Module 1 · The Pluripotent State
Embryonic Stem Cells & Pluripotency
An embryonic stem cell (ESC) is a cell line derived from the inner cell mass (ICM) of a pre-implantation blastocyst that can be propagated indefinitely in culture while retaining the ability to generate all three germ layers. ESCs collapse the enormous diversity of an embryo down to a single cell type that a laboratory can grow, freeze, electroporate, and differentiate on demand. Their derivation in mouse (Evans & Kaufman 1981; Martin 1981) and human (Thomson 1998) are two of the watershed moments of modern biology.
1. From Blastocyst to Cell Line
Three days after fertilisation, the mouse embryo is a ~32-cell morula. Differential adhesion (E-cadherin) and asymmetric Hippo signalling sort cells into an outer trophectoderm (future placenta) and an inner cell mass (ICM, future embryo). A further day of signalling sorts the ICM into epiblast (embryo proper) and primitive endoderm (yolk sac). ESCs are derived by isolating the ICM at E3.5 (mouse) or day 5–6 (human) and plating on mitotically inactivated feeders or defined matrix.
The critical insight (Smith 1988; Williams 1988) was that leukaemia inhibitory factor (LIF) / JAK-STAT signalling, or BMP4 + FGF withdrawal, holds mouse ESCs in a ground state of pluripotency. The modern “2i” medium (Ying & Smith 2008) uses small-molecule inhibitors of MEK and GSK3 to achieve the same.
2. The Core Pluripotency Circuit
Three transcription factors — Oct4 (Pou5f1), Sox2, and Nanog — bind co-operatively at ESC super-enhancers and co-regulate essentially every pluripotency-associated gene. They each bind the promoters of the other two (and their own), forming a mutually reinforcing triad that implements a bistable switch. ChIP-seq from Young’s and Ng’s labs (2005–2008) revealed the “core circuitry”.
Downstream effectors include Klf4, Esrrb, Tbx3, Sall4, and a set of microRNAs (miR-290 cluster) that repress differentiation-inducing genes. The whole network sits on top of a distinctive chromatin state: low H3K9me3 heterochromatin, bivalent H3K4me3/H3K27me3 domains at developmental genes (Bernstein 2006), and active X (in female cells). These bivalent domains are the molecular substrate for “poised” developmental regulators.
Simulation: Bistable Pluripotency Circuit
Click Run to execute the Python code
Code will be executed with Python 3 on the server
3. Naive vs Primed Pluripotency
Mouse ESCs and human ESCs look superficially similar but differ in a fundamental way (Nichols & Smith 2009). Mouse ESCs derived in LIF/2i represent the naive state: uniform colonies, two active X chromosomes in female cells, LIF-dependent, able to contribute to chimaeras after blastocyst injection. Standard human ESCs and post-implantation mouse EpiSCs represent the primed state: flat colonies, one active X, FGF/Activin-dependent, cannot chimaerise standard blastocysts.
The primed state is closer to E5.5–E6.5 post-implantation epiblast. Methods to derive genuine human naive ESCs (t2iLGöY, Theunissen 2014; 5iLAF, Takashima 2014; PXGL, Guo 2017) are now standard. Whether primed or naive better supports differentiation depends on the target: hematopoietic, cardiac, and some neural protocols favour primed starting material; extra-embryonic lineages and some early models favour naive.
The existence of two stable pluripotent states tells us that the “pluripotency plateau” of the Waddington landscape has sub-structure — at least two well-defined attractors, and possibly a continuum.
4. The Teratoma Assay & Germline Transmission
Demonstrating pluripotency requires a functional test. Three standard assays:
- Embryoid body formation: in suspension, ESCs aggregate and spontaneously form all three germ layers. Simplest, most permissive; not definitive.
- Teratoma: subcutaneous injection into an immunocompromised mouse; histological evidence of ecto/meso/endoderm in the resulting tumour is the field’s workhorse proof.
- Blastocyst chimaerism & germline transmission: ESCs injected into a blastocyst contribute to the chimaera, including the germline. Only for mouse (and recently rat); ethically impossible for human.
The “tetraploid complementation” assay — ESCs injected into a tetraploid host blastocyst (whose cells cannot contribute to the embryo proper) produce an embryo derived entirely from the injected ESCs — is the most stringent test of pluripotency, satisfied by naive mouse ESCs and by very few human lines.
5. Directed Differentiation
The practical usefulness of ESCs depends on guiding them down specific developmental trajectories. The modern recipe: mimic embryonic signalling in stepwise fashion. For neural (ectoderm): dual-SMAD inhibition (Noggin + SB431542, Chambers 2009). For mesoderm: BMP4 + Activin + FGF. For endoderm (pancreas, liver): high Activin (nodal-mimetic). For cardiomyocytes: the Murry protocol (CHIR then IWR-1, 2012).
Protocol efficiency ranges from >90% (cardiomyocytes, neurons) to <10% (renal tubule, true haematopoietic stem cells). The bottlenecks are well-characterised for some (haematopoietic: cannot reproduce the AGM/aorta endothelium-to-HSC transition) and mysterious for others.
6. Why ESCs Matter Medically
Human ESCs are the first cell type in history that can, in principle, be differentiated to any of the ~200 human cell types under laboratory control. Every current iPSC clinical therapy — RPE for AMD, dopaminergic neurons for Parkinson’s, islet cells for diabetes, retinal photoreceptors, cardiomyocytes for post-MI cardiac patches — uses protocols originally worked out on ESCs. The ethical controversies that shadowed ESC work in the 2000s have largely been sidestepped by iPSC technology (Module 2), but the underlying biology and every differentiation protocol were established in ESCs.