SorCS1-mediated Sorting of Neurexin in Dendrites Maintains Presynaptic Function

The pre- and postsynaptic membranes comprising the synaptic junction differ in protein composition. The mechanisms that maintain the polarized distribution of synaptic membrane proteins are poorly understood. The sorting receptor SorCS1 is a critical trafficking regulator of neuronal receptors, including neurexin (Nrxn), a presynaptic adhesion molecule essential for synaptic transmission. We find that SorCS1 controls a balance between axonal and dendritic Nrxn1α surface levels. Newly synthesized Nrxn1α traffics to the somatodendritic surface, followed by endocytosis. SorCS1 interacts with the Rab11 effector protein Rab11FIP5/Rip11 to facilitate the transition of internalized Nrxn1α from early to recycling endosomes and bias Nrxn1α surface polarization toward the axon. In the absence of SorCS1, Nrxn1α accumulates in early endosomes and mis-polarizes to the dendritic surface, impairing presynaptic function. The axonal/dendritic balance of Nrxn1α surface distribution is activity-dependent, indicating that SorCS1-mediated sorting in somatodendritic endosomes dynamically controls Nrxn1α axonal surface polarization required for proper presynaptic function.


Abstract
The pre-and postsynaptic membranes comprising the synaptic junction differ in protein composition. The mechanisms that maintain the polarized distribution of synaptic membrane proteins are poorly understood.
The sorting receptor SorCS1 is a critical trafficking regulator of neuronal receptors, including neurexin (Nrxn), a presynaptic adhesion molecule essential for synaptic transmission. We find that SorCS1 controls a balance between axonal and dendritic Nrxn1α surface levels. Newly synthesized Nrxn1α traffics to the somatodendritic surface, followed by endocytosis. SorCS1 interacts with the Rab11 effector protein Rab11FIP5/Rip11 to facilitate the transition of internalized Nrxn1α from early to recycling endosomes and bias Nrxn1α surface polarization toward the axon. In the absence of SorCS1, Nrxn1α accumulates in early endosomes and mis-polarizes to the dendritic surface, impairing presynaptic function. The axonal/dendritic balance of Nrxn1α surface distribution is activity-dependent, indicating that SorCS1-mediated sorting in somatodendritic endosomes dynamically controls Nrxn1α axonal surface polarization required for proper presynaptic function.

Introduction
Synapses connect neurons into neural circuits that underlie brain function. Synapses are organized in three major compartments: the presynaptic nerve terminal, the synaptic cleft, and the postsynaptic specialization 1 , which differ widely in their protein composition 2,3 . Reliable trafficking mechanisms that control compartment-specific protein composition during formation, maturation, and plasticity of the synapse are thus of key importance for assembly and function of neural circuits [4][5][6] .
Vacuolar protein sorting 10 (VSP10P)-receptor family proteins have emerged as important regulators of intracellular trafficking of numerous ligands in neurons 7,8 . The SorCS1-3 (Sortilin-related CNS expressed) members of the VSP10P family are predominantly expressed in the vertebrate nervous system and regulate the intracellular trafficking of key synaptic surface proteins, including glutamate receptors, neurotrophin receptors, and adhesion molecules [9][10][11][12][13] . These studies identify SorCS proteins as critical regulators of synaptic function and plasticity, but the cellular and molecular mechanisms by which they sort their synaptic cargo proteins remain poorly understood.
SorCS1 is an endosomal sorting receptor that cycles between the plasma membrane and the endosomal pathway [7][8][9] . A key SorCS1 cargo protein is the synaptic adhesion molecule neurexin (Nrxn). SorCS1 directly binds Nrxn via an extracellular domain cis interaction and is required for surface and synaptic localization of Nrxn 13 . Nrxns, in mammals expressed from three genes as α-, β-, and γ-Nrxns, are presynaptic adhesion molecules that engage in a network of interactions with multiple pre-and postsynaptic ligands 14 . Nrxns act in a cell type-specific manner to regulate synapse number and function [15][16][17] . Mutations in NRXNs have been associated with multiple neuropsychiatric disorders, especially schizophrenia and autism [18][19][20] , underscoring the importance of Nrxns for normal brain function. Despite their essential role, little is known about the mechanisms that control presynaptic localization and abundance of Nrxns 21 .
Here, we show that the sorting receptor SorCS1 controls a balance between axonal and dendritic Nrxn surface levels. In mature neurons, newly synthesized Nrxn1α traffics to the somatodendritic plasma membrane, followed by endocytosis, endosomal sorting and translocation to the axon via a transcytotic pathway. SorCS1 interacts with the Rab11 effector protein Rab11FIP5/Rip11 to facilitate the transition of internalized Nrxn1α from early to recycling endosomes and bias surface polarization of Nrxn1α toward the axon. In the absence of SorCS1, Nrxn1α accumulates in early endosomes and mis-polarizes to the dendritic surface, impairing synapse formation and presynaptic function. The balance between the axonal/dendritic surface distribution of Nrxn1α is activity-dependent, indicating that SorCS1-mediated sorting in dendrites maintains the axonally polarized distribution of Nrxn1α that is required for proper presynaptic function.

SorCS1-mediated sorting in dendrites regulates Nrxn1α axonal surface polarization
We first determined the subcellular distribution of endogenous SorCS1 in neurons, which has remained uncertain due to a lack of suitable antibodies for the immunodetection of SorCS1. To this end, we generated a SorCS1 knock-in (KI) mouse (Sorcs1 HA/HA ), using CRISPR/Cas9 to insert a hemagglutinin (HA) epitope tag in the Sorcs1 locus ( Supplementary Fig. 1a,b). HA immunodetection in Sorcs1 HA/HA cortical neurons revealed prominent punctate SorCS1 immunoreactivity in soma and dendrites (Fig. 1a,c), which was mimicked by HA-tagged SorCS1 exogenously expressed in wild-type (WT) hippocampal neurons (Fig. 1b,c). These results show a preferentially somatodendritic distribution for SorCS1, suggesting that SorCS1-mediated sorting of cargo proteins occurs in this compartment.
To investigate the mechanism by which SorCS1 controls the surface distribution of Nrxn, we used primary cultures of cortical neurons, which strongly express Sorcs1 22,23 , and focused on Nrxn1α as one of the most abundant Nrxn isoforms in the brain 15,24 . Cortical Sorcs1 flox/flox mouse neurons were electroporated with GFP-tagged Cre recombinase (Cre) to remove SorCS1 (Sorcs1 KO), transfected with HA-tagged Nrxn1α (HA-Nrxn1α), and live-labeled with an HA antibody in combination with Ankyrin-G and MAP2 immunodetection to analyze the surface distribution of Nrxn1α in the axonal and dendritic compartments, respectively. Loss of SorCS1 caused a decrease in Nrxn1α axonal surface levels and a concomitant increase in dendritic surface levels compared to control cells expressing GFP, changing Nrxn1α surface polarization from axonal to dendritic (Fig. 1d,e). Re-expression of WT SorCS1 (SorCS1 WT ) in Cre-positive Sorcs1 flox/flox neurons restored Nrxn1α surface polarization (Fig. 1d,e), indicating that the defect occurred cell-autonomously. Rescue with the endocytosis-defective mutant 25 of SorCS1 (SorCS1 Y1132A ), which remains on the plasma membrane ( Supplementary Fig. 1c,d), did not rescue Nrxn1α surface polarization (Fig. 1d,e), suggesting that SorCS1 is either required for endocytosis of Nrxn1α or needs to localize to endosomes to regulate the surface balance of Nrxn1α between axon and dendrites.
Taken together, these results indicate that SorCS1 acts in the somatodendritic compartment to control a balance between dendritic and axonal surface expression of Nrxn1α and bias Nrxn1α surface polarization toward the axon.

Indirect axonal trafficking of Nrxn1α
Our results suggests that Nrxn1α traffics through both dendritic and axonal compartments. To monitor Nrxn1α trafficking in cortical neurons, we employed the retention using selective hooks (RUSH) system 26 to retain Nrxn1α in the endoplasmic reticulum (ER) and induced synchronous release and transport through the secretory pathway by application of biotin (Fig. 2a). We performed live-cell imaging in [8][9][10] days in vitro (DIV) rat cortical neurons co-expressing SBP-GFP-Nrxn1α (reporter) and streptavidin-KDEL (ER hook), and live-labeled these with an antibody against the axon initial segment (AIS) protein To analyze the dynamics of Nrxn1α-positive vesicles in both neuronal compartments, we analyzed vesicle trafficking at early (20-34 min) and late (1-2 hr) time points after Nrxn1α release from the ER.
Kymograph analysis showed a small increase in the mean number of vesicles present in dendrites at late compared to early time-points, as a consequence of an increased number of motile vesicles (Fig. 2g,h; Supplementary Movies 5,6). Shortly after ER release, the majority of vesicles in dendrites moved in the anterograde direction, shifting to a retrograde movement at the later time-point (Fig. 2i). Strikingly, motile Nrxn1α vesicles were absent in axons shortly after ER release but increased dramatically at late time- In a complementary approach, we transfected mouse cortical neurons with HA-Nrxn1α and treated these with a reversible blocker of anterograde Golgi trafficking ( Supplementary Fig. 2e), brefeldin A (BFA) 28 , to allow synchronized release of Nrxn1α. Following BFA wash-out and resumption of trafficking, Nrxn1α was first detected on the dendritic surface before appearing on the axonal surface ( Supplementary   Fig. 2d,f-h). Together, these results show that following transport through the secretory pathway, Nrxn1α is first inserted in the somatodendritic plasma membrane and appears on the axonal surface with a marked delay.

Endocytosis and endosomal transport are required for axonal polarization of Nrxn1α
Our observations suggest an indirect, dendrite-to-axon trafficking route for Nrxn1α. This is reminiscent of the transcytotic trafficking pathway 29 , in which cargo is first delivered to the somatodendritic plasma membrane, followed by internalization into endosomes and trafficking to the axonal compartment via endosome-derived carriers. To monitor Nrxn1α intracellular transport, mouse cortical neurons transfected with HA-Nrxn1α were incubated with an anti-HA antibody, washed and returned to the incubator for 15 or 40 min of chase (antibody pulse-chase assay; Supplementary Fig. 3a). Internalized Nrxn1α was largely restricted to the dendritic compartment ( Supplementary Fig. 3a,b) and shifted from early endosomes (EEs; EEA1-and Rab5-positive) to recycling endosomes (REs; Rab11-positive) over time ( Supplementary Fig.  3c-h). Blocking of dynamin-dependent endocytosis, either by expressing a dominant-negative mutant of Dynamin1 (K44A) ( Supplementary Fig. 4a,b) or with Dynasore ( Supplementary Fig. 4c,d), caused a decrease in Nrxn1α axonal surface levels and a concomitant increase in dendritic surface levels .
Expression of GTP binding-deficient mutants of Rab5 (S34N) and Rab11 (S25N), to interfere with transport to early and recycling endosomes, respectively, mimicked the effect of blocking endocytosis (Supplementary Fig. 4e-h). However, expression of Rab7 (T22N) to interfere with transport to late endosomes did not affect the axonal surface polarization of Nrxn1α ( Supplementary Fig. 4i,j). Thus, endocytosis, early and recycling endosomal transport, but not late endosomal transport, are required for accumulation of Nrxn1α on the axonal plasma membrane. Consistent with these observations, neurons transfected with HA-Nrxn1α and chased for 90 min with an anti-HA antibody showed increased levels of internalized Nrxn1α in the axon compared to neurons chased for 15 min ( Supplementary Fig. 3i,j). As Nrxn1α endocytosis is largely restricted to the dendritic compartment ( Supplementary Fig. 3a,b,i,j), dendritic plasma membrane-derived Nrxn1α is the most likely source for sustaining this increase in internalized Nrxn1α detected in the axon. Taken together, these results indicate that Nrxn1α is trafficked to the axon via the transcytotic route.

SorCS1-mediated sorting controls the subcellular distribution of endogenous Nrxn
Transcytotic trafficking implies that endogenous Nrxns will also localize to dendrites, at least transiently.
Several independent studies have reported the presence of a dendritic pool of Nrxn 13,30-33 , in addition to axonal localization, but this has remained controversial due to the lack of suitable antibodies for the detection of endogenous Nrxns. To unequivocally determine the subcellular localization of endogenous Nrxn1α in neurons, we generated a Nrxn1α KI mouse (Nrxn1α HA/HA ) ( Supplementary Fig. 5a,b). We cultured Nrxn1α HA/HA cortical neurons together with WT mouse cortical neurons, such that the majority of presynaptic inputs onto Nrxn1α HA/HA neurons are from WT, unlabeled cells, enabling the reliable detection of HA-Nrxn1α in the somatodendritic and axonal compartment ( Supplementary Fig. 5c). In permeabilized DIV3 Nrxn1α HA/HA neurons, total Nrxn1α localized to a perinuclear compartment and displayed a punctate distribution in dendrites and axons, with prominent accumulation at dendritic tips and axonal growth cones ( Fig. 3a). In DIV7 and DIV11 neurons, dendritic localization became more pronounced (Fig. 3a,b;  Supplementary Fig. 5d). Quantification of the axonal/dendritic ratio of the total pool of endogenous Nrxn1α indicated a shift from axonal to dendritic enrichment of Nrxn1α as neurons mature (Fig. 3b). Surface Nrxn1α showed a polarized distribution toward the axon at all developmental time points analyzed (Fig.   3c,d). The localization pattern observed in Nrxn1α HA/HA neurons was recapitulated in WT neurons labeled with a pan-Nrxn antibody 34 (Fig. 3e,f), which shows specific labeling of endogenous Nrxn in cultured mouse cortical neurons ( Supplementary Fig. 5g,h), or transfected with exogenous HA-Nrxn1α ( Supplementary   Fig. 5e,f). Thus, endogenous Nrxn1α is an axonally polarized surface protein that accumulates in dendrites as neurons mature.
To test the role of transcytosis in the subcellular distribution of endogenous Nrxn1α, we blocked endocytosis in Nrxn1α HA/HA neurons using the K44A Dynamin1 mutant. Expression of K44A-Dynamin1 reduced the axonal intensity of total endogenous Nrxn1α and decreased the balance between axonal and dendritic Nrxn1α levels in Nrxn1α HA/HA cortical neurons (Fig. 3g,h), indicating that endocytosis is required for axonal polarization of endogenous Nrxn1α.
To determine whether SorCS1-mediated sorting controls the subcellular distribution of endogenous Nrxn, we immunostained Sorcs1 flox/flox neurons electroporated with Cre or GFP with the pan-Nrxn antibody.
Loss of SorCS1 reduced axonal intensity of total endogenous Nrxn and decreased the axonal/dendritic ratio for endogenous Nrxn compared to control neurons (Fig. 3i,j). Taken together, these findings demonstrate that the axonally polarized distribution of endogenous Nrxn is controlled by transcytosis and SorCS1-mediated sorting.

Selective mis-sorting of transcytotic cargo in the absence of SorCS1
We next determined whether the surface polarization of other axonal membrane proteins was affected in the absence of SorCS1. The cell adhesion molecules L1/NgCAM and Caspr2 are both targeted to the axonal surface, but via distinct trafficking routes. L1/NgCAM is transcytosed from dendrites to axon 35,36 .
Caspr2 is indiscriminately delivered to both axon and dendrites but is selectively endocytosed from the somatodendritic surface, resulting in axonal polarization 37 . Sorcs1 flox/flox cortical neurons electroporated with Cre or GFP were transfected with myc-L1 or HA-Caspr2 and immunostained for surface myc ( Supplementary Fig. 6a) or HA ( Supplementary Fig. 6b). Consistent with a general role for SorCS1 in regulating the transcytotic pathway, surface polarization of L1/NgCAM, but not of Caspr2, was perturbed by SorCS1 loss (Supplementary Fig. 6a-c). The polarity index of two somatodendritic proteins (GluA2 and MAP2), was unchanged in Sorcs1 KO neurons ( Supplementary Fig. 7), indicating that loss of SorCS1 does not generally perturb the polarized distribution of neuronal proteins. Together, these results show that loss of SorCS1 selectively impairs the transcytotic route, while keeping other neuronal polarity mechanisms intact.

A SorCS1-Rab11FIP5 interaction controls Nrxn1α transition from early to recycling endosomes
The cellular and molecular mechanisms by which SorCS1 sorts its cargo remain unclear. Our previous SorCS1 interactome analysis 13 identified multiple components of the endocytic machinery associated with SorCS1. To dissect which sorting step of Nrxn1α is disrupted in Sorcs1 KO neurons, we followed the fate of internalized Nrxn1α throughout the endosomal pathway ( Supplementary Fig. 8e).
To determine whether SorCS1 is required for endocytosis of Nrxn1α, we repeated the antibody pulse-chase assay in Sorcs1 flox/flox cortical neurons. Internalization of HA-Nrxn1α in dendrites after 20 min of chase was not affected in Sorcs1 KO neurons compared to control cells ( Supplementary Fig. 8a,b), indicating that SorCS1 is not required for Nrxn1α endocytosis. REs (Rab11) in dendrites, and REs (Rab11) in the axon, respectively. Quantification revealed an increase in the colocalization of internalized Nrxn1α with EEs and Rab4-fast REs in dendrites of Sorcs1 KO neurons compared to control neurons (Fig. 4b,d). Colocalization of internalized Nrxn1α with Rab11-REs in Sorcs1 KO neurons was decreased in dendrites ( Fig. 4f) and even more strongly reduced in the axon (Fig. 4h).
Thus, in the absence of SorCS1-mediated sorting, Nrxn1α accumulates in EEs and is mis-sorted to Rab4fast REs. These observations are consistent with our finding that dendritic surface levels of Nrxn1α are increased in Sorcs1 KO neurons (Fig. 1a,b), which is likely due to increased recycling of Nrxn1α from Rab4-positive endosomes back to the dendritic plasma membrane. Similarly, the reduced axonal surface levels of Nrxn1α (Fig. 1a,b) likely result from decreased sorting of Nrxn1α to Rab11-REs. Together, these results indicate that SorCS1 plays a critical role in the transition of Nrxn1α from EEs to Rab11-REs ( Supplementary Fig. 8e).
We reasoned that SorCS1 interacts with additional proteins to facilitate Nrxn1α sorting from early to recycling endosomes. Rab11 family-interacting protein 5 (Rab11FIP5/Rip11; from here on Rip11), is prominently present in the raw mass spectrometric (MS) data set obtained after affinity purification (AP) of SorCS1 complexes from rat brain extracts with two independent antibodies 13 (AP-MS data available online). Rip11, which belongs to the family of Rab11-interacting proteins 38 , localizes to REs in polarized epithelial cells and regulates transcytosis of proteins from the basolateral to the apical plasma membrane 39,40 . A possible role of Rip11 in neuronal transcytosis has not been reported. Western blot analysis of immunoprecipitated HA-SorCS1 from postnatal Sorcs1 HA/HA KI cortical extracts showed strong HA-SorCS1 enrichment and a robust Rip11 band, which are absent in the mouse IgG control (Fig. 4i), validating the AP-MS data. Like SorCS1, Rip11 predominantly localized to the somatodendritic compartment ( Supplementary Fig. 8c,d), where it displayed a punctate distribution and robustly colocalized with SorCS1 (Fig. 4j,k). Expression of a dominant-negative (DN) form of Rip11 that inhibits the transport from early to recycling endosomes 41 reduced axonal surface levels of HA-Nrxn1α and increased dendritic surface levels, shifting Nrxn1α surface polarization from axonal to dendritic (Fig. 4l,m). Together, these results indicate that a SorCS1-Rip11 interaction facilitates sorting of Nrxn1α from early to recycling endosomes, thus biasing trafficking to the axon while preventing mis-sorting to fast recycling and late endosomes ( Supplementary Fig. 8e).

Loss of SorCS1 impairs Nrxn-mediated synaptogenesis
Mis-sorting of Nrxn to the dendritic surface in Sorcs1 KO neurons would be expected to impair synapse formation on heterologous cells expressing a postsynaptic ligand for Nrxn 42,43 . We infected Sorcs1 flox/flox cortical neurons with lentivirus (LV) to express Cre-T2A-mCherry or mCherry as control and co-cultured these with HEK293T cells expressing FLAG-Neuroligin 1 (Nlgn1). FLAG-Nlgn1 induced strong clustering of the presynaptic marker Synapsin1 in control axons contacting the HEK293T cell surface (Fig. 5a). Loss of SorCS1 reduced Nlgn1-mediated Synapsin1 clustering (Fig. 5a). This defect was specific, as presynaptic differentiation induced by Netrin-G Ligand 3 (NGL-3), which requires leukocyte common antigen-related protein (LAR) in axons 44 , was not affected in Sorcs1 KO neurons (Fig. 5b). Infection with a lentiviral vector expressing shRNAs against all Nrxns (Nrxn TKD) 42,45 mimicked the defect in Sorcs1 KO neurons (Fig. 5a), suggesting that Nlgn1-induced presynaptic differentiation is impaired in Sorcs1 KO neurons due to decreased axonal surface levels of Nrxn. To test this, we expressed Nrxn1α in Sorcs1 KO neurons using LV. Expression of Nrxn1α, which mis-polarizes to the dendritic surface in Sorcs1 KO neurons, did not rescue the impaired Nlgn1-mediated synaptogenic activity (Fig. 5a). However, expression of a Nrxn1α deletion mutant lacking the cytoplasmic 4.1-binding motif (Δ4.1), which bypasses the transcytotic route and SorCS1-mediated sorting ( Supplementary Fig. 9), rescued the synaptogenic defect caused by SorCS1 loss (Fig. 5a). These results show that SorCS1-mediated sorting of Nrxns, by promoting Nrxn accumulation on the axonal surface, is required for normal synaptogenesis onto Nlgn1-expressing cells.

SorCS1-mediated sorting is required for presynaptic function
In the absence of SorCS1, mis-sorting of Nrxns, which regulate neurotransmitter release [15][16][17] , would be expected to impair presynaptic function. To assess the consequences of loss of SorCS1 on synaptic function, we recorded spontaneous miniature excitatory postsynaptic currents (mEPSCs) from Sorcs1 flox/flox autaptic cortical cultures electroporated with Cre or GFP. SorCS1 loss strongly decreased mEPSC frequency ( Fig. 6a,b). Decay kinetics and amplitude of mEPSCs were not altered by loss of SorCS1 ( Fig.   6a,b), suggesting that the amount of neurotransmitter release per vesicle and postsynaptic receptor properties were not affected. The amplitude and total charge transfer of single evoked EPSCs (eEPSCs) were reduced in Sorcs1 KO neurons (Fig. 6c,d). The mEPSC frequency and eEPSC amplitude defects can be attributed to a decrease in the readily releasable pool (RRP) size or in the vesicular release probability (Pves). To assess these parameters, we applied a hyperosmotic sucrose stimulus. The amplitude and total charge transfer of synaptic responses (Fig. 6e,f) and RRP size (Fig. 6g) induced by single application of 0.5 M sucrose were strongly decreased in Sorcs1 KO neurons. Because eEPSC amplitude and RRP size were proportionally reduced in Sorcs1 KO cells, the Pves (evoked EPSC charge/initial sucrose charge) was unchanged (Fig. 6h). To further examine the synaptic defects in Sorcs1 KO neurons, we performed train stimulations of 20 stimuli at different frequencies to probe for short-term . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; plasticity defects that cannot be detected by induction of single EPSC and single sucrose application.
Repeated stimulation at 10 Hz produced a more pronounced rundown of normalized evoked responses (synaptic depression) in Sorcs1 KO neurons compared to control cells (Fig. 6i,j). Furthermore, pairedpulse ratio (PPR) measurement at different inter-stimuli intervals further showed the increase in synaptic depression caused by SorCS1 loss (Fig. 6k), suggesting that active excitatory synapses in Sorcs1 KO cells have reduced neurotransmitter release. Indeed, the RRP size and release kinetics during 10 Hz-train stimulation ( Supplementary Fig. 10) calculated by back-extrapolation were reduced in Sorcs1 KO neurons, indicating a presynaptic defect independent of the silent synapses phenotype that we described previously 13 . Together, these observations indicate that loss of SorCS1 impairs neurotransmitter release, reminiscent of Nrxn loss-of-function [15][16][17] , supporting the notion that SorCS1-mediated transcytosis of Nrxns is required for normal presynaptic function.

The axonal/dendritic balance of Nrxn surface polarization is activity-dependent
Finally, we tested whether the axonal/dendritic balance of Nrxn surface polarization is regulated in response to changes in activity. Cortical cultures were treated either with picrototoxin (PTX) or tetrodotoxin (TTX) for 48 hr. PTX (GABA A receptor blocker) and TTX (voltage-gated sodium channel blocker) cause global changes in neuronal activity, by increasing or decreasing the neuronal firing rates, inducing homeostatic scaling down or scaling up 46 , respectively. PTX treatment led to a loss of exogenous Nrxn1α from the axonal surface, concomitant with an increase on the dendritic surface, changing Nrxn1α surface polarization from axonal to dendritic (Fig. 7a,b). The axonal/dendritic balance of endogenous Nrxns, detected with a pan-Nrxn antibody, was also shifted toward dendritic polarization in PTX-treated neurons (Fig. 7c,d). Conversely, TTX-treated neurons showed an increase in axonal surface polarization of Nrxn1α (Fig. 7e,f). After SorCS1 depletion from cortical neurons, TTX treatment failed to increase the axonal surface polarization of Nrxn1α (Fig. 7g,h). In conclusion, these observations indicate that neurons can dynamically change the axonal/dendritic surface balance of Nrxns in response to bidirectional changes in activity, likely in a SorCS1-mediated transcytosis-dependent manner.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Discussion
The mechanisms by which neurons control the polarized distribution and abundance of key synaptic membrane proteins are poorly understood. Here, we show that SorCS1-mediated sorting in somatodendritic endosomes controls a balance between axonal and dendritic surface levels of Nrxn and dynamically maintains Nrxn axonal surface polarization required for proper presynaptic function.

SorCS1 sorts cargo to the transcytotic pathway
The mechanism by which SorCS1 regulates neuronal cargo trafficking has remained unclear. We find that SorCS1 controls a balance between axonal and dendritic surface polarization of Nrxn. In mature neurons, newly synthesized Nrxn1α traffics to the somatodendritic surface. Following endocytosis, SorCS1 mediates Nrxn1α sorting from early (Rab5-positive) to recycling endosomes (Rab11-positive). Nrxn1αcontaining REs are subsequently transcytosed from dendrites to the axon. In agreement, blocking of endocytosis, EE-, and RE-mediated transport all shift the axonal/dendritic surface balance of Nrxn1α toward the dendrite. These effects are mimicked by Sorcs1 KO. We find that SorCS1 is required for the transition of endocytosed Nrxn1α from EEs to REs and identify the Rab11-interacting protein Rab11FIP5/Rip11 as a novel SorCS1 interactor. Interference with Rip11 function alters the axonaldendritic surface balance of Nrxn1α in the same way that Sorcs1 KO does. Rab11FIPs function as linkers between Rab11 and motor proteins to promote sorting of cargo from EEs to REs 38,47 and transcytosis in polarized epithelial cells 39,40 , but their function in neuronal transcytosis has not been explored. Our data suggest that SorCS1/Rip11 form a protein complex that localizes to dendritic endosomes and sorts internalized Nrxn1α from early to recycling endosomes, thus biasing the trafficking of Nrxn1α-containing REs to the axon while preventing Nrxn1α mis-sorting to lysosomes and the dendritic surface ( Supplementary Fig. 8e).
The transcytotic pathway has been proposed as the canonical mechanism for protein and lipid . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; and BACE1 52 ) in neurons have been shown to be translocated from dendrites to the axon. We show that SorCS1 loss impairs trafficking of Nrxns and NgCAM, but not of Caspr2, which traffics via the selective endocytosis/retention pathway 37 , supporting a role for SorCS1 in regulating the transcytotic pathway in the maintenance of neuronal polarity.

SorCS1-mediated sorting of Nrxn and synaptic function
Presynaptic differentiation in Sorcs1 KO axons contacting Nlgn1-expressing HEK293T cells, but not NGL-3-expressing cells, is impaired. Nrxn TKD in neurons mimics this defect, in agreement with previous observations 42 . Expression of the Nrxn1α deletion mutant lacking the 4.1-binding motif in Sorcs1 KO neurons rescues the defect in presynaptic differentiation on Nlgn1-expressing heterologous cells.
Together, these results suggest that the selective impairment in Nlgn1-induced presynaptic differentiation in Sorcs1 KO neurons is due to a decrease in the abundance of axonal surface Nrxns.
Nrxns play key roles in neurotransmitter release [15][16][17] . Alpha-Nrxn KO in neocortical cultured slices reduces mEPSC/mIPSC frequency and eIPSC amplitude 17 . Beta-Nrxn KO in cortical neurons decreases mEPSC frequency and eEPSC amplitude 15 . Similarly, loss of SorCS1 in autaptic cortical neurons results in decreased mEPSC frequency and eEPSC amplitude. In addition, loss of SorCS1 in autaptic cortical . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; neurons decreased the amplitude of sucrose-evoked EPSC and increased the synaptic depression of normalized eEPCSs during stimulus trains, similar to alpha-Nrxn KO neurons 17 . Together, these results suggest that mis-trafficking of Nrxn is a major contributor to the defects in presynaptic function in Sorcs1 KO neurons. These defects are likely attributed to a decrease in neurotransmitter release, rather than a decrease in excitatory synapse density, which is unaffected in Sorcs1 KO cortical neurons 13 and in Nrxn TKD hippocampal neurons 42 .

A dynamic axonal/dendritic surface balance of Nrxn
Using CRISPR/Cas9-mediated epitope tagging of endogenous Nrxn1α, we demonstrate unequivocally that Nrxn1α is an axonally polarized surface protein with a substantial presence in dendrites. The axonaldendritic balance of Nrxn1α is developmentally regulated. Early in neuronal development, endogenous Nrxn1α is enriched in the axon and traffics directly to the axon after TGN exit. As neurons mature, Nrxn1α trafficking changes to a dendrite-exclusive insertion after TGN exit, and accumulation of Nrxn1α at the axonal surface is sustained via a SorCS1-dependent transcytotic pathway. The biological function of this developmental regulation and circuitous trafficking route is not clear. One hypothesis that has been put forward is that transcytosis in neurons provides an efficient way of delivering receptors from a reservoir of readily-synthesized proteins. Similar to other transcytotic axonal cargo, a dendritic pool of Nrxn may supply the demand for receptors along the axonal surface, either constitutively 50,51 or in response to ligandreceptor interactions and signalling 50,55 . Supporting this idea, we find that the axonal/dendritic surface balance of Nrxn is activity-dependent. SorCS1-mediated sorting could thus dynamically regulate Nrxn surface distribution to adjust presynaptic release properties in response to signaling and neuronal activity.
Dendritically localized Nrxns might have a function in this compartment. Postsynaptically expressed Nrxn1 decreases Nlgn1's synaptogenic effect in hippocampal neurons, likely by cis-inhibition of Nlgn1 that prevents trans-synaptic interaction with presynaptic Nrxns 32 . In retinal ganglion cells, a shift in Nrxn localization away from dendrites has been proposed to allow dendritic innervation 30 . In cortical neurons we observe the opposite: Nrxn surface expression in dendrites increases with maturation. Possibly, dendritic Nrxns modulate the function of postsynaptic Nrxn ligands, such as Nlgns, which control postsynaptic neurotransmitter receptor function 56 , or LRRTM1, which shapes presynaptic properties 57 .
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; Consistent with such a modulatory role, loss of alpha-Nrxns in cortical neurons results in a cell-autonomous decrease in NMDA receptor-dependent postsynaptic currents, likely reflecting a change in the postsynaptic localization of NMDA receptors 58 . At the C. elegans neuromuscular junction, the ectodomain of postsynaptic Nrxn is proteolytically cleaved and binds to the presynaptic α2δ calcium channel subunit to inhibit presynaptic release 59 .
Our observation that critical cargo stalls in EEs in the absence of SorCS1-mediated sorting is reminiscent of the enlarged EEs that are an early hallmark of neurodegenerative diseases 60 , especially in light of SORCS1's link to late-onset Alzheimer's disease 61,62 . Given SorCS1's role in sorting neuronal receptors, impaired trafficking of these critical cargo proteins in the absence of SorCS1 might also contribute to the pathophysiology of the neurodevelopmental disorders with which SORCS1 has been associated 63-66 . These observations underscore the notion that intracellular sorting is a key contributor to the proper maintenance of synaptic protein composition and function.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Declaration of Interests
The authors declare no competing interests.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.   . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  (e) Quantification of (d): surface HA-Nrxn1α fluorescence intensity in axon and dendrites relative to total surface levels and normalized to cells expressing EGFP and ratio of axonal/dendritic surface HA intensity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Figure 2
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.     (m) Quantification of (l): surface HA-Nrxn1α fluorescence intensity in axon and dendrites relative to total surface levels and normalized to cells expressing EGFP and ratio of axonal/dendritic surface HA intensity (n = 30 for each group). *p < 0.05; ***p < 0.001 (Kruskal-Wallis test followed by Dunn's multiple comparisons test, 3 independent experiments).
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.    (h) Quantification of (g): surface HA-Nrxn1α fluorescence intensity in axon and dendrites relative to total surface levels and normalized to cells expressing EGFP untreated and ratio of axonal/dendritic surface HA intensity. EGFP_Ctr (n = 35 neurons); EGFP_TTX (n = 35); Cre_Ctr (n = 36); Cre_TTX (n = 35). ***p < 0.001 (Kruskal-Wallis test followed by Dunn's multiple comparisons test, 4 independent experiments).
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  (b) Detection of HA-SorCS1 by western blot in total brain extracts prepared from Sorcs1 HA/HA KI mice (P60).
Total protein staining by using Ponceau method was used as loading control.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  (d) Quantification of (c): internalized SorCS1 fluorescence intensity relative to total levels and normalized to cells expressing WT-SorCS1; surface SorCS1 fluorescence intensity relative to total levels and normalized to cells expressing WT-SorCS1. WT (n = 28 neurons); Y1132A (n = 30). ***p < 0.001 (Mann-Whitney test,

