Carbon Nanotube Wins the Competitive Binding over Proline-Rich

ARTICLE pubs.acs.org/JPCC

Carbon Nanotube Wins the Competitive Binding over Proline-Rich Motif Ligand on SH3 Domain Guanghong Zuo,†,‡ Wei Gu,§ Haiping Fang,*,†,‡ and Ruhong Zhou*,||,^ †

Shanghai Institute of Applied Physics, Chinese Academy of Sciences, P.O. Box 800-204, Shanghai 201800, P. R. China T-Life Research Center, Department of Physics, Fudan University, Shanghai 200433, P. R. China § Zentrum f€ur Bioinformatik, Universit€at des Saarlandes, 66041 Saarbr€ucken, Germany IBM Thomas J. Watson Research Center, Yorktown Heights, New York 10598, United States ^ Department of Chemistry, Columbia University, New York, New York 10027, United States

)

‡

bS Supporting Information ABSTRACT: The binding competition between a proline-rich motif (PRM) ligand and a hydrophobic nanoparticle, the singlewall carbon nanotube (SWCNT), at the binding pocket of SH3 domain, has been investigated by molecular dynamics simulations. It is found that the SWCNT has a very high probability of occupying the binding pocket of the SH3 domain, which prevents the PRM ligand from binding to the pocket. The binding free energy landscapes show that the SWCNT has ∼0.6 kcal/mol stronger binding affinity than the ligand in the three-way binding competition (SWCNT þ ligand þ protein). The potent binding affinity between the SWCNT and the SH3 domain is shown to be mainly from the π
π stacking interactions between the CNT and aromatic residues in the binding pocket. Our findings show that the existence of hydrophobic particles can greatly reduce the possibility of the regular binding of the ligand with the target protein, suggesting potential toxicity to proteins by hydrophobic nanoscale particles.

’ INTRODUCTION The understanding of the properties of nanoscale particles and its interactions with various materials plays a key role in nanoscience and nanotechnology. Particularly, the interactions between the nanoscale particles and biomolecules are essential to the nanoparticle-based biotechnology and biomedical applications, such as gene delivery,1 cellular imaging,2 tumor therapy,3 and biological experimental technology.4,5 On the other hand, there are also growing concerns on the biosafety of these nanoparticles.6
11 There have been extensive recent studies on the adsorption of proteins on nanoparticles, both experimentally and theoretically. It has been found that the adsorption usually changes the structures of the proteins. For example, GoldbergOppenheimer and Regev explored the conformational changes of bovine serum albumin (BSA) binding with a single-walled carbon nanotube (SWCNT) via cryogenic temperature transmission electron microscopy.12 A recent molecular dynamics simulation showed that the conformation of a subdomain of human BSA could be significantly affected by its adsorption onto carbon nanotube surfaces.13 There are also some discussions on the protein function disruptions due to the existence of the nanoscale particles. In 2003, Park et al. showed experimentally that the SWCNTs with suitable sizes and shapes were effective in blocking some biological membrane ion channel.14 Karajanagi et al. also investigated the conformational and active site changes r 2011 American Chemical Society

of proteins after adsorbing onto the SWCNT surfaces, using atomic force microscope and Fourier transform infrared spectroscopy.15 Very recently, with molecular dynamics simulations, we found that the carbon nanotubes could plug into the hydrophobic cores of WW domains to form stable complexes, which disrupted and blocked the active sites of WW domains from binding to the corresponding ligands, thus leading to the loss of the original function.16 All the aforementioned studies focus on the direct interactions of nanoscale particles with proteins. However, the effect of nanoscale particles on the ligand
receptor binding, i.e., a three way SWCNT
ligand
protein competitive binding, is much less studied. In this paper, we present an approach toward this direction by studying the interaction among a nanoparticle, a ligand, and a target protein with molecular dynamics simulations. As a model system, we take a proline-rich ligand (PPPVPPRR) and its binding module, the SH3 domain, together with a pristine SWCNT as an example to illustrate the idea. Here the SH3 domain is a protein domain in signaling and regulatory proteins as the functional module to identify and bind proline-rich motif (PRM) of their binding partners,17
23 which is one of the most Received: March 21, 2011 Revised: May 5, 2011 Published: June 02, 2011 12322

dx.doi.org/10.1021/jp2026303 | J. Phys. Chem. C 2011, 115, 12322–12328

The Journal of Physical Chemistry C abundant protein interaction domains.24 Thus, if the existence of the SWCNTs shows a significant impact on the recognition between PRMs and the SH3 domain, it would indicate the potential toxicity of SWCNTs to these proteins. Our results show that the SWCNT has a higher probability than the PRM ligand to occupy the binding pocket of the SH3 domain. The state where the SWCNT occupies the binding pocket of the SH3 domain is found to be the global minimum of the binding free energy landscape, with a ∼0.6 kcal/mol lower binding affinity than the equivalent state of the RPM ligand binding. The stronger binding affinity between the SWCNT and the SH3 domain is shown to be mainly from the π
π stacking interactions between the CNT and aromatic residues in the binding pocket. Essentially, the SWCNT prevents the RPM ligand from binding to the active site of the SH3 domain, thus disrupting the biological function of the SH3 domain.

