Craig Stuart Pikaard

Understanding how the three-dimensional (3D) organization of the genome relates to chromatin and epigenetic landscapes requires methods capable of measuring chromosomal DNA interactions and chromatin modifications in an integrated manner. Current approaches typically use separate assays to assess chromatin conformation and epigenetic state, thus these features are not assessed on the same DNA molecules. Here, we describe Epi-PoreC, an Oxford Nanopore long-read sequencing-based method that combines the chromosome conformation capture method, PoreC with Fiber-seq, a method for detecting accessible chromatin regions based on their ability to be labeled using an exogenous adenosine methyltransferase. The positions of chromosome contacts, 6-methyladenosines, and endogenous 5-methylcytosines are then detected by nanopore sequencing, all on the same DNA molecules. Using Arabidopsis thaliana nuclei, we demonstrate that Epi-PoreC yields genome-wide chromatin contact maps comparable to those obtained by standard PoreC and DNA methylation and accessibility profiles comparable to standard Fiber-seq experiments. Epi-PoreC is thus an efficient means for conducting epigenetic and 3D genome profiling on individual DNA sequences in a single assay. Multidisciplinary Understanding genome regulation is aided by integrating information about chromosome architecture, DNA methylation, and chromatin accessibility. However, most experimental methods measure these features separately, precluding their detection on the same DNA molecules. Here we describe Epi-PoreC, an nanopore sequencing–based method that simultaneously measures chromatin contacts, DNA methylation, and chromatin accessibility in a single assay. The approach integrates chromosome conformation capture (PoreC) with Fiber-seq–based adenine methyltransferase labeling of accessible DNA. Using Arabidopsis thaliana nuclei, we demonstrate that Epi-PoreC produces chromatin contact maps highly concordant with those obtained by standard PoreC assays and DNA methylation and accessibility profiles comparable to those obtained by Fiber-seq. Because long nanopore sequencing reads can traverse multi-way chromatin ligation products, Epi-PoreC enables detection of chromatin interactions and modifications on the same DNA molecules, yielding genome folding and epigenetic landscape information relevant to studies of gene and genome function. Epi-PoreC therefore expands the toolkit for multi-omic singlemolecule genome analysis, particularly for complex or repetitive genomic loci where long-read sequencing is needed to identify sub-repeats or sequence elements that map to specific chromosomal locations. Method summary: Epi-PoreC integrates Fiber-seq–based adenine methyltransferase labeling of accessible DNA with the PoreC chromosome conformation capture protocol. Long-read Oxford Nanopore sequencing then allows the simultaneously detection of chromatin contacts, cytosine methylation and chromatin accessibility on individual DNA molecules. The method thus provides an efficient workflow for integrated single-molecule analyses of genome architecture and epigenetic state.

Arabidopsis thaliana has two ribosomal RNA (rRNA) gene loci, nucleolus organizer regions NOR2 and NOR4 , whose complete sequences are missing in current genome assemblies. Ultralong DNA sequences assembled using an unconventional approach yielded ~5.5- and 3.9-Mbp sequences for NOR2 and NOR4 in the reference strain, Col-0. The distinct rRNA gene subtype compositions of the NORs enabled the positional mapping of their active and inactive regions, using RNA sequencing to identify subtype-specific transcripts and DNA sequencing to identify subtypes associated with flow-sorted nucleoli. Comparisons of wild-type and silencing-defective plants revealed that most rRNA gene activity occurs in the central region of NOR4 , whereas most, but not all, genes of NOR2 are epigenetically silenced. Intervals of low CG and CHG methylation overlap regions where gene activity and gene subtype homogenization are high. Collectively, the data reveal the genetic and epigenetic landscapes underlying nucleolar dominance (differential NOR activity) and implicate transcription as a driver of rRNA gene concerted evolution.

Arabidopsis thaliana has two ribosomal RNA gene loci, nucleolus organizer regions NOR2 and NOR4 , whose complete sequences remain undefined. Ultra-long DNA sequences assembled using an unconventional approach yielded 5.5 and 3.9 Mbp sequences for NOR2 and NOR4 (in the reference strain, Col-0), revealing their distinct gene subtype compositions. RNA sequencing and identification of genes associated with flow-sorted nucleoli of wild-type or silencing-defective mutant plants shows that most of NOR4 is comprised of active genes whereas most, but not all, NOR2 genes are epigenetically silenced. Long intervals of low CG and CHG methylation overlap regions of gene activity and gene subtype homogenization. Collectively, the data reveal the genetic and epigenetic landscapes of the NORs and implicate transcription in rRNA gene concerted evolution. One Sentence Summary: NOR2 and NOR4 sequences fill genome gaps and enable megabase-scale analyses of rRNA gene regulation and concerted evolution.

