Models

Molecular Models

Click and drag to move the molecule. Scroll to zoom

Note that some larger models will take more time to render, thanks for your patience!

Modified APOE

Modified APOE consists of the addition of a hydrazone bond that connects O-glycosylated APOE to MPBH which is attached to DSPE-PEG-2000-Amine.

The steps for generating this molecule in silico are as follows:
1. A FASTA sequence is provided for the APOE molecule used in experiments (see below):
    MKVLWAALLVTFLAGCQAKVEQAVETEPEPELRQQTEWQSGQRWELALGRFWDYLRWVQTLSEQVQEELL
ATVGSLAGQPLQERAQAWGERLRARMEEMGSRTRDRLDEVKEQVAEVRAKLEEQAQQIRLQAEAFQARLK
SWFEPLVEDMQRQWAGLVEKVQAAVGTSAAPVPSDNH
2. SWISS-MODEL homology modelling is used to generate a protein representative of the FASTA sequence 1. Results show a GMQE score of 0.97 on a scale of 0-1 which indicates high quality. The model output page is shown below:
3. The homology modelled protein is the basis for further modifications.
4. N-Acetyl-d-glucosamine is joined to APOE on Ser-308 (a major glycosylation site)2.using UCSF Chimera.
5. Ser-308 is typically a site for O-glycosylation with core 1 glycans2. Hence, beta-d-galactose and subsequently sialic acid were added onto glucosamine using UCSF Chimera3.
6. DSPE-PEG-Amine is thiolated with Traut’s reagent using UCSF Chimera.
7. MPBH is joined to the DSPE-PEG polymer using UCSF Chimera.
8. A hydrazone bond is formed between sialic acid and MPBH using UCSF Chimera.
9. The final modified and unmodified APOE structures can be found in the models tab under experiments.

1. K. Arnold, L. Bordoli, J. Kopp and T. Schwede, "The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling", Bioinformatics, vol. 22, no. 2, pp. 195-201, 2005.

2. "Protein Annotations for Apolipoprotein E; ARRAY(0x159d298) - GenScript ProtBank Protein Database", Genscript.com, 2016. [Online]. Available: http://www.genscript.com/cgi-bin/protein/protein.pl?id=APOE_HUMAN. [Accessed: 22- Oct- 2016].

3. A. Sørensen, H. Clausen and H. Wandall, "Carbohydrate clearance receptors in transfusion medicine", Biochimica et Biophysica Acta (BBA) - General Subjects, vol. 1820, no. 11, pp. 1797-1808, 2012.

Modified EGF

Modified consists of the addition of DSPE-PEG-2000 MAL which is connected to EGF through a thioether bond.

The steps for generating this molecule in silico are as follows:
1. The starting PDB file for EGF is 1EGF.
2. Traut’s reagent is attached to the N-terminus of EGF using the join models option in UCSF Chimera.
1. The 3D structure of Traut’s reagent is determined using CHEMDRAW’s 3D model preview.
2. For the purposes of this project, we are assuming that only the N-terminus primary amine is thiolated.
3. DSPE-PEG-2000 Maleimide is the joined with thiolated EGF to form the full EGF ligand. The 3D structure of DSPE-PEG-2000 Maleimide is determined using CHEMDRAW’s 3D model preview. This model is quite inaccurate as the polymer chain actually folds on itself according to the 3D image provided by the supplier. The MOL2 file for the actual 3D structure was unobtainable. Only a 2D MOL2 file was obtainable from the supplier. Hence a straight chain polymer was used to represent DSPE-PEG(2000) Maleimide. It should not significantly affect the docking results as DSPE is typically embedded in the liposome bilayer.
4. The final modified and unmodified EGF structures can be found in the models tab under experiments.

EGFR

This model represents the extracellular domain of EGFR. The PDB ID for this molecule is 1NQL.

LDLR

The model represents the extracellular domain of LDLR. The PDB ID for this molecule is 1N7D.

APOE

The model represents unmodified APOE. This model was produced through homology modelling described in the Modified APOE tab.

EGF

The model represents unmodified EGF. This PDB ID for this molecule is 1egf.

Modified APOE Complexes

This model represents the top three complexes determined by ZDOCK for the docking of modified APOE to LDLR. The three complexes are superimposed. The various positions of modified APOE can be identified by the straight polymer chain (PEG).

Unmodified APOE Complexes

This model represents the top three complexes determined by ZDOCK for the docking of unmodified APOE to LDLR. The three complexes are superimposed. The various positions of unmodified APOE can be identified by comparing the unmodified APOE/LDLR complex molecule to the receptor in the LDLR tab.

Modified EGF Complex

This model represents the top three complexes determined by ZDOCK for the docking of modified EGF to EGFR. The three complexes are superimposed. The various positions of modified EGF can be identified by the straight polymer chain (PEG).

Unmodified EGF Complex

This model represents the top three complexes determined by ZDOCK for the docking of unmodified EGF to EGFR. The three complexes are superimposed. The various positions of unmodified EGF can be identified by comparing the unmodified EGF/EGFR complex molecule to the receptor in the EGFR tab.

Rendered with PV

ZDOCK

ZDOCK is an initial stage rigid body protein-protein docking system. Modified EGF/APOE ligands were docked onto their receptors using ZDOCK. Specific residues were blocked from and included in the active sites to improve the accuracy of the docking results. The residues to block and include are determined from literature 1 2.

Modified and unmodified EGF were docked onto the EGF receptor. Modified and unmodified APOE were docked onto the LDL receptor using ZDOCK. The ZDOCK docking results are shown below.

Docking Preparation

Modified and unmodified EGF/APOE and their corresponding receptors are prepared for ZDOCK by using UCSF Chimera. The steps used to prepare the PDB files for docking include:

1. Deleting solvent molecules and ions.
2. Adding all hydrogens using default Chimera settings.
3. The Add Charge function to assign partial charges to atoms.

These modifications are necessary to improve the accuracy of ZDOCK as they affect the desolvation and electrostatics energies determined by ZDOCK3.

EGF Docking Results

The scoring value outputs for modified and unmodified EGF are relatively similar. The average % difference $$\frac{\text{differential}}{\text{EGF}}$$ is 5.55%. Thus there isn’t a significant difference between scoring function values for modified and unmodified EGF. The top five models from the ZDOCK output indicate slight differences in the binding configurations for modified and unmodified EGF to EGFR.

The orientations of EGF to the receptor vary between modified and unmodified EGF as shown in Figure 2. The configurations tend to all be bound to the same location on EGF (domain III), with the exception of one configuration in modified EGF being bound to the domain I location. In reality, EGF should be interacting with both domains I and III. Thus, the ZDOCK results aren’t accurate compared to studies in literature. Binding residues were specified for both Domain I and III in the docking process. Since ZDOCK runs a rigid-body scan, the conformational changes in EGFR required for an accurate EGF configuration (interactions with domain I and III) are not represented by ZDOCK. However, the ZDOCK scoring function results do show that the modifications to EGF do not significantly affect it’s interaction with the active site in domain III of EGFR. Rather, the results seem to indicate that there is a slightly improved interaction (larger scoring function value) for modified versus unmodified EGF. This difference is likely due to the additional favourable electrostatic interactions generated by the additional charges and mass from the modifications. Thus, the modifications to EGF will likely not significantly affect how EGF binds to EGFR.

APOE Docking Results

The scoring value outputs for modified and unmodified APOE are relatively similar. The average % difference $$\frac{\text{differential}}{\text{APOE}}$$ is 7.16%. Thus there isn’t a significant difference between scoring function values for modified and unmodified APOE. The top five models from the ZDOCK output indicate small differences in the binding configurations for modified and unmodified APOE to EGFR.

The orientations of APOE to the receptor vary between modified and unmodified APOE as shown in Figure 4. The binding configurations of APOE to the LDL Receptor are quite similar with the exception of one configuration for modified APOE. APOE’s active site consist of residues 135-151. It seems to primarily interact with the region from residues 150-180 on the LDLR receptor. The larger scoring function values in Figure 3 are likely due to the increased electrostatic interactions due to the presence of modifications with partial charges in the non-cutoff region of ZDOCK. Thus, it seems that the modification does not seem to significantly affect the interaction between APOE and the LDLR receptor as there isn’t a significant difference between the binding configurations and scoring function values in the top five ZDOCK models.

Modelling Conclusion

An initial stage rigid body docking using ZDOCK was conducted for modified and unmodified APOE/EGF with LDLR and EGFR respectively. The models used in the docking software were generated using CHEMDRAW and UCSF Chimera to a reasonable degree of accuracy. The modifications to EGF and APOE do not seem to significantly affect the scoring function values and the binding configurations in ZDOCK relative to unmodified EGF and APOE. The docked complexes may not represent the actual complexes for modified and unmodified EGF and APOE as ZDOCK is an initial stage in docking studies. However, the results do indicate that for the particular set of active sites on the receptors and ligands, the modifications do not significantly affect the interaction between the ligand and receptor. Thus, it is likely that the modifications will not significantly affect the active site interactions for the correct ligand-receptor configuration. However, there are several improvements that can be made to this analysis:

Energy minimization with CHARMM, addition of polar hydrogens with CHARMM and re-ranking complexes using ZRANK.

Step one will improve the accuracy of the ranked list of complexes that represent potential ligand-receptor configurations for EGF and APOE. In addition the scoring function values will be more accurate. Thus we can analyze the effect of modifications to APOE and EGF with more certainty.

Obtain structures of reagents used for the modifications to EGF and APOE from experiments (NMR, XRD).

Step two will improve the accuracy of the models built online. Some of the components used to construct modified EGF and APOE such as DSPE-PEG-2000 maleimide were approximated three dimensional structures based off of bond angles using CHEMDRAW.

Use other docking programs to cross reference ZDOCK.

Rosetta DOCK and PatchDOCK are alternatives to ZDOCK. It is important to cross reference docking results with other docking algorithms to ensure that the results are relatively similar between reputable scoring functions.

1. K. Arnold, L. Bordoli, J. Kopp and T. Schwede, "The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling", Bioinformatics, vol. 22, no. 2, pp. 195-201, 2005.

2. "Protein Annotations for Apolipoprotein E; ARRAY(0x159d298) - GenScript ProtBank Protein Database", Genscript.com, 2016. [Online]. Available: http://www.genscript.com/cgi-bin/protein/protein.pl?id=APOE_HUMAN. [Accessed: 22- Oct- 2016].

3. R. Chen, L. Li and Z. Weng, "ZDOCK: An initial-stage protein-docking algorithm",Proteins: Structure, Function, and Genetics, vol. 52, no. 1, pp. 80-87, 2003.