Structure-based macrocyclization of α-ketoamides leads to potent inhibitors of coronaviral and enteroviral proteases

Abstract

Viral proteases represent validated targets for direct-acting antivirals and the treatment of associated infections. In co-crystal structures of M^pro of SARS-CoV-2 with peptidomimetic inhibitors, we noticed a spatial proximity of sidechains filling the S1’ and S2 pockets, as well as those filling S3 and S1 pockets. To enhance molecular rigidity, the proximal residues were conformationally fixed by macrocyclization. We report the synthesis of two macrocyclic series, i.e. exocyclic nitriles with linked P3 and P1 residues and endocyclic α-ketoamides with linked P1’ and P2 residues, and characterize their binding modes and bioactivities. The 17-membered macrocyclic α-ketoamide 20 f inhibited M^pro (IC₅₀ = 370 nM) and exerted anti-SARS-CoV-2 effects (EC₅₀ = 1.9 μM). Leveraging structural similarities between M^pro and the 3C^pro of enterovirus D68, we describe with two co-crystal structures how α-ketoamide macrocycles bound to and inhibited the enteroviral protease. Notably, 20 f exhibited very potent antiviral activities with EC₅₀‘s of 33, 133, and 146 nM against EV-D68, EV-A71, and CVB3, respectively. The study demonstrates how broad-spectrum activity can be achieved with direct-acting antivirals.

Viral proteases are key targets for developing direct-acting antivirals to combat infections like SARS-CoV-2. Here, the authors synthesize macrocyclic α-ketoamide inhibitors, demonstrating potent inhibition of SARS-CoV-2 and enteroviruses, highlighting macrocyclization’s potential to enhance antiviral efficacy and achieve broad-spectrum activity.

Introduction

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), the etiological agent of the COVID-19 pandemic, critically depends on its main protease (M^pro), also referred to as 3C-like protease (3 CL^pro), for viral replication and maturation^1,2. M^pro facilitates the proteolytic cleavage of the viral polyprotein at conserved recognition sites, generating functional proteins essential for the assembly and propagation of the virus³. Given its pivotal role in the viral life cycle and the absence of homologous proteases in humans, M^pro constitutes an attractive and selective target for the development of antiviral therapeutics. Significant efforts to inhibit M^pro have led to the discovery and development of numerous potent compounds that bind either covalently or non-covalently to the protease^2,4,5. Covalent inhibitors exploit the highly nucleophilic catalytic cysteine residue (Cys145) within the active site of M^pro, forming either reversible or irreversible covalent bonds that effectively block enzymatic activity. This binding mode led to approved drugs such as nirmatrelvir⁶, leritrelvir⁷, simotrelvir⁸, and other potent compounds in clinical trials^9–12, notably these compounds either employ a nitrile or an α-ketoamide as a reactive warhead (Fig. 1). Also, an inhibition through non-covalent interactions is possible, as shown by the approved drug ensitrelvir¹³ and a series of other potent compounds^14–18. Despite these advances, the search for improved M^pro inhibitors continues to enhance their selectivity, pharmacokinetic and safety profiles. The use of nirmatrelvir and the co-administered drug ritonavir is complicated due to drug-drug interactions and might be impaired due to resistant variants featuring the E166V and E166A mutation^19,20. Recently, it has been reported that resistance to one M^pro inhibitor might increase susceptibility to another, suggesting that finding additional inhibitors is valuable²¹. By leveraging insights from existing compounds^{7,9,14,22–23} and expanding towards new chemical formats²⁴, the pursuit of effective M^pro inhibitors remains a cornerstone of therapeutics against emerging SARS-CoV-2 variants and for a preparedness against future pandemic threats by coronaviruses.

Fig. 1. Structures of selected previously reported SARS-CoV-2 M^pro inhibitors.

The covalent binders nirmatrelvir and leritrelvir and the non-covalent binder ensitrelvir have been licensed for COVID-19. The linear compounds 13b-K, 12c, 11a, and 11r were starting points of our work. Compounds 84 and 26 represent the first macrocyclic M^pro inhibitors. The year of first publication is given in brackets.

Moreover, there are opportunities to expand such protease inhibitors to broader spectrum antivirals, because the 3C-like protease M^pro of coronaviruses shares structural similarities to the 3C proteases of enteroviruses (EV), including the 3C of the Coxsackievirus B3 (CVB3). Finding broad-spectrum inhibitors is non-obvious, because of pronounced differences in size, domain structure, S2 pocket specificity, or catalytic residues between the proteases²⁵. But yet, we previously reported the peptidomimetics 11a and 11r as antiviral agents with activity against both coronaviruses and enteroviruses²⁶, motivating us to investigate possible line extensions for such protease inhibitors^26,27.

Macrocyclic drugs, first discovered from natural sources, have been recognized as essential medicines, e.g., due to their immunosuppressant, antibiotic or antitumor activities^28–31. Their increased structural rigidity and diverse functionality within a pre-organized ring structure minimizes entropy loss upon binding, potentially enhancing target affinity. Furthermore, macrocyclization has been demonstrated to confer resistance to proteolytic degradation, facilitate intramolecular hydrogen bonding, and improve bioavailability, thus enabling compounds to effectively reach intracellular targets. These advantages have led to an intense exploitation of macrocycles beyond natural products by designing synthetic scaffolds in the past two decades^28,32,33. Because macrocyclic peptides typically exhibit lower off-target interactions, higher membrane permeability, and improved proteolytic stability compared to linear peptides. A first successful application of the concept to antivirals has been demonstrated by ciluprevir, an inhibitor of the NS3/4A protease of hepatitis C virus (HCV) that was derived from a linear substrate analog. Further optimization led to the approved anti-HCV drugs simeprevir and grazoprevir³⁴. Also, macrocycles against 3C proteases of hRV³⁵ and enteroviruses have been reported³⁶. Recently, the groups of Gütschow, Müller and Sofia found macrocyclic compounds inhibiting M^pro of SARS-CoV-2, i.e., the covalently binding azapeptide nitrile 84 and the noncovalently binding peptidomimetic 26, respectively (Fig. 1)^37,38. Our groups have investigated α-ketoamides as reversible, covalent binders of 3C proteases such as M^pro using structural biology, and reported the ketoamide 13b as one of the first M^pro inhibitors against SARS-CoV-2 in early 2020. Building on insights from co-crystal structures of related ketoamides^26,39,40, such as 13b-K or 11a (Fig. 1), we designed two series of covalently binding, macrocyclized compounds. We prepared four nitriles with linked P3 and P1 residues and eight α-ketoamides with linked P1’ and P2 residues, and studied their binding modes by co-crystal structures, as well as their structure-activity relationships (SAR). Finally, we investigated their ability to inhibit enteroviral 3C proteases and found a second indication for the macrocycles as antivirals against enteroviruses. A first profiling of ADME properties suggests potential for further preclinical development.

Results

Design rationale

A close inspection of crystallographic data on 12c bound to M^pro (PDB: 8AIZ) revealed an 8.6 Å spatial separation between the γ-carbon of the lactam side chain in P1 and the benzylic position of the P4 substituent (Fig. 2A, B). This observation suggested that a P1–P3 linkage by macrocyclization to enhance conformational rigidity and preorganization for target binding might be feasible. For this purpose, glutamine was chosen to occupy the P1 position, as in natural substrates of M^pro. We selected nitrile as a less reactive covalent warhead, because its high efficacy has been clinically validated (e.g., in nirmatrelvir⁶ or simotrelvir⁸), and because the synthesis could build on readily available L-glutamate. This led to the P1–P3-linked macrocyclic nitriles as target structures for synthesis (Fig. 3A and Scheme 1).

Fig. 2. Co-crystal structure of M^pro in complex with 12c (PDB: 8AIZ).

A Compound 12c is colored in brown, and M^pro residues are colored in cyan. H-bonds between 12c and the corresponding M^pro residues are colored in red; yellow dashes highlight all other interactions. B The γ-carbon of the lactam side chain in P1 and the benzylic position of the P4 substituent are separated by a distance of 8.6 Å. C The β-carbon of the cyclopropyl residue from the P2 site and the amide nitrogen atom of the P1’ site are separated by a distance of 6.2 Å.

Fig. 3. Structures and retrosynthesis of M^pro-targeting macrocycles.

A P1–P3-linked macrocycles, exemplified by 9b, with an exocyclic nitrile warhead. B P2-P1’-linked macrocycles, exemplified by 20f, with an endocyclic α-ketoamide warhead.

Scheme 1.

Synthesis of macrocyclic nitriles 9a-b^a. ^aReagents and conditions: a 1. (BOC)₂O/Py-NH₄CO₃, 1,4-dioxane, 2. TFA, CH₂Cl₂, 0 °C–rt, 5 h, quant. over two steps; b HATU/TEA, DMF, 0 °C–rt, 6 h, 90%; c TFA, CH₂Cl₂, 0 °C–rt, 5 h, quant.; d 5-(((benzyloxy) carbonyl) amino)pentanoic (or hexanoic) acid, HATU/TEA, DMF, 0 °C–rt, 10 h, 65–74%; e Pd/C, H₂ gas, CH₃OH, 5 h, 80%; f HATU/TEA, DMF, 0 °C–rt, 16 h, 75%; g Burgess reagent, DCM, rt, 4 h, 51–75%; e Pd/C, H₂ gas, CH₃OH, 5 h, 82–85%; h HATU/TEA, DMF, 0 °C–rt, 3 h, 43–64%.

When assuming a different perspective on the co-crystal structure of 12c to M^pro (PDB: 8AIZ), it became evident that also the β-carbon of the cyclopropyl residue from the P2 site and the amide nitrogen atom of the P1’ site were separated by a distance of only 6.2 Å (Fig. 2C). This implied that a second macrocyclization was possible between those sites. For practical reasons, we replaced the cyclopropane residue of P2 with a phenyl ring. Such a change was tolerated according to previous data obtained with the peptidomimetic 11a²⁶ on the M^pro of SARS-CoV. Moreover, a co-crystal structure between 11a and M^pro suggested that a meta-substitution of the aryl ring would be optimal for a linkage to the P1’ site. For the latter, structural data, with our as well as other ketoamides (e.g., leritrelvir, Fig. 1), gave confidence that an alkyl chain was tolerated at the ketoamide nitrogen. In order to optimize interactions with the P4 site, the N-terminal amine was “capped” with different moieties, such as phenylacetyl that were identified in our previous optimization studies⁴⁰. We therefore aimed synthesizing the P2-P1’-linked macrocycles as depicted in Fig. 3B.

Synthesis of macrocyclic nitriles with linked P3 and P1 residues

Our retrosynthetic plan for the synthesis of macrocyclic nitriles envisioned an amide coupling with HATU under basic conditions for the ring closure and a dehydration reaction with Burgess reagent to convert the exocyclic amide to a nitrile (Fig. 3A). The central intermediate 5a would be derived from an amide coupling reaction with the protected pyridone-derived dipeptide acid and the glutamatic acid amide.

In the forward direction, N-Boc-D-Glu(OBn)-OH was converted to the corresponding amide by treating it with Boc anhydride, pyridine, and ammonium carbonate in 1,4-dioxane at room temperature for 16 h (Scheme 1)⁴¹. Subsequently, the amide was deprotected to afford (S)-1-amino-5-(benzyloxy)-1,5-dioxopentan-2-aminium trifluoro acetate in quantitative yields. The latter⁴² was treated with the pyridone-containing peptidomimetic dipeptide 1⁴⁰ to the tripeptide 2 in 90% yield using HATU/TEA coupling conditions in DMF. Following deprotection of the Boc-group with TFA in DCM to give 3, the tetrapeptides 4a, b were obtained under optimized acid–amine coupling conditions with the ω-amino-substituted carboxylic acids.

The fully protected amide 4a was transformed into the corresponding de-protected intermediates by treatment with Pd/C under a hydrogen atmosphere for 5 h, affording 5a in 80% yield. Intermediate 5a was then cyclized to the macrocyclic amide using standard peptide coupling conditions under high dilution to minimize dimer formation, yielding the desired macrocycle 6a in 75% yield. However, attempts to convert the exocyclic amide into the corresponding nitrile using Burgess reagent in various solvents were unsuccessful, likely due to the high polarity of the compounds. Also, the use of other dehydrating agents such as TFFA/DMF was unsuccessful. Consequently, we adopted an alternative synthetic approach, converting the fully protected amides 4a, b to the corresponding nitriles 7a, b in 51–75% yields. The nitriles were de-protected to the corresponding free amino acids (82–85% yield) and then successfully cyclized to the desired macrocyclic nitriles 9a, b with 43–64% yields. At the final stage, the respective diastereomers were separated and isolated using preparative HPLC purifications.

A stereochemical assignment was made with the help of a co-crystal structure of 9b-K in complex with M^pro, that revealed an (S,S)-configuration for that diastereoisomer (Fig. 4). In consequence, 9b-H possesses an (R,S)-configuration. We then identified the stereochemistry of the 9a derivatives using NMR spectroscopy. The 9b-K macrocycle displayed a triplet at 5.06 ppm (t, J = 7.7 Hz, 1H) together with a doublet of doublets at 4.97 ppm (dd, J = 9.9, 4.7 Hz, 1H). In contrast, the 9b-H macrocycle showed two distinct doublets of doublets at 4.99 ppm (dd, J = 9.7, 5.8 Hz, 1H) and 4.65 ppm (dd, J = 10.2, 4.0 Hz, 1H) in the same region. The corresponding amide NH protons exhibited only minor chemical shift perturbations upon a change of stereoconfiguration (8.21 and 8.17 ppm). Similarly, in the 9a-K macrocycle, the two CH protons appeared as a single multiplet at 4.93 ppm (2H), whereas in the 9a-H isomer, these protons resonated as two separate doublets of doublets at 4.99 and 4.65 ppm (1H each), respectively. Thus, matching the chemical shifts observed in the 9b pair to those in the 9a pair allowed the assignment of the stereochemistry. The downfield NH proton at ~8.2 ppm was present in both 9a isomers as well, with only minor differences in chemical shift and coupling pattern (Supporting Information, S11–14).

Fig. 4. Co-crystal structure of M^pro in complex with 9b-K (PDB: 9SSN).

(A Shown as sticks, B as surface) Amino acids (sticks) and water molecules (spheres) are colored in cyan, compound 9b-K is colored in gray, and all H-bonds between the inhibitor and the corresponding M^pro residues are colored in red. Yellow dashes highlight all other interactions. C, D Superposition of the 9b-K (colored in gray) with 13b-K (colored in light blue), with key differences in the placement given by orange dashed lines.

Functional and structural characterization of interactions between macrocyclic nitriles and M^pro

The macrocyclic nitriles were characterized with respect to their ability to inhibit the proteolytic activity of SARS-CoV-2 M^pro. For this purpose, the compound-mediated reduction in the cleavage rate of a model substrate was measured. Among the tested compounds, only 9b-K exhibited a measurable inhibition of 62% at 20 µM, while 9a-H, 9a-K, and 9b-H were inactive (enzyme inhibitory activities in Table S1). In spite of the moderate potency observed for 9b-K, we determined its co-crystal structure in complex with M^pro at a resolution of 1.5 Å, in order to shed light on its binding mode (Fig. 4).

The association of 9b-K was strengthened by 11 hydrogen bonds to M^pro residues. Moreover, the sulfur atom of Cys145 bound covalently to the P1´carbon of 1.8 Å in the S1´ subsite. An overall structural alignment of the 9b-K (9SSN) with 13b-K (PDB: 6Y2G) reveals root mean square (r.m.s.) deviations of 1.4, 1.8, and 2.7 Å for the Cα positions in the S2, S3, and S4 subsites, respectively (Fig. 4D). The P1´ imine occupied the oxyanion hole and make three interactions with the backbone amides of Cys145 (2.9 Å), Ser144 (3.5 Å; 3.0 Å for 13b-K) and Gly143 (3.5 Å; 3.0 Å for 13b-K). The adjacent amide oxygen formed a hydrogen bond with a water molecule HOH208 (2.9 Å) and a weak interaction with His41 Nε2 (3.2 Å; 2.7 Å for 13b-K). In the S1–S2 subsite, hydrogen-bonding interactions to His163 Nε2 (2.8 Å; 2.6 Å for 13b-K), the backbone carbonyl oxygen of Phe140 (3.1 Å; 3.0 Å for 13b-K) and the Glu166 Cε2 remained despite the absence of the γ-lactam. In the S3 subsite, the pyridone played the same role as 13b-K and make strong interaction with the backbone amide (2.9 Å) of Glu166 and weak interaction with the backbone carbonyl (3.6 Å; 3.4 Å for 13b-K) of the same residue. However, the interaction to Gln189 became weaker in 9b-K (4.5 Å vs. 4.2 Å for 13b-K). Finally, the P3 amide nitrogen of 9b-K interacted also with main chain carbonyl of Glu166 (3.3 Å; 3.0 Å for 13b-K). In summary, the macrocyclic ring of 9b-K showed only partial fits to the S3-S4 subsite, and multiple interactions were weaker in comparison to for 13b-K; this experimental binding mode thus provides an explanation for the poor inhibitory potency of 9b-K.

