Nanophysiology approach unveils spatiotemporal properties of local Ca2+ signaling in retinal RBC terminals.

A. Left panel. Single projection from a series of confocal optical sections through a zebrafish RBC synaptic terminal. A synaptic terminal was voltage-clamped using a whole-cell pipette with an internal solution containing TAMRA-RBP (magenta) to label synaptic ribbons (a, magenta). RBP fluorescence was concentrated at ribbons and also filled the entire terminal, allowing visualization of the terminal border. Experiments were carried out on ribbons that could be distinguished from adjacent ribbons (white asterisks). Scale, 5 μm. Right panel, Close-up view of a single synaptic ribbon. The outside of the cell is to the right, and x-t scan lines (dotted lines) were positioned perpendicular to the plasma membrane, extending from the intracellular side of the ribbon to the extracellular space. A rapid x-t line scan was taken at a ribbon location perpendicular to the plasma membrane across a ribbon (b) with sequential dual laser scanning performed at rates of 1.51 milliseconds per line per channel (3.02 milliseconds per line for both channels). The resulting x-t raster plots were used to measure the fluorescence intensity profiles of RBP (magenta) and the Ca2+ transient (cyan; Cal520HA) shown in panels C and E. Scale, 1.6 μm.

B. Voltage-clamp recording of a RBC terminal. Terminals were held at -65 mV and stepped to 0 mV (t0) for 10 ms (red) to evoke a brief Ca2+ current (black). A typical experiment began with a voltage command (VH= -65 mV), and a TTL pulse generated by the Patch Master software (t1) triggers image acquisition (t2). t0 is the time of depolarization.

C. Illustration of the approach used to obtain the spatial location of Ca2+ signals with respect to the ribbon. Example of an x-t raster plot that is oriented to illustrate the x-axis intensity profiles of RBP (magenta) and Ca2+ signal (cyan) fluorescence during a brief depolarization. Sequential dual laser scanning was performed at 1.51 milliseconds per line for one channel (3.02 milliseconds per line for both channels).

D. Fluorescence intensity profiles along the x-axis for RBP (magenta) and Cal520HA before stimuli (gray line), during 10 ms depolarization (cyan line), and after stimuli (black line) depolarization were obtained by averaging three pixels along the time-axis. RBP (magenta) and Ca2+ signals during (light cyan) were fit with a Sigmoid-Gaussian function 1. The centroid (x-axis position) of the RBP (magenta) and Ca2+ signals during (cyan) were taken as the peak of the Gaussian fit (x0). The parameter x1/2 (dotted magenta line) from the Sigmoid fit to the RBP fluorescence (magenta trace) was used to estimate the location of the plasma membrane.

E. The x-t raster plot shown was from the same recording as in panel C but re-oriented to demonstrate the t-axis intensity profiles of RBP (magenta) and Ca2+ transient (cyan).

F. Spatially averaged Cal520HA fluorescence as a function of time at ribbon proximal (light cyan), distal (dark cyan), and cytoplasmic (gray) locations from the single ribbon shown in D, upper panel.

Kinetics of Ca2+ transients in response to brief stimuli recorded with freely diffusible indicators.

A. Spatially averaged Cal520HA fluorescence as a function of time at ribbon proximal (light cyan) and distal (dark cyan) locations from a single ribbon, as shown in Fig.1 (n=24 ribbons from 7 different RBCs). The corresponding maximum values of trial-averaged ΔF/Frest were significantly higher at ribbon-proximal locations than those at ribbon-distal locations, with mean ± SEM of 1.9 ± 0.3 and 1.5 ± 0.25, respectively (paired t-test, p< 0.001, N=24).

A inset. The temporal profile of events between 20-60 ms is shown in an expanded view for better visualization. Scale bars: vertical, 0.5 (ΔF/Frest); horizontal,10 ms.

B. Spatially averaged Cal520LA fluorescence as a function of time at ribbon proximal (light cyan) and distal (dark cyan) locations. Data points show the average intensity (± SEM) in each horizontal row of 5 pixels for three 10 ms depolarizations at distinct ribbon locations (see Materials and Methods and Fig.1). Fluorescence intensity is normalized with respect to the baseline fluorescence before stimulation, and averaged over all pixels (i.e., over space and time). The arrow indicates the onset of the 10-ms depolarizing stimulus. (n=21 ribbons from 4 different RBCs). The peak ΔF/Frest at the membrane after 10-ms depolarization was significantly larger for proximal than distal signals (paired t-test, ΔF/Frest: 3.1 ± 0.4 and 1.9 ± 0.2, respectively, P=0.001, N=21). The currents were not significantly different between the Cal520HA and Cal520LA conditions (unpaired t-test, Cal520HA average current = 45.1 ± 4.5 pA; Cal520LA average current = 51.4 ± 4.4 pA; p = 0.33).

