FEBS Letters xxx (2014) xxx–xxx
journal homepage: www.FEBSLetters.org
Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase Olivier Bornet a, Matthieu Nouailler b, Michaël Feracci a,1, Corinne Sebban-Kreuzer b, Deborah Byrne a, Hubert Halimi b, Xavier Morelli c, Ali Badache c, Françoise Guerlesquin b,⇑ a b c
IMM, CNRS, Aix-Marseille Université, 31 chemin J. Aiguier, 13402 Marseille Cedex 20, France LISM, CNRS UMR 7255, Aix-Marseille Université, 31 chemin J. Aiguier, 13402 Marseille Cedex 20, France CRCM, CNRS UMR7258, Inserm U1068, Institut Paoli-Calmettes, Aix-Marseille Université, UM105, F-13009 Marseille, France
a r t i c l e
i n f o
Article history: Received 7 October 2013 Revised 8 April 2014 Accepted 18 April 2014 Available online xxxx Edited by Angel Nebreda Keywords: IDP Proline rich motifs ErbB2 Src-kinase SH3 domain and NMR spectroscopy
a b s t r a c t Overexpression of the ErbB2 receptor tyrosine kinase is associated with most aggressive tumors in breast cancer patients and is thus one of the main investigated therapeutic targets. Human ErbB2 Cterminal domain is an unstructured anchor that recruits specific adaptors for signaling cascades resulting in cell growth, differentiation and migration. Herein, we report the presence of a SH3 binding motif in the proline rich unfolded ErbB2 C-terminal region. NMR analysis of this motif supports a PPII helix conformation and the binding to Fyn-SH3 domain. The interaction of a kinase of the Src family with ErbB2 C-terminal domain could contribute to synergistic intracellular signaling and enhanced oncogenesis. Structured summary of protein interactions: ErbB2 and Fyn-SH3 bind by nuclear magnetic resonance (View interaction) Fyn-SH3 binds to ErbB2 by surface plasmon resonance (View interaction) Ó 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
1. Introduction Overexpression of the ErbB2 receptor tyrosine kinase is associated with the most aggressive tumors in breast cancer [1], inducing a constitutive ligand-independent activation of intracellular signaling. Experimental studies have demonstrated that ErbB2 receptor tyrosine kinase confers many properties specific of a cancerous cell, such as the uncontrolled proliferation, the resistance to apoptosis and the increase in motility. ErbB2 is composed of four domains: a globular extracellular domain (sequence: 23–646; structure, PDB code: 1S78), a helical membrane-spanning domain (sequence: 650–684; structure, PDB code: 1IIJ), a tyrosine-kinase intracellular domain (sequence: 696–987; structure, PDB code: 3POZ) and an unfolded C-terminal domain, CtErbB2 (sequence: 988–1255). ErbB2 C-terminal domain contains five specific phosphorylated tyrosine residues which serve as binding sites for PTB and SH2 domains found in signaling molecules or adaptors among which Shc, Grb2 or Grb7 [2]. Recruitment of these adaptors is the ⇑ Corresponding author. Fax: +33 491 164 540. E-mail address: [email protected] (F. Guerlesquin). Present address: Department of Biochemistry, University of Leicester, Leicester LE1 9HN, UK. 1
first event instructing the cell to proliferate, differentiate, survive or migrate. While ErbB2 signaling has been extensively studied, the structure, the dynamics and the molecular mechanisms of the unstructured C-terminal domain are not elucidated. CtErbB2, predicted as an intrinsically disordered domain, is the hub of a protein–protein interaction network associated with the ErbB2 receptor and thus plays a crucial role in signaling processes. Compared to structured proteins, intrinsically disordered proteins are enriched in P, Q, M, K, R, S and E residues which are associated with a lack of tertiary structure resulting in a highly flexible main chain. They are usually described as an ensemble of interconverting heterogeneous conformations which allow them to interact with various targets [3]. Linear motifs are short segments forming structured binding site within a disordered protein [4]. In these motifs, a great number of proline residues are found either in the conserved identities or in the flanking regions. The side chain atoms of proline form a pyrrolidine ring with the backbone atoms facilitating sequence-specific recognition in the absence of high affinity interaction [5]. Such specific and weak interactions are essential for cell signaling and communication that require fast reversible interactions. Polyproline stretches can adopt two unique helical conformations. Polyproline type I, PPI, is a right-handed helix consisting of cis-prolines. PPII helix is a left-handed helix
http://dx.doi.org/10.1016/j.febslet.2014.04.029 0014-5793/Ó 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
2
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
which contains prolines in trans-conformation and may accommodate other amino acids such as glutamine, serine and arginine. PPII helix is an extended structure accommodating three residues per turn [6]. These linear motifs are largely found in the eukaryotic proteome and have a great biological importance [7]. CtErbB2 contains 16% of proline residues and the role of the proline-rich motifs has not yet been investigated. We have performed a bioinformatics analysis of the human CtErbB2 and found a SH3 binding motif (R1146PQPPSP1152). Using NMR, we have investigated this linear motif of the CtErbB2 domain and give evidence that it represents a new binding site for Src family kinases (SFKs). 2. Experimental procedures 2.1. Bioinformatics analysis Human CtErbB2 sequence was analyzed using Eukaryotic Linear Motif resource (http://elm.eu.org). A Lig_SH3_1 motif was found in the sequence from 1146 to 1152 and will be called here-after [RPQPPSP]-site. 2.2. Protein production and peptide synthesis A human ErbB2 derived peptide of [RPQPPSP]-site (sequence R1P2Q3P4P5S6P7R8E9) and two mutant peptides (sequence A1P2Q3A4P5S6A7R8E9 and sequence R1A2Q3P4A5S6P7A8E9, respectively) were synthesized by Schaffer Company. A 13C-Pro labeled peptide of [RPQPPSP]-site (sequence R1P2Q3P4P5S6P7R8E9) was synthesized using U-13C, 15N LproN FMOC provided by CortecNet company. For Grb2-SH3N domain a plasmid construct was produced using the pETite N-His SUMO vector (Lucigen Corporation, Middleton, Wisconsin, United states), an enzyme free PCR directional cloning system. This vector system has an N-terminal His-tag followed by a SUMO fusion tag with a SUMO protease site, resulting in a protein with an additional 107 residues added to the synthetic gene of Grb2-SH3N domain (MHHHHHHGSLQDSEVNQEAKPEVKPEVKPETHINLKVSDGSSEIFFKIKKTTPLRRL MEAFAKRQGKEMDSLT FYDGIEIQADQTPEDLDMEDNDIIEAHREQIGG). The plasmid pETite– N-His-SUMO-Grb2-SH3_1 was transformed into HI-Control BL21 (DE3) chemically competent cells for protein expression (Lucigen). Transformed cells were grown in 4L of M9 minimal medium containing 15NH4Cl (1 g/L) as the only source of nitrogen, and the expression was induced with 500 mM IPTG (isopropyl b-D-1-thiogalactopyranoside) at an optical density of 0.6 at 30 °C for 16 h to a final optical density of 1.8. Protein was purified using a Hitrap Ni IMAC FS 1 mL column (GE Healthcare) with a gradient of imidazole. The SH3 domain after cleavage with the SUMO protease was purified using the same column. Cleaved GrbB2 SH3 domain was buffer exchanged to 25 mM phosphate buffer, 150 mM NaCl pH 7 using a HiPrep™ 26/10 Desalting column (GE Healthcare). Finally, gel filtration was performed using a Superdex 75 10/300 GL (GE Healthcare) in 25 mM NaPi, 150 mM NaCl pH 7, 25% glycerol. Fyn-SH3 domain was produced in Escherichia coli BL21 (DE3), according to [8]. 15N-Fyn-SH3 was obtained using 1 L of M9 minimal medium containing 15NH4Cl (1 g/L) as the only source of nitrogen and 15N/13C-Fyn-SH3 was obtained using 2 L of M9 minimal medium containing 15NH4Cl (1 g/L) as the only source of nitrogen and 13C-glucose (2 g/L) as the only source of carbon. Protein purification was achieved using Ni–NTA affinity chromatography (Qiagen) and imidazole gradient. 2.3. Surface plasmon resonance experiments The dissociation constant value was obtained from titration experiments performed on Biacore 3000 equipment. Fyn SH3 domain was covalently immobilized on a carboxymethyl-
dextran-coated sensor chip by amine coupling. R1P2Q3P4P5S6P7R8E9 peptide was injected at 23.4; 46.8; 93.7; 187.5; 375; 750; 1500; 3000; 6000 and 12 000 lM concentration. The dilution series sensograms were evaluated using the BIAevaluation software. 2.4. NMR spectroscopy experiments All the experiments were recorded at 303 K on a Bruker Avance III 600 MHz spectrometer equipped with a TCI cryoprobe. For ErbB2 peptide assignment, natural abundance 13C-HSQC was obtained on the unlabeled R1P2Q3P4P5S6P7R8E9 peptide, and 13CHSQC was carried out on 13C-labeled R1P2Q3P4P5S6P7R8E9 peptide. TOCSY (80 ms) and NOESY (300 ms) experiments were recorded on a 3 mM unlabeled R1P2Q3P4P5S6P7R8E9 peptide. All the samples were in 50 mM KPO4, pH 7.0 and 150 mM NaCl and 90% H2O, 10% D2O. Structure calculations of the R1P2Q3P4P5S6P7R8E9 peptide were carried out with CYANA using interproton distances and backbone dihedral restraints. The accuracy of the NMR models could be assessed based on the criteria for successful structure calculation using the program CYANA. Each of the 25 lowest energy structures chosen for the final structure ensemble were subjected to restrained molecular dynamics using Amber 10.
