Nitrile-Based Inhibitors Targeting both SARS-CoV‑2 Main Protease and Human Cathepsin L

Abstract

The SARS-CoV-2 main protease (M^pro) and host cysteine protease cathepsin L (CatL) are both attractive targets for antiviral intervention. Herein we describe the rational design and synthesis of dual M^pro/CatL inhibitors derived from the previously reported leads SM141 and SM142. Optimization focused on replacement of the acrylate ester warhead in SM141 and SM142 with a nitrile, modifying the P2, P3, and P4 capping groups, and variation of the P1 lactam ring size. In addition, conformational restriction was achieved through macrocyclization between the P1 and P4 side chains to enhance binding affinity. These efforts identified multiple dual-target inhibitors with nanomolar potency against both M^pro and CatL and potent antiviral activity as well as selective M^pro inhibitors with strong enzymatic activity. Targeting both a viral protease and a host factor may provide more durable antiviral efficacy by reducing susceptibility to resistance arising from viral evolution of M^pro.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of coronavirus disease 2019 (COVID-19), has posed an unprecedented global health challenge since its emergence in late 2019. As of September 2024, more than 700 million cases and over 7 million deaths have been reported worldwide, underscoring the profound public health impact of this pandemic. − Intensive research efforts have led to the development and approval of multiple vaccines and the clinical evaluation of numerous antiviral agents, with only a limited number achieving market authorization. Despite these advances, the continued emergence of viral variants threatens the efficacy of existing vaccines and therapeutics. − Although most infections result in mild symptoms comparable to seasonal influenza, COVID-19 remains potentially fatal for vulnerable populations. Furthermore, the growing prevalence of post-COVID-19 condition (long COVID), , characterized by persistent complications such as chronic fatigue and cognitive impairment, highlights the need for novel therapeutic strategies. Consequently, the development of alternative antiviral agents remains an urgent priority to address these evolving challenges.

The SARS-CoV-2 main protease (M^pro), also referred to as 3C-like protease (3CL^pro), is indispensable for viral replication and maturation. This cysteine protease orchestrates the cleavage of viral polyproteins pp1a and pp1ab into functional nonstructural proteins that assemble the replication–transcription complex. − Catalysis is mediated by a highly conserved catalytic dyad − comprising cysteine 145 (C145) and histidine 41 (H41), wherein H41 facilitates deprotonation of C145 to generate a nucleophilic thiolate that drives peptide bond hydrolysis. Inhibitors that target C145 leverage an electrophilic warhead (e.g., α-keto amides, α,β-unsaturated ketones, aldehydes, dihaloacetamides, and vinyl sulfones) that forms a covalent adduct with this active-site cysteine, thereby abrogating enzymatic function. Numerous such inhibitors have been identified as promising antiviral agents targeting M^pro. −

Beyond viral proteases, host factors critically influence SARS-CoV-2 pathogenesis. Human cathepsin L (hCatL), a lysosomal cysteine protease, facilitates viral entry by cleaving the spike protein within endosomal compartments. − Consequently, dual inhibition of M^pro and hCatL offers a compelling therapeutic paradigm that simultaneously targets viral replication and entry to enhance antiviral efficacy, mitigate resistance arising from viral mutations, and provide a complementary approach to existing interventions aimed at curbing infection and transmission.

Despite the therapeutic promise of simultaneously targeting viral and host proteases, few studies have systematically explored dual inhibitors of M^pro and hCatL. ,− In our previous work, we identified SM141 and SM142 as potent dual inhibitors derived from the parent scaffolds D-FFRCMKyne and D-FFCitCMKyne through rational optimization. These compounds exhibit a distinctive binding mode within the M^pro active site while maintaining selectivity by sparing papain-like protease (PL^pro), − another cysteine protease integral to the viral life cycle. Functionally, SM141 and SM142 demonstrated robust antiviral activity in A549 cells expressing human ACE2 (hACE2), with EC₅₀ values of 8.2 ± 0.9 nM and 14.7 ± 2.2 nM respectively and exhibited nanomolar potency against hCatL. The unique structural and functional attributes of these inhibitors underscore their potential as a foundation for next-generation antiviral agents that leverage dual-target engagement to enhance efficacy and mitigate resistance.

The primary objective of this study was to design dual inhibitors of M^pro and hCatL with improved biochemical potency through rational modification of the SM141 and SM142 scaffolds. Our strategy involved replacing the acrylate ester warhead with a moderately reactive nitrile group, systematically varying the P2, P3, and P4 capping groups, and adjusting the size of the P1 lactam ring (Figure ). Nitriles have attracted increasing interest as electrophilic warheads because their mild electrophilicity, selectivity, and chemical stability reduce off-target reactivity and enable the development of safer therapeutics. , To enhance conformational rigidity and binding affinity, we also introduced cyclization between the P1 and P4 side chains (Figure ). These approaches yielded multiple dual inhibitors with nanomolar potency as well as M^pro-selective compounds that exhibit robust biochemical activity.

1.

Design strategies for novel derivatives and analogs of the lead compounds SM141 and SM142, featuring a nitrile warhead, targeting SARS-CoV-2 M^pro and cathepsin L.

Using SM141 as a template, we replaced the acrylate ester with a nitrile warhead while retaining the P2 and P3 positions. The size of the γ-lactam glutamine derivative at P1 was varied, and a systematic structure–activity relationship (SAR) analysis was performed to evaluate different P4 capping groups (Table ). Incorporation of a Cbz group yielded UM-004, which exhibited submicromolar inhibitory activity (IC₅₀ < 1 μM) against both M^pro and hCatL, with approximately 2-fold greater potency for M^pro (Table ). Substitution of Cbz with a trifluoroamide produced UM-005, which maintained comparable submicromolar activity (Table ). Introduction of Boc and methyl carbamate groups at the N-terminus generated UM-006 and UM-158, respectively; both retained submicromolar potency, although UM-158 was ∼2.5-fold more potent than UM-006, indicating that less bulky groups are better tolerated at the P4 N-terminus (Table ).

1. SAR Analysis of P4 Cap Modifications in SM141/SM142 Analogs.

^aIC₅₀ values are reported as mean ± SEM (n = 2).

^bM^pro inhibition and hCatL inhibition were determined by monitoring the fluorescence generated upon cleavage of the following substrates: DABCYL-KTS­AVL­QSGF­RKME-EDANS-NH₂ and Z-FR-AMC, respectively.

To further probe P4 diversity, we increased the γ-lactam ring to six members and incorporated trifluoro, methyl, ethyl, isopropyl, and trifluoro sulfonamide groups at the N-terminus, yielding UM-059, UM-061, UM-062, UM-063, and UM-188, respectively (Table ). UM-063 (isopropyl) displayed IC₅₀ values >10 μM for both targets, suggesting poor tolerance for bulky substituents at this position (Table ). By contrast, UM-059, UM-061, UM-062, and UM-188 retained submicromolar activity. Notably, elongating the P4 cap by one methylene unit (UM-062) resulted in a 2-fold increase in M^pro potency compared to UM-061, while UM-188 (trifluoro sulfonamide) exhibited balanced activity against both targets (Table ). UM-059 demonstrated a 7-fold increase in M^pro potency relative to UM-005, while maintaining similar activity against hCatL (Table ). Collectively, these findings indicate that enlarging the lactam ring enhances M^pro activity and that steric constraints at P4 strongly influence inhibitory potency.

Next, we varied the substitution patterns at the P2 and P3 positions, incorporating selected tolerated P4 capping groups from Table while maintaining the γ-lactam glutamine derivative at P1 and the nitrile warhead (Table ). Our initial compound, UM-075, contained m-difluoro substitution on the l-phenylalanine ring at P2, a d-phenylalanine residue at P3, and an N-terminal trifluoroamide cap. UM-075 exhibited submicromolar activity against both targets, with ∼10-fold greater potency toward M^pro compared to hCatL (Table ). Replacing the P4 cap with methyl carbamate yielded UM-161, which showed a 4-fold reduction in M^pro activity but retained hCatL potency (Table ). Introducing o-fluoro substitution at P2 produced UM-179, resulting in a 6-fold decrease in M^pro activity while maintaining hCatL activity. Next, UM-159 was synthesized with l-phenylalanine at P2, p-fluoro-substituted d-phenylalanine at P3, and an N-terminal methyl carbamate cap; this compound demonstrated submicromolar activity against both targets, with an ∼2-fold increase in hCatL potency (Table ). Substituting P3 in UM-159 with m-difluoro-substituted d-phenylalanine generated UM-160, which showed a 2-fold increase in M^pro activity while retaining hCatL potency (Table ). Collectively, these results suggest that m-fluoro substitution is preferred over p-fluoro substitution at P3, and similarly, meta substitution on l-phenylalanine at P2 is favored over ortho substitution.