’ COMPUTATIONAL METHODS The simulation system includes a SH3 domain, a proline-rich motif (PPPVPPRR) ligand, and an armchair SWCNT. The SH3 domain has a characteristic β-barrel fold which consists of five β-strands arranged as two tightly packed antiparallel β-sheets.25 The PRM ligand PPPVPPRR is a specific ligand which binds on active site of the SH3 domain. The binding structure of the SH3 domain and this ligand was obtained from Protein Data Bank (PDB code: 1CKB,25 shown in Figure 1b), and the binding behavior of the SH3 domain and this ligand had studied by the MD simulation.26 It has been suggested experimentally that residues F8, W36, P50, and Y53 are the key residues in binding with the ligand,17,27 and the hydrophobic interaction is the dominant force for binding as the simulation of the binding of the SH3 domain and this ligand.26 The pristine SWCNT is in the armchair form of (m, n), where m = n = 3, corresponding to a tube diameter of 4.04 Å, with a length of 19.54 Å. A recent experimental study has shown that long SWCNTs can be biodegraded into very small fragments by the enzyme horseradish peroxidase (HRP),28 and another experiment showed that the shorter SWCNT may induce more toxicity.29 Thus, short SWCNTs might exist, and studying SWCNTs with relatively short lengths might be of significant importance to human health as well. The SH3 domain and the ligand were prepared from the Protein Data Bank (PDB code: 1CKB,25 residues 134
190 for the SH3 domain, and residue 1
8 for the ligand) and modeled by AMBER03 force field.30 The carbon atoms of the SWCNT were modeled as uncharged Lennard-Jones particles with a cross section of σcc = 3.40 Å and a depth of the potential well of εcc = 0.36 kJ/mol.31,32 Two three-way (SWCNT þ ligand þ protein) binding complex systems were set up. In the first system, the initial separations of the geometric center of the SWCNT and the ligand from that of the SH3 domain were both set at 30 Å, a distance long enough to avoid the starting point bias during the binding process (more below). The initial orientations of the SWCNT and the RPM ligand versus the SH3 domain were set at different directions (shown in Supporting Information Figure S1). The resulting complex was then solvated in a rhombic dodecahedral periodic box, with the distance between the solutes and the boundary of the box at least 10 Å. The final complex system size is 12 378 atoms. For the second complex system, a similar procedure was followed, but with the SWCNT and ligand swapped in space (thus in the opposite orientations versus the SH3 domain as in the first system) for further enhancement of

ARTICLE

Figure 1. Structures of the three-way binding complex: SH3 domain, ligand and SWCNT. (a) A typical binding snapshot in which the SWCNT occupies the binding pocket and contacts the key residues of the SH3 domain, while the ligand binds partly with the SWCNT and partly with the SH3 domain, and (b) the SH3 domain binding with the PRM ligand (PDB: 1CKB) in its native state. Here the SH3 domain is shown in cartoon with yellow β-strands and green loops, and the key residues (F8, W36, P50, and Y53) in the pocket are identified by red sticks. The atoms of the SWCNT are shown in wheat-colored spheres. The ligand with the sequence PPPVPPRR is shown in cyan cartoon with the key residues identified by blue sticks. The solvated surfaces are shown for SWCNT, the ligand and the SH3 domain in (a).

sampling. The final system size is 13 367 atoms (see Figure S1). Both complex systems are sampled at least 10 times with slightly different initial conformations to simulate the binding process (see below). The TIP3P33 water model was used for the salvation, and three Naþ were added into solution to neutralize each system. The solvated systems were simulated with molecular dynamics, which is widely used in the studies of biomolecules34
52 and nanoscale systems.31,32,53
55 We used the Gromacs package 4.0.56 In the simulations, the covalent bonds involving H atoms were constrained by the LINCS algorithm, which allowed a time step of 2 fs. The long-range electrostatic interactions were treated with the particle-mesh Ewald method (PME) with a grid spacing of 1.2 Å, while the van der Waals interactions were handled with the usual smooth cutoff, with a cutoff distance set at 10 Å. After energy minimization, both the SWCNT and RPM ligand complex systems were equilibrated by MD simulations for 200 ps at a constant pressure of 1 bar and temperature of 298 K using Berendsen coupling. The production runs were performed with the NVT ensemble at 298 K. Ten trajectories, with 200þ ns each, were obtained for both the three-way (SWCNT þ ligand þ SH3 domain) binding complexes.

’ RESULTS AND DISCUSSION We found that, in 13 out of the 20 total simulation trajectories, the SWCNT occupies the binding pocket and directly contacts with those key residues at the active site of SH3 domain, which prevents the ligand from binding to the pocket, as one typical example shows in Figure 1a. In one extreme case, the SWCNT directly contacts with nine residues near the binding pocket, including three key residues W36, P50, and Y53. In the other 12 cases, the SWCNT contacts at least one of the key residues of the binding pocket. On the other hand, there is only one case out of total 20 where the PRM ligand occupies the SH3 domain binding pocket, while the SWCNT does not bind to the pocket. But even in this single case, no native binding mode between the ligand and the SH3 domain was observed during our 200 ns simulation length. These results indicate that the SWCNT has a higher 12323

dx.doi.org/10.1021/jp2026303 |J. Phys. Chem. C 2011, 115, 12322–12328

The Journal of Physical Chemistry C

Figure 2. One representative trajectory in which the SWCNT adsorbs onto the binding pocket and makes direct contacts with the key residues of the SH3 domain, while the ligand binds partly with the SWCNT and partly with the SH3 domain. (a) Some representative snapshots. Here the proteins are shown in cartoons with yellow β-strands and green loops. The key residues of the protein domain are noted by red sticks, and the ligands with the sequence PPPVPPRR are shown in cyan cartoon with the key residues identified by blue sticks. The SWCNTs are shown in wheat-colored sticks. (b) The protein
SWCNT, protein
ligand, and ligand
SWCNT interface areas as a function of time.

