Replication timing uncovers a two-compartment nuclear architecture of interphase euchromatin

H S Akram; E E Wear; L Mickelson-Young; Z M Turpin; L Hanley-Bowdoin; W F Thompson; L Concia; H W Bass

doi:10.1093/plcell/koag042

. 2026 Feb 20;38(3):koag042. doi: 10.1093/plcell/koag042

Replication timing uncovers a two-compartment nuclear architecture of interphase euchromatin

H S Akram ¹, E E Wear ², L Mickelson-Young ³, Z M Turpin ⁴, L Hanley-Bowdoin ⁵, W F Thompson ⁶, L Concia ⁷, H W Bass ^8,^2,^✉,³

PMCID: PMC13017713 PMID: 41719277

Abstract

Genome replication is temporally regulated during S phase, with specific genomic regions replicating at defined times in a process that is known as replication timing (RT). Based on 3D cytology in replicating nuclei, we previously proposed a model in which maize euchromatin is subdivided into subcompartments distinguished by chromatin condensation and RT. However, whether this compartmentalization reflects a general nuclear architecture that persists throughout the cell cycle was unclear. To test this model, we conducted two orthogonal assays—Hi-C for genome-wide interaction data and 3D FISH for direct visualization of chromatin organization in maize (Zea mays L.). Hi-C analyses revealed distinct patterns of early-S regions exhibited negative insulation scores with long-range contacts, whereas middle-S regions showed the opposite. Early-S regions showed the strongest correlation with epigenomic signatures of open, transcriptionally active chromatin. 3D oligo FISH painting confirmed that early-S and middle-S replicating regions occupy adjacent but largely non-overlapping nucleoplasmic sub-territories throughout interphase stages, including G1. Together, our findings redefine the maize euchromatin “A” compartment as two spatially distinct subcompartments derived from high-frequency RT transitions between early and middle S along the linear genome. These findings have implications for chromatin-templated processes and underscore the importance of RT as a defining feature of genome organization.

Replication timing delineates the spatial organization within euchromatin regions of maize nuclei, revealing key features of intermediate-scale genome architecture.

Introduction

The eukaryotic cell cycle consists of a widely conserved series of events, including cell growth, DNA replication, and division into two daughter cells. DNA replication occurs during the S phase, and the specific time within S phase when each genomic region replicates can be measured and annotated as replication timing (RT) (Rhind and Gilbert 2013; Wear et al. 2017; Zhao et al. 2020). The fact that some cells or tissues have different RT profiles reveals the existence of an underlying RT program coupled to development or differentiation (Hiratani et al. 2008; Rivera-Mulia et al. 2015). The RT program is generally recognized as ensuring the faithful reproduction and transmission not only of the nucleotide sequence but also of the chromatin state, contributing to the propagation of epigenomic cell type identity and functions (Rivera-Mulia and Gilbert 2016).

Most eukaryotic RT knowledge comes from yeast and mammals, where complex temporal programs have been described and extensively studied (Prioleau and MacAlpine 2016; Vouzas and Gilbert 2023). However, the regulatory mechanisms that govern temporal RT programs remain understudied in plants (Sanchez et al. 2012). Notably, plants lack homologs of key RT regulators such as geminin and Rif1, which are essential for coordinating the cell cycle and RT in yeast and mammals (reviewed by Patmanidi et al. 2017; Richards et al. 2022). Similarly, plants and mammals exhibit some key differences in their chromatin organization as characterized by proximity ligation assays. A prominent feature of mammalian chromatin architecture is the presence of topologically associating domains (TADs), which partition the genome into domains spanning hundreds of kilobases (Rao et al. 2014). These TAD boundaries are typically enriched for architectural proteins such as CTCF and cohesin. In contrast, plants are reported to lack any clear CTCF homolog and cohesin-enriched TAD boundaries (Liu et al. 2016; Golicz et al. 2020; Lee and Seo 2023). Together, these observations strongly suggest that there are fundamental differences in the higher-order organization of plant chromatin. These differences, along with the vast evolutionary distance between plants and opisthokonts, highlight the importance of studying DNA RT in plant models.

Early studies of plants concluded that genome replication is essentially biphasic, with early-replicating and late-replicating regions (Sparvoliet al. 1994; Samaniego et al. 2002). Later studies greatly improved the resolution of the RT profile by using nucleotide analogs such as 5-bromo-2′-deoxyuridine (BrdU) (Lee et al. 2010) and 5-ethynyl-2′-deoxyuridine (EdU) for pulse-labeling replicating DNA. EdU has become the analog of choice because it does not require acid or heat denaturation for detection and, instead, is directly conjugated with click chemistry to add a fluorescent dye used for both fluorescence-activated nuclei sorting (FANS) and microscopy (Salic and Mitchison 2008). These EdU-labeled nuclei can be separated by flow cytometry from unlabeled G1 and G2 nuclei and further divided by increasing DNA content into subpopulations representing sequential stages of S phase. Three separated nuclei populations, early-S, middle-S, and late-S, have been used for microscopic analysis and sequencing of replicated DNA (Repli-seq) (Bass et al. 2014). The first genome-wide RT maps in plants were produced for Arabidopsis thaliana and maize (Zea mays L.) using these techniques (Wear et al. 2017; Concia et al. 2018). The comparative analysis of early-S, middle-S, and late-S replication in Arabidopsis revealed that early-S and middle-S replicating regions are quite similar to each other and distinct from late-S regions. Early and middle replicating domains are enriched with open chromatin as defined by nuclease sensitivity and histone acetylation markers, with early-replicating regions being gene-rich, while late-replicating domains are characterized by an enrichment of transposons and repressive epigenetic markers (Lee et al. 2010; Wang et al. 2015; Liu et al. 2016; Concia et al. 2018). Although Arabidopsis has provided a foundational understanding of plant RT, its small genome size and high gene density are unusual among plant genomes. Plants with larger genome sizes, such as maize (2n = 2X = 20) with a 1C value of 2.3 Gb, have expanded intergenic content, which generally tends to replicate primarily during middle-S and exists in a slightly more compacted state compared to the early-S replicating chromatin (Bass et al. 2015; Liu et al. 2016; Arifulin et al. 2024).

The maize genome, which is about two-thirds the size of the human genome and contains a complex array of transposons, is more typical of higher plants. This fact, together with the fact that maize has been the subject of detailed genetic and cytogenetic studies for more than 100 years (Freeling and Walbot 1996; Nannas and Dawe 2015), makes it an attractive system for exploring the control of replication in plants with complex genomes and chromatin architecture. Furthermore, as one of the world's most agriculturally significant crops, insights into the replication process in maize are relevant not only for fundamental research but also for strategies to enhance crop performance and resilience.

Our group has developed a powerful experimental system to characterize replication in actively growing, intact maize root tip meristems. This system has provided key insights and enables examination of replication patterns, genome features, and associated chromatin structure parameters at different stages of the cell cycle (Bass et al. 2014; Savadel and Bass 2017). The maize root tip system examines nuclei from cells growing in their native, organismal, and developmental context, unlike nearly all of the mammalian models that depend primarily on cell culture systems. Although heterochromatic regions in maize replicate late, and many genes replicate early as they do in most other eukaryotic systems, the spatiotemporal patterns of early and middle S-phase replication in maize differ significantly from those in mammalian and yeast systems. In mammals, DNA synthesis occurs at different nuclear regions throughout S phase, classified as 5 sequential stages referred to as patterns 1 to 5 (O’Keefe et al. 1992) or Types I-V (Zink 2006), in which the first, third, and fifth stages represent the earliest, the middle-most, and the latest stages of S phase, respectively. Despite the different naming conventions, there is agreement on the nuclear distribution of replication in mammalian nuclei, which starts with broad distribution in euchromatic portions of the nucleoplasm at early S phase, then in perinuclear and perinucleolar regions during middle S phase, followed by heterochromatic portions of the nucleoplasm at late S phase (O’Keefe et al. 1992; Dimitrova and Berezney 2002; Panning and Gilbert 2005; Zink 2006; Rivera-Mulia and Gilbert 2016). Many mammalian RT studies compare early to late RT, and genomic regions where they switch, with relatively few studies focusing directly on middle-S (Hiratani et al. 2008; Desprat et al. 2009; Ryba et al. 2010). In maize, DNA synthesis is thoroughly dispersed throughout the nucleus during both early-S and middle-S phase and becomes more punctate and clustered only in late-S phase. Analysis of DNA RT in maize shows that early-S phase replicative labeling primarily colocalizes with regions of relatively weak 4′,6-diamidino-2-phenylindole (DAPI) fluorescence, indicating a low DNA density, while middle-S phase EdU labeling aligns with regions of euchromatin showing stronger DAPI staining. Based on these findings, we proposed a “mini-domain chromatin fiber RT model”, which states that maize euchromatin exists as a closely interspersed mixture of two subcompartments, each characterized by distinct RT and chromatin morphology (Bass et al. 2015).

To test the mini-domain model, we employed a combination of proximity ligation (Hi-C) and quantitative in situ hybridization (3D-FISH) techniques in this study. Here, we provide new evidence supporting this model of maize euchromatin, showing that the described relationships hold throughout the cell cycle. Our results support a view of the plant euchromatin global “A” compartment as largely composed of two spatially and epigenetically distinct, but closely interspersed, compartments.

Results

Repli-seq scheme and mini-domain chromatin fiber RT model

The overall experimental scheme and replicative labeling approach that previously led to the two-compartment mini-domain RT model (Bass et al. 2015) is summarized in Fig. 1. Maize (Zea mays L.) roots are labeled in vivo with EdU, which is subsequently clicked with Alexa Fluor 488 (AF-488), allowing for bivariate FANS sorting into three populations of S-phase nuclei, early-S, middle-S, and late-S (Fig. 1a). Representative examples illustrate the canonical labeling patterns for each stage (Fig. 1b), with early-S and middle-S synthesis broadly dispersed throughout the nucleoplasm, whereas late-S synthesis is limited to discrete patchy heterochromatic regions. In maize, the cytological similarity observed between the early-S and middle-S phases in their DNA labeling patterns is one of several factors that support the classification of middle-S-replicating chromatin as euchromatin. Although replicative labeling in both early S and middle S nuclei is widely distributed, it overlaps with chromatin that differs in condensation (Bass et al. 2015).

For image description, please refer to the figure legend and surrounding text. — Repli-seq scheme and mini-domain chromatin fiber RT model. a) DAPI and EdU-stained nuclei were analysed via flow cytometry using 355 nm (UV) and 488 nm (blue) lasers. The bivariate plot displays DNA content (DAPI fluorescence with emission filter 460 ± 50 nm) and EdU incorporation (AF-488 fluorescence with emission filter 530 ± 40 nm). Rectangles indicate the G1, early-S, middle-S, late-S, and G2 nuclei that were sorted b) Cytology of EdU-labeled early-S, middle-S, and late-S sorted nuclei. Images were collected using 3D deconvolution microscopy, corrected for wavelength-dependent chromatic aberration, and displayed as grayscale. Each nucleus is shown with intensity-averaged projections spanning 2 µm (10 Z-sections) from the center. Scale bars are 2 micrometers. c) DAPI-stained image of a maize interphase nucleus. The nucleolus is labeled “n”, and one of the knobs is highlighted as “K”. The yellow box indicates the euchromatin region magnified in the next panel. d) A zoomed-in area of the euchromatin region with fiber-like structures. e) Illustration depicting the proposed two euchromatin sub-compartments: early-S (thin/gray) and middle-S (thick/black). f) A schematic representation of euchromatin and heterochromatin distributed along chromosome 5, including a centromere (CEN) (Gent et al. 2015; Hufford et al. 2021) and knob (K) on the long arm. The Repli-seq profiles and Repliscan segments spanning the entire chromosome are shown with early RT (E, blue), middle RT (M, green), and late RT (L, red). g) UCSC Genomaize browser view zoomed into two 2-Mb non-heterochromatin regions with Repli-seq profiles and RT segments defined by Repliscan.

Based on this cytology, the previously proposed “mini-domain chromatin fiber RT model” (Bass et al. 2015) is illustrated (Fig. 1c and d). According to this model, maize euchromatin, which is not uniformly stained with DAPI (Fig. 1c and d), exists as two closely interspersed subcompartments that are evident in S phase, corresponding to early-S and middle-S replicating regions. The early-S replicating chromatin regions are relatively weakly stained with DAPI, whereas the middle-S replicating chromatin regions exhibit stronger DAPI staining, as depicted in the interpretive diagram (Fig. 1e).

At the molecular level, RT is visualized as Repli-seq coverage profiles from early-S, middle-S, or late-S replicating DNA libraries as previously described (Wear et al. 2017). For this study, Repli-seq data were mapped to the current maize B73 genome assembly (B73v5) and used to annotate and segment the genome into discrete DNA RT classes using Repliscan (Zynda et al. 2017).

