cAMP controls rod photoreceptor sensitivity via multiple targets in the phototransduction cascade

In early studies, both cyclic AMP (cAMP) and cGMP were considered as potential secondary messengers regulating the conductivity of the vertebrate photoreceptor plasma membrane. Later discovery of the cGMP specificity of cyclic nucleotide–gated channels has shifted attention to cGMP as the only secondary messenger in the phototransduction cascade, and cAMP is not considered in modern schemes of phototransduction. Here, we report evidence that cAMP may also be involved in regulation of the phototransduction cascade. Using a suction pipette technique, we recorded light responses of isolated solitary rods from the frog retina in normal solution and in the medium containing 2 µM of adenylate cyclase activator forskolin. Under forskolin action, flash sensitivity rose more than twofold because of a retarded photoresponse turn-off. The same concentration of forskolin lead to a 2.5-fold increase in the rod outer segment cAMP, which is close to earlier reported natural day/night cAMP variations. Detailed analysis of cAMP action on the phototransduction cascade suggests that several targets are affected by cAMP increase: (a) basal dark phosphodiesterase (PDE) activity decreases; (b) at the same intensity of light background, steady background-induced PDE activity increases; (c) at light backgrounds, guanylate cyclase activity at a given fraction of open channels is reduced; and (d) the magnitude of the Ca²⁺ exchanger current rises 1.6-fold, which would correspond to a 1.6-fold elevation of [Ca²⁺]_in. Analysis by a complete model of rod phototransduction suggests that an increase of [Ca²⁺]_in might also explain effects (b) and (c). The mechanism(s) by which cAMP could regulate [Ca²⁺]_in and PDE basal activity is unclear. We suggest that these regulations may have adaptive significance and improve the performance of the visual system when it switches between day and night light conditions.     

The kinetics of inactivation of the rod phototransduction cascade with constant Ca2+i

A rich variety of mechanisms govern the inactivation of the rod phototransduction cascade. These include rhodopsin phosphorylation and subsequent binding of arrestin; modulation of rhodopsin kinase by S- modulin (recoverin); regulation of G-protein and phosphodiesterase inactivation by GTPase-activating factors; and modulation of guanylyl cyclase by a high-affinity Ca(2+)-binding protein. The dependence of several of the inactivation mechanisms on Ca2+i makes it difficult to assess the contributions of these mechanisms to the recovery kinetics in situ, where Ca2+i is dynamically modulated during the photoresponse. We recorded the circulating currents of salamander rods, the inner segments of which are held in suction electrodes in Ringer's solution. We characterized the response kinetics to flashes under two conditions: when the outer segments are in Ringer's solution, and when they are in low-Ca2+ choline solutions, which we show clamp Ca2+i very near its resting level. At T = 20-22 degrees C, the recovery phases of responses to saturating flashes producing 10(2.5)-10(4.5) photoisomerizations under both conditions are characterized by a dominant time constant, tau c = 2.4 +/- 0.4 s, the value of which is not dependent on the solution bathing the outer segment and therefore not dependent on Ca2+i. We extended a successful model of activation by incorporating into it a first-order inactivation of R*, and a first-order, simultaneous inactivation of G-protein (G*) and phosphodiesterase (PDE*). We demonstrated that the inactivation kinetics of families of responses obtained with Ca2+i clamped to rest are well characterized by this model, having one of the two inactivation time constants (tau r* or tau PDE*) equal to tau c, and the other time constant equal to 0.4 +/- 0.06 s.     

Evolution of vertebrate rod and cone phototransduction genes

Vertebrate cones and rods in several cases use separate but related components for their signal transduction (opsins, G-proteins, ion channels, etc.). Some of these proteins are also used differentially in other cell types in the retina. Because cones, rods and other retinal cell types originated in early vertebrate evolution, it is of interest to see if their specific genes arose in the extensive gene duplications that took place in the ancestor of the jawed vertebrates (gnathostomes) by two tetraploidizations (genome doublings). The ancestor of teleost fishes subsequently underwent a third tetraploidization. Our previously reported analyses showed that several gene families in the vertebrate visual phototransduction cascade received new members in the basal tetraploidizations. We here expand these data with studies of additional gene families and vertebrate species. We conclude that no less than 10 of the 13 studied phototransduction gene families received additional members in the two basal vertebrate tetraploidizations. Also the remaining three families seem to have undergone duplications during the same time period but it is unclear if this happened as a result of the tetraploidizations. The implications of the many early vertebrate gene duplications for functional specialization of specific retinal cell types, particularly cones and rods, are discussed.     

The retinal phototransduction process: enzymatic cascade and regulation

Among cellular systems performing the transduction of an external stimulus, phototransduction in vertebrate rod cells is a unique case which allows convergent approaches to electrophysiological, biophysical and biochemical analyses. The framework of the molecular processes involved in the corresponding enzymatic cascade is now elucidated and can be considered as a model for G protein mediated transductions. We present here the main features of this cascade, its amplification and regulation properties. The mode of stimulation by the aluminofluoride ion is particularly addressed.     

Mathematical and computational modelling of spatio-temporal signalling in rod phototransduction

Rod photoreceptors are activated by light through activation of a cascade that includes the G protein-coupled receptor rhodopsin, the G protein transducin, its effector cyclic guanosine monophosphate (cGMP) phosphodiesterase and the second messengers cGMP and Ca2+. Signalling is localised to the particular rod outer segment disc, which is activated by absorption of a single photon. Modelling of this cascade has previously been performed mostly by assumption of a well-stirred cytoplasm. We recently published the first fully spatially resolved model that captures the local nature of light activation. The model reduces the complex geometry of the cell to a simpler one using the mathematical theories of homogenisation and concentrated capacity. The model shows that, upon activation of a single rhodopsin, changes of the second messengers cGMP and Ca2+ are local about the particular activated disc. In the current work, the homogenised model is computationally compared with the full, non-homogenised one, set in the original geometry of the rod outer segment. It is found to have an accuracy of 0.03% compared with the full model in computing the integral response and a 5200-fold reduction in computation time. The model can reconstruct the radial time-profiles of cGMP and Ca2+ in the interdiscal spaces adjacent to the activated discs. Cellular electrical responses are localised near the activation sites, and multiple photons sufficiently far apart produce essentially independent responses. This leads to a computational analysis of the notion and estimate of 'spread' and the optimum distribution of activated sites that maximises the response. Biological insights arising from the spatio-temporal model include a quantification of how variability in the response to dim light is affected by the distance between the outer segment discs capturing photons. The model is thus a simulation tool for biologists to predict the effect of various factors influencing the timing, spread and control mechanisms of this G protein-coupled, receptor-mediated cascade. It permits ease of simulation experiments across a range of conditions, for example, clamping the concentration of calcium, with results matching analogous experimental results. In addition, the model accommodates differing geometries of rod outer segments from different vertebrate species. Thus it represents a building block towards a predictive model of visual transduction.     

Stochastic simulation of activation in the G-protein cascade of phototransduction.

Activation of the G-protein cascade underlying phototransduction has been modeled by simulating the two-dimensional diffusional interactions that occur at the rod disc membrane between the three reacting protein species, which are the activated rhodopsin (R*), the G-protein (G), and the effector protein (E, the phosphodiesterase, PDE). The stochastic simulations confirm the main predictions of a simplified analytical model (Lamb, T. D., and E. N. Pugh, 1992, Journal of Physiology 449:719-758), and extend that treatment to more complicated cases, where there is a finite probability of reaction or a finite time for reaction. The simulations also provide quantitative estimates of the efficiency of coupling from activated G-protein (G*) to activated effector (E*) in terms of the concentrations, lateral diffusion coefficients, and binding rate constants of the participating molecules; the efficiency of coupling from G* to E* is found to be not as high as in the previous simplified analytical theory. The findings can be extended to other G-protein cascades, provided that the physical parameters of those cascades are specified.     

Autophagy supports survival and phototransduction protein levels in rod photoreceptors

Damage and loss of the postmitotic photoreceptors is a leading cause of blindness in many diseases of the eye. Although the mechanisms of photoreceptor death have been extensively studied, few studies have addressed mechanisms that help sustain these non-replicating neurons for the life of an organism. Autophagy is an intracellular pathway where cytoplasmic constituents are delivered to the lysosomal pathway for degradation. It is not only a major pathway activated in response to cellular stress, but is also important for cytoplasmic turnover and to supply the structural and energy needs of cells. We examined the importance of autophagy in photoreceptors by deleting the essential autophagy gene Atg5 specifically in rods. Loss of autophagy led to progressive degeneration of rod photoreceptors beginning at 8 weeks of age such that by 44 weeks few rods remained. Cone photoreceptor numbers were only slightly diminished following rod degeneration but their function was significantly decreased. Rod cell death was apoptotic but was not dependent on daily light exposure or accelerated by intense light. Although the light-regulated translocation of the phototransduction proteins arrestin and transducin were unaffected in rods lacking autophagy, Atg5-deficient rods accumulated transducin-α as they degenerated suggesting autophagy might regulate the level of this protein. This was confirmed when the light-induced decrease in transducin was abolished in Atg5-deficient rods and the inhibition of autophagy in retinal explants cultures prevented its degradation. These results demonstrate that basal autophagy is essential to the long-term health of rod photoreceptors and a critical process for maintaining optimal levels of the phototransduction protein transducin-α. As the lack of autophagy is associated with retinal degeneration and altered phototransduction protein degradation in the absence of harmful gene products, this process may be a viable therapeutic target where rod cell loss is the primary pathologic event.Autophagy is an intracellular pathway where cytoplasmic constituents are delivered to the lysosomes for degradation. Defective autophagy can contribute to the age-dependent accumulation of damaged proteins and organelles leading to altered cellular homeostasis and loss of function.^{1, 2, 3, 4, 5} The metabolic roles of autophagy can be classified into two types, basal and induced. In nutrient-rich conditions, autophagy is suppressed but still occurs at low levels (basal autophagy); however, when cells are subjected to stress (starvation, injury, hypoxia), autophagy is activated immediately (induced autophagy).⁶ Induced autophagy maintains the amino acid pool inside cells to adapt to starvation while constitutive autophagy has been shown to function as a cell-repair mechanism that is important for long-lived postmitotic cells.^{7, 8, 9, 10, 11} Defects in autophagy have been associated with neurodegenerative diseases,^{12, 13, 14, 15} diabetes,^{16, 17} lysosomal storage disease¹⁸ and the loss of vision.¹⁹ In addition to macroautophagy, microautophagy and chaperone-mediated autophagy (CMA) have been described. Although little is known about microautophagy in mammalian cells, macroautophagy (hereafter autophagy) is a major pathway for bulk degradation of cytoplasmic components. CMA is a more selective pathway for degradation of cytosolic proteins that can compensate for the loss of macroautophagy.^{2, 20, 21, 22}Inherited retinal degenerative diseases such as retinitis pigmentosa or Leber''s congenital amaurosis are characterized by premature and progressive death of rod and cone photoreceptor cells.²³ These diseases are characterized by the loss of night vision due to the death of rods followed by the loss of cones leading to diminished visual acuity and a reduction in the quality of life for patients. Disease is typically associated with the production of harmful gene products that promote pathology by inhibiting critical pathways resulting in cell death.^{24, 25, 26} Strategies to prevent photoreceptor death during retinal degenerative disease such as gene replacement therapies or inhibition of cell death pathways have been undertaken with some success;^{27, 28, 29} however, effective treatments for these blinding disorders are lacking.Another strategy that could be used in conjunction with other therapies might be to enhance survival by stimulating autophagy. Augmenting autophagy would increase the supply of nutrients to stressed cells and accelerate removal of damaged proteins thereby prolonging cell survival beyond what would be possible by only preventing cell death. Although canonical^{22, 30, 31, 32, 33} and noncanonical autophagic mechanisms³⁴ have been detected in the eye, our knowledge of basic autophagy functions in this organ is still limited. In order to understand how autophagy maintains retinal homeostasis and function, we undertook studies to examine the consequences of deleting the essential autophagy gene Atg5 in rod photoreceptors.     

The gain of rod phototransduction: reconciliation of biochemical and electrophysiological measurements

We have resolved a central and long-standing paradox in understanding the amplification of rod phototransduction by making direct measurements of the gains of the underlying enzymatic amplifiers. We find that under optimized conditions a single photoisomerized rhodopsin activates transducin molecules and phosphodiesterase (PDE) catalytic subunits at rates of 120-150/s, much lower than indirect estimates from light-scattering experiments. Further, we measure the Michaelis constant, Km, of the rod PDE activated by transducin to be 10 microM, at least 10-fold lower than published estimates. Thus, the gain of cGMP hydrolysis (determined by kcat/Km) is at least 10-fold higher than reported in the literature. Accordingly, our results now provide a quantitative account of the overall gain of the rod cascade in terms of directly measured factors.     

Effect of 11-cis 13-demethylretinal on phototransduction in bleach-adapted rod and cone photoreceptors

We used 11-cis 13-demethylretinal to examine the physiological consequences of retinal's noncovalent interaction with opsin in intact rod and cone photoreceptors during visual pigment regeneration. 11-Cis 13-demethylretinal is an analog of 11-cis retinal in which the 13 position methyl group has been removed. Biochemical experiments have shown that it is capable of binding in the chromophore pocket of opsin, forming a Schiff-base linkage with the protein to produce a pigment, but at a much slower rate than the native 11-cis retinal (Nelson, R., J. Kim deReil, and A. Kropf. 1970. Proc. Nat. Acad. Sci. USA. 66:531-538). Experimentally, this slow rate of pigment formation should allow separate physiological examination of the effects of the initial binding of retinal in the pocket and the subsequent formation of the protonated Schiff-base linkage. Currents from solitary rods and cones from the tiger salamander were recorded in darkness before and after bleaching and then after exposure to 11-cis 13-demethylretinal. In bleach-adapted rods, 11-cis 13-demethylretinal caused transient activation of phototransduction, as evidenced by a decrease of the dark current and sensitivity, acceleration of the dim flash responses, and activation of cGMP phosphodiesterase and guanylyl cyclase. The steady state of phototransduction activity was still higher than that of the bleach-adapted rod. In contrast, exposure of bleach-adapted cones to 11-cis 13-demethylretinal resulted in an immediate deactivation of transduction as measured by the same parameters. These results extend the validity of a model for the effects of the noncovalent binding of a retinoid in the chromophore pockets of rod and cone opsins to analogs capable of forming a Schiff-base and imply that the noncovalent binding by itself may play a role for the dark adaptation of photoreceptors.     

The carbohydrates in frog retinal rod outer segments

Frog retinal rod outer segments appear to contain uncharacterized chemical components whose mass is roughly equivalent to 12--51% of the rhodopsin mass. Available data suggest that such components include soluble proteins and complex polysaccharides, and that hyaluronic acid accounts for a substantial fraction of this mass. Electron microscopic histochemical staining studies suggest that these polysaccharide components are located within the ROS disks. The oligosaccharide moieties of rhodopsin also appear localized within the disks. The interdisk cytoplasm may contain carbohydrates, but their quantity and identity are uncertain. Rhodopsin oligosaccharides as well as some fraction of the intradisk polysaccharide appear to have extended saccharide chains preferentially oriented perpendicular to the surface of the disk membrane. Possible roles for these polysaccharides in disk development and photoexcitation are discussed. The immediate need for complete rod outer segment chemical composition data is emphasized.     

Protein complement of rod outer segments of frog retina

Rod outer segments (ROS) from frog retina have been purified by Percoll density gradient centrifugation, a procedure that preserves their form and intactness. One- and two-dimensional electrophoretic analysis reveals a smaller number of proteins than is observed in many cell organelles and permits quantitation of the 20 most abundant polypeptides. Rhodopsin accounts for 70% of the total protein (3 X 10(9) copies/outer segment), and approximately 70 other polypeptides are present at more than 6 X 10(4) copies/outer segment. Another 17% of the total protein is accounted for by the G-protein (3 X 10(8) copies/outer segment) that links rhodopsin bleaching and the activation of cyclic GMP phosphodiesterase (PDE). The phosphodiesterase accounts for 1.5% of the protein (1.5 X 10(7) copies/outer segment), and a 48,000-dalton component that binds to the membrane in the light accounts for a further 2.6%. The function of approximately 90% of the total protein in the outer segment is known, and two-thirds of the non-rhodopsin protein is accounted for by enzyme activities associated with cyclic GMP metabolism. The relative molar abundance of rhodopsin, G-protein, and PDE is 100:10:1. Apart from these major membrane-associated proteins, most of the other proteins are cytosolic. Thirteen other polypeptides are found at an abundance of one or more copies per 1000 rhodopsins, nine soluble and four membrane-bound, and their abundance relative to rhodopsin has been quantitated. ROS have been separated into subcellular fractions which resolve three classes of soluble, extrinsic membrane, and integral membrane proteins. A listing of the proteins that are phosphorylated and their subcellular localization is given. Approximately 25 phosphopeptides are detected, and most are in the soluble fraction. Fewer phosphorylated proteins are associated with the purified outer segments than with crude ROS. Distinct patterns of phosphorylation are associated with intact rods incubated with [32P]Pi and broken rods incubated with [gamma-32P]ATP.     

Kinetics of the classical complement activation cascade

Transient-phase and steady-state equations have been derived for the classical complement activation mechanism. From these equations the corresponding ones for the rapid equilibrium conditions have been derived. We propose an experimental design to determine the kinetic parameters.     

Calcium content of frog rod outer segments and discs

To test the “Ca²⁺ hypothesis of visual excitation”, we measured the total Ca²⁺ content of freshly isolated bullforg rod outer segments, and have compared the total Ca²⁺ contents of fully dark-adapted discs with discs exposed to small amounts of light. Discs were prepared by hypotonically lysing outer segments under conditions expected to remove Ca²⁺ from the cytoplasm but not from the discs. Ca²⁺ was assayed by atomic absorption spectrophotometry. We find that both discs and outer segments contain a total of about 0.1–0.2 Ca²⁺ per rhodopsin molecule. Thus, each frog disc retains about $\text{2 · 10}{}^5\text{Ca}{}^{\mathrm{2+}}$. If most of this Ca²⁺ were free in the aqueous space inside the intact discs, the Ca²⁺ activity would be a few mM. Since the light-regulated Na⁺ channels have been reported to be highly sensitive to cytoplasmic Ca²⁺, this store of Ca²⁺ in the discs is far more than required by the Ca²⁺ hypothesis. However, despite several variations in experimental conditions, we did not observe any light-activated release of Ca²⁺ from discs in response to stimuli that photoactivated a small fraction of the rhodopsin, as required by the Ca²⁺ hypothesis. In the 26 experiments reported here we could have detected a release as small as 20–30% of the Ca²⁺ content of the disc.     

Superoxide dismutase of bovine and frog rod outer segments.

Bovine rod outer segments (ROS) contain soluble superoxide dismutase (SOD) which from cyanide sensitivity and electrophoretic mobility appears identical to CuZn SOD of erythrocytes. Enzyme activity of ROS extracts is 200–400 times as much as remainder of retina. Frog ROS also contains a cyanide-sensitive SOD which is not due to erythrocyte contamination since the retina is avascular. SOD in ROS may inhibit free radical oxidation of polyunsaturated fatty acids. In light, high oxygen concentrations in developing retina may activate lipid peroxidation leading to retrolental fibroplasia. High concentrations of ascorbic acid in the retina may act as a protective mechanism against superoxide.     

Two forms of intermediates of frog rhodopsin in rod outer segments

Using frog rod outer segments, we measured changes of the absorption spectrum during the conversion of rhodopsin to a photosteady-state mixture composed of rhodopsin, isorhodopsin and bathorhodopsin by irradiation with blue light (440 nm) at ? 190°C and during the reversion of bathorhodopsin to a mixture of rhodopsin and isorhodopsin by irradiation with red light (718 nm) at ? 190°C. The reaction kinetics was expressed by one exponential in the former case and by two exponentials in the latter. These results suggest that rhodopsin is composed of a single molecular species, while bathorhodopsin is composed of two kinds of molecular species designated as batho₁-rhodopsin and batho₂-rhodopsin. On warming the two forms of bathorhodopsin, each bathorhodopsin converted to its own lumirhodopsin, metarhodopsin I and finally a free all-trans-retinal plus opsin. The absorption spectra of the two forms of bathorhodopsin, lumirhodopsin and metarhodopsin I were measured at ? 190°C. We infer that a rhodopsin molecule in the excited state relaxes to either batho₁-rhodopsin or batho₂-rhodopsin, and then converts to its own intermediates through one of the two parallel pathways.