The protein memory: how keratins control embryonic cell fate

By Andres Hernandez Maduro

The process of cell specialisation is intricate and dynamic, varying across cell types. Pathways for stem cell differentiation are dependent on both their surrounding extracellular matrix and mitotic parent. However, the exact process of controlling cell fate is less certain – so how does a fertilised egg know to become a multicellular organism?

When a new mammalian zygote begins to divide by mitosis, the resulting daughter cells remain totipotent until the embryo has grown to contain 8 to 16 cells in total. The embryo is thereby known as a morula1 – here, the first cells start to differentiate into what will later become the foetus’ placenta. In particular, as the morula becomes larger, cells with different amounts of exposure to the extracellular environment are seen to divide asymmetrically (as sibling cells are made with distinct identities).2 

Cells at the periphery of the morula (known as the trophectoderm1) achieve this through polarised protein concentrations, with the apical domain (facing outward) being filled with cytoskeleton constituents like actin.2 Conversely, their basal layer (facing inward) is devoid of large polypeptides. Curiously, it has been observed that this polarity is actually lost during cell division, and only reemerges once daughter cells have split and positioned themselves relative to the axis along which division occurs.3 Should a daughter cell appear at the core of the morula, for example, it will not exhibit the polarity we see in peripheral cells. One might argue that this property is again solely due to the surrounding environment embryonic cells are exposed to after division. But, then, how does the structure maintain spherical symmetry? And how can apical protein domains consistently be positioned on one side of peripheral cells, while they do not appear in core cells at all?

Using fast live imaging on mouse cells, researchers found that keratin filaments – large intermediate protein fibres known for their function in hair and nails – play a key role in maintaining apical-basal polarity throughout early embryonic development.3 As the apical domain disintegrates during cell division, keratin remains localised on the apical side of the parent cell when it splits via cytokinesis. This apical retention is attributed to the large molecular mass of the filaments, rendering them unable to diffuse rapidly away. Furthermore, keratin and other proteins have been shown to associate with the cytoplasmic actin meshwork in the cell4, which holds them in place during mitosis. As a result, keratin filaments are not present in the daughter cell at the core, and instead are retained at the apical pole of the daughter cell at the periphery. Simply retaining the filaments, however, will not produce the complex network of proteins that form beside the apical membrane in the trophectoderm. Instead, keratin simply facilitates this transition, acting as memory storage that allows the cell to ‘know’ how to position the protein domain. This occurs via a number of multi-step signalling pathways that involve Hippo and Yap proteins5, in which keratin filaments can accelerate the reformation of the apical domain complex as it was before division.

The organisation of embryonic cells continues until the embryo is fully polarised by the 32-cell mark (when it has become a blastocyst), at which point most of the constituent keratin is found at its periphery. Besides acting as a guide for cell specialisation, this feature serves another purpose. As we see in mature skin cells, keratin can be used to stiffen and harden cells – and so we start to see how the trophectoderm becomes the capsule that encompasses the future foetus organs. The same properties allow peripheral cells to gradually shape the embryo as it develops – maintaining its spherical symmetry and keeping inner cells organised and compacted together. This also means that core cells, lacking in keratin, will be more flexible and able to morph into the multilayered system required to produce complex multicellular life. 

Worth mentioning is that keratin filaments are only ever used as fate determinants in these kinds of ‘inside-out’ structures. In comparison, most other specialisation pathways in mammalian tissues are directed by the mitotic machinery itself. Here, fate determinants are actually brought into specified poles during mitosis for asymmetric division to take place.6 Additionally, trophectoderm formation is thought to involve transcription factors that activate trophoblast genes differentially in peripheral cells (e.g. Tead4, phosphorylated in peripheral cells, binds to and activates trophoblast gene Cdx2, though it is inhibited in inner cells), independent of the presence of keratin.5 It is also not clear how the fate of cells at the embryo’s core is assigned without the use of keratin.

Nevertheless, such mechanisms indicate that protein allocation can determine the identity of a cell, and that proteins have roles in storing and carrying information in biological systems. This memory capability is seen not only in mice, but in bacteria, fungi and other multicellular organisms that use protein aggregates to record cell state along differentiation pathways.7 Such findings might suggest alternative methods of manipulating asymmetric differentiation pathways in the lab, with possible applications to artificial organ synthesis and stem cell therapies.


  1. Wiley, LM. Trophectoderm: the first epithelium to develop in the mammalian embryo. Scanning Microsc. 1988;2(1):417-26. Available from:
  1. Zenker J, White MD, Gasnier M, Alvarez YD, Lim HYG, Bissiere S, Biro M, Plachta N. Expanding Actin Rings Zipper the Mouse Embryo for Blastocyst Formation. Cell 2018 Apr 19;173(3):776-791.e17. Available from:
  1. Lim HYG, Alvarez YD, Gasnier M, et al. Keratins are asymmetrically inherited fate determinants in the mammalian embryo. Nature 2020;585:404–409. Available from:
  1. Serres MP, Samwer M, Truong Quang BA, Lavoie G, Perera U, Görlich D, Charras G, Petronczki M, Roux PP, Paluch EK. F-Actin Interactome Reveals Vimentin as a Key Regulator of Actin Organization and Cell Mechanics in Mitosis. Dev Cell. 2020 Jan 27;52(2):210-222.e7. doi: 10.1016/j.devcel.2019.12.011. Epub 2020 Jan 9. PMID: 31928973; PMCID: PMC6983945.
  1. Nishioka N, Inoue K, Adachi K, et al. The Hippo signaling pathway components Lats and Yap pattern Tead4 activity to distinguish mouse trophectoderm from inner cell mass. Dev Cell. 2009 Mar;16(3):398-410. Available from:
  1. Knoblich JA. Asymmetric cell division: recent developments and their implications for tumour biology. Nat Rev Mol Cell Biol. 2010 Dec;11(12):849-60. Available from:
  1. Otzen D, Riek R. Functional Amyloids. Cold Spring Harb Perspect Biol. 2019 Dec 2;11(12):a033860. Available from:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s