A high-resolution 3D genome map of kiwifruit provides insights into chromatin architecture and transcriptional activity

Shuangling Xie, Tong Li, Jing Yang, Jingrui Wang, Jinli Gong, Juan Wang, Xiaoli Hu, Xiaolong Li, Xuepeng SunHorticulture 2026 · DOI

TL;DR

Researchers have created the first high-resolution 3D genome map of kiwifruit, integrating it with epigenomic and transcriptomic data. This map reveals how the kiwifruit genome is organized in leaf and fruit tissues, showing broad conservation but also subtle, tissue-specific variations that offer insights into transcriptional regulation in a perennial woody crop.

What Problem Does This Paper Address?

Understanding how a cell's DNA is physically organized in three dimensions (3D genome architecture) is crucial for deciphering how genes are turned on and off (transcriptional regulation). While we have good 3D genome maps for model organisms and annual crops, such detailed resources are scarce for perennial woody plants, especially fruit crops. These plants have complex life cycles and developmental programs that might involve unique genome organization strategies. Kiwifruit, a commercially important woody perennial, presents an excellent opportunity to investigate these aspects. Previous studies hinted at conserved 3D genome features but lacked the resolution and multi-tissue integration needed to understand dynamic regulation and its link to epigenetics and gene expression. This paper aims to fill that gap by providing a comprehensive, tissue-resolved 3D genome atlas for kiwifruit.

Methods

The core of this study relies on in situ Hi-C sequencing, a standard technique for mapping DNA-DNA interactions within the nucleus. They performed this on both kiwifruit leaf and fruit tissues, generating high-resolution contact maps. To interpret these maps, they integrated them with a suite of other "omics" data:

  • Chromatin accessibility (ATAC-seq): Identifies regions of the genome that are more open and potentially active.
  • Whole-genome DNA methylation: Maps methylation patterns, which are often associated with gene silencing.
  • Seven histone modifications: These are chemical tags on histone proteins that package DNA, and they play key roles in regulating gene expression.
  • RNA-seq: Measures gene expression levels across various tissues and developmental stages.

They employed computational tools like HiCExplorer for data processing and normalization, and Calder for inferring hierarchical subcompartments within the genome. Mustache and hicDetectLoops were used for chromatin loop detection. Statistical analyses, including Pearson correlation, principal component analysis (PCA), and Jaccard similarity, were used to compare data and identify significant patterns.

Datasets

The study utilized the following datasets:

  • Newly generated Hi-C data: From kiwifruit (Actinidia chinensis cv. 'Hongyang') leaf and fruit tissues (three biological replicates each).
  • Newly generated CUT&Tag data: For seven histone modifications (H3K4me3, H3K4me, H3K9ac, H3K27ac, H3K36me3, H3K27me3, H3K9me3) from kiwifruit leaf tissue.
  • Newly generated ATAC-seq data: From kiwifruit leaf tissue.
  • Newly generated RNA-seq data: From multiple kiwifruit tissues (root, leaf, stem, flower) and 12 fruit developmental time points.
  • Existing whole-genome bisulfite sequencing (WGBS) data: For DNA methylation analysis.

The study also references publicly available datasets for comparison and validation, including those from pepper, human, and other plant species.

Key Insights and Ideas

  • Broad Conservation, Subtle Variation: The fundamental organization of the kiwifruit genome into A/B compartments, TAD-like domains, and chromatin loops is largely conserved between leaf and fruit tissues. However, at finer scales, there are quantitative differences in subcompartment organization, domain insulation, and loop detection. This suggests a stable underlying framework with dynamic fine-tuning for tissue-specific functions.
  • Subcompartments as Functional Units: The study identified hierarchical subcompartments within the broader A/B compartments. These subcompartments show strong correlations with specific histone modifications and gene expression levels, suggesting they represent functional units of genome organization.
  • TADs and Epigenetic Signatures: Kiwifruit TAD-like domains can be classified into "Active," "Intermediate," and "Heterochromatin-driven Folding" (HDF) types, each with distinct epigenetic profiles (e.g., active domains with active marks, HDF domains with repressive marks and TEs). This classification provides a framework for understanding how chromatin structure relates to gene activity and repression.
  • Fruit-Specific Genes in Repressive Contexts: Interestingly, many fruit-specific genes were found within B compartments and insulated domains in fruit tissues. This suggests that their specific expression might be regulated by being "kept quiet" in a generally repressive environment, rather than being actively promoted by open chromatin or specific loops.
  • Loops Play a Limited Role in Tissue-Specific Expression: While chromatin loops are detected, their overlap with differentially expressed genes or tissue-specific genes is relatively low. This implies that for many genes, transcriptional changes are more closely tied to broader domain-level organization rather than specific loop interactions.

