Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Letter
  • Published:

Self-organization of the in vitro attached human embryo


Implantation of the blastocyst is a developmental milestone in mammalian embryonic development. At this time, a coordinated program of lineage diversification, cell-fate specification, and morphogenetic movements establishes the generation of extra-embryonic tissues and the embryo proper, and determines the conditions for successful pregnancy and gastrulation. Despite its basic and clinical importance, this process remains mysterious in humans. Here we report the use of a novel in vitro system1,2 to study the post-implantation development of the human embryo. We unveil the self-organizing abilities and autonomy of in vitro attached human embryos. We find human-specific molecular signatures of early cell lineage, timing, and architecture. Embryos display key landmarks of normal development, including epiblast expansion, lineage segregation, bi-laminar disc formation, amniotic and yolk sac cavitation, and trophoblast diversification. Our findings highlight the species-specificity of these developmental events and provide a new understanding of early human embryonic development beyond the blastocyst stage. In addition, our study establishes a new model system relevant to early human pregnancy loss. Finally, our work will also assist in the rational design of differentiation protocols of human embryonic stem cells to specific cell types for disease modelling and cell replacement therapy.

This is a preview of subscription content, access via your institution

Access options

Buy this article

Prices may be subject to local taxes which are calculated during checkout

Figure 1: D.p.f. 6 human blastocyst embryos display human-specific transcriptional profiles.
Figure 2: Attached d.p.f. 8 human embryos begin transcriptional and morphological self-organization.
Figure 3: D.p.f. 10 embryo cell lineages diversify and self-organize amniotic and yolk sac cavities.
Figure 4: D.p.f. 12 embryos exhibit characteristic of CS5C in vivo including TE cellular phenotypes.

Similar content being viewed by others


  1. Bedzhov, I., Leung, C. Y., Bialecka, M. & Zernicka-Goetz, M. In vitro culture of mouse blastocysts beyond the implantation stages. Nature Protocols 9, 2732–2739 (2014)

    Article  CAS  Google Scholar 

  2. Bedzhov, I. & Zernicka-Goetz, M. Self-organizing properties of mouse pluripotent cells initiate morphogenesis upon implantation. Cell 156, 1032–1044 (2014)

    Article  CAS  Google Scholar 

  3. Rossant, J. Mouse and human blastocyst-derived stem cells: vive les differences. Development 142, 9–12 (2015)

    Article  CAS  Google Scholar 

  4. Rossant, J., Chazaud, C. & Yamanaka, Y. Lineage allocation and asymmetries in the early mouse embryo. Phil. Trans. R. Soc. Lond. B 358, 1341–1349 (2003)

    Article  CAS  Google Scholar 

  5. Berg, D. K. et al. Trophectoderm lineage determination in cattle. Dev. Cell 20, 244–255 (2011)

    Article  CAS  Google Scholar 

  6. Niakan, K. K. & Eggan, K. Analysis of human embryos from zygote to blastocyst reveals distinct gene expression patterns relative to the mouse. Dev. Biol. 375, 54–64 (2013)

    Article  CAS  Google Scholar 

  7. Li, Y. et al. BMP4-directed trophoblast differentiation of human embryonic stem cells is mediated through a DeltaNp63+ cytotrophoblast stem cell state. Development 140, 3965–3976 (2013)

    Article  CAS  Google Scholar 

  8. Roode, M. et al. Human hypoblast formation is not dependent on FGF signalling. Dev. Biol. 361, 358–363 (2012)

    Article  CAS  Google Scholar 

  9. O’Leary, T. et al. Tracking the progression of the human inner cell mass during embryonic stem cell derivation. Nature Biotechnol. 30, 278–282 (2012)

    Article  Google Scholar 

  10. Schrode, N., Saiz, N., Di Talia, S. & Hadjantonakis, A. K. GATA6 levels modulate primitive endoderm cell fate choice and timing in the mouse blastocyst. Dev. Cell 29, 454–467 (2014)

    Article  CAS  Google Scholar 

  11. Blakeley, P. et al. Defining the three cell lineages of the human blastocyst by single-cell RNA-seq. Development 142, 3151–3165 (2015)

    Article  CAS  Google Scholar 

  12. Shakiba, N. et al. CD24 tracks divergent pluripotent states in mouse and human cells. Nature Commun. 6, 7329 (2015)

    Article  ADS  CAS  Google Scholar 

  13. O’Rahilly, R. & Müller, F. Developmental stages in human embryos: revised and new measurements. Cells Tissues Organs 192, 73–84 (2010)

    Article  Google Scholar 

  14. National Research Council Human Embryonic Stem Cell Research Advisory Committee, Board on Life Sciences, Board on Health Sciences Policy, National Research Council, and Institute of Medicine. Final Report of the National Academies’ Human Embryonic Stem Cell Research Advisory Committee and 2010 Amendments to the National Academies’ Guidelines for Human Embryonic Stem Cell Research (National Academies Press, 2010)

  15. International Society for Stem Cell Research. Guidelines for the Conduct of Human Embryonic Stem Cell Research, Version 1 (ISSCR, Northbrook, Illinois, 2006)

Download references


We thank the members of the Brivanlou laboratory for their advice and criticisms, in particular C. Nchako, S. Tse for technical assistance, and members of the Zernicka-Goetz laboratory for their advice on how to culture embryos through attachment. We also thank A.K. Hadjantonakis for discussions, A. Wilkerson for support, and A. Brivanlou and P. Carleton-Evans for their comments on the manuscript. This work was supported by a STARR Foundation grant (number 2013-026) and Rockefeller Private funds. Images were obtained using instrumentation in The Rockefeller University Bio-Imaging Resource Center purchased with grant funds from the Sohn Conference Foundation. The Carnegie stage images are used with permission from the Virtual Human Embryo Project ( We give special thanks for technical advice on imaging to A. North, K. Thomas, and P. Ariel, and on image analysis and rendering to T. Tong. This work would not have been possible without the generosity of the people who consented to donate their embryos to research, to whom we are indebted.

Author information

Authors and Affiliations



A.D., G.C., and L.P. performed experiments; A.D. and G.C. analysed experiments; M.Z.-G. was instrumental in teaching and transferring knowledge on the mouse technology to A.D.; E.S. provided criticism of the work and manuscript; A.H.B. conceived and designed the project, established contact with the source of the biological material, provided guidance and advice throughout the work, and interfaced with the Institutional Review Board at The Rockefeller University; all authors contributed to the manuscript.

Corresponding author

Correspondence to Ali H. Brivanlou.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Quantification of cells positive for transcriptional markers over d.p.f. 6–12 of in vitro culture.

a, Time course of total cell numbers via DAPI quantification. Confocal data sets for each embryo were scored for total cell numbers in Imaris. The mean ± s.e.m. is shown for d.p.f. 6, 8, 10, 12 (n = 8, 4, 4, 2, embryos per time point respectively over n = 2, 2, 2, 1 independent experiments). A total of 91 embryos were thawed and 64 (70%) showed developmental progression and attached; 26 of 64 (40%) embryos showed normal size and morphology and were analysed for markers of cell fate. The remaining embryos all showed abnormal development, represented by few (0–5) and scattered OCT4 cells, and were excluded from analysis as obvious failures of development. Subsequent to staining and analysis of this set of embryos with these core markers, six d.p.f. 7 and five d.p.f. 10 embryos showing normal OCT4 development were also stained and analysed for additional markers. b, Total cells expressing cell type markers, mean ± s.e.m. for OCT4, GATA6, CDX2, and GATA3: d.p.f. 6 (n = 8, 7, 5, 3 respectively); d.p.f. 8 and 10 (n = 4 embryos per marker); d.p.f. 12 (n = 2 embryos per marker). The line subdividing the OCT4 population identifies the two subpopulation of OCT4+ cells: OCT4HI (bottom of the bar), passing a threshold for high-intensity staining; and OCT4LO (top of the bar), passing a low threshold (but not the higher one for OCT4HI) for staining. At d.p.f. 6, OCT4HI cells were confined to the ICM while the OCT4LO cells represented the TE; between d.p.f. 8 and 12, OCT4LO cells were also co-expressing CDX2 (ysTE cells), while OCT4HI cells were confined to the Epi. GATA6 was scored for high or low (HI/LO) expression at d.p.f. 6 only, where GATA6HI cells (bottom of the bar) localized to the ICM and GATA6LO cells localized to the TE. See also Supplementary Table. c, Number of cells expressing lineage-specific markers expressed as a percentage of total nuclei; each bar shows mean ± s.e.m. for OCT4, GATA3, GATA6, and CDX2. The line subdividing the OCT4 bars at d.p.f. 6–12 and the GATA6 bar at d.p.f. 6 represents the high or low (HI/LO) expressing cells as explained above.

Source data

Extended Data Figure 2 Human-specific transcriptional profiles of ICM and TE in d.p.f. 6 human blastocysts.

a, Additional three-dimensional rendering of the front half of a d.p.f. 6 blastocyst (left), and ICM-zoom (right). DAPI (white) identifies all nuclei and phalloidin (magenta, actin, virtual channel) shows intercellular boundaries between TE cells. A three-dimensional segmentation mask on the DAPI channel was subtracted from the raw GATA6–phalloidin channel to yield a phalloidin-only virtual channel for these panels. This embryo is the same as displayed in Fig. 1d–j. Scale bar, 100 μm for whole embryos, 20 μm for ICM-zooms. bf, Top, whole embryo, and bottom, ICM-zooms, stained for OCT4 (green), GATA6 (red, virtual channel), CDX2 (cyan), and serial merges. b, c, Arrows show that ICM cells have high-intensity OCT4 with or without GATA6 staining; arrowheads show low-level expression of OCT4 and GATA6 in TE cells; e, f, CDX2 staining is weak and predominantly cytoplasmic. The GATA6-only channel is the result of an inverse mask using three-dimensional segmentation on DAPI to remove extra-nuclear phalloidin staining, resulting in nuclear the GATA6-only virtual channel. This embryo is the same as displayed in Fig. 1k–l. Scale bar, 100 μm for whole embryos, 20 μm for ICM-zooms. g, No HCGB and very weak CK7 signal was observed in d.p.f. 7 embryos (n = 6). Shown is a section through a d.p.f. 7 embryo stained with DAPI (white), OCT4 (green), HCGB (yellow), and CK7 (magenta). Scale bar, 100 μm.

Extended Data Figure 3 Molecular signature of d.p.f. 8–10 embryos.

a, Embryo reconstruction of a d.p.f. 8 embryo oriented at a 30° pitch; individual channels are shown in greyscale and then merged as OCT4 (green), GATA6–phalloidin (red), GATA3 (blue). Phalloidin staining of all d.p.f. 8 embryos was exceptionally weak compared with all other d.p.f. (all embryos analysed in two separate experiments); scale bar, 50 μm. b, No HCGB staining was detected at d.p.f. 8 (n = 4). Section through a d.p.f. 8 embryo showing DAPI (white), OCT4 (green), HCGB (yellow), and CK7 (magenta). Scale bar, 50 μm. c, No CD24 staining was detected at d.p.f. 8 (n = 4). Section through a d.p.f. 8 embryo showing DAPI (white), OCT4 (green), CD24 (red). Scale bar, 50 μm. d, Embryo reconstruction of a d.p.f. 10 embryo oriented at a 30° pitch; individual channels are shown in greyscale and then merged as OCT4 (green), GATA6–phalloidin (red), GATA3 (blue). Phalloidin staining of this embryo was so intense that the original underlying GATA6 nuclear stain is not visible in this reconstruction; scale bar, 50 μm.

Extended Data Figure 4 Human-specific transcriptional profile of d.p.f. 10 embryos.

D.p.f. 10 in vitro attached embryos recapitulate the principal landmarks of Carnegie stage 5b (in vivo d.p.f. 9). ak, Confocal z-sections of a d.p.f. 10 embryo stained for DAPI (white), OCT4 (green), GATA6 and phalloidin (GATA6-Ph, red), CDX2 (cyan), GATA3 (blue), and combinatorial merges. The box in f indicates the area including amniotic and yolk sac cavities, which is shown in Fig. 3d; scale bar, 50 μm. lr, Additional confocal z-section of the boxed region in f, stained for DAPI (white), OCT4 (green), GATA6–phalloidin (red), CDX2 (cyan), and combinatorial merges, showing amniotic cavity formed by OCT4+-only Epi cells (arrow in m and n), flanking GATA6+-only PE cells (white arrowhead in n), and CDX2+/GATA6+/OCT4LO ysTE cells lining the yolk sac cavity (yellow arrowhead in n, o); the CDX2+/GATA6+/OCT4LO merge of this series is shown in Fig. 3d. Scale bar, 20 μm. sx, Confocal z-section of the Epi/PE area of a d.p.f. 10 embryo stained for DAPI (white), OCT4 (green), GATA6 and CD24 (red), NANOG (magenta). This is the same embryo as in Fig. 3v, w. Scale bar, 20 μm.

Extended Data Figure 5 Full z-series d.p.f. 10 amniotic and yolk sac cavities.

Series of confocal z-sections of the embryo from Fig. 3c stained for OCT4 (green), GATA6–phalloidin (red), and CDX2 (cyan), showing the amniotic and yolk sac cavities, with ysTE cells showing expression of varying levels of all three markers. The scale is the same as in Fig. 3d.

Extended Data Figure 6 Tilted three-dimensional reconstruction of a d.p.f. 12 embryo.

The embryo reconstruction is oriented at a 30° pitch, with individual channels shown in greyscale and then merged as OCT4 (green), GATA6–phalloidin (red), GATA3 (blue). Phalloidin staining of this embryo was so intense upon re-staining that the original underlying GATA6 nuclear stain is not visible in this data set. Scale bar, 50 μm.

