2-(Arylamino)-6-(trifluoromethyl)nicotinic Acid Derivatives: New HIV-1 RT Dual Inhibitors Active on Viral Replication

The persistence of the AIDS epidemic, and the life-long treatment required, indicate the constant need of novel HIV-1 inhibitors. In this scenario the HIV-1 Reverse Transcriptase (RT)-associated ribonuclease H (RNase H) function is a promising drug target. Here we report a series of compounds, developed on the 2-amino-6-(trifluoromethyl)nicotinic acid scaffold, studied as promising RNase H dual inhibitors. Among the 44 tested compounds, 34 inhibited HIV-1 RT-associated RNase H function in the low micromolar range, and seven of them showed also to inhibit viral replication in cell-based assays with a selectivity index up to 10. The most promising compound, 21, inhibited RNase H function with an IC50 of 14 µM and HIV-1 replication in cell-based assays with a selectivity index greater than 10. Mode of action studies revealed that compound 21 is an allosteric dual-site compound inhibiting both HIV-1 RT functions, blocking the polymerase function also in presence of mutations carried by circulating variants resistant to non-nucleoside inhibitors, and the RNase H function interacting with conserved regions within the RNase H domain. Proving compound 21 as a promising lead for the design of new allosteric RNase H inhibitors active against viral replication with not significant cytotoxic effects.


Introduction
The acquired immunodeficiency syndrome (AIDS) pandemic is still one of the major world health concerns. Presently approved treatments do not allow immunization of people or the eradication of the infection once acquired, and among the estimated 37.9 million people living with HIV worldwide in 2019, only 24.5 million people were accessing the current antiretroviral therapy [1]. While the highly-active antiretroviral therapy (HAART) allows patients who receive optimum treatment a life-expectation comparable to that of the uninfected population [2,3], the sub-optimal adherence and coverage of seropositive people, coupled with life-long treatment, still cause emergence of drug-resistant variants, that could lead to therapy failure [4], and whose transmission to naïve patient is a major concern and requires an optimized treatment [5][6][7]. These data remark the constant need to expand the antiretroviral armamentarium with new drugs, with novel targets, to be employed on

Evaluation of RNase H inhibitory activity
To establish the potential biologic effect of the synthesized compounds we tested them on the HIV-1 RT-associated RNase H activity, using as a control the RNase H inhibitors RDS1759 [25] and beta-Thujaplicinol (BTP) [49] (Table 1)

Evaluation of RNase H Inhibitory Activity
To establish the potential biologic effect of the synthesized compounds we tested them on the HIV-1 RT-associated RNase H activity, using as a control the RNase H inhibitors RDS1759 [25] and beta-Thujaplicinol (BTP) [49] (Table 1). The ethyl nicotinate 9 showed promising inhibitory activity (IC 50    OEt The amides showed a RNase H inhibitory activity rather different from esters. The nicotinamide 6 was completely inactive, while the oxime 25 showed more potent anti-RNase H activity, being the best inhibitor with an IC 50 value of 0.7 µM. The monosubstituted aryl amides 28-30 and 37 inhibited RNase H activity with IC 50 in the 5.6-20 µM range. The replacement of the 3-methyl group of amide 29 with a 3-methoxy group to give compound 33 abolished the activity, while the presence of two or three methoxy groups as in compounds 34-36 repristinated the anti-RNase H efficacy. The benzylamides 38-47 showed RNase H inhibitory activity (IC 50 in the 9-24 µM concentration range). The arylpiperazino amides 49 and 51 displayed some RNase H inhibitory potency (IC 50 values = 18 and 20 µM, respectively). The introduction of a second chlorine atom on amide 49, to give the 3,4-dichlorophenyl piperazine analog 52, produced an increase in potency (IC 50 7 µM).

Evaluation of Antiviral Activity
The 28 compounds that were able to inhibit the HIV-1 RNase H activity with an IC 50 value below 30 µM were tested for the ability to inhibit HIV-1 replication in TZMbl cells ( Table 2). The 4-chlorophenyl ester 13 showed to be active on the HIV replication in the same range of concentration of the IC 50 value, with an EC 50 value of 10 µM, with a selectivity index (SI) of 3.6. The 3,4-dimethoxyphenyl ester 21 resulted as the best ester inhibitor of the HIV replication activity with an EC 50 value of 5 µM and a SI greater than 10. Amides 34 and 36 showed to be able to inhibit HIV-1 replication (EC 50 values = 2 and 1.8 µM, respectively) although residual cytotoxic effect activity caused a poor selectivity (SI = 2.6 and 1.8, respectively), and also the amides 49 and 52 displayed HIV inhibitory activity (EC 50 values = 10 µM and SI = 2.7 and 1.6, respectively). Noteworthy, the most active RNase H inhibitor in biochemical assays, compound 25, was inactive in the HIV replication assay.

Mode of Action Studies
In order to confirm our hypothesis of allosteric binding mode, we firstly examined the possible interaction between selected compounds and the magnesium, used as a cofactor by both the RT-associated enzymatic functions. The UV spectra of the seven compounds able to inhibit viral replication were measured in the absence and in the presence of magnesium ions, using as a positive control the active site inhibitor RDS1759. Results ( Figure 2) showed that, while the positive control displayed a typical shift in the maximum of absorbance of the spectra, among the tested compounds only amide 29 showed a shift of 17 nm in the peak of absorbance, while amides 34 and 52 showed a moderate hyperchromic effect, but not a shift. Differently, the spectra of compounds 13, 21, 36, and 49 were not affected by the presence of the Mg 2+ . Overall, the data seem to exclude that Mg 2+ coordination might be involved in the mechanism of action of the selected compounds. In order to further dissect the mode of action the most promising compound, we selected ester 21 that inhibited the RNase H function with an IC50 value of 14 μM, the HIV-1 replication in cell culture with EC50 value of 5 μM and showed a SI  10, along with the amide derivative 49, that was In order to further dissect the mode of action the most promising compound, we selected ester 21 that inhibited the RNase H function with an IC 50 value of 14 µM, the HIV-1 replication in cell culture with EC 50 value of 5 µM and showed a SI ≥ 10, along with the amide derivative 49, that was also active against HIV-1 replication but showed a higher toxicity in TZMbl cells. Both compounds were tested against the RDDP function (Table 3), showing IC 50 values (16.1 and 22.9 µM, respectively), very similar to the ones obtained on the RNase H function, indicating a dual-function allosteric inhibitory profile. To clarify their binding mode, and also ascertain the possibility that these compounds could be active against circulating drug-resistant variants, compounds 21 and 49 were tested against both RT-associated functions of a panel of enzymes carrying single amino acid substitutions in the polymerase domain: mutations K103N, Y181C that are related to resistance to the NNRTIs [29], and mutation V108A that is related to the binding of the allosteric RNase H inhibitors on the pocket close to the NNRTI binding site [42]. Besides, both compounds were tested also on RTs mutated in two amino acid residues within the RNase H domain, Q475A and A502F, part of conserved regions of RNase H domain [50], that have been shown to be involved in the binding of RNase H inhibitors into a pocket close to the RNase H catalytic site but acting as allosteric RNase H inhibitors [25,37,40].
We tested the inhibition of the HIV-1 RDDP activity of mutated HIV-1 RTs by compounds 21 and 49 (Table 3), using as control the NNRTI efavirenz (EFV). Interestingly, the two compounds displayed a profile very different from EFV. Indeed, compounds 21 and 49 were able to inhibit the K103N and Y181C RT-associated RDDP activities. In particular, compound 21 showed no difference in potency of inhibition towards the two enzymes, while compound 49 showed a 2.5-fold decrease in potency of inhibition for K103N RT and a decrease of two-fold against Y181C RT. As expected, the IC 50 value of EFV showed an increase of 7.6-fold against K103N, while it did not show change in potency of inhibition when tested against the V108A RT RDDP activity. Differently, compounds 21 and 49 showed a consistent increase in IC 50 value (5.1-and >4.3-fold for compound 21 and 49, respectively) on the V108A RT RDDP function. Of note, all the inhibitors were able to inhibit the RT-associated RDDP activity of A502F and Q475A RTs, except for 49 that showed to increase its RDDP IC 50 value on Q475A RT of 2.5-fold.
The HIV-1 RNase H activity of the mutated HIV-1 RTs in the presence of compounds 21 and 49 was analyzed using as control the RNase H active site inhibitor BTP (Table 4). Results showed that the positive control equally inhibited all the tested enzymes, while compound 21 had a 3.7-fold increase in IC 50 value on the K103N RT RNase H function, and it completely lost its RNase H inhibitory efficacy towards V108A, Q475A, and A502F RTs. Differently, compound 49 showed a moderate decrease in its potency of inhibition on the RNase H activity of the K103N, Y181C, and A502F RTs, being completely ineffective against the RNase H function of V108A and Q475A RTs. Overall, results obtained on the V108A RT, whose RDDP and RNase H functions were both non-susceptible to inhibition by compounds 21 and 49, indicate a major role played by the pocket close to the NNRTI binding site on inhibition of these compounds of both the enzymatic activities, with a combination of short-and long-range effects, and suggest a key role of residue V108 for the correct interaction between the inhibitors and the binding pocket. On the second hand, the fact that amino acidic substitutions Q475A and A502F caused a loss of potency only for inhibition of their RNase H functions suggest that this second allosteric pocket exerts a short-range influence only within the RNase H domain.

Discussion
Dual inhibitors are a new class of molecules that aim to increase the efficacy of the drug and raise the genetic barrier to overcome the selection of resistant strains by hitting two targets at the same time [44,51,52]. In the constant need to identify and develop new anti-HIV drugs, possibly active against circulating drug-resistant variants, this approach has been applied to RDDP/Integrase and RNase H/Integrase dual inhibitors [53][54][55][56][57][58][59][60][61], that target two different viral enzymes, and to RNase H/RDDP dual inhibitors, that target two distinct, but strongly interconnected, enzymatic functions located in the same enzyme [9,36,62].
In this work, we designed a set of 44 new compounds following the pharmacophoric requirements previously identified for RNase H/RDDP dual-site dual inhibitors [38][39][40][41][42]63]. In this case, the central core was represented by the 6-(trifluoromethyl)nicotinic acid, and, differently from the previous attempt to design nicotinate-based antimicrobial agents that had two aromatic functions [45,46], we decorated the core, making ester or amide of the nicotinic acid and introducing an arylamino moiety at the 2-position. The scaffold maintained a sp2 hybridization state of the linker [40,42], that in this case involved the aryl ring, with a steric hindrance slightly bigger than previously reported compounds. For the synthesis of the compounds, we successfully applied our previously reported method for the synthesis of 2-arylamino-6-trifluoromethyl-3-pyridine carboxylic acids [47], and the final esters and amides were obtained by EDCI method in good yields and high purity without use of a tedious purification procedure [48].
The designed molecules showed a consistent inhibition of the HIV-1 RT-associated RNase H function, with the amide derivatives being overall slightly more potent than the ester counterparts. Cell-based assays revealed that the scaffold is more promising than the previous reported attempts of RNase H/RDDP dual inhibitors. Among the 44 2-(arylamino)-6-(trifluoromethyl)nicotinic acid derivatives, 7 compounds inhibited viral replication, while no active compound has been identified so far among the isatin-derivatives [41,42]. Furthermore, among the pyrrole and pyrazole derivatives, only one dual inhibitor was active against viral replication [40], being less active (EC 50 = 25 µM) and more toxic (CC 50 = 44 µM).
The 3,5-dimethoxyphenyl ester 21 was able to block HIV-1 replication with a SI > 10. The mode of action studies revealed that the compound acts differently form NNRTI inhibitors, with an activity profile against RDDP activity different from EFV on K103N and Y181C RTs, since these amino acid substitutions had a lower impact on its efficacy as compared to previously reported compounds [40,42,64]. Site-directed mutagenesis experiments showed that compound 21 binds independently to two different allosteric sites previously described: S first pocket (pocket 1), contiguous with the NNRTI binding pocket, was described for hydrazones [35], isatines [41,42] and pyrazoles [40]. A second pocket (pocket 2), within the RNase H domain, was firstly reported for selective allosteric RNase H inhibitors, such as hydrazones derivatives [36,37], and later for isatines [41,42] and pyrazoles [40]. For years the contribution of the two pockets to the inhibition of the two enzymatic functions was controversial. Our results confirmed that pocket 2 is an allosteric site for selective inhibition of HIV-1 RT-associated RNase H, since mutations A502F and Q475A completely abrogated the inhibitory effect of compound 21 on that sole function, proving also that these two conserved residues [50] play a key role on compound activity. Moreover, we showed that binding to pocket 1 is sufficient for inhibition of both enzymatic activities, with a combination of longand short-range effects, as previously hypothesized [36,42], since the introduction of amino acidic substitution V108A had a strong impact on the inhibition of both RDDP and RNase H functions by compound 21.
These results provide useful insights for the design of novel and more potent HIV-1 RT RNase H/RDDP allosteric dual inhibitors, active against selected drug-resistant variants. With this aim, a major focus should be pointed on pocket 1 as the most suitable to be exploited for drug design, and on compound 21 as a promising good hit to be developed.

Chemistry
All commercially available solvents and reagents were used without further purification. NMR spectra were recorded on an Inova 500 spectrometer (Varian, Palo Alto, CA, USA). The chemical shifts (δ) are reported in parts per million downfield from tetramethylsilane (TMS), which was used as internal standard, and the spectra were recorded in hexadeuteriodimethylsulphoxide (DMSO-d 6 ). Infrared spectra were recorded on a Vector 22 spectrometer (Bruker, Bremen, Germany) in Nujol mulls. The main bands are given in cm −1 . Positive-ion electrospray ionization (ESI) mass spectra were recorded on a double-focusing MAT 95 instrument (Finnigan, Waltham, MA, USA) with BE geometry. Melting points (mp) were determined on a SMP1 Melting Point apparatus (Stuart Scientific, Stone, UK) and were uncorrected. All reported products showed NMR spectra in agreement with the assigned structures. The purity of the tested compounds was determined by combustion elemental analyses conducted by the Microanalytical Laboratory of the Department of Chemical and Pharmaceutical Sciences of the University of Ferrara with a MT-5 CHN recorder elemental analyzer (Yanagimoto, Kyoto, Japan), and the values found were within 0.4% of theoretical values. Compounds 6, 9 [47] 11-15, 19, 21, 23, 27, 29-37 and 49-51 were prepared as previously described [48].

General Procedure for the Synthesis of Esters (11-24)
A mixture of acid 10 (0.33 g, 1 mmol), EDCI (1.92 g, 1.1 mmol), and HOBt (0.13 g, 1 mmol) in dry MeCN (10 mL) was stirred at room temperature for 30 min and then treated with the appropriate phenol (1 mmol). The mixture was stirred at room temperature for an additional 24 h. Then the solution was evaporated to dryness in vacuo. The residue was dissolved in ethyl acetate (20 mL) and washed with brine (2 × 5 mL), 5% aqueous sodium hydroxide (2 × 5 mL), and water (2 × 5 mL). The organic layer was dried over anhydrous sodium sulfate. Concentration of the dried extract yielded a residue which was triturated with di-isopropyl ether. The formed precipitate was filtered off and purified by crystallization from the adequate solvent to give the ester derivatives 11-24.

Mg 2+ Coordination
The inhibitors were diluted in water to a final concentration of 100 µM. UV spectra recorded in the absence and after addition of 6 mM MgCl 2 in a water solution, using a Nano Drop ONE reader (Thermo Scientific). Results are plotted as absorbance vs wavelength using Prisme 6, version 6.01.

Site-Directed Mutagenesis
Amino acid substitutions were introduced into the p66 HIV-1 RT subunit coded in a p6HRT-prot plasmid using the QuikChange protocol (Agilent Technologies Inc., Santa Clara, CA).

Antiviral Assay
Drug-mediated inhibition of virus-induced cytotoxicity was assayed in TZM-BL cells. Triplicate wells of 96-well plates containing 1 × 10 4 TZMbl cells were infected with HIV-1 IIIB strain at a multiplicity of infection of 0.1. Serial dilutions of drugs were added immediately after infection. Luciferase activity was measured as reported [67]. Two days after infection, the culture medium was removed from each well and 100 µl of Bright Glo reagent (Promega, Luis Obispo, CA) was added to the cells for measurement of luminescence using a Victor 2 luminometer (Perkin). The 50% effective concentration (EC 50 ) was defined as the concentration that caused a 50% reduction of luciferase activity (relative light units) compared to virus control wells. Cell viability in CEM cells was quantified 5 days after infection with the MTT-dye reduction method.

Conclusions
We identified a new promising 2-amino-6-(trifluoromethyl)nicotinic acid scaffold for compounds that can inhibit both HIV-1 RT activities and are also active against viral replication. Compound 21, active against viral replication in low micromolar range with no significant cytotoxic effects, showed to be a dual inhibitor active also against the RNA dependent DNA polymerase activity of HIV-1 RT, being active against RTs carrying NNRTI-resistant mutations. Mode-of-action and single-site mutagenesis studies indicated that compound 21 potency of inhibition is strongly impaired by the substitutions of amino acidic residues in two different pockets, located in two different domains of RT enzyme, hence supporting the hypothesis of a dual-site binding mode. Overall compound 21 is a promising lead for the design of new allosteric RNase H inhibitors more potent against HIV-1 replication.  Funding: This research was funded by the Sardinian Regional Government grant LR07/17 (F76C18000800002).

Conflicts of Interest:
The authors declare no conflicts of interest.