Eph receptor tyrosine kinases participate in a variety of normal and pathogenic processes during development and throughout adulthood. This versatility is likely facilitated by the ability of Eph receptors to signal through diverse cellular signalling pathways: primarily by controlling cytoskeletal dynamics, but also by regulating cellular growth, proliferation, and survival. Despite many proteins linked to these signalling pathways interacting with Eph receptors, the specific mechanisms behind such links and their coordination remain to be elucidated. In a proteomics screen for novel EPHB2 multi-effector proteins, we identified human MYC binding protein 2 (MYCBP2 or PAM or Phr1). MYCBP2 is a large signalling hub involved in diverse processes such as neuronal connectivity, synaptic growth, cell division, neuronal survival, and protein ubiquitination. Our biochemical experiments demonstrate that the formation of a complex containing EPHB2 and MYCBP2 is facilitated by FBXO45, a protein known to select substrates for MYCBP2 ubiquitin ligase activity. Formation of the MYCBP2-EPHB2 complex does not require EPHB2 tyrosine kinase activity and is destabilised by binding of ephrin-B ligands, suggesting that the MYCBP2-EPHB2 association is a prelude to EPHB2 signalling. Paradoxically, the loss of MYCBP2 results in increased ubiquitination of EPHB2 and a decrease of its protein levels suggesting that MYCBP2 stabilises EPHB2. Commensurate with this effect, our cellular experiments reveal that MYCBP2 is essential for efficient EPHB2 signalling responses in cell lines and primary neurons. Finally, our genetic studies in C. elegans provide in vivo evidence that the ephrin receptor VAB-1 displays genetic interactions with known MYCBP2 binding proteins. Together, our results align with the similarity of neurodevelopmental phenotypes caused by MYCBP2 and EPHB2 loss of function, and couple EPHB2 to a signaling effector that controls diverse cellular functions.
This valuable study identifies an Ephrin type-B Receptor 2 (EPHB2) interactor, MYCBP2, as a potential regulator of EPHB2 stability and function. In contrast to expectations, based on MYCBP2 function in the ubiquitin pathway, loss of function of MYCBP2 resulted in less EPHB2 receptor and defective EPHB2 function. The paper is supported by a largely convincing set of biochemical, cell culture and in vivo experiments.
Eph receptor tyrosine kinases and their membrane-tethered ligands, the ephrins, elicit short distance cell-cell signals that regulate many biological processes. Ephrin-Eph signalling primarily impacts the cytoskeleton with the immobilization of highly dynamic axonal growth cones being a classic example. Other processes that involve changes in transcription, growth, and survival such as angiogenesis, synaptic plasticity, stem cell fate, tumorigenesis and neurodegeneration also involve the Eph/ephrin system. Many proteins are postulated to couple Eph receptors to different intracellular effectors, but the molecular logic of this diversity remains fragmented (Kania and Klein, 2016;Bush, 2022).
EphB subfamily members preferentially bind transmembrane ephrin-Bs and although both molecules participate in bidirectional signalling, ephrin-B activation of EphB signalling cascades is more thoroughly studied (Gale et al., 1996;Mellitzer et al., 1999). To elicit robust Eph receptor forward signalling, ephrins multimerise in signalling clusters by intercalating with Ephs on a signal-recipient cell with array size correlating with signal amplitude (Kullander et al., 2001;Schaupp et al., 2014). Signalling initiation involves the activation of the receptor tyrosine kinase and phosphorylation of tyrosines proximal to the EphB transmembrane domain (Soskis et al., 2012;Binns et al., 2000). Eph-evoked cytoskeletal effects such as cell contraction and growth cone collapse result from changes in small GTPase activity modulated by EphB Guanine nucleotide exchange factors (GEFs) and GTPase activating proteins (GAPs) (Margolis et al., 2010;Shi et al., 2007). Eph signalling has also been linked to fundamental cellular pathways such as the Ras-MAPK pathway, mTOR-regulated protein synthesis, cell division and survival (Bush and Soriano, 2010;Nie et al., 2010;Fawal et al., 2018;Genander et al., 2009;Depaepe et al., 2005). Eph forward signalling eventually leads to their internalisation and either recycling or degradation via endosome/lysosome and ubiquitination/proteasome pathways (Zimmer et al., 2003;Okumura et al., 2017). While identification of a growing number of proteins interacting with Eph receptors has moved the field forward, we have yet to clarify the question of how Eph receptors activate various fundamental cellular processes, often within the same cell type.
Myc-binding protein 2 (MYCBP2), also known as Protein Associated with Myc (PAM) and Highwire, RPM-1, or Phr1 in different species, is a large signalling hub that regulates cytoskeletal dynamics, neuronal development, and axonal degeneration (Guo et al., 1998;Grill et al., 2016;Virdee, 2022). It has an atypical RING ubiquitin ligase activity that inhibits the p38/MAP and JNK kinase pathways thereby regulating cytoskeletal dynamics underlying axonal development and synaptic growth (Nakata et al., 2005;Collins et al., 2006;Pao et al., 2018;Wan et al., 2000;Lewcock et al., 2007;Borgen et al., 2017). MYCBP2 further regulates the Tuberin Sclerosis Complex linked to cell growth (Han et al., 2012), initiation of autophagy via ULK (Crawley et al., 2019) and NMNAT2-regulated neuronal survival and axonal degeneration (Babetto et al., 2013;Xiong et al., 2012). Biochemical mapping has shown that human MYCBP2 and its C. elegans ortholog RPM-1 rely upon the FBD1 domain to bind the F-box protein FBXO45 that acts as a ubiquitination substrate selector (Desbois et al., 2018;Sharma et al., 2014). Intriguingly, the neurodevelopmental phenotypes caused by MYCBP2 and EPHB2 loss of function are similar, raising the possibility that these two molecules could function in the same pathway (Henkemeyer et al., 1996;Lewcock et al., 2007;Dalva et al., 2000). Nonetheless, a biochemical or genetic interaction between MYCBP2 and EPHB2 has not been demonstrated in any system.
To shed light on how EphB receptors fulfil their multitude of functions, we used mass spectrometry (MS)-based proteomics to identify multi-effector proteins that bind EPHB2. One of our proteomic hits was MYCBP2, which we demonstrated forms a complex with EPHB2 using a combination of biochemical and cellular assays. Furthermore, we show that this interaction is required for efficient EPHB2 signalling in cell lines and primary neurons. Consistent with these findings, we observed in vivo genetic interactions in C. elegans between the Eph receptor, VAB-1, and known RPM-1 binding proteins. Our collective results indicate that the relationship between EPHB2 and MYCBP2 does not appear to involve the ubiquitin ligase activity of MYCBP2, and raise the possibility that MYCBP2 links EPHB2 to diverse fundamental cellular functions.
Proteomics identifies MYCBP2 as a putative EPHB2-interacting protein
To understand the molecular logic underlying EPHB2 signalling diversity, we performed affinity purification coupled to MS (AP-MS) in order to identify EPHB2-interacting proteins, and prioritised those known to be signalling hubs. We used a stable HeLa cell line with tetracycline-inducible expression of BirA-linked EPHB2-FLAG, that we previously used to study EPHB2 signalling (Lahaie et al., 2019). To identify ephrin ligand-dependent EPHB2 protein complexes, we stimulated EPHB2-FLAG-overexpressing cells with pre-clustered Fc control or ephrinB2-Fc (eB2-Fc). We then harvested and lysed the cells, performed anti-FLAG immunoprecipitation, and used mass spectrometry to identify EPHB2 protein complexes (Fig. 1A).
To identify EPHB2-specific interactions and remove background contaminants, we performed Significance Analysis of INTeractome (SAINTexpress analysis (Teo et al., 2014)) using our EPHB2-related controls (Lahaie et al., 2019). To better visualize the changes in putative EPHB2 binding partners, we compared the average spectral counts of the identified proteins using ProHits-viz tool (Knight et al., 2017). Comparison of Fc and ephrin-B2-treated samples did not yield any significant differences. Since we applied the ligand and collected the samples on a time scale comparable to known EphB signalling dynamics, this limitation could be potentially due to autoactivation of Eph receptors when they are overexpressed, obscuring some ligand-dependent effects (Lackmann et al., 1998). However, the resulting scatter plot confirmed the presence of several known EPHB2 interactors, such as FYN and YES1 Src kinases and members of the Eph receptor family (Fig. 1B) (Banerjee et al., 2022). One of the most prominent, novel hits was the E3 ubiquitin ligase and signalling hub protein MYCBP2, and its binding partner FBXO45. FBXO45 was previously identified as a putative EPHB2 interacting protein in large-scale, cell-based interactome studies (Huttlin et al., 2021;Salokas et al., 2022).
Biochemical validation of MYCBP2 binding to EPHB2
To confirm that MYCBP2 can indeed form a molecular complex with EPHB2, we tested whether endogenous MYCBP2 co-immunoprecipitates (co-IP) with FLAG-tagged EPHB2 in HEK 293T cells. Based on differences in EPHB2 and EPHA3 interactomes, we reasoned that EPHA3 may serve as a negative control for the EPHB2-MYCBP2 association (Huttlin et al., 2021). We found that MYCBP2 coprecipitated with affinity-purified EPHB2-FLAG, but not EPHA3 (Fig. 1C). This confirmed MYCBP2 binding to EPHB2 and suggested MYCPB2 displays EPH receptor subtype specificity. Importantly, the EPHB2-MYCBP2 interaction was reduced by 24.5% and 53% following ephrin-B1 and ephrin-B2 treatment respectively in HeLa EPHB2 cells, suggesting the involvement of MYCBP2 in ephrin-B:EPHB2 signalling (Fig. 1D, E, eB1-Fc, p=0.1365; eB2-Fc, p=0.0002). To test this interaction in vivo, we performed co-IP using dissociated rat cortical neurons, which further confirmed MCYBP2 association with EPHB2 (Fig. 1F).
