Year 4 (NCE)

The goal of our work has been to derive new human embryonic stem cells lines from embryos that are donated by couples at the completion of in vitro fertilization cycles. Many laboratories including our own, derived lines using the method that was originally described. Embryos that were grown for 5-6 days in the laboratory were transferred to a lawn of cells that supported the eventual outgrowth of human embryonic stem cells. Using standard methods to perform derivations, we were struck by ways in which we thought this method could be improved. For example, the cells of the embryo that are destined to form the placenta die. Therefore, it seemed likely that the products they release could have untoward influences on the stem cells. In addition, we observed the formation of morphologically distinct cell types before tightly packed “islands” of stem cells finally emerged in their midst. We speculated that these extraneous cells could also have undue effects. Finally, we examined, at very high magnification, embryos that were grown for 5-6 days. Although the cells that give rise to the offspring are supposed to be nearly equivalent at this stage, we saw that they no longer look alike. We took this as possible evidence of specialization, meaning they might already be starting down the road toward becoming one of the cell types of the body. Therefore, we began to consider new approaches for producing human embryonic stem cells. In particular, we wanted to derive lines from single cells of earlier stage embryos that might be less specified and more plastic in terms of their ability to differentiate into progeny that are difficult to obtain from conventional human embryonic stem cell lines.

In preliminary experiments, we worked with other scientists from the biotechnology sector to develop methods for deriving lines from single cells (termed blastomeres) that were removed from embryos that were grown for approximately 3 days in the laboratory. From the work of other scientists we knew that around this time the intrinsic genetic programs of the egg fade and those of the embryo start to unfold. Therefore, it was quite possible that human embryonic stem cell lines derived at this stage would be less specified than their counterparts, which were established by conventional methods. However, our initial work showed that this new approach was not very efficient. Although a few blastomeres formed lines, the majority did not. Therefore, we had to make this a more robust procedure. We took our clues for improving this method from the architecture of the embryo. Specifically, the cells are always grouped together and never found in isolation. Thus, to mimic this arrangement, we surrounded each blastomere with cells and proteins to simulate the geometry and molecular environment of the early embryo. The result was a much more efficient derivation procedure. Accordingly, we had developed the tools to perform the desired experiments.

For the derivation procedure, we used 5 embryos that were donated by a single couple. From single blastomeres removed at the 3-day stage we derived a total of 9 human embryonic stem cell lines. Thus, the lines were either genetically identical or related. First, we had to prove that they had the characteristics that researchers think are common to all human embryonic stem cells. Specifically, they expressed a panel of markers that are associated with this state. Also, in a laboratory dish, they made descendants of the 3 lineages that give rise to all the cell types in the body. The most rigorous test of the stem cell state is transplantation into a mouse where these cells are able to differentiate further. Like those derived by conventional means, our collection formed many structures such as bone, muscle and glands. Therefore, we concluded that they had the same general properties as human embryonic stem cells that were derived from later stage embryos.

Thus, we turned to the interesting question of whether they had unique properties. We characterized the cells at the genetic, epigenetic and functional level. At the level of genes, we compared the lines that were derived from a single cell of a 3-day old embryo to human embryonic stem cell lines that were derived by conventional means. There were striking differences in the gene expression patterns between the two. Epigenetic marks decorate DNA designating functional from nonfunctional areas in terms of gene expression. The lines that were derived from single cells had many fewer of these modifications suggesting less specialization. Finally, at a functional level, we showed that a subset of the lines we produced could spontaneously form cell types that are found in both the offspring and the placenta. Since these lineages diverge early in development, we took this as evidence of a greater degree of plasticity than is associated with the conventional lines. We think this property could be very useful for devising regenerative medicine therapies.