View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Developmental Cell 10, 497–508, April, 2006 ª2006 Elsevier Inc. DOI 10.1016/j.devcel.2006.02.004 Systematic Analysis of the Transcriptional Switch Inducing Migration of Border Cells Lodovica Borghese,1 Georgina Fletcher,1 ferent for different situations. To understand how ex- Juliette Mathieu,1 Ann Atzberger,1,3 William C. Eades,2 actly cells become migratory, it is therefore necessary Ross L. Cagan,2 and Pernille Rørth1,* to understand which genes are regulated downstream 1 European Molecular Biology Laboratory of the transcription factors that induce the switch. Meyerhofstrasse 1 Once this is understood for specific, well-characterized 69117 Heidelberg transitions, the patterns can be compared to determine Germany whether there is a common gene expression cassette 2 Washington University School of Medicine that is regulated for cells to become migratory. Alterna- 660 South Euclid Avenue tively, each cell type may employ a different strategy to St. Louis, Missouri 63110 release itself from the constraints of a tissue and move away. So far, a complete expression profile reflecting the transcriptional switch has not been obtained from Summary any controlled cell migration process. While controlled cell migration behavior is useful for Cell migration within a natural context is tightly con- animal development, it also has pathological correlates. trolled, often by specific transcription factors. How- If tumor cells originating in an epithelium become migra- ever, the switch from stationary to migratory behavior tory, this will likely contribute to their ability to metasta- is poorly understood. Border cells perform a spatially size and therefore be dangerous. A large number of and temporally controlled invasive migration during studies have analyzed metastasis-associated gene ex- Drosophila oogenesis. Slbo, a C/EBP family transcrip- pression. This has been done by comparison of cell lines tional activator, is required for them to become migra- with different migratory or metastatic potential, which, if tory. We purified wild-type and slbo mutant border done in vitro, may be problematic (Tatenhorst et al., cells as well as nonmigratory follicle cells and per- 2005). Another, more physiological approach is to com- formed comparative whole-genome expression profil- pare gene expression profiles of groups of metastatic ing, followed by functional tests of the contributions of and nonmetastatic tumors of a particular type (van ’t identified targets to migration. About 300 genes were Veer et al., 2002; Dyrskjot et al., 2003). The latter ap- significantly upregulated in border cells, many depen- proach appears to be very useful for collecting predic- dent on Slbo. Among these, the microtubule regulator tive markers for metastasis, but, given the complexity Stathmin was strongly upregulated and was required of metastasis, it is likely less useful for understanding for normal migration. Actin cytoskeleton regulators how cells become migratory. A more refined analysis were also induced, including, surprisingly, a large to identify genes specifically involved in cell migration cluster of ‘‘muscle-specific’’ genes. We conclude that in this context was done by capturing the cells within Slbo induces multiple cytoskeletal effectors, and that a tumor that are most motile and analyzing their expres- each contributes to the behavioral changes in border sion profiles (Wang et al., 2004). As metastasizing tumor cells. cells may need to reactivate a developmental program normally used by cells to become migratory, studies of normal migratory switches are also likely to contribute Introduction to our understanding of metastasis. To analyze how cells become migratory in response to In complex, multicellular animals, cell migration is a very a normal transcriptional switch, we have analyzed gene tightly controlled process. Specific cells migrate at spe- expression patterns in border cells of the Drosophila cific times during development and, to a lesser extent, in ovary. Border cells are a small group of cells that per- adult animals. While cells of the immune system may be form a well-defined and controlled migration in vivo considered professional migratory cells, most other and have become a useful model for studying invasive cells are only migratory in a specific phase; their migra- cell migration (Montell, 2001; Rørth, 2002). Border cells tory behavior needs to be activated and later inacti- delaminate from the follicular epithelium, invade into vated. Examples are neurons, glial cells, neural crest de- the underlying germline tissue, and migrate to the oo- rivatives, germline cells, as well as certain muscle cells. cyte. One of the key transcription factors inducing If the migratory cells arise in an epithelium, the cells may migration of border cells is Slbo (Drosophila C/EBP). Ex- undergo an epithelial-to-mesenchymal transition to be- pression of Slbo requires a spatial signal (unpaired) from come migratory. Like other cell fate changes, induction the preexisting anterior polar cells to induce activation of migratory behavior can be triggered by specific sig- of the JAK/STAT pathway in the prospective outer bor- nals and transcription factors and is associated with der cells. The anterior polar cells will become the central changes in gene expression (Birchmeier and Brohmann, two cells of the border cell cluster. Together with tempo- 2000; Gammill and Bronner-Fraser, 2002; Montell, 2001; ral signals, this leads to the determination of border Rørth, 2002). The transcription factors are, however, dif- cells, and thus specific expression of Slbo, at the ante- rior tip of an egg chamber at a specific time. The effect *Correspondence: [email protected] of Slbo, in turn, appears to be relatively direct, as the 3 Present address: Weatherall Institute of Molecular Medicine, John border cells migrate a few hours after Slbo is expressed. Radcliffe Hospital, Headington, OX39DS, United Kingdom. Slbo expression is essential for border cell migration. Developmental Cell 498 Transcriptional Switch Inducing Cell Migration 499 Slbo is also expressed in centripetal cells at a slightly A protocol for fluorescence-activated cell sorting later stage, but it is not required for their migration. To (FACS) had been established for isolating GFP-labeled understand how border cells become migratory in re- follicle cells (Bryant et al., 1999). We modified the proto- sponse to Slbo, we need to know which genes are reg- col to minimize handling time, but we found in either ulated by Slbo in border cells. In order to determine case that border cells were not recovered from the this, we have carried out genome-wide transcription FACS. Visual inspection of the sample showed that profiling of border cells purified directly from ovaries. while follicle cells were dissociated to single cells, bor- We compared wild-type border cells with their slbo mu- der cells tended to remain as a cluster (Figure 1D), sug- tant counterparts and also compared border cells with gesting that adhesion between border cells is particu- neighboring follicle cells of the same developmental larly strong. By modifying the FACS sorting such that stages. Further functional studies with mutations in the pressure on the cells was reduced, we were able these genes identified a number of genes that affect to purify these rare border cell clusters to high purity the process, but they have also revealed that the pro- without disrupting them (Figures 1E and 1F). Normal cess is fairly robust. and slbo mutant border cells were purified in this man- ner, as were the follicle cells for comparison (Figure 1G). Results and Discussion Sorted cells from multiple samples were pooled and RNA was extracted (Figure 1H) and subjected to linear Gene Expression Profiling of Border Cells mRNA amplification (Figure 1I) and labeling for hybrid- and Validation ization to Drosophila Affymetrix arrays. One technical Border cells arise from the follicular epithelium at stage replicate, using the same border cell total RNA pool as 8/9 of oogenesis and then delaminate and migrate. In or- starting material, demonstrated the reproducibility of der to identify genes specifically expressed in this small the method (Figure 1J). Multiple biological replicates (in- cluster of invasive cells, we decided to compare the dependent pools of collected material) allowed us to gene expression profile of border cells to that of all other identify which expression changes could truly be con- follicle cells from the same stages. This was done to en- sidered cell type and genotype dependent (Figures 1K sure that the cells compared were from the same envi- and 1L; see tables in the Supplemental Data available ronment and of the same general maturation stage. with this article online). In total, about 5000 genes were For example, all of these cells are postmitotic. Border reproducibly detected on the array. The remaining cells were labeled in vivo by expression of GFP with a 9000 genes represented on the array may not be detect- specific GAL4 ‘‘driver’’ called c522 that drives expres- ably expressed in this cell type. In addition, some genes sion only in border cells (Figure 1A). The GAL4 driver may have been missed if the probe set was not optimal. c323 similarly allowed labeling of the other follicle cells The array results were externally validated as follows: (Figure 1C). To identify genes regulated by Slbo in bor- we selected 20 genes with different levels