We next asked whether the kinase activity of EPHB2 is required for the formation of the EPHB2-MYCBP2 complex. To test this, we expressed GFP-tagged wild type (WT) EPHB2 or EPHB2 with a kinase-dead mutation (KD) in HeLa cells. We found that MYCBP2 showed comparable coprecipitation with WT and KD EPHB2 (Fig. 1G). Thus, EPHB2 kinase activity is not required for the formation of MYCBP2 - EPHB2 complexes, suggesting that MYCBP2 association with EPHB2 may be a prelude to ephrin-B-evoked EPHB2 signalling.
FBXO45 enhances the association between MYCBP2 and EPHB2
Our proteomics screen for EPHB2 interactors also identified FBXO45, the F-box protein that forms a ubiquitin ligase complex with MYCBP2 (Sharma et al., 2014;Saiga et al., 2009). Thus, we initially reasoned that FBXO45 might perform a similar role in the formation of the MYCBP2-EPHB2 complex. We first tested whether FBXO45 binds to EPHB2 by co-expressing MYC-tagged FBXO45 with EPHB2-FLAG or EPHA3-FLAG in HEK293T cells. Co-IP revealed that FBXO45 can associate with EPHB2 but not EPHA3, suggesting that EPHB2, FBXO45 and MYCBP2 could form a ternary complex (Fig. 2A). To test this idea, we co-expressed EPHB2-FLAG and GFP-MYCBP2 in the presence or absence of MYC-FBXO45 and examined co-IP efficiency between EPHB2 and MYCBP2. Interestingly, FBXO45 enhanced the interaction between EPHB2 and MYCBP2 (Fig. 2B, C, p=0.0068). Together, these data suggest that EPHB2 can form a complex with both MYCBP2 and FBXO45, and that FBXO45 increases the efficiency of MYCBP2-EPHB2 interaction.
Biochemical mapping of EPHB2-MYCBP2 interaction
To identify the MYCBP2 protein domain(s) required for the formation of the ternary complex with EPHB2 and FBXO45, we co-expressed EPHB2-FLAG and three GFP-MYCBP2 fragments in HEK293T cells. Co-IP revealed that the central region of MYCBP2 was sufficient for binding with EPHB2 (Fig. 2D, E). In addition, co-expression of MYC-FBXO45 demonstrated that the association of the central domain of MYCBP2 with EPHB2 is enhanced by FBOX45 and is consistent with the presence of an FBXO45 binding site within this MYCBP2 fragment (Fig. 2E) (Sharma et al., 2014).
To identify the domains of EPHB2 required for the formation of the tripartite complex, we took advantage of the observation that EPHA3 does not readily form a complex with MYCBP2 or FBXO45. Thus, we performed domain swapping experiments between EPHB2 and EPHA3 reasoning that placing EPHA3-specific sequences in EPHB2 would inhibit the formation of the tripartite complex. We constructed a series of FLAG-EPHB2/EPHA3 chimeras and determined whether they bound MYCBP2 in the presence of FBXO45 (Fig. 2F). Co-IP revealed that the ability of a particular chimera to associate with MYCBP2 was correlated with its association with FBXO45, in line with the MYCBP2-FBXO45 complex interacting with EPHB2. Surprisingly, we found that EPHB2-EPHA3 chimeras with an EPHA3 identity of intracellular juxtamembrane, kinase, SAM or PDZ binding domains retained the ability to associate with FBXO45 and MYCBP2, suggesting that the formation of the tripartite complex is driven by the extracellular domain and/or the transmembrane domain of EPHB2 (Fig. 2G). However, these results are also consistent with the possibility that the EPHB2 identity of the extracellular fragments could alter the conformation of the intracellular domains of EPHA3 identity, allowing the interaction with FBXO45-MYCBP2 to occur. To exclude this possibility, we created EPHB2 mutants lacking the intracellular or extracellular domains and tested their ability to complex with MYCBP2 and FBXO45 by co-IP (Fig. 2H). In these experiments only the deletion mutant lacking the intracellular domain retained its ability to form the tripartite complex. Collectively, these results argue that the combination of extracellular and transmembrane domains of EPHB2 are necessary and sufficient for formation of the MYCBP2–FBXO45–EPHB2 complex (Fig. 2I).
MYCBP2 is required for EPHB2-mediated cellular responses
Given the decrease in MYCBP2–EPHB2 association evoked by ephrin-B treatment (Fig. 1D, E), we next sought to determine whether MYCBP2 fulfils a specific function in ephrin-B:EPHB2 forward signalling. Thus, we infected EPHB2-FLAG HeLa cells using lentivirus containing CRISPR sgRNA targeting the MYCBP2 exon 6, and pooled MYCBP2CRISPR cells after puromycin selection (Desbois et al., 2018). Using EPHB2-FLAG HeLa cells carrying a stably integrated empty expression vector (CTRLCRISPR) as controls, we found that MYCBP2CRISPR led to a reduction in endogenous MYCBP2 protein levels (Fig. 3A).
To study the role of MYCBP2 in EphB signalling, we took advantage of an experimental paradigm in which exposure to ephrin-B2 evokes cytoskeletal contraction of HeLa cells expressing EPHB2 (Lahaie et al., 2019). Thus, following induction of EPHB2 expression, MYCBP2CRISPR and CTRLCRISPR cells were stimulated with pre-clustered Fc control or ephrin-B2 for 15 minutes and scored as collapsed or uncollapsed. While the proportion of collapsed cells for the two cell lines treated with Fc was similar (CTRLCRISPR, 5.4%; MYCBP2CRISPR, 5.7%; p=0.9841), ephrin-B2 treatment resulted in the collapse of 19.0% of CTRLCRISPR cells but only 12.1% of MYCBP2CRISPR cells (n=9 coverslips; Fig. 3B, C; p=0.0017). Moreover, when compared to the Fc conditions, MYCBP2CRISPR cells exhibited a less drastic change in collapse rate upon ephrin-B2 treatment (Fig. 3B, C; eB2 vs Fc: CTRLCRISPR, p<0.0001; MYCBP2CRISPR, p=0.0115).
In addition, time-lapse imaging of MYCBP2CRISPR and CTRLCRISPR cells transiently transfected with an EPHB2-GFP expression plasmid revealed a similar attenuation of ephrin-B2-induced cellular contraction (Fig. 3D, E; p=0.0268). These data argue that MYCBP2 regulates a short-term cellular response evoked by ephrin-B2:EPHB2 signalling.
To study longer-term cellular responses evoked by ephrin-B2:EPHB2 signalling, we turned to a stripe assay in which the preference of CTRLCRISPR and MYCBP2CRISPR cells for immobilized ephrin-B2 or Fc was measured. To do this, cells of either line were deposited over alternating stripes of ephrin-B2 or Fc, and stripe preference was scored for individual cells after overnight incubation. While only 33.2% of CTRLCRISPR cells resided on ephrin-B2 stripes, this proportion was significantly increased to 47.4% for MYCBP2CRISPR cells, suggesting the loss of MYCBP2 function led to a decreased repulsion from ephrin-B2 stripes (Fig. 3F, G; n=5 and 7 carpets respectively, p=0.0109). When cells were plated on Fc:Fc stripes, CTRLCRISPR and MYCBP2CRISPR cells exhibited no preference over cy3-conjugated Fc stripes (49.18% vs 51.88%, p=0.5386, images not shown). Native HeLa cells only respond to ephrin-B2 once they are made to express EphB2. Thus, our data suggest that MYCBP2 is required for EPHB2-mediated cellular responses in HeLa cells.
Loss of MYCBP2 decreases cellular levels of EPHB2 protein
The association of EPHB2 with MYCBP2 and its substrate recognition protein FBXO45 suggests that the MYCBP2 ubiquitin ligase complex could target EPHB2 for degradation, a mechanism frequently deployed to terminate transmembrane receptor signalling (Foot et al., 2017). However, the results of our cellular assays contradicted this model, and rather suggested that loss of MYCBP2 function decreased EPHB2 signalling. To further evaluate these two scenarios, we first compared EPHB2 protein levels in HeLa MYCBP2CRISPR cells and CTRLCRISPR cells by using tetracycline to induce the EPHB2 overexpression. We found that MYCBP2 loss reduced EPHB2 protein levels (Fig. 4A). To further confirm this, instead of inducing EPHB2 overexpression with tetracycline, we transfected EPHB2-FLAG plasmid into both CTRLCRISPR and MYCBP2CRISPR cells and examined EPHB2-FLAG levels after two days. As shown in Fig. 4B, C, levels of EPHB2-FLAG were significantly lowered by 26.1% in HeLa MYCBP2CRISPR cells compared to CTRLCRISPR cells (p=0.0046). We further investigated whether MYCBP2 affects EPHB2 protein turnover when cycloheximide is added to prevent new protein synthesis. EPHB2-FLAG expression was induced by tetracycline for 12 hours followed by protein synthesis inhibition with cycloheximide. We found that EPHB2 half-life was reduced in HeLa cells lacking MYCBP2 compared to control (Fig. 4D, E; 8h treatment, p=0.0474).
