Melanopsin-mediated optical entrainment regulates circadian rhythms in vertebrates

Abstract

Melanopsin (OPN4) is a light-sensitive protein that plays a vital role in the regulation of circadian rhythms and other nonvisual functions. Current research on OPN4 has focused on mammals; more evidence is needed from non-mammalian vertebrates to fully assess the significance of the non-visual photosensitization of OPN4 for circadian rhythm regulation. There are species differences in the regulatory mechanisms of OPN4 for vertebrate circadian rhythms, which may be due to the differences in the cutting variants, tissue localization, and photosensitive activation pathway of OPN4. We here summarize the distribution of OPN4 in mammals, birds, and teleost fish, and the classical excitation mode for the non-visual photosensitive function of OPN4 in mammals is discussed. In addition, the role of OPN4-expressing cells in regulating circadian rhythm in different vertebrates is highlighted, and the potential rhythmic regulatory effects of various neuropeptides or neurotransmitters expressed in mammalian OPN4-expressing ganglion cells are summarized among them.

This review looks at melanopsin across vertebrate species and discusses melanopsin-positive cells in the regulation of circadian rhythm.

Introduction

In the vertebrate retina, the sensitivity of dim-light vision is supported by rod photoreceptors, whereas cone photoreceptors mediate color discrimination and high visual acuity at higher light intensities^1,2. Compared with visual forming visual pathways, the regulation of non-image forming visual pathways is performed by intrinsically photosensitive retinal ganglion cells (ipRGCs), such as circadian entrainment³, pupillary light reflex^4–6, and time-restricted feeding⁷. Although ipRGCs are less represented in mammalian retina (mice: ~1–5%, human: ~0.4–1.5%)^8–12, melanopsin (OPN4), as an opsin, gives it powerful non-image forming function^13,14.

OPN4 is a G protein-coupled receptor initially identified in the dermal melanocytes of Xenopus laevis. It includes an extracellular amino-terminal and seven transmembrane domains with high homology to invertebrate opsins¹⁵. The OPN4 gene has been detected in most vertebrates and analyzed in two lineages, xenopus (OPN4x) and mammalian (OPN4m) orthologs¹⁶. It was also found that there are two OPN4m splice variants in mice and humans, the short (OPN4-S) and long (OPN4-L) isoforms, which differ mainly in the number of phosphorylatable serines and threonines in the C-terminus, which may lead to differences in the inactivation dynamics of OPN4 in different species^16,17. Regarding photosensitivity, the λ_max (peak sensitivity) of OPN4 was 480 nm as measured directly in light response to ipRGCs, confirmed by mouse models lacking rods and cones^13,18,19. Notably, the peak sensitivity of OPN4 shows some minor differences in many studies depending on differences in the detection methods, technology, or species^20,21.

Here, we collate the distribution of OPN4 in mammals, birds, and teleost fish based on published evidence. We then highlight the mechanisms by which the non-visual photosensitization of OPN4 mediates in vertebrate circadian rhythm regulation. Taking the photosensitive activation of OPN4 as a starting point, our review focuses on the mechanism of OPN4-mediated photoentrainment action in circadian rhythm regulation in vertebrates. Admittedly, the other OPN4 activations of G-protein coexist and have been summarized in recent relevant reviews^22–24. In this article, the G_q/11 pathway in the OPN4-mediated phototransduction was mainly described due to its widespread presence in vertebrate ipRGCs^25–28. In addition, OPN4-mediated light entrainment impacts melatonin secretion in the vertebrate retina and pineal gland. This leads to more diverse rhythmic regulatory pathways in non-mammalian vertebrates than mammals.

Distribution of OPN4 in vertebrates

Mammalian

Mammalian OPN4 is derived from a single OPN4 gene with two splice variants in ipRGCs, which has been localized and accurately classified by much evidence. OPN4 is also distributed in mammalian peripheral tissues (Table 1), but its functions remain to be further investigated. Therefore, OPN4 in mammalian ipRGCs will be discussed first.

Table 1.

Photosensitivity of vertebrate OPN4 in peripheral organs.

Taxa	Species	Location	Related effects	References
Amphibia	Xenopus laevis	Melanocyte	Skin pigmentation	Provencio et al.15
Reptile	Hydrophiinae	Skin	Tail phototaxis	Crowe-Riddell et al.27
Mammal	Mouse	Aortas, pulmonary arteries, airway smooth muscle	Light-dependent relaxation	Sikka et al.172; Barreto et al.173; Yim et al.174
Mammal	Mouse	Melanocytes	Pigmentation	de Assis et al.175
Mammal	Human	Mesenchymal stem cells	Angiogenesis	Yang et al.176
Mammal	Human	Subcutaneous white adipose tissue	Lipolysis of lipid droplets	Ondrusova et al.177

The OPN4-expressing ipRGCs were previously thought to be a homogeneous cell population with sparsely branched dendritic trees on the outermost layer of the inner plexiform layer in mammals⁸. Subsequently, the expression of OPN4 in M1-M6 ipRGCs in the mouse retina has been identified^12,29. This OPN4 in ipRGCs with various morphological and physiological characteristics can provide complete light-dark discrimination and partial vision in rodless/coneless (rd/rd cl) mice^30,31. Among these subtypes of ipRGCs, M1 expressed the highest content of OPN4, and it mainly exerts OPN4-induced photoentrainment^32–34. Correspondingly, the suprachiasmatic nucleus (SCN) is innervated primarily by M1-subtype ipRGCs (~80%), and OPN4 in M1-subtype ipRGCs significantly regulates rhythmic regulation in mammals^35,36.

Birds

Mammals lost OPN4x during evolution and chromosomal re-arrangements^37,38, which accompanied mammal adaptation to the nocturnal niche^39,40. In a bird’s retina, two lineages for OPN4 are expressed⁴¹. OPN4m is stably expressed in the retinal ganglion cells (RGCs) during the development of birds, whereas OPN4x was limited to the forming RGCs at embryonic 8 (E8), but mainly expressed in PROX1-positive horizontal cells (HCs) at E15⁴². These OPN4-expressing horizontal cell precursors continue to express OPN4x after migrating and developing into horizontal cells⁴³.

In contrast to mammals, the distribution of OPN4 in birds is no longer concentrated in the retina (Fig. 1)⁴⁴. Bird pinealocytes are directly photosensitive⁴⁵, and the reconstitution of the recombinant proteins with 11-cis-retinal demonstrated that it expresses two lineages of melanopsins⁴⁶. The transcriptional levels of OPN4 in the pineal gland showed a more robust diurnal feature than that in the retina and were significantly increased at night⁴¹. Although avian pinealocytes possess both OPN4m and OPN4x (also called OPN4-1 and OPN4-2 in chickens), their distribution is not cell-specific. It may activate different types of G proteins to perform light-sensing functions⁴⁷. In addition, multiple nuclei composing deep brain photoreceptors in birds also express OPN4 (Fig. 1), including the lateral septal organ, premammillaris nucleus, paraventricular nucleus (PVN), and paraventricular organ⁴⁸. OPN4-positive dopaminergic neurons in these nuclei can respond to daytime length⁴⁹.

Fig. 1. Distribution of OPN4 in teleost fish, amphibian, reptiles, birds, and mammals on generalized sagittal sections.

In teleost fish and birds, two orthologs of OPN4 are distributed in the retina, brain, and pineal gland^{44,61,64,166–168}. For teleosts, amphibians, and reptiles, splice variants of OPN4 are classified as OPN4m and OPN4x. In the reptile brain, OPN4x expression has been detected in the telencephalon, mesencephalon, and rhombencephalon, but the specific nuclei are still unclear. It is important to note that the available evidence does not determine the cell type of OPN4x in the inner nuclear layer or whether OPN4m is present in RGCs in reptiles. Mammalian OPN4 is mainly expressed in the retina, which integrates more complex photosensitive functions and widely projects to different brain regions through different subtypes of ipRGCs to regulate various physiological functions¹⁴. Brackets indicate representative species. AC amacrine cell, AVT area ventralis of tsai, BC bipolar cell, Dm the medial zone of the dorsal telencephalic region, DT dorsal thalamus, HA habenula, HC horizontal cell, ipRGC intrinsically photosensitive retinal ganglion cell, LH lateral hypothalamic nucleus, LSO lateral septal organ, LVII facial lobe, LX vagal lobe, ME median eminence, MGC Muller glial cell, mPON magnocellular preoptic nucleus, nTS nucleus tractus solitarius, PG preglomerular area, PH plexus of horsley, PMM nucleus premammillaris, POA preoptic area, POM medial preoptic nucleus, PTN posterior tuberal nucleus, PVN periventricular nucleus, R raphe nucleus, SCN suprachiasmatic nucleus, SL nucleus septalis lateralis, SP subpallium, VT ventral thalamus, VTA ventral tegmental area.

Reptiles

The studies on OPN4 in reptiles has mainly focused on lizards, sea snakes, and turtles, but there still needs to be more evidence to locate the expression sites of OPN4 and its isoforms accurately. To date, OPN4m was not detected in sea snakes, while OPN4x was mainly expressed in RGCs and cone cells⁵⁰. Although OPN4x-positive staining was also observed in the inner nuclear layer⁵⁰, the cell type could not be determined. In freshwater turtles, OPN4m is highly expressed in the retina, but it is not yet certain whether OPN4m is localized in RGCs⁵¹. In extraretinal photoreceptors, OPN4x is also expressed in the lateral eye and brain of ruin lizards but has not been detected in the pineal gland⁵².

Amphibians

When it was discovered, melanopsin was found in the retina, melanophores, and deep brain photoreceptors of Xenopus laevis¹⁵. Both OPN4m and OPN4x have been localized in RGCs, horizontal cells, and pineal complex^53,54. OPN4-expressing RGCs have been shown to participate in the melanocyte pigmentation process by producing alpha-melanocyte stimulating hormone in the pituitary gland⁵⁵. OPN4 in the pineal complex may participate in the change of skin color through the neuroendocrine pathway⁵⁴. These photosensitive neuroendocrine circuits enable Xenopus to maintain rapid physiological pigmentation change.

Teleost fish

Five splice variants were detected in zebrafish (OPN4.1, OPN4a, OPN4b, OPN4xa, and OPN4xb), which are derived from two melanopsin lineages (OPN4m and OPN4x) to confer overall photosensitivity to the teleost retina and to adapt to the dynamic light environments in the aquatic habitats⁵⁶. Similar to birds, both lineages of OPN4 are expressed in RGCs, and partial OPN4 splice variants are distributed in horizontal cells^57–59, which independently mediates the role of HCs in photosensitive signaling⁶⁰. In extraretinal tissue, OPN4m was detected in the dorsal thalamus, ventral hypothalamus, and nucleus lateralis tuberis pars lateralis; OPN4x was evident in the SCN and habenular nucleus⁵⁹. Evidence for functional partitioning suggests that OPN4m mediates the light-seeking behavior in larvae distributed in the preoptic area⁶¹, whereas OPN4x regulates circadian rhythms in the SCN⁶². The zebrafish pineal gland is a photosensitive structure with various opsins, a subpopulation of pinealocytes capable of sensing shorter wavelength light, characterized by the expression of OPN4x⁶³. In addition, two splice variants of OPN4, OPN4.1 and OPN4xb, were detected in the pineal gland, which is responsible for inhibiting melatonin synthesis during the day and maintaining voluntary movements in a state of absolute arousal⁶⁴.

Overall, current evidence has shown that OPN4 is mainly distributed in the retina of mammals, while it is also widely expressed in the brains of teleost fish, amphibians, reptiles, and birds (Fig. 1).

Light activation of OPN4 in the retina

OPN4 is a G protein-coupled receptor with 11-cis retinal as a covalently bound protonated Schiff base (PSB11)⁶⁵. Under the induction of light, the conformation of 11-cis retinal changed with the transformation of PBS11 to its all-trans isomer, which changed the state of PSB11 to a 7-cis state^66–68. In this series of changes, the 11-cis and 7-cis retinal indicate OPN4’s silent state, while the all-trans structure indicates light signaling conversion⁶⁸. This tristability confers on OPN4 a sustained response to light and a broader spectrum of its own⁶⁹. Following this reaction, the G_q/11 class of G-proteins will become active and further trigger the activation of phospholipase C-beta 4 (PLCβ4). This leads to the hydrolysis of phosphatidylinositol 4,5-bisphosphate to form inositol triphosphate and diacylglycerol through the transient receptor potential cation channel subfamily C member 6/7 (TrpC6/7) nonselective cation channels in the cell membrane and finally increases the intracellular Ca²⁺ concentration (Fig. 2)^14,21,23. Using calcium ion probes, Sekaran et al. consistently found that OPN4 can specifically respond to a wavelength of 470 nm with a significant increase in Ca²⁺ concentration in a mouse model lacking cone and rod photoreceptors⁷⁰. In addition, a recent study demonstrated that internally released Ca²⁺ marks the opening of the OPN4-mediated light-sensitive pathway⁷¹, which is the opposite of the hyperpolarization of rods and cones¹⁴. In the opsin photosensitive response termination, OPN4 is subject to C-terminal phosphorylation. Its phosphorylation process preferentially interacts with G protein-coupled receptor, kinase 2/3 (GRK2/3), preventing OPN4-expressing ipRGCs from generating sustained action potentials after light stimulation^72–74. Meanwhile, arrestin is also involved in the inhibition and reactivation of the light response of OPN4. When the C-terminus of OPN4 is phosphorylated, it can bind to arrestin. β-arrestin 2 primarily regulates the deactivation of OPN4, whereas β-arrestin 1 initiates regeneration of OPN4^75,76. The above responses allow ipRGCs to sustain responses under prolonged illumination (Fig. 2).

