Based on our analysis of XCI in
Based on our analysis of XCI in hESCs and hiPSCs we propose a model in which pluripotent stem papain show three states of XCI: lines with no XCI, lines with full XCI, and lines with partial XCI (Fig. 4). hESCs are derived from early embryos and therefore start out with two active X chromosomes. Thus, any change in XCI state implies an active process in which cells can inactivate or reactivate their X chromosomes (Shen et al., 2008; Silva et al., 2008). In contrast, hiPSCs are derived from somatic cells in which XCI is well established. During reprogramming the hiPSCs may immediately acquire a different XCI status, thus activating the silent chromosome entirely or partially at the tips of the chromosome, far from the XIST transcription site. However, it is also possible that at the pluripotent stage the cells can alter their XCI state in culture, similar to hESCs. Despite the similarity observed between hESCs and hiPSCs in the XCI status, one must consider the possibility that the biological process leading to the variation in XCI may be remarkably different between the two cell types. Thus, hiPSC lines that show partial or full XCI may have either undergone reactivation followed by partial or full inactivation, may have only been partially reactivated, or not activated at all on reprogramming (Fig. 4). Tchieu et al. (2010) have recently described female hiPSC lines which retain an inactive X chromosome. Supporting our method, analysis of the expression profiles of four of these cell lines categorized them as lines with full inactivation (data not shown).
In summary, by using meta-analysis we were able to show that the XCI status of both hESCs and hiPSCs can be divided into three categories. As for hiPSCs, this categorization has an added value showing that the X chromosome can be activated on reprogramming. Our analysis shows that determining XCI status of cells according to XIST expression alone may be misleading and include only dichotomic conclusions ignoring the new category of partial XCI suggested here. However, the importance of XIST in the inactivation process can be seen in our \"inactivation maps,\" showing that inactivation always includes areas adjacent to the transcription site of XIST. The findings presented here show that the XCI status of both hESCs and hiPSCs is not homogeneous, and suggest a more accurate classification of these cells. We believe that this classification, in addition to the suggested inactivation maps, might direct us to better understand the epigenetics and reprogramming of the X chromosome.
Materials and methods