Discovery of selective small molecule type III phosphatidylinositol 4-kinase alpha (PI4KIIIα) inhibitors as anti hepatitis C (HCV) agents

Article abstract of DOI:10.1021/jm400781h

Hepatitis C virus (HCV) assembles many host cellular proteins into unique membranous replication structures as a prerequisite for viral replication, and PI4KIIIα is an essential component of these replication organelles. RNA interference of PI4KIIIα results in a breakdown of this replication complex and cessation of HCV replication in Huh-7 cells. PI4KIIIα is a lipid kinase that interacts with the HCV nonstructural 5A protein (NS5A) and enriches the HCV replication complex with its product, phosphoinositol 4-phosphate (PI4P). Elevated levels of PI4P at the endoplasmic reticulum have been linked to HCV infection in the liver of HCV infected patients.1 We investigated if small molecule inhibitors of PI4KIIIα could inhibit HCV replication in vitro. The synthesis and structure-activity relationships associated with the biological inhibition of PI4KIIIα and HCV replication are described. These efforts led directly to identification of quinazolinone 28 that displays high selectivity for PI4KIIIα and potently inhibits HCV replication in vitro.

Full text of DOI:10.1021/jm400781h

Article
pubs.acs.org/jmc
Discovery of Selective Small Molecule Type III Phosphatidylinositol
4‑Kinase Alpha (PI4KIIIα) Inhibitors as Anti Hepatitis C (HCV) Agents
Anna L. Leivers,† Matthew Tallant,† J. Brad Shotwell,† Scott Dickerson,‡ Martin R. Leivers,†
Octerloney B. McDonald,‡ Jeff Gobel,‡ Katrina L. Creech,‡ Susan L. Strum,‡ Amanda Mathis,†,§
Sabrinia Rogers,† Chris B. Moore,† and Janos Botyanszki*,†
†Infectious Diseases Medicines Discovery Unit, GlaxoSmithKline, 5 Moore Drive, Research Triangle Park, North Carolina
27709-3398, United States
‡Platform Technology and Science, GlaxoSmithKline, 5 Moore Drive, Research Triangle Park, North Carolina 27709-3398, United
States
S
* Supporting Information
ABSTRACT: Hepatitis C virus (HCV) assembles many host cellular proteins into unique
membranous replication structures as a prerequisite for viral replication, and PI4KIIIα is an
essential component of these replication organelles. RNA interference of PI4KIIIα results in a
breakdown of this replication complex and cessation of HCV replication in Huh-7 cells.
PI4KIIIα is a lipid kinase that interacts with the HCV nonstructural 5A protein (NS5A) and
enriches the HCV replication complex with its product, phosphoinositol 4-phosphate (PI4P).
Elevated levels of PI4P at the endoplasmic reticulum have been linked to HCV infection in the
liver of HCV infected patients.1 We investigated if small molecule inhibitors of PI4KIIIα could
inhibit HCV replication in vitro. The synthesis and structure−activity relationships associated
with the biological inhibition of PI4KIIIα and HCV replication are described. These efforts led directly to identification of
quinazolinone 28 that displays high selectivity for PI4KIIIα and potently inhibits HCV replication in vitro.
against other viruses that utilize the same cellular process. A
potential challenge of interfering directly with a cellular protein
is a higher likelihood of an on target related adverse phenotype.
The most advanced example of inhibiting an HCV host target is
the inhibition of cyclophilin B with cyclosporin A analogues.
Cyclophilin B is a host factor necessary for HCV replication
through the stimulation of NS5A’s binding to the RNA. The
cyclosporin A analogues have been advanced to clinical
evaluation8,9 and have proven to be effective inhibitors of
HCV replication with an increased genetic barrier to resistance.
Recently, a search for other host factors by multiple research
groups10−13 suggested that type III phosphatidylinositol 4-
kinase alpha (PI4KIIIα) is an essential host factor for HCV
replication. The relationship between PI4KIIIα and HCV
replication was further strengthened when a small molecule that
was previously thought to inhibit HCV replication through
NS5A inhibition was recently shown to inhibit HCV replication
through PI4KIIIα inhibition.14 PI4KIIIα is a lipid kinase that
catalyzes the phosphorylation of phosphatidylinositol at the 4-
position. The product, phosphatidylinositol 4-phosphate
(PI4P), is an intermediate toward the formation of
phosphatidylinositol 4,5-diphosphate (PIP2) which also plays
an important role in intracellular membrane trafficking. It was
shown that siRNA silencing of the PI4KIIIα gene (PI4KCA) in
INTRODUCTION
■
Hepatitis C virus (HCV), a positive-sense single-stranded RNA
virus within the flaviviridae virus family, was identified almost a
quarter century ago.2 An estimated 130−170 million people,
about 3% of the world population, are chronically infected, and
80% of the infections will progress to liver cirrhosis which
ultimately can lead to liver carcinoma.3 HCV infection is the
most common cause of liver transplantation and poses a major
health burden. After the recent approval of the HCV protease
inhibitors (PIs) boceprevir4 and telaprevir,5 the standard of care
for HCV infection now involves a triple combination therapy of
a PI + pegylated interferon-α (INF α) + ribavirin. Although this
regimen is more effective than the previous interferon−ribavirin
only treatment, it is still less than ideal because of significant
side effects attributed to both INF α and the protease inhibitors
themselves.6 A safe, well tolerated, all oral drug regime with a
high (>90%) cure rate remains an important scientific and
societal goal.
Traditional approaches toward developing an HCV treat-
ment have mainly involved targeting the viral genome directly,
including both structural and nonstructural viral proteins.
Multiple molecules are in various phases of preclinical or
clinical development.7 A complementary approach is to achieve
therapeutically relevant inhibition of a host cellular component
that is used by the HCV virus throughout its replication cycle.
Inhibiting a host target (as opposed to a viral protein) could
provide additional benefits including a higher barrier to viral
resistance, activity against all the HCV genotypes, and activity
Special Issue: HCV Therapies
Received: May 26, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
b
Figure 1. Chemical starting point (pIC50 and pEC50). pIC50 = −log IC50. pEC50 = −log EC50.
Huh-7 cells prevents replication with 99% efficiency. It was
proposed that PI4KIIIα plays an essential role in intracellular
membrane alterations, specifically by regulating the NS5A
phosphorylation which leads to modulating the morphology of
the viral replication site leading to the formation of the
functional HCV replication complex.10,15
On the basis of these results, GlaxoSmithKline initiated a
program to explore the therapeutic use of PI4KIIIα inhibition
for the treatment of HCV. In this report, we present the
discovery of highly selective, small molecule PI4KIIIα inhibitors
that effectively inhibit HCV replication in vitro.
One of the most promising representative chemotypes that
emerged from the HTS campaign was compound 1 (Figure 1),
which showed high PI4KIIIα activity (pIC50 = 8.8). Although
this molecule did show selectivity against PKs (data not
shown), it suffered from the lack of selectivity against the other
LKs we profiled. It was minimally selective against PI4KIIIβ
(pIC50 = 7.7) and was almost equipotent to the entire family of
PI3Ks (pIC50 all >8.6). Interestingly, the quinazoline core
contained within compound 1 was a motif of a small molecule,
a 4-anilinoquinazoline derivative that had been identified by
another group as an inhibitor of PIK4IIIα.14 Initial optimization
of 1 led to the development of compound 2 that had good
activity against PI4KIIIα (pIC50 = 7.4) along with moderate-to-
good selectivity over PI4KIIIβ (pIC50 = 5.5) and PI3Kα, -β, -δ,
-γ (pIC50 of 5.6, 5.0, 6.2, 6.2, respectively). This molecule also
exhibited high in vitro potency against multiple HCV genotypes
(GTs 1a, 1b, 2a). The key structural change that led to the
improved PI4KIIIα selectivity was the incorporation of the 2-
amino functionality on the quinoline ring.21
The lead molecule 1 originated from an ATP competitive LK
inhibitor library and was predicted to inhibit PI4KIIIα through
binding to the ATP binding site. To confirm this assumption,
we conducted an ATP competition study using compound 2.
The study clearly indicated that 2 inhibits PI4KIIIα in an ATP
competitive manner (see “Ki and Mode of Inhibition Study” in
Supporting Information (section VIII)).
In an attempt to gain insight into the SAR, a homology
model was constructed from a proprietary PI3Kγ structure,
since a crystal structure of PI4KIIIα was not available. The
sequence homology of the two proteins allowed PI4KIIIα
residue assignments to be made with high confidence for most
of the ATP binding site with the exception of the P-loop
containing a β sheet. To mitigate this, multiple P-loop models
were constructed. The gross binding mode of 2, as illustrated in
Figure 2, was predicted with high confidence. The amino-
quinoline moiety is predicted to form hydrogen bonds to the
hinge residue Ile1841. The sulfonamide hydrogen likely forms a
hydrogen bond to the catalytic Lys1792. A water molecule,
which is often conserved in PI3K structures, is also predicted to
be present in PI4KIIIα and forms a hydrogen bond to the
pyridyl nitrogen. The methoxy group is not expected to have a
SCREENING HITS AND HOMOLOGY MODELING
■
We conducted a focused high throughput screening (HTS)
campaign of GlaxoSmithKline’s lipid kinase (LK) inhibitor
collection against an HCV 1b replicon assay to identify a
suitable chemical starting point. The collection contained
molecules prepared for previous lipid kinase programs, in total
over 10 000 molecules. The aim was to find molecules that
inhibited HCV replication that were suitable for optimization
and that might already exhibit selectivity versus protein kinases
(PK).
As PI4KIIIα belongs to a family of 19 phosphoinositide lipid
kinases and it shares a sequence homology of the active site
with PI4KIIIβ and PI3Ks,16,17 the active compounds from the
HTS campaign were then tested for PI4KIIIα, PI4KIIIβ,
PI3Kα, PI3Kβ, PI3Kγ, and PI3Kδ activity. This screen would
allow us to find compounds selective for PI4KIIIα. Compounds
with poor kinase selectivity could have adverse effects,
especially since the closely related PI3Ks are common
chemotherapeutic targets and are involved in cellular activities
including proliferation, survival, and glucose homeostasis.18−20
In addition, we tested the most interesting molecules for other
(type II) LKs and a panel of PKs. Overall, the inhibitors of
PI4KIIIα from the HTS hits had decent to high selectivity
against PKs but showed poor or no selectivity against type III
PI4Kβ and type I PI3Ks. Thus, as the primary screening
strategy, we measured PI4KIIIα activity (along with HCV
inhibitory potency) and closely monitored the biochemical
selectivity against PI4KIIIβ, PI3Kα, PI3Kβ, PI3Kγ, and PI3Kδ
while occasionally testing the lead molecules for other (type II)
LKs and a panel of PKs.
B
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
conditions using bis(pinacolato)diboron, potassium acetate,
and PdCl2(dppf)−DCM in dioxane gave boronic esters 37−39.
For late stage diversification of the headgroup, compound 32
was directly borylated to give compound 40.
To explore a variety of cores (Scheme 2), 2-phenyl-
acetonitrile (41) was used as a common intermediate. The
reaction of 2-phenylacetonitrile and 2-amino-5-bromonicotin-
aldehyde with piperidine in dioxane at 100 °C gave an
incomplete reaction; however, the addition of solid sodium
hydroxide followed by heating at 80 °C gave complete
conversion to 42. Treatment of the nitrile 41 and 5-amino-2-
chloroisonicotinaldehyde with sodium hydride gave the 1,7-
naphthyridine intermediate (43). The 1,5-naphthyridin-2-
amine intermediate (44) was synthesized through the
cyclization of 2-phenylacetonitrile and 3-amino-6-
bromopicolinaldehyde in the presence of NaOH at 65 °C.
The final cross-coupling reactions of the bromo or chloro
intermediates (42−44) with compound 37 afforded com-
pounds 4, 5, and 6 in moderate yields.
Figure 2. Cutaway view of 2 bound to a PI4Kα homology model.
Hydrogen bonds are shown in yellow. Image was created with Maestro
molecular modeling software (Schrodinger LLC).
̈
direct interaction with the protein: the model did not predict
specific role in the inhibitors’ binding.
The quinoxaline core (Scheme 3) was synthesized via a
three-component cyclization reaction involving 4-bromo-1,2-
benzenediamine (45), benzaldehyde, and 1,1,3,3-tetramethyl-
butyl isocyanide. The primary adduct was oxidized with DDQ
to give a mixture of 46a and 46b in about a 1:2 ratio. The two
regioisomers were treated with HCl in dioxanes at 70 °C to
give 47a and 47b, which were subsequently cross-coupled with
compound 37 to afford 7 and 48, respectively. The isomers
were separated by HPLC, and absolute structural assignments
were made based on 13C chemical shifts and ab initio IR
analysis.
The binding model led us to hypothesize that changes to
both the phenylsulfonamide and morpholinophenyl moieties
might impact selectivity over the PI3Ks. Changes to the core
could impact binding deep within the pocket and could
indirectly change the protein interactions made by the pyridyl
moiety and sulfonamide headgroup. The morpholinophenyl
group itself appeared quite solvent exposed. The model
generally provided a significant framework for testing SAR
hypotheses and evaluating findings.
The 2-aminoquinoline series, represented by 2, was a step in
the right direction with respect to selectivity; however, the
potency against PI4KIIIα was significantly lower than the
original hit (1) and the overall selectivity was still not
satisfactory. In addition, the entire 2-aminoquinoline series
suffered from poor DMPK properties (Table 2, entry 1). We
hoped that further exploration of the SAR around the “head”,
“linker”, “core”, and “tail” regions of the structure (Figure 3)
would enable us to identify a molecule with pIC50 of >8.0
against PI4KIIIα, selectivity of >100× against other LKs and
PKs, and a substantially improved DMPK profile.
Synthesis of the quinazolinone derivatives (Scheme 4) was
achieved through the cyclization of commercially available
isothiocyanates with 2-aminobenzoic acids (49). Chlorination
at position 2 was affected by refluxing in neat POCl3 in the
presence of PCl5 to give intermediates 50−55. The 2-amino
functionality was installed by displacement of the chlorine with
(4-methoxyphenyl)methanamine in DMF, followed by cleavage
of the para-methoxybenzyl (PMB) group by heating in neat
TFA to give intermediates 56−61. Products 8, 9, and 11 were
made by the cross-coupling reaction of 56 with the
intermediates 37, 38, and 39 described in Scheme 1.
Quinazolinone 56 (Y = Br) was borylated to give intermediate
62, which upon cross-coupling with intermediate 35 gave
product 10. A cross-coupling reaction of 56 with 40 gave the
diamino intermediate 63. It was possible to take advantage of
the different reactivities of the 3-pyridyl and 2-quinazolinonyl-
amino groups and incorporate a set of sulfonyl groups using the
corresponding sulfonyl chlorides at the last step, allowing for
late stage diversification of 63 to final products 12−19.
Although N-arylquinazolinones were easily accessed from the
aryl isothiocyanates, this route did not work using alkyl
isothiocycanates. The initial cyclization with cyclopentyl
isothiocyanate failed to give the desired product; therefore, an
alternative route was devised (Scheme 5) based on a literature
precedent.22 The O-phenylisourea intermediate 65 was formed
from methyl 5-bromoanthranilate (64), which was treated with
an alkylamine in hot isopropanol to give the corresponding
tetrahydroquinazolines 66 and 67. Hydrolysis with concen-
trated HCl in DMSO gave the free amines 68 and 69 in
quantitative yield.
Figure 3. Optimization strategy.
CHEMISTRY
■
The head sulfonamides 37−40 were synthesized as boronic
ester building blocks (Scheme 1). Compounds 33−35 were
made from sulfonylation of commercially available 5-bromo-3-
anilines (30−32) with 2,4-difluorobenzene-1-sulfonyl chloride
in pyridine. Treatment of compound 33 with iodomethane
gave the N-methylated sulfonamide 36. Standard borylation
The “reverse” sulfonamide head was synthesized via an
adaptation of a literature procedure23 (Scheme 6). Upon
formation of a diazonium ion of 70, treatment with thionyl
C
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of the Sulfonamide Head Groups
a
Reagents and conditions: (a) 2,4-difluorobenzene-1-sulfonyl chloride, pyridine, room temperature, overnight, 72% yield; (b) MeI, K2CO3, DMF,
room temperature, 48 h, 79% yield; (c) (BPin)2, KOAc, PdCl2(dppf)−DCM, dioxane, 80−100 °C, 2−18 h, 56−89% yield.
a
Scheme 2. Synthesis of the Alternative Cores
a
Reagents and conditions: (a) 2-amino-5-bromonicotinaldehyde, piperidine, dioxane, 100 °C, 3 h, then NaOH, 80 °C, overnight, 74% yield; (b) 5-
amino-2-chloroisonicotinaldehyde, NaH, THF, room temperature, 30 min, 27% yield; (c) 3-amino-6-bromopicolinaldehyde, NaOH, dioxane, 65 °C,
2.5 h, 90% yield; (d) 37, KOAc or 2 M K2CO3, Pd(PPh3)2Cl2, dioxane, 120 °C, microwave, 1−1.75 h, 5−13% yield.
D
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 3. Synthesis of the Quinoxaline Core
a
Reagents and conditions: (a) benzaldehyde, 1,1,3,3-tetramethylbutyl isocyanide, concentrated HCl, MeOH, room temperature, overnight, 100%
yield, 72% purity; (b) DDQ, benzene, room temperature, overnight, 29.3% yield; (c) HCl, DCM, 70 °C, 94% yield; (d) 37, K2CO3, PdCl2(dppf)−
DCM, dioxane/water, 80 °C, overnight, 69% yield.
a
Scheme 4. General Route to the Aryl Tail Derivatives of the Quinazolinone Core
a
Reagents and conditions: (a) Ar(R) isothiocyanate, TEA, dioxane or EtOH, reflux, 6 h to overnight, 51−93% yield; (b) PCl5, POCl3, reflux, 7 h to
overnight, 70−100% yield; (c) (4-methoxyphenyl)methanamine, DIEA, DMF, 80 °C, overnight to 48 h, 96−100% yield; (d) TFA, microwave, 140
°C, <30 min or TFA, reflux, 2 days, 45−100%, yield; (e) 37, 38, 39, Cs2CO3, PdCl2(dppf)−DCM, dioxane/water, 80 °C, 1 h to overnight, 54−65%
yield; (f) bis(pinacolato)diboron, PdCl2(dppf)−DCM, KOAc, dioxane, 90 °C, 3 h; (g) 35, PdCl2(dppf)−DCM, dioxane, saturated NaHCO3,
microwave, 120 °C, 20 min, 6% yield; (h) 40, Cs2CO3, PdCl2(dppf)−DCM, THF/water, 65 °C, 1 h, 40% yield; (i) Z-SO2Cl, pyridine, 0 °C, room
temperature, 2−24 h, 11−72% yield.
chloride and copper(I) chloride gave the sulfonyl chloride (71).
Subsequent reaction with 2,4-difluoroaniline in pyridine,
followed by displacement of the chlorine with sodium
methoxide gave the “head” sulfonamide intermediate 72.
Borylation was carried out with bis(pinacolato)diboron in
dioxane in the presence of potassium acetate and
PdCl2(dppf)−DCM to give the corresponding boronic ester
(73). Finally the 3-arylquinazolinones (56−61, Scheme 4) and
the 3-alkylquinazolinones (68, 69, Scheme 5) were cross-
coupled with compound 73 to give products 20−27 in
moderate to high yield.
RESULTS AND DISCUSSION
■
In an effort to identify a molecule with pIC50 of >8.0 against
PI4KIIIα, selectivity of >100× against other LKs and PKs, and
good DMPK profile, we decided to explore analogues
E
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 5. Synthesis of 3-Alkyl Derivatives of the 2-Aminoquinazolinone Core
a
Reagents and conditions: (a) diphenyl cyanocarbodiimidate, pyridine, 50 °C, overnight, 64% yield; (b) R1-NH2, i-PrOH, 85 °C, 1 h, 58−100%
yield; (c) concentrated HCl, DMSO, 100 °C, 1−2.5 h, 100% yield.
a
Scheme 6. Synthesis of the Reverse Sulfonamide Analogues
a
Reagents and conditions: (a) SOCl2, water, Cu(I)Cl, HCl, NaNO2, −5 to 30 °C, overnight, 77% yield; (b) (1) 2,4-difluoroaniline, pyridine, DCM,
room temperature, overnight, 93% yield; (2) NaOMe/MeOH, 50 °C, 4 h, 93% yield over two steps; (c) (BPin)2, KOAc, PdCl2(dppf)−DCM,
dioxane, 95 °C, 2 h, 60% yield; (d) 56−61, 68, 69, Cs2CO3 or K2CO3 or KOAc, PdCl2(dppf)−DCM, dioxane/water, 80−100 °C, 1 h to overnight,
26−65% yield.
containing alternative cores (Table 1). We replaced the central
quinoline ring of 3 with isosteric ring systems containing an
additional heteroatom in hopes of creating a potential new
interaction site with the target enzyme. In the case of the 1,8-
naphthyridine (4), this modification resulted in an almost 10-
fold decrease in PI4KIIIα potency (pIC50 = 6.2). Moving the
nitrogen to the 7-position (5, pIC50 = 6.7) provided a
compound with slightly lower PI4KIIIα potency relative to 3
and comparable selectivity (0.5−1.0 vs 0−1.1). In contrast,
inserting the nitrogen at the 5-position (6, pIC50 = 5.1) was not
well tolerated resulting in a 2 log decrease in potency. When we
moved the nitrogen to the 4-position to give the quinoxaline
core (7), a 1.3 log improvement in PI4KIIIα potency was
observed (pIC50 = 8.3). The replicon activity and the selectivity
was maintained. Given the high tolerability of the substitution
at the 4-position for PI4KIIIα potency, a carbonyl derivative 8,
a quinazolin-4-one core, was evaluated. A significant increase in
PI4KIIIα potency was observed (pIC50 = 8.8), and the
inhibition of HCV 1b replicon reached the picomolar level
(pIC50 = 9.3). Compound 8 also offered a solid improvement
in selectivity against PI3Ks (0.4−1.8). While the selectivity did
improve, the absolute potencies against all PI3Ks were the
highest within the alternative core series.
improved selectivity, and promising rat DMPK, we chose 8 as
our new lead for optimization.
Comprehensive SAR related to the “head”, “linker”, and “tail”
around this new quinazolinone core of compound 8 was
developed. The results of the head and linker optimization are
summarized in Table 3. The nitrogen in the linker pyridine ring
was crucial for high PI4KIIIα potency, as removing it (i.e., 9)
resulted in a significant decrease (2.3 log). This is consistent
with our model in which the pyridine-N forms an interaction
with the backbone of the enzyme through a water bridge.
Replacing the methoxy group with an amino group (10)
slightly improved the PI4KIIIα activity and validated the
model’s predictions that this group serves just as a space filling
function. Interestingly, this change did not translate to HCV
inhibition, as reflected by a significant decrease in the replicon
activity (pIC50 = 7.8). N-Methylation of the sulfonamide (11)
was not tolerated and decreased the PI4KIIIα potency over
>1000× (3.4 log). These findings were consistent with the
molecule’s predicted binding mode, which was thought to
involve the sulfonamide hydrogen forming a H-bond with the
catalytic Lys1792 (Figure 2, homology model).
In order to decrease the molecular weight, we tried to
minimize the headgroup by truncating parts of the molecule. In
the case of the 2,4-difluorobenzenesulfonamide moiety,
eliminating this group and replacing it with a methyl group
The DMPK of 8 was also evaluated (Table 2). To our
satisfaction, the low-dose rat DMPK study revealed that 8 had
substantially better properties relative to the best molecule in
the 2-amionquinoline series (2) in terms of exposure (AUC)
and the percent bioavailability (% F) (10× and 2×,
respectively). Also, the half-life has increased almost 3-fold
(6.70 h vs 2.39 h). On the basis of the high PI4KIIIα potency,
(12) resulted in a 10-fold loss in PI4KIIIα potency (pIC50
=
7.7), although the selectivity was improved (5× to 10×). We
wondered if slightly larger alkyl groups might restore PI4KIIIα
potency; however, the isopropyl group (13) led to a further
decrease in the potency of PI4KIIIα (pIC50 = 6.9) and a loss of
F
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 1. In Vitro Enzyme and Replicon Activity Profile of Analogues with Alternative Cores (pIC50
)
a
b
pIC50 = −log IC50. Selectivity range: smallest to highest differences of pIC50(PI4Kα) − pIC50(PIKx), assigned to 0 if negative.
substituted 8. Furthermore, an electron deficient heteroar-
omatic ring (19) proved to be the most potent analogue against
PI4KIIIα in the series (pIC50 = 9.5). Although these molecules
showed higher selectivity, the absolute potencies against the
other lipid kinases were still high.
Table 2. Comparison of DMPK Parameters of 2 and 8 after
Single Oral 5 mg/kg Solution Dose in Rat
compd AUC (ng·h/mL) Cmax (ng/mL) T1/2 (h) Tmax (h) F (%)
2
8
4996
1392
3575
2.39
6.7
0.67
6
38
60
60177
The breakthrough toward improving selectivity came when
the sulfonamide moiety of 8 was reversed to give 20. There was
a minor decrease in PI4KIIIα potency (pIC50 of 8.2) but a
significantly larger decrease in PI4KIIIβ (pIC50 of 5.5) and in
PI3Kα,β,δ,γ activities (pIC50 of 6.0, 4.9, 6.3, 5.9, respectively),
thereby improving the selectivity to the desirable range (1.9−
3.3). On the basis of the homology model (Figure 2), we
theorized that reversing the sulfonamide the nature of the H-
bond between the inhibitor and Lys1792 was likely weakened
and this change was better tolerated by PI4KIIIα than by other
lipid kinases.
selectivity as well (0.2−1.6). An unsubstituted phenyl ring (14)
improved the PI4KIIIα potency to the single digit nanomolar
range (pIC50 = 8.6) and the initial selectivity was gained back.
These molecules clearly suggested that an aromatic ring at this
position is favorable for high PI4KIIIα potency and selectivity.
To further explore the electronics and sterics around the
aromatic ring, an electron-donating methyl group was
introduced at various positions (15, 16, 17). In general, these
changes were well tolerated, especially the 2-position (15)
In the context of the new “reverse sulfonamide” structure, we
investigated the tail region of the molecule with the goal of
further optimization. The results of the tail SAR are
summarized in Table 4. Replacing the phenyl group with
smaller alkyl groups was not well tolerated. As exemplified with
the ethyl derivative (21), a decrease in PI4KIIIα activity was
observed (pIC50 = 7.4), and it was coupled with diminishing
selectivity. Larger alkyl groups, like the tetrahydropyran (22),
did maintain potency (pIC50 = 8.0) against PI4KIIIα, but the
selectivity decreased below the starting molecule’s (20) range.
where the PI4KIIIα activity was slightly improved (pIC50
=
9.0). It was also the best in the series in terms of selectivity
(1.4−2.6). A slight loss in potency and selectivity was observed
placing the methyl group at the 3- or 4-position (pIC50 of 8.4
and 7.8, respectively). This trend was also observed in the
corresponding replicon potencies as well (pIC50 of 9.4, 8.9, and
8.5, respectively).
A similar exercise with electron-withdrawing substitutions
was also conducted. The addition of a cyano group to the para
position (18) resulted in activity on par with the 2,4-difluoro-
G
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 3. In Vitro Enzyme and Replicon Activity Profile of Analogues Optimized on the Linker and Head Regions (pIC50
)
H
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. continued
a
b
pIC50 = −log IC50. Selectivity range: smallest to highest differences of pIC50(PI4Kα) − pIC50(PIKx), assigned to 0 if negative.
Larger alkyl groups, like the tetrahydropyran (22), did maintain
potency (pIC50 = 8.0) against PI4KIIIα, but the selectivity
decreased compared with the starting lead molecule’s (20)
range (1.4−2.6 vs 1.9−3.3, respectively). As the alkyl groups
offered no significant improvement over the aryl groups, we
optimized the aryl ring in hopes of finding greater improve-
ments in activity and selevtivity.
quinazolinone 2- and 4-positions, respectively. In such
circumstances, one of the enantiomers could have a better fit
within the chiral environment of the enzyme binding site which
could translate to higher potency and better selectivity. To
investigate this possibility, 23 was analyzed by means of chiral
chromatography, and two distinct peaks were observed. The
two enantiomers were separated to give 28 and 29 with a chiral
purity of >99% for each. The absolute configurations of the
enantiomers were determined by vibrational circular dichroism
(VCD)24 and spectroscopy assigned as 28 (aS) and 29 (aR)
(Figure 4). Upon heating at 80 °C in DMSO for 6 days, no
interconversion or decomposition was observed suggesting that
they would be thermodynamically stable in the conditions of a
biological system.
Table 5 summarizes the head-to-head comparison of the in
vitro enzyme and replicon data of the isomers. As we expected,
they had different profiles with 28 exhibiting 25× higher
PI4KIIIα potency than 29 (pIC50 of 8.3 vs 6.9) and an
improved selectivity range (2.7−3.2 vs 1.4−2.1) against the
other lipid kinases. Comparing the anti-HCV activity, the
difference in PI4KIIIα potencies was reflected in the replicon
activities: 28 was about 1 order of magnitude more potent
against all three genotypes measured (1a, 1b, 2a). We also
compared the iv and po DMPK properties in rat of 28 and 29
(Table 6). The study revealed that 28 had acceptable
properties, although the less potent 29 was superior in this
aspect. In addition, we ran 28 against a 300-protein kinase
panel in order to get a broad picture of the overall selectivity. It
showed an extremely clean profile against all the measured PKs
as well (see data in section VII of Supporting Information).
On the basis of the above results, compound 28 was selected
as a precandidate and was promoted for additional animal
In the course of investigating substitutions on the tail phenyl
ring, an electron withdrawing trifluoromethyl group was added
to all possible positions (23, 24, 25). In general, the PI4KIIIα
potency suffered, although the ortho isomer (23, pIC50 = 7.8)
still retained significant activity. The selectivity showed the
same trend: slight decrease and ortho position being the best.
In fact, 23 displayed an excellent selectivity profile, reaching the
target >100-fold selectivity against PI4KIIIβ and all PI3Ks with
a selectivity range of 160- to 600-fold (2.2−2.8). The ortho-
methyl derivative (26) showed improved PI4KIIIα activity,
suggesting that an electron donating substitution was favorable
for potency, although this resulted in poorer selectivity against
PI3Ks (1.3−2.3). To test the size aspect of substituting at the
ortho position, the isopropyl analogue was synthesized.
Displaying only marginally improved selectivity (1.5−2.7),
this analogue (27) had a decreased PI4KIIIα activity (pIC50
=
7.5). Upon further examination of the SAR data, we concluded
that 23 fulfilled our starting criteria (low nanomolar PI4KIIIα
potency and >100-fold selectivity over PI4KIIIβ and
PI3Kα,β,δ,γ); therefore, it became our candidate for further
profiling.
The structure of 23 suggested a possibility of atropisomer-
ism. It appeared likely that the ortho-trifluoromethyl group
could prevent the free rotation of the phenyl group because of
the proximity of the amino and carbonyl groups in the
I
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 4. In Vitro Enzyme and Replicon Activity Profile of Analogues Optimized on the Tail Region (pIC50
)
a
b
pIC50 = −log IC50. Selectivity range: smallest to highest differences of pIC50(PI4Kα) − pIC50(PIKx), assigned to 0 if negative.
Figure 4. Resolution and Assignment of Atropisomers.
Table 5. Comparison of in Vitro Enzyme and Replicon Potencies of Racemic 23 and the Atropisomers 28 and 29
a
a
pIC50
pIC50
b
compd
PI4Kα
PI4Kβ
PI3Kα
PI3Kβ
PI3Kδ
PI4Kγ
selectivity range
rep 1a
rep 1b
rep 2a
23
28
29
7.8
8.3
6.9
5.8
6.0
5.3
5.6
5.6
5.5
5.0
5.1
4.8
5.5
5.6
5.0
5.5
5.6
5.5
2.2−2.8
2.7−3.2
1.4−2.1
7.7
7.9
6.9
8.5
8.6
7.8
7.9
7.9
7.1
a
b
pIC50 = −log IC50. Selectivity range: smallest to highest differences of pIC50(PI4Kα) − pIC50(PIKx), assigned to 0 if negative.
studies. It had a maximum tolerated dose of 50 mg/kg,
following single dose administration, and was poorly tolerated
in the 14-day mouse toxicity study. Animals treated with 28 at
40 mg kg−1 day−1 had morbidity and mortality and test-article-
related changes in the stomach, thymus, spleen, and clinical
pathology parameters. A detailed description of the study will
J
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 6. DMPK Comparison of 28 and 29, Rats Single iv (1 mg/kg) and Oral (5 mg/kg) Administration
compd
route
CL (mL min−1 kg−1
2.5
)
T1/2 (h)
Vdss (L/kg)
0.55
Cmax (ng/mL)
Tmax (h)
DNAUC (h kg ng mL−1 mg−1
)
F (%)
29
iv
3.5
2.7
1.3
4.0
7113
9985
3695
3626
po
iv
4429
1556
3
>100
98
28
4.5
0.34
po
4.25
a
Formulation: DMSO/cyclodextrin.
tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine-3-sulfonamide (37) (18.1
g, 89%). 1H NMR (400 MHz, DMSO-d6) δ ppm 10.11 (s, 1 H), 8.56
(d, J = 1.76 Hz, 1 H), 8.09 (d, J = 1.76 Hz, 1 H), 7.20−7.34 (m, 2 H),
7.00−7.09 (m, 1 H), 3.97 (s, 3 H), 1.28 (s, 12 H); LC−MS m/z =
427.20 (M + H)+.
be reported in a subsequent publication (manuscript in
preparation).
CONCLUSION
■
Targeting a host protein to achieve antiviral activity is a
relatively unexplored area and carries the risk of higher
likelihood of off-target activity. However, the potential benefits
of wide-spectrum inhibition, higher barrier to resistance, and a
novel mechanism of action are clear advantages of host−target
inhibition. Starting with a focused HTS of a 10K lipid kinase
library we targeted the host cell lipid kinase PI4KIIIα that was
shown by siRNA knockout experiments to be vital for HCV
replication. To determine if PI4KIIIα inhibition with a small
molecule translates to an anti-HCV effect, we synthesized
multiple series of inhibitors. The plotted PI4KIIIα vs replicon
potencies clearly indicated a correlation (R2 = 0.79) that
strongly supports the original hypothesis (section V of
Supporting Information). Through extensive optimization, we
developed a potent host enzyme inhibitor (28) that is the first
extremely selective PI4KIIIα inhibitor against other lipid and
protein kinases, coupled with the in vitro inhibition of HCV.
N-(5-(7-Amino-6-phenyl-1,8-naphthyridin-3-yl)-2-methoxy-
pyridin-3-yl)-2,4-difluorobenzenesulfonamide (4). 2-Amino-5-
bromonicotinaldehyde (0.5g, 2.49 mmol), 2-phenylacetonitrile (41)
(0.574 mL, 4.97 mmol), and piperidine (0.491 mL, 4.97 mmol) were
dissolved in dioxane (10 mL) and heated to 100 °C for 3 h. The
mixture was then treated with sodium hydroxide (0.099 g, 2.49 mmol)
and was heated at 80 °C overnight. The solution was concentrated
under reduced pressure and the residue was purified by silica gel
chromatography (0−15% MeOH/DCM) to give 6-bromo-3-phenyl-
1
1,8-naphthyridin-2-amine (42) (0.549g, 74%) as a brown solid. H
NMR (400 MHz, DMSO-d6) δ ppm 8.74 (d, J = 2.54 Hz, 1 H), 8.39
(d, J = 2.74 Hz, 1 H), 7.87 (s, 1 H), 7.41−7.56 (m, 5 H), 6.63 (b.s., 2
H); LC−MS m/z = 300.2, 302.2 (M + H)+.
A microwave vial was charged with 6-bromo-3-phenyl-1,8-
naphthyridin-2-amine (42) (0.2 g, 0.666 mmol), 2,4-difluoro-N-(2-
methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-yl)-
benzenesulfonamide (37) (0.284 g, 0.666 mmol), KOAc (0.196 g, 2.00
mmol), and 1,4-dioxane (10 mL). To the vial was then added
Pd(PPh3)2Cl2 (0.047 g, 0.067 mmol), and the mixture was placed
under a N2 atmosphere. The vial was heated to 100 °C in the
microwave for 25 min and then heated for an additional hour at 120
°C. The solution was concentrated and the residue was prepurified by
silica gel chromatography (0−5% MeOH/DCM) and was then
purified by silica gel chromatography (100% EtOAc) to give N-[5-(7-
amino-6-phenyl-1,8-naphthyridin-3-yl)-2-(methyloxy)-3-pyridinyl]-
2,4-difluorobenzenesulfonamide (4) (18 mg, 5%) as a pale yellow
EXPERIMENTAL SECTION
■
Compound Purities and Identity. All compounds underwent an
immediate processing quality control protocol to confirm identity and
determine relative purity immediately before processing in the enzyme
and replicon assays. Analysis is by UPLC−MS with UV diode array
detection to determine purity and MS used to confirm molecular
weight. A Waters Acquity UPLC system comprising binary solvent
manager, sample manager, PDA detector, Waters ZQ or SQD mass
spectrometer, Waters Acquity evaporative light scattering detector or
Polymer Laboratories evaporative light scattering detector was
employed. Mobile phases were the following: acetonitrile + 0.1%
formic acid; water + 0.1% formic acid. Wash solutions were the
following: strong wash 100% acetonitrile + 0.1% formic acid; weak
wash 50:50 acetonitrile/water + 0.1% formic acid. All individual lots of
compounds tested ≥95% pure according to this protocol.
1
solid. H NMR (400 MHz, MeOH-d4) δ ppm 8.85 (d, J = 2.5 Hz, 1
H), 8.23 (dd, J = 6.3, 2.4 Hz, 2 H), 8.02 (d, J = 2.1 Hz, 1 H), 7.89 (s, 1
H), 7.84 (td, J = 8.5, 6.2 Hz, 1 H), 7.50−7.56 (m, 4 H), 7.42−7.50 (m,
2 H), 7.14−7.23 (m, 1 H), 6.99−7.07 (m, 1 H), 4.88 (s, 1 H), 3.78 (s,
3 H); LC−MS m/z = 520.4 (M + H)+; HRMS calculated for
C26H20N5O3F2S (M + 1)+, 520.1255; found 520.1255.
N-(5-(2-Amino-3-phenyl-1,7-naphthyridin-6-yl)-2-methoxy-
pyridin-3-yl)-2,4-difluorobenzenesulfonamide (5). To a solution
of 5-amino-2-chloroisonicotinaldehyde (344 mg, 2.20 mmol) and 2-
phenylacetonitrile (41) (258 mg, 2.20 mmol) in THF (15 mL) at
room temperature was added 60% sodium hydride (264 mg, 6.60
mmol). The mixture was stirred for 30 min. The solvents were
evaporated under reduces pressure and the residue was dissolved in
EtOAc, washed with saturated NH4Cl, brine, dried over Na2SO4, and
concentrated to give 6-chloro-3-phenyl-1,7-naphthyridin-2-amine (43)
(0.150 g, 27%). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.68 (s, 1 H),
7.87 (s, 1 H), 7.78 (s, 1 H), 7.44−7.61 (m, 5 H), 6.58 (br s, 2 H);
LC−MS m/z = 256.2 (M + H)+.
N-(2,4-Difluorophenyl)-2-methoxy-5-(4,4,5,5-tetramethyl-
1,3,2-dioxaborolan-2-yl)pyridine-3-sulfonamide (37). A solution
of 5-bromo-2-methoxypyridin-3-amine (30) (24.8 g, 117 mmol) in
pyridine (45 mL) at room temperature was treated by the dropwise
addition of 2,4-difluorobenzene-1-sulfonyl chloride (15.8 mL, 117
mmol) over 15 min. The mixture was stirred at room temperature
overnight and was loaded onto a silica gel column and purified (0−
30% EtOAc/hexane) to give N-(5-bromo-2-methoxypyridin-3-yl)-2,4-
1
difluorobenzenesulfonamide (33) as a yellow solid (32 g, 72%). H
A microwave vial was charged with 6-chloro-3-phenyl-1,7-
naphthyridin-2-amine (43) (130 mg, 0.508 mmol), 2,4-difluoro-N-
(2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-
yl)benzenesulfonamide (37) (130 mg, 0.305 mmol), 2 M aquoeus
K2CO3 (0.8 mL), and 1,4-dioxane (4 mL). The mixture was degassed,
treated with Pd(PPh3)2Cl2 (35.7 mg, 0.051 mmol), and placed under a
nitrogen atmosphere. The mixture was heated in the microwave at 120
°C for 1 h. The solution was concentrated under reduced pressure and
prepurified by silica gel chromatography (0−40% EtOAc/DCM) and
was then purified with reverse phase chromatography (0−10%ACN/
H2O) to give N-[5-(2-amino-3-phenyl-1,7-naphthyridin-6-yl)-2-(meth-
NMR (400 MHz, DMSO-d6) δ ppm 10.46 (s, 1 H), 8.13 (d, J = 2.35
Hz, 1 H), 7.72−7.82 (m, 2 H), 7.58 (ddd, J = 10.65, 9.19, 2.44 Hz, 1
H), 7.23 (td, J = 8.50, 1.76 Hz, 1 H), 3.61 (s, 3 H); LC−MS m/z =
379.21, 381.20 (M + H)+.
A solution of N-(5-bromo-2-methoxypyridin-3-yl)-2,4-difluoroben-
zenesulfonamide (33) (18g, 47.5 mmol), 4,4,4′,4′,5,5,5′,5′-octamethyl-
2,2′-bi(1,3,2-dioxaborolane) (15.7 g, 61.7 mmol), KOAc (14.0 g, 142
mmol) in 1,4-dioxane (100 mL) was treated with PdCl2(dppf)−
CH2Cl2 (0.775 g, 0.949 mmol). The mixture was heated at 100 °C for
4 h and was then loaded onto silica gel column and purified (0−30%
EtOAc/hexane) to give N-(2,4-difluorophenyl)-2-methoxy-5-(4,4,5,5-
K
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
yloxy)-3-pyridinyl]-2,4-difluorobenzenesulfonamide (5) (20 mg, 13%)
as a pale yellow oil. 1H NMR (400 MHz, chloroform-d) δ ppm 9.14 (s,
1 H), 8.60 (d, J = 1.8 Hz, 1 H), 8.40 (d, J = 1.8 Hz, 1 H), 7.84−7.96
(m, 1 H), 7.82 (s, 1 H), 7.78 (s, 1 H), 7.44−7.60 (m, 6 H), 6.87−6.99
(m, 2 H), 5.26 (br s, 2 H), 3.95 (s, 3 H); LC−MS m/z = 520.4 (M +
H)+; HRMS calculated for C26H20N5O3F2S (M + 1)+, 520.1255; found
520.1254.
N-(5-(6-Amino-7-phenyl-1,5-naphthyridin-2-yl)-2-methoxy-
pyridin-3-yl)-2,4-difluorobenzenesulfonamide (6). A solution of
3-amino-6-bromopicolinaldehyde (0.13 g, 0.647 mmol), 2-phenyl-
acetonitrile (41) (0.076 g, 0.647 mmol), and sodium hydroxide (0.078
g, 1.94 mmol) in 1,4-dioxane (10 mL) was heated at 65 °C for 2.5 h.
The mixture was concentrated under reduced pressure and the residue
was purified by silica gel chromatography (0−30% EtOAc/hexanes) to
give 6-bromo-3-phenyl-1,5-naphthyridin-2-amine (44) (0.175 g, 90%)
as a yellow solid. LC−MS m/z = 300.2, 302.2 (M + H)+.
A flask containing the 35:65 mixture of 6-bromo-3-phenyl-2-
quinoxalinamine (47a) and 7-bromo-3-phenylquinoxalin-2-amine
(47b) (0.086 g, 0.287 mmol), 2,4-difluoro-N-[2-(methyloxy)-5-
(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-3-pyridinyl]-
benzenesulfonamide (37) (0.134 g, 0.315 mmol), K2CO3 (0.119 g,
0.860 mmol), PdCl2(dppf)−DCM (0.023 g, 0.029 mmol) in 1,4-
dioxane (5.97 mL) and water (5.97 mL) was heated at 80 °C, and the
mixture was stirred overnight. After cooling to room temperature, the
mixture was diluted with EtOAc and water. The combined organics
were washed with saturated NaHCO3, brine, dried over Na2SO4,
filtered, and concentrated. The residue was purified by reverse phase
chromatography (10−90% ACN/H2O) to give an unassigned 38:62
mixture of N-[5-(2-amino-3-phenyl-6-quinoxalinyl)-2-(methyloxy)-3-
pyridinyl]-2,4-difluorobenzenesulfonamide (7) and N-(5-(3-amino-2-
phenylquinoxalin-6-yl)-2-methoxypyridin-3-yl)-2,4-difluorobenzene-
sulfonamide (48) (0.103 g, 69%) as a light yellow solid. LC−MS m/z
= 520.2 (M + H)+. The isomers were separated on an Aglient
preparative HPLC instrument using 7.5 mL/min 30% IPA in hexane
on 10 mm ODH column to afford the first eluting isomer N-(5-(2-
amino-3-phenylquinoxalin-6-yl)-2-methoxypyridin-3-yl)-2,4-difluoro-
benzenesulfonamide (7) (12 mg): 1H NMR (400 MHz, DMSO-d6) δ
ppm 10.31 (br s, 1 H), 8.44 (s, 1 H), 7.95−8.05 (m, 2 H), 7.72−7.88
(m, 4 H), 7.63−7.68 (m, 1 H), 7.50−7.62 (m, 4 H), 7.17−7.26 (m, 1
H), 6.69 (br s, 2 H), 3.66 (s, 3 H); LC−MS m/z = 520.3 (M + H)+;
HRMS calculated for C26H20N5O3F2S (M + 1)+, 520.1255; found
520.1254. The second eluting isomer was N-(5-(3-amino-2-phenyl-
quinoxalin-6-yl)-2-methoxypyridin-3-yl)-2,4-difluorobenzenesulfona-
mide (48) (24 mg): 1H NMR (400 MHz, DMSO-d6) δ ppm 10.34 (br
s, 1 H), 8.45 (s, 1 H), 7.98−8.01 (m, 1 H), 7.87−7.91 (m, 1 H), 7.74−
7.83 (m, 4 H), 7.61−7.66 (m, 1 H), 7.52−7.60 (m, 4 H), 7.17−7.26
(m, 1 H), 6.67 (br s, 2 H), 3.68 (s, 3 H); LC−MS m/z = 520.3(M +
H)+; HRMS calculated for C26H20N5O3F2S (M + 1)+, 520.1255; found
520.1253. Absolute stereochemical assignments were made by using
13C chemical shifts and using ab initio IR analysis.
A microwave vial was charged with 6-bromo-3-phenyl-1,5-
naphthyridin-2-amine (44) (170 mg, 0.566 mmol), 2,4-difluoro-N-
(2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-
yl)benzenesulfonamide (37) (241 mg, 0.566 mmol), KOAc (167 mg,
1.699 mmol), and 1,4-dioxane (10 mL). The solution was degassed,
treated with Pd(PPh3)2Cl2 (39.7 mg, 0.057 mmol), placed under a
nitrogen atmosphere, and heated in the microwave at 120 °C for 95
min. The mixture was washed with water and extracted into EtOAc.
The organics were concentrated and prepurified by silica gel
chromatography (0−100% EtOAc/DCM) and was then purified by
reverse phase chromatography (10−100% ACN/H2O) to give N-[5-
(6-amino-7-phenyl-1,5-naphthyridin-2-yl)-2-(methyloxy)-3-pyridinyl]-
2,4-difluorobenzenesulfonamide (6) (20 mg, 7%) as a yellow solid. 1H
NMR (400 MHz, chloroform-d) δ ppm 8.46 (d, J = 8.8 Hz, 1 H), 8.21
(d, J = 1.4 Hz, 1 H), 8.01 (d, J = 1.6 Hz, 1 H), 7.80−7.91 (m, 2 H),
7.65−7.74 (m, 2 H), 7.56 (d, J = 9.2 Hz, 1 H), 7.47 (dd, J = 5.2, 1.7
Hz, 3 H), 7.38 (br s, 1 H), 6.90−7.00 (m, 2 H), 6.57 (br s, J = 8.0 Hz,
2 H), 3.97 (s, 3 H); LC−MS m/z = 520.4 (M + H)+; HRMS
calculated for C26H20N5O3F2S (M + 1)+, 520.1255; found 520.1257.
N-(5-(2-Amino-3-phenylquinoxalin-6-yl)-2-methoxypyridin-
3-yl)-2,4-difluorobenzenesulfonamide (7). A solution of 4-
bromo-1,2-benzenediamine (45) (6.7 g, 35.8 mmol) was dissolved
in degassed anhydrous MeOH (166 mL) and was purged with
nitrogen. The solution was treated with concentrated HCl (2.94 mL,
35.8 mmol), followed by benzaldehyde (3.62 mL, 35.8 mmol) and
1,1,3,3-tetramethylbutyl isocyanide (6.24 mL, 35.8 mmol). The
mixture was stirred at room temperature overnight and then
concentrated under reduced pressure. The residue was partitioned
between saturated NaHCO3 and CHCl3. The combined organics were
washed with brine, dried over MgSO4, filtered, and concentrated to
give 6-bromo-3-phenyl-N-(1,1,3,3-tetramethylbutyl)-1,4-dihydro-2-
quinoxalinamine (16.1 g, 100%, 72% purity) which was used without
further purification. LC−MS m/z = 414.3, 416.3 (M + H)+. A solution
of 6-bromo-3-phenyl-N-(1,1,3,3-tetramethylbutyl)-1,4-dihydro-2-qui-
noxalinamine (16.1 g, 38.8 mmol) in benzene (100 mL) was treated
by the addition of DDQ (8.80 g, 38.8 mmol) in benzene (200 mL).
The mixture was stirred at room temperature overnight. The dark
solution was concentrated onto Celite and purified by silica gel
chromatography (0−10% EtOAc/hexanes) to give an unassigned
38:62 mixture of 6-bromo-3-phenyl-N-(1,1,3,3-tetramethylbutyl)-2-
quinoxalinamine (46a) and 7-bromo-3-phenyl-N-(2,4,4-trimethylpen-
tan-2-yl)quinoxalin-2-amine (46b) (4.68 g, 29%) as a yellow oil. LC−
MS m/z = 412.3, 414.2 (M + H)+.
N-(5-(2-Amino-4-oxo-3-phenyl-3,4-dihydroquinazolin-6-yl)-
2-methoxypyridin-3-yl)-2,4-difluorobenzenesulfonamide (8).
A solution of isothiocyanatobenzene (22.7 mL, 190 mmol), 2-
amino-5-iodobenzoic acid (49, Y = I) (50 g, 190 mmol), and TEA
(39.7 mL, 285 mmol) in 1,4-dioxane (750 mL) was maintained at
reflux for 6 h. The mixture was cooled to room temperature, and solids
were filtered. The solids were resuspended in Et2O and filtered to
afford 6-iodo-3-phenyl-2-thioxo-2,3-dihydroquinazolin-4(1H)-one
1
(67.8 g, 178 mmol) as a white solid. H NMR (400 MHz, DMSO-
d6) δ ppm 13.10 (s, 1 H), 8.17 (d, J = 1.95 Hz, 1 H), 8.07 (dd, J =
8.63, 2.00 Hz, 1 H), 7.44−7.52 (m, 2 H), 7.37−7.44 (m, 1 H), 7.18−
7.30 (m, 3 H); LC−MS m/z = 380.83 (M + H)+. 6-Iodo-3-phenyl-2-
thioxo-2,3-dihydroquinazolin-4(1H)-one (34.8 g, 92 mmol) was added
to POCl3 (205 mL, 2197 mmol) at room temperature. The mixture
was treated with PCl5 (33.4 g, 160 mmol) in one portion and was
stirred at room temperature for 15 min, then heated at 110 °C
overnight. The mixture was allowed to cool to room temperature and
was concentrated under reduced pressure. The residue was slurried in
EtOAc, added portionwise to stirred ice−water, and stirred until all
solids had dissolved. The organic phase was separated, washed with
brine, dried over sodium sulfate, and concentrated. The residue was
triturated by stirring in Et2O and then filtered to give 2-chloro-6-iodo-
3-phenylquinazolin-4(3H)-one (50, Y = I) (24.5 g, 70%) as a solid. 1H
NMR (400 MHz, DMSO-d6) δ ppm 8.38 (d, 1 H), 8.20 (dd, 1 H),
7.44−7.68 (m, 6 H); LC−MS m/z = 382.9 (M + H)+.
The 38:62 mixture of 6-bromo-3-phenyl-N-(1,1,3,3-tetramethylbu-
tyl)-2-quinoxalinamine (46a) and 7-bromo-3-phenyl-N-(2,4,4-trime-
thylpentan-2-yl)quinoxalin-2-amine (46b) (4.68 g, 11.4 mmol) in
DCM (51.1 mL) was treated by the addition of 4 N in dioxanes HCl
(33 mL) and was heated to 70 °C overnight. The solvents were
evaporated under reduced pressure, and the residue was taken up in
EtOAc and treated with saturated NaHCO3. The combine extracts
were washed with brine, dried over Na2SO4, filtered, and concentrated
to give an unassigned 35:65 mixture of 6-bromo-3-phenyl-2-
quinoxalinamine (47a) and 7-bromo-3-phenylquinoxalin-2-amine
(47b) (3.20 g, 94%). LC−MS m/z = 300.1, 302.2 (M + H)+.
DIEA (22.4 mL, 128 mmol) and 4-methoxybenzylamine (10.9 mL,
83 mmol) were dissolved in DMF (200 mL), and 2-chloro-6-iodo-3-
phenylquinazolin-4(3H)-one (50, Y = I) (24.5 g, 64.0 mmol) was
added portionwise. The mixture was heated at 80 °C for 4 h and then
cooled to room temperature. The mixture was concentrated to 75 mL
and diluted with EtOAc. The organic phase was washed with 5% LiCl
(2×), brine, dried over sodium sulfate, and concentrated. The residue
was purified by silica chromatography (1:6:3 EtOAc/hexanes/DCM)
to give the 6-iodo-2-((4-methoxybenzyl)amino)-3-phenylquinazolin-
4(3H)-one (29.7 g, 96%) as a white solid. 1H NMR (400 MHz,
DMSO-d6) δ ppm 8.13 (d, 1 H), 7.85 (dd, 1 H), 7.50−7.67 (m, 3 H),
L
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
7.34−7.46 (m, 2 H), 7.22 (d, 2 H), 7.09 (d, 1 H), 6.78−6.91 (m, 2 H),
6.46 (s, 1 H), 4.42 (d, 2 H), 3.70 (s, 3 H); LC−MS m/z = 484.1 (M +
H)+. 6-Iodo-2-((4-methoxybenzyl)amino)-3-phenylquinazolin-4(3H)-
one (29.7 g, 61.5 mmol) was added to TFA (300 mL) and heated at
reflux for 2 days. The mixture was cooled to room temperature and
was concentrated. The residue was taken up in DCM and poured
portionwise into an ice/saturated NaHCO3. The organic phase, with
suspended solids, was separated and the aqueous phase extracted with
DCM. The combined organic phases were diluted with Et2O to a ratio
of 1:1 DCM/Et2O. The solids were collected by filtration, washed with
Et2O, and dried under vacuum to give 2-amino-6-iodo-3-phenyl-
quinazolin-4(3H)-one (56, Y = I) (10.2 g, 98% purity, 45%). The
filtrate was concentrated and the residue slurried in 1:1 DCM/Et2O.
The solids were collected and dried under vacuum to give additional 2-
amino-6-iodo-3-phenylquinazolin-4(3H)-one (56, Y = I) (7.9 g, 95%
purity, 34%). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.12 (d, J = 2.15
Hz, 1 H), 7.85 (dd, J = 8.63, 2.20 Hz, 1 H), 7.48−7.63 (m, 3 H),
7.30−7.39 (m, 2 H), 7.05 (d, J = 8.68 Hz, 1 H), 6.42 (br s, 2 H); LC−
MS m/z = 363.90 (M + H)+.
A solution of 2-amino-6-iodo-3-phenylquinazolin-4(3H)-one (56, Y
= I) (0.150 g, 0.413 mmol), 2,4-difluoro-N-(2-methoxy-5-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-yl)benzenesulfonamide
(37) (0.194 g, 0.454 mmol), Cs2CO3 (0.404 g, 1.24 mmol), and
PdCl2(dppf)-DCM (0.034 g, 0.041 mmol) in 1,4-dioxane (1.65 mL)
and water (0.413 mL) was heated at 80 °C for 1 h. After cooling to
room temperature, the solution was diluted with EtOAc and washed
with water, brine, dried over Na2SO4, and concentrated. The residue
was loaded onto Celite and purified by silica gel chromatography (50−
100% EtOAc/hexanes) and was triturated with Et2O to give N-(5-(2-
amino-4-oxo-3-phenyl-3,4-dihydroquinazolin-6-yl)-2-methoxypyridin-
3-yl)-2,4-difluorobenzenesulfonamide (8) (0.143 g, 65%) as a pale tan
solid. 1H NMR (400 MHz, DMSO-d6) δ ppm 10.30 (br s, 1 H), 8.33
(d, J = 2.15 Hz, 1 H), 8.02 (d, J = 2.34 Hz, 1 H), 7.89 (dd, J = 8.59,
2.34 Hz, 1 H), 7.85 (d, J = 2.34 Hz, 1 H), 7.72−7.80 (m, 1 H), 7.50−
7.64 (m, 4 H), 7.38 (d, J = 7.22 Hz, 2 H), 7.34 (d, J = 8.59 Hz, 1 H),
7.21 (t, 1 H), 6.39 (br s, 2 H), 3.65 (s, 3 H); LC−MS m/z = 536.14
(M + H)+; HRMS calculated for C26H20N5O4F2S (M + 1)+, 536.1204;
found 536.1202.
1 H), 6.31 (br s, 2 H), 5.09 (s, 2 H), 3.89 (s, 3 H); LC−MS m/z =
360.1 (M + H)+.
A solution of MsCl (128 mg, 1.12 mmol) and 2-amino-6-[5-amino-
6-(methyloxy)-3-pyridinyl]-3-phenyl-4(3H)-quinazolinone (63) (50
mg, 0.139 mmol) in pyridine (1 mL) was stirred at room temperature
for 1 day. The resulting mixture was treated with MeOH (1 mL) and
concentrated. The residue was dissolved in DMF and purified by
reverse phase chromatography (10−90% ACN/H2O) to obtain N-[5-
(2-amino-4-oxo-3-phenyl-3,4-dihydro-6-quinazolinyl)-2-(methyloxy)-
3-pyridinyl]methanesulfonamide (12) (17 mg, 28%) as a white solid:
1H NMR (400 MHz, DMSO-d6) δ ppm 3.07 (s, 3 H), 3.95 (s, 3 H),
6.37 (br s, 2 H), 7.29−7.42 (m, 3 H), 7.48−7.63 (m, 3 H), 7.87 (d, J =
2.24 Hz, 1 H), 7.92 (dd, J = 8.58, 2.34 Hz, 1 H), 8.05 (d, J = 2.24 Hz, 1
H), 8.29 (d, J = 1.76 Hz, 1 H), 9.38 (br s, 1 H); LC−MS m/z = 438.4
(M + H)+; HRMS calculated for C21H20N5O4S (M + 1)+, 438.1236;
found 438.1237.
5-(2-Amino-4-oxo-3-phenyl-3,4-dihydroquinazolin-6-yl)-N-
(2,4-difluorophenyl)-2-methoxypyridine-3-sulfonamide (20).
Step a. SOCl2 (14 mL, 192 mmol) was added dropwise over 60 min
to water (83 mL) and cooled to 0 °C, maintaining the temperature of
the mixture at 0−7 °C. The solution was allowed to warm to 18 °C
over 17 h. Copper(I) chloride (51 mg, 0.515 mmol) was added to the
mixture, and the resultant yellow-green solution was cooled to −3 °C
using an acetone/ice bath.
Step b. HCl (45 mL, 1480 mmol) was added, with agitation, to 5-
bromo-2-chloro-3-pyridinamine (70) (9.3 g, 44.8 mmol), maintaining
the temperature of the mixture below 30 °C with ice cooling. The
reaction mixture was cooled to −5 °C using an ice−acetone bath, and
a solution of sodium nitrite (3.33 g, 48.3 mmol) in water (13 mL) was
added dropwise over 15 min, maintaining the temperature of the
reaction mixture between −5 to 0 °C. The resultant slurry was cooled
to −2 °C and stirred for 10 min.
Step c. The slurry from step b was cooled to −5 °C and added to the
solution obtained from step a over 30 min, maintaining the
temperature of the reaction mixture between −3 to 0 °C (the slurry
from step b was maintained at −5 °C throughout the addition). As the
reaction proceeded, a solid began to precipitate. When the addition
was complete, the reaction mixture was agitated at 0 °C for 75 min.
The suspended solid was filtered, washed with water, and dried under
reduced pressure to give 5-bromo-2-chloro-3-pyridinesulfonyl chloride
(71) (10 g, 77%). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.52 (d, J =
2.54 Hz, 1 H), 8.30 (d, J = 2.54 Hz, 1 H).
A solution of 5-bromo-2-chloro-3-pyridinesulfonyl chloride (71)
(20.0 g, 68.7 mmol) in DCM (200 mL) was treated with pyridine
(6.12 mL, 76 mmol) and 2,4-difluoroaniline (7.61 mL, 76 mmol). The
mixture was stirred at room temperature overnight, and the volatiles
were removed under reduced pressure. The residue was taken up in
MeOH (200 mL), and 25% NaOMe/MeOH (200 mL, 896 mmol)
solution was added. The mixture was heated at 50 °C for 4 h. The
volatiles were removed under reduced pressure. To the dark oily
residue icy water was added (400 mL) and was extracted with EtOAc
(200 mL and 2 × 100 mL). The combined organics were washed with
brine (2 × 100 mL), dried over MgSO4, and evaporated to give 5-
bromo-N-(2,4-difluorophenyl)-2-methoxypyridine-3-sulfonamide (72)
(24.26g, 93%). 1H NMR (400 MHz, DMSO-d6) δ = 10.27 (br s, 1 H),
8.48 (d, J = 2.4 Hz, 1 H), 8.06 (d, J = 2.4 Hz, 1 H), 7.24 (td, J = 6.2,
9.1 Hz, 1 H), 7.19−7.10 (m, 1 H), 7.00−6.89 (m, 1 H), 3.89 (s, 3 H);
LC−MS m/z = 377.1, 379.1 (M − H)−.
N-[5-(2-Amino-4-oxo-3-phenyl-3,4-dihydro-6-quinazolinyl)-
2-(methyloxy)-3-pyridinyl]methanesulfonamide (12). A solu-
tion of 5-bromo-2-(methyloxy)-3-pyridinamine (30) (Small Mole-
cules, Inc., Hoboken, NJ) (5.0 g, 24.6 mmol), 4,4,4′,4′,5,5,5′,5′-
octamethyl-2,2′-bi-1,3,2-dioxaborolane (7.5 g, 29.6 mmol),
PdCl2(dppf)−CH2Cl2 (2.0 g, 2.46 mmol), and KOAc (7.25 g, 73.9
mmol) in anhydrous dioxane (120 mL) was purged with nitrogen and
heated to 100 °C for 18 h. The mixture was concentrated in vacuo.
The residue was diluted with EtOAc (500 mL) and filtered through
Celite. The filtrate was washed with cold water (250 mL), brine, dried
over MgSO4, and concentrated in vacuo. The residue was purified by
silica gel chromatography (70−100% EtOAc/hexanes) to afford 2-
methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-
1
amine (40) (4.38 g, 71%). H NMR (400 MHz, DMSO-d6) δ ppm
7.64 (d, J = 1.51 Hz, 1 H), 7.12 (d, J = 1.76 Hz, 1 H), 4.88 (s, 2 H),
3.87 (s, 3 H), 1.26 (s, 12 H); LC−MS m/z = 250.8 (M + H)+.
A solution 2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-
yl)pyridin-3-amine (40) (4.38 g, 17.6 mmol), 2-amino-6-iodo-3-
phenyl-4(3H)-quinazolinone (56, Y = I) (5.8 g, 16.0 mmol), Cs2CO3
(15.6 g, 47.9 mmol), and PdCl2(dppf)−CH2Cl2 (1.30 g, 1.6 mmol) in
THF (60 mL) and water (20 mL) was purged with nitrogen and
heated to 65 °C for 1 h. The reaction was concentrated in vacuo, and
the residue was diluted with ethyl acetate (500 mL) and filtered
through Celite. The filtrate was washed with water, brine, dried over
MgSO4, and concentrated in vacuo. The residue was triturated in hot
ACN (60 mL), cooled to room temperature, and stirred for 30 min.
The solids were filtered to give 2-amino-6-(5-amino-6-methoxypyr-
idin-3-yl)-3-phenylquinazolin-4(3H)-one (63) (2.30 g, 40%). 1H
NMR (400 MHz, DMSO-d6) δ ppm 8.00 (d, J = 2.3 Hz, 1 H), 7.85
(dd, J = 8.6, 2.3 Hz, 1 H), 7.68 (d, J = 2.1 Hz, 1 H), 7.46−7.64 (m, 3
H), 7.35−7.41 (m, 2 H), 7.31 (d, J = 8.4 Hz, 1 H), 7.18 (d, J = 2.1 Hz,
A flask was charged with 5-bromo-N-(2,4-difluorophenyl)-2-
methoxypyridine-3-sulfonamide (72) (13.12 g, 34.6 mmol),
4,4,4′,4′,5,5,5′,5′-octamethyl-2,2′-bi(1,3,2-dioxaborolane) (10.54 g,
41.5 mmol), KOAc (10.19 g, 104 mmol), PdCl2(dppf)−CH2Cl2
(2.83 g, 3.46 mmol), and 1,4-dioxane (173 mL). The mixture was
degassed by bubbling nitrogen through for 15 min, and then the
mixture was heated 95 °C for 2 h. The black liquid was filtered
through a Celite pad and was evaporated to dryness. The black
semisolid product was purified by silica gel chromatography (0−40%
EtOAc/hexane) to give N-(2,4-difluorophenyl)-2-methoxy-5-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine-3-sulfonamide (73)
1
(8.89g, 60%) as a pale yellow solid. H NMR (400 MHz, DMSO-
M
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
d6) δ 10.12 (s, 1H), 8.57 (d, J = 1.86 Hz, 1H), 8.10 (d, J = 1.76 Hz,
1H), 7.17−7.40 (m, 2H), 6.93−7.13 (m, 1H), 3.97 (s, 3H), 1.28 (s,
12H); LC−MS 344.89 m/z = (M + H)+.
DMSO-d6) δ ppm 8.72 (d, J = 2.1 Hz, 1 H), 8.10−8.23 (m, 2 H), 8.02
(d, J = 2.3 Hz, 1 H), 7.85 (dd, J = 8.7, 2.2 Hz, 1 H), 7.31 (td, J = 9.0,
6.2 Hz, 1 H), 7.24 (d, J = 8.6 Hz, 2 H), 7.17 (s, 2 H), 6.97−7.07 (m, 1
H), 3.99−4.09 (m, 2 H), 3.96 (s, 3 H), 1.17 (t, J = 6.9 Hz, 3 H); LC−
MS m/z = 488.3 (M + H)+; HRMS calculated for C22H20N5O4F2S (M
+ 1)+, 488.1204; found 488.1200.
N-(2,4-Difluorophenyl)-2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-di-
oxaborolan-2-yl)pyridine-3-sulfonamide (73) (117 mg, 0.275 mmol),
2-amino-6-iodo-3-phenyl-4(3H)-quinazolinone (56) (100 mg, 0.275
mmol), PdCl2(dppf)-CH2Cl2 (22.49 mg, 0.028 mmol) were purged
with nitrogen. The mixture was treated with 2 M sodium carbonate
(500 μL, 1.00 mmol) and 1,4-dioxane (2 mL), and nitrogen was
bubbled through the mixture. The mixture was heated at 80 °C for 2 h.
After the mixture was cooled, silica gel was added and the volatiles
were removed under reduced pressure. The residue was purified by
column chromatography (DCM/EtOAc) to afford 5-(2-amino-4-oxo-
3-phenyl-3,4-dihydroquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-me-
thoxypyridine-3-sulfonamide (20) (111 mg, 75%). 1H NMR (DMSO-
d6) δ: 10.19 (s, 1H), 8.76 (d, J = 2.3 Hz, 1H), 8.17 (d, J = 2.5 Hz, 1H),
8.02 (d, J = 2.3 Hz, 1H), 7.91 (dd, J = 8.7, 2.2 Hz, 1H), 7.47−7.69 (m,
4H), 7.29−7.41 (m, 4H), 6.97−7.11 (m, 1H), 6.42 (br s, 2H), 3.97 (s,
3H); LC−MS m/z = 536.4 (M + H)+.
5-(2-Amino-4-oxo-3-(2-(trifluoromethyl)phenyl)-3,4-dihy-
droquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-methoxypyri-
dine-3-sulfonamide (23) and (aS)-5-(2-Amino-4-oxo-3-(2-
(trifluoromethyl)phenyl)-3,4-dihydroquinazolin-6-yl)-N-(2,4-
difluorophenyl)-2-methoxypyridine-3-sulfonamide (28) and
(aR)- 5-(2-Amino-4-oxo-3-(2-(trifluoromethyl)phenyl)-3,4-dihy-
droquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-methoxypyri-
dine-3-sulfonamide (29). A solution of 2-amino-5-iodobenzoic acid
(49, Y = I) (35.0 g, 133 mmol), 1-isothiocyanato-2-(trifluoromethyl)-
benzene (24.1 mL, 160 mmol), and TEA (26.0 mL, 186 mmol) in
EtOH (507 mL) was heated at reflux overnight. After the mixture was
cooled to room temperature, a small amount of black solids were
filtered, rinsed with EtOH, and the solids were discarded. The filtrate
was concentrated and the brown residue was taken up in DCM and
washed with saturated NaHCO3. Solids crashed out and were filtered
and rinsed with DCM. The gray solids were taken up in EtOH and
sonicated. The solids were filtered, rinsed with cold EtOH, and dried
under reduced pressure at 45 °C overnight to give 6-iodo-2-thioxo-3-
(2-(trifluoromethyl)phenyl)-2,3-dihydroquinazolin-4(1H)-one (27.28
5-(2-Amino-3-ethyl-4-oxo-3,4-dihydro-6-quinazolinyl)-N-
(2,4-difluorophenyl)-2-(methyloxy)-3-pyridinesulfonamide
(21). Methyl 2-amino-5-bromobenzoate (64) (6.63 g, 28.8 mmol) in
pyridine (53 mL) was treated with diphenyl cyanocarbodimidate (6.87
g, 28.8 mmol) and heated to 50 °C overnight. The mixture was
concentrated under reduced pressure. The residue was taken up in
DCM and 5% K2CO3 (200 mL). The aqueous portion was extracted
with DCM. The combined organics were washed with brine (2×),
dried over MgSO4, filtered, and concentrated to give an oil that began
to solidify. The residue was then treated with 1:1 Et2O/hexanes (100
mL). The solids were filtered and rinsed with 1:1 Et2O/hexanes to
give methyl 5-bromo-2-{[(cyanoamino)(phenyloxy)methylidene]-
1
g, 46%). H NMR (400 MHz, DMSO-d6) δ ppm 13.28 (br s, 1 H),
8.16 (d, J = 1.95 Hz, 1 H), 8.04 (dd, J = 8.54, 1.51 Hz, 1 H), 7.75−7.85
(m, 2 H), 7.59−7.67 (m, 1 H), 7.50 (d, J = 7.80 Hz, 1 H), 7.20 (d, J =
8.68 Hz, 1 H); LC−MS m/z = 499.0 (M + H)+. A suspension of 6-
iodo-2-thioxo-3-(2-(trifluoromethyl)phenyl)-2,3-dihydroquinazolin-
4(1H)-one (15 g, 33.5 mmol) in POCl3 (74.9 mL, 803 mmol) was
treated with PCl5 (11.15 g, 53.3 mmol), and the mixture was heated to
reflux overnight. A second flask containing a suspension of 6-iodo-2-
thioxo-3-(2-(trifluoromethyl)phenyl)-2,3-dihydroquinazolin-4(1H)-
one (12.28 g, 27.4 mmol) in POCl3 (61.3 mL, 658 mmol) was treated
with PCl5 (9.13 g, 43.8 mmol), and the mixture was heated to reflux
overnight. The two reactions were concentrated under reduced
pressure. The residues were taken up in EtOAc and combined. The
combined organics were washed with water (3×), brine, dried over
MgSO4, filtered, and concentrated to give 2-chloro-6-iodo-3-(2-
(trifluoromethyl)phenyl)quinazolin-4(3H)-one (51, Y = I) (26.1 g,
95%) as a light beige solid. LC−MS m/z = 450.8 (M + H)+.
A solution of 2-chloro-6-iodo-3-(2-(trifluoromethyl)phenyl)-
quinazolin-4(3H)-one (51, Y = I) (26.1 g, 57.9 mmol) and DIEA
(15.2 mL, 87 mmol) in DMF (265 mL) was treated by the addition of
4-methoxybenzylamine (9.07 mL, 69.4 mmol) and was heated to 80
°C overnight. The mixture was diluted with EtOAc and washed with
5% LiCl (5×), brine, dried over MgSO4, filtered, and concentrated to
give a dark brownish yellow solid. This residue was purified with silica
gel chromatography (100% DCM) to give 6-iodo-2-((4-
methoxybenzyl)amino)-3-(2-(trifluoromethyl)phenyl)quinazolin-
4(3H)-one (14.34 g, 94% purity) as a solid. LC−MS m/z = 552.0 (M
+ H)+. The impure fractions (∼16 g) were purified by silica gel
chromatography (10−30% then 50% EtOAc/hexanes) to give
additional 6-iodo-2-((4-methoxybenzyl)amino)-3-(2-
(trifluoromethyl)phenyl)quinazolin-4(3H)-one (11.08 g). LC−MS
m/z = 552.0 (M + H)+. The combined overall recovery was 25.4 g
and 80% yield. A solution 6-iodo-2-((4-methoxybenzyl)amino)-3-(2-
(trifluoromethyl)phenyl)quinazolin-4(3H)-one (21.8 g, 39.5 mmol) in
TFA (198 mL) was split among 12 microwave flasks and was heated to
140 °C in the microwave for 20 min. After cooling to room
temperature, the portions were combined and the solvents were
removed under reduced pressure and the residue was taken up in
DCM. The solution was washed with saturated NaHCO3 (3×) during
which solids formed, the solution/solid was washed with water, and
the white solids were filtered to give 2-amino-6-iodo-3-(2-
(trifluoromethyl)phenyl)quinazolin-4(3H)-one (57, Y = I) (13.5 g,
79%). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.11 (d, J = 1.95 Hz, 1
H), 7.94 (d, J = 7.82 Hz, 1 H), 7.84−7.90 (m, 2 H), 7.72−7.78 (m, 1
1
amino}benzoate (65) as a white solid (6.95 g, 64%). H NMR (400
MHz, DMSO-d6) δ ppm 10.83 (br s, 1 H), 8.02 (d, J = 2.15 Hz, 1 H),
7.89 (dd, J = 8.60, 2.34 Hz, 1 H), 7.63 (d, J = 8.60 Hz, 1 H), 7.43−7.49
(m, 2 H), 7.30−7.37 (m, 1 H), 7.24 (d, J = 8.01 Hz, 2 H), 3.87 (s, 3
H); LC−MS m/z = 374.1, 376.1 (M + H)+.
Methyl 5-bromo-2-(((cyanoimino)(phenoxy)methyl)amino)-
benzoate (65) (200 mg, 0.534 mmol) was placed in a sealed tube
with i-PrOH (2.5 mL) and 2 M ethylamine in THF (0.267 mL, 1.07
mmol). The mixture was heated to 85 °C for 1 h. The solvents were
removed under reduced pressure to give N-(6-bromo-3-ethyl-4-oxo-
3,4-dihydroquinazolin-2-yl)cyanamide (66) (164 mg, 100%) which
was used without further purification. 1H NMR (400 MHz, DMSO-d6)
δ ppm 7.92 (br s, 1 H), 7.11−7.19 (m, 1 H), 6.70−6.80 (m, 1 H), 4.01
(q, J = 6.91 Hz, 2 H), 1.08−1.16 (m, 3 H); LC−MS m/z = 305.3,
307.3 (M + H)+.
N-(6-Bromo-3-ethyl-4-oxo-3,4-dihydroquinazolin-2-yl)cyanamide
(66) (156 mg, 0.534 mmol) was placed in a sealed tube followed by
DMSO (1.79 mL) and concentrated HCl (3.56 mL). The mixture was
heated to 100 °C for 1 h. The mixture was poured into water (20 mL),
treated with K2CO3 until gas evolution ceased, and extracted with
DCM (2 × 25 mL). The solvents were then removed under reduced
pressure, and the residue was dissolved into EtOAc (25 mL), washed
with water (12 mL), brine, dried over MgSO4, and concentrated to
give 2-amino-6-bromo-3-ethylquinazolin-4(3H)-one (68) (0.143 g,
100%). 1H NMR (400 MHz, DMSO-d6) δ ppm 7.94 (d, J = 2.34 Hz, 1
H), 7.66 (dd, J = 8.74, 2.49 Hz, 1 H), 7.19 (br s, 2 H), 7.11 (d, J = 8.89
Hz, 1 H), 3.97−4.05 (m, 2 H), 1.10−1.21 (m, 3 H); LC−MS m/z =
268.3, 270.3 (M + H)+.
PdCl2(dppf)−CH2Cl2 (9.14 mg, 0.011 mmol), KOAc (43.9 mg,
0.448 mmol), and N-(2,4-difluorophenyl)-2-methoxy-5-(4,4,5,5-tetra-
methyl-1,3,2-dioxaborolan-2-yl)pyridine-3-sulfonamide (73) (52.5 mg,
0.123 mmol) were combined with 2-amino-6-bromo-3-ethylquinazo-
lin-4(3H)-one (68) (30 mg, 0.112 mmol) in dioxane (0.8 mL) and
water (0.267 mL). The mixture was degassed and heated to 100 °C in
a sealed tube. After 1 h, the mixture was cooled and the solution was
filtered to remove any solids and purified by reverse phase
chromatography (10−90% ACN/water) to give 5-(2-amino-3-ethyl-
4-oxo-3,4-dihydroquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-methoxy-
pyridine-3-sulfonamide (21) (30 mg, 52%). 1H NMR (400 MHz,
N
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
H), 7.62 (d, J = 7.82 Hz, 1 H), 7.06 (d, J = 8.70 Hz, 1 H), 6.69 (br s, 2
H); LC−MS m/z = 432.0 (M + H)+.
Present Address
§A.M.: Salix Pharmaceuticals Inc., 8510 Colonnade Center
Drive, Raleigh, NC 27615
A mixture of 2-amino-6-iodo-3-(2-(trifluoromethyl)phenyl)-
quinazolin-4(3H)-one (7.75 g, 18.0 mmol) (57, Y = I), N-(2,4-
difluorophenyl)-2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-
2-yl)pyridine-3-sulfonamide (8.43 g, 19.8 mmol) (73), PdCl2(dppf)−
CH2Cl2 (1.47 g, 1.80 mmol), Cs2CO3 (17.6 g, 53.9 mmol) was treated
with 1,4-dioxane (71.9 mL) and water (18.0 mL) and heated to 80 °C
for 1 h. After cooling to room temperature, the mixture was diluted
with EtOAc and water. The combine organics were washed with brine,
dried over Na2SO4, filtered through a plug of silica, and concentrated
to give dark brownish/black residue. The residue was taken up in
DCM, loaded directly onto a silica gel column, and prepurified (1−3%
MeOH/DCM) to give a peach solid (9.47 g) and the mixed fractions
(4.378 g) as an orange solid. The residues, the peach solid (8.1 g) and
the orange solid (4.378 g), were separately loaded onto silica gel and
separately purified by silica gel chromatography (50−100% EtOAc/
hexane) to give a pale yellow solid foam (6 and 3 g), respectively. The
batches were combined (9 g), and the yellow foam was sonicated in
Et2O for 15 min. The solid was filtered off, washed with ether, and
dried overnight at room temperature under high vacuum to give a
white fluffy powder. The product was dried overnight under reduced
pressure at 60 °C to give 5-(2-amino-4-oxo-3-(2-(trifluoromethyl)-
phenyl)-3,4-dihydroquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-me-
Author Contributions
Chemistry experiments: A.L.L., M.R.L., M.T., and J.B..
Modeling and compound design: S.D. Assay development
and assays: O.B.M., J.G., and K.L.C. Enzymology experiments:
S.L.S. DMPK experiments: A.M. and S.R. Contribution to or
assistance in preparation of the manuscript: A.L.L., J.B., S.D.,
O.B.M., S.L.S., C.B.M., S.R., and J.B.S.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank George Adjabeng, Maosheng Duan, Xiaofei Li,
Robert McFadyen, Wenyan Mo, Aniko Redman, and Jianjun
Xue for the synthetic work; Douglas Minick and Timothy
Spitzer for structure assignments; Rachel Schmidt, Eric Wentz,
and Yang Zhao for separation and purification of key
compounds; Zhi Hong, Robert Hamatake, Christopher
Roberts, and Andrew Spaltenstein for overall contribution of
project design.
1
thoxypyridine-3-sulfonamide (23) (8.1 g, 75%) as a white solid. H
NMR (400 MHz, DMSO-d6) δ ppm 10.21 (br s, 1 H), 8.74 (s, 1 H),
8.17 (d, J = 2.54 Hz, 1 H), 8.01 (d, J = 2.34 Hz, 1 H), 7.85−7.98 (m, 3
H), 7.73−7.79 (m, 1 H), 7.62 (d, J = 7.80 Hz, 1 H), 7.26−7.36 (m, 2
H), 7.18−7.26 (m, 1 H), 6.97−7.05 (m, 1 H), 6.68 (br s, 2 H), 3.96 (s,
3 H); LC−MS m/z = 604.4 (M + H)+; HRMS calculated for
C27H19N5O4F5S (M + 1)+, 604.1078; found 604.1077.
ABBREVIATIONS USED
■
LK, lipid kinase; PI4KIIIα, type III phosphatidylinositol 4-
kinase α; PI4KIIIβ, type III phosphatidylinositol 4-kinase β;
PI3Kα, type I phosphatidylinositol 3-kinase α; PI3Kβ, type I
phosphatidylinositol 3-kinase β; PI3Kδ, type I phosphatidyli-
nositol 3-kinase δ; PI3Kγ, type I phosphatidylinositol 3-kinase
γ; PK, protein kinase
Separation of 5-(2-amino-4-oxo-3-(2-(trifluoromethyl)phenyl)-3,4-
dihydroquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-methoxypyridine-3-
sulfonamide (23) was performed by dissolving 5 g of material in 100
mL of CHCl3/MeOH (3:1) with a couple drops of formic acid, then
separated by a preparative SFC method, 40% MeOH/CO2, 100 bar,
40 °C, 90 g/min, 220 nm, AD 25 cm × 3 cm, 83 mg/inj, to afford first
eluting isomer (aS)-5-(2-amino-4-oxo-3-(2-(trifluoromethyl)phenyl)-
3,4-dihydroquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-methoxypyri-
dine-3-sulfonamide (28) (2.1 g, >99% ee): 1H NMR (400 MHz,
DMSO-d6) δ ppm 10.19 (br s, 1 H), 8.76 (d, J = 2.44 Hz, 1 H), 8.18
(d, J = 2.44 Hz, 1 H), 8.02 (d, J = 2.15 Hz, 1 H), 7.86−7.98 (m, 3 H),
7.73−7.81 (m, 1 H), 7.63 (d, J = 7.82 Hz, 1 H), 7.29−7.36 (m, 2 H),
7.20−7.28 (m, 1 H), 6.98−7.07 (m, 1 H), 6.67 (br s, 2 H), 3.98 (s, 3
H); LC−MS m/z = 604.2 (M + H)+. There was also a second eluting
isomer (aR)-5-(2-amino-4-oxo-3-(2-(trifluoromethyl)phenyl)-3,4-di-
hydroquinazolin-6-yl)-N-(2,4-difluorophenyl)-2-methoxypyridine-3-
sulfonamide (29) (2.22 g, >99% ee): 1H NMR (400 MHz, DMSO-d6)
δ ppm 10.18 (br s, 1 H), 8.75 (d, J = 2.34 Hz, 1 H), 8.17 (d, J = 2.44
Hz, 1 H), 8.01 (d, J = 2.15 Hz, 1 H), 7.86−7.98 (m, 3 H), 7.72−7.80
(m, 1 H), 7.59−7.66 (m, 1 H), 7.27−7.36 (m, 2 H), 7.18−7.27 (m, 1
H), 6.97−7.05 (m, 1 H), 6.67 (br s, 2 H), 3.97 (s, 3 H); LC−MS m/z
= 604.2 (M + H)+); HRMS calculated for C27H19N5O4F5S (M + 1)+,
604.1078; found 604.1077. Absolute stereochemical assignments were
made by VCD Analysis.24
REFERENCES
■
(1) Reiss, S.; Rebhan, I.; Backes, P.; Romero-Brey, I.; Erfle, H.;
Matula, P.; Kaderali, L.; Poenisch, M.; Blankenburg, H.; Hiet, M. S.;
Longerich, T.; Diehl, S.; Ramirez, F.; Balla, T.; Rohr, K.; Kaul, A.;
Buhler, S.; Pepperkok, R.; Lengauer, T.; Albrecht, M.; Eils, R.;
̈
Schirmacher, P.; Lohmann, V.; Bartenschlager, R. Recruitment and
Activation of a Lipid Kinase by Hepatitis C Virus NS5A Is Essential for
Integrity of the Membranous Replication Compartment. Cell Host
Microbe 2011, 9, 32−45.
(2) Choo, Q. L.; Kuo, G.; Weiner, A.; Overby, L.; Bradley, D.;
Houghton, M. Isolation of a cDNA Clone Derived from a Flood-
Borne Non-A, Non-B Viral Hepatitis Genome. Science 1989, 244,
359−362.
(3) Lavanchy, D. The Global Burden of Hepatitis C. Liver Int. 2009,
29 (Suppl. 1), 74−81.
(4) Njoroge, F.; Chen, K.; Shih, N.; Piwinski, J. Challenges in
Modern Drug Discovery: A Case Study of Boceprevir, an HCV
Protease Inhibitor for the Treatment of Hepatitis C Virus Infection.
Acc. Chem. Res. 2008, 41, 50−59.
(5) Lin, C.; Kwong, A.; Perni, R. Discovery and Development of VX-
950, a Novel, Covalent, and Reversible Inhibitor of Hepatitis C Virus
NS3/4A Serine Protease. Infect. Disord.: Drug Targets 2006, 6, 3−16.
(6) Pearlman, B. Hepatitis C Treatment Update. Am. J. Med. 2004,
117, 344−352.
(7) Meanwell, N.; Belema, M. Hepatitis C VirusProgress toward
Inhibiting the Nonenzymatic Viral Proteins. Annu. Rep. Med. Chem.
2011, 46, 263−282.
(8) Gregory, M.; Bobardt, M.; Obeid, S.; Chatterji, U.; Coates, N.;
Foster, T.; Gallay, P.; Leyssen, P.; Moss, S.; Neyts, J.; Nur-e-Alam, M.;
Paeshuyse, J.; Piraee, M.; Suthar, D.; Warneck, T.; Zhang, M. Q.;
Wilkinson, B. Preclinical Characterization of Naturally Occurring
Polyketide Cyclophilin Inhibitors from the Sanglifehrin Family.
Antimicrob. Agents Chemother. 2011, 55, 1975−1981.
ASSOCIATED CONTENT
■
S
* Supporting Information
Procedures for enzyme and replicon assays, DMPK studies, and
the synthesis of compounds not described under the
Experimental Section. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Phone +919-483-3745. E-mail: janos.4.botyanszki@gsk.com.
O
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(9) Meanwell, N.; Kadow, J.; Scola, P. Progress towards the
Discovery and Development of Specifically Targeted Inhibitors of
Hepatitis C Virus. Annu. Rep. Med. Chem. 2009, 44, 379−440.
(10) Berger, K.; Cooper, J.; Heaton, N.; Yoon, R.; Oakland, T.;
Jordan, T.; Mateu, G.; Grakoui, A.; Randall, G. Roles for Endocytic
Trafficking and Phosphatidylinositol 4-Kinase III Alpha in Hepatitis C
Virus Replication. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 7577−7582.
(11) Vaillancourt, F. H.; Pilote, L.; Cartier, M.; Lippens, J.; Luizzi,
M.; Bethell, R.; Cordingley, M. G.; Kukolj, G. Identification of a Lipid
Kinase as a Host Factor Involved in Hepatitis C Virus RNA
Replication. Virology 2009, 387, 5−10.
(12) Li, Q.; Brass, A.; Ng, A.; Hu, Z.; Xavier, R.; Liang, T.; Elledge, S.
A Genome-Wide Genetic Screen for Host Factors Required for
Hepatitis C Virus Propagation. Proc. Natl. Acad. Sci. U.S.A. 2009, 106,
16410−16415.
(13) Tai, A.; Benita, Y.; Peng, L.; Kim, S.; Sakamoto, N.; Xavier, R.;
Chung, R. A Functional Genomic Screen Identifies Cellular Cofactors
of Hepatitis C Virus Replication. Cell Host Microbe 2009, 5, 298−307.
(14) Bianco, A.; Reghellin, V.; Donnici, L.; Fenu, S.; Alvarez, R.;
Baruffa, C.; Peri, F.; Pagani, M.; Abrignani, S.; Neddermann, P.; De
Francesco, R. Metabolis of Phosphatidylinositol 4-Kinase IIIα-
Dependent PI4P Is Subverted by HCV and Is Targeted by 4-Anilino
Quinazoline with Antiviral Activity. PLoS Pathog. 2012, 8 (3),
e1002576.
(15) Reiss, S.; Harak, C.; Romero-Brey, I.; Radujkovic, D.; Klein, R.;
Ruggieri, A.; Rebhan, I.; Bartenschlager, R.; Lohmann, V. The Lipid
Kinase Phosphatidylinositol-4 Kinase III Alpha Regulates the
Phosphorylation Status of Hepatitis C Virus NS5A. PLoS Pathog.
2013, 9, e1003359.
(16) Gehrmann, T.; Heilmeyer, L., Jr. Phosphatidylinositol 4-Kinase.
Eur. J. Biochem. 1998, 253, 357−370.
(17) Brown, J.; Auger, K. Phylogenomic of Phosphoinositide Lipid
Kinases: Perspectives on the Evolution of Second Messenger Signaling
and Drug Discovery. BMC Evol. Biol. 2011, 11 (4), 1471−2148.
(18) Courtney, K.; Corcoran, R.; Engleman, J. The PI3K Pathway as
a Drug Target in Human Cancer. J. Clin. Oncol. 2010, 28 (6), 1075−
1083.
(19) Katso, R.; Okkenhaug, K.; Ahmandi, K.; White, S.; Timms, J.;
Waterfield, M. Cellular Function of Phosphoinositide 3-Kinase:
Implications for Development, Immunity, Homeostasis, and Cancer.
Annu. Rev. Cell Dev. Biol. 2001, 17, 615−675.
(20) Engelman, J.; Luo, J.; Cantley, L. The Evolution of
Phosphatidylinositol 3-Kinases as Regulators of Growth and
Metabolism. Nat. Rev. Genet. 2006, 7, 606−619.
(21) Botyanszki, J.; Moore, C.; Keicher, J.; DeAnda, F.; Duan, M.;
Tallant, M.; Redman, A.; McFadyen, R.; Shotwell, J. B.; Banka, A.;
Leivers, M.; Xue, J.; McDonald, B.; Strum, S.; Creech, K.; Gobel, J.;
Roberts, C. Discovery of PI4Ka Selective Inhibitors as Anti-HCV
Agents. Presented at 64th Southeast Regional Meeting of the
American Chemical Society, Raleigh, NC, U.S., Nov 14−17, 2012;
SERM-258.
(22) Garratt, P.; Hobbs, C.; Wrigglesworth, R. One-Carbon
Compounds as Synthetic Intermediates. The Synthesis of Hydro-
pyrimidines and Hydroquinazolines by Sequential Nucleophilic
Addition to Diphenyl Cyanocarbonimidate with Concomitant
Cyclization. J. Org. Chem. 1989, 54 (5), 1062−1069.
(23) Hogan, P.; Cox, B. Aqueous Process Chemistry: The
Preparation of Aryl Sulfonyl Chlorides. Org. Process Res. Dev. 2009,
13, 875−879.
(24) He, Y.; Wang, B.; Dukor, R.; Nafie, L. Determination of
Absolute Configuration of Chiral Molecules Using Vibrational Optical
Activity: A Review. Appl. Spectrosc. 2011, 65 (7), 699−723.
P
dx.doi.org/10.1021/jm400781h | J. Med. Chem. XXXX, XXX, XXX−XXX