Fig. 2. Activation and termination of OPN4 in M1-subtype ipRGCs of mammals.

The OPN4-mediated light-sensitive pathways are predominantly triggered by the downstream G_q/11, PLCβ4, and TRPC6/7 cation channels in mammals. Retinaldehyde is covalently bonded to the transmembrane structure in OPN4, and light (especially near 480 nm) can change its conformation from an 11-cis to an all-trans state to a 7-cis state (silent state). It will trigger downstream G_q/11 coupling, causing PLCβ4 to break down PIP2 into DAG and IP3, where DAG activates the opening of the TRPC6/7 cation channels. The activated C-terminus of OPN4 is phosphorylated in response to GRK2/3, resulting in inactivation. This process may also involve β-Arrestin 2. In addition, β-Arrestin 1 leads to the isomer regeneration of OPN4, which serves subsequent light activation. DAG diacylglycerol, G_q/11 G protein subunit alpha q/11, GRK2/3 G protein-coupled receptor kinase 2/3, PLCβ4 phospholipase C-beta 4, PIP2 phosphatidylinositol bisphosphate, IP3 inositol triphosphate, TRPC6/7 transient receptor potential cation channel subfamily C member 6/7.

Due to the complexity of the G protein family and the variation of OPN4 subtypes in different species, the optical signal transduction of the OPN4 pathway is mainly dependent on G_q/11 in M1-subtype ipRGCs in mammalian and partial OPN4-expressing cells in non-mammalian vertebrates^{25,28,47,77,78}. Recent studies have shown that adenylyl cyclase 2 and cAMP mediate the phototransduction of OPN4 in M4-subtype ipRGCs²⁴. Considering the ability of retinal adenosine to influence photosensitive electrophysiological activity in the retina⁷⁹, the effect of cAMP on OPN4 phototransduction cannot be ignored. Admittedly, OPN4-mediated phototransduction mechanisms have also been implicated in species that involve G_i/o (human, mouse, and amphioxus), Gs (chicken), or Gt (chicken) activity^26,47,80.

Photosensitive regulation of circadian rhythms by OPN4

OPN4-induced non-visual photosensitive signals can target numerous nuclei, including the SCN, intergeniculate leaflet, ventral lateral geniculate nucleus, and olivary pretectal nucleus (OPN)⁸¹. The SCN is the center for orchestrating mammalian circadian rhythms, while the OPN is essential for regulating the pupillary light reflex^82,83. The synaptic structures at the ends of these tracts specialize in different nucleus regions, leading to differences in threshold sensitivity, speed, and accuracy of visual responses in these nuclei⁸⁴. As we mentioned above, M1-subtype ipRGCs form the core of non-visual photosensitivity. According to the molecularly defined Brn3b transcription factor expression, the M1 subtype consists of two distinct subpopulations. The majority of projections from M1 ipRGCs to the thalamus and midbrain are Brn3b-positive M1-subtype ipRGCs⁸⁵, which regulate OPN4-dependent pupillary light reflexes and light-induced acute body temperature changes^83,86,87. The SCN is innervated by Brn3b-negative M1-type ipRGCs⁸³, and it is here that the SCN orchestrates multiple oscillators with a duration of almost 24 h^88,89. Therefore, the non-visual function of OPN4 contributes to controlling central rhythms in mammals via Brn3b-negative M1-subtype ipRGCs projections in the SCN region.

