The impact of reporter kinetics on the interpretation of data gathered with fluorescent reporters

Fluorescent reporters of biological functions are used to monitor biochemical events and signals in cells and tissue. For neurobiology, these have been particularly useful for monitoring signals in the brains of behaving animals. In order to enhance signal-to-noise, fluorescent reporters typically have kinetics that are slower than that of the underlying biological process. This low-pass filtering by the reporter renders the fluorescence transient a leaking integrated version of the biological signal. Here I discuss the effects that low-pass filtering, or more precisely of integrating by convolving with an exponentially decaying kernel, has on the interpretation of the relationship between the reporter fluorescence transient and the events that underlie it. Unfortunately, when the biological events being monitored are impulse-like, such as the firing of an action potential or the release of neurotransmitter, filtering greatly reduces the maximum correlation coefficient that can be found between the events and the fluorescence signal. This can erroneously support the conclusion that the fluorescence transient and the biological signal that it reports are only weakly related. Furthermore, when examining the encoding of behavioral state variables by nervous system, filtering by the reporter kinetics will favor the interpretation that fluorescence transients encode integrals of measured variables as opposed to the variables themselves. For these reasons, it is necessary to take into account the filtering effects of the indicator by deconvolving with the convolution kernel and recovering the underlying biological events before making conclusions about what is encoded in the signals emitted by fluorescent reporters.