Sec3 exocyst component knockdown inhibits axonal formation and cortical neuronal migration during brain cortex development

Neurons are the largest known cells, with complex and highly polarized morphologies and consist of a cell body (soma), several dendrites, and a single axon. The establishment of polarity necessitates initial axonal outgrowth in concomitance with the addition of new membrane to the axon's plasmalemma. Axolemmal expansion occurs by exocytosis of plasmalemmal precursor vesicles primarily at the neuronal growth cone membrane. The multiprotein exocyst complex drives spatial location and specificity of vesicle fusion at plasma membrane. However, the specific participation of its different proteins on neuronal differentiation has not been fully established. In the present work we analyzed the role of Sec3, a prominent exocyst complex protein on neuronal differentiation. Using mice hippocampal primary cultures, we determined that Sec3 is expressed in neurons at early stages prior to neuronal polarization. Furthermore, we determined that silencing of Sec3 in mice hippocampal neurons in culture precluded polarization. Moreover, using in utero electroporation experiments, we determined that Sec3 knockdown affected cortical neurons migration and morphology during neocortex formation. Our results demonstrate that the exocyst complex protein Sec3 plays an important role in axon formation in neuronal differentiation and the migration of neuronal progenitors during cortex development.

2002). Diverse studies in yeast and mammals have determined that the exocyst plays a key role in the polarized membrane traffic. It is well established that a tethering complex, the exocyst complex (EC), drives spatial location and specificity of vesicle fusion at the plasma membrane where the PPVs are directed (Grindstaff et al., 1998;Hazuka et al., 1999;Hsu et al., 1999;Martin-Urdiroz et al., 2016;TerBush & Novick, 1995). The exocyst is a well conserved octameric complex usually comprised by Sec3,Sec5,Sec6,Sec8,Sec10,Sec15,Exo70,and Exo84. While several studies have been made to determine the structure of the complex as well as the assembly process in several cell types (Lepore et al., 2018;Mei & Guo, 2018, 2019Munson & Novick, 2006;Picco et al., 2017), it still remain unclear the role of the EC components in the neuron.
Since the neuron is a highly polarized cell which requires extremely rapid membrane expansion during differentiation, the participation of the EC could be critic for vesicle delivery and exocytosis for axonal outgrowth.
Only a few components of the EC have been studied using in vitro models giving a glimpse of the major role in polarity of this octameric complex. In this context, we have previously reported that, in cultured hippocampal neurons, (a universally employed method to study neuronal development and polarization), TC10 (a small GTP binding protein) activation by IGF-1 and the consequent assembling of Exo70 to the growth cone plasmalemma are critical for the control of PPVs exocytosis and, hence, pioneer axonal outgrowth and the establishment of neuronal polarity (Dupraz et al., 2009). Also, a NGF-induced Exo70-TC10 complex has been shown to modulate neurite outgrowth in PC12 cells (Pommereit & Wouters, 2007). It has been also shown by other colleagues that, exocyst proteins, such as Sec6 and Sec8, are concentrated at growth cones of hippocampal neurons Lalli & Hall, 2005;Vega & Hsu, 2001). Moreover, cortical neurons expressing undetectable levels of Sec6, Sec8, and Exo84 failed to grow a fully developed axon (Lalli, 2009). Additionally, the EC (especially the Sec6, Sec8, and Exo84 proteins) interacts with the polarity complex (Par3/Par6), a major controller of neuronal polarization (Das et al., 2014;Lalli, 2009).
However, besides the references cited before, it remains to be determined which of the EC proteins are essential components that could participate in the regulation of initial axonal outgrowth and the establishment of neuronal polarity. It has not yet been determined if all the eight EC components have any influence in polarization acquisition. Between the proteins of this octameric complex, we decided to study the Sec3 protein since several reasons suggested a critic role of this particular protein on the regulation of neuronal polarity.
In yeast, Sec3 is placed in the bud tip even in the absence of actin cables, which are necessarily to the polarized exocytosis and is the only EC component capable of recruiting vesicles ectopically to surrogated organelles (Finger et al., 1998;Luo et al., 2014). This activity may be achieved because of the spatial localization of this protein in the EC since Sec3 binds to the plasma membrane, acting as anchor to the other components of the EC and the vesicles (Baek et al., 2010;Pleskot et al., 2015). In mammals the exocyst in a highly dynamic structure formed by two subcomplexes and Sec3 is necessary for the formation of one of these subcomplexes (Ahmed et al., 2018).
Besides the structural role, Sec3 may also be involved in the regulation of the vesicle fusion, since it is weakly attached to the complex and departs from the complex earlier presumably to allow the recruiting of the SNARE complex (Ahmed et al., 2018). The relationship between the EC and the SNARE complex is poorly understood, but it has been described that the N terminal domain of Sec3 binds to the Snare Sso1/2 and increase the rate of fusion (Mei & Guo, 2019;Yue et al., 2017). On the other hand, just few in vivo experiments have been performed but strikingly indicate the potential central role that Sec3 may have on this process, since homozygous mice with Sec3 null mutation suffer peri-implantation lethality (Mizuno et al., 2015).
Altogether, these results point out that the exocyst complex could be critical for neurite development and, hence, neuronal polarization and also strongly suggests that Sec3 has an important regulatory activity on exocyst function, thus, in membrane expansion which in essential for the polarization process.
Despite the structural studies about Sec3 and functional studies of different proteins of the exocyst that have been performed, the possible direct participation of this protein in neuronal axonal formation has yet to be studied. Here, we examined the expression and distribution of Sec3 in hippocampal neurons in culture and its participation in the establishment of neuronal polarity. Our results showed that Sec3 was expressed early in developing neurons and present in the growth cone and distal axon, prominent sites for new membrane addition. In these cells, loss of function of Sec3 repressed the establishment of neuronal polarity. Using in utero electroporation we also studied the consequences of Sec3 loss of function on pyramidal cortex neuron migration and changes in polarity during cortical formation. Neurons electroporated with a shRNA targeting Sec3 failed to migrate to the upper cortical layers and accumulated mainly at the ventricular/subventricular zones. Knocking down Sec3, also abrogated the morphological change from multipolar to bipolar and cells were arrested as multipolar forming an atypical tissue organization. In summary, the results reported in this study show that Sec3 is necessary for axonal outgrowth and the establishment of neuronal polarity in hippocampal neurons in culture and the early polarity switch of cortical plate neurons during cortical formation in utero.

| Animals
All animal procedures were done according to the "Guide for the care and use of laboratory animals" (8th Edition, USA), Federation were achieved in animals from 2 to 6 months of age. Animals were under standard cages and companion, ad libitum access to food and water, with a 12-h light/dark cycle and controlled humidity (40-70%) and temperature (20-26°C). A total of 40 pregnant young female mice were used providing each one between 4 and 10 embryos. At least three female mice and their embryos were used to carry on each experiment. Each female pregnant animal is treated as an independent experiment. For in vitro experiments one female was used for each experiment dissecting hippocampus from at least four embryos. Two animals died during/after in situ experiments.

| Hippocampal culture and transfection
Dissociated hippocampal pyramidal neurons were prepared from fetal mouse brain and cultured as described previously in (Banker & Cowan, 1977). Briefly, pregnant C57BL/6CIQBC mice were asphyxiated by CO 2 at 18 days of post-fertilization and subsequent cervical dislocation. The embryos were recovered creating an opening in the mid-ventral side of the mouse with large sterile surgical scissors removed from their individual placenta and then decapitated with sterile scissors in a laminar flow hood. The removed heads were placed in plates with sterile Hank's Balance Salt Solution (HBSS, Sigma-Aldrich, RRID:SCR_008988) at 4°C under a dissecting microscope in order to remove the brain, grasp the meninges and remove the hippocampus. Isolated hippocampus was enzymatically dissociated with trypsin (0.25% w/v for 18 min at 37ºC Life Technologies, RRID: SCR_008817) followed by mechanical trituration with firepolished Pasteur pipettes. After resuspension the cells were plated immediately onto polyl-lysine (1 mg/ml Sigma-Aldrich, RRID: SCR_008988) coated glass coverslips and maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco Thermo Fisher Scientific, RRID: SCR_008452) in a rich serum environment completed by adding 10% horse serum (Gibco Thermo Fisher Scientific) for 2 h.
For those experiments involving expression of shRNA sequences in early stages of neuronal development, a protocol of transfection of neurons in suspension, before plating, was used, similar to the procedure described in (Halterman et al., 2009)  were placed to 37°C in a humidified 5% CO 2 incubator. After 1.15 h, transfecting complex was removed carefully from each coverslip (at this time most of neurons were already attached) and serum-free medium plus the N2 mixture was added to cultures. Cultures were placed in a humidified 37°C incubator with 5% CO 2 for 24-48 h before fixation. Shortly after plating, hippocampal neurons first extend lamellipodia (stage 1) and afterward several minor neurites that are initially indistinguishable (stage 2). Then, at stage 3, one of these initially equivalent neurites grows more rapidly than the others and becomes the axon, whereas the other neurites subsequently develop into dendrites (stage 4). Neurons are considered to be at stage 3 when the length of the axon exceeds that of the average minor neurite by at least 20 μm.

