Human Immunodeficiency Virus type 1 (HIV-1) is responsible of the infection of 30 million people worldwide. The ssRNA HIV-1 genome retrotranscription into proviral dsDNA is a fundamental step in the HIV-1 replication cycle. This process is carried out by the viral coded RT, a multifunctional protein that catalyzes different reactions combining the RNA- and DNA-dependent DNA polymerase and RNase H activities, both essential for viral replication (Esposito and Tramontano, 2014). After the retrotranscription process, the viral enzyme IN allows the integration of the HIV-1 genome into the host cell chromosome, through two essential catalytic reactions, named 3′-processing and strand-transfer (Esposito et al., 2015). Despite an effective antiretroviral therapy has been developed, the incidence of HIV infection continues to raise. The development of new drugs with new mode of actions, such as single molecules that may act on two different target enzymes or two different catalytic functions (Patyar et al., 2011), would reduce the number of administered drugs, their chronic toxicity and the chance of selecting drug resistant viruses (Costi et al., 2014; Esposito and Tramontano, 2014); hence it remains one of the current main goals in drug discovery.

Metal-chelating molecules have been extensively studied as antivirals, resulting in the individuation of important classes of metal-dependent enzyme inhibitors (Ronconi and Sadler, 2007; Hocharoen and Cowan, 2009; Liao et al., 2010; Rogolino et al., 2012; Barry and Sadler, 2013). In particular, drugs targeting HIV-1 IN have evolved into an important component of the currently used clinical protocols. Their mechanism of action involves the chelation of the magnesium cofactors within the active site of the HIV-1 IN enzyme. The first clinically developed IN inhibitor (INI), raltegravir (Isentress^®), was approved in late 2007 (Summa et al., 2008; Hicks and Gulick, 2009). In 2012, elvitegravir (Sato et al., 2006) and in 2013 dolutegravir (Kawasuji et al., 2013) jointed the therapeutic pool, and other chelating IN inhibitors are currently in clinical trials (Karmon and Markowitz, 2013).

Since HIV IN catalytic domain exhibits striking structural similarities with the viral RT-associated RNase H domain (Rogolino et al., 2012), the development of dual-acting drugs targeting both viral IN and HIV-1 RT-associated RNase H function has been proposed as an interesting strategy (Didierjean et al., 2005; Billamboz et al., 2008; Marchand et al., 2008; Corona et al., 2014a; Costi et al., 2014; Cuzzucoli Crucitti et al., 2015). In fact, both HIV-1 IN and RNase H domains belong to the polynucleotidyl transferases superfamily (Esposito and Tramontano, 2014); these similarities allow to investigate the structural features that a compound should hold to achieve a dual IN/RNase H inhibition, taking advantage of the interactions that an inhibitory compound can make with the amino-acidic lateral chains that surround the catalytic core of both IN and RNase H (Corona et al., 2016). It is worth noting that, despite the RNase H function is a promising target for drug development (Corona et al., 2014b) and different classes of compounds were reported as potent inhibitors of the HIV-1 RNase H function (Esposito et al., 2012; Corona et al., 2013), no RNase H inhibitors have moved toward an advanced stage of (pre)clinical development. N′-acylhydrazones represent an interesting class of chelating ligands with a broad spectrum of antiviral activities (Borkow et al., 1997; Rollas and Küçükgüzel, 2007), including HIV (Gong et al., 2011; Sanchez et al., 2013; Herschhorn et al., 2014), vaccinia virus (Abdel-Aal et al., 2006) and influenza virus (Chen et al., 2014). Recently, we focused our attention on a family of N′-acylhydrazones which proved to be versatile chelating inhibitors of the influenza virus PA endonuclease (Carcelli et al., 2016). We also identified some N′-acylhydrazone metal complexes with promising in vitro antiviral activity against various DNA- and RNA-viruses (Rogolino et al., 2015). Based on the functional similarity between HIV IN, HIV RNase H and the influenza PA endonuclease (Rogolino et al., 2012), we decided to evaluate a panel of N′-acylhydrazones against both HIV IN and RNase H, with the aim of identifying a unique chelating motif effective across metal-dependent enzymes of diverse viruses. Herein we present the biological evaluation in both enzymatic and cellular assays of a series of chelating N′-acylhydrazones (compounds 1–23, Figure 1).

The structure-activity relationship for these compounds was analyzed, based on their presumed metal binding modes, thus identifying the chelating features involved in dual inhibition of both HIV IN and RNase H. Interestingly, some of these compounds (i.e., 18, 20, 21 and 23) inhibit the RT-associated RNase H function in the micromolar and high nanomolar range. In particular, compound 18 is also able to inhibit viral replication in cell-based antiviral assays in the low micromolar range, representing one of the best inhibition profiles so far reported. Furthermore, computational modeling studies were conducted for compound 18 and some model ligands to explore the binding mode of these compounds within the active sites of the two HIV enzymes.

All reagents of commercial quality were used without further purification. The purity of the compounds was determined by elemental analysis and verified to be ≥95%. NMR spectra were recorded at 25°C on a Bruker Avance 400 FT spectrophotometer. The attenuated total reflectance IR spectra were recorded by means of a Nicolet-Nexus (Thermo Fisher) spectrophotometer by using a diamond crystal plate in the range of 4000–400 cm^-1. Elemental analyses were performed by using a FlashEA 1112 series CHNS/O analyzer (Thermo Fisher) with gas-chromatographic separation. Electrospray mass spectral analyses (ESI-MS) were performed with an electrospray ionization (ESI) time-of-flight Micromass 4LCZ spectrometer. MS spectra were acquired in positive EI mode by means of a direct exposure probe mounting on the tip of a Re-filament with a DSQII Thermo Fisher apparatus, equipped with a single quadrupole analyzer.

The coordinating versatility of N′-acylhydrazone ligands is well-known (Albrecht et al., 2009; Ray et al., 2009; Rogolino et al., 2015). If a 2-hydroxy substituted phenyl ring is present on the backbone of the ligand, it can coordinate one or, depending on denticity, two metal centers (M₁ and M₂ in Figure 2). Moreover, if a gallic moiety is introduced as B (Figure 2), an additional coordinating mode arises (M_I and M_II). In our recent studies, this versatility was demonstrated to be at the basis of efficient inhibition of influenza virus PA endonuclease, due to chelation of the two metal ions in the catalytic center of the enzyme (Carcelli et al., 2016). Hence we wanted to verify whether this versatility could be exploited for the development of effective dual chelating inhibitors of other metal-dependent enzymes, like HIV-1 RNase H and IN, with two Mg(II) ions in the catalytic site.

In fact, while most of the tested compounds were able to inhibit the RNase H function, compounds 7–9 showed also to be able to inhibit the HIV-1 IN, envisaging the possibility to design dual inhibitors, with a synergic effect against both HIV RNase H and IN. Of note, the activity of compound 10 toward RNase H is particularly interesting. In fact, compound 10, In fact, since 10 does not carry the 2-hydroxyphenyl group in A, it lacks the possibility to chelate in a tridentate ONO fashion (Mode A in Figure 2); however, it can coordinate two cations by means of its three OH groups. This chelating mode was indeed observed in the crystal structure of 23 complexed with the N-terminal part of the influenza PA endonuclease, which showed that 23 chelates the two manganese (II) cofactors through its gallic moiety (Carcelli et al., 2016).

Results on compounds 10–23 indicated that more stringent requirements should be met to obtain efficient inhibition of HIV IN with these ligands, when none (11), one (13) or two (14–18) hydroxyl groups are present on aromatic ring A, or even when this is replaced by a pyridine moiety (12). Compounds 11–17 proved to be selective RNase H inhibitors. The goal of identifying dual chelating inhibitors of both HIV-1 RNase H and IN, was fulfilled in compound 18 and in the series 20–23, where the N′-acylhydrazone scaffold carries three hydroxyl substituents on the aromatic ring A. In particular, compound 18 inhibits RNase H function ninefold more potently that IN activity and the HIV replication in the same micromolar range. Compounds 20–23 inhibit both HIV enzymes more potently than compound 18 and their inhibitory ability strongly depends on the position of the hydroxyl substituents, but a different trend has been detected for each enzyme, presumably attributable to the interactions established within the two different active sites. Compounds 20–23, unfortunately, showed poor reproducibility in HIV replication assay, compared to compounds 18. This can be due to lower solubility in cell culture, which prevented from obtaining reliable data for this series. Compounds 18, on the contrary, with inhibition of HIV replication of 17 μM, can be considered a better hit. For compounds 11–18, chelation of the metal cofactors according to mode B (Figure 2) could be hypothesized. This is consistent with (i) the activity of compound 11 that cannot exhibit the chelating mode A because it lacks the 2-OH group; (ii) the inactivity of compound 19, where the gallic moiety is replaced by a 3,4,5-trimethoxy benzene ring. Evidently, the 3,4,5-trihydroxybenzyl moiety in B is essential, but not sufficient to ensure potent enzyme inhibition, since the interactions of ring A, for example with the amino acid side chains of the protein, appear to be crucial in modulating the inhibitory activity.

For the compounds 20–23, we hypothesize that the inhibitory activity of this gallic moiety-containing compound series is determined by (i) the ability of chelating the metal ions in the active site of the enzyme; and (ii) the presence and position of the hydroxyl substituents in A, which may possibly yield relevant ligand–protein interactions (e.g., through hydrogen bonds).

It has been reported that hydrazones have an interesting dual inhibition profile, since they inhibit both RT-associated RNase H and RDDP functions (Borkow et al., 1997). In particular, hydrazones were reported to inhibit the RNase H function (i) binding to an allosteric site located between the polymerase active site and the non-nucleoside reverse transcriptase inhibitor (NNRTI) pocket (Himmel et al., 2006); (ii) binding to a site located between the RNase H active site and the connection domain (Gong et al., 2011). Differently, our data support the hypothesis that the present metal-chelating N′-acylhydrazone series bind as chelating agents into the RNase H active site.

Intriguingly, striking similarities can be seen in the coordination modes proposed for the best compound 18 and compounds 15 and 23 compared to those of some 3-hydroxypyrimidine-2,4-dione-5-N-benzylcarboxamides, recently described as potent inhibitors of HIV-1 IN and RNase H (Wu et al., 2016). In fact, compounds 15, 18 and 23 are predicted to establish secondary interactions with PFV residues Asp185, Tyr212, and Glu221, which correspond to HIV IN Asp116, Tyr143 and Glu152 (Dayer, 2016). Moreover, compound 23 establish additional secondary interactions with the hydroxyl group of Tyr212 and with the nitrogen of the peptidic bridge between Asp185 and Gln186. Importantly, while residue Tyr143 is known to be involved in the interaction with the first generation IN Inhibitors (Corona et al., 2016), both the catalytic Asp116 and Asn117 residues are part of a short, highly conserved region (H114-G118) (Ceccherini-Silberstein et al., 2009), that bridges the β4 sheet, with the α5 helix. Also of note the fact that among the residues hypothesized to be involved in the interaction between N′-acylhydrazone derivatives and the HIV-1 RT RNase H domain, some exhibit a high degree of conservation, even among patients treated with RT inhibitors (Santos et al., 2008). In particular, residue Asn474 is part of the RNase H primer grip motif and its mutation has been shown to determine to a strong decrease in viral fitness (Dash et al., 2008). Interestingly, residue Asn474 has been previously shown to play a crucial role also for pyrrolyl diketoacids binding (Corona et al., 2014a) and hence it seems to be involved in the binding of chelating agents to the RNase H active site.

Finally, aiming to develop dual HIV-1 RNase H – IN inhibitors, it is also worth noting that the predicted interactions obtained for the model ligands toward IN and RNase H would suggest two different coordinating modes: while for IN it can reasonably proposed the binding mode B (Figure 3B), in the case of RNase H an alternative coordinative hypothesis can be invoked (Figure 4B). In fact, docking results for ligands 15, 18 and 23 within the RNase H active site reveal that the gallic moiety is oriented perpendicularly to the plane formed by metal cofactors and the coordinating residues, with only two hydroxyl groups involved in the chelation (Figure 4B). This unusual orientation could depend on the presence, in the RNase H active site, of a fourth residue (i.e., Asp549), which is able to provide for an additional coordinating bond with a metal ion, thus blocking a coordination site of the metal. This essential information is clearly to be taken into account for further drug development of dual target inhibitors.

The metal-chelating N′-acylhydrazones derivatives are a good source to design dual inhibitors targeted to both HIV-1 RT-associated RNase H and IN activities. The best identified analog, compound 18, inhibits in the micromolar range both RNase H and IN functions and is also able to inhibit viral replication in the cell-based antiviral assay. The computational docking studies performed support the hypothesis that the N′-acylhydrazones derivatives can chelate the metal ions in both HIV-1 RNase H and IN active sites, and that the functional determinants for their efficacy are different with respect to the two enzymes. This information, together with the one that the binding of these compounds in the enzyme catalytic sites can be stabilized by the interaction with conserved amino acid residues, will be essential for the development of further analog optimization.

All authors have given approval to the final version of the manuscript. MC, DR, and AG: drug design and chemical synthesis. NP: docking studies. FE, AnC, AlC, CP, LN, and ET: biological studies. MC, DR, ET, and FE wrote the paper.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer SH and handling Editor declared their shared affiliation and the handling Editor states that the process nevertheless met the standards of a fair and objective review.