To implant in the uterus, the mammalian embryo first specifies two cell lineages: the pluripotent inner cell mass that forms the fetus, and the outer trophectoderm layer that forms the placenta1. In many organisms, asymmetrically inherited fate determinants drive lineage specification2, but this is not thought to be the case during early mammalian development. Here we show that intermediate filaments assembled by keratins function as asymmetrically inherited fate determinants in the mammalian embryo. Unlike F-actin or microtubules, keratins are the first major components of the cytoskeleton that display prominent cell-to-cell variability, triggered by heterogeneities in the BAF chromatin-remodelling complex. Live-embryo imaging shows that keratins become asymmetrically inherited by outer daughter cells during cell division, where they stabilize the cortex to promote apical polarization and YAP-dependent expression of CDX2, thereby specifying the first trophectoderm cells of the embryo. Together, our data reveal a mechanism by which cell-to-cell heterogeneities that appear before the segregation of the trophectoderm and the inner cell mass influence lineage fate, via differential keratin regulation, and identify an early function for intermediate filaments in development.
This is a preview of subscription content
Subscription info for Chinese customers
We have a dedicated website for our Chinese customers. Please go to naturechina.com to subscribe to this journal.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Source data are provided with this paper.
Code for apical surface curvature analysis has been published in a publicly available repository at https://github.com/gracelhy/Analysis-of-embryo-parameters.
White, M. D., Zenker, J., Bissiere, S. & Plachta, N. Instructions for assembling the early mammalian embryo. Dev. Cell 45, 667–679 (2018).
Knoblich, J. A. Asymmetric cell division: recent developments and their implications for tumour biology. Nat. Rev. Mol. Cell Biol. 11, 849–860 (2010).
Tarkowski, A. K. & Wróblewska, J. Development of blastomeres of mouse eggs isolated at the 4- and 8-cell stage. J. Embryol. Exp. Morphol. 18, 155–180 (1967).
Torres-Padilla, M. E., Parfitt, D. E., Kouzarides, T. & Zernicka-Goetz, M. Histone arginine methylation regulates pluripotency in the early mouse embryo. Nature 445, 214–218 (2007).
White, M. D. et al. Long-lived binding of Sox2 to DNA predicts cell fate in the four-cell mouse embryo. Cell 165, 75–87 (2016).
Wang, J. et al. Asymmetric expression of LincGET biases cell fate in two-cell mouse embryos. Cell 175, 1887–1901.e18 (2018).
Biase, F. H., Cao, X. & Zhong, S. Cell fate inclination within 2-cell and 4-cell mouse embryos revealed by single-cell RNA sequencing. Genome Res. 24, 1787–1796 (2014).
Goolam, M. et al. Heterogeneity in Oct4 and Sox2 targets biases cell fate in 4-cell mouse rmbryos. Cell 165, 61–74 (2016).
Shi, J. et al. Dynamic transcriptional symmetry-breaking in pre-implantation mammalian embryo development revealed by single-cell RNA-seq. Development 142, 3468–3477 (2015).
Casser, E. et al. Totipotency segregates between the sister blastomeres of two-cell stage mouse embryos. Sci. Rep. 7, 8299 (2017).
Johnson, M. H. & Ziomek, C. A. The foundation of two distinct cell lineages within the mouse morula. Cell 24, 71–80 (1981).
Maître, J. L. et al. Asymmetric division of contractile domains couples cell positioning and fate specification. Nature 536, 344–348 (2016).
Anani, S., Bhat, S., Honma-Yamanaka, N., Krawchuk, D. & Yamanaka, Y. Initiation of Hippo signaling is linked to polarity rather than to cell position in the pre-implantation mouse embryo. Development 141, 2813–2824 (2014).
Korotkevich, E. et al. The Apical domain is required and sufficient for the first lineage segregation in the mouse embryo. Dev. Cell 40, 235–247.e7 (2017).
Nishioka, N. et al. The Hippo signaling pathway components Lats and Yap pattern Tead4 activity to distinguish mouse trophectoderm from inner cell mass. Dev. Cell 16, 398–410 (2009).
Zenker, J. et al. Expanding actin rings zipper the mouse embryo for blastocyst formation. Cell 173, 776–791.e17 (2018).
