Development of Behavioral Assays to Assess ASD-like Behaviors in a Drosophila Model

Background: Autism spectrum disorder (ASD) is a genetically diverse disorder characterized by social deficits and repetitive behaviors. Research has focused on molecular mechanisms and phenotypic consequences of ASD-associated genetic mutations using transgenic animal models. The fruit fly (Drosophila melanogaster) is beginning to be used to model genetic mutations associated with neurodevelopmental disorders, including ASD, due to its short gestation period and simple genome. Several assays exist to assess molecular mechanisms in drosophila. However, less work has focused on behavioral assays. The recent use of drosophila in the study of ASD necessitates the development of ASD-relevant behavioral assays in drosophila.

Objectives: We aim to (a) develop a set of behavioral assays and coding schemes to measure ASD-like behavior in drosophila, and (b) estimate reliability, stability, and sensitivity-to-change of the assays. These include assays for stereotyped motor behavior, activity level, and social interaction.

Methods: We used male, transgenic flies expressing either a healthy or mutated copy of the human dopamine transporter (DAT) gene to test our assays. This mutation is a recently discovered rare variant associated with ASD. For the stereotypy and activity assay, flies (9 healthy, 9 mutant) were tested individually in an enclosed arena that permitted locomotion but not flying. The fly’s movements were recorded with a high-speed camera (1000 frames/second) during a baseline period and a challenge period. During the challenge period an ecologically relevant predatory wasp sound was presented to elicit stress-induced effects on locomotor and social behaviors. Trained observers coded the flies’ behavior by segmenting the videos into 150ms intervals and assigning each interval a behavior code (locomotion, idle, or stereotypy). The social assay used the same arena and stimulus presentation; however, in this assay, 4 flies were present in the arena concurrently (24 healthy and 24 mutants in total). Distance to nearest neighbor preceding and following stimulus onset was used as a measure of social response to the predatory sound. Previous work in drosophila has demonstrated that a distancing response to stress is adaptive and socially relevant.

Results: Our assays indicated that rates of stereotypy were low for both groups and did not change in response to the stimulus. Idle behavior was higher in the mutants than in the healthy flies (F = 6.43, p = .022) but was not significantly affected by stimulus onset. Locomotion was lower in mutants than in healthy flies (F = 8.32, p = .011) and demonstrated a trend for a group by time interaction (F = 3.09, p = .059) suggesting altered locomotor response to stress in the mutants. The social assay indicated that at baseline, the groups did not differ in social distance. After stimulus onset, mutants remained closer together than healthy flies (F = 6.57, p = .0297) suggesting a failure of mutant flies to display a socially adaptive response.

Conclusions: Our assays were reliable and sensitive for detecting a variety of ASD-relevant behavioral endpoints. Future work is needed to refine these behavioral assays and examine their effectiveness in a larger variety of ASD-related genetic mutations.