| Line cell culture
For sh validation frozen vials of Neuro-2a mus Musculus (ATCC® CCL-131™) were recovered. Cells were cultured in DMEM (Dulbecco's modified Eagle medium) supplemented with 10% of fetal bovine serum (FBS) (Gibco Thermo Fisher Scientific) and maintained in a humidified 37°C incubator with 5% CO 2 . Transient transfection of cultured neurons was performed as described previously, and the constructs used at concentration of 1.5 or 2.5 μg per 60 mm dish (Rosso et al., 2005).

| Isolation of growth cones
Axonal growth cones were isolated from developing brain as described previously (Letinic et al., 2009;Liu et al., 2014). In brief, brains of 18 days of post-fertilization embryos were homogenized (H). A low-speed supernatant (LSS) was prepared, loaded onto a discontinuous sucrose density gradient with steps of 0.83, 1 and 2.66 M sucrose, and spun to equilibrium at 242 000 g. The fraction at the load/0.83 m interface (designated "A") contained the isolated growth cones or growth cone particles (GCPs).

| Immunofluorescence microscopy
Neurons were fixed for 20 min at 25°C in phosphate-buffered saline (PBS) solution containing paraformaldehyde (4% w/v Sigma-Aldrich, RRID: SCR_008988) and sucrose (4% w/v). After a thorough washing, the cultures were permeabilized with Triton X-100 (0.2% v/v Sigma-Aldrich, RRID:SCR_008988) in PBS for 6 min. A 1 h blocking step at 25°C was done before labeling with primary antibodies (overnight at 4°C). After subsequent washes with PBS, cultures were incubated with fluorescent secondary antibodies conjugated to Alexa Fluor 488, 546, or 633 (1 h at 25°C). Coverslips were mounted in slides. The membranes were rinsed with Tris-buffered saline (TBS) (Tris, pH 7.5 10 mM, NaCl 150 mM) and 1 h blocking step at 25°C in TBS containing 5% w/v non-fat milk was done before labeling with primary antibodies (overnight at 4°C in TBS containing 1% w/v non-fat milk).
The embryos were placed back into the abdominal cavity and the abdominal cavity was sutured. After this procedure, an analgesic, tramadol (5 mg/kg. Finadiet), was administered. Female mice were allowed to recover from anesthesia on a warm plate.

| Immunohistochemistry
Eighteen days post fertilization and also 3 days after IUE, mice were anesthetized with 90 mg/kg Ketamine (Holliday) and 10 mg/kg Xylazine (Richmond) mixture and subsequent cervical dislocation.
The embryo brains were removed and post-fixed in paraformalde- The brains were cryoprotected by overnight immersion in sucrose (30% w/v in PBS) and embedded in OCT. Then, the brains were cryopreserved using liquid nitrogen and store at −80°C for 2 days. UPLANSAPO 0.75 NA or 40× UPLANSAPO 0.9. For 3D z-stack 10-15 confocal sections were collected every 2 μm step size. Images were analyzed using Fiji software (NIH). All images were processed, and the graphic art was prepared using Adobe PhotoShop (Adobe Photoshop 13.0 version 6.1, RRID: SCR_014199 Adobe Systems).

| Statistical analysis and study design
Statistical analyses were performed using GraphPad Prism software (GraphPad Prism version 5.00 for Windows, GraphPad Software, San Diego California USA GraphPad Prism RRID: SCR_002798).
Data are presented as mean ± SEM, as indicated in the figure legends. In all the cases, at least three independent experiments from different female mice were carried out. No sample calculation was performed; the maximum possible number of subjects was used in each study. The number of subjects allocated in each group was determined following previous published data (Nieto Guil et al., 2017;Sosa et al., 2006). design was not pre-registered. No randomization was performed to allocate subjects in this study. No blinding was performed. No sample calculation was performed; the maximum number of subjects possible was used in each study. The study was exploratory.
No exclusion criteria were pre-determined. We did not run tests to identify outliers and no data points were excluded from the analysis.
All animal experiments were made between 9 a.m. and 1 p.m.

