X-Reactivation Impacts Human iPSC Differentiation Potential Towards Blood

To determine novel key regulators that direct ES/iPS cell differentiation to hematopoietic lineages, we compared the gene expression profiles of multiple iPS cell lines with defferential blood forming capacity. We generated multiple iPS cell lines from amniotic fluid derived mesenchymal stromal cells (AF-iPS) which differentiated towards hematopoietic lineages using our standardized and highly reproducible differentiation protocol. Of the 9 AF-iPS cell lines derived from an individual female patient, the average efficiency of CD45+ hematopoietic cells was 14.2 +/- 9% (range 1.6 to 26.3%).

To elucidate the possible reasons for this diversity in efficiency, we grouped the AF-iPS cell lines on the basis of lowest and highest blood differentiation capacity and compared their gene expression profiles by microarray. We found very few changes above 1.5-fold, but interestingly, among the 11 genes that were over-expressed in the AF-iPSC lines with poor blood differentiation efficiency 10 were located on X chromosome, and the remaining one reported to be involved in Notch signalling.

A combination of cumulative sum analysis and the location of differentially expressed genes on the X chromosome identified putative regions of reactivation at multiple, but distinct locations. The possibility of X-reactivation in these female lines was reinforced further where lower levels of XIST were seen in AF-iPSC lines shown to have low blood forming potential, however only half of the iPS cell lines with high blood differentiation capacity showed normal XIST expression when compared to the amniotic fluid mesenchymal starting cell material.

To determine whether the block in differentiation was tissue specific we tested the differentiation capacity of the AF-iPSC lines towards neuronal lineages. Intriguingly we found neural cell differentiation was not hampered within all lines with poor blood potential suggesting that the over-expression of genes as a consequence of X-reactivation can impart a specific negative effect on differentiation towards the blood lineages from pluripotency stage, while not having an effect on neuronal cell development.

To further define the source of this block, we have begun working knocking down the overexpressed genes on X chromosome in lines with poor blood differentiation potential to determine whether the efficiency can be increased (or fully rescued) with one, or a combination of these 11 candidate genes. These results have implications for the identification and selection of female iPS lines suitable for therapeutic purposes.