Stochastic Modeling of Single-Cell Gene Expression Adaptation Reveals Non-Genomic Contribution to Evolution of Tumor Subclones

Immunology Biochemistry Life Sciences Cell Biology Oncology Genetics Medicine
stochastic modeling gene expression adaptation tumor subclones Wnt signaling immunotherapy non-genomic contribution cellular evolution

Stochastic Modeling of Single-Cell Gene Expression Adaptation Reveals Non-Genomic Contribution to Tumor Subclone Evolution

Academic Background

Cancer is a complex disease whose progression is driven by the evolution of cells that are selected through adaptive advantages for growth. Traditionally, cancer evolution research has focused primarily on genetic mutations, considering them the main driver of tumor evolution. However, increasing evidence suggests that non-genetic factors, such as epigenetic changes and gene expression variations, also play a significant role in cancer evolution. Adaptive changes in gene expression can directly influence cellular functions, and uncovering these changes can enhance our understanding of selective pressures in tumor evolution, providing a foundation for designing more effective treatment protocols. However, due to technological limitations, especially in bulk sequencing studies that struggle to distinguish cell-level expression changes from variations in cell population composition, single-cell RNA sequencing (scRNA-seq) offers new opportunities to study the evolution of gene expression.

The research team in this study constructed a stochastic model to analyze the adaptive evolution of gene expression in mouse melanoma subclones, revealing its relationship with tumor phenotypes, particularly in response to immunotherapy. By integrating single-cell gene expression data and mutation-based phylogenetic trees, the researchers identified adaptive gene expression patterns associated with different phenotypes, especially those related to the Wnt signaling pathway, providing novel insights into the evolution of tumor subclones.

Source of the Paper

The paper was co-authored by M.G. Hirsch, Soumitra Pal, Farid Rashidi Mehrabadi, and other researchers from multiple renowned institutions, including the National Library of Medicine (NLM) and the National Cancer Institute (NCI) at the National Institutes of Health (NIH), among others. The study was published on January 15, 2025, in the journal Cell Systems, titled “Stochastic Modeling of Single-Cell Gene Expression Adaptation Reveals Non-Genomic Contribution to Evolution of Tumor Subclones”, with the DOI 10.1016/j.cels.2024.11.013.

Research Process and Results

Research Process

  1. Data Generation and Subclone Isolation
    The researchers first isolated 24 single cells from a highly heterogeneous melanoma cell line (B2905) and expanded them into independent subclonal lines through culture. These subclonal lines represent subclones derived from a common origin of melanoma cells. Subsequently, whole-exome sequencing (WES) and single-cell RNA sequencing (scRNA-seq) were performed on these subclones to obtain gene expression data and mutation information. Based on the mutation data, the researchers constructed a mutation-based phylogenetic tree for subsequent gene expression evolution analysis.

  2. Phenotypic Trait Analysis
    Each subclone was implanted into genetically identical mice, and their in vivo growth kinetics and response to immunotherapy (anti-CTLA4 treatment) were recorded. Based on growth rates and immunotherapy responsiveness, the researchers divided the subclones into three groups: highly aggressive and resistant (HA-R), secondarily aggressive and sensitive (SA-S), and mixed aggression and sensitive (MA-S).

  3. Stochastic Model Construction and Adaptive Gene Expression Analysis
    The researchers used the Ornstein-Uhlenbeck (OU) process (a stochastic model) to simulate gene expression evolution. The model assumes that changes in gene expression may result from neutral evolution, constrained evolution, or adaptive evolution. Using the Evogenex software, the researchers analyzed the gene expression data for each subclone group to identify genes with adaptive expression in specific subclone groups.

  4. Functional Analysis and Validation
    The identified genes with adaptive expression were subjected to KEGG pathway enrichment analysis to determine their biological functions. Additionally, the researchers validated the role of these genes in treatment response by implanting the parental melanoma cell line into mice and subjecting them to immunotherapy.

Key Results

  1. Adaptive Evolution of Gene Expression
    The research team identified 812, 1277, and 616 genes with adaptive expression in the HA-R, SA-S, and MA-S subclone groups, respectively. These gene expression patterns were closely associated with the phenotypes of the subclones. For example, genes in the HA-R subclones were primarily related to cell invasion and non-canonical Wnt signaling, while genes in the SA-S subclones were associated with cell proliferation and canonical Wnt signaling.

  2. Functional Enrichment
    Downregulated genes in the HA-R subclones and upregulated genes in the SA-S subclones were enriched in ribosome-related pathways, which are associated with cell growth. Downregulated genes in the SA-S subclones were enriched in the Rap1 signaling pathway and bacterial invasion of epithelial cells, pathways associated with cell migration.

  3. Validation Experiments
    Analysis of tumors from the parental melanoma cell line following immunotherapy revealed significant overlap between genes with expression changes and those identified as adaptively expressed, further validating the reliability of the results.

Conclusions and Significance

The findings demonstrate that adaptive evolution of gene expression plays a critical role in the phenotypic differentiation of mouse melanoma subclones. By integrating single-cell RNA sequencing data and mutation-based phylogenetic trees, the researchers were able to identify gene expression patterns associated with subclone phenotypes, particularly the roles of different branches of the Wnt signaling pathway in tumor proliferation and invasion. This discovery not only provides new insights into the non-genetic mechanisms of tumor evolution but also offers potential targets for designing personalized treatment strategies tailored to specific phenotypes.

Furthermore, the stochastic model developed by the research team provides a novel method for analyzing gene expression evolution in single-cell data, overcoming the limitations of traditional differential expression analysis, and holds broad applicability.

Research Highlights

  1. Innovative Methodology: This study is the first to apply the Ornstein-Uhlenbeck process to single-cell cancer data, successfully identifying adaptive evolutionary patterns in gene expression.
  2. Revelation of Non-Genetic Mechanisms: The research highlights the importance of non-genetic factors in tumor evolution, particularly the role of gene expression changes in subclone phenotypic differentiation.
  3. Potential Applications: By identifying gene expression patterns associated with immunotherapy responsiveness, the study provides new insights into personalized cancer treatment.

Additional Valuable Information

The research team also explored the role of epigenetic regulation in the adaptive evolution of gene expression, particularly the expression changes in histone modification enzyme genes identified in the HA-R subclones, offering directions for future related research.