Development of the bi-partite Gal4-UAS system in the African malaria mosquito, Anopheles gambiae

Abstract

Functional genetic analysis in Anopheles gambiae would be greatly improved by the development of a binary expression system, which would allow the more rapid and flexible characterisation of genes influencing disease transmission, including those involved in insecticide resistance, parasite interaction, host and mate seeking behaviour. The Gal4-UAS system, widely used in Drosophila melanogaster functional genetics, has been significantly modified to achieve robust application in several different species. Towards this end, previous work generated a series of modified Gal4 constructs that were up to 20 fold more active than the native gene in An. gambiae cells. To examine the Gal4-UAS system in vivo, transgenic An. gambiae driver lines carrying a modified Gal4 gene under the control of the carboxypeptidase promoter, and responder lines carrying UAS regulated luciferase and eYFP reporter genes have been created. Crossing of the Gal4 and UAS lines resulted in progeny that expressed both reporters in the expected midgut specific pattern. Although there was minor variation in reporter gene activity between the different crosses examined, the tissue specific expression pattern was consistent regardless of the genomic location of the transgene cassettes. The results show that the modified Gal4-UAS system can be used to successfully activate expression of transgenes in a robust and tissue specific manner in Anopheles gambiae. The midgut driver and dual reporter responder constructs are the first to be developed and tested successfully in transgenic An. gambiae and provide the basis for further advancement of the system in this and other insect species.

Figure 1. Gal4 driver and UAS responder constructs and eYFP nuclear localization.

Illustrations of (A) the bi-partite Gal4-UAS system and (B) the Gal4 driver and UAS responder constructs used for transformation of An. gambiae mosquitoes. The driver cassette (upper) consists of the An. gambiae CP promoter upstream of the 147aa DNA binding domain from Gal4 fused in frame to three VP16 minimal activation domains, contained within a piggyBac vector carrying the dsRed marker gene under control of the 3×P3 promoter. Grey arrows show piggyBac repeats. The responder cassette (lower) consists of UAS regions upstream of both LUC and eYFPnls, and flanked by gypsy insulator sequences. These are contained within a piggyBac vector marked with eCFP under control of the 3×P3 promoter. (C to E) An. gambiae cell line co-transfection with Gal4 driver (pSL*LRIM1-Gal4Δ), responder (pSL*UAS-eYFPnls-g) and a control plasmid expressing cytoplasmic dsRed under the Drosophila Actin5C promoter. (C) Localisation of eYFP to the cell nucleus observed with fluorescein filter and bright field (D) Cytoplasmic expression of dsRed observed with a rhodamine filter; (E) A composite image of C and D. Scale bars are 5 µM.

Figure 2. Expression of fluorescent protein in the midgut.

Images A to F show a representative image of eYFP expression in the female midgut of driver lines Cln, Drt, Dgl, F, G and Ivr respectively crossed with the responder line Mbl, photographed through a GFP-B filter. (G) Midgut eYFP expression at 100× magnification photographed through a fluorescein filter. Images H to J are confocal microscope images taken with YFP filtering (H), transmitted light (I), and merged (I) demonstrating localization to midgut nuclei. All guts are from sugarfed mosquitoes Scale bar is 10 µm.

Figure 3. Luciferase activity in midguts and carcasses of female mosquitoes.

Mean luciferase activity from dissected midguts and remaining carcass is shown for progeny heterozygous for Gal4 and UAS cassettes (RLU = Relative Light Units, N = 6). All samples are from sugarfed mosquitoes. Controls for both responder lines are homozygous for UAS cassette (N = 3), and the control driver lines F and Dgl are homozygous for Gal4 cassette (N = 6). The responder line (W = Wnd, M = Mbl), the tissue examined (gut = midgut, car = carcass) and the driver line utilized in each cross and in the controls, are indicated along the X-axis. Error bars show standard errors. Significance differences (Mann Whitney, p<0.05) in midgut specific expression between crosses involving different driver lines but the same responder line are indicated by letters above each bar. u* and x* indicates significant difference between progeny from these crosses compared to all others. Other letters marked * (ie v*, w*, y*, z*) indicate significant difference only to those crosses bearing the same letter (ie v* is significantly different to v).

Figure 4. Luciferase activity in male and female mosquitoes.

Mean luciferase activity in whole male and female sugarfed mosquitoes for all crosses between six driver lines and two responder lines (RLU = Relative Light Units). Driver lines, responder lines and sex are indicated along the X axis. All mosquitoes were heterozygous for Gal4 and UAS cassettes (N = 6). Error bars show standard errors. There are significant differences between male and female luciferase activities in all crosses (Mann Whitney, p<0.01). Luciferase activities in males from crosses involving F and Ivr driver lines are significantly different to those males from all other driver line crosses (Mann Whitney, p<0.05).

Figure 5. Semi-quantitative RT-PCR of midgut transgene transcription in response to blood feeding.

RT-PCR of RNA from female mosquito midguts heterozygous for the Gal4 and UAS transgene cassettes. PCR reactions from two RNA preparations are shown for each timepoint after bloodfeeding. Results from unfed (0 hour), 3, 6 and 24 hour timepoints are shown plus a cDNA free negative control (−). A DNA ladder is shown with sizes indicated (M). Two independent amplicons were amplified for the native carboxypeptidase gene (CP), the Gal4 transgene (Gal4) and the luciferase transgene (LUC). Ribosomal S60 (rS60) was used to standardize for RNA quantity and a DNA contamination control was performed using rS60 primers (-RT).