Acceleration of Key Reactions as a Strategy to Elucidate the Rate-Limiting Chemistry Underlying Phototransduction Inactivation

Abstract

purpose. A reconstituted system was used to establish a strategy to determine the rate-limiting chemistry responsible for recovery of the dim-flash response in rod photoreceptors.

methods. A general approach for identifying the rate-limiting step in a series of reactions is to evaluate the consequences of accelerating each step separately, while monitoring the rate of formation of the end product of the series. This strategy was applied to the reactions involved in quenching phototransduction in bovine rod outer segment (bROS) homogenates. The decay of photoactivated rhodopsin (R*) and inactivation of transducin by guanosine triphosphate (GTP) hydrolysis are the leading candidates for limiting the rate of phototransduction turn-off. These reactions were accelerated separately and together by adding hydroxylamine and/or the regulator of G-protein signaling-9 catalytic domain (RGS9d) while monitoring phosphodiesterase (PDE) activity triggered by a pulse of light in bROS homogenates.

results. PDE activity in bROS homogenates triggered by a flash of light returned to its dark value with a rate constant of 0.087 ± 0.002 seconds in this system. The rate of PDE recovery increased to 0.11 ± 0.004 seconds when R* decay was accelerated with 10 to 50 mM hydroxylamine, suggesting that R* inactivation limits the rate of phototransduction turn-off under these conditions. Adding both hydroxylamine and RGS9d, a factor that accelerates transducin inactivation, increased the rate of PDE decay even further. RGS9d had no effect on PDE recovery kinetics unless quenching of R* was also accelerated.

conclusions. Under in vitro conditions in bROS homogenates, the quenching of R* normally limits the rate of phototransduction shut-off. If R* decay is accelerated, inactivation of transducin by GTP hydrolysis becomes rate limiting. This study offers a general approach that could be used to investigate the rate-limiting chemistry of phototransduction turn-off in vivo.

The time course of a G-protein signaling cascade reflects the balance between a series of activation and recovery reactions. The cGMP hydrolysis cascade of vertebrate phototransduction is the G-protein-mediated pathway whose time course is currently the most precisely defined. Light triggers a signaling cascade in vertebrate rod photoreceptors in which photoactivated rhodopsin stimulates transducin to bind guanosine triphosphate (GTP) and promote hydrolysis of the intracellular second-messenger cGMP. Phototransduction hyperpolarizes the rod cell and inhibits release of neurotransmitter at the rod synaptic terminal. Temporal resolution of information carried by light requires rapid activation and rapid deactivation of phototransduction. 

After exposure to intense illumination, rod photoreceptors recover slowly toward their dark state through a process that requires both quenching of phototransduction and regeneration of rhodopsin. ^{1 2 3} After exposure to weaker stimuli, rods recover to their dark state much more quickly. Recovery of single photoreceptors after a dim flash occurs with highly stereotyped kinetics with a time constant of approximately 0.2 seconds for mammalian rods at 37°C and approximately 2 seconds for amphibian rods at 24°C to 25°C. ^{4 5 6} Recovery under these conditions requires both cGMP synthesis by guanylyl cyclase and termination of each of the phototransduction reactions. Biochemical studies and analyses of genetically manipulated mice have shown that several reactions are required for normal recovery kinetics. These reactions include Ca²⁺-dependent modulation of guanylyl cyclase activity by guanylyl cyclase-activating protein (GCAP), rhodopsin phosphorylation, arrestin binding to phosphorylated rhodopsin and GTP hydrolysis by transducin. ^{7 8 9 10}  

Although each of these reactions is essential for normal recovery kinetics, the chemical step that limits the rate of photoreceptor recovery has not yet been clearly defined. ¹¹ For example, both the hydrolysis of GTP by transducin and the deactivation of rhodopsin, have been implicated as the rate-determining step for photoreceptor recovery. ^{12 13} Alternatively, the identity of the rate-limiting step may depend on experimental conditions. For example, the lifetime of active transducin may be limiting with very dim flashes, whereas the quenching of photoactivated rhodopsin (R*) may limit recovery after more intense flashes. ¹⁴ Additional mechanisms for inactivation of phototransduction that are not dependent on hydrolysis of GTP have been postulated, including phosphorylation of the phosphodiesterase (PDE) inhibitory subunit and a target enzyme inhibitory factor. ^{15 16}  

None of these models has been tested by direct biochemical analysis, because the fast kinetics of the photoresponse cannot be duplicated in vitro. Decay of phototransduction requires seconds to tens of seconds in vitro, several times slower than in vivo. ^{17 18 19} This is probably due to the dilution of rhodopsin-containing membranes and loss of soluble factors during rod outer segment preparation. For example, the inactivation of rhodopsin is much faster in rod outer segment preparations that retain more rhodopsin kinase. ²⁰  

Ultimately, the identity of the rate-limiting step for inactivation of phototransduction must be determined in vivo. However, to help devise an effective strategy for these types of studies we performed a study in an in vitro model to demonstrate an experimental approach that can identify which reaction limits the rate of phototransduction inactivation. The strategy was as follows: It is clear that photoresponses depend ultimately on the reactions governing the production and destruction of the second messenger, cGMP. Most studies suggest that quenching cGMP PDE activity limits the rate of photoreceptor recovery. ^{6 21} To elucidate the rate-limiting step in PDE inactivation, we can monitor cGMP hydrolysis continuously in vitro and accelerate each reaction that contributes to switching off phototransduction. Only acceleration of the rate-limiting chemistry will result in faster PDE inactivation kinetics. 

In this study, we tested the applicability of this strategy by accelerating inactivation of R* and transducin, the two leading candidates for limiting the rate of phototransduction decay. To monitor the rate at which phototransduction is quenched after a flash stimulus, we measured the decay of light-activated PDE activity in homogenates of bovine rod outer segments (bROS). Quenching of photoactivated rhodopsin was accelerated by treating bROS homogenates with hydroxylamine, a compound that rapidly reacts with the chromophore of photoactivated rhodopsin to form retinal oxime and opsin. Hydroxylamine has been used in numerous studies with ROS homogenates and with intact photoreceptors to quench rhodopsin photoproducts artificially. ^{22 23 24 25} Inactivation of transducin was accelerated by supplementing ROS homogenates with the catalytic domain of the photoreceptor GTPase accelerating protein (GAP), RGS9. ²⁶ RGS9 accelerates transducin inactivation by stabilizing the transition state of the GTP hydrolysis reaction. ^{27 28 29}  

Using this strategy in vitro, we found that we could effectively determine the rate-limiting chemistry underlying PDE inactivation under conditions in which either R* decay or transducin GTPase activity was limiting. This strategy could be used to elucidate the rate-limiting step of photoreceptor inactivation in vivo by using transgenic technology to induce overexpression of factor(s) that accelerate these same reactions. 

Materials and Methods

ROS Preparation

bROS were prepared by using a flotation technique on a discontinuous sucrose gradient. ³⁰ The ROS membranes were diluted to a concentration of 12 mg/mL rhodopsin, aliquoted and frozen at −70°C in 20 mM 3-(N-morpholino)propanesulfonic acid (MOPS; pH 7.2), 2 mM MgCl₂, 60 mM KCl, 30 mM NaCl, 100 μM phenylmethylsulfonyl fluoride (PMSF), and 1 mM dithiothreitol (DTT). 

Expression and Purification of RGS9d and Recoverin

Glutathione S-transferase (GST)-tagged RGS9 catalytic domain (amino acids 276–431) was overexpressed in Escherichia coli and purified as described. ³¹ Recombinant recoverin was overexpressed in E. coli and purified over phenyl sepharose as described. ³² Fully myristoylated recoverin was used in these experiments. 

PDE Assay

cGMP hydrolysis was assayed by continuously monitoring the pH of a 10-μL drop of ROS suspension using a micro pH electrode (Microelectrodes, Inc. Londonderry, NH). The ROS suspension contained 30 μM rhodopsin, 500 μM adenosine triphosphate (ATP), 500 μM GTP, 5 mM cGMP, 1 mM DTT, 4 mM MgCl₂, 60 mM KCl, 30 mM NaCl, 10 mM HEPES (pH 7.2), except where noted otherwise. Hydroxylamine was prepared fresh by dissolving in water and adjusting the pH to 7.2 with NaOH immediately before the experiment. Hydroxylamine increased the buffer capacity of the reaction mixture. To correct for this, we determined the relative buffer capacities of the assay mixtures by measuring the change in pH resulting from the addition of a small aliquot of 0.01 N HCl in the presence or absence of hydroxylamine. The ratio of the values was used to linearly scale the pH responses. All assays were performed at room temperature. PDE activity was monitored in response to a flash of light that bleached 0.005% to 0.022% of the rhodopsin. To ensure that the rate of PDE decay was not limited by depletion of substrate or product inhibition, the same flash was delivered after PDE activity had recovered to its basal rate. The output from the pH meter was digitized with an analog-to-digital converter (Universal Laboratory Interface; Vernier Software, Beaverton, OR) and collected using Logger Pro software (Vernier Software) at 10 Hz. The rate of inactivation of PDE was approximated with a single exponential fit, using Igor Pro software (WaveMetrics, Lake Oswego, OR). PDE traces in all figures represent averages of at least three test results. PDE decay rates under single GTP turnover conditions were measured by monitoring the pH of 100 μL of fully bleached ROS (30 μM rhodopsin) in response to the addition of GTP (200 nM final concentration). The reaction was continuously stirred in a 96-well plate. All assays were performed at least in triplicate. 

Single-Turnover GTPase Assay

The rate of GTP hydrolysis was measured using a protocol previously described. ³³ Briefly, GTPase assays were initiated by adding GTP (γ-³²P labeled; 500,000 dpm/time point) to a final concentration of 200 nM to fully bleached bovine ROS (30 μM rhodopsin) under the same conditions used for PDE assays. Reactions were quenched by the addition of trichloroacetic acid (15% wt/vol final concentration) at various time points from 3 to 35 seconds. Nucleotides were separated from inorganic phosphate by incubation with activated charcoal. The fraction of GTP hydrolyzed was calculated by subtracting background counts in prequenched samples, and dividing by the average total counts per minute in fully hydrolyzed samples. The fraction of hydrolyzed GTP was plotted as a function of time. A control experiment was performed to test whether hydroxylamine affects the hydrolysis rate of transducin. Transducin was preactivated by adding GTP to a final concentration of 200 nM 3 seconds before addition of either hydroxylamine (30 mM final concentration) or H₂O. Reactions were then quenched at 10 seconds and processed as described earlier. 

Results

To monitor inactivation of phototransduction after a brief stimulus, we measured PDE activity in bROS homogenates by continuously monitoring the pH of an ROS suspension containing 30 μM rhodopsin, as originally described by Yee and Liebman. ³⁴ Phototransduction was stimulated by a brief flash of light that bleached between 0.005% and 0.022% of the rhodopsin. Light-activated PDE activity is proportional to the first derivative of the pH-versus-time plot (Fig. 1A) . ³⁵ Inactivation of phototransduction was approximated by using a single exponential function to fit the declining phase of the PDE response (Fig. 1A , bottom). The rate constant (k) for PDE turn-off in our system was 0.087 ± 0.002 seconds at 20°C. ROS concentration had little effect on the inactivation kinetics from 15 to 60 μM rhodopsin. As expected, omission of ATP from the assay resulted in a larger amplitude response, slower time-to-peak and slower inactivation kinetics (k = 0.024 ± 0.001 seconds) presumably because rhodopsin is not phosphorylated under these conditions (Fig. 1B) . 

Effect of Addition of Hydroxylamine on PDE Inactivation

Hydroxylamine quenches photoactivated rhodopsin (R*) by reacting with its chromophore to produce all-trans retinal oxime and opsin. To test whether quenching of R* limits the rate of PDE inactivation under the conditions of our in vitro assays, we added 50 mM hydroxylamine to the assay. Hydroxylamine accelerated the kinetics of PDE decay from 0.0870 ± 0.002 seconds to 0.113 ± 0.004 seconds in response to a flash that bleached approximately 0.005% of the rhodopsin (Fig. 2A) . The rate of PDE turn-off became saturated in the presence of 10 to 20 mM hydroxylamine, suggesting that the rate of the next slowest step in phototransduction decay is determined (Fig. 2B) at these concentrations. As confirmation of this, we found that ATP had little effect on PDE inactivation kinetics in the presence of more than 10 mM hydroxylamine, proving that hydroxylamine quenches photoactivated rhodopsin faster than endogenous rhodopsin kinase (Fig. 2C) . We were surprised to find that addition of hydroxylamine to our assay did not decrease the amplitude of the PDE response as would be expected if R* decay were accelerated. We found that the amplitude of PDE responses were unchanged or even increased in the presence of hydroxylamine, suggesting that hydroxylamine may also have an effect on the activation of phototransduction. 

Effect of Hydroxylamine on Recoverin’s Prolongation of PDE Inactivation

When bound to Ca²⁺, recoverin inhibits phosphorylation of rhodopsin and prolongs PDE activity in vitro. ³⁶ To determine whether Ca²⁺-recoverin modulates PDE lifetime in ways that are masked by its effect on R* lifetime, we added recoverin to our assay under conditions where R* decay was accelerated by hydroxylamine. Recoverin (5 μM) slowed PDE inactivation roughly fourfold and delayed the time-to-peak roughly twofold in the absence of hydroxylamine but in the presence of hydroxylamine, recoverin had little effect (Table 1) . Therefore, recoverin does not modulate other phototransduction inactivation steps downstream from inactivation of R* under these in vitro conditions. 

It should be noted that the effect of recoverin was dependent on Ca²⁺ with a k _1/2 of approximately 2 μM Ca²⁺ (data not shown) consistent with other reports of recoverin’s effect on PDE inactivation and photoresponse recovery in truncated rods infused with recombinant recoverin. ^{36 37} A different study showed that recoverin affects response amplitude but not inactivation kinetics of PDE in vitro. ¹⁷ We can duplicate this effect by performing the assay under conditions in which the free Ca²⁺ concentration is near 1 μM, similar to the Ca²⁺ concentration used in that study. However, when higher concentrations of Ca²⁺ are used, both the response amplitude and inactivation kinetics are slowed (data not shown). 

Effect of Addition of RGS9 Catalytic Domain on PDE Inactivation

To determine whether the hydrolysis of GTP is a rate-limiting step for inactivation, we analyzed the effects of adding RGS9d, the catalytic domain (amino acids 276–431) of the photoreceptor GTPase-accelerating protein (GAP), RGS9. PDE inactivation was not substantially affected by addition of up to 10 μM RGS9d in the absence of hydroxylamine (Fig. 3A) . In the presence of hydroxylamine, PDE decay was accelerated significantly by addition of micromolar amounts of RGS9d (Fig. 3B) . The addition of 5 μM RGS9d stimulated the rate of PDE decay from 0.116 ± 0.006 seconds to 0.192 ± 0.004 seconds. Figure 3C shows the dependence of the PDE inactivation time constant (τ = 1/k) on the concentration of RGS9d in the presence and absence of hydroxylamine. The amplitude of the PDE response in the presence of hydroxylamine was smaller when RGS9d was included in the assay. This was expected, because when R* decay was accelerated with hydroxylamine, transducin GTPase activity controlled the rate of recovery. Accelerating GTP hydrolysis with RGS9d yields a response with a shorter time-to-peak, thereby reducing the response amplitude. 

PDE Decay under Single GTP Turnover Conditions

For a direct comparison of the rate of GTP hydrolysis with PDE inactivation, we monitored PDE activity under single-turnover conditions in which [GTP] ≪ [transducin]. This simplifies the kinetic analysis, because any activated transducin hydrolyzes only one GTP molecule and is not recycled for another round of PDE activation. ³³ PDE activity decayed at 0.21 ± 0.03 seconds after addition of 200 nM GTP to fully bleached ROS membranes (Fig. 4A) . The rate of PDE decay under these conditions is sensitive to RGS9d. The addition of 7 μM RGS9d accelerated the rate of PDE decay to 0.34 ± 0.02 seconds. We also measured the hydrolysis rate of GTP under identical conditions using γ-³²P labeled GTP. GTP hydrolysis closely matches the rate of PDE decay. Figure 4B shows the fraction of GTP hydrolyzed as a function of time in the presence and absence of RGS9. The smooth lines represent the decay rate of PDE activity under identical conditions. 

The rate of transducin GTPase activity under these single-turnover conditions was faster (rate constant k = 0.21 ± 0.03 seconds) than the rate measured after a pulse of light in the presence of hydroxylamine (k = 0.116 ± 0.006 seconds). Under single-turnover conditions, all the transducin was activated synchronously. Under our flash conditions, lingering R* may stimulate transducin during the recovery phase to cause more complex kinetics that give the appearance of a slower rate of inactivation. To test whether this discrepancy is due to a nonspecific effect of hydroxylamine on transducin’s GTPase activity, we measured GTP hydrolysis rates under single-turnover conditions, with and without hydroxylamine, using a preactivation method previously described. ³³ The presence of 30 mM hydroxylamine in the assay did not significantly affect the rate of GTP hydrolysis (data not shown). 

Discussion

After exposure to a stimulus, a signaling system must respond, integrate, and adapt to the stimulus for its duration and then return to the quiescent state when the stimulus is terminated. For rod phototransduction, activation and inactivation of the transduction machinery must occur very rapidly to achieve temporal resolution of light signals. Photoactivation of rhodopsin triggers a signaling cascade in which transducin and PDE are activated to hydrolyze cGMP. To return the photoreceptor to its resting state, the concentration of cGMP in the cell must be restored. This is accomplished by stimulation of cGMP synthesis by the decrease in Ca²⁺ that accompanies phototransduction and by quenching each of the activated phototransduction components. Although several reactions that participate in recovery are known, the identity of the reaction that dominates the rate of phototransduction inactivation has not been decisively established. 

An effective way to evaluate whether a particular reaction in a series is rate-limiting is to test the consequences of specifically accelerating it. We tested this strategy by accelerating key reactions involved in quenching phototransduction in an in vitro system. Our strategy can be extended to elucidate the rate-limiting chemistry underlying the photoresponse in vivo by overexpression of factors that accelerate these same reactions. 

Effect of R* Decay on Inactivation of Phototransduction In Vitro

Kinetic modeling studies have suggested that phosphorylation of R* is the chemical reaction that limits inactivation of phototransduction after a flash of light. ¹³ To investigate whether R* inactivation is rate-limiting in our standard in vitro assay conditions we accelerated it with hydroxylamine. Addition of hydroxylamine accelerates phototransduction turn-off in our assays, even when as little as 0.005% of rhodopsins were stimulated by light. This confirmed that inactivation of R* can limit phototransduction turn-off. Inactivation accelerates with increasing hydroxylamine concentrations up to 10 mM. Above 10 mM hydroxylamine, the time constant of PDE inactivation becomes constant at 9 seconds. This suggests that because R* is quenched more quickly, a different reaction begins to dominate the rate of phototransduction inactivation. To test whether hydroxylamine quenches R* faster than the endogenous kinase, we omitted ATP from the assay. We found that PDE inactivation did not depend on ATP when hydroxylamine was present. 

Activity of Recoverin in Prolonging PDE Activity

The Ca²⁺-binding protein recoverin inhibits rhodopsin kinase in vitro and prolongs the lifetime of PDE but this effect requires micromolar concentrations of Ca²⁺. ^{36 37 38} The range of free Ca²⁺ concentration in an intact photoreceptor, roughly 50 to 550 nM, seems inconsistent with the idea that recoverin plays a dynamic role in modulating rhodopsin kinase activity in vivo. ³⁹ With this in mind, we tested whether recoverin has any other effect that is masked by prolonged R* lifetime in vitro. Under our standard assay conditions, recoverin-Ca²⁺ prolongs PDE lifetime approximately three- to fourfold and delays the time to peak roughly twofold. Under conditions where R* decay is accelerated with hydroxylamine and no longer dependent on rhodopsin kinase, Ca²⁺-recoverin has little effect on PDE inactivation. Therefore, Ca²⁺-recoverin prolongs the lifetime of PDE only by slowing the decay of R* under these conditions. Our data support the traditional model for recoverin in which Ca²⁺-bound recoverin sensitizes the photoreceptor in the dark by inhibiting rhodopsin kinase. ³⁶ In the light, when levels of Ca²⁺ in the photoreceptor are low, this inhibition is relieved. 

Effect of GTP Hydrolysis by Transducin

By accelerating the decay of photoactivated rhodopsin, we determined the rate of the next slow step in phototransduction inactivation. To test whether this reaction was transducin GTP hydrolysis, we added the catalytic domain of RGS9 (RGS9d) to our assay. In the absence of hydroxylamine, RGS9d had no effect on the kinetics of phototransduction turn-off, confirming that a reaction upstream from transducin inactivation normally determines the rate of phototransduction decay. When R* decay was accelerated by hydroxylamine, RGS9d accelerated PDE inactivation, showing directly that the next slowest step in phototransduction inactivation is the hydrolysis of GTP. It is notable that RGS9d accelerated GTP hydrolysis, even when the endogenous complement of RGS9 was present. Therefore, transducin must not be saturated with RGS9 at the concentration of ROS membranes used in our study. This finding suggests that overexpression of RGS9 in a transgenic mouse model may help define the rate-limiting chemistry for photoreceptor inactivation in vivo. 

The effect of RGS9d on the rate of PDE inactivation is a consequence of its effect on GTP hydrolysis. We confirmed this by showing that the rate of PDE decay matched closely the rate of GTP hydrolysis under single-turnover conditions, consistent with a previous study. ⁴⁰ Furthermore, both rates were accelerated to the same extent when RGS9d was added to the assay. Therefore, the rate of GTP hydrolysis determined the rate at which PDE was quenched. These findings exclude the formal possibility that there is an intervening slow step between GTP hydrolysis and PDE turn-off, in agreement with a previous study. ⁴¹ Our results do not support previous suggestions that PDE is turned off by mechanisms independent of GTP hydrolysis—at least under the conditions of our in vitro assays. ^{15 16}  

Conclusions

After a brief stimulus, photoreceptors return to their dark state by accelerating cGMP synthesis and slowing the rate of cGMP hydrolysis. Which branch of cGMP metabolism limits the rate of photoreceptor recovery is still a complex question. Guanylyl cyclase activity stimulated by the decrease in Ca²⁺ that accompanies the photoresponse may be strong enough to cause a net increase in the level of cGMP even in the presence of activated PDE. Two studies have suggested that elevated PDE activity remains even after photocurrent has recovered to its dark level after a dim flash. ^{42 43} In addition, measurements of cytoplasmic free Ca²⁺ levels show that Ca²⁺ remains low throughout the photoresponse suggesting that guanylyl cyclase remains maximally activated even as inward current recovers to its dark level. ³⁹  

Alternatively, recovery to the dark state may be limited by quenching of activated PDE. Experimental evidence for this is given in a study in which responses to dim flash were measured under conditions in which the level of Ca²⁺ was fixed at its dark level throughout the photoresponse, blocking light-stimulated cGMP synthesis. ²¹ The rate of photocurrent recovery under these conditions was the same as the rate measured when the level of Ca²⁺ was allowed to decrease after photostimulation. This argues that the quenching of PDE activity, rather than guanylyl cyclase stimulation, is responsible for the kinetics of photocurrent recovery. 

Our study directly identified two slow steps that can control the inactivation rate of PDE: R* quenching and transducin GTP hydrolysis. With the smallest stimulus that we could use in our assay, R* quenching limited inactivation of PDE. However, when R* decay was accelerated, inactivation of transducin was the next reaction to become rate limiting. PDE inactivation rates measured under the in vitro conditions of our assays are much slower than rates measured in vivo for photoreceptor recovery in mouse rods. This discrepancy may be due to loss of soluble factors, dilution of phototransduction components, and the temperature at which the assays were performed. These experiments are meant to lay the groundwork for additional in vivo studies and should not be interpreted as a quantitative description of photoresponse recovery. A similar set of experiments accelerating R* decay and transducin GTPase activity by overexpressing rhodopsin kinase or RGS9 in mouse rods will be necessary to resolve unambiguously which of these reactions dictates the rate of phototransduction decay in the intact photoreceptor. 

 Supported by The Howard Hughes Medical Institute, National Eye Institute Grants EY06641 (JBH) and EY11900 (TGW), the University of Washington Molecular and Cellular Biology Training Grant 5T32 GM07270 (MJK), and the National Library of Medicine Training Grant LM07093 (MES).

 Submitted for publication July 9, 2002; revised September 12 and October 10, 2002; accepted October 12, 2002.

 Disclosure: M.J. Kennedy, None; M.E. Sowa, None; T.G. Wensel, None; J.B. Hurley, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

pH assay for PDE activity. (A) The pH of an ROS suspension was continuously monitored. A pulse of light (0.005% bleach) delivered at time 0 resulted in a decrease in pH (curve shown in the upward direction), due to cGMP hydrolysis. The first derivative (bottom) is proportional to the rate of PDE activity. PDE activity recovered to its dark level as phototransduction was quenched. The rate of PDE decay was approximated with a single exponential fit of the declining phase (bold trace) of the response. Under these conditions, PDE activity was quenched with a k of 0.087 ± 0.002 seconds. (B) ATP was essential for fast inactivation of phototransduction. The decay of PDE activity was monitored in the presence (solid trace) and absence (broken trace) of ATP. The rate constant for PDE recovery in the absence of ATP was 0.024 ± 0.001 seconds.

Figure 2.

 

Hydroxylamine accelerated the inactivation of PDE activity. (A) The rate of PDE inactivation in the presence and absence of 50 mM hydroxylamine (NH₂OH) in response to a pulse of light at time 0. Hydroxylamine (50 mM) accelerated the rate constant for PDE recovery from k = 0.087 ± 0.002 to 0.113 ± 0.004 seconds. (B) The time constant for PDE inactivation (τ = 1/k) was measured in response to increasing concentrations of hydroxylamine. The rate of inactivation of PDE saturated at τ = 9 seconds. Each point was determined using a single exponential fit to three separate traces and averaging τs. Error bars, SD. (C) PDE inactivation was monitored in the presence of 50 mM hydroxylamine with 500 μM ATP (solid line) or without ATP (broken line) in response to a pulse of light at time 0. Approximately 0.011% of the rhodopsin was bleached in these experiments.

Table 1.

 

Kinetic Parameters of Phototransduction Inactivation in the Presence and Absence of Hydroxylamine and Recoverin

	Time-to-Peak (sec)*	Inactivation Rate (sec), †
−Hydroxylamine	5.65 ± 0.15	0.074 ± 0.009
−Recoverin
−Hydroxylamine	11.25 ± 0.15	0.022 ± 0.001
+5 μM Recoverin
+20 mM Hydroxylamine	5.2 ± 0.1	0.112 ± 0.014
−Recoverin
+20 mM Hydroxylamine	5.75 ± 0.75	0.0855 ± 0.009
+5 μM Recoverin

Figure 3.

 

Effect of RGS9d on PDE inactivation. (A) The kinetics of PDE decay without hydroxylamine (NH₂OH) in the presence (solid line) and absence (broken line) of 10 μM RGS9d. (B) The kinetics of PDE decay with hydroxylamine in the presence (solid line) and absence (broken line) of 10 μM RGS9d. Bold line: single exponential fit to the data. RGS9d (10 μM) accelerated the decay of PDE from k = 0.116 ± 0.006 to 0.192 ± 0.004 seconds. (C) RGS9d was titrated in the presence and absence of hydroxylamine. τ for PDE decay was estimated from single exponential fits to the data and was averaged from three different traces. In these experiments the flash bleached approximately 0.01% of the rhodopsin. Error bars, SD.

Figure 4.

 

Single-turnover GTPase activity (A) PDE activity monitored during addition of 200 nM GTP to fully bleached ROS membranes (30 μM rhodopsin) at time 0. The rate of PDE decay was determined by a single exponential fit to the data, with (solid trace) and without (broken trace) 8 μM RGS9d. RGS9d accelerated the rate of PDE decay from k = 0.21 ± 0.03 to 0.34 ± 0.02 seconds. (B) The rate of GTP hydrolysis was measured directly using γ-labeled GTP. The smooth lines represent the rate of decay of PDE activity under identical conditions, either in the presence or absence of 8 μM RGS9d. Each point was determined in triplicate. Error bars, SD.

The authors thank Fred Rieke and Peter Detwiler for useful discussions during preparation of the manuscript.