Fluorescence Fluctuations and Equivalence Classes of Ca (2+) Imaging Experiments

[Image: see text] release into the cytosol through inositol 1,4,5-trisphosphate receptors (IP(3)Rs) plays a relevant role in numerous physiological processes. IP(3)R-mediated [Image: see text] signals involve [Image: see text]-induced [Image: see text]-release (CICR) whereby [Image: see text] release through one open IP(3)R induces the opening of other channels. IP(3)Rs are apparently organized in clusters. The signals can remain localized (i.e., [Image: see text] puffs) if CICR is limited to one cluster or become waves that propagate between clusters. [Image: see text] puffs are the building blocks of [Image: see text] waves. Thus, there is great interest in determining puff properties, especially in view of the current controversy on the spatial distribution of activatable IP(3)Rs. [Image: see text] puffs have been observed in intact cells with optical techniques proving that they are intrinsically stochastic. Obtaining a correct picture of their dynamics then entails being able to detect the whole range of puff sizes. [Image: see text] puffs are observed using visible single-wavelength [Image: see text] dyes, slow exogenous buffers (e.g., EGTA) to disrupt inter-cluster CICR and UV-photolyzable caged IP(3). Single-wavelength dyes increase their fluorescence upon calcium binding producing images that are strongly dependent on their kinetic, transport and photophysical properties. Determining the artifacts that the imaging setting introduces is particularly relevant when trying to analyze the smallest [Image: see text] signals. In this paper we introduce a method to estimate the expected signal-to-noise ratio of [Image: see text] imaging experiments that use single-wavelength dyes. The method is based on the Number and Brightness technique. It involves the performance of a series of experiments and their subsequent analysis in terms of a fluorescence fluctuation model with which the model parameters are quantified. Using the model, the expected signal-to-noise ratio is then computed. Equivalence classes between different experimental conditions that produce images with similar signal-to-noise ratios can then be established. The method may also be used to estimate the smallest signals that can reliably be observed with each setting.