Characterization of sleep and arousal threshold in female Ae. aegypti.

(A-D) Representative images of female Ae. aegypti either in active state or during sleeping. These images illustrate postural differences of the hind legs in active or inactive states, which were recorded via videography under the daytime or nighttime. (E) Quantification of averaged time interval that individual animal required to change postures. Averaged time interval to shift postures was significantly longer during the day (clear bar) than that at night (gray-filled bar). Data were assessed for normality using Shapiro-Wilk test. As the Day group violated the normality assumption (P < 0.05), comparisons between Day and Night groups were analyzed using a two-tailed Mann-Whitney U test: P = 0.0021 and U = 174.50. Individual data points are shown as gray or black for selective bouts during daytime (N=16) or black for bouts during nighttime (N=13). (F) Schematic of modified Drosophila ARousal Threshold (DART) system using for assessment of sleep responsiveness in mosquitoes. The mechanical stimuli were delivered once per hour for one LD cycle, beginning at ZT0, with protocol shown in bottom left box. (G) Behavioral threshold assessment in mosquitoes across a single light-dark cycle (LD, 15 h:9 h) using a modified DART system. To determine the behavioral threshold, mosquitoes were subjected to vibratory stimuli of incrementally increasing intensity [0.6g–1.0g (acceleration)] once per hour throughout one complete LD cycle. For each stimulation event, the duration of the ongoing inactive bout prior to stimulation was recorded. Responses were normalized to the mean response within each stimulus intensity bin, such that values represent proportional changes relative to the individual baseline. The data indicated that mosquitoes exhibiting shorter periods of inactivity (1–9 min) frequently responded to the stimulus, whereas those in longer inactive bouts (≥10 min) were significantly less responsive, which suggested the onset of deeper sleep following approximately 10 minutes of continuous inactivity, thereby validating the use of a 10-minute immobility threshold for defining sleep behaviorally in this context. One-way ANOVA F4,140 = 14.93, (1-5 min) vs. (6-10 min) P = 0.2460; (1-5 min) vs. (11-15 min) P = 0.0451. N=29. (H-I) Individual mosquito trajectories showing increased sleep duration (H, paired t test, P = 0.0478) and averaged bout length (ABL) during nighttime (I, paired t test, P = 0.0041). Error bars represent the standard error of the mean (±SEM). Here and in subsequent figures, asterisks indicate the level of significance: *P < 0.05, **P < 0.01, ***P < 0.001.

Reduced day and night sleep duration in AeCyc−/− mosquitoes.

(A) Schematic of sleep monitoring in wide type (WT) and AeCyc−/− mosquitoes via infrared-based LAM10 (tube outside diameter = 10 mm, length = 100 mm). (B) Activity profile across day and night cycles (15 h:9 h) for WT (gray) and AeCyc−/−(pink) mosquitoes. AeCyc−/− mutants showed elevated activity during both day and night periods compared to WT animals. Light-colored shadows indicate the ±SEM error bar, and white and black boxes indicate the daytime and nighttime, respectively. (C) Two-way ANOVA revealed a significant interaction between genotype and ZT time (F1,70 = 36.56, P < 0.0001). WT mosquitoes showed a sharp decrease in activity from ZT15 to ZT16 (paired t-test, P < 0.0001), whereas AeCyc−/− mutants maintained high activity levels at ZT16 (paired t-test, P = 0.4084). At ZT15, AeCyc−/− activity was significantly lower than WT (unpaired t-test, P = 0.0045, Cohen’s d = 1.00), while at ZT16 AeCyc−/− activity was significantly higher than WT (unpaired t-test, P < 0.0001, Cohen’s d = 1.78), suggesting the dusk activity peak was largely lost in AeCyc−/− mosquitoes. (D) Sleep profile across day and night cycles for WT (gray) and AeCyc−/− (pink) mosquitoes. AeCyc−/−mutants showed decreased sleep during both day and night periods compared to WT animals. (E) Two-way ANOVA revealed significant main effects of genotype (F1,70 = 515.66, P < 0.0001) and time of day (F1,70 = 653.43, P < 0.0001), but no significant interaction (F1,70 = 0.98, P = 0.3261). Both WT and AeCyc−/−mosquitoes slept more during the night than during the day (paired t-test, WT: P < 0.0001; AeCyc−/− : P < 0.0001). However, AeCyc−/− mutants exhibited significantly reduced sleep duration compared to WT controls during both the day (Welch’s t-test, P < 0.0001) and the night (unpaired t-test, P < 0.0001). (F) Two-way ANOVA revealed significant main effects of genotype (F1,70 = 16.23, P = 0.0001) and time of day (F1,70 = 92.29, P < 0.0001), with a marginally significant interaction (F1,70 = 17.35, P < 0.0001). Both WT and AeCyc−/− mosquitoes showed higher P(Wake) during the day than at night (WT: Wilcoxon signed-rank test, P < 0.0001; AeCyc−/−: paired t-test, P = 0.0021). Notably, AeCyc−/− mutants exhibited a significantly elevated P(Wake) during the night compared to WT controls (Mann-Whitney U test, U = 41.00, P < 0.0001), whereas daytime P(Wake) did not differ between genotypes (unpaired t-test, P = 0.9081). These results indicated that AeCyc−/− was specifically required for maintaining low wake probability during the night. (G) Two-way ANOVA revealed significant main effects of genotype (F1,70 = 16.80, P < 0.0001) and time of day (F1,70 = 35.69, P < 0.0001), with a significant interaction (F1,70 = 9.38, P = 0.0031). Both WT and AeCyc−/− mosquitoes showed higher P(Doze) during the night than during the day (WT: paired t-test, P < 0.0001; AeCyc−/−: paired t-test, P = 0.0428). AeCyc-/- mutants also exhibited a significantly reduced P(Doze) during the night compared to WT controls (unpaired t-test, P < 0.0001). Error bars represent the standard error of the mean (±SEM). Data points in the bar graphs represent individual animals; N=20 for WT in gray and N=17 for AeCyc−/− in pink.

Effects of blood feeding on sleep patterns in Ae. aegypti.

(A) Schematic representation of the experimental design. Female mosquitoes, aged 3-5 days post-eclosion (dpe), were either sugar-fed or blood-fed. Following oviposition, individual animals were loaded to the infrared-based activity monitors (LAM10, diameter=10 mm) used for tracking sleep and activity. (B) Sleep profile (min per hour) across the 24 h cycle on Day 4 post-oviposition. Blood-fed mosquitoes (red line, N=31) displayed a consistently elevated sleep profile compared to sugar-fed counterparts (gray line, N=31). Light-colored shadows indicate the ±SEM error bar, and white and black boxes indicate the daytime and nighttime, respectively. (C) Comparison of total sleep of blood-fed mosquitoes (red cycles) on Day 4 post blood meal to sugar-fed controls (gray diamonds). Blood-fed mosquitoes showed a significant increase in total sleep compared to sugar-fed mosquitoes (Unpaired t-test, P < 0.0001). (D) Total sleep duration across Day 4 to 9 post-blood meal. Two-way ANOVA revealed significant effects of time, feeding status, and their interaction (all P < 0.0001). Blood-fed mosquitoes exhibited significantly increased sleep compared to sugar-fed controls on Days 4-6 (Day 4: P < 0.0001; Day 5: P < 0.0001; Day 6: P = 0.021), but not on Days 7-9. (E-G) Changes of sleep architecture of blood-fed mosquitoes (red cycles) on Day 4 post blood meal compared to sugar-fed controls (gray diamonds). Blood-fed mosquitoes demonstrated a significantly lower waking propensity, P(Wake), compared to sugar-fed mosquitoes (E, Mann-Whitney U test = 726, P = 0.0006). Blood-fed mosquitoes exhibited a significantly higher sleeping propensity, P(Doze), compared to sugar-fed controls on Day 4 post-blood feeding (F, Unpaired t-test, P < 0.0001). Error bars represent the standard error of the mean (±SEM). Data points in the bar graphs represent individual animals. Waking activity (activity per waking minute) was reduced on Day 4 post blood meal compared to that of sugar-fed animals, indicating lower movement intensity after blood feeding (G, Unpaired t-test, P = 0.0063).

Effects of blood feeding on sleep in Ae. aegypti females.

(A) Total sleep duration in blood-fed Ae. aegypti versus sugar-fed controls in 30-min time-bin of immobility. Sleep of blood-fed mosquitoes (red cycles) was significantly increased compared to sugar-fed controls (gray diamonds, unpaired t-test, P < 0.0001). (B) Comparisons of the averaged sleep for either sugar-fed or blood-fed individual subjects during daytime and nighttime periods. Two-way ANOVA revealed significant main effects of feeding status (F1,120 = 61.60, P < 0.0001) and time of day (F1,120 = 39.04, P < 0.0001), with no significant interaction (F1,120 = 2.43, P = 0.122). Both sugar-fed and blood-fed mosquitoes slept more during the night than during the day (sugar-fed: paired t-test, P = 0.010; blood-fed: paired t-test, P < 0.0001). Blood-fed mosquitoes showed significantly higher sleep levels during both daytime and nighttime compared to sugar-fed controls (daytime: unpaired t-test, P < 0.0001; nighttime: Mann-Whitney U test, U = 174, P < 0.0001). (C) Daily activity across Days 4 to 9 post-blood meal. Two-way ANOVA revealed significant effects of time, feeding status, and their interaction (all P < 0.0001). Blood-fed mosquitoes (red line) exhibited significantly reduced activity compared to sugar-fed controls (dark gray line) on Days 4 and 5 (Day 4: unpaired t-test, P = 0.0045; Day 5: unpaired t-test, P = 0.0144), but not on Days 6-9 (Day 6: P = 0.337; Day 7: P = 0.304; Day 8: P = 0.431; Day 9: P = 0.431). (D) Total daily activity for sugar-fed and blood-fed mosquitoes on Day 4 post blood meal. Blood-fed mosquitoes displayed significantly lower daily activity levels compared to sugar-fed controls (unpaired t-test, P = 0.0045). (E) Hourly activity profiles across one LD cycle for sugar-fed and blood-fed mosquitoes. Blood-fed mosquitoes showed a distinct reduced activity during both day and night periods. Light-colored shadows indicate the ±SEM error bar, and white and black boxes indicate the daytime and nighttime, respectively. Each dot represents one individual mosquito, dots in gray indicate sugar-fed mosquitoes (N=31) and red dots are for blood-fed animals (N=31). Error bars represent the standard error of the mean (±SEM), and asterisks indicate the level of significance: *P < 0.05, **P < 0.01, ***P < 0.001.

Validation of blood feeding-induced sleep in female Aedes aegypti using EthoVision XT.

(A) Schematic of the EthoVision XT-based behavioral recording assay. Female mosquitoes at 1–3 days post-eclosion (dpe) were provided with either a sugar meal or a blood meal and then individually transferred into a 6-well plate immediately after feeding for continuous behavioral monitoring over one LD cycle. EthoVision XT was used for automated locomotor tracking, trajectory visualization, and sleep analysis. (B) Sleep profiles across the 24 h LD cycle in sugar-fed and blood-fed mosquitoes. Blood-fed females exhibited increased sleep relative to sugar-fed controls during both the light and dark phases. Lines represent mean values, and shaded areas indicate ± SEM. White and black bars denote the light and dark phases, respectively. (C) Total daily sleep duration in sugar-fed and blood-fed mosquitoes. Blood-fed females showed a significant increase in total sleep duration compared with sugar-fed controls (unpaired t-test, P < 0.0001). (D) Representative locomotor trajectories of sugar-fed and blood-fed mosquitoes recorded by EthoVision XT. Red lines indicate movement paths, illustrating reduced activity in blood-fed females. (E) Blood feeding significantly reduced total locomotor distance compared to sugar feeding (unpaired t-test, P = 0.015), reflecting the concomitant increase in sleep behavior. Gray denotes sugar-fed mosquitoes (N=13) and red denotes blood-fed mosquitoes (N=17). Error bars represent ± SEM, and each dot in the bar graphs represents one individual mosquito.

Representative velocity traces and characterization of inactivity/sleep bout duration in sugar-fed and blood-fed female Aedes aegypti.

(A) Velocity traces (per frame) of sugar-fed and blood-fed female mosquitoes recorded by EthoVision XT. Gray and red traces represent sugar-fed and blood-fed females, respectively. The dashed horizontal line indicates the velocity threshold of 0.4 mm/s used to classify behavioral states: velocity > 0.4 mm/s was defined as wake/activity, and velocity < 0.4 mm/s as doze/sleep. Shaded regions indicate the dark phase. The upper panel shows a zoomed-in view of representative time windows at higher resolution. (B) Velocity traces (per second) displayed with rolling smoothing over one second to illustrate locomotor events (wake bouts) in sugar-fed and blood-fed female Aedes aegypti. (C) Distribution of wake/active bouts in sugar-fed and blood-fed mosquitoes across one light/dark (LD) cycle. (D) Number of detected wake/active bouts in sugar-fed and blood-fed mosquitoes using different minimum bout duration thresholds (5 s, 10 s, and 30 s). As expected, the number of detected bouts decreased as the minimum bout duration criterion increased. (E) Blood-fed mosquitoes exhibited a significantly lower probability of wakefulness [P(Wake)] compared to sugar-fed controls (unpaired t-test, P < 0.0001), consistent with their increased total sleep duration (Figure 4C). Behavioral state occupancy was quantified based on a velocity threshold of 0.4 mm/s: periods with velocity > 0.4 mm/s were classified as wake/activity, whereas periods with velocity < 0.4 mm/s were classified as doze/sleep. (F) Blood feeding significantly increased the probability of dozing [P(Doze)] compared to sugar feeding (P < 0.0001), inversely mirroring the reduced P(Wake) shown in panel E.