Coulombe, P. A. & Wong, P. Cytoplasmic intermediate filaments revealed as dynamic and multipurpose scaffolds. Nat. Cell Biol. 6, 699–706 (2004).
Jackson, B. W. et al. Formation of cytoskeletal elements during mouse embryogenesis. Intermediate filaments of the cytokeratin type and desmosomes in preimplantation embryos. Differentiation 17, 161–179 (1980).
Paulin, D., Babinet, C., Weber, K. & Osborn, M. Antibodies as probes of cellular differentiation and cytoskeletal organization in the mouse blastocyst. Exp. Cell Res. 130, 297–304 (1980).
Oshima, R. G., Howe, W. E., Klier, F. G., Adamson, E. D. & Shevinsky, L. H. Intermediate filament protein synthesis in preimplantation murine embryos. Dev. Biol. 99, 447–455 (1983).
Hesse, M., Franz, T., Tamai, Y., Taketo, M. M. & Magin, T. M. Targeted deletion of keratins 18 and 19 leads to trophoblast fragility and early embryonic lethality. EMBO J. 19, 5060–5070 (2000).
Tamai, Y. et al. Cytokeratins 8 and 19 in the mouse placental development. J. Cell Biol. 151, 563–572 (2000).
Baribault, H., Price, J., Miyai, K. & Oshima, R. G. Mid-gestational lethality in mice lacking keratin 8. Genes Dev. 7 (7A), 1191–1202 (1993).
Lu, H., Hesse, M., Peters, B. & Magin, T. M. Type II keratins precede type I keratins during early embryonic development. Eur. J. Cell Biol. 84, 709–718 (2005).
Chisholm, J. C. & Houliston, E. Cytokeratin filament assembly in the preimplantation mouse embryo. Development 101, 565–582 (1987).
Emerson, J. A. Disruption of the cytokeratin filament network in the preimplantation mouse embryo. Development 104, 219–234 (1988).
Ralston, A. & Rossant, J. Cdx2 acts downstream of cell polarization to cell-autonomously promote trophectoderm fate in the early mouse embryo. Dev. Biol. 313, 614–629 (2008).
Yoon, K. H. et al. Insights into the dynamic properties of keratin intermediate filaments in living epithelial cells. J. Cell Biol. 153, 503–516 (2001).
Fleming, T. P., Garrod, D. R. & Elsmore, A. J. Desmosome biogenesis in the mouse preimplantation embryo. Development 112, 527–539 (1991).
Den, Z., Cheng, X., Merched-Sauvage, M. & Koch, P. J. Desmocollin 3 is required for pre-implantation development of the mouse embryo. J. Cell Sci. 119, 482–489 (2006).
Alarcon, V. B. Cell polarity regulator PARD6B is essential for trophectoderm formation in the preimplantation mouse embryo. Biol. Reprod. 83, 347–358 (2010).
Doi, M. & Edwards, S. F. The Theory of Polymer Dynamics Vol. 73 (Oxford Univ, Press, 1988).
Almonacid, M., Terret, M. É. & Verlhac, M. H. Actin-based spindle positioning: new insights from female gametes. J. Cell Sci. 127, 477–483 (2014).
Käs, J., Strey, H. & Sackmann, E. Direct imaging of reptation for semiflexible actin filaments. Nature 368, 226–229 (1994).
Samarage, C. R. et al. Cortical tension allocates the first inner cells of the mammalian embryo. Dev. Cell 34, 435–447 (2015).
Hirate, Y. et al. Polarity-dependent distribution of angiomotin localizes Hippo signaling in preimplantation embryos. Curr. Biol. 23, 1181–1194 (2013).
Leung, C. Y. & Zernicka-Goetz, M. Angiomotin prevents pluripotent lineage differentiation in mouse embryos via Hippo pathway-dependent and -independent mechanisms. Nat. Commun. 4, 2251 (2013).
Magin, T. M. et al. Lessons from keratin 18 knockout mice: formation of novel keratin filaments, secondary loss of keratin 7 and accumulation of liver-specific keratin 8-positive aggregates. J. Cell Biol. 140, 1441–1451 (1998).
Leonavicius, K. et al. Mechanics of mouse blastocyst hatching revealed by a hydrogel-based microdeformation assay. Proc. Natl Acad. Sci. USA 115, 10375–10380 (2018).
