Epigenomic Heterogeneity of Non-Functional Pancreatic Neuroendocrine Tumors Uncovered by Single nucleus and Spatial ATAC Profiling

Dejiang Wang, Xiangjun Di, Fei Gao, Guangqiang Li, Liang‐In Lin, Shaohui He, Di Zhang, Junjielong Xu, Jin Yu, Yan Liang, Michael Cecchini, Jill Lacy, John W. Kunstman, Pamela L. Kunz, Jing DubioRxiv 2025 · DOI

TL;DR

This study uses single-nucleus and spatial ATAC-seq to map the epigenetic landscape of non-functional pancreatic neuroendocrine tumors (NF-PanNETs), revealing how chromatin accessibility drives tumor progression and is influenced by the tumor's spatial microenvironment. The findings highlight distinct epigenetic states associated with tumor grade and identify specific tumor-stroma interactions that promote either proliferation or invasion, offering new avenues for therapeutic targeting.

What Problem Does This Paper Address?

Non-functional pancreatic neuroendocrine tumors (NF-PanNETs) are a heterogeneous group of cancers that often present with advanced disease and poor prognosis. While genetic mutations are known to play a role, the epigenetic mechanisms and spatial organization that drive their progression and heterogeneity remain poorly understood. Existing bulk ATAC-seq studies provide an averaged view, masking crucial cellular and spatial variations. Single-cell ATAC-seq (scATAC-seq) offers single-cell resolution but loses spatial context. This paper aims to bridge this gap by integrating snATAC-seq and spatial ATAC-seq to provide a high-resolution, spatially resolved epigenomic map of NF-PanNETs, uncovering the regulatory programs that govern their development and microenvironment.

Methods

The researchers employed a multi-pronged approach combining single-nucleus ATAC-seq (snATAC-seq) and spatial ATAC-seq.

  • snATAC-seq: This technique was used to profile chromatin accessibility at the single-nucleus level across multiple NF-PanNET samples. This allowed for the identification of distinct cell types (tumor cells, immune cells, fibroblasts, exocrine cells) and their associated regulatory programs.
  • Spatial ATAC-seq: This cutting-edge technology maps chromatin accessibility directly within intact tissue architecture. This is crucial for understanding how the spatial context and interactions between different cell types (e.g., tumor cells and stromal cells) influence epigenetic states and cellular behavior.
  • Bioinformatics Analysis: A suite of computational tools was used for data processing, including read alignment, peak calling, cell type annotation (based on gene activity scores of known markers), and differential chromatin accessibility analysis. Techniques like UMAP and LSI were employed for dimensionality reduction and visualization. TF activity was inferred using chromVAR, and motif enrichment analysis was performed to identify key regulatory factors. Pseudotime trajectory analysis was used to reconstruct the evolutionary path of tumor progression.

Datasets

The study utilized primary tumor specimens from 12 untreated NF-PanNET patients.

  • snATAC-seq: Performed on 9 tumor specimens.
  • Spatial ATAC-seq: Performed on 12 tumor specimens, with 6 matched to the snATAC-seq cohort.

Key Insights and Ideas

  • Epigenetic roadmap of tumor progression: The study reveals a stepwise, TF-driven trajectory of NF-PanNET progression, moving from lineage fidelity in low-grade tumors to proliferative reprogramming in intermediate grades, and finally to stemness-associated regulatory dominance in high-grade disease.
  • Spatial context is critical: Cellular behavior in NF-PanNETs is not solely determined by intrinsic epigenetic states but is profoundly influenced by spatial context and interactions with the tumor microenvironment.
  • Distinct tumor-stroma niches: Two key ecological niches were identified: a "proliferative niche" where tumor cells interact with myofibroblastic CAFs (myCAFs) and activate MYC/mTOR signaling, and an "invasive niche" where tumor cells are spatially segregated from myCAFs and exhibit KRAS signaling and EMT-associated programs.
  • Intra-tumor heterogeneity: Beyond inter-tumor differences, the study found significant intra-tumor heterogeneity, with distinct tumor subtypes co-existing within the same tumor mass, exhibiting complementary functional programs.
  • Novel biomarkers for stratification: The identified TF signatures and spatial epigenetic patterns offer potential for molecular stratification, prognosis prediction, and targeted therapeutic interventions.

Key Results

  • Cell type identification: snATAC-seq successfully identified and annotated distinct cell populations within NF-PanNETs, including tumor cells, immune cells, fibroblasts, and exocrine cells.
  • Grade-specific TF activity: Grade 1 tumors showed enrichment for lineage-defining TFs (HNF1A, HNF4A, FOXA2), while Grade 2 tumors exhibited increased activity of proliferative regulators (E2F family, AP-1 complex). Grade 3 tumors showed a sharp increase in YAP1 activity.
  • Spatial niches identified: Spatial ATAC-seq revealed a proliferative niche associated with MYC/FOX family TFs and an invasive niche enriched for Snail family TFs and KRAS pathway activity.
  • CAF heterogeneity: Five distinct CAF subpopulations were identified, with specific subtypes linked to EMT, mechanobiology, and immune modulation. High-grade tumors showed enrichment of pro-metastatic TFs in CAFs.
  • Immune landscape characterization: The study detailed the chromatin landscape of intratumoral immunity, identifying distinct myeloid and lymphoid compartments and their functional roles.
  • Intra-tumor spatial subtypes: In samples like PNET8 and PNET9, distinct tumor subtypes were found spatially associated with myCAFs (proliferative) or spatially separated (KRAS-driven).

Limitations and Future Work

  • Sample Size: The authors acknowledge the relatively small sample size and limited representation of Grade 3 tumors as a limitation.
  • Future Directions:
    • Multi-modal data integration: Future research should focus on integrating chromatin accessibility data with spatial transcriptomics, proteomics, and metabolomics for a comprehensive view of the tumor ecosystem.
  • Therapeutic targeting: The identified TF networks (e.g., YAP1-HIPPO, FOXA2-HNF, SNAIL-EMT) present compelling opportunities for targeted intervention through pharmacological inhibition or transcriptional modulation.
  • Spatially informed precision oncology: The spatial epigenetic information can inform surgical margins, guide biopsy site selection, and enable targeted delivery of therapeutics to aggressive tumor compartments.

Summary updated

Abstract

Non-functional pancreatic neuroendocrine tumors (NF-PanNETs) account for the majority of neuroendocrine neoplasms arising in the pancreas and exhibit substantial clinical and biological heterogeneity, yet their epigenetic regulation and spatial architecture remain poorly understood. Here, we present an integrative study of NF-PanNETs across multiple tumor grades using single-nucleus ATAC-seq (snATAC-seq) and spatial ATAC-seq. snATAC-seq delineates the chromatin accessibility landscapes of distinct tumor subtypes, immune cells, and cancer-associated fibroblasts (CAFs), revealing key transcription factor (TF) programs that drive tumor progression and shape microenvironmental interactions. Spatial ATAC-seq further identifies two distinct tumor-stroma ecological niches: a proliferative niche marked by MYC and FOX family, and an invasive niche enriched for Snail family TFs and KRAS pathway activity. These findings demonstrate that cellular behavior in NF-PanNETs is governed not only by intrinsic epigenetic states but also by spatial context. Together, our study provides a spatially resolved epigenomic framework for dissecting NF-PanNET heterogeneity and evolution, offering new biomarkers and regulatory axes for molecular stratification and precision therapy.
BibTeX
@misc{wang2025epigenomic,
  author = {Dejiang Wang and Xiangjun Di and Fei Gao and Guangqiang Li and Liang‐In Lin and Shaohui He and Di Zhang and Junjielong Xu and Jin Yu and Yan Liang and Michael Cecchini and Jill Lacy and John W. Kunstman and Pamela L. Kunz and Jing Du},
  title = {{Epigenomic Heterogeneity of Non-Functional Pancreatic Neuroendocrine Tumors Uncovered by Single nucleus and Spatial ATAC Profiling}},
  year = {2025},
  doi = {10.1101/2025.09.11.675640},
  howpublished = {bioRxiv},
}
Figure