A switch from TE-like heterochromatin to euchromatin underlies activation of protein storage genes in maize endosperm

This study reveals that a specific subset of maize protein storage genes, typically silenced by TE-like heterochromatin in most tissues, undergoes maternal demethylation in the endosperm to switch to a euchromatic state, enabling exceptionally high and often imprinted expression.

The Gist

The Big Picture: The Maize "Silent Library"

Imagine the DNA in a corn plant as a massive library containing millions of books (genes). In most parts of the plant (the leaves, the roots, the stem), the library is very strict. If a book has a "Do Not Read" stamp on it, it stays locked in a dark, dusty basement. This is usually how the plant handles Transposable Elements (TEs)—junk DNA that can jump around and cause chaos. The plant silences them by wrapping them in "heterochromatin," which is like a heavy, lead blanket that keeps them shut down.

Usually, if a gene gets this heavy blanket, it stays silent forever. But this paper discovered a fascinating exception in the corn kernel (the endosperm).

The Plot Twist: Unlocking the Basement

The researchers found a special group of genes that should be locked in the basement because they have that "Do Not Read" stamp (methylation). In the leaves, these genes are indeed silent. But when the corn kernel starts to grow, something magical happens: the plant actively rips the heavy blanket off these specific genes.

Once the blanket is removed, these genes don't just whisper; they scream. They become the loudest, most active voices in the entire kernel, producing massive amounts of protein.

The Cast of Characters

1. The "TE-like" Genes (The Rebels):
   These are the genes the paper focuses on. They look like junk DNA (TEs) because they are covered in methylation (the "Do Not Read" stamp) in the rest of the plant. However, in the kernel, they are the stars of the show.

   • What do they do? About one-third of them are Zeins. Think of Zeins as the "bricks" used to build the corn's protein storage warehouse. They are the main reason corn kernels are so nutritious. The others are "security guards" (antifungal proteins) or "delivery drivers" (secreted proteins) that help the kernel grow and protect itself.

2. The "Demolition Crew" (DNGs):
   How does the plant remove the heavy blanket? It uses special enzymes called DNGs (DNA glycosylases). You can think of them as a demolition crew with sledgehammers. In the kernel, this crew is hired specifically to knock down the "Do Not Read" stamps on these Zein genes so they can start working.

3. The "Imprinting" Mystery (Mom vs. Dad):
   Corn kernels are unique because they have three sets of chromosomes: two from the mother (the corn plant) and one from the father (the pollen).

   • The Surprise: For most of these "Rebel" genes, the plant turns on both the mom's and the dad's copies. They work together to build the kernel.
   • The Exception: For a few specific genes, the plant only turns on the mom's copy. Why? The researchers found that if the "Do Not Read" stamp is still stuck on the front door (the promoter) of the dad's gene, the demolition crew can't get in. But if the stamp is only on the inside of the book (the gene body), the crew can still open it. This explains why some genes are "maternally imprinted" (only mom's copy works) while others are not.

The Two Types of "Active" Genes

The paper distinguishes between two types of genes that get turned on in the kernel:

1. The Classic Imprinted Genes: These are like standard books that were locked in the basement. The demolition crew removes the lock on the cover (the promoter), and the book opens. They work moderately well.
2. The "TE-like" Genes (The Focus of this paper): These are like books that were entirely wrapped in lead blankets, even inside the pages. The demolition crew has to strip the blanket off the entire book (the whole gene body). The result? These genes don't just open; they go into overdrive, producing huge amounts of protein.

The "Secret Sauce" Analogy

Think of the corn kernel as a construction site building a skyscraper (the mature seed).

• The Leaves (Sporophyte): The construction site is quiet. The blueprints for the skyscraper are locked in a vault because they aren't needed yet.
• The Kernel (Endosperm): The construction site goes into "All Hands on Deck" mode.
• The Zeins: These are the steel beams and concrete. The plant realizes it needs millions of tons of them.
• The Mechanism: Instead of just unlocking the vault, the plant realizes, "Hey, these blueprints are marked 'Junk' by mistake!" So, it sends a team to scrub off the "Junk" label from the entire blueprint. Once the label is gone, the factory goes into overdrive, churning out steel beams at a rate 100 times faster than normal.

Why Does This Matter?

This discovery changes how we understand plant genetics. We used to think that if a gene looked like "junk DNA" (had TE-like methylation), it was broken or silenced forever. This paper shows that for corn, this "junk" status is actually a switch.

It's a biological strategy:

1. Keep it silent in the leaves so the plant doesn't waste energy or get confused by jumping genes.
2. Switch it on in the kernel with extreme power to create the massive food reserves needed to feed the baby plant when it sprouts.

In short, the corn plant has evolved a way to turn its "junk" DNA into a high-powered protein factory, ensuring that the next generation of corn has enough food to survive.

Technical Summary

1. Problem Statement

In plant genomes, DNA methylation in the CG and CHG contexts is typically associated with transcriptional repression, particularly when found in Transposable Elements (TEs). Genes exhibiting this "TE-like methylation" (teM) pattern in their coding bodies are usually poorly expressed, misannotated, or silenced in sporophytic tissues.

• The Paradox: While some teM genes are known to be highly expressed in pollen, it was unclear if a similar mechanism exists in the maize endosperm, a tissue critical for seed nutrition and development.
• The Gap: It is not fully understood how genes with constitutive heterochromatic marks (teM) in the sporophyte can be dynamically activated to extremely high levels in the endosperm, specifically regarding the major seed storage proteins (zeins). Furthermore, the distinction between classical imprinted genes (regulated by promoter demethylation) and these highly expressed teM genes remains unclear.

2. Methodology

The authors employed a multi-omics approach combining epigenomics, transcriptomics, and chromatin profiling in Zea mays (maize).

