Properties of ion channels closed by light and opened by guanosine 3',5'-cyclic monophosphate in toad retinal rods.

1. In patch-clamp recordings from outer segments of dark-adapted rod photoreceptors, single-channel recordings were obtained from the light-sensitive conductance when divalent cations were omitted from the pipette solution bathing the extracellular face of the recorded patch of membrane. 2. Activity of the light-sensitive channel was suppressed by light within the normal response range of the dark-adapted rod. During dim, steady illumination, the rate of opening of the channel fluctuated dramatically, as expected qualitatively from statistical fluctuations in the number of photoisomerizations occurring within the effective collecting area of the recorded patch. 3. The light-sensitive channel flickered rapidly in the open state, so that individual events appeared as a burst of openings and closings. The average duration of a burst was 0.78 +/- 0.03 ms (mean +/- S.E.). The average duration of an individual opening was 0.18 +/- 0.008 ms. The average closed duration within a burst was 0.37 +/- 0.02 ms. 4. Hyperpolarization of the recorded patch had no effect on average burst or open duration, although opening frequency increased slightly (+18.6 +/- 4.9%; n = 13; mean +/- S.E.). Average single-channel current increased linearly with hyperpolarization, giving an estimated single-channel conductance of 20.5 +/- 1.1 pS. By extrapolation of the relation between channel current and hyperpolarization, the dark driving force was estimated to be about 48 mV. 5. In addition to reducing the rate of channel events, dim non-saturating light also reduced the average duration of a burst of openings and the average duration of openings within a burst. 6. About 50% of cell-attached patches showed no channel activity in darkness. Light-suppressable channel activity could be induced in these silent patches by perfusing the outer segment with low-Ca2+ Ringer solution. Similarly, activity could be increased dramatically by low-Ca2+ Ringer solution in patches that did show channel activity in the dark. From the maximal channel activity observed during low-Ca2+ perfusion, the lower limit for the number of channels per patch was 20-70, corresponding to an estimated channel density of 100-350 channels micron-2. 7. After recording light-sensitive channel activity in the intact rod, the patch of membrane was excised, exposing the intracellular membrane face. Application of guanosine 3',5'-cyclic monophosphate (cyclic GMP) to the intracellular face activated channels (Haynes, Kay & Yau, 1986; Zimmerman & Baylor, 1986; Matthews, 1986d, 1987) whose properties could then be compared directly with the light-sensitive channels recorded earlier in the same patch of membrane.(ABSTRACT TRUNCATED AT 400 WORDS)