The maize genome comprises 10 metacentric chromosomes, each ranging from approximately 150 to 300 Mb in size. In this study, we use chromosome 5 as a representative example based on the fact that it represents a typical maize metacentric chromosome, and that its RT profiles reflect those of the whole genome as previously defined by Repli-seq (Wear et al. 2017). The RT profiles along the entire length of maize chromosome 5 (Fig. 1f) illustrate regional enrichments for early, middle, and late replication. The early-S RT profiles show the highest coverage in the more distal regions of the chromosome arms, where gene density tends to be higher in maize chromosomes (Schnable et al. 2009). The middle-S RT profiles show a unique pattern, being relatively evenly distributed across most of the chromosome. In contrast, late-S RT profiles show the highest coverage near the center of the chromosome at the centromeric and pericentromeric regions. A distinct non-centromeric block of late-S replicating chromatin is also seen at the heterochromatic knob, embedded in the long arm of chromosome 5. To gain further insight into the relationships among the different RT classes, we inspected 2-Mb non-heterochromatin regions on chromosome 5 to specifically focus on the distribution of early-S versus middle-S (Fig. 1g, blue and green) RT segment classes. Zooming in on 2-Mb regions outside of the centromeric heterochromatin illustrates a striking pattern of RT, in which closely interspersed alternating early-S and middle-S segments are evident, switching back and forth every ∼100 kb. This early and middle segment interspersion pattern is typical of most of the chromosome arms in maize, as previously shown by Wear et al. (2017). This mini-domain pattern is notably distinct from the large Mb-size RT domains commonly described for mammals (Pope et al. 2010; Baker et al. 2012). A genomic composition analysis of early-S, middle-S, and late-S RT segment regions revealed that all three classes were predominantly composed of transposable elements (TEs), which comprise ∼85% of the maize B73 genome (Schnable et al. 2009). Given their abundance and distribution across all three RT classes (Fig. S1a), it is clear that TEs or repeats alone are not a defining feature of RT-associated genome organization.

We aimed to test whether the mini-domain model, originally defined based on two distinct chromatin types during S phase in root tip cells, represents a broader principle of maize nuclear architecture, applying not only to nuclei in S phase but also throughout interphase and across different maize tissues. This idea was first tested using high-throughput conformation capture (Hi-C) analysis with nuclei from the entire mitotic cell cycle (Fig. S1b–e).

Hi-C data reveal covariance and correlation with early and middle RT in maize euchromatin

We used formaldehyde-fixed root tip nuclei from 3-day-old seedlings to prepare Hi-C libraries in biological replicates of 1.25 million nuclei each (Fig. S1b–e). For comparison, we also made Hi-C libraries from immature earshoot nuclei. The analysis of the resulting Hi-C libraries provided genome coverage and valid pair contact frequencies comparable to that previously reported for maize bundle sheath cells (Dong et al. 2017). This confirmed that the contact patterns remained intact following the nuclei isolation, flow sorting, and library preparation steps (Table S1).

To compare nuclear architecture to DNA RT, we performed nuclear compartment analyses using Hi-C, as shown in Fig. 2. To examine early-S versus middle-S replicating regions, we looked at three different non-heterochromatin 2-Mb regions from the short arm of chromosome 5 (Fig. 2a). We plotted the RT profiles and segments along with the chromosome-wide eigenvector (EV) and insulation score (IS) analyses with a 50-kb bin size for all four biological replicates (Fig. 2b–d). EVs from principal component analysis (PCA) of Hi-C data are used to identify genomic regions that interact more frequently with each other, thus revealing features of large-scale chromatin organization, such as compartments (see Methods). The IS of a genomic region reports its frequency of interaction with the neighboring regions. A low or negative IS indicates a scarcity of local interactions relative to longer-range contacts (Crane et al. 2015). Strikingly, most of the points at which the EV and IS values switched from positive to negative mirrored the locations where RT segment switches occurred (Fig. 2b–d). For instance, the light-blue shaded regions, representing early-S replicating segments, showed positive EV and negative IS values. The low IS of early-S RT regions denotes a greater tendency for long-range contacts. Relatedly, the green-shaded regions, indicating middle-S replicating segments, correspond to negative EV and positive IS values, suggesting more confined interactions (Crane et al. 2015). Finally, the red-shaded areas show late-S replicating segments with corresponding negative EV and positive IS values. These Hi-C patterns uncover smaller-scale nuclear architectural features that describe more accurately the global A (early) and B (late) compartments by differentiating at the local scale early-S from middle-S or late-S regions. We next pooled the root tip biological replicates based on their similarity (Fig. S2a) and used the pooled data to generate the contact matrix plots that are shown in relation to the Repliscan segment classification (Fig. 2e–g, Fig. S2b). The contact matrices showed a strong diagonal signal for the expected abundant interactions among adjacent loci. Beyond the diagonal, interaction hotspots were observed (examples highlighted in blue and pink circles). Blue circles highlight contacts between two non-adjacent early RT regions, while pink circles mark interactions between early and middle RT regions. These distal contact hotspots represent long-range interactions within these representative 2-Mb regions. Notably, early-S replicating segments appeared to display more long-reaching interactions with other early-S replicating segments compared to adjacent middle-S replicating segments. Some of these early-to-early contacts extend beyond 1 Mb, skipping multiple intervening RT segments consistent with the low IS of early-S RT regions (Fig. 2b–d). In these regions, the middle-to-middle and late-to-late contacts show only minimal off-diagonal signals (eg, see the green and red boxes in Fig. 2f). Importantly, most of the long-distance contacts in these 2 Mb windows are early-to-early (Fig. 2b–d).

To determine if these relationships between nuclear architecture and RT hold true for the rest of the genome, we carried out a global-scale correlation analysis between RT and Hi-C. For this, we combined the chromosome-wide Hi-C data (EV and IS) from all ten chromosomes. The early-S Repli-seq coverage values showed a strong positive correlation (r = + 0.69) with the EVs, whereas the middle-S Repli-seq coverage showed a weak positive correlation (r = + 0.07). The late-S Repli-seq coverage values showed a strong negative correlation (r = −0.61) with the EVs (Fig. 2h). When doing a similar genome-wide correlation with IS, we observed a strong negative correlation with early-S Repli-seq coverage (r = −0.67), a weak negative correlation with middle-S Repli-seq coverage (r = −0.04), and a strong positive correlation with late-S Repli-seq coverage (r = + 0.66) (Fig. 2i). For both EV and IS, the strongest but opposite correlations were observed for early-S compared to late-S chromatin. The middle-S chromatin showed the same direction of correlation as that of the early-S chromatin, but with much smaller correlation values. Together, these data show that genomic regions distinguished by their timing of replication also correlate with distinct Hi-C-defined chromatin and genomic organization across the whole genome.

To examine the impact of resolution on the observed correlation patterns, we performed similar correlation analyses at multiple resolutions. Chromosome-wide eigenvector (CEG) analyses at 25 and 100 kb, as well as block-based eigenvector (BEG) analysis at 50 kb, confirmed similar (although slightly weaker for BEG) trends across all resolutions at both the chromosome and genome-wide levels (Fig. S3). In addition to EV analysis, we conducted IS analysis at a range of resolutions (1, 10, and 25 kb), using two sliding window sizes (2.5 and 1 Mb; Fig. S4). All of these additional IS analyses have a similar correlation with our RT data as observed at the 50 kb CEG analysis.

We next compared our Hi-C data for root tips to the EV and IS data from earshoots analysed in parallel (Table S1) and to maize bundle sheath cells from a published study (Dong et al. 2017). Pairwise comparisons between tissues (Fig. S5) revealed strong genome-wide similarity between root tip and earshoot Hi-C profiles (r = 0.85 for both EV and IS), both of which are actively growing tissues. We also observed slightly weaker but significant correlations between root tip and bundle sheath Hi-C profiles (r = 0.73 for EV and 0.66 for IS), possibly reflecting some level of cell-type independent global conservation of genome organization.

We also investigated whether any of the ten maize chromosomes showed more or less tissue-specific variation compared to the whole genome (Fig. S6). Interestingly, we found that the root tip and earshoot Hi-C data showed a strong positive correlation across all chromosomes, with the notable exception of chromosome 6, which contains the nucleolar organizing region near the end of the short arm. The correlation coefficient for chromosome 6 was r = 0.62, compared to all the others, which ranged from r = + 0.78 to +0.95 (Fig. S6a). We found a similar situation for root tip versus bundle sheath, with chromosome 6 showing the lowest correlation (Fig. S6b). In addition, most of the other chromosomes showed relatively weaker correlations with bundle sheath data, likely due to variations in Hi-C library preparation or cell type or both. Using a stratum-adjusted correlation coefficient (SCC) analysis (Yang et al. 2017), we also found that root tip to earshoot similarity was higher than that of either one compared to bundle sheath (Fig. S7).

Given the strong correlations between RT and Hi-C defined nuclear architecture in root tip, we wondered if the RT classification alone from root tip could be predictive of Hi-C nuclear architecture in other tissues. We only have Repli-seq data for the root tip tissue, whereas Hi-C data is available for the root tip and the other two tissues. Therefore, we performed a correlation analysis between the root tip Repli-seq data and the Hi-C EV data from the other two tissues, for which RT data is not available. These results are summarized in Fig. 3. The root tip Repli-seq-defined genome structure closely mirrored the Hi-C EVs of the other tissues, which can be seen locally (Fig. 3b–e) and globally (Fig. 3f–h). These analyses demonstrated that RT in root tip predicts chromatin structure across other tissues, including earshoot and bundle sheath, suggesting a robust association between RT and chromatin organization across the plant.

Hi-C contacts are enriched within RT classes

We used the pooled Hi-C valid contacts to examine the degree to which moderate to long range contacts were occurring between regions with the same RT assignments, ie within-class contacts, as predicted by the two-compartment mini-domain model for euchromatin (Fig. 4). We assigned RT class annotations to each of the two members of the valid pair contacts and calculated the frequencies of contacts with matching and differing pairs for early-S, middle-S, or late-S regions. The analysis revealed that of all the valid pair contacts, early-to-early accounted for 28% (2.93 million), middle-to-middle for 24% (2.47 million), and late-to-late for 16% (1.65 million) (Fig. 4a). Contacts between two different RT classes were less frequent, with 16% early-to-middle, 13% middle-to-late, and only 3% early-to-late. Contacts which had either hybrid RT segment categories (such as EM, ML, EL, or EML) or no RT data were excluded. These findings indicate the “same-class” categories are the most frequent, even after normalizing to account for their relative abundance (Fig. 4b).

To examine more closely the same-RT-class contacts, we conducted a virtual 4C analysis of two early and two middle RT genomic regions, each approximately 150 kb, located on chromosome 5 (Fig. 4c). This approach allowed for the identification of long-range interactions centered around a specific genomic region of interest, known as the “bait” region, or virtual bait in this case. Two early and two middle RT segments were designated as baits, and their same-class contacts were plotted at 10 kb resolution along the entire chromosome 5 (Fig. 4d). All four baits showed long distance contacts, including to regions on the other arm of the chromosome, but the early-to-early contacts extended further, even beyond the distal knob on the long arm of chromosome 5. This difference can not be explained by the location of the baits, because they come from the same local region (Fig. 4c), which represents less than 1% of the 226 Mb chromosome 5. The per-bait analysis also showed that the same class contacts were the predominant type. The least frequent contacts were between early or middle baits and late-S regions for all four segments analysed (Fig. 4e–h). We observed consistent results when we decomposed the contact matrices for all of chromosome 5 into early-to-early (Fig. 4i) and middle-to-middle RT contacts only (Fig. 4j). As expected, the early-to-middle contacts showed minimal off-diagonal contacts (Fig. 4k).

We next investigated the persistence of these same class contact tendencies as a function of distance relative to the linear genome. The results demonstrated the expected distance-dependent decay of all contact classes, with early-to-early RT class contacts showing the highest frequency at any given distance (Fig. S8a). This enrichment was not a result of the relative abundance of the RT classes per se, as shown using a control with the RT positions randomly shuffled (Fig. S8b).

Overall, the contact class pair enrichments and the virtual 4C bait results showed that same-class contacts were more frequent than cross-class contacts, and that early RT segments exhibit more of the longest-range interactions than the middle RT segments.

Hi-C-based chromatin architecture and Repli-seq align with epigenomic and transcriptional features

To test how our Hi-C data correlated to genomic and genetic features, we checked the correlation of our 50-kb Hi-C EVs with epigenetic and transcriptional features known to be associated with chromatin structure (Fig. 5). We first examined how the Hi-C data correlated with histone post-translational modifications using ChIP-seq data from B73 maize root tips (Wear et al. 2017, 2020; Cahn et al. 2024).

