Va
max and Vb
max are
frequently used to correlate with ONL cell counts as an indication of
the stage of retinal degeneration.
Figure 4 shows data from lines 1 and 3 plotted together. The change in log
V
max of both the a- and b-waves was proportional
to ONL cell loss (a-wave,
R = 0.94; b-wave,
R = 0.89;
Fig. 4 ), but Va
max slope was approximately double that of Vb
max,
indicating that Va
max is more sensitive for
detecting photoreceptor cell loss. Note that in end-stage disease the
lowest two points deviate downward from these straight lines, as
expected, because when all cells are lost, both
Va
max and Vb
max must head
toward “zero.” What seems remarkable is the good fit of these
regression lines across data pooled from both lines 1 and 3 comprising
a more than 80% range of cell loss. This suggests that the
degeneration process is similar across these two lines for all but
end-stage ONL cell loss.
The a-wave threshold also showed a log increase with ONL cell loss
(
R = 0.93, not shown). As described for
Figure 3b -wave threshold remained essentially static with age within each of
lines 1 and 3, and by threshold measurements, the a-wave was more
sensitive than the b-wave for tracking cell loss.
In a previous study, saturated amplitude of fast PIII was proportional
to ROS length for albino rats reared under a different lighting
intensity.
23 In that study, ROS length was manipulated
through the process of photostasis to two light-rearing intensities.
The P23H rhodopsin mutant animals provide an opportunity to extend this
observation to include the case of cell loss in addition to ROS
shortening. A-wave saturation occurs at the flash intensity sufficient
to close all cGMP-gated channels and to completely interrupt the rod
dark-current. If one assumes that channel density per unit area of ROS
plasma membrane remains constant despite a change in ROS length, then
the total dark-current per rod would be proportional to ROS length, as
found by Reiser et al.
23 However, in the P23H rat, total
retinal photocurrent will also be proportional to the number of rod
cells remaining. Consequently, for the P23H rat, the maximum
photocurrent change on complete channel closure with a saturating ERG
flash would be proportional to the product of ROS length and ONL cell
count (ROS × ONL).
Figure 5A provides support for this idea by the tight linear correlation between
Va
max and ROS × ONL across both lines 1 and
3 (
R = 0.96). Note that, like the Ganzfeld ERG, the
histology represents a global retinal assessment, because ROS and ONL
were averaged across the entire retina. These data are well fitted
across a 100-fold range for both Va
max and
ROS × ONL, and the line regresses to zero a-wave voltage for zero
ROS × ONL. This suggests that ROS plasma membrane channel density
is maintained during degeneration. Neither ROS nor ONL alone gives a
direct relationship of Va
max. ONL is related to
the logarithm of Va
max (Fig. 4) and
consequently deviates considerably from direct proportionality. The
same is true for ROS versus Va
max (not shown but
as one might anticipate from the linear relationship of ROS with ONL
shown in
Fig. 2B ).
Figure 5B shows that b-wave amplitude is linearly correlated with the
a-wave (
R = 0.95), and tracking the b-wave may provide
a substitute for a-wave recordings, particularly for higher amplitude
responses. The tight correlation indicates linearity of visual signal
transfer across the rod to bipolar synapse. However, this relationship
fails for end-stage disease when the a-wave is greatly reduced, and the
b-wave persists even when the a-wave voltage approaches zero. The
b-wave offset from zero must be interpreted carefully, because this
cannot be taken to imply that rod signals reach the proximal retina
despite a complete loss of rods. This may reflect signal divergence
from the very few rods that remain and signal amplification by the
second and third order neurons.
24 Alternatively, signals
might derive from remaining cone cells, because cones contribute very
little to the a-wave. In any case, the b-wave amplitude appears to be a
more complex and consequently less direct measure of residual rod
function in end-stage disease.