Wirtz, D. Particle-tracking microrheology of living cells: principles and applications. Annu. Rev. Biophys. 38, 301–326 (2009).
Zenker, J. et al. A microtubule-organizing center directing intracellular transport in the early mouse embryo. Science 357, 925–928 (2017).
Piotrowska-Nitsche, K., Perea-Gomez, A., Haraguchi, S. & Zernicka-Goetz, M. Four-cell stage mouse blastomeres have different developmental properties. Development 132, 479–490 (2005).
Panamarova, M. et al. The BAF chromatin remodelling complex is an epigenetic regulator of lineage specification in the early mouse embryo. Development 143, 1271–1283 (2016).
Wang, L. et al. CARM1 methylates chromatin remodeling factor BAF155 to enhance tumor progression and metastasis. Cancer Cell 30, 179–180 (2016).
Plachta, N., Bollenbach, T., Pease, S., Fraser, S. E. & Pantazis, P. Oct4 kinetics predict cell lineage patterning in the early mammalian embryo. Nat. Cell Biol. 13, 117–123 (2011).
Kaur, G. et al. Probing transcription factor diffusion dynamics in the living mammalian embryo with photoactivatable fluorescence correlation spectroscopy. Nat. Commun. 4, 1637 (2013).
Tabansky, I. et al. Developmental bias in cleavage-stage mouse blastomeres. Curr. Biol. 23, 21–31 (2013).
Daniels, B. R., Masi, B. C. & Wirtz, D. Probing single-cell micromechanics in vivo: the microrheology of C. elegans developing embryos. Biophys. J. 90, 4712–4719 (2006).
Park, H. Y., Trcek, T., Wells, A. L., Chao, J. A. & Singer, R. H. An unbiased analysis method to quantify mRNA localization reveals its correlation with cell motility. Cell Rep. 1, 179–184 (2012).
Skamagki, M., Wicher, K. B., Jedrusik, A., Ganguly, S. & Zernicka-Goetz, M. Asymmetric localization of Cdx2 mRNA during the first cell-fate decision in early mouse development. Cell Rep. 3, 442–457 (2013).
We thank X. Liang for assistance with human embryo work. This work was supported by grants from ASTAR, EMBO, and HHMI to N.P., EMBL Australia to M.B., and the ASTAR Graduate Scholarship to H.Y.G.L.
The authors declare no competing interests.
Peer review information Nature thanks Magdalena Zernicka-Goetz and the other anonymous reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
a, 3D views of mouse embryos at multiple developmental stages, stained for K18. K18 expression and localization resemble that of K8. Note the initial assembly of filaments in a specific subset of cells in the 8-cell embryo. Data are from five independent experiments. b, Double immunofluorescence for K8 and K18 shows their colocalization in filament structures within the same embryo. Data are from three independent experiments. c, Double immunofluorescence using a pan-keratin antibody and K18 shows colocalization in filament structures within the same embryo. Data are from three independent experiments. d, High-magnification views highlight keratin filament organization at multiple developmental stages (top). Surface render of computationally-segmented cells and keratin filaments with top and side views show the changes in cell morphology and keratin filament organization at different developmental stages. The density of the keratin filament network increases over time and the filaments become enriched at cell-cell junctions. Data are from five independent experiments. e, Live imaging of embryos expressing K18-Emerald. A subset of cells begins to assemble keratin filaments at the eight-cell stage, similar to observations from immunofluorescence for endogenous keratins. No keratin filaments are detected in four-cell or early uncompacted eight-cell embryos. Data are from three independent experiments. f, g, Colocalization of K18-Emerald and immunofluorescence against K8 (f) or K18 (g). Bottom panels show zoomed views of single cells expressing keratin filaments, with arrows pointing to an example of signal colocalization. Data are from three independent experiments. Scale bars, 10 μm.
