Drosophila ATF-2 regulates sleep and locomotor activity in pacemaker neurons

Abstract

Stress-activated protein kinases such as p38 regulate the activity of transcription factor ATF-2. However, the physiological role of ATF-2, especially in the brain, is unknown. Here, we found that Drosophila melanogaster ATF-2 (dATF-2) is expressed in large ventral lateral neurons (l-LN(v)s) and also, to a much lesser extent, in small ventral lateral neurons, the pacemaker neurons. Only l-LN(v)s were stained with the antibody that specifically recognizes phosphorylated dATF-2, suggesting that dATF-2 is activated specifically in l-LN(v)s. The knockdown of dATF-2 in pacemaker neurons using RNA interference decreased sleep time, whereas the ectopic expression of dATF-2 increased sleep time. dATF-2 knockdown decreased the length of sleep bouts but not the number of bouts. The ATF-2 level also affected the sleep rebound after sleep deprivation and the arousal threshold. dATF-2 negatively regulated locomotor activity, although it did not affect the circadian locomotor rhythm. The degree of dATF-2 phosphorylation was greater in the morning than at night and was enhanced by forced locomotion via the dp38 pathway. Thus, dATF-2 is activated by the locomotor while it increases sleep, suggesting a role for dATF-2 as a regulator to connect sleep with locomotion.

FIG. 1.

Expression of dATF-2 in the pacemaker neurons. (A) dATF-2 is expressed predominantly in l-LN_vs and, to a much lesser extent, in s-LN_vs. Fly brains expressing GFPnls (green) using the pdf-GAL4 driver were stained with anti-dATF-2 antibody (red). The two images showing GFPnls and dATF-2 are merged in the right panel. Arrows and arrowheads indicate l- and s-LN_vs, respectively. Scale bar, 100 μm. (B) dATF-2 expression in the brain. The wider region of the brain stained with anti-dATF-2 antibody (red) is shown. The region indicated in panel A is surrounded by a square. Punctate staining, which might be neurites extending from LN_vs, are shown by double arrowheads. The asterisk indicates a nonspecific signal that also was obtained using preimmune serum. Scale bar, 300 μm. (C) Preparation of the P-dATF-2-specific antibody. S2 cells were transfected with the Flag-dATF-2 expression plasmid and treated with or without sorbitol. Whole-cell lysates were prepared and analyzed by Western blotting using anti-Flag or anti-P-dATF-2 antibody. IB, immunoblot. (D) Phosphorylated dATF-2 is expressed only in l-LN_vs. Flies expressing GFPnls (green) using the pdf-GAL4 driver were maintained on a 12-h LD cycle. At ZT5, the brains were stained with anti-P-dATF-2 antibody and analyzed using confocal microscopy. Arrows and arrowheads indicate l-LN_vs and s-LN_vs, respectively. Scale bar, 50 μm. (E) P-dATF-2 is not expressed in s-LN_vs of larval brain. The larval brains expressing GFPnls (green) using the pdf-GAL4 driver were stained with anti-dATF-2 (upper) or anti-P-dATF-2 (lower) antibody and analyzed as described above.

FIG. 2.

dATF-2 is not expressed in mushroom bodies. Fly brains that expressed GFPnls (green) in mushroom bodies using the c253 driver were stained with anti-dATF-2 antibody (red). The same confocal microscopy optical sections are shown. The two images showing GFPnls and dATF-2 are merged in the right panel.

FIG. 3.

Up- and downregulation of dATF-2 in l-LN_vs. (A) Analysis of the brains of dATF-2 knockdown flies, in which dATF-2 dsRNA was expressed using the pdf-GAL4 driver, and of dATF-2-overexpressing flies. Genotypes are as follows. dATF-2 knockdown flies (pdf>IR), pdf-GAL4/+ UAS-dATF-2IR/+; dATF-2-overexpressing flies (pdf>dATF-2WT), pdf-GAL4/+ UAS-dATF-2/+; control flies (pdf-G/+), pdf-GAL4/+. Arrows indicate l-LN_vs. Scale bar, 50 μm. (B) Signal intensities from immunostaining images were quantified, and the average of the total intensity of 10 to 16 independent hemispheres was obtained. The values of the dATF-2 RNAi line and the dATF-2 overexpression line are indicated relative to that of the control lines by a bar graph, with SEM. ***, P < 0.001.

