Phospshoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) dual inhibitors: Discovery and structure-activity relationships of a series of quinoline and quinoxaline derivatives

ARTICLE

pubs.acs.org/jmc

Phospshoinositide 3-Kinase (PI3K)/Mammalian Target of Rapamycin

(mTOR) Dual Inhibitors: Discovery and StructureꢀActivity

Relationships of a Series of Quinoline and Quinoxaline Derivatives^£

Nobuko Nishimura,*^,†Aaron Siegmund,^†Longbin Liu,^†Kevin Yang,^†Marian C. Bryan,^†Kristin L. Andrews,^||

Yunxin Bo,^†Shon K. Booker,^†Sean Caenepeel,^‡Daniel Freeman,^‡Hongyu Liao,^†John McCarter,^{^}

Erin L. Mullady,^{^}Tisha San Miguel,^{^}Raju Subramanian,^§Nuria Tamayo,^†Ling Wang,^‡

Douglas A. Whittington,^||Leeanne Zalameda,^{^}Nancy Zhang,^‡Paul E. Hughes,^‡and Mark H. Norman^†

^†Department of Chemistry Research and Discovery, ^‡Department of Oncology Research, ^§Pharmacokinetics & Drug Metabolism,

Department of Molecular Structure, and ^{^}Department of High-Throughput Screening/Molecular Pharmacology, Amgen Inc.,

One Amgen Center Drive, Thousand Oaks, California 91320-1799, United States

S Supporting Information

b

ABSTRACT: The phosphoinositide 3-kinase (PI3K) family

catalyzes the ATP-dependent phosphorylation of the 3⁰-hydro-

xyl group of phosphatidylinositols and plays an important role

in cell growth and survival. There is abundant evidence demon-

strating that PI3K signaling is dysregulated in many human

cancers, suggesting that therapeutics targeting the PI3K path-

way may have utility for the treatment of cancer. Our eﬀorts to

identify potent, eﬃcacious, and orally available PI3K/mamma-

lian target of rapamycin (mTOR) dual inhibitors resulted in the

discovery of a series of substituted quinolines and quinoxalines derivatives. In this report, we describe the structureꢀactivity

relationships, selectivity, and pharmacokinetic data of this series and illustrate the in vivo pharmacodynamic and eﬃcacy data for a

representative compound.

’ INTRODUCTION

converts ptdIns (3,4,5)P3 to ptdIns (4,5)P2, and gain of function

mutations in the PI3K catalytic subunit, p110R. These observa-

tions suggest that inhibitors targeting PI3K and mTOR may have

utility as cancer therapeutics.⁵

The phosphoinositide 3-kinase (PI3K) family catalyzes the

ATP-dependent phosphorylation of the 3⁰-hydroxyl group of

phosphatidylinositols. PI3Ks are subdivided into three classes,

depending on sequence homology and substrate preferences.

The class I PI3K family, comprising four isoforms (p110R,

p110β, p110δ, and p110γ), generate phosphatidylinositol

3,4,5-triphosphate, (ptdIns (3,4,5)P3), a potent secondary mes-

senger that triggers the activation of several downstream eﬀec-

tors, including the serine-threonine kinase, AKT (also known as

protein kinase B or PKB). In response to PI3K activation, AKT is

phosphorylated, which results in the triggering of a signal

transduction cascade that ultimately stimulates mammalian target

of rapamycin (mTOR) containing complex 1 (mTORC1).

mTORC1 plays a key role in regulating cell growth, survival and

proliferation by integrating diverse signaling inputs including

growth factors, nutrient availability, and cellular energy levels.^1,2

There is abundant evidence demonstrating that PI3K signal-

ing is dysregulated in human cancers, with many tumors harbor-

ing somatic genetic alterations, that result in constitutive

activation of the PI3K signaling network.^3,4The most common

genetic changes associated with constitutive PI3K signaling in

cancer are loss of function mutations in the phosphatase and

tensin homologue gene (PTEN), a lipid phosphatase that

In our eﬀorts to inhibit the PI3K pathway, we identiﬁed an

orally active dual PI3K/mTOR kinase inhibitor, N-(6-(6-chloro-

5-(4-ﬂuorophenylsulfonamido)pyridin-3-yl)benzo[d]thiazol-2-yl)-

acetamide (1; Figure 1), which showed potent activities in both

the biochemical (K_ivalues of 1, 2, 5, and 1 nM for PI3KR, β, δ, γ,

respectively, and an IC₅₀value of 2 nM for mTOR) and cellular

(IC₅₀= 6.3 nM) assays.^6,7In addition, compound 1 was active at

low doses in vivo as determined by the inhibition of AKT (Ser

473) phosphorylation in a mouse liver pharmacodynamic assay

(EC₅₀= 399 ng/mL) and by its ability to inhibit the growth of

U-87 MG tumor xenografts (ED₅₀= 0.26 mg/kg). However,

further investigations revealed that signiﬁcant amounts of the

deacetylated metabolite (2-aminobenzothiazole, 2; Figure 1) were

formed in both isolated hepatocytes and in vivo.⁸For example,

when 1 was dosed orally in rats for 4 days at 0.3 mg/kg (q.d.),

the ratio of N-acylamine 1 to amine 2 was ∼1.2 ([AUC]_0ꢀ24). The

development of compound 1 was not desirable as the circulating

Received: April 1, 2011

Published: May 25, 2011

r

2011 American Chemical Society

4735

dx.doi.org/10.1021/jm200386s J. Med. Chem. 2011, 54, 4735–4751

|

Journal of Medicinal Chemistry

ARTICLE

Figure 2. Interactions of the benzothiazole and quinoline linker binder

groups with Val 882 in PI3KR.

interaction to the NH of Val 882 would be maintained and the

C(2)H would be closer to the Val 882 carbonyl.¹²Furthermore,

the geometry of the C(2)Hꢀcarbonyl hydrogen bond appeared

to be more optimal with the use of a 6,6-heterocyclic linker

binder than with the benzothiazole group (Figure 2C vs B). More

importantly, we postulated that it would be possible to modulate

the physicochemical properties of such inhibitors by accessing

the ribose pocket at either the 3- or 4-position of a 6,6-hetero-

cyclic linker binder group thereby providing a means to improve

both the enzyme and cellular potencies (Figure 2C; R-group). In

this paper, we report the structure activity relationship (SAR)

studies of the 6,6-heterocyclic linker binder systems that led to

the identiﬁcation of two new series of potent and selective PI3K/

mTOR dual inhibitors with good pharmacokinetic and pharma-

codynamic proﬁles. Initially, we examined several 6,6-hetero-

cyclic systems as alternative linker binder groups and, subse-

quently, focused our SAR investigations on the 3-and 4-positions

of quinoline and quinoxoline series. Optimizing the compounds

in this way led to the discovery of quinoline 16d, whose in vivo

evaluation is described herein.

Figure 1. Key interactions of compounds 1ꢀ3 with PI3Kγ protein.

metabolite 2 is also a potent PI3K inhibitor (e.g., PI3KR K_i= 5.6

nM), and the potential exists that reacetylation of 2 could occur

in vivo. To circumvent this deacetylation problem, we investi-

gated two diﬀerent approaches. In the ﬁrst approach, we sought

to attenuate the in vivo stability of the N-acetyl group through the

replacement of the benzothiazole ring with various 6,5-bicyclic

heterocycles.⁹Those investigations led to the identiﬁcation of the

N-(imidazo[1,2-b]pyridazin-2-yl)acetamide series that proved to

be stable toward deacetylation both in vitro and in vivo. In this

investigation, we examined an alternative approach where we

replaced the N-acetyl 2-aminobenzothiazole moiety in 1 with

several 6,6-heterocyles that lack the N-acetyl moiety, therefore

completely eliminating the potential of deacetylation.¹⁰

To aid in the design of novel inhibitors, we ﬁrst analyzed the

crystal structure of compound 1 bound to the active site of the

PI3Kγ protein, which is highly homologous to the other class I

PI3K isoforms.¹¹The key interactions of compound 1 to the

enzyme are illustrated in Figure 1. The N-acetyl benzothiazole

group, also referred to as the linker binder moiety, forms two key

hydrogen bonds with the backbone NH and carbonyl of Val 882.

In this interaction, the thiazole nitrogen acts as the hydrogen

bond acceptor and the NH of the N-acetyl group serves as the

hydrogen bond donor. Other key interactions include a hydro-

gen bond to one of the sulfonamide oxygens from the Lys 833

side chain, and a hydrogen bond network between the nitrogen

on the central pyridine ring and a water molecule that is bridged

between the Asp 841 carboxylate and the Tyr 867 hydroxyl

groups within the aﬃnity pocket. Finally, the para-ﬂuorophenyl

group occupies the upper region of the ribose pocket.

Designing a linker binder without an N-acetamide group or a

free exocyclic amine was crucial to resolving the deacetylation

issue. To assess the binding contribution of the exocyclic amino

group in 2, compound 3 (Figure 2B) was prepared and its

inhibitory activity against PI3KRwas determined to be K_i=12.9nM.

The 10-fold decrease in activity was attributed to the diﬀe-

rence in both the strength and orientation of the hydrogen

bond interaction between the NH of 1 and CH of 3 with the

carbonyl of Val 882 (Figure 2A vs B). When tested in the U-87

MG cellular assay, compound 3 showed an IC₅₀of more than

700 nM. The loss in cellular activity of 3 vs 1 (∼200-fold) was

more than the loss in enzyme activity (10-fold), suggesting that

changes in the physicochemical properties of the compound led

to reduced cell penetration or higher protein binding. We

reasoned that by replacing the benzothiazole with a 6,6-hetero-

cycle, such as a quinoline (Figure 2C), the key hydrogen bonding

’ CHEMISTRY

Scheme 1 outlines the synthesis of benzothiazole 3 and

compounds with various unsubstituted 6,6-heterocycles as shown

in Table 1. The compounds were prepared through Suzuki

coupling reactions¹³with either the bromide 4a or the corres-

ponding boronic acid pinacol ester 4b, and either 5-bromoben-

zothiazole 5 or the appropriate 6,6- heterocycles. The requisite

linker binder intermediates were either commercially available

(5, 6a, 6b, 6e, 6g, and 6h) or prepared as described below (6c, 6d,

and 6f) in Scheme 2.

Scheme 2 shows the synthesis of the requisite heterocyclic

intermediates 6c, 6d, and 6f. The 2-chloro-1,5-naphthyridine

intermediate 6c was prepared from unsubstituted 1,5-naphthyr-

idine (8) via N-oxide formation followed by chlorination with

POCl₃(Scheme 2A). This reaction gave a 1:1 mixture of the

4- and 2-chloro regioisomers 9 and 6c, which were separated by

silica gel chromatography. The synthesis of 1,7-naphthyridine

derivative 6d started by bromination of 2-cyano-3-methylpyr-

idine (10) followed by a one carbon homologation using

potassium cyanide to yield dicyanopyridine 11 (Scheme 2B).

The key cyclization was performed in acetic acid in the presence

of HBr to yield the 8-bromo-1,7-naphthyridin-6-amine 12. The

bromine of 12 was removed via catalytic hydrogenation and the

amino group was converted directly to a triﬂate by diazotization

of the amine followed by displacement with triﬂic acid to aﬀord

6d. The pinacol ester of quinoxaline 6f was prepared from the

corresponding bromide (13) by treatment with bis(pinacolato)-

diborane (Scheme 2C).

4736

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Scheme 1. Synthesis of Compounds in Table 1^a

^aReagents and conditions: (a) PdCl₂dppf, bis(pinacolato)diborane, potassium acetate, 1,4-dioxane, 120 ꢀC; (b) PdCl₂dppf or Pd(PPh₃)₄, potassium or

sodium carbonate, 1,4-dioxane, water, 90ꢀ100 ꢀC.

Scheme 3 shows the synthesis of 4-substituted quinolines that

are illustrated in Table 2. In general, the targeted compounds

16aꢀf were prepared by Suzuki reactions of 4a or 4b with amino-

substituted quinolines 15bꢀ15e that were derived from 14 via

S_N2⁰reactions. Alternatively, the chloro intermediate 16a was

treated with a variety of secondary amines to give derivatives

16fꢀ16i or with aryl boronic acids or esters to provide the

corresponding aromatic derivatives 16jꢀ16m. In the case of

compound 16e, the requisite Suzuki coupling partner was

boronic acid 15e. This intermediate was synthesized by ﬁrst

converting 14 to the corresponding boronic ester (15a) and then

substituting the 4-chloro with 1-(pyridine-4-ylmethyl)piperidine.

The boronic ester was hydrolyzed to the boronic acid during the

substitution reaction to produce 15e.

Scheme 4 outlines the synthesis of 3-substituted quinoxalines

(18aꢀ18s). In the majority of cases, the compounds were

prepared from the intermediates 18a or 18b that were obtained

by the coupling of boronic ester 4b with either 7-bromo-2-

chloroquinoxaline (17a)¹⁴or the more reactive 7-bromo-2-

ﬂuoroquinoxaline (17b). The latter compound, 17b, was ob-

tained by treating 17a with tetra-n-butylammonium ﬂuoride

(TBAF) in DMSO.¹⁵The ﬁnal compounds were prepared either

via S_N2⁰reactions of ﬂuoroquinoxaline 18b with the appropriate

amines to give 18fꢀr or via a Suzuki reaction of 18b with pyridin-

4-yl boronic acid to provide 18e. Compounds 18c and 18d were

prepared by coupling 4b with 17c and 17d, which were derived

from 7-bromo-2-ﬂuoroquinoxaline (17b). Deprotection of the

Boc-protected piperidine derivative 18r with triﬂuoroacetic acid

(TFA) provided the corresponding piperidine 18s.

the following SAR discussion will focus on inhibition of PI3KR as

well as activities in the U-87 MG cellular assay that measures the

inhibition of pAKT (Ser 473). Selectivity proﬁles of the most potent

derivatives are reported in Table 4 and in the Supporting Information.

To test the hypothesis that 6,6-bicyclic heterocycles could

serve as replacements for the benzothiazole linker binder, quino-

line, isoquinoline, quinoxaline, quinazoline, cinnoline, and

naphthyridine analogues of compound 1 were prepared and

evaluated. The results are shown in Table 1. Quinoline 7a was

∼2-fold more potent in both enzyme and cellular assays than

benzothiazole 3 and achieved a similar enzyme potency as

aminobenzothiazole 2, thereby illustrating the ability of a 6,6-

heterocycle to serve as an excellent linker binder moiety. How-

ever, moving the nitrogen to the 2-position, as in isoquinoline 7b,

resulted in dramatic loss of potency (K_i= 375 nM), indicating

that the position of the hydrogen bond acceptor was extremely

important. For the remainder of the derivatives shown in Table 1,

we maintained the quinoline nitrogen of 7a and systematically

added one additional nitrogen at each of the other six positions of

the bicyclic ring. When the additional nitrogen was incorporated

at the 3, 4, or 5 positions of the quinoline ring system, similar

levels of enzyme activities were obtained (i.e., quinazoline 7g,

quinoxaline 7f, and 1,5-naphthyridine 7c). On the other hand,

adding the additional nitrogen to the 7 or 8 positions gave two

isomeric naphthyridines (7d and 7e) that exhibited >100-fold

drop in enzyme potency compared to 7a. These results demon-

strated that having an additional hydrogen bond acceptor at the

same side as N1-nitrogen was detrimental to the molecule’s

ability to bind to the enzyme. For example, the N-8 nitrogen of 7e

would have an unfavorable interaction with the backbone

carbonyl of Glu 880. In addition, the aromatic CH at the 8

position of 7a also can form a favorable interaction to the

carbonyl of Glu 880, which is lacking in 7e. In the case of 7d,

’ RESULTS AND DISCUSSION

In general, the compounds described in this paper were

inhibitors of PI3KR, β, δ, γ, and mTOR; however, for simplicity,

4737

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Table 1. Enzyme and Cellular Assay Results of Various 6,6-

Bicyclic Heterocyclic Derivatives^a

Scheme 2. Synthesis of Intermediates 6c, 6d, and 6f ^a

^aReagents and conditions: (a) 3-chloroperoxybenzoic acid, DCM, room

temperature; (b) POCl₃, 100 ꢀC; (c) NBS, CCl₄, reﬂux; (d) KCN,

MeOH, room temperature; (e) HBr, AcOH, 0 ꢀC; (f) Pd/C, KOH, H₂,

room temperature; (g) NaNO₂, triﬂuoromethanesulfonic acid, DMF,

room temperature; (h) PdCl₂dppf, bis(pinacolato)diborane, potassium

acetate, 1,4-dioxane, 120 ꢀC.

optimization based on the following considerations: (1) We

found that quinazoline 7g was unstable in weakly acidic condi-

tions such as 0.1% TFA in water. It is likely that the quinazoline

ring undergoes hydrate formation, a phenomenon well-known

for this type of system.^16,17(2) Modiﬁcations at either the 3- or

4-position of the 6,6-bicyclic systems were more readily acces-

sible from the quinoline and the quinoxaline derivative (7a and

7f) relative to 1,5-naphthyridine analogue (7c). (3) Further-

more, the additional nitrogen at the 5-position of 7c seemed to

oﬀer no signiﬁcant advantage over quinoline 7a.