We previously anticipated that modifications associated with active genes would be positively correlated with early-S RT regions within euchromatin, whereas repressive marks would not be or less so (Savadel and Bass 2017). Indeed, we found significant correlations between our root tip Hi-C EV data analysed at 50-kb resolution on chromosome 5 and the active gene-associated marks (Fig. 5a–d). Among the marks with positive correlations, H3K27ac and H3K4me3 had the strongest correlations with r = + 0.73 and +0.65, respectively. The positively sloped, linear clouds reflect concordance between EV maxima and ChIP-seq coverage enrichment for these active chromatin features. The values for the histone marks and the EV varied along the length of the chromosome, showing a general trend of higher values in the chromosome arms for all of the marks tested, except for H3K9me2 (Fig. S9a–d). Consistent with our findings, regions enriched for H3K56ac (EV r = + 0.45) and H3K4me3 (EV r = + 0.65) were previously classified as early-S replicating regions, which also show enrichment for genes with high median expression levels (Wear et al. 2017). The broader and more scattered distributions with positive correlation are still associated with open chromatin but presumably spread more diffusely across broader regions rather than colocalized to EV maxima (Fig. 5c and d). Regions enriched for H3K27me3, on the other hand, exhibited a weakly positive association with Hi-C EVs (Fig. 5e, r = + 0.13). The H3K27me3 mark is known to be associated with poised genes (Zhang et al. 2007; Makarevitch et al. 2013) and also found to be enriched in middle-S replicating regions and genes with lower median expression levels (Wear et al. 2017). The only histone mark we examined that showed a negative correlation with our root tip EV was H3K9me2 (Fig. 5f, r = −0.20), which is known to be associated with heterochromatin (Montgomery et al. 2020; Xu and Jiang 2020). H3K9me2 is typically associated with transposable elements (TEs) (West et al. 2014), which are broadly distributed across the whole maize genome, consistent with the distribution observed across chromosome 5 (Fig. S9f).

We also investigated the correlation between Hi-C EVs and transcriptional activity estimates from the RAMPAGE transcription start site mapping data, and it showed a positive correlation (Fig. 5g, r = + 0.53) (Cahn et al. 2024). The correlation plots for the RAMPAGE data show substantial dispersion, which we attribute to the natural sparsity of the data. Finally, we compared Hi-C EV with chromatin accessibility using micrococcal nuclease (MNase) sensitivity profiling datasets from maize root tip nuclei. The read coverage levels from a light MNase digestion (Turpin et al. 2018) showed a positive correlation of r = + 0.48 with Hi-C EVs (Fig. 5h, Fig. S9h). Likewise, the open chromatin defined by sequencing small fragments from light MNase digests in purified G1 nuclei showed a positive correlation of r = + 0.42 with Hi-C EVs (Fig. 5i, Fig. S9i). G1 MMOA-seq data dispersion is also linked to natural data sparsity. These were calculated for chromosome 5 because it was selected for additional cytological analysis, but similar trends were observed genome-wide for correlations between these same epigenomic features and both Hi-C and RT (Fig. 5j).

Given that RT is intimately coupled to a multitude of chromatin features, we tested the effect of including Repli-seq data inputs for chromatin state (CS) analyses. For this analysis, we performed high-resolution (200 bp) chromatin states analysis using ChromHMM (Ernst and Kellis 2010), employing two input sets: one combining eight chromatin features with three Repli-seq datasets, and the other with the same eight features excluding Repli-seq data. The eight common chromatin features used in both training sets included ChIP-seq for H3K27ac, H3K4me3, H3K4me1, H3K56ac, H3K27me3, and H3K9me2, plus RAMPAGE for transcription and light digest MNase for open chromatin. Among the tested models, from 3-state to 12-state, we selected the 6-state models for further analysis based on the biological interpretability and the transition parameters, which provide a measure of state resolution quality (Figs. S10, S11). The inclusion of Repli-seq resulted in a more balanced distribution of chromatin state abundances (Fig. S10a, b). Including Repli-seq data as input, CS1 and CS2 are distinguished by the absence (CS1) or presence (CS2) of active gene histone marks, H3K4me3, H3K4me1, and H3K27ac. The middle-S replicating regions are similarly distinguished by the absence (CS4) or presence (CS3) of these same active gene histone marks (Fig. S10c–e).

We were able to integrate RT data at high resolution along with epigenomic features of chromatin structure, using transition state data. We found that including Repli-seq data improved the resolution across the entire range of trained models (Fig. S10i). Overall, we conclude that ChromHMM better resolves chromatin states when RT data is included. Similar to the replicative labeling patterns (Fig. 1b), the tendency for early and middle S in similar chromatin states (see emission heatmaps of 3, 4, and 5 state models; Fig. S12) provides additional evidence that these two RT classes represent functionally unique subtypes of euchromatin (Fig. S12, and compare Figs. S10c CS2 to CS3).

3D cytological evidence for the spatial bifurcation of euchromatin into early-S and middle-S chromatin regions

Having shown that Hi-C data aligns well with the two-compartment mini-domain model (Fig. 1h), we undertook follow-up experiments to further validate our findings using a different type of evidence, 3D cytology with chromosome painting. For this approach, we used FISH probes designed to detect early-S and middle-S RT sequences within fixed nuclei (Fig. 6). We found that these two RT segment class frequencies for the short arm of chromosome 5 (chr5S) closely matched those of the whole genome (Fig. S13a). For instance, early-S segments were 30% of chr5S and 30% of the whole genome, whereas middle-S segments were 38% and 35% of chr5S and the genome, respectively. Therefore, we designed RT-class-specific oligo FISH painting probes at comparable densities for early-S and middle-S segments spanning the entire chr5S (Fig. S13b). These collectively account for 68% of chr5S, and the probe distributions match those of the RT segment classes (Fig. 6a). These are better seen by focusing on a 2-MB region from the middle of chr5S (Fig. 6b).

For this RT class oligo FISH painting experiment, early-S and middle-S segment probe sets were labeled with different fluorophores and hybridized to a mixed population of nuclei from the entire cell cycle, made by pooling G1, early-S, middle-S, late-S, and G2 nuclei (Fig. S14d). Representative images from 3D deconvolution image datasets are shown as 1-µm Z-stack projections, a five-optical-section projection spanning the FISH probe signal regions. For quantitative analysis, we selected nuclei in which two similarly sized paint signals were observed for diploid chromosome 5 (Fig. 6c, circles). In these nuclei, non-specific background staining of the nucleolus (labeled “n”) was commonly seen, but the maize nucleolus organizing region (NOR) is not located on chromosome 5, and this non-specific signal was disregarded. To quantify the DAPI signal in the FISH-painted regions, we manually segmented the FISH paint signals to create volumetric polygon objects to obtain DAPI density (average photon counts per cubic micron). As a control, random non-FISH euchromatin regions were selected, avoiding the nucleolus and bright knobs (described in Methods). In this way, we could obtain the DAPI values for chromatin in painted RT class probe regions relative to bulk euchromatin control regions (Fig. 6d). By plotting the log₂ ratio of DAPI density in the FISH painted versus control regions, we showed that the average DAPI concentration in early-S FISH chromatin (mean of −0.26) was less than the control regions, whereas the middle-S FISH chromatin (mean of +0.21) was higher. The ratios of RT painted versus control region values do not overlap for early-S versus middle-S probe sets (Fig. 6e, box plot). Because we observed this pattern for measured, individual nuclei from a mixture of G1, S, and G2 phases, where only ∼40% of the nuclei were in S phase, we concluded that the relationship between RT and DAPI density applies to each and every nucleus measured (Fig. 6e), and not just those in S-phase.

To more definitively establish the spatial bifurcation of euchromatin outside of S phase, we performed dual-probe FISH experiments using early-S and middle-S probe sets on purified G1 nuclei, as summarized in Fig. 7. Representative images of dual-labeled G1 nuclei (Fig. 7a) showed spatial separation of early-S and middle-S probe regions corresponding to the same unreplicated chromosome. We repeatedly observed paired but spatially separated signals for the early-S versus middle-S probe sets, in contrast to the intermixed, linear arrangement of early-S and middle-S segments along the chromosome (Fig. 1d). By spatially segmenting the FISH signal regions, we could directly compare DAPI concentration in early-S chromatin (green outlines) relative to middle-S (red outlines) chromatin from the same chromosome, avoiding any uncertainty in selecting “control” regions from other parts of the nucleus. We quantified the DAPI intensity for each chromosome 5S-based pair of FISH signals (Fig. 7b), providing a log₂ ratio that captures the relative chromatin compaction of the two FISH-delineated areas. We found that the average DAPI concentration ratios for all the early/middle FISH region pairs were 0.86 (sd 0.07) in the G1 nuclei (Fig. 7c; expressed as log₂ ratio), clearly establishing that RT defined regions are spatially separated and differentially condensed, even outside of S phase.

These results validate the mini-domain chromatin fiber RT model, demonstrating that maize euchromatin exists as an interspersed mixture of two compartments distinguishable by condensation state and RT. Early-S FISH probes are preferentially localized in regions with low DAPI signals, whereas middle-S FISH probes coincide with regions of higher DAPI intensity. 3D dual-label RT FISH painting in G1 nuclei provided compelling evidence, consistent with Hi-C data, that the two euchromatin compartments are integral features of nuclear organization.

Discussion

This study sheds light on the close relationship between DNA RT and chromatin organization in the maize genome, focusing particularly on nuclei in the terminal 1 mm of actively growing root tips and their euchromatin compartments. By using two orthogonal methods for defining nuclear architecture, Hi-C and 3D FISH, we demonstrated how RT reveals a spatial bifurcation of maize euchromatin (the global A) into separate early-S and middle-S compartments. These findings support our earlier mini-domain chromatin fiber RT model (Bass et al. 2015; Savadel and Bass 2017) and extend it into G1 (Fig. 7a–c), outside of S phase.

Other studies in eukaryotic systems point to the fact that simple global A and B compartments are widely conserved but are insufficient to understand the complexity of the structural and functional organization of chromatin. Mammalian chromatin, for instance, has been divided into a variable number of states or subcompartments, ranging from three to eight or more, based on various approaches including epigenomic marks, modularity-based analyses, trans contacts, hierarchical clustering of domains, and data from TSA-seq, DamID, and Hi-C (Rao et al. 2014; Nichols and Corces 2021; Wang et al. 2021; Wen et al. 2022; Kalluchi et al. 2023). Similarly, plant chromatin, primarily from Arabidopsis studies, has also been shown to possess multiple chromatin states defined by distinct features (Roudier et al. 2011; Sequeira-Mendes et al. 2014; Concia et al. 2018; Liu et al. 2018). These studies tend to agree that the global A compartment can be split into two subcompartments, A1 and A2, which replicate in that order in plants and animals (Rao et al. 2014; Dong et al. 2017; Wen et al. 2022). We speculate that the maize early-S and middle-S chromatin described here reflects an A1 and A2 type of subdivision, for which a single assay, RT profiling, can be used to map. Consistent with this, a recent light and EM cytological analysis of the very large plant genome, Nigella damascena, also revealed that early-S versus middle-S replication partitioned into more open versus less open chromatin, respectively (Arifulin et al. 2024).

Here, we show fine-scale interspersion of replicon-size chromosome segments with strikingly different RT, and those timings correlate well with both Hi-C and 3D molecular cytology (Figs 2, 6, and 7). This type of finely interspersed early-S versus middle-S chromatin exhibits a clear tendency to alternate correspondingly with Hi-C measures of chromatin structure. The close interspersion of these patterns highlights differences in RT and chromatin features within what has previously been regarded as undifferentiated euchromatin.

Evidence for the nonrandom spatial distribution of early, middle, and late RT domains is supported by genomic RT profiling with Repli-seq (Wear et al. 2017) and 3D quantitative microscopy. Within the S phase, using replicative labeling, we know that early replication predominantly occurs in distal, gene-rich euchromatic regions with open chromatin features, whereas middle replication corresponds to denser euchromatin associated with repetitive elements and less active genes (Bass et al. 2015). We now show that the close interspersion of early and middle RT domains (e.g. Figure 1g) correlates with alternating chromatin states and persists in interphase nuclei.

We have also documented the robust covariance between chromatin interactions and RT annotations. Early RT regions exhibit higher contact frequencies with other early regions, even across long genomic distances, as evidenced by the EV and IS analyses (Figs 2 and 4). Middle RT regions exhibit a lesser tendency for long-reaching interaction (Figs 2e–g and 4d, j) and show a more condensed chromatin structure compared to early RT regions (see Figs 6e and 7c). These “same-RT-class” interactions may reflect different functions. For example, early-to-early interactions might be largely based on transcriptional activity, while middle-to-middle interactions, occurring in regions with fewer active genes and more abundant retroelements, might reflect a chromatin packaging function.

In general, it is known that the chromatin regions that replicate at different time points in S phase, such as early versus late, also show distinct enrichments for epigenomic features in plants and animals (Lee et al. 2010; Dileep et al. 2015; Wear et al. 2017). For instance, early-S replicating regions generally exhibit open and accessible chromatin, enrichment for active chromatin marks such as histone acetylation, and relatively higher levels of transcriptional activity. On the other hand, late-S replicating regions generally exhibit the opposing features, such as closed and repressive chromatin marks.

In contrast to the established binary framework of S phase, which divides RT into early and late stages, our study highlights a tripartite RT model consisting of early-, middle-, and late-S phase. We focused specifically on the early and middle stages, as both exhibit broad nucleoplasmic distribution and map to the global Hi-C-defined A compartment, yet they display distinct molecular features. Notably, the chromatin state CS3 (Fig. S10c) includes middle-S replicating regions and active epigenomic marks, similar to the chromatin state CS2, which includes early-S replicating regions and active epigenomic marks (Fig. S10c; emission plot, compare CS2 and CS3). However, the association of middle-S with active histone marks appears contradictory to our genome-wide correlation results (Fig. 5). We attribute this apparent discrepancy to the fact that the CS3 is only a small proportion (∼4%, Figs. S10a, CS3) of the total genome, whereas all the middle-S replicating regions occupy nearly one-third of the genome. The active gene chromatin histone acetylation mark H3K27ac had correlation values of r = + 0.72 for early-S versus r = −0.61 for middle-S chromatin. Similarly, but in the opposite direction, the gene-repressive chromatin mark H3K9me2 had correlation values of r = −0.19 for early-S versus r = + 0.044 for middle-S chromatin (Fig. 5), consistent with previous work (Wear et al. 2017). In addition, transcriptional characteristics of genomic regions, including transcription start site mapping and transcriptomics, also tend to mirror the pattern of epigenomic marks with regard to gene activity. Specifically, analyzing expression levels for promoters using the RAMPAGE assay (Batut and Gingeras 2013; Cahn et al. 2024), we observed a correlation of r = + 0.55 for early-S chromatin and r = −0.12 for middle-S chromatin.

Both chromatin state and genome-wide correlation analyses indicate that middle-S shares more characteristics with early-S than it does with late-S. This result is consistent with the replicative labeling patterns that initially led us to classify middle-S as euchromatin. However, chromatin state analysis with RT inclusion also revealed that some early-S regions (CS1, 14% of the genome) as well as some middle-S regions (CS4, 18% of the genome) lack the active epigenomic mark associations and have a high overlap with transposable elements (Fig. S10e). These findings further illustrate the fact that TEs in the maize genome can populate regions considered by multiple criteria to be euchromatic.

The FISH experiments also provided new and compelling cytological evidence for the bifurcation of the global A compartment into early and middle chromatin. Unlike our previous work, which was limited to replicative labeling within early-S or middle-S, our RT-FISH approach allowed us to examine the spatial location of the RT-defined sequences outside of S phase and independent of replicative labeling. We carried out two related types of experiments. The first experiment used early and middle RT FISH probes separately on total nuclei and quantified the DAPI concentration in the FISH-defined regions relative to a similarly sized chromatin control region with properties typical of bulk euchromatin (Fig. 6). The second experiment combined early and middle probe sets labelled with different fluorophores, hybridizing them together on purified G1 nuclei. This procedure enabled us to simultaneously visualize the spatial separation and nuclear localization of both RT-FISH signals, as well as to quantify their relative DAPI densities, without the need for an external control region (Fig. 7). In every case, we observed signal partitioning that validates the two-compartment architecture, which we now know is present in G1 nuclei. This finding demonstrates that RT-related nuclear architecture exists prior to S phase, and is not simply a result of processes in S phase itself.

Given that Hi-C compartments and RT are tightly coupled in the root tip, we reasoned that one could ask to what extent RT programs might be conserved in other tissues, using Hi-C structure as a proxy for RT data. We showed that biologically vastly different organs, earshoot versus root tip, had strong correlations for Hi-C-based compartmentalization genome-wide, with r = + 0.85 for both EV and IS (Fig. S5). When correlating RT from root tip with Hi-C measures from earshoot or bundle sheath, the correlations were similar in strength and direction to those from within-tissue comparisons (Fig. 3). These findings provide evidence in support of the idea that the two-compartment architecture of euchromatin (as depicted in Fig. 1c–e) not only extends throughout the cell cycle in root tips but appears to exist in other tissues as well. This observation raises the intriguing possibility that this pattern of nuclear architecture may reflect a broader organizational principle in other large-genome eukaryotes. The practical implications of understanding RT in model crop species are many. Knowledge of replication programs and their control also raises the possibility of finding new ways of modifying genome function for agronomic purposes, including the addition of RT regulators, self-replicating elements, recombination control for breeding, and artificial chromosomes.

Materials and methods

.

Plant materials

Experiments were done using maize inbred B73 root tips or ear shoots. For root tips, as described in a previous protocol (Mickelson-Young et al. 2022), seeds were surface sterilized and imbibed overnight with constant stirring and aeration. Seeds were surface-sterilized again with 0.5% sodium hypochlorite, 0.05% Tween-20 solution, and then rinsed with water. Seeds were germinated in boxes with paper towels moistened with sterile water at 28 °C for 3 days under continuous light (Feit Electric OneSync LED light system, 3,000 K, 300–400 lux). The material for biological replicates was grown independently and harvested on different days. For earshoots, field-grown maize plants were harvested between 9 and 11 am on sunny days, and earshoots ranging in size from 0.5 to 1 cm were flash frozen in liquid nitrogen and pooled (15 to 20 earshoots) for storage at −80 °C as described in a previous paper (Savadel et al. 2021).

Repli-seq realigned to B73v5

For this study, we used previously published Repli-seq data generated from maize B73 root tips that were EdU labeled, dissected (terminal 1 mm), and fixed as described (Wear et al. 2017). In isolated nuclei, the EdU was conjugated to Alexa Fluor 488 (AF-488), and total DNA was stained with DAPI for flow sorting of early-S, middle-S, and late-S nuclei, as well as non-replicating G1 nuclei, based on both EdU incorporation and DNA content (like in Fig. 1a). After DNA extraction, the EdU-labeled DNA was immunoprecipitated from the S-phase fractions and sequenced. The raw sequence data from early-S, middle-S, or late-S and the G1 reference libraries were aligned to the current working version of the maize genome, Zm-B73-REFERENCE-NAM-5.0 (B73v5), using BWA mem (v0.7.15) (Heng 2013) with default parameters. Aligned sequences that were properly paired and uniquely aligned (MAPQ score >10) were retained with SAMtools (Li et al. 2009) and duplicates removed using Picard MarkDuplicates (v2.7.1). For genome segmentation and annotation, we used the previously custom-built Repliscan software as described (Zynda et al. 2017). Highly correlated replicates (Wear et al. 2017) were pooled to produce read densities in different bin sizes (1, 3, 5, and 10 kb) across the genome. The various bin sizes were used for resolution analyses, and all four bin sizes for all three RT samples are available as bigwig files (File S1). Repliscan normalization was achieved in two ways; (1) by sequence depth scaling to 1 × coverage using the reads per genomic content (RPGC) method, and (2) by normalization to a non-replicating reference (G1) to correct for sequencing biases, read mappability variation, and copy number. Genomic bins with the lowest 1% from the reads per bin distribution in G1 were also removed. The final replication profile data for each of the S-phase samples was calculated as a ratio of the S-phase sample to the G1 reference value in each bin and represents an estimate of the intensity of replication activity in that bin (“replication signal”) for that portion of S phase. Finally, replication signal profiles for each S-phase sample were Haar wavelet smoothed (Percival and Walden 2013). The final step in the Repliscan segmentation incorporates the replication signal for each S-phase portion in each bin and makes a qualitative call of the predominant replication time in that bin. The largest replication signal value is always classified as replicating, and the algorithm allows multiple time classifications (e.g. early-mid or middle-late) when another signal is within 50% of the highest value. The parameters for maize B73 described in Mickelson-Young et al. (2022) were used except for the following minor modification: each bin was assigned to a RT segment class using a segmentation threshold of 1.0 (−threshold value, −value 1.0), so that the segmentation algorithm included any window with early, middle or late replication signals that are greater than the G1 reference signal (> ratio of 1.0). The resulting Repliscan segmentation file assigned unique and mutually exclusive RT classes across the genome. We primarily used the early (E), middle (M), or late (L) segments, which accounted for ∼30%, ∼35%, or ∼24% of the maize genome, respectively.

Root tip and earshoot nuclei for Hi-C

For Hi-C on seedling root nuclei, the terminal 0 to 1 mm of the primary and first two seminal root tips were dissected and crosslinked with 1% formaldehyde for 15 min. Nuclei were isolated, stained with DAPI (No EdU), and flow sorted on a BD FACSAria III equipped with UV (355 nm) and blue (488 nm) lasers and using flow cytometer settings and doublet discrimination gating as previously described (Wheeler et al. 2025). The sorted nuclei were collected in ∼1.8 mL CLB⁺ buffer (cell lysis buffer with 1 mL DTT and protease inhibitor tablet) as described (Mickelson-Young et al. 2022). After excluding doublets, debris, and endocycling nuclei, approximately 1.3 to 1.5 million nuclei were collected as input for each biological replicate of the Hi-C assay. A representative FANS profile is shown in Fig. S1d.

For Hi-C on field-grown earshoot nuclei, we used the method of Savadel et al. (2021) to isolate nuclei. The frozen tissues were ground in liquid nitrogen to a fine powder, and the nuclei were crosslinked in 1% formaldehyde. The fixation reaction was stopped by the addition of 0.1 vol (∼1 mL) of 2.5 M glycine. The tissue was then mechanically disrupted with a Polytron (3 × 10 seconds at ∼ 1/5 maximum speed) to liberate nuclei. Nuclei were pelleted by centrifugation at 2,000 × g for 15 min at 4 °C and resuspended in a 15 mL CLB⁺ buffer. Total earshoot nuclei were used for the Hi-C assay.

Preparation and sequencing of hi-C libraries from maize root tip and ear shoot

The Hi-C libraries were generated essentially following the protocol first described for maize leaf by Dong et al. (2017) and later detailed in a methods chapter by Dong and Zhong (2020). We made some minor modifications to accommodate the library kits currently used by the Molecular Cloning Facility (Department of Biological Science, Florida State University, Tallahassee, FL) and reduced washing steps to avoid nuclei loss as described in the wet bench full protocol (File S2). Among the key steps were permeabilization of nuclei with 0.5% (w:v) SDS for 5 min at 65 °C, DNA digestion with Dpn-II restriction enzyme, overhang repair with Klenow and biotinylated nucleotides, and ligation with T4 DNA ligase. Ligated DNA was biotin affinity purified, crosslinks reversed and sheared to 300–500 bp to yield DNA fragments for library construction using NEBNext Adapters for Illumina. Libraries from both root tips and earshoots were sequenced using Illumina Hi-Seq 2500 paired-end 150-bp chemistry (Translational Science Lab at College of Medicine, Florida State University, Tallahassee, FL) to obtain library sequences at a depth ranging from 127 to 309 M (Table S1) per replicate, with four replicates for root tips and three for earshoots. We used FastQC (v0.12.1) to assess the read quality and then performed trimming using trimmomatic (v0.39) to remove low-quality bases and adapter sequences. The trimmed reads were used in the HiC-pro pipeline as described in the HiC-Pro Data Processing Pipelines file (File S3).

Nuclei isolation for MMOA-seq

For Mononucleosomal MNase-defined Cistrome-Occupancy Analysis (MMOA-seq), nuclei preparation was done using our previously published protocol (Bass et al. 2014; Wear et al. 2017). Briefly, maize seedling roots were labeled in 25 µM EdU for 20 min, rinsed, and placed in 100 µM thymidine to stop labeling. The terminal 1-mm maize root tip segments were excised, fixed with 1% formaldehyde, and lysed to isolate nuclei for MMOA-seq. Incorporated EdU was conjugated to AF-488 using a Click-iT EdU Alexa Fluor 488 imaging kit (Life Technologies). The nuclei were then counterstained with DAPI using a cell lysis buffer containing 2 μg/mL DAPI and 40 μg/mL Ribonuclease A and filtered through a 20-μm nylon mesh filter (Partec). Flow-sorted G1 nuclei were used for MMOA-seq library preparation.

