In recent years, the interplay of leukemic cells with the bone marrow niche has emerged as a potential contributing factor for the survival, therapy resistance and relapse-initiating capacity of leukemic cell subsets in ALL. This is known as early, environment-mediated drug resistance (EMDR). The role for leukemia-derived extracellular vesicles (LEVs) in this interplay remains largely unexplored. In vitro studies and PDX models have demonstrated that LEVs are internalized by bone marrow stromal cells (BMSC), leading to BMSC reprogramming and metabolic switch to glycolysis, with beneficial effects on leukemic cell survival as a result. This suggests that LEVs could also play a role in leukemic cell survival in ALL during chemotherapy, however, studies examining LEVs in patients are warranted.

We performed fractionation by differential ultracentrifugation and subsequent mass spectrometry analysis of bone marrow plasma samples (n=12) collected at time of diagnosis (n=6, one sample was processed at 30, 60 and 120 min after sampling to assess the potential effect of contamination with platelet-derived EVs), during (n=2) and after induction therapy (n=4) in patients with B-cell precursor ALL (n=4). We isolated four fractions: A (2,000g pellet): Enriched in platelets/apoptotic bodies, M (20,000g pellet, enriched in microvesicles: large, membrane-derived EVs), E (150,000g pellet, enriched in exosomes), and S (150,000g supernatant, mostly soluble proteins), as well as R (raw untreated plasma), and we subjected these to mass spectrometry-based quantitative proteomics in 3 technical replicates. We examined protein composition and enriched pathways in the fractions and investigated associations of EV protein levels with the size of the leukemic cell fraction in the bone marrow. Flowcytometric data was used to confirm cell distributions/surface marker expression.

Examining the protein composition of the five fractions (A, M, E, S, and R) showed that although the highest number of proteins was found in the A fraction (2803 proteins), the M fraction (2477 proteins) had the highest proteome diversity (Figure 1). Focusing on the M fraction, we compared diagnostic with follow up samples (blast percentage day 0: 92, 95, and 72, day 15: 0.4 and 1.3, day 29: 0.05, 0.2, 0, and 0.001, one day 0 sample excluded due to poor readout quality). We found significant upregulation of B-cell associated proteins at diagnosis, incl. HLA-DR, CD19, and CD22, along with the stem cell marker CD34, while T-cell and platelet-associated proteins did not show this pattern. Flow cytometry demonstrated only very few healthy B-cells in the samples at time of diagnosis (5, 0.8, and 0.6% vs. 25, 3, 32 and 21% on day 29) and showed all the leukemic clones to be CD34positive, confirming the BCP-ALL leukemic origin of the EVs.

Examining the proteins most closely correlated to the percentage of blasts across all the samples, we found several proteins associated with poor survival, cell migration/adhesion and chemoresistance in acute leukemia, including CD97 (r=0.85, p=0.006), PLXNB2 (r=0.72, P=0.037), and ITGA4 (r=0.6, p=0.097).

Gene set enrichment analysis of the proteins showing the most prominent differences between day 0 and day 15/29 using Reactome identified a variety of metabolic processes, but also oncogenic MAPK signaling (NES 1.94, p-adj=0.005), integrin signaling (NES 1.94, p-adj=0.003), and regulation of expression of SLITs and ROBOs (NES 2.1, p-adj=3x10-7) as enriched among the day 0 samples. Interestingly, we recently found genes within the latter two selectively upregulated in leukemias with poor induction therapy response (Modvig et al, Molecular Oncology, 2022). Taken together, this suggests that LEVs indeed could play a role in leukemic cell adherence and EMDR in the bone marrow niche.

In conclusion, LEV content can be reliably characterized in bone marrow plasma of patients with ALL through differential ultracentrifugation and mass spectrometry-based proteomics. LEV protein composition reflects processes such as leukemic adhesion/migration, survival and drug resistance, carefully suggesting a role for LEVs in EMDR, but larger studies in patients are warranted. The identification of LEVs and key pathways in the leukemic-stromal cell interplay/EMDR can lead to development of new drugs blocking this cross-talk, potentially contributing to an improved relapse-free survival in patients with ALL.

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution