Single-cell chromatin accessibility profiling of glioblastoma identifies an invasive cancer stem cell population associated with lower survival

Paul Guilhamon, Charles Chesnelong, Michelle Kushida, Ana Nikolić, Divya Singhal, Graham MacLeod, Seyed Ali Madani Tonekaboni, Florence M.g. Cavalli, Christopher Arlidge, Nishani Rajakulendran, Naghmeh Rastegar, Xiaoguang Hao, Rozina Hassam, Laura J. Smith, Heather Whetstone, Fiona J. Coutinho, Bettina Nadorp, Katrina Ellestad, H. Artee Luchman, Jennifer A. Chan, Molly S. Shoichet, Michael D. Taylor, Benjamin Haibe‐Kains, Samuel Weiss, Stéphane Angers, Marco Gallo, Peter B. Dirks, Mathieu LupieneLife 2021 · DOI

TL;DR

This study uses single-cell chromatin accessibility (scATAC-seq) to dissect the heterogeneity within glioblastoma (GBM) cancer stem cells (GSCs), identifying three distinct states: Reactive, Constructive, and Invasive. The "Invasive" GSC state is strongly associated with poorer patient survival and increased invasiveness, offering a potential new avenue for identifying high-risk patients and developing targeted therapies.

What Problem Does This Paper Address?

Glioblastoma (GBM) is a highly aggressive brain cancer with a dismal prognosis. A key challenge in treating GBM is the presence of cancer stem cells (GSCs), which are thought to drive tumor progression, recurrence, and resistance to therapy. While previous studies have identified different cellular states within GBM tumors using RNA sequencing, the functional and molecular underpinnings of GSC heterogeneity, particularly their role in invasiveness and patient outcome, remain poorly understood. This paper aims to characterize the functional diversity of GSCs at a single-cell level by examining their chromatin accessibility, a key determinant of gene expression and cellular identity, to identify GSC states that correlate with disease aggressiveness and survival.

Methods

The researchers employed single-cell ATAC-sequencing (scATAC-seq) to profile the accessible chromatin landscape of cells from four primary GBM tumors and 27 patient-derived GSC lines. This technique allows for the identification of regulatory elements that are open and thus potentially active in individual cells.

Key analytical steps included:

  • scATAC-seq Data Processing: Standard pipelines were used for read alignment, peak calling, and dimensionality reduction (UMAP).
  • Cell State Assignment: Cells were assigned to known GBM cellular states (NPC, OPC, AC, MES) based on chromatin accessibility enrichment scores for signature genes.
  • GSC Identification: Putative cancer stem cells were identified by focusing on chromatin accessibility at promoters of genes known to be involved in self-renewal and tumor propagation.
  • GSC State Clustering: Spectral clustering was applied to the ATAC-seq profiles of the 27 GSC lines to identify distinct GSC states.
  • Functional Characterization: Gene set enrichment analysis (GSEA) was used to identify biological processes and pathways associated with each GSC state. Motif enrichment analysis was performed to identify transcription factors driving these states.
  • In Vitro and In Vivo Assays: Invasion assays were conducted using collagen matrices to assess the migratory and invasive properties of different GSC states. Orthotopic xenografts in mice were used to evaluate the survival impact of GSC states.
  • TCGA Data Analysis: Publicly available TCGA GBM data was used to correlate GSC state signatures with patient survival.

Datasets

  • Primary GBM Tumors: scATAC-seq data from four adult IDH wild-type GBM tumors.
  • Patient-Derived GSC Lines: Bulk ATAC-seq data from 27 patient-derived GSC lines.
  • TCGA GBM Cohort: RNA-seq and microarray data from 144 IDH wild-type GBM patients for survival analysis.

Key Insights and Ideas

  • Chromatin Accessibility as a GSC Discriminator: The study demonstrates that scATAC-seq is a powerful tool for resolving heterogeneity within the GSC population, going beyond what RNA-seq alone can reveal.
  • Three Distinct GSC States: The research uncovers three functionally distinct GSC states—Reactive, Constructive, and Invasive—each characterized by unique chromatin accessibility profiles and associated transcription factor dependencies.
  • The Invasive State and Poor Prognosis: The "Invasive" GSC state is not only functionally more aggressive in vitro and in vivo but is also directly linked to significantly lower patient survival in both xenograft models and TCGA data. This suggests that the proportion and characteristics of invasive GSCs within a tumor could be a critical prognostic marker.
  • Transcription Factor Signatures: Each GSC state is governed by specific sets of transcription factors, highlighting potential therapeutic targets. For instance, SP1 is crucial for Reactive GSCs, while FOXD1 is uniquely essential for Invasive GSCs.
  • Intra-Tumor Heterogeneity: The presence of these distinct GSC states within individual primary GBM tumors underscores the complex cellular landscape of GBM and suggests that therapeutic strategies may need to target multiple GSC populations.

