Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Aug 30.
Published in final edited form as: J Mol Recognit. 2012 Aug;25(8):427–434. doi: 10.1002/jmr.2205

Dimerization in the Grb7 Protein

Tabitha A Peterson a, Renee L Benallie a, Andrew M Bradford a, Sally C Pias a, Jaron Yazzie a, Siamee N Lor b, Zachary M Haulsee b, Chad K Park c,a, Dennis L Johnson a, Larry R Rohrschneider d, Anne Spuches b, Barbara A Lyons a,*
PMCID: PMC3758474  NIHMSID: NIHMS403013  PMID: 22811067

Abstract

In previous studies, we showed that the tyrosine phosphorylation state of growth factor receptor–bound protein 7 (Grb7) affects its ability to bind to the transcription regulator FHL2 and the cortactin-interacting protein, human HS-1-associated protein-1. Here, we present results describing the importance of dimerization in the Grb7–Src homology 2 (SH2) domain in terms of its structural integrity and the ability to bind phosphorylated tyrosine peptide ligands. A tyrosine phosphorylation-mimic mutant (Y80E–Grb7–SH2) is largely dimerization deficient and binds a tyrosine-phosphorylated peptide representative of the receptor tyrosine kinase (RTK) erbB2 with differing thermodynamic characteristics than the wild-type SH2 domain. Another dimerization-deficient mutant (F99R–Grb7–SH2) binds the phosphorylated erbB2 peptide with similarly changed thermodynamic characteristics. Both Y80E–Grb7–SH2 and F99R–Grb7–SH2 are structured by circular dichroism measurements but show reduced thermal stability relative to the wild type–Grb7–SH2 domain as measured by circular dichroism and nuclear magnetic resonance. It is well known that the dimerization state of RTKs (as binding partners to adaptor proteins such as Grb7) plays an important role in their regulation. Here, we propose the phosphorylation state of Grb7–SH2 domain tyrosine residues could control Grb7 dimerization, and dimerization may be an important regulatory step in Grb7 binding to RTKs such as erbB2. In this manner, additional dimerization-dependent regulation could occur downstream of the membrane-bound kinase in RTK-mediated signaling pathways.

Keywords: adaptor proteins, Grb7, tyrosine kinase, erbB2, HER2, neu, cell migration, oligomerization, calorimetry

INTRODUCTION

Growth factor receptor–bound protein 7, 10 and 14 (Grb7, Grb10 and Grb14, respectively) comprise a family of Src homology 2 (SH2) domain-containing Grb proteins discovered through CORT screening (Ooi et al., 1995; Daly et al., 1996). A homologous Caenorhabditis elegans protein, Mig-10, is involved in neuronal migration during development (Manser et al., 1997). The members of this family all contain an N-terminal Pro-rich region, an Ras associating–like domain (Wojcik et al., 1999), a pleckstrin homology domain, a between pleckstrin and Src (BPS) domain and a C-terminal SH2 domain. The Ras associating, pleckstrin homology and BPS domains are also known as the Grbs and Mig domain. Mig-10 does not contain a C-terminal SH2 domain but instead replaces this with a Pro-rich sequence at the C-terminal (Manser et al., 1997).

Grb7 has been shown to bind to erbB2 (avian erythroblastosis oncogene B 2) and focal adhesion kinase (FAK) through its SH2 domain. ErbB2 and Grb7 are both overexpressed in a subset (20–30%) of breast cancer patients who share a poor long-term survival rate and a statistically high incidence of metastases (Slamon, 1987; Seshadri et al., 1993). The FAK/Grb7 association results in FAK phosphorylation of Grb7 at tyrosine residues and enhanced cell migration effects (Han et al., 2000). Chu et al. (Chu et al., 2009) have also shown that the tyrosine phosphorylation state of Grb7 is important in Grb7-mediated cell migration. The overexpression of Grb7 can lead to increased cell migration and development of metastases in cancer, as seen in the overexpression of Grb7 in lymph node metastases (Tanaka et al., 1998). In addition, a Grb7 protein variant, lacking its C-terminal region and termed Grb7V, has been linked to invasive esophageal carcinoma (Tanaka et al., 1998) and high-grade ovarian cancer (Wang et al., 2010).

The tyrosine phosphorylation state of Grb7 has been shown important in Grb7’s ability to interact with other signaling molecules. We have shown that the Grb7 protein binds to both FHL2 and human HS-1-associated protein X-1, and the tyrosine phosphorylation state of Grb7 is an important determinant in binding to both proteins (Siamakpour-Reihani et al., 2009; Siamakpour-Reihani et al., 2011). Tsai et al. (2007) showed that Grb7 acts as a translational repressor by binding to the mRNA 5′ UTR of the kappa opioid receptor and Grb7 binding to kappa opioid receptor mRNA is controlled by its tyrosine phosphorylation state.

There is evidence that members of the Grb7 protein family can exist in a multimeric form, and this form could participate in a regulatory mechanism. The Grb10-SH2 domain crystal structure (PDB ID No. 1NRV) indicates the SH2 domain exists as a dimer (Stein et al., 2003), whereas the gel filtration of Chinese hamster ovary cell lysates overexpressing full-length Grb10 implied the protein may exist in a tetrameric form, that is, a dimer of dimers (Dong et al., 1998). Depetris et al. (2005) showed the high-affinity binding of the Grb14 protein to the insulin receptor requires the BPS domain and a dimerized SH2 domain, based on the decreased inhibitory action of an SH2 dimerization defective Grb14 mutant.

Our own previous results show that the Grb7–SH2 domain undergoes a competing dimerization reaction in solution (PDB ID No. 1MW4; Ivancic et al., 2005), and a Grb7 crystal structure also presents the Grb7–SH2 domain as a dimer (PDB ID No. 2QMS; Porter et al., 2007). The presence of dimerized molecules by X-ray crystallographic methods can often result from crystallization artifacts and not correlate with the existence of actual dimers. In the case of the Grb7 protein family, however, the previous studies substantiate multimeric (primarily dimeric) forms of these proteins exist in solution.

The studies presented here seek to lay the groundwork for clarifying the role of dimerization and tyrosine phosphorylation state in Grb7 regulation and function.

MATERIALS AND METHODS

Protein expression and purification

The expression and the purification of the human Grb7–SH2 domain have been described previously (Brescia et al., 2002). Uniformly 15N-labeled Grb7–SH2 domain was produced in minimal media containing 1 g/L (15NH4)2SO4 (Cambridge Isotope Labs, Woburn, MA, USA). Typical yields of purified Grb7–SH2 domain are 4–6 mg/l of culture. The expression and the purification of the Y80E and F99R Grb7–SH2 domain mutants are identical to that of the wild type (WT)–Grb7–SH2 domain.

Mutagenesis

The site-directed mutagenesis of the F99 and Y80 residues of the Grb7–SH2 domain was performed by polymerase chain reaction according to the manufacturer’s guidelines, using the Quick-Change II mutagenesis kit (Invitrogen, Grand Island, NY, USA). Mutations were verified by DNA sequencing. The tyrosine 80 residue was selected for mutation based on its prediction as a possible phosphorylation site using the software NetPhos 2.0 (www.cbs.dtu.dk/zservices/NetPhos/).

Sample preparation for nuclear magnetic resonance, isothermal titration calorimetry and circular dichroism

Sample conditions for nuclear magnetic resonance (NMR) spectroscopy were as follows: 0.1–0.2 mM of protein, 1× phosphate-buffered saline and 5 mM of dithiothreitol, pH 7.4. Sample conditions for the isothermal titration calorimetry (ITC) studies were as follows: 30–50 μM Grb7–SH2 domain and 300–500 μM pY1139 peptide ligand, with replacement of dithiothreitol with tris(2-carboxyethyl)phosphine. The ITC conditions for the F99R and Y80E Grb7–SH2 domain mutants are the same as that for WT. The circular dichroism (CD) sample conditions were 0.5 mg/ml SH2 domain (WT, F99R, Y80D or Y80E) in 1× phosphate-buffered saline, pH 7.4.

Circular dichroism

CD spectra were recorded on a Jasco (Nantes, France) J-810 spectropolarimeter using 1-mm-thick quartz cells. In general, the Grb7–SH2 domain samples (WT, Y80E, F99R and Y80D) were analyzed in 1× phosphate buffer saline, at pH 7.4 and 25 °C. CD spectra were recorded with 0.5 mg/ml Grb7SH2 domain sample between 185 and 250 nm using a wavelength width of 2.0 nm and a scanning rate of 50 nm/min. For all spectra, the background spectrum of the buffer was subtracted.

Size exclusion chromatography

Size exclusion chromatography was performed on a Sephacryl S-200 HR (Amersham Biosciences, Pittsburgh, PA, USA) column pre-equilibrated with 1× phosphate-buffered saline, pH 7.40 at 4 °C. A volume of 0.5 ml of protein sample was injected onto the column and eluted with the same buffer at 0.5 ml/min. The absorbance of the eluate was monitored at 280 nm. The molecular weight of the SH2 domains was estimated by comparison of the elution volumes with those of gel filtration standards (Bio-Rad, Hercules, CA, USA). Initial concentrations on application to the column were as follows: WT, 756 μM; Y80F, 666 μM; Y80E, 974 μM; F99R, 678 μM; and Y80D, 1120 μM. Note that at these concentrations, more than 99.9% of the WT–Grb7–SH2 domain is in the dimerized form (Porter et al., 2005).

Isothermal titration calorimetry

Titrations were performed using a VP-ITC ultrasensitive isothermal titration calorimeter (MicroCal Corp., Northampton, MA, USA) (Wiseman et al., 1989). The Grb7–SH2 domain (WT, F99R or Y80E) was dialyzed with three exchanges of buffer (1/1000 v/v). To reduce errors arising from heats of dilution due to buffer differences between samples in the syringe and the reaction vessel, the pY1139 peptide ligand was dialyzed concurrently with the Grb7–SH2 sample. Both peptide and protein concentrations were determined using measured A280 values before loading into the syringe and reaction vessel. For each titration, 10 μl of peptide ligand (300–500 μM) was delivered into the protein sample (30–50 μM) during the 20-s duration with an adequate interval (250 s) allowing complete equilibration. Binding curves involved the addition of 30 injections, which enabled 50% saturation to occur by the fifteenth injection. The heats of dilution, obtained by titrating the identical peptide solution into the reaction cell containing only the sample buffer, were subtracted before analysis. Data were recorded at a baseline of 20 μcal/s and at a high gain mode, thus providing the fastest re-equilibration between injections. The syringe mixing speed was 300 rpm to ensure sufficient mixing while keeping baseline noise at a minimum. The data were collected automatically and analyzed with a one-site model via the Origin software provided from Microcal (ITC Origin, V7.0). The origin uses a nonlinear least-squares algorithm (minimization of χ2), the concentration of the titrant and sample and an equilibrium equation to provide best fit values of the binding curve thus providing stoichiometry (n), change in enthalpy (ΔH) and binding constant K (Zhang et al., 2000).

NMR spectroscopy

All NMR spectra of the 15N-labeled Grb7–SH2 domains (WT, F99R or Y80E) were obtained at the annotated temperatures (Results section) on a Varian Unity Plus 500 MHz spectrometer. The spectrometer was equipped with a 1H/13C/15N resonance pulsed-field gradient probe. Spectra were recorded in the States-TPPI mode for quadrature detection with carrier frequencies for 1H and 15N at 4.73 and 120.0 ppm, respectively. In the direct dimension, 1024 actual data points were collected, and in the indirect dimension, 128 actual data points were collected, with 64 scans per increment. Spectral widths were 2200 Hz in the F1 dimension and 8000 Hz in the F2 dimension. All 1H–15N correlation spectra were processed identically using instrument-supplied Varian VNMR software with sinebell squared (direct dimension) or sinebell (indirect dimension) apodization functions and zero filling to the next power of 2 in both dimensions.

Molecular dynamics simulations

Molecular dynamics simulations were carried out using chain A of the ligand-free Grb7–SH2 domain crystal structure (2QMS) (Porter et al., 2007) to generate starting models. The F99R and the Y80E point mutations were introduced in silico using the Swiss-PdbViewer software (Guex and Peitsch, 1997). All-atom molecular dynamics simulations were performed using AMBER software (version 9) (Case et al., 2006). Production simulations were conducted with constant pressure at 300 K (using Langevin dynamics to control temperature; Loncharich et al., 1992), following energy minimization and equilibration of the starting models. Simulations of the F99R–Grb7–SH2 mutant used a 1-femtosecond time step with a modified (Onufriev et al., 2000; Onufriev et al., 2004) generalized Born implicit solvent model (Still et al., 1990) with a distance cutoff of 16 Å for nonbonded interactions and a salt concentration of 150 mM (determined based on the solution conditions used to determine the wild-type Grb7–SH2 domain crystal structure; Porter et al., 2007). Production simulations for the F99R–Grb7–SH2 mutant were carried out with constant temperature for 1.8 ns, after 1000 cycles of energy minimization, 20 ps of heating and equilibration (from 0 to 300 K). Independent simulations of the Y80E–Grb7–SH2 mutant were conducted (i) in implicit solvent, as described previously for the F99R–Grb7–SH2 mutant, and (ii) in explicit solvent, with a 10-Å water box and with neutralizing ions. The additional simulations with explicit solvent were carried out for the Y80E mutant but not for the F99R mutant because we identified a putative stabilizing salt bridge involving the Y80E residue, and the generalized Born implicit solvent model is known to overestimate the strength of salt bridges (Okur et al., 2008). Explicit solvent production simulations of the Y80E–Grb7–SH2 mutant were carried out with constant pressure and temperature for 5 ns, after 3500 cycles of energy minimization, 20 ps of heating from 0 to 300 K and 200 ps of unrestrained equilibration. In all simulations, a Langevin thermostat, with a collision frequency of 1 per picosecond, was used for temperature control.

Analytical ultracentrifugation

Sedimentation velocity experiments were run in a Beckman XL-I analytical ultracentrifuge. Sample protein and buffer solution (400 μl each) were loaded into the appropriate sides of a two-sector shaped cell. After allowing thermal equilibration at 4 °C for >1 h, the sample was spun at 40 000 rpm. Absorbance scans were taken continuously for 17 h. The data were analyzed with SEDFIT, allowing for a continuous c(s) distribution (Schuck et al., 2002). The partial specific volume of the protein and the buffer density and viscosity were calculated from SEDNTERP (Harding et al., 1992; Lebowitz et al., 1998). Initial concentrations were 93 μM for Y80E and 79 μM for F99R. Note that at these concentrations, more than 99.9% of the WT–Grb7–SH2 domain is in the dimerized form (Porter et al., 2005).

RESULTS

A tyrosine phosphorylation mimic, Y80E–Grb7–SH2, is dimerization deficient

The Y80 residue of the Grb7–SH2 domain (Y495 in the full-length protein) is a predicted tyrosine phosphorylation site. Gel filtration studies of the Y80E–Grb7–SH2 mutant indicate that the domain is at least partially monomeric (Figure 1). Specifically, the Y80E–Grb7–SH2 domain elutes from a gel filtration column in a discreet band, at 71.1 ml elution volume, which converts to a molecular weight of 19 000. This is in comparison with the WT–Grb7–SH2 domain, which elutes over a broad volume range (65–76 ml), with a mean elution volume (65.9 ml) corresponding to an average molecular weight of 29 000. The previously characterized Grb7–SH2 dimerization-deficient mutant, F99R–Grb7–SH2 (Porter et al., 2005), eluted from the same column at 74.6 ml, corresponding to a molecular weight of 11 000 (molecular weight of WT–Grb7–SH2, 13 800). The conservative mutation, Y80F, elutes at the same volume as the WT–Grb7–SH2 domain, demonstrating that loss of dimerization is specific to the glutamate mutation (i.e. negating the possibility that any mutation at the Y80 position causes loss of dimerization).

Figure 1.

Figure 1

Open in a new tab

Size exclusion chromatography results for the WT–Grb7–SH2, F99R–Grb7–SH2, Y80E–Grb7–SH2, Y80F–Grb7–SH2 and Y80D–Grb7–SH2 domains. The bottom panel contains the elution profile for each Grb7–SH2 domain. The y-axis is the absorption signal at 280 nm, whereas the x-axis is the ratio of the protein elution volume divided by the column void volume (34 ml). The top panel shows the calibration curve determined by running the following molecular weight standards through the column: conalbumin, 75 000 MW; carbonic anhydrase, 29 000 MW; ribonuclease A, 13 700 MW. The y-axis plots the molecular weight, whereas the x-axis shows the molecular weight standard protein elution volume divided by the column void volume.

Analytical ultracentrifugation studies of the Y80E–Grb7–SH2 and F99R–Grb7–SH2 domain mutants agree with the gel filtration findings. Figure 2 shows the sedimentation velocity results for the F99R–Grb7–SH2 (dotted line) and Y80E–Grb7–SH2 (solid line), as a function of the sedimentation coefficient distribution (c(s), y-axis) versus the Svedberg constant ([S], x-axis). Fitting the data yields an apparent MW of 12.2 kDa for the F99R–Grb7–SH2 domain and 13.3 kDa for the Y80E–Grb7–SH2 domain. In sedimentation equilibrium experiments (data not shown), both mutants can be fit to a monomer–dimer equilibrium; however, the determined Kds are 1600 μM (F99R) and 300–590 μM (Y80E). These Kds are considerably larger than the Kd determined for the WT–Grb7–SH2 domain dimerization (11–22 μM; Porter et al., 2005; Porter et al., 2007). A small peak at a Svedberg value [S] of approximately 2.9 (indicated by an asterisk) is due to a glutathione S-transferase-dimer impurity, resulting from the expression and purification of the recombinant Grb7–SH2 domains (band at approximately 27–28 kDa in the denaturing sodium dodecyl sulfate–polyacrylamide gel electrophoresis; lane 4, Figure 2).