B inset. Temporal profile of events between 20-60 ms was expanded for better visualization. Scale bars: vertical, 0.5 (ΔF/Frest); horizontal, 10 ms.

Temporal properties of Ca2+ transients recorded with free and ribeye-bound Ca2+ indicators.

A-B. Confocal images of the isolated RBCs that were whole-cell voltage-clamped using an internal solution containing the TAMRA-RBP (magenta) and either (A) Cal520HA-RBP (cyan) or (B) Cal520LA-RBP (cyan). Note prominent spots in both TAMRA-RBP and ribeye-bound Ca2+ indicators (A) Cal520HA-RBP (cyan) or (B) Cal520LA-RBP, showing the location of the ribbon. Scale bars, 2 μm.

C-D. Spatially averaged fluorescence of (C) Cal520HA-RBP (n=19) or (D) Cal520LA-RBP (n=30) as a function of time at ribbon proximal (light cyan), and distal (dark cyan) locations. Data points show the average intensity (± SEM) in each horizontal row of five pixels for 10 ms depolarizations at distinct ribbon locations. Fluorescence intensity at the onset of the 10 ms depolarizing stimulus (arrow) was normalized to the baseline fluorescence before stimulation and averaged over all pixels (i.e., over space and time). The amplitude differences between ribbon-proximal and ribbon-distal Ca2+ signals were well-resolvable using the ribbon-bound Cal520HA-RBP indicator (Fig. 3C, light vs. dark cyan traces, paired t-test: ΔF/Frest = 3.0 ± 0.4 vs. 1.9 ± 0.3, respectively, p=0.001) and Cal520LA-RBP indicator (Fig. 3D, light vs. dark cyan traces, paired t-test: ΔF/Frest = 5.5 ± 0.9 vs. 3.3 ± 0.8, respectively, p=0.003). The current amplitudes were not significantly different between Cal520HA-RBP and Cal520LA-RBP readings (mean current amplitudes: Cal520HA-RBP = 49.8 ± 2.2 pA, Cal520LA-RBP = 49.9 ± 2.9 pA; p = 0.99).

C-D inset. Temporal profile of events between 10-50 ms were expanded for better visualization. Scale bars: vertical, 1 (ΔF/Frest, C inset) or 2 (ΔF/Frest, D inset); horizontal, 10 ms.

E-F. Average fluorescence intensity of (E) proximal and (F) distal Ca2+ signals obtained with Cal520LA-RBP (light cyan and dark cyan, respectively) and Cal520L-free (grey). The amplitudes of ribbon-proximal Ca2+ signals were higher when measured with Cal520LA-RBP than with Cal520LA-free (Fig. 3E, Cal520LA-RBP (light cyan) vs. Cal520LA-free (gray), unpaired t-test: ΔF/Frest = 5.5 ± 0.9, n=30 vs. 3.0 ± 0.4, n=19, p=0.04) but this was not the case for distal Ca2+ signals (Fig. 3F, Cal520LA-RBP (light cyan) vs. Cal520LA-free (gray), unpaired t-test: ΔF/Frest = 3.3 ± 0.8, n=30 vs. 1.9 ± 0.2, n=21, p=0.15).

E-F inset. Events between 10-50 ms were expanded for better visualization. Scale bars: vertical, 2 (ΔF/Frest, E inset) and 1 (ΔF/Frest, F inset); horizontal, 10 ms.

Ca2+ signals at synaptic ribbon at different distances from the plasma membrane.

A. Ultrastructure of a zebrafish RBC with kinetically distinct vesicle pools, as described in Supplementary Fig. 1, UFRP (green vesicles), and RRP (yellow vesicles) are primarily released via ribbon sites (R) at the cytomatrix of the active zone (arrowhead). Recycling pool (RP, orange vesicles) in the cytoplasm (C), likely to be released via non-ribbon (NR) sites. Scale bar: 500 nm.

