Structure-based lead optimization of peptide-based vinyl methyl ketones as SARS-CoV-2 main protease inhibitors

Abstract

Despite several major achievements in the development of vaccines and antivirals, the fight against SARS-CoV-2 and the health problems accompanying COVID-19 are still ongoing. SARS-CoV-2 main protease (M^pro), an essential viral cysteine protease, is a crucial target for the development of antiviral agents. A virtual screening analysis of in-house cysteine protease inhibitors against SARS-CoV-2 M^pro allowed us to identify two hits (i.e., 1 and 2) bearing a methyl vinyl ketone warhead. Starting from these compounds, we herein report the development of Michael acceptors targeting SARS-CoV-2 M^pro, which differ from each other for the warhead and for the amino acids at the P2 site. The most promising vinyl methyl ketone-containing analogs showed sub-micromolar activity against the viral protease. SPR38, SPR39, and SPR41 were fully characterized, and additional inhibitory properties towards hCatL, which plays a key role in the virus entry into host cells, were observed. SPR39 and SPR41 exhibited single-digit micromolar EC₅₀ values in a SARS-CoV-2 infection model in cell culture.

Graphical abstract

1. Introduction

At the end of 2019, a few cases of a new coronavirus (CoV) were recorded in Wuhan (China) [1]. In February 2020, the International Committee on Taxonomy of Viruses named the newly emerged human pathogen severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which is responsible for the ‘coronavirus disease 19’ (COVID-19) [2,3]. To date (November 2022), over 643 million cases of COVID-19 were reported worldwide, resulting in dramatic consequences in terms of public health, economy, and social life [[4], [5], [6]]. The impressive commitment made by the scientific community led to several vaccines preventing the severe form of the disease and a small set of drugs for inpatient treatment [[7], [8], [9], [10]]. Despite the SARS-CoV-2 pandemic is now better controlled in terms of the number of cases, hospitalization, and deaths, the fight against COVID-19 is still ongoing. The risk of COVID-19 reinfection, the possible onset of new variants resistant to vaccines and drugs, and the effectiveness and durability of both vaccine- and infection-induced antibodies remain relevant concerns for the coming years [[11], [12], [13], [14], [15], [16]].

In the last two years, many articles reported the inhibitory properties of several classes of molecules towards SARS-CoV-2 main protease, also known with the acronyms of SARS-CoV-2 M^pro and 3-CL^pro (i.e., picornavirus 3C-like protease) [[17], [18], [19], [20], [21]]. The enormous efforts led to precious information concerning the minimum requirements for the inhibition, and several crystal structures of SARS-CoV-2 M^pro bound with potent inhibitors were deposited in the Protein Data Bank (PDB) [[22], [23], [24]]. Structurally, SARS-CoV-2 M^pro is a homodimer composed of two protomers that consist of three domains, namely I, II, and III, with the catalytic dyad Cys145-His41 located in a cleft between the domains I and II [25].

Recently, we reported the results of a virtual screening campaign against SARS-CoV-2 M^pro using an our in-house library of peptidic and non-peptidic ligands characterized by different types of electrophilic warheads [26], initially developed as inhibitors of rhodesain, a cysteine protease of Trypanosoma brucei rhodesiense [[27], [28], [29], [30], [31], [32], [33]]. Starting from the N3-SARS-CoV-2 M^pro complex (PDB: 7BQY) [23], the 106 in-house compounds were docked into the binding site of SARS-CoV-2 M^pro. In accordance with the predicted binding free energy values, fifteen of them were selected for biological evaluation [26]. Among them, the dipeptidyl inhibitors 1 and 2 showed IC₅₀ values in the micromolar range towards the viral cysteine protease (Chart 1 ).

Chart 1.

The best inhibitors identified from our in-house database. IC₅₀ values against SARS-CoV-2 M^pro are indicated.

The two compounds share the methyl vinyl ketone warhead, homophenylalanine (hPhe), and an aliphatic amino acid (i.e., cyclohexylalanine (Cha) and Leu) at the P1 and P2 positions, respectively, and a phenyl ring at the P3 site bearing an electron-withdrawing group (EWG) at the para position. Covalent docking-based experiments showed that the methyl vinyl ketone warhead of both dipeptides lends itself well to the formation of a covalent adduct with the catalytic Cys145 (Fig. 1 ).

Fig. 1.

Predicted binding mode of compound 1. The protein is represented as a yellow surface while the ligands as blue sticks.

The CO group could establish an H-bond with the NH group of Gly143 of SARS-CoV-2 M^pro, meanwhile, the peptide backbone of the two inhibitors is involved in a series of H-bonds and van der Waals interactions with the target, as we reported [26]. Docking analysis confirmed the importance of hydrophobic residues at the P2 site, while the aromatic ring at the P3 position fits well in the S3 pocket. Conversely, the hPhe residue lodged in the P1 site did not seem to establish tight interaction with the P1 site, as confirmed by molecular dynamics (MD) simulations [26].

Our compounds mainly differ from the potent SARS-CoV-2 M^pro inhibitors in an aromatic residue being present at the P1 site, which substitutes the Gln residue known to be required for strong binding affinity [21]. We, therefore, decided to design a new series of peptide-based Michael acceptors (SPR35-SPR44) bearing a Gln pentatomic surrogate at the P1 site, since this unnatural amino acid is common in the most potent SARS-CoV-2 M^pro inhibitors (Chart 2 ). A panel of seven aliphatic amino acids was introduced at the P2 position, to explore the effect of different aliphatic side chains on the binding affinity. In particular, linear, branched, and cyclic amino acids were inserted, in agreement with the most promising results reported in the literature [34]. Along with Leu (SPR37), the bulkier tert-butyl alanine (Tba) was introduced (SPR40), as well as the rigid analog tert-Leu (Tle, SPR36). To evaluate the role of the branched side chain of Leu, the linear isomer of Leu and Val, namely Norleucine (Nle, SPR38) and Norvaline (Nva, SPR35), respectively, were inserted. Considering its linear side chain, Nle could fit well into the S2 pocket, while Nva perfectly mimics the side chain length of Leu without ramification. The presence of the cyclic analog of Leu, namely cyclopropylalanine (Cpa), as well as the bulkier Cha, provided inhibitors with great inhibitory properties [24,35], and for this reason, both residues were incorporated in SPR39 and SPR40, respectively. The carbobenzyloxy (Cbz) group was introduced as the N-capping at the P3 position, both because its presence was found to be well-tolerated in our virtual screening analysis (Fig. 1) and is part of potent SARS-CoV-2 M^pro inhibitors, such as GC376, UAWJ9-36-3, and UAWJ9-36-1 [36,37]. Lastly, the methyl vinyl ketone warhead of our lead compounds 1 and 2 was maintained: to our knowledge, no Michael acceptors carrying the methyl vinyl ketone warhead were rationally developed for SARS-CoV-2 M^pro inhibition to date. Validation of this electrophilic moiety as a valuable warhead capable of inhibiting SARS-CoV-2 M^pro was one of the purposes of this structure-activity relationship (SAR) study from the outset. Considering the promising predicted binding free energy values of some methyl vinyl ester warhead-containing compounds in our in-house database [26], this warhead was also evaluated in derivatives carrying the bulkier residues at the P2 position, such as Nle, Tba, and Cha (i.e., SPR42, SPR43, and SPR44, respectively).

Chart 2.

Design of SPR35-SPR44. Abbreviations: Nva, norvaline; Tle, tert-leucine; Leu, leucine; Nle, nor-leucine; Cpa, cyclopropyl alanine; Tba, tert-butyl alanine; Cha, cyclohexylalanine.

Therefore, we herein report the synthesis, biological evaluation, and docking studies of a series of peptide-based Michael acceptors SPR35-SPR44 targeting SARS-CoV-2 M^pro.

2. Results and discussions

2.1. Synthesis of SPR35-SPR44

The novel Michael acceptors SPR35-SPR44 were synthesized in solution following the Boc-chemistry procedures. With this approach, the peptide backbone was initially synthesized from C- to N-terminus, and the warheads were introduced in the last step (Scheme 1 ). Indeed, the Weinreb amide at the C-terminal portion is well suited to be reduced in the corresponding aldehyde, which by Wittig reaction provided the desired final compounds, as we recently reported [38]. The commercially available ester 3 was hydrolyzed in alkaline conditions and the resulting acid 4 was coupled with N,O-dimethylhydroxylamine hydrochloride in the presence of TBTU and NMM as coupling reagent and base, respectively. The use of the most common bases employed in coupling reactions, such as DIPEA and TEA, gave lower yields. Subsequently, the Boc protecting group was removed by treatment with TFA, and the resulting trifluoroacetate 6 was coupled with the N-Cbz amino acids 7a-g. Lastly, the N-Cbz-dipeptides 8a-g were treated with LiAlH₄, and the Wittig reaction between the resulting aldehydes and the appropriate Wittig reagents (i.e., 1-(triphenylphosphoranylidene)-2-propanone and methyl (triphenylphosphoranylidene)acetate for ketone and ester warheads, respectively) gave the final compounds SPR35-SPR44.

