SNAP23–Kif5 complex controls mGlu1 receptor trafficking

Fabrice, Raynaud,; Vincent, Homburger,; Martial, Seveno,; Oana, Vigy,; Enora, Moutin,; Laurent, Fagni,; Julie, Perroy,

doi:10.1093/jmcb/mjy031

SNAP23–Kif5 complex controls mGlu1 receptor trafficking

Fabrice, Raynaud,;Vincent, Homburger,;Martial, Seveno,;Oana, Vigy,;Enora, Moutin,;Laurent, Fagni,;Julie, Perroy, 2018-10-01 00:00:00 Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 doi:10.1093/jmcb/mjy031 Journal of Molecular Cell Biology (2018), 10(5), 423–436 j 423 Published online September 14, 2018 Article SNAP23–Kif5 complex controls mGlu1 receptor trafﬁcking 1,3 1 2 1 1 Fabrice Raynaud , Vincent Homburger , Martial Seveno , Oana Vigy , Enora Moutin , 1 1, Laurent Fagni , and Julie Perroy IGF, CNRS, INSERM, Univ. Montpellier, F-34094 Montpellier, France BioCampus Montpellier, CNRS, INSERM, Univ. Montpellier, F-34094 Montpellier, France Present address: INSERM U1046 PhyMedExp, CNRS UMR9214, Univ. Montpellier, F-34094 Montpellier, France * Correspondence to: Julie Perroy, E-mail: Julie.Perroy@igf.cnrs.fr Edited by Xuebiao Yao Metabotropic glutamate receptors are expressed at excitatory synapses and control synaptic transmission in mammalian brain. These receptors are involved in numerous patho-physiological functions. However, little is known about the molecular determi- nants responsible for their intracellular transport and membrane targeting. Here we investigated the nature of the molecular motor and adaptor protein responsible for trafﬁcking and membrane localization of the group I metabotropic glutamate mGlu1 postsynaptic receptor in cultured hippocampal neurons. In proteomic studies, we identiﬁed the synaptosome-associated protein 23 (SNAP23) and the molecular motor Kif5 kinesin as proteins interacting with mGlu1 receptor. We showed that SNAP23, but not Kif5, directly interacts with mGlu1 receptor carboxyl terminus. Using a recombination approach to impair or enhance the inter- action between SNAP23 and Kif5, we found that the SNAP23–Kif5 complex controls the trafﬁcking of mGlu1 receptor along micro- tubules. Additional ﬂuorescence recovery after cleavage experiments allowed us to identify a role of the complex in the receptor cell surface targeting. In conclusion, our study indicates that along dendritic processes Kif5–SNAP23 complex contributes to prop- er mGlu1 receptor trafﬁcking and cell surface expression. Keywords: mGlu1, SNARE, kinesin, neuron, trafﬁc have been of intense interest in the last 10 years, but relatively Introduction little is known about the mechanisms controlling the trafﬁcking of Metabotropic glutamate (mGlu1–mGlu8) receptors are mGlu receptors. G-protein-coupled receptors that are located at brain excitatory Several proteins have been shown to control the functional synapses to regulate glutamatergic neurotransmission. The expression of mGlu1/5 receptors at postsynaptic terminals. mGlu1 and mGlu5 receptor subtypes activate the extracellular These are postsynaptic intracellular scaffolding proteins asso- signal-regulated kinase (ERK)/mitogen-activated protein kinase ciated with cytoskeleton, plasma membrane, or fusion proteins. (MAPK) pathway, as well as phospholipase C, which leads to Thus, Homer proteins (Ango et al., 2000), Tamalin (Kitano et al., intracellular calcium release and protein kinase C activation. 2002), Calmodulin (Choi et al., 2011), Neuregulin (Ledonne Regulation of functional expression of mGlu1/5 receptors plays a et al., 2015), VPS26A-SNX27 complex (Lin et al., 2015), and critical role in brain functions and neurological diseases. For Numb (Zhou et al., 2015) bind to mGlu1 and/or mGlu5 receptors instance, mGlu1 receptor is critically involved in neuronal devel- and control their organization in the postsynaptic element and opment (Ichise et al., 2000; Hannan et al., 2001), synaptic plasti- recycling at the synaptic cell surface. This is critical for synaptic city and associated learning and memory (Aiba et al., 1994a, b; plasticity and related cognitive and motor functions. However Luscher and Huber, 2010), addiction to drugs of abuse (Kenny nothing is known about carrier proteins involved in the long- and Markou, 2004), and several neuropsychiatric and neurode- range delivery of mGlu receptors before their insertion into the generative disorders (Ferraguti et al., 2008). Molecular mechan- postsynaptic membrane. isms underlying vesicle trafﬁcking of neurotransmitter receptors Soluble N-ethylmaleimide-sensitive-factor Attachment protein REceptor (SNARE) proteins are a class of membrane-associated Received December 8, 2017. Revised April 9, 2018. Accepted May 11, 2018. proteins known to regulate the process of synaptic vesicle © The Author(s) (2018). Published by Oxford University Press on behalf of Journal of Molecular Cell Biology, IBCB, SIBS, CAS. All rights reserved. fusion with the presynaptic plasma membrane (Lin and Scheller, Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 424 j Raynaud et al. 2000; Jahn and Scheller, 2006). However, these proteins are interaction in the IntAct database (Figure 1). We identiﬁed Kif5b also involved in microtubule-dependent postsynaptic trafﬁcking (kinesin-1 heavy chain), a molecular motor involved in the ves- of glutamate receptors. For instance, SNAP23 is a homolog icular trafﬁcking of AMPA receptors along microtubules SNARE protein of SNAP25 (Oyler et al., 1989; Duc and Catsicas, (Hoerndli et al., 2013), and SNAP23, a SNARE protein involved 1995; Tao-Cheng et al., 2000; Washbourne et al., 2002b) that is in the vesicular transport and exocytosis of various membrane ubiquitously expressed in neuronal and non-neuronal cells proteins (Lin and Scheller, 2000; Jahn and Scheller, 2006). To (Ravichandran et al., 1996). In hippocampal neurons, SNAP23 is address the existence of mGlu1–SNAP23–Kif5b complexes in liv- expressed in both soma and dendrites and regulates the surface ing cells, we used C6 glioma cells, because the ﬂat shape of expression, recycling, and synaptic function of the ionotropic these cells was more convenient for microtubule dynamics stud- N-methyl-D-aspartate (NMDA) glutamate receptor (Washbourne ies. Moreover, these cells endogenously express mGlu1 receptor et al., 2004; Suh et al., 2010). Other studies have shown that (Luis Albasanz et al., 2002; Viwatpinyo and Chongthammakun, NMDA receptors are transported in clusters along microtubules 2009; Castillo et al., 2010) and we show that they also express (Washbourne et al., 2002a) in vesicles containing the kinesin pro- SNAP23 and Kif5 proteins (Figure 2C and D). Immunoﬂuorescent tein Kif17 (Setou et al., 2000; Guillaud et al., 2003). The experiments showed that mcherry-tagged mGlu1a receptor colo- microtubule-dependent motor Kif5 and the associated kinesin calized with YFP-tagged Kif5b or YFP-tagged SNAP23 in vesicles, light chain Klc2 would mediate the transport of the alpha-amino- in C6 glioma cell line (Figure 2A and B). Furthermore, YFP- 3-hydroxy-5-methylisoazol-4-propionate (AMPA) glutamate recep- tagged mGlu1a receptor could be immunoprecipitated with Kif5 tor subunit GLR-1 to the synapse, in Caenorhabditis elegans.This (Figure 2C) and SNAP23 in transfected neuroblastoma cell line Kif5-dependent transport would control rapid delivery, removal (Figure 2D). Altogether these results suggested that mGlu1a and redistribution of functional postsynaptic AMPA receptors, receptor can form a complex with Kif5 and SNAP23 in model and thus synaptic strength (Hoerndli et al., 2013). cell line. At present nothing is known about the transport of mGlu receptors along microtubules. Here we report that mGlu1 recep- Interaction between mGlu1a receptor and Kif5–SNAP23 complex tor can form a cargo with SNAP23 in microtubule-associated involves mGlu1a receptor and SNAP23 carboxyl termini vesicles conveyed by Kif5 along dendrites of cultured hippocam- Kif5 and SNAP23 were known to form a complex (Diefenbach pal neurons. This result provides the ﬁrst evidence for a molecu- et al., 2002), and the above results showed that mGlu1a recep- lar mechanism of vesicular transport of postsynaptic mGlu tor belongs to the Kif5–SNAP23 complex. We therefore investi- receptors in neurons. gated the molecular determinants of the interaction between mGlu1a receptor and Kif5–SNAP23 complex in neuroblastoma Results cell line. Deletion of the C-terminus of mGlu1areceptor(YFP- The Kif5–SNAP23 complex interacts with mGlu1a receptor mGlu1aΔC) strongly decreased co-immunoprecipitation of the In order to identify multiprotein complexes interacting with receptor mutant with both Kif5 and SNAP23 (Figure 2C and D). the intracellular mGlu1a receptor C-terminus, we performed dif- These experiments showed that the C-terminal domain of mGlu1a ferential proteomic studies in an easily transfectable neuron- receptor was required for the interaction between the receptor derived cell line, the N2a neuroblastoma cell line expressing and Kif5–SNAP23 complex. either full-length YFP-mGlu1a or C-terminus-truncated mGlu1a We then investigated whether mGlu1a receptor could directly (YFP-mGlu1aΔC) receptors. The mGlu1 sequences coverage interact with SNAP23 or Kif5. We measured bioluminescence found using mGlu1aΔC as a bait conﬁrmed the absence of resonance energy transfer (BRET) signals between these pro- C-terminus mapping (Supplementary Figure S1). Differential teins tagged with BRET donor and acceptor compatible entities analysis of proteins that were immunoprecipitated with these (mGlu1a-Rluc8 and YFP-SNAP23 or YFP-Kif5). For a constant receptors indicated that 173 proteins interact with the full- expression level of mGlu1a-Rluc8, the BRET signal increased length receptor, but not with YFP-mGlu1aΔC (Supplementary hyperbolically as a function of the YFP-SNAP23 expression level, Table S1). Only proteins with a fold enrichment of six (based on indicating that mGlu1a receptor speciﬁcally interacts with SNAP23 the MSMS ratio) and a minimum of two unique peptides per (Figure 3A). In contrast, only a weak and linear, non-speciﬁc protein were indeed considered as potential interactors of the BRET signal could be detected between mGlu1a and Kif5 pro- carboxyl terminus of the mGlu1 receptor. To determine if these teins (Figure 3B). These results suggested that SNAP23 but not proteins belong to speciﬁc physiological categories, we next Kif5 directly interacts with mGlu1a receptor. This SNAP23 direct performed a gene ontology (GO) analysis based on both bio- interaction could also be detected with mGlu5 but not mGlu7 recep- logical process and cellular component domains. Interestingly, tors (Supplementary Figure S2), suggesting that the interaction numerous proteins are involved in signaling pathways, trans- was speciﬁc to group I mGlu receptors (mGlu1aand mGlu5). port, and neuronal functions. In search for proteins involved in We then searched for the interaction domains responsible for mGlu1 receptor trafﬁcking, we focused on proteins annotated to the mGlu1a receptor–SNAP23 interaction by measuring BRET the ‘intracellular transport’ GO term (GO:0046907). Among the signals between C-terminal deletion mutants of these proteins, 173 potential interactors of mGlu1 C-tail, 29 proteins belong to in transfected neuroblastoma cell line. Successive deletions of this functional category, 9 with known protein–protein the distal part of mGlu1 C-terminus did not affect its BRET Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 425 Figure 1 mGlu1a receptor-associated protein network. Searching for proteins involved in mGlu1 receptor trafﬁcking, the network is focused on the nine proteins displayed in the internal circle that are annotated to the ‘intracellular transport’ GO term and have protein–protein interactions in the IntAct database. It consists of 28 nodes and 22 edges. Node colors reﬂect the MS/MS ratio and vary from red (high ratio) to light yellow (low ratio). Thick node border highlights proteins annotated to the ‘intracellular transport’ GO term. Node shapes represent different protein taxonomies: circle (mouse), diamond (human), square (rat), and hexagon (bovine). The edge width is proportional to the number of published manuscripts—curated in the IntAct database—that have reported molecular interaction data. The edge stroke color maps the conﬁdence value: from 0.6 (light gray) to the local maximum 0.92 (dark gray). signals with wild-type SNAP23 until the receptor lacks all but SNAP23 may bind to Kif5 via its N-terminus (Diefenbach et al., the 21 proximal amino acids of the tail (mGlu1aΔ341, 2002) and serves as a cargo for the transport of mGlu1areceptor Figure 3F). This result showed that interaction with SNAP23 by the molecular motor Kif5. requires the proximal C-terminus of mGlu1a receptor until amino acid position 971 (mGlu1aΔ229, Figure 3E). BRET signals were Kif5 controls microtubule trafﬁcking and membrane targeting of also detected between wild-type mGlu1a receptor and SNAP23 the mGlu1a receptor mutants deleted from the last 82 C-terminal amino acids, but Since we found that mGlu1a receptor is part of a Kif5- were lost after deletion of the last 122 C-terminal amino acids of containing complex and because Kif5 is involved in vesicular the protein (SNAP23Δ122, Figure 3J). These results indicated transport of proteins (Hirokawa et al., 2010), this molecular that interaction of SNAP23 with mGlu1a receptor required the motor could regulate trafﬁcking along microtubules and cell sur- amino acids 88–128 of SNAP23. Altogether these results show face expression of the mGlu1a receptor. We investigated this that mGlu1a receptor directly interacts with SNAP23 and that this hypothesis using rapalog-induced reconstruction of Kif5 protein interaction depends on C-terminal domains of both proteins. (see Jenkins et al., 2012 for the methodology) in C6 glioma cell Thus these results reinforce the above tentative hypothesis that line. Brieﬂy, this approach, based on the reconstruction of a Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 426 j Raynaud et al. Figure 2 The mGlu1a receptor forms a complex with Kif5b and SNAP proteins. (A and B) Co-localization of mcherry-mGlu1a receptor with YFP- Kif5 (A) or YFP-SNAP23 (B) proteins in C6 glioma cells. Following a temperature shift at 20°C to synchronize post-Golgi trafﬁc and thrombin- mediated cleavage of cell surface receptor’s mcherry tag, time-lapse images of intracellular protein trafﬁcking were taken at 34°C as described in Material and methods section. Arrows indicate mcherry-mGlu1a receptor and YFP-Kif5 (A) or YFP-SNAP23 (B) vesicle co-localization. (C and D) Kif5–SNAP23–mGlu1a receptor complex revealed by co-immunoprecipitation experiments. Neuroblastoma cells were transfected with plas- mids encoding wild-type or C-terminus-truncated YFP-mGlu1areceptors and Kif5 (C)orSNAP23 (D) proteins. Cells were washed, solubilized, and subjectedtoimmunoprecipitation(IP)withGFP-trap andrevealedbywestern blots(WB)witheither anti-Kif5 (C)oranti-SNAP23 (D) antibodies. split kinesin by chemical dimerization, enables the identiﬁcation that mGlu1a receptor association with Kif5 complex is essential of motor proteins that interact with speciﬁc vesicle populations. for the receptor plasma membrane targeting. To characterize the trapping of SNAP23 and mGlu1 receptor by We further investigated this hypothesis by measuring the Kif5 on microtubules, we used the rigor mutant of Kif5, which is dominant-negative effects of two Kif5 mutants, Kif5 rigor and stalled on microtubules. Figure 4A shows that mcherry-tagged Kif5 unbound, in C6 glioma cell line. Kif5 rigor is an active form cargoless-Kif5 recombined with its YFP-tagged cargo binding of Kif5 that binds to microtubules and blocks microtubule- domain (CBD) decorates microtubules. The YFP-tagged SNAP23 dependent protein transport. This mutant inhibited cell surface protein colocalized with recombined Kif5 protein on ﬁlamentous expression of YFP-tagged mGlu1a receptor (Figure 5B), suggest- structures (Figure 4B), suggesting the recruitment of SNAP23 by ing that membrane expression of the receptor involves Kif5 on microtubules. Similarly, forced interaction between microtubule-dependent vesicular trafﬁc of proteins. The Kif5 SNAP23 and Kif5 using rapalog (Figure 4C and Supplementary unbound mutant is an inactive form of Kif5 that lacks ability to Movie S1) recruited mcherry-tagged mGlu1a receptor to micro- interact with microtubules and therefore aborts the transport of tubule like structures (Figure 4D). Together, these results sug- Kif5-speciﬁc cargo. This mutant also inhibited membrane expres- gest that mGlu1a receptor can form a complex with SNAP23 and sion of YFP-tagged mGlu1a receptor (Figure 5B), thus suggesting Kif5 on microtubules, and that SNAP23 may serve as cargo for involvement of Kif5 in membrane trafﬁcking of the receptor. We the transport of mGlu1a receptor by the molecular motor Kif5. then investigated the speciﬁcity of the effect of Kif5 on mGlu1a We then investigated whether mGlu1a receptor-Kif5 interaction receptor membrane trafﬁcking by comparing the effects of Kif5 could control membrane targeting of the receptor. The mGlu1a and Kif3 unbound mutants on membrane expression of YFP- receptor was tagged with an extracellular-thrombin cleavable YFP tagged mGlu1a receptor in FRAC experiments. Only Kif5 mutant tag and transfected in C6 glioma cells. Following thrombin- signiﬁcantly inhibited mGlu1a receptor plasma membrane induced extinction of cell surface receptor ﬂuorescence, we could recovery, 1 h after cleavage of the receptor extracellular YFP measure intracellular receptor trafﬁcking efﬁciency and kinetics tag with thrombin (Figure 5C).Wealsotestedthe effects of from the Golgi to the cell surface, by ﬂuorescence recovery Kif5 rigor and Kif5 unbound mutants on cell surface targeting after cleavage (FRAC) experiments. Interestingly, deletion of the of N-terminal YFP-tagged mGlu7 receptor in FRAC experiments. C-terminus of mGlu1a receptor (YFP-th/mGlu1aΔC) signiﬁcantly Unlike YFP-tagged mGlu1 receptors, the membrane targeting of delayed plasma membrane targeting of the mutant receptor over YFP-mGlu7 receptor was unaffected by Kif5 mutants (Figure 5D). aperiodof 120 min (Figure 5A). Knowing that the C-terminus These experiments suggested that Kif5 controls membrane tar- of mGlu1a is required for interaction of the receptor to the geting of mGlu1a receptor via microtubule-dependent transport Kif5-containing complex (Figure 3), these experiments suggested in C6 glioma cell line. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 427 Figure 3 Identiﬁcation of mGlu1a receptor and SNAP23 interaction domains by BRET analysis. (A and B) mGlu1a directly interacts with SNAP23, but not Kif5. BRET signal between mGlu1a-Rluc8 and YFP-SNAP23 (A) or mGlu1a-Rluc8 and YFP-Kif5 (B) constructs. Neuroblastoma cells were co-transfected with a ﬁxed concentration of mGlu1a-Rluc8 and increasing concentrations of YFP-SNAP23 (A) or YFP-Kif5 (B). (C–F) mGlu1a receptor proximal C-terminus (until amino acid position 971) is required for its interaction with SNAP23. BRET signal between trun- cated mGlu1a-Rluc8 receptors and YFP-SNAP23. Neuroblastoma cells were co-transfected with a ﬁxed concentration of C-terminus-truncated mGlu1a receptor fused to Rluc8 constructs as illustrated, and increasing concentrations of YFP-SNAP23 plasmids. (G–J) SNAP-23 amino acids 88–128 are required for its interaction with mGlu1a receptor. BRET signal between mGlu1a-Rluc8 receptors and C-terminus-truncated YFP-SNAP23 proteins. Neuroblastoma cells were co-transfected with a ﬁxed concentration of Rluc8-tagged mGlu1a receptor and increasing concentrations of YFP-tagged truncated SNAP23 mutants as illustrated. The data shown are representatives of ﬁve independent experiments performed in triplicate and expressed as mean ± SEM. The curves were ﬁtted using a nonlinear regression equation, assuming a single bind- ing site. Note that mGlu1a-Rluc8 receptor did not interact with YFP-Kif5 (B), mGlu1aΔ341 receptor mutant did not interact with YFP-SNAP23 (F), and YFP-SNAP23Δ122 mutant did not interact with mGlu1a-Rluc8 receptor (J), as indicated by a BRET signal that increased linearly (rather than hyperbolically) with the YFP-tagged protein expression level, which most likely reﬂected random collision. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 428 j Raynaud et al. Figure 4 Kif5–SNAP23 complex controls microtubule trafﬁcking of the mGlu1a receptor. Visualization of Kif5–SNAP23–mGlu1a receptor com- plex after rapalog treatment in C6 glioma cell line. Images were obtained in the absence (upper line) or presence (bottom line) of the rapa- log dimerizer. The framed area is magniﬁed in the right bottom corner. (A) Kif5 rigor-stained microtubules after rapalog-induced reconstruction of split Kif5 (arrows). Cells were transfected with Kif5 cargoless/mcherry-FKBP and YFP-tagged rapamycin-binding domain (FRB) associated with kinesin CBD (FRB/YFP-CBD) constructs. (B) SNAP23-stained microtubules after rapalog-induced Kif5 rigor reconstruc- tion (arrows). Cells were transfected with split Kif5 cargoless/FKBP, FRB/mcherry-CBD, and YFP-SNAP23 constructs. (C) YFP-SNAP23-decorated micro- tubules after rapalog-induced forced interactions between SNAP23 and Kif5. Cells were transfected with Kif5 rigor cargoless/mcherry-FKBP and FRB/YFP-SNAP23 constructs. (D) mcherry-mGlu1a-stained microtubules induced by rapalog-forced interactions between SNAP23 and Kif5. Cells were transfected with the split Kif5 rigor cargoless/FKBP, FRB/YFP-SNAP23, and mcherry-mGlu1a receptor constructs. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 429 Figure 5 Kif5 is involved in mGlu1a receptor post-Golgi trafﬁc. (A) mGlu1a receptor cell surface targeting depends on the receptor C-terminus. Schematic view of FRAC experiments (upper part). YFP-th/mGlu1a wild-type receptor (YFP-th/mGlu1a) or C-terminus-truncated receptor (YFP-th/mGlu1aΔC) constructs were transfected in C6 glioma cells. Cell surface recovery of each receptor construct was expressed at different time points after thrombin-induced cleavage of the extracellular YFP tag and expressed as percentage of control wild-type recep- tor surface expression obtained in the absence of thrombin treatment (nc white bar). For this and following panels, values are mean ± SEM of three independent experiments. (B) Negative kif5 mutants inhibit cell surface mGlu1a receptor targeting. C6 glioma cells were co- transfected with YFP-th/mGlu1a receptor and kif5 rigor or kif5 unbound mutant constructs and subjected to FRAC experiments (same as in A). Results indicate a 50% inhibition of maximal cell surface receptor recovery in the presence of dominant-negative kif5 mutants. (C) Dose– response relationship and kif5-speciﬁc inhibition of cell surface YFP-th/mGlu1a receptor targeting. C6 glioma cells were co-transfected with a ﬁxed concentration of YFP-th/mGlu1a receptor construct and increasing concentrations of dominant-negative Kif5 or Kif3 unbound mutant constructs. Cells were then subjected to FRAC experiments to measure cell surface recovery of YFP-th/mGlu1a receptor 1 h after cleavage of the YFP tag with thrombin (same as in A). Results indicate inhibition of cell surface receptor recovery with increasing concentrations of Kif5, but not Kif3 unbound constructs. (D) Dominant-negative Kif5 mutants did not modify mGlu7a receptor cell surface expression. Cells were co- transfected with YFP-th/mGlu7a receptor and Kif5 rigor or Kif5 unbound mutants and subjected to FRAC experiments to measure cell surface YFP-th/mGlu7a receptor recovery after cleavage of its external YFP tag (same as in A). Results indicate no modiﬁcation in receptor recovery in the presence of dominant-negative Kif5 rigor and unbound mutants. The Kif5–SNAP23 complex controls mGlu1a receptor membrane investigated the role of this complex in mGlu1a receptor mem- targeting in neurons brane targeting. Deletion of the C-terminus of mGlu1a receptor We then investigated whether the role of Kif5–SNAP23 com- (YFP-th/mGlu1aΔC, Figure 7A) delayed membrane targeting of plex in membrane targeting of mGlu1a receptor that we found YFP-tagged mGlu1a receptor in neurons, indicating a role of the in C6 glioma cell line also applied to neurons. Transfected C-terminus of the receptor in this process. As Kif5 was found to mcherry-tagged mGlu1a receptor colocalized with YFP-tagged control membrane targeting of mGlu1a receptor in C6 glioma cells Kif5 or YFP-tagged SNAP23 in mouse cultured hippocampal neu- (Figure 5), we investigated whether this also applied to neurons. rons (Figure 6A and B; Supplementary Movies S2−S4). These We found that Kif5 unbound mutants altered membrane recovery results suggest that the Kif5–SNAP23 complex is involved in the of YFP-tagged mGlu1areceptor(Figure 7B) in FRAC experiments, vesicular transport of mGlu1a receptor in neurons. We then in neurons. Conversely, deletion of postsynaptic density protein, Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 430 j Raynaud et al. Figure 6 Co-localization of mcherry-mGlu1a receptor with Kif5 and SNAP23 proteins in hippocampal neurons. Arrows indicate co-localization of mcherry-mGlu1a receptor with Kif5 (A) and SNAP23 (B). Drosophila disc large tumor suppressor, zonula occludens-1 protein Our study was focused on mGlu1 receptor trafﬁcking along (PDZ) and homer interaction motifs included in the C-terminus of dendrites, because we found that mGlu1 receptor in cultured mGlu1a receptor did not affect the membrane targeting of the mouse hippocampal neurons is only expressed in dendrites receptor (Figure 7C). Together these experiments suggest that the (data not shown; Das and Banker, 2006). However, in other cell molecular motor Kif5 binds to the cargo protein SNAP23 and types, mGlu1 receptors can be expressed at presynaptic sites as recruits mGlu1a receptors to control the receptor membrane target- well. The nature of molecular motors and adaptors involved in ing in neurons. the presynaptic trafﬁcking of mGlu1 receptors is still unknown. Interaction with other protein partners might trigger speciﬁc Discussion subcellular sorting and trafﬁcking of the receptor. Here we revealed aspects of the molecular mechanisms of A proteomic approach following mGlu1 receptor immunopreci- postsynaptic mGlu1 receptor transport in neurons. We ﬁrst iden- pitation experiments allowed us to identify a large number of tiﬁed Kif5 and SNAP23 as molecular motor and adaptor protein proteins including nine proteins known to be involved in cellular associated with mGlu1 receptor. We show that SNAP23–Kif5 transport (Figure 1). Among these proteins, we focused on Kif5 complex controls mGlu1 receptor dendritic trafﬁcking via and SNAP23. Kif5 and SNAP23 were known to form a complex microtubule-dependent transport. Previous studies have shown (Diefenbach et al., 2002). Here we corroborated this result and that the molecular motor myosin VI and Kif17 (Wu et al., 2002), further showed that Kif5–SNAP23 complex could in turn interact and vesicular fusion protein SNAP23 and SNAP25 (Osten et al., with the carboxyl terminus of mGlu1 receptor thus forming a 1998; Noel et al., 1999; Washbourne et al., 2004; Suh et al., Kif5–SNAP23–mGlu1 receptor complex. We found that this com- 2010) regulate synaptic delivery of ionotropic AMPA and NMDA plex was recruited on microtubule like structures, which was con- glutamate receptors. Thus the present and previous results fur- sistent with the hypothesis that the molecular motor Kif5 could ther argue in favor of an important role of kinesins and SNARE be responsible for the vesicular transport of the SNAP23–mGlu1 proteins in the postsynaptic delivery of glutamate receptor fam- receptor cargo along dendritic microtubules. As SNAP23 belongs ily, therefore participating to the formation and possibly matur- to the large family of SNARE proteins involved in membrane ation of glutamatergic synapses. fusion processes, SNAP23 may also control exocytosis of the Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 431 Figure 7 mGlu1a receptor C-terminus- and Kif5-dependent cell surface targeting of YFP-th/mGlu1a receptor in cultured hippocampal neu- rons. (A) Deletion of YFP-th/mGlu1 receptor C-terminus impairs cell surface targeting of YFP-th/mGlu1a receptor. Wild-type YFP-mGlu1a receptor (mGlu1a) or C-terminus-truncated receptor (mGlu1aΔC) constructs were transfected in hippocampal neurons (same as Figure 5A). For this and following panels, values are mean ± SEM of three independent experiments. (B) Kif5 unbound mutant prevents YFP-th/mGlu1 receptor cell surface recovery. FRAC experiments performed in neurons co-transfected with wild-type YFP-th/mGlu1a receptor (mGlu1a) and Kif5 unbound mutant. (C) PDZ and homer interaction motifs are not involved in cell surface YFP-th/mGlu1a receptor targeting. Wild-type YFP-mGlu1a receptor (mGlu1a) or receptor deleted from its PDZ (mGlu1aΔ-PDZ) or homer (mGlu1Δ-homer) binding motif was transfected in hippocampal neurons. Cells were then processed as in B. mGlu1 receptor to the cell surface. This hypothesis was consist- neuronal cells. In the brain, SNAP23 and SNAP25 are both ent with our ﬁnding that the deletion of the carboxyl terminus of located in the somato-dendritic neuronal compartment, but dis- mGlu1 receptor delayed the membrane targeting of the receptor. play non-overlapping distribution. SNAP23 mediates the mem- The existence of a Kif5–SNAP23–mGlu5 complex suggests that brane delivery of NMDA receptors (Suh et al., 2010). Here we Kif5–SNAP23-mediated transport could apply to both mGlu1 and show that SNAP23 regulates the dendritic trafﬁcking and postsy- mGlu5 receptors. naptic membrane delivery of mGlu1 receptors. Thus consistent SNARE proteins mediate membrane fusion events in various with the somato-dendritic location of SNAP23, both mGlu1 and cell types (Lin and Scheller, 2000; Jahn and Scheller, 2006). For NMDA receptors bind to SNAP23. However, these receptors instance, SNAP25 is a neuron-speciﬁc SNARE protein that regu- seem to display distinct vesicular transport. The NMDA receptor lates synaptic vesicle fusion with the presynaptic plasma mem- was shown to be transported by the molecular motor Kif17 brane (Lin and Scheller, 2000; Jahn and Scheller, 2006), and (Setou et al., 2000; Guillaud et al., 2003). Whether SNAP23 and membrane delivery of postsynaptic AMPA receptors (Hoerndli Kif17 belong to the same complex to transport NMDA receptor et al., 2013). Postsynaptic translocation of AMPA receptors has is still an open question. If so, this would imply that mGlu1 and also been shown to depend on the actin motor myosin VI and NMDA receptors are sorted to two different motors (Kif5 and the adaptor protein SAP97 (Wu et al., 2002). The SNAP25 homo- Kif17) through the same adaptor. An interesting topic would log, SNAP23, is ubiquitously expressed in neuronal and non- then be to understand which molecular determinant beside the Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 432 j Raynaud et al. adaptor itself could specify the interaction with selective Rluc8, and p-mGlu1a-Δ62-Rluc8 were obtained by inserting molecular motors, such as speciﬁc cargo-adaptor conformation Rluc8 ended by a stop codon after positions 859, 971, 1029, or post-translational modiﬁcations. An attractive alternative and 1138 of p-mGlu1a coding sequence (1200 amino acids). In hypothesis would be that adaptors would specify the nature of the plasmids p-mGlu1a-Δ229-Rluc8, p-mGlu1a-Δ171-Rluc8, and the molecular motor, which can be supported by the fact that p-mGlu1a-Δ62-Rluc8, the retention motif RRKK in the C-terminal mLin-10 has been identiﬁed as adaptor protein for Kif17 in the coding sequence of mGlu1 was exchanged for AAAA to avoid transport of NMDA receptors (Setou et al., 2000). Hence more endoplasmic reticulum retention as previously described for this explorations are needed to fully understand these molecular receptor (Chan et al., 2001). mechanisms, but Kif5, Kif17, and myosin VI molecular motors as The plasmid YFP-th-mGlu7a was obtained by introducing YFP-th well as SAP97, SNAP23, and SNAP25 proteins seem to play before the coding sequence of p-mGlu7a. essential roles in glutamate receptors speciﬁc trafﬁcking. Kif5band Kif5c constructs were generated from the mouse Metabotropic glutamate receptors modulate efﬁcacy of synap- Kif5band mouseKif5cIMAGE clones (#30543821 and #30536079, tic transmission mediated by the ionotropic AMPA and NMDA respectively; Gene Service, Source BioScience). The rigor muta- glutamate receptors. Regulation of mGlu receptors delivery at tion was obtained using point mutation-containing primers to postsynaptic sites is essential for an optimal functioning of the substitute the threonine 92 by a valine. We created the unbound synapse. The strength of synaptic transmission also depends on mutation using point mutation-containing primers to substitute the number of functional postsynaptic ionotropic AMPA and the serine 204 by a glycine and the histidine 205 by a glycine. NMDA glutamate receptors. For instance, long-term potentiation The plasmids p-YFP-Kif5b and p-YFP-Kif5c were obtained by PCR and long-term depression of synaptic transmission can result ampliﬁcation of the EYFP coding sequence from p-EYFP and from an increased and decreased number of active postsynaptic insertion into p-Kif5b and p-Kif5c. ionotropic glutamate receptors, respectively. Parallel modiﬁca- To obtain p-FRB-YFP-CBD, we ampliﬁed by PCR the Kif5b cod- tion in the number of mGlu receptors should therefore be ing sequence (amino acids 816–964) from p-Kif5b and inserted required to maintain an optimal tuning of either the potentiated into p-FRB-YFP-snap23. The sequence 816–964 contains the or depressed synapse. Thus kinesins and SNARE proteins would CBD and the minimal domain for SNAP23 binding (814–907)as play essential roles not only during synaptogenesis, but also previously described (Diefenbach et al., 2002). during synaptic plasticity by controlling appropriate delivery of p-Kif5 rigor cargoless-mcherry was obtained by introducing metabotropic receptors to maintain adequate synaptic tuning. It mcherry after position 560 in Kif5c coding sequence. This plas- is striking that glutamate receptors use different proteins for mid was further transformed into p-Kif5 rigor cargoless-mcherry- their intracellular trafﬁcking although they may be translocated FKBP using pC M-F2E (ARIAD Pharmaceuticals, Inc); and p-Kif5 to same postsynaptic sites. One possible reason for such differ- rigor cargoless-FKBP was obtained by mcherry coding sequence ent mechanisms is that each receptor may need speciﬁc regula- removal. tion for appropriate tuning of the synapse. Therefore, this p-Kif3a was generated from the human Kif3a IMAGE clone implies ﬁne regulation of the glutamate receptor assembly with (#5298675; Gene Service, Source BioScience). From p-Kif3a, we their cargo proteins, a process that remains largely elusive created the unbound mutation using point mutation-containing currently. primers to substitute the serine 220 by an alanine and the histi- dine 221 by an alanine. Materials and methods SNAP23 constructs were generated with the mouse SNAP23b DNA constructs IMAGE clone (#30606358; Gene Service, Source BioScience). YFP-mGlu1a, mGlu1aΔC(equivalent to mGlu1aΔ341), mGlu1a- The plasmid p-YFP-SNAP23 was obtained by subcloning the YFP, and mGlu7a plasmids were gifts from Dr Jean-Philippe Pin PCR ampliﬁed SNAP23 coding sequence (210 amino acids) into (IGF, CNRS, INSERM, Universite´ Montpellier). p-EYFP (Clontech); p-YFP-SNAP23Δ8, Δ49, Δ82, and Δ122 were The plasmid YFPth-mGlu1a was obtained by introducing the obtained by introducing a stop codon after positions 202, 161, thrombin cleavage sequence GGLVPRGSGG immediately after the 128, and 88 of the SNAP23b coding sequence. YFP sequence by site-directed mutagenesis using oligonucleotide: The plasmid p-FRB-YFP-SNAP23 was obtained by insertion of 5′-ATACACGCGTACCACCACTGCCTCTCGGAACTAAACCACCCTTGTACA the FRB coding sequence from pC -R E (ARIAD Pharmaceuticals, 4 H GCTCGTCCAT-3′ and subcloning. The plasmid mcherry-th-mGlu1a Inc.) into p-YFP-SNAP23, after introduction of a Spe1 restriction was obtained by exchanging YFP for mcherry. The plasmid site just after the initiation codon in p-YFP-SNAP23. All mutants p-YFP-th-mGlu1aΔC was obtained by introducing YFP-th in were generated by PCR and veriﬁed by sequencing. p-mGlu1aΔC. The plasmid p-mGlu1a-Rluc8 was obtained by exchanging YFP from p-mGlu1a-YFP for Rluc8 from pcDNA-Rluc8. Neuroblastoma cells, C6 glioma cells, and hipppocampal The plasmid p-YFP-th-mGlu1aΔPDZ was obtained by mutation neuronal culture of PDZ binding motif (STL) in p-YFP-th-mGlu1a. The plasmid Neuroblastoma N2a cells were obtained from Dr A. Varrault p-YFP-th-mGlu1aΔhomer was obtained by mutation of homer (IGF, Montpellier). They were cultured as previously described binding motif (PPSPFR) in p-YFP-th-mGlu1a. The plasmids (Brabet et al., 1991). Cells were transfected with lipofectamine p-mGlu1a-Δ341-Rluc8, p-mGlu1a-Δ229-Rluc8, p-mGlu1a-Δ171- 2000 (Invitrogen) according to the manufacturer’s instructions. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 433 C6 glioma cells were cultured and transfected as described pre- (M) was speciﬁed as variable modiﬁcation. Database searches viously (Giau et al., 2005). Wistar rat hippocampal cultures were were performed with a mass tolerance of 20 ppm for precursor prepared from E-17/18 embryos of either sex and grown in ion for mass calibration, and with a 4.5-ppm tolerance after cali- Neurobasal medium supplemented with 0.5mM L-glutamine, 2% bration. The maximum false peptide and protein discovery rate B27, and 10% heat-inactivated horse serum (all from Invitrogen). was speciﬁed as 0.01. Seven amino acids were required as min- Hippocampal neurons were transfected at DIV-9 with lipofecta- imum peptide length. The MaxQuant software generates several mine 2000 (Invitrogen) according to the manufacturer’s stand- output ﬁles that contain information about identiﬁed peptides ard protocol. For immunoﬂuorescence and FRAC experiments, and proteins. The ‘proteinGroups.txt’ ﬁle is dedicated to the iden- cell lines were plated onto coverslips (12 or 35 mm diameter) tiﬁed proteins: each single row collapses into protein groups all and hippocampal neurons onto coverslips coated with 50 μg/ml proteins that cannot be distinguished based on identiﬁed pep- poly-D-lysine (Sigma). tides. An in-house bioinformatics tool has been developed to automatically select a representative protein ID in each protein Protein separation and identiﬁcation by LC–MS/MS group. First, proteins with the most identiﬁed peptides are iso- YFP-mGlu1aorYFP-mGlu1aΔC receptors immunoprecipitated lated in a so called ‘match group’ (proteins from the ‘Protein IDs’ with GFP-Trap-A were separated on SDS-PAGE gels (12% polyacryl- column with the maximum number of ‘peptides counts (all)’). For amide, Mini-PROTEAN TGX™ Precast Gels, Bio-Rad) and stained the remaining match groups where more than one protein ID with Protein Staining Solution (Euromedex). Gel lanes were cut existed after ﬁltering, the ‘leading’ protein has been chosen as into ﬁve gel pieces and destained with 50 mM triethylammonium the best annotated protein in UniProtKB (reviewed entries rather biCarbonate (TEABC) and three washes in 100% acetonitrile. than automatic ones, the one with the highest protein existence Proteins were digested in gel using trypsin (500 ng/band, Gold, evidence or the most annotated protein according to the number Promega) as previously described (Thouvenotetal., 2008). Digest of GO annotations dated from 05/09/2017). products were dehydrated in a vacuum centrifuge and reduced to 3 μl. The generated peptides were analyzed online by nano- Bioinformatics and network analysis ﬂowHPLC–nanoelectrospray ionization using on a LTQ-Orbitrap XL Only proteins with a fold enrichment of six (based on the MS/MS mass spectrometer (Thermo Scientiﬁc) coupled to an Ultimate 3000 ratio) and a minimum of two unique peptides per protein were con- HPLC (Thermo Fisher Scientiﬁc). Desalting and pre-concentration of sidered as potential interactors of the carboxyl terminus of the samples were performed online on a Pepmap pre-column mGlu1 receptor (Supplementary Table S1). Protein–protein interac- (0.3 mm × 10 mm, Dionex). A gradient consisting of 0–40%B tions of these 173 proteins have been downloaded from the IntAct for 60 min and 80% B for 15 min (A = 0.1% formic acid, 2% database (Orchard et al., 2014)(October 12, 2017)and ﬁltered acetonitrile in water; B = 0.1% formic acid in acetonitrile) at 300 based on an edge score set to 0.6 (high conﬁdence). Protein net- nl/min was used to elute peptides from the capillary reverse- work analysis was performed using Cytoscape (Shannon et al., phase column (0.075 mm × 150 mm, Acclaim Pepmap 100 2003)(version 3.5.1) and the ClueGO application (Bindea et al., C18, Thermo Fisher Scientiﬁc). Eluted peptides were electro- 2009)(v2.3.4) was used to handle GO annotations (from EBI- sprayed online at a voltage of 2.4 kV into a LTQ-Orbitrap XL QuickGO-GOA, September 6–October 16, 2017). A focus was made mass spectrometer. A cycle of one full scan mass spectrum on proteins annotated to the ‘intracellular transport’ GO term, (400–2000 m/z) at a resolution of 60000 (at 400 m/z), followed GO:0046907 (29/173 proteins; 9/29 with known PPI in IntAct). by ﬁve data-dependent MS/MS spectra was repeated continu- ously throughout the nanoLC separation. All MS/MS spectra Immunoprecipitation and western blots were recorded using normalized collision energy (35%, activa- Neuroblastoma cells were transfected with YFP-mGlu1aor tion) with an isolation window of 3 m/z. Data were acquired C-terminus-truncated YFP-mGlu1a (YFP-mGlu1aΔC) receptor. using the Xcalibur software (v 2.0.7). For all full scan measure- Twenty-four hours after transfection, cells (3 × 10-cm diameter ments with the Orbitrap detector, a lock-mass ion from ambient plates per condition) were washed and subjected to 20°C tem- air (m/z 445.120024) was used as an internal calibrant as perature blockade for 2 h without thrombin and then shifted to described (Olsen et al., 2005). Analysis of MS data was per- 37°C for 30 min as described for FRAC in C6 glioma cells (see formed using the MaxQuant software package (v 1.5.5.1)(Cox below). Cells were lysed and lysates processed as previously and Mann, 2008). Tandem mass spectra (MS/MS) were searched described (Moutin et al., 2014) using GFP-Trap-A (ChromoTek) in by the Andromeda search engine (Cox et al., 2011) against the place of RFP-Trap-A. After washing, the solid phase was incu- UniProtKB Reference Proteome UP000000589 database for the bated in Laemmli buffer at 60°C and protein samples resolved Mus musculus taxonomy (release 2017_08, 22277 canonical by PAGE on a 10% gel, transferred onto nitrocellulose sheet and entries) and the speciﬁc sequence of interest using the following subjected to immunoblotting using anti-kif5 antibody (1/2000,H2, parameters: enzyme speciﬁcity was set as Trypsin/P, and a max- MAB1614, Chemicon Int., Millipore) and anti-snap23 (1/2000, imum of two missed cleavages and a mass tolerance of 0.5 Da for Synaptic Systems or Abcam) for 2 h. Blots were then washed fragment ion were applied. A second database of known contami- three times with PBS containing 0.1% Tween-20 (PBST) and nants provided with the MaxQuant suite was also employed. The incubated with donkey anti-rabbit or anti-mouse peroxidase- ‘match between runs’ and ‘iBAQ’ options were checked. Oxidation conjugated antibodies (1/2000, Rockland) for 1 h. Blots were Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 434 j Raynaud et al. washed ﬁve times with PBST and proteins were visualized with FRB-tagged protein consisted of CBD from Kif5 tail (FRB-mcherry- ECL Westen blotting detection reagent (Amersham, GE Healthcare CBD or FRB-YFP-CBD) and YFP-SNAP23 (FRB/YFP-SNAP23). Life Sciences) on ChemiDoc Imaging System (Bio-Rad). Twenty-four hours after cell transfection, heterodimerization of the FRB-tagged and FKBP-tagged proteins was induced by the BRET measurements addition of 1 μM heterodimerizer AP21967 (Ariad Pharmaceuti- We have previously described BRET measurements in cell cals), a rapamycin analog, for 15–60 min to drive the protein population (Perroy et al., 2004; Raynaud et al., 2013). Brieﬂy, co- complex onto microtubules. Control (not treated with AP21967) expression of mGlu1a-Rluc8 and YFP-SNAP23 or YFP-Kif5 in neuro- and treated cells were then ﬁxed and processed for microscopy. blastoma cells was monitored using a spectrophotometric plate reader (MITHRAS LB 940, Berthold Technologies). BRET signal Microscopy was plotted as a function of total ﬂuorescence over total lumines- Fixed cell images were acquired on a Zeiss Axiovert 200 TV cence signal, and the ratio used as an index of the concentration inverted microscope equipped with a software-driven ﬁlter of YFP-tagged over Rluc8-tagged proteins. Co-expression of wheel, a 63× oil immersion objective [1.3 numerical aperture increasing concentrations of YFP-tagged proteins with a con- (NA); Zeiss], or a 25× oil immersion objective (0.8 NA; Zeiss). A stant level of Rluc8-tagged proteins gives rise to a hyperbolical MicroMax 1300 charge-device camera (Princeton Instruments) increase and saturation in BRET signal, only if the two proteins driven by MetaMorph imaging software (version 4.17; Universal speciﬁcally interact. Random collisions between tagged proteins Imaging Corporation) was used for cell imaging. Filter sets gave a linear and non-saturating signal (bystander BRET), indic- (Zeiss) for YFP-tagged proteins were 500 ± 20 and 535 ± 30 nm ating non-speciﬁc protein–protein interactions. band pass ﬁlters for excitation and emission, respectively. For mcherry-tagged proteins, we used ﬁlters (Omega Optical) with a Fluorescence recovery after cleavage 555 ± 50 nm band pass and a 625 ± 50 nm for excitation The mGlu1a receptor was tagged on its extracellular N-terminal and emission, respectively. Z series of images were acquired (31 domain with a thrombin cleavable tag fused to YFP protein. The planes) from −3 μmto +3 μm to image cells. An equatorial plane recombinant protein was transfected in C6 glioma cell line, and was selected from each series. 24 h later cells were washed twice with DMEM medium (ref Imaging of live cells was performed on the same Zeiss setup. 61965, Thermo Fisher Scientiﬁc), supplemented with 10mM A homemade temperature-regulated chamber was used at 34°C HEPES and 1 mg/ml chicken egg albumin (Sigma). Cells were to slightly reduce vesicle motion. For time-lapse imaging, expos- then incubated at 20°C in the same medium to block post-Golgi ure times were 300–500 ms for YFP-tagged proteins, 0.5–1 sec for trafﬁc, followed by thrombin treatment (5 U/ml, Calbiochem) mcherry-tagged proteins. Images from the MicroMax camera were (Rosenberg et al., 2001) to cleave the external N-terminal YFP tag converted to TIFF ﬁles that were edited using ImageJ 1.47 (NIH of cell surface mGlu1a receptors. After extensive washing, cells Image) and compiled into QuickTime (Apple Computers) movies. were returned to their initial medium at 37°C for cell surface receptor recovery. Cells were then washed and ﬁxed at different Supplementary material time periods and quantiﬁcation of cell surface receptor was deter- Supplementary material is available at Journal of Molecular mined using rabbit polyclonal anti-GFP antibodies (Invitrogen). For Cell Biology online. quantitative measurements, coverslips were transferred in 24-well microplates and processed for in-cell western assay (Odyssey, Acknowledgements LI-COR Biosciences). Brieﬂy, cells were blocked with gelatin, Mass spectrometry experiments were carried out using facil- washed, incubated with anti-GFP antibodies, washed again and ities of the Functional Proteomics Platform of Montpellier and ﬁnally incubated with IRDye800CW-conjugated donkey anti-rabbit BRET experiments using facilities of Arpege pharmacological polyconal antibodies (Rockland). Infrared ﬂuorescence was deter- screening platform, at the Institute of Functional Genomics. mined using the 700 nm channel of the Odyssey Infrared Imaging System (LI-COR Biosciences). Fluorescence intensity was measured Funding using Image J software from circular gates placed in the middle of This work was supported by the European Research Council (ERC) the wells. Background values measured in wells containing mock- under the European Union’s Horizon 2020 research and innovation transfected cells were subtracted from the signal. All measure- programme (grant agreement no. 646788 to J.P.), the Agence ments were performed in triplicate and expressed as mean ± SEM. NationaledelaRecherche (ANR-13-JSV4-0005-01 to J.P.), and the Re´gion Languedoc-Roussillon (Chercheur d’Avenir 146090 to J.P.). Rapalog-mediated protein heterodimerization Conventional kinesin1 (Kif5) rigor mutant tightly bound to Conﬂict of interest: none declared. microtubules in C6 glioma cells (data not shown) has been pre- viously described for mouse ﬁbroblast L cells (Nakata and References Hirokawa, 1995). Presently, we replaced the tail of the protein, Aiba, A., Chen, C., Herrup, K., et al. (1994a). Reduced hippocampal long-term which binds the cargo, by FKBP-tagged or not with mcherry potentiation and context-speciﬁc deﬁcit in associative learning in mGluR1 (Kif5 cargoless/FKBP and Kif5 cargoless/mcherry-FKBP). The mutant mice. Cell 79, 365–375. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 435 Aiba, A., Kano, M., Chen, C., et al. (1994b). Deﬁcient cerebellar long-term Kitano, J., Kimura, K., Yamazaki, Y., et al. (2002). Tamalin, a PDZ domain- depression and impaired motor learning in mGluR1 mutant mice. Cell 79, containing protein, links a protein complex formation of group 1 metabo- 377–388. tropic glutamate receptors and the guanine nucleotide exchange factor Ango, F., Pin, J.P., Tu, J.C., et al. (2000). Dendritic and axonal targeting of cytohesins. J. Neurosci. 22, 1280–1289. type 5 metabotropic glutamate receptor is regulated by homer1 proteins Ledonne, A., Nobili, A., Latagliata, E.C., et al. (2015). Neuregulin 1 signalling and neuronal excitation. J. Neurosci. 20, 8710–8716. modulates mGluR1 function in mesencephalic dopaminergic neurons. Mol. Bindea, G., Mlecnik, B., Hackl, H., et al. (2009). ClueGO: a Cytoscape plug-in Psychiatry 20, 959–973. to decipher functionally grouped gene ontology and pathway annotation Lin, T.B., Lai, C.Y., Hsieh, M.C., et al. (2015). VPS26A-SNX27 interaction- networks. Bioinformatics 25, 1091–1093. dependent mGluR5 recycling in dorsal horn neurons mediates neuropathic Brabet, P., Pantaloni, C., Bockaert, J., et al. (1991). Metabolism of two Go pain in rats. J. Neurosci. 35, 14943–14955. alpha isoforms in neuronal cells during differentiation. J. Biol. Chem. 266, Lin, R.C., and Scheller, R.H. (2000). Mechanisms of synaptic vesicle exocyt- 12825–12828. osis. Annu. Rev. Cell Dev. Biol. 16, 19–49. Castillo, C.A., Leon, D.A., Ballesteros-Yanez, I., et al. (2010). Glutamate differ- Luis Albasanz, J., Fernandez, M., and Martin, M. (2002). Internalization of ently modulates metabotropic glutamate receptors in neuronal and glial metabotropic glutamate receptor in C6 cells through clathrin-coated vesi- cells. Neurochem. Res. 35, 1050–1063. cles. Brain Res. Mol. Brain Res. 99, 54–66. Chan, W.Y., Soloviev, M.M., Ciruela, F., et al. (2001). Molecular determinants Luscher, C., and Huber, K.M. (2010). Group 1 mGluR-dependent synaptic of metabotropic glutamate receptor 1B trafﬁcking. Mol. Cell. Neurosci. 17, long-term depression: mechanisms and implications for circuitry and dis- 577–588. ease. Neuron 65, 445–459. Choi, K.Y., Chung, S., and Roche, K.W. (2011). Differential binding of cal- Moutin, E., Compan, V., Raynaud, F., et al. (2014). The stoichiometry of scaf- modulin to group I metabotropic glutamate receptors regulates receptor fold complexes in living neurons—DLC2 functions as a dimerization engine trafﬁcking and signaling. J. Neurosci. 31, 5921–5930. for GKAP. J. Cell Sci. 127, 3451–3462. Cox, J., and Mann, M. (2008). MaxQuant enables high peptide identiﬁcation Nakata, T., and Hirokawa, N. (1995). Point mutation of adenosine rates, individualized p.p.b.-range mass accuracies and proteome-wide pro- triphosphate-binding motif generated rigor kinesin that selectively blocks tein quantiﬁcation. Nat. Biotechnol. 26, 1367–1372. anterograde lysosome membrane transport. J. Cell Biol. 131, 1039–1053. Cox, J., Neuhauser, N., Michalski, A., et al. (2011). Andromeda: a peptide Noel, J., Ralph, G.S., Pickard, L., et al. (1999). Surface expression of AMPA search engine integrated into the MaxQuant environment. J. Proteome receptors in hippocampal neurons is regulated by an NSF-dependent Res. 10, 1794–1805. mechanism. Neuron 23, 365–376. Das, S.S., and Banker, G.A. (2006). The role of protein interaction motifs in Olsen, J.V., de Godoy, L.M., Li, G., et al. (2005). Parts per million mass accur- regulating the polarity and clustering of the metabotropic glutamate acy on an Orbitrap mass spectrometer via lock mass injection into a receptor mGluR1a. J. Neurosci. 26, 8115–8125. C-trap. Mol. Cell. Proteomics 4, 2010–2021. Diefenbach, R.J., Diefenbach, E., Douglas, M.W., et al. (2002). The heavy Orchard, S., Ammari, M., Aranda, B., et al. (2014). The MIntAct project— chain of conventional kinesin interacts with the SNARE proteins SNAP25 IntAct as a common curation platform for 11 molecular interaction data- and SNAP23. Biochemistry 41, 14906–14915. bases. Nucleic Acids Res. 42,D358–D363. Duc, C., and Catsicas, S. (1995). Ultrastructural localization of SNAP-25 Osten, P., Srivastava, S., Inman, G.J., et al. (1998). The AMPA receptor GluR2 within the rat spinal cord and peripheral nervous system. J. Comp. Neurol. C terminus can mediate a reversible, ATP-dependent interaction with NSF 356, 152–163. and alpha- and beta-SNAPs. Neuron 21, 99–110. Ferraguti, F., Crepaldi, L., and Nicoletti, F. (2008). Metabotropic glutamate 1 Oyler, G.A., Higgins, G.A., Hart, R.A., et al. (1989). The identiﬁcation of a receptor: current concepts and perspectives. Pharmacol. Rev. 60, novel synaptosomal-associated protein, SNAP-25, differentially expressed 536–581. by neuronal subpopulations. J. Cell Biol. 109, 3039–3052. Giau, R., Carrette, J., Bockaert, J., et al. (2005). Constitutive secretion of pro- Perroy, J., Pontier, S., Charest, P.G., et al. (2004). Real-time monitoring of tease nexin-1 by glial cells and its regulation by G-protein-coupled recep- ubiquitination in living cells by BRET. Nat. Methods 1, 203–208. tors. J. Neurosci. 25, 8995–9004. Ravichandran, V., Chawla, A., and Roche, P.A. (1996). Identiﬁcation of a novel Guillaud, L., Setou, M., and Hirokawa, N. (2003). KIF17 dynamics and regula- syntaxin- and synaptobrevin/VAMP-binding protein, SNAP-23, expressed in tion of NR2B trafﬁcking in hippocampal neurons. J. Neurosci. 23, non-neuronal tissues. J. Biol. Chem. 271, 13300–13303. 131–140. Raynaud, F., Janossy, A., Dahl, J., et al. (2013). Shank3-Rich2 interaction reg- Hannan, A.J., Blakemore, C., Katsnelson, A., et al. (2001). PLC-beta1, acti- ulates AMPA receptor recycling and synaptic long-term potentiation. vated via mGluRs, mediates activity-dependent differentiation in cerebral J. Neurosci. 33, 9699–9715. cortex. Nat. Neurosci. 4, 282–288. Rosenberg, M., Meier, J., Triller, A., et al. (2001). Dynamics of glycine Hirokawa, N., Niwa, S., and Tanaka, Y. (2010). Molecular motors in neurons: receptor insertion in the neuronal plasma membrane. J. Neurosci. 21, transport mechanisms and roles in brain function, development, and dis- 5036–5044. ease. Neuron 68, 610–638. Setou, M., Nakagawa, T., Seog, D.H., et al. (2000). Kinesin superfamily motor Hoerndli, F.J., Maxﬁeld, D.A., Brockie, P.J., et al. (2013). Kinesin-1 regulates protein KIF17 and mLin-10 in NMDA receptor-containing vesicle transport. synaptic strength by mediating the delivery, removal, and redistribution of Science 288, 1796–1802. AMPA receptors. Neuron 80, 1421–1437. Shannon, P., Markiel, A., Ozier, O., et al. (2003). Cytoscape: a software envir- Ichise, T., Kano, M., Hashimoto, K., et al. (2000). mGluR1 in cerebellar onment for integrated models of biomolecular interaction networks. Purkinje cells essential for long-term depression, synapse elimination, and Genome Res. 13, 2498–2504. motor coordination. Science 288, 1832–1835. Suh, Y.H., Terashima, A., Petralia, R.S., et al. (2010). A neuronal role for Jahn, R., and Scheller, R.H. (2006). SNAREs—engines for membrane fusion. SNAP-23 in postsynaptic glutamate receptor trafﬁcking. Nat. Neurosci. 13, Nat. Rev. Mol. Cell Biol. 7, 631–643. 338–343. Jenkins, B., Decker, H., Bentley, M., et al. (2012). A novel split kinesin assay Tao-Cheng, J.H., Du, J., and McBain, C.J. (2000). Snap-25 is polarized to axons identiﬁes motor proteins that interact with distinct vesicle populations. and abundant along the axolemma: an immunogold study of intact neu- J. Cell Biol. 198, 749–761. rons. J. Neurocytol. 29, 67–77. Kenny, P.J., and Markou, A. (2004). The ups and downs of addiction: role Thouvenot, E., Urbach, S., Dantec, C., et al. (2008). Enhanced detection of of metabotropic glutamate receptors. Trends Pharmacol. Sci. 25, CNS cell secretome in plasma protein-depleted cerebrospinal ﬂuid. 265–272. J. Proteome Res. 7, 4409–4421. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 436 j Raynaud et al. Viwatpinyo, K., and Chongthammakun, S. (2009). Activation of group I meta- Washbourne, P., Thompson, P.M., Carta, M., et al. (2002b). Genetic ablation botropic glutamate receptors leads to brain-derived neurotrophic factor of the t-SNARE SNAP-25 distinguishes mechanisms of neuroexocytosis. expression in rat C6 cells. Neurosci. Lett. 467, 127–130. Nat. Neurosci. 5, 19–26. Washbourne, P., Bennett, J.E., and McAllister, A.K. (2002a). Rapid recruit- Wu, H., Nash, J.E., Zamorano, P., et al. (2002). Interaction of SAP97 with ment of NMDA receptor transport packets to nascent synapses. Nat. minus-end-directed actin motor myosin VI. Implications for AMPA receptor Neurosci. 5, 751–759. trafﬁcking. J. Biol. Chem. 277, 30928–30934. Washbourne, P., Liu, X.B., Jones, E.G., et al. (2004). Cycling of NMDA recep- Zhou, L., Yang, D., Wang, D.J., et al. (2015). Numb deﬁciency in cerebellar tors during trafﬁcking in neurons before synapse formation. J. Neurosci. Purkinje cells impairs synaptic expression of metabotropic glutamate recep- 24, 8253–8264. tor and motor coordination. Proc. Natl Acad. Sci. USA 112, 15474–15479. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Molecular Cell Biology Oxford University Press http://www.deepdyve.com/lp/oxford-university-press/snap23-kif5-complex-controls-mglu1-receptor-trafficking-pMwmFA7sFR

