ARMS/Kidins220 and synembryn-B levels regulate NGF-mediated secretion

Development of the nervous system requires a temporally and spatially orchestrated set of events. Among the different growth factors that participate in this process, neurotrophins represent important players. Neurotrophins constitute a family of growth factors that regulate neuronal survival and differentiation, synapse formation and plasticity in the nervous system. In addition, they have been implicated in different neurodegenerative diseases such as Alzheimer, Huntington and Parkinson disease and in neuropsychiatric disorders such as depression and schizophrenia (Autry and Monteggia, 2012; Chao, 2003). The neurotrophin family includes nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3, also known as NTF3) and neurotrophin-4 (NT-4, also known as NTF4), which exert their functions through binding to two different receptors, the Trk neurotrophin receptors (TrkA, TrkB and TrkC; also known as NTRK1, NTRK2 and NTRK3, respectively) and the p75 pan-neurotrophin receptor (also known as NGFR). NGF, the prototype of neurotrophin, acts through TrkA and p75 (Arévalo and Wu, 2006; Reichardt, 2006). It has been described that NGF enhances depolarization- and ionomycin-induced neurotransmitter release in PC12 cells, an effect that depends on TrkA, MAPK proteins (i.e. ERK1/2, also known as MAPK3 and MAPK1) and phosphoinositide 3-kinase (PI3K) activation (Amino et al., 2002). However, it is unknown whether other proteins are involved in NGF-mediated secretion.

ARMS (also known as Kidins220) is a scaffold protein that was originally identified as a protein kinase D (PKD) substrate (Iglesias et al., 2000) and as an interacting protein of p75 and Trk neurotrophin receptors (Kong et al., 2001). Among other functions, ARMS has been implicated in the sustained activation of ERK1/2 in response to NGF (Arévalo et al., 2006, 2004; Hisata et al., 2007; Neubrand et al., 2010), in synaptic modulation (Arévalo et al., 2010; Cesca et al., 2015; Cortés et al., 2007; Lopez-Menendez et al., 2009; Sutachan et al., 2010; Wu et al., 2009), in dendritic arborization (Cesca et al., 2012; Chen et al., 2012; Wu et al., 2009), in excitotoxicity (Lopez-Menendez et al., 2009), in tumor progression in melanoma cells (Liao et al., 2011), in T cell activation (Deswal et al., 2013) and in the regulation of B cell development (Fiala et al., 2015). In addition, ARMS has been implicated in the secretion of the neurotensin hormone in BON cells, a cell line derived from a human pancreatic carcinoid tumor, in response to phorbol esters located downstream of PKD (Li et al., 2008, 2012). However, it is unknown whether ARMS plays any role in secretion in response to neurotrophins in neuronal cells.

Resistant inhibitor of cholinesterase-8 (Ric-8) was originally described in Caenorhabditis elegans mutants resistant to Aldicarb (Miller et al., 1996) and later on implicated in the release of neurotransmitters in the worm (Miller et al., 2000). Ric-8 has also been implicated in Drosophila melanogaster where it is known to regulate synapse number (Romero-Pozuelo et al., 2014). Ric-8 works as a non-canonical guanidine exchange factor (GEF) for monomeric Gα proteins (Klattenhoff et al., 2003; Tall et al., 2003). In mammals, the orthologs of Ric-8, named synembryn-A and synembryn-B, exhibit different specificities for Gα proteins (Chan et al., 2011; Romo et al., 2008). In addition to modulating phospholipase C (PLC)β through Gαq, the C. elegans Ric-8 activates a parallel pathway through activation of Trio, a GEF for Rho GTPases (Williams et al., 2007). Although the involvement of Ric-8 in neurotransmitter release is well established in C. elegans, it is unknown whether synembryn proteins participate in the secretion of proteins.

Here, we describe an interaction between ARMS and synembryn-B involving the polyproline-rich region of ARMS and the N-terminus of synembryn-B. In addition, we have identified a new signaling pathway implicated in the regulated secretion mediated by NGF that involves the TrkA receptor, ARMS, synembryn-B and Rac1. Different levels of ARMS and synembryn-B proteins regulate NGF-mediated secretion, through modulation of Rac1 activity. Gαq and Trio proteins link ARMS–synembryn-B to Rac1. NGF-induced Rac1 activity is sensitive to either decreases in Gαq or Trio activity. Therefore, NGF-mediated secretion requires two scaffold proteins, ARMS and synembryn-B, which can control Rac1 activity.

To identify proteins that can interact with ARMS, we performed a yeast two-hybrid screening using the polyproline-rich region of ARMS as bait and a cDNA library made from rat dorsal root ganglion (DRG) neurons (Kong et al., 2001). The only interacting protein identified was synembryn-B, a vertebrate homolog of C. elegans Ric-8. The putative interaction of ARMS and synembryn-B observed in the yeast two-hybrid screening was confirmed by co-immunoprecipitation experiments carried out in transfected HEK293 cells (Fig. 1A). Moreover, the interaction was observed in DRG neurons expressing endogenous levels of ARMS and synembryn-B and, interestingly, increased in response to NGF stimulation (Fig. 1B). Therefore, ARMS displays an interaction with synembryn-B. To assess the regions of synembryn-B and ARMS involved in the interaction, mapping experiments were performed. The N-terminal region of synembryn-B was required for the association with ARMS, as assessed by co-immunoprecipitation experiments (Fig. 1C), whereas the last 20 amino acids of ARMS (bait 7.6) were responsible for the interaction with synembryn-B, as assessed by the yeast two-hybrid experiment (Fig. 1D,E). In addition, pulldown assays using a bacterially purified GST–7.6-ARMS fragment and cell lysates expressing Flag–synembryn-B indicated a direct interaction between both proteins (Fig. 1F). Taken together, these results indicate that ARMS and synembryn-B directly interact using specific regions of each protein.

ARMS has been directly linked to NGF-mediated signaling (Arévalo et al., 2006, 2004; Neubrand et al., 2010) and NGF elicits regulated secretion (Canossa et al., 1997; Kruttgen et al., 1998). To investigate the role of ARMS, we utilized PC12 cells because they respond to NGF (Greene and Tischler, 1976) and are a model widely used to study secretion (Westerink and Ewing, 2008). A transfected human growth hormone (hGH) was used as a reporter of secretion (Sugita, 2004; Wick et al., 1993). We observed that acute NGF stimulation for 30 min evoked a significant increase in the secretion of transfected hGH in non-differentiated PC12 cells (Fig. 2A). Given that it has been described that hGH is packed in large dense-core vesicles (LDCVs) (Lowe et al., 1988; Schweitzer and Kelly, 1985), we decided to validate these results by performing electron microscopy studies (Fig. S1A). We found a significant reduction in the amount of LDCVs observed in cells in response to NGF stimulation (Fig. 2B). Therefore, NGF induces hGH secretion in non-differentiated PC12 cells. Subsequently, we sought to identify the signaling pathways involved in NGF-induced hGH release, and performed secretion experiments in non-differentiated PC12 cells in the presence of different inhibitors to block TrkA, ERK1/2 or PI3K activation. Blocking the activation of TrkA or ERK1/2 halted NGF-mediated secretion, whereas inhibition of PI3K activation had no observable effect (Fig. 2C). The presence of these compounds did not alter basal hGH secretion (Fig. S1B) and the specificity of the inhibitors was tested by western blot analysis (Fig. 2D). Therefore, our results indicate that NGF induces hGH secretion that directly depends on the TrkA and ERK1/2 pathways.

In order to study how NGF mediates regulated secretion, we manipulated the level of ARMS to address the effect upon NGF-mediated secretion. Stable PC12 clones overexpressing GFP or GFP–ARMS were generated (Fig. 3A, left panel), and three independent clones expressing each protein were used to perform hGH secretion assays. Interestingly, overexpression of GFP–ARMS blocked NGF-mediated hGH release (Fig. 3A, right panel), without affecting basal secretion (Fig. S2C). These results were further supported by the presence of a similar number of LDCVs in PC12 clones expressing GFP–ARMS in both basal and stimulated conditions (Fig. 3B). Considering that ARMS protein levels can affect the insertion of AMPA receptors into the plasma membrane (Arévalo et al., 2010), we wondered whether membrane targeting of TrkA could be altered upon overexpression of ARMS. After performing biotinylation experiments, similar TrkA membrane targeting was observed in GFP and GFP–ARMS-expressing cells (Fig. S2A). Biotinylated ARMS was used as a control of membrane targeting and to show the overexpression of the protein, and tubulin was used as a negative control (Fig. S2A). Moreover, the activation of TrkA and downstream signaling pathways in response to NGF were checked. No differences were observed in the levels of phosphorylated (p)TrkA, pAkt or pERK1/2 in cells expressing GFP or GFP–ARMS (Fig. S2A). Given that the activation of TrkA was weak in whole-cell lysates we performed additional experiments immunoprecipitating TrkA before checking activation. The levels of pTrkA were comparable in GFP and GFP–ARMS cells (Fig. S2B). Therefore, ARMS overexpression does not alter TrkA targeting to the plasma membrane in PC12 cells.

To address whether depletion of ARMS had an effect on NGF-mediated secretion, we took advantage of a previously validated short hairpin RNA (shRNA) against ARMS (Arévalo et al., 2010; Cortés et al., 2007). Downregulation of ARMS protein levels resulted in a significant increase in NGF-mediated hGH release when compared to control cells (Fig. 3C), with no effect on basal secretion (Fig. S2B). Taken together, these results suggest that the levels of ARMS control NGF-mediated hGH secretion.

Given that ARMS has been directly implicated in ERK1/2 activation in response to NGF (Arévalo et al., 2006, 2004), we addressed the discrepancies observed in hGH secretion in response to NGF between the levels of ARMS and ERK1/2 activation. Similar experiments to those described in Fig. 3C, were carried out in the presence of the ERK1/2 inhibitor U0126. Inhibition of ERK1/2 activation did not block the increase of NGF-mediated hGH release mediated by ARMS protein depletion, whereas it halted regulated secretion in control cells (Fig. 3D). Basal secretion was again not altered by the presence of U0126 (Fig. S2C). Taken together, these results suggest that ARMS regulates NGF-mediated secretion through a mechanism that seems to be independent of ERK1/2 activation.