B. Representative x-t plots show the fluorescence intensity of Cal520LA-RBP (cyan) and TAMRA-RBP (magenta) as a function of distance (vertical axis) and time (horizontal axis) at (Ba) ribbon sites and (Bb) non-ribbon (NR) locations. The darker region at the upper edge of each plot is the extracellular space and the arrowheads show the timing of depolarization. Scale bars: vertical, 1.6 μm and horizontal, 75 ms.

C. Spatially averaged Cal520LA-RBP as a function of time at ribbon proximal (light cyan), distal (dark cyan), and non-ribbon locations (blue) (n=26 (proximal ribbons), 26 (distal ribbons), and 15 (non-ribbon) from 5∼8 different RBCs, respectively).

C inset. Temporal profile of events between 0-100 ms were expanded for better visualization. Scale bars: vertical, 0.2 (ΔF/Frest); horizontal, 20 ms.

D. Ca2+ measurements along the ribbon axis using the nanophysiology approach demonstrate that the proximal Ca2+ signals can go as high as 26.4 ± 3.1 µM (light cyan, N=26) and distal as 15.6 ± 1.5 µM (dark cyan, N=26), and non-ribbon 10.4 ± 0.4 µM (NR, blue, N=15), respectively, in response to 10 ms stimuli. Error bars show standard errors. All conditions were significantly different from each other as assessed by paired t-test when comparing proximal vs. Distal and unpaired t-test when comparing non-ribbon to proximal or distal (proximal vs. distal p = 0.004; proximal vs. non-ribbon p < 0.001; distal vs. non-ribbon p = 0.002). The currents were not significantly different between conditions (Mean current: 0.2 mM EGTA Cal520L-RBP proximal and distal = 51.8 ± 3.2 pA, 0.2 mM EGTA Cal520L-RBP non-ribbon = 47.3 ± 3.0 pA; p = 0.35).

Effect of exogenous Ca2+ buffers on spatiotemporal properties of Ca2+ microdomains in RBC terminal recorded with low-affinity ribbon-bound dye.

A-D. Average temporal fluorescence intensity (normalized to ΔF/Frest) of proximal (light cyan) and distal (dark cyan) Ca2+ signals with Cal520LA-RBP as a function of time at distinct ribbon locations with pipette solution containing (A) 0.2 mM EGTA, (B) 2 mM EGTA, (C) 10 mM EGTA, or (D) 2 mM BAPTA. Proximal measurements were significantly higher than distal measurements in all conditions as assessed using paired t-tests (0.2 mM EGTA: p = 0.0027, n = 29; 2 mM EGTA: p = 0.034, n = 21; 10 mM EGTA: p < 0.001, n = 42; 2 mM BAPTA: p = 0.0073, n = 21). The currents between conditions were not significantly different from each other (mean current amplitudes in 0.2 mM EGTA: 50.6 ± 3.0 pA, 2 mM EGTA: 49.7 ± 3.1 pA, 10 mM EGTA: 47.3 ± 3.1 pA, 2 mM BAPTA: 56.1 ± 2.6 pA; 0.2 mM EGTA vs 2 mM EGTA: p = 0.84, 0.2 mM EGTA vs 10 mM EGTA: p = 0.46, 0.2 mM EGTA vs 2 mM BAPTA: p = 0.19).

Inset. The temporal profiles of events between 0-50 ms were expanded for better visualization. Scale bars: vertical, 2 (ΔF/Frest; panels A and B) and 1 (ΔF/Frest; panels C and D); horizontal, 10 ms.

E. Average temporal fluorescence intensity (normalized to ΔF/Frest) of proximal Ca2+ signals measured with Cal520LA-RBP (cyan) and Cal520LA-free (gray) as a function of time with pipette solution containing 2 mM BAPTA. The corresponding maximum values of trial-averaged ΔF/Frest was significantly higher with 2mM Cal520L bound BAPTA than 2 mM Cal520L free BAPTA, with mean ± SEM of 3.6 ± 0.9 and 1.3 ± 0.2, respectively (paired t-test, p = 0.015; Cal520L-free: n = 20; Cal520L-bound: n = 21). The currents between conditions were not significantly different from each other (mean current amplitudes in 2 mM BAPTA Cal520L-RBP: 56.1 ± 2.6 pA; 2 mM BAPTA Cal520L-Free: 53.3 ± 2.2 pA; p = 0.42).

Geometry of the computational model of intra-terminal Ca2+ dynamics.

Simulation domain is a box with dimensions (1.28 × 1.28 × 1.1) µm3, approximating the fraction of synaptic terminal volume per single ribbon. Ca2+ ions enter near the base of the ellipsoidal ribbon at four locations marked by black disks representing Ca2+ channels or their clusters. The semi-transparent gray coordinate plane Y=0 corresponds to the section used for the pseudo-color 2D [Ca2+] plots in Figure 7. Ca2+ is extruded on all surfaces of this domain, simulating combined clearance by pumps and exchangers on the plasmalemmal as well as internal endoplasmic reticulum and mitochondrial membranes.

Simulation of the effect of an endogenous immobile buffer of different concentrations on [Ca2+] dynamics in response to a 10ms pulse.

(A1-C2) In each sub-plot, the left-hand panels show the [Ca2+] time course in response to a 10 ms constant current pulse (total current of 1 pA) at five select locations marked by colored circles in the right panel. The right-hand panels show a pseudo-color plot of [Ca2+] in a 2D section of the 3D simulation volume in Fig. 6, at a fixed point in time corresponding to the end of the current pulse. Concentration values for each level curve are indicated in the color bar. Endogenous buffer is immobile, with a concentration of 200 µM in panels A1-C1 (resting buffering capacity 100 µM), vs. 1.44 mM in panels A2-C2 (resting buffering capacity 720). Exogenous buffer concentrations were 0.2 mM EGTA (panels A1, A2), 10 mM EGTA (panels B1, B2), 2 mM BAPTA (panels C1, C2).

Heterogeneity in the spatiotemporal properties of Ca2+ microdomains in RBC terminal.

A. Top panel. Cartoon of two representative RBCs (cell A and cell B), each containing differently colored ribbons. Bottom panel. Ribbon-to-ribbon variability was measured by recording local Ca2+ signals near different ribbons (yellow and purple traces) for each RBC. If an RBC had multiple readings for a single ribbon, averages were obtained for comparisons as described in Supplementary Fig. 6. Bottom panel inset: sample Ca2+ currents for the cells from which the Ca2+ signal sample traces mentioned above were obtained (purple and yellow traces). Currents were similar across the different cells. Vertical scale = 80 pA, horizontal scale = 5 ms.

B-C. Variability of Ca2+ signals in different ribbons across different cells in the presence of 10 mM EGTA, recorded using Cal520LA-RBP. Proximal Ca2+ amplitude values were significantly different between cells A and B (p = 0.022), cells A and D (p = 0.023), and cells B and D (p = 0.005) but similar between A and C (n = 4 cells, 4 fish), as assessed by unpaired t-tests. (C) In distal locations, ribbon amplitude values were significantly different between cells A and B (p = 0.031), but similar across all other cell comparisons (n = 4 cells, 4 fish), as assessed by unpaired t-tests. The currents were not significantly different across the different cells, as assessed by unpaired t-tests (mean currents in RBC a: 47.7 ± 2.9 pA, RBC b: 43.7 ± 4.0 pA, RBC c: 43.4 ± 2.0 pA, RBC d: 40.9 ± 2.4 pA; RBC a vs. RBC b: p = 0.45, RBC a vs. RBC c: p = 0.22, RBC a vs. RBC d: p = 0.14, RBC b vs. RBC c: p = 0.93, RBC b vs. RBC d: p = 0.54, RBC c vs. RBC d: p = 0.52).

D. Top panel. Illustration of a RBC containing three ribbons (numbered 1-3). Bottom panel. Ca2+ signal measurements from three distinct ribbons (black, gray, and blue traces) were compared to determine the ribbon-to-ribbon variability within each RBC, as described in Supplementary Fig. 7. Bottom panel inset: sample Ca2+ currents for the cells from which the Ca2+ signal sample traces mentioned above were obtained (black, gray, and blue traces). Currents were similar across the different cells. Vertical scale = 80 pA, horizontal scale = 5 ms.

E-F. Box plot illustrating [Ca2+] across various ribbons of an individual RBC, which is shown as RBC a. Ribbon variability within individual cells was measured with 10 mM EGTA using Cal520LA-RBP at (E) proximal and (F) distal locations. (E) Proximal Ca2+ amplitude values were significantly different among all ribbons (paired t-test, p < 0.001) (n = 5 ribbons, 1 RBCs, 1 fish). (F) Distal Ca2+ amplitudes were significantly different among all ribbon comparisons (paired t-test, p < 0.001) except for ribbons 2 and 5 (n = 5 ribbons, 1 RBCs, 1 fish). Similar analyses were conducted in two more cells and found similar observations (data not shown). The currents were similar across all ribbons since these were readings from the same cell. Given that some ribbons only have one reading, it is not possible to conduct a paired t-test to statistically compare them; however, the average current ± standard error for the cell shown was 47.7 ± 2.9 pA.

Distribution of measured distance between synaptic ribbons in three zebrafish RBCs.

A. Reconstruction of three RBC terminals closest to the ganglion cell layer resembling the shape and size of the mammalian RBC1. The ribbons are shown in magenta. Note that the total number of ribbons includes the “floating” ribbons detached from the plasma membrane. RBC II contained 7 floating ribbons. Scale bar, 5 µm. RBCs, rod bipolar cells.

B. Overview of the distance measurements. Top. 3D rendering of a single RBC terminal is shown in light green with ribbons in magenta. Black lines show an example of the different SBF-SEM layers that are cut to obtain individual images. Bottom. Four sample SBF-SEM layers are shown with two example ribbons (magenta) located near each other but in different layers. If two ribbons were on the same layer, their linear distance was taken, whereas if they were located in different layers, their linear distance and height difference were used to calculate their actual distance using the Pythagorean Theorem. Each SBF-SEM layer has a thickness of 50 nm. A 3D volume movie of the synaptic terminal of a zebrafish bipolar neuron synaptic ribbon distribution and measurement is provided in Video 1.

C. Histograms showing the ribbon distance for the three RBCs. x0 shows the mean of the distribution.

D. Histogram of the distances between each ribbon and its five nearest ribbons for all ribbons contained in the three RBCs. x0 shows the mean of the distribution. Please note that the y-axis for B and C have different sizes, given that they represent the number of occurrences of each event.

Serial block-face scanning electron microscopy analysis reveals heterogenous RBC ribbon shape, size and area of the ribbon facing the plasma membrane.

A-C. EM images of zebrafish RBC ribbon structures (A) and their respective 3D reconstructions to illustrate different shapes and sizes of synaptic ribbons (A, B, and C, magenta), plasma membrane (A and B, yellow), and the area of the ribbon facing the plasma membrane (A, B, and C, cyan). Each of the 5 rows of images illustrates one ribbon synapse from a zebrafish retinal RBC. A 3D reconstruction of the RBC synaptic terminal and ribbon from SBF-SEM stacks is provided in Video 2.

D. 3D reconstruction of the RBC terminal closest to the ganglion cell layer resembles the shape and the size of the mammalian RBC1 2. The ribbons are colored magenta.

E. Summary of the active zone size, the area of the ribbon associated with the plasma membrane measured in serial sections across the three RBCs from Fig 9A. The individual distribution of the three RBCs active zone sizes is provided in Supplementary Fig. 8. The z-size of the SBF-SEM sections is 50 nm, and each ribbon spans 2-5 consecutive sections. The solid cyan circles in the violin plots show individual synaptic ribbon measurements from three RBCs, with the average measurements shown in the solid black circle.

Heterogeneity of Ca2+ microdomains in RBC terminals.

Larger ribbons have stronger maximal Ca2+ influx and larger CAZ. A. Images show voltage-clamped RBC filled with a solution containing TAMRA-RBP to label ribbons before depolarization (left, magenta) and Cal520LA-RBP to measure the amplitude of Ca2+ influx (middle, cyan) during depolarization and superimposed of the two (right) to compare the size or ribbon vs. maximal Ca2+ influx. A time-lapse movie of the synaptic terminal of a zebrafish bipolar neuron during Ca2+ influx is provided in Video 3. B. Scatter plot of maximal Ca2+ influx (F/Frest) vs. TAMRA-RBP. Dashed lines are linear regressions, and r is Pearson’s correlation coefficient. C. Scatter plot of maximal synaptic ribbon length vs. maximal synaptic ribbon width as estimated from SBF-SEM images. Dashed lines are linear regressions, and r is Pearson’s correlation coefficient. D. Scatter plot of area of AZ vs. ribbon width (filled black circles) and of area of AZ vs. ribbon length (filled grey circles). Dashed lines are linear regressions, and r is Pearson’s correlation coefficient. AZ, Active Zone.