SUMMARY: Nucleolus organizer regions (NORs) are eukaryotic chromosomal loci where ribosomal RNA (rRNA) genes are clustered, typically in hundreds to thousands of copies. Transcription of these rRNA genes by RNA polymerase I and processing of their transcripts results in the formation of the nucleolus, the sub‐nuclear domain in which ribosomes are assembled. Approximately 90 years ago, cytogenetic observations revealed that NORs inherited from the different parents of an interspecific hybrid sometimes differ in morphology at metaphase. Fifty years ago, those chromosomal differences were found to correlate with differences in rRNA gene transcription and the phenomenon became known as nucleolar dominance. Studies of the past 30 years have revealed that nucleolar dominance results from selective rRNA gene silencing, involving repressive chromatin modifications, and occurs in pure species as well as hybrids. Recent evidence also indicates that silencing depends on the NOR in which an rRNA gene is located, and not on the gene's sequence. In this perspective, we discuss how our thinking about nucleolar dominance has shifted over time from the kilobase scale of individual genes to the megabase scale of NORs and chromosomes and questions that remain unanswered in the search for a genetic and biochemical understanding of the off switch.

The expression of genomically-encoded information is not error-free. Transcript-error rates are dramatically higher than DNA-level mutation rates, and despite their transient nature, the steady-state load of such errors must impose some burden on cellular performance. However, a broad perspective on the degree to which transcript-error rates are constrained by natural selection and diverge among lineages remains to be developed. Here, we present a genome-wide analysis of transcript-error rates across the Tree of Life using a modified rolling-circle sequencing method, revealing that the range in error rates is remarkably narrow across diverse species. Transcript errors tend to be randomly distributed, with little evidence supporting local control of error rates associated with gene-expression levels. A majority of transcript errors result in missense errors if translated, and as with a fraction of nonsense transcript errors, these are underrepresented relative to random expectations, suggesting the existence of mechanisms for purging some such errors. To quantitatively understand how natural selection and random genetic drift might shape transcript-error rates across species, we present a model based on cell biology and population genetics, incorporating information on cell volume, proteome size, average degree of exposure of individual errors, and effective population size. However, while this model provides a framework for understanding the evolution of this highly conserved trait, as currently structured it explains only 20% of the variation in the data, suggesting a need for further theoretical work in this area.

Arabidopsis thaliana has two nucleolus organizer regions (NORs), consisting of long tandem arrays of 45S rRNA genes on chromosomes 2 ( NOR2 ) and 4 ( NOR4 ). The rRNA gene subtypes mapping to NOR2 are mostly silenced during development, whereas most of those mapping to NOR4 are active. Cytosine methylation of promoters plays a significant role in rRNA gene silencing, but recent findings demonstrate that gene body methylation also play a key role in gene silencing. Previous studies have demonstrated that CG methylation plays a role in rRNA gene silencing. However, it remains to be determined whether Chromomethylase 2 (CMT2)-mediated, RdDM-independent CHH methylation plays any role in rRNA gene silencing. To find out the relative importance of cytosine methylation in each of the sequence contexts (CG, CHG, and CHH, where H=A, T, or C), we studied multiple DNA methylation-deficient Arabidopsis mutants. Primarily, we show that the mutants that displayed higher losses of CG or CHH methylation at rDNA loci, display higher degrees of NOR2 rRNA gene silencing disruption. Using whole genome bisulfite sequencing data, we show that the loss of RdDM-independent CMT2, not the loss of RdDM-dependent DRM2, erases most of the CHH methylation at rDNA loci, while retaining CG methylation levels comparable to those of wild-type Col-0, demonstrating the sufficiency of CHH methylation loss for releasing rRNA gene silencing. Analysis of Arabidopsis and tomato NOR sequences revealed that in 45S rRNA gene sequences, cytosines mostly occur in the CHH context, followed by CG and the lowest being in CHG contexts. The predominance of CHH sites is even more pronounced when the rRNA gene promoter and spacer promoter(s) are considered (0-7% CG, 14-25% CHG, and 79-80% CHH), suggesting a mechanism to explain the stronger disruption of NOR2 gene silencing in mutants that display higher losses of CHH methylation. Our data also reveal a role for gene body methylation in rRNA gene silencing. We show that NOR2 genes are relatively hypermethylated compared to NOR4 , consistent with findings from recent studies. By contrast, human rRNA genes have more CG sites than CHH and CHG, correlating with the evolutionary loss of CMTs in mammals. CMTs are plant-specific, which primarily methylate CHG and CHH sites. Our data define a critical role for CMT2-mediated RdDM-independent CHH methylation, in combination with MET1-mediated CG methylation, in rRNA gene silencing, which is entirely mediated by CG methylation in mammals.

