On the reproducibility of single photon responses (SRPs): The gordian knot of rod phototransduction perseveres

Photoreceptors able to reliably signal the absorption of a single photon of light occur in virtually all vertebrate and invertebrate species examined. This sensitivity is critical to species' survival. The variability of SPRs is much lower than expected based on any chemical scheme in which the critical shutoff reaction inactivates with a single step, since such reactions have an exponential (highly variable) distribution of lifetimes. Since the single photon response results from a biochemical cascade initiated by activation of a single rhodopsin molecule (R*,), if R*, were shut off in a single step, SPRs would have highly variable durations and amplitudes. This would severely degrade single-photon detectability. We evaluated theories of reproducibility with Monte-Carlo simulations of SPRs using a detailed stochastic model of the phototransduction cascade. The candidates are: feedback, local saturation and some form of multi-step shutoff of R*,. Local saturation as the sole mechanism seems unlikely (Rieke &Baylor, 1998; Detwiler et al., 2002; Field & Rieke, 2002), and our present analyses show that Ca++-feedback per se is not adequate. However, we show that observed SPR reproducibility can be achieved with models in which R* activity decreases (by sequential phosphorylation) in as few as 6 or 7 stages, concomitant with a graded increase in the probability of R*-quenching by arrestin (Gibson et al., 2000). Nevertheless, our analyses show that it is surprisingly difficult to account for other key aspects of low-light-level rod responses without introducing mechanisms outside the bounds of known phototransduction biochemistry.