Synthesis of macrocyclic α-ketoamides with linked P1’ and P2 residues

To obtain macrocycles with an endocyclic α-ketoamide linking P1’ and P2 residues, our retrosynthesis envisaged the attachment of the P4 residue in the last step (Fig. 3B). The ring closure should involve a Grubbs metathesis reaction from the open-chain intermediate 15a. The latter should be obtained through a Passerini reaction between the aldehyde 14 and a terminally unsaturated isonitrile. A variation of the isonitrile chain length would allow an adjustment of the macrocyclic ring size.

In the forward direction, the synthesis started with the allylation of Boc- and methyl-protected m-bromophenylalanine⁴³ through a Suzuki–Miyaura cross-coupling to give the intermediate 10 (Scheme 2). The use of CsF as a base was important for an efficient outcome of the reaction⁴⁴. Reaction conditions were further optimised to use m-bromo phenylalanine ester with less catalyst loading. Alternatively, a Stille coupling with allylstannane, conducted on a gram scale, gave comparable yields (Supporting Information S5). Following hydrolysis with lithium hydroxide, the carboxylic acid 11 was obtained in a yield of 80%. The intermediate 11 was coupled with methyl-(S)-2-amino-3-((S)-2-oxopyrrolidin-3-yl)propanoate HCl salt to give the dipeptide ester 12 using HATU in TEA as a base (74% yield). The ester functionality in 12 was further converted to the aldehyde 14 in two steps⁴⁰. The aldehyde 14 represented the last common intermediate before diversification. It was reacted with pent-4-en-isonitrile or hept-5-en-isonitrile and acetic acid in a Passerini reaction to give the acetoxy amides 15a and 15b, respectively (Scheme 3). Macrocyclization was achieved by a ring-closing metathesis reaction using the Hoveyda–Grubbs-II catalyst, which converted 15a to the macrocyclic 16b as the major product as a mixture of E/Z diastereomers in a yield of 54%. In addition, 16a, whose cycle was shortened by one methylene unit, was formed as a side product in a yield of 9%. We assume that the Ru catalyst induced an isomerization of the allyl-substituted phenyl ring to a styrene prior to the metathesis reaction⁴⁵. The equivalent reaction with 15b gave the products 16d and 16c. Because main and side product could be separated by prep HPLC in both pairs, macrocycles of four different ring sizes 16a–d were available for SAR studies. The alkenes 16a–d were hydrogenated at room temperature, catalyzed by palladium on carbon (10%) to the corresponding alkanes 17a–d in yields of 82–87%. Upon removal of the Boc group to give 18a–d, a partial hydrolysis of the α-acetoxy moiety to α-hydroxyamides occurred, which were tolerated at this stage. The free primary amine was then equipped with P4 substituents. For this purpose, phenylacetic acid, a P4 group that was favorable in linear α-ketoamides⁴⁰, was activated with HATU and TEA over 3.5 h and reacted with all four macrocycles to give 19a–d. Three more analogs were prepared starting from intermediate 18b. The introduction of a larger phenylpropionyl moiety as an S4 group led to 19e, the use of trifluoroacetyl-tert-leucine, reflecting the P3/P4 groups of nirmatrelvir, gave 19f, and the reaction with 1,1,1-trifluoro-2-isocyanatoethane gave the urea 19g. The last steps comprised the complete basic hydrolysis of the acetoxyamides and a Dess–Martin oxidation to the final α-ketoamides 20a–f in yields of 53–69% over two steps. In order to probe whether the reduction of the double bond from 16 to 17 was beneficial or detrimental, it was kept in 20h. Analog 20h carried the phenylpropionyl side chain and was prepared as an E/Z mixture as shown in Scheme 4. All compounds were analytically characterized by NMR, HR-MS, and HPLC. In sum, macrocyclic α-ketoamides were prepared in 11–12 steps with typical overall yields of 3.2–1.5%.

Scheme 2.

Synthesis of intermediate aldehyde 14. ^aReagents and conditions: a Pd(PPh₃)₄, CsF, 2-allyl-4,4,5,5-tetramethyl-1,3,2-dioxaborolane, THF, reflux, 17 h, 93%; b LiOH.H₂O, aq.CH₃OH, 2 h, 80%; c HATU/TEA, DMF, 0 °C–rt, 16 h, 74%; d NaBH₄, CH₃OH, 5 h, 82%; e DMP, CH₂Cl₂, rt, 3 h, 65%. Note that 11, 12, and 13 are different from the previously reported compounds 11a, 11r, 12c, or 13b-K shown in Fig. 1.

Scheme 3.

Synthesis of macrocyclic α-ketoamides 20a–g. ^aReagents and conditions: a alkenyl isonitriles, CH₃COOH, CH₂Cl₂, rt, 17 h, 55–57%; b Hoveyda–Grubbs II, ClCH₂CH₂Cl, 18 h, 43–54%; c Pd-C (10%), H₂ gas, rt, 5 h, 82–87%; d Conc. HCl, CH₂Cl₂, 0 °C–rt, 5 h, 65–93%; e R-COOH, HATU, TEA, DMF, 0 °C–rt, 3 h, 35–63%; f (i) LiOH.H₂O, CH₃OH, H₂O, 0 °C–rt, 1.5 h; (ii) DMP, DMF, rt, 2 h, 53–69% over two steps; g CF₃CH₂NCO, DIPEA, DMF, 0 °C–rt, 24 h, 53%.

Scheme 4.

Synthesis of macrocyclic α-ketoamide 20 h. ^aReagents and conditions: a 4 M HCl, CH₂Cl₂, 5 h, 73%; b 3-phenylpropanoic acid, HATU, TEA, DMF, 0 °C–rt, 5 h, 48%; c (i) LiOH.H₂O, CH₃OH, H₂O, 0 °C– rt, 1.5 h; (ii) DMP, DMF, rt, 2 h, 65% over two steps.

Structural characterization of interactions between macrocyclic α-ketoamides and M^pro

Purified SARS-CoV-2 M^pro was co-crystallized with the macrocycles 20a, 20b, and 20e–h (Figs. 5 and 6, data collection and refinement statistics in Tables S2 and S3). The crystal structures of all inhibitors in complex with M^pro revealed that the sulfur atom of Cys145 formed a covalent bond (1.7–1.9 Å) with the electrophilic carbon of the P1´, establishing an R-configuration. Within the S1´ subsite of the enzyme, the P1´ carbonyl oxygen occupied the oxyanion hole and accepted three hydrogen bonds (2.8–3.3 Å) from the backbone amides of Cys145, Ser144, and Gly143 for all inhibitors (Figs. 5 and 6). The thiohemiketal formed a strong hydrogen bond (2.5–2.7 Å) with His41 Nε2 for 20b and 20e–g, but a weak interaction for 20a and 20h. All macrocyclic inhibitors showed a weak interaction of 3.3–3.6 Å between the macrocyclic ring and the Cγ2 of Thr25. Furthermore, a weak interaction (3.3–3.6 Å) was observed between the oxygen atom of the carbonyl group of Thr26 and the macrocyclic ring of all inhibitors except 20b and 20h. This suggests a relative decrease in the flexibility of the macrocyclic ring due to the additional double bond in 20h. While 20h was prepared as a mixture of E- and Z- diastereomers, only the E-isomer co-crystallized with M^pro.

Fig. 5. Overall co-crystal structure of Mpro in complex with 20a (A PDB: 9T5E) and with 20g (B PDB: 9SSO).

A Amino acids (sticks) and B water molecules (spheres) are colored in cyan, 20g is colored in gold and 20a is colored orange. All H-bonds between the inhibitor and the corresponding M^pro residues are colored in red. Yellow dashes highlight all other interactions.

Fig. 6. Co-crystal structure of macrocyclic ketoamides 20a (PDB: 9T5E), 20h (PDB: 9T72), 20b (PDB: 9T52), 20e (PDB:9T55), 20f (PDB: 9T5F) and 20g (PDB: 9SSO).

The views illustrate the occupation of the S4 pockets on the left sides by 20e–g.

With regards to the S1 and S2 subsites, all macrocycles formed interactions to His163 Nε2 (2.6 Å), to the backbone carbonyl oxygen of Phe140 (3.1–3.3 Å) and to the carboxylate of Glu166 (3.1–3.3 Å). However, interactions between the endocyclic main chain amide and the carbonyl oxygen of His164 (3.0–3.1 Å) as well as the hydrogen bond between the carbonyl oxygen of the inhibitor and the backbone amide of Glu166 (2.7–2.9 Å), were only found for 20f, 20g, and 20e.

Finally, only 20e–g occupied the S4 subsite, although 20e did not engage in specific hydrogen-bonding interactions (Fig. 6). 20g showed additional hydrogen-bonding interactions to Gln189 Oε1 (2.9 Å) and Gln189 Nε2 (3.0 Å). In contrast, a water molecule (HOH 98) formed a bifurcated hydrogen-bond with the main-chain amide of 20f (3.0 Å) and the Gln189 Oε1 (2.7 Å).

In vitro activities

The ability of the α-ketoamides to inhibit the enzymatic activity of M^pro was tested next. The analogs 20a–d all had a phenylacetyl residue, but the alkyl linker between the aryl and the amide moieties varied systematically, covering pentyl, hexyl, heptyl or octyl groups. These four analogs inhibited the enzyme with IC₅₀ values of 8.2, 1.37, 3.0, and 4.57 µM, respectively (Table 1). Thus, the size of the macrocycle had a clear impact on activity, and the hexyl linker, corresponding to a 17-membered macrocycle, was optimal. Therefore, subsequent analogs were prepared with this ring size. Also, the nature of the P4 residue turned out to be relevant for activity: The elongation of the phenylacetyl group to phenylpropionyl as in 20e led to an increase of activity from 1.37 to 0.5 µM. The urea 20g and the tert-leucine analog 20f were the most potent analogs in the series with IC₅₀’s of 0.36 and 0.37 µM, respectively. This increase in potency can be traced back to their ability to populate the S4 site, as observed in the co-crystal structures. An unsaturated hexenyl linker was incorporated in analog 20h, carrying a phenylpropionyl moiety in the P4 position. With an activity of 2.2 µM, the analog was weaker than the saturated analog 20e (IC₅₀ = 0.5 µM). This may reflect the unfavorable orientation of alkene functionality in the S1’ and S2 subsites and the reduced number of interactions observed in the co-crystal structure. In summary, the SAR studies showed that a 17-membered, saturated macrocycle was an optimal M^pro inhibitor, with analogs populating S4 sites such as 20e–g reaching activities in the nanomolar range.

Table 1.

Enzyme inhibition, antiviral activities and cytotoxicity of macrocyclic α-ketoamides

Cmpd	L	R	Ring size	Mpro IC50a (µM)	SARS-CoV-2 EC50b (µM)	CC50c (µM)	EV-D68 3Cpro IC50a (µM)	EV-D68 EC50d (µM)
20a	pentyl		16	8.2 ± 1.9	n.d.	n.d.	17% inh. @100 µM	n.d.
20b	hexyl		17	1.37 ± 0.36	23.3 ± 3.9	>50	2.8 ± 0.6	5.0 ± 0.7
20c	heptyl		18	3.0 ± 0.69	n.d.	n.d.	10% inh. @100 µM	n.d.
20d	octyl		19	4.57 ± 0.73	n.d.	n.d.	2.9 ± 0.4	n.d.
20e	hexyl		17	0.50 ± 0.09	7.0 ± 2.7	>50	1.4 ± 0.2	4.3 ± 0.3
20f	hexyl		17	0.37 ± 0.07	1.9 ± 0.3	>50	2.7 ± 0.2	0.033 ± 0.004
20g	hexyl		17	0.36 ± 0.12	>50	>50	2.9 ± 0.4	1.5 ± 0.1
20h	hex-2-enyle		17	2.2 ± 0.68	47.6 ± 1.7	>50	n.d.	n.d.
Nirmatrelvirf				0.015 ± 0.07	0.07 ± 0.02	>10	46.8 ± 25.5	n.d.
Rupintrivir				>30	n.d.	78.9 ± 11.0	0.28 ± 0.03	0.005 ± 0.001

^a ± standard deviation (SD), n = 3.

^b tested on SARS-CoV-2-ZG/A549^ACE2+TMPRSS2 cells.

^c tested on A549^ACE2+TMPRSS2 cells by Cell Counting Kit-8; n.d.-not determined, ± standard deviation (SD).

^d tested on RD cells, ± standard deviation (SD).

^e tested as an E/Z mixture.

^f Taken from ref. ⁴¹.

Antiviral activities against SARS-CoV-2

Based on the above SAR studies, we further evaluated the antiviral activities of selected macrocyclic α-ketoamides. For this purpose, human A549 lung alveolar epithelial cells transfected with ACE2 + TMPRSS2⁴⁶, referred to as A549-AT cells herein, were infected with a SARS-CoV-2 wildtype strain. Their ATP content was quantified by CellTiterGlo reagent to detect virus-mediated cytopathic effects and their prevention by the inhibitors. We found that 20b, equipped with a hexyl linker and an N-terminal phenylacetyl group, exhibited a moderate activity with an EC₅₀ of 23.3 µM. The change of the N-terminal group had drastic effects on activity: A phenylpropionyl group was considerably more potent (EC₅₀ = 7.0 µM). In view of its high in vitro potency, we were surprised to find the urea 20g as inactive at concentrations up to 50 µM. Analog 20h, carrying a phenylpropionyl group and a hexenyl linker, also displayed little activity (EC₅₀ = 47.6 µM). The most potent macrocycle was the trifluoroacetyl-tert-leucine analog 20f with an EC₅₀ of 1.9 µM (Table 1).

Cellular accumulation

To get a deeper understanding of the links of biochemical and cellular activity, intracellular concentrations of selected macrocyclic α-ketoamides were determined in A549-AT cells and utilized to calculate cellular accumulation (K_p values) by taking the cell number and extracellular compound concentration into account. Additionally, nonspecific binding during the experiments was considered to receive the final K_p values (Table 2 and cellular accumulation data in Table S6). We observed intracellular accumulation for inhibitors 20b, 20e, 20f, and 20h, as reflected by K_p values greater than 1, with compounds 20e (K_p = 3.57) and 20h (K_p = 3.50) showing the highest intracellular levels. Similar to the cellular antiviral activity, the N-terminal phenylpropionyl group in 20e increased cellular accumulation in comparison to the phenylacetyl group contained in 20b. Despite its potent antiviral activity, the tested intracellular concentration of 20f was moderate with a resulting K_p value of 1.67. The urea analog 20g, inactive in cellular antiviral studies against SARS-CoV-2 despite potent biochemical activity, was found to be the only derivate which did not accumulate in cells (K_p = 0.85). This potentially explains the weak cellular activity. Overall, the cellular accumulation were not predicted in the same rank order as cellular antiviral activities. However, cellular accumulation did not account for protein binding in the cytosol or distribution into sub-cellular organelles. It is also conceivable that uptake is altered under infection conditions. These non-quantified contributors to target engagement might account for the differences in rank order.

Table 2.

Cellular accumulation of selected macrocyclic α-ketoamides

Cmpd	Kp valuea
20b	2.42
20e	3.57
20f	1.67
20g	0.85
20h	3.50

^atested in A549^ACE2+TMPRSS2 cells. See also Table S6.

Expansion of enzymatic and antiviral activities of macrocycles to enteroviruses

The main protease M^pro of SARS-CoV-2 and the 3C proteases (3C^pro) of enteroviruses share structural similarities, are both cysteine proteases and cleave after glutamine residues. A distinction lies within the S2 binding pocket, which is enclosed in M^pro, whereas it is comparatively open in 3C^pro and preferentially recognizes phenylalanine rather than leucine residues at this position²⁵. Because the macrocyclic ketoamides incorporate a phenylalanine residue at P2, we hypothesized that tight binding interactions to 3C^pro should be feasible. To verify this, the purified EV-D68 3Cpro was also co-crystallized with the macrocycles 20e and 20g (Fig. 7, data collection and refinement statistics in Table S3). The crystal structures of both inhibitors in complex with 3C^pro of enterovirus D68, solved at resolutions of 1.91 and 2.08 Å, respectively, reveal that the sulfur atom of Cys147 formed a covalent bond (1.8 Å each) with the electrophilic carbon of the P1´, analogous to the SARS-CoV-2 M^pro complex. Within the S1´ subsite of the enzyme, the P1´ carbonyl oxygen occupied the oxyanion hole and accepted three hydrogen bonds (2.8–3.4 Å) from the backbone amides of Cys147, Gln146 and Gly145 for the complex with 20g, but only two hydrogen bonds for the complex with 20e (from Cys147 and Gly145) (Fig. 7). The strong hydrogen bond between the thiohemiketal and His41 shown in complex with M^pro was lost for corresponding His40 residue in the 3C^pro complex, because His40 was oriented outside the pocket. The γ-lactam in the S1 pocket had only two hydrogen bonds with His161 (3.0–3.3 Å) and Thr142 (3.0–3.3 Å) for 20e and 20g, whereas the complex with M^pro was stabilized by three hydrogen bonds (to His163, Phe140, and Glu166). In the S2 subsite, the endocyclic phenyl groups of 20e and 20g extend more deeply into the pocket compared to the M^pro complexes, where the phenyl groups face outwards: The distance between the endocyclic phenyl groups of 20g or 20e and the first residue in the S2 subsite was larger for M^pro (5.1 and 4.9 Å to Met49, respectively) than for 3C^pro (4.0 and 3.8 Å to Pro38, respectively) (Structures with comparisons of the binding modes in Fig. S2). This improved occupation reflects the substrate specificity for Phe in 3C^pro, while M^pro recognized Leu at this position. In the S2 and S3 subsites, interactions between the carbonyl oxygen of the inhibitor and the main chain amide of Val162 (3.0–3.1 Å) as well as a hydrogen bond to Gly164 (3.0–3.4 Å) were observed for both 20e and 20g. M^pro-complexes had corresponding interactions to His164 and His163, respectively. Finally, two weaker interactions between the trifluoroacetyl group of 20g and Gly128 (3.8 Å) and Asn126 (3.4 Å) were present in the S3-S4 subsite, but not with the phenylpropionyl group of 20e. In summary, the structural characterization of ketoamide macrocycles demonstrated multiple similarities in their binding to 3C^pro and M^pro. While less interactions were exerted at the S1 and S1’ sites of 3C^pro, the occupancy of the S2 site was improved. This provides a rational and understanding why the compounds may act as dual inhibitors of these distinct viral proteases.

