VISUAL TRANSDUCTION IN RETINA 
King-Wai Yau, Ph.D., Investigator 
The main focus of Dr. Yau's laboratory continues 
to be on the mechanism of visual transduction in 
retinal photoreceptors. Three projects were carried 
out over the past year: 1 ) measurement of dark rod 
phosphodiesterase activity in rod outer segments, 
2) identification of a new cGMP-gated cation chan- 
nel subunit in rod cells, and 3)immunocytochemi- 
cal study of transducin subunits in cones. 
Rod Phosphodiesterase Activity in Darkness 
In darkness there is a steady influx of cations into 
the outer segment of rod and cone photoreceptors 
through a cGMP-activated conductance. Isomeriza- 
tion of the visual pigment by light activates, via a G 
protein called transducin, a cGMP phosphodiester- 
ase that hydrolyzes cGMP, thus leading to the clo- 
sure of the cGMP-gated conductance. This stops the 
cation influx and produces a membrane hyperpolar- 
ization as the neural response to light. In darkness 
there is also a basal level of cGMP turnover due to 
dark cyclase and phosphodiesterase activities. This 
basal cGMP turnover can be an important factor in 
determining a cell's sensitivity to light as well as the 
kinetics of the light response; essentially, a high 
dark turnover rate can translate into a low sensitivity 
to light and fast response kinetics. Thus a difference 
in the dark cGMP turnover rate may underlie some 
of the differences in sensitivity and response ki- 
netics known to exist between rods and cones. Dr. 
Yau's group is comparing the dark phosphodiester- 
ase activities in the two kinds of receptors. So far 
they have measured the activity in rods, by examin- 
ing the ability of exogenous cGMP in darkness to 
activate the cation conductance in a rod outer seg- 
ment (that has been truncated at the base to allow 
entry of externally applied cGMP) in the presence 
and the absence of the drug IBMX, which inhibits 
phosphodiesterase activity. The difference between 
the two situations should thus reflect dark phospho- 
diesterase activity. This comes to a rate of ~l-2/s 
in rods. The dark rate in cones remains to be 
measured. 
New cGMP-gated Cation Channel Subunit 
in Rods 
The cGMP-gated channel, which is unusual 
among ion channels for its choice of ligand, was first 
purified and cloned by others from bovine eye. It is 
a 78-kDa protein with putative transmembrane do- 
mains and a characteristic cyclic nucleotide- 
binding site. Immunocytochemical studies have fur- 
ther confirmed its presence in the rod outer seg- 
ment, and expression in Xenopus oocytes has veri- 
fied its function as a cation channel that is gated by 
cGMP. Thus it appears that the native channel may 
be a homo-oligomeric complex. Recently Dr. Yau's 
group has cloned a protein from human retina that 
shows ~90% homology to the bovine channel. 
When expressed in a kidney cell line, this protein 
also shows cGMP-gated channel activity, suggesting 
that it is probably the human homologue of the rod 
channel. Dr. Yau's group then screened a human 
retinal cDNA library with lower stringency for re- 
lated proteins. The aim was to clone the homolo- 
gous channel in cone cells, which from previous 
work is known to be functionally distinct from the 
rod channel. A distinct clone was obtained that 
showed considerable homology to the rod channel 
clone. The surprise was that when expressed by it- 
self in a cell line, this new clone did not form a 
functional cation channel. On the other hand, when 
coexpressed with the rod channel clone, channel 
activity could be observed that more closely resem- 
bled the native channel activity (recorded from am- 
phibian rods) than with the rod clone alone; 
namely, the flickering openings and closings that 
are characteristic of the native channel are repro- 
duced by the expressed channel in the presence of 
this new clone. Evidence from immunocytochemis- 
try using an antipeptide antibody has further veri- 
fied that the protein encoded by this new clone in- 
deed resides in the outer segments of rods rather 
than cones in the human retina. Thus it appears that 
the native cGMP-gated channel in rods is a hetero- 
oligomeric complex. In retrospect, this is not too 
surprising, because practically all other ligand- 
gated ion channels studied so far appear to be made 
up of more than one subunit species. The hunt for 
the true cone channel subunit(s) is still in progress. 
Cell Biological Studies of the Cone 
Transducin Subunits 
G proteins that mediate signal transduction path- 
ways, such as the phototransduction cascade stud- 
ied by Dr. Yau's group, are heterotrimers composed 
of a, (8, and 7 subunits. The functional specificity of 
a particular G protein has long been thought to re- 
side in the a subunit, with the and 7 subunits 
(particularly the |8 subunit) supposed to play a more 
or less supporting role. Rods and cones have been 
known to have different G protein (transducin) a 
subunits, but the makeup of their other subunits re- 
448 
