A genetically encoded sensor measures temporal oxytocin release from different neuronal compartments

Abstract

Oxytocin (OT), a peptide hormone and neuromodulator, is involved in diverse physiological and pathophysiological processes in the central nervous system and the periphery. However, the regulation and functional sequences of spatial OT release in the brain remain poorly understood. We describe a genetically encoded G-protein-coupled receptor activation-based (GRAB) OT sensor called GRAB_OT1.0. In contrast to previous methods, GRAB_OT1.0 enables imaging of OT release ex vivo and in vivo with suitable sensitivity, specificity and spatiotemporal resolution. Using this sensor, we visualize stimulation-induced OT release from specific neuronal compartments in mouse brain slices and discover that N-type calcium channels predominantly mediate axonal OT release, whereas L-type calcium channels mediate somatodendritic OT release. We identify differences in the fusion machinery of OT release for axon terminals versus somata and dendrites. Finally, we measure OT dynamics in various brain regions in mice during male courtship behavior. Thus, GRAB_OT1.0 provides insights into the role of compartmental OT release in physiological and behavioral functions.

Extended Data Fig. 1 |. Characterization of GRAB_OT sensors in HEK293T cells.

a. Representative expression and response images of OT1.0 to 100 nM OT in HEK293T cells. Scale bar, 20 μm. b. Example traces (left) and summary data (right) of OT1.0-expressing HEK293T cells pre-incubated with saline, 10 μM YM022, or 10 μM Atosiban in response to OT. Saline: n = 30 cells from 3 coverslips [30/3], YM022: n = 39/4, Atosiban: n = 44/3. Two-tailed Student’s $t$-tests, p = 0.84 (between Saline and YM022); p = 9.8 × 10⁻¹⁸ (between Saline and Atosiban). c. Excitation (Ex) and emission (Em) spectra of the OT1.0 sensor with or without 100 nM OT. d. Representative images of OT1.0 and OTmut expressed in HEK293T cells in saline and in the presence of 100 nM OT. Also shown is RFP-CAAX expression, showing localization at the plasma membrane. The images at the right show the change in OT1.0 and OTmut fluorescence in response to OT application. White rectangle with enhanced contrast showing OTmut expressing HEK293T cells in saline. Scale bars, 20 μm. e. Summary of the peak change in OT1.0 and OTmut fluorescence measured in HEK293T cells in response to 100 nM OT. OT1.0: n = 45 cells from 3 coverslips; OTmut: n = 31 cells from 4 coverslips. Two-tailed Student’s $t$-tests were performed (p = 5.5 × 10⁻³¹). f. Dose–response curves for OT1.0 and OTmut expressed in HEK293T cells in response to the indicated concentrations of OT and AVP, with the corresponding EC₅₀ values shown. The data were normalized to the maximal response measured in OT group. The dosage curves of OT1.0 to OT/AVP were averaged from 9 individual trials, with 3–4 wells per trial. g. Dose-response curves for OT1.0 expressed in HEK293T cells in response to the indicated concentrations of OT and its orthologous peptides, with amino acid sequence alignment shown. n = 4 coverslips for each group. h. Summary of the peak change in OT1.0 fluorescence measured in HEK293T cells in response to the indicated compounds applied at 1 μM (CRF, NTS, NPY, and VIP) or 10 μM (Glu, GABA, Gly, DA, NE, and 5-HT), normalized to the peak response measured in OT; n = 4 wells per group. CRF, corticotropin-releasing factor; NTS, neurotensin; NPY, neuropeptide Y; VIP, vasoactive intestinal peptide; Glu, glutamate; GABA, $\gamma$-aminobutyric acid; Gly, glycine; DA, dopamine; NE, norepinephrine; and 5-HT, 5-hydroxytryptamine (serotonin). One-way ANOVA test was performed for all groups (F (10, 3.88) =132.3, p = 2.0 × 10⁻⁴); Dunnett’s T3 multiple comparisons tests were performed (p = 4.6 × 10⁻³, 3.8 × 10⁻³, 4.4 × 10⁻³, 4.0 × 10⁻³, 4.2 × 10⁻³, 4.2 × 10⁻³, 4.6 × 10⁻³, 4.8 × 10⁻³, 4.5 × 10⁻³, and 4.5 × 10⁻³ (between OT and CRF, NTS, SST, NPY, CCK, VIP, Glu, GABA, ACh, DA)). i. Representative traces of the OT1.0 signal evoked by OT puffing at indicated concentrations. j. Summary of the OT1.0 signal time constant at indicated OT concentrations (n = 19 cells for 30 nM OT, n = 11 cells for 100 nM OT and n = 9 cells for 1000 nM OT). k. Representative traces and summary $\text{ΔF}/{\text{F}}_0$ of OT1.0 when bath application of OT at saturated concentration using line-scanning mode. n = 6 trials. l. The calibration curve of OT1.0 dose-dependent fluorescence response in line-scanning mode, which is used to estimate the local OT concentration reaching the cells during local puffing experiments. m. The association rate constant of the OT1.0 sensor for OT. Local OT concentrations were estimated from i (n = 19 cells for 30 nM OT, n = 11 cells for 100 nM OT and n = 9 cells for 1000 nM OT). **p < 0.01, and ***p < 0.001. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 2 |. Negligible downstream signaling coupling of OT1.0 sensor in HEK293T cells.

