Creation of a GAL4/UAS-coupled inducible gene expression system for use in Drosophila cultured cell lines.

Abstract

The usefulness of the GAL4/UAS system as a tool for targeted gene expression in Drosophila melanogaster has been widely recognized (Duffy, 2002, this issue). By combining tissue-specific GAL4 drivers with stably integrated pUAST-based responder constructs, the GAL4/ UAS system (Brand and Perrimon, 1993) transformed gene misexpression into an even more precise and powerful tool for functional analysis. Over the past 30 years, the versatility of Drosophila as a model organism has been extended as well by the development of cultured cell lines from a variety wildtype and mutant fly strains and a variety of tissues within those strains (Cherbas and Cherbas, 1998). Traditionally, regulated gene expression in Drosophila cultured cell lines has been achieved by placing coding sequences of interest under direct control of a constitutive promoter, such as pAct5C from the actin5C gene (e.g., Krasnow et al., 1989), or an inducible promoter, such as pMT from the metallothionein (mt) gene (Bunch et al., 1988), to create “direct-drive” systems. Because endogenous mt gene expression is the result of a systemic response to heavy metal toxicity in Drosophila (Maroni et al., 1987), inducible mt gene expression can be achieved by the introduction of heavy metals, normally copper ion (Cu ) in the form of copper sulfate, into transiently or stably transfected cell lines (Bunch et al., 1988). The first approach to adapting the GAL4/UAS system for use in Drosophila cultured cell lines involved cotransfection of constructs that encoded specific proteins of interest under control of the GAL4-responsive UAS cassette with a constitutive GAL4 driver (e.g., under control of the armadillo promoter, our unpublished results, or the act5C promoter, Johnson et al., 2000) to create a more flexible “indirect-drive” system. We have taken this system another step forward, introducing inducibility by placing GAL4 expression under control of the mt promoter in the widely used vector pRMHa-3 (Bunch et al., 1988; “pMT” hereafter). Through the use of this pMT:GAL4 construct and constructs that encode components of the Drosophila Notch (reviewed in Baron et al., 2002) and EGF receptor (dEGFR1, reviewed in Klambt, 2000) signal transduction pathways, we demonstrate that the GAL4/UAS system can be employed to craft inducible protein expression in Drosophila cell lines. This approach also benefits from the growing library of pUAST responder constructs available for the engineering of regulated gene expression in Drosophila. The use of an inducible GAL4/UAS system in cultured cells will provide control over the levels and timing of expression, allowing for lower expression of proteins that may cause toxicity when constitutively expressed. We used expression constructs encoding the wildtype Delta protein (DeltaWT; Kopczynski et al., 1988) and a constitutively active form of the Notch protein (NotchICD; Fortini et al., 1993) to compare the efficacy of the “indirect-drive” pMT:GAL4/pUAST system we have developed with that of the “direct-drive” pMT expression system (Bunch et al., 1988). The pMT:DeltaWT and pMT:NotchICD constructs have been used extensively in Drosophila cultured cells to study Delta–Notch interactions and downstream signaling events (Fehon et al., 1990; Klueg et al., 1998; Klueg and Muskavitch, 1999; Lieber et al., 2002). The pMT:DeltaWT or pMT: GAL4 and pUAST:DeltaWT constructs were transfected into Drosophila S3 cells, continuously induced, and levels of protein expression were assayed over time by Western blot analysis (Fig. 1). We found that when equimolar amounts of Delta-encoding constructs were used, expression levels driven by pMT:DeltaWT were higher than those driven by pMT:GAL4/pUAST:DeltaWT (Fig. 1). At 2 h postinduction, expression based on pMT: DeltaWT was detected in 4% of the cells and expression based on pMT:GAL4/pUAST:DeltaWT was seen in 2% of cells transiently transfected. By 8 h postinduction, these percentages increased to 11% for pMT:DeltaWT and 4% for pMT:GAL4/pUAST:DeltaWT (unpubl. obs.). Expression based on either construct set was easily detected by immunohistochemistry at 8 h postinduction (Fig. 3A,B). To confirm that the lower levels of expression using pMT:GAL4/pUAST:DeltaWT are inherent characteristics