probability of occupying the binding pocket of SH3 domain than the PRM ligand. Interestingly, in 11 of those 13 cases, the ligand also binds partly with the SWCNT and partly with the surface of the SH3 domain (near the binding pocket), and in the remaining two cases, the ligand only binds partly with the surface of the SH3 domain (far from the binding pocket) but does not bind with the SWCNT, indicating that the ligand still maintains some binding capability, though lower than that of the SWCNT (more below). Even though the other 7 out of the 20 total trajectories do not show a direct binding of the SWCNT with the SH3 domain binding pocket, the SWCNT does display some tendency of the binding. For example, in 3 out of the 7 trajectories, the SWCNT was getting close to the W37 residue in the binding pocket, and in the other two trajectories, it was getting close to the Y03 residue. The incomplete binding might be related to the fact that even though we have tried fairly long simulation lengths of 200þ ns, they are still not long enough. Figure 2a shows some representative snapshots of one typical trajectory in which the SWCNT first binds to the SH3 domain and then the ligand binds to the SWCNT. The interface area between the SH3 domain and the SWCNT, denoted by SPC, is also shown to illustrate this process (see Figure 2b). Here the interface area is defined as half of the difference between the solvent accessible surface area of the complex of the SH3 domain and SWCNT and the sum of the solvent accessible surface areas of the two objects separately.57 It shows that there are three stages for the binding between the SH3 domain and the SWCNT in this case. Initially, at t = 0, the center of the SH3 domain and the SWCNT are 30 Å apart and SPC = 0 (see the snapshot at t = 0 ns). The SWCNT and SH3 domain remains separated for about 3 ns (initial stage). Then, SPC rises quickly to about 150 Å2 around t = 3 ns, indicating that the SH3 domain and the SWCNT approach each other (see the snapshot at t = 3 ns, where the SWCNT is adsorbed onto the SH3 domain near the binding pocket). After that, the SWCNT keeps adjusting its position and orientation (second stage). At ∼35 ns, there is a significant jump

ARTICLE

Figure 3. Numbers of contacts between the SWCNT (top) and the ligand (bottom), with the residues of SH3 domain. The number is colorcoded, with the “blank” white meaning no contact. There is a contact if the distance between a non-hydrogen atom in the bound object and a non-hydrogen atom in the residue is less than 5 Å.

in SPC from ∼150 to ∼250 Å2 (see the snapshot at t = 35 ns, where the SWCNT is now bound right onto the binding pocket of the SH3 domain). From t = ∼35 ns up to 200 ns, there were only minor fluctuations of SPC at ∼250 Å (third stage), indicating a fairly high stability of the adsorption of SWCNT on the SH3 domain. In Figure 2b, we also display the interface areas of the ligand with the SH3 domain (Protein
Ligand) and the ligand with the SWCNT (Ligand
SWCNT), denoted by SPL and SLC, respectively. At t = 0, SPL = SLC = 0. Within the first 5 ns, SPL increases to about 200 Å2 very quickly. The corresponding snapshot shows that the ligand is adsorbed onto the SH3 domain (but not at the binding pocket). However, SPL then decreases remarkably from 200 Å2 to a very small value in a very short period of time (∼87
90 ns), and it finally recovers back to ∼150 Å2 after t = 100 ns. Meanwhile, SLC increases to about 100 Å2 at t = ∼90 ns, and it keeps rising to about 200 Å2 until it saturates around t = ∼120 ns. A detailed examination reveals that the ligand is temporarily separated from the SH3 domain around t = 88 ns. It then diffuses to the other side of the SH3 domain and binds to both the SWCNT and the SH3 domain (see the snapshot at 100 ns). After that, both SPL and SLC remain at these values with normal fluctuations, indicating the stability of the final threeway bound structure among the SWCNT, the ligand, and the SH3 domain. To further characterize the adsorption between the SWCNT, the ligand, and the SH3 domain, we have computed the contact numbers of the SWCNT and the ligand with SH3 domain. Here we used the protein residue as a unit base in accounting (see Figure 3), and a contact is counted if the distance between a nonhydrogen atom in the bound object and a non-hydrogen atom in the protein residue is less than 5 Å. Generally speaking, the closer the object is to the protein residue, the larger the contact number is between the protein and the object, and the stronger the object binds to the protein residue. Similarly, from the contacts of the SWCNT on the SH3 domain, we found that there were three stages in the binding process. In the initial stage, from 0 to ∼3 ns, the SH3 domain and the SWCNT were separated; in the second stage, from ∼3 to ∼35 ns, the SWCNT was adsorbed onto four residues of the SH3 domain; in the third stage, from ∼35 ns to the end of the simulation, the SWCNT bound closely with many residues (up to 9) of the SH3 domain, including the key residues W36, P50, and Y53. Interestingly, the contacts between the SWCNT and the ligand are found to be quite stable; i.e., these contacts remain 12324

dx.doi.org/10.1021/jp2026303 |J. Phys. Chem. C 2011, 115, 12322–12328

The Journal of Physical Chemistry C

Figure 4. Binding free energy landscapes of the SWCNT (a) and the ligand (b) with the SH3 domain. SPC and SPL denote the interface area of the SWCNT and ligand with the SH3 domain, respectively. DKC and DKL are the minimal distance of the SWCNT and the ligand from these key residues, respectively. The unit of the free energy is kcal/mol.

in contact once they appear (see Figure S2a). On the contrary, the contacts between the ligand and the protein (both ligand on protein (Figure 3b) and protein on ligand (Figure S2b) vary with time, indicating that the relative positions between the ligand and the SH3 domain change with time much more often. Further, the great changes of the ligand
protein contacts happen at t = ∼88 ns (Figure 3b and Figure S2b), corresponding to the previous observation that the ligand temporarily separates from the SH3 domain, then diffuses to the other side of the SH3 domain, and finally readsorbs onto both the SWCNT and the SH3 domain. Conformational fluctuations of the ligand on the SWCNT and SH3 domain are observed after the final adsorption. For example, from 90 to 150 ns, the ligand binds both the SH3 domain (on the residues 12
13, 16, and 36) and the SWCNT by the C-terminal residues, while from 150 to 200 ns, the ligand binds the SH3 domain (on the residues 9
11 and 53) by the N-terminal and the SWCNT by C-terminal (see also Figure S2). The free energy landscape can be determined by first calculating the normalized probability distribution function P(X) from a histogram analysis of conformations sampled from the MD simulation. Since the potential of mean force (PMF) W(X), or equivalently the free energy, is related to this probability distribution function through the relation PðXÞ ¼ 1=Z expð
βWðXÞÞ where X is the specified choice set of reaction coordinates (RCs) and Z is the equi-partition function, the relative free energy change corresponding to a change in RC can be obtained easily from WðX2Þ
WðX1Þ ¼
RT log½PðX2Þ=PðX1Þ which implicitly includes the conformational entropy contributions. Figure 4a shows the free energy landscape as a function of two reaction coordinates: the interface area between the SH3 domain and the SWCNT (SPC) and the minimal distance between the SWCNT and the key residues of the SH3 domain

ARTICLE

Figure 5. Interactions between the SWCNT and some of the aromatic residues of SH3 domain. The top panel shows the key residues of the binding pocket, including F08, W36, and Y53. The bottom panel shows the aromatic residues which are exposed but do not belong to the binding pocket, including Y03 and W37. Here the SH3 domain is shown in gray cartoon, with the contacting aromatic residues shown in red sticks and labeled in green text. The SWCNT is shown in wheat-colored sticks, and the ligand in dark-green cartoon.

(including F08, W36, P50, and Y53), denoted by DKC. Clearly, if the SWCNT binds to one of the key residues, it will prevent the regular binding of the ligand to the SH3 domain. In the computation, we sampled the structures with the time step of 20 ps in the last 180 ns of all the 20 trajectories. The global minimum in this landscape is found at SPC ∼ 220 Å2 and DKC < 5.0 Å with a binding free energy of
6.08 kcal/mol, which corresponds to the state where the SWCNT binds to the SH3 domain binding pocket. Two local minima, in which SPC ∼ 150 Å2, DKC ∼ 8.0 Å and SPC ∼ 180 Å2, DKC ∼ 18.0 Å, are also observed. Detail studies show that these two states correspond to the SWCNT binding to the side (the region around the residue W37) and the opposite (the region around the residue Y03) of the binding pocket of the SH3 domain, respectively (data not shown). Our statistical analyses based on the χ2 hypothesis test also show that the SWCNT is remarkably favored in binding with the hydrophobic residues of the SH3 domain (see Figure S3 in the Supporting Information). It was found that the p values for the SWCNT contacting with hydrophobic residues, aromatic residues, and the hydrophobic aromatic residues were 6.51  10
3, 2.42  10
12, and 2.00  10
19, respectively. These results indicate that the interactions between the SWCNT and the hydrophobic residues, particularly the aromatic residues (π
π stacking interactions), play an important role in the binding of SWCNT and proteins. Figure 5 shows some of the representative local snapshots for the π
π stacking interactions between the SWCNT and key residues of the binding pocket. We note that favorable interactions of the carbon nanotube with the hydrophobic, particularly the aromatic residues, have also been observed in recent carbon nanotube
peptide experiments.58
60 That is, these interactions are independent of the size of proteins and CNTs. Therefore, even though the SWCNT used in our simulation was quite small, the observation of the high probability of occupying the binding pocket of the protein by the hydrophobic nanoscale particles is extendable to the hydrophobic nanoscale particles of larger sizes, which is expected to be 12325

dx.doi.org/10.1021/jp2026303 |J. Phys. Chem. C 2011, 115, 12322–12328

The Journal of Physical Chemistry C observed using much more extensive computations, with the advancement of both computational resources and efficient sampling algorithms. It should be also noted that the π
π stacking interactions might be underestimated in standard force fields due to the lack of polarization, and more advanced techniques such as quantum mechanics simulations might be needed to fully catch the effect; however, a recent study61 shows that classical simulations can sometimes even better capture these π
π stacking interactions than the quantum mechanics simulations because the classical force fields are often directly fit from experimental data, and the quantum mechanics simulations on the other hand might suffer from the limited size and boundary conditions. The binding of the ligand to the SH3 domain is also described by the free energy landscape as a function of the two equivalent reaction coordinates: the interface area between the ligand and the SH3 domain (SPL) and the distance between the ligand and the key residues of the SH3 domain (DKC) (shown in Figure 4b). The lowest binding affinity for the ligand is
5.51 kcal/mol, which is 0.57 kcal/mol less than the SWCNT shown above. This explains why the PRM ligand loses the competition to the SWCNT in the binding to the SH3 domain. Interestingly, the binding free energy landscape also shows that there are two basins, with the free energy of
5.51 kcal/mol (the lowest for the ligand binding) and
5.16 kcal/mol. Both states have similar values of SPL ∼ 180 Å2, i.e., adsorbing onto the SH3 domain with similar contacting surface areas. The difference lies in the DKL, i.e., the distance to the key residues of the SH3 domain binding pocket. The deeper one has a very small value of DKL ∼ 5.0 Å, which indicates that the ligand is around the binding pocket. The other basin has a larger value of DKL ∼ 12.0 Å, indicating that the ligand is far away the binding pocket. We note that the same DKL (∼12 Å) does not mean that there is only one binding site for the ligand in this case. A detailed study shows that the ligand binds to a broad region of the SH3 domain with residues 11
19, 23
25, and 43
47 involved. It is worth noting that one previous MD simulation showed that the ligand could recognize the binding pocket and form the native binding mode with the SH3 domain, in the absence of nanoscale particles.26 It was found that the long-range electrostatic interactions played an important role in the ligand recognition of the binding pocket, through its two arginine residues on the C-terminal.26 The PRM ligand’s R7 and R8 residues have favorable long-range electrostatic interactions with SH3-domain’s acidic residues, D14, E16, D17, and E33 on RT and n-sCr loops, near one end of the binding pocket, which provides guidance to an accelerated ligand binding mode search. This electrostatically accelerated association of proteins was also previously observed in experiments.62,63 The stability of the final binding, on the other hand, is driven by the structural match and hydrophobic interactions between the ligand (prolines and V4) and the SH3 domain (F8, F10, W36, P50, P52, and Y53), in addition to the salt bridges formed by R7/R8. In our current complex system with the presence of hydrophobic nanoparticles, the SWCNT wins the competition over the ligand for the binding pocket even without any guidance from the long-range electrostatic interactions (since SWCNT is uncharged in our simulations), indicating a very strong hydrophobic interaction between the SWCNT and the SH3 domain. Interestingly, the ligand still manages to locate the area of the active site and bind with both the SWCNT and SH3 domain (though no longer in the binding pocket), which implies the PRM ligand still maintains some

ARTICLE

recognition capability due to the long-range electrostatic interactions through its two C-terminal arginine residues. On the other hand, for the optimal binding of the ligand (the native binding mode), there is a structural match between the hydrophobic clusters of the ligand (prolines and V4) and the active site of the SH3 domain (F8, F10, W36, P50, P52, and Y53), as shown in Figure 1b for the native structure. This raises the demand for an extensive conformational space mapping in order to reach this native binding mode even with the electrostatic acceleration. Moreover, the hydrophilic residues in the ligand (the two terminal arginines) often shield the hydrophobic ones in the water environment to “inhibit” their exposure to the active site. While, on the other hand, the hydrophobic interactions between the SWCNT and the hydrophobic residues of the binding pocket, particularly those aromatic ones, are strong and nonspecific. This makes the SWCNT more straightforward to be adsorbed onto the binding pocket and explains why it takes about 35 ns for the SWCNT to reach a stable binding state at the active site, while it takes ∼130 ns for the ligand to just come near the SH3 domain in our MD simulations.26 This is also consistent with the above free energy landscape analysis, which reveals that the SWCNT is more favorable than the PRM ligand to be bound to the active site of the SH3 domain. Thus, in the competition for the binding pocket of the SH3 domain, the SWCNT has advantages over the PRM ligand in both kinetics and thermodynamics. The SWCNT essentially occupies the binding pocket of the SH3 domain and interrupts its native binding with the PRM ligand.

’ CONCLUSION We have studied the binding process of a proline-rich motif (PRM) ligand with its target protein, SH3 domain, in the presence of a hydrophobic SWCNT, using molecular dynamics simulations. We found that the SWCNT has a very high probability to occupy the binding pocket of the SH3 domain, with about 0.6 kcal/mol more favorable binding affinity than the original PRM ligand. The presence of the hydrophobic nanoparticle essentially blocks the PRM ligand from its native binding mode. In most of the simulation trajectories, the ligand is found to be adsorbed onto the SWCNT near the active site of the SH3 domain. We also noticed an interesting adsorption
separation
diffusion
readsorption phenomenon for the ligand on SH3 domain, with which the ligand first adsorbs onto a non-native binding site, then separates from the site and diffuses toward the native binding pocket, and finally binds close to the native binding pocket (occupied by the SWCNT). This implies that the ligand still has the capability to recognize the binding pocket of the SH3 domain, even though it loses the competition to the SWCNT in binding to the SH3 domain. This can be explained by the relatively close binding affinities with the SH3 domain:
6.08 kcal/mol for the SWCNT versus
5.51 kcal/mol for the PRM ligand. However, the ligand cannot achieve the native binding mode with the SH3 domain in any of the 20 cases we studied because the binding pocket was held by the SWCNT with priority. We found that the interactions between the SWCNT and the hydrophobic residues, particularly the aromatic residues (π
π stacking interactions), play an important role in the binding of SWCNT with proteins. These observations indicate that the presence of the SWCNT will occupy the binding pocket of SH3 domain, which prevents the regular binding of the PRM ligand, thus disrupting the biological function of the SH3 domain. 12326

dx.doi.org/10.1021/jp2026303 |J. Phys. Chem. C 2011, 115, 12322–12328

The Journal of Physical Chemistry C

’ ASSOCIATED CONTENT

bS

Supporting Information. List of all of our simulation systems, the binding site of protein and SWCNT on ligand, and distribution of the number of residues in contact with the SWCNT. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Tel 86-21-59554785, Fax 86-21-59552394, e-mail fanghaiping@ sinap.ac.cn (H.F.); Tel (914) 945-3591, Fax (845) 489-9512, e-mail [email protected] (R.Z.).