Key Results

  • scATAC-seq successfully identified distinct chromatin accessibility modules across four primary GBM tumors, with some modules reflecting known GBM cellular states.
  • Clustering of 27 patient-derived GSC lines based on ATAC-seq data revealed three distinct GSC states: Reactive, Constructive, and Invasive.
  • The Invasive GSC state exhibited significantly higher invasiveness in vitro compared to Reactive and Constructive states.
  • Orthotopic xenografts showed that mice injected with Invasive GSCs had significantly shorter survival times compared to those injected with Reactive or Constructive GSCs.
  • Analysis of TCGA data revealed that a signature derived from the Invasive GSC state was predictive of lower patient survival, particularly in an "Invasive-high" subgroup.
  • Each GSC state was associated with enrichment of specific transcription factor families and essential genes, suggesting state-specific vulnerabilities.

Limitations and Future Work

  • Cohort Size: While the study analyzed multiple tumors and GSC lines, the number of samples within each identified GSC state (especially the Invasive state) was relatively small, particularly for the xenograft survival analysis. Larger cohorts would strengthen the statistical power of these findings.
  • GSC State Assignment in Primary Tumors: The scATAC-seq analysis of primary tumors showed that Reactive and Invasive GSCs were present in much lower proportions than Constructive GSCs. Further validation is needed to confirm the precise representation and functional significance of these rarer states in situ.
  • Therapeutic Targeting: The paper identifies potential therapeutic targets (e.g., SP1, FOXD1) and suggests combination therapies. Future work would involve testing the efficacy of targeting these specific GSC states in preclinical models and eventually in clinical trials.
  • Translational Potential: The identification of an "Invasive GSC signature" that predicts survival is a significant step towards clinical application. Further validation and refinement of this signature for patient stratification are warranted.

Summary updated

Abstract

Chromatin accessibility discriminates stem from mature cell populations, enabling the identification of primitive stem-like cells in primary tumors, such as glioblastoma (GBM) where self-renewing cells driving cancer progression and recurrence are prime targets for therapeutic intervention. We show, using single-cell chromatin accessibility, that primary human GBMs harbor a heterogeneous self-renewing population whose diversity is captured in patient-derived glioblastoma stem cells (GSCs). In-depth characterization of chromatin accessibility in GSCs identifies three GSC states: Reactive, Constructive, and Invasive, each governed by uniquely essential transcription factors and present within GBMs in varying proportions. Orthotopic xenografts reveal that GSC states associate with survival, and identify an invasive GSC signature predictive of low patient survival, in line with the higher invasive properties of Invasive state GSCs compared to Reactive and Constructive GSCs as shown by in vitro and in vivo assays. Our chromatin-driven characterization of GSC states improves prognostic precision and identifies dependencies to guide combination therapies.
BibTeX
@inproceedings{guilhamon2021single,
  author = {Paul Guilhamon and Charles Chesnelong and Michelle Kushida and Ana Nikolić and Divya Singhal and Graham MacLeod and Seyed Ali Madani Tonekaboni and Florence M.G. Cavalli and Christopher Arlidge and Nishani Rajakulendran and Naghmeh Rastegar and Xiaoguang Hao and Rozina Hassam and Laura J. Smith and Heather Whetstone and Fiona J. Coutinho and Bettina Nadorp and Katrina Ellestad and H. Artee Luchman and Jennifer A. Chan and Molly S. Shoichet and Michael D. Taylor and Benjamin Haibe‐Kains and Samuel Weiss and Stéphane Angers and Marco Gallo and Peter B. Dirks and Mathieu Lupien},
  title = {{Single-cell chromatin accessibility profiling of glioblastoma identifies an invasive cancer stem cell population associated with lower survival}},
  booktitle = {eLife},
  year = {2021},
  doi = {10.7554/elife.64090},
}
Figure