henhao
Sign in
Evolutionary BiologybioRxiv
Preprint

Dynamic evolution of EZHIP, an inhibitor of the Polycomb Repressive Complex 2 in mammals

Pravrutha RamaniD, Hana Khan, Janet M. YoungiD, Rebecca Ferreira Alves, Maria Toro Moreno, Alice L. Li, Alice H. BergeriD, Toshio TsukiyamaiD, Harmit Singh MalikiD

Abstract

The Polycomb Repressive Complex 2 (PRC2) is an ancient, conserved chromatin-interacting complex that controls gene expression, facilitating differentiation and cellular identity during development. Its regulation is critical in most eukaryotes. EZHIP was recently characterized as a PRC2 inhibitor and oncohistone mimic in mammals. Although EZHIP expression is typically restricted to the germline, its aberrant expression in pediatric brain tumors inhibits PRC2-mediated H3K27 methylation, driving disease progression. To gain a deeper understanding of its normal functions, we systematically examined EZHIP evolution across mammals using comparative genomics, synteny analyses, and motif discovery. Extending previous work, we find that EZHIP originated on the X chromosome and was retained there in most placental mammals, except in Afrotheria. In addition to the highly conserved H3K27M-like histone mimic motif, our motif analyses reveal six previously unidentified EZHIP motifs, including a putative nuclear localization signal, a serine-enriched region, and tandem repeats that are largely well-conserved in placental mammals. We hypothesize that these motifs are also critical to EZHIPs functions, including in PRC2 interaction and inhibition. We show that EZHIP has evolved under strong diversifying selection in primates and underwent dynamic expansions and losses across species. Some paralogs, such as EZHIP2 in primates, also evolved under positive selection. Based on its evolutionary attributes and germ-cell expression, we propose that EZHIP arose and evolved rapidly due to inter-parental conflict over fetal development in utero in placental mammals. Our work provides a foundation for further investigations into EZHIPs multifaceted roles in mammalian reproduction and disease.

PDFDOICode
Full-text digest ↓

02 /Specification & Classification

Click any tag to browse matching papers
Citations
2
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Opus 5.5 · xhigh effort · 7 Nov 2026 · from the full text of version 3 (17 Oct 2026) · CC BY 4.0

EZHIP, a PRC2 inhibitor, arose on the placental mammal X chromosome, evolves under positive selection in primates, and duplicated, including a testis-expressed primate paralog, EZHIP2, whose gorilla copy inhibits H3K27me3.

EZHIP is well studied as a driver in cancers, but its normal germline roles, origin and divergence were unclear, and its orthologs were thought to share only a 14-amino-acid KLP motif. This work traces EZHIP across more than 80 mammalian genomes and finds additional conserved motifs, repeats, positive selection and recurrent duplications.

Key results

  1. ShownEZHIP originated in the last common ancestor of placental mammals and sits at a conserved X-linked syntenic locus. It was not found in opossum or platypus. Many Afrotherian species have lost it or lack an intact KLP.

    Searches of more than 80 placental genomes plus opossum and platypus; synteny checked in 15 placental mammals with flanking genes BMP15, NUDT10/11 and MAGED1. No KLP match anywhere in the elephant genome; EZHIP in five Afrotherian species (tenrecs, golden moles) is under purifying selection. · EZHIP arose and has been retained in placental…; Identification of EZHIP orthologs

  2. ShownBesides the KLP (inside motif 9), EZHIP carries further conserved motifs, and tandem repeats with a ~33 bp ancestral unit. Repeat copy number varies widely across species and even within humans.

    MEME found 10 motifs, all E-value < 10^-5; seven occur once per protein (1-4, 8-10). Repeat copies range from 3 in some bats and carnivores to 21 in some ungulates; in gnomAD the common human allele has 6 repeats, others 5. · EZHIP orthologs encode seven well-conserved…; Motif analyses

  3. ShownEZHIP diverges far faster than its partner EZH2 and shows site-specific positive selection in simian primates, despite weak purifying selection across the whole gene.

