Unraveling Heterogeneity of Aged Hematopoietic Stem Cells By Single-Cell RNA Sequence Analysis

During aging, hematopoietic stem cells (HSCs) alter quantitatively as well as qualitatively due to accumulating damages induced by intrinsic and extrinsic stresses. Functional decline of HSCs causes deregulated hematopoiesis resulting in anemia, immune dysfunction, and increased risk of hematologic malignancies. The absolute number of HSCs in aged mice evidently increases compared to young mice. In addition, aged HSCs show abnormal hematopoiesis such as myeloid-biased differentiation accompanied by low production of lymphocytes. However, the molecular mechanisms underlying age-associated changes in hematopoiesis remain largely unknown. In this study, we performed single-cell RNA sequence analysis (scRNA-seq) of HSCs from young (10-week-old), middle-aged (12-month-old) and aged (20-month-old) mice to gain insight into the dynamics of HSC aging.

scRNA-seq analysis revealed three major clusters (A, B, C) and three minor clusters (D, E, F) in young HSCs. Of interest, aged HSCs also showed similar cluster formation. One of the minor clusters was characterized by gene signature associated with inflammatory response and significantly increased with age, while the remaining two minor clusters, which showed cell cycle gene signature, did not change in proportion. One of the major clusters, Cluster C, which showed the strongest expression of HSC-specific gene sets compared with other clusters, moderately increased with age.

Our scRNA-seq analysis confirmed upregulation of age-related genes previously reported, such as Selp, Mt1, and Vwf, in a considerable portion of aged HSCs. von Willebrand factor (Vwf) encodes a blood glycoprotein produced by megakaryocytes and endothelial cells. Several groups have reported that Vwf-expressing HSCs show myeloid/platelet-biased differentiation (Joana C et al., Nature 2013; Sandra P et al., Developmental Cell 2018). Importantly, our scRNA-seq data identified that Clusterin (Clu) is rarely expressed in young HSCs and is dramatically upregulated in aged HSCs. Clu expression was predominantly increased in one of the major clusters, Cluster C. Clusterin encodes a secreted chaperone involved in clearance of cellular debris and regulation of apoptosis. We hypothesized that Clu would be useful as a marker of a unique subpopulation of aged HSCs, and thus conducted further analysis by using Clu reporter mice with Clu BAC clone, in which an EGFP reporter gene was inserted at the initiating ATG codon of the Clu gene so that EGFP expression is driven by the regulatory sequences of the BAC gene.

Clu/GFP was preferentially expressed in HSCs and at lower frequencies in MPP1 than HSCs in Lineage ^-Sca-1 ⁺c-Kit ⁺ (LSK) cell fraction. Clu-positive HSCs expressed high levels of CD150 and were detected in 10% and 60% of HSCs in 10-week-old young mice and 8-month-old middle-aged mice, respectively, indicating that Clu-positive HSCs increase with aging. We next assessed the function of Clu-positive HSCs in young mice. Clu-positive young HSCs established significantly lower chimerism than Clu-negative young HSCs and preferentially differentiated into myeloid cells in competitive transplantation assays. RNA-seq analysis of Clu-positive and Clu-negative HSCs from young mice confirmed that Clu-positive HSCs show the gene signature of myeloid-biased HSCs. Characterization of Clu-positive and negative subpopulations in aged HSCs is currently underway. These results suggest that Clu-positive HSCs represent myeloid-biased HSCs which expand with aging, thus Clu expression serves as a novel marker to monitor the alterations in HSC heterogeneity with aging.