The circadian activity rhythm is reset by nanowatt pulses of ultraviolet light

Abstract

The circadian pacemaker synchronizes to the Earth's rotation by tracking step-by-step changes in illumination that occur as the sun passes the horizon. While twilight progressions of irradiance and colour are considered important stimuli in this process, comparably less thought has been given to the possibility that ultraviolet A (UVA) radiation might actually play a more formative role given its evolutionary significance in shaping 24 h timekeeping. Here, we show that Drosophila activity rhythms can be phase-shifted by UVA light at an energy range seated well below that of the visible spectrum. Because the energy threshold for this resetting matches the incident amount of UVA on the human retina at twilight, our results suggest that UVA light has the potential to function as a similar time cue in people.

1. Background

Sunrise and sunset offer scenes of natural beauty that have long-held emotional and symbolic meaning for humanity. This aesthetic appreciation is likely oriented by a brain system designed to capture and steer visual attention towards objects perceived to confer survival advantage, such as edible fruits or healthy mates [1,2]. In a similar manner, while awe-inspiring, the dynamic colour and irradiance changes that accompany the twilight sky also have an important communication value. The sun's steady emergence or disappearance from the horizon gives way to a stream of photic information that multicellular animals track to entrain their endogenous circadian rhythms to the Earth's rotation [3]. Though food availability and social interactions can provide secondary time cues [4], the metazoan circadian pacemaker ultimately depends on replays of twilight at dawn—and mirror images of this replay at dusk—to synchronize the body's physiology to a resource-optimal temporal niche rooted in either day or night [5].

Irradiance detection and colour discrimination at twilight are thought to be the key sensory mechanisms favoured by the circadian pacemaker to infer time of day [6–8]. Illumination within the twilight zones is fraught with various signal-to-noise problems, but a consistent feature of these transition periods is the relative enrichment of shorter, ‘bluer’ wavelengths in the visible light spectrum produced when the sun is seated below the horizon (often referred to as the ‘blue hour’ in dim-light photography) [8–10]. Changes in the quantity and quality of ambient light at dawn and dusk are so salient that little consideration has been given to other factors within the solar cycle that might contribute to pacemaker photoentrainment. One factor that often gets overlooked is ultraviolet (UV) radiation. The Earth's atmosphere did not filter UV light 3–4 billion years ago, and recent conjecture posits that biological rhythms first evolved in early photosynthesizing organisms to create cycles of DNA regulation that protected against UV damage during the day and facilitated cell replication at night [11–14]. It stands to reason that UV exposure provided a selective pressure for establishing approximately 24 h timekeeping. Furthermore, UVA light (315–400 nm) in particular, owing to its continuing lack of absorption by the ozone layer and its virtual absence from moonlight or starlight [10,15,16], would be an adaptive temporal cue that the evolving pacemaker could consult to tell it when the sun was out.

At present, animals ranging from Drosophila to humans maintain UVA-sensitive retinal photoreceptors [17–24]. In traditional laboratory models with UVA-transmitting ocular lenses (or no lens at all), UVA has been generally shown to elicit pacemaker resetting [25–30]. In humans, the logic for how UV light is used by the circadian system is not understood, though many assume that the approximately 2% of incident UVA that survives filtration through the lens has little-to-no influence on retinal and brain function [31]. The likelihood that UVA from sunlight has a direct, evolutionarily conserved impact on the human circadian system would be increased if the pacemaker were shown to respond to very small doses of UVA below the range of its sensitivity to monochromatic visible light. We evaluated this possibility—that UVA light might sit atop a hierarchy of information guiding pacemaker photoentrainment—in Drosophila ananassae. This particular species of fruit fly tends to shun more natural habitats in favour of cohabitation with humans [32]. Accordingly, the animals show a consolidated pattern of locomotor activity during the day and consolidated sleep at night that mimics the diurnal sleep/wake patterns of people and offers a realistic model of human circadian behaviour [33].

2. Material and methods

The ananassae were derived from an isofemale line maintained at the Drosophila Species Stock Center (DSSC) at Cornell University (stock # 14024–0371.16; NSF Award #1351502). Stocks were reared at 25°C in DigiTherm^® incubators (Tritech Research, Inc., Los Angeles, CA) and entrained to a 12 L : 12 D cycle (broad-spectrum light source: 4 W cold-cathode fluorescent light tube with step-up inverter, freely mounted with no fixture, illuminance at rack level = 887.7 lux, irradiance = 309.5 µW cm⁻²; Tritech model DT2-LB-F12IN/CIRC-L-INV; lights-on at 07.00, MST). The stocks were transferred daily to generate a steady supply of offspring. For phase-shifting experiments, female flies were selected as late-stage, ‘pharate adult’ pupae, moved onto fresh food and housed in groups of five to six in a secondary DigiTherm^® incubator. This secondary incubator, in which the collected pupae eclosed, was programmed to run a 12 L : 12 D cycle with lights-on at 01.00, MST, to accommodate subsequent phase-delaying treatments at CT13 (i.e. 14.00, the first hour after lights-off).

An Aschoff Type II paradigm was used to quantify the effects of millisecond pulses on phase resetting of the flies' locomotor activity rhythms. Animals were entrained to the 12 L : 12 D schedule under which they enclosed for 3 days. Prior to lights-off on the last day of the schedule, the flies were grouped into disposable cotton-plugged Pyrex tubes (approx. 8–10 flies per tube, 13 mm outside diameter and 100 mm long). For light administration at CT13, two to four of these tubes were placed side-by-side onto a titanium dioxide paint-coated platform and exposed to one of the following treatments: 120 ms pulses of either (i) UV (λ_max 365 nm, half-bandwidth ≤ 10 nm; irradiance = 0.019, 0.033, 0.069, 0.140 or 0.210 µW cm⁻²), (ii) blue (λ_max 452 nm, half-bandwidth ≤ 21 nm; irradiance = 0.47, 0.98 or 15.6 µW cm⁻²), (iii) cyan (irradiance = 3.88 or 34.1 µW cm⁻²), (iv) green (λ_max 525 nm, half-bandwidth ≤ 35 nm; irradiance = 3.83 or 32.60 µW cm⁻²), (v) red (λ_max 640 nm, half-bandwidth ≤ 20 nm; irradiance = 8.6 or 32.3 µW cm⁻²) or (vi) white (irradiance = 1.10, 3.56 or 5.70 µW cm⁻²) light delivered on the second (1 Hz) for 15 min with a ColorDome LED Ganzfeld lamp (Diagnosys LLC, Lowell, MA). Cyan emission was accomplished by simultaneous activation of blue and green LEDs; white emission was accomplished by simultaneous activation of blue, green as well as red LEDs. Stimulation instructions were sent by Diagnosys’ software-interfaced Espion E³, an amplifier console capable of producing pulse width modulation (PWM) intensity-controlled LED flashes as short as 4 ms. Lamp output was specified in candelas per square metre (cd m⁻²; i.e. a photometric measure of luminous intensity) and quantified—photometrically (lux, lumens m⁻²) and radiometrically (µW cm⁻²)—with the ILT950 spectroradiometer (International Light Technologies, Peabody, MA). Irradiance measures were used to calculate photon flux (photons cm⁻² s⁻¹) according to the equation: photon flux = irradiance (µW cm⁻²)/energy per photon (hc/λ_max). All flash protocols began precisely at CT13, ended by CT13.25, and were conducted in complete darkness with the aid of night vision headgear. Independent sets of naive animals were used for each flash protocol.

Following photic treatment, flies were housed singly in glass chambers (5 mm outside diameter and 65 mm long) containing a plug of corn flour-nutritional yeast-agar medium on one end (identical to the formulation of culture tubes used for colony maintenance—0.8% agar, 3.5% sucrose, 1.7% glucose, 6% fine-grained masa and 1% yeast) and a cotton fitting on the other, and loaded into Trikinetics DAM2 Drosophila Activity Monitors (TriKinetics, Inc., Waltham, MA). Their motion was independently tracked for the next 3 days under constant darkness (DD) by cross-sectioned infrared beams, which transmitted movement information over modem/USB to a computer acquisition software every 30 s. DAM2 units were situated in climate-controlled vivariums identical to the ones used in colony management and under the same ambient conditions.

Phase shifts of behaviour were calculated for each fly (one fly per one 5 × 65 mm tube) by determining the horizontal distance between the time of lights-on in the previous light--dark (LD) schedule (CT0, 01.00) and the software-called activity onset on the second day after millisecond flash exposure (ClockLab Analysis v.6, Actimetrics, Wilmette, IL). The activity onsets of ananassae are always phase-locked to the timing of lights-on within an LD schedule. Post-pulse in DD, transients are observed for a day, but the flies' behavioural rhythms stably reset by the second DD cycle. To correct for phase movements that might simply accompany transitions from LD to DD, a control group was transferred into DD without any light treatment at CT13. Net calculations of onset shifts were normalized for the effects of LD schedule removal. The general protocol for our assessment is consistent with the standard (long-held) practices of the Drosophila phase response curve (PRC) literature. Here, Aschoff II or ‘anchored’ paradigms are routinely used to measure the effects of night-time light on locomotor activity phase. The position of the phase reference point 2 days post-pulse is often the benchmark for quantifying the magnitude of a shift [34]. The performance of each light treatment was evaluated by a one-sample t-test to determine whether the net delay shift it produced was greater than zero, a score that indicates no phase movement. Significance was set at p = 0.05.

3. Results

Pacemaker photoresponses can be compared through a series of phase-shifting experiments where animals are given a night-time light signal that either delays or advances their endogenous rhythms [34–36]. We have recently confirmed that ananassae exhibit phase-resetting responses to electrical light that are typical of Drosophila melanogaster, rodents and humans [34–37] and, like these animals, have the capacity to integrate photic information presented across a series of millisecond flashes [37–39]. These background data allowed us to quantify the phase-shifting effects of 1 Hz millisecond exposure to narrowband, blue (λ_max 452 nm), green (λ_max 525 nm), red (λ_max 640 nm) or UVA (λ_max 365 nm) light delivered at CT13, a point in the subjective evening where light administration is thought to be compatible with stable circadian entrainment [40].

Consistent with previous observations [25,41], the action spectra for pacemaker resetting of behavioural and physiological rhythms were biased towards shorter wavelengths (figure 1). Ananassae receiving blue LED stimulation showed delays in locomotor onset that were still significantly above baseline when the flashes were lowered to levels of illuminance below the dark limits of civil twilight and within the limits of nautical twilight (approx. 0.3–11 lux, when the sun is more than 6° below the horizon; pulse irradiance = 0.98–15.6 µW cm⁻²; figure 1b, second column, and table 1) [42]. This remained so when blue LED activation was combined with green and red to produce white light emission at civil twilight illuminance (figure 1b, last column). While millisecond light exposure from green LED units resulted in significant phase-shifting of the flies' activity rhythm, the delay responses asymptoted below the larger magnitude responses seen with blue and white light administration. This deficit could not be overcome when the irradiance received from the green LED exposure was adjusted higher to match the approximate energy of blue light at dawn/dusk (rows 8 and 12, table 1), but could be increased slightly when green and blue LED activation were combined to produce cyan emission (rows 8–10, table 1). In keeping with previous results in flies, rodents and humans [25,43,44], red light at 11–51 lux (pulse irradiance = 8.6–32.3 µW cm⁻²) did not trigger any circadian responses in ananassae beyond that noted with removal of the LD schedule (second to last column, figure 1b and table 1).

Figure 1.

Comparing pacemaker responses to the non-visible and visible light spectrum. (a) Flies received 120 ms pulses of 1 Hz UVA or visible LED light for 15 min starting 1 h after lights-off (CT13) on the last day of the light–dark schedule (12 L : 12 D). Irradiance spectra for the UVA, blue, green and red LED units, alone, and when used in combination to produce cyan or white light pulses, are shown across the bottom panels of (a). Enumerated within are the dominant emission wavelengths for each output. (b) Post-pulse, animals were loaded into individual activity monitoring chambers and tracked for 3 days in constant darkness (DD). A population actogram for each of the treatment groups (1–17; table 1) is provided, overhung with grey and black bars that show the timing of the previous LD schedule. Asterisks mark delivery of the light stimulation protocol. The actograms are organized in columns situated underneath the colour of light that was administered and positioned in rough approximation to how much energy was delivered (reference the logarithmic scale to the far left). To help visualize shifts in locomotor rhythms, a population actogram from a non-light exposed control group (black) is situated beneath each monochromatic set. A dot, where present, defines the threshold at which a significant behavioural shift is achieved. BG, blue–green; BGR, blue–green–red.

Table 1.

Summary of light treatments and their effects on locomotor activity phase. Luminance was specified by the experimenters with a computer-interfaced Ganzfeld lamp (ColorDome, Diagnosys LLC, Lowell, MA). Spectral energy data from the lamp's emission was quantified with an ILT950 spectroradiometer (International Light Technologies, Peabody, MA). For behavioural data, mean ± s.e.m. was calculated from the delay shifts observed in individual animals (numbers in parentheses indicate the number of animals sampled per condition).

^aLight treatments not producing a significant delay shift (evaluated by a one-sample t-test to determine whether the delay magnitude was greater than 0.0 h; significance set at p = 0.05).

Owing to the restraints in power output from UVA-LED chips, we began our circadian analysis of UVA-LED stimulation at an irradiance approximately 4.5× lower than that necessary to produce a significant behavioural shift with blue millisecond flashes at CT13. Flies exposed to UVA flash protocols delivering a total energy payload of 3–23 µJ cm⁻² exhibited significant delays in locomotor activity rhythm that scaled with the amount of light received (figures 1 and 2). Delivery below this energy range, however, did not trigger any net phase resetting (table 1, first row). Our data speak to past efforts to establish UVA light as a time-giver or ‘zeitgeber’. At the same time, they provide the first evidence to suggest that the pacemaker might be capable of orienting to extremely low levels of UVA light that exceed its sensitivity to visible spectrum light.

Figure 2.

Pacemaker responses to UVA versus blue light exposure. Phase shifts of locomotor activity (h) were quantified for flies stimulated with one of several intensities of UVA (magenta) or blue LED light at CT13 (120 ms pulses, 1 Hz, 15 min). Individual values are shown in scatter plot as a function of the total energy content of the light (µJ, logarithmic scale). The average response observed in each treatment group (#1–8; figure 1 and table 1) connects the trend lines, which culminate at the energy level producing the greatest mean behavioural shift (top broken dotted line). For both UVA and blue light, larger phase shifts are observed with increasing energy. However, the dose–response curve for UVA is completely segregated—i.e. seated in a lower, non-overlapping range—from the blue dose–response curve when light is delivered according to the 120 ms, 1 Hz protocol.

As the energy response curves for UVA and blue were both monotonic (one-way ANOVA, F > 7.7, p < 0.0001), but did not overlap (figure 2), it is worth asking how the threshold for UVA-induced shifting in our study compares with the visible red, green and blue (RGB) thresholds that have been reported elsewhere. This direct comparison is complicated by the fact that measurements of circadian photosensitivity are invariably scheduled at points of the subjective evening where light has been shown to trigger the most resetting (e.g. CT14–16). In many cases, these measurements are also done with light-sensitized animals that have been housed in constant darkness for 7–14 days before the probe pulse (see, for example, [43,45–47]). Here, we used a circadian procedure that targeted an earlier portion of the subjective evening that is less inherently circadian photosensitive and stimulated animals on the last night of their LD schedule, so that their responses were more consistent with the natural field shape of the PRC vis-à-vis photoentrainment [48]. In spite of these factors, the quanta density threshold we report for UVA-induced resetting (6.50 × 10¹² photons cm⁻²; table 1) is the lowest ever recorded to produce a behavioural phase shift in an animal. All other investigations with visible RGB light in Drosophila and rodents have reported quanta density thresholds between 9.00 × 10¹² and 6.50 × 10¹³ photons cm⁻² [38,43–47,49].

4. Discussion

Drosophila use various characteristics of the light spectrum to organize behavioural responses to sunlight. For instance, the skyward position of the sun and the intensity and polarization gradients created by atmospheric scattering of light are among the stable references that the animals consult to maintain a particular heading while dispersing over short distances or travelling between nearby locations [50,51]. Over wider terrain (approx. 10 miles), light-dependent magnetoreception or ‘compass orientation’ likely offers an additional resource for achieving accurate navigation [52–55]. Wavelength-guided responses to narrowband RGB light have also long been scrutinized to better understand how Drosophila process colour or entrain to solar LD cycles (e.g. [25,56]). However, relatively few studies have explored the informative properties of light in the UVA band. Those that have suggest that UVA is a relevant signal for calibrating approach or avoidance to sunlight at midday, a time when flies are at risk for thermal injury and desiccation [20].

In the current study, we report for the first time that the Drosophila circadian activity rhythm is shifted by UVA light at an energy level seated below that of the visible RGB spectrum. Owing to evolutionary convergences in the photoentrainment pathways of flies and mammals [5], including similar photosensitivity thresholds and action spectra for circadian resetting [6,25,41,49], our findings raise the possibility that UVA from crystalline lens-filtered sunlight plays an equally important synchronizing role for the human circadian system. While this generalization is complicated by the fact that Drosophila possess different light sensing hardware from mammals and a different organizational layout to their brain's pacemaker network [57], it is made more tangible by human data showing distinct UVA visual sensitivity in children and young adults [58,59]. A few calculations would further bolster the case that UVA is a temporal cue pertinent to human circadian physiology. Assuming 2 h exposure to the midday sun, the dose of solar UVA that would be expected to reach the human lens is ≥2 J cm⁻² [31,60]. If we correct for the general difference in ambient illumination between the midday sun and twilight (a scaling factor anywhere between approx. 1/10 000 and 1/25 000), this means that the amount of lenticular UVA exposure at dawn or dusk could be as high as 200 µJ cm⁻². If we further assume that 98% of this light is filtered, then this means that the retina would be subject to a maximum twilight UVA dose near 4 µJ cm⁻², which matches the threshold we have calculated for UVA phase-shifting in the current study (3.54 µJ cm⁻²; table 1, row 2). These calculations place UVA light in an energy range that has the potential to impact the human retina and therefore circadian timekeeping.

The potential of UVA light to convey time-of-day estimates to the human pacemaker offers a parsimonious explanation for why circadian impairment might afflict people as they get older. From childhood to about late adolescence, the human lens is capable of transmitting photic signals down to the UVA–UVB spectrum [61,62]. However, with age, spectral transmission of shorter wavelength light, especially UVA, is reduced to the retina because of yellowing of lens proteins [63]. This curtailment might have functional consequences for the non-image forming system, such as weakened entrainment to the solar cycle [64,65]. Lens yellowing is widely thought to protect the older retina from phototoxic damage, but complete elimination of the UVA band might pressure the pacemaker to orient to longer wavelength visible light that has much less phase-shifting capacity. Future phototherapies seeking to rehabilitate age-related circadian dysfunction might benefit from the design of strategies that compensate for UVA signalling loss.

Ethics

All experimental procedures were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

Data accessibility

Original data have been uploaded as the electronic supplementary material.

Authors' contributions

F.F. developed the study concept, oversaw its experimental design, drafted the paper and procured funding for all aspects of the work. Behavioural testing and data collection were performed by D.N. and S.K. All the authors contributed to discussing and interpreting the findings and approved the final version of the manuscript for submission.

Competing interests

The authors have no competing interests.

Funding

We are indebted to Science Foundation Arizona (SFAz), the BIO5 Institute at the University of Arizona and the State of Arizona and Arizona Department of Health Services for their financial support.