Extended Data Figure 7 Additional images from a d.p.f. 12 embryo.

ae, Confocal z-section of the d.p.f. 12 embryo in Fig. 4c stained for DAPI (white), OCT4 (green), GATA6–phalloidin (GATA6-Ph, red), GATA3 (blue), and CDX2 (cyan). Scale bar, 100 μm. fo, Confocal sections of the same d.p.f. 12 embryo stained and imaged first (fj) for DAPI (white), OCT4 (green), GATA6–phalloidin (red), and GATA3 (blue) and then (ko) re-stained and re-imaged for DAPI (white), OCT4 (green), CK7 (magenta), and HCGB (yellow). DAPI and OCT4 were used as landmarks to identify the same z-plane and the same cells between the two staining and imaging rounds. The arrow in j and o indicates an example of nascent lacuna, typical of ST cells at Carnegie stage 5c in vivo; the box indicates an example of multinucleated cells characteristic of ST lineage progression; zoom-ins of the area in the box are presented in Fig. 4r–v. Scale bar, 20 μm.

Supplementary information

Supplementary Table

This file contains Supplementary Table 1, which shows Cell scoring by marker and cell type. (PDF 109 kb)

Timecourse of in vitro human embryo attachment from DPF6 to DPF14

OCT4 (green), GATA3 (blue), GATA6-Phalloidin (red). For the 3D rendering of DPF14 only, OCT4 and GATA3 channels were thresholded using the spot-finding algorithm in Imaris and then masked to remove non-nuclear background. (MOV 29120 kb)

Example of DPF6 embryo

DAPI (blue), OCT4 (green), CDX2 (cyan), GATA6 (red), Phalloidin (white). For this 3D rendering animation, a DAPI mask was generated by Imaris spot finding and then used to remove nuclear GATA6 from the original GATA6-phalloidin channel (creating Phalloidin-only and GATA6-only virtual channels). The Phalloidin virtual channel was then normalized across z using default settings in Imaris. A DAPI nuclear mask was also used to isolate nuclear OCT4 signal from non-nuclear background staining. A few pieces of debris outside of the embryo were manually cropped in 3D. This embryo was imaged at 40x. (MOV 28454 kb)

Example of DPF6 embryo

OCT4 (green), GATA3 (blue), GATA6-Phalloidin (red) (MOV 24352 kb)

Example of DPF8 embryo.

OCT4 (green), GATA3 (blue), GATA6-Phalloidin (red) (MOV 17985 kb)

Example of DPF10 embryo and cavitation

DAPI (blue), OCT4 (green), CDX2 (cyan), GATA6-Phalloidin (red). A 3D DAPI mask was used to exclude non-nuclear CDX2 and OCT4 stainings from the raw channels. Video breakdown by seconds: 0-12, 3D render of DAPI, nuclear CDX2, nuclear OCT4, and GATA6-Phalloidin; 12-16, z-sections (2µm step) from the bottom to the top of stack; 16-35, OCT4 (unmasked channel) and GATA6-Phalloidin as z-sections from the top to the bottom of the stack; 35-45, CDX2 (unmasked channel) is added at the bottom z-plane and the slice traverses from the bottom to the top of the stack showing bright CDX2 co-staining with GATA6 and OCT4 in ysTE cells; 45-50, CDX2 (unmasked channel) was replaced with nuclear (masked) CDX2, the z-slice returns to the bottom of the stack leaving progressive overlaying sections on display; 50-55 zoom out of 3D render; 55-58 DAPI staining faded back in. (MOV 25859 kb)

Example of amniotic and yolk sac cavities in a DPF10 embryo

OCT4 (green), CDX2 (cyan), GATA6-Phalloidin (red). 3D reconstruction of a cropped volume containing EPI, amniotic and yolk sac cavities, and ysTE cells. An oblique, 1.5µm virtual section is imposed and the 3D rendering is removed to visualize the signal from the plane passing through the structure and back. This image was obtained at 40x. (MOV 26976 kb)

Example of amniotic cavity in a DPF10 embryo

OCT4 (green), CDX2 (cyan), GATA6-Phalloidin (red); Z-series fly-through reconstruction. (AVI 23591 kb)

Example of cytotrophoblast and syncytiotrophoblast phenotypes in a DPF12 embryo

OCT4 (green), GATA3 (blue), GATA6-Phalloidin (red). (MOV 24149 kb)

PowerPoint slides

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Deglincerti, A., Croft, G., Pietila, L. et al. Self-organization of the in vitro attached human embryo. Nature 533, 251–254 (2016).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing