Genetically encoded green fluorescent Ca2+ indicators with improved detectability for neuronal Ca2+ signals

Abstract

Imaging the activities of individual neurons with genetically encoded Ca(2+) indicators (GECIs) is a promising method for understanding neuronal network functions. Here, we report GECIs with improved neuronal Ca(2+) signal detectability, termed G-CaMP6 and G-CaMP8. Compared to a series of existing G-CaMPs, G-CaMP6 showed fairly high sensitivity and rapid kinetics, both of which are suitable properties for detecting subtle and fast neuronal activities. G-CaMP8 showed a greater signal (F(max)/F(min) = 38) than G-CaMP6 and demonstrated kinetics similar to those of G-CaMP6. Both GECIs could detect individual spikes from pyramidal neurons of cultured hippocampal slices or acute cortical slices with 100% detection rates, demonstrating their superior performance to existing GECIs. Because G-CaMP6 showed a higher sensitivity and brighter baseline fluorescence than G-CaMP8 in a cellular environment, we applied G-CaMP6 for Ca(2+) imaging of dendritic spines, the putative postsynaptic sites. By expressing a G-CaMP6-actin fusion protein for the spines in hippocampal CA3 pyramidal neurons and electrically stimulating the granule cells of the dentate gyrus, which innervate CA3 pyramidal neurons, we found that sub-threshold stimulation triggered small Ca(2+) responses in a limited number of spines with a low response rate in active spines, whereas supra-threshold stimulation triggered large fluorescence responses in virtually all of the spines with a 100% activity rate.

Figure 1. Characterization of G-CaMPs in vitro and in HeLa cells.

A, Schematic representation. Mutations are indicated with respect to G-CaMP2. RSET and M13 are tags that encode hexahistidine and a target peptide for Ca²⁺-bound CaM derived from MLCK, respectively. The amino-acid numbers of EGFP and CaM are indicated in parentheses. B, Dynamic range (F _max/F _min) and Ca²⁺ affinity (K _d). Error bars, s.d. (n = 3 each). C, Ca²⁺ titration curve. Curves were fit according to the Hill equation. K _d is shown in B. D, Normalized fluorescence and absorbance (inset) spectra of G-CaMP6 and G-CaMP8 in 1 µM Ca²⁺ or 1 mM EGTA. E, Fluorescence images of HeLa cells expressing G-CaMPs. Bars, 30 µm. F, Time course of the changes (ΔF/F) in G-CaMP fluorescence in response to 100 µM ATP. Error bars, s.d. G, Baseline fluorescence and peak responses (ΔF/F) to ATP application in HeLa cells.

Figure 2. Characterization of G-CaMPs in cultured hippocampal slices.

A, Expression of G-CaMP6 in hippocampal CA3 pyramidal neurons. Inset: Higher-magnification views are shown in the bottom panels. B, Baseline fluorescence of hippocampal neurons expressing G-CaMP3, G-CaMP6 and G-CaMP8. No significant differences in variance were detected among the three groups (P>0.05, χ² = 2.90, Bartlett’s test). Error bars, s.d. (n = 7 each, P>0.05, Tukey’s test). C, Representative traces of the response (ΔF/F) to spike trains. The frequency of stimuli was 50 Hz. Right: Magnified views of single spikes. D, Mean responses (ΔF/F) and SNRs of G-CaMP3 (black), G-CaMP6 (red) and G-CaMP8 (blue). Inset: Magnified views of 1–2 spikes. Error bars, s.e.m. (n = 7 each). E, Rise and decay time constants for the responses to single spikes. Error bars, s.e.m. (n = 7 each; *P<0.05 in Tukey’s post-hoc test following one-way ANOVA). F, Trial-averaged responses of G-CaMP6 to spike trains. Gray, individual traces (n = 10 trials); red, averaged traces. Bars indicate stimulus timing. Inset: Magnified views.

Figure 3. Comparison of G-CaMP responses in acute cortical slices.

A, Confocal image of G-CaMP6-expressing cortical pyramidal cells. The expression of G-CaMP6 was driven by the CAG promoter via in utero plasmid electroporation. Inset: Higher-magnification views are shown in the right panels. B, Representative ΔF/F traces in response to 1–4 spikes evoked at 50 Hz. C, Mean responses (ΔF/F) of G-CaMP3 and G-CaMP6 to spike trains. Error bars, s.e.m. (G-CaMP3, n = 4 cells; G-CaMP6, n = 5 cells).

Figure 4. Temperature dependence of G-CaMP6 signals.

A, Representative traces of the fluorescence response (ΔF/F) of G-CaMP6 to a single spike at 25–28°C and at 37°C. B, Mean responses (ΔF/F) of G-CaMP6 to spike trains. Error bars, s.e.m. (n = 6 each). C, Rise and decay time constants of the responses of G-CaMP6 to single spikes. (*P<0.05, paired t-test).

Figure 5. Electrophysiological properties of hippocampal neurons expressing G-CaMP6.

A, Left, input resistance. Middle, membrane capacitance. Right, resting potential. Error bars, s.e.m. (n = 6 each). There were no significant differences between the control and G-CaMP6 groups for any of the parameters (P>0.05, Student’s t-test). B, Left, spontaneous current under the voltage clamp at –70 mV. Middle, amplitude of the excitatory postsynaptic current. Right, frequency of the excitatory postsynaptic current. Error bars, s.e.m. (n = 6 each, P>0.05, Student’s t-test).

Figure 6. Ca²⁺ imaging of cholinergic DA motoneurons in freely moving C. elegans.

A, Confocal images of L1 larvae expressing G-CaMP6 (jqEx97) or G-CaMP3 (jqEx216) in the DA motoneurons. In both transgenic strains, DsRed-Express-1 is co-expressed in the DA motoneurons. TL, transmitted-light image. Arrows indicate the DA7 motoneuron analyzed in B. B, Representative spontaneous fluorescence responses (ΔR/R) of G-CaMPs from DA7 cholinergic neurons in transgenic worms during locomotion. C, Mean peak responses (ΔR/R). Error bars, s.e.m. (n = 10 each from 4 worms, *P = 0.0020, Student’s t-test). Movies of the recordings are available as supplementary information (Movies S1 and S2).

Figure 7. Ca²⁺ Imaging of individual spines in cultured hippocampal slices.

A, Schematic representation of G-CaMP6-actin. B, Schematic drawing of the placement of a stimulation electrode and a patch pipette in a cultured hippocampal slice. C, Z-projection of a representative CA3 pyramidal neuron expressing G-CaMP6-actin. The position of the patch pipette is indicated by dotted lines. Two spines of interest (S1 and S2) in the striatum lucidum are indicated by yellow circles. Inset: High-magnification views of rectangular windows. D, Changes in fluorescence at S1and S2 and membrane potential (V _m) upon sub- or supra-threshold electrical stimulation (Stim). E, Mean responses of active spines. Error bars, s.d. (n = 63 and 131 spines for sub- and supra-threshold stimulation, respectively, from 5 slices); *P<0.05, Student’s t-test. The average ΔF/F of the soma in response to supra-threshold stimulation was 15±6.5%.

Figure 8. Long-term imaging of Ca²⁺ activity in spines in a cultured hippocampal pyramidal neuron.

A, Z-projection of a representative CA3 pyramidal neuron expressing G-CaMP6-actin at 8 (upper) and 29 (lower) days in vitro (Div). After 7 days in vitro, the G-CaMP6-actin plasmid was introduced into the neuron via single-cell electroporation. Two spines of interest (S1, S2) are indicated by yellow circles. B, Changes in fluorescence at S1 and S2 upon supra-threshold electrical stimulation (Stim). The average spine ΔF/F ratios in response to supra-threshold stimulation were 253±30.5% and 201±46.6% at 8 Div and 29 Div, respectively (n = 25 spines, P>0.05, Student’s t-test).