In the present work we demonstrate a circadian rhythm in consensual PLRs in a nonmammalian model of blindness, the GUCY1* chickens, kept under constant illumination conditions (DD). These birds carry an autosomal recessive mutation in the photoreceptor-specific guanylate cyclase 1 (GC1) gene
17 that severely affects both phototransduction and survival of cones and rods.
18,19 This retinopathy causes blindness at hatch, similar to Leber's congenital amaurosis in humans.
20 However, these birds still retain the ability to detect light that regulates a variety of nonimage forming functions such as the photic entrainment of feeding rhythms and PLRs.
6,12 Furthermore, as previously shown in WT RGCs,
21 in GUCY1* birds these cells are still able to give circadian expression to the mRNA for aralkylamine N-acetyltransferase (AA-NAT), a key regulatory enzyme in the synthesis of melatonin, albeit with a completely shifted phase.
22
In most studies on PLRs, pupil amplitude (either absolute or relative to the starting pupil diameter) is the main parameter for characterizing pupil constriction in response to light.
1,3,23,24,27 However, other variables could also be considered, including latency, velocity, or acceleration.
25,26
Here we have examined two of these PLR variables in GUCY1* chickens: the relative amplitude (% constriction) and velocity (kinetics) of the responses. A clear daily change was visualized in pupil amplitude during white light stimulation (
Fig. 1); this amplitude variation was observed in both the NSP (
Fig. 2A) and the SP (
Fig. 4A) of the PLR, indicating that both phases contribute to the whole response and that the daily control of pupil size is measurable after only a few seconds of the bright white light stimulus onset.
Similar daily changes in whole PLR amplitude were observed in response to a 475-nm blue light stimulus (12.05 mW/cm
2;
Fig. 1), but in this case the changes appear to be entirely due to differences observed during the SP (
Fig. 4B), since no clear daily differences were seen during the NSP (
Fig. 2). Thus, with a bright 475-nm blue light stimulus, the daily control of PRL would seem to be measurable only when the pupil has reached its stable minimum size.
As expected, levels of constriction during the SP for a 475-nm blue light pulse are higher than those for a white light pulse (
Fig. 4B) since the opsin photopigment driving the PLRs in GUCY1* birds is most sensitive around 484 nm. This also explains why, when using a 475-nm blue light stimulus, the percentage of constriction obtained during the NSP throughout the circadian times (
Fig. 2A) was more pronounced than that observed with a white light stimulus at both subjective day and night (
Figs. 2A,
2B). By contrast, no daily differences in the PLR amplitude under a 430-nm light stimulus (0.35 mW/cm
2) were found after the whole statistical analysis (
Fig. 1,
middle) or the analysis of the SP and NSP conducted separately (
Figs. 2,
4,
middle). Nevertheless, a small percentage of pupil constriction was observed in the controls exposed to a 430-nm blue light stimulus (at both SP and NSP). It is worth noting here that although the intensity of the 430-nm light stimulus used in the experiments produces a 40% to 50% maximum constriction (see the 430-nm dose-response curve in Ref. 6,
Fig. 3), we cannot completely rule out the possibility that circadian control of the PLR amplitude might occur at higher intensities for this wavelength. Furthermore, because the 475 and 430 nm lights used here were not equiluminent, we cannot directly compare their daily profiles. Nevertheless, the comparisons between circadian phases for the same light sources (white or 475nm) clearly demonstrates a circadian control of the PLR in GUCY1* chickens.
We have previously determined the dose-response curves and action spectrum for the PLR in GUCY1* chickens,
6 which revealed that an opsin-based photopigment, with maximum absorption at 484 nm, mediates these responses. In addition, we have also demonstrated that the inner retina of GUCY1* chickens expresses the two melanopsin genes (
Opn4x and
Opn4m) as well as their corresponding proteins.
6,15 These observations, together with the fact that melanopsin is expressed in a circadian manner with higher levels during the light phase in chicken retinal cells,
34 suggest that the circadian PLR shown in this work could be driven by opsin 4. Cryptochromes, which display maximal absorbance at 430 nm, have been associated with some nonvisual photoperception tasks at early developmental stages in chickens, though these responses disappear after birth.
6,28 Moreover, our previous studies suggest that cryptochromes do not act as functional photopigments directly responsible for conducting the PLRs in adult GUCY1* chickens.
