Cryo-correlative light and electron tomography of dopaminergic axonal varicosities reveals non-synaptic modulation of cortico-striatal synapses

Cryo-correlative light and electron tomography of dopaminergic axonal varicosities reveals non-synaptic modulation of cortico-striatal synapses | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Cryo-correlative light and electron tomography of dopaminergic axonal varicosities reveals non-synaptic modulation of cortico-striatal synapses David Perrais, Paul Lapios, Robin Anger, Vincent Paget-Blanc, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6081416/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 13 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Dopamine is an essential brain neuromodulator involved in reward and motor control. Dopaminergic (DA) neurons project to most brain areas, with particularly dense innervation in the striatum. DA varicosities bind to target striatal synapses and form dopamine hub synapses (DHS). However, the basic features of dopamine release sites are still largely unknown. Here we studied the ultrastructure of fluorescent DA and glutamatergic (GLU) synaptosomes isolated from the striatum of adult mice with cryo-correlative light and electron microscopy and cryo-electron tomography. We observed that DA synaptosomes display ~ 10 times fewer vesicles than GLU ones. DA vesicles are bigger and less round. Vesicle organization at single nanometer scale indicates that most GLU synaptosomes have tethered and primed vesicles, indicative of a readily releasable pool, while only 39% of DA synaptosomes have tethered vesicles, which appear not to be primed. In addition, GLU terminals contacted by DA terminals in DHS have more primed vesicles than others. While DA varicosities do not form genuine synapses, their adhesion to cortico-striatal synapses may convey a local regulation of synaptic release properties. Biological sciences/Neuroscience/Synaptic transmission/Vesicle trafficking Biological sciences/Neuroscience/Synaptic transmission/Neurotransmitters Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Neuromodulation adjusts network activity and has major impacts on behavior. Among neuromodulators, dopamine acts in the basal ganglia network to encode reward prediction and participate to the initiation of movement. Dopaminergic (DA) projections to basal ganglia originate from two midbrain nuclei, the substantia nigra and the ventral tegmental area. These projections densely innervate the striatum to regulate the activity of spiny projections neurons (SPNs), which are central to functions like motor control, reward prediction and motivation 1 . Yet, at the ultrastructural level, the organisation of DA transmission is not clear. This situation is in stark contrast with the detailed characterization of neurotransmission machineries at forebrain glutamatergic (GLU) synapses. At GLU terminals, synaptic vesicles (SVs) are organized in a cluster polarized towards a portion of the plasma membrane called the active zone (AZ) which faces the post-synaptic density (PSD). The AZ contains specific proteins such as RIM1/2, bassoon and ELKS 2 . Prior to fusion, proximal SVs follow a series of steps, which can be observed with cryo-electron tomography (cryo-ET), from initial tethering, where SVs are tethered to the plasma membrane by one filament of approximately 10–25 nm 3 , 4 , to the formation of multiple short tethers, likely comprising the SNARE complex, which brings SVs closer than 5 nm to the plasma membrane 5 . These steps strongly depend on the AZ proteins RIM1 and Munc13a, which are known to control fast SV exocytosis 6 . Thus, the presence of docked and primed vesicles may constitute a hallmark of readily releasable vesicles in axons. Dopamine is also exocytosed from vesicles within milliseconds after stimulation 7 , 8 . This process relies on the calcium sensor Synaptotagmin-1 9,10 . Dopamine release shows strong paired pulse depression which is in part controlled by Synaptotagmin 7 11 . Moreover, DA axons contain AZ proteins, RIM1, Munc13 and ELKS, which are important for fast dopamine release 12 , 13 . However, only ~ 30% of DA varicosities contain such assemblies 12 . This observation is supported by functional evidence showing that around a quarter of dopaminergic varicosities are active, as assessed by release of fluorescent neurotransmitter analogues 14 . Furthermore, serial electron microscopy studies of dopaminergic axon terminals revealed heterogeneous vesicular content, with terminals containing varying combinations of small and large vesicles, while some terminals appeared to lack vesicles altogether 15 . Overall, DA terminals exhibit a non-stereotyped vesicular organization which may explain the functional diversity of release. However, observations of the accurate spatial arrangement of DA vesicles prior to fusion has been hampered by chemical fixation, staining procedures and difficulties to identify rare DA axons which negatively affects the preservation of cellular morphology and precludes interpretation of molecular details 16 . The relationship between DA release sites and target cells remains poorly documented. Upon release, dopamine binds to G-protein coupled receptors of the D1 group (D1/5R) or D2 group (D2-4R) whose signals respectively increase or decrease the excitability of target cells. On pre-synaptic terminals they influence SV release probability, while at the post-synapse they act on ion channels and glutamate receptors (reviewed in 17 ). In particular, neurons projecting from motor and prefrontal cortical regions form GLU cortico-striatal (CS) synapses responsible for the activation of SPNs. Interestingly, the stimulation of dopamine release in acute slices attenuates the release kinetic from a subset but not all GLU terminals 18 , showing that dopamine influences the activity of CS synapses. However, the precise ultrastructure through which dopaminergic terminals interact with their synaptic targets is unclear. DA synapses showing classical pre and post-synaptic features represent a minority 15 , 19 – 21 . Nevertheless, DA boutons are often found in close apposition with either pre or post-synaptic elements of CS synapses 22 . This proximity is functionally relevant, because DA release generates 2 µm wide hotspots of dopamine as estimated using carbon nanotube sensors 23 . Moreover, the synchronisation of DA phasic release with local glutamate uncaging induces structural plasticity at spines 24 . Isolation of DA synaptosomes from striatal tissue and sorting by fluorescence activated synaptosome sorting (FASS) revealed that most DA terminals are in close contact with other terminals, such as GLU presynapses 25 . These interactions are conserved even though tissue homogenisation and droplet-based fluorescence activated sorting exposed structures to significant mechanical shearing forces. We termed these multipartite DA containing synapses “dopamine hub synapses” (DHS). Among them, 25% were formed with CS synapses marked by the vesicular glutamate transporter VGLUT1 22,25,26 . Importantly, CS-DHS contain an increased signal for the presynaptic proteins VGLUT1 and bassoon compared to other CS synapses. Collectively, these findings highlight the importance of a local dopaminergic signalling for the regulation of CS synapses. However, there is still a major lack of ultrastructural observations in close-to-native conditions of DA terminals associated with their target synapses. Cryo-correlative light and electron microscopy (cryo-CLEM) and cryo-electron tomography (cryo-ET) enable identification of vitrified fluorescently labeled terminals and a close-to-native 3D observations of cellular ultrastructure and protein complexes at a nanometer scale 27 – 29 . This allows determination of the vesicular organisation in different synaptic types and characterization of protein complexes 30 . Synaptosomes are a suitable model for cryo-EM because they can be vitrified by plunge freezing, imaged in transmission EM without thinning, and because they preserve association between terminals and important pre- and postsynaptic function 16 , 29 , 31 . Here, we applied cryo-CLEM and cryo-ET to fluorescently labeled synaptosomes in order to determine the ultrastructural features of DA terminals, CS synapses and CS-DHS. We reveal the spatial organization of DA vesicles, as well as the structural association with GLU synapses in DHS. Importantly, we observed tethered SVs in DA terminals and quantified differences in SV organization of CS that are correlated to the association with DHS. Results Identification and observation of GLU and DA synaptosomes by cryo-CLEM and cryo-ET We assessed the ultrastructure of identified glutamatergic (GLU) and dopaminergic (DA) terminals extracted from the striatum of adult mice using cryo-CLEM combined with cryo-ET as described in Fig. 1 . We prepared synaptosomes from the striata of 12 to 18 week-old mice as previously described 25 and detailed in Methods and Fig. 1 A,B. We used 3 mouse lines. For GLU elements we took advantage of the knock-in (KI) mouse line in which the VGLUT1 open reading frame is tagged with the sequence of the fluorescent protein Venus 32 . In the striatum, VGLUT1 venus specifically labels cortico-striatal terminals 22 , 26 . For DA synaptosomes we injected an adeno associated viral vector carrying sequences for Cre-dependent mNeonGreen expression (AAV1 pCAG-Flex-mNeonGreen) in the midbrain of dopamine transporter promoter (DAT)-Cre transgenic mice, which specifically labels dopaminergic neurons 25 . To identify CS-DHS we used dual tagging of GLU (green) and DA (red) synaptosomes by crossing VGLUT1 venus mice with DAT-Cre 33 and the reporter line Ai14-tdTomato 34 . Striatal synaptosomes were incubated at 37°C for 15 minutes before use. They are capable of depolarization-evoked exocytosis and recycling, as shown by uptake and release of the membrane dye FM4-64 (Figure S1 ), in accordance with previous work 35 , 36 . This suspension was mixed with electron dense fluorescent fiducial beads for alignment and gold beads for tomogram reconstruction. The mixture was applied on an EM grid, plunge frozen into liquid ethane (Fig. 1 C) and kept in liquid nitrogen for further use. We first observed the grids with a cryo-fluorescence microscope. We selected isolated fluorescent spots with nearby fiducial beads for alignment (Fig. 1 D). In mice with both fluorescent markers for GLU and DA synaptosomes, we selected either an individual isolated spot, or pairs of spots where the two colours were separated by less than 1 µm corresponding to putative DHS. We then observed the same grids with a cryo-transmission electron microscope (Talos Arctica), and located regions of interest based on bead patterns (Fig. 1 E). We used the images of neighbouring independent fiducial beads to measure the pointing precision between cryo-fluorescence and cryo-electron microscopy (Figure S2; 117 ± 82 nm, n = 64). This pointing precision is smaller than the radii of GLU and DA synaptosomes (see below). Therefore, we can identify unambiguously the terminals of interest in cryo-EM, either GLU or DA. Finally, we acquired tilt series of selected synaptosomes and reconstructed tomograms with weighted back projection method using IMOD 37 . We obtained tomograms from 19 preparations (Table 1): 4 from VGLUT1-Venus mice (39 GLU synaptosomes), 5 from DAT-Cre + AAV-mNeonGreen mice (43 DA synaptosomes) and 10 from VGLUT1-Venus*DAT-Cre-Ai14-tdTomato (62 GLU and 51 DA synaptosomes, among which 32 form DHSs). We thus have a dataset of tomograms of 101 GLU synaptosomes and 94 DA synaptosomes with 32 CS-DHSs (Fig. 1 F). We created 3D models of these synaptosomes, in which we segmented the plasma membrane of the synaptosome, internal membranes and, when applicable, identified adhering structures such as a post-synaptic element (PSE), as illustrated in Fig. 1 G. GLU synaptosomes are larger and contain more SVs than DA synaptosomes The mean size of GLU synaptosomes, as measured by their maximal extension, is 823 nm, which corresponds to a visible volume of 0.0915 µm 3 (Fig. 2 A,C,D). They all contain small round SVs (Fig. 2 A). Each synaptosome contains 13 to 872 vesicles, 191 on average (Fig. 2 E), for an average density of 1956 vesicles/µm 3 (Fig. 2 F). In addition, 21/101 synaptosomes contained a mitochondrion (Fig. 1 F, 2 A) and some contained large organelles or pleiotropic organelles such as round vacuoles, multivesicular bodies, as well as other intracellular features, such as filaments and clathrin-coated vesicles (Figure S3A-F). The majority of GLU presynaptic elements (64/101, 63%) are connected to a PSE, which also contains occasionally intracellular organelles (Figure S3G-I). A gallery of representative tomographic slices and 3D models of GLU synaptosomes is shown in Figure S4. DA synaptosomes (example Fig. 2 B) have a mean size of 575 nm, which corresponds to a visible volume of 0.0341 µm 3 , significantly smaller than GLU synaptosomes (Fig. 2 C,D, p-value < 0.0001). DA synaptosomes contain on average 30 vesicles but their number is quite variable: some are almost empty (23/94 synaptosomes have less than 5 vesicles) while others contain tens or even hundreds of vesicles (Fig. 2 E and Figure S5, S6A-B). Overall, the vesicle density of DA synaptosomes is about 3-fold smaller (723 per µm 3 ) than for GLU synaptosomes (Fig. 2 F). DA synaptosomes also contain other organelles, such as vacuoles, mitochondria, multivesicular bodies (Fig. 1 F, 2 B, S6C-H). Table 2 reports the number of DA synaptosomes with these organelles. While SVs in GLU synaptosomes are spherical, small and uniform in size (outer diameter 39.8 ± 5.8 nm, n = 18 897, with only 2.7% vesicle larger than 60 nm), vesicles in DA synaptosomes are significantly larger (44.9 ± 9.9 nm, n = 2782, p < 0.0001) with 8.8% of vesicles having a diameter in the 60–100 nm range (Fig. 2 G). In total, 66/94 DA synaptosomes have at least one vesicle bigger than 60 nm, with a proportion of 23.5 ± 19.6% big vesicles in these synaptosomes. Moreover, vesicles in DA synaptosomes are often elongated or pleomorphic (Fig. 2 H,J). We used Wadell’s index WI (see Methods) to quantify vesicle sphericity. Most vesicles in GLU terminals have an index close to 1 (perfect sphere). However, vesicles in DA synaptosomes are significantly less round than the ones in GLU synaptosomes (Fig. 2 H, p-value = 0.0083). These elongated vesicles are spread across the synaptosomes (Fig. 2 J, S5). Overall, there are 65/94 DA synaptosomes containing at least one elongated vesicle (WI < 0.95) (Fig. 2 J). In these synaptosomes there are 29.2 ± 26.9% of elongated vesicles. Contact zones between GLU, DA synaptosomes and PSEs In GLU synaptosomes we identified a clear PSE separated from a presynaptic terminal by a clearly defined synaptic cleft in 64/101 (63%) synaptosomes. Typically, the PSE is defined by a sealed plasma membrane compartment (Fig. 1 F, 3 A,C, S4) but in some cases it had an open membrane adhering to the presynaptic terminal forming a synaptic cleft (Fig. 1 F,S4). The area of contact between GLU and PSE is characterized by roughly parallel membranes defining a wide cleft (31.8 ± 6.2 nm, n = 34) with dark material around the midline (Fig. 3 D,E), as observed previously 38 , 39 . The active zone (AZ), defined as the membrane region of GLU synaptosome in contact with the PSE (outlined in red in Fig. 3 A), has an area of 0.079 ± 0.055 µm 2 (n = 33) (Fig. 3 F). We analysed the presence of protein density in the pre- and post-synaptic elements of GLU synaptosomes. The normalized densities decrease by 20% in both compartments (Fig. 3 J,K). However, we detected a clear increase in density 10 to 25 nm from the PSE membrane, which corresponds to the post-synaptic density (PSD), as detected with other modalities of electron microscopy for synapses in situ 40 and synaptosomes with cryo-ET 29 . In contrast, only two out of 94 DA terminals were separated by a wide cleft from a characteristic PSE containing PSD, as in GLU synaptosomes. Interestingly, one of these DA synaptosomes has the most vesicles (610), all of them small and round, which makes it indistinguishable from classical GLU synaptosomes (Figure S6A). The other one has 40 vesicles (Figure S6B). In these two synaptosomes, the distribution of vesicle sizes and sphericity matches distributions seen for GLU synaptosomes (dotted line in Fig. 2 G,H). Another DA synaptosome has 295 vesicles (see two upper dots in Fig. 2 E) and also resembles classical GLU synaptosomes, even though it was not connected to a PSE but engaged into a CS-DHS. These DA synaptosomes could originate from neurons co-expressing VGLUT2 and which are able to release glutamate and generate AMPAR-mediated post-synaptic potentials 41 , 42 . Nevertheless, many DA terminals are in close apposition with other structures which are occasionally resembling a GLU terminal (Figure S5A) or other, harder to characterize structures which could be parts of postsynaptic spines. To get a better insight on the features of contact zones between DA and GLU terminals, we analyzed 32 tomograms from apposed GLU and DA synaptosomes which correspond to CS-DHSs 25 . The area of contact between GLU and DA synaptosomes have a similar size as GLU active zones (Fig. 3 B,F). The width of the cleft between GLU and DA terminals is 12.1 ± 2.4 nm, much smaller than the synaptic cleft of GLU synaptosomes (Fig. 3 G, p < 0.0001). Along this cleft, dense material is also observed, albeit not as pronounced as for GLU/PSE synaptic clefts (Fig. 3 H,I). The intracellular side of DA and GLU terminals around the adhesion site are devoid of visible densities such as PSDs (Fig. 3 L,M). In 2 tomograms, the DA terminal is not in contact with the presynaptic GLU synaptosome but with the PSE (Fig. 3 C,S7B). The contact area (0.062 and 0.066 µm 2 ) and cleft sizes (10.9 and 15.6 nm) are in the same range as for DA/preGLU contact sites (Fig. 3 F,G). Similarly, the densities around the contact site are in the same range, without signs of a PSD (Fig. 3 N,O). More examples of DHS virtual planes and models are available in Figure S7. Vesicles in GLU and DA terminals are tethered to the plasma membrane and are interconnected. Synaptic vesicles in GLU and GABAergic synapses are divided into functional pools 43 . SVs which fuse first with the plasma membrane upon calcium stimulation comprise the readily releasable pool, which was defined morphologically by EM of chemically fixed, dehydrated samples as the vesicles which appear docked, that is in direct contact with the plasma membrane 44 , 45 . Cryo-ET revealed that in close to native state, SVs are not docked but are tethered to the plasma membrane via one or several electron dense filaments, as observed in synaptosomes and dissociated neuronal cultures 3 , 4 , 29 , 30 , 46 . We found at least one tethered vesicle at the active zone in almost all quantified striatal GLU synaptosomes (24/26, 92%) (Fig. 4 A). Tethers, as well as connectors (bridges that interconnect SVs) were detected using an automated, template-free method by the hierarchical connectivity algorithm 47 . Among proximal vesicles, defined as those localized less than 45 nm from the plasma membrane, 40 ± 7% have one or more tethers (Fig. 4 C). These tethers have a length of 13.9 ± 7.5 nm (Fig. 4 D). The number of tethers increases as vesicles are located closer to the plasma membrane (Fig. 4 E). Remarkably, vesicles located less than 5 nm from the plasma membrane have on average 3 tethers, which corresponds to synaptic vesicles that are functionally primed for fusion 3 , 4 . In DA synaptosomes, we found clear examples of tethers between vesicles and the plasma membrane, as well as vesicles connected by more than one tether (Fig. 4 B). We detected vesicles tethered to the plasma membrane in 39% (13/33) of DA synaptosomes. The percentage of tethered vesicles among proximal vesicles in DA is 25 ± 6%, which was lower, but not significantly different from GLU synaptosomes (Fig. 4 C). However, tether length is on average significantly larger in DA versus GLU (DA: 22.43 nm; std. 12.21; GLU: 13.89 nm; std: 7.47; p-value < 0.001) (Fig. 4 D) likely a direct consequence of the complete absence of tethered DA vesicles localized less than 5 nm from the plasma membrane (Fig. 4 E). Importantly, for DA and GLU vesicles located further away, the number of tethers per SV is not different (Fig. 4 E). Another morphological hallmark of the polarized exocytosis of SV in cortical synaptosomes is the larger fraction of volume occupied by proximal synaptic vesicles towards the active zone 3 , 4 , 29 . GLU synaptosomes showed a mean peak occupancy around 20 nm from the plasma membrane that is clearly visible in individual occupancy profiles (Fig. 4 F,G) and is similar to those observed for cortical synaptosomes 3 , 4 , 29 . In DA synaptosomes, the occupancy lacked the peak in the proximal vesicles region, but occasional enrichment close to the whole plasma membrane was observed (Fig. 4 H,I). However, the absence of a morphologically identifiable DA active zone, may have biased the measurement of occupancy, because it included a larger, possibly irrelevant volume in our analysis. The percentage of inter-connected SVs was significantly higher in GLU (53%) than in DA synaptosomes (20%) (Fig. 4 J) and the length of connectors was significantly different (DA: 18.50 nm; std. 9.18 and GLU: 16.33 nm; std. 13.07; p-value = 0.0021) (Fig. 4 K). Finally, among proximal vesicles, there were more SVs that were both tethered and connected in GLU (17%) than in DA synaptosomes (6.4%). In the latter, the majority of vesicles were neither tethered nor connected (Fig. 4 L). Tethered SVs highlight a subpopulation of DA terminals We compared DA synaptosomes that contain (T+) to ones that do not contain (T-) tethered vesicles (Fig. 5 A). We excluded from the analysis the 2 DA synaptosomes with a PSE (see Figure S6AB). The T + DA synaptosomes contain significantly (twice) more SVs (Fig. 5 B), and are significantly larger than the T- synaptosomes (Fig. 5 C). On average, T + synaptosomes contain 2 tethered vesicles, and a maximum of 5 (Fig. 5 D). In T + DA terminals we do not observe a clear AZ, which corresponds in GLU synapses to the location facing the synaptic cleft where tethered synaptic vesicles concentrate and undergo exocytosis upon stimulation. We looked for a putative DA active zone in two ways. First, the active zone could locate at or right next to the contact zone with GLU terminals. However, tethered vesicles are almost never found at the contact site. Second, an active zone is expected to concentrate tethered vesicles. Therefore, we analyzed DA synaptosomes with multiple tethered vesicles. Among the 13 T + DA synaptosomes, 6 had multiple tethered vesicles, for a total of 19 tethered vesicles (Fig. 5 D,E). The shortest arc distance (i.e. staying on the plasma membrane) between these vesicles is on average 210 nm (Fig. 5 F). This is smaller than the average distance between random points on the synaptosome plasma membrane (estimated at 567 nm, see Methods for derivation). This suggests that tethered DA vesicles may gather in a preferential zone of the plasma membrane, a putative active zone. Moreover, in T + DA synaptosomes, vesicles are located closer to the plasma membrane than in T- DA synaptosomes (Fig. 5 G), reinforcing the idea that in DA synaptosomes vesicle location is polarized to a putative active zone where vesicle may tether and fuse. We also compared DA synaptosomes forming CS-DHS (that is adhering to a GLU terminal) or not. We found no significant differences in their vesicle number, spatial organization and tethering (Figure S9). This suggests that the adhesion to a GLU terminal does not affect the propensity of DA synaptosomes to contain tethered vesicles. Correlation of DHS connection with alterations of CS terminals We wondered whether the adhesion of a DA terminal to a GLU synapse affects the ultrastructure of the GLU presynapse. We compared GLU synaptosomes that were not part of DHS (GLU DA- synapses, Fig. 6 A, n = 30) with GLU synaptosomes engaged into a DHS (GLU DA + synapses or DHSs, Fig. 6 B, n = 32), both from the dual color model. The total number of vesicles are not different between GLU DA- and GLU DA + terminals, on average 205.7 and 222.9 vesicles (Fig. 6 C; p-value = 0.864), as well as, the density of vesicles, 1893 per µm 3 and 1943 per µm 3 , respectively (Fig. 6 D; p-value = 0.577). Similarly, the AZ areas are not different (Fig. 6 E). Together, these argue that the morphology of terminals and overall distribution of SVs is the same. Neverthless, the fraction of the volume occupied by proximal SVs was significantly higher in GLU DA + than in GLU DA- synaptosomes, and showed a peak that signifies a higher concentration of proximal SVs, while the occupancy peak was absent in GLU DA- terminals (Fig. 6 F,G). A precise characterization of SV location and tether length showed that the results for GLU DA + terminals were similar to those of unperturbed hippocampal glutamatergic synapses reported previously 3 , while GLU DA- terminals had a lower proportion of SVs located 5–10 nm to the AZ membrane but, surprisingly, a higher proportion of tethers 6–12 nm in length (Figure S10). Additionally, proximal vesicles in GLU DA + synaptosomes are significantly more connected to other vesicles than in GLU DA- synaptosomes (46.7% vs. 21.6%; p-value = 0.0167) (Fig. 6 J). The fraction of tethered vesicles was about 70% bigger in GLU DA + terminals but did not reach significance. However, the fraction of proximal SVs that are both tethered and connected was significantly higher in GLU DA + terminals (26.7% vs. 5.4%; p-value = 0.0435) (Fig. 6 K,L). Together with the observation that the number of tethers per SV was similar in GLU DA + and GLU DA-terminals at all distances to the plasma membrane (Fig. 6 K), our results argue that connectors are primarily responsible for the observed difference in the proximal SV distribution. Discussion Here, we report the ultrastructure of fully hydrated, close-to-native DA terminals from adult mouse striatum at a single nanometer resolution, as well as of DHS, which comprise DA terminals in contact with GLU synapses and were characterized previously with immunofluorescence 25 . Synaptosomes obtained from mouse brain constitute a reliable model to investigate the spatial configuration of synapses 16 , 38 . They are particularly amenable for cryo-ET because they can be directly observed by cryo-EM without further processing such as cryo-sectioning or cryo-focused ion beam milling 16 . Moreover, they retain functionality, such as SV exocytosis and endocytosis, protein composition and post-synaptic calcium signalling 3 , 29 , 35 , 38 . Likewise, we show with FM4-64 labelling that our preparation of synaptosomes undergoes stimulation dependent SV cycling. Moreover, the GLU synaptosomes we obtained from the mouse striatum are qualitatively and quantitatively comparable to forebrain and hippocampal synaptosomes (likely glutamatergic), and dissociated cultures observed previously with cryo-ET 3 – 5 , 29 , 30 . The presynaptic element contains hundreds of small round SVs (~ 40 nm diameter) and contact a PSE with a clearly defined PSD and a 32 nm wide synaptic cleft filled with dense material. This defines an AZ of ~ 0.08 µm 2 , in the range of reported sizes varying between 0.04 µm 2 and 0.10 µm 23 , 48 , 49 . Moreover, SVs are polarized towards the AZ, with a peak of volume occupancy ~ 25 nm from the plasma membrane, which reflects the enrichment in tethered and primed vesicles, similar to forebrain synapses and was proposed to be necessary for proper neurotransmitter release 50 . Nevertheless, this peak was less pronounced in our sample of CS synaptosomes than in forebrain synaptosomes 3 , 4 , which may reflect genuine differences in relative SV pool sizes in cortical/hippocampal vs striatal synapses. Remarkably, proximal SVs located less than 5 nm from the AZ have in both CS and cortical synapses on average 3 tethers 4 , 5 , which likely corresponds to the primed state of SV and to functionally defined readily releasable SVs. In neurons lacking Munc13-1 and Munc13-2, which abolishes priming 51 , proximal SVs have only 1 tether 4 . Therefore, we propose that CS synapses have primed vesicles with the same structural hallmark as the ones defined in hippocampal synapses. The most striking feature of DA synaptosomes is the relative sparsity of small synaptic-like vesicles, about 3-fold less dense than in GLU synaptosomes. Because DA synaptosomes are also smaller than GLU synaptosomes, there are 10 times less vesicles in DA synaptosomes than in GLU synaptosomes. These observations are consistent with ultrastructure of DA axons determined with serial electron microscopy reconstruction in chemically fixed striatum tissue in which axonal varicosities display very heterogenous ultrastructure: some have only few vesicles, others have small synaptic-like vesicles, larger ones, or both types of vesicles 15 . Likewise, recent studies have shown qualitatively similar results in cultured DA neurons observed with cryo-CLEM 52 and DA neurons derived from human induced pluripotent stem cells 53 . Vesicles in DA synaptosomes are also significantly more elongated than vesicles in GLU synaptosomes. Elongated-shaped vesicles have also been documented at GABAergic inhibitory synapses with conventional electron microscopy 54 and also with cryo-CLEM 30 . Moreover, SVs in GLU synapses become elongated in the absence of VGLUT1, which reflects a lower luminal osmotic pressure 55 , 56 . Therefore, DA vesicles, like GABA vesicles, experience different osmotic pressure than GLU vesicles. Importantly, we show that, similar to GLU synaptosomes, vesicles in DA synaptosomes are frequently linked together by connectors and to the plasma membrane by tethers. Nevertheless, only 39% of the DA synaptosomes contain at least one tethered vesicle. The DA synaptosomes with at least one tethered vesicle are bigger and contain twice as many vesicles as the ones without any tethered vesicle. Interestingly, other studies have shown that only 30% of DA axonal varicosities contain active zone proteins bassoon, RIM or ELKS 12 , and this proportion is also found for functional DA varicosities able to release fluorescent dopamine analog 14 . Therefore, we propose that synaptosomes containing tethered vesicles correspond to release-competent DA terminals. Moreover, the number of tethers linked to proximal vesicles (less than 45 nm from the plasma membrane) is the same in GLU and DA synaptosomes. This suggests that similar molecules are involved in vesicle tethering in both types of synapses. Indeed, proteomic analysis of striatal DA synapses show similar degree of enrichment of the major proteins involved in vesicle tethering 25 , 57 . However, DA synaptosomes lacked very proximal vesicles (less than 5 nm from the plasma membrane), which have on average 3 tethers, a hallmark of vesicles primed for release in GLU synapses. This difference points to a fundamental difference between these two types of terminals. We thus predict that the kinetics of vesicle exocytosis will be much slower in DA neurons than in GLU neurons, where SV exocytosis occurs in less than 1 ms after calcium entry. So far, the kinetics of DA vesicle exocytosis has been determined with fast amperometry or fluorescence imaging of dopamine sensors 7 , 8 .The rise time of these events is greater than 10 ms but it could be dominated by diffusion of ligand to the detector micrometres away from the dopamine release site. More precise investigation of dopamine release is needed to determine if the kinetics of dopamine release is genuinely slower, and whether this is due to differences in vesicle priming. We found a clear PSE at only 2 out of 94 DA terminals. These are among the synaptosomes which contain many small round vesicles, and are thus indistinguishable from GLU synaptosomes. They may correspond to terminals of DA/GLU neurons co-expressing the vesicular glutamate transporter VGLUT2 41,42,58 . In the adult mouse, these terminals are mostly concentrated in the shell of the nucleus accumbens. Interestingly, in these axons VGLUT2 is segregated from the vesicular dopamine transporter VMAT2, suggesting that glutamate and dopamine release sites are segregated 58 . In the other 92 DA terminals reconstructed, no clear PSE was identified. We identified previously that DA terminals can strongly interact with other synaptic elements in a so-called DHS 25 . In the 32 reconstructed DHS, we found that 28 directly adhere to a presynaptic VGLUT1 element and 4 contacting the GLU PSE. On the other hand, DA terminals, identified in conventional transmission EM with immuno-labelling of tyrosine hydroxylase, are as likely to contact the pre- or post-synaptic side of CS synapses 22 . The selection of DHSs by cryo-fluorescence with the presynaptic marker VGLUT1-Venus likely biased our sampling towards interaction with presynaptic markers. Nevertheless, we could characterize both pre- and post-synaptic DA/GLU adhesion sites in their native state with cryo-ET. We found that the adhesion sites had similar areas, around 0.05 µm 2 , as well as the cleft sizes, around 12 nm. However, these adhesions did not define a privileged tethering site for DA vesicles. Overall, we did not identify a localized site for DA vesicle tethering, suggesting a large AZ for exocytosis. Finally, we showed that the morphology of GLU terminals and overall distribution of SVs are very similar in GLU DA + and GLU DA- terminals. Also, we detected tethers of different lengths, which were previously associated with different molecular composition of tethers 3 , in both types of presynaptic terminals. These suggest that the GLU terminal formation and their protein composition does not depend on the presence of DA terminals. Nevertheless, the distribution of proximal SVs in GLU DA + terminals was consistent with those of non-perturbed glutamatergic synapses, while GLU DA- terminals showed a flat proximal SV distribution profile previously associated with a reduced neurotransmitter release 3 , 4 . Furthermore, we correlated these differences with changes in SV connectivity and possibly also with tethering. Therefore, our data indicate that the contact between DA and GLU terminals at DHS modulates the SV organization at the single nanometer scale, as well as release properties of GLU terminals, and that this modulation is mediated by SV connectors. We could hypothesize that the formation of DHSs in the striatum is an important feature modulating glutamate release. Overall, DHSs could be an important substrate for the modulation of glutamate release and striatal activity in vivo. These multi-partite assemblies offer a close proximity between dopamine release sites and specific glutamatergic synapses. By this mechanism, the spatial and temporal synchronization between dopamine and glutamate activity would be maximal, which could play a major role in the plasticity of excitatory input to striatal neurons 24 . Materials and Methods Animal models Three mouse models have been used. The VGLUT1 venus knock-in line to label CS synapses. The DAT-cre BAC transgenic mouse line 33 in which we transduced VTA/SNc neurons with an AAV1 carrying pCAG-Flex-mNeongreen coding sequences to label dopaminergic neurons projecting to the striatum (coordinates from bregma are A/P: 2.9mm; M/L: 1.6 mm; D/V: 4.6 mm with 12° angle). To co-detect CS synapses with DA terminals in the same sample, we used a double labelling approach. DAT-cre crossed with the reporter Ai14TdTomato mouse line were crossed with VGLUT1 venus KI mice. Adult mice of both genders, aged 12 to 18 weeks were used. We refined the experimental design and the procedures to reduce as possible the number of animals used and their suffering. All procedures were in accordance with the European guide for the care and use of laboratory animals and approved by the ethics committee of Bordeaux University (CE50) and the French Ministry of Research under the APAFIS n° #21132 and #38144. Preparation of synaptosomes The preparation of synaptosomes was adapted from a previously published protocol 59 . Briefly, animals were euthanized by cervical dislocation, decapitated and the head was immersed in liquid nitrogen for 5 seconds for rapid cooling but not freezing of the tissue. The striata were subsequently dissected under an epi-fluorescence stereomicroscope (Leica Microsystems, Germany). Samples were then homogenized in 1.5 ml of ice-cold isosmolar buffer (0.32 M sucrose, 4 mM HEPES pH7.4, protease inhibitor cocktail Set 3 EDTA-free (EMD Millipore Corp.)) using a 2 ml-glass-Teflon® homogenizer with 12 strokes at 900 rpm. The homogenate (H) was centrifuged at 1000 g for 5 min at 4°C in a benchtop microcentrifuge. The supernatant (S1) was separated from the pellet (P1) and centrifuged at 12500 g for 8 min at 4°C. The crude synaptosomes pellet (P2) was suspended in 350 µL of isosmolar buffer and layered on a two-step ficoll density gradient (5 mL of 13% Ficoll and 5 mL of 7.5% ficoll, both in 0.32 M sucrose, 4 mM HEPES). The gradient was centrifuged at 50,000 × g for 1 h and 10 min at 4°C (Optima L100XP Beckman Coulter with SW32Ti rotor). The synaptosome fraction was recovered at the 7.5 and 13% ficoll interface using a 0.5 ml syringe. An additional centrifuge of the collected fraction was performed in 2ml of HEPES-buffered Krebs like solution (HBK) 36 composed of : 143 mM NaCl, 4,7 mM KCl, 1,3mM MgSO 4 , 1,2 mM CaCl 2 , 20 mM HEPES, 0,1mM Na 2 HPO 4 and 10 mM D-glucose, pH = 7,4) at 12,500g for 5min in order to wash the excess of ficoll and sucrose residues. The barely visible pellet was resuspended in 200µl HBK leftover and placed for 15 minutes at 37°C for physiological recovery before plunge-freezing. Alternatively, S1 fraction diluted in HBK has been used allowing for faster preparation with comparable results. FM4-64 uptake and release assay To assess synaptosome integrity, we performed a fluorogenic vesicle exo-endocytosis assay using the amphiphilic styryl dye FM4-64 (Thermo Scientific T13320). Synaptosomes were prepared fresh with the same protocol as for cryo-CLEM. Then they have been diluted in HBK and centrifuged 34 min at 6750g on 12 mm coverslips coated with 1 mg/ml Poly-L-Lysine. Imaging was performed right after on a wide field epifluorescence microscope Leica DMI8 equipped with an inverted 63X/1.4 oil objective, a Hamamatsu Flash 4.0 v2 camera and a controlled 37°C/CO2 chamber. Coverslips were placed in a recording chamber (Ludin), incubated with 500 µl of prewarmed HBK and imaged in TRITC and GFP channels at different registered positions over a 5 µm stack (1st acq). HBK was removed and FM4-64 (6 ng/µl final concentration), KCl (40 mM final concentration) and 37°C/CO2 HBK mixture was added inducing exo-endocytosis cycle and staining membrane and internalized vesicles (see supplementary Fig. 1A). After 1 min 30 s, KCl was rinsed twice with HBK and FM4-64 with HBK was added staining the remaining plasma membrane. After 1 min 30 s, second acquisition was launched (2nd acq). After 5 rinses with HBK, a third acquisition captured signal from internalized vesicles only (3rd acq). KCl-HBK mixture was added to induce depolarization and exocytosis and distaining was imaged 1 min 30 s after (4th acq). Loading was measured by subtracting FM intensity of single synaptosome at acquisition n°3 (vesicles stained) versus acquisition n°1 (background). Release was measured by subtracting intensity of single synaptosome at acquisition n°3 (vesicles stained) with acquisition n°4 (non exocytosed vesicles and background dye binding). Control experiments with no KCl for loading or no KCl for release were performed. Analysis was performed on Fiji 60 using a home-built macro-command. Plunge-freezing Quantifoil R2/2 Cu 200 or R3.5/1 Cu 200 mesh grids (Q-R2_2-2C100 or Q-3.5_1-3C100 Delta Microscopies) were glow-discharged for 30 seconds at 2.5 mA using a ELMO glow discharge system (Cordouan Technologies) to enhance their hydrophilicity. Following this, 4 µl of freshly purified synaptosomes, mixed with with 10 nm colloidal gold beads (Sigma 741957) and 100 nm FluoSpheres™ beads (F8797 ThermoFischer) were applied to the grids. The excess sample was immediately blotted away from the opposite side for 5 seconds in a Vitrobot Mark IV (Thermo Fisher Scientific), maintained at a temperature of 4°C and at 100% humidity. Finally, grids were plunge-frozen in liquid ethane and stored