2. SAR Analysis of P₂, P₃, and P₄ Cap Modifications in SM141/SM142 Analogs.

^aIC₅₀ values are reported as mean ± SEM (n = 2).

^bM^pro inhibition and hCatL inhibition were determined by monitoring the fluorescence generated upon cleavage of the following substrates: DABCYL-KTS­AVL­QSGF­RKME-EDANS-NH₂ and Z-FR-AMC, respectively.

Next, we synthesized UM-067, which incorporated l-cyclopropylalanine at P2, d-phenylalanine at P3, and a trifluoroamide P4 cap. This modification at P2 resulted in a dramatic increase in M^pro potency, with UM-067 achieving an IC₅₀ of 11 nM while losing activity against hCatL (IC₅₀ > 10,000 nM) (Table ). Replacing the P4 cap with methyl carbamate produced UM-162, which maintained similar M^pro potency but regained some of its activity against hCatL, highlighting the role of P4 capping groups in modulating dual inhibition (Table ). Introducing p-fluoro substitution on the d-phenylalanine at P3 in UM-162 yielded UM-163, which displayed submicromolar activity for both targets and 42-fold activity for M^pro over hCatL (Table ). Overall, incorporation of the l-cyclopropylalanine moiety at P2 strongly favored M^pro activity over hCatL.

Next, we maintained P1 as the γ-lactam glutamine derivative and P3 as d-phenylalanine, while keeping the trifluoroamide group at the P4 cap unchanged. To rigidify the P2–P3 backbone amide, spirocyclic and bicycloproline moieties were introduced (Table ). The first analog, UM-056, incorporated spirocyclic cyclopropylproline at P2 and exhibited submicromolar potency with an IC₅₀ of 740 nM against M^pro but lacked activity against hCatL. Substitution of cyclopropylproline with bicycloproline derivativescyclopentylproline and dimethylcyclopropylprolineyielded UM-057 and UM-068, respectively. Both compounds demonstrated comparable submicromolar potency against M^pro (IC₅₀ = 110–120 nM) but similarly lost activity against hCatL (Table ). Collectively, these findings indicate that rigidifying the P2–P3 backbone amide enhances M^pro activity over hCatL.

3. Rigidification of the P₂–P₃ Amide Bond and Cyclization between the P₁ Side Chain and P₄ Carbamate.

^aIC₅₀ values are reported as mean ± SEM (n = 2).

^bM^pro inhibition and hCatL inhibition were determined by monitoring the fluorescence generated upon cleavage of the following substrates: DABCYL-KTS­AVL­QSGF­RKME-EDANS-NH₂ and Z-FR-AMC, respectively.

To enhance conformational rigidity and target engagement, we designed the macrocyclic inhibitors UM-034 and UM-050 by linking the P1 and P4 side chains. Macrocyclization is a well-established strategy in medicinal chemistry that preorganizes ligands into bioactive conformations, often improving affinity, selectivity, and activity. , Consistent with recent studies on M^pro inhibitors, this approach was employed to generate UM-034 and UM-050 with the potential for enhanced target engagement relative to their linear counterparts (Table ). Using the SM141 template, P2 and P3 were kept constant. At P1, the position was replaced with a C-terminal nitrile-containing glutamine, whose side chain was linked to a carbamate at the P4 cap via hexanyl (UM-034) or butanyl (UM-050) alkyl chains (Table ). UM-034 exhibited an IC₅₀ of ∼2 μM against both targets, whereas UM-050 showed a 2-fold decrease in M^pro activity and a 4.5-fold decrease in hCatL activity (Table ). Overall, these findings indicate that the linker length connecting the P1 side chain and P4 carbamate significantly influences activity, with the longer alkyl chain providing better and more balanced potency against both targets.

