Cell polarity is essential both for cell function and for several key developmental processes, such as cell migration, axis determination and asymmetric cell division, whereas loss of polarity is a critical step in the formation of tumours. We use the Drosophila ovary to analyse how cells become polarised, using a combination of cell-biological, genetic and molecular approaches.

Much of our work uses the oocyte as a model, since the localisation of bicoid and oskar mRNAs to opposite ends of this very large cell defines the anterior-posterior axis of the embryo. We are using proteomic and biochemical approaches to elucidate how conserved polarity proteins regulate the organisation of the cytoskeleton, and we are investigating the mechanisms of mRNA transport by making time-lapse movies of mRNA particles in wildtype and mutant oocytes. We are also performing large scale screens for mutants that affect the localisations of bicoid and oskar mRNAs, and are analysing novel polarity and mRNA localisation factors that these identify.

We are also examining the establishment of apical-basal polarity in epithelial cells using the follicle cells and the adult gut as models. We have recently discovered that the tumour suppressor, LKB1, and the energy sensor, AMPK, are specifically required for epithelial polarity under conditions of energetic stress, revealing the existence of a distinct low energy polarity pathway. We have now identified several other components of this pathway, all of which have also been implicated in cancer. We are therefore performing RNAi screens for new genes that are required for polarity under either high or low energy conditions.

Starvation-dependent tumour formation. Removal of the AMP-dependent protein kinase from clones of follicle cells (marked by the absence of GFP; green) causes the cells to lose their polarity and over-proliferate, resulting in small tumours. This phenotype is only observed under starvation conditions.

Drosophila anterior-posterior axis formation. A stage 10A egg chamber showing the localisation of PAR-6 (red) and PAR-1 (green) to complementary cortical domains in the oocyte. The nuclei are stained in blue. These PAR proteins control the polarity of the microtubule cytoskeleton to define where bicoid and oskar mRNAs are localised

Novel rod-like structures in the epithelial follicle cells. We have performed a large-scale protein trap screen, in which a transposon containing an artificial exon encoding GFP and two protein affinity tags is jumped around the genome. When this inserts in the right frame between protein coding exons of a gene, the endogenous protein is tagged with GFP. The image shows the GFP-tagged protein produced by one these lines, which labels a novel structure in each follicle cell.

The localisation of bicoid mRNA (green) and oskar mRNA (red) in a stage 10A Drosophila oocyte. Bicoid mRNA has been labelled with MS2-GFP and oskar mRNA with RFP-Staufen

Plain English:
Most cells are polarised, with one end being different from the other. For example, a motor neuron receives inputs into dendrites and transmits these signals down an axon for many centimetres to the other end of the cell, where chemical messengers are released to stimulate muscle contraction. Since polarised cells like neurons perform different functions at each end of the cell, they need to localise the proteins that perform these functions, and this is often achieved by localising the mRNAs from which the proteins are translated. One of the best examples of this is provided by the egg of the fruitfly, Drosophila, where the anterior localisation of bicoid mRNA determines where the head will form, and the posterior localisation of oskar mRNA defines where the abdomen develops. Because it is easy to make mutants in Drosophila and the egg is a very large cell, we are using this system to investigate the conserved molecular mechanisms that polarise cells and target mRNAs to the right place. Although this is basic research, it is relevant to several medical problems. For example, most tumour cells lose polarity, and one of the genes we have characterised is mutated in both spontaneous and inherited cancers.

Selected publications:

•Zimaynin VL, Belaya K, Pecreaux J, Gilchrist MJ, Clark A, Davis I and St Johnston D (2008) In vivo imaging of oskar mRNA transport reveals the mechanism of posterior localization. Cell 134. 843-853

• Mirouse V, Christoforou CP, Fritsch C, St Johnston D and Ray R (2009) Dystroglycan and Perlecan provide a basal cue that is required for epithelial polarity during energetic stress. Dev Cell [in press]

• Mirouse V, Swick LS, Kazgan N, St Johnston D and Brenman JE (2007) LKB1 and AMPK maintain epithelial cell polarity under energetic stress. J Cell Biol 77, 387-392

Drosophila oogenesis. A Drosophila ovariole, containing a series of germline cysts (green, BicD) that progress through oogenesis as they move posteriorly. The cysts are born at the anterior of the ovariole, and become surrounded by somatic follicle cells (red, FasIII) as they exit the germarium. Each cyst contains 16 germ cells, and one of these is selected to become the oocyte and accumulates higher levels of BicD protein.