Turbulence Enhances Fine-Tuning of Mitochondria Delivery in Megakaryocyte Maturation Contributing to Biogenesis of High Quality Platelets

Platelet (PLT) products have a short shelf life of 4-5 days but blood donations are limited. As a solution, we established induced pluripotent stem cell (iPSC)-derived immortalized megakaryocyte cell lines (imMKCLs) (Nakamura, Cell Stem Cell, 2014), which generated platelet-like particles (iPSC-PLTs) under static conditions, albeit with low yield and function. Subsequently, by developing turbulence-flow bioreactors, imMKCLs generated high yield of competent iPSC-PLTs (Ito, Cell, 2018). Meanwhile, it is still unclear the reason why turbulence stimulation into imMKCLs allow higher production efficiency and improved function.

Herein we found turbulence increases CD42b (GPIbα) expression in iPSC-PLTs, which was inversely correlated with phosphatidylserine (PS) expression levels (as represented by Annexin V binding). Only CD42b⁺ population of iPSC-PLTs responded to agonist stimulation, detected by P-selectin expression or PAC1 binding. We demonstrated functional CD42b⁺ iPSC-PLTs are determined by mitochondria activity in iPSC-PLTs, but not in imMKCLs.

Regarding CD42b expression, it is known that ADAM17 causes CD42b shedding upon exposure of phosphatidylserine on the cell surface (Sommer, Nature communications, 2016). Therefore, we examined whether mitochondria activity influence CD42b shedding. To test this, we added FCCP, a mitochondrial uncoupler depleting mitochondria membrane potential, in the absence or presence of shedding inhibitors in imMKCL culture. Depleting mitochondria activity decreased CD42b expression levels on iPSC-PLTs, independently of culture conditions. Pre-incubation with ADAM17 inhibitor or different type shedding inhibitors prior to FCCP treatment allowed to retain CD42b levels, suggesting that mitochondria function might regulate downstream ADAM17 activity or shedding mechanisms. This was confirmed by GPIX deficiency in imMKCLs, as no difference in mitochondria activity was found compared to the control.

Differences in CD42b levels were not seen in MKs during maturation, independent of culture conditions, therefore we suspected turbulence promote intact mitochondria delivery in imMKCLs. To examine this hypothesis, we used Mdivi-1, which inhibit Drp1 activity and thereby suppress mitochondria fission, to mimic impaired mitochondria delivery. Administration of Mdivi-1 to turbulence-exposed imMKCLs decreased CD42b expression in subsequently generated iPSC-PLTs. Using confocal microscopy, we also visualized fused mitochondria in imMKCLs upon addition of Mdivi-1, strongly suggesting that negative CD42b expression in iPSC-PLTs is due to impaired mitochondria delivery.

Next, to gain mechanistic insight at molecular levels, we differentiated imMKCLs in static or turbulence conditions and swapped the culture conditions on day 3 of the total 6 days of imMKCL maturation. It revealed that turbulence is required only after day 3 onwards, but not before. scRNA analysis and pharmacological examination results strongly suggested that increased HDAC5, 6, and 9 levels by turbulence stimulation are crucial for efficient mitochondria delivery into iPSC-PLTs.

We conclude that turbulence increases ‘fine-tuning’ of mitochondria distribution in imMKCLs and efficient delivery to PLTs with relatively sufficient activity via an HDAC mediated mechanism, where mitochondria activity is required for inhibition of phosphatidylserine exposure on the cell surface and of ADAM17 activity.