Model
Background
With great Covid delay, the Oxybegone team is in possession of most of our synthesized DNA. Alberta has been the Canadian province with the highest per capita Covid cases for most of the pandemic. Provincial Covid restrictions have prevented the team from getting into the lab during this season. Our team’s parts are made available in the registry for future iGEM teams that want to take advantage of the time and expense required for their design and synthesis.
A pivot was made to modeling late in the season since the team was Covid denied access to the lab space required to characterize our constructs. The GEM estrogen receptor designed for transformation and functioning in yeast did not have an available 3D structure in the Protein Data Bank. These two small town high school teams do not have access to x-ray crystallography or the above average computing power required to effectively predict the 3D structure of the synthetically designed GEM protein. In the end, an hER (human estrogen receptor) was used since it was the most similar in structure and active binding site to our GEM construct.
Our Goal
The team’s goal was to conduct 3D modeling in order to further understand the bonding of oxybenzone molecules to both the hER and the GEM protein of our plasmid. Through the use of computer simulation, we visually demonstrate the fitting of the oxybenzone into the hER and the GEM protein to prove the effectiveness of oxybenzone detection within our system. The main focus was to look at the interactions between estradiol and the hER, as well as the interactions between oxybenzone and the hER since we did not have the definitive structure of GEM. It was hypothesized that the binding of estradiol to the hER would give a clearer understanding on how ligand-receptor interactions work and it would provide a model for what a successful binding of oxybenzone to the hER should look like. Based on this first model, a 3D model of oxybenzone binding to the hER was generated to illustrate how oxybenzone could act as an endocrine disruptor and replace the estradiol ligand. The binding of oxybenzone to our plasmid’s GEM protein was also demonstrated. This showcases oxybenzone’s ability to act as an endocrine disruptor, solidifying its capability to negatively affect both aquatic ecosystems and human beings. Using the GEM protein, it would be demonstrated how the interactions occur within our actual oxybenzone detecting device implementation.
Binding of Estradiol to the Human Estrogen Receptor
![binding of estradiol to the human estrogen receptor](https://static.igem.org/mediawiki/2021/0/0f/T--CTR_Alberta_Canada--modelgreenandyelllow.jpg)
Binding of Oxybenzone to the Human Estrogen Receptor
![binding of oxybenzone to the human estrogen receptor](https://static.igem.org/mediawiki/2021/6/6c/T--CTR_Alberta_Canada--modelgreenandblue.jpg)
Locating our GEM protein
Our initial steps in 3D modeling included locating the GEM protein within our plasmid, as we wanted to see if it would be possible to generate a 3D model. Within our system, the plasmid that uses the GEM protein is pHES839, and the sequence was copy and pasted from Addgene into a document for further analysis. We based the sequence on the publication “Robust Synthetic Circuits for Two-Dimensional Control of Gene Expression in Yeast.” From this publication, the sequences that code for GAL4-BS, hER-LBD, and MSN2-AD (GEM) were identified and collected.
![Sequences for GAL4, hERLBD, and MSN2](https://static.igem.org/mediawiki/2021/1/11/T--CTR_Alberta_Canada--mathmodelling_1stimage_wikigeneexpression.jpg)
![Sequences for GAL4, hERLBD, and MSN2](https://static.igem.org/mediawiki/2021/a/a6/T--CTR_Alberta_Canada--neonlist_mathmodelling_sequenceCTR.jpg)
This paper GEM sequence was used to locate the GEM within our own plasmid, to ensure that they matched. However, when attempting to control the sequence on our plasmid, there were noticeable changes in the sequence. Specifically, it looked like part of the n-terminal MSN2-SD region was removed (the last 20-30 base pairs) in the iGEM plasmid. An estimation of where the GEM was involved pinpointing the start and the end of the protein, even though there may have been changes in between. The start of the GAL4-BS region and the end region of the MSN2-AD was used to attain our estimated plasmid GEM protein sequence.
After this, our next step was to find out if there were any significant changes between the two GEM sequences. The DNA was translated into amino acid sequences for a more in depth analysis. The paper GEM amino acid sequence was found using ExPASY and the plasmid GEM amino acid sequence using ORFfinder.
![GEM amino acid sequence using ORFfinder](https://static.igem.org/mediawiki/2021/0/04/T--CTR_Alberta_Canada--bluesheet_modelling_orf_CTR.jpg)
The two sequences were aligned in BLAST (Basic Local Alignment Search Tool) to analyze their differences. They were found to have 5 amino acid missense changes and 7 amino acid deletions.
![5 amino acid missense changes and 7 amino acid deletions](https://static.igem.org/mediawiki/2021/b/bb/T--CTR_Alberta_Canada--modelling_long_list_of_genes_mathmodelling.jpg)
Next, we investigated if any of these changes entailed a disparity in polarity and hydrophobicity. Three primary changes seen on the alignments tab were analyzed. The first difference involved a change from arginine to cysteine. Arginine is polar and hydrophilic, and cysteine is polar and hydrophobic. Thus, there is a change in the hydrophobicity of this acid from the two amino acid sequences. The second difference involved a change from aspartic acid, which is polar and hydrophilic, to glycine, which is non-polar and neutral. Thus, there was a change in both polarity and hydrophobicity. Finally, the third difference analyzed was a change from glycine to valine. Both are non-polar, so the change that occurred was a change in hydrophobicity from neutral to hydrophobic.
Are the identified differences predicted to be impactful? 3D models were generated for each of these amino acid sequences using PHYRE 2 and converted to PDB files that could be imported into Open Source PyMOL. When importing the two GEM structures into PyMOL, the align function was used to compare the two structures. Both of these sequences were mostly identical, except a minor change at the beginning of the sequence in which the plasmid GEM sequence was slightly longer than the paper GEM sequence. Due to the almost identical nature of the sequences, it was concluded that it would be feasible to use the plasmid GEM sequence for the purposes of our 3D modeling. In the future, a greater analysis of the disparities between this paper GEM sequence and the plasmid GEM sequence would need to be conducted in order to attain a more in depth understanding of how these changes would affect the folding of the protein and the corresponding binding processes and affinities, even if the changes are in fact minimal.
Binding of Estradiol to the Human Estrogen Receptor
After attaining a GEM protein model that could be potentially fitted with oxybenzone, the next step was to generate a 3D model of estradiol binding to the hER. As we were unable to locate the specific complex used, a search on the Protein Data Bank was attempted in order to find an estradiol and hER complex to utilize instead. The 1ERE (human estrogen receptor binding domain in complex with 17beta-estradiol) structure was used according to its description on the Protein Data Bank. It was the most viable option for proving adequate representation of the ligand-receptor interactions. By inputting this PDB file into PyMOL, a visual representation of this type of binding was attained, one that we would mimic for our own plasmid design. How the ligand binds to the receptor within this complex was of specific interest.
1ERE complex with estradiol ligand
![1ERE complex with estradiol ligand](https://static.igem.org/mediawiki/2021/f/f3/T--CTR_Alberta_Canada--greenmodell_mathmateical_closeup_.jpg)
In order to attain a clearer picture on ligand-receptor interactions, the ligand preset setting was used and then ribbons were hidden in order to tailor the image to only show the amino acid residues that were interacting with estradiol. Within these images, a few hypotheses were made about the binding behaviour of oxybenzone to the hER. Multiple bindings were seen within 1ERE and thus, there is a potential that oxybenzone would also be able to bind multiple times to the hER. In addition, in order for oxybenzone to properly bind, the analysis of the interactions and orientation of it and the surrounding amino acid residues are especially predominant.
1ERE complex with Estradiol Ligand (Active sites)
![1ERE complex with estradiol ligand (Active sites)](https://static.igem.org/mediawiki/2021/d/d8/T--CTR_Alberta_Canada--colorsofmathmodelling.jpg)
Binding of Oxybenzone
After attaining a model demonstrating the bonding of estradiol and the hER, further investigation was required to replicate this bonding with oxybenzone. The oxybenzone file was found on Molin Instincts. The hER receptor was isolated from the 1ERE structure within PyMOL. This was done because the receptor is often found in complex with estradiol. The fit of oxybenzone and hER bonding should not be interfered with by another ligand.
hER Receptor (seperated from 1ERE)
![hER receptor seperated from 1ERE](https://static.igem.org/mediawiki/2021/5/59/T--CTR_Alberta_Canada--green_model_math_faraway.jpg)
In an attempt to demonstrate the binding of a chosen ligand to a chosen receptor, a process of molecular docking was done. Unfortunately, to conduct thorough molecular docking and analysis through software such as Autodock Vina and PyMOL was not successful. However, valuable insight was gained through utilizing the blind docking web server, CB-Dock. Its features include locating specific binding regions of a protein, and ranking different binding modes.
Molecular docking of Oxybenzone and the Human Estrogen Receptor
In the docking of oxybenzone and the hER receptor, the oxybenzone file from Molin Instincts and the hER receptor isolated from 1ERE was uploaded into CB-Dock. The images below depict the first binding mode of this ligand and receptor interaction.
![First binding mode of this ligand and receptor interaction](https://static.igem.org/mediawiki/2021/1/17/T--CTR_Alberta_Canada--severalcolorsmodelling.jpg)
![First binding mode of this ligand and receptor interaction](https://static.igem.org/mediawiki/2021/d/d5/T--CTR_Alberta_Canada--cyanmodelling.jpg)
CB-Dock also provided an analysis table with each binding mode's Vina Score, the empirical scoring function which calculates the affinity of protein-ligand binding. Upon research, we discovered that CB-Dock will display the five best poses of the ligand through which it is bonded with the protein and their energies. If just analyzing Vina Score, according to CB-Dock, the most effective pose is that with an energy of -7.4. Docking of oxybenzone and the GEM protein, utilizing the PHYRE 2 generated structure from before were also conducted.
![Docking of oxybenzone and the GEM protein](https://static.igem.org/mediawiki/2021/a/a9/T--CTR_Alberta_Canada--7.4bindingpicture.jpg)
Molecular Docking of Oxybenzone and the GEM Protein
![Docking of ozybenzone and the GEM protein](https://static.igem.org/mediawiki/2021/1/17/T--CTR_Alberta_Canada--purplemodelfarway.jpg)
![Docking of oxybenzone and GEM protein](https://static.igem.org/mediawiki/2021/3/3b/T--CTR_Alberta_Canada--purplemodellingcloseup.jpg)
According to CB-Dock, the most effective pose in this interaction is that with an energy of -7.8.
![The most effective post in this interaction is that with an energy of-7.8](https://static.igem.org/mediawiki/2021/7/7c/T--CTR_Alberta_Canada--7.8bindingpicture.jpg)
Ultimately, the use of CB-Dock provided us with an introduction into the realm of molecular docking. A general visual exemplification of how oxybenzone can bind to both the hER and the GEM protein was obtained. An exposure to the concept of molecular docking was gained, as well as the concepts of binding affinity and Vina Scores. In the future, further research into the analysis of binding affinities when utilizing software such as CB-Dock will be required to go beyond a visual demonstration of bonding. Molecular docking with other software such as Autodock Vina in accordance with PyMOL should be pursued to gain a more in depth analysis of these ligand-receptor interactions.
Mutagenesis
The modeling team gained an introductory experience on the process of mutagenesis when working with Alina Arvisais (University of Waterloo) and Dia Michailidou-Koupan (University of Lethbridge). Homology modeling was conducted by constructing an atomic resolution model of our GEM protein and 1ERE structure. Clustal Omega was used in order to align the two sequences and when analyzing the aligning receptor regions, it was observed that the GEM protein contained a valine amino acid that the 1ERE did not. Thus, within PyMOL, the process of analyzing a point mutation with the substitution of the valine amino acid was completed.
Mutagenesis of 1ERE and GEM protein.
![Mutagenesis of 1ERE and GEM protein](https://static.igem.org/mediawiki/2021/d/d9/T--CTR_Alberta_Canada--RED_PROTIEN.jpg)
Future Modeling
Future modeling will include more intensive analysis with Autodock Vina and PyMOL, further mutagenesis work, and greater research into the measurements of binding affinities as described previously. In addition, it is hypothesized that conducting mathematical modeling would also be useful to the project. Mathematical modeling to predict the electrochemical signal levels produced in relation to certain oxybenzone concentrations would be conducted using the significant addition of LacZ to our receptor. This data would be used for the further development of the project’s systems-level and would be used in future lab work done with S. cerevisiae.