The retinohypothalamic tract (RHT) is a retinofugal neuronal pathway which, in mammals, mediates nonimage-forming vision to various areas in the brain involved in circadian timing, masking behavior, and regulation of the pupillary light reflex. The RHT costores the two neurotransmitters glutamate and pituitary adenylate cyclase activating peptide (PACAP), which in a rather complex interplay are mediators of photic adjustment of the circadian system. To further characterize the role of PACAP/PACAP receptor type 1 (PAC1) receptor signaling in light entrainment of the clock and in negative masking behavior, we extended previous studies in mice lacking the PAC1 receptor (PAC1 KO) by examining their phase response to single light pulses using Aschoff type II regime, their ability to entrain to non-24-h light-dark (LD) cycles and large phase shifts of the LD cycle (jet lag), as well as their negative masking response during different light intensities. A prominent finding in PAC1 KO mice was a significantly decreased phase delay of the endogenous rhythm at early night. In accordance, PAC1 KO mice had a reduced ability to entrain to T cycles longer than 26 h and needed more time to reentrain to large phase delays, which was prominent at low light intensities. The data obtained at late night indicated that PACAP/PAC1 receptor signaling is less important during the phase-advancing part of the phase-response curve. Finally, the PAC1 KO mice showed impaired negative masking behavior at low light intensities. Our findings substantiate a role for PACAP/PAC1 receptor signaling in nonimage-forming vision and indicate that the system is particularly important at lower light intensities.
- circadian rhythm
- suprachiasmatic nucleus
- jet lag
the retinohypothalamic tract (RHT) is the principal neuronal pathway that mediates nonimage-forming vision to the mammalian brain. Nonimage-forming vision is involved in the entrainment of the biological clock, regulation of masking behavior, control of light-regulated secretion of melatonin from the pineal gland, and regulation of the pupillary light reflex (8). We have previously shown that the neuropeptide pituitary adenylate cyclase activating peptide (PACAP) is costored with glutamate in the RHT (13) and that the two neurotransmitters are mediators of photic adjustment of the circadian system in a rather complex interplay (9). After the demonstration of PACAP in the RHT, (12) a number of studies have delineated the role of PACAP in light entrainment, which primarily seems to be exerted via the PACAP-specific PACAP receptor type 1 (PAC1) receptor (3, 4, 11, 14, 17, 23). In vitro, PACAP has been shown to modulate glutamatergic signaling in a time- and concentration-dependent manner targeting the core clock genes Per1 and Per2 via cyclic AMP/PKA and the inositotriphospate/phospholipase C signaling pathway (reviewed in Ref. 10). In studies on mice lacking either PACAP or the PAC1 receptor, using different regimes of light duration and intensity, they were found to have altered response to photic stimulation at night (4, 11, 17). In the first study by Kawaguchi et al. (17), mice lacking PACAP showed a slightly diminished phase delay but a significantly attenuated phase advance to a single light pulse at early and late night, respectively. In a later study by Colwell et al. (4), PACAP-deficient mice exhibited significant impairment in the magnitude of both light-induced phase delay and phase advance (4). We found that the PAC1 receptor-deficient mice displayed a larger phase delay at early night and decreased phase advance at late night (11). Mice lacking either PACAP or the PAC1 receptor are able to entrain to the light-dark (LD) cycle (4, 11, 17), and mice lacking PACAP respond as wild-type (WT) mice to reentrainment of large (8-h) phase shifts of the external LD cycle. Also, masking behavior has been reported to be undisturbed in PACAP knockout (KO) animals (4).
To further characterize the role of PACAP/PAC1 receptor signaling in nonimage-forming vision, we examined PAC1 receptor-deficient mice (PAC1 KO) under the light stimulation regime known as the Aschoff type II regime [in which the animal is kept in a LD cycle, which is then changed to constant darkness (DD) after the phase-shifting light pulse] (1). This regime, which eliminates the confounding effects of chronic DD, is more natural (20). Furthermore, it takes advantage of simultaneous light stimulation (since all animals are entrained) at a given clock time, thus avoiding stress of the animal when moving it due to different circadian cycles of the individual animal. The ability of the PAC1 KO mice to light entrain was also examined during non-24-h LD cycles (T cycles), under which conditions, the animals are forced to phase shift to achieve stable entrainment. Because the amplitude of the phase shift is proportional with the total energy of the light pulse available, (6, 22, 30), the influence of different light intensities during each T cycle was also examined. Finally, the ability to reentrain to large phase shifts (8 h) and masking behavior during different light intensities was investigated.