in liquid nitrogen until observation. Cryo-fluorescence microscopy We used a commercially available system to perform cryo-fluorescence microscopy (Leica – DM6 FS Cryo CLEM). We monitored on test grids with TetraSpecks™ beads (Thermo Scientific T7279) shifts between channels and drift artifacts in cryogenic conditions. No significant shift was noticed, while noticeable mechanical drift occurs for acquisitions of more than 20 z-steps, which may be necessary when the grid is not flat and exactly parallel to the focal plane. Therefore, we limited our acquisitions to flat areas and imaged consecutive z planes with an increment of 0.5 to 1 µm (20 z-steps max). A custom-made Plexiglas® chamber was built around the set-up to maintain relative humidity under 40% reducing contaminations by water. Temperature of the objective chamber was maintained at -190°C (83K) using a pump projecting vapor of fresh liquid nitrogen. Each grid was observed separately reducing the time spent in the chamber to 25 minutes maximum. Sample was observed with a 50X objective (Leica – HC PL APO 50x/0.90NA Dry 11566064) and z-stacks were acquired through a Hamamatsu ORCA-Flash 4.0 camera, in 4 channels, in the following order: Brightfield (Empty); Tx Red (Em: BP560/40 nm; Ex: BP630/75 nm); GFP (Em: BP470/40 nm; Ex: BP525/50 nm); DAPI (Em: BP360/40 nm; Ex: BP470/40 nm). Large, flat area of the grids were acquired using the LAS X Navigator mosaic imaging tool. Grids were stored in liquid nitrogen until the cryo-EM session. Images from regions of interest were stacked in a maximal intensity Z-projection using Fiji software for each fluorescence channels 60 . Minimal Z projection was performed for the brightfield channel allowing the detection of holes in the carbon layer. Contrast has been adjusted and allowed detection of all synaptosomes (dim and bright), as well as fluorescent beads. TIFF images have been saved in PNG file format for editing. Each square of interest (with one or more synaptosome in a hole) was identified and marked with a unique number. Cropping of each corresponding area into a single image was done and resulting PNG files were saved-back into TIFF files for compatibility with Serial EM 61 . Correlation between fluorescence and electron microscopy The correlation process was carried out in 2 stages. The initial “approximate” correlation focused on locating the grid squares captured through cryo-fluorescence microscopy. After inserting the grid into the cryo-transmission electron microscope, a low-magnification montage (34x) was constructed using SerialEM 61 to visualize the entire grid. The fluorescence image was then imported into the software via the “Import Map” function. Areas of interest were identified using obvious landmarks, such as carbon holes and the center of the grid, which were visible in both imaging modalities. Once the squares were identified, a high-magnification montage (1600x) was created for each area of interest. The corresponding fluorescence image for each square was imported separately. Precise correlation in SerialEM was done using the fluorescent fiducial beads as registration points, detectable in both fluorescence and electron microscopy. Beads were selected and assigned specific identifiers on the fluorescence image, and the same beads were subsequently located on the electron microscopy image. A total of 5 to 10 surrounding beads were used for the transformation to correlate with a single target for tomography. Both images were then correlated using the “Transform Item” function. Cryo-electron tomography Tilt series of the synaptosomes were recorded using a Talos Arctica microscope (Thermo Fisher Scientific), operating at 200 kV and equipped with a K2 Summit direct electron detector (Gatan). The tilt series were acquired in a bidirectional scheme using SerialEM 4.0, covering angles from − 60° to + 60° with 2° increments, at a magnification of 11,000x, resulting in a pixel size of 3.987 Å. Imaging was performed with an underfocus of -8 µm, and the cumulative electron dose ranged between 80–90 e⁻/Ų. Tilt series alignment and tomogram reconstruction were carried out using EMAN2 62 for quick visualization and data selection. IMOD software package 4.11.25 37 integrated into the SCIPION framework 63 was used to reconstruct selected tomograms for segmentation in 3dmod (see below). Reconstruction was achieved using weighted back-projection combined with a SIRT-like filter. Binning of 4 was applied and resulted in a pixel size of 1.5588 nm. Segmentation and production of models We selected tomograms with accurate correlation, sufficient contrast and intact synaptic features for segmentation (numbers provided in Table 1). Other tomograms were discarded (examples in Figure S8). Manual segmentation was performed using 3dmod, a software from the IMOD package 64 . Synaptosome plasma membranes were segmented until disappearance and meshed, therefore no interpolation was used and missing wedge volume was not corrected for the synaptosomes. Organelles and vesicles were interpolated using spherical interpolation, resulting in closed objects. Active zone area was defined as plasma membrane portion facing the post-synaptic density. Contact area (DHS) was the plasma membrane portion of the DA synaptosomes following tight apposition (< 15 nm) to the GLU plasma membrane. Synaptic vesicles were defined as objects smaller than 80 nm with round or elongated shapes. Mitochondria were easily detected as dense folded membranes are present in the lumen. Vesicular bodies were observed thanks to the presence of one or more vesicle-like structure inside. Objects with folded, irregular shapes and bigger diameter have been categorized separately and correspond to unidentified structures. Data analysis We obtained geometrical information such as volume and maximum diameter from the 3D models using imodinfo command lines (IMOD package). Sphericity was calculated using Wadell’s index 65 (W i ) with the following formula W i = \(\:{\pi\:}^{\frac{1}{3}}{\left(6V\right)}^{\frac{2}{3}}/S\) where V represents the volume and S the surface of the vesicle. Thus W i = 1 for a perfect sphere and decreases for non-spherical objects. Vesicle density inside synaptosomes was obtained dividing the number of vesicles by the volume available inside the synaptosome. In Fiji, cytoplasmic density profiles were obtained averaging a 40 nm stack and pixel intensities were measured using the built-in commands. In Fiji, synaptic cleft and interspace density profiles were measure at 3 different single z planes (low, middle and high) in the stack and averaged for each tomogram. Values were normalized to the neighboring non-cleft area corresponding to background. We excluded tomograms in which gold beads were present as it artificially alters the density. SV tethers and connectors were detected in and automated, template-free manner using the hierarchical connectivity algorithm, and their morphology, localization and interrelationship was analyzed by Pyto package (version 1.10, available at https://github.com/vladanl/Pyto ), as described before 29 , 47 . Briefly, for the analysis of vesicle distribution (volume occupancy), the presynaptic cytoplasm (including SVs) was divided into 1-pixel-thick layers according to the distance to the AZ membrane, and the fraction of the layer volume occupied by SVs was measured. Connector and tether lengths were calculated as the minimal edge-to-edge distance between connector / tether voxels contacting an SV or plasma membrane that takes into account central regions of tethers and connectors. In this way, the ambiguity inherent to the measurement of length of 3D objects is resolved and curvature of tethers and connectors contributed to their calculated lengths. However, possible extended protein-lipid binding regions are not considered. All image processing and statistical analysis software procedures were written in Python and implemented in Pyto package [37]. Pyto uses NumPy and SciPy packages and graphs are plotted using Matplotlib 66 – 68 Statistical analysis was performed between the experimental groups using only planned, orthogonal comparisons. For the analysis of properties pertaining to individual SVs, connectors and tethers (such as the SV distance to the AZ membrane, tether length and fraction of tethers/connectors having a certain property), values within experimental groups were combined. Bars on the graphs show mean values and error bars the standard error of the mean (sem). In cases a fraction of SVs or tethers is shown, the error bars represent sem between synapse means. We used Student's t test for statistical analysis of values that appeared to be normally distributed (e.g., vesicle diameter) and K-W test (nonparametric) for values deviating from the normal distribution (e.g., number of tethers and connectors per vesicle). For frequency data (e.g., fraction of connected and non-connected vesicles), χ2 test was used. In all cases, confidence levels were calculated using two-tailed tests. The confidence values were indicated in the graphs by a single asterisk for P < 0.05, double for P < 0.01, and triple for P < 0.001. All values of statistical tests are presented in Table 3 (Supplementary data). We focused on tomograms originating from the VGLUT1 venus x DAT-cre x Ai14 tdTomato because it is the only model where we could reliably identify both GLU and DA and thus determine CS-DHS and non-CS-DHS conditions with certainty (see Table 2). Moreover, we selected tomograms where the contrast was optimal and sufficient information was available. We picked GLU synaptosomes where we identified the active zone and for DA where there were more than 2 vesicles. Thus, it ensures an optimal detection and a reliable comparison of the filaments between both conditions. We normalized tomogram density and applied a Kernel Gaussian filter (sigma = 2) to improve detection. Tomograms used for the glutamatergic active zone were cropped on the dedicated region and analysis was performed blind to the DHS, non-DHS condition. Estimation for shortest distance between tethered vesicles We measured the arc distance between two tethered DA vesicles d (that is the distance while remaining in the plasma membrane) by measuring the distance in the projection in the tomogram plane measured along the arc of the plasma membrane (d xy ) and the distance between the two tomogram planes in which the tethers connect the plasma membrane (d z ) as d = \(\:\sqrt{\left({d}_{xy}\right)²+\left({d}_{z}\right)²}\) . We estimated the distance between random points by taking the average maximum diameter of T + DA synaptosomes D (722 ± 205 nm, Fig. 5 C). For the distance between random points, we approximated the projection of synaptosomes by a circle of radius R = D/2 = 361 nm. The average arc distance between random points in a circle is half the length of a half circle (by symmetry), or \(\:\frac{\pi\:}{2}R\) = 567 nm and its variance is \(\:\frac{\pi\:²}{12}R²\) that is a standard deviation of 327 nm. This estimate is a lower limit of the true value, because it neglects the axial distance d z and approximates synaptosomes as circles, minimizing distance. Declarations Competing interests The authors declare no competing financial interests. Author contributions P.L. performed stereotaxic AAV injections. synaptosome preparations and live imaging of FM4-64 uptake with the help of V. P-B. P.L. performed synaptosome freezing, cryo-fluo imaging, cryo-EM and cryo-ET with the help of R.A. E.M. and R.F. supervised the cryo-ET methodology. P.L. did all reconstructions and annotations of tomograms under the supervision of E.H. and D.P. P.L. performed quantitative analysis of all data together with V.L. P.L., R.F., E.H. and D.P. acquired funding. P.L., E.H. and D.P. wrote the manuscript and all other authors edited it. Acknowledgements We thank Peter Vanhoutte and Nicolas Heck for providing the DAT-cre * Ai14 tdTomato line, the Pôle in Vivo for animal breeding, husbandry, help with stereotaxic injections. We thank members of the Bordeaux Imaging Center, namely Monica Fernandez-Monreal from for help with cryo-fluorescence microscopy and Fabrice Cordelières for the analysis of FM4-64 loading experiments. We thank student interns Timothé Lapha, Jade Giraud and Solène Hospital for reconstruction of some of the tomograms. This work was supported by ANR (DopamineHub ANR-19-CE16-0003 to E.H. and D.P., FrontoFAT ANR-20-CE14-0020-03 to E.H. and UltraDopa ANR-24-CE16-5973-01 to E.H., D.P. and R.F.), the Fondation Recherche Médicale (to P.L., end of PhD and D.P., FRM team), the European Research Council (ERC consolidator Grant PneumoTransfo to R.F.) and the Regional Council of Nouvelle Aquitaine (ParkSynGraft 205024) to E.H. and D.P. This work was supported by the French Government through the France 2030 program [grant number 21-ESRE-0024] managed by the French National Research Agency (ANR) as part of the "Investissements d'avenir" program. References Costa KM, Schoenbaum G, Dopamine (2022) Curr Biol 32:R817–R824 Emperador-Melero J, Kaeser PS (2020) Assembly of the presynaptic active zone. Curr Opin Neurobiol 63:95–103 Fernández-Busnadiego R et al (2013) Cryo–electron tomography reveals a critical role of RIM1α in synaptic vesicle tethering. J Cell Biol 201:725–740 Papantoniou C et al (2023) Munc13- and SNAP25-dependent molecular bridges play a key role in synaptic vesicle priming. Sci Adv 9:eadf6222 Radecke J et al (2023) Morphofunctional changes at the active zone during synaptic vesicle exocytosis. EMBO Rep 24:e55719 Imig C et al (2014) The Morphological and Molecular Nature of Synaptic Vesicle Priming at Presynaptic Active Zones. Neuron 84:416–431 Patriarchi T et al (2018) Ultrafast neuronal imaging of dopamine dynamics with designed genetically encoded sensors. Science 360:eaat4422 Sun F et al (2018) A Genetically Encoded Fluorescent Sensor Enables Rapid and Specific Detection of Dopamine in Flies, Fish, and Mice. Cell 174:481–496e19 Banerjee A, Lee J, Nemcova P, Liu C, Kaeser P (2020) S. Synaptotagmin-1 is the Ca2 + sensor for fast striatal dopamine release. eLife 9, e58359 Delignat-Lavaud B et al (2023) Synaptotagmin-1-dependent phasic axonal dopamine release is dispensable for basic motor behaviors in mice. Nat Commun 14:4120 Lebowitz JJ et al (2024) Synaptotagmin-7 Counteracts Short-Term Depression during Phasic Dopamine Release. eNeuro 11 Liu C, Kershberg L, Wang J, Schneeberger S, Kaeser PS (2018) Dopamine Secretion Is Mediated by Sparse Active Zone-like Release Sites. Cell 172:706–718 Banerjee A et al (2022) Molecular and functional architecture of striatal dopamine release sites. Neuron 110:248–265e9 Pereira DB et al (2016) Fluorescent false neurotransmitter reveals functionally silent dopamine vesicle clusters in the striatum. Nat Neurosci 19:578–586 Wildenberg G et al (2021) Partial connectomes of labeled dopaminergic circuits reveal non-synaptic communication and axonal remodeling after exposure to cocaine. eLife 10:e71981 Zuber B, Lučić V (2022) Neurons as a model system for cryo-electron tomography. J Struct Biology: X 6:100067 Tritsch NX, Sabatini BL (2012) Dopaminergic Modulation of Synaptic Transmission in Cortex and Striatum. Neuron 76:33–50 Bamford NS et al (2004) Heterosynaptic Dopamine Neurotransmission Selects Sets of Corticostriatal Terminals. Neuron 42:653–663 Arluison M, Dietl M, Thibault J (1984) Ultrastructural morphology of dopaminergic nerve terminals and synapses in the striatum of the rat using tyrosine hydroxylase immunocytochemistry: a topographical study. Brain Res Bull 13:269–285 Descarries L, Watkins KC, Garcia S, Bosler O, Doucet G (1996) Dual character, asynaptic and synaptic, of the dopamine innervation in adult rat neostriatum: a quantitative autoradiographic and immunocytochemical analysis. J Comp Neurol 375:167–186 Uchigashima M, Ohtsuka T, Kobayashi K, Watanabe M (2016) Dopamine synapse is a neuroligin-2–mediated contact between dopaminergic presynaptic and GABAergic postsynaptic structures. Proc. Natl. Acad. Sci. U.S.A. 113, 4206–4211 Moss J, Bolam JP (2008) A Dopaminergic Axon Lattice in the Striatum and Its Relationship with Cortical and Thalamic Terminals. J Neurosci 28:11221–11230 Beyene AG et al (2019) Imaging striatal dopamine release using a nongenetically encoded near infrared fluorescent catecholamine nanosensor. Sci Adv 5:eaaw3108 Yagishita S et al (2014) A critical time window for dopamine actions on the structural plasticity of dendritic spines. Science 345:1616–1620 Paget-Blanc V et al (2022) A synaptomic analysis reveals dopamine hub synapses in the mouse striatum. Nat Commun 13:3102 Herzog E et al (2001) The Existence of a Second Vesicular Glutamate Transporter Specifies Subpopulations of Glutamatergic Neurons. J Neurosci 21:RC181–RC181 Sartori A et al (2007) Correlative microscopy: Bridging the gap between fluorescence light microscopy and cryo-electron tomography. J Struct Biol 160:135–145 Schwartz CL, Sarbash VI, Ataullakhanov FI, Mcintosh JR, Nicastro D (2007) Cryo-fluorescence microscopy facilitates correlations between light and cryo-electron microscopy and reduces the rate of photobleaching. J Microsc 227:98–109 Fernandez-Busnadiego R et al (2010) Quantitative analysis of the native presynaptic cytomatrix by cryoelectron tomography. J Cell Biol 188:145–156 Tao C-L et al (2018) Differentiation and Characterization of Excitatory and Inhibitory Synapses by Cryo-electron Tomography and Correlative Microscopy. J Neurosci 38:1493–1510 Nicholls DG, Sihra TS (1986) Synaptosomes possess an exocytotic pool of glutamate. Nature 321:772–773 Herzog E et al (2011) In Vivo Imaging of Intersynaptic Vesicle Exchange Using VGLUT1Venus Knock-In Mice. J Neurosci 31:15544–15559 Turiault M et al (2007) Analysis of dopamine transporter gene expression pattern – generation of DAT-iCre transgenic mice. FEBS J 274:3568–3577 Madisen L et al (2010) A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci 13:133–140 Verhage M et al (1991) Differential release of amino acids, neuropeptides, and catecholamines from isolated nerve terminals. Neuron 6:517–524 Daniel JA, Malladi CS, Kettle E, McCluskey A, Robinson PJ (2012) Analysis of synaptic vesicle endocytosis in synaptosomes by high-content screening. Nat Protoc 7:1439–1455 Mastronarde DN, Held SR (2017) Automated tilt series alignment and tomographic reconstruction in IMOD. J Struct Biol 197:102–113 Martinez-Sanchez A et al (2021) Trans-synaptic assemblies link synaptic vesicles and neuroreceptors. Sci Adv 7:eabe6204 Lučić V, Yang T, Schweikert G, Förster F, Baumeister W (2005) Morphological Characterization of Molecular Complexes Present in the Synaptic Cleft. Structure 13:423–434 Korogod N, Petersen CC, Knott GW (2015) Ultrastructural analysis of adult mouse neocortex comparing aldehyde perfusion with cryo fixation. Elife 4:e05793 Chuhma N (2004) Dopamine Neurons Mediate a Fast Excitatory Signal via Their Glutamatergic Synapses. J Neurosci 24:972–981 Tritsch NX, Ding JB, Sabatini BL (2012) Dopaminergic neurons inhibit striatal output through non-canonical release of GABA. Nature 490:262–266 Rizzoli SO, Betz WJ (2005) Synaptic vesicle pools. Nat Rev Neurosci 6:57–69 Schikorski T, Stevens CF (2001) Morphological correlates of functionally defined synaptic vesicle populations. Nat Neurosci 4:391–395 Watanabe S et al (2013) Ultrafast endocytosis at mouse hippocampal synapses. Nature 504:242–247 Schrod N et al (2018) Pleomorphic linkers as ubiquitous structural organizers of vesicles in axons. PLoS ONE 13:e0197886 Lučić V, Fernández-Busnadiego R, Laugks U, Baumeister W (2016) Hierarchical detection and analysis of macromolecular complexes in cryo-electron tomograms using Pyto software. J Struct Biol 196:503–514 Schikorski T, Stevens CF (1997) Quantitative ultrastructural analysis of hippocampal excitatory synapses. J Neurosci 17:5858–5867 Siksou L et al (2007) Three-Dimensional Architecture of Presynaptic Terminal Cytomatrix. J Neurosci 27:6868–6877 Zuber B, Lučić V (2019) Molecular architecture of the presynaptic terminal. Curr Opin Struct Biol 54:129–138 Varoqueaux F et al (2002) Total arrest of spontaneous and evoked synaptic transmission but normal synaptogenesis in the absence of Munc13-mediated vesicle priming. Proceedings of the National Academy of Sciences 99, 9037–9042 Lycas MD, Morado DR, Gether U, Briggs JAG, Erlendsson S (2024) Ultrastructural Dynamics of Dopaminergic Presynaptic Release Sites revealed by Cryo-correlative Light and Electron Microscopy. 