Loading next page...

References (108)

C. Lüscher, K. Huber (2010)
Group 1 mGluR-Dependent Synaptic Long-Term Depression: Mechanisms and Implications for Circuitry and Disease
Neuron, 65
Kyu-Yeong Choi, Seungsoo Chung, K. Roche (2011)
Differential Binding of Calmodulin to Group I Metabotropic Glutamate Receptors Regulates Receptor Trafficking and Signaling
The Journal of Neuroscience, 31
R. Jahn, R. Scheller (2006)
SNAREs — engines for membrane fusion
Nature Reviews Molecular Cell Biology, 7
Mitsutoshi Setou, T. Nakagawa, D. Seog, N. Hirokawa (2000)
Kinesin superfamily motor protein KIF17 and mLin-10 in NMDA receptor-containing vesicle transport.
Science, 288 5472
Noel (1999)
365
Neuron, 23
Hannan (2001)
282
Nat. Neurosci., 4
Guillaud (2003)
131
J. Neurosci., 23
Hongju Wu, J. Nash, P. Zamorano, C. Garner (2002)
Interaction of SAP97 with Minus-end-directed Actin Motor Myosin VI
The Journal of Biological Chemistry, 277
Ango (2000)
8710
J. Neurosci., 20
Ravichandran (1996)
13300
J. Biol. Chem., 271
R. Diefenbach, E. Diefenbach, M. Douglas, A. Cunningham (2002)
The heavy chain of conventional kinesin interacts with the SNARE proteins SNAP25 and SNAP23.
Biochemistry, 41 50
Lin (2015)
14943
J. Neurosci., 35
P. Washbourne, Xiao-Bo Liu, E. Jones, A. McAllister (2004)
Cycling of NMDA Receptors during Trafficking in Neurons before Synapse Formation
The Journal of Neuroscience, 24
Jenkins (2012)
749
J. Cell Biol., 198
C. Duc, S. Catsicas (1995)
Ultrastructural localization of SNAP‐25 within the rat spinal cord and peripheral nervous system
Journal of Comparative Neurology, 356
Hirokawa (2010)
610
Neuron, 68
J. Cox, M. Mann (2008)
MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification
Nature Biotechnology, 26
J. Noel, G. Ralph, L. Pickard, Jackie Williams, E. Molnár, J. Uney, G. Collingridge, J. Henley (1999)
Surface Expression of AMPA Receptors in Hippocampal Neurons Is Regulated by an NSF-Dependent Mechanism
Neuron, 23
Brian Jenkins, H. Decker, Marvin Bentley, Julie Luisi, G. Banker (2012)
A novel split kinesin assay identifies motor proteins that interact with distinct vesicle populations
The Journal of Cell Biology, 198
E. Thouvenot, S. Urbach, C. Dantec, J. Poncet, M. Séveno, E. Demettre, P. Jouin, J. Touchon, J. Bockaert, P. Marin (2008)
Enhanced detection of CNS cell secretome in plasma protein-depleted cerebrospinal fluid.
Journal of proteome research, 7 10
F. Ango, J. Pin, J. Tu, B. Xiao, P. Worley, J. Bockaert, L. Fagni (2000)
Dendritic and Axonal Targeting of Type 5 Metabotropic Glutamate Receptor Is Regulated by Homer1 Proteins and Neuronal Excitation
The Journal of Neuroscience, 20
Viwatpinyo (2009)
127
Neurosci. Lett., 467
Lin (2000)
19
Annu. Rev. Cell Dev. Biol., 16
F. Raynaud, A. Janossy, Jan-Ulrik Dahl, F. Bertaso, J. Perroy, A. Varrault, M. Vidal, P. Worley, T. Boeckers, J. Bockaert, P. Marin, L. Fagni, V. Homburger (2013)
Shank3-Rich2 Interaction Regulates AMPA Receptor Recycling and Synaptic Long-Term Potentiation
The Journal of Neuroscience, 33
J. Olsen, Lyris Godoy, Guoqing Li, B. Maček, P. Mortensen, R. Pesch, A. Makarov, Oliver Lange, S. Horning, M. Mann (2005)
Parts per Million Mass Accuracy on an Orbitrap Mass Spectrometer via Lock Mass Injection into a C-trap*S
Molecular & Cellular Proteomics, 4
José Albasanz, M. Fernández, Mairena Martín (2002)
Internalization of metabotropic glutamate receptor in C6 cells through clathrin-coated vesicles.
Brain research. Molecular brain research, 99 1
Richard Lin, R. Scheller (2000)
Mechanisms of synaptic vesicle exocytosis.
Annual review of cell and developmental biology, 16
(2020)
differential binding
Catalysis from A to Z
Castillo (2010)
1050
Neurochem. Res., 35
Raynaud (2013)
9699
J. Neurosci., 33
Ichise (2000)
1832
Science, 288
Olive Lloyd-Baker (1964)
IDENTIFICATION OF NOVEL
Notes and Queries, 11
A. Ledonne, A. Nobili, E. Latagliata, V. Cavallucci, E. Guatteo, S. Puglisi-Allegra, M. D’Amelio, N. Mercuri (2014)
Neuregulin 1 signalling modulates mGluR1 function in mesencephalic dopaminergic neurons
Molecular Psychiatry, 20
Nakata (1995)
1039
J. Cell Biol., 131
Oyler (1989)
3039
J. Cell Biol., 109
Hoerndli (2013)
1421
Neuron, 80
Washbourne (2004)
8253
J. Neurosci., 24
Luscher (2010)
445
Neuron, 65
Jahn (2006)
631
Nat. Rev. Mol. Cell Biol., 7
Das (2006)
8115
J. Neurosci., 26
S. Orchard, M. Ammari, Bruno Aranda, L. Breuza, Leonardo Briganti, F. Broackes-Carter, N. Campbell, G. Chavali, Carol Chen, N. del-Toro, M. Duesbury, M. Dumousseau, Eugenia Galeota, U. Hinz, M. Iannuccelli, Sruthi Jagannathan, R. Jiménez, J. Khadake, A. Lægreid, L. Licata, R. Lovering, B. Meldal, A. Melidoni, Miladys Milagros, Daniele Peluso, L. Perfetto, P. Porras, A. Raghunath, S. Ricard-Blum, B. Roechert, A. Stutz, M. Tognolli, K. Roey, G. Cesareni, H. Hermjakob (2013)
The MIntAct project—IntAct as a common curation platform for 11 molecular interaction databases
Nucleic Acids Research, 42
Ronald Giau, J. Carrette, J. Bockaert, V. Homburger (2005)
Constitutive Secretion of Protease Nexin-1 by Glial Cells and Its Regulation by G-Protein-Coupled Receptors
The Journal of Neuroscience, 25
Chan (2001)
577
Mol. Cell. Neurosci., 17
Taeko Ichise, M. Kano, K. Hashimoto, Dai Yanagihara, Kazuki Nakao, R. Shigemoto, Motoya Katsuki, A. Aiba (2000)
mGluR1 in cerebellar Purkinje cells essential for long-term depression, synapse elimination, and motor coordination.
Science, 288 5472
Laurent Guillaud, M. Setou, N. Hirokawa (2003)
KIF17 Dynamics and Regulation of NR2B Trafficking in Hippocampal Neurons
The Journal of Neuroscience, 23
Osten (1998)
99
Neuron, 21
Aiba (1994)
365
Cell, 79
Diefenbach (2002)
14906
Biochemistry, 41
Kittikun Viwatpinyo, S. Chongthammakun (2009)
Activation of group I metabotropic glutamate receptors leads to brain-derived neurotrophic factor expression in rat C6 cells
Neuroscience Letters, 467
Moutin (2014)
3451
J. Cell Sci., 127
N. Hirokawa, Shinsuke Niwa, Yosuke Tanaka (2010)
Molecular Motors in Neurons: Transport Mechanisms and Roles in Brain Function, Development, and Disease
Neuron, 68
Ledonne (2015)
959
Mol. Psychiatry, 20
G. Bindea, B. Mlecnik, H. Hackl, P. Charoentong, M. Tosolini, A. Kirilovsky, W. Fridman, F. Pagès, Z. Trajanoski, J. Galon (2009)
ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks
Bioinformatics, 25
J. Cox, Nadin Neuhauser, Annette Michalski, R. Scheltema, J. Olsen, M. Mann (2011)
Andromeda: a peptide search engine integrated into the MaxQuant environment.
Journal of proteome research, 10 4
W. Chan, M. Soloviev, F. Ciruela, R. Mcilhinney (2001)
Molecular Determinants of Metabotropic Glutamate Receptor 1B Trafficking
Molecular and Cellular Neuroscience, 17
Sonal Das, G. Banker (2006)
The Role of Protein Interaction Motifs in Regulating the Polarity and Clustering of the Metabotropic Glutamate Receptor mGluR1a
The Journal of Neuroscience, 26
F. Ferraguti, L. Crepaldi, F. Nicoletti (2008)
Metabotropic Glutamate 1 Receptor: Current Concepts and Perspectives
Pharmacological Reviews, 60
Ferraguti (2008)
536
Pharmacol. Rev., 60
(2002)
Interaction of SAP97 with minus-end-directed actin motor myosin VI. Implications for AMPA receptor trafficking
J. Biol. Chem., 277
Madelaine Rosenberg, J. Meier, A. Triller, C. Vannier (2001)
Dynamics of Glycine Receptor Insertion in the Neuronal Plasma Membrane
The Journal of Neuroscience, 21
Suh (2010)
338
Nat. Neurosci., 13
P. Washbourne, J. Bennett, A. McAllister (2002)
Rapid recruitment of NMDA receptor transport packets to nascent synapses
Nature Neuroscience, 5
Atsu Alba, M. Kano, Chong Chen, M. Stanton, G. Fox, K. Herrup, T. Zwingman, S. Tonegawa (1994)
Deficient cerebellar long-term depression and impaired motor learning in mGluR1 mutant mice
Cell, 79
Kitano (2002)
1280
J. Neurosci., 22
Duc (1995)
152
J. Comp. Neurol., 356
Orchard (2014)
D358
Nucleic Acids Res., 42
P. Osten, S. Srivastava, G. Inman, F. Vilim, L. Khatri, L. Lee, B. States, S. Einheber, T. Milner, P. Hanson, E. Ziff (1998)
The AMPA receptor GluR2 C terminus can mediate a reversible, ATP-dependent interaction with NSF and alpha- and beta-SNAPs.
Neuron, 21 1
C. Castillo, D. León, I. Ballesteros-Yáñez, I. Iglesias, Mairena Martín, J. Albasanz (2010)
Glutamate Differently Modulates Metabotropic Glutamate Receptors in Neuronal and Glial Cells
Neurochemical Research, 35
P. Brabet, C. Pantaloni, J. Bockaert, V. Homburger (1991)
Metabolism of two Go alpha isoforms in neuronal cells during differentiation.
The Journal of biological chemistry, 266 20
Aiba (1994)
377
Cell, 79
Young Suh, A. Terashima, R. Petralia, R. Wenthold, J. Isaac, K. Roche, P. Roche (2010)
A neuronal role for SNAP-23 in postsynaptic glutamate receptor trafficking
Nature neuroscience, 13
V. Ravichandran, Ashish Chawla, P. Roche (1996)
Identification of a Novel Syntaxin- and Synaptobrevin/VAMP-binding Protein, SNAP-23, Expressed in Non-neuronal Tissues*
The Journal of Biological Chemistry, 271
Perroy (2004)
203
Nat. Methods, 1
Choi (2011)
5921
J. Neurosci., 31
Jun Kitano, K. Kimura, Y. Yamazaki, T. Soda, R. Shigemoto, Y. Nakajima, S. Nakanishi (2002)
Tamalin, a PDZ Domain-Containing Protein, Links a Protein Complex Formation of Group 1 Metabotropic Glutamate Receptors and the Guanine Nucleotide Exchange Factor Cytohesins
The Journal of Neuroscience, 22
Kenny (2004)
265
Trends Pharmacol. Sci., 25
P. Washbourne, Peter Thompson, Peter Thompson, M. Carta, E. Costa, E. Costa, J. Mathews, G. López-Bendito, Z. Molnár, M. Becher, C. Valenzuela, L. Partridge, Michael Wilson (2002)
Genetic ablation of the t-SNARE SNAP-25 distinguishes mechanisms of neuroexocytosis
Nature Neuroscience, 5
Wu (2002)
30928
J. Biol. Chem., 277
P. Kenny, A. Markou (2004)
The ups and downs of addiction: role of metabotropic glutamate receptors.
Trends in pharmacological sciences, 25 5
(2015)
Numb deficiency in cerebellar Purkinje cells impairs synaptic expression of metabotropic glutamate receptor and motor coordination
Proc. Natl Acad. Sci. USA, 112
Cox (2011)
1794
J. Proteome Res., 10
Tzer-Bin Lin, C. Lai, M. Hsieh, Hsueh-Hsiao Wang, Jen-Kun Cheng, Y. Chau, Gin-Den Chen, H. Peng (2015)
VPS26A–SNX27 Interaction-Dependent mGluR5 Recycling in Dorsal Horn Neurons Mediates Neuropathic Pain in Rats
The Journal of Neuroscience, 35
Giau (2005)
8995
J. Neurosci., 25
Bindea (2009)
1091
Bioinformatics, 25
Washbourne (2002)
19
Nat. Neurosci., 5
Liang Zhou, Dong Yang, De-Juan Wang, Ya-Jun Xie, Jia-Huan Zhou, Lin Zhou, Hao Huang, Shuo Han, Chong‐Yu Shao, Hua-shun Li, J. Zhu, M. Qiu, C. Zeeuw, Ying Shen (2015)
Numb deficiency in cerebellar Purkinje cells impairs synaptic expression of metabotropic glutamate receptor and motor coordination
Proceedings of the National Academy of Sciences, 112
Paul Shannon, Andrew Markiel, Owen Ozier, Nitin Baliga, Jonathan Wang, Daniel Ramage, Nada Amin, Benno Schwikowski, T. Ideker (2003)
Cytoscape: a software environment for integrated models of biomolecular interaction networks.
Genome research, 13 11
Tao-Cheng (2000)
67
J. Neurocytol., 29
J. Perroy, S. Pontier, Pascale Charest, M. Aubry, M. Bouvier (2004)
Real-time monitoring of ubiquitination in living cells by BRET
Nature Methods, 1
Setou (2000)
1796
Science, 288
J. Tao-Cheng, Jing Du, C. McBain (2000)
Snap-25 is polarized to axons and abundant along the axolemma: an immunogold study of intact neurons
Journal of Neurocytology, 29
Rosenberg (2001)
5036
J. Neurosci., 21
Washbourne (2002)
751
Nat. Neurosci., 5
Cox (2008)
1367
Nat. Biotechnol., 26
F. Hoerndli, Dane Maxfield, P. Brockie, J. Mellem, E. Jensen, Rui Wang, D. Madsen, A. Maricq (2013)
Kinesin-1 Regulates Synaptic Strength by Mediating the Delivery, Removal, and Redistribution of AMPA Receptors
Neuron, 80
P. Osten, S. Srivastava, G. Inman, F. Vilim, L. Khatri, L. Lee, B. States, S. Einheber, T. Milner, P. Hanson, E. Ziff (1998)
The AMPA Receptor GluR2 C Terminus Can Mediate a Reversible, ATP-Dependent Interaction with NSF and α- and β-SNAPs
Neuron, 21
C. Kleinewächter (1999)
On Identification
T. Nakata, N. Hirokawa (1995)
Point mutation of adenosine triphosphate-binding motif generated rigor kinesin that selectively blocks anterograde lysosome membrane transport
The Journal of Cell Biology, 131
Olsen (2005)
2010
Mol. Cell. Proteomics, 4
A. Aiba, Chong Chen, K. Herrup, Christian Rosenmund, S. Tonegawa (1994)
Reduced hippocampal long-term potentiation and context-specific deficit in associative learning in mGluR1 mutant mice
Cell, 79
G. Oyler, G. Higgins, R. Hart, E. Battenberg, M. Billingsley, F. Bloom, Michael Wilson (1989)
The identification of a novel synaptosomal-associated protein, SNAP-25, differentially expressed by neuronal subpopulations
The Journal of Cell Biology, 109
Zhou (2015)
15474
Proc. Natl Acad. Sci. USA, 112
Luis Albasanz (2002)
54
Brain Res. Mol. Brain Res., 99
A. Hannan, C. Blakemore, A. Katsnelson, T. Vitalis, K. Huber, M. Bear, J. Roder, Daesoo Kim, Hee-Sup Shin, P. Kind (2001)
PLC-β1, activated via mGluRs, mediates activity-dependent differentiation in cerebral cortex
Nature Neuroscience, 4
Brabet (1991)
12825
J. Biol. Chem., 266
Thouvenot (2008)
4409
J. Proteome Res., 7
Enora Moutin, V. Compan, F. Raynaud, C. Clerté, Nathalie Bouquier, G. Labesse, M. Ferguson, L. Fagni, C. Royer, J. Perroy (2014)
The stoichiometry of scaffold complexes in living neurons – DLC2 functions as a dimerization engine for GKAP
Journal of Cell Science, 127
Shannon (2003)
2498
Genome Res., 13

Publisher: Oxford University Press
Copyright: © The Author(s) (2018). Published by Oxford University Press on behalf of Journal of Molecular Cell Biology, IBCB, SIBS, CAS. All rights reserved.
ISSN: 1674-2788
eISSN: 1759-4685
DOI: 10.1093/jmcb/mjy031
Publisher site: See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 doi:10.1093/jmcb/mjy031 Journal of Molecular Cell Biology (2018), 10(5), 423–436 j 423 Published online September 14, 2018 Article SNAP23–Kif5 complex controls mGlu1 receptor trafﬁcking 1,3 1 2 1 1 Fabrice Raynaud , Vincent Homburger , Martial Seveno , Oana Vigy , Enora Moutin , 1 1, Laurent Fagni , and Julie Perroy IGF, CNRS, INSERM, Univ. Montpellier, F-34094 Montpellier, France BioCampus Montpellier, CNRS, INSERM, Univ. Montpellier, F-34094 Montpellier, France Present address: INSERM U1046 PhyMedExp, CNRS UMR9214, Univ. Montpellier, F-34094 Montpellier, France * Correspondence to: Julie Perroy, E-mail: Julie.Perroy@igf.cnrs.fr Edited by Xuebiao Yao Metabotropic glutamate receptors are expressed at excitatory synapses and control synaptic transmission in mammalian brain. These receptors are involved in numerous patho-physiological functions. However, little is known about the molecular determi- nants responsible for their intracellular transport and membrane targeting. Here we investigated the nature of the molecular motor and adaptor protein responsible for trafﬁcking and membrane localization of the group I metabotropic glutamate mGlu1 postsynaptic receptor in cultured hippocampal neurons. In proteomic studies, we identiﬁed the synaptosome-associated protein 23 (SNAP23) and the molecular motor Kif5 kinesin as proteins interacting with mGlu1 receptor. We showed that SNAP23, but not Kif5, directly interacts with mGlu1 receptor carboxyl terminus. Using a recombination approach to impair or enhance the inter- action between SNAP23 and Kif5, we found that the SNAP23–Kif5 complex controls the trafﬁcking of mGlu1 receptor along micro- tubules. Additional ﬂuorescence recovery after cleavage experiments allowed us to identify a role of the complex in the receptor cell surface targeting. In conclusion, our study indicates that along dendritic processes Kif5–SNAP23 complex contributes to prop- er mGlu1 receptor trafﬁcking and cell surface expression. Keywords: mGlu1, SNARE, kinesin, neuron, trafﬁc have been of intense interest in the last 10 years, but relatively Introduction little is known about the mechanisms controlling the trafﬁcking of Metabotropic glutamate (mGlu1–mGlu8) receptors are mGlu receptors. G-protein-coupled receptors that are located at brain excitatory Several proteins have been shown to control the functional synapses to regulate glutamatergic neurotransmission. The expression of mGlu1/5 receptors at postsynaptic terminals. mGlu1 and mGlu5 receptor subtypes activate the extracellular These are postsynaptic intracellular scaffolding proteins asso- signal-regulated kinase (ERK)/mitogen-activated protein kinase ciated with cytoskeleton, plasma membrane, or fusion proteins. (MAPK) pathway, as well as phospholipase C, which leads to Thus, Homer proteins (Ango et al., 2000), Tamalin (Kitano et al., intracellular calcium release and protein kinase C activation. 2002), Calmodulin (Choi et al., 2011), Neuregulin (Ledonne Regulation of functional expression of mGlu1/5 receptors plays a et al., 2015), VPS26A-SNX27 complex (Lin et al., 2015), and critical role in brain functions and neurological diseases. For Numb (Zhou et al., 2015) bind to mGlu1 and/or mGlu5 receptors instance, mGlu1 receptor is critically involved in neuronal devel- and control their organization in the postsynaptic element and opment (Ichise et al., 2000; Hannan et al., 2001), synaptic plasti- recycling at the synaptic cell surface. This is critical for synaptic city and associated learning and memory (Aiba et al., 1994a, b; plasticity and related cognitive and motor functions. However Luscher and Huber, 2010), addiction to drugs of abuse (Kenny nothing is known about carrier proteins involved in the long- and Markou, 2004), and several neuropsychiatric and neurode- range delivery of mGlu receptors before their insertion into the generative disorders (Ferraguti et al., 2008). Molecular mechan- postsynaptic membrane. isms underlying vesicle trafﬁcking of neurotransmitter receptors Soluble N-ethylmaleimide-sensitive-factor Attachment protein REceptor (SNARE) proteins are a class of membrane-associated Received December 8, 2017. Revised April 9, 2018. Accepted May 11, 2018. proteins known to regulate the process of synaptic vesicle © The Author(s) (2018). Published by Oxford University Press on behalf of Journal of Molecular Cell Biology, IBCB, SIBS, CAS. All rights reserved. fusion with the presynaptic plasma membrane (Lin and Scheller, Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 424 j Raynaud et al. 2000; Jahn and Scheller, 2006). However, these proteins are interaction in the IntAct database (Figure 1). We identiﬁed Kif5b also involved in microtubule-dependent postsynaptic trafﬁcking (kinesin-1 heavy chain), a molecular motor involved in the ves- of glutamate receptors. For instance, SNAP23 is a homolog icular trafﬁcking of AMPA receptors along microtubules SNARE protein of SNAP25 (Oyler et al., 1989; Duc and Catsicas, (Hoerndli et al., 2013), and SNAP23, a SNARE protein involved 1995; Tao-Cheng et al., 2000; Washbourne et al., 2002b) that is in the vesicular transport and exocytosis of various membrane ubiquitously expressed in neuronal and non-neuronal cells proteins (Lin and Scheller, 2000; Jahn and Scheller, 2006). To (Ravichandran et al., 1996). In hippocampal neurons, SNAP23 is address the existence of mGlu1–SNAP23–Kif5b complexes in liv- expressed in both soma and dendrites and regulates the surface ing cells, we used C6 glioma cells, because the ﬂat shape of expression, recycling, and synaptic function of the ionotropic these cells was more convenient for microtubule dynamics stud- N-methyl-D-aspartate (NMDA) glutamate receptor (Washbourne ies. Moreover, these cells endogenously express mGlu1 receptor et al., 2004; Suh et al., 2010). Other studies have shown that (Luis Albasanz et al., 2002; Viwatpinyo and Chongthammakun, NMDA receptors are transported in clusters along microtubules 2009; Castillo et al., 2010) and we show that they also express (Washbourne et al., 2002a) in vesicles containing the kinesin pro- SNAP23 and Kif5 proteins (Figure 2C and D). Immunoﬂuorescent tein Kif17 (Setou et al., 2000; Guillaud et al., 2003). The experiments showed that mcherry-tagged mGlu1a receptor colo- microtubule-dependent motor Kif5 and the associated kinesin calized with YFP-tagged Kif5b or YFP-tagged SNAP23 in vesicles, light chain Klc2 would mediate the transport of the alpha-amino- in C6 glioma cell line (Figure 2A and B). Furthermore, YFP- 3-hydroxy-5-methylisoazol-4-propionate (AMPA) glutamate recep- tagged mGlu1a receptor could be immunoprecipitated with Kif5 tor subunit GLR-1 to the synapse, in Caenorhabditis elegans.This (Figure 2C) and SNAP23 in transfected neuroblastoma cell line Kif5-dependent transport would control rapid delivery, removal (Figure 2D). Altogether these results suggested that mGlu1a and redistribution of functional postsynaptic AMPA receptors, receptor can form a complex with Kif5 and SNAP23 in model and thus synaptic strength (Hoerndli et al., 2013). cell line. At present nothing is known about the transport of mGlu receptors along microtubules. Here we report that mGlu1 recep- Interaction between mGlu1a receptor and Kif5–SNAP23 complex tor can form a cargo with SNAP23 in microtubule-associated involves mGlu1a receptor and SNAP23 carboxyl termini vesicles conveyed by Kif5 along dendrites of cultured hippocam- Kif5 and SNAP23 were known to form a complex (Diefenbach pal neurons. This result provides the ﬁrst evidence for a molecu- et al., 2002), and the above results showed that mGlu1a recep- lar mechanism of vesicular transport of postsynaptic mGlu tor belongs to the Kif5–SNAP23 complex. We therefore investi- receptors in neurons. gated the molecular determinants of the interaction between mGlu1a receptor and Kif5–SNAP23 complex in neuroblastoma Results cell line. Deletion of the C-terminus of mGlu1areceptor(YFP- The Kif5–SNAP23 complex interacts with mGlu1a receptor mGlu1aΔC) strongly decreased co-immunoprecipitation of the In order to identify multiprotein complexes interacting with receptor mutant with both Kif5 and SNAP23 (Figure 2C and D). the intracellular mGlu1a receptor C-terminus, we performed dif- These experiments showed that the C-terminal domain of mGlu1a ferential proteomic studies in an easily transfectable neuron- receptor was required for the interaction between the receptor derived cell line, the N2a neuroblastoma cell line expressing and Kif5–SNAP23 complex. either full-length YFP-mGlu1a or C-terminus-truncated mGlu1a We then investigated whether mGlu1a receptor could directly (YFP-mGlu1aΔC) receptors. The mGlu1 sequences coverage interact with SNAP23 or Kif5. We measured bioluminescence found using mGlu1aΔC as a bait conﬁrmed the absence of resonance energy transfer (BRET) signals between these pro- C-terminus mapping (Supplementary Figure S1). Differential teins tagged with BRET donor and acceptor compatible entities analysis of proteins that were immunoprecipitated with these (mGlu1a-Rluc8 and YFP-SNAP23 or YFP-Kif5). For a constant receptors indicated that 173 proteins interact with the full- expression level of mGlu1a-Rluc8, the BRET signal increased length receptor, but not with YFP-mGlu1aΔC (Supplementary hyperbolically as a function of the YFP-SNAP23 expression level, Table S1). Only proteins with a fold enrichment of six (based on indicating that mGlu1a receptor speciﬁcally interacts with SNAP23 the MSMS ratio) and a minimum of two unique peptides per (Figure 3A). In contrast, only a weak and linear, non-speciﬁc protein were indeed considered as potential interactors of the BRET signal could be detected between mGlu1a and Kif5 pro- carboxyl terminus of the mGlu1 receptor. To determine if these teins (Figure 3B). These results suggested that SNAP23 but not proteins belong to speciﬁc physiological categories, we next Kif5 directly interacts with mGlu1a receptor. This SNAP23 direct performed a gene ontology (GO) analysis based on both bio- interaction could also be detected with mGlu5 but not mGlu7 recep- logical process and cellular component domains. Interestingly, tors (Supplementary Figure S2), suggesting that the interaction numerous proteins are involved in signaling pathways, trans- was speciﬁc to group I mGlu receptors (mGlu1aand mGlu5). port, and neuronal functions. In search for proteins involved in We then searched for the interaction domains responsible for mGlu1 receptor trafﬁcking, we focused on proteins annotated to the mGlu1a receptor–SNAP23 interaction by measuring BRET the ‘intracellular transport’ GO term (GO:0046907). Among the signals between C-terminal deletion mutants of these proteins, 173 potential interactors of mGlu1 C-tail, 29 proteins belong to in transfected neuroblastoma cell line. Successive deletions of this functional category, 9 with known protein–protein the distal part of mGlu1 C-terminus did not affect its BRET Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 425 Figure 1 mGlu1a receptor-associated protein network. Searching for proteins involved in mGlu1 receptor trafﬁcking, the network is focused on the nine proteins displayed in the internal circle that are annotated to the ‘intracellular transport’ GO term and have protein–protein interactions in the IntAct database. It consists of 28 nodes and 22 edges. Node colors reﬂect the MS/MS ratio and vary from red (high ratio) to light yellow (low ratio). Thick node border highlights proteins annotated to the ‘intracellular transport’ GO term. Node shapes represent different protein taxonomies: circle (mouse), diamond (human), square (rat), and hexagon (bovine). The edge width is proportional to the number of published manuscripts—curated in the IntAct database—that have reported molecular interaction data. The edge stroke color maps the conﬁdence value: from 0.6 (light gray) to the local maximum 0.92 (dark gray). signals with wild-type SNAP23 until the receptor lacks all but SNAP23 may bind to Kif5 via its N-terminus (Diefenbach et al., the 21 proximal amino acids of the tail (mGlu1aΔ341, 2002) and serves as a cargo for the transport of mGlu1areceptor Figure 3F). This result showed that interaction with SNAP23 by the molecular motor Kif5. requires the proximal C-terminus of mGlu1a receptor until amino acid position 971 (mGlu1aΔ229, Figure 3E). BRET signals were Kif5 controls microtubule trafﬁcking and membrane targeting of also detected between wild-type mGlu1a receptor and SNAP23 the mGlu1a receptor mutants deleted from the last 82 C-terminal amino acids, but Since we found that mGlu1a receptor is part of a Kif5- were lost after deletion of the last 122 C-terminal amino acids of containing complex and because Kif5 is involved in vesicular the protein (SNAP23Δ122, Figure 3J). These results indicated transport of proteins (Hirokawa et al., 2010), this molecular that interaction of SNAP23 with mGlu1a receptor required the motor could regulate trafﬁcking along microtubules and cell sur- amino acids 88–128 of SNAP23. Altogether these results show face expression of the mGlu1a receptor. We investigated this that mGlu1a receptor directly interacts with SNAP23 and that this hypothesis using rapalog-induced reconstruction of Kif5 protein interaction depends on C-terminal domains of both proteins. (see Jenkins et al., 2012 for the methodology) in C6 glioma cell Thus these results reinforce the above tentative hypothesis that line. Brieﬂy, this approach, based on the reconstruction of a Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 426 j Raynaud et al. Figure 2 The mGlu1a receptor forms a complex with Kif5b and SNAP proteins. (A and B) Co-localization of mcherry-mGlu1a receptor with YFP- Kif5 (A) or YFP-SNAP23 (B) proteins in C6 glioma cells. Following a temperature shift at 20°C to synchronize post-Golgi trafﬁc and thrombin- mediated cleavage of cell surface receptor’s mcherry tag, time-lapse images of intracellular protein trafﬁcking were taken at 34°C as described in Material and methods section. Arrows indicate mcherry-mGlu1a receptor and YFP-Kif5 (A) or YFP-SNAP23 (B) vesicle co-localization. (C and D) Kif5–SNAP23–mGlu1a receptor complex revealed by co-immunoprecipitation experiments. Neuroblastoma cells were transfected with plas- mids encoding wild-type or C-terminus-truncated YFP-mGlu1areceptors and Kif5 (C)orSNAP23 (D) proteins. Cells were washed, solubilized, and subjectedtoimmunoprecipitation(IP)withGFP-trap andrevealedbywestern blots(WB)witheither anti-Kif5 (C)oranti-SNAP23 (D) antibodies. split kinesin by chemical dimerization, enables the identiﬁcation that mGlu1a receptor association with Kif5 complex is essential of motor proteins that interact with speciﬁc vesicle populations. for the receptor plasma membrane targeting. To characterize the trapping of SNAP23 and mGlu1 receptor by We further investigated this hypothesis by measuring the Kif5 on microtubules, we used the rigor mutant of Kif5, which is dominant-negative effects of two Kif5 mutants, Kif5 rigor and stalled on microtubules. Figure 4A shows that mcherry-tagged Kif5 unbound, in C6 glioma cell line. Kif5 rigor is an active form cargoless-Kif5 recombined with its YFP-tagged cargo binding of Kif5 that binds to microtubules and blocks microtubule- domain (CBD) decorates microtubules. The YFP-tagged SNAP23 dependent protein transport. This mutant inhibited cell surface protein colocalized with recombined Kif5 protein on ﬁlamentous expression of YFP-tagged mGlu1a receptor (Figure 5B), suggest- structures (Figure 4B), suggesting the recruitment of SNAP23 by ing that membrane expression of the receptor involves Kif5 on microtubules. Similarly, forced interaction between microtubule-dependent vesicular trafﬁc of proteins. The Kif5 SNAP23 and Kif5 using rapalog (Figure 4C and Supplementary unbound mutant is an inactive form of Kif5 that lacks ability to Movie S1) recruited mcherry-tagged mGlu1a receptor to micro- interact with microtubules and therefore aborts the transport of tubule like structures (Figure 4D). Together, these results sug- Kif5-speciﬁc cargo. This mutant also inhibited membrane expres- gest that mGlu1a receptor can form a complex with SNAP23 and sion of YFP-tagged mGlu1a receptor (Figure 5B), thus suggesting Kif5 on microtubules, and that SNAP23 may serve as cargo for involvement of Kif5 in membrane trafﬁcking of the receptor. We the transport of mGlu1a receptor by the molecular motor Kif5. then investigated the speciﬁcity of the effect of Kif5 on mGlu1a We then investigated whether mGlu1a receptor-Kif5 interaction receptor membrane trafﬁcking by comparing the effects of Kif5 could control membrane targeting of the receptor. The mGlu1a and Kif3 unbound mutants on membrane expression of YFP- receptor was tagged with an extracellular-thrombin cleavable YFP tagged mGlu1a receptor in FRAC experiments. Only Kif5 mutant tag and transfected in C6 glioma cells. Following thrombin- signiﬁcantly inhibited mGlu1a receptor plasma membrane induced extinction of cell surface receptor ﬂuorescence, we could recovery, 1 h after cleavage of the receptor extracellular YFP measure intracellular receptor trafﬁcking efﬁciency and kinetics tag with thrombin (Figure 5C).Wealsotestedthe effects of from the Golgi to the cell surface, by ﬂuorescence recovery Kif5 rigor and Kif5 unbound mutants on cell surface targeting after cleavage (FRAC) experiments. Interestingly, deletion of the of N-terminal YFP-tagged mGlu7 receptor in FRAC experiments. C-terminus of mGlu1a receptor (YFP-th/mGlu1aΔC) signiﬁcantly Unlike YFP-tagged mGlu1 receptors, the membrane targeting of delayed plasma membrane targeting of the mutant receptor over YFP-mGlu7 receptor was unaffected by Kif5 mutants (Figure 5D). aperiodof 120 min (Figure 5A). Knowing that the C-terminus These experiments suggested that Kif5 controls membrane tar- of mGlu1a is required for interaction of the receptor to the geting of mGlu1a receptor via microtubule-dependent transport Kif5-containing complex (Figure 3), these experiments suggested in C6 glioma cell line. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 427 Figure 3 Identiﬁcation of mGlu1a receptor and SNAP23 interaction domains by BRET analysis. (A and B) mGlu1a directly interacts with SNAP23, but not Kif5. BRET signal between mGlu1a-Rluc8 and YFP-SNAP23 (A) or mGlu1a-Rluc8 and YFP-Kif5 (B) constructs. Neuroblastoma cells were co-transfected with a ﬁxed concentration of mGlu1a-Rluc8 and increasing concentrations of YFP-SNAP23 (A) or YFP-Kif5 (B). (C–F) mGlu1a receptor proximal C-terminus (until amino acid position 971) is required for its interaction with SNAP23. BRET signal between trun- cated mGlu1a-Rluc8 receptors and YFP-SNAP23. Neuroblastoma cells were co-transfected with a ﬁxed concentration of C-terminus-truncated mGlu1a receptor fused to Rluc8 constructs as illustrated, and increasing concentrations of YFP-SNAP23 plasmids. (G–J) SNAP-23 amino acids 88–128 are required for its interaction with mGlu1a receptor. BRET signal between mGlu1a-Rluc8 receptors and C-terminus-truncated YFP-SNAP23 proteins. Neuroblastoma cells were co-transfected with a ﬁxed concentration of Rluc8-tagged mGlu1a receptor and increasing concentrations of YFP-tagged truncated SNAP23 mutants as illustrated. The data shown are representatives of ﬁve independent experiments performed in triplicate and expressed as mean ± SEM. The curves were ﬁtted using a nonlinear regression equation, assuming a single bind- ing site. Note that mGlu1a-Rluc8 receptor did not interact with YFP-Kif5 (B), mGlu1aΔ341 receptor mutant did not interact with YFP-SNAP23 (F), and YFP-SNAP23Δ122 mutant did not interact with mGlu1a-Rluc8 receptor (J), as indicated by a BRET signal that increased linearly (rather than hyperbolically) with the YFP-tagged protein expression level, which most likely reﬂected random collision. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 428 j Raynaud et al. Figure 4 Kif5–SNAP23 complex controls microtubule trafﬁcking of the mGlu1a receptor. Visualization of Kif5–SNAP23–mGlu1a receptor com- plex after rapalog treatment in C6 glioma cell line. Images were obtained in the absence (upper line) or presence (bottom line) of the rapa- log dimerizer. The framed area is magniﬁed in the right bottom corner. (A) Kif5 rigor-stained microtubules after rapalog-induced reconstruction of split Kif5 (arrows). Cells were transfected with Kif5 cargoless/mcherry-FKBP and YFP-tagged rapamycin-binding domain (FRB) associated with kinesin CBD (FRB/YFP-CBD) constructs. (B) SNAP23-stained microtubules after rapalog-induced Kif5 rigor reconstruc- tion (arrows). Cells were transfected with split Kif5 cargoless/FKBP, FRB/mcherry-CBD, and YFP-SNAP23 constructs. (C) YFP-SNAP23-decorated micro- tubules after rapalog-induced forced interactions between SNAP23 and Kif5. Cells were transfected with Kif5 rigor cargoless/mcherry-FKBP and FRB/YFP-SNAP23 constructs. (D) mcherry-mGlu1a-stained microtubules induced by rapalog-forced interactions between SNAP23 and Kif5. Cells were transfected with the split Kif5 rigor cargoless/FKBP, FRB/YFP-SNAP23, and mcherry-mGlu1a receptor constructs. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 429 Figure 5 Kif5 is involved in mGlu1a receptor post-Golgi trafﬁc. (A) mGlu1a receptor cell surface targeting depends on the receptor C-terminus. Schematic view of FRAC experiments (upper part). YFP-th/mGlu1a wild-type receptor (YFP-th/mGlu1a) or C-terminus-truncated receptor (YFP-th/mGlu1aΔC) constructs were transfected in C6 glioma cells. Cell surface recovery of each receptor construct was expressed at different time points after thrombin-induced cleavage of the extracellular YFP tag and expressed as percentage of control wild-type recep- tor surface expression obtained in the absence of thrombin treatment (nc white bar). For this and following panels, values are mean ± SEM of three independent experiments. (B) Negative kif5 mutants inhibit cell surface mGlu1a receptor targeting. C6 glioma cells were co- transfected with YFP-th/mGlu1a receptor and kif5 rigor or kif5 unbound mutant constructs and subjected to FRAC experiments (same as in A). Results indicate a 50% inhibition of maximal cell surface receptor recovery in the presence of dominant-negative kif5 mutants. (C) Dose– response relationship and kif5-speciﬁc inhibition of cell surface YFP-th/mGlu1a receptor targeting. C6 glioma cells were co-transfected with a ﬁxed concentration of YFP-th/mGlu1a receptor construct and increasing concentrations of dominant-negative Kif5 or Kif3 unbound mutant constructs. Cells were then subjected to FRAC experiments to measure cell surface recovery of YFP-th/mGlu1a receptor 1 h after cleavage of the YFP tag with thrombin (same as in A). Results indicate inhibition of cell surface receptor recovery with increasing concentrations of Kif5, but not Kif3 unbound constructs. (D) Dominant-negative Kif5 mutants did not modify mGlu7a receptor cell surface expression. Cells were co- transfected with YFP-th/mGlu7a receptor and Kif5 rigor or Kif5 unbound mutants and subjected to FRAC experiments to measure cell surface YFP-th/mGlu7a receptor recovery after cleavage of its external YFP tag (same as in A). Results indicate no modiﬁcation in receptor recovery in the presence of dominant-negative Kif5 rigor and unbound mutants. The Kif5–SNAP23 complex controls mGlu1a receptor membrane investigated the role of this complex in mGlu1a receptor mem- targeting in neurons brane targeting. Deletion of the C-terminus of mGlu1a receptor We then investigated whether the role of Kif5–SNAP23 com- (YFP-th/mGlu1aΔC, Figure 7A) delayed membrane targeting of plex in membrane targeting of mGlu1a receptor that we found YFP-tagged mGlu1a receptor in neurons, indicating a role of the in C6 glioma cell line also applied to neurons. Transfected C-terminus of the receptor in this process. As Kif5 was found to mcherry-tagged mGlu1a receptor colocalized with YFP-tagged control membrane targeting of mGlu1a receptor in C6 glioma cells Kif5 or YFP-tagged SNAP23 in mouse cultured hippocampal neu- (Figure 5), we investigated whether this also applied to neurons. rons (Figure 6A and B; Supplementary Movies S2−S4). These We found that Kif5 unbound mutants altered membrane recovery results suggest that the Kif5–SNAP23 complex is involved in the of YFP-tagged mGlu1areceptor(Figure 7B) in FRAC experiments, vesicular transport of mGlu1a receptor in neurons. We then in neurons. Conversely, deletion of postsynaptic density protein, Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 430 j Raynaud et al. Figure 6 Co-localization of mcherry-mGlu1a receptor with Kif5 and SNAP23 proteins in hippocampal neurons. Arrows indicate co-localization of mcherry-mGlu1a receptor with Kif5 (A) and SNAP23 (B). Drosophila disc large tumor suppressor, zonula occludens-1 protein Our study was focused on mGlu1 receptor trafﬁcking along (PDZ) and homer interaction motifs included in the C-terminus of dendrites, because we found that mGlu1 receptor in cultured mGlu1a receptor did not affect the membrane targeting of the mouse hippocampal neurons is only expressed in dendrites receptor (Figure 7C). Together these experiments suggest that the (data not shown; Das and Banker, 2006). However, in other cell molecular motor Kif5 binds to the cargo protein SNAP23 and types, mGlu1 receptors can be expressed at presynaptic sites as recruits mGlu1a receptors to control the receptor membrane target- well. The nature of molecular motors and adaptors involved in ing in neurons. the presynaptic trafﬁcking of mGlu1 receptors is still unknown. Interaction with other protein partners might trigger speciﬁc Discussion subcellular sorting and trafﬁcking of the receptor. Here we revealed aspects of the molecular mechanisms of A proteomic approach following mGlu1 receptor immunopreci- postsynaptic mGlu1 receptor transport in neurons. We ﬁrst iden- pitation experiments allowed us to identify a large number of tiﬁed Kif5 and SNAP23 as molecular motor and adaptor protein proteins including nine proteins known to be involved in cellular associated with mGlu1 receptor. We show that SNAP23–Kif5 transport (Figure 1). Among these proteins, we focused on Kif5 complex controls mGlu1 receptor dendritic trafﬁcking via and SNAP23. Kif5 and SNAP23 were known to form a complex microtubule-dependent transport. Previous studies have shown (Diefenbach et al., 2002). Here we corroborated this result and that the molecular motor myosin VI and Kif17 (Wu et al., 2002), further showed that Kif5–SNAP23 complex could in turn interact and vesicular fusion protein SNAP23 and SNAP25 (Osten et al., with the carboxyl terminus of mGlu1 receptor thus forming a 1998; Noel et al., 1999; Washbourne et al., 2004; Suh et al., Kif5–SNAP23–mGlu1 receptor complex. We found that this com- 2010) regulate synaptic delivery of ionotropic AMPA and NMDA plex was recruited on microtubule like structures, which was con- glutamate receptors. Thus the present and previous results fur- sistent with the hypothesis that the molecular motor Kif5 could ther argue in favor of an important role of kinesins and SNARE be responsible for the vesicular transport of the SNAP23–mGlu1 proteins in the postsynaptic delivery of glutamate receptor fam- receptor cargo along dendritic microtubules. As SNAP23 belongs ily, therefore participating to the formation and possibly matur- to the large family of SNARE proteins involved in membrane ation of glutamatergic synapses. fusion processes, SNAP23 may also control exocytosis of the Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 431 Figure 7 mGlu1a receptor C-terminus- and Kif5-dependent cell surface targeting of YFP-th/mGlu1a receptor in cultured hippocampal neu- rons. (A) Deletion of YFP-th/mGlu1 receptor C-terminus impairs cell surface targeting of YFP-th/mGlu1a receptor. Wild-type YFP-mGlu1a receptor (mGlu1a) or C-terminus-truncated receptor (mGlu1aΔC) constructs were transfected in hippocampal neurons (same as Figure 5A). For this and following panels, values are mean ± SEM of three independent experiments. (B) Kif5 unbound mutant prevents YFP-th/mGlu1 receptor cell surface recovery. FRAC experiments performed in neurons co-transfected with wild-type YFP-th/mGlu1a receptor (mGlu1a) and Kif5 unbound mutant. (C) PDZ and homer interaction motifs are not involved in cell surface YFP-th/mGlu1a receptor targeting. Wild-type YFP-mGlu1a receptor (mGlu1a) or receptor deleted from its PDZ (mGlu1aΔ-PDZ) or homer (mGlu1Δ-homer) binding motif was transfected in hippocampal neurons. Cells were then processed as in B. mGlu1 receptor to the cell surface. This hypothesis was consist- neuronal cells. In the brain, SNAP23 and SNAP25 are both ent with our ﬁnding that the deletion of the carboxyl terminus of located in the somato-dendritic neuronal compartment, but dis- mGlu1 receptor delayed the membrane targeting of the receptor. play non-overlapping distribution. SNAP23 mediates the mem- The existence of a Kif5–SNAP23–mGlu5 complex suggests that brane delivery of NMDA receptors (Suh et al., 2010). Here we Kif5–SNAP23-mediated transport could apply to both mGlu1 and show that SNAP23 regulates the dendritic trafﬁcking and postsy- mGlu5 receptors. naptic membrane