a, b. Representative expression images in BFP channel, pseudocolor images (top) and $\text{ΔF}/{\text{F}}_0$ traces (bottom) showing the Ca²⁺ response to the indicated concentrations of OT or ATP in HEK293T cells expressing bOTR-BFP (a) or OT1.0-BFP (b). c. Summary of peak Ca²⁺ $\text{ΔF}/{\text{F}}_0$ for bOTR or OT1.0 expressed HEK293T cells corresponding to (a and b) at indicated OT concentrations, with the corresponding EC₅₀ value shown. The data were normalized to the peak response measured in 100 μM ATP. n = 3 coverslips for each group. Two-tailed Student’s $t$-tests were performed between bOTR and OT1.0 (p = 0.56, 0.29, 0.018, and 0.013 for 0.01, 0.1, 1, and 10 nM OT). d. $\text{β}$-arrestin coupling was measured using the Tango assay in cells expressing the bovine OTR (bOTR), OT1.0, bOTR and OT1.0, or no receptor (Control). n = 3 wells each. Two-tailed Student’s $t$-tests, p = 0.12 (between bOTR and bOTR+OT1.0); p = 6.8 × 10⁻⁷ (between bOTR and OT1.0); p = 2.9 × 10⁻³ (between OT1.0 and Control). e. Representative images (left) and summary (right) of the fluorescence change in OT1.0-expressing neurons in response to a 2-hour continuous OT application. n = 5 cultures with >30 cells each. Row matched one-way ANOVA, with the Geisser-Greenhouse correction, F (1.509,6.037) =0.16, p = 0.79. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s., not significant. All scale bars represent 20 μm. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 3 |. Chemogenetic activation of oxytocinergic neurons induces OT release.

a. Schematic diagram depicting the chemogenetic activation experiments. A mixture of AAVs (EF1$\text{α}$-dio-hM3Dq-mCherry and hSyn-OT1.0) was injected into the PVN of OT-Cre mice. As a control, hSyn-OT1.0 was injected into the PVN of OT-Cre x Ai14 mice (no-hM3Dq). The PVN and third ventricle (3 V) are indicated. b. Left: representative 2-photon microscopy merged images of OT1.0 (green channel) and the RFP channel (red, mCherry expression for OT-hM3Dq and tdTomato for no-hM3Dq). Right: responses of the OT1.0 sensor measured in ACSF (baseline), 60 nM DCZ, and 100 nM OT. Scale bars, 100 μm. c, d. Example OT1.0 traces (c) and peak change (d) in OT1.0 fluorescence; where indicated, DCZ or OT were applied to the slices. n = 7 slices from 2 mice for OT-hM3Dq and n = 5 slices from 2 mice for no-hM3Dq. Two-tailed Student’s $t$-tests were performed (p = 8.9 × 10⁻⁶ (left) and 0.41 (right)). ***p < 0.001 and n.s., not significant. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 4 |. Immunostaining of OT neurites in the VTA and PVN.

a. Representative images of OT-positive axons in the VTA of WT and OT-KO mice. Green, OT antibody; blue, DAPI. Scale bars: left, 1 mm; right, 100 μm. b-d. Schematic drawings depicting the experimental strategy (b), representative images (c), and quantification data (d) showing the colocalization of PSD95-EGFP with OT-positive neurites in the PVN or VTA of coronal slices. Red, OT (indicated by OT antibody in the VTA or tdTomato in the PVN); green, PSD95 (indicated by GFP antibody); blue, DAPI. n = 3 slices from 1 mouse. Two-tailed Student’s t-tests were performed (p = 0.014). Scale bars: left, 200 μm; right, 20 μm. *p < 0.05. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 5 |. Electrical stimulation evoked OT and Glutamate release in the PVN, S1 and dStr.