Figure 2.

Figure 2

Open in a new tab

Left panel: sedimentation velocity results for the F99R–Grb7–SH2 (dotted line) and Y80E–Grb7–SH2 (solid line) domains. Right panel: gradient sodium dodecyl sulfate–polyacrylamide gel electrophoresis analysis of the F99R–Grb7–SH2 domain sample for analytical ultracentrifugation study. Lane 4 is the essentially pure F99R–Grb7–SH2 domain.

Dimerization-deficient mutants of the Grb7–SH2 (F99R–Grb7–SH2 and Y80E–Grb7–SH2) domain are structurally destabilized

CD spectra were recorded for the WT–Grb7–SH2, F99R–Grb7–SH2 and Y80E–Grb7–SH2 domains. The spectra are similar (Figure S2, top panel), indicate all three protein domains are folded and consist of a mixture of α-helical and β-sheet secondary structure. The results for the WT–Grb7–SH2 and F99R–Grb7–SH2 are in agreement with a previous report that the F99R–Grb7–SH2 domain mutant is structured (Porter et al., 2005). In addition, the 15N–1H correlation [nitrogen heteronuclear single quantum coherence (Nhsqc)] NMR spectra of all three SH2 domains show that the proteins are structured, displaying well-defined resonances and large chemical shift dispersion (Figure S1).

CD spectra were recorded in 10 °C increments from 25 °C to 65 °C for all three Grb7–SH2 domains (WT, F99R or Y80E). An accepted approach for monitoring protein denaturation (structural loss) by CD temperature melting is to plot the normalized ellipticity at 222 and/or 218 nm (Greenfield, 2006; Orwig and Lieberman, 2011) (Figure 3). In addition, the full CD traces for all three SH2 domains at different temperatures are provided in Figure S2. Residual secondary structure remains at 55 °C for the WT–Grb7–SH2 domain (Figures 3 and S2, center panel), whereas all secondary structure is absent for the Y80E–Grb7–SH2 and F99R–Grb7–SH2 domain mutants at this temperature. By 65 °C, all three Grb7–SH2 domains (WT, F99R or Y80E) are unstructured by CD measurement (Figures 3 and S2, lower panel). It is unlikely the changes in ellipticity for any of these SH2 domains is simply due to aggregation in that all three SH2 domains reach a constant ellipticity value at higher temperatures (i.e. the denaturation curve fiattens at the top; John and Weeks, 2000).

Figure 3.

Figure 3

Open in a new tab

CD temperature denaturation study of the WT–Grb7–SH2, F99R–Grb7–SH2 and Y80E–Grb7–SH2 domains monitored at 222 nm wavelength. Experimental details can be found in the Materials and Methods section. X, WT–Grb7–SH2 domain; O, F99R–Grb7–SH2 domain; +, Y80E–Grb7–SH2 domain.

An Nhsqc NMR analysis of all three Grb7–SH2 domains (WT, F99R or Y80E) was performed at temperatures ranging in 5 °C increments from 25 °C to 60 °C and revealed residues that experienced early loss of structure at lower temperatures (40 °C–45 °C) versus residues whose resonances remained at higher temperature (50 °C–55 °C) (Figure S1). Early loss of structure is indicated in residues 1–27 and 60–77, whereas residues 30–52 remained structured at 50 °C (Figure 4). The regions showing early loss of structure are the N-terminus, βD and βD′ strands and D′E loop of the SH2 domain. The regions of the SH2 domain that remain structured at 50 °C include most of the αA helix and βB strand. By 60 °C, the distinctive Nhsqc fingerprint representative of a folded polypeptide had essentially collapsed and/or disappeared for WT–Grb7–SH2, indicating a primarily random coil and/or denatured peptide (Figure S1). This loss of the Nhsqc finger-print pattern occurred approximately 10 °C earlier for the F99R–Grb7–SH2 and Y80E–Grb7–SH2 domains, indicating reduced thermal stability for these mutants that corroborated the CD thermal denaturation results.

Figure 4.

Figure 4

Open in a new tab

Ribbon representation of the NMR thermal denaturation study results for the WT–Grb7–SH2 domain. Two different orientations of the Grb7–SH2 are shown. Left panel: side view; right panel: top view. Red coloration represents regions that showed early loss of resonances (before 40 °C, Figure S1): residues 1–27 and 60–77. Blue coloration represents regions that retained resonances at 50 °C: residues 30–52. Grey coloration represents regions of mixed early loss of resonances at 40 °C and retention of resonances at 50 °C.

The binding of a tyrosine phosphorylated peptide to dimerization-deficient Grb7–SH2 domain mutants is entropically disfavored

Quantitative binding affinity measurements by ITC were recorded for the F99R–Grb7–SH2 (Figure 5A) and Y80E–Grb7–SH2 (Figure 5B) domains binding to a 10 amino acid phosphorylated tyrosine peptide (pY1139) representative of the Grb7 binding site on erbB2 at 25 °C (Table 1).

Figure 5.

Figure 5

Open in a new tab

(A, top) Exothermic heats of reaction (μcal/s) measured for 35 injections of the pY1139 peptide into an F99R–Grb7–SH2 domain sample. (Bottom) The solid line is the best fit to the data for a single binding site model using a nonlinear least squares fit to solve for Kb (equilibrium binding constant, 1/Kd) and ΔH (change in enthalpy of reaction), as reported in the Results section and Table 1. (B, top) Exothermic heats of reaction (μcal/s) measured for 35 injections of the pY1139 peptide into a Y80E–Grb7–SH2 domain sample. (Bottom) The solid line is the best fit to the data for a single binding site model using a nonlinear least squares fit to solve for Kb (1/Kd ) and ΔH, as reported in the Results section and Table 1.

Table 1.

Thermodynamic parameters for the binding of the pY1139 peptide to the Grb7–SH2 domains at 25 °C

Grb7–SH2 domain Kd (μM) ΔG (kcal mol−1) ΔH (kcal mol−1) −TΔS (kcal mol−1)
WTa 2.28 ± 0.15 −7.70 ± 0.04 −4.66 ± 0.05 −3.04 ± 0.09
F99R–Grb7–SH2b 9.29 ± 2.17 −6.89 ± 0.13 −34.0 ± 10.2 26.9 ± 10.3
Y80E–Grb7–SH2c 2.29 ± 0.43 −7.69 ± 0.10 −46.4 ± 12.5 38.7 ± 12.6
Open in a new tab
a

The reported values are the average for five separate experiments.

b

The reported values are the average for three separate experiments.

c

The reported values are the average for two separate experiments.

The F99R–Grb7–SH2 domain binds the erbB2 peptide with somewhat lower affinity (average Kd = 9.29 μM), whereas the Y80E–Grb7–SH2 binds the erbB2 peptide with similar affinity (average Kd = 2.29 μM) compared with the WT–Grb7–SH2 domain (Kd = 2.28 μM; Ivancic et al., 2005). For both mutants, the interaction is enthalpically favored (F99R–Grb7–SH2, ΔH = −34.0 kcal/mol; Y80E–Grb7–SH2, ΔH = −46.4 kcal/mol) and entropically disfavored (F99R–Grb7–SH2, −TΔS = +26.9 kcal/mol; Y80E–Grb7–SH2, −TΔS = +38.7 kcal/mol). This is in contrast to the WT–Grb7–SH2 domain binding pY1139, which is both enthalpically and entropically favored (ΔH = −4.66 kcal/mol, −TΔS = −3.04 kcal/mol; Ivancic et al., 2005).

Because the Grb7 protein (and its SH2 domain in particular for this study) has no enzymatic activity, no enzymatic assay exists to assess the functionality of the SH2 domain mutants. However, the ability of the mutants F99R and Y80E to bind tyrosine phosphorylated peptides can act as a demonstration of functionality. Therefore, Table 1 demonstrates that both Y80E–Grb7–SH2 and F99R–Grb7–SH2 domains bind the erbB2-representative peptide pY1139 within a factor of 5 compared with the WT–Grb7–SH2 domain.

Furthermore, it is assured that the WT–Grb7–SH2 domain exists essentially as a dimer (i.e. no monomeric WT–Grb7–SH2 domain present) for the ITC studies. The Kd for dimerization of the Grb7–SH2 domain is 11 μM (Porter et al., 2005). All ITC experiments were performed at concentrations considerably greater than this, for example, 30 μM. At 11 μM greater than 99.9% of the WT–Grb7–SH2 domain is in the dimerized form.

Finally, these studies provide the first steps for establishing the dimerization characteristics of these Grb7–SH2 domain mutants. It is acknowledged that not all protein for the Y80E mutant may be binding the pY1139 peptide (as evidenced by the binding stoichiometry), and the F99R mutant may suffer from aggregation effects (Porter et al., 2005).

Molecular dynamics simulations indicate phosphorylation could favor the monomeric form of the Grb7–SH2 domain

The 2QMS Grb7–SH2 crystal structure (Porter et al., 2007) provided a starting model for a dimerized form of the SH2 domain, as opposed to the monomeric NMR-derived structure (Ivancic et al., 2003). The computational mutation of the tyrosine 80 residue (tyrosine 492 in the 2QMS Grb7–SH2 domain structure) to glutamate and energy minimization was performed on chain A before molecular dynamics simulations (described in Materials and Methods section). Molecular dynamics simulations in implicit solvent (1.8 ns) and in explicit solvent (independent simulations, 5 ns) suggest that conformational fluctuations in the βD′/EF-loop region may be minimized because of the formation of a salt-bridge interaction between Arg89 and Glu80 (Figure 6B). Such salt-bridge stabilization would discourage opening in the region necessary for establishment of a possible domain-swapped dimeric form (proposed by Pias et al., 2010) and would favor the monomeric conformation of the SH2 domain.

Figure 6.

Figure 6

Open in a new tab

Comparison of interactions involving residue R89 in the wild-type Grb7–SH2 domain (green) and in the Y80E–Grb7–SH2 mutant (orange). (A) Amino–pi stacking interaction involving the Y80 and R89 side chains in the wild-type domain (chain A of the Grb7–SH2 domain crystal structure, 2QMS). (B) Salt bridge formed by side chain atoms of residues E80 and R89 in the Y80E–Grb7–SH2 mutant. Shown is a snapshot collected at the end of a 5-ns explicit solvent molecular dynamics simulation. Figure rendered with PyMOL (http://www.pymol.org/).

We additionally modeled the F99R-Grb7-SH2 domain mutant using in silico mutation along with molecular dynamics simulations (1.8 ns in implicit solvent, described in Materials and Methods section). The in silico introduction of the F99R mutation demonstrates that the substitution of arginine for phenylalanine acts as a partially ‘conservative’ mutation. In the wild-type protein, a key hydrophobic interaction occurs between F90 and F99, in which the F99R–Grb7–SH2 mutant is replaced with a hydrophobic interaction involving F90 and the aliphatic region of the R99 side chain (Figure 6A).

DISCUSSION

The identification of a drug target is not sufficient for development of a successful treatment. This is amply demonstrated by the hundreds of ‘targets’ identified, coupled with the small number of actual targeted therapies available. For successful drug development, a basic understanding of how the target molecule functions and is regulated in its physiological state is needed. With this in mind, it is logical to think that the oligomerization state of a molecule can affect its ability to engage in signaling and also its binding to inhibitors. Our studies here have sought to begin understanding the role of dimerization in the Grb7 protein and provide a foundation for exploitation of this knowledge in the development of Grb7 signaling modulators.

The wild-type Grb7 protein exists primarily in dimeric form, both as the full-length protein and in a truncated form consisting of only the Grb7–SH2 domain (Ivancic et al., 2003; Porter et al., 2005; Porter et al., 2007). The binding of the Grb7–SH2 domain to the pY1139 peptide representative of erbB2 results in at least partial loss of dimerization (Ivancic et al., 2005) and is both an enthalpically and entropically favored process. It can be argued that pY1139 binding, driving monomerization of the SH2 domain, results in the entropically favored situation of change from a one-molecule form of Grb7 (i.e. dimerized) to a two-molecule Grb7 form. This favorable increase in entropy could potentially offset the disfavorable change in entropy that must result from ordering of residues at the erbB2/Grb7 binding interface. In the Grb7–SH2 dimerization-deficient mutants F99R and Y80E, binding to the pY1139 erbB2 peptide is enthalpically favored but entropically disfavored. This is consistent with an inability to monomerize on erbB2 peptide binding and subsequent entropic compensation for residue ordering at the Grb7/erbB2 binding interface. Thus, the loss of Grb7 dimerization may supply favorable entropic compensation during or after binding to the erbB2 receptor tyrosine kinase.

Dimerization itself and the structural stability afforded by dimerization over the monomeric state (as evidenced by our CD and NMR denaturation studies) may provide a form of Grb7 autoregulation by enhancing the lifetime of the nonsignaling form of the protein. In this model, upon binding to a signaling partner, such as erbB2, phosphorylation drives the Grb7 protein to the monomeric state, with the binding interaction to erbB2 potentially compensating for the loss of dimerization structural stability. The unequivocal demonstration of the effect of phosphorylation on dimerization state in the Grb7–SH2 domain would be demonstrated by successful in vitro tyrosine phosphorylation followed by the characterization of the SH2 domain molecular weight by size exclusion chromatography and/or AUC. In vitro phosphorylation of the full-length Grb7 protein has been achieved by Tsai et al. (2007). In vitro phosphorylation of the Grb7–SH2 domain alone using both FAK and Src kinases is presently underway in our laboratory.

We have recently performed molecular dynamics simulations on the wild-type Grb7–SH2 domain, using the 2QMS Grb7–SH2 domain crystal structure (Porter et al., 2007) as a starting model for the dimerized form (Pias et al., 2010). These calculations showed that the Grb7–SH2 domain could potentially undergo a conformational change resulting in an energetically stable domain-swapped form, akin to the domain-swapped conformational form of the Grb2-SH2 domain (Schiering et al., 2000; Nioche et al., 2002; Benfield et al., 2007). Further molecular dynamics simulations on the Y80E–Grb7–SH2 domain, in this study, provide an argument for the preferred monomeric state of this mutant. From these simulations, we have hypothesized that a salt bridge formed between Arg89 and Glu80 stabilizes the βD′/EF-loop region of the Grb7–SH2 domain, pushing the reaction equilibrium towards the monomeric instead of the domain-swapped dimeric form (Figure 6B).

It is not readily apparent how the Y80E phosphorylation mimic mutation would disrupt ‘unswapped’ dimerization. The dimerization interface in the 2QMS Grb7–SH2 crystal structure appears to be composed almost entirely of the F99 and F90 residues (F502 and F511 in the 2QMS structure numbering) in both the WT and Y80E Grb7–SH2 domains. These four residues (two Phe residues from each monomer) form a hydrophobically packed dimer interface and may also engage in stabilizing pi-stacking interactions. The molecular dynamics simulations on the Y80E–Grb7–SH2 mutant suggest that these interactions are maintained in the unswapped dimer form, suggesting no stabilization of the monomer over the unswapped dimer. The results are unequivocal, however, that the Y80E–Grb7–SH2 domain phosphorylation mimic is dimerization deficient. These simulation results do not preclude the possibility that Grb7 may simultaneously exist in three forms: monomer, unswapped dimer and swapped dimer.

Molecular dynamics simulations on the F99R–Grb7–SH2 domain indicate that the dimerized form would be disfavored in both the swapped and unswapped dimeric cases because of the disruption of the hydrophobic packing among the four phenylalanine residues that form the dimer interface.

Another SH2-containing protein, Grb2, has been the subject of targeted therapy development, with little clinical success. Inhibitors to the Grb2-SH2 domain suffer from low physiological binding and/or lack of specificity. It has been shown that the dimerized form of the Grb2-SH2 domain binds with lower affinity than the monomer form to tyrosine phosphorylated peptides representative of intracellular binding partners (Benfield et al., 2007). In addition, the existence of a domain-swapped dimerized form of the Grb2-SH2 domain has been shown to exist (Schiering et al., 2000; Nioche et al., 2002). The potentially important contribution of the dimerized form of the Grb2-SH2 domain to binding has been largely disregarded in the design and development of Grb2 inhibitors.

Our study suggests that dimerization in the Grb7 protein may be controlled by its tyrosine phosphorylation state. The monomeric versus dimeric state of the Grb7–SH2 domain binds the pY1139 erbB2 peptide with differing thermodynamic characteristics and demonstrates decreased structural stability. Such factors may be important considerations in the development of effective therapeutics targeting Grb7.

Abbreviations and symbols

erbB2

erythroblastic leukemia viral (v-erb-b) oncogene homolog 2, also known as HER2

CD

circular dichroism

FAK

focal adhesion kinase

Grb7,10,14

growth factor receptor–bound protein 7, 10 and 14 protein, respectively

ITC

isothermal titration calorimetry

Nhsqc

nitrogen heteronuclear single quantum coherence

NMR

nuclear magnetic resonance

pY

phosphorylated tyrosine

SH2

Src homology 2

Footnotes

Supporting information may be found in the online version of this paper.

References

  1. Benfield AP, Whiddon BB, Clements JH, Martin SF. Structural and energetic aspects of Grb2-SH2 domain-swapping. Arch Biochem Biophys. 2007;462:47–53. doi: 10.1016/j.abb.2007.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Brescia PJ, Ivancic M, Lyons BA. Assignment of backbone 1H, 13C, and 15N resonances of human Grb7–SH2 domain in complex with a phosphorylated peptide ligand. J Biomol NMR. 2002;23:77–78. doi: 10.1023/a:1015345302576. [DOI] [PubMed] [Google Scholar]
  3. Case DA, Darden TA, Cheatham TE, Simmerling CL, III, Wang J, Duke RE, Luo R, Merz KM, Pearlman DA, Crowley M, Walker RC, Zhang W, Wang B, Hayik S, Roitberg A, Seabra G, Wong KF, Paesani F, Wu X, Brozell S, Tsui V, Gohlke H, Yang L, Tan C, Mongan J, Hornak V, Cui G, Beroza P, Mathews DH, Schafmeister C, Ross WS, Kollman PA. AMBER. Vol. 9. University of California; San Francisco: 2006. [Google Scholar]
  4. Chu PY, Huang LY, Hsu CH, Liang CC, Guan JL, Hung TH, Shen TL. Tyrosine phosphorylation of Grb7 by FAK in the regulation of cell migration, proliferation, and tumorigenesis. J Biol Chem. 2009;284(30):20215–26. doi: 10.1074/jbc.M109.018259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Daly RJ, Sanderson GM, Janes PW, Sutherland RL. Cloning and characterization of GRB14, a novel member of the GRB7 gene family. J Biol Chem. 1996;271:12502–12510. doi: 10.1074/jbc.271.21.12502. [DOI] [PubMed] [Google Scholar]
  6. Depetris RS, Hu J, Gimpelevich I, Holt LJ, Daly RJ, Hubbard SR. Structural basis for inhibition of the insulin receptor by the adaptor protein Grb14. Mol Cell. 2005;20:325–333. doi: 10.1016/j.molcel.2005.09.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dong LQ, Porter S, Hu D, Liu F. Inhibition of hGrb10 binding to the insulin receptor by functional domain-mediated oligomerization. J Biol Chem. 1998;273:17720–17725. doi: 10.1074/jbc.273.28.17720. [DOI] [PubMed] [Google Scholar]
  8. Greenfield N. Using circular dichroism collected as a function of temperature to determine the thermodynamics of protein unfolding and binding interactions. Nat Protoc. 2006;1:2527–2535. doi: 10.1038/nprot.2006.204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Guex N, Peitsch MC. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis. 1997;18:2714–2723. doi: 10.1002/elps.1150181505. [DOI] [PubMed] [Google Scholar]
  10. Han DC, Shen TL, Guan JL. Role of Grb7 targeting to focal contacts and its phosphorylation by focal adhesion kinase in regulation of cell migration. J Biol Chem. 2000;275:28911–28917. doi: 10.1074/jbc.M001997200. [DOI] [PubMed] [Google Scholar]
  11. Harding SE, Rowe AJ, Horton JC, editors. Analytical Ultracentrifugation in Biology and Polymer Science. Royal Society of Chemistry; Cambridge: 1992. [Google Scholar]
  12. Ivancic M, Daly RJ, Lyons BA. Solution structure of the human Grb7–SH2 domain/erbB2 peptide complex and structural basis for Grb7 binding to ErbB2. J Biomol NMR. 2003;27:205–219. doi: 10.1023/a:1025498409113. [DOI] [PubMed] [Google Scholar]
  13. Ivancic M, Spuches AM, Guth EC, Daugherty MA, Wilcox DE, Lyons BA. Backbone nuclear relaxation characteristics and calorimetric investigation of the human Grb7–SH2/erbB2 peptide complex. Protein Sci. 2005;14:1556–1569. doi: 10.1110/ps.041102305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. John D, Weeks K. van’t Hoff enthalpies without baselines. Protein Sci. 2000;9:1416–1419. doi: 10.1110/ps.9.7.1416. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Lebowitz J, Teale M, Schuck PW. Analytical band centrifugation of proteins and protein complexes. Biochem Soc Trans. 1998;26:745–749. doi: 10.1042/bst0260745. [DOI] [PubMed] [Google Scholar]
  16. Loncharich RJ, Brooks BR, Pastor RW. Langevin dynamics of peptides: The frictional dependence of isomerization rates of N-acetylalanyl-N-prime-methylamide. Biopolymers. 1992;32:523–535. doi: 10.1002/bip.360320508. [DOI] [PubMed] [Google Scholar]
  17. Manser J, Roonprapunt C, Margolis B. C. elegans cell migration gene mig-10 shares similarities with a family of SH2 domain proteins and acts cell nonautonomously in excretory canal development. Dev Biol. 1997;184:150–164. doi: 10.1006/dbio.1997.8516. [DOI] [PubMed] [Google Scholar]
  18. Nioche P, Liu WQ, Broutin I, Charbonnier F, Latreille MT, Vidal M, Roques B, Garbay C, Ducruix A. Crystal structures of the SH2 domain of Grb2: highlight on the binding of a new high-affinity inhibitor. J Mol Biol. 2002;315:1167–1177. doi: 10.1006/jmbi.2001.5299. [DOI] [PubMed] [Google Scholar]
  19. Okur A, Wickstrom L, Simmerling C. Evaluation of Salt Bridge Structure and Energetics in Peptides Using Explicit, Implicit, and Hybrid Solvation Models. Journal of Chemical Theory and Computation. 2008;4:488–498. doi: 10.1021/ct7002308. [DOI] [PubMed] [Google Scholar]
  20. Onufriev A, Bashford D, Case DA. Modification of the Generalized Born Model Suitable for Macromolecules. J Phys Chem B. 2000;104:3712–3720. [Google Scholar]
  21. Onufriev A, Bashford D, Case DA. Exploring protein native states and large-scale conformational changes with a modified generalized born model. Proteins. 2004;55:383–394. doi: 10.1002/prot.20033. [DOI] [PubMed] [Google Scholar]