Scheme 1.

Reagents and conditions: i) NaOH, MeOH, H₂O, −5 °C–0 °C, 1 h; ii) TBTU, NMM, 0 °C, 30 min, then N,O-dimethylhydroxylamine hydrochloride, rt, on; iii) 30% TFA in DCM, 0 °C–rt, TLC monitoring; iv) 7a-g, TBTU, NMM, 0 °C, 30 min, then 6, rt, on; v) LiAlH₄, dry THF, −10 °C, TLC monitoring; vi) DCM, appropriate Wittig reagent, rt, 2 h.

2.2. Biological investigation

After an initial screening for the inhibitory activity against SARS-CoV-2 M^pro at 20 μM, K _i values were determined for compounds with a percentage of inhibition >90% (Table 1 ). Nirmatrelvir and SARS-CoV-2 M^pro inhibitor 11a [24] were used as positive controls, and DMSO was used as negative control (solvent control). The K _i values of 3.15 ± 0.42 nM and 8.32 ± 0.90 nM obtained for Nirmatrelvir and 11a, respectively, were in agreement with those reported in the literature [24,39], which proved the accuracy of these biological assays. With the exception of SPR35 and SPR36, the vinyl ketone derivatives SPR37–SPR41 showed K _i values in the sub-micromolar range against SARS-CoV-2 M^pro, ranging from 0.184 μM to 0.416 μM. In particular, the Cha-analogs SPR41 exhibited the best K _i value towards the target, meanwhile, SPR38 and SPR39, which carry Nle and Cpa, respectively, exhibited comparable binding affinity. The presence of Leu and Tba (i.e., SPR37 and SPR40, respectively) led to a slight loss of affinity. The similar activities shown by SPR38 and SPR41, which carry two deeply different side chains in terms of the steric hindrance (i.e., linear and cyclic, respectively) at the P2 site, suggest a limited specificity towards bulky aliphatic amino acids at the S2 pocket. On the other hand, the Nva-containing analog SPR35 showed a higher K _i value that lies in the low micromolar range, while the Tle derivative SPR36 exhibited only a low percentage of inhibition at the screening concentration. It is interesting to note as the reduction of the side chain size negatively influences the inhibitory properties towards SARS-CoV-2 M^pro: in fact, Nva derivative SPR35 exhibited a binding affinity almost 7-fold lower with respect to the Nle analog SPR38, which only differs for the length of the side chain (n-propyl vs n-butyl, respectively). Furthermore, an increased loss of affinity was observed in the presence of a rigid tert-butyl side chain of SPR36. All in all, the incorporation of aliphatic amino acids with bulky and hindering side chains at the P2 site, irrespective of whether they are linear, cyclic, or branched, led to inhibitors with sub-micromolar binding affinities towards SARS-CoV-2 M^pro. In contrast, the presence of small or constrained side chains, incorporated both in SPR35 and SPR36, resulted in loss or lack of activity, indicating that Nva and Tle do not fit well into the S2 pocket.

Table 1.

Biological evaluation of novel SPRs towards SARS-CoV-2 M^pro and selectivity assessment.

Compounds	EWG	AA	Ki (μM) or % of inhibition at 20 μM
			SARS-CoV-2 Mpro	hCatL	hCatB
SPR35		Nva	1.77 ± 0.16	–	–
SPR36		Tle	33 ± 2%	–	–
SPR37		Leu	0.386 ± 0.055	–	–
SPR38		Nle	0.260 ± 0.066	1.92 ± 0.10	11.1 ± 1.2
SPR39		Cpa	0.252 ± 0.028	3.38 ± 0.20	7.88 ± 0.65
SPR40		Tba	0.416 ± 0.058	–	–
SPR41		Cha	0.184 ± 0.025	0.252 ± 0.018	14.4 ± 1.2
SPR42		Nle	29 ± 4%	–	–
SPR43		Tba	50 ± 5%	–	–
SPR44		Cha	47 ± 2%	–	–
Nirmatrelvir	–	–	0.003 ± 0.0004	–	–
11a	–	–	0.008 ± 0.0009	–	–

Completely different results were obtained with the vinyl methyl ester derivatives SPR42–SPR44: none of the three analogs showed significant inhibition at the screening concentration, which suggests that the vinyl methyl ester warhead poorly reacts with the catalytic cysteine. The ester analogs SPR42–SPR44 share the binding site recognition with potent inhibitors SPR38–SPR39 and SPR41, thus, the lack of activity is clearly related to the different warheads.

Considering the appreciable inhibitory properties of vinyl methyl ketone derivatives, the most promising Michael acceptors were selected for the biological evaluation towards a panel of cysteine proteases. SPR38, SPR39, and SPR41 showed no inhibition at 100 μM against SARS-CoV-2 papain-like protease (PL^pro), which plays a crucial role in processing viral polyproteins [40]. This result was expected as the peptide-based recognition moiety of the most potent PL^pro inhibitors incorporates different amino acids [41]. Similarly, no inhibitory activity against dengue virus NS2B/NS3 serine protease was observed at 100 μM. Regarding to human cathepsins (hCats), the three selected compounds were assayed towards hCatL and hCatB: both human cysteine proteases resulted to be sensitive to SPRs inhibition in the micromolar range, meanwhile, SPR41 exhibited hCatL inhibition in the sub-micromolar range. As well-known, hCatL mediates the cleavage of SARS-CoV spike protein, which is necessary for the endosomal entry route of the virus into host cells [42,43]. Similarly, hCatL plays the same role in the SARS-CoV-2 infection, enhancing the virus entry [[42], [43], [44], [45]]. In COVID-19, the plasma levels of hCatL resulted to be higher in patients with severe disease with respect to the ones with non-severe form [42]. Hence, high plasma levels of circulating hCatL were found to be directly correlated with the disease progression and its severity.

The antiviral effect of hCatL inhibitors was widely reported in the literature [[46], [47], [48]], meanwhile, dual inhibition of SARS-CoV-2 M^pro and hCatL could provide a synergistic antiviral effect in vivo [49,50]. Simultaneous inhibition of two valid targets involved in two different cellular and/or viral pathways could provide huge advantages compared to the antiviral agents directed against a single target, as widely reported in the literature in several research fields of drug discovery and medicinal chemistry [[51], [52], [53], [54]]. In light of all these data, the inhibitory properties of SPR38–SPR39 and SPR41 towards hCatL could lead to positive effects in terms of antiviral activity in vivo, similarly to the potent dual inhibitor MPI8 [49]. Lastly, the three assayed Michael acceptors showed moderate inhibition against hCatB: whilst SPR38 and SPR39 exhibited K _i values almost 2- and 6-fold higher compared to those observed for the hCatL inhibition, SPR41 resulted to be mildly selective towards hCatL, with a selectivity index (SI = K _i hCatB/K _i hCatL) of 57.

To test the nature of the ligand's mode of inhibition of M^pro and hCatL, dilution assays were performed as described previously [55,56]. Solutions with inhibitor concentrations of 5 x IC₅₀ were prepared to ensure potent inhibition. After incubation, samples were diluted to achieve inhibitor concentrations of 0.1 x IC₅₀. Enzyme activities were recorded before and after dilution (Fig. 2 ). In the case of non-covalent or reversible-covalent binding, the enzyme activity should recover after dilution. Since neither M^pro nor hCatL showed recovering enzyme activities for none of the tested compounds, we propose a covalent irreversible binding mode. Based on these findings, we sought to determine the kinetic parameters for this mechanism of inhibition. Despite the results, indicating a covalent irreversible binding mode, time-dependent measurements did not show a curvature significant enough to be deconvoluted to the kinetic parameters K _i, k _inact, and k _2nd. (Fig. S1). Therefore, we hypothesized, that this binding mechanism depends on a bi-phasic mechanism of inhibition as previously published for dipeptidyl enoates [57]. In this case, high inhibitor concentrations led to an irreversible inhibition, whereas lower concentrations yielded the observation of a reversible binding mode.

Fig. 2.

A) Dilution assay with SPR38 (blue), SPR39 (green), SPR41 (purple), and a DMSO (white) treated reference against SARS-CoV-2 M^pro. Bar chart of the relative enzyme activities, demonstrating no significant activity recovering after dilution. B) Dilution assay with SPR38 (blue), SPR39 (green), SPR41 (purple), and a DMSO (white) treated reference against hCatL. Bar chart of the relative enzyme activities, demonstrating no significant recovering activity after dilution.

