WO2018046929A1 - Methods and compositions for co-culturing pluripotent and extra-embryonic cells - Google Patents

Abstract

The present invention relates to methods for culturing mammalian embryonic stem cells and culture media used in such methods.

Description

METHODS AND COMPOSITIONS FOR CO-CULTURING PLURIPOTENT AND

EXTRA-EMBRYONIC CELLS

The present invention relates to methods for culturing mammalian embryonic stem cells and culture media used in such methods.

Morphogenetic transformations during implantation development are critical for mammalian embryo patterning and yet are poorly understood. The first such major transformation is the generation of the pro-amniotic cavity that precedes symmetry breaking to establish the anterior- posterior axis leading to formation of the germ layers and primordial germ cells. These events are achieved through signalling between the pluripotent epiblast, that will generate the foetus, and its enveloping extra-embryonic tissues, trophoblast and primitive endoderm, progenitors of the placenta and yolk sad -6.

It would be beneficial to re-create these early embryonic developmental stages in vitro to enable further study of the early stages of embryonic development. This may enable the discovery of genes that are involved in or essential to this process. Such discoveries may lead to the development of new therapies for myriad conditions, including fertility treatment. It may also lead to new methods for establishing stem cell lines with propensities to differentiate into specific tissues. Previous attempts to re-create these early events in mammalian embryo development in vitro have not achieved pro-amniotic cavity formation and symmetry breaking, although there have been some models (Ten Berge et al 2008; Van den Brink et al, 2014, Warmflash et al, 201 ; full references in Examples - refs 12, 14 and 15) where ES-cell aggregates were able to initiate polarised gene expression. In Bedzhov & Zernicka-Goetz (2014) Cell, 156(5), 1032-44, Meinhardt et al (2014) Stem Cell Reports, 3, 1-13, and Lee et al (2007) Nature Methods, 4(A), 359-65, single mouse embryonic stem cells (ES cells), or in the case of Lee et al (2007), breast epithelial cells, were embedded in a three-dimensional Matrigel® matrix where they divided to form spheroids with a rosette/cystlike structure, possessing a central lumen, which mimics the early-post-implantation epiblast of the mouse embryo. However, when left to develop past 48 hours in culture these ES-rosette structures fail to resemble the embryo and instead develop into neuroepithelium, a default pathway (Meinhardt et al, 2014).

In both Bedzhov & Zernicka-Goetz (2014) and Meinhardt et al (2014) mouse ES cells were cultured using N2B27 medium, which is a defined medium developed for two-dimensional culture of mouse ES cells (Ying et al, 2003, Nature Biotechnology, 21(2), 183-6). This medium was originally developed as a way to differentiate mouse ES cells into neuroectoderm (Ying et al, 2003) but also sustains mouse ES cell pluripotency when combined with 2i (CHIRON and ERK inhibitors) (Ying et al, 2008, Nature, 453(7194), 519-24).

The present inventors have now surprisingly found that it is possible to culture mammalian pluripotent stem cells, such as ES cells, in vitro such that they organise into embryo-like structures undertaking morphogenetic events leading to lumenogenesis, symmetry breaking and mesoderm specification, just as in developing embryos.

Thus, in a first aspect, the present invention provides an in vitro method of culturing a mammalian pluripotent stem cell comprising contacting a mammalian pluripotent stem (PS) cell with an extraembryonic stem cell. In a particularly preferred embodiment, the extra-embryonic stem cell is a trophoblast stem (TS) cell. The extra-embryonic stem cell may alternatively be a primitive- endoderm (PES) stem cell. By "pluripotent stem cell" we mean a stem cell that has the potential to differentiate into any of the three germ layers: endoderm (which forms structures such as the gastrointestinal tract and the respiratory system), mesoderm (which forms structures such as the musculoskeletal system, the vascular system and the urogenital system), or ectoderm (which forms epidermal tissues and the nervous system). The pluripotent stem cells may be derived from any mammalian source, such as early stage embryos, established cell lines derived from mammalian cells, or parthenogenesis of mammalian cells, for example. The mammalian source may be a human, an ape, a monkey, a rat, a mouse, a rabbit, a goat, a sheep, a pig, cattle, a horse or a dog, for example. The extra-embryonic stem cells in the methods described herein are generally intact cells which are capable of growth and are capable of self renewal.

By "trophoblast stem cell" we mean stem cells derived from the trophoblast lineage of the embryo. The trophoblast stem cells of the invention are preferably not extra-embryonic cells derived from the two cell types which are precursors of the human placenta: the cytotrophoblast and the syncitiotrophoblast. In a preferred embodiment these cells may be derived from mouse embryos. These cells are preferably self-renewing stem cells which represent the extraembryonic stem cell compartment of the early embryo. They can be derived at late pre- implantation stages E4.5 or early post-implantation stages (E5.5) but the resulting cell lines are equivalent to the stem cell compartment existing in the extra-embryonic ectoderm of the post- implantation mouse egg cylinder. Transcription factors such as Elf5, Eomes, and Tfap2C mark this lineage. TS cells can also be considered as cells that are the precursors of the differentiated cells of the placenta.

In the mouse, TS cells can be derived from outgrowths of either blastocyst polar trophectoderm or extraembryonic ectoderm, which originates from polar trophectoderm after implantation. TS cells were first derived from the mouse embryo by Tanaka et al (1998) Science, 282(5396), 2072-2075. Rai et al (2015) Developmental Biology, 398(1), 110-119 developed a system to culture TS cells in three dimensions. This method involved culturing the cells in hanging drops or in low-attachment dishes to form aggregates. More recently, Kubaczka et al (2014) Stem Cell Reports, 2(2), 232-242 developed defined culture conditions which permitted the growth and proliferation of TS cells in two dimensions on a Matrigel® layer, without the support of feeder cells. A further method for culturing TS cells in defined conditions is using N2B27 plus Activin A, FGF2, XAV939 (a wnt inhibitor) and Rock inhibitor. In this method, cells are maintained in 2 dimensions on fibronectin. The present inventors have found that these conditions are not suitable for culturing ES and TS cells together.

In an embodiment, the pluripotent stem (PS) cell may be an induced pluripotent stem (iPS) cell. By "induced pluripotent stem (iPS)" we include any somatic (adult) cell that has been genetically reprogrammed to an embryonic stem cell-like state. As would be understood by a person of skill in the art, this genetic reprogramming can involve the cell being forced to express genes and factors important for maintaining the defined properties of an embryonic stem cell. Genetic reprogramming can also involve epigenetic changes such as the changing of chromatin conformation to a more open state. It is envisaged that an iPS cell may ideally have the open chromatin state of a pluripotent embryo. Mouse iPS cells were first reported by Takahashi and Yamanaka (2006) Cell 126(40): 663-76, who demonstrated the ability to induce cells into a pluripotent state using mouse fibroblasts and four transcription factors, Oct4, Sox2, Klf4 and c- Myc. Human iPS cells were first reported by Takahashi et al. (2007) Cell 131(5):861-72 and Yu et al. 2007 Science 318(5858): 1917-20. Mouse iPS cells demonstrate important characteristics of pluripotent stem cells, including the expression of stem cell markers, the formation of tumors containing cells from all three germ layers, and the ability to contribute to many different tissues when injected into mouse embryos at a very early stage in development. Human iPS cells also express stem cell markers and are capable of generating cells characteristic of all three germ layers (endoderm, mesoderm or ectoderm). It is envisaged that the iPS cells may express markers of naive pluripotency, such as Nanog, Oct4, Sox2, Klf4 etc. Further, Oct4 expression, which is under the control of a different enhancer depending on whether a cell is in state of naive or primed pluripotency, is also envisaged as a marker of pluripotency. The cell may express Oct4 under the control of the proximal enhancer. The iPS may be a cell from any mammalian species but it is particularly preferred that it is a human, ape, monkey, rat, mouse, rabbit, goat, sheep, cow, pig, horse or dog derived cell, for example.

In an alternative embodiment, the pluripotent stem (PS) cell may be a mammalian embryonic stem (ES) cell. By "embryonic stem cell" we mean a pluripotent stem cell derived from the inner cell mass of a blastocyst, which is an early-stage preimplantation embryo. It is envisaged that such cells may express genes involved in the naive pluripotency network (Oct4/ Nanog, Sox2, Klf4 etc). Such cells may also have Oct4 proximal enhancer activity. They may contribute to all embryonic tissues in chimeras. It is envisaged that the methods may be carried out with both Naive and Primed pluripotent cells. The ES cells may be derived from mammalian embryos, obtained from iPS cells or obtained from appropriate cell lines. As would be understood by a person of skill in the art, ES cells may be obtained from 'banks' such as the UK stem cell bank from which you can acquire human stem cell lines for research. The Jackson Laboratory, US (who provide Jax mice) also stores and derives mouse ES cells which are available for purchase. It is preferred that the ES cells are obtained or are obtainable by a method that does not involve the destruction of human or non-human animal embryos.

In one aspect, the invention provides an in vitro method of preparing an embryo-like structure wherein said method comprises contacting one or more mammalian pluripotent stem (PS) cell with one or more extra-embryonic stem cells, and contacting said cells with a medium and a substrate which are capable of supporting growth of both the PS cells and the extra-embryonic stem cells in culture. Preferably, said one or more extra-embryonic stem cells are intact. The extra-embryonic stem cells comprise more than one class of extra-embryonic stem cells, optionally wherein said extra-embryonic stem cells comprise trophoblast stem cells.

It is envisaged that in any aspect or embodiment of the present invention, the mammalian pluripotent stem cell may be a mouse or a human pluripotent stem cell.

It is envisaged that in any aspect or embodiment of the present invention, the PS (e.g. iPS or ES) cells and the TS cells may be derived from different species than one another. The resulting embryo-like structure would thus be a chimera of the two species. The embryo-like structure can be a chimeric structure comprising stem cells from two or more sources, wherein said sources comprise at least one species of mammal. In a further embodiment, the method may involve culturing the PS cell and TS cell with one or more further classes of extra-embryonic cell (i.e. in addition to the TS cell). It is envisaged that this will enable the growth of a structure that resembles an embryo even more closely. For example, the additional extra-embryonic stem cell may be a primitive-endoderm (PES) stem cell (known as cultured XEN cells in mouse). Naturally developing mammalian embryos have this second type of extra-embryonic stem cell that helps to specify anterior structures. In an embodiment of the method of the invention, the PS cells may be contacted with a pluripotent stem cell-trophoblast stem cell (TPS) culture medium and a substrate which are capable of supporting growth of both the PS cells and the TS cells in culture. In embodiments where ES cells and TS cells are co-cultured, the medium used could be termed embryonic stem cell- trophoblast stem cell (TES) culture medium. By "substrate" we mean any matrix, network, scaffold etc. that is capable of supporting growth of the cells in the culture medium.

In an embodiment of the culture method of the present invention the method may comprise the earlier step of maintaining the PS cells and TS cells in separate culture for at least one passage. This important step increases the efficiency of the formation of the embryo-like structures of the invention. It is particularly beneficial to maintain the TS cells on a layer of feeder cells which support the growth of the TS cells but do not proliferate themselves. Further, maintaining the cells for at least one passage before use in the methods of the invention allows inspection of the cells to check they are healthy and of good quality. The pluripotency of the PS cells can be verified at this stage by checking cell morphology and by marker staining. Further, passaging of the cells prior to the methods allows for removal of any non-viable cells following thawing of the source cells from liquid nitrogen.

In an embodiment of the invention, the PS cells and TS cells may be obtained from blastocysts. In an alternative embodiment, they may be obtained from a cell line. It is particularly preferred if obtaining the cells does not involve the destruction of a human or animal embryo.

In an embodiment of the culture method of the invention, the PS and TS cells may be removed from culture (i.e. separate culture), washed separately and then combined and contacted with the substrate. For example, a non-limiting example protocol for the preparation of and culturing of ES cells and TS cells in the methods of the invention is as follows:

On day 0 of culture:

Wash ES cells with PBS (phosphate buffered saline) and treat with 0.05% trypsin-EDTA (ethylenediaminetetraacetic acid) for 5-10 mins, then inactivate the trypsin using serum-based culture medium to obtain a single-cell suspension. Treat TS cells with 0.05% trypsin-EDTA for 4 mins and inactivate the trypsin to obtain small clumps of TS cells in suspension. Centrifuge ES cell and TS cell suspensions, remove the supernatant, and wash each pellet of cells using PBS. Centrifuge ES cells and TS cells again, then resuspend each cell-pellet in PBS for a second time.

Using a haemocytometer, count 10,000-40,000 cells/ml for both ES cell and TS cell suspensions. Mix 50 μΙ of ES cells in suspension with 50 μΙ of TS cells in suspension.

Centrifuge this mixture, remove the supernatant, then re-suspend the pellet in ice cold liquid Matrigel®.

Plate the Matrigel®-cell suspension and allow the matrix to solidify for 2 mins at 37 degrees Celsius.

- Once the Matrigel® is solid, flood the wells of the plate with TES culture medium and culture at 5% C02 and 37 degrees Celsius for up to 120 hours / five days.

On days 1-5 of culture: feed cultures with fresh ES-TS culture medium every other day.

Any of these parameters could be adjusted as appropriate by a person of skill in the art as necessary to obtain an equivalent result. For example, alternative buffers and chelators may be used, along with alternative purification methods, as would by understood by a person of skill in the art.

Thus, the method of the invention may comprise the further step of incubating the combined cells in TPS medium, as defined herein.

In an alternative to the 3 dimensional "3D" method described above, a '3D on top'/ sandwich approach (Described in Lee et al, 2007) may be used. This involves seeding cells onto a layer of Matrigel® and then a mixture of medium and 10% Matrigel® is put on top once the cells have attached. This produces the equivalent results as the 3D embedded method described above.

