Characterisation of the megakaryocyte proteome in patients with Philadelphia negative myeloproliferative neoplasms

Abstract Introduction: Megakaryocytes are considered to drive pathology in the classical Philadelphia-negative myeloproliferative neoplasms (MPNs). The distinct cellular and spatial morphological abnormalities of megakaryocytes in each MPN (primary myelofibrosis – PMF; polycythemia vera – PV; essential thrombocythemia – ET) aid their diagnostic classification, yet a characterisation of molecular differences is lacking. Their limited numbers within bone marrow (<0.05% of nucleated bone marrow cells) have to date precluded unbiased proteomic analysis of in-vivo megakaryocytes and previous studies have relied on analyses of cultured megakaryocytes. We undertook the first mass spectrometry-based proteomic study of in-vivo megakaryocytes isolated from humans with MPNs. Methods: We employed a deep visual proteomics (DVP) strategy combining high-resolution tissue imaging, laser-capture micro-dissection (LCM) and data-independent acquisition (DIA) mass spectrometry to profile megakaryocytes from archived formalin-fixed paraffin-embedded (FFPE) bone marrow trephine samples. Cohorts included 20 patients each with PMF, PV, and with ET, as well as age- and sex- matched controls (uninvolved bone marrow acquired during routine disease staging for B-cell lymphoma). Using artificial intelligence-assisted image analysis (AIVIA), we isolated 250 megakaryocytes per sample with the Leica LMD7 system. Proteomic analysis was conducted on an Orbitrap Astral mass spectrometer (Thermo Fisher) and the data subjected to a library-free DirectDIA search in Spectronaut (v 19.5). Differential protein expression between the groups was determined with ANOVA and pair-wise comparisons using limma (v.3.62.1), and pathway enrichment analysis was performed with Reactome software. Results: A total of 4875 (80%) proteins were reproducibly identified and carried through for statistical analysis. Principal component analysis revealed a clear disease-specific separation between megakaryocyte proteomic signatures. When compared to control megakaryocytes, there were 948, 57 and 83 differentially expressed proteins in PMF, PV and ET megakaryocytes, respectively. Two growth factors (EGFL6 and FGF2) were exclusively detected across all MPN megakaryocytes but absent in controls. Several proteins enriched in PMF — such as the senescence biomarker CDKN1A; the actin severing protein SCIN; the TGF-β related proteins LTBP4 and CSRP2; and the immune checkpoint marker CD276 (B7-H3) — were also undetected in control megakaryocytes. Five proteins (TOE1, HPSE, CCL5, SLC2A3 and ELOVL7) were increased in all MPN subgroup megakaryocytes when compared to control megakaryocytes, while Factor V, a plasma protein that is primarily endocytosed by megakaryocytes, was consistently reduced in all MPN subgroups. Reactome enrichment analysis identified differentially activated biological pathways, particularly those involving innate immunity in PMF and ET megakaryocytes (FDR 9.17x10-39 and FDR 4.52 x 10-11, respectively). This was characterized by upregulated complement cascade proteins (C1QBP, C9, C2, C1QB, C7, C8A, CFH, C8G, C1QC, C1QA, CFB) and interferon signaling proteins (PML, OAS3, HLADRA, FLNB, MAVS, NUP85, DHX9, CD44 and ISG15) in PMF megakaryocytes, and downregulated proteins involved in antigen processing and cross presentation of phagosomes in PMF and ET megakaryocytes (NCF2, S100A8, S100A9, CTSS, CYBB, NCF4, IKBKG). Additionally, PMF megakaryocytes showed upregulated RNA metabolism and extracellular matrix protein pathways (FDR 3.28 x 10-45 and FDR 2.11 x 10-16, respectively). The hemostasis pathway was upregulated in both PMF and ET megakaryocytes (FDR 9.37x10-8 and FDR 0.002, respectively). Lipid metabolism pathways, particularly PPAR-γ regulated proteins, were downregulated in PMF and PV megakaryocytes (FDR 0.03 and FDR 1.8x 10-4, respectively), compared to control megakaryocytes. Conclusion: This study is the first to conduct a comprehensive proteomic analysis of human in-vivo differentiated megakaryocytes, and apply it to confirm that the morphological differences seen across the megakaryocytes of PMF, PV and ET are reflected by distinct disease-specific signatures. Our data identified pathways that can be further explored and uncovered potential biomarkers that could aid diagnostic decisions and lead to new therapeutic targets for MPNs.