MATERIAL AND METHODS
A total of 51 PAC1 KO and 51 WT male mice (8 to 10 wk old when included in the study) of the 129/Sv PAC1 strain of mice (11, 16) were used in the study. Group A (11 WT and 11 KO mice) was used for light-induced phase shift at zeitgeber time 16 (ZT16) vs. circadian time 16 (CT16). Group B (eight WT and eight KO mice) was used for light-induced phase shift at ZT22 vs. CT22. Group C (eight WT and eight KO mice) was used to determine a light-phase response curve (PRC) using the Aschoff type II regime. Group D (eight WT and eight KO mice) was used for T-cycle experiment of 26 h and 27 h at 300, 70, and 10 lux and a T-cycle experiment of 23 h at 300 and 10 lux. Group E (eight WT and eight KO mice) was used for T-cycle experiment of 22 h and 21 h at 300 and 10 lux, masking experiments (180 min) at 300 and 10 lux and for jet lag experiments 8-h advance and 8-h delay at 300, 70, and 10 lux. Group F (eight WT and eight KO mice) was used for masking experiments (180 min) at 300, 70, and 10 lux and for jet lag experiments 8-h advance and 8-h delay at 300, 70, and 10 lux. Animals were maintained in a 12:12-h LD cycle in individual cages with food and water available ad libitum. Animals were treated according to the principles of Laboratory Animal Care (Law on Animal Experiments in Denmark, publication 382, June 10, 1987). The experimental protocols in this study were conducted under a license from the Ministry of Justice (Denmark).
Animals were transferred to individual cages equipped with a running wheel in ventilated, light-tight chambers with controlled white lighting adjustable from 10 to 900 lux via a resistance and a switch (IDEC SmartRelay and IDEC SmartRelay software WindLGC V4.0) to generate T cycles of T = 21 h to 27 h. Wheel-running activity was monitored by an online PC connected via a magnetic switch to the Minimitter Running Wheel activity system (11) [consisting of QA-4 activity input modules, DP-24 data ports and Vital View data acquisition system, MiniMitter, Sunriver, OR; ver. 4.1]. Wheel revolutions were collected continuously in 10-min bins. Animals were entrained to a 12:12-h LD cycle [lights on at 7:00 A.M. designated ZT = 0, off at 7:00 P.M. = ZT12]) at 300 lux for at least 14 days before start of experiments.
Light Source and Light Intensity Measurements
White lightning was delivered from fluorescent tubes placed on top of each cage. The light intensity could be adjusted from 10 to 900 lux via a resistance. Light intensity was measured using an Advantest Optical Power meter TQ8210 (MetricTest, Hayward, CA), and measurements were determined at settings of 514 nm: 300 lux (115.0 μW/cm2), 70 lux (19.1 μW/cm2), 10 lux (4.3 μW/cm2).
Endogenous period tau.
Free-running period tau (τ) was assessed during days 4–18 in constant darkness (DD) using ClockLab (ActiMetric Software, Coulbourn Instruments, Wilmette, IL).
Light-induced phase-response curve (PRC) using Aschoff type II regime.
In the present study, we used the Aschoff type II regime, in which the animal is kept in a LD cycle, which is then changed to DD after the phase-shifting light pulse. All animals were light stimulated for 30 min at 300 lux in their home cages in separate experiments at ZT14, ZT16, ZT18, ZT21, ZT22, and ZT23, respectively, whereafter the lights were turned off for the next 5–7 days followed by 14 days of reentrainment in LD before the next light pulse experiment. The light-induced phase shift was determined as described previously (11), using the difference in phase from regression lines drawn through the activity onset of the entrained (LD) onset immediately before the day of stimulation and the onset from 5 to 7 days after light stimulation (to avoid being misled by transients).
Comparison of Light-Induced Phase Shift with Aschoff Type I vs. Aschoff Type II Regime
Because the light-induced PRC in PAC1 KO mice was significantly different using the type II regime (see results) compared with our previous data with the type I regime in the same genotype (see Ref. 11), we decided to compare the response in the same mice of each genotype using the two regimes. WT and PAC1 KO mice were entrained to a 12:12-h LD cycle and released into constant darkness (DD) for at least 14 days. Hereafter, they were light stimulated for 30 min (300 lux) at CT16 [CT16 was determined according to Dann and Pittendrigh (5)], and the phase shift was determined as described above. During the light stimulation, each animal cage was removed from the home container (each container holds 16 cages) and placed in a separate room for illumination. After the end of the illumination period, the cage was returned to the home container, and the animals were left in DD for at least 10 days. The animals were then reentrained to the 12:12-h LD cycle for another 14 days, whereafter they received a 30-min (300 lux) light pulse at ZT16 within the home container. The lights were turned off for the next 5–7 days followed by 14 days of reentrainment in LD. The same procedure was performed at CT22 and ZT22, respectively. As control, phase shifts at both early and late night were also determined in the same animals during the transition from LD to DD. Phase shift was measured as described above.
T cycles with Different Light Intensities
To test the functional sensitivity of the light entrainment system in PAC1 KO and WT mice, we used non-24-h LD cycles (T cycles). Such experiments demand daily phase shifting of defined size and direction to achieve stable entrainment. Animals were exposed for T cycles of T = 26 and T = 27 h followed by T cycles at T = 23 h and T = 22 h and 21 h. Because the amplitude of the phase shift is proportional with the total energy of the light pulse available (6, 22, 30), we evaluate the influence of different light intensities during each T cycle. At the start of each T cycle, all animals kept in T cycles >24 h received light of 300 lux for 14 days, whereafter light intensity was reduced to 70 lux for another 10–14 days followed by a further reduction to 10 lux for another 1 to 3 wk. Animals kept in T cycles <24 h were tested at 300 and 10 lux (pilot experiments found no difference at these light intensities and 70 lux). The ability to entrain was evaluated by examining the activity onset relative to light onset during the T cycles (i.e., the phase angle), and the periodogram function in ClockLab was used to determine τ during the various light conditions (Fig. 2).
Eight-Hour Phase Delays/Advances (Jet Lag) at Different Light Intensities
To examine whether the changed sensitivity to light influenced the time of reentrainment during large shifts of the external LD cycle, animals were exposed for 8-h phase delays of the external LD cycle followed by an 8-h phase advance after reentrainment to the new LD cycle. Reentrainment was defined as the first day of consecutive days, in which the onset occurred within 30 min in phase with the new LD cycle. These experiments were performed at light intensities of 300, 70, and 10 lux.
Negative Masking During Changing Light Intensities
Negative masking behavior in PAC1KO animals was examined during 3 h of light at three different intensities (300 lux, 70 lux, and 10 lux) exposed from ZT14 to ZT17. Each experiment was performed followed by ∼7 days of entrainment in a 12:12-h LD cycle. Baseline activity for each animal was determined as the activity measured as 10-min bins during the 3 h (ZT14–ZT17) on the night before the 3 h of light. Masking activity was determined as the activity measured during the 3-h light pulse as 10-min bins.
Statistics were performed using GraphPad Prism v 4.0 (San Diego, CA). For comparison of two independent groups, Mann-Whitney U-test was used. For comparison of two dependent groups (ZT vs. CT), Wilcoxon signed rank test was used. Two-way ANOVA followed by Bonferroni post hoc test was used for the comparison of data from jet lag and masking experiments. P < 0.05 was considered statistically significant.
The endogenous period (τ).
In the backcross strain of PAC1 129/Sv KO mice used in the present study, the τ was significantly shorter compared with the WT (23.24 vs. 23.84 h, n = 8 animals in each group, P = 0.0012, Mann Whitney U-test).
Light-induced phase-response curve (PRC) using Aschoff type II regime.
