Light-initiated changes of cyclic-GMP levels in the frog retina measured with quick-freezing techniques

Abstract

Although there is good agreement that light reduces the amount of cyclic GMP (cGMP) in the retina, the exact time-course of this decrease is not well established. Bullfrog retinal sections were isolated under infrared light and quick-frozen with liquid nitrogen-cooled, metal hammers after exposure to various intensities of continuous illumination. This quick-freezing should stop the degradation of cGMP within 50-100 ms. The frozen retinal sections were then slowly warmed up in the presence of perchloric acid to denature enzymes involved in cGmp metabolism. cGMP was determined by radioimmunoassay and comparison was made between light- and dark-adapted retinal sections from the same animal. The average cGMP concentration was 44.3 +/- 0.7 pmol cGMP/mg protein or 170.9 +/- 3.2 pmol cGMP/retina. After 1 s of illumination no significant change in cGMP concentration was found even with the brightest light used (approximately 7 x 10(7) rhodopsins bleached/second per rod. At this intensity the first significant decrease in cGMP from dark-adapted levels was detected 3-5 s after the initiation of illumination; cGMP decayed to 70-75% of the dark-adapted value after approximately 30 s. With lower intensity illumination the cGMP levels recovered to dark-adapted levels after the initial decrease even though the bleaching light remained on.

Full text

Light-Initiated Changes of Cyclic
Guanosine Monophosphate Levels in
the Frog Retina Measured with
Quick-Freezing Techniques
PAUL KILBRIDE and THOMAS G. EBREY
Department of Physiology and Biophysics, University of Illinois at Urbana-Champaign, Urbana,
Illinois 61801
ABSTRACT Although there is good agreement that light reduces the amount
of cyclic GMP (cGMP) in the retina, the exact time-course of this decrease is
not well established. Bullfrog retinal sections were isolated under infrared light
and quick-frozen with liquid nitrogen-cooled, metal hammers after exposure to
various intensities of continuous illumination. This quick-freezing should stop
the degradation of cGMP within 50-100 ms. The frozen retinal sections were
then slowly warmed up in the presence of perchloric acid to denature enzymes
involved in cGMP metabolism, cGMP was determined by radioimmunoassay
and comparison was made between light- and dark-adapted retinal sections
from the same animal. The average cGMP concentration was 44.3:1:0.7 pmol
cGMP/mg protein or 170.9:1:3.2 pmol cGMP/retina. After 1 s of illumination
no significant change in cGMP concentration was found even with the brightest
light used (~ 7 • 10 7 rhodopsins bleached/second per rod. At this intensity the
first significant decrease in cGMP from dark-adapted levels was detected 3-5 s
after the initiation of illumination; cGMP decayed to 70-75% of the dark-
adapted value after ~ 30 s. With lower intensity illumination the eGMP levels
recovered to dark-adapted levels after the initial decrease even though the
bleaching light remained on.
INTRODUCTION
Several observations implicate cyclic guanosine 3', 5' monophosphate (cGMP)
as an important factor in the response of a vertebrate photoreceptor cell to
light. The highest concentration of cGMP in any mammalian tissue has been
found in the retina (Farber and Lolley, 1974; Gordis et al., 1974, 1977;
Ferrendelli and Cohen, 1976; Krishna et al., 1976). Most of this cGMP is
found in rod outer segment preparations (ROS) (Fletcher and Chader, 1976;
Woodruff et al., 1977). Moreover, microdissection of retinal layers has located
as much as 90% of retinal cGMP in the photoreceptor layers (Farber and
J. GEN. PHYSIOL. 9 The Rockefeller University Press 9 0022-1295/79/09/0415/12 $1.00 415
Volume 74 September 1979 415-426
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

416 THE JOURNAL OF GENERAL PHYSIOLOGY 9 VOLUME 74 9 1979
Lolley, 1974; Orr et al., 1976). In addition, mice that have retinas deficient in
photoreceptors have much lower amounts ofcGMP than normal mice (Father
and Lolley, 1976; Ferrendelli and Cohen, 1976). Light can lower these high
cGMP concentrations. A light-adapted retina has about one-half the amount
of cGMP as a dark-adapted one (Goridis et al., 1974, 1977; Ferrendelli and
Cohen, 1976; Cohen et al., 1978; Mitzel et al., 1978). This decrease is believed
to be due to a light-activated, cGMP-specific and GTP-dependent phospho-
diesterase (PDE) (Krishna et al., 1976; Wheeler and Bitensky, 1977; Bitensky
et al., 1978). Interestingly, inhibitors of PDE activity can dramatically affect
the amplitude of the rod receptor potential (Ebrey and Hood, 1973; Lipton et
al., 1977).
Although there is good agreement that light reduces the amount of cGMP
in the retina, the exact time-course of this decrease is not well established.
Goridis et al. (1974) found that a light exposure of 5 and 15 s caused 30 and
70% drops, respectively, of the cGMP levels of bovine retinas; the cGMP-
consuming reaction was stopped by dropping the retina in perchloric acid
(PCA). Goridis et al. (1977), using frog retinas, found that 10 s after a light
flash that bleached 60% of the rhodopsin, there was no significant change in
cGMP levels, but a large decrease was attained by 1 min. They stopped the
cGMP-consuming reaction by immersion of the retina in PCA. With an
attenuated light flash they found a significant drop of cGMP levels in only 3
s and a near maximal decrease of the cGMP level after 30 s. Mitzel et al.
(1978), using mouse retinas, found a 70% decrease in cGMP levels within 30-
60 s after turning on a continuous bleaching light (250 ft-c); they stopped the
reaction by freezing the retinas on dry ice.
Light is also known to cause a decrease in cGMP levels in isolated ROS
(Brodie and Bownds, 1976; Fletcher and Chader, 1976; Krishna et al., 1976;
Woodruffet al., 1977). Woodruffet al. (1977) found with continuous saturat-
ing illumination of frog ROS that cGMP levels decreased by 40-50% within
6 s; approximately half of this decrease occurred within 200-400 ms. At
appropriate times they stopped the cGMP-consuming reactions by rapidly
mixing the ROS with PCA.
Recently, Yee and Liebman (1978), using rod disc membrane preparations,
found that increased PDE activity in the presence of high GTP concentrations
could be detected within 100 ms by a bright flash, while very low intensities
of light or a low GTP concentration could cause delays up to seconds for full
activation.
The present experiments were undertaken to determine a more precise rate
of decrease of cGMP in the retina. By comparison to similar techniques on
muscle (see Brooker, 1975) the use of rapid-freezing techniques with the retina
should stop the PDE activity within 50-100 ms. If, as indicated above, most
of the cGMP is in the photoreceptors, then this change in retinal cGMP levels
should fairly accurately reflect changes in photoreceptor cGMP levels.
METHODS
Bullfrogs
(Rana catesbeiana)
were dark-adapted for at least 12 h before use. All the
following operations were performed under infrared illumination, using an infrared
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

KILBRIDE AND EBREY
Quick-Freezing in Measurement of cGMP Levels
tn
Frog Retina
417
image converter (FJW Industries, Mt. Prospect, II). After pithing, an eye was excised
and hemisected with a razor blade. The posterior half of the eye was cut in two and
the sections were placed into freshly oxygenated frog Ringer's solution. The Ringer's
solution contained 108 mM NaCI, 2.5 mM KCI, 0.6 mM Na2SO4, 0.13 mM NaHCO3,
1.6 mM CaCI2, 1.2 mM MgSO4, 3.0 mM HEPES (N-2 hydroxyethylpiperazine N'-2
ethanesulphonic acid), and .5.6 mM glucose adjusted to pH 7.5 with NaOH (modified
from Brown and Pinto, 1974). The retina was gently teased away from the pigment
epithelium and transferred to a petri dish containing about 25 ml of freshly oxygenated
Ringer's solution, keeping the retinal sections for each frog in separated petri dishes.
These sections were stored approximately 20-2.5 min in a light-tight chamber before
being used.
Pairs of retinal sections from the same frog were removed from the light-tight
chamber, and by using a small glass spoon, placed along with a volume ('-0.4 mi) of
Ringer's solution into pairs of large test tubes (30 ml Corex, Corning Glass Works,
Coming, N.Y.). Hammers made of aluminum were designed to just fit inside the
large tubes; the hammer's bottom was curved to match the tube's bottom curve. The
hammers were kept in liquid nitrogen (77~ until just prior to use. One retinal
section of the pair was placed in the test tube and had a "cold" hammer dropped on
it; this tube was then placed in an ethanol-dry ice bath (-75~ and served as a
dark-adapted control. The other tube with retinal section was placed in the automated
"cold" hammer dropping apparatuS. A "cold" hammer was suspended by an electro-
magnet above the tube, and air was blown between the hammer and the top of the
tube to prevent cold air drafts from precooling the retinal section. The temperature
in the bottom of the tube was within 0.5~ of the ambient temperature. A triggering
system both opened a shutter to expose the retinal section to a continuous light source
and at the appropriate time dropped the "cold" hammer on the retinal section. The
tube containing the frozen retinal section and the hammer was put into the ethanol-
dry ice bath (-75~
At the conclusion of a set of experiments, after all samples tubes were in the -75~
bath, the room lights were turned on. To the tube-hammer assembly was added 1.5
ml of a PCA (10%): methanol, 1:1 mixture prechiiled to -20~ The tube-hammer
assembly was allowed to warm up slowly (several hours) until the PCA-methanol had
melted. Then the hammer was removed and rinsed with 1 ml PCA-methanol (-20~
the frozen retinal section was homogenized with a syringe, and finally the syringe was
washed with 0.5 ml PCA-methanol. We assumed that after this treatment all the
enzymes in the preparation had been denatured and that samples could be worked
with at room temperature. Although this assumption seems reasonable, partial
controls were performed by adding radioactive eGMP to frozen test tubes with and
without retinas and then adding PCA-methanol as outlined above. At the the end of
the pH neutralization process described below the samples contained identical (4-5%)
amounts of cGMP measured by a modification of the method of Thompson and
Appleman (1971) indicating no metabolism of cGMP after the retinas were frozen.
Even this control did not rule out the somewhat unlikely possibility that when the
samples were slowly warmed up from -75~ in the presence of PCA, some cGMP
metabolism might occur which the added exogenous cGMP did not monitor.
The homogenized retina was centrifuged at 36,000 g (4~ for 1 h. The pellet was
saved for protein determination after adding 0.5 N NaOH and heating in a 40~
water bath for 12 h (Hess and Lewin, 1965) along with standard protein solutions.
Protein was measured by the method of Lowry et al. (1951). The supernate was
neutralized with 6 N KOH using a PH indicator dye to mark the end point. This
mixture was centrifuged at 25,000 g for 20 rain to remove the PCA precipitate. 1 ml
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