Key Results

  • High-Resolution Maps: Generated Hi-C contact maps at resolutions as fine as 2-kb, enabling detailed analysis of genome architecture.
  • Subcompartment Diversity: Identified eight subcompartment classes, with an average length of ~225 kb at 40-kb resolution.
  • TAD-like Domain Conservation: Found that ~49% of TAD-like domains were shared between leaf and fruit tissues, indicating significant conservation.
  • Loop Numbers: Identified 1660 nonredundant loops in leaf tissue and 3117 in fruit tissue.
  • Fruit-Specific Gene Localization: 92.6% of identified fruit-specific genes were located within TAD-like domains in both tissues. Over 60% of these genes were in B compartments in fruit tissue.
  • Limited DEG-Loop Overlap: Found that only a small subset of differentially expressed genes (DEGs) overlapped with chromatin loops.

Limitations and Future Work

  • Tissue Specificity: The study focused on leaf and fruit tissues. Expanding to other tissues (e.g., root, stem, flower) and developmental stages would provide a more complete picture of kiwifruit's 3D genome dynamics.
  • Cell Type Heterogeneity: The current Hi-C and epigenomic data represent population averages. Single-cell Hi-C or Micro-C would be necessary to resolve cell-type-specific variations.
  • Algorithmic Dependence: The definition and strength of TADs and compartments can be influenced by algorithmic choices.
  • Experimental Validation: The study highlights correlative associations. Future work should involve experimental validation of the functional roles of specific chromatin structures (e.g., using CRISPR-based perturbations) in regulating gene expression.
  • Broader Perennial Crop Comparisons: The authors suggest that this work provides a framework for future comparative studies in other perennial crops, which would be valuable for understanding evolutionary aspects of 3D genome organization.

Summary updated

Abstract

Abstract Comprehensive three-dimensional (3D) genome maps are essential for understanding transcriptional regulation, yet such resources remain limited for perennial woody crops. Here, we present a high-resolution, tissue-resolved 3D genome atlas of kiwifruit (Actinidia chinensis). Using in situ Hi-C, we generated chromatin contact maps from leaf and fruit tissues and integrated these data with epigenomic and transcriptomic datasets, including chromatin accessibility, whole-genome DNA methylation, seven histone modifications, and RNA-seq profiles spanning multiple tissues and fruit developmental stages. This integrated dataset enables systematic annotation of genome architecture across multiple spatial scales, including A/B compartments, hierarchical subcompartments, TAD-like domains, and chromatin loops. Global features of 3D genome organization are broadly similar between tissues, while quantitative variation is observed at finer scales, such as subcompartment rank, domain insulation strength, and loop detection frequency. Integration with genomic and epigenomic features reveals consistent associations between chromatin states and spatial organization, providing a reference framework for interpreting plant genome architecture in a perennial context. We further map tissue-specific gene sets onto the 3D genome landscape and describe their spatial distributions relative to compartments, domains, and loop anchors, offering a view of how transcriptional programs relate to higher-order chromatin organization. Together, this work establishes an integrative, high-resolution 3D genome resources for a woody perennial fruit crop, and supports future functional, evolutionary, and applied research in kiwifruit and other perennial species.
BibTeX
@inproceedings{xie2026high,
  author = {Shuangling Xie and Tong Li and Jing Yang and Jingrui Wang and Jinli Gong and Juan Wang and Xiaoli Hu and Xiaolong Li and Xuepeng Sun},
  title = {{A high-resolution 3D genome map of kiwifruit provides insights into chromatin architecture and transcriptional activity}},
  booktitle = {Horticulture},
  year = {2026},
  doi = {10.1093/hr/uhag076},
}
Figure