| Expression and distribution of Sec3 in developing neurons
In order to be involved in neuronal polarization, a protein should be expressed early before this phenomenon occurs, so we studied whether Sec3 protein is expressed in cultured hippocampal pyramidal neurons at early stages of development. In this model, cells undergo stereotypical changes in the morphology from round cells with lamellipodial structures (stage 1), growing to symmetrical or not polarized cells with equivalent projections (stage 2) (between 12 and 18 h of differentiation in vitro (H.I.V)) and, later, to exhibit a discernible axon (20-24 h in culture) showing from that moment a polarized morphology (stage 3). At stage 4 the dendrites begin to develop and finally, at differentiation stage 5, the neuron suffers axonal and dendritic arborization leading to a maturation process. Stages 4 and 5 are out of the scope of the present work because the polarization has already occurred at this time (Dotti et al., 1988). A graphical time line is included in Figure 1a Figure 1b showed that besides the expected band of Sec3 (apparent molecular weight 102 kDa) a lower band was observed. As this pattern was observed in all the western blots analyzed it may represent alternative splicing products since this protein has been described to have at least eight different isoforms (UniProt, 2021) Also this double-band pattern has formerly found in several cell types (Arasaki et al., 2018;Lin et al., 2018). Full length gels are showed in Figure S2a).
Preceding data exhibit that, in neurons, some components of the exocyst such as Sec6, Sec8, and Exo70 are localized in the axon and axon growth cones of hippocampal neurons; but it has not been determined specifically for Sec3 (Dupraz et al., 2009;Hazuka et al., 1999;Vega & Hsu, 2001). The immunofluorescence of the stage3 neuron showed in Figure 1a suggests an increased level of immunolabeling of Sec3 in the axon, especially in the middle region of the axon. Since Tau1 is a microtubule-associated protein which is known to be enriched at the distal axons, we decided to evaluate the co-distribution with this protein. Thereby, stage 3 neurons were stained with Sec3 (green channel) and Tau1 antibody (red channel) (Figure 3a-top). As we can observe in the magnification figure, more yellow pixels are seen at the neuronal axon, suggesting a co-distribution between these two proteins (Figure 3a-bottom). We also determined the intensity profile of the green and red channel along the axon, selected in the picture (Figure 3b-left). The histogram reveals a high spatial concordance between the two antibody marks (Figure 3b To determine the presence of Sec3 at the growth cone, a neuronal domain that leads the axonal growth and a prominent site for the addition of new membrane in developing neurons, was essential (Craig & Banker, 1994;Futerman & Banker, 1996;Pfenninger & Friedman, 1993); so we then evaluated the expression of Sec3 at the growth cone. Hippocampal cells were examined by immunofluorescence using Sec3 and Tyrosinated Tubulin antibody performed at late stage2/early stage 3 (when the growth cone is more prominent) showing a robust expression of the former at the growth cone (Figure 3d). Tyrosinated Tubulin was selected because it has been described to be associated with highly dynamic structures and would assess the cytoskeleton dynamics at the growth cone accurately (Arregui et al., 1991). Furthermore, by performing fluorescent microscopy and imaging colocalization analysis, we demonstrated that Sec3 has a strong co-distribution with the cytoskeletal protein Tubulin (Figure S1b-d). The presence of this protein at the growth cone was also analyzed by subcellular fractioning of fetal brain in vesicles with ultrastructural similarities to the growth cone called growth cone particles (GCPs) (Lohse et al., 1996;Pfenninger et al., 1983). Our results showed that Sec3 is present in the GCPs (Figure 3e). Full length gels are showed in Figure S2b). These results further confirmed that Sec3 is present in important structures for the establishment of polarity such as the axon and the growth cone.

| Sec3 is necessary for the in the establishment of neuronal polarity
Further investigation regarding the possible involvement of Sec3 on the establishment of neuronal polarity was conducted.
We silenced the expression of Sec3 protein in mice hippocampal neuron cultures using targeted shRNAs inserted into biscistronic plasmids which has a second promoter within the same vector that expressed enhanced green fluorescent protein (GFP).
When the polarity is impaired, the neurons remain mostly at early stages (1 and 2), lack of one long prolongation (the axon-at least 20 µm larger than the others prolongations) and the absence of H.I.V. Also we relate co-transfected neurons with Sec3FL and shSec3 or ssCtrl. We found that over 80% of the Sec3-targeted shRNAs transfected neurons remained at stages 1 or 2 of differentiation, and only 12.63 ± 1.83% of neurons had formed a discernible axon. In contrast, over 52.28 ± 2.25% of the control cells exhibited a relatively long Tau-1-containing axon. Regarding the cells co-transfected with shSec3 and Sec3FL or ssCtrl and Sec3FL about 35.10 ± 3.08% and 40.20 ± 2.25% of the cells, respectively, exhibited a discernible Tau-1-enriched axon (Figure 4d).

