Shedding light on photoperiodism

In mammals, photoreception is restricted to the eye. By contrast, nonmammalian vertebrates possess photoreceptors in a wide array of extraocular tissues, including the pineal complex and brain (1). The identity of the deep brain photoreceptors has remained elusive. They continue to be the least studied of all nonvisual photoreceptor classes, although their existence in the vertebrates has been known for nearly a century. In this issue of PNAS, the Yoshimura laboratory presents a study that represents a significant stride toward identifying and characterizing encephalic photoreceptors (2).

History of Encephalic Photoreception

In 1911, von Frisch (3) recognized in the European minnow that light-induced changes in skin coloration occurred even in the absence of eyes and the photoreceptive pineal gland. However, crude lesions of the diencephalon abolished this photoresponse, leading him to propose the existence of encephalic photoreceptors. In birds, the presence of such photoreceptors was conclusively shown in 1944 through the work of Benoit and Ott (4), who showed that direct illumination of the hypothalamus through fine quartz rods (the fiber optics of the day) could induce testicular growth in blinded ducks. It was not until the late 1960s that the pioneering research of Menaker (5) indicated that light-driven reproductive responses in birds could be mediated entirely by photoreceptors outside of the retina and pineal. His group showed that, in addition to regulating avian reproduction, encephalic photoreceptors also synchronize the phase of daily activity rhythms to the day:night cycle (6, 7).

Precisely pinpointing the anatomical site of the avian encephalic photoreceptors and identifying the photopigments that render these photoreceptors light-sensitive have proven difficult. Several groups have attempted to localize these cells immunohistochemically. Rhodopsin-like immunoreactivity was initially observed in the tuberal hypothalamus and septal forebrain regions of the pigeon (8, 9). Later studies indicated that melanopsin and vertebrate ancient opsin (VA-opsin) were found in the avian brain (10, 11).

Role in Photoperiodism

Determining the photopigment responsible for a given physiological response to light can serve as a guidepost by which the cells harboring the photopigment can be identified and characterized. Action spectrum analysis (i.e., establishing which wavelengths of light optimally elicit a photoresponse) is one method by which photopigments can be linked to particular physiology. Foster et al. (12) showed in the quail that the light-dependent increase in serum luteinizing hormone (LH), a prerequisite for testicular growth, was greatest in response to blue-green wavelengths, suggesting the involvement of a rhodopsin-like photopigment (12).

Nakane et al. aggregate the findings of the past into a realistic working model of photoperiodism in birds.

LH is released into the circulation from the anterior pituitary in response to gonadotropin-releasing hormone (GnRH). GnRH arises from a sparse cohort of neurons that terminate in the median eminence of the mediobasal hypothalamus. Here, the GnRH is secreted into the hypophyseal portal system, ultimately leading to the release of LH from the anterior pituitary. Importantly, the secretion of GnRH from nerve terminals in the median eminence is gated by glial cells called tanycytes (13, 14). The Yoshimura laboratory had suggested previously that this gating is dependent on the light-induced activation of thyroid hormone through the enzymatic conversion of the inactive T4 form to the active T3 form (15, 16). The enzyme responsible for this conversion, type 2 deiodinase, is expressed in tanycytes in response to thyroid stimulating hormone (TSH). TSH is sensed by the endfeet of tanycytes after it has been secreted from thyrotropes located in the pars tuberalis.

The big question remained: how is light regulating TSH and consequently initiating this complex hormonal signaling cascade that ultimately leads to an activated reproductive axis? In PNAS, Nakane et al. (2) describe cells in the paraventricular organ of the hypothalamus that express the photopigment Opn5. They show that these presumed photoreceptor cells send projections to the external zone of the median eminence, which is juxtaposed to the par tuberalis, the site of TSH release. Consolidating these findings, Nakane et al. (2) propose a model of photoperiodism in birds that encompasses the entire process from photoreception to gonadal induction.

