Investigating lipid mechanisms of MSC immune modulation using metabolic and phenotypic single cell profiling

Abstract

Background & Aim The immunosuppressive capabilities of human mesenchymal stromal cells (MSCs) have led researchers to explore their potential as therapeutics targeting numerous autoimmune diseases. However, MSC therapies have had limited success so far in clinical trials. Their immune effects are mediated through a complex combination of direct receptor-mediated interactions, vesicle communication and cytokine release mechanisms, making it difficult to identify a unifying underlying biological mechanism of action that could be leveraged to monitor and enhance cell potency. Here, we investigate the role of membrane lipids, a key cell membrane component, associated with these immunosuppressive mechanisms. We aim to evaluate the lipid metabolic profiles of MSCs with known variability in immune regulation and explore the phenotypic changes that occur as a result of differences in lipid signaling. We use longitudinal imaging strategies to directly monitor the associations of cell-cell receptor binding, vesicle release, cytoskeletal network and secretory pathways with immune-related cell activities. Overall, this study aims to directly link lipid metabolism with MSC immunosuppressive behavior using mass spectrometry and image-based approaches, which could be further used to monitor and predict immune response during biomanufacturing of therapeutic cells. Methods, Results & Conclusion MSCs were obtained from cultures with varying immune suppressive activity and secretory profiles. Lipidomic analysis conducted on bulk MSC populations using HPLC/MS/MS revealed several key differences between high and low potency donors. High potency lines showed a negative correlation between sphingomyelin (SM) and ceramide (Cer) whereas the low potency line had a positive correlation indicating that conversion between SM and Cer was associated with cell immune suppression. Multivariate phenotypic imaging analysis of cytoskeletal and vesicle markers showed a strong correlation to cell SM conversion rates. To study heterogeneity of the culture subpopulations, we used Matrix Assisted Laser Desorption/Ionization Mass Spectrometry Imaging to obtain single cell lipid profiles and used advanced deep learning on single cell fluorescent imaging features to understand phenotypic outcomes of lipid signaling. Our findings could have profound implications in understanding immunosuppressive activity of MSCs that could be tuned to enhance the therapeutic efficacy of MSC cultures.