Fig. 7. Co-crystal structure of macrocyclic ketoamides in complex with 3C^pro of EV-D68.

Upper panels show the complex with 20e (A shown as sticks, B as surface; PDB:29CT), lower panels show the complex with 20g (C shown as sticks, D as surface; PDB:29DA).

In order to provide functional evidence, we probed the ability of the macrocycles to inhibit the 3C^pro of the enterovirus D68 in a FRET-based enzymatic assay. Five compounds inhibited the 3C^pro of D68 with IC₅₀ values of 1.4–2.9 µM (Table 1). For 20b and 20e, we could also show inhibition of the 3C^pro of the enterovirus EV-A71 with IC₅₀ values of 6.9 and 4.4 µM, respectively (enzymatic inhibition data in Table S7). The most potent inhibitor was 20e, carrying a hexyl linker and a phenylpropionyl group at the P4 position. Thus, a dual inhibition of coronaviral and enteroviral proteases could be achieved. Beyond that, we also observed inhibition of human cathepsins, structurally related cysteine proteases, for the tested inhibitors 20b, 20e, and 20h (inhibition of human cathepsins in Table S8). The inhibition was particularly strong for cathepsin L. Because cathepsin L plays a role in the entry of SARS-CoV-2 Omicron variants, such as the BA.5 strain, its inhibition has been regarded as favorable to enhance the antiviral effect^9,47; other cathepsins are rather seen as off-targets.

We also tested macrocycles from the first series with an exocyclic nitrile warhead. Both 9b-K and 9a-K were inactive against EV-D68 3C^pro. Thus, the insufficient activity of the exocyclic nitriles against M^pro was also reflected in the related enteroviral protease.

We finally tested the macrocycles 20b and 20e–g in cellular assays for anti-enteroviral activity. RD cells were infected with EV-D68, and the activity of the compounds was determined by a cell viability assay using crystal violet. The analogs 20b, 20e and 20 g inhibited this virus with EC₅₀‘s of 5.0 ± 0.6, 4.3 ± 0.3, and 1.5 ± 0.1 µM, respectively. To our surprise, 20f exhibited very potent activity with an EC₅₀ of 0.033 ± 0.004 µM against EV-D68 (Fig. 8A). We therefore expanded the profiling of 20f in HeLa-R19 cells that were infected with EV-D68, EV-A71 and Coxsackievirus CVB3 reporter viruses. Again, 20 f was able to potently inhibit these viruses with EC₅₀‘s of 0.035, 0.13, and 0.15 µM, respectively (Fig. 8B, antiviral effects of 20f on EV-D68, EV-A71 and Coxsackievirus-B3 in Supplementary Fig. 3). Given that the cellular cytotoxicity of the compound was >100 µM, the selectivity indices (SI) of 20 f for the antiviral activities against EV-D68, EV-A71 and Coxsackievirus CVB3 were >1030, >769, and >667.

Fig. 8. Antiviral effects of macrocyclic ketoamides on EV-D68, EV-A71 and Coxsackievirus-B3.

A Concentration-response curves of the macrocyclic inhibitors 20b and 20e–g. The antiviral effect against EV-D68 in RD cells was measured by a cell viability assay using crystal violet. B Antiviral effects of 20f detected by the luciferase activity of corresponding reporter viruses. All measurements were performed with n = 3 biologically independent experiments. The mean values are depicted with error bars representing the standard deviation (SD).

In summary, an extension of macrocycles rationally designed M^pro inhibitors towards enteroviral 3C proteases was possible. It was associated with some loss of in vitro potency, but a remarkably high cellular antiviral activity in different assay systems for the analog 20f.

In vitro ADME properties of 20f

As 20f exhibited broader spectrum activity against SARS-CoV-2 and three distinct enteroviruses, we were interested in its in vitro ADME properties. We found that 20f was metabolically stable in both mouse and human microsomes and exhibited excellent plasma stability. 20f showed high plasma protein binding of around 99% in mouse and 99.1% in human plasma (Table 3). Future studies would aim to lower plasma protein binding while preserving the excellent bioactivity profile.

Table 3.

ADME properties of 20f

Property	Value
Clint [µl/min/mg protein] mouse	< 23
t1/2 [h] mouse microsomes	> 60
Clint [µl/min/mg protein] human	< 23
t1/2 [h] human microsomes	> 60
t1/2 [h] mouse plasma	> 240
t1/2 [h] human plasma	> 240
plasma protein binding mouse [%]	99.0 ± 0.3
plasma protein binding human [%]	99.1 ± 0.4

Discussion

We explored the feasibility of macrocyclization for rationally designed inhibitors of the coronaviral main protease. Based on the previously reported linear compound 12c, a first series of analogues bearing an exocyclic nitrile warhead and a macrocycle bridging P1 and P3 was designed. Four compounds were isolated as pure diastereomers by preparative HPLC and displayed only low inhibitory activity against M^pro of SARS-CoV-2, and none against 3C^pro of EV-D68. In spite of this, a co-crystallization of 9b-K with M^pro at a high resolution of 1.5 Å was successful. The structure confirmed that the compound occupied the P1′ to P2 subsites as expected, and it also showed the intended covalent binding to the nitrile warhead. Thus, the structure provided support for our initial design concept. However, it also revealed important limitations. Multiple hydrogen bonds, namely to Gly143, Ser144, His41, Glu189, and Glu166, were elongated by 0.3 Å or more. Thus, the suboptimal fit or the lack of substituents engaging the S1/S3/S4 subsites significantly diminished overall potency. These observations provided clear direction for subsequent optimization efforts. On a general note, they show that characterizing protein-ligand structures even for poorly active compounds is both feasible and valuable. This implies to re-consider drug profiling workflows that often place structural biology experiments after filtering for biochemical activity, rather than conducting them in parallel.

Our second design hypothesis was that linking the P1′–P2 residues to a macrocycle while retaining an endocyclic ketoamide warhead could yield potent inhibitors. This led to the synthesis of macrocyclic α-ketoamides 20a–d, featuring linkers of varying lengths from pentyl to octyl. Among these 20b with a hexyl linker demonstrated an IC₅₀ of 1.37 µM against M^pro and an antiviral EC₅₀ of 23.3 µM. A key observation from co-crystal structures of 20a–d was that a phenacetyl group did not properly occupy the S3/S4 subsites. We therefore designed 20e, bearing a phenylpropionyl capping group, which showed an improved IC₅₀ of 0.5 µM and a threefold increase in anti-SARS-CoV-2 activity compared with 20b. Subsequent modifications as in 20f and 20g yielded improved IC₅₀ values of 0.37 and 0.36 µM, respectively. Interestingly, 20g failed to show measurable antiviral activity at 50 µM, highlighting that structural features beyond enzyme inhibition play an essential role. We assume that the presence of additional hydrogen-bonding groups impaired cell permeability, as observed experimentally (Table 2). A similar observation was made in our previous P3/P4 capping group optimization studies⁴⁰, where most—but not all—ureas were inactive in cellular assays. In contrast, 20f exhibited an antiviral EC₅₀ of 1.9 µM against SARS-CoV-2, underscoring the importance of an optimally configured P3/P4 capping unit. The unsaturated cycle 20h was less active than its saturated analogs 20e. Moreover, because E/Z stereoselectivity could not be achieved in the synthetic route, saturated macrocycles were preferred. In sum, a successful macrocyclization of ketoamide-based M^pro inhibitors could be achieved. A clear advantage over the best linear compounds is not obvious at this stage. However, the macrocyclic scaffold itself has not been fully explored. While optimization efforts in this study mostly focused on finding the optimal length of an alkyl linker, discontinuous activity trends in both cellular and enzymatic assays as well as the crystal structures imply high conformational plasticity of the macrocyclic scaffold. The decoration of the cycle with heteroatoms or substituents thus offers further potential for improvements in activity through conformational restrictions, as shown in systematic macrocycle optimization programs against different targets^48,49.

Given the (limited) similarities of 3CL^pro and 3C^pro, we further evaluated the activities of macrocyclic α-ketoamides against the EV-D68 protease as well as in corresponding antiviral assays. Enteroviruses belong to the Picornaviridae and cause a variety of clinical manifestations that include also serious diseases, such as meningitis, viral encephalitis, myopericarditis, acute flaccid myelitis (AFM), hand, or neonatal sepsis⁵⁰. EV-D68 was the primary causative agent in the 2014 outbreak, which was associated with a surge in cases of AFM. Although AFM occurs only in a subset of cases, EV-D68 infections more commonly present as respiratory illness⁵¹. EV-A71 is the principal causative agent of hand, foot, and mouth disease (HFMD), a self-limiting febrile illness that has caused large-scale outbreaks in the Asia–Pacific region⁵². Coxsackievirus B3 (CVB3), a strain belonging to the EV-B species, is the etiological agent of viral myocarditis and sudden cardiac death, and can be particularly fatal in newborns and young children, where it often manifests with septic-like disease. Given the severity and diversity of diseases caused by these pathogens, the development of broad-spectrum antiviral agents covering enteroviruses would be highly valuable.

The principal feasibility of a re-purposing of coronalviral protease inhibitors towards enteroviruses has been demonstrated before^26,27. However, while our macrocycles were weaker inhibitors than nirmatrelvir for SARS-CoV-2 M^pro and rupintrivir for EV 3C^pro, the two positive controls did not exert broad-spectrum activity against both proteases. This demonstrates that dual target inhibition is non-trivial. In previous studies, the activities of compounds in the “second” indication was mostly weaker or equal to the one that guided their design. In contrast, we found inhibitors whose enteroviral activity surpassed those against the coronavirus. In particular, 20f demonstrated the highest potency, exhibiting EC₅₀ values of 0.033–0.15 µM against three enteroviruses, thereby highlighting its promise as a broad-spectrum antiviral candidate. The macrocycle 20f incorporates a phenylalanine residue at the S2 pocket, which we identified as an optimal substituent in earlier studies. The large and open architecture of the 3C^pro S2 pocket accommodates this residue effectively, thereby facilitating proper alignment of the inhibitor and enabling covalent interaction with the catalytic cysteine of the Cys–His-Asp triad. In addition, the trifluoroacetyl–tert-leucine capping group of 20f seems to engage in favorable interactions within the 3C^pro active site, given its superiority over other P4 residues. A first ADME profiling indicated sufficient metabolic and chemical stability of 20f. These findings highlight the structural features critical for potent protease inhibition and encourage us to generate further druggable macrocyclic antiviral compounds based on the promising hit 20f.

Conclusion

A structure-based design of macrocyclic inhibitors targeting the SARS-CoV-2 main protease was achieved by two different approaches, that were both successful in terms of target binding. However, the inactivity of the nitrile series showcases how subtle design flaws induce drastic functional impairments. In contrast, the endocyclic α-ketoamides bound, were functionally active and exerted anti-coronaviral activity. Notably, 20f could be “repurposed” to inhibit the related enteroviral 3C proteases and exert potent antiviral activity in the nM range. While most studies aim to achieve broad-spectrum antiviral activity with compounds acting on host targets, the macrocycles prepared here demonstrate that this is also possible with direct-acting antivirals.

Methods

General information on chemical synthesis

Commercial reagents were used as received, and all other reagents were prepared using known literature procedures. All solvents used for reactions, workups, and purifications had the HPLC purity grade. Dried solvents were purchased in water-free form (99.5%, extra dry, absolute, AcroSeal^TM, ACROS Organics^TM) and used as received. Reactions were either monitored by Liquid Chromatography-coupled Mass Spectrometry (LCMS) analysis or thin-layer chromatography (TLC) on “TLC Silica gel 60 F254” plates (Merck) and visualized by staining with aqueous basic KMnO₄. LCMS was conducted with an Agilent® 1260 HPLC System with a DAD detector and an Agilent® 6130 quadrupole mass detector with Electrospray Ionization (ESI) (MeCN/H₂O + 0.1% HCOOH). Flash chromatography was performed on silica gel 60 (technical grade, pore size 60 Å, 40–63 μm, 230–400 mesh, Supelco®). Reverse-Phase High Pressure Liquid Chromatography (RP-HPLC) was performed with a Phenomenex Luna C18 RP-column 00G-4252-P0-AX, 5 μm, 100 Å, 250 × 21.2 mm (flow rate 10 mL/min, max. loading 100 mg crude) coupled to a Thermo Fisher Scientific® Dionex Ultimate 3000 HPLC-System using a ternary solvent mixture (MeCN/H₂O + 0.1% CF₃ COOH). Product-containing fractions were identified by LCMS and lyophilized to dryness. Purity of final compounds was measured on SHIMADZU HPLC (Model Number: CBM-20A) with Gemini® 3 μm NX-C18-110 Å, LC-Column 50 × 2 m.m (flow rate 1 mL/min). For Nuclear Magnetic Resonance (NMR) spectroscopic analysis, Bruker Avance III or Bruker Avance III HD spectrometers were employed and ¹H NMR spectra were recorded at 500 or 700 MHz, respectively 126 and 176 MHz for ¹³C NMR spectra. Chemical shifts are reported in parts per million (ppm) using the residual non deuterated solvent resonance for proton measurements and the solvent resonance for carbon measurements as internal standards (D₃COD: ¹H = 3.31 ppm and ¹³C = 49.00 ppm; DMSO-d₆: ¹H = 2.50 ppm and ¹³C = 39.52 ppm, CD₃CN: ¹H = 1.94 ppm and ¹³C = 1.32, 118.26 ppm). Data are reported as follows: chemical shift (multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet), coupling constant(s) in Hz, integration). High resolution mass spectrometry (HRMS) was performed using a Dionex Ultimate 3000 HPLC system equipped with a DAD detector and a Bruker maXis HD QTOF mass detector with electrospray ionization (ESI). Samples were injected via an Ultimate 3000RS auto sampler (Thermo Fisher Scientific) and data are reported in the form of mass-to- charge ratio (m/z). All the final compounds used for biological studies are purified based on prep-HPLC and having purity more than 95%. Analytical HPLC-MS traces of several compounds exhibit a second peak with a mass that is +18 Da heavier than the expected mass; this peak is ascribed to a reversible hydrate formation upon the addition of water to the keto group.

Synthesis of (S)-1-amino-5-(benzyloxy)-1,5-dioxopentan-2-aminium trifluoro acetate

N-Boc-D-Glu(OBn)-OH (3 g, 8.9 mmol) was dissolved in 1,4-dioxane, and then NH₄HCO₃ (1.1 g, 14.08 mmol), pyridine (0.9 mL), Boc₂O (2.7 mL, 12.38 mmol) were added into solution under rt for 16 h. 1,4-Dioxane was removed by rotary evaporation. The residue was dissolved in EtOAc and extracted with water. Organic layer was collected and dried over Na₂SO₄, followed by the concentration and purification by flash column chromatography, eluting with 20–30% of ethylacetate/cyclohexane afforded benzyl (S)-5-amino-4-((tert-butoxycarbonyl) amino)-5-oxopentanoate a white solid. ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 7.42–7.30 (m, 5H), 7.32–7.22 (m, 1H), 7.00 (s, 1H), 6.81 (d, J = 8.2 Hz, 1H), 5.17–5.01 (s, 2H), 3.95 (ddd, J = 13.9, 7.8 Hz, 1H), 2.45–2.30 (m, 2H), 1.96–1.69 (m, 2H), 1.45–1.28 (m, 9H). ESI-MS (m/z): 337 [M + H]⁺.

To the stirred solution of benzyl (S)-5-amino-4-((tert-butoxycarbonyl) amino)-5-oxopentanoate (1.3 g, 3.389 mmol) in CH₂Cl₂ (10 mL), Trifluoro acetic acid (4 mL, excess) was added at ice-cold temperature. The reaction mixture was stirred for 6 h. After confirming the completion of the reaction based on LCMS, it was concentrated to dryness and co-evaporated with methanol. This resulting solid was used for further reaction without purification (1.35 g, 100%). ESI-MS Da (m/z): 237 Da [M + H]⁺.