Unlike mammals, chick retinal ganglion cells were classified into six subgroups according to their somal and dendritic characteristics (subgroups Ic, Is, IIc, IIs, IIIs, and IVc)^90,91. The subgroups IIs and IIIs had a more significant proportion of thalamic projection⁹². Identifying the function of RGCs from chickens is still challenging, despite similarities in RGC projection pathways to the brain between birds and mammals. Brn3b molecular markers commonly used in mammals may not be suitable for birds. All types of Brn3 factors (Brn3a, Brn3b, and Brn3c) can promote the differentiation of chick RGCs and are not mainly regulated by Brn3b as in mammals⁷⁷. Furthermore, species differences make it challenging to directly administer current antibodies and viral vectors to birds’ retinas or central nervous systems. Nevertheless, studying retinal ganglion cell subtypes in chicks may be more effectively accomplished using in vivo transfection or electroporation transfection^93–95. For non-mammalian vertebrates, the pineal gland of birds and teleost fish has rhythmic pacing functions and is involved in constituting the multi-oscillatory circadian timing system^96,97. The photosensitization of photoreceptors in the retina by OPN4 may have limited effects on circadian rhythms in these species. Therefore, when discussing OPN4-mediated non-visual photosensitive functions, the pineal gland of non-mammalian vertebrates will also be emphasized.

It should be noted that cones and rods can affect not just the local biological clock of the retina^98,99, but also the master clock of the SCN¹⁰⁰. During development, ipRGCs form functional connections with the cone/rod system in the inner reticular layer, allowing them to serve as relays to transmit collected rod and cone information to the brain while retaining their intrinsic photosensitivity^30,101. Photoentrainment induced by rods can influence the master clock via cone circuits, which may complement the function of photoentrainment in ipRGCs in dim light¹⁰². Accordingly, the light power required to activate OPN4 (>1 μW) under in vitro conditions is higher than conventional retinoids (~0.2 μW)¹⁰³. At the same time, ultraviolet (λ_max 365 nm) and green (λ_max 505 nm) sensitive cone cells are also able to indirectly influence the electrophysiological activity of the neurons in the SCN via ipRGCs, contributing to photoentrainment^100,104. Additionally, harmonizing the photosensitive signals from the cones, rods, and ipRGCs also plays a crucial role in ensuring the pupillary light reflex functions properly¹⁰⁵. Thus, the influence of cones and rods on circadian rhythm regulation should not be undervalued.

Contribution of OPN4 to mammalian circadian rhythms

Exposure to monochromatic blue light (460 nm) can suppress human melatonin levels and interfere with resetting circadian rhythm^106,107. As part of this regulation, the photosensitive signal of OPN4 is first transmitted to the SCN through the retino-hypothalamic tract (RHT), followed by the paraventricular nucleus and the intermediolateral nucleus via the polysynaptic circuit distributed in the SCN region, and finally to the release of melatonin innervated by the sympathetic nerve in the superior cervical ganglion (SCG)^108,109. In addition to this approach, OPN4-positive ipRGCs can rely on self-synthesized neurotransmitters and neuropeptides to more directly and rapidly affect the SCN master clock.

Retinal glutamatergic signals are responsible for transmitting external light information to the SCN, and binocular enucleation induced a significant decrease in vesicular glutamate transporter 2 (Vglut2) immunoreactivity in the ventrolateral part of the SCN¹¹⁰. The experiments in OPN4^Cre/+::Vglut2^flox/flox transgenic mice proved that the glutamate transmission from ipRGCs is necessary for light to entrain circadian rhythms in dim light¹¹¹. Regarding synaptic connections, glutamatergic ipRGCs have neural projections with many photosensitive neurons in the SCN. Many Vglut2-immunoreactive axons were observed to be in synaptic contact with vasoactive intestinal peptide (VIP)- and gamma-aminobutyric acid (GABA)-positive neurons¹¹². ipRGCs have direct synaptic connections with arginine vasopressin (AVP) neurons in the dorsal SCN¹¹³. Glutamatergic signaling primarily controls the expression of clock genes concerning the regulation of the SCN master clock. The glutamatergic activation of the N-methyl-d-aspartic acid (NMDA) receptor leads to an influx of extracellular Ca²⁺, followed by Ca²⁺/calmodulin-dependent kinase II and nitric oxide synthase activation^114,115. Then, the increased nitric oxide levels activate ryanodine receptors (RyRs) in the intracellular endoplasmic reticulum¹¹⁶. Finally, intracellular Ca²⁺ is released by activated RyR, leads to phosphorylation of cAMP response element-binding (CREB) protein, and regulates transcription of period and cryptochrome by CLOCK and BMAL¹¹⁷. During the maintenance of the circadian rhythm, the transcription factor CREB can integrate photosensitive information and mediate the reset of the circadian rhythm¹¹⁸. It is undeniable that the strength of this OPN4-mediated glutamatergic signaling is different in species with diurnal activity patterns, which is also reflected in their nonidentical phase-response curves (PRC). The projection of ipRGCs-SCN in the Nile rat (Arvicanthis niloticus) is comparable to that of the Syrian hamsters¹¹⁹. However, there are differences in sensitivity to phase movement between the two species on the NMDA-induced PRC^120,121, which is also reflected in the strong resistance of Arvicanthis niloticus to NMDA¹²².

