Deciphering spatio-temporal fluorescence changes using multi-threshold event detection (MTED)

Abstract

Recent achievements in indicator optimization and imaging techniques promote the exploration of Ca2+ activity patterns as a main second messenger in many organs. Astrocytes are important regulators of brain activity and well known for their Ca2+-dependent modulation of neurons. However, standardized methods to analyze and interpret Ca2+ activity recordings are missing and hindering global comparisons. Here, we present a biophysics-based concept to analyze Ca2+signals, which includes multiple thresholds and provides the experimenter with a comprehensive toolbox for a differentiated and in-depth characterization of Ca2+ signals. We analyzed various ex vivo and in vivo imaging datasets and verify the validity of our multi-threshold event detection (MTED) algorithm across Ca2+ indicators, imaging setups, and model systems from primary cell culture to awake, head-fixed mice. Applying our MTED concept enables standardized analysis and advances research using optical readouts of cellular activity to decrypt brain function. It allowed us to obtain new insights into the complex dependence of Ca2+activity patterns on temperature and neuronal activity.Highlights→We present a robust pixel-based algorithm to analyze multidimensional fluorescence data.→Automated multiple-threshold analysis accurately quantifies changes in fluorescence across magnitudes.→It characterizes the complexity of dynamic and overlapping activity patterns of Ca2+ activity of astrocytes in vitro, in situ, and in vivo.→Its application provides quantitative parameters how temperature and neuronal activity determine astrocytic Ca2+ activity.

Novel algorithms for improved detection and analysis of fluorescent signal fluctuations

“…Three to four FOVs per animal were recorded in the somatosensory cortex, resulting in a total of 23 image sequences. Similar studies, investigating the effects of common anesthetics as well as natural sleep on astroglial Ca 2+ signaling in the neocortex had previously been performed (17, 18, 25) and served as reference.…”

“…The capabilities of PBasE, CoRoDe as well as MSparkles are demonstrated by analyzing differences in astroglial Ca 2+ events, recorded in the somatosensory cortex of awake and anesthetized mice. These differences have been studied before (17,20,21), serving as reference to our results. Next, the results obtained by MSparkles are compared to those obtained with three other Ca 2+ analysis tools, CHIPS (22), CaSCaDe (7) and AQuA (9).…”

“…These so-called microdomain events are elementary signals, serving autonomous functions (16), they are not location specific and can occur at all regions of the complex process network. Occurring in functionally independent cellular sub compartments, they occupy volumes in the sub-µm³ range and cause changes in fluorescence close to noise level (14, 17), rendering these events particularly difficult to characterize, requiring careful detection and removal of fluorescence background (17).…”

“…Fluorescence events are classically analyzed using stationary regions of interest (ROIs). Recently, new approaches utilizing so-called dynamic events (9, 17, 18) to extract and analyze non-stationary fluorescence events have been proposed. The analysis of dynamic events is an important extension and can reveal new insights.…”

“…The aim of PBasE is to accurately estimate basal fluorescence levels ( F 0 ) while simultaneously preserving fluorescent transients originating from phasic signaling. While PBasE builds upon the idea of penalizing potential fluorescence signals for the purpose of background estimation (17), an additional design goal is the ability to preserve slow, and long lasting fluorescence elevations. Building upon the background corrected dataset (( F − F 0 )/ F 0 = Δ F / F 0 ), CoRoDe is aimed at extracting and segmenting fluorescently active regions by explicitly leveraging temporal information.…”