Targeting SARS-CoV-2 Main Protease: A Bacteria-Based Colorimetric Assay for Screening Natural Antiviral Inhibitors

Abstract

SARS-CoV-2 main protease (M^pro) is essential for viral polyprotein processing and represents a prime target for antiviral drug discovery. However, most available screening strategies rely on computational predictions or cell-free biochemical approaches that provide limited functional context and often require specialized instrumentation, while mammalian cell-based models remain costly and require high biosafety levels. Accordingly, there remains a shortage of simple, rapid, and biosafe functional screening tools suitable for early-stage prioritization of potential M^pro inhibitors, particularly those derived from natural sources and in urgent situations such as the COVID-19 pandemic. In this study, a bacterial colorimetric reporter assay was developed that directly links SARS-CoV-2 M^pro activity to β-galactosidase function in Escherichia coli. To the best of our knowledge, the developed assay represents the first bacterial colorimetric model for functional detection of SARS-CoV-2 M^pro inhibition based on a phenotypic readout. The assay enables the rapid visual detection of protease inhibition on X-gal-containing medium and provides a cost-effective and biosafe platform for prioritizing candidate inhibitors, under standard laboratory conditions, prior to further validation. Due to its bacterial expression context, this assay is intended for functional screening to provide the most promising candidate compounds and/or extracts for subsequent biochemical or mammalian cell-based validation; it is not intended to determine quantitative potency or to replace further validation approaches. It should be noted that the selective compound uptake in E. coli restricts the range of chemical compositions that can be evaluated using this method. Therefore, proof-of-concept application was demonstrated using pomegranate juice, a representative natural inhibitor source, rather than most currently known specific M^pro inhibitors. In addition, other plant-derived preparations, including rhubarb, grape, and red/black currant juices, were tested demonstrating the assay’s applicability to diverse natural matrices.

1. Introduction

The main protease of novel coronavirus (SARS-CoV-2 M^pro), also known as the 3C-like protease (3CL^pro) or non-structural protein 5 (Nsp5), plays a key role in viral replication by cleaving polyproteins pp1ab and pp1a into functional proteins [1,2]. Accordingly, it is considered a primary drug target in the search for antiviral agents against COVID-19. In addition to its importance in the viral life cycle, SARS-CoV-2 M^pro has no homologs in the human host and exhibits a distinct substrate specificity, recognizing and cleaving at the motif Leu-Gln↓(Ser, Ala, Gly), which minimizes the risk of off-target effects, toxicity, or cross-reactivity in human cells [3,4]. SARS-CoV-2 M^pro is also highly conserved among coronaviruses, reducing the possibility that resistance mutations will affect the efficacy of its inhibitors, making these inhibitors promising candidates for broad-spectrum antiviral development [4].

The discovery and validation of SARS-CoV-2 M^pro inhibitors typically involve a series of screening steps. Initial candidate selection is often performed using computational approaches or in vitro biochemical assays with purified recombinant M^pro to identify compounds that interact with or inhibit the protease and measure their inhibitory potency (IC₅₀ values). During the COVID-19 pandemic, the urgent need to identify anti-SARS-CoV-2 drugs, coupled with the requirement for biosafety level 3 facilities, led most research on M^pro inhibitors to rely on these in silico or in vitro biochemical approaches rather than more complex cell-based or in vivo models [5,6,7]. Promising candidates would then move to cell-based systems, where their antiviral activity and cytotoxicity are tested, often reported as EC₅₀ values, under conditions that more closely resemble host cell physiology. Although these strategies are useful for identifying promising candidates, they have their limitations [8]. For instance, computational docking and molecular dynamics simulations can predict the binding affinity of a potential inhibitor, but do not confirm functional inhibition [9], and their accuracy depends heavily on the data source, necessitating further experimental validation [10]. Similarly, in vitro biochemical assays, using purified SARS-CoV-2 M^pro under highly controlled conditions provide quantitative measurements of enzyme inhibition but often require specialized instrumentation and can be costly. Moreover, such assays have limited capacity to evaluate complex matrices like crude plant-derived preparations [11]. In contrast, mammalian cell-based assays more closely reflect host cell environments but remain costly and require high biosafety levels. Consequently, there remains a need for simple, rapid, and biosafe functional screening tools that can support early-stage prioritization of candidate M^pro inhibitors before more resource-demanding validation steps.

Plants have long served as sources of therapeutic agents, and to this day, tens of thousands of medicinal species remain underexplored, offering a vast reservoir for drug discovery [12]. Their secondary metabolites, such as flavonoids, alkaloids, terpenoids, and phenolics, exhibit significant structural diversity shaped by evolutionary adaptation [13]. This diversity can enable broad biological activity, target selectivity, multitarget potential, and generally favorable safety profiles, making plant-derived compounds attractive candidates for early-stage antiviral research [13,14]. Several phytochemicals, such as quercetin, baicalein, and epigallocatechin gallate (EGCG), have demonstrated inhibitory effects against SARS-CoV-2 M^pro in biochemical and computational studies [5]. In recent years, specifically since COVID-19 pandemic, plants have also been increasingly recognized not only as sources of pharmacologically active compounds but as functional foods that may contribute to disease prevention and immune support through their complex mixtures of bioactive metabolites. A growing body of research highlights the role of polyphenol-rich fruits, vegetables, and plant-based products in modulating antiviral, anti-inflammatory, and immunomodulatory responses, positioning functional foods as promising complementary resources for managing viral infections and highlighting the effect of diet and proper nutrition on response to infections [15,16,17,18,19]. Importantly, such nutritional and prophylactic benefits should not be equated with direct antiviral efficacy or therapeutic potency, and plant-derived compounds identified in this context are generally regarded as starting points or lead structures that may require further optimization within standard drug discovery workflows. Therefore, given their accessibility, sustainability, and established safety, plant-derived preparations/compounds remain valuable resources for the initial exploration and prioritization of potential broad-spectrum antiviral leads. However, despite numerous computational and biochemical findings, their functional activity against SARS-CoV-2 M^pro remains largely unexplored under cellular conditions. The lack of simple, affordable, and biosafe functional assays that can efficiently prioritize candidates before more resource-intensive biochemical and mammalian cell-based steps has impeded this stage of validation [5,20,21].

To address this gap, we developed a bacterial reporter-based colorimetric assay that enables functional assessment of SARS-CoV-2 M^pro inhibition using a phenotypic readout in Escherichia coli cells. The assay is based on insertion of an engineered M^pro cleavage site into the lacZα reporter gene, such that proteolytic activity of M^pro disrupts β-galactosidase function, while inhibition of M^pro restores color development on X-gal-containing medium. To the best of our knowledge, this represents the first bacterial colorimetric model designed for this purpose. The assay offers a simple visual readout that enables rapid evaluation of tested crude preparations/compounds and provides a cost-effective, scalable platform that can be applied under standard laboratory conditions. Importantly, considering the bacterial expression conditions of the assay, it is intended as a simple, first-pass screening and prioritization tool within the broader M^pro inhibitor discovery workflow, rather than as a replacement for established biochemical assays or mammalian cell-based antiviral models. Candidate hits identified using this assay are expected to undergo subsequent validation using conventional in vitro and eukaryotic cell-based approaches. This study describes the development of the assay and following functional validation using pomegranate juice, a natural source of several bioactive compounds with reported antiviral properties, in addition to other different plant-derived preparations including rhubarb, grape, and red/black currant juices, demonstrating the assay’s applicability to diverse natural matrices.

2. Materials and Methods

2.1. De novo Assembly and Sequence Verification of Codon-Optimized SARS-CoV-2 M^pro Gene

The nucleotide sequence of the SARS-CoV-2 M^pro gene was obtained from the isolate SARS-CoV-2/human/USA/WI-CDC-ASC210624442/2021 (GenBank accession no. OM036858.1), corresponding to the B.1.617.2 (Delta) lineage. The sequence was manually codon-optimized, based on E. coli codon usage bias. The primary aim was to achieve stable, moderate intracellular expression suitable for inhibition screening, rather than maximal translational efficiency that could result in false-negative inhibition readouts due to excessive M^pro abundance in bacterial cells. Codon substitutions were introduced selectively to eliminate rare codons clusters that could cause premature termination, while avoiding excessive over-optimization. The applied frequency thresholds (≥0.16 for most amino acids and ≥0.14 for six-codon families) were chosen to exclude only the rarest variants. As a result, some non-optimal but moderately used synonymous codons were intentionally retained to maintain balanced translation kinetics and physiologically relevant expression levels. The final optimized nucleotide sequence is provided in the Appendix A. De novo assembly of the gene was performed using a PCR-based overlapping fragment synthesis strategy. The entire gene was divided into 35 partially overlapping fragments, which were gradually assembled through 16 consecutive PCR reactions. Oligonucleotides and primers for intermediate amplicons were designed using UGENE (Unipro UGENE v40.1) and Vector NTI software (version 1.2.0.Thermo Fisher Scientific/Invitrogen, Carlsbad, CA, USA). Amplifications were carried out using a commercial PCR mix containing Taq DNA polymerase (Thermo Fisher Scientific, Vilnius, Lithuania). The final product was analyzed by electrophoresis on a 1% agarose gel, and verified by Sanger sequencing with terminal primers. Any detected point mutations were corrected by site-directed mutagenesis, followed by re-sequencing to confirm the integrity and accuracy of the final assembled sequence. A schematic representation of the gene assembly process is provided in Figure A1.

2.2. Plasmid Construction

The E. coli colorimetric reporter assay was constructed using the high-copy plasmid pUC19 as the vector backbone. This plasmid was selected for its strong replication in E. coli and its functional lacZα gene, which enables β-galactosidase-mediated hydrolysis of X-gal into a blue chromogenic product. To enable M^pro-dependent color switching, a specific SARS-CoV-2 M^pro cleavage site was inserted within the lacZα coding sequence without disrupting its open reading frame. The inserted nucleotide sequence (AAAACCAGCGCGGTGCTGCAGAGCGGCTTCCGCAAAATGGAG) encodes the amino acid sequence KTSAVLQ↓SGFRKME, corresponding to a SARS-CoV-2 M^pro recognition motif, with cleavage occurring between the Gln and Ser residues. The insert length was maintained as a multiple of three nucleotides, and the flanking codons were preserved to prevent any amino acid substitutions or premature stop codons. This design ensured that β-galactosidase activity remained intact unless the cleavage site was cleaved by M^pro.

The assembled codon-optimized M^pro gene was placed under the control of the lac promoter/operator to allow co-regulated expression of both M^pro and lacZα, and terminated by a T7 terminator sequence. The resulting lacP/O-M^pro-T7 terminator cassette was inserted into the same pUC19 plasmid. All cloning steps were verified by restriction digestion and Sanger sequencing. A schematic representation of the main genetic components is shown in Figure 1.

2.3. Bacterial Strain, Transformation, and Expression Conditions

Chemically competent E. coli DH5α cells (ΔlacZ58(M15)) were used as the host strain for all cloning and expression experiments. This strain was selected for its high transformation efficiency, compatibility with high-copy-number plasmids such as pUC19, and reduced background protease and endonuclease activities, which support stable maintenance of recombinant constructs [22]. In addition, DH5α carries a chromosomal deletion in the lacZ gene, so the strain does not produce functional β-galactosidase on its own. This deletion allows functional complementation with the lacZα fragment encoded by pUC19, making DH5α ideal for colorimetric reporter assays based on β-galactosidase activity [22,23]. Transformation was performed by the standard heat-shock method, followed by recovery in LB medium and plating on LB agar supplemented with ampicillin (100 µg/mL). Plates were incubated at 37 °C for 16 h to obtain recombinant colonies. For expression experiments, single colonies were inoculated into 5 mL of LB broth containing ampicillin and grown overnight (~16 h) at 37 °C with shaking at 250 rpm. The following day, cultures were refreshed into fresh LB medium (1:100 dilution) to maintain exponential growth. Cell density was monitored spectrophotometrically at 600 nm using a Hitachi U-2900 UV–Visible spectrophotometer (Hitachi, Tokyo, Japan). When the optical density (OD₆₀₀) reached 0.6–0.9, expression was induced by adding isopropyl β-D-1-thiogalactopyranoside (IPTG; Helicon, Moscow, Russia) to a final concentration of 0.2 mM. At the same time, 5-bromo-4-chloro-3-indolyl β-D-galactopyranoside (X-gal; dissolved in DMSO, SibEnzyme, Novosibirsk, Russia) was added to a final concentration of 0.2 mg/mL to enable colorimetric visualization of lacZα activity. Cultures were subsequently incubated at 26 °C with shaking (250 rpm) for an additional 16–18 h to allow color development.

2.4. SDS–PAGE Analysis of M^pro Expression

Cell lysates from E. coli DH5α carrying the recombinant pUC19–M^pro construct and from control cells containing the empty pUC19 vector were analyzed by SDS–PAGE. Prior to lysis, cell cultures were normalized by OD₆₀₀ to ensure comparable protein loading. Samples were mixed with Laemmli SDS sample buffer containing 2% SDS (Bio-Rad, Hercules, CA, USA), boiled for 5 min, and loaded onto a 12% polyacrylamide gel together with a PageRuler™ Plus Prestained Protein Ladder (Thermo Fisher Scientific) as a molecular weight marker. Electrophoresis was carried out using standard protocols. Following electrophoresis, proteins were visualized by staining with Coomassie Brilliant Blue (Bio-Rad).

2.5. Preparation of Plant Juices

Fresh pomegranates (Punica granatum) (local variety) and fresh guelder rose berries (Viburnum opulus) were obtained from a local market during their natural season (mid-November). Fresh rhubarb stalks (Rheum × hybridum.), red currants (Ribes rubrum), and black currants (Ribes nigrum) were harvested from a locally cultivated garden in mid-July, and black grapes (Vitis vinifera) were obtained from a local market in late October. With the exception of pomegranate and guelder rose, all plant materials were frozen immediately after harvesting or purchase in liquid nitrogen and stored at −80 °C until use. All manipulations involving the juices were performed under sterile conditions. Fresh plants (fruit/stalk) were homogenized to obtain raw juices using either manual or mechanical methods, depending on tissue characteristics. For pomegranate and guelder rose, approximately 25 g of pomegranate arils and 25 g of guelder rose berries were manually homogenized separately using a sterile mortar and pestle. While plants with higher fiber content or larger tissue volume were processed using a laboratory blender: a single green rhubarb stalk (97 g), approximately 50 g of black grape berries, and approximately 30 g and 70 g of red and black currants, respectively, were homogenized accordingly, with higher masses used for black currant due to differences in berry size and tissue density. The homogenates were clarified by centrifugation at 4200 rpm for 15 min at 4 °C, and the clear supernatants were collected. The pH of the supernatants was measured and adjusted to 4.0–4.5 with 10 N NaOH to reduce excessive acidity while preserving the potential contribution of their organic acids to the anti-M^pro activity. The juices were shortly stored at 4 °C until use.

2.6. Design and Functional Basis of the Colorimetric Screening Assay

The colorimetric screening assay was designed to function on a gain-of-signal basis. As mentioned above, in E. coli DH5α, functional β-galactosidase activity is obtained through α-complementation, as the plasmid-encoded LacZα fragment associates with the chromosomally encoded LacZω fragment to form an active enzyme [23]. Accordingly, the assay is based on the functional relationship between M^pro and the lacZα reporter gene. In the absence of M^pro, cells expressing lacZα alone produce a blue coloration in LB medium supplemented with X-gal and IPTG, due to β-galactosidase-mediated hydrolysis of X-gal yielding 5-bromo-4-chloro-indole, which oxidizes and dimerizes into a blue product. However, co-expression of M^pro leads to cleavage of its recognition site inserted within the lacZα fragment, disrupting β-galactosidase activity and preventing color formation. The addition of an inhibitory compound blocks M^pro activity, restoring β-galactosidase function and producing a gain of blue signal. The assay principle and expected color outcomes are illustrated in Figure 2.