ipRGCs also express a peptide neurotransmitter called pituitary adenylate cyclase-activating peptide (PACAP) and colocalize with glutamate at the terminals of RHT in the SCN^123,124. Previous studies have shown that adding PACAP to SCN slices in wild mice at circadian time (CT) 6 can advance the peak of the SCN activity rhythm in this and subsequent circadian rhythms¹²⁵. However, the phase and amplitude of the neuronal firing rhythm do not change in Adcyap1 (adenylate cyclase activating polypeptide 1, encoding PACAP) knockout mice at CT6 and CT7 in the SCN¹²⁶. In addition, light stimulation in the early night (CT15) delayed the phase, while light stimulation in the late night (CT21) advanced the phase¹²⁷. Consistently, the influence of PACAP on the circadian rhythm depends on glutamate in the late night (phase advance), and the independent regulation of circadian rhythms by glutamate occurs in the early night (phase delay)^126,128,129. The time-dependence phase shift at night may be due to PACAP and glutamate acting on different SCN neuronal subpopulations. Compared to glutamate, the positive signals for PACAP were mainly distributed in the dorsomedial SCN and a small amount in the central/ventral SCN^126,130. Using c-Fos to mark neuronal activity, neurons with significant light responses during the subjective daytime were distributed in the dorsal SCN, and light did not affect the rhythm phase of mice¹³¹. This phenomenon is consistent with the evidence that circadian rhythms are not altered in Adcyap1 knockout mice. Regarding regulatory mechanisms, PACAP has a regulatory effect on glutamatergic calcium signaling and has a different time window from glutamate in regulating CREB phosphorylation^132,133. PACAP can regulate circadian rhythm by differentially regulating mitogen- and stress-activated protein kinase 1 phosphorylation downstream of p42/44 mitogen-activated protein kinase between day and night¹³⁴. Therefore, in terms of the autonomous rhythm of SCN neurons, PACAP may be a supplementary factor to OPN4-mediated SCN mastering circadian clock rhythms in response to the risk of potential circadian imbalance underlying Vglut2 deficiency when glutamate stimulation alone is insufficient.

Most OPN4-containing cells also expressed vasopressin (VP), which has glutamatergic nerve fibers projecting to the non-visual nuclei of the brain, and the application of VP receptor antagonist decreases the response of SCN neurons to photic entrainment of the RHT¹³⁵. Additionally, vasopressinergic axons can affect the activity of ventral SCN cells in a VP-dependent manner¹³⁶. Applying the antagonists of vasopressin V1a and V1b receptors to the SCN can promote (near instantaneous) re-entrainment to the new light/dark cycle¹³⁷. Current evidence suggests that VP⁺ ipRGCs have synaptic co-localization with gastrin releasing peptide (GRP)- and VIP-positive neurons, but VP⁺ ipRGCs are not directly connected in AVP neurons¹³⁵. According to single-nucleus RNA sequencing assays, all AVP clusters expressed glutamate receptor subunits with minimal expression of GABA receptors. However, some of their AVP nonlight-responsive clusters could express VIP receptor type 2¹³⁸. Therefore, we speculate that photosensitive AVP, GRP, and VIP neurons may be downstream neurons of ipRGCs when OPN4-expressing ipRGCs secrete both glutamate and VP.

