Sub-Picomolar Detection of SARS-CoV-2 RBD via Computationally-Optimized Peptide Beacons

Soumya Pratap Tripathy, Manvitha Ponnapati, Joseph M. Jacobson, Pranam ChatterjeebioRxiv 2021 · DOI

TL;DR

This paper introduces novel peptide beacons engineered using deep learning for highly sensitive detection of the SARS-CoV-2 spike protein's receptor-binding domain (S-RBD). These beacons achieve sub-picomolar detection limits, paving the way for rapid, point-of-care diagnostic platforms.

What Problem Does This Paper Address?

The ongoing threat of SARS-CoV-2 necessitates rapid and sensitive diagnostic tools. While existing methods like RT-PCR are accurate, they are often slow, expensive, and require complex laboratory procedures. Antigen tests are faster and cheaper but suffer from lower sensitivity. There's a clear need for diagnostic solutions that are both rapid and highly sensitive, capable of detecting the virus at very low concentrations. This paper aims to fill this gap by developing a molecular assay that can directly detect a key viral protein, the S-RBD, with exceptional sensitivity.

Methods

The core of this work lies in the design and engineering of "peptide beacons." These are molecular constructs composed of two peptides that form a heterodimer. Crucially, they incorporate a binding ligand specific to the SARS-CoV-2 S-RBD and a fluorophore-quencher pair.

Here's a breakdown of their approach:

  • Deep Learning-Based Protein Design: They leveraged deep learning tools (trRosetta, Rosetta) for protein structure prediction and energy-based modeling to design the peptide sequences. This allowed them to computationally optimize the beacon's structure and function.
  • Peptide Beacon Architecture: The beacons are designed to act as conformational switches. In the absence of S-RBD, they adopt a "closed" conformation where a fluorophore and quencher are held close together, resulting in low fluorescence due to Förster Resonance Energy Transfer (FRET). Upon binding to S-RBD, the beacon "opens," separating the fluorophore and quencher, leading to a significant increase in fluorescence.
  • Computational Docking: They used HDOCK to computationally verify that their designed peptide sequences could indeed bind to the S-RBD.
  • In Vitro Characterization: Promising candidates were synthesized and tested in vitro. This involved measuring fluorescence signals in the presence of varying concentrations of S-RBD and a control protein (Influenza HA) to assess sensitivity and specificity.
  • In Cell Validation: To further validate binding, they adapted a cellular assay where the peptide beacons were fused to an E3 ubiquitin ligase. This allowed them to measure S-RBD binding indirectly through the degradation of an S-RBD-GFP fusion protein.

Datasets

Not explicitly discussed as distinct datasets for training. The paper focuses on designing new molecular entities rather than training a model on existing data. The S-RBD protein structure used for computational docking was derived from PDB 6MOJ.

Key Insights and Ideas

  • Deep Learning for Molecular Design: This work demonstrates the power of deep learning not just for analyzing existing biological structures but for designing novel functional molecular entities like peptide beacons.
  • Conformational Switching for Sensitivity: The peptide beacon design cleverly utilizes a conformational change triggered by S-RBD binding to generate a strong fluorescence signal. This "molecular switch" mechanism is key to achieving high sensitivity.
  • Sub-Picomolar Detection is Achievable: The successful demonstration of sub-picomolar detection limits for S-RBD is a significant achievement, surpassing the sensitivity of many current rapid diagnostic tests.
  • Integration Potential: The authors envision these beacons being integrated into on-chip optical sensors, which could lead to truly portable and rapid point-of-care diagnostic devices.

Key Results

  • Two candidate beacons, C17LC21 and C21LC21, were identified and validated.
  • Sub-picomolar limits of detection (LoD) were achieved:
    • C17LC21: LoD of ~20 fM (femtomolar) with a dissociation constant (Kd) of 1.615 × 10⁻¹² M.
  • C21LC21: LoD of ~400 fM with a Kd of 6.766 × 10⁻¹³ M.
  • Low cross-reactivity: The beacons showed negligible response to Influenza HA, indicating good specificity.
  • In-cell validation confirmed binding: The peptide beacon constructs successfully induced degradation of S-RBD-sfGFP in human cells, validating their binding capability.

Limitations and Future Work

  • Integration into a Point-of-Care Device: While the authors envision integration with on-chip sensors, this paper focuses on the beacon design and initial validation. The actual development and testing of a functional point-of-care device is future work.
  • Clinical Validation: The current study primarily involves in vitro and in-cell experiments. Further validation in clinical samples is necessary to confirm their utility as a diagnostic tool for actual patient samples.
  • Broader Viral Target Detection: The authors suggest that their deep learning pipeline could be adapted to design beacons for other viral species, which is a promising avenue for future research.

Summary updated

Abstract

The novel coronavirus SARS-CoV-2 continues to pose a significant global health threat. Along with vaccines and targeted therapeutics, there is a critical need for rapid diagnostic solutions. In this work, we employ deep learning-based protein design to engineer molecular beacons that function as conformational switches for high sensitivity detection of the SARS-CoV-2 spike protein receptor binding domain (S-RBD). The beacons contain two peptides, together forming a heterodimer, and a binding ligand between them to detect the presence of S-RBD. In the absence of S-RBD (OFF), the peptide beacons adopt a closed conformation that opens when bound to the S-RBD and produces a fluorescence signal (ON), utilizing a fluorophore-quencher pair at the two ends of the heterodimer stems. Two candidate beacons, C17LC21 and C21LC21, can detect the S-RBD with limits of detection (LoD) in the sub-picomolar range. We envision that these beacons can be easily integrated with on-chip optical sensors to construct a point-of-care diagnostic platform for SARS-CoV-2.
BibTeX
@misc{tripathy2021sub,
  author = {Soumya Pratap Tripathy and Manvitha Ponnapati and Joseph M. Jacobson and Pranam Chatterjee},
  title = {{Sub-Picomolar Detection of SARS-CoV-2 RBD via Computationally-Optimized Peptide Beacons}},
  year = {2021},
  doi = {10.1101/2021.06.04.447114},
  howpublished = {bioRxiv},
}
Figure