Ephrin ligand treatment eventually results in Eph receptor degradation, a process associated with signalling termination. We therefore asked whether ligand-mediated EPHB2 receptor degradation depends on MYCBP2. We induced EPHB2 expression in CTRLCRISPR and MYCBP2CRISPR cells, and exposed them to ephrin-B2 for a different length of time. Western blotting revealed that 4-8 hours stimulation with ephrin-B2 reduced EPHB2 levels in CTRLCRISPR cells, but this effect was more drastic in MYCBP2CRISPR cells (Fig. 4F, G). Taken together, our results are not consistent with MYCBP2 ubiquitinating EPHB2 and causing its degradation. Unexpectedly, our data indicate that MYCBP2 stabilizes EPHB2 in HeLa cells under both naïve and ligand-challenged conditions.
Loss of MYCBP2 enhances ligand-induced EPHB2 receptor ubiquitination
Ubiquitination of receptor tyrosine kinases, including Eph receptors, can herald their degradation via the proteasome and thus termination of signalling (Haglund and Dikic, 2012;Sabet et al., 2015). This model is not supported by our results, which suggest that MYCBP2 is required for EPHB2 protein maintenance. Nonetheless, we investigated whether EPHB2 receptor ubiquitination is altered in HeLa cells depleted of MYCBP2. HeLa CTRLCRISPR and MYCBP2CRISPR cells were transfected with HA-tagged ubiquitin, EPHB2 expression was induced with tetracycline, and cells were treated with ephrin-B2 for 30min. We observed that EPHB2 receptor ubiquitination was not significantly increased in CTRLCRISPR cells after this short-term ligand treatment (Fig. 4H, I CTRLCRISPR, p=0.1349). In contrast, EPHB2 ubiquitination was significantly increased in MYCBP2CRISPR cells (Fig. 4H, I MYCBP2CRISPR, p=0.0195). This effect argues against the concept that EPHB2 is a MYCBP2 ubiquitination substrate, and suggests that in the absence of MYCBP2 degradation of the EPHB2 receptor is enhanced due to increased ubiquitination.
Loss of MYCBP2 promotes EPHB2 degradation through the lysosomal pathway
EPHB2 can be degraded by either a proteasomal or lysosomal pathway depending on the cellular context (Cissé et al., 2011; Litterst et al., 2007;Fasen et al., 2008). Thus, we wanted to shed light on how EPHB2 is degraded and understand why EPHB2 degradation is enhanced by MYCBP2 loss of function. To do so, we induced EPHB2 expression in CTRLCRISPR and MYCBP2CRISPR cells and applied the S26 proteasome inhibitor MG132, or the lysosomal inhibitors BafilomycinA1 or Chloroquine. We found that MG132 did not have significant effects on EPHB2 levels in both cell types (Fig. 4J, K). However, we found that BafilomycinaA1 (BafA1) significantly increased EPHB2 protein levels in both HeLa CTRLCRISPR cells and MYCBP2CRISPR cells by 19% and 40%, respectively (Fig. 4J, K). We also observed a trend towards increased EPHB2 levels with Chloroquine (coQ) treatment in CTRLCRISPR (14%) and MYCBP2CRISPR (35%), further suggesting a role for lysosomal degradation (Fig. 4J, K). These observations suggest that the loss of MYCBP2 promotes EPHB2 receptor degradation through the lysosomal pathway, which expands our understanding of how lysosomal degradation of EPHB2 is regulated.
Loss of MYCBP2 does not decrease EPHB1, EPHB3, or EPHA3 receptor levels
The above experiments raise the question of whether MYCBP2 is a general regulator of Eph receptor stability. Since EPHA3 does not form a complex with MYCBP2, and EphA receptor levels are controlled by the proteasomal pathway (Sharfe et al., 2003;Walker-Daniels et al., 2002), we hypothesized that MYCBP2 might regulate the levels of the entire EphB receptor class. To test this, we co-transfected plasmids encoding GFP-tagged EPHB1 and HA-tagged EPHB3 into HeLa CTRLCRISPR and MYCBP2CRISPR cells. Compared to CTRLCRISPR cells, EPHB1 and EPHB3 levels were not significantly reduced in MYCBP2CRISPR cells (Fig. 4L, M, EPHB1, p=0.0588; EPHB3, p=0.4253). Likewise, FLAG-EPHA3 levels were similar in CTRLCRISPR and MYCBP2CRISPR cells (Fig. 4N, O, p=0.5369). Taken together, these data suggest that MYCBP2 specifically affects EPHB2 receptor stability.
Loss of MYCBP2 attenuates the magnitude of EPHB2 cellular signalling
EPHB2 receptor activation evokes signal transduction events such as tyrosine phosphorylation of EPHB2, activation of EPHB2 tyrosine kinase function and phosphorylation of the ERK1/2 downstream effector (Poliakov et al., 2008). We therefore measured EPHB2 and ERK1/2 phosphorylation in CTRLCRISPR and MYCBP2CRISPR cells for up to 8 hours after ephrin-B2 application (Fig. 5A). P-EPHB2 (pY20) and p-ERK1/2 signals were normalised to GAPDH and ERK1/2, respectively. In HeLa CTRLCRISPR cells, pTyr-EPHB2 response reached a plateau after 1-2 h treatment and remained up 8 hours post-stimulation (Fig. 5A-C). On the contrary, ephrin-B2-evoked phosphorylation of EPHB2 was reduced in MYCBP2CRISPR cells with quantitative results showing a significant reduction by 8 hours of treatment (Fig. 5A, C; 8h p=0.0331). We were also able to detect significantly lower p-ERK1/2 levels in MYCBP2CRISPR cells relative to CTRLCRISPR cells, although activation of ERK1/2 by ephrin-B2 was variable (Fig. 5D. 4h, p=0.0494; 8h, p=0.0078; n=6). We again noted significantly enhanced EPHB2 degradation in MYCBP2CRISPR cells (Fig. 5E; 8h, p=0.0437). This decrease was also observed when CTRLCRISPR and MYCBP2CRISPR cells were treated with ephrin-B1 (data not shown).
Next, we asked whether the decrease in ligand-evoked EPHB1/2 and ERK1/2 activation in MYCBP2CRISPR cells reflects a requirement for MYCBP2 in the EPHB2 signalling cascade per se, or whether it is explained by decreased EPHB2 protein levels caused by MYCBP2 loss. We thus normalized p-EPHB2 and p-ERK1/2 signal to the levels of EPHB2 protein at all time points, which revealed that kinetics and magnitude of EPHB2 and ERK1/2 phosphorylation are similar between CTRLCRISPR and MYCBP2CRISPR cells (Fig. 5F, G). Although these data argue against a direct role of MYCBP2 in the early events of EPHB2 signalling, they nevertheless indicate that MYCBP2 loss results in marked attenuation of EPHB2 activation and its downstream pERK1/2 signalling in line with decreased cellular responses to ephrin-B2.
Exogenous Fbxo45 binding domain of MYCBP2 disrupts the EPHB2-MYCBP2 interaction
Previous studies showed that FBXO45 binds to the FBD1 domain of MYCBP2, and exogenous FBD1 overexpression can disrupt the FBXO45-MYCBP2 association (Sharma et al., 2014). We thus tested whether FBD1 overexpression can interfere with the formation of the EPHB2-MYCBP2 complex (Fig. 6A). Indeed, the expression of GFP-FBD1 wild-type (WT) in HEK cells expressing EPHB2-FLAG and MYC-FBXO45 reduced binding between EPHB2 and MYCBP2 (Fig. 6B). This effect was not observed with a GFP-FBD1 mutant (mut) fragment that harbours three point mutations that inhibit binding to FBXO45. Reduction of the EPHB2-MYCBP2 interaction was also observed in cells that were not overexpressing FBXO45 (Fig. 6C). Thus, exogenous FBD1 can specifically disrupt the EPHB2-MYCBP2 association. This suggests that FBXO45 binding to MYCBP2 may facilitate for formation of the MYCBP2-EPHB2 complex.
FBD1 expression impairs EPHB2-mediated neuronal responses
To determine whether the EPHB2-MYCBP2 interaction is required for EPHB2 function, we disrupted binding via FBD1 ectopic expression in cell lines or primary neurons, and studied cellular responses to ephrin-B treatment. We introduced GFP-FBD1 mut or GFP-FBD1 WT into HeLa cells with inducible EPHB2 expression and cultured these cells on ephrin-B2 and Fc stripes. Following overnight culture, only 19.2% of cells expressing GFP-FBD1 mut resided on ephrin-B2 stripes. In contrast, 28.6% of cells expressing GFP-FBD1 WT were found on ephrin-B2 stripes indicating that FBD1 expression can dampen ephrin-B2:EPHB2 mediated cell repulsion (Fig. 6D, E; p=0.0107). In contrast, cells expressing GFP-FBD1 mut or GFP-FBD1 WT displayed no preference for either one of Fc:Fc control stripes (49.63% vs 49.85%, p=0.9560, images not shown).
Since EPHB2 is expressed in the embryonic chicken spinal cord, we performed in ovo electroporation to introduce plasmids encoding GFP-FBD1 mut or GFP-FBD1 WT into the spinal cords of 2.5-day old embryonic chickens (Hamburger-Hamilton; (HH) Stage 15-17; (Hamburger and Hamilton, 1951;Luria et al., 2008)). Two days later spinal cords were dissected, divided into explants, cultured overnight on alternating stripes containing ephrin-B2 or Fc, and axonal GFP signal present over ephrin-B2 versus Fc stripes was determined. When explants expressed GFP-FBD1 mut (negative control), we observed 26.5% of axonal GFP signal resided on ephrin-B2 stripes (Fig. 7A, B). In contrast, we found a significant increase of 32.2% of neurites expressing FBD1 WT on ephrin-B2 stripes (Fig. 7A, B; p=0.0410). Thus, FBD1 expression impairs long-term repulsive responses to ephrinB2-EPHB2 signalling in spinal explants.
To study short-term neuronal responses to ephrin-Bs, we turned to mouse hippocampal neurons and a growth cone collapse assay (Srivastava et al., 2013). Here, we electroporated GFP-FBD1 mut or GFP-FBD1 WT expression plasmids into dissociated hippocampal neurons and treated them with pre-clustered Fc, ephrin-B1 or ephrin-B2 for one hour (Fig. 7C). Neurons expressing GFP-FBD1 mut showed 33.9% growth cone collapse with Fc treatment, while Ephrin-B1 and Ephrin-B2 treatment elicited significant increases to 57.9% and 53.3% collapse, respectively (Fig. 7C, D). Importantly, both ephrin-B ligands significantly increased the frequency of growth cone collapse with respect to Fc treatment when FBD1 mut was expressed (Fig. 7D; p<0.0001, and p=0.0008). In contrast, this effect was either significantly reduced or less evident in neurons expressing GFP-FBD1 WT: with only 48.1% and 57.5% growth cones being collapsed by ephrin-B1 and ephrin-B2, respectively, compared to 41.7% being collapsed by Fc treatment (Fig. 7D; p=0.4357 and p=0.0065, respectively). Together, these data indicate that impairing MYCBP2 function via FBD1 expression disrupts ephrin:B-EPHB2 signalling in axonal guidance.
C. elegans ephrin receptor VAB-1 interacts genetically with known RPM-1/MYCBP2 binding proteins
Next, we sought to test genetic interactions between an ephrin receptor and the MYCBP2 signalling network using an in vivo animal model. To do so, we turned to C. elegans which has a sole Eph family receptor called VAB-1(George et al., 1998), and a single MYCBP2 ortholog called RPM-1(Grill et al., 2016). Previous studies have shown that RPM-1/MYCBP2 is required to terminate axon outgrowth with failed termination defects in rpm-1 mutants arising due to impaired growth cone collapse(Schaefer et al., 2000;Borgen et al., 2017). Prior studies have also demonstrated that RPM-1 functions via numerous downstream pathways, and that RPM-1 binding proteins display genetic enhancer interactions (Grill et al., 2007;Grill et al., 2012;Tulgren et al., 2014;Baker et al., 2014). Given our cell-based biochemical results showing that MYCBP2 binds EphB2, we opted to evaluate two known RPM-1 binding proteins for genetic interactions with the VAB-1 Ephrin receptor. 1) The Rab GEF GLO-4 is an RPM-1 binding protein that is orthologous to mammalian SERGEF. GLO-4 functions via the GLO-1/RAB32 small GTPase and is not involved in RPM-1 ubiquitin ligase activity (Grill et al., 2007;Sharma et al., 2014)(Fig 8A). 2) We also tested the C. elegans FBXO45 ortholog, FSN-1, which is the F-box protein of the RPM-1 ubiquitin ligase complex (Liao et al., 2004;Sharma et al., 2014;Baker and Grill, 2017;Desbois et al., 2022) (Fig 8A).
To evaluate genetic interactions between the VAB-1 Eph receptor and RPM-1/MYCBP2 binding proteins, we used the PLM mechanosensory neurons of C. elegans. There are two PLM neurons with one on either side of the body. Each PLM neuron extends a primary axon that grows anteriorly and terminates at a precise location before the cell body of another mechanosensory neuron called ALM (Fig. 8B). Consistent with previous studies (Mohamed and Chin-Sang, 2006), vab-1/EphR loss-of-function (lf) mutants presented moderate failed termination defects where the PLM axon overextends beyond the ALM cell body (Fig. 8B).
Quantitation indicated that this phenotype occurred at a low, but significant frequency (Fig. 8C). We then examined the genetic relationship between vab-1 and the RPM-1 binding protein glo-4/SERGEF. In vab-1; glo-4 double mutants, we observed both moderate overextension defects and more severe failed termination defects where the axon hooked ventrally (Fig. 8B). The frequency of both types of axon termination defects were significantly enhanced in vab-1; glo-4 double mutants (Fig. 8B, C). Overextension defects were significantly rescued by transgenic expression of VAB-1 in vab-1; glo-4 double mutants (Fig. 8C). We also tested the relationship between the F-box protein fsn-1/FBXO45 and vab-1, and observed enhanced defects in vab-1; fsn-1 double mutants (Fig. 8D). Finally, we evaluated vab-1; rpm-1 double mutants and did not observe a significant change in phenotypic severity or penetrance compared to rpm-1 single mutants (Fig. 8E).
Collectively, these genetic experiments support two conclusions. 1) Enhanced defects in vab-1; glo-4 and vab-1; fsn-1 double mutants indicate that the VAB-1 Ephrin receptor functions in parallel to known RPM-1 binding proteins to facilitate axon termination. These results are consistent with the VAB-1 Ephrin receptor functioning within the RPM-1 signalling network. 2) Lack of suppression in vab-1; fsn-1 and vab-1; rpm-1 double mutants indicates that VAB-1 is not inhibited or degraded by the RPM-1/FSN-1 ubiquitin ligase complex. Thus, our findings in C. elegans are consistent with our biochemical studies in mammalian cells that demonstrate MYCBP2 stabilizes EPHB2 rather than degrading it. We note that because C. elegans contains only a single Ephrin receptor, our findings with C. elegans do not necessarily pertain specifically to EPHB2. Nonetheless, these results add further evidence indicating there are substantial genetic interactions between known MYCBP2 binding proteins and Eph receptors.
Our MS-based proteomics efforts to identify EPHB2 interacting proteins yielded MYCBP2, a signalling hub and ubiquitin ligase that is functionally linked to many of the cellular processes also mediated by EPHB2, including cellular growth, proliferation, synapse formation, and axon development. Our experiments argue against EPHB2 being a MYCBP2 ubiquitination substrate. Instead, we envisage a model where MYCBP2 controls EPHB2 signalling indirectly by preventing its lysosomal degradation and maintaining EPHB2 protein levels sufficiently high to mediate efficient cellular and axonal growth cone repulsion from ephrin-B ligands. The interaction between MYCBP2 and EPHB2 may allow the coupling EPHB2 to fundamental cellular processes that control growth, proliferation, and survival. Here, we discuss the molecular logic of MYCBP2 and EPHB2 association in the context of Eph receptor signalling, its potential implications for neural development and diversification of EphB2 signalling.
Functional significance of MYCBP2-EPHB2 complex formation
Our biochemical experiments validate the formation of a MYCBP2-EPHB2 complex and suggest that this association is decreased following ligand application. Because PHR proteins like MYCBP2 are large signalling hubs, MYCBP2 association with EPHB2 might sterically hinder the formation of EPHB2 multimers and clusters necessary for signalling. Thus, ligand-induced dissociation of the MYCBP2-EPHB2 complex could be a prelude to signalling. Our findings indicate that MYCBP2 association with EPHB2 is enhanced by FBXO45, a subunit of the MYCBP2/FBXO45 complex that mediates ubiquitination substrate binding. Previous work showed that MYCBP2 functions to polyubiquitinate specific protein substrates, targeting them for degradation and inhibition (Crawley et al., 2019;Han et al., 2012;Nakata et al., 2005;Desbois et al., 2022). We considered the possibility that the ubiquitin ligase activity of MYCBP2 is important for the termination of EPHB2 signalling, but several lines of evidence argue against this idea: (1) the application of ephrin-B2 ligand results in the dissociation of the MYCBP2-EPHB2 complex (Walker-Daniels et al.), (2) loss of MYCBP2 function results in decreased cellular levels of EPHB2 protein, and (3) impairing MYCBP2 increases ligand-stimulated EPHB2 ubiquitination. In addition, our quantification of EPHB2 signalling suggested that decreased EPHB2 and ERK phosphorylation seen in MYCBP2-deficient cells can be accounted for by the decrease in EPHB2 receptor levels. Thus, a more plausible model that is consistent with our results is that MYCBP2 association with EPHB2 protects EPHB2 from turnover by lysosome-mediated degradation and prevents EPHB2 ubiquitination by the action of unidentified ubiquitin ligases. Interestingly, studies in C. elegans have shown that RPM-1/MYCBP2 regulates lysosome biogenesis via the GLO-4/GLO-1 pathway (Grill et al., 2007). While this established a link between MYCBP2 signalling and the endo-lysosomal degradation system, our results now indicate that MYCBP2 can also influence turnover of EPHB2 via lysosomal degradation. Consistent with these findings, we observed that the Eph receptor VAB-1 functions in parallel genetic pathways with the known RPM-1 binding protein GLO-4/SERGEF that regulates lysosome biogenesis.
