Extended Data Fig. 9 | Epiblast development during embryo culture in the
3D condition. This figure relates to Fig. 4. a–d, Dynamic expressions of
pluripotent genes over human embryo development in the 3D condition. a, The
dynamic expressions of OCT4, TFCP2L1 and KLF4 during human embryo
development (3 out of 3 embryos). b, The dynamic expressions of OCT4, SOX17
and KLF17 during human embryo development (3 out of 3 embryos). Loss of
TFCP2L1, KLF4 and KLF17 at 10 d.p.f. (implantation stage) indicate the
pluripotent state transition of epiblasts. c, Expression of NANOG and PRDM14
at 10-d.p.f. human embryos (2 out of 2 embryos). d, Dynamic expressions of
OCT4, CD24 (a primed pluripotency gene) and KLF17 during human embryo
development (2 out of 2 embryos). e, t-SNE of pluripotent stem cells. To exclude
cells from the AME and intermediate state in the epiblast cluster, we excluded
the NANOG-negative cells and maintained 136 cells with high expression of
NANOG. f, The violin plots show the dynamics of naive, primed and common
pluripotency genes in pre-gastrulation embryos. All violins have the same
maximum width, black dot denotes the mean. AME and intermediate state cells
were excluded in the synthesis. *P < 0.05, **P < 0.01, ***P < 0.001, two-sided
Wilcoxon rank-sum test. g–j, PluriNetWork analysis of EPIs from ICM (g), pre-
EPI (h), post-EPI (i) and PSA-EPI (j) stages revealed that key pluripotency
regulators dominated the networks. k, DEGs during EPI development. GO
terms and representative genes in DEGs in the pairwise comparisons are
indicated. l–n, Scatter-plot comparison of the gene-expression levels between
ICM and pre-EPI (l), pre-EPI and post-EPI (m), and post-EPI and PSA-EPI (n). Key
genes are annotated (Supplementary Table 6). Red denotes upregulated; green
denotes downregulated; >twofold difference, uncorrected P < 0.01 (likelihood
ratio test) and median FPKM > 1 in one group. Scale bars, 50 μm (a, b), 100 μm (c)
or 25 μm (d).