Hematopoietic Clonal Evolution Goes Spatial

Summary:

The spatial distribution of cells carrying clonal hematopoiesis mutations in the bone marrow and the potential role of interactions with the microenvironment are largely unknown. This study takes clonal evolution to the spatial level by describing a novel technique examining the spatial location of mutated clones in the bone marrow and the first evidence that mutated hematopoietic clones are spatially constrained and have heterogenous locations within millimeters of distance.

See related article by Young et al., p. 153 (10).

Clonal hematopoiesis (CH) refers to the age-associated phenomenon whereby mutations are acquired over the course of a person's life and can increase the risk for developing hematologic malignancies (1). Mutations associated with CH can occur in genes including DNMT3A, TET2, ASXL1, and JAK2, many of which have also been identified in blood cancers including the myeloproliferative neoplasms (MPN) and acute myeloid leukemia (1). MPNs are clonal disorders in which mutations arise in hematopoietic stem and progenitor cells (HSPC; ref. 1). Traditionally, clonal evolution has been inferred through bulk genomic sequencing. However, single-cell genotyping has greatly improved the temporal tracking of clonal evolution in hematologic malignancies and highlight their expansion overtime. The Mission Bio Tapestri platform combines multiomic single cell DNA genotyping with cell-surface protein immunophenotyping, revealing deeper complexities of clonal evolution, suggesting that bulk sequencing is not sufficient to fully understand such dynamic events (2).

As blood cancers are traditionally thought of as “liquid tumors,” freely moving throughout the body, less importance has been placed on identifying the spatial location of malignant hematopoietic clones as compared with solid tumors. Studies in some solid tumors have shown that progression and changes in spatial locality of the tumor can correlate with distinct oncogenic events (3–5). This highlights that cancer cell interactions with the surrounding microenvironment cells in solid organs may be essential to overcome tissue barriers as the cancer spreads.

Although solid organs and the bone marrow (BM) are very different tissues, these observations raise the question whether similar or different niche-related mechanisms are playing a role in the development and progression of blood cancers. The BM is a unique niche critical for healthy hematopoiesis, and we have now seen at single-cell resolution how disruption of the niche microenvironment is detrimental to maintenance of hematopoiesis (6).

The nongenetic factors including microenvironmental niche interactions that may facilitate the competitive advantage of CH mutations are under investigation but remain largely unknown. Interestingly, it has recently been shown that different parts of the skeleton respond heterogeneously to stress responses, potentially indicating specialized niches in certain bones (7). With regard to hematologic malignancies, the BM niche undergoes substantial remodeling, suggesting that survival of malignant clones depends on reshaping the niche to support malignant transformation (8). Constraint of CH clones was suggested in a previous study showing that clones detected in a femur head were not detected in the peripheral blood after hip arthroplasty (9).

Many questions remain regarding the role of interactions of CH clones within the BM niche. Do certain parts of the BM niche harbor mutated clones and/or do mutated cones initiate remodeling of the BM niche? Is this mutation dependent? Do these interactions control which CH mutations arise first? Could understanding these mechanisms enable the creation of therapeutics to eradicate CH clones before they transform? We are at the tip of the iceberg in our investigations of this emerging field. The creation of a spatial map of mutated clones in pre- and postleukemic transformation will be essential to addressing these questions.

In this issue of Blood Cancer Discovery, Young and colleagues start to address these hypotheses by performing comprehensive mutational analysis of the BM and peripheral blood (PB) of a patient diagnosed with polycythemia vera (PV) and known driver clonal mutation JAK2V617F (10). They combine results from bulk targeted genomic sequencing (Myeloseq assay) and digital droplet PCR (ddPCR; Fig. 1A). They then perform single-cell genotyping and immunophenotyping of mononuclear BM cells using the Mission Bio Tapestri platform combined with flow cytometry validation (Fig. 1B).

Figure 1.

Combining bulk, single-cell, and spatial genomic genotyping to dissect clonal hematopoiesis. A, Bulk targeted DNA sequencing, targeted Myeloseq panel, and digital droplet PCR (ddPCR) were performed on peripheral blood (PB) and bone marrow (BM) from a polycythemia vera (PV) patient showing increased variant allele frequency (VAF) of mutations in the BM. B, Single-cell genotyping and immunophenotyping using the Mission Bio Tapestri platform of BM revealed separate expansion of clones undetectable by bulk sequencing. C, Genotyping of BM sections from the femur head revealed spatial heterogeneity of hema­topoietic clones and clustering of CD71⁺ cells. Created with BioRender.com.

Importantly, the authors demonstrate proof-of-principle of a novel technique to investigate the spatial constraints of clonal mutations in the BM (Fig. 1C). To demonstrate this technique, they leverage a femur head explant following hip arthroplasty from the PV patient. The technique involved partitioning the femur head using a band saw and hand tools into approximately 400 2-mm³ fragments. The original spatial localization of the fragments was recorded before using them for further mutational analysis by targeted panel DNA sequencing and/or ddPCR. The authors show that different mutated clones cluster in different sections of the BM niche.

Bulk genomic sequencing of both the PB and BM by Myeloseq assay and ddPCR identified multiple clonal mutations in addition to the JAK2V617F driver mutation, including TET2 L1081*, DNMT3A R882C, GNAS R844H (Fig. 1A). Importantly, all the mutations were found to have a higher clonal mutational burden in the BM of the patient at the time of hip arthroplasty; for example, JAK2V617F variant allele frequency was BM = 0.64 and PB = 0.27. This demonstrates the difficulty of tracking CH via PB as the actual mutational burden may be greatly underestimated. Nevertheless, expansion of the TET2 L1081* was observed in the PB of the patient over time.

Single-cell genomic sequencing and immunophenotyping of the BM revealed a deeper understanding of the clonal evolution of this patient. Although it could have been inferred from the bulk sequencing results that the majority of BM cells were heterozygous for JAK2V617F, single-cell genotyping identified a very different clonal composition comprising of wild-type cells (48.6%), a large JAK2V617F^HOM clone (39.4%), and JAK2V167F^HOM/TET2^HET (5.3%) derived from the original JAK2V617F^HET clone (3.2%; Fig. 1B). The JAK2V617F neoplastic clones were enriched in the erythroblasts (CD71⁺ CD141⁺) and myeloid (CD33⁺) populations. Interestingly, DNMT3A^HET (1.9%) and GNAS^HET (1.6%) were each separate clones and were detected across all populations, including T and B cells, and especially enriched in the HSPC CD34⁺ compartment.

The novel partitioning technique revealed the location of the clones within the BM niche showing spatial heterogeneity across the femur head within millimeters of space. GNAS^het and DNMT3A^het clones were found to potentially colocalize with the neoplastic JAK2V617F and TET2^HET clones, whereas the JAK2V617F clone without TET2^het clustered separately. Using imaging mass cytometry, the authors also visualized distinct clustering of CD71⁺ erythroblasts in the patient BM, possibly indicating separate localization of the neoplastic clone within the BM niche (Fig. 1C).

Although this article profiled one PV patient, the authors highlight that it provides a proof of concept of this novel technique to expand their investigation to additional samples. Indeed, this paper offers intriguing insights into both the clonal evolution of PV at the single-cell level and the spatial clonal architecture in the BM. This study opens up exciting new avenues to explore regarding understanding the complexities of clonal localization in the BM niche.

Authors’ Disclosures

No disclosures were reported.