Machine learning-assisted elucidation of CD81–CD44 interactions in promoting cancer stemness and extracellular vesicle integrity

Erika K. Ramos, Chia‐Feng Tsai, Yuzhi Jia, Yue Cao, Megan Manu, Rokana Taftaf, Andrew D. Hoffmann, Lamiaa El-Shennawy, Marina Gritsenko, Valery Adorno-Cruz, Emma J. Schuster, David Scholten, Dhwani Patel, Xia Liu, Priyam Patel, Brian Wray, Youbin Zhang, Shanshan Zhang, Ronald Moore, Jeremy V Mathews, Matthew J. Schipma, Tao Liu, Valerie Tokars, Massimo Cristofanilli, Tujin Shi, Yang Shen, Nurmaa K. Dashzeveg, Huiping LiueLife 2022 · DOI

TL;DR

This study leverages machine learning and experimental validation to uncover a critical interaction between CD81 and CD44 proteins that drives breast cancer stemness and metastasis. The findings reveal CD81 as a novel driver of tumor progression, highlighting its role in promoting tumor cell aggregation, extracellular vesicle integrity, and ultimately, the spread of cancer.

What Problem Does This Paper Address?

Triple-negative breast cancer (TNBC) is a particularly aggressive form of breast cancer with a high propensity for metastasis and low survival rates. A key challenge in understanding and treating such cancers lies in elucidating the complex molecular networks that govern tumor-initiating cells (cancer stem cells), which are responsible for tumor relapse, therapy resistance, and metastatic spread. While CD44 has been implicated in these processes, the precise molecular mechanisms, particularly concerning protein-protein interactions and their role in extracellular vesicle (EV) function, remain poorly understood. This paper aims to unravel the intricate interplay between CD81 and CD44 in promoting cancer stemness and metastasis, particularly in the context of TNBC.

Methods

The researchers employed a multi-pronged approach combining cutting-edge computational and experimental techniques:

  • Machine Learning-Assisted Protein Structure Modeling: Tools like iTasser, ClusPro, and Bayesian Active Learning (BAL) were used to predict the structural interactions between CD44 and CD81, focusing on their extracellular regions. This allowed for the identification of potential binding hotspots.
  • Guided Mutagenesis: Based on the structural predictions, specific mutations were introduced into CD81 (CD81d) to disrupt its interaction with CD44, providing experimental validation of the predicted interaction sites.
  • Proteomic and Phosphoproteomic Analyses: Global proteomic and phosphoproteomic profiling of tumor cells with depleted CD81 or CD44 was performed using mass spectrometry. This helped identify altered signaling pathways, particularly those related to endocytosis.
  • RNA Sequencing: Transcriptome analysis was conducted to understand the broader molecular changes induced by CD81 depletion.
  • Experimental Validation: A suite of experimental techniques was used, including:
  • Mammosphere Assays: To assess cancer stem cell self-renewal capacity.
  • Immunofluorescence Microscopy: To visualize protein localization and co-localization.
  • Co-immunoprecipitation (Co-IP): To confirm direct protein-protein interactions.
  • Transmission Electron Microscopy (TEM) and Cryo-EM: To examine the ultrastructure of cells and extracellular vesicles.
  • Nanoparticle Tracking Analysis (NTA) and Microflow Vesiculometry (MFV): To quantify EV release and size.
  • In Vivo Tumorigenesis and Metastasis Models: Using mouse models to assess the role of CD81 in tumor initiation, growth, and lung metastasis.
  • Circulating Tumor Cell (CTC) Analysis: Using flow cytometry and immunofluorescence to examine CD81 expression in patient-derived CTCs.

Datasets

  • Public Databases: Analyses of existing datasets were performed, including those for patient survival based on CD81 expression (Liu_2014 cohort) and single-cell RNA sequencing of Parsotix-filtered CTCs.
  • Newly Generated Datasets:
    • RNA sequencing data from MDA-MB-231 cells with siCon and siCD81 treatments.
  • Mass spectrometry-based global proteomic and phosphoproteomic data from tumor cells and exosomes/EVs.
  • Data from in vivo experiments involving tumor implantation and metastasis in mice.
  • Data from flow cytometry and immunofluorescence analysis of patient-derived CTCs.

