Structure-Based Design of Potent Peptidomimetic Inhibitors Covalently Targeting SARS-CoV-2 Papain-like Protease

The papain-like protease (PLpro) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) plays a critical role in the proteolytic processing of viral polyproteins and the dysregulation of the host immune response, providing a promising therapeutic target. Here, we report the structure-guide design of novel peptidomimetic inhibitors covalently targeting SARS-CoV-2 PLpro. The resulting inhibitors demonstrate submicromolar potency in the enzymatic assay (IC50 = 0.23 μM) and significant inhibition of SARS-CoV-2 PLpro in the HEK293T cells using a cell-based protease assay (EC50 = 3.61 μM). Moreover, an X-ray crystal structure of SARS-CoV-2 PLpro in complex with compound 2 confirms the covalent binding of the inhibitor to the catalytic residue cysteine 111 (C111) and emphasizes the importance of interactions with tyrosine 268 (Y268). Together, our findings reveal a new scaffold of SARS-CoV-2 PLpro inhibitors and provide an attractive starting point for further optimization.


Introduction
The ongoing pandemic of COVID-19 (Coronavirus Disease-2019) has inflicted an unprecedented number of infections and fatalities worldwide [1]. This highly contagious disease is caused by a novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) which could result in severe respiratory diseases as well as unclear sequelae [2]. Although current vaccines and treatments have demonstrated some effectiveness in preventing COVID-19 or reducing its severity [3,4], a multitude of noteworthy Omicron variants has still emerged in many countries [5][6][7], necessitating the continued development of therapeutic strategies and drug candidates to combat SARS-CoV-2 [8].
The optimal antiviral drugs prefer to target the viral proteins essential for the SARS-CoV-2 life cycle. The papain-like protease (PL pro ) and chymotrypsin-like protease (called 3CL pro , also referred to as main protease) are two essential cysteine proteases encoded by the SARS-CoV-2 genome. Both of them are required to cleave two viral polyproteins (pp1a and pp1ab) to generate 16 mature nonstructural proteins for genome transcription and replication [9,10]. In addition, SARS-CoV-2 PL pro recognizes and cleaves the C-terminal LxGG sequence of ubiquitin (Ub) and ubiquitin-like proteins (UbLs) such as interferonstimulated gene product 15 (ISG15) to remove the Ub and UbL modifications from host proteins, respectively [11,12]. The deubiquitinase (DUB) activity of SARS-CoV-2 PL pro is Int. J. Mol. Sci. 2023, 24, 8633 2 of 15 thought to induce dysregulation in the host immune response against viral infection and replication [13][14][15]. Therefore, targeting SARS-CoV-2 PL pro is an attractive strategy for both suppressing the virus replication and preventing the disruption of the host antiviral immunity [16].
Over the past decade, rational covalent drug design has attracted extensive attention from pharmaceutical chemists and led to the discovery of numerous approved covalent drugs [17,18]. Both SARS-CoV-2 3CL pro and PL pro have a catalytic cysteine that can be attacked by electrophiles, serving as ideal targets for covalent inhibitor design [19]. At present, two covalent inhibitors of SARS-CoV-2 3CL pro , nirmatrelvir and simnotrelvir, have been approved for the treatment of COVID-19 [3]. In contrast, there are few covalent inhibitors of SARS-CoV-2 PL pro reported, which only include two peptidomimetic inhibitors (VIR250 and VIR251) [20] and GRL0617 analogs [21]. Although the GRL0617 analogs were much more potent relative to peptidomimetic inhibitors, their crystal structures in complex with SARS-CoV-2 PL pro are not available yet. Accordingly, the peptidomimetic inhibitors were selected as the starting point for further optimization with the objective of identifying novel covalent inhibitors targeting SARS-CoV-2 PL pro .
In this work, we have identified a new class of covalent peptidomimetic inhibitors of SARS-CoV-2 PL pro using a structure-guide design strategy. The most potent inhibitor displays a submicromolar potency in an enzymatic assay. Moreover, a crystal structure of SARS-CoV-2 PL pro in complex with compound 2 was determined to understand the explicit binding mode of these congeneric inhibitors. In addition, the cell-based inhibitory activity of all designed compounds against SARS-CoV-2 PL pro was tested by the PL-FlipGFP assay, a cell-based protease inhibition assay. Overall, our study comprehensively explored the in vitro inhibitory activities of novel covalent peptidomimetic inhibitors of SARS-CoV-2 PL pro through biochemical and biophysical approaches.

A Fusion of VIR251 and Compound 1 Led to the Design of Compound 2
VIR250 and VIR251, as the first reported peptidomimetic inhibitors, covalently modify the catalytic cysteine (C111) and occupy the substrate binding pocket of SARS-CoV-2 PL pro ( Figure 1A). However, these two compounds show micromolar potency with a half-maximal inhibitory concentration (IC 50 ) value of~50 µM [20]. After a thorough investigation of the crystal structures of SARS-CoV-2 PL pro in complex with VIR250 and VIR251, we found an open conformation for the side chain of tyrosine 268 (Y268) which has no interactions with the aromatic ring at the P4 position of these peptidomimetic inhibitors. In contrast, a number of crystal structures of SARS-CoV-2 PL pro in complex with potent noncovalent inhibitors reveal a closed side-chain conformation of Y268 which forms distinct π-π interactions with the inhibitors in an edge-to-face manner ( Figure 1B) [22][23][24][25]. Therefore, it's reasonable to speculate that the replacement of the P4 group of VIR251 with the portion of noncovalent inhibitors could achieve more interactions with Y268 in the closed side-chain conformation and subsequently improve the potency of peptidomimetic inhibitors.
To test this prediction, we first overlaid the crystal structures of SARS-CoV-2 PL pro in complex with VIR251 as well as previously reported noncovalent inhibitors to dig out a more suitable scaffold. To our delight, we found that compound 1 could be a good starting point [21], as its amide bond is proximal to the amide bond between the P3 and P4 groups of VIR251 (red circles shown in Figure 1C). As a result, compound 2 is designed and synthesized based on the fusion of VIR251 and compound 1.

Compound 2 Showed Strong Inhibition against SARS-CoV-2 PL pro In Vitro
A short fluorogenic peptide substrate (RLRGG-AMC) was used to measure the inhibitory potency of compound 2 against SARS-CoV-2 PL pro by an enzymatic assay [24,26]. As anticipated, compound 2 significantly inhibited the protease's activity in a dose-dependent manner. The resulting IC50 value of compound 2 is 0.46 μM (Figure 2A), which is about 100-fold and 5-fold improvement in potency compared to that of VIR251 and compound 1, respectively. We also performed the same assay to determine the IC50 value (1.47 μM) of GRL0617, the positive control ( Figure S1), which is consistent with the one published [27,28]. Accordingly, the potency of compound 2 even exhibits a 2-fold increase compared to the positive control. It has been suggested in previous literature that reducing reagents such as dithiothreitol (DTT) in the enzymatic assay buffer can rule out promiscuous compounds that have non-specific inhibition towards many cysteine proteases [29,30]. For this reason, we carried out the enzymatic inhibition assay again in a DTTcontaining buffer. As expected, there is no apparent change in the inhibitory activities of compound 2 at varied concentrations in the presence and absence of 4 mM DTT ( Figure  2B), indicating its high specificity for SARS-CoV-2 PL pro . Additionally, considering the DUB activity of SARS-CoV-2 PL pro , an enzymatic inhibition assay was conducted using a second fluorogenic substrate (Ub-AMC) derived from Ub, which was synthesized according to the literature [31]. The result shows that compound 2 repeats the dose-dependent inhibition against the protease with an IC50 value of 0.46 μM (Figure 2A) which is identical to the one resulting from the enzymatic assay using RLRGG-AMC. These results together The red circles highlight the amide bonds of VIR251 and compound 1, and this amide bond of VIR251 is proximal to the one of compound 1.

Compound 2 Showed Strong Inhibition against SARS-CoV-2 PL pro In Vitro
A short fluorogenic peptide substrate (RLRGG-AMC) was used to measure the inhibitory potency of compound 2 against SARS-CoV-2 PL pro by an enzymatic assay [24,26]. As anticipated, compound 2 significantly inhibited the protease's activity in a dose-dependent manner. The resulting IC 50 value of compound 2 is 0.46 µM (Figure 2A), which is about 100-fold and 5-fold improvement in potency compared to that of VIR251 and compound 1, respectively. We also performed the same assay to determine the IC 50 value (1.47 µM) of GRL0617, the positive control ( Figure S1), which is consistent with the one published [27,28]. Accordingly, the potency of compound 2 even exhibits a 2-fold increase compared to the positive control. It has been suggested in previous literature that reducing reagents such as dithiothreitol (DTT) in the enzymatic assay buffer can rule out promiscuous compounds that have non-specific inhibition towards many cysteine proteases [29,30]. For this reason, we carried out the enzymatic inhibition assay again in a DTT-containing buffer. As expected, there is no apparent change in the inhibitory activities of compound 2 at varied concentrations in the presence and absence of 4 mM DTT ( Figure 2B), indicating its high specificity for SARS-CoV-2 PL pro . Additionally, considering the DUB activity of SARS-CoV-2 PL pro , an enzymatic inhibition assay was conducted using a second fluorogenic substrate (Ub-AMC) derived from Ub, which was synthesized according to the literature [31]. The result shows that compound 2 repeats the dose-dependent inhibition against the protease with an IC 50 value of 0.46 µM (Figure 2A) which is identical to the one resulting from the enzymatic assay using RLRGG-AMC. These results together suggest that compound 2 is a potent peptidomimetic inhibitor of SARS-CoV-2 PL pro while the fusion of VIR251 and compound 1 results in a strong synergetic effect on the inhibitory potency improvement. from two other highly pathogenic human coronaviruses, SARS-CoV and middle east respiratory syndrome coronavirus (MERS-CoV). Compound 2 inhibits SARS-CoV PL pro with an IC50 value of 0.39 μM ( Figure 2F) which is similar to its potency against SARS-CoV-2 PL pro (IC50 = 0.47 μM). However, compound 2 is basically ineffective towards MERS-CoV PL pro even at a concentration of 100 μM ( Figure 2F). The reason for such difference may be that SARS-CoV-2 PL pro shares a sequence identity of 82.9% with SARS-CoV PL pro while only 32.9% sequence identity with MERS-CoV PL pro [16]. In this regard, compound 2 serves as a novel potent peptidomimetic inhibitor targeting both SARS-CoV and SARS-CoV-2 PL pro .  The time-dependent inhibition assay was also performed to validate the covalent binding of compound 2. Specifically, different concentrations of compound 2 were incubated with protease at the final concentration of 50 nM for the indicated time, and the activity of the protease in each reaction was measured after incubation. It turns out that compound 2 shows time-dependent inhibition of SARS-CoV-2 PL pro ( Figure 2C), which was in line with the model of covalent binding. Covalent ligand binding often involves a two-step process, an initial reversible binding event followed by the formation of a covalent bond, which is characterized by the binding affinity (K i ) and the rate constant of covalent bond formation (k inact ), respectively [32]. The resulting K i and k inact values of compound 2 are 5.32 µM and 0.051 min −1 , respectively, and the ratio of these two parameters (k inact /K i ) is 9609 M −1 min −1 ( Figure 2D). Taken together, compound 2 is characterized as an effective peptidomimetic inhibitor targeting SARS-CoV-2 PL pro in a covalent manner.
To confirm the specific binding between inhibitors and SARS-CoV-2 PL pro , a thermal shift assay (TSA) was also applied after incubation of 40 µM protease with 200 µM compounds. As shown in Figure 2E, the melting temperature (T m ) of SARS-CoV-2 PL pro incubated with compound 2 significantly increases compared to that of apo protease (68.0 • C vs. 43.3 • C) and it is also much higher than that of the protease incubated with GRL0617 (68.0 • C vs. 50.4 • C). These results demonstrate that compound 2 directly binds to SARS-CoV-2 PL pro to greatly improve protein stability.
In addition, we also tested the inhibitory activity of compound 2 against the PL pro s from two other highly pathogenic human coronaviruses, SARS-CoV and middle east respiratory syndrome coronavirus (MERS-CoV). Compound 2 inhibits SARS-CoV PL pro with an IC 50 value of 0.39 µM ( Figure 2F) which is similar to its potency against SARS-CoV-2 PL pro (IC 50 = 0.47 µM). However, compound 2 is basically ineffective towards MERS-CoV PL pro even at a concentration of 100 µM ( Figure 2F). The reason for such difference may be that SARS-CoV-2 PL pro shares a sequence identity of 82.9% with SARS-CoV PL pro while only 32.9% sequence identity with MERS-CoV PL pro [16]. In this regard, compound 2 serves as a novel potent peptidomimetic inhibitor targeting both SARS-CoV and SARS-CoV-2 PL pro .