Fig. 1. (A) 15N-HSQC spectra of Fyn SH3 domain at 0.1 mM concentration, performed in the absence (in black) and in the presence of R1P2Q3P4P5S6P7R8E9 peptide at 0.280 mM concentration (in red), (B) 15N-HSQC spectra of Fyn SH3 domain at 0.1 mM concentration, performed in the absence (in black) and in the presence of A1P2Q3A4P5S6A7R8E9 peptide at 0.280 mM concentration (in red), (C) 15 N-HSQC spectra of Fyn SH3 domain, at 0.1 mM concentration, performed in the absence (in black) and in the presence of R1A2Q3P4A5S6P7A8E9 peptide at 0.280 mM concentration (in red).
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
3
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
Assignment of 15N-HSQC of Fyn SH3 domain was obtained using a N–13C-labeled Fyn SH3 sample at 1 mM in 50 mM KPO4, pH 7.0; 150 mM NaCl and 90% H2O, 10% D2O, and HNCA, HNCOCA, CBCANH, CBCACONH, HNCO and HNCACO heteronuclear experiments. Chemical shift mapping of [RPQPPSP]-site on Fyn-SH3 was obtained from 15N-HSQC recorded on 15N-labeled Fyn-SH3 sample at 0.1 mM in 50 mM phosphate buffer, pH 7.0 and 150 mM NaCl and 90% H2O, 10% D2O, in the presence of 0; 0.05; 0.1; 0.2 and 0.4 mM ErbB2 peptide. These experiments were recorded at 303 K on a Bruker Avance III 500 MHz spectrometer equipped with a QXI probe. 15
3. Results and discussion 3.1. Identification of a class I SH3 binding motif in ErbB2 C-terminal domain Human CtErbB2 contains 46 proline residues over 275 amino acids. The importance of these proline residues in protein–protein interactions and/or domain flexibility is one of our main focal points in CtErbB2 structural studies. Human CtErbB2 sequence
(988–1255) was analyzed using the Eukaryotic Linear Motif (EML) resource to identify functional sites associated to the ErbB2 diseasome. A [RPQPPSP]-site motif was identified at position (1146–1152) as a class I SH3 domain binding site (RxxPxxP). In order to characterize this polyproline motif, we have synthesized a peptide corresponding to the ErbB2 sequence (R1P2Q3P4P5S6P7R8E9). Given that this sequence is located close to one of the phosphotyrosine, pY1139, a ligand of the Grb2 and Grb7-SH2 domains, we decided first to verify if the [RPQPPSP]-site could bind one of the SH3 domains of Grb2 (Grb2-SH3N, for the N-terminal SH3 domain of Grb2). On the basis of 15N-HSQC of the Grb2-SH3N domain, we did not observe any chemical shift variation of Grb2-SH3 amides in the presence of the R1P2Q3P4P5S6P7R8E9 peptide, suggesting that there is no interaction between the two molecules (data not shown). Subsequently, on the basis of the predicted data obtained from the ELM resource, we performed an NMR titration of FynSH3/R1P2Q3P4P5S6P7R8E9 peptide interactions. 15N-HSQC of FynSH3 domain recorded in the presence of the ErbB2 peptide revealed that this peptide induced chemical shifts on some NH correlation peaks of Fyn-SH3 (Fig. 1A). The fast exchange observed between the free and bound forms were in agreement with a KD of 0.935 mM as determined by NMR and of 0.909 mM as deter-
A
B
50
40
50
Response [RU]
Req [RU]
30
20
10
40 30 20 10 0 -10
0
0
100
200
300
400
Time [s] 0
2000
4000
6000
Peptide [μM] Fig. 2. (A) NMR titration of FynSH3 domain/R1P2Q3P4P5S6P7R8E9 peptide complex formation. Black is FynSH3 domain HSQC spectrum, at 0.1 mM concentration and in the absence of peptide, blue is the same sample with 80 lM peptide, red with 150 lM peptide, green with 210 lM peptide, pink with 280 lM peptide and orange with 330 lM peptide. The plot of 96R chemical shift variations in the presence of R1P2Q3P4P5S6P7R8E9 peptide is shown, Dd ¼ ðDdH2 þ ðDdN2 =25Þ1=2 Þ [26], (B) SPR data for FynSH3 domain/ R1P2Q3P4P5S6P7R8E9 peptide complex formation.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
4
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
mined by surface plasmon resonance experiments (Fig. 2). These values are in the range of the weak interactions already reported for others SH3/ligand complexes [9]. 3.2. Structural analysis of ErbB2 polyproline motif The high content in proline residues in the R1146PQPPSPRE1152 sequence suggested a possible left-handed polyproline type II (PPII) helix organization of the peptide. In order to confirm that this peptide has an extended structure consistent with a PPII helix, we have performed a series of NMR experiments. Our first observations from the 13C-HSQC spectra, shows the 13Cc of proline residues around 27 ppm and the 13Cb around 32 ppm (Fig. 3A). These values are in agreement with a trans-conformation for all proline residues of the peptide, considering that for a cis-conformation these values are 24.4 ppm for the 13Cc and 33.8 ppm for the 13Cb [10]. Moreover, on the basis of the sequential assignment of the TOCSY and NOESY homonuclear spectra of the R1P2Q3P4P5S6P7R8E9 peptide, we could observe Hai1/Hdi nOes between residues R1 and P2; Q3 and P4 and S6 and P7 (data not shown). These sequential nOes strongly validate the trans-conformation of proline residues. Then, from NMR experiments we could observe that the peptide contained a single conformation and with the assumption that the PPII conformation allowed a rigid structure to the peptide, we used a general relationship between the size of spin–spin coupling constant 3J and the intervening torsion angle h. The dependence of 3 JHNa on the torsion angle u in polypeptides was obtained from the equation 3 JHNa ¼ 6:4 cos2 h 1:4 cos h þ 1:9 [11]. From 1Dproton experiments, we could measure the 3JHNa coupling
R1P2Q3P4P5S6P7R8E9
13
C
A
constants for Q3 (6.5 Hz), S6 (6.4 Hz), R8 (6.9 Hz) and E9 (7.5 Hz). The plot of 3JHNa versus the torsion angle h provided values around 78° for Q3 and S6 suggesting their implication in an elongated structure that may be a PPII helix. Nonetheless, these spin–spin couplings are valuable data in complementation of NOE distances. The analysis of the sequential nOes gave evidence of strong Hai/ HNi+1 correlation for P2/Q3, P5/S6, P7/R8 and R8/E9 (Fig. 3B) in perfect agreement with a PPII conformation. Finally, we calculated the conformation of the peptide using CYANA software. A PPII helix was clearly obtained from our NMR restraints (Fig. 4A). The structure of R1P2Q3P4P5S6P7R8E9 peptide contains three residues per turn and residues R1, P4 and P7 lie on the same face of the helix, each containing side chains with roughly the same orientations than the peptides 91–104 of the p85 subunit of PI3-Kinase within the complex with Fyn-SH3 (Fig. 4B). We investigated the binding of a mutant of the R1P2Q3P4P5S6P7R8E9 peptide where R1, P4 and P7 were substituted by alanine residues (A1P2Q3A4P5S6A7R8E9). NMR titration experiments revealed that these three residues are essential for the interaction since their substitution abolished the complex formation with Fyn SH3 (Fig. 1B). Multiple prolines are favorable to the interaction with SH3 domains since the pyrrolidone rings conformationally restrict the polypeptide and promote the formation of a sterically driven left handed PPII helix. In particular additional prolines are commonly found in P2 and P5 positions as it is the case for the CtErbB2 [RPQPPSP]-site. The presence of additional proline residues facilitates SH3 binding by stabilizing the PPII conformation and inducing the optimized spacing and orientation of the key residues for the interaction. We thus investigated the binding of a mutant of the R1P2Q3P4P5S6P7R8E9 peptide where P2, P5 and R8 were substituted by alanine (R1A2Q3P4A5S6P7A8E9). NMR titration performed with this mutant indicated that residues P2, P5 and R8 were not essential for the interaction of the peptide with Fyn-SH3 (Fig. 1C). Moreover, the 3JHNa coupling constants found in this mutant were in agreement with a conserved PPII helix in the R1A2Q3P4A5S6P7A8E9 peptide.
A
1
H 1
H
B
B
C CtErbB2 [RPQPPSP]-site: 1
H
Fig. 3. (A) Natural abundance 1H–13C HSQC of unlabeled R1P2Q3P4P5S6P7R8E9 peptide at 2 mM concentration, (B) HaHN correlations in 2D homonuclear NOESY of R1P2Q3P4P5S6P7R8E9 peptide at 2 mM concentration.
Peptide 91-104 of the p85 subunit :
RPQPPS PRE PP RPL PVAPGSSLT
Fig. 4. (A) 25 lowest energy structures of ErbB2 R1P2Q3P4P5S6P7R8E9 peptide, (B) 25 lowest energy structures of the peptide 91–104 of p85 from the complex with Fyn SH3 domain (PDB code: 1A0N), (C) Sequences of CtErbB2 [RPQPPSP]-site peptide and peptide 91–104 of p85 subunit of PI3-Kinase.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
A
0.3 0.25
5
Δδ
0.2 0.15 0.1 0.05 86 LEU 88 VAL 90 LEU 92 ASP 94 GLU 96 ARG 98 GLU 100ASP 102SER 104HIS 106GLY 108LYS 110GLN 112LEU 114SER 116GLU 118ASP 120TRP 122ALA 124SER 126THR 128GLY 130THR 132TYR 134PRO 136ASN 138VAL 140PRO
0
B
Fig. 5. (A) Weighted average chemical shift variations (Dd) observed on resonances of Fyn SH3 domain in the presence of R1P2Q3P4P5S6P7R8E9 peptide, Dd ¼ ðDdH2 þ ðDdN2 =25Þ1=2 Þ [26], (B) chemical shift mapping of R1P2Q3P4P5S6P7R8E9 peptide interacting site on Fyn SH3 domain within the structure of p85 peptide/Fyn SH3 complex (PDB code: 1A0N). Fyn SH3 residues are in grey, Fyn SH3 residues affected by the presence of R1P2Q3P4P5S6P7R8E9 peptide are in green and peptides 91–104 of p85 subunit of PI3-Kinase residues are in red.
Fig. 6. Src kinase SH3 domain sequence alignment. C-Src, Fyn, c-Yes, Fgr, Blk, Hck, Lck and Lyn sequences of SH3 domains are represented. In red are the Fyn SH3 residues affected by human CtErbB2 R1P2Q3P4P5S6P7R8E9 peptide.
3.3. Fyn SH3/CtErbB2 [RPQPPSP]-site complex formation 15
N HSQC recorded on free and bound Fyn SH3 domain indicated a specific chemical shift variation induced by the R1P2Q3P4P5S6P7R8E9 peptide. We obtained sequential NH assignments of the free and R1P2Q3P4P5S6P7R8E9 peptide bound Fyn SH3 domain using heteronuclear experiments. On the basis of chemical shift variations, we could map the interacting surface (Fig. 5A). The interacting interface is very similar to that observed in the structure of p85 peptide 91–104/Fyn SH3 complex (PDB code: 1A0N) (Fig. 5B). The residues affected by the complex formation are respectively located in the RT-loop, the n-Src loop, the distal loop and the ring of the W119, as it was previously reported for numerous Fyn SH3/ polyproline motif complexes [12].