The “protective” effect of MYCBP2 vis-à-vis EPHB2 ubiquitination and lysosomal degradation might be secondary to effects on other aspects of EPHB2 signalling that we have not explored experimentally. For example, since MYCBP2 is a signalling hub with multiple substrates and binding proteins, it could bring EphB2 into close proximity to components of the MYCBP2 signalling network integrating cell-cell communication via ephrin:Eph signals with MYCBP2 intracellular signalling to influence fundamental cellular processes. One example of this is suggested by a recent study demonstrating a link between EPHB2 and mTOR-mediated cell growth signalling pathways (Gagne et al., 2021). One potential mechanism could involve Ephrin binding-induced dissociation of the EPHB2-MYCBP2 interaction, allowing MYCBP2 ubiquitination of the TSC complex that regulates mTOR function (Han et al., 2012).
The formation of a MYCBP2-EPHB2 complex that includes other signalling receptors could also explain the curious result that the extracellular domain of EPHB2 is critical for MYCBP2 association. The finding would be consistent with the extracellular domain of EPHB2 interacting with the extracellular domains of other receptors, whose intracellular domains are linked more directly with MYCBP2 and FBXO45. An alternative explanation may be a non-classical extracellular MYCBP2-EPHB2 interaction similar to the one proposed for the extracellular domain of N-Cadherin and FBXO45 (Na et al., 2020).
MYCBP2 and EPHB2 functions in the developing nervous system
Our proteomic, biochemical, and genetic experiments indicate that MYCBP2 and EPHB2 function in the same pathway. This is also supported by the striking similarity of MYCBP2 and EPHB2 loss of function phenotypes in the mouse nervous system. In the context of developing neuronal connections, EphB2 and Mycbp2/Phr1 mouse mutants display similar axon guidance phenotypes such as abnormal limb nerve trajectories by motor axons, defective growth cone crossing of the midline at the level of the optic chiasm and decreased connectivity between the two cortical hemispheres through the corpus callosum (Lewcock et al., 2007;Luria et al., 2008;Williams et al., 2003;Henkemeyer et al., 1996;D’Souza et al., 2005). Both proteins also function in synaptic development and their loss of function leads to decreased numbers of synapses and altered synapse morphology (Wan et al., 2000;Bloom et al., 2007;Dalva et al., 2000;Zhen et al., 2000). Furthermore, outside the nervous system, both are involved in a variety of cancers with recent evidence linking them to esophageal adenocarcinoma and c-MYC-dependent control of cell proliferation (Venkitachalam et al., 2022;Han et al., 2012;Genander et al., 2009). Our proteomic, biochemical, and cellular experiments together with genetic interaction studies in C. elegans now provide a new framework in which to consider phenotypic and disease links between EPHB2 and MYCBP2.
Importantly, the biomedical relevance of our findings are heightened by a recent study that identified genetic variants in MYCBP2, which cause a neurodevelopmental disorder termed MYCBP2-related Developmental delay with Corpus callosum Defects (MDCD) (Al-Abdi et al., 2022). MDCD features defective neuronal connectivity including a hypoplastic or absent corpus callosum, neurobehavioral deficits including intellectual disability and epilepsy, and abnormal craniofacial development. This constellation of comorbidities in MDCD closely resembles some of the phenotypes observed in mice with deficient EPHB2 signalling. Given our finding that MYCBP2 loss reduces EPHB2 levels and influences Eph receptor effects on axon development, the question of whether EPHB2 expression levels are normal in MYCBP2 patients remains pertinent.
In conclusion, our study has revealed numerous biochemical and genetic links between MYCBP2 and EPHB2. Our findings indicate that the MYCBP2/FBXO45 complex protects EPHB2 from degradation, and these are functionally integrated signalling players with an evolutionarily conserved role in axonal development. Future studies will be needed to address how the EPHB2-MYCBP2 interaction affects nervous system development in mammals in vivo and to identify further regulators of EPHB2 degradation. Additionally, another idea worthy of closer examination in the future is the possibility that MYCBP2 signalling could provide routes through which EPHB2-initiated signals access numerous fundamental cellular functions.
We thank Dominic Fillion for assistance with microscopy, Denis Flaubert and the IRCM Proteomics Core for proteomics analysis, Gu Hua for HA-Ubiquitin plasmid, Sylvie Lahaie, Louis-Philippe Croteau and Farin B. Bourojeni for discussions, and Meirong Liang for technical support. This work was supported by project grants from the Canadian Institutes of Health Research (PJT-162225, MOP-77556, PJT-153053, and PJT-159839) to AK and a grant from the National Institutes of Health (R01 NS072129) to BG. CC received IRCM Jacques-Gauthier Scholarship. S. L. B. is a Fonds de recherche du Québec – Santé (FRQS) scholar. SSP is funded by IRCM Jean-Coutu Scholarship.
All authors: Conceptualisation, Writing – Editing and Review.
CC: Methodology, Investigation, Data Curation, Writing – Original Draft, Visualisation.
SLB: Methodology, Investigation, Data Curation, Visualisation, Supervision, Writing – Original Draft.
SSP: Methodology, Investigation, Data Curation, Visualisation. XZ: Methodology, Investigation.
DCW: Methodology, Investigation.
MD: Methodology, Investigation, Resources, Data Curation, Visualisation, Supervision, Writing – Original Draft.
BG: Supervision, Funding Acquisition, Writing – Original Draft.
AK: Writing – Original Draft, Supervision, Project Administration, Funding Acquisition.
Conflict of interest
The authors declare no competing financial interests.
Materials and Methods
All animal experiments were carried out in accordance with the Canadian Council on Animal Care guidelines and approved by the IRCM Animal Care Committee. Fertilized chicken eggs (FERME GMS, Saint-Liboire, QC, Canada) were incubated at 38-39°C and staged according to Hamburger and Hamilton (HH) (Hamburger and Hamilton, 1951). C57BL/6 mice were used for hippocampal neuron collapse assay. Timed mating vaginal plug was designated as E0.5.
HeLa and HEK293T cells were maintained in DMEM (Thermo Fisher Scientific, #11965092) supplemented with 10% Fetal Bovine Serum (FBS; Wisent Bioproducts, #080-150) and 1% Penicillin/Streptomycin (Wisent Bioproducts, #450-200-EL) at 37°C with 5% CO2. Tetracycline inducible HeLa EPHB2-FLAG cells were generated by transfecting Flp-In T-REx HeLa cells with EPHB2-BirA*-FLAG expression plasmid using Lipofectamine 3000 (Invitrogen, #L3000015) followed by hygromycin selection (200µg/ml). Stable but not clonal CTRLCTRISPR or MYCBP2CRISPR HeLa EPHB2-FLAG cell lines were generated by infecting cells with packaged lentivirus, followed by puromycin selection (1µg/ml). Lentivirus particles were packaged using MYCBP2 sgRNA CRISPR plasmid designed to target the MYCBP2 exon6 (pLentiCRISPR2-sgMYCBP2) and pLentiCRISPR empty vector was used as Ctrl CRISPR. EPHB2-FLAG overexpression was induced using 1µg/ml tetracycline simultaneously with cell starvation in DMEM supplemented with 0.5%FBS, 1% penicillin/streptomycin for 12-20h. Prior to cell stimulation, Fc control (Millipore, #401104), ephrinB1-Fc (R&D, #473-EB) or ephrinB2-Fc (R&D, #496-EB) were pre-clustered using goat anti-human Fc IgG (Sigma, #I2136) in 4:1 ratio for 30min.
Affinity purification - mass spectrometry
HeLa EPHB2-FLAG cells cultured in DMEM supplemented with 0.5% FBS and 1µg/ml tetracycline in 15cm cell culture plates for 20h, were treated with pre-clustered Fc control or ephrinB2-Fc for 15min. After treatment, cells were washed twice with PBS and lysed using a lysis buffer (50mM Tris, pH7.4; 150mM NaCl; 1% NP-40) supplemented with protease (Roche, #11836153001) and phosphatase inhibitors (Roche, #04906837001). The lysates were collected in 1.5ml Eppendorf tubes and centrifuged at 13,000rpm for 15min at 4°C. Supernatants were transferred to new tubes with prewashed anti-FLAG agarose beads (Sigma, #A2220) and incubated on a rotator overnight at 4°C. The following day, beads were washed four times using 50mM Ammonium Bicarbonate. The on-bead proteins were diluted in 2M Urea/50mM ammonium bicarbonate and on-bead trypsin digestion was performed overnight at 37°C. The samples were then reduced with 13 mM dithiothreitol at 37°C and, after cooling for 10min, alkylated with 23 mM iodoacetamide at room temperature for 20min in the dark. The supernatants were acidified with trifluoroacetic acid and cleaned from residual detergents and reagents with MCX cartridges (Waters Oasis MCX 96-well Elution Plate) following the manufacturer’s instructions. After elution in 10% ammonium hydroxide /90% methanol (v/v), samples were dried with a Speed-vac, reconstituted under agitation for 15min in 12µL of 2%ACN-1%FA and loaded into a 75μm i.d. × 150 mm Self-Pack C18 column installed in the Easy-nLC II system (Proxeon Biosystems). Peptides were eluted with a two-slope gradient at a flowrate of 250nL/min. Solvent B first increased from 2 to 35% in 100min and then from 35 to 80% B in 10min. The HPLC system was coupled to Orbitrap Fusion mass spectrometer (Thermo Scientific) through a Nanospray Flex Ion Source. Nanospray and S-lens voltages were set to 1.3-1.7 kV and 60 V, respectively. Capillary temperature was set to 225°C. Full scan MS survey spectra (m/z 360-1560) in profile mode were acquired in the Orbitrap with a resolution of 120,000 with a target value at 3e5. The 25 most intense peptide ions were fragmented in the HCD collision cell and analyzed in the linear ion trap with a target value at 2e4 and a normalized collision energy at 29 V. Target ions selected for fragmentation were dynamically excluded for 15 sec after two MS2 events.
