SPONTANEOUS, LIGAND-INDEPENDENT ACTIVITY OF THE CGMP-GATED ION CHANNELS IN CONE PHOTORECEPTORS OF FISH

1. We studied the electrical conductance of membrane patches detached from the outer segment of single cone photoreceptors isolated from str iped bass retina. 2. Only a single class of ion channels exists in the plasma membrane of the cone outer segments; they are gated by cytopla smic cGMP and select cations over anions, but distinguish poorly among cations. In the absence of added cGMP and of divalent cations, howeve r, membrane patches detached from the outer segments exhibit a small c onductance that ideally selects cations over anions, but distinguishes poorly between Na+ and Li+. 3. The cGMP-independent conductance does not arise from the effect of residual cGMP that may remain associated with the detached membrane, because treatment of the patch with cGMP-s pecific phosphodiesterase does not affect this conductance. 4. The cGM P-independent conductance is pharmacologically indistinguishable from that activated by cGMP. Ca2+ and L-cis-diltiazem block both conductanc es at comparable concentrations and with similar quantitative characte ristics. 5. We analysed the noise of Ca2+- or L-cis-diltiazem-dependen t macroscopic currents both in the presence and in the absence of cGMP . In the presence of cGMP, the power density spectrum of the noise is well fitted by the sum of two Lorentzian components. The same function with similar corner frequencies fits the noise of the cGMP-independen t currents. However, the total power in the current fluctuations is sm aller in the absence of cGMP than in its presence; also, the ratio of the zero frequency asymptotes of the low over the high frequency compo nents, S-1(O)/S-h(O), is larger in the absence of cGMP than in its pre sence. In contrast, the noise of the current through the leak conducta nce between electrode and membrane patch does not exhibit a Lorentzian component over the frequency tested. 6. The maximum value of the cGMP -dependent conductance over all membrane patches sampled varies over a range of about 10-fold. Over this range, however, there is a linear r elationship between the cGMP-dependent and cGMP-independent conductanc es. The sum of our results suggest that the cGMP-independent conductan ce arises from the spontaneous activity of the ion channels ordinarily gated by cGMP. The cGMP-independent conductance is about 4.5% of that measured under saturating cGMP concentrations. 7. In the absence of d ivalent cations and cGMP, the spontaneous channel activity sustains a mean (+/- S.D.) membrane conductance of 0.18 +/- 0.11 nS in the detach ed patches. This predicts an outer segment conductance in the intact c ell that is about 500-fold larger in magnitude than the conductance ac tually measured in physiological solutions under bright light, when th e cytoplasmic cGMP concentration is expected to be nil. This differenc e cannot be explained by the known effect of divalent cations on the u nitary conductance of cGMP-gated channels alone. Divalent cations, the refore, must also control the probability of channel opening and, prob ably, prevent spontaneous channel activity.