In an embodiment, the method of the invention may involve maintaining the cells in culture for up to 5 days. Thus, the cells may be maintained in culture for 0, 1 , 2, 3 or 4 to 5 days, or 1 , 2, 3, 4 or 5 days, or any fraction thereof. The cells may be maintained in culture for more than 5 days depending on the intended use of the cells, the size of the plate used for culture and the density of the starting culture, as would be understood by a person of skill in the art.

It is particularly preferred that in the methods of the invention, the substrate provides a three- dimensional culture environment. Thus, the substrate may comprise a network/matrix/scaffold. The network/matrix/scaffold may comprise at least one extracellular matrix (ECM) protein, or analogue thereof. It is preferred if the ECM protein is collagen or an analogue thereof, laminin or an analogue thereof, vitronectin or an analogue thereof, fibronectin or an analogue thereof and/or gelatin. These proteins provide a three-dimensional scaffold on which the cells may grow and interact with other cells in all directions, which allows them to form networks more alike those formed in vivo. It is particularly preferred that the ECM protein is collagen or an analogue thereof and/or laminin or an analogue thereof.

In an embodiment, the substrate may comprise laminin, collagen IV, heparin sulphate proteoglycans, entactin/nidogen, and growth factors. This substrate may be Matrigel® Matrix, which is available from Corning B.V. Life Sciences, for example: BD Matrigel Matrix (Basement membrane matrix) Ref: 354230.

The present inventors have devised a co-culture system which supports the growth and development of the two different cell types. The inventors have found that by culturing pluripotent cells such as ES cells in combination with TS cells in a specially developed medium, the cells surprisingly organise themselves into unified embryo-like structures. The cells were grown on a Matrigel® Matrix, which aided in the formation of these structures.

In contrast to 3D aggregates of embryonic stem cells, sometimes known as embryoid bodies, containing only one type of stem cell e.g. 3D structures of cells of a particular type present in a particular organ (organ-specific stem cells), the embryo-like structures of the methods described herein generally comprise two or more distinct stem cell types, preferably one of which is of embryonic origin and at least one of which is extra-embryonic origin. The embryo-like structures preferably comprise one or more mammalian pluripotent stem (PS) cells with one or more extraembryonic stem cells, optionally wherein said extra-embryonic stem cells comprise trophoblast stem cells. The embryo-like structure may express one or more embryonic lineage markers, optionally wherein said embryonic lineage markers comprise Oct4. The embryo-like structure can comprise at least one ES compartment, and at least one TS compartment optionally wherein each of said compartments comprises one or more cavities. The embryo-like structure can comprise one or more cavities. In contrast to embryoid bodies comprising cells from one embryonic tissue only, the culture methods and the embryo-like structures of the invention generally provide an in vitro model which captures the development of the whole organism, rather than just that of one organ.

Pluripotent cells such as ES cells and TS cells have not been combined in culture previously, neither in two dimensional nor three dimensional culture. Each cell type requires different cell culture media and conditions and would not have been expected to thrive in combination. The present invention is particularly surprising since when TS cells are embedded in Matrigel® and cultured alone in their conventional medium, they are capable of growing into aggregates but do not mimic the architecture of the extra-embryonic part of the mouse egg-cylinder. The present inventors have found that it is the combination of both PS cells and TS cells that is required to form these embryo-like structures.

The embryo-like structures generally have a reproducible, characteristic morphology which mimics the mouse embryo. These embryo-like structures undertake morphogenetic events leading to lumenogenesis, symmetry-breaking and mesoderm specification in an ES-derived embryonic compartment bordering a TS-derived extra-embryonic compartment, just as in developing embryos. This symmetry breaking is linked to unification of ES- and TS-cavities and canonical Wnt signalling. When these embryo-like structures develop further, they

spontaneously generate primordial germ cells on the border between ES and TS

compartments in response to BMP signalling, as in the developing embryo. An embryo-like structure may comprise an extra-embryonic compartment abutting an embryonic compartment. An embryo-like structure may comprise a shared fluid-filled cavity at the centre of the cylinder. The embryo-like structure can be surrounded by extracellular matrix, optionally wherein said extracellular matrix comprises laminin. The present invention demonstrates a remarkable inherent ability of two stem cell types to assemble themselves and communicate to specify mesoderm and germline. The accuracy with which these structures recapitulate the spatio-temporal events of embryogenesis in vivo made the inventors term the embryo-like structures derived from a combination of ES cells and TS cells, Trophoblast and Embryonic Stem cell (TES) embryos or synthetic-ET-embryoids.

The embryo-like structure can comprise at least 120 cells. The embryo-like structure can comprise from 200-500 cells, optionally at about 4 to 6 days of culture. The number of cells in each embryo-like structure is in line with the number of cells that make up the post-implantation mouse embryo at comparable stages of development.

The novel medium used by the present inventors to co-culture the two cell types was key to the development of the embryo-like structures (synthetic ET embryoids). This medium was specially developed to support these cells in co-culture and is referred to herein as TPS or TES medium. The medium generally supports the co-culture of two stem cell types at the same time, embryonic stem cells and trophoblast stem cells, in three dimensions in extracellular matrix. This medium allows the embryonic stem cells to exit naive pluripotency and differentiate into an epiblast-like state, whilst maintaining the trophoblast stem cells in a state which permits the formation of an ExE-like compartment.

The medium can contain common ingredients such as insulin, progesterone, transferrin, L- glutamine, sodium pyruvate and DMEM F/12. These promote cell/ tissue growth and survival. These ingredients are used in many conventional cell culture media to provide basic support to cells.

The medium may contain neurobasal A, and may not contain 2i/LIF. This can help promote ESC differentiation into ectoderm. The medium may contain FGF4 and/or heparin. These can help to maintain the self-renewing capacity of the trophoblast stem cells. The medium may contain a small amount of serum. This can allow the promotion of tissue development via signalling factors, without inducing precocious cell differentiation. In an embodiment of the invention, the TPS medium may comprise a basal culture medium supplemented with a non-human serum or serum substitute thereof; L-glutamine or a derivative or analogue thereof; a reducing agent; Fibroblast Growth Factor (FGF), an FGF analogue or FGF receptor agonist; insulin, an insulin analogue or insulin receptor agonist; and progesterone, a progesterone analogue or progesterone receptor agonist. It will be understood that the inclusion of FGF is important for TS cell growth and self-renewal and is therefore a component of conventional media to culture TS cells (Tanaka et al, 1998. Science, 282(5396), 2072-2075). In an embodiment, it is envisaged that the FGF4 could be replaced with basic FGF (FGF2). It will be understood that L-glutamine, or a derivative or analogue thereof, is an essential amino acid required by cells in culture. It is envisaged that any suitable analogue of L-glutamine may be used in its place, such as GlutaMAX Supplement, available from ThermoFisher Scientific. Preferably the L-glutamine is used at a concentration of about 2 mM.

The term "analogue" is used in this specification to refer to a biologically active analogue of any of the components of the culture medium. Such an analogue may be natural or synthetic. The term "analogue" may refer to a compound which may be structurally related to the relevant molecule. The term "agonist" may refer to a compound which might not be structurally related to the relevant molecule. For example, an agonist may activate the relevant receptor by altering the conformation of the receptor. Nevertheless, in both cases the terms are used in this specification to refer to compounds or molecules which can mimic, reproduce or otherwise generally substitute for the specific biological activity of the relevant molecule. It will be understood that the basal medium may comprise water, salts, amino acids, a carbon source, vitamins, lipids and a buffer. The medium may also comprise an albumin. Suitable carbon sources may be assessed by one of skill in the art from compounds such as glucose, sucrose, sorbitol, galactose, mannose, fructose, mannitol, maltodextrin, trehalose dihydrate, and cyclodextrin. Basal media are commercially available, for example, under the trade names Advanced DMEM/F12 (Gibco, 12634-010) and CMRL-1066 (Invitrogen or Sigma).

It will be understood that the non-human serum or serum substitute thereof may be included in the culture medium at a concentration of about 5% to about 50%, about 10% to about 30%, about 10% to about 20%, or about 15% to about 25%, e.g. about 10%, about 15% or about 20%.

The progesterone and/or insulin may be included in the medium as constituents of a neuronal supplement such as "N2 supplement". "N2 supplement" can be mixed from base ingredients or is available commercially. N2 is an important component of N2B27 medium used to support ES cells (Ying et al, 2003 Nature Biotechnology, 21(2), 183-6). A standard representative example of an appropriate N2 supplement recipe is as follows, for 10 ml N2 supplement: 5.357ml (or balance) DM EM F-12 (Gibco, 21331-020) as a base for the supplement; 2.5-3 ml Insulin (Sigma, I9278-5ML); 1 ml Apo-transferrin solution (100mg in 1ml TC grade H20) (Sigma Aldrich, T1147- 100MG); 1 ml Bovine Albumin fraction V (Gibco, 15260-037); 33-60 μΙ Progesterone (0.006g in 10 ml ethanol) (Sigma-Aldrich, P8783-1G); optionally, 100 μΙ Putrescine dihydrochloride (0.8 g in 5 ml TC grade H20) (Sigma-Aldrich, P5780-5G); and 10 μΙ Sodium selenite (0.006 g in 10 ml TC grade H20) (Sigma Aldrich, S5261-10G).

Transferrin is an iron carrier and it may also help to reduce toxic levels of oxygen radicals and peroxide in the culture medium. Putrescine is a precursor of spermidine which binds to NDMA receptors on neural cells. Its inclusion in the N2 supplement for the present invention is optional. Sodium selenite is used as an anti-oxidant. It is a co-factor for glutathione peroxidase and other proteins. Bovine Albumin fraction V is the fifth fraction of bovine albumin which has undergone the Cohn method of fractionating serum proteins. It provides proteins for the culture medium.

In an embodiment, the Insulin, Apo-transferrin, and sodium selenite may be included as ITS-G (Insulin-transferrin-selenium, made by Gibco).

In an embodiment, the TPS medium may comprise about 0.1 mg/l to about 200 mg/l insulin or insulin analogue or about 0.05 ng/ml to about 300 ng/ml insulin receptor agonist. The insulin receptor agonist may be one or more of insulin, IGF-I, and/or IGF-II, and/or an analogue thereof. The concentration of the insulin receptor agonist (e.g. insulin) in the culture medium may be about 0.1 mg/l to about 200 mg/l, about 0.5 mg/l to about 100 mg/l, about 1 mg/l to about 50 mg/l, about 2 mg/l to about 25 mg/l, or about 5 mg/l to about 12.5 mg/l, e.g. about 10 mg/l. The concentration of the insulin receptor agonist (e.g. IGF-1 or IGF-2) in the culture medium may be about 0.05 ng/ml to about 300 ng/ml, about 0.25 ng/ml to about 200 ng/ml, about 1 ng/ml to about 150 ng/ml, about 5 ng/ml to about 100 ng/ml, or about 25 ng/ml to about 75 ng/ml, e.g. about 50 ng/ml. Insulin is an important ingredient in the medium. It promotes glucose and amino acid uptake, lipogenesis, intracellular transport, and the synthesis of proteins and nucleic acids and is essential for the growth of neural stem cells. The inventors found that by increasing the amount of insulin in the medium, they observed a surprising increase in culture efficiency.

In an embodiment, the TPS medium may comprise about 1 ng/ml to about 2 μg/ml progesterone, progesterone analogue or progesterone receptor agonist. The progesterone receptor agonist may be progesterone and/or an analogue thereof. The concentration of the progesterone receptor agonist, or an analogue thereof, in the culture medium may be about 1 ng/ml to about 2 g/ml, about 5 ng/ml to about 1.5 μg/ml, about 10 ng/ml to about 1 μg/ml, about 20 ng/ml to about 750 ng/ml, about 50 ng/ml to about 500 ng/ml, or about 100 ng/ml to about 300 ng/ml, e.g. about 200 ng/ml. Progesterone is also an important ingredient in the medium. Progesterone is a steroid hormone involved in pregnancy, which also has an effect on cell growth. The inventors found that by increasing the amount of progesterone in the medium, they observed an increase in culture efficiency.

It is envisaged that the reducing agent may be 2-mercaptoethanol (2-ME) (β-mercaptoethanol), N-acetyl-L-cysteine, glutathione or dithiothreitol or any other suitable reducing agent. These potent reducing agents may be used in the medium to prevent the build-up of toxic reactive oxygen species in culture, as would be understood by a person of skill in the art. The concentration of the reducing agent in the culture medium may be about 0.5 μΜ to about 250 μΜ, about 5 μΜ to about 200 μΜ, about 7.5 μΜ to about 150 μΜ, about 10 μΜ to about 100 μΜ, about 15 μΜ to about 50 μΜ, about 17.5 μΜ to about 40 μΜ, or about 20 μΜ to about 30 μΜ e.g. about 25 μΜ. Preferably about 100 μΜ.

In an embodiment, the TPS medium may further comprise sodium pyruvate. Sodium pyruvate may be included as an additional carbon source for cells to metabolise as an alternative to glucose. This is a particularly beneficial for the growth of the TS cells in culture. It is envisaged that the inclusion of sodium pyruvate may avoid a lag in TS cell growth that may result from altering culture conditions abruptly following the washing and re-suspension steps in the method highlighted above. Sodium pyruvate may be used at a concentration of at about 2 mM to about 0.1 mM, preferably about 1 mM to about 0.25 mM, even more preferably about 1 mM to about 0.5 mM, particularly preferably at 0.5 mM.

In an embodiment, the TPS medium may further comprise further non-essential amino acids (NEAA). For example, the NEAA may be selected from any one or all of L-glycine, L-alanine, L- asparagine, L-aspartic acid, L-glutamic acid, L-proline and L-serine each at a concentration as further defined herein. The non-essential amino acids that may be included in the culture medium, for example, comprise glycine (about 1 mg/l to about 25 mg/l or about 5 mg/l to about 10 mg/l e.g. about 7.5 mg/l), L-alanine (about 1 mg/l to about 25 mg/l or about 5 mg/l to about 10 mg/l e.g. about 9 mg/l), L-asparagine (about 5 mg/l to about 30 mg/l or about 10 mg/l to about 15 mg/l e.g. about 13.2 mg/l), L-aspartic acid (about 5 mg/l to about 30 mg/l or about 10 mg/l to about 15 mg/l e.g. about 13 mg/l), L-glutamic acid (about 5 mg/l to about 50 mg/l or about 10 mg/l to about 20 mg/l e.g. about 15 mg/l), L-proline (about 5 mg/l to about 30 mg/l or about 10 mg/l to about 15 mg/l e.g. about 11 mg/l) and/or L-serine (about 5 mg/l to about 30 mg/l or about 10 mg/l to about 15 mg/l e.g. about 11 mg/l). Preferably, the culture medium may further comprise L-glycine at a concentration of about 7.5 mg/l, L-alanine at a concentration of about 9 mg/l, L-asparagine at a concentration of about 13 mg/l, L-aspartic acid at a concentration of about 13 mg/l, L-glutamic acid at a concentration of about 14.5 mg/l, L-proline at a concentration of about 11.5 mg/l and L-serine at a concentration of about 10.5 mg/l.

In an embodiment of the invention, the non-human serum may be foetal bovine serum (FBS). It will be understood that foetal bovine serum is important as it contains growth factors and morphogens to enhance cell growth and proliferation. The serum or serum replacement may be included in the culture medium at about 5% v/v to about 60% v/v, about 10% v/v to about 50% v/v, about 15% v/v to about 45% v/v, or about 20% v/v to about 40% v/v. It is particularly preferred if the TPS medium comprises 10 % v/v FBS. FBS is serum isolated from the blood of calves, which contains many growth factors in undefined quantities which enhance cell survival in culture. The FBS may be replaced with a specifically defined serum replacement, such as defined 'Knockout serum replacement' KSR. Such serum replacement media are commercially available under the trade names KSR (KnockOutTM Serum Replacement, Invitrogen, 10828- 010) and N2B27 (e.g. Invitrogen, ME100137L1). Alternatively, the culture medium may comprise a serum replacement medium as described in WO 98/30679 (in particular, Tables 1 to 3), the contents of which is expressly incorporated by reference. The serum replacement medium may be included in the culture medium at about 5% to about 60%, about 10% to about 50%, about 15% to about 45%, or about 20% to about 40%, e.g. about 30%. The culture medium may further comprise one or more of transferrin, selenium (for example sodium selenite, in this case provided as a salt), and/or ethanolamine, and/or an analogue thereof. Preferably, the culture medium comprises transferrin, selenium (for example sodium selenite, in this case provided as a salt) and ethanolamine. For example, the culture medium may comprise ITS-X (Invitrogen, 51500-056). The concentration of transferrin, or an analogue thereof, in the culture medium may be about 0.01 mg/l to about 500 mg/l, about 0.05 mg/l to about 250 mg/l, about 0.1 mg/l to about 100 mg/l, about 0.5 mg/l to about 25 mg/l, about 1 mg/l to about 10 mg/l, or about 2.5 mg/l to about 7.5 mg/l, e.g. about 5.5 mg/l. The concentration of selenium (for example sodium selenite), or an analogue thereof, in the culture medium may be about 0.0001 mg/l to about 0.1 mg/l, about 0.0002 mg/l to about 0.05 mg/l, about 0.0005 mg/l to about 0.02 mg/l, about 0.001 mg/l to about 0.01 mg/l, or about 0.005 mg/l to about 0.0075 mg/l, e.g. about 0.0067 mg/l. The concentration of ethanolamine, or an analogue thereof, in the culture medium may be about 0.01 mg/l to about 500 mg/l, about 0.025 mg/l to about 250 mg/l, about 0.05 mg/l to about 100 mg/l, about 0.1 mg/l to about 50 mg/l, about 0.25 mg/l to about 25 mg/l, about 0.5 mg/l to about 10 mg/l, or about 1 mg/l to about 5 mg/l e.g. about 2 mg/l.

It is envisaged that the basal culture medium may be Dulbecco's Modified Eagle Medium (DMEM), Roswell Park Memorial Institute (RPMI) medium 1640, or Neurobasal® or Neurobasal® A or a mixture thereof. It is particularly preferred if the basal culture medium is a mixture of DMEM, RPMI and Neurobasal® A. Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 is a standard cell culture medium used to support a wide variety of cell types. This was a base component of N2B27 used to maintain and differentiate ES cells (Ying et al, 2003 Nature Biotechnology, 21(2), 183-6). Roswell Park memorial Institute medium is traditionally used to culture lymphoid cells and has a high concentration of phosphate. It is the standard base medium for culturing TS cells. It provides the basic nutrients required by the cells. Neurobasal® A is a base medium conventionally used to culture neural cell types. This was a base component of N2B27 used to maintain and differentiate ES cells (Ying et al, 2003 Nature Biotechnology, 21 (2), 183-6). In a particularly preferred embodiment, the basal culture medium of the invention is a mixture of 50 % RPMI, 25 % DMEM and 25 % Neurobasal® A. The mixture may alternatively be 40 % RPMI, 30 % DMEM and 30 % Neurobasal® A, or 30% RPMI, 35% DMEM, and 35% Neurobasal® A or any other appropriate combination as could be determined by a person of skill in the art. In an embodiment, the TPS/TES medium may further comprises an antibiotic(s) and/or another antimicrobial, for example antibacterial, compound. The antibiotic may be penicillin and/or streptomycin. These antibacterial compounds are included to prevent contamination of the culture with bacteria, which would adversely affect the developing embryo-like structure. Penicillin may be included in the culture medium at a concentration of about 1 unit/ml to about 500 units/ml, about 2 units/ml to about 250 units/ml, about 5 units/ml to about 100 units/ml, about 10 units/ml to about 50 units/ml, or about 20 units/ml to about 30 units/ml e.g. about 25 units/ml. Streptomycin may be included in the culture medium at a concentration of about 1 μg/ml to about 500 μg/ml, about 2 pg/ml to about 250 μg/ml, about 5 pg/ml to about 100 μg/ml, about 10 μg/ml to about 50 μg/ml, or about 20 μg/ml to about 30 μg/ml e.g. about 25 μg/ml. Preferably, the culture medium may comprise penicillin at a concentration of about 25 units/ml and/or streptomycin at a concentration of about 25 μg/ml.

In a further embodiment, the TPS/TES medium may further comprise heparin. Heparin is used to assist binding of FGF4, required for TS cells.

In a particularly preferred embodiment, the TPS/TES medium comprises 50 % RPMI, 25 % DMEM F-12 and 25 % Neurobasal A, supplemented with 10 % FBS, 2mM L-glutamine, 0.1 mM 2ME, 0.5 mM sodium pyruvate, 0.25x N2 supplement, 0.5x B27® supplement, 12.5 ng/ml FGF4 and 50mg/ml heparin.

B27® supplement is a commercial supplement originally developed for supporting neural lineages such as retinal ganglion cells and is available from ThermoFisher Scientific. It would be understood by a person of skill in the art that alternative supplements are available and may be substituted for this in the methods and media of the invention.

Heparin is a cofactor known to facilitate the uptake of FGFs and stabilizes these proteins so it may be included in the media for this purpose but is not essential.

In a particularly preferred embodiment, the TES medium of the invention may be made using the following recipe (for 10 ml): 5 ml RPMI 1640 (Sigma Aldrich, M3817); 2.5 ml DMEM F-12 (Gibco, 21331-020); 2.5 ml Neurobasal A (Gibco, 10888-022); 50 μΙ N2 supplement; 100 μΙ B27 supplement (Gibco, 10889-038); Supplemented with: 10 % foetal bovine serum (Stem cell institute); 2 mM L-glutamine (Gibco, 25030-024); 0.5 mM sodium pyruvate (Gibco, 11360039); 0.1 mM 2-mercaptoethanol (Gibco, 31350-010); FGF4 (12.5ng/ml) (Peprotech); and Heparin sodium salt (500ng/ml) (Soluble to 50mg/ml. I make 1000x (1 mg/ml) stock by dissolving powder in 1xPBS.) (Sigma).

In an alternative embodiment of the medium of the invention, both the B27 and N2 substituents of the TES medium may be replaced with SOS® (Cat No M09-50, Cell Guidance Systems). Other neuronal supplements are also available and would be expected to be effective, as would be understood by a person of skill in the art.

It is intended that the method of the invention will be capable of allowing PS cells, in particular ES cells, and TS cells to organise into embryo-like structures. These embryo-like structures may have a number of commercial uses, such as research tools for studying embryo development and identifying genes that are important in this process and in fertility, tools for drug testing, material for cell based therapies, etc. It is envisaged that the embryo-like structures will undertake lumenogenesis, symmetry breaking and mesoderm specification.

In an embodiment of the methods of the invention, the PS cells may be genetically modified. This may include alterations such as the deletion or insertion of genes or chromosomes, or alteration in epigenetic markers, or transcription factor addition or removal, or alteration in extra- chromosomal DNA, such as mitochondrial DNA. Such methods may be used to study the role of certain genetic markers or traits in embryo development or in the functioning of the cell or embryo. Modified cells may be used to establish cell lines for use in methods of treatment or diagnosis.

The present invention thus provides an in vitro cell culture medium comprising the TPS medium as described herein. The medium has the special property of enabling embryo-like structures for form when used to culture PS cells, such as ES cells, in combination with TS cells, as herein described.

The invention also provides a culture medium supplement for producing the in vitro culture medium of the invention comprising an insulin receptor agonist and a progesterone receptor agonist. The culture medium supplement may comprise insulin, or analogue thereof. The medium may comprise progesterone, or an analogue thereof.

The culture medium supplement can be constituted such that when converted to the final medium for use in the in vitro culturing of embryos, any of the in vitro culture media embodiments defined herein are produced. In all cases, upon conversion, the final medium thereby produced is capable of supporting development of an embryo-like structure from co-cultured PS and TS cells in the presence of a suitable substrate as defined herein. Any of the optional additional components, such as defined herein, may be included in the culture medium supplement or may be provided as separate supplements. Components of the supplement may be provided in amounts such that when reconstituted any of the working amounts defined herein are produced, provided that the medium is capable of supporting development of an embryo-like structure from co-cultured PS and TS cells in the presence of a suitable substrate as defined herein. The culture medium supplement may be constituted such that the individual components are concentrated relative to the final in vitro culture medium by between about x5 to about x500, about x25 to about x250, about x50 to about x200, or about x75 to about x150 e.g. about x100.

The present invention also provides a kit for culturing a mammalian cell comprising the in vitro cell culture medium of the present invention and a substrate (e.g. ECM) as defined herein. The kit may also comprise any of the culture medium supplements, as defined herein, for producing the in vitro culture medium of the invention and a basal medium, as defined herein, and/or one or more separate supplements comprising one or more of the components as defined herein. The kit may further comprise one or more receptacles suitable for containing a culture comprising the substrate and medium in combination with PS and TS cells. The receptacle may be a solid support made of a plastics material or glass. Preferably, the receptacle is suitable for imaging, for example time-lapse imaging. The methods of the invention may further comprise the step of recording one or more images of the embryo-like structure. Additionally, or alternatively, the methods may further comprise the steps of contacting the embryo-like structure with a test agent and determining the effect of the test agent on development of said embryo-like structure. Thus, the present invention also provides an embryo-like structure obtainable by the methods of the invention. Such structures may be isolated or maintained in culture to produce stable cell lines.

The present invention also provides for the use of the in vitro culture medium of the invention for culturing PS cells in combination with TS cells. Preferably, the PS cells and TS cells are cultured in a substrate (e.g. ECM) as described herein.

The present invention also permits the means of taking specific cell types from the synthetic-ET- embryoid structures, for example primordial germ cells, so that this specific cell type can be expanded for other applications in regenerative medicine. It also permits the removal of pluripotent cells from the ET-embryoid so that they can be differentiated in vitro into other cell types. The invention also provides a method of investigating mechanisms involved in embryogenesis, comprising the in vitro method of culturing a mammalian pluripotent stem cell of the invention. The methods of the invention may thus be used to study the morphological development of the embryo by providing an in vitro model system. This can be used to study the impact of particular genes in the process by deleting, modifying or overexpressing them, or to study the effect of external factors such as hormones, nutrients etc. on the growth and development of the embryo. The cells may be monitored using techniques described in the Examples. Thus, the present invention also provides a method of identifying a compound useful for treating or preventing a disease, the method comprising contacting a cell or embryo-like structure obtainable by the in vitro method of culturing a mammalian pluripotent stem cell of the invention, with the compound and determining the effect of the compound on the cell or embryo-like structure. Such test compounds may be added to the culture media of the invention during or after the establishment of the embryo-like structure. The step of determining the effect of the compound on the cell or embryo-like structure may comprise comparing a phenotype or a genotype in the present of said compound with the phenotype or genotype in the absence of the compound. The invention also provides the use of any compound identified by such methods in the treatment or prevention of said disease.

The invention also provides a cell or embryo-like structure obtainable by the in vitro method of culturing a mammalian pluripotent stem cell of the invention, for use in a method of diagnosing, preventing or treating a disease in a patient in need thereof. For example, cells obtainable from the present invention may be used in stem cell therapies, such as treatments for cancers, replacement tissue, reconstructive surgery, tissue repair, wound healing, bone marrow transplantation, stroke, baldness, blindness, deafness, diabetes, heart disease, bowel disease, arthritis, skeletal injury, teeth replacement, neuronal disease and any other condition where replacement cells or tissues may be advantageous. The cells may also be utilised for screening therapeutic compounds for efficacy and safety, as would be understood by a person of skill in the art.

