Platelet Transcriptome Yields Progressive Markers in Chronic Myeloproliferative Neoplasms and Identifies Putative Targets of Therapy

Zhu Shen, Wenfei Du, Cecelia Perkins, Lenn Fechter, Vanita Natu, Holden T. Maecker, Jesse W. Rowley, Jason Gotlib, James L. Zehnder, Anandi KrishnanbioRxiv 2021 · DOI

TL;DR

This paper leverages platelet RNA sequencing to identify progressive molecular markers in chronic myeloproliferative neoplasms (MPNs), enabling accurate prediction of disease progression, particularly to myelofibrosis (MF). The study also uncovers impaired protein homeostasis as a key driver of MPN evolution and suggests novel therapeutic targets.

What Problem Does This Paper Address?

Predicting disease progression in chronic degenerative disorders like MPNs is a significant challenge. This difficulty hinders early detection, effective risk stratification, and the development of timely preventive interventions. Current methods often rely on genetic information, but the complex interplay of molecular changes that drive disease progression, especially the transition from earlier MPN subtypes (ET, PV) to the more advanced myelofibrosis (MF), remains incompletely understood. The authors aim to identify more sensitive biomarkers for tracking this progression and to uncover underlying biological mechanisms.

Methods

The researchers employed a multi-pronged approach:

  • Platelet RNA Sequencing (RNA-seq): This is the core technology used to analyze gene expression profiles in purified platelets.
  • Machine Learning (Lasso-penalized regression): This statistical technique was used to build predictive models. Lasso is particularly useful for high-dimensional data as it performs feature selection, identifying a minimal set of significant predictors.
  • Unsupervised Learning (Principal Component Analysis - PCA): PCA was used to visualize and explore the overall patterns of gene expression and to identify distinct clusters of patients based on their transcriptomic profiles.
  • Differential Gene Expression Analysis (DGEA): This standard technique was used to identify genes that are significantly up- or down-regulated between different groups of patients (e.g., MPN subtypes vs. controls, treated vs. untreated).
  • Gene Set Enrichment Analysis (GSEA): GSEA was used to identify biological pathways that are significantly altered in MPN patients, providing mechanistic insights.
  • External Validation: The predictive models were rigorously tested on independent datasets, both temporally (using a separate cohort collected years later) and geographically (using data from other institutions), to ensure robustness.

Datasets

The study utilized the following datasets:

  • Stanford MPN Cohorts: Two independent, time-separated cohorts of MPN patients and healthy controls collected at Stanford University.
  • Cohort 1: 71 MPN patients (24 ET, 33 PV, 40 MF) and healthy controls.
  • Cohort 2: 49 MPN patients (24 ET, 33 PV, 40 MF) and healthy controls.
  • Independently Published Datasets:
    • A dataset from Rondina et al. (n=31 healthy donors) used for integrated training.
  • A dataset from Guo et al. (n=25 MF, n=15 healthy donors) used for geographical validation.

Key Insights and Ideas

  • Platelets as a "Liquid Biopsy": The study strongly supports the idea that platelet transcriptomes can serve as a sensitive and accessible proxy for molecular changes occurring in MPNs, reflecting disease status and progression.
  • Progressive Molecular Signatures: The researchers identified a clear, incremental shift in gene expression across the MPN spectrum (from ET/PV to MF), suggesting a continuous progression of molecular alterations.
  • Impaired Protein Homeostasis: A major mechanistic insight is the prominent role of impaired protein homeostasis and the integrated stress response in MPN progression, particularly in MF. This suggests that cellular stress management pathways are dysregulated.
  • Beyond Genetics: The findings demonstrate that platelet transcriptomic markers can predict disease progression independently of, or in addition to, genetic mutations, offering a more comprehensive view of disease biology.
  • Therapeutic Targets: The study identified specific gene signatures associated with JAK inhibitor treatment (ruxolitinib) and other pathways (interferon, proliferation, proteostasis) that could serve as targets for future MPN therapies.