An intriguing question is why are proline-rich motifs widely distributed in eukaryote cells? [13]. PPII helices are stable and resistant structures, that are relatively permissive to amino acid substitutions. Therefore various combinations of non-proline and proline residues do not compromise these structural frames. The non-proline residues located in positions 1, 3 and 6 are at the contacting interfaces and among ligands of Src kinase SH3 domains, except R1, the other residues are variable playing an important role in the ligand specificity (Fig. 4C). SH3 domains constitute one of the best-known protein-interaction modules [14]. These domains play critical roles in a great number of biological processes, binding Pro-rich motifs with low affinity [15]. SH3 domains from the Src kinase family are interesting examples of the balance between promiscuity and specificity.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
6
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
Src kinase family contains nine members c-Src, c-Yes, Fyn, Lyn, Lck, Hck, Blk, Fgr and Yrk which share a SH3 domain responsible for the recruitment of polyproline motifs. A sequential alignment of SH3 domains of SFKs is provided in Fig. 6. One can observe that the residues in Fyn-SH3 domain affected by the presence of the R1P2Q3P4P5S6P7R8E9 peptide are highly conserved among the SFKs. This data is in agreement with the redundant functions of SFKs in the cell. However, there is an increasing number of evidence for SFK specificity in signaling. 3.4. Src family kinases in ErbB2 positive breast cancer SFKs association with ErbB2 has been shown in human breast cancer cell lines and tumors [16–18]. For instance, high Src activity correlates with ErbB2 positivity and high tumor grade in ductal carcinoma in situ [19]. This interaction results in enhanced Src activity and stability [20,21]. The recent observations that various SFKs, including c-Src, c-Yes, Fyn and Lck, contribute to resistance to ErbB2-targeting drugs [22,23] further support the importance of delineating SFK contribution to ErbB2 signaling. Src physically associates with ErbB2, but the mechanism is still unclear, as studies involved either the SH2 or the catalytic domain [16,24,25]. Herein, we have identified a PPII motif in human ErbB2 C-terminus that interacts with Fyn-SH3 domain and thus could contribute to SFKs interaction with ErbB2. In the future, it will be important to determine, beyond Fyn, which SFK SH3 domains actually interact with ErbB2, and if additional SFK domains or other motifs in CtErbB2 contribute to define SFK affinity and specificity for the ErbB2 receptor tyrosine kinase. Acknowledgements The authors thank the CAPM from UMR7286 for SPR measurements and D. Salaün for technical assistance. AB was supported by INCa-DGOS-Inserm 6038. References [1] Badache, A. and Gonçalves, A. (2006) The ErbB2 signaling network as a target for breast cancer therapy. J. Mammary Gland Biol. Neoplasia 11, 13–25. [2] Jones, R.B., Gordus, A., Krall, J.A. and MacBeath, G. (2006) A quantitative protein interaction network for the ErbB receptors using protein microarrays. Nature 439, 168–174. [3] Goh, K.I., Cusick, M.E., Valle, D., Childs, B., Vidal, M. and Barabasi, A.L. (2007) The human disease network. Proc. Natl. Acad. Sci. U.S.A. 104, 8685–8690. [4] Srinivasan, M. and Dunker, A.K. (2012) Proline rich motifs as drug targets in immune mediated disorders. Int. J. pept. 2012, 634769. [5] Kay, B.K., Williamson, M.P. and Sudol, M. (2000) The importance of being proline: the interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J. 14, 231–241. [6] Adzhubei, A.A., Stenberg, M.J.E. and Makarov, A.A. (2013) Polyproline-II helix in proteins: structure and function. J. Mol. Biol. 425, 2100–2132.
[7] Mansiaux, Y., Joseph, A.P., Gelly, J.C. and de Breven, A.G. (2011) Assignment of PolyProline II conformation and analysis of sequence–structure relationship. PLoS One 6 (3), e18401. [8] Shi, X., Betzi, S., Lugari, A., Opi, S., Restouin, A., Parrot, I., Martinez, J., Zimmermann, P., Lecine, P., Huang, M., Arold, S.T., Collette, Y. and Morelli, X. (2012) Structural recognition mechanisms between human Src homology domain 3 (SH3) and ALG-2-interacting protein X (Alix). FEBS Lett. 586 (13), 1759–1764. [9] Yu, H., Chen, J.K., Feng, S., Dalgarno, D.C., Brauer, A.W. and Schreiber, S.L. (1994) Structural basis for the binding of proline-rich peptides to SH3 domains. Cell 76, 933–945. [10] Shen, Y. and Bax, Ad. (2010) Prediction of Xaa-Pro peptide bond conformation from sequence and chemical shifts. J. Biomol. NMR 46, 199–204. [11] Wuthrich, K. (1986) NMR of Proteins and Acid Nucleics, Wiley-Interscience Publication. [12] Renzoni, D.A., Pugh, D.J., Siligardi, G., Das, P., Morton, C.J., Rossi, C., Waterfield, M.D., Campbell, I.D. and Ladbury, J.E. (1996) Structural and thermodynamic characterization of the interaction of the SH3 domain from Fyn with the proline-rich binding site on the p85 subunit of PI3-kinase. Biochemistry 35 (49), 15646–15653. [13] Li, S.S. (2005) Specificity and versatility of SH3 and other proline-recognition domains: structural basis and implications for cellular signal transduction. Biochem. J. 390, 641–653. [14] Saksela, K. and Permi, P. (2012) SH3 domain ligand binding: what’s the consensus and where’s the specificity? FEBS Lett. 586, 2609–2614. [15] Mayer, B.J. (2001) SH3 domains: complexity in moderation. J. Cell Sci. 114, 1253–1263. [16] Muthuswamy, S.K., Siegel, P.M., Dankort, D.L., Webster, M.A. and Muller, W.J. (1994) Mammary tumors expressing the neu proto-oncogene possess elevated c-Src tyrosine kinase activity. Mol. Cell. Biol. 14, 735–743. [17] Sheffield, L.G. (1998) C-Src activation by ErbB2 leads to attachmentindependent growth of human breast epithelial cells. Biochem. Biophys. Res. Commun. 250, 27–31. [18] Belsches-Jablonski, A.P., Biscardi, J.S., Peavy, D.R., Tice, D.A., Romney, D.A. and Parsons, S.J. (2001) Src family kinases and HER2 interactions in human breast cancer cell growth and survival. Oncogene 20, 1465–1475. [19] Wilson, G.R., Cramer, A., Welman, A., Knox, F., Swindell, R., Kawakatsu, H., Clarke, R.B., Dive, C. and Bundred, N.J. (2006) Activated c-SRC in ductal carcinoma in situ correlates with high tumour grade, high proliferation and HER2 positivity. Br. J. Cancer 95, 1410–1414. [20] Vadlamudi, R.K., Sahin, A.A., Adam, L., Wang, R.A. and Kumar, R. (2003) Heregulin and HER2 signaling selectively activates c-Src phosphorylation at tyrosine 215. FEBS Lett. 543, 76–80. [21] Tan, M., Li, P., Klos, K.S., Lu, J., Lan, K.H., Nagata, Y., Fang, D., Jing, T. and Yu, D. (2005) ErbB2 promotes Src synthesis and stability: novel mechanisms of Src activation that confer breast cancer metastasis. Cancer Res. 65, 1858–1867. [22] Rexer, B.N., Ham, A.J., Rinehart, C., Hills, S., Granja-Ingram, M., GonzalesAngulo, A.M., Mills, G.B., Dave, B., Chang, J.C., Liebeler, D.C. and Arteaga, C.L. (2011) Phosphoproteomic mass spectrometry profiling links Src family kinases to escape from HER2 tyrosine kinase inhibition. Oncogene 30, 4163– 4174. [23] Zhang, S., Huang, W.C., Li, P., Guo, H., Poh, S.B., Brady, S.W., Xiong, Y., Tseng, L.M., Li, S.H., Ding, Z., Sahin, A.A., Esteva, F.J., Hortobagyi, G.N. and Yu, D. (2011) Combating trastuzumab resistance by targeting SRC, a common node downstream of multiple resistance pathways. Nat. Med. 17, 461–469. [24] Biscardi, J.S., Ishizawar, R.C., Silva, C.M. and Parsons, S.J. (2000) Tyrosine kinase signalling in breast cancer: epidermal growth factor receptor and c-Src interactions in breast cancer. Breast Cancer Res. 2, 203–210. [25] Marcotte, R., Zhou, L., Kim, H., Roskelly, C.D. and Muller, W.J. (2009) C-Src associates with ErbB2 through an interaction between catalytic domains and confers enhanced transforming potential. Mol. Cell. Biol. 29, 5858–5871. [26] Grzesiek, S., Bax, A., Clore, G.M., Gronenborn, A.M., Hu, J.S., Kaufman, J., Palmer, I., Stahl, S.J. and Wingfield, P.T. (1996) The solution structure of HIV-1 Nef reveals an unexpected fold and permits delineation of the binding surface for the SH3 domain of Hck tyrosine proteinkinase. Nat. Struct. Biol. 3, 340–345.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
journal homepage: www.FEBSLetters.org
Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase Olivier Bornet a, Matthieu Nouailler b, Michaël Feracci a,1, Corinne Sebban-Kreuzer b, Deborah Byrne a, Hubert Halimi b, Xavier Morelli c, Ali Badache c, Françoise Guerlesquin b,⇑ a b c
IMM, CNRS, Aix-Marseille Université, 31 chemin J. Aiguier, 13402 Marseille Cedex 20, France LISM, CNRS UMR 7255, Aix-Marseille Université, 31 chemin J. Aiguier, 13402 Marseille Cedex 20, France CRCM, CNRS UMR7258, Inserm U1068, Institut Paoli-Calmettes, Aix-Marseille Université, UM105, F-13009 Marseille, France
