IMPROVED IN VIVO RESPONSE KINETICS OF ENGINEERED GCaMP VARIANTS

Abstract

The use of genetically encoded calcium indicator proteins (GECIs) has become a fundamental component of neuroscience research. Despite the improvements made in the brightness and dynamic range of fluorescence, however, they continue to face limitations on performance relative to synthetic calcium indicators. One problem is their relatively slow response kinetics (up to 100-fold slower than synthetic indicators), which can debilitate their usefulness in reliably tracking single action potentials or detecting subtle changes in firing rates of neurons. Another problem arises from the high cooperativity of GECIs, which cause individual GECIs to only be sensitive to a very narrow range of calcium concentrations ([Ca2+]) whereas biologically relevant concentrations associated with neural activity can span a much wider range (0.1-10 μM). Improvements in these limitations on the temporal kinetics and range of GECIs will allow for greater resolution and applicability in studying neural activity.

Our laboratory has used a structure-based design approach to address the improvement of these properties in the popular GECI, GCaMP3. By targeting mutations on GCaMP3’s calcium-sensitive domains, calmodulin and its partner peptide RS20, we have developed over 50 new variants that exhibit a wide range of binding affinities and improved response kinetics over GCaMP3 without compromising brightness and fluorescent range. In the current study, several of the most promising variants from the in vitro assays have been expressed and functionally characterized in the Drosophila nervous system.

Fluorescent responses evoked by electrical stimulation in presynaptic boutons of the neuromuscular junction reveal that these variants show significant improvements in both fluorescence change and response speeds over GCaMP3. In addition, the structure-based approach used to generate these variants has given insight into the molecular mechanism of GCaMP fluorescence. Building on a molecular dynamic model of GCaMP3 activation constructed by our laboratory and from the observations of this research, this work proposes an adjusted model for the deactivation of GCaMP3.