Figure 14.6. Drawing by Ramon y Cajal of the cell layers in a vertebrate retina.
The top of this drawing is the part of the retina closest to the interior of the
eyeball and the vitreous humor. Light enters the retina from the top and
passes through several layers of cells before interacting with the rod and cone
photoreceptors. At the fovea, these other cell layers are folded back, allowing
incoming light to have unobstructed access to the photoreceptor cells—the
pit structure of the fovea.Even a single photon of light is enough to generate a signal from a
rod cell. However, because the rhodopsin molecules are so sensitive
to light, they are right at the threshold for having isomerization of
retinal occasionally occur randomly as a result of thermal agitation.
Thus, ganglion cells will often receive signals that have not been
generated by light but, rather, are background noise. To avoid sending
signals to the brain that are not related to actual visual information,
the ganglion cells wait for coincidences of several signals from one or
more photoreceptors. This increases the minimum number of pho-
tons necessary to generate an action potential from a ganglion cell
that is passed along the optic nerve to something between five and
ten. This is still a small number of photons, corresponding to very,
very dim light.
Relatively recently, in the 1990s, another kind of photoreceptive
retinal opsin protein was discovered in light-sensitive cells in the
skin of frogs. This protein, named melanopsin, was shortly thereafter
found in certain ganglion cells in the vertebrate (including human)
retina. These photosensitive retinal ganglion cells send their axons
into regions of the brain involved in the regulation of pupil size and
the synchronization of circadian rhythms (see Chapter 20).
An important property of neural cells in the visual system—begin-