Key Insights and Ideas

  • CD81 as a Novel Driver of TNBC Metastasis: The study firmly establishes CD81 as a critical player in TNBC progression, moving beyond its known roles in immune cells.
  • CD81-CD44 Direct Interaction is Crucial: The research demonstrates a direct physical interaction between CD81 and CD44 in their extracellular regions, which is essential for their pro-metastatic functions.
  • EV Integrity and Function are CD81-Dependent: CD81 plays a vital role in maintaining the structural integrity of extracellular vesicles, and these EVs, in turn, promote cancer stemness in recipient cells.
  • Machine Learning as a Powerful Discovery Tool: The integration of machine learning for protein structure prediction and interaction analysis significantly accelerated the identification of the CD81-CD44 interaction and guided experimental validation.
  • CD81 as a Potential Biomarker: The enrichment of CD81 in clustered CTCs suggests its potential as a biomarker for aggressive disease and metastasis.

Key Results

  • Mammosphere Formation: CD81 depletion significantly reduced mammosphere formation, indicating a role in cancer stemness.
  • Tumor Cell Clustering: CD81 KO cells showed compromised cluster formation, while CD81 reintroduction rescued this phenotype.
  • EV Biogenesis and Integrity: CD81KO cells secreted more EVs, but these EVs exhibited impaired membrane integrity.
  • Tumorigenesis and Metastasis: CD81 deficiency led to significantly reduced tumor initiation, growth, and, most notably, abrogated lung metastasis in mouse models. Mice with CD81KO tumors showed 100% survival compared to 0% for WT tumors.
  • CTC Enrichment: CD81 was found to be highly expressed in human circulating tumor cells (CTCs), particularly in clustered CTCs, which are known to be more metastatic.
  • Clinical Correlation: High CD81 expression in breast tumors correlated with unfavorable patient survival outcomes.

Limitations and Future Work

  • EV Concentration: The study acknowledges that higher concentrations of EVs were used for education experiments than might be found physiologically, which is a limitation of current in vitro models.
  • Other Tetraspanins: While CD81 is highlighted, the paper notes that other tetraspanins (e.g., TSPAN8, CD9) also play roles in cancer stemness, suggesting future research could explore their interplay with CD44 and CD81.
  • Therapeutic Strategies: The study suggests that blocking CD81 could be a potential therapeutic strategy, but further investigation is needed to explore this in detail.
  • Microenvironment Interactions: Future studies should investigate the effects of cancer exosomes on tumor stromal cells and immune cells in various microenvironment niches.
  • Broader Cancer Types: The findings in TNBC warrant investigation into the role of CD81-CD44 interactions in other cancer types.

Summary updated

Abstract

Tumor-initiating cells with reprogramming plasticity or stem-progenitor cell properties (stemness) are thought to be essential for cancer development and metastatic regeneration in many cancers; however, elucidation of the underlying molecular network and pathways remains demanding. Combining machine learning and experimental investigation, here we report CD81, a tetraspanin transmembrane protein known to be enriched in extracellular vesicles (EVs), as a newly identified driver of breast cancer stemness and metastasis. Using protein structure modeling and interface prediction-guided mutagenesis, we demonstrate that membrane CD81 interacts with CD44 through their extracellular regions in promoting tumor cell cluster formation and lung metastasis of triple negative breast cancer (TNBC) in human and mouse models. In-depth global and phosphoproteomic analyses of tumor cells deficient with CD81 or CD44 unveils endocytosis-related pathway alterations, leading to further identification of a quality-keeping role of CD44 and CD81 in EV secretion as well as in EV-associated stemness-promoting function. CD81 is coexpressed along with CD44 in human circulating tumor cells (CTCs) and enriched in clustered CTCs that promote cancer stemness and metastasis, supporting the clinical significance of CD81 in association with patient outcomes. Our study highlights machine learning as a powerful tool in facilitating the molecular understanding of new molecular targets in regulating stemness and metastasis of TNBC.
BibTeX
@inproceedings{ramos2022machine,
  author = {Erika K. Ramos and Chia‐Feng Tsai and Yuzhi Jia and Yue Cao and Megan Manu and Rokana Taftaf and Andrew D. Hoffmann and Lamiaa El-Shennawy and Marina Gritsenko and Valery Adorno-Cruz and Emma J. Schuster and David Scholten and Dhwani Patel and Xia Liu and Priyam Patel and Brian Wray and Youbin Zhang and Shanshan Zhang and Ronald Moore and Jeremy V Mathews and Matthew J. Schipma and Tao Liu and Valerie Tokars and Massimo Cristofanilli and Tujin Shi and Yang Shen and Nurmaa K. Dashzeveg and Huiping Liu},
  title = {{Machine learning-assisted elucidation of CD81–CD44 interactions in promoting cancer stemness and extracellular vesicle integrity}},
  booktitle = {eLife},
  year = {2022},
  doi = {10.7554/elife.82669},
}
Figure