This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Looking for help with molecular dynamics simulation of EEF1A2 D91N variant vs wild-type

Hello everyone,

I am the parent of a child carrying a heterozygous EEF1A2 D91N (Asp91Asn) variant.

I have been trying to understand whether this variant may primarily affect protein stability rather than completely disrupting function.

My current hypothesis is: • D91 is a highly conserved buried residue. • The mutation replaces Aspartate (negatively charged) with Asparagine (neutral). • Structural models suggest a salt bridge may be replaced by a weaker hydrogen-bond network. • Because the residue is buried, I suspect the mutation could subtly destabilize the folded state without causing complete misfolding. • This could potentially increase local flexibility (“protein breathing”), partial unfolding events, or susceptibility to proteasomal degradation.

I would like to compare wild-type EEF1A2 and D91N using molecular dynamics simulations.

Questions:

1.  Would MD simulations be suitable for detecting potential stability differences between WT and D91N?
2.  Which metrics would be most informative?
•   RMSD
•   RMSF
•   Hydrogen bond occupancy
•   Solvent accessibility
•   Salt bridge persistence
•   Free energy calculations
3.  How long would simulations likely need to be (100 ns, 500 ns, 1 µs)?
4.  Would anyone be interested in helping perform or set up such a comparison?

My main goal is to determine whether D91N behaves like a mildly destabilizing buried variant rather than a complete loss-of-function mutation.

Any advice would be greatly appreciated.

Thank you!

protein eef1a2 simulation
• 409 views • link
updated by Michael Dondrup Michael Dondrup • written by fossy217 • fossy217

Hi, I was wondering if you were still interested in going forward with this or if you have more questions. It is possible to set up such a simulation in CharmmGUI from the PDB structure, but it needs maybe some significant trimming to the binding site. One also needs to include replication. You would start with a short simulation run of the wild-type to assess the time required for up to one microsecond. Then inspect the simulation to assess whether the GDP dissociates from the complex at all.

• 0 views • link
written by Michael Dondrup Michael Dondrup

2 answers

D -> N mutation is generally not disruptive, but it can be troublesome if the Asp negative charge is required. To put it differently: it is a modest substitution structurally, but electrostatically it could be a major one.

Based on your description of being buried, I am going to assume that D91 is not involved in metal coordination. That leaves salt bridges. I suggest you do predictors like AlphaMissense or FoldX before doing a molecular dynamics simulation. These methods may answer your question if there is a major energy change, and both will take less time and resources than simulations.

• 0 views • link
written by Mensur Dlakic Mensur Dlakic

I checked the AlphaMissense predictions for EEF1A2 D91 and found something quite interesting.

At position D91, AlphaMissense gives a mean score of 1.0, which is essentially the maximum possible pathogenicity score.

Even more interesting, all tested substitutions at this position are predicted to be pathogenicThis suggests that residue 91 is extremely sensitive to amino acid changes.

What caught my attention is that even D-E (Aspartate to Glutamate) is predicted to be pathogenic. Since Asp and Glu are both negatively charged and chemically very similar, this may indicate that not only the charge but also the precise geometry and local interactions at this position are important.

My working hypothesis is that D91 is a structurally important buried residue. The D91N variant replaces a negatively charged Aspartate with a neutral Asparagine, potentially disrupting a salt bridge or other stabilizing interactions. If so, the mutation might lead to local destabilization, altered dynamics, or reduced protein stability rather than complete loss of folding.

• 0 views • link
written by fossy217 • fossy217

My working hypothesis is that D91 is a structurally important buried residue. The D91N variant replaces a negatively charged Aspartate with a neutral Asparagine, potentially disrupting a salt bridge or other stabilizing interactions.

That may be spot-on according to the described function of D91, even though it is not making contact with the phosphate group of GDP itself, it may be essential for releasing it.

• 0 views • link
written by Michael Dondrup Michael Dondrup

I did a quick literature survey, starting from the PDB structure of EEF1A2.

Overview

EEF1A2 is a gene encoding eEF1A, a eukaryotic translation elongation factor that is expressed in nerve and muscle cells. In simple tems the functions of the protein is to support the protein expression machinery by bringing tRNA loaded with amino acids to the ribosome. eEF1A is essential for proper neuron-development and function. The reason why there is a neuron-specific gene for this factor is unknown.

Mutations in EEF1A2 are rare with very few cases described in the literature, all leading to neuron-related disorders with a prevalence estimated at only 2.92 per 100,000. These mutations are believed to be de novo mutations (not inherited) and all described cases I found in the literature were heterozygous. As far as I could see, there is only one case described in the literature of a c.271G->A (p.D91N) mutation.

Based on the heterozygous genotype (only one copy of a missense mutation is sufficient to cause the phenotype), animal and cell experiments, researchers have proposed that the effect of some mutations in EEF1A2 is in fact due to a toxic gain of function, as opposed to loss of function. Note that in this study, the D91N mutation was not investigated. One hypothesis is that the missense protein may sequester or block its binding partners, thereby inhibiting the proper protein translation.

Role of D91

The X-ray structure of mammalian eEF1A has been resolved and some attention has been given to the D91 residue of the protein. eEF1A binds GTP/GDP, a nucleotide and often found in molecular switches. It is believed that D91 is itself part of a molecular switch (Switch II), D91 changes its conformation – upon interaction with the recycling factor eEF1Ba – by 180 degrees, then forms a bond with another residue (K20), and thereby allows the release of bound GDP (See Figure 4C&D in Crepin et al. (2014))

Hypothesis

The exchange of the Aspartate (negatively charged) with Asparagine (neutral) may interfere with the function of molecular switching (switch II in Figure. 4CD) e.g. by not being able to form a K20-N91 salt-bridge and possibly inhibiting the release of GDP and further locking in the conformation, which again could have consequences for the dissociation from tRNA, the recycling factor, and ribosomes. This would explain why a heterozygous mutation could be sufficient to cause the effect, because the missense protein is not only non-functional but also sequesters its binding partners.

For structural analysis and possibly MD simulations, this means that the binding partners should be considered. In particular, one should focus on the conformation of switch II and the K20-D91/N91 bridge, and the effect on the release of GDP.

Disclaimer

I am not an expert in this field, just a bioinformatician who skimmed the literature. I hope the information is helpful. I am not an expert in MD simulations myself, but several researchers in my group are. I may be able to assist you with further analysis; protein-ligand simulations will require substantial computational resources, though. Please feel free to contact me (email in profile). Note: LLM/AI tools were deliberately not used to write this post.

References

  • Thibaut Crepin, Vyacheslav F. Shalak, Anna D. Yaremchuk, Dmytro O. Vlasenko, Andrew McCarthy, Boris S. Negrutskii, Michail A. Tukalo, Anna V. El'skaya, Mammalian translation elongation factor eEF1A2: X-ray structure and new features of GDP/GTP exchange mechanism in higher eukaryotes, Nucleic Acids Research, Volume 42, Issue 20, 10 November 2014, Pages 12939–12948, https://doi.org/10.1093/nar/gku974

  • Lam, W.W.K., Millichap, J.J., Soares, D.C., Chin, R., McLellan, A., FitzPatrick, D.R., Elmslie, F., Lees, M.M., Schaefer, G.B., DDD study and Abbott, C.M. (2016), Novel de novo EEF1A2 missense mutations causing epilepsy and intellectual disability. Mol Genet Genomic Med, 4: 465-474. https://doi.org/10.1002/mgg3.219

  • Mohamed MS, Klann E. Autism- and epilepsy-associated EEF1A2 mutations lead to translational dysfunction and altered actin bundling. Proc Natl Acad Sci U S A. 2023 Sep 19;120(38):e2307704120. doi: 10.1073/pnas.2307704120.

• 0 views • link
written by Michael Dondrup Michael Dondrup

Log in to answer this question.

More posts like this