RESEARCH NEWS & VIEWS DEVELOPMENTAL BIOLOGY hESCs and hiPSCs) that are grown in conven- tional culture conditions in the laboratory are molecularly equivalent to cells in an early-stage embryo that are ready to undergo gastrulation7. A model of human Thus, hESCs and hiPSCs can be grown indefi- nitely in this state in a tissue-culture incubator, or cryopreserved in liquid nitrogen and thawed embryo implantation and cultured again many years later. Currently, hESCs and hiPSCs are used in An innovative microfluidic device has enabled the modelling of the events that different ways to study the various types of occur in human embryos when they implant in the wall of the uterus. It could be cell that make up human embryos and fetuses. used to help understand early pregnancy loss. See Letter p.421 They can be allowed to spontaneously differ- entiate to create aggregates of cells known as embryoid bodies8. Alternatively, their differ- AMANDER T. CLARK of cell-differentiation events in the human entiation can be directed by incubation with embryo around the time of implantation. The specific cocktails of growth factors, to create ach year, millions of women around the events involved include the formation of the simplified 3D models of fetal organs or spe- world experience early pregnancy loss1. amniotic sac and a disc of cells inside it called cific germ layers9,10. hiPSC lines derived from Most pregnancy losses occur in the first the epiblast; the initial specification of pri- patients can be used for ‘disease in a dish’ mod- Etrimester, around the time that the embryo mordial germ cells (PGCs; the eventual sperm elling. Such strategies give rise to the same implants in the lining of the uterus. Our lack of or egg cells); and the start of a process called cell types as those found in late-stage embryos understanding of why some early pregnancies gastrulation. In gastrulation, the epiblast dif- and fetuses; however, the initial germ-layer fail is frustrating to women and their partners. ferentiates into the three germ layers of cells differentiation seen in these models is dis- To fill this fundamental gap in our knowledge, — the ectoderm, mesoderm and endoderm organized, and bears no resemblance to the models of human embryo implantation are — that will form the fetus. spatially organized differentiation seen in greatly needed. On page 421, Zheng et al.2 Crucially, this model does not involve the human embryos during implantation. There- report the creation of an advanced model of use of intact human embryos or an in vitro fore, such models have not been very helpful human embryo development that recapitulates attachment system for human embryos6. for studying the peri-implantation stage of many of the key events that happen around the Instead, it relies on a simple bioengineered human pregnancy. time of implantation (that is, during the peri- microfluidic device and on the differentiation To create a 3D environment in which the implantation stage). of pluripotent cells — which can differentiate stem cells could develop into embryonic- The authors’ model involves the genera- into any type of cell in the body. These pluri- like sac structures, Zheng and colleagues tion of synthetic structures resembling the potent cells can be derived from human engineered a microfluidic device containing embryonic sac — which, later in development, embryos (human embryonic stem cells; three channels: a central channel for loading a becomes a fluid-filled bag and the embryo. It hESCs) or generated through the conversion matrix material, another for loading hESCs or is a substantial advance on previous exciting of differentiated human cells (human induced hiPSCs, and a third containing flowing liquids models developed by this group3 and others4,5 pluripotent stem cells; hiPSCs). carrying molecules called morphogens that because it reveals the spatial coordination Human pluripotent stem cells (including induce stem-cell differentiation (Fig. 1a). The a b c Anteriorized Amniotic hESC Embryonic-sac-like ectoderm-like cell Pluripotent structure cell Morphogens hESC 18 h 36 h Matrix Matrix pouch Posteriorized PGC-like cell Primitive- streak-like cell Figure 1 | Modelling human embryo development during implantation. influence the development of pluripotent cells on either side of each sac. a, Zheng et al.2 developed a three-channel microfluidic device containing c, After 36 hours of exposure to various morphogens, the sacs show signs a cell-loading channel, a matrix-filled channel and an induction channel of anteriorization or posteriorization — that is, features of the anterior or through which morphogens (molecules that affect development and growth) posterior ends of an embryo. Anteriorized and posteriorized sacs have cells can be pumped. Human pluripotent cells (which can differentiate into all resembling those that form the developing amniotic membrane (amniotic the cell types of the body) that are fed into the cell-loading channel settle ectoderm-like cells), and posteriorized sacs show cells that resemble in matrix pouches. b, After 18 hours, the pluripotent cells (in this case, primordial germ cells (PGCs; the eventual fetus’s sperm or egg cells) and cells human embryonic stem cells; hESCs) form embryonic-like sac structures. of the posterior primitive streak (which, in human embryos, form the basis of Morphogens pumped through the induction and cell-loading channels the body’s bilateral symmetry). 350 | NATURE | VOL 573 | 19 SEPTEMBER 2019©2019 Spri nger Nature Li mited. All ri ghts reserved. ©2019 Spri nger Nature Li mited. All ri ghts reserved. NEWS & VIEWS RESEARCH microfluidic device was lined with trapezium- neuronal cells would have formed. Therefore, extra information about embryonic develop- shaped posts 80 micrometres apart that created Zheng and colleagues’ micro fluidic model ment. In turn, we might be able to use this evenly spaced matrix pouches in which the might technically be exempt from the 14-day information to help millions of women around stem cells could grow and differentiate. rule because it does not represent an intact the world to better understand or even avoid The authors loaded stem cells into the device embryo in the way that we consider an embryo early pregnancy loss. ■ and, 18 hours later, pumped morphogen-con- donated for research from an in vitro fertili- taining fluids into either the liquid-containing zation clinic to be intact. However, the eth- Amander T. Clark is in the Department channel on one side or the cell-loading channel ics of running experiments using human of Molecular, Cell and Developmental on the other (Fig. 1b). After 36 hours, struc- embryonic-like sacs for longer than 14 days Biology, and the Broad Stem Cell Research tures with 3D organization that mimicked should be evaluated, with particular focus on Center, University of California Los Angeles, a peri-implantation human embryonic sac whether the structure should at any point be Los Angeles, California 90095, USA. had formed in each pocket. Furthermore, considered a human being. e-mail: [email protected] the authors could use different morphogens Finally, I urge policymakers to consider 1. Wang, X. et al. Fertil. Steril. 79, 577–584 (2003). to induce anteriorization or posterioriza- this device to be a useful addition to the 2. Zheng, Y. et al. Nature 573, 421–425 (2019). tion — that is, the development of character- human-stem-cell toolbox, as it can be used 3. Shao, Y. et al. Nature Commun. 8, 208 (2017). 4. Christodoulou, N. et al. Nature Cell Biol. 20, istic features of the anterior or posterior ends to examine a period of human embryonic 1278–1289 (2018). of a normal embryo (Fig. 1c). However, the development that is currently inaccessible for 5. Simunovic, M. et al. Nature Cell Biol. 21, 900–910 embryonic-like sacs did not anteriorize and in vivo research. The model can be scaled up (2019). 6. Deglincerti, A. et al. Nature 533, 251–254 (2016). posteriorize at the same time, which is crucial to become high throughput, and thus could be 7. Nakamura, T. et al. Nature 537, 57–62 (2016). for further embryo development. invaluable to environmental toxicologists. The 8. Itskovitz-Eldor, J. et al. Mol. Med. 6, 88–95 (2000). In posteriorized embryonic-like sacs meas- length of time for which these structures can 9. Spence, J. R. et al. Nature 470, 105–109 (2011). 10. Irie, N. et al. Cell 160, 253–268 (2015). uring just over 100 micrometres in diameter, be studied in the device is relatively short; cur- 11. Aach, J., Lunshof, J., Iyer, E. & Church, G. M. eLife 6, the authors observed cell populations that rently, most of them collapse within a few days. e20674 (2017). resembled a developing amniotic membrane In future, if these embryonic-like sacs could be (the membrane that lines the walls of the cultured for longer, they could provide crucial This article was published online on 11 September 2019. amniotic-fluid-filled sac), a posterior primi- tive streak (which forms the basis of the body’s bilateral symmetry), mesoderm-like cells and, COMPUTING remarkably, PGC-like cells. In the anteriorized embryonic-like sacs, the authors saw anterior primitive streak-like cells and endoderm-like cells. Thus, this is one of the first humanized Stochastic magnetic bits models that can enable study of the awesome complexity of spatial relationships between cells in the peri-implantation window of rival quantum bits development. At this point, readers might be asking Circuits based on the stochastic evolution of nanoscale magnets have been used to whether these embryonic-like sacs are, in fact, split large numbers into prime-number factors — a problem that only quantum human embryos. The structures can be consid- computers were previously expected to solve efficiently. See Letter p.390 ered imitations of the real things, but crucially, they are not viable (they could not develop into a normal fetus) and do not have certain DMITRI E.