GCaMP indicators targeted to pre- and post-synaptic compartments.

(A) Schematic of the presynaptic ratiometric Syt::mScarlet::GCaMP (Scar8f/Scar8m) Ca2+ indicators showing localization to synaptic vesicles via fusion to the Ca2+ sensor Synaptotagmin (SYT). (B) Representative images of NMJs expressing the indicated reporter driven in motor neurons with the OK319-GAL4 driver (w1118; OK319-GAL4/UAS-Scar8f) immunostained with anti-GFP (GCaMP) and anti-SYT. Note that endogenous mScarlet signals were obtained without antibody labeling. (C) Schematic of the BRP::mScarlet::GCaMP8f/8m (Bar8f/Bar8m) ratiometric Ca2+ indicator, which targets GCaMP to active zones via fusion to the BRP-short protein (Schmid et al., 2008). (D) Representative images of NMJs expressing the indicated reporter driven in motor neurons (w;OK6-GAL4/Bar8f and w;OK6-GAL4/Bar8m) immunostained with anti-GFP (GCaMP) and anti-BRP. Note that native mCherry or mScarlet signals were obtained without antibody labeling. (E) Schematic of the SynapGCaMP indicator, which targets GCaMP to postsynaptic compartments via a Shaker PDZ domain.(Newman et al., 2017) (F) Representative images of NMJs expressing the indicated reporter (w;MHC-CD8-GCaMP6f-Sh/+;+, w;;MHC-CD8-GCaMP8f-Sh/+, w;;MHC-CD8-GCaMP8m-Sh/+) immunostained with anti-GFP (GCaMP), -GluRIIC (glutamate receptors), and -DLG (postsynaptic density). All confocal images in this figure were deconvolved in Huygens (CMLE) prior to display (see methods).

“CaFire” - a Python-based analysis program for quantifying synaptic Ca2+ imaging data.

(A) Workflow showing how Ca2+ imaging data and downstream analysis is performed. Raw timelapse movies are processed with SVI Huygens software to correct and deconvolve image artifacts. ROIs are then selected in ImageJ Fiji software and intensities are extracted. Data are exported to Excel for manual inspection or directly analyzed using CaFire software. (B) Screenshots of the CaFire user interface. The left panels allow users to specify input file properties and analysis settings, and the second panel from the left shows the peak-detection criteria. The central panel displays fluorescence traces with automatically detected Ca²⁺ events overlaid and the event-based partition results. Detected events and their extracted parameters are listed in the data table below. The right panel shows the partition specifications used to segment evoked responses. (C) Examples of two distinct analysis pipelines implemented in CaFire. Evoked event analysis: (1) Thresholds are set for automatic peak detection; (2) Events are automatically partitioned based on stimulus intervals (e.g., 1 Hz and 5 Hz); (3) Parameters such as peak amplitude, rise time constant (τrise), and decay time constant (τdecay) are calculated using exponential fits. Mini event analysis: (1) Users define amplitude thresholds for event detection; (2) CaFire automatically identifies candidate events; (3) Missed or misidentified events can be manually corrected; (4) Event parameters are exported for each validated event.

Benchmarking presynaptic GCaMP8 to synthetic dyes and cytosolic sensors.

(A) Confocal images showing presynaptic loading of the synthetic dye OGB-1 and presynaptic expression of Scar8f at MN-Ib boutons. (B) Example single-AP evoked presynaptic Ca²⁺ transients reported by OGB-1 (black) and GCaMP8f (red). Traces represent the average of 30 APs from a single preparation acquired at 303 fps. (C) Quantification of response amplitude (ΔF/F), rise time, and decay times from single-AP Ca²⁺ transients reported by OGB-1 and Scar8f. (D) Example train-evoked Ca²⁺ transients reported by OGB-1 (black) and Scar8f (red) during three trains of 10 APs delivered at 21.3 Hz. (E) Quantification of train-evoked Ca²⁺ signals showing amplitude (ΔF/F) of the final Ca²⁺ transient in the train (17th), measured from baseline; frequency facilitation, calculated as the amplitude of the 17th transient divided by the amplitude of the 1st; and decay time constant (τdecay) measured after the final transient in the train. (F) Confocal NMJs images showing presynaptic expression of the cytosolic sensor RSET-8m (Cytosolic GCaMP8m) and Scar8m. (G) Example single-AP evoked Ca²⁺ transients reported by RSET-8m (black) and Scar8m (red) acquired at ∼115 fps. (H) Quantification of amplitude (ΔF/F), rise time (τrise) and decay time constants (τdecay) from single-AP Ca²⁺ transients reported by RSET-8m (w;OK319-GAL4/+; UAS-RSET-GCaMP8m/+) and Scar8m (w;OK319-GAL4/+; UAS-Syt::mScarlet::GCaMP8m/+). Shaded traces and bars show mean ±SEM. Statistical comparisons are unpaired two-tailed t-tests; significance is indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, not significant. Detailed statistics including p-values are provided in Table S1.

Scar8m is an optimal presynaptic Ca2+ indicator.

(A) Schematic of the UAS-Syt::mScarlet3::GCaMP8m (Scar8m) ratiometric Ca²⁺ indicator consisting of mScarlet3 and GCaMP8m targeted to synaptic vesicles via fusion to Synaptotagmin (SYT). (B) Representative images of a MN-Ib bouton expressing Scar8m (w;OK319-GAL4/+;Scar8m/+) resonant scanned at ∼115 fps. Fluorescence from GCaMP8m (green) and mScarlet3 (magenta) is shown at baseline, peak, decay, and recovery to baseline. (C) Normalized Ca²⁺ signals following single action potential (AP) stimulation. Averaged traces compare the kinetics of GCaMP6s, GCaMP8f, and GCaMP8m. (D) Representative GCaMP and mScarlet signals recorded from Syt::GCaMP6s (GCaMP only), Scar8f, and Scar8m (GCaMP plus mScarlet) in response to single AP stimuli. Thin traces are sequential single-trial sweeps from the same NMJ; the thick colored trace is the mean trace after stimulus alignment. The black curve indicates a single-exponential decay fit used to estimate τ. (E) Quantification of average peak amplitude (ΔR/R, GCaMP/mScarlet ratios), rise time constant (τrise), and decay time constant (τdecay) from the indicated sensors. Scar8m yields significantly higher peak ΔR/R signals, similar rise time kinetics, and a modestly slower decay compared to Scar8f. (F) Presynaptic Ca²⁺ responses of the indicated sensors to 5 Hz and 10 Hz stimulation trains. Vertical black ticks above the traces indicate the timing of stimulation pulses. Error bars represent ±SEM. Significance is indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, not significant; detailed statistics including p-values are presented in Table S1.

Scar8m captures differences in Ca2+ levels between motor neuron subtypes and after plasticity.

(A,B) Representative images of Scar8m expressed at both MN-Ib (A) and MN-Is (B) motor neuron subtypes immunostained with anti-GFP. (C) ΔF/F traces of GCaMP8m and mScarlet3 responses from single AP stimulation at MN-Ib and MN-Is, with ∼2x higher responses observed at MN-Is over -Ib, as expected. Thin traces are sequential single-trial sweeps from the same NMJ; the thick colored trace is the mean trace after stimulus alignment. The black curve indicates a single-exponential decay fit used to estimate τ. (D) Quantification of ΔR/R responses from the two inputs. (E) ΔF/F traces of GCaMP8m and mScarlet3 responses from single AP stimulation at MN-Ib in wild type (w;OK319-GAL4/+;Scar8m/+) and GluRIIA mutants (w;OK319-GAL4,GluRIIAPV3/GluRIIAPV3;Scar8m/+), which express presynaptic homeostatic plasticity (PHP). Note the enhanced presynaptic Ca2+ levels induced after PHP plasticity. (F) Quantification of ΔR/R responses from the indicated genotypes. Error bars represent ±SEM. Significance is indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, not significant; detailed statistics including p-values are presented in Table S1.

Evaluating the ability of Bar8f and Bar8m to capture active zone–specific Ca²⁺ changes.

(A) Schematic of the BRP::mScarlet3::GCaMP8m (Bar8m) ratiometric Ca²⁺ indicator, consisting of GCaMP8m and mScarlet3 fused to the BRP-short domain, which traffics to individual AZs. (B) Live confocal image of a single MN-Ib bouton expressing Bar8m. Dashed circles indicate active zones used for area-scan measurements, and the yellow dotted line indicates the line-scan ROI used for high-speed imaging. (C) Example Ca²⁺ transients from individual AZs recorded using Bar8f resonant area scans (∼60 fps; left), Bar8m resonant area scans (∼120 fps; middle), and Bar8m line scans (∼1000 fps; right). Black traces show the mean ratiometric response (ΔR/R) for Bar8f and Bar8m area scans and the mean ΔF/F response for Bar8m line scans from a representative NMJ; green shading denotes ±SEM across repeated trials. (D) Quantification of mean evoked response amplitude and τrise for the three imaging conditions shown in (C). (E) Bar graph showing peak ΔR/R (or ΔF/F) values from individual AZs collected across multiple boutons from different NMJs for Bar8f area scans, Bar8m area scans, and Bar8m line scans. Each bar represents a single active zone (sorted by amplitude within each condition). Light red shaded boxes indicate the interquartile range (IQR; Q1–Q3) of individual data points for each active zone, and active zones whose mean responses fall outside this IQR range are highlighted in pink. Significance is indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, not significant; detailed statistics including p-values are presented in Table S1.

SynapGCaMP8m is an optimal postsynaptic Ca²⁺ indicator.

(A) Schematic of the SynapGCaMP8m reporter, with GCaMP targeted to postsynaptic compartments near glutamate receptors via a Shaker-PDZ motif. Super resolution image using STED microscopy showing the GCaMP8m reporter is localized outside of GluRs. (B) Live confocal images of muscle 6 NMJ boutons expressing SynapGCaMP8m were performed using resonant area scans. The indicated ROI shows representative frames at baseline and peak quantal Ca²⁺ transients acquired at ∼115 fps. (C) Averaged single miniature Ca²⁺ events recorded from SynapGCaMP6f, -8f, and -8m. The thick green trace shows the mean ΔF/F waveform and the light green shading indicates ±SEM; the black curve over the decay phase is a single-exponential fit used to estimate τdecay. SynapGCaMP8m yields the highest peak signal and maintains rapid kinetics. The corresponding heatmaps below were generated from a single vertical line scan extracted from the representative miniature-event ROI, and visualize a spatiotemporal fluorescence dynamics (ΔF/F) along that line over time. (D) Representative Ca2+ events reporting individual miniature transmission in the indicated sensors. (E) Quantification of ΔF/F peak amplitude, and rise and decay time constants (τrise and τdecay) for each SynapGCaMP variant. All comparisons in bar graphs are statistically significant unless “ns” is shown. Error bars indicate ±SEM. Significance is indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; ns, not significant; detailed statistics including p-values are presented in Table S1.

SynapGCaMP8m quantal resolution approaches that of electrophysiology.

(A) Simultaneous recordings of quantal events at MN-Ib boutons using the indicated SynapGCaMP variant (green) and electrophysiology (black) after silencing MN-Is (w;Is-GAL4/+;UAS-BoNT-C/SynapGCaMP). Red circles indicate mEPSP events not captured by SynapGCaMP, and asterisks mark Ca²⁺ minis detected optically. SynapGCaMP8m captures quantal events with high sensitivity, comparable to electrophysiology. (B) Quantification of the proportion of quantal events captured by the indicated SynapGCaMP variant as a proportion of the total mEPSP events recorded by electrophysiology. SynapGCaMP6f detects only about half of electrophysiological events, while both SynapGCaMP8f and SynapGCaMP8m capture nearly all mEPSPs. (C) Scatter plot of paired miniature event amplitudes recorded simultaneously by SynapGCaMP6f and -8f and electrophysiology. Each point represents a single matched event; Pearson’s correlation coefficients (r=0.46 for 6f and r=0.73 for 8f) are indicated. (D) Scatter plot of paired miniature event amplitudes recorded simultaneously by SynapGCaMP8m and electrophysiology. Each point represents a single matched event; a strong linear relationship is observed (Pearson’s r=0.81), indicating that optical signals scale proportionally with quantal amplitude. (E) Bar plots showing average mEPSP amplitudes (left) and ΔF/F amplitudes of quantal Ca²⁺ events (right) in the indicated genotypes (same conditions as in (A) but including GluRIIA or GluRIIB mutant alleles). SynapGCaMP8m accurately resolves quantal size differences with similar resolution as the electrophysiological data, with quantal amplitudes in both datasets exhibiting the expected relationship (GluRIIA-/-<WT<GluRIIB-/-). Bars show mean ±SEM, dots represent individual boutons. (F) Cumulative frequency distributions of mEPSP amplitudes (left) and Ca2+ mini event amplitudes (right); each are significantly different using the Kolmogorov–Smirnov Test. See Table S1 for full statistical details including p-values.