Barley grain filling follows a changing pattern of parental gene activity

A changing role for parental gene copies

Barley grains store starch and proteins in the endosperm, the tissue that supplies much of their nutritional value. A study led by Martin Kovacik and colleagues examines how genomic imprinting—gene activity that depends on whether a copy came from the mother or father—helps regulate this storage process.

Their central finding concerns LYS3, a gene encoding a transcription factor, a protein that regulates other genes. The researchers report that its maternal copy is active first, while its paternal copy becomes partially active later. That change accompanies a shift in how storage proteins, starch granules and gene-expression patterns depend on parental inheritance.

The work was posted on bioRxiv on November 4, 2026. It is a preprint and has not been certified by peer review.

Mapping imprinting in developing grains

The team sequenced RNA from barley endosperm at eight and 12 days after pollination. RNA measurements reveal which genes are being expressed. They studied reciprocal crosses, reversing which parent supplied pollen, between the cultivar Morex and either an Ethiopian landrace or wild barley.

The screen identified 255 candidate imprinted genes: 150 classified as maternally expressed and 105 as paternally expressed. Independent checks of parent-of-origin expression gave highly consistent results, the authors report. Some differences between crosses reflected whether the researchers had suitable genetic markers to distinguish the parental copies.

The team then examined these genes in existing datasets covering grain development over time and gene activity across different tissues. One group showed particularly strong expression in endosperm transfer cells. These nutrient-absorbing cells sit beside a maternal tissue that transports nutrients toward the developing endosperm.

This places a concentration of imprinted-gene activity at an important nutrient-transfer interface. However, the spatial analysis mapped expression of genes identified as imprinted; it did not establish their parental expression pattern at every location and developmental stage.

Why LYS3 presented a puzzle

LYS3 was already known to regulate genes involved in endosperm filling and maturation. Loss-of-function mutants have shrunken endosperm, reduced starch and hordein storage proteins, and higher lysine content.

Yet mature mutant grains follow a conventional recessive inheritance pattern. In offspring produced by self-pollinating plants carrying one normal and one mutant copy, the expected pattern is three grains without the recessive phenotype for every one with it. That seemed difficult to reconcile with a gene expressed initially only from its maternal copy.

To investigate, the researchers crossed plants carrying normal or mutant LYS3 in different parental combinations and followed developing grains.

At 12 days after pollination, overall endosperm gene-expression patterns separated according to whether the mother supplied normal or mutant LYS3. By day 20, the pattern instead distinguished grains carrying only mutant copies from the other combinations. The authors describe this as a transition from an early maternal-dependent pattern to one consistent with recessive inheritance.

Storage proteins and starch follow the transition

The biochemical and microscopic observations supported that timing shift. At day 16, hordein levels were lower when the mother supplied mutant LYS3, even if the father supplied a normal copy. By day 20, that cross resembled the normal comparison sample.

Starch granules also differed. Early in development, grains with maternal mutant copies contained smaller, less uniform granules. At day 20, more uniform, lens-shaped granules appeared in every cross combination except the one carrying only mutant copies.

Measurements of mature grain thickness supported two detectable phenotype groups rather than an additional intermediate group. The observed distribution was consistent with the expected three-to-one recessive pattern.

Partial paternal activity provides an explanation

RNA sequencing detected only maternal LYS3 expression at day eight. At day 12, paternal reads reached up to 18% of the total, below the 33.3% expected without parental expression bias. Additional measurements across days eight to 20 found an increasing paternal contribution, reaching 10–23% of measured amplification products.

The authors interpret this partial paternal activation as progressively compensating for an early lack of functional maternal LYS3. Importantly, the paternal copy did not need to reach an unbiased expression share for later traits to resemble the normal condition.

The broader screen assessed imprinting at two developmental time points, while the detailed functional investigation focused on LYS3. The findings therefore distinguish sustained gene activity from sustained imprinting. The authors also interpret the results through parental conflict theory, but that evolutionary interpretation is separate from the measured changes in gene expression and grain development.