Synaptotagmin-7 enhances phasic dopamine release

Dopamine released from substantia nigra pars compacta (SNc) neurons modulates movement, motivation, and reward. In addition to their tonic firing pattern, dopamine neurons also fire high-frequency bursts that cause superlinear increases in dopamine release. To study this poorly understood form of short-term plasticity, we used the fluorescent dopamine sensor dLight1.3b to examine the role of the calcium-binding protein synaptotagmin-7 (SYT7). We report that SYT7 mediates a hidden component of facilitation, which was unmasked by lowering initial release probability, or by low-frequency stimulation of nerve terminals. In Syt7 KO neurons, there was profound synaptic depression that significantly reduced release during stimulations that mimic in vivo firing patterns of SNc neurons. D2-mediated inhibitory postsynaptic currents in the SNc revealed a similar role for SYT7 in somatodendritic release. Our results indicate that SYT7 drives short-term facilitation of release from dopamine neurons, which likely underlies frequency-dependence of dopamine signaling in vivo.

is expressed throughout the brain, but is enriched in SNc neurons (Saunders et al.,47 2018). SYT7 partially mediates release from dendrites within the SNc, and is present in 48 axon terminals of cultured dopamine neurons (Mendez et al., 2011), but whether SYT7 49 regulates release from dopamine axons remains unknown. Measurement of dopamine 50 release at sub-second intervals is required to test this hypothesis (Covey et al., 2013). 51 Here, we used high-speed photometry of a fluorescent dopamine sensor and whole-cell 52 electrophysiology in Syt7 knockout mice to probe the role of SYT7 in dopamine release. 53 Release evoked by high-frequency stimulation was reduced in Syt7 KO animals in 54 nerve terminals and in somatodendritic compartments. SYT7 enhanced dopamine 55 release during transitions from tonic to phasic firing. Our data establish an important 56 role for SYT7 in short-term plasticity and phasic dopamine release. 57 58 Results 59 We immunolabeled coronal brain sections with antibodies against SYT7 and tyrosine 60 hydroxylase (TH) ( Figure 1A). In high-resolution images obtained using 3D structured To probe the impact of SYT7 in striatal dopamine release, we monitored dopamine in 79 acute striatal slices from wildtype (WT) and Syt7 -/-(Syt7 KO) mice (Chakrabarti et al.,80 2003) using the genetically-encoded fluorescent dopamine sensor dLight1.3b (Patriarchi 81 et al., 2018). Adeno-associated viral vector was injected unilaterally into the dorsal 82 striatum to drive dLight expression in striatal neurons (Figure 2A-B). Sagittal brain slices 83 (270 µm) were subsequently prepared for imaging and electrophysiology. dLight 84 fluorescence was excited with dim light (mean intensity = 0.96 mW/mm 2 ) and recorded 85 at 10 kHz using a high-speed photodiode detector ( Figure 2C). Dopamine axons were 86 activated by local extracellular stimulation, in artificial cerebrospinal fluid (ACSF) 87 containing 2 mM Ca 2+ . To isolate action potential-driven dopamine release, recordings 88 were performed in the presence of antagonists for nicotinic, D2 and GABA receptors. 89 Electrical stimulation elicited rapid, low noise transients ( Figure 2D). The use of dim 90 excitation light resulted in minimal photobleaching, permitting prolonged imaging over 91 the same region of the striatum ( Figure 2E). Bath application of the D1 antagonist SKF-92 83566 (10 μM) reduced the responses by 95 ± 3% (n = 4), confirming that fluorescence 93 was driven by binding of dopamine to dLight ( Figure 2F). Lowering external Ca 2+ to 0. to decay to 50% of the peak were altered in the absence of SYT7 ( Figure 3A-B), 121 suggesting that SYT7 does not prolong the presence of dopamine in the extracellular 122 space by promoting asynchronous release. experiments was determined by unpaired Student's t-test after normal distribution and 131 homoscedasticity were confirmed. For comparisons of more than 2 samples, two-way 132 ANOVA tests were performed and Šidák corrected. Significance is shown as *: p < 0.05, 133 **: p < 0.01, ***: p < 0.001, ****: p < 0.0001.