The effect of GABA on the master clock can be excitatory or inhibitory in different contexts, but there is no doubt about its importance^139,140. The light information transmitted by ipRGCs can induce oscillations in the GABAergic system in the SCN. This may be because GABA_B receptors are highly localized ventral to the SCN and are closely related to the signal afferents and the terminal synaptic remodeling of RGCs¹⁴¹. Studies in hamsters have shown that the antagonism of either GABA_A (ionotropic) or GABA_B receptors (metabotropic) in the SCN significantly increases the phase-shifting effects of light induction before a light pulse is provided in the early night rather than in the late night, suggesting that the inhibition of phase shift by extracellular GABA occurs mainly in the early night^142,143. These data proved that changes in GABA in the SCN region are synchronized with non-visual light signals in the RGCs. Through GABAergic signaling, OPN4-expressing ipRGCs can also preserve circadian stability. Some ipRGCs co-expressing Gad2 and OPN4 can transmit GABAergic signals to the SCN to inhibit excessive light entrainment, and neurons receiving these GABAergic signals contain some VIP neuronal subsets¹⁴⁴. Correspondingly, VIP neurons maintain the regular operation of the circadian rhythm, and inhibiting VIP neurons leads to increased phase shift¹⁴⁵.

Therefore, the optical signal mediated by OPN4 in ipRGCs is transmitted to the SCN via RHT. Light entrainment is mainly determined by glutamatergic transmitters and supplemented by multiple neuropeptides in the SCN to adjust the phase shift and intensity (Fig. 3). Simultaneously, GABAergic neurotransmitters may act as inhibitors in this terminal region. These inputs may prevent unnecessary adjustments of the master circadian clock in the SCN by external environmental light. Notably, the expression time of neuropeptides does not match the timing of the phase shift caused by it (such as PACAP)¹²⁶. Considering that neuropeptides need to undergo an extended length of RHT (mice: ~10 mm; rat: >20 mm) after they are synthesized from the cell body to the SCN region, they are transported only about ~140 mm per day along axons^84,146. Therefore, when researching circadian rhythms, it would be interesting to look into the rate of transmission in the RHT and the rhythm of these neuropeptides’ expression in the retina.

Fig. 3. The light entrainment of OPN4 on circadian rhythms may involve multiple neural projection pathways, including neurotransmitters or neuropeptides.

ipRGCs are a class of retinal ganglion cells that express OPN4 (red) and can transmit OPN4-mediated photosensitive signals to the SCN via RHT projections. a Vglut2, but not Vglut1, packages glutamate (solid blue circles) into synaptic vesicles in these axons. These ipRGCs axons mainly make synaptic contacts with VIP neurons (green), AVP (pink), and other light-responsive neurons (gray) in the SCN^169,170. The VIP neurons form part of the SCN core region and may communicate with AVP neurons via VIP receptor type 2. b Some ipRGC axons can also release PACAP (solid brown circles) to regulate VIP neurons via VPAC₂ and PAC₁ receptors¹⁷¹. In addition, some OPN4-expressing ipRGCs also expressed VP (solid pink circles). c The axons of these ipRGCs are glutamatergic and VP-positive, and light stimulation can affect their secretion of VP. VP⁺ ipRGCs showed synaptic co-localization with GRP (yellow) and VIP neurons, but VP⁺ ipRGCs were not directly connected to AVP neurons¹³⁵. d Some ipRGCs expressed GAD2 and could transmit GABA (solid purple circles) to the SCN regions. These GABAergic signals can excite or inhibit some SCN neurons, including VIP neurons, and maintain the homeostasis of the central rhythms¹⁴⁴. AVP arginine vasopressin, GAD2 glutamic acid decarboxylase 2, GABA γ-aminobutyric acid, GRP gastrin releasing peptide, ipRGC intrinsically photosensitive retinal ganglion cell, RHT retinal hypothalamic tract, SCN supra-chiasmatic nucleus, VIP vasoactive intestinal peptide, VP vasopressin, PACAP pituitary adenylate cyclase-activating peptide.

