Uncovering perturbations in human hematopoiesis associated with healthy aging and myeloid malignancies at single cell resolution

Marina Ainciburu, Teresa Ezponda, Nerea Berastegui, Ana Alfonso Piérola, Amaia Vilas‐Zornoza, Patxi San Martín‐Úriz, Diego Alignani, Jose Lamo de Espinosa, Mikel San-Julián, Tamara Jiménez, Félix López, Sandra Muntión, Fermín Sánchez‐Guijo, Antonieta Molero, Julia Montoro, Guillermo Serrano, Aintzane Díaz-Mazkiaran, Miren Lasaga, David Gómez-Cabrero, María Díez‐Campelo, David Valcárcel, Mikel Hernáez, Juan P. Romero, Felipe PrósperbioRxiv 2021 · DOI

TL;DR

This study uses single-cell RNA sequencing to map the complex landscape of human hematopoietic stem and progenitor cells (HSPCs) across aging and myelodysplastic syndromes (MDS). By integrating computational analyses like trajectory inference and gene regulatory network construction, the researchers uncover age-related shifts in cell populations and identify specific molecular signatures associated with MDS, paving the way for personalized therapeutic strategies.

What Problem Does This Paper Address?

Hematopoiesis, the process of blood cell formation, is a dynamic and intricate system. While traditionally viewed as a linear differentiation pathway, recent single-cell studies suggest a more continuous and heterogeneous landscape of hematopoietic stem and progenitor cells (HSPCs). Aging is known to alter this system, leading to a decline in immune function and an increased risk of myeloid malignancies like myelodysplastic syndromes (MDS). However, the precise molecular mechanisms underlying these age-related changes and their contribution to MDS pathogenesis remain incompletely understood. This paper aims to dissect these complexities by characterizing the transcriptional dynamics and regulatory networks of HSPCs in healthy aging and in MDS patients, seeking to identify key drivers of these perturbations.

Methods

The researchers employed a multi-faceted computational approach combined with single-cell RNA sequencing (scRNA-seq). Key techniques include:

  • scRNA-seq: This technology allowed for the profiling of individual cells, revealing heterogeneity within HSPC populations.
  • Unsupervised Clustering and Cell Type Annotation: Standard bioinformatics pipelines were used to group cells based on their transcriptional profiles and assign them to specific cell types.
  • Classification Model (Logistic Regression with Elastic-Net Regularization): An in-house classifier was developed to assign cell identities to elderly and MDS samples based on a reference dataset of young cells. This method is noted for its suitability for sparse data and variable selection.
  • Gene Set Enrichment Analysis (GSEA): Used to identify functional differences between cell populations and conditions by assessing the enrichment of predefined gene sets.
  • Trajectory Inference (STREAM and Palantir): These algorithms were crucial for reconstructing the differentiation pathways of HSPCs, allowing the researchers to study how gene expression changes dynamically during cell development and how these trajectories are altered by aging and disease.
  • Gene Regulatory Network (GRN) Reconstruction (SCENIC): This approach was used to infer the regulatory relationships between transcription factors and their target genes, providing insights into the control mechanisms governing hematopoiesis and how they are perturbed in aging and MDS.

Datasets

  • Human Bone Marrow CD34+ Cells:
    • Young healthy donors (n=5, age 19-23 years)
    • Elderly healthy donors (n=3, age 61-84 years)
    • Newly diagnosed MDS patients (n=4, with multilineage dysplasia)
  • Publicly Available Dataset: A dataset of 8,176 human CD34+ progenitors was used for benchmarking the in-house cell classifier.

Key Insights and Ideas

  • Aging Alters HSPC Proportions and Function: The study reveals significant shifts in the proportions of different progenitor cell types with age, alongside functional changes indicated by altered gene expression pathways related to stress response and inflammation in elderly individuals.
  • Trajectory Inference Reveals Age-Related Differentiation Impairments: By tracing cell differentiation paths, the researchers found that while elderly HSPCs might show an initial bias towards certain lineages, they often lose the capacity for mature differentiation later in the process.
  • Gene Regulatory Networks Highlight Age-Associated Dysregulation: The analysis of GRNs demonstrates a decrease in the activity of specific regulons in elderly HSPCs, suggesting a loss of co-regulatory mechanisms that guide normal differentiation.
  • MDS Exhibits Distinct Transcriptional Signatures and GRN Alterations: The study identified specific molecular lesions in MDS patients, including altered erythroid differentiation dynamics and the activation of specific regulons (e.g., SMAD1, HOXA6, POU2F2, RUNX1) that are implicated in the pathogenesis of the disease.
  • Personalized Therapeutic Potential: The ability to identify patient-specific transcriptional alterations and deregulated GRNs in MDS opens avenues for developing personalized therapeutic strategies.

Key Results

  • Cellular Composition Changes with Age: A significant increase in HSCs and a reduction in committed lymphoid and myeloid lineages were observed in elderly individuals. Erythroid-committed cells also showed an increased proportion.
  • Functional Pathways Differ with Age: GSEA revealed enrichment of stress-related pathways (e.g., TNF-α, TGF-β, p53, apoptosis) in elderly HSPCs, while young HSPCs showed enrichment in metabolic and cell cycle progression pathways.
  • Monocytic Differentiation Impairment in Elderly: Trajectory analysis indicated that while early progenitors in elderly individuals showed a stronger bias towards monocytic differentiation, more mature progenitors lost this capacity.
  • MDS Patients Show Aberrant Erythroid Differentiation: Analysis of MDS patients revealed altered gene expression dynamics in genes involved in erythroid differentiation, with some genes showing patient-specific alterations.
  • Identification of Key MDS-Associated Regulons: Several transcription factors, including SMAD1, HOXA6, POU2F2, and RUNX1, were identified as having aberrant activity in MDS patients, suggesting their role in disease development.

Limitations and Future Work

  • Heterogeneity of MDS: The authors acknowledge the significant clinical and molecular heterogeneity of MDS, which led them to focus on a specific subtype (multilineage dysplasia) for their analysis. This highlights the need for further studies to understand the full spectrum of MDS.
  • In-house Classifier Benchmarking: While the in-house classifier performed well, minor differences were noted when applied to external data, suggesting ongoing refinement might be beneficial.
  • Therapeutic Exploitation: The identification of abnormal regulatory mechanisms in MDS is presented as a potential starting point for personalized therapeutic approaches. Future work would involve validating these findings and developing targeted interventions.
  • Broader Application of Computational Tools: The study demonstrates the power of combining scRNA-seq with advanced computational methods for dissecting complex biological processes, suggesting this approach can be applied to other diseases and biological systems.

Summary updated

Abstract

ABSTRACT Early hematopoiesis is a continuous process in which hematopoietic stem and progenitor cells (HSPCs) gradually differentiate toward specific lineages. Aging and myeloid malignant transformation are characterized by changes in the composition and regulation of HSPCs. In this study, we used single cell RNA sequencing (scRNAseq) to characterize an enriched population of human hematopoietic stem and progenitor cells (HSPC) obtained from young and elderly healthy individuals. Based on their transcriptional profile, we identified changes in the proportions of progenitor compartments during aging, and differences in their functionality, as evidenced by gene set enrichment analysis. Trajectory inference revealed that altered gene expression dynamics accompanied cell differentiation, which could explain age-associated changes in hematopoiesis. Next, we focused on key regulators of transcription by constructing gene regulatory networks and detected regulons that were specifically active in elderly individuals. Using previous findings in healthy cells as a reference, we analyzed scRNA-seq data obtained from patients with myelodysplastic syndrome (MDS) and detected specific alterations of the expression dynamics of genes involved in erythroid differentiation in all patients with MDS such as TRIB2. In addition, the comparison between transcriptional programs and GRN regulating normal HSPCs and MDS HSPCs allowed identification of regulons that were specifically active in MDS cases such as SMAD1, HOXA6, POU2F2 and RUNX1 suggesting a role of these TF in the pathogenesis of the disease. In summary, we demonstrate that the combination of single cell technologies with computational analysis tools enable the study of a variety of cellular mechanisms involved in early hematopoiesis and can be used to dissect perturbed differentiation trajectories associated with aging and malignant transformation. Furthermore, the identification of abnormal regulatory mechanisms associated with myeloid malignancies could be exploited for personalized therapeutic approaches in individual patients.
BibTeX
@misc{ainciburu2021uncovering,
  author = {Marina Ainciburu and Teresa Ezponda and Nerea Berastegui and Ana Alfonso Piérola and Amaia Vilas‐Zornoza and Patxi San Martín‐Úriz and Diego Alignani and Jose Lamo de Espinosa and Mikel San-Julián and Tamara Jiménez and Félix López and Sandra Muntión and Fermín Sánchez‐Guijo and Antonieta Molero and Julia Montoro and Guillermo Serrano and Aintzane Díaz-Mazkiaran and Miren Lasaga and David Gómez-Cabrero and María Díez‐Campelo and David Valcárcel and Mikel Hernáez and Juan P. Romero and Felipe Prósper},
  title = {{Uncovering perturbations in human hematopoiesis associated with healthy aging and myeloid malignancies at single cell resolution}},
  year = {2021},
  doi = {10.1101/2021.07.30.454542},
  howpublished = {bioRxiv},
}
Figure