmental processes it regulates, including endosperm cell proliferation, cell type differentiation, and nutrient transport to the embryo. The significant reduction in seed size observed in *ded1* mutants highlights the crucial role of this gene in maize grain development.

This research identifies *Ded1*, a paternally expressed R2R3-MYB transcription factor, as a major regulator of maize seed size. The imprinting of *Ded1* and its dosage-sensitive effects represent a clear example of xenia. The detailed characterization of *Ded1*’s regulatory network provides insights into the molecular mechanisms governing early endosperm development and nutrient allocation. Future research could focus on exploring the evolutionary conservation of DED1 homologs in other species and investigating the regulatory interactions between DED1 and other key players in maize seed development.

The study's limitations include the potential underestimation of DEGs due to the mixed endosperm genotypes in the control samples and the fact that steady-state transcript levels reflect a combination of direct and indirect responses. The focus on 12 DAP endosperm tissue might not fully capture the dynamic changes in gene expression throughout the entire seed development process. Furthermore, the generalizability of the findings to other plant species might be limited due to the low conservation of imprinted gene expression among angiosperms.