ARMS protein levels are tightly regulated during nervous system development (Cortés et al., 2007; Wu et al., 2009) and considering the above results, we decided to assess ARMS levels in PC12 cells that had been treated with NGF for different times. A trend associated with the increase in ARMS expression, during the course of up to 2 days of NGF treatment (Fig. 3E), was observed when cells were in the process of differentiation. By contrast, there was a statistically significant reduction in the levels of ARMS protein 5 days post-treatment (Fig. 3E), when the cells were already differentiated (Fig. S2D). Subsequently, we addressed whether regulated secretion in response to NGF depended on cell differentiation. PC12 cells expressing hGH were treated with NGF for different periods and depleted of NGF for 4 h. These cells were then exposed to acute levels of NGF for 30 min before performing secretion assays. Cells previously treated with NGF for 1 day did not show hGH secretion above basal levels, whereas those cells treated for 5 days with NGF did exhibit a very strong secretory response to an NGF stimulation of 30 min (Fig. 3F). Basal secretion was downregulated in response to differentiation (Fig. S2E). These results suggest that ARMS levels are tightly regulated depending on the differentiation status of PC12 cells and, together with the overexpression and depletion data, indicate the relevance of ARMS protein levels on NGF-mediated secretion.

ARMS has been implicated in stimuli-dependent secretion of the neurotensin hormone (Li et al., 2008) and can regulate NGF-mediated secretion (Fig. 3). Because of the direct interaction of ARMS and synembryn-B (Fig. 1), we asked whether synembryn-B had a similar impact upon NGF-mediated secretion. The C. elegans homolog of synembryn, Ric-8, is capable of modulating neurotransmitter release (Miller et al., 2000). We first overexpressed synembryn-B along with hGH in PC12 cells and assayed for secretion (Fig. 4A, left panel). We observed that NGF-mediated hGH release was blocked in cells overexpressing synembryn-B (Fig. 4B), without affecting basal hGH secretion (Fig. S3A). These data were corroborated after counting similar numbers of LDCVs in cells overexpressing synembryn-B in both the presence and absence of NGF (Fig. 4B). To further address the role of synembryn-B in secretion, we infected PC12 cells with lentivirus expressing a specific shRNA against synembryn-B that had been previously validated (Nagai et al., 2010) (Fig. 4C, left panels). Reduction of synembryn-B levels enhanced the NGF-mediated hGH release as compared to control cells (Fig. 4C), but no effect was observed on basal secretion (Fig. S3B). Taken together, these results suggest that synembryn-B levels regulate NGF-mediated secretion in a similar way to ARMS.

To address whether both ARMS and synembryn-B levels might be required for the secretion in response to NGF treatment, we performed additional experiments. Depletion of synembryn-B remedied the NGF-mediated secretion defect caused by ARMS overexpression, resulting in the secretion levels of control cells (Fig. 4D). Similarly, depletion of ARMS rescued the NGF-induced secretion defect observed in cells overexpressing synembryn-B (Fig. 4E). In none of the previous experiments was the hGH basal secretion altered (Fig. S3C,D). Therefore, ARMS and synembryn-B protein levels are both required to control regulated secretion in response to NGF.

In order to determine how ARMS and synembryn-B could modulate secretion, we took into account that NGF treatment leads to Rac1 activation (Nusser et al., 2002; Yamaguchi et al., 2001) and that ARMS has been shown to modulate Rac1 activation through the GEF Trio (Neubrand et al., 2010). Because Ric-8 regulates Trio through Gαq (Williams et al., 2007), we assessed whether ARMS and synembryn-B regulated Rac1 activity in response to NGF. In these experiments, we stimulated with NGF for 2.5 min because it has been reported that Rac1 is activated after 1–3 min of treatment (Yamaguchi et al., 2001). In PC12 cells expressing GFP or control shRNA, we observed similar results to those previously reported (Fig. 5A). Cells overexpressing GFP–ARMS or GFP–synembryn-B showed a significant increase in the amount of GTP-bound Rac1 in NGF-stimulated conditions when compared to control cells (Fig. 5A, left panels), whereas reduction of ARMS or synembryn-B levels resulted in significant impairment of Rac1 activation (Fig. 5A, right panels). To extend these results, we tested whether expression of the dominant-negative mutant Rac1 (Rac1-N17) had any consequence on NGF-mediated hGH secretion deficiency in cells overexpressing ARMS or synembryn-B. Expression of this GTPase mutant rescued the NGF-mediated secretion defects observed in cells overexpressing ARMS or synembryn-B to the secretion levels present in GFP control cells (Fig. 5B). Expression of Rac1-N17 resulted in a reduced basal secretion in all samples (Fig. S3E). These results suggest that ARMS and synembryn-B regulate NGF-mediated secretion by modulating Rac1 GTPase activity.

