Comparison of ES cells with relevant compartments of early embryos

Embryonic stem cells (ES cells) are pluripotent stem cells that can contribute to all lineages of the embryo proper and were first isolated in 1981 in this laboratory (Evans and Kaufmann, 1981). They have contributed immeasurably to biology, both in terms of studying pluripotent stem cell maintenance in culture and in vitro differentiation, and because they can be easily genetically manipulated in culture, allowing for the production of mice with designed specific mutations (targeted mice), via homologous recombination. The recent isolation of human ES cell lines is promising in terms of stem cell therapy development. Mouse ES cell lines are isolated from the inner cell masses of mouse blastocysts between days 3.5 and 5.0 post coitum, or from the inner cell masses of delayed blastocysts. However, it has been suggested that some aspects of their biology may be more similar to embryonic ectoderm from early post implantation embryos, than to pre-implantation blastocysts. It has also been suggested that human embryonic stem cells are more similar to mouse stem cells derived from early embryonic ectoderm, than to mouse ES cells. Until recently, however, it has been difficult to answer these questions definitively, as the quantity of tissue available from early embryos is extremely limiting. However, we have used two rounds of RNA amplification on early embryonic samples and individual colonies picked from ES cell culture to compare the transcriptomic profiles of mouse ES cells with the embryonic tissues from with these cells are commonly derived, and with slightly later embryonic stages. We have found that ES cells are notably most similar in profile to embryonic ectoderm at day 5.5 (EE5.5) . ES cells cannot be derived from this post implantation embryonic tissue, so this result is in some ways surprising. However, ES cells show key differences from EE5.5, including the expression of markers of pluripotency such as Oct4 and Nanog (see gene lists acompanying this record), which explain how their ability to contribute to all tissues of the embryo proper is retained. Keywords: cell type comparison We used the NIA 15K array set, which is perfect for this study as the cDNA clones that comprise this set are derived from germ line tissues, germ cells and various embryonic stages. The pooled control used in this study was assembled using all of the samples used originally when isolating clones for the NIA15K set. This reference sample should therefore hybridize to most of the clones in the NIA set, providing a good background signal with which to compare samples from specific embryonic stages. This pooled control was always labelled with Cy5, allowing all of the experimental samples (IMT11 ES cells, ICM88, ICM105, DICM136, DICM180, EE5.5 and EE6.5) to be labelled with Cy3. Since these samples were always labelled with cy3, they could be compared with each other without risk of dye bias, since they were all labelled with the same dye. 6 repetitions were carried out for all samples except DICM180 (which was extremely difficult to isolate), for which there are 4 repetitions. 3 biological replicates were carried out for each stage (2 slides per biological replicate gives 6 reps total). RNA was isolated using the Zymo Mini RNA isolation kit, and then amplified twice using the Arcturus RiboAmp kit according to the manufacturer's protocol. This yielded enough amplified RNA for target labelling and array hybridization. The Atlas™ PowerScript™ Fluorescent Labelling Kit and 5μg of amplified RNA were used for labelling according to the manufacturer’s protocol (Promega). This labelling method is indirect (amino-allyl), reducing the chances of direct cyanine dye incorporation affecting the efficiency of hybridization.