Nutmeg and Ylang-Ylang Genomes Reveal How Plants Rebuilt the Centromere From Scratch

Centromeres are the unsung heroes of cell division. In the overwhelming majority of plants and animals, each chromosome funnels all of its centromeric activity into one discrete region, the narrow waist where spindle fibers grab and pull duplicated strands apart. Yet a handful of evolutionary mavericks have abandoned this design altogether. In certain members of the order Magnoliales — the botanical lineage that gives us nutmeg, magnolias and ylang-ylang — centromeric function is smeared along the entire length of every chromosome, a configuration known as holocentricity. Understanding how such a radical reorganization of the genome’s architecture could evolve has long frustrated biologists, because most holocentric species sit on distant branches of the tree of life, far from convenient monocentric relatives.

A new study offers the clearest view yet of that transition. A research team from Northeast Forestry University and BGI Research assembled chromosome-level genomes for two closely related Magnoliales species with opposite centromere designs: Myristica fragrans, the nutmeg tree, whose chromosomes are holocentric, and Cananga odorata, the ylang-ylang, whose chromosomes follow the classic monocentric plan. By comparing the two assemblies with an integrated battery of genomic, epigenomic and phylogenetic analyses, the researchers were able to isolate the molecular changes that accompany the shift from a single centromere per chromosome to a distributed one. The work was published in the Journal of Systematics and Evolution.

The scale of the assemblies alone represents a technical achievement. The team generated a 637-million-base-pair genome for M. fragrans spanning 22 chromosomes and a 717-million-base-pair genome for C. odorata spanning 8 chromosomes, with completeness scores between 97.5 and 98.6 percent on the BUSCO benchmark of conserved single-copy genes. Hi-C contact maps, which chart how chromatin folds in three-dimensional space, confirmed that each C. odorata chromosome carries a single prominent heterochromatic block — the cytological signature of a conventional centromere. M. fragrans, by contrast, showed no such concentrated signal. Two independent computational centromere-prediction tools independently recapitulated the pattern, calling one centromeric locus per chromosome in the ylang-ylang genome and multiple loci per chromosome in nutmeg.

Interestingly, the regions associated with centromeric function in both species were not chemically alien to one another. In each genome, centromere-associated regions correlated positively with methylation of cytosines in the CG and CHG sequence contexts, and both fell within the so-called B compartments of chromatin, the heterochromatic zones that are generally transcriptionally silent. Whatever else changed during the holocentric transition, the deep epigenetic identity of centromeric DNA appears to have been conserved. The differences emerged elsewhere in the chromosome landscape, and they were dramatic.

Transposable elements — the jumping genes that make up a large fraction of most plant genomes — turned out to be the most conspicuous point of divergence. C. odorata displayed tightly clustered heterochromatin and a dominant plateau of LTR/Gypsy-family retrotransposons, including the Tekay, Athila and SIRE lineages, heaped around its concentrated centromeric regions. M. fragrans instead showed a dispersed chromatin organization studded with multiple discrete peaks of LTR/Copia elements, particularly the Ale and SIRE families. The methylation mark mCHH, which in plants typically fortifies heterochromatin against transposon activity, was depleted within the centromeric regions of nutmeg but enriched around those of ylang-ylang — a mirror-image pattern in the epigenetic armor of two sister lineages.

Tandem repeat satellite DNA, long regarded as the canonical stuff of centromeres, told a similarly divergent story. The two major tandem-repeat families in the nutmeg genome, designated MF_351 and MF_269, preferentially localized to the chromosome termini, where they carried elevated levels of CG and CHG methylation — a positioning that suggests they may contribute to structural stability at the chromosome ends rather than to a concentrated centromere. In ylang-ylang, tandem repeats accumulated predominantly in euchromatic, gene-rich regions, with negligible occupancy of heterochromatin. In other words, holocentricity in Myristica did not arise simply by enlarging a conventional centromere; the entire repeat landscape was rearranged.

Detective work at the evolutionary timescale helped rule out the most obvious suspect. Phylogenomic analysis and comparisons of synonymous substitution rates, so-called Ks analyses, showed that the two species share an ancient whole-genome duplication, dubbed LAURb, that occurred roughly 115 million years ago. Crucially, that duplication is shared, meaning it predates the split between the holocentric and monocentric lineages and cannot by itself explain the transition. The team also found evidence that M. fragrans and the related species M. yunnanensis experienced two distinct waves of transposon expansion, pointing instead to repeat dynamics as a lineage-specific driver.

The proteins that physically attach chromosomes to the spindle machinery offered few dramatic answers. CENH3, the core kinetochore histone that epigenetically defines centromeric identity, is highly conserved between the two species, differing only in subtle lineage-specific substitutions within the centromere-recognition loop 1 and the flexible N-terminal tail. An elongated N-terminus in NASP, another centromere-associated protein, turned out to be a Magnoliales-wide feature rather than a signature of holocentricity, and the researchers detected no large-scale structural changes in any of the other core kinetochore proteins. The machinery of chromosome segregation, in short, appears essentially intact on both sides of the divide.

The synthesis that emerges is one of lineage-specific rewiring rather than a single master event. According to first author Jinyin Xing, comparing a holocentric species with a closely related monocentric one at chromosome-level resolution was precisely what allowed the team to dissect the genomic changes underlying what he calls this remarkable reorganization. The evidence converges on remodeling of repetitive DNA and epigenetic states as the proximate cause of the centromere’s dispersion in Myristica, while whole-genome duplication and major kinetochore protein evolution can be largely excluded as sufficient explanations.

Beyond resolving a longstanding puzzle in chromosome biology, the study delivers a resource with practical and conceptual reach. Chromosome-level reference genomes for nutmeg and ylang-ylang — both economically and culturally significant species — will support breeding, conservation and comparative work in a lineage that anchors some of the earliest branches of flowering plant evolution. More broadly, the comparative framework demonstrates how holocentricity, one of the most striking alternative solutions to the problem of chromosome inheritance, can emerge through the slow accumulation of repeat and chromatin changes rather than through a single evolutionary leap. For a structure as fundamental as the centromere, that lesson — that even the most essential genomic architecture is negotiable over deep time — is a reminder of how inventive evolution can be with the rules of life.

Subject of Research: Comparative genomics of monocentric-to-holocentric centromere transition in Magnoliales plants

Article Title: Chromosome-level genomes of two Magnoliales species provide new insights into the holocentric transition

Article References: Chromosome-level genomes of two Magnoliales species provide new insights into the holocentric transition. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: centromere evolution, holocentric chromosomes, Magnoliales, comparative genomics, Myristica fragrans, Cananga odorata, transposable elements, DNA methylation, CENH3, whole-genome duplication, genome assembly, Journal of Systematics and Evolution