a r t i c l e
i n f o
Article history: Received 7 October 2013 Revised 8 April 2014 Accepted 18 April 2014 Available online xxxx Edited by Angel Nebreda Keywords: IDP Proline rich motifs ErbB2 Src-kinase SH3 domain and NMR spectroscopy
a b s t r a c t Overexpression of the ErbB2 receptor tyrosine kinase is associated with most aggressive tumors in breast cancer patients and is thus one of the main investigated therapeutic targets. Human ErbB2 Cterminal domain is an unstructured anchor that recruits specific adaptors for signaling cascades resulting in cell growth, differentiation and migration. Herein, we report the presence of a SH3 binding motif in the proline rich unfolded ErbB2 C-terminal region. NMR analysis of this motif supports a PPII helix conformation and the binding to Fyn-SH3 domain. The interaction of a kinase of the Src family with ErbB2 C-terminal domain could contribute to synergistic intracellular signaling and enhanced oncogenesis. Structured summary of protein interactions: ErbB2 and Fyn-SH3 bind by nuclear magnetic resonance (View interaction) Fyn-SH3 binds to ErbB2 by surface plasmon resonance (View interaction) Ó 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
1. Introduction Overexpression of the ErbB2 receptor tyrosine kinase is associated with the most aggressive tumors in breast cancer [1], inducing a constitutive ligand-independent activation of intracellular signaling. Experimental studies have demonstrated that ErbB2 receptor tyrosine kinase confers many properties specific of a cancerous cell, such as the uncontrolled proliferation, the resistance to apoptosis and the increase in motility. ErbB2 is composed of four domains: a globular extracellular domain (sequence: 23–646; structure, PDB code: 1S78), a helical membrane-spanning domain (sequence: 650–684; structure, PDB code: 1IIJ), a tyrosine-kinase intracellular domain (sequence: 696–987; structure, PDB code: 3POZ) and an unfolded C-terminal domain, CtErbB2 (sequence: 988–1255). ErbB2 C-terminal domain contains five specific phosphorylated tyrosine residues which serve as binding sites for PTB and SH2 domains found in signaling molecules or adaptors among which Shc, Grb2 or Grb7 [2]. Recruitment of these adaptors is the ⇑ Corresponding author. Fax: +33 491 164 540. E-mail address: [email protected] (F. Guerlesquin). Present address: Department of Biochemistry, University of Leicester, Leicester LE1 9HN, UK. 1
first event instructing the cell to proliferate, differentiate, survive or migrate. While ErbB2 signaling has been extensively studied, the structure, the dynamics and the molecular mechanisms of the unstructured C-terminal domain are not elucidated. CtErbB2, predicted as an intrinsically disordered domain, is the hub of a protein–protein interaction network associated with the ErbB2 receptor and thus plays a crucial role in signaling processes. Compared to structured proteins, intrinsically disordered proteins are enriched in P, Q, M, K, R, S and E residues which are associated with a lack of tertiary structure resulting in a highly flexible main chain. They are usually described as an ensemble of interconverting heterogeneous conformations which allow them to interact with various targets [3]. Linear motifs are short segments forming structured binding site within a disordered protein [4]. In these motifs, a great number of proline residues are found either in the conserved identities or in the flanking regions. The side chain atoms of proline form a pyrrolidine ring with the backbone atoms facilitating sequence-specific recognition in the absence of high affinity interaction [5]. Such specific and weak interactions are essential for cell signaling and communication that require fast reversible interactions. Polyproline stretches can adopt two unique helical conformations. Polyproline type I, PPI, is a right-handed helix consisting of cis-prolines. PPII helix is a left-handed helix
http://dx.doi.org/10.1016/j.febslet.2014.04.029 0014-5793/Ó 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
2
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
which contains prolines in trans-conformation and may accommodate other amino acids such as glutamine, serine and arginine. PPII helix is an extended structure accommodating three residues per turn [6]. These linear motifs are largely found in the eukaryotic proteome and have a great biological importance [7]. CtErbB2 contains 16% of proline residues and the role of the proline-rich motifs has not yet been investigated. We have performed a bioinformatics analysis of the human CtErbB2 and found a SH3 binding motif (R1146PQPPSP1152). Using NMR, we have investigated this linear motif of the CtErbB2 domain and give evidence that it represents a new binding site for Src family kinases (SFKs). 2. Experimental procedures 2.1. Bioinformatics analysis Human CtErbB2 sequence was analyzed using Eukaryotic Linear Motif resource (http://elm.eu.org). A Lig_SH3_1 motif was found in the sequence from 1146 to 1152 and will be called here-after [RPQPPSP]-site. 2.2. Protein production and peptide synthesis A human ErbB2 derived peptide of [RPQPPSP]-site (sequence R1P2Q3P4P5S6P7R8E9) and two mutant peptides (sequence A1P2Q3A4P5S6A7R8E9 and sequence R1A2Q3P4A5S6P7A8E9, respectively) were synthesized by Schaffer Company. A 13C-Pro labeled peptide of [RPQPPSP]-site (sequence R1P2Q3P4P5S6P7R8E9) was synthesized using U-13C, 15N LproN FMOC provided by CortecNet company. For Grb2-SH3N domain a plasmid construct was produced using the pETite N-His SUMO vector (Lucigen Corporation, Middleton, Wisconsin, United states), an enzyme free PCR directional cloning system. This vector system has an N-terminal His-tag followed by a SUMO fusion tag with a SUMO protease site, resulting in a protein with an additional 107 residues added to the synthetic gene of Grb2-SH3N domain (MHHHHHHGSLQDSEVNQEAKPEVKPEVKPETHINLKVSDGSSEIFFKIKKTTPLRRL MEAFAKRQGKEMDSLT FYDGIEIQADQTPEDLDMEDNDIIEAHREQIGG). The plasmid pETite– N-His-SUMO-Grb2-SH3_1 was transformed into HI-Control BL21 (DE3) chemically competent cells for protein expression (Lucigen). Transformed cells were grown in 4L of M9 minimal medium containing 15NH4Cl (1 g/L) as the only source of nitrogen, and the expression was induced with 500 mM IPTG (isopropyl b-D-1-thiogalactopyranoside) at an optical density of 0.6 at 30 °C for 16 h to a final optical density of 1.8. Protein was purified using a Hitrap Ni IMAC FS 1 mL column (GE Healthcare) with a gradient of imidazole. The SH3 domain after cleavage with the SUMO protease was purified using the same column. Cleaved GrbB2 SH3 domain was buffer exchanged to 25 mM phosphate buffer, 150 mM NaCl pH 7 using a HiPrep™ 26/10 Desalting column (GE Healthcare). Finally, gel filtration was performed using a Superdex 75 10/300 GL (GE Healthcare) in 25 mM NaPi, 150 mM NaCl pH 7, 25% glycerol. Fyn-SH3 domain was produced in Escherichia coli BL21 (DE3), according to [8]. 15N-Fyn-SH3 was obtained using 1 L of M9 minimal medium containing 15NH4Cl (1 g/L) as the only source of nitrogen and 15N/13C-Fyn-SH3 was obtained using 2 L of M9 minimal medium containing 15NH4Cl (1 g/L) as the only source of nitrogen and 13C-glucose (2 g/L) as the only source of carbon. Protein purification was achieved using Ni–NTA affinity chromatography (Qiagen) and imidazole gradient. 2.3. Surface plasmon resonance experiments The dissociation constant value was obtained from titration experiments performed on Biacore 3000 equipment. Fyn SH3 domain was covalently immobilized on a carboxymethyl-
dextran-coated sensor chip by amine coupling. R1P2Q3P4P5S6P7R8E9 peptide was injected at 23.4; 46.8; 93.7; 187.5; 375; 750; 1500; 3000; 6000 and 12 000 lM concentration. The dilution series sensograms were evaluated using the BIAevaluation software. 2.4. NMR spectroscopy experiments All the experiments were recorded at 303 K on a Bruker Avance III 600 MHz spectrometer equipped with a TCI cryoprobe. For ErbB2 peptide assignment, natural abundance 13C-HSQC was obtained on the unlabeled R1P2Q3P4P5S6P7R8E9 peptide, and 13CHSQC was carried out on 13C-labeled R1P2Q3P4P5S6P7R8E9 peptide. TOCSY (80 ms) and NOESY (300 ms) experiments were recorded on a 3 mM unlabeled R1P2Q3P4P5S6P7R8E9 peptide. All the samples were in 50 mM KPO4, pH 7.0 and 150 mM NaCl and 90% H2O, 10% D2O. Structure calculations of the R1P2Q3P4P5S6P7R8E9 peptide were carried out with CYANA using interproton distances and backbone dihedral restraints. The accuracy of the NMR models could be assessed based on the criteria for successful structure calculation using the program CYANA. Each of the 25 lowest energy structures chosen for the final structure ensemble were subjected to restrained molecular dynamics using Amber 10.
