Molecular Dynamics of the Rotavirus Calcium Pore
Problem and Introduction
Figure 1. Graphic taken from Troeger et al., (2018)
Current oral vaccines for the human rotavirus are ineffective in malnourished children leading to 200,000 deaths yearly. Valuable research has not been done due to lack of funding which is typical for diseases not found in the West. Another difficulty with rotavirus research has been the nature of the non structural protein 4, which causes the deadly symptoms of a rotavirus infection. It is technically challenging to work with as it is amphipathic and partially disordered. Its inherent instability has led to the general structure of NSP4 being unknown. Because of the global nature of rotavirus, the virus has mutated which in turn has caused the amino acid sequence of NSP4 to be different in various strains of the virus. In order to consider NSP4 as a drug target and to make the most positive global impact, more complete structural information is needed for lab and mutated versions of NSP4. This experiment will will bring awareness to global health disparities by educating listeners on the lack of research being done on rotavirus along with the communities affected by it, while also informing listeners about novel computational methods for biomedical research.
Research
Figure 2. Graphic made by Banks Lab, (2024), Model of Rotavirus Pathogenesis
Rotavirus kills 200,000 children ranging from ages 0-5 yearly especially in Sub-Saharan Africa & South Asia. Spreads through contaminated feces, surfaces, and water. Causes severe diarrhea & vomiting, leading to dehydration and death. Oral vaccines significantly reduced deaths but are much less effective in malnourished children due to their weakened immune responses. During an infection, the rotavirus sends one strand of ribonucleic acid (RNA) into the cells of the villi of the small intestine, producing NSP4 proteins. Four NSP4 come together and form a calcium pore embedded in the membrane of the ER, which disrupts water absorption by releasing Ca2+ into the cytoplasm, disrupting the chloride pump in the cell membrane, leading to the fatal symptoms of rotavirus. This process also stabilizes new viral particles, leading to increased infection throughout the small intestine. Due to recent advancements in artificial intelligence, AlphaFold2, a program that uses machine learning to accurately model known protein structures, was identified as a tool that could be used to develop a model of NSP4, which has an unknown structure. If the NSP4 amino acid sequence runs in the AlphaFold2 program, then a significant confidence level can be achieved for the full-length protein model. The model could then be used for drug design studies.
Hypothesis
If the NSP4 amino acid sequence runs in the AlphaFold2 program, then a significant confidence level can be achieved for the full-length protein model. The model could then be used for drug design studies.
If mutated NSP4 amino acid sequences run in the AlphaFold2 program, then the structures can be analyzed and compared to each other to determine if the mutations significantly change protein structure. Results can then be used to determine differences in calcium pore behavior.
If a tetramer comprised of confident monomer models of NSP4 can be placed in an ER-like membrane, then its folding and dynamics can be assessed and compared to existing biochemical data through molecular dynamics (MD) simulations.
Methods and Experimental Design
Figure 3. Graphic made by Banks Lab, (2024), Computational Workflow for Modeling NSP4 Dynamics
Materials:
Protein Modeling Steps:
sequence
Data
NSP4 SA11 Runs:
CHARMM-GUI Bilayer Set-Up With Tetramer Configuration of SA11 Monomer:
NSP4 MX05-126 Runs:
NSP4 MX04-27 Runs:
NSP4 MW013274 Runs:
NSP4 ALN69879 Runs:
NSP4 ALN69860 Runs:
Photos generated from AlphaFold2 and CHARMM-GUI
Abstract
Non-structural protein 4 (NSP4) is found to interact with the endoplasmic reticulum (ER) and is involved in rotavirus (RV) infection, causing alterations in Ca2+ ion localization that lead to disease. Thus, NSP4 is vital in RV replication, morphogenesis, and pathogenesis. It is recognized that the ER plays a key role in rotavirus replication and disease by being the target of NSP4’s activity. It is shown that NSP4 alters the endoplasmic reticulum Ca2+ ion concentration by releasing it from the ER lumen into the cytoplasm. This alteration in the electrochemical gradient between the ER and cytoplasm is a major contributor to rotavirus disease. By assessing the natural conditions of cytoplasm and ER of an enterocyte during the naturally occurring infection, it is possible to understand the activity of NSP4 and to design drugs to target RV. However, due to the limitations of in vitro methods to study NSP4 structure (including cost and time), computational analysis of the protein is a viable alternative to simulate NSP4 dynamics. This project sought to obtain a confident model of NSP4 and assess the behavior of NSP4 conformation and dynamics in a membrane model that mimics the human enterocyte ER built using CHARMM-GUI. Using the AI system AlphaFold2, the 3D structure of NSP4 was modeled. Subsequently, a tetramer configuration created in ClusPro 2.0 of the confident model with glycans was placed in a membrane model that mimics the human enterocyte ER using CHARMM-GUI.
Discussion
The monomer models of NSP4 consistently have high confidence, but the C-terminus is consistently unconfident. This could be due to the nature of the protein and how it functions in its naturally occurring state or it could be due to limitations in the AlphaFold2 program. AlphaFold2 can accurately predict the structures of over 200 million proteins that have known structural data. The structure of NSP4 is largely unknown, therefore the AlphaFold2 program is more limited with what can be predicted.
The mutated monomer models of NSP4 have similar structures, consistently having a very high confidence with a low confident C-terminus. The structure of mutated proteins is very important to rotavirus research to identify potential drug target because, with proteins, structure determines function. Depending on how mutated the the NSP4 protein is, the forming and function of the calcium pore could drastically change, changing what could be considered a drug target. While doing the runs for MX04-27, AlphaFold2 underwent an update which led to faster runs of the program and changes in how the protein structure is represented. This update may have led to some inconsistencies in protein representations or changes in how AlphaFold2 predicts structure.
The creation of a tetramer model of the NSP4 monomer models from the SA11 strain of the rotavirus that is embedded in the ER-like layer is a large and necessary step for continuing research. Due to the limitations of NSP4 for in vitro studies, the computational approach to study and assessing behavior in the ER membrane is a viable option for identifying drug targets, leading to life-saving therapies for the children impacted by rotavirus.
Future directions for this project include performing MD simulations of the glycosylated tetramer model of the NSP4 protein to obtain data about how the calcium pore forms and functions within the cell. Future research could also include identifying the structures of more mutated proteins and optimizing the confidence of the structures of the proteins included in this experiment.
Results
Molecular Dynamics Set-Up Steps:
Independent Variable - the amino acid sequences
Dependent Variable - confidence levels of the protein
Scan for better photos of data
Special thanks to Blessing Akinmade, Jada Henry, and Frank Williams!
Acknowledgments