Characterization of live imaging reporters to simultaneously monitor presynaptic and excitatory postsynaptic compartments.

A & B, lentiviral-delivered HaloTag-Syb2 in primary hippocampal neurons labeled with cell-permeant JF646 HaloTag ligand and subsequently immunostained for Syn1/2 (A; presynaptic) or Homer1 (B; excitatory postsynaptic). C, quantification of Pearson’s correlation coefficient from data in Figure 1A & B. D & E, representative dendrites from primary hippocampal neurons transduced with lentivirus encoding mClover3-Homer1c immunostained for SHANK2 (D; excitatory postsynaptic) or Gephyrin (E; inhibitory postsynaptic). F, quantification of Pearson’s correlation coefficient from experiments in Figure 1D & E. G, diagram of experimental strategy to label Schaffer collateral synapses ex vivo. H, representative viral labeling in the CA1 region from experiments outlined in Figure 1G. I, high magnification images of the indicated sub-regions of CA1 neurons expressing mClover3-Homer1 and receiving HaloTag-Syb2-positive CA3 Schaffer collateral inputs. J, representative live primary hippocampal neuron cultures expressing mClover3-Homer1c, HaloTag-Syb2 and mTagBFP2 before (left) and after (right) a 15-hr imaging period. K, visualization of nascent excitatory synapses in primary cultures. Representative HaloTag-Syb2-labeled growth cone forming a co-cluster with a mClover3-Homer1c punctum along a mTagBFP2-filled dendrite. Numerical data are means ± SEM from 4 independent biological replicates. Statistical significance was assessed with a two-tailed t-test (***, p<0.001). See Movie S1 for representative live imaging.

Visualization of distinct populations of excitatory synapses.

A, live hippocampal cultures expressing mClover3-Homer1c, HaloTag-Syb2, and mTagBFP2 before (left) and after (right) a 15-hr imaging period. Boxes indicate zoomed-in regions shown in panels C and D. B, example tracked pre- and postsynaptic HaloTag-Syb2 and mClover3-Homer1c puncta and synapse tracks. Cyan crosshairs mark current object position, while cyan trails indicate previous 30 frames (5 mins per frame). Tracks were further classified as indicated in the flow chart. C, Top, representative puncta tracking of co-clustered HaloTag-Syb2 (magenta) and mClover3-Homer1c (green) with synapse tracks (cyan) over 15 hrs. Bottom, synapse categories for the same tracks shown. Trails indicate previous 30 frames (2.5 hrs). D, as in C, with nascent synapses along an extending dendrite. E, histogram of synapse track durations with color-coded track categories. F, Left, data averaged across time and replicate cultures showing total relative abundance of each track type. Right, percentage of each synapse track category at each frame during the imaging session. Each bar represents synapses active at each sampling interval (5 min). N=5 independent cultures. G, speed of track motion for Transient, Stable-like, and Stable synapses. One-way RM ANOVA (P=0.0031) with Tukey’s post hoc. N=5 independent cultures. Gray dots are individual track speed; colored points are average track speed per culture. (*, P < 0.05). Post hoc test comparisons were made between all groups, but only comparisons where adj P < 0.05 are shown. See Movie S2 and S3 for representative tracking examples and drift correction. See Figures S1 and S2 for additional characterization of tracking approaches.

Spatiotemporal dynamics of excitatory synapse populations during maturation.