Dietary protein increases sleep in female Ae. aegypti via BSA feeding.

(A) Schematic overview of the experimental design. Female mosquitoes, aged 3-5 dpe, were either sugar-fed or fed Bovine Serum Albumin (BSA). Following oviposition, sleep was monitored using custom-built infrared activity monitors over a six-day recording period. (B) Comparison of total sleep of BSA-fed mosquitoes (blue squares) on Day 4 post BSA meal to sugar-fed controls (gray diamonds). Total sleep duration of BSA-fed mosquitoes showed a significant increase compared to sugar-fed mosquitoes (Welch’s t-test, P < 0.0001). (C) Sleep distribution across one LD 15 h:9 h cycle on Day 4 post BSA meal. BSA-fed mosquitoes (blue line) demonstrated elevated sleep levels during both day and night periods compared to sugar-fed mosquitoes (gray line). Light-colored shadows indicate the ±SEM error bar, and white and black boxes indicate the daytime and nighttime, respectively. (D) Total sleep duration across Days 4 to 9 post BSA feeding. Two-way ANOVA revealed significant effects of time, feeding status, and their interaction (P = 0.009). BSA-fed mosquitoes (blue) exhibited significantly increased sleep compared to sugar-fed controls (gray) on Days 4 and 5 post BSA feeding (Day 4: Welch’s t-test, P < 0.0001; Day 5: unpaired two-tailed t-test, P = 0.0006), but not on Days 6-9 (Day 6: P = 0.252; Day 7: P = 0.526; Day 8: P = 0.130; Day 9: P = 0.317). (E-G) Changes of sleep architecture of BSA-fed mosquitoes (blue) on Day 4 compared to sugar-fed mosquitoes (gray). Significantly lower P(Wake) (E, unpaired t-test, P < 0.0001) and higher P(Doze) (F, unpaired t-test, P = 0.0003) was showed in BSA-fed population compared to sugar-fed group on Day 4 post-BSA feeding. BSA-fed mosquitoes also exhibited reduced waking activity compared to sugar-fed controls (G, unpaired t-test, P = 0.0082). Each dot in the bar graphs represents one individual mosquito, and blue squares indicate BSA-fed mosquitoes (N=31) and gray diamonds sugar-fed individuals (N=31).

Role of the leucokinin receptor (Lkr) in sleep regulation of female Ae. aegypti.

(A) Schematic overview of the experimental design for sleep monitoring in dsRNA-injected mosquitoes. (B) Sleep profiles of female mosquitoes injected with either dsLKR or dsEGFP on Day 1 post injection. Mosquitoes with Lkr knockdown (pink) displayed a distinct sleep pattern, with increased sleep observed during both daytime and nighttime periods compared to dsEGFP group (green). Light-colored shadows indicate the ±SEM error bar, and white and black boxes indicate the light and dark periods, respectively. (C) Daily sleep duration was significantly induced in dsLKR-injected mosquitoes (pink squares) compared to dsEGFP controls (green diamonds) (unpaired two-tailed t-test, P = 0.0004). (D) Total sleep duration across Days 1 to 6 post-siRNA injection. Mosquitoes injected with dsLKR (red line with pink squares) exhibited significantly increased sleep compared to controls injected with dsEGFP (green line with diamond markers) on Days 1-3 (Day 1: unpaired t-test, , P = 0.0004; Day 2: Welch’s t-test, P = 0.0005; Day 3: Welch’s t-test, P = 0.012), but not on Days 4-6 (Day 4: P = 0.184; Day 5: P = 0.511; Day 6: P = 0.477). (E-G) Comparisons of P(Wake), P(Doze) and waking activity of mosquitoes injected with dsLKR to dsEGFP controls on Day 1 post injection. After knockdown of Lkr, female mosquitoes demonstrated a significantly lower probability of waking compared to dsEGFP injected controls (E, unpaired t-test, P = 0.0011). In contrast, the sleeping propensity, P(Doze), was significantly induced in dsLKR-injected mosquitoes compared to dsEGFP controls (F, unpaired t-test, P = 0.0003). dsLKR injected mosquitoes exhibited significantly reduced waking activity compared to dsEGFP controls (G, unpaired t-test with Welch’s correction, P = 0.0094). Error bars represent the standard error of the mean (±SEM), and asterisks indicate the level of significance: *P < 0.05, **P < 0.01, ***P < 0.001. Each dot represents one individual mosquito, and green diamonds indicate dsEGFP-injected mosquitoes (N=22) and pink squares with cross makers indicate individuals injected with dsLKR (N=22).

Primers used to amplify dsRNA templates Capital letters indicate gene-specific annealing sequences, and lowercase letters indicate the T7 promoter sequence.