Synthesis of benzyl (S)-5-amino-4-((S)-2-(3-((tert-butoxycarbonyl) amino)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-5-oxopentanoate (2)

To a solution of (S)-2-(3-((tert-butoxycarbonyl) amino)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanoic acid 1 (750 mg, 2.33 mmol) in DMF (10 mL), HATU (1.8 g, 4.66 mmol) was added at 0 °C. The reaction mixture was stirred at 0 °C under a nitrogen atmosphere, and after few minutes, a pre-stirred solution of (S)-1-amino-5-(benzyloxy)-1,5-dioxopentan-2-aminium trifluoro acetate (730 mg, 3.09 mmol) and triethylamine (1.41 g,13.97 mmol) in DMF (5 mL) at 0 °C was added. The reaction was continued for 6 h at 0 °C. Progress of the reaction was monitored based on TLC and LCMS. The reaction mixture was concentrated, diluted with CH₂Cl₂ (50 mL), and washed with water. The resulting organic layer was washed with brine solution and dried over Na₂SO₄. The resulting organic was evaporated to dryness and purified on a flash column over a C18 column (25 − 45% CH₃CN/H₂O). The product was isolated with a 90% yield (1.13 g). ¹H NMR (600 MHz, DMSO-d₆, 298 K): δ 8.51 (d, J = 8.1 Hz, 0.5 H), 8.30 (d, J = 8.0 Hz, 0.5 H), 7.85–7.66 (m, 2H), 7.47–7.27 (m, 6H), 7.07 (s, 1H), 6.31–6.21 (m, 1H), 5.53 (t, J = 14.9 Hz, 1H), 5.07 (sx2, 2H), 4.26–4.13 (m, 1H), 2.42–2.26 (m, 2H), 2.04–1.92 (m, 2H), 1.81 (d, J = 15.0 Hz, 2H), 1.46 (sx2, 9H), 0.48 (t, J = 14.8 Hz, 1H), 0.35–0.24 (m, 2H), 0.15–0.06 (m, 1H), −0.02–0.102 (m, 1H). ESI-MS (m/z): 541 Da [M + H]⁺.

Synthesis of benzyl (S)-5-amino-4-((S)-2-(3-amino-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-5-oxopentanoate TFA salt (3)

To the stirred solution of (S)-5-amino-4-((S)-2-(3-((tert-butoxycarbonyl) amino)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-5-oxopentanoate 2 (650 mg, 1.20 mmol) in CH₂Cl₂ (10 mL), Trifluoro acetic acid (2 mL, excess) was added at ice-cold temperature. The reaction mixture was stirred for 6 h. After confirming the completion of the reaction based on LCMS, it was concentrated to dryness and co-evaporated with methanol. This resulting solid of 3 was used for further reaction without purification (663 mg, 100%). ESI-MS (m/z): 441 Da [M + H]⁺.

Synthesis of benzyl (S)-5-amino-4-((S)-2-(3-(5-(((benzyloxy)carbonyl) amino) pentanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-5-oxopentanoate (4a)

Following the procedure described for the preparation of compound 2 with a slight modification of conducting the reaction with 3 (300 mg, 0.541 mmol) and 5-(((benzyloxy) carbonyl) amino) pentanoic acid (252 mg, 1.083 mmol), afforded 4a (236 mg, 65%) as a white solid. ¹H NMR (500 MHz, CD₃CN 298 K) δ: 8.54–8.44 (m, 1H), 8.37–8.27 (m, 1H), 7.56–7.23 (m, 10H), 6.95 (s, 0.5H), 6.70 (s, 0.5H), 6.36–6.22 (m, 1H), 6.01–6.13 (m, 1H), 5.82–5.70 (m, 1H), 5.37 (t, J = 7.6 Hz, 0.3H), 5.16 (dd, J = 8.4, 6.9 Hz, 0.68H), 5.10–5.01 (m, 4H), 4.35–4.27 (m, 1H), 3.20–3.04 (m, 2H), 2.50–2.25 (m, 4H), 2.22–2.04 (m, 1H), 1.93–1.74 (m, 3H), 1.68–1.52 (m, 2H), 1.51–1.39 (m, 2H), 0.60–0.48 (m, 1H), 0.41–0.27 (m, 2H), 0.11–0.02 (m, 1H), 0.00 to −0.09 (m, 1H). ESI-MS (m/z): 674 Da [M + H]⁺.

Synthesis of benzyl (S)-5-amino-4-((S)-2-(3-(6-(((benzyloxy)carbonyl) amino) hexanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-5-oxopentanoate (4b)

Following the procedure described for the preparation of compound 2 with a slight modification of conducting the reaction with 3 (660 mg, 1.19 mmol) and 6-(((benzyloxy)carbonyl) amino) hexanoic acid (632 mg, 2.385 mmol) afforded 4b (605 mg, 74%) as a white solid. ¹H NMR (500 MHz, CD₃CN, 298 K) δ: 8.50–8.28 (m, 2H), 7.44–7.32 (m, 7H), 7.25–7.00 (m, 2H), 6.81–6.65 (m, 2H), 6.41–6.21 (m, 2H), 5.88–5.56 (m, 2H), 5.22 (ddd, J = 15.3, 12.5, 6.6 Hz, 1H), 5.11–4.98 (m, 4H), 4.32–4.16 (m, 1H), 3.19–3.00 (m, 2H), 2.51–2.23 (m, 4H), 2.14–2.00 (m, 2H), 1.92–1.79 (m, 2H), 1.70–1.55 (m, 2H), 1.53–1.42 (m, 2H), 1.39–1.25 (m, 2H), 0.63–0.51 (m, 1H), 0.43–0.28 (m, 2H), 0.16–0.06 (m, 1H), 0.04 to −0.10 (m, 1H). ESI-MS (m/z): 688 Da [M + H]⁺.

Synthesis of (S)-5-amino-4-((S)-2-(3-(5-aminopentanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-5-oxopentanoic acid (5a)

To a solution of 4a (45 mg, 0.0669 mmol) in methanol (1.5 mL) was added Pd/C (5 mg, 10% on carbon) under nitrogen. The suspension was degassed under vacuum and purged with hydrogen several times. The mixture was stirred hydrogen atmosphere at 20 °C for 5 h. LCMS showed the starting material was consumed completely. The mixture was filtered, the solid was washed with methanol (5 mLX3). The combined filtrate was concentrated in vacuum to give 5a (24 mg, 80%) as white solid, which was used for the next step without further purification. ¹H NMR (600 MHz, CD₃OD, 298 K) δ: 8.57–8.16 (m, 1H), 7.55–7.40 (m, 1H), 7.40–7.27 (m, 1H), 6.38 (dt, J = 11.5, 7.2 Hz, 1H), 5.48–5.31 (m, 1H), 4.37–4.23 (m, 1H), 3.05–2.90 (m, 2H), 2.82–2.43 (m, 3H), 2.37–1.90 (m, 8H), 1.83–1.53 (m, 6H), 0.62 (dt, J = 12.1, 7.3 Hz, 1H), 0.52–0.35 (m, 2H), 0.20–0.13 (m, 1H), 0.08–0.01 (m, 1H). ESI-MS (m/z): 450 Da [M + H]⁺.

Synthesis of cyclic amide (6a)

A solution of 5a (50 mg, 0.111 mmol) in DMF (20 mL), HATU (64 mg, 0.167 mmol) and TEA (33.8 mg, 0.335 mmol) was added at ice cold temp. Reaction was stirred at ice cold temp under dry conditions. Reaction was continued for 5 h in the same temp. Progress of the reaction was monitored based on TLC and LCMS. After 5 h, the reaction was stopped and neutralized with 10% TFA in acetonitrile. Resulted crude was purified on prep-HPLC to give the 6a as separated diastereomers (16 + 20 mg, 75%) as white solid. Dia-1: ¹H NMR (700 MHz, CD₃OD, 298 K) δ: 8.19 (d, J = 7.1 Hz, 1H), 7.40 (t, J = 9.3 Hz, 1H), 6.42 (t, J = 7.2 Hz, 1H), 5.14–5.05 (m, 1H), 4.37 (dd, J = 9.9, 3.1 Hz, 1H), 2.97–2.86 (m, 1H), 2.56–2.24 (m, 4H), 2.24–2.15 (m, 1H), 2.14–1.97 (m, 3H), 1.90–1.41 (m, 6H), 1.37–1.24 (m, 1H), 0.71–0.61 (m, 1H), 0.50–0.40 (m, 2H), 0.16 (d, J = 8.7 Hz, 1H), 0.07 (t, J = 12.7 Hz, 1H). ESI-MS (m/z): 432 Da [M + H]⁺.

Dia-2: ¹H NMR (700 MHz, CD₃OD, 298 K) δ: 8.36–8.28 (m, 1H), 7.45 (d, J = 5.9 Hz, 1H), 6.45–6.41 (m, 1H), 4.84–4.78 (m, 1H), 4.34–4.26 (m, 1H), 3.37–3.33 (m, 1H), 3.24–3.17 (m, 1H), 2.86–2.79 (m, 1H), 2.61–2.53 (m, 1H), 2.43–2.41 (m, 1H), 2.38–2.23 (m, 3H), 2.07–2.13 (m, 1H), 2.07–1.98 (m, 2H), 1.88–1.79 (m, 1H), 1.77–1.69 (m, 1H), 1.64–1.51 (m, 2H), 0.59–0.51 (m, 1H), 0.42 (ddd, J = 9.6, 7.4, 5.0 Hz, 1H), 0.39–0.31 (m, 1H), 0.15–0.09 (m, 1H), −0.04 to −0.10 (m, 1H). ESI-MS (m/z): 432 Da [M + H]⁺.

Synthesis of benzyl (S)-4-((S)-2-(3-(5-(((benzyloxy)carbonyl) amino) pentanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-4-cyanobutanoate (7a)

Compound 4a (180 mg, 0.282 mmol), inner salt (Burgess Reagent; 135 mg, 0.567 mmol) was added to a solution in CH₂Cl₂ (5 mL). After the reaction mixture had been stirred at 25 °C for 4 h, the reaction mixture was diluted and quenched by saturated aqueous sodium bicarbonate solution (2 mL). The separated organic phase was concentrated and purified on reverse phase flash system flash column over C18 column (30–50% H₂O/ACN), which gave 7a (140 mg, 75%) of as a white solid. ¹H NMR (700 MHz, DMSO-d₆, 298 K) δ : 9.18–8.99 (m, 2H), 8.22 (t, J = 5.8 Hz, 1H), 7.45–7.18 (m, 11H), 6.29 (dt, J = 12.6, 7.2 Hz, 1H), 5.57–5.43 (m, 1H), 5.14–4.96 (sX2, 4H), 4.88–4.76 (m, 1H), 3.01–2.96 (m, 2H), 2.48–2.35 (m, 4H), 2.14–1.95 (m, 3H), 1.81–1.70 (m, 1H), 1.63–1.35 (m, 4H), 0.55–0.43 (m, 1H), 0.39–0.26 (m, 2H), 0.10 (dd, J = 9.2, 4.3 Hz, 1H), 0.02 (dd, J = 13.3, 11.3 Hz, 1H). ESI-MS (m/z): 656 Da [M + H]⁺.

Synthesis of benzyl (S)-4-((S)-2-(3-(6-(((benzyloxy) carbonyl) amino) hexanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-4-cyanobutanoate (7b)

Following the procedure described for the preparation of compound 7a with a slight modification of conducting the reaction with 4b (605 mg, 0.88 mmol) and inner salt (Burgess Reagent; 419 mg, 1.76 mmol) afforded 7b (302 mg, 51%) as a white solid. ¹H NMR (500 MHz, CD₃CN, 298 K) δ: 8.43–8.28 (m, 2H), 7.43–7.27 (m, 10H), 7.23–7.18 (m, 1H), 6.34–6.21 (m, 1H), 5.63 (s, 1H), 5.45–5.26 (m, 1H), 5.14–4.98 (m, 4H), 4.86–4.69 (m, 1H), 3.16–2.95 (m, 2H), 2.54–2.30 (m, 4H), 2.15–2.01 (m, 2H), 1.92–1.87 (m, 2H), 1.61 (dq, J = 14.2, 7.3 Hz, 2H), 1.51–1.41 (m, 2H), 1.39–1.23 (m, 2H), 0.61–0.49 (m, 1H), 0.42–0.30 (m, 2H), 0.12–0.04 (m, 1H), 0.03 to −0.06 (m, 1H), ESI-MS (m/z): 670 Da [M + H]⁺.

Synthesis of (S)-4-((S)-2-(3-(5-aminopentanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-4-cyanobutanoic acid (8a)

Following the procedure described for the preparation of 5a with a slight modification of conducting the reaction with 7a (130 mg, 0.198 mmol) and Pd/C (15 mg, 10% on charcoal) afforded 8a (70 mg, 82%) as a white solid which was used for further reaction without purification. ESI-MS (m/z): 432 Da [M + H]⁺.

Synthesis of (S)-4-((S)-2-(3-(6-aminohexanamido)-2-oxopyridin-1(2H)-yl)-3-cyclopropylpropanamido)-4-cyanobutanoic acid (8b)

Following the procedure described for the preparation of 5a with a slight modification of conducting the reaction with 7b (300 mg, 0.4484 mmol) and Pd/C (32 mg, 10% on charcoal), afforded 8b (170 mg, 85%) as a white solid. ¹H NMR (500 MHz, CD₃OD, 298 K) δ : 8.34–8.24 (m, 1H), 7.53–7.42 (m, 1H), 6.38 (tt, J = 9.3, 7.3 Hz, 1H), 5.47–5.36 (dt, J = 17.3, 8.6 Hz, m, 1H), 4.85 (dd, J = 8.5, 6.1 Hz, 1H), 4.79 (t, J = 7.3 Hz, 1H), 2.96–2.87 (m, 2H), 2.56–2.43 (m, 2H), 2.38–2.26 (m, 2H), 2.17–1.89 (m, 4H), 1.78–1.63 (m, 4H), 1.52–1.39 (m, 2H), 0.70–0.57 (m, 1H), 0.50–0.39 (m, 2H), 0.22–0.13 (m, 1H), 0.11–0.03 (m, 1H). ESI-MS (m/z): 446 Da [M + H]⁺.

Synthesis of macrocyclic nitrile (9a)

Following the procedure described for the preparation of compound 6a with a slight modification of conducting the reaction with compound 8a (70 mg, 0.162 mmol) and HATU (95 mg, 0.2517 mmol), afforded compound 9a as separated diastereomers (12 + 17.8 mg, 43%) as a white solid.

(2 R,5S)-2-(cyclopropylmethyl)-3,8,14,20-tetraoxo-1,4,9,15-tetraazabicyclo[14.3.1]icosa-16,18-diene-5-carbonitrile (9a-H)

HPLC purity: 95.5%,¹H NMR (500 MHz, CD₃OD, 298 K): δ 8.17 (dd, J = 7.3, 1.3 Hz, 1H), 7.40 (dd, J = 7.0, 1.3 Hz, 1H), 6.43–6.30 (m, 1H), 4.99 (dd, J = 9.7, 5.8 Hz, 1H), 4.65 (dd, J = 10.5, 3.5 Hz, 1H), 3.15–3.05 (m, 1H), 2.98 (s, 1H), 2.60–2.35 (m, 3H), 2.32–2.16 (m, 2H), 2.11–1.97 (m, 2H), 1.92–1.69 (m, 2H), 1.60–1.43 (m, 2H), 0.68–0.39 (m, 2H), 0.45 (tdd, J = 13.3, 9.0, 5.1 Hz, 2H), 0.17 (dt, J = 8.8, 4.9 Hz, 1H), 0.09–0.01 (m, 1H). ¹³C NMR (176 MHz, CD₃OD, 298 K) δ: 174.91, 172.09, 159.69, 132.67, 126.33, 119.37, 108.03, 66.20, 57.86, 41.68, 39.70, 34.16, 31.75, 29.07, 28.38, 24.41, 17.65, 8.76, 5.37, 4.54, 0.10. HRMS: calcd. for C₂₁H₂₈N₅O₄ [M + H]⁺, m/z = 414.2141 Da; found, m/z = 414.2136 Da.

(2S,5S)-2-(cyclopropylmethyl)-3,8,14,20-tetraoxo-1,4,9,15-tetraazabicyclo[14.3.1]icosa-16,18-diene-5-carbonitrile (9a-K)

HPLC purity: 96.5%, ¹H NMR (500 MHz, CD₃OD, 298 K) δ: 8.18 (dt, J = 8.7, 4.4 Hz, 1H), 7.41 (dd, J = 7.0, 1.4 Hz, 1H), 6.43–6.39 (m, 1H), 4.98–4.85 (m, 2H), 3.38–3.32 (m, 1H), 2.92 (ddd, J = 13.4, 11.0, 7.1 Hz, 1H), 2.54–2.33 (m, 4H), 2.27–1.94 (m, 4H), 1.85–1.69 (m, 2H), 1.59–1.41 (m, 2H), 0.70–0.60 (m, 1H), 0.53–0.38 (m, 2H), 0.28–0.12 (m, 1H), 0.06–0.01 (m, 1H). ¹³C NMR (126 MHz, CD₃OD, 298 K) δ: 173.21, 172.71, 169.93, 157.74, 128.27, 124.78, 118.15, 106.22, 64.14, 56.23, 37.95, 36.67, 32.64, 29.94, 27.64, 23.14, 16.03, 7.02, 3.67, 3.20, −1.45. ESI-HRMS: calcd. for C₂₁H₂₈N₅O₄ [M + H]⁺, m/z = 414.2141 Da; found, m/z = 414.2138 Da.

Synthesis of macrocyclic nitrile (9b)

Following the procedure described for the preparation of 6a with a slight modification of conducting the reaction with 8b (85 mg, 0.191 mmol) and HATU (108 mg, 0.286 mmol) afforded 9b as separated diastereomers (24 + 28 mg, respectively, 64% in sum) as a white solid.

(2R,5S)-2-(cyclopropylmethyl)-3,8,15,21-tetraoxo-1,4,9,16-tetraazabicyclo[15.3.1]henicosa-17,19-diene-5-carbonitrile (9b-H):

¹H NMR (500 MHz, CD₃OD, 298 K): δ 8.17 (dd, J = 7.4, 1.7 Hz, 1H), 7.44 (dd, J = 7.0, 1.7 Hz, 1H), 6.45–6.35 (m, 1H), 4.99 (dd, J = 9.7, 5.8 Hz, 1H), 4.65 (dd, J = 10.2, 4.0 Hz, 1H), 3.08 (ddd, J = 13.5, 6.9, 3.8 Hz, 1H), 2.96 (d, J = 5.9 Hz, 1H), 2.52–2.35 (m, 3H), 2.30–2.19 (m, 2H), 2.14–2.04 (m, 1H), 2.02–1.94 (m, 1H), 1.73 (qd, J = 13.9, 7.1 Hz, 2H), 1.57–1.37 (m, 4H), 0.67–0.55 (m, 1H), 0.51–0.36 (m, 2H), 0.17 (dt, J = 9.3, 5.0 Hz, 1H), 0.07 to −0.00 (m, 1H). ¹³C NMR (176 MHz, CD₃OD, 298 K): δ 174.00, 173.07, 170.01, 158.15, 132.14, 128.28, 126.24, 117.78, 106.09, 65.16, 57.78, 39.72, 38.30, 36.53, 32.22, 30.64, 27.96, 25.18, 16.08, 7.15, 3.76, 2.90, −1.46. HRMS: calcd. for C₂₂H₃₀N₅O₄ [M + H]⁺, m/z = 428.2297 Da; found, m/z = 428.2298 Da.

(2S,5S)-2-(cyclopropylmethyl)-3,8,15,21-tetraoxo-1,4,9,16-tetraazabicyclo[15.3.1]henicosa-17,19-diene-5-carbonitrile (9b-K)

HPLC purity: 97.6%, ¹H NMR (500 MHz, CD₃OD, 298 K): δ 8.21 (dd, J = 7.4, 1.7 Hz, 1H), 7.45 (dd, J = 7.1, 1.7 Hz, 1H), 6.39 (t, J = 7.2 Hz, 1H), 5.06 (t, J = 7.7 Hz, 1H), 4.97 (dd, J = 9.9, 4.7 Hz, 1H), 3.49–3.40 (m, 1H), 3.01–2.90 (m, 2H), 2.53–2.30 (m, 3H), 2.22–2.04 (m, 3H), 1.97–1.63 (m, 4H), 1.61–1.27 (m, 5H), 0.75–0.63 (m, 1H), 0.57–0.40 (m, 2H), 0.22–0.06 (m, 2H). ¹³C NMR (176 MHz, MeOD, 298 K) δ: 173.74, 172.82, 170.13, 157.89, 130.95, 128.45, 125.69, 118.11, 105.91, 62.89, 56.55, 55.82, 38.71, 36.17, 33.24, 30.53, 28.82, 27.91, 24.92, 16.08, 15.64, 15.54, 7.09, 3.59, -1.49. HRMS: calcd. for C₂₂H₃₀N₅O₄ [M + H]⁺, m/z = 428.2297 Da; found, m/z = 428.2297 Da.

Synthesis of methyl (S)-3-(3-allylphenyl)-2-((tert-butoxycarbonyl) amino) propanoate (10)

A two-neck round-bottom flask was charged (S)-3-(3-bromophenyl)-2-((tert-butoxycarbonyl) amino) propanoate (1.14 g, 3.19 mmol), CsF (1.93 g, 12.69 mmol), in THF (50 mL), de-gassed with argon gas for 15 min. 2-allyl-4,4,5,5-tetramethyl-1,3,2-dioxaborolane 2a (1.61 g, 8.39 mmol) and Pd(PPh₃)₄ (0.3 g, 0.259 mmol) were added and the resulting reaction mixture was heated to reflux for 24 h. At the end of the reaction (TLC monitoring), crude was dissolved in 100 ml of ethyl acetate and washed with water. Aqueous layer was extracted ethyl acetate (2 × 50 ml) and combined organic layer was washed with brine and dried over Na₂SO₄. Resulted solution was evaporated to dryness and purified over column chromatography with 1:9 mixture EA/Cyclohexane on silica gel, which gave methyl (S)-3-(3-allylphenyl)-2-((tert-butoxycarbonyl) amino) propanoate (10). Yield: 2.7 g, 93%. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 7.31–7.18 (m, 2H), 7.10–7.01 (m, 3H), 5.97–5.90 (m, 1H), 5.09–5.03 (m, 2H), 4.18–4.09 (m, 1H), 3.65–3.55 (m, 3H), 3.33 (s, 2H), 2.98–2.90 (m, 1H), 2.88–2.80 (m, 1H), 1.37–1.28 (s, 9H). ESI-MS (m/z): 320 Da [M + H]⁺.

Synthesis of (S)-3-(3-allylphenyl)-2-((tert-butoxycarbonyl) amino) propanoic acid (11)

A solution of methyl (S)-3-(3-allylphenyl)-2-((tert-butoxycarbonyl) amino) propanoate (10) (1.1 g, 3.34 mmol) in methanol (24 mL) and water (6 mL) was added to LiOH.H₂O (0.29 g, 6.95 mmol). The reaction mixture was stirred for 2 h at 25 °C. The reaction mixture pH was adjusted to 4–5 with saturated aqueous citric acid solution and concentrated minimum solution. Resulted crude was re-dissolved in CH₂Cl₂ and washed with water and brine solution. Resulted solution was evaporated to dryness which gave (S)-3-(3-allyl phenyl)-2-((tert-butoxy carbonyl) amino) propanoic acid (11) as a white solid. Yield: 841 mg, 80%. ¹H NMR (300 MHz, DMSO-d₆, 298 K): δ 7.20 (t, J = 12.4 Hz, 1H), 7.13–6.83 (m, 4H), 6.04–5.86 (m, 1H), 5.15–4.99 (m, 2H), 4.01 (dd, J = 16.9, 6.0 Hz, 1H), 3.38–3.27 (m, 2H), 3.01 (dd, J = 18.1, 8.8 Hz, 1H), 2.87–2.69 (m, 1H), 1.32 (s, 9H). ESI-MS (m/z): 306 Da [M + H]⁺. Crude was sufficient pure to perform further transformation. We have repeated the reaction in larger gram scale reactions.

Synthesis of methyl (S)-2-((S)-3-(3-allylphenyl)-2-((tert-butoxycarbonyl) amino) propanamido)-3-((S)-2-oxopyrrolidin-3-yl) propanoate (12)

To a solution of 11 (1.43 g, 4.68 mmol) in DMF (30 mL) was added Et₃N (3.83 mL, 27.5 mmol) and HATU (3.5 g, 9.2 mmol). The mixture was stirred at 25 °C for 5 min. Then the (S)-methyl 2-amino-3-((S)-2-oxopyrrolidin-3-yl) propanoate (1.56 g, 7.0 mmol) was added to the above mixture. Progress of the reaction was monitored based on TLC and LCMS. After completion of reaction, the mixture was added to ice-cold water and extracted in CH₂Cl₂ (3 × 50 mL). Resulted organic layer was washed with brine solution and dried over Na₂SO₄. Resulted organic was evaporated to dryness and purified on Flash Chromatography on C18 column (25–45% of H₂O/ACN). Product 12 was isolated with yield: 1.61 g, 74%. ¹H NMR (700 MHz, CD₃CN, 298 K): δ 7.91–7.70 (m, 1H), 7.25–7.17 (m, 1H), 7.12–6.81 (m, 3H), 6.15–6.05 (m, 1H), 6.0–5.93 (m, 1H), 5.50–5.38 (m, 1H), 5.12–5.00 (m, 2H), 4.41–4.07 (m, 3H), 3.68 (sx2, 3H), 3.36 (s, 2H), 3.25–3.13 (m, 2H), 3.07–2.95 (m, 1H), 2.88–2.53 (m, 2H), 2.36–2.16 (m, 2H), 2.09–1.98 (m, 1H), 1.87–1.65 (m, 3H), 1.35 (sx2, 9H). ESI-MS (m/z): 474 Da [M + H]⁺.

Synthesis of tert-butyl ((S)-3-(3-allylphenyl)-1-(((S)-1-hydroxy-3-((S)-2-oxopyrrolidin-3-yl) propan-2-yl) amino)-1-oxopropan-2-yl) carbamate (13)

The compound 12 (1.28 g, 2.71 mmol) and NaBH₄ (0.844 g, 21.6 mmol) were dissolved in methanol (50 mL) and the reaction mixture was stirred at 20 °C for 7 h. The mixture was treated with 10% ammonium chloride solution and on concentration to give as residue, then extracted with DCM (100 mL × 3). The organic phase was dried with Na₂SO₄ and concentrated. The residue was purified on a silica gel column to give 13. Yield: 1.2 g, 100%, which was sufficient pure to perform further transformation. ESI-MS (m/z): 446 Da [M + H]⁺.

Synthesis of tert-butyl ((S)-3-(3-allylphenyl)-1-oxo-1-(((S)-1-oxo-3-((S)-2-oxopyrrolidin-3-yl) propan-2-yl) amino) propan-2-yl)carbamate (14)

To a solution of 13 (1 g, 2.25 mmol) in DCM (25 mL) was added DMP (1.23 g, 2.9 mmol). The reaction mixture was stirred for 3 h at 25 °C. The reaction mixture was concentrated to give a residue. It was purified on a silica gel column to give 14. Yield: 633 mg, 65%. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 9.30 (s, 1H), 8.48–8.28 (m, 1H), 8.09–7.84 (m, 1H), 7.68–7.38 (m, 2H), 7.20–6.83 (m, 3H), 6.06–5.88 (m, 1H), 5.68 (s, 1H), 5.14–5.0 (m, 1H), 4.75–4.48 (m, 1H), 4.24–4.05 (m, 1H), 3.77–3.61 (m, 0.5H), 3.33–3.25 (m, 1H), 2.86–2.65 (m, 1H), 2.29 – 1.76 (m, 3H), 1.64–1.45 (m, 1H), 1.42–1.18 (m, 10H). ESI-MS (m/z): 444 Da [M + H]⁺.

(6S,9S)-6-(3-allylbenzyl)-2,2-dimethyl-4,7,11-trioxo-9-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-oxa-5,8,12-triazaheptadec-16-en-10-yl acetate (15a)

To a solution of 14 (200 mg, 0.451 mmol) in DCM (20 mL) was added AcOH (52 μL, 0.916 mmol). Then the mixture was stirred at 25 °C for 10 min. Then 5-isocyanopent-1-ene (88 mg, 0.902 mmol) was added. The reaction was stirred for 12 h. The mixture was purified on a silica gel column to give the acetoxy amide 15a as mixture of diastereomers (148 mg, 55%) as a white solid.

Dia-1: ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.35–7.81 (m, 2H), 7.79–7.51 (m, 1H), 7.28–6.80 (m, 5H), 6.49–6.15 (m, 1H), 6.03–5.70 (m, 2H), 5.11–4.85 (m, 4H), 4.70 (dx2, J = 6.6 Hz, 1H), 4.20–4.04 (m, 1H), 3.98–3.64 (m, 1H), 3.22–2.98 (m, 4H), 2.91–2.69 (m, 1H), 2.28–1.72 (m, 8H), 1.59–1.39 (m, 4H), 1.41–1.13 (m, 9H), ESI-MS (m/z): 599 Da [M + H]⁺.

Dia-2: ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.14–7.81 (m, 2H), 7.63–7.49 (m, 9.2 Hz, 1H), 7.25–6.86 (m, 5H), 6.51–6.22 (m, 1H), 5.95 (ddt, J = 16.8, 10.1, 6.7 Hz, 1H), 5.86–5.73 (m, 1H), 5.14–4.80 (m, 4H), 4.37–4.25 (m, 1H), 4.21–4.04 (m, 1H), 3.25–2.97 (m, 4H), 2.94–2.58 (m, 2H), 2.40–1.79 (m, 8H), 1.66–1.39 (m, 3H), 1.37–0.97 (m, 10H). ESI-MS (m/z): 599 Da [M + H]⁺.

(6S,9S)-6-(3-allylbenzyl)-2,2-dimethyl-4,7,11-trioxo-9-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-oxa-5,8,12-triazaoctadec-17-en-10-yl acetate (15b)

Following the procedure described for the preparation of 15a with a slight modification of conducting the reaction with acetic acid (81 mg, 1.22 mmol) and 7-isocyanohept-1-ene (156 mg, 1.22 mmol) at rt afforded the acetoxy amide 15b (290 mg, 57%) as a white solid.¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.06–7.74 (m, 2H), 7.57 (t, J = 32.7 Hz, 1H), 7.23–6.83 (m, 6H), 5.99–5.89 (m, 1H), 5.86–5.68 (m, 1H), 5.12–4.76 (m, 4H), 4.36–4.17 (m, 2H), 3.20–2.96 (m, 5H), 2.93–2.55 (m, 3H), 2.28–2.05 (m, 5H), 1.98–1.80 (m, 5H), 1.68–1.52 (m, 2H), 1.47–1.14 (m, 15H). ESI-MS (m/z): 627 Da [M + H]⁺.

(3S,6S,E/Z)-3-((tert-butoxycarbonyl)amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl)methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphan-13-en-7-yl acetate (16b)

Hoveyda–Grubbs IInd generation catalyst (20.3 mg, 0.0246 mmol) was added to a deoxygenated and magnetically stirred solution of 15a (123 mg, 0.205 mmol) in DCE (100 mL) maintained under a nitrogen atmosphere. The resulting solution was heated at reflux for 18 h, then cooled and concentrated under reduced pressure. The residue so-obtained was subjected to flash chromatography to give the acetoxy amides 16b, 16a (71 mg, 12 mg, in total 71%) as mixture of diastereomers.¹H NMR (700 MHz, DMSO-d₆, 298 K): Major Dia-1: δ 7.96–7.79 (m, 1H), 7.59–7.32 (m, 2H), 7.17 (h, J = 7.5 Hz, 1H), 7.11–6.95 (m, 4H), 6.93–6.85 (m, 2H), 5.54–5.41 (m, 1H), 4.81 (d, J = 9.9 Hz, 1H), 4.11–4.06 (m, 1H), 3.91 (s, 1H), 3.31–3.18 (m, 2H), 3.17–2.94 (m, 4H), 2.91–2.74 (m, 1H), 2.56–2.51 (m, 3H), 2.19–1.91 (m, 7H), 1.72–1.45 (m, 3H), 1.42–1.32 (s, 9H), 1.30–1.20 (m, 2H), ESI-MS (m/z): 571 Da [M + H]⁺.

Major Dia-2: δ 7.66–7.51 (m, 3H), 7.16–6.97 (m, 4H), 6.60 (d, J = 7.6 Hz, 1H), 5.57–5.42 (m, 2H), 4.91 (d, J = 2.5 Hz, 1H), 4.27–4.21 (m, 1H), 4.11 (dd, J = 13.3, 5.8 Hz, 1H), 3.31–3.23 (m, 2H), 3.21–3.11 (m, 2H), 3.07–3.01 (m, 1H), 2.94–2.89 (m, 2H), 2.78–2.69 (m, 1H), 2.56–2.51 (m, 3H), 2.29–2.11 (m, 3H), 2.11–1.90 (m, 2H), 1.69–1.51 (m, 2H), 1.45–1.32 (m, 10H), 1.32–1.18 (m, 2H). ESI-MS (m/z): 571 Da [M + H]⁺.

(3S,6S,E/Z)-3-((tert-butoxycarbonyl)amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl)methyl)-5,9-diaza-1(1,3)-benzenacyclotetradecaphan-12-en-7-yl acetate (16a)

¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 7.97–7.78 (m, 1H), 7.57–7.41 (m, 1H), 7.20–6.91 (m, 4H), 6.83–6.68 (m, 1H), 5.70–5.60 (m, 1H), 5.45–5.35 (m, 1H), 4.50–4.43 (m, 1H), 4.34–4.25 (m, 1H), 3.94–3.85 (m, 1H), 3.28–2.94 (m, 5H), 2.94–2.64 (m, 2H), 2.34–2.09 (m, 5H), 1.66–1.05 (m, 11H). ESI-MS (m/z): 557 Da [M + H]⁺.

