Differential specificity of SARS‐CoV‐2 main protease variants on peptide versus protein‐based substrates

The SARS‐CoV‐2 main protease (Mpro) holds significant importance as a biological target in combating coronaviruses due to its importance in virus replication. Considering the emergence of novel SARS‐CoV‐2 variants and the mutations observed in the Mpro sequence, we hypothesized that these mutations may have a potential impact on the protease's specificity. To test this, we expressed Mpro corresponding to the original strain and variants Beta1, Beta2, and Omicron and analyzed their activity on protein‐based and peptide substrates. Although we observed differential activity on the protein‐based substrate, there was very little difference when analyzed on the peptide substrate. We conclude that mutations on the Mpro sequence, despite having a minor effect on a peptide substrate cleavage, did not change the catalytic site environment enough to build resistance to inhibition. Therefore, we propose that inhibitors initially designed for the Mpro of the original strain will be effective in all the variants. Thus, Mpro is likely to continue to be a target of therapeutic interest as mutations in its sequence are rare and, as we show here, have a minor effect on the protease's recognition of peptide‐based molecules.


Introduction
In 2019, a new and highly infective coronavirus SARS-CoV-2 emerged, which made the World Health Organization declare a world pandemic on 11 March 2020 [1].The human coronaviruses (HCoVs) are positive-stranded RNA viruses that are the leading causes of upper respiratory tract illness in humans.The HCoV genome encodes several nonstructural proteins that are synthesized as two large polyproteins pp1a and pp1ab, which are cleaved by the virus proteases named papain-like protease and the main protease (M pro ; EC3.4.22.69)[2].The protease's primary role is to process the polyproteins pp1a and pp1ab, which mediate the functions required for viral replication and transcription [1,[3][4][5], and possibly also cleavage of host proteins [6,7].The enzyme has a very stringent cleavage specificity, recognizing preferentially the sequence Leu-Gln↓ (Ser, Ala, Gly) [3,8].Thus, due to its vital role and cleavage specificity, M pro has become a key target for developing antiviral drugs [9,10].Paxlovid was the first approved drug for the treatment of COVID-19 that targets M pro with high affinity [11].The drug is composed of the M pro inhibitor Nirmatrelvir [12] and Ritonavir that inhibits cytochrome P4503A4 [13].
The SARS-CoV-2 virus has undergone various mutations since its first appearance, generating several variants of concern (VOCs): Alpha, Beta, Gamma, Delta, and Omicron [14].The variants have generally shown a higher transmission efficiency and disease severity, which is related to mutations in the virus genome, more specifically in the spike protein [14].The genomic sequences of the variants can be found in the Global Initiative on Sharing All Influenza website (https://gisaid.org).Although the majority of mutations are found in the spike protein, some mutations have also occurred in the M pro sequence [15].For this study, we specifically focused on two of the most widespread VOCs, Beta and Omicron.We sought to understand whether these mutations affect catalysis or specificity of M pro using both synthetic peptide and protein-based substrates.

SARS-CoV-2 M pro construct
The M pro mutations were performed by site-directed mutagenesis of the original strain (OS) M pro construct (Fig. 1A).The beta mutations K90R and A193V are present at the domain I interface and, in a loop connecting domains II and III, respectively.The omicron mutation is present at domain II of the protein (Fig. 1B).

Kosmotropic salts lead to an increase in the original strain M pro catalytic activity
As M pro is an obligate dimer, we assumed that dimerization is concentration-dependent, obtaining satisfactory catalytic activity is necessary to increase the enzyme concentration.Kosmotropic salts have been employed routinely to investigate the dimeric status of enzymes in vitro, including other viral proteases and caspases [17,18].To improve the enzyme's catalytic activity without having to work with high concentrations, the OS M pro was tested against a series of kosmotropic salts, which are ions that have a stabilizing effect on proteins [19], and can lead to stabilization of dimeric conformations.The results showed that some kosmotropic salts, such as sodium citrate, were able to substantially enhance M pro activity without the need for high concentrations of the enzyme in the assays.
The catalytic activity of M pro is significantly increased in the presence of kosmotropes (Fig. 2).The most effective kosmotropic salt was sodium citrate, which is a known surrogate dimerization reagent [20,21].The highest activity increase was observed when using 1.4 M of sodium citrate, which led to an increase in activity of 36.5-fold for the OS M pro (Fig. 2B).Although, as the presence of sodium citrate can decrease the solubility of the inhibitors, a lower concentration of 0.7 M was chosen to use in the subsequent assays.Therefore, assays were performed in 20 mM PIPES pH 7.2, 100 mM NaCl and 1 mM EDTA buffer supplemented with 0.7 M citrate (assay buffer).

Mechanism of enhancement of activity by kosmotropes
To better understand the effects of sodium citrate on the catalytic activity of OS M pro , the protease was assayed against the peptide substrate QS1 [22] using two concentrations of sodium citrate.The kinetic constants obtained for the OS M pro using 1.0 and 0.7 M Fig. 1.Graphical representation and structure of SARS-CoV-2 M pro .(A) SARS-CoV-2 M pro construct with the red arrow indicating the autocleavage during expression of the protein.The table displays the M pro variants with the respective mutations on the protein sequence.(B) M pro dimer structure, where the catalytic dyad C145 and H41 are displayed at protomer A (red), and the protein's mutations are highlighted in protomer B (green).The protein structure was generated using UCSF CHIMERA software [16].PDB: 6WTM.
of sodium citrate revealed changes in both K M and k cat .The K M values decreased by 3.6-fold as the concentration of sodium citrate increased, suggesting that the protein binds to the substrate with higher affinity.Similarly, the k cat increases by 45-fold with the increase of sodium citrate.The data are consistent with the notion that sodium citrate enhances both the binding and catalysis, suggesting that the presence of sodium citrate leads to a better stabilization of the active conformation of the protein.

M pro variants displayed similar catalytic efficiency on a peptide substrate
The M pro variants were assayed against the peptide substrate QS1 to evaluate the effect of the mutations on the catalytic efficiency of the protease.The kinetic constants obtained for the variants showed that the mutations at the M pro sequence led to a minimally improved catalytic efficiency for the QS1 substrate, with the Beta2 variant exhibiting at most a 1.5-fold increase in catalysis when compared to the OS M pro (Fig. 3).We conclude that M pro variants exhibit only minor differences in catalysis on the peptide substrate.

M pro variants showed significantly increased catalytic efficiency on a protein substrate when compared to the original strain
To determine the influence of the M pro mutations on a natural protein substrate, we constructed a protein that closely simulates the natural polyprotein environment, namely the catalytic mutant M pro .This construct contains the catalytic C145 substituted by an alanine residue.The C145A M pro has the same construct as the variants, with a sequence that mimics the natural substrate of M pro .The assay is based on the fact that active M pro will hydrolyze the catalytic  mutant M pro at its recognition sequence SALVQ↓S, releasing the GST domain (Fig. 1A) which we visualized using SDS/PAGE (Fig. 4A).The SDS/PAGE gel was scanned, and through a densitometric analysis of the bands, the relative band intensities were plotted against the active protease concentration for the determination of the amount of enzyme required to cleave 50% of the substrate during the incubation period (EC50; Fig. 4B).
The results obtained for the protein-based substrate were considerably different than those for the peptide substrate (Fig. 5).The Beta2 variant had a higher catalytic efficiency than the Omicron variant with a substantial sevenfold increase in cleavage rate against the protein-based substrate when compared to the OS M pro .

Inhibition of the M pro variants is comparable with the original strain M pro
We investigated the affinity of the M pro variants with Nirmatrelvir (Neq1183) [12].Binding was so tight that we were not able to observe equilibrium, and therefore, no K i determination was possible.This is consistent with the covalent nature of the inhibitor [23].Indeed, as the inhibitor had a similar behavior for all the M pro variants and the enzyme concentrations were all equal to 25 nM, it was considered that the K i for Neq1183 against the M pro variants is below 25 nM, yielding a pK i higher than 8.0, which is consistent with a previous publication [24].All variants exhibited identical tight binding behavior with the inhibitor; therefore, we conclude that the inhibitor cannot distinguish between the variants.

Discussion
The objective of this study was to investigate the effects of different mutations in the SARS-CoV-2 sequence of the protease M pro on the catalytic activity and specificity of this protease.Initial test with a fluorogenic peptide substrate (QS1) revealed that activity was substantially enhanced in the presence of kosmotropic salts.This is consistent with the need of M pro to act as an obligate dimer [20,21].In our study, the kosmotrope sodium citrate was able to improve the catalytic activity (k cat /K M ) of the protease by 160fold when compared to the M pro activity in its conventional assay buffer (Table 1).This enhancement in activity resulted from an increase in both binding affinity (K M ) and turnover rate (k cat ).In optimal buffer, there were no substantial changes in the kinetic constants of the protease variants toward the peptide substrate QS1, suggesting that the mutations did not change the active site space enough to affect the enzymes' function [7].In contrast, the protease variants were noticeably different when analyzed using a protein substrate.Thus, the mutations appear to have more impact on the overall structure of the protease, focusing on protein-protein interactions with its natural substrates, the nsps.
Since the activity of the variants on the peptide substrate was very similar, we hypothesized that peptidebased molecules would react similarly with all variants.This hypothesis was validated by testing with the inhibitor Neq1183.These results suggest that the mutations on the OS M pro sequence, despite having a minor effect on a peptide substrate cleavage, did not change the catalytic site environment enough to build resistance to inhibition.Therefore, we propose that inhibitors initially designed for the M pro of the original strain will be effective in all the variants.Hence, M pro is likely to continue to be a target of therapeutic interest as mutations in its sequence are rare and, as we show here, have a minor effect on the protease's recognition of peptide-based molecules.

Synthetic chemistry
Neq1183 was synthesized as per literature procedure [25] with minor modifications along route steps.The synthetic Scheme 1 and stepwise synthesis procedures are as follows.
The methyl ester 3 (0.5 mmol) in 1 : 1 mixture of THF (3 mL) and aqueous 2 M LiOH (3 mL) was stirred at room temperature for 2 h.Then, the suspension was acidified with 2 M HCl to pH = 1 and extracted with ethyl acetate (3 × 25 mL).The combined organic phase was dried over sodium sulfate, filtered, and evaporated, and the crude product was used for the next step without purification.
Triethylamine (3 equiv.) was added under argon to a solution of R-N-boc-3-methyl piperazine (1 equiv.) in dry dichloromethane, and the reaction mixture was cooled down to À78 °C.Trifluoroacetic anhydride (1 equiv.) was added dropwise, and the solution was stirred at À78 °C for 3 h.The reaction was quenched with a saturated aqueous solution of NaHCO 3 at À78 °C and was allowed to warm to room temperature.The aqueous phase was extracted with dichloromethane (3 × 25 mL), and the combined organic phase was washed with water (25 mL) dried over sodium sulfate, filtered, and concentrated under reduced pressure to obtain the product.
The resulting mixture was allowed to reach room temperature and was stirred at room temperature for 16 h.Then, this mixture was diluted with ethyl acetate (30 mL) and washed with water and brine.The organic phase was washed with 1 M HCl, brine, dried over sodium sulfate, filtered, and evaporated to afford the product.
The amide 7 (1 equiv.) was dissolved in anhydrous dichloromethane and stirred at room temperature under argon.Burgess reagent (3.5 equiv.) was added in 50 mg portions over 2 h.The mixture was stirred for an additional 15 min, and the mixture was applied directly to silica gel for purification.

Mutagenesis of SARS-CoV-2 M pro variants
The SARS-CoV-2 M pro plasmid was designed based on Zhang et al. [1] and inserted in a pGEX-6p-1 vector for expression in Escherichia coli.Mutations C145A, K90R (Beta1, Beta2), P132H (Omicron), and A193V (Beta2) were constructed by overlap PCR mutagenesis.The two internal, partially complementary, mutagenic primers were (forward, primer-1, 3, 5, and 7) and (reverse, primer-2, 4, 6, and 8; Table S1).The two flanking primers were (forward, primer-9) and (reverse, primer-10; Table S1).PCR was carried out on a GeneAmp PCR System 2400 (Perkin-Elmer, Waltham, MA, USA) with Phusion polymerase (Thermo Scientific, Waltham, MA, USA).The first round PCR generated two fragments using primer-9/primers-1,3,5,7 and primer-10/primers-2,4,6,8. The PCR products were purified by agarose gel, and the bands were excised and further purified using QIAquick PCR purification kit (Qiagen, Hilden, Germany).The purified products were combined and used as the template for the second round PCR with the two flanking primers.The PCR product was purified and restricted with BamHI/XhoI, purified again, and ligated to pGEX-6P-1 restricted with BamHI/XhoI.Competent cells of E. coli strain DH5α were transformed with the ligated vector.The full sequence of the insert was checked by DNA sequencing (Eton Biosciences, San Diego, CA, USA) to verify the success of the mutagenesis and the absence of any other mutations.The expression plasmid was then subcloned into E. coli strain BL21(DE3) and expressed and purified as the original strain.

SARS-CoV-2 M pro expression and purification
The construct has a GST tag at the N terminus that is connected to a linker with the recognition sequence that mimics the viral environment of the nonstructural proteins (nsps) for auto-cleavage during expression [8].A C-terminal histidine tag was added to aid in purification.
The proteins were expressed using E. coli BL21 (DE3).Cultures were grown in 2× YT media containing 100 μgÁmL À1 ampicillin (Sigma-Aldrich, Darmstadt, Germany).When cells reached an OD 600 of 0.6, protein expression was induced with 0.5 mM of isopropyl-D-thiogalactoside (Sigma-Aldrich) at 37 °C, 100 g for 5 h.The cells were collected by centrifugation for 15 min at 9954 g and 4 °C.The obtained pellets were resuspended using lysis buffer A (20 mM Tris, 150 mM NaCl pH 7.8) in 5 mLÁg À1 of pellet, and sonicated for 5 min.The lysate was centrifuged at 4 °C, 20 000 g for 1 h, and the obtained supernatant was added to a Ni 2+ -chelating Sepharose resin (GE Healthcare Life Sciences, Chicago, IL, USA) on a chromatography column.Washed with 50 mL of buffer A and 10 mL of buffer B (20 mM Tris, 500 mM NaCl pH 7.8).The M pro -His was obtained by imidazole elution using different percentages of buffer C (20 mM Tris pH 7.8, 150 mM NaCl, 500 mM Imidazole).The fractions were analyzed by SDS/PAGE.The protein samples were stored at À80 °C.

Enzyme kinetics
All assays were performed in a CLARIOstar microplate reader (BMG Labtech, Ortenberg, Germany) in a 96-well white plate (Corning New York, NY, USA), by following the release of 7-amino-4-carbamoylmethylcoumarin (ACC) fluorophore from the substrate Ac-Abu-Tle-Leu-Gln-ACC (QS1) [22] in an excitation and emission wavelength of 355 and 460 nm, respectively.The substrate hydrolysis rates (RFUÁs À1 ) were determined and analyzed according to the methods below, which describe equilibrium and rate constants for the inhibition of M pro .

Effect of kosmotropic salts on the M pro activity
The OS M pro activity was evaluated using different kosmotropic salts: sodium citrate (NaCitrate), ammonium citrate (NH 4 Citrate), sodium sulfate (NaSO 4 ), ammonium sulfate ((NH 4 ) 2 SO 4 ), and sodium chloride (NaCl).The assay was performed in buffer 20 mM PIPES, 100 mM NaCl, 1 mM EDTA, and 4 mM DTT pH 7.2.The enzyme (0.2 μM) samples were incubated into the buffer with different concentrations of kosmotropes in a range from 0 to 1.4 M. The substrate final concentration was 20 μM.The enzymatic assay was performed at 37 °C.The fluorescence emitted by the release of the ACC fluorophore was measured for approximately 15 min.Only the linear portion of each progress curve was used to determine the substrate hydrolysis rate in RFUÁs À1 .
Determination of the Michaelis-Menten constant (K M ) for OS SARS-CoV-2 M pro and variants The OS M pro and variants were prepared in the assay buffer (20 mM PIPES pH 7.2, 100 mM NaCl, 1 mM EDTA, 0.7 M sodium citrate, 4 mM DTT).The final concentration of M pro (25 nM) was titrated with serial dilutions of substrate.The rate of hydrolysis was monitored for approximately 15 min at 37 °C.Data were analyzed using (GRAPHPAD PRISM 8, San Diego, CA, USA).

Analysis of inhibition kinetics
The previously described M pro inhibitor Neq1183 (Nirmatrelvir [12]) was titrated against the enzyme in assay buffer.The inhibitor was preincubated with M pro (25 nM) for 30 min at 37 °C.The reaction was started with the addition of the QS1 substrate with a final concentration of 20 μM.

Cleavage of protein-based substrate
Cleavage of the protein substrate (catalytic mutant M pro ) was analyzed by observing the effect of a range of M pro variant concentrations, by SDS/PAGE.The catalytic mutant M pro was added to the protein solutions with a fixed concentration of 2.0 μM.The solutions were incubated for 1 h at 37 °C.After 1 h, the reaction was stopped by the addition of 30% TCA.Samples were boiled and added to an SDS/PAGE gel for the separation of the reaction products.The catalytic efficiency was obtained using Eqn (1) [21] by determining the M pro concentration at which 50% of substrate cleavage was observed (EC50).The gels were scanned at 700 nm using a LI-COR Odyssey CLx.
Equation 1: k cat /K M determination for natural substrate cleavage.EC50 is the active M pro concentration required to cleave half the substrate in the incubation period t.

Fig. 2 .
Fig. 2. Effect of kosmotropic salts on the activity of OS M pro .(A) Rate of substrate QS1 cleavage over different concentrations of kosmotropes.(B) Fold increase in activity of OS M pro at a salt concentration of 1.5 M. The fold increase was calculated by comparing the activity of the protease in a 20 mM PIPES pH 7.2 buffer with 100 mM of NaCl, 1 mM EDTA, and 4 mM DTT.

Fig. 3 .
Fig.3.Kinetic constant comparison of M pro variants on a peptide substrate.Kinetic constants were obtained for the OS M pro and variants in assay buffer against the substrate QS1 as described in Materials and methods.Values show the mean and standard deviation of three technical replicates.

Fig. 4 .
Fig. 4. Protein substrate cleavage by the omicron variant M pro .(A) Densitometric analysis of the extent of cleavage of 2 μM protein substrate (catalytic mutant M pro ) by a twofold dilution series of the omicron variant in assay buffer for 1 h at 37 °C.This concentration was chosen for optimal visualization in the Coomassie-stained gel.(B) Relationship of relative band intensities against the omicron variant concentration reveals the EC50.Catalytic efficiency was calculated through Eqn (1) (see Materials and methods).

Fig. 5 .
Fig. 5. Comparison of rate of cleavage of a protein-based substrate by M pro variants.Two micromolar of the recombinant protein-based substrate was cleaved in vitro with a dilution series of the M pro variants and analyzed by SDS/PAGE (as in Fig.4).The k cat /K M values represent the mean and range of two to three independent measurements.

Table 1 .
Effects of sodium citrate on kinetic constants.Kinetic constants were obtained for OS M pro using PIPES pH 7.2 buffer with different concentrations of sodium citrate assayed with peptide substrate QS1.Different enzyme concentrations were used for differing sodium citrate concentrations as described in the table.The table shows the mean and standard deviation of three technical replicates.