Department of Biology & Biochemistry

Professor of Stem Cell and Developmental Genetics

4 South 0.63

Tel: +44 (0) 1225 383828 


Anyone interested in opportunities for PhD or post-doctoral study, using zebrafish or medaka, should contact me by e-mail in the first instance.

Current lab members

  • Dr Deeya Ballim
  • Dr Yusuke Nagao
  • Ruqaiya Al-Jabri
  • Karen Camargo-Sosa
  • Jennifer Owen
  • Kleio Petratou
  • Marc Shedden

Prof Robert Kelsh


Current research

The big picture – neural crest cells as a model of stem cell development and human disease

We are interested in three fundamental questions in developmental biology:

  • How do multipotent stem cells become specified to one of several distinct fates?
  • How do these specified cells reorganise their gene regulatory network (GRN) to achieve stable differentiation?
  • how is cell migration through the embryo patterned?

These questions are equally important in the related fields of stem cell biology and regenerative medicine, since answering them will help us to understand and control the differentiation of stem cells for therapeutic use, and get them to the correct locations. Likewise, the processes under consideration – maintenance of multipotency, specification of cell fate, genetic control of differentiation, and guided cell migration – are exactly those that are defective in many congenital diseases, so that understanding of the normal process and the disease state go hand-in-hand.

The vertebrate neural crest is an attractive model system in which to examine all three questions. Neural crest cells are multipotent, forming many diverse cell types, including pigment cells, neurons and glia. Additionally, crest cells undergo extensive migrations and yet form a stereotypic distribution of each cell type within the embryo (Kelsh et al., 2008; Kelsh and Erickson, 2013). Furthermore, understanding crest development has important medical implications since defects in crest development are the basis of many human syndromes, including Waardenburg-Shah syndrome and Hirschsprung's disease, and since neural crest stem cells offer significant promise for therapies (Delfino-Machin et al., 2007).

Zebrafish pigmentation - neural crest development in ‘glorious Technicolor’!

We focus largely on neural crest-derived pigment cells as a highly tractable ‘model-within-a-model’ (Lapedriza et al, 2014; Schartl et al, 2016), but also are looking at neuronal derivatives (e.g. Elworthy et al., 2005; Carney et al., 2008; Delfino-Machin et al., 2017). We use the zebrafish as our model system because the embryo is exquisitely suited to their direct study – the transparency of the embryo makes studying these cells beautifully straight-forward, and the three different pigmented cell-types, all believed to derive from a common progenitor (the chromatoblast), provide a simple model of neural crest development within the neural crest.

We are studying the fundamental issue of how stem cells function – does fate choice follow a Direct Fate Restriction or Progressive Fate Restriction Model – and aim to reconcile these two conflicting views of neural crest development. We are using classic genetics, single cell transcriptional profiling and various mathematical modelling approaches to understand how GRNs reorganise to allow selection of alternative fate choices (Greenhill et al., 2011; Nagao et al., 2014; Vibert et al., 2017; Nagao et al., in prep.; Subkhankulova et al., in prep.). We are using genetics, clonal analysis and imaging to understand larval pigment pattern formation and the mechanisms driving adult pigment stem cell segregation and quiescence (Camargo-Sosa et al., in prep.). Finally, we are using the zebrafish mutants to model human disease, currently focusing on understanding genotype-phenotype correlations in SOX10-associated conditions.

My laboratory has experience of the majority of techniques applicable in the zebrafish system. More advanced techniques that we are currently developing/utilising include single cell transcriptomics, Cre-lox recombination for clonal analysis, mathematical modelling of both GRNs and pigment pattern formation, and high efficiency phenotypic rescue using microinjection.


Read publications by Robert Kelsh

View more publications »