Taste Preference Assay: A Method for Measuring Feeding Behavior in Drosophila

This protocol is an excerpt from Bantel and Tessier, Taste Preference Assay for Adult Drosophila, J. Vis. Exp. (2016).

1. Starvation

1. Prepare fly starvation vials by saturating a cotton ball with 18.2 MΩ water at the bottom of a standard fly vial. Alternatively, similarly saturate a small strip of filter paper with 18.2 MΩ water and place at an angle within the vial.
2. Collect flies into sets of ~100 animals on a CO₂ pad and then add the flies to a prepared vial.
   NOTE: Best results are obtained with animals that are less than 5 days old. However, the exact age of the animals can be controlled as an experimental variable to determine changes in taste preference over time.
3. Use a cotton ball or foam stopper to secure the vials closed. Place vials on their side in an environmentally controlled incubator. Maintain the temperature at 25 °C, and the humidity above 70%. Leave vials untouched for 24 hr.

2. Taste Preference Assay

1. Prepare all tastants for the assay on the same day as testing.
   NOTE: The exact tastants to be used will vary depending on experimental question being asked. The following are example tastants used in this protocol. See section 4 for optimizations.
   1. Prepare control tastant (1 mM sucrose) by combining 10 µl of 100 mM sucrose solution, 13 µl of red food coloring, and 977 µl of 18.2 MΩ water.
   2. Prepare experimental tastant (5 mM sucrose) by combining 50 µl of 100 mM sucrose solution, 10 µl of blue food coloring, and 940 µl of 18.2 MΩ water.
2. Make assay chambers using a standard 100 mm x 15 mm plastic petri dish prepared in the following manner:
   1. Place three 10 µl drops of control tastant nearest the edge of the plate at 12 o'clock and another 3 drops at 6 o'clock. Ensure that the spacing between drops is similar.
   2. Place three 10 µl drops of experimental tastant nearest the edge of the plate at 3 o'clock and another 3 drops at 9 o'clock. Ensure that the spacing between drops is similar.
   3. Repeat steps 2.2.1 and 2.2.2 for as many replicates as desired.
3. Empty 1 vial of ~100 starved flies onto a CO₂ pad just long enough to anesthetize all animals (approximately 10 sec). Brush the animals into the middle of a prepared assay chamber and cover with the dish lid.
   NOTE: Longer periods of CO₂ exposure should be avoided to improve recovery time and limit interference with the feeding behavior. Exposure to ice (~5 min) may be used for anesthetizing to avoid CO₂ behavioral effects that may arise from even limited exposure.
4. Place the assay chamber in an opaque cardboard box. Be sure to label the outside of the box with the condition and genotype being tested.
5. Place the entire setup (assay chamber contained within cardboard box from step 2.4) into a 25 °C incubator with at least 70% humidity for 2 hr.
6. Repeat steps 2.3 through 2.5 for all replicates.
7. After 2 hours, place the assay chambers, still contained within cardboard boxes, directly into a -20 °C freezer until ready for quantitation.

3. Taste Preference Assay Quantification

1. Allow a single assay chamber to warm up to room temperature (approximately 5 min).
2. Under a dissection microscope, using a brush or pair of forceps, group animals based on the color of their abdomen: red, blue, purple or clear (Figure 1).
3. Record the number of animals in each grouping. Consider clear animals to have not participated in the assay and therefore do not include them in any calculations.
4. Calculate the preference index according to one of the following equations:
   1. If the experimental tastant of interest is added to the red dye, then use (N_red + 0.5N_purple)/(N_red + N_blue + N_purple).
   2. If the experimental tastant is added to the blue dye, then adjust the equation to (N_blue + 0.5N_purple)/(N_blue + N_red + N_purple).
5. Repeat the calculations for all experimental conditions and replicates.

4. Optimization of Taste Preference Assay

1. Empirically determine the concentration of food coloring indicators to be used so food coloring does not affect the outcome of the taste assay, as follows:
   1. Prepare 4 tastants using the same base compound (e.g. 5 mM sucrose) as indicated in step 2.1, but omit the food coloring.
   2. Add 1.3% red food coloring to one of the tastants. Make the remaining 3 tastants with blue food coloring of varying concentrations in each tube (e.g. 0.6%, 1%, and 1.3%).
   3. Complete protocol steps 2.2 through 3.4 for each tastant pair: 1.3% red vs. 0.6% blue; 1.3% red vs. 1% blue and 1.3% red vs. 1.3% blue.
   4. Repeat step 4.1.1-4.1.3 with different percentages of blue food coloring until the preference index averages a value of 0 (Figure 2).
      NOTE: As a starting point, 1.3% red food coloring coupled with 1% blue food coloring typically yields good results. If no satisfactory concentration of blue food coloring can be matched to 1.3% dye, then step 4.1.1 through 4.1.3 can be repeated with varying concentrations of red coloring and a constant concentration of blue food coloring.
   5. Analyze all conditions to be tested with the same optimized food coloring concentrations.

Figure 1: Taste preference assay results. Some examples in the variation of abdominal coloring are shown. Dark red ingested (A). Light red ingested (B). Dark blue ingested (C). Light blue ingested (D). Purple abdomens are considered when the entire coloration appears purple (E), or when distinct regions of the abdomen show portions of red (arrowhead) and separate portions of blue (arrow) (F). Please click here to view a larger version of this figure.

Figure 2: Controlling for food coloration effects. The addition of food coloring to the tastants should not have any effect on the animals' taste preference. Varying the concentration of blue dye while maintaining a constant concentration of red dye revealed an optimum combination of 1.3% red to 1.0% blue. This is indicated by a preference index value near 0.5. Values are the mean ± standard deviation. *p < 0.05, ***p < 0.001 from two-sided student's t-test. Please click here to view a larger version of this figure.