Single Cell Transcriptomics on Bone Marrow Biopsies Unveils Peculiar Bone Microenvironment Cell Subtypes and Highlights Alterations Correlated with High-Risk Features in Precursor Monoclonal Gammopathies and Multiple Myeloma Patients

Multiple myeloma (MM) can be preceded by monoclonal gammopathy of undetermined significance (MGUS) and smoldering MM (SMM). In the pathophysiology of MM, the bone microenvironment (BME) plays a pivotal role, but how BME changes contribute to tumor progression in MGUS and SMM remains unresolved. Previous data on BME are mainly produced in vitro from bone marrow (BM) aspirates, and few are on MGUS/SMM patients. To date, no single-cell RNA sequencing database of BME cells has been described in MM. The aim of this project was to characterize, for the first time at single cell level, the BME in patients with MGUS/SMM and newly diagnosed MM (NDMM) to identify high-risk features and alterations involved in tumor progression.

From 16 bone biopsies of MGUS, SMM, NDMM patients, we depleted CD235a⁺, CD45⁺, CD31⁺, and CD138⁺ cells to enrich the rare BME non-hemopoietic cells. The negative cell fractions were analyzed by scRNAseq. Data were generated on Chromium 10X Genomics. Seurat package in R was used for quality check and to identify differentially expressed markers across cell types and clinical conditions. Pathway and biological process (BP) activity inference was performed using EnrichR and GSVA R packages. Trajectory analysis was run with Monocle3 R package.

A total of 43154 BME cells were profiled (12.717 from MGUS, 17.230 from SMM and 13.207 from NDMM samples). Within the whole “first-of-its-kind” BME dataset, we identified 16 cell type clusters, all belonging to mesenchymal lineage: 10 clusters showed mesenchymal-stromal features (MSCs) and 4 osteoblast-like (OBs). The peculiar expression of several genes (DKK1, ITGA6, OPG, LEPR, IBSP, WISP2, PDPN, CTSK, CHI3L1) and increased specific BP activity (glycolysis-gluconeogenesis, arachidonic-acid-metabolism, direct ossification, regulation of hematopoietic stem cell proliferation) characterize subtypes of MSCs and OBs. Two further clusters included cells with marked expression of the lncRNA FTX and by IRF1 gene, respectively.

Notably, significantly decreased proportion of OB precursors and Mature OB cells, and concomitant significant increase of Proliferating MSC, were found in NDMM. In SMM samples we found an increase of a mesenchymal-cell cluster (MSC1) overexpressing NEAT1 compared to MGUS and a decrease of OB precursors clusters overexpressing LEPR compared to MGUS and MMD samples.

Functional analyses revealed that MGUS samples, in MSC clusters, negatively regulated IL-6 production and upregulated immune-stimulating pathways compared to MM. In NDMM samples, the Proliferating MSC cluster showed an increase in BPs involved in hematopoietic stem cell maintenance; concomitant, in MSC osteogenic and OB precursors LEPR+ cell clusters, they upregulated processes involved in mature B cell homeostasis and in the suppression of OBs development. The SMM samples presented a peculiar cell cluster, MSC osteogenic, with immune-modulatory features and also, interestingly, they showed in several cell clusters, an upregulation of BPs involved in osteoblast development, non-canonical Wnt signaling but also in bone remodeling, which supports the idea that SMM patients attempt to contrast the establishment of an pro-osteolytic bone microenvironment. Moreover, we have explored if the high-risk pre-malignant patients are characterized by BME alterations. High-risk MGUS/SMM samples are characterized by a downregulation of BPs involved the suppression of T-helper 17 cells and in the stimulation of the T-helper 1 cells and exhibited, in the MSCs clusters, an alteration of the glutamine/glutamate metabolism; on the other hand, high-risk SMM showed an increase of Proliferating MSC cells cluster, like NDMM, and an upregulation of ATP-synthase binding cassette transporters. Overall, these data support the hypothesis of a more immune-stimulatory and less MM-supportive BME in the precursor monoclonal gammopathies, which is remodeled with the increase of malignant clone and in the presence of high-risk features.

Finally, pseudotime analyses highlighted, for the first time in MM setting from bone biopsies, several and complex trajectories of differentiation from more immature cell clusters (undifferentiated MSC) to Mature OB cluster, confirming the ability of the dataset to snapshot the in vivo complexity of patient’s bone microenvironment, also highlighting a significant reduction of the MSC routes of differentiation from MGUS samples to NDMM.