Extended Data Fig. 2 Tracking and quantitative analysis of keratin filament movement during interphase and mitosis.
a, Time series of an embryo expressing K18-Emerald and RFP-Utrophin, with the corresponding major cellular events labelled in the left column. Separate K18-Emerald and RFP-Utrophin channels are shown. Right panels show 2D views through a single cell that assembles keratin filaments, for better visualization of keratin distribution within the cell, relative to the apical domain. Keratin filament assembly is initiated before the formation of the apical domain. When the apical domain forms, keratin filaments become enriched apically in close association with F-actin. During mitosis, the apical domain disassembles but keratin filaments remain apically localized, resulting in their asymmetric inheritance by the outer daughter cell. Data are from three independent experiments. b, Immunofluorescence of endogenous keratins in embryos fixed at different stages of apical domain formation recapitulates the pattern and localization of keratin filaments relative to the apical domain observed in live imaging experiments. Data are from three independent experiments. c, Computationally-rendered filaments obtained from live imaging data. In this example, five individual filaments were tracked over time with a 10-min interval between frames. Data are from three independent experiments. d, The log mean square displacement (MSD) versus log lag time graph indicates that the movement of keratin filaments is unconfined and diffusive (slope > 1). Pearson’s correlation. e, Volume of an individual keratin filament and total filament volume within a single tracked cell increase linearly over time. Pearson’s correlation. f, Quantification of filament speed, volume of filaments, and polarization index before apical domain formation, after apical domain formation, and during mitosis. After the formation of the apical domain, keratin filaments move more slowly, display a larger total volume, and become more apically polarized than before apical domain formation. During mitosis, keratin filaments move faster, but retain a large volume and high apical polarization. ***P = 0.0002; **P = 0.001; Kruskal–Wallis test for filament speed; **P = 0.003; ANOVA test for filament volume; **P = 0.003; Kruskal–Wallis test for polarization index. Scheme shows the parameters used for calculation of the polarization index. d1 is the distance between the volume-weighted centre of mass of the keratin filaments and the centre of mass of the cell. d2 is the length of the apical-basal axis of the cell. g, High-resolution immunofluorescence images show that keratin filaments align specifically along actin filaments extending from the apical domain. Green arrows indicate examples of keratin–actin colocalization. Data are from three independent experiments. h, Differences in F-actin accumulation at the apical domain of control embryos and embryos treated with cytochalasin D or a high concentration of SiR-Actin. Insets show zoomed views of individual 8-cell blastomeres, highlighting the loss of the apical domain in cytochalasin D-treated embryos, and a dense accumulation of apical F-actin in SiR-Actin-treated embryos. Data are from three independent experiments. Scale bars, 10 μm.
a–c, Immunofluorescence of endogenous plakoglobin (a), plakophilin (b), and desmoglein2 (c) before and after apical domain formation. Apical accumulation of all three desmosome components is observed after apical domain formation. d, Live imaging of an embryo expressing desmoglein2-Emerald, RFP-utrophin and H2B-RFP recapitulates the endogenous desmoglein2 expression, both before and after apical domain formation. e, Time series of embryo expressing desmoglein2-Emerald, RFP-utrophin and H2B-RFP. Desmoglein2-Emerald accumulates with the apical domain (labelled by RFP-utrophin) during interphase. When the cell enters mitosis, desmoglein2-Emerald disassembles from the apical surface together with the apical domain. White arrows indicate two different mitotic events within the same embryo. f, Live embryo expressing desmoglein2-Ruby and K18-Emerald shows the enrichment of keratin filaments at the site of apical desmosome accumulation. g, Embryos injected with siRNAs against desmosome components do not accumulate desmoglein2 apically with the apical domain. h, Desmosome knockdown causes a more homogenous distribution of keratin filaments, as measured by a polarization index. **P = 0.01; unpaired, two-tailed Mann–Whitney U-test. Data are from three independent experiments. Scale bars, 10 μm.
Extended Data Fig. 4 Keratin filaments are stably retained during mitosis, and become asymmetrically inherited by outer daughter cells.