In an embodiment, the cell or embryo-like structure for use in a method of diagnosing, preventing or treating a disease in a patient in need thereof as described herein, may be used for transplantation into the patient. It is envisaged that in certain embodiments, the pluripotent cell used to obtain the cell or embryo-like structure may have been obtained from the patient originally, thus reducing the likelihood of rejection by the patient's immune system. Thus, a pluripotent stem cell, for example an induced pluripotent stem cell, obtained from a patient may be cultured using the methods of the invention to provide material for transplantation back into that patient to prevent or treat a condition. For example, the embryo-like cell may be used to grow replacement organs or tissues for the patient to regain function of such organs or tissues in the patient following loss of function through degeneration, ageing and/or disease.

The present invention also includes a method of providing a transgenic non-human animal, comprising gestating an embryo derived from a cell cultured using an in vitro method of culturing a mammalian pluripotent stem cell of the invention. Other extra-embryonic cell types such as primitive-endoderm stem cells (XEN cells in mice) may be incorporated into the embryoid culture to facilitate the development of an embryo capable of development to term. Such transgenic non- human animals may be useful in drug screening or in the study of disease. For example, model animals may be produced to study specific conditions. It is envisaged that the novel methods provided herein could be used to more efficiently develop transgenic and chimeric embryos (which currently relies for example, on the labour-intensive process of harvesting blastocysts and manually replacing the inner cell mass).

The invention also provides a method of elucidating the role of a gene in embryo development, the method comprising obtaining a pluripotent cell where the gene has been modified or knocked out and culturing said cell using the in vitro method of culturing a mammalian pluripotent stem cell of the invention. Thus, the methods may aid in the development of treatments for conditions relating to embryo development, such as fertility treatment.

The invention also provides a method of imaging an embryo during development comprising culturing a mammalian embryo-like structure using the methods of the invention and imaging apparatus, and recording an image of said embryo. The image may be a two dimensional or three dimensional image. A plurality of images may be recorded of the same embryo.

The invention also provides an imaging apparatus comprising a kit of the invention, microscopy apparatus and suitable recording apparatus. An imaging apparatus may further comprise image processing apparatus. Additionally, an imaging apparatus may further comprise a fluorescent microscope. Additionally, or alternatively, an imaging apparatus may further comprise a confocal microscope.

The invention also includes any novel method of culture or culture medium described herein with reference to the Description, Examples and Figures. FIGURES

Figure 1 : Self-assembly of mouse ES cells and TS cells generates a structure

that mimics the mouse embryo, a. Graphical scheme of the protocol used for TESembryo generation. 1. ES cells and TS cells are maintained in adherent culture in

standard conditions. 2. Single ES cells and small clumps of TS cells are suspended in

Matrigel and plated in drops which are allowed to solidify. 3. Solid drops are cultured in TES-embryo medium (Methods). 4. Embryo-like structures emerge within 96 hours of culture, b. TES-embryo after 96 hours of development and stained to reveal: red,

Oct4 (equivalent to mid grey in greyscale); green, Eomes (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale). From left hand side, 1^st Panel: Brightfield image; 2^nd Panel: Oct4 (mid grey, e.g. in bottom section) and DAPI (dark grey, e.g. in top section); 3^rd Panel: Eomes (light grey) and DAPI (dark grey). 4^th Panel: composite image produced through merging of 2^nd and 3^rd panels. Scale bar=20 m. n=20. c. TES-embryo expressing the embryonic lineage marker, Oct4. Red, Oct4 (equivalent to mid grey in greyscale); Blue, DAPI (equivalent to dark grey in greyscale). From left hand side, 1^st Panel: Brightfield image; 2nd Panel: DAPI (dark grey); 3^rd Panel: Oct4 (mid grey, e.g. in bottom section); 4^th Panel: composite image produced through merging of 2^nd and 3^rd panels. Scale bar =20μιη. d. Embryo cultured in w^'iro for 48 hours from blastocyst stage and stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); green, Eomes (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale). From left hand side, 1^st Panel: Brightfield image; 2^nd Panel: Oct4 (mid grey, e.g. in bottom section) and DAPI (dark grey); 3^rd Panel: Eomes (light grey) and DAPI (dark grey). 4^th Panel: composite image produced through merging of 2^nd and 3^rd panels. n=20. Scale bar= 20μπι. e. Post-implantation embryo recovered at embryonic day E5.5 and stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale). From left hand side, 1^st Panel: Brightfield image; 2^nd Panel: DAPI (dark grey); 3^rd Panel: Oct4 (mid grey). 4^th Panel (iv) composite image produced through merging of 2^nd and 3^rd panels. Scale bar =20 m. f. Comparison of cell number. Quantification of mean cell number in embryonic and extra embryonic compartments for TES-embryos and egg cylinder stage embryos cultured for 48 hours in vitro from blastocyst stage. From left hand side, 1^st Column, TES Embryo; 2^nd Column, Post-implantation embryo; Y-axis: Number of cells. OCT4 positive cells in grey and EOMES positive cells in white. TES-embryos after 96 hours of development have similar number of cells to natural E5.5 embryos in embryonic and extra-embryonic compartments (Student' s t- test of means, n=20 per group, (2 separate experiments) not significant). Error bars = SEM. g. Comparison of area. Quantification of mean area of embryonic and extra-embryonic compartments at the middle Z plane for TES-embryos after 96 hours of development and egg cylinder stage embryos cultured for 48 hours in vitro from late blastocyst stage. From left hand side, 1^st Column, TES Embryo; 2^nd Column, Post-implantation embryo, Y-axis, Area (μηΊ₂). Embryonic cells in grey; Extra-embryonic cells in white. Areas of embryonic and extra-embryonic parts are similar (Student's t-test of means, n=20 per group (2 separate experiments), not significant). Error bars=SEM. NB: Area occupied by the visceral endoderm was excluded from quantification in natural embryos as there is no equivalent tissue for comparison in TES-embryos. h. Developmental time-course showing TES-embryos from 72 to 96 hours in culture, (i) Brightfield; (ii) Stained to reveal ES cells (red, equivalent to mid grey in greyscale) and TS cells (blue, equivalent to dark grey in greyscale). (iii) Illustration of 2D nuclear segmentation over time: From left hand side, 1^st panel: ES cavity; 2^nd panel: Multiple cavities in ES and TS; 3^rd panel: Continuous cavity. ES cells (red) form a cavity first, followed by TS cells (blue). Scale bar=20pm. Yellow (equivalent to white in greyscale in (i) and (ii),) dotted/black lines (equivalent to dark grey in (iii)) highlight cavitation progression over time. 2D segmentation at the middle Z-plane was performed to better visualise nuclear shape.

Figure 2: TES-embryos cavitate similarly to natural mouse embryos, a. A TES embryo after 72 hours of development compared with an E5.5 embryo recovered from the mother stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); green, aPKC (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale). Panels (i) and (ii) E5.5 Embryo ; Panels (iii) and (iv) TES Embryo. White arrows and insets show the cells of the epiblast enriched with aPKC on the cavity-side in both cases. Scale bar = 20μιη. n=20, 2 separate experiments, b. Time-course of TES embryos during cavitation, and accompanying quantification of PCX intensity. Schematics alongside indicate morphology at 72 hours, 84 hours, 96 hours (Red, equivalent to mid grey in greyscale, embryonic compartment; Blue, equivalent to dark grey in greyscale, extra-embryonic compartment). Staining indicates red, Oct4/ aPKC (equivalent to mid grey in greyscale); green, PCX( equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale).. Scale bar= 20 m. n=30, 3 separate experiments. Adjacent panel: PCX staining intensity (x-axis, distance (Mm), y-axis, pixel intensity) over the part indicated by the dotted lines across ES compartments and TS compartments respectively at the middle Z plane. White arrows indicate the co-incidence of aPKC and PCX in the magnified inserts. Asterisks indicate multiple small cavities in the TS- compartment at 84 hours.

c. A TES-embryo after 72 hours showing the shape of cells in ES- and TS compartments. Stained to reveal red, Oct4 (equivalent to mid grey in greyscale); green, E-cadherin (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale). Insets and yellow dotted lines (equivalent to white in greyscale) highlight cell shape in central and two adjacent panels on right hand side. White dotted lines in bottom panel highlight cavities (equivalent of grey in greyscale). Scale bar= 20μπι. n=30, 3 separate experiments, d. Mean cell aspect ratio. Column: A, TS edge; B, TS boundary and C, ES. Quantification showing mean cell aspect ratio in TES-embryos is significantly different between ES- and TS-compartments after 72 hours of development (ANOVA test of means, P<0.001 , n=30 per group, pooled from 3 separate experiments, error bars= SEM).Y-axis: Mean cell aspect ratio. Scale from 0 to 1.2 in increments of 0.2.

Figure 3: Breakdown of basal lamina between compartments of TES-embryos and onset of mesoderm specification a. TES-embryos during cavitation and stained to reveal: red, Oct4 (Equivalent to mid grey in greyscale); yellow/black, Laminin (equivalent to light grey in greyscale in row A, black in row B); blue, DAPI (equivalent to dark grey in greyscale). Upper Row A: stained for Oct4, DAPI and Laminin; Lower Row: stained for laminin. Colouring for antibodies in greyscale corresponds to the colouring of the corresponding antibody labels. Time in matrigel from 72 hours (left hand side to 96 hours (right hand side) From left hand side, 1^st panel: Before joining; 2nd and 3^rd panels: Mid-joining; 4^th panel: Joined. White arrowheads indicate cells from the ES-compartment extending into the TS-compartment (Upper row). Yellow arrows (equivalent to white in greyscale) and insets indicate residual laminin at the compartment boundary (Lower row). n=20 TES-embryos at each time point; 2 separate experiments. Scale Bar= 20μιη. b. Two fused ES compartments after 84 hours of development and stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); yellow (equivalent to white in greyscale )/black, Laminin. Scale bar= 20μιη. Inset shows residual laminin that is not displaced towards either of the fusing compartments. n=12; 2 separate experiments

c. Angle of laminin boundary breakdown. First column: ES + ES; Second column: ES + TS. Angular displacement of laminin as the boundary is broken in ES-ES structures (n=12, mean Θ = 91.05^° , pooled from 2 separate experiments) and TES-embryos (n=13, mean θ= 80.3^° pooled from 2 separate experiments) is significantly different.= (Student' s t test of means, P<0.01 , Error bars= SEM). For a description of how Θ was measured, see Methods. Y-axis: Angle of displacement. Scale from 50 to 110 in 10 degree increments.

d. A TES-embryo during cavitation stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); cyan, Laminin (equivalent to very light grey in greyscale); white, DAPI. Insets highlight the boundary region. White arrowheads indicate residual laminin. Yellow (equivalent to white in greyscale) dotted lines trace the shape of the chimeric floret-like cell arrangements at the boundary. n=15 TES-embryos, 2 separate experiments. Scale bar= 20 m.

e. A TES-embryo during cavitation stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); cyan, Laminin (equivalent to very light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale). Colouring for antibodies in greyscale corresponds to the colouring of the corresponding antibody labels. Yellow (equivalent to white in greyscale) line indicates chimeric cell arrangements, and white dotted lines trace the outline of the cavity. Scale bar= 20μπι. n=15, 2 separate experiments. Right- a schematic drawing to represent a TES-embryo during cavitation with a chimeric cell floret at the boundary.

f. Confocal snapshots images of ES cell-structures alone in Matrigel and TES-embryos expressing T:GFP (green, equivalent to light grey in greyscale).

Scale bar=20pm. From left hand side, 1^st panel: ES cells in Matrigel; 2^nd panel: TES Embryo; 3^rd panel: Gastrula stage embryo. White arrowheads and white dotted lines indicate the outline of each structure (outer) and a cavity (inner). n=100 TES-embryos analysed, 4 separate experiments. Right hand pair: a gastrula-stage embryo expressing T/Brachyury

(green, equivalent to light grey in greyscale). Scale bar= 20μητ

g. Percentage of T:GFP positive structures. Columns: A, ES-TS, and B, ES alone. Percentage of T:GFP +ve in grey and T:GFP -ve in white. Quantification showing that the proportion of TES embryos expressing T.GFP after 96 hours in culture is significantly increased compared to expression in ES cell-structures alone in Matrigel. (Fisher's exact test, P<0.001 , n=108, 64 TES-embryos and 44 ES-cell structures counted, 2 separate experiments. Error bars=SEM, data were normalised with a square-root transformation before a parametric statistical test was performed.)

h. The primitive streak region of a gastrula stage embryo, compared to the T:GFP-positive region of a TES-embryo. Left hand side, E6.5 embryo (early streak); Right hand side TES Embryo. Blue, DAPI (equivalent to dark grey in greyscale); green, Brachyury/ T:GFP

(equivalent to light grey in greyscale). Zoom and insets highlight the mesoderm region in the TES-embryo and in the natural embryo. Scale bar= 20μηη. XZ panel highlights the asymmetry in T/Brachyury to one side of the structure in both cases.

i. TES Embryo (43%). A TES-embryo expressing T. GFP in the OCT4-positive region after 100 hours of development. Stained to reveal: red, Oct (equivalent to mid grey in greyscale); green, T:GFP (equivalent to light grey in greyscale) ; blue, DAPI (equivalent to dark grey in greyscale). n=20.

j. Percentage of structures with asymmetric T:GFP expression. Columns: A, ES-TE, and B, ES alone. The proportion of T/Brachyury-positive TES-embryos with a polarised/ asymmetric domain of expression with respect to the long axis of the structure is significantly increased compared with ES cells-structures alone in Matrigel (Student's t-test, P<0.001 , n=100

TES-embryos and n=100 ES cell-alone structures counted per experiment, mean

proportion calculated 4 separate experiments, Error bars= SEM). Y-axis: Percentage of structures with asymmetric T:GFP expression, scale 0 to 50% in 10% increments. k. TES Embryo mid-cavitation. A TES-embryo during cavitation and the initiation of

T/Brachyury expression, stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); green, T:GFP (equivalent to light grey in greyscale); yellow/black, Laminin (equivalent to white (panels (ii) and (Hi)) / black (panel (i) in greyscale). Scale bar= 20pm. White asterisk indicates T:GFP positive cell at the boundary. White arrows and inset indicate displaced residual laminin at the opposite side to the T:GFP-positive cell. n=10 (2 separate experiments). Panel i:

Brightfield view, laminin (black). Panel ii: Stained for T-GFP (light grey), laminin (white). Panel iii: Merge; stained for T-GFP (Light grey); Laminin (white); Oct4 (mid grey); DAPI (dark grey). Figure 4: TES-embryos develop mesoderm in response to Wnt signalling and

primordial germ cell-like cells in response to BMP signalling, a. TES-embryos

expressing the Wnt reporter H2B-GFP:Tcf/LEF and T/brachyury at 90 hours, 96 hours, and 102 hours (Columns from Left to right: 90 hours (left); 96 hours (middle); 102 hours (right)). Stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); green, H2B-GFP:Tcf/LEF (equivalent to light grey in greyscale); white, T/Brachyury. Top row: Stained for Oct4, DAPI; Middle row stained for H2B-GFP:Tcf/LEF, DAPI, Bottom row: Stained for Brachyury, DAPI. Scale bar = 20 m. Insets highlight cells co- expressing the Wnt reporter and T/Brachyury. n=15 per each time-point. Scale bar in zoom= 10 μιη.

b. No. of co-expressing cells (Y-axis, Number of cells). Quantification of the mean number of Wnt/ Brachyury co-expressing cells detected in the ES-compartment of TES-embryos after 90 hours (A), 96 hours (B), and 102 hours (C) of culture. n= 15 per time-point (3 separate experiments). Error bars = SEM. Y-Scale: -10 to 50 in increments of 10.

c. TES-embryos cultured in the presence of 200ng/ml DKK1 (B, right hand side) and in control conditions (A, left hand side) for 96 hours and stained to reveal: red, Oct4 (Equivalent to mid grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); white, T/Brachyury. Top row: Brachyury only. Bottom row: Oct4, DAPI and Brachyury. Yellow (equivalent to white in greyscale) arrows indicate the presence of T/Brachyury-positive mesodermal cells in control conditions, which are undetectable in the presence of DKK1. Scale bar= 20μιτι. n=15 per group (2 separate experiments).

d. A TES-embryo after 120 hours of development showing an asymmetric domain of mesoderm and putative PGCs, stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); white, Brachyury; green, Tfap2c (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale)). From left hand side, 1^st Panel: Brachyury (white) and DAPI (dark grey); 2^nd Panel: Oct4 (mid grey, e.g. in bottom section) and TFAP2C stained (light grey, e.g. in top right-hand side); 3^rd Panel: Max. proj image. Insets highlight the Oct4-Tfap2c double positive cells which occupy the boundary in the Brachyury-positive region of the TES- embryo. Scale bar= 20μηι. n=13 (2 separate experiments).

e. A TES-embryo after 120 hours of development, expressing Stella:GFP concomitantly with phospho- SMAD1. (Green, Stella-GFP (equivalent to light grey in greyscale); White, p-SMAD1 ; Red, Oct4 (equivalent to mid grey in greyscale); Blue, DAPI (equivalent to dark grey in greyscale). From left hand side, 1^st Panel: SMAD1 (white) and Stella-GFP (light grey); 2^nd Panel: Oct4 (mid grey), DAPI (dark grey) and Stella-GFP (light grey) (e.g. in box and corresponding inset); 3^rd Panel Max.proj image. Scale bar= 20μηη. n=15 (3 separate experiments). Insets indicate Stella:GFP-positive cells in the ES-compartment.

f. Percentage of ST:GFP positive structures. Column A, TES Embryo; Column B, ES cells in Matrigel. Within Columns A and B, ST:GFP +ve in dark grey and ST:GFP -ve in light grey. Quantification showing the proportion of TES-embryos expressing Stella:GFP is significantly increased compared to expression in ES cell-structures alone in Matrigel. (Fisher' s exact test, P<0.05, n=30 TES-embryos, n=23 ES-cell structures (2 separate experiments), Error bars= SEM). Y-scale 0 to 100% in increments of 20%.

g. TES-embryos cultured for 96 hours in control conditions (top row) or in the presence of BMP antagonist Noggin (50ng/ml) (bottom row) and stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); white, P-SMAD1 ; green, Stella-GFP (equivalent to light grey in greyscale). From left hand side, 1^st Panel: P-SMAD1 ; 2^nd Panel: P-SMAD1 , OCT4 and Stella:GFP; 3^rd Panel: OCT4, DAPI and Stella:GFP. Scale bar=20μm. n=15 TES-embryos analysed per group (2 separate experiments).

h. Schematic of development of the mouse embryo through implantation up to

mesoderm specification compared to TES-embryos. TES-embyros shown in top row, embryos in bottom row. Red cells=ES/epiblast cells in each case (A; equivalent to dark grey in greyscale); Dark Blue cells=TS/Trophectoderm/ Extra-embryonic ectoderm cells (B; equivalent to mid grey in greyscale); Light green cells=Brachyury-positive cells (C; equivalent to mid to light grey in greyscale); Purple cells=Primordial Germ Cells (Equivalent to darkest grey in bottom row in greyscale). Yellow= basement membrane/ECM (Equivalent to light grey in bottom row in greyscale). In the embryo, Dark green cells=Primitive endoderm /Visceral endoderm cells. In the TES-embryo, ECM (laminin) surrounds the entire TES-embryo, similar to the basement membrane secreted by the visceral endoderm in natural embryos. White dotted line highlights a chimeric cell arrangement at the boundary between ES- and TS- compartments at cavity fusion. Mesodermal domain occupies a similar area of the embryonic compartment in both TES-embryos and natural embryos. After 50 hours we detect clustering in the ES cells before compartments join at 70 hours. After joining, cavitation occurs progressively in the ES compartment, then in the TS compartment. Unification then occurs by 96 hours, which is followed by mesodermal and PGC specification in TES-embryos. E4.5, E4.75: Epiblast polarization; E5.0: Lumen formation; E5.25, E5.5: Trophoblast re-organisation; E5.75, E 6.0: Pro-amniotic cavity formation; E6.25, E6.5: Primitive streak formation. At 50 hours, ES rosette lumenogenesis; TS cavity formation after 70 hours, Cavity merge, symmetry breaking, mesoderm specification after 90 hours, PGC-LC induction after 100 hours. Window of brachyury expression indicated by bar below timeline.

Figure 5. Graphical scheme of peri-implantation mouse development going from pre- implantation on left hand side to post-implantation on right hand side. Red, epiblast (EPI) (A); Dark blue, polar trophectoderm (TE)/extraembryonic ectoderm (ExE) (D); Light blue, Mural TE (E); Green, primitive endoderm (PE) / visceral endoderm (VE) (C). Yellow, parietal endoderm (paE) (B). (i) Apolar EPI (ii) Polarised EPI lumenogenesis; (iii) joined cavity.

Figure 6. Frames from a time-lapse sequence showing epiblast-like cavitation and

morphogenesis in the TES-embryo, which occurs after fusion with the TS-compartment. Time in matrigel increasing in time from left-hand side to right-hand side of figure. From left hand side: 1^st Panel: 70h; 2^nd Panel: 72h; and 3^rd Panel: 77h. Obj=20x, Zoom=2.5, Timestep= 30 minutes. Scale bar = 20pm. CAG:GFP reporter ES cells have green membranes (equivalent to light grey in greyscale), TS cells are unlabelled. Filled white arrowheads indicate cell clusters before cavitation. Top row: CAG:GFP Stained; Bottom row: Merged. Open double-headed arrowheads indicate cavity opening and expansion.

b. Frames from a time-lapse sequence showing epiblast-like cavitation of the ES-compartment of the TES-embryo. Time in matrigel increasing in time from left-hand side to right-hand side. From left hand side, 1^st Panel: 70h, Cells cluster; 2^nd Panel: 72h, Cavity opens; 3^rd Panel: 77h, Cavity expands. Obj= 20x, Zoom=2.5, Timestep= 30 minutes. Scale bar = 20pm. CAG:GFP reporter ES cells are labelled with green membranes (equivalent to light grey in greyscale). Row 1 : Middle Z-section of the CAG:GFP labelled ES compartment at indicated timepoints during cavitation. Scale bar = 20pm. Row 2: 2D segmentation of cell membranes in the ES- compartment to highlight cell shape at different timepoints during cavitation. Row 3: 3D rendering of the CAG-GFP labelled ES- compartment of a TES-embryo at indicated points during cavitation. Solid white arrowheads indicate cell clusters before cavitation. Open double- headed arrows indicate cavity opening and expansion

c. Frames from a brightfield acquisition of a time-lapse sequence showing cell shape changes in the TS-compartment of a TESembryo during its development. Panel (i), Smooth; Panels (ii) and (iii), Cells change shape. Obj= 20x, Zoom=2.5, Timestep= 30 minutes. Scale bar = 20pm. Figure 7. A TS cell-aggregate grown for 84 hours in co-culture with ES cells, but not in contact with them. Stained to reveal: Panel 1 : green; Cdx2 (equivalent to light grey in greyscale), blue, DAPI (equivalent to dark grey in greyscale); red, aPKC (equivalent to mid grey in greyscale). Panel 2: green, F-actin (equivalent to light grey in greyscale), blue, DAPI (equivalent to dark grey in greyscale); red, aPKC (equivalent to mid grey in greyscale). White arrowheads indicate cavities within the structure. Inset shows a small cavity opening where aPKC and F-actin are enriched. Scale bar= 30pm. n=20 (2 separate experiments), b. A TS cell aggregate grown alone in Matrigel for 84 hours, and stained to reveal green, Cdx2 (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); red, aPKC (equivalent to mid grey in greyscale). No cavities or polarised localisation of aPKC could be detected. n=20, (2 separate experiments). Scale bar= 30pm. Figure 8. a. A TES-embryo after 96 hours of development with a continuous cavity, stained to reveal: red, aPKC (equivalent to dark grey in greyscale); green, PCX (equivalent to light grey in greyscale). Scale Bar= 20pm. White dotted lines highlight the cavity and white arrows indicate PCX lining the lumen. n=30 (3 separate experiments), b. Lumen area at middle Z (x-axis: Timepoint, scale from 60 to 100 hours in 10 hour increments; y-axis: Area (pm²), scale from 0 to 10000 in 2000 μιη² increments. Quantification showing the mean estimated lumen area of TES-embryos (measured at middle Z-plane) at sequential time points in their development. Lumen area increases with time in culture. n=30 per group, (3 separate experiments).

Figure 9. TES-embryos cultured presence of SB431542 for 96 hours and stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); green, PCX (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); gray, P-SMAD2. Scale bar= 20μιη. Adjacent panel: quantification of PCX intensity (x-axis, Distance (pm) from 0 to 100; y- axis, Pixel intensity from 0 to 300)) over the part indicated by dotted lines across the ES and TS compartments in TES-embryos. Distribution of PCX intensity indicates that ES- compartment cavitation is unaffected in the absence of Nodal/activin signalling. n=10 (2 separate experiments), b. A TES-embryo made from Nodal -/- ES cells and allowed to develop for 96 hours and stained to reveal: red, Oct4 (equivalent to mid grey in greyscale); green, PCX (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); gray, P-SMAD2 (equivalent to dark grey in greyscale). Scale bar= 20pm. Adjacent panel:

quantification of PCX intensity (x-axis, Distance (pm) from 0 to 100; y-axis, Pixel intensity from 0 to 120 for TS, from 0 to 300 for ES)) over the part indicated by dotted lines across the ES and TS compartments. PCX intensity distributions indicate that ES compartment cavitation is unaffected in the absence of Nodal/Activin signalling. n=10 (2 separate experiments), c. A TES-embryo cultured in control conditions for 96 hours and stained to reveal: red, Oct4 (equivalent to median grey in greyscale); green, PCX (equivalent to light grey in greyscale); blue, DAPI (equivalent to dark grey in greyscale); gray, P-SMAD2 (equivalent to dark grey in greyscale). Scale bar=20pm. Adjacent panel: quantification of PCX intensity (x-axis, Distance (μηι) from 0 to 60 for TS, from 0 to 80 for ES; y-axis, Pixel intensity from 0 to 150 for TS, from 0 to 250 for ES) over the part indicated by dotted lines across the ES-and TS- compartments. PCX intensity distributions indicate that ES compartment cavitation is unaffected in the absence of Nodal/Activin signalling. n=10 (2 separate experiments), d. Snapshots of living TES-embryos comprising Nodal -/- ES cells and TS_EGFP TS cells after 72 hours (left hand side) and 96 hours (right hand side). Colouring for antibody (TS-EGFP) light grey. Scale bar= 20μπι. e. A TS cell aggregate cultured in the presence of Activin A (50ng/ml) for 72 hours, stained to reveal: Panel V. green; Cdx2 (equivalent to light grey in greyscale), blue, DAPI (equivalent to dark grey in greyscale); red, aPKC (equivalent to mid grey in greyscale). Panel 2: green, F-actin (equivalent to light grey in greyscale), blue, DAPI (equivalent to dark grey in greyscale); red, aPKC (equivalent to mid grey in greyscale). Inset shows a small cavity opening where aPKC and F-actin are enriched. Scale bar = 20μίτι. n= 20 (3 separate experiments).

Figure 10. T/Brachyury expression in the primitive streak region of a gastrula stage embryo, compared to a similar region of the TES-embryo, stained to reveal: blue, DAPI (Equivalent to dark grey in greyscale), green, Brachyury/ T:GFP (Equivalent to light grey in greyscale). Upper row: E6.5 Embryo and Lower row: TES Embryo. Zooms highlight the Texpressing region in the TES-embryo, and in the natural embryo. Scale bar=20 m. Images are maximum projections, b. Mean ratio of mesdendodermal area/total embryonic area. From left hand side, 1^st column: E6.5 Embryo; 2^nd Column: TES Embryo. Y-Axis: Mean ratio of ES-compartment

area/mesodermal area from 0 to 0.35 in increments of 0.05. Quantification showing TES- embryos and E6.5 embryos are similar in the size of the region of mesoderm. (n= 10 per group, Mean ratio of primitive streak area/ total epiblast area of E6.5 embryo=0.219. Mean ratio of ES- compartment area/mesodermal area TES-embryo after 96h of development^ 0.279. Student' s t test of means, not significant), c. A TES-embryo stained to reveal T:GFP expression overlaid with endogenous T/Brachyury expression. (Panel 1 : Red, Oct4 (Equivalent to mid grey in greyscale); green, T:GFP (Equivalent to light grey in greyscale); blue DAPI (Equivalent to dark grey in greyscale)). White boxes highlight T:GFP overlaid with T/Brachyury (Panel 2: Red, T:GFP (Equivalent to mid grey in greyscale); Green, Brachyury (Equivalent to light grey in greyscale)).

Figure 11. a. Percentile quantification showing the mean number of TES-embryos which developed detectable mesoderm was significantly reduced in the presence of DKK1 compared with control conditions (Student's t-test of means,

P<0.001 , n=400, pooled data from 4 separate experiments). From left hand side, 1^st Column: Control; 2^nd Column: DKK1. Y-axis: Mean number of structures with detectable mesoderm expressed as a percentage, scale from 0 to 100 in increments of 20. b. A TES-embryo (row 1), in vitro cultured egg-cylinder (row 2) and an embryo recovered from the mother at embryonic day 5.5 (row 3) stained to reveal P-SMAD1 , white; Oct4, red (Equivalent to mid grey in greyscale); DAPI blue (Equivalent to dark grey in greyscale). Scale bar= 20μπι. EXAMPLES

Example 1 : Self-assembly of embryonic and trophoblast stem cells recapitulates embryo architecture in vitro Morphogenetic transformations during implantation development are critical for mammalian embryo patterning and yet are poorly understood. The first such major transformation is the generation of the pro-amniotic cavity that precedes symmetry breaking to establish the anterior- posterior axis leading to formation of the germ layers and primordial germ cells. These events are achieved through signalling between the pluripotent epiblast, that will generate the foetus, and its enveloping extra-embryonic tissues, trophoblast and primitive endoderm, progenitors of the placenta and yolk sac (refs1-6). Here, we establish an in vitro model system that recapitulates these spatio-temporal events of embryogenesis. We describe three-dimensional culture conditions that enable mouse embryonic stem (ES) cells and trophoblast stem (TS) cells to organise themselves into unified embryo-like structures. These structures undertake morphogenetic events leading to lumenogenesis, symmetry-breaking and mesoderm specification in an ES-derived embryonic compartment bordering a TS-derived extra-embryonic compartment, just as in developing embryos. This symmetry breaking is linked to unification of ES- and TS-cavities and canonical Wnt signalling. When these embryo-like structures develop further, they spontaneously generate primordial germ cells on the border between ES and TS compartments in response to BMP signalling, as in the developing embryo.

Together, these findings demonstrate a remarkable inherent ability of two stem cell types to assemble themselves and communicate to specify mesoderm and germline. The accuracy with which these structures recapitulate the spatial temporal events of embryogenesis in vivo made us term them Trophoblast and Embryonic Stem cell (TES) embryos.

Organoids derived from progenitor cells present an invaluable system to recapitulate many events in organ formation in vitro (refs 7-1 1). Structures derived from ES cells can also develop to initiate the polarised expression of genes associated with gastrulation, but they do not replicate the spatial events that culminate in symmetry breaking and positioning of germ layers (refs 12- 15). We hypothesised that this might be because such ES cell-derived structures lack the organization of the embryo with its distinct embryonic and extra-embryonic compartments and the complex signalling interactions between them (Figure 5). To test this hypothesis, we sought to establish an in vitro model system more akin to the development of the embryo. We postulated that by allowing ES cells to interact with extra-embryonic stem cells, the cell-cell communication events necessary for correct spatial development would take place. We began by establishing a three-dimensional (3D) culture environment and medium that would support development of ES cells side-by-side with extra-embryonic TS cells. We plated single-cell suspensions of ES cells with small clumps of TS cells in Matrigel®, which we allowed to solidify in drops before culture in a medium we developed for this purpose (Fig. 1a; Methods). We found that ES and TS cells proliferate in this environment to form first two separate aggregates that become joined between two and three days of culture. On the fourth day, these ES and TS cells organize themselves into a single continuous cell layer with an elongated morphology of approximately 100um wide x 200um long (Fig. 1 b, 1c). Examination of the expression of key embryonic and extra-embryonic molecular markers (Oct4 and Eomes) revealed discrete ES- and TS-derived compartments within a single structure that has a central fluid-filled cavity (Fig.1b). These ES-TS structures were remarkably similar to early post-implantation embryos developing in vivo and in vitro for two days after blastocyst plating (Fig. 1 b-e, Supplementary movie 1) as assessed by several criteria including: morphology, size, cell numbers and expression of lineage markers (Fig. 1 b-g). These results suggest that in this culture system ES and TS cells communicate sufficiently well to self- assemble into structures that mimic early postimplantation embryos.

One of the first critical events in post-implantation development is formation of the proamniotic cavity that begins in the embryonic compartment of the embryo. It has recently been shown that this process is initiated by an apical cellular constriction followed by lumenogenesis (ref 16) and is occurring in the epiblast of both the mouse (ref 16) and human embryo (ref 17). We therefore wished to examine the cellular events leading to cavitation in TES-embryos. We found that a cavity emerged during a 30-hour time window between three and five days after the initial plating of ES and TS cells (Fig. 1 h). When we followed this process in living TES-embryos, we found that cavitation was initiated in the ES-derived embryonic compartment, as in vivo (ref 16) (Fig. 6). The onset of cavitation was preceded by elongated cells clustering around one or more meeting points (Figure 6). Subsequently cells rearranged to form a cavity in the centre of their ES-compartment (Fig. 6). This cavity expanded over the next few hours during which time cells in the TS-derived compartment started to change their shape leading to the formation of multiple small cavities (Fig. 6). The ES-cavity and the TS-cavity then finally united (Fig. 1h). This physical process and its timing are very similar to cavitation in the embryo where the epiblast cavity first forms, subsequently expands, and then unites with growing extra-embryonic cavities (ref 16). It is currently unknown how the cavity forms in the extra-embryonic compartment and how it becomes unified with the embryonic cavity (ref 18). We therefore wished to use the TES-embryos to gain insight into this question. To this end, we generated TESembryos and examined the sequence of events leading to cavitation in the extraembryonic (TS)-compartment. We analysed the localisation of aPKC and the sialoglycoprotein, Podocalyxin (PCX), as markers for apical polarisation and lumenogenesis respectively. After 72 hours of culture, a cavity was present only in the ES-compartment and aPKC was enriched on the cavity facing sides of cells, resembling its distribution in the natural embryo (Fig. 2a). PCX staining was maximal in the ES-compartment along the sides of cells lining the lumen, whereas in the TS-compartment, where a cavity had yet to form, it associated with all cell membranes but had no clear pattern (Fig. 2b). By 84 hours, TES-embryos had developed one cavity in the ES-compartment and one or more additional cavities within the TS-compartment (Fig. 2b). PCX was now also enriched at the lining of each of these cavities (Fig. 2b).

At this stage, the embryonic and extra-embryonic cavities were still distinct. To determine whether cavitation was an inherent property of TS cell aggregates, we cultured TS cells in 3D without ES cells. We were unable to detect any cavity in TS cells cultured in this way (Figure 7) suggesting that cavitation in TS compartments of TES-embryos occurs in response to their association with ES cells.

By 96 hours, all TES-embryos had developed a single, continuous cavity extending between the ES- and TS-compartments (Fig. 2b, Fig. 8) that had expanded continually in total volume since 72 hours (Fig 8). Cells lining both compartments had now become arranged to resemble a continuous epithelium and PCX could be detected where the cells faced the common lumen (Fig. 2b, Fig 8).

To gain insight into how the ES and TS cavities become joined, we focused on the boundary between embryonic and extra-embryonic compartments and analysed progressive cell shape changes. At 72 hours, cells in the ES-compartment were wedge-shaped and apically constricted, whereas boundary cells in the TS compartment exhibited a rounded morphology, as revealed by E-cadherin staining (Fig. 2c). The cell aspect ratio (width/length) showed that TS cells at the boundary between the two compartments differed significantly in shape from cells in the ES compartment at this stage (Fig. 2d) prior to the formation of a continuous cavity. We also analysed the distribution of laminin, which is present in the basal membrane at the boundary between embryonic and extra-embryonic compartments in the early post-implantation embryo (refs 16, 19). When the cavities between the ES- and TS compartments were fully expanded, no laminin could be detected between the two compartments (Fig. 3a). However, at intermediate time-points during cavity merging, we could detect a discontinuous layer of ES cells with associated residual laminin between the two compartments. Some of the ES cells in this layer appeared to be enveloped in the TS-compartment (Fig. 3a, top, white arrows). Residual laminin at this border between compartments showed weaker, and in places broken staining, sometimes extending into the TS-compartment (Fig. 3a, bottom, yellow arrows). In contrast to the TES- embryo, we could not detect any extension of laminin staining into one or other side when two ES-compartments underwent fusion (Fig. 3b-c). Thus, the preferential displacement of laminin towards the extra-embryonic part during cavitation is characteristic of junction between ES and TS-compartments. We also observed characteristic floret-like arrangements of both ES and TS cells at the ES-TS boundary, where there was no residual laminin between compartments and at the point at which cavities were merging (Fig. 3d-e). Interestingly, only one such chimeric cell floret could be detected per TES-embryo (n=15). These chimeric cell arrangements might facilitate tissue morphogenesis via cell rearrangements during ES-TS cavity fusion, as suggested in other systems of epithelial morphogenesis (ref 20). Together, our observations indicate that the formation of a single cavity requires the disintegration of the extracellular matrix border between ES and TS-compartments and associated cell rearrangement. Given that basement membrane breakdown also has to occur between the embryonic and extra-embryonic compartments during cavity formation in vivo, our results suggest that the TES-embryos can mimic this process, at least to some extent.

Because TS cells did not organise into structures that cavitate when cultured alone (Fig. 7), we hypothesised that cells of the ES-compartment might secrete a signal to induce cavitation in the TS-compartment. One such signalling candidate is Nodal, known to be secreted by ES cells in culture (ref 21), and which is essential in vivo (ref 22). Furthermore, TGF-beta signalling is required for trophoblast self-renewal (ref 23).

To test the potential role of Nodal signalling in this process, we generated TES-embryos in the presence of the Activin/TGF-beta receptor inhibitor, SB43154224, added to the culture at 48 hours ( Fig. 9). We verified the inhibition of the Nodal/Activin pathway by p-SMAD2 staining and analysed the intensity and distribution of PCX staining after 96 hours of development (Fig. 9). In contrast to control TES-embryos, we detected no cavities in the TS-compartment of SB431542- treated TES-embryos (Fig. 9 a,b). In contrast, the cavitation of the ES-compartment was not affected, but we noted a reduction in Oct4 staining. To validate this result, we also generated TES-embryos using tamoxifen inducible-knockout Nodal ES cells (ref 25). We fixed these Nodal- /-ES TES-embryos and stained them to reveal PCX after 96 hours of development (Fig. 9b). We found that Nodal-/- TES embryos failed to cavitate in their TS-compartments and, as in the presence of SB431542, a cavity was still present in the ES-compartment (Fig. 9 b,c). As further confirmation, we monitored development of living TES-embryos generated from GFP-expressing TS cells (ref 26) and Nodal-/- ES cells.

Once again, the resulting TES-embryos had clearly disorganised extra-embryonic compartments that failed to cavitate, even after 96 hours (Fig. 9d). In support of these data, we found that culture of TS cells alone in 3D in the presence of exogenous Activin A lead to cavity formation (Fig. 9e). These results indicate that TES-embryos mimic post-implantation pro-amniotic cavity formation and suggest a potential novel role for Nodal/Activin signalling in inducing cavitation in the extraembryonic compartment. Once the pro-amniotic cavity has formed, the next developmental event is the breaking of the embryo' s symmetry to initiate gastrulation. In normal development, a second extra-embryonic tissue, the visceral endoderm, participates in this process by providing the anterior visceral endoderm (AVE) (ref 3) signalling centre. To determine whether TES-embryos, that lack visceral endoderm, could undergo a symmetry breaking event, we allowed them to develop for up to 120 hours when we determined if they express the T/Brachyury, which marks mesoderm formation as the primitive streak forms (refs 4,27). To this end, we utilised ES cells that express a T:GFP reporter (ref 28) . At 96 hours of development, the TES-embryos expressed T:GFP in a similar pattern to natural embryos (Fig. 3f) with the number of cells present in the embryonic compartment at the onset of T/Brachyury expression ranging between 80.5 cells (n=10 TES- embryos). The proportion of TES-embryos (65.6%, 42/64) expressing T:GFP at this time was significantly higher than in structures derived solely from ES cells (27.2%, 12/44) (Fig. 3g). Strikingly, T:GFP expression became confined to a discrete domain on one side of the ES- compartment extending from the boundary with the TS compartment.

This domain of T/Brachyury expression occupied a similar proportion of the ES-compartment of TES-embryos (Fig. 3h,i) as the equivalent domain in the epiblast of embryos in vivo (Fig. 10 a, b). This asymmetric T:GFP expression domain was seen in a significantly higher proportion of TES-embryos (43%, 43/100) compared to T:GFP expression in structures derived solely from ES cells developing for the same period of 120 hours (14%, 14/100) (Fig. 3j). We confirmed that the T:GFP expression overlaid with endogenous Brachyury expression (Fig. 10c). Thus, the regionalised induction of mesoderm is promoted when ES cells develop in the presence of TS cells. Interestingly, the T:GFP domain was located not only at the ES-TS boundary, but it was also always opposite an area of residual laminin at the junction of the two compartments in all TES-embryos examined that had both features (Fig. 3k). These results suggest that the TS compartment in TES-embryos promotes T/Brachyury expression in a manner that mimics the role of the extra-embryonic ectoderm in the embryo, leading to more efficient, self-organised symmetry breaking.

In natural development, Wnt3 expression precedes the induction of T/Brachyury expression and mesoderm specification (ref 5). We therefore hypothesised that Wnt signalling might become active in TES-embryos leading to mesodermal specification.

To test this, we generated TES-embryos using H2B-GFP:Tcf/LEF reporter ES cells (ref 29) to monitor the activity of Wnt signalling. Embryos were fixed at different time points of their development and counter-stained to examine T/Brachyury expression. After 90 hours, we detected H2B-GFP:Tcf/LEF expression at the ES-TS boundary, but T/Brachyury was not expressed at that time (Fig. 4a). However, after 96 hours H2BGFP: Tcf/LEF expression at the ES-TS boundary co-localised with T/Brachyury expression (Fig. 4a). This domain of T/Brachyury and H2B-GFP:Tcf/LEF-positive cells increased in size and cell number over the next 12 hours (Fig.4a, b). These results suggest that canonical Wnt signalling activity precedes mesodermal specification in TES-embryos. To determine whether Wnt signalling is required for T/Brachyury expression, we allowed TES-embryos to develop in the presence of the canonical Wnt antagonist DKK1 (ref 30), added after 48 hours of culture. In contrast to controls, the proportion of TES- embryos specifying mesoderm was significantly reduced after 96 hours (Fig. 4c, Fig. 11a). These results indicate that the Wnt signalling pathway is crucial to specify mesoderm in TES-embryos as is the case in vivo.

The next major step in post-implantation development is the specification of primordial germ cells (PGCs). In vivo, PGCs are specified at the proximal end of the primitive streak, at the boundary between embryonic and extra-embryonic compartments (ref 6). This requires Prdm14, Blimpl (Prdml) and Tfap2c (AP2y) that act in synergy to specify germ-cell fate (ref 31). Since both Prdm14 and Blimpl are direct targets of T/Brachyury, we hypothesised that our TES-embryos might also be able to specify PGC-like cells. To test this hypothesis, we fixed TES-embryos after 120 hours of development and stained them to reveal Tfap2c. We detected Tfap2c-Oct4 double- positive cells at the ES-compartment boundary in the domain where T/Brachyury was expressed (Fig. 4d).

This is a remarkably similar site to the location of PGCs specification in vivo. To further investigate PGC-like cell specification, we generated TES-embryos using ES cells that express GFP-tagged Stella (Stella:GFP) as a PGC reporter (ref 32). In accord with our earlier result, we observed expression of the GFP-tagged Stella after 120 hours in culture (Fig. 4e). Strikingly, on average 5 Stella:GFP cells could be detected at the boundary between the ES and TS- compartments (Fig. 4e). When compared to ES cells cultured alone in Matrigel, the proportion of Stella:GFP-positive structures was significantly increased (Fig. 4f), suggesting that the presence of the TS compartment promotes PGC-like cell specification. Since specification of PGCs is induced by BMP signalling from the extra-embryonic compartment (ref 6), we stained our TES- embryos for p-SMAD1 to mark BMP signalling pathway activity (Fig. 4e, 4g). We found that its expression in TES-embryos was as in embryos developing in vitro and in vivo, indicating that they were competent to specify PGC-like cells (Fig. 4e, 4g, top; Fig. 11b). To determine whether BMP signalling was required for PGC-like cell specification in the TES-embryos, we cultured them in the presence of the BMP antagonist, Noggin33, which abrogates BMP signalling as revealed by staining for p-SMAD1 (Fig. 4g, bottom). In contrast to controls, in the presence of BMP inhibitor TES-embryos failed to express Stella:GFP. These results indicate that after the specification of mesoderm cells, TES-embryos develop to specify germ cells as normally developing embryos.

Previous attempts to replicate events of mouse embryogenesis using ES cells alone succeeded in achieving polarised expression of mesoderm markers upon activation of Wnt signalling (refs 12,14). However, the resulting structures did not recapitulate embryo morphology. Perhaps because in normal development, the extra-embryonic tissues are key to pattern the epiblast by inducing the anterior-posterior axis and germ layers specification. This is thought to require the AVE to restrict signalling from the extraembryonic ectoderm to break embryo symmetry (refs 34- 35). However, we show here that morphogenetic transformations leading to symmetry breaking and positional specification of mesoderm can occur without the AVE, and in fact without any visceral endoderm tissue. It will be of future interest to determine exactly how addition of visceral endoderm might replicate additional aspects of development. As it is, the system we have developed allowing self-assembly of only ES and TS cells is sufficient to recapitulate early post- implantation embryogenesis including: (1) the architectural changes leading to cavity formation in the embryonic (ES) and the extra-embryonic (TS) compartments before their unification into the equivalent of the embryo' s proamniotic cavity; (2) the symmetry-breaking events leading to localised expression of genes required to form mesoderm next to the boundary between embryonic and extraembryonic compartments, as in the embryo; and (3) the specification of PGC-like cells in a position equivalent to the site of their development in the embryo. Additionally, we show that these events utilise the same specific signalling pathways and remarkably follow a similar spatial and temporal profile as in natural embryos. Thus, our approach of using two types of stem cell significantly extends models using ES cells alone (refs 12,14).

In contrast to these earlier models, the system we describe permits cell-cell interactions at the boundaries of well-established compartments that replicate embryo architecture and enable the timing and positioning of signalling events to refine the mesoderm and germline expression domains.

In conclusion, our results indicate that spatially regulated morphogenesis and signalling during implantation and early post-implantation development can be recapitulated by self-assembly of just two cell types: ES cells and TS cells. This model represents a powerful platform to study interactions between the embryonic and extraembryonic compartments, and to identify the signalling pathways that lead to symmetry-breaking, and mesoderm and germline specification in a spatial context that mimics the natural embryo.

References

1. Stern, CD. Gastrulation: from cells to embryo. Ed. New York: Cold Spring

Harbor Laboratory (2004).

2. Tarn, P. P. L, & Loebel, D. F. Gene function in mouse embryogenesis: get set

for gastrulation. Nature Reviews. Genetics, 8(5), 368-81 (2007).

3. Arnold, S. J., & Robertson, E. J. Making a commitment: cell lineage allocation

and axis patterning in the early mouse embryo. Nature Reviews. Molecular Cell

Biology, 10(2), 91 -103 (2009). 4. Wilkinson, D. Expression of the mouse T gene and its role in mesoderm formation. Letters To Nature, 346, 183-187 (1990).

5. Rivera-Perez, J. A, & Magnuson, T. Primitive streak formation in mice is preceded by localized activation of Brachyury and Wnt3. Developmental

Biology, 288(2), 363-71 (2005).

6. Lawson, K. A., et al. Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes & Development, 13, 424-436 (1999).

7. Sato, T., et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature, 459(7244), 262-5 (2009).

8. Eiraku, M., et al. Self-organizing optic-cup morphogenesis in three-dimensional culture. Nature, 472(7341), 51 -6 (2011).

9. Lancaster, M. A. et al. Cerebral organoids model human brain development and microcephaly. Nature, 29(2), 185- 185 (2013).

10. Xia, Y., et al. The generation of kidney organoids by differentiation of human pluripotent cells to ureteric bud progenitor-like cells. Nature Protocols, 9(11),

2693-704 (2014).

11. Takasato, M., et al. Directing human embryonic stem cell differentiation towards a renal lineage generates a self-organizing kidney. Nature Cell Biology, 16(1), 118-26 (2014).

12. ten Berge, D.,et al. Wnt signalling mediates self-organization and axis formation in embryoid bodies. Cell Stem Cell, 3(5), 508- 18 (2008).

13. Fuchs, C, et al. Self-organization phenomena in embryonic stem cell-derived embryoid bodies: axis formation and breaking of symmetry during

cardiomyogenesis. Cells, Tissues, Organs, 195(5), 377-91 (2012).

14. van den Brink, S. C, et al. Symmetry breaking, germ layer specification and axial organisation in aggregates of mouse embryonic stem cells. Development (Cambridge, England), 141 (22), 4231 -42 (2014).

15. Warmflash, A., Sorre, B., Etoc, F., Siggia, E. D., & Brivanlou, A. H. A method to recapitulate early embryonic spatial patterning in human embryonic stem cells. Nature Methods, 11 (8) (2014).

16. Bedzhov, I., & Zernicka-Goetz, M. Self-Organizing Properties of Mouse Pluripotent Cells Initiate Morphogenesis upon Implantation. Cell, 156(5), 1032- 44 (2014).

17. Shahbazi, M. N., Jedrusik, A., Vuoristo, S., et al. Self-organization of the human embryo in the absence of maternal tissues. Nature Cell Biology 18(6), 700-8 (2016).

18. Bedzhov, I., Graham, S. J. L, Leung, C. Y., Zernicka-Goetz, M.

Developmental plasticity, cell fate specification and morphogenesis in the early mouse embryo. Philosophical Transactions of the Royal Society (2014). 19. Miner, J. H., Li, C, Mudd, J. L, Go, G., & Sutherland, A. E. Compositional and structural requirements for laminin and basement membranes during mouse embryo implantation and gastrulation. Development (Cambridge, England),

131 (10), 2247-2256 (2004).

20. Villasenor, A., Chong, D. C, Henkemeyer, M., & Cleaver, O. Epithelial dynamics of pancreatic branching morphogenesis. Development (Cambridge, England), 137(24), 4295-305 (2010).

21. Watabe, T., & Miyazono, K. Roles of TGF-beta family signalling in stem cell renewal and differentiation. Cell Research, 19(1), 103- 115 (2009).

22. Brennan, J., et al. Nodal signalling in the epiblast patterns the early mouse embryo. Nature, 411 (6840), 965-9 (2001).

23. Guzman-Ayala, M., Ben-Haim, N., Beck, S., Constam, D. B. Nodal protein processing and fibroblast growth factor 4 synergize to maintain a trophoblast stem cell microenvironment. Proceedings of the National Academy of Sciences of the United States of America, 101 (44), 15656- 15660 (2004).

24. Inman, G. J., et al. SB-431542 is a potent and specific inhibitor of transforming growth factor-beta superfamily type I activin receptor-like kinase (ALK)

receptors ALK4, ALK5, and ALK7. Molecular Pharmacology, 62(1), 65-74

(2002).

25. Wu, Q., et al. Nodal/activin signalling promotes male germ cell fate and suppresses female programming in somatic cells. Development (Cambridge,

England), 140(2), 291 -300 (2013).

26. Tanaka, S. Promotion of Trophoblast Stem Cell Proliferation by FGF4.

Science, 282(5396), 2072-2075 (1998).

27. Herrmann, B. G. Expression pattern of the Brachyury gene in whole-mount TWis/TWis mutant embryos. Development (Cambridge, England), 113(3), 913-

917 (1991).

28. Fehling, H. J. et al. Tracking mesoderm induction and its specification to the hemangioblast during embryonic stem cell differentiation. Development,

130(17), 4217-4227 (2003).

29. Ferrer-Vaquer, A., et al. A sensitive and bright single-cell resolution live imaging reporter of Wnt/B-catenin signaling in the mouse. BMC Developmental Biology,

10, 121 (2010).

30. Niida, A., et al. DKK1 , a negative regulator of Wnt signaling, is a target of the beta-catenin/TCF pathway. Oncogene, 23(52), 8520-8526 (2004).

31. Gunesdogan, U., Magnusdottir, E., Surani, M. A. Primoridal germ cell

specification : a context-dependent cellular differentiation event, (October), 2- 11 (2014).

32. Payer, B., et al. Generation of stella-GFP transgenic mice: a novel tool to study germ cell development. Genesis (New York, N. Y. : 2000), 44(2), 75-83 (2006).

33. Zimmerman, L. B., De Jesus-Escobar, J. M. & Harland, R. M. The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4.

Ce// 86, 599-606 (1996).

34. Thomas, P. Q., Brown, A, & Beddington, R. S. Hex: a homeobox gene

revealing peri-implantation asymmetry in the mouse embryo and an early

transient marker of endothelial cell precursors. Development (Cambridge,

England), 125, 85 94 (1998).

35. Yamamoto, M., et al. Nodal antagonists regulate formation of the

anteroposterior axis of the mouse embryo. Nature, 428(6981) 387-392, (2004).

36. Bedzhov, I., Leung, C. Y., Bialecka, M., & Zernicka-Goetz, M. In vitro culture of mouse blastocysts beyond the implantation stages. Nature Protocols, 9(12),

2732-2739 (2014).

37. Tanaka, S. Derivation and culture of mouse trophoblast stem cells in vitro.

Methods in Molecular Biology (Clifton, N.J.), 329, 35-44 (2006).

38. Ying, Q.-L, Stavridis, M., Griffiths, D., Li, M., & Smith, A. Conversion of

embryonic stem cells into neuroectodermal precursors in adherent

monoculture. Nature Biotechnology, 21 (2), 183-6 (2003).

39. Pollard, S. M., Benchoua, A., & Lowell, S. Neural Stem Cells, Neurons, and

Glia. Methods in Enzymology, 418(06), 151 -169 (2006).

40. Faure, E., et al. A workflow to process 3D+time microscopy images of

developing organisms and reconstruct their cell lineage. Nature

Communications, 78674 (2016).