a. Schematic illustration depicting the experimental setup for panels (b-g). b and d. Representative fluorescence images (left) and $\text{ΔF}/{\text{F}}_0$ pseudocolor images (right) showing the expression and electrical stimulation-induced response of OT1.0 or iGluSnFR in the PVN, S1, and dStr. c and e. Representative traces of electrical stimulation evoked and 100 nM OT or 100 μM Glu perfusion induced OT1.0 or iGluSnFR signals. f and g. Summary of the change in OT1.0 or iGluSnFR fluorescence in response to 100 pulses or ligand application (n = 5 slices from 4 mice [5/4], 6/3, and 6/3 mice for OT1.0 in the PVN, S1, and dStr, respectively; n = 12 slices from 4 mice [12/4], 6/2, and 7/2 mice for iGluSnFR in the PVN, S1, and dStr, respectively.) Two-tailed Student’s t-tests were performed (f: for electrical stimulation, p = 2.9 × 10⁻³, 0.11, and 0.034 between basal and peak $\text{ΔF}/{\text{F}}_0$ for PVN, S1, and dStr, respectively; p = 3.1 × 10⁻³ between PVN and S1; p = 9.5 × 10⁻³ between PVN and dStr; for OT perfusion, p = 0.037 between PVN and S1; p = 0.27 between PVN and dStr; g: for electrical stimulation, p = 0.50 between PVN and S1; p = 0.63 between PVN and dStr; p = 0.77 between S1 and dStr; for Glu perfusion, p = 0.92 between PVN and S1; p = 0.12 between PVN and dStr; p = 0.26 between S1 and dStr). The data of OT1.0 and iGluSnFR in the PVN are reused from Fig. 5. *p < 0.05, **p < 0.01, and n.s., not significant. All scale bars represent 100 μm. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 6 |. Application of OT with different concentrations in OT1.0 expressing slices.

a-c. Example pseudocolor images (top) and fluorescence traces (bottom) of OT1.0-expressing slices containing the PVN (a) or VTA (b, c) before and after application of 100 nM OT (a, b) or AVP (c). d, f. Representative fluorescence images showing the expression of OT1.0 in the VTA (d) and PVN (f). e, g. Example traces of OT1.0 signals in response to 1, 10, 100, and 1000 nM OT application in the VTA (e) and PVN (g). h, i. Summary of peak and post-application (wash) $\text{ΔF}/{\text{F}}_0$ in response to 1, 10, 100, and 1000 nM OT in the VTA (h) and PVN (i). n = 3 slices from 3 mice [3/3] and 3/2 mice for OT1.0 in the VTA and PVN, respectively. Two-tailed Student’s t-tests were performed (h: p = 0.057, 0.024, 0.18, and 0.31 (between Peak and Wash) for 1, 10, 100, and 1000 nM OT, respectively; i: p = 0.011, 8.5 × 10⁻³, 0.021, and 9.2 × 10⁻³ (between Peak and Wash) for 1, 10, 100, and 1000 nM OT, respectively). *p < 0.05, **p < 0.01, and n.s., not significant. All scale bars represent 100 μm. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 7 |. Dissecting the Ca²⁺ sources underlying somatodendritic OT release.

a. Top left: representative image of OT1.0 expressed in the PVN (left). Also shown are example pseudocolor images (top row), corresponding traces (bottom row), and summary of the peak OT1.0 response (right) to 100 electrical stimuli delivered at 20 Hz in ACSF, nimodipine (Nim; 10 μM), Cd²⁺ (200 μM), or 10 nM OT. n = 5 slices from 3 mice for ACSF, Nim, and Cd²⁺; n = 3 slices from 2 mice for OT. Paired two-tailed Student’s $t$-tests were performed (p = 9.7 × 10⁻⁴ (left) and 0.031 (right)). b. Representative pseudocolor images (top row) and corresponding traces (bottom row) of OT1.0 expressed in the PVN in response to 100 electrical stimuli delivered at 20 Hz in ACSF, $\text{ω}$-Agx-IVA (0.3 μM), $\text{ω}$-CTx (1 μM), or SNX-482 (100 nM) to block P/Q-, N-, and R-type VGCCs, respectively. c. Representative fluorescence image of iGluSnFR (top left) and schematic drawing depicting the experimental strategy (bottom left), related to Fig. 3e, f. Example pseudocolor images (top) and traces (bottom) of the change in iGluSnFR fluorescence in response to 100 electrical pulses delivered at 20 Hz in ACSF, $\text{ω}$-Agx-IVA (0.3 μM), $\text{ω}$-CTx (1 μM), nimodipine (Nim; 10 μM), or Cd²⁺ (200 μM) to block P/Q-, N-, L-type or all VGCCs, respectively (slices were sequentially perfused with the indicated blockers). *p < 0.05 and **p < 0.01. All scale bars represent 100 μm. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 8 |. Somatodendritic OT release is insensitive to cell type-specific expression of TeNT.

a-j. Schematic drawings depicting the experimental strategy (a, f), representative images showing the expression (b, g) and peak response of OT1.0 sensor (c, h), average traces (d, i), and summary (e, j) of OT1.0 in response to 100 pulses stimulation delivered at 20 Hz or OT perfusion applied under the indicated conditions. n = 7 slices from 2 mice [7/2] and 5/2 mice for PVN and VTA, respectively. Two-tailed Student’s t-tests were performed (e: p = 0.83 and 0.47 between Ctrl and DIO-TeNT for electrical stimulation and OT perfusion; j: p = 0.038 and 0.78 between Ctrl and DIO-TeNT for electrical stimulation and OT perfusion). The data of Ctrl groups in the PVN and VTA are reused from Fig. 5. Scale bars, 100 μm. *p < 0.05 and n.s., not significant. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 9 |. OT1.0 can detect intraventricularly injected OT in the PVN of rats.

a. Schematic diagram depicting the experimental strategy for in vivo recording of OT1.0 in rats. An AAV expressing hSyn-OT1.0 was injected into the PVN of WT Sprague Dawley female rats; optical fibers were placed in the above PVN 2 weeks later, 10 mM OT (1 μL) was injected into the lateral ventricle during recording, and 470-nm light was used to excite the OT1.0 sensor together with isosbestic control signal (405 nm). b. Exemplary histological verification of the optic fiber placement and the OT1.0 expression in the periPVN area. OT1.0 was stained with anti-GFP antibody for visualization. OF, optic fiber. Scale bar, 400 μm. c. Average trace and quantification of OT1.0 signal. The OT1.0 and isosbestic signals were sampled at 1 Hz. n = 4 rats. Paired two-tailed Student’s $t$-tests were performed (p = 1.3 × 10⁻³). **p < 0.01. Summary data are presented as the mean±s.e.m.

Extended Data Fig. 10 |. Optogenetic activation of neurons induces somatodendritic and axonal OT release in vivo in freely moving rats.

a. Schematic illustrations depicting the optogenetic activation experiments with both the OT sensor and the optogenetic stimulation by ChrimsonR in the PVN. b. Representative traces recorded in the rat PVN of changes in normalized fluorescent emission $\text{ΔF}/{\text{F}}_0$ (525 nm) during excitation at isosbestic control (405 nm, in purple) or sensor wavelength (465 nm, in green) before, during and after stimulation of ChrimsonR with pulses (at 593 nm, in orange) of 10 ms at a frequency of 1, 7 or 30 Hz for a total duration of 5 s). c. Summary of the peak changes in OT1.0 fluorescence emission (at 525 nm) during excitation at the sensor wavelength (465 nm, in green) or isosbestic control (405 nm, in purple) in rats PVN (at 1, 7, or 30 Hz photostimulation). n = 4 rats per group. Paired two-tailed Student’s $t$-tests were performed between 465 nm and 405 nm (p = 9.1 × 10⁻³, 6.1 × 10⁻⁴, and 6.6 × 10⁻⁴ for 1, 7, and 30 Hz, respectively). d. Summary of the rise time (‘on’) and decay time (‘off’) constants (${\text{T}}_{1/2}$) of the OT1.0 response to photostimulation. n = 4 rats per group. e. Schematic illustrations depicting the optogenetic activation experiments with OT sensor expressed in the SON and the optogenetic stimulator ChrimsonR expressed in the PVN. f. Representative traces recorded in the rat SON of changes in normalized fluorescent emission $\text{ΔF}/{\text{F}}_0$ (525 nm) during excitation at isosbestic control (405 nm, in purple) or sensor wavelength (465 nm, in green) before, during and after stimulation of ChrimsonR (at 593 nm, in orange) with pulses of 10 ms at a frequency of 1, 7, or 30 Hz for a total duration of 5 s). g. Summary of the peak change in OT1.0 fluorescence emission (at 525 nm) during excitation at the sensor wavelength (465 nm, in green) or isosbestic control (405 nm, in purple) in rats SON (at 1, 7, or 30 Hz photostimulation). n = 4 rats per group. Paired two-tailed Student’s $t$-tests were performed between 465 nm and 405 nm (p = 0.043, 8.8 × 10⁻³, and 0.034 for 1, 7, and 30 Hz, respectively). h. Summary of the rise time (‘on’) and decay time (‘off’) constants (${\text{T}}_{1/2}$) of the OT1.0 response to photostimulation. n = 4 rats per group. *p < 0.05, **p < 0.01 and ***p < 0.001. Summary data are presented as the mean±s.e.m.

Fig. 1 |. Development of GRAB_OT sensors and activity-dependent OT release in brain slices.

a, Schematic diagram depicting the principle behind the GRAB_OT sensor. b, Selection of a candidate sensor for further development in HEK293T cells by screening OTRs cloned from the indicated species. c, Linker length optimization of GRAB_OT sensors based on OT0.5. d, Left, representative images of OT1.0 expression and the response to 100 nM OT in cultured neurons. Right, representative soma (indicated by the white arrowhead) and neurite images. e, Example traces (left) and summary data (right) of OT1.0-expressing neurons pretreated with saline or 10 μM atosiban in response to OT. Soma, n = 19 regions of interest (ROIs) from 3 dishes (19/3); neurite, n = 90/3; atosiban, n = 36/3. Two-tailed Student’s $t$ tests: P = 0.50 (between soma and neurite), P = 2.3 × 10⁻¹⁸ (between soma and atosiban), P = 7.6 × 10⁻⁵¹ (between neurite and atosiban). f, Normalized dose–response curve of OT1.0-expressing neurons in response to the indicated concentrations of OT; n = 4 dishes. g, Summary of normalized $\Delta F/F_0$ measured in OT1.0-expressing neurons in response to the indicated compounds. CRF, corticotropin-releasing factor; NTS, neurotensin; NPY, neuropeptide Y; VIP, vasoactive intestinal peptide; Glu, glutamate; Gly, glycine; DA, dopamine; NE, norepinephrine; 5-HT, 5-hydroxytryptamine. n = 3 wells per group. One-way ANOVA: F(11, 4.446) = 69, P = 2.0 × 10⁻⁴; Dunnett’s T3 multiple-comparisons tests, P = 0.030, 0.018, 0.022, 0.020, 0.016, 0.029, 0.030, 0.029, 0.029, 0.029 and 0.029 (from left to right, between OT and other groups, respectively). h, Summary of the kinetics of the OT1.0 response: illustration of the local puff system (left), representative response trace (middle) and group data for ${\tau }_{\text{on}}$ (right); n = 18 cells from 3 cultures. i, Schematic illustration depicting the experimental design in the VTA for j–l. j, Example fluorescence and pseudocolor images of OT1.0- or OTmut-expressing brain slices at baseline and in response to the indicated stimuli in the presence of artificial cerebrospinal fluid (ACSF) or 5 μM L368. The dashed red circles indicate the ROI used to calculate the response, and the approximate location of the stimulating electrode is indicated. k, Representative traces (left) and summary data (right) for the change in OT1.0 or OTmut fluorescence in response to the indicated stimuli in ACSF or L368. n = 7 slices from 5 mice (7/5), 3/2 and 3/1 for ACSF, L368 and OTmut, respectively. Two-tailed Student’s $t$ tests: P = 1.2 × 10⁻⁵ (between ACSF and L368), P = 3.2 × 10⁻⁵ (between ACSF and OTmut). l, Summary of the rise and decay time constants (${\tau }_{\text{on}}$ and ${\tau }_{\text{off}}$) of the electrically evoked fluorescence increase in OT1.0-expressing slices in response to the indicated stimuli. n = 7, 7 and 9 slices for the 20-, 50- and 100-pulse groups, respectively. m, Schematic illustration depicting the experimental design in the PVN for n–p. n–p, Similar to j–l. In o, n = 9 slices from 5 mice (9/5), 9/5 and 3/1 for ACSF, L368 and OTmut, respectively; in p, n = 2, 3 and 17 slices for the 20-, 50- and 100-pulse groups, respectively. Paired two-tailed Student’s $t$ test, P = 3.7 × 10⁻⁵ (between ACSF and L368); two-tailed Student’s $t$ test, P = 3.8 × 10⁻⁶ (between ACSF and OTmut). *P < 0.05, ***P < 0.001; NS, not significant. Scale bars: 10 μm (d), 25 μm (h) and 100 μm (j,n). Summary data are presented as the mean ± s.e.m.