a, Immunofluorescence shows the extensive remodelling of cortical F-actin and microtubules during different stages of mitosis. Data are from six independent experiments. b, Immunofluorescence for K8 shows endogenous keratin filaments retained within mitotic cells in embryos fixed at multiple stages of development. Data are from six independent experiments. c, FRAP experiments for K18-Emerald and mRuby2-Actin performed in whole live embryos. All cells selected for FRAP were at the eight-cell stage and in interphase, when the actin ring is visible. 3D views of entire pre-FRAP embryos (left), zoomed views of the photobleached regions of interest (middle), and kymographs of pre- and post-FRAP fluorescence intensities (right). Data are from three independent experiments. d, Analysis of FRAP experiments. Left graphs show fluorescence recovery of K18-Emerald (green) and mRuby2-Actin (red) over time. Thinner lines represent raw data after normalization, and thicker lines indicate fitted exponential curves. Right graph shows that K18-Emerald has a larger immobile fraction than mRuby2-Actin. ***P < 0.0001; unpaired, two-tailed Student’s t-test. e, Live imaging of embryos expressing K8-Emerald show a similar pattern of expression and inheritance as K18-Emerald. The outer daughter cell inherits most of the keratin filaments during an outer-inner division (top). Computational segmentation of the same cell at each stage of mitosis (bottom). Quantification of proportion of keratin filaments inherited by outer and inner cells in live embryos expressing K8-Emerald shows a comparable asymmetry in keratin inheritance as K18-Emerald. **P = 0.001; unpaired, two-tailed Student’s t-test. f, Time series of a cell expressing K18-Emerald undergoing a symmetric outer-outer division. Keratin filaments are uniformly inherited by both daughter cells during divisions producing two outer cells (top). Computational segmentation of the same cells at each time point (bottom), and whole embryo inset highlighting the outer location of both daughter cells (right). Data are from four independent experiments. Scale bars, 5 μm.
Extended Data Fig. 5 A dense F-actin meshwork within mitotic cells hinders the movement of keratin filaments away from the apical cortex.
a, Immunofluorescence of embryos fixed specifically when a cell was undergoing mitosis or cytokinesis. Top, keratin filaments remain apically-localized throughout different mitotic stages, and become inherited by the prospective outer cell. Bottom, computational segmentation of the same cells highlighting the apical keratin distribution and asymmetric keratin inheritance. Quantification of proportion of endogenous keratin filaments present in the apical and basal regions of mitotic cells, and between prospective outer and inner daughter cells, showing a comparable asymmetry in endogenous keratin localization and inheritance as K18-Emerald dynamics in live embryos. ***P < 0.0001; unpaired, two-tailed Student’s t-test. b, Embryos microinjected with Pard6b siRNAs do not form an apical F-actin ring in the eight-cell embryo. Data are from three independent experiments. c, Mitotic cell within a fixed human embryo also displays an apical localization of keratins. d, A dense cytoplasmic F-actin meshwork is maintained throughout interphase and all stages of mitosis. Data are from three independent experiments. e, The F-actin meshwork is also present in cells across different stages of development. Representative images of a 3-cell, compacted 8-cell, and 16-cell embryo with all cells displaying a dense cytoplasmic F-actin meshwork. Data are from three independent experiments. f, Analysis of keratin filament movement during mitosis reveals that filament speed is inversely related to filament volume. n = 8 filaments; Pearson’s correlation. g, Acute cytochalasin D treatment for 15 min specifically during mitosis disrupts the F-actin meshwork, reduces the apical localization of keratins, and increases keratin filament speed. Cells treated with MG132 for 3 h retain an F-actin meshwork, but keratin apical localization is reduced and filament speed is unchanged. **P = 0.002 for CytoD; **P = 0.01 for MG132; Kruskal–Wallis test for polarization index; ***P = 0.0005; ANOVA test for filament speed. h, Scheme of a cell division producing an inner (green) and an outer (blue) cell. Keratin filaments localize close to the apical cortex of the forming outer daughter cell. The distance between the apical cortex and cytokinetic furrow, time between disassembly of the apical F-actin domain and cytokinesis, and the mean speed of keratin filament movement are indicated. Scale bars, 5 μm.