FIG. 4.

dATF-2 regulates locomotor activity and sleep. (A) Increased locomotor activity of dATF-2 knockdown flies. The locomotor activity records of male control (pdf-G/+), dATF-2 knockdown (pdf>IR), and dATF-2-overexpressing flies (pdf>dATF-2WT) are shown. Adult flies were entrained to cycles of 12-h LD, and their locomotor activity was measured for 3 days in 12-h LD. The white and black horizontal bars above the panels indicate light and dark periods, respectively. The curves connect mean values ± SEM (pdf-G/+, n = 29; pdf>IR, n = 25; pdf>dATF-2WT, n = 31). Asterisks indicate the values that are statistically different from those of control flies. *, P < 0.05. On the right, the locomotor activity of three types of flies is shown with SEM. ***, P < 0.001. (B) Decreased sleep of dATF-2 knockdown flies. Daily time course (30-min intervals) of the amount of sleep in male control (pdf-G/+), dATF-2 knockdown (pdf>IR), and dATF-2-overexpressing flies (pdf>dATF-2WT). Curves connect mean values ± SEM (pdf-G/+, n = 29; pdf>IR, n = 25; pdf>dATF-2WT, n = 31). *, P < 0.05. (C) Shortened sleep bout duration of dATF-2 knockdown flies. (Left) Sleep bout number per 24 h in 12-h LD is indicated, with SEM. (Right) Sleeping episodes during a 24-h period were categorized based on their duration, and the total amount of sleep in each category per day is indicated, with SEM. *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

FIG. 5.

dATF-2 levels affect sleep in DD conditions. (A) Daily time course (30-min intervals) of the amount of sleep, defined here as behavioral immobility lasting 5 min or more, in male control (pdf-G/+), dATF-2 knockdown (pdf>IR), and dATF-2-overexpressing flies (pdf>dATF-2WT). Adult flies were entrained to cycles of 12-h LD, and their locomotor activity then was measured in DD for 3 days. Curves connect the mean values ± SEM. (pdf-G/+, n = 29; pdf>IR, n = 25; pdf>dATF-2WT, n = 31). *, P < 0.05. (B and C) Daily time course (30-min intervals) of the amount of sleep, defined here as behavioral immobility lasting 30 min or more. Experiments were performed as described above in 12-h LD (B) or DD (C).

FIG. 6.

Phosphorylation of dATF-2 in l-LN_vs shows a diurnal rhythm and is enhanced by forced locomotion. (A) The diurnal rhythm of dATF-2 phosphorylation in l-LN_vs. Flies expressing GFPnls (green) using the pdf-GAL4 driver were maintained on a 12-h LD cycle, indicated by the white and black horizontal bars, respectively, below the panel on the right. At various times, the brains were stained with anti-P-dATF-2 or anti-dATF-2 antibodies and analyzed using confocal microscopy. Typical staining patterns at ZT5 and ZT17 are shown on the left. Arrows indicate l-LN_vs. Scale bar, 50 μm. The signals of P-dATF-2 and dATF-2 in l-LN_vs at various times were measured (data not shown), and the ratios are indicated by bar graphs on the right. (B) Forced locomotion enhances the phosphorylation of dATF-2. The amount of P-dATF-2 in l-LN_vs at ZT8 or ZT20 was measured as described above using flies that had 1 h of forced locomotion or no forced locomotion. Typical staining patterns at ZT8 are shown on the left, and the amount of P-dATF-2 in l-LN_vs is indicated as the value relative to that of flies without forced locomotion by the bar graph (with SEM) on the right. Arrows and arrowheads indicate l-LN_vs and s-LN_vs, respectively. Scale bar, 50 μm. (C) Effect of different lengths of forced locomotion on the phosphorylation of dATF-2. The amount of P-dATF-2 in l-LN_vs at ZT10, when flies are relatively active and have small amounts of sleep, was measured using flies that had different lengths of forced locomotion.

FIG. 7.

Role of dATF-2 in sleep homeostasis and arousal threshold. (A and B) Increased sleep rebound response after sleep deprivation in flies overexpressing dATF-2. Sleep in male control (pdf-G/+), dATF-2 knockdown (pdf>IR), and dATF-2-overexpressing flies (pdf>dATF-2WT) early in the morning after a 12-h sleep deprivation (ZT12 to ∼ZT0) (red) or before sleep deprivation (gray) is shown with SEM (30-min intervals). (A) Curves connect mean values ± SEM (n = 32 for each). (B) The amount of sleep gains for 2 h immediately after sleep deprivation is shown, with SEM. *, P < 0.05. (C) Alterations in the arousal threshold depending on dATF-2 expression. Three types of flies (n = 32 of each) were treated with different intensities of vibration once (1 s at ZT19) during the night, and the number of flies that did not resume sleep behavior (a rest of more than 5 min) until 30 min after vibration is indicated with SEM. Since all flies tested were exhibiting sleep behavior before vibration, the number of flies that had no sleep after vibration is correlated with the arousal threshold. *, P < 0.05; **, P < 0.01.

FIG. 8.

Phosphorylation of dATF-2 is regulated by the dp38 pathway. (A) dp38 pathway mutants show decreased dATF-2 phosphorylation. The amount of P-dATF-2 in l-LN_vs at ZT5 was measured as described in the legend to Fig. 3 using flies with the indicated genotype. The amounts of P-dATF-2 are indicated as values relative to that of wild-type flies (with SEM) in the bar graph. ***, P < 0.001. (B) Effect of dp38 and Mekk1 on the phosphorylation of dATF-2. The amount of P-dATF-2 in l-LN_vs at ZT5 was measured as described in the legend to Fig. 2 using flies expressing dMekk1 or the dominant-negative (DN) form of dp38b or Basket using the pdf-GAL4 driver. The amounts of P-dATF-2 are indicated as the values relative to that of control flies (pdf>Gal4), with SEM, in the bar graph. *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

FIG. 9.

Transcriptional activation of the tim gene by dATF-2. (A) Analysis of the tim promoter. A reporter plasmid in which the mutated tim promoter region was linked to the luciferase gene was transfected into S2 cells together with the dATF-2 expression plasmid or a control plasmid. The luciferase activity was measured. The reporter plasmids used are shown schematically on the left. X indicates the mutated sites containing CRE-like sequences, in which the sequence matched with the consensus CRE is indicated by boldface letters. The ratio of luciferase activity in the presence of dATF-2 to the activity without dATF-2 is indicated as the change (n-fold) in activation. The averages and standard deviations from three experiments are indicated. (B) dATF-2 activates the tim promoter together with dClk. A reporter plasmid in which the 1.9-kb tim promoter region was linked to the luciferase gene was transfected into S2 cells together with the dATF-2 expression plasmid (0, 0.5, or 1.0 μg) and the indicated amount (+, 50 ng; ++, 100 ng) of the dClk expression plasmid or control plasmid. The luciferase activity was measured. The averages and standard deviations from three experiments are indicated. (C and D) Decreased level of TIM in dATF-2 knockdown flies. Control, dATF-2 knockdown, and dATF-2-overexpressing flies, all of which expressed GFPnls (green) using the pdf-GAL4 driver, were maintained on a 12-h LD cycle. At various times, fly brains were stained with anti-TIM antibody and analyzed using confocal microscopy. (C) Typical staining patterns at ZT2 are shown. Arrows indicate l-LN_vs. Scale bar, 50 μm. (D) TIM signals in l-LN_vs at various times were measured, and the average of the total intensity of 10 to 16 independent hemispheres is indicated using a bar graph, with SEM. Asterisks indicate values that are statistically different from those for control flies at each time point. **, P < 0.01; ***, P < 0.001.

FIG. 10.

Levels of TIM in dATF-2 knockdown flies. Fly brains were stained with anti-TIM antibody as described in the legend to Fig. 8B. Typical staining patterns at ZT8, ZT14, and ZT20 are shown. Arrows indicate l-LN_vs. Scale bar, 50 μm.