The PRC of the PAC1 KO mice during Aschoff type II regime displayed a markedly reduced light sensitivity at early night compared with WT animals (Fig. 1). Thus, at ZT14 and ZT16, the light responses were reduced −86 min and −129 min in WT vs. −39 min and −24 min in PAC1 KO mice, respectively. At late night, no difference was found between PAC1 KO mice and WT mice (Fig. 1).
Comparison of light-induced phase shift with Aschoff type I vs. Aschoff type II regime.
We examined the light-induced phase response in the same mice of each genotype using the type I and the type II regimes (Fig. 1, B and C). In agreement with previous findings (11), a light pulse at early subjective night induced a phase delay in PAC1 KO mice at CT16 (type I condition), which was significantly larger than in WT animals (Fig. 1B). When the same animals were exposed to a light pulse at ZT16 (type II condition), PAC1 KO mice displayed significantly smaller phase delay compared with the wild type (Fig. 1B). WT animals showed similar phase delays to the 30-min light pulse in the two regimes, whereas the responses in the PAC1 KO mice were significantly different (Fig. 1B). At late subjective night, a similar pattern was found. The responses in WT mice were independent of the regime used (Fig. 1C), whereas the PAC1 KO mice showed a significantly smaller phase advance during the type I regime than in the type II regime (Fig. 1C). Neither during early nor late night did the WT and the PAC1-deficient mice differ in their response to the LD to DD transition, and the responses had low variability (Fig. 1, B and C).
Entrainment to T > 24 h at different light intensities.
Both WT and PAC1 KO mice entrained at τ = 26 h at light intensities of 300 and 70 lux. When the light intensity was reduced to 10 lux WT mice were still able to entrain, whereas the PAC1 KO mice established stable activity rhythms with a period length of 25.4 ± 0.2 h (Fig. 2A). When WT mice were transferred into cycles of T = 27 h (13.5:13.5 h LD), they entrained at light intensities of 300 and 70 lux, but at 10 lux, the animals could no longer entrain and continued with an activity cycle corresponding to a period length of 25.73 ± 0.16 h (Fig. 2, B and C). PAC1 KO mice, in contrast, were unable to fully entrain to a 27-h cycle at any of the light intensities. The synchronized activity of 300, 70, and 10 lux corresponded to cycles of 26.7 ± 0.12 h, 26.2 ± 0.31 h, and 25.0 ± 0.25 h, respectively (Fig. 2, B and D).
Entrainment to T < 24 h at different light intensities.
WT and PAC1 KO mice were able to entrain to short T cycles of 23 and 22 h at both 300 and 10 lux (Fig. 3) light intensities (data not shown). At T cycles of 21 h, six WT mice entrained, whereas two mice were unable to entrain to the 21-h cycle. These mice showed abrupt changes in phase, so-called “phase jumping” (Fig. 3). Only one PAC1 KO mouse entrained to the T cycle, whereas six showed phase jumping (Fig. 3).
Entrainment after 8-h phase delay (jet lag west) at different light intensities.
WT mice exposed to 8-h phase delay at 300 lux entrained to the new LD cycle within two cycles (Fig. 4, A and B), which was significantly faster than the four cycles needed for the PAC1 KO mice (Fig. 4, A and C) (P < 0.001, two way ANOVA followed by Bonferroni post hoc test). When light intensity was reduced to 10 lux, WT mice used six cycles to reentrain after the 8-h delay of the LD cycle (Fig. 4, D and E), while the PAC1 KO used 10 cycles (Fig. 4, D and F) (P < 0.0001, two-way ANOVA followed by Bonferroni post hoc test).
Entrainment after 8-h phase advance (jet lag east) at different light intensities.
WT mice exposed to 8-h phase advance at 300 lux light intensity used five cycles to reentrain to the new LD cycle (Fig. 4, G and H), while PAC1 KO mice used significantly longer time, namely eight cycles (P = 0.035, two-way ANOVA followed by Bonferroni post hoc test) (Fig. 4, G and I). When light intensity was reduced to 10 lux, one group of both WT (n = 9) and PAC1 KO mice (n = 9) used ∼15 cycles for readjusting (advance) to the new LD cycle (Fig. 4, J–L). Six WT and five PAC1 KO mice respond to the 8-h phase advance at 10 lux light intensity by a phase delay. The WT mice used 15 cycles, whereas PAC1 KO mice used 21 cycles for reentrainment to the 8-h shift of the LD cycle (P = 0.05, two-way ANOVA followed by Bonferroni post hoc test).