To improve the cellular potency of compound 7a, a variety of

4-substituted quinolines were synthesized and evaluated. The

results of the PI3KR enzyme and U-87 MG cellular assays are

shown in Table 2. The methoxyethylamine derivative 16c showed a

35-fold decrease in the enzyme potency relative to 7a. We

postulated this loss of potency was due to the presence of a

tautomerizable NH proton on the substituent that would diminish

the ability of the quinoline nitrogen to serve as a hydrogen bond

acceptor. On the other hand, tertiary amines at the 4-position,

such as the dimethyl amino analogue 16b, were well tolerated.

Even bulkier groups were tolerated in the ribose pocket as shown

by the single-digit nanomolar enzyme activity from piperidine

16f. More importantly, introducing polar functionalities on the

piperidine ring proved to be fruitful in improving the cellular

potency. For example, the 4-hydroxylpiperidine analogue 16g

was 3-fold more potent in the cellular assay than 16f and the

morpholine derivative 16d was 30-fold more potent, with an IC₅₀

of 16 nM. Compound 16d represented the ﬁrst example of a

modiﬁed linker binder derivative showing an IC₅₀< 20 nM in the

cellular assay. The piperazine analogues 16h and 16e also showed

cellular potencies comparable to that of 16d, indicating that polar

groups were preferred in this region. As lipophilicity increased, as

was the case for the N-benzyl piperazine analogue 16i, so did the

enzyme-to-cell shift. This trend also held true for aromatic

substituents; as the lipophilicity of the substituents decreased

in going from phenyl (16j) to 4-dimethylaminophenyl (16m) to

3-pyridyl (16k) and 4-pyridyl (16l), incremental improvements

in cellular potencies (127 to 4 nM) were observed.

^aThe inhibition of PI3KR activity was determined using a modiﬁed in

vitro AlphaScreen assay with human p110R enzyme. Cellular IC₅₀values

were determined by AKT(Ser 473) phosphorylation assay using U-87

MG cells. ^bThe results are reported as average ( SD. The results

without SD are from a single measurement.

the nitrogen at the 7 position is adjacent to the face of Tyr 867,

which is also a less favorable interaction than the CH of 7a.

Similarly cinnoline 7h, which lacks the CH as hydrogen bond

donor to Val 882, also showed a 15-fold drop in potency. It is

likely that the nitrogen at the 2-position of the cinnoline core

causes an unfavorable interaction with the carbonyl of Val 882

and therefore leads to diminished aﬃnity.

Having identiﬁed four 6,6-heterocyclic systems that showed K_i

values of <10 nM (compounds 7a, 7c, 7f, and 7g), we focused on

the quinoline and quinoxaline scaﬀolds (7a and 7f) for further

With the success in the 4-substituted quinoline series, we

turned our attention to 3-substituted quinoxalines to probe the

4738

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Scheme 3. Synthesis of Compounds with 4-Substituted Quinolines^a

^aReagents and conditions: (a) amine, DMF or DMSO, 90ꢀ120 ꢀC; (b) PdCl₂dppf, bis(pinacolato)diborane, potassium acetate, 1,4-dioxane, 120 ꢀC;

(c) PdCl₂dppf, potassium or sodium carbonate, 1,4-dioxane, water, 90ꢀ100 ꢀC; (d) For 16j: phenylboronic acid, FibreCat, sodium carbonate, 1,4-

dioxane, water, 120 ꢀC; for 16kꢀ16m: aryl boronic acid or aryl boronic ester, dichlorobis(t-butylphenylphosphine)palladium, potassium acetate, n-

butanol, water, 100 ꢀC.

structure activity relationships around a diﬀerent area of the

ribose pocket. The results are shown in Table 3. When the 3

position of quinoxaline 7f was substituted by a methoxyethyl-

amine (18f), the enzymatic activity was increased in contrast to

the observed decrease in the corresponding quinoline analogue

16c. This is likely due to the fact that the 3-methoxyethylamine

NH proton in 18f cannot tautomerize with the N-1 nitrogen that

is involved in the crucial hydrogen bonding interaction with Val

882, unlike the corresponding quinoline analogue 16c. As a

result, there was little change in potency in going from a

secondary amine to a tertiary amine (18f vs 18g) in this series.

In addition, both 18f and 18g showed more than a 25-fold

improvement in the cellular potencies compared to the parent

quinoxaline 7f. Similar to the quinoline series, bulkier groups

were tolerated in the quinoxaline series as well, as seen by

analogues 18h, 18i, and 18j. However, these analogues also

showed relatively large enzyme-to-cell shifts, which may be

attributed to the compounds’ overall increased lipophilicity.

Introducing polar functionalities in this area improved cellular

potency, as shown by analogues 18k and 18l. Surprisingly, the

3-piperidine derivative 18s showed a signiﬁcant loss in cellular

activity compared to 18h.

18q, and 18d. The 4-pyridyl derivative 18e also showed good

activity in the cellular assay. In general, the cellular potency in the

3-substituted quinoxaline series was less sensitive to substitution

than in the case of the 4-substituted quinoline series. This

observation may be rationalized in that substituents in the

3-position of the quinoxalines project away from the binding

pocket and more toward the solvent front than do groups on the

4-position of the quinolines.

To examine this in more detail, we obtained an X-ray crystal

structure of quinoline 16d bound to PI3Kγ protein (Figure 3A).

The binding mode of 16d was similar to the benzothiazole

compound 1. As was observed in the crystal structure of

benzothiazole compound 1 (illustrated in Figure 1), the quino-

line nitrogen forms a hydrogen bond with the Val 882 backbone

NH and a water molecule is bridged between the Asp 841

carboxylate, the Tyr 867 hydroxyl group, and the nitrogen on

the center pyridine ring. The morpholine ring in 16d occupies

the ribose pocket (the space above the quinoline ring) that was

unoccupied in the crystal structure of compound 1. The quinoxa-

line compound 18d was modeled into the X-ray structure of the

enzyme (Figure 3B). The space-ﬁlling model on the right shows

that the morpholine ring extends toward the solvent front rather

than into the ribose pocket. This is consistent with the observa-

tion that the cellular potency in the 3 substituted quinoxalines

series was less sensitive to substitution than the corresponding

quinoline analogues.