MMOA-seq library preparation and sequencing

MMOA-seq libraries were prepared using nuclei in 500-µL aliquots per replicate. Briefly, 60 µL of nuclei were distributed to each of four 1.5 mL screw cap tubes. 10 × MNase working dilutions (25, 12.5, and 6.25 U/mL) were prepared from a 20,000 U/mL stock. 6.7 µL of each working dilution (or MDB) was added to each of the 60 µL nuclei aliquots and immediately vortexed and briefly centrifuged. Complete digestion reactions were transferred to a 37 °C shaking water bath and incubated for 15 min. Reactions were promptly stopped after 15 min by the addition of 5 µL of 0.5 M EGTA (pH 8.0). Digested chromatin was then decrosslinked overnight at 65 °C in 1% SDS, 150 mM NaCl, 20 µg/mL Proteinase K. DNA was purified by 25:24:1 Phenol:Chloroform:Isoamyl alcohol (pH 8.0). Nucleic acids were precipitated,d and the pellets were washed with ice cold 70% ethanol and air-dried before being redissolved in 100 µL of 10 mM Tris EDTA buffer.

For each library, 75 ng of selected digests (1.25 and 2.5 U/mL MNase) were combined and diluted to 50 µL with 0.1 × TE. Sequencing libraries were made according to the NEBNext Ultra II DNA Library Prep Kit for Illumina with the following modifications to the size selection protocol for retention of small fragments (<200 bp): the first bead addition used 50 µL of beads, the second used 100 µL of beads. These size selection steps were chosen to collect both mononucleosome-sized DNA and smaller sub-nucleosome-sized DNA fragments for sequencing library preparation. A total of 8 cycles of PCR were completed at the barcoding step.

Libraries were quantified by Qubit dsDNA HS fluorimetry and KAPA qPCR. DNA fragment size distribution was analysed on an Agilent High Sensitivity D1000 Screentape assay. An equimolar pool of all 24 MMOA-seq libraries was prepared and sequenced using 50-bp paired-end reads on a Novaseq S1 flow cell at the College of Medicine Translational Science Lab, Florida State University, Tallahassee, FL.

HiC-pro analysis of Hi-C data

HiC-Pro (Servant et al. 2015) is a custom bioinformatic software for the analysis of Hi-C datasets. The pipeline performs sequential steps, including short read mapping, detection of valid ligation fragments, and various quality controls. The output consists of inter- and intra-chromosomal contact maps at various resolutions.

Trimmed and filtered paired-end reads were aligned to the B73v5 maize reference genome using bowtie-build software (v2.4.4). Then HiC-pro (v3.1.0) was used to generate the contact matrices (File S3). The genome was divided into non-overlapping bins of equal size at different resolutions (1 kb, 5 kb,10 kb, 50 kb, 250 kb, 500 kb, and 1 mb), and contacts were scored in each bin. The frequency of interaction between bins was represented by bi-dimensional heatmaps (“contact matrices”), containing both inter- and intra-chromosomal contacts. Contact matrices were visualized using Juicebox (Durand et al. 2016) and used to calculate a Pearson correlation matrix for observed/expected intra- and inter-chromosomal interactions. The KR normalization method was applied to all Hi-C matrices. KR is an implicit normalization method through matrix balancing, and is used as a valid alternative to explicit normalization of sequencing biases (Knight and Ruiz 2013; Ay and Noble 2015).

Insulation score and eigenvector analysis

To systematically identify folding structures at the local scale, we applied several techniques, including PCA, the Insulation Index method, and Virtual 4C analysis. We ran chromosome-wide PCA to calculate the first EVs (principal components) for three binning resolutions (25, 50, and 100 kb), and processed files are available on figshare (File S4). The EV values are commonly used to infer genomic compartments as initially described in Lieberman-Aiden et al. (2009). Briefly, Hi-C data is first processed into a matrix where each entry represents the frequency of contact between two genomic loci. The raw contact matrix is normalized to correct for biases, and PCA is performed on the normalized contact matrix. The PCA gives every bin a score (positive, negative, or near zero). If two bins each have high scores of the same sign (positive or negative), they tend to interact more. If they have high scores of opposite signs, they tend to interact less. This way, it reveals finer details about chromatin organization. The assignment of the EV sign is arbitrary, but here we oriented it (based on other genomic features) such that positive EVs correspond to open chromatin and negative EVs correspond to closed chromatin. We also performed chromosomal block-based EV (BEG) analysis at 50-kb resolution, using the constrained hierarchical clustering method as described for maize by Dong et al. in 2017 (Grimm 1987; Dong et al. 2017). A series of IS analyses (Crane et al. 2015; Kruse et al. 2020) was calculated, quantifying the cumulative frequency of interactions in each bin size series (1 kb, 10, 25, and 50 kb) and visualized within 1-mb or 2.5-mb sliding windows using the bin sizes as step sizes (File S5). We calculated the log₂ of the observed/expected interaction frequencies using the median scores for the expected values. Valleys/minima (negative values) indicate loci of reduced frequency of interaction with flanking regions, whereas local maxima (positive values) reveal folding structures.

Virtual 4C analysis (Yardımcı and Noble 2017) was centered on “viewpoints” of interest (“baits”), chosen in our case among replication segments of different classes, early-S or middle-S. We scored all the valid interactions between each viewpoint and the rest of the chromosome and compared the frequency of early-to-early and middle-to-middle versus early-to-middle interactions.

Chromatin state analysis using a hidden Markov model

For inputs used with chromatin state analysis, we obtained published data sets from B73 maize root tip samples, including histone modification ChIP-seq (Wear et al. 2017; Cahn et al. 2024), MNase-seq (Turpin et al. 2018), RAMPAGE (Cahn et al. 2024), and Repli-seq (Wear et al. 2017). The data source files are cited in Table S2, and the raw fastq files were aligned to B73v5 as described in GitHub (See data availability statement).

Briefly, fastQC software (v0.12.1) was used for quality control, and low-quality reads were filtered out. Adapter cleanup was done by trimmomatic (trimmomatic_latest.sif). Bowtie2 software (v2.4.4) was used to align sequencing reads to the reference genome (B73v5 in maize) with the default parameters. Duplicate reads were removed by the samtools software (v1.9-4). Bam files were converted into bed files (File S6) and used as an input for chromHMM binarization.

The Hidden Markov Model was applied to aggregate the multi-dimensional matrices into chromatin states (Ernst and Kellis 2010; Ernst et al. 2011) using ChromHMM software (v1.26-0) with a bin size set to 200 bp. The LearnModel program of ChromHMM was used to learn the chromatin state model with the numstates being set as 3 to 10 and 12 states. The resulting output files for states with and without RT are available (File S7), and they include bed files for genome browser visualization.

Statistical analysis

For virtual 4C arc plots, we used plotGardner v1.0.17 (Kramer et al. 2022) package for the R statistical suite (The R Project for Statistical Computing n.d.) for the visualization of genomic regions. Two similarly sized early and middle RT segments were selected as 4C baits, and then their corresponding 10-kb bins were extracted. Each 10-kb bin was annotated with its RT class, and all bins in contact with the bait bins were retrieved. Hexbin plots were also drawn using the R package ggplot2 v4.1.2 (Wickham 2016) with Hi-C EV versus different chromatin marks and early-S, middle-S RT profiles versus chromatin marks at 50-kb resolution. R scripts used for plots are available on GitHub (See data availability statement).

Root tip nuclei for 3D-FISH

Maize seedlings were grown, and nuclei were prepared and sorted as described above for MMOA-seq. For 3D quantitative FISH with the mixed population nuclei, we combined 30 µL each of the flow-sorted EdU-labeled nuclei from flow sorting gates (Fig. 1a) for G1 (1.5 M/mL, 16%), early-S (1.9 M/mL, 21%), middle-S (1.5 M/mL, 16%), late-S (0.36 M/mL, 4%), and G2 (3.9 M/mL, 41%). In this mix, the S-phase nuclei were ∼41% of the total. For 3D-FISH with G1 nuclei, only flow-sorted G1 nuclei were used (Fig. 7c, inset panel).

Probe labeling for 3D-FISH

Oligonucleotide probe libraries were designed by DAICEL Arbor Biosciences using three sets of sequences as inputs, early-S, middle-S, and late-S RT segments. These annotation segments were based on the Repli-seq data from Wear et al. (2017), realigned to B73v5 using the Repliscan pipeline (Zynda et al. 2017) to produce a DNA RT class annotation file, RT_class_ALL_27800_b73v5_9colBed_vhsf521e.BED (File S8). From this annotation file, we derived coordinates for two sets of genomic regions (early or middle) within the first 105 Mb on chromosome 5, which corresponds to the short arm. These were used as input to obtain two libraries of oligos that were T_m-matched ∼45-mers, high-density, and uniformly spaced at high density across the chromosome arm. The resulting sets of oligos included a total of 22,460 or 27,390 oligos from the early-S or middle-S segment regions, respectively, and the coordinates of the uniquely named oligos are provided as bed files (File S9).

A series of experiments was conducted to convert these dsDNA libraries into fluorescently labeled single-stranded DNA (ssDNA) libraries (File S10). Briefly, the dsDNA library underwent PCR amplification to obtain a DNA yield sufficient for in vitro transcription. The purified DNA was processed using the Qiagen QIAquick PCR Purification Kit and quantified using a spectrophotometer (NanoDrop). After DNA purification, in vitro transcription was carried out using the MEGAshortscript TM T7 Kit (Thermo Fisher), using 480 ng/μL DNA as template per reaction. Following transcription, RNA purification was performed using the RNA Clean-Up Mini Kit (Macherey-Nagel). RNA was quantified by NanoDrop to verify that the concentration was above 1 μg/μL, as required by the reverse transcription PCR (RT PCR) step. Single-stranded DNA (ssDNA) was generated through a reverse transcription reaction using SuperScript IV Reverse Transcriptase, with the addition of 52 μg of RNA to obtain a sufficient yield of ssDNA. At this stage, ssDNA probe pools were labeled with any of three different fluorophores (Eurofins Genomics US; green, ALP_a488b: [Alexa488]; red ALP_a594a: [Alexa594]; or far red, ALP_a647n: [ATTO647N]). The process resulted in RNA-DNA hybrids and some unincorporated primers. Exonuclease-I was used to digest unincorporated primers, and RNase was used to digest the RNA from RNA-DNA hybrids. After RNA and primer digestion, labeled ssDNA was purified using the Zymo Quick-RNA purification kit and quantified.

Fluorescent in situ hybridization

The FISH protocol was adapted from Bass et al. (1997) as detailed by Howe et al. (2013), using a polyacrylamide embedding technique for 3D FISH analysis. Fixed EdU-labeled nuclei were embedded in a thin layer of optically clear 3 × acrylamide mix on a glass slide. Then, wash buffers and prehybridization buffers were exchanged by addition and removal of 200-µL drops. A buffer containing RT class-labeled probes, 50% formamide, and 2 × SSC was added and incubated at 37 °C for 30 min for prehybridization. After prehybridization, incubation slides were placed on a hot plate at 65 °C for 30 min for genomic DNA denaturation. After denaturation, the slides were placed back in the 37 °C incubator overnight for hybridization. The following day, the slides underwent serial washing with buffers of increasing stringency. Finally, slides were mounted with mounting media and sealed with Sally Hansen Hard as Nails polish.

Deltavision microscopy and image analysis

Following the FISH experiment, 3D images were captured with 0.2-micron projections by multiple wavelength iterative deconvolution microscopy. The raw data were subsequently subjected to 3D iterative deconvolution and chromatic aberration correction (CAC). For analysis purposes, after 3D data collection, individual nuclei with probe signals were cropped to allow quantitative signal colocalization analysis using the in-built solid object builder polygons method.

We used the DeltaVision software (softWoRx v6.5.2) programs EditPolygon and VolumeBuilder to segment the FISH signal regions in 3D. We used EditPolygon to trace the edges of the FISH signals manually with a mouse, and VolumeBuilder to connect the polygon series into a single 3D volumetric object with a closed and continuous surface area. Using VolumeBuilder, we obtained the object volumes summed intensities (sum of all voxels within the object) for DAPI, which we used to calculate the DAPI concentration. For these analyses, the control regions were similarly built in non-FISH “bulk euchromatin” areas in the same nucleus, avoiding heterochromatic and nucleolar regions. Quantitative segmentation data analysis was performed on at least 50 nuclei for each of the early and middle probe sets. All 3D images were stored on the image repository omero.bio.fsu.edu, allowing for additional segmentation analysis.

Supplementary Material

koag042_Supplementary_Data

koag042_supplementary_data.zip^{(20.2MB, zip)}

Acknowledgments

The authors would like to acknowledge help with early versions of the data analysis by Dr. Jawon Song and Joshua Urrutia at the Texas Advanced Computing Center.

Contributor Information

H S Akram, Department of Biological Science, Florida State University, Tallahassee, FL 32303, United States.

E E Wear, Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695, United States.

L Mickelson-Young, Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695, United States.

Z M Turpin, Department of Biological Science, Florida State University, Tallahassee, FL 32303, United States.

L Hanley-Bowdoin, Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695, United States.

W F Thompson, Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695, United States.

