The purpose of this study was to determine whether the ON response
of the ERG of the cone system is reduced differentially relative to the
OFF response in XLRS. The patients with XLRS who were included in this
study represented a relatively broad range of b-wave amplitudes of the
ON response, from response amplitudes that were within normal limits to
amplitudes that were markedly attenuated
(Fig. 6) . Regardless of the
extent of b-wave attenuation, however, all the patients with XLRS
showed a reduction in the b-wave to d-wave ratio compared with the
control subjects
(Fig. 7) . The reduced b-wave to d-wave ratios
demonstrate that there is a greater deficit in the ON response than in
the OFF response of the cone system in XLRS.
For the primate cone system, the initial portion of the b-wave of the
ON response represents primarily the activity of depolarizing bipolar
cells (DBCs), with the later portion of the b-wave modulated by the
response of hyperpolarizing bipolar cells (HBCs).
32 Intravitreal administration of 2-amino-4-phosphonobutyric acid (APB),
which blocks the DBC light response, virtually eliminates the
b-wave.
32 33 Therefore, the significantly reduced b-wave
to d-wave ratio of the cone system that we observed in these patients
with XLRS likely represents a predominant response attenuation within
the DBC pathway.
Of note, these patients with XLRS did not show an enhancement of the
d-wave that is frequently observed in other forms of retinal disease in
which there is a differentially reduced ON response.
19 The
initial portion of the d-wave of the OFF response represents the
activity of HBCs together with the offset of the cone photoreceptor
response.
32 However, the waveform characteristics are also
modulated by the activity of DBCs,
32 so that a reduced DBC
response would be expected to enhance the amplitude of the d-wave. The
absence of an enhanced d-wave, together with the slightly reduced
d-wave amplitude observed in two of these patients with XLRS
(Fig. 6) ,
suggests that there may be some impairment within the HBC pathway as
well as within the DBC pathway.
In addition to an ON-response defect, the patients with XLRS in this
study had significantly increased implicit times for all ERG waveform
components
(Fig. 5B) . Increased implicit times of the a-wave and b-wave
of the cone system have been noted before in XLRS,
9 10 and
an increased implicit time of the d-wave was reported
recently.
34 Based on a recent study,
35 it is
likely that the increased ERG implicit times of patients with XLRS
result from a “sluggish” cone photoreceptor response. According to
that study, the frequency-response function of an early linear filter
(presumed to represent the cone photoreceptor response) showed a lower
corner frequency in patients with XLRS than in control subjects. A
lower corner frequency corresponds to a broadened impulse–response
function, with a delayed response peak.
36 Such a delayed
cone photoreceptor response could account not only for the increased
implicit time of the a-wave, which represents primarily cone
photoreceptor activity,
37 but also the prolonged implicit
times of the other ERG components that depend on the cone photoreceptor
response.
The precise explanation for the ON-pathway defect that we observed in
the ERG of the cone system in XLRS remains to be resolved. It is
possible that the attenuated b-wave amplitude of the ON response
directly reflects dysfunctional DBCs, given current evidence that the
b-wave represents the summed activity of DBCs.
11 12 An
alternative possibility is that the reduced ON response is a
consequence of the abnormal cone photoreceptor response described
above. Although there are no apparent differences in the response gains
of DBCs and HBCs, there are systematic differences in their temporal
response properties, as well as differences in their signal
transduction mechanisms.
38 Therefore, it is conceivable
that an abnormal cone photoreceptor input could have a differential
effect on the response properties of the two bipolar cell types, so
that there is an apparently greater response deficit within the DBC
pathway.
It is also possible that the ON-response deficit in XLRS is a secondary
effect of Müller cell dysfunction on the DBC pathway. As noted in
the introduction, histologic studies have reported evidence of abnormal
Müller cells in XLRS.
2 3 4 Müller cells have an
important role in maintaining the homeostasis of extracellular
glutamate and potassium levels (reviewed by Reference 39). It has been
proposed that the golden-white fundus sheen that has been observed in
some patients with XLRS,
28 29 and was apparent in one of
our patients (patient 6), results from an abnormal accumulation of
extracellular potassium.
28 Changes in extracellular
potassium levels can influence neuronal responses, and high levels of
glutamate are potentially toxic to neural elements.
39 Therefore, the relatively greater attenuation of the b-wave of the ON
response in XLRS may represent a greater susceptibility of DBCs to
abnormalities in the extracellular environment, in which Müller
cell dysfunction may play a role.
The relationship between the abnormal ON response of the ERG of the
cone system in XLRS and mutations in the XLRS1 gene remains to be
identified. As noted in the Introduction, the XLRS1 gene is expressed
primarily in photoreceptor cells.
13 14 15 The gene product
is a protein (retinoschisin) that is secreted by the photoreceptors and
is involved in cell–cell interactions, particularly within the inner
retina and probably including Müller cells.
15 This
protein has also been reported to be associated with photoreceptor and
cone bipolar cell membranes.
14 Thus, there are a number of
candidate mechanisms by which mutations in a gene that is expressed in
photoreceptors could affect the ON response of the cone system ERG in
XLRS.
An ON-response defect of the cone system has been observed in several
other forms of retinal disorders in which, as in XLRS, the brief-flash
ERG of both rod and cone systems shows a reduced b-wave to a-wave
ratio.
16 17 18 19 It has been proposed that an overall
impairment in the response properties of DBCs could account not only
for the relative attenuation of the b-wave of the cone system in these
disorders, but also for the b-wave attenuation of the rod
system,
17 18 21 because the primary rod pathway involves
signal transmission through rod bipolar cells, which are of the
depolarizing type.
40 A similar explanation may account for
the relative attenuation of the b-wave that has been observed for both
rod and cone systems in XLRS.
5 6 7 8 9 10 However, there are
important differences between XLRS and these other disorders that show
an ON-response deficit. For example, the b-wave attenuation of rod and
cone systems is not as severe in patients with XLRS as it is in
patients with other disorders, such as CSNB and MAR.
18 Further, in contrast to patients with these other disorders, patients
with XLRS typically have normal or near-normal rod-mediated absolute
thresholds,
6 8 despite the moderately attenuated rod
b-wave amplitude and a markedly reduced scotopic threshold
response.
8 Therefore, the relationship between the
abnormal ERG responses of the rod system and the ON-response deficit of
the cone system in XLRS requires further clarification.
Sawtooth flicker was used as a stimulus for eliciting ERG ON and OFF
responses in the present study, with the goal of minimizing the
potential eye movement artifacts that can particularly affect the ERG
OFF response to long-duration flashes.
21 22 The sawtooth
stimulus was successful in this regard. The ON and OFF responses
elicited by sawtooth flicker showed good intrasubject consistency
across stimulus cycles. This stimulus also has several properties that
are potentially useful in assessing the characteristics of the ERG ON
and OFF responses. For example, sawtooth flicker has the same
time-average luminance for both increments and decrements, so that an
equivalent adaptation state is maintained for each. The temporal
frequency of the stimulus can be varied to assess the
frequency–response characteristics of ON and OFF responses. The two
waveforms are inversely related but otherwise identical, so that
differences in the ERG responses to the two stimuli can be used to
investigate retinal nonlinearities. Thus, sawtooth flicker may be of
value in quantitative studies of ERG ON and OFF responses in various
forms of retinal disease.
In summary, the patients with XLRS in this study showed an attenuation
of the b-wave of the ON response of the cone system that was greater
than the attenuation of the d-wave of the OFF response, regardless of
overall ERG amplitude. This finding indicates that there is a
relatively greater impairment within the DBC pathway than within the
HBC pathway. The precise explanations for the attenuated ON response in
XLRS and the relationship to mutations in the XLRS1 gene remain to be
determined.
The authors thank Deborah J. Derlacki for assistance in subject
testing.