A system to analyze the initiation of random X-chromosome inactivation using time-lapse imaging of single cells

Abstract

In female eutherian mammal development, X-chromosome inactivation (XCI) of one of the two X chromosomes is initiated early. Understanding the relationship between the initiation of XCI and cell fate is critical for understanding early female development and requires a system that can monitor XCI in single living cells. Traditional embryonic stem cells (ESCs) used for XCI studies often lose X chromosomes spontaneously during culture and differentiation, making accurate monitoring difficult. Additionally, most XCI assessment methods necessitate cell disruption, hindering cell fate tracking. We developed the Momiji (version 2) ESC line to address these difficulties, enabling real-time monitoring of X-chromosome activity via fluorescence. We inserted green and red fluorescent reporter genes and neomycin and puromycin resistance genes into the two X chromosomes of PGK12.1 ESCs, creating a female ESC line that retains two X chromosomes more faithfully during differentiation. Momiji (version 2) ESCs exhibit a more stable XX karyotype than other ESC lines, including the parental PGK12.1 line. This new tool offers valuable insights into the relationship between XCI and cell fate, improving our understanding of early female development.

Fig. 1

X-chromosome instability in Momiji (version 1) ESCs during culture for differentiation and maintenance. (A) Momiji (version 1) ESCs were maintained with repeated passages without drug selection (upper panels, total passage number 19) and after drug selection using puromycin and neomycin (lower panels, total passage number 19 including 6 days with drug selection). (B) A scheme for inducing Momiji (version 1) ESC differentiation. Drugs (neomycin and puromycin) were added only during ESC preculture and were removed from the medium after cell differentiation induction. This is because rXCI occurs after differentiation induction and one of the X chromosomes is inactivated, causing the cells to become drug resistant but only to puromycin or neomycin. MEK and GSK3 inhibitors are referred to as 2i. (C) Differentiation of Momiji (version 1) ESCs at passages 5, 10, and 20. (D) FACS analysis of Momiji (version 1) ESCs on day 6 of differentiation. (E) DNA-FISH analysis with an X-specific probe (No. MXO-10; Chromosome Science Labo) in Momiji (version 1) ESCs before and after differentiation. Three cell culture samples were hybridized independently. At least 100 nuclei were counted in each sample.

Fig. 2

Targeting of fluorescence reporter genes in PGK12.1. (A) Schematic diagram of the Momiji (version 2) ESC genotype. (B) Strategy for targeting eGFP and mCherry reporters at the Pgk1 locus. Genotyping of eGFP (C) and mCherry (D) in isolated cell lines (with grouping of gels cropped from different parts of the same gel or from different gels). Uncropped electrophoresis gel images are included in Figs. S6 and S7. (E) qPCR analysis of eGFP and mCherry in the genomic DNA of Momiji (versions 1 and 2). The genomic DNA obtained from Momiji (version 1) ESCs cultured under drug selection for 3 days was used as a control; this DNA had one copy insertion of the reporter gene in the X chromosome. (F) DNA-FISH analysis of Momiji (version 2) and parental PGK 12.1 ESCs using an X-specific probe (No. MXO-10, Chromosome Science Labo). Momiji (version 2) ESCs (P56) were cultured under undifferentiated conditions with drug selection (neomycin and puromycin) for 3 days and subsequently analyzed using DNA-FISH. The analysis used parental PGK12.1 ESCs (P31) without drug selection. Three cell culture samples were hybridized independently. At least 100 nuclei were counted in each sample. Differences between Momiji (version 2) and PGK12.1 on day 0 were determined using a Fisher exact test.

Fig. 3

Improved X-chromosome stability of Momiji (version 2) ESCs during ESC differentiation. (A) A scheme for inducing differentiation in Momiji ESCs (versions 1 and 2). Momiji (version 1) ESCs were established under FBS + 2i/LIF conditions for efficiency. The cells (P5) were then used for experiments in which differentiation was induced without large expansions to avoid an increase in the passage number. By contrast, the parental strain PGK12.1 of Momiji (version 2) ESCs was established under conventional culture conditions, FBS-LIF, as described in the original report. The same conditions were used for the Momiji (version 2) ESC culture experiments. Momiji (version 2) ESCs (P55) were used in this study. Temporal fluorescence analysis of eGFP and mCherry in Momiji (versions 1 and 2) ESCs over the period of differentiation by (B) fluorescence microscopy and (C) flow cytometry. (D) Temporal analysis of differentiation marker genes in Momiji (versions 1 and 2) ESCs on days 0, 6, and 9 of differentiation by RT‒qPCR (n = 3 or 4). (E) We conducted temporal analysis of Xist cloud formation in Momiji (versions 1 and 2) ESCs throughout differentiation using RNA-FISH. At least 40 nuclei were counted in each sample. Designations of the Xist pattern in each nucleus: None, no detectable Xist signal; S1, single spot signal; S2, two spot signals; C1′, large and faint single cloud; C1, single cloud; C2, two clouds. (F) DNA-FISH analysis of the X-chromosome number in Momiji (versions 1 and 2) ESCs using an X-specific probe (No. MXO-10, Chromosome Science Labo) before and after the period of differentiation. Three cell culture samples were hybridized independently. At least 100 nuclei were counted in each sample.

Fig. 4

Random XCI lineage tree from a single cell of Momiji (version 2) ESCs. (A) Still images captured from video sequences of live-cell imaging during differentiation of Momiji (version 2) ESCs. (B) Lineage tree of Momiji (version 2) ESCs during differentiation. The cell division was tracked, and the cell color was determined manually. All progeny cells, which were derived from a single cell marked with *, died during cell differentiation. All differentiated cells derived from one yellow cell (marked with *1) turned green, indicating that Xi choice was already determined in the *1 cell. Similarly, Xi choice was determined for yellow cells (*2), from which all the derived cells turned red. Xi choice was not determined in yellow cells (*3), which are progenitors of yellow cells *1 and *2. These patterns predict that the Xi choice is determined during cell proliferation between cells *1/*2 and *3 (i.e., approximately the first and second days after inducing differentiation). (C) Summary of the relationship between the rXCI process and cell fate in a single cell.