Several of our key findings are listed below, but of these I consider the discovery of non-rod, non-cone retinal photoreceptors to be the most important to date. 

1) The discovery and characterisation of a third class of photoreceptor in the vertebrate eye.

Little more than a decade ago discussion that the eye might contain a non-rod, non-cone photoreceptor generated either polite amusement or hostile rebuttal by most vision scientists. Since the eye had been the subject of intense research for two centuries it seemed inconceivable that such a system could have been overlooked. The dogma was that all photoreception took place in the rods and cones of the outer retina whilst the cells of the inner retina provide the initial stages of signal processing prior to complex visual processing in the brain. However, two separate lines of study from my group showed that the inner retina also contains photosensory neurones.

1A – Studies in Mammals: In the early 1990’s we used retinally degenerate and transgenic animal models to understand how the circadian system in rodents is regulated by light.  The approach used mice which carry gene defects resulting in blindness (greatly diminished or undetectable visual responses) and monitor the impact of this loss on the ability of these rodents to adjust (entrain) their circadian rhythm system to a light/dark cycle. We showed that despite massive rod and cone photoreceptor loss these mice were not only able to entrain their circadian rhythms, but could do so with the same sensitivity as fully sighted animals (4). These, and a host of subsequent experiments (1), showed that the processing of light information by the circadian and visual systems was different and suggested that there might be another class of photoreceptor within the eye.  But because we could not preclude the possibility that only a very small number of rods and/or cones are required for photoentrainment, we engineered mice (rd/rd cl) in which all the rods and cones were functionally ablated. Circadian entrainment (6), the regulation of pineal melatonin (10, 11), and a variety of other responses to environmental irradiance (e.g. pupil constriction, (9), were preserved in rd/rd cl mice and these data provided unambiguous evidence for a third class of ocular photoreceptor, quite different from the rods and cones, within the mammalian retina. To localise these photoreceptors Mark Hankins, Sum Sakaran, Rob Lucas and I utilised the isolated rd/rd cl mouse retina in combination with Ca²⁺ - imaging techniques. This approach showed that the retina contains a plexus of electrically coupled photosensitive ganglion cells (pRGCs) and that Ca²⁺ is likely to play an important role in the phototransduction cascade (16).  Most recently this approach has shown that these pRGCs are photosensitive at birth and convey light information to the brain (17).  Parallel studies on the rd/rd cl mouse employed action spectroscopy to characterise a novel opsin/vitamin A photopigment (OP) with a maximum sensitivity in the “blue” part of the spectrum (_λmax 479nm/ OP⁴⁷⁹) (7, 9). Although we had deduced the biochemistry of the photopigment, the molecular identity of OP⁴⁷⁹ remained a mystery. In collaboration with collegues in the USA, notably King-Wai Yau and Samar Hatter we then showed that melanopsin is likely to be this photopigment. Melanopsin is expressed in the pRGCs, and its genetic ablation in mice lacking all functional rods and cones abolishes circadian and pupillary responses to light (7, 12). The functional properties of melanopsin have been assessed very recently by members of our group (and independently by two other laboratories in the USA) by combining the expression of melanopsin protein with physiological assays of cellular photosensitivity. Remarkably, Mark Hankins has shown that melanopsin can confer photosensitivity to a variety of non-photosensitive cell types (see Melyan et. al. 2005). Our recent and unpublished work has shown that pRGCs regulate multiple areas of physiology including sleep propensity and cardiac function (work in progress). Finally, novel microarray approaches developed by Stuart Peirson and studies in gene knock-out models by Henrik Oster in the group have identified new and unexpected proteins in the melanopsin phototransduction cascade (15).