    Primate, ungulate and carnivore EZHIP orthologs reach ~60% identity, versus >90% for primate EZH2; rodent EZHIP drops to 40% over 20 MY versus >98% for rodent EZH2. Whole-gene dN/dS ~0.85; in 25 simian primates PAML puts 12.2% of sites at average dN/dS 4.08. · EZHIP has evolved under diversifying selection

  4. ShownEZHIP has duplicated recurrently. Mouse has 15 X-linked paralogs, and two paralogs were retained for long periods: a Caniformia paralog (~60 million years) and the autosomal primate EZHIP2 (~40 million years).

    Of the 15 mouse paralogs, 4 are pseudogenes and 11 are intact but lack a KLP. EZHIP2 arose in the Haplorhini ancestor on human chromosome 5, apparently by retrotransposition; it is lost or pseudogenized in bonobo/chimpanzee and marmoset/squirrel monkey. · EZHIP has been recurrently duplicated in…

  5. ShownEZHIP2 is expressed only in testis and shows signs of selection. Gorilla EZHIP2 lowers H3K27me3 in HEK293T cells, whereas the truncated human EZHIP2 does not.

    Public RNA-seq shows EZHIP2 in testes only, human and rhesus. In 17 non-human simian primates dN/dS is 1.35 (p=0.02 versus neutral), with 11.3% of codons at dN/dS 4.97. Western blots: ΔKLP or M298K restores H3K27me3 for gorilla EZHIP2. · EZHIP has been recurrently duplicated in…; RNA-seq analyses; Western blot analyses

Methods, data and reproducibility

Homology and synteny searches identified EZHIP orthologs and paralogs across mammalian genomes. These were analysed with de novo motif discovery, tandem repeat detection, maximum-likelihood phylogenies and PAML/FUBAR selection tests. Public RNA-seq quantified expression, and transgene expression in HEK293T cells was checked by H3K27me3 Western blot.

  • Genome assemblies of more than 80 placental mammals plus opossum and platypus (searched for EZHIP orthologs and paralogs) · multiple mammals · n = more than 80 placental species plus 2 outgroups
  • Public RNA-seq of human and rhesus macaque tissues, human spermatogenesis and human embryos (Table S5) · human; rhesus macaque
  • TCGA tumor and normal RNA-seq BAM files (EZHIP2 counting) · human · n = n ≈ 11,000
  • HEK293T cells expressing human EZHIP, gorilla EZHIP2 or human EZHIP2, with KLP deletion or M-to-K mutants · human (cell line)
  • CCLE cancer cell line expression and gnomAD human repeat alleles · human
Methods & tools
TBLASTN, MUSCLE, PhyML, MEME suite (MEME, MAST, FIMO), HMMER 3.4, Tandem Repeat Finder, PAML codeml, HyPhy (FUBAR, GARD), STAR, featureCounts, AlphaFold, TimeTree
Code
https://github.com/jayoung/EZHIP_evolution
Software & environment
Geneious Prime 2025.0.3; HMMER 3.4; Subread featureCounts 2.0.3-GCC-11.2.0.
Reproducibility gaps
Supplementary Data files hold the sequences, but no download URL or accession is printed in the text.
Limitations (5)
  • AlphaFold gives very low confidence for EZHIP, so no testable structural hypotheses follow from it. · stated by the authors
  • How loss of the newly found motifs changes EZHIP function is unknown, and the evolutionary origin of EZHIP and the KLP remains unresolved. · stated by the authors
  • GARD gave weak evidence for recombination in EZHIP (evidence ratio 85.4); selection results are reported only for the full-length alignment, excluding repetitive regions. · stated by the authors
  • The parental-conflict model for EZHIP is not directly tested here; it rests on parallels with short H2A variants and on sequence evolution. · noted in this summary
  • PRC2 inhibition was tested only by overexpression in one cell line (HEK293T), with H3K27me3 Western blots and no quantification reported in the text. · noted in this summary

Who should read:Read this if you study PRC2 regulation, germline epigenetics or rapidly evolving gene families, or want conserved EZHIP motifs and paralog catalogues across mammals.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISRagazzini '19Yette '20Jain '19Krabbenhoft '25Jain '20Sharaf '21
68% semantic matchDevelopmental Biology
Roberta Ragazzini, Raquel Pérez-Palacios, Irem H. Baymaz+15 more·Nature Communications·Aug 26, 2019· 107 citations