’ ACKNOWLEDGMENT We thank Peng Xiu, Chunlei Wang, Wenpeng Qi, Payel Das, and Bruce Berne for helpful discussions. This research is supported in part by grants from NNSFC (10825520), NBRPC (973 Program) (2007CB936000 and 2007CB814800), Shanghai Leading Academic Discipline Project (B111), and Shanghai Supercomputer Center of China. R.Z. acknowledges the support from the IBM BlueGene Science Program. ’ REFERENCES (1) Rosi, N. L.; Giljohann, D. A.; Thaxton, C. S.; Lytton-Jean, A. K. R.; Han, M. S.; Mirkin, C. A. Science 2006, 312, 1027. (2) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005, 307, 538. (3) Wang, X.; Yang, L. L.; Chen, Z.; Shin, D. M. Ca-Cancer J. Clin. 2008, 58, 97. (4) Li, H. K.; Huang, J. H.; Lv, J. H.; An, H. J.; Zhang, X. D.; Zhang, Z. Z.; Fan, C. H.; Hu, J. Angew. Chem., Int. Ed. 2005, 44, 5100. (5) Hu, W.; Peng, C.; Luo, W.; Lv, M.; Li, X.; Li, D.; Huang, Q.; Fan, C. ACS Nano 2010, 4, 4317. (6) Service, R. F. Science 2000, 290, 1526. (7) Donaldson, K.; Aitken, R.; Tran, L.; Stone, V.; Duffin, R.; Forrest, G.; Alexander, A. Toxicol. Sci. 2006, 92, 5. (8) Gilbert, N. Nature 2009, 460, 937. (9) Nel, A.; Xia, T.; Madler, L.; Li, N. Science 2006, 311, 622. (10) Zhao, Y.; Xing, G.; Chai, Z. Nature Nanotechnol. 2008, 3, 191. (11) Chen, Z.; Meng, H.; Xing, G. M.; Chen, C. Y.; Zhao, Y. L.; Jia, G.; Wang, T. C.; Yuan, H.; Ye, C.; Zhao, F.; Chai, Z. F.; Zhu, C. F.; Fang, X. H.; Ma, B. C.; Wan, L. J. Toxicol. Lett. 2006, 163, 109. (12) Goldberg-Oppenheimer, P.; Regev, O. Small 2007, 3, 1894. (13) Shen, J.-W.; Wu, T.; Wang, Q.; Kang, Y. Biomaterials 2008, 29, 3847. (14) Park, K. H.; Chhowalla, M.; Iqbal, Z.; Sesti, F. J. Biol. Chem. 2003, 278, 50212. (15) Karajanagi, S. S.; Vertegel, A. A.; Kane, R. S.; Dordick, J. S. Langmuir 2004, 20, 11594. (16) Zuo, G. H.; Huang, Q.; Wei, G. H.; Zhou, R. H.; Fang, H. P. ACS Nano 2010, 4, 7508. (17) Ball, L. J.; Kh€une, R.; Schneider-Mergener, J.; Oschkinat, H. Angew. Chem., Int. Ed. 2005, 44, 2852. (18) Macias, M. J.; Wiesner, S.; Sudol, M. FEBS Lett. 2002, 513, 30. (19) Bucciantini, M.; Giannoni, E.; Chiti, F.; Baroni, F.; Formigli, L.; Zurdo, J. S.; Taddei, N.; Ramponi, G.; Dobson, C. M.; Stefani, M. Nature 2002, 416, 507. (20) Gorina, S.; Pavletich, N. P. Science 1996, 274, 1001. (21) Zarrinpar, A.; Bhattacharyya, R. P.; Lim, W. A. Sci. STKE 2003, 2003, RE8.

ARTICLE

(22) Li, X. F.; Chen, Y. S.; Liu, Y. W.; Gao, J.; Gao, F.; Bartlam, M.; Wu, J. Y.; Rao, Z. H. J. Biol. Chem. 2006, 281, 28430. (23) Kobashigawa, Y.; Sakai, M.; Naito, M.; Yokochi, M.; Kumeta, H.; Makino, Y.; Ogura, K.; Tanaka, S.; Inagaki, F. Nat. Struct. Mol. Biol. 2007, 14, 503. (24) Venter, J. C.; Adams, M. D.; Myers, E. W.; Li, P. W.; Mural, R. J.; Sutton, G. G.; Smith, H. O.; Yandell, M.; Evans, C. A.; Holt, R. A.; Gocayne, J. D.; Amanatides, P.; Ballew, R. M.; Huson, D. H.; Wortman, J. R.; Zhang, Q.; Kodira, C. D.; Zheng, X. Q. H.; Chen, L.; Skupski, M.; Subramanian, G.; Thomas, P. D.; Zhang, J. H.; Miklos, G. L. G.; Nelson, C.; Broder, S.; Clark, A. G.; Nadeau, C.; McKusick, V. A.; Zinder, N.; Levine, A. J.; Roberts, R. J.; Simon, M.; Slayman, C.; Hunkapiller, M.; Bolanos, R.; Delcher, A.; Dew, I.; Fasulo, D.; Flanigan, M.; Florea, L.; Halpern, A.; Hannenhalli, S.; Kravitz, S.; Levy, S.; Mobarry, C.; Reinert, K.; Remington, K.; Abu-Threideh, J.; Beasley, E.; Biddick, K.; Bonazzi, V.; Brandon, R.; Cargill, M.; Chandramouliswaran, I.; Charlab, R.; Chaturvedi, K.; Deng, Z. M.; Di Francesco, V.; Dunn, P.; Eilbeck, K.; Evangelista, C.; Gabrielian, A. E.; Gan, W.; Ge, W. M.; Gong, F. C.; Gu, Z. P.; Guan, P.; Heiman, T. J.; Higgins, M. E.; Ji, R. R.; Ke, Z. X.; Ketchum, K. A.; Lai, Z. W.; Lei, Y. D.; Li, Z. Y.; Li, J. Y.; Liang, Y.; Lin, X. Y.; Lu, F.; Merkulov, G. V.; Milshina, N.; Moore, H. M.; Naik, A. K.; Narayan, V. A.; Neelam, B.; Nusskern, D.; Rusch, D. B.; Salzberg, S.; Shao, W.; Shue, B. X.; Sun, J. T.; Wang, Z. Y.; Wang, A. H.; Wang, X.; Wang, J.; Wei, M. H.; Wides, R.; Xiao, C. L.; Yan, C. H.; Yao, A.; Ye, J.; Zhan, M.; Zhang, W. Q.; Zhang, H. Y.; Zhao, Q.; Zheng, L. S.; Zhong, F.; Zhong, W. Y.; Zhu, S. P. C.; Zhao, S. Y.; Gilbert, D.; Baumhueter, S.; Spier, G.; Carter, C.; Cravchik, A.; Woodage, T.; Ali, F.; An, H. J.; Awe, A.; Baldwin, D.; Baden, H.; Barnstead, M.; Barrow, I.; Beeson, K.; Busam, D.; Carver, A.; Center, A.; Cheng, M. L.; Curry, L.; Danaher, S.; Davenport, L.; Desilets, R.; Dietz, S.; Dodson, K.; Doup, L.; Ferriera, S.; Garg, N.; Gluecksmann, A.; Hart, B.; Haynes, J.; Haynes, C.; Heiner, C.; Hladun, S.; Hostin, D.; Houck, J.; Howland, T.; Ibegwam, C.; Johnson, J.; Kalush, F.; Kline, L.; Koduru, S.; Love, A.; Mann, F.; May, D.; McCawley, S.; McIntosh, T.; McMullen, I.; Moy, M.; Moy, L.; Murphy, B.; Nelson, K.; Pfannkoch, C.; Pratts, E.; Puri, V.; Qureshi, H.; Reardon, M.; Rodriguez, R.; Rogers, Y. H.; Romblad, D.; Ruhfel, B.; Scott, R.; Sitter, C.; Smallwood, M.; Stewart, E.; Strong, R.; Suh, E.; Thomas, R.; Tint, N. N.; Tse, S.; Vech, C.; Wang, G.; Wetter, J.; Williams, S.; Williams, M.; Windsor, S.; Winn-Deen, E.; Wolfe, K.; Zaveri, J.; Zaveri, K.; Abril, J. F.; Guigo, R.; Campbell, M. J.; Sjolander, K. V.; Karlak, B.; Kejariwal, A.; Mi, H. Y.; Lazareva, B.; Hatton, T.; Narechania, A.; Diemer, K.; Muruganujan, A.; Guo, N.; Sato, S.; Bafna, V.; Istrail, S.; Lippert, R.; Schwartz, R.; Walenz, B.; Yooseph, S.; Allen, D.; Basu, A.; Baxendale, J.; Blick, L.; Caminha, M.; Carnes-Stine, J.; Caulk, P.; Chiang, Y. H.; Coyne, M.; Dahlke, C.; Mays, A. D.; Dombroski, M.; Donnelly, M.; Ely, D.; Esparham, S.; Fosler, C.; Gire, H.; Glanowski, S.; Glasser, K.; Glodek, A.; Gorokhov, M.; Graham, K.; Gropman, B.; Harris, M.; Heil, J.; Henderson, S.; Hoover, J.; Jennings, D.; Jordan, C.; Jordan, J.; Kasha, J.; Kagan, L.; Kraft, C.; Levitsky, A.; Lewis, M.; Liu, X. J.; Lopez, J.; Ma, D.; Majoros, W.; McDaniel, J.; Murphy, S.; Newman, M.; Nguyen, T.; Nguyen, N.; Nodell, M.; Pan, S.; Peck, J.; Peterson, M.; Rowe, W.; Sanders, R.; Scott, J.; Simpson, M.; Smith, T.; Sprague, A.; Stockwell, T.; Turner, R.; Venter, E.; Wang, M.; Wen, M. Y.; Wu, D.; Wu, M.; Xia, A.; Zandieh, A.; Zhu, X. H. Science 2001, 291, 1304. (25) Wu, X. D.; Knudsen, B.; Feller, S. M.; Zheng, J.; Sali, A.; Cowburn, D.; Hanafusa, H.; Kuriyan, J. Structure 1995, 3, 215. (26) Ahmad, M.; Gu, W.; Helms, V. Angew. Chem., Int. Ed. 2008, 47, 7626. (27) Nguyen, J. T.; Turck, C. W.; Cohen, F. E.; Zuckermann, R. N.; Lim, W. A. Science 1998, 282, 2088. (28) Allen, B. L.; Kichambare, P. D.; Gou, P.; Vlasova, I. I.; Kapralov, A. A.; Konduru, N.; Kagan, V. E.; Star, A. Nano Lett. 2008, 8, 3899. (29) Becker, M. L.; Fagan, J. A.; Gallant, N. D.; Bauer, B. J.; Bajpai, V.; Hobbie, E. K.; Lacerda, S. H.; Migler, K. B.; Jakupciak, J. P. Adv. Mater. 2007, 19, 939. (30) Duan, Y.; Wu, C.; Chowdhury, S.; Lee, M. C.; Xiong, G. M.; Zhang, W.; Yang, R.; Cieplak, P.; Luo, R.; Lee, T. S.; Caldwell, J.; Wang, J. M.; Kollman, P. J. Comput. Chem. 2003, 24, 1999. 12327

dx.doi.org/10.1021/jp2026303 |J. Phys. Chem. C 2011, 115, 12322–12328

The Journal of Physical Chemistry C

ARTICLE

(31) Gong, X. J.; Li, J. Y.; Lu, H. J.; Wan, R. Z.; Li, J. C.; Hu, J.; Fang, H. P. Nature Nanotechnol. 2007, 2, 709. (32) Hummer, G.; Rasaiah, J. C.; Noworyta, J. P. Nature 2001, 414, 188. (33) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. J. Chem. Phys. 1983, 79, 926. (34) Duan, Y.; Kollman, P. A. Science 1998, 282, 740. (35) Levy, Y.; Wolynes, P. G.; Onuchic, J. N. Proc. Natl. Acad. Sci. U. S.A. 2004, 101, 511. (36) Snow, C. D.; Nguyen, N.; Pande, V. S.; Gruebele, M. Nature 2002, 420, 102. (37) Zhou, R. H.; Huang, X. H.; Margulis, C. J.; Berne, B. J. Science 2004, 305, 1605. (38) Liu, P.; Huang, X. H.; Zhou, R. H.; Berne, B. J. Nature 2005, 437, 159. (39) Zuo, G.; Hu, J.; Fang, H. Phys. Rev. E 2009, 79, 031925. (40) Mayor, U.; Guydosh, N. R.; Johnson, C. M.; Grossmann, J. G.; Sato, S.; Jas, G. S.; Freund, S. M. V.; Alonso, D. O. V.; Daggett, V.; Fersht, A. R. Nature 2003, 421, 863. (41) Mirny, L.; Shakhnovich, E. Annu. Rev. Biophys. Biomol. Struct. 2001, 30, 361. (42) Garcia, A. E.; Paschek, D. J. Am. Chem. Soc. 2008, 130, 815. (43) Miyashita, N.; Straub, J. E.; Thirumalai, D. J. Am. Chem. Soc. 2009, 131, 17843. (44) Hua, L.; Huang, X.; Zhou, R.; Berne, B. J. J. Phys. Chem. B 2006, 110, 3704. (45) Liu, P.; Huang, X.; Zhou, R.; Berne, B. J. J. Phys. Chem. B 2006, 110, 19018. (46) Zhou, R. J. Mol. Graphics Modell. 2004, 22, 451. (47) Zhou, R. H.; Eleftheriou, M.; Royyuru, A. K.; Berne, B. J. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5824. (48) Gao, Y. Q.; Yang, W.; Karplus, M. Cell 2005, 123, 195. (49) Karplus, M.; Gao, Y. Q.; Ma, J.; van der Vaart, A.; Yang, W. Philos. Trans. R. Soc. London, A 2005, 363, 331. (50) Roitberg, A. E.; Okur, A.; Simmerling, C. J. Phys. Chem. B 2007, 111, 2415. (51) Li, W. F.; Zhang, J.; Wang, J.; Wang, W. J. Am. Chem. Soc. 2008, 130, 892. (52) Krone, M. G.; Hua, L.; Soto, P.; Zhou, R.; Berne, B. J.; Shea, J. E. J. Am. Chem. Soc. 2008, 130, 11066. (53) Tu, Y. S.; Xiu, P.; Wan, R. Z.; Hu, J.; Zhou, R. H.; Fang, H. P. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 18120. (54) Lindman, S.; Lynch, I.; Thulin, E.; Nilsson, H.; Dawson, K. A.; Linse, S. Nano Lett. 2007, 7, 914. (55) Giovambattista, N.; Lopez, C. F.; Rossky, P. J.; Debenedetti, P. G. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 2274. (56) Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E. J. Chem. Theory Comput. 2008, 4, 435. (57) Nadassy, K.; Wodak, S. J.; Janin, J. Biochemistry 1999, 38, 1999. (58) Wang, S.; Humphreys, E. S.; Chung, S. Y.; Delduco, D. F.; Lustig, S. R.; Wang, H.; Parker, K. N.; Rizzo, N. W.; Subramoney, S.; Chiang, Y. M.; Jagota, A. Nature Mater. 2003, 2, 196. (59) Zheng, L. F.; Jain, D.; Burke, P. J. Phys. Chem. C 2009, 113, 3978. (60) Brown, S.; Jespersen, T. S.; Nygard, J. Small 2008, 4, 416. (61) Tomasio, S. D.; Walsh, T. R. Mol. Phys. 2007, 105, 221. (62) Selzer, T.; Albeck, S.; Schreiber, G. Nat. Struct. Biol. 2000, 7, 537. (63) Schreiber, G.; Fersht, A. R. Nat. Struct. Biol. 1996, 3, 427.

12328

dx.doi.org/10.1021/jp2026303 |J. Phys. Chem. C 2011, 115, 12322–12328