2.7. Assay Setup and Inhibitor Treatment (Plant Juices)

Freshly prepared, slightly pH-adjusted juices were tested at a range of final concentrations (v/v). Pomegranate, guelder rose, and grape juices were tested at 2%, 5%, 10%, and 20%. For red and black currant juices, concentrations of 2%, 5%, 15%, and 30% were tested to account for differences in matrix composition and consistency. While rhubarb stalk juice was tested at higher concentrations of 11.25%, 22.5%, and 45%, due to its high intrinsic water content and lower acidity. All bacterial cultures were standardized to an identical OD₆₀₀ of 0.8 prior to treatment (as achieved by refreshing overnight cultures, see Section 2.3.). Juices were added at the time of induction simultaneously with IPTG and X-gal, and cultures were incubated at 26 °C for 16–18 h with shaking at 250 rpm to allow color development. To enable the use of high juice fractions (≥10% v/v) without diluting the cell density, OD-standardized cultures were centrifuged (2500 rpm, 2 min), supernatants discarded, and pellets immediately resuspended to 1.00 mL with a premix adjusted to yield a final 1× LB composition at the target juice fraction. Antibiotic, IPTG, and X-gal were added at the time of resuspension. All experimental groups were handled identically to maintain equal initial OD and consistent medium composition across treatments.

A defined set of internal controls was included to ensure the validity and interpretability of the colorimetric screening readout, accounting for responsiveness, baseline β-galactosidase activity, and potential sources of experimental variability. The detailed composition and purpose of control groups are summarized in

Table 1. A summary of internal controls used in the developed colorimetric reporter assay.

Control	Genetic Composition	LB Additives	Purpose
LB-only control	No bacteria, only LB medium	Antibiotic.	Serves as a contamination control and spectrophotometric blank.
Reporter-only control	Cells carrying only the modified reporter (lacZα with inserted Mpro cleavage site)	Antibiotic + IPTG + X-gal.	Expected to turn blue; confirms baseline β-galactosidase activity.
Vehicle control	Cells carrying the full construct (lacZα + Mpro)	Antibiotic + IPTG + X-gal + solvent (e.g., DMSO) or NaOH (10 N) used for pH adjustment.	Expected to remain white; rules out effects of solvent or pH adjustment on bacterial growth or reporter function.
Untreated control	Cells carrying the full construct (lacZα + Mpro)	Antibiotic + IPTG + X-gal.	Expected to remain white; confirms baseline Mpro activity in the absence of inhibition.
Tested-juice-only control (no X-gal)	Cells carrying the full construct (lacZα + Mpro)	Antibiotic + IPTG + plant juice (premixed with concentrated LB for ≥10% v/v). No X-gal.	Should show only the natural pigment of the tested juice; used to rule out interference of plant pigmentation with visual or spectrophotometric readout.
Reporter-only + Plant juice control	Cells carrying only the modified reporter (lacZα with inserted Mpro cleavage site)	Antibiotic + IPTG + X-gal + plant juice (premixed with concentrated LB for ≥10% v/v).	Expected to remain blue; assesses plant-derived effects on the reporter alone. Later considered redundant since both lacZα and Mpro share the same regulatory elements, but potentially useful for color reference when comparing juice coloration with blue signal.

Cell growth inhibition by the tested juices was evaluated by comparing OD₆₀₀ values measured immediately before and after incubation with the respective concentrations.

.

2.8. Quantification of Color Development

Following overnight incubation with the tested juices, bacterial cells were harvested by centrifugation at 4200 rpm for 15 min at 4 °C. The resulting pellets were resuspended in lysis buffer (8 M urea, 1.5 M Tris-HCl, pH 8.0), added at a ratio of 15 volumes of lysis buffer per gram of wet pellet, and incubated with gentle rotation at room temperature for 30 min. To ensure complete release of intracellular pigment, cell suspensions were subjected to ultrasonic disruption using a Q700 sonicator (Qsonica, Newtown, CT, USA) under the following conditions: 50% amplitude, 1 min on/2 min off cycles. The lysates were then centrifuged at 3600 rpm for 5 min to remove cell debris, and the clarified supernatants were analyzed spectrophotometrically. The intensity of the developed blue chromogen was quantified by measuring absorbance at 650 nm. A schematic representation of the experimental workflow is shown in Figure A3.

2.9. Statistical Analysis and Assay Quality Evaluation

All samples were tested in triplicates. Given the small sample size (n = 3), no assumptions of normal distribution were made; therefore, a non-parametric Mann–Whitney U test was applied to compare independent groups. Statistical significance was accepted at p < 0.05. Calculations were performed in Microsoft Excel. Data were normalized to the “tested-juice-only” control (no X-gal) (see Table 1).

Assay quality was evaluated using the Z-factor, calculated from replicate measurements of maximum inhibition samples (treated with pomegranate juice) and negative (untreated) controls based on absorbance at 650 nm, according to the method of Zhang et al. (1999) [24].

3. Results

3.1. Verification of the Assembled Genetic Components of the Assay

Following correction of all detected sequence errors in the assembled, codon-optimized M^pro gene by iterative site-directed mutagenesis, agarose gel electrophoresis confirmed the expected amplicon size of approximately 1163 bp, corresponding to the full assembled expression cassette containing the assembled codon-optimized M^pro gene flanked with the upstream lac promoter/operator region and the downstream T7 terminator (Figure 3A). Sanger sequencing verified the integrity of the open reading frame, revealing no frameshift or nonsense mutations.

Restriction digestion analysis confirmed the correctness of the final recombinant plasmid containing both the M^pro and modified lacZα genes. Sequencing further verified that insertion of the M^pro cleavage site preserved the lacZα reading frame. The complete construct comprising the lac promoter/operator-M^pro-T7 terminator cassette and the modified lacZα reporter was successfully assembled. A schematic plasmid map generated using Benchling software (Benchling Inc., San Francisco, CA, USA) is presented in Figure A2. The verified pUC19-M^pro construct was used in all subsequent functional assays.

3.2. Expression Validation of M^pro Protein in E. coli DH5α

Expression and functionality of the M^pro protein were primarily verified phenotypically (see Section 3.3.), and SDS–PAGE analysis was performed as an additional supportive confirmation of recombinant M^pro expression. Comparison of total cell lysates from E. coli DH5α carrying the recombinant pUC19–M^pro plasmid and the corresponding empty-vector control, showed an additional distinct protein band in the recombinant sample, absent in the control. This band migrated with an apparent molecular weight of approximately 35 kDa, close to the expected size of SARS-CoV-2 M^pro (33.8 kDa), consistent with successful expression under the applied induction conditions (Figure 3B). The presence of this band supports intracellular expression of M^pro and is consistent with the phenotypically observed loss of β-galactosidase activity resulting from proteolytically active M^pro.

3.3. Functional Validation of the Assay in Living Cells

The functionality of the colorimetric reporter assay was verified by comparing the phenotypes of E. coli DH5α cultures expressing both M^pro and the modified lacZα reporter with those expressing only the modified reporter. Following induction with IPTG in the presence of X-gal, cultures carrying only the reporter developed a vivid blue coloration, confirming intact β-galactosidase activity. In contrast, cultures co-expressing M^pro remained uncolored, indicating that the protease successfully cleaved its recognition site within lacZα, thereby disrupting β-galactosidase function (as observed in the visual readouts presented later). These results confirmed that the assay operated as designed and produced a distinct, visually detectable, and quantifiable phenotypic output suitable for inhibitor screening.

3.4. Proof-of-Concept Application of the Screening Assay Using Pomegranate and Guelder Rose Juices

For cultures incubated with pomegranate juice, absorbance at 650 nm, demonstrated a significant, concentration-dependent increase in blue coloration at all tested concentrations (p = 0.000) compared with untreated control, indicating inhibition of M^pro activity within living cells. The relative blue signal intensities (A₆₅₀ values normalized to the control containing juice without X-gal) are presented in Figure 4A, and visual readouts are shown in Figure 5.

Assessment of optical density (OD₆₀₀) before and after incubation with pomegranate juice showed a significant concentration-dependent reduction in growth at concentrations (5%, 10%, and 20%), compared to untreated control (Figure 4B). Despite this growth inhibition, blue coloration remained clearly detectable, indicating that the residual viable cell population retained sufficient β-galactosidase activity for the assay to generate a positive colorimetric readout.

In contrast, cultures incubated with guelder rose juice also exhibited strong and consistent growth inhibition across tested concentrations (significant reduction in OD₆₀₀ values); however, no increase in blue coloration was observed at any concentration. Because the cell population was substantially reduced and no measurable color development occurred, the juice was excluded from quantitative analysis.

These results confirm the ability of the developed E. coli-based assay to detect intracellular M^pro inhibition using plant preparations even in conditions of partial growth inhibition/mild toxicity.

3.5. Application of the Assay to Additional Edible Plant Juices

For further evaluation of assay’s applicability, additional edible plant juices (green rhubarb, black grape, red and black currant) were tested using the same experimental workflow and readout criteria described above. Cultures incubated with these juices showed differential outcomes across multiple concentrations (

Table 2. Qualitative screening outcomes for additional juices tested using the developed assay.

Plant Juice	Observed Outcome (Blue Coloration 1) at a Range of Tested Concentrations
	1st Concentration	2nd Concentration	3rd Concentration	4th Concentration
Green rhubarb (Rheum × hybridum)	11.25%	22.5%	45%
	−	+	+
Black grapes (Vitis vinifera)	2%	5%	10%	20%
	–	–	–	–
Red currant (Ribes rubrum)	2%	5%	15%	30%
	–	–	–	–
Black currant (Ribes nigrum)	2%	5%	15%	30%
	–	–	+	+

¹ Blue coloration refers to visually detectable and spectrophotometrically confirmed increases in A₆₅₀ values relative to untreated controls. (+) indicates observed blue coloration; (–) indicates no blue coloration.

).

For cultures incubated with rhubarb and black currant juices, absorbance at 650 nm, demonstrated a significant, concentration-dependent increase in blue coloration at their higher concentrations (p = 0.000) compared with untreated controls, indicating inhibition of M^pro activity within bacterial cells. The relative blue signal intensities (A₆₅₀ values normalized to the control containing juice without X-gal) are presented in Figure 6A. Additionally, comparison of OD₆₀₀ values before and after incubation with these two juices showed a significant concentration-dependent reduction in growth at higher concentrations, compared to untreated controls (Figure 6B). However, a clear colorimetric signal was retained under these conditions, indicating again that M^pro inhibition could be detected in the assay despite partial growth suppression.

In contrast, cultures incubated with grape and red currant juices showed significant reduction in OD₆₀₀ values, but no detectable signal was observed at any concentration. Accordingly, they were excluded from quantitative analysis.

Taken together, these findings further show the developed assay’s broader applicability, and support its use as a simple, initial-screening, prioritization tool for the search of potential inhibitory effects in diverse plant-derived samples.

3.6. Internal Controls Validation and Assay Reliability

Performance of the set of internal controls, included to ensure the assay’s reliability, is summarized in

Table 3. A summary of observed outcomes for all internal controls.

Control	Observed Outcome
LB-only control	No signal or growth, indicating a clean background.
Reporter-only control	Strong blue coloration, confirming β-galactosidase activity (positive baseline).
Vehicle control	No effect on color or growth, confirming solvent neutrality.
Untreated control	No coloration observed with X-gal, confirming Mpro activity.
Tested-juice-only control (no X-gal)	Only native plant pigment observed; no blue coloration (negative baseline).
Reporter-only + Plant juice control	Blue coloration mixed with natural pigment, serving as color reference for positive inhibition effects.

.

The Z-factor for the assay, calculated using the highest tested concentration of pomegranate juice and the negative (untreated) control, was 0.69, indicating a reliable separation between inhibited and uninhibited states and confirming the suitability of the assay for reproducible screening applications.

4. Discussion

This study demonstrates the successful development and functional validation of an E. coli-based colorimetric reporter assay for detecting inhibition of the SARS-CoV-2 M^pro directly in a bacterial cellular context by plant crude preparations (juices) and other potential inhibitors. The developed assay links M^pro functionality to a quantifiable visual phenotypic output, enabling assessment of protease inhibition by tested juices in a safe, simple, and cost-effective way. Such a tool offers a practical complement to existing computational, biochemical, and eukaryotic models, that fills the practical gap for early-stage functional prioritization of candidate M^pro inhibitors, particularly under constrained laboratory conditions.

The correctness and integrity of the assembled genetic components were first validated by sequencing and restriction analyses, confirming that the codon-optimized M^pro gene along with its regulatory elements and the modified lacZα reporter were correctly constructed and maintained in frame. This step was essential to ensure the functional linkage between both expressed proteins, and, thereby, between M^pro enzymatic activity and the resulting phenotypic output. Sequence errors introduced during initial synthesis and amplification were corrected by site-directed mutagenesis. Codon optimization was performed to minimize the presence of rare codons and reduce the risk of premature translational termination, while maintaining moderate expression levels suitable for functional inhibition screening and without altering the encoded amino acid sequence. Such strategies have been reported to support reliable heterologous viral protein production in bacterial systems, supporting the reliability of this design for intracellular bacterial expression of SARS-CoV-2 M^pro for the assay’s purpose [25].

Expression and functional activity of recombinant M^pro were primarily demonstrated phenotypically through the loss of blue coloration upon induction, which was noticed exclusively when M^pro was co-expressed with the modified lacZα reporter. This verified that M^pro reserved its catalytic activity in the bacterial cytoplasm, and specifically cleaved its specific site within the modified lacZα gene product. The retention of blue coloration in strains expressing the modified reporter alone confirmed that insertion of the M^pro cleavage site did not compromise β-galactosidase activity, while the disappearance of the signal upon M^pro expression directly reflected intracellular proteolytic activity.

SDS–PAGE analysis was also conducted as a supportive validation step revealing an additional protein band in recombinant DH5α lysates with an apparent molecular mass close to the theoretical size of M^pro (33.8 kDa). Detection of this band in DH5α lysates indicated that even a cloning host strain, which is not optimized for protein overexpression, can produce sufficient protease levels for measurable activity, further confirming the efficiency of the construct.

The obtained binary color change: Blue coloration in the absence of protease activity and its loss when M^pro is expressed provides a clear qualitative readout with additional semi-quantitative information suitable for comparative screening, derived from normalized absorbance measurements. Compared to in silico, biochemical, and mammalian cell-based assays, this assay offers an efficient middle ground for initial screening of potential inhibitors. More specifically, in silico screening, while cost-effective, can only predict or filter candidates without functional confirmation, whereas biochemical assays often require purified protein and specialized instrumentation and can have limited capacity to screen complex matrices. Mammalian cell-based systems, despite physiological relevance, remain costly, less accessible, and demand advanced biosafety infrastructure. The E. coli–based assay, within its bacterial context, simultaneously enables functional validation, rather than unverified predictions, while remaining experimentally simple, safe, broadly accessible, and applicable for complex matrices such as crude plant preparations/juices. Importantly, considering the assay’s design provides a semi-quantitative readout of M^pro inhibition, it enables rapid identification of samples that exceed a functional inhibition threshold, rather than precise determination of inhibitory potency or IC₅₀ values. Also, because intracellular M^pro expression levels cannot be reduced to the low enzyme concentrations used in cell-free biochemical assays, the assay is not intended to resolve fine differences in inhibitor potency or rank strong inhibitors quantitatively. Instead, the observed concentration-dependent color restoration reflects relative inhibitory strength within a biologically constrained bacterial context, sufficient for prioritization decisions at an early screening stage.

Proof-of-concept application of the assay was first shown using pomegranate (Punica granatum) and guelder rose (Viburnum opulus) juices. For pomegranate juice, a significant increase in blue coloration at higher concentrations indicated successful inhibition of M^pro activity within bacterial cells. Pomegranate was selected due to its reported strong antiviral, antioxidant, and polyphenol-rich profile, as well as several studies demonstrating inhibition of SARS-CoV-2 M^pro by its ellagitannins and punicalagins [26,27]. In contrast, guelder rose juice, selected for its high polyphenolic content and traditional medicinal use (particularly in Eastern Europe and Russia, where it is widely applied against respiratory conditions) [28], induced strong growth suppression without detectable color restoration, suggesting a lower or absent content of active M^pro-targeting phytochemicals at the tested concentrations. Importantly, this differential outcome demonstrates the assay’s ability to distinguish plant-derived crude preparations with stronger inhibitory potentials from those with absent or lower potentials, even when both affect bacterial growth, as only juices with functional M^pro-inhibitory activity result in restoration of the blue phenotype, while unrelated phytochemical profiles do not generate false-positive signals. At the same time, reduced cell viability does not result in false-negative inhibition outcomes.

To further examine the general applicability of the assay beyond a single illustrative example, additional plant-derived preparations, including juices of rhubarb (Rheum × hybridum), black grape (Vitis vinifera), and red and black currant (Ribes rubrum and Ribes nigrum, respectively), were subsequently screened using the same workflow. These samples exhibited distinct response patterns, with rhubarb and black currant juices producing detectable M^pro inhibition at higher concentrations, while grape and red currant juices showed no inhibitory signal at all tested concentrations. These additional tests, although do not constitute a comprehensive natural product library, demonstrate that the assay can be applied to diverse plant matrices with varying composition, supporting its applicability and suitability for rapid functional prioritization of crude samples.

Juices, rather than purified extracts, were used because they offered the simplest test material without the need for extraction, further demonstrating the assay’s simplicity. Importantly, juices also allowed for the assessment of potential synergistic effects among multiple phytochemicals, rather than isolated compounds. From a nutritional and practical perspective, evaluating dietary juices was relevant, as the focus in urgent pandemic situations should prioritize safe, accessible, and naturally derived antiviral sources [19]. The tested concentration range was selected to represent moderate exposure levels; sufficient to deliver active compounds while avoiding excessive sugar or polyphenol content that could affect bacterial physiology or optical readings. Strong pH adjustment was avoided to examine potential effects of naturally occurring acids (while still relatively preserving bacterial tolerance to acidity), such as ascorbic, caffeic, chlorogenic, ellagic, ferulic, gallic, protocatechuic, and ursolic acids [29,30], many of which have been reported to inhibit SARS-CoV-2 M^pro [26,31].

The comprehensive set of internal controls verified assay readout specificity and robustness by ensuring the validity and interpretability of the visual outcome. These included treated reporter-only controls to assess assay responsiveness, reporter-only and untreated controls to confirm baseline activity and reporter integrity, and juice-only, vehicle, and contamination controls to rule out non-specific color or growth effects. Together, these controls establish clear interpretive baselines for distinguishing true inhibitory activity from unrelated visual or metabolic changes.

In addition to the set of internal controls, assay performance was further assessed using the Z-factor, a widely accepted statistical parameter for evaluating screening assay quality and signal separation. In the context of this proof-of-concept study, Z-factor calculation was performed using a high-contrast inhibitory condition (the highest tested concentration of pomegranate juice) and the untreated control. Importantly, this approach was applied solely for assay quality evaluation and does not imply the use of pomegranate juice as a standardized positive control. The obtained Z-factor value demonstrates that the assay is robust and suitable for reproducible first-pass functional screening.

During assay optimization, several technical challenges were encountered and systematically resolved. Initially, a dual-plasmid configuration was tested in which E. coli cells were co-transformed with one plasmid encoding SARS-CoV-2 M^pro under the lac promoter/operator and another containing the lacZα reporter gene interrupted by the engineered M^pro cleavage site. However, unbalanced gene expression, unequal plasmid copy-numbers, and potential incompatibility of replication origins introduced substantial variability in color development, complicating interpretation. Transition to a single-plasmid construct unified gene regulation and eliminated variability. Additionally, early assays were performed on solid LB-agar supplemented with IPTG, X-gal, and ampicillin, however, yielded uneven color distribution and inconsistent quantification, which were resolved by adopting liquid cultures that ensured uniform distribution and allowed quantitative spectrophotometric assessment. Induction at higher cell densities (OD₆₀₀ = 0.6–0.9) enhanced assay sensitivity toward weaker inhibitory effects, whereas lower induction ODs produced less distinct color differentiation. IPTG concentrations of 0.1–0.5 mM and X-gal concentrations below 0.2 mg/mL were evaluated; the selected combination of 0.2 mM IPTG and 0.2 mg/mL X-gal offered optimal balance between induction efficiency and visual contrast. Temperature optimization revealed that incubation at 26 °C provided the most consistent color development with minimal growth inhibition. Importantly, the timing of colorimetric readout was also empirically optimized to ensure reliable differentiation between inhibited and uninhibited states. More specifically, the readout window bounds were determined using two of the internal controls: The reporter-only control, and untreated control (see Table 1). Cultures of the former showed a progressive blue color development following induction, reaching a stable plateau at approximately 16 h, which was selected as the lower bound of readout window to avoid false negativities. While cultures of the latter remained white until approximately 18–19 h, after which gradual background color development was observed, so the upper bound of readout window was selected at 18 h to avoid false positivity. Thus, partial inhibition of M^pro remains detectable within this controlled temporal window, ensuring consistent interpretation across samples with different inhibitory strengths. Additionally, several buffer formulations were tested for efficient cell lysis without altering the chromogenic pigment. The final buffer composition achieved reproducible solubilization and color stability. These troubleshooting steps collectively enhanced assay reproducibility and demonstrated its adaptability to complex sample types, ensuring optimal performance.

As discussed above, the developed assay presents practical advantages as an early-stage screening tool. It assesses inhibition in a living bacterial system, thereby accounting for protease expression and functional stability under conditions relevant to E. coli, and enabling validation of actual protease binding. Its gain-of-signal design minimizes false negatives, since color restoration reflects inhibition rather than signal loss, which in other assays could result from toxicity or metabolic disruption. The assay is rapid, experimentally simple, and compatible with standard laboratory equipment, avoiding the need for advanced instrumentation or costly eukaryotic models. It is also biosafe and noninfectious, as it does not involve live virus handling, and allows straightforward visual interpretation supported by semi-quantitative measurements. The assay also enables built-in monitoring of bacterial cytotoxicity through bacterial growth observations. Importantly, it has been proven capable of testing complex plant preparations, not only purified compounds, supporting its use in preliminary evaluation of synergistic and cumulative effects of dietary plants as potential preventive or supportive measures against viral infections.

Despite these strengths, several limitations should be acknowledged. The E. coli outer membrane restricts permeability to large molecules, limiting the assay’s ability to assess macromolecular or poorly permeable inhibitors such as drugs based on nirmatrelvir or ensitrelvir. As a result, these drugs cannot be used as positive controls for M^pro inhibition in the proposed screening assay. The bacterial cytoplasmic environment also differs substantially from that of mammalian host cells, protein folding, post-translational regulation, and tested compounds behavior may not fully reflect human cellular conditions. Furthermore, highly colored or viscous extracts could interfere with optical readings at extreme concentrations, although internal normalization largely compensates for this. Therefore, the assay is not intended to replace high-resolution biochemical or antiviral efficacy assays but rather to serve as a preliminary, prioritization-stage screening tool within a broader M^pro inhibitor discovery pipeline. Candidate hits from this assay are expected to undergo further validation using established biochemical and eukaryotic cell-based approaches. A concise comparison of the developed assay with existing approaches is provided in

Table 4. Comparison of the key features, advantages and limitations of the developed assay with existing approaches.

Approach	Key Features	Advantages	Limitations
In silico (computational docking, molecular dynamics)	Virtual prediction of binding affinities based on Mpro structure	Fast, inexpensive, suitable for large-scale screening	Lacks biological context; no data on cellular uptake, solubility, or toxicity
E. coli-based colorimetric reporter assay (present study)	Mpro activity linked to β-galactosidase reporter within living bacterial cells	Simple visual and semi-quantitative readout, biosafe, cost-effective, compatible with complex natural matrices, enables first-pass functional prioritization under standard laboratory conditions, adaptable for large-scale screening and for other proteases/mutants	Limited by bacterial permeability and potential interference from highly colored extracts, intracellular environment differs from mammalian host cells
Biochemical (cell-free enzyme assays, fluorescence/luminescence readouts)	Recombinant purified Mpro tested with substrates in vitro	Quantitative, high specificity, allows kinetic analysis and IC50 determination	Requires protein purification and specialized equipment, ignores cellular uptake, limited compatibility with complex mixtures
Mammalian cell-based models	Mpro expression or viral replication assays in human cell lines	Physiologically relevant, integrates host metabolism	Expensive, time-consuming, requires high-biosafety facilities and expertise, potential safety risks

, summarizing their main features, advantages, and limitations.

In a broader perspective, beyond SARS-CoV-2 M^pro, the presented design can be readily adapted to other viral or bacterial proteases/mutant variants by substituting the corresponding protease-specific cleavage motifs within the reporter sequence without altering the core assay design. More specifically, for emerging M^pro mutant variants, such adaptation would primarily involve replacement of the cleavage site sequence to match variant-specific substrate preferences, while maintaining identical expression and readout conditions. Adaptation to other viral or bacterial proteases would similarly require insertion of validated cleavage motifs and, where necessary, adjustment of promoter strength or induction parameters to account for differences in protease expression levels or activity. It can also be optimized for high-throughput applications, enabling faster primary screening. The low biosafety requirements, minimal instrumentation needs, and visual readout make it particularly suitable for preliminary research or educational settings where advanced cell culture facilities are unavailable. Future work may also include the use of membrane-permeabilizing controls to expand the range of testable compounds.

5. Conclusions

Overall, the results confirm that the developed assay successfully couples protease activity to a visible phenotypic change, enabling the reliable functional identification of protease inhibition in a biosafe and accessible bacterial system. While not intended for the precise quantitative determination of inhibitory potency, the assay represents a practical and biosafe tool for preliminary evaluation of natural M^pro inhibitors and can be adapted for broader antiviral screening purposes. The selective permeability of the E. coli membrane determines the range of compounds that can be effectively tested. Consequently, the assay, in its current format, is particularly well-suited for screening small molecules capable of penetrating the bacterial cell.