Calcium is a key signaling molecule in all cells. In muscle, the rise and fall of intracellular calcium concentration [Ca]i triggers contraction and relaxation, and many other subcellular signaling pathways. Dysfunctional calcium handling can cause lethal cardiac arrhythmia, as well as numerous other diseases of the heart and other tissues. To better understand normal function and disease mechanisms, fluorescent calcium indicators and microscopic imaging are widely used to measure [Ca]i changes in detail.
Cardiac contraction is triggered by the synchronous release of calcium from many thousands of local release sites throughout each cell. Between beats, there are rare spontaneous calcium release events that appear as tiny sparks of fluorescent light from the indicator, confined around individual release sites at around 1 micrometer in diameter. These are called “Ca Sparks”, and under pathological conditions, these sparks can propagate between sites to produce Ca waves that can lead to cardiac arrhythmias. Ca sparks are also formed in smooth muscle cells, and their dysfunction is associated with pathology, such as hypertension.
Image analysis of these Ca sparks and waves is important in understanding what goes wrong in disease states and for testing novel therapeutic strategies and potential drugs. However, Ca spark data files are large and challenging to analyse by hand, so customised software analysis tools are critical.
© Jakub TomekA research team led by Drs Jakub Tomek, Christopher Ko, and Donald M. Bers in collaboration with Drs Manuel F. Navedo and Madeline Nieves-Cintron at the Department of Pharmacology, UC Davis School of Medicine have developed a new user-friendly software tool called SparkMaster 2 for automated analysis of Ca sparks in microscopy data.
The research team said: “The software is highly accurate, robust, and can distinguish between distinct patterns of calcium release. For example, when calcium sparks initiate large “Ca waves” propagating through the cell, prior analysis tools could not handle this well, but it has become possible now. SparkMaster 2 is also open source, easy to adopt, and completely free to use. It has been designed to work rapidly and with maximum degree of automation, making it suitable for applications in high-throughput settings, such as for analysing results of compound screening. It also allows the analysis of calcium sparks recorded using different microscopy approaches as well as cells from different tissues and genetically-modified mice. Thus, the tool will have broad applicability.
Dr Tomek added: “One challenge we faced was to measure how much of an improvement SparkMaster 2 brings with regards to the accuracy of detecting calcium sparks in the data. In the end, we created realistic-looking synthetic calcium spark data, which enables us to with certainty where the sparks are. It surprised us that SparkMaster 2 not only clearly outperformed the previous gold standard software, but it also surpassed the detection performance of our human colleagues (for whom it took much more time as well). We don’t claim that SparkMaster 2 is superior to human annotators for all the data it may encounter, but nevertheless, we consider this result to be encouraging.
“It was great to be able to present SparkMaster 2 at recent conferences before it was published and see the enthusiastic response. We have learned that others were keen to have a new analysis tool like SparkMaster 2, that overcomes many of the limitations they have experienced with prior tools.”
More information about the software can be found in ‘SparkMaster 2: A new software for automatic analysis of calcium spark data’, which is available to read in Circulation Research.
© UC Davis-based team: Donald M. Bers (sitting), Manuel F. Navedo (standing left), Christopher Ko (standing middle), Madeline Nieves-Cintron (right)
*Full description of SparkMaster 2 outputs image: On the left is an example of data collected using a line-scan confocal microscope. A single line through the cell is scanned within repeatedly (ca. 500 times per second), and calcium indicator fluorescence is recorded. Time goes from top to bottom and the location along the cell length is left-to-right. When single sparks recruit further sparks, this can lead to the emergence of propagating calcium wave, as shown starting near the middle of the cell and propagating to the left. The right panel shows how SparkMaster 2 detects calcium sparks, miniwaves, and waves. For each of the recorded events, the software records numerous event properties, such as location, size, amplitude, dynamics of rise and fall of fluorescence to a spreadsheet, for further analyses.