Doi:10.1016/j.cub.2006.01.016

Current Biology Vol 16 No 2
R44
Viagra slows the
visual response
to flicker
Andrew Stockman1,
Lindsay T. Sharpe1,
Adnan Tufail2, Philip D. Kell3
and Glen Jeffery1
As an undesirable side effect,
sildenafil citrate (Viagra) partially
Target radiance (log quanta s–1 deg–2) Target radiance (log quanta s–1 deg–2) inhibits the phosphodiesterase
PDE6 [1], which plays an essential
role in phototransduction
(reviewed in [2,3]). PDE6 not only
activates the visual transduction
cascade, but also controls the
time over which the visual
response is integrated, thereby
helping to maintain the size of the
visual response within an optimal
range as the light level is
increased [4]. Consistent with an
increase in temporal integration
following Viagra ingestion,
performance improvements have
Figure 1. Critical flicker fusion (A,B) and modulation sensitivity (C,D) data for L.T.S.
been reported for an identification
(A,C) and G.J. (B,D).
task that relies on the integration
A centrally fixated flickering target of 4° of visual angle in diameter and 650 nm in wave-
of two temporally separated
length was presented in the centre of a 9° background field of 481 nm. The background
stimuli [5]. Using more
of 8.26 log quanta s–1 deg–2 at the cornea was there primarily to saturate the rods. The
conventional psychophysical
mean radiance of the 650 nm target was varied for the c.f.f. measurements and fixed at
measures of temporal sensitivity
9.52 log quanta s–1 deg–2 for the modulation sensitivity measurements. In each panel,
the pre-Viagra baseline data (dotted open circles) are the averages of three or more
and resolution [6,7], we find that
separate measurements made before drug ingestion. The shapes of the coloured
therapeutic 100 mg doses of
symbols denote the dose numbers: dose 1 (triangles); dose 2 (circles); dose 3 (inverted
Viagra can cause mild to
triangles); and dose 4 (hexagons). The time after dose ingestion is coded by the colour
moderate transient losses in
of the symbols (see key). The times noted are the midpoints of the time periods after
human temporal visual sensitivity.
dose ingestion during which a given curve was measured (from the first setting made
These losses imply that in the
at the end of a 3 min pre-adaptation period to the last setting). A c.f.f. curve required on
average 12 min to complete, whereas a modulation sensitivity curve required 8 min.
more affected observers, Viagra
Crossed symbols distinguish between two runs made during the same colour-coded
may provide a unique tool for
time window.
pharmacologically manipulating
the activity of PDE6 in humans in
known as the temporal resolution
Below the plateau, the losses
vivo, and thus for investigating the
limit or critical fusion frequency
increase slightly with target
role of PDE6 in light adaptation. At
(c.f.f.) — as a function of the
radiance, reaching ~10 Hz for
the light levels tested, the
radiance of the 650 nm target.
L.T.S. and 5 Hz for G.J. (coloured
frequency-dependent sensitivity
Under these conditions of
symbols).
losses caused by the inhibition of
adaptation, flicker detection is
To characterize the losses at
PDE6 by Viagra are consistent
mediated by long-wavelength
frequencies below the c.f.f., we
with an almost doubling of the
sensitive (L–) cones, and to a
measured modulation thresholds
time over which visual events are
lesser extent by middle-
as a function of temporal
normally integrated (from ~6.9 to
wavelength sensitive cones (M–)
frequency. Subjects were asked
12.6 ms, assuming a single
cones. The results are shown in
to adjust the target modulation
integration stage).
Figure 1 for observers L.T.S. (A)
until the apparent flicker was just
Observers were presented with
and G.J. (B). With increasing
visible. Modulation is adjusted by
a flickering red (650 nm) target
target radiance, their baseline
varying the fraction of the light
superimposed in the centre of a
(pre-Viagra) c.f.f. functions (open
that flickers, while keeping the
moderate-intensity blue (481 nm)
dotted circles) rise steadily until
time-average intensity of the light
background field. They were
reaching a plateau between 38
constant. Figure 1 shows
asked to set the highest temporal
and 52 Hz. Ingestion of Viagra
temporal modulation sensitivities
frequency at which the flicker
induces moderate losses in c.f.f.
for L.T.S. (C) and G.J. (D)
appeared to be just visible — also
for both subjects at all levels.
measured at a 650 nm radiance
Magazine
R45
change occurs at a single
By inhibiting PDE6, Viagra
integrating stage with an
lengthens the time over which the
exponential decay (which we
visual response is integrated,
assume reflects the activity of
thus impairing the overall ability
some biochemical process or
of the eye to light adapt and to
processes, such as the
regulate its sensitivity.
hydrolysis of cGMP catalyzed by
PDE6). We model this change by
Acknowledgements
applying a standard equation for
This work was supported by a
a leaky integrator [8] and
Wellcome Trust awarded to A.S. We
determining the change in
thank Sabine Apitz, Bruce Henning and
integration time that will best
Rhea Eskew for helpful advice.
account for our data (see
Supplemental Data available on-
Supplemental data
line). Given that the mean data
Supplemental data including
for L.T.S. (dotted circles) and G.J.
Experimental Procedures are available
(grey dotted squares) are similar,
at http://www.current-biology.com/
we have fitted this model
cgi/content/full/16/2/R44/DC1/
simultaneously to both mean
data sets. The model has three
References
parameters: a baseline
Wallis, R.M., Casey, J., Howe, L.,
integration time, a Viagra-
Leishman, D., and Napier, C.M. (1998).
induced integration time and a
Characterisation of retinal
frequency-independent
phosphodiesterase (PDE) isozymes and
sensitivity scaling factor.
the effects of sildenafil in vitro.
Ophthalmic Res. 30 (suppl. 1), 111.
The best-fitting model is shown
Stryer, L. (1988). Molecular basis of
by the continuous red line in the
visual excitation. Cold Spring Harb.
lower panel. The best-fitting
Symp. Quant. Biol. 53, 283–294.
parameters (± the asymptotic
Pugh, E.N., Jr., Nikonov, S., and Lamb,
Figure 2. Modulation sensitivity losses
standard error for each
T.D. (1999). Molecular mechanisms of
vertebrate photoreceptor light
for L.T.S. (A) and G.J. (B).
parameter) are 6.91(±2.64) ms for
adaptation. Curr. Opin. Neurobiol. 9,
The data, from Figure 1, are plotted rela-
the baseline integration time,
410–418.
tive to the mean baseline measurements
12.57(±3.79) ms for the Viagra-
Nikonov, S., Lamb, T.D., and Pugh, E.N.,
(horizontal lines). Symbol and symbol
induced integration time, and
Jr. (2000). The role of steady
colours are as in Figure 1. The continu-
0.34(±0.05) log
unit for the
phosphodiesterase activity in the
ous lines are the mean losses, plotted in
kinetics and sensitivity of the light-
(C) for L.T.S. (dotted circles) and G.J.
scaling factor. The scaling factor
adapted Salamander rod
(grey dotted squares), which also shows
is needed to compensate for
photoresponse. J. Gen. Physiol. 116,
the model fit (red line).
changes in sensitivity caused by
795–824.
the background becoming more
Mollon, J.D., Regan, B.C., Foo, R., and
Morris, B.J. (2003). Sildenafil increases
of 9.52 log quanta s–1 deg–2
effective when the integration
persistence of vision. Invest.
(Figure 1, vertical dashed lines).
time is lengthened (by 0.26 log10
Ophthalmol. Vis. Sci. 44 (suppl), 4097.
Consistent with the changes in
unit for a 6.91–12.57 ms change
Hecht, S., and Verrijp, C.D. (1933). The
their c.f.f. data, both L.T.S. and
in time constant). This increase in
influence of intensity, color and retinal
G.J. show some losses of
effectiveness is partially offset by
location on the fusion frequency of
intermittent illumination. Proc. Natl.
modulation sensitivity after
the comparable increase in
Acad. Sci. USA 19, 522–535.
ingesting Viagra at most
sensitivity to low frequency
De Lange, H. (1958). Research into the
frequencies. The relative losses
flicker, but not entirely so (a
dynamic nature of the human fovea-
compared with baseline are
complete offset would be
cortex systems with intermittent and
highlighted in Figure 2 for LTS (A)
consistent with Weber’s Law).
modulated light. I. Attenuation
characteristics with white and colored
and GJ (B) for data collected
The fact that the overall
light. J. Opt. Soc. Am. 48, 777–784.
between 60 and 166 min after
modulation sensitivity drops by
Watson, A.B. (1986). Temporal
Viagra ingestion, in which it
0.08 log10 represents a partial
sensitivity. In Handbook of Perception
becomes clear that the losses for
failure of light adaptation as
and Human Performance, Volume 1, K.
both subjects increase with
Viagra presumably drives the
Boff, L. Kaufman and J. Thomas, eds.
(New York: Wiley), pp. 6–1-6–43.
temporal frequency. This
system into a less optimal
steepening suggests that Viagra
(nonlinear) part of its operating
is associated with a lengthening
range. The data are clearly
of the visual integration time.
consistent with a lengthening of
1Institute of Ophthalmology, University
We can model the mean losses
the integration time, but the
College London, 11-43 Bath Street,
London EC1V 9EL, UK. 2Moorfields Eye
(continuous lines in Figure 2A,B
precise values of the time
Hospital, 162 City Road, London EC1V
and symbols in 2C) and thus
constants are only moderately
2PD, UK. 3Archway Sexual Health
obtain a rough estimate of the
well constrained by the model,
Clinic, Whittington Hospital Highgate
change in the time constant of
and should be considered as
Hill and Institute of Urology, Middlesex
integration by assuming that the
preliminary estimates.
Hospital, London N19 5NF, UK.

Source: http://cvrl.ucl.ac.uk/people/Stockman/pubs/2006%20Viagra%20time%20CB%20SSTKJ.pdf