A Crystal Structure of SARS-CoV-2 PL pro Covalently Bound with Compound 2
To gain insights into the precise ligand binding mode, a crystal structure of SARS-CoV-2 PL pro in complex with compound 2 was determined at 2.5 Å resolution ( Figure 3A, Table 1). As expected, the contiguous electron density is clearly shown between C111 and the P1 group of compound 2 ( Figure 3B), demonstrating the covalent bond formed between the protease and 2. In addition, the P1-P3 groups of compound 2 form multiple hydrogen bonds (H-bonds) with the main chain of N109, G163, Y268, and G271 as well as the side chain of Y264 ( Figure 3A). The superimposed crystal structures of SARS-CoV-2 PL pro in complex with compound 2 and VIR251 reveal that the protein-ligand interaction patterns of the P1-P3 groups of two inhibitors are highly conserved ( Figure 3C). Meanwhile, the side chain of Y268 is in a closed conformation as anticipated so as to form π-π interactions with the P4-naphthyl group of compound 2; a salt bridge is also established between the side chain of D164 and the piperidine ring of the P4-position ( Figure 3A). The superimposition of crystal structures of SARS-CoV-2 PL pro bound with compounds 1 and 2 reveals that the P4 group of the two compounds employs a highly similar binding pose in the complexes ( Figure 3D). Therefore, in the SARS-CoV-2 PL pro -2 complex, two fragments individually adopted from VIR251 and compound 1, bind to the pocket as we desired, highlighting the elegance and utility of the structure-based drug design in significantly facilitating the development of new inhibitors.

A Crystal Structure of SARS-CoV-2 PL pro Covalently Bound with Compound 2
To gain insights into the precise ligand binding mode, a crystal structure of SARS-CoV-2 PL pro in complex with compound 2 was determined at 2.5 Å resolution ( Figure 3A, Table 1). As expected, the contiguous electron density is clearly shown between C111 and the P1 group of compound 2 ( Figure 3B), demonstrating the covalent bond formed between the protease and 2. In addition, the P1-P3 groups of compound 2 form multiple hydrogen bonds (H-bonds) with the main chain of N109, G163, Y268, and G271 as well as the side chain of Y264 ( Figure 3A). The superimposed crystal structures of SARS-CoV-2 PL pro in complex with compound 2 and VIR251 reveal that the protein-ligand interaction patterns of the P1-P3 groups of two inhibitors are highly conserved ( Figure 3C). Meanwhile, the side chain of Y268 is in a closed conformation as anticipated so as to form π-π interactions with the P4-naphthyl group of compound 2; a salt bridge is also established between the side chain of D164 and the piperidine ring of the P4-position ( Figure 3A). The superimposition of crystal structures of SARS-CoV-2 PL pro bound with compounds 1 and 2 reveals that the P4 group of the two compounds employs a highly similar binding pose in the complexes ( Figure 3D). Therefore, in the SARS-CoV-2 PL pro -2 complex, two fragments individually adopted from VIR251 and compound 1, bind to the pocket as we desired, highlighting the elegance and utility of the structure-based drug design in significantly facilitating the development of new inhibitors.

Substitution of the Warhead Leading to the Design of Compound 4
To further improve the potency of compound 2, we borrowed the warhead from compound 3 as it is a GRL0617 analog and possesses strong inhibition against SARS-CoV-2 PL pro (IC 50 = 94 nM) [21]. As a result, compound 4 was designed and synthesized ( Figure 4A). The enzymatic inhibition assay displays that the inhibitory potency of compound 4 is~1-fold higher than 2. The exact IC 50 values measured by using RLRGG-AMC and Ub-AMC substrates are 0.23 µM and 0.32 µM, respectively ( Figure 4B). In addition, compound 4 inhibits SARS-CoV-2 PL pro in a time-dependent manner ( Figure 4C), suggesting that a covalent binding mode is utilized by compound 4. Kinetic parameters of compound 4 binding with SARS-CoV-2 PL pro were also calculated and the resulting value of k inact /K i is 15,914 M −1 min −1 which also showed an increase compared to compound 2 ( Figure 4D). This improvement is mainly attributed to the substitution of the warhead as the k inact value of compound 4 is higher than that of compound 2 (0.074 min −1 vs. 0.051 min −1 ) while the K i values of the two compounds are similar (4.65 µM vs. 5.32 µM, Figures 2D and 4D). Moreover, the T m value of SARS-CoV-2 PL pro incubated with compound 4 has a 15.1 • C increase compared to that of apo protease, which confirms the direct binding between the compound and the protease ( Figure 4E). Consequently, compound 4 is a new covalent peptidomimetic inhibitor of SARS-CoV-2 PL pro with improved potency compared to compound 2.

The Cell-Based Protease and Cytotoxicity Assay of Two Designed Compounds
The intracellular inhibition of two designed compounds was tested using the PL-FlipGFP assay, a cell-based protease assay initially developed by Wang et al. [25]. In this assay, a fluorogenic GFP containing a cleavage site (LRGGAPTK) for SARS-CoV-2 PL pro , namely PL-FlipGFP, and a red fluorescent protein (mCherry) were constructed together. The ratio of GFP/mCherry fluorescent signal was used to represent the activity of the protease in cells and reveal the intracellular inhibition of compounds. The results show that compound 2 displays potent intracellular inhibition with the half-maximal effective concentration (EC50) of 3.61 μM ( Figure 5A), which exhibits ~4.4-fold improvement compared to GRL0617 (EC50 = 19.56 μM, Figure 5B) and comparable potency relative to compound 3 (EC50 = 2.97 μM, Figure 5C) that is the so far most potent covalent inhibitor of SARS-CoV-2 PL pro . Unfortunately, compound 4 shows remarkable inhibition against the protease in the enzymatic assay, but it hardly has an inhibitory effect in this cell-based protease assay ( Figure 5D). In addition, we tested the cytotoxicity of these mentioned inhibitors in HEK293T cells by the cell counting kit-8 (CCK8) assay. The results show that the halfmaximal cytotoxic concentration (CC50) value of compound 2 is ~180 μM ( Figure 5A). Collectively, compound 2 is a promising covalent peptidomimetic inhibitor of SARS-CoV-2 PL pro with potent intracellular inhibition and low cytotoxicity.

The Cell-Based Protease and Cytotoxicity Assay of Two Designed Compounds
The intracellular inhibition of two designed compounds was tested using the PL-FlipGFP assay, a cell-based protease assay initially developed by Wang et al. [25]. In this assay, a fluorogenic GFP containing a cleavage site (LRGGAPTK) for SARS-CoV-2 PL pro , namely PL-FlipGFP, and a red fluorescent protein (mCherry) were constructed together. The ratio of GFP/mCherry fluorescent signal was used to represent the activity of the protease in cells and reveal the intracellular inhibition of compounds. The results show that compound 2 displays potent intracellular inhibition with the half-maximal effective concentration (EC 50 ) of 3.61 µM ( Figure 5A), which exhibits~4.4-fold improvement compared to GRL0617 (EC 50 = 19.56 µM, Figure 5B) and comparable potency relative to compound 3 (EC 50 = 2.97 µM, Figure 5C) that is the so far most potent covalent inhibitor of SARS-CoV-2 PL pro . Unfortunately, compound 4 shows remarkable inhibition against the protease in the enzymatic assay, but it hardly has an inhibitory effect in this cell-based protease assay ( Figure 5D). In addition, we tested the cytotoxicity of these mentioned inhibitors in HEK293T cells by the cell counting kit-8 (CCK8) assay. The results show that the half-maximal cytotoxic concentration (CC 50 ) value of compound 2 is~180 µM ( Figure 5A). Collectively, compound 2 is a promising covalent peptidomimetic inhibitor of SARS-CoV-2 PL pro with potent intracellular inhibition and low cytotoxicity.

Discussion
The persistent pandemic COVID-19 together with previous epidemics of SARS and MERS has raised great awareness of the increasing infection risks of highly pathogenic CoVs. This concern calls for a huge demand for the discovery and development of anti-CoV drugs. In this work, using a fusion strategy guided by the structural information, we identified the novel peptidomimetic inhibitors covalently targeting SARS-CoV-2 PL pro , which is also well verified by the determined crystal structure. Meanwhile, the inhibitors such as compound 2 exhibit a submicromolar potency (IC50 = 0.46 μM) in the enzymatic assay and a significant inhibition (EC50 = 3.61 μM) in the cell-based protease assay ( Figure  S2).
The best inhibitor reported in the previous publication was also tested in our study [21], and it is referred to as compound 3. We found that the potency of our inhibitors is comparable to compound 3 (IC50 = 0.094 μM, EC50 = 2.97 μM). In addition, our compound has lower cytotoxicity in HEK293T cells (CC50 = 179.9 μM vs. 58.3 μM), indicating a higher safety profile of the compound. Therefore, a class of novel and potent inhibitors against SARS-CoV-2 PL pro has been identified in our study, which is valuable for further development.
It is noteworthy that our work presented here used a structure-based drug design approach to successfully design a cysteine-targeted covalent ligand. The structure information reveals the importance of Y268 side-chain conformation in protein-ligand interactions and allows us to efficiently obtain a potent covalent peptidomimetic inhibitor using a fusion strategy. The resulting compound 2 is more potent than both parental compounds, which indicates the strong synergetic effect of two fragments individually from two parental compounds to improve the potency of the new compound is achieved. To gain further insight into the protein-ligand interactions involving Y268 and other sur-

Discussion
The persistent pandemic COVID-19 together with previous epidemics of SARS and MERS has raised great awareness of the increasing infection risks of highly pathogenic CoVs. This concern calls for a huge demand for the discovery and development of anti-CoV drugs. In this work, using a fusion strategy guided by the structural information, we identified the novel peptidomimetic inhibitors covalently targeting SARS-CoV-2 PL pro , which is also well verified by the determined crystal structure. Meanwhile, the inhibitors such as compound 2 exhibit a submicromolar potency (IC 50 = 0.46 µM) in the enzymatic assay and a significant inhibition (EC 50 = 3.61 µM) in the cell-based protease assay ( Figure S2).
The best inhibitor reported in the previous publication was also tested in our study [21], and it is referred to as compound 3. We found that the potency of our inhibitors is comparable to compound 3 (IC 50 = 0.094 µM, EC 50 = 2.97 µM). In addition, our compound has lower cytotoxicity in HEK293T cells (CC 50 = 179.9 µM vs. 58.3 µM), indicating a higher safety profile of the compound. Therefore, a class of novel and potent inhibitors against SARS-CoV-2 PL pro has been identified in our study, which is valuable for further development.
It is noteworthy that our work presented here used a structure-based drug design approach to successfully design a cysteine-targeted covalent ligand. The structure information reveals the importance of Y268 side-chain conformation in protein-ligand interactions and allows us to efficiently obtain a potent covalent peptidomimetic inhibitor using a fusion strategy. The resulting compound 2 is more potent than both parental compounds, which indicates the strong synergetic effect of two fragments individually from two parental compounds to improve the potency of the new compound is achieved. To gain further insight into the protein-ligand interactions involving Y268 and other surrounding residues, computational studies are needed for the optimization of the inhibitors. Accordingly, the present study highlights the power and advantage of the structure-based drug design strategy.
In summary, we report a class of novel and potent peptidomimetic inhibitors that covalently target SARS-CoV-2 PL pro . Our findings present a new scaffold of SARS-CoV-2 PL pro inhibitors and provide a promising starting point for further optimization.

Materials and Methods for Synthesis and Characterization of Compound 2 and 4
All chemical reagents and compounds were used as supplied by standard suppliers (Bidepharm, Shanghai, China) without further purification. All reactions were monitored by thin layer chromatography (TLC), and visualization was achieved by using ultraviolet light (254 nm). The column chromatography was performed using flash chromatography. 1 H NMR and 13 C NMR spectra were recorded on a BRUKER AVANCE NEO 500 or 400 (Billerica, MA, USA) at 500 or 400 MHz and 126 or 100 MHz, respectively. Coupling constants (J) are expressed in hertz. Chemical shifts (δ) of NMR spectra are reported in parts per million (ppm) units. Mass spectra (MS) of compounds were measured using a Thermo Fisher FINNIGAN LTQ spectrometer (Thermo Finnigan, San Jose, CA, USA). The purity of all tested compounds was confirmed to be ≥95% by HPLC. Analysis was performed on a EClassical 3100 HPLC system (Elite, Dalian, China) under the following analytical method: column, WondaSil Superb C18 (5 µm, 4.6 mm × 250 mm); solvent A: water containing 1‰ trifluoroacetic acid (TFA); solvent B: acetonitrile containing 1‰ TFA; gradient, 10% B to 100% B over 15 min, 100% B for 5 min; flow rate, 1 mL/min; detective wavelength, 220 nm, 254 nm; column temperature, 25 • C. The spectroscopic graphs are provided in the Supplementary Materials (Figures S3-S16).

Plasmids Construction
The cDNA encoding SARS-CoV-2 PL pro , SARS-CoV PL pro and MERS-CoV PL pro with a 6× His tag at N-terminus were cloned into a pET-15b-SUMO vector, respectively. Lb pro  from foot-and-mouth disease virus with an 8× His tag at N-terminus was cloned into a pET-11d vector. The human Ub gene with a 6× His tag at C-terminus was cloned into a pET-28a vector. All cDNAs mentioned above were E. coli. codon-optimized and synthesized by GenScript (Shanghai, China).
For transfection of mammalian cells, the cDNA encoding SARS-CoV-2 PL pro with mammalian codon optimization was also ordered from GenScript and cloned into the pcDNA3.1 with a C-terminal FLAG tag. The sequence of pcDNA3-PL-flipGFP-T2A-mCherry was designed based on the plasmid of pcDNA3-TEV-flipGFP-T2A-mCherry (Addgene catalog NO.124429) where TEV cleave site was replaced by SARS-CoV-2 PL pro cleavage site (amino acid sequence: LRGGAPTK) and ordered from GenScript.

Expression and Purification of SARS-CoV, SARS-CoV-2 and MERS-CoV PL pro
The expression and purification processes of SARS-CoV-2 PL pro , SARS-CoV PL pro and MERS-CoV PL pro were similar as described below. Briefly, the expression plasmids were transformed into E. coli. Rosetta (DE3) competent cells. The cells were grown in LB medium to an OD 600 of 0.8 and induced by IPTG at a final concentration of 0.5 mM and shaken at 18 • C overnight. The 6× His-SUMO2-SARS-CoV-2 PL pro was first purified by the Ni-NTA column (GE Healthcare, Marlborough, MA, USA) and cleaved by SUMO Specific Peptidase 2. The resulting protein samples were further purified by SP-Sepharose (GE Healthcare) and Superdex75 (GE Healthcare). The eluted proteins were stored in a solution containing 25 mM HEPES (pH 7.5) and 2 mM DTT for the subsequent experiments.

Crystallization and Data Collection
SARS-CoV-2 PL pro (1 mg/mL) was incubated with 100 µM peptidomimetic inhibitors in 25 mM HEPES (pH 7.5) at 10 • C for 2 h. The mixture was supplemented with 10 mM DTT and concentrated to 12 mg/mL for crystallization. The crystal of SARS-CoV-2 PL pro in complex with compound 2 was grown by mixing equal volumes of the protein/compound and a reservoir (0.2 M sodium bromide, 20% w/v PEG 3350) on a 96-well sitting plate at 20 • C. Before data collection, crystals were flash-frozen in liquid nitrogen directly in the presence of the reservoir solution. X-ray diffraction data were collected at beamline BL19U1 at the Shanghai Synchrotron Radiation Facility and processed with the program autoPROC 1.0 [33]. The structure was solved by the program PHASER 2.8.3 [34] and refined with the program PHENIX 1.17.1-3660 [35]. The refined structure was deposited into the Protein Data Bank with accession code 8IHO. The complete statistics as well as the quality of the solved structures are given in Table 1.

Synthesis of the Fluorogeneic Substrate Ub-AMC Based on Ub
The expression process of Ub and Lb pro was the same as that of SARS-CoV-2 PL pro . Both Ub and Lb pro were first purified by a Ni-NTA column (GE Healthcare), followed by buffer exchange using a Superdex75 (GE Healthcare) pre-equilibrated with a buffer containing 50 mM HEPES, 150 mM NaCl, pH 8.0. The fractions were flash-frozen in liquid nitrogen and stored at −80 • C for further use. Ub-AMC was synthesized according to previous references [31]. Specifically, Gly-Gly-AMC was dissolved in DMSO to a saturated solution and mixed with an equal volume of Ub stock solution (5 mM). After that, Lb pro (40 µM) was added to the reaction solution at room temperature. After incubating for about 6 h, the reaction mixture was purified by a Ni-NTA column (GE Healthcare). The flow-through was further purified by Superdex 75 (GE Healthcare) pre-equilibrated with a buffer containing 50 mM HEPES, 150 mM NaCl, and pH 8.0. The fraction of the second main peak in the chromatogram was collected and stored at −80 • C for further use.

Enzymatic Assays of SARS-CoV, SARS-CoV-2 and MERS-CoV PL pro
The enzymatic inhibition assays of all compounds were performed using 96-well plates at room temperature. The fluorogenic substrates used in the assay were RLRGG-AMC [36] which was synthesized by GenScript or Ub-AMC synthesized as described above. The assays were performed in a total volume of 120 µL, which contained the following components: 50 mM HEPES pH 7.0, 0.1 mg/mL BSA, 50 nM SARS-CoV-2 PL pro , indicated concentrations of the compound or an equal volume of solvent (DMSO or H 2 O). After 60 min incubation, reactions were initiated with the addition of RLRGG-AMC or Ub-AMC to reach a final concentration of 20 µM or 1.5 µM, respectively. After that, the fluorescent signal (excitation: 360 nm, emission: 460 nm) was immediately measured every 1 min for 5 min with a BioTek H1 plate reader (Winooski, VT, USA). The initial velocity of the reaction was obtained by fitting the linear portion of the curve into a straight line and the inhibition effect of compounds was evaluated by calculating the change of the initial rate. Half maximal inhibitory concentration (IC 50 ) was determined by nonlinear regression analysis of the dose-response curves using GraphPad Prism 9.0. Three independent experiments were performed.

Kinetic Analysis of Compounds Covalently Binding to SARS-CoV-2 PL pro
The binding of covalent compounds with SARS-CoV-2 PL pro can be described in two steps: an initial reversible binding event (K i ) followed by the formation of the covalent bond (k incat ). For determination of K i and k inact , SARS-CoV-2 PL pro at a final concentration of 50 nM was incubated with various concentrations of the compound for the indicated time. At each time point, the enzymatic assay was carried out as mentioned above. Relative enzymatic activity was calculated by the initial velocity ratio of reactions added with the compound over to the reaction added with solvent. Relative enzymatic activity for various concentrations of the compound over a time course was fitted to the semilogarithmic plot equation to generate observed the rate constant value (k obs ) for each concentration. The resulting k obs values were then plotted versus compound concentrations ([C]), then k inact and K i values were calculated according to the equation: k obs = k inact × ([C]/([C] + K i )) using GraphPad Prism 9.0. For each compound, three independent experiments were performed for the determination of k inact and K i values. Three independent experiments were performed.

Thermal Shift Assay of SARS-CoV-2 PL pro Incubated with Compounds
Inhibitors binding to SARS-CoV-2 PL pro was estimated by differential scanning fluorimetry via a Bioer Real-Time PCR QuantGene 9600 machine (Hangzhou, China). Specifically, SARS-CoV-2 PL pro was diluted to a final concentration of 40 µM. Then, the diluted protein was added with 200 µM inhibitor and treated at 25 • C for 30 min. After that, 1 × SYPRO orange dye was added to the mixture, and the fluorescence was monitored under a temperature gradient range from 25 • C to 90 • C with 0.03 • C/s incremental step. The measured T m was calculated from a melt curve with the program Gene-9660. Three independent experiments were performed.

Cell-Based PL-FlipGFP Assay
The cell-based PL-FlipGFP assay was designed stemming from previous reports [11,25]. Briefly, HEK293T cells were seeded in 6-well plates. The following day, cells at 60-80% confluency were co-transfected with 2 µg pcDNA3-PL-flipGFP-T2A-mCherry plasmid and 0.2 µg protease expression plasmid in each well with 6.6 µg polyethylenimine (PEI). Five hours later, the cells were digested and divided into 96-well plates to a confluence of 30-40%. Different concentrations of the compounds were added and then the cells were incubated at 37 • C for 48 h. On the third day, the supernatant was removed and 60 µL lysis buffer (100 mM Tris pH 7.5, 100 mM NaCl, 0.5% Triton X-100) was added. After 5 min of incubation at 4 • C, 40 µL cell lysate was transferred to a 384-well plate and the fluorescent signal of GFP (excitation: 475 nm, emission: 505 nm) and mCherry (excitation: 580 nm, emission: 610 nm) were measured. The inhibition ratio was calculated based on the ratio of GFP/mCherry fluorescent signal. The EC 50 value was calculated by plotting the inhibition ratio over the applied compound concentration in GraphPad Prism 9.0. Three independent experiments were performed.

Cytotoxicity Assay
The cytotoxicity of compounds on the HEK293T cells was determined by cell counting kit-8 (CCK8) assays. HEK293T cells were treated with compounds at different concentrations. DMSO-treated cells were used as the control. After 24 h, the cytotoxicity was measured using CCK8 according to the manufacturer's instructions (Yeasen, Shanghai, China). At final, the absorbance at a wavelength of 450 nm was measured using a BioTek H1 plate reader and the CC 50 value was calculated in GraphPad Prism 9.0.