HHMI-NIH Research Scholars
Learn about the HHMI-NIH Research Scholars Program, also known as the Cloister Program. More

Janelia Farm Research Campus
Learn about the new HHMI research campus located in Virginia. More

Summary: King-Wai Yau is interested in the detection of sensory stimuli by the nervous system. His laboratory focuses particularly on the first stage of this process, namely, the transduction of sensory stimuli by the respective receptor neurons.

Retinal Rod and Cone Photoreceptors
Vision begins in the rod and cone receptors of the retina. Light triggers a membrane hyperpolarization in these cells, which in turn modulates the release of synaptic transmitter from these cells. The hyperpolarization of rods and cones to light is now well understood. In darkness, nonselective cation channels on the plasma membrane of the receptor's outer segment (the part of the cell that contains the visual pigment) are kept open by the cyclic nucleotide guanosine 3',5'-cyclic monophosphate (cGMP), letting both Na⁺ and Ca²⁺ into the cell. This "dark" current depolarizes the cell and maintains a high level of neurotransmitter (glutamate) release from the cell's synaptic terminal in darkness. Light activates the reaction cascade: light —> photoisomerization of visual pigment —> G protein activation —> cGMP-phosphodiesterase stimulation —> cGMP hydrolysis. As a result, the cGMP level falls, leading to the closure of the cGMP-gated channels, which results in membrane hyperpolarization and reduced glutamate release.

	Mouse retina stained to show ganglion cells...

Rods, which are extremely sensitive to light, can signal the absorption of a single photon. They are responsible for dim-light vision. Cones, which are much less photosensitive, are responsible for bright-light vision. Cones also have a much higher ability to adapt to steady light. The cellular mechanisms underlying these differences in rod and cone physiology are still unclear. Because the various proteins involved in phototransduction have distinct rod and cone versions, the most likely explanation for these differences is that these proteins have somewhat different properties in rods and cones. Recently, we have focused on the different functions of the rod and cone pigments. By expressing human red cone pigment as a transgene in Xenopus, we were able to produce frogs in which the rods express both endogenous rod pigment and transgenic cone pigment. By using monochromatic light that biases the stimulation of either the rod pigment or the cone pigment, we can compare the functions of the two side by side in the same cell. We have found, surprisingly, that rod and cone pigments produce essentially identical responses when present in the same cell. This indicates that, with respect to signaling downstream, rod and cone pigments function identically.

We have also found, however, that cone pigment is much more prone to spontaneous (thermal) isomerization than rod pigment (for red cone pigment, about 10⁴ times more likely). This spontaneous activity is equivalent to a steady light even in darkness, and adapts the cone cell as a result. In other words, a cone behaves like a rod already adapted in darkness. Primate rods and cones differ in sensitivity by about 100-fold in darkness. In principle, the spontaneous activity of red cone pigment can desensitize the red cone by about 10-fold, accounting for about half of the overall difference in rod and cone sensitivity.

In addition, it has long been known that, unlike rod pigment, cone pigment has some tendency to redissociate into the apoprotein (opsin) and the chromophore (11-cis-retinal) in darkness without isomerization of the chromophore to all-trans-retinal. Because free opsin (i.e., without chromophore) has a weak ability to activate the transduction process, this pigment dissociation likewise leads to some steady adaptation of the cone cell, even in darkness. With single-cell measurements, we have found that this pigment dissociation contributes about a factor of 2 to the overall difference in rod and cone sensitivity.

The remaining rod/cone difference in sensitivity, not accounted for by the pigment, most likely resides in the transduction proteins downstream.

Nonrod/Noncone Photoreceptors in the Retina
In addition to detecting and tracking objects in the visual world (what we normally associate with "vision"), our eyes also receive light for other functions (what we call "accessory visual functions"), such as pupillary light reflex and photoentrainment of the body circadian rhythm. It has long been assumed that rods and cones are also exclusively responsible for these functions, because no other retinal photoreceptors have ever been reported. In recent years, however, evidence has suggested the contrary. For example, mice genetically engineered to have lost all rods and cones by degeneration still show the pupil reflex and can be photoentrained. In parallel, several opsin-like proteins have been cloned from the eye, but these proteins are not located in rod and cone photoreceptors. One of these proteins is melanopsin, first cloned from frog skin melanophores and subsequently found by in situ hybridization to be also present in the inner retina of frog and mammals.

With immunocytochemistry and an antibody against rat melanopsin, we have found that melanopsin is present in a small subset (approximately 1 percent) of retinal ganglion cells (RGCs) throughout the rat/mouse retina. Melanopsin is present on the cell body, throughout the dendrites, and even on the proximal part of the axon coursing over the retina. The melanopsin-positive RGC processes form a network that covers the entire retina. Similar immunocytochemistry with an antibody against human melanopsin has shown that these cells are also present in primate retina. To trace the projections of these RGCs, we targeted the tau-LacZ marker gene to the melanopsin gene locus, which allowed us to label the respective axons with X-Gal. In both heterozygous and homozygous knockout animals, the X-Gal–labeled RGC axons project primarily to the suprachiasmatic nucleus (SCN), the intergeniculate leaflet (IGL), and the olivary pretectal nucleus (OPN). The SCN is the seat of the central circadian pacemaker, the IGL is an important relay station of the photoentrainment circuit, and the OPN is a control center for the pupil reflex. Thus, the melanopsin-positive RGCs project to the key brain centers involved in accessory visual functions. By labeling these RGCs retrogradely in live rats by injecting fluorescent rhodamine beads into the SCN (in collaboration with David Berson's laboratory, Brown University), we found that the RGCs are intrinsically photosensitive (i.e., they remain sensitive to light even in the presence of blockers that eliminate all synaptic transmission in the retina); furthermore, these intrinsically photosensitive RGCs are immunoreactive to melanopsin.

In heterozygous mice that still retain one copy of the melanopsin gene (the other replaced by tau-LacZ), the SCN-projecting RGCs are still intrinsically photosensitive. In homozygous knockout animals, however, these cells lose their sensitivity to light, confirming that melanopsin is required for photosensitivity. Furthermore, the knockout animals have an impaired pupil reflex: even in bright light, the pupil I s unable to constrict fully as in wild type. Finally, we have generated mice that lack melanopsin and have nonfunctional rod and cone phototransduction mechanisms. The pupil reflex is essentially eliminated in these mice, and they also fail to show any circadian photoentrainment.

We conclude that melanopsin is part of a bona fide light-detecting system that functions with the rod/cone system to signal light for various accessory functions. Besides the melanopsin-associated system and the rod/cone system, there does not appear to be any other photodetection system in the retina. We do not yet know whether melanopsin itself is a light-detecting pigment or merely an indispensable component of the light-detecting system.

Olfactory Receptor Neurons and Transduction
Odors are detected by olfactory receptor neurons in the nasal epithelium. Olfactory transduction involves binding of odorants to specific odorant receptor proteins on the membrane of the cilia of these neurons. This binding by odorants activates the receptor proteins, which, via a G protein, activate an adenylate cyclase that synthesizes cAMP. As a result, a cAMP-gated, nonselective cation channel opens, depolarizing the cell to firing threshold. Thus, there is remarkable homology between visual and olfactory transductions, despite the great difference in the nature of the sensory stimulus. Whereas visual transduction is understood in quantitative detail, however, olfactory transduction is only understood qualitatively. Little is known beyond the identifications of the main proteins involved in the process. We are using single-cell electrophysiology combined with mouse genetic engineering to dissect the olfactory transduction process in detail.

Some of this work is supported by grants from the National Institutes of Health and the Human Frontier Science Program.

Last updated June 09, 2004