Mouse model of chromosome mosaicism reveals lineage-specific depletion of aneuploid cells and normal developmental potential

Abstract

Most human pre-implantation embryos are mosaics of euploid and aneuploid cells. To determine the fate of aneuploid cells and the developmental potential of mosaic embryos, here we generate a mouse model of chromosome mosaicism. By treating embryos with a spindle assembly checkpoint inhibitor during the four- to eight-cell division, we efficiently generate aneuploid cells, resulting in embryo death during peri-implantation development. Live-embryo imaging and single-cell tracking in chimeric embryos, containing aneuploid and euploid cells, reveal that the fate of aneuploid cells depends on lineage: aneuploid cells in the fetal lineage are eliminated by apoptosis, whereas those in the placental lineage show severe proliferative defects. Overall, the proportion of aneuploid cells is progressively depleted from the blastocyst stage onwards. Finally, we show that mosaic embryos have full developmental potential, provided they contain sufficient euploid cells, a finding of significance for the assessment of embryo vitality in the clinic.

Figure 1. Reversine treatment induces aneuploidy in mouse embryos.

(a) Uniform aneuploid embryos contain a single identical abnormality in their chromosome status and mosaic aneuploid embryos contain blastomeres with differing chromosome constitutions caused by mitotic error post fertilization. (b) Blastomeres treated with reversine during division (n=81) have significantly greater rates of chromosome missegregation during mitosis than controls (n=72; ***P<0.001, χ²-test). (c) Blastomeres treated with reversine have a shorter metaphase than controls (***P<0.001, Mann–Whitney U-test). These effects are reversible on washout of the drug (n=32). (d) Validation of aneuploidy in reversine-treated embryos, by single-cell genome sequencing. Each circos plot represents the genomic constitution of an embryo in all its cells, with blastomeres presented as rings and the chromosomes as segments. All results of the single-cell genome sequencing are provided in Supplementary Data 1.

Figure 2. Preimplantation development of aneuploid embryos.

(a) Embryos were treated with reversine (n=20) and analysed for expression of lineage-associated genes at the early blastocyst stage, compared with controls (n=24). No difference in the number of Cdx2-positive and Nanog-positive cells was observed. (b) Embryos were treated with reversine (n=26) and analysed for lineage specification at the expanded blastocyst stage, compared with controls (n=31). Reversine-treated embryos have a significantly reduced cell number in all lineages (*P<0.05 and ***P<0.001; Student's t-test), but the segregation of cell lineages was not affected. (c) Embryos were injected with Mad2 siRNA (n=22) or control siRNA (n=20) and analysed at the expanded blastocyst stage. Mad2 siRNA-injected embryos had significantly fewer cells than control siRNA-injected embryos (*P<0.05; Student's t-test), but this effect was not as clear as that seen with reversine treatment. Scale bars, 20 μm. Error bars represent s.e.m.

Figure 3. Pre-implantation development of chimeric mosaic embryos.

(a) Chimeras containing a 1:1 ratio of control and reversine-treated blastomeres were generated at the eight-cell stage. (b) Chimeras expressing Histone H2B-GFP (n=20) were imaged for 2 days to the late blastocyst (BC) stage and the movements and divisions of all cells were tracked (n=1,079 cells). Scale bar, 20 μm. (c) Example of lineage tree showing the fate of control and reversine-treated blastomeres in a chimeric embryo. Cell tracking data are provided in Supplementary Data 2. (d) The number of abnormal cells decreases in the ICM during blastocyst expansion (*P<0.05 and **P<0.01; Student's t-test). (e) Average lengths of the fifth and sixth cell cycles for control and abnormal cells in inside and outside positions. Outside cells have a significantly longer sixth cell cycle than controls (***P<0.001; Student's t-test). (f) The proportion of slow outliers is greatly increased in the outside cells of the abnormal clone, compared with controls (***P<0.001, χ²-test). Error bars represent s.e.m.

Figure 4. Elimination of abnormal cells by apoptosis during blastocyst development.

(a) Sequential images from a time-lapse series of a 1:1 reversine chimera (control cells in this embryo are labelled with Tomato fluorescent protein), showing apoptosis of a reversine-treated cell (arrow), followed by engulfment of the apoptotic debris into an efferosome by a neighbouring control cell (star). Scale bar, 20 μm. (b) The percentage of blastomeres undergoing apoptosis is significantly increased in the abnormal clone (n=36 embryos, 634 ICM cells, 1,346 TE cells; ***P<0.001; Fisher's test) Error bars represent s.e.m. (c) Sister blastomeres (n=321 pairs) are more likely to have similar fates than would be expected by chance (**P<0.01 and ***P<0.001; binomial test).

Figure 5. Loss of reversine-treated cells in postimplantation chimeric mosaic embryos.

Chimeras that had been transferred to recipient females were recovered at early postimplantation stages (E6.5–E7.5) and the contribution of reversine-treated cells to the egg cylinder assessed. In these examples, the control clone is marked by the expression of Histone H2B-GFP. In controls, both clones contribute to the embryo. In the reversine chimeras, the majority of each embryo originates from the control clone (demonstrated by the lack of DAPI-positive/GFP-negative cells). Scale bars, 50 μm.

Figure 6. Successful fetal development of chimeric mosaic embryos.

Chimeras that had been transferred to recipient females were recovered at E13.5. (a) Examples of whole-mount images (Histone H2B-GFP expression marks the reversine-treated clone). Scale bars, 2 mm. Note: the fetuses were imaged using different microscopes, the difference in colour is reflective of the microscope used, not differences between the two groups. (b) Fetal and placental biopsies from the E13.5 conceptuses shown in A. Scale bars, 100 μm. In all cases, where a Histone H2B-GFP clone was present, the cells were evenly distributed throughout the tissues. Both placental and fetal biopsy findings were consistent with the whole-mount GFP images.

Figure 7. Effects of pre-implantation chromosome mosaicism on embryo development and survival.

Reversine-treated embryos formed blastocysts but failed to develop past implantation. Increasing the proportion of control blastomeres in the embryo rescued the lethal phenotype. Numbers represent the viability of early postimplantation embryos that had successfully implanted. Early postimplantation rescue resulted in complete rescue of developmental potential, with E13.5 and live birth rates equivalent to viability at E6.5–7.5.