Research

Stem cell renewal and lineage selection

One of our ongoing interests is in how stem cell behaviour is regulated by extrinsic signals from the local microenvironment, or niche. Two different and complementary approaches can be taken to investigate this: observing stem cells in vivo and recreating the niche in vitro. Stem cell behaviour in vivo is a composite response to all niche signals, whereas in vitro it is possible to parse out the response to individual signals.

For our in vitro studies we have developed, in collaboration with Wilhelm Huck (Department of Chemistry, University of Cambridge and Radboud University, Nijmegen), micropatterned extracellular matrix (ECM)-coated glass substrates that selectively capture single human epidermal stem cells. The substrates are amenable to microscopic analysis of living cells, allowing us to perform FRET and image cytoskeletal dynamics. In addition, we can perform single cell gene expression profiling of cells on these substrates (Gautrot et al., Biomaterials 2010; 31: 5030; Connelly et al., Nat Cell Biol 2010; 12: 711). We found that when spreading is restricted on small circular islands, cells exit the stem cell compartment and differentiate. When cells can spread on large circular islands, however, they do not differentiate (Connelly et al., Nat Cell Biol 2010; 12: 711). We also found that differentiation does not depend on ECM composition or density. Instead, the state of assembly of the actin cytoskeleton regulates differentiation by controlling serum response factor (SRF) transcriptional activity. Differentiation requires SRF and its co-factor MAL, and is also dependent on the presence of growth factors. SRF target genes, the AP1 transcription factors FOS and JUNB, are required for differentiation. c-Fos mediates serum responsiveness, while Jun-B is regulated by actin and MAL. We are now investigating how stem cells respond to differences in substrate stiffness and topology, and whether environmental responsiveness is altered in cells from squamous cell carcinomas.

Complementing the in vitro studies, we have continued to investigate the stem cell compartment in vivo, using genetically modified mice. One of the key pathways that regulates epidermal stem cells is the Wnt pathway. We have compared the responses of different stem cell populations to activation of Wnt signalling (Baker et al., Dev Biol 2010; 343: 40). We have found that activation of beta-catenin in the stem cells of the hair follicle bulge stimulates proliferation, but not the formation of additional hair follicles. In contrast, when we target cells at the base of the sebaceous gland they readily form ectopic follicles. We are currently investigating whether this reflects intrinsic differences between the cells, or differences in their local microenvironment.

Stem cells and tumour formation

Interactions between epidermal cells and stromal cells profoundly influence normal differentiation and tumour formation, and we are using a variety of approaches to study these interactions (Jensen et al., Nat Protoc 2010; 5: 898). In a recent study we found that activation of Notch signalling in the epidermis leads to profound changes in the underlying dermis, with the accumulation of cells that express markers of the neural crest, melanocytes, smooth muscle and peripheral nerve (Ambler and Watt, Development 2010; 137: 3569). These effects are dependent on epidermal upregulation of the Notch ligand Jagged 1. Gene expression profiling reveals that epidermal Notch activation results in upregulation of a number of growth factors and cytokines, including TNFalpha, whose expression is dependent on epidermal Jagged 1 expression. Since the Notch pathway is activated in a large number of different tumour types, the effects that we have observed may provide a general mechanism for how Notch signalling alters the tumour stroma.

Figure 1
Section of tumour from a mouse that is chimeric for eGFP expression (green) and expression of activated MEK1 (red) (see Arwert et al., Proc Natl Acad Sci USA 2010; 107: 19903). Ki67 positive nuclei are shown in yellow. Section was also labelled with a nuclear counter stain (blue). Photograph courtesy of Heather Zecchini (Light Microscopy core facility).

Another example of how aberrant activation of a signalling pathway in one cell population can have a profound impact on neighbouring cells comes from our ongoing studies of the consequences of aberrant integrin expression. Integrin expression is normally confined to the basal epidermal layer, but in many squamous cell carcinomas expression extends to the suprabasal cell layers. Suprabasal integrin expression results in upregulation of Erk mitogen-activated protein kinase (MAPK) signalling and we have modelled this by expressing an activated MAPK kinase 1 (MEK1) transgene in the suprabasal, non-dividing, differentiated epidermal cell layers (InvEE transgenics). We have found that wounding induces benign skin tumours in InvEE mice (Arwert et al., Proc Natl Acad Sci USA 2010; 107: 19903). Differentiating, non-dividing cells that express MEK1 stimulate adjacent cells to divide and become incorporated into the tumour (Figure 1). Tumour formation is associated with epidermal expression of IL1α and can be inhibited by blocking inflammation using dexamethasone. Depletion of γδ T cells and macrophages also reduces tumour formation. Our results are quite unexpected, because they show that differentiated epidermal cells can trigger tumorigenesis without re-acquiring the ability to divide.