Mass spectrometry data analysis
The peak list files were generated with Proteome Discoverer (version 2.3) using the following parameters: minimum mass set to 500 Da, maximum mass set to 6000 Da, no grouping of MS/MS spectra, precursor charge set to auto, and minimum number of fragment ions set to 5. Protein database searching was performed with Mascot 2.6 (Matrix Science) against the UniProt Human protein database. The mass tolerances for precursor and fragment ions were set to 10 ppm and 0.6 Da, respectively. Trypsin was used as the enzyme allowing for up to 1 missed cleavage. Cysteine carbamidomethylation was specified as a fixed modification, and methionine oxidation as variable modification. Data interpretation was performed using Scaffold (version 4.8) and further statistical analysis was performed through ProHits integrated with SAINT (Significance Analysis of INTeractome) (Liu et al., 2010).
Cell lysis, co-immunoprecipitation and Western blotting
Cells were washed with PBS, lysed with RIPA buffer (50mM Tris, pH 7.4; 150mM NaCl; 1% NP-40; 0.1% SDS) supplemented with protease and phosphatase inhibitors. For co-IP experiments, cells were lysed with co-IP buffer (50mM Tris, pH 7.4; 150mM NaCl; 0.1% NP-40) with protease and phosphatase inhibitors. Cell lysates were centrifuged at 13,000 rpm for 20min at 4°C, then the supernatants were collected, and total protein concentrations were quantified using BCA kit (Thermo Scientific, #23225). For FLAG co-IP, 500-1000µg of total protein was incubated with 20-40 µl of prewashed anti-FLAG agarose beads (Sigma, #A2220) for 3h at 4°C. After incubation, the beads were centrifuged at 2,600rpm for 1min at 4℃ and washed three times with the co-IP buffer. The beads were resuspended in 2xSDS-PAGE loading buffer (5x loading buffer: Tris, 150mM, pH 6.8; SDS, 10%; Glyercol, 30%; b-Mercaptoethanol, 5%; Bromophenol Blue, 0.02%). For western blotting, proteins were separated on 6-10% SDS-PAGE gels and transferred to methanol pre-activated PVDF membranes (Millipore, #IPVH00010). For MYCBP2 blots, gels were wet transferred overnight at 4°C using 33V. Membranes were incubated in blocking buffer (TBST: 20mM Tris, pH 7.6; 150mM NaCl; 0.1% Tween 20; 5% skim milk) for 1h at room temperature, followed by primary antibody incubation (1-2h at room temperature or overnight at 4°C) and corresponding secondary antibody incubation (1h at room temperature). Primary antibodies were: rabbit polyclonal anti-MYCBP2 (Abcam, #ab86078; RRID:AB_1925230), rabbit anti-pERK1/2 (Thr202/Tyr204; Cell Signaling Technology, #9101; RRID:AB_331646), rabbit anti-ERK1/2 (Cell Signaling Technology, #9102; RRID:AB_330744), goat polyclonal anti-EPHB2 (R&D Systems, #AF467; RRID:AB_355375), mouse monoclonal anti-Actin (Sigma-Aldrich, #A5441; RRID:AB_476744), mouse monoclonal anti-pTyr (PY20; Santa Cruz Biotechnology, #sc-508; RRID:AB_628122), mouse monoclonal anti-GAPDH (Santa Cruz Biotechnology, #sc-47724; RRID:AB_627678), mouse monoclonal anti-FLAG-HRP (Sigma-Aldrich, #A8592; RRID:AB_439702), rabbit monoclonal anti-HA (Cell Signaling Technology, #3724; RRID:AB_1549585), mouse monoclonal anti-MYC (Santa Cruz Biotechnology, #sc-40; RRID:AB_627268), rabbit polyclonal anti-GFP (Thermo Fisher Scientific, #A-11122; RRID:AB_221569). Secondary antibodies were: Donkey anti-Goat HRP (Jackson, 705-035-003), Donkey anti-Mouse HRP (Jackson, 715-035-151), Donkey anti-Rabbit HRP (Jackson, 711-035-152). After three washes with TBST, membranes were incubated with ECL reagent (Cytiva, RPN2106) for 1min and chemiluminescence signal was acquired using film or Bio-Rad ChemiDoc Imaging machine. Band intensity was quantified using ImageJ or Bio-Rad Image Lab software.
HeLa cell collapse assay
HeLa-EPHB2 CTRLCRISPR and MYCBP2CRISPR cells were seeded on glass coverslips (Electron Microscopy Sciences, #7223101) in 24 well plates at a density of 20 000 cells/well. After 24 hours, cell media was changed for DMEM supplemented with 0.5% FBS, 1% P/S and 1µg/ml tetracycline to starve the cells and induce EPHB2 expression for 16-20 hours. Cells were then stimulated with 1.5µg/ml of pre-clustered Fc control or ephrin-B2-Fc for 15min. Cells were fixed with 3.2% paraformaldehyde (Lewcock et al.), 6% sucrose in PBS for 12-15min. Nuclei were stained with DAPI and F-actin with Phalloidin Alexa Fluor 568 conjugate (Thermo Fisher, #A12380). Images were acquired using Zeiss LSM710 confocal microscope and 20x objective. Fully rounded cells are scored as collapsed cells.
For time lapse imaging experiments, CTRLCRISPR or MYCBP2CRISPR cells were plated on Poly-D-Lysine coated glass bottom 35mm dishes (MATTEK, #P35GC-1.5-10-C) at a density of 300,000 cells/dish. The next day, cells were transfected with 1.5µg of EPHB2-GFP plasmid for 4-5 h, using lipofectamine 3000 in opti-MEM (ThermoFisher, #31985070), and then media was changed for DMEM supplemented with 0.5% FBS. The following day, the images were acquired under Zeiss Spinning Disk Microscope using a 20x objective. During the imaging, the cells were maintained at 37°C with 5% CO2. Pre-clustered ephrinB2-Fc was added to a final concentration of 2µg/mL directly into the dishes at the beginning of each experiment. The images were acquired every minute for 1h.
HeLa cell stripe assay
Alternative ephrin-B2-Fc or Fc stripes were prepared using silicon matrices with a micro-well system (Poliak et al., 2015). HeLa EPHB2 CTRLCRISPR and MYCBP2CRISPR cells, or HeLa EPHB2 cells transiently transfected with GFP-FBD1 wild-type or GFP-FBD1 mutant (GRR/AAA: G2404A, R2406A, R2408A), were cultured with tetracycline for 20h, trypsinized and plated on stripes (∼10 000 cells per carpet); see (Desbois et al., 2018) for specific sequences. The next day, cells were fixed with paraformaldehyde (3.2% PFA, 6% sucrose in PBS), and stained with DAPI and Phalloidin Alexa Fluor 488; or DAPI, Rabbit anti-GFP (Thermo Fisher Scientific, # A-11122; RRID:AB_221569) and Phalloidin Alexa Fluor 647 (Abcam, #ab176759). Images were acquired using Zeiss LSM710 confocal microscope and 20x objective (three vision fields for each carpet). Cells on or off the ephrin-B2 stripes were decided by the nuclei localization of the cells.
HeLa EphB2 CTRLCRISPR and MYCBP2CRISPR cells were seeded in 6-well plates at a density of 0.5 million cells per well. Next day, cells were transfected with 1.2µg HA-Ubiquitin (gift of Gu Hua) using lipofectamine 3000 (2µl/well) for 4-5h in opti-MEM (ThermoFisher, #31985070), then media was changed to DMEM with 10%FBS. The following day, EPHB2 expression was induced using 1µg/ml tetracycline in DMEM with 0.5% FBS for 12h followed by 2µg/ml Fc or eB2-Fc treatment for 30min. After IP using anti-FLAG beads, precipitates were eluted with 2xSDS loading buffer, resolved using 8% SDS gel and transferred onto PVDF membranes. Membranes were blocked in 5% milk following an incubation with anti-HA antibody (Cell Signaling Technology, #3724) to detect EPHB2 ubiquitination levels. The membrane was then stripped using mild stripping buffer (1L: 15g glycine, 1g SDS, 10mL Tween 20, pH 2.2) and probed with anti-FLAG antibody to reveal EPHB2 levels.
Lysosome and proteasome inhibition
HeLa EPHB2 CTRLCRISPR and MYCBP2CRISPR cells were seed in 6-well plates at a density of 0.5 million cells per well. Next day, EPHB2 overexpression was induced using 1µg/ml tetracycline in DMEM with 10% FBS for 16h, followed by 26S proteasome inhibitor (MG132, 50 µM, Sigma, #474790) or lysosome inhibitor treatment (BafilomycinA1, 0.2 µM, Sigma, #B1793; Chloroquine, 50 µM, Tocris, #4109) for 6hr.
Chick in ovo electroporation
Fertilized eggs (FERME GMS, Saint-Liboire, QC) were incubated in an incubator (Lyon Technologies, model PRFWD) at 39°C with a humidity level of around 40%-60% according to standard protocols. At HH stage.15-17, chick embryo spinal neural tubes were electroporated with expression constructs (TSS20 Ovodyne electroporator at 30V, 5 pulses, 50ms wide, 1000ms interval). Following electroporation, eggs were sealed with double layer of parafilm (Pechiney Plastic Packaging Company) and incubated till HH stage 24-26.
Chick spinal explants stripe assay
Alternative eprhinB2-Fc/Fc stripes were prepared using silicon matrices with a micro-well system and pre-coated with laminin. At HH stage 24-26, chick embryos were harvested, and the lumbar part neural tubes were dissected with tungsten needles (World Precision Instruments) in MN medium (20ml motor neuron medium: 19.2ml Neurobasal medium, 400µl B27 supplement, 200µl 50mM L-glutamic acid, 200µl 100xP/S antibiotics,73mg L-glutamine). The lumbar neural tube was then cut into around 20 explants and the explants were plated on stripes (a 1cmx1cm square covers the whole stripe area). After overnight incubation, the explants were fixed with 4% PFA for 12min at 37°C, washed once with PBS, and incubated with blocking buffer, primary antibodies, and secondary antibodies. After three times PBS washes, the explants on stripes were mounted and explants with extensions were imaged using LSM710 (at least three explants per carpet). The percentage of branches on ephrinB2 stripes was calculated using the total length of branches localized on ephrinB2 stripes divided by the total length of the branches of the explant.
Dissociated mouse hippocampal neuron culture and electroporation
Primary hippocampal neurons were cultured from wild-type C57BL/6 mice at embryonic day 16–18 (E16-18). The hippocampi were dissected out and collected in 4.5ml dissection buffer (calcium-and magnesium-free Hank’s BSS: 500 ml distilled water (Gibco, #15230162), 56.8 ml 10xHBSS (Gibco, #14185052), 5.68ml 1M HEPES (Gibco, #15630080), 2.84ml HyClone (Thermo Scientific, #SV30010)). Hippocampi were added with 0.5ml 2.5%Tyrpsin and incubated at 37°C for 13-15min. After 5 times of thorough wash with dissection buffer, hippocampal neurons were dissociated in 0.8ml DMEM (Gibco, #11965118) with 10% FBS (Wisent Bioproducts, 080-150) by pipetting 10 times up and down. Then cell numbers were counted and desired number of neurons were directed for electroporation. After spin down at 2,000 rpm for 2min, dissociated hippocampal neurons (1×106/condition) were resuspended with 100µl homemade nucleofection solution, mixed with 5ug of DNA, and transferred into the aluminum cuvettes (AMAXA/Lonza). Electroporation was achieved by Nucleofector I (AMAXA/Lonza) using program O-05 (Mouse CNS neurons). 1ml Plating Medium (PM; 500ml MEM (Sigma, #M4655), 17.5ml 20% Glucose (Sigma, #G8270), 5.8ml 100mM pyruvate (Sigma, #P2256), 58ml heat-inactivated horse serum (Thermo Scientific, #26050088)) was added to the cuvettes immediately, and desired number of neurons were plated on 1mg/ml Poly-L-Lysine (Sigma, #P2636) coated coverslips in 12-well plate. At 1 day in vitro (DIV1), medium was replaced with Neuron growth and maintenance medium (NBG; 500ml Neurobasal medium (Gibco, #21103049), 10ml SM1 neuronal supplement (Stemcell, #05711), and 1.25ml GlutaMAX-I (Gibco, #35050)).
Hippocampal neuron growth cone collapse assay
Electroporated hippocampal neurons were cultured on glass coverslips (18 mm, 100 thousand neurons/well in a 12-well plate). At DIV2, the neurons were treated with pre-clustered ephrinB1-Fc, ephrinB2-Fc or Fc control in NBG medium at a final concentration of 2µg/ml for 60min at 37°C. After treatment, neurons were fixed with paraformaldehyde (3.2% PFA,6% sucrose in PBS) for 12min, followed by two PBS washes, and blocked with Blocking Buffer (PBS, 0.15% TritonX-100, 2% FBS) for 60min at room temperature. Neurons were then incubated with rabbit anti-GFP antibody (1:5000, Thermo Fisher Scientific, #A11122; RRID:AB_221569) in Blocking Buffer for 90min at room temperature. Followed by three PBS washes, neurons were incubated with DAPI, donkey anti-rabbit IgG Alexa Fluor 488 conjugate (1:1000, Jacksonimmuno, #711545152) and Phalloidin Alexa Fluor 568 conjugate (1:300, Invitrogen, #A12380) in Blocking Buffer for 60min. Neurons on coverslips were then washed and mounted on microscope slides (Fisherbrand, #1255015). Collapsed growth cones were scored using followed criteria.
Collapsed hippocampal neuron growth cone quantification
Neuron selection: only neurons with moderate GFP staining (neurons with strong GFP signalling were excluded) and only neurons with more than three branches were scored. Branch selection: branches shorter than the diameter of the neuron cell body were excluded; branches intermingling with others were excluded. Collapsed growth cone: growth cones of a fan shape are scored as full, growth cones with a width smaller than that of the branches are scored as collapsed, and growth cones’ size in between are scored as “hard to tell”. The collapse rate was calculated using collapse growth cone numbers divided by the total growth cone numbers.
Microscopy and imaging
HeLa cell collapse assay images were acquired using Leica DM6 or Zeiss LSM710 confocal microscopy. Stripe assay and growth cone collapse assay images were acquired with a Zeiss LSM710 or LSM700 confocal microscope.
Quantification and statistical analysis
All cell counts, collapsed/uncollapsed visualization and explant neurite length measurements were performed with ImageJ 2.9.0 (Schindelin et al., 2012). All numbers are illustrated in figure legends. In western blotting, each n represents one independent experiment; in neuron growth cone collapse assay, each n represents one independent experiment with neurons pooled from multiple embryos. All data statistical analyses were performed using GraphPad Prism 9.1.1. Test methods and p values were described in figure legends, with p value 0.05 as a significance threshold.
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository (Perez-Riverol et al., 2019) with the dataset identifier PXD041786 and 10.6019/PXD041786.
C. elegans genetics and strains
C. elegans N2 isolate was used for all experiments. Animals were maintained using standard procedures. The following mutant alleles were used: vab-1(dx31) II, fsn-1(gk429) III, rpm-1(ju44) V, glo-4(ok623) V. The integrated transgene used to evaluate axon termination was muIs32 [Pmec-7GFP] II. For genetic analysis the animals were grown at 23 °C.
Transgenic extrachromosomal arrays were generated using standard microinjection procedures for C. elegans. vab-1 minigene (pCZ47) was injected at either 25 ng/µL or 50 ng/µL, and co-injection markers used for transgene selection were either neomycin resistance (pBG-264) or Pttx-3::RFP (pBG-41). pBluescript (pBG-49) was used to reach a final concentration of 100 ng/µl in all injection mixes. pCZ47 was a gift from Andrew Chisholm (Addgene plasmid # 128414 ; RRID:Addgene_128414)
C. elegans axon termination analysis and imaging
Axon termination defects were defined as PLM axons that extended beyond the normal termination point adjacent to the ALM cell body. Two different failed termination phenotypes were scored: axon overextension (moderate phenotype) where the PLM axon grew beyond ALM cell body, axons that overextend and form a ventral hook (severe phenotype). To quantify axon termination defects, 20-30 young adult animals were anesthetized (10 μM levamisole in M9 buffer) on a 2% agar pad on glass slides and visualized with a Leica DM5000 B (CTR5000) epifluorescent microscope (40x oil-immersion objective). For image acquisition, young adult animals were mounted on a 3% agarose pad and a Zeiss LSM 710 (40x oil-immersion objective) was used to generate z stacks.
For statistical analysis of axon termination defects, comparisons were done using a Student’s t-test with Bonferroni correction on GraphPad Prism software. Error bars represent standard error of the mean (SEM). Significance was defined as p < 0.05 after Bonferroni correction. Bar graphs represent averages from 8 to 10 counts (20–30 animals/count) obtained from eight or more independent experiments for each genotype. For transgenic rescue experiments, data shown in Figure 8B was obtained from two, independently derived transgenic extrachromosomal arrays.