Next, we evaluated the antiviral activity of UM-005, UM-067, and UM-163 against live SARS-CoV-2 replication in hACE2-expressing A549 cells. These compounds possess a range of different M^pro and hCatL potencies: UM-005 is a dual inhibitor that exhibits comparable submicromolar potency against both M^pro and hCatL (IC₅₀ ≈ 350 nM); UM-067, is our most potent M^pro-selective inhibitor with an IC₅₀ of 11 nM; and UM-163, is a highly potent M^pro inhibitor (IC₅₀ = 18 nM) with significant activity toward hCatL (IC₅₀ = 760 nM) (Table ). Nirmatrelvir served as a positive control (IC₅₀ = 11 ± 2 nM). In hACE2-expressing A549 cells, UM-005, UM-067, and UM-163 exhibited antiviral activity with IC₅₀ values of 330 ± 100, 40 ± 18, and 65 ± 13 nM, respectively. UM-067 was the most potent analog in this series, displaying activity within ∼4-fold of nirmatrelvir (Figure A–D). Furthermore, UM-005, UM-067, and UM-163 showed <5% cytotoxicity at concentrations up to 50 μM, indicating favorable selectivity in this assay (Figure E–G).

2.

Antiviral activities and cytotoxicities of M^pro/hCatL inhibitors. (A–D) Antiviral activities of (A) nirmatrelvir (IC₅₀ = 11 ± 2 nM), (B) UM-005 (IC₅₀ = 330 ± 100 nM), (C) UM-067 (IC₅₀ = 40 ± 18 nM), and (D) UM-163 (IC₅₀ = 65 ± 13 nM) against live SARS-CoV-2 infection in hACE2-expressing A549 cells. The y-axis scale in (A) differs from those in (B–D). (E–G) Cytotoxicities of (E) UM-005, (F) UM-067, and (G) UM-163 assessed in A549-hACE2 cells after 24 h treatment at the indicated concentrations. Lactate dehydrogenase (LDH) release into culture supernatants was quantified as a measure of cytotoxicity. Data are presented as mean ± SEM from three independent experiments.

Next, to determine the structural basis of inhibition of our lead inhibitors, UM-005 and UM-067, we solved high-resolution cocrystal structures of each compound bound to M^pro. The UM-005 structure was determined at 2.71 Å in space group P1 with two monomers in the asymmetric unit (ASU). The UM-067 structure was solved at 1.87 Å in the monoclinic space group C2, with one monomer in the ASU. Data collection and refinement metrics are summarized in Table S1. Although the UM-005 costructure contained two molecules in the ASU, strong electron density for the inhibitor was observed only in monomer A. Monomer B displayed weak and discontinuous density at the active site, precluding reliable placement of the inhibitor. We therefore docked UM-005 in monomer A and water molecules in the residual density of monomer B. In both structures, continuous electron density was observed between the catalytic C145 residue and the electrophilic nitrile warhead of the inhibitors, consistent with covalent bond formation between M^pro and the small molecules.

UM-005 and UM-067 differ only at the P2 position, where UM-005 contains an l-phenylalanine group and UM-067 contains an l-cyclopropylalanine moiety. In both structures, the nitrile warhead interacts with the backbone amines of G143, S144, and C145, as well as nearby water molecules. Notably, in the UM-067 structure, additional positive density was observed in the difference map adjacent to the warhead, which is absent in the UM-005 structure. After some investigation, a sodium ion was assigned to this density that is usually occupied by a water molecule. Interestingly, H41 of the S2 subsite is oriented such that the τ nitrogen of the imidazole ring lies within 3 Å of the sodium ion. By contrast, H41 in the UM-005 structure is rotated 180° and forms a hydrogen bond with the backbone carbonyl of H163 (Figure A-B).

3.

Crystal structures of (A) UM-005- and (B) UM-067-bound SARS-CoV-2 M^pro. Hydrogen bonds are displayed as dark-gray dashed lines, red spheres are water molecules, and the purple sphere in the UM-067 structure is a sodium ion.

