Pluripotency is dependent on the maintenance of a proper epigenetic landscape (Gaspar-Maia et al., 2011; Orkin and Hochedlinger, 2011). Bivalent domains, which are defined by the paradoxical coexistence of a permissive histone mark (H3K4me3) and a repressive mark (H3K27me3), are thought to play an important role in pluripotency by keeping developmental genes in a silenced state poised for activation upon differentiation (Azuara et al., 2006; Bernstein et al., 2006). In support of this model, bivalent domains have been shown to be present preferentially at developmental regulatory genes in undifferentiated embryonic stem cells (ESCs) and several adult tissues in vivo including sperm, testis, the cerebellum, and the hematopoietic compartment (Mikkelsen et al., 2007; Cui et al., 2009, 2012; Hammoud et al., 2009).The moderate de-repression of developmental regulators in eed mutant ESCs, which have reduced levels of H3K27me3, further supports this model (Azuara et al., 2006; Boyer et al., 2006). However, it has recently been proposed that the concomitant presence of H3K4me3 and H3K27me3 at developmental genes in ESCs is an in vitro artifact resulting from suboptimal culture conditions (Hong et al., 2011; Marks et al., 2012). In addition, direct evidence for bivalency in the developing mammalian embryo is limited and conflicting due to technical difficulties associated with the low amounts of material available (Alder et al., 2010; Dahl et al., 2010; Rugg-Gunn et al., 2010). The universal nature of bivalency has been further questioned due to conflicting reports from non-mammalian species, with bivalent domains demonstrated to exist in zebrafish (Vastenhouw et al., 2010) but not detected in Xenopus or Drosophila embryos (Akkers et al., 2009; Schuettengruber et al., 2009). We therefore lack a clear understanding of whether bivalency exists in embryonic cells in vivo, and how it relates to pluripotency.
Of particular interest to pluripotency are primordial germ cells (PGCs), the embryonic precursors of the germline. PGCs have a transcriptional profile similar to ESCs (Qin et al., 2012), including the silencing of developmental regulators, and can give rise to pluripotent stem cells when cultured in vitro. However, unlike ESCs, PGCs are unipotent and unable to contribute to chimeras. Additionally, PGCs are thought to undergo a period of ‘epigenetic erasure’ or reprogramming via global removal of both DNA methylation and histone modifications that may prevent transmission of epimutations (Greer and Shi, 2012; Seisenberger et al., 2013). Due to the low number of PGCs present during development, nearly all evidence for this model is from immunofluorescence (IF)-based methodologies that lack gene level resolution and have shown conflicting results (Seki et al., 2007; Hajkova et al., 2008; Kagiwada et al., 2012). To characterize the chromatin state of PGCs, we developed a low cell number Chromatin Immunoprecipitation (ChIP) for the analysis of histone marks using less than 10,000 cells per IP without the need for carrier chromatin or pre-amplification. Using this technique, we performed ChIP-Seq and ChIP-qPCR to show that bivalent domains are present at developmental regulatory genes at multiple stages of PGC development.