A, representative dendrites at DIV8-9 with HaloTag-Syb2 and mClover3-Homer1 with overlayed tracks (left) and track categorizations (right). Track trails indicate previous 30 frames (2.5 hrs). B, breakdown of synapse category averages per frame at DIV8-9 for synapse tracks, mClover3-Homer1c tracks, and HaloTag-Syb2 tracks. Synapse tracks are reproduced from Figure 2F for comparison. C, duration of synapse tracks at DIV8-9, mClover3-Homer1c tracks, and HaloTag-Syb2 tracks. One-way RM ANOVA (P=0.001) with Tukey’s post hoc. N=5 independent cultures. Gray dots are individual track duration; colored points are average track duration per culture. D, as in A, but with DIV11-14 cultures. E, as in B, but with DIV11-14 cultures. F, as in C, but with DIV11-14 cultures. One-way RM ANOVA (P=0.03) with Tukey’s post hoc. N=4 independent cultures. G, puncta density of mClover3-Homer1c tracks, HaloTag-Syb2, and paired puncta (synapses) across the imaging period for DIV8-9 cultures. Data are Mean ± SD. H, as in G, but with DIV11-14 cultures. I, average puncta density of DIV8-9 and DIV11-14 cultures. Two-way ANOVA (DIV, P=0.114; Puncta type, P=0.0004; interaction, P=0.74) with Tukey post hoc. J, puncta speed of paired and unpaired tracks for HaloTag-Syb2 paired with mClover3-Homer1c. Two-way RM ANOVA (DIV, P=0.31; Pairing, P=0.0116; interaction, P=0.73) with Fisher’s LSD post hoc. N=4-5 independent cultures. Gray dots are individual tracks speed and colored points are average track speed per culture. K, track duration of paired and unpaired tracks for HaloTag-Syb2 paired with mClover3-Homer1c. Two-way RM ANOVA (DIV, P=0.78; Pairing, P=0.029; interaction, P=0.82) with Fisher’s LSD post hoc. N=4-5 independent cultures. Gray dots are individual tracks duration and colored points are average track duration per culture. L, as in J, but for mClover3-Homer1c paired with Syb2. Two-way RM ANOVA (DIV, P=0.41; Pairing, P=0.0462; interaction, P=0.94) with Fisher’s LSD post hoc. N=4-5 independent cultures. M, as in K, but for mClover3-Homer1c paired with Syb2. Two-way RM ANOVA (DIV, P=0.63; Pairing, P=0.0006; interaction, P=0.49) with Fisher’s LSD post-hoc. N=4-5 independent cultures. Post hoc test comparisons were made between all groups, but only comparisons where adj P < 0.05 are shown for brevity. (*, P<0.05, ** P<0.01).

Labeling approaches for inhibitory synapses in vitro and ex vivo.

A, co-localization of the tdTomato-Gephyrin lentiviral reporter with GABARα1 (inhibitory postsynaptic; left) or Homer1 (excitatory postsynaptic; right). B, diagram of experimental strategy to virally label OLM interneuron-CA1 synapses. C, representative CA1 neurons expressing tdTomato-Gephyrin and receiving HaloTag-Syb2-positive presynaptic terminals from OLM interneurons. Right, high magnification image of the stratum lacunosum-moleculare region.

Live imaging inhibitory synapse spatiotemporal dynamics.

A, representative dendrites at DIV8-9 with HaloTag-Syb2 and tdTomato-Gephyrin with overlayed tracks (left) and track categorizations (right). Track trails indicate previous 30 frames (2.5 hrs). B, breakdown of synapse category averages per frame across all cultures DIV8-9 for synapse tracks, tdTomato-Gephyrin tracks, and HaloTag-Syb2 tracks. C, duration of synapse tracks at DIV8-9, tdTomato-Gephyrin tracks, and HaloTag-Syb2 tracks. One-way RM ANOVA (P=0.0379) with Tukey’s post hoc. N=3 independent cultures. Gray dots are individual tracks duration and colored points are average track duration per culture. D, as in A, but with DIV11-14 cultures. E, as in B, but with DIV11-14 cultures. F, as in C, but with DIV11-14 cultures. One-way RM ANOVA (P=0.09) with Tukey’s post hoc test. G, puncta density of tdTomato-Gephyrin tracks, HaloTag-Syb2, and paired puncta (synapses) across the imaging period for DIV8-9 cultures. Data are Mean ± SD. H, as in G, but with DIV11-14 cultures. I, average puncta density of DIV8-9 and DIV11-14 cultures with tdTomato-Gephyrin and HaloTag-Syb2. Two-way ANOVA (DIV, P=0.39; Puncta type, P<0.0001; interaction, P=0.32) with Tukey post hoc. J, puncta speed of paired and unpaired tracks for HaloTag-Syb2 paired with Gephyrin. Two-way RM ANOVA (DIV, P=0.095; Puncta type, P=0.088; interactions P=0.21). K, track duration of paired and unpaired tracks for HaloTag-Syb2 paired with Gephyrin. Two-way RM ANOVA (DIV, P=0.63; Puncta type, P=0.15; interaction, P=0.67). L, as in J, but for tdTomato-Gephyrin paired with Syb2. Two-way RM ANOVA (DIV, P=0.61, Puncta type, P=0.0592, interaction, P=0.86). M, as in K, but for tdTomato-Gephyrin paired with Syb2. Two-way RM ANOVA (DIV, P=0.67; Puncta type, P=0.0084; interaction, P=0.42) with Fisher’s LSD post hoc. Post hoc test comparisons were made between all groups, but only comparisons where adj P < 0.05 are shown. (*, P<0.05, ** P<0.01, *** P<0.001). See Movie S4 for representative live imaging data and Figure S3 for additional characterization of tracking approaches.

