Team:UNSW Australia/Model/MD

Molecular Dynamics

Why did we do Physical Modeling?

Modeling the physical structure of a protein can reveal how its various domains and subunits arrange themselves in space. A protein’s structure is critical to its function, thus by modeling a protein’s structure, we can better understand its role1. The exact combination of proteins used to build our scaffold has not been previously tried, and thus knowledge of its physical arrangement is limited. We ran molecular dynamics simulations of the prefoldin complex in order to visualize how flexible its components are1. From this we were able to estimate the feasibility of being able to attach linkers, tags and other proteins to the prefoldin to form a scaffold. We believe that the simulation run on prefoldin paves the way for more complex simulations of the entire scaffold to be run in future. We were also able to estimate physical parameters important for our mathematical model.

Aim of Modeling

We aimed to determine how the various appendages of the prefoldin molecule move in space. From this we aimed to estimate how close the enzymes could be clustered together when attached to the prefoldin complex.

Molecular Dynamics

We ran molecular dynamics on half of the alpha-beta prefoldin hexamer, with PDB ID 1FXK (Video 1). Due to our inexperience with running MD we enlisted the help of Donald Thomas from UNSW’s School of Chemistry. The amber software suite was used to run the molecular dynamics simulations, with a 1000 picoseconds simulation of the prefoldin molecule produced4; 5. Prior to running the simulation, the PDB file had to be manually corrected by removing unwanted ions and incorrect amino acids. This was required as many crystal structures are published with molecules left over from the crystallization, such as glycerol.

INSERT VID/GIF of MD from Daniel’s page

Video 1: Molecular dynamics simulation of the half-hexamer of prefoldin. The half hexamer contains 1 alpha and 2 beta subunits of the prefoldin hexamer. The two beta subunits flank the central alpha subunit.

From this simulation it is clear the prefoldin complex is flexible. The beta subunits can be seen branching away from the central alpha subunit, reaching a distance of 60 angstroms, whilst also collapsing in to a distance of 25 angstroms (Figure 1). Due to the flexible nature of prefoldin we believe it would be possible to attach a range of different enzymes to the scaffold, without being limited by the size of the enzyme to be attached.

Insert Figures of the 60 angstrom distance and 25 angstrom distance (on the folder Daniel uploaded to). Put 60 on the left, and 25 on the right (see legend)

Figure 1: The straight line distance between the C-termini of the alpha and beta subunits was measured in VMD. The image on the left shows the subunits as they expand apart, the image on the right shows the subunits as they close together.

Furthermore, our actual system contains linkers composed of nine small residues (serines and glycines) attached to the C-termini of each subunit, to which the catcher part of the Spy and Snoop Tag-Catcher systems is fused. The longer linkers should allow for even more flexibility of the scaffold, allowing large enzymes to be attached6. The C-terminal coil of one of the beta prefoldin subunits can be seen extending and coiling in the simulation, and pointing in different directions (Figure 2). Thus we believe that the linkers we have designed in between the prefoldin subunits and the catcher domains will provide further flexibility and space to attach enzymes of varying sizes.

Insert Figure titled “abpfd randomcoil conformations” in the google drive where Daniel uploaded MD

Figure 2: The C-termini coil can be seen retracted on the left image, whilst extended on the right image. This indicates flexibility at the linker regions on the prefoldin molecule.

We believe that these simulations show how the prefoldin molecule can be employed as a scaffold. In future we aim to run more complex simulations involving the tag-catcher system, and enzymes attached to prefoldin.

How this Informed the Mathematical Modeling

From the results of molecular dynamics we were able to estimate what distance the enzymes would be when clustered on the prefoldin complex. We were able to extract the distance between subunits over time, finding that the distance between subunits ranged between 20 and 70 angstroms (Figure 3).3.

Figure 3: The distance between a single alpha and beta subunit over time, measured in angstroms.

From the graph, we estimate that over this short time period the subunits average a separation distance of about 50 angstroms. We used this distance to calculate the diffusion of substrate between the enzymes, and subsequent product yield at this distance in our mathematical model of enzyme kinetics and diffusion.

Discussion

Running molecular dynamics served as an effective method of modelling the physical nature of our system. We were able to run MD on half of the prefoldin hexamer; PDB ID 1FXK. The major obstacle to running the simulations on the full scaffold was the need to edit the pdb files generated from crystal structures. Considering that we assembled the prefoldin-catcher-tag-enzyme complex manually in Chimera our structure likely had more errors. Generally speaking, the major errors that need to be corrected are erroneous ions and amino acids that need to be removed 4,5. When running the simulation on a half hexamer, we have to acknowledge that any results used to inform our use of the full scaffold are extrapolations. The flexibility shown by the half hexamer does lend support to prefoldin's usability as protein scaffold. However without the tag-catcher system and enzymes attached, we cannot know exactly how the molecule will move. We believe that by adding a series of 9 GSG linkers to the prefoldin C-termini, we created more flexibility in the scaffold 6. The added flexibility of the GSG linkers, combined with the flexibility shown of the half-hexamer in molecular dynamics, supports the ability of our scaffold to support a range of enzymes of different sizes (Figure 1, 2).

References

  1. Hegyi, H. and Gerstein, M. The relationship between protein structure and function: a comprehensive survey with application to the yeast genome. JMB 288 147-164 (1999).
  2. Karplus, M and Kuriyan, J. Molecular dynamics and protein function. PNAS 102 6679-6685 (2005).
  3. Siegert, R., Leroux, M., Scheufler, C., UlrichHartl, F., and Moarefi, I. Structure of the Molecular Chaperone Prefoldin: Unique Interaction of Multiple Coiled Coil Tentacles with Unfolded Proteins. Cell 103 621-632 (2000).
  4. R. Salomon-Ferrer, D.A. Case, R.C. Walker. An overview of the Amber biomolecular simulation package. WIREs Comput. Mol. Sci. 3, 198-210 (2013).
  5. D.A. Case, T.E. Cheatham, III, T. Darden, H. Gohlke, R. Luo, K.M. Merz, Jr., A. Onufriev, C. Simmerling, B. Wang and R. Woods. The Amber biomolecular simulation programs. J. Computat. Chem. 26, 1668-1688 (2005).
  6. Chen, X., Zaro, J., and Shen, W. Fusion Protein Linkers: Property, Design and Functionality. Adv Drug Deliv Rev. 65 1357-1369 (2013).