Negative masking at different light intensities.
WT and PAC1 KO mice were tested for negative masking behavior between ZT14 and ZT17 at light intensities of 300, 70, and 10 lux. At 300 lux, both the WT and the PAC1 KO mice showed identical negative masking behavior with an almost complete inhibition of running-wheel activity during the light period (Fig. 5, A and B and Fig. 6, A and B) (P = 0.008 and P = 0.004, respectively. Two-way ANOVA followed by Bonferroni post hoc test). When light intensity was reduced to 70 lux, masking behavior in WT mice was unaffected (Figs. 5C and 6C) (P = 0.04, two-way ANOVA followed by Bonferroni post hoc test), while PAC1 KO mice no longer showed significant masking but continued their running wheel behavior not significantly different from baseline activity (Figs. 5D and 6D) (P = 0.24, two-way ANOVA followed by Bonferroni post hoc test). At 10 lux, WT mice were still able to mask (Figs. 5E and 6E) (P = 0.041, two-way ANOVA followed by Bonferroni post hoc test), whereas the PAC1 KO mice showed no masking behavior (Figs. 5F and 6F) (P = 0.72, two-way ANOVA followed by Bonferroni post hoc test).
The present study was undertaken to further characterize the role of PACAP/PAC1 receptor signaling in light entrainment of the clock and in negative masking behavior, another nonimage-forming vision regulated behavior-mediated by the melanopsin/PACAP containing intrinsic photosensitive retinal ganglion cells (8). We preferred the Aschoff type II to the type I procedure (20) to generate a light PRC, since it eliminates possible confounding effects of chronic DD, including a more light-sensitive system compared with the LD cycle experiment (25–27, 29) and permits simultaneous light stimulation (since all animals are entrained) at a given clock time. Using the type II procedure, we found that the response to light stimulation in PAC1 KO mice differed from our previously obtained results using the type I regime (11). Most striking was the attenuated phase delay found in the type II at early night (increased in type I), a result, however, which accords with that found in two different strains of mice lacking PACAP (4, 17). Although the response in WT mice was independent of the regime used, the PAC1 KO mice displayed different light responses at both early and late subjective night. Although this is not readily explained, it is likely that mice removed from their home container during the type I regime exhibit stress and anxiety. Both PACAP-deficient mice and PAC1 receptor-deficient mice have an altered response to stressful stimuli (15, 24), which might influence their response to light, as previously reported in rodents using restrained stress (19).
To further substantiate the altered responsiveness in PAC1 KO mice, we performed T-cycle experiments. When changing the length of the LD cycle, as well the light intensity, the limits of (light-) entrainment could be investigated. The ability to entrain to light is determined by several factors, including the animals' free-running period, the PRC, and the overall sensitivity to light. Our results of the T-cycle experiment with T > 24 h showed that WT mice could entrain to T = 26 h at all light intensities, whereas the PAC1 KO mice were unable to entrain during low light intensities. This could, to a large extent, be explained by the significantly shorter τ of the PAC1 mice (23.2 h) than the τ of WT mice (23.8 h). The PAC1 mice, therefore, need to make larger phase shifts than the WT mice. When the T cycle was extended to 27 h and the light intensity simultaneously reduced to 10 lux, the PAC1 KO mice reached their limit of entrainment. The inability to entrain to T = 27 h could no longer be explained by the difference in free-running period length between the two genotypes but suggests that the PAC1 KO mice are less responsive to the delaying light stimulation at early night. The attenuated light sensitivity in PAC1 KO mice accords well with the prolonged time for reentrainment after a shift of the LD cycle, corresponding to an 8-h east-west trans-Atlantic travel. Such a large phase shift of the external LD cycle affected the PAC1 KO mice even more under reduced light intensities. While entrainment to T cycles of 23 and 22 h at high and low light intensities and the PRC at late night were similar in the two genotypes, most of the PAC1 KO mice (six out of seven animals) reached their limits of entrainment at T cycles of 21 h (Fig. 4). This was unexpected, considering the shorter free-running period in the PAC1 KO mice. Theoretically, these mice should make smaller daily phase shifts to entrain to the 21-h cycle compared with the WT mice. It was also expected that the PAC1 KO would reentrain to 8-h phase advance of the LD cycle using fewer cycles than the WT mice, but this was not the case. The data suggest that the PAC1 KO mice have also attenuated response to light at late night, although this was not disclosed by the single light pulse experiment of the PRC.
