A genetically encoded BRET-based SARS-CoV-2 Mpro protease activity sensor

COVID-19, caused by SARS-CoV-2, poses a significant global health threat. SARS-CoV-2's main protease, Mpro (also known as 3CLpro), is essential for processing viral polyproteins (ppla and pplab) into functional non-structural proteins, making it a crucial factor in the viral life cycle. Mpro functions as a homodimer, utilizing a conserved Cys-His catalytic dyad to cleave the pplab polyprotein at 11 sites. Its unique cleavage sequence specificity, not recognized by human proteases, makes it an attractive target for antiviral therapies. Existing assays to monitor Mpro activity include fluorescence resonance energy transfer (FRET)-based assays and split-luciferase assays, some of which utilize peptide substrates containing Mpro cleavage sequences. While effective, these methods have limitations. Bioluminescence resonance energy transfer (BRET), a technique relying on energy transfer from a luciferase donor to a fluorescent acceptor, offers advantages such as high signal-to-noise ratio and extended dynamic range. This study aimed to develop highly sensitive and specific BRET-based sensors for real-time monitoring of Mpro activity in live cells and in vitro, addressing the need for improved tools in antiviral drug discovery and functional genomics.

Several studies have used FRET-based in vitro assays to identify antiviral molecules targeting SARS-CoV-2 Mpro. These assays typically employ peptide substrates encompassing Mpro cleavage sequences. Examples include the identification of inhibitors like Boceprevir, GC376, and calpain inhibitors. Other studies have utilized live cell-based assays, such as those employing FlipGFP, which undergoes a fluorescence change upon Mpro-mediated cleavage. Combining FlipGFP and luciferase assays has also proven successful in identifying Mpro inhibitors. BRET technology has been successfully applied to create a range of genetically encoded sensors for live cell studies, detecting various biological events, including proteolytic cleavage. The mNeonGreen (mNG) and NanoLuc (NLuc) protein pair has emerged as a highly effective donor-acceptor pair for BRET due to its excellent spectral overlap and other desirable properties. However, a dedicated, highly sensitive, and specific BRET-based sensor for Mpro activity was still needed.

The researchers designed two BRET-based Mpro sensors by inserting Mpro's N-terminal autocleavage sequences (short: AVLQSGFR; long: KTSAVLQSGFRKME) between mNG (acceptor) and NLuc (donor) proteins. The rationale was that the close proximity of mNG and NLuc in the intact sensor would result in efficient energy transfer, while Mpro-mediated cleavage would separate the proteins, reducing energy transfer. The chosen sequences are highly conserved among SARS-CoV-2 isolates. To assess the peptides' structural flexibility, the researchers used Modeller for structural modeling and performed all-atom molecular dynamics (MD) simulations using NAMD software. HEK293T cells were transfected with the sensors, alone or with plasmids expressing either wild-type Mpro or a catalytically inactive C145A mutant. BRET assays were performed by measuring emission at wavelengths corresponding to NLuc and mNG. The effect of Mpro DNA concentration and time post-transfection were evaluated. The performance of the BRET sensors was compared to a previously reported FlipGFP-based Mpro sensor. Pharmacological inhibition of Mpro by GC376 was assessed in live cells. In vitro assays utilized recombinantly purified Mpro with the sensor lysates to analyze protease activity and GC376 inhibition, with and without PEG 8000 to simulate molecular crowding.

Both short and long BRET sensors showed robust cleavage in live cells co-expressing wild-type Mpro, but not with the C145A mutant, demonstrating high specificity. Cleavage was dependent on Mpro DNA concentration and time, with a discernable decrease in BRET ratio observed at a minimum of 1.25 ng/well of Mpro plasmid DNA. The BRET sensors exhibited faster kinetics and higher specificity compared to the FlipGFP sensor. The BRET assay successfully measured GC376-mediated Mpro inhibition in live cells, with IC50 values in general agreement with previously reported values. In vitro assays also showed concentration-dependent Mpro cleavage of the sensors. Notably, the presence of 25% (w/v) PEG 8000 (simulating molecular crowding) significantly increased the rate of Mpro activity and decreased GC376's inhibitory potency (IC50 values increased substantially). The enzyme kinetic studies revealed cooperativity in the Mpro protein's action, influenced by its dimerization.

The development of these BRET-based Mpro sensors provides a significant advancement in monitoring Mpro activity. Their high sensitivity and specificity, coupled with their applicability in live cells and in vitro, makes them valuable tools for drug discovery. The observed increase in Mpro activity and reduced inhibitor potency in a crowded environment highlight the importance of considering intracellular conditions during drug development. The differences observed between the BRET and FlipGFP sensors suggest that the BRET-based approach offers several advantages in terms of sensitivity and specificity. This work contributes to the toolbox of methods for studying Mpro and evaluating potential antiviral agents against SARS-CoV-2.

This study successfully developed genetically encoded, BRET-based sensors for detecting SARS-CoV-2 Mpro protease activity. These sensors displayed high sensitivity and specificity, enabling the accurate measurement of Mpro activity in both live cells and in vitro assays. The sensors effectively monitored the inhibition of Mpro by GC376. Furthermore, the study revealed the impact of molecular crowding on Mpro activity and inhibitor efficacy. These sensors provide a powerful new tool for antiviral drug discovery and functional genomics research, particularly for studying the effects of Mpro sequence variations.

The study primarily focused on a single Mpro inhibitor, GC376. Further studies are needed to evaluate the sensors' performance with a broader range of inhibitors. The in vitro assays, while using PEG to mimic molecular crowding, might not perfectly replicate the complex intracellular environment. Although the chosen Mpro cleavage sequences are highly conserved, future work could investigate the sensors' performance with other Mpro variants or mutations that may arise during viral evolution.