In intact mammalian photoreceptors, Ca2+-dependent modulation of cGMP-gated ion channels is detectable in cones but not in rods

Abstract

In the mammalian retina, cone photoreceptors efficiently adapt to changing background light intensity and, therefore, are able to signal small differences in luminance between objects and backgrounds, even when the absolute intensity of the background changes over five to six orders of magnitude. Mammalian rod photoreceptors, in contrast, adapt very little and only at intensities that nearly saturate the amplitude of their photoresponse. In search of a molecular explanation for this observation we assessed Ca2+-dependent modulation of ligand sensitivity in cyclic GMP-gated (CNG) ion channels of intact mammalian rods and cones. Solitary photoreceptors were isolated by gentle proteolysis of ground squirrel retina. Rods and cones were distinguished by whether or not their outer segments bind PNA lectin. We measured membrane currents under voltage-clamp in photoreceptors loaded with Diazo-2, a caged Ca2+ chelator, and fixed concentrations of 8Br-cGMP. At 600 nM free cytoplasmic Ca2+ the midpoint of the cone CNG channels sensitivity to 8BrcGMP, 8BrcGMPK1/2, is approximately 2.3 microM. The ligand sensitivity is less in rod than in cone channels. Instantly decreasing cytoplasmic Ca2+ to <30 nM activates a large inward membrane current in cones, but not in rods. Current activation arises from a Ca2+ -dependent modulation of cone CNG channels, presumably because of an increase in their affinity to the cyclic nucleotide. The time course of current activation is temperature dependent; it is well described by a single exponential process of approximately 480 ms time constant at 20-21 degrees C and 138 ms at 32 degrees C. The absence of detectable Ca2+-dependent CNG current modulation in intact rods, in view of the known channel modulation by calmodulin in-vitro, affirms the modulation in intact rods may only occur at low Ca2+ concentrations, those expected at intensities that nearly saturate the rod photoresponse. The correspondence between Ca2+ dependence of CNG modulation and the ability to light adapt suggest these events are correlated in photoreceptors.

Figure 1.

Micrographs of isolated ground squirrel photoreceptors. Side-by-side images of the same cells labeled with FITC-PNA captured either under DIC contrast enhancement or epifluorescent illumination. The top, left is a cone, characterized by a brightly fluorescent outer segment and a discrete fluorescent band in the inner segment. The top, right is of a rod, characterized by the absence of fluorescence in the outer segment and diffuse fluorescence throughout the inner segment. The images in the lower row were captured under DIC illumination and depict the range of cytological structure of isolated photoreceptors. OS outer segment IS inner segment, N nuclear region, and S synaptic terminal. Bar, 5 μm.

Figure 2.

Time continuous record of membrane current at −40 mV holding voltage in isolated ground squirrel cones (A) or rod (B). The arrow labeled WCM indicates the moment whole-cell mode was attained, defined as time zero. (A) Data from two different cones, one studied with a tight seal electrode filled with 0.8 μM 8Br-cGMP with 400 μM zaprinast and 600 nM free Ca²⁺ and the other with a solution lacking the nucleotide, as labeled. In the absence of nucleotide, current slowly changed to a final, steady outward value. In the presence of 8Br-cGMP, the inward current reflects activation of CNG ion channels with a time course determined by the rate of exchange/replacement between the cell's cytoplasmic content and the electrode-filling solution, as well as the cell's adjustment to the new, imposed Ca²⁺ and nucleotide concentrations. (B) Membrane current in a rod studied with a tight seal electrode filled with 8 μM 8Br-cGMP with 400 μM Zaprinast and 600 nM free Ca²⁺. After reaching WCM, the current changed with an exponential time course to a final, steady inward value. C illustrates the amplitude of the steady-state difference current measured in individual photoreceptors before and after attaining whole-cell mode in the presence of various 8Br-cGMP concentrations in the electrode filling solution. The continuous line depicts the Hill equation (text Eq. 1) optimally fit to the cone data with Imax 620 pA, n = 2.5 and ^8BrcGMPK_1/2 = 2.3 μM. The concentration range tested is too small to accurately define the features of the Hill function for rods.

Figure 3.

