Novel PF74-like small molecules targeting the HIV-1 capsid protein: Balance of potency and metabolic stability

Abstract

Of all known small molecules targeting human immunodeficiency virus (HIV) capsid protein (CA), PF74 represents by far the best characterized chemotype, due to its ability to confer antiviral phenotypes in both early and late phases of viral replication. However, the prohibitively low metabolic stability renders PF74 a poor antiviral lead. We report herein our medicinal chemistry efforts toward identifying novel and metabolically stable small molecules targeting the PF74 binding site. Specifically, we replaced the inter-domain-interacting, electron-rich indole ring of PF74 with less electron-rich isosteres, including imidazolidine-2,4-dione, pyrimidine-2,4-dione, and benzamide, and identified four potent antiviral compounds (10, 19, 20 and 26) with markedly improved metabolic stability. Compared to PF74, analog 20 exhibited similar submicromolar potency, and much longer (51-fold) half-life in human liver microsomes (HLMs). Molecular docking corroborated that 20 binds to the PF74 binding site, and revealed distinct binding interactions conferred by the benzamide moiety. Collectively, our data support compound 20 as a promising antiviral lead.

Graphical abstract

HIV-1 capsid protein (CA)-targeting PF74 suffers from prohibitively low metabolic stability. Structural modifications have yet to produce analogs with balanced potency and metabolic stability. Replacing the inter-domain interacting indole ring with less electron-rich isosteres allowed the identification of compound 20 with drastically improved metabolic stability while largely maintaining antiviral potency.

1. Introduction

Human immunodeficiency virus type 1 (HIV-1) remains a global healthcare challenge despite the successful development of many antiviral drugs¹. Until an HIV-1 cure^2,3 is developed, current antiretroviral therapy (ART) is expected to be lifelong, and the virus will eventually select strains resistant to current drug classes. This necessitates the development of new drugs with novel molecular targets and distinct resistance profiles. Toward this end, HIV-1 drug discovery research targeting the multifunctional HIV-1 capsid protein (CA) has drawn increasing interest4, 5, 6. CA and its assembled product, the capsid core7, 8, 9, play critical roles in multiple steps within the HIV-1 replication cycle10, 11, 12 via CA‒CA interactions or CA‒host factor interactions. The capsid core provides a protected environment¹³ for viral reverse transcription and shields viral genome products from host nucleic acid sensing^14,15. CA‒CA interactions drive capsid assembly^7,9,16,17 and impact capsid disassembly and core stability¹⁸. Perturbation of core stability results in premature uncoating, impaired reverse transcription, and attenuated infection^19,20. Importantly, HIV-1 core stability and uncoating are also tightly regulated via the interactions between CA and various cellular factors to enable most post entry steps such as cytoplasmic trafficking, nuclear entry, integration, and host restriction^12,21. CA-interacting host factors include TRIM5α^22,23, cleavage and polyadenylation specific factor 6 (CPSF6)^24,25, nucleoporins 15326, 27, 28 and 358^29,30 (NUP153 and NUP358), MxB^31,32, and cyclophilin A33, 34, 35. Therefore, in addition to potentially inhibiting viral strains resistant to current drug classes, CA-targeting small molecules also have the advantage of conferring both early stage and late stage antiviral phenotypes by disrupting CA‒CA and CA‒host interactions.

HIV-1 CA consists of seven helices in the CA_NTD and four helices in the CA_CTD (Fig. 1A)36, 37, 38. A few small molecule-binding sites have been identified to accommodate distinct CA-targeting chemotypes⁴, of which the PF74 binding site³⁶ represents a particularly attractive drug target. Located at the CA dimer interface, this binding site is formed by many residues (lemon surface) within H3 and H4 of the CA_NTD (green), and a few residues (gray surface) in the H8 and H9 of the adjacent CA_CTD (blue, Fig. 1A). The detailed CA-binding mode of PF74 is shown in Fig. 1B³⁶. Key molecular interactions include: 1) three H-bonds by the phenylalanine core of PF74 (two with N57, and one with K70); 2) interactions of the aniline moiety with the N-methyl group N53, and the phenyl ring with A105, T107, and Y130; 3) hydrophobic interactions by the phenyl ring of the phenylalanine core with residues M66 and L69; and 4) inter-CA domain interactions by the indole moiety with the CA_NTD (via H-bond with Q63 and π−cation interaction with K70) and the adjacent CA_CTD (via Y169, R173, and K172)³⁶. Significantly, the PF74 site also binds host co-factors, such as nucleoporin Nup15326, 27, 28 and CPSF6^24,25,39,40. The binding features and the competition against host factors likely account for the unique concentration-dependent dual antiviral mechanisms of PF74: competing against host factors for capsid binding at low concentrations, and inducing premature uncoating at high concentrations¹⁹.

Figure 1.

Design of novel analogs targeting the PF74 binding site of HIV-1 CA. (A) Structure of HIV-1 CA dimer. The PF74 binding pocket is lined by numerous residues (lemon surface) of CA_NTD, as well as a few residues (gray surface) of the adjacent CA_CTD. Dimer structure and PF74 binding were reproduced in PyMOL based on PDB ID: 4XFZ³⁶. (B) Detailed binding interactions of PF74. The indole moiety of PF74 is involved in inter-CA domain interactions: with the CA_NTDvia H-bond to Q63 (dotted line) and π−cation interaction with K70 (arrow), and with the adjacent CA_CTD through interactions involving K182, Y169 and R173. (C) Novel analogs binding to the PF74 site were designed by replacing the indole ring with monocyclic rings with the aim to improve metabolic stability while retaining antiviral potency.

Despite all the favorable characteristics, including well-established binding mode, unique dual mechanism of action, good synthetic tractability and high amenability for optimization, PF74 as an antiviral lead is severely flawed due to its extremely poor metabolic stability41, 42, 43, 44. We have previously identified a compound showing much improved (44-fold) microsomal stability when compared to PF74⁴³. However, the compound features an altered backbone which conferred moderately reduced antiviral potency. Recently reported efforts also involved replacing the indole moiety with a 1,2,3-triazole ring^41,45 or benzenesulfonamide moiety⁴⁶, though these studies resulted in compounds with only marginally improved (<3-fold) microsomal stability. Therefore, the need to identify novel small molecules targeting the PF74 binding site with potent antiviral activity and improved metabolic stability remains. Toward this end, our current work is centered around replacing the indole ring with a few distinct 5- and 6-membered scaffolds (Fig. 1C). The rationale is that the susceptibility to oxidative metabolism could be effectively mitigated by replacing the electron-rich indole ring with an electron-deficient isostere. Such a scaffold-hopping strategy has been widely used to address metabolic issues associated with electron-rich aromatic rings⁴⁷. Our efforts led to the synthesis and testing of 26 analogs, of which 4 demonstrated potent anti-HIV-1 activity, significant effects on CA hexamer stability, and more importantly, substantially improved metabolic stability over PF74.

2. Results and discussion

2.1. Chemistry

Synthesis of analogs 1−15 and 17−26 is outlined in Scheme 1. Briefly, Boc-protected phenylalanine (28) was reacted with various amines under a well-established method using T₃P to afford 29. After removal of Boc protecting group using TFA, intermediate 30 was obtained, which was further reacted with commercially available acid derivative 31 to produce 1−15 and 18−26. Meanwhile, analog 17 was produced from the reaction of 30a and acid 32. The synthetic procedure of 16 is similar to the synthesis of 1−15 and 17−26 except amino acid 33 was used in the first step (Scheme 2).

Scheme 1.

Synthesis of analogs 1−15 and 17−26. Reagents and conditions: (i) amine, T₃P, DIPEA, DMF, rt, 12 h; (ii) TFA; DCM, rt, 4−6 h; (iii) HATU, DIPEA, DMF, rt, 12 h.

Scheme 2.

Synthesis of analog 16. Reagents and conditions: (i) N-methylaniline, T₃P, DIPEA, DMF, rt, 12 h; (i) TFA; DCM, rt, 6 h; (iii) HATU, DIPEA, DMF, rt, 12 h.

The preparation of analog 27 is described in Scheme 3. Boc-protected phenylalanine 28 was treated with l-proline methyl ester to afford intermediate 37 which was converted to carboxylic acid intermediate 38 in the presence of lithium hydroxide. Intermediate 38 was reacted with aniline to afford 39 which was converted to 40 via a TFA-mediated hydrolysis. Finally, 40 was treated with acid intermediate 41 to produce analog 27.

Scheme 3.

Synthesis of analog 27. Reagents and conditions: (i) l-proline methyl ester hydrochloride, T₃P, DIPEA, DMF, rt, 12 h; (ii) LiOH, MeOH/H₂O, rt, overnight; (iii) aniline, HATU, DIPEA, DMF, rt, 12 h; (iv) TFA; DCM, rt, 6 h; (v) HATU, DIPEA, DMF, rt, 12 h.

2.2. Biological assays and SAR

All final compounds were first tested in a thermal shift assay which measures the change of protein melting point (ΔT_m) upon compound binding. A positive ΔT_m value indicates a stabilizing effect and a negative value indicates a destabilizing effect on the protein. All analogs were also subjected to a cell-based antiviral screening at 20 μmol/L against HIV-1. Active compounds were further tested at 2 μmol/L, and those with significant inhibition were selecetd for dose response antiviral testing. In parallel, cytotoxicity was also measured for all final compounds. PF74 was used as a control in these assays and the values were consistent with previous reports^43,44.

The 26 analogs synthesized in this work belong to three distinct subtypes, each featuring a monocyclic ring (R₁) to engage in the inter-CA domain interactions in lieu of the indole ring. Within each subtype, we explored the effect of the aniline substituent (R₂), typically including 4-Me, 4-Cl, 3-Cl, 4-F, 3-F, 3-Br, and 3-CF₃ (Table 1) The first subtype contains the 5-imidazolidine-2,4-dione which has multiple H-bond donors (HBDs) and H-bond acceptors (HBAs). Compounds of this subtype (2−7) did not exhibit significant antiviral activity against HIV-1, with the exception of 4, which inhibited HIV-1 weakly at 2 μmol/L and almost completely at 20 μmol/L. The next series bears a 1-pyrimidine-2,4-dione ring as the indole replacement (8−17). Of these analogues, the ones closely mimicking PF74 (8−15) all strongly inhibited HIV-1 at 20 μmol/L (85%–100%). A few (8, 10, 11, 14 and 15) also exhibited inhibition at 2 μmol/L. The most potent analog of this series, compound 10, inhibited HIV-1 with an EC₅₀ of 1.7 μmol/L. When the benzyl of the phenylalanine core was changed to a phenyl ring with reversed stereochemistry, the resulting compound 16 did not inhibit HIV-1. The lack of antiviral activity was observed also with compound 17, which features the same pyrimidine-2,4-dione, albeit with a different linker and a different site of connection (C5 vs. N1). In the thermal shift assay, analogs of the pyrimidine-2,4-dione subtype did not significantly impact the stability of CA hexamer (ΔT_m = 0−1.7 °C).

Table 1.

Anti-HIV-1 activity, cytotoxicity, and CA hexamer stability profiles of all analogs.

Compd.	R1	R2	EC50a (μmol/L)	Inhibition at 2 μmol/L:20 μmol/L (%)	CC50b (μmol/L)	TSA ΔTmc (°C)
PF74d (1)		H	0.61±0.2	–	76±9	7.4
2		H	>20	–	>50	0
3		4-Me	–	2/62	>50	1.0
4		4-Cl	–	17/96	>50	‒1.2
5		3-Cl	>20	–	>50	0
6		4-F	>20	–	>50	0
7		3-F	>20	–	>50	0
8		H	–	35/95	>50	0.5
9		4-Me	–	7/85	>50	0.6
10		4-Cl	1.7±0.2	–	>50	1.7
11		3-Cl	–	37/95	>50	0.6
12		4-F	–	6/94	>50	0
13		3-F	–	0/74	>50	0.9
14		3-Br	–	22/100	>50	0.5
15		3-CF3	–	28/95	>50	0
16			>20	–	>50	1.3
17			>20	–	>50	0
18		H	–	15/96	>50	1.3
19		4-Me	1.6±0.02	–	>50	5.0
20		4-Cl	0.88±0.02	–	>50	4.1
21		3-Cl	–	40/95	>50	2.0
22		4-F	–	12/97	>50	2.2
23		3-F	–	3/94	>50	0
24		3-Br	–	38/97	>50	2.2
25		3-CF3	–	19/89	>50	1.2
26			2.5±0.03	–	>50	−2.5
27			>20	–	>50	1.1

^aHalf maximal effective concentration: mean ± standard deviation (SD) from at least two independent experiments.

^b50% cytotoxic concentration: mean ± SD from at least two independent experiments.

^cΔT_m: melting point change of CA hexamer in presence of a compound compared to DMSO control.

^dValues for PF74 also reported in Refs. 43 and 44. −Not applicable.

By far the most interesting series is the benzamide subtype (18−27). With the exception of 23 and 27, all analogs of this subtype demonstrated discernible HIV-1 inhibition at 2 μmol/L and complete inhibition at 20 μmol/L. Three analogs potently inhibited HIV-1 in dose response fashion (19: EC₅₀ = 1.6 μmol/L; 20: EC₅₀ = 0.88 μmol/L; 26: EC₅₀ = 2.5 μmol/L) and moderately impacted the stability of CA hexamer (19: ΔT_m = 5.0; 20: ΔT_m = 4.1; 26: ΔT_m = −2.5). It is not clear why 26 destabilized while 19 and 20 stabilized CA hexamer. Notably compound 27 has a proline insert between the phenylalanine core and the aniline terminus, rendering a substantially different backbone, which evidently does not favor binding to HIV-1 CA. Overall, scaffold-hopping on the indole ring allowed us to identify four analogs (10, 19, 20 and 26) of two different subtypes with low to sub micromolar antiviral activities (EC₅₀ = 0.88−2.5 μmol/L). Particularly important is the benzamide subtype, which generally conferred significant antiviral activity, with compound 20 (EC₅₀ = 0.88 μmol/L) exhibiting submicromolar potency comparable to PF74 (EC₅₀ = 0.61 μmol/L).

2.3. Metabolic stability in liver microsomes

A serious drawback of PF74 is the lack of metabolic stability^41,42. Cytochrome P450 (CYPs)-mediated oxidative metabolism constitutes a major pharmacokinetic barrier for oral drugs⁴⁸. Since the vast majority of metabolizing enzymes are found in the liver, measuring in vitro metabolic stability in liver microsomes can reliably predict drug oral bioavailability⁴⁹. Low stability in liver microsomes, as signified by short half-life (t_1/2), predicts excessive first pass liver metabolism and poor oral bioavailability. To assess the metabolic stability of our analogs, we tested the four most potent antiviral compounds (10, 19, 20 and 26), along with PF74 and two other analogs (18 and 23), in both human liver microsomes (HLMs) and mouse liver microsomes (MLMs). In addition, we also measured the microsomal stability of the four best compounds in the presence of cobicistat (Cobi)⁵⁰, a CYP3A inhibitor. CYP3A is a liver metabolizing enzyme responsible for the metabolism of more than half of all drugs⁵¹. Peptidomimetics, such as PF74, are well-known substrates for CYP3A⁵² and are highly susceptible to oxidative metabolism. By inhibiting CYP3A, Cobi acts as a pharmacokinetic (PK) enhancer⁵³ to boost the plasma concentration of peptidomimetic drugs, such as HIV-1 protease inhibitors⁵³. The metabolic stability results are summarized in Table 2. In our assays, the half-life of PF74 was less than 1 min in both HLMs and MLMs, consistent with previous reports^41,42. Remarkably, all six of our tested analogs demonstrated hugely improved metabolic stability in both HLMs (t_1/2 = 12−36 min, 17−50-fold increase over PF74) and MLMs (t_1/2 = 2.4−13 min, 4−22-fold increase over PF74). These results indicate that scaffold-hopping of the indole ring with more electron-deficient moieties could lead to resistance to phase I metabolism. When measured in the presence of Cobi, the half-lives were drastically increased, except for compound 19 in HLMs. Based on our metabolic stability studies, compound 20, with half-lives of 36 min (51-fold over PF74) in HLMs and 13 min (22-fold over PF74) in MLMs, is predicted to have favorable oral bioavailability.

Table 2.

Half-life (t_1/2, min) of selected compounds in liver microsomes.

Compd.	t1/2 (min)
	HLM	HLM (+Cobia)	MLM	MLM (+Cobia)
PF74b	0.7	91	0.6	34
10	12	>120	4.6	>120
18	15	–	3.6	–
19	22	42	5.0	34
20	36	>120	13	52
23	12	–	3.6	–
26	14	>120	2.4	28

^aMeasured in the presence of CYP3A inhibitor Cobi.

^bValues for PF74 were also reported in our previous publications^43,44. –Not applicable.

2.4. Molecular modeling

To study the potential binding mode of our analogs and understand the SAR, we conducted molecular docking for a few representative compounds, including potent compounds (10 and 20), and less potent (4) or inactive (2 and 17) analogs. The modeling used the co-crystal structure of PF74-bound HIV-1 CA (PDB code: 4XFZ³⁶). Compound 2 (yellow, Fig. 2A) which bears an imidazolidine-2,4-dione in place of indole ring, was completely inactive, probably because of the potential loss of key H-bonds with N57 and Q63 and the cation‒π interaction with K70, resulting a very poor docking score (Glide score: −2.7 for 2 vs. −5.8 for PF74). Compound 4 (pink, Fig. 2A), the aniline 4-Cl congener of 2, adopted a conformation and binding mode similar to 2, except that the 4-Cl formed a halogen-bond with N74, which likely contributed to the activity observed with 4. Significantly, a 6-membered pyrimidine-2,4-dione in compound 10 is predicted to restore key interactions with N57 and Q63 (Fig. 2B). These interactions also allowed compound 10 to fit in the PF74 pocket and interact with all other key residues, which could explain the observed improved potency (17% inhibition at 2 μmol/L for 4 vs. EC₅₀: 1.7 μmol/L for 10). However, when the same pyrimidine-2,4-dione scaffold is connected to the phenylalanine core via the C5 site (vs. the N-1 site in compound 10) and the methylene linker deleted, the resulting compound 17 (Fig. 2C) appeared to be forced out of the PF74 pocket and bound away from the preferred site (circled). This may explain why compound 17 was completely inactive. Finally, the most interesting modeling results were observed with compound 20 which features a benzamide moiety as the indole replacement. Important docking observations for 20 (Fig. 2D) included 1) the aniline moiety and the phenylalanine core appear bound in a very similar way as in PF74; 2) the 4-Cl of compound 20 formed a halogen bond⁵⁴ with N74; 3) the benzamide moiety adopted an orientation similar to the indole ring of PF74 and formed a similar H-bond with Q63 (via the amide NH); 4) the interaction with K70 was via an H-bond by the amide carbonyl in 10 as opposed to the π−cation interaction in PF74. This potential binding mode for compound 20 is likely to contribute to the observed potent antiviral activity and increased CA hexamer stabilization (Table 1).

Figure 2.

Docking of selected compounds into PF74-bound HIV-1 CA (PDB ID: 4XFZ³⁶). Predicted binding modes of: (A) compounds 2 (yellow sticks) and 4 (pink sticks). Glide scores (kcal/mol): −2.7 for 2 and −3.0 for 4; (B) compound 10. Glide score: −6.4 kcal/mol; (C) compound 17, away from the binding site (circled). Glide score: −4.7 kcal/mol; (D) compound 20. Glide score: −6.7 kcal/mol. Hydrogen-bonding, halogen-bonding, and cation‒π interactions are depicted as black dotted lines, pink dashed lines, and double-headed arrows, respectively. CA is shown in cyan cartoon with key residues around binding site shown as sticks, and ligands 10, 17, and 20 are shown as pink sticks. The nitrogen, oxygen, and chlorine atoms are colored blue, red, and green, respectively.

3. Conclusions

Although PF74 is a potent and well-characterized CA-targeting antiviral small molecule, it lacks metabolic stability to be an antiviral lead. The indole ring of PF74 is a particularly interesting handle for exploring both the potency and metabolic stability for two reasons: 1) it interacts with both the CA_NTD and the adjacent CA_CTD; 2) it is highly electron-rich, and thus could be susceptible to oxidative metabolism. In this work, we used a scaffold-hopping approach to replace the inter-CA domain-interacting and electron-rich indole ring with electron-deficient moieties, including imidazolidine-2,4-dione, pyrimidine-2,4-dione, and benzamide, and identified four potent antiviral compounds (10, 19, 20 and 26) with dramatically improved metabolic stability. Our best compound, analog 20, exhibited submicromolar potency comparable to PF74 and much longer half-lives than PF74 in HLMs (51-fold) and MLMs (22-fold). Considerably longer half-life was also observed when metabolic stability was measured in the presence of PK enhancer, Cobi. Molecular modeling corroborated the binding of 20 to the PF74 pocket and revealed interesting potential binding interactions conferred by the benzamide moiety. Altogether, these data suggest that analog 20 can be a potentially viable antiviral lead.

4. Experimental

4.1. Chemistry

All commercial chemicals were used as supplied unless indicated otherwise. Compounds were purified via flash chromatography using a Combiflash RF-200 (Teledyne ISCO, Lincoln, NE, USA) with RediSep columns (silica) and indicated mobile phase. ¹H and ¹³C NMR spectra were recorded on a Varian 600 MHz (Agilent Technologies, Santa Clara, CA, USA) or Bruker 400 spectrometer (Bruker, Billerica, MA, USA). Diastereomeric ratio was determined by ¹H NMR analysis. Mass data were acquired using an Agilent 6230 TOF LC/MS spectrometer (Agilent Technologies). PF74 was synthesized according to reported procedures^43,44.

4.1.1. General procedure for synthesis of 2−27

HATU (2 equiv.) and DIPEA (3 equiv.) were added to a solution of an acid derivative (1 equiv.) in DMF. After the mixture was stirred for 20 min at room temperature, amine (1.2 equiv.) was added, and the resulting mixture was further stirred overnight at room temperature. To the reaction mixture was added H₂O and the mixture was extracted with ethyl acetate three times. The organic layers were combined, washed with brine, dried over anhydrous MgSO₄, and concentrated in vacuum. The crude mixture was purified using Combi-flash on silica gel (EtOAc in hexanes as eluent).

4.1.1.1. (2S)-2-(2-(2,5-Dioxoimidazolidin-4-yl)acetamido)-N-methyl-N,3-diphenylpropanamide (2)

Yield 70%, dr 1.1:1. ¹H NMR (600 MHz, CD₃OD) δ 8.18−8.15 (m, 1H), 7.28−7.24 (m, 3H), 7.11−7.09 (m, 3H), 6.93−6.76 (m, 4H), 4.58−4.52 (m, 1H), 4.23−4.18 (m, 1H), 3.10−3.07 (m, 3H), 2.87−2.83 (m, 1H), 2.70−2.58 (m, 2H), 2.47−2.41 (m, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 177.4, 177.2, 173.5, 173.2, 171.0, 170.5, 159.9, 159.7, 144.0, 143.9, 138.1, 130.8, 130.3, 130.2, 129.5, 129.3, 128.7, 128.6, 127.9, 56.9, 56.6, 53.8, 53.4, 39.4, 39.1, 38.3, 38.2, 38.1, 38.0; HRMS-ESI (−) m/z Calcd. for C₂₁H₂₁N₄O₄ [M−H]⁻ 393.1568, Found 393.1573.

4.1.1.2. (2S)-2-(2-(2,5-Dioxoimidazolidin-4-yl)acetamido)-N-methyl-3-phenyl-N-(p-tolyl)propanamide (3)

Yield 65%, dr 1.3:1. ¹H NMR (600 MHz, CDCl₃) δ 9.52−9.36 (m, 1H), 7.52−7.36 (m, 1H), 7.22−7.14 (m, 4H), 6.91−6.78 (m, 4H), 4.85−4.77 (m, 1H), 4.35−4.32 (m, 1H), 2.92−2.81 (m, 2H), 2.74−2.70 (m, 1H), 2.51−2.45 (m, 1H), 2.37−2.36 (m, 3H); ¹³C NMR (150 MHz, CDCl₃) δ 174.7, 174.4, 172.1, 171.7, 169.1, 168.5, 157.3, 157.2, 139.6, 139.5, 138.4, 138.3, 136.2, 130.4, 129.4, 129.2, 128.5, 128.4, 126.9, 126.9, 55.8, 55.5, 51.9, 51.3, 39.1, 38.5, 38.0, 37.9, 37.7, 37.4, 21.1; HRMS-ESI (−) m/z Calcd. for C₂₂H₂₃N₄O₄ [M−H]⁻ 407.1725, Found 407.1730.

4.1.1.3. (2S)-N-(4-Chlorophenyl)-2-(2-(2,5-dioxoimidazolidin-4-yl)acetamido)-N-methyl-3-phenylpropanamide (4)

Yield 82%, dr 2.3:1. ¹H NMR (600 MHz, CD₃OD) δ 7.34−7.24 (m, 6H), 6.98−6.86 (m, 3H), 4.61−4.55 (m, 1H), 4.33−4.30 (m, 1H), 3.14−3.12 (m, 3H), 2.98−2.92 (m, 1H), 2.79−2.72 (m, 2H), 2.58−2.54 (m, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 177.5, 173.4, 173.1, 170.9, 170.4, 160.0, 142.6, 142.5, 138.0, 137.9, 134.9, 130.7, 130.4, 130.3, 129.6, 128.1, 56.8, 56.7, 53.6, 53.2, 39.7, 39.4, 38.2, 38.1, 37.9; HRMS-ESI (−) m/z Calcd. for C₂₁H₂₀ClN₄O₄ [M−H]⁻ 427.1179, Found 427.1183.

4.1.1.4. (2S)-N-(3-Chlorophenyl)-2-(2-(2,5-dioxoimidazolidin-4-yl)acetamido)-N-methyl-3-phenylpropanamide (5)

Yield 83%, dr 1.2:1. ¹H NMR (600 MHz, CD₃OD) δ 7.34−7.25 (m, 6H), 6.97−6.65 (m, 3H), 4.61−4.54 (m, 1H), 4.35−4.32 (m, 1H), 3.14−3.12 (m, 3H), 2.97−2.92 (m, 1H), 2.80−2.71 (m, 2H), 2.58−2.53 (m, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 177.4, 177.3, 173.3, 173.1, 171.0, 170.5, 159.9, 159.8, 145.1, 137.9, 137.8, 135.9, 135.8, 131.8, 130.3, 129.7, 129.4, 128.7, 128.2, 127.3, 56.8, 56.7, 53.8, 53.4, 39.7, 39.4, 38.2, 38.1, 37.9; HRMS-ESI (−) m/z Calcd. for C₂₁H₂₀ClN₄O₄ [M−H]⁻ 427.1179, Found 427.1184.

4.1.1.5. (2S)-2-(2-(2,5-Dioxoimidazolidin-4-yl)acetamido)-N-(4-fluorophenyl)-N-methyl-3-phenylpropanamide (6)

Yield 66%, dr 1.6:1. ¹H NMR (600 MHz, CD₃OD) δ 7.26−7.23 (m, 3H), 7.06−6.93 (m, 6H), 4.61−4.54 (m, 1H), 4.33−4.29 (m, 1H), 3.15−3.12 (m, 3H), 2.98−2.93 (m, 1H), 2.79−2.69 (m, 2H), 2.58−2.53 (m, 1H); HRMS-ESI (−) m/z Calcd. for C₂₁H₂₀FN₄O₄ [M−H]⁻ 411.1474, Found 411.1475.

4.1.1.6. (2S)-2-(2-(2,5-Dioxoimidazolidin-4-yl)acetamido)-N-(3-fluorophenyl)-N-methyl-3-phenylpropanamide (7)

Yield 75%, dr 1.7:1. ¹H NMR (600 MHz, CD₃OD) δ 7.36−7.24 (m, 5H), 7.10−7.06 (m, 1H), 6.96−6.81 (m, 3H), 4.64−4.58 (m, 1H), 4.35−4.31 (m, 1H), 3.16−3.14 (m, 3H), 2.98−2.93 (m, 1H), 2.80−2.70 (m, 2H), 2.58−2.52 (m, 1H); HRMS-ESI (−) m/z Calcd. for C₂₁H₂₀FN₄O₄ [M−H]⁻ 411.1474, Found 411.1479.

4.1.1.7. (S)-N-Methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-N,3-diphenylpropanamide (8)

Yield 75%. ¹H NMR (600 MHz, CDCl₃) δ 9.35 (s, 1H), 7.50 (s, 1H), 7.38−7.35 (m, 3H), 7.19−7.17 (m, 3H), 6.94−6.86 (m, 4H), 4.84 (q, J = 7.8 Hz, 1H), 4.43 (d, J = 15.8 Hz, 1H), 4.21 (d, J = 15.8 Hz, 1H), 3.26 (s, 3H), 2.92 (dd, J = 13.4, 6.8 Hz, 1H), 2.74 (dd, J = 13.4, 7.6 Hz, 1H), 1.89 (s, 3H); ¹³C NMR (150 MHz, CDCl₃) δ 171.2, 166.0, 164.1, 151.1, 142.2, 140.4, 136.0, 129.9, 129.3, 128.4, 128.4, 127.2, 126.9, 110.9, 51.5, 49.6, 38.9, 37.8, 12.4; HRMS-ESI (−) m/z Calcd. for C₂₃H₂₃N₄O₄ [M−H]⁻ 419.1725, Found 419.1730.

4.1.1.8. (S)-N-Methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenyl-N-(p-tolyl)propanamide (9)

Yield 75%. ¹H NMR (600 MHz, CDCl₃) δ 9.47 (s, 1H), 7.61 (d, J = 8.5 Hz, 1H), 7.19−7.14 (m, 5H), 6.95−6.89 (m, 3H), 6.80 (brs, 1H), 4.87–4.83 (m, 1H), 4.45 (d, J = 15.8 Hz, 1H), 4.20 (d, J = 15.8 Hz, 1H), 3.23 (s, 3H), 2.92 (dd, J = 13.4, 6.9 Hz, 1H), 2.74 (dd, J = 13.4, 7.5 Hz, 1H), 2.36 (s, 3H), 1.89 (s, 3H); ¹³C NMR (150 MHz, CDCl₃) δ 171.4, 166.0, 164.2, 151.1, 140.5, 139.6, 138.3, 136.1, 130.4, 129.4, 128.4, 126.9, 126.8, 110.9, 51.4, 49.5, 39.0, 37.9, 21.1, 12.4; HRMS-ESI (−) m/z Calcd. for C₂₄H₂₅N₄O₄ [M−H]⁻ 433.1881, Found 433.1890.

4.1.1.9. (S)-N-(4-Chlorophenyl)-N-methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenylpropanamide (10)

Yield 79%. ¹H NMR (600 MHz, CDCl₃) δ 9.48 (s, 1H), 7.56 (d, J = 8.4 Hz, 1H), 7.28 (d, J = 8.6 Hz, 2H), 7.23–7.22 (m, 3H), 6.98–6.97 (m, 1H), 6.94–6.92 (m, 2H), 6.73 (brs, 1H), 4.78 (dd, J = 15.3, 8.0 Hz, 1H), 4.43 (d, J = 15.8 Hz, 1H), 4.24 (d, J = 15.8 Hz, 1H), 3.20 (s, 3H), 2.93 (dd, J = 13.2, 8.0 Hz, 1H), 2.78 (dd, J = 13.2, 6.9 Hz, 1H), 1.91 (s, 3H); ¹³C NMR (150 MHz, CDCl₃) δ 171.3, 166.1, 164.2, 151.1, 140.6, 140.4, 135.8, 134.1, 129.9, 129.4, 128.6, 128.5, 127.0, 111.0, 51.5, 49.6, 39.2, 37.8, 12.4; HRMS-ESI (−) m/z Calcd. for C₂₃H₂₂ClN₄O₄ [M−H]⁻ 453.1335, Found 453.1338.

4.1.1.10. (S)-N-(3-Chlorophenyl)-N-methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenylpropanamide (11)

Yield 70%. ¹H NMR (600 MHz, CD₃OD) δ 7.33–7.26 (m, 7H), 6.96 (m, 3H), 4.59 (t, J = 7.6 Hz, 1H), 4.40 (q, J = 16.4 Hz, 2H), 3.13 (s, 3H), 2.98 (dd, J = 13.0, 8.7 Hz, 1H), 2.79 (dd, J = 13.1, 6.6 Hz, 1H), 1.86 (s, 3H); ¹³C NMR (150 MHz, CD₃OD) δ 172.9, 168.9, 167.0, 153.0, 145.0, 143.6, 137.7, 135.9, 131.9, 130.3, 129.7, 129.4, 128.7, 128.2, 127.3, 111.0, 53.5, 50.5, 39.6, 38.0, 12.2; HRMS-ESI (−) m/z Calcd. for C₂₃H₂₂ClN₄O₄ [M−H]⁻ 453.1335, Found 453.1340.

4.1.1.11. (S)-N-(4-Fluorophenyl)-N-methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenylpropanamide (12)

Yield 90%. ¹H NMR (600 MHz, CD₃OD) δ 7.25–7.23 (m, 4H), 7.03–6.95 (m, 6H), 4.58 (t, J = 7.5 Hz, 1H), 4.41–4.35 (m, 2H), 3.14 (s, 3H), 2.98 (dd, J = 13.2, 8.1 Hz, 1H), 2.77 (dd, J = 13.2, 7.0 Hz, 1H), 1.86 (s, 3H); ¹³C NMR (150 MHz, CD₃OD) δ 173.1, 168.9, 167.0, 163.4 (d, J_CF = 247.1 Hz), 153.0, 143.6, 139.9 (d, J_CF = 3.0 Hz), 137.9, 130.7, 130.4, 129.6, 128.1, 117.4 (d, J_CF = 23.0 Hz), 111.0, 53.4, 50.5, 39.4, 38.1, 12.2; HRMS-ESI (−) m/z Calcd. for C₂₃H₂₂FN₄O₄ [M−H]⁻ 437.1631, Found 437.1635.

4.1.1.12. (S)-N-(3-Fluorophenyl)-N-methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenylpropanamide (13)

Yield 82%. ¹H NMR (600 MHz, CDCl₃) δ 9.77 (s, 1H), 7.79 (d, J = 8.2 Hz, 1H), 7.32−7.29 (m, 1H), 7.25−7.21 (m, 3H), 7.04−7.00 (m, 2H), 6.93−6.92 (m, 2H), 6.77 (s, 1H), 6.37 (s, 1H), 4.84−4.80 (m, 1H), 4.46−4.28 (m, 2H), 3.21 (s, 3H), 2.94 (dd, J = 13.2, 8.1 Hz, 1H), 2.80 (dd, J = 13.2, 6.9 Hz, 1H), 1.91 (s, 3H); ¹³C NMR (100 MHz, CDCl₃) δ 171.4, 166.2, 164.3, 162.8 (d, J_CF = 249.4 Hz), 151.3, 143.5 (d, J_CF = 9.4 Hz), 140.6, 135.9, 130.9 (d, J_CF = 9.1 Hz), 129.3, 128.5, 127.1, 123.2, 115.5 (d, J_CF = 20.9 Hz), 114.7 (d, J_CF = 22.3 Hz), 111.0, 51.7, 49.6, 39.2, 37.7, 12.4; HRMS-ESI (−) m/z Calcd. for C₂₃H₂₂FN₄O₄ [M−H]⁻ 437.1631, Found 437.1632.

4.1.1.13. (S)-N-(3-Bromophenyl)-N-methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenylpropanamide (14)

Yield 70%. ¹H NMR (600 MHz, CD₃OD) δ 7.47 (d, J = 7.9 Hz, 1H), 7.28−7.22 (m, 6H), 7.01−6.95 (m, 3H), 4.60–4.57 (m, 1H), 4.40 (q, J = 16.5 Hz, 2H), 3.13 (s, 3H), 2.97 (dd, J = 13.1, 8.7 Hz, 1H), 2.79 (dd, J = 13.1, 6.6 Hz, 1H), 1.86 (s, 3H); ¹³C NMR (150 MHz, CD₃OD) δ 172.8, 168.9, 167.0, 153.0, 145.1, 143.6, 137.6, 132.4, 132.1, 131.5, 130.3, 129.7, 128.2, 127.7, 123.6, 111.0, 53.5, 50.5, 39.6, 38.0, 12.2; HRMS-ESI (−) m/z Calcd. for C₂₃H₂₂BrN₄O₄ [M−H]⁻ 497.0830, Found 497.0835.

4.1.1.14. (S)-N-Methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-3-phenyl-N-(3-(trifluoromethyl)phenyl)propanamide (15)

Yield 70%. ¹H NMR (600 MHz, CD₃OD) δ 7.61 (d, J = 7.9 Hz, 1H), 7.51 (t, J = 7.8 Hz, 1H), 7.34–7.22 (m, 6H), 6.95–6.93 (m, 2H), 4.55 (t, J = 8.1 Hz, 1H), 4.40 (q, J = 16.5 Hz, 2H), 3.16 (s, 3H), 2.98 (dd, J = 13.0, 8.7 Hz, 1H), 2.79 (dd, J = 13.1, 6.5 Hz, 1H), 1.87 (s, 3H); ¹³C NMR (150 MHz, CD₃OD) δ 172.9, 168.9, 167.0, 153.0, 144.5, 143.6, 137.6, 132.9 (q, J_CF = 33 Hz), 132.8, 131.7, 130.2, 129.7, 128.2, 127.4 (q, J_CF = 3.5 Hz), 125.4, 124.9 (q, J_CF = 271.7 Hz), 111.0, 53.5, 50.5, 39.6, 38.0, 12.2; HRMS-ESI (−) m/z Calcd. for C₂₄H₂₂F₃N₄O₄ [M−H]⁻ 487.1599, Found 487.1604.

4.1.1.15. (R)-N-Methyl-2-(2-(5-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)acetamido)-N,2-diphenylacetamide (16)

Yield 69%. ¹H NMR (600 MHz, CDCl₃) δ 8.73 (s, 1H), 7.63 (t, J = 7.1 Hz, 1H), 7.33−7.28 (m, 3H), 7.23−7.15 (m, 3H), 6.99 (s, 1H), 6.92−6.88 (m, 3H), 5.51 (d, J = 7.1 Hz, 1H), 4.43−4.25 (m, 2H), 3.27 (s, 3H), 1.88 (s, 3H); ¹³C NMR (150 MHz, CDCl₃) δ 169.5, 165.5, 163.9, 150.9, 141.7, 140.7, 136.6, 129.7, 128.6, 128.5, 128.2, 127.9, 127.8, 110.7, 55.0, 49.6, 38.1, 12.3; HRMS-ESI (−) m/z Calcd. for C₂₂H₂₁N₄O₄ [M−H]⁻ 405.1569, Found 405.1568.

4.1.1.16. (S)-N-(1-(Methyl(phenyl)amino)-1-oxo-3-phenylpropan-2-yl)-2,4-dioxo-1,2,3,4-tetrahydropyrimidine-5-carboxamide (17)

Yield 59%. ¹H NMR (600 MHz, DMSO-d₆) δ 11.64 (s, 1H), 11.62 (s, 1H), 9.05 (d, J = 7.6 Hz, 1H), 8.00 (s, 1H), 7.42–7.38 (m, 3H), 7.17–7.14 (m, 5H), 6.81 (d, J = 6.6 Hz, 2H), 4.65–4.62 (m, 1H), 3.11 (s, 3H), 2.88 (dd, J = 13.6, 5.6 Hz, 1H), 2.63 (dd, J = 13.6, 8.5 Hz, 1H); ¹³C NMR (100 MHz, DMSO-d₆) δ 170.8, 164.5, 161.7, 150.8, 148.3, 143.2, 137.3, 130.1, 129.3, 128.8, 128.4, 127.9, 127.1, 104.2, 51.5, 38.7, 37.6; HRMS-ESI (−) m/z Calcd. for C₂₁H₁₉N₄O₄ [M−H]⁻ 391.1413, Found 391.1415.

4.1.1.17. (S)-N-(2-((1-(Methyl(phenyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (18)

Yield 79%. ¹H NMR (600 MHz, CDCl₃) δ 7.80 (d, J = 7.4 Hz, 2H), 7.50 (t, J = 7.3 Hz, 1H), 7.42 (t, J = 7.7 Hz, 1H), 7.37–7.35 (m, 3H), 7.20–7.18 (m, 3H), 6.91–6.90 (m, 3H), 6.83 (brs, 1H), 4.85 (q, J = 7.4 Hz, 1H), 4.15–4.07 (m, 2H), 3.24 (s, 3H), 2.94 (dd, J = 13.3, 7.2 Hz, 1H), 2.76 (dd, J = 13.3, 7.1 Hz, 1H); ¹³C NMR (150 MHz, CDCl₃) δ 171.1, 168.0, 167.4, 142.3, 136.0, 133.7, 131.7, 129.8, 129.3, 128.5, 128.5, 128.3, 127.3, 127.1, 127.0, 51.4, 43.2, 39.2, 37.7; HRMS-ESI (−) m/z Calcd. for C₂₅H₂₄N₃O₃ [M−H]⁻ 414.1824, Found 414.1823.

4.1.1.18. (S)-N-(2-((1-(Methyl(p-tolyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (19)

Yield 82%. ¹H NMR (600 MHz, CDCl₃) δ 7.80 (d, J = 7.4 Hz, 2H), 7.50 (t, J = 7.4 Hz, 1H), 7.43−7.41 (m, 2H), 7.21−7.20 (m, 3H), 7.14 (d, J = 8.0 Hz, 2H), 6.94−6.90 (m, 4H), 6.76 (brs, 1H), 4.86 (q, J = 7.3 Hz, 1H), 4.14−4.08 (m, 2H), 3.22 (s, 3H), 2.95 (dd, J = 13.3, 7.2 Hz, 1H), 2.77 (dd, J = 13.3, 7.0 Hz, 1H), 2.37 (s, 3H); ¹³C NMR (150 MHz, CDCl₃) δ 171.2, 167.9, 167.3, 139.7, 138.2, 136.1, 133.8, 131.7, 130.4, 129.3, 128.5, 128.4, 127.1, 127.0, 126.9, 51.3, 43.1, 39.2, 37.8, 21.1; HRMS-ESI (−) m/z Calcd. for C₂₆H₂₆N₃O₃ [M−H]⁻ 428.1988, Found 428.1989.

4.1.1.19. (S)-N-(2-((1-((4-Chlorophenyl)(methyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (20)

Yield 90%. ¹H NMR (600 MHz, CDCl₃) δ 7.81 (d, J = 7.2 Hz, 2H), 7.51 (t, J = 7.4 Hz, 1H), 7.43 (t, J = 7.7 Hz, 2H), 7.28−7.22 (m, 5H), 7.03–6.96 (m, 4H), 6.69 (s, 1H), 4.79 (dd, J = 14.7, 8.2 Hz, 1H), 4.14 (d, J = 7.4 Hz, 2H), 3.18 (s, 3H), 2.96 (dd, J = 13.1, 8.4 Hz, 1H), 2.81 (dd, J = 13.1, 6.4 Hz, 1H); ¹³C NMR (150 MHz, CDCl₃) δ 171.2, 168.1, 167.4, 140.8, 135.9, 134.0, 133.7, 131.8, 129.8, 129.4, 128.7, 128.6, 128.6, 127.1, 127.1, 51.4, 43.2, 39.5, 37.7; HRMS-ESI (−) m/z Calcd. for C₂₅H₂₃ClN₃O₃ [M−H]⁻ 448.1433, Found 448.1440.

4.1.1.20. (S)-N-(2-((1-((3-Chlorophenyl)(methyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (21)

Yield 79%. ¹H NMR (600 MHz, CD₃OD) δ 7.87 (d, J = 7.4 Hz, 2H), 7.54 (t, J = 7.4 Hz, 1H), 7.46 (t, J = 7.7 Hz, 2H), 7.33–7.22 (m, 6H), 6.96–6.94 (m, 3H), 4.62 (t, J = 7.8 Hz, 1H), 4.04 (s, 2H), 3.13 (s, 3H), 2.96 (dd, J = 13.1, 8.5 Hz, 1H), 2.79 (dd, J = 13.1, 6.6 Hz, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 173.0, 171.0, 170.4, 145.1, 137.7, 135.9, 135.0, 132.9, 131.8, 130.3, 129.7, 129.6, 129.4, 128.7, 128.5, 128.1, 127.3, 53.4, 43.7, 39.6, 38.0; HRMS-ESI (−) m/z Calcd. for C₂₅H₂₃ClN₃O₃ [M−H]⁻ 448.1433, Found 448.1439.

4.1.1.21. (S)-N-(2-((1-((4-Fluorophenyl)(methyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (22)

Yield 79%. ¹H NMR (600 MHz, CD₃OD) δ 7.86 (d, J = 7.2 Hz, 2H), 7.55 (t, J = 8.4 Hz, 1H), 7.48–7.45 (m, 3H), 7.22–7.21 (m, 3H), 7.06–7.03 (m, 3H), 6.95–6.94 (m, 2H), 4.62 (t, J = 7.5 Hz, 1H), 4.02 (s, 2H), 3.14 (s, 3H), 2.97 (dd, J = 13.2, 7.9 Hz, 1H), 2.77 (dd, J = 13.2, 7.0 Hz, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 173.2, 171.0, 170.4, 163.4 (d, J_CF = 246.9 Hz), 140.0, 137.9, 135.0, 132.9, 130.7 (d, J_CF = 7.5 Hz), 130.3, 129.6, 129.5, 128.5, 128.0, 117.3 (d, J_CF = 23.1 Hz), 53.2, 43.7, 39.4, 38.1; HRMS-ESI (−) m/z Calcd. for C₂₅H₂₃FN₃O₃ [M−H]⁻ 432.1729, Found 432.1734.

4.1.1.22. (S)-N-(2-((1-((3-Fluorophenyl)(methyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (23)

Yield 80%. ¹H NMR (600 MHz, CDCl₃) δ 7.82 (d, J = 7.6 Hz, 2H), 7.51 (t, J = 7.4 Hz, 1H), 7.44−7.42 (m, 2H), 7.32−7.28 (m, 1H), 7.25−7.21 (m, 3H), 7.04−6.95 (m, 5H), 6.73 (s, 1H), 6.33 (s, 1H), 4.85−4.81 (m, 1H), 4.17−4.11 (m, 2H), 3.20 (s, 3H), 2.96 (dd, J = 13.1, 8.4 Hz, 1H), 2.82 (dd, J = 13.1, 6.4 Hz, 1H); ¹³C NMR (100 MHz, CDCl₃) δ 171.2, 168.2, 167.5, 162.8 (d, J_CF = 249.3 Hz), 143.7 (d, J_CF = 9.5 Hz), 135.9, 133.7, 131.8, 130.8 (d, J_CF = 9.1 Hz), 129.3, 128.6, 128.6, 127.2, 123.2 (d, J_CF = 2.9 Hz), 115.4 (d, J_CF = 20.9 Hz), 114.7 (d, J_CF = 22.2 Hz), 51.5, 43.2, 39.5, 37.6; HRMS-ESI (−) m/z Calcd. for C₂₅H₂₃FN₃O₃ [M−H]⁻ 432.1729, Found 432.1732.

4.1.1.23. (S)-N-(2-((1-((3-Bromophenyl)(methyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (24)

Yield 80%. ¹H NMR (600 MHz, CD₃OD) δ 7.87 (d, J = 7.3 Hz, 2H), 7.54 (t, J = 7.4 Hz, 1H), 7.48–7.45 (m, 3H), 7.25–7.22 (m, 5H), 6.98–6.95 (m, 3H), 4.62 (t, J = 7.8 Hz, 1H), 4.04 (s, 2H), 3.13 (s, 3H), 2.96 (dd, J = 13.1, 8.6 Hz, 1H), 2.79 (dd, J = 13.1, 6.6 Hz, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 173.0, 171.0, 170.4, 145.2, 137.7, 135.0, 132.9, 132.4, 132.1, 131.5, 130.3, 129.7, 129.6, 128.5, 128.2, 127.7, 123.6, 53.4, 43.7, 39.6, 38.0; HRMS-ESI (−) m/z Calcd. for C₂₅H₂₃BrN₃O₃ [M−H]⁻ 492.0928, Found 492.0933.

4.1.1.24. (S)-N-(2-((1-(Methyl(3-(trifluoromethyl)phenyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (25)

Yield 89%. ¹H NMR (600 MHz, CD₃OD) δ 7.87 (d, J = 7.5 Hz, 2H), 7.63 (d, J = 7.7 Hz, 1H), 7.56–7.46 (m, 5H), 7.24–7.19 (m, 4H), 6.94 (d, J = 6.8 Hz, 2H), 4.60–4.57 (m, 1H), 4.04 (s, 2H), 3.17 (s, 3H), 2.97 (dd, J = 13.0, 8.6 Hz, 1H), 2.80 (dd, J = 13.1, 6.6 Hz, 1H); ¹³C NMR (150 MHz, CD₃OD) δ 173.1, 171.1, 170.4, 144.6, 137.7, 135.1, 132.9, 132.8, 131.7, 130.2, 129.7, 129.6, 128.5, 128.2, 126.0, 125.4, 124.9 (q, J_CF = 272.6 Hz), 124.0, 53.4, 43.7, 39.6, 38.0; HRMS-ESI (−) m/z Calcd. for C₂₆H₂₃F₃N₃O₃ [M−H]⁻ 482.1697, Found 482.1700.

4.1.1.25. (S)-N-(2-((1-(Ethyl(phenyl)amino)-1-oxo-3-phenylpropan-2-yl)amino)-2-oxoethyl)benzamide (26)

Yield 70%. ¹H NMR (600 MHz, CDCl₃) δ 7.81 (d, J = 7.6 Hz, 2H), 7.49−7.45 (m, 2H), 7.40−7.34 (m, 5H), 7.26−7.25 (m, 1H), 7.19−7.16 (m, 3H), 6.93−6.92 (m, 2H), 4.73 (q, J = 7.4 Hz, 1H), 4.13−4.08 (m, 2H), 3.85−3.79 (m, 1H), 3.66−3.60 (m, 1H), 2.97 (dd, J = 13.3, 7.1 Hz, 1H), 2.78 (dd, J = 13.3, 7.3 Hz, 1H), 1.09 (t, J = 7.2 Hz, 3H); ¹³C NMR (100 MHz, CDCl₃) δ 170.8, 168.3, 167.4, 140.7, 136.3, 133.8, 131.6, 129.6, 129.4, 128.5, 128.4, 128.4, 127.2, 126.9, 51.8, 44.8, 43.1, 39.1, 12.8; HRMS-ESI (−) m/z Calcd. for C₂₆H₂₆N₃O₃ [M−H]⁻ 428.1988, Found 428.1990.

4.1.1.26. (S)-1-(Benzoylglycyl-L-phenylalanyl)-N-phenylpyrrolidine-2-carboxamide (27)

Yield 68%. ¹H NMR (600 MHz, CDCl₃) δ 9.73 (s, 1H), 8.31 (d, J = 9.1 Hz, 1H), 7.82 (d, J = 7.3 Hz, 2H), 7.59 (d, J = 7.7 Hz, 2H), 7.54 (t, J = 7.4 Hz, 1H), 7.47−7.44 (m, 2H), 7.29−7.27 (m, 2H), 7.12−7.07 (m, 3H), 7.04−7.00 (m, 3H), 5.13−5.10 (m, 1H), 4.85−4.83 (m, 1H), 4.49−4.45 (m, 1H), 4.14–4.11 (m, 1H), 3.85–3.81 (m, 1H), 3.59–3.56 (m, 1H), 3.12 (dd, J = 13.7, 6.0 Hz, 1H), 2.85 (dd, J = 13.7, 7.8 Hz, 1H), 2.27–2.17 (m, 3H), 2.02–1.98 (m, 1H); ¹³C NMR (100 MHz, CDCl₃) δ 170.7, 169.6, 168.1, 167.9, 138.7, 135.8, 133.5, 132.0, 129.6, 129.0, 128.7, 128.4, 127.1, 126.9, 123.9, 119.6, 61.0, 51.9, 47.9, 43.4, 39.0, 29.2, 24.9; HRMS-ESI (−) m/z Calcd. for C₂₉H₂₉N₄O₄ [M−H]⁻ 497.2194, Found 497.2192.

4.2. Thermal shift assays (TSAs)

TSAs used purified covalently-crosslinked hexameric CA^{A14C/E45C/W184A/M185A} (CA121). CA121 cloned in a pET11a expression plasmid was kindly provided by Dr. Owen Pornillos (University of Virginia, Charlottesville, VA, USA). CA121 was expressed in Escherichia coli BL21(DE3)RIL and purified according to reported protocols³⁷. The TSAs were conducted as previously described55, 56, 57, with each reaction containing 7.5 μmol/L CA121 in 50 mmol/L sodium phosphate buffer (pH 8.0), 1× Sypro Orange Protein Gel Stain (Life Technologies, Carlsbad, CA, USA), and either 1% DMSO (control) or 20 μmol/L compound (1% DMSO final). The plate was heated from 25 to 95 °C with a heating rate of 0.2 °C every 10 s in the PikoReal real-time PCR system (Thermo Fisher Scientific, Waltham, MA, USA) or the QuantStudio 3 real-time PCR system (Thermo Fisher Scientific). The fluorescence intensity was measured with an Ex range of 475–500 nm and Em range of 520–590 nm. The difference in the melting temperature (ΔT_m) of CA121 in DMSO (T₀) versus in the presence of compound (T_m) were calculated using the following Eq. (1):

ΔT_m (°C)=T_m−T₀. (1)

4.3. Virus production

The wild-type laboratory HIV-1 strain, HIV-1_NL4-3⁵⁸, was produced using a pNL4-3 vector (NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH, Bethesda, MD, USA). HIV-1_NL4-3 was generated by transfecting HEK 293FT cells with 10 μg of pNL4-3 vector and FuGENE®HD Transfection Reagent (Promega, Madison, WI, USA) in a T75 flask. Supernatant was harvested 48–72 h post-transfection and transferred to MT2 cells for viral propagation. Virus was harvested upon observation of syncytia formation, typically after 3–5 days. The viral supernatant was then concentrated using 8% w/v PEG 8000 overnight at 4 °C, followed by centrifugation for 40 min at 3500 rpm (Sigma 4-16S, Sigma Laborzentrifugen, Osterode am Harz, Germany). The resulting viral-containing pellet was concentrated 10-fold by resuspension in DMEM without FBS and stored at −80 °C.

4.4. Anti-HIV-1 and cytotoxicity assays

Anti-HIV-1 activity of PF74 and related analogs was examined in TZM-GFP cells. The potency of HIV-1 inhibition was determined based on the inhibition of viral LTR-activated GFP expression in the presence of compounds compared to DMSO controls. Briefly, TZM-GFP cells were plated at density of 1 × 10⁴ cells per well in a 96-well plate. After 24 h, media was replaced with increasing concentrations of compound. Cells were exposed to HIV-1_NL4-3 (MOI = 1) 24 h post treatment. After 48 h incubation, anti-HIV-1 activity was determined by counting the amount of GFP positive cells on a Cytation™ 5 Imaging Reader (BioTek, Winooski, VT, USA) and 50% effective concentration (EC₅₀) values were determined.

The cytotoxicity of each compound was also determined in TZM-GFP cells. Cells plated at a density of 1 × 10⁴ cells per well in a 96-well plate were continuously exposed to increasing concentrations compounds over a period of 72 h. The number of viable cells in each well was determined using an XTT Cell Proliferation Kit (R&D Systems, Inc., Minneapolis, MN, USA), and 50% cytotoxicity concentration (CC₅₀) values were determined. All cell-based assays were conducted in duplicate and in at least two independent experiments.

To obtain EC₅₀ and CC₅₀ dose response curves, values were plotted in GraphPad Prism 5 and analyzed with the log_inhibitor vs. normalized response–variable slope equation as Eq. (2):

Y = 100/(1+10^{(((LogIC₅₀)‒X)×HillSlope)}) (2)

where Y is the percent response, X is the logarithm of the concentration (μmol/L) of a compound, IC₅₀ is the concentration (μmol/L) of the compound that gives a response half-way between the top and bottom of the curve, and HillSlope describes the steepness of the curve. Final values were calculated for each independent assay and average values for all assays were calculated. Statistical analysis (calculation of standard deviation) was performed using Microsoft Excel.

4.5. Microsomal stability assay

The in vitro microsomal stability assay was performed using commercially available human and mouse liver microsomes (Sekisui XenoTech, Kansas City, KS, USA). Briefly, the test compound (1 μmol/L final concentration) was pre-incubated with or without 0.5 μmol/L cobicistat (medchemexpress.com, Monmouth Junction, NJ, USA) in a 0.1 mol/L potassium phosphate buffer solution (pH 7.4, 37 °C) containing liver microsomal protein (0.5 mg/mL final concentration) and MgCl₂ (1 mmol/L final concentration). After a 15 min pre-incubation, the enzyme cofactor nicotinamide adenine dinucleotide phosphate (NADPH, 1 mmol/L final concentration) was added to initiate the reaction. All reactions were carried out in duplicate and a negative control without NADPH was performed in parallel to assess any chemical instability or non-NADPH dependent enzymatic degradation. A 50 μL of reaction mixture was taken at different time points (0, 5, 15, 30, and 60 min), and quenched with 150 μL of acetonitrile containing an appropriate internal standard and 0.1% formic acid. The samples were then vortexed and centrifuged at 15,000 rpm (Eppendorf centrifuge 5424R, Enfield, CT, USA) for 5 min at 4 °C. The supernatants were collected and analyzed by a LC−MS/MS system consisting of an Agilent 1260 Infinity HPLC (Agilent Technologie) and an AB Sciex QTrap 5500 mass spectrometer (AB Sciex LLC., Toronto, Canada) to determine the in vitro metabolic half-life (t_1/2).

4.6. Molecular modeling

PF74-bound full length native HIV-1 CA (PDB ID: 4XFZ)³⁶ was utilized to perform molecular modeling using the Schrödinger small molecule drug discovery suite 2019-1⁵⁹ (Schrödinger Inc., New York, NY, USA). Maestro⁶⁰ (Schrödinger Inc.) was employed to analyze the above crystal structure. A standard docking protocol entailed preparation of the protein of interest, grid generation, ligand preparation, and docking as key steps. Preparation of the protein involved refinement of the crystal structure using the protein preparation wizard⁶¹ (Schrödinger Inc.), in which the missing hydrogen atoms, side chains, and loops were added using prime, followed by minimization using the OPLS 3e force field⁶² to optimize the hydrogen bonding network and converge the heavy atoms to an rmsd of 0.3 Å. The binding site around the native ligand PF74, covering all the key residues within 12 Å, was defined using the receptor grid generation tool in Maestro (Schrödinger Inc.). After drawing compounds in Maestro, different conformers of those compounds were generated using LigPrep⁶³ at pH of 7±2 to serve as input for docking process. Finally, Glide XP⁶⁴(Glide, version 8.2, New York, NY, USA) was used to perform the docking with the van der Waals radii of nonpolar atoms for each of the ligands scaled by a factor of 0.8. Protein flexibility was accounted by post docking refinement and minimization under implicit solvent. Numbering of residues of HIV-1 CA, used in the discussion and the figures, was based on the full length native HIV-1 CA.

Acknowledgments

This research was supported by the National Institute of Allergy and Infectious Diseases, the National Institutes of Health, USA, grant number R01AI120860 (to Stefan G. Sarafianos and Zhengqiang Wang). We thank the Minnesota Supercomputing Institute (Minneapolis, MN, USA) for molecular modeling resources.

Author contributions

Stefan G. Sarafianos and Zhengqiang Wang conceptualized the research. Lei Wang and Sanjeev Kumar V. Vernekar designed, synthesized and characterized all compounds. Rajkumar Lalji Sahani conducted molecular docking. Jiashu Xie performed the metabolic stability assays. Mary C. Casey, Karen A. Kirby, Haijuan Du, Huanchun Zhang, and Philip R. Tedbury performed TSA, antiviral and cytotoxicity assays. Lei Wang and Zhengqiang Wang wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article: