Stochastic Treatment of Bump Latency and Temporal Overlapping in Limulus Ventral Photoreceptors

Abstract

We present quantum bumps obtained from flash experiments at the Limulus ventral nerve photoreceptor under voltage clamp conditions. The results are shown and discussed in form of histograms for the latency, amplitude and net charge transfer (current time integral) of the bump current responses. We argue that the experimental latency histogram s cannot be described satisfactorily by chemical models if one assumes that not more than one photon is captured per flash. Instead of, one has to take into account the Poisson statistics of the captures of 0,1,2 ,... photons released by a single flash. We show that the inclusion of Poisson statistics makes the effective latency histograms of flash responses typically asymmetric and skewed to wards short latencies as compared to that of model histograms for one-photon responses. Our conjecture also implies that under our experimental conditions a fraction of up to 20% of the bump responses evoked by a flash should be suspected to be superpositions of two ore more one-photon responses which cannot be separated by any kind of evaluation analysis. Consequently, the average values of amplitudes and net charge transfers of the light-evoked bump responses are expected to be overestimated as compared to that of true one-photon responses. This hypothesis is confirm ed by a numerical simulation of light-evoked bump responses using experimentally recorded spontaneous bumps (at times larger than 1 s after the flash) as the simulation material. We show that the superposition of one-photon events in the light-evoked bump responses due to Poisson statistics settles the question why their amplitudes and net charge transfers are found to be larger than that of the spontaneous bumps. We suggest that true one-photon responses evoked by a light flash and spontaneous bumps start from the same activated rhodopsin state and take the same biochemical pathway.