1B – Studies in Teleost Fish:  In parallel with our studies in mammals, we discovered non-rod, non-cone ocular photoreception in fish by using a very different set of approaches. In 1997 we isolated a novel opsin gene family from teleost fish that we termed the VA (vertebrate ancient)-opsins (18). This gene encodes a functional VA opsin photopigment and is expressed in a sub-set of retinal horizontal and amacrine/ganglion cells (19). Since this discovery pre-dated our findings in mammals it represents the first demonstration of a non-rod, non-cone ocular photopigment in any vertebrate. Subsequent collaborative work with Mark Hankins has combined electrophysiological, molecular and anatomical approaches to study the cell biology of these novel retinal photoreceptors. They respond to environmental irradiance, integrate light information from rods and cones, and contain a photopigment with a maximum sensitivity in the “blue” part of the spectrum (_λmax 477nm) (8). 

Significance of these discoveries: The eye has been considered the best characterised part of the central nervous system. The fundamental questions about the eye were considered answered, with only the details left to resolve. Our discovery of novel ocular photoreceptors in mammals and fish has forced a major reassessment of how the eye processes light information to regulate a variety of different photosensory tasks, and it is likely that much of this work will have important clinical implications. Not least on the classification of human blindness. Ophthalmologists are now beginning to appreciate the full consequences of eye loss, a state that deprives an individual of both their sense of space and time (also see 5 – below). 

2) The discovery of opsin/vitamin A based photopigments in extraretinal photoreceptors. 

Non-mammalian vertebrates possess a diverse complement of photoreceptors including the pineal organ, deep encephalic photoreceptors, and dermal/peripheral tissue photoreceptors. Although these photoreceptors were first recognised in the early part of the 19^th century, the basis for their photosensitivity remained (and to a degree still remains) poorly understood. My work provided overwhelming evidence that these diverse photoreceptors use broadly conserved mechanisms based upon opsin/vitamin A photopigments. Two of my early Nature papers, summarise these conceptual breakthroughs: (i) the photosensitive dermal pigment cells of certain teleost fish utilize an opsin-based photopigment system (13). This was shown by generating antibodies against visual pigment opsins, and localising opsin protein in the photosensitive membranes of the dermal iridophores.  In 2003 we isolated a new opsin gene family (tmt-opsin) from fish that probably encodes the specific opsin for this peripheral tissue photosensitivity (14). This work is ongoing;  (ii)  A study published in 1985, addressed a long-standing question in seasonal physiology. In the 1930’s birds had been shown to use a photoreceptor located deep within the brain to detect daylength changes and regulate seasonal physiology, but nothing was known of how this photoreceptor might function. By using action spectroscopy we established that these encephalic receptors utilize an opsin/vitamin A based photopigment system (2). We were also able to show how the detection of these daylength changes are translated into neuroendocrine changes of the reproductive axis (3).

3) Circadian rhythms and schizophrenia.

Patients with schizophrenia frequently complain of poor sleep and are commonly observed to shift their rest-activity cycle. Despite the introduction of new antipsychotic drugs, sleep-wake abnormalities remain a major complaint. Of the few previous studies undertaken, all have suggested that abnormal rest-activity or EEG sleep profiles in schizophrenic patients are associated with an increase in severity of psychotic symptoms, a decrease in cognitive performance, and poor psychosocial outcomes. Katharina Wulff and I, working with Eileen Joyce (UCL) and Derk-Jan Dijk (Surrey) have initiated a detailed study of this phenomenon. Our results show that the circadian timing of the sleep-wake profile and endocrine rhythms are either severely delayed or free-running with respect to time of day in patients with schizophrenia. This work is currently being prepared for publication. We feel that a greater understanding of circadian disturbance in schizophrenia will not only increase our understanding of the neurobiology and neurogenetics of the disorder, but also provide the substrate for the development of clinical and pharmacological interventions. This will improve rehabilitation potential and the quality of life of patients and their families. Future studies will address whether a correction of these sleep-wake abnormalities will result in a reduction in psychotic episodes and the abnormalities of neurocognition, emotion and social isolation that are intrinsic to the disorder (5).