04.15.589543 Preprint at https://doi.org/10.1101/2024.04.15.589543 (2024) Fujise K, Rosenfeld MS, Rafiq NM (2024) Synaptic vesicle characterization of iPSC-derived dopaminergic neurons provides insight into distinct secretory vesicle pools. 02.22.581435 Preprint at https://doi.org/10.1101/2024.02.22.581435 (2024) Uchizono K (1965) Characteristics of Excitatory and Inhibitory Synapses in the Central Nervous System of the Cat. Nature 207:642–643 Valdivia O (1971) Methods of fixation and the morphology of synaptic vesicles. J Comp Neurol 142:257–273 Siksou L et al (2013) A role for vesicular glutamate transporter 1 in synaptic vesicle clustering and mobility. Eur J Neurosci 37:1631–1642 van Oostrum M et al (2023) The proteomic landscape of synaptic diversity across brain regions and cell types. Cell 0 Zhang S et al (2015) Dopaminergic and glutamatergic microdomains in a subset of rodent mesoaccumbens axons. Nat Neurosci 18:386–392 De-Smedt-Peyrusse V, Darriet L, Trifilieff P, Herzog E, Angelo MF (2018) Subcellular Fractionation of Brain Tissue from Small Tissue Explants. In: Murphy KM (ed) Synaptosomes. Springer, New York, NY, pp 75–84. doi: 10.1007/978-1-4939-8739-9_5 . Schindelin J et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682 Mastronarde DN (2005) Automated electron microscope tomography using robust prediction of specimen movements. J Struct Biol 152:36–51 Tang G et al (2007) EMAN2: an extensible image processing suite for electron microscopy. J Struct Biol 157:38–46 de la Rosa-Trevín JM et al (2016) Scipion: A software framework toward integration, reproducibility and validation in 3D electron microscopy. J Struct Biol 195:93–99 Kremer JR, Mastronarde DN, McIntosh JR (1996) Computer Visualization of Three-Dimensional Image Data Using IMOD. J Struct Biol 116:71–76 Wadell H, Volume (1935) Shape, and Roundness of Quartz Particles. J Geol 43:250–280 Harris CR et al (2020) Array programming with NumPy. Nature 585:357–362 Virtanen P et al (2020) SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat Methods 17:261–272 Hunter JD, Matplotlib (2007) A 2D Graphics Environment. Comput Sci Eng 9:90–95 Tables Tables 1 to 3 are available in the Supplementary Files section. Additional Declarations There is NO Competing Interest. Supplementary Files Lapiosetal.suppv22.docx Supplementary Figures and Tables Cite Share Download PDF Status: Published Journal Publication published 13 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6081416","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":435736314,"identity":"9aa84ec4-fa8e-4e4d-8b7a-a8f26114c0d3","order_by":0,"name":"David Perrais","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-5878-5408","institution":"Interdisciplinary Institute for NeuroScience","correspondingAuthor":true,"prefix":"","firstName":"David","middleName":"","lastName":"Perrais","suffix":""},{"id":435736315,"identity":"7cf3af76-e05b-4a73-b2bd-7c1ca357d30b","order_by":1,"name":"Paul Lapios","email":"","orcid":"","institution":"Centre national de la recherche scientifique","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Lapios","suffix":""},{"id":435736316,"identity":"db584898-322c-47c2-8216-d092b8fd5b7a","order_by":2,"name":"Robin Anger","email":"","orcid":"https://orcid.org/0009-0008-1448-2019","institution":"European Institute for Chemistry and Biology, IECB","correspondingAuthor":false,"prefix":"","firstName":"Robin","middleName":"","lastName":"Anger","suffix":""},{"id":435736317,"identity":"e20e5e16-db5a-4e41-a31b-8a058346a44f","order_by":3,"name":"Vincent Paget-Blanc","email":"","orcid":"https://orcid.org/0000-0001-8773-3538","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Vincent","middleName":"","lastName":"Paget-Blanc","suffix":""},{"id":435736318,"identity":"9c4c78e8-9ae6-4439-b7c9-8f559f9ef7e4","order_by":4,"name":"Esther Marza","email":"","orcid":"","institution":"European Institute for Chemistry and Biology, IECB","correspondingAuthor":false,"prefix":"","firstName":"Esther","middleName":"","lastName":"Marza","suffix":""},{"id":435736319,"identity":"bbd04e4e-8163-4e65-b470-eba9cb79f8bb","order_by":5,"name":"Vladan Lučić","email":"","orcid":"https://orcid.org/0000-0003-3698-7436","institution":"Max Planck Institute of Biochemistry","correspondingAuthor":false,"prefix":"","firstName":"Vladan","middleName":"","lastName":"Lučić","suffix":""},{"id":435736320,"identity":"bda9ca76-46b2-4fcc-9874-89e1f8ccc8e2","order_by":6,"name":"Remi Fronzes","email":"","orcid":"","institution":"European Institute for Chemistry and Biology, IECB","correspondingAuthor":false,"prefix":"","firstName":"Remi","middleName":"","lastName":"Fronzes","suffix":""},{"id":435736321,"identity":"2b34a038-675d-4f08-adfb-9861aa35e43c","order_by":7,"name":"Etienne Herzog","email":"","orcid":"https://orcid.org/0000-0002-0058-6959","institution":"Centre national de la recherche scientifique","correspondingAuthor":false,"prefix":"","firstName":"Etienne","middleName":"","lastName":"Herzog","suffix":""}],"badges":[],"createdAt":"2025-02-21 17:41:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6081416/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6081416/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-66355-x","type":"published","date":"2025-12-13T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80999543,"identity":"248b2389-b85f-4331-80f9-94cb08b71f4f","added_by":"auto","created_at":"2025-04-21 06:01:37","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":752564,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCryo-CLEM and cryo-ET of cortico-striatal and dopaminergic synaptosomes.\u003c/strong\u003e \u003cstrong\u003eA\u003c/strong\u003e, \u003cem\u003eLeft\u003c/em\u003e, schematic of the expected organization between DA varicosities (magenta) and GLU synapses (green), forming DHSs. \u003cem\u003eRight top\u003c/em\u003e, VGLUT1\u003csup\u003evenus\u003c/sup\u003e fluorescence in a sagittal slice. CS synapses generate an intense fluorescence signal in the striatum (dotted line). \u003cem\u003eRight bottom\u003c/em\u003e, DA projections produce a bright striatal fluorescence in a coronal slice of a DAT-Cre mouse injected with the AAV-Flex-mNeonGreen (magenta). Scale bars: 3 mm \u003cstrong\u003eB\u003c/strong\u003e, Scheme of synaptosomes preparation from fresh striatum obtained after differential centrifugations. The fractions (H, S1, P2, B, see Methods) are increasingly enriched in synaptosomes. After discontinuous density gradient centrifugation, synaptosomes are diluted and centrifuged to remove Ficoll. \u003cstrong\u003eC\u003c/strong\u003e, A suspension of synaptosomes mixed with fluorescent and gold fiducial beads is layered on the grid, blotted and plunge-frozen in liquid ethane. \u003cstrong\u003eD\u003c/strong\u003e, Example of a cryo-fluorescence microscopy image of a single grid square from a VGLUT1\u003csup\u003e venus \u003c/sup\u003ex DAT-Cre x Ai14-tdTomato with 4 channels: bright field, fluorescence of tdTomato, Venus and blue fiducial beads. These images were used to select synaptosomes of interest, green (VGLUT1\u003csup\u003evenus\u003c/sup\u003e) or magenta (DAT+) next to several blue beads, for observation with electron microscopy. Scale bar 10 µm. \u003cstrong\u003eE\u003c/strong\u003e, Left cryo-EM image of the same square as displayed in D. Scale bar 10 µm. Right, subset of cryo-fluorescence (top) and EM (bottom) of the same portion of grid with 3 fiducial blue beads visible in both modalities (yellow circles) for fine registration. Scale bar: 3 µm. Synaptosomes with clearly visible beads and good ice quality were further imaged with high magnification tilted series. \u003cstrong\u003eF\u003c/strong\u003e, Reconstructed tomograms of fluorescent synaptosomes obtained by cryo-electron tomography. \u003cem\u003eLeft\u003c/em\u003e, overlay of fluorescence and bright field channels showing tagged synaptosomes of interest and fiducial beads. Scale bars: 5 µm. \u003cem\u003eRight\u003c/em\u003e, representative single tomographic slice of 1.558 nm of thickness showing a clear structure outline and intracellular organelles. Magenta or green cross marks point to the registered centre of the fluorescent tag (GLU or DA synaptosome), Scale bars: 500 nm. \u003cstrong\u003eG\u003c/strong\u003e, 3D models from the synaptosomes displayed in F obtained by segmentation of membranes. The display of 3D models are as follows: plasma membrane of synaptosomes in green for GLU or magenta for DA; Small (synaptic) vesicles in blue; elongated endoplasmic-reticulum-like structures in orange; large endosome-like organelles in yellow; mitochondrion in magenta. For both GLU synaptosomes, a clear synaptic cleft was detected, with either a closed post-synaptic element (PSE) in grey (top), or an opened post-synaptic membrane (bottom). Both of them exhibit a post-synaptic density. The presynaptic GLU active zones are shown in red. Scale bar: 500 nm\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/43bf88ab3f7739d4972a29f6.png"},{"id":80997597,"identity":"ff1123d9-d32d-4de5-9112-7f0b5f1a7925","added_by":"auto","created_at":"2025-04-21 05:37:37","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":724541,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of synaptosome size and vesicular content for GLU and DA synaptosomes. A,B\u003c/strong\u003e. Example of a single tomogram plane of 1.558 nm thickness, \u0026nbsp;and corresponding to 3D models of GLU (A) and DA (B) synaptosomes. Same color code as in Figure 1. In addition, multivesicular bodies are visible in the tomogram slices (blue arrows) and in the models (dark blue). Scale bars 500 nm \u003cstrong\u003eC-F\u003c/strong\u003e, Distributions and average diameters (C), visible volumes (D) number of vesicles (E) and vesicle density (F) of 101 GLU and 94 DA synaptosomes. The distributions are all different (t test p \u0026lt; 0.0001). \u003cstrong\u003eG\u003c/strong\u003e, Cumulative distribution of vesicle diameters in GLU (green line, n = 18897) and DA (magenta line, n = 2782) synaptosomes. The diameters are significantly different (Kolmogorov-Smirnov test p \u0026lt; 0.0001). The green dotted line (n = 650) corresponds to vesicles in DA terminals with a PSE sharing the features of asymmetric synapses (examples in S6). \u003cstrong\u003eH\u003c/strong\u003e, Cumulative distribution of vesicle sphericity (WI). GLU vesicles are significantly more spheric than DA vesicles (Kolmogorov-Smirnov test p \u0026lt; 0.0083). We show examples of DA vesicles with WI = 0.85, 0.89, 0.95 and 0.99 from left to right above the plot. \u003cstrong\u003eI\u003c/strong\u003e,\u003cstrong\u003eJ\u003c/strong\u003e, Cropped zoomed in tomograms of GLU (I) and DA (J) synaptosomes display the variety of vesicle sizes and shapes. Scale bars 100 nm.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/2e8336fd0d923cd6d94fc97c.png"},{"id":80997601,"identity":"ebf5d95f-996f-4ca4-9252-963f4635c80f","added_by":"auto","created_at":"2025-04-21 05:37:37","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":898936,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eContact zones between GLU, DA synaptosomes and PSEs. A-C\u003c/strong\u003e, 3D models of a GLU presynaptic elements (green) connected to a PSE (grey). The active zones are drawn in red. In B, a DA element connects the GLU presynaptic element. In C, a DA element connects the PSE. The contact zones with DA are drawn in yellow. \u003cstrong\u003eD\u003c/strong\u003e, single plane showing the synaptic cleft between a GLU synaptosome (outlined in green) and a PSE (outlined in grey). The electron dense material in the middle of the cleft is clearly visible (red arrows). \u003cstrong\u003eE\u003c/strong\u003e, Pixel intensities along the length of the synaptic cleft, normalized to area outside the synaptic cleft (n = 20). \u003cstrong\u003eF\u003c/strong\u003e, Area of active zones of GLU synaptosomes and contact zone between DA and GLU synaptosomes or DA and PSEs. \u003cstrong\u003eG\u003c/strong\u003e, Membrane to membrane distance of the cleft of: GLU with PSE (synaptic cleft), DA with GLU synaptosomes, and DA with PSE. \u003cstrong\u003eH\u003c/strong\u003e, Single plane showing the cleft between a DA (magenta) and a GLU synaptosome (green). \u003cstrong\u003eI\u003c/strong\u003e, Pixel intensities along the length of the DA/GLU contact site, normalized to region outside the contact site (n = 24). \u003cstrong\u003eJ\u003c/strong\u003e, Image of a contact site (synaptic cleft) between GLU and PSE. \u003cstrong\u003eK\u003c/strong\u003e, Normalized intensity profiles in the cytoplasm starting 5 nm away from the plasma membrane, towards the center of GLU (green) and PSE (grey). In both cases, the averaged intensity reached a plateau around 80% 40 nm away. In the PSE, the intensity clearly increases 20 nm away, which corresponds to the PSD. We observe no clear increase on the presynaptic side of GLU. \u003cstrong\u003eL-M\u003c/strong\u003e, Same as J-K for DA-GLU contact sites. \u003cstrong\u003eN-O\u003c/strong\u003e, Same as J-K for DA-PSE contact sites.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/c88169c2a0325e8562f4daba.png"},{"id":80997599,"identity":"97d0922d-eb42-4988-8adb-c6ebc9a22e3e","added_by":"auto","created_at":"2025-04-21 05:37:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":578002,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVesicles with tethers and connectors in GLU and DA synaptosomes. A\u003c/strong\u003e, Tomogram planes showing vesicles tethered to the plasma membrane in GLU synaptosomes (white arrowheads). Some vesicles are connected via multiple tethers. Connectors between vesicles are also visible (light blue arrowheads). Scale bar 100 nm \u003cstrong\u003eB\u003c/strong\u003e, Same as A for DA synaptosomes. \u003cstrong\u003eC\u003c/strong\u003e, Percentage of proximal vesicles (located less than 45 nm) connected by at least one tether in GLU and DA synapses (t-test; p-value = 0,133). \u003cstrong\u003eD\u003c/strong\u003e, Length of tethers connecting vesicles to the plasma membrane in GLU and DA synapses (t-test; p-value \u0026lt; 0.001). \u003cstrong\u003eE\u003c/strong\u003e, Average number of tethers for vesicles at various distances from the plasma membrane (t-tests; p-values = 0.934; 0.777; 0.770). \u003cstrong\u003eF\u003c/strong\u003e, Fraction of volume occupied by vesicles vs distance to the AZ for GLU synapses. The mean distribution is shown as a thick line. \u003cstrong\u003eG\u003c/strong\u003e, 3D model of a GLU synaptosome around the active zone showing the concentration of proximal vesicles that are tethered.\u003cstrong\u003eH\u003c/strong\u003e, Fraction of volume occupied by vesicles vs distance to the plasma membrane for DA synapses. The mean distribution is shown as a thick line. \u003cstrong\u003eI\u003c/strong\u003e, 3D model of a DA synaptosome with the zone of proximal vesicles. \u003cstrong\u003eJ\u003c/strong\u003e, Percentage of all vesicles connected to another in GLU and DA synaptosomes (t-test; p-value = 0.0004). \u003cstrong\u003eK\u003c/strong\u003e, Average length of connectors (t-test; p-value = 0.0021). \u003cstrong\u003eL\u003c/strong\u003e, Proportion of proximal vesicles that are tethered and/or connected in GLU and DA synaptosomes (t-test; p-value = 0.0008).\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/fdb0437c3d40c47f055a6bbc.png"},{"id":80999554,"identity":"947f290c-e35d-4062-98b4-c8742e92405d","added_by":"auto","created_at":"2025-04-21 06:02:10","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":428544,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of tethered vesicles in DA synaptosomes. A\u003c/strong\u003e, 3D models of DA synaptosomes with (T+, left) and without (T-, right) tethered vesicles. Scale bars, 500 nm. Color coding as in Figure 1. \u003cstrong\u003eB\u003c/strong\u003e, Total number of vesicles in DA synaptosomes with (T+) or without (T-) tethered SV (Mann-Whitney; p-value = 0.0068).\u003cstrong\u003e C\u003c/strong\u003e, Diameter of DA synaptosomes with (T+) or without (T-) tethered vesicles (Mann-Whitney; p-value = 0.0401). \u003cstrong\u003eD\u003c/strong\u003e, Number of tethered vesicles per T+ DA synaptosomes.\u003cstrong\u003e E\u003c/strong\u003e, 3D model of a DA synaptosome with a cluster of 3 tethered SVs on one side of the plasma membrane. Scale bar 500 nm. On the right, detail of the synaptosome with the 3 tethered vesicles and corresponding tomogram images. Scale bar 50 nm. \u003cstrong\u003eF\u003c/strong\u003e, Nearest neighbour distance between tethers on the plasma membrane. \u003cstrong\u003eG\u003c/strong\u003e, Number of vesicles per µm\u003csup\u003e2\u003c/sup\u003e of plasma membrane at increasing distances from the plasma membrane for T+ and T- DA synaptosomes (t-tests; p-values = 0.0266; 0.131; 0.159; 0.697).\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/4ae939552c5d1804c080c5f9.png"},{"id":80999553,"identity":"8c28f569-9961-4ffc-ad01-27eb229ee610","added_by":"auto","created_at":"2025-04-21 06:02:07","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":732487,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of GLU synaptosomes connected or not to a DA terminal. A-B\u003c/strong\u003e, 3D models of a DA- GLU synaptosome (A) and a DA+ GLU synaptosome or DHS (B). Scale bars 500 nm. \u003cstrong\u003eC-D\u003c/strong\u003e, Number (C) (Mann-Whitney; p-value = 0.864) and density (D) of vesicles in DA- and DA+ GLU synaptosomes (Mann-Whitney; p-value = 0.577) \u003cstrong\u003eE\u003c/strong\u003e, Active zone area of DA- and DA+ GLU synaptosomes (t-test; p-value = 0.497). \u003cstrong\u003eF\u003c/strong\u003e, Average fraction of volume occupied by SVs vs distance to the plasma membrane for DA- and DA+ GLU synaptosomes (t-tests; p-values: 0 to 45 nm = 0.0215; 75 to 150 nm = 0.0195; 150 to 250 nm = 0.0016) . \u003cstrong\u003eG\u003c/strong\u003e, 3D models of DA- and DA+ synaptosomes showing the number of proximal vesicles at the active zone. \u003cstrong\u003eH\u003c/strong\u003e, Percentage of proximal vesicles (\u0026lt; 45 nm from plasma membrane) that are tethered in DA- and DA+ GLU synaptosomes (t-test; p-value = 0.1602). \u003cstrong\u003eI\u003c/strong\u003e, Number of tethers per vesicle at various distances from the active zone (t-tests; p-values: \u0026lt;5 nm = 0.4638; 5 to 10 nm = 0.5421; \u0026gt;10 nm = 0.4816). \u003cstrong\u003eJ\u003c/strong\u003e¸ Percentage of proximal vesicles that are connected to other vesicles in DA- and DA+ GLU synaptosomes (t-test; p-value = 0.0167). \u003cstrong\u003eK\u003c/strong\u003e, Proportion of proximal vesicles that are tethered and/or connected in DA- and DA+ GLU synaptosomes. The proportion of tethered and connected vesicles is significantly higher in DA+ GLU synaptosomes (chi-squared test; p-value = 0.0435). \u003cstrong\u003eL\u003c/strong\u003e, 3D models of DA- and DA+ GLU synaptosomes. In the DA- GLU synaptosome, a single SV is tethered and not connected to other vesicles. In the DA+ GLU synaptosome, four SVs are tethered and connected to one more vesicle.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/ad9d592caa4be543e900ea2a.png"},{"id":99212706,"identity":"cc969899-9521-45b6-935a-c7fd45e09736","added_by":"auto","created_at":"2025-12-30 08:27:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5290749,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/0e633914-9f24-466b-8da2-58b915c8db87.pdf"},{"id":80997603,"identity":"312370c9-7597-4763-9826-3bdb9f53aa5e","added_by":"auto","created_at":"2025-04-21 05:37:37","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5970489,"visible":true,"origin":"","legend":"Supplementary Figures and Tables","description":"","filename":"Lapiosetal.suppv22.docx","url":"https://assets-eu.researchsquare.com/files/rs-6081416/v1/410dfea70ba176b08fd79aff.