A deep dive into the bone marrow plasma proteome by optimized extracellular vesicle enrichment Free

Communication between leukemic cells and the bone marrow microenvironment is thought to play a pivotal role in acute lymphoblastic leukemia (ALL), contributing to leukemic cell survival, therapy resistance, and relapse. This communication is often mediated by proteins via direct cell–cell interactions, secreted signaling molecules, and by extracellular vesicles (EVs).

Mass spectrometry (MS)-based proteomics is a powerful analytical technique for identifying and quantifying proteins from complex biological matrixes. However, the high dynamic range of protein levels in complex biofluids like bone marrow plasma - spanning several orders of magnitude - limits the detection of low-abundance proteins. To address this challenge, EV enrichment prior to MS analysis efficiently reduces levels of high abundance proteins, allowing for improved detection of proteins of lower abundance.

This positions EVs as a dual-purpose mediator in proteomic studies of the leukemic microenvironment; as components of the pre-analytical workflow aiming for the most comprehensive proteome, and as biologically relevant messengers involved in leukemic-niche signaling. However, the choice of EV isolation method is a key determinant of data quality with direct impact on the sensitivity for protein detection and on the biological interpretability of results.

Here, we present a comparative evaluation of EV enrichment strategies for optimal proteomic profiling of bone marrow plasma from patients diagnosed with ALL.

Three EV enrichment techniques were evaluated: Differential ultracentrifugation (UC), size exclusion chromatography (SEC), and strong anion exchange (SAX) magnetic bead-based isolation. For each technique, pooled bone marrow plasma collected on day 71 post-diagnosis and processed within 30 minutes after collection to minimize contamination from platelet-derived EVs, was used in triplicates. Firstly, the amount of input (50 µL, 200 µL or 500 µL) was evaluated to determine the minimal volume required without compromising results as bone marrow plasma is a limited sample type. The 50 µL input resulted in a significantly reduced number of identified proteins, whereas no substantial difference was observed between the 200 µL and 500 µL inputs. Thus, 200 µL was applied for subsequent evaluations except for SEC where 150 µL was applied due to column specifications. Unprocessed plasma (Raw) was analyzed in parallel triplicates for comparison.

Isolated EVs were digested by trypsin and analyzed using a state-of-the-art Orbitrap Astral mass spectrometer in data-independent acquisition (DIA) mode. Among the enrichment strategies, UC yielded the highest median number of quantified proteins (5058), followed by SAX (3333) and SEC (2856). In contrast, the Raw plasma samples yielded a significantly lower number of proteins (median: 915). Unique protein profiles were observed for each method: 1724 proteins were exclusively detected in UC-enriched samples, compared to only 850 and 108 unique proteins in SAX and SEC samples, respectively.

Important EV markers (CD9, CD63 and CD81) were consistently detected in all EV-enriched samples, regardless of the method used, but were not detected in Raw plasma. Additional biologically relevant markers, including the stem cell marker CD34 and the B-cell associated protein HLA-DR, were also quantified in the EV-enriched samples, but not detected in Raw plasma. The detection of membrane-bound proteins such as CD markers and HLA molecules in EV enriched samples confirms the successful isolation of EVs.

Notably, the B-cell marker CD19 and the chemokine receptor CXCR4 - key to cell migration within the bone marrow niche - were quantified exclusively in UC-enriched samples, supporting that UC provides the highest sensitivity, enabling the quantification of a broader range of proteins of interest.

In conclusion, this study demonstrates enrichment of EVs as a successful strategy for more comprehensive proteome analysis of bone marrow plasma in ALL and underlines the EV enrichment method as central for the discovery of biologically relevant proteins. Among the tested methods, differential ultracentrifugation provided the most comprehensive proteomic coverage, highlighting a potential for both biomarker discovery and mechanistic studies of leukemic cell–niche interactions. In addition to the full, optimized enrichment strategy, corresponding proteomic findings from diagnostic single patient samples will be presented.