Synthetic miRNA Switch Technology Elucidates Heterogeneity in Regulation of Immortalized Megakaryocyte Cell Lines, Associated with Improvement of Platelet Generation Efficiency for Clinical Use

We have developed a robust in vitro platelet generation system using a self-renewing megakaryocyte (MK) cell line, imMKCL, which was established by introducing three doxycycline (DOX)-inducible transgenes, c-MYC, BMI1, and BCL-XL into human induced pluripotent stem cells (iPSCs)-derived megakaryocyte-erythroid progenitor (MEP) cells for immortalization (Nakamura et al. Cell Stem Cell 2014). This system is suitable for clinical application since imMKCLs proliferate from 10⁶ to over 10¹⁰ cells in only 20 days in the presence of DOX (self-renewing stage), and when DOX is removed, imMKCLs enlarge, form proplatelets and release platelets (maturation stage). However, although imMKCLs display homogenous appearance at their self-renewing stage, only a minor population eventually produces platelets at a level comparable to MKs in vivo. Therefore, it is critical to resolve this heterogeneity in terms of maturation to improve the efficiency of platelet production for their practical use.

In that regard, we herein apply microRNA (miRNA) switch technology to identify a subpopulation of imMKCLs with high capacity to produce platelets upon maturation. The synthetic miRNA responsive switches could be used for detecting and efficiently purifying cell populations based on endogenous miRNA activity (Miki et al. Cell Stem Cell 2015).

First we performed a screening of miRNA switches to investigate the miRNA activities in imMKCLs at the self-renewing stage. Twenty-six miRNA candidates turned out to have some activities in imMKCLs in the presence of DOX. Among those miRNAs, we noticed that Let-7a-5p showed a clear distinguishable pattern with about 10% of the cells in the lower Let-7a-5p activity group (Group L), while other large majority in the higher activity group (Group H). When we sorted the two groups by flow cytometer even in the presence of DOX (self-renewing stage), Group L population in imMKCL revealed to produce significantly more platelets than Group H population (average of 9.9 vs 6.0 platelets per each MK, p<0.01) after DOX depletion. Transfecting Let-7a-5p inhibitor into Group H cells increased platelet production equivalent to the level of Group L, suggesting that Let-7a-5p suppresses the maturation of imMKCLs.

Next, we found that the expression level of LIN28A but not LIN28B, a putative downstream target of c-MYC for self-renewal in imMKCL, in Group L was 8.3 times higher than that in Group H (p<0.01), where LIN28A is reported to downregulate the expression of Let-7a-5p by inhibiting the processing of Let-7 miRNA precursors by DICER (Piskounova et al. Cell 2011). This difference was not affected by Let-7a-5p inhibitor, in line with the role of LIN28A as the upstream regulator of Let-7.

Finally, to address the targets of Let-7a-5p, we then performed RNA-seq analysis to compare gene expression profiles between Group L and Group H. We found that 691 genes were upregulated more than twice in Group L than in Group H. Among these, 48 genes were predicted to be the targets of Let-7a-5p according to three miRNA target prediction databases, miRSystem, miRDB, and miRSearch. Validation of 48 genes by qPCR confirmed the upregulation of several genes selectively in platelet-producing imMKCLs. Accordingly, transfecting Let-7a-5p inhibitor into Group H cells upregulated the expressions of these confirmed genes, strongly suggesting that they are the candidate targets of Let-7a-5p regulating maturation of imMKCLs.

In conclusion, miRNA switch technology enabled us to purify an imMKCL population with high platelet production capability based on the activity of Let-7a-5p. Furthermore, this study revealed the heterogeneity of miRNA expressions in imMKCLs at their self-renewal stage, and suggests that miRNA-mediated gene regulation at this stage affect their maturation afterwards.