Key Results

  • High Predictive Accuracy for MF: A machine learning model using a core set of progressive platelet transcriptomic markers achieved high accuracy in predicting MF, with an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.96 for temporal validation and 0.97 for geographical validation.
  • Subtype Discrimination: The platelet transcriptome effectively discriminated between healthy controls and the three MPN subtypes (ET, PV, MF), with distinct and progressively divergent gene expression patterns observed.
  • Identification of Key Genes: The study identified a small set of (<5) progressive transcriptomic markers that are highly predictive of disease progression.
  • Mechanistic Pathways: GSEA revealed significant enrichment of pathways related to immune modulation, interferon response, proliferation, and notably, impaired protein homeostasis and integrated stress response in MPNs, especially in MF.

Limitations and Future Work

  • Cross-sectional Design: The study's design is cross-sectional, not longitudinal. While it provides snapshots of gene expression, it doesn't directly track disease progression over time. The authors suggest that longitudinal evaluation of disease and treatment markers is warranted.
  • Platelet Transcriptome Focus: The study focused solely on the platelet transcriptome. Future work should investigate overlap with other cell types (megakaryocytes, CD34+ cells, immune cells, whole blood) to gain a more complete picture of bone marrow pathology.
  • Ribosomal RNA Depletion: The choice of ribosomal RNA depletion might have excluded some diversity in the platelet RNA repertoire, which could be explored in future studies.
  • Validation Needs: While external validation was performed, further biological and computational validation is needed to translate these findings into clinical decision-making and personalized medicine.
  • Future Directions: The authors suggest exploring the utility of platelet RNA signatures for understanding MPN complications (thrombosis, bleeding, fibrosis) and for designing therapies. They also propose investigating the potential of machine learning algorithms for predicting the probability and timing of progression from ET/PV to MF.

Summary updated

Abstract

Abstract Predicting disease progression remains a particularly challenging endeavor in chronic degenerative disorders and cancer, thus limiting early detection, risk stratification, and preventive interventions. Here, profiling the spectrum of chronic myeloproliferative neoplasms (MPNs) as a model, we identify the blood platelet transcriptome as a proxy for highly sensitive progression biomarkers that also enables prediction of advanced disease via machine learning algorithms. Using RNA sequencing (RNA-seq), we derive disease-relevant gene expression in purified platelets from 120 peripheral blood samples constituting two time-separated cohorts of patients diagnosed with one of three MPN subtypes at sample acquisition – essential thrombocythemia, ET (n=24), polycythemia vera, PV (n=33), and primary or post ET/PV secondary myelofibrosis, MF (n=42), and healthy donors (n=21). The MPN platelet transcriptome reveals an incremental molecular reprogramming that is independent of patient driver mutation status or therapy and discriminates each clinical phenotype. Leveraging this dataset that shows a characteristic progressive expression gradient across MPN, we develop a machine learning model (Lasso-penalized regression) and predict advanced subtype MF at high accuracy and under two conditions of external validation: i) temporal: our two Stanford cohorts, AUC-ROC of 0.96; and ii) geographical: independently published data of an additional n=25 MF and n=46 healthy donors, AUC-ROC of 0.97). Lasso-derived signatures offer a robust core set of &lt; 5 MPN transcriptome markers that are progressive in expression. Mechanistic insights from our data highlight impaired protein homeostasis as a prominent driver of MPN evolution, with persistent integrated stress response. We also identify JAK inhibitor-specific signatures and other interferon, proliferation, and proteostasis-associated markers as putative targets for MPN-directed therapy. Our platelet transcriptome snapshot of chronic MPNs demonstrates a proof of principle for disease risk stratification and progression beyond genetic data alone, with potential utility in other progressive disorders. Highlights Leveraging two independent and mutually validating MPN patient cohorts, we identify progressive transcriptomic markers that also enable externally validated prediction in MPNs. Our platelet RNA-Seq data identifies impaired protein homeostasis as prominent in MPN progression and offers putative targets of therapy. VISUAL ABSTRACT
BibTeX
@misc{shen2021platelet,
  author = {Zhu Shen and Wenfei Du and Cecelia Perkins and Lenn Fechter and Vanita Natu and Holden T. Maecker and Jesse W. Rowley and Jason Gotlib and James L. Zehnder and Anandi Krishnan},
  title = {{Platelet Transcriptome Yields Progressive Markers in Chronic Myeloproliferative Neoplasms and Identifies Putative Targets of Therapy}},
  year = {2021},
  doi = {10.1101/2021.03.12.435190},
  howpublished = {bioRxiv},
}
Figure