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Cryo-correlative light and electron tomography of dopaminergic axonal varicosities reveals non-synaptic modulation of cortico-striatal synapses","fulltext":[{"header":"Introduction","content":"\u003cp\u003eNeuromodulation adjusts network activity and has major impacts on behavior. Among neuromodulators, dopamine acts in the basal ganglia network to encode reward prediction and participate to the initiation of movement. Dopaminergic (DA) projections to basal ganglia originate from two midbrain nuclei, the substantia nigra and the ventral tegmental area. These projections densely innervate the striatum to regulate the activity of spiny projections neurons (SPNs), which are central to functions like motor control, reward prediction and motivation \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Yet, at the ultrastructural level, the organisation of DA transmission is not clear. This situation is in stark contrast with the detailed characterization of neurotransmission machineries at forebrain glutamatergic (GLU) synapses.\u003c/p\u003e \u003cp\u003eAt GLU terminals, synaptic vesicles (SVs) are organized in a cluster polarized towards a portion of the plasma membrane called the active zone (AZ) which faces the post-synaptic density (PSD). The AZ contains specific proteins such as RIM1/2, bassoon and ELKS \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Prior to fusion, proximal SVs follow a series of steps, which can be observed with cryo-electron tomography (cryo-ET), from initial tethering, where SVs are tethered to the plasma membrane by one filament of approximately 10\u0026ndash;25 nm \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, to the formation of multiple short tethers, likely comprising the SNARE complex, which brings SVs closer than 5 nm to the plasma membrane \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. These steps strongly depend on the AZ proteins RIM1 and Munc13a, which are known to control fast SV exocytosis \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Thus, the presence of docked and primed vesicles may constitute a hallmark of readily releasable vesicles in axons.\u003c/p\u003e \u003cp\u003eDopamine is also exocytosed from vesicles within milliseconds after stimulation \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This process relies on the calcium sensor Synaptotagmin-1 \u003csup\u003e9,10\u003c/sup\u003e. Dopamine release shows strong paired pulse depression which is in part controlled by Synaptotagmin 7 \u003csup\u003e11\u003c/sup\u003e. Moreover, DA axons contain AZ proteins, RIM1, Munc13 and ELKS, which are important for fast dopamine release \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. However, only\u0026thinsp;~\u0026thinsp;30% of DA varicosities contain such assemblies \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. This observation is supported by functional evidence showing that around a quarter of dopaminergic varicosities are active, as assessed by release of fluorescent neurotransmitter analogues \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Furthermore, serial electron microscopy studies of dopaminergic axon terminals revealed heterogeneous vesicular content, with terminals containing varying combinations of small and large vesicles, while some terminals appeared to lack vesicles altogether \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Overall, DA terminals exhibit a non-stereotyped vesicular organization which may explain the functional diversity of release. However, observations of the accurate spatial arrangement of DA vesicles prior to fusion has been hampered by chemical fixation, staining procedures and difficulties to identify rare DA axons which negatively affects the preservation of cellular morphology and precludes interpretation of molecular details \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe relationship between DA release sites and target cells remains poorly documented. Upon release, dopamine binds to G-protein coupled receptors of the D1 group (D1/5R) or D2 group (D2-4R) whose signals respectively increase or decrease the excitability of target cells. On pre-synaptic terminals they influence SV release probability, while at the post-synapse they act on ion channels and glutamate receptors (reviewed in \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e). In particular, neurons projecting from motor and prefrontal cortical regions form GLU cortico-striatal (CS) synapses responsible for the activation of SPNs. Interestingly, the stimulation of dopamine release in acute slices attenuates the release kinetic from a subset but not all GLU terminals \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, showing that dopamine influences the activity of CS synapses. However, the precise ultrastructure through which dopaminergic terminals interact with their synaptic targets is unclear. DA synapses showing classical pre and post-synaptic features represent a minority \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Nevertheless, DA boutons are often found in close apposition with either pre or post-synaptic elements of CS synapses \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. This proximity is functionally relevant, because DA release generates 2 \u0026micro;m wide hotspots of dopamine as estimated using carbon nanotube sensors \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Moreover, the synchronisation of DA phasic release with local glutamate uncaging induces structural plasticity at spines \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIsolation of DA synaptosomes from striatal tissue and sorting by fluorescence activated synaptosome sorting (FASS) revealed that most DA terminals are in close contact with other terminals, such as GLU presynapses \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. These interactions are conserved even though tissue homogenisation and droplet-based fluorescence activated sorting exposed structures to significant mechanical shearing forces. We termed these multipartite DA containing synapses \u0026ldquo;dopamine hub synapses\u0026rdquo; (DHS). Among them, 25% were formed with CS synapses marked by the vesicular glutamate transporter VGLUT1 \u003csup\u003e22,25,26\u003c/sup\u003e. Importantly, CS-DHS contain an increased signal for the presynaptic proteins VGLUT1 and bassoon compared to other CS synapses. Collectively, these findings highlight the importance of a local dopaminergic signalling for the regulation of CS synapses. However, there is still a major lack of ultrastructural observations in close-to-native conditions of DA terminals associated with their target synapses.\u003c/p\u003e \u003cp\u003eCryo-correlative light and electron microscopy (cryo-CLEM) and cryo-electron tomography (cryo-ET) enable identification of vitrified fluorescently labeled terminals and a close-to-native 3D observations of cellular ultrastructure and protein complexes at a nanometer scale \u003csup\u003e\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. This allows determination of the vesicular organisation in different synaptic types and characterization of protein complexes \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Synaptosomes are a suitable model for cryo-EM because they can be vitrified by plunge freezing, imaged in transmission EM without thinning, and because they preserve association between terminals and important pre- and postsynaptic function \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Here, we applied cryo-CLEM and cryo-ET to fluorescently labeled synaptosomes in order to determine the ultrastructural features of DA terminals, CS synapses and CS-DHS. We reveal the spatial organization of DA vesicles, as well as the structural association with GLU synapses in DHS. Importantly, we observed tethered SVs in DA terminals and quantified differences in SV organization of CS that are correlated to the association with DHS.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eIdentification and observation of GLU and DA synaptosomes by cryo-CLEM and cryo-ET\u003c/h2\u003e \u003cp\u003eWe assessed the ultrastructure of identified glutamatergic (GLU) and dopaminergic (DA) terminals extracted from the striatum of adult mice using cryo-CLEM combined with cryo-ET as described in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. We prepared synaptosomes from the striata of 12 to 18 week-old mice as previously described \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e and detailed in Methods and Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA,B. We used 3 mouse lines. For GLU elements we took advantage of the knock-in (KI) mouse line in which the VGLUT1 open reading frame is tagged with the sequence of the fluorescent protein Venus \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In the striatum, VGLUT1\u003csup\u003evenus\u003c/sup\u003e specifically labels cortico-striatal terminals \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. For DA synaptosomes we injected an adeno associated viral vector carrying sequences for Cre-dependent mNeonGreen expression (AAV1 pCAG-Flex-mNeonGreen) in the midbrain of dopamine transporter promoter (DAT)-Cre transgenic mice, which specifically labels dopaminergic neurons \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. To identify CS-DHS we used dual tagging of GLU (green) and DA (red) synaptosomes by crossing VGLUT1\u003csup\u003evenus\u003c/sup\u003e mice with DAT-Cre \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e and the reporter line Ai14-tdTomato \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Striatal synaptosomes were incubated at 37\u0026deg;C for 15 minutes before use. They are capable of depolarization-evoked exocytosis and recycling, as shown by uptake and release of the membrane dye FM4-64 (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), in accordance with previous work \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. This suspension was mixed with electron dense fluorescent fiducial beads for alignment and gold beads for tomogram reconstruction. The mixture was applied on an EM grid, plunge frozen into liquid ethane (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC) and kept in liquid nitrogen for further use.\u003c/p\u003e \u003cp\u003eWe first observed the grids with a cryo-fluorescence microscope. We selected isolated fluorescent spots with nearby fiducial beads for alignment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). In mice with both fluorescent markers for GLU and DA synaptosomes, we selected either an individual isolated spot, or pairs of spots where the two colours were separated by less than 1 \u0026micro;m corresponding to putative DHS. We then observed the same grids with a cryo-transmission electron microscope (Talos Arctica), and located regions of interest based on bead patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). We used the images of neighbouring independent fiducial beads to measure the pointing precision between cryo-fluorescence and cryo-electron microscopy (Figure S2; 117\u0026thinsp;\u0026plusmn;\u0026thinsp;82 nm, n\u0026thinsp;=\u0026thinsp;64). This pointing precision is smaller than the radii of GLU and DA synaptosomes (see below). Therefore, we can identify unambiguously the terminals of interest in cryo-EM, either GLU or DA. Finally, we acquired tilt series of selected synaptosomes and reconstructed tomograms with weighted back projection method using IMOD \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. We obtained tomograms from 19 preparations (Table\u0026nbsp;1): 4 from VGLUT1-Venus mice (39 GLU synaptosomes), 5 from DAT-Cre\u0026thinsp;+\u0026thinsp;AAV-mNeonGreen mice (43 DA synaptosomes) and 10 from VGLUT1-Venus*DAT-Cre-Ai14-tdTomato (62 GLU and 51 DA synaptosomes, among which 32 form DHSs). We thus have a dataset of tomograms of 101 GLU synaptosomes and 94 DA synaptosomes with 32 CS-DHSs (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF). We created 3D models of these synaptosomes, in which we segmented the plasma membrane of the synaptosome, internal membranes and, when applicable, identified adhering structures such as a post-synaptic element (PSE), as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eG.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eGLU synaptosomes are larger and contain more SVs than DA synaptosomes\u003c/h3\u003e\n\u003cp\u003eThe mean size of GLU synaptosomes, as measured by their maximal extension, is 823 nm, which corresponds to a visible volume of 0.0915 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA,C,D). They all contain small round SVs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Each synaptosome contains 13 to 872 vesicles, 191 on average (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE), for an average density of 1956 vesicles/\u0026micro;m\u003csup\u003e3\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF). In addition, 21/101 synaptosomes contained a mitochondrion (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA) and some contained large organelles or pleiotropic organelles such as round vacuoles, multivesicular bodies, as well as other intracellular features, such as filaments and clathrin-coated vesicles (Figure S3A-F). The majority of GLU presynaptic elements (64/101, 63%) are connected to a PSE, which also contains occasionally intracellular organelles (Figure S3G-I). A gallery of representative tomographic slices and 3D models of GLU synaptosomes is shown in Figure S4.\u003c/p\u003e \u003cp\u003eDA synaptosomes (example Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB) have a mean size of 575 nm, which corresponds to a visible volume of 0.0341 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, significantly smaller than GLU synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC,D, p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). DA synaptosomes contain on average 30 vesicles but their number is quite variable: some are almost empty (23/94 synaptosomes have less than 5 vesicles) while others contain tens or even hundreds of vesicles (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE and Figure S5, S6A-B). Overall, the vesicle density of DA synaptosomes is about 3-fold smaller (723 per \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e) than for GLU synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF). DA synaptosomes also contain other organelles, such as vacuoles, mitochondria, multivesicular bodies (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB, S6C-H). Table\u0026nbsp;2 reports the number of DA synaptosomes with these organelles. While SVs in GLU synaptosomes are spherical, small and uniform in size (outer diameter 39.8\u0026thinsp;\u0026plusmn;\u0026thinsp;5.8 nm, n\u0026thinsp;=\u0026thinsp;18 897, with only 2.7% vesicle larger than 60 nm), vesicles in DA synaptosomes are significantly larger (44.9\u0026thinsp;\u0026plusmn;\u0026thinsp;9.9 nm, n\u0026thinsp;=\u0026thinsp;2782, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) with 8.8% of vesicles having a diameter in the 60\u0026ndash;100 nm range (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG). In total, 66/94 DA synaptosomes have at least one vesicle bigger than 60 nm, with a proportion of 23.5\u0026thinsp;\u0026plusmn;\u0026thinsp;19.6% big vesicles in these synaptosomes. Moreover, vesicles in DA synaptosomes are often elongated or pleomorphic (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eH,J). We used Wadell\u0026rsquo;s index WI (see Methods) to quantify vesicle sphericity. Most vesicles in GLU terminals have an index close to 1 (perfect sphere). However, vesicles in DA synaptosomes are significantly less round than the ones in GLU synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eH, p-value\u0026thinsp;=\u0026thinsp;0.0083). These elongated vesicles are spread across the synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eJ, S5). Overall, there are 65/94 DA synaptosomes containing at least one elongated vesicle (WI\u0026thinsp;\u0026lt;\u0026thinsp;0.95) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eJ). In these synaptosomes there are 29.2\u0026thinsp;\u0026plusmn;\u0026thinsp;26.9% of elongated vesicles.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eContact zones between GLU, DA synaptosomes and PSEs\u003c/h3\u003e\n\u003cp\u003eIn GLU synaptosomes we identified a clear PSE separated from a presynaptic terminal by a clearly defined synaptic cleft in 64/101 (63%) synaptosomes. Typically, the PSE is defined by a sealed plasma membrane compartment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF,\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA,C, S4) but in some cases it had an open membrane adhering to the presynaptic terminal forming a synaptic cleft (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF,S4). The area of contact between GLU and PSE is characterized by roughly parallel membranes defining a wide cleft (31.8\u0026thinsp;\u0026plusmn;\u0026thinsp;6.2 nm, n\u0026thinsp;=\u0026thinsp;34) with dark material around the midline (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD,E), as observed previously \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. The active zone (AZ), defined as the membrane region of GLU synaptosome in contact with the PSE (outlined in red in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA), has an area of 0.079\u0026thinsp;\u0026plusmn;\u0026thinsp;0.055 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e (n\u0026thinsp;=\u0026thinsp;33) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF). We analysed the presence of protein density in the pre- and post-synaptic elements of GLU synaptosomes. The normalized densities decrease by 20% in both compartments (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eJ,K). However, we detected a clear increase in density 10 to 25 nm from the PSE membrane, which corresponds to the post-synaptic density (PSD), as detected with other modalities of electron microscopy for synapses in situ \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e and synaptosomes with cryo-ET \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. In contrast, only two out of 94 DA terminals were separated by a wide cleft from a characteristic PSE containing PSD, as in GLU synaptosomes. Interestingly, one of these DA synaptosomes has the most vesicles (610), all of them small and round, which makes it indistinguishable from classical GLU synaptosomes (Figure S6A). The other one has 40 vesicles (Figure S6B). In these two synaptosomes, the distribution of vesicle sizes and sphericity matches distributions seen for GLU synaptosomes (dotted line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG,H). Another DA synaptosome has 295 vesicles (see two upper dots in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE) and also resembles classical GLU synaptosomes, even though it was not connected to a PSE but engaged into a CS-DHS. These DA synaptosomes could originate from neurons co-expressing VGLUT2 and which are able to release glutamate and generate AMPAR-mediated post-synaptic potentials \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eNevertheless, many DA terminals are in close apposition with other structures which are occasionally resembling a GLU terminal (Figure S5A) or other, harder to characterize structures which could be parts of postsynaptic spines. To get a better insight on the features of contact zones between DA and GLU terminals, we analyzed 32 tomograms from apposed GLU and DA synaptosomes which correspond to CS-DHSs \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. The area of contact between GLU and DA synaptosomes have a similar size as GLU active zones (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB,F). The width of the cleft between GLU and DA terminals is 12.1\u0026thinsp;\u0026plusmn;\u0026thinsp;2.4 nm, much smaller than the synaptic cleft of GLU synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). Along this cleft, dense material is also observed, albeit not as pronounced as for GLU/PSE synaptic clefts (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH,I). The intracellular side of DA and GLU terminals around the adhesion site are devoid of visible densities such as PSDs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eL,M). In 2 tomograms, the DA terminal is not in contact with the presynaptic GLU synaptosome but with the PSE (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC,S7B). The contact area (0.062 and 0.066 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e) and cleft sizes (10.9 and 15.6 nm) are in the same range as for DA/preGLU contact sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF,G). Similarly, the densities around the contact site are in the same range, without signs of a PSD (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eN,O). More examples of DHS virtual planes and models are available in Figure S7.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eVesicles in GLU and DA terminals are tethered to the plasma membrane and are interconnected.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eSynaptic vesicles in GLU and GABAergic synapses are divided into functional pools \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. SVs which fuse first with the plasma membrane upon calcium stimulation comprise the readily releasable pool, which was defined morphologically by EM of chemically fixed, dehydrated samples as the vesicles which appear docked, that is in direct contact with the plasma membrane \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Cryo-ET revealed that in close to native state, SVs are not docked but are tethered to the plasma membrane via one or several electron dense filaments, as observed in synaptosomes and dissociated neuronal cultures \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. We found at least one tethered vesicle at the active zone in almost all quantified striatal GLU synaptosomes (24/26, 92%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Tethers, as well as connectors (bridges that interconnect SVs) were detected using an automated, template-free method by the hierarchical connectivity algorithm \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Among proximal vesicles, defined as those localized less than 45 nm from the plasma membrane, 40\u0026thinsp;\u0026plusmn;\u0026thinsp;7% have one or more tethers (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). These tethers have a length of 13.9\u0026thinsp;\u0026plusmn;\u0026thinsp;7.5 nm (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). The number of tethers increases as vesicles are located closer to the plasma membrane (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE). Remarkably, vesicles located less than 5 nm from the plasma membrane have on average 3 tethers, which corresponds to synaptic vesicles that are functionally primed for fusion \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn DA synaptosomes, we found clear examples of tethers between vesicles and the plasma membrane, as well as vesicles connected by more than one tether (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). We detected vesicles tethered to the plasma membrane in 39% (13/33) of DA synaptosomes. The percentage of tethered vesicles among proximal vesicles in DA is 25\u0026thinsp;\u0026plusmn;\u0026thinsp;6%, which was lower, but not significantly different from GLU synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). However, tether length is on average significantly larger in DA versus GLU (DA: 22.43 nm; std. 12.21; GLU: 13.89 nm; std: 7.47; p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD) likely a direct consequence of the complete absence of tethered DA vesicles localized less than 5 nm from the plasma membrane (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE). Importantly, for DA and GLU vesicles located further away, the number of tethers per SV is not different (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE).\u003c/p\u003e \u003cp\u003eAnother morphological hallmark of the polarized exocytosis of SV in cortical synaptosomes is the larger fraction of volume occupied by proximal synaptic vesicles towards the active zone \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. GLU synaptosomes showed a mean peak occupancy around 20 nm from the plasma membrane that is clearly visible in individual occupancy profiles (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF,G) and is similar to those observed for cortical synaptosomes \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. In DA synaptosomes, the occupancy lacked the peak in the proximal vesicles region, but occasional enrichment close to the whole plasma membrane was observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eH,I). However, the absence of a morphologically identifiable DA active zone, may have biased the measurement of occupancy, because it included a larger, possibly irrelevant volume in our analysis.\u003c/p\u003e \u003cp\u003eThe percentage of inter-connected SVs was significantly higher in GLU (53%) than in DA synaptosomes (20%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eJ) and the length of connectors was significantly different (DA: 18.50 nm; std. 9.18 and GLU: 16.33 nm; std. 13.07; p-value\u0026thinsp;=\u0026thinsp;0.0021) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eK). Finally, among proximal vesicles, there were more SVs that were both tethered and connected in GLU (17%) than in DA synaptosomes (6.4%). In the latter, the majority of vesicles were neither tethered nor connected (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eL).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eTethered SVs highlight a subpopulation of DA terminals\u003c/h3\u003e\n\u003cp\u003eWe compared DA synaptosomes that contain (T+) to ones that do not contain (T-) tethered vesicles (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). We excluded from the analysis the 2 DA synaptosomes with a PSE (see Figure S6AB). The T\u0026thinsp;+\u0026thinsp;DA synaptosomes contain significantly (twice) more SVs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB), and are significantly larger than the T- synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). On average, T\u0026thinsp;+\u0026thinsp;synaptosomes contain 2 tethered vesicles, and a maximum of 5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD). In T\u0026thinsp;+\u0026thinsp;DA terminals we do not observe a clear AZ, which corresponds in GLU synapses to the location facing the synaptic cleft where tethered synaptic vesicles concentrate and undergo exocytosis upon stimulation. We looked for a putative DA active zone in two ways. First, the active zone could locate at or right next to the contact zone with GLU terminals. However, tethered vesicles are almost never found at the contact site. Second, an active zone is expected to concentrate tethered vesicles. Therefore, we analyzed DA synaptosomes with multiple tethered vesicles. Among the 13 T\u0026thinsp;+\u0026thinsp;DA synaptosomes, 6 had multiple tethered vesicles, for a total of 19 tethered vesicles (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD,E). The shortest arc distance (i.e. staying on the plasma membrane) between these vesicles is on average 210 nm (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eF). This is smaller than the average distance between random points on the synaptosome plasma membrane (estimated at 567 nm, see Methods for derivation). This suggests that tethered DA vesicles may gather in a preferential zone of the plasma membrane, a putative active zone. Moreover, in T\u0026thinsp;+\u0026thinsp;DA synaptosomes, vesicles are located closer to the plasma membrane than in T- DA synaptosomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eG), reinforcing the idea that in DA synaptosomes vesicle location is polarized to a putative active zone where vesicle may tether and fuse.\u003c/p\u003e \u003cp\u003eWe also compared DA synaptosomes forming CS-DHS (that is adhering to a GLU terminal) or not. We found no significant differences in their vesicle number, spatial organization and tethering (Figure S9). This suggests that the adhesion to a GLU terminal does not affect the propensity of DA synaptosomes to contain tethered vesicles.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eCorrelation of DHS connection with alterations of CS terminals\u003c/h3\u003e\n\u003cp\u003eWe wondered whether the adhesion of a DA terminal to a GLU synapse affects the ultrastructure of the GLU presynapse. We compared GLU synaptosomes that were not part of DHS (GLU DA- synapses, Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA, n\u0026thinsp;=\u0026thinsp;30) with GLU synaptosomes engaged into a DHS (GLU DA\u0026thinsp;+\u0026thinsp;synapses or DHSs, Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB, n\u0026thinsp;=\u0026thinsp;32), both from the dual color model. The total number of vesicles are not different between GLU DA- and GLU DA\u0026thinsp;+\u0026thinsp;terminals, on average 205.7 and 222.9 vesicles (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC; p-value\u0026thinsp;=\u0026thinsp;0.864), as well as, the density of vesicles, 1893 per \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e and 1943 per \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD; p-value\u0026thinsp;=\u0026thinsp;0.577). Similarly, the AZ areas are not different (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE). Together, these argue that the morphology of terminals and overall distribution of SVs is the same.\u003c/p\u003e \u003cp\u003eNeverthless, the fraction of the volume occupied by proximal SVs was significantly higher in GLU DA\u0026thinsp;+\u0026thinsp;than in GLU DA- synaptosomes, and showed a peak that signifies a higher concentration of proximal SVs, while the occupancy peak was absent in GLU DA- terminals (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eF,G). A precise characterization of SV location and tether length showed that the results for GLU DA\u0026thinsp;+\u0026thinsp;terminals were similar to those of unperturbed hippocampal glutamatergic synapses reported previously \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, while GLU DA- terminals had a lower proportion of SVs located 5\u0026ndash;10 nm to the AZ membrane but, surprisingly, a higher proportion of tethers 6\u0026ndash;12 nm in length (Figure S10).\u003c/p\u003e \u003cp\u003eAdditionally, proximal vesicles in GLU DA\u0026thinsp;+\u0026thinsp;synaptosomes are significantly more connected to other vesicles than in GLU DA- synaptosomes (46.7% vs. 21.6%; p-value\u0026thinsp;=\u0026thinsp;0.0167) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eJ). The fraction of tethered vesicles was about 70% bigger in GLU DA\u0026thinsp;+\u0026thinsp;terminals but did not reach significance. However, the fraction of proximal SVs that are both tethered and connected was significantly higher in GLU DA\u0026thinsp;+\u0026thinsp;terminals (26.7% vs. 5.4%; p-value\u0026thinsp;=\u0026thinsp;0.0435) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eK,L). Together with the observation that the number of tethers per SV was similar in GLU DA\u0026thinsp;+\u0026thinsp;and GLU DA-terminals at all distances to the plasma membrane (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eK), our results argue that connectors are primarily responsible for the observed difference in the proximal SV distribution.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eHere, we report the ultrastructure of fully hydrated, close-to-native DA terminals from adult mouse striatum at a single nanometer resolution, as well as of DHS, which comprise DA terminals in contact with GLU synapses and were characterized previously with immunofluorescence \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSynaptosomes obtained from mouse brain constitute a reliable model to investigate the spatial configuration of synapses \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. They are particularly amenable for cryo-ET because they can be directly observed by cryo-EM without further processing such as cryo-sectioning or cryo-focused ion beam milling \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Moreover, they retain functionality, such as SV exocytosis and endocytosis, protein composition and post-synaptic calcium signalling \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Likewise, we show with FM4-64 labelling that our preparation of synaptosomes undergoes stimulation dependent SV cycling. Moreover, the GLU synaptosomes we obtained from the mouse striatum are qualitatively and quantitatively comparable to forebrain and hippocampal synaptosomes (likely glutamatergic), and dissociated cultures observed previously with cryo-ET \u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. The presynaptic element contains hundreds of small round SVs (~\u0026thinsp;40 nm diameter) and contact a PSE with a clearly defined PSD and a 32 nm wide synaptic cleft filled with dense material. This defines an AZ of ~\u0026thinsp;0.08 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, in the range of reported sizes varying between 0.04 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e and 0.10 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e ,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. Moreover, SVs are polarized towards the AZ, with a peak of volume occupancy\u0026thinsp;~\u0026thinsp;25 nm from the plasma membrane, which reflects the enrichment in tethered and primed vesicles, similar to forebrain synapses and was proposed to be necessary for proper neurotransmitter release \u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. Nevertheless, this peak was less pronounced in our sample of CS synaptosomes than in forebrain synaptosomes \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, which may reflect genuine differences in relative SV pool sizes in cortical/hippocampal vs striatal synapses. Remarkably, proximal SVs located less than 5 nm from the AZ have in both CS and cortical synapses on average 3 tethers \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, which likely corresponds to the primed state of SV and to functionally defined readily releasable SVs. In neurons lacking Munc13-1 and Munc13-2, which abolishes priming \u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e, proximal SVs have only 1 tether \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Therefore, we propose that CS synapses have primed vesicles with the same structural hallmark as the ones defined in hippocampal synapses.\u003c/p\u003e \u003cp\u003eThe most striking feature of DA synaptosomes is the relative sparsity of small synaptic-like vesicles, about 3-fold less dense than in GLU synaptosomes. Because DA synaptosomes are also smaller than GLU synaptosomes, there are 10 times less vesicles in DA synaptosomes than in GLU synaptosomes. These observations are consistent with ultrastructure of DA axons determined with serial electron microscopy reconstruction in chemically fixed striatum tissue in which axonal varicosities display very heterogenous ultrastructure: some have only few vesicles, others have small synaptic-like vesicles, larger ones, or both types of vesicles \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Likewise, recent studies have shown qualitatively similar results in cultured DA neurons observed with cryo-CLEM \u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e and DA neurons derived from human induced pluripotent stem cells \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Vesicles in DA synaptosomes are also significantly more elongated than vesicles in GLU synaptosomes. Elongated-shaped vesicles have also been documented at GABAergic inhibitory synapses with conventional electron microscopy \u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e and also with cryo-CLEM \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Moreover, SVs in GLU synapses become elongated in the absence of VGLUT1, which reflects a lower luminal osmotic pressure \u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e,\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. Therefore, DA vesicles, like GABA vesicles, experience different osmotic pressure than GLU vesicles.\u003c/p\u003e \u003cp\u003eImportantly, we show that, similar to GLU synaptosomes, vesicles in DA synaptosomes are frequently linked together by connectors and to the plasma membrane by tethers. Nevertheless, only 39% of the DA synaptosomes contain at least one tethered vesicle. The DA synaptosomes with at least one tethered vesicle are bigger and contain twice as many vesicles as the ones without any tethered vesicle. Interestingly, other studies have shown that only 30% of DA axonal varicosities contain active zone proteins bassoon, RIM or ELKS \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, and this proportion is also found for functional DA varicosities able to release fluorescent dopamine analog \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Therefore, we propose that synaptosomes containing tethered vesicles correspond to release-competent DA terminals. Moreover, the number of tethers linked to proximal vesicles (less than 45 nm from the plasma membrane) is the same in GLU and DA synaptosomes. This suggests that similar molecules are involved in vesicle tethering in both types of synapses. Indeed, proteomic analysis of striatal DA synapses show similar degree of enrichment of the major proteins involved in vesicle tethering \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. However, DA synaptosomes lacked very proximal vesicles (less than 5 nm from the plasma membrane), which have on average 3 tethers, a hallmark of vesicles primed for release in GLU synapses. This difference points to a fundamental difference between these two types of terminals. We thus predict that the kinetics of vesicle exocytosis will be much slower in DA neurons than in GLU neurons, where SV exocytosis occurs in less than 1 ms after calcium entry. So far, the kinetics of DA vesicle exocytosis has been determined with fast amperometry or fluorescence imaging of dopamine sensors \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.The rise time of these events is greater than 10 ms but it could be dominated by diffusion of ligand to the detector micrometres away from the dopamine release site. More precise investigation of dopamine release is needed to determine if the kinetics of dopamine release is genuinely slower, and whether this is due to differences in vesicle priming.\u003c/p\u003e \u003cp\u003eWe found a clear PSE at only 2 out of 94 DA terminals. These are among the synaptosomes which contain many small round vesicles, and are thus indistinguishable from GLU synaptosomes. They may correspond to terminals of DA/GLU neurons co-expressing the vesicular glutamate transporter VGLUT2 \u003csup\u003e41,42,58\u003c/sup\u003e. In the adult mouse, these terminals are mostly concentrated in the shell of the nucleus accumbens. Interestingly, in these axons VGLUT2 is segregated from the vesicular dopamine transporter VMAT2, suggesting that glutamate and dopamine release sites are segregated \u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. In the other 92 DA terminals reconstructed, no clear PSE was identified.\u003c/p\u003e \u003cp\u003eWe identified previously that DA terminals can strongly interact with other synaptic elements in a so-called DHS \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. In the 32 reconstructed DHS, we found that 28 directly adhere to a presynaptic VGLUT1 element and 4 contacting the GLU PSE. On the other hand, DA terminals, identified in conventional transmission EM with immuno-labelling of tyrosine hydroxylase, are as likely to contact the pre- or post-synaptic side of CS synapses \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. The selection of DHSs by cryo-fluorescence with the presynaptic marker VGLUT1-Venus likely biased our sampling towards interaction with presynaptic markers. Nevertheless, we could characterize both pre- and post-synaptic DA/GLU adhesion sites in their native state with cryo-ET. We found that the adhesion sites had similar areas, around 0.05 \u0026micro;m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, as well as the cleft sizes, around 12 nm. However, these adhesions did not define a privileged tethering site for DA vesicles. Overall, we did not identify a localized site for DA vesicle tethering, suggesting a large AZ for exocytosis.\u003c/p\u003e \u003cp\u003eFinally, we showed that the morphology of GLU terminals and overall distribution of SVs are very similar in GLU DA\u0026thinsp;+\u0026thinsp;and GLU DA- terminals. Also, we detected tethers of different lengths, which were previously associated with different molecular composition of tethers \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, in both types of presynaptic terminals. These suggest that the GLU terminal formation and their protein composition does not depend on the presence of DA terminals. Nevertheless, the distribution of proximal SVs in GLU DA\u0026thinsp;+\u0026thinsp;terminals was consistent with those of non-perturbed glutamatergic synapses, while GLU DA- terminals showed a flat proximal SV distribution profile previously associated with a reduced neurotransmitter release \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Furthermore, we correlated these differences with changes in SV connectivity and possibly also with tethering. Therefore, our data indicate that the contact between DA and GLU terminals at DHS modulates the SV organization at the single nanometer scale, as well as release properties of GLU terminals, and that this modulation is mediated by SV connectors. We could hypothesize that the formation of DHSs in the striatum is an important feature modulating glutamate release. Overall, DHSs could be an important substrate for the modulation of glutamate release and striatal activity in vivo. These multi-partite assemblies offer a close proximity between dopamine release sites and specific glutamatergic synapses. By this mechanism, the spatial and temporal synchronization between dopamine and glutamate activity would be maximal, which could play a major role in the plasticity of excitatory input to striatal neurons \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eAnimal models\u003c/h2\u003e \u003cp\u003eThree mouse models have been used. The VGLUT1\u003csup\u003evenus\u003c/sup\u003e knock-in line to label CS synapses. The DAT-cre BAC transgenic mouse line\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e in which we transduced VTA/SNc neurons with an AAV1 carrying pCAG-Flex-mNeongreen coding sequences to label dopaminergic neurons projecting to the striatum (coordinates from bregma are A/P: 2.9mm; M/L: 1.6 mm; D/V: 4.6 mm with 12\u0026deg; angle). To co-detect CS synapses with DA terminals in the same sample, we used a double labelling approach. DAT-cre crossed with the reporter Ai14TdTomato mouse line were crossed with VGLUT1\u003csup\u003evenus\u003c/sup\u003e KI mice. Adult mice of both genders, aged 12 to 18 weeks were used.\u003c/p\u003e \u003cp\u003eWe refined the experimental design and the procedures to reduce as possible the number of animals used and their suffering. All procedures were in accordance with the European guide for the care and use of laboratory animals and approved by the ethics committee of Bordeaux University (CE50) and the French Ministry of Research under the APAFIS n\u0026deg; #21132 and #38144.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003ePreparation of synaptosomes\u003c/h2\u003e \u003cp\u003eThe preparation of synaptosomes was adapted from a previously published protocol \u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e. Briefly, animals were euthanized by cervical dislocation, decapitated and the head was immersed in liquid nitrogen for 5 seconds for rapid cooling but not freezing of the tissue. The striata were subsequently dissected under an epi-fluorescence stereomicroscope (Leica Microsystems, Germany). Samples were then homogenized in 1.5 ml of ice-cold isosmolar buffer (0.32 M sucrose, 4 mM HEPES pH7.4, protease inhibitor cocktail Set 3 EDTA-free (EMD Millipore Corp.)) using a 2 ml-glass-Teflon\u0026reg; homogenizer with 12 strokes at 900 rpm. The homogenate (H) was centrifuged at 1000 g for 5 min at 4\u0026deg;C in a benchtop microcentrifuge. The supernatant (S1) was separated from the pellet (P1) and centrifuged at 12500 g for 8 min at 4\u0026deg;C. The crude synaptosomes pellet (P2) was suspended in 350 \u0026micro;L of isosmolar buffer and layered on a two-step ficoll density gradient (5 mL of 13% Ficoll and 5 mL of 7.5% ficoll, both in 0.32 M sucrose, 4 mM HEPES). The gradient was centrifuged at 50,000 \u0026times; g for 1 h and 10 min at 4\u0026deg;C (Optima L100XP Beckman Coulter with SW32Ti rotor). The synaptosome fraction was recovered at the 7.5 and 13% ficoll interface using a 0.5 ml syringe. An additional centrifuge of the collected fraction was performed in 2ml of HEPES-buffered Krebs like solution (HBK) \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e composed of : 143 mM NaCl, 4,7 mM KCl, 1,3mM MgSO\u003csub\u003e4\u003c/sub\u003e, 1,2 mM CaCl\u003csub\u003e2\u003c/sub\u003e, 20 mM HEPES, 0,1mM Na\u003csub\u003e2\u003c/sub\u003eHPO\u003csub\u003e4\u003c/sub\u003e and 10 mM D-glucose, pH\u0026thinsp;=\u0026thinsp;7,4) at 12,500g for 5min in order to wash the excess of ficoll and sucrose residues. The barely visible pellet was resuspended in 200\u0026micro;l HBK leftover and placed for 15 minutes at 37\u0026deg;C for physiological recovery before plunge-freezing.\u003c/p\u003e \u003cp\u003eAlternatively, S1 fraction diluted in HBK has been used allowing for faster preparation with comparable results.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eFM4-64 uptake and release assay\u003c/h2\u003e \u003cp\u003eTo assess synaptosome integrity, we performed a fluorogenic vesicle exo-endocytosis assay using the amphiphilic styryl dye FM4-64 (Thermo Scientific T13320). Synaptosomes were prepared fresh with the same protocol as for cryo-CLEM. Then they have been diluted in HBK and centrifuged 34 min at 6750g on 12 mm coverslips coated with 1 mg/ml Poly-L-Lysine. Imaging was performed right after on a wide field epifluorescence microscope Leica DMI8 equipped with an inverted 63X/1.4 oil objective, a Hamamatsu Flash 4.0 v2 camera and a controlled 37\u0026deg;C/CO2 chamber. Coverslips were placed in a recording chamber (Ludin), incubated with 500 \u0026micro;l of prewarmed HBK and imaged in TRITC and GFP channels at different registered positions over a 5 \u0026micro;m stack (1st acq). HBK was removed and FM4-64 (6 ng/\u0026micro;l final concentration), KCl (40 mM final concentration) and 37\u0026deg;C/CO2 HBK mixture was added inducing exo-endocytosis cycle and staining membrane and internalized vesicles (see supplementary Fig.\u0026nbsp;1A). After 1 min 30 s, KCl was rinsed twice with HBK and FM4-64 with HBK was added staining the remaining plasma membrane. After 1 min 30 s, second acquisition was launched (2nd acq). After 5 rinses with HBK, a third acquisition captured signal from internalized vesicles only (3rd acq). KCl-HBK mixture was added to induce depolarization and exocytosis and distaining was imaged 1 min 30 s after (4th acq). Loading was measured by subtracting FM intensity of single synaptosome at acquisition n\u0026deg;3 (vesicles stained) versus acquisition n\u0026deg;1 (background). Release was measured by subtracting intensity of single synaptosome at acquisition n\u0026deg;3 (vesicles stained) with acquisition n\u0026deg;4 (non exocytosed vesicles and background dye binding). Control experiments with no KCl for loading or no KCl for release were performed. Analysis was performed on Fiji\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e using a home-built macro-command.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003ePlunge-freezing\u003c/h2\u003e \u003cp\u003eQuantifoil R2/2 Cu 200 or R3.5/1 Cu 200 mesh grids (Q-R2_2-2C100 or Q-3.5_1-3C100 Delta Microscopies) were glow-discharged for 30 seconds at 2.5 mA using a ELMO glow discharge system (Cordouan Technologies) to enhance their hydrophilicity. Following this, 4 \u0026micro;l of freshly purified synaptosomes, mixed with with 10 nm colloidal gold beads (Sigma 741957) and 100 nm FluoSpheres\u0026trade; beads (F8797 ThermoFischer) were applied to the grids. The excess sample was immediately blotted away from the opposite side for 5 seconds in a Vitrobot Mark IV (Thermo Fisher Scientific), maintained at a temperature of 4\u0026deg;C and at 100% humidity. Finally, grids were plunge-frozen in liquid ethane and stored in liquid nitrogen until observation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eCryo-fluorescence microscopy\u003c/h2\u003e \u003cp\u003eWe used a commercially available system to perform cryo-fluorescence microscopy (Leica \u0026ndash; DM6 FS Cryo CLEM). We monitored on test grids with TetraSpecks\u0026trade; beads (Thermo Scientific T7279) shifts between channels and drift artifacts in cryogenic conditions. No significant shift was noticed, while noticeable mechanical drift occurs for acquisitions of more than 20 z-steps, which may be necessary when the grid is not flat and exactly parallel to the focal plane. Therefore, we limited our acquisitions to flat areas and imaged consecutive z planes with an increment of 0.5 to 1 \u0026micro;m (20 z-steps max). A custom-made Plexiglas\u0026reg; chamber was built around the set-up to maintain relative humidity under 40% reducing contaminations by water. Temperature of the objective chamber was maintained at -190\u0026deg;C (83K) using a pump projecting vapor of fresh liquid nitrogen. Each grid was observed separately reducing the time spent in the chamber to 25 minutes maximum. Sample was observed with a 50X objective (Leica \u0026ndash; HC PL APO 50x/0.90NA Dry 11566064) and z-stacks were acquired through a Hamamatsu ORCA-Flash 4.0 camera, in 4 channels, in the following order: Brightfield (Empty); Tx Red (Em: BP560/40 nm; Ex: BP630/75 nm); GFP (Em: BP470/40 nm; Ex: BP525/50 nm); DAPI (Em: BP360/40 nm; Ex: BP470/40 nm). Large, flat area of the grids were acquired using the LAS X Navigator mosaic imaging tool. Grids were stored in liquid nitrogen until the cryo-EM session.\u003c/p\u003e \u003cp\u003eImages from regions of interest were stacked in a maximal intensity Z-projection using Fiji software for each fluorescence channels \u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. Minimal Z projection was performed for the brightfield channel allowing the detection of holes in the carbon layer. Contrast has been adjusted and allowed detection of all synaptosomes (dim and bright), as well as fluorescent beads. TIFF images have been saved in PNG file format for editing. Each square of interest (with one or more synaptosome in a hole) was identified and marked with a unique number. Cropping of each corresponding area into a single image was done and resulting PNG files were saved-back into TIFF files for compatibility with Serial EM \u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eCorrelation between fluorescence and electron microscopy\u003c/h2\u003e \u003cp\u003eThe correlation process was carried out in 2 stages. The initial \u0026ldquo;approximate\u0026rdquo; correlation focused on locating the grid squares captured through cryo-fluorescence microscopy. After inserting the grid into the cryo-transmission electron microscope, a low-magnification montage (34x) was constructed using SerialEM \u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e to visualize the entire grid. The fluorescence image was then imported into the software via the \u0026ldquo;Import Map\u0026rdquo; function. Areas of interest were identified using obvious landmarks, such as carbon holes and the center of the grid, which were visible in both imaging modalities.\u003c/p\u003e \u003cp\u003eOnce the squares were identified, a high-magnification montage (1600x) was created for each area of interest. The corresponding fluorescence image for each square was imported separately. Precise correlation in SerialEM was done using the fluorescent fiducial beads as registration points, detectable in both fluorescence and electron microscopy. Beads were selected and assigned specific identifiers on the fluorescence image, and the same beads were subsequently located on the electron microscopy image. A total of 5 to 10 surrounding beads were used for the transformation to correlate with a single target for tomography. Both images were then correlated using the \u0026ldquo;Transform Item\u0026rdquo; function.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eCryo-electron tomography\u003c/h2\u003e \u003cp\u003eTilt series of the synaptosomes were recorded using a Talos Arctica microscope (Thermo Fisher Scientific), operating at 200 kV and equipped with a K2 Summit direct electron detector (Gatan). The tilt series were acquired in a bidirectional scheme using SerialEM 4.0, covering angles from \u0026minus;\u0026thinsp;60\u0026deg; to +\u0026thinsp;60\u0026deg; with 2\u0026deg; increments, at a magnification of 11,000x, resulting in a pixel size of 3.987 \u0026Aring;. Imaging was performed with an underfocus of -8 \u0026micro;m, and the cumulative electron dose ranged between 80\u0026ndash;90 e⁻/\u0026Aring;\u0026sup2;.\u003c/p\u003e \u003cp\u003eTilt series alignment and tomogram reconstruction were carried out using EMAN2 \u003csup\u003e62\u003c/sup\u003e for quick visualization and data selection. IMOD software package 4.11.25 \u003csup\u003e37\u003c/sup\u003e integrated into the SCIPION framework \u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e was used to reconstruct selected tomograms for segmentation in 3dmod (see below). Reconstruction was achieved using weighted back-projection combined with a SIRT-like filter. Binning of 4 was applied and resulted in a pixel size of 1.5588 nm.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eSegmentation and production of models\u003c/h2\u003e \u003cp\u003eWe selected tomograms with accurate correlation, sufficient contrast and intact synaptic features for segmentation (numbers provided in Table\u0026nbsp;1). Other tomograms were discarded (examples in Figure S8).\u003c/p\u003e \u003cp\u003eManual segmentation was performed using 3dmod, a software from the IMOD package \u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. Synaptosome plasma membranes were segmented until disappearance and meshed, therefore no interpolation was used and missing wedge volume was not corrected for the synaptosomes. Organelles and vesicles were interpolated using spherical interpolation, resulting in closed objects. Active zone area was defined as plasma membrane portion facing the post-synaptic density. Contact area (DHS) was the plasma membrane portion of the DA synaptosomes following tight apposition (\u0026lt;\u0026thinsp;15 nm) to the GLU plasma membrane. Synaptic vesicles were defined as objects smaller than 80 nm with round or elongated shapes. Mitochondria were easily detected as dense folded membranes are present in the lumen. Vesicular bodies were observed thanks to the presence of one or more vesicle-like structure inside. Objects with folded, irregular shapes and bigger diameter have been categorized separately and correspond to unidentified structures.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eData analysis\u003c/h2\u003e \u003cp\u003eWe obtained geometrical information such as volume and maximum diameter from the 3D models using \u003cem\u003eimodinfo\u003c/em\u003e command lines (IMOD package). Sphericity was calculated using Wadell\u0026rsquo;s index \u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e (W\u003csub\u003ei\u003c/sub\u003e) with the following formula W\u003csub\u003ei\u003c/sub\u003e = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\pi\\:}^{\\frac{1}{3}}{\\left(6V\\right)}^{\\frac{2}{3}}/S\\)\u003c/span\u003e\u003c/span\u003e where V represents the volume and S the surface of the vesicle. Thus W\u003csub\u003ei\u003c/sub\u003e = 1 for a perfect sphere and decreases for non-spherical objects.\u003c/p\u003e \u003cp\u003eVesicle density inside synaptosomes was obtained dividing the number of vesicles by the volume available inside the synaptosome. In Fiji, cytoplasmic density profiles were obtained averaging a 40 nm stack and pixel intensities were measured using the built-in commands. In Fiji, synaptic cleft and interspace density profiles were measure at 3 different single z planes (low, middle and high) in the stack and averaged for each tomogram. Values were normalized to the neighboring non-cleft area corresponding to background. We excluded tomograms in which gold beads were present as it artificially alters the density.\u003c/p\u003e \u003cp\u003eSV tethers and connectors were detected in and automated, template-free manner using the hierarchical connectivity algorithm, and their morphology, localization and interrelationship was analyzed by Pyto package (version 1.10, available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/vladanl/Pyto\u003c/span\u003e\u003cspan address=\"https://github.com/vladanl/Pyto\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), as described before \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Briefly, for the analysis of vesicle distribution (volume occupancy), the presynaptic cytoplasm (including SVs) was divided into 1-pixel-thick layers according to the distance to the AZ membrane, and the fraction of the layer volume occupied by SVs was measured. Connector and tether lengths were calculated as the minimal edge-to-edge distance between connector / tether voxels contacting an SV or plasma membrane that takes into account central regions of tethers and connectors. In this way, the ambiguity inherent to the measurement of length of 3D objects is resolved and curvature of tethers and connectors contributed to their calculated lengths. However, possible extended protein-lipid binding regions are not considered. All image processing and statistical analysis software procedures were written in Python and implemented in Pyto package [37]. Pyto uses NumPy and SciPy packages and graphs are plotted using Matplotlib\u003csup\u003e\u003cspan additionalcitationids=\"CR67\" citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e Statistical analysis was performed between the experimental groups using only planned, orthogonal comparisons. For the analysis of properties pertaining to individual SVs, connectors and tethers (such as the SV distance to the AZ membrane, tether length and fraction of tethers/connectors having a certain property), values within experimental groups were combined. Bars on the graphs show mean values and error bars the standard error of the mean (sem). In cases a fraction of SVs or tethers is shown, the error bars represent sem between synapse means. We used Student's t test for statistical analysis of values that appeared to be normally distributed (e.g., vesicle diameter) and K-W test (nonparametric) for values deviating from the normal distribution (e.g., number of tethers and connectors per vesicle). For frequency data (e.g., fraction of connected and non-connected vesicles), χ2 test was used. In all cases, confidence levels were calculated using two-tailed tests. The confidence values were indicated in the graphs by a single asterisk for P\u0026thinsp;\u0026lt;\u0026thinsp;0.05, double for P\u0026thinsp;\u0026lt;\u0026thinsp;0.01, and triple for P\u0026thinsp;\u0026lt;\u0026thinsp;0.001. All values of statistical tests are presented in Table\u0026nbsp;3 (Supplementary data).\u003c/p\u003e \u003cp\u003eWe focused on tomograms originating from the VGLUT1\u003csup\u003evenus\u003c/sup\u003e x DAT-cre x Ai14 tdTomato because it is the only model where we could reliably identify both GLU and DA and thus determine CS-DHS and non-CS-DHS conditions with certainty (see Table\u0026nbsp;2). Moreover, we selected tomograms where the contrast was optimal and sufficient information was available. We picked GLU synaptosomes where we identified the active zone and for DA where there were more than 2 vesicles. Thus, it ensures an optimal detection and a reliable comparison of the filaments between both conditions. We normalized tomogram density and applied a Kernel Gaussian filter (sigma\u0026thinsp;=\u0026thinsp;2) to improve detection. Tomograms used for the glutamatergic active zone were cropped on the dedicated region and analysis was performed blind to the DHS, non-DHS condition.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eEstimation for shortest distance between tethered vesicles\u003c/h2\u003e \u003cp\u003eWe measured the arc distance between two tethered DA vesicles d (that is the distance while remaining in the plasma membrane) by measuring the distance in the projection in the tomogram plane measured along the arc of the plasma membrane (d\u003csub\u003exy\u003c/sub\u003e) and the distance between the two tomogram planes in which the tethers connect the plasma membrane (d\u003csub\u003ez\u003c/sub\u003e) as d = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\sqrt{\\left({d}_{xy}\\right)\u0026sup2;+\\left({d}_{z}\\right)\u0026sup2;}\\)\u003c/span\u003e\u003c/span\u003e. We estimated the distance between random points by taking the average maximum diameter of T\u0026thinsp;+\u0026thinsp;DA synaptosomes D (722\u0026thinsp;\u0026plusmn;\u0026thinsp;205 nm, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). For the distance between random points, we approximated the projection of synaptosomes by a circle of radius R\u0026thinsp;=\u0026thinsp;D/2\u0026thinsp;=\u0026thinsp;361 nm. The average arc distance between random points in a circle is half the length of a half circle (by symmetry), or \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{\\pi\\:}{2}R\\)\u003c/span\u003e\u003c/span\u003e = 567 nm and its variance is \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{\\pi\\:\u0026sup2;}{12}R\u0026sup2;\\)\u003c/span\u003e\u003c/span\u003e that is a standard deviation of 327 nm. This estimate is a lower limit of the true value, because it neglects the axial distance d\u003csub\u003ez\u003c/sub\u003e and approximates synaptosomes as circles, minimizing distance.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing financial interests.\u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eP.L. performed stereotaxic AAV injections. synaptosome preparations and live imaging of FM4-64 uptake with the help of V. P-B. P.L. performed synaptosome freezing, cryo-fluo imaging, cryo-EM and cryo-ET with the help of R.A. E.M. and R.F. supervised the cryo-ET methodology. P.L. did all reconstructions and annotations of tomograms under the supervision of E.H. and D.P. P.L. performed quantitative analysis of all data together with V.L. P.L., R.F., E.H. and D.P. acquired funding. P.L., E.H. and D.P. wrote the manuscript and all other authors edited it.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eWe thank Peter Vanhoutte and Nicolas Heck for providing the DAT-cre * Ai14 tdTomato line, the P\u0026ocirc;le in Vivo for animal breeding, husbandry, help with stereotaxic injections. We thank members of the Bordeaux Imaging Center, namely Monica Fernandez-Monreal from for help with cryo-fluorescence microscopy and Fabrice Cordeli\u0026egrave;res for the analysis of FM4-64 loading experiments. We thank student interns Timoth\u0026eacute; Lapha, Jade Giraud and Sol\u0026egrave;ne Hospital for reconstruction of some of the tomograms.\u003c/p\u003e \u003cp\u003eThis work was supported by ANR (DopamineHub ANR-19-CE16-0003 to E.H. and D.P., FrontoFAT ANR-20-CE14-0020-03 to E.H. and UltraDopa ANR-24-CE16-5973-01 to E.H., D.P. and R.F.), the Fondation Recherche M\u0026eacute;dicale (to P.L., end of PhD and D.P., FRM team), the European Research Council (ERC consolidator Grant PneumoTransfo to R.F.) and the Regional Council of Nouvelle Aquitaine (ParkSynGraft 205024) to E.H. and D.P. This work was supported by the French Government through the France 2030 program [grant number 21-ESRE-0024] managed by the French National Research Agency (ANR) as part of the \"Investissements d'avenir\" program.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eCosta KM, Schoenbaum G, Dopamine (2022) Curr Biol 32:R817\u0026ndash;R824\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEmperador-Melero J, Kaeser PS (2020) Assembly of the presynaptic active zone. Curr Opin Neurobiol 63:95\u0026ndash;103\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFern\u0026aacute;ndez-Busnadiego R et al (2013) Cryo\u0026ndash;electron tomography reveals a critical role of RIM1α in synaptic vesicle tethering. J Cell Biol 201:725\u0026ndash;740\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePapantoniou C et al (2023) Munc13- and SNAP25-dependent molecular bridges play a key role in synaptic vesicle priming. Sci Adv 9:eadf6222\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRadecke J et al (2023) Morphofunctional changes at the active zone during synaptic vesicle exocytosis. EMBO Rep 24:e55719\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eImig C et al (2014) The Morphological and Molecular Nature of Synaptic Vesicle Priming at Presynaptic Active Zones. Neuron 84:416\u0026ndash;431\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatriarchi T et al (2018) Ultrafast neuronal imaging of dopamine dynamics with designed genetically encoded sensors. Science 360:eaat4422\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun F et al (2018) A Genetically Encoded Fluorescent Sensor Enables Rapid and Specific Detection of Dopamine in Flies, Fish, and Mice. Cell 174:481\u0026ndash;496e19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanerjee A, Lee J, Nemcova P, Liu C, Kaeser P (2020) S. Synaptotagmin-1 is the Ca2\u0026thinsp;+\u0026thinsp;sensor for fast striatal dopamine release. eLife 9, e58359\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDelignat-Lavaud B et al (2023) Synaptotagmin-1-dependent phasic axonal dopamine release is dispensable for basic motor behaviors in mice. Nat Commun 14:4120\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLebowitz JJ et al (2024) Synaptotagmin-7 Counteracts Short-Term Depression during Phasic Dopamine Release. \u003cem\u003eeNeuro\u003c/em\u003e 11\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu C, Kershberg L, Wang J, Schneeberger S, Kaeser PS (2018) Dopamine Secretion Is Mediated by Sparse Active Zone-like Release Sites. Cell 172:706\u0026ndash;718\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanerjee A et al (2022) Molecular and functional architecture of striatal dopamine release sites. Neuron 110:248\u0026ndash;265e9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePereira DB et al (2016) Fluorescent false neurotransmitter reveals functionally silent dopamine vesicle clusters in the striatum. Nat Neurosci 19:578\u0026ndash;586\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWildenberg G et al (2021) Partial connectomes of labeled dopaminergic circuits reveal non-synaptic communication and axonal remodeling after exposure to cocaine. eLife 10:e71981\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuber B, Lučić V (2022) Neurons as a model system for cryo-electron tomography. J Struct Biology: X 6:100067\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTritsch NX, Sabatini BL (2012) Dopaminergic Modulation of Synaptic Transmission in Cortex and Striatum. Neuron 76:33\u0026ndash;50\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBamford NS et al (2004) Heterosynaptic Dopamine Neurotransmission Selects Sets of Corticostriatal Terminals. Neuron 42:653\u0026ndash;663\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArluison M, Dietl M, Thibault J (1984) Ultrastructural morphology of dopaminergic nerve terminals and synapses in the striatum of the rat using tyrosine hydroxylase immunocytochemistry: a topographical study. Brain Res Bull 13:269\u0026ndash;285\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDescarries L, Watkins KC, Garcia S, Bosler O, Doucet G (1996) Dual character, asynaptic and synaptic, of the dopamine innervation in adult rat neostriatum: a quantitative autoradiographic and immunocytochemical analysis. J Comp Neurol 375:167\u0026ndash;186\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUchigashima M, Ohtsuka T, Kobayashi K, Watanabe M (2016) Dopamine synapse is a neuroligin-2\u0026ndash;mediated contact between dopaminergic presynaptic and GABAergic postsynaptic structures. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 113, 4206\u0026ndash;4211\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoss J, Bolam JP (2008) A Dopaminergic Axon Lattice in the Striatum and Its Relationship with Cortical and Thalamic Terminals. J Neurosci 28:11221\u0026ndash;11230\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeyene AG et al (2019) Imaging striatal dopamine release using a nongenetically encoded near infrared fluorescent catecholamine nanosensor. Sci Adv 5:eaaw3108\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYagishita S et al (2014) A critical time window for dopamine actions on the structural plasticity of dendritic spines. Science 345:1616\u0026ndash;1620\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaget-Blanc V et al (2022) A synaptomic analysis reveals dopamine hub synapses in the mouse striatum. Nat Commun 13:3102\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerzog E et al (2001) The Existence of a Second Vesicular Glutamate Transporter Specifies Subpopulations of Glutamatergic Neurons. J Neurosci 21:RC181\u0026ndash;RC181\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSartori A et al (2007) Correlative microscopy: Bridging the gap between fluorescence light microscopy and cryo-electron tomography. J Struct Biol 160:135\u0026ndash;145\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwartz CL, Sarbash VI, Ataullakhanov FI, Mcintosh JR, Nicastro D (2007) Cryo-fluorescence microscopy facilitates correlations between light and cryo-electron microscopy and reduces the rate of photobleaching. J Microsc 227:98\u0026ndash;109\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFernandez-Busnadiego R et al (2010) Quantitative analysis of the native presynaptic cytomatrix by cryoelectron tomography. J Cell Biol 188:145\u0026ndash;156\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao C-L et al (2018) Differentiation and Characterization of Excitatory and Inhibitory Synapses by Cryo-electron Tomography and Correlative Microscopy. J Neurosci 38:1493\u0026ndash;1510\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNicholls DG, Sihra TS (1986) Synaptosomes possess an exocytotic pool of glutamate. Nature 321:772\u0026ndash;773\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerzog E et al (2011) In Vivo Imaging of Intersynaptic Vesicle Exchange Using VGLUT1Venus Knock-In Mice. J Neurosci 31:15544\u0026ndash;15559\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTuriault M et al (2007) Analysis of dopamine transporter gene expression pattern\u0026thinsp;\u0026ndash;\u0026thinsp;generation of DAT-iCre transgenic mice. FEBS J 274:3568\u0026ndash;3577\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMadisen L et al (2010) A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci 13:133\u0026ndash;140\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerhage M et al (1991) Differential release of amino acids, neuropeptides, and catecholamines from isolated nerve terminals. Neuron 6:517\u0026ndash;524\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaniel JA, Malladi CS, Kettle E, McCluskey A, Robinson PJ (2012) Analysis of synaptic vesicle endocytosis in synaptosomes by high-content screening. Nat Protoc 7:1439\u0026ndash;1455\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMastronarde DN, Held SR (2017) Automated tilt series alignment and tomographic reconstruction in IMOD. J Struct Biol 197:102\u0026ndash;113\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartinez-Sanchez A et al (2021) Trans-synaptic assemblies link synaptic vesicles and neuroreceptors. Sci Adv 7:eabe6204\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLučić V, Yang T, Schweikert G, F\u0026ouml;rster F, Baumeister W (2005) Morphological Characterization of Molecular Complexes Present in the Synaptic Cleft. Structure 13:423\u0026ndash;434\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorogod N, Petersen CC, Knott GW (2015) Ultrastructural analysis of adult mouse neocortex comparing aldehyde perfusion with cryo fixation. Elife 4:e05793\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChuhma N (2004) Dopamine Neurons Mediate a Fast Excitatory Signal via Their Glutamatergic Synapses. J Neurosci 24:972\u0026ndash;981\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTritsch NX, Ding JB, Sabatini BL (2012) Dopaminergic neurons inhibit striatal output through non-canonical release of GABA. Nature 490:262\u0026ndash;266\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRizzoli SO, Betz WJ (2005) Synaptic vesicle pools. Nat Rev Neurosci 6:57\u0026ndash;69\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchikorski T, Stevens CF (2001) Morphological correlates of functionally defined synaptic vesicle populations. Nat Neurosci 4:391\u0026ndash;395\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatanabe S et al (2013) Ultrafast endocytosis at mouse hippocampal synapses. Nature 504:242\u0026ndash;247\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchrod N et al (2018) Pleomorphic linkers as ubiquitous structural organizers of vesicles in axons. PLoS ONE 13:e0197886\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLučić V, Fern\u0026aacute;ndez-Busnadiego R, Laugks U, Baumeister W (2016) Hierarchical detection and analysis of macromolecular complexes in cryo-electron tomograms using Pyto software. J Struct Biol 196:503\u0026ndash;514\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchikorski T, Stevens CF (1997) Quantitative ultrastructural analysis of hippocampal excitatory synapses. J Neurosci 17:5858\u0026ndash;5867\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiksou L et al (2007) Three-Dimensional Architecture of Presynaptic Terminal Cytomatrix. J Neurosci 27:6868\u0026ndash;6877\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuber B, Lučić V (2019) Molecular architecture of the presynaptic terminal. Curr Opin Struct Biol 54:129\u0026ndash;138\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVaroqueaux F et al (2002) Total arrest of spontaneous and evoked synaptic transmission but normal synaptogenesis in the absence of Munc13-mediated vesicle priming. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 99, 9037\u0026ndash;9042\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLycas MD, Morado DR, Gether U, Briggs JAG, Erlendsson S (2024) Ultrastructural Dynamics of Dopaminergic Presynaptic Release Sites revealed by Cryo-correlative Light and Electron Microscopy. 04.15.589543 Preprint at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2024.04.15.589543\u003c/span\u003e\u003cspan address=\"10.1101/2024.04.15.589543\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFujise K, Rosenfeld MS, Rafiq NM (2024) Synaptic vesicle characterization of iPSC-derived dopaminergic neurons provides insight into distinct secretory vesicle pools. 02.22.581435 Preprint at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2024.02.22.581435\u003c/span\u003e\u003cspan address=\"10.1101/2024.02.22.581435\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUchizono K (1965) Characteristics of Excitatory and Inhibitory Synapses in the Central Nervous System of the Cat. Nature 207:642\u0026ndash;643\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValdivia O (1971) Methods of fixation and the morphology of synaptic vesicles. J Comp Neurol 142:257\u0026ndash;273\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiksou L et al (2013) A role for vesicular glutamate transporter 1 in synaptic vesicle clustering and mobility. Eur J Neurosci 37:1631\u0026ndash;1642\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Oostrum M et al (2023) The proteomic landscape of synaptic diversity across brain regions and cell types. Cell 0\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S et al (2015) Dopaminergic and glutamatergic microdomains in a subset of rodent mesoaccumbens axons. Nat Neurosci 18:386\u0026ndash;392\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe-Smedt-Peyrusse V, Darriet L, Trifilieff P, Herzog E, Angelo MF (2018) Subcellular Fractionation of Brain Tissue from Small Tissue Explants. In: Murphy KM (ed) Synaptosomes. Springer, New York, NY, pp 75\u0026ndash;84. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-1-4939-8739-9_5\u003c/span\u003e\u003cspan address=\"10.1007/978-1-4939-8739-9_5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchindelin J et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676\u0026ndash;682\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMastronarde DN (2005) Automated electron microscope tomography using robust prediction of specimen movements. J Struct Biol 152:36\u0026ndash;51\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang G et al (2007) EMAN2: an extensible image processing suite for electron microscopy. J Struct Biol 157:38\u0026ndash;46\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede la Rosa-Trev\u0026iacute;n JM et al (2016) Scipion: A software framework toward integration, reproducibility and validation in 3D electron microscopy. J Struct Biol 195:93\u0026ndash;99\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKremer JR, Mastronarde DN, McIntosh JR (1996) Computer Visualization of Three-Dimensional Image Data Using IMOD. J Struct Biol 116:71\u0026ndash;76\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWadell H, Volume (1935) Shape, and Roundness of Quartz Particles. J Geol 43:250\u0026ndash;280\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarris CR et al (2020) Array programming with NumPy. Nature 585:357\u0026ndash;362\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVirtanen P et al (2020) SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat Methods 17:261\u0026ndash;272\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHunter JD, Matplotlib (2007) A 2D Graphics Environment. Comput Sci Eng 9:90\u0026ndash;95\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 3 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6081416/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6081416/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDopamine is an essential brain neuromodulator involved in reward and motor control. Dopaminergic (DA) neurons project to most brain areas, with particularly dense innervation in the striatum. DA varicosities bind to target striatal synapses and form dopamine hub synapses (DHS). However, the basic features of dopamine release sites are still largely unknown. Here we studied the ultrastructure of fluorescent DA and glutamatergic (GLU) synaptosomes isolated from the striatum of adult mice with cryo-correlative light and electron microscopy and cryo-electron tomography. We observed that DA synaptosomes display\u0026thinsp;~\u0026thinsp;10 times fewer vesicles than GLU ones. DA vesicles are bigger and less round. Vesicle organization at single nanometer scale indicates that most GLU synaptosomes have tethered and primed vesicles, indicative of a readily releasable pool, while only 39% of DA synaptosomes have tethered vesicles, which appear not to be primed. In addition, GLU terminals contacted by DA terminals in DHS have more primed vesicles than others. While DA varicosities do not form genuine synapses, their adhesion to cortico-striatal synapses may convey a local regulation of synaptic release properties.\u003c/p\u003e","manuscriptTitle":"Cryo-correlative light and electron tomography of dopaminergic axonal varicosities reveals non-synaptic modulation of cortico-striatal synapses","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-21 05:37:33","doi":"10.21203/rs.3.rs-6081416/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"b3644b52-4124-4cd7-9e52-2634fe5efda4","owner":[],"postedDate":"April 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":46397637,"name":"Biological sciences/Neuroscience/Synaptic transmission/Vesicle trafficking"},{"id":46397638,"name":"Biological sciences/Neuroscience/Synaptic transmission/Neurotransmitters"}],"tags":[],"updatedAt":"2025-12-30T08:27:22+00:00","versionOfRecord":{"articleIdentity":"rs-6081416","link":"https://doi.org/10.1038/s41467-025-66355-x","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-12-13 05:00:00","publishedOnDateReadable":"December 13th, 2025"},"versionCreatedAt":"2025-04-21 05:37:33","video":"","vorDoi":"10.1038/s41467-025-66355-x","vorDoiUrl":"https://doi.org/10.1038/s41467-025-66355-x","workflowStages":[]},"version":"v1","identity":"rs-6081416","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6081416","identity":"rs-6081416","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}