Dynamic assembly of ribbon synapses and circuit maintenance in a vertebrate sensory system

Abstract

Ribbon synapses transmit information in sensory systems, but their development is not well understood. To test the hypothesis that ribbon assembly stabilizes nascent synapses, we performed simultaneous time-lapse imaging of fluorescently-tagged ribbons in retinal cone bipolar cells (BCs) and postsynaptic densities (PSD95-FP) of retinal ganglion cells (RGCs). Ribbons and PSD95-FP clusters were more stable when these components colocalized at synapses. However, synapse density on ON-alpha RGCs was unchanged in mice lacking ribbons (ribeye knockout). Wildtype BCs make both ribbon-containing and ribbon-free synapses with these GCs even at maturity. Ribbon assembly and cone BC-RGC synapse maintenance are thus regulated independently. Despite the absence of synaptic ribbons, RGCs continued to respond robustly to light stimuli, although quantitative examination of the responses revealed reduced frequency and contrast sensitivity.

Fig. 1

Number of axonal ribbons of distinct ON BC types across development. a Schematics showing the key neurons of this study and their ribbon synapses. Ph photoreceptors, BC bipolar cell, IPL inner plexiform layer, GC ganglion cell, PSD postsynaptic density. b Confocal images showing individual axon terminals and ribbons of three different types of ON BCs across development. Ribbons were labelled by anti-CtBP2 and those outside the axon terminals were excluded digitally (see Methods for details). c Axonal arbor volume of the three BC types. d Developmental increase in the total number of ribbons per axon, and e the fraction of ribbons (CtBP2 positive puncta) normalized to the average total ribbon number per axon at P21, for each BC type studied. Data were collected from 6–10 cells from 4–8 retinas for each age-group. Data are presented as mean ± sem

Fig. 2

RIBEYE-tagRFP-T mice enable visualization of ribbons in live retina. a A schematic drawing of the RIBEYE-RIBEYE:tagRFP-T transgene. b Volumetric view of a two-photon reconstruction of RIBEYE-tagRFP-T live retina. OPL outer plexiform layer. IPL inner plexiform layer. c Vertical section of CtBP2 labeling in control (ctrl) retina. d Vertical section of CtBP2 labeling, RIBEYE-tagRFP-T (tagRFP-T), and overlay of both channels. e Colocalization of CtBP2 labeling and RIBEYE-tagRFP-T signal on a type 6 BC axon terminal. f Comparison of number of CtBP2 puncta on T6 BC axon terminal in control (ctrl; Grm6-YFP; eifht cells from eight retinas) and RIBEYE-tagRFP-T (RE-RFP) (six cells from four retinas) mice. g An example of light-evoked EPSC on ON-alpha RGCs in control and RIBEYE-tagRFP-T mice. Quantification of the amplitude of the transient (h) and sustained (i) components of the light-evoked EPSC. (f, h, and i: n number of cells, from three retinas for each group). P21 mice were used in all experiments. Data are presented as mean ± sem

Fig. 3

Dynamics of ribbons on axon terminals of BCs in developing and mature retinas. a Live imaging of ribbons within individual T6 BC axon terminals in RIBEYE-tagRFP × Grm6-YFP retinas. All RIBEYE-tagRFP-T puncta that appeared or disappeared during the imaging session were identified. Yellow arrow: punctum appeared; Green arrow: punctum disappeared. b Quantification of the number of RIBEYE-tagRFP-T puncta that were formed or were eliminated at P10 (five cells from three retinas) or at P40 (four cells from two retinas). Filled circle and yellow bar: puncta formed; Open circle and green bar: puncta eliminated. Data are presented as mean ± sem

Fig. 4

RIBEYE puncta are more stable when apposed to PSD95-FP. a Monitoring ribbons and postsynaptic specialization (PSD95) simultaneously in a P10 retina from RIBEYE-tagRFP-T/PSD95-mVenus double transgenic mice by confocal microscopy. Shown here are the maximum intensity projections of image stacks that were acquired every 5 min up to 20 min. An example of a relatively stable RIBEYE-tagRFP-T punctum apposed to PSD95-mVenus punctum (arrow and cyan box) and an unstable RIBEYE-tagRFP-T punctum not apposed to PSD95-mVenus punctum (arrowhead and magenta box). b Survival rate of ribbon puncta apposed (PSD95+) or not apposed (PSD95−) to PSD95-mVenus puncta (three cells from three retinas). Data are presented as mean ± sem