- 1.Loss-of-function variants in MYCBP2 cause neurobehavioural phenotypes and corpus callosum defects. Brainawac 364
- 2.The Phr1 ubiquitin ligase promotes injury-induced axon self-destructionCell Reports 3:1422–1429
- 3.Defining Minimal Binding Regions in Regulator of Presynaptic Morphology 1 (RPM-1) Using Caenorhabditis elegans Neurons Reveals Differential Signaling ComplexesJournal of Biological Chemistry 292:2519–2530
- 4.RPM-1 uses both ubiquitin ligase and phosphatase-based mechanisms to regulate DLK-1 during neuronal developmentPLoS Genet 10
- 5.EPH receptor tyrosine kinases phosphorylate the PAR-3 scaffold protein to modulate downstream signaling networksCell Rep 40
- 6.Phosphorylation of tyrosine residues in the kinase domain and juxtamembrane region regulates the biological and catalytic activities of Eph receptorsMol Cell Biol 20:4791–805
- 7.The requirement for Phr1 in CNS axon tract formation reveals the corticostriatal boundary as a choice point for cortical axonsGenes Dev 21:2593–606
- 8.RPM-1 regulates axon termination by affecting growth cone collapse and microtubule stabilityDevelopment 144:4658–4672
- 9.Cellular and molecular mechanisms of EPH/EPHRIN signaling in evolution and developmentCurr Top Dev Biol 149:153–201
- 10.Ephrin-B1 forward signaling regulates craniofacial morphogenesis by controlling cell proliferation across Eph-ephrin boundariesGenes Dev 24:2068–80
- 11.Reversing EphB2 depletion rescues cognitive functions in Alzheimer modelNature 469:47–52
- 12.Highwire restrains synaptic growth by attenuating a MAP kinase signalNeuron 51:57–69
- 13.Autophagy is inhibited by ubiquitin ligase activity in the nervous systemNat Commun 10
- 14.Formation of the retinotectal projection requires Esrom, an ortholog of PAM (protein associated with Myc)Development 132:247–56
- 15.EphB receptors interact with NMDA receptors and regulate excitatory synapse formationCell 103:945–56
- 16.Ephrin signalling controls brain size by regulating apoptosis of neural progenitorsNature 435:1244–50
- 17.PAM forms an atypical SCF ubiquitin ligase complex that ubiquitinates and degrades NMNAT2J Biol Chem 293:13897–13909
- 18.Ubiquitin ligase activity inhibits Cdk5 to control axon terminationPLoS Genet 18
- 19.Ligand binding induces Cbl-dependent EphB1 receptor degradation through the lysosomal pathwayTraffic 9:251–66
- 20.Cross Talk between One-Carbon Metabolism, Eph Signaling, and Histone Methylation Promotes Neural Stem Cell DifferentiationCell Reports 23:2864–2873
- 21.Ubiquitination and the Regulation of Membrane ProteinsPhysiol Rev 97:253–281
- 22.Tyrosine phosphorylation of DEPTOR functions as a molecular switch to activate mTOR signalingJ Biol Chem 297
- 23.Eph receptors and ligands comprise two major specificity subclasses and are reciprocally compartmentalized during embryogenesisNeuron 17:9–19
- 24.Dissociation of EphB2 signaling pathways mediating progenitor cell proliferation and tumor suppressionCell 139:679–692
- 25.The VAB-1 Eph receptor tyrosine kinase functions in neural and epithelial morphogenesis in C. elegansCell 92:633–43
- 26.C. elegans RPM-1 regulates axon termination and synaptogenesis through the Rab GEF GLO-4 and the Rab GTPase GLO-1Neuron 55:587–601
- 27.RAE-1, a novel PHR binding protein, is required for axon termination and synapse formation in Caenorhabditis elegansJ Neurosci 32:2628–36
- 28.The PHR proteins: intracellular signaling hubs in neuronal development and axon degenerationNeural development 11:8–18
- 29.Identification of a large Myc-binding protein that contains RCC1-like repeatsProc Natl Acad Sci U S A 95:9172–7
- 30.The role of ubiquitylation in receptor endocytosis and endosomal sortingJ Cell Sci 125:265–75
- 31.A series of normal stages in the development of the chick embryoJ Morphol 88:49–92
- 32.The E3 ubiquitin ligase protein associated with Myc (Pam) regulates mammalian/mechanistic target of rapamycin complex 1 (mTORC1) signaling in vivo through N- and C-terminal domainsJ Biol Chem 287:30063–72
- 33.Nuk controls pathfinding of commissural axons in the mammalian central nervous systemCell 86:35–46
- 34.Dual proteome-scale networks reveal cell-specific remodeling of the human interactomeCell 184:3022–3040
- 35.Mechanisms of ephrin-Eph signalling in development, physiology and diseaseNature reviews Molecular cell biology 17:240–256
- 36.ProHits-viz: a suite of web tools for visualizing interaction proteomics dataNat Methods 14:645–646
- 37.Kinase-dependent and kinase-independent functions of EphA4 receptors in major axon tract formation in vivoNeuron 29:73–84
- 38.Distinct subdomains of the EphA3 receptor mediate ligand binding and receptor dimerizationJ Biol Chem 273:20228–37
- 39.The endosomal sorting adaptor HD-PTP is required for ephrin-B:EphB signalling in cellular collapse and spinal motor axon guidanceScientific reports 9
- 40.The Ubiquitin Ligase Phr1 Regulates Axon Outgrowth through Modulation of Microtubule DynamicsNeuron 56:604–620
- 41.An SCF-like ubiquitin ligase complex that controls presynaptic differentiationNature 430:345–350
- 42.Ligand binding and calcium influx induce distinct ectodomain/gamma-secretase-processing pathways of EphB2 receptorJ Biol Chem 282:16155–63
- 43.ProHits: integrated software for mass spectrometry-based interaction proteomicsNat Biotechnol 28:1015–7
- 44.Specification of motor axon trajectory by ephrin-B:EphB signaling: symmetrical control of axonal patterning in the developing limbNeuron 60:1039–53
- 45.EphB-mediated degradation of the RhoA GEF Ephexin5 relieves a developmental brake on excitatory synapse formationCell 143:442–55
- 46.Eph receptors and ephrins restrict cell intermingling and communicationNature 400:77–81
- 47.Characterization of loss-of-function and gain-of-function Eph receptor tyrosine kinase signaling in C. elegans axon targeting and cell migrationDev Biol 290:164–76
- 48.Fbxo45 Binds SPRY Motifs in the Extracellular Domain of N-Cadherin and Regulates Neuron Migration during Brain DevelopmentMol Cell Biol 40
- 49.Regulation of a DLK-1 and p38 MAP kinase pathway by the ubiquitin ligase RPM-1 is required for presynaptic developmentCell 120:407–20
- 50.Tsc2-Rheb signaling regulates EphA-mediated axon guidanceNature Neuroscience 13:163–172
- 51.Ubiquitin ligase SPSB4 diminishes cell repulsive responses mediated by EphB2Molecular biology of the cell 28:3532–3541
- 52.Activity-based E3 ligase profiling uncovers an E3 ligase with esterification activityNature 556:381–385
- 54.The PRIDE database and related tools and resources in 2019: improving support for quantification dataNucleic Acids Res 47:D442–D450
- 55.Synergistic integration of Netrin and ephrin axon guidance signals by spinal motor neuronseLife 4
- 56.Regulation of EphB2 activation and cell repulsion by feedback control of the MAPK pathwayThe Journal of cell biology 183:933–947
- 57.Ubiquitination switches EphA2 vesicular traffic from a continuous safeguard to a finite signalling modeNat Commun 6
- 58.Fbxo45 forms a novel ubiquitin ligase complex and is required for neuronal developmentMolecular and Cellular Biology 29:3529–3543
- 59.Physical and functional interactome atlas of human receptor tyrosine kinasesEMBO Rep 23
- 60.. rpm-1, a conserved neuronal gene that regulates targeting and synaptogenesis in C. elegansNeuron 26:345–56
- 61.The composition of EphB2 clusters determines the strength in the cellular repulsion responseJ Cell Biol 204:409–22
- 62.Nat Methods:676–82
- 63.Ephrin-A1 induces c-Cbl phosphorylation and EphA receptor down-regulation in T cellsJ Immunol 170:6024–32
- 64.Identification of a peptide inhibitor of the RPM-1 . FSN-1 ubiquitin ligase complexJ Biol Chem 289:34654–66
- 65.Alpha2-chimaerin interacts with EphA4 and regulates EphA4-dependent growth cone collapseProc Natl Acad Sci U S A 104:16347–52
- 66.A chemical genetic approach reveals distinct EphB signaling mechanisms during brain developmentNat Neurosci 15:1645–54
- 67.EphB2 receptor forward signaling controls cortical growth cone collapse via Nck and PakMol Cell Neurosci 52:106–16
- 68.SAINTexpress: improvements and additional features in Significance Analysis of INTeractome softwareJ Proteomics 100:37–43
- 69.The Nesprin family member ANC-1 regulates synapse formation and axon termination by functioning in a pathway with RPM-1 and beta-CateninPLoS Genet 10
- 70.The Ephrin B2 Receptor Tyrosine Kinase Is a Regulator of Proto-oncogene MYC and Molecular Programs Central to Barrett’s NeoplasiaGastroenterology 163:1228–1241
- 71.An atypical ubiquitin ligase at the heart of neural development and programmed axon degenerationNeural Regen Res 17:2347–2350
- 72.c-Cbl-dependent EphA2 protein degradation is induced by ligand bindingMol Cancer Res 1:79–87
- 73.Highwire regulates synaptic growth in DrosophilaNeuron 26:313–29
- 74.Ephrin-B2 and EphB1 mediate retinal axon divergence at the optic chiasmNeuron 39:919–35
- 75.The Highwire ubiquitin ligase promotes axonal degeneration by tuning levels of Nmnat proteinPLoS Biol 10
- 76.Regulation of presynaptic terminal organization by C. elegans RPM-1, a putative guanine nucleotide exchanger with a RING-H2 finger domainNeuron 26:331–43
- 77.EphB-ephrinB bi-directional endocytosis terminates adhesion allowing contact mediated repulsionNature Cell Biology 5:869–878