RNA-directed DNA methylation in plants is guided by 24-nt siRNAs generated in parallel with 23-nt RNAs of unknown function. We show that 23-nt RNAs function as passenger strands during 24-nt siRNA incorporation into AGO4. The 23-nt RNAs are then sliced into 11- and 12-nt fragments, with 12-nt fragments remaining associated with AGO4. Slicing recapitulated with recombinant AGO4 and synthetic RNAs reveals that siRNAs of 21–24 nt, with any 5′-terminal nucleotide, can guide slicing, with sliced RNAs then retained by AGO4. In vivo, RdDM target locus RNAs that copurify with AGO4 also display a sequence signature of slicing. Comparing plants expressing slicing-competent versus slicing-defective AGO4 shows that slicing elevates cytosine methylation levels at virtually all RdDM loci. We propose that siRNA passenger strand elimination and AGO4 tethering to sliced target RNAs are distinct modes by which AGO4 slicing enhances RNA-directed DNA methylation.

The study of RNAs has become one of the most influential research fields in contemporary biology and biomedicine. In the last few years, new sequencing technologies have produced an explosion of new and exciting discoveries in the field but have also given rise to many open questions. Defining these questions, together with old, long-standing gaps in our knowledge, is the spirit of this article. The breadth of topics within RNA biology research is vast, and every aspect of the biology of these molecules contains countless exciting open questions. Here, we asked 12 groups to discuss their most compelling question among some plant RNA biology topics. The following vignettes cover RNA alternative splicing; RNA dynamics; RNA translation; RNA structures; R-loops; epitranscriptomics; long non-coding RNAs; small RNA production and their functions in crops; small RNAs during gametogenesis and in cross-kingdom RNA interference; and RNA-directed DNA methylation. In each section, we will present the current state-of-the-art in plant RNA biology research before asking the questions that will surely motivate future discoveries in the field. We hope this article will spark a debate about the future perspective on RNA biology and provoke novel reflections in the reader.

RNA-directed DNA methylation in plants is guided by 24 nt siRNAs generated in parallel with 23 nt RNAs of unknown function. We show that 23 nt RNAs function as passenger strands during 24 nt siRNA incorporation into AGO4. The 23 nt RNAs are then sliced into 11 and 12 nt fragments, with 12 nt fragments remaining associated with AGO4. Slicing recapitulated with recombinant AGO4 and synthetic RNAs reveals that siRNAs of 21-24 nt, with any 5’ terminal nucleotide, can guide slicing, with sliced RNAs then retained by AGO4. In vivo, RdDM target locus RNAs that copurify with AGO4 also display a sequence signature of slicing. Comparing plants expressing slicing-competent versus slicing-defective AGO4 shows that slicing elevates cytosine methylation levels at virtually all RdDM loci. We propose that siRNA passenger strand elimination and AGO4 tethering to sliced target RNAs are distinct modes by which AGO4 slicing enhances RNA-directed DNA methylation.

In plants, selfish genetic elements, including retrotransposons and DNA viruses, are transcriptionally silenced by RNA-directed DNA methylation. Guiding the process are short interfering RNAs (siRNAs) cut by DICER-LIKE 3 (DCL3) from double-stranded precursors of ~30 bp that are synthesized by NUCLEAR RNA POLYMERASE IV (Pol IV) and RNA-DEPENDENT RNA POLYMERASE 2 (RDR2). We show that Pol IV’s choice of initiating nucleotide, RDR2’s initiation 1–2 nt internal to Pol IV transcript ends and RDR2’s terminal transferase activity collectively yield a code that influences which precursor end is diced and whether 24 or 23 nt siRNAs are produced. By diversifying the size, sequence, and strand specificity of siRNAs derived from a given precursor, alternative patterns of DCL3 dicing allow for maximal siRNA coverage at methylated target loci.