The Adhesion Mechanism of Marine Mussel’s Foot Protein (mefp):

Adsorption of L-Dopa on α- & β-cristobalite Silica using Density Functional

Theory (DFT)

Shabeer Ahmad Mian*, Younas Khan

Department of Physics University of Peshawar, Pakistan

Electronic Supplementary Information (ESI)

In marine environment the mussels are surrounded by water and the same for the protein

it secreted to adhere to the rocks. In order to mimic this wet environment we first model the

water molecules for its adsorption on cristobalite surface. The 31 water molecules and

cristobalite surface were optimized before their interaction with each other. The same process

was done for the catechol and water and cristobalite surface previously published14-16. Initially we

perform the dry adhesion of ldopa on cristobalite surface in order to compare with the catechol

adsorption on dry surface. The reason for doing such an adsorption is to first confirm the dry

adhesion and their comparative strength. The wet environment we actually introduce in the next

step in order to develop more interest of the readers.

We perform this study in the presence of 31 water molecules to observe the bond strength

of ldopa and water on hydrophilic surface. To do so we consider thirty one (31) water molecules

on the cristobalite surface and also the co-adsorption of water molecule with ldopa amino acid on

the same surface. We optimized the geometries of both molecules and surface separately. The

optimized geometries were considered in such a way that the ldoap is surrounded by thirty one

(31) water on the cristobalite surface. It is much clear from the graphical representation of figure

S1 that the thirty one (31) water molecule surround the adsorbed ldopa on cristobalite surface. It

is observed that bonding sites of adsorbed ldopa are in contact with surface silanols without

being affected by surrounding water. On the dry cristobalite surface the ldoap make 4 hydrogen

bonds with the surface silanols. On the wet surface ldopa make four (4) hydrogen bonds with

surface silanols while two hydrogen bonds with the surrounding water molecule.

In the case of dry adsorption of ldopa on the surface the average hydrogen bond length is

approximately 1.853 Å but after the addition of thirty one (31) water molecules adsorption the

average hydrogen bond length with surface silanols become 1.713 Å. In addition to these four

hydrogen with surface silanols ldopa the upper part of ldopa i.e. alanine make two hydrogen

bonds with the surrounding water. This way ldopa make total of six hydrogen bonds with the

surface and surrounded water molecules. This indicates that introducing of water medium to the

ldopa adsorption further strengthen the binding of ldopa to the cristobalite surface.

In figure S1 the bonding of ldopa with surface silanols and water is clearly shown to

understand the adsorption mechanism. All the optimized geometries of thirty one (31) water

molecules co-adsorbed with the ldopa on the cristobalite surface, thirty one (31) water molecules

adsorbed on the surface, thirty one (31) water molecule and thirty one (31) water molecules

along with ldopa are given in figure S1, S2, S3 and S4 respectively.

Molecular Dynamics Simulation:

The molecular mechanism underlying water-resistant adhesion, an MD simulation for the

L-Dopa adsorption on a wet silica surface containing thirty one (31) water molecules was

performed. The L-Dopa molecule was placed on top of the wet silica surface covered with 31

water molecules (Fig. S5). A long MD trajectory of 10 ps was propagated using the Verlet

algorithm with a time step of 1 fs. The forces on atoms were calculated on the fly using the DFT

based on the Perdew–Zunger local density approximation (LDA) for the exchange–correlation

functional. We used only the Γ point in the integration of the brillouin zone; a 2.72 keV mesh

cutoff was used for the atomic orbitals. The initial velocities of atoms were sampled from the

Boltzmann distribution at temperature of 300 K. Unlike our previous study and similar to our

current optimization of the geometry process the bottom silicon and hydrogen atoms were

constrained to maintain the bulk properties of the silica. As in our previous study we only allow

the surface silanols and silicon atoms while the rest of the system below the surface was under

constrain. As shown in Fig. S5, l-dopa relocated the pre-adsorbed water molecules onto the

surface and finally made direct contact with the surface. We did not observe any protonation of

the water molecule or silanols of the surface. The MD simulations we ran at room temperature

(300 K) using LDA methods. The previous study shows that deprotonation occurred in every

case, and was always started from one of the surface silanols. We did not find any deprotonation

of mutually interacting molecules or for l-dopa interacting with water molecules. There have

been reports that the surface silanols are acidic (as acidic as vinegar) in the presence of a local

strain in geometry. To avoid the acidic nature of surface silanols this time we did not consider

the surface silica rigid due to which no proton formation is observed in the whole MD trajectory.

Therefore, the deprotonation presumably arises from the shortcoming of fixing the most of the

surface in the previous MD simulation.

Figure S1: L-dopa adsorbed in the presence of thirty one (31) water molecules forming an

intermolecular distance of 1.71 Å with Cristobalite surface.

Figure S2: Optimized geometry of thirty one (31) Water molecules adsorbed on Cristobalite

surface

Figure S3: Optimized Geometry of thirty one (31) Water molecules

Figure S4: Optimized Geometry of thirty one (31) Water molecules surrounds L-Dopa

Figure S5: Snapshot taken from 10ps MD simulation of L-Dopa adsorption on a wet silica

surface of thirty one water molecules. The top panel shows the top view and bottom panel

represent side view.