a, FRAP experiments for mRuby2-Actin performed at the apical domain of interphase cells with keratins, cells without keratins, and cells microinjected with desmosome siRNAs. Selected photobleached regions of interest (left) and kymographs of pre- and post-FRAP fluorescence intensities (right) are shown. Data are from three independent experiments. b, Analysis of FRAP experiments. Graphs show the fluorescence recovery of mRuby2-Actin over time for each condition. Thinner red lines indicate raw data after normalization, thicker red lines are fitted exponential curves, and thick black lines represent the mean fitted exponential curves. c, Cells lacking keratins and cells with reduced desmosome expression show a smaller immobile fraction of mRuby2-Actin compared to cells with keratins. **P = 0.0002 for without keratins; **P = 0.003 for desmosome KD; Kruskal–Wallis test. d, Immunofluorescence of 16-cell stage control embryos and embryos treated with cytochalasin D and SiR-Actin. Disruption of actin stability using cytochalasin D reduces accumulation of apical polarity markers PARD6B and PKCζ. By contrast, increasing actin stability using SiR-Actin increases apical polarity levels. *P = 0.03; ***P = 0.0009; ANOVA test for PARD6B; *P = 0.03; **P = 0.003; Kruskal–Wallis test for PKCζ. e, Desmosome knockdown in 16-cell stage embryos reduces levels of apical polarity markers PARD6B and PKCζ. ***P = 0.0002 for PARD6B; ***P = 0.001 for PKCζ; Unpaired, two-tailed Mann–Whitney U-test. f, Live imaging of K18-Emerald in an embryo displaying a cell division. After division, the daughter cell that did not inherit keratins (cyan) undergoes apical constriction to form the pluripotent inner cell mass35, whereas the outer daughter cell that inherited keratins (yellow) does not internalize. Data are from three independent experiments. g, Analysis of internalization events in cells that inherited (K+) or did not inherit (K−) keratin filaments after division. ***P < 0.0001; two-tailed Fisher’s exact test. h, Immunofluorescence of endogenous K8 and AMOT in a 16-cell stage embryo. Right panels indicate zoomed views of the apical region of cells with and without keratins, with separate K8 and AMOT channels for better visualization. Cells with keratins display higher levels of apical AMOT than cells lacking keratins and cells with K8 and K18 knockdown. *P = 0.04; ***P < 0.0001; Kruskal–Wallis test. Scale bars, 10 μm.
Extended Data Fig. 7 Experimental manipulations of keratin levels show that keratins regulate CDX2 to specify the first trophectoderm cells of the embryo.
a, Immunofluorescence for K8 in embryos microinjected with siRNAs for K8 and K18 at the one-cell stage, or into only one cell at the two-cell stage. This double-knockdown approach extensively eliminates keratin filament assembly. Data are from five independent experiments. b, Knockdown of K8 and K18 in half of the embryo also eliminates filament formation by K19. White arrowheads show knockdown cells. Data are from three independent experiments. c, Keratin overexpression causes a premature and widespread assembly of a keratin network within the 8- to 16-cell stage embryo. Images show examples of embryos microinjected with high levels of K8 and K18 RNA at the 1-cell stage, or into one cell of the 2-cell embryo. Data are from three independent experiments. d, Keratin overexpression causes some filaments to be inherited by inner cells of the 16-cell stage embryo (yellow segmented cell indicated by arrow in left panel). 2D view shows keratin filament organization within outer and inner cells of keratin overexpressing embryos (right). Data are from three independent experiments. e, Inner cells in keratin overexpressing embryos express lower levels of nuclear YAP and CDX2 than outer cells. *P = 0.04; ***P < 0.0001; unpaired, two-tailed Mann–Whitney U-test. f, Knockdown of YAP using siRNAs microinjected into one cell of the two-cell embryo reduces CDX2 levels, in both keratin-positive and keratin-negative cells. H2B-RFP was co-injected with the siRNAs to identify the knockdown cells (white arrowheads). ***P < 0.0001; ANOVA test for CDX2. Right graph shows that our knockdown approach using YAP siRNAs effectively reduced YAP levels. *P = 0.03, unpaired, two-tailed Mann–Whitney U-test for YAP. g, Scheme depicting cloning strategy to generate rescue constructs for K8 and K18. The coding regions of K8 and K18 are indicated by thick yellow arrows, and the targeted sequence locations for the keratin siRNAs used in this study are indicated by the red arrows. The specific siRNA target sequences are highlighted in yellow, corresponding to the keratin wild-type (WT) sequence (top rows). The rescue construct sequences are indicated (bottom rows). Note the conservation of amino acid sequence despite the scrambling of DNA bases throughout the siRNA target sequence. In each experiment, H2B-RFP was co-injected with the siRNAs and/or mRNAs to label the injected half of the embryo, and 100% of H2B-positive cells displayed keratin filaments when injected with the rescue construct. K8/K18-knockdown cells express lower levels of CDX2 than control cells with keratins, but this phenotype is rescued when the keratin rescue constructs are co-injected with keratin siRNAs. ***P < 0.0001; ANOVA test. Scale bars, 10 μm.