135
In addition to asynchronous release, SYT7 mediates short-term synaptic facilitation. 136 Such facilitation increases synchronous release for several hundred milliseconds after 137 an action potential, as assessed by the paired-pulse ratio (PPR) (Jackman & Regehr, 138 2017). Because dLight transients lasted for more than 100 ms, for short inter-stimulus 139 intervals (ISI) the amplitude of the second response was calculated by subtracting the 140 response driven by a single stimulus ( Figure 3C). In WT animals the PPR was ~1 at 141 short intervals (0.87 ± 0.07 for ISI = 20 ms), followed by marked depression, and slow 142 recovery from depression ( Figure 3D). In contrast, Syt7 KOs displayed the hallmarks of 143 high release probability (p) synapses, depressing at short intervals (PPR = 0.34 ± 0.05 144 for ISI = 20 ms) and recovering slowly. This suggests that SYT7 serves to offset vesicle 145 depletion, resulting in sustained striatal dopamine release at short inter-stimulus 146 intervals.

148
Synapses that show short-term depression can still possess a hidden component of was 4.6-fold larger than the pre-depressed IPSC, a greater ratio than that seen without 220 a series of pre-pulses. However, in Syt7 KO slices the ratio of pre-pulse to phasic IPSC 221 was significantly lower than in WT slices, and did not differ from that seen without pre-222 depression in KO slices (WT: 4.6 ± 0.2, KO: 2.6 ± 0.2, p < 0.0001) ( Figure 6E). Thus, 223 Syt7 also mediates facilitation of dopamine release at somatodendritic sites. Here, we report that SYT7 drives short-term facilitation of dopamine release. SYT7 241 significantly increased release during phasic bursts, despite the fact that both axonal 242 and dendritic synapses exhibit a high initial release probability and overt synaptic 243 depression. All electrophysiological recordings and analysis were done blind to the genotype of the 292 animal, but for other experiments blinding and randomization were not performed.

294
Immunohistochemistry 295 Mice were anaesthetized with ketamine/xylazine (100/10 mg/kg) and transcardially 296 perfused with 4% paraformaldehyde (PFA) in PBS. Brains were removed and post-fixed 297 overnight at 4 °C. 50 μm slices were prepared using a Leica VT1000S vibratome, and 298 permeabilized for 1 h in vehicle (PBS, 10% normal goat serum, 0.3% Triton X-100), 299 then incubated overnight at 4 °C in vehicle with primary antibodies (mouse anti-Syt7, 300 GABA-A, and GABA-B conductances respectively. Internal solution used for recording 354 contained (in mM): 100 K-methanesulphonate, 20 NaCl, 1.5 MgCl2, 10 HEPES (K), 10 355 BAPTA, 2 ATP, 0.2 GTP, and 10 phosphocreatine. Giga-ohm seals were obtained using 356 glass electrodes (1.1-1.4 MOhm resistance) and allowed to stabilize for 1-2 min before 357 break-in. No overt difference in cell firing behavior was observed between genotypes in 358 cell attached configuration (data not shown). Series resistance, capacitance, and 359 membrane resistance were measured shortly after break-in using the average of 20 5 360 mV test pulses, and were monitored during and upon completion of all recordings. Cells 361 were voltage-clamped at -58 mV using an AxoPatch1B amplifier and current was 362 recorded at 1k Hz. In addition to spontaneous pacemaker activity in cell-attached 363 configuration, DA cell identity was verified by the presence of an Ih current in response 364 to a 50 mV hyperpolarizing step. D2-IPSCs were evoked using a platinum bipolar 365 stimulator connected to a stimulus isolator (Warner Precision instruments) placed in the 366 slice caudal to recorded cell bodies. All recordings are the average of 3-4 repetitions 367 conducted once per minute that are baselined to the 500 ms immediately preceding 368 stimulation before averaging. All recording and analysis of D2-IPSC experiments was 369 conducted using AxoGraph software (Berkeley CA).

371
Acknowledgements 372 We thank Stefanie Kaech Petrie, Hannah Bronstein, and the Advanced Imaging Core at 373 the Jungers Center for microscopy assistance, and Gary Westbrook and Dennis 374 Weingarten for helpful comments on the manuscript. This work was supported by a 375 Whitehall Foundation Grant (SLJ), the National Institutes of Health (RO1 DA004523 to 376 JTW), the National Institute on Drug Abuse (T32DA007262 to JJL), and the National 377 Institute of Neurological Disorders and Stroke (P30 NS061800 to the OHSU Imaging 378 Center).

380
Competing Interests 381 The authors have no competing interests to declare. 382 383