Synaptic Actions of Amyotrophic Lateral Sclerosis-Associated G85R-SOD1 in the Squid Giant Synapse

Presynaptic infusion of G85R-SOD1-YFP altered synaptic vesicle dynamics. A, Series of six continuous trains of HFS (each train is 50 Hz for 5 s with 5 s between trains) was applied to the synapses infused with SOD1 proteins for 15 min before HFS trains. WT-SOD1-GFP-injected synapses showed constant EPSP slope at the beginning of each train suggesting robust neurotransmission (>200, also see Extended Data Fig. 2-1 for the sixth train). However, G85R-SOD1-YFP-injected synapse started to show reductions in synaptic transmission, as evidenced by the dramatic decrease in EPSP slopes taken during the first train both at the beginning and at the steady state, suggesting severe depletion of synaptic vesicles from both the RRPs and the RPs. This inhibition was more obvious in the sixth train (Extended Data Fig. 2-1). B, EPSPs of the first train from WT-SOD1-YFP-injected synapses (black, n = 11) and from G85R-SOD1-YFP-injected synapses (red, n = 6) were integrated and averaged, followed by linear fit. The intersection with y-axis indicated the size of the RRP and the slope indicated the mobilization rate of vesicles from the RP to RRP. C, Normalized to baseline values before the injection of SOD1-YFP, RRP size and mobilization rate from the first train were plotted as individual biological replicates (n = 11 for WT-SOD1-infused synapses and n = 6 for G85R-SOD1-infused synapses) to show significant reductions in both RRP and mobilization of vesicles from the RP by G85R-SOD1-YFP, but not by WT-SOD1-YFP.

Presynaptic infusion of G85R-SOD1 inhibited synaptic vesicle (SV) availability. Representative EM images illustrate morphology of AZs (labeled with red *) and numbers of SVs in fixed synapses infused with WT-SOD1 (A, n = 3) and G85R-SOD1 (B, n = 3). G85R-SOD1-infused synapses showed vacant AZs and occasional abnormal membranous structures (indicated by blue # and purple ^⋀). C–F, Quantification of averaged vesicle number across all AZs from 102 WT- and 102 G85R-SOD1-YFP-infused synapses showed statistically significant reductions in the total vesicle number and in the electron lucid vesicle number by G85R-SOD1-YFP as compared with WT-SOD1. The clathrin-coated vesicles and the large electron lucid vesicles were comparable between WT- and G85R-SOD1-YFP-infused synapses. Nested one-way ANOVA was performed to compare WT and G85R-infused synapses (p < 0.0001) as well as across biological triplicates within each group (p > 0.05) ns: not significant. G, Averaged number of SVs per AZ were plotted along the distance from AZ (binned by 50 nm). H, Distance distribution plot of SVs in each 50-μm bin showed a global reduction of SVs from G85R-SOD1-YFP-infused synapses regardless of the distance from the AZs. Because of the drastically reduced numbers of SVs far (>675 nM) from the AZs in both WT and G85R synapses, the reduction seemed to disappear or even be reversed, however, the differences far from the AZ may not be significant due to the substantially decreased numbers of vesicles in that area for both WT and G85R synapses. I, Cumulative SV distribution plots showed similar distribution patterns of vesicles in WT (solid line) and G85R (dashed line) synapses, confirming the even inhibition of SV availability by G85R-SOD1 independent of the distance from the AZs.

iHFS prevented synaptic transmission deficits caused by G85R-SOD1-YFP. A, iHFS (5 s of 50 Hz applied every 30 min) was first applied to the synapse at time 0 before G85R injection, both presynaptic and postsynaptic membrane potentials were recorded 1 min after HFS (P1). B, If iHFS trains were applied every 30 min before and during continuous SOD1-YFP infusion into the synapse, G85R-SOD1 no longer inhibited synaptic transmission and the synapse continued to fire even after 7 h without significant changes in either presynaptic or postsynaptic membrane potentials (P15). C, EPSP slopes remained constant and comparable under basal stimulation with iHFS in WT and G85R synapses. D, EPSP slope from one single train of iHFS showed no significant difference between WT and G85R synapses except for occasional augmentation seen in G85R-SOD1-YFP-infused synapses. E, To evaluate the vesicle dynamics, EPSP slopes were measured and integrated during HFS trains applied either with 2 h WT- or G85R-SOD1-YFP infusion in the presence of iHFS every 30 min (n = 6). There were no significant differences in RRP size and mobilization rate of vesicles trafficking from RP to RRP by G85R-SOD1-YFP. F, EM showed normal presynaptic structure with normal numbers of vesicles at the AZ (indicated by red asterisk) in G85R-SOD1-YFP-infused synapses when iHFS was applied. Interestingly, in three out of five synapses, HFS restored firing in G85R-SOD1-YFP-infused synapses, where EPSPs were significantly inhibited (Extended Data Fig. 4-1).

Presynaptic infusion of G85R-SOD1-His increased Ca²⁺ levels in the presynaptic terminal. Ratiometric live Ca²⁺ imaging was performed every 30 s after electrophoretic (100 nA current) infusion of a Ca²⁺ indicator, 1 mM fura-2 in 100 mM KCl, into the presynaptic axon through the first presynaptic stimulation electrode inserted into the palm (Fig. 1A). Fura-2 injection continued for 10 min, and the synapse was allowed to equilibrate for 30 min before SOD1 injection through the second electrode to ensure fura-2 was diffused evenly throughout the whole presynaptic terminal at roughly 100 μM. Ratiometric images were taken at Ex360 nm and Ex390 nm assisted by ultra-high-speed wavelength switching λ DG-4/5 xenon arc lamp system. Ratios at various locations (A) were calculated as (fluorescent intensity^360nm)/(fluorescent intensity^390nm) to provide a measure of intracellular Ca²⁺ concentrations. The blue arrow indicates the injection site for SOD1 proteins. 1: palm, 2–6: PreG, 4: infusion site, 7: PreS, an adjacent presynaptic axon branch which is smaller in size. B, G85R-SOD1-His induced Ca²⁺ increases while WT-SOD1-His had no effect on Ca²⁺ concentration. Raw Ca²⁺ concentrations were derived from the equation [Ca²⁺] = and averaged across 12 WT- and 19 G85R-SOD1-injected presynaptic terminals (PreG) respectively. Kd′, Rmin, and Rmax were calculated from standards as described in Materials and Methods. Although different synapses varied in their baseline Ca²⁺ concentrations, G85R consistently increased [Ca²⁺] at around 60 min after protein injection whereas WT had no effect. C, D, Baseline normalized Ca²⁺ ratio defined as (R₁₂₀ – R₀)/R₀ (R₁₂₀: 60 min after SOD1 injection, R₀: before SOD1 injection) were plotted at four sites for WT (n = 12) and G85R (n = 19). G85R-SOD1 caused increases in [Ca²⁺] at all sites including the palm where Ca²⁺ channels are sparse or absent. The increase in [Ca²⁺] correlated roughly with G85R-SOD1 concentration, with the highest levels at the injection site and lowest in PreS, a smaller presynaptic axonal branch infused with fura-2 at a comparable concentration but with a lower SOD1 concentration due to slower diffusion from the injection site in PreG. E, F, iHFS applied at 30 min after SOD1 injection seemed to restore Ca²⁺ homeostasis in G85R-SOD1-injected synapses (n = 19), and this equilibrium lasted at least 2 h after iHFS, similar to that in WT (n = 12), this suggested the possibility that redistribution of Ca²⁺ is induced by iHFS. As expected, Ca²⁺ levels increased during HFS in synapses infused with either WT or G85R-SOD1 proteins (Extended Data Fig. 5-1).