a, Punctate desmosome structures labelled using immunofluorescence for desmoplakin (Dsp) colocalize with K8 along the trophectoderm cell–cell junctions of the blastocyst. Data are from three independent experiments. b, Analysis of apical surface curvature in control and K8/K18-knockdown blastocysts. Individual cells within the intact embryo were computationally-segmented in 3D. Single cells (blue) are selected for apical surface analysis. Middle panels show rendering of the apical surfaces (orange) of the selected cells. The right panels show fitting of the cell apical surface to a sphere for calculation of radius of apical surface curvature. Data are from three independent experiments. c, K8/K18-knockdown blastocysts display morphogenetic defects, revealed by smaller blastocyst volume, higher junctional tortuosity, and trophectoderm cells with lower radius of apical surface curvature. **P = 0.004 for blastocyst volume; ***P < 0.0001 for junctional tortuosity; ***P < 0.0001 for surface curvature; unpaired, two-tailed Mann–Whitney U-test. d, 2D confocal planes of live control and K8/K18-knockdown blastocysts, microinjected with fluorescent nanoparticles (yellow). Data are from three independent experiments. e, Images show nanoparticles within single trophectoderm cells, in control and K8/K18-knockdown embryos. Middle panels show representative trajectories of nanoparticle movement. Graph shows their mean squared displacement (MSD) over lag time. Thicker lines represent the mean of individual curves. The graph has two phases revealing different cytoskeletal properties: a time-independent (short lag times) and a time-dependent (long lag times) phase. These phases are associated with elasticity and viscosity, respectively40. Differences in MSD during the time-independent phase reveal higher elasticity, indicative of lower cytoplasmic stiffness, in the K8/K18-knockdown cells. f, Co-injection of keratin rescue constructs with K8/K18 siRNAs can restore blastocyst morphology to control conditions. Unpaired, two-tailed Student’s t-test for blastocyst volume and surface curvature; unpaired, two-tailed Mann–Whitney U-test for junction tortuosity; NS, not significant. Scale bars, 10 μm.
Extended Data Fig. 9 Heterogeneities in BAF155 and CARM1 within the early embryo trigger differential expression of keratins at the eight-cell stage.
a, Live-imaging of an embryo expressing K8-Emerald, H2B-RFP and RFP-Utrophin confirms that the first cells to assemble keratin filaments are sister cells. The microtubule bridge connecting sister cells can be identified by RFP-Utrophin accumulation (white arrowheads)41. Data are from three independent experiments. b, Scheme shows the stereotypical 3D organization of a tetrahedral four-cell embryo. The vegetal blastomere is located distal from the polar body. c, Selective photoactivation of the vegetal blastomere. The vegetal blastomere is identified based on its distal position from the polar body. The vegetal cell nucleus is then targeted with a two-photon laser (820 nm light) to photoactivate H2B-paGFP. 2D confocal planes show efficient photoactivation immediately after 820 nm light illumination. Data are from three independent experiments. d, The first cells to form keratin filaments are unrelated to the order of cell divisions during the 4- to 8-cell stage transition. χ2 test. e, BAF155 knockdown reduces BAF155 immunofluorescence levels relative to control blastomeres, while BAF155 overexpression increases them. Embryos were microinjected with BAF155 siRNAs or high levels of BAF155 RNA respectively at the one-cell stage. ***P < 0.0001; ANOVA test. f, Embryos treated with trichostatin A (TSA) display extensive keratin filament formation, while embryos treated with actinomycin D (Act D) do not form filaments. *P = 0.0489; ***P < 0.0001; ANOVA test. g, Microinjection of K8 and K18 mRNA into the one-cell embryo causes premature assembly of an extensive keratin filament network throughout early blastomeres before the eight-cell stage. ***P < 0.0001; two-sided Fisher’s exact test. h, BAF155-overexpressing embryos treated with actinomycin D do not form keratin filaments at the eight-cell stage. Data are from three independent experiments. i, Dimethyl-BAF155 is lowest in the vegetal blastomere. **P = 0.004; unpaired, two-tailed Student’s t-test. j, CARM1 overexpression increases CARM1 immunofluorescence levels relative to control blastomeres. Embryos were microinjected with high levels of Carm1 RNA at the 1-cell stage. ***P < 0.0001; unpaired, two-tailed Mann–Whitney U-test. k, CARM1 overexpression reduces keratin filament assembly. ***P = 0.0007; unpaired, two-tailed Student’s t-test. l, Overexpression of BAF155 or mutant BAF155(R1064K) causes premature keratin filament assembly at the four-cell stage. **P = 0.009 for BAF155 overexpression; **P = 0.005 for BAF155(R1064K); two-sided Fisher’s exact test. m, BAF155-knockdown blastomeres (white arrowheads) display lower levels of CDX2 than control cells (orange arrowheads) at the same stage. BAF155 siRNAs were microinjected into only one cell of the two-cell embryo. ***P < 0.0001; ANOVA test. n, CARM1-overexpression blastomeres (white arrowheads) display lower levels of CDX2 than control blastomeres (orange arrowheads) at the same stage. High levels of Carm1 RNA were microinjected into only one cell at the two-cell stage. **P = 0.005 for control inner cells; *P = 0.04 for CARM1 overexpression outer cells; **P = 0.006 for CARM1-overexpression inner cells; ANOVA test. Scale bars, 10 μm.