Fig. 5

PSD95 puncta are more stable when apposed to ribbons. a RGCs were biolistically labeled with CFP together with PSD95-YFP. Live-cell imaging: every 2 h, up to 10 h. b From top to bottom: An example of PSD95 formation with (w/) or without (w/o) RIBEYE-tagRFP-T and PSD95 elimination with or without RIBEYE-tagRFP-T puncta. c The rate of PSD95 puncta formation with and without RIBEYE-tagRFP-T puncta at P10 and P40. d Survival rate of PSD95 puncta with and without RIBEYE-tagRFP-T puncta at P10 and at P40. Five live imaging experiments are performed at both P10 (five cells from four retinas) and P40 (four cells from two retinas). Data are presented as mean ± sem

Fig. 6

PSD95 density is unperturbed in ribeye-ko mice. Visualization of PSD95 puncta on dendrites of ON-alpha RGCs in littermate control (a) and ribeye-ko (b) mice. Loss of ribbon apposed to PSD95 puncta was confirmed by CtBP2 immunostaining (A and B insets). Wholemount view of a ON-alpha RGC biolistically labeled in Grm6td-Tomato ctrl (c) or ribeye-ko (d) retina at P21. Magnified views of a T6 BC synapsing onto the RGC in ctrl (c, 1) or KO (d, 2) retina are provided in the lower panels. Arrowheads indicate synapses between the T6 BC and the RGC. e Quantification of PSD95 density of ON-alpha RGCs in P11 and P21 retinas of littermate control (ctrl) and ribeye-ko (KO) mice. f Number of synapses between a pair of T6 BC and ON-alpha RGC. g Dynamics of PSD95 puncta on P11 ON-alpha RGCs during a 2 h recording period in ctrl and ribeye-ko mice. Number of cells is indicated in each bar graph. For each experiment, 2–4 retinas are used for each group. Data are presented as mean ± sem

Fig. 7

Light response properties and synaptic transmission in ribeye-ko mice. a (Left) Action potentials of ON-alpha RGCs evoked by a 500 ms light flash. (Right) Average firing rates during the stimulation duration. b (top) Whole-cell voltage-clamp recording of spontaneous excitatory postsynaptic currents (sEPSCs) from ctrl and KO. (bottom) Mean amplitude and frequency of sEPSCs. c Chirp stimulus (top) and example excitatory synaptic inputs to ON-alpha RGCs from ribeye knockout (red) and littermate control (black) retinas. d Increment and decrement responses from first section of chirp stimulus. The increment/decrement ratio determines how large the tonic excitatory input (i.e., the current suppressed by the decrement) is compared to the maximal excitatory input (i.e., the current in response to the increment). e Responses to frequency ramp. The amplitude of the response to each cycle of the stimulus is plotted against temporal frequency. Frequency tuning was measured from the frequency (F_1/2) at which the response fell to half of its maximal value. f Responses to contrast ramp. The amplitude of the response to each cycle of the stimulus is plotted against contrast. Contrast sensitivity was measured from the contrast (C_1/2) at which responses reached half maximum. g Summary of increment/decrement ratios (left), frequency tuning (middle) and contrast sensitivity (right) across cells. Low rod light level is 2 R*/rod/s, medium rod is 50 R*/rod/s and cone is 2000 R*/cone/s. *** denotes p < 0.001 (unpaired t-test). Data are presented as mean ± sem

Fig. 8

BCs form ribbon-containing and ribbon-free contacts with wild-type RGCs. a (Top panels) Relationship of PSD95-YFP puncta on ON-alpha RGCs (blue) with ribbons (CtBP2 labeling) at P9 and P32. Contact sites with PSD95-YFP and CtBP2 apposed are indicated as ‘colocalized’ (magenta) whereas PSD95-YFP sites without CtBP2 are marked as ‘non-colocalized’ (yellow). (Bottom panels) Magnified views of the stretch of dendrite within the rectangles. Red arrows indicate PSDS95 puncta not apposed to CtBP2 puncta. b Developmental increase of PSD95 density on the dendrites of ON-alpha RGCs. c Percentage of CtBP2 puncta apposed to PSD95 puncta on the dendrites of ON-alpha RGCs across development. N = 4–6 cells from 3–6 retinas. d Maximum intensity projections of confocal image stacks of en face views of T6 BC axon terminals and the dendrites and PSD95 puncta of an ON-alpha RGC. Two example synaptic contacts are marked with the asterisks. Magnified views for both contacts are shown on the right. (1) ribbon synaptic contact; (2) ribbonless synaptic contact. e Electron micrograph showing a typical ribbon (arrowhead) synapse between a T6 BC axon and a dyad of ON-alpha RGC and amacrine cell (AC) dendrites. Both GC and amacrine cell membranes are line with PSDs (brackets). f Two consecutive electron micrographs (S129 and S130) showing a non-ribbon contact (arrows) for the same T6 BC and ON-alpha GC pair shown in e. The GC membrane is slightly concave and clearly lined by a postsynaptic density (brackets). Data are presented as mean ± sem