Single‐channel measurement from the cyclic GMP‐activated conductance of catfish retinal cones.

Publisher Website
Google Scholar
Abstract
1. A patch of plasma membrane was excised, in the inside-out configuration, from the outer segment tip of a catfish cone and recorded electrically with a patch pipette. A solution of 118 mM-NaCl was present on both sides of the membranes. 2. With the solution outside of the pipette containing a low concentration (typically several micromoles per litre) of cyclic GMP and the membrane potential held at a non-zero level, brief steps of current indicative of the openings of single ion channels could be detected. There was no sign of desensitization to the ligand over a period of tens of seconds. 3. The prominent openings were associated with a conductance near 50 pS and an open-time constant of 0.5 ms or less. There was also an indication of sub-state openings. 4. The conductance of the large openings appeared to be invariant between -50 mV and +50 mV. However, the macroscopic current-voltage relation measured at a saturating concentration of cyclic GMP showed a slight upward curvature, which we attribute to a voltage dependence in the open probability of the fully liganded channel. 5. The relation between mean current and cyclic GMP concentration had an average Hill coefficient of about 2.4. The Hill coefficient was not affected by membrane voltage, but the conductance was activated by cyclic GMP slightly more readily at depolarizations; this could be adequately explained by a higher open probability of the fully liganded channel at positive voltages. 6. In several experiments, the membrane patch apparently contained a single cyclic GMP-activated channel, in that the measured current never rose above that for a single channel even at high concentrations of cyclic GMP. In these cases, a high concentration of the ligand simply engaged the channel in a literally continuous burst of openings, with an open probability of 0.8-0.9 at between -30 mV and +30 mV. The amplitude distribution of the burst under these conditions could be described by a beta distribution, consistent with the channel switching predominantly between a single closed state and a single open state when fully liganded. 7. Estimates of channel density on the cone membrane ranged from about 2 to 130 .mu.m-2, with an average of 20 .mu.m-2. This observed density is about ten times lower than the density of the homologous channel on rod membrane, being roughly in inverse relation to the tenfold larger surface membrane area of the cone outer segment. 8. Computer simulation based on kinetic information extracted from the above results and a scheme in which the channel can enter into a large open state when fully liganded and into a sub-open state when one less cyclic (GMP molecule is bound is able to produce channel activity that resembles the recorded activity in temporal characteristics. 9. The similarities between the rod conductance and the cone conductance are discussed.