Gap-junctional coupling and absolute sensitivity of photoreceptors in macaque retina

Abstract

We investigated gap-junctional coupling of rods and cones in macaque retina. Cone voltage responses evoked by light absorption in neighboring rods were briefer and smaller than responses recorded in the rods themselves. Rod detection thresholds, calculated from noise and response amplitude histograms, closely matched the threshold for an ideal detector limited by quantal fluctuations in the stimulus. Surprisingly, cone thresholds were only approximately two times higher. Amplitude fluctuations in cones could be explained by a Poisson distribution of photoisomerizations within a pool of seven or more coupled rods. Neurobiotin coupling between rods and cones was consistent with our electrical recordings, with approximately six rods labeled per injected cone. The spatial distribution of tracer-coupled rods matched the light-evoked cone receptive field. The gap junction inhibitor carbenoxolone abolished both electrical and tracer coupling. Amplitude fluctuations in most rods were accounted for by the expected rate of light absorption in their outer segments. The fluctuations in some rods, however, were consistent with a summation pool of up to six rods. When single rods were injected with Neurobiotin, up to 10 rods were labeled. Rod-rod and rod-cone electrical coupling is expected to extend the range of scotopic vision by circumventing saturation at the rod to rod-bipolar cell synapse; however, because coupling also renders the rod synapse less effective at separating out photon signals from dark noise, coupling is expected to elevate the absolute threshold of dark-adapted observers.

Figure 1.

Rod signals measured in cones and rods. A, B, Photovoltage responses in a rod (A) or cone (B) to flashes of increasing flash strength ranging from 0.3 to 803 photons μm^-2 in A and 2.0 to 3874 photons μm^-2 in B. Traces are averages of one to four responses. A, Bandwidth, DC-20 Hz. B, Bandwidth, DC-50 Hz. C, D, Response fluctuations to dim flashes in a rod (C) or cone (D). The smooth curves through the noisy measured responses are the scaled response templates from matched filtering. Flash photon densities are 1.0 photons μm^-2 (C) and 2.0 photons μm^-2 (D); bandwidth, DC-5 Hz.

Figure 2.

High-pass filtering of rod signals in cones. A, Dim flash responses averaged across 7 rods (dashed trace) and 13 cones (solid trace) and normalized by photoisomerizations per rod. See Materials and Methods for averaging procedure. Bandwidth, DC-10 Hz. B, Same responses as in A on an expanded time scale, with cone response rescaled to match rod response peak. C, Amplitude density of the Fourier components of the flash responses in A.

Figure 3.

Response fluctuations in dim light. Ensemble variance (solid line) and square of ensemble mean (dashed line) recorded in a rod (A) and cone (B) from 100-147 flashes. Flashes = 0.5 photons μm^-2 (A) and 2.0 photons μm^-2 (B). Bandwidth, DC-5 Hz.

Figure 4.

Evidence of rod-rod signal coupling. A, Average rod response at three flash strengths. Bandwidth, DC-5 Hz; 142-147 responses per average. B, Circles plot peak amplitude of mean responses in A as a function of flash strength. The line is the least squares fit of the data points to a line intersecting the origin (slope, 1.17 mV per R^*/rod). C, Ensemble variance of the same responses as in A. D, Circles plot peak amplitude of the variance in C as a function of flash strength. The solid line is the best linear fit through the origin (slope, 0.265 mV² per R^*/rod). Dashed line indicates the expected function of an uncoupled rod. From the slopes in B and D, a = 0.23 mV and N = 5.2 rods.

Figure 5.

Probability density distribution of response amplitudes of a rod (A, C) and cone (B, D). Dark noise histograms p_N (A, B) and signal histograms p_S (C, D), obtained from matched filtering, are plotted by the bars. Number of trials = 675-805 (A, B) and 143-149 (C, D). Flash strength = 1.0 photons μm^-2. Smooth curves drawn through noise histograms are Gaussian distributions with SD σ₀ = 0.43 mV (A) and 0.12 mV (B). Smooth curves through signal histograms are from Equation 1, with the constants in C and D, respectively: i = 1.0, 1.0 photons μm^-2; a = 0.23, 0.049 mV; N = 5.2, 3.1 rods; σ₀ = 0.43, 0.12 mV; and σ₁ = 0.03, 0 mV. P_C = 0.95 (A, C), and P_C = 0.82 (B, D).

Figure 6.

Detection thresholds of rods and cones. A, Frequency distribution of thresholds (at P_C = 0.73) from 10 rods (striped bars) and 16 cones (shaded bars). B, Electrical coupling reduces detection threshold. Thresholds versus rod pool size N in 10 rods (•) and 4 cones (○). N was obtained from variance and means analysis (see Materials and Methods). Smooth curve is for an ideal (noiseless) detector, given by -[ln(0.54)] N^-1 R^*/rod.

Figure 7.

Tracer and signal coupling between rods and cones is blocked by the gap junction inhibitor carbenoxolone. A, B, Combined confocal and Nomarski contrast images of the outer nuclear layer after cone injection of Neurobiotin (green). The cone in A, injected in control solution, is coupled to neighboring rods. The cone in B, injected in the presence of 100 μm carbenoxolone, was not tracer coupled to other cells. C, Rod responses measured in a cone in control solution (top) and again after application of 100 μm carbenoxolone (bottom). Flashes = 28 photons μm^-2. Traces are averages of four to six responses. Bandwidth, DC-5 Hz. D, Three-dimensional reconstruction of Neurobiotin-labeled rods and cones after injection in a cone. OS, Outer segment; ONL, outer nuclear layer; OPL, outer plexiform layer. Scale bars: A, B, D, 10 μm. Results in A-D were obtained from four separate cone recordings.

Figure 8.

Correspondence of tracer coupling and receptive field. A, Neurobiotin labeling (green) of rods coupled to injected cone (asterisk). Scale bar, 10 μm. B, Rod responses in the cone shown in A to flashes of light covering a hemifield rotated around the recorded cone. Traces are averages of 6-12 responses to flashes of 16 photons μm^-2. Bandwidth, DC-5 Hz. Cone membrane potential was recorded in whole-cell mode. Symbols to left of traces denote orientation of the stimulus hemifield relative to the recorded cone.

Figure 9.

Tracer and signal coupling between rods. A, B, Combined confocal and Nomarski contrast images after Neurobiotin injection of single rods. Only the injected rod is labeled in A; four rods were labeled in B. C, Three-dimensional reconstruction of rods in B. OS, Outer segment; ONL, outer nuclear layer; OPL, outer plexiform layer. Scale bars: A-C, 10 μm. D, Frequency distribution of rod pool size (N). Tracer coupling is indicated by green bars. Signal coupling, determined from mean and variance analysis of dim flash responses, is indicated by striped bars. N = 1 indicates an uncoupled rod.

Figure 10.

Effect of rod-rod coupling on psychophysical thresholds of dark-adapted human observers. A, Psychophysical threshold (R^*) in a two-alternative forced-choice experiment, calculated from model simulations, as a function of rod pool size N (for details of the model, see Materials and Methods). The shaded region represents the range of experimentally determined electrical pool sizes of macaque rods (N = 1-6). The dashed line indicates the expected threshold in the absence of synaptic nonlinearities. B, Psychophysical threshold intensity of coupled rods (R_C ^*) relative to uncoupled rods (R_U ^*) as a function of stimulus diameter, with N = 2. Coupling lowered threshold (R_C ^*/R_U ^* < 1) for stimulus diameters <0.066° (arrow).