Introduction

Our project on designing NeoFvs as a therapy for Dengue triggered multiple questions that we found worth exploring through in-silico methods.

The answers to these questions influence our design choices about how we proceeded in our laboratory work.

In-silico modeling has been quite an important aspect of our project, and we have tried to integrate our dry lab and wet lab workflow to the best we could.

Design:

The questions that we investigated are:

• Stability of our NeoFv
  We predicted its structure through DeepMind’s AlphaFold 2.0. Once we predicted the structure of our drug, we wanted to investigate whether it is stable in the conditions at its site of action. We simulated an MD to and verified minimal structural fluctuations over 100ns. We configured it at a temperature of 40°C, which is the body temperature the body reaches under a severe dengue infection.
• Demonstrating pH dependent binding
  One of the features of our peptide is the pH dependent interaction with the human FcRn receptor at pH 6. We wanted to investigate interaction by setting up a more realistic molecular dynamics simulation where the residues were protonated to mimic the endosomal environment of pH 6.
• What is the ideal position for the FcRn binding peptide on our scFv?
  The 16aa peptide that binds to the human FcRn receptor can be placed at either the C-terminus, the N-terminus or the linker between the domains of the scFv. We wanted to investigate what would be the ideal position to place this peptide and we used molecular docking methods on predicted structures of our NeoFv variants to try to solve this problem.
• What is the best conformation of our peptide?
  The FcRn binding peptide has two variants: Cyclic and Linear, where the difference is the presence of a disulfide bridge in the cyclic variant, which loops the peptide onto itself, generating a more stable conformation (hypothesized). We wanted to investigate the difference in the binding affinity of both of these peptides. We used molecular docking methods in combination with molecular dynamics simulations to investigate this.
• Optimizing factors for production of NeoFvs
  There are various factors that control the yield of production of a biologic drug. For industrial scale production of any drug, these values need to be identified and any interactions between various governing factors should be characterized. To do just that, we set up statistically designed experiments to execute in the lab and identified the optimal values of production of our drug, and also characterized how these factors interact.

Details of our modeling efforts are available on the following pages: