3-Hydroxypyrimidine-2,4-diones as an Inhibitor Scaffold of HIV Integrase

Abstract

Integrase (IN) represents a clinically validated target for the development of antivirals against human immunodeficiency virus (HIV). Inhibitors with a novel structure core are essential for combating resistance associated with known IN inhibitors (INIs). We have previously disclosed a novel dual inhibitor scaffold of HIV IN and reverse transcriptase (RT). Here we report the complete structure-activity relationship (SAR), molecular modeling and resistance profile of this inhibitor type on IN inhibition. These studies support an antiviral mechanism of dual inhibition against both IN and RT and validate 3-hydroxypyrimidine-2,4-diones as an IN inhibitor scaffold.

Introduction

HIV infection and the establishment of proviral latency depend critically on the insertion of viral DNA into host genome.^1–3 This integration process comprises two distinct steps: 3′ processing (3′P) in which a sequence-specific endonucleolytic cleavage of the viral DNA generates a CA-OH at the 3′ end; and strand transfer (ST) where the processed viral DNA is inserted into the host genome through a transesterification reaction using the 3′ CA-OH as the nucleophile.⁴ The virally encoded enzyme IN catalyzes both reactions; thus represents an important target for HIV chemotherapeutic intervention. IN as an antiviral target is rendered particular attractive by the lack of host cellular counterpart. On the other hand, IN uses the same active site (DD35E motif) for both steps where either a viral DNA (in 3′P) or a cellular DNA (in ST) serves as the endogenous substrate. Therefore, inhibitors of IN could also benefit from a potentially high genetic barrier to resistance development. In the past two decades, IN-targeted antiviral research has led to the identification of various inhibitor types, of which the most prominent scaffolds all feature a diketoacid (DKA) functionality or its heterocyclic bioisosteres.^5–9 These chemotypes can accommodate highly selective ST inhibition and have played an essential role in unraveling the pharmacophore requirements for IN binding.^10–15 Extensive research on these chemotypes culminated with a number of investigational drugs (Figure 1), including raltegravir (1)^{16, 17}, the first and only INI presently approved for clinical use; elvitegravir (2)¹⁸, a promising candidate in phase III development; and S/GSK-1349572 (3)^{19, 20}, a second-generation IN inhibitor in phase IIb trials.²¹ In the meantime, HIV integration also involves a cellular co-factor LEDGF/p75.^22–24 Targeting the protein-protein interaction between IN and LEDGF/p75 represents an emerging approach to HIV chemotherapeutic intervention.²⁵ Collectively inhibitors targeting IN can substantially enhance the current highly active anti-retroviral therapy (HAART), the standard chemotherapy for HIV/AIDS.²⁶ However, mutations conferring resistance to these inhibitors have been generated in cell culture and emerged in clinical studies^27–29, suggesting that INIs will be confronted with the same resistance issue that has plagued HIV/AIDS chemotherapy. Continued search for structurally novel INIs to combat resistance is therefore urgently needed. Previously we have reported that N-3 hydroxylation of pyrimidine-2,4-diones yielded inhibitors dually active against HIV reverse transcriptase (RT) and IN (4, figure 1).³⁰ Since this scaffold represents a novel chemotype for INIs, further SAR efforts on IN inhibition are warranted.

Figure 1.

Structures of best known INIs (1–3) and a novel INI (4).

HIV IN is a 32-kDa protein encoded by viral pol gene consisting of three functional domains³¹: the N-terminal domain (NTD) with a conserved “HH-CC” zinc-binding motif; the catalytic core domain (CCD) containing the key D64-D116-E152 catalytic triad; and the C-terminal domain (CTD) important for DNA binding. Crystal structures of each single domain as well as double domains have been resolved^32–34, albeit without the binding of viral DNA substrate. Detailed understanding on HIV IN catalysis requires structural information on a DNA-bound full-length IN complex, or an intasome, which currently remains elusive due to major technical barriers³⁵ in obtaining stable HIV intasome. Recently disclosed crystal structures of a homologous prototype foamy virus (PFV) intasome^{36, 37} have revealed the requisite catalytic molecular network comprising IN, viral and/or host DNA substrates and the inhibitor, allowing a largely improved understanding on the mechanism of catalysis and inhibition of retroviral IN. Homologous models^{38, 39} constructed based on these PFV crystal structures have provided valuable details into HIV IN mechanism of action and formed the basis for structure-based INI design. Significantly these models corroborate the minimal pharmacophore embedded in major INIs: a chelating triad capable of binding two Mg²⁺ ions, and a hydrophobic benzyl moiety. Our novel IN inhibitors are constructed around a pyrimidine-2,4-dione ring, a pharmacophore core that has found wide applications in medicinal chemistry.^40–43 Central to our design is to introduce an OH group on the N-3 position which, along with the two flanking carbonyl groups, would yield a chelating triad of C(O)N(OH)C(O) that mimics a DKA functionality (Figure 2, chelating domain). On the other hand, the hydrophobic benzyl moiety can be easily installed via an N-1 alkylation, and the mono- or di-halogenation of this terminal benzyl group will allow us to explore the halogen effect IN inhibition (Figure 2, aromatic domain). Furthermore, to ensure that this benzyl group can fit in the hydrophobic cavity, its spatial placement over the chelating triad has to be optimized. This can be achieved through a linker of different length (Figure 2, linker domain). Finally, the overall hydrophobicity of the inhibitor could be modified by altering the C5 and C6 substituents (Figure 2, hydrophobic domain).

Figure 2.

SAR design of lead compound 4. Major structural domains are highlighted.

Results and Discussion

Chemistry

The chemical synthesis of aromatic and chelating domains starts with a known 5,6-disubstituted pyrimidine-2,4-dione 5^{44, 45} (Scheme 1). Subsequent selective N-1 alkylation of 5 via a bis-silylated pyrimidine intermediate with an in situ generated chloromethyl ether allows the incorporation of the aromatic domain with an ether linkage (6–10, Scheme 1). As previously reported ³⁰, the key N-3 hydroxylation was achieved via a base-mediated meta-chloroperbenzoic acid (mCPBA) oxidation. To assess the significance of this hydroxyl group on IN inhibition, an N-3 amino analogue (15) as well as analogues with an N-3 methyl or methoxy group were also synthesized (10 and 14, Scheme 1).

Scheme 1.

^a Synthesis of aromatic and chelating domains

^aReagents and conditions: a) CH₃C(OTMS)=NTMS (BSA), substituted benzyl alcohol, (HCHO)_n, TMSCl, TBAI ( cat.), CH₂Cl₂, rt, 70–90%; b) Cs₂CO₃, THF, MeI, 40 °C, 53–63%; c) NaH, m-CPBA, THF, rt, 53–85%; d) NaH, MSH, THF, rt, 54%.

The synthesis of the linker domain is described in Scheme 2. Analogues with an ether linkage were easily prepared through a selective N-1 alkylation with a chloromethyl ether (Scheme 2, B). The requisite alcohols (19–21) were synthesized from commercial carboxylic acids via a direct reduction with LAH. The construction of all-carbon tethers in the linker domain, however, proved much more challenging. The BSA-assisted selective N-1 alkylation of pyrimidine-2,4-diones generally involves a chloromethyl ether as the alkylating agent. Same reaction with benzyl bromide was effected only under forcing conditions (microwave, high temperature and long reaction time; see Scheme 2, C). Generic alkyl halides fail to undergo alkylation with the bis-silylated pyrimidine intermediate, and base-mediated direct alkylation normally leads to a mixture of region-isomers. As it turned out, the introduction of all-carbon chain into N-1 requires multiple protection and deprotection steps (Scheme 2, D). In this event, the N-1 position was first protected with an easily introduced ethoxymethyl group, which allowed the protection of N-3 position with an acetophenone group. The N-1 ether linkage was then cleaved under acidic condition to yield the N-3 protected intermediate 35. Base-mediated direct alkylation led to the incorporation of the carbon tethers followed by a Mg mediated N-3 deprotection to yield key N-1 alkylated intermediates 39–41. All N-1 alkylated pyrimidine-2,4-diones were then oxidized with mCPBA to produce desired N-3 hydroxylated compounds (Scheme 2).

Scheme 2.

^a Synthesis of linker domain

^aReagents and conditions: a) LiAlH₄, THF, rt, 67–72%; b) CBr₄, PPh₃, THF, rt, 73–80%; c) CH₃C(OTMS)=NTMS (BSA), CH₂Cl₂; 19–21, (HCHO)_n, TMSCl, TBAI ( cat.), rt, 83–89%; d) NaH, m-CPBA, THF, rt, 62–72%; e) BSA, benzyl bromide, chlorobenzene, microwave, 160 °C, 50%; f) BSA, chloromethyl ethyl ether, TBAI ( cat.), CH₂Cl₂, rt, 89%; g) K₂CO₃, PhCOCH₂Br, MeCN, reflux, 84%; h) TFA:H₂O (9:1), reflux, 76%; i) Cs₂CO₃, 4-F-Ph(CH₂)_nBr, DMF, 80 °C, 31–63%; j) Mg, AcOH, MeOH, rt, 48–57%.

The synthesis of hydrophobic domain with variable C-5 and C-6 substituents (Scheme 3, A) was achieved through a similar N-3 hydroxylation of N-1 alkylated intermediates (45–50 and 57–63). Compounds with a slightly different scaffold, the 5,6-benzopyrimidine-2,4-diones (75–76), were also prepared to probe the effect of a fused benzene ring on IN inhibition. These analogues were synthesized via a route depicted in Scheme 3, C.

Scheme 3.

^a Synthesis of hydrophobic domain

^aReagents and conditions: a) NaH, m-CPBA, THF, rt 32–72%; b) (i) CDI/toluene, reflux; then NH₂OBn, reflux; (ii) NaOH/EtOH, reflux, 53%; c) benzyl bromide or 4-Flurobenzyl bromide, Cs₂CO₃, DMF, 80 °C, 58–67%; d) HBr (48% in H₂O), reflux, 77–78%.

SAR on IN Inhibition

All synthetic analogues were first tested against recombinant HIV-1 IN using a gel assay. Both 3′P and ST functions were evaluated. As clearly shown in Tables 1–3, our scaffold tends to selectively inhibit ST as none of the analogues demonstrated inhibitory activity against 3′P.

Table 1.

Effects of aromatic and chelating domains on IN inhibition

Compd	R1	R2	IN IC50 (μM)a
			3′P	STb
11c	H	OH	>111	21 ± 2
12	2,4-F	OH	>111	4.8 ± 0.7
13	3-Cl-2-F	OH	>111	2.3 ± 0.6
4c	4-F	OH	>111	3.5 ± 0.6
9	4-F	H	>111	>111
10	4-F	Me	>111	>111
14	4-F	OMe	>111	>111
15c	H	NH2	>111	>111

^aConcentration inhibiting enzymatic function by 50%.

^bMean value ± standard deviation from triplicate experiments.

^cRef ³⁰.

Table 3.

Effects of hydrophobic domain on IN inhibition

Compd	R1	linker	R2	R3	IN IC50 (μM)a
					3′P	STb
11c	H	CH2OCH2	iPr	benzyl	>111	21 ± 2
51	H	CH2OCH2	iPr	4-Fbenzyl	>111	7.3 ± 0.8
52	F	CH2OCH2	iPr	4-Fbenzyl	>111	5.5 ± 0.7
53	H	CH2OCH2	Et	benzyl	>111	8.2 ± 1.0
54	H	CH2OCH2	Me	H	>111	75 ± 7
55	H	CH2OCH2	benzyl	Et	>111	36 ± 6
56	F	CH2OCH2	benzyl	Et	>111	11 ± 0.8
64	H	CH2	iPr	4-Fbenzyl	>111	32 ± 6
65	F	CH2	iPr	4-Fbenzyl	>111	73 ± 12
66	H	CH2	iPr	benzyl	>111	28 ± 2
67	H	CH2	Et	benzyl	>111	>111
68	F	CH2	Et	benzyl	>111	24 ± 3
69	H	CH2	Me	H	>111	>111
70	F	CH2	Me	H	>111	>111
75	H	CH2	benzo (fused)		>111	>111
76	F	CH2	benzo (fused)		>111	>111

^aConcentration inhibiting enzymatic function by 50%.

^bMean value ± standard deviation from triplicate experiments.

^cRef ³⁰.

Aromatic and Chelating Domains

We have shown previously³⁰ that removing the phenyl ring at the end of the N-1 pendant resulted in a significant loss of inhibition against IN, and that without the OH group the N(3)H analogue is completely inactive against IN. These findings corroborate the pharmacophore hypothesis that HIV-1 IN inhibition requires minimally a chelating domain and an aromatic domain. In this study, additional analogues were synthesized to further study these two domains. As shown in Table 1 and Figure 3, halogen substitution of the N-1 benzyl seems to significantly benefit IN binding. A fluorination at the para position (inhibitor 4) leads to a 6-fold improvement in IN inhibition over the unsubstituted compound 11. Further improvement in activity was only marginal with a 3-chloro-2-fluoro substituted analogue 13, whereas a 2,4-difluoro compound 12 showed slightly reduced inhibition when compared to 4 (Table 1). On the other hand, replacing the N-3 OH group with H, Me, OMe or even NH₂ yielded compounds completely devoid of inhibitory activity against IN (9–10 and 14–15, Table 1), confirming the critical requirement of the N-3 OH for chelating Mg²⁺. The lack of potency with the amino analogue 15 also reflects an inferior chelating ability of NH₂ for Mg²⁺ when compared to OH.³⁰

Figure 3.

Dose response curves of compounds 4, 9–10 and 12–14.

The linker domain

Having established the chelating triad and the N-1 benzyl group as the two major structural determinants of our novel INI scaffold, our SAR efforts were then centered around the linker domain as it dictates the angle and distance between the chelator and benzyl group. Toward this end, we have designed linkers of different nature and length. Compounds 4 and 28–30 all have an ether linkage with incremental methylene groups (Table 2). Interestingly, these analogues all exhibited inhibitory activity in ST assay at low micromolar range, and demonstrated a discernible binding dependence on the length of the linker. As shown in Table 2, the 4-atom linker (CH₂)₂OCH₂ (inhibitor 28) conferred the highest inhibition within this series, and a reduced inhibition was achieved with compounds having a shorter (4) or longer linker (29–30). A similar albeit more sensitive trend was observed with compounds having an all carbon N-1 linker. In this case, short linkages, such as CH₂ (32) and (CH₂)₂ (42), appeared to pose severe binding difficulties while the 3- and 4-atom linkers (CH₂)₃ (43) and (CH₂)₄ (44) conferred a drastically improved inhibition against ST (Table 2). These findings strongly imply that the optimal IN inhibition entails a linker of 3–4 atoms. The nature of the linker (ether linkage vs all-carbon linkage) was also found to substantially impact IN inhibition as inhibitors 43 and 44 demonstrated significantly better inhibition than compound 28. Detailed docking studies on the inhibitor binding may provide a molecular basis for this observation.

Table 2.

Effects of linker domain on IN inhibition

Compd	Linker	IN IC50 (μM)a
		3′P	STb
4c	CH2OCH2	>111	3.5 ± 0.6
28	(CH2)2OCH2	>111	1.6 ± 0.2
29	(CH2)3OCH2	>111	3.2 ± 0.5
30	(CH2)4OCH2	>111	4.5 ± 0.4
32	CH2	>111	22 ± 0.5
42	(CH2)2	>111	46 ± 6
43	(CH2)3	>111	0.44 ± 0.1
44	(CH2)4	>111	0.56 ± 0.04

^aConcentration inhibiting enzymatic function by 50%.

^bMean value ± standard deviation from triplicate experiments.

^cRef ³⁰.

The hydrophobic domain

Synthetic analogues in these domains vary mainly in the size of the C-5 alkyl group (iPr, Et and Me) and fluorine substitution on the C-6 benzyl. These changes impact most likely the steric hindrance and the overall hydrophobicity of these inhibitors. Interestingly, a C-5 ethyl group was found to confer significantly better inhibition than isopropyl group (53 vs 11, Table 3), indicating that less steric hindrance at C-5 could benefit IN binding. However, a reversed effect was observed in compounds with an N-1 linker of CH₂, wherein a C-5 ethyl substitution (67) caused a complete loss of inhibition when compared to a C-5 isopropyl substitution (66). On the other hand, swapping C5 and C6 substituents of inhibitor 53 seems to harm IN binding (55 vs 53) as the IC₅₀ value increased by 4-fold. Notably, a para-fluoro substitution at the C-6 benzyl clearly favors IN inhibition (53 vs 11, 56 vs 55). With a more dramatic change, a C-5 and C-6 fused benzene ring was found to render compounds 75–76 completely inactive against IN inhibition (Table 3). These findings indicate that the hydrophobic domain affects IN binding in a less comprehensible way and changes in this domain impact the activity most likely in concert with the linker domain.

Modeling

The common mode of inhibitory binding was explored through molecular docking using a homologous HIV IN model³⁸ constructed based on the crystal structure of prototype foamy virus (PFV) IN.³⁶ Significantly, the new inhibitor 43 fits perfectly into the IN binding site through two major binding domains also featured in raltegravir (1): the chelation of two Mg²⁺ ions, and the placement of the 4-F-benzyl group into the protein-DNA interfacial hydrophobic pocket (Figure 4). The high degree of similarity shown between the binding of 1 and 43 confirms that our new molecular scaffold can effectively engage with HIV IN. Interestingly, while residue Y143 contributes siginificantly to binding of 1 through a π-stacking interaction (Figure 4, A), the same interaction was not observed with our inhibitor 43 (Figure 4, B), suggesting that our inhibitor could be less susceptible to resistance associated with Y143 mutation. It is also noteworthy that the linker domain is situated right in the hydrophobic cleft between residue P145 and viral DNA nucleobases. Therefore, an ether linkage may provide a less favorable hydrophobic interaction than an all-carbon linker, thus the observed reduced inhibition with the ether linkage (Table 2, 43 Vs 28).

Figure 4.

Docking of raltegravir (1, A) and inhibitor 43 (B) in HIV-1 IN CCD in complex with Mg²⁺ and DNA. The 3-N hydroxyl group simultaneously chelates to both Mg²⁺ ions while allowing the placement of the benzyl group into the protein-DNA interfacial hydrophobic pocket.

Further validation was achieved by correlating the SAR of docked inhibitors 4, 11, 28, 29, 30, 32, 42, 43, 44, 66 as shown in Figure 5. The outstanding correlation observed between these two sets of IC₅₀ data (exp. vs calc., correlation coefficiency = 0.82) supports our docking model with a common mode of binding as a valid platform for HIV IN inhibitor design.

Figure 5.

Strong correlation of docked 3-hydroxypyrimidine-2,4,-diones analogues supports a common mode of binding for HIV IN STI design

Antiviral Assay

To evaluate the antiviral activity of newly synthesized INIs, all compounds with an IC₅₀ of 10 μM or lower from IN ST assay were tested against HIV-1IIIB in CEM-SS cells using an assay based on viral cytopathic effect (CPE). In this assay, the reduction of CPE was used to indicate the antiviral activity of a compound. All selected compounds were first screened under a single concentration (10 μM). Inhibitors with excellent CPE reduction and cell viability from this screening assay were further tested in a dose response fashion, from which the EC₅₀ and CC₅₀ were determined. Table 4 summarizes the antiviral testing results.

Table 4.

Anti-HIV-1 activity of IN inhibitors

Compd	Single Concentration (10 μM)		Dose Response
	CPE Reduction (%)a	Cell Viability (%)	EC50 (μM)b	CC50 (μM)c	TId
4e	100	100	0.024	>20	>833
11e	100	100	0.0080	>20	>2,500
12	100	100	2.1	>100	>47
13	100	100	0.95	>100	>105
28	27	83	NDf	ND	--
29	30	92	ND	ND	--
30	12	88	ND	ND	--
43	12	97	ND	ND	--
44	6.0	94	ND	ND	--
49	48	100	ND	ND	--
50	10	92	ND	ND	--
51	100	100	1.1	>20	>18

^aReduction of viral cytopathic effect, indicating inhibitory activity of a compound.

^bConcentration inhibiting virus replication by 50%.

^cConcentration resulting in 50% cell death.

^dTherapeutic index, defined by CC₅₀/EC₅₀.

^eRef ³⁰.

^fNot determined.

Remarkably, compounds 43 and 44, the two most potent INIs from IN ST assay with submicromolar IC₅₀s, have shown only marginal activity against HIV-1 at 10 μM (Table 4). This finding strongly suggests that despite their ability to confer exceptional IN inhibition in biochemical assays, all-carbon linkers lack the proper physicochemical properties for the compounds to reach their target in cell culture. By contrast, with the exception of compounds having a para-fluorine at the C-6 benzyl group (49–50, Table 4), INIs with an ether linkage of CH₂OCH₂ (4, 11–13 and 51) showed antiviral activity at the low micromolar to low nanomolar range without appreciable cytotoxicity (Table 4). Longer ether linkages that produced excellent anti-IN activity for compounds 28–30 did not generate similar activity in the antiviral assay, further demonstrating that additional CH₂s in the linker domain hinder antiviral activity.

Resistance

To establish the resistance profile of our inhibitors, we first tested two representative inhibitors 4 and 13 against HIV IN mutants. In this study, major mutants associated with raltegravir resistance^{23, 24}, namely a G140S/Q148H double mutant and two single mutants Y143R and N155H were employed. The results are summarized in Table 5. Notably the G140S/Q148H double mutant which causes high resistance to raltegravir (1) yielded a much lower fold resistance to 4 and 13, whereas the two single mutants confer comparable resistance to 4, 13 and 1, suggesting a potentially common binding mode for these inhibitors.

Table 5.

Fold-resistance of inhibitors 4 and 13 against IN mutants

IN Mutants	Fold-resistancea, b
	4c	13	1
G140S/Q148H	10/13	14/17	148/139
Y143R	7/4	7/5	24/28
N155H	5/6	5/7	10/8

^aFold-resistance is defined by IC_50(mutant)/IC_{50 (WT)}.

^bValues from two independent experiments.

^cRef ³⁰.

Based on these biochemical data, inhibitors 4 and 13 were further evaluated in cell culture against raltegravir-resistant HIV strains. Unlike the initial antiviral assay wherein HIV-1 IIIB was used in CEM-SS cells, the resistance studies were carried out with NL4-3 based viral strains in MT-4 cells. Selected strains contain major mutations corresponding to the IN mutants (Table 6 Vs Table 5). As shown in Table 6, all three viral strains show a high level resistance to raltegravir (1). Interestingly, although likely within experimental variations, the small fold-resistances observed with inhibitor 4 (Table 6) correlate closely with resistance observed with IN mutants in biochemical assays (Table 5). The testing of inhibitor 13 was complicated by its toxicity observed in MT-4 cells. Notably the same compound did not show toxicity in CEM-SS cells (TC₅₀ > 100 μM, Table 4). On the other hand, antiviral assay with two HIV-1_NL4-3 strains containing major mutations associated with NNRTI resistance, K103N or Y181C, revealed a fold-resistance of 16 and 48, respectively, to inhibitor 4.⁴⁶ These results suggest that the overall antiviral activity of 4 could reflect a concerted yet not necessarily balanced inhibition to both RT and IN.

Table 6.

Fold-resistance of inhibitors 4 and 13 against raltegravir-resistant HIV-1 strains

Viruses	Major Mutations	Fold-resistancea
		4b	13c	1
8070-2	G140S/Q148H	7.3	>2.0	>855
1556-1	Y143C	5.9	>1.7	>855
4736-2	N155H	3.1	>5.1	149

^aFold-resistance is defined by EC_50(mutant)/EC_{50 (WT)}.

^bRef ³⁰.

^cToxicity observed (TC₅₀ = 2.0 μM).

In summary, through complete SAR and docking studies we have demonstrated that the N-3 OH group and the 4-F-benzyl moiety on the N-1 side chain are the key structural determinants for IN inhibition, and that the linker domain sits in a hydrophobic cleft between DNA nucleobases and P145. Therefore, both the nature and the length of the linker greatly impact IN binding, with the optimal IN inhibition entailing a 3–4 atom all-carbon linkage. However, studies in cell culture revealed that inhibitors with an all-carbon linker generally lack antiviral activity, presumably due to unfavored physicochemical properties. Resistance studies with raltegravir-resistant IN mutants as well as HIV-1_NL4-3 viral strains demonstrated low level resistances toward inhibitors 13 and 4, likely due to their anti-RT activity. In the end, compounds (4, 11 and 13) featuring a CH₂OCH₂ linker demonstrate low micromolar IN inhibition and nanomolar anti-HIV activity, and are suitable for further development.

Experimental

Chemistry

General Procedures

All commercial chemicals were used as supplied unless otherwise indicated. Dry solvents (THF, Et₂O, CH₂Cl₂ and DMF) were dispensed under argon from an anhydrous solvent system with two packed columns of neutral alumina or molecular sieves. Flash chromatography was performed on a Teledyne Combiflash RF-200 with RediSep columns (silica) and indicated mobile phase. All reactions were performed under inert atmosphere of ultra-pure argon with oven-dried glassware. ¹H and ¹³C NMR spectra were recorded on a Varian 600 MHz spectrometer. Mass data were acquired on an Agilent TOF II TOS/MS spectrometer capable of ESI and APCI ion sources. Analysis of sample purity was performed on a Varian Prepstar SD-1 HPLC system with a Varian Microsorb-MW 100-5 C18 column (250mm × 4.6 mm): solvent A = H₂O, solvent B = MeCN; flow rate = 1.0 mL/min; Method: linear 30–95% (B) over 25 min. All tested compounds have a purity ≥ 96%.

6-Benzyl-1-(benzyloxymethyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (6).⁴⁴

To a suspension of paraformaldehyde (144 mg, 4.80 mmol) in TMSCl (2.0 mL), was added benzyl alcohol (518 mg, 4.80 mmol) at room temperature. The reaction mixture was stirred until a clear solution was formed. The solution was concentrated under reduced pressure to give benzyl chloromethyl ether as oil, which was taken to the next step without further purification. To a suspension of pyrimidine 5 (976 mg 4.00 mmol) in 16 mL of anhydrous DCM was added BSA (2.16 mL, 8.80 mmol) at room temperature. The resulting mixture was stirred until a clear solution was achieved. The freshly prepared benzyl chloromethy ether was added to this solution followed by the addition of a catalytic amount of TBAI. The reaction mixture was kept overnight and then quenched by adding a saturated aqueous solution of NaHCO_3. The aqueous phase was extracted with CH₂Cl₂ (20 mL × 3). The combined organic extracts were dried over Na₂SO₄ and concentrated under reduced pressure. The resultant residue was subjected to combiflash (Hex/EtOAc, 2:1) to afford compound 6 (1.25 g, 86 %) as a white solid. ¹H NMR (600 MHz, CDCl₃) δ 7.28 (m, 6H), 7.20 (m, 2H), 7.00 (d, J = 7.8 Hz, 2H), 5.14 (s, 2H), 4.59 (s, 2H), 4.10 (s, 2H), 2.79 (septet, J = 6.6 Hz, 1H), 1.21 (d, J = 6.6 Hz, 6H); MS (ESI+) m/z: 365.17 (M+1).

6-Benzyl-1-((2,4-difluorobenzyloxy)methyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (7)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 88 %; ¹HNMR (600 MHz, CDCl₃) δ 9.32 (s, 1H), 7.34 (m, 3H), 7.26 (t, J = 7.8 Hz, 1H), 7.07 (d, J = 7.8 Hz, 2H), 6.86 (t, J = 9.0 Hz, 1H), 6.81 (d, J = 9.0 Hz, 1H), 5.20 (s, 2H), 4.64 (s, 2H), 4.16 (s, 2H), 2.87 (m, 1H), 1.28 (d, J = 6.6 Hz, 6H); MS (ESI−) m/z: 399.42 (M-1).

6-Benzyl-1-((3-chloro-2-fluorobenzyloxy)methyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (8)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 90 %; ¹ HNMR (600 MHz, CDCl₃) δ 8.96 (b, 1H), 7.36 (m, 3H), 7.33 (t, J = 7.8 Hz, 2H), 7.10 (m, 3H), 5.21 (s, 2H), 4.69 (s, 2H), 4.16 (s, 2H), 2.87 (m, 1H), 1.28 (d, J = 6.6 Hz, 6H); MS (ESI−) m/z: 415.87 (M-1).

6-Benzyl-1-((4-fluorobenzyloxy)methyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (9)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 89 %; ¹HNMR (600 MHz, CDCl₃) δ 8.97 (b, 1H), 7.34 (t, J = 7.2 Hz, 2H), 7.30 (m, 3H), 7.07 (d, J = 7.8 Hz, 2H), 7.03 (m, 2H), 5.20 (s, 2H), 4.61 (s, 2H), 4.16 (s, 2H), 2.88 (septet, J = 7.2 Hz, 1H), 1.28 (d, J = 7.2 Hz, 6H); MS (ESI+) m/z: 383.18 (M+1).

6-Benzyl-1-((4-fluorophenethoxy)methyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (25)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 89 %; ¹HNMR (600 MHz, CDCl₃) δ 8.89 (b, 1H), 7.27 (t, J = 7.2 Hz, 2H), 7.20 (m, 1H), 7.08(m, 2H), 6.99 (d, J = 7.8 Hz, 2H), 6.91 (t, J = 8.4 Hz, 2H), 5.04 (s, 2H), 3.94 (s, 2H), 3.71 (t, J = 6.6 Hz, 2H), 2.74 (m, 3H), 1.20 (d, J = 6.6 Hz, 6H); MS (ESI−) m/z: 395.45 (M-1).

6-Benzyl-1-((3-(4-fluorophenyl)propoxy)methyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (26)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 83 %; ¹HNMR (600 MHz, CDCl₃) δ 8.97 (b, 1H), 7.29 (t, J = 7.8 Hz, 2H), 7.21 (t, J = 7.8 Hz, 1H), 7.05 (m, 4H), 6.89 (t, J = 9.0 Hz, 2H), 5.05 (s, 2H), 4.11 (s, 2H), 3.50 (t, J = 6.0 Hz, 2H), 2.87 (septet, J = 7.2 Hz, 1H), 2.56 (t, J = 7.8 Hz, 2H), 1.78 (m, 2H), 1.23 (d, J = 6.6 Hz, 6H); MS (ESI−) m/z: 409 (M-1).

6-Benzyl-1-((4-(4-fluorophenyl)butoxy)methyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (27)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 84 %; ¹HNMR (600 MHz, CDCl₃) δ 10.01 (b, 1H), 7.34 (t, J = 7.8 Hz, 2H), 7.26 (t, J = 7.2 Hz, 1H), 7.10 (m, 4H), 6.94 (d, J = 9.0 Hz, 2H), 5.23 (s, 2H), 4.16 (s, 2H), 3.58 (t, J = 6.6 Hz, 2H), 2.86 (septet, J = 7.2 Hz, 1H), 2.58 (t, J = 7.2 Hz, 2H), 1.63 (m, 2H), 1.56 (m, 2H), 1.27 (d, J = 7.2 Hz, 6H); MS (ESI−) m/z: 423.51 (M-1).

6-Benzyl-1-(ethoxymethyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (33)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 89 %; ¹HNMR (600 MHz, CDCl₃) δ 8.68 (b, 1H), 7.29 (t, J = 7.8 Hz, 2 H), 7.22 (m, 2H), 7.06 (d, J = 7.8 Hz, 1H), 5.05 (s, 2H), 4.11 (s, 2H), 3.56 (q, J = 7.2 Hz, 2H), 2.81 (septet, J = 7.2 Hz, 1H), 1.22 (d, J = 6.6 Hz, 6H), 1.11 (t, J = 7.2 Hz, 3H); MS (ESI−) m/z: 301.37 (M-1).

1-(Benzyloxymethyl)-5-methylpyrimidine-2,4(1H,3H)-dione (48)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 82 %; ¹HNMR (600 MHz, CDCl₃) δ 9.21 (b, 1H), 7.34 (m, 5H), 7.10 (s, 1H), 5.21 (s, 2H), 4.66 (s, 2H), 4.62 (s, 2H), 1.91 (s, 3H); MS (ESI−) m/z: 245.10 (M-1).

5-Benzyl-1-(benzyloxymethyl)-6-ethylpyrimidine-2,4(1H,3H)-dione (49)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 78 %; ¹HNMR (600 MHz, CDCl₃) δ 9.64 (s, 1H), 7.31 (m, 4H), 7.24 (m, 3H), 6.99 (m, 3H), 5.43 (s, 2H), 4.67 (s, 2H), 4.08 (s, 2H), 2.73 (q, J = 6.6 Hz, 2H), 1.06 (t, J = 6.6 Hz, 3H); MS (ESI−) m/z: 349.62 (M-1).

5-Benzyl-6-ethyl-1-((4-fluorobenzyloxy)methyl)pyrimidine-2,4(1H,3H)-dione (50)

This compound was prepared as a white solid following the procedure described for the preparation of 6. Yield: 80 %; ¹ HNMR (600 MHz, CDCl₃) δ 9.51 (s, 1H), 7.29 (t, J = 7.2 Hz, 4H), 7.20 (d, J = 6.6 Hz, 2H), 7.00 (t, J = 6.6 Hz, 2H), 5.41 (s, 2H), 4.62 (s, 2H), 4.08 (s, 2H), 2.72 (q, J = 6.6 Hz, 2H), 1.07 (t, J = 6.6 Hz, 3H); MS (ESI−) m/z: 367.15 (M-1).

6-Benzyl-1-((4-fluorobenzyloxy)methyl)-5-isopropyl-3-methylpyrimidine-2,4(1H,3H)-dione (10)

A mixture of compound 9 (60 mg, 0.16 mmol), methyl iodide (27 mg, 0.19 mmol) and Cs₂CO₃ (61 mg, 0.19 mmol) in 2 mL of THF was stirred under 40 °C for 3 h, and then allowed to cool to room temperature. The reaction mixture was quenched by adding 5 mL of H₂O and extracted with EtOAc (10 mL × 3). The combined organic phases were dried over Na₂SO₄, and then concentrated under reduced pressure. The resultant residue was purified with combiflash to give compound 10 (40 mg, 63%) as clean oil: ¹HNMR (600 MHz, CDCl₃) δ 7.25 (t, J = 7.8 Hz, 2H), 7.21 (m, 3H), 6.99 (d, J = 7.2 Hz, 2H), 6.95 (t, J = 8.4 Hz, 2H), 5.16 (s, 2H), 4.53 (s, 2H), 4.09 (s, 2H), 3.28 (s, 3H), 2.87 (septet, J = 7.2 Hz, 1H), 1.22 (d, J = 7.2 Hz, 6H); ¹³C NMR (150 MHz, CDCl₃) δ 163.2, 161.8, 161.5, 152.6, 146.0, 135.4, 133.2 (d, J = 3.3 Hz), 129.6 (d, J = 7.8 Hz), 129.2, 127.3 (d, J = 7.4 Hz), 118.9, 115.4 (d, J = 21.2 Hz), 73.7, 71.0, 33.4, 28.2, 27.9, 20.4; MS (ESI−) m/z: 395.45 (M-1).

6-Benzyl-1-((4-fluorobenzyloxy)methyl)-5-isopropyl-3-methoxypyrimidine-2,4(1H,3H)-dione (14)

This compound was prepared following the procedure described for the preparation of 10. Yield: 53 %; ¹HNMR (600 MHz, CDCl₃) δ 7.27 (t, J = 8.4 Hz, 2H), 7.24 (m, 3H), 6.99 (d, J = 7.2 Hz, 2H), 6.96 (t, J = 8.4 Hz, 2H), 5.14 (s, 2H), 4.55 (s, 2H), 4.08 (s, 2H), 3.94 (s, 3H), 2.85 (septet, J = 7.2 Hz, 1H), 1.22 (d, J = 7.2 Hz, 6H); MS (ESI−) m/z: 411.45 (M-1).

3-Amino-6-benzyl-1-(benzyloxymethyl)-5-isopropylpyrimidine-2,4(1H,3H)-dione (15).³⁰

To a solution of 6 (60 mg, 0.16 mmol) in l.0 mL of THF was added NaH (31 mg, 0.82 mmol) at 0 °C. This reaction mixture was allowed to warm to room temperature over 1h., and then cooled to 0 °C followed by the addition of O-(Mesitylsulfonyl)hydroxylamine⁴⁷ (MSH, 66 mg, 0.32 mmol). After stirring for 0.5 h, this reaction mixture was stirred at rt overnight. The reaction was quenched by adding 10 mL of H₂O; the aqueous phase was then extracted with ethyl acetate (10 mL × 3). The combined organic extracts were dried over Na₂SO₄ and concentrated under reduced pressure. The resultant residue was subjected to combiflash (Hex/EtOAc, 1:1) to give compound 15 (22 mg, 54 % based on consumed start material) as a white solid: ¹HNMR (600 MHz, CDCl₃) δ 7.33-7.26 (m, 8H), 7.04-7.03 (d, J = 7.2 Hz, 2H), 5.24 (s, 2H), 4.66 (s, 2H), 4.19 (s, 2H), 2.90 (septet, J = 7.2 Hz, 1H), 1.29 (d, J = 6.6 Hz, 6H); HRMS (ESI−) calcd. for C₂₂H₂₅N₃O₃ [M−H]⁻ 380.1969, found 380.1976 (E = −1.9 ppm).

6-Benzyl-1-((4-fluorobenzyloxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (4).⁴⁷

To a solution of 9 (100 mg, 0.26 mmol) in 6 mL of THF was added NaH (31 mg, 1.30 mmol) at 0 °C. This reaction mixture was allowed to warm to room temperature over 1 h., and then cooled to 0 °C followed by the addition of m-CPBA (135 mg, 0.79 mmol). This reaction mixture was allowed to warm to room temperature and stirred overnight. After the reaction was quenched by adding 10 mL of H₂O, the aqueous phase was acidified with 1N HCl to pH = 7, and then extracted with ethyl acetate (10 mL × 3). The combined organic extracts were dried over Na₂SO₄ and concentrated under reduced pressure. The resultant residue was subjected to combiflash (Hex/EtOAc, 1:1) to give compound 4 (38 mg, 72 % based on consumed start material) as a white solid. ¹HNMR (600 MHz, CDCl₃) δ 7.33-7.26 (m, 5H), 7.04-6.99 (m, 4H), 5.03 (s, 2H), 4.64 (s, 2H), 4.19 (s, 2H), 2.90 (septet, J = 7.2 Hz, 1H), 1.29 (d, J = 6.6 Hz, 6H); ¹³C NMR (150 MHz, CDCl₃) δ 163.3, 162.4, 161.6, 152.1, 148.3, 135.2, 133.1(d, J = 3.3 Hz), 129.7 (d, J = 7.8 Hz), 129.2, 127.2(d, J = 6.8 Hz), 119.9, 115.4 (dd, J = 21.2 Hz), 72.8, 70.9, 33.5, 28.2, 20.4; HRMS (ESI−) calcd. for C₂₂H₂₃FN₂O₄ [M−H]⁻ 399.1715, found 399.1705 (E = 2.4 ppm).

6-Benzyl-1-(benzyloxymethyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (11)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 53 %; ¹HNMR (600 MHz, CDCl₃) δ 7.32-7.30 (m, 7H), 7.02 (J = 7.2 Hz, 2H), 5.30 (s, 2H), 4.67 (s, 2H), 4.19 (s, 2H), 2.90 (septet, J = 7.2 Hz, 1H), 1.28 (d, J = 6.6 Hz, 6H); ¹³C NMR (150 MHz, CDCl₃) δ 162.2, 151.9, 148.4, 137.3, 135.3, 129.1, 128.5, 128.0, 127.8, 127.3, 127.2, 119.8, 73.0, 71.8, 33.5, 28.2, 20.4; HRMS (ESI−) calcd. for C₂₂H₂₄N₂O₄ [M−H]⁻ 381.1809, found 381.1803 (E = 1.5 ppm).

6-Benzyl-1-((2,4-difluorobenzyloxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (12)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 73 %; ¹HNMR (600 MHz, CDCl₃) δ 7.18 (m, 1H). 7.20 (m, 2H), 7.14 (t, J = 7.2 Hz, 1H), 6.97 (d, J = 7.8 Hz, 2H), 6.73 (td, J = 1.8, 9.6 Hz, 1H), 6.67 (td, J = 1.8, 9.6 Hz, 1H), 5.13 (s, 2H), 4.55 (s, 2H), 4.04 (s, 2H), 2.75 (m, 1H), 1.13 (d, J = 6.6 Hz, 6H); ¹³C NMR (150, CDCl₃) δ 163.8 (d, J = 11.7 Hz), 162.1(dd, J = 7.8, 19.5 Hz), 160.4 (d, J = 11.7 Hz), 148.3, 135.2, 131.5 (d, J = 7.8 Hz), 129.2, 127.2, 120.4 (d, J = 15.6 Hz), 120.0, 111.2 (dd, J = 3.9, 21.2 Hz), 104.1 (t, J = 25.2 Hz), 72.7, 64.9, 33.4, 28.3, 20.3; HRMS (ESI−) calcd. for C₂₂H₂₂F₂N₂O₄[M−H]⁻ 415.1475, found 415.1473 (E = 0.45 ppm).

6-Benzyl-1-((3-chloro-2-fluorobenzyloxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (13)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 85 %; ¹HNMR (600 MHz, CD₃OD) δ 7.32 (t, J = 6.6 Hz, 1H), 7.23 (m, 3H), 7.18 (t, J = 7.2 Hz, 1H), 7.05 (t, J = 7.8 Hz, 1H), 7.00 (d, J = 7.8 Hz, 2H), 5.29 (s, 2H), 4.67 (s, 2H), 4.11 (s, 2H), 2.87 (m, 1H), 1.22 (d, J = 7.2 Hz, 6H); ¹³C NMR (150, CDCl₃) 163.7 (d, J = 11.7 Hz), 162.1 (t, J = 11.7 Hz), 160.4 (t, J = 11.7 Hz), 159.7, 144.2, 135.5, 131.8 (d, J = 11.7 Hz), 129.1, 127.3(t, J = 16.2 Hz), 120.5 (d, J = 11.1 Hz), 118.2, 116.6, 111.1 (dd, J = 3.5, 21.3 Hz), 103.9 (t, J = 25.1 Hz), 72.9, 65.5, 33.4, 28.3, 20.4; HRMS (ESI+) calcd. for C₂₂H₂₂ClFN₂O₄ [M−H]⁻ 431.1179, found 431.1181 (E = −0.83 ppm).

6-Benzyl-1-((4-fluorophenethoxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (28)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 66 %; ¹HNMR (600 MHz, CDCl₃) δ 7.34 (t, J = 7.2 Hz, 2H), 7.29 (d, J = 7.2 Hz, 1H), 7.14 (m, 2H), 7.03 (d, J = 7.8 Hz, 2H), 6.97 (t, J = 8.4 Hz, 2H), 5.19 (s, 2H), 4.04 (s, 2H), 3.82 (t, J = 6.6 Hz, 2H), 2.88 (septet, J = 7.2 Hz, 1H), 2.81 (t, J = 6.6 Hz, 2H), 1.28 (d, J = 6.6 Hz, 6H); ¹³C NMR (150 MHz, CDCl₃) δ 162.5 (d, J = 16.2 Hz), 160.7, 155.5, 152.1, 148.4, 135.4, 134.2 (d, J = 2.9 Hz), 130.3 (d, J = 7.8 Hz), 129.2, 127.2, 119.8, 115.2 (d, J = 21.2 Hz), 72.9, 70.0, 35.1, 33.3, 28.2, 20.4; HRMS (ESI+) calcd. for C₂₃H₂₅FN₂O₄ [M+H]⁺ 413.1871, found 413.1868 (E = 0.76 ppm).

6-Benzyl-1-((3-(4-fluorophenyl)propoxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (29)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 71 %; ¹ HNMR (600 MHz, CDCl₃) δ 7.33 (m, 2H), 7.26 (m, 1H), 7.07 (m, 2H), 7.01 (m, 2H), 6.95 (m, 2H), 5.19 (s, 2H), 4.18 (s, 2H), 3.58 (m, 2H), 2.89 (septet, J = 7.2 Hz, 1H), 2.60 (m, 2H), 1.82 (m, 2H), 1.29 (d, J = 6.6 Hz, 6H); ¹³C NMR (150 MHz, CDCl₃) δ 162.9, 162.1, 160.5, 152.4, 148.5, 137.2 (d, J = 3.3 Hz), 135.4, 129.7 (d, J = 7.8 Hz), 129.2, 127.3 (d, J = 7.8 Hz), 119.9, 115.2 (d, J = 20.6 Hz), 73.0, 68.4, 33.5, 31.3, 31.1, 28.2, 20.4; HRMS (ESI+) calcd. for C₂₄H₂₇FN₂O₄ [M+H]⁺ 427.2028, found 427.2026 (E = 0.38 ppm).

6-Benzyl-1-((4-(4-fluorophenyl)butoxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (30)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 64 %; ¹HNMR (600 MHz, CDCl₃) δ 7.34 (t, J = 7.2 Hz, 2H), 7.28 (d, J = 7.2 Hz, 1H), 7.11 (dd, J = 1.8, 7.8 Hz, 2H), 7.07 (d, J = 7.8 Hz, 2H), 6.96 (t, J = 8.4 Hz, 2H), 5.21 (s, 2H), 4.19 (s, 2H), 3.60 (t, J = 6.0 Hz, 2H), 2.91 (septet, J = 7.2 Hz, 1H), 2.58 (t, J = 7.2 Hz, 2H), 1.63 (m, 2H), 1.58 (m, 2H), 1.29 (d, J = 6.6 Hz, 6H); ¹³C NMR (150, CDCl₃) δ 162.7, 162.0, 160.4, 152.2, 148.5, 137.8 (d, J = 3.3 Hz), 135.4, 129.7 (d, J = 7.8 Hz), 129.2, 127.3 (d, J = 7.8 Hz), 119.8, 115.0 (d, J = 20.7 Hz), 73.1, 69.3, 34.7, 33.4, 29.2, 28.9, 28.1, 20.3; HRMS (ESI+) calcd. for C₂₅H₂₉FN₂O₄ [M+H]⁺ 441.2184, found 441.2164 (E = 4.6 ppm).

6-Benzyl-1-(4-fluorobenzyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (32)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 72 %; ¹HNMR (600 MHz, CDCl₃) δ 7.30 (t, J = 7.2 Hz, 2H), 7.05-7.03 (m, 3H), 6.98 (d, J = 7.2 Hz, 2H), 6.94 (m, 2H), 4.88 (s, 2H), 3.83 (s, 2H), 2.81 (septet, J = 7.2 Hz, 1H), 1.22 (d, J = 6.6 Hz, 6H,); HRMS (ESI+) calcd. for C₂₁H₂₁FN₂O₃ [M+H]⁺ 369.1609, found 369.1608 (E = 0.26 ppm).

6-Benzyl-1-(4-fluorophenethyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (42)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 62 %; ¹HNMR (600 MHz, CDCl₃) δ 7.26–7.24 (m, 2H), 7.21 (m, 1H), 6.97(m, 2H), 6.89-6.88 (m, 2H), 6.85-6.84 (m, 2H), 3.79 (s, 2H), 3.73 (t, J = 7.8 Hz, 2H), 2.82 (septet, J = 7.2 Hz, 1H), 2.68 (t, J = 7.8 Hz, 2H), 1.29 (d, J = 6.6 Hz, 6H); HRMS (ESI+) calcd. for C₂₃H₂₃FN₂O₃ [M+H]⁺ 383.1765, found 383.1768 (E = 0.38 ppm).

6-Benzyl-1-(3-(4-fluorophenyl)propyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (43)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 72 %; ¹HNMR (600 MHz, CDCl₃) δ 7.21 (m, 3H), 6.94 (m, 2H), 6.87 (t, J = 8.4 Hz, 2H), 6.75 (d, J = 7.8 Hz, 2H), 3.71 (s, 2H), 3.58 (t, J = 7.2 Hz, 2H), 2.78 (septet, J = 7.2 Hz, 1H), 2.46 (t, J = 7.2 Hz, 2H), 1.79 (m, 2H), 1.29 (d, J = 6.6 Hz, 6H); HRMS (ESI+) calcd. for C₂₃H₂₅FN₂O₃[M+H]⁺ 397.1922, found 397.1924 (E = −0.51 ppm).

6-Benzyl-1-(4-(4-fluorophenyl)butyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (44)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 65 %; ¹HNMR (600 MHz, CDCl₃) δ 7.34 (t, J = 7.2 Hz, 2H), 7.28 (d, J = 7.8 Hz, 1H), 7.05 (m, 4H), 6.94 (t, J = 8.4 Hz, 2H), 3.92 (s, 2H), 3.73 (m, 2H), 2.87 (septet, J = 7.2 Hz, 1H), 2.51 (t, J = 7.2 Hz, 2H), 1.56 (m, 2H), 1.51 (m, 2H), 1.27 (d, J = 7.2 Hz, 6H); HRMS (ESI+) calcd. for C₂₄H₂₇FN₂O₃[M+H]⁺ 411.2078, found 411.2081 (E = −0.62 ppm).

1-(Benzyloxymethyl)-6-(4-fluorobenzyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (51)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 74%; ¹HNMR (600 MHz, CDCl₃) δ 7.32-7.27 (m, 5H), 7.00-6.98 (m, 4H), 5.28 (s, 2H), 4.67 (s, 2H), 4.14 (s, 2H), 2.81 (septet, J = 7.2 Hz, 1H), 1.27 (d, J = 6.6 Hz, 6H); HRMS (ESI−) calcd. for C₂₂H₂₃FN₂O₄[M−H]⁻ 399.1715, found 399.1706 (E = 2.1 ppm).

6-(4-Fluorobenzyl)-1-((4-fluorobenzyloxy)methyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (52)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 72%; ¹HNMR (600 MHz, CDCl₃) δ 7.28-7.27 (m, 6H), 7.00 (d, J = 7.2 Hz, 2H), 5.24 (s, 2H), 4.61 (s, 2H), 4.12 (s, 2H), 2.83 (septet, J = 7.2 Hz, 1H), 1.25 (d, J = 6.6 Hz, 6H); HRMS (ESI−) calcd. for C₂₂H₂₂F₂N₂O₄[M−H]⁻ 417.1620, found 417.1620 (E = 0.10 ppm).

6-Benzyl-1-(benzyloxymethyl)-5-ethyl-3-hydroxypyrimidine-2,4(1H,3H)-dione (53)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 70 %; ¹HNMR (600 MHz, CDCl₃) δ 7.33-7.27 (m, 8H), 7.03 (d, J = 7.8 Hz, 2H), 5.26 (s, 2H), 4.65 (s, 2H), 4.15 (s, 2H), 2.51 (q, J = 7.8 Hz, 2H), 1.06 (t, J = 7.2 Hz, 3H); HRMS (ESI−) calcd. for C₂₁H₂₂N₂O₄[M−H]⁻ 367.1652, found 367.1641 (E = 3.1 ppm).

1-(Benzyloxymethyl)-3-hydroxy-5-methylpyrimidine-2,4(1H,3H)-dione (54)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 37 %; ¹ HNMR (600 MHz, CDCl₃) δ 7.32-7.28 (m, 5H), 7.06 (s, 1H), 5.25 (s, 2H), 4.60 (s, 2H), 1.93 (s, 3H); HRMS (ESI−) calcd. for C₁₃H₁₄N₂O₄ [M−H]⁻ 263.1026, found 263.1023 (E = 1.3 ppm).

5-Benzyl-1-(benzyloxymethyl)-6-ethyl-3-hydroxypyrimidine-2,4(1H,3H)-dione (55)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 53%; ¹HNMR (600 MHz, CDCl₃) δ 7.22 (m, 4H), 7.19 (m, 3H), 7.12 (m, 3H), 5.42 (s, 2H), 4.61 (s, 2H), 3.76 (s, 2H), 2.68 (q, J = 7.2 Hz, 2H), 0.99 (t, J = 7.2 Hz, 3H); HRMS (ESI−) calcd. for C₂₁H₂₂N₂O₄ [M−H]⁻ 365.1507, found 365.1484 (E = 6.2 ppm).

5-Benzyl-6-ethyl-1-((4-fluorobenzyloxy)methyl)-3-hydroxypyrimidine-2,4(1H,3H)-dione (56)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 59%; ¹HNMR (600 MHz, CDCl₃) δ 7.13 (m, 3H), 7.07 (m, 4H), 6.83 (t, J = 6.6 Hz, 2H), 5.28 (s, 2H), 4.48 (s, 2H), 3.67 (s, 2H), 2.72 (q, J = 6.6 Hz, 2H), 1.07 (t, J = 6.6 Hz, 3H); HRMS (ESI−) calcd. for C₂₁H₂₁FN₂O₄[MH]⁻ 383.1485, found 383.1394 (E = 4.8 ppm).

1-Benzyl-6-(4-fluorobenzyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (64)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 70 %; ¹HNMR (600 MHz, CDCl₃) δ 7.33 (t, J = 7.2 Hz, 2H), 7.30 (m, 1H), 7.11 (d, J = 7.2 Hz, 2H), 7.06-7.04 (m, 2H), 7.02-7.00 ( m, 2H), 4.98 (s, 2H), 3.86 (s, 2H), 2.84 (septet, J = 7.2 Hz, 1H), 1.29 (d, J = 6.6 Hz, 6H); HRMS (ESI+) calcd. for C₂₁H₂₁FN₂O₃ [M+H]⁺ 369.1609, found 369.1606 (E = −1.9 ppm).

1,6-Bis(4-fluorobenzyl)-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (65)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 70 %; ¹HNMR (600 MHz, CDCl₃) δ 7.10-7.08 (m, 2H), 7.06-7.04 (m, 2H), 7.02-6.99 (m, 4H), 4.94 (s, 2H), 3.89 (s, 2H), 2.84 (septet, J = 7.2 Hz, 1H), 1.28 (d, J = 6.6 Hz, 6H); ¹³C NMR (150 MHz, CDCl₃) δ 163.1 (d, J = 31.2 Hz), 162.2, 161.4 (d, J = 31.2 Hz), 151.8, 148.3, 132.2 (d, J = 3.3 Hz), 130.4 (d, J = 2.9 Hz), 128.6 (d, J = 7.8 Hz), 127.6 (d, J = 8.4 Hz), 119.7, 116.4 (d, J = 21.8 Hz), 116.1 (d, J = 21.8 Hz), 46.6, 33.6, 28.2, 20.4; HRMS (ESI+) calcd. for C₂₁H₂₀F₂N₂O₃ [M+H]⁺ 387.1515, found 387.1502 (E = 3.3 ppm).

1,6-Bibenzyl-3-hydroxy-5-isopropylpyrimidine-2,4(1H,3H)-dione (66)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 53 %; ¹HNMR (600 MHz, CDCl₃) δ 7.39-7.34 (m, 4H), 7.32-7.30 (m, 2H), 7.12 (d, J = 7.2 Hz, 2H), 7.06 ( d, J = 7.8 Hz, 2H), 5.00 (s, 2H), 3.91 (s, 2H), 2.90 (septet, J = 7.2 Hz, 1H), 1.29 (d, J = 7.2 Hz, 6H); HRMS (ESI+) calcd. for C₂₁H₂₂N₂O₃ [M+H]⁺ 351.1703, found 351.1689 (E = 4.1 ppm).

1,6-Bibenzyl-5-ethyl-3-hydroxypyrimidine-2,4(1H,3H)-dione (67)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 43 %; ¹HNMR (600 MHz, CDCl₃) δ 7.31-7.27 (m, 3H), 7.26-7.23 (m, 3H), 7.04 (d, J = 6.6 Hz, 2H), 7.00 (d, J = 6.6 Hz, 2H), 4.91 (s, 2H), 3.80 (s, 2H), 2.47 (q, J = 7.2 Hz, 2H), 1.00 (t, J = 7.2 Hz, 3H); HRMS (ESI+) calcd. for C₂₀H₂₀N₂O₃ [M+H]⁺ 337.1547, found 337.1534 (E = 3.8 ppm).

6-Benzyl-5-ethyl-1-(4-fluorobenzyl)-3-hydroxypyrimidine-2,4(1H,3H)-dione (68)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 48 %; ¹HNMR (600 MHz, CDCl₃) δ 7.38-7.35 (m, 2H), 7.32 (m, 1H), 7.15-7.13 (m, 2H), 7.07 (d, J = 7.8 Hz, 2H), 7.04 (m, 2H), 4.94 (s, 2H), 3.88 (s, 2H), 2.52 (q, J = 7.2 Hz, 2H), 1.07 (t, J = 7.2 Hz, 3H); HRMS (ESI+) calcd. for C₂₀H₁₉FN₂O₃ [M+H]⁺ 355.1452, found 355.1442 (E = 3.0 ppm).

1-Benzyl-3-hydroxy-5-methylpyrimidine-2,4(1H,3H)-dione (69)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 32 %; ¹HNMR (600 MHz, CDCl₃) δ 7.35-7.31 (m, 3H), 7.28 (d, J = 7.2 Hz, 2H), 6.94 (s, 1H), 4.94 (s, 2H), 1.90 (s, 3H); HRMS (ESI+) calcd. for C₁₂H₁₂N₂O₃ [M+H]⁺ 233.0921, found 233.0926 (E = −2.3 ppm).

1-(4-Fluorobenzyl)-3-hydroxy-5-methylpyrimidine-2,4(1H,3H)-dione (70)

This compound was prepared as a white solid following the procedure described for the preparation of 4. Yield: 41 %; ¹ HNMR (600 MHz, CDCl₃) δ 7.30-7.27 (m, 2H), 7.08-7.05 (m, 2H), 6.96 (s, 1H), 4.85 (s, 2H), 1.89 (s, 3H); HRMS (ESI−) calcd. for C₁₂H₁₁FN₂O₃[M−H]⁻ 251.0826, found 251.0822 (E = 1.8 ppm).

Biology

In Vitro Integrase Catalytic Assays

Expression and purification of the recombinant IN in Escherichia coli were performed as previously reported^{28, 48} with addition of 10% glycerol to all buffers. Preparation of oligonucleotide substrates has been described.⁴⁹ Integrase reactions were performed in 10 μL with 400 nM of recombinant IN, 20 nM of 5′-end [³²P]-labeled oligonucleotide substrate and inhibitors at various concentrations. Solutions of 10% DMSO without inhibitors were used as controls. Reactions were incubated at 37 °C (60 minutes) in buffer containing 50 mM MOPS, pH 7.2, 7.5 mM MgCl₂, and 14.3 mM 2-mercaptoethanol. Reactions were stopped by addition of 10 μL of loading dye (10 mM EDTA, 98% deionized formamide, 0.025% xylene cyanol and 0.025% bromophenol blue). Reactions were then subjected to electrophoresis in 20% polyacrylamide–7 M urea gels. Gels were dried and reaction products were visualized and quantitated with a Typhoon 8600 (GE Healthcare, Little Chalfont, Buckinghamshire, UK). Densitometric analyses were performed using ImageQuant from Molecular Dynamics Inc. The concentrations at which enzyme activity was reduced by 50% (IC₅₀) were determined using “Prism” software (GraphPad Software, San Diego, CA) for nonlinear regression to fit dose-response data to logistic curve models.

HIV-1 Assay

The HIV Cytoprotection assay used CEM-SS cells and the IIIB strain of HIV-1. Briefly virus and cells were mixed in the presence of test compound and incubated for 6 days. The virus was pre-titered such that control wells exhibit 70 to 95% loss of cell viability due to virus replication. Therefore, antiviral effect or cytoprotection was observed when compounds prevent virus replication. Each assay plate contained cell control wells (cells only), virus control wells (cells plus virus), compound toxicity control wells (cells plus compound only), compound colorimetric control wells (compound only) as well as experimental wells (compound plus cells plus virus). Cytoprotection and compound cytotoxicity were assessed by MTS (CellTiter® 96 Reagent, Promega, Madison WI) and the EC₅₀ (concentration inhibiting virus replication by 50%), CC₅₀ (concentration resulting in 50% cell death) and a calculated TI (therapeutic index CC₅₀/EC₅₀) were provided. Each assay included the HIV RT inhibitor AZT as a positive control.

Modeling and docking

All modeling study was carried out using schrodinger modeling suite⁵⁰ based on our previous homology model³⁸ of HIV-1 IN CCD-DNA complex using Glide with standard precision protocol.⁵¹ The Mg²⁺ ions and the interfacial hydrophobic pocket between the HIV-1 IN and DNA were defined as required constraints. The van der Waals radii of non-polar atoms for each of the ligands were scaled by a factor of 0.8 to account for structure variability to specific ligand binding. Validation of the docking protocol was carried out by re-docking both raltegravir and elvitegravir into the ligand binding site with structure comparison to the earlier reported mode of binding in PFV IN complex.³⁶ To further validate our structural model for HIV IN inhibitor design, correlation of the experimental and predicted IC₅₀s for a common mode of binding was performed for the most active compounds (4, 11, 28, 29, 30, 32, 42) based on linear reposnse method⁵² implemented within Liason.⁵⁰

Abbreviations List

IN

integrase

HIV

human immunodeficiency virus

INI

integrase inhibitor

SAR

structure-activity-relationship

3′P

3′ processing

ST

strand transfer

DKA

diketoacid

HAART

highly active antiretroviral therapy

AIDS

acquired immunodeficiency syndrome

RT

reverse transcriptase

NTD

N-terminal domain

CCD

catalytic core domain

CTD

C-terminal domain

PFV

prototype foamy virus

mCPBA

meta-chloroperbenzoic acid

CPE

cytopathic effect