1) Confirm habituation in visual response: To demonstrate that Drosophilamelanogaster shows a progressive reduction in jump probability across repeated dimming pulses, indicative of habituation, and to distinguish the strength of habituation by the slope of response decline tested with Generalized Linear Mixed Models (GLMMs).

2) Investigate stimulus parameter effects: To examine how interstimulus interval or delay (ISI of 2 s, 3 s or 5s) and light intensity (levels 3 or 4 in arbitrary units) modulate habituation rates.

3) Assess aging’s impact on habituation: To determine whether aging impairs habituation, we tested the expectation that 30-day-old flies maintained at 25˚C would show a slower habituation rate compared to 10-day-old flies, reflecting age-related declines in neural plasticity.

4) Evaluate potassium channel effects on habituation:

To test the hypothesis that eag and Shmutations increase habituation rates compared to cn bw controls due to hyperexcitability from disrupted voltage-gated potassium channels, potentially resulting in a steeper response decrement.

Two mutant stocks of Drosophila melanogaster were used. The first was cinnabarbrown (cn bw), which had white eyes due to the lack of screening pigments in their eyes. This makes them sensitive to a simple light-off stimulus, which triggers their escape jump. The cn bw stock was obtained from Chun-Fang Wu.

The second strain, ether-à-go-go Shaker; cinnabar brown (eag Sh; cn bw) adds potassium channel mutations (eag and Sh) to this background, allowing us to test how hyperexcitability affects habituation in the same visual circuit. The eag Sh; cn bw mutant stock was made by crossing virgin females of eag Sh and males of cinnabar brown(cn bw). Males of the F1 offspring were crossed with fresh stocks of virgin females of eag Sh. F2 offspring were all homozygous for eag and Sh alleles with white-eyes. White-eyed F2 were selected and crossed to establish the eag Sh; cn bw stock.

Flies were kept in standard vials containing a cornmeal-based food medium (Jazz-mix Drosophila food, Fisher Scientific, Hanover Park, Illinois, USA) prepared at a ratio of 347 g per liter of water. Newly emerged flies were collected every 1 to 5 days and kept at densities of up to 20 - 30 flies per vial to prevent overcrowding. Flies were transferred to fresh food vials at 6 to 10-day intervals to ensure optimal health and prevent progeny from appearing. All stocks were maintained in a 25˚C incubator.

Food vials containing a high density of late-stage, mature (dark) pupae were selected. All existing adult flies were removed and discarded, ensuring only pupae remained. The cleared vials were maintained under standard conditions for 5 days to allow adult eclosion, generating a cohort of flies aged 0 - 5 days. These flies were then transferred to fresh food vials and aged for an additional 5 days, resulting in the 10-day-old experimental cohort with a final age range of 5 - 10 days post-eclosion. Using the same initial 0 - 5-day cohort, the 30-day-old group was generated through sequential transfers and extended aging for a further 25 days, resulting in a final age range of 25 - 30 days post-eclosion.

Using an aspirator, a fly was transported into a spherical glass flask of diameter 5 cm with a foam plugged in the truncated neck of the flask. This glass chamber was transported into a testing enclosure to test the startle response. The chamber was inverted (neck down), and placed in an enclosure with dimensions of 57 × 53 cm. The enclosure was covered with heavy black plastic to block out ambient light, that could be opened in the front to allow access and observation. A steady background light was supplied by a classroom-type microscope illuminator (Bausch and Lomb) microscope illuminator positioned 35cm directly above the chamber, shining straight down onto it.

The stimulus light was provided by a fiber-optic lamp (PL900-A, Dolan-Jenner Industries, Inc., Lawrence, Massachusetts, USA). The two ends of the Y-shaped fiber-optic light guide (with detachable lenses removed) were positioned horizontally 5.5 cm from opposite sides of the testing chamber. The lamp’s dimmer knob has six factory-marked indicator dots, where 0 indicates fully off, 3 indicates approximately 50% of maximum output, and 6 indicates 100% output. For the purposes of this experiment, these dial positions are referred to as “3 units” and “4 units” (with “4 units” obtained by rotating the knob approximately two-thirds of the way between the 3 and 6 marks). These intensity levels are arbitrary and instrument-specific; we did not measure absolute light output in lux or irradiance. However, the relative difference between levels is consistent across experiments, and the full lamp configuration (model, distance, and chamber geometry) is provided here to allow approximate replication.

