Great question! I am guessing that you are keeping up with current news about stem cell research and how scientists are trying to "instruct" cells in terms of their identities (as a skin or nerve cell in your example) so as to use them in therapeutic ways.

Under normal circumstances, as an animal develops from egg to embryo to fetus, cells become more specialized, taking on specific identities (or "fates") as they become part of specialized tissues, organs or structures. We think of these cells as "terminally differentiated" - once they lock in to a certain fate pathway, they cannot revert or deviate from that pathway, and that is probably the case most of the time. But recent studies have suggested that the developmental arrow does not always point in one direction. Now, there is also a special set of cells stashed away in different areas of the body that are "multi-potent" - they can give rise to lots of different cell types as needed. These are the stem cells. They are difficult to isolate, grow and study, but we have some idea about what allows them to have this special ability (more on this below).

Back to the skin and nerve cell. One of the first things to appreciate is that all of the cells in your body contain the same DNA (genes), yet you have many different hundred of cell types (fates). This is because each cell type is expressing (using) a unique combination of genes that encode proteins that carry out the special functions. Your retina, for example, is expressing unique proteins that allow the rods and cones to form and function while your heart cells express genes encoding ion channels and specialized proteins for heart muscle to work. So, for a skin cell to change to a nerve cell, these gene expression combinations have to change - and change dramatically. And under most circumstances, that just won't happen (you can imagine why our bodies would try to avoid that - we don't want retinal cells suddenly appearing in our hearts or on our skin!). BUT, some very recent work that was published at the end of 2007 was very exciting, because it suggested that under very special circumstances, a skin cell could be re-instructed to take on a different fate/identity. Researchers took skin cells, which we consider to be "terminally differentiated" and cultured them in a dish. Then, these researchers did something very difficult - they "tricked" the skin cells into turning "on" a few genes that only STEM CELLS express. These skin cells stopped expressing their "skin" genes and they "reverted" (the fate arrowhead pointed the opposite direction) to become stem cells (or at least stem cell like - they are still being studied). Then, the researchers provided a few key growth factors to the cells to "instruct" them what identity to take on - we have some vague ideas of how to tell stem cells to become neuron-like, and that is what the researchers did (now the arrow head points in the forward direction again). Now the cells stop expressing the "stem cell" genes and switch on the "neuron genes." So, the skin cells became neurons in a kind of two step process. First, they revert to a more "plastic" stem cell type, and then they switch gears and go down a different fate pathway. This is as opposed to just a one step "switch" from skin to neuron. This is still all very difficult and very preliminary, but super exciting. You can find lay-press articles on this research in Time and Newsweek, but the papers were of course, published in research journals: