Dynamic changes of genomic and transcriptomic tumor diversity during melanoma progression

Ariane L. Moore, Christos Dimitrakopoulos, Stefanie Flückiger-Mangual, Lukas Frischknecht, Christian Beisel, Wilhelm Krek, Niko BeerenwinkelbioRxiv 2021 · DOI

TL;DR

This study uses a mouse model to track melanoma progression and metastasis over time, integrating genomic and transcriptomic data to identify key signaling pathways and genes involved. The research highlights the increasing genomic and transcriptomic diversity of tumors as they progress, identifying Vegfc and Rock1 as crucial players in metastasis and suggesting potential new therapeutic targets.

What Problem Does This Paper Address?

Metastatic melanoma is a deadly cancer that is notoriously difficult to treat. While we know that genetic mutations drive melanoma, understanding how tumors evolve and become metastatic over time, especially in the absence of external mutagens like UV radiation, is crucial for developing better therapies. This paper aims to dissect the dynamic genomic and transcriptomic changes that occur during melanoma progression and metastasis, providing a more comprehensive picture of tumor evolution.

Methods

The researchers employed a multi-omics approach, combining whole-exome sequencing (WES) and RNA sequencing (RNA-seq) on melanoma samples collected at different time points from a BrafV600E-driven, Pten-deficient mouse model. Key techniques and analyses included:

  • Time-course sampling: Tumors were collected from seven mice at seven different time points, allowing for the tracking of changes over time.
  • Genomic analysis: WES was used to identify single-nucleotide variants (SNVs), indels, and copy number variants (CNVs).
  • Transcriptomic analysis: RNA-seq was used to measure gene expression levels.
  • Network-based integration: The NetICS tool was used to integrate genomic and transcriptomic data, identifying "mediator genes" that link genetic alterations to gene expression changes. This approach helps to understand how mutations propagate through signaling pathways.
  • Principal Component Analysis (PCA): Used to visualize the relationships between samples based on gene expression.
  • Differential Gene Expression Analysis: To identify genes that are significantly up- or down-regulated between different sample groups (e.g., primary vs. metastatic, early vs. late stage).
  • Pathway Overrepresentation Analysis: To identify biological pathways that are significantly enriched among the differentially expressed genes.
  • Genomic Diversity Measurement: A metric was adopted from ecology to quantify the genetic divergence between samples within a mouse.

Datasets

The study utilized data generated from a BrafV600E-driven, Pten-deficient mouse model. Specifically, they collected:

  • Seven primary melanoma tumor samples.
  • Nine metastasis samples from lymph nodes.
  • One normal tail sample as a control.

All samples underwent whole-exome sequencing (WES) and RNA sequencing (RNA-seq). The generated WES and RNA-seq data are available at the European Nucleotide Archive under accession number PRJEB28651.

Key Insights and Ideas

  • Tumor diversity increases over time: Even without external mutagens, melanoma tumors become genetically more diverse as they progress. This heterogeneity is a major challenge for treatment.
  • Vegfc as a key signaling hub: The gene Vegfc, involved in lymphatic vessel growth and metastasis, emerged as a critical mediator gene. Its upregulation correlates with melanoma progression and metastasis.
  • Rock1 as a metastasis driver: The gene Rock1 was found to be overexpressed in metastatic samples and is implicated in metastasis and cell migration, potentially driven by PI3K/AKT pathway activation.
  • MAPK pathway deregulation: The study highlights the increasing deregulation of genes within the MAPK signaling pathway during tumor evolution, suggesting its central role in melanoma progression.
  • Nonsense-mediated decay in metastasis: This RNA degradation process was found to be significantly enriched and upregulated in metastatic samples, suggesting a role in regulating gene expression during metastasis.

Key Results

  • Gene expression changes: Many genes were increasingly upregulated during progression, including those involved in extracellular matrix organization and Vegfc.
  • Genomic alterations: The number of genes affected by SNVs and indels generally increased with tumor progression. Two samples were identified as "hypermutators" with significantly higher mutation loads.
  • Vegfc network: Identified Vegfc as a highly ranked mediator gene, linking upstream mutations to downstream differentially expressed genes, many of which are in the MAPK pathway.
  • Rock1 network: Identified Rock1 as a significant mediator in metastasis, with upstream mutations often involving the PI3K/AKT pathway.
  • Genomic diversity: The mean pairwise genetic diversity score increased with tumor growth time, indicating growing heterogeneity within tumors.
  • Transcriptomic differences: Metastasis samples showed significant differences in gene expression compared to primary tumors, with pathways like nonsense-mediated decay being upregulated.

Limitations and Future Work

  • Mouse model limitations: The authors acknowledge that their study, using a Braf-activated and Pten-deficient mouse model, focuses on a specific subset of melanomas. Induced alterations in the model might influence observed changes, and certain genes common in human melanoma (like CDKN2A) were not lost in their model.
  • Sample size for diversity: The authors suggest that a larger sample size might be necessary to definitively determine if diversity decreases after very long tumor evolution periods (e.g., >140 days).
  • Coverage depth for phylogenetic trees: The reliability of phylogenetic tree inference might be limited by the average coverage of 65x.
  • Future studies: The authors suggest that future studies with different mouse models or human samples are needed to explore the full spectrum of melanoma. They also highlight Vegfc and Rock1 as promising drug targets for reducing invasive and metastatic cancer cell activity.

Summary updated

Abstract

Abstract Metastatic melanoma is an aggressive disease that is notoriously difficult to treat. Here, we used a BRAF V600E -driven mouse model to study melanoma progression and metastasis in the absence of extrinsic mutagenic factors such as UV radiation. The tumors were grown for different time periods in seven different mice such that disease progression was tracked over time. We sequenced the whole exome and whole transcriptome of 16 melanoma samples from these mice. The timed samples revealed that many genes are increasingly upregulated during progression, such as genes involved in extracellular matrix organization, and Vegfc , a factor that induces lymphatic vessel growth and metastases to lymph nodes, which was identified as an important signaling hub in our analysis. The Vegfc gene interaction network highlighted multiple genes from the MAPK pathway that are increasingly deregulated during tumor evolution. The genomic diversity of primary and metastatic tumors showed an increase over time. Additionally, our integrative multi-omics analysis adds further evidence that Rock1 is a crucial player driving metastasis formation, and may be upregulated due to mutations in the PI3K/AKT pathway. In summary, our study contributes towards an improved understanding of melanoma which may eventually lead to improved treatment strategies.
BibTeX
@misc{moore2021dynamic,
  author = {Ariane L. Moore and Christos Dimitrakopoulos and Stefanie Flückiger-Mangual and Lukas Frischknecht and Christian Beisel and Wilhelm Krek and Niko Beerenwinkel},
  title = {{Dynamic changes of genomic and transcriptomic tumor diversity during melanoma progression}},
  year = {2021},
  doi = {10.1101/2021.03.18.435947},
  howpublished = {bioRxiv},
}
Figure