Replacing the rod with the cone transducin subunit decreases sensitivity and accelerates response decay

Abstract

Cone vision is less sensitive than rod vision. Much of this difference can be attributed to the photoreceptors themselves, but the reason why the cones are less sensitive is still unknown. Recent recordings indicate that one important factor may be a difference in the rate of activation of cone transduction; that is, the rising phase of the cone response per bleached rhodopsin molecule (Rh*) has a smaller slope than the rising phase of the rod response per Rh*, perhaps because some step between Rh* and activation of the phosphodiesterase 6 (PDE6) effector molecule occurs with less gain. Since rods and cones have different G-protein alpha subunits, and since this subunit (Talpha) plays a key role both in the interaction of G-protein with Rh* and the activation of PDE6, we investigated the mechanism of the amplification difference by expressing cone Talpha in rod Talpha-knockout rods to produce so-called GNAT2C mice. We show that rods in GNAT2C mice have decreased sensitivity and a rate of activation half that of wild-type (WT) mouse rods. Furthermore, GNAT2C responses recover more rapidly than WT responses with kinetic parameters resembling those of native mouse cones. Our results show for the first time that part of the difference in sensitivity and response kinetics between rods and cones may be the result of a difference in the G-protein alpha subunit. They also indicate more generally that the molecular nature of G-protein alpha may play an important role in the kinetics of G-protein cascades for metabotropic receptors throughout the body.

Figure 1. Generation and characterization of GNAT2C mice

A, transgenic construct. B, a representative Western blot showing relative levels of cone transducin α subunit (GNAT2) signal from 1 μg of GNAT2C and 9 μg of WT retinal extracts derived individually from four retinas of each genotype. (His)₆-tagged recombinant GNAT2 protein (in ng) was used as standards for measurement of absolute amount in each retina. C, the pixel value of each recombinant protein band shown in B in arbitrary units (AU) was plotted against the amount of recombinant protein (open triangles) loaded to construct a standard curve. The 16 ng data point was excluded due to apparent signal saturation. The averaged levels of native GNAT2 per retina from four different GNAT2C (open square) and WT (filled square) retinal samples were determined by extrapolation to be 2220 ± 140 ng and 31 ± 4 ng (measurements of 16 samples from 4 mice). D, a representative Western blot showing the level of rod transducin α subunit (GNAT1) from seven WT mouse retinal samples (9 μg) against known amounts of (His)₆-tagged recombinant GNAT1 protein. E, the pixel value of each native GNAT1 signal (filled triangles) in arbitrary units (AU) was plotted against signals from duplicate recombinant GNAT1 proteins (open triangles). The averaged level of GNAT1 per retina was determined by extrapolation to be 2190 ± 60 ng (measurements of 35 samples from 5 mice).

Figure 2. Expression levels of proteins and retinal structure

A, a representative Western blot of 10 μg each of WT, GNAT2C and GNAT2C/R9AP95 retinal extracts, probed with a cocktail of primary antibodies consisting of A670 (1:2000, anti-GC-F, from David Garbers, UT Southwestern), PA1-720 (1:1000, anti-PDE6α, Affinity Bioreagents), MA1-720 (G8, 1:5000, anti-GRK1, Affinity Bioreagents), CT318 (1:2500, anti-RGS9-1, from Melvin Simon, Caltech), UUTA2 (1:3000, anti-GNAT2), UUTA1 (1:5000, anti-GNAT1), BN-1 (1:10,000, anti-GNB1, from Melvin Simon, Caltech), 14C10 (1:20,000, anti-GAPDH, Cell Signaling Technology), Gerti (1:5000, anti-phosducin, from Rehwa Lee, UCLA) and DSC-Rv (1:10,000, anti-recoverin) to simultaneously detect proteins involved in phototransduction. Shown here is an autoradiogram exposed for 30 s immediately following enhanced chemiluminescence. Note the increased expression of RGS9-1 in GNAT2C/R9AP95 retinas and different mobilities of GNAT1 and GNAT2 under our experimental conditions. B, semi-thin sections of WT (left) and GNAT2C (right) mouse retinas. OS, outer segment; IS, inner segment; ONL, outer nuclear layer; INL, inner nuclear layer, OPL, outer plexiform layer. Scale bar, 20 μm. C, confocal images of GNAT2C retinas under dark-adapted conditions (left) and after exposure to room light for 30 min after pupil dilatation (right). Redistribution of cone transducin outside of the outer segment layer is evident after light exposure. OS, outer segments; IS, inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer. Scale bar, 20 μm.

Figure 5. Differences in rate of activation and decay of WT and GNAT2C rods

A, mean small-amplitude responses of 21 WT rods and 9 GNAT2C rods to flashes of intensities 17 (WT) and 79 (GNAT2C) photons μm⁻². Same responses as in Fig. 3. Responses have been normalized rod by rod to peak amplitude of response to compare waveform of response decay. Responses have been fitted with single exponentials (grey traces) of 258 ms (WT) and 122 ms (GNAT2C). B, black traces are mean initial time courses of responses of 16 WT rods to 10 ms flashes at intensities of 8.6, 21 and 79 photons μm⁻², after filtering with an 8-pole Bessel filter with bandwidth set to 70 Hz and sampled at 200 s⁻¹. Responses have been normalized rod by rod to the peak amplitude of the photocurrent (r_max). Grey traces are fits to data with the same mean values of A of 20.5 s⁻² and t_eff of 18 ms at all three intensities. C, black traces are mean initial time courses of responses recorded and normalized as in B of 14 GNAT2C rods to 10 ms flashes at intensities of 21, 79 and 227 photons μm⁻². Grey traces are fits to data with A of 10.2 s⁻² and t_eff of 19.3 ms. Single dashed grey curve gives prediction for brightest intensity with rod value of A (20.5 s⁻²).

Figure 4. Sensitivity to flashes and backgrounds

A, mean peak amplitude of responses averaged from same rods as in Fig. 3 and plotted as function of flash intensity for WT (filled squares), GNAT2C (open squares), and GNAT2C/R9AP95 (open triangles) mice. Data have been fitted with , where r is the peak amplitude of the response to the flash, r_max is the maximum amplitude of r, I is the flash intensity, and k is a constant. Best-fitting values of k were 0.026 for WT, 0.0057 for GCAT2C and 0.0073 for GNAT2C/R9AP95. Only curves for WT and GNAT2C are shown in figure. Note that fewer rods were used for calculating the response–intensity functions than for Table 1, accounting for small differences in sensitivity and peak response amplitude. B, increment sensitivity of response to 20 ms flash in the presence of a steady background light. Sensitivity in pA photon⁻¹μm² was calculated in darkness () and in the presence of background light (S_F) as the peak amplitude of the response in the linear range divided by the flash intensity. Ordinate gives mean sensitivity in background light divided by sensitivity in the absence of a background as a function of background intensity, averaged from 10 WT rods (filled squares), 11 GNAT2C rods (open squares), and 10 GNAT2C/R9AP95 (open triangles) rods. Continuous lines are best-fitting Weber–Fechner functions given by eqn (1) for WT rods with I₀= 77 photons μm⁻² s⁻¹ and for GNAT2C rods with I₀= 546 photons μm⁻² s⁻¹. A fit of the data for GNAT2C/R9AP95 rods to eqn (1) gave a best-fitting value of I₀= 585 photons μm⁻² s⁻¹ (not shown).

Figure 3. Suction-electrode recordings of rod light responses to graded series of flashes showing acceleration of response decay after expression of cone transducin

A, mean response waveforms averaged from 21 WT rods to 20 ms flashes at intensities of 4, 17, 43, 159, 453 and 863 photons μm⁻². B, mean response waveforms averaged from 9 GNAT2C rods to 10 ms flashes at intensities of 8.6, 21, 79, 227, 561, 1220 and 2110 photons μm⁻². C, mean response waveforms averaged from 12 GNAT2C/R9AP95 rods to 10 ms flashes at intensities of 8.6, 21, 79, 227, 561 and 1220 photons μm⁻².