| Sec3 is necessary for the polarized insertion of the IGF1 receptor
One of the earliest events in the differentiation of hippocampal neurons is the enrichment of phosphorylated (active) insulin like growth factor type 1 receptor (IGF-1r) in a single minor neurite.
Upon IGF1r phosphorylation, cells do not exhibit a noticeable axon (stage 2 of development) nonetheless the future axon is discernible since the active receptor is specifically segregated to one neurite (Sosa et al., 2006). Hence, we next analyzed the distribution apparent molecular weight 102 kDa) and tubulin (second row) as a loading control and secondary antibodies suitable for near infrared fluorescence. Note that besides the expected band of Sec3 (apparent molecular weight 102 kDa) a lower band was observed. As this pattern was observed in all the western blots analyzed it may represent alternative splicing products since this protein has been described to have at least eight different isoforms (UniProt, 2021) Also this double-band pattern has formerly found in several cell types (Arasaki et al., 2018;Lin et al., 2018). The obtained images were converted to gray scale and inverted. The results of one of three independent experiments are shown. The blots belong to the same experiment but were cropped for a better understanding. (c) Quantification of western blot for Sec3. Bars represent mean ± SEM of the density measured in each band relative to tubulin density. A two-way ANOVA Test with Bonferroni posttest was performed. *p ≤ 0.05 n = 3 independent gels from 3 different cell cultures. Merged panels were pseudocolored (Figure 5b-top) labeling of the active IGF-1r was not enriched in any particular minor process. The active IGF-1r polarization index (for a description see Figure 4 legend) was significantly higher (p ≤ 0.001) in the neurons transfected with the ssCtrl (Mean value P.I. ± SEM = 2.14 ± 0.17) sequence than in the Sec3-suppressed neurons (P.I. = 1.42 ± 0.05) (Figure 5c). Taken together, these results indicate that Sec3 is essential for axonal growth and the establishment of polarity in hippocampal neurons.

| Sec3 is necessary for the regulation of cortical neuron migration
Changes in polarity and morphology have been largely analyzed in cells in culture, so we wanted to verify our findings in an in situ model. The most studied in situ system for neuronal differentiation and polarity is to measure cortical neurons migration using in utero electroporation. As cortex formation proceeds, newborn neurons in the ventricular zone (VZ) migrate radially to reach the cortical plate (CP). Using in utero electroporation experiments it was described that the EC proteins Exo70 and Exo84 harm neuronal migration (Guo et al., 1999;Letinic et al., 2009). We next investigated a possible role of Sec3 in cortical migration by using in utero electroporation of cortical progenitors at embryonic day (E) 14.5 to visualize neurons destined to comprise layers II-IV of the cortex, allowing analysis of the location and morphology of the progeny after further 3 days of in situ development. We used a plasmid encoding the shRNA-Sec3 sequence co-electroporated with a plasmid encoding a red fluorescent protein (DsRed) containing a CAG promoter. In this way, the co-electroporation efficiency is high while we can easily follow the electroporated cells (because the CAG promoter has high level of expression) lasting until the shRNA sequence is expressed (Liu et al., 2014;Saito & Nakatsuji, 2001). The co-electroporation   Figure 7b). This demonstrates that the migration is impaired when Sec3 is suppressed.

