An improved measurement of the Ca-binding affinity of fluorescent Ca indicators.

Fluorescent Ca indicators are widely used to measure the intracellular Ca concentration ([Ca]) in living cells, including neurons. By calibrating an indicator in solutions that mimic the main ionic constituents of the actual cytoplasm, [Ca] can be determined from the measured fluorescence intensity. However, different studies have reported considerably different Ca-binding affinities (K) for the same indicator, even though they used calibrating solutions with similar compositions. In this paper, we present a method to accurately determine the K values of non-ratiometric Ca indicators in solutions that mimicked a standard patch-clamp internal solution. The free Ca concentration ([Ca]) in these solutions, which was set by either EGTA or HEDTA, was measured with a Ca-selective macroelectrode. We found that such a measurement was critical for an accurate calibration of the Ca indicators. The K values of OGB-1, OGB-6F, fluo-5F, and fluo-4FF were 0.26 ± 0.01, 8.7 ± 0.4, 1.00 ± 0.05, and 23.0 ± 0.7 μM, respectively. Calculating [Ca] with Maxchelator, a widely used computer program, led to a significant underestimation of the K values of OGB-6F, fluo-5F, and fluo-4FF. This is because the purity of EGTA was considerably less than that advertised by the manufacturer. In addition, the K value of HEDTA was overestimated by Maxchelator. Therefore, besides batch-to-batch variations, the fact that [Ca] in the calibrating solutions of many studies was estimated with Maxchelator is very likely a reason for the different published values of K of Ca indicators. Using a reaction-diffusion model to reproduce Ca rises in a nerve terminal, we further showed that incorrect calibration of fluorescent Ca indicators can underlie the large variation of the endogenous Ca binding ratio between different types of excitatory synapses.