6 Nevertheless, we cannot completely rule out this possibility on the basis of the results shown here.
Despite the white and 475-nm blue light stimuli not being equiluminent, the kinetic responses obtained were similar at any CT examined, reflecting that both light conditions were equally efficient for driving daily variations in the PLRs.
Our results also suggest that under the light conditions presented in this work, there are no daily changes in the slope (
τ) of the PLR time courses for GUCY1* chickens, despite
τs being calculated via different fitting approaches.
29 As we mentioned above, there is indeed a daily rhythm in the pupil amplitude during the NSP when the stimulus was white light. A parsimonious explanation for these observations could be that both phases contribute to the whole PLR under white light, reflecting the fact that the amplitude of the whole PLR time courses for some CTs were higher than others, maintaining the slopes.
It is therefore likely that the small PLR observed with a 430-nm light stimulus is because the light source used for these experiments exhibited an emission spectrum peaking at 430 nm, but with a tail overlapping the absorption spectrum for melanopsin (
Supplementary Fig. S1). This observation also explains why the characteristic lifetime (
τ) for the PRL time courses under a 430-nm light stimulus are longer and the PLRs slower (
Fig. 3) than those obtained under the white and 475-nm blue light conditions. In this connection, under white or 475-nm blue light stimuli, the kinetic responses did not differ statistically at any CT examined, reflecting that the two light pulses are equally efficient at driving the PLRs.
Another interesting finding is that GC1 is not essential for the functioning of the photocascade operating in nonclassical photoreceptors located in the inner retina. This strongly suggests that the biochemical photocascade operating in the nonvisual inner retinal photoreceptors does not require the synthesis of cGMP by GC1 as occurs in isolated ipRGCs.
30–32 In the chicken and mammalian retina, GC2 (also known as GC-F) is also expressed; however, levels of cGMP in GUCY1* chicken retinas are 6-fold lower than those in WT animals.
17
It is known that the chicken retina works as an autonomous oscillator independent of the suprachiasmatic nuclei, displaying circadian rhythms in melatonin, dopamine, and photopigments synthesis, variation in the amplitude of electroretinogram components, and glycerophospholipid synthesis—among others
21,22,33–41—reviewed in Guido et al.
42 Moreover, we have previously shown that chicken RGCs also contain autonomous oscillators capable of generating self-sustained rhythms in melatonin synthesis and in the expression and activity of AA-NAT with higher levels during the day under DD, LL, or regular LD cycles.
21,22,31 Strikingly, a number of activities controlled by retinal clocks presumably located in the RGCs were shown to peak during the day, such as the highest sensitivity to white and blue (475 nm) light for the PLRs described in this study and specific ERG parameters: the b-wave amplitude and implicit time which were also higher during the day,
37 when RGC melatonin peaks.
21,22,31
These facts lead us to believe that our results, showing daily changes in pupillary responses under constant dark conditions, reflect truly circadian control of this nonimage photic response in GUCY1* chickens. Recently, a daily rhythm in PLR was also reported in the retinal degenerate mice rd/rd. However, these experiments were performed under a 12-hour light/12-hour dark cycle,
2 so that a light-driven PLR rhythm cannot be completely excluded. Nevertheless, this PLR rhythm, together with the fact that clock mutant backgrounds reduced the PLR sensitivity of rd/rd mice, strongly suggests the circadian modulation of PLR in mice.
2
On the basis of previous results found in the literature, we suggest that at least two independent mechanisms could be involved in the circadian modulation shown here. The first is at the level of the retina itself, since the ipRGCs involved in the control of pupil size may act as independent oscillators controlling this response in humans.
14 Furthermore, the melanopsin photopigment in the chicken inner retina is expressed in a circadian manner with higher levels during the subjective day,
34 regulated by dopamine
43 that, in turn, also exhibit a daily rhythm with higher levels during the light phase.
40,41 We have previously demonstrated that GUCY1* inner retinas express the two genes of melanopsin (
Opn4x and
Opn4m) at the mRNA and protein levels
6,15 and here we observed stronger pupillary responses during the subjective day than the subjective night.
The second mechanism could be indirect circadian control exerted by the suprachiasmatic nuclei on the Edinger-Westphal nuclei, which are critical in controlling pupil size.
44
Overall, our results clearly demonstrate for the first time the circadian control of the PLRs in nonmammalian vertebrates with no functional cones or rods (GUCY1* chickens) and support the hypothesis that the circadian control of PLR may occur in these blind animals, mediated by a vitamin A-based photopigment with a maximum sensitivity around 480 nm.