The stimulus consisted of a brief darkening of the fiber illuminator light, achieved by closing a shutter (VMM-T1, Vincent Associates, Rochester, NY, USA) placed between the lamp and the entrance of the fiber optic light guide. Since a constant level of background light was maintained by the microscope illuminator above the test chamber, this reduction in light was referred to as a “dimming pulse.” The shutter control allowed precise adjustment of both the stimulus duration and the time between stimuli (interstimulus interval). The pulse duration of each dimming pulse was set to 20 milliseconds. Responses were observed and recorded using a smartphone video camera (iPhone). Videos were analyzed manually to score jumps as binary outcomes (1 = jump, 0 = no jump).

To assess jump probability, flies were given at least 2 minutes to acclimate to the chamber before a series of stimuli was administered. Each stimulus train included 15 dimming pulses, each lasting 20 milliseconds with varying baseline light intensity and interstimulus interval. To assess how changes in light intensity and interstimulus interval (stimulus frequency) affect habituation results, light intensity was maintained at 3 units and 4 units while interstimulus interval was changed from 2 s, 3 s to 5 s.

For some trials, we quantified two criteria for habituation: spontaneous recovery (the passive, time dependent return of a habituated response after the repetitive stimulus has stopped) and dishabituation (the active restoration of a habituated response by presenting a novel, strong stimulus different from the habituating one). For one set of trials. after the 15-pulse habituation train, the same fly was subjected first to dishabituation, then another habituation trial, then to spontaneous recovery, then to a final habituation trial. Immediately after pulse 15, the fly was subjected to a mechanical disturbance by placing its spherical glass flask in a vortexer for 30 seconds at medium speed setting. This vortexing provided a novel mechanical stimulus. Following vortexing, the flies were immediately returned to the testing chamber and subjected to a second 15-pulse train to assess the magnitude of response restoration. Following this, to test for spontaneous recovery, the same flies were allowed to rest undisturbed for one minute before receiving a final 15-pulse habituation train.

To prevent potential confounding between the dishabituation stimulus and the measurement of spontaneous recovery, the experiment was replicated with an independent-measures design. In this refined protocol, different cohorts of flies were used for each restoration test. One cohort was subjected to a 1-minute dishabituating stimulus (vortexing) followed by re-testing. A separate, time-matched control cohort experienced a 1-minute rest period in the chamber prior to re-testing. This allowed the processes of stimulus-induced recovery and spontaneous recovery to be quantified independently.

For each combination of genotype (cnbw or eagSh;cnbw), age (10-day or 30-day), interstimulus interval (2 s, 3 s, or 5 s), and light intensity (3 or 4 arbitrary units), a separate cohort of 20 flies was tested (n = 20 per condition). Sex was not recorded, but both males and females were present in each cohort. No flies were excluded from the analysis, and every fly included in the study responded at least once during the 15-pulse train.

Flies were never reused across different stimulus conditions. Each individual contributed to only one combination of ISI, intensity, and age. For dishabituation and spontaneous recovery experiments that used an independent-measures design, separate cohorts of n = 10 flies per genotype were used, as indicated in the corresponding figure legends.

Jump probability was assessed by scoring whether a fly jumped (1) or did not jump (0) for each of the 15 dimming pulses in a stimulus train. Statistical analysis was performed using a Generalized Linear Mixed Model (GLMM) implemented with the glmer() function from the lme4 package in R to quantify the rate of habituation (slope) across the 15 stimulus pulses. The model specification was Response ~ Age * as.numeric (Pulse) + (1|FlyID), where the binary response (jump = 1, no jump = 0) was modeled with a binomial distribution and logit link function. The fixed effects were Age (a categorical factor with two levels: 10-day and 30-day) and Pulse number (1 - 15), with their interaction term to test for differences in habituation rates. For trials with K+ channel mutants, genotype replaced age as a fixed effect (all flies were 10 days old in those experiments). A random intercept for individual flies (FlyID) was included to account for within-subject correlations across the 15 repeated trials.

To account for repeated measurements within individual flies, we compared several mixed-effect model structures using the lme4 package in R. The candidate models included: (1) a null model with random intercept only (1|FlyID); (2) a model with random intercept and random slope for Pulse (Pulse|FlyID); and (3) a model with random intercept only but including an autocorrelation structure. All models were fitted with the same fixed effects (Age * as.numeric(Pulse) for age experiments; Genotype * as.numeric(Pulse) for genotype experiments). Models were compared using AICc (corrected for small sample size), BIC, and log-likelihood ratios, as well as model weights derived from AICc. The random-intercept-only model consistently yielded the lowest AICc and BIC values and the highest model weight (AICc weight > 0.95 in all cases). Models with random slopes either failed to converge or produced singular fits with higher AICc values. Therefore, the simpler random-intercept-only structure was retained for all final analyses.

Habituation is frequency dependent, showing that the time between stimuli is a critical factor determining whether habituation can occur. Habituation is stronger at higher stimulus frequency (Table 1).This can be observed in Figure 1(a) when interstimulus interval (ISI) is 2 s (0.5Hz), resulting in steep decline with transformed slope −0.347 and p < 0.001 in young flies. Conversely, at longer interstimulus intervals (lower frequency), such as 5 s (0.2 Hz), the long intervals allow spontaneous recovery for neural circuits to reset, preventing habituation altogether as observed in Figure 3(a) and Figure 3(b). The medium ISI of 3 seconds (0.33 Hz) falls in an intermediate range, where the balance between stimulation frequency and recovery time creates a threshold effect, making habituation dependent on the light intensity.

Light intensity worked as a complex modulator, demonstrating interaction with the frequency of stimulation (Table 1). At the medium ISI of 3 s, habituation was absent at light intensity at 3 units (Figure 2(a)), indicating insufficient stimulus strength to overcome the recovery period and induce cumulative synaptic depression. In contrast, at light intensity at level 4, a significant decline emerged as seen in Figure 2(b), suggesting that higher intensity enhances the neural activation, driving a stronger response to habituation.

Table 1.Summary of habituation success.

Keys: Red = 10 days old flies; Blue = 30 days old flies.

There was a significant increase in jump probability after performing dishabituation process in the habituated response (Table 2). There was a substantial restoration of jump probability in eag sh; cn bwand cn bw (Figure 4). In addition, there was a significant increase in jump response in both eag sh; cn bwand cn bw in the spontaneous recovery test from 0 and 0 at the end of habituation following dishabituation to 0.85 and 0.75 respectively. To avoid potential confounding effect between dishabituation influencing the later spontaneous recovery test we implemented an independent assessment design (Figures 5 and 6). Separate cohorts of 10 habituated flies each were used to assess dishabituation (through mechanical vortex) and spontaneous recovery (rest). In Figure 5, there was a significant increase in jump probability after performing dishabituation process in the habituated response from 0 and 0.2 to 1 and 0.8 respectively. Likewise, in Figure 6, there was a significant increase in jump probability after performing spontaneous recovery process in the habituated response from 0 and 0.1 to 0.7 and 0.5 respectively.

Table 2. Summary of potassium ion channel genotype effect on habituation.

(a)

(b)

(c)

Figure 4.(a)Habituation rate for eag Sh; cn bw and cn bw tested under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 4 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.(b) Dishabituation in eag sh; cn bw and cn bw under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 4 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.(c) Spontaneous recovery in eag Sh; cn bw and cn bw tested under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 3 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.