(3S,6S,E)-3-((tert-butoxycarbonyl)amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl)methyl)-5,9-diaza-1(1,3)-benzenacycloheptadecaphan-15-en-7-yl acetate (16d)

Following the procedure described for the preparation of 16b with a slight modification of conducting the reaction with Hoveyda–Grubbs IInd generation catalyst (16.1 mg, 0.01956 mmol) at 80 °C afforded the acetoxy amide 16d, 16c (35 mg, 7 mg, in total 43%) as a white solid.¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 7.86–7.75 (m, 1H), 7.63–7.52 (m, 1H), 7.19–7.08 (m, 2H), 7.06–6.93 (m, 2H), 5.63–5.57 (m, 1H), 5.51–5.38 (m, 1H), 4.86–4.66 (m, 1H), 4.18–4.11 (m, 1H), 3.94 (t, J = 15.8 Hz, 1H), 3.38–3.27 (m, 2H), 3.18–2.95 (m, 2H), 2.90–2.74 (m, 2H), 2.66–2.54 (m, 1H), 2.29–2.01 (m, 5H), 1.87–1.80 (m, 5H), 1.62–1.16 (m, 12H). ESI-MS (m/z): 599 Da [M + H]⁺.

(3S,6S,E)-3-((tert-butoxycarbonyl) amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl)methyl)-5,9-diaza-1(1,3)-benzenacyclohexadecaphan-14-en-7-yl acetate (16c)

¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 7.95 (s, 1H), 7.70–7.52 (m, 2H), 7.26–6.90 (m, 3H), 6.49 (s, 1H), 6.36 (s, 1H), 6.29–6.12 (m, 1H), 4.70 (s, 1H), 4.38–4.04 (m, 2H), 3.2 (d, J = 10 Hz, 2H), 3.18–2.95 (m, 2H), 2.83–2.62 (m, 2H), 2.38–1.79 (m, 7H), 1.60–1.12 (m, 11H), 0.95–0.78 (m, 2H). ESI-MS (m/z): 585 Da [M + H]⁺.

(3S,6S)-3-((tert-butoxycarbonyl) amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclotetradecaphane-7-yl acetate (17a)

To a mixture of 16a (50 mg, 0.089 mmol) in MeOH (5 mL) was added Pd/C (5 mg, 10% on carbon). The mixture was stirred under a hydrogen atmosphere at room temperature until the starting material was consumed completely. The mixture was filtered through highflow and the filtrate was concentrated to give the crude product 17a (30 mg, 60%) as a white solid. ¹H NMR (600 MHz, DMSO-d₆, 298 K): δ 8.10–7.92 (m, 1H), 7.63–7.47 (m, 2H), 7.42–7.27 (m, 2H), 7.20–6.85 (m, 4H), 6.75–6.51 (m, 2H), 4.79–4.61 (m, 2H), 4.23–4.10 (m, 2H), 3.99–3.87 (m, 2H), 3.15–3.29 (m, 2H), 3.20–2.87 (m, 3H), 2.79–2.51 (m, 3H), 2.26–1.69 (m, 5H), 1.68–1.51 (m, 3H), 1.46–1.05 (m, 10H). ESI-MS (m/z): 559 Da [M + H]⁺.

(3S,6S)-3-((tert-butoxycarbonyl) amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-7-yl acetate (17b)

Following the procedure described for the preparation of 17a with a slight modification of conducting the reaction with Pd/C (12 mg, 10% on carbon) in CH₃OH at rt afforded the acetoxy amide 17b (82 mg, 82%) as a white solid.¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 7.87 (s, 1H), 7.56 (s, 1H), 7.30–7.18 (m, 2H), 7.16–7.07 (m, 3H), 7.03–6.93 (m, 2H), 4.80 (t, J = 16.5 Hz, 1H), 4.15–4.02 (m, 2H), 3.29 (s, 2H), 3.19–2.97 (m, 4H), 2.84 (d, J = 19.0 Hz, 2H), 2.55–2.51 (m, 2H), 2.19–1.87 (m, 5H), 1.74–1.45 (m, 4H), 1.45–0.99 (m, 12H). ESI-MS (m/z): 573 Da [M + H]⁺.

(3S,6S)-3-((tert-butoxycarbonyl) amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclohexadecaphane-7-yl acetate (17c)

Following the procedure described for the preparation of 17a, with a slight modification in which the reaction was conducted with Pd/C (6 mg, 10% on carbon) in CH₃OH at room temperature, acetoxy amide 17c (12 mg, 85%) was obtained as a white solid. The product was used for further reactions without purification. ESI-MS (m/z): 587 [M + H]⁺.

(3S,6S)-3-((tert-butoxycarbonyl) amino)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacycloheptadecaphane-7-yl acetate (17d)

Following the procedure described for the preparation of 17a with a slight modification of conducting the reaction with Pd/C (8 mg, 10% on carbon) in CH₃OH at rt, afforded the acetoxy amide 17d (23 mg, 87%) as a white solid. ¹H NMR (600 MHz, DMSO-d₆, 298 K): δ 7.99–7.75 (m, 2H), 7.73–7.48 (m, 2H), 7.17–6.97 (m, 4H), 6.85 (t, J = 13.0 Hz, 1H), 5.00–4.87 (m, 1H), 4.78 (d, J = 5.4 Hz, 1H), 4.19–4.27 (m, 1H), 4.17–3.86 (m, 2H), 3.29–3.15 (m, 1H), 3.16–2.69 (m, 5H), 2.62–2.55 (m, 1H), 2.29–2.10 (m, 2H), 2.05 (d, J = 8.9 Hz, 4H), 1.97–1.67 (m, 2H), 1.66–1.37 (m, 6H), 1.37 (sx2, m, 9H), 1.27–1.06 (m, 6H). ESI-MS (m/z): 601 Da [M + H]⁺.

(3S,6S)-3-amino-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclotetradecaphane-7-yl acetate hydrochloride (18a)

To a solution of the compound 17a (30 mg, 0.053 mmol) in CH₂Cl₂ (3 mL), cooled to 0 °C was added TFA (0.2 ml, excess). After the consumption of the starting material based on LCMS, the solvent was removed under vacuum, to afforded the acetoxy amine salt 18a along with it’s De-acylated amine salt (18a-OH) (16 mg, 65.5% in total) as a white solid. ¹H NMR (600 MHz, DMSO-d₆, 298 K): δ 8.41–8.25 (m, 2H), 8.01–7.54 (m, 2H), 7.35–7.28 (m, 1H), 7.20–7.00 (m, 3H), 6.99–6.87 (m, 1H), 4.83 (t, J = 15.1 Hz, 1H), 4.37–4.29 (m, 1H), 4.19–4.01 (m, 1H), 3.67–3.42 (m, 1H), 3.19–2.90 (m, 4H), 2.90–2.68 (m, 2H), 2.64–2.51 (m, 2H), 2.38–2.05 (m, 4H), 1.87–1.44 (m, 6H), 1.43–1.02 (m, 5H). ESI-MS (m/z): 417 Da [M + H]⁺.

(3S,6S)-3-amino-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-7-yl acetate hydrochloride (18b)

Following the procedure described for the preparation of compound 17a with a slight modification of conducting the reaction with 4 M HCl in dioxane (0.354 ml, 10 eq.) in DCM at rt, afforded the acetoxy amine 18b and hydroxy amine 18b-OH (43 mg, 20 mg, in total 93%) after purification Prep-HPLC as a white solid. 18b-OH- ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.33 (s, 1H), 7.60–7.32 (m, 2H), 7.16–6.86 (m, 3H), 6.54 (s, 1H), 5.81 (d, J = 5.7 Hz, 1H), 4.06–3.92 (m, 1H), 3.70 (dd, J = 5.7, 2.3 Hz, 1H), 3.47–3.36 (m, 1H), 3.20–2.79 (m, 4H), 2.65–2.52 (m, 1H), 2.39–1.83 (m, 5H), 1.73–1.00 (m, 8H). ESI-MS (m/z): 431 Da [M + H]⁺.

(3S,6S)-3-amino-7-hydroxy-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclohexadecaphane-4,8-dione hydrochloride (18c)

Following the procedure described for the preparation of compound 18a with a slight modification of conducting the reaction with 4 M HCl in dioxane (0.1 ml, excess) in DCM at rt, afforded the hydroxy amine salt 18c-OH (8.1 mg, 75%) as a white solid. ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.22–7.90 (m, 2H), 7.70–7.51 (m, 2H), 7.23–6.90 (m, 3H), 6.12 (d, J = 5.1 Hz, 1H), 4.08–3.97 (m, 1H), 3.71 (s, 1H), 3.21–2.77 (m, 6H), 2.60–2.54 (m, 2H), 2.35–1.79 (m, 7H), 1.71–1.12 (m, 8H). ESI-MS (m/z): 445 Da [M + H]⁺.

(3S,6S)-3-amino-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacycloheptadecaphane-7-yl acetate hydrochloride (18 d)

Following the procedure described for the preparation of compound 18a with a slight modification of conducting the reaction with 4 M HCl in dioxane (0.2 ml, excess) in DCM (0.5 ml) at rt, afforded the 18d along with hydroxy amine salt 18d-OH (10 + 4 mg, 78%, in total) as a white solid. 18d-OH_Dia-1: ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.23 (s, 1H), 7.69 (t, J = 12.5 Hz, 1H), 7.60–7.47 (m, 1H), 7.29–6.93 (m, 3H), 6.10 (broad s, 1H), 5.52–5.28 (m, 1H), 4.10–4.01 (m, 1H), 3.95–3.48 (m, 5H), 3.25–3.00 (m, 4H), 2.94–2.52 (m, 4H), 2.39–2.25 (m, 1H), 2.19–1.85 (m, 2H), 1.74–1.04 (m, 10H). ESI-MS (m/z): 459 Da [M + H]⁺.

18d-OH_Dia-2: ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.23–7.65 (m, 2H), 7.51 (dt, J = 10.7, 6.8 Hz, 2H), 7.20–6.91 (m, 4H), 6.20–6.06 (m, 1H), 4.19–4.02 (m, 1H), 3.96–3.81 (m, 1H), 3.77–3.41 (m, 2H), 3.20–2.78 (m, 6H), 2.70–2.52 (m, 3H), 2.35–1.84 (m, 5H), 1.71–1.0 (m, 12H). ESI-MS (m/z): 459 Da [M + H]⁺.

Synthesis of (3S,6S,E/Z)-3-amino-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl)methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphan-13-en-7-yl acetate Hydrochloride (18e)

Following the procedure described for the preparation of compound 18a with a slight modification of conducting the reaction with 4 M HCl in dioxane (0.2 ml, excess) in DCM (0.5 ml) at rt, afforded the acetoxy amine salt 18e (7.1 mg, 79.5%) as a white solid. ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.57–7.97 (m, 2H), 7.95–7.51 (m, 2H), 7.45–6.97 (m, 3H), 6.48–6.17 (m, 1H), 6.09–5.29 (m, 2H), 5.13–4.81 (m, 1H), 4.41–3.94 (m, 2H), 3.42–2.71 (m, 6H), 2.32–1.76 (m, 4H), 1.76–1.36 (m, 3H), 1.36–0.71 (m, 5H). ESI-MS (m/z): 471 Da [M + H]⁺.

(3S,6S)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-(2-phenylacetamido)-5,9-diaza-1(1,3)-benzenacyclotetradecaphane-7-yl acetate (19a)

To a solution of phenyl acetic acid (5.3 mg, 0.04 mmol) in DMF, HATU (15 mg, 0.0397 mmol) was added at 0 °C. The reaction mixture was stirred at 0 °C under a nitrogen atmosphere, and after few minutes, a pre-stirred solution of amine TFA salt 18a (13 mg, 0.0281 mmol) and triethylamine (17 mg, 0.168 mmol) in DMF at 0 °C was added. The reaction was continued for 5 h at 0 °C. Progress of the reaction was monitored based on TLC and LCMS. The reaction mixture was acidified with 10%TFA, filtered and crude was further purified on reverse-phase preparative HPLC which afforded acetoxy amide 19a (10.2 mg, 63.4%) as a white solid. ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.60–8.50 (m, 0.5H), 8.41–8.28 (m, 0.5H), 8.09–8.01 (m, 1H), 7.62–7.45 (m, 1H), 7.38–6.84 (m, 9H), 6.86–6.32 (t, J = 9.5 Hz, 1H), 4.74–4.59 (m, 1H), 4.58–4.46 (m, 1H), 4.29–4.03 (m, 1H), 3.71–3.41 (m, 3H), 3.37–3.18 (m, 2H), 3.15–2.87 (m, 3H), 2.83–2.55 (m, 2H), 2.22–1.93 (m, 4H), 1.86–1.49 (m, 3H), 1.47–0.84 (m, 8H). ESI-MS (m/z): 577 Da [M + H]⁺.

(3S,6S)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-(2-phenylacetamido)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-7-yl acetate (19b)

Following the procedure described for the preparation of compound 19a with a slight modification of conducting the reaction with compound 18b (50 mg, 0.1207 mmol) and 2-phenylacetic acid (20 mg, 0.147 mmol) afforded compound 19b (21 mg, 35%) as a white solid. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.30 (dx2, J = 7.6 Hz, 1H), 7.98–7.87 (m, 1H), 7.75–7.50 (m, 3H), 7.35–6.85 (m, 8H), 4.91–4.77 (m, 1H), 4.53–4.35 (m, 1H), 4.24–4.02 (m, 1H), 3.58–3.44 (m, 2H), 3.18–2.66 (m, 6H), 2.22–1.96 (m, 4H), 1.95–1.46 (m, 5H), 1.41–0.80 (m, 9H). ESI-MS (m/z): 591 Da [M + H]⁺.

N-((3S,6S)-7-hydroxy-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclohexadecaphane-3-yl)-2-phenylacetamide (19c)

Following the procedure described for the preparation of compound 19a with a slight modification of conducting the reaction with compound 18c-OH (6 mg, 0.0107 mmol) and 2-phenylacetic acid (2.2 mg, 0.0161 mmol) afforded hydroxy compound 19c (3.5 mg, 58%) as a white solid. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.77 (d, J = 4.2 Hz, 1H), 8.57–8.38 (m, 1H), 8.13 (s, 1H), 7.81–7.41 (m, 2H), 7.39–7.14 (m, 3H), 7.13–6.84 (m, 3H), 6.02–5.71 (m, 2H), 4.44–4.35 (m, 1H), 4.08–4.03 (m, 1H), 3.84–3.74 (m, 1H), 3.61–3.43 (m, 2H), 3.18–2.70 (m, 5H), 2.31–1.86 (m, 5H), 1.73–1.33 (m, 5H), 1.33–0.80 (m, 8H). ESI-MS (m/z): 563 Da [M + H]⁺.

(3S,6S)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-(2-phenylacetamido)-5,9-diaza-1(1,3)-benzenacycloheptadecaphane-7-yl acetate (19d)

Following the procedure described for the preparation of compound 19a with a slight modification of conducting the reaction with compound 18 d (10 mg, 0.0186 mmol) and 2-phenylacetic acid (3.8 mg, 0.02830 mmol) afforded compound 19 d (7.1 mg, 61%) as a white solid. ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.46–8.40 (m, 1H), 8.02–7.92 (m, 2H), 7.81–7.53 (m, 3H), 7.43–6.94 (m, 7H), 4.87–4.70 (m, 2H), 4.52–4.25 (m, 1H), 4.24–3.91 (m, 2H), 3.61–3.48 (m, 2H), 3.25–2.91 (m, 4H), 2.99–2.70 (m, 2H), 2.64–2.57 (m, 1H), 2.17–1.99 (m, 3H), 1.81–0.80 (m, 10H). ESI-MS (m/z): 619 Da [M + H]⁺.

(3S,6S)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-(3-phenylpropanamido)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-7-yl acetate (19e)

Following the procedure described for the preparation of compound 19a with a slight modification of conducting the reaction with compound 18b (52 mg, 0.102 mmol) and 3-phenylpropanoic acid (23 mg, 0.153 mmol) afforded compound 19e (23 mg, 40%) as a white solid. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.30–8.11 (m, 1H), 7.76–7.52 (m, 2H), 7.31–6.83 (m, 10H), 4.84 (dx2, J = 5.1 Hz, 1H), 4.56–4.33 (m, 1H), 4.22–3.99 (m, 1H), 3.25–2.92 (m, 4H), 2.86–2.57 (m, 5H), 2.49–2.27 (m, 4H), 2.26–1.80 (m, 6H), 1.71–0.96 (m, 9H), ESI-MS (m/z): 605 Da [M + H]⁺.

(3S,6S)-3-((S)-3,3-dimethyl-2-(2,2,2-trifluoroacetamido) butanamido)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-7-yl acetate (19 f)

Following the procedure described for the preparation of compound 19a with a slight modification of conducting the reaction with compound 18b-OH (30 mg, 0.0696 mmol) and (S)-3,3-dimethyl-2-(2,2,2-trifluoroacetamido) butanoic acid (23.7 mg, 0.1044 mmol) afforded compound 19f (12 mg, 27%) as a white solid. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 9.27–9.19 (m, 1H), 8.62–8.37 (m, 1H), 7.84–7.39 (m, 2H), 7.23–6.91 (m, 4H), 6.19–5.73 (broad s, 1H), 4.53–4.29 (m, 2H), 4.02–3.53 (m, 4H), 3.17–2.95 (m, 3H), 2.95–2.70 (m, 3H), 2.64–2.54 (m, 2H), 2.46–2.28 (m, 2H), 2.21–2.07 (m, 1H), 2.01–1.78 (m, 2H), 1.73–1.45 (m, 3H), 1.44–0.77 (m, 12H). ESI-MS (m/z): 640 Da [M + H]⁺.

(3S,6S)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-3-(3-(2,2,2-trifluoroethyl) ureido)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-7-yl acetate (19g)

To a solution of amine salt 18b (35 mg, 0.0597 mmol) in dry CH₂Cl₂, N,N-diisopropylethyl amine (42 μL, 0.240 mmol) was added at 0 °C. After stirring for few minutes, 2-trifluoroethyl isocyanate (15 mg, 0.12 mmol) was added at the same temperature and stirred at rt under inert conditions. The reaction was continued for 24 h, and the progress of the reaction was monitored based on TLC and LCMS. After completion of the reaction, it was quenched with 0.1 mL of methanol and evaporated to dryness. The resulting crude was purified based on a flash column over silica gel (5–10% MeOH/CH₂Cl₂), which gave the desired product (19 mg, 53.3%). ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 7.92–7.76 (m, 2H), 7.61–7.54 (m, 1H), 7.22–6.78 (m, 5H), 6.56–6.21 (m, 1H), 4.94–4.76 (m, 1H), 4.55–4.30 (m, 1H), 4.26–4.05 (m, 1H), 3.88–3.74 (m, 2H), 3.19–2.95 (m, 4H), 2.83–2.57 (m, 2H), 2.24–2.04 (m, 5H), 2.09–1.92 (m, 1H), 1.80–1.09 (m, 11H). ESI-MS (m/z): 598 Da [M + H]⁺.

(3S,6S,E/Z)-4,8-dioxo-6-(((S)-2-oxopyrrolidin-3-yl)methyl)-3-(3-phenylpropanamido)-5,9-diaza-1(1,3)-benzenacyclopentadecaphan-13-en-7-yl acetate (19h)

Following the procedure described for the preparation of compound 18a with a slight modification of conducting the reaction with compound 18e (7 mg, 0.0138 mmol) and 3-phenylpropanoic acid (4.3 mg, 0.0266 mmol) afforded compound 19 h (4.1 mg, 48%) as a white solid. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.51–8.06 (m, 1H), 8.02–7.84 (m, 1H), 7.73–7.48 (m, 2H), 7.38–6.80 (m, 8H), 6.43–5.97 (m, 1H), 5.69–5.19 (m, 1H), 5.02–4.56 (m, 1H), 4.49–4.19 (m, 1H), 4.04–3.78 (m, 1H), 3.25–2.97 (m, 6H), 2.94–2.66 (m, 3H), 2.37–1.78 (m, 6H), 1.72–1.12 (m, 5H), ESI-MS (m/z): 603 Da [M + H]⁺.

2-phenyl-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclotetradecaphane-3-yl) acetamide (20a)

To the stirred solution of compound 19a (10 mg, 0.017 mmol) in aq. CH₃OH, LiOH·H₂O (1.45 mg, 0.034 mmol) was added and stirred at rt for 1.5 h. The progress of the reaction was checked on LCMS. After completion of the reaction, it was further purified on reverse-phase preparative HPLC, which afforded of a hydroxyl product (4.5 mg, 48%) as a white solid. ESI-MS (m/z): 535 Da [M + H]⁺.

Dess–Martin periodinane (DMP) (5.39 mg, 0.0126 mmol) was added to a solution of the above hydroxy amide compound (4.5 mg, 0.0084 mmol) in DMF (2 mL). The mixture was stirred at rt for 2 h. After completion of the reaction, it was treated with few drops of a Na₂S₂O₃ solution and concentrated. The product was further purified on reverse-phase preparative HPLC, which afforded macrocyclic α-ketoamide 20a (3.7 mg, 58%) as a white solid. HPLC purity: 97.4%, ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.62–8.42 (m, 2H), 8.35–8.19 (m, 2H), 7.63–7.44 (sx2, 1H), 7.33–7.17 (m, 4H), 7.14–6.71 (m, 2H), 5.22–5.16 (m, 1H), 4.58–4.41 (m, 1H), 3.62–3.42 (m, 2H), 3.19–2.84 (m, 5H), 2.25–1.91 (m, 3H), 1.80–1.11 (m, 9H). ¹³C NMR (176 MHz, DEPT-135, DMSO-d₆) δ: 129.04, 128.89, 128.00, 127.83, 126.81, 126.32, 53.62, 49.68, 41.80, 39.87, 39.18, 39.12, 38.37, 37.71, 37.14, 34.60, 32.24, 29.98, 27.35, 26.83, 25.40, −0.04. HRMS: calcd. for C₃₀H₃₇N₄O₅ [M + H]⁺, m/z = 533.2763 Da; found, m/z = 533.2768 Da.

2-phenyl-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-3-yl) acetamide (20b)

Compound 20b was prepared in two steps in a 41.3% yield by deprotection of acetoxy amide 19b (21 mg, 0.0359 mmol), followed by oxidation with DMP (21.66 mg, 0.051 mmol) utilizing the same procedure as described for compound 20b. HPLC purity: 99.1%, ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.89–8.69 (m, 1H), 8.65–8,52 (m, 1H), 8.26–7.91 (m, 1H), 7.72–7.59 (m, 1H), 7.36–7.17 (m, 4H), 7.17–6.86 (m, 3H), 6.76–6.63 (m, 2H), 5.44–5.25 (m, 1H), 4.90–4.53 (m, 1H), 4.30–4.22 (m, 1H), 3.58–3.37 (m, 3H), 3.16–3.0 (m, 2H), 2.97–2.73 (m, 3H), 2.46–2.02 (m, 4H), 2.01–1.78 (m, 1H), 1.72–1.21 (m, 9H). ¹³C NMR (176 MHz, DMSO-d₆, 298 K): δ 197.49, 178.29, 170.68, 170.23, 160.39, 158.62, 142.14, 136.80, 136.72, 129.97, 129.48, 128.60, 128.25, 127.45, 126.75, 53.82, 50.62, 42.43, 39.27, 38.04, 35.11, 32.60, 29.93, 27.81, 27.25, 25.45. HRMS: calcd. for C₃₁H₃₉N₄O₅ [M + H]⁺, m/z = 547.2920 Da; found, m/z = 547.2915 Da.

2-phenyl-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclohexadecaphane-3-yl) acetamide (20c)

Compound 20c was prepared in 63% yield from hydroxy amide 19c (3.5 mg, 0.0062 mmol) by oxidation with DMP (3.96 mg, 0.00931 mmol). The reaction mixture was stirred at room temperature for 2 h. After completion of the reaction, a few drops of Na₂S₂O₃ solution were added, and the mixture was concentrated. The crude product was purified by reverse-phase preparative HPLC to afford macrocyclic α-ketoamide 20c. HPLC purity: 95%. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.80–8.61 (m, 2H), 7.84 (d, J = 7.6 Hz, 1H), 7.65–7.47 (m, 2H), 7.37–6.94 (m, 6H), 6.75–6.64 (m, 1H), 5.37–5.29 (m, 1H), 4.61 (dd, J = 12.0, 6.1 Hz, 1H), 3.61–3.41 (m, 3H), 3.20–3.04 (m, 2H), 3.01–2.56 (m, 4H), 2.49–2.32 (m, 2H), 2.31–2.14 (m, 2H), 1.95–1.45 (m, 7H), 1.40–1.00 (m, 7H). ¹³C NMR (176 MHz, DMSO-d₆, DEPT-Q, 298 K): δ 196.97, 179.15, 177.82, 170.49, 169.81, 160.11, 141.84, 136.43, 136.36, 129.14, 128.82, 128.10, 128.07, 127.91, 127.68, 127.15, 126.59, 126.42, 126.31, 126.18, 54.81, 53.19, 53.00, 50.98, 41.91, 37.96, 37.66, 37.52, 37.43, 34.63, 31.98, 29.49, 27.24, 26.83, 26.58, 26.45, 26.23, 26.03, 23.67, 23.38. HRMS: calcd. for C₃₂H₄₁N₄O₅ [M + H]⁺, m/z = 561.3076 Da; found, m/z = 561.3074 Da.

2-phenyl-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacycloheptadecaphane-3-yl) acetamide (20 d)

Compound 20d was prepared in two steps in a 59% yield by de-protection of acetoxy amide 19d (7 mg, 0.0113 mmol), followed by oxidation with DMP (5.8 mg, 0.0138 mmol) utilizing the same procedure as described for compound 20a. HPLC purity: 97.8%, ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.76–8.63 (m, 1H), 8.59–8.46 (m, 1H), 8.20–7.92 (m, 1H), 7.74–7.45 (m, 2H), 7.33–7.15 (m, 3H), 7.14–6.78 (m, 3H), 6.92–6.79 (m, 1H), 5.13–5.21 (m, 1H), 4.63–4.46 (m, 1H), 3.60–3.40 (m, 3H), 3.20–2.97 (m, 2H), 2.93–2.76 (m, 3H), 2.67–2.42 (m, 2H), 2.33–1.94 (m, 3H), 1.96–1.78 (m, 1H), 1.69–1.40 (m, 5H), 1.40–1.04 (m, 8H). ¹³C NMR (176 MHz, DMSO-d₆, 298 K): δ 197.47, 179.48, 178.64, 171.48, 170.28, 161.57, 141.94, 137.43, 137.05, 130.30, 129.55, 129.07, 128.43, 128.32, 127.56, 126.73, 54.45, 51.92, 42.60, 42.42, 38.29, 37.88, 34.85, 31.99, 30.33, 28.59, 27.59, 27.26, 25.80. HRMS: calcd. for C₃₃H₄₃N₄O₅ [M + H]⁺, m/z = 575.3233 Da; found, m/z = 575.3231 Da.

3-phenyl-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-3-yl) propanamide (20e)

Compound 20e was prepared in two steps in a 65% yield by deprotection of acetoxy amide 19e (23 mg, 0.04 mmol), followed by oxidation with DMP (21 mg, 0.049 mmol) utilizing the same procedure as described for compound 20a. HPLC purity: 97.2%, ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.79–8.70 (m, 1H), 8.57 (t, J = 9.9 Hz, 1H), 7.82–7.79 (m, 1H), 7.66–7.59 (m, 1H), 7.31–7.14 (m, 5H), 7.07 (t, J = 7.5 Hz, 1H), 6.95 (t, J = 10.8 Hz, 1H), 6.73 (d, J = 7.6 Hz, 1H), 6.68 (s, 1H), 5.29 (s, 1H), 4.61–4.58 (m, 1H), 3.40 (s, 1H), 3.17–3.07 (m, 2H), 2.90 (d, J = 13.5 Hz, 1H), 2.85–2.73 (m, 4H), 2.48–2.36 (m, 3H), 2.31–2.15 (m, 2H), 1.85 (d, J = 17.1 Hz, 1H), 1.68 (d, J = 12.3 Hz, 1H), 1.56–1.19 (m, 2H). ¹³C NMR (176 MHz, DMSO-d₆, 298 K): δ 197.94, 177.33, 171.51, 170.78, 160.36, 158.97, 141.98, 137.15, 129.91, 128.75, 127.75, 126.67, 126.29, 53.84, 50.72, 40.01, 39.99, 39.84, 39.31, 38.75, 38.64, 38.24, 37.98, 35.08, 32.67, 31.51, 29.90, 27.33, 27.29, 25.46. HRMS: calcd. for C₃₂H₄₁N₄O₅ [M + H]⁺, m/z = 561.3076 Da; found, m/z = 561.3071 Da.

(S)-3,3-dimethyl-2-(2,2,2-trifluoroacetamido)-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-3-yl) butanamide (20f)

Compound 20c was prepared in 67% yield from hydroxy amide 19f (15 mg, 0.022 mmol) by oxidation with DMP (9.9 mg, 0.0234 mmol). The reaction mixture was stirred at room temperature for 2 h. After completion of the reaction, a few drops of Na₂S₂O₃ solution were added, and the mixture was concentrated. The crude product was purified by reverse-phase preparative HPLC to afford macrocyclic α-ketoamide 20f. HPLC purity: 95.4%, ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 9.38–9.16 (m, 1H), 8.77–8.62 (m, 1H), 8.58–8.29 (m, 2H), 7.69–7.54 (m, 1H), 7.17–6.99 (m, 3H), 6.81–6.61 (m, 1H), 5.28 (d, J = 17.8 Hz, 1H), 4.89–4.52 (m, 1H), 4.54–4.31 (m, 2H), 3.18–2.74 (m, 5H), 2.48–2.08 (m, 4H), 2.033–1.73 (m, 2H), 1.70–0.77 (m, 15H). ¹³C NMR (126 MHz, DMSO-d₆, 298 K): δ 197.27, 178.53, 170.80, 168.71, 160.64, 141.95, 137.18, 129.80, 128.27, 127.10, 126.84, 61.45, 60.64, 55.31, 50.23, 40.55, 40.22, 40.13, 39.96, 39.79, 39.46, 37.61, 35.17, 35.04, 32.94, 29.98, 28.25, 27.91, 27.06, 26.89, 25.80. HRMS: calcd. for C₃₁H₄₃F₃N₅O₆ [M + H]⁺, m/z = 638.3165 Da; found, m/z = 638.3162 Da.

1-(2,2,2-trifluoroethyl)-3-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphane-3-yl) urea (20g)

Compound 20g was prepared in two steps in a 69% yield by deprotection of acetoxy amide 19g (18 mg, 0.030 mmol), followed by oxidation with DMP (16 mg, 0.0378 mmol) utilizing the same procedure as described for compound 20a. HPLC purity: 99.4%, ¹H NMR (500 MHz, DMSO-d₆, 298 K): δ 8.99–8.91 (m, 1H), 8.82–8.68 (m, 1H), 7.75–7.62 (sx2, 1H), 7.14 (t, J = 15.6 Hz, 1H), 7.02–6.92 (m, 2H), 6.91–6.78 (m, 1H), 6.58 (sx2, 1H), 6.19–6.13 (m, 1H), 5.48–5.34 (m, 1H), 4.82–4.35 (m, 1H), 3.97–3.78 (m, 2H), 3.17–3.10 (m, 2H), 2.99–2.74 (m, 3H), 2.49–2.26 (m, 3H), 2.24–2.10 (m, 1H), 2.02–1.82 (m, 2H), 1.74–1.63 (m, 1H), 1.61–1.18 (m, 8H). ¹³C NMR (126 MHz, DMSO-d₆, 298 K): δ 197.81, 178.15, 171.05, 159.85, 156.78, 142.12, 136.65, 136.63, 130.32, 128.09, 128.06, 127.69, 126.64, 53.5, 50.5, 39.96, 38.06, 35.35, 32.12, 30.05, 27.89, 27.71, 26.72, 25.24. HRMS: calcd. for C₂₆H₃₅F₃N₅O₅ [M + H]⁺, m/z = 554.2590 Da; found, m/z = 554.2585 Da.

3-phenyl-N-((3S,6S)-4,7,8-trioxo-6-(((S)-2-oxopyrrolidin-3-yl) methyl)-5,9-diaza-1(1,3)-benzenacyclopentadecaphan-13-en-3-yl) propanamide (20h)

Compound 20h was prepared in two steps in a 53% yield by deprotection of acetoxy amide 19h (4.1 mg, 0.0068 mmol), followed by oxidation with DMP (3.4 mg, 0.008 mmol) utilizing the same procedure as described for compound 20a. HPLC purity: 99.8%, E/Z ratio = 1:4. ¹H NMR (700 MHz, DMSO-d₆, 298 K): δ 8.66–8.45 (m, 2H), 8.26–8.11 (m, 1H), 8.11–7.99 (m, 1H), 7.96–7.84 (m, 1H), 7.70–7.59 (m, 1H), 7.31–7.06 (m, 5H), 7.06–6.90 (m, 2H), 6.90–6.64 (m, 2H), 5.56–5.33 (m, 2H), 5.02 (dd, J = 15.3 Hz, 1H), 4.68–4.14 (m, 2H), 3.33–3.18 (m, 3H), 3.17–2.95 (m, 3H), 2.91–2.69 (m, 4H), 2.49–2.36 (m, 2H), 2.29–1.83 (m, 6H), 1.69–1.44 (m, 4H). ¹³C NMR (176 MHz, DMSO-d₆, DEPT-135, 298 K): δ139.85, 132.47, 129.48, 128.13, 126.55, 125.68, 53.99, 50.86, 47.94, 39.87, 39.26, 39.14, 38.63, 37.76, 37.40, 36.37, 32.51, 31.79, 30.86, 29.43, 27.30, 26.59, 23.39. HRMS: calcd. for C₃₂H₃₉N₄O₅ [M + H]⁺, m/z = 559.2920 Da; found, m/z = 559.2917 Da.

Antiviral activity assays against SARS-CoV-2 in A549^ACE2+TMPRSS2 cells

The antiviral activity against SARS-CoV-2 Zagreb isolate (SARS-CoV2/ZG/297-20, University Hospital for Infectious Diseases, Zagreb, Croatia, GISAID database ID: EPI_ISL_451934)⁵³ on A549^ACE2+TMPRSS2 (obtained from the Goethe-University Frankfurt)⁴⁶ was assessed as described previously⁴⁰. In brief, compounds were added at a final volume of 100 µl in 10 different concentrations starting at 50 µM diluted by twofold. Each compound was tested in three to four technical replicates and at least two biological replicates. The cells were infected at BSL-3 level with SARS-CoV-2 wild-type at an MOI of 0.005. On day three post infection, cellular viability was analyzed by measuring the luminescence after addition of cell titer glo reagent (CellTiter-Glo® Substrate (lyophilized, #G755A, Promega) and CellTiter-Glo® Buffer (#G756A, Promega)). EC₅₀ values were calculated in GraphPad Prism using a nonlinear regression—exponential decay ([Inhibitor] vs. response with variable slope for 20b, 20e and 20f and normalized response for 20g and 20h).

Cytotoxicity measurements in A549^ACE2+TMPRSS2 cells

A549^ACE2+TMPRSS2 cells were seeded at a density of 2.0 × 10⁴ cells/well in incubation media (DMEM + 2% FBS) in a 96-well flat-bottom cell culture plate (Corning, REF 353075) and incubated at 37 °C and 5% CO₂. On the following day, the compounds were serially diluted in DMEM + 2% FBS and added at a final volume of 100 µl in 10 different concentrations starting at 50 µM diluted by twofold. Each compound was tested in three to four technical replicates. Additionally, one column per plate was treated with 0.5% DMSO in incubation media without any test compound used as negative control. After 71 h of incubation, 0.1% Triton reagent (×-100, Sigma-Aldrich, STBH3768) was added to three wells per plate as positive control. Cell viability was determined after 72 h of incubation by measuring the absorbance at 450 nm using the Infinite® 200 PRO Tecan Microplate Reader. For that, 10 µl of Cell Counting Kit-8 reagent (Dojindo Laboratories, CK04-11) was added to each well and incubated at 37 °C and 5% CO₂ for 2 h_. The measured sample absorbance (A_sample) was normalized as followed: ((A_sample – A_triton)/(A_untreated – A_triton)) × 100%, where A_triton is the average absorbance of all trition treated wells (positive control) and A_untreated is the average absorbance of all untreated wells per plate (negative control).

Cell culture

HeLa-R19 (ATCC, CCL-2) and HEK293T (ATCC, CRL-3216) were seeded and maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% (V/V) fetal bovine serum (FBS) and 100 μg/ml penicillin/streptomycin (P/S).

Production of recombinant luciferase reporter viruses

CVB3 (Nancy strain), EV-D68 (Fermon strain) and EV-A71 (BrCr strain) cDNA sequences were cloned in an expression vector under the transcriptional control of a T7 polymerase promoter. In the case of CVB3, the Renilla luciferase gene was inserted upstream of the VP4 gene and separated by a 3C/3D cleavage site. For EV-D68 and EV-A71, the Renilla luciferase gene was inserted between the VP1 and 2A gene, flanked by 2A cleavage sites. Viral RNA was transcribed in vitro from these plasmids (RiboMAX, Promega) and transfected into HEK293T cells with Lipofectamine2000 (Invitrogen). When total cytopathic effect (CPE) was reached (2–4 days), medium was harvested and subjected to three freeze-thaw cycles, followed by centrifugation at 4000 × g for 15 min to pellet cell debris. Supernatant was collected and stored as virus stocks at −80 °C. Viral titers (TCID₅₀) were determined on HeLa-R19 cells and calculated with the Spearman-Kärber method using the average of three independent titrations.

Infection and antiviral activity assays using reporter viruses

HeLa-R19 cells were seeded in flat-bottom 96-well plates (10⁴ cells per well) at least 16 h prior to infection experiments. Cells were infected with recombinant luciferase reporter viruses (MOI 0.1) for 30 min at 37 °C (CVB3 and EV-A71) or 33 °C (EV-D68). After 30 min of infection, the viral inoculum was aspirated and the cells were incubated with medium containing varying concentrations of 20f for 6.5 h at 37 °C (CVB3 and EV-A71) or 33 °C (EV-D68). All medium was removed 7 h post-infection and the cells were lysed with Renilla luciferase lysis buffer (Promega). Renilla luciferase activity was measured using the Renilla luciferase assay system (Promega) with a GloMax Explorer luminometer (Promega). Cell viability assays were performed parallel to infection experiments, but did not include virus infection. Cell viability was measured using the MTS assay with Aqueous One Solution (Promega) according to manufacturer’s protocol.

Antiviral activity assays measuring EV-D68-induced cytopathic effects

To test for antiviral activity against enterovirus D68, RD cells (ATCC, CCL-136) were seeded at a density of 2 × 10⁴ cells per well in DMEM supplemented with 10% FBS in a 96-well flat-bottom cell culture plate, and incubated at 37 °C in a 5% CO₂ atmosphere. The following day, the inhibitors were serially diluted in DMEM containing 10% FBS to a final volume of 100 µl. The RD cells were treated using ten different concentrations, starting at 100 µM and diluted by a factor of two. Each inhibitor was tested in three technical replicates. After 24 h of treatment, the compounds were removed, and the cells were infected with EV-D68 virus (ATCC, VR-1824) at a multiplicity of infection of 0.1 in DMEM containing 2% FBS and incubated at 33 °C in a 5% CO₂ atmosphere for one hour. One hour after infection, the virus was removed and the same concentrations of compounds were added to the cells, this time prepared in DMEM supplemented with 2% FBS. After 72 h, the incubation was stopped and cell viability was assessed using crystal violet.

To measure cell viability, the medium was removed and the cells were washed once with 100 µl of PBS. The cells were then fixed by adding 50 µl of methanol to each well for one minute. This was then replaced with 100 µl of a 0.1% crystal violet solution prepared in distilled water and the plate was shaken for 30 min at room temperature on a rocking platform, protected from light. The cells were then washed three times with 100 µl of PBS. After air-drying the plate for at least two hours, 100 µl of methanol was added to dissolve the crystal violet. The plate was then shaken for 30 min on a rocking platform, protected from light. Absorbance was measured at 570 nm using the Tecan Sunrise plate reader. The measured absorbance was normalized using the following formula: -(1-(A_infected – A_sample)/(A_infected – A_uninfected) × 100% and EC₅₀ values were calculated in GraphPad Prism 10 using nonlinear regression (dose-response–Asymmetric (five parameter), X is log(concentration)).

Cytotoxicity assessments in RD cells

RD cells (ATCC, CCL-136) were seeded at a density of 2 × 10⁴ cells per well in DMEM supplemented with 10% FBS in a 96-well flat-bottom cell culture plate, and incubated at 37 °C in a 5% CO₂ atmosphere. The following day, the inhibitors were serially diluted in DMEM containing 10% FBS to a final volume of 100 µl. Ten different concentrations were used, starting at 100 µM and diluted by a factor of two. Each inhibitor was tested in three technical replicates. Cell viability was measured after five days of incubation by determining the crystal violet absorbance at 570 nm using the Tecan Sunrise plate reader. Cell viability was calculated as followed: (A_{compound treated cells}/A_{untreated cells}) × 100%.

Cloning and recombinant protein production

A gene encoding the SARS-CoV-2 main protease (M^pro wild type, GenBank reference: MN908947.3 (ORF1ab polyprotein residues 3264–3569)) with E. coli codon usage was produced by MWG Eurofins. The amplified DNA fragment was digested with the restriction enzymes BamHI and XhoI and then ligated into the plasmid pGEX-6P-1. The WT M^pro coding gene with BamHI and XhoI sites was amplified from the M^pro construct described previously⁵⁴ and was used as a template. The gene sequence of the M^pro was verified by sequencing (MWG Eurofins, Ebersberg, Germany). For recombinant protein production, the SARS-CoV-2 M^pro PGEx6p-1 plasmid was transformed into BL21 (DE3) competent cells and the protein was produced as described before³.

Determination of inhibition concentrations (IC₅₀’s)

A fluorescent substrate considering the cleavage site of SARS-CoV2 M^pro (Dabcyl-KTSAVLQ ↓ SGFRKM-E (Edans)-NH₂), the cleavage site of Enterovirus D68 (Dabcyl-KTATVQ ↓ GPSLD-E (Edans)-NH₂ × TFA, and buffer composed of 20 mM HEPES, 120 mM NaCl, 0.4 mM EDTA, 4 mM DTT, 20% Glycerol, pH 7.0 was used for the inhibition assay. In the fluorescence resonance energy transfer (FRET)-based cleavage assay⁵⁵, the fluorescence signal of the Edans generated due to the cleavage of the substrate by the M^pro was observed at an emission wavelength of 460 nm with excitation at 360 nm fluorescence spectrophotometer (BioTek). Tecan—The Spark Control—Multimode Microplate reader was used for the inhibition test. The compound was diluted with 100% DMSO to prepare a stock solution. IC₅₀ was determined with the incubation of 50 nM of SARS-CoV2 M^pro, 1 µM of Enterovirus D68, and compound at various concentrations from 0 to 100 µM in reaction buffer at 37 °C for 10 min. In the end, the FRET substrate at a final concentration of 50, 10, and 10 µM was added to each well for M^pro, EV-D68 3C^pro, and EV-71 3C^pro, respectively, at a final total volume of 100 µl to initiate the reaction. The IC₅₀ value was calculated with the help of GraphPad Prism software. Inhibitory activity of the compound was measured in triplicates.

Determination of inhibition rates against host cell proteases cathepsin B, K, L

Cathepsin B (CTSB) from human liver was provided from Enzo Life Sciences (BML-SE198) and diluted to a 500 nM stock solution in 50 mM NaOAc, pH 5.0, 1 mM EDTA and 2.5 mM DTT. The benzyloxycarbonyl-protected dipeptide Z-Phe-Arg-MCA (4-methylcoumaryl-7-amide) was used as a fluorogenic substrate for cathepsin B and L (Peptanova 3095-v).

Active human cathepsin K (CTSK) from Sigma Aldrich (SRP6561) was diluted to a 50 nM stock solution in 100 mM NaOAc, pH 5.5, 2.5 mM Na₂EDTA and 2.5 mM DTT. The fluorogenic substrate Z-Leu-Arg-MCA was also provided from Peptanova (3210-v) and is widely used as a substrate for cathepsin K.

All compounds were dissolved in 100% DMSO. To determine the inhibition rates at fixed concentrations of 1 and 10 µM, predilutions of 0.1 and 1 mM were made in 100% DMSO. Since 1 µl of compound is used in a total volume of 100 µl, the final DMSO concentration is 1%. 10 µl of stock solutions of either cathepsin B, K, or L (final concentrations: 50 nM CTSB, 5 nM CTSK, 25 nM CTSL) were preincubated with 1 or 10 µM compound in the same buffer condition recommended for each protease stock. After 10 min incubation at 37 °C, the proteolytic reaction was started by adding 30 µl of the respective substrate (final concentration 10 µM; exception 5 µM for CTSK reaction). The increase of fluorescence was detected using a Multimode Microplate Reader, BioTek Synergy H1 from Agilent (λ_ex = 360 nm; λ_em = 460 nm). All compounds were tested in triplicates. DMSO instead of compound served as a negative control and was set as the highest (100%) activity rate. Taking the initial slope, the activity rate was calculated related to the DMSO control. The inhibition rate (difference of activity rate to 100%) is shown in Table S8.

Crystallization and diffraction data collection of M^pro in complex with macrocyclic α-ketoamides

The purified SARS-CoV-2 M^pro and Enterovirus D68-3C^pro proteins were concentrated to 14 and 6 mg/mL, respectively. Proteins were then incubated overnight with a fivefold molar excess of inhibitor, and the mixture was incubated overnight at 4 °C. After being centrifuged at 13,000 rpm to remove precipitates. The supernatants of the SARS-CoV-2 M^pro compound mixture were then subjected to co-crystallization screening with commercially available kits (PACT premierTM HT-96, Morpheus and LFS Screen from Molecular Dimensions, Rotherham, UK), using an Art Robbins Instruments robot and the vapor diffusion sitting drop method at 20 °C. The Enterovirus D68-3C^pro were subjected to manual vapor diffusion sitting drop co-crystallization gradient screening of 0.1 M Tris-HCl pH 7.2–8.5, 0.2 M ammonium acetate, 15–35% w/v PEG 3350 at 4 °C. Diffraction data from the M^pro in complex with inhibitors were collected from crystals grown under the conditions of LSF No.-H5 (0.1 M SPG, pH 6.5, 25% w/v PEG 1500) for 9b-K, LSF No.-H7 (0.1 M Bis-Tris, pH 6.5, 25% w/v PEG 3350) for 20b, Morpheus No.-E5 ((0.1 M Sodium HEPES; MOPS (acid) pH 7.5 30% w/v (40% v/v PEG 500 MME; 20% w/v PEG 20000)), 0.1 M Ethylene glycol) for 20f and 20g, PACT No.-A4 (0.1 M SPG, pH 9.0, 25% w/v PEG 1500) for 20e, PACT No.-A3 (0.1 M SPG, pH 8.0, 25% w/v PEG 1500) for 20h, LSF No.-D10 (0.1 M PTCP, pH 8.0, 30% w/v PEG 1000) for 20a. Diffraction data from the Enterovirus D68-3C^pro in complex with inhibitors were collected from crystals grown under the conditions of 0.1 M Tris-HCl pH 7.6–8, 0.2 M Ammonium acetate, 20–25% w/v PEG 3350. Crystals were then cryo-protected using 20% glycerol. Fished crystals were flash-cooled using liquid nitrogen. Diffraction data were collected at 100 K at DESY PETRA III beamline P11 (Hamburg, Germany) using a Pilatus 6 M detector (Dectris) and with synchrotron radiation at a wavelength of 1.0332 Å.

Diffraction data processing, structure solution, and refinement

The programs XDSapp⁵⁶, Pointless^55,57,, and Scala⁵⁸ were used for dataset processing, with Molrep^57,58 to solve the phase problem via the molecular replacement method using the SARS-CoV-2 M^pro free enzyme (PDB code: 6Y2E) and the Enterovirus D68-3C^pro free enzyme (PDB code: 3ZV8), All compounds were built into Fo–Fc difference densities using the Coot software 0.9.8.91⁵⁹ and structures were refined using the Refmac5 program⁶⁰. All graphical illustrations and RMSD values were calculated using PyMOL 4.6⁶¹. Diffraction data and model refinement statistics are displayed in Tables S2 and S3.

ADME in vitro studies

The plasma stability assay, the metabolic stability assay as well as the plasma protein binding assay were conducted as described previously⁶².

HPLC-MS/MS analysis

Samples were analyzed using an Agilent 1290 Infinity II HPLC system coupled to an AB Sciex QTrap 6500plus mass spectrometer using the same LC conditions as described previously⁶³. Mass transitions for controls and compounds are depicted in Table S4.

Additionally, an Agilent 1290 Infinity II HPLC system coupled to an AB SCIEX Triple Quad^TM 7500 mass spectrometer was used with previously described LC conditions⁶³. Mass transitions for controls and compounds are depicted in Table S5.

Cellular accumulation assay on A549^ACE2+TMPRSS2 cells

Cellular accumulation was assessed as described previously⁶³. In brief, the compounds 20b and 20e–h were added at a final concentration of 25 µM to A549-AT cells in a 96-well format. After a 1 h incubation period at 37 °C and 5% CO₂, intracellular concentrations were received by HPLC-MS/MS analysis. Mass transitions can be found in Tables S4 and S5. A calibration curve ranging from 0.1 to 10000 nM as well as quality control samples (QCs; concentrations 1, 10, 100, and 1000 nM) were prepared for all compounds and measured simultaneously. Finally, K_p values were calculated after accounting for nonspecific binding and determining the cellular volume per well.

Author contributions

∇.R.K.A. and H.E.K. contributed equally to this work. R.K.A. Designed and synthesized the molecules and drafted the manuscript. H.E.K. contributed and analyzed data, administered the project, supervised biochemical data, conducted X-ray crystallographic studies, and wrote and edited the manuscript. A.M. developed bioanalytical MS/MS methods, conducted BSL-3 antiviral and cellular accumulation assays, analyzed the data and wrote and edited the manuscript. S.J. conducted biochemical assays. H.D. conducted EV-D68 antiviral assays, analyzed the data and wrote and edited the manuscript. R.V. conducted EV-D68, EV-A71 & CV antiviral assays, analyzed the data and wrote and edited the manuscript. J.R. conducted biochemical assays and analyzed the data. D.L.H. and F.V.K. acquired funding for and supervised the EV-D68, EV-A71 & CV antiviral assays, analyzed the data and wrote and edited the manuscript. K.R. analyzed data, administered the project, supervised antiviral, bioanalytical, ADME and cellular accumulation work, acquired funding and wrote and edited the manuscript. R.H. conceptualized the study, administered the project, analyzed the data, and acquired funding. M.B. conceptualized the study, analyzed the data, administered the project, acquired funding and wrote and edited the manuscript.

Peer review

Peer review information

Communications Chemistry thanks Felix Hausch, Florenci González and Haixia Su for their contribution to the peer review of this work. A peer review file is available.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its Supplementary Information files. Should any raw data files be needed in another format, they are available from the corresponding author upon reasonable request. Crystallographic coordinates and structure factors are available from the PDB under accession codes 9T5F (M^pro complex with 20f in space group C2), 9SSO (M^pro complex with 20g in space group C2), 9T72 (M^pro complex with 20h in space group C2), 9T5E (M^pro complex with 20a in space group C2), 9SSN (M^pro complex with 9b-K in space group C2), 9T52 (M^pro complex with 20b in space group C2), 9T55 (M^pro complex with 20e in space group C2), 29CT (3C^pro in complex with 20e) and 29DA (3C^pro in complex with 20g)

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42004-026-02151-y.