Fig. 1. (A) 15N-HSQC spectra of Fyn SH3 domain at 0.1 mM concentration, performed in the absence (in black) and in the presence of R1P2Q3P4P5S6P7R8E9 peptide at 0.280 mM concentration (in red), (B) 15N-HSQC spectra of Fyn SH3 domain at 0.1 mM concentration, performed in the absence (in black) and in the presence of A1P2Q3A4P5S6A7R8E9 peptide at 0.280 mM concentration (in red), (C) 15 N-HSQC spectra of Fyn SH3 domain, at 0.1 mM concentration, performed in the absence (in black) and in the presence of R1A2Q3P4A5S6P7A8E9 peptide at 0.280 mM concentration (in red).
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
3
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
Assignment of 15N-HSQC of Fyn SH3 domain was obtained using a N–13C-labeled Fyn SH3 sample at 1 mM in 50 mM KPO4, pH 7.0; 150 mM NaCl and 90% H2O, 10% D2O, and HNCA, HNCOCA, CBCANH, CBCACONH, HNCO and HNCACO heteronuclear experiments. Chemical shift mapping of [RPQPPSP]-site on Fyn-SH3 was obtained from 15N-HSQC recorded on 15N-labeled Fyn-SH3 sample at 0.1 mM in 50 mM phosphate buffer, pH 7.0 and 150 mM NaCl and 90% H2O, 10% D2O, in the presence of 0; 0.05; 0.1; 0.2 and 0.4 mM ErbB2 peptide. These experiments were recorded at 303 K on a Bruker Avance III 500 MHz spectrometer equipped with a QXI probe. 15
3. Results and discussion 3.1. Identification of a class I SH3 binding motif in ErbB2 C-terminal domain Human CtErbB2 contains 46 proline residues over 275 amino acids. The importance of these proline residues in protein–protein interactions and/or domain flexibility is one of our main focal points in CtErbB2 structural studies. Human CtErbB2 sequence
(988–1255) was analyzed using the Eukaryotic Linear Motif (EML) resource to identify functional sites associated to the ErbB2 diseasome. A [RPQPPSP]-site motif was identified at position (1146–1152) as a class I SH3 domain binding site (RxxPxxP). In order to characterize this polyproline motif, we have synthesized a peptide corresponding to the ErbB2 sequence (R1P2Q3P4P5S6P7R8E9). Given that this sequence is located close to one of the phosphotyrosine, pY1139, a ligand of the Grb2 and Grb7-SH2 domains, we decided first to verify if the [RPQPPSP]-site could bind one of the SH3 domains of Grb2 (Grb2-SH3N, for the N-terminal SH3 domain of Grb2). On the basis of 15N-HSQC of the Grb2-SH3N domain, we did not observe any chemical shift variation of Grb2-SH3 amides in the presence of the R1P2Q3P4P5S6P7R8E9 peptide, suggesting that there is no interaction between the two molecules (data not shown). Subsequently, on the basis of the predicted data obtained from the ELM resource, we performed an NMR titration of FynSH3/R1P2Q3P4P5S6P7R8E9 peptide interactions. 15N-HSQC of FynSH3 domain recorded in the presence of the ErbB2 peptide revealed that this peptide induced chemical shifts on some NH correlation peaks of Fyn-SH3 (Fig. 1A). The fast exchange observed between the free and bound forms were in agreement with a KD of 0.935 mM as determined by NMR and of 0.909 mM as deter-
A
B
50
40
50
Response [RU]
Req [RU]
30
20
10
40 30 20 10 0 -10
0
0
100
200
300
400
Time [s] 0
2000
4000
6000
Peptide [μM] Fig. 2. (A) NMR titration of FynSH3 domain/R1P2Q3P4P5S6P7R8E9 peptide complex formation. Black is FynSH3 domain HSQC spectrum, at 0.1 mM concentration and in the absence of peptide, blue is the same sample with 80 lM peptide, red with 150 lM peptide, green with 210 lM peptide, pink with 280 lM peptide and orange with 330 lM peptide. The plot of 96R chemical shift variations in the presence of R1P2Q3P4P5S6P7R8E9 peptide is shown, Dd ¼ ðDdH2 þ ðDdN2 =25Þ1=2 Þ [26], (B) SPR data for FynSH3 domain/ R1P2Q3P4P5S6P7R8E9 peptide complex formation.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
4
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
mined by surface plasmon resonance experiments (Fig. 2). These values are in the range of the weak interactions already reported for others SH3/ligand complexes [9]. 3.2. Structural analysis of ErbB2 polyproline motif The high content in proline residues in the R1146PQPPSPRE1152 sequence suggested a possible left-handed polyproline type II (PPII) helix organization of the peptide. In order to confirm that this peptide has an extended structure consistent with a PPII helix, we have performed a series of NMR experiments. Our first observations from the 13C-HSQC spectra, shows the 13Cc of proline residues around 27 ppm and the 13Cb around 32 ppm (Fig. 3A). These values are in agreement with a trans-conformation for all proline residues of the peptide, considering that for a cis-conformation these values are 24.4 ppm for the 13Cc and 33.8 ppm for the 13Cb [10]. Moreover, on the basis of the sequential assignment of the TOCSY and NOESY homonuclear spectra of the R1P2Q3P4P5S6P7R8E9 peptide, we could observe Hai1/Hdi nOes between residues R1 and P2; Q3 and P4 and S6 and P7 (data not shown). These sequential nOes strongly validate the trans-conformation of proline residues. Then, from NMR experiments we could observe that the peptide contained a single conformation and with the assumption that the PPII conformation allowed a rigid structure to the peptide, we used a general relationship between the size of spin–spin coupling constant 3J and the intervening torsion angle h. The dependence of 3 JHNa on the torsion angle u in polypeptides was obtained from the equation 3 JHNa ¼ 6:4 cos2 h 1:4 cos h þ 1:9 [11]. From 1Dproton experiments, we could measure the 3JHNa coupling
R1P2Q3P4P5S6P7R8E9
13
C
A
constants for Q3 (6.5 Hz), S6 (6.4 Hz), R8 (6.9 Hz) and E9 (7.5 Hz). The plot of 3JHNa versus the torsion angle h provided values around 78° for Q3 and S6 suggesting their implication in an elongated structure that may be a PPII helix. Nonetheless, these spin–spin couplings are valuable data in complementation of NOE distances. The analysis of the sequential nOes gave evidence of strong Hai/ HNi+1 correlation for P2/Q3, P5/S6, P7/R8 and R8/E9 (Fig. 3B) in perfect agreement with a PPII conformation. Finally, we calculated the conformation of the peptide using CYANA software. A PPII helix was clearly obtained from our NMR restraints (Fig. 4A). The structure of R1P2Q3P4P5S6P7R8E9 peptide contains three residues per turn and residues R1, P4 and P7 lie on the same face of the helix, each containing side chains with roughly the same orientations than the peptides 91–104 of the p85 subunit of PI3-Kinase within the complex with Fyn-SH3 (Fig. 4B). We investigated the binding of a mutant of the R1P2Q3P4P5S6P7R8E9 peptide where R1, P4 and P7 were substituted by alanine residues (A1P2Q3A4P5S6A7R8E9). NMR titration experiments revealed that these three residues are essential for the interaction since their substitution abolished the complex formation with Fyn SH3 (Fig. 1B). Multiple prolines are favorable to the interaction with SH3 domains since the pyrrolidone rings conformationally restrict the polypeptide and promote the formation of a sterically driven left handed PPII helix. In particular additional prolines are commonly found in P2 and P5 positions as it is the case for the CtErbB2 [RPQPPSP]-site. The presence of additional proline residues facilitates SH3 binding by stabilizing the PPII conformation and inducing the optimized spacing and orientation of the key residues for the interaction. We thus investigated the binding of a mutant of the R1P2Q3P4P5S6P7R8E9 peptide where P2, P5 and R8 were substituted by alanine (R1A2Q3P4A5S6P7A8E9). NMR titration performed with this mutant indicated that residues P2, P5 and R8 were not essential for the interaction of the peptide with Fyn-SH3 (Fig. 1C). Moreover, the 3JHNa coupling constants found in this mutant were in agreement with a conserved PPII helix in the R1A2Q3P4A5S6P7A8E9 peptide.
A
1
H 1
H
B
B
C CtErbB2 [RPQPPSP]-site: 1
H
Fig. 3. (A) Natural abundance 1H–13C HSQC of unlabeled R1P2Q3P4P5S6P7R8E9 peptide at 2 mM concentration, (B) HaHN correlations in 2D homonuclear NOESY of R1P2Q3P4P5S6P7R8E9 peptide at 2 mM concentration.
Peptide 91-104 of the p85 subunit :
RPQPPS PRE PP RPL PVAPGSSLT
Fig. 4. (A) 25 lowest energy structures of ErbB2 R1P2Q3P4P5S6P7R8E9 peptide, (B) 25 lowest energy structures of the peptide 91–104 of p85 from the complex with Fyn SH3 domain (PDB code: 1A0N), (C) Sequences of CtErbB2 [RPQPPSP]-site peptide and peptide 91–104 of p85 subunit of PI3-Kinase.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
A
0.3 0.25
5
Δδ
0.2 0.15 0.1 0.05 86 LEU 88 VAL 90 LEU 92 ASP 94 GLU 96 ARG 98 GLU 100ASP 102SER 104HIS 106GLY 108LYS 110GLN 112LEU 114SER 116GLU 118ASP 120TRP 122ALA 124SER 126THR 128GLY 130THR 132TYR 134PRO 136ASN 138VAL 140PRO
0
B
Fig. 5. (A) Weighted average chemical shift variations (Dd) observed on resonances of Fyn SH3 domain in the presence of R1P2Q3P4P5S6P7R8E9 peptide, Dd ¼ ðDdH2 þ ðDdN2 =25Þ1=2 Þ [26], (B) chemical shift mapping of R1P2Q3P4P5S6P7R8E9 peptide interacting site on Fyn SH3 domain within the structure of p85 peptide/Fyn SH3 complex (PDB code: 1A0N). Fyn SH3 residues are in grey, Fyn SH3 residues affected by the presence of R1P2Q3P4P5S6P7R8E9 peptide are in green and peptides 91–104 of p85 subunit of PI3-Kinase residues are in red.
Fig. 6. Src kinase SH3 domain sequence alignment. C-Src, Fyn, c-Yes, Fgr, Blk, Hck, Lck and Lyn sequences of SH3 domains are represented. In red are the Fyn SH3 residues affected by human CtErbB2 R1P2Q3P4P5S6P7R8E9 peptide.
3.3. Fyn SH3/CtErbB2 [RPQPPSP]-site complex formation 15
N HSQC recorded on free and bound Fyn SH3 domain indicated a specific chemical shift variation induced by the R1P2Q3P4P5S6P7R8E9 peptide. We obtained sequential NH assignments of the free and R1P2Q3P4P5S6P7R8E9 peptide bound Fyn SH3 domain using heteronuclear experiments. On the basis of chemical shift variations, we could map the interacting surface (Fig. 5A). The interacting interface is very similar to that observed in the structure of p85 peptide 91–104/Fyn SH3 complex (PDB code: 1A0N) (Fig. 5B). The residues affected by the complex formation are respectively located in the RT-loop, the n-Src loop, the distal loop and the ring of the W119, as it was previously reported for numerous Fyn SH3/ polyproline motif complexes [12].
