Abstract
Background
The genome of rye, Secale cereale, is distinguished by large repetitive regions including subtelomeric heterochromatin and retrotransposon-dominant centromeres, which contrast with the satellite-repeat-based centromeres in most characterized plant genome assemblies. This study aims to decode the architecture and evolution of these elusive regions through high-resolution genome assembly, with a focus on centromere dynamics and chromatin regulation.
Results
Using PacBio HiFi and Nanopore sequencing, we generate a chromosome-scale assembly encompassing three complete centromeres and resolving subtelomeric heterochromatin. We identify terminal tandem repeat arrays as key determinants in establishing specialized chromatin environments linked to retrotransposon deposition. Notably, rye centromeres exhibit an unconventional epigenetic signature depleted of conventional activation and repression marks but displaying unique DNA hypomethylation patterns. This retrotransposon-enriched landscape promotes both the integration of young LTR retrotransposons and the recruitment of CENH3. Cross-species CENH3 ChIP-seq analyses reveal that Cereba retrotransposons are associated with enhanced CENH3 loading in cultivated and wild rye lineages, particularly through their conserved protease and integrase domains, suggesting a potential positive feedback loop for centromere evolution.
Conclusions
Our findings establish retrotransposons as autonomous organizers of centromere chromatin and identity in rye, challenging the paradigm of satellite-dependent centromere specification. The dual role of retrotransposons in maintaining CENH3 recruitment while facilitating genomic innovation provides a mechanistic basis for centromere plasticity. This work advances functional genomics of Triticeae crops and opens new avenues for centromere engineering to manipulate meiotic stability and chromosome transmission in crop breeding.
Similar content being viewed by others
Background
Centromeres are essential chromosomal elements ensuring faithful chromosome segregation during cell division [1]. Their function relies on CENH3-containing nucleosomes, where the centromere-specific histone H3 variant CENH3 epigenetically determines centromere identity by binding repetitive DNA sequences [1,2,3]. Despite this conserved functional role, centromeric DNA sequences diverge rapidly across species—a paradox central to eukaryotic genome evolution [4, 5]. In species with well-characterized centromeres—including humans, Arabidopsis, rice, and Forsythia suspensa—centromeres typically comprise species-specific tandem repeat arrays [6,7,8,9]. Advances in telomere-to-telomere assemblies have uncovered “hybrid” centromeres in maize (Zea mays), cotton (Gossypium spp), and watermelon (Citrullus lanatus), where tandem repeats intermingle with retrotransposons [10,11,12,13,14]. However, retrotransposon-dominated centromeres remain poorly due to technological limitations in resolving repetitive architectures and deciphering their epigenetic regulation. This knowledge gap impedes our understanding of how sequence plasticity and functional conservation coexist in these critical genomic regions.
Similar to centromeres, subtelomeres represent another genomic frontier where repetitive elements orchestrate chromosomal dynamics. These regions maintain genome integrity through mechanisms involving telomere homeostasis and heterochromatin formation [15,16,17,18,19]. Rye (Secale cereale L.), a diploid Triticeae species with a 7.9 Gb genome [20, 21], provides a unique system to investigate retrotransposon-dominated centromeres. Its subtelomeres harbor massive heterochromatic blocks composed of species-specific tandem repeats (pSc200, pSc250) occupying 8–12% of the genome [22,23,24], while its centromeres were dominated by long terminal repeat (LTR) retrotransposons rather than tandem arrays [25,26,27]. This striking structural contrast between subtelomeric and centromeric domains establishes rye as a natural experimental system to dissect how divergent repeat classes jointly orchestrate chromosomal organization. Despite rye’s agricultural significance in marginal environments (e.g., tolerance to poor soil fertility, salinity, and alkalinity) [28,29,30], its genome assembly has been hindered by extreme repetitiveness. Previous assemblies fragmented subtelomeric tandem arrays and centromeres, obscuring critical insights into their organization [20, 28, 31]. Additionally, the 1RS-1BL translocation in wheat breeding also highlights the practical importance of rye centromere stability and our limited understanding of its sequence basis.
Centromere presents a conserved functional paradox: despite essential roles in chromosome segregation retained throughout eukaryotes, centromeric DNA sequences exhibit rapid evolutionary divergence and structural plasticity [32,33,34]. In Arabidopsis, cycles of retrotransposon proliferation and satellite DNA homogenization drive centromere diversification, while soybean domestication demonstrates centromere repositioning [5]. Maize centromeres reveal retrotransposons influencing CENH3 binding patterns through sequence-specific interactions [35]. These findings demonstrate that repetitive sequences evolve dynamically, with epigenetic regulation (DNA methylation, histone modifications) shaping their functional roles [5, 25, 32, 36]. However, the mechanisms enabling retrotransposons to drive centromere formation and stability in the absence of satellite repeats remain unknown— a problem central to understanding genome organization.
Here, we generated a highly contiguous rye reference genome assembly using PacBio HiFi and Nanopore ultra-long read technologies, achieving complete assembly of three centromeres and subtelomeric heterochromatin domains. Through comprehensive analysis of this assembly, we investigated how tandem repeat arrays in terminal heterochromatin regions shape retrotransposon distribution and evolution. Moreover, we revealed unique epigenetic signatures and sequence characteristics of rye centromeres, which create an environment conducive to young LTR insertion and CENH3 binding. Further analysis across different rye subspecies uncovered a positive feedback mechanism where Cereba retrotransposons promote CENH3 loading through specific modifications in their INT (integrase) and PROT (protease) domains, which in turn facilitates further Cereba accumulation during domestication. Our findings provide novel insights into centromere evolution during rye domestication.
Results
Assembly and annotation of the rye genome
By leveraging 720.8 Gb (~ 92 × coverage) of PacBio HiFi reads with an N50 length of 16.8 kb, 164.8 Gb (~ 21 × coverage) of Oxford Nanopore ultra-long reads with an N50 length of 74 kb, and 506.5 Gb (~ 64.6 × coverage) of Illumina paired-end reads, we generated a high-quality chromosome-scale assembly of the Weining rye genome (Weining_v2.0), achieving complete assembly of three centromeres, terminal repeat arrays, and rDNA regions. The assembly strategy integrated a phased multi-platform approach. Initial contig assemblies generated with Hifiasm [37] (N50 = 70.26 Mb) and Verkko [38] (N50 = 48.34 Mb) were scaffolded into chromosome-scale structures using Ragtag [39], Chromap-Yahs [40, 41], and Juicer with 3D-DNA [42]. Post-scaffolding gap resolution was systematically performed by the following: (1) cross-assembly alignment between Hifiasm and Verkko scaffolds to identify consensus gaps; (2) targeted closure of unresolved gaps using TGS-GapCloser [43] with hybrid Oxford Nanopore ultra-long reads and PacBio HiFi data to leverage both read length and accuracy advantages. This hierarchical polishing strategy yielded a final 7.748 Gb assembly, with 99.52% sequences (7.711 Gb) anchored to seven chromosomes and a scaffold N50 of 1.107 Gb (Table 1, Fig. 1). Additionally, compared to previous assemblies, Weining_v2.0 incorporates approximately 460 Mb of newly resolved sequences, primarily in centromeric regions (~ 200 Mb), terminal repeat arrays (~ 150 Mb), and other repetitive elements (~ 110 Mb). Taken together, the integrated approach generated a greatly improved genome assembly that enhances genome continuity, especially in highly repetitive regions of the rye genome.
Genomic features and epigenetic landscapes of the rye genome. Circos plot representing the assembled rye genome (Weining_v2.0). From the outermost to innermost tracks: The outer gray track shows the seven chromosomes of rye (Chr1R-Chr7R) with their physical positions in megabases (Mb), followed by CENH3 ChIP-seq signals in centromeric regions (A), gene density (B), terminal inverted repeat (TIR) density (C), full-length long LTR retrotransposons density (D), and DNA methylation levels in the CG (E), CHG (F), and CHH contexts (G). The intensity of color (from light yellow to dark red/orange) indicates the relative abundance or signal strength of each feature. This high-quality assembly spans 7.748 Gb with 99.52% of sequences anchored to seven chromosomes, providing comprehensive insights into the genomic and epigenetic organization of the rye genome
Multiple metrics were used to assess the assembly quality. The Long Terminal Repeat Assembly Index (LAI) reached 20.02, meeting gold-standard quality criteria [44] (Additional file 1: Table S1). BUSCO analysis identified 98.0% of expected single-copy orthologs [45], exceeding the ness of previous rye assemblies (Additional file 1: Table S2). Merqury analysis showed a Quality Value (QV) score of 64.00, indicating high base-level accuracy [46] (Additional file 1: Table S3). These metrics collectively confirmed the high quality and completeness of the assembly.
Repetitive elements were annotated using homology-based and ab initio prediction methods. TEs occupied 91.23% (7.07 Gb) of the genome, consisting of DNA transposons (4.72%, 365.68 Mb) and retrotransposons (85.30%, 6.61 Gb) (Additional file 1: Table S4). Analysis of repetitive sequences identified 251 Mb of tandem repeats containing 1,128,554 units, predominantly located in proximal telomeric regions (Additional file 2: Fig. S1). Gene prediction was performed by integrating de novo prediction, homology-based annotation, and RNA-seq data [31]. We annotated 51,033 high-confidence protein-coding genes, including 5437 genes not found in previous assemblies. The extensive annotation of both repetitive and genic regions provides a foundation for investigating genome organization and evolution in rye. Together with the improved genome assembly, these annotations establish a high-quality genomic resource for rye research.
Tandem repeat arrays shape LTR retrotransposon distribution and organization in terminal heterochromatin
Rye chromosomes are characterized by distinctive large blocks of subtelomeric heterochromatin, a unique feature among cereal crops where other major cereals, including wheat, barley, oat, and cultivated maize, lack such prominent terminal heterochromatic regions (Fig. 2A). Using our high-quality genome assembly, we precisely characterized three distinct classes of tandem repeats within these terminal heterochromatic blocks, corresponding to the previously reported pSc119.2, pSc200, and pSc250 sequences [47,48,49,50], refining their consensus sequences and providing more accurate unit lengths (Fig. 2B, Additional file 2: Fig. S2A, B, Additional file 1: Table S5). Phylogenetic analysis revealed contrasting evolutionary patterns among these families: pSc119.2 showed clear chromosome-specific clustering, suggesting independent evolution on different chromosomes, while pSc200 and pSc250 displayed no obvious chromosome-specific patterns, indicating frequent sequence exchanges between chromosomes (Additional file 2: Fig. S2C). These distinct evolutionary trajectories suggested different functional constraints among the tandem repeat families.
Tandem repeat arrays shape LTR retrotransposon distribution in rye. A FISH analysis showing the distinct chromosomal distribution of heterochromatic blocks in different cereal species. Representative mitotic metaphase chromosomes from rye, oat, barley, wheat, and maize (from left to right) were stained with DAPI. Note the prominent terminal heterochromatic blocks (bright signals) unique to rye chromosomes (orange arrows indicate subtelomeric heterochromatin). Scale bar = 10 µm. B Length distribution and sequence variation analysis of the pSc250 tandem repeat family. The upper panel shows the length distribution of individual repeat units, while the lower panel displays the sequence variation frequency at each nucleotide position. C Box plot showing the frequency distribution of full-length LTR retrotransposons (fl-LTR) across different genomic regions: subtelomere (Subt), chromosome arms (Arm), pericentromeric (Peri), and centromeric (Cent) regions. Asterisks indicate statistical significance (***P < 0.001). D Sequence similarity analysis of LTR retrotransposons in a representative region of chromosome 7R (7.3–7.8 Mb). The upper panel shows the distribution of sequence identity percentages, while the lower heatmap displays pairwise sequence similarities between LTR elements, with different LTR subfamilies indicated. HOR (Higher-Order Repeat): A hierarchical structure composed of smaller repeat units arranged into larger tandem repeat arrays. E Phylogenetic analysis of tandem repeat sequences. The outer ring is color-coded to indicate different repeat families (pSc119.2, pSc200, and pSc250) and genomic locations (Cent, centromeric; Random, randomly selected regions). F Model illustrating the relationship between tandem repeat arrays and LTR copies in subtelomeric regions, showing the process of array-specific LTR evolution and tandem repeat homogenization
Given the distinct evolutionary trajectories among these tandem repeat families, which reflect potential differences in functional constraints, we further investigated their potential roles in genome organization. We first divided the genome into four distinct regions (centromeres, pericentromeric regions, chromosome arms, and subtelomeric regions containing tandem repeat arrays) and found that LTRs are significantly less abundant in the subtelomeric regions containing tandem repeat arrays compared to other regions, suggesting that these areas may be unfavorable for LTR integration or may actively clear LTR insertions (Fig. 2C). Additionally, we observed that LTRs located within the same tandem repeat array exhibit significantly higher sequence similarity compared to those located in different arrays, further hinting at localized evolutionary dynamics (Fig. 2D, Additional file 2: Fig. S3A, B). Phylogenetic analysis revealed that LTRs located in subtelomeric tandem repeat arrays formed specific clusters distinct from those located on chromosome arms (Fig. 2E), and sequence analysis supported this pattern, showing that terminal region-associated LTRs displayed significantly lower sequence variation compared to their counterparts on chromosome arms (Additional file 2: Fig. S3C, D). Together, these findings suggest that LTRs within subtelomeric tandem repeat arrays may evolve under distinct evolutionary pressures, potentially driven by localized genomic environments or functional constraints associated with these regions.
To understand the mechanism underlying these patterns, we investigated the chromatin landscape of these regions. Despite previous reports showing that these subtelomeric regions appear as condensed heterochromatic blocks through cytological observations[22,23,24], we surprisingly found that heterochromatic marks (H3K9me3 and H3K27me3) were primarily localized to the regions between tandem repeat arrays rather than on the arrays themselves (Additional file 2: Fig. S4A). To quantify potential DNA structure-mediated effects, we analyzed the secondary structure stability of tandem repeats using RNAfold-derived folding energy scores (lower/more negative values indicate higher stability), following the analytical approach described in our previous study [32]. These sequences exhibited strong propensities to form stable non-B DNA structures (Additional file 2: Fig. S4B), which might contribute to their condensed state independent of conventional histone modifications. Moreover, the intensity of heterochromatic modifications showed positive correlation with LTR density in the inter-array regions (Additional file 2: Fig. S4C), suggesting that the unique structural properties of tandem repeat arrays may contribute to both the establishment of heterochromatic domains and the regulation of LTR insertion sites.
This chromatin-level organization was further reflected in distinct methylation patterns of LTRs in subtelomeric regions. Compared to their chromosome arm counterparts, these LTRs showed significantly higher CG methylation levels, increased CHH methylation, similar CHG levels, and slightly lower CHH methylation (Additional file 2: Fig. S4D). Together with the specific distribution of histone modifications, this characteristic epigenetic landscape suggests that tandem repeat arrays help establish and maintain specialized chromatin environments that facilitate LTR integration and evolution, effectively serving as organizational units that shape both the distribution and evolutionary trajectory of LTRs in these terminal regions (Fig. 2F). These findings collectively reveal a previously unrecognized role of tandem repeat arrays in genome organization, where they may create distinct evolutionary compartments through epigenetic mechanisms, thereby influencing both the distribution and evolution of retrotransposons in terminal chromosomal regions.
Retrotransposon-rich centromeres in rye maintain their identity through distinct epigenetic markers
Whereas the centromeres of canonical plant models such as Arabidopsis thaliana and rice are tandem repeat-dominated, and the centromeres of Zea mays contain both tandem repeats and retrotransposons [6,7,8], rye centromeres are strikingly devoid of tandem repeats and instead rely predominantly on retrotransposons [26], providing an excellent model to study how retrotransposon-based centromeres maintain their identity and function. Using previously published CENH3 ChIP-seq data [26], we mapped all seven centromeric regions in our assembly and notably achieved complete sequence assembly for three of these centromeres, validated by CENH3 ChIP-seq profiles showing singular, continuous signal peaks without bifurcation, multimerization, or disruptions (Additional file 1: Table S6). Whole-genome collinearity analysis of all peri-centromeres revealed substantially improved continuity in the current assembly compared to Weining_v2.0 (Fig. 3A, Additional file 2: Fig. S5), with fewer alignment breaks and tighter concordance between contiguous scaffolds and CENH3 signal domains. Analysis of sequence composition in these well-defined centromeric regions revealed that retrotransposons constitute over 95% of the sequences in these regions, with a marked enrichment of Gypsy retrotransposons and a corresponding decrease in Copia elements within the CENH3-enriched zone (Fig. 3B, Additional file 2: Fig. S6). Further supporting this structural divergence, analysis of CENH3 ChIP-seq data using RepeatExplorer corroborated the absence of tandem repeat arrays in these regions. This distinct composition pattern sets the stage for understanding the molecular mechanisms underlying centromere specification and maintenance.
Retrotransposon-rich centromeres maintain their identity through distinct epigenetic signatures. A Comparison of centromere assembly between the previous and current versions of the Weining rye genome. Upper panel shows CENH3 ChIP-seq signals, and the synteny dot plot demonstrates improved contiguity in the centromeric region of chromosome 1R. The color scale indicates sequence similarity percentage. B Distribution of genes and TEs across a representative centromeric region (chr1: 310–350 Mb). From top to bottom: gene distribution (blue bars), CENH3 ChIP-seq signals (red), and relative abundance of different TE families (Gypsy in blue, Copia in purple, other TEs in brown). C Correlation between CENH3 enrichment and Gypsy content, with dots colored by insertion time (from red-young to purple-old). D ChromHMM analysis showing nine distinct chromatin states defined by histone modifications (H3K27me3, H3K9me3, H3K4me3) and R-loop signals. E Genomic features across the centromeric region of chromosome 1R, showing CENH3 signals, Gypsy abundance, Gypsy insertion time, and chromatin states. F DNA methylation profiles (CG, CHG, and CHH contexts) across centromeric regions. G Detailed view of DNA methylation patterns and sequence distributions around LTR integration sites. The CG methylation pattern is shown in two shades of blue (dark blue for methylation level, light blue for sequence distribution), while CHH methylation is shown in two shades of red (dark red for methylation level, light red for sequence distribution). The left y-axis shows DNA methylation percentage, and right y-axis indicates sequence distribution. H Cumulative distribution of CG and CHH methylation levels in different genomic contexts: chromosome arms (Arm), centromeric regions with different LTRs (Cent-diff), and centromeric regions with similar LTRs (Cent-same). P-values indicate statistical significance of the differences
Given this specific distribution of retrotransposon families in centromeric regions, we next examined the characteristics of these centromere-associated retrotransposons. Notably, we found a preferential accumulation of younger retrotransposons in centromeric regions (Additional file 2: Fig. S7A, B), with a strong positive correlation between CENH3 signal intensity and Gypsy content, particularly for recently inserted elements (Fig. 3C, Additional file 2: Fig. S7C, D). Phylogenetic analysis further supported this observation, as Gypsy elements showed distinct clustering patterns between centromeric and non-centromeric regions (Additional file 2: Fig. S7E). While centromere identity is epigenetically maintained rather than sequence-defined [4, 51], we observed that Gypsy INT domains within centromeric regions exhibit significantly lower Ka/Ks ratios than their non-centromeric counterparts (Additional file 2: Fig. S7F), indicating intensified purifying selection on these retrotransposons’ protein-coding functions in genomic niches associated with centromeres. This aligns with evidence that retrotransposons invading functional centromeres require intact replication machinery to persist.