delivery of mGlu1 receptors. Thus consistent SNARE proteins mediate membrane fusion events in various with the somato-dendritic location of SNAP23, both mGlu1 and cell types (Lin and Scheller, 2000; Jahn and Scheller, 2006). For NMDA receptors bind to SNAP23. However, these receptors instance, SNAP25 is a neuron-speciﬁc SNARE protein that regu- seem to display distinct vesicular transport. The NMDA receptor lates synaptic vesicle fusion with the presynaptic plasma mem- was shown to be transported by the molecular motor Kif17 brane (Lin and Scheller, 2000; Jahn and Scheller, 2006), and (Setou et al., 2000; Guillaud et al., 2003). Whether SNAP23 and membrane delivery of postsynaptic AMPA receptors (Hoerndli Kif17 belong to the same complex to transport NMDA receptor et al., 2013). Postsynaptic translocation of AMPA receptors has is still an open question. If so, this would imply that mGlu1 and also been shown to depend on the actin motor myosin VI and NMDA receptors are sorted to two different motors (Kif5 and the adaptor protein SAP97 (Wu et al., 2002). The SNAP25 homo- Kif17) through the same adaptor. An interesting topic would log, SNAP23, is ubiquitously expressed in neuronal and non- then be to understand which molecular determinant beside the Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 432 j Raynaud et al. adaptor itself could specify the interaction with selective Rluc8, and p-mGlu1a-Δ62-Rluc8 were obtained by inserting molecular motors, such as speciﬁc cargo-adaptor conformation Rluc8 ended by a stop codon after positions 859, 971, 1029, or post-translational modiﬁcations. An attractive alternative and 1138 of p-mGlu1a coding sequence (1200 amino acids). In hypothesis would be that adaptors would specify the nature of the plasmids p-mGlu1a-Δ229-Rluc8, p-mGlu1a-Δ171-Rluc8, and the molecular motor, which can be supported by the fact that p-mGlu1a-Δ62-Rluc8, the retention motif RRKK in the C-terminal mLin-10 has been identiﬁed as adaptor protein for Kif17 in the coding sequence of mGlu1 was exchanged for AAAA to avoid transport of NMDA receptors (Setou et al., 2000). Hence more endoplasmic reticulum retention as previously described for this explorations are needed to fully understand these molecular receptor (Chan et al., 2001). mechanisms, but Kif5, Kif17, and myosin VI molecular motors as The plasmid YFP-th-mGlu7a was obtained by introducing YFP-th well as SAP97, SNAP23, and SNAP25 proteins seem to play before the coding sequence of p-mGlu7a. essential roles in glutamate receptors speciﬁc trafﬁcking. Kif5band Kif5c constructs were generated from the mouse Metabotropic glutamate receptors modulate efﬁcacy of synap- Kif5band mouseKif5cIMAGE clones (#30543821 and #30536079, tic transmission mediated by the ionotropic AMPA and NMDA respectively; Gene Service, Source BioScience). The rigor muta- glutamate receptors. Regulation of mGlu receptors delivery at tion was obtained using point mutation-containing primers to postsynaptic sites is essential for an optimal functioning of the substitute the threonine 92 by a valine. We created the unbound synapse. The strength of synaptic transmission also depends on mutation using point mutation-containing primers to substitute the number of functional postsynaptic ionotropic AMPA and the serine 204 by a glycine and the histidine 205 by a glycine. NMDA glutamate receptors. For instance, long-term potentiation The plasmids p-YFP-Kif5b and p-YFP-Kif5c were obtained by PCR and long-term depression of synaptic transmission can result ampliﬁcation of the EYFP coding sequence from p-EYFP and from an increased and decreased number of active postsynaptic insertion into p-Kif5b and p-Kif5c. ionotropic glutamate receptors, respectively. Parallel modiﬁca- To obtain p-FRB-YFP-CBD, we ampliﬁed by PCR the Kif5b cod- tion in the number of mGlu receptors should therefore be ing sequence (amino acids 816–964) from p-Kif5b and inserted required to maintain an optimal tuning of either the potentiated into p-FRB-YFP-snap23. The sequence 816–964 contains the or depressed synapse. Thus kinesins and SNARE proteins would CBD and the minimal domain for SNAP23 binding (814–907)as play essential roles not only during synaptogenesis, but also previously described (Diefenbach et al., 2002). during synaptic plasticity by controlling appropriate delivery of p-Kif5 rigor cargoless-mcherry was obtained by introducing metabotropic receptors to maintain adequate synaptic tuning. It mcherry after position 560 in Kif5c coding sequence. This plas- is striking that glutamate receptors use different proteins for mid was further transformed into p-Kif5 rigor cargoless-mcherry- their intracellular trafﬁcking although they may be translocated FKBP using pC M-F2E (ARIAD Pharmaceuticals, Inc); and p-Kif5 to same postsynaptic sites. One possible reason for such differ- rigor cargoless-FKBP was obtained by mcherry coding sequence ent mechanisms is that each receptor may need speciﬁc regula- removal. tion for appropriate tuning of the synapse. Therefore, this p-Kif3a was generated from the human Kif3a IMAGE clone implies ﬁne regulation of the glutamate receptor assembly with (#5298675; Gene Service, Source BioScience). From p-Kif3a, we their cargo proteins, a process that remains largely elusive created the unbound mutation using point mutation-containing currently. primers to substitute the serine 220 by an alanine and the histi- dine 221 by an alanine. Materials and methods SNAP23 constructs were generated with the mouse SNAP23b DNA constructs IMAGE clone (#30606358; Gene Service, Source BioScience). YFP-mGlu1a, mGlu1aΔC(equivalent to mGlu1aΔ341), mGlu1a- The plasmid p-YFP-SNAP23 was obtained by subcloning the YFP, and mGlu7a plasmids were gifts from Dr Jean-Philippe Pin PCR ampliﬁed SNAP23 coding sequence (210 amino acids) into (IGF, CNRS, INSERM, Universite´ Montpellier). p-EYFP (Clontech); p-YFP-SNAP23Δ8, Δ49, Δ82, and Δ122 were The plasmid YFPth-mGlu1a was obtained by introducing the obtained by introducing a stop codon after positions 202, 161, thrombin cleavage sequence GGLVPRGSGG immediately after the 128, and 88 of the SNAP23b coding sequence. YFP sequence by site-directed mutagenesis using oligonucleotide: The plasmid p-FRB-YFP-SNAP23 was obtained by insertion of 5′-ATACACGCGTACCACCACTGCCTCTCGGAACTAAACCACCCTTGTACA the FRB coding sequence from pC -R E (ARIAD Pharmaceuticals, 4 H GCTCGTCCAT-3′ and subcloning. The plasmid mcherry-th-mGlu1a Inc.) into p-YFP-SNAP23, after introduction of a Spe1 restriction was obtained by exchanging YFP for mcherry. The plasmid site just after the initiation codon in p-YFP-SNAP23. All mutants p-YFP-th-mGlu1aΔC was obtained by introducing YFP-th in were generated by PCR and veriﬁed by sequencing. p-mGlu1aΔC. The plasmid p-mGlu1a-Rluc8 was obtained by exchanging YFP from p-mGlu1a-YFP for Rluc8 from pcDNA-Rluc8. Neuroblastoma cells, C6 glioma cells, and hipppocampal The plasmid p-YFP-th-mGlu1aΔPDZ was obtained by mutation neuronal culture of PDZ binding motif (STL) in p-YFP-th-mGlu1a. The plasmid Neuroblastoma N2a cells were obtained from Dr A. Varrault p-YFP-th-mGlu1aΔhomer was obtained by mutation of homer (IGF, Montpellier). They were cultured as previously described binding motif (PPSPFR) in p-YFP-th-mGlu1a. The plasmids (Brabet et al., 1991). Cells were transfected with lipofectamine p-mGlu1a-Δ341-Rluc8, p-mGlu1a-Δ229-Rluc8, p-mGlu1a-Δ171- 2000 (Invitrogen) according to the manufacturer’s instructions. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 433 C6 glioma cells were cultured and transfected as described pre- (M) was speciﬁed as variable modiﬁcation. Database searches viously (Giau et al., 2005). Wistar rat hippocampal cultures were were performed with a mass tolerance of 20 ppm for precursor prepared from E-17/18 embryos of either sex and grown in ion for mass calibration, and with a 4.5-ppm tolerance after cali- Neurobasal medium supplemented with 0.5mM L-glutamine, 2% bration. The maximum false peptide and protein discovery rate B27, and 10% heat-inactivated horse serum (all from Invitrogen). was speciﬁed as 0.01. Seven amino acids were required as min- Hippocampal neurons were transfected at DIV-9 with lipofecta- imum peptide length. The MaxQuant software generates several mine 2000 (Invitrogen) according to the manufacturer’s stand- output ﬁles that contain information about identiﬁed peptides ard protocol. For immunoﬂuorescence and FRAC experiments, and proteins. The ‘proteinGroups.txt’ ﬁle is dedicated to the iden- cell lines were plated onto coverslips (12 or 35 mm diameter) tiﬁed proteins: each single row collapses into protein groups all and hippocampal neurons onto coverslips coated with 50 μg/ml proteins that cannot be distinguished based on identiﬁed pep- poly-D-lysine (Sigma). tides. An in-house bioinformatics tool has been developed to automatically select a representative protein ID in each protein Protein separation and identiﬁcation by LC–MS/MS group. First, proteins with the most identiﬁed peptides are iso- YFP-mGlu1aorYFP-mGlu1aΔC receptors immunoprecipitated lated in a so called ‘match group’ (proteins from the ‘Protein IDs’ with GFP-Trap-A were separated on SDS-PAGE gels (12% polyacryl- column with the maximum number of ‘peptides counts (all)’). For amide, Mini-PROTEAN TGX™ Precast Gels, Bio-Rad) and stained the remaining match groups where more than one protein ID with Protein Staining Solution (Euromedex). Gel lanes were cut existed after ﬁltering, the ‘leading’ protein has been chosen as into ﬁve gel pieces and destained with 50 mM triethylammonium the best annotated protein in UniProtKB (reviewed entries rather biCarbonate (TEABC) and three washes in 100% acetonitrile. than automatic ones, the one with the highest protein existence Proteins were digested in gel using trypsin (500 ng/band, Gold, evidence or the most annotated protein according to the number Promega) as previously described (Thouvenotetal., 2008). Digest of GO annotations dated from 05/09/2017). products were dehydrated in a vacuum centrifuge and reduced to 3 μl. The generated peptides were analyzed online by nano- Bioinformatics and network analysis ﬂowHPLC–nanoelectrospray ionization using on a LTQ-Orbitrap XL Only proteins with a fold enrichment of six (based on the MS/MS mass spectrometer (Thermo Scientiﬁc) coupled to an Ultimate 3000 ratio) and a minimum of two unique peptides per protein were con- HPLC (Thermo Fisher Scientiﬁc). Desalting and pre-concentration of sidered as potential interactors of the carboxyl terminus of the samples were performed online on a Pepmap pre-column mGlu1 receptor (Supplementary Table S1). Protein–protein interac- (0.3 mm × 10 mm, Dionex). A gradient consisting of 0–40%B tions of these 173 proteins have been downloaded from the IntAct for 60 min and 80% B for 15 min (A = 0.1% formic acid, 2% database (Orchard et al., 2014)(October 12, 2017)and ﬁltered acetonitrile in water; B = 0.1% formic acid in acetonitrile) at 300 based on an edge score set to 0.6 (high conﬁdence). Protein net- nl/min was used to elute peptides from the capillary reverse- work analysis was performed using Cytoscape (Shannon et al., phase column (0.075 mm × 150 mm, Acclaim Pepmap 100 2003)(version 3.5.1) and the ClueGO application (Bindea et al., C18, Thermo Fisher Scientiﬁc). Eluted peptides were electro- 2009)(v2.3.4) was used to handle GO annotations (from EBI- sprayed online at a voltage of 2.4 kV into a LTQ-Orbitrap XL QuickGO-GOA, September 6–October 16, 2017). A focus was made mass spectrometer. A cycle of one full scan mass spectrum on proteins annotated to the ‘intracellular transport’ GO term, (400–2000 m/z) at a resolution of 60000 (at 400 m/z), followed GO:0046907 (29/173 proteins; 9/29 with known PPI in IntAct). by ﬁve data-dependent MS/MS spectra was repeated continu- ously throughout the nanoLC separation. All MS/MS spectra Immunoprecipitation and western blots were recorded using normalized collision energy (35%, activa- Neuroblastoma cells were transfected with YFP-mGlu1aor tion) with an isolation window of 3 m/z. Data were acquired C-terminus-truncated YFP-mGlu1a (YFP-mGlu1aΔC) receptor. using the Xcalibur software (v 2.0.7). For all full scan measure- Twenty-four hours after transfection, cells (3 × 10-cm diameter ments with the Orbitrap detector, a lock-mass ion from ambient plates per condition) were washed and subjected to 20°C tem- air (m/z 445.120024) was used as an internal calibrant as perature blockade for 2 h without thrombin and then shifted to described (Olsen et al., 2005). Analysis of MS data was per- 37°C for 30 min as described for FRAC in C6 glioma cells (see formed using the MaxQuant software package (v 1.5.5.1)(Cox below). Cells were lysed and lysates processed as previously and Mann, 2008). Tandem mass spectra (MS/MS) were searched described (Moutin et al., 2014) using GFP-Trap-A (ChromoTek) in by the Andromeda search engine (Cox et al., 2011) against the place of RFP-Trap-A. After washing, the solid phase was incu- UniProtKB Reference Proteome UP000000589 database for the bated in Laemmli buffer at 60°C and protein samples resolved Mus musculus taxonomy (release 2017_08, 22277 canonical by PAGE on a 10% gel, transferred onto nitrocellulose sheet and entries) and the speciﬁc sequence of interest using the following subjected to immunoblotting using anti-kif5 antibody (1/2000,H2, parameters: enzyme speciﬁcity was set as Trypsin/P, and a max- MAB1614, Chemicon Int., Millipore) and anti-snap23 (1/2000, imum of two missed cleavages and a mass tolerance of 0.5 Da for Synaptic Systems or Abcam) for 2 h. Blots were then washed fragment ion were applied. A second database of known contami- three times with PBS containing 0.1% Tween-20 (PBST) and nants provided with the MaxQuant suite was also employed. The incubated with donkey anti-rabbit or anti-mouse peroxidase- ‘match between runs’ and ‘iBAQ’ options were checked. Oxidation conjugated antibodies (1/2000, Rockland) for 1 h. Blots were Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 434 j Raynaud et al. washed ﬁve times with PBST and proteins were visualized with FRB-tagged protein consisted of CBD from Kif5 tail (FRB-mcherry- ECL Westen blotting detection reagent (Amersham, GE Healthcare CBD or FRB-YFP-CBD) and YFP-SNAP23 (FRB/YFP-SNAP23). Life Sciences) on ChemiDoc Imaging System (Bio-Rad). Twenty-four hours after cell transfection, heterodimerization of the FRB-tagged and FKBP-tagged proteins was induced by the BRET measurements addition of 1 μM heterodimerizer AP21967 (Ariad Pharmaceuti- We have previously described BRET measurements in cell cals), a rapamycin analog, for 15–60 min to drive the protein population (Perroy et al., 2004; Raynaud et al., 2013). Brieﬂy, co- complex onto microtubules. Control (not treated with AP21967) expression of mGlu1a-Rluc8 and YFP-SNAP23 or YFP-Kif5 in neuro- and treated cells were then ﬁxed and processed for microscopy. blastoma cells was monitored using a spectrophotometric plate reader (MITHRAS LB 940, Berthold Technologies). BRET signal Microscopy was plotted as a function of total ﬂuorescence over total lumines- Fixed cell images were acquired on a Zeiss Axiovert 200 TV cence signal, and the ratio used as an index of the concentration inverted microscope equipped with a software-driven ﬁlter of YFP-tagged over Rluc8-tagged proteins. Co-expression of wheel, a 63× oil immersion objective [1.3 numerical aperture increasing concentrations of YFP-tagged proteins with a con- (NA); Zeiss], or a 25× oil immersion objective (0.8 NA; Zeiss). A stant level of Rluc8-tagged proteins gives rise to a hyperbolical MicroMax 1300 charge-device camera (Princeton Instruments) increase and saturation in BRET signal, only if the two proteins driven by MetaMorph imaging software (version 4.17; Universal speciﬁcally interact. Random collisions between tagged proteins Imaging Corporation) was used for cell imaging. Filter sets gave a linear and non-saturating signal (bystander BRET), indic- (Zeiss) for YFP-tagged proteins were 500 ± 20 and 535 ± 30 nm ating non-speciﬁc protein–protein interactions. band pass ﬁlters for excitation and emission, respectively. For mcherry-tagged proteins, we used ﬁlters (Omega Optical) with a Fluorescence recovery after cleavage 555 ± 50 nm band pass and a 625 ± 50 nm for excitation The mGlu1a receptor was tagged on its extracellular N-terminal and emission, respectively. Z series of images were acquired (31 domain with a thrombin cleavable tag fused to YFP protein. The planes) from −3 μmto +3 μm to image cells. An equatorial plane recombinant protein was transfected in C6 glioma cell line, and was selected from each series. 24 h later cells were washed twice with DMEM medium (ref Imaging of live cells was performed on the same Zeiss setup. 61965, Thermo Fisher Scientiﬁc), supplemented with 10mM A homemade temperature-regulated chamber was used at 34°C HEPES and 1 mg/ml chicken egg albumin (Sigma). Cells were to slightly reduce vesicle motion. For time-lapse imaging, expos- then incubated at 20°C in the same medium to block post-Golgi ure times were 300–500 ms for YFP-tagged proteins, 0.5–1 sec for trafﬁc, followed by thrombin treatment (5 U/ml, Calbiochem) mcherry-tagged proteins. Images from the MicroMax camera were (Rosenberg et al., 2001) to cleave the external N-terminal YFP tag converted to TIFF ﬁles that were edited using ImageJ 1.47 (NIH of cell surface mGlu1a receptors. After extensive washing, cells Image) and compiled into QuickTime (Apple Computers) movies. were returned to their initial medium at 37°C for cell surface receptor recovery. Cells were then washed and ﬁxed at different Supplementary material time periods and quantiﬁcation of cell surface receptor was deter- Supplementary material is available at Journal of Molecular mined using rabbit polyclonal anti-GFP antibodies (Invitrogen). For Cell Biology online. quantitative measurements, coverslips were transferred in 24-well microplates and processed for in-cell western assay (Odyssey, Acknowledgements LI-COR Biosciences). Brieﬂy, cells were blocked with gelatin, Mass spectrometry experiments were carried out using facil- washed, incubated with anti-GFP antibodies, washed again and ities of the Functional Proteomics Platform of Montpellier and ﬁnally incubated with IRDye800CW-conjugated donkey anti-rabbit BRET experiments using facilities of Arpege pharmacological polyconal antibodies (Rockland). Infrared ﬂuorescence was deter- screening platform, at the Institute of Functional Genomics. mined using the 700 nm channel of the Odyssey Infrared Imaging System (LI-COR Biosciences). Fluorescence intensity was measured Funding using Image J software from circular gates placed in the middle of This work was supported by the European Research Council (ERC) the wells. Background values measured in wells containing mock- under the European Union’s Horizon 2020 research and innovation transfected cells were subtracted from the signal. All measure- programme (grant agreement no. 646788 to J.P.), the Agence ments were performed in triplicate and expressed as mean ± SEM. NationaledelaRecherche (ANR-13-JSV4-0005-01 to J.P.), and the Re´gion Languedoc-Roussillon (Chercheur d’Avenir 146090 to J.P.). Rapalog-mediated protein heterodimerization Conventional kinesin1 (Kif5) rigor mutant tightly bound to Conﬂict of interest: none declared. microtubules in C6 glioma cells (data not shown) has been pre- viously described for mouse ﬁbroblast L cells (Nakata and References Hirokawa, 1995). Presently, we replaced the tail of the protein, Aiba, A., Chen, C., Herrup, K., et al. (1994a). Reduced hippocampal long-term which binds the cargo, by FKBP-tagged or not with mcherry potentiation and context-speciﬁc deﬁcit in associative learning in mGluR1 (Kif5 cargoless/FKBP and Kif5 cargoless/mcherry-FKBP). The mutant mice. Cell 79, 365–375. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 mGlu1 receptor trafﬁcking j 435 Aiba, A., Kano, M., Chen, C., et al. (1994b). Deﬁcient cerebellar long-term Kitano, J., Kimura, K., Yamazaki, Y., et al. (2002). Tamalin, a PDZ domain- depression and impaired motor learning in mGluR1 mutant mice. Cell 79, containing protein, links a protein complex formation of group 1 metabo- 377–388. tropic glutamate receptors and the guanine nucleotide exchange factor Ango, F., Pin, J.P., Tu, J.C., et al. (2000). Dendritic and axonal targeting of cytohesins. J. Neurosci. 22, 1280–1289. type 5 metabotropic glutamate receptor is regulated by homer1 proteins Ledonne, A., Nobili, A., Latagliata, E.C., et al. (2015). Neuregulin 1 signalling and neuronal excitation. J. Neurosci. 20, 8710–8716. modulates mGluR1 function in mesencephalic dopaminergic neurons. Mol. Bindea, G., Mlecnik, B., Hackl, H., et al. (2009). ClueGO: a Cytoscape plug-in Psychiatry 20, 959–973. to decipher functionally grouped gene ontology and pathway annotation Lin, T.B., Lai, C.Y., Hsieh, M.C., et al. (2015). VPS26A-SNX27 interaction- networks. Bioinformatics 25, 1091–1093. dependent mGluR5 recycling in dorsal horn neurons mediates neuropathic Brabet, P., Pantaloni, C., Bockaert, J., et al. (1991). Metabolism of two Go pain in rats. J. Neurosci. 35, 14943–14955. alpha isoforms in neuronal cells during differentiation. J. Biol. Chem. 266, Lin, R.C., and Scheller, R.H. (2000). Mechanisms of synaptic vesicle exocyt- 12825–12828. osis. Annu. Rev. Cell Dev. Biol. 16, 19–49. Castillo, C.A., Leon, D.A., Ballesteros-Yanez, I., et al. (2010). Glutamate differ- Luis Albasanz, J., Fernandez, M., and Martin, M. (2002). Internalization of ently modulates metabotropic glutamate receptors in neuronal and glial metabotropic glutamate receptor in C6 cells through clathrin-coated vesi- cells. Neurochem. Res. 35, 1050–1063. cles. Brain Res. Mol. Brain Res. 99, 54–66. Chan, W.Y., Soloviev, M.M., Ciruela, F., et al. (2001). Molecular determinants Luscher, C., and Huber, K.M. (2010). Group 1 mGluR-dependent synaptic of metabotropic glutamate receptor 1B trafﬁcking. Mol. Cell. Neurosci. 17, long-term depression: mechanisms and implications for circuitry and dis- 577–588. ease. Neuron 65, 445–459. Choi, K.Y., Chung, S., and Roche, K.W. (2011). Differential binding of cal- Moutin, E., Compan, V., Raynaud, F., et al. (2014). The stoichiometry of scaf- modulin to group I metabotropic glutamate receptors regulates receptor fold complexes in living neurons—DLC2 functions as a dimerization engine trafﬁcking and signaling. J. Neurosci. 31, 5921–5930. for GKAP. J. Cell Sci. 127, 3451–3462. Cox, J., and Mann, M. (2008). MaxQuant enables high peptide identiﬁcation Nakata, T., and Hirokawa, N. (1995). Point mutation of adenosine rates, individualized p.p.b.-range mass accuracies and proteome-wide pro- triphosphate-binding motif generated rigor kinesin that selectively blocks tein quantiﬁcation. Nat. Biotechnol. 26, 1367–1372. anterograde lysosome membrane transport. J. Cell Biol. 131, 1039–1053. Cox, J., Neuhauser, N., Michalski, A., et al. (2011). Andromeda: a peptide Noel, J., Ralph, G.S., Pickard, L., et al. (1999). Surface expression of AMPA search engine integrated into the MaxQuant environment. J. Proteome receptors in hippocampal neurons is regulated by an NSF-dependent Res. 10, 1794–1805. mechanism. Neuron 23, 365–376. Das, S.S., and Banker, G.A. (2006). The role of protein interaction motifs in Olsen, J.V., de Godoy, L.M., Li, G., et al. (2005). Parts per million mass accur- regulating the polarity and clustering of the metabotropic glutamate acy on an Orbitrap mass spectrometer via lock mass injection into a receptor mGluR1a. J. Neurosci. 26, 8115–8125. C-trap. Mol. Cell. Proteomics 4, 2010–2021. Diefenbach, R.J., Diefenbach, E., Douglas, M.W., et al. (2002). The heavy Orchard, S., Ammari, M., Aranda, B., et al. (2014). The MIntAct project— chain of conventional kinesin interacts with the SNARE proteins SNAP25 IntAct as a common curation platform for 11 molecular interaction data- and SNAP23. Biochemistry 41, 14906–14915. bases. Nucleic Acids Res. 42,D358–D363. Duc, C., and Catsicas, S. (1995). Ultrastructural localization of SNAP-25 Osten, P., Srivastava, S., Inman, G.J., et al. (1998). The AMPA receptor GluR2 within the rat spinal cord and peripheral nervous system. J. Comp. Neurol. C terminus can mediate a reversible, ATP-dependent interaction with NSF 356, 152–163. and alpha- and beta-SNAPs. Neuron 21, 99–110. Ferraguti, F., Crepaldi, L., and Nicoletti, F. (2008). Metabotropic glutamate 1 Oyler, G.A., Higgins, G.A., Hart, R.A., et al. (1989). The identiﬁcation of a receptor: current concepts and perspectives. Pharmacol. Rev. 60, novel synaptosomal-associated protein, SNAP-25, differentially expressed 536–581. by neuronal subpopulations. J. Cell Biol. 109, 3039–3052. Giau, R., Carrette, J., Bockaert, J., et al. (2005). Constitutive secretion of pro- Perroy, J., Pontier, S., Charest, P.G., et al. (2004). Real-time monitoring of tease nexin-1 by glial cells and its regulation by G-protein-coupled recep- ubiquitination in living cells by BRET. Nat. Methods 1, 203–208. tors. J. Neurosci. 25, 8995–9004. Ravichandran, V., Chawla, A., and Roche, P.A. (1996). Identiﬁcation of a novel Guillaud, L., Setou, M., and Hirokawa, N. (2003). KIF17 dynamics and regula- syntaxin- and synaptobrevin/VAMP-binding protein, SNAP-23, expressed in tion of NR2B trafﬁcking in hippocampal neurons. J. Neurosci. 23, non-neuronal tissues. J. Biol. Chem. 271, 13300–13303. 131–140. Raynaud, F., Janossy, A., Dahl, J., et al. (2013). Shank3-Rich2 interaction reg- Hannan, A.J., Blakemore, C., Katsnelson, A., et al. (2001). PLC-beta1, acti- ulates AMPA receptor recycling and synaptic long-term potentiation. vated via mGluRs, mediates activity-dependent differentiation in cerebral J. Neurosci. 33, 9699–9715. cortex. Nat. Neurosci. 4, 282–288. Rosenberg, M., Meier, J., Triller, A., et al. (2001). Dynamics of glycine Hirokawa, N., Niwa, S., and Tanaka, Y. (2010). Molecular motors in neurons: receptor insertion in the neuronal plasma membrane. J. Neurosci. 21, transport mechanisms and roles in brain function, development, and dis- 5036–5044. ease. Neuron 68, 610–638. Setou, M., Nakagawa, T., Seog, D.H., et al. (2000). Kinesin superfamily motor Hoerndli, F.J., Maxﬁeld, D.A., Brockie, P.J., et al. (2013). Kinesin-1 regulates protein KIF17 and mLin-10 in NMDA receptor-containing vesicle transport. synaptic strength by mediating the delivery, removal, and redistribution of Science 288, 1796–1802. AMPA receptors. Neuron 80, 1421–1437. Shannon, P., Markiel, A., Ozier, O., et al. (2003). Cytoscape: a software envir- Ichise, T., Kano, M., Hashimoto, K., et al. (2000). mGluR1 in cerebellar onment for integrated models of biomolecular interaction networks. Purkinje cells essential for long-term depression, synapse elimination, and Genome Res. 13, 2498–2504. motor coordination. Science 288, 1832–1835. Suh, Y.H., Terashima, A., Petralia, R.S., et al. (2010). A neuronal role for Jahn, R., and Scheller, R.H. (2006). SNAREs—engines for membrane fusion. SNAP-23 in postsynaptic glutamate receptor trafﬁcking. Nat. Neurosci. 13, Nat. Rev. Mol. Cell Biol. 7, 631–643. 338–343. Jenkins, B., Decker, H., Bentley, M., et al. (2012). A novel split kinesin assay Tao-Cheng, J.H., Du, J., and McBain, C.J. (2000). Snap-25 is polarized to axons identiﬁes motor proteins that interact with distinct vesicle populations. and abundant along the axolemma: an immunogold study of intact neu- J. Cell Biol. 198, 749–761. rons. J. Neurocytol. 29, 67–77. Kenny, P.J., and Markou, A. (2004). The ups and downs of addiction: role Thouvenot, E., Urbach, S., Dantec, C., et al. (2008). Enhanced detection of of metabotropic glutamate receptors. Trends Pharmacol. Sci. 25, CNS cell secretome in plasma protein-depleted cerebrospinal ﬂuid. 265–272. J. Proteome Res. 7, 4409–4421. Downloaded from https://academic.oup.com/jmcb/article-abstract/10/5/423/4995849 by Ed 'DeepDyve' Gillespie user on 13 November 2018 436 j Raynaud et al. Viwatpinyo, K., and Chongthammakun, S. (2009). Activation of group I meta- Washbourne, P., Thompson, P.M., Carta, M., et al. (2002b). Genetic ablation botropic glutamate receptors leads to brain-derived neurotrophic factor of the t-SNARE SNAP-25 distinguishes mechanisms of neuroexocytosis. expression in rat C6 cells. Neurosci. Lett. 467, 127–130. Nat. Neurosci. 5, 19–26. Washbourne, P., Bennett, J.E., and McAllister, A.K. (2002a). Rapid recruit- Wu, H., Nash, J.E., Zamorano, P., et al. (2002). Interaction of SAP97 with ment of NMDA receptor transport packets to nascent synapses. Nat. minus-end-directed actin motor myosin VI. Implications for AMPA receptor Neurosci. 5, 751–759. trafﬁcking. J. Biol. Chem. 277, 30928–30934. Washbourne, P., Liu, X.B., Jones, E.G., et al. (2004). Cycling of NMDA recep- Zhou, L., Yang, D., Wang, D.J., et al. (2015). Numb deﬁciency in cerebellar tors during trafﬁcking in neurons before synapse formation. J. Neurosci. Purkinje cells impairs synaptic expression of metabotropic glutamate recep- 24, 8253–8264. tor and motor coordination. Proc. Natl Acad. Sci. USA 112, 15474–15479.

Journal

Journal of Molecular Cell Biology – Oxford University Press

Published: Oct 1, 2018

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

SNAP23–Kif5 complex controls mGlu1 receptor trafficking

SNAP23–Kif5 complex controls mGlu1 receptor trafficking

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

SNAP23–Kif5 complex controls mGlu1 receptor trafficking

SNAP23–Kif5 complex controls mGlu1 receptor trafficking

References (108)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies