Multi-Phase Ex Vivo Generation of Platelet-like Particles from CD34⁺ Cord Blood Cell-Derived Megakaryocytes

Currently, transfusable platelet units are solely derived from volunteer donors, but limited shelf-life and risk of infectious disease transmission present several logistical challenges for platelet storage and distribution at hospitals and blood centers. Thus, there is a growing interest in producing transfusable platelets in culture. Ex vivo platelet production is currently limited by the small quantity of input cells available for megakaryocytic (MK) differentiation, and low efficiency of generating platelets in vitro from differentiated MK cells. We present a serum-free, stroma-free, multi-phase culture process for production of CD41⁺/CD42⁺ platelet-like-particles (PLPs) from CD34⁺-selected umbilical cord blood (CB) cells. Our method greatly expands the number of input CD34⁺ cells towards the MK lineage, thus increasing the total number of CD41⁺/CD42⁺ mature MK cells and PLPs per input CD34⁺ cell by several fold compared to cultures without pre-expansion.

CD34⁺ CB cells were expanded using histone deacetylase inhibitor valproic acid (VPA) and cytokines (primary culture),transferred to low oxygen conditions for further expansion and MK commitment (secondary culture), and then transferred to a final stage where MK cells mature and are fragmented under shear conditions (tertiary culture). We demonstrate that expanded pools of CD34⁺ CB cells differentiate into mature MK cells and produce functional CD41⁺/CD42⁺ PLPs under in vitro shear conditions.

The length of the primary pre-expansion culture had a strong effect on CD41⁺/CD42⁺ presentation in subsequent culture steps. We evaluated the growth and differentiation of CD34⁺ cells for 16 independent CB units pre-expanded for 5 or 7 days in the presence of VPA. While 7-day VPA expansion decreased the peak percentage of CD41⁺/CD42⁺ cells in comparison to 5-day VPA expansion (E7: 21 ± 1.6% vs. E5: 30 ± 1.9% CD41⁺/CD42⁺ p=0.001), total MK cell production tended to be greater for 7-day vs. 5-day VPA expansion due to increased cell proliferation (E7: 381 ± 90 vs. E5: 290 ± 52 CD41⁺/CD42⁺ MK cells per input CD34⁺ cell). In comparison, unexpanded cultures produced 118 ± 29 CD41⁺/CD42⁺ cells per input CD34⁺ cell. Also, DNA content, as measured by percent high ploidy CD41⁺ cells, was positively correlated with length of pre-expansion (7-day VPA expanded: 13 ± 0.2%, 5-day VPA: 10 ± 0.5%, unexpanded: 7.1 ± 0.4% percent ≥8N CD41⁺ cells, p<0.001).

In secondary and tertiary culture, we evaluated the effect of shear force on proliferation, MK maturation, and PLP generation. Culturing cells in an orbital shaker increased the percentage of CD41⁺/CD42⁺ cells for VPA treated cultures (shear 35 ± 2.8% vs. static 24 ± 2.6%, p = 0.02), but decreased the number of total cells in culture (shear 1270 ± 298 vs. static 2030 ± 323 total nucleated cells per CD34⁺ input cell). Individual MK cells within a heterogenous bulk culture were shown to release CD41⁺/CD42⁺ PLPs asynchronously over the 6-day tertiary phase. Time-course studies of PLP sampling every 24 hours revealed that 2200± 640 CD41⁺/CD42⁺ PLPs were generated per input CD34⁺ cell (7-day VPA expansion, n=3) at peak concentration. The CD41⁺/CD42⁺ PLP production rate of 7-day VPA cultures was greater than that for unexpanded CB CD34⁺ cells (1100± 120 CD41⁺/CD42⁺ PLPs per input CD34⁺ cell, n=3). We analyzed the morphology and functionality of generated PLPs and showed that generated PLPs became PAC-1⁺ and CD62P⁺ when activated by thrombin receptor agonists. PLPs also showed typical formations of stretched microtubule coils and actin stress fibers when spread on fibrinogen-coated glass surfaces upon activation. Interestingly, PLPs released by 7-day VPA cultures were found to be smaller than PLPs released by unexpanded and 5-day VPA cultures, and more comparable in size to donor platelets. Our study presents a proof-of-concept that the number of functional PLPs can be significantly increased via a multi-phase MK production process.