Given the unique composition and organization of these retrotransposon-rich centromeres, we sought to understand how their functional identity is maintained at the epigenetic level. Through comprehensive epigenetic profiling including ChIP-seq for histone modifications H3K9me3 (constitutive heterochromatin/transposon silencing), H3K27me3 (dynamic transcriptional repression), and H3K4me3 (transcriptionally active regions), as well as R-loop profiling (RNA–DNA hybrid-associated genomic instability), followed by chromHMM analysis, we found that these retrotransposon-rich functional centromeres are predominantly characterized by an unmarked chromatin state (State 8), depleted of conventional activation/repression marks (H3K4me3, H3K9me3, and H3K27me3) while maintaining exclusive CENH3 enrichment (Fig. 3D, E, Additional file 2: Fig. S8). The detection of State 7 (R-loop-enriched) within centromeres points to transcription-coupled R-loop formation at retrotransposon loci, although these regions lack the H3K4me3 signature characteristic of canonical transcriptional initiation (Fig. 3D, E). In contrast, pericentromeric regions displayed a more complex epigenetic landscape, primarily composed of State 9 (Weak heterochromatin) and State 7, with occasional State 8 patches. This distinctive chromatin state distribution reveals a specialized regulatory environment specific to retrotransposon-rich centromeres.
The unexpected finding of an unmarked chromatin state in functional centromeres led us to investigate other potential epigenetic mechanisms that might contribute to centromere identity and function. Through whole-genome bisulfite sequencing (WGBS), we discovered that centromeric regions exhibited a DNA methylation signature similar to that reported in wheat [52], characterized by reduced CG methylation and elevated CHH methylation. Unlike wheat, rye centromeres also showed reduced CHG methylation (Fig. 3F). Retrotransposon-rich centromeres therefore exhibit a distinct methylation profile compared with satellite-based centromeres of other plants, which are typically characterized by higher CG methylation [6, 7, 53]. In pericentromeric regions, we observed increasing CG and CHG methylation levels proximal to centromeres, with a gradual decrease as distance from the centromere increased. When examining full-length LTRs (fl-LTRs) in centromeric regions, we found their integration sites displayed higher levels of CHH methylation and lower levels of CG methylation compared to fl-LTRs on chromosome arms (Fig. 3G). Importantly, this distinctive methylation profile remained consistent regardless of estimated insertion time, as evidenced by comparison between recently inserted (cent-same) and more ancient (cent-diff) elements (Fig. 3H). This conserved methylation pattern indicates that the unique epigenetic landscape of centromeric regions is established prior to and maintained independently of LTR integration events.
Together, these findings reveal that rye centromeres are characterized by a distinct combination of sequence composition and epigenetic features that differs markedly from typical satellite-based plant centromeres.
Cereba-mediated CENH3 loading shows domain-specific variation among rye subspecies
Recent studies have characterized centromere evolution in several plant species with satellite repeat-dominated centromeres, including Arabidopsis, rice, and soybean [5, 36, 54]. However, how retrotransposon-rich centromeres evolve during domestication remains largely unknown. To address this question, we investigated centromere changes across different rye subspecies, including wild relatives (S. strictum, S. sylvestre), the putative progenitor (S. vavilovii), and three cultivated ryes (Weining, Lo7, and Imperial). Through graph-based clustering of centromeric reads using RepeatExplorer2 de novo pipelines, we generated subspecies-specific repeat profiles and quantified their divergence through comparative cluster abundance analysis, revealing dynamic architectural reorganization of centromeric repeats (Additional file 2: Fig. S9A). While centromere positions and sizes were largely conserved between cultivated ryes and S. vavilovii, wild rye subspecies showed marked differences. Additionally, S. sylvestre exhibited notably narrower and shifted centromere peaks on chromosomes 2R and 3R, while S. strictum showed almost no enrichment peaks (Fig. 4A). These altered ChIP-seq profiles may be associated with underlying DNA structural rearrangements. Among cultivated varieties, we observed distinct patterns, particularly on chromosome 3R where Imperial and Lo7 rye displayed a characteristic depression in their centromeric peaks (Fig. 4A, S9A). These findings suggest substantial centromere sequence divergence between wild and cultivated ryes.
Cereba-mediated CENH3 loading shows domain-specific variation among rye subspecies. A CENH3 ChIP-seq profiles across centromeric regions of chromosomes 2R and 3R in different rye subspecies and cultivars (Weining, Imperial, Lo7, Vavilov, S. sylvestre, and S. strictum). B Principal component analysis (PCA) of k-mer profiles from CENH3 ChIP-seq data across different rye subspecies. C Hierarchical clustering analysis showing the relationships among rye subspecies based on centromeric sequence composition. D Distribution of different retrotransposon families across rye subspecies. Upper panel shows cluster sizes of different repeat families, and lower panel displays their distribution patterns across subspecies, with Cereba elements highlighted in the pink box. E Heatmap showing the abundance (RPKM) of different retrotransposon families in CENH3 ChIP-seq data across rye subspecies. Color scale indicates RPKM values from low (blue) to high (red). F CENH3 binding patterns across Cereba elements in different rye subspecies. Upper panel shows fold change and statistical significance, middle panel displays CENH3 signal distribution across the Cereba element, and bottom panel indicates the structural domains of Cereba (GAG, PROT, RT, RH, INT, CHDCR)
To quantitatively assess these differences, we performed k-mer analysis of the CENH3 ChIP-seq datasets. Principal component analysis (PCA) and phylogenetic analysis of the k-mer profiles clearly separated wild and cultivated ryes, with the three cultivated subspecies clustering closely together, S. vavilovii as the next closest group, and S. strictum and S. sylvestre showing the greatest divergence (Fig. 4B, C). This clustering pattern aligns with the established evolutionary relationships among rye subspecies [55]. RepeatExplorer analysis revealed that the major source of variation in repeat abundance among subspecies stems from Cereba sequences, a subfamily of Chromoviridae LTR retrotransposons (Fig. 4D). Strikingly, cultivated ryes showed substantially higher CENH3 nucleosome enrichment on Cereba sequences compared to their wild relatives S. strictum and S. sylvestre (Fig. 4E). Further analysis of CENH3 ChIP-seq signals between Cereba insertion regions and random genomic regions in Weining rye revealed significantly stronger CENH3 binding in Cereba-associated loci (Additional file 2: Fig. S9B).
To investigate the potential mechanisms underlying these CENH3 binding differences, we examined CENH3 binding patterns across Cereba elements in different subspecies. While the overall CENH3 binding profiles were similar among subspecies, we identified notable differences in binding intensity within specific regions, particularly the INT and PROT domains of Cereba elements (Fig. 4F). The Cereba elements within centromeres exhibit differing levels of enrichment, with newer insertions showing higher CENH3 occupancy, while older insertions have weaker enrichment, indicating that new insertions match this CENH3 enrichment profile (Additional file 2: Fig. S10). Notably, these domains encode functional retrotransposition modules—prior studies highlight that Gypsy integrase domains can mediate chromatin targeting [56], and recent work in Triticum monococcum suggests similar INT-driven centromere localization via CR motifs binding CENH3 [57]. Sequence analysis revealed that the Cereba of cultivated rye harbor altered nucleosome positioning signals within INT/PROT (Fig. 4F), potentially stabilizing CENH3 deposition. In contrast, the LTR and GAG regions showed relatively uniform CENH3 occupancy across rye subspecies, with no obvious differences detected among them (Fig. 4F). Collectively, our data are consistent with a lineage-specific co-option of INT/PROT domains during rye centromere evolution. We hypothesize that this process may follow a two-step mechanistic model (Fig. 5): INT first targets pre-insertion CENH3 hotspots, and this may synergize with PROT-LTR-mediated post-insertion stabilization to form a self-reinforcing loop driving Cereba’s adaptive co-evolutionary trajectory.
INT/PROT co-option drives lineage-specific centromere evolution. Schematic model depicting the coevolution of tandem repeats and retrotransposon activity in shaping centromeric adaptation during rye domestication. Step 1 (targeting): INT domains selectively integrate into pre-existing CENH3-enriched regions, priming centromeric hotspots; Step 2 (stabilization): PROT-associated LTRs enforce chromatin cohesion, suppressing post-insertion structural erosion. Evolutionary outcome: These steps form a self-reinforcing loop, propelling Cereba retrotransposons along an adaptive trajectory during rye domestication
Discussion
Rye has long been recognized for its superior agronomic traits, particularly its exceptional disease resistance and notable resilience against abiotic stressors such as poor soil fertility, salinity, and alkalinity [28,29,30]. However, its large genome, characterized by prominent heterochromatic blocks and highly repetitive centromeric regions, has presented significant challenges in understanding its genomic architecture. By leveraging third-generation sequencing technologies, including ONT ultra-long reads and PacBio HiFi reads, we have generated a high-quality reference genome that enables unprecedented insights into rye’s genome organization, particularly its centromeres and heterochromatic regions.
A distinctive feature of the rye genome is the distribution pattern of its major tandem repeats (pSc119.2, pSc200, and pSc250), which notably differs from other well-studied plant species such as maize, rice, Arabidopsis, Brachypodium, and soybean [5, 7, 8, 13, 58,59,60,61,62,63]. While these species typically concentrate their tandem repeats in centromeric regions, rye’s major tandem repeats are predominantly located in proximal chromosomal regions. Our analysis reveals that these terminal tandem repeat arrays create unique evolutionary compartments that significantly influence genome organization in two ways. Our analysis reveals that these terminal tandem repeat arrays form evolutionary compartments that may significantly influence genome organization in two ways. Second, they create distinct evolutionary niches, evidenced by the high sequence similarity of LTRs within the same array but divergent patterns between different arrays. This compartmentalization effect on LTR evolution has not been previously reported in other plant genomes and may represent a novel mechanism for maintaining genome stability in regions with high repeat content.
The epigenetic landscape associated with these tandem repeat arrays is particularly noteworthy. Our observation of distinct methylation patterns and histone modifications suggests a complex regulatory mechanism controlling both the stability of the arrays themselves and their influence on nearby chromatin regions. The higher levels of H3K9me3 and H3K27me3 modifications in regions between arrays, rather than within them, indicate that these structures might serve as boundaries for heterochromatin domains, as previous research had suspected [47].
The centromeric regions of rye exhibit several intriguing features that provide new insights into centromere evolution. Consistent with findings in Arabidopsis and wheat [7, 36, 64], we observed that centromeric retrotransposons show more recent insertion times compared to those in other genomic regions. Our study reveals a temporal relationship between LTR insertion and CENH3 loading, suggesting that LTR integration precedes CENH3 nucleosome establishment. This conserved methylation landscape, characterized by a reduction in CG/CHG methylation alongside elevated CHH methylation, has been observed in centromeres of other species [6, 7, 53]. Furthermore, the distinct DNA methylation patterns of centromeric LTRs— characterized by lower CG and higher CHH methylation compared to arm-located LTRs—suggest an epigenetic environment that actively facilitates both LTR integration and subsequent CENH3 loading.
One of the key discoveries in our study is the identification of a positive feedback mechanism between Cereba elements and CENH3 during rye domestication. Cereba sequences, members of the Chromoviridae clade of LTR retrotransposons, play a crucial role in centromere function [65]. Our analysis reveals that cultivated rye varieties show significantly higher CENH3 nucleosome enrichment on Cereba sequences compared to wild species, with particularly striking differences in the INT and PROT domains. This pattern suggests that modifications in these domains may have been crucial for centromere evolution during domestication, potentially contributing to enhanced chromosomal stability in cultivated varieties.
The evolutionary dynamics of rye centromeres appear to be driven by repeated insertions and rearrangements of Cereba sequences, leading to rapid sequence evolution and diversification among rye subspecies. The Cereba-CENH3 feedback mechanism demonstrates how TEs can transition from potential genomic parasites to essential functional elements during crop domestication. While similar mechanisms have been observed in cotton [12], the extent of Cereba’s influence on centromere architecture appears to be particularly pronounced in rye.
Our findings have significant implications for both basic research and crop improvement. The unique organization of rye’s genome, particularly the relationship between tandem repeats and LTR evolution in terminal regions, provides new insights into mechanisms of genome stability in regions with high repeat content. The identification of specific Cereba sequence modifications associated with enhanced centromere function could guide efforts to engineer more stable chromosomes in hybrid crops. This is especially relevant for various rye translocation lines, notably the widely recognized 1RS-1BL translocation line, where centromere stability is essential for chromosome engineering [66]. In addition, understanding the epigenetic requirements for centromere establishment could facilitate the development of artificial chromosomes for crop biotechnology applications. The distinctive epigenetic features we discovered in retrotransposon-rich centromeres, including their unmarked chromatin state and specific DNA methylation patterns, provide new perspectives for centromere engineering. However, we acknowledge that other factors, including chromosomal rearrangements and broader epigenetic changes, may also contribute to centromeric divergence.
Conclusions
Through high-resolution genome assembly, we demonstrate that rye centromere are dominated by retrotransposons rather than satellite repeats and exhibit a distinctive epigenetic profile characterized by an “unmarked” chromatin state lacking conventional activation/repression marks and displaying reduced CG/CHG but elevated CHH methylation. This specialized environment promotes young LTR retrotransposon insertion and facilitates CENH3 binding. Cross-subspecies analyses further reveal that Cereba retrotransposons—particularly through conserved INT and PROT domains—enhance CENH3 loading, suggesting a positive feedback mechanism that reinforces centromere stability during rye domestication.
Methods
Plant material and DNA extraction
The rye inbred line Weining was subjected to force self-pollination to decrease the genomic heterozygosity and grown in a greenhouse with conditions of 25 °C for 16 h under light and 20 °C for 8 h in the dark. Fluorescence in situ hybridization (FISH) was employed on root tip tissues to confirm the chromosome number, revealing a total of seven pairs of chromosomes. After 14 days, the fresh young leaf tissue was collected and frozen immediately in liquid nitrogen for DNA extraction, and high molecular weight genomic DNA was prepared by the SDS method followed by purification with QIAGEN® Genomic kit (Cat#13343, QIAGEN) according to the standard operating procedure provided by the manufacturer. The DNA degradation and contamination of the extracted DNA was monitored on 1% agarose gels. DNA purity was then detected using NanoDrop™ One UV–Vis spectrophotometer (Thermo Fisher Scientific, USA), of which OD260/280 ranged from 1.8 to 2.0 and OD 260/230 is between 2.0 and 2.2. At last, DNA concentration was further measured by Qubit® 3.0 Fluorometer (Invitrogen, USA).
Library preparation and sequencing (ONT)
A total amount of 2 µg DNA per sample was used as input material for the ONT library preparations. After the sample was qualified, size-select of long DNA fragments were performed using the BluePippin system (Sage Science, USA). Next, the ends of DNA fragments were repaired, and A-ligation reactions were conducted with NEBNext Ultra II End Repair/dA-tailing Kit (Cat# E7546). The adapter in the LSK109 kit was used for further ligation reaction and Qubit® 3.0 Fluorometer (Invitrogen, USA) was used to quantify the size of library fragments. Sequencing was then performed on a Nanopore GridION X5/PromethION sequencer (Oxford Nanopore Technologies, UK) instrument on BENAGEN.
Library preparation and sequencing (PacBio)
SMRTbell target size libraries were constructed for sequencing according to PacBio’s standard protocol (Pacific Biosciences, CA, USA) using either 10 kb or 20 kb preparation solutions. The main steps for library preparation are as follows: (1) gDNA shearing, (2) DNA damage repair, (3) blunt end-ligation with hairpin adapters from the SMRTbell Express Template Prep Kit 2.0 (Pacific Biosciences), (4) size selection, and (5) binding to polymerase. Briefly, a total amount of 2 µg DNA per sample was used for the DNA library preparations. The genomic DNA sample was sheared by g-TUBEs (Covaris, USA) according to the expected size of the fragments for the library. Single-strand overhangs were then removed, and DNA fragments were damage repaired, end polished, and ligated with the stem-loop adaptor for PacBio sequencing. Link-failed fragments were further removed by exonuclease, and target fragments were screened by the BluePippin (Sage Science, USA). The SMRTbell library was then purified using AMPure PB, and Agilent 2100 Bioanalyzer (Agilent technologies, USA) was used to detect the size of library fragments. Sequencing was performed on a PacBio Sequel Ⅱ instrument with Sequel II Sequencing Kit 2.0 on Nextomics.
Library preparation and sequencing (Illumina)
The process for preparing genomic DNA samples for short-read sequencing was performed as follows: Genomic DNA was extracted from the leaves of 2-week-old seedlings using the CTAB method. Library construction was conducted using the MGIEasy Universal DNA Library Prep Kit V1.0 (CAT#1,000,005,250, MGI) according to its standard protocol. In this process, 1 μg of genomic DNA was randomly fragmented with Covaris. The fragmented DNA was then size-selected using MGIEasy DNA Clean beads (CAT#1,000,005,279, MGI) to achieve an average fragment size of 200–400 bp. Subsequent steps included end-repair and 3′ adenylation of the selected fragments, followed by adapter ligation. The DNA underwent PCR amplification, and the resultant products were purified using the MGIEasy DNA Clean beads. The double-stranded PCR products were heat-denatured and circularized using the splint oligo sequence provided in the MGIEasy Circularization Module (CAT#1,000,005,260, MGI). The resulting single-strand circular DNA (ssCir DNA) was then formatted into the final library and underwent QC. The qualified libraries were sequenced using the DNBSEQ-T7RS platform.
Data quality control
Nanopore sequencers output FAST5 files containing signal data, basecalling were first performed to convert the FAST5 files to FASTQ format with Guppy. The raw reads of FASTQ format with mean_qscore_template < 7 were then filtered, resulting in pass reads.
Genome assembly
For 164.8 Gb ultra-long ONT data, 720.8 Gb PacBio HiFi data, and 422.5 Gb Hi-C data, both hifiasm[37] (0.19.8-r603) and verkko [67] (v1.4.1) were used for whole genome assembly with the default parameters. Then we aligned contigs generated by verkko and hifiasm to the chloroplast genome from Secale cereale and Triticum aestivum, and mitochondrion genome from Triticum aestivum using BLAST (2.15.0 +) with default parameters. Contigs with a minimum of 80% of their bases covered by alignments were identified as chloroplast or mitochondrial genome sequences and subsequently excluded from the assembly.
We next split the remaining contigs larger than 1 Mb into 50 kb fragments and utilized mmseq2 [68] (version 14.7e284) to align the contigs against the NT (Nucleotide Sequence Database), providing parameters: easy-search –threads 128 –format-mode 0. Contigs in which 80% of the sequences aligned to the bacterial domain were considered as exogenous contaminants and subsequently removed.
For 506.5 Gb NGS Illumina paired-end data, trim-galore (0.6.10) and trimmomatic [69] (0.39) were used to remove the adapter and low-quality reads. Then trimmed reads were mapped to the remaining contigs with bwa-mem [70] (version 0.7.17-r1188) software using default parameters, and the coverage of these contigs was calculated with samtools coverage [71] (version 1.17). The percentage of covered bases below 1 in contigs was eliminated.
