Overview 

Nearly all of the current work in the lab is devoted to a rare type of retinal ganglion cell that is a true photoreceptor, capable of autonomously transducing light energy into neural responses. These intrinsically photosensitive retinal ganglion cells (ipRGCs) use the photopigment melanopsin to generate remarkably sustained depolarizing light responses that encode the level of illumination. Further background information on these cells is provided here.

Phototransduction. 

Phototransduction is the complex biochemical process by which a photoreceptor cell absorbs light energy and emits a physiological response.  In the case of ipRGCs, this process begins with the absorption of photons by the photopigment melanopsin and culminates in the gating of a cationic conductance in the plasma membrane.  This results in membrane depolarization and the generation of action potentials.  A continuing focus in the laboratory is to understand this process in detail, from the behavior of melanopsin itself, through the second messenger systems it activates, through the light-gated conductance.  We have recently been able to show that this phototransduction cascade bears an uncanny resemblance to that in invertebrate photoreceptors, especially that in the fruit fly Drosophila.   Light causes melanopsin to activate a G protein in the G_q/11 family, which in turn activates the effector enzyme phospholipase C (PLC).  The entire cascade appears to be within or tightly linked to the plasma membrane, because even excised patches of ipRGC membrane retain their photosensitivity. Many aspects of the cascade remain poorly understood and we are working on three facets of this problem:

Photopigment bistability  -  The photopigments in animal photoreceptors consist of a protein, called an opsin, and a chromophore, which is derived from Vitamin A.  Absorption of a photon triggers the isomerization of the chromophore from 11-cis retinaldehyde to all-trans retinaldehyde, causing a conformational change in the opsin which initiates the G protein signaling cascade.  In vertebrate rods and cones, the activated photopigment is unstable and quickly dissociates into its opsin and retinaldehyde components.  Regeneration of the photopigment requires a complex sequence of enzymatic reactions many of which are carried out in the retinal pigment epithelium (RPE) a tissue directly adjacent to the neural retina.  Melanopsin, however, has been reported to differ from rod and cone opsins in retaining its all-trans retinaldehyde after photoexcitation, just as invertebrate opsins are known to do.  In the excited form of such ‘bistable’ photopigments, light has been shown to trigger a reisomerization of the chromophore from all-trans to 11-cis, thus restoring the original photosensitive pigment (‘photoreversal’).  We are currently developing physiological evidence that this process does indeed occur under physiological conditions in ipRGCs and that is has functional consequences for their behavior.   

Light-gated channel  -   There has been much speculation that the ion channel(s) responsible for the photocurrent in ipRGCs belong to the TRP family.  This would be in keeping with the similarities of phototransduction in ipRGCs and Drosophila photoreceptors, because the light-gated channels in fly photoreceptors are TRP and TRPL, the founding members of this diverse family of channels.  However, we still do not know with certainty whether photocurrent in ipRGCs is carried by TRP channels and, if so, which specific members of this channel family are responsible.  Nor do we understand how these channels are gated.  The involvement of PLC as the effector enzyme would seem to implicate either inositol trisphosphate (IP3) or diacylglycerol (DAG), but we have obtained evidence against both of these as second messengers.  This raises the possibility that the key signaling event is actually the reduction in the effector enzyme’s substrate, PIP₂, a membrane phospholipid. 

Inactivation mechanisms  -   All sensory receptors must have mechanisms to shut off their responses once the stimulus is withdrawn.  By analogy with other animal photoreceptors, we presume that in ipRGCs this process involves phosphorylation of the melanopsin protein by a G-protein receptor kinase (GRK) followed by the binding of arrestin to uncoupling the activated photopigment from its cognate G protein.  We are working to determine whether this does indeed occur and, if so, the identity of the specific GRKs and arrestins. 

Synaptic inputs of ipRGCs. 

While they are not dependent upon synaptic inputs for their photosensitivity, ipRGCs do, in fact, receive such inputs in the inner plexiform layer (IPL), the synaptic layer in which ganglion cells receive synaptic contacts.  The ipRGCs clearly receive both excitatory glutamatergic inputs from bipolar cells and inhibitory GABAergic and glycinergic inputs from amacrine cells.  A peculiarity of the excitatory inputs is that they are dominated by the ON channel (although most cells also receive a weaker OFF input).  This appears to violate the dogma that the dendrites of ON center ganglion cells stratify in the inner (ON) sublayer of the IPL, where ON bipolars terminate.  We are working to understand this paradox, and to provide a detailed accounting of the types of bipolar cells that influence ipRGCs.  We also want to understand which types of amacrine cells modulate the ipRGCs.  We are particularly interested in the dopaminergic amacrine cells, because their dendrites costratify with those of the ipRGCs and because these amacrine cells play a key role in mechanisms of modulation of retinal physiology by lighting history (adaptation) and by circadian phase. 

Intraretinal outputs of ipRGCs. 

The conventional view of ganglion cell organization is that these cells integrate signals from lower order retinal neurons and communicate the resulting output signal through the optic nerve to visual centers in the brain.  However, several convergent lines of evidence have begun to suggest that ipRGCs may actually give rise to a ‘centrifugal’ signal within the retina, capable of modulating lower order retinal neurons.  In collaborative studies with the laboratories of Doug McMahon at Vanderbilt and Gary Pickard at Colorado State, we have obtained strong evidence that ipRGCs drive light responses in a subset of dopaminergic amacrine cells. Because dopaminergic amacrines are central to retinal mechanisms of adaptation and circadian modulation, ipRGCs are in a position to regulate these mechanisms through their outputs to dopaminergic amacrine cells.  We are working to bolster the evidence for this influence, to probe its synaptic basis, and to understand its functional implications.

Multiple types of ipRGCs. 

Nearly all prior studies of ipRGCs in rodents have described a single morphological type with a broad, sparse dendritic arbor largely restricted to the outermost part of the inner plexiform layer.  However, in collaboration with Iggy Provencio’s group at Univ. Virginia, we have found that sensitive anti-melanopsin antibodies stain an additional morphological type stratifying in the inner IPL. The presence of two plexuses of melanopsin immunoreactive dendrites at the inner and outer margins of the IPL has been observed in primates as well by Dacey and colleagues.  We are working to characterize the morphology of the inner-stratifying type of melanopsin expressing cells and to determine whether they might be functional photoreceptors  -   a second type of ipRGCs.  If they are photoreceptors, we plan to compare and contrast their functional properties and central projections with those of the established type of ipRGCs.  

Circadian modulation of ipRGCs. 

The ipRGCs are perhaps best known for their role in synchronizing the master circadian pacemaker in the hypothalamus to the rising and setting of the sun.  However, the retina is an autonomously circadian tissue, with daily oscillations in the expression of core clock genes and strong circadian rhythms in physiology.  There have been suggestions that melanopsin expression may be under circadian control, but as yet there has been no attempt to determine whether this translates into circadian variations in the sensitivity of these key players in central circadian entrainment.  We are testing this by characterizing the irradiance-response functions of ipRGCs at different circadian times.

Whole-cell voltage clamp recording of an isolated cultured ipRGC. In control conditions (black), light stimulus (bar) evoked a large inward current. The response was abolished by bath application of an inhibitor of the effector enzyme PLC (red).

Recordings from excised patches of the plasma membrane of ipRGCs. Image at top left shows the recording pipette. The excised patch at the tip is essentially invisible. Light evoked a marked inward current (middle trace) and a train of action potentials (bottom trace), depending on the recording method. Inset at upper right shows three spikes on a faster time base. The spikes are abolished by the selective sodium-channel blocker tetrodotoxin (not shown).