Systematic Structure–Activity Relationship Investigation on Novel 5‑Alkyl-6-benzyl-2-{[(methylthio/phenylthio)alkyl]thio}-pyrimidin-4(3H)‑ones Endowed with Potent Anti-HIV‑1 Activity

Abstract

Four series of 5-alkyl-6-benzyl-2-[(methylthio/phenylthio)­alkyl]­thiopyrimidin-4­(3H)-ones were systematically diversified at C2, C5, and C6 and evaluated as anti-HIV-1 agents. In MT-4 cells, these compounds displayed low-to-moderate cytotoxicity and subnanomolar-to-submicromolar inhibition of WT HIV-1, with potency governed by a balanced SAR. Ortho-dihalogenated C6-benzyl groups, particularly 2,6-F₂ and 2,6-Cl₂, were generally favored. In the 2,6-dihalobenzyl series, optimal activity required C5 methyl combined with para-substituted phenylthiomethyl groups at C2, whereas 3,5-Me₂-benzyl analogues benefited from C5 isopropyl paired with methylthioethyl (24l) or unsubstituted phenylthiomethyl (24m) C2 side chains. The p-methoxyphenyl­thiomethyl thiothymine derivatives 21e (2,6-F₂) and 23e (2,6-Cl₂) emerged as lead inhibitors, showing high selectivity, potent activity against WT and clinically relevant HIV-1 variants, superior efficacy compared with nevirapine and efavirenz, dapivirine-like potency, and improved resistance tolerance relative to phenethyl-S-DABO bioisosteres. Together, biological and QSAR and docking data contribute to defining optimization strategies aimed at achieving durable anti-HIV-1 activity within the investigated chemotype and support further development of next-generation NNRTIs with improved efficacy against clinically relevant HIV-1 strains.

Introduction

The 3,4-dihydro-2-alkoxy-6-benzyl-4-oxopyrimidines (DABOs), , originally described by our group as HIV-1 NNRTIs structurally related to 1-[(2-hydroxyethoxy)­methyl]-6-(phenylthio)­thymine (HEPT), have long served as a platform for antiviral optimisation. Extensive SAR and molecular modeling studies identified the C2, C5, and C6 positions of the pyrimidin-4­(3H)-one scaffold as the main determinants for anti-HIV-1 activity. −

At the C2 position, appropriately sized lipophilic substituents are required for optimal anti-HIV-1 activity. ,, In particular, alkoxy (DABOs, 1), , alkylthio (S-DABOs, 2), ,− alkylamino (NH-DABOs, 3), , dialkylamino/aza­(hetero)­cycloalkyl (N,N-DABOs, 4) − or cycloalkyl (C-DABOs, 5), side chains consistently confer the highest HIV-1 inhibitory activity.

At the C6 position, the 2,6-difluorobenzyl moiety is generally preferred, as exemplified by F₂-S-DABOs (6), ,− , conformationally restricted S-DABOs (7), ,,,− F₂-NH-DABOs (3), , F₂-N,N-DABOs (4), − and DAPY-DABO hybrids (8). , However, this trend is context-dependent: among conformationally restricted (cyclo)­alkyl-S-DABOs certain 6-[(2-Cl-6-F-phenyl)­ethyl]-2-(cycloalkylthio)-5-methylpyrimidin-4­(3H)-ones (9) are more potent than the corresponding 2,6-F₂-phenyl-substituted (10) counterparts (Figure ), while among conformationally flexible phenylalkyl-S-DABOs there are some 2,6-Cl₂ derivatives (11) that are more potent against both WT and mutant HIV-1 strains than their 2,6-F₂ counterparts (12) (Figure ).

1.

Most relevant DABO family compound series (1–10) disclosed over the years.

2.

Background and rational design of the MTM-, MTE-, and PTM-S-DABOs (21–24) object of the present study.

Optimization at the C5 position confirmed that a methyl group is generally optimal in most DABO series, whereas pyrimidin-4­(3H)-one, 5-ethylpyrimidin-4­(3H)-one, and 5-isopropylpyrimidin-4­(3H)-one analogues typically exhibit reduced potency compared to the corresponding thymine congeners (Figure ). ,,,,−

SAR studies on C2 have generated several series of S-DABO analogues with potent anti-HIV-1 activity. ,,,,,,− Among them, 2-methylthiomethyl-S-DABOs (MTM-S-DABOs, 13, Figure ), ,,, bearing a methylthiomethyl (MTM) group on the exocyclic sulfur atom of the pyrimidine ring, and 2-oxophenethyl-S-DABOs (14, Figure ) displayed a hybrid SAR profile combining features of HEPTs and (cyclo)­alkyl-S-DABOs. ,,, Like HEPTs, their potency often increased upon introduction of bulky lower alkyl substituents at C5 (i-Pr > Et > Me > H), ,,, but, at least for MTM-S-DABOs, declined with allyl, benzyl or phenyl groups at this position. Like (cyclo)­alkyl-S-DABOs, they still retained modest activity with a C5 hydrogen, , and were sensitive to the C6 benzyl substitution, ,, performing best in terms of potency and selectivity in the presence of a 2,6-dihalo substitution. Indeed, MTM-S-DABOs were inactive with C6 benzhydryl or aryl moieties, while the 2,6-F₂ substitution on the C6 benzyl group and, in some cases, the 2-Cl-6-F one, markedly increased potency. Among 2-oxophenethyl-S-DABOs (14), ,− the C5 ethyl/iso-propyl derivatives bearing a 2,6-dichlorobenzyl/2-chloro-6-fluorobenzyl moiety at C6 gave the highest broad-spectrum anti-HIV-1 activity and the lowest fold-resistance ratio. The 2-chloro-6-fluoro substitution was likewise optimal in the related phenylaminocarbonylmethyl-S-DABOs (16). , The highly potent 2-phenylethyl-S-DABOs (11, 12) of Botta and co-workers , inspired numerous bioisosteric/regioisomeric analogues, including 2-oxophenethyl-S-DABOs (14), ,− 2-oxophenethyl-S-DACOs (15), − phenylaminocarbonylmethyl-S-DABOs (16), , phenylpyrazol-S-DACOs (17), and 2-hydroxyphenethyl-S-DACOs (18), many of which inhibited HIV-1 replication at low nanomolar concentrations with low cytotoxicity (Figure ), confirming that modulation of the C2 sulfur-linked side chain strongly affects antiviral activity and selectivity.

Previous work on dihydro-alkyl-thio-naphthylmethyl-oxopyrimidine (DATNO) thiouracil derivatives (19), showed that homologation from S-[(methylthio)­methyl] to S-[(ethylthio)­methyl] enhanced anti-HIV-1 activity, while 6-(thiophen-2-ylmethyl)-2-thiothymines (20) showed comparable potency for S-[(methylthio)­methyl] and S-[(phenylthio)­methyl] groups, with twofold higher activity for S-[(methylthio)­ethyl] analogues. However, these compounds remained less active than NNRTIs such as nevirapine (NVP), likely because replacing the C6 benzyl group with thiophen-2-ylmethyl reduces the activity as observed with (cyclo)­alkyl-S-DABOs. ,,

To the best of our knowledge, comprehensive SAR analyses on MTM-S-DABO analogues addressing the combined influence of C2-substituted thioether side chains, C5 alkyl substitution, and C6 arylmethyl patterns have not been reported yet. Moreover, biological evaluation against clinically relevant HIV-1 mutant strains and mutated RT remains scarce, representing a significant gap in the current understanding of this privileged S-DABO series.

We therefore carried out a systematic SAR study, against both WT and mutant HIV-1 strains and RTs, of MTM-S-DABOs (21a,f,k, 22a,f,k, 23a,f,k, 24a,f,k) and new related derivatives (21b–e,g–j,l–o, 22b–e,g–j,l–o, 23b–e,g–j,l–o, 24b–e,g–j,l–o, Figure ), with modifications at several positions of the pyrimidinone core. At C2, alongside the parent MTM group, we introduced the superior homologue (methylthio)­ethyl substituent (MTE-S-DABOs) as well as (un)­substituted (phenylthio)­methyl (PTM-S-DABOs) side chains on the exocyclic sulfur atom to probe steric, conformational, and electronic effects. The PTM chains were designed as bioisosteres of both the 2-oxophenethyl and the 2-phenylethyl , side chains and were decorated with p-F and p-OMe groups (Figure ), that are known to confer high anti-HIV-1 potency across these S-DABOs. ,, At the C6, we retained the highly effective benzyl moiety and included the most favorable substitution patterns, namely 2,6-F₂, 2-Cl-6-F, and 2,6-Cl₂. ,,,,, The 3,5-Me₂-benzyl substitution was also comprised for comparative purposes, given its established contribution to antiviral activity both in HEPT derivatives and in selected 2-oxophenethyl-S-DABOs. Finally, C5 was varied with Me, Et, and i-Pr to modulate the stereoelectronic effects within the non-nucleoside binding site (NNBS) (Figure ).

Chemistry

Final compounds 21a, b–e,f, g–j,k, l–o, 22a–o, 23a–o, and 24a–j,k, l–o were prepared as outlined in Scheme . All 5-alkyl-6-(arylmethyl)-2-thiouracils intermediates 25a, b, c, d–h, i, j, k, l were prepared according to previously reported procedures. Subsequent S-alkylation of intermediates 25a–l with the proper alkyl chlorides afforded the final compounds 21-24 via two alternative routes, depending on the nature of the chloride used. For the chloromethyl methyl sulfide, the appropriate 5-alkyl-6-(arylmethyl)-2-thiouracils (25a–l) were reacted with a two-fold molar excess of the alkylating agent in dry DMF at room temperature providing the final compounds 21a, f, k, 22a,f,k, 23a,f,k, and 24a,f,k. In contrast, chloroethyl methyl sulfide and (un)­substituted chloromethyl phenyl sulfides were reacted with the thiouracils 25a–l in an ethanolic potassium hydroxide solution, first at room temperature and then under reflux, providing the final compounds 21b–e,g–j,l–o, 22b–e,g–j,l–o, 23b–e,g–j,l–o and 24b–e,g–j,l–o. All alkylating agents were commercially available except for the chloromethyl p-fluorophenyl sulfide and the chloromethyl p-methoxyphenyl sulfide, which were prepared according to the published Böhme reaction procedures. Chemical and physical characterization data of the newly-synthesized final compounds 21b–e,g–j,l–o, 22a–o, 23a–o, and 24a–j,l–o, together with the corresponding elemental analyses, are provided in Tables S1 and S2, respectively (Supporting Information Part 1).

1. Synthesis of Compounds 21–24 .

^a Reagents and conditions: (a) chloromethyl methyl sulfide, dry DMF, rt; (b) potassium hydroxide, 95% ethanol, chloromethyl ethyl sulfide or properly (un)­substituted chloromethyl phenyl sulfide, rt then reflux.

Results and Discussion

Cytotoxicity and Wild Type HIV-1 Inhibiting Activity

The four series of 5-substituted 6-benzyl-2-phenyl/methylthioalkyl-S-DABO derivatives 21a–o (2,6-F₂), 22a–o (2-Cl-6-F), 23a–o (2,6-Cl₂), and 24a–o (3,5-Me₂) were tested to evaluate their cytotoxicity (CC₅₀) and capability to inhibit the WT HIV-1-induced cytopathic effect in lymphoid MT-4 cells (EC₅₀; HIV-1 strain, NL4-3). Nevirapine (NVP), efavirenz (EFV), and the “second generation” NNRTI dapivirine were added as reference drugs. Since the new PTM-S-DABOs (21c–e,h–j,m–o, 22c–e,h–j,m–o, 23c–e,h–j,m–o, and 24c–e,h–j,m–o) can be considered bioisosteric analogues of the 2-(p-methoxyphenyl)­ethyl-S-DABOs 11 ¹³ and 12,¹⁴ previously reported data under the same experimental conditions for those compounds were also included for comparison (Table ).

1. Cytotoxicity and Anti-HIV-1 Activity (WT HIV-1 Strain NL4-3) of S-DABOs 21–24 .

Compd	R	R1	X	EC50, μM	CC50, μM	Selectivity Index
21a	Me	2,6-F2	CH2SMe	0.006 ±0.0002	37±2	6183
21b	Me	2,6-F2	(CH2)2SMe	0.0029±0.0001	67±2	23161
21c	Me	2,6-F2	CH2SPh	0.004±0.0002	>64±3	>16000
21d	Me	2,6-F2	CH2S(4-FPh)	0.002±0.0001	>61±2	>61200
21e	Me	2,6-F2	CH2S(4-MeOPh)	0.0008±0.00005	>59±3	>74250
21f	Et	2,6-F2	CH2SMe	0.002±0.00007	38±1	19250
21g	Et	2,6-F2	(CH2)2SMe	0.006±0.0003	>70±4	>70100
21h	Et	2,6-F2	CH2SPh	0.005±0.0003	>62±3	>12360
21i	Et	2,6-F2	CH2S(4-FPh)	0.009±0.0005	>59±2	>6578
21j	Et	2,6-F2	CH2S(4-MeOPh)	0.003±0.0001	>57±2	>19167
21k	i-Pr	2,6-F2	CH2SMe	0.0009±0.00004	54±1	60111
21l	i-Pr	2,6-F2	(CH2)2SMe	0.007±0.0003	>67±3	>9643
21m	i-Pr	2,6-F2	CH2SPh	0.012±0.0006	>59±1	>4975
21n	i-Pr	2,6-F2	CH2S(4-FPh)	0.025±0.0010	>57±1	>2292
21o	i-Pr	2,6-F2	CH2S(4-MeOPh)	0.033±0.001	>55±1	>1688
22a	Me	2-Cl-6-F	CH2SMe	0.029±0.002	>72±3	>2500
22b	Me	2-Cl-6-F	(CH2)2SMe	0.018±0.0007	>70±3	>3872
22c	Me	2-Cl-6-F	CH2SPh	0.020±0.001	>61±2	>3070
22d	Me	2-Cl-6-F	CH2S(4-FPh)	0.009±0.0003	>59±2	>6533
22e	Me	2-Cl-6-F	CH2S(4-MeOPh)	0.004±0.0002	>57±2	>19067
22f	Et	2-Cl-6-F	CH2SMe	0.020±0.001	>70±4	>3485
22g	Et	2-Cl-6-F	(CH2)2SMe	0.019±0.001	>67±3	>3526
22h	Et	2-Cl-6-F	CH2SPh	0.406±0.02	>59±4	>146
22i	Et	2-Cl-6-F	CH2S(4-FPh)	0.729±0.03	>57±3	>78
22j	Et	2-Cl-6-F	CH2S(4-MeOPh)	0.510±0.02	>55±2	>109
22k	i-Pr	2-Cl-6-F	CH2SMe	0.010±0.0004	>47±2	>4700
22l	i-Pr	2-Cl-6-F	(CH2)2SMe	0.594±0.02	>64±3	>109
22m	i-Pr	2-Cl-6-F	CH2SPh	0.874±0.04	>57±3	>66
22n	i-Pr	2-Cl-6-F	CH2S(4-FPh)	1.01±0.05	>55±2	>55
22o	i-Pr	2-Cl-6-F	CH2S(4-MeOPh)	0.743±0.03	>54±2	>72
23a	Me	2,6-Cl2	CH2SMe	0.111±0.005	>69±4	>623
23b	Me	2,6-Cl2	(CH2)2SMe	0.079±0.003	>66±3	>843
23c	Me	2,6-Cl2	CH2SPh	0.015±0.0006	>59±3	>3933
23d	Me	2,6-Cl2	CH2S(4-FPh)	0.005±0.0002	>57±3	>11320
23e	Me	2,6-Cl2	CH2S(4-MeOPh)	0.002±0.0001	>55±2	>27550
23f	Et	2,6-Cl2	CH2SMe	0.190±0.011	>67±2	>392
23g	Et	2,6-Cl2	(CH2)2SMe	0.565±0.022	>64±3	>114
23h	Et	2,6-Cl2	CH2SPh	0.594±0.025	30±2	51
23i	Et	2,6-Cl2	CH2S(4-FPh)	0.329±0.029	>55±2	167
23j	Et	2,6-Cl2	CH2S(4-MeOPh)	0.207±0.012	>53±3	>258
23k	i-Pr	2,6-Cl2	CH2SMe	0.283±0.007	>64±3	>227
23l	i-Pr	2,6-Cl2	(CH2)2SMe	0.445±0.02	>62±3	>139
23m	i-Pr	2,6-Cl2	CH2SPh	0.452±0.02	>55±3	>126
23n	i-Pr	2,6-Cl2	CH2S(4-FPh)	0.831±0.04	>53±2	>64
23o	i-Pr	2,6-Cl2	CH2S(4-MeOPh)	0.561±0.03	>52±2	>93
24a	Me	3,5-Me2	CH2SMe	1.84±0.11	>78±4	>42
24b	Me	3,5-Me2	(CH2)2SMe	0.209±0.01	>75±4	>357
24c	Me	3,5-Me2	CH2SPh	0.736±0.03	>65±2	>89
24d	Me	3,5-Me2	CH2S(4-FPh)	3.2±0.20	>62±3	>20
24e	Me	3,5-Me2	CH2S(4-MeOPh)	4.1±0.20	>61±3	>15
24f	Et	3,5-Me2	CH2SMe	0.778±0.03	>75±3	>96
24g	Et	3,5-Me2	(CH2)2SMe	0.180±0.01	>72±3	>179
24h	Et	3,5-Me2	CH2SPh	0.380±0.02	>63±2	>166
24i	Et	3,5-Me2	CH2S(4-FPh)	0.705±0.03	>59±3	>83
24j	Et	3,5-Me2	CH2S(4-MeOPh)	0.720±0.03	>60±2	>83
24k	i-Pr	3,5-Me2	CH2SMe	0.115±0.006	>72±3	>623
24l	i-Pr	3,5-Me2	(CH2)2SMe	0.027±0.001	>69±4	>2554
24m	i-Pr	3,5-Me2	CH2SPh	0.049±0.002	>61±2	>1242
24n	i-Pr	3,5-Me2	CH2S(4-FPh)	0.056±0.002	27±1	474
24o	i-Pr	3,5-Me2	CH2S(4-MeOPh)	0.082±0.003	33±1	396
11	Me	2,6-Cl2	CH2CH2(4-MeOPh)	<0.00014±0.000007	>57±3	>407142
12	Me	2,6-F2	CH2CH2(4-MeOPh)	0.0049±0.0003	>62±3	>12694
NVP				0.12±0.006	>7	>58
EFV				0.007±0.0003	>3	>429
Dapivirine				0.0006±0.00003	3	5000

^aValues are means ± SD determined from at least two independent experiments, each one done in triplicate.

^bEffective concentration 50, concentration needed to inhibit by 50% the HIV-1-induced cytopathic effect, evaluated with the MTT method in MT-4 cells (HIV-1 strain: NL4-3).

^cCytotoxic concentration 50, concentration needed to induce 50% death of noninfected cells, evaluated with the MTT method in MT-4 cells.

^dSelectivity index (SI), CC₅₀/EC₅₀.

^eHigher concentrations could not be tested due to precipitation of compounds in the culture medium.

All tested compounds 21-24 exhibited low-to-mild cytotoxicity, with CC₅₀ values ranging from 27 μM to >78 μM and selectivity indexes (SI = CC₅₀/EC₅₀) between >15 and >74250 (Table ). Most compounds showed good-to-excellent activity against WT HIV-1 with EC₅₀ values ranging from sub-nanomolar to sub-micromolar in MT-4 infected cells, except for 3,5-Me₂-benzyl compounds 24a,d,e which displayed single-digit micromolar values.

Evaluation of compounds 21–24 against WT HIV-1 NL4-3 strain revealed a highly substitution-dependent antiviral profile, in which the nature of the C6 benzyl moiety, the C5 alkyl substituent, and the C2 thioether side chain sensibly influenced anti-HIV-1 potency and selectivity. Indeed, overall data showed that for any given substitution pattern at both C2 and C5 positions the highest anti-HIV-1 activity was displayed by the compounds of the C6 2,6-F₂-benzyl series 21a–o, while the rank order of potency across the remaining series varied depending on the profile of substitution at both C2 and C5 positions. The higher activity of the 2,6-F₂-benzyl series 21a–o is consistent with the results of the QSAR analysis, which highlighted a favorable contribution of van der Waals interactions associated with this substitution pattern (see 2D-QSAR section). More specifically, the higher potency of C6-2,6-dihalobenzyl analogues compared with their 3,5-dimethylbenzyl analogues is consistent with optimized steric and electronic complementarity within the NNBS, wherein the electron-deficient benzyl ring is expected to engage in favorable π-stacking with the electron-rich aromatic side chain lining the pocket. ,, Within the 2,6-dihalobenzyl subset, the superior activity of the 2,6-F₂ derivatives over the corresponding 2,6-Cl₂ and 2-Cl-6-F congeners further agrees with the QSAR findings and the smaller van der Waals volume of fluorine, which may allow tighter packing against lipophilic pocket boundaries. ,

The influence of the C5 substituent was likewise highly context-dependent and closely coupled to the stereoelectronic properties of the C6 benzyl fragment and the C2 thioether substituent. In the 6-(2,6-dihalobenzyl)-2-thiouracil series 21 and 22, all MTE and PTM analogues followed the classical SAR trend of (cyclo)­alkyl-S-DABO, with activity increasing in the sequence i-Pr < Et < Me, while the MTM analogues (21a,f,k and 22a,f,k, respectively) showed the already known opposite “HEPT-like” order of potency (i-Pr > Et > Me) confirming the favorable effect of bulky C5 substituents in this context (Table ). In the 2,6-Cl₂-benzyl series 23 the C5 position SAR trend was i-Pr < Et < Me with the MTM, MTE, and PTM derivatives, with the only partial exception of the C5 ethyl derivatives 23g and 23h that resulted less potent than the corresponding C5 isopropyl counterparts 23l and 23m, respectively (Table ). In contrast, the 3,5-Me₂-benzyl series (24a–o) consistently exhibited a “HEPT-like” trend, irrespective of the substitution at the C2 position, in which potency progressively increased with the size of the C5 substituents from 2-thiothymines (24a–e) derivatives to the corresponding 5-isopropyl-2-thiouracils (24k–o) (Table ).

In this context, the activity trend reflects a balance between steric volume and binding orientation. C5-Me is optimal in the 2,6-dihalobenzyl/PTM congeners, in which the combined steric bulk of the C6 aryl and extended C2 thioether approaches the cavity limit, whereas larger alkyls incur localized crowding. ,, When C2 is occupied by the smaller MTM group, or when the C6 aryl ring is a 3,5-Me₂-benzyl motif, the C5-adjacent subpocket is underfilled, and i-Pr restores van der Waals contact without prohibitive penalty, ,,, consistent with the observed “HEPT-like” behavior. ,

The exocyclic C2 thioether substituent influenced the antiviral potency in a context-dependent manner. For any given pattern of C5/C6 substitution, the extension by one methylene unit of the C2 side chain of MTM derivatives (21a,f,k, 22a,f,k, 23a,f,k, and 24a,f,k) to afford the closest superior homologous MTE S-DABO counterparts (21b,g,l, 22 b,g,l, 23 b,g,l, and 24b,g,l), resulted in a slight increase of activity, with the only exceptions of the 5-ethyl/isopropyl 6-(2,6-dihalobenzyl)-2-thiouracil derivatives (21l, 22l, and 23g,l), and in a markedly different SAR profile. In contrast, the substitution of the MTM side chain at C2 with phenylthiomethyl groups provided PTM-S-DABO derivatives that, if substituted with a MeO group or F atom at para position of the phenyl ring, displayed the highest potency and selectivity in the 5-methyl 6-(2,6-dihalobenzyl)-2-thiouracil series 21c,d,e, 22c,d,e, and 23c,d,e. Indeed, among the 2-thiothymine analogues of the 2,6-F₂- and 2-Cl-6-F-benzyl series (21a–e and 22a–e) activity decreased in the order pMeOPhSCH₂ > pFPhSCH₂ > MTE > PhSCH₂ > MTM, while among the 2-thiothymines of the 2,6-Cl₂-benzyl series (23a–e) the order was pMeOPhSCH₂ > pFPhSCH₂ > PhSCH₂ > MTE > MTM, and considering the only PTM moiety substitution profile the trend of potency was always pMeOPhSCH₂ > pFPhSCH₂ > PhSCH₂. In contrast, within the 6-(3,5-Me₂-benzyl)-2-thiouracils series (24), anti-HIV-1 activity decreased in the order MTE > PhSCH₂ > pFPhSCH₂ > pMeOPhSCH₂ > MTM for any given substitution at C5. Notably, within the PTM subset, the potency trend was reversed compared with the 2,6-dihalobenzyl series, following the order PhSCH₂ > p-FPhSCH₂ > p-MeOPhSCH₂.

In this regard, the modest potency gain from MTM to MTE reflects improved side-chain accommodation within the lipophilic NNBS tunnel, whereas the larger gain by PTM is consistent with additional aromatic interactions. In the 2,6-dihalobenzyl series, the p-MeO-PTM moiety markedly enhances potency through a synergistic combination of increased π-electron density, a critical hydrogen-bond acceptor site at the methoxy oxygen, and enhanced local Me lipophilicity, ,, in accordance with the MFTA-QSAR analysis (see below). The p-fluoro analogue supports intermediate activity by retaining aromatic and hydrophobic character while lacking the MeO hydrogen-bond acceptor and resonance donation. In the 3,5-Me₂ series, altered scaffold geometry repositions the C2 vector, penalizing bulky para-substituted PTMs via localized steric repulsion and shifting preference toward compact MTE or unsubstituted PTM chains. ,,

Overall, the data showed that optimal anti-HIV-1 activity requires precise balancing of steric volume and electronic properties across all three substitution vectors. Except for the previously reported MTM-S-DABOs 21a, f, k, the most potent among the new compounds were all p-MeO- and, to a lesser extent, p-F-substituted PTM-S-DABOs belonging to the 6-(2,6-dihalobenzyl)-2-thiothymine series (21d,e, 22d,e, and 23d,e), with the greatest potency displayed by the 2,6-F₂-benzyl and, to a lesser extent, 2,6-Cl₂-benzyl p-MeO PTM-S-DABO derivatives 21e and 23e, respectively. Compared with the reference drugs, 21e and 23e exhibited subnanomolar to low single-digit nanomolar activity against WT HIV-1 (EC₅₀ = 0.8–2 nM), outperforming both NVP and EFV in terms of potency, cytotoxicity, and selectivity, and showing activity substantially comparable to that of dapivirine. Interestingly, the 2,6-F₂-benzyl derivative 21e was more than 6-fold more potent than its bioisosteric counterpart 12,¹⁴ whereas the 2,6-Cl₂-benzyl analogue 23e was more than 10-fold less active than the corresponding bioisostere 11 ¹³ (Table ).

The 3,5-Me₂-benzyl series (24) was generally the least potent and selective. However, 24l and 24m, which bear an isopropyl group at the C5 position and an MTE or PTM side chain at C2, respectively, partially restored anti-HIV-1 activity, with potency in the low double-digit nanomolar range. This activity was superior to that of NVP, but lower than those of both EFV and dapivirine (Table ).

Inhibitory Activity against Clinically Relevant HIV-1 Mutant Strains

To evaluate their resistance profile, MTM-, MTE-, and PTM-S-DABOs 21–24 were then tested against a panel of clinically relevant HIV-1 mutant strains (Y181C, Y188L, and K103N). Selected compounds (21d,e, 22d,e, 23d,e, and 24l,m), exhibiting a favorable activity against WT HIV-1, were also assayed for their efficacy against the HIV-1 clinical isolate IRLL98 and the double-mutant K103N+Y181C. Finally, the performance of 21–24 was compared to that of NVP, EFV, dapivirine, and the reference phenethyl S-DABOs 11 ¹³ and 12 ¹⁴ (Table ).

2. Anti-HIV-1 Activity of 21-24 against Clinically Relevant HIV-1 Mutant Strains .

				EC50, μM (fold resistance)
Compd	R	R1	X	IRLL98	Y181C	Y188L	K103N	K103N+Y181C
21a	Me	2,6-F2	CH2SMe	ND	8±0.4(1283)	>37±2	6.5±0.3 (1083)	ND
21b	Me	2,6-F2	(CH2)2SMe	ND	2±0.1 (827)	9.5±0.4 (3276)	0.42±0.02 (145)	ND
21c	Me	2,6-F2	CH2SPh	ND	1±0.05 (275)	3±0.1 (775)	0.1±0.005 (26)	ND
21d	Me	2,6-F2	CH2S(4-FPh)	0.008±0.0004 (4)	0.75±0.03 (375)	2±0.1 (1050)	0.05±0.002 (27)	0.15±0.006 (75)
21e	Me	2,6-F2	CH2S(4-MeOPh)	0.003±0.0002 (4)	0.54±0.02 (675)	1±0.05 (1375)	0.009±0.0004 (11)	0.062±0.003 (77)
21f	Et	2,6-F2	CH2SMe	ND	2.5±0.1(1250)	>38±3	2.2 ±0.1(1100)	ND
21g	Et	2,6-F2	(CH2)2SMe	ND	3±0.1 (517)	10±0.6 (1717)	0.9±0.04 (150)	ND
21h	Et	2,6-F2	CH2SPh	ND	>62±2	>62±3	>62±3	ND
21i	Et	2,6-F2	CH2S(4-FPh)	ND	2±0.1 (2322)	>59±3	15±1 (1667)	ND
21j	Et	2,6-F2	CH2S(4-MeOPh)	ND	1.5±0.07 (500)	>57±2	2.7±0.1 (900)	ND
21k	i-Pr	2,6-F2	CH2SMe	ND	2±0.1(2333)	28±2(31444)	2±0.1(2111)	ND
21l	i-Pr	2,6-F2	(CH2)2SMe	ND	6±0.3 (800)	>68±4	37±2 (5300)	ND
21m	i-Pr	2,6-F2	CH2SPh	ND	>60±3	>60±4	>60±3	ND
21n	i-Pr	2,6-F2	CH2S(4-FPh)	ND	2.6±0.1 (104)	>57±2	>57±3	ND
21o	i-Pr	2,6-F2	CH2S(4-MeOPh)	ND	4±0.2 (124)	>58±3	5.8 ±0.3 (176)	ND
22a	Me	2-Cl-6-F	CH2SMe	ND	6.5±0.3 (224)	>73±4	3.8±0.2 (131)	ND
22b	Me	2-Cl-6-F	(CH2)2SMe	ND	5±0.3 (289)	40±3 (2222)	2.8±0.2 (155)	ND
22c	Me	2-Cl-6-F	CH2SPh	ND	9.5±0.5 (475)	13±0.6 (665)	2.2±0.1 (110)	ND
22d	Me	2-Cl-6-F	CH2S(4-FPh)	0.027±0.0002 (3)	1.2±0.04 (133)	5±0.3 (589)	0.095±0.004 (10)	0.25±0.01 (28)
22e	Me	2-Cl-6-F	CH2S(4-MeOPh)	0.008±0.0004 (2)	0.75±0.04 (187)	2.2±0.1 (550)	0.025±0.001 (6)	0.075±0.003 (19)
22f	Et	2-Cl-6-F	CH2SMe	ND	4.4±0.2 (220)	>70±4	3.2±0.1 (160)	ND
22g	Et	2-Cl-6-F	(CH2)2SMe	ND	3.8±0.2 (200)	38±3 (1995)	2±0.1 (111)	ND
22h	Et	2-Cl-6-F	CH2SPh	ND	12±0.6 (30)	>59±4	>59±3	ND
22i	Et	2-Cl-6-F	CH2S(4-FPh)	ND	8±0.4 (11)	18±1 (25)	>57±3	ND
22j	Et	2-Cl-6-F	CH2S(4-MeOPh)	ND	3.0±0.2 (6)	15±2	>55±3	ND
22k	i-Pr	2-Cl-6-F	CH2SMe	ND	3±0.1 (330)	>47±3	2.5±0.1 (250)	ND
22l	i-Pr	2-Cl-6-F	(CH2)2SMe	ND	9±0.4 (15)	47±2 (79)	>65±3	ND
22m	i-Pr	2-Cl-6-F	CH2SPh	ND	>58±3	>58±2	>58±3	ND
22n	i-Pr	2-Cl-6-F	CH2S(4-FPh)	ND	14±0.3 (14)	19±2 (19)	>65±3	ND
22o	i-Pr	2-Cl-6-F	CH2S(4-MeOPh)	ND	4±0.2 (6)	2.6±0.1 (3)	>54±3	ND
23a	Me	2,6-Cl2	CH2SMe	ND	>69±4	>69±4	>69±4	ND
23b	Me	2,6-Cl2	(CH2)2SMe	ND	7.0±0.3 (89)	34±2 (432)	48±3 (611)	ND
23c	Me	2,6-Cl2	CH2SPh	ND	1.0±0.05 (87)	3.1±0.1 (207)	0.1±0.005 (7)	ND
23d	Me	2,6-Cl2	CH2S(4-FPh)	0.017±0.0009 (3)	0.97±0.04 (195)	2.4 ±0.1 (480)	0.065 ±0.002 (13)	0.21±0.01 (42)
23e	Me	2,6-Cl2	CH2S(4-MeOPh)	0.004±0.0002 (2)	0.32±0.01 (160)	0.85±0.04 (425)	0.008±0.0004 (4)	0.042±0.002 (21)
23f	Et	2,6-Cl2	CH2SMe	ND	29±2 (151)	>67±3	>67±4	ND
23g	Et	2,6-Cl2	(CH2)2SMe	ND	2.5±0.1 (4)	4.0±0.2 (6)	28±2 (49)	ND
23h	Et	2,6-Cl2	CH2SPh	ND	5.0±0.2 (8)	17±1 (29)	>30±2	ND
23i	Et	2,6-Cl2	CH2S(4-FPh)	ND	3.5±0.2 (11)	6.1±0.3 (18)	>55±2	ND
23j	Et	2,6-Cl2	CH2S(4-MeOPh)	ND	3.2±0.2 (15)	4.1±0.4 (19)	>53±3	ND
23k	i-Pr	2,6-Cl2	CH2SMe	ND	27±2 (96)	>64±3	>64±4	ND
23l	i-Pr	2,6-Cl2	(CH2)2SMe	ND	8.0±0.4 (18)	>62±3	>62±4	ND
23m	i-Pr	2,6-Cl2	CH2SPh	ND	10±0.5 (23)	>55±2	>55±3	ND
23n	i-Pr	2,6-Cl2	CH2S(4-FPh)	ND	4.6±0.2 (6)	16±0.6 (19)	>53±3	ND
23o	i-Pr	2,6-Cl2	CH2S(4-MeOPh)	ND	4.5±0.2 (8)	8±0.4 (14)	>52±3	ND
24a	Me	3,5-Me2	CH2SMe	ND	12±0.7 (6)	>74±4	64±3 (35)	ND
24b	Me	3,5-Me2	(CH2)2SMe	ND	2.3±0.1 (12)	57±3 (273)	58±2 (278)	ND
24c	Me	3,5-Me2	CH2SPh	ND	10±0.5 (14)	28±2 (38)	>65±3	ND
24d	Me	3,5-Me2	CH2S(4-FPh)	ND	>62±3	31±2 (10)	>62±4	ND
24e	Me	3,5-Me2	CH2S(4-MeOPh)	ND	>61±2	>61±3	>61±3	ND
24f	Et	3,5-Me2	CH2SMe	ND	25±2 (51)	>75±3	>75±4	ND
24g	Et	3,5-Me2	(CH2)2SMe	ND	3±0.1 (17)	>72±4	>72±4	ND
24h	Et	3,5-Me2	CH2SPh	ND	4±0.2 (10)	18±1 (46)	>63±3	ND
24i	Et	3,5-Me2	CH2S(4-FPh)	ND	26±1 (37)	27±1 (37)	>60±3	ND
24j	Et	3,5-Me2	CH2S(4-MeOPh)	ND	>59±3	16±1 (22)	>59±3	ND
24k	i-Pr	3,5-Me2	CH2SMe	ND	1±0.05 (12)	>72±4	7±0.4 (63)	ND
24l	i-Pr	3,5-Me2	(CH2)2SMe	0.054±0.002 (2)	0.52±0.02(19)	1±0.05(39)	6.2±0.3 (230)	1.25±0.05 (46)
24m	i-Pr	3,5-Me2	CH2SPh	0.095±0.005 (2)	0.7±0.03(15)	1.3±0.05 (26)	7.3±0.4 (149)	1.35±0.054 (28)
24n	i-Pr	3,5-Me2	CH2S(4-FPh)	ND	1±0.04 (18)	2.3±0.1 (40)	9.5±0.6 (170)	ND
24o	i-Pr	3,5-Me2	CH2S(4-MeOPh)	ND	2.4±0.1 (29)	2.6±0.2 (32)	9.9±0.5 (121)	ND
11	Me	2,6-Cl2	CH2CH2(4-MeOPh)	0.00022±0.00003 (2)	2.3±0.2 (16428)	6.9±0.4 (49286)	0.87±0.05 (6214)	0.95±0.08(6787)
12	Me	2,6-F2	CH2CH2(4-MeOPh)	0.035±0.002 (7)	6.3±0.3 (1286)	8.9±0.5 (1735)	4.1±0.2 (837)	1.2±0.1(245)
NVP				>7	>7	>7	5.6±0.3 (47)	>7
EFV				0.63±0.03 (90)	0.044±0.003 (6)	1.6±0.1 (224)	0.39±0.02 (56)	0.40±0.016 (57)
Dapivirine				0.006±0.0003 (10)	0.012±0.0008 (20)	0.33±0.02 (550)	0.003±0.0001 (5)	0.054±0.002 (90)

^aValues are means ± SD determined from at least two independent experiments, each one done in triplicate.

^bEffective concentration 50, concentration needed to inhibit by 50% the HIV-1-induced cytopathic effect, evaluated with the MTT method in MT-4 cells.

^cFold resistance: ratio of EC₅₀ value against drug-resistant strain and EC₅₀ of the WT NL4-3 strain (Table ).

^dClinical isolate resistant to NRTIs (lamivudine, emtricitabine) and NNRTIs (NVP, EFV, delavirdine).

^eND, not determined.

^fHigher concentration could not be tested due to precipitation of compounds in the culture medium.

As shown in Table , compounds 21–24 mostly retained the SAR patterns regarding the effects on the antiviral activity of the substitution profile at C2, C5, and C6 positions observed against WT HIV-1 (Table ). The MTM-S-DABOs across all four series (21a,f,k, 22a,f,k, 23a,f,k, and 24a,f,k) displayed weak activity in the micromolar range against all tested mutant strains, yet demonstrating WT-like SAR trends, with two notable exceptions. The 2,6-Cl₂-benzyl MTM-S-DABOs (23a,f,k), contrary to what observed against WT HIV-1, exhibited HEPT-like behavior with the antiviral activity slightly increasing in the order Me < Et < i-Pr at C5 position, while remaining extremely weak against all mutants (Table ). In the 3,5-Me₂-benzyl series (24a,f,k), the preferred profile of C5 substitution versus the entire panel of mutants followed the order Et < Me < i-Pr, slightly differing from the trend Me < Et < i-Pr registered against WT HIV-1 (Table ). Similarly, the MTE-S-DABOs (21b,g,l, 22b,g,l, 23b,g,l, and 24b,g,l) mostly retained the SAR observed with WT HIV-1. The Me and i-Pr substitutions at C5 position of the pyrimidinone core remained the most effective against all evaluated mutant strains in the case of 2,6-F₂-benzyl (21b,g,l) and 3,5-Me₂-benzyl (24b,g,l) derivatives, respectively. In contrast, the 2-Cl-6-F- (22b,g,l) and 2,6-Cl₂-benzyl-substituted (23b,g,l) series showed an atypical pattern requiring an ethyl moiety at C5 (Et > Me > i-Pr) for the maximal activity against all mutants (Table ). Among the PTM-S-DABOs, the 6-(2,6-dihalobenzyl)-2-thiothymines (21c–e, 22c–e, and 23c–e) and the 6-(3,5-Me₂-benzyl)-5-i-Pr-2-thiouracil 24m showed the strongest activity against all mutants, aligning with the results observed with WT HIV-1. For any substitution at the para phenyl ring position the three series of 2,6-dihalobenzyl PTM-S-DABOs (21c–e,h–j,m-o, 22c–e,h–j,m-o, and 23c–e,h–j,m–o) showed the following trend of antiviral potency against all mutants Me > Et > i-Pr, while among the 3,5-Me₂benzyl PTM derivatives (24c–e,h–j,m–o) the SAR trend about the C5 substitution was the opposite (i-Pr > Et > Me), but identical to what observed against WT HIV-1 (Table ). Irrespective of the substitution at C5 position, among the 2,6-dihalobenzyl PTM-S-DABOs the trend of antiviral potency against all variants, including IRLL98 and the double mutant K103N+Y181C, was always pMeOPhSCH₂ > pFPhSCH₂ > PhSCH₂, while among 3,5-Me₂benzyl analogues (24c–e,h–j,m–o) was reverted to PhSCH₂ > pFPhSCH₂ > pMeOPhSCH₂.

Among all tested derivatives, the p-MeO- and, to a lesser extent, p-F-substituted PTM-S-DABOs belonging to the 6-(2,6-dihalobenzyl)-2-thiothymine series (21d,e, 22d,e, and 23d,e), the 5-isopropyl-3,5-Me₂-benzyl MTE derivative 24l, and its unsubstituted PTM analogue 24m had the highest antiviral potencies toward the entire panel of HIV-1 mutants and isolates, outperforming NVP against all variants, except K103N where 24l and 24m displayed comparable activity. Although always less active than EFV against Y181C (EC₅₀ = 0.32–1.2 μM vs 0.044 μM), these compounds were all superior to EFV against the clinical isolate IRLL98 (EC₅₀ = 0.003-0.095 μM vs 0.63 μM for EFV), comparable against Y188L (EC₅₀ = 0.85-5.3 μM vs 1.6 μM for EFV) and, with the exception of 24l and 24m, markedly more potent against both K103N (EC₅₀ = 0.008-0.095 μM vs 0.39 μM for EFV) and the double mutant K103N+Y181C (EC₅₀ = 0.042-0.25 μM vs 0.40 μM for EFV). Strikingly, the two most active derivatives, the 5-methyl-6-(2,6-F₂-benzyl) p-MeO PTM-S-DABO 21e and its 2,6-Cl₂-benzyl analogues 23e, with the only exception of IRLL98 where 21e outperformed 12, but 23e was worse than 11, consistently showed stronger inhibition across all HIV-1 variants than their phenethyl-S-DABO counterparts 11 and 12 suggesting a superiority of the PTM-S-DABO over the phenethyl-S-DABO in enhancing the resistance tolerance. Additionally, 21e and 23e, despite weaker than dapivirine against Y181C and Y188L, were substantially comparable against K103N and the double mutant K103N+Y181C, and slightly more active against the clinical isolate IRLL98.

Enzyme Inhibitory Activity on HIV-1 WT and NNRTI-Resistant Mutated RT

The MTM-, MTE-, and PTM-S-DABOs 21–24, along with the reference compounds NVP, EFV, dapivirine, and phenylethyl-S-DABOs 11 ¹³ and 12,¹⁴ were also evaluated against recombinant WT HIV-1 RT and against a panel of NNRTI-resistant RT variants (K103N, Y181I, and L100I). The concentration required to achieve 50% inhibition (ID₅₀), and the corresponding fold-resistance values are reported in Table .

3. Inhibitory Activity of 21-24 against HIV-1 WT and NNRTI-Resistant rRTs .

				ID50, μM (fold resistance)
Compd	R	R1	X	WT RT	K103N	Y181I	L100I
21a	Me	2,6-F2	CH2SMe	0.105±0.005	>20	>40	13.6±0.6 (129)
21b	Me	2,6-F2	(CH2)2SMe	0.008±0.0004	2.5±0.1 (312)	>40	3.3±0.1 (412)
21c	Me	2,6-F2	CH2SPh	0.021±0.001	0.45±0.02 (21)	9.2±0.4 (438)	1.5±0.08 (71)
21d	Me	2,6-F2	CH2S(4-FPh)	0.006±0.0003	0.09±0.005 (15)	5.1±0.3 (850)	0.68±0.03 (113)
21e	Me	2,6-F2	CH2S(4-MeOPh)	0.002±0.0001	0.044±0.002 (22)	4.3±0.2 (2150)	0.55±0.03 (275)
21f	Et	2,6-F2	CH2SMe	0.074±0.004	13.8±0.5 (187)	>40	5.2±0.3 (70)
21g	Et	2,6-F2	(CH2)2SMe	0.027±0.001	11.9±0.4 (441)	>40	2.5±0.1 (93)
21h	Et	2,6-F2	CH2SPh	0.079±0.004	16.8±0.5 (213)	12.8±0.5 (162)	2.1± 0.09 (27)
21i	Et	2,6-F2	CH2S(4-FPh)	0.081±0.005	7.5±0.3 (92)	13.4±0.6 (166)	1.7±0.07 (21)
21j	Et	2,6-F2	CH2S(4-MeOPh)	0.071±0.004	12.6±0.5 (177)	>40	1.3±0.05 (18)
21k	i-Pr	2,6-F2	CH2SMe	0.043±0.002	5.3±0.3 (124)	>40	4.2±0.2 (97)
21l	i-Pr	2,6-F2	(CH2)2SMe	0.095±0.005	>20	>40	4.5±0.2 (47)
21m	i-Pr	2,6-F2	CH2SPh	0.105±0.005	>40	>40	4.1±0.2 (39)
21n	i-Pr	2,6-F2	CH2S(4-FPh)	0.145±0.007	>40	>40	4.2±0.2 (29)
21o	i-Pr	2,6-F2	CH2S(4-MeOPh)	0.185±0.009	>40	>40	8.5±0.4 (46)
22a	Me	2-Cl-6-F	CH2SMe	0.269±0.01	>40	>40	6.7±0.3 (25)
22b	Me	2-Cl-6-F	(CH2)2SMe	0.028±0.001	18.2±0.8 (650)	>40	3.3±0.1 (118)
22c	Me	2-Cl-6-F	CH2SPh	0.031±0.001	8.4±0.4 (271)	10.3±0.5 (332)	2.6±0.1 (84)
22d	Me	2-Cl-6-F	CH2S(4-FPh)	0.011±0.0004	0.635±0.03 (58)	6.5±0.3 (591)	2.4 ± 0.1(218)
22e	Me	2-Cl-6-F	CH2S(4-MeOPh)	0.006±0.0003	0.078±0.003 (13)	2.9±0.1 (483)	1.2±0.05 (200)
22f	Et	2-Cl-6-F	CH2SMe	0.145±0.006	>40	>40	>40
22g	Et	2-Cl-6-F	(CH2)2SMe	0.079±0.004	11.9±0.6 (151)	>40	2.6±0.1 (33)
22h	Et	2-Cl-6-F	CH2SPh	0.095±0.005	16.7±0.7 (176)	12.8±0.5 (135)	2.1±0.1 (22)
22i	Et	2-Cl-6-F	CH2S(4-FPh)	0.081±0.004	7.4±0.3 (91)	13.4±0.6 (165)	1.6±0.06 (20)
22j	Et	2-Cl-6-F	CH2S(4-MeOPh)	0.071±0.004	12.6±0.6 (177)	>40	1.3±0.05 (18)
22k	i-Pr	2-Cl-6-F	CH2SMe	0.075±0.004	>40	>40	>20
22l	i-Pr	2-Cl-6-F	(CH2)2SMe	0.125±0.005	>20	>40	5.5±0.2 (44)
22m	i-Pr	2-Cl-6-F	CH2SPh	0.119±0.005	>40	>40	4.2±0.2 (35)
22n	i-Pr	2-Cl-6-F	CH2S(4-FPh)	0.169±0.007	>40	>40	4.2±0.2 (25)
22o	i-Pr	2-Cl-6-F	CH2S(4-MeOPh)	0.397±0.02	>40	>40	13.5±0.7 (34)
23a	Me	2,6-Cl2	CH2SMe	0.390±0.02	>40	>40	>40
23b	Me	2,6-Cl2	(CH2)2SMe	0.093±0.005	>40	>40	1.7±0.07 (18)
23c	Me	2,6-Cl2	CH2SPh	0.042±0.002	0.965±0.05 (23)	7.8±0.3 (186)	2.3±0.1 (55)
23d	Me	2,6-Cl2	CH2S(4-FPh)	0.010±0.002	0.445±0.02 (44)	4.3±0.2 (430)	1.8±0.09 (180)
23e	Me	2,6-Cl2	CH2S(4-MeOPh)	0.005±0.0002	0.045±0.002 (9)	1.4±0.06 (280)	1.1±0.05 (217)
23f	Et	2,6-Cl2	CH2SMe	0.52±0.03	>40	>40	>40
23g	Et	2,6-Cl2	(CH2)2SMe	0.65±0.03	>40	>40	>40
23h	Et	2,6-Cl2	CH2SPh	0.36±0.002	>40	>40	5.7±0.2 (16)
23i	Et	2,6-Cl2	CH2S(4-FPh)	0.26±0.01	>40	>40	>40±1.9
23j	Et	2,6-Cl2	CH2S(4-MeOPh)	0.18±0.009	17.9±1 (99)	>40	3±0.1 (17)
23k	i-Pr	2,6-Cl2	CH2SMe	0.60±0.03	>40	>40	>40
23l	i-Pr	2,6-Cl2	(CH2)2SMe	0.396±0.02	>20	>40	>30
23m	i-Pr	2,6-Cl2	CH2SPh	0.419±0.02	>40	>25	>40
23n	i-Pr	2,6-Cl2	CH2S(4-FPh)	0.925±0.05	>20	>40	15.7±0.8 (17)
23o	i-Pr	2,6-Cl2	CH2S(4-MeOPh)	0.438±0.02	>40	>40	12.2±0.6 (28)
24a	Me	3,5-Me2	CH2SMe	0.345±0.02	>40	5.5±0.2 (16)	5.5±0.3 (16)
24b	Me	3,5-Me2	(CH2)2SMe	0.05±0.002	>20	2.3±0.1 (47)	2.98±0.1 (60)
24c	Me	3,5-Me2	CH2SPh	0.098±0.005	>20	1.6±0.08 (16)	>40
24d	Me	3,5-Me2	CH2S(4-FPh)	0.106±0.005	>20	8.05±0.4 (76)	>40
24e	Me	3,5-Me2	CH2S(4-MeOPh)	0.129±0.006	>20	11.9±0.6 (92)	7.2±0.3 (56)
24f	Et	3,5-Me2	CH2SMe	0.048±0.002	>40	3.5±0.2 (73)	4.1±0.2 (85)
24g	Et	3,5-Me2	(CH2)2SMe	0.084±0.007	9.85±0.5 (117)	6.6±0.3 (79)	2.91±0.12 (35)
24h	Et	3,5-Me2	CH2SPh	0.019±0.0009	>20	7.5±0.4 (390)	1.9±0.1 (100)
24i	Et	3,5-Me2	CH2S(4-FPh)	0.020±0.001	>20	7.97±0.5 (398)	2.39±0.1 (119)
24j	Et	3,5-Me2	CH2S(4-MeOPh)	0.026±0.001	>20	8.74±0.5 (336)	1.65±0.08 (63)
24k	i-Pr	3,5-Me2	CH2SMe	0.026±0.001	7.3±0.4 (281)	19.4±1 (746)	1.8±0.09 (69)
24l	i-Pr	3,5-Me2	(CH2)2SMe	0.012±0.0006	5.61±0.3 (467)	2.13±0.19 (177)	0.44±0.02 (37)
24m	i-Pr	3,5-Me2	CH2SPh	0.014±0.0007	6.35 ± 0.3 (794)	2.21±0.09(158)	0.98±0.05 (122)
24n	i-Pr	3,5-Me2	CH2S(4-FPh)	0.037±0.002	>40	>25	>40
24o	i-Pr	3,5-Me2	CH2S(4-MeOPh)	0.058±0.003	>40	>25	>40
11	Me	2,6-Cl2	CH2CH2(4-MeOPh)	0.026±0.001	3.2± 0.2 (123)	1.5±0.07 (57)	ND
12	Me	2,6-F2	CH2CH2(4-MeOPh)	0.029±0.001	3.5±0.2 (127)	>20	ND
NVP				0.4±0.02	6.8± 0.3 (17)	33±2 (82)	8.5±0.4 (21)
EFV				0.03±0.001	3 ±0.2 (100)	0.09±0.005 (3)	ND
Dapivirine				0.007±0.0003	0.1±0.04 (14)	>1	0.09±0.005 (13)

^aValues are means ± SD determined from at least three experiments.

^bInhibitory dose 50, compound dose required to inhibit the HIV-1 rRT activity by 50%.

^cFold resistance: ratio of ID₅₀mut/ID₅₀WT values.

^dHigher concentration could not be tested due to precipitation of compounds in the culture medium.

^eND, not determined.

Overall, the enzymatic data for 21–24 strongly correlated with the cellular results, supporting HIV-1 RT as their primary target and substantially confirming the SAR profiles regarding the substitutions at positions C2, C5, and C6 observed in cellular assays.

The only small deviation was observed in the 3,5-Me₂-benzyl derivatives bearing a p-F/p-MeO substitution at the PTM side chain (24d,e,i,j,n,o), where an Et group at C5 instead of the i-Pr conferred the maximum inhibitory potency against both WT and mutated rRTs. However, consistently with the cellular data, in the 3,5-Me₂-benzyl series 24 the best inhibitory potency against WT and mutated enzymes was obtained with the 5-isopropyl-3,5-Me₂-benzyl MTE derivative 24l and its unsubstituted PTM analogue 24m (Table ). Again, in line with cellular results, in the 2,6-dihalobenzyl series 21–23, the introduction of a Me group at C5 and of a MeO group or, to lesser extent, a F atom at para position of the PTM side chain in C2 provided the highest inhibition against all tested RTs, leading to the two most potent derivatives 21e and 23e. Endowed with single-digit nanomolar inhibitory potency against WT RT and (sub)­micromolar potency toward mutated RTs, 21e and 23e resulted sensibly more potent than NVP, EFV, and the phenethyl-S-DABOs 11 and 12 versus all tested RTs, except Y181I in case of EFV. Moreover, both were slightly more active than dapivirine against WT and K103N RTs, comparable against Y181I, and weaker versus L100I mutated enzymes (Table ).

2D-QSAR Analysis

To relate the structural features of these anti-HIV-1 agents to their activity, we performed a systematic QSAR analysis based on molecular field topology analysis (MFTA). In addition to conventional SAR, QSAR provides a quantitative mapping of descriptors across the entire molecular scaffold. By projecting these descriptors onto molecular supergraphs, MFTA can reveal both universal and mutation-dependent activity “hot-spots”. Based on the EC₅₀ and ID₅₀ data (Tables –), we calculated the pID₅₀ = −Log­(ID₅₀) and pEC₅₀ = −Log­(EC₅₀) values. These data and the corresponding structures of organic compounds were entered into an SDF file using DataWarrior. Inactive compounds were included together with active analogues, whereas IRLL98 and K103N+Y181C were excluded due to low amount of data. The SDF files were used in the MFTA program, and the molecular supergraphs (Figure ) were constructed for each compound series and each strain.

3.

Compound 21b and its superposition on a molecular supergraph.

For each molecular supergraph, the spatial distribution of molecular descriptors was calculated, including: (i) the effective atomic charge (Q), estimated through electronegativity equalization; (ii) the effective van der Waals radius of the first atomic environment (Re), reflecting the steric contributions of the central non-hydrogen atom and other bonded atoms; (iii) group lipophilicity (Lipo), accounting for the contributions of the central non-hydrogen atom and the hydrogen atoms bonded to it; and (iv) the hydrogen-bond donor (Hd) and acceptor (Ha) capacities of an atom.

In the resulting contour maps, red regions indicate favorable contributions of a given descriptor to the modeled property, whereas blue regions denote unfavorable descriptor localization. Color intensity reflects the magnitude of the contribution. Charge descriptors correspond exclusively to the analysis of σ-bond electronic density. Tables and show data obtained based on ID₅₀ and EC₅₀ values, respectively.

4. Molecular Supergraphs, Constructed Based on ID₅₀ (Fold Resistance) Values.

5. Molecule Supergraphs, Constructed Based on EC₅₀ (Fold Resistance) Values.

Charge Descriptor Analysis

The analysis was focused on the structural fragments positively contributing to the activity across all HIV-1 strains studied (Tables and ). The correspondence between the numbering scheme and the molecular positions is as follows: number 2 corresponds to C2 of the pyrimidinone ring, number 10 corresponds to C5, while number 7 to C6. Numbers 18 and 20 correspond to positions 2 and 6, respectively, of the benzene ring of the benzyl substituent at the C6 atom, whereas atoms 32 and 33 correspond to positions 3 and 5 of the benzene ring. In the ID₅₀ models (Table ), the most recurrent charge pattern involved positions 10-14-30: positions 10 and 14 correspond to a partially negative charge, while position 30 is preferable for a partially positive charge (electronic density donors). This corresponds to R = alkyl substituents from Tables and . The same preferable positions can be seen for charge descriptors in EC₅₀ data (Table ). Another common position is position 6 with a partially positive charge, which is the same in the case of EC₅₀. Positions 23 and 24, which are part of the benzene ring in the X substituent, exhibit preferable partial negative charges that are common across all enzyme variants in the ID₅₀ data, indicating that these positions serve as favorable electronic density acceptors. EC₅₀ data for charges (Table ) shows a common pattern for the NL4-3 WT, Y181C, and K103N strains for the benzene ring in positions 21-17-15-16-19-22 (benzene ring of C6 benzyl substituent), where positive charges are preferred. This indicates that the benzene ring is a preferable electronic density donor here, which corresponds to preferable electronic density acceptors in positions 18 and 20. This corresponds to the 21-17-15-16-19-22 positions of the ring in the case of ID₅₀ data (Table ), and the picture looks somewhat different. Finally, positions 32 and 33 show a recurrent preference for increased electron density across all strains (except for Y181I), which corresponds to R₁ = 3,5-Me₂ and indicates a preference for electron-density donors.

VdW Descriptor Analysis

Positions 14-30 show preferable small VdW substituents for both ID₅₀ and EC₅₀, consistent with optimal activity for R = Me (Tables –). Regions 33-21 and 19-33 preferentially accommodate sterically small groups across most viral strains, with the notable exception of Y181I in ID₅₀ assays. This corresponds to R₁ = 3,5-Me₂ in Tables –. Conversely, EC₅₀ data indicates that the symmetric positions 20 and 18 indeed favor bulkier substituents for the NL-43WT, Y181C, and Y188L strains. This trend corresponds to R₁ = 2,6-Cl₂ substitution, suggesting that increased steric effects at these positions enhance antiviral potency.

Hydrogen Bond Analysis

Only hydrogen-bond acceptors (HBAs) were analyzed, since the variable structural regions did not contain hydrogen-bond donors. An HBA in position 28 is favorable for antiviral activity across all tested strains, except for Y181C and Y181I, as supported by both EC₅₀ and ID₅₀ datasets. This corresponds to the methoxy oxygen in the X = p-MeOPhSCH₂ substituent at C2, highlighting the importance of a polar electron-rich site in this region for productive ligand-target interactions. Likewise, position 9 is favorable for HBAs against all strains based on EC₅₀ values and for all, but the Y181I variant in ID₅₀ assays. This position is associated with the sulfur atom in X = (CH₂)₂SMe; notably, oxygen substitution at this site seems tolerated, suggesting some flexibility in the nature of the HBA while preserving activity. Position 6 is unfavorable for HBAs for all the strains across all data.

Lipophilicity Analysis

Positions 33-21 and 19-32 are unfavorable for lipophilicity for all strains in both ID₅₀ and EC₅₀, with the sole exception of the Y181I strain in the ID₅₀ dataset. Interestingly, position 14 is favorable for lipophilic substituents for all strains across all the data, consistent with the enhanced activity observed for R = Me. Position 29 also preferentially accommodates lipophilic substituents, particularly against WT NL4-3, Y181C, and K103N strains in EC₅₀ data, and against K103N and L100I in ID₅₀ data; this effect can be attributed to the methyl group of the para-methoxy substituent (X = p-MeOPhSCH₂). On the other hand, position 30 is intolerant to lipophilic substitution across all datasets, indicating a sterically or polarity-constrained region.

Overall, differences between the ID₅₀ and EC₅₀-based descriptor distributions likely reflect the distinct experimental readouts and the additional biological factors influencing cellular antiviral activity. Nonetheless, the QSAR models identified consistent electronic, steric, hydrogen-bonding, and lipophilicity patterns that determine inhibitor activity across HIV-1 strains. Key positions (e.g., 10-14-30, 14-30, 28, 29) were shown to strongly influence potency and explain why certain substituents improve or weaken activity. The analysis revealed both common structural requirements for broad-spectrum activity and mutation-dependent preferences. The QSAR results provide descriptor-based guidelines for further optimization of the series.

Complementary 3D-QSAR, ligand- and structure-based docking, and COMBINE analyses further supported the observed SAR by showing, for the representative highly active derivative 21e, accommodation of the 2,6-difluorobenzyl moiety within the aromatic-rich NNBS pocket, preservation of the key hydrogen bond between the pyrimidinone ring and Lys101, and extension of the C2 thioether side chain into an adjacent channel where additional van der Waals contacts contribute to ligand accommodation (see Figures S5–S8 and Table S10 in Supporting Information Part 2).

Conclusion

Four series of MTM-, MTE-, and PTM-S-DABO derivatives (21–24) were evaluated against WT and drug-resistant HIV-1 variants. Cellular and enzymatic data confirmed the established S-DABOs SAR, − ,− ,,,,,− highlighting the cooperative roles of the C6 benzyl, C5 alkyl, and C2 thioether substituents in antiviral activity (Figure ).

4.

SAR graphical summary of MTM-, MTE-, and PTM-S-DABOs 21–24.

Overall, ortho-dihalogenation at C6 was preferred over 3,5-Me₂ substitution, with a general potency trend 2,6-F₂ > 2,6-Cl₂ > 2-Cl-6-F > 3,5-Me₂, although specific inversions depend on C2 and C5 substitutions. In the 2,6-dihalobenzyl series, a C5 methyl group was generally optimal, whereas larger alkyl groups reduced activity against WT and HIV-1 mutants. However, selected derivatives showed an “HEPT-like” behavior, with C5 isopropyl becoming favorable, particularly in MTM-S-DABOs bearing 2,6-F₂ or 2-Cl-6-F benzyl groups, and in 3,5-Me₂-benzyl analogues irrespective of the C2 substituent. The C2 thioether side chains further modulated activity in a context-dependent manner. The best antiviral profile was observed for PTM-S-DABOs of the 2,6-dihalobenzyl thiothymine series (21c–e, 22c–e, and 23c–e), that always followed the activity order pMeOPhSCH₂ > pFPhSCH₂ > PhSCH₂. Conversely, the 3,5-Me₂-benzyl S-DABOs 24 showed the opposite trend, with PhSCH₂ outperforming para-substituted PTM groups, suggesting a distinct binding mode. Among these compounds, the C5 isopropyl derivatives 24l and 24m were the most active (Figure ).

The two most active derivatives, the 5-methyl-6-(2,6-F₂-benzyl) p-MeO PTM-S-DABO 21e and its 2,6-Cl₂-benzyl analogue 23e, showed subnanomolar-to-low-nanomolar activity against WT HIV-1 replication and rRT, combined with minimal cytotoxicity, excellent selectivity, and submicromolar-to-single digit nanomolar activity against drug-resistant HIV-1 variants and clinically relevant mutated rRT (Tables –). Both compounds outperformed NVP and EFV across most clinically relevant strains, except Y181C, and displayed potency comparable to dapivirine against WT, K103N, IRLL98, and K103N+Y181C. They also consistently surpassed the phenethyl-S-DABO analogues 11 and 12 across nearly all mutant variants, supporting the phenylthiomethyl side chain as superior to the 2-phenylethyl group for improving resistance tolerance. Although the best 3,5-Me₂-benzyl-substituted derivatives (24l,m) were less potent than 21e and 23e against WT, IRLL98, K103N, and K103N+Y181C strains, they retained comparable activity against Y181C and Y188L, with minimal fold resistance.

QSAR and molecular docking investigations further supported these SAR trends, linking broad-spectrum anti-HIV-1 activity and resistance tolerance to a defined combination of electronic, steric, hydrogen bonding, and lipophilic features. Electron-rich para-substituted phenylthiomethyl fragments, optimized C5 substituents, and properly positioned hydrogen-bond acceptors emerged as key determinants of productive RT inhibition.

In conclusion, this study identifies PTM-S-DABOs, particularly 21e and 23e, as promising next-generation NNRTI leads, defining a clear optimization map for the S-DABO chemotype and supporting the phenylthiomethyl thiothymine motif as a robust platform for developing NNRTIs with improved efficacy against clinically relevant HIV-1 strains.

Experimental Section

Chemistry

Melting points were determined on a Buchi 530 melting point apparatus and are uncorrected. ¹H-NMR spectra were recorded at 400 MHz on a Bruker AC400 spectrometer, reporting chemical shifts in δ (ppm) units relative to the internal reference tetramethylsilane. Mass spectra were recorded on an API-TOF Mariner by Perspective Biosystems (Stratford, TX, USA), and samples were injected by a Harvard pump using a flow rate of 5-10 μL/min and infused into the electrospray system. Elemental analyses were obtained by a PE 2400 (PerkinElmer) analyzer and were used to determine the purity of the described compounds that is >95%. Analytical results are within ±0.40% of the theoretical values. All compounds were routinely checked by TLC and ¹H-NMR. TLC was performed on aluminum-backed silica gel plates (Merck DC, Alufolien Kieselgel 60 F₂₅₄) with spots visualized by UV light. Yields of reactions refer to purified products. All chemicals were purchased either from Aldrich Chimica, Milan (Italy), or from TCI Europe NV, Zwijndrecht (Belgium), and were of the highest purity. As a rule, samples prepared for physical and biological studies were dried in high vacuum over P₂O₅ for 20 h at temperatures ranging from 25 to 40 °C, depending on the sample’s melting point.

General Procedure for the Preparation of 5-Alkyl-6-benzyl-2-{[(methylthio)­methyl]­thio}­pirimidin-4­(3H)-ones. Example: 6-(2,6-Dichlorobenzyl)-5-ethyl-2-{[(methylthio)­methyl]­thio}­pyrimidin-4­(3H)-one (23f, MC2569)

A solution of commercially available chloromethyl methyl sulfide (369 mg, 0.32 mL, 3.8 mmol) and 6-(2,6-dichlorobenzyl)-5-ethyl-2-thiouracil 25h (0.6 g, 1.9 mmol) in anhydrous DMF (6 mL) was stirred overnight at room temperature. After completion the resulting mixture was quenched with cold water (60 mL), stirred for 1 h, and extracted with ethyl acetate (3×75 mL). The organic layer was washed in sequence with water (2×50 mL) and brine (3×25 mL), filtered through a TLC-grade silica gel pad, dried with sodium sulfate, and evaporated to dryness in vacuo. The residue was recrystallized from a mixture of acetonitrile and THF to give pure 23f as a white solid. Yield, 53%; mp, 203-204°C. ¹H-NMR (CDCl₃) δ: 1.22 (t, 3H, CH₂CH ₃), 2.01 (s, 3H, SCH ₃), 2.70 (q, 2H, CH ₂CH₃), 3.94 (s, 2H, CH ₂Ar), 4.30 (s, 2H, SCH ₂S), 7.16-7.20 (m, 1H, C₄-H Ar), 7.34-7.36 (m, 2H, C_3,5-H Ar), 12.03 (br s, 1H, CONH) ppm. MS (ESI), m/z: 375 [M + H]⁺.

General Procedure for the Preparation of 5-Alkyl-6-benzyl-2-{[2-(methylthio)­ethyl]­thio}­pirimidin-4­(3H)-ones. Example: 6-(3,5-Dimethylbenzyl)-5-isopropyl-2-{[2-(methylthio)­ethyl]­thio}­pyrimidin-4­(3H)-one (24l, MC2947)

6-(3,5-Dimethylbenzyl)-5-isopropyl-2-thiouracil 25l (1.2 g, 4.2 mmol) was added to a stirred solution of 84.5% potassium hydroxide (277 mg, 4.2 mmol) in 20 mL of 95% ethanol. The resulting mixture was stirred for an additional hour at room temperature, treated with commercially available chloroethyl methyl sulfide (0.55 g, 0.49 mL, 5.0 mmol), and stirred overnight at room temperature. The next day, the mixture was further stirred at reflux under protection from atmospheric CO₂ until the complete conversion of the starting 2-thiouracil derivative. Subsequently, the reaction mixture was chilled to room temperature, diluted with water, stirred for 1 h, and extracted with ethyl acetate (3×75 mL). The organic layer was washed successfully with water (2×50 mL) and brine (3×25 mL), filtered through a TLC-grade silica gel pad, dried over sodium sulfate, and evaporated to dryness in vacuo. The resulting residue was recrystallized from n-hexane to afford pure 24l as a white solid. Yield, 35%, mp 106–107°C. ¹H-NMR (CDCl₃) δ 1.16-1.32 (d, 6H, CH­(CH ₃)₂), 2.01 (s, 3H, SCH ₃), 2.16-2.32 (s, 6H, C_3,5–CH ₃ ), 2.63-2.78 (t, 2H, SCH₂CH ₂ SCH₃), 3.00-3.08 (m, 1H, CH(CH₃)₂), 3.16-3.25 (t, 2H, SCH ₂CH ₂ SCH₃), 3.81 (s, 2H, CH ₂Ar), 6.83 (s, 1H, C₄-H Ar), 6.89 (s, 2H, C_2,6-H Ar), 11.95 (br s, 1H, CONH). MS (ESI), m/z: 363 [M + H]⁺.

General Procedure for the Preparation of 5-Alkyl-6-benzyl-2-{[(arylthio)­methyl]­thio}­pirimidin-4­(3H)-ones. Example: 6-(2-Chloro-6-fluorobenzyl)-5-methyl-2-{[(phenylthio)­methyl]­thio}­pyrimidin-4­(3H)-one (22c, MC2955)

6-(2-Chloro-6-fluorobenzyl)-2-thiothymine 25d (1.2 g, 4.2 mmol) was added to a stirred solution of 84.5% potassium hydroxide (277 mg, 4.2 mmol) in 20 mL of 95% ethanol. The resulting mixture was stirred for an additional hour at room temperature, treated with commercially available chloromethyl phenyl sulfide (0.79 g, 0.67 mL, 5.0 mmol), and stirred overnight at room temperature. The next day, the mixture was stirred at reflux under protection from atmospheric CO₂ until the complete conversion of the starting 2-thiothymine derivative. Subsequently, it was chilled to room temperature and diluted with water. The resulting precipitate was filtered off, washed with water and air-dried. Recrystallization of the crude product from 95% ethanol until the constant melting point furnished the target compound 22c. Yield, 64%; mp 181-183°C; ¹H-NMR (CDCl₃) δ 2.09 (s, 3H, CH ₃), 4.00 (s, 2H, ArCH ₂), 4.24 (s, 2H, SCH ₂S), 6.89-6.95 (m, 1H, C₄ -H Ar), 7.05-7.20 (m, 7H, C_3,5 -H and CH Ar rings), 12.45 (br s, 1H, CONH). MS (ESI), m/z: 407 [M + H]⁺

Synthetic and analytical characterization data for the remaining final compounds of series 21–24, including yields, melting points, ¹H-NMR and ESI-MS analyses, are reported in the Supplementary Information Part 1.

Biology. Anti-HIV-1 Activity in Lymphoid Cells

Biological activity of the compounds was tested in the lymphoid MT-4 cell line (received from the NIH AIDS Reagent Program) against the WT HIV-1 NL4-3 strain and five different HIV-1 strains, as described before. − Briefly, MT-4 cells were infected with the appropriate HIV-1 strain (or mock-infected to determine cytotoxicity) in the presence of different drug concentrations. On day five postinfection, a tetrazolium-based colorimetric method (MTT method) was used to evaluate the number of viable cells. The IRLL98 HIV-1 strain contains the following mutations in the RT coding sequence: M41L, D67N, Y181C, M184 V, R211 K, and T215Y (conferring resistance to NRTIs) and mutations K101Q, Y181C, and G190A (conferring resistance to NNRTIs). The HIV-1 K103N, Y181C, Y188L, and the K103N+Y181C mutants were received from the Medical Research Council Centralized Facility for AIDS Reagents, Herfordshire, UK.

RT Inhibition Assays

Chemicals:[³H]­dTTP (40 Ci/mmol) was purchased from Hartmann Analytic and unlabeled dNTPs from Promega. PerkinElmer was the supplier of GF/C filters. All other reagents were of analytical grade and purchased from Merck. Nucleic acid substrate: The homopolymer poly­(rA) was mixed at weight ratios in nucleotides of 10:1 to the oligomer oligo­(dT)­18in 20 mM Tris-HCl (pH 8.0) containing 20 mM KCl and 1 mM EDTA, heated at 65 °C for 5 min and then slowly cooled at room temperature. Enzymatic assay: The co-expression vectors pUC12N/p66­(His)/p51with the wild-type or the mutant forms of HIV-1 RT p66 were kindly provided by Dr. S. H. Hughes (NCI-Frederick Cancer Research and Development Center). Proteins were expressed in Escherichia coli and purified as described. RNA-dependent DNA polymerase activity was performed in a final volume of 15 μL containing reaction buffer (50 mM Tris-HCl pH 7.5, 1 mM DTT, 0.2 mg/mL BSA, 4% glycerol), 10% DMSO, 10 mM MgCl₂, 0.5 μgof poly­(rA)/oligo­(dT)­10:1 (0.3 μM3′–OH ends), 10 μM [³H]-dTTP (1 Ci/mmol), and 2-4 nM RT. Reactions were incubated at 37 °C for 20 min. Then 10 μL aliquots were spotted on glass fiber filters GF/C that were immediately immersed in 5% ice-cold TCA. Filters were washed twice in 5% ice-cold TCA and once in ethanol for 5 min and dried, and acid-precipitable radioactivity was quantified by scintillation counting. Data obtained from monitoring the incorporation of radioactive [³H]­dTTP into poly­(rA)/oligo­(dT) in the presence of increasing doses of inhibitors were plotted using Graphpad Prism 5.0 software.

Computational Investigations

See the separate Supporting Information Part 2 file.

Glossary

Abbreviations

CC₅₀

compound concentration toxic for 50% of cells

DABO

3,4-dihydro-2-alkoxy-6-benzyl-4-oxo-pyrimidine

DACO

dihydro-aryl/alkylthio-cyclohexylmethyl-oxopyrimidine

DATNO

dihydro-alkylthio-naphthylmethyl-oxopyrimidine

DAPY

diarylpyrimidine

EC₅₀

effective concentration 50, concentration needed to protect 50% of cells from the HIV-1 induced cytopathogenicity

EFV

efavirenz

F₂-NH-DABO

2-alkyl/arylamino-6-(2,6-difluorobenzyl)-4-(3H)-pyrimidinone

F₂-N,N-DABO

2-(N,N-disubstituted) amino-6-(2,6-difluorobenzyl)-4-(3H)-pyrimidinone

Ha

hydrogen-bond acceptor

HBDs

hydrogen bond donors

HBAs

hydrogen bond acceptors

Hd

hydrogen-bond donor

HEPT

1-[(2-hydroxyethoxy)­methyl]-6-(phenylthio)­thymine

HIV-1

human immunodeficiency virus 1

ID₅₀

50% inhibiting dose

MFTA

molecular field topology analysis

MTE-S-DABOs

6-(arylmethyl)-2-{[(methylthio)­ethyl]­thio}­pyrimidin-4­(3H)-ones

MTM-S-DABOs

6-(arylmethyl)-2-{[(methylthio)­methyl]­thio}­pyrimidin-4­(3H)-ones

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

NNBS

non-nucleoside binding site

NVP

nevirapine

PTM-S-DABOs

6-(arylmethyl)-2-{[(phenylthio)­methyl]­thio}­pyrimidin-4­(3H)-ones

QSAR

quantitative structure–activity relationship

RT

reverse transcriptase

SI

selectivity index

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c08020.

• Synthetic and analytical characterization of compounds 21–24: yields, melting points, ¹H NMR and ESI-MS data; physical and chemical data for the newly synthesized final compounds (Table S1); elemental analyses of the newly synthesized final compounds (Table S2) (PDF)

• Computational investigations on compounds 21–24: ligand-based 3D-QSAR/Py-CoMFA and structure-based molecular docking and Py-ComBinE/COMBINE approaches; computational modeling results and validation; experimental computational procedures; data and software availability; molecular structures, SMILES, pID50 values, and dataset assignment (Table S1); conformational analyses (Table S2); molecular alignment methods and scoring functions (Table S3); molecular properties used for template selection (Table S4); molecular parameters of compounds 21–24 (Table S5); Py-CoMFA models and alignment rules (Table S6); statistical performance of the selected 3D-QSAR model (Table S7); experimental, recalculated, cross-validated, and predicted pID50 values (Table S8); AutoDock Vina re-docking assessment (Table S9); optimized Py-ComBinE models (Table S10); experimental and predicted pID50 values for WT and mutant HIV-1 RTs (Table S11); 3D-QSAR model performance (Figure S1); steric and electrostatic activity-contribution maps (Figures S2 and S3); residue-interaction contributions from the Py-ComBinE model (Figure S4); representative binding modes of compound 21e in WT, K103N, Y181I, and L100I HIV-1 RTs (Figures S5–S8) (PDF)

#.

G.S.H., E.D., and E.F. contributed equally. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.