Cycles of circadian illuminance are sufficient to entrain and maintain circadian locomotor rhythms in Drosophila

Abstract

Light at night disrupts the circadian clock and causes serious health problems in the modern world. Here, we show that newly developed four-package light-emitting diodes (LEDs) can provide harmless lighting at night. To quantify the effects of light on the circadian clock, we employed the concept of circadian illuminance (CIL). CIL represents the amount of light weighted toward the wavelengths to which the circadian clock is most sensitive, whereas visual illuminance (VIL) represents the total amount of visible light. Exposure to 12 h:12 h cycles of white LED light with high and low CIL values but a constant VIL value (conditions hereafter referred to as CH/CL) can entrain behavioral and molecular circadian rhythms in flies. Moreover, flies re-entrain to phase shift in the CH/CL cycle. Core-clock proteins are required for the rhythmic behaviors seen with this LED lighting scheme. Taken together, this study provides a guide for designing healthful white LED lights for use at night, and proposes the use of the CIL value for estimating the harmful effects of any light source on organismal health.

Figure 1. White LEDs with high circadian illuminance (CIL) abolish circadian rhythm more rapidly than LEDs with low CIL.

(A) Schematic illustration of color-tunable and high-color quality four-package white LEDs using B semiconductor-type LEDs and long-wavelength pass dichroic filter (LPDF)-capped monochromatic green, amber, and red phosphor-converted LEDs. (B) Spectral power distribution of CH (cool white, 10,000 K), CL (warm white, 2,000 K), and amber LEDs are shown. (C to E) Locomotor activity of w¹¹¹⁸ flies under different light conditions. w¹¹¹⁸ flies were entrained by daily cycles of 12 h 10,000 K light followed by 12 h complete darkness (LD). After 4 days of entrainment, flies were exposed to constant light (LL) from the three kinds of LED. Light blue, pink, and amber horizontal bars indicate 10,000 K, 2,000 K, and amber LED lights whose qualities are shown in (B). Black horizontal bars indicate complete darkness. Each vertical bar represents average relative activity during a 30-min bin. During the LD cycles, light gray vertical bars represent activity during lights-on periods and black vertical bars represent activity during lights-off periods, whereas during LL conditions, light gray vertical bars represent locomotor activity during subjective day and dark gray vertical bars represent activity during subjective night. ZT, zeitgeber time; ZT0, lights-on time; CT, circadian time; CT0, subjective lights-on time. (F) Morning anticipation values and (G) entrainment indices were calculated for the three different LED light conditions. The numbers of flies used for the analyses shown in (C–G) are 31, 30, and 31 for 10,000 K, 2,000 K, and amber, respectively. Error bars denote standard error of the mean (SEM). Data were compared by pairwise Student’s t-test. *p < 0.05, ***p < 0.0001. (H,I) Flies were exposed to constant CH light or CL light, and heads were collected at indicated time points. Protein extracts were analyzed by immunoblotting with anti-CRY (Rb2) antibody. Note that CRY levels were higher in CL light than in CH light at every time point tested. n.s. denotes for non specific band.

Figure 2. 12 h:12 h cycles of white LED light with high and low circadian illuminance (CIL) produce robust behavioral and molecular rhythms.

Flies were exposed to conventional constant light (LL) conditions for 2 days and then were maintained on 12 h:12 h cycles of 10,000 K white LED light with high CIL (CH) and 2,000 K white LED light with low CIL (CL) for 11 days. (A and B) Averaged actograms of flies of the indicated genotypes are shown. Each row of the actogram was double-plotted. The numbers of flies used for the analyses were 61 (CS), 60 (w¹¹¹⁸). Light blue and pink horizontal bars indicate CH and CL, respectively. (C to H) Head extracts from CS flies on day 5 of the CH/CL cycle were obtained and processed for immunoblotting with anti-PER (C), anti-TIM (D), and anti-CRY (E) antibodies. Actin served as a loading control. Relative levels of PER (F), TIM (G), and CRY (H) were determined by measuring band densities using Image J software. Three replicates were used for the calculation. Values represent mean ± SEM.

Figure 3. CH/CL-cycle–induced circadian locomotor activity is dependent on core circadian-clock genes.

Flies were exposed to conventional constant light (LL) conditions for 2 days and then maintained on 12 h:12 h cycles of 10,000 K white LED light with high CIL (CH) and 2,000 K white LED light with low CIL (CL) for 11 days. (A and B) Averaged actograms of clock mutants w; per⁰¹ (A) and Clk^out(B) flies are shown. Each row of the actogram was double-plotted. The numbers of flies used for the analyses were 30 (per⁰¹) and 27 (Clk^out). Light blue and pink horizontal bars indicate CH and CL, respectively. (C) Daily average activity profiles of flies of indicated genotype on days 5 and 6 of 12 h:12 h CH/CL cycles are shown. The numbers of flies used for the analyses were 61 (CS), 60 (w¹¹¹⁸), 30 (per⁰¹) and 27 (Clk^out). Light gray vertical bars represent activity during the CH period and black vertical bars represent activity during the CL period. (D and E) Entrainment indices were calculated for flies of each genotype under CH/CL cycles using time windows ZT6–12 (D) and ZT9–15 (E). ZT, zeitgeber time. Values represent mean ± SEM and are compared by pairwise t-test. ***p < 0.0001.

Figure 4. Difference in CIL, not absolute CIL level, is important to drive circadian rhythmicity in CH/CL cycles.

(A) Flies were first exposed to constant light (LL) for 2 days and then entrained to CH/CL cycles for 7 days, followed by constant CL conditions for 7 days. Daily activity profiles of CS flies from day 6 of CH/CL to day 3 of CL/CL are shown. Activity peaks became lower and disappeared beginning on day 3 of constant CL. The numbers of CS flies used for the analyses were 56. (B) Averaged actograms of flies of indicated genotypes are shown. Each row of the actogram was double-plotted. The numbers of flies used for the analyses were 56 (CS), 63 (w¹¹¹⁸), and 18 (per⁰¹). VIL and CIL of 10,000 K (CH) light are increased to 1,570 lx and 1,734 blx, respectively. Those of 2,000 K (CL) light are increased to 1,470 lx and 338 blx, respectively. Dark blue and red horizontal bars indicate brighter CH and CL, respectively. Flies were first exposed to constant light for 2 days and then exposed to 12 h:12 h cycles of brighter CH/CL. Whereas CS and w¹¹¹⁸ flies show circadian rhythmicity, per⁰¹ flies show no rhythmic pattern, which is similar to observations made previously for CH/CL cycles with low light intensity. (C and D) Entrainment indices were calculated for flies of each genotype under brighter CH/CL cycles using time windows ZT6–12 (C) and ZT9–15 (D). ZT, zeitgeber time. Values represent mean ± SEM and are compared by pairwise t-test. **p < 0.001. *p < 0.05.

Figure 5. Wild-type flies can be re-entrained to a 6-h phase shift in the CH/CL cycle.

(A to F) Averaged actograms of flies of indicated genotypes are shown. Each row of the actogram was double-plotted. Flies were first exposed to constant light (LL) for 3 days and then entrained to either a conventional laboratory LD cycle (A–C) or CH/CL cycle (D–F) for 6 days. The properties of CH and CL lights used for this experiment were the same as those of the 10,000 K and 2,000 K lights in Fig. 1B. On day 7, the lights-on time was advanced by 6 h for both LD and CH/CL. The number of flies used for the LD shift was 67 (CS), 31 (w¹¹¹⁸), and 32 (cry⁰). For the CH/CL shift, 53 (CS), 62 (w¹¹¹⁸), and 38 (cry⁰) flies were used. (G) Day 0 indicates the day before the shift and day 1 indicates the first day of the shift. The time of peak activity of each fly for days 0, 1, 2, 3, and 4 was determined using ActogramJ after smoothing. The difference in peak time between day 0 and the indicated day was calculated, averaged, and plotted for each condition. In the LD phase shift, CS flies completely adjusted to the new LD regimen by day 3 of the shift, whereas it took more than 4 days for cry⁰ flies to advance the phase by 6 h. In the CH/CL phase shift, CS flies shifted phase gradually and reached 6 h of phase shift on day 4. cry⁰ flies did not follow the phase shift of CH/CL cycles.