An intriguing question is why are proline-rich motifs widely distributed in eukaryote cells? [13]. PPII helices are stable and resistant structures, that are relatively permissive to amino acid substitutions. Therefore various combinations of non-proline and proline residues do not compromise these structural frames. The non-proline residues located in positions 1, 3 and 6 are at the contacting interfaces and among ligands of Src kinase SH3 domains, except R1, the other residues are variable playing an important role in the ligand specificity (Fig. 4C). SH3 domains constitute one of the best-known protein-interaction modules [14]. These domains play critical roles in a great number of biological processes, binding Pro-rich motifs with low affinity [15]. SH3 domains from the Src kinase family are interesting examples of the balance between promiscuity and specificity.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
6
O. Bornet et al. / FEBS Letters xxx (2014) xxx–xxx
Src kinase family contains nine members c-Src, c-Yes, Fyn, Lyn, Lck, Hck, Blk, Fgr and Yrk which share a SH3 domain responsible for the recruitment of polyproline motifs. A sequential alignment of SH3 domains of SFKs is provided in Fig. 6. One can observe that the residues in Fyn-SH3 domain affected by the presence of the R1P2Q3P4P5S6P7R8E9 peptide are highly conserved among the SFKs. This data is in agreement with the redundant functions of SFKs in the cell. However, there is an increasing number of evidence for SFK specificity in signaling. 3.4. Src family kinases in ErbB2 positive breast cancer SFKs association with ErbB2 has been shown in human breast cancer cell lines and tumors [16–18]. For instance, high Src activity correlates with ErbB2 positivity and high tumor grade in ductal carcinoma in situ [19]. This interaction results in enhanced Src activity and stability [20,21]. The recent observations that various SFKs, including c-Src, c-Yes, Fyn and Lck, contribute to resistance to ErbB2-targeting drugs [22,23] further support the importance of delineating SFK contribution to ErbB2 signaling. Src physically associates with ErbB2, but the mechanism is still unclear, as studies involved either the SH2 or the catalytic domain [16,24,25]. Herein, we have identified a PPII motif in human ErbB2 C-terminus that interacts with Fyn-SH3 domain and thus could contribute to SFKs interaction with ErbB2. In the future, it will be important to determine, beyond Fyn, which SFK SH3 domains actually interact with ErbB2, and if additional SFK domains or other motifs in CtErbB2 contribute to define SFK affinity and specificity for the ErbB2 receptor tyrosine kinase. Acknowledgements The authors thank the CAPM from UMR7286 for SPR measurements and D. Salaün for technical assistance. AB was supported by INCa-DGOS-Inserm 6038. References [1] Badache, A. and Gonçalves, A. (2006) The ErbB2 signaling network as a target for breast cancer therapy. J. Mammary Gland Biol. Neoplasia 11, 13–25. [2] Jones, R.B., Gordus, A., Krall, J.A. and MacBeath, G. (2006) A quantitative protein interaction network for the ErbB receptors using protein microarrays. Nature 439, 168–174. [3] Goh, K.I., Cusick, M.E., Valle, D., Childs, B., Vidal, M. and Barabasi, A.L. (2007) The human disease network. Proc. Natl. Acad. Sci. U.S.A. 104, 8685–8690. [4] Srinivasan, M. and Dunker, A.K. (2012) Proline rich motifs as drug targets in immune mediated disorders. Int. J. pept. 2012, 634769. [5] Kay, B.K., Williamson, M.P. and Sudol, M. (2000) The importance of being proline: the interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J. 14, 231–241. [6] Adzhubei, A.A., Stenberg, M.J.E. and Makarov, A.A. (2013) Polyproline-II helix in proteins: structure and function. J. Mol. Biol. 425, 2100–2132.
[7] Mansiaux, Y., Joseph, A.P., Gelly, J.C. and de Breven, A.G. (2011) Assignment of PolyProline II conformation and analysis of sequence–structure relationship. PLoS One 6 (3), e18401. [8] Shi, X., Betzi, S., Lugari, A., Opi, S., Restouin, A., Parrot, I., Martinez, J., Zimmermann, P., Lecine, P., Huang, M., Arold, S.T., Collette, Y. and Morelli, X. (2012) Structural recognition mechanisms between human Src homology domain 3 (SH3) and ALG-2-interacting protein X (Alix). FEBS Lett. 586 (13), 1759–1764. [9] Yu, H., Chen, J.K., Feng, S., Dalgarno, D.C., Brauer, A.W. and Schreiber, S.L. (1994) Structural basis for the binding of proline-rich peptides to SH3 domains. Cell 76, 933–945. [10] Shen, Y. and Bax, Ad. (2010) Prediction of Xaa-Pro peptide bond conformation from sequence and chemical shifts. J. Biomol. NMR 46, 199–204. [11] Wuthrich, K. (1986) NMR of Proteins and Acid Nucleics, Wiley-Interscience Publication. [12] Renzoni, D.A., Pugh, D.J., Siligardi, G., Das, P., Morton, C.J., Rossi, C., Waterfield, M.D., Campbell, I.D. and Ladbury, J.E. (1996) Structural and thermodynamic characterization of the interaction of the SH3 domain from Fyn with the proline-rich binding site on the p85 subunit of PI3-kinase. Biochemistry 35 (49), 15646–15653. [13] Li, S.S. (2005) Specificity and versatility of SH3 and other proline-recognition domains: structural basis and implications for cellular signal transduction. Biochem. J. 390, 641–653. [14] Saksela, K. and Permi, P. (2012) SH3 domain ligand binding: what’s the consensus and where’s the specificity? FEBS Lett. 586, 2609–2614. [15] Mayer, B.J. (2001) SH3 domains: complexity in moderation. J. Cell Sci. 114, 1253–1263. [16] Muthuswamy, S.K., Siegel, P.M., Dankort, D.L., Webster, M.A. and Muller, W.J. (1994) Mammary tumors expressing the neu proto-oncogene possess elevated c-Src tyrosine kinase activity. Mol. Cell. Biol. 14, 735–743. [17] Sheffield, L.G. (1998) C-Src activation by ErbB2 leads to attachmentindependent growth of human breast epithelial cells. Biochem. Biophys. Res. Commun. 250, 27–31. [18] Belsches-Jablonski, A.P., Biscardi, J.S., Peavy, D.R., Tice, D.A., Romney, D.A. and Parsons, S.J. (2001) Src family kinases and HER2 interactions in human breast cancer cell growth and survival. Oncogene 20, 1465–1475. [19] Wilson, G.R., Cramer, A., Welman, A., Knox, F., Swindell, R., Kawakatsu, H., Clarke, R.B., Dive, C. and Bundred, N.J. (2006) Activated c-SRC in ductal carcinoma in situ correlates with high tumour grade, high proliferation and HER2 positivity. Br. J. Cancer 95, 1410–1414. [20] Vadlamudi, R.K., Sahin, A.A., Adam, L., Wang, R.A. and Kumar, R. (2003) Heregulin and HER2 signaling selectively activates c-Src phosphorylation at tyrosine 215. FEBS Lett. 543, 76–80. [21] Tan, M., Li, P., Klos, K.S., Lu, J., Lan, K.H., Nagata, Y., Fang, D., Jing, T. and Yu, D. (2005) ErbB2 promotes Src synthesis and stability: novel mechanisms of Src activation that confer breast cancer metastasis. Cancer Res. 65, 1858–1867. [22] Rexer, B.N., Ham, A.J., Rinehart, C., Hills, S., Granja-Ingram, M., GonzalesAngulo, A.M., Mills, G.B., Dave, B., Chang, J.C., Liebeler, D.C. and Arteaga, C.L. (2011) Phosphoproteomic mass spectrometry profiling links Src family kinases to escape from HER2 tyrosine kinase inhibition. Oncogene 30, 4163– 4174. [23] Zhang, S., Huang, W.C., Li, P., Guo, H., Poh, S.B., Brady, S.W., Xiong, Y., Tseng, L.M., Li, S.H., Ding, Z., Sahin, A.A., Esteva, F.J., Hortobagyi, G.N. and Yu, D. (2011) Combating trastuzumab resistance by targeting SRC, a common node downstream of multiple resistance pathways. Nat. Med. 17, 461–469. [24] Biscardi, J.S., Ishizawar, R.C., Silva, C.M. and Parsons, S.J. (2000) Tyrosine kinase signalling in breast cancer: epidermal growth factor receptor and c-Src interactions in breast cancer. Breast Cancer Res. 2, 203–210. [25] Marcotte, R., Zhou, L., Kim, H., Roskelly, C.D. and Muller, W.J. (2009) C-Src associates with ErbB2 through an interaction between catalytic domains and confers enhanced transforming potential. Mol. Cell. Biol. 29, 5858–5871. [26] Grzesiek, S., Bax, A., Clore, G.M., Gronenborn, A.M., Hu, J.S., Kaufman, J., Palmer, I., Stahl, S.J. and Wingfield, P.T. (1996) The solution structure of HIV-1 Nef reveals an unexpected fold and permits delineation of the binding surface for the SH3 domain of Hck tyrosine proteinkinase. Nat. Struct. Biol. 3, 340–345.
Please cite this article in press as: Bornet, O., et al. Identification of a Src kinase SH3 binding site in the C-terminal domain of the human ErbB2 receptor tyrosine kinase. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.04.029