Our experimental setup of dilution assays is not suitable for Nirmatrelvir as a reversible control substance. Since it exhibits a low IC₅₀ value, the resulting concentrations of Nirmatrelvir would be way lower than the enzyme concentration. Therefore, to further demonstrate the covalent irreversible reaction mode of SPRs, we performed a dialysis assay as previously described [55,56]. To do so, M^pro was incubated for 60 min with inhibitor concentrations of 10-fold the IC₅₀ value for the most potent inhibitors SPR39, SPR41, and Nirmatrelvir as a reversible control substance and performed dialysis for 25 h. Enzyme activities of samples drawn at several time points were recorded (Fig. 3 ). Due to its high affinity and covalent reversible binding mode, Nirmatrelvir was dialyzed very slowly, so that after 25 h of dialysis, only 50% of enzymatic activity was recovered. However, dialysis of SPR39 and SPR41 did not show recovering enzymatic activities, further supporting their covalent irreversible binding mode.

Fig. 3.

Dialysis assay of SPR39, SPR41, and Nirmatrelvir, the latter used as a covalent reversible control substance. Means of relative activities are depicted as bar chart, errors of technical duplicates are indicated.

Lastly, the antiviral activities of the three selected M^pro inhibitors were assessed using Huh-7-ACE2 cells infected with SARS-CoV-2. Except for SPR38, which was proven to be unstable and toxic, SPR39 and SPR41 exhibited single-digit micromolar EC₅₀ values when SARS-CoV-2 replication was tested in cell culture so that the observed antiviral activity is consistent with the M^pro inhibition (Table 2 ).

Table 2.

Antiviral and cytotoxicity evaluation of the most promising SARS-CoV-2 M^pro inhibitors.

Cmpd	Huh-7-ACE2 cells infected EC50 (μM)	Huh-7-ACE2 cells CC50 (μM)	SI
SPR38	18.5 ± 6.5	60.9 ± 11.5	3
SPR39	1.5 ± 0.3	100	66.6
SPR41	1.8 ± 0.1	14.5 ± 3.4	8
Nirmatrelvir	<0.01	>100	>10.000

As can be noted, the antiviral activities of the three assessed compounds were not in perfect correlation with the SARS-CoV-2 M^pro inhibition. Despite that, our results are comparable with those reported in several SAR studies aimed at the development of novel anti-COVID-19 agents [20,58,59]. The discrepancy between the enzymatic inhibitory properties and cellular effects is usually ascribed to the low permeability of cell membranes. The design of suitable carrier-linked prodrugs and the incorporation into liposomes or nanocarriers could improve the membrane crossing, resulting in increased concentrations of inhibitors within cells, where SARS-CoV-2 M^pro is located [60,61].

Lastly, the cytotoxicity towards the same uninfected cell line was evaluated: SPR39, which carries Cpa at the P2 site, exhibited a CC₅₀ value of 100 μM, resulting in a SI of 66.6, whereas the Cha-containing analog SPR41 was shown to be more cytotoxic.

2.3. Docking studies

Molecular modeling studies were indeed instrumental to identify the lead compounds 1 and 2 as SARS-CoV-2 M^pro inhibitors [26] that were subjected to a first round of optimization in the present study. Here, the same theoretical procedure was also used to provide at molecular level information on the binding pose of the most interesting compound and rationalize the obtained structure-activity relationship (SAR) data. In particular, the covalent docking protocol available with the AutoDock4 (AD4) docking software named “flexible side-chain method” [62] was employed to predict the binding pose of SPR39 in the M^pro X-ray structure having PDB code 7BQY [23]. This compound was selected considering its promising biological activity in in vitro and cell-based experiments. As expected, the calculated binding position strongly resembles to one already reported for its close lead analog 6. More precisely, the predicted covalent adduct with the enzyme Cys145 residue allows to form an additional H-bond with the backbone NH of Gly143 (Fig. 4 ). The ligand Gln pentatomic surrogate at the P1 site is lodged in the S1 pocket forming a double H-bond with His163 and Glu166 just like the N3 peptide co-crystallized in the 7BQY structure. On the other hand, the P2 Cpa residue is well inserted in the S2 pocket establishing favorable hydrophobic interactions with the side chains of His41, Met49, Met165, and Asp187. In this position, different chemically related substituents were explored (Chart 1) resulting in comparable potencies (Table 1) except for the Tle one, that proved to be detrimental for enzyme inhibition. In this case, it is possible to postulate that the intrinsic rigidity of the Tle side chain might negatively influence the ligand/enzyme recognition or formation of the covalent adduct. The P3 side chain is lodged in a cleft made up of Met165, Leu167, Pro168, Gln189, and Gln192. Here the terminal phenyl ring might establish hydrophobic contacts with the aforementioned residues as well as π-stacking interactions with the surrounding π-faces of the backbone amides.

Fig. 4.

Predicted binding mode of SPR39 into the X-ray SARS-CoV-2 M^pro structure. The protein is represented as yellow surface and white sticks, and the ligand as green sticks. H-bond interactions are represented as dashed yellow lines.

To probe the stability of these interactions molecular dynamics (MD) simulations were also attained of the AD4 SPR39-SARS-CoV-2 M^pro predicted complex. This structure was subjected to a 100 ns long MD simulation and results analyzed by examining the ligand root mean square deviations and fluctuations (L-rmsd and L-rmsf, respectively) to profile the modifications in the ligand atom positions.

As reported in Fig. 5 A, the AD4 predicted binding mode is very stable throughout the entire production run. Analysis of the main ligand fluctuations broken by atom demonstrates that the P3 site is the most flexible one while the rest of the molecule, including the P2 site, is adopting a stable conformation (Fig. 5B). Interestingly, the same site in the parent compound 6 was experiencing a high degree of flexibility in previously run MD simulations [26] thereby further underscoring the viability of our design hypothesis. The reason for this stable binding conformation can be ascribed to the stability of the ligand-enzyme interactions as evidenced when plotting the interaction fraction of the protein-ligand contacts throughout the simulation (Fig. 5C and D).

Fig. 5.

(A) rmsd (Å) plot of SPR39 over time (ns). (B) L-rmsf plot broken down by atom corresponding to the reported two-dimensional structure of SPR39. (C and D) Protein interactions of SPR39 throughout the simulation categorized into H-bonds, hydrophobic, ionic, and water bridges.

3. Conclusion

In this paper, we report the development of novel peptide-based Michael acceptors targeting SARS-CoV-2 M^pro. With a few exceptions, analog bearing a vinyl methyl ketone warhead showed K _i values in the sub-micromolar range against the target (SPR38–SPR41), while the vinyl methyl ester derivatives SPR42–SPR44 did not inhibit the viral cysteine protease. These data highlight the key role played by the vinyl methyl ketone warhead, which strongly reacts with the catalytic Cys of SARS-CoV-2 M^pro. The side chain size of the aliphatic amino acids at the P2 site was found to be crucial for the binding affinity: analogs containing relatively bulky amino acids, such as Cha, Nle, Cpa, and Tba proved to be more potent compared to derivatives carrying small and rigid side chains at this position. Interestingly, appreciable inhibitory properties against hCatL were observed for SPR38, SPR39, and SPR41. Considering the key role played by hCatL in the viral entry into cells, the dual inhibition of hCatL and SARS-CoV-2 M^pro could lead to a synergistic antiviral effect in vivo. The three fully characterized vinyl methyl ketone Michael acceptors showed a bi-phasic mechanism of inhibition towards both SARS-CoV-2 M^pro and hCatL: in fact, whilst reversible binding mode was detected at low inhibitor concentration at a 30 min timescale, irreversible inhibition at higher concentrations was observed. Lastly, SPR39 and SPR41 showed single-digit micromolar antiviral activity, and SPR39 exhibited moderate selectivity towards infected cells (SI = 66.6). All in all, peptide-based Michael acceptors bearing a vinyl methyl ketone warhead and bulky aliphatic amino acids at the P2 site were found to strongly inhibit SARS-CoV-2 M^pro, and an antiviral effect in the low micromolar range was observed. Based on the promising findings described herein, further SAR studies aiming to improve binding affinity, antiviral activity, and cytotoxicity will be carried out.

4. Experimental section

4.1. Chemistry

Reagents and solvents were purchased from several commercial suppliers. TBTU, N,O-dimethylhydroxylamine hydrochloride, TFA, and Cbz-amino acids were obtained from Fluorochem. N-Methyl morpholine were purchased from VWR. TFA, LiAlH₄, methyl (triphenylphosphoranylidene) acetate, and methyl (triphenylphosphoranylidene)-2-propanone were obtained from Merck, as well as silica gel 60 F254 plates and silica gel (200–400 mesh) employed for TLC and column chromatography, respectively. All the TLC were treated with an ethanol solution of phosphomolybdic acid hydrate (15%), which was obtained by Fluorochem, meanwhile, TLC in which aldehydes were evaluated, were treated with 2,4-dinitrophenylhydrazine TLC stain. All the ¹H and ¹³C spectra were performed on a Varian 500 MHz, operating at 499.74 and 125.73 MHz for ¹H and ¹³C, respectively. Deuterated solvents, namely CDCl₃ and MeOD, were obtained from Merck and the signal of the solvents was used as the internal standard. Splitting patterns are described as singlet (s), doublet (d), doublet of doublet (dd), triplet (t), quartet (q), multiplet (m), and broad singlet (bs). Chemical shifts are expressed in ppm and coupling constants (J) in Hz. Elemental analyses were performed on a C. Erba model 1106 (elemental analyzer for C, H, and N) apparatus, and ±0.4% of the theoretical values were found.

4.1.1. (S)-2-((tert-Butoxycarbonyl)amino)-3-((S)-2-oxopyrrolidin-3-yl)propanoic acid (4)

In a round bottom flask (A), the commercially available Boc-cGln-OMe 3 (1 eq.) was dissolve in MeOH (0.7 mL/mmol) and cooled down up to −5 °C with ice/salt bath. Meanwhile, an aqueous solution of NaOH (4 eq., same volume of MeOH) was prepared, cooled down and added dropwise to the flask A over 10 min, keeping the temperature below 0 °C. The reaction was maintained in stirring and the temperature was constantly monitored (not more than 2–3 °C). After 1 h, TLC monitoring (EtOAc/light petroleum, 7:3. R _f starting material: 0.30 in this mixture) showed the disappearance of the starting material, the pH was neutralized up to 7 with 1 M HCl and methanol was removed in vacuo. Subsequently, 1 M HCl was added up to pH 1, and the organic phase was extracted with EtOAc (x 3), washed with brine (x 3), dried over Na₂SO₄ and concentrated in vacuo. The resulting residue was used for the next step without further purification. Consistency = white foamy powder. R _f = 0.0 in EtOAc/light petroleum (7:3). Yield: 74%. ¹H NMR (500 MHz) in CDCl₃, δ = 1.44 (s, 9H), 1.83–1.97 (m, 2H), 2.13–2.24 (m, 1H), 2.38–2.49 (m, 1H), 2.57–2.66 (m, 1H), 3.34–3.46 (m, 2H), 4.33–4.41 (m, 1H), 5.69 (d, J = 7.6 Hz, 1H), 6.98 (s, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 27.74, 28.31, 33.79, 37.98, 41.00, 52.05, 80.05, 155.79, 174.82, 181.17. NMR data are in agreement with those already reported in the literature [63].

4.1.2. tert-Butyl ((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)carbamate (5)

In a round-bottom flask, the acid 4 (1 eq.) was dissolved in DCM (10 mL/mmol) and stirred at 0 °C. TBTU (1.2 eq.) and NMM (2 eq.) were added, and the reaction was maintained in stirring for 30 min at 0 °C. After this time, N,O-dimethylhydroxylamine hydrochloride (1.1 eq.) was added portion-wise, and the pH was checked (>8). The reaction was vigorously stirred at rt on. After this time, DCM vas removed in vacuo, and the resulting residue was dissolved in EtOAc washed with 1 M HCl (x 2), NaHCO₃ saturated solution (x 2), and brine (x 2), dried over Na₂SO₄, and concentrated. The crude was purified using EtOAc/MeOH 9:1 as the eluent mixture. Consistency = pale yellow poder; Yield = 86%; R _f = 0.62 in EtOAc/MeOH 9:1.¹H NMR (500 MHz) in CDCl₃, δ = 1.43 (s, 9H, Boc), 1.63–1.71 (m, 1H), 1.78–1.88 (m, 1H), 2.10 (t, J = 11.1 Hz, 2H), 2.44–2.55 (m, 2H), 3.21 (s, 3H), 3.31–3.36 (m, 2H), 3.79 (s, 3H), 4.67 (t, J = 8.3 Hz, 1H), 5.48 (d, J = 8.8 Hz, 1H), 6.44 (s, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 28.07, 28.48, 32.37, 34.45, 38.13, 40.42, 49.44, 61.74, 79.75, 155.91, 172.69, 179.95. NMR data are in agreement with those already reported in the literature [63].

4.1.3. (S)-1-(Methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-aminium 2,2,2-trifluoroacetate (6)

In a round-bottom flask, intermediate 5 was dissolved in DCM (5 mL/mmol) at 0 °C and an equivalent volume of TFA was added dropwise and the resulting solution was vigorously stirred. The reaction was monitored through TLC (eluent mixture EtOAc/MeOH 9:1) and maintained in stirring until the disappearance of the starting material (around 1 h). After that, DCM was easily removed in vacuo, and the resulting suspension was further evaporated with toluene, chloroform, and diethyl ether. The obtained white powder was used for the next step without purification. Consistency = white powder; Yield = 93%; R _f = 0.0 in EtOAc/MeOH 9:1.¹H NMR (500 MHz) in MeOD, δ = 1.79–1.89 (m, 1H), 1.89–1.98 (m, 1H), 2.04 (ddd, J = 15.0, 5.0, 2.8 Hz, 1H), 2.43–2.36 (m, 1H), 2.74–2.83 (m, 1H), 3.26 (s, 3H), 3.36–3.41 (m, 2H), 3.82 (s, 3H), 4.40 (dd, J = 9.5, 2.8 Hz, 1H). ¹³C NMR (125 MHz) in MeOD, δ = 29.58, 32.59, 33.24, 41.92, 42.06, 52.45, 62.38, 161.26, 181.62.

4.1.4. General procedure for the synthesis of intermediates 8a-g

In a round bottom flask (A), the commercially available Cbz-amino acids 7a-g (1.5 eq.) were dissolved in DCM (10 mL/mmol) and cooled down up to 0 °C with an ice bath. TBTU (1.5 eq.) and NMM (2 eq.) were added, and the reaction was kept in vigorously stirring for 30 min. Meanwhile, the trifluoroacetate salt 6 (1 eq.) was suspended in DCM (10 mL/mmol) and NMM (2eq.) was added at 0 °C. The pH was checked (>8) and the resulting solution was added dropwise to the flask A. After that, the reaction was left in stirring at rt on. Subsequently, DCM vas removed in vacuo, and the residue was dissolved in EtOAc and washed with 1 M HCl (x 2), NaHCO₃ saturated solution (x 2), and brine (x 2), dried over Na₂SO₄, and concentrated in vacuo. The obtained crude was purified using the appropriate eluent mixture below described.

4.1.4.1. Benzyl ((S)-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-1-oxopentan-2-yl)carbamate (8a)

In this reaction Cbz-Nva-OH 7a was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 71%; R _f = 0.24 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.80–0.93 (m, 3H), 1.26–1.43 (m, 2H), 1.49–1.84 (m, 4H), 2.05–2.20 (m, 1H), 2.26–2.48 (m, 2H), 31.7 (s, 3H), 3.18–3.28 (m, 2H), 3.78 (s, 3H), 4.21–4.32 (m, 1H), 4.79–4.91 (m, 1H), 5.05 (d, J = 12.1 Hz, 1H), 5.08 (d, J = 12.0 Hz, 1H), 5.74 (d, J = 8.5 Hz, 1H), 6.77 (d, J = 11.2 Hz, 1H), 7.18–7.36 (m, 5H), 7.72 (d, J = 7.5 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 13.85, 18.74, 28.18, 33.23, 35.50, 38.40, 40.48, 48.59, 53.54, 54.76, 61.62, 66.84, 128.06, 128.10, 128.53, 136.47, 156.09, 171.83, 172.46, 180.02.

4.1.4.2. Benzyl ((S)-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-3,3-dimethyl-1-oxobutan-2-yl)carbamate (8b)

In this reaction Cbz-Tle-OH 7b was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 60%; R _f = 0.33 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 1.01 (s, 9H), 2.10–2.21 (m, 1H), 2.30–2.48 (m, 2H), 3.19 (s, 3H), 3.21–3.33 (m, 2H), 3.80 (s, 3H), 4.12 (d, J = 9.5 Hz, 1H), 4.74–4.84 (m, 1H), 5.06 (d, J = 12.4 Hz, 1H), 5.09 (d, J = 12.3 Hz, 1H), 5.67 (d, J = 9.5 Hz, 1H), 6.68 (s, 1H), 7.27–7.36 (m, 5H), 7.87 (d, J = 6.3 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 26.66, 28.46, 32.43, 32.79, 35.10, 38.71, 40.57, 49.36, 61.67, 62.63, 66.89, 128.08, 128.15, 128.57, 136.56, 156.38, 171.02, 171.81, 180.01.

4.1.4.3. Benzyl ((S)-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-4-methyl-1-oxopentan-2-yl)carbamate (8c)

In this reaction Cbz-Leu-OH 7c was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 76%; R _f = 0.45 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.86 (t, J = 6.0 Hz, 6H), 1.40–1.51 (m, 1H), 1.52–1.75 (m, 4H), 2.03–2.14 (m, 1H), 2.22–2.43 (m, 2H), 3.12 (s, 3H), 3.14–3.23 (m, 2H), 3.72 (s, 3H), 4.22–4.30 (m, 1H), 4.78–4.86 (m, 1H), 4.99 (d, J = 12.6 Hz, 1H), 5.03 (d, J = 11.9 Hz, 1H), 5.66 (d, J = 8.6 Hz, 1H), 6.86 (s, 1H), 7.16–7.30 (m, 5H), 7.66 (d, J = 7.6 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 21.88, 23.09, 24.68, 28.11, 32.25, 33.26, 38.36, 40.49, 42.28, 48.48, 53.51, 61.58, 66.81, 127.99, 128.05, 128.48, 136.47, 156.11, 171.83, 172.92, 180.11.

4.1.4.4. Benzyl ((S)-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-1-oxohexan-2-yl)carbamate (8d)

In this reaction Cbz-Nle-OH 7d was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 79%; R _f = 0.36 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.80 (t, J = 6.6 Hz, 3H), 1.20–1.30 (m, 4H), 1.50–1.59 (m, 1H), 1.60–1.68 (m, 1H), 1.68–1.80 (m, 2H), 2.05–2.14 (m, 1H), 2.26–2.42 (m, 2H), 3.12 (s, 3H), 3.73 (s, 3H), 4.17–4.24 (m, 1H), 4.79–4.87 (m, 1H), 5.00 (d, J = 12.6 Hz, 1H), 5.04 (d, J = 12.1 Hz, 1H), 5.65 (d, J = 8.4 Hz, 1H), 6.64 (s, 1H), 7.17–7.31 (m, 5H), 7.59 (d, J = 7.6 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 14.00, 22.43, 27.56, 28.23, 33.06, 33.31, 38.39, 40.49, 48.59, 53.53, 55.00, 61.64, 66.87, 128.07, 128.11, 128.54, 136.52, 156.11, 171.86, 172.46, 180.03.

4.1.4.5. Benzyl ((S)-3-cyclopropyl-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-1-oxopropan-2-yl)carbamate (8e)

In this reaction Cbz-Cpa-OH 7e was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 77%; R _f = 0.35 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.01–0.11 (m, 2H), 0.32–0.48 (m, 2H), 0.65–0.79 (m, 1H), 1.51–1.82 (m, 4H), 2.05–2.20 (m, 1H), 2.25–2.49 (m, 2H), 3.07–3.28 (m, 2H), 3.17 (s, 3H), 3.78 (s, 3H), 4.27–4.39 (m, 1H), 4.80–4.94 (m, 1H), 5.06 (d, J = 13.1 Hz, 1H), 5.09 (d, J = 12.3 Hz, 1H), 5.79 (d, J = 8.0 Hz, 1H), 6.67–6.75 (bs, 1H), 7.20–7.36 (m, 5H), 7.65 (d, J = 7.3 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 4.39, 4.51, 7.31, 28.25, 33.44, 38.11, 38.38, 40.47, 48.57, 53.53, 55.50, 61.65, 66.82, 127.76, 128.07, 128.53, 136.55, 155.97, 171.80, 172.17, 180.00.

4.1.4.6. Benzyl ((S)-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-4,4-dimethyl-1-oxopentan-2-yl)carbamate (8f)

In this reaction Cbz-Tba-OH 7f was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 73%; R _f = 0.34 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.80–0.97 (s, 9H), 1.37–1.50 (m, 1H), 1.60–1.84 (m, 2H), 2.05–2.18 (m, 1H), 2.23–2.47 (m, 2H), 2.67–2.75 (m, 1H), 3.15 (s, 3H), 3.14–3.35 (m, 2H), 3.75 (s, 3H), 4.22–4.35 (m, 1H), 4.76–4.91 (m, 1H), 5.03 (d, J = 12.3 Hz, 1H), 5.08 (d, J = 11.8 Hz, 1H), 5.73 (d, J = 8.8 Hz, 1H), 6.92 (bs, 1H), 7.12–7.33 (m, 5H), 7.61 (d, J = 7.6 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 28.09, 29.71, 30.57, 33.39, 38.31, 38.62, 40.44, 46.35, 48.38, 52.97, 61.56, 66.80, 127.93, 128.02, 128.46, 136.50, 155.80, 171.79, 173.37, 180.02.

4.1.4.7. Benzyl ((S)-3-cyclohexyl-1-(((S)-1-(methoxy(methyl)amino)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl)propan-2-yl)amino)-1-oxopropan-2-yl)carbamate (8g)

In this reaction Cbz-Cha-OH 7g was used as the acid. Eluent mixture: EtOAc/MeOH, 19:1. Consistency = pale yellow solid; Yield = 83%; R _f = 0.40 in EtOAc/MeOH 19:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.76–1.00 (m, 2H), 1.03–1.26 (m, 3H), 1.26–1.40 (m, 1H), 1.41–1.52 (m, 1H), 1.53–1.85 (m, 7H), 2.06–2.23 (m, 2H), 2.29–2.49 (m, 2H), 3.18 (s, 3H), 3.20–3.29 (m, 2H), 3.78 (s, 3H), 4.22–4.37 (m, 1H), 4.81–4.94 (m, 1H), 5.05 (d, J = 12.5 Hz, 1H), 5.12 (d, J = 12.0 Hz, 1H), 5.54 (d, J = 8.3 Hz, 1H), 6.51 (s, 1H), 7.18–7.39 (m, 5H), 7.54 (d, J = 7.8 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 26.14, 26.32, 26.51, 28.27, 32.43, 33.39, 33.85, 34.03, 38.36, 38.69, 40.49, 40.79, 48.51, 52.92, 61.65, 66.89, 128.05, 128.13, 128.56, 136.53, 156.13, 171.84, 172.99, 180.00.

4.1.5. General procedure for the warhead incorporation in SPR35-SPR44

In a round-bottom flask, the intermediate 8a–g (1 eq.) was solubilized in dry THF (20 mL/mmol), cooled down to −10 °C with an ice/salt bath, and vigorously stirred. LiAlH₄ (1 eq.) was added each 30 min until the TLC monitoring (DCM/MeOH 19:1) did not show the presence of starting material (usually, 2 or 3 eq. of LiAlH₄ were added). Cold temperature was kept. Aldehydes were detected by the treatment of the TLC with 2,4-dinitrophenylhydrazine TLC stain. Afterwards, the unreacted LiAlH₄ was quenched with 1 M KHSO₄ and the ice-bath was removed. The suspension was moved in a separatory funnel, DCM was added, and the two phases were separated. The organic phase was further extracted with DCM (x 3). Subsequently, the merged organic phases were washed with NaHCO₃ saturated solution (x 2), 1 M KHSO₄ (x 2) and brine (x 2), dried over Na₂SO₄ and concentrated. The obtained residues were used for the next step without further purification. In a round-bottom flask, the aldehydes were solubilized in DCM (5 mL/mmol) and the appropriate Wittig reagent (1 eq.) was added in one portion. The reaction was stirred at rt for 2 h. After that, the solvent was removed in vacuo and the desired products were purified by column chromatography using the appropriate eluent below described.

4.1.5.1. Benzyl ((S)-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)pentan-2-yl)carbamate (SPR35)

In this reaction 8a and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 41% (two steps). R _f = 0.30 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.89–0.97 (m, 3H), 1.30–1.44 (m, 2H), 1.55–1.70 (m, 2H), 1.72–1.87 (m, 2H), 1.99–2.11 (m, 1H), 2.22 (s, 3H), 2.29–2.39 (m, 1H), 2.39–2.48 (m, 1H), 3.21–3.35 (m, 2H), 4.28–4.38 (m, 1H), 4.53–4.61 (m, 1H), 5.09 (s, 2H), 5.64 (d, J = 8.7 Hz, 1H), 6.16 (d, J = 15.9 Hz, 1H), 6.41 (s, 1H), 6.65 (dd, J = 16.0, 6.0 Hz, 1H), 7.26–7.36 (m, 5H), 7.93 (d, J = 7.4 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 13.88, 18.90, 27.66, 28.66, 35.08, 35.71, 38.64, 40.66, 49.44, 54.92, 67.01, 128.08, 128.26, 128.65, 129.88, 136.49, 146.26, 156.25, 172.53, 180.31, 198.33. Elemental analysis for C₂₃H₃₁N₃O₅, calculated: C, 64.32; H, 7.27; N, 9.78; found: C, 63.98; H, 7.33; N, 9.86.

4.1.5.2. Benzyl ((S)-3,3-dimethyl-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)butan-2-yl)carbamate (SPR36)

In this reaction 8b and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. In this reaction methyl (triphenylphosphoranylidene)-2-propanone was used as the Wittig reagent. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 27% (two steps). R _f = 0.25 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.93 (s, 9H), 1.43–1.55 (m, 1H), 1.63–1.74 (m, 1H), 2.04–2.13 (m, 1H), 2.13–2.20 (m, 1H), 2.16 (s, 3H), 2.20–2.38 (m, 2H), 4.12 (d, J = 9.7 Hz, 1H), 4.39–4.50 (m 1H), 5.02 (s, 1H), 5.63 (d, J = 9.7 Hz, 1H), 6.15 (d, J = 16.2 Hz, 1H), 6.59 (dd, J = 16.0, 6.2 Hz, 1H), 6.63 (s, 1H), 7.21–7.32 (m, 5H), 7.99 (d, J = 6.9 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 26.81, 27.50, 28.46, 29.82, 34.99, 38.45, 40.65, 49.50, 62.61, 67.00, 127.99, 128.26, 128.65, 130.40, 136.49, 146.32, 156.54, 171.14, 180.17, 198.43. Elemental analysis for C₂₄H₃₃N₃O₅, calculated: C, 64.99; H, 7.50; N, 9.47; found: C, 65.24; H, 7.67; N, 9.34.

4.1.5.3. Benzyl ((S)-4-methyl-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)pentan-2-yl)carbamate (SPR37)

In this reaction 8c and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 31% (two steps). R _f = 0.30 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.94 (s, 3H), 0.96 (s, 3H), 1.44–1.59 (m, 1H), 1.59–1.74 (m, 2H), 1.73–1.85 (m, 2H), 1.99–2.09 (m, 1H), 2.23 (s, 3H), 2.31–2.50 (m, 2H), 4.30–4.39 (m,1H), 4.49–4.59 (m, 1H), 5.10 (s, 2H), 5.47 (d, J = 8.7 Hz, 1H), 6.16 (d, J = 15.8 Hz, 1H), 6.24 (s, 1H), 6.65 (dd, J = 16.0, 6.7 Hz, 2H), 7.27–7.35 (m, 5H), 7.94 (d, J = 7.4 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 22.07, 23.17, 24.97, 27.68, 28.78, 35.07, 38.64, 40.67, 42.62, 49.57, 53.74, 67.06, 128.06, 128.27, 128.66, 129.92, 136.49, 146.24, 156.28, 172.97, 180.29, 198.36. Elemental analysis for C₂₄H₃₃N₃O₅, calculated: C, 64.99; H, 7.50; N, 9.47; found: C, 65.17; H, 7.41; N, 9.30.

4.1.5.4. Benzyl ((S)-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)hexan-2-yl)carbamate (SPR38)

In this reaction 8d and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 19% (two steps). R _f = 0.27 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.86–0.91 (m, 3H), 1.29–1.35 (m, 3H), 1.37–1.50 (m, 1H), 1.58–1.69 (m, 1H), 1.72–1.86 (m, 4H), 1.99–2.06 (m, 1H), 2.22 (s, 3H), 2.31–2.48 (m, 1H), 3.26–3.32 (m, 2H), 4.25–4.33 (m, 1H), 4.51–4.61 (m, 1H), 5.10 (s, 2H), 5.57 (d, J = 8.2 Hz, 1H), 6.16 (d, J = 15.9 Hz, 1H), 6.19 (s, 1H), 6.65 (dd, J = 15.9, 5.3 Hz, 1H), 7.26–7.37 (m, 5H), 7.93 (d, J = 7.3 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 14.06, 22.49, 27.61, 27.68, 28.49, 33.28, 35.10, 38.55, 40.64, 49.21, 55.06, 66.97, 128.03, 128.24, 128.63, 129.84, 136.46, 146.31, 156.24, 172.53, 180.31, 198.34. Elemental analysis for C₂₄H₃₃N₃O_5, calculated: C, 64.99; H, 7.50; N, 9.47; found: C, 64.73; H, 7.67; N, 9.28.

4.1.5.5. Benzyl ((S)-3-cyclopropyl-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)propan-2-yl)carbamate (SPR39)

In this reaction 8e and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. In this reaction methyl (triphenylphosphoranylidene)-2-propanone was used as the Wittig reagent. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 43% (two steps). R _f = 0.21 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.06–0.16 (m, 2H), 0.41–0.50 (m, 2H), 0.66–0.75 (m, 1H), 1.59–1.69 (m, 2H), 1.75–1.88 (m, 1H), 1.98–2.08 (m, 1H), 2.23 (s, 3H), 2.30–2.50 (m 2H), 3.24–3.38 (m, 2H), 4.31–4.40 (m, 1H), 4.52–4.61 (m, 1H), 5.11 (s, 2H), 5.63 (d, J = 9.0 Hz, 1H), 6.06 (s, 1H), 6.17 (d, J = 16.1 Hz, 1H), 6.65 (d, J = 15.7 Hz, 1H), 7.28–7.39 (m, 5H), 7.95 (d, J = 8.5 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 4.54, 4.66, 7.44, 27.69, 28.96, 35.01, 38.08, 38.69, 40.69, 49.83, 55.78, 67.04, 128.12, 128.27, 128.66, 130.03, 136.50, 146.07, 156.16, 172.31, 180.34, 198.34. Elemental analysis for C₂₄H₃₁N₃O_5, calculated: C, 65.29; H, 7.08; N, 9.52; found: C, 65.40; H, 7.19; N, 9.37.

4.1.5.6. Benzyl ((S)-4,4-dimethyl-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)pentan-2-yl)carbamate (SPR40)

In this reaction 8f and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 56% (two steps). R _f = 0.20 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.96 (s, 9H), 1.45 (dd, J = 14.4, 9.2 Hz, 1H),1.57–1.64, (m, 1H), 1.74–1.82 (m, 2H), 2.05 (ddd, J = 14.2, 11.9, 5.0 Hz, 1H), 2.21 (s, 3H), 2.28–2.37 (m, 1H), 2.38–2.46 (m, 1H), 3.22–3.32 (m, 2H), 4.32–4.40 (m, 1H), 1.45–4.57 (m, 1H), 5.10 (s, 2H), 5.51 (d, J = 8.8 Hz, 1H), 6.15 (d, J = 16.0 Hz, 1H), 6.46 (s, 1H), 6.63 (dd, J = 16.0, 5.5 Hz, 1H), 7.27–7.34 (m, 5H), 7.88 (d, J = 7.0 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 27.68, 28.62, 29.85, 30.71, 35.12, 38.56, 40.65, 46.83, 49.41, 53.17, 67.03, 127.96, 128.24, 128.64, 129.98, 136.52, 146.17, 155.97, 173.52, 180.32, 198.40. Elemental analysis calculated for C₂₅H₃₅N₃O₅: C, 65.62; H, 7.71; N, 9.18; found: C, 65.79; H, 7.57; N, 9.10.

4.1.5.7. Benzyl ((S)-3-cyclohexyl-1-oxo-1-(((S,E)-5-oxo-1-((S)-2-oxopyrrolidin-3-yl)hex-3-en-2-yl)amino)propan-2-yl)carbamate (SPR41)

In this reaction 8g and 1-(triphenylphosphoranylidene)-2-propanone were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 25:1. Consistency = pale yellow solid; Yield = 45% (two steps). R _f = 0.23 in EtOAc/MeOH 25:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.84–1.03 (m, 2H), 1.09–1.25 (m 3H), 1.31–1.39 (m, 1H), 1.46–1.54 (m, 1H), 1.61–1.72 (m, 5H), 1.77–1.85 (m, 3H), 1.99–2.08 (m, 1H), 2.24 (s, 3H), 2.31–2.40 (m, 1H), 2.40–2.48 (m, 1H), 3.24–3.36 (m, 2H), 4.30–4.39 (m, 1H), 4.50–4.62 (m, 1H), 5.09 (d, J = 12.5 Hz, 1H), 5.13 (d, J = 12.3 Hz, 1H), 6.17 (d, J = 16.0 Hz, 1H), 6.21 (s, 1H), 6.66 (dd, J = 16.0, 5.0 Hz, 1H), 7.28–7.38 (m, 5H), 7.91 (d, J = 6.9 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 26.21, 26.39, 26.51, 27.68, 28.81, 32.73, 33.81, 34.28, 35.06, 38.63, 40.67, 41.01, 49.56, 53.17, 67.03, 128.04, 128.26, 128.66, 129.91, 136.55, 146.22, 156.24, 173.05, 180.33, 198.32. Elemental analysis calculated for C₂₇H₃₇N₃O₅: C, 67.06; H, 7.71; N, 8.69; found: C, 67.32; H, 7.53; N, 8.44.

4.1.5.8. (S,E)-Methyl 4-((S)-2-(((benzyloxy)carbonyl)amino)hexanamido)-5-((S)-2-oxopyrrolidin-3-yl)pent-2-enoate (SPR42)

In this reaction 8d and methyl (triphenylphosphoranylidene)-acetate were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 100:1. Consistency = pale yellow solid; Yield = 38% (two steps). R _f = 0.34 in EtOAc/MeOH 100:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.81–0.92 (m, 3H), 1.27–1.38 (m, 4H), 1.55–1.68 (m, 2H), 1.73–1.87 (m, 4H), 1.97–2.07 (m, 1H), 2.30–2.48 (m, 2H), 3.71 (s, 3H), 4.24–4.34 (m, 1H), 4.52–4.60 (m, 1H), 5.09 (s, 2H), 5.61 (d, J = 8.2 Hz, 1H), 5.93 (d, J = 15.7 Hz, 1H), 6.23 (s, 1H), 6.83 (dd, J = 15.7, 5.3 Hz, 1H), 7.27–7.37 (m, 5H), 7.85 (d, J = 6.9 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 14.04, 22.52, 27.67, 28.64, 33.37, 35.10, 38.60, 40.64, 49.35, 51.75, 55.07, 67.01, 120.84, 128.13, 128.25, 128.64, 136.50, 147.80, 156.21, 166.79, 172.41, 180.29. Elemental analysis calculated for C₂₄H₃₃N₃O₆: C, 62.73; H, 7.24; N, 9.14; found: C, 62.87; H, 7.02; N, 8.96.

4.1.5.9. (S,E)-Methyl 4-((S)-2-(((benzyloxy)carbonyl)amino)-4,4-dimethylpentanamido)-5-((S)-2-oxopyrrolidin-3-yl)pent-2-enoate (SPR43)

In this reaction 8f and methyl (triphenylphosphoranylidene)-acetate were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 100:1. Consistency = pale yellow solid; Yield = 29% (two steps). R _f = 0.33 in EtOAc/MeOH 100:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.96 (s, 9H), 1.43 (dd, J = 14.4, 9.3 Hz, 1H), 1.57–1.66 (m, 1H), 1.74–1.83 (m 2H), 2.02 (ddd, J = 14.3, 11.9, 5.3 Hz, 1H), 2.30–2.46 (m, 2H), 3.20–3.35 (m, 2H), 3.72 (s, 3H), 4.25–4.36 (m, 1H), 4.47–4.58 (m, 1H), 5.10 (s, 2H), 5.40 (d, J = 8.7 Hz, 1H), 5.93 (d, J = 15.6 Hz, 1H), 6.15 (s, 1H), 6.82 (dd, J = 15.7, 5.6 Hz, 1H), 7.27–7.36 (m, 5H), 7.77 (d, J = 7.2 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 28.78, 29.87, 30.74, 35.14, 38.57, 40.63, 46.83, 49.52, 51.75, 53.20, 67.08, 120.99, 128.08, 128.24, 128.63, 136.53, 147.70, 155.98, 166.84, 173.43, 180.27.43Elemental analysis calculated for C₂₅H₃₅N₃O₆: C, 63.41; H, 7.45; N, 8.87; found: C, 63.59; H, 7.60; N, 8.72.

4.1.5.10. (S,E)-Methyl 4-((S)-2-(((benzyloxy)carbonyl)amino)-3-cyclohexylpropanamido)-5-((S)-2-oxopyrrolidin-3-yl)pent-2-enoate (SPR44)

In this reaction 8g and methyl (triphenylphosphoranylidene)-acetate were used as the intermediate and Wittig reagent, respectively. Eluent mixture: EtOAc/MeOH, 100:1. Consistency = pale yellow solid; Yield = 35% (two steps). R _f = 0.26 in EtOAc/MeOH 100:1.¹H NMR (500 MHz) in CDCl₃, δ = 0.84–0.99 (m, 3H), 1.09–1.24 (m, 3H), 1.29–1.38 (m, 1H), 1.57–1.73 (m, 7H), 1.73–1.84 (m, 1H), 2.02 (ddd, J = 14.2, 12.0, 5.3 Hz, 1H), 2.30–2.46 (m, 2H), 3.22–3.34 (m, 2H), 3.72 (s, 3H), 4.29–4.38 (m, 1H), 4.50–4.60 (m, 1H), 5.08 (d, J = 12.6 Hz, 1H), 5.11 (d, J = 12.0 Hz, 1H), 5.45 (d, J = 8.4 Hz, 1H), 5.93 (d, J = 15.7 Hz, 1H), 6.16 (s, 1H), 6.83 (dd, J = 15.6, 5.3 Hz, 1H), 4.27–4.37 (m, 5H), 7.80 (d, J = 6.9 Hz, 1H). ¹³C NMR (125 MHz) in CDCl₃, δ = 26.19, 26.34, 26.54, 28.72, 32.76, 33.79, 34.26, 35.12, 38.56, 40.64, 41.10, 49.36, 51.75, 53.13, 67.02, 120.85, 128.10, 128.25, 128.64, 136.55, 147.79, 156.22, 166.81, 172.97, 180.31. Elemental analysis calculated for C₂₇H₃₇N₃O₆: C, 64.91; H, 7.46; N, 8.41; found: C, 64.83; H, 7.22; N, 8.67.

4.2. Biological evaluations

4.2.1. Enzyme expression and preparation

4.2.1.1. SARS-CoV-2 M^pro

The expression of SARS-CoV-2 M^pro was performed exactly as described previously [26]. Briefly, pMal-c2 plasmid DNA (New England Biolabs) containing the entire SARS-CoV-2 M^pro coding sequence flanked, at the 5′ end, by a short sequence specifying the 5 C-terminal residues of nonstructural protein 4 and, at the 3′ end, 6 histidine codons at the 3’ end. The plasmid construct was transformed in Escherichia coli (E. coli) BL21-Gold (DE3) (Agilent Technologies, Santa Clara, CA, USA) cells. After growing the bacterial culture in LB medium with ampicillin to an OD₆₀₀ of ∼0.5 and induction with isopropyl-β-d- thiogalactopyranoside (IPTG), M^pro was produced at 18 °C for 16 h. Cell pellets obtained by centrifugation were resuspended in lysis buffer (20 mM Tris−HCl pH 7.8, 150 mM NaCl, 20 mM imidazole) and lysed by sonication (Sonoplus HD 2200; Bandelin, Berlin, Germany). The cleared lysate was subjected to immobilized metal affinity chromatography (IMAC) on a HisTrap HP 5 ml column (Cytiva Europe GmbH, Freiburg im Breisgau. Germany). After washing with IMAC buffer A (20 mM Tris−HCl pH 7.8, 200 mM NaCl, 20 mM imidazole), M^pro was eluted with IMAC buffer B (20 mM Tris−HCl pH 7.8, 200 mM NaCl, 500 mM imidazole). The collected fractions containing M^pro, were subjected to a gel filtration step (HiLoad 16/600 Superdex 75 pg column; GE Healthcare, Chicago, IL, USA) in SEC buffer (20 mM Tris−HCl pH 7.8, 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT)). After dilution to 10 μM and adjustment to 10% (v/v) glycerol, M^pro was shock frozen in liquid N₂ and stored at −80 °C.

4.2.1.2. SARS-CoV-2 PL^pro

The SARS-CoV-2 PL^pro was prepared exactly as described previously [64].

4.2.1.3. Dengue virus NS2B/NS3

The glycine linked dengue virus 2 NS2B/NS3 protease was prepared exactly as described previously [65].

4.2.1.4. hCatL and hCatB

Human cathepsins B in this study was purchased from Calbiochem (Merck Merck KGaA, Darmstadt, Germany). Human Cathepsin L was purchased by Sigma Adrich (St. Louis, Missouri, USA).

4.2.2. Enzyme activity assays

4.2.2.1. Enzyme inhibition assays

Nirmatrelvir was purchased from AOBIUS (Gloucester, Massachusetts, USA). 11a was purchased from BIOMOL GmbH (Hamburg, Germany). Inhibitory activity was determined using either a FRET-substrate (SARS-CoV-2 M^pro) [26] or fluorogenic AMC-substrates (NS2B/NS3 [65], PL^pro [64], hCatL [66], and hCatB [67]). Assays were performed in white flat-bottom 96-well microtiter plates (Greiner bio-one, Kremsmünster, Austria) on a TECAN Infinite F2000 PRO plate reader (Agilent Technologies, Santa Clara, USA) for SARS-CoV-2 M^pro or a TECAN Spark 10 M (Agilent Technologies) for assays using AMC-substrates.

As a general procedure, inhibitors were dissolved as 20 mM DMSO-stock solutions. Substrates were also dissolved in DMSO. For more detailed information for each assayed enzyme see (Table S2). After an initial screening at 20 μM for SARS-CoV M^pro or 100 μM for all other proteases, IC₅₀ values of active inhibitors were determined. Therefore. half-logarithmic dilution series of active inhibitors were prepared (eg. final concentrations: 100, 30, 10, 3, 1, 0.3, 0.1 μM, and DMSO as control). For each well, 185 μL of the respective buffer was supplemented with 5 μL of the enzyme-solutions, followed by 10 μL of the Inhibitors. Reactions were initiated without further incubation by addition of 5 μL of the substrate-solutions and vigorous mixing. Measurements were performed in at technical triplicates. The fluorescence was recorded in intervals of 30 s for 10 min at 25 °C (EDANS: λ _ex 335 nm; λ _em 493 nm; AMC: λ _ex 380 nm; λ _em 460). IC₅₀ values were calculated with GraFit (Version 6.0.12; Erithacus Software Limited, East Grinstead, West Sussex, UK) [68] by fitting the enzymatic activities against the respective inhibitor concentration to the four-parameter equation. To correct for substrate competition, K _i values calculated by the Cheng-Prusoff equation.

4.2.2.2. Dilution assay

Experiments shifting the inhibitor concentrations from 5-fold the respective IC₅₀ to 0.1-fold the IC₅₀, were performed for SARS-CoV-2 M^pro and hCatL mainly as described previously [55,56]. The 5-fold IC₅₀ solutions were as described for enzyme activity assays without addition of the substrate but with 50-fold the SARS-CoV-2 M^pro or 10-fold the hCatL concentration. Samples were incubated for 60 min at rt to ensure potent inhibition. After incubation, one sample per enzyme was diluted 50-fold to achieve inhibitor concentrations of 0.1-fold the IC₅₀. Enzymatic activities of 58.5 μL samples initiated with 1.5 μL of the respective substrate concentration were recorded in triplicates before and after dilution and were normalized to similar treated control experiment with DMSO instead of inhibitor solutions.

4.2.2.3. Dialysis assay

Experiments extracting unbound or reversibly bound inhibitors from SARS-CoV-2 M^pro were performed using a custom-built dialysis chamber allowing the parallel examination of five samples mainly as described previously [55,56]. In brief, a dialysis membrane (cut off 13 kDa MW) connected the sample containing wells with a chamber of continuously flowing assay buffer supplemented with 7.5% (v/v) DMSO (flowrate: ∼200 mL/h). Samples were prepared similar to the enzymatic activity assay conditions without substrate (5-fold the volumes). To potently inhibit SARS-CoV-2 M^pro activity, inhibitors were used at a concentration 10-fold the IC₅₀. To compensate for loss of enzymatic activity during dialysis, SARS-CoV-2 M^pro was used in a final concentration of 250 nM, exceeding the concentration of Nirmatrelvir, so that activity could not be fully inhibited by Nirmatrelvir. Activity control measurements were performed using pure DMSO instead of the inhibitor solutions. Hence, 975 μL reaction mixtures were incubated for 60 min at rt to allow covalent reaction (if possible). After that, the first samples (t = 0 min) were drawn, and the rest was put in the wells of the dialysis device. Samples of 58.5 μL were drawn in duplicates at seven different time points (0, 30, 60, 150, 300, 600 and 1500 min). Enzymatic cleavage reactions were initiated by the addition of 1.5 μL of substrate solution in a final concentration of 25 μM. Fluorescence was recorded over 10 min as described for enzymatic activity assays.

4.2.3. Cell-based antiviral activity and cytotoxicity assays

4.2.3.1. Cells and viruses

Huh-7 cells overexpressing human angiotensin-converting enzyme 2 (ACE2) (Huh-7-ACE2; kindly provided by Friedemann Weber, Institute of Virology, Justus Liebig University Giessen) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 U/mL penicillin and 100 μg/mL streptomycin) at 37 °C in an atmosphere containing 5% CO₂. The SARS-CoV-2 isolate Munich 929 [69] was kindly provided by Christian Drosten (Institute of Virology, Charité-Universitätsmedizin, Berlin).

4.2.3.2. Cell toxicity

Cytotoxic concentrations 50% (CC₅₀) of the compounds used in antiviral activity assays were determined using MTT assays as described previously [70].

4.2.3.3. Antiviral activity

To determine effective concentrations 50% (EC₅₀) of the respective compounds, Huh-7-ACE2 cells were inoculated with SARS-CoV-2 at a multiplicity of infection (MOI) of 0.1 plaque-forming units (pfu) per cell. After incubation for 1 h at 33 °C, the virus inoculum was replaced with fresh cell culture medium containing the test compounds at the indicated concentration. After 23 h at 33 °C, the cell culture supernatants were collected and virus titers were determined by virus plaque assay as described previously [64].

4.3. Molecular modeling methods

4.3.1. Docking

AutoDock4 (AD4) was employed for molecular docking calculations [71]. The formation of the covalent adduct between ligand and protein was modeled using the covalent docking protocol devised by Bianco et al. known as the “flexible side chain method” [62]. Using the Maestro suite, the ligand SPR39 was modeled with two extra atoms where the alkylation would take place. Namely, a sulfur and a carbon atom, in order to match the corresponding atoms Cys145 of the protein. The M^pro X-ray structure having PDB code 7BQY was downloaded from the RCSB PDB database and prepared for docking using the protein preparation wizard, part of the Schrödinger suite [72]. The overlay of the ligand with the reactive cysteines was attained using the scripts offered by the AD4 website. The AutoGrid4 software was used to prepare the protein grid maps using the ligand atom types as probes. The enzyme grid box with a size of 60 Å × 60 Å × 60 Å and 0.375 Å spacing was centered on the coordinates of the cognate c7BQY o-crystal ligand. The docking calculations were performed by treating the modified cysteine/ligand residue as flexible. The Lamarckian Genetic Algorithm (LGA) was employed for the docking simulations. 100 runs of LGA were executed. The docking run consisted of 20 million energy evaluations using the Lamarckian genetic algorithm local search (GALS) method. The GALS method evaluates a population of possible docking solutions and propagates the most successful individuals from each generation into the subsequent generation of possible solutions. A low-frequency local search according to the method of Solis and Wets is applied to docking trials to ensure that the final solution represents a local minimum. The docking experiment was performed with a population size of 150, and 300 rounds of Solis and Wets local search were applied with a probability of 0.06. A mutation rate of 0.02 and a crossover rate of 0.8 were used to generate new docking trials for subsequent generations, and the best individual from each generation was propagated over the next generation. All the other settings were left at their default value. The docking results from the calculation were clustered on the basis of root-mean-square deviation (solutions differing by less than 2.0 Å) between the Cartesian coordinates of the atoms and were ranked on the basis of free energy of binding (ΔG_AD4). Finally, the SPR39 docking pose with the best-predicted ΔG_AD4 was selected. All the images were rendered using the UCSF Chimera X software [73].

4.3.2. Molecular dynamics simulations

The complex obtained from the docking results was subjected to a molecular dynamics (MD) simulation by means of the Desmond module of the Schrödinger software package [74,75]. As a first step, the system builder panel was used to prepare the system for the MD calculation. Each complex was embedded in a parallelepiped box by solvating it with TIP3P water model [76]. The initial −3 negative charge was neutralized using 3 Na⁺ ions. Then, the system was equilibrated by employing the NPT ensemble with the default Desmond protocol that includes eight steps. The first 7 are short simulations known as the equilibration phase, where the system temperature is gradually increased and the solute is partially restrained. After the first 7 steps, the equilibrated systems were subjected to the 100 ns MD final production run with PBC conditions and NPT ensemble. The system was set to 300° K temperature and 1 atm pressure throughout the simulation utilizing the Martyna−Tobias−Klein barostat [77] and Nose−Hoover chain thermostat [78]. The OPLSe force field [79] was used for all the MD simulation steps.

Author contribution

SP: Investigation, synthesis, methodology, characterization, writing—original draft, and validation; RE: Validation, review, and editing; EC: Synthesis; SDM: Design, molecular docking, editing, validation and funding acquisition; SH: Enzyme preparation, enzymatic activity assays, dilution assays, dialysis assays; CM: cytotoxicity and antiviral assays; JZ: review and editing, funding acquisition; T.S.: Validation, review, and editing; SC: Design, molecular docking, editing, validation and funding acquisition; MZ: Validation, Funding acquisition, Supervision, review, and editing. All authors have read and agreed to the published version of the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.