Characterization of TKIT CRISPR/Cas9-based reporters to label endogenous pre- or postsynaptic compartments.

A-E, characterization of CRISPR/Cas9 tagging strategy for endogenous presynaptic Bassoon. A, diagram of TKIT CRISPR/Cas9 approach to tag endogenous Bassoon via AAV-mediated delivery of TKIT CRISPR components. B, example neurons transduced with AAVs encoding an HA-Bassoon DNA donor, sgRNAs, without (left) or with (right) Cas9. Neurons were co-stained for HA, endogenous Bassoon, and the somatodendritic marker MAP2. C-E, co-localization of HA-tagged Bassoon with endogenous Bassoon (C), another presynaptic marker Syn1/2 (D), and the excitatory postsynaptic marker Homer1 (E). F-J, TKIT tagging of endogenous postsynaptic excitatory Homer1c. F, diagram of the CRISPR/Cas9-mediated tagging strategy for endogenous Homer1c. G, example primary hippocampal neurons transduced with AAVs encoding the HA-Homer1c DNA donor and sgRNAs without (left) or with (right) Cas9. Neurons were co-stained for HA tag, endogenous Homer1, and MAP2. H-J, co-localization of HA-tagged Homer1c with endogenous Homer1 (H), another excitatory postsynaptic marker SHANK2 (I), and the inhibitory postsynaptic marker Gephyrin (J). K-O, validation of TKIT-based tagging of endogenous inhibitory postsynaptic Gephyrin. K, diagram of endogenous Gephyrin tagging via TKIT CRISPR/Cas9. L, example primary hippocampal neurons transduced with AAVs encoding the Gephyrin tagging donor and sgRNAs without (left) or with (right) Cas9. Neurons were co-stained for GFP, endogenous Gephyrin, and MAP2. M-O, immunostaining for GFP-tagged Gephyrin together with endogenous Gephyrin (M), another inhibitory postsynaptic component GABARα1 (N), and the excitatory postsynaptic marker Homer1 (O). P-R, TKIT-mediated Gephyrin tagging in the CA1 region. P, diagram of experimental strategy to label endogenous Gephyrin in CA1 neurons. Q, GFP tagging of endogenous Gephyrin in the CA1 region. mTagBFP2 was used as an injection site marker. R, representative CA1 neuron dendrite labeled with GFP-tagged endogenous Gephyrin. See Figures S4-S6 for additional characterization of TKIT CRISPR/Cas9 tagging approaches.

Long-term imaging of endogenous inhibitory postsynaptic Gephyrin.

A, example primary hippocampal neuron harboring tdTomato-tagged endogenous Gephyrin before and after a 15-hr imaging period. B, example dendrite with tdTomato-labeled endogenous Gephyrin and lentiviral expressed HaloTag-Syb2 time-lapse imaged over a 15-hr period, with overlayed tracks (left) and track categorizations (right). C, breakdown of synapse category averages per frame across all cultures DIV8-9 for synapse tracks, tdTomato-Gephyrin tracks, and HaloTag-Syb2 tracks. D, duration of synapse tracks, tdTomato-Gephyrin tracks, and HaloTag-Syb2 tracks. One-way RM ANOVA (P=0.1213). N=5 independent cultures. Gray dots are individual tracks duration and colored points are average track duration per culture. E, puncta density of endogenous tdTomato-Gephyrin tracks, HaloTag-Syb2, and paired puncta (synapses) across the imaging period. Data are Mean ± SD. F, average puncta density of cultures with endogenously labeled tdTomato-Gephyrin and lentiviral HaloTag-Syb2. One-way ANOVA (P=0.01) with Tukey’s post hoc. G, puncta speed (left) and track duration (right) of paired and unpaired tracks for HaloTag-Syb2 paired with endogenous Gephyrin. Paired t-tests, N=5 independent cultures. H, as in G, with endogenous tdTomato-Gephyrin paired with HaloTag-Syb2. Paired t-tests, N=5 independent cultures. Post hoc test comparisons were made between all groups, but only comparisons where adj P < 0.05 are shown. (**, P<0.01). See Movie S5 for representative live imaging data and Figure S7 for additional characterization of tracking approach.

Monitoring excitatory and inhibitory synapse ratios and functional maturation during synaptogenesis.