The glutamine mimicking γ-lactam group at the P1 position forms multiple hydrogen bond interactions within the S1 subsite of M^pro, namely, S144, H163, E166, and H172. Backbone hydrogen bonds are also observed between the γ-lactam moiety and F140, L141, and E166. The S2 subsite, comprising H41, M49, Y54, and M165, prefers residues similar in size to leucine at the P2 position, as observed for inhibitors GC-376 and PF-00835231. In both UM-005 and UM-067, hydrophobic interactions within the S2 subsite stabilize the respective P2 groups (Figure ).

Both inhibitors also contain a d-phenylalanine moiety at P3 and a trifluoramide cap at P4. The d-phenylalanine group forms hydrophobic interactions with alkyl atoms of Q189 (S5 subsite) and P168, as well as crystal lattice contacts with P252. The side chain of Q189 forms a hydrogen bond with the inhibitors at the peptide bond linking P2 and P3. The trifluoroamide group is solvent exposed and interacts with water molecules, as well as the backbone of E166, L167, and P168. The fluorine atoms of the trifluoroacetamide group at P4 form close F···O contacts with both the backbone carbonyl oxygen and the side-chain carboxylate oxygen atoms of E166. Similar CF···O contacts have been reported , and, in some cases, interpreted as fluorine-mediated halogen-bonding interactions (Figure ). Collectively, these structures elucidate the molecular basis of inhibition of the M^pro, shedding light into how these small molecules block viral maturation and propagation.

Although UM-005 and UM-067 differ only in their P2 substituent, replacement of the phenyl group with a cyclopropyl group resulted in a 33-fold increase in M^pro inhibitory potency and a substantial improvement in selectivity over hCatL. Previous studies have shown that the S2 subsite of M^pro preferentially accommodates relatively small hydrophobic substituents similar in size to leucine, ,,,,− whereas larger hydrophobic groups can be less well tolerated because of the restricted dimensions of the pocket. Consistent with this trend, analysis of the cocrystal structures revealed that the P2 substituents of UM-005 and UM-067 adopt distinct orientations within the S2 pocket. In UM-005, the P2 phenyl group is positioned such that it does not participate in the edge-to-face aromatic interaction between the P2 and P3 substituents observed in SM141 (Figures and S23), suggesting a less favorable binding geometry. In contrast, the smaller cyclopropyl group of UM-067 appears to be more effectively accommodated within the S2 pocket, potentially reducing steric constraints and enabling a more favorable overall ligand conformation. These structural differences likely contribute to the enhanced M^pro potency of UM-067. While the molecular basis for the observed hCatL selectivity cannot be definitively established in the absence of a hCatL cocrystal structure, the results suggest that the steric and conformational properties of the P2 substituent play a critical role in determining both potency and selectivity within this inhibitor series. Accordingly, optimization of the P2 substituent represents an important strategy for future inhibitor design.

Given the nanomolar biochemical potency UM-067 and its strong bias as an M^pro inhibitor, we evaluated its human and mouse plasma protein binding and liver microsomal stability (Table ). UM-067 showed moderate binding to both human and mouse plasma protein (Table ). The percent unbound value for the internal standard warfarin fell within its established historical range, while that of carbamazepine was slightly below its historical benchmark. In liver microsomes, UM-067 exhibited comparable metabolic stability relative to the positive control diclofenac, indicating that this compound shows favorable metabolic stability.

4. Combined Liver Microsomal Stability and Plasma Protein Binding Data.

parameter	verapamil (neg. control)	diclofenac (pos. control)	carbamazepine	warfarin	UM-067
human liver microsomal T 1/2 (min)	6.8	14.0	na	na	27.0
mouse liver microsomal T 1/2 (min)	2.4	30.8	na	na	15.7
human % bound drug	na	na	83.2	99.4	80.3
human % recovery	na	na	94.4	106.5	100.2
human % stability	na	na	95.9	107.2	99.2
mouse % bound drug	na	na	77.2	93.3	90.7
mouse % recovery	na	na	105.9	103.1	103.7
mouse % stability	na	na	110.4	106.1	102.8

^ana = not applicable.

In this study, we report the design and synthesis of novel inhibitors targeting both the M^pro and hCatL, as well as compounds exhibiting selective inhibition of M^pro. Among these, UM-005 and UM-067 demonstrated significant antiviral activity in hACE2-expressing A549 cells infected with SARS-CoV-2. Of the newly synthesized compounds, 14 displayed submicromolar potency against both targets, while six exhibited selective inhibition of M^pro. Notably, UM-067 emerged as the most potent M^pro-selective inhibitor, highlighting its potential as a lead compound for further optimization. SAR analysis revealed that less bulky substituents at the P4 N-terminus were better tolerated, m-fluoro substitution on the phenyl ring at the P2 and P3 positions was favored over o-fluoro substitution, and incorporation of aliphatic alkyl small groups at P2 enhanced M^pro activity over hCatL. Increasing the size of the γ-lactam glutamine derivative further improved potency. Additionally, cyclization between the P1 and P4 side chains produced macrocyclic inhibitors whose activity was dependent on linker length.

Relative to the reference inhibitor SM141, the compounds described herein generally exhibited increased M^pro potency and reduced hCatL inhibition, indicating that modifications within this series preferentially enhanced engagement of the viral protease while maintaining measurable activity against the host target. SAR analysis revealed that both the P2 substituent and P4 capping group play important roles in modulating potency and selectivity across the two proteases. Most notably, UM-005 and UM-067 differ only at the P2 position; replacement of a phenyl group with cyclopropyl resulted in a 33-fold improvement in M^pro inhibitory potency together with enhanced selectivity over hCatL, highlighting the critical role of the P2 substituent in determining protease preference. These findings demonstrate that subtle structural changes can markedly influence the balance between viral and host protease inhibition, providing valuable guidance for future optimization efforts. Although optimization toward M^pro potency reduced hCatL inhibition relative to earlier analogs such as SM141 and SM142, dual-target activity was retained across the scaffold. UM-160 displayed the most balanced inhibitory profile, with IC₅₀ values of 140 ± 0 nM against M^pro and 110 ± 25 nM against hCatL, identifying it as a promising lead for further development. Collectively, these results define key structural determinants governing M^pro and hCatL inhibition and establish a framework for the rational design of next-generation antiviral agents with improved potency, selectivity, and therapeutic potential.

Experimental Procedures

Reagents and Compound Synthesis

Synthetic methods and characterization of the dual and selective M^pro inhibitors described herein are reported in the Supporting Information (Schemes S1–S24 and Figures S1–S22). All other reagents were obtained from commercial sources, and their specific sources are indicated below or in the Supporting Information.

Ethical Statement

All animal experiments were conducted in accordance with institutional and applicable regulatory guidelines and were approved by the Institutional Animal Ethics Committee (IAEC) of Syngene International Ltd. (protocol numbers SYNGENE/IAEC/1632/10-2024 and SYNGENE/IAEC/1681/03-2025).

Determination of IC₅₀ Values for SARS-CoV-2 M^Pro

Inhibition assays were performed by monitoring the cleavage of a fluorogenic M^Pro substrate (DABCYL-KTS­AVL­QSGF­RKME-EDANS-NH₂) in the presence and absence of the inhibitors. Cleavage of this substrate C terminal to the glutamine (Q) produces an increase in fluorescence. M^Pro (12.5 nM final) in assay buffer (20 mM HEPES, 120 mM NaCl, 0.4 mM EDTA, 4 mM DTT and 20% glycerol pH 6.5) was treated with either test compound (0–100 μM) or 5 μL of assay buffer with DMSO (1% final DMSO) for 30 min at 30 °C. Next, the peptide-substrate was added (20 μM final), and the reaction mixture was incubated for 60 min at 30 °C. The total volume of the assay mixture was 20 μL. End-point analysis was performed by recording the fluorescence at excitation/emission wavelengths of 360 nm/490 nm using a TecanSpark instrument. The IC₅₀ values were determined from the GraphPad analysis tool [log­(inhibitor) vs response; variable slope, four parameters]. All measurements were done in duplicate.

Determination of IC₅₀ Values for Cathepsin L

Inhibition assays were performed using a fluorogenic substrate (Z-FR-AMC) that produces fluorescence upon cleavage by cathepsin L. Cathepsin L (0.5 nM final) in assay buffer (100 mM sodium acetate, 100 mM NaCl, 1 mM EDTA, 0.01% Brij 35 and 5 mM DTT pH 5.5) was treated with either test compound (0–10 μM) or 5 μL of assay buffer with DMSO (1% final DMSO) for 15 min at RT. Next, the peptide-substrate was added (5 μM final), and the reaction mixture was incubated for 60 min at RT. The total volume of the assay mixture was 20 μL. End-point analysis was performed by recording the fluorescence at excitation/emission wavelengths of 360 nm/460 nm using a TecanSpark instrument. The IC₅₀ values were determined from the GraphPad analysis tool [log­(inhibitor) vs response; variable slope, four parameters]. All measurements were done in duplicate.

Cell Culture

Human ACE2-A549 cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% v/v fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin.

Biosafety and Antiviral Efficacy Assay

All experiments with SARS-CoV-2 were performed in a biosafety level 3 laboratory by personnel equipped with powered air-purifying respirators. The antiviral efficacy assay was performed as previously described. Briefly, A549-hACE2 cells were infected with SARS-CoV-2 at a multiplicity of infection (MOI) of 0.05, and the cells were incubated with the indicated concentrations of inhibitors. The inhibitors were maintained in the medium for the duration of the experiment. At 24 h postinfection (hpi), total RNA was isolated from the cells using TRIzol reagent (Invitrogen), followed by purification with the Direct-zol RNA Miniprep Kit (Zymo Research) according to the manufacturer’s protocols. Equal amounts of RNA were reverse-transcribed into cDNA using the iScript cDNA Synthesis Kit (Bio-Rad). The resulting cDNA was diluted 1:100 and subjected to quantitative PCR (qPCR) using iQ SYBR Green Supermix (Bio-Rad). The mRNA expression level of the SARS-CoV-2 nucleocapsid (N) gene was quantified and normalized to the endogenous reference gene Gapdh. Relative gene expression was calculated using the comparative 2^–ΔΔC(t) method. The specificity of the qPCR amplification was confirmed by melting curve analysis for each sample. The half-maximal inhibitory concentration (IC₅₀) and 90% inhibitory concentration (IC₉₀) values were determined by nonlinear regression analysis using GraphPad Prism software (version 8.0).

Crystallization of M^pro with UM-005 and UM-067

M^pro cocrystallization, data collection, molecular replacement, and structure refinement were performed as previously described. , Briefly, 500 μM UM-005 and UM-067 were mixed with 150 μM M^pro and incubated for 1 h on ice prior to crystallization. Diffraction-quality crystals were obtained with 10–20% w/v PEG 3350, 0.2–0.3 M NaCl, and 0.1 M Bis-Tris methane (pH 5.5) using hanging-drop vapor diffusion at ambient room temperature after 1 week. Data were collected at 100 K at the National Synchrotron Light Source II (NSLS-II) AMX beamline. Diffraction data were processed using XDS. Molecular replacement was performed with Phaser using PDB entry 7L0D as the search model. Structural refinement was carried out using PHENIX Refine, , and model building was performed in COOT. Pseudomerohedral twinning was observed in the UM-005 dataset and was accounted for during refinement using the twin law −h, k, −k – l. Inhibitor geometry optimization was performed with Gauss View 6 using DFT at the B3LYP/6-311++G­(d,p) level. Both crystal structures are available for download in the Protein Data Bank under PDB codes pdb_000011RO (M ^pro –UM-005) and pdb_000012AF (M ^pro –UM-067).

Safety Statement

No unexpected or unusually high safety hazards were encountered.

Glossary

Abbreviations

SAR

structure–activity relationship

hCatL

human cathepsin L

SARS-CoV-2

severe acute respiratory syndrome coronavirus 2

M^pro

main protease

COVID-19

coronavirus disease 2019

nsp

nonstructural protein

PL^pro

papain-like protease

ASU

asymmetric unit

MOI

multiplicity of infection

PDB

Protein Data Bank

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.6c00236.

• Detailed description of the synthesis of the compounds (Schemes S1–S24), comprehensive analytical data, including NMR spectra, HPLC traces, and ESI-MS spectra of the final compounds (Figures S1–S22), and crystallographic data metrics (Table S1 and Figure S23) (PDF)

This work was supported in part by the National Institutes of Health (Grant GM118112).

The authors declare the following competing financial interest(s): K.A.F. and P.R.T. are scientific founders of Danger Bio.