418
THE JOURNAL OF GENERAL PHYSIOLOGY 9 VOLUME 74 *
1979
of this supernate was added to 1 ml of Tris-EDTA buffer (0.2 M Tris and 16 mM
EDTA) at pH 7.5. cGMP in the Tris-EDTA-buffered supernate was determined by
a cGMP-specific radioimmunoassay (RIA) (Amersham/Seale Corp., Arlington
Heights, Ill.). Standard samples of cGMP (Sigma Chemical Co., St. Louis Mo.) were
added to PCA-methanol and carried through the procedure to obtain a standard
curve, cGMP concentrations are expressed as picomoles cGMP/milligram of total
retinal protein.
The light source used was from a tungsten lamp (Sylvania 500W DAK projector
lamp, GTE Sylvania, Inc., Stamford, Conn.) which was passed through various
neutral density (ND) filters and an electromagnetic shutter. The unattenuated light
(ND - 0) was found to bleach a rhodopsin solution (2% LO, phosphate buffer) in the
automated hammer-dropping apparatus at the rate of 2.4%/s. We estimate this is
equivalent to bleaching -7 • 107 rhodopsins/second per ROS.
In order to estimate how fast a retinal section was cooled by the hammer and tube
assembly, a 36-gauge copper-constantan thermocouple was inserted through a small
hole in the bottom of one of the 30-ml test tubes used in the experiments. A retinal
section with Ringer's solution was placed over the thermocouple and a "cold" hammer
dropped into the tube. The thermocouple registered -20~ within 100 ms after the
hammer hit the retina.
The condition and responsiveness of the retinal sections were monitored by record-
ing their electroretinograms (ERGs) using an apparatus similar to one already
described (Ebrey and Hood, 1973; Hood and Hock, 1973). In brief, the isolated
retinal sections were placed in the above Ringer's solution in which 15 mM NaCI was
replaced with 15 mM Na aspartate to isolate the receptor response. The retinal
sections were placed in about 25 ml freshly oxygenated Ringer's solution in the same
type petri dish as that used to store the retinal sections in the quick-freeze experiments.
A pair of cotton-wick electrodes was placed above and below the retinal section while
it was emersed in the Ringer's solution. The retinal sections were exposed to broad
field illumination from a strobe flash after passing the light through a 500 nm
interference filter and suitable neutral density filters. The ERGs appeared normal.
After 25 min, the maximum time the isolated retinal sections were stored for the
quick-freezing experiments, the amplitude of the ERGs of the retinal sections had
decreased slightly, about 0.3 neutral density units from the maximum dark-adapted
initial response. The latency of a response using light intensity comparable to a ND
-- 2 light intensity in the cGMP experiment was -50 ms with the time to peak of
-300 ms. Brighter lights gave shorter latencies. Although the ERG cannot be used to
derive the exact kinetics of the rod outer segment's conductance changes, it should
give a rough indication of how fast these changes must be.
RESULTS
Control experiments in which a retinal section was placed in a small amount
of Ringer's solution in the bottom of a tube as was done in a quick-freezing
experiment showed no significant decay in cGMP in the dark for 20 min
(Table I). In our experiments the retinal sections were in a tube for less than
2 min. A comparison of dark-adapted values of picomole cGMP/milligram
protein in retinal sections at the beginning and end of 12 sets of experiments
showed a decrease of only 5% in those retinal sections that were stored longer.
To account for animal-to-animal variation, the dark-adapted controls along
with the retinal sections for each data point were done from the same frog;
the control was frozen immediately prior to a light-exposed experiment. The
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

KILBRIDE AND EBREY
Quzck-Fretzmg m Measurement of cGMP Ltr~ls in Frog Retina
419
values of cGMP (picomole/milligram protein) were normalized to one so a
comparison could be made among frogs. The actual dark value for the frog
retinal cGMP concentrations in these experiments was found to be 44.3 3:0.7
pmol cGMP/mg protein (+ SEM; 210 retinal samples) which corresponds to
170.9 + 3.2 pmol cGMP/retina.
The measurement of cGMP in retinal sections exposed to light and then
quick-frozen to stop enzyme activity is shown in Fig. 1. Five different contin-
uous light intensities were obtained by placing neutral density filters (ND ---
0,2,3,4, and 5) in the light path. After a l-s light exposure there is no significant
change in cGMP levels at any light intensities used. Even at the highest
intensity (ND = 0), the first significant change in cGMP is not seen until 3-
5 s. The time needed for one-half of the maximum decrease of cGMP is
estimated to be 4 3:1 s for this light intensity (ND ffi 0) and 14 3:3 s for the
next most intense light (ND -- 2). These two intensities of light (ND - 0 and
TABLE I
CONCENTRATION cGMP IN RETINAL SECTIONS TO CHECK
FOR SPONTANEOUS DECAY IN DARK
Time in dark % of initial
rain pmol cGMP/mg protein*
0 100+6 (3)
10 102 + 7 (3)
20 96+3 (3)
30
893= 12
(2)
* Mean =l= SEM (number of retinal sections).
2) caused the cGMP to decay to about 70-75% of the dark-adapted level by
60s.
Another striking feature of Fig. 1 is that at the lower light intensities (ND
-- 3 or 4) which cause an initial decrease in cGMP, the cGMP level returned
to its dark-adapted value even though the bleaching light remained on. With
the ND -- 4 light intensity, the cGMP level recovered by 30 s while at the
higher intensity (ND = 3) it required about 60 s. At the time of minimal
cGMP levels, about 20 s, the higher light intensity (ND ffi 3) had about 75%
of the dark-adapted level of cGMP while the lower intensity light dropped to
only about 86%. With the lowest intensity light (ND -- 5) shown in Fig. 1,
there is no significant change in cGMP levels from the dark levels with the
exception of the point at 5 s. The value of the 5-s point is 0.86 3:0.05 (10
samples) of the normalized dark-adapted level, and probably indicates a
significant decrease. If so, then the decrease in cGMP recovers by 10 s (see
Fig. 1). These observations imply that the recovery of cGMP levels toward
dark-adapted levels can occur in the light but is slower the greater the light
intensity.
Using very dim light (ND ffi 7) the value for the cGMP level after 5 s of
illumination was 0.95 + 0.11 (five samples) of the normalized dark-adapted
level, indicating no significant decrease. This result and the small change for
the ND =ffi 5 filter indicate that the threshold for a measureable decrease is
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

420
THE JOURNAL OF GENERAL PHYSIOLOGY 9
VOLUME
74
9 1979
somewhere between these two intensities which we estimate to bleach 7 x 10 2
and 7 rhodopins/second per ROS.
DISCUSSION
Orr et al. (1976) found over 90% of the total cGMP in the rabbit retina
concentrated in the photoreceptor layer. We estimate from their data that
01.35
I0
15 20
.30
40 50 60
i i i ! , l i , i i
%
LO h
~o.B t"'-,I,,...~,..... ~t~_ '
"
~0.6
,•I0
08
"~0.6
"~ 1.0
~08
~0.8
0.6
~0.8
~0.6
o,
,;
Seconds
E)ND.=5
, I
40 ~0 60
FIGURE 1. Concentrations of cGMP (pmol cGMP/mg protein) as a function
of light exposure (seconds) in the retina normalized so that dark-adapted values
are one. The light intensity was attenuated by five different neutral density
filters (A)ND =0, (B)ND =2, IC)ND = 3, (D)ND = 4, and (E)ND = 5. ND
=0 corresponds to ~ 7 • 10 rhodopsins bleached# per ROS. The points
represent the mean • SEM with an average of seven light exposed retinal
sections assayed per point.
about 50% of the cGMP in the retina is in the ROS. We find about 170 pmol
eGMP/frog retina, so assuming that about 50% of this is in the ROS and that
there are about 106 ROS per retina (Gordon and Hood, 1976), then we find
about 5 • 107 eGMPs/ROS. Woodruff et al. (1977) estimate that in isolated
bullfrog ROS there are 1-2 eGMPs per 100 rhodopsins and 3 X 109 rhodop-
sins/ROS which gives about 3-6 X l0 T cGMPs/ROS. Moreover, Goridis et
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

KILaXZOE ^NO EaxEY
Quick-Frtezzng ~n Measurement of cGMP Levels in Frog Retina
421
al. (1977) found 25.0 pmol cGMP/mg protein in
Rana esculenta.
We find a
similar value, 44.3 + 0.7 pmol cGMP/mg protein in our frogs,
Rana catesbeiana.
Recently, de Azeredo et al. (1978) report 43.50:1:2.26 pmol cGMP]mg dry
weight in the ROS layer upon microdissection of frog
(Rana pipiens)
retina.
We estimate this value leads to about 11 #M cGMP in the ROS or about 1
• 107 cGMPs/ROS. Thus, our levels of cGMP in the ROS are in fair
agreement with these others despite different manipulations of the ROS.
How much of the cGMP decrease we see in the retina is due to decreases in
the ROS's cGMP is unclear. Orr et al. (1976) found that cGMP in the ROS
layer of rabbit retinas decreased about 30% after 1 h in the light; we estimate
from their data that about 40% of the total decrease of cGMP in the retina is
from the decrease in the ROS. de Azeredo et al. (1978), using frog
(Rana
pipiens)
retinas, found a larger decrease of cGMP levels (70%) in the ROS
layer. Although these results give some indication that a substantial percent
of the total decrease of eGMP of the retina is due to decrease in the ROS's
cGMP, they do not reflect any differences in the rate of decrease in cGMP in
the various retinal layers. For example the ROS's cGMP could be a dispro-
portionately large fraction of the retinal cGMP decrease that occurs during
the initial second after illumination. Moreover, there may be two distinct
pools of cGMP in the ROS, a small one which changes rapidly and a larger
one which changes slowly. Our results must be viewed cautiously in light of
these considerations. We find the retinal eGMP levels decrease about 25%
after 15 s of illumination with a decay half-time of about 4 s with the brightest
light (ND -- 0). If all the initial decrease we see in retinal cGMP is entirely
in the ROS, then the decay rate is about 6% of the ROS's cGMP per second,
or about 3 • 106 cGMPs/ROS per second. If only 50% of the initial decrease
we see in retinal cGMP is from the ROS, then this decay would be reduced
by half.
We detected a significant decrease from the dark-adapted level of cGMP
only after several seconds. This results is surprising because it appears there
should be enough light-stimulated PDE activity in the ROS to hydrolyze all
the cGMP in the ROS much more rapidly than this. Miki et al. (1975) found
a Vm~ of 48,000 mol cGMP hydrolyzed/min per tool enzyme with a ratio of
930 rhodopsins per PDE and with a Km of 70 #M. From these values we
estimate
a Vmax
of 2 • 10 a cGMPs/s per ROS and a Km of 3 • 107 cGMPs/
ROS. These numbers are probably conservative estimates of the maximum
PDE activity, a Using them in the integrated Michaelis-Menten equation and
assuming all the PDE is activated immediately, it can be shown that it should
take <25 ms for the cGMP to be reduced by 50% from an initial concentration
of 5 • l07 cGMPs/ROS. We are unable to explain this apparent discrepancy
between the calculated rate of cGMP hydrolysis and what we actually
measure. Perhaps our result is due to there being only a small pool of cGMP
accessible to the high activity PDE. Another possiblity is that the activation
of the PDE may not solely be due to the availability of bleached rhodopsin,
but in vivo is controlled by some rate-limiting factor which is not rate-limiting
in the biochemical experiments. For example, PDE activation in rod disk
I Hurley, J., and T. G. Ebrey. Manuscript submitted for publication,
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

422 THE JOURNAL OF OF-,NE, RAL PHYSIOLOGY* VOLUME, 74. 1979
preparations has been shown to be delayed for several seconds with low GTP
concentrations (Yee and Liebman, 1978).
Our resuh of no significant decay in cGMP after 1 s of illumination is in
apparent disagreement with the experiments of Woodruff et al. (1977). In
isolated frog ROS they found decreases of 20-25% in cGMP levels within
200-400 ms. They stopped the cGMP-consuming reactions by rapidly mixing
the ROS with PCA; this procedure may not rapidly stop an enzyme reaction
occurring inside the rod's plasma membrane. However, a more recent report
by that group (Bownds et al., 1978) confirms a latency for cGMP decrease of
50 ms and a half-time of 125 ms. Another difference between our experiments
and Woodruffet al. (1977) is that they used low-calcium (i.e., EGTA) Ringer's
solution; Cohen et al. (1978) has shown this can affect the cGMP levels in the
retina. Yet another reason for the difference in results could be the difference
between the isolated frog ROS preparation of Woodruff et al., which may
have alterations in its normal pools of metabolites and our retinal preparation
which may more accurately reflect in vivo conditions.
Fig. 1 shows cGMP levels recover to dark-adapted levels within 30-60 s
under continuous low intensity illumination (ND == 3,4, and 5). Previous
work has not found recovery of cGMP levels in the dark or in the light in the
retina (Goridis et al., 1974; Goridis et al., 1977) perhaps because of higher
light levels than those needed to observe recovery was used. The origin of the
recovery we observed is unknown, but it might be due to the switching of the
PDE from a high activity form to low activity form even though the light
remains on. In such a case, the cGMP level in the ROS wouldn't go to zero
if the hypothetical decay of PDE activity is rapid enough. Recently, Yee and
Liebman (1978) and Liebman (1979) in their ROS disk experiments have
presented some evidence for a lowering of PDE activity after initial activation.
In the rod photoreceptor, Yoshikami and Hagins (1973) have proposed an
internal excitation transmitter which carries the light signal between the
bleached rhodopsin and the permeability sites on the plasma membrane.
Yoshikami and Hagins (1973) suggested calcium as a good candidate for the
internal transmitter. Cyclic nucleotides have also been suggested as an internal
transmitter in photoreceptors (see review by Hubbell and Bownds, 1979). One
of the criteria a candidate for an internal transmitter should fulfill is that its
concentration should undergo significant changes within several hundred
milliseconds after light absorption, the approximate time range of the receptor
potential (Baylor et al., 1974). As noted above, we found only significant
cGMP level changes only after more than I s of illumination. Moreover, with
our brightest light, which would saturate the electrical response, we estimate
the maximum rate of decrease to be 1-3 • 106 cGMPs hydrolyzed/s per
ROS, or about 0.6% of the cGMP in the ROS hydrolyzed during 100 ms. So
the cGMP decrease we observed is probably too slow by more than an order
of magnitude to be involved as an internal excitation transmitter.
Another property an internal transmitter must have is a high signal-to-noise
ratio at low light intensities, since a photoreceptor can respond to single
photon events (Hecht et al., 1942). When an attempt is made in the ROS to
create a low concentration of cGMP due to local high PDE activity triggered
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

KILBRIDE AND EBREY
Quick-Freezing in Measurement of cGMP Levels in Frog Retina
423
by a single photon, cGMP from the surrounding region will diffuse into this
area of low concentration (see negative transmitter in Fig. 2 A). This diffusion
will tend to wash out any large decrease in local transmitter concentration.
Thus, it is somewhat difficult to see how one photon, even if it activated
enzymes which lead to the hydrolysis of all the cGMP between two disks,
could affect the sodium permeability sites associated with the "--20-60 disks
that are shut off by one photon (see Cone, 1973; Yoshikami and Hagins,
photon
.(,I
! 9 e.,
~"
~
I
A) Negative transmitter
photon
oo,,vo,,o i
I
B) Positive transmitter
9 e.e I
FmuRE 2. Schematic representation of (A) a negative transmitter and (B) a
positive transmitter. The figures represent stylized ROS with the dots repre-
senting the transmitters.
1973). In the case of the release of a positive transmitter released against a low
background (see Fig. 2 B) it is easier to obtain a high signal-to-noise ratio.
However, the possibility of a negative transmitter cannot yet be dismissed.
In summary, there appears to be some serious problems with cGMP as a
candidate for an internal excitation transmitter in vertebrate rods. However,
cGMP as well as other transmitter candidates such as 5' GMP should be
considered as possible adaptive signal transmitters. The time scale of cGMP
recovery that we found in dim light is reminiscent of light adaptation in the
cold-blooded vertebrate rod (gecko: Kleinschmidt and Dowling, 1975; toad:
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

424
THE JOURNAL OF GENERAL PHYSIOLOGY 9 VOLUME
74 9 1979
Fain, 1976). Moreover, a depletion signal like cGMP which is averaged over
a large portion of the ROS could serve adaptative processes well.
We thank James Hurley for many helpful discussions and Dr. Peter H. Hartline for some of the
original suggestions for the experiments.
This work was supported in part by grant EY-07005 from the National Institutes of Health.
Paul Kilbride was a National Eye Institute National Institutional Research Service Award
Predoctoral Fellow.
Received for publicatton 8January 1979.
REFERENCES
BAYLOR, D. A.,
A. L.
HODGKIN,
and T. LAMa. 1974. Reconstruction of the electrical responses
of turtle cones to flashes and steps of light.J.
Physiol. (Lond.).
242:759-791.
BITENSKY, M. W., G. L. WHEELER, B. ALOm, S. VETURY, and Y. MATUO. 1978. Light- and
GTP-activated photoreceptor phosphodiesterase: Regulation by a light-activated GTPase
and identification of rhodopsin as the phosphodiesterase binding site.
Adv. Cyclic Nucleotide
Res.
9:553-572.
BOWNDS, M. D., M. L. WOODRUFF, M. S. BIERNaAUM, A. S. POLANS, J. HERMOLIN, P. R.
ROaINSON, and B. P. AaRAMSON. 1978. A role for cyclic GMP in vertebrate photoreceptors.
Society for Neuroscience 8th Annual Meeting, St. Louis, Mo. 621.
BROmE, A. E., and D. BowsDs. 1976. Biochemical correlates of adaptation processes in isolated
frog photoreceptor membranes.
J. Gen. Physzol. 68:1-11.
BROOKER, G. 1975. Implications of cyclic nucleotide oscillations during the myocardial contrac-
tion cycle.
Adv, Cyclic Nucleotide Res.
5"435-452.
BROWN, J. E,, and L. H. PINTO. 1974. Ionic mechanism for the photoreceptor potential of Bufo
marinus.J.
Physiol. (Lond.).
236"575-591.
COHEN, A. I., I. A. HALL, andJ. A. FERRENDELLI. 1978. Calcium and cyclic nucleotide regulation
in incubated mouse retinas.J.
Gen. Physiol.
7h595-612.
CONE, R. A. 1973. The internal transmitter model for visual excitations: Some quantitative
implications.
In
Biochemistry and Physiology of Visual Pigments. H./.,anger, editor. Springer-
Verlag New York, Inc. 257-282.
DE AZEREDO, F. A. M., W. D. LUST, and J. V. PASSONNEAU. 1978. Guanine nucleotide
concentrations
in vivo
in outer segments of dark and light adapted frog retina.
Biochem.
Bwphys. Res. Commun.
85:293-300.
EBREY, T. G., and D. C. HooD. 1973. The effects of cyclic nucleotide phosphodiesterase
inhibitors on the frog rod receptor potential.
In
Biochemistry and Physiology of Visual
Pigments. H. Langer, editor. Springer-Verlag New York, Inc. 341-350.
FAIN, G. L. 1976. Sensitivity of toad rods: dependence on wave-length and background
illumination. J.
Physiol. (Lond.).
261:71-101.
FARBER, D. B., and R. N. LOLLEY. 1974. Cyclic guanosine monophosphate: elevation in
degenerating photoreceptor cells of the C3H mouse retina.
Science (Wash. D.C.).
186:449-
451.
FARBER, D. B., and R. N. LOLLEY. 1976. Enzyme basis for cyclic GMP accumulation in
degenerative photoreceptor cells of mouse retina.
J. Cyclic Nucleotide Res.
2:139-148.
FERRENDELLI, J. A., and A. I. COHEN. 1976. The effects of light and dark adaptation on the
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

KILBRIDE AND EBREY Quick.Freezing in Measurontnt of cGMP Levels in Frog Retina 425
levels of cyclic neucleotides in retinas of mice heterozygous for a gene for photoreceptor
dystrophy. Biochem. Biophys. Res. Commun. 73:421-427.
FLETCHER, R. T., and G. J. CH^DER. 1976. Cyclic GMP: control of concentration by light in
retinal photoreceptors. Biochem. Biophys. Res. Cornmun. 70:. 1297-1302.
GORDON, J., and D. C. HooD. 1976. Anatomy and physiology of the frog retina. In The
Amphibian Visual System: A Muhidisciplinary Approach. K. V. Fite, editor. Academic
Press, Inc., New York. 29-86.
GORIDIS, C., P. F. URBAN, and P. MANDEL. 1977. The effects of flash illumination on the
endogenous cyclic GMP content of isolated frog retinas. Exp. Eye Res. 24:171-177,
Goglms, C., N. VmMAUX, H. L. CAILLA, and M. A. DEr-~OE. 1974. Rapid, light-induced
changes of retinal cyclic GMP levels. FEBS (Fed. Fur. Biochem. Sot.) Lett. 49:,167-169.
HECHT, S., S. SHLAER, and M. H. PIRENEE. 1942. Energy, quanta, and vision.J. Gen. Physiol. 25:
819-840.
HEss, H. H., and E. LEWIN. 1965. Microassay of biochemical structural components in nervous
tissues-II Methods for cerebrosides, proteolipid proteins and residue proteins. J. Neurochem.
12:205-211.
HooD, D. C., and P. A. HocK. 1973. Recovery of cone receptor activity in the frog's isolated
retina. Vision Res. 13:1943-1951.
HUnBELL, W. L., and M. D. BOWNDS. 1979. Visual transduction in vertebrate photoreceptors.
Annu. Rev. Neuroscience. 2:17-34.
KLEINSCHMIDT, J., and J. E. DowusG. 1975. Intracellular recordings from gecko photoreceptors
during light and dark adaptation. J. C, en. Physiol. 66:.617-648.
KRISHNA,
G., N.
KRISHNAN,
R. T.
FLETCHER,
and G.
CHADER,
1976. Effects of light on cyclic
GMP metabolism in retinal photoreceptors. J. Neurochem. 27:717-722.
LIEBM^N, P. A. 1979. One rhodopsin activates many photoreceptor esterase molecules by lateral
diffusion. Biophys. J. 25:1a. (Abstr.).
LIPTON, S. A., H. RASMUSSEN, and J. E. DOWUNG. 1977. Electrical and adaptive properties in
Bufo marinus: II. Effects of cyclic nucleotides and prostaglandins. J. Gen. Physiol. 70:.771-791.
LowRY, O. H., N. J. ROSEBROUOH, A. L. FARR, and R. J. RANDALL. 1951. Protein measurement
with the Folin phenol reagent. J. Biol. Chem. 193:265-275.
MIKX, N., J. M. BARAB^N, J. J. I~mNS, J. j. BoYcE, and M. W. BITENSgV. 1975. Purification
and properties of the light activated cyclic nucleotide phosphodiesterase of rod outer
segments. J. Biol. Chem. 25:6320-6327.
MITZEL, D, L., I. A. HALL, G. W. DEVRxES, A. I. COHEN, and J. A. FERRENDELLL 1978.
Comparison of cyclic nucleotide and energy metabolism of intact mouse retina in situ and in
vitro. Exp. Eye Res. 27:27-37.
ORR, H. T., O. H. LowRY, A. I. COHEN, and J. A. FEgRENDELLL 1976. Distribution of 3':5'-
cyclic AMP and 3':5'-cyclic GMP in rabbit retina in vivo: selective effects of dark and light
adaptation and ischemia. Proc. Natl. Acad. Sci. U. S. A. 73:4442-4445.
SzuTs, E. A., and R. A. CONE. 1977. Calcium content of frog rod outer segments and discs.
B,ochim. Biophys. Acta. 468:194-208.
THOMPSON, W. J., and M. M. APPLEM^N. 1971. Multiple cyclic nucleotide phosphodiesterase
activities from rat brain. Biochemist~. 10:.311-316.
WHEELER, G. L., and M. W. BITENSK~'. 1977. A light-activated GTPase in vertebrate photore-
ceptors: regulation of light-activated cyclic GMP phosphodiesterase. Proc. Natl. Acad. Sci.
U.S.A. 74:4238-4242.
WOODRUFF, M. L., D. BOWNDS, S. H. GREEN, J. L. MORRISSE'f, and A. SHEDLOVSKY, 1977.
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979

426
THE JOURNAL OF GENERAL PHYSIOLOGY 9 VOLUME 74 9
1979
Guanosine 3', 5'-cyclic monophosphate and the
in vitro
physiology of frog photoreceptor
membranes,
f. C, en. Physiol.
69:.667-679.
YEE, R., and P. A. LIEBMAN. 1978. Light activated phosphodiesterase of the rod outer segment;
kinetics and parameters of activation and deactivation,
f. Biol. Chem.
253:8902-8909.
YOSmKAMI, S., and W. A. HAOINS. 1973. Control of the dark current in vertebrate rods and
cones.
In
Biochemistry and Physiology of Visual Pigments. H. Langer, editor. Springer-
Verlag New York, Inc. 245-255
on July 9, 2011jgp.rupress.orgDownloaded from
Published September 1, 1979