The Notch pathway provides the most direct route for exchange of gene regulatory signals between adjacent cells. It is critical for coordination of gene expression in almost all animal tissues. Feedback loops built from Notch pathway components drive cell diversification and generate pattern both in space and in time. To understand how tissues become organised during development, maintain their organisation during tissue renewal, and suffer disorganisation in cancer and other diseases, we need to understand the logic and dynamics of the Notch control circuitry, where and when it is brought into play, and what functions it performs in specific situations. We have continued to explore these questions in a variety of tissues, both experimentally and by mathematical modelling, using the zebrafish as our main experimental organism.
Ear and lateral line
The sensory epithelium of the inner ear and lateral line provides a classic example of Notch-mediated lateral inhibition, giving rise to an alternating array of sensory hair cells and supporting cells. We have found that Notch signalling controls not only the spatial pattern of cell types, but also the timing of differentiation. We have generated transgenic zebrafish in which production of specific components of the Notch pathway can be switched on acutely by heat shock, or conditionally by means of a UAS promoter, and we are using these fish to investigate the dynamics of the Notch signalling machinery.
In birds, the array of auditory hair cells and supporting cells, once formed, is normally stable, with no cell turnover; but if hair cells are destroyed, the surviving supporting cells resume proliferation and serve as stem cells for regeneration of hair cells. In collaboration with the laboratories of N Daudet (University College London) and J Stone (Washington University, Seattle), we have investigated the role of Notch signalling in these switches of cell behaviour. The results show that Notch signalling is not required for maintenance of the quiescent state ¿ which evidently depends on some other type of signal - but is reactivated during regeneration.
Gut
In the lining of the intestine, Notch signalling controls the differentiation of secretory and absorptive cells, again through lateral inhibition. In this system, however, new differentiated cells are continually being produced from stem cells, and the sequence of developmental decisions is more complex than in the ear. Our current work in the zebrafish indicates that Notch signalling is critical not only in controlling the basic choice between stem, secretory, and absorptive cell fates, but also in controlling choices between different secretory subtypes. From our conditional knock-out of the Delta1 gene in the mouse intestine, it seems that the same is true in mammals. But the details remain unclear, for lack of knowledge of the precise lineage relationships of the different classes of cells in the gut lining. This has prompted us to develop a new tool for clonal analysis of cell lineage and gene function in the zebrafish.
Cell lineage analysis
Clonal analysis hitherto has been difficult in vertebrates. Our new line of transgenics, TgMAZe, for Mosaic Analysis of gene function in the Zebrafish, should make the task easier. The transgenic construct incorporates a heat-shock promoter that drives expression of Cre. Upon induction, the Cre gene excises itself, and this brings a Gal4-VP16 construct under the control of a constitutive EF1a promoter. The construct also incorporates a UAS:mRFP gene, giving heritable expression of a nuclear-targeted mRFP marker in the clone of cells where the heat shock has triggered self-excision of Cre. By crossing TgMAZe fish with fish carrying an additional UAS driven transgene or combination of transgenes, we can create marked clones of cells in which the Notch pathway is activated or blocked, either by itself or in combination with other pathways. We can then compare their behaviour with that of clones that simply express the RFP marker.
Our new methods of clonal analysis promise to have many applications, not only to tissues such as the lining of the gut and the inner ear, but also to cell populations such as the neural crest and vascular endothelial cells, whose migratory behaviour we have found to depend on Notch signalling.
Somite segmentation clock
Feedback loops in the regulation of genes in the Notch pathway underlie the remarkable temporal oscillations of gene expression that control the segmentation of the vertebrate body. In the zebrafish, our past work has led us to a quantitative theory of this 'segmentation clock', according to which the oscillations in individual cells arise from delayed negative feedback in the autoregulation of the Notch target genes her1 and her7, while neighbouring cells are kept in synchrony through communication via the Notch pathway. By blocking Notch signalling with DAPT, we have now shown that in this system Notch-mediated cell communication is needed only for synchronisation. We have been continuing our efforts to measure more of the key parameters and to clarify the role of certain other members of the her gene family that have been shown to oscillate in parallel with her1 and her7.
Each body segment corresponds to one tick of the segmentation clock, but what dictates the total number of segments? In a collaboration with the laboratory of O Pourquié (Stowers Institute, Kansas City), we have compared four vertebrate species: zebrafish, chick, mouse, and corn snake. We find that the striking differences in segment number between chick, mouse, and snake mainly reflect differences in the rate of the segmentation clock relative to the rate of cell division in the presomitic mesoderm at the tail end of the body; in the zebrafish, however, the relatively small number of body segments is a reflection of a much briefer program of cell divisions in this growth zone and an earlier halting of the segmentation clock. These findings raise fundamental questions about the long-term programming of growth in the embryo - questions that are highly relevant to the problem of growth control in cancer.
Figure 1. TgMAZe - a new tool for clonal analysis of gene function in the zebrafish. Confocal images of ectoderm overlying the eye (A-C) and the tail (D-G) of 5-day transgenic zebrafish embryos containing TgMAZe plus UAS-NICD (activated Notch) constructs. The embryos were heat-shocked 2 days earlier to trigger recombination in a subset of cells. These cells have given rise to clones marked by nuclear-localised RFP (red) and expressing NICD (green). [Images by Russell Collins]
For a list of refereed research papers, see Publications (in navigation on left).
