We’re excited to share our newest publication @nature. With #scRNAseq we investigated 10 key #epigenetic regulators in #embryogenesis. Interesting early task for #Polycomb! Meissner lab @MPI_MolGen & @HSCRB /w @helenekretzmer & Zachary D. Smith (1/13) nature.com/articles/s4158…
What is normal in #devbio? First, we generated a single-cell reference of WT mouse #gastrulation and early organogenesis to which we later compared mutant phenotypes. We identified 42 embryonic and extraembryonic cell states and placed them in a lineage tree. (2/13) 
Our WT reference is time-resolved covering #embryo stages E6.5-8.5. Variation is information: We pooled multiple embryos with unique SNP profiles per single-cell experiment; cell-to-embryo assignment is possible from scRNA data. (3/13)
Next, we mutated DNA- and histone methyltransferases individually at the zygote stage using #CRISPR Cas9, which we confirmed by inspecting scRNA reads over cut sites. Many mutants isolated at E8.5 are developmentally delayed as indicated by their cell composition. (4/13)
But the loss of epigenetic regulator function can result in abnormalities beyond delayed #development. For example, mutants of the essential PRC2 component Eed contain earlier cell states that are present in strikingly abnormal proportions. (5/13)
For example, female EED mutants show strongly diminished extraembryonic ectoderm, in line with Terry Magnuson’s work on PRC2 in X chromosome inactivation (nature.com/articles/ng574z). Of note, defective imprinted #XCI is detectable at the level of few remaining single cells. (6/13)
Without EED not only extraembryonic mesoderm but surprisingly also primordial germ cells (#PGC) are overproduced. We put this scRNAseq discovery back into the whole-embryo context using a PGC-reporter generated by Mitinori Saitou's lab (nature.com/articles/ng.186). (7/13)
At least initially, PGCs form at the expected time and location in PRC2 mutants, in numbers comparable to WT. Interestingly, #germline/#pluripotency markers are broadly derepressed and co-expressed in cells, indicative of an insufficiently silenced pluripotency network. (8/13)
Now all-in-one: PRC2 mutants appear comparatively normal at the onset of gastrulation (left), then overproduce posterior-proximal structures, while the embryo proper remains underdeveloped and forms cell states that strongly deviate in expression (right). (9/13)
Is this phenotype mainly caused by deregulated signalling between cells? No. We exposed EED KO embryonic stem cells to various growth factors. Pluripotency & mesodermal genes are derepressed/responsive; neuroectodermal genes not, indicating strong cell-autonomous effects. (10/13)
Feel like sth is missing for a Meissner lab study? DNA methylation in the #epiblast is substantially altered not only for PRC2 but also other polycomb mutants; without correlation between affected genes & phenotype. Broad epigenetic changes manifest prior to gastrulation! (11/13)
In sum, our strategy recovers robust morphological and transcriptional information for complex mutant phenotypes, and uncovered a dominant role for PRC2 in restricting the germline. Moving towards a fully quantitative view of how cell diversity emerges during development. (12/13)
This work wouldn’t be there without my brilliant colleagues at the @MPI_MolGen, in particular @helenekretzmer, Zachary D. Smith @HSCRB, @AbhiSampKumar, @hetzel_s, and Alex Meissner. Many questions remain and we are up for the challenge! (13/13)
