Whole Animal Feeding FLat (WAFFL): a complete and comprehensive validation of a novel, high-throughput fly experimentation system

Abstract

Non-mammalian model organisms have been essential for our understanding of the mechanisms that control development, disease, and physiology, but they are underutilized in pharmacological and toxicological phenotypic screening assays due to their low throughput in comparison with cell-based screens. To increase the utility of using Drosophila melanogaster in screening, we designed the Whole Animal Feeding FLat (WAFFL), a novel, flexible, and complete system for feeding, monitoring, and assaying flies in a high-throughput format. Our 3D printed system is compatible with inexpensive and readily available, commercial 96-well plate consumables and equipment. Experimenters can change the diet at will during the experiment and video record for behavior analysis, enabling precise dosing, measurement of feeding, and analysis of behavior in a 96-well plate format.

Keywords: Drosophila; feeding behavior; high-throughput; screening.

Fig. 1.

WAFFL apparatus. a) Each WAFFL houses flies in a 96-well plate format and provides food through a capillary in the bottom of the well. b) Food is provided in a 96-well plate that sits underneath the WAFFL, allowing up to 96 parallel experiments in a single WAFFL. The top is sealed with a reusable silicone mat to maintain humidity. c) To visualize fly behavior during experiments, an acrylic cover can be substituted for the silicone mat. d) To transfer flies out of the WAFFL while maintaining their well positions on the 96-well plate, a transfer adapter and receiver plate is used.

Fig. 2.

Fly survival in the WAFFL. a) 7 d time course of fly survival in WAFFL with liquid 10:5 food using conventionally reared OregonR flies. Inset depicts enlarged area. $N=311$. b) 48 h time course of fly survival in WAFFL with liquid 10:5 food using Conventionally reared OregonR ($N=456$), Conventionally reared CantonS ($N=264$), and germ-free CantonS ($N=264$). Inset depicts enlarged area to better visualize differential data sets.

Fig. 3.

Fly feeding rate in the WAFFL by food type. A modified EX-Q assay using liquid food in the WAFFL was used to determine fly feeding behavior, comparing 10:5 food ($N=538$), 10% glucose ($N=506$), or $1\times$ PBS ($N=502$). Mann–Whitney test ($P=0.013$) with pairwise Wilcoxon rank sum between each pair. Dunn’s correction for $\alpha$. ***P < 0.001.

Fig. 4.

MUFFIN video recording system. Exploded diagram of the MUFFIN. a) Slotted aluminum frame, b) acrylic covers, c) monitoring chamber, d) acrylic base plate, e) camera sensors and lenses, f) acrylic wall, g) mounting cutout with NIR light (700 m) diffuser, h) printed circuit board, i) 3D printed camera board mount, j) acrylic support, k) support stand-offs, l) 24 RPI modules, m) USB power ports for RPIs, n) 24-port Ethernet switch and a 5-port Ethernet switch, o) output cable, p) fans, q) power entry module, and r) sliding lift door.

Fig. 5.

Fly feeding behavior using the MUFFIN. Plotted smoothed data collected with the MUFFIN, compared females to males, fed flies versus starved, and flies provided a high- versus low-caloric content diet a) Average distance travel within 1 h of recording. Each point along the smoothed trend line is a fly averaged across 1 s. Because $z$-movements are not tracked, the distance is given in relative units normalized to the maximum distance observed in a time step. b) Proportion of time near the food, defined as the proportion of frames for every 2 min of video in which the fly was in the food region circle (as defined in “Materials and methods”). Each point along the smoothed trend line is a fly averaged across 1 s. c) Box plots showing the measurement of relative food content per fly after a 1 h feeding with CDF with sulforhodamine B dye using the WAFFL (as defined in “Materials and methods”).

Fig. 6.

WAFFL overview. a) Flies are treated individually in wells with small amounts of food. b) Flies are monitored while feeding. c) Flies can be macerated for downstream analysis.