Contig scaffolding
For these two versions of contigs (hifiasm and verkko), different scaffold strategies were further employed. We indexed the corrected contigs and performed Hi-C data alignment using Chromap [40] (0.2.4-r467) with parameters (chromap index; chromap –preset hic –remove-pcr-duplicates –SAM), followed by genome scaffolding using Yahs [41] (version 1.2a.1) with default parameters. Subsequently, we utilized juicebox [72] for validation and refinement, and employed the juicer-post function to generate the final genome sequence from the scaffolding results. Juicer [42] (version 1.6) and 3dDNA [73] (version 180,114) were employed in another strategy to scaffold the corrected contigs, enhancing resolution and precision, and “run-asm-pipeline” command specifies one iterative round for mis-join correction using the parameters (-r 1).
Additionally, we utilized a homology-based scaffolding approach for contig placement and orientation determination. We ran Ragtag [39] (v2.1.0) using published Weining genome sequence as the reference sequence, but excluded the un-scaffolded sequence (GWHASIY00000008) with the command (ragtag.py scaffold –aligner minimap2 –t 120 –e exclude.txt). After scaffolding, the contigs were ordered and oriented, and we ensured the consistency between our assembly and the previously published Weining rye genome with regard to chromosome numbering.
Gap close
The initial genome assembly was performed using hifiasm, followed by a primary scaffolding with ragtag. Subsequent secondary scaffolding was accomplished using Chromap and Yahs, producing a reference genome for assembly-to-reference alignment. We conducted alignments between the reference genome and other versions of the genome (verkko, hifiasm + 3dDNA, and the published genome) using minimap2 [74] (2.26-r1175) with parameters: -ax asm5.
We then filtered the PAF records for alignments with a mapping quality equal to 60 (within the range of 0–255). Among these filtered records, we selectively chose those that spanned genomic gaps. In cases where multiple records spanned the same gap, we prioritized records based on the CIGAR value, evaluating the number of matches in the 500 bp, 1000 bp, and 2000 bp regions flanking the gap. The record with the highest mapping quality was selected for subsequent gap closure steps.
We aligned the sequences intended for gap filling against the un-scaffolded contigs from the reference genome using minimap2 with the same parameters. Subsequently, we conducted a final filtering step based on a mapping quality equal to 60 and the CIGAR value, removing contigs that exhibited a high degree of similarity with the gap-filling sequences.
TE annotation
Both de novo prediction and homology-based search methods were employed simultaneously for the identification and annotation of repetitive sequences in the genome. In the homology-based annotation approach, tools such as RECON, RepeatScout, LTRHarvest, CD-HIT, LTR_retriever, MAFFT, Ninja, UCSC TwoBit Tools, and RepeatModeler [75] were utilized for de novo searching, alignment, and prediction of repetitive sequences. The ab initio repetitive sequences obtained from the analysis were aligned with the Repbase database, and subsequently, the RepeatClassifier tool within RepeatModeler was utilized for classification.
The CLARI-TE tool [76] was employed for the annotation of repetitive sequences using homology-based sequence search methods. The manually curated cereal repetitive elements sequence clariTeRep.fna was utilized for TE annotation with RepeatMasker, providing the following parameters: “-lib clariTeRep.fna -gff -xm -pa 30”. The EMBLmyGFF3 tool was subsequently used for converting the results into an alternate data format. The output results were split based on chromosome numbers, and the highly similar overlapping repetitive sequences were corrected, and classification of the repetitive sequences was conducted using clariTE. The final annotation of repetitive sequences in the genome was obtained by merging the results from both de novo prediction methods and homology-based search methods.
Additionally, LTR-finder [77] (v1.07), LTR-retriever [78] (v2.9.0), and LTR-harvest [79] (v1.1) software were employed for de novo full-length LTR prediction within the genome. To filter out false positives and integrate the final results, the Extensive de novo TE Annotator [80] (v2.0.1) software was used with default parameters.
Gene annotation
The gene annotation of the rye genome was conducted using three main approaches: ab initio prediction, evidence-based prediction using RNA-seq data, and homology-based sequence alignment. Prior to gene prediction, the genome was first subjected to repeat sequence masking using RepeatMasker. For evidence-based prediction, three methods were employed. Initially, RNA-seq data underwent adapter trimming and low-quality filtering using Cutadapt and Trimmomatic. Subsequently, the processed RNA-seq reads were aligned to the rye reference genome using BWA and STAR [81] (version = 2.7.11a) separately. The resulting alignments were then used for transcriptome assembly with Scallop [82] (v0.10.5), StringTie [83] (v2.2.1), and Cufflinks [84] (v2.2.1). Finally, Mikado [85] was employed to generate an optimal transcript set based on various pieces of evidence, including three different transcript assemblies, high-confidence splice sites, ORFs identified by TransDecoder, and homologous transcripts from the Plants Gramene database [86], including Arabidopsis thaliana, Secale cereal, Triticum aestivum, and Hordeum vulgare.
In the homology-based sequence alignment approach, CDS sequences were obtained based on the annotation of a previously published Weining Rye genome and annotation using gffread [87] (v0.12.1). A database was constructed using GMAP, and the CDS sequences were aligned to the rye genome to generate homology-based genome annotations. The Gemoma was also employed for gene annotation through homology-based sequence searches, utilizing protein-coding sequences from Arabidopsis thaliana, Triticum aestivum, and Hordeum vulgare.
In the ab initio prediction method, gene annotation of the genome was performed using the GeneMark-ES [88] (version 4.33) and AUGUSTUS [89] (version 3.2.3) prediction tools. The script, autoAugTrain.pl, was used to train gene prediction models using previously published annotations for rye, and predictions were also made using the default wheat training set provided by the software.
All gene models generated through these methods were integrated using EVidenceModeler [90] (EVM) to produce a non-redundant gene annotation set. The annotation process also included Gene Ontology (GO) annotation using InterProScan, filtering for genes related to TEs, and the final identification of high-confidence coding protein genes.
CENH3 and histone modification ChIP-seq library
Nuclei were isolated from 3 to 5 g of young rye leaves ground carefully in liquid nitrogen, using a lysis buffer containing 1 mM EDTA, protease inhibitor cocktail, 10 mM Tris–HCl (pH 7.5), 10 mM NaCl, 0.5% NP-40, 5 mM β-ME, and 0.1 mM PMSF. Pelleted nuclei underwent two wash steps with lysis buffer and centrifugation at 1200 g for 5 min at 4 °C. The chromatin pellet was resuspended in MNase buffer (10% sucrose, 50 mM Tris–HCl pH 7.5, 1 mM CaCl2, 1 mM MgCl2, 0.1 mM PMSF) and digested with 0.5 U MNase (New England BioLabs) at 37 °C for 20 min. EDTA quenched the reaction, followed by centrifugation at 12,000 rpm for 10 min at 4 °C. Fifty microliters of protein A beads precleared the supernatant at 4 °C for 3–4 h with rotation. An input sample was separated, while the CENH3 or histone modification antibody incubated with the remainder overnight at 4 °C. Immunocomplexes recovered by 50 µL protein A beads underwent washes with low, medium, and high salt buffers. Elution occurred in buffer (50 mM NaCl, 20 mM Tris–HCl, 5 mM EDTA, 2% SDS) at 65 °C for 55 min.
The anti-CENH3 antibody utilized for the study of centromere in rye was originally derived from wheat, targeting the RTKHPAVRKTK peptide sequence. Additionally, our analysis utilized antibodies against H3K4me3, H3K9me3, and H3K27me3 from Millipore.
Identification of centromeres
Raw sequence reads first underwent quality control using the FastQC software. Trimmomatic v.0.36 then removed adaptor sequences and low-quality bases. The high-quality trimmed reads were mapped to our assembly genome utilizing the BWA-MEM software with default parameters. Mapped reads were extracted using SAMtools view -F 20, and duplicates were removed via SAMtools v.1.3.1. For visualization, the alignment BAM files were converted to normalized 10 bp binned coverage files using deepTools. Finally, data snapshots were constructed employing the Integrative Genomics Viewer visualization tool, and the final centromeric regions were determined by visual inspection of the distribution of CENH3 ChIP–seq peaks.
Identification of CENH3-enrich/depleted subdomains
CENH3-binding domains were identified by comparing the ChIP and input data using deepTools, which is designed to identify ChIP-enriched signals that tend to form specific subdomains. The input and CENH3 ChIP-seq alignment data were processed separately using deepTools with parameters -bs 1000, –of bedgraph, and –normalizeUsing RPKM. The input data was utilized to compute the random signal distribution, and the p-value for each interval was calculated using the pnorm function from the base package in R. Intervals with p-values < = 0.0005 were defined as CENH3-enriched regions, indicating that the CENH3 signal was significantly higher than the random signal in these regions. Conversely, intervals with p-values > 0.0005 were defined as CENH3-depleted regions, implying that the CENH3 signal was not significantly higher than the random signal. Subsequently, intervals not belonging to the centromeric regions were filtered out based on the definition of centromere regions.
ssDRIP-seq library
The methodology for the DNA:RNA immunoprecipitation assay was derived from a prior study [91]. Nuclei isolated from 5 g of rye seedlings were resuspended in 1.6 to 4 mL TE and lysed through SDS/proteinase K digestion overnight at 37 °C. Extracts containing nucleic acids were mixed with phenol–chloroform, and genomic DNA was precipitated using salt-ethanol precipitation, incubated at room temperature for 30 min. The precipitated DNA was gently spooled, washed with 70% ethanol, air-dried, and dissolved in 100 µL TE buffer. Fifty microliters of DNA was sonicated on a Covaris system (10% Duty Factor, 200 cycles/burst, 140 peak incident power, and 60 s) to yield ~ 300 bp fragments on average. RNase H (Takara 2151) treatment at 37 °C overnight was performed on half of each sample. DRIP was carried out with the S9.6 antibody (Kerafast ENH001), and the DRIPed DNA was detected using fragment analysis in NGS mode. The ssDRIP-seq libraries were constructed from ~ 200 bp DRIPed DNA fragments using the Accel-NGS 1S Plus DNA Library Kit (Swift Biosciences) per the manufacturer’s protocol. Library DNA was analyzed on a fragment analyzer and sequenced on an Illumina NovaSeq system (2 × 150 bp). qPCR of the eluted DNA using SYBR Green Mix (Roche 04887352001) was performed on a Roche LightCycler 480 Instrument II and analyzed via the comparative CT method. DNA:RNA hybrid enrichment was calculated based on the IP/input ratio.
ssDRIP-seq data analysis
Similar to ChIP-seq, high-quality reads after Trimmomatic v.0.36 adaptor removal and quality filtering were mapped to the reference genome using BWA-MEM. Mapped reads were extracted with SAMtools view -F −20 and de-duplicated. All ssDRIP-seq mapped reads were categorized into forward (wR-loops), representing R-loop structures with a DNA:RNA hybrid on the Crick strand and ssDNA on the Watson strand; and reverse reads (cR-loops). The MACS2 peak-calling software detected enriched peaks from the alignments. For visualization, BAM files were converted to normalized 10 bp binned coverage files. Lastly, the Integrative Genomics Viewer constructed data snapshots.
FISH
During the FISH assay, mitotic chromosomes were isolated from root apical meristems [92]. Oligonucleotide probes were synthesized with a green fluorophore (5′6-FAM) labeling at their 5′ ends (BGI Tech). In addition, the ChIPed DNA and PCR-amplified DNAs were labeled with Alexa Fluor 594–5-dUTP (red) or Alexa Fluor 488–5-dUTP (green) as needed using the nick-translation method. Images were captured using an Olympus BX61 microscope and processed with Adobe Photoshop CS 5.0.
Phylogenetic analysis of non-overlapping tandem repeats
Tandem repeat arrays (pSc119.2, pSc200, and pSc250) were detected genome-wide by BLASTN searches of family consensus monomers against the rye assembly (e-value ≤ 1e-10, identity ≥ 90%, query coverage ≥ 80%). Overlapping or nested hits were collapsed by retaining the match with the highest bit score, yielding non-overlapping monomers. Monomers were kept if their lengths were within ± 10% of the family consensus and contained < 5% ambiguous bases (N). For computational tractability, sequences were randomly subsampled to 2000 per family without replacement. A maximum-likelihood phylogeny was inferred with IQ-TREE (multicore, v2.0.3) with default parameters and 1000 ultrafast bootstrap replicates to assess node support [93]. Trees were visualized and annotated in iTOL (v6.8).
DNA secondary structure predictions using ViennaRNA
DNA secondary structure prediction using ViennaRNA was performed using a 300 nt sliding window with 1 nt step size across all chromosomes, and the analysis was performed using RNAfold v2.1.5 with parameter file dna_mathews2004.par, specifically for folding single-stranded DNA sequences from the ViennaRNA package v2.1.5.
Comparative analysis of rye relatives
Raw CENH3 ChIP-seq reads from different rye subgenomes were mapped to the Weining rye reference genome using BWA-MEM with default parameters. Duplicate reads were removed, and alignments were filtered to retain only high-quality mappings (MAPQ ≥ 20) using SAMtools. Processed BAM files were converted to BigWig format (bamCoverage, deepTools) for visualization in Integrative Genomics Viewer (IGV).
Centromeric repeat elements were predicted de novo from CENH3-enriched regions using RepeatExplorer2’s graph-based clustering approach. Default settings were applied to classify repeat families, with a focus on identifying centromere-specific tandem repeats and retrotransposons. For comparative analysis of subgenomes, we employed RepeatExplorer’s comparative mode to cluster repeat families shared across subgenomes, identify lineage-specific centromeric repeats, and quantify divergence in repeat composition using pairwise similarity thresholds. Repetitive element dynamics were further examined by calculating differential abundance of major repeat classes (e.g., CRR-like elements) among subgenomes, supported by hierarchical clustering of repeat profiles.
Statistical analysis
All statistical analyses were performed in R (v4.1.0) unless stated otherwise. Unless otherwise noted, tests were two-sided, and the significance threshold was α = 0.05. For pairwise comparisons of signal intensities, methylation levels, or k-mer abundances between two groups (e.g., centromere vs. arm; Cereba vs. random loci), we used the Wilcoxon rank-sum test (wilcox.test), two-sided. ECDFs were drawn with stat_ecdf (ggplot2).
Associations between variables (e.g., CENH3 signal vs. Gypsy content, methylation vs. distance to centromere) were evaluated with Spearman’s ρ (cor.test, method = “spearman”). For genomic-context enrichment, we additionally performed permutation tests with 1000 randomizations using region-preserving shuffles, reporting empirical P values.
For WGBS, methylation fractions at CG/CHG/CHH contexts were computed as mC/(mC + uC) per cytosine; regional methylation values represent coverage-weighted means. Comparisons of methylation between genomic compartments used Wilcoxon tests with BH correction.
k-mer and PCA analyses. k-mer counts from CENH3 ChIP-seq were normalized to counts-per-million and log2 transformed where indicated. PCA was performed with prcomp on centered and scaled matrices; clustering was evaluated by Euclidean distance and complete linkage (hclust).
To assess centromere-specific CENH3 enrichment differences between cultivated rye (Group A) and wild relatives (Group B), we compared normalized ChIP-seq signals (RPKM) at cereba repeat loci. Group A comprised Weining, Imperial, and Lo7 genomes, while Group B included S. vavilovii, S. sylvestre, and S. strictum. Differential CENH3 occupancy was evaluated using the non-parametric Wilcoxon rank-sum test (implemented in R v4.2.1) between Group A and Group B signal distributions. Adjusted p-values (< 0.05) were computed to account for multiple testing (Benjamini–Hochberg correction).
Plots were generated with ggplot2 (v3.5.0). All code used to run the statistical analyses is available in the accompanying repository (see Availability of data and materials).
Data availability
The PacBio HiFi, Oxford Nanopore ultra-long, and Illumina sequencing data generated in this study are available from the National Genomics Data Center (NGDC), China, under BioProject PRJCA041046 [94]. The genome assembly has been deposited under the same BioProject, while the corresponding genome annotation is available at Figshare [95]. Histone ChIP-seq and whole-genome bisulfite sequencing data are available from NGDC under BioProject PRJCA041081 [96]. Public RNA-seq data used for gene annotation were retrieved from NCBI under accession PRJNA680499 [97]. Public ssDRIP-seq data were obtained from NGDC under accession CRA011914 [98]. The code used in this project is released under the MIT License and has been archived on Zenodo (DOI: https://doi.org/10.5281/zenodo.17072245) [99]. The repository is also mirrored on GitHub (https://github.com/CongyangY/RyeAssembly_centromere) [100].
References
Westhorpe FG, Straight AF. Functions of the centromere and kinetochore in chromosome segregation. Curr Opin Cell Biol. 2013;25:334–40.
Feng C, Yuan J, Bai H, Liu Y, Su H, Liu Y, et al. The deposition of CENH3 in maize is stringently regulated. Plant J. 2020;102:6–17.
Evtushenko EV, Elisafenko EA, Gatzkaya SS, Lipikhina YA, Houben A, Vershinin AV. Conserved molecular structure of the centromeric histone CENH3 in Secale and its phylogenetic relationships. Sci Rep. 2017;7:17628.
Henikoff S, Ahmad K, Malik HS. The centromere paradox: stable inheritance with rapidly evolving DNA. Science. 2001;293:1098–102.
Liu Y, Yi C, Fan C, Liu Q, Liu S, Shen L, et al. Pan-centromere reveals widespread centromere repositioning of soybean genomes. Proc Natl Acad Sci U S A. 2023;120:e2310177120.
Altemose N, Logsdon GA, Bzikadze AV, Sidhwani P, Langley SA, Caldas GV, et al. Complete genomic and epigenetic maps of human centromeres. Science. 2022;376:eabl4178.
Naish M, Alonge M, Wlodzimierz P, Tock AJ, Abramson BW, Schmucker A, et al. The genetic and epigenetic landscape of the Arabidopsis centromeres. Science. 2021;374:eabi7489.
Song JM, Xie WZ, Wang S, Guo YX, Koo DH, Kudrna D, et al. Two gap-free reference genomes and a global view of the centromere architecture in rice. Mol Plant. 2021;14:1757–67.
Cui J, Zhu C, Shen L, Yi C, Wu R, Sun X, et al. The gap-free genome of Forsythia suspensa illuminates the intricate landscape of centromeres. Hortic Res. 2024;11:uhae185.
Deng Y, Liu S, Zhang Y, Tan J, Li X, Chu X, et al. A telomere-to-telomere gap-free reference genome of watermelon and its mutation library provide important resources for gene discovery and breeding. Mol Plant. 2022;15:1268–84.
Zhang Y, Zhao M, Tan J, Huang M, Chu X, Li Y, et al. Telomere-to-telomere Citrullus super-pangenome provides direction for watermelon breeding. Nat Genet. 2024;56:1750–61.
Chang X, He X, Li J, Liu Z, Pi R, Luo X, et al. High-quality Gossypium hirsutum and Gossypium barbadense genome assemblies reveal the landscape and evolution of centromeres. Plant Commun. 2024;5:100722.
Chen J, Wang Z, Tan K, Huang W, Shi J, Li T, et al. A complete telomere-to-telomere assembly of the maize genome. Nat Genet. 2023;55:1221–31.
Huang G, Bao Z, Feng L, Zhai J, Wendel JF, Cao X, et al. A telomere-to-telomere cotton genome assembly reveals centromere evolution and a mutator transposon-linked module regulating embryo development. Nat Genet. 2024;56:1953–63.
Hocher A, Ruault M, Kaferle P, Descrimes M, Garnier M, Morillon A, et al. Expanding heterochromatin reveals discrete subtelomeric domains delimited by chromatin landscape transitions. Genome Res. 2018;28:1867–81.
Tashiro S, Nishihara Y, Kugou K, Ohta K, Kanoh J. Subtelomeres constitute a safeguard for gene expression and chromosome homeostasis. Nucleic Acids Res. 2017;45:10333–49.
van Emden TS, Forn M, Forne I, Sarkadi Z, Capella M, Martin Caballero L, et al. Shelterin and subtelomeric DNA sequences control nucleosome maintenance and genome stability. EMBO Rep. 2019. https://doi.org/10.15252/embr.201847181.
Berthiau AS, Yankulov K, Bah A, Revardel E, Luciano P, Wellinger RJ, et al. Subtelomeric proteins negatively regulate telomere elongation in budding yeast. EMBO J. 2006;25:846–56.
Kwapisz M, Morillon A. Subtelomeric transcription and its regulation. J Mol Biol. 2020;432:4199–219.
Li G, Wang L, Yang J, He H, Jin H, Li X, et al. A high-quality genome assembly highlights rye genomic characteristics and agronomically important genes. Nat Genet. 2021;53:574–84.
Bartos J, Paux E, Kofler R, Havrankova M, Kopecky D, Suchankova P, et al. A first survey of the rye (Secale cereale) genome composition through BAC end sequencing of the short arm of chromosome 1R. BMC Plant Biol. 2008;8:95.
Vershinin AV, Schwarzacher T, Heslop-Harrison JS. The large-scale genomic organization of repetitive DNA families at the telomeres of rye chromosomes. Plant Cell. 1995;7:1823–33.
McIntyre CL, Pereira S, Moran LB, Appels R. New Secale cereale (rye) DNA derivatives for the detection of rye chromosome segments in wheat. Genome. 1990;33:635–40.
Vershinin AV, Elisafenko EA, Evtushenko EV. Genetic redundancy in rye shows in a variety of ways. Plants. 2023;12(2):282. https://doi.org/10.3390/plants12020282.
Liu C, Huang Y, Guo X, Yi C, Liu Q, Zhang K, et al. Young retrotransposons and non-B DNA structures promote the establishment of dominant rye centromere in the 1RS.1BL fused centromere. New Phytol. 2024;241:607–22.
Liu C, Fu S, Yi C, Liu Y, Huang Y, Guo X, et al. Unveiling the distinctive traits of functional rye centromeres: minisatellites, retrotransposons, and R-loop formation. Sci China Life Sci. 2024. https://doi.org/10.1007/s11427-023-2524-0.
Wang J, Liu Y, Su H, Guo X, Han F. Centromere structure and function analysis in wheat-rye translocation lines. Plant J. 2017;91:199–207.
Bauer E, Schmutzer T, Barilar I, Mascher M, Gundlach H, Martis MM, et al. Towards a whole-genome sequence for rye (Secale cereale L.). Plant J. 2017;89:853–69.
Schreiber M, Himmelbach A, Borner A, Mascher M. Genetic diversity and relationship between domesticated rye and its wild relatives as revealed through genotyping-by-sequencing. Evol Appl. 2019;12:66–77.
Martis MM, Zhou R, Haseneyer G, Schmutzer T, Vrana J, Kubalakova M, et al. Reticulate evolution of the rye genome. Plant Cell. 2013;25:3685–98.
Rabanus-Wallace MT, Hackauf B, Mascher M, Lux T, Wicker T, Gundlach H, et al. Chromosome-scale genome assembly provides insights into rye biology, evolution and agronomic potential. Nat Genet. 2021;53:564–73.
Yi C, Liu Q, Huang Y, Liu C, Guo X, Fan C, et al. Non-B-form DNA is associated with centromere stability in newly-formed polyploid wheat. Sci China Life Sci. 2024;67:1479–88.
Liu Q, Yi C, Zhang Z, Su H, Liu C, Huang Y, et al. Non-B-form DNA tends to form in centromeric regions and has undergone changes in polyploid oat subgenomes. Proc Natl Acad Sci U S A. 2023;120:e2211683120.
Huang Y, Shi Q, Zhou C, Wang C, Liu Y, Yi C, Su H, Han F. Wide hybridizations reveal the robustness of functional centromeres in Triticum-Aegilops species complex lines. J Genet Genomics. 2024;51(5):570–3. https://doi.org/10.1016/j.jgg.2023.12.001.
Wolfgruber TK, Sharma A, Schneider KL, Albert PS, Koo DH, Shi J, et al. Maize centromere structure and evolution: sequence analysis of centromeres 2 and 5 reveals dynamic loci shaped primarily by retrotransposons. PLoS Genet. 2009;5:e1000743.
Wlodzimierz P, Rabanal FA, Burns R, et al. Cycles of satellite and transposon evolution in Arabidopsis centromeres. Nature. 2023;618(7965):557–65. https://doi.org/10.1038/s41586-023-06062-z.
Cheng H, Concepcion GT, Feng X, Zhang H, Li H. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nat Methods. 2021;18:170–5.
Rautiainen M, Nurk S, Walenz BP, et al. Telomere-to-telomere assembly of diploid chromosomes with Verkko. Nat Biotechnol. 2023;41(10):1474–82. https://doi.org/10.1038/s41587-023-01662-6.
Alonge M, Lebeigle L, Kirsche M, Jenike K, Ou S, Aganezov S, et al. Automated assembly scaffolding using RagTag elevates a new tomato system for high-throughput genome editing. Genome Biol. 2022;23:258. https://doi.org/10.1038/s41587-023-01662-6.
Zhang H, Song L, Wang X, Cheng H, Wang C, Meyer CA, et al. Fast alignment and preprocessing of chromatin profiles with Chromap. Nat Commun. 2021;12:6566.
Zhou C, McCarthy SA, Durbin R. Yahs: yet another hi-c scaffolding tool. Bioinformatics. 2023. https://doi.org/10.1093/bioinformatics/btac808.
Durand NC, Shamim MS, Machol I, Rao SS, Huntley MH, Lander ES, et al. Juicer provides a one-click system for analyzing loop-resolution Hi-C experiments. Cell Syst. 2016;3:95–8.
Xu M, Guo L, Gu S, Wang O, Zhang R, Peters BA, et al. TGS-gapcloser: a fast and accurate gap closer for large genomes with low coverage of error-prone long reads. Gigascience. 2020. https://doi.org/10.1093/gigascience/giaa094.
Ou S, Chen J, Jiang N. Assessing genome assembly quality using the LTR assembly index (LAI). Nucleic Acids Res. 2018;46:e126.
Seppey M, Manni M, Zdobnov EM. Busco: assessing genome assembly and annotation completeness. Methods Mol Biol. 2019;1962:227–45.
Rhie A, Walenz BP, Koren S, Phillippy AM. Merqury: reference-free quality, completeness, and phasing assessment for genome assemblies. Genome Biol. 2020;21:245.
Evtushenko EV, Levitsky VG, Elisafenko EA, Gunbin KV, Belousov AI, Safar J, et al. The expansion of heterochromatin blocks in rye reflects the co-amplification of tandem repeats and adjacent transposable elements. BMC Genomics. 2016;17:337.
Taketa S, Ando H, Takeda K, Harrison GE, Heslop-Harrison JS. The distribution, organization and evolution of two abundant and widespread repetitive DNA sequences in the genus Hordeum. Theor Appl Genet. 2000;100:169–76.
Contento A, Heslop-Harrison JS, Schwarzacher T. Diversity of a major repetitive DNA sequence in diploid and polyploid Triticeae. Cytogenet Genome Res. 2005;109:34–42.
Vershinin AV, Alkhimova EG, Heslop-Harrison JS. Molecular diversification of tandemly organized DNA sequences and heterochromatic chromosome regions in some Triticeae species. Chromosoma Res. 1996;4:517–25.
Talbert PB, Masuelli R, Tyagi AP, Comai L, Henikoff S. Centromeric localization and adaptive evolution of an Arabidopsis histone H3 variant. Plant Cell. 2002;14:1053–66.
Zhao J, Xie Y, Kong C, Lu Z, Jia H, Ma Z, et al. Centromere repositioning and shifts in wheat evolution. Plant Commun. 2023;4:100556.
Hofstatter PG, Thangavel G, Lux T, Neumann P, Vondrak T, Novak P, et al. Repeat-based holocentromeres influence genome architecture and karyotype evolution. Cell. 2022;185(3153–3168):e3118.
Xie L, Huang Y, Huang W, Shang L, Sun Y, Chen Q, Bi S, Suo M, Zhang S, Yang C, et al. Genetic diversity and evolution of rice centromeres. bioRxiv 2024:2024.2007.2028.605524.
Sun Y, Shen E, Hu Y, Wu D, Feng Y, Lao S, et al. Population genomic analysis reveals domestication of cultivated rye from weedy rye. Mol Plant. 2022;15:552–61.
Malik HS, Eickbush TH. Modular evolution of the integrase domain in the Ty3/Gypsy class of LTR retrotransposons. J Virol. 1999;73:5186–90.
Heuberger M, Koo DH, Ahmed HI, Tiwari VK, Abrouk M, Poland J, et al. Evolution of Einkorn wheat centromeres is driven by the mutualistic interplay of two LTR retrotransposons. Mob DNA. 2024;15:16.
International Rice Genome Sequencing P. The map-based sequence of the rice genome. Nature. 2005;436:793–800.
Zhang C, Xie L, Yu H, Wang J, Chen Q, Wang H. The T2T genome assembly of soybean cultivar ZH13 and its epigenetic landscapes. Mol Plant. 2023. https://doi.org/10.1016/j.molp.2023.10.003.
Wang L, Zhang M, Li M, Jiang X, Jiao W, Song Q. A telomere-to-telomere gap-free assembly of soybean genome. Mol Plant. 2023;16:1711–4.