(a)

(b)

Figure 5.(a) Habituation rate before dishabituation (independent measures) in eag Sh; cn bw and cn bw tested under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 3 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.(b) Dishabituation rate following vortex stimuli (independent measures) in eag Sh; cn bw and cn bw tested under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 3 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.

(a)

(b)

Figure 6.(a) Habituation rate before spontaneous recovery (independent measures) in eag Sh; cn bw and cn bw tested under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 3 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.(b) Spontaneous recovery following rest (independent measures) in eag Sh; cn bw and cn bw tested under the stimulus condition of Interstimulus interval (ISI) = 3 seconds and Light Intensity = 3 units. The probability of the light-off jump response (y-axis) is plotted against the stimulus number (x-axis, Pulse). Fixed effects are interpreted as described in results.

Aging significantly impaired the habituation process, but only under conditions that produced robust habituation,Table 1. Under optimal habituation conditions, interstimulus interval of 2 s and light intensity at level 3 units, 30-day-old flies showed a significantly shallower habituation slope than 10-day-old flies (Age30day:as.numeric) (Pulse = 0.194, p < 0.001; Figure 1(a)). The older flies maintained higher response probabilities throughout the pulse train. Under moderate habituation conditions (ISI 3s, Intensity 4), there was also a significant age-related impairment with old flies habituating more slowly than the young flies (interaction = 0.149, p < 0.001; Figure 2(b)). When young flies showed weak or no habituation, there were no significant age differences in slope or habituation process as seen in Figure 1(b), ISI 2s Intensity 4: p = 0.152; and Figure 3(a) and Figure 3(b) ISI 5s: p > 0.32 and p > 0.57.

The results presented in Figure 4 detail the comparison of habituation process between hyperexcitable mutant eag Sh; cn bw and cn bw. Conceptually, cn bw (the genotype used throughout these experiments) is the control, but to analyze the habituation of eag sh; cn bw, that genotype was taken as the baseline for analyzing these experiments.

All comparisons were performed under the interstimulus interval of 3 seconds and light intensity of 4 units. Mutations in voltage-gated potassium channels significantly accelerated the habituation process, Table 2. The eag Sh;cn bw mutants showed a significantly steeper negative slope than cn bw controls as seen Figure 4(a) and Figure 4(b). Disruption of potassium channels through eag Sh mutations increases neuronal excitability, which in turn may accelerate short-term synaptic depression, leading to faster behavioral habituation. This provides a direct relationship between specific ion channel function, synaptic plasticity mechanisms, and learning behavior.

Our systematic approach confirms that the Drosophila light-off jump response is a valid model for studying short-term, non-associative learning, while also defining the stimulus parameters that optimize its manifestation. It is important to carefully characterize habituation to distinguish the observed reduction in jump probability from non-learning phenomena including sensory adaptation and motor fatigue. We confirmed habituation by demonstrating a statistically significant decline in jump probability across 15 sequential 20 millisecond dimming pulses, quantified using generalized linear mixed models (GLMMs).

To confirm the behavioral decrement was genuine learning, we rigorously assessed response recovery through spontaneous recovery and dishabituation [2]. Recognizing the fundamental risk that the stimulus required for dishabituation affects the measurement of spontaneous recovery, we implemented an independent measures design. The dishabituation test, using mechanical vortexing, resulted in a recovery of the jump response that was significantly higher than the recovery observed in the time-matched spontaneous recovery control group The spontaneous recovery tests the return of the response after a rest period confirmed that the habituation was reversible. This differential recovery suggests that dishabituation actively recruits a separate, arousal-driven system to override the habituation-induced inhibition.

The strength of habituation was profoundly shaped by the characteristics of the stimulus train. Our findings show a specific effective range of stimulation defined by the interaction of stimulus timing and light intensity.

The interstimulus interval (ISI), the time delay between successive dimming pulses, emerged as the primary determinant of habituation success. This parameter influences the frequency of synaptic depression underlying habituation. Longer interstimulus interval allowed recovery between pulses, preventing depression from accumulating [18]. The requirement for interstimulus interval to induce habituation aligns with short-term synaptic depression, the primary neural mechanism for non-associative learning. High interstimulus interval of 2 s leads to depletion of the readily releasable vesicle pool, causing a steep behavioral decline (habituation). At a long interstimulus interval, synaptic vesicles replenishment between pulses prevents any accumulating depression.

Light intensity proved to be a critical modulator of stimulus efficacy. This was clearly demonstrated at the intermediate interstimulus interval of 3 s. At lower intensity (3 units), habituation failed to occur in both age treatment groups. The light intensity strength was insufficient to overcome the recovery period provided by the 3 s interstimulus interval. At high intensity (level 4), the stimulus strength increased. The increased stimulus strength was sufficient to drive the circuit into habituation in both groups. It was only when stimulus strength was adequate that the age-related impairment became apparent with older flies habituating significantly more slowly.

Aging significantly impaired the capacity for habituation, supporting the hypothesis that aging reduces neural plasticity in the Drosophila visual-motor circuit. At shorter ISIs (2 s) and lower intensity (3 units), 30-day-old flies exhibited a shallower habituation slope compared to 10-day-old flies, indicating slower adaptation to repetitive stimuli. Similarly, at ISI 3 s and higher intensity (4), older flies showed reduced habituation. These findings suggest age-related declines in neural plasticity, potentially due to reduced synaptic efficacy, or altered gene expression in learning pathways [13]. In contrast, no significant age effects occurred at longer ISIs (5 s), where habituation was absent overall. Our finding implies that aging reduces habituation, reflecting cognitive declines in higher organisms, including humans, where habituation deficits are link to disorders like Parkinson’s disease [7]. The difficulty of older (30-days) flies to ignore repetitive stimuli mirrors the behavioral difficulties faced by elderly individuals who experience reduced or diminished sensory filtering, leading to increased distractibility and struggle processing information amongst background noise.

The ether-à-go-go(eag) Shaker(Sh) strain showed a significantly accelerated habituation compared to the cnbw control, providing a direct connection between potassium ion channel function, and neuronal excitability.

The accelerated habituation observed in eagSh;cnbw mutants supports the predicted effects of reduced potassium currents. The ether-à-go-go (eag) and Shaker (Sh) genes encode voltage-gated potassium channels responsible for neuronal repolarization [19]. Their combined mutation produces a hyperexcitable process characterized by an increased action potential [10]. Broader action potentials prolong depolarization, leading to excessive calcium influx per stimulus at synaptic release sites. This heightened calcium load drives accelerated depletion of the neurotransmitters of synaptic vesicles, resulting in the steeper behavioral decline we quantified. Our results directly confirm and extend earlier neurogenetic findings in the Drosophilagiant fiber escape pathway [20], demonstrating that mutations in genes encoding specific potassium channels significantly affect habituation.

The findings from this study go beyond the simple confirmation of habituation in Drosophila. They establish a robust, quantifiable model for understanding the interaction between intrinsic (genetic, age) and extrinsic (stimulus parameters) factors in neural plasticity. This work provides a foundational framework for understanding the specific genetic and circuit-level mechanisms underlying short-term memory decay and restoration in a simple model organism.

By systematically varying interstimulus interval and light intensity, we demonstrated that habituation is not a fixed genetic trait, but a dynamic process shaped by sensory input parameters.

The results provide a clear demonstration in Drosophila that aging impairs habituation in a circuit-specific manner. This positions the fly as a powerful model for aging neuroscience, allowing rapid genetic screens for interventions that restore neural plasticity in older animals.

The slower habituation in 30-day-old flies mirrors the sensory filtering deficits seen in human aging, cognitive impairment, and Parkinson’s disease [12]. This assay could be used to test drugs or genetic interventions aimed at preserving or restoring habituation-like plasticity.

The accelerated habituation in eag Sh mutants’ models conditions of neuronal hyperexcitability, such as some forms of epilepsy, autism spectrum disorder, where sensory habituation is often altered [5]. This provides a genetically tractable system to explore how ion channel mutations lead to altered information filtering at the circuit level.