AGO104 is a RdDM effector of paramutation at the maize b1 locus

Abstract

Although paramutation has been well-studied at a few hallmark loci involved in anthocyanin biosynthesis in maize, the cellular and molecular mechanisms underlying the phenomenon remain largely unknown. Previously described actors of paramutation encode components of the RNA-directed DNA-methylation (RdDM) pathway that participate in the biogenesis of 24-nucleotide small interfering RNAs (24-nt siRNAs) and long non-coding RNAs. In this study, we uncover an ARGONAUTE (AGO) protein as an effector of the RdDM pathway that is in charge of guiding 24-nt siRNAs to their DNA target to create de novo DNA methylation. We combined immunoprecipitation, small RNA sequencing and reverse genetics to, first, validate AGO104 as a member of the RdDM effector complex and, then, investigate its role in paramutation. We found that AGO104 binds 24-nt siRNAs involved in RdDM, including those required for paramutation at the b1 locus. We also show that the ago104-5 mutation causes a partial reversion of the paramutation phenotype at the b1 locus, revealed by intermediate pigmentation levels in stem tissues. Therefore, our results place AGO104 as a new member of the RdDM effector complex that plays a role in paramutation at the b1 locus in maize.

Fig 1. b1TR siRNAs and AGO104 interact in reproductive tissues.

A) Identified (colored) and putative as based on homology with A. thaliana proteins (grey) RdDM members involved in small interfering RNAs biogenesis (left) and de novo DNA methylation (right) in maize. RdDM proteins involved in paramutation are shown in red. B) Stem-loop PCR for RdDM-dependent R3 siRNAs in immature (im) and mature (ma) reproductive tissues of Mm and mm B’ plants. Arrows indicate the 67-bp expected band generated by R3. Ladder: Promega 100bp DNA Ladder Molecular Weight Marker. C) Stem-loop PCR of siRNAs extracted from IPs of AGO104 performed in tissues of B’ Mop1/Mop1 plants. Control + are small RNAs extracted directly from reproductive tissues. AbAGO104 are small RNAs extracted from the IPs of AGO104. -Ab correspond to the mock immunoprecipitation samples (without Ab). Arrows indicate the 67-bp expected bands. Ladder: Promega 100bp DNA Ladder Molecular Weight Marker.

Fig 2. RNA-IP sequencing (RIP-seq) of AGO104-loaded small RNAs in mature ears of B73 (b), Mm (B’) and mm (B-I like) individuals.

A) Size distribution of reads normalized to 1. B) Distribution of 24-nt reads within the 100-kb region that includes the b1TR (red box) and the b1 gene (blue box). x axis shows the B73/b1TRs composite reference map used for aligning reads. Vertical black bars indicate normalized read counts (CPM: Counts Per Million).

Fig 3. Crossing scheme used for a reverse genetic screen designed to investigate AGO104 contribution to paramutation.

Alleles in genotype descriptions are as follows: M: mop1; m: mop1-1; A: ago104; a: ago104-5; b: neutral b1 allele; B-I like: epiallele from a mm plant. Pigmentation phenotypes are indicated in squared brackets.

Fig 4. Pigmentation phenotypes observed at 46 and 56 days post seedling (dps).

A) Pictures of stem phenotypes at 46 dps in (left-to-right): mm control plants (B-I like, dark purple pigmentation); Mm control plants (B’, light pigmentation); Mm;Aa cross 2 progenies (intermediate pigmentation). The white arrow indicates a typical unpigmented node. B and C) Pigmentation phenotypes in control plants (with n > 25 for each control) and in cross 2 progenies, respectively. Numbers in bars are percentages. D) Absorbance at 550 nm of anthocyanins extracted using 1 g of stem tissue from 56 dps plants. std is the standard deviation. y axis scale is shown at .05 intervals between 0 < DO < 0.25 and 0.25 intervals between 1 < DO < 2.