Liu Y, Du H, Li P, Shen Y, Peng H, Liu S, et al. Pan-genome of wild and cultivated soybeans. Cell. 2020;182(162–176):e113.
Shen Y, Liu J, Geng H, Zhang J, Liu Y, Zhang H, et al. De novo assembly of a Chinese soybean genome. Sci China Life Sci. 2018;61:871–84.
Chen C, Wu S, Sun Y, Zhou J, Chen Y, Zhang J, et al. Three near-complete genome assemblies reveal substantial centromere dynamics from diploid to tetraploid in Brachypodium genus. Genome Biol. 2024;25:63.
Zhao J, Xie Y, Kong C, Lu Z, Jia H, Ma Z, et al. Centromere repositioning and shifts in wheat evolution. Plant Commun. 2023. https://doi.org/10.1016/j.xplc.2023.100556.
Gorinsek B, Gubensek F, Kordis D. Evolutionary genomics of chromoviruses in eukaryotes. Mol Biol Evol. 2004;21:781–98.
Liu Y, Liu Q, Yi C, Liu C, Shi Q, Wang M, et al. Past innovations and future possibilities in plant chromosome engineering. Plant Biotechnol J. 2024. https://doi.org/10.1111/pbi.14530.
Rautiainen M, Nurk S, Walenz BP, Logsdon GA, Porubsky D, Rhie A, et al. Telomere-to-telomere assembly of diploid chromosomes with verkko. Nat Biotechnol. 2023;41:1474–82.
Steinegger M, Soding J. MMseqs2 enables sensitive protein sequence searching for the analysis of massive data sets. Nat Biotechnol. 2017;35:1026–8.
Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30:2114–20.
Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25:1754–60.
Danecek P, Bonfield JK, Liddle J, Marshall J, Ohan V, Pollard MO, Whitwham A, Keane T, McCarthy SA, Davies RM, Li H. Twelve years of SAMtools and BCFtools. Gigascience. 2021;10:giab008. https://doi.org/10.1093/gigascience/giab008.
Durand NC, Robinson JT, Shamim MS, Machol I, Mesirov JP, Lander ES, et al. Juicebox provides a visualization system for Hi-C contact maps with unlimited zoom. Cell Syst. 2016;3:99–101.
Dudchenko O, Batra SS, Omer AD, Nyquist SK, Hoeger M, Durand NC, et al. De novo assembly of the Aedes aegypti genome using Hi-C yields chromosome-length scaffolds. Science. 2017;356:92–5.
Li H. New strategies to improve minimap2 alignment accuracy. Bioinformatics. 2021;37:4572–4.
Flynn JM, Hubley R, Goubert C, Rosen J, Clark AG, Feschotte C, et al. RepeatModeler2 for automated genomic discovery of transposable element families. Proc Natl Acad Sci U S A. 2020;117:9451–7.
Daron J, Glover N, Pingault L, Theil S, Jamilloux V, Paux E, et al. Organization and evolution of transposable elements along the bread wheat chromosome 3B. Genome Biol. 2014;15:546.
Xu Z, Wang H. Ltr_finder: an efficient tool for the prediction of full-length LTR retrotransposons. Nucleic Acids Res. 2007;35:W265-268.
Ou S, Jiang N. Ltr_retriever: a highly accurate and sensitive program for identification of long terminal repeat retrotransposons. Plant Physiol. 2018;176:1410–22.
Ellinghaus D, Kurtz S, Willhoeft U. LTRharvest, an efficient and flexible software for de novo detection of LTR retrotransposons. BMC Bioinformatics. 2008;9:18.
Ou S, Su W, Liao Y, Chougule K, Agda JRA, Hellinga AJ, et al. Benchmarking transposable element annotation methods for creation of a streamlined, comprehensive pipeline. Genome Biol. 2019;20:275.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29:15–21.
Shao M, Kingsford C. Accurate assembly of transcripts through phase-preserving graph decomposition. Nat Biotechnol. 2017;35:1167–9.
Kovaka S, Zimin AV, Pertea GM, Razaghi R, Salzberg SL, Pertea M. Transcriptome assembly from long-read RNA-seq alignments with StringTie2. Genome Biol. 2019;20:278.
Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene and transcript expression analysis of RNA-seq experiments with tophat and cufflinks. Nat Protoc. 2012;7:562–78.
Venturini L, Caim S, Kaithakottil GG, Mapleson DL, Swarbreck D. Leveraging multiple transcriptome assembly methods for improved gene structure annotation. GigaScience. 2018;7(8):giy093. https://doi.org/10.1093/gigascience/giy093.
Tello-Ruiz MK, Jaiswal P, Ware D. Gramene: a resource for comparative analysis of plants genomes and pathways. Methods Mol Biol. 2022;2443:101–31.
Pertea G, Pertea M. GFF utilities: GffRead and GffCompare. F1000Research. 2020;9:ISCB Comm J–304. https://doi.org/10.12688/f1000research.23297.2.
Ter-Hovhannisyan V, Lomsadze A, Chernoff YO, Borodovsky M. Gene prediction in novel fungal genomes using an ab initio algorithm with unsupervised training. Genome Res. 2008;18:1979–90.
Hoff KJ, Stanke M. Predicting genes in single genomes with AUGUSTUS. Curr Protoc Bioinformatics. 2019;65:e57.
Haas BJ, Salzberg SL, Zhu W, Pertea M, Allen JE, Orvis J, et al. Automated eukaryotic gene structure annotation using EVidenceModeler and the program to assemble spliced alignments. Genome Biol. 2008;9:R7.
Liu Y, Liu Q, Su H, et al. Genome-wide mapping reveals R-loops associated with centromeric repeats in maize. Genome Res. 2021;31(8):1409–18. https://doi.org/10.1101/gr.275270.121.
Han F, Lamb JC, Birchler JA. High frequency of centromere inactivation resulting in stable dicentric chromosomes of maize. Proc Natl Acad Sci U S A. 2006;103:3238–43.
Letunic I, Bork P. Interactive tree of life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 2021;49:W293–6.
Yi C, Liu Q, Zhu C, Liu C, Zhou C, He W, Wang C, Yuan J, Liu Y, Han F. Rye genome assembly. Datasets. National Genomics Data Center (NGDC), PRJCA041046. https://bigd.big.ac.cn/bioproject/browse/PRJCA041046 (2025).
Yi C, Liu Q, Zhu C, Liu C, Zhou C, He W, Wang C, Yuan J, Liu Y, Han F. Genome assembly and annotation of rye (Secale cereale, Weining v2.0). Datasets. Figshare. https://doi.org/10.6084/m9.figshare.30067834 (2025).
Yi C, Liu Q, Zhu C, Liu C, Zhou C, He W, Wang C, Yuan J, Liu Y, Han F. Rye epigenome profiling. Datasets. National Genomics Data Center (NGDC), PRJCA041081. https://bigd.big.ac.cn/bioproject/browse/PRJCA041081 (2025).
Li G, Wang L, Yang J, He H, Jin H, Li X, Ren T, Ren Z, Li F, Han X, et al. NGS transcriptome data for Weining rye genome gene annotation. Datasets. National Center for Biotechnology Information (NCBI), PRJNA680499. https://www.ncbi.nlm.nih.gov/bioproject/PRJNA680499 (2020).
Liu C, Fu S, Yi C, Liu Y, Huang Y, Guo X, Zhang K, Liu Q, Birchler JA, Han F. R-loops of rye. Datasets. National Genomics Data Center (NGDC), CRA011914. https://ngdc.cncb.ac.cn/gsa/browse/CRA011914 (2023).
Yi C, Liu Q, Zhu C, Liu C, Zhou C, He W, Wang C, Yuan J, Liu Y, Han F. High-resolution genome assembly reveals retrotransposon-mediated centromere dynamics in rye. Zenodo. https://doi.org/10.5281/zenodo.17072244 (2025)
Yi C, Liu Q, Zhu C, Liu C, Zhou C, He W, Wang C, Yuan J, Liu Y, Han F. High-resolution genome assembly reveals retrotransposon-mediated centromere dynamics in rye. Github. https://github.com/CongyangY/RyeAssembly_centromere (2025).
Acknowledgements
We thank Chengzhi Liang (Institute of Genetic and Developmental Biology, Chinese Academy of Sciences) for suggestions on the manuscript and rye assembly. We thank Nathan D. Han from The Edison Family Center for critical reading of the article and helpful comments.
Peer review information
Wenjing She was the primary editor of this article and managed its editorial process and peer review in collaboration with the rest of the editorial team. The peer-review history is available in the online version of this article.
Funding
This work was supported by the National Key Research and Development Program of China (2022YFF1003303) and the National Natural Science Foundation of China (31991212).
Author information
Authors and Affiliations
Contributions
Congyang Yi, Yang Liu, and Fangpu Han conceived and designed the study. Congyang Yi carried out the genome assembly and annotation. Congyang Yi and Qian Liu performed the data analyses. Qian Liu and Congle Zhu conducted the key experiments, including FISH and CENH3 ChIP-seq. Chang Liu, Chen Zhou, Wanna He, Chunhui Wang, and Jing Yuan contributed to sample preparation and data collection. All authors revised and approved the final manuscript. Congyang Yi, Qian Liu, and Congle Zhu contributed equally to this work. All authors read and approved the final manuscript.
Corresponding authors
Ethics declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests. Fangpu Han is a guest editor of the Centromere structure and evolution collection for Genome Biology but was not involved in the editorial process of this manuscript.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Yi, C., Liu, Q., Zhu, C. et al. High-resolution genome assembly reveals retrotransposon-mediated centromere dynamics in rye. Genome Biol 26, 308 (2025). https://doi.org/10.1186/s13059-025-03792-3
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1186/s13059-025-03792-3