A, representative primary hippocampal neurons expressing lentiviral-delivered mClover3-Homer1c, tdTomato-Gephyrin, HaloTag-Syb2, and mTagBFP2. B, puncta density of Homer1c:Syb2 and Gephyrin:Syb2 pairs at indicated timepoints. Two-way ANOVA (DIV, 0.66; puncta type, 0.0027; interaction, 0.74) with Fisher’s LSD post hoc. Mean ± SEM, N=3 cultures. C, excitatory/inhibitory ratios calculated from same cultures. Welch’s t-test (P=0.62). Mean ± SEM, N=3 cultures. D, representative traces for mEPSC/mIPSC ratio measurements recorded from DIV6, 10 and 16 primary hippocampal neurons. E & F, cell capacitance (E) and membrane resistance (F) from recordings at indicated timepoints. G, cumulative probability plot of mEPSC amplitude measurements. H, cumulative probability plot of mEPSC inter-event intervals. I & J, similar to G and H except for mIPSC measurements. K, average mEPSC/mIPSC amplitudes at indicated timepoints. L, average ratio of mEPSC/mIPSC amplitudes. One-way ANOVA with Tukey post hoc (ns – not significant). M & N, similar to K and L except for mEPSC/mIPSC frequency measurements. One-way ANOVA with Tukey post hoc (ns – not significant).

CRISPR KI sequences, AAV and Lenti titering primers, and primers for estimating the tagging efficiency of the TKIT method.

Primers sequences for analysis of CRISPR/CAS9 off-target modifications.

Tracking neuron cultures with heterogenous motion.

A, representative images of cultures classified as having moving neurites (top) and still neurites (bottom). Gray and white boxes in moving neurites depict a neurite that moves throughout the imaging session. B, duration of puncta in movies binned into moving or still neurites for mClover3-Homer1c, HaloTag-Syb2, or synapse pairs. Two-way ANOVA (Moving-Still, P=0.0074; Puncta-type, P=0.0002; interaction P=0.9) with Tukey’s post hoc. C, as in B but with puncta speed. Two-way ANOVA (Moving-Still, P=0.0001; Puncta-type, P=0.28; interaction P=0.57) with Tukey’s post hoc. D, as in B but with net displacement (start to end distance). Two-way ANOVA (Moving-Still, P=<0.0001; Puncta-type, P=0.044; interaction P=0.79) with Tukey’s post hoc. E, as in B, but with path length (total distance traveled). Two-way ANOVA (Moving-Still, P=0.17; Puncta-type, P=<0.0001; interaction P=0.52) with Tukey’s post hoc.

Additional parameters for excitatory synapse dynamics across development, related to Figures 2 and 3.

A, example HaloTag-Syb2 (top) and mClover3-Homer1c (bottom) tracks with track categories (as in Figure 2). B & C, histogram of track duration for mClover3-Homer1c (B) and HaloTag-Syb2 (C). D, histogram of inter-puncta centroid distances for mClover3-Homer1c and HaloTag-Syb2 puncta. E, histogram of puncta pairing duration for mClover3-Homer1c (green) and HaloTag-Syb2 (magenta). Tracks were classified as unpaired (pairing duration < 5% total time) or paired (pairing duration > 75% total time). F, pairing analysis of data shown in Figure 3. Counts of paired puncta / the total puncta of that type are shown. For all puncta, this represents all puncta pairs relative to total number of puncta of any type. Two-way ANOVA (DIV, P=0.99; puncta type, P=0.0035; interaction, P=0.98) with Tukey’s post hoc. Mean ± SEM, N=4-5 independent cultures. G, total puncta count of all puncta types for mClover3-Homer1c, HaloTag-Syb2, and puncta pairs. Two-way ANOVA (DIV, P=0.55; puncta type, P=0.0019; interaction, P=0.98) with Tukey’s post hoc. Mean ± SEM, N=4-5 independent cultures.

Additional parameters for inhibitory synapse dynamics across development, related to Figure 5.