independent experiments). Graphs show mean ± SEM.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.   (g and h) Quantification of (d and f): internal and surface Nrxn1α fluorescence intensity in dendrites and axons relative to total levels in 3 independent experiments (n = 30 neurons for each group).
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.   . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. (j) Quantification of (i); Ctr (n = 30 neurons); T22N (n = 29).
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; (HDR) allowed for precise HA-tagging of the Nrxn1 locus right after the signal peptide (SP). An AatII restriction site was introduced in the DNA sequence coding for the HA-tag, by taking advantage of the redundancy of the genetic code, in order to facilitate the distinction between homozygous and heterozygous . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. (c) Quantification of (a and b): surface L1 and Caspr2 fluorescence intensity in axon and dendrites relative to total surface levels and normalized to cells expressing EGFP and ratio of axonal/dendritic surface L1 and Caspr2 intensity. Ctr_L1 (n = 29 neurons); Cre_L1 (n = 26); Ctr_Caspr2 (n = 28); Cre_Caspr2.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.    Cargo destined for either axon or dendrites is sorted via Rab11-REs and transported to its respective final destination 3,4 . SorCS1/Rab11FIP5 sorting complex localizes to EEs and sorts newly internalized Nrxn into cytoskeleton-associated vesicles that carry cargo from EEs to Rab11-REs. Nrxn-containing Rab11-REs are subsequently transcytosed from dendrites to the axon. In Sorcs1 KO cells, Nrxn is not sorted to Rab11-REs, accumulating in EEs and mis-trafficking to Rab4-REs. Consequently, axonal surface levels of Nrxn are decreased and dendritic surface levels are increased, which is likely due to decreased sorting of Nrxn to Rab11-REs and increased recycling of Nrxn from Rab4-REs back to the dendritic plasma membrane, respectively.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. (b) Estimation of the RRP size used during neuronal activity by back-extrapolating a linear fit of the steady state current towards the ordinate axis intercept, which represents the initial RRP size before train stimulations (10 Hz) 5 . RRP size of active synapses was reduced in DIV14-DIV16 Sorcs1 KO neurons.
(c) Kinetics of activity dependent depletion of the RRP, using repetitive train stimulations (10 Hz). We normalized eEPSC synchronous charge -to assess release kinetics independent of the difference in RRP pool size -and calculated the initial eEPSC charge before stimulation. DIV14-DIV16 Sorcs1 KO neurons showed a smaller initial eEPSC charge, suggesting a reduced efficacy of calcium dependent synaptic vesicle release during repetitive stimulation. Control (n = 18 neurons); Cre (n = 13). *p < 0.05 (Mann-Whitney test, 3 independent experiments).
Graphs show mean ± SEM.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019;    Co-culture assays were performed as previously described 10   (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Primary neuronal cultures for imaging
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; and Technology, South Korea); and myc-tagged L1 (Dan P. Felsenfeld, CHDI Foundation, USA). All DNA constructs used in this study were verified by sequencing.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Surface immunostaining of endogenous and exogenous extracellular HA-tagged Nrxn1α and exogenous HAtagged Caspr2
For surface immunostaining of exogenous HA-Nrxn1α and HA-Caspr2, live mouse cortical and hippocampal neurons were incubated with rabbit anti-HA (1:1000 dilution; Sigma-Aldrich, Cat #H6908) diluted in conditioned neuronal culture medium for 15 min at room temperature. For surface immunostaining of endogenous HA-Nrxn1α live mouse cortical

Nrxn1α
HA/HA KI neurons were incubated with rabbit anti-HA (1:100 dilution; Cell Signaling Technology, Cat #3724) diluted in conditioned neuronal culture medium for 20 min at room temperature. Neurons were then fixed in 4% (wt/vol) sucrose and 4% (wt/vol) paraformaldehyde in PBS for 10 min at room temperature, followed by several washes in PBS and blocking in 10% (wt/vol) BSA in PBS for 1 hr at room temperature. Neurons were then incubated with anti-rabbit secondary antibody diluted in 3% (wt/vol) BSA in PBS (1 hr, room temperature). Following permeabilization, neurons were processed for immunocytochemistry as described above.

Surface immunostaining of extracellular myc-tagged L1
To allow surface immunostaining of myc-L1 (live labeling proved to be impossible), neurons were fixed first, followed by several washes and blocking, and then incubated with mouse anti-myc (1:1000 dilution; Santa Cruz Biotechnology, Cat #sc-40) diluted in 3% (wt/vol) BSA in PBS overnight at 4 °C. Subsequently, neurons were incubated with the respective secondary antibody for 1 hr at room temperature, and processed for immunocytochemistry as described above.

Antibody pulse-chase experiments
Cultured living neurons were incubated at room temperature for 10 min in the presence of a high concentration (1:250) of a mouse anti-HA antibody (Covance, Cat #MMS-101P), against extracellular HA-tagged Nrxn1α and SorCS1, diluted in conditioned medium. Neurons were then washed with pre-warmed PBS at 37 °C to remove the unbound antibody, and were further incubated in antibody free conditioned medium in a 37 °C, 5% (vol/vol) CO 2 /95% (vol/vol) air incubator (for different periods) to allow the internalization of antibody-bound receptors. After this incubation, neurons were fixed in 4% (wt/vol) sucrose and 4% (wt/vol) paraformaldehyde in PBS for 10 min at room temperature.
Next, neurons were either exposed to a super-saturating concentration (1:300) of the first of two secondary antibodies, to label the primary antibody-bound surface pool of protein, and/or incubated overnight with Fab fragments anti-mouse [0.25 mg/mL (Santa Cruz Biotechnology, Cat #715-007-003) in 5% (wt/vol) BSA in PBS] to block all primary antibodybound receptors that were not internalized and/or not labeled by the first secondary antibody. After permeabilization, cells were processed for immunocytochemistry as described above, and the pool of internalized receptors was labeled by incubation with the second secondary antibody (1:1000) for 1 hr at room temperature. This strategy allows differential labeling of cell surface and internalized pools of protein.

Brefeldin A (BFA) assay
Mouse cortical neurons expressing extracellular HA-tagged Nrxn1α for 4 hr were treated with 0.75 μg/mL brefeldin A (Sigma-Aldrich, Cat #B5936) added to the coverslips in a 12-well plate containing conditioned neuronal culture medium. After 16 hr of treatment, coverslips were washed twice in pre-warmed neuronal culture medium, transferred to their original dishes containing conditioned neuronal culture medium and fixed for 10 min in 4% (wt/vol) sucrose and . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/552695 doi: bioRxiv preprint first posted online Feb. 17, 2019; 4% (wt/vol) paraformaldehyde in PBS at various intervals thereafter (1, 2, and 4 hr after BFA washout). Internal and surface pools of Nrxn1α were then differentially labeled. First, neurons were blocked in 10% (wt/vol) bovine serum albumin (BSA) in PBS for 1 hr at room temperature and incubated (overnight, 4 °C) with mouse anti-HA (1/1000) (Covance, Cat #MMS-101P) diluted in 3% (wt/vol) BSA in PBS. After, several washes neurons were incubated with super-saturating concentration (1/300) of anti-mouse secondary antibody diluted in 3% (wt/vol) BSA in PBS 1 hr at room temperature. Remaining unlabelled primary antibody-bound surface HA-Nrxn1α was blocked with overnight incubation with Fab fragments anti-mouse [0.13 mg/mL (Santa Cruz Biotechnology, Cat #715-007-003) in 5% (wt/vol) BSA in PBS]. After permeabilization, cells were processed for immunocytochemistry as described above to label the axonal and somatodendritic compartments and the pool of internal HA-Nrxn1α, labeled with an incubation with the same anti-HA (1/1000) primary antibody, followed by incubation with the respective secondary antibodies for 1 hr at room temperature. This strategy allows differential labeling of cell surface and internal pools of HA-Nrxn1α.

Dynasore treatment
Mouse cortical or hippocampal neurons expressing extracellular HA-tagged Nrxn1α for 30 hr were either treated with 20 µM Dynasore (Sigma-Aldrich, Cat #D7693) or DMSO, added to the coverslips in a 12-well plate containing conditioned neuronal culture medium, for 18 hr. Afterwards, surface HA-Nrxn1α was labeled in live neurons as described above.

Image Analysis and Quantification
All images obtained from immunocytochemistry experiments in fixed cells were captured on a Leica SP8 laserscanning confocal microscope (Leica Micro-systems). The same confocal acquisition settings were applied to all images taken from a single experiment. Parameters were adjusted so that the pixel intensities were below saturation.
Fiji analysis software was used for quantitative imaging analysis. Z-stacked images were converted to maximal intensity projections and thresholded using constant settings per experiment.

Quantification of axonal and dendritic immunofluorescence intensity and ratio of axonal/dendritic immunofluorescence intensity
Fluorescence intensity was measured as the sum of integrated intensity in representative portions of axons and dendrites using Ankyrin-G and MAP2 as guides, respectively. Axonal and dendritic intensities were divided by neuritic length and total intensity (axonal + dendritic) ('Relative Axonal Intensity' and 'Relative Dendritic Intensity') to adjust measurements across cells with varying expression levels (exogenous expression). The A/D ratio was then calculated by dividing the values of axonal and dendritic intensities obtained for every cell ('Axonal/Dendritic Intensity'). A uniformly distributed protein yields an A/D ratio of around 1. A preferentially dendritically localized protein yields an A/D ratio < 1, whereas a preferentially axonally localized protein yields an A/D ratio > 1. To quantify the fluorescence . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Classical polarity index measurement
In some cases polarization of cargo was determined by using the classical polarity index originally described by 22,23 .
This index does not provide a direct measurement of the axonal and dendritic fluorescent intensities, but it allows a quick estimation of the compartmentalized polarization of proteins of interest. One-pixel-wide lines were traced along three dendrites and representative portions of the axon, using MAP2 and Ankyrin-G as guides, in non-thresholded images. The mean intensities (dendritic was averaged from three dendrites) were used to calculate the dendrite:axon (D:A) 'Polarity Index'. D:A = 1, uniform staining; D:A < 1, preferential axonal staining; D:A > 1, preferential dendritic staining.

BFA-assay and long-term live-cell imaging
Fluorescence intensity was measured as the sum of integrated intensity in representative portions of axons and dendrites (using Ankyrin-G and MAP2 as guides); and cell bodies. In the case of long-term live-cell imaging, axonal and dendritic intensities were divided by neuritic length, and somatic intensities by somatic area ('Nrxn Intensity' or 'TfR Intensity'). In case of the BFA-assay, axonal and dendritic intensities were divided by neuritic length and total intensity [surface (axonal and dendritic) + internal (axonal and dendritic)] ('Relative Axonal Intensity' and 'Relative Dendritic Intensity').

Antibody pulse-chase experiments
Fluorescence intensity was measured as the sum of integrated intensity in representative portions of axons and dendrites using Ankyrin-G and MAP2 as guides, respectively. Intensities of internalized SorCS1 and Nrxn were divided by neuritic length and by the total intensity (surface + internal) ('Relative Intensity of Internalized Nrxn' or 'SorCS') or by the total internalized intensity (axonal + dendritic) ('Relative Intensity of Internalized Nrxn'). Intensities of surface SorCS1 were also divided by neuritic length and by the total intensity (surface + internal) ('Relative Intensity of Surface SorCS1').

Manders coefficient
Manders coefficient was measured by using the Fiji plugin JACoP 24 . Manders coefficient measures the proportion of the signal from channel 'a' that coincides with the signal in channel 'b' over the total intensity of 'a' -M1 coefficient 25 .
For our measurements internalized Nrxn was defined as channel 'a' and endosomal markers ( Figure S3) or SorCS1 ( Figure S6) as channel 'b'; and Rip11 as channel 'a' and SorCS1 as channel 'b' (Figure 6).

Colocalization of internalized Nrxn with endosomal markers in Sorcs1 KO cells
The colocalization of internalized Nrxn with endosomal markers was evaluated by measuring the density of doublepositive puncta for internalized Nrxn and endosomes and by measuring the intensity of Nrxn present in these puncta.
Fluorescence intensity was measured as the sum of mean intensity of internalized Nrxn puncta (defined as 0.02 µm 2infinite) in representative portions of axons and dendrites (using Ankyrin-G and MAP2 as guides), colocalizing with endosomal markers. Intensities of internalized Nrxn were divided by neuritic length ('Nrxn Intensity in Double Positive . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Live-Cell Imaging
For live-cell imaging of fluorescently tagged Nrxn1α-and TfR-positive vesicles, neurons were transfected with a bicistronic expression plasmid encoding Streptavidin-KDEL and SBP-GFP-Nrxn1α or Streptavidin-KDEL and TfR-SBP-GFP, respectively, using the RUSH system 6

Live labeling of the axon initial segment (AIS)
Before every live-cell imaging experiment primary rat cortical cultured neurons were live-labeled with an anti-pan-Neurofascin antibody (NeuroMab, Cat #75-172) to distinguish the axon from the dendrites. Live-cell imaging experiments were performed with rat cortical neurons because live-labeling of the AIS with the anti-pan-Neurofascin antibody did not work in mouse cultured neurons. Briefly, coverslips with neurons were quickly rinsed in pre-warmed neuronal culture medium. Neurons were incubated with anti-pan-Neurofascin antibody (1/500) diluted in conditioned neuronal culture medium for 10 min in a 37 °C, 5% (vol/vol) CO 2 /95% (vol/vol) air incubator. Coverslips were then quickly washed twice with pre-warmed neuronal culture medium. Finally, neurons were incubated with an Alexa-555 anti-mouse (1/400) (Invitrogen, Cat #A31570) secondary antibody diluted in conditioned neuronal culture medium and incubated for 10 min in a 37 °C, 5% (vol/vol) CO 2 /95% (vol/vol) air incubator. Neurons were quickly rinsed twice with pre-warmed neuronal culture medium and used for live-cell imaging experiments.

Analysis of Nrxn1α-and TfR-positive vesicle transport
Dendrites and axon were imaged from the same neuron using spinning disk confocal microscopy. Time-lapses were performed by sequential capture of 200-ms images every second for 60-120 s. Acquisitions were analyzed using an . CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. 10 HEPES, 10 Glucose (300 mOsm, pH 7.30). Cells were whole-cell voltage clamped at −70 mV with a double EPC-10 amplifier (HEKA Elektronik) under control of Patchmaster v2x32 software (HEKA Elektronik). Currents were lowpass filtered at 3 kHz and stored at 20 kHz. Patch pipettes were pulled from borosilicate glass using a multi-step puller (P-1000; Sutter Instruments). Pipette resistance ranged from 3 to 5 MΩ. The series resistance was compensated to ~75%. Only cells with series resistances below 15 MΩ were included for analysis. All recordings were made at room temperature. Spontaneous glutamatergic release was (sEPSC) was recorded at -70 mV. Evoked release was induced using brief depolarization of the cell soma (from 70 to 0 mV for 1 ms) to initiate action potential-dependent glutamatergic release (eEPSCs). A fast local multi-barrel perfusion system (Warner SF-77B, Warner Instruments) was used determine the RRP size using external recording solution containing 500 mM sucrose. A custom analysis procedure in Igor Pro (Wavemetrics Inc.) was used for offline analysis of evoked and sucrose responses. Spontaneous events were detected using Mini Analysis program (Synaptosoft).

Statistical Analysis
The results are shown as average or as average ± s.e.m., with n referring to the number of analyzed neurons for each group. For most experiments at least 3 independent cultures were included for analysis. Datasets were tested either using Mann-Whitney U test or Kruskal-Wallis test by Dunn's multiple comparisons test. Statistical testing was performed using GraphPad Prism (GraphPad Software). In all instances, statistical significance was defined as follows: n.s. -not significant (p > 0.05), * p< 0.05, ** p < 0.01, *** p < 0.001.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

SUPPLEMENTARY VIDEOS and LEGENDS
Supplementary Movie 1 DIV8 WT rat cortical neuron co-expressing the ER-hook (Streptavidin-KDEL) and SBP-GFP-Nrxn1α (grayscale, inverted for clarity) and live-stained for the AIS marker Neurofascin to label the axon. Biotin was added 10 min after the beginning of the imaging session. Cell was recorded every 5 min for 2,5 hr. The axon is indicated. Frame rate: 2 fps.

Supplementary Movie 2
DIV9 WT rat cortical neuron co-expressing the ER-hook (Streptavidin-KDEL) and TfR-SBP-GFP (grayscale, inverted for clarity) and live-stained for the AIS marker Neurofascin to label the axon.
Biotin was added 10 min after the beginning of the imaging session. Cell was recorded every 5 min for 2,5 hr. The axon is indicated. Frame rate: 2 fps.

Supplementary Movie 3
DIV10 WT rat cortical neuron co-expressing the ER-hook (Streptavidin-KDEL) and SBP-GFP-Nrxn1α (grayscale, inverted for clarity) and live-stained for the AIS marker Neurofascin to label the axon. Cell was recorded every second for 30 s. The axon is indicated. Frame rate: 4 fps.

Supplementary Movie 4
DIV3 WT rat cortical neuron co-expressing the ER-hook (Streptavidin-KDEL) and SBP-GFP-Nrxn1α (grayscale, inverted for clarity) and live-stained for the AIS marker Neurofascin to label the axon. Biotin was added 10 min after the beginning of the imaging session. Cell was recorded every 5 min for 2,5 hr. The axon is indicated. Frame rate: 2 fps.
. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.