Extended experimental methods:

Embryo recovery and culture: 6-week old F1 female mice were naturally mated and sacrificed at midday after 5 days post-coitum. The uterus was recovered and embryos were manually dissected from deciduae in M2 medium using fine forceps. Embryos were cultured as described in Bedzhov et al, 2014 (ref 36). Blastocysts were recovered from the mother at 4.5 days post coitum by uterine flushing with M2 medium. Recovered blastocysts had their mural trophectoderm manually dissected away, before plating in ibiTreat microscopy plastic m-plates (Ibidi) and cultured in IVC1 medium (Advanced DMEM/F12 supplemented with 20% heat-inactivated FBS, 2 mM l-glutamine, penicillin-streptomycin ((25 pg/ml)), 1 ^χ ITS-X ((10 mg per litre insulin, 5.5 mg per litre transferrin, 0.0067 mg per litre sodium selenite and 2 mg per litre ethanolamine)), 8 nM β -estradiol, 200 ng/ml progesterone and 25 μΜ N-acetyl-l-cysteine). After 48 hours in culture, the medium was changed to IVC2 medium (Advanced DMEM/F12

supplemented with 30% KSR, 2 mM l-glutamine, penicillin-streptomycin ((25 Mg/ml)), 1 ITS-X ((10 mg per litre insulin, 5.5 mg per litre transferrin, 0.0067 mg per litre sodium selenite and 2 mg per litre ethanolamine)), 8 nM β -estradiol, 200 ng/ml progesterone and 25 μΜ N-acetyl-l-cysteine).

Cell culture: ES cells were cultured at 37^°C and 5% C02 on gelatinized tissue-culture grade plates and passaged once they reached confluency. Cells were cultured in DMEM with 15% FBS, 2mM L-glutamine, 0.1 mM 2-ME, 0.1mM NEAA, 1mM sodium pyruvate, and 1 % penicillin-streptomycin) supplemented with PD0325901 (1uM), CHIR99021 (3uM) (2i) and leukaemia inhibitory factor (0.1 mM, LIF).

TS cells were cultured at 37 ^°C and 5% C02, in RPMI 1640 (Sigma) with 20% FBS, 2mM L-glutamine, 0.1 mM 2-ME, 1 mM sodium pyruvate, and 1% penicillin streptomycin, plus FGF4 (Peprotech) and heparin (Sigma) in the presence of

inactivated DR4 MEFs37. Cells were passaged when wells became 80% confluent. Cell Lines used in the study: All experiments were performed using E14 or 129 mouse ES cells, Inducible Nodal knockout ES cells25, T:GFP ES cells28, H2BGFP: Tcf/LEF ES cells29, Stella:GFP32 ES cells, wild-type TS cells, and TS_EGFP26 cells.

'3D embedded' culture: ES cell colonies were dissociated to single cells by

incubation with 0.05% trypsin-EDTA at 37 ^°C. Cells were pelleted by centrifugation for 5 min/1 ,000 rpm, washed with PBS, and re-pelleted. This was repeated twice, then ES cell and TS cell suspensions were mixed and re-pelleted. The pellet was resuspended in Matrigel (BD, 356230). The cell suspension was plated on ibiTreat

microscopy plastic m-plates (Ibidi) and incubated at 37 ^°C until the Matrigel solidified. The plate was then filled with pre-warmed medium. Cells were cultured at 37 ^°C and 5% C02. TES-embryo culture medium used was: 50% RPMI, 25% DMEM F-12 and 25% Neurobasal A, (supplemented with 10% FBS, 2mM L-glutamine, 0.1 mM 2ME, 0.5mM sodium pyruvate, 0.25x N2 supplement, 0.5x B27 supplement, FGF4 (12.5 ng/ml) and heparin (Sigma) 50mg/ml. Immunofluorescence: Cells/ embryos were fixed with 4% paraformaldehyde for 15 mins at room-temperature, then rinsed in PBS. Permeabilization was performed with 0.3% Triton-X-100, 0.1 % Glycin in PBS for 10 minutes at room-temperature. Primary antibody incubation was performed overnight at 4^°C. The following day, cells were washed, then incubated overnight in secondary antibody at 4^°C. DAPI in PBS (5mg/ml) was added prior to confocal imaging. For antibodies used, see Table 1.

Table 1

Imaging, Processing and Analysis: All images and time-lapse movies were acquired using a Leica SP5 confocal microscope, and all analyses were carried out using opensource image analysis software ' Fiji ' , or 'Bioemergences' software (ref 40). Where appropriate, images were de-noised using the 'Smart Denoise' plugin written for ' Fiji ' (Dr. Richard Butler, Gurdon Institute, Cambridge).

Statistics: All statistical tests were performed on GraphPad Prism 7.0 software. Data were checked for normal distribution and equal variances before each parametric

statistical test was performed. If data did not conform to a normal distribution, data

were normalised using a square-root transformation. Error bars represent standard

error of the mean in all cases, unless otherwise specified. The number of independent experiments performed in each analysis is indicated in the appropriate figure legend.

Measurement of laminin displacement angle: The angle of displacement of residual laminin was calculated using images of TES-embryos during cavity fusion. A vertical line was drawn from ES compartment to TS compartment, and a line across the

boundary was drawn perpendicular. Laminin adjacent to the boundary would therefore have an angle of 90^°. Θ is equal to the angle of the laminin extending into the TS

compartment (θ< 90^°).

Antibody (species) Vendor, Catalogue Number, Dilution

Oct 3/4 (mouse) Santa cruz sc-5279 1 :200

Tbr2/Eomes (rabbit) abeam ab23345 1 :400

aPKC (rabbit) Santa cruz sc-17781 1 :200

Podocalyxin (rat) R&D systems MAB1556 1 :400

Brachyury/T (goat) Santa cruz sc-17745 1 :50

GFP (rat) Nacalai biochemicals 04404-84 1 :2000

Tfap2c (rabbit) Santa cruz sc-8977 1 :200

Laminin (rabbit) Sigma L9393 1 :400

Cdx2 (mouse) Launch diagnostics MU392-UC 1 :200

E-cadherin (rat) Life Technologies (Thermofisher scientific) 13-1900 1 :400

Phospho-SMAD2/3 Cell signalling technologies 8685P 1 :200

Phospho-SMAD 1/5/9 Cell signalling technologies 13820P 1 :200

F-actin (Phalloidin 488) Life Technologies (Thermofisher scientific) A12379 1 : 1000

Alexa 488 (Donkey anti-rat) Life Technologies (Thermofisher scientific) A21208 1 :500

Alexa 568 (Donkey anti-mouse) Life Technologies (Thermofisher scientific) A10037 1 :500

Alexa 647 (Donkey anti-rabbit) Life Technologies (Thermofisher scientific) A31573 1 :500

Alexa 647 (Donkey anti-goat) Life Technologies (Thermofisher scientific) A21447 1 :500

While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention. All documents cited herein are expressly incorporated by reference.

Claims

1. An in vitro method of culturing a mammalian pluripotent stem cell comprising contacting a mammalian pluripotent stem (PS) cell with an extra-embryonic stem cell.

2. The method of Claim 1 , wherein the extra-embryonic stem cell is a trophoblast stem (TS) cell.

3. The method of Claim 1 , therein the extra-embryonic stem cell is a primitive-endoderm (PES) stem cell.

4. The method of any preceding claim, wherein the pluripotent stem (PS) cell is an induced pluripotent stem (iPS) cell.

5. The method of any of Claims 1 to 3, wherein the pluripotent stem (PS) cell is a mammalian embryonic stem (ES) cell.

6. The method of any preceding claim, wherein the mammalian pluripotent stem cell is a mouse or human pluripotent stem cell.

7. The method of any preceding claim, wherein the PS cells are contacted with a pluripotent stem cell-trophoblast stem cell (TPS) culture medium and a substrate which are capable of supporting growth of both the PS cells and the TS cells in culture.

8. The method of any preceding claim comprising the earlier step of maintaining the PS cells and TS cells in separate culture for at least one passage.

9. The method of any preceding claim, wherein the PS cells and TS cells are obtained from blastocysts.

10. The method of any of preceding claim wherein the PS cells are contacted with both TS cells and PES cells.

11. The method of Claim 8 or 9, wherein the PS and TS cells are removed from culture, washed separately and then combined and contacted with the substrate.

12. The method of Claim 11 , comprising the further step of incubating the combined cells in TPS medium.

13. The method of Claim 12, wherein the cells are maintained in culture for up to 5 days.

14. The method of any of Claims 7 to 13, wherein the substrate provides a three-dimensional culture environment.

15. The method of any of Claims 7 to 14, wherein the substrate comprises a matrix.

16. The method of Claim 15, wherein the matrix comprises at least one extracellular matrix (ECM) protein, or analogue thereof.

17. The method of Claim 16, wherein the ECM protein is collagen or an analogue thereof, laminin or an analogue thereof, fibronectin or an analogue thereof and/or gelatin.

18. The method of Claim 17, wherein the ECM protein is collagen or an analogue thereof and/or laminin or an analogue thereof.

19. The method of any of Claims 7 to 18, wherein the substrate comprises laminin, collagen IV, heparin sulphate proteoglycans, entactin/nidogen, and growth factors.

20. The method of Claim 19, wherein the substrate is Matrigel® Matrix.

21. The method of any of Claims 7 to 20, wherein the TPS medium comprises a basal culture medium supplemented with a non-human serum or serum substitute thereof; L-glutamine or a derivative or analogue thereof; a reducing agent; Fibroblast Growth Factor (FGF), an FGF analogue or FGF receptor agonist; insulin, an insulin analogue or insulin receptor agonist; and progesterone, a progesterone analogue or progesterone receptor agonist.

22. The method of Claim 21 , wherein the TPS medium comprises about 0.1 mg/l to about 200 mg/l insulin or insulin analogue, or about 0.05 ng/ml to about 300 ng/ml insulin receptor agonist.

23. The method of Claim 21 or 22, wherein the TPS medium comprises about 1 ng/ml to about 2 μg/ml progesterone, progesterone analogue or progesterone receptor agonist.

24. The method of any of Claims 21 to 23, wherein the reducing agent is 2-mercaptoethanol (2-ME), N-acetyl-L-cysteine or dithiothreitol.

25. The method of any of Claims 21 to 24, wherein the TPS medium further comprises sodium pyruvate.

26. The method of any of Claims 21 to 25, wherein the TPS medium further comprises nonessential amino acids (NEAA).

27. The method of Claim 26, wherein the NEAA are selected from any one or all of L-glycine, L-alanine, L-asparagine, L-aspartic acid, L-glutamic acid, L-proline and L-serine.

28. The method of any of Claims 21 to 27, wherein the non-human serum is foetal bovine serum (FBS).

29. The method of Claim 28, wherein the TPS medium comprises 10 % v/v FBS.

30. The method of any of Claims 21 to 29, wherein the basal culture medium is Dulbecco's Modified Eagle Medium (DMEM), Roswell Park Memorial Institute (RPMI) medium 1640, or Neurobasal® or Neurobasal® A or a mixture thereof.

31. The method of Claim 30, wherein the basal culture medium is a mixture of DMEM, RPMI and Neurobasal® A.

32. The method of Claim 31 , wherein the mixture is 50 % RPMI, 25 % DMEM and 25 % Neurobasal® A.

33. The method of any of Claims 21 to 32, wherein the TPS medium further comprises an antibiotic(s) and/or another antimicrobial, for example antibacterial, compound.

34. The method of Claim 33, wherein the antibiotic is penicillin and/or streptomycin.

35. The method of any of Claims 21 to 34, wherein the TPS medium further comprises heparin

36. The method of any of Claims 21 to 35, wherein the TPS medium comprises 50 % RPMI, 25 % DMEM F-12 and 25 % Neurobasal A, supplemented with 10% FBS, 2mM L-glutamine, 0.1 mM 2ME, 0.5 mM sodium pyruvate, 0.25x N2 supplement, 0.5x B27 supplement, 12.5 ng/ml FGF4 and 50mg/ml heparin.

37. The method of any preceding claim, wherein the method is capable of allowing PS cells and TS cells, and optionally PES cells, to organise into embryo-like structures.

38. The method of Claim 37, wherein the embryo-like structures undertake lumenogenesis, symmetry breaking and mesoderm specification.

39. The method of any preceding claim, wherein the PS cells have been genetically modified.

40. An in vitro cell culture medium comprising the TPS medium of any of Claims 21 to 36.

41. A kit for culturing a mammalian cell comprising:

i) the in vitro cell culture medium of Claim 40; and

ii) a substrate according to any of Claims 7 to 20.

42. The kit of Claim 41 further comprising one or more receptacles suitable for containing a culture comprising the substrate and medium in combination with PS and TS cells.

43. An embryo-like structure obtainable by the method of any of Claims 1 to 39.

44. Use of the in vitro culture medium according to Claim 40 for culturing PS cells in combination with TS cells.

45. The use of Claim 44, wherein the PS cells and TS cells are cultured in a substrate according to any of Claims 7 to 20.

46. A method of investigating mechanisms involved in embryogenesis, comprising the methods of any of Claims 1 to 39.

47. A method of identifying a compound useful for treating a disease, the method comprising contacting a cell or embryo-like structure obtainable by the method of any of Claims 1 to 39 with the compound.

48. A cell or embryo-like structure obtainable by the method of any of Claims 1 to 39 for use in a method of diagnosing or treating a disease in a patient in need thereof.

49. The cell or embryo-like structure for use according to Claim 48, wherein the cell or embryo-like structure is for transplantation into the patient.

50. The cell or embryo-like structure for use according to Claim 48 or 49, wherein the pluripotent cell was obtained from the patient.

51. A method of providing a transgenic non-human animal, comprising gestating an embryo derived from a cell cultured using a method according to any of Claims 1 to 39.

52. A method of elucidating the role of a gene in embryo development, the method comprising obtaining a pluripotent cell where the gene has been modified or knocked out and culturing said cell using the methods of any of Claims 1 to 39.