• Gene Identification:
  • Defined 67 "endosperm teM genes" based on strict criteria: high CG and CHG methylation (≥40%) in leaf (sporophyte) coding DNA, and expression at least 5-fold higher in endosperm than in nine other sporophytic tissues.
  • Filtered for "core genes" (syntenic across 26 NAM founder genomes) to exclude misannotated TEs.
• Methylation Analysis:
  • Utilized EM-seq data from W22 inbred lines (endosperm at 15 DAP, embryo, and mdr1 mutants) and B73 leaf/pollen.
  • Compared methylation levels between sporophyte and endosperm, and between wild-type and mdr1 (a DNA glycosylase/lyase mutant) to assess active demethylation.
  • Introduced N-normalized methylation (mC/total nucleotides) to account for cytosine density in promoter regions, distinguishing it from standard C-normalized methylation.
• Expression Profiling:
  • Analyzed RNA-seq time courses (6–38 Days After Pollination, DAP) to cluster genes by expression patterns.
  • Assessed imprinting status using allele-specific RNA-seq from reciprocal crosses (W22 × B73).
• Chromatin Profiling:
  • Performed CUT&Tag for histone modifications (H3K27me3, H3K56ac) and ChIP-seq/MNase-seq for other marks (H3K4me3, H3K9me2, H3K27me2, H2A.Z) in endosperm vs. sporophyte tissues.
• Comparative Analysis:
  • Contrasted endosperm teM genes with Flank-Methylated Endosperm Genes (FMEGs) (classical imprinted genes) and pollen teM genes.

3. Key Contributions & Results

A. Identification of a Unique Gene Class

The study identified 67 endosperm teM genes that are:

• TE-like in Sporophytes: Heavily methylated (CG/CHG) and heterochromatic in leaves.
• Demethylated in Endosperm: Show significant reduction in CG and CHG methylation in the endosperm, particularly in gene bodies and flanks.
• Exceptionally Highly Expressed: They constitute >40% of the transcriptome at 16 DAP. The median expression is 258-fold higher in endosperm than in sporophytes (compared to 12-fold for general endosperm-specific genes).
• Structurally Distinct: They encode short proteins (median CDS 465 bp), often lack introns, and are enriched for secreted proteins.

B. Functional Enrichment: The Zein Connection

A striking finding is that 23 of the 67 genes are zeins (major seed storage prolamin genes), including 19-kDa and 22-kDa $\alpha$-zeins and the 50-kDa $\gamma$-zein.

• Cluster 1 (Early): 39 genes peaking early (0–8 DAP), enriched for secreted proteins in the embryo surrounding region (ESR) and basal endosperm transfer layer (BETL).
• Cluster 2 (Late): 28 genes peaking later (8–30 DAP), dominated by zeins and the transcription factor pbf1, expressed in the starchy endosperm.

C. Chromatin State Switching

The study demonstrates a dramatic chromatin transition:

• Sporophyte: teM genes possess heterochromatic marks (H3K9me2, H3K27me2) and lack active marks (H3K4me3, H3K56ac), resembling TEs.
• Endosperm: Upon demethylation, these genes acquire euchromatic marks (H3K4me3, H3K56ac, H3K27me3) and chromatin accessibility.
• Implication: The removal of TE-like methylation allows the recruitment of activating histone modifiers, enabling massive transcriptional output.

D. Mechanism of Imprinting and Demethylation

• Role of DNGs: Demethylation is driven by DNA glycosylases/lyases (DNGs), specifically MDR1 and DNG102. In mdr1 mutants, methylation increases and expression decreases.
• Promoter vs. Gene Body Methylation:
  • Biparental Expression: Some teM genes are expressed from both parental alleles in endosperm despite paternal methylation in gene bodies.
  • Maternal Imprinting: Genes with strong maternal preference (imprinting) correlate with high methylation in the promoter region (TSS), not the gene body.
  • N-Normalization: Using N-normalized methylation (accounting for cytosine density), the authors showed that genes with high promoter methylation (>5% N-normalized) are maternally imprinted, while those with low promoter methylation (<0.5%) are biparentally expressed.
  • This mechanism explains why some zeins are imprinted while others are not.

E. Conservation with Pollen

The study draws parallels to pollen teM genes, which are also highly expressed, short, intron-poor, and dependent on DNGs for activation. In both tissues, the removal of TE-like methylation (specifically near TSS) is the key switch for high expression.

4. Significance

• Redefining Gene Regulation: The paper challenges the dogma that TE-like methylation in gene bodies is solely a sign of pseudogenes or misannotation. It reveals a functional class of genes that utilize this "silencing" mark in the sporophyte to ensure they are only expressed in specific reproductive tissues (pollen/endosperm).
• Mechanism of Storage Protein Synthesis: It provides the epigenetic basis for the massive accumulation of zeins in maize endosperm, linking DNA demethylation directly to the activation of the plant's primary protein storage machinery.
• Imprinting Nuance: It refines the understanding of genomic imprinting, showing that while classical imprinting relies on promoter demethylation, the degree of promoter methylation determines whether a teM gene is imprinted or biparentally expressed.
• Evolutionary Insight: The conservation of this mechanism in both pollen and endosperm suggests a convergent evolutionary strategy for regulating highly specific, high-output secretory pathways in plant reproductive tissues.

Conclusion

The authors demonstrate that a subset of maize genes, including the critical zein storage proteins, exist in a repressed, heterochromatic state in the sporophyte due to TE-like methylation. In the endosperm, active demethylation by DNGs triggers a switch to an active euchromatic state, allowing for exceptionally high, tissue-specific expression. The presence or absence of methylation specifically at the promoter (rather than the gene body) dictates whether these genes are imprinted or expressed from both parental genomes.