Myeloma heterogeneity at cell resolution

In this issue of Blood, Poos et al¹ show how integrative multiomics at the single-cell level can track and characterize distinct subclones at single-cell resolution to study the emergence of resistance and identify new targets. The authors combined bulk whole genome sequencing data with single-cell RNA and assay for transposase-accessible chromatin sequencing (ATAC-seq) data with longitudinal samples from 15 patients with relapsed/refractory disease to reconstruct the clonal structures at multiple time points.

For more than a decade, scientists have evaluated the multiple myeloma (MM) genome from patients with precursor conditions to heavily treated relapsed refractory disease using genomic interrogation.^2,3 Although these studies have helped clinicians and scientists identify new biomarkers, targets, and resistance pathways, they have also revealed significant patient heterogeneity.^4,5 When does this heterogeneity arise? Are the alterations present at initial diagnosis or are they acquired? Because these studies used bulk sequencing platforms, in which all cells from a patient are pooled together before profiling, it was difficult for scientists to explore these questions. However, the tools used for genomic studies are also evolving,^6,7 as demonstrated by Poos et al.

The authors identified mitochondrial mutations using the chromatin accessibility data (single-cell ATAC sequencing) and used those mutations to refine the subclone evolution patterns predicted based on copy number alterations and mitochondrial DNA mutations.⁸ The branches in the resulting evolution tree provide a clearer picture of the epigenomic and transcriptomic differences across clones and time points, and it suggest that converging adaptation of existing subclones before treatment might be responsible for acquired resistance, which is independent of treatment.

By comparing shrinking and growing clonal populations after treatment, the authors identified the upregulation of CD44, a cell adhesion molecule, as a potential target for the MCL1 inhibitor.⁹ This finding was validated in another patient data set and by analyzing the interaction between microenvironment cells and MM cells. The critical point was that not all subclones had the same level of interactions with different cell types in the bone marrow.

This study showed that combining techniques provides substantial advantages for in-depth evaluations^7,10 and opens doors to many future applications, such as evaluating cells from precursors to newly diagnosed MM, evaluating the effects of specific treatments, evaluating the effect of MM cell targeting therapies vs immunotherapies, and comparing in vivo and in vitro models. However, there are still technical areas needing refinement. The authors had difficulty studying branching evolution patterns because of the limitations of existing single-cell platforms. Some of these difficulties may improve by simply sequencing more cells or improving strategies for combining MM cells and bone marrow microenvironment interactions. Alternatively, spatial technologies may be found to be useful in future studies. Mutations, which may have a significant role in MM subclonal structures, were not studied here because of the limitations of single-cell methodologies; however, combining whole-transcriptome sequencing with ATAC-seq or combining DNA and RNA data at a single-cell level may help future studies of mutation-driven subclones.¹⁰ This study is a glimpse of the future, in which single-cell studies will help in overcoming MM.

Conflict-of-interest disclosure: The author declares no competing financial interests.