  22. Ooi J, Yajnik V, Immanuel D, Gordon M, Moskow JJ, Buchberg AM, Margolis B. The cloning of Grb10 reveals a new family of SH2 domain proteins. Oncogene. 1995;10:1621–1630. [PubMed] [Google Scholar]
  23. Orwig S, Lieberman R. Biophysical Characterization of the Olfactomedin Domain of Myocilin, an Extracellular Matrix Protein Implicated in Inherited Forms of Glaucoma. PLoS One. 2011;6:e16347. doi: 10.1371/journal.pone.0016347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pias S, Peterson TA, Johnson DL, Lyons BA. The intertwining of structure and function: proposed helix-swapping of the SH2 domain of Grb7, a regulatory protein implicated in cancer progression and inflammation. Crit Rev Immunol. 2010;30:299–304. doi: 10.1615/critrevimmunol.v30.i3.70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Porter CJ, Matthews JM, Mackay JP, Pursglove SE, Schmidberger JW, Leedman PJ, Pero SC, Krag DN, Wilce MC, Wilce JA. Grb7 SH2 domain structure and interactions with a cyclic peptide inhibitor of cancer cell migration and proliferation. BMC Struct Biol. 2007;7:58. doi: 10.1186/1472-6807-7-58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Porter CJ, Wilce MC, Mackay JP, Leedman P, Wilce JA. Grb7–SH2 domain dimerisation is affected by a single point mutation. Eur Biophys J. 2005;34:454–460. doi: 10.1007/s00249-005-0480-1. [DOI] [PubMed] [Google Scholar]
  27. Schiering N, Casale E, Caccia P, Giordano P, Battistini C. Dimer formation through domain swapping in the crystal structure of the Grb2-SH2-Ac-pYVNV complex. Biochemistry. 2000;39:13376–13382. doi: 10.1021/bi0012336. [DOI] [PubMed] [Google Scholar]
  28. Schuck P, Perugini MA, Gonzales NR, Howlett GJ, Schubert D. Size-distribution analysis of proteins by analytical ultracentrifugation: strategies and application to model systems. Biophys J. 2002;82:1096–1111. doi: 10.1016/S0006-3495(02)75469-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Seshadri R, Firgaira FA, Horsfall DJ, McCaul K, Setlur V, Kitchen P. Clinical significance of HER-2/neu oncogene amplification in primary breast cancer. The South Australian Breast Cancer Study Group. J Clin Oncol. 1993;11:1936–1942. doi: 10.1200/JCO.1993.11.10.1936. [DOI] [PubMed] [Google Scholar]
  30. Siamakpour-Reihani S, Argiros HJ, Wilmeth LJ, Haas LL, Peterson TA, Johnson DL, Shuster CB, Lyons BA. The cell migration protein Grb7 associates with transcriptional regulator FHL2 in a Grb7 phosphorylation-dependent manner. J Mol Recognit. 2009;22:9–17. doi: 10.1002/jmr.916. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Siamakpour-Reihani S, Peterson TA, Bradford AM, Argiros HJ, Haas LL, Lor SN, Haulsee ZM, Spuches AM, Johnson DL, Rohrschneider LR, Shuster CB, Lyons BA. Grb7 binds to Hax-1 and undergoes an intramolecular domain association that offers a model for Grb7 regulation. J Mol Recognit. 2011;24:314–321. doi: 10.1002/jmr.1062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Slamon DJ. Proto-oncogenes and human cancers. N Engl J Med. 1987;317:955–957. doi: 10.1056/NEJM198710083171509. [DOI] [PubMed] [Google Scholar]
  33. Stein EG, Ghirlando R, Hubbard SR. Structural basis for dimerization of the Grb10 Src homology 2 domain. Implications for ligand specificity. J Biol Chem. 2003;278:13257–13264. doi: 10.1074/jbc.M212026200. [DOI] [PubMed] [Google Scholar]
  34. Still WC, Tempczyk A, Hawley RC, Hendrickson T. Semianalytical treatment of solvation for molecular mechanics and dynamics. J Am Chem Soc. 1990;112:6127–6129. [Google Scholar]
  35. Tanaka S, Mori M, Akiyoshi T, Tanaka Y, Mafune K, Wands JR, Sugimachi K. A novel variant of human Grb7 is associated with invasive esophageal carcinoma. J Clin Invest. 1998;102:821–827. doi: 10.1172/JCI2921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tsai NP, Bi J, Wei LN. The adaptor Grb7 links netrin-1 signaling to regulation of mRNA translation. EMBO J. 2007;26:1522–1531. doi: 10.1038/sj.emboj.7601598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wang Y, Chan DW, Liu VW, Chiu P, Ngan HY. Differential functions of growth factor receptor–bound protein 7 (GRB7) and its variant GRB7v in ovarian carcinogenesis. Clin Cancer Res. 2010;16:2529–2539. doi: 10.1158/1078-0432.CCR-10-0018. [DOI] [PubMed] [Google Scholar]
  38. Wiseman T, Williston S, Brandts JF, Lin LN. Rapid measurement of binding constants and heats of binding using a new titration calorimeter. Anal Biochem. 1989;179:131–137. doi: 10.1016/0003-2697(89)90213-3. [DOI] [PubMed] [Google Scholar]
  39. Wojcik J, Girault JA, Labesse G, Chomilier J, Mornon JP, Callebaut I. Sequence analysis identifies a ras-associating (RA)-like domain in the N-termini of band 4.1/JEF domains and in the Grb7/10/14 adapter family. Biochem Biophys Res Commun. 1999;259:113–120. doi: 10.1006/bbrc.1999.0727. [DOI] [PubMed] [Google Scholar]
  40. Zhang Y, Akilesh S, Wilcox DE. Isothermal titration calorimetry measurements of Ni(II) and Cu(II) binding to His, GlyGlyHis, HisGlyHis, and bovine serum albumin: a critical evaluation. Inorg Chem. 2000;39:3057–3064. doi: 10.1021/ic000036s. [DOI] [PubMed] [Google Scholar]

RESOURCES