A research team has computationally designed a synthetic peptide that enables modular detection of the enzymatic activity of coronavirus 3C-like protease (3CLPro), an enzyme essential for viral replication and maturation. When incorporated into bioluminescent or fluorescent reporter proteins, the optimized peptide produced clear loss-of-signal or gain-of-signal responses corresponding to different 3CLPro expression levels. Unlike conventional tests that primarily indicate whether viral molecules are present, this proof-of-concept strategy measures protease activity, which may more closely reflect functional viral processes.
Current coronavirus diagnostics mainly rely on nucleic acid amplification, antibody detection, or viral antigen detection. Although real-time quantitative polymerase chain reaction (RT-qPCR) is highly sensitive, viral RNA can remain detectable long after infectivity has declined, making it difficult to distinguish active infection from residual genetic material. Antibody tests may respond too late to diagnose current infection, while antigen tests generally detect the presence of structural proteins rather than viral activity. By contrast, 3CLPro cleaves viral polyproteins at multiple sites and is indispensable for coronavirus replication. Its highly conserved substrate-binding pocket therefore offers a promising basis for developing activity-sensitive analytical tools that can complement existing detection methods.
A study (DOI: 10.48130/els-0026-0002 ) published in Engineering in Life Sciences on 22 June 2026 by Tao Wang's & Cheng Zhu's team, Tianjin University, reports an optimized synthetic substrate that substantially improves the responsiveness of luciferase- and green fluorescent protein-based 3CLPro sensors.
The researchers began with the crystal structure of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 3CLPro bound to its native eight-amino-acid substrate, SAVLQSGF. Because the enzyme shows more than 96% sequence identity among global coronavirus strains and retains a particularly conserved substrate-recognition pocket, they used Rosetta FastDesign and the deep-learning model ProteinMPNN to generate alternative peptide sequences while preserving the cleavage site. AlphaFold3 predicted that the candidates occupied the catalytic pocket in conformations resembling the native substrate. Of four selected designs, the peptide PVILQYTT, designated P1, showed a Rosetta energy score of −75.8 Rosetta energy units, compared with −56.1 for the native sequence. A sequence-similarity search also indicated that P1 was distinct from eukaryotic proteins, supporting its orthogonality. The team then expressed 3CLPro at different levels in human embryonic kidney 293T cells to test three reporter architectures. In the first, the peptide was inserted into Gaussia luciferase (Gluc), so protease cleavage reduced luminescence. The native substrate produced a 23% reduction, whereas P1 produced a 67% decrease. Across increasing 3CLPro levels, P1-IN-Gluc displayed a dynamic response range of 52%–80%, compared with 18%–30% for the native sequence. In a second design, Gluc was anchored inside the cell by a transmembrane helix. Cleavage released the enzyme for secretion, creating a gain-of-signal response. A flexible linker generated a 47% signal increase and a dynamic range of 42%–55%, outperforming a rigid linker. For fluorescence-based detection, P1 was incorporated into FlipGFP, which switches from a dark to a fluorescent state after protease cleavage. 3CLPro increased fluorescence 3.9-fold with P1 but only 1.2-fold with the native substrate. As 3CLPro expression rose sixfold, the P1 sensor produced a 2.0- to 5.0-fold fluorescence increase. Finally, P1-IN-Gluc responded dose-dependently to the 3CLPro inhibitor Paxlovid: luminescence rose from a baseline of 7.5×10 4 to 1.5×10 5 and 3.0×10 5 after treatment with 50 and 100 nanomolar, respectively.
Overall, the study demonstrates how computational protein design can transform a conserved viral enzyme substrate into several interchangeable activity-reporting modules. The results establish a cell-based proof of concept rather than a clinical diagnostic test, and further work must evaluate the sensors in virus-infected samples and determine their specificity, sensitivity, and practical detection limits. Nevertheless, the approach provides a flexible foundation for measuring coronavirus protease activity, comparing candidate inhibitors, and developing future assays that report functional viral activity alongside conventional measurements of viral presence.
###
References
DOI
Original Source URL
https://doi.org/10.48130/els-0026-0002
Funding information
This work was supported by the National Key Research and Development Program of China (Grant No. 2024YFC3407002), and the National Natural Science Foundation of China (Grant No. 22577090).
About Engineering in Life Sciences
Engineering in Life Sciences (e-ISSN 1618-2863; p-ISSN 1618-0240) is an international source on bioengineering principles and innovations in life sciences and biotechnology, spanning biochemical engineering, process engineering, industrial chemistry. As a fully open access journal, we aim to promote global relationships among biologists, biotechnologists and bioengineers.
Engineering in Life Sciences
Not applicable
Computational design of 3C-like protease substrate peptide for modular detection of protease activity of coronavirus
22-Jun-2026
The authors declare that they have no competing interests.