17Background: Multiple myeloma (MM) is a hematological malignancy characterized by the 18 clonal expansion of malignant plasma cells. Though durable remissions are possible, MM is 19 considered incurable, with relapse occurring in almost all patients. There has been limited 20 data reported on the lipid metabolism changes in plasma cells during MM progression. Here, 21we evaluated the feasibility of concurrent lipidomics and proteomics analyses from patient 22 plasma cells, and report these data on a limited number of patient samples, demonstrating the 23 feasibility of the method, and establishing hypotheses to be evaluated in the future. 24Methods: Plasma cells were purified from fresh bone marrow aspirates using CD138 25 microbeads. Proteins and lipids were extracted using a bi-phasic solvent system with 26 methanol, methyl tert-butyl ether, and water. Untargeted proteomics, untargeted and targeted 27 lipidomics were performed on 7 patient samples using liquid chromatography-mass 28 spectrometry. Two comparisons were conducted: high versus low risk; relapse versus newly 29 diagnosed. Proteins and pathways enriched in the relapsed group was compared to a public 30 transcriptomic dataset from Multiple Myeloma Research Consortium reference collection 31 (n=222) at gene and pathways level. 32Results: From one million purified plasma cells, we were able to extract material and 33 complete untargeted (~6000 and ~3600 features in positive and negative mode respectively) 34 and targeted lipidomics (313 lipids), as well as untargeted proteomics analysis (~4100 35 reviewed proteins). Comparative analyses revealed limited differences between high and low 36 risk groups (according to the standard clinical criteria), hence we focused on drawing 37 comparisons between the relapsed and newly diagnosed patients. Untargeted and targeted 38 lipidomics indicated significant down-regulation of phosphatidylcholines (PCs) in relapsed 39 MM. Although there was limited overlap of the differential proteins/transcripts, 76 40 Page 3 of 29 significantly enriched pathways in relapsed MM were common between proteomics and 41 transcriptomics data. Further evaluation of transcriptomics data for lipid metabolism network 42 revealed enriched correlation of PC, ceramide, cardiolipin, arachidonic acid and cholesterol 43 metabolism pathways to be exclusively correlated among relapsed but not in newly-44 diagnosed patients. 45
Conclusions:This study establishes the feasibility and workflow to conduct integrated 46 lipidomics and proteomics analyses on patient-derived plasma cells. Potential lipid 47 metabolism changes associated with MM relapse warrant further investigation. 48