L Concia, University of Texas at Austin, Texas Advanced Computing Center, Austin, TX 78758, United States.

H W Bass, Department of Biological Science, Florida State University, Tallahassee, FL 32303, United States.

Author contributions

H.W.B., L.H-B., and W.F.T. conceived the study. H.S.A., H.W.B., L.C., L.H-B., and W.F.T. designed the experiments. H.S.A. prepared Hi-C libraries and performed 3D FISH experiments, collected and analysed the data. Z.M.T. prepared MMOA-seq libraries. L.C. and H.W.B. contributed to the analysis and interpretation of data. E.E.W. and L.M-Y. provided maize root tip nuclei for the study. H.S.A. and H.W.B. drafted the manuscript, and all authors critically revised and approved the final manuscript.

Supplementary material

Supplementary material is available at The Plant Cell online.

Funding

This work was supported by a grant from the National Science Foundation Plant Genome Research (PGRP) (NSF IOS 2025811 to L.H-B., L.C., W.F.T., and H.W.B.), awards from Florida State University (Dean's award for Doctoral Excellence to HSA; Ben and Karen Thrower award to HSA), and funds from North Carolina State University.

Data availability

Raw Hi-C data, MMOA-seq data, processed all valid pairs files, Hi-C files, and bigwig files can be accessed at NCBI GEO under the accession number GSE287128. Scripts for Hi-C data analysis can be accessed at

https://github.com/Sarachaudry/HiC-contacts-with-RT-annotations

https://github.com/Sarachaudry/Correlation_Hexbin_plots

https://github.com/Sarachaudry/virtual-4C-interactions-with-Sushi-in-R

https://github.com/Sarachaudry/ChromHMM-Chromatin-States-Analysis

https://github.com/Sarachaudry/R-script-for-scatter-plot

GEO accession GSE287128: (reviewer token azojyygqvfyhjcl)

Material and correspondence

Material request from the corresponding author, Hank W. Bass.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

root tip AmiGo: PO:0000025

References

Arifulin EA et al. 2024. Global nuclear reorganization during heterochromatin replication in the giant-genome plant Nigella Damascena L. Plant J. 120:1508–1521. 10.1111/tpj.17063. [DOI] [PubMed] [Google Scholar]
Ay F, Noble WS. 2015. Analysis methods for studying the 3D architecture of the genome. Genome Biol. 16:183. 10.1186/s13059-015-0745-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Baker A et al. 2012. Replication fork polarity gradients revealed by megabase-sized U-shaped replication timing domains in human cell lines. PLoS Comput Biol. 8:e1002443. 10.1371/journal.pcbi.1002443. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bass HW et al. 2014. A maize root tip system to study DNA replication programmes in somatic and endocycling nuclei during plant development. J Exp Bot. 65:2747–2756. 10.1093/jxb/ert470. [DOI] [PubMed] [Google Scholar]
Bass HW et al. 2015. Defining multiple, distinct, and shared spatiotemporal patterns of DNA replication and endoreduplication from 3D image analysis of developing maize (Zea Mays L.) root tip nuclei. Plant Mol Biol. 89:339–351. 10.1007/s11103-015-0364-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bass HW, Marshall WF, Sedat JW, Agard DA, Cande WZ. 1997. Telomeres cluster de novo before the initiation of Synapsis: a three-dimensional spatial analysis of telomere positions before and during meiotic prophase. J Cell Biol. 137:5–18. 10.1083/jcb.137.1.5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Batut P, Gingeras TR. 2013. RAMPAGE: promoter activity profiling by paired-End sequencing of 5′-complete cDNAs. Curr Protoc Mol Biol. 104:Unit 25B.11. 10.1002/0471142727.mb25b11s104. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cahn J, et al. MaizeCODE reveals bi-directionally expressed enhancers that harbor molecular signatures of maize domestication. Nat Commun. 2024:15:10854. 10.1038/s41467-024-55195-w. Erratum in: Nat Commun. 2025:16:6790. https://doi.org/10.1038/s41467-025-62179-x. PMID: 39738013; PMCID: PMC11685423. [DOI] [PMC free article] [PubMed] [Google Scholar]
Concia L et al. 2018. Genome-wide analysis of the Arabidopsis replication timing program. Plant Physiol. 176:2166–2185. 10.1104/pp.17.01537. [DOI] [PMC free article] [PubMed] [Google Scholar]
Crane E et al. 2015. Condensin-driven remodelling of X chromosome topology during dosage compensation. Nature. 523:240–244. 10.1038/nature14450. [DOI] [PMC free article] [PubMed] [Google Scholar]
Desprat R et al. 2009. Predictable dynamic program of timing of DNA replication in human cells. Genome Res. 19:2288–2299. 10.1101/gr.094060.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dileep V et al. 2015. Topologically associating domains and their long-range contacts are established during early G1 coincident with the establishment of the replication-timing program. Genome Res. 25:1104–1113. 10.1101/gr.183699.114. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dimitrova DS, Berezney R. 2002. The spatio-temporal organization of DNA replication sites is identical in primary, immortalized and transformed mammalian cells. J Cell Sci. 115(Pt 21):4037–4051. 10.1242/jcs.00087. [DOI] [PubMed] [Google Scholar]
Dong P et al. 2017. 3D chromatin architecture of large plant genomes determined by local A/B compartments. Mol Plant. 10:1497–1509. 10.1016/j.molp.2017.11.005. [DOI] [PubMed] [Google Scholar]
Dong P, Zhong S. 2020. Characterization of plant 3D chromatin architecture, in situ hi-C library preparation, and data analysis. Methods Mol Biol. 2093:147–157. 10.1007/978-1-0716-0179-2_11. [DOI] [PubMed] [Google Scholar]
Durand NC et al. 2016. Juicebox provides a visualization system for hi-C contact maps with unlimited zoom. Cell Syst. 3:99–101. 10.1016/j.cels.2015.07.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ernst J et al. 2011. Mapping and analysis of chromatin state dynamics in nine human cell types. Nature. 473:43–49. 10.1038/nature09906. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ernst J, Kellis M. 2010. Discovery and characterization of chromatin states for systematic annotation of the human genome. Nat Biotechnol. 28:817–825. 10.1038/nbt.1662. [DOI] [PMC free article] [PubMed] [Google Scholar]
Freeling M, Walbot V. editors. 1996. The maize handbook. Springer Lab Manuals. Springer. [Google Scholar]
Gent JI, Wang K, Jiang J, Dawe RK. 2015. Stable patterns of CENH3 occupancy through maize lineages containing genetically similar centromeres. Genetics. 200:1105–1116. 10.1534/genetics.115.177360. [DOI] [PMC free article] [PubMed] [Google Scholar]
Golicz AA, Bhalla PL, Edwards D, Singh MB. 2020. Rice 3D chromatin structure correlates with sequence variation and meiotic recombination rate. Commun Biol. 3:235. 10.1038/s42003-020-0932-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
Grimm EC. 1987. CONISS: a FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Comput Geosci. 13:13–35. 10.1016/0098-3004(87)90022-7. [DOI] [Google Scholar]
Heng L. 2013. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM [preprint]. arXiv 3997. 10.48550/arXiv.1303.3997. [DOI]
Hiratani I et al. 2008. Global reorganization of replication domains during embryonic stem cell differentiation. PLoS Biol. 6:e245. 10.1371/journal.pbio.0060245. [DOI] [PMC free article] [PubMed] [Google Scholar]
Howe ES, Murphy SP, Bass HW. 2013. Three-dimensional acrylamide fluorescence in situ hybridization for plant cells. Methods Mol Biol. 990:53–66. 10.1007/978-1-62703-333-6_6. [DOI] [PubMed] [Google Scholar]
Hufford MB et al. 2021. De Novo assembly, annotation, and comparative analysis of 26 diverse maize genomes. Science. 373:655–662. 10.1126/science.abg5289. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kalluchi A, Harris HL, Reznicek TE, Rowley MJ. 2023. Considerations and caveats for analyzing chromatin compartments. Front Mol Biosci. 10:1168562. 10.3389/fmolb.2023.1168562. [DOI] [PMC free article] [PubMed] [Google Scholar]
Knight PA, Ruiz D. 2013. A fast algorithm for matrix balancing. IMA J Numerical Anal. 33:1029–1047. 10.1093/imanum/drs019. [DOI] [Google Scholar]
Kramer NE et al. 2022. Plotgardener: cultivating precise multi-panel figures in R. Bioinformatics. 38:2042–2045. 10.1093/bioinformatics/btac057. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kruse K, Hug CB, Vaquerizas JM. 2020. FAN-C: a feature-rich framework for the analysis and visualisation of chromosome conformation capture data. Genome Biol. 21:303. 10.1186/s13059-020-02215-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lee H, Seo PJ. 2023. Accessible gene borders establish a core structural unit for chromatin architecture in Arabidopsis. Nucleic Acids Res. 51:10261–10277. 10.1093/nar/gkad710. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lee T-J et al. 2010. Arabidopsis Thaliana chromosome 4 replicates in two phases that correlate with chromatin state. PLoS Genet. 6:e1000982. 10.1371/journal.pgen.1000982. [DOI] [PMC free article] [PubMed] [Google Scholar]
Li H et al. ; 1000 Genome Project Data Processing Subgroup . 2009. The sequence alignment/map format and SAMtools. Bioinformatics. 25:2078–2079. 10.1093/bioinformatics/btp352. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lieberman-Aiden E et al. 2009. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science. 326:289–293. 10.1126/science.1181369. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liu C et al. 2016. Genome-wide analysis of chromatin packing in Arabidopsis Thaliana at single-gene resolution. Genome Res. 26:1057–1068. 10.1101/gr.204032.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liu Y et al. 2018. PCSD: a plant chromatin state database. Nucleic Acids Res. 46:D1157–D1167. 10.1093/nar/gkx919. [DOI] [PMC free article] [PubMed] [Google Scholar]
Makarevitch I et al. 2013. Genomic distribution of maize facultative heterochromatin marked by trimethylation of H3K27. Plant Cell. 25:780–793. 10.1105/tpc.112.106427. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mickelson-Young L et al. 2022. A protocol for genome-wide analysis of DNA replication timing in intact root tips. Methods Mol Biol. 2382:29–72. 10.1007/978-1-0716-1744-1_3. [DOI] [PubMed] [Google Scholar]
Montgomery SA et al. 2020. Chromatin organization in early land plants reveals an ancestral association between H3K27me3, transposons, and constitutive heterochromatin. Curr Biol. 30:573–88.e7. 10.1016/j.cub.2019.12.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nannas NJ, Dawe RK. 2015. Genetic and genomic toolbox of Zea Mays. Genetics. 199:655–669. 10.1534/genetics.114.165183. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nichols MH, Corces VG. 2021. Principles of 3D compartmentalization of the human genome. Cell Rep. 35:109330. 10.1016/j.celrep.2021.109330. [DOI] [PMC free article] [PubMed] [Google Scholar]
O’Keefe RT, Henderson SC, Spector DL. 1992. Dynamic organization of DNA replication in mammalian cell nuclei: spatially and temporally defined replication of chromosome-specific alpha-satellite DNA sequences. J Cell Biol. 116:1095–1110. 10.1083/jcb.116.5.1095. [DOI] [PMC free article] [PubMed] [Google Scholar]
Panning MM, Gilbert DM. 2005. Spatio-temporal organization of DNA replication in murine embryonic stem, primary, and immortalized cells. J Cell Biochem. 95:74–82. 10.1002/jcb.20395. [DOI] [PubMed] [Google Scholar]
Patmanidi AL et al. 2017. Concise review: geminin-a tale of two tails: DNA replication and transcriptional/epigenetic regulation in stem cells. Stem Cells. 35:299–310. 10.1002/stem.2529. [DOI] [PubMed] [Google Scholar]
Percival DB, Walden AT. 2013. Cambridge Series in statistical and probabilistic mathematics: wavelet methods for time series analysis series number 4. Cambridge University Press. 10.1017/cbo9780511841040. [DOI] [Google Scholar]
Pope BD, Hiratani I, Gilbert DM. 2010. Domain-wide regulation of DNA replication timing during mammalian development. Chromosome Res. 18:127–136. 10.1007/s10577-009-9100-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Prioleau M-N, MacAlpine DM. 2016. DNA replication origins-where do we begin? Genes Dev. 30:1683–1697. 10.1101/gad.285114.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rao SSP et al. 2014. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 159:1665–1680. 10.1016/j.cell.2014.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rhind N, Gilbert DM. 2013. DNA replication timing. Cold Spring Harb Perspect Biol. 5:a010132. 10.1101/cshperspect.a010132. [DOI] [PMC free article] [PubMed] [Google Scholar]
Richards L, Das S, Nordman JT. 2022. Rif1-dependent control of replication timing. Genes (Basel). 13:550. 10.3390/genes13030550. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rivera-Mulia JC et al. 2015. Dynamic changes in replication timing and gene expression during lineage specification of human pluripotent stem cells. Genome Res. 25:1091–1103. 10.1101/gr.187989.114. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rivera-Mulia JC, Gilbert DM. 2016. Replication timing and transcriptional control: beyond cause and effect-part III. Curr Opin Cell Biol. 40:168–178. 10.1016/j.ceb.2016.03.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
Roudier F et al. 2011. Integrative epigenomic mapping defines four main chromatin states in Arabidopsis. EMBO J. 30:1928–1938. 10.1038/emboj.2011.103. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ryba T et al. 2010. Evolutionarily conserved replication timing profiles predict long-range chromatin interactions and distinguish closely related cell types. Genome Res. 20:761–770. 10.1101/gr.099655.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
Salic A, Mitchison TJ. 2008. A chemical method for fast and sensitive detection of DNA synthesis in vivo. Proc Natl Acad Sci U S A. 105:2415–2420. 10.1073/pnas.0712168105. [DOI] [PMC free article] [PubMed] [Google Scholar]
Samaniego R, de la Torre C, de la Espina SMD. 2002. Dynamics of replication foci and nuclear matrix during S phase in Allium Cepa L. Cells. Planta. 215:195–204. 10.1007/s00425-002-0741-5. [DOI] [PubMed] [Google Scholar]
Sanchez MdlP, Costas C, Sequeira-Mendes J, Gutierrez C. 2012. Regulating DNA replication in plants. Cold Spring Harb Perspect Biol. 4:a010140. 10.1101/cshperspect.a010140. [DOI] [PMC free article] [PubMed] [Google Scholar]
Savadel SD et al. 2021. The native cistrome and sequence motif families of the maize ear. PLoS Genet. 17:e1009689. 10.1371/journal.pgen.1009689. [DOI] [PMC free article] [PubMed] [Google Scholar]
Savadel SD, Bass HW. 2017. Take a Look at plant DNA replication: recent insights and new questions. Plant Signal Behav. 12:e1311437. 10.1080/15592324.2017.1311437. [DOI] [PMC free article] [PubMed] [Google Scholar]
Schnable PS et al. 2009. The B73 maize genome: complexity, diversity, and dynamics. Science. 326:1112–1115. 10.1126/science.1178534. [DOI] [PubMed] [Google Scholar]
Sequeira-Mendes J et al. 2014. The functional topography of the Arabidopsis genome is organized in a reduced number of linear motifs of chromatin states. Plant Cell. 26:2351–2366. 10.1105/tpc.114.124578. [DOI] [PMC free article] [PubMed] [Google Scholar]
Servant N et al. 2015. HiC-pro: an optimized and flexible pipeline for hi-C data processing. Genome Biol. 16:259. 10.1186/s13059-015-0831-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sparvoli E, Levi M, Rossi E. 1994. Replicon clusters may form structurally stable complexes of chromatin and chromosomes. J Cell Sci. 107(Pt 11):3097–3103. 10.1242/jcs.107.11.3097. [DOI] [PubMed] [Google Scholar]
The R Project for Statistical Computing . n.d.. Accessed December 3, 2024. https://www.r-project.org/.
Turpin ZM et al. 2018. Chromatin structure profile data from DNS-seq: differential nuclease sensitivity mapping of four reference tissues of B73 maize (L). Data Brief. 20:358–363. 10.1016/j.dib.2018.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vouzas AE, Gilbert DM. 2023. Replication timing and transcriptional control: beyond cause and effect—part IV. Curr Opin Genet Dev. 79:102031. 10.1016/j.gde.2023.102031. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wang C et al. 2015. Genome-wide analysis of local chromatin packing in Arabidopsis Thaliana. Genome Res. 25:246–256. 10.1101/gr.170332.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wang Y et al. 2021. SPIN reveals genome-wide landscape of nuclear compartmentalization. Genome Biol. 22:36. 10.1186/s13059-020-02253-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wear EE et al. 2017. Genomic analysis of the DNA replication timing program during mitotic S phase in maize (zea mays) root tips. Plant Cell. 29:2126–2149. 10.1105/tpc.17.00037. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wear EE et al. 2020. Comparing DNA replication programs reveals large timing shifts at centromeres of endocycling cells in maize roots. PLoS Genet. 16:e1008623. 10.1371/journal.pgen.1008623. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wen Z et al. 2022. Extensive chromatin structure-function associations revealed by accurate 3D compartmentalization characterization. Front Cell Dev Biol. 10:845118. 10.3389/fcell.2022.845118. [DOI] [PMC free article] [PubMed] [Google Scholar]
West PT et al. 2014. Genomic distribution of H3K9me2 and DNA methylation in a maize genome. PLoS One. 9:e105267. 10.1371/journal.pone.0105267. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wheeler E et al. 2025. A comparison of genomic methods to assess DNA replication timing. Sci Rep. 15:17761. 10.1038/s41598-025-02699-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wickham H. 2016. Ggplot2. PDF. Use R!. 2nd ed. Springer International Publishing. [Google Scholar]
Xu L, Jiang H. 2020. Writing and Reading histone H3 lysine 9 methylation in Arabidopsis. Front Plant Sci. 11:452. 10.3389/fpls.2020.00452. [DOI] [PMC free article] [PubMed] [Google Scholar]
Yang T et al. 2017. HiCRep: assessing the reproducibility of hi-C data using a stratum-adjusted correlation coefficient. Genome Res. 27:1939–1949. 10.1101/gr.220640.117. [DOI] [PMC free article] [PubMed] [Google Scholar]
Yardımcı GG, Noble WS. 2017. Software tools for visualizing hi-C data. Genome Biol. 18:26. 10.1186/s13059-017-1161-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhang X et al. 2007. Whole-genome analysis of histone H3 lysine 27 trimethylation in Arabidopsis. PLoS Biol. 5:e129. 10.1371/journal.pbio.0050129. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhao PA, Sasaki T, Gilbert DM. 2020. High-resolution repli-seq defines the temporal choreography of initiation, elongation and termination of replication in mammalian cells. Genome Biol. 21:76. 10.1186/s13059-020-01983-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zink D. 2006. The temporal program of DNA replication: new insights into old questions. Chromosoma. 115:273–287. 10.1007/s00412-006-0062-8. [DOI] [PubMed] [Google Scholar]
Zynda GJ et al. 2017. Repliscan: a tool for classifying replication timing regions. BMC Bioinformatics. 18:362. 10.1186/s12859-017-1774-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

