Distinct Pathogenesis of Clonal Hematopoiesis Revealed By Single Cell RNA Sequencing Integrated with Highly Sensitive Genotyping Method

Background

Clonal hematopoiesis (CH) in apparently healthy individuals has been drawing an increasing attention of recent years due to its implication in the risk of hematopoietic malignancies and also in cardiovascular complications. However, our knowledge on CH has been largely based on genetic studies, while few functional analyses have been performed using human materials with most studies being confined to artificial models using mice. The major challenge here is the difficulty of isolating CH clones from wild-type (WT) cells in human bone marrow (BM) to elucidate the effect of CH-mutations.

Methods

To investigate cellular phenotypes of mutated and WT cells separately in CH, we developed a Fluidigm C1-based single cell-sequencing platform for simultaneous genotyping and gene expression analysis. We analyzed a total of 10,178 hematopoietic stem/progenitors (HSPCs) derived from BM of patients with (n=11) and without (n=17) CH.

Results

In the analysis of HSPCs from CH(−) elderly individuals, we found a significant positive correlation between age and expression of gene sets implicated in inflammatory responses including TGFβ signalling and IL-6. We next analysed CH(+) samples, including those with mutations in TET2, DNMT3A, SF3B1 and IDH1 in which we investigated the pathways enriched in differentially expressed genes between mutated and unmutated cells. Regardless of mutation type, mutant cells showed an upregulation of genes implicated in an enhanced cell proliferation, while genes related to inflammatory responses were significantly downregulated. These results suggest that mutant HSPCs show an enhanced cell proliferation and an attenuated response to an inflammatory microenvironment in aged BM, compared with endogenous WT counterparts. Based on these observations, we further investigated the role of the BM microenvironment in CH, in which we compared the phenotype of WT cells in age-matched CH(+) and CH(−) cases. Compared with those from CH(−) cases, WT cells from these CH(+) cases showed an enhanced response to proinflammatory cytokines, including IL-6, interferons and TNF, suggesting a possibility that CH(+) BM might be characterized by a more enhanced inflammatory microenvironment, compared with CH(−) BM.

To understand the different phenotype of WT cells between CH(+) and CH(−) cases, we next investigated a possible effect of mutant cells on endogenous WT cells. Such an effect was first documented in an IDH1-mutated CH case. IDH1-mutated cells showed a gene expression profile suggestive of an enhanced cell proliferation compared with endogenous IDH1-WT cells (WT ^IDH1). However, on the basis of comparison of gene expression between WT cells from CH(−) cases, it was better explained by suppressed proliferation of endogenous WT cells, rather than enhanced proliferation of mutant cells. In agreement with this, mouse BM cells treated with 2-hydroxyglutalate (2HG), an oncometabolite produced by IDH1-mutant cells, mimicked the endogenous WT cells in IDH1-mutated cases, including downregulated E2F target genes and upregulated inflammation-related genes, compared with control BM cells. A non-cell autonomous effect of mutations was also seen in cases with TET2-mutated CH, in which not only TET2-mutated cells but also endogenous WT cells exhibited a significantly different gene expression profile, compared with those from CH(−) cases. Of interest, when BM cells from WT Ly5.1/5.2 mice were co-transplanted with those from heterozygous conditional Tet2 knock-out (Ly5.2) or WT (Ly5.2) mice into lethally irradiated mice (Ly5.1) and flow-sorted Lin ⁻ WT donor competitor cells were analysed using single cell sequencing, the WT HSC-like cells co-transplanted with Tet2-mutant competitors exhibited enhanced cell proliferation and IFNα and IFNγ pathway genes, compared with those co-transplanted with WT competitors.

Conclusions

Taken together, these results suggest that mutant cells in CH(+) BM have non-cell autonomous effects on endogenous WT cells, which might be responsible for an accelerated inflammatory microenvironment of aged BM, favor positive selection of CH-clones, and also affect the phenotype of endogenous WT cells, contributing to the pathogenesis of CH.