Previous in vitro studies have shown that PACAP can modulate the effects of the other RHT transmitter, glutamate (13). When applied together with glutamate, PACAP amplifies the phase-shifting properties of glutamate at early subjective night, and in accordance, application of a PACAP receptor antagonist decreases the glutamate-induced phase delay (3). Furthermore, infusion of anti-PACAP antibodies near the suprachiasmatic nucleus (SCN) in hamsters attenuates light-induced phase delays (2). In vitro studies at late night have shown that PACAP attenuates the effects of glutamate (3). It was, therefore, to be expected that a light pulse resulted in a larger phase shift in mice lacking PACAP/PAC1 receptor signaling, but this was not the case, either in the present study or in mice lacking PACAP (17). The discrepancy between the result obtained in vitro and in vivo may be explained by the omission of other neuronal pathways known to modulate RHT signaling in the denervated in vitro preparation (see discussion in Ref. 28). Studies of glutamate as a mediator of light-induced phase shift have also demonstrated that results obtained in vitro (7) cannot always be reproduced in vivo (18).
The behavioral data of the present study agree with our results on gene expression in the SCN of PAC1 KO mice (11). At early night, light induced the expression of the immediate early gene c-Fos and the two light-responsive clock genes, Per1 and Per2, in WT mice, while their expression was significantly attenuated in PAC1 KO mice (11). At late night, no difference between the genotypes was found in light-induced expression of c-Fos and Per gene expression (11), as found in the behavioral light response in the present study.
It is well established that animals can confine their behavior (either nocturnal or diurnal) by synchronizing the endogenous clock to the LD cycle and by responding directly to light or darkness, the latter process known as masking (21). The demonstration that the negative masking behavior in PAC1 KO mice is impaired at lower light intensities indicated that the PACAP/PAC1 receptor-mediated light signaling is important for normal masking behavior. The response pattern was similar to that seen for the ability to synchronize the clock and could be a general pattern for nonimage-forming vision-regulated functions. Negative masking has also been examined in mice lacking PACAP (4). In this study, a paradigm of 1 h of light at ZT14–ZT15 using three different clearly defined light intensities was used (4). In contrast to the present study in PAC1 receptor-deficient mice, no difference in negative masking behavior could be demonstrated. To solve this difference in masking response, studies using the same paradigm in both genotypes need to be performed.
Using PAC1 receptor-deficient mice, Aschoff type II regime, T cycles, and different light intensities, the present study demonstrates a role of PACAP/PAC1 receptor signaling in light entrainment and in negative masking behavior. This was most prominent at low levels of irradiance during early night stimulation where PAC1 receptor-deficient mice showed reduced phase shifting ability and decreased negative masking behavior. Studies at late night suggest that PACAP/PAC1 receptor signaling is less important during the phase advancing portion of the phase-response curve. Our findings substantiate a role for PACAP/PAC1 receptor signaling in nonimage forming vision and indicate that the system is particularly important at lower light intensities.
Perspectives and Significance
The present study extends previous studies and substantiates the role of PACAP/PAC1 receptor signaling in nonimage-forming vision to the brain in mammals. The PACAP/PAC1 signaling system seems to be of particular importance when light intensity is low, suggesting a particular role as mediator of light information at dusk and dawn. Since PACAP-containing retinal projections of the RHT also reach areas in the brain involved in sleep regulation it is likely that PACAP/PAC1 signaling is involved in light-regulated sleep functions. One might also speculate that the PACAP system is affected in individuals suffering from seasonal affective disorder, a change in mood behavior occurring at low light intensities during the winter season.
This study was supported by The Lundbeck Foundation, The Danish Biotechnology Center for Cellular Communication, (Grants 0001716 and 22020345, respectively) and the Danish Medical Research Council.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society