koag042_Supplementary_Data

koag042_supplementary_data.zip^{(20.2MB, zip)}

Data Availability Statement

https://github.com/Sarachaudry/HiC-contacts-with-RT-annotations

https://github.com/Sarachaudry/Correlation_Hexbin_plots

https://github.com/Sarachaudry/virtual-4C-interactions-with-Sushi-in-R

https://github.com/Sarachaudry/ChromHMM-Chromatin-States-Analysis

https://github.com/Sarachaudry/R-script-for-scatter-plot

GEO accession GSE287128: (reviewer token azojyygqvfyhjcl)

[koag042-B1] Arifulin EA et al. 2024. Global nuclear reorganization during heterochromatin replication in the giant-genome plant Nigella Damascena L. Plant J. 120:1508–1521. 10.1111/tpj.17063. [DOI] [PubMed] [Google Scholar]

[koag042-B2] Ay F, Noble WS. 2015. Analysis methods for studying the 3D architecture of the genome. Genome Biol. 16:183. 10.1186/s13059-015-0745-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B3] Baker A et al. 2012. Replication fork polarity gradients revealed by megabase-sized U-shaped replication timing domains in human cell lines. PLoS Comput Biol. 8:e1002443. 10.1371/journal.pcbi.1002443. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B4] Bass HW et al. 2014. A maize root tip system to study DNA replication programmes in somatic and endocycling nuclei during plant development. J Exp Bot. 65:2747–2756. 10.1093/jxb/ert470. [DOI] [PubMed] [Google Scholar]

[koag042-B5] Bass HW et al. 2015. Defining multiple, distinct, and shared spatiotemporal patterns of DNA replication and endoreduplication from 3D image analysis of developing maize (Zea Mays L.) root tip nuclei. Plant Mol Biol. 89:339–351. 10.1007/s11103-015-0364-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B6] Bass HW, Marshall WF, Sedat JW, Agard DA, Cande WZ. 1997. Telomeres cluster de novo before the initiation of Synapsis: a three-dimensional spatial analysis of telomere positions before and during meiotic prophase. J Cell Biol. 137:5–18. 10.1083/jcb.137.1.5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B7] Batut P, Gingeras TR. 2013. RAMPAGE: promoter activity profiling by paired-End sequencing of 5′-complete cDNAs. Curr Protoc Mol Biol. 104:Unit 25B.11. 10.1002/0471142727.mb25b11s104. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B8] Cahn J, et al. MaizeCODE reveals bi-directionally expressed enhancers that harbor molecular signatures of maize domestication. Nat Commun. 2024:15:10854. 10.1038/s41467-024-55195-w. Erratum in: Nat Commun. 2025:16:6790. https://doi.org/10.1038/s41467-025-62179-x. PMID: 39738013; PMCID: PMC11685423. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B9] Concia L et al. 2018. Genome-wide analysis of the Arabidopsis replication timing program. Plant Physiol. 176:2166–2185. 10.1104/pp.17.01537. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B10] Crane E et al. 2015. Condensin-driven remodelling of X chromosome topology during dosage compensation. Nature. 523:240–244. 10.1038/nature14450. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B11] Desprat R et al. 2009. Predictable dynamic program of timing of DNA replication in human cells. Genome Res. 19:2288–2299. 10.1101/gr.094060.109. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B12] Dileep V et al. 2015. Topologically associating domains and their long-range contacts are established during early G1 coincident with the establishment of the replication-timing program. Genome Res. 25:1104–1113. 10.1101/gr.183699.114. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B13] Dimitrova DS, Berezney R. 2002. The spatio-temporal organization of DNA replication sites is identical in primary, immortalized and transformed mammalian cells. J Cell Sci. 115(Pt 21):4037–4051. 10.1242/jcs.00087. [DOI] [PubMed] [Google Scholar]

[koag042-B14] Dong P et al. 2017. 3D chromatin architecture of large plant genomes determined by local A/B compartments. Mol Plant. 10:1497–1509. 10.1016/j.molp.2017.11.005. [DOI] [PubMed] [Google Scholar]

[koag042-B15] Dong P, Zhong S. 2020. Characterization of plant 3D chromatin architecture, in situ hi-C library preparation, and data analysis. Methods Mol Biol. 2093:147–157. 10.1007/978-1-0716-0179-2_11. [DOI] [PubMed] [Google Scholar]

[koag042-B16] Durand NC et al. 2016. Juicebox provides a visualization system for hi-C contact maps with unlimited zoom. Cell Syst. 3:99–101. 10.1016/j.cels.2015.07.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B17] Ernst J et al. 2011. Mapping and analysis of chromatin state dynamics in nine human cell types. Nature. 473:43–49. 10.1038/nature09906. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B18] Ernst J, Kellis M. 2010. Discovery and characterization of chromatin states for systematic annotation of the human genome. Nat Biotechnol. 28:817–825. 10.1038/nbt.1662. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B19] Freeling M, Walbot V. editors. 1996. The maize handbook. Springer Lab Manuals. Springer. [Google Scholar]

[koag042-B20] Gent JI, Wang K, Jiang J, Dawe RK. 2015. Stable patterns of CENH3 occupancy through maize lineages containing genetically similar centromeres. Genetics. 200:1105–1116. 10.1534/genetics.115.177360. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B21] Golicz AA, Bhalla PL, Edwards D, Singh MB. 2020. Rice 3D chromatin structure correlates with sequence variation and meiotic recombination rate. Commun Biol. 3:235. 10.1038/s42003-020-0932-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B22] Grimm EC. 1987. CONISS: a FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Comput Geosci. 13:13–35. 10.1016/0098-3004(87)90022-7. [DOI] [Google Scholar]

[koag042-B23] Heng L. 2013. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM [preprint]. arXiv 3997. 10.48550/arXiv.1303.3997. [DOI]

[koag042-B24] Hiratani I et al. 2008. Global reorganization of replication domains during embryonic stem cell differentiation. PLoS Biol. 6:e245. 10.1371/journal.pbio.0060245. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B25] Howe ES, Murphy SP, Bass HW. 2013. Three-dimensional acrylamide fluorescence in situ hybridization for plant cells. Methods Mol Biol. 990:53–66. 10.1007/978-1-62703-333-6_6. [DOI] [PubMed] [Google Scholar]

