I have been fascinated by the neural crest and the pigment cells it forms since I first started working with zebrafish, as a post-doc in the Nüsslein-Volhard lab, in 1991. Neural crest (stem) cells are multipotent progenitors of an astonishing diversity of cell-types – including both those associated with ectoderm (neuronal, glial and pigment cell-types) and those associated with mesoderm (skeletogenic fates - although this is becoming controversial - see post!) – and for this reason the neural crest has been nicknamed ‘the fourth germ layer’.
My lab's goals have been two fold, to understand how:
1) these different fate choices are made, stably, and in the correct proportions, and
2) they become distributed in the correct way i.e. how does pigment patterning occur?
Our primary approach has been a genetic one, beginning with the Tübingen 1996 screen. During this exciting period, I helped identify nearly 100 genes controlling pigment cell development in the zebrafish embryo. Over the following years, first as a post-doc with Judith Eisen in Eugene, Oregon, and since 1997 in my own lab at the University of Bath, I have focussed on identifying those genes crucial to fate specification and patterning, and then on studying how they control neural crest development at a cellular and molecular level.In all our work, pigment cells form a ‘model within a model’, but we do not lose sight of the important roles some of these genes play in neuronal and glial development, and the impact they have on human congenital disease.
In the meantime, whilst mutants are the gold standard for understanding gene function, we have come to realise the value of a complementary chemical biology approach, principally for its allowing temporal control of protein function. We have performed substantial (1400 compound) screens to assess effects on pigment cells and have identified compounds generating phenotypes reminiscent of those identified in our genetic screens, albeit usually in combinations.
As our knowledge of individual genes has increased, and as they began to be slotted into gene regulatory networks (GRNs), we began to question whether we really understand the diagrams we were proposing as models - how good is our intuitive understanding of these networks? This triggered a very enjoyable period, which we are very much still in, of exploring the contribution that a mathematical modelling approach can contribute. Inspired by the pioneering zebrafish studies of Julian Lewis, Nick Monk and Andy Oates, and teaming up with biological mathematicians, principally Andrea Rocco (originally in Bath, but now at the University of Surrey) and more recently with Hartmut Schwetlick (Dept of Mathematics, University of Bath), has allowed us to develop an iterative approach combining experimental genetics, mathematical modelling and simulation. This adds clarity to our hypotheses and rigour to our evaluations of our GRNs, and we are now exploring how far this ‘bottom-up’ systems biology approach can take us.
Another approach usually conjured up by the systems biology label is ‘big data’, derived from microarray or deep sequencing studies. What I would love to be able to do, is to use such data to recreate de novo the GRN for pigment cell development. This ‘top-down’ approach risks being untestable (you will conjure up some sort of GRN, but will it have any resemblance to reality?), but as our ‘bottom-up’ approach progresses we have a growing core network to look for in any ‘top-down’ driven model.