| Sec3 is necessary for the polarity switch in cortical neurons
During radial migration of cortical neurons many multipolar neurons can be observed at the lower intermediate zone. These cells extend and retract thin processes at random directions not following the radial glia (Bielas et al., 2004;Noctor et al., 2004;Tabata & Nakajima, 2003). After this stage, as cells approach the middle of the intermediate zone one neurite grows faster to form the leading process (the future apical dendrite), followed by the appearance of a trailing process (the future axon), acquiring a bipolar morphology. (Yogev & Shen, 2017).
After the axon emerges, the centrosomes and Golgi of the neurons get reoriented toward the cortical plate surface, as they move to the upper part of the intermediate zone (Gupta et al., 2002). We next analyzed if the EC protein Sec3 participates in the early transition from multipolar to bipolar morphology. Animals were electroporated at E14.5 and observed at E18. At this time, previous studies have shown that most cells in the upper intermediate zone and cortical plate are bipolar (Bai et al., 2003;Namba et al., 2014). Our results show that in control conditions 76.76 ± 1.42% of the neurons migrating through the middle intermediate zone exhibited a bipolar morphology, as defined by the absence of more than two projections (Figure 8a-Top, arrow, quantification F I G U R E 4 Sec3 is necessary for the establishment of neuronal polarity in mouse hippocampal neurons. (a) Representative images of double immunofluorescence micrographs showing the localization of Sec3 (red, third column) or Tau-1 (blue, second column), and GFP as a transfection marker (first column) in hippocampal pyramidal neurons in culture at stage 3 after 24 H.I.V transfected with Sec3-targeted shRNA (shSec3, first row) and a scrambled RNA sequence (ssCtrl, second row). Note that the neurons transfected with Sec3-targeted shRNA did not develop axons and did not target Tau-1 to any particular neurite (arrow) and the diminished levels of endogenous Sec3 protein (arrow head). In contrast, cells transfected with an ssCtrl exhibit an axon-like process enriched in Tau-1 (arrow). Calibration bar = 20 μm. (b) Representative images of double immunofluorescence micrographs showing the localization of Tau-1 (blue, second column), Sec3 (red, third column), and GFP as a transfection marker (first column) in hippocampal pyramidal neurons in culture at stage 3 after 24 H.I.V transfected with Sec3-targeted shRNA and Sec3FL (shSec3/Sec3FL, first row) or scrambled RNA sequence and Sec3FL (ssCtrl, second row). Note that in the two conditions the neurons transfected with shSec3 or ssCtrl and Sec3FL develop axons. Calibration bar = 20 μm. (c) Western blot of mouse Neuro-2a harvested at 24 H.I.V showing protein levels in culture in the presence of ssCtrl or shSec3 (first row; apparent molecular weight 102 kDa) and tubulin (second row) as a loading control and secondary antibodies suitable for near infrared fluorescence. The obtained images were converted to gray scale and inverted. The blots belong to the same experiment but were cropped for a better understanding. (d) Quantification of neuronal stages of cells transfected with Sec3-targeted shRNA, Sec3-targeted shRNA or ssCtrl with a full length Sec3 protein or control sequence after 24 H.I.V. Bars represent mean ± SEM of the percentage. A two-way ANOVA Test with Bonferroni post-test was performed. ***p ≤ 0.001, ns: non-significance. n = 3 independent culture preparations from different animals. At least 100 neurons were scored for each condition. Merged panels were pseudocolored shown in Figure 8b Figure 8b). This demonstrates that the multipolar/bipolar switch is diminished when Sec3 expression is suppressed.

| DISCUSS ION
For the last years a lot of effort has been made toward the study of the exocyst structure. It has been recently published that the first F I G U R E 5 Sec3 is necessary for the polarized insertion of the IGF1 receptor. (a) Representative images of triple immunofluorescence micrographs showing the distributions of Sec3 (red, first and fourth columns from left, second row) or phosphorylated IGF-1r (red, second and fifth column from left, second row) or tyrosinated tubulin (third and sixth columns from left, second row) and the transfection marker GFP (first row, all columns) in hippocampal pyramidal neurons in culture at stage 2 after 20 H.I.V transfected with Sec3-targeted shRNA (shSec3, left panel) or a scrambled RNA sequence (ssCtrl, right panel). Neurons were deprived of growth factors for 4 h and stimulated with 20 nM IGF-1 for 2 min before fixation. Note the polarization of active (membrane-inserted) IGF-1r to one of the minor neurites of the cell transfected with ssRNA (arrow). In contrast, neurons transfected with shSec3 showing virtually no Sec3 detected failed to polarize the active IGF-1r to any neurite. Calibration bar = 20 μm. (b) Quantification of a polarization index of active IGF-1r (P.I. IGF-1r.) calculated as the fluorescence intensity (A.U.) of the brightest minor neurite/average fluorescence intensity (A.U.) of the other minor neurites of the same cell. Bars represent mean ± SEM of the P.I. Neurons were processed as in c. The polarization index is significantly higher in neurons transfected with scrambled sequence compared to those transfected with shSec3. A t-test was performed ***p ≤ 0.001. ns: non-significant n = 3 independent culture preparations prom different animals. At least 10 cells were scored for each condition. Merged panels were pseudocolored step in the exocyst complex formation involves the formation of two subcomplexes: (i) The I subcomplex containing Sec3,Sec5,Sec6,and Sec8,and (ii) The II subcomplex holding Sec10, Sec15, Exo 70, and Exo84. After contacting the plasmalemma these two subcomplexes fit together forming the whole exocyst complex to anchor the vesicle to the plasmalemma. Before fusion occurs, the Sec3 subunit is liberated from the exocyst complex. After fusion has started, the remaining exocyst subunits get freed from the fusion point. A striking F I G U R E 6 Electroporation with shSec3 reduces expression Sec3. (a) Representative images of brains co-electroporated at E14,5 and analyzed at E18 with Sec3-targeted shRNA (top) or control ssCtrl (bottom). The cortical plate (CP) is labeled (arrow head) for orientation. A higher magnification of the framed area is displayed showing the expression of DsRed-positive cells (second row) and Sec3 (green, third row). Note that the neurons transfected with Sec3-targeted shRNA showed diminished levels of endogenous Sec3 protein (arrow). In contrast, cells transfected with an ssCtrl exhibit normal levels of protein (arrow). Merged panels were pseudocolored F I G U R E 7 Sec3 is necessary for the regulation of cortical neuron migration. (a) Representative images of brains co-electroporated at E14.5 and analyzed at E18 with control ssCtrl (first row), Sec3-targeted shRNA (second row), and Sec3-targeted shRNA/Sec3FL or ssCtrl/ Sec3FL and in all cases with DsRed. Few DsRed-positive cells were located in the cortical plate (CP) and the marginal zone (MZ) when Sec3targeted shRNA expression was knocked down compared to control. Calibration bar = 10 μm. (b) Quantification of the distribution of DsRedpositive cells in Cortical Plate (CP), intermediate zone (IZ), and ventricular/subventricular zones (VZ/SVZ) as indicated in (a) transfected with Sec3-targeted shRNA, Sec3-targeted shRNA/Sec3FL or ssCtrl /Sec3FL or control sequence at E18. Bars represent mean ± SEM of the percentage. A Two-way ANOVA Test with Bonferroni post-test was performed. ***p ≤ 0.001, **p ≤ 0.01, ns: non-significant, n = 3 number of animals. At least 30 slices of brain were scored for each condition. All panels were pseudocolored conclusion raised from these observations, although Sec3 is necessary to form the exocyst complex, this subunit is softly bond to the complex and gets liberated before fusion (Ahmed et al., 2018).
Also some interesting interactions of Sec3 have been mapped at the level of specific residues, indicating that the N-terminal part of this protein attaches to the SNARE Sso1 increasing the rate for SNARE complex formation (Yue et al., 2017). It has been also shown a direct interaction between Sec3 and the small GTPase Rho1 in yeast (Munson & Novick, 2006). Moreover, it has been determined that Rho GTPases are fundamental not only for cytoskeleton remodeling but also for cell polarity. In numerous tissues and organs, including the CNS, this protein family plays a key role in cell migration, forecasting as good candidates for involvement in neuronal migration disorders (Cappello et al., 2012).
Many findings point to the EC as important for polarized delivery of membrane in several cell types (Novick et al., 1980). For example, the yeast EC marks regions of new membrane addition during budding and cytokinesis (Finger et al., 1998;Guo et al., 1999;TerBush & Novick, 1995). In multi-cellular organisms the EC has been also involved in different processes including exocytosis, such as the establishment of polarity in epithelial cells (Grindstaff et al., 1998), the insertion of Glut4, the glucose transporter, into the plasmalemma of adipocytes (Inoue et al., 2003) and post-synaptic NMDA and AMPA receptor trafficking in dendrites (Sans et al., 2003). Developing neurons necessitates to add a significant amount of newly synthetized membrane to their plasmalemma to support axonal formation and elongation. This is achieved by exocytosis of PPVs specially at the axonal growth cone. We have previously published that PPV´s exocytosis is triggered by IGF-1 at the third distal axon and the growth cone (Pfenninger et al., 2003). The activation of the small GTPase TC10 and the exocyst component Exo70 is also involved in this pathway required for the control of plasmalemmal enlargement (Dupraz et al., 2009). Moreover, in PC12 cells the Exo70-TC10 complex modulates neurite outgrowth induced by the neurotrophic factor NGF (Pommereit & Wouters, 2007). Although it has become increasingly evident the possible role of the EC in polarization, beside these data, no functional studies about the role of individual components of the exocyst complex on neuronal differentiation and establishment of neuronal polarity have been yet performed.
The goal of the experiments shown in the present report was to determine the possible participation of the exocyst protein Sec3 on the regulation of axonal initial outgrowth necessary for the establishment of neuronal polarity. In a first set of experiments, we showed the early temporal expression of Sec3 in developing neurons at stages 1 and 2 before polarization and its localization preferentially at distal axons and growth cones. Up-regulation experiments using Sec3-targeted shRNA revealed that this exocyst protein is necessary for the regulation of initial axonal outgrowth and hippocampal neuron polarization. We extend this initial observation to an in situ model to study the role of Sec3 on migration and F I G U R E 8 Sec3 is necessary for the polarity switch in cortical neurons. (a) Representative images of brains co-electroporated with control ssCtrl (top), Sec3-targeted shRNA (middle) and Sec3-targeted shRNA/Sec3FL (bottom), in all cases with DsRed at E14.5 and analyzed at E18. DsRed-positive cells with Sec3-targeted shRNA located in the intermediate zone showed predominantly a multipolar morphology in contrast to the ssCtrl, Sec3-targeted shRNA/Sec3FL or ssCtrl/Sec3FL cells that exhibit bipolar morphology. Calibration bar =50 μm. (b) Quantification of the morphology of DsRed-positive cells as multipolar or Unipolar/bipolar in the intermediate zone (IZ). Bars represent mean ± SEM of the percentage. A Two-way ANOVA Test with Bonferroni post-test was performed ***p ≤ 0.001 n = 3, number of animals. At least 30 images were scored for each condition. Micrograh images were pseudocolored differentiation of pyramidal cortical neurons by in utero electroporation. During brain cortex formation, the migrating young neurons transiently exhibit a multipolar morphology at the VZ/SVZ region, at the time the neurons penetrate the IZ the cell morphology changes and the neurons become bipolar, with a leading process (the future axial dendrite) and a training process (the future axon) (Bai et al., 2003;Namba et al., 2014;Noctor et al., 2004;Tabata & Nakajima, 2003). This change in morphology is an essential step in neuron migration indispensable for the accurate lamination of the neocortex (Evsyukova et al., 2013;Miyoshi & Fishell, 2012;Noctor et al., 2004;Ohshima et al., 2007;Pacary et al., 2011). Our results show that loss of function of Sec3 (see above) maintain most of the neurons as multipolar, inhibiting the polarity change at the IZ. It follows that most neurons with knocked-down expression of Sec3 remain arrested at the SVZ/VZ/IZ and are unable to form an axon, abolishing neuronal polarity. Co-transfection whit a human form of Sec3 rescued the morphology allowing near to normal migration and differentiation.
Although our study focused on Sec3, the exocyst consists of one copy of each of its eight subunits (845 kDa in total): Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70, and Exo84 (Heider & Munson, 2012;Hsu et al., 1999;TerBush & Novick, 1995). In spite of the ideas about the dynamics of the formation of the exocyst complex have change along the years, Sec3 protein is consistently located in contact with the plasma membrane as well as the exocyst partner Exo70 (Boyd et al., 2004;Dong et al., 2005;Finger et al., 1998;Picco et al., 2017;Yamashita et al., 2010). With the following results, we have uncovered that Sec3 is critical for the establishment of neuronal polarity both in vitro and in situ experiments. Intriguingly Sec3 and Exo70 (which is also needed for neuron polarity as has been shown for our laboratory (Dupraz et al., 2009)), are associated to the plasma membrane. The question of how the different exocyst proteins are involved in neuronal polarization and subsequent in cortical migration is intricate. One straightforward possibility is that the localization of the different components within the exocyst may define the role in the polarity, but certainly the study of this correlation will be one of the major challenges for the future.

ACK N OWLED G M ENTS
We acknowledge the technical and imaging assistance from Dr Cecilia Sampedro and Dr Carlos Más, Centro de Micro y Nanoscopía All experiments were conducted in compliance with the ARRIVE guidelines.

CO N FLI C T S O F I NTE R E S T
The authors have no conflict of interest to declare.

AUTH O R S CO NTR I B UTI O N S
All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by FBP.
LJS discussed the results and participated in the preparation of the paper. SQ and FBP planned experiments, discussed the results and participated in the preparation of the paper, SQ. All authors commented on the manuscript. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on request from the authors. A preprint of this article was posted on ResearchSquare: https://www.resea rchsq uare.