Keratin expression is regulated by early heterogeneities in the BAF complex. During inner–outer cell segregation, apically localized keratin filaments are asymmetrically inherited by outer daughter cells, where they stabilize apical F-actin to promote apical polarity and acquisition of a trophectoderm fate. At late stages, keratins also support blastocyst morphogenesis. The numbers indicate the key events.
Mouse and human blastocysts display an extensive keratin filament network in the trophectoderm 3D render of fixed mouse and human blastocysts stained for K8 (orange) and DAPI (cyan). Data are representative of 3 independent experiments.
Keratin filaments assemble at the apical surface of an 8-cell blastomere Time-lapse imaging of a live embryo expressing K18-Emerald (orange) and RFPMAP2c (white) at the 8-cell stage. One cell (top) begins to assemble filaments before the other visible cells of the embryo. The filaments accumulate in the subcortical apical region of the cell. Data are representative of 3 independent experiments.
Tracking computationally segmented filaments shows that keratins associated with the apical domain become more static High magnification timelapse imaging of a single cell in a live embryo expressing K18-Emerald (orange), RFPUtrophin (white), and H2B-RFP (white). The dynamic behavior of keratin filament movement can be observed throughout the cell. The second and third playback show individual computationally segmented filaments. The keratin filament in contact with the apical domain (blue) is more static than the filament away from the apical surface (yellow). Data are representative of 3 independent experiments.
Keratin filaments are asymmetrically inherited during an outerinner cell division Time-lapse imaging of a live embryo expressing K18-Emerald (orange), RFP-Utrophin (white), and H2B-RFP (white). The first cell to form keratin filaments is indicated. Keratin filaments begin to accumulate apically in that cell, in close association with the F-actin ring. Upon entry to mitosis, the F-actin ring disassembles, but keratin filaments remain apically localized, and subsequently become asymmetrically inherited by the outer daughter cell. The inner daughter cell is devoid of keratins. Data are representative of 3 independent experiments.
Keratin-inheriting cells remain in the outer layer of the embryo, while keratin-negative cells can undergo internalization Time-lapse imaging of a live embryo expressing K18-Emerald (orange), RFP Utrophin (white), and H2B-RFP (white). A cell (bottom left) divides and produces two daughter cells, initially positioned in the outer layer of the embryo. The outer daughter cell on the left inherits keratin filaments and remains in the outer layer. By contrast, the outer cell that did not inherit keratins subsequently internalizes to contribute to the inner mass of the embryo. Data are representative of 3 independent experiments.
About this article
Cite this article
Lim, H.Y.G., Alvarez, Y.D., Gasnier, M. et al. Keratins are asymmetrically inherited fate determinants in the mammalian embryo. Nature 585, 404–409 (2020). https://doi.org/10.1038/s41586-020-2647-4
The keratin 17/YAP/IL6 axis contributes to E-cadherin loss and aggressiveness of diffuse gastric cancer
Nature Reviews Molecular Cell Biology (2021)
Nature Reviews Molecular Cell Biology (2021)