To address further how Rac1 activity is regulated by ARMS and synembryn-B, we used the inhibitor ITX-3, which specifically targets the Trio–RhoG–Rac1 pathway (Bouquier et al., 2009). Application of ITX-3 resulted in a significant reduction of Rac1 activation in cells in which ARMS or synembryn-B were overexpressed (Fig. 5C). To ascertain whether Gαq is the link between synembryn-B and Trio, we took advantage of the Gαq inhibitor YM254890 (Takasaki et al., 2004), which significantly eliminated the enhanced Rac1 activation in response to NGF in cells overexpressing synembryn-B, but not completely in cells overexpressing ARMS (Fig. 5D). Taken together, these results suggest that ARMS and synembryn-B regulate Rac1 activity and that Trio and Gαq link ARMS and synembryn-B with the aforementioned GTPase.

Given that ARMS and synembryn-B protein levels are instrumental for NGF-mediated secretion, we were curious to know whether high levels of ARMS or synembryn-B could be maintained in differentiated PC12 cells. Thus, we used the clones overexpressing GFP–ARMS and GFP–synembryn-B described above (Figs 3A and 4A). In response to NGF treatment, we observed an accumulation of GFP–ARMS and GFP–synembryn-B proteins that was maintained for up to 1–2 days (Fig. 6A, top). However, a strong downregulation of GFP–ARMS and GFP–synembryn-B expression was observed after 2 days of NGF treatment (Fig. 6A, top). This pattern, although more intense, resembled the previous one observed for ARMS in wild-type PC12 cells treated with NGF (Fig. 3F). These results suggest that both ARMS and synembryn-B levels, independently of the initial amounts of each protein, are tightly regulated in response to NGF treatment. To address whether this was a specific effect of NGF, we used FGFb (also known as FGF2) in place of NGF given that it is a known factor that also differentiates PC12 cells (Rydel and Greene, 1987). A similar kinetics of ARMS and synembryn-B expression was obtained upon FGFb treatment in wild-type PC12 cells (Fig. S4A). Taken together, these results indicate that ARMS and synembryn-B levels are tightly regulated in response to different stimuli capable of differentiating PC12 cells.

To assess the secretion capabilities of PC12 cells overexpressing ARMS or synembryn-B at different time points of differentiation, we performed secretion experiments at 1 and 5 days upon NGF treatment, similar to those described in Fig. 3F. None of the cells overexpressing GFP–ARMS or GFP–synembryn-B, previously treated with NGF for 1 day, responded to an acute NGF 30 min treatment in terms of secretion (Fig. 6A, bottom). However, the same overexpressing cells treated with NGF for 5 days did display a very strong hGH secretion response to the same acute treatment (Fig. 6A, bottom). These results correlated with the reduced levels of ARMS and synembryn-B observed at this time point (Fig. 6A top). Basal secretion was downregulated in response to differentiation (Fig. S4B). These data resembled the data obtained with wild-type PC12 cells and indicate that, once the cells differentiate, there is a physiological downregulation of ARMS and synembryn-B levels accompanied by a potentiation of NGF-mediated secretion.

There is strong support for the role of ARMS in the differentiation of PC12 cells in response to NGF (Arévalo et al., 2006, 2004; Chen et al., 2012; Neubrand et al., 2010; Wu et al., 2009). Considering that synembryn-B levels are very high during differentiation and follow the same pattern as ARMS protein levels, we addressed whether synembryn-B levels were involved in this process. Downregulation of synembryn-B prevented the differentiation of PC12 cells when treated for 5 days with NGF, exhibiting a result similar to that obtained for ARMS-depleted cells (Fig. 6B). However, overexpression of GFP–synembryn-B or GFP–ARMS did not increase the amount of differentiated cells when compared to cells only expressing GFP cells at the same time point (data not shown). In contrast, an acceleration in cell differentiation was observed for cells overexpressing GFP–ARMS after being treated with NGF for 40 h. This result was not observed for cells overexpressing synembryn-B (Fig. 6C). Taken together, these results indicate that synembryn-B is required for PC12 differentiation in response to NGF.

In the present work, we identified an interaction between ARMS and synembryn-B with important consequences. In addition, we provide evidence that ARMS and synembryn-B proteins control regulated secretion in response to NGF without affecting basal secretion. These findings are based on: (1) the blocking effect on NGF-mediated secretion with high levels of ARMS or synembryn-B, (2) a potentiation in regulated secretion in response to depletion of ARMS or synembryn-B levels, and (3) the requirement of both proteins for regulated secretion. Downstream of ARMS and synembryn-B in the regulation of NGF-mediated secretion is the small GTPase Rac1, and its activation in response to NGF requires both of the aforementioned proteins together with Gαq and Trio. Interestingly, there is a strong, physiological downregulation of ARMS and synembryn-B protein levels upon PC12 differentiation that coincides with a very potent regulated secretion in response to NGF.

ARMS is a well-recognized scaffold protein that is essential for different signaling pathways (Neubrand et al., 2012). In addition, ARMS seems to play a positive role in neurotensin secretion from BON cells in response to phorbol 12-myristated 13-acetate downstream of the PKD signaling pathway (Li et al., 2008). The different approaches used to obtain our results clearly demonstrate that the ARMS protein has a negative role in NGF-mediated secretion in PC12 cells. The discrepancies observed between previously published results and ours might be due to the fact that the effects of ARMS on secretion depend on the cellular model and/or the stimulus used to elicit secretion. It is worth noting that ARMS has also been postulated to be a cargo protein for the kinesin-1 motor (Bracale et al., 2007), and it is associated with acetylated α-tubulin and microtubule-regulating proteins, such as stathmin and MAP1, that actively control the microtubule network (Higuero et al., 2010). Microtubules have been directly implicated in the reorganization of the cytoskeleton that takes place during neuronal polarization and differentiation (Kapitein and Hoogenraad, 2015). In addition, other proteins related to microtubule dynamics, such as CLASP2, have been recently involved in neurotransmitter release (Beffert et al., 2012). It might be possible that different microtubule-regulating proteins that interact with ARMS could affect regulated secretion.

Our laboratory, in addition to others, has previously shown that ARMS positively regulates ERK1/2 activation (Arévalo et al., 2004; Hisata et al., 2007). ARMS exclusively activates the sustained ERK1/2 activation that occurs after 40–60 min of NGF treatment (Arévalo et al., 2004). In this work, we have observed that inhibition of ERK1/2 activation did not prevent the effect of ARMS depletion on NGF-mediated secretion. Although secretion experiments have been performed after 30 min of NGF stimulation, we know that this effect occurs very rapidly because 2.5 min of NGF treatment is enough to modulate ARMS and synembryn-B levels and Rac1 activation. These data suggest to us that a different signaling pathway mediates the effect of ARMS on NGF-regulated secretion. The identification of synembryn-B in the yeast two-hybrid screening performed opened the possibility that this protein might be part of the signaling complex involved in the action of ARMS on regulated secretion. The relationship of ARMS and synembryn-B has been demonstrated because both interact, their levels have a similar effect on secretion and both proteins are regulated similarly in response to physiological differentiation. Ric-8 proteins function exclusively on dissociated monomeric Gα subunits (Tall et al., 2003) and controls neurotransmitter release at the neuromuscular junction in C. elegans (Miller et al., 2000). The data obtained in the PC12 cell model used to address the role of ARMS and synembryn-B in regulated secretion require additional validation using a more relevant, physiological model such as primary cultured neurons or even in vivo. Future analysis is required in order to be able to extrapolate these interesting results to neurons.

Downstream of ARMS and synembryn-B, we have found that Rac1 is a key player because ARMS and synembryn-B levels are instrumental for Rac1 activation (Fig. 5). It has been independently demonstrated that Rac1 is downstream of ARMS and synembryn-B (Neubrand et al., 2010; Williams et al., 2007). Moreover, our results suggest that inhibition of Gαq and Trio totally blocks the enhanced Rac1 activation obtained by synembryn-B overexpression and only partially blocks the one in response to ARMS overexpression (Fig. 5). Rac1 activation has a positive role in NGF-mediated differentiation (Neubrand et al., 2010) and a negative role in regulated secretion (Frantz et al., 2002). It has been shown that expression of the dominant-negative mutant Rac1-N17, with a similar localization to that of the wild-type Rac1, enhanced stimulated secretion in response to K⁺ depolarization, whereas the constitutively active mutant Rac1-V12 resulted in a decrease in K⁺-induced secretion (Frantz et al., 2002). Our results also support a negative role of Rac1 activation in PC12 cells in NGF-mediated secretion and a positive role in NGF-mediated differentiation.

The analysis of the levels of ARMS and synembryn-B in response to the physiological differentiation of PC12 cells upon NGF or FGFb treatment together with the regulated secretion results indicates a very strong inverse correlation between the expression of both proteins and the regulated secretion in response to NGF (Fig. 6). Taken together, the results obtained allow us to propose a model in which high levels of ARMS and synembryn-B early during development promote neuronal differentiation and are reduced once the cells have differentiated. Once cells undergo differentiation, high levels of ARMS and synembryn-B keep regulated secretion in check until proper connections are formed. It is at this time that the levels of ARMS and synembryn-B become downregulated. High levels of ARMS during neuronal differentiation and brain maturation have been previously reported (Cortés et al., 2007; Wu et al., 2009) and ARMS has been widely demonstrated to regulate neuronal differentiation (Arévalo et al., 2006, 2004; Bracale et al., 2007; Cesca et al., 2012; Chen et al., 2012; Higuero et al., 2010; Hisata et al., 2007; Neubrand et al., 2010; Wu et al., 2009). During the preparation of this manuscript an article has been published identifying different regions of ARMS cleaved by calpain (Gamir-Morralla et al., 2015). Those authors suggest that ARMS could be used as a therapeutic target for excitotoxicity by blocking its degradation. Considering the importance that the levels of ARMS have in the regulated secretion, further studies need to be performed before attempting the modulation of its levels in the treatment of different pathologies such as brain ischemia.

In summary, we have found that ARMS and synembryn-B levels regulate NGF-mediated secretion and are tightly regulated during development promoting differentiation and keeping regulated secretion under control.

Mouse Nerve Growth Factor 2.5S form (NGF) was from Alomone Labs, K252a (100 nM) from Sigma, LY294002 (20 µM) from Cayman Chemicals, U0126 (10 µM) from Calbiochem, ITX-3 (100 µM) from Calbiochem, YM-254890 (1 µM) a generous gift from Taiho Pharmaceutical and isopropyl-β-D-thiogalactopyranoside (IPTG) from Prolabo. The following antibodies were used at the following concentrations: rabbit anti-ARMS C-terminus polyclonal (1 µg/ml) and rabbit anti-synembryn C-terminus polyclonal (1.5 µg/ml), generated in our laboratory (see below); mouse anti-GAPDH monoclonal (1:10,000, Sigma, catalog no G8795), mouse β-actin monoclonal (1:5000, Sigma, catalog no A5441), rabbit β-tubulin III polyclonal (1:10,000, Sigma, catalog no T2200) and rabbit Flag polyclonal (1:500, Sigma, catalog no F7425), mouse phospho-ERK1/2 monoclonal (1:2000, Cell Signaling, catalog no 9106), mouse phospho-Ser473-Akt1 (PKBα; 1:1000, Upstate, catalog no 05-669), mouse monoclonal Rac1 (1:100, BD Transduction Laboratories, catalog no 610651) and mouse GFP monoclonal (1:2000, Clontech, catalog no 632380).

The pFLAG-ARMS/Kidins220 construct was used as previously described (Arévalo et al., 2004), full-length synembryn-B, SynΔC-term (coding amino acids 1–261) and SynΔN-term (coding amino acids 262–520) were cloned into pFlag-CMV-2 (Sigma). Control shRNA (5′-GCGCGCTTTGTAGGATTCG-3′), ARMS shRNA (5′-GCCACCAAGATGAGAAATA-3′) (Yu et al., 2011; Cortes et al., 2007) and synembryn-B shRNA (5′-CAGTTGGAAGGTGCATAAT-3′) (Nagai et al., 2010) were cloned into the pLVTHM vector. To establish cell lines, the H1 promoter followed by the corresponding shRNA were subcloned into the pERFPc3 vector. The complete coding sequences of synembryn-B and ARMS were cloned into the pEGFPC3 vector (Clontech). Glutathione S-transferase (GST)-fused, the Cdc42/Rac-interactive-binding domain of rat αPAK (GST–CRIB) and the dominant negative form of human Rac1 were provided by Xosé Bustelo [Centro de Investigación del Cáncer, Instituto de Biología Molecular y Celular del Cáncer, Consejo Superior de Investigaciones Científicas (CSIC)-University of Salamanca, Salamanca, Spain]. Vector pCMV-hGH-SV40 [a generous gift of Sabine Hilfiker, Institute of Parasitology and Biomedicine “López-Neyra”, Consejo Superior de Investigaciones Científicas (CSIC), Spain] was used to clone the complete coding sequences of synembryn-B and dominant-negative form of human Rac1. Bait 7, containing amino acids from P1057 to L1151 of proline-rich region of ARMS was amplified by PCR and cloned into the pEG202 vector in frame with the DNA-binding domain LexA. For mapping protein interactions, an additional six baits were generated in the same way as bait 7: bait 7.1 (P1057-G1131), bait 7.2 (P1057-G1111), bait 7.3 (P1057-G1189), bait 7.4 (P1090-L1151), bait 7.5 (P1112-L1151) and bait 7.6 (P1133-L1151). Bait 2.1 (S1304-L1715) was used as a negative control. In addition, bait 7.6 was subcloned into the pGEX-6P vector (Pharmacia). All constructs were verified by sequencing. Expression of the corresponding protein of each construct was verified by western blot analysis.

PC12 cells (provided by Sabine Hilfiker) were cultured in RPMI 1640 (Lonza) supplemented with 10% heat-inactivated horse serum, 5% fetal bovine serum (FBS), 2 mM L-glutamine and 100 U/ml penicillin-streptomycin (Gibco) on collagen-coated plates (BD Bioscience). HEK293 and 293FT cells were grown in DMEM (Gibco) supplemented with 10% bovine serum, 1% non-essential amino acids, 2 mM L-glutamine and 100 U/ml penicillin-streptomycin. E15.5 embryos from Wistar rats were used to obtain DRGs that were incubated, and dissociated with 0.25% trypsin in L-15 medium for 45 min at 37°C as previously described (Yu et al., 2011). All animal experiments were performed according to approved guidelines by the Bioethics Committee of the University of Salamanca and following European Community guidelines. Briefly, cells were plated using plating medium (MEM, 10% FBS, 0.4% glucose, 2 mM glutamine, 100 U/ml Pen/Strep) and NGF (50 ng/ml) overnight on plates coated with growth-factor-reduced Matrigel (BD Biosciences) as substrate. On the following day, the medium was changed to NB (Neurobasal-A medium, B-27, 0.4% glucose, 2 mM glutamine), NGF and 5-fluorodeoxyuridine (5FU) (2.44 µg/ml) and uridine (2.44 µg/ml). All cells were cultured at 37°C in 5% CO₂.

Polyclonal antibodies against ARMS and synembryn-B were generated in rabbits using recombinant proteins produced in bacteria consisting of GST–ARMS (last 180 amino acids of ARMS) and 6×His tagged to the first 261 amino acids of synembryn-B. The protocol used to immunize the rabbits and to purify the antibodies was carried out as previously described (Calvo et al., 2015). In brief, 3 weeks after an initial first injection of 500 μg of recombinant protein, boosts of 200 μg of recombinant protein were administered every 2 weeks. Bleedings were obtained 7–10 days after the second boost. Antibodies against GST were isolated by passing the serum through a GST–glutathione–agarose column twice. The remaining ARMS antibodies were purified using a GST–ARMS–glutathione–agarose column. Antibodies against synembryn-B were purified with an affinity column generated with the His–synembryn-B and Niquel beads. Antibody elution was performed with 0.1 M glycine, pH 2.5. The pH was neutralized with 1 M Tris-HCl, pH 9.0, and the solution was supplemented with 50 mM NaCl.

The yeast two-hybrid assay was performed with bait 7 (Fig. 1D) and a cDNA library from rat dorsal root ganglia neurons (DRG) as previously described (Kong et al., 2001). Approximately 3×10⁷ transformants were analyzed after being grown in absence of leucine and by measuring β-galactosidase activity. To map the ARMS sequence involved in the interaction with synembryn-B, baits 2.1 (corresponding to amino acids 1304–1715 of ARMS), 7.1, 7.2, 7.3, 7.4, 7.5 and 7.6 were used (Fig. 1D).

Cells were lysed in a lysis buffer (10 mM Tris-HCl pH 7.4, 150 mM NaCl, 2 mM EDTA, 1% NP- 40, 1 mM PMSF, 1 μg/ml Aprotinin, 2 μg/ml Leueptine, 1 mM Vanadate, 10 mM NaF and 20 mM β-glycerophosphate) for 40 min at 4°C with gentle shaking and centrifuged at 14,000 g for 15 min to eliminate the debris. SDS-buffer was added to lysates and boiled for 5 min to denature the proteins. Proteins were resolved by SDS-PAGE, and western blots were performed using antibodies against the different proteins. To avoid problems with the Ig chains we used Protein-A- or Protein-G-conjugated horseradish peroxidase (HRP) when same species antibodies were used for both immunoprecipitation and western blotting.

Surface labeling assays were performed using PC12 cells. Briefly, to detect the amount of surface proteins, GFP- and GFP–ARMS-expressing cells were used. Cells were stimulated with or without NGF (100 ng/ml) for 30 min. Subsequently, cells were washed sequentially using room temperature PBS and cold PBS, chilled on ice, and incubated in 0.5 µg/ml Sulfo-NHS-LC-biotin (Pierce) dissolved in biotinylation buffer (PBS, 1 mM CaCl₂, 0.5 mM MgCl₂) for 15 min at 4°C to label the membrane proteins. Free biotin was quenched with 0.1 M glycine for 15 min at 4°C. Cells were then washed twice with cold PBS and lysed as indicated above. Biotinylated proteins were isolated from the total cell lysate by immobilization on NeutroAvidin beads (Pierce) overnight at 4°C. The beads were washed three times with lysis buffer and 20 µl of 2×SDS sample buffer was added before boiling for 7 min. Proteins were subjected to SDS-PAGE and immunoblotted with the corresponding antibodies.

Expression of GST- and histidine-fused peptides was induced in DH5α E. coli with 0.2–0.5 mM IPTG for 5–6 h at 30°C. Bacteria were collected, resuspended in lysis buffer (PBS and 0.4 mg/ml lysozime, 1 mM PMSF, 1 mg/ml aprotinin and 2 mg/ml leupeptine), sonicated and a final concentration of 1% Triton X-100 was added. Supernatant was incubated with glutathione Sepharose 4B (Pharmacia^®) or nickel beads overnight at 4°C for 30 min at room temperature and washed three times with cold PBS.

Lysates of HEK293 cells transfected with a vector encoding Flag–synembryn-B were incubated for 4 h at 4°C with GST or GST–ARMS-7.6 pre-bound to agarose–glutathione conjugates. Beads were washed, boiled for 5 min in SDS-PAGE sample buffer and pulldown proteins were resolved by 10% SDS-PAGE and immunoblotted using anti-Flag antibodies.

The lentivirus used in this study was generated by co-transfection using calcium phosphate in 293FT cells as described previously (Wiznerowicz and Trono, 2003). 2.5×10⁶ 293FT cells, seeded on a 10-cm plate the day before, were transfected with 20 μg of pLVTHM containing the specific shRNA sequence together with 15 μg of psPAX2 and 6 μg of pMD.2G plasmids. The medium was replaced by fresh medium without antibiotics after 8 h. After 48 h, the supernatant containing the lentivirus was collected, centrifuged at 4700 g for 10 min, passed through a 0.45 μM filter, and stored in aliquots at −80°C. PC12 cells were infected using 50 μl of supernatant containing lentivirus per 1×10⁵ cells. The expression or the reduction in the expression of the corresponding protein was assessed by western blot analysis.

PC12 cells with a density of 80–90% were transfected with plasmids encoding GFP or RFP and the corresponding cDNA or shRNA using LipofectAMINE 2000 (Invitrogen^®) following the manufacturer's protocol. Medium containing the transfection mixture was replaced after 4 h by fresh medium and, after 24 h in culture, the cells were diluted 1:10 using complete medium supplemented with 0.5 mg/ml G418 (Gibco). Resistant clones to G418 were isolated after 3–5 weeks, propagated and maintained in the presence of 0.25 mg/ml of G418. Positive clones were identified under fluorescence microscope and the corresponding expression was tested by western blot analysis.

Wild-type or different clones of PC12 cells (2.5×10⁶ cells/well) were seeded on six-well plates 24 h prior to transfection. Cells were transfected with a construct expressing the human growth hormone (hGH) using LipofectAMINE 2000 and, on the following day, the cells were divided into two wells to measure both basal constitutive release and regulated evoked release of hGH from the same transfected cells. Regulated secretion of hGH was measured 72 h after transfection and cells were incubated for 30 min with a physiological saline solution (PSS; 15 mM Hepes-NaOH, pH 7.4, 145 mM NaCl, 5.6 mM KCl, 2.2 mM ClCa₂, 0.5 mM ClMg₂, 5.6 mM glucose) as the control or PSS and NGF (100 ng/ml) as the experiment. Plates were placed on ice for 15 min to effectively stop hGH release in all of the samples, the supernatants were then collected and the cells were lysed by conducting five cycles of freeze-thaw. The concentration of hGH was measured in both extracellular medium and cell lysates using an ELISA assay kit (Roche^®) according to the manufacturer's instructions. The amount of hGH secreted was calculated as the percentage of total hGH detected in the medium and cell lysates. To compare data from different experiments, the basal secretion was set to 100% and the secretion in response to NGF was expressed as a percentage of basal secretion. The data of each experiment was the result of the average of two samples processed independently (in duplicate).

PC12 cells were serum starved for 4 h, and treated or not with NGF for 30 min and then fixed with 2% paraformaldehyde and 2% glutaraldehyde (Electron Microscopy Sciences) in PBS. Cells were pelleted, embedded in agar and postfixed with 1% OsO₄ in distilled water containing 1% potassium ferricyanide for 1 h. Before dehydration, the pieces were subjected to block staining with 1% uranyl acetate in distilled water. Dehydration was performed using a graded series of cold ethanol. EmBed 812 (Taab; Reading, UK) was employed as embedding resin. Semithin sections (1-μm thick) were stained with Toluidine Blue and ultrathin sections were mounted on Collodion-coated one-hole grids, contrasted with uranyl acetate and lead citrate, and examined with a Zeiss EM900 electron microscope. Micrographs were taken with a coupled digital camera using the ImageSP software.

PC12 cells (1×10⁴ per well) were seeded in 12-well plates in medium with reduced serum (1% horse serum, 0.5% FBS) and 100 ng/ml NGF. Neurite-bearing cells were quantified at 40 h or 5 days post-treatment. The cells were considered to be differentiated when the length of at least one neurite was more than twice the length of the cell body.

Rac1 activity was measured as previously described (Katoh et al., 2000). PC12 cells were serum starved for 4 h, lysed on ice for 10 min in Fishing buffer (50 mM Tris-HCl pH 7.5, 10% glycerol, 1% NP-40, 200 mM NaCl, 2 mM MgCl₂) and lysates were clarified by centrifugation at 16,000 g for 10 min at 4°C. Lysates (600 μg) were incubated for 45 min at 4°C with 30 μg of GST–CRIB prebound to agarose–glutathione conjugates. The beads were washed three times in 1 ml of cold fishing buffer and boiled for 5 min in SDS-PAGE sample buffer. Pulled down proteins were resolved by 12% SDS-PAGE and immunoblotted using anti-Rac1 antibodies.

Data are presented as mean±s.e.m. and analyzed using Prism or Microsoft Excel software. Comparisons between means of different groups were performed using two-tailed Student's t-test after having checked normality of data and equality of variances. In secretion experiments, statistical analysis was performed using a paired two-tailed Student's t-test.

We thank Pilar Pérez for helpful discussions, Xosé Bustelo for the plasmids encoding GST–CRIB and dominant-negative form of human Rac1, Sabine Hilfiker for the PC12 cells and the hGH plasmid, and Taiho Pharmaceutical for the YM-254890 compound.