Fig. 2 |. Probing the spatial and temporal dynamics of OT release in the axonal and somatodendritic compartments.

a, Example time-lapse pseudocolor images of OT1.0 expressed in the VTA and PVN and of iGluSnFR expressed in the PVN in acute brain slices. The dashed lines were used to analyze spatial and temporal dynamics. Similar results were observed for more than six slices for each group. Scale bars, 100 μm. b, Spatial profile of the evoked change in fluorescence shown in a. Each heatmap shows the average of three trials conducted in one slice. c, Temporal dynamics of the data shown in b measured at the indicated distance from the release center. The data were processed with 5× binning. d, Spatial dynamics of the data shown in b measured 1, 2, 5 and 10 s after the start of stimulation. Each curve was fitted with a Gaussian function. The data were processed with 5× binning and normalized to the peak response of each curve. e, Representative diffusion coefficients ($D$) were determined by plotting the square of FWHM (full width at half maximum) against time on the basis of the data shown in d. The diffusion coefficients were obtained by fitting a linear function using the FWHM² calculated from the first 10 s ((i) and (ii)) or the first 0.2 s (iii). In b–e, data shown are for OT1.0 in the VTA (i) and PVN (ii) and for iGluSnFR in the PVN (iii). f, Summary of ${\tau }_{\text{on}}$ and ${\tau }_{\text{off}}$ for the evoked OT1.0 response in the VTA and PVN and for the iGluSnFR response in the PVN at a distance of 10 μm from the release center (n = 6, 10 and 4 slices, respectively). Two-tailed Student’s $t$ tests: ${\tau }_{\text{on}}$: P = 0.95 (between OT in the VTA and OT in the PVN), 5.4 × 10⁻⁴ (between OT in the PVN and glutamate in the PVN); ${\tau }_{\text{off}}$: P = 0.046 (between OT in the VTA and OT in the PVN), 4.6 × 10⁻⁴ (between OT in the PVN and glutamate in the PVN). g, Summary of the FWHM of activity-dependent OT and glutamate signals measured in d at the indicated time points; n = 7 slices from 4 mice (7/4), 11/5 and 4/1 for OT in the VTA, OT in the PVN and glutamate in the PVN, respectively. Two-tailed Student’s $t$ tests were performed for 1, 2, 5 and 10 s (P = 0.82, 0.70, 0.42 and 0.70 (between OT in the VTA and OT in the PVN); P = 0.93, 0.031, 4.5 × 10⁻⁴ and 1.2 × 10⁻⁵ (between OT in the PVN and glutamate in the PVN)). h, Summary of the diffusion coefficients measured in e; n = 7 slices from 4 mice (7/4), 11/5 and 4/1 for OT in the VTA, OT in the PVN and glutamate in the PVN, respectively. Two-tailed Student’s $t$ tests: P = 0.93 (between OT in the VTA and OT in the PVN), 0.015 (between OT in the PVN and glutamate in the PVN), 0.014 (between OT in the VTA and glutamate in the PVN). The data for OT in the VTA and OT in the PVN were reanalyzed from raw images used in Fig. 1k,o. *P < 0.05, ***P < 0.001; NS, not significant. Summary data are presented as the mean ± s.e.m.

Fig. 3 |. N-type and L-type VGCCs support axonal and somatodendritic OT release, respectively.

a, Pseudocolor two-photon images of evoked OT release in the PVN (top) and average $\Delta F/F_0$ traces of OT1.0 in the PVN, OT1.0 in the VTA and iGluSnFR in the VTA (bottom) in the indicated concentrations of extracellular Ca²⁺. b, Summary of the normalized peak $\Delta F/F_0$ measured in a at the indicated Ca²⁺ concentrations (top) and summary of EC₅₀ for OT1.0 in the PVN and VTA and iGluSnFR in the VTA (bottom); the data in the upper panel were normalized to the peak response measured in 4 mM Ca²⁺. n = 9 slices from 4 mice (9/4), 5/3 and 5/2 for OT in the PVN, OT in the VTA and glutamate in the VTA, respectively. Two-tailed Student’s $t$ tests: P = 0.55 (between OT in the PVN and OT in the VTA), P = 0.21 (between OT in the VTA and glutamate in the VTA), P = 2.6 × 10⁻³ (between OT in the PVN and glutamate in the VTA). c, Representative fluorescence image of OT1.0 (top left) and schematic drawing depicting the experimental strategy (bottom left). Shown are example pseudocolor images (top) and traces (bottom) of the evoked change in OT1.0 and iGluSnFR fluorescence in ACSF, the L-type VGCC blocker nimodipine (Nim; 10 μM), the P/Q-type VGCC blocker $\text{ω}$-Agx-IVA (0.3 μM), the N-type VGCC blocker $\text{ω}$-CTx-GVIA (1 μM) or 200 μM Cd²⁺ to block all VGCCs (slices were sequentially perfused with the indicated blockers). d, Normalized peak responses for the data measured in c; n = 4 slices from 2 mice for OT1.0 and n = 4 slices from 2 mice for iGluSnFR. Paired two-tailed Student’s $t$ tests: P = 7.6 × 10⁻³, 9.8 × 10⁻³, 2.5 × 10⁻³ and 3.2 × 10⁻³ (left); P = 0.075, 3.6 × 10⁻³, 1.1 × 10⁻³ and 0.20 (right) (between ACSF, nimodipine, $\text{ω}$-Agx, $\text{ω}$-CTx and Cd²⁺). e,f, Same as c and d, respectively, for OT1.0 and iGluSnFR expressed in the PVN; n = 7 slices from 3 mice for OT1.0 and n = 5 slices from 3 mice for iGluSnFR. Paired two-tailed Student’s $t$ tests: P = 0.41, 3.8 × 10⁻³, 1.8 × 10⁻³ and 4.4 × 10⁻⁴ (left); P = 0.041, 0.065, 0.18 and 0.08 (right) (between ACSF, $\text{ω}$-Agx, $\text{ω}$-CTx, nimodipine and Cd²⁺). *P < 0.05, **P < 0.01, ***P < 0.001; NS, not significant. All scale bars represent 100 μm. Summary data are presented as the mean ± s.e.m.

Fig. 4 |. SNARE proteins have distinct roles in axonal and somatodendritic OT release.

a–e, Schematic drawings depicting the experimental strategy (i), representative images showing the expression and peak response of the OT1.0 sensor or iGluSnFR (ii), average traces (iii) and summary (iv) of OT1.0 and/or iGluSnFR in response to a 100-pulse stimulation delivered at 20 Hz or OT or glutamate perfusion applied under the indicated conditions. For the summary data, in control (Ctrl) groups, n = 9 slices from 6 mice (9/6) (a,d), 5/4 (b,e) and 12/4 (c,e) for OT1.0 in the VTA and PVN and iGluSnFR in the PVN; in BoNT/A groups, n = 4/2 (a), 6/2 (b) and 9/3 (c) for OT1.0 in the VTA and PVN and iGluSnFR in the PVN; in TeNT groups, n = 8/4 (d), 11/4 (e) and 6/2 (e) for OT1.0 in the VTA and PVN and iGluSnFR in the PVN; and in the TeNT + VAMP2vw group, n = 6/3 (d). Two-tailed Student’s $t$ tests: P = 2.3 × 10⁻³ (left) and 0.24 (right) (a); P = 5.3 × 10⁻³ (left) and 0.39 (right) (b); P = 2.5 × 10⁻⁵ (left) and 3.6 × 10⁻⁴ (right) (c); P = 3.8 × 10⁻⁶, 9.1 × 10⁻⁴, 0.18 (left) and 0.89, 0.76, 0.84 (right) (between Ctrl and TeNT, TeNT and TeNT + VAMP2vw, TeNT + VAMP2vw and Ctrl) (d); P = 0.46 (left), 3.4 × 10⁻⁴ (middle) and 0.18 (right) (e). **P < 0.01, ***P < 0.001; NS, not significant. All scale bars represent 100 μm. Summary data are presented as the mean ± s.e.m.

Fig. 5 |. OT1.0 can be used to monitor OT release in vivo during male mating.

a, Schematic illustrations depicting the in vivo intracerebroventricular (ICV) infusion experiments (left) and a fluorescence image of OT1.0 expressed in the BNST (right). Scale bar, 1 mm. b,c, Averaged traces (b) and summary (c) of the change in OT1.0 or OTmut fluorescence in response to the indicated concentrations of OT in mice. n = 4 mice per group. Paired two-tailed Student’s $t$ tests: OT1.0: P = 0.20, 0.75, 2.8 × 10⁻³, 2.4 × 10⁻³ and 2.1 × 10⁻³; OTmut: P = 0.91, 0.26, 0.12, 0.80 and 0.95 (between saline and 1 μM, 10 μM, 100 μM, 1 mM and 10 mM OT). d, Averaged traces (left) and summary (right) of $\Delta F/F_0$ in OT1.0- expressing mice in response to the application of 500 nl of 1 mM OT with or without (reused from b) 500 nl of 50 mM atosiban. n = 4 mice per group. Two-tailed Student’s $t$ test, P = 2.3 × 10⁻³. e, Averaged traces (left) and summary (right) of $\Delta F/F_0$ in OT1.0-expressing mice in response to the application of 500 nl of 100 μM OT (reused from b) or 500 nl of 100 μM AVP. n = 4 mice per group. Two-tailed Student’s $t$ test, P = 2.7 × 10⁻³. f, Schematic illustrations depicting the optogenetic activation experiments. g,h, Averaged traces of five trials (g) and peak response summary (h) of OT1.0 and OTmut in the PFC following photostimulation of the PVN in OT-Cre mice that received an intraperitoneal injection of saline or L368. In h, n = 6 mice per group. Paired two-tailed Student’s $t$ tests were performed between saline and L368: P = 0.39, 0.014, 8.6 × 10⁻⁴ and 1.1 × 10⁻⁶ for 0.25, 1, 5 and 10 s, respectively. Two-tailed Student’s $t$ tests were performed between saline and OTmut: P = 0.41, 8.5 × 10⁻³, 5.8 × 10⁻⁴ and 1.4 × 10⁻⁶ for 0.25, 1, 5 and 10 s, respectively. i, Summary of the rise and decay time constants ($T_{1/2}$) of the optogenetically evoked OT1.0 response; n = 6 mice per group. j, Schematic diagram depicting the experimental strategy for in vivo recording of OT1.0 in mice. k, Cartoon illustration of the three principal behaviors during male mating. l–o, OT1.0 or OTmut $\Delta F/F_0$ measured in the VTA (top), PVN (middle) or mPFC (bottom) during male mating. Shown are representative traces of a single recording (l), average time-locked traces from five individual behaviors (m), a summary of the peak responses (n) and the time constants (o) measured during the indicated mating behaviors; n = 6 mice per group. Two-tailed Student’s $t$ tests were performed between OT1.0 and OTmut for each region: P = 0.031, 0.45 and 0.024; 0.51, 4.2 × 10⁻⁴ and 0.56; and 0.017, 0.67 and 7.6 × 10⁻⁶ for VTA, PVN and PFC, respectively. Paired two-tailed Student’s $t$ tests were performed between sniffing and intromission, between sniffing and ejaculation and between intromission and ejaculation for each region: P = 0.23, 0.021 and 0.023; 9.3 × 10⁻⁴, 0.99 and 6.1 × 10⁻⁴; and 0.014, 0.045 and 1.1 × 10⁻⁵ for VTA, PVN and PFC, respectively. *P < 0.05, **P < 0.01, ***P < 0.001; NS, not significant. Summary data are presented as the mean ± s.e.m.

Fig. 6 |. Model showing the molecular basis for axonal versus somatodendritic OT release.

In the male brain during mating, OT is released from different compartments and via different mechanisms during the various behaviors. Top, during sniffing, OT is released primarily from the mPFC; during intromission, OT is released from the PVN; and, finally, during ejaculation, OT is released from the VTA and mPFC. Bottom, axonal OT release is mediated primarily by N-type VGCCs and the SNARE proteins SNAP25 and VAMP2 (left). In contrast, somatodendritic OT release is mediated primarily by L-type VGCCs and SNAP25, but does not require VAPM2 (right). Note that the release of classic neurotransmitters such as glutamate from small presynaptic vesicles is mediated by P/Q-type VGCCs, N-type VGCCs, SNAP25 and VAMP2.