Elaine Fuchs of Rockefeller University in New York City studies how stem cells in the skin maintain their orderly but complicated cycle of skin renewal, how they adjust so effectively to heal wounds and what goes wrong in disease.

If you could list only five markers or transcription factors that people who study skin stem cells really need to know, what would they be?

Key factors for the hair follicle stem cells are Sox9, Tcf3, K15, CD34 and Lgr5. Key markers that encompass epidermal and hair follicle stem cells are p63, K14 and K5, but these proteins are expressed by all keratinocytes with proliferative capacity, irrespective of whether they are stem cells or transit amplifying cells.

One thing you've done a lot of is marking and following cells. How do you decide what labels to use?

There is no crystal ball as to how to accomplish this, as you always run the risk of altering the biology of the protein. You just have to do the controls to test for preservation of protein function. Let me offer a few examples:

Elaine Fuchs, Rockefeller University

If we just want to look at Sox9 expression, but only in cells that express Sox9, then we would engineer the Sox9 locus to express GFP [green fluorescent protein] under the control of the Sox9 promoter, so that GFP expression will be exclusive to the cells that express endogenous Sox9. Here it would be important to use a short-lived GFP, such that the minute the cell switches off endogenous Sox9 expression, you'll lose GFP protein.

If you want to follow what happens to progeny of Sox9-expressing cells, then you want to first engineer a gene encoding Cre-lox recombinase driven by a Sox9 promoter. Then, mate this mouse to another mouse harbouring a gene containing a constitutively active promoter followed by a stop codon that's flanked by lox recombinant sites and then [a] GFP coding sequence. In that way, whenever the cells express Sox9 they activate the recombinase, the recombinase removes the stop codon and those cells start expressing GFP. Then all the progeny will always be marked by GFP, because from this point onward, GFP expression is regulated by a constitutively active promoter.

And then if you control the timing of when cells express the Cre (using [a] progesterone or tamoxifen inducible system), you can choose when to start marking and following the GFP-expressing progeny that came from cells that express Sox9.

I wanted to talk to you about the orientation of skin cell division. It seems like in the embryo, skin cells are dividing out to cover the embryo, whereas in the adult skin, cells are dividing up to be shed from the surface. What are the big questions in understanding this transition?

One surprising finding for us really built off an observation made many years back when I was still at the University of Chicago. We noticed that the plane of the spindle orientation undergoes a quite dramatic change during embryonic development, where it goes from being parallel to the basement membrane to [being] perpendicular. We now have some insights on the molecular features that enable the epithelium to undergo these changes in the division plane.

A simple epithelium doesn't have to think about this — it's always dividing parallel to the basement layer. But if you're going to have a stratified epithelium, organized in layers, there has to be some mechanism that allows this stratification. When embryonic epidermal cells divide asymmetrically relative to the basement membrane, they do so by sensing polarity cues. Genetics has told us that flies use a set of evolutionarily conserved proteins to polarize divisions in the neuroblasts and produce divisions that yield one undifferentiated daughter and one daughter committed to make a neuron. We've found that this group of proteins is there at the right place in time to play a similar role in orienting divisions in the skin epithelium, perhaps to generate one proliferative daughter and one daughter committed to stratify and terminally differentiate. What's interesting about the skin is that the epidermis doesn't restrict divisions to be exclusively symmetric or asymmetric. It does both. When and how the skin controls these division choices will be interesting to decipher in the future.

Editor's note: These are some more specialized questions than the ones discussed in a  broader interview.