The effect of lowering cytoplasmic Ca²⁺ on cyclic nucleotide-dependent membrane currents. (A) The membrane current measured in a cone at −40 mV holding voltage with a tight-seal electrode filled with 0.8 μM 8BrcGMP, 1 mM dark Diazo-2, and 600 nM free Ca²⁺. 2.5 min after attaining whole-cell mode, free Ca²⁺ was lowered by uncaging Diazo-2 with a bright Xe flash. The moment the flash was presented defines time zero in the graphs. The flash discharge caused a fast (<10 ms) electrical artifact observed even with the electrode alone. Rapidly lowering Ca² (<50 ms) increased the inward current to a peak followed by a slow drift back toward its starting, steady-state value, to which it eventually returned (not depicted). (B) The membrane current measured in a rod at −40 mV with an electrode filled with 8 μM 8BrcGMP, 1 mM total dark Diazo-2, and 600 nM free Ca²⁺. Lowering cytoplasmic Ca²⁺ with the uncaging flash was without any effect on membrane current. (C and D) Membrane currents measured in the same cone and rod after flash illumination. Currents were activated by stepping membrane voltage to between −90 and 90 mV in 10-mV steps. The voltage- and time-dependent electrical behavior of rods and cones is indistinguishable and demonstrate both cells were normal and electrically intact.

Figure 4.

Current activation by flash uncaging Diazo-2 requires the presence of cyclic nucleotides and does not arise from an effect of light alone. Currents measured at −40 mV holding voltage in three different cones. The currents shown were measured at the steady holding current reached 2.5 min after attaining whole-cell mode. At time zero an uncaging Xe flash was presented and the lamp discharge caused a transient artifact. Every cone shown was loaded with solutions that contained 600 nM free Ca²⁺ and 1 mM free Mg²⁺, but differed in their 8Br-cGMP or Diazo-2 concentrations. The cone loaded with 0.8 μM 8Br-cGMP and 1 mM total dark Diazo-2 (as labeled) exhibited an exponential increase in the inward current amplitude upon flash illumination. The flash-dependent current enhancement was not observed if we omitted either Diazo-2 (replaced with 1 mM total BAPTA) or 8Br-cGMP from the cell-filling solution (as labeled).

Figure 5.

Kinetics of Ca²⁺-dependent modulation of CNG currents in cones. Cells studied in a temperature-controlled chamber at either 32°C (loaded with 1.5 μM 8Br-cGMP) or 20°C (loaded with 0.8 μM 8Br-cGMP). Currents measured at −40 mV at the steady holding current reached 2.5 min after attaining whole-cell mode. At that moment, a flash to uncage Diazo-2 was delivered and the instant of delivery defined the origin of the time axis. At either temperature, the flash-induced decrease in cytoplasmic-free Ca²⁺ activated an inward current of comparable amplitude, but different time course. The upper and lower tracings in each column are the same data displayed at different time resolution. Superimposed on the experimental data are first order exponential decay curves optimally fit to the results (text Eq. 2). At 32°C, Ihold = −75.3 pA, Imax = −123.3 pA, ΔICa = 48 pA, τ = 91.8 ms. At 20°C, Ihold = −71 pA, Imax = −103.5 pA, ΔICa = 32.5 pA, τ = 456 ms.

Figure 6.

CNG current modulation is lost over time from individual cells. Cones studied in a temperature-controlled chamber at either 32°C (loaded with 1.5 μM 8Br-cGMP) or 20°C (loaded with 0.8 μM 8Br-cGMP). (A and B) Currents repeatedly measured at −40 mV in the same cell at various times after first attaining whole-cell mode, as labeled in minutes. At each time, a flash to uncaged Diazo-2 was delivered and the instant of delivery defined the zero time in the scales in seconds. The holding current at −40 mV drifted in value to the extent illustrated in C and D for each cell. To compare repeated flash measurements in the same cell and because the holding current drifts, currents in A and B were DC shifted so that the holding current immediately before a flash a zero value. Each uncaging flash caused a change in inward current, but the extent of the change was progressively smaller as time elapsed after establishing whole-cell mode. Superimposed on the current measured after 2.5 min are first order exponential decay curves optimally fit to the results (text Eq. 2). At 32°C, Ihold = −94 pA, Imax = −134 pA, ΔICa = 40 pA, τ = 85.6 ms. At 20°C, Ihold = −65 pA, Imax = −93.7 pA, ΔICa = 28.8 pA, τ = 475 ms. The rate of modulation loss, represented by the slope of the line that passes through the points in C and D, is higher at 32°C than 20°C.