Contribution of OPN4 to circadian rhythms in birds

The central biological clock system of the bird is formed by the hypothalamus, retina, and pineal gland¹⁴⁷. The non-visual photosensitization in GUCY1* chickens, a null mutation chicken model that causes blindness at hatching, is more complex than that in rd/rd cl mice. When blocking the input of light signals from the head, blocking the perception of light by opsins in the SCN and pineal gland directly reduces the effect of light drive on circadian rhythms^6,148. On this basis, GUCY1* chickens showed a feeding rhythm disruption after enucleation⁶. This finding implies that the retina plays an essential role in maintaining the circadian rhythm in chickens. Additionally, when only hypothalamic photoreceptors were retained, GUCY1* chickens were still able to maintain a brief circadian phase shift to adapt to the next light-dark cycle, indicating that photoreceptors from the hypothalamus may play a role in light regulation of circadian feeding behavior¹⁴⁸. Although the evidence presented above does not exclude the possibility that other opsins have non-visual effects, the unique modulation of eating rhythm in GUCY1* chicken is particular to the light with wavelengths near the maximum absorption peak of OPN4. Another noteworthy example under light stimulation is that the pupillary light responses of chickens follow a circadian rhythm comparable to that of mammals¹⁴⁹. It has been observed that GUCY1* chickens can maintain the circadian rhythm of pupillary light responses and reach maximum sensitivity at 480 nm¹⁵⁰. Considering the light-absorbing properties of OPN4 and the distinct nonvisual photosensitivity function, the above findings strongly indicate that chicken retinal OPN4 regulates circadian rhythm. Notably, the photosensitivity function of OPN4 in the chicken retina may be more complex than is currently known. Chicken horizontal cells express OPN4x, which controls the release of GABA and regulates the membrane potential of photoreceptors following photosensitive activation⁴³. Although this function is oriented more towards vision modulation, it cannot be excluded that OPN4x-expressing HCs may also affect the non-visual photosensitivity of OPN4-expressing retinal ganglion cells (RGCs) through signaling crosstalk.

The pineal gland of birds shows a robust melatonin secretion rhythm in vivo and in vitro. Monochromatic blue light (480 nm) can advance the phase of the rhythm-negative regulatory genes and inhibit the mRNA levels of Cry1 and Aanat (a key enzyme in melatonin synthesis) in the pineal gland, both in vivo and in vitro^151,152. Since the specific membrane receptors for melatonin are distributed in the SCN region, melatonin can act directly on the SCN in an endocrine form to regulate the clock rhythms in the SCN¹⁵³. Compared with the pineal gland, the chicken retina is a relatively independent organ in the circadian rhythm, and pinealectomy does not alter the circadian oscillations in the retina¹⁵⁴. The main effect of monochromatic blue light on the retinal circadian clock is to delay the phase of OPN4 rather than the phase shifts of clock genes or the mRNA levels of Aanat¹⁴⁷. The SCN is the primary retinorecipient hypothalamic structure in birds¹⁵⁵. When OPN4 in the chicken retina is excited by light, its non-visual light signals are mainly transmitted to the SCN. Then, the axons emitted from the SCN regulate downstream nuclei, such as the PVN and the infundibular nucleus (similar to the mammalian arcuate nucleus)¹⁵⁶. It has been demonstrated that the hypothalamic appetite-related genes show a circadian rhythm¹⁵⁷. Is it possible that ipRGCs expressing OPN4 might indirectly regulate the appetite of broilers through their neural projections to the SCN? Further investigation of the relationship between the non-visual photosensitive function of OPN4 and the feeding rhythm will help answer this question.

Contribution of OPN4 to circadian rhythms in teleost fish

Although the light-sensing mechanism of an extraretinal photoreceptor is unclear, it may represent the most basic approach to light-sensing^44,158. The eye and pineal gland are the main central clock structures for zebrafish, which conduct autonomous oscillations, photoreception, and melatonin production¹⁵⁹. The eomesa-expressing RGCs and pineal gland in zebrafish both express Opn4.1 and Opn4xb^160,161. When knocked out the Opn4.1 and Opn4xb in the zebrafish, genes involved in phototransduction and tryptophan metabolism were significantly altered, resulting in increasing melatonin synthesis⁶⁴. Meanwhile, by affecting the synaptic plasticity in hypothalamic neurons or directly acting on melatonin receptors distributed in the hypothalamic SCN, the light signals can control the circadian rhythms through OPN4 in zebrafish^162–165. Therefore, OPN4 expressed in the retina and central nervous system constructs photosensitive sensing in zebrafish and regulates circadian rhythms through melatonin.

Conclusions

OPN4 is a member of the G protein-coupled receptor family. Mammalian OPN4-expressing ipRGCs also express a variety of neurotransmitters and neuropeptides, which together with OPN4 regulate circadian rhythms. In contrast to mammals, teleost fish and birds have a more complicated system for controlling their circadian rhythms, and OPN4, which is expressed in the retina, brain, and pineal gland, is crucial for photosensitivity.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

Conceptualization, J.C., D.P.; writing—original draft, D.P.; supervision, Z.W., Y.C.; project administration, J.C., Y.C. All authors have read and agreed to the published version of the manuscript.

Peer review

Peer review information

Communications Biology thanks Mario Guido and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary handling editor: Manuel Breuer. A peer review file is available.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-023-05432-7.