A, example HaloTag-Syb2 (top) and tdTomato-Gephyrin (bottom) tracks with track categories (as in Figure 2). B & C, histogram of track duration for tdTomato-Gephyrin (B) and HaloTag-Syb2 (C). D, histogram of inter-puncta centroid distances for tdTomato-Gephyrin and HaloTag-Syb2 puncta. E, histogram of puncta pairing duration for tdTomato-Gephyrin (green) and HaloTag-Syb2 (magenta). Tracks were classified as unpaired (pairing duration < 5% total time) or paired (pairing duration > 75% total time). F, pairing analysis of data shown in Figure 5. Counts of paired puncta / the total puncta of that type are shown. For all puncta, this represents all puncta pairs relative to total number of puncta of any type. Two-way ANOVA (DIV, P=0.50; puncta type, P<0.0001; interaction, P=0.98) with Tukey’s post hoc. Mean ± SEM, N=4-5 independent cultures. G, Total puncta count of all puncta types for tdTomato-Gephyrin, HaloTag-Syb2, and puncta pairs. Two-way ANOVA (DIV, P=0.51; puncta type, P=0.0008; interaction, P=0.94) with Tukey’s post hoc. Mean ± SEM, N=4-5 independent cultures.

CRISPR/Cas9 tagging of endogenous GluA2 using the TKIT approach.

A & B, example of the efficacy of the TKIT tagging approach with the SEP-GluA2 tagging constructs from Fang et al., 202122. A, low magnification overviews illustrating the prevalence of SEP-tagged GluA2 neurons across a neuronal culture. B, high magnification representative dendritic stretches of endogenously tagged GluA2. Cells were co-labeled for GFP together with excitatory presynaptic vGLUT1 and the somatodendritic marker MAP2.

Additional characterization of CRISPR/Cas9 TKIT-mediated approach to label endogenous pre- and postsynaptic markers.

A-C, quantification of Pearson’s correlation coefficient from results in Figure 6. A, Pearson’s correlation coefficient measurements of HA-tagged endogenous presynaptic Bassoon co-immunostained for indicated synaptic markers. B, similar to A, except for the HA-tagged Homer1c TKIT. C, similar to A, except for GFP-tagged Gephyrin. D, example TKIT tagging of both Gephyrin and Homer1c. E-H, RT-qPCR analysis of CRISPR/Cas9 tagging efficiency. E, qPCR probes detecting the HA-tagged Bassoon transcript are only amplified when neurons are co-transduced with AAVs encoding the DNA donor/sgRNAs, and Cas9. Copy numbers were calculated via standard curve qPCR using a plasmid containing the tagged sequence. F, similar to E, except for HA-tagged Homer1c transcript. G, similar to E, except for GFP-tagged Gephyrin transcript. H, efficiency estimates based on standard curve qPCR with primers detecting tagged Homer1c or Gephyrin transcript specifically, or primers detecting both tagged and non-tagged transcripts. Numerical data are means ± SEM from 3-4 independent culture replicates indicated as open circles. Statistical significance was assessed via two-tailed t-test or one-way ANOVA with post hoc Tukey tests (*, p<0.05; ***, p<0.001).

Analysis of TKIT off-target effects.

A, chromatogram depicting HA insertion into the 5’ of endogenous Homer1c. B, Homer1c sgRNA1 sequence compared to the top two predicted off-target sites. Chromatograms depict sequencing of off-target sites. C, similar to B, except for Homer1c sgRNA2. D, chromatogram depicting GFP insertion into the 5’ of endogenous Gephyrin. E, Gephyrin sgRNA1 sequence compared to the top two predicted off-target sites. Chromatograms depict sequencing of off-target sites. F, similar to E, except for Gephyrin sgRNA2. G, chromatogram depicting HA insertion into the 5’ of endogenous Bassoon. H, Bassoon sgRNA1 sequence compared to the top two predicted off-target sites. Chromatograms depict sequencing of off-target sites. I, similar to H, except for Bassoon sgRNA2. Data depicts representative chromatograms from three independent experiments.

Additional parameters for tracking endogenous Gephyrin, related to Figure 7.

A, histogram of inter-puncta centroid distances for endogenous tdTomato-Gephyrin and HaloTag-Syb2 puncta. B, histogram of puncta pairing duration for endogenous tdTomato-Gephyrin (green) and HaloTag-Syb2 (magenta). Tracks were classified as unpaired (pairing duration < 5% total time) or paired (pairing duration > 75% total time). C, pairing analysis of data shown in Figure 7. Counts of the number of paired puncta / the total puncta of that type are shown. For all puncta, this represents all puncta pairs relative to total number of puncta of any type. One-way ANOVA (P=0.013) with Tukey’s post hoc. Mean ± SEM, N=5 independent cultures. D, total puncta counts of all puncta for endogenous tdTomato-Gephyrin, HaloTag-Syb2, and puncta pairs. Kruskall-Wallis test (P=0.07). Mean ± SEM, N=5 independent cultures.