Single-molecule protein sequencing using biological nanopores

Abstract

The last few years have shown promising developments in single-molecule protein sequencing using a biological nanopore. Using only one membrane pore in a salt buffer, an applied potential creates a measurable ion current through the pore’s constriction. If a small molecule passes through the pore, this disruption is visible as a characteristic current pattern based on the molecular structure of the molecule. Recent developments by Henry in 2021 show that single amino acid substitution in short proteins can be detected using this method. Unfortunately, due to the complexity of the reading, de novo sequencing does not yet apply to these single molecule reads. Nevertheless, detecting different variants of a protein is possible. In current work this method is being developed to detect phosphorylation sites on a peptide of biological significance. For the latter, an immunopeptide called IRS2 is chosen to be once and twice phosphorylated. This phosphorylation causes the peptide to be a cancer biomarker. If this low-cost method of post-translational modification detection is successful, it would be an improvement compared to the standard mass spectrometry sequencing approach. In this BEP research, problems faced before and after sequencing the IRS2 peptide are approached from two different perspectives. The first aim is to adapt the data acquisition software to automate the workflow. The second aim is to adapt and model the Freely Jointed Chain (FJC) Model to a heterogeneously charged peptide. Both aims are defined to be applicable to other peptides as well. For the first aim, a LabVIEW plugin was developed using insights from data processing in MATLAB. This tool detects the state of the nanopore reading, responds using voltage control in a closed feedback loop and frequently allows for calibration checks. All features were tested with training sequences from real data and show promising results. However, more testing in the lab is required to determine its accuracy compared to a human operator. For the second aim, the FJC model was analytically adapted to the IRS2 peptide inside the nanopore’s electric field. The energetically most favourable configuration was then sought with a Metropolis Algorithm. As a result, the electrostatic potential was calculated and implemented into the Metropolis Algorithm. Despite the simplification of this method, it is still expandable in a modular way to incorporate additional potential such as charge-charge interactions or springs to simulate backbone flexibility. Finally, improving this model further would eventually lead to the relative positions of all chain elements inside the pore to develop an understanding of the (phosphorylated) IRS2 readings.