Dihydroxyphenyl- and Heteroaromatic-Based Thienopyrimidinones to Tackle HIV-1 LEDGF/p75-Dependent IN Activity

The spread of Human Immunodeficiency Virus (HIV) still represents a global public health issue of major concern, and would benefit from unveiling unique viral features as targets for drug design. In this respect, HIV-1 integrase (IN), due to the absence of homologs in human cells, is a popular target for the synthesis of novel selective compounds. Moreover, as drug-resistant viral strains are rapidly evolving, the development of novel allosteric inhibitors is acutely required. Recently, we have observed that Kuwanon-L, quinazolinones and thienopyrimidinones containing at least one polyphenol unit, effectively inhibited HIV-1 IN activity. Thus, in the present research, novel dihydroxyphenyl-based thienopyrimidinone derivatives were investigated for their LEDGF/p75-dependent IN inhibitory activity. Our findings indicated a close correlation between the position of the OH group on the phenyl moiety and IN inhibitory activity of these compounds. As catechol may be involved in cytotoxicity, its replacement by other aromatic scaffolds was also exploited. As a result, compounds 21–23, 25 and 26 with enhanced IN inhibitory activity provided good lead candidates, with 25 being the most selective for IN. Lastly, UV spectrometric experiments suggested a plausible allosteric mode of action, as none of the thienopirimidinones showed Mg2+ chelation properties otherwise typical of IN strand transfer inhibitors (INSTIs).


Introduction
According to the World Health Organization, HIV transmission still represents an important worldwide public health problem, with around 1.3 million new infections worldwide by 2022 [1] mainly related to HIV-1, the agent responsible for the acquired immunodeficiency syndrome (AIDS).
Although many therapeutic strategies have been developed to strengthen the cART arsenal, the evolution of drug-resistant viral strains remains a serious issue.Thus, identifying novel viral factors as targets for drug design remains a priority.In this regard, HIV-1 integrase (IN) is an attractive target for new systematic drug discovery efforts, especially because of the absence of a cellular homolog that may result in a reduced occurrence of adverse effects [2].
IN is a multimeric enzyme and, together with reverse transcriptase (RT) and protease, is one of the three viral enzymes necessary for HIV-1 replication.It catalyzes the incorporation of double-stranded proviral DNA into the host cell genome [3] and comprises three domains: the N-terminal region, the central core domain containing the catalytic triad DDE that chelates two Mg 2+ ions, and the C-terminal region [4].In particular, the integration process requires the following two steps: (1) in the cytosol, the IN dimer catalyzes cleavage of a guanine-thymine (GT) dinucleotide from the 3 -terminus of each viral DNA end (3processing) and (2) in the nucleus, tetrameric IN mediates integration of recessed viral DNA ends into the cellular DNA via a concerted hydroxy-mediated nucleophilic reaction (DNA strand transfer).
The advent of a convenient in vitro functional assay based on recombinant IN [5] prompted drug discovery efforts to inhibit its DNA strand transfer activity by targeting the active site.As a result, strand transfer inhibitors (INSTIs) Raltegravir (RAL) [6,7], Elvitegravir (EVG) [8], Dolutegravir (DTG) [9], Cabotegravir [10] and Bictegravir [11] have been developed and approved for clinical use.Regrettably, the rapid emergence of viral variants displaying cross-resistance to the FDA-licensed INSTIs highlights the need to develop inhibitors with unique modes of action.
Molecules 2023, 28, x FOR PEER REVIEW 2 of 16 IN is a multimeric enzyme and, together with reverse transcriptase (RT) and protease, is one of the three viral enzymes necessary for HIV-1 replication.It catalyzes the incorporation of double-stranded proviral DNA into the host cell genome [3] and comprises three domains: the N-terminal region, the central core domain containing the catalytic triad DDE that chelates two Mg 2+ ions, and the C-terminal region [4].In particular, the integration process requires the following two steps: (1) in the cytosol, the IN dimer catalyzes cleavage of a guanine-thymine (GT) dinucleotide from the 3′-terminus of each viral DNA end (3′-processing) and (2) in the nucleus, tetrameric IN mediates integration of recessed viral DNA ends into the cellular DNA via a concerted hydroxymediated nucleophilic reaction (DNA strand transfer).
The advent of a convenient in vitro functional assay based on recombinant IN [5] prompted drug discovery efforts to inhibit its DNA strand transfer activity by targeting the active site.As a result, strand transfer inhibitors (INSTIs) Raltegravir (RAL) [6,7], Elvitegravir (EVG) [8], Dolutegravir (DTG) [9], Cabotegravir [10] and Bictegravir [11] have been developed and approved for clinical use.Regre ably, the rapid emergence of viral variants displaying cross-resistance to the FDA-licensed INSTIs highlights the need to develop inhibitors with unique modes of action.
Interestingly, a recent high throughput screening study provided a series of different promising hit to lead IN-inhibitor scaffolds from which the (1-(tert-butoxycarbonyl)-4phenylpiperidine-4-carbonyl)phenylalanine 4 was selected and used for further optimization [25] (Figure 1).In addition, the inhibitory activity of various natural flavonols containing polyphenol moieties, e.g., angoletin, quercetin, myricetin and kaempferol has been previously reported [26][27][28].The increasing interest in novel allosteric IN binding sites prompted some of us to investigate the recently discovered sucrose binding pocket, identifying Kuwanon-L (Figure 2) as a novel allosteric inhibitor.In silico investigation supported the biological The increasing interest in novel allosteric IN binding sites prompted some of us to investigate the recently discovered sucrose binding pocket, identifying Kuwanon-L (Figure 2) as a novel allosteric inhibitor.In silico investigation supported the biological findings, also shedding light on the possible binding mode of the polyphenolic scaffold within the sucrose binding pocket [29].findings, also shedding light on the possible binding mode of the polyphenolic scaffold within the sucrose binding pocket [29].Recently, we also reported on the ability of some 3,4-catechol-thienopyrimidinone [30,31] and quinazolinone derivatives [32], originally designed to allosterically inhibit the HIV-1 RT-associated Ribonuclease H (RNase H) function, to hamper LEDGF/p75dependent IN activity.Interestingly, with regard to thienopyrimidinones, we observed that the 2-(2,3-dihydroxyphenyl)-6-methyl-5-propylthieno [2,3-d]pyrimidin-4(3H)-one (5), bearing a 2,3-catecol moiety, displayed the best IN inhibitory activity [30].Figure 2 shows the most active compounds.
Thus, in the present paper, we investigated the LEDGF/p75-dependent IN activity of a series of dihydroxyphenyl-thienopyrimidinones using the strand transfer inhibitor Raltegravir as a positive control.Our results demonstrated a strict relationship between the position of the OH groups on the phenyl moiety and their IN inhibitory activity.Moreover, UV spectrometric experiments demonstrated that all of the thienopirimidinones were unable to chelate Mg 2+ , thereby suggesting a possible allosteric mode of action.As the catechol scaffold-although present in many biologically active compounds-may be associated with cytotoxicity [39,40], we also synthesized and explored the inhibitory activity of differently substituted thienopyrimidinones.They were also examined for their RT-RNase H activity using 2-(3,4-dihydroxyphenyl)-5,6dimethylthieno [2,3-d]pyrimidin-4(3H)-one (GZ 510) [30] as a positive control.
Thus, in the present paper, we investigated the LEDGF/p75-dependent IN activity of a series of dihydroxyphenyl-thienopyrimidinones using the strand transfer inhibitor Raltegravir as a positive control.Our results demonstrated a strict relationship between the position of the OH groups on the phenyl moiety and their IN inhibitory activity.Moreover, UV spectrometric experiments demonstrated that all of the thienopirimidinones were unable to chelate Mg 2+ , thereby suggesting a possible allosteric mode of action.As the catechol scaffold-although present in many biologically active compounds-may be associated with cytotoxicity [39,40], we also synthesized and explored the inhibitory activity of differently substituted thienopyrimidinones.They were also examined for their RT-RNase H activity using 2-(3,4-dihydroxyphenyl)-5,6-dimethylthieno [2,3-d]pyrimidin-4(3H)-one (GZ 510) [30] as a positive control.

Results
Methods of investigation are shown in Figure 3.

Chemistry
Figure 4 presents thienopyrimidinones investigated in the present research.As reported in Scheme 1, the novel thienopyrimidinones were synthesized according to our previously reported procedure for preparation of compounds 8, 9, 13, 14, 15-18, 23 and 26 [30].In brief, a Gewald reaction between cyanoacetamide, the appropriate carbonyl derivative, and elemental sulfur in the presence of diethylamine (DEA) as a base, was firstly carried out to afford the correspondent 2-amino-4,5-disubstituted-thiophene-3-carboxamide.Subsequently, the thiophene intermediate reacted via direct oxidative condensation with the proper aldehyde in the presence of molecular iodine, to give rise to the final products 10-12, 19-22, 24, 25, 27 and 28.As reported in Scheme 1, the novel thienopyrimidinones were synthesized according to our previously reported procedure for preparation of compounds 8, 9, 13, 14, 15-18, 23 and 26 [30].In brief, a Gewald reaction between cyanoacetamide, the appropriate carbonyl derivative, and elemental sulfur in the presence of diethylamine (DEA) as a base, was  As reported in Scheme 1, the novel thienopyrimidinones were synthesized according to our previously reported procedure for preparation of compounds 8, 9, 13, 14, 15-18, 23 and 26 [30].In brief, a Gewald reaction between cyanoacetamide, the appropriate carbonyl derivative, and elemental sulfur in the presence of diethylamine (DEA) as a base, was

Biology
All of the thienopyrimidinones were tested for their ability to affect the IN-LEDGF/p75 dependent activity.
From the analysis of the first series of compounds, it was immediately evident how the relative OH position on the aromatic core influenced the inhibitors' IC 50 (Table 1).Remarkably, moving the OH from 4-to 2-position seemed determinant for the improved activity of compound 10.Other possible combinations were explored, but the distancing of the second OH group negatively affected the inhibitory activity of compounds 12 and 13.The 3,5-dihydroxyphenyl core appeared to be the best combination, and 2-(3,5dihydroxyphenyl)-5,6-dimethylthieno [2,3-d]

Biology
All of the thienopyrimidinones were tested for their ability to affect the IN-LEDGF/p75 dependent activity.
From the analysis of the first series of compounds, it was immediately evident how the relative OH position on the aromatic core influenced the inhibitors' IC50 (Table 1).Remarkably, moving the OH from 4-to 2-position seemed determinant for the improved activity of compound 10.Other possible combinations were explored, but the distancing of the second OH group negatively affected the inhibitory activity of compounds 12 and 13.The 3,5-dihydroxyphenyl core appeared to be the best combination, and 2-(3,5dihydroxyphenyl)-5,6-dimethylthieno [2,3-d]pyrimidin-4(3H)-one 14 displayed the best IC50 value.

Biology
All of the thienopyrimidinones were tested for their ability to affect the IN-LEDGF/p75 dependent activity.
From the analysis of the first series of compounds, it was immediately evident how the relative OH position on the aromatic core influenced the inhibitors' IC50 (Table 1).Remarkably, moving the OH from 4-to 2-position seemed determinant for the improved activity of compound 10.Other possible combinations were explored, but the distancing of the second OH group negatively affected the inhibitory activity of compounds 12 and 13.The 3,5-dihydroxyphenyl core appeared to be the best combination, and 2-(3,5dihydroxyphenyl)-5,6-dimethylthieno [2,3-d]pyrimidin-4(3H)-one 14 displayed the best IC50 value.

Biology
All of the thienopyrimidinones were tested for their ability to affect the IN-LEDGF/p75 dependent activity.
From the analysis of the first series of compounds, it was immediately evident how the relative OH position on the aromatic core influenced the inhibitors' IC50 (Table 1).Remarkably, moving the OH from 4-to 2-position seemed determinant for the improved activity of compound 10.Other possible combinations were explored, but the distancing of the second OH group negatively affected the inhibitory activity of compounds 12 and 13.The 3,5-dihydroxyphenyl core appeared to be the best combination, and 2-(3,5dihydroxyphenyl)-5,6-dimethylthieno [2,3-d]pyrimidin-4(3H)-one 14 displayed the best IC50 value.

Biology
All of the thienopyrimidinones were tested for their ability to affect the IN-LEDGF/p75 dependent activity.
From the analysis of the first series of compounds, it was immediately evident how the relative OH position on the aromatic core influenced the inhibitors' IC50 (Table 1).Remarkably, moving the OH from 4-to 2-position seemed determinant for the improved activity of compound 10.Other possible combinations were explored, but the distancing of the second OH group negatively affected the inhibitory activity of compounds 12 and 13.The 3,5-dihydroxyphenyl core appeared to be the best combination, and 2-(3,5dihydroxyphenyl)-5,6-dimethylthieno[2,3-d]pyrimidin-4(3H)-one 14 displayed the best IC50 value.

Biology
All of the thienopyrimidinones were tested for their ability to affect the IN-LEDGF/p75 dependent activity.
From the analysis of the first series of compounds, it was immediately evident how the relative OH position on the aromatic core influenced the inhibitors' IC50 (Table 1).Remarkably, moving the OH from 4-to 2-position seemed determinant for the improved activity of compound 10.Other possible combinations were explored, but the distancing of the second OH group negatively affected the inhibitory activity of compounds 12 and 13.The 3,5-dihydroxyphenyl core appeared to be the best combination, and 2-(3,5dihydroxyphenyl)-5,6-dimethylthieno[2,3-d]pyrimidin-4(3H)-one 14 displayed the best IC50 value.Introducing more bulky substituents on the thiophene scaffold showed a different trend.In this case, distancing of the second OH group from the 2-hydroxy moiety improved inhibition of the IN LEDGF/p75-dependent activity, as demonstrated by the 3,5,6,7,8,9-hexahydro-4H-cyclohepta [4,5]thieno [2,3-d]pyrimidin-4-ones 16 and 17.Once again, the 3,5-dihydroxyphenyl unit proved the optimal core, with compound 18 being the best inhibitor of both series of thienopyrimidinones (Table 2 and Figure 5).Introducing more bulky substituents on the thiophene scaffold showed a different trend.In this case, distancing of the second OH group from the 2-hydroxy moiety improved inhibition of the IN LEDGF/p75-dependent activity, as demonstrated by the 3,5,6,7,8,9-hexahydro-4H-cyclohepta [4,5]thieno [2,3-d]pyrimidin-4-ones 16 and 17.Once again, the 3,5-dihydroxyphenyl unit proved the optimal core, with compound 18 being the best inhibitor of both series of thienopyrimidinones (Table 2 and Figure 5).Introducing more bulky substituents on the thiophene scaffold showed a different trend.In this case, distancing of the second OH group from the 2-hydroxy moiety improved inhibition of the IN LEDGF/p75-dependent activity, as demonstrated by the 3,5,6,7,8,9-hexahydro-4H-cyclohepta [4,5]thieno [2,3-d]pyrimidin-4-ones 16 and 17.Once again, the 3,5-dihydroxyphenyl unit proved the optimal core, with compound 18 being the best inhibitor of both series of thienopyrimidinones (Table 2 and Figure 5).Introducing more bulky substituents on the thiophene scaffold showed a different trend.In this case, distancing of the second OH group from the 2-hydroxy moiety improved inhibition of the IN LEDGF/p75-dependent activity, as demonstrated by the 3,5,6,7,8,9hexahydro-4H-cyclohepta [4,5]thieno [2,3-d]pyrimidin-4-ones 16 and 17.Once again, the 3,5-dihydroxyphenyl unit proved the optimal core, with compound 18 being the best inhibitor of both series of thienopyrimidinones (Table 2 and Figure 5).These results prompted us to explore substitution on the thiophene ring, and the most interesting compounds are reported in Table 3. Replacing the cyclohepthyl ring of 9 with a branched cyclohexane dramatically improved the inhibitory activity of compound 19 (Figure 6).Similarly, substitution of methyl groups of 8 with the bulkier and more flexible isobutyl moiety generated the more active thienopyrimidinone 20 (Table 3).As the presence of dihydroxyphenyls may induce cytotoxicity [39,40], we addressed the possibility of replacing the dihydroxyphenyl moiety with other aromatic scaffolds, testing the inhibitory activity on both IN and RNase H functions ( Introducing more bulky substituents on the thiophene scaffold showed a different trend.In this case, distancing of the second OH group from the 2-hydroxy moiety improved inhibition of the IN LEDGF/p75-dependent activity, as demonstrated by the 3,5,6,7,8,9-hexahydro-4H-cyclohepta [4,5]thieno [2,3-d]pyrimidin-4-ones 16 and 17.Once again, the 3,5-dihydroxyphenyl unit proved the optimal core, with compound 18 being the best inhibitor of both series of thienopyrimidinones (Table 2 and Figure 5).These results prompted us to explore substitution on the thiophene ring, and the most interesting compounds are reported in Table 3. Replacing the cyclohepthyl ring of 9 with a branched cyclohexane dramatically improved the inhibitory activity of compound 19 (Figure 6).Similarly, substitution of methyl groups of 8 with the bulkier and more flexible isobutyl moiety generated the more active thienopyrimidinone 20 (Table 3).These results prompted us to explore substitution on the thiophene ring, and the most interesting compounds are reported in Table 3. Replacing the cyclohepthyl ring of 9 with a branched cyclohexane dramatically improved the inhibitory activity of compound 19 (Figure 6).Similarly, substitution of methyl groups of 8 with the bulkier and more flexible isobutyl moiety generated the more active thienopyrimidinone 20 (Table 3).As the presence of dihydroxyphenyls may induce cytotoxicity [39,40], we addressed the possibility of replacing the dihydroxyphenyl moiety with other aromatic scaffolds, testing the inhibitory activity on both IN and RNase H functions (Table 4).These results prompted us to explore substitution on the thiophene ring, and the most interesting compounds are reported in Table 3. Replacing the cyclohepthyl ring of 9 with a branched cyclohexane dramatically improved the inhibitory activity of compound 19 (Figure 6).Similarly, substitution of methyl groups of 8 with the bulkier and more flexible isobutyl moiety generated the more active thienopyrimidinone 20 (Table 3).As the presence of dihydroxyphenyls may induce cytotoxicity [39,40], we addressed the possibility of replacing the dihydroxyphenyl moiety with other aromatic scaffolds, testing the inhibitory activity on both IN and RNase H functions (Table 4).Compounds 21-23 and 26 inhibited both IN and RNase-H activity, with 22 and 23 being more selective against IN, while introducing a -NO2 group on the thiophene core of 24 totally reversed its activity, making compound 25 completely IN selective (Table 4 and Figure 7).Intriguingly, both pyrrole compounds 27 and 28 were completely IN inactive, while the indole derivative 26 showed a good activity on both IN and RNase-H function (Table 4 and Figure 7).These results prompted us to explore substitution on the thiophene ring, and the most interesting compounds are reported in Table 3. Replacing the cyclohepthyl ring of 9 with a branched cyclohexane dramatically improved the inhibitory activity of compound 19 (Figure 6).Similarly, substitution of methyl groups of 8 with the bulkier and more flexible isobutyl moiety generated the more active thienopyrimidinone 20 (Table 3).As the presence of dihydroxyphenyls may induce cytotoxicity [39,40], we addressed the possibility of replacing the dihydroxyphenyl moiety with other aromatic scaffolds, testing the inhibitory activity on both IN and RNase H functions (Table 4).Compounds 21-23 and 26 inhibited both IN and RNase-H activity, with 22 and 23 being more selective against IN, while introducing a -NO2 group on the thiophene core of 24 totally reversed its activity, making compound 25 completely IN selective (Table 4 and Figure 7).Intriguingly, both pyrrole compounds 27 and 28 were completely IN inactive, while the indole derivative 26 showed a good activity on both IN and RNase-H function (Table 4 and Figure 7).As the presence of dihydroxyphenyls may induce cytotoxicity [39,40], we addressed the possibility of replacing the dihydroxyphenyl moiety with other aromatic scaffolds, testing the inhibitory activity on both IN and RNase H functions (Table 4).Compounds 21-23 and 26 inhibited both IN and RNase-H activity, with 22 and 23 being more selective against IN, while introducing a -NO 2 group on the thiophene core of 24 totally reversed its activity, making compound 25 completely IN selective (Table 4 and Figure 7).Intriguingly, both pyrrole compounds 27 and 28 were completely IN inactive, while the indole derivative 26 showed a good activity on both IN and RNase-H function (Table 4 and Figure 7).Finally, to investigate the possible IN inhibition mechanism, active thienopyrimidinones were tested for their metal-chelating properties regarding Mg 2+ ions implicated as cofactors in the mode of action of INSTIs [41].Interestingly, although catechol-based compounds are especially notorious for their metal complexing properties, our UV spectrometric experiments excluded the participation of Mg 2+ cations in their mode of action (see Supplementary Material).Finally, to investigate the possible IN inhibition mechanism, active thienopyrimidinones were tested for their metal-chelating properties regarding Mg 2+ ions implicated as cofactors in the mode of action of INSTIs [41].Interestingly, although catechol-based compounds are especially notorious for their metal complexing properties, our UV spectrometric experiments excluded the participation of Mg 2+ cations in their mode of action (see Supplementary Material).

Discussion
A series of thienopyrimidinones have been tested for their in vitro inhibitory activity on IN catalytic function.We chose the LEDGF/p75-dependent integration assay, as it is a versatile tool for detecting alteration of various IN functions such as multimerization, LEDGF/p75 binding and catalytic activity [25].
Structurally, a 3,4-catechol unit at the C-2 position of the thienopyrimidinone scaffold, regardless of any other substitution on the thiophene core, was a key factor in consistently fostering HIV-1 RT-RNase H but not IN activity [30].As a matter of fact, in our first experiment we observed that switching from 2,3-to 3,4-catechol turned 5 in the IN inactive compound 29 (IC 50 > 50 mM) (Figure 8).

Discussion
A series of thienopyrimidinones have been tested for their in vitro inhibitory activity on IN catalytic function.We chose the LEDGF/p75-dependent integration assay, as it is a versatile tool for detecting alteration of various IN functions such as multimerization, LEDGF/p75 binding and catalytic activity [25].
Structurally, a 3,4-catechol unit at the C-2 position of the thienopyrimidinone scaffold, regardless of any other substitution on the thiophene core, was a key factor in consistently fostering HIV-1 RT-RNase H but not IN activity [30].As a ma er of fact, in our first experiment we observed that switching from 2,3-to 3,4-catechol turned 5 in the IN inactive compound 29 (IC50 > 50 mM) (Figure 8).These observations prompted us to further investigate the relationship between the position of OH substitutions and IN LEDGF/p75-dependent activity.
Firstly, compounds 8 and 9 were selected as a baseline, since they were the most active against RNAse H and very poorly active against IN (Figure 9), and several derivatives exploring all the possible OH combinations were investigated.The results shown in Tables 1 and 2 are suggestive of a cooperative relationship between the OH relative position and the thiophene substitution.In particular, as also evidenced in Table 3  These observations prompted us to further investigate the relationship between the position of OH substitutions and IN LEDGF/p75-dependent activity.
Firstly, compounds 8 and 9 were selected as a baseline, since they were the most active against RNAse H and very poorly active against IN (Figure 9), and several derivatives exploring all the possible OH combinations were investigated.
The results shown in Tables 1 and 2 are suggestive of a cooperative relationship between the OH relative position and the thiophene substitution.In particular, as also evidenced in Table 3, bulkier and branched alkyl substitution in 5,6 positions may enhance IN inhibitory activity.
Moreover, compound 11, bearing only a 2-hydrophenyl core, was slightly more active than 8, hypothesizing a cooperative intramolecular hydrogen bonding between amide and hydroxy moiety.
Although the catechol unit is present in a wide variety of biologically active compounds, there is no doubt that its redox reactivity may be related to cytotoxicity [39,40].Thus, a series of thienopyrimidinones supporting different heteroaromatic cores have been investigated (Table 4).Our first candidate, the electron-rich furan ring often used as a bioisosteres of benzene core [42] provided a valid alternative, leading to compound 21 (RNase H IC 50 of 2.5 ± 0.1 and IN IC 50 of 5.75 ± 0.85).Introducing an electron-donating methyl group in 5-position resulted in a reduced activity of 22, even showing a higher IN selectivity.Conversely, the presence of the strong electron withdrawing unit NO 2 , also able to act as an H-bond acceptor as in compound 23, was a more successful substitution for IN activity while still maintaining selectivity against RNase H function.The presence of the more aromatic thiophene ring led to the inactive compound 24.As already observed with 23, once again the introduction of a NO 2 moiety appeared related to the IN inhibitory activity, as compound 25 is the most active and selective of the five member ring series.
consistently fostering HIV-1 RT-RNase H but not IN activity [30].As a ma er of fact, in our first experiment we observed that switching from 2,3-to 3,4-catechol turned 5 in the IN inactive compound 29 (IC50 > 50 mM) (Figure 8).These observations prompted us to further investigate the relationship between the position of OH substitutions and IN LEDGF/p75-dependent activity.
Firstly, compounds 8 and 9 were selected as a baseline, since they were the most active against RNAse H and very poorly active against IN (Figure 9), and several derivatives exploring all the possible OH combinations were investigated.The results shown in Tables 1 and 2 are suggestive of a cooperative relationship between the OH relative position and the thiophene substitution.In particular, as also evidenced in Table 3, bulkier and branched alkyl substitution in 5,6 positions may enhance IN inhibitory activity.
Moreover, compound 11, bearing only a 2-hydrophenyl core, was slightly more active than 8, hypothesizing a cooperative intramolecular hydrogen bonding between amide and hydroxy moiety.
Although the catechol unit is present in a wide variety of biologically active compounds, there is no doubt that its redox reactivity may be related to cytotoxicity [39,40].Thus, a series of thienopyrimidinones supporting different heteroaromatic cores have been investigated (Table 4).Our first candidate, the electron-rich furan ring often used as a bioisosteres of benzene core [42] provided a valid alternative, leading to compound 21 (RNase H IC50 of 2.5 ± 0.1 and IN IC50 of 5.75 ± 0.85).Introducing an electrondonating methyl group in 5-position resulted in a reduced activity of 22, even showing a higher IN selectivity.Conversely, the presence of the strong electron withdrawing unit  Interestingly, and conversely to what was previously observed for RNase H [30,31], the replacement of the dihydroxyphenyl core with other heteroaromatic nuclei had a greater impact on IN inhibitory activity.
As the indole moiety proved to be a key factor in some recently published IN allosteric inhibitors [43], we also examined compound 26, whose activity against RT-RNase H has been reported [30].Interestingly, it was most active against IN, although more selective for the RNase H function (Table 3), making it a good candidate for further optimization.Replacement of the indole moiety with pyrrole or 1-methyl pyrrole, as in compounds 27 and 28, caused a loss of the IN activity and also affected RNAse H inhibition.
Finally, our attempt to investigate the mechanism of IN inhibition revealed the inability of the tested compounds to chelate Mg 2+ ions.In fact, no Mg 2+ effect was observed, as the absorbance peaks of our compounds did not show a significant shift with respect to control compound RDS1643, even at a high concentration of Mg 2+ ions, suggesting a possible allosteric mode of action.

Chemistry
Unless otherwise specified, all reagents and solvents were obtained from commercial suppliers and used without further purification.

MgCl 2 Coordination Assay
The ability of the compounds to coordinate Mg 2+ were determined by means of previously reported protocols [31].The UV-Vis spectrum from 250 to 600 nm was recorded before and after the addition of 6 mM MgCl 2 and plotted using Prism9.

RNase H inibitory Assay
The previously reported procedure was used for preparation and purification of wild type p66/p51 HIV-1 RT.Enzyme was stored in a 50% glycerol-containing buffer at −20 • C [32,46].An 18-nt 3 -fluorescein-labeled RNA annealed to a complementary 18-nt 5dabsyl-labeled DNA was used to determine IC 50 values, following a previously described procedure [47].In brief, 1 µL of each inhibitor dissolved in DMSO was added to a 96-well plate, followed by 10 µL of the proper RT (15−80 ng/mL) in reaction buffer.The addition of 10 µL of RNA/DNA hybrid (2.5 µM) initiated the hydrolysis.Then, 50 mM Tris-HCl, pH 8.0, 60 mM KCl, 10 mM MgCl 2 , 1% DMSO, 150−800 ng of RT, 250 nM substrate, and increasing concentrations of inhibitor were chosen as final assay conditions.As negative control, wells containing only DMSO were used.Subsequently, the plates were incubated in a Spectramax Gemini EM fluorescence spectrometer for 10 min at 37 • C, measuring the fluorescence (λex = 485 nm; λem = 520 nm) at 1 min intervals in order to determine the linear initial rates in the presence (vi) and absence (vo) of inhibitor.Percent inhibition was assessed as 100(vo − vi)/vo and plotted against log[I].Prism5 (GraphPad Software 5.0) was used to calculate the IC 50 values.All assays were carried out in triplicate.

Conclusions
Since the dihydroxyphenyl unit has been evidenced in various allosteric IN inhibitors, we investigated here the activity of novel thienopyrimidinones bearing a dihydrohyphenyl core at C-2.Our findings showed a strict correlation between the OH position and their IN inhibitory activity, and coordination assays evidenced no Mg 2+ effect, suggesting an allosteric mode of action.
In conclusion, the thienopyrimidinone skeleton is confirmed as a privileged framework in HIV drug discovery.Furthermore, as catechol may be implicated in cytotoxicity, we also explored its substitution with other heteroaromatic scaffolds.In particular, our findings indicated that variation at C-2 position may be a key factor for compound selectivity.In this regard, compounds 21-23 and 25-26 with an improved activity were revealed to be a good hit to lead candidates.In particular, compound 26 showed very good activity on both RNase H and IN and is proposed as a dual-inhibitor, while 5,6-dimethyl-2-(5-nitrothiophen-2-yl)thieno [2,3-d]pyrimidin-4(3H)-one 25, which was completely inactive on RNase H but with a good IN IC 50 , was the most selective against IN.
Substitutions at C-5 and C-6 also proved to be vital.In this regard, the crucial introduction of a branched cyclohexyl moiety instead of a cycloheptane ring offered the 2-(3,4-dihydroxyphenyl)-5-methyl-5,6,7,8-tetrahydrobenzo [4,5]thieno [2,3-d]pyrimidin-4(3H)-one 20 as a hit compound for further exploitation.Encouraged by the results, novel derivatives are now under consideration, also using an in silico approach, to gain more insight into the allosteric binding site (manuscript in preparation).

Figure 1 .
Figure 1.Examples of chemical scaffolds obtained from high throughput screening approach.

Figure 1 .
Figure 1.Examples of chemical scaffolds obtained from high throughput screening approach.

Figure 2 .
Figure 2. Chemical structure of Kuwanon-L and of most active thienopyrimidinones and quinazolinones.

Figure 2 .
Figure 2. Chemical structure of Kuwanon-L and of most active thienopyrimidinones and quinazolinones.

Figure 4 .
Figure 4.Chemical structures of thienopyrimidinones tested for their effect on the IN LEDGF/p75dependent activity.

Figure 4 .
Figure 4.Chemical structures of thienopyrimidinones tested for their effect on the IN LEDGF/p75dependent activity.

9 OHFigure 4 .
Figure 4.Chemical structures of thienopyrimidinones tested for their effect on the IN LEDGF/p75dependent activity.

Figure 5 .
Figure 5.Chemical structure of dihydroxyphenyl thienopyrimidinones with the best IN inhibitory activity.

Figure 5 .
Figure 5.Chemical structure of dihydroxyphenyl thienopyrimidinones with the best IN inhibitory activity.

19 ICFigure 6 .
Figure 6.Effect of the replacement of cyclohepthyl ring with a branched cyclohexane on IN inhibitory activity.
pyrimidin-4(3H)-one 14 displayed the best IC 50 value.Subsequently, the thiophene intermediate reacted via direct oxidative condensation with the proper aldehyde in the presence of molecular iodine, to give rise to the final products 10-12, 19-22, 24, 25, 27 and 28.

Table 3 .
Effect of the thiophene ring substitution on the HIV-1 IN LEDGF/p75-dependent activity.
Figure 6.Effect of the replacement of cyclohepthyl ring with a branched cyclohexane on IN inhibitory activity.

Table 4 )
. Compounds 21-23 and 26 inhibited both IN and RNase-H activity, with 22 and 23 being more selective

Table 3 .
Effect of the thiophene ring substitution on the HIV-1 IN LEDGF/p75-dependent activity.
Figure 6.Effect of the replacement of cyclohepthyl ring with a branched cyclohexane on IN inhibitory activity.

Table 3 .
Effect of the thiophene ring substitution on the HIV-1 IN LEDGF/p75-dependent activity.

Table 3 .
Effect of the thiophene ring substitution on the HIV-1 IN LEDGF/p75-dependent activity.
Figure 6.Effect of the replacement of cyclohepthyl ring with a branched cyclohexane on IN inhibitory activity.

Table 3 .
Effect of the thiophene ring substitution on the HIV-1 IN LEDGF/p75-dependent activity.
Figure 6.Effect of the replacement of cyclohepthyl ring with a branched cyclohexane on IN inhibitory activity.