Identifying GMP-grade hPSC lines with good MK output for future clinical-grade production. (A) Refining the seeding of hPSCs for FoP. The original protocol (Moreau et al,⁶  top panel) uses a 3D embryoid body (EB) to promote the initial mesoderm induction step (top panel). We progressed this toward a 2D system more suited to large-scale manufacturing, first using 2D cell clumps (middle panel), then single-cell seeding (bottom panel). Images (left) show light microscope images of day 0 cultures for each method. Scale bars, 500 μm. (B) Transitioning from 3D to 2D single-cell seeding does not compromise MK yield or purity. Bar graphs for 3 hPSC lines (QOLG, FFDK, and A1ATD1) seeded as clumps or single cells showing the number of MKs (expressed per undifferentiated starting hiPSC) obtained by day 20 of culture (left graph) and their purity by percentage of mature CD41⁺CD42⁺ cells (right). Mean ± 1× standard error of the mean (SEM); n = 4. (C) Screening GMP hPSC lines for MK output by FoP. Eighteen GMP hPSC lines were tested for their MK output at day 20 using the culture protocol shown in panel A (bottom panel) alongside 2 control iPSC lines. Bar graph showing MK output for the 9 GMP lines that had over 50% CD41⁺CD42⁺ cells, including 5 hESCs (lavender bars), 4 hiPSCs (light blue bars), and 2 control hiPSC lines (orange bars). Yields are expressed per 1.00E+05 undifferentiated starting cells (dotted line) plotted on a log₁₀ axis, Mean ± 1× SEM where n > 1. Numbers above each column are the mean yield of MKs per starting cell. (D) Cytospins of the control line QOLGand DELTA-3 at day 24 of differentiation showing large multinucleated MKs stained using Rapid Romanowsky. Scale bars, 50 μm. ns, not significant; SCF, stem cell factor; TPO, thrombopoietin.