[koag042-B26] Hufford MB et al. 2021. De Novo assembly, annotation, and comparative analysis of 26 diverse maize genomes. Science. 373:655–662. 10.1126/science.abg5289. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B27] Kalluchi A, Harris HL, Reznicek TE, Rowley MJ. 2023. Considerations and caveats for analyzing chromatin compartments. Front Mol Biosci. 10:1168562. 10.3389/fmolb.2023.1168562. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B28] Knight PA, Ruiz D. 2013. A fast algorithm for matrix balancing. IMA J Numerical Anal. 33:1029–1047. 10.1093/imanum/drs019. [DOI] [Google Scholar]

[koag042-B29] Kramer NE et al. 2022. Plotgardener: cultivating precise multi-panel figures in R. Bioinformatics. 38:2042–2045. 10.1093/bioinformatics/btac057. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B30] Kruse K, Hug CB, Vaquerizas JM. 2020. FAN-C: a feature-rich framework for the analysis and visualisation of chromosome conformation capture data. Genome Biol. 21:303. 10.1186/s13059-020-02215-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B31] Lee H, Seo PJ. 2023. Accessible gene borders establish a core structural unit for chromatin architecture in Arabidopsis. Nucleic Acids Res. 51:10261–10277. 10.1093/nar/gkad710. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B32] Lee T-J et al. 2010. Arabidopsis Thaliana chromosome 4 replicates in two phases that correlate with chromatin state. PLoS Genet. 6:e1000982. 10.1371/journal.pgen.1000982. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B33] Li H et al. ; 1000 Genome Project Data Processing Subgroup . 2009. The sequence alignment/map format and SAMtools. Bioinformatics. 25:2078–2079. 10.1093/bioinformatics/btp352. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B34] Lieberman-Aiden E et al. 2009. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science. 326:289–293. 10.1126/science.1181369. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B35] Liu C et al. 2016. Genome-wide analysis of chromatin packing in Arabidopsis Thaliana at single-gene resolution. Genome Res. 26:1057–1068. 10.1101/gr.204032.116. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B36] Liu Y et al. 2018. PCSD: a plant chromatin state database. Nucleic Acids Res. 46:D1157–D1167. 10.1093/nar/gkx919. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B37] Makarevitch I et al. 2013. Genomic distribution of maize facultative heterochromatin marked by trimethylation of H3K27. Plant Cell. 25:780–793. 10.1105/tpc.112.106427. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B38] Mickelson-Young L et al. 2022. A protocol for genome-wide analysis of DNA replication timing in intact root tips. Methods Mol Biol. 2382:29–72. 10.1007/978-1-0716-1744-1_3. [DOI] [PubMed] [Google Scholar]

[koag042-B39] Montgomery SA et al. 2020. Chromatin organization in early land plants reveals an ancestral association between H3K27me3, transposons, and constitutive heterochromatin. Curr Biol. 30:573–88.e7. 10.1016/j.cub.2019.12.015. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B40] Nannas NJ, Dawe RK. 2015. Genetic and genomic toolbox of Zea Mays. Genetics. 199:655–669. 10.1534/genetics.114.165183. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B41] Nichols MH, Corces VG. 2021. Principles of 3D compartmentalization of the human genome. Cell Rep. 35:109330. 10.1016/j.celrep.2021.109330. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B42] O’Keefe RT, Henderson SC, Spector DL. 1992. Dynamic organization of DNA replication in mammalian cell nuclei: spatially and temporally defined replication of chromosome-specific alpha-satellite DNA sequences. J Cell Biol. 116:1095–1110. 10.1083/jcb.116.5.1095. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B43] Panning MM, Gilbert DM. 2005. Spatio-temporal organization of DNA replication in murine embryonic stem, primary, and immortalized cells. J Cell Biochem. 95:74–82. 10.1002/jcb.20395. [DOI] [PubMed] [Google Scholar]

[koag042-B44] Patmanidi AL et al. 2017. Concise review: geminin-a tale of two tails: DNA replication and transcriptional/epigenetic regulation in stem cells. Stem Cells. 35:299–310. 10.1002/stem.2529. [DOI] [PubMed] [Google Scholar]

[koag042-B45] Percival DB, Walden AT. 2013. Cambridge Series in statistical and probabilistic mathematics: wavelet methods for time series analysis series number 4. Cambridge University Press. 10.1017/cbo9780511841040. [DOI] [Google Scholar]

[koag042-B46] Pope BD, Hiratani I, Gilbert DM. 2010. Domain-wide regulation of DNA replication timing during mammalian development. Chromosome Res. 18:127–136. 10.1007/s10577-009-9100-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B47] Prioleau M-N, MacAlpine DM. 2016. DNA replication origins-where do we begin? Genes Dev. 30:1683–1697. 10.1101/gad.285114.116. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B48] Rao SSP et al. 2014. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 159:1665–1680. 10.1016/j.cell.2014.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B49] Rhind N, Gilbert DM. 2013. DNA replication timing. Cold Spring Harb Perspect Biol. 5:a010132. 10.1101/cshperspect.a010132. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B50] Richards L, Das S, Nordman JT. 2022. Rif1-dependent control of replication timing. Genes (Basel). 13:550. 10.3390/genes13030550. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B51] Rivera-Mulia JC et al. 2015. Dynamic changes in replication timing and gene expression during lineage specification of human pluripotent stem cells. Genome Res. 25:1091–1103. 10.1101/gr.187989.114. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B52] Rivera-Mulia JC, Gilbert DM. 2016. Replication timing and transcriptional control: beyond cause and effect-part III. Curr Opin Cell Biol. 40:168–178. 10.1016/j.ceb.2016.03.022. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B53] Roudier F et al. 2011. Integrative epigenomic mapping defines four main chromatin states in Arabidopsis. EMBO J. 30:1928–1938. 10.1038/emboj.2011.103. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B54] Ryba T et al. 2010. Evolutionarily conserved replication timing profiles predict long-range chromatin interactions and distinguish closely related cell types. Genome Res. 20:761–770. 10.1101/gr.099655.109. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B55] Salic A, Mitchison TJ. 2008. A chemical method for fast and sensitive detection of DNA synthesis in vivo. Proc Natl Acad Sci U S A. 105:2415–2420. 10.1073/pnas.0712168105. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B56] Samaniego R, de la Torre C, de la Espina SMD. 2002. Dynamics of replication foci and nuclear matrix during S phase in Allium Cepa L. Cells. Planta. 215:195–204. 10.1007/s00425-002-0741-5. [DOI] [PubMed] [Google Scholar]

[koag042-B57] Sanchez MdlP, Costas C, Sequeira-Mendes J, Gutierrez C. 2012. Regulating DNA replication in plants. Cold Spring Harb Perspect Biol. 4:a010140. 10.1101/cshperspect.a010140. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B58] Savadel SD et al. 2021. The native cistrome and sequence motif families of the maize ear. PLoS Genet. 17:e1009689. 10.1371/journal.pgen.1009689. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B59] Savadel SD, Bass HW. 2017. Take a Look at plant DNA replication: recent insights and new questions. Plant Signal Behav. 12:e1311437. 10.1080/15592324.2017.1311437. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B60] Schnable PS et al. 2009. The B73 maize genome: complexity, diversity, and dynamics. Science. 326:1112–1115. 10.1126/science.1178534. [DOI] [PubMed] [Google Scholar]

[koag042-B61] Sequeira-Mendes J et al. 2014. The functional topography of the Arabidopsis genome is organized in a reduced number of linear motifs of chromatin states. Plant Cell. 26:2351–2366. 10.1105/tpc.114.124578. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B62] Servant N et al. 2015. HiC-pro: an optimized and flexible pipeline for hi-C data processing. Genome Biol. 16:259. 10.1186/s13059-015-0831-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B63] Sparvoli E, Levi M, Rossi E. 1994. Replicon clusters may form structurally stable complexes of chromatin and chromosomes. J Cell Sci. 107(Pt 11):3097–3103. 10.1242/jcs.107.11.3097. [DOI] [PubMed] [Google Scholar]

[koag042-B64] The R Project for Statistical Computing . n.d.. Accessed December 3, 2024. https://www.r-project.org/.

[koag042-B65] Turpin ZM et al. 2018. Chromatin structure profile data from DNS-seq: differential nuclease sensitivity mapping of four reference tissues of B73 maize (L). Data Brief. 20:358–363. 10.1016/j.dib.2018.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B66] Vouzas AE, Gilbert DM. 2023. Replication timing and transcriptional control: beyond cause and effect—part IV. Curr Opin Genet Dev. 79:102031. 10.1016/j.gde.2023.102031. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B67] Wang C et al. 2015. Genome-wide analysis of local chromatin packing in Arabidopsis Thaliana. Genome Res. 25:246–256. 10.1101/gr.170332.113. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B68] Wang Y et al. 2021. SPIN reveals genome-wide landscape of nuclear compartmentalization. Genome Biol. 22:36. 10.1186/s13059-020-02253-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B69] Wear EE et al. 2017. Genomic analysis of the DNA replication timing program during mitotic S phase in maize (zea mays) root tips. Plant Cell. 29:2126–2149. 10.1105/tpc.17.00037. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B70] Wear EE et al. 2020. Comparing DNA replication programs reveals large timing shifts at centromeres of endocycling cells in maize roots. PLoS Genet. 16:e1008623. 10.1371/journal.pgen.1008623. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B71] Wen Z et al. 2022. Extensive chromatin structure-function associations revealed by accurate 3D compartmentalization characterization. Front Cell Dev Biol. 10:845118. 10.3389/fcell.2022.845118. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B72] West PT et al. 2014. Genomic distribution of H3K9me2 and DNA methylation in a maize genome. PLoS One. 9:e105267. 10.1371/journal.pone.0105267. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B73] Wheeler E et al. 2025. A comparison of genomic methods to assess DNA replication timing. Sci Rep. 15:17761. 10.1038/s41598-025-02699-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B74] Wickham H. 2016. Ggplot2. PDF. Use R!. 2nd ed. Springer International Publishing. [Google Scholar]

[koag042-B75] Xu L, Jiang H. 2020. Writing and Reading histone H3 lysine 9 methylation in Arabidopsis. Front Plant Sci. 11:452. 10.3389/fpls.2020.00452. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B76] Yang T et al. 2017. HiCRep: assessing the reproducibility of hi-C data using a stratum-adjusted correlation coefficient. Genome Res. 27:1939–1949. 10.1101/gr.220640.117. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B77] Yardımcı GG, Noble WS. 2017. Software tools for visualizing hi-C data. Genome Biol. 18:26. 10.1186/s13059-017-1161-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B78] Zhang X et al. 2007. Whole-genome analysis of histone H3 lysine 27 trimethylation in Arabidopsis. PLoS Biol. 5:e129. 10.1371/journal.pbio.0050129. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B79] Zhao PA, Sasaki T, Gilbert DM. 2020. High-resolution repli-seq defines the temporal choreography of initiation, elongation and termination of replication in mammalian cells. Genome Biol. 21:76. 10.1186/s13059-020-01983-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[koag042-B80] Zink D. 2006. The temporal program of DNA replication: new insights into old questions. Chromosoma. 115:273–287. 10.1007/s00412-006-0062-8. [DOI] [PubMed] [Google Scholar]

[koag042-B81] Zynda GJ et al. 2017. Repliscan: a tool for classifying replication timing regions. BMC Bioinformatics. 18:362. 10.1186/s12859-017-1774-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Replication timing uncovers a two-compartment nuclear architecture of interphase euchromatin

H S Akram

E E Wear

L Mickelson-Young

Z M Turpin

L Hanley-Bowdoin

W F Thompson

L Concia

H W Bass

Abstract

Introduction

Results

Repli-seq scheme and mini-domain chromatin fiber RT model

Figure 1 .

Hi-C data reveal covariance and correlation with early and middle RT in maize euchromatin

Figure 2 .

Figure 3 .

Hi-C contacts are enriched within RT classes

Figure 4 .

Hi-C-based chromatin architecture and Repli-seq align with epigenomic and transcriptional features

Figure 5 .

3D cytological evidence for the spatial bifurcation of euchromatin into early-S and middle-S chromatin regions

Figure 6 .

Figure 7 .

Discussion

Materials and methods

.

Plant materials

Repli-seq realigned to B73v5

Root tip and earshoot nuclei for Hi-C

Preparation and sequencing of hi-C libraries from maize root tip and ear shoot

Nuclei isolation for MMOA-seq

MMOA-seq library preparation and sequencing

HiC-pro analysis of Hi-C data

Insulation score and eigenvector analysis

Chromatin state analysis using a hidden Markov model

Statistical analysis

Root tip nuclei for 3D-FISH

Probe labeling for 3D-FISH

Fluorescent in situ hybridization

Deltavision microscopy and image analysis

Supplementary Material

Acknowledgments

Contributor Information

Author contributions

Supplementary material

Funding

Data availability

Material and correspondence

Dive Curated Terms

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases