Fluorescence intensity is a poor predictor of saturation effects in two-photon microscopy: artifacts in fluorescence correlation spectroscopy.

Fluorescence correlation spectroscopy (FCS) has become an increasingly important measurement tool for biological and biomedical investigations, with the capability to assay molecular dynamics and interactions both in vitro and within living cells. Information recovery in FCS requires an accurate characterization and calibration of the observation volume. A number of recent reports have demonstrated that the calibration of the observation volume is excitation power dependent, a complication that arises due to excitation saturation. While quantitative models are now available to account for these volume variations, many researchers attempt to avoid saturation issues by working with low nonsaturating excitation intensities. For two-photon excited fluorescence, this is typically thought to be achievable by working with excitation powers for which the total measured fluorescence signal maintains its quadratic dependence on excitation intensity. We demonstrate that observing only the power dependence of the fluorescence intensity will tend to underestimate the importance of saturation, and explain these findings in terms of basic physical models. Copyright (c) 2007 Wiley-Liss, Inc.