Simulations of Peptide-Graphene Interactions in Explicit Water - The

Aug 21, 2013 - ... Minyu Xiao , Qiuming Wang , Charles L. Brooks III , and Zhan Chen. Journal of the American Chemical Society 2017 139 (5), 1928-1936...
0 downloads 0 Views 967KB Size
Subscriber access provided by University of Chicago Library

Article

Simulations of Peptide-Graphene Interactions in Explicit Water Aerial N Camden, Stephen Austin Barr, and Rajiv J Berry J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/jp403505y • Publication Date (Web): 21 Aug 2013 Downloaded from http://pubs.acs.org on August 24, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Simulations of Peptide-Graphene Interactions in Explicit Water Aerial N. Camden, Stephen A. Barr, Rajiv J. Berry* Air Force Research Laboratory, Materials & Manufacturing Directorate, WPAFB, OH 45433

*Corresponding Author 1-937-255-2467 [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 23

Abstract The interaction of graphene with biomolecules has a variety of useful applications. In particular, graphitic surfaces decorated with peptides are being considered for high performance biochemical sensors. The interaction of peptides with graphene can also provide insight into the binding behavior of larger biomolecules. In this investigation, we have computed the binding enthalpies of a series of GXG tripeptides with graphene using classical molecular dynamics. Explicit water molecules were included to capture the effect of solvent. Of the twenty amino acid residues examined (X in GXG), arginine, glutamine, and asparagine exhibit the strongest interactions with graphene. Analysis of the trajectories shows that the presence of graphene affects the peptide conformation relative to its conformation in solution. We also find that the peptides favor the graphene interface predominantly due to the influence of the solvent, with hydrophilic residues binding more strongly than hydrophobic residues.

These results

demonstrate the need to include explicit solvent atoms when modeling peptide-graphene systems to mimic experimental conditions. Furthermore, the scheme outlined herein may be widely applicable for the determination and validation of surface interaction parameters for a host of molecular fragments using a variety of techniques, ranging from coarse-grained models to quantum mechanical methods.

KEYWORDS: Binding enthalpy, tripeptide, TEAM force field, LAMMPS, hydropathy

2 ACS Paragon Plus Environment

Page 3 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Introduction Graphene is a two-dimensional sheet of sp2 hybridized carbon atoms packed into a honeycomb-shaped lattice. Its large surface area and the high degree of aromaticity give rise to notable electronic, thermal, and mechanical properties.1 The unique properties of graphene have raised interest for its use in various applications, such as targeted drug delivery2,3, radio frequency switches4, supercapacitors5,6, and even the removal of pollutants7-9 and pathogens10 from the environment. Molecular adsorption on graphitic surfaces induces changes in the electronic properties of the surface which can be readily detected. Due to this phenomenon, graphene is considered a prime material for use in ultrasensitive biochemical sensors.11-15

Graphene is particularly

sensitive in detection because all of its atoms are on the surface and exposed to the surroundings.4

Graphene’s electronically low noise13,16-17 and high electrical conductivity4

further enhance its capacity as a sensor. The detection of peptides by graphene has many applications in the fields of nanotechnology and biosensing.18 The specificity of the sensing surface can be increased via functionalization; to this end, several experimental studies have investigated peptide chains to identify sequences that bind to graphitic surfaces.11,19-22 Theoretical studies have also been performed to examine the binding of peptide chains and single amino acids with graphitic surfaces.12,23

Molecular dynamics is a tool that is widely used to study the energetics of

biomolecule-surface interactions.23-36

In the present work, we utilize classical molecular

dynamics (MD) simulations to investigate the interactions of graphene with amino acid residues in a peptide, as opposed to an isolated amino acid which is not part of a chain.

The MD

trajectories to determine the effect of graphene on peptide structure and conformation are also

3 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 23

examined. This method provides insight into the mechanism of binding at a molecular level. The computed surface interactions of the individual residues are useful for the parameterization of coarse-grained models, which enables the extension of our simulations to larger time and length scales. Such models would facilitate rapid computational screening of larger peptide sequences for tailored biochemical sensors.

Theoretical Methods The interaction of amino acid residues with a pure graphene sheet was evaluated by conducting MD simulations for the zwitterionic tripeptides, GXG. This model was chosen to examine the interaction of individual amino acid residues, X, in a typical peptide chain. The glycine termini have been used in other studies as well,37,38 where they have been shown to be a “minimally invasive host,”38 since glycine is the smallest and simplest amino acid. For this reason, the GXG model was deemed appropriate for the calculation of binding enthalpy of amino acid residues in a peptide chain. The binding enthalpy (BE) of peptides to graphene was calculated as the difference in energy of the system when the peptide is near the surface (ENEAR) and when it is infinitely far away (EINF). The binding enthalpies of the GXG peptides with graphene were evaluated using a four box approach, illustrated in Figure 1. This method has been previously developed and validated using peptide-gold and peptide-mica interactions.29 As shown in ref. 29, this approach is both efficient and accurate, provided that the systems are sufficiently large to eliminate size effects and have been equilibrated. Instead of evaluating the energy of two large cells, the energies of four smaller systems were calculated: a water cell (EWAT), graphene and peptide solution (ESURF+SOLN), peptide solution (ESOLN), and a graphene-water cell (ESURF+WAT). For the BE computation of the GXG series, the small cell simulations were conducted to evaluate ESOLN 4 ACS Paragon Plus Environment

Page 5 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

and ESURF+SOLN energies for each residue X, while the simulations devoid of peptide (EWAT and ESURF+WAT) were required only once for the entire series.

This considerably reduces the

computational resources required. Five independent simulation cells were constructed for each system. For each peptide, five binding enthalpies (BEi) were computed from the energies of the five independent “SURF+SOLN” cells and the average energies of the “WAT”, “SOLN” and “SURF+WAT” cells: BEi = (EiSURF+SOLN+ EavgWAT) – (EavgSOLN + EavgSURF+WAT)

WAT

SURF + SOLN

(1)

SOLN

SURF + WAT

BE = ENEAR – EFAR BE = (ESURF+SOLN + EWAT) – (ESOLN + ESURF+WAT)

Figure 1. Schematic of the efficient four box method to calculate binding enthalpies.29

Each simulation cell was constructed five independent times using the Amorphous Cell module within the Materials Studio GUI.39 Care was taken to ensure that the chiral residues were built in their naturally occurring L-stereoisomer. Each cell contained 660 explicit water molecules to establish an aqueous environment for the peptides. Peptides with a net charge were neutralized by including an oppositely charged counterion, either a sodium or a chloride ion. A graphene surface containing 240 carbons was constructed with initial dimensions of 26.680 Å x 27.726 Å. The simulated systems represent free-standing graphene, since water molecules are 5 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 23

present on both sides of the surface. In addition, the GGG tripeptide was investigated in a system with no water below the surface, i.e. nonperiodic in the z direction. This was found to weaken the binding from -14.6±0.9 kcal/mol to -12.4±0.6 kcal/mol. Although different substrates will affect energetics, this study focuses on free-standing graphene. The peptide-solution cells and the water cells were placed atop the surface to create the initial “SURF+SOLN” and the “SURF+WAT” models, respectively. All of the constructed cells were minimized to reduce close atomic contacts and imported into the Direct Force Field GUI,40 where the TEAM force field parameters were assigned (see Figure S1 in Supporting Information for point charges). Next, the models were exported as formatted data files for subsequent MD simulations using the isothermal-isobaric (NPT, T = 298.15 K, P = 1 atm, timestep = 1 fs) ensemble in the LAMMPS41 software package. Each system was equilibrated for a minimum of 4 ns followed by at least 10 ns of production time. Upon completion, the surface and peptide solution trajectories were examined to ensure that the system had equilibrated and that the peptide had reached the graphene-water interface within the first 4 ns. Once the system had equilibrated, the peptide remained at the surface and moved freely parallel to the interface for the duration of the simulation. The effect of system size was examined for the GGG tripeptide by constructing larger cells with 2,500 water molecules on a 40.020 Å x 41.589 Å surface containing 540 carbon atoms. Binding enthalpies computed for the larger systems were within 1 kcal/mol of the small systems, which is within statistical error.

Results/Discussion As previously mentioned, the aim of this work is to determine the binding enthalpy of amino acid residues found in a typical peptide. However, the calculation of average binding

6 ACS Paragon Plus Environment

Page 7 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

enthalpy in the GXG series includes the effect of the terminal glycines. To evaluate this effect, the binding enthalpy contribution of the zwitterionic terminal glycine residues was investigated in a series of small glycine chains, G-(G)n-G, where n=0, 1, 2 and 3. These results are shown in Figure 2, which clearly depicts a linear relationship between binding enthalpy and chain length. A linear fit was then used to determine the binding enthalpy contribution of the terminal glycines, which corresponds to the intercept at n=0 (GG). From this fit, a value of -7.10 kcal/mol was calculated for the terminal GG contribution in the “small” system.

The

contribution of the central residue, X, was then extracted by subtracting the contribution of the terminal glycines from the total binding enthalpy. For longer peptides, a linear relationship between BE and chain length is not expected due to protein folding.

Figure 2. Binding enthalpy as a function of chain length in G-(G)n-G, n=0-3. R2 =0.99 for both linear fits. Some error bars are smaller than the point size.

All residues (X) were attracted to the graphene surface, as indicated by the negative value of the binding enthalpy contributions in Table 1. Table S1 in the Supporting Information lists the binding enthalpies of each of the five independent runs for each peptide. The reported standard errors of the mean (SEM) in Table 1 were computed as follows:

 

 √

(2)

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 23

where s is the sample standard deviation of the binding enthalpies and n = 5 is the number of independent samples for each peptide. Values of s range from 0.47 kcal/mol for GQG to 3.31 kcal/mol for GRG. The binding enthalpy contributions of X are plotted as a function of the Kyte-Doolittle hydropathy index42 in Figure 3, where a decrease in hydropathy is associated with stronger binding.

These results are consistent with numerous studies on peptide surface

interactions on several points.13,14,19,43-48 First, the phenomenon of protein adsorption is known to occur on hydrophobic surfaces.43 Second, solvent effects have been shown to play a major role in protein adsorption due to the water structure at the biomaterial interface.43,44 Further, many experimental studies have determined that hydrophilic molecules bind to graphitic surfaces.13,14,19,45-47

Finally, experimental evidence exists demonstrating that hydrophilic

proteins bind to graphitic surfaces more readily than hydrophobic proteins.48

8 ACS Paragon Plus Environment

Page 9 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table 1. Computed binding enthalpy of GXG and the contributions due to central residue X.

Nonpolar

Polar, Uncharged

Aromatic

Polar, acidic Polar, basic

GXG

BE [kcal/mol]

BE of Central X [kcal/mol]

Std. Err. [kcal/mol]

Kyte-Doolittle Hydropathy Index42

GIG

-10.0

-2.9

1.4

4.5

GVG

-8.1

-1.0

0.7

4.2

GLG

-10.2

-3.1

0.8

3.8

GCG

-11.1

-4.0

0.3

2.5

GMG

-12.5

-5.4

0.4

1.9

GAG

-11.7

-4.6

0.8

1.8

GGG

-14.7

-7.6

0.9

-0.4

GTG

-10.8

-3.7

0.3

-0.7

GSG

-11.7

-4.6

1.1

-0.8

GPG

-11.6

-4.5

0.4

-1.6

GNG

-15.3

-8.2

0.2

-3.5

GQG

-15.9

-8.8

0.2

-3.5

GFG

-10.3

-3.2

0.7

2.8

GWG

-14.2

-7.1

0.5

-0.9

GYG

-13.4

-6.3

0.7

-1.3

GHG

-13.4

-6.3

0.7

-3.2

GDG

-12.7

-5.6

1.3

-3.5

GEG

-11.8

-4.7

0.8

-3.5

GKG

-14.9

-7.8

0.8

-3.9

GRG

-17.5

-10.4

1.5

-4.5

Figure 3. Binding enthalpy contribution of each residue as a function of the hydropathy index.

9 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 23

To better understand the nature of binding, the concentration profiles were examined to illustrate how molecules arrange themselves with respect to the graphene surface. We examined the time averaged concentration profiles of both the water-oxygen atoms and the peptide atoms normal to the graphene surface. Figure 4 shows the concentration profiles for all atoms of four peptides, each with four different types of central residues (amide, aromatic, acidic, and basic). The first solvation shell of water was approximately 2.8 Å away from the graphene surface; a second and a third shell were also apparent. A common trend observed was that the peptide backbone and the side group of the central residue tended to concentrate within the first shell of water. Alternatively, the NH3+ and COO- termini groups remained in the space between the water solvation shells.

Figure 4. Concentration profiles of water oxygen atoms (red) and peptide atoms (blue) on a graphene surface. Each X side chain concentrates within the primary shell of water oxygen atoms, while the N and C termini tend to reside in the region between the first two shells.

10 ACS Paragon Plus Environment

Page 11 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

The concentration profiles also aid in understanding the relationship between hydropathy index and BE. The layered structure of water provides an opportunity for the peptide atoms to arrange themselves in an energetically favorable location at the interface. For instance, the central residue, X, primarily resides in the first shell of water, which has a higher density than bulk water. Thus, it is more favorable for hydrophilic residues to be near the surface compared to hydrophobic residues. For E, D, N, and Q, which all have equal hydropathy indices of -3.5, a scatter in BEs is observed. This scatter can be explained by the shelling of water on the graphene. Residues with a carboxyl side chain experience unfavorable interactions with the negatively charged oxygen atoms in the primary shell. Therefore, D and E exhibit weaker BE relative to the neutral residues, N and Q. Similarly, the positively charged residues, R and K, are amongst the stronger binders due to their favorable interactions with the negatively charged water oxygen atoms in the first shell. To help understand the driving forces for peptide binding, the calculated binding enthalpies were broken down into their components: kinetic, electrostatic, van der Waals, and bonded interactions. For all peptides, it was found that the major energy contributor to the binding enthalpy was the change in electrostatic energy. Since the graphitic carbons have a fixed charge of zero in the TEAM force field, the electrostatic contribution cannot be directly due to peptide-graphene interactions. Instead, we find that solvation effects play a major role in the binding enthalpy of a peptide on a graphene surface. To further evaluate this idea, we examined the contribution to the binding enthalpy of three different molecular groups: the solvent, the surface, and the peptide. Table 2 summarizes the results for GAG, GQG, and GYG. For each of these example peptides, the computed binding enthalpy is dominated by solvent contributions. Remarkably, the potential energy changes of the surface and peptide were minimal. From these

11 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 23

results, we conclude that the water forces the peptide towards the surface-water interface, an observation consistent with computational studies.44 Further, the water structure is less disturbed when the peptide is at the interface compared to when the peptide is in solution. This shows that water plays a significant role in peptide-graphene interactions, which is also observed in studies of peptide-carbon nanotube interactions.49 These results also demonstrate the importance of using explicit water models for the simulation of peptide-graphene systems.

Table 2. Potential energy component analysis (in kcal/mol) for select GXG-graphene models. Note the significant energy contribution of the solvent.

Peptide GAG GQG GYG

BE -11.5 -15.2 -15.1

ΔPEsolvent -9.0 -15.3 -12.4

ΔPEsurface -3.5 -0.6 -0.1

ΔPEpeptide +1.4 +1.1 -2.3

ΔKE -0.4 -0.3 -0.3

As shown previously in Table 1, the four strongest binders to graphene did not contain aromatic residues. This is interesting to note, since many other computational studies indicate that the π-π interactions of aromatic rings play a dominant role in interactions with graphitic surfaces. Tryptophan, especially, has been determined to be the strongest binder to graphitic surfaces in several computational investigations.12,20,31 However, explicit water molecules were not included in these studies. In the present work, where the effect of water is explicitly considered, GWG was computed to be the strongest binding aromatic peptide, followed by GYG and GHG. Indeed, we observe that these aromatic residues exhibit a relatively strong interaction with the surface. However, the binding enthalpy of GWG was slightly surpassed by other peptides that did not contain an aromatic side chain (GRG, GQG, GNG, GKG, and GGG). Another noteworthy observation from this study is that the four strongest binders to graphene were GRG, GQG, GNG, and GKG. These peptides contain either a positively charged 12 ACS Paragon Plus Environment

Page 13 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

side chain (GRG and GKG) or an amide group (GQG and GNG). GQG and GNG had relatively high binding enthalpies with the graphene surface (-15.9 kcal/mol and -15.3 kcal/mol, respectively). Upon visualization of the simulation trajectories, it was observed that the side chain amide groups of GQG and GNG tend to align parallel to the graphene surface, as shown in Figure 5. Animations of the trajectories also show that only one of the peptide bonds aligns close to the surface at a given time. Throughout the simulation, the aligned bond was observed to alternate between the two peptide bonds. A tendency for only one of the two peptide bonds to align parallel to the graphene was also observed in the other tripeptides.

Figure 5. Typical conformation for GQG (left) and GNG (right). Water molecules hidden for clarity.

It is also illustrative to inspect the peptide structure both in solution and on the surface. For this, Ramachandran plots from each of the five peptide-surface trajectories were examined in order to visualize the conformation of the peptide backbone. For the GXG-solution (with no graphene present) these plots do not display any significant differences in the conformational behavior of the five independent simulations. However, for the GXG-graphene simulations, there were distinct conformational differences between the five systems. For each system, the presence of graphene induced a change in peptide conformation. As a representative example, the Ramachandran plots of GAG are shown in Figure 6. The distinct conformational differences

13 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 23

are related to the distinct variations in the computed binding enthalpies, listed in Table 3. Similar conformational variations are observed in all other peptide systems, except for GGG, which has little steric hindrance. This indicates that the systems are trapped in local minima, which accounts for the statistically significant differences in binding enthalpy between the runs. These variations are summarized in Table S1 of the Supporting Information for the twenty peptides examined.

Other simulation techniques, such as parallel tempering,50-52 could be

utilized to fully explore the free energy landscape.

Ψ

Ψ

ϕ

Ψ

Ψ

ϕ

Ψ

ϕ

ϕ

Ψ

ϕ

ϕ

Figure 6. Ramachandran plots displaying conformational distribution for the tripeptide GAG. Points represent evenly spaced snapshots of the MD trajectory. The blue and green regions represent low-energy conformations and allowable conformations, respectively.

14 ACS Paragon Plus Environment

Page 15 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table 3. Computed binding enthalpies for the five independent GAG simulations.

GAG Run 1 Run 2 Run 3 Run 4 Run 5

BE [kcal/mol] -11.5 -11.3 -10.9 -10.0 -14.7

Conclusions Classical molecular dynamics simulations have been utilized to examine the binding of the GXG tripeptides with a pure graphene surface. It was found that the layered water near the interface directs the hydrophobic and hydrophilic components of the peptide to energetically favorable locations. In this way, solvent plays a significant role in these systems by influencing peptide binding and conformation. Both an implicit solvent model and a vacuum environment fail to capture the effect of solvent.

Thus, the inclusion of an explicit solvent model is

imperative to a more complete understanding of peptide-graphene systems. Further, it was observed that binding enthalpy can be related to the hydropathy index. This provides insight into the interaction of a typical amino acid residue within a peptide chain with graphitic surfaces. Our calculations clearly demonstrate that the solvent effects have a direct impact on the binding enthalpy. However, the solvent will also affect the entropy associated with binding. Free energy calculations present an opportunity for future research. In conjunction with the present work, these calculations would enable the entropy-enthalpy breakdown and, thus, provide deeper insight into the mechanism of graphene-peptide binding.

Quantum mechanical (QM)

simulations also present an area of future research, as these methods are more accurate, albeit much slower, than classical simulations. QM also inherently includes polarizability, which would help determine if this effect is important in peptide binding, an outstanding question in the 15 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 23

field. Because the explicit water molecules play a substantial role in mediating the interaction of peptides with graphene, they cannot be excluded from the model. Therefore, the system size required makes MD simulations using accurate QM methods computationally impractical with current technology.

Even QM/MM methods may still require a very large QM region to

appropriately model solvent effects. This region would include the graphene, the peptide, and an adequate number of water layers. These challenges present tradeoffs that must be considered in future computational studies of these complex interactions. An efficient scheme allowed us to examine all 20 amino acids and compute accurate peptide-surface interactions with the TEAM force field. This scheme can be utilized for rapid screening of biomolecular-surface interactions by other computational methods, such as additional force fields, quantum mechanical simulations, or QM/MM to obtain and validate parameter sets for use in improved coarse-grained models, facilitating the development of tailored peptide-functionalized graphene for sensitive biochemical sensors.

Acknowledgment We gratefully acknowledge the use of the Department of Defense’s supercomputing resources and the Consolidated Customer Assistance Center. We also sincerely thank Dr. Pedro Derosa, Dr. Gary Kedziora, and Dr. Peter Mirau for useful discussions.

Supporting Information Full table of individual binding enthalpies and force field charge schematics. This material is available free of charge via the Internet at http://pub.acs.org.

16 ACS Paragon Plus Environment

Page 17 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

References (1) Geim, A.; Novoselov, K. The Rise of Graphene. Nat. Mat. 2007, 6, 183-191. (2) Depan, D.; Shah, J.; Misra, R. Controlled Release of Drug from Folate-Decorated and Graphene Mediated Drug Delivery System: Synthesis, Loading Efficiency, and Drug Release Response. Materials Science and Engineering. 2011, 31, 1305-1312. (3) Sun, X.; Liu, Z.; Welsher, K.; Robinson, J.; Goodwin, A.; Zaric, S.; Dai, H. NanoGraphene Oxide for Cellular Imaging and Drug Delivery. Nano Res. 2008, 1, 203-212. (4) Yang, Z.; Gao, R.; Hu, N.; Chai, J.; Cheng, Y.; Zhang, L.; Wei, H.; Kong, E.; Zhang, Y. The Prospective 2D Graphene Nanosheets: Preparation, Functionalization and Applications. Nano-Micro Lett. 2011, 3, 272-280. (5) Lin, Z.; Liu, Y.; Yao, Y.; Hildreth, O.; Li, Z.; Moon, K.; Wong, C. Superior Capacitance of Functionalized Graphene. J. Phys. Chem. C. 2011, 115, 7120-7125. (6) Liu, C.; Yu, Z.; Neff, D.; Zhamu, A.; Jang, B. Graphene-Based Supercapacitor with an Ultrahigh Energy Density. Nano. Lett. 2010, 10, 4863-4868. (7) Ramraj, A.; Hillier, I. Binding of Pollutant Aromatics on Carbon Nanotubes and Graphite. J. Chem. Inf. Model. 2010, 50, 585-588. (8) Li, Y.; Du, Q.; Liu, T.; Sun, J.; Jiao, Y.; Xia, Y.; Xia, L.; Wang, Z.; Zhang, W.; Wang, K; Zhu, H.; Wu, D. Equilibrium Kinetic and Thermodynamic Studies on the Adsorption of Phenol onto Graphene. Mat. Res. Bull. 2012, 47, 1898-1904. (9) Xu, J.; Wang, L.; Zhu, Y. Decontamination of Bisphenol A from Aqueous Solution by Graphene Adsorption. Langmuir. 2012, 28, 8418-8425. (10) Krishnamoorthy, K.; Veerapandian, M.; Zhang, L.; Yun, K.; Kim, S. Antibacterial Efficiency of Graphene Nanosheets against Pathogenic Bacteria via Lipid Peroxidation. J. Phys. Chem. C. 2012, 116, 17280-17287. 17 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 23

(11) Cui, Y.; Kim, S.; Jones, S.; Wissler, L.; Naik, R.; McAlpine, M. Chemical Functionalization of Graphene Enabled by Phage Displayed Peptides. Nano Lett. 2010, 10, 45594565. (12) Rajesh, C.; Majumder, C.; Mizuseki, H.; Kawazoe, Y. A Theoretical Study on the Interaction of Aromatic Amino Acids with Graphene and Single Walled Carbon Nanotube. J. Chem. Phys. 2009, 130, 124911- 124916. (13) Schedin, F.; Geim, A.; Morozov, S.; Hill, E.; Blake, P.; Katsnelson, M.; Novoselov, K. Detection of Individual Gas Molecules Adsorbed on Graphene. Nat. Mat. 2007, 6, 652-655. (14) Guo, C.; Ng, S.; Khoo, S.; Zheng, X.; Chen, P.; Li, C. RGD-Peptide Functionalized Graphene Biomimetic Live-Cell Sensor for Real-Time Detection of Nitric Oxide Molecules. ACS Nano. 2012, 6, 6944-6951. (15) Tang, L.; Wang, J.; Lim, T.; Bi, X.; Lee, W.; Lin, Q.; Chang, Y.; Lim, C.; Loh, K. HighPerformance Graphene-Titania Platform for Detection of Phosphopeptides in Cancer Cells. Analytical Chem. 2012, 84, 6693-6700. (16) Novoselov, K.; Geim, A.; Morozov, S.; Jiang, D.; Katsnelson, M.; Grigorieva, I.; Dubonos, S.; Firsov, A. Two-Dimensional Gas of Massless Dirac Fermions in Graphene. Nature. 2005, 438, 197-200. (17) Novoselov, K.; Jiang, Z.; Zhang, Y.; Morozov, S.; Stormer, H.; Zeitler, U.; Maan, J.; Boebinger, G.; Kim, P.; Geim, A. Room-Temperature Quantum Hall Effect in Graphene. Science. 2007, 315, 1379. (18) Wang, Y.; Li, Z.; Wang, J.; Li, J.; Lin, Y. Graphene and Graphene Oxide: Biofunctionalization and Applications in Biotechnology. Trends in Biotech. 2011, 29, 205-212.

18 ACS Paragon Plus Environment

Page 19 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(19) Kim, S.; Kuang, Z.; Slocik, J.; Jones, S.; Cui, Y.; Farmer, B.; McAlpine, M.; Naik, R. Preferential Binding of Peptides to Graphene Edges and Planes. J. Am. Chem. Soc. 2011, 133, 14480-14483. (20) Salzmann, C.; Ward, M.; Jacobs, R.; Tobias, G.; Green, M. Interaction of Tyrosine-, Tryptophan-, and Lysine-Containing Polypeptides with Single-Wall Carbon Nanotubes and Its Relevance for the Rational Design of Dispersing Agents. J. Phys. Chem. C. 2007, 111, 18520-18. (21) Xie, H.; Becraft, E.; Baughman, R.; Dalton, A.; Dieckmann, G. Ranking the Affinity of Aromatic Residues for Carbon Nanotubes by Using Designed Surfactant Peptides. J. Peptide Sci. 2008, 14, 139-151. (22) Li, X.; Chen, W.; Zhan, Q.; Dai, L.; Sowards, L.; Pender, M.; Naik, R. Direct Measurements of Interactions between Polypeptides and Carbon Nanotubes. J. Phys. Chem. B. 2006, 110, 12621-12625. (23) Pandey, R.; Kuang, Z.; Farmer, B.; Kim, S.; Naik, R. Stability of Peptide (P1 and P2) Binding to a Graphene Sheet via an All-Atom to All-Residue Coarse-Grained Approach. Soft Mat. 2012, 8, 9101-9109. (24) Caldwell, J.; Agard, D.; Kollman, P. Free Energy Calculation on Binding and Catalysis by α-lytic Protease: The Role of Substrate Size in the P1 Pocket. Proteins. 1991, 10, 140-148. (25) Gilson, M.; Given, J.; Bush, B.; McCammon, J. The Statistical-Thermodynamic Basis for Computation of Binding Affinities: A Critical Review. Biophys. J. 1997, 72, 1047-1069. (26) Karplus, M.; McCammon, J. Molecular Dynamics Simulations of Biomolecules. Nat. Struct. Mol. Biol. 2002, 9, 646-652. (27) Cheng, Y.; Liu, G.; Li, Z.; Lu, C. Computational Analysis of Binding Free Energies between Peptides and Single-Walled Carbon Nanotubes. Physica A. 2006, 367, 293-304.

19 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 23

(28) Shen, J.; Wu, T.; Wang, Q.; Pan, H. Molecular Simulation of Protein Adsorption and Desorption on Hydroxyapatite Surfaces. Biomaterials. 2008, 29, 513-532. (29) Heinz, H. Computational Screening of Biomolecular Adsorption and Self-Assembly on Nanoscale Surfaces. J. Comput Chem. 2010, 31, 1564-1568. (30) Heinz, H.; Farmer, B.; Pandey, R.; Slocik, J.; Patnaik, S.; Pachter, R.; Naik, R. Nature of Molecular Interactions of Peptides with Gold, Palladium, and Pd-Au Bimetal Surfaces in Aqueous Solution. J. Amer. Chem. Soc. 2009, 131, 9704-9714. (31) Tomasio, S.; Walsh, T. Modeling the Binding Affinity of Peptides for Graphitic Surfaces. Influences of Aromatic Content and Interfacial Shape. J. Phys. Chem. C. 2009, 113, 8778-8785. (32) Walsh, T.; Tomasio, S. Investigation of the Influence of Surface Defects on Peptide Adsorption onto Carbon Nanotubes. Mol. Biosys. 2010, 6, 1707-1718. (33) Yancey, J.; Vellore, N.; Collier, G.; Stuart, S.; Latour, R. Development of Molecular Simulation Methods to Accurately Represent Protein-Surface Interactions: The Effect of Pressure and Its Determination for a System with Constrained Atom. Biointerphases. 2010, 5, 85-95. (34) Hoefling, M.; Iori, F.; Corni, S.; Gottschalk, K. Interaction of Amino Acids with the Au(111) Surface: Adsorption Free Energies from Molecular Dynamics Simulations. Langmuir. 2010, 26, 8347-8351. (35) Patwardhan, S.; Emami, F.; Berry, R.; Jones, S.; Naik, R.; Deschaume, O.; Heinz, H.; Perry, C. Chemistry of Aqueous Silica Nanoparticle Surfaces and the Mechanism of Selective Peptide Adsorption. J. Am. Chem. Soc. 2012, 134, 6244-6256.

20 ACS Paragon Plus Environment

Page 21 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(36) Mijajlovic, M.; Penna, M.; Biggs, M. Free Energy of Adsorption for a Peptide at a Liquid/Solid Interface via Nonequilibrium Molecular Dynamics. Langmuir. 2013, 29, 29192926. (37) Wang, Y.; Ai, H. Theoretical Insights into the Interaction Mechanism between Proteins and SWCNTs: Adsorptions of Tripeptides GXG of SWCNTs. J. Phys. Chem. B. 2009, 113, 9620-9627. (38) Beck, D.; Alonso, D.; Inoyama, D.; Daggett, V. The Intrinsic Conformational Propensities of the 20 Naturally Occurring Amino Acids and Reflection of These Propensities in Proteins. Proc. Natl. Acad Sci. 2008, 105, 12259-12264. (39) Materials Studio (Version 5.5) [Software]. (2010). San Diego, CA: Accelrys. (40) Direct Force Field (Version 7.0) [Software]. (2011). San Diego, CA: Aeon Technology Inc. (41) LAMMPS (March 2011 Version) [Software]. (2011). Albuquerque, NM: Sandia National Laboratories. (42) Kyte, J.; Doolittle, R. A Simple Method for Displaying the Hydropathic Character of a Protein. J. Mol. Biol. 1982, 157, 105-132. (43) Vogler, E. Structure and Reactivity of Water at Biomaterial Surfaces. Adv. Colloid. Interface Sci. 1998, 74, 69-117. (44) Skelton, A.; Liang, T.; Walsh, T. Interplay of Sequence, Conformation, and Binding at the Peptide-Titania Interface as Mediated by Water. App. Mater. Interfaces. 2009, 1, 1482-1491. (45) Wegenhart, B.; Tan, L.; Held, M.; Kieliszewski, M.; Chen, L. Aggregate Structure of Hydroxyproline-rich Glycoprotein (HRGP) and HRGP Assisted Dispersion of Carbon Nanotubes. Nanoscale Res. Lett. 2006, 1, 154-159.

21 ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 23

(46) Huang, K.; Li J.; Wu, Y.; Liu, Y. Amperometric Immunobiosensor for α-Fetoprotein Using

Au

Nanoparticles/Chitosan/TiO2-Graphene

Composite

Based

Platform.

Bioelectrochemistry. 2013, 90, 18-23. (47) Wang, J.; Zhao, Y.; Ma, F.; Wang, K.; Wang, F.; Xia, X. Synthesis of a Hydrophilic PolyL-lysine/Graphene Hybrid through Multiple Non-covalent Interactions for Biosensors. J. Mater. Chem. B. 2013, 1, 1406-1413. (48) Nepal, D.; Geckeler, K. Proteins and Carbon Nanotubes: Close Encounter in Water. Small. 2007, 3, 1259-1265. (49) Anversa, J.; Piquini, P. The Effects of an Explicit Water Environment on the Interaction of a Single Wall Carbon Nanotube with Amino Acids: A Theoretical Study. Chem. Phys. Lett. 2011, 518, 81-86. (50) Lyubartsev, A.; Martsinovski, A.; Shevkunov, S.; Vorontsov-Vel’yaminov. New Approach to Monte Carlo Calculation of the Free Energy: Method of Expanded Ensembles. J. Chem. Phys. 1992, 96, 1776-1783. (51) Marinari, E; Parisi, G. Simulated Tempering: A New Monte Carlo Scheme. Europhys. Lett. 1992, 19, 451-458. (52) Geyer, C.; Thompson, E. Annealing Markov Chain Monte Carlo with Applications to the Ancestral Inference. J. Am. Stat. Soc. 1995, 90, 909-920.

22 ACS Paragon Plus Environment

Page 23 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table of Contents Image

23 ACS Paragon Plus Environment