The high amino acid sequence homology of Opn5 to other opsin sequences certainly suggests that it functions as a photopigment protein (1, 17). Nakane et al. (2) confirm that Opn5 is indeed light-sensitive by heterologously expressing it in Xenopus oocytes and recording light-induced ionic currents (2). Oocytes not expressing Opn5 showed no photoresponses. They completed an action spectrum and determined the sensitivity of Opn5 to peak in the violet wavelengths (420 nm). This is significantly more blue-shifted than the sensitivity of the encephalic photoreceptors described by Foster et al. (12) (492 nm). Nakane et al. (2) measure testicular size in response to various wavelength exposures, and their analysis shows that UV (UVB and UVA) and blue wavelengths stimulate gonadal growth as effectively as white light, a finding consistent with the involvement of Opn5 (2). These results, however, do not rule out the possibility that a blue-sensitive photopigment, such as melanopsin or perhaps, VA-opsin, mediates the photoperiodic response. Opsins maximally sensitive to blue light still absorb significantly in the violet and UV wavelengths (18). Therefore, it remains a possibility that the UV and blue wavelengths tested by Nakane et al. (2) stimulated testis growth through photopigments other than Opn5. As previously mentioned, VA-opsin has indeed been identified in the developing (19) and adult avian brain (11); it has also been suggested (albeit not tested) that a TSH-mediated photoperiodic response, virtually identical to the one presented by Nakane et al. (2), is initiated by VA-opsin activation (11).

Role of Opn5

Therefore, the historical pendulum has swung from a time when no encephalic photopigments were definitively identified to the current state where at least three photopigments are known to reside in the brain. What are all of these deep-brain photopigments doing? Those localized to distinct areas of the brain may mediate particular responses to light. For example, photopigments found in the tuberal hypothalamus may mediate the photoperiodic reproductive responses, whereas those found in the septal forebrain mediate the photoentrainment of circadian activity rhythms. This argument, however, cannot be applied to Opn5 and VA-opsin; these opsins are expressed in the same anatomical neighborhood in the bird brain (2, 11) and may be coexpressed.

Possessing multiple photopigments could broaden the spectral range to which a photoreceptor cell is sensitive, thereby increasing the probability of photon capture under diverse ambient-light conditions. Alternatively, one of the photopigments may be playing an auxiliary role in light reception. Opsin-based photosensory pigments are G protein-coupled receptors that detect light, because they possess a covalently bound 11-cis-retinaldehyde chromophore that functions as a photolabile ligand (20). On absorption of a photon, the 11-cis-retinaldehyde is photoisomerized to all-trans-retinaldehyde (21). This conformational change activates the photopigment and initiates a signaling cascade. After this occurs, the photopigment is spent and rendered useless for additional signaling. The chromophore must be regenerated by replacing the all-trans isomer with another 11-cis chromophore. Photopigments of invertebrates are capable of photoconverting all-trans-retinoids back to 11-cis in situ through the absorption of another photon (21). The photopigments of vertebrates are incapable of such in situ regeneration and must relinquish their all-trans-retinaldehyde and incorporate a new 11-cis-retinaldehyde molecule.

Multiple mechanisms have evolved to do this. A specialized class of opsin has emerged whose sole function is to incorporate the spent all-trans chromophore and use the energy of an absorbed photon to photoisomerize the retinoid back to the 11-cis configuration (22). The 11-cis chromophore eventually is used to regenerate the photosensory opsins, rendering them capable of initiating another round of phototransduction.

Peropsin (23) and retinal pigment epithelium-retinal G protein receptor opsin (RGR-opsin) (24) are two examples of such auxiliary opsins (referred to as photoisomerases). Interestingly, several published comparisons of opsin amino acid sequences have grouped Opn5 among the photoisomerases rather than the photosensory opsins (17, 22, 25). Furthermore, the intron:exon structure of the Opn5 gene more closely resembles that of peropsin and RGR-opsin (17). These findings raise the possibility that, in the avian brain, Opn5 does not function as the photopigment directly mediating the photoperiodic response but rather, maintains the sensitivity of an attendant photosensory protein, perhaps VA-opsin, by ensuring that sufficient 11-cis retinoid is available.

The relative importance of the various opsins in encephalic photoperiodic signaling remains to be established. In the meantime, the photosensory/photoisomerase capacity of Opn5 can be determine empirically by analyzing its chromophore isomer preference as has been done for other opsins (23, 26). Nevertheless, the strides made by the Yoshimura lab (2) over the last decade to understanding the molecular and cellular basis for avian photoperiodism have been significant. Nakane et al. (2) aggregate the findings of the past into a realistic working model of photoperiodism in birds. This model is likely to shed light on photoperiodic responses of seasonally breeding mammals (27, 28), which, despite lacking encephalic photoreceptors, also respond to changing day lengths.