While dimethylamine at the 3-position in the quinoxaline core

was well tolerated (18c; U-87 MG cellular IC₅₀= 44 nM), larger

cell shifts with highly lipophilic substituents tended to be associated

with compounds bearing cyclic amines, such as 18m, 18n, and

18o. When heteroatoms were introduced to these groups, the

cellular potencies were improved, as shown with analogues 18p,

The SAR investigations described above enabled the identiﬁ-

cation of several potent PI3K/mTOR dual inhibitors with IC₅₀

4739

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Table 2. Enzyme and Cellular Assay Results of 4-Substituted

Quinoline Derivatives.^a

values of e20 nM in the U-87 MG cellular assay (16d, 16e, 16k,

16l, 18d, 18k, and 18q).¹⁸The in vitro proﬁles of these

derivatives are reported in Table 4. All compounds were potent

against all of the class I PI3K isoforms, mTOR, and hVPS34, a

class III PI3K. The compounds were also potent inhibitors of the

protein kinase DNA-PK (DNA-dependent serine/threonine

protein kinase), which contains a catalytic domain homologous

to the PI3Ks, but were selective against other protein kinases.¹⁹

The pharmacokinetic properties of these compounds were

also examined and the results are shown in Table 5. All but one

compound, 16e, showed good stability in rat and human liver

microsomes, and therefore, they were evaluated in rat in vivo

pharmacokinetic studies. In general, all compounds tested

showed low in vivo clearance (0.02ꢀ0.17 L/h/kg; 0.6ꢀ1% of

rat liver blood ﬂow rate), low to moderate volume of distribution

(0.2ꢀ1.2 L/kg), and a moderate to long mean residence time

(MRT; 2ꢀ15 h). When dosed orally, all except compound 18k,

demonstrated excellent bioavailability (%F = 70ꢀ92) and high

exposures (AUC).

Compounds with acceptable pharmokinetic properties (16d,

16k, 16l, 18d, and 18q) were tested in a mouse liver pharmaco-

dynamic (PD) model in which the inhibition of HGF induced

AKT (Ser 473) phosphorylation was measured.²⁰All com-

pounds were dosed orally at 0.3, 1, and 3 mg/kg, and after 3 or

4 h, the mice were injected with 12 μg of hepatocyte growth

factor (HGF) to activate PI3K-dependent AKT phosphorylation

in the liver. The levels of AKT (Ser 473) phosphorylation were

determined ﬁve minutes post HGF administration by a quanti-

tative electrochemiluminescence immunoassay. The results are

shown in Table 6. All ﬁve compounds showed dose-dependent

inhibition of AKT (Ser 473) phosphorylation, indicating that

they eﬀectively inhibited PI3K in vivo. The doseꢀresponse data

for a representative example, quinoline 16d, is shown graphically

in Figure 4A. In addition to testing the compounds at the 3 or 4 h

time points, we were interested in evaluating the duration of the

compound’s inhibitory eﬀect in vivo. The time-course PD for

quinoline 16d is illustrated in Figure 4B. In mice dosed orally at 3

mg/kg, compound 16d inhibited the phosphorylation of AKT

(Ser 473) at 96% after 1 h and 72% after 6 h, and the eﬀect was

signiﬁcantly diminished after 24 h (29% inhibition).

The promising pharmacodynamic results of compound 16d in

the mouse liver model prompted us to evaluate it in a PTEN-null

U-87 MG glioblastoma xenograft model.²¹Compound 16d was

dosed orally at 10 mg/kg daily, 30 mg/kg daily, or at 10 mg/kg

twice daily for 12 days (Figure 5).²²At 10 mg/kg daily dosing, the

tumor growth was suppressed 62% compared to the vehicle

group. Furthermore, tumor regression was observed when dosed

either at 10 mg/kg bid or 30 mg/kg qd, showing that compound

16d was eﬃcacious in this model.

’ CONCLUSION

We have identiﬁed two series of potent PI3K/mTOR dual

inhibitors with excellent pharmacokinetic properties and in vivo

eﬃcacies. These compounds were designed to avoid the issue of

deacetylation encountered with the early lead compound 1. By

adopting 6,6-heterocycles as linker binders, we were able to

eliminate the acetyl amino hydrogen bonding donor without

sacriﬁcing the enzyme potency. Furthermore, we demonstrated

that by incorporating suitable substituents either at the 4-

position of the quinoline or the 3-position of the quinoxaline rings,

excellent cellular potencies could also be achieved. In addition,

^aThe inhibition of PI3KR activity was determined using a modiﬁed in

vitro AlphaScreen assay with human p110R enzyme. Cellular IC₅₀values

were determined by AKT(Ser 473) phosphorylation assay using U-87

MG cells. ^bThe results are reported as average ( SD. The results

without SD are from a single measurement.

4740

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Scheme 4. Synthesis of 3-Substituted Quinoxalines^a

^aReagents and conditions: (a) TBAF, DMSO, room temperature; (b) amine, DMSO, or DMF, 60ꢀ100 ꢀC; (c) PdCl₂dppf or dichlorobis(di-t-

butylphenylphosphine)palladium, potassium carbonate, 1,4-dioxane, water, 90ꢀ100 ꢀC; (d) pyridin-4-yl boronic acid, dichlorobis(di-t-butylphenylpho-

sphine)palladium, potassium acetate, 1,4-dioxane, water, 90 ꢀC; (e) TFA, DCM, room temperature.

these results supported the hypothesis that the ribose pocket of

theenzymecanbeeﬀectively utilized in optimizing both the potency

and the physicochemical properties of PI3K inhibitors. Further-

more, we demonstrated that a representative compound, 16d,

successfully blocked the targeted PI3K pathway in a mouse PD

model and inhibited tumor growth in a U-87 MG xenograft model.

reported in parts per million (ppm, δ units). All final compounds were

purified to >95% purity as determined by LC/MS obtained on an

Agilent 1100 spectrometer using the following methods: [A] Agilent SB-

C18 column (50 ꢁ 3.0 mm, 2.5 μ) at 40 ꢀC with a 1.5 mL/min flow rate

using a gradient of 5ꢀ95% [0.1% TFA in acetonitrile] in [0.1% TFA in

water] over 3.5 min; [B] Phenomenex Gemni NX C18 column (50 ꢁ

3.0 mm, 3 μ) at 40 ꢀC with a 1.5 mL/min flow rate using a gradient of

5ꢀ95% [0.1% formic acid in acetonitrile] in [0.1% formic acid in water]

over 3.5 min. Low-resolution mass spectral (MS) data were obtained at

the same time of the purity determination on the LC/MS instrument

using ES ionization mode (positive). High-resolution exact mass

measurements were performed using electrospray ionization (positive

mode) on a quadrupole time-of-flight (QTOF) mass spectrometer

(Xevo Q-TOF, Waters Inc.). The instrument was calibrated immedi-

ately before performing the exact mass measurements using sodium

formate cluster ions; the resulting mass accuracy for all calibrant ions was

better than 1.5 ppm. DMSO stock solutions (10 mM) were diluted 2000

times with 50/50 acetonitrile/water. Each sample (2 μL) was introduced

to the mass spectrometer by flow injection using 1:1 water/acetonitrile as

the carrier solvent with a flow rate of 100 μL/min. MS data were acquired

from 100 to 1000 m/z with an acquisition rate of 1.0 scan/s. Reserpine

(∼40 fmol/μL) was used as a lock mass reference and was ionized using

a second orthogonal sprayer.

’ EXPERIMENTAL SECTION

Chemistry. General. Unless otherwise noted, all materials were

obtained from commercial suppliers and used without further purifica-

tion. Anhydrous solvents were obtained from Aldrich, Acros, or EM

Science and used directly. All reactions involving air- or moisture-

sensitive reagents were performed under a nitrogen or argon atmo-

sphere. Microwave-assisted reactions were conducted with either an

initiator from Biotage, Uppsala, Sweden, or Explorer from CEM,

Matthews, North Calorina. Silica gel chromatography was performed

using either glass columns packed with silica gel (200ꢀ400 mesh,

Aldrich Chemical) or prepacked silica gel cartridges (Biotage or Redisep).

Preparative HPLC was performed with Varian Prostar using method:

[A] Phenomenex Synergi, MAX-RP column (150 ꢁ 50 mm, 10 μ)

with a 80 mL/min flow rate using a gradient of 5ꢀ100% [0.1% TFA

in acetonitrile] in [0.1% TFA in water] over 20 min; [B] Phenomenex

Gemni, C18 column (100 ꢁ 30 mm, 5 μ) with a 40 mL/min flow

rate using a gradient of 5ꢀ100% [0.1% TFA in acetonitrile] in [0.1%

TFA in water] over 20 min. NMR spectra were determined with a

Bruker 300 MHz or DRX 400 MHz spectrometer. Chemical shifts were

General Boronic Ester Formation Method. To an appropriately sized

reaction vessel was added a halide (1 equiv), bis(pinacolato)diborane

(1.1 equiv), potassium acetate (2.0 equiv), and PdCl₂dppf (0.05 equiv)

in 1,4-dioxane. The mixture was deoxygenated by bubbling nitrogen

through it for 5 min. The reaction mixture was heated at 120 ꢀC by

4741

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Table 3. Enzyme and Cellular Assay Results of 3-Substituted

Quinoxaline Derivatives^a

conventional heating or irradiated with microwave at 120 ꢀC, until the

starting halide was consumed. The reaction mixture was allowed to cool

to room temperature then partitioned between water and EtOAc. The

aqueous phase was extracted with EtOAc. The combined organic phases

were washed with saturated aqueous NaCl. The organic phase was dried

over sodium sulfate, filtered, and concentrated in vacuo. If necessary, the

crude product was purified by silica gel column chromatography.

General Suzuki Coupling Method. To an appropriately sized reaction

vessel was added a halide (1 equiv), a boronic ester or acid (1.1 equiv),

potassium or sodium carbonate (2 M aqueous solution, 3 equiv), palladium

catalyst (PdCl₂dppf, Pd(PPh₃)₄, or FibreCat, 0.05 to 0.1 equiv) in 1,

4-dioxane. The mixture was deoxygenated by bubbling nitrogen through

it for 5 min. The mixture was heated at 90 ꢀC by conventional heating or

irradiated with microwave at 100 ꢀC, until the starting halide was

consumed. The reaction mixture was allowed to cool to room tempera-

ture and was partitioned between water and EtOAc. The aqueous phase

was extracted with EtOAc. The combined organic phases were washed

with saturated aqueous NaCl. The organic phase was dried over

sodium sulfate, filtered, and concentrated in vacuo. If necessary, the

crude product was purified by silica gel column chromatography.

General S_N2⁰Method on Aryl Halides. To a solution of aryl halide

(1 equiv) in DMF or DMSO was added an amine (2ꢀ10 equiv). The

reaction mixture was heated to 60ꢀ100 ꢀC until the starting aryl halide

was consumed. The reaction mixture was allowed to cool to room

temperature then directly subjected to chromatographic purification or

partitioned between water and organic solvent, extracted with more

organic solvent. The combined organic phases were dried, filtered, and

concentrated, then purified by silica gel column chromatography.

N-(5-(Benzo[d]thiazol-6-yl)-2-chloropyridin-3-yl)-4-fluorobenze-

nesulfonamide (3). This compound was prepared from 4b (0.212 g,

0.514 mmol) and 6-bromobenzo[d]thiazole (0.100 g, 0.467 mmol)

according to the general Suzuki coupling method to obtain the title

compound (0.150 g, 77%) as a tan solid. MS (ESI pos. ion) m/z: calc’d

for C₁₈H₁₁ClFN₃O₂S₂: 419.0; found 419.9 (Mþ1). HRMS calc’d for

1

C₁₈H₁₁ClFN₃O₂S₂(M þ H): 420.0037; found 420.0045. H NMR

(400 MHz, DMSO-d₆) δ 7.43 (t, J = 8.71 Hz, 2 H), 7.77ꢀ7.91 (m, 3 H),

8.08 (d, J = 2.15 Hz, 1 H), 8.22 (d, J = 8.41 Hz, 1 H), 8.55 (d, J = 1.37 Hz,

1 H), 8.63 (d, J = 1.96 Hz, 1 H), 9.48 (s, 1 H), 10.54 (br s, 1 H).

N-(5-Bromo-2-chloropyridin-3-yl)-4-fluorobenzenesulfonamide

(4a). (a). N-(5-bromo-2-chloro-3-pyridinyl)-4-fluoro-N-((4-fluoro-

phenyl)sulfonyl)benzenesulfonamide. To a solution of 5-bromo-2-

chloropyridin-3-amine (20.0 g, 96.4 mmol) in pyridine (150 mL) was

added 4-fluorobenzene-1-sulfonyl chloride (41.3 g, 212 mmol). The

resulting mixture was heated to 100 ꢀC under N₂. The reaction was

allowed to cool to room temperature. Water (100 mL) was added and the

mixture was stirred at room temperature for 1 h. The resulting precipitate

was collected by filtration, washed with water, and air-dried to afford the title

compound (43.0 g, 85%) as an off-white solid. MS (ESI pos. ion) m/z:

calc’d for C₁₇H₁₀BrClF₂N₂O₄S₂: 521.9; found 522.7, 524.7 (Mþ1, Mþ3).

¹H NMR (300 MHz, CDCl₃) δ 7.23ꢀ7.35 (m, 4 H), 7.61 (d, J = 2.19 Hz,

1 H), 7.95ꢀ8.06 (m, 4 H), 8.55 (d, J = 2.34 Hz, 1 H).

(b). N-(5-Bromo-2-chloropyridin-3-yl)-4-fluorobenzenesulfona-

mide (4a). To a suspension of N-(5-bromo-2-chloro-3-pyridinyl)-4-

fluoro-N-((4-fluorophenyl)sulfonyl)benzenesulfonamide (36.6 g, 69.9

mmol) in MeOH (400 mL) was added potassium carbonate (21.2 g, 154

mmol) and water (5 mL). The resulting mixture was stirred at room

temperature for 20 h. The solvent was removed in vacuo. The residue

was suspended in water and the pH was adjusted to 6 using 5 N HCl. The

aqueous phase was extracted with EtOAc (4 ꢁ 200 mL). The combined

organic phases were dried over magnesium sulfate, filtered, and con-

centrated. Crystallization from 30% acetone in hexanes afforded the title

compound (21.5 g) as an off-white solid. The mother liquor was con-

centrated and purified by silica gel chromatography (20% acetone in

hexanes) to afford another batch of the title compound (2.7 g) as a tan

^aThe inhibition of PI3KR activity was determined using a modiﬁed in

vitro AlphaScreen assay with human p110R enzyme. Cellular IC₅₀values

were determined by AKT(Ser 473) phosphorylation assay using U-87

MG cells. ^bThe results are reported as average ( SD. The results

without SD are from a single measurement.

4742

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Figure 3. (A) X-ray crystal structure of 16d (green) in PI3Kγ. (B) Modeled structure of 18d (orange) in PI3Kγ.

Table 4. In Vitro Proﬁles of Selected Compounds^a

PI3K IC₅₀(nM)

compd number

R

β

γ

δ

mTOR IC₅₀(nM)

hVPS34 IC₅₀(nM)

DNA-PK IC₅₀(nM)

16d

16e

16k

16l

4.6 ( 3

4.6 ( 2

5.4 ( 3

6.5 ( 4

3.1 ( 0.2

12 ( 0.8

5.4 ( 1

13 ( 10

15 ( 0.7

10 ( 7

8.1 ( 3

6.5 ( 1

7.7 ( 3

13 ( 1

2.4 ( 0.3

16 ( 1.

13 ( 2

4.3 ( 2

3.9 ( 1

3.9 ( 2

11 ( 1

ꢀ

2.3 ( 0.5

6.0 ( 0.1

5.9 ( 3

ꢀ

1.6 ( 2

2.9 ( 0.08

5.3 ( 0.05

5.0 ( 2

ꢀ

15 ( 0.3

15 ( 12

9.8 ( 6

9.6 ( 4

6.0 ( 0.4

4.4 ( 0.4

6.0 ( 0.5

5.6 ( 0.5

0.41 ( 0.2

5.0 ( 1

18d

18l

3.4 ( 1

ꢀ

5.3 ( 0.5

37 ( 19

18q

ꢀ

^aThe results are reported as average ( SD. The results without SD are from single measurement. Details of the assays are reported in the Supporting

Information.

solid. The total combined yield was 95%. MS (ESI pos. ion) m/z: calc’d

2-Chloro-1,5-naphthyridine (6c). To a 100 mL round-bottomed

flask was added 1,5-naphthyridine (8, 0.260 g, 2.00 mmol), DCM

(10 mL), 3-chloroperoxybenzoic acid (0.517 g, 3.00 mmol). The

reaction mixture was stirred at room temperature overnight. The solvent

was removed in vacuo, and the residue was purified by silica gel

chromatography (20% MeOH/EtOAc) to give 1,5-naphthyridine

N-(1)-oxide (0.223 g, 76% yield). This material (0.198 g, 1.36 mmol)

was dissolved in POCl₃(2 mL) and the mixture was stirred at 100 ꢀC for

8 h. The solvent was removed in vacuo. Saturated aqueous NaHCO₃

(2 mL) was added to the residue and the aqueous phase was extracted

with EtOAc (2 ꢁ 20 mL). The combined organic phases were washed

with saturated aqueous NaCl (2 mL), dried over sodium sulfate, filtered,

and concentrated in vacuo. The crude product was purified by silica gel

1

for C₁₁H₇BrClFN₂O₄S: 364.9; found 365.9, 367.9 (Mþ1, Mþ3). H

NMR (300 MHz, MeOD) δ 7.09ꢀ7.20 (m, 2 H), 7.63 (d, J = 2.19 Hz,

1 H), 7.71 (d, J = 2.34 Hz, 1 H), 7.84ꢀ7.94 (m, 2 H).

N-(2-Chloro-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-

pyridin-3-yl)-4-fluorobenzenesulfonamide (4b). This compound was

prepared from 4a (2.01 g, 5.51 mmol) using the general boronic ester

formation method to afford the title compound (1.53 g, 67%) as a

viscous oil, which solidified upon standing. MS (ESI pos. ion) m/z:

calc’d for boronic acid C₁₁H₉BClFN₂O₄S: 330.0; found 331.0 (boronic

acid Mþ1). ¹H NMR (300 MHz, CDCl₃) δ 1.33ꢀ1.41 (m, 12 H), 6.88

(s, 1 H), 7.10ꢀ7.19 (m, 2 H), 7.75ꢀ7.83 (m, 2 H), 8.31 (d, J = 1.6 Hz,

1 H), 8.46 (d, J = 1.6 Hz, 1 H).

4743

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Table 5. Pharmacokinetic Properties of Selected Compounds

in vitro microsomal stability^a

in vivo rat PK^b

MRT (h)

rat

human

IV^c

PO^d

Compd Number

CL μL/(min*mg)

CL (L/kg/h)

Vd,ss (L/kg)

%F

AUC (ng*h/mL)

16d

16e

16k

16l

108

>399

33

47

247

<14

25

0.17

ꢀ

0.13

0.04

0.02

0.59

0.08

0.44

ꢀ

0.81

0.52

0.19

1.21

0.40

3.0

ꢀ

92^e

ꢀ

9554

ꢀ

5630

14700

29560

271

6.3

15

73^f

70^e

88^e

16^f

72^e

20

18d

18k

18q

16

9.0

2.1

5.4

<14

32

31

10750

^aSingle experimental value; estimated from parent compound (1 μM) remaining following a 30 min incubation in liver microsome (0.25 mg/mL) and

NADPH (1 mM). ^bPharmacokinetic parameters following administration in male SpragueꢀDawley rat: 3 animals per study. ^cDosed at 1 mg/kg as a

solution in DMSO. ^dPO doses was 2 mg/kg for 16d and were 1 mg/kg for all other compounds. ^eDosed as a suspension in 1% Tween 80, 2% HPMC, pH

2.2 with methanesulfonic acid. ^fDosed as a suspension in 1% Tween 80, 2% HPMC, pH 2.0 with HCl.

Table 6. Inhibition of AKT (Ser 473) Phosphorylation in a Mouse Liver Pharmacodynamic Model^a

0.3 mg/kg

conc^b(ng/mL)

1 mg/kg

conc^b(ng/mL)

3 mg/kg

conc^b(ng/mL)

compd number

time (h)

% inh

16d

16k

16l

3

4

65

70

89

55

50

145

210

326

448

155

75

97

98

86

492

899

97

99

1310

2330

2470

4470

1790

950

>99

97

18d

18q

1730

417

95

^a% Inhibition was calculated using the level of AKT (Ser 473) phosphorylation of veh/HGF as 100%. ^bTotal drug concentration in plasma.

chromatography (60% EtOAc/hexanes) to give the title compound

(0.078 g, 35%). MS (ESI pos. ion) m/z: calc’d for C₈H₅ClN₂: 164.0;

found 165.0 (Mþ1). ¹H NMR (300 MHz, CDCl₃) δ 7.64 (d, J = 8.77

Hz, 1 H), 7.66ꢀ7.72 (m, 1 H), 8.35 (t, J = 8.77 Hz, 2 H), 8.99 (dd, J =

4.17, 1.68 Hz, 1 H). Also isolated was 4-chloro-1,5-naphthyridine

(9, 0.086 g, 39%). MS (ESI pos. ion) m/z: calc’d for C₈H₅ClN₂:

(b). 8-Bromo-1,7-naphthyridin-6-amine (12). To a 50 mL round-

bottomed flask was added hydrobromic acid, 30% in acetic acid

(0.32 mL, 5.9 mmol), and compound 11 (0.280 g, 1.96 mmol) in

AcOH (0.5 mL) at 0 ꢀC. The reaction mixture was stirred at 0 ꢀC for 30

min. The resulting solid was collected and washed with 50% EtOAc/

hexanes, then saturated aqueous NaHCO₃(5 mL) was added and the

mixture was extracted with EtOAc (2 ꢁ 50 mL). The organic extract was

washed with saturated aqueous NaCl (5 mL), dried over sodium sulfate,

filtered, and concentrated in vacuo and the residue was purified by silica

gel chromatography (40% EtOAc/hexanes) to give the title compound

(0.312 g, 71%). MS (ESI pos. ion) m/z: calc’d for C₈H₆BrN₃: 223.0;

found 224.0, 226.0 (Mþ1, Mþ3). ¹H NMR (300 MHz, CDCl₃) δ 4.63

(s, 2 H), 6.61 (s, 1 H), 7.42 (dd, J = 8.48, 4.09 Hz, 1 H), 7.85 (dd, J = 8.48,

1.61 Hz, 1 H), 8.78 (dd, J = 3.95, 1.61 Hz, 1 H).

(c). 1,7-Naphthyridin-6-yl trifluoromethanesulfonate (6d). To a

50 mL round-bottomed flask was added compound 12 (0.224 g, 1.00

mmol), potassium hydroxide (0.067 g, 1.2 mmol), 10% palladium on

carbon (0.011 g, 0.1 mmol), and EtOH (2 mL). The mixture was stirred

under a hydrogen balloon for 4 h. The mixture was filtered through a pad

of Celite and washed with EtOAc. The filtrate was concentrated in vacuo

and the crude product was purified by silica gel chromatography (100%

EtOAc) to give 1,7-naphthyridin-6-amine (0.108 g, 74% yield). To a

solution of this, 1,7-naphthyridin-6-amine (0.102 g, 0.70 mmol) in DMF

(1.6 mL), sodium nitrite (0.097 g, 1.41 mmol), and trifluoromethane-

sulfonic acid (0.8 mL, 9.0 mmol) were added. The reaction mixture was

stirred at room temperature for 2 h and then diluted with EtOAc

(40 mL) and washed with water (5 mL), saturated NaHCO₃(5 mL),

and saturated NaCl (5 mL). The organic phase was dried over sodium

sulfate, filtered, and concentrated in vacuo. The crude product was

1

164.0; found 165.0. H NMR (300 MHz, CDCl₃) δ 7.75 (dd, J =

8.55, 4.17 Hz, 1 H), 7.79 (d, J = 4.68 Hz, 1 H), 8.47 (dd, J = 8.62, 1.61 Hz,

1 H), 8.87 (d, J = 4.68 Hz, 1 H), 9.11 (dd, J = 4.09, 1.61 Hz, 1 H).

1,7-Naphthyridin-6-yl trifluoromethanesulfonate (6d). (a). 3-(Cya-

nomethyl)picolinonitrile (11). To a 100 mL round-bottomed flask was

added 3-methylpicolinonitrile (10, 2.36 g, 20.0 mmol), NBS (7.82 g,

43.9 mmol), and CCl₄(50 mL). The reaction mixture was stirred at

reflux for 16 h and allowed to cool to room temperature. The resulting

solid was removed and the filter cake was washed with 50% EtOAc/

hexanes. The combined filtrate was concentrated in vacuo and the crude

product was purified by silica gel chromatography (30% EtOAc/hexane)

to give 3-(bromomethyl)picolinonitrile (2.56 g, 65% yield). To a

solution of this 3-(bromomethyl)picolinonitrile (1.56 g, 7.92 mmol)

in MeOH (50 mL), potassium cyanide (0.773 g, 1.19 mmol) was added.

The reaction mixture was stirred at room temperature for 6 h. The

solvent was removed in vacuo and the residue was dissolved in EtOAc

(50 mL), washed with water (10 mL), saturated aqueous NaCl (10 mL),

dried over sodium sulfate, filtered, and concentrated in vacuo. The crude

product was purified by silica gel chromatography (40% EtOAc/

hexanes) to give 11 (0.328 g, 29% yield). MS (ESI pos. ion) m/z: calc’d

for C₈H₅N₃: 143.0; found 144.0 (Mþ1). ¹H NMR (300 MHz, CDCl₃)

δ 4.06 (s, 2 H), 7.63 (dd, J = 8.11, 4.75 Hz, 1 H), 8.01ꢀ8.08 (m, 1 H),

8.73 (dd, J = 4.75, 1.39 Hz, 1 H).

4744

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Figure 4. Eﬀect of compound 16d in a mouse liver pharmacodynamic model measuring the inhibition of HGF-stimulated AKT (Ser 473)

phosphorylation:a (A) Doseꢀresponse study at 3 h post dose. (B) Time-course study at 3 mg/kg.

^aAsterisks denote p < 0.05 compared with the vehicle/HGF group. Statistical signiﬁcance was evaluated by Dunnett’s method. Bars represent the

average ( SD (n = 3). Diamonds represent mean plasma concentrations.

purified by silica gel chromatography (50% EtOAc/hexanes) to give the

title compound (0.138 g, 71% yield) as a yellow solid. MS (ESI pos. ion)

(400 MHz, CDCl₃) δ 1.40 (s, 12 H), 8.00ꢀ8.18 (m, 2 H), 8.60 (s, 1 H),

8.86 (d, J = 5.02 Hz, 2 H).

1

m/z: calc’d for C₉H₅F₃N₂O₃S: 278.0; found 279.0 (Mþ1). H NMR

N-(2-Chloro-5-(quinolin-6-yl)pyridin-3-yl)-4-fluorobenzenesulfo-

namide (7a). This compound was prepared from 4a (0.250 g, 0.684

mmol) and quinoline-6-boronic acid (0.177 g, 1.03 mmol) according to

the general Suzuki coupling method to afford the title compound (0.140 g,

50%) as a white solid. MS (ESI pos. ion) m/z: calc’d for C₂₀H₁₃ClFN₃-

O₂S: 413.0; found 414.0 (Mþ1). HRMS calc’d for C₂₀H₁₃ClFN₃O₂S

(M þ H) 414.0472: found; 414.0492. ¹H NMR (400 MHz, DMSO-d₆)

δ 7.45 (t, J = 8.80 Hz, 2 H), 7.62 (dd, J = 8.22, 4.11 Hz, 1 H), 7.84 (dd,

(300 MHz, CDCl₃) δ 7.61 (s, 1 H), 7.71 (dd, J = 8.48, 4.24 Hz, 1 H),

8.21ꢀ8.30 (m, 1 H), 9.12 (dd, J = 4.24, 1.61 Hz, 1 H), 9.36 (s, 1 H).

6-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)quinoxalines (6f).

This compound was prepared from 6-bromoquinoxaline (1.00 g, 4.78

mmol) using the general boronic ester formation method to afford the

title compound (1.01 g, 82%) as a viscous amber oil. MS (ESI pos. ion)

1

m/z: calc’d for C₁₄H₁₇BN₂O₂: 256.1; found 257.2 (Mþ1). H NMR

4745

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

Figure 5. In vivo eﬃcacy of compound 16d in a PTEN Null U-87 MG glioblastoma xenograft model. Data represent the mean (n = 10) ( standard

deviation. Statistical signiﬁcance was evaluated by Repeated Measures ANOVA followed by Dunnett post hoc test.

J = 8.61, 5.09 Hz, 2 H), 8.05ꢀ8.11 (m, 1 H), 8.12ꢀ8.20 (m, 2 H), 8.35

(s, 1 H), 8.47 (d, J = 8.02 Hz, 1 H), 8.76 (d, J = 1.96 Hz, 1 H), 8.97 (d, J =

4.11 Hz, 1 H), 10.54 (s, 1 H).

C₁₉H₁₂ClFN₄O₂S (M þ H): 415.0425; found 415.0410. ¹H NMR

(300 MHz, MeOD) δ 7.30 (t, J = 8.77 Hz, 2 H), 7.76 (dd, J = 8.18, 4.38 Hz,

1 H), 7.89 (dd, J = 8.99, 5.04 Hz, 2 H), 8.44 (d, J = 2.34 Hz, 1 H), 8.62

(dd, J = 8.18, 1.90 Hz, 1 H), 8.70 (d, J = 2.34 Hz, 1 H), 8.79 (d, J = 2.48

Hz, 1 H), 9.15 (dd, J = 4.38, 1.90 Hz, 1 H), 9.39 (d, J = 2.48 Hz, 1 H).

N-(2-Chloro-5-(quinoxalin-6-yl)pyridin-3-yl)-4-fluorobenzenesul-

fonamide (7f). This compound was prepared from 4a (0.300 g, 0.82

mmol) and 6f (0.252 g, 0.98 mmol) according to the general Suzuki

coupling method to afford the title compound (0.150 g, 44%) as a light-

brown solid. MS (ESI pos. ion) m/z: calc’d for C₁₉H₁₂ClFN₄O₂S: 414.0;

found 415.0 (Mþ1). HRMS calc’d for C₁₉H₁₂ClFN₄O₂S (M þ H):

415.0425; found 415.0410. ¹H NMR (400 MHz, DMSO-d₆) δ 7.44 (t,

J = 8.78 Hz, 2 H), 7.84 (dd, J = 8.78, 5.27 Hz, 2 H), 8.13ꢀ8.29 (m, 3 H),

8.42 (s, 1 H), 8.83 (s, 1 H), 9.03 (d, J = 11.04 Hz, 2 H), 10.58 (s, 1 H).

N-(2-Chloro-5-(quinazolin-6-yl)pyridin-3-yl)-4-fluorobenzenesul-

fonamide (7g). This compound was prepared from 4b (0.110 g, 0.27

mmol) and 6-bromoquinazoline (0.062 g, 0.30 mmol) according to the

general Suzuki coupling method to afford the title compound (0.060 g,

54%) as a white solid. MS (ESI pos. ion) m/z: calc’d for C₁₉H₁₂FClN₄-

O₂S: 414.0; found 414.9 (Mþ1) observed in basic LC/MS condition

(NH₄OH as additive). In acidic LC/MS condition with 0.1% TFA, the

hydrate mass (Mþ19) was observed. HRMS calc’d for C₁₉H₁₂ClFN₄O₂S

(M þ H): 415.0425; found 415.0410. ¹H NMR (400 MHz, DMSO-d₆)

δ 7.45 (t, J = 8.80 Hz, 2 H), 7.84 (dd, J = 8.71, 5.18 Hz, 2 H), 8.17 (d, J =

8.80 Hz, 1 H), 8.23 (d, J = 2.35 Hz, 1 H), 8.38 (dd, J = 8.80, 1.76 Hz,

1 H), 8.55 (d, J = 1.37 Hz, 1 H), 8.78 (d, J = 2.15 Hz, 1 H), 9.37 (s, 1 H)

9.71 (s, 1 H), 10.59 (br s, 1 H).

N-(2-Chloro-5-(isoquinolin-6-yl)pyridin-3-yl)-4-fluorobenzenesulfon-

amide (7b). This compound was prepared from 4b (0.262 g, 0.63 mmol)

and 6-bromoisoquinoline (0.120 g, 0.58 mmol) according to the general

Suzuki coupling method to afford the title compound (0.020 g, 8.4%) as

a white solid. MS (ESI pos. ion) m/z: calc’d for C₂₀H₁₃ClFN₃O₂S: 413.0;

found: 414.0 (Mþ1). HRMS calc’d for C₂₀H₁₃ClFN₃O₂S (M þ H):

1

414.0472; found 414.0483. H NMR (400 MHz, DMSO-d₆) δ 7.44

(t, J = 8.80 Hz, 2 H), 7.82ꢀ7.85 (m, 2 H), 7.93 (d, J = 5.87 Hz, 1 H), 8.00

(dd, J = 8.61, 1.37 Hz, 1 H), 8.16 (d, J = 2.15 Hz, 1 H), 8.29 (d, J = 8.61

Hz, 1 H), 8.33 (s, 1 H), 8.58 (d, J = 5.67 Hz, 1 H), 8.75 (s, 1 H), 9.40 (s,

1 H), 10.58 (s, 1 H).

N-(2-Chloro-5-(1,5-naphthyridin-2-yl)pyridin-3-yl)-4-fluorobenzene-

sulfonamide (7c). This compound was prepared from 4b (0.135 g, 0.33

mmol) and 6c (0.054 g, 0.33 mmol) according to the general Suzuki

coupling method to afford the title compound (0.078 g, 57%) as a white

solid. MS (ESI pos. ion) m/z: calc’d for C₁₉H₁₂ClFN₄O₂S: 414.0; found

415.0 (Mþ1). HRMS calc’d for C₁₉H₁₂ClFN₄O₂S (M þ H): 415.0425;

found 415.0420. ¹H NMR (300 MHz, CDCl₃) δ 7.07 (s, 1 H), 7.16 (t,

J = 8.55 Hz, 2 H), 7.73 (dd, J = 8.55, 4.17 Hz, 1 H), 7.88 (dd, J = 8.92,

4.97 Hz, 2 H), 8.11 (d, J = 8.77 Hz, 1 H), 8.49 (d, J = 8.48 Hz, 1 H), 8.55

(d, J = 8.77 Hz, 1 H), 8.83 (d, J = 2.19 Hz, 1 H), 8.94 (d, J = 2.19 Hz,

1 H), 9.03 (dd, J = 4.09, 1.61 Hz, 1 H).

N-(2-Chloro-5-(1,7-naphthyridin-6-yl)pyridin-3-yl)-4-fluoroben-

zenesulfonamide (7d). This compound was prepared from 4b (0.062 g,

0.15 mmol) and 6d (0.042 g, 0.15 mmol) according to the general

Suzuki coupling method to afford the title compound (0.046 g, 73%) as a

white solid. MS (ESI pos. ion) m/z: calc’d for C₁₉H₁₂ClFN₄O₂S: 414.0;

found 415.0 (Mþ1). HRMS calc’d for C₁₉H₁₂ClFN₄O₂S (M þ H):

N-(2-Chloro-5-(cinnolin-6-yl)pyridin-3-yl)-4-fluorobenzenesulfo-

namide (7h). This compound was prepared from 4b (0.296 g, 0.72

mmol) and 6-bromocinnoline (0.150 g, 0.72 mmol) according to the

general Suzuki coupling method to afford the title compound (0.015 g, 5%)

as a yellow solid. MS (ESI pos. ion) m/z: calc’d for C₁₉H₁₂ClFN₄O₂S:

414.0; found 415.0 (Mþ1). HRMS calc’d for C₁₉H₁₂ClFN₄O₂S (M þ H):

1

415.0425; found 415.0430. H NMR (300 MHz, CDCl₃) δ 7.01 (s,

1 H), 7.10ꢀ7.21 (m, 2 H), 7.68 (dd, J = 8.33, 4.24 Hz, 1 H), 7.82ꢀ7.91

(m, 2 H), 8.11 (d, J = 0.73 Hz, 1 H), 8.27 (d, J = 7.75 Hz, 1 H), 8.78 (d,

J = 2.19 Hz, 1 H), 8.95 (d, J = 2.19 Hz, 1 H), 9.08 (dd, J = 4.17, 1.68 Hz,

1 H), 9.62 (s, 1 H).

N-(2-Chloro-5-(1,8-naphthyridin-3-yl)pyridin-3-yl)-4-fluoroben-

zenesulfonamide (7e). This compound was prepared from 4b (0.140 g,

0.35 mmol) and 3-bromo-1,8-naphthyridine (0.073 g, 0.35 mmol) accord-

ing to the general Suzuki coupling method to afford the title compound

(0.098 g, 67%) as a white solid. MS (ESI pos. ion) m/z: calc’d for

C₁₉H₁₂ClFN₄O₂S: 414.0; found 415.0 (Mþ1). HRMS calc’d for

1

415.0425; found 415.0410. H NMR (400 MHz, DMSO-d₆) δ 7.44

(t, J = 8.78 Hz, 2 H), 7.83 (dd, J = 8.78, 5.27 Hz, 2 H), 8.24 (s, 1 H),

8.28ꢀ8.31 (m, 2 H), 8.46 (s, 1 H), 8.60 (d, J = 9.03 Hz, 1 H), 8.79

(s, 1 H), 9.46 (d, J = 5.52 Hz, 1 H), 10.60 (s, 1 H).

4-Chloro-6-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)quinoline

(15a). This compound was prepared from 6-bromo-4-chloroquinoline

(1.5 g, 6.2 mmol) using the general boronic ester formation method to

afford the title compound (1.45 g, 81%) as a brown solid. MS (ESI pos. ion)

m/z: calc’d for C₁₅H₁₇BClNO₂: 289.1; found 290.3 (Mþ1). ¹H NMR

4746

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

(300 MHz, CDCl₃) δ 1.41 (s, 12 H), 7.51 (d, J = 4.82 Hz, 1 H), 8.05ꢀ

8.12 (m, 1 H), 8.14 (s, 1 H), 8.73 (s, 1 H), 8.81 (d, J = 4.68 Hz, 1 H).

6-Bromo-N,N-dimethylquinolin-4-amine (15b) and 6-bromo-

N-(2-methoxyethyl)quinolin-4-amine (15c). These compounds were

prepared from 6-bromo-4-chloroquinoline (0.2 g, 0.8 mmol) and

2-methoxyethylamine (0.7 mL, 8 mmol) according to the general

S_N2⁰reaction method to afford the following compounds. (i) 6-

Bromo-N,N-dimethylquinolin-4-amine (15b) (0.150 g, 72%) as a yellow

oil. MS (ESI pos. ion) m/z: calc’d for C₁₁H₁₁BrN₂: 250.0; found 251.0,

253.0 (Mþ1, Mþ3). ¹H NMR (300 MHz, CDCl₃) δ 3.03 (s, 6 H), 6.78

(d, J = 5.12 Hz, 1 H), 7.70 (dd, J = 8.92, 2.19 Hz, 1 H), 7.90 (d, J = 9.06

Hz, 1 H), 8.22 (d, J = 2.19 Hz, 1 H), 8.66 (d, J = 5.12 Hz, 1 H).

(ii) 6-Bromo-N-(2-methoxyethyl)quinolin-4-amine (15c) (0.050 g, 22%)

as a light-brown solid. MS (ESI pos. ion) m/z: calc’d for C₁₂H₁₃BrN₂:

280.0; found 281.0, 283.0 (Mþ1, Mþ3). ¹H NMR (300 MHz, CDCl₃)

δ 3.42ꢀ3.55 (m, 5 H), 3.75 (t, J = 5.12 Hz, 2 H), 5.28 (br s, 1 H), 6.46

(d, J = 5.41 Hz, 1 H), 7.70 (dd, J = 9.06, 2.05 Hz, 1 H), 7.86 (d, J = 9.06 Hz,

1 H), 7.92 (d, J = 2.05 Hz, 1 H), 8.57 (d, J = 5.26 Hz, 1 H).

6-Bromo-4-morpholinoquinoline (15d). This compound was pre-

pared from 6-bromo-4-chloroquinoline (3.0 g, 12 mmol) and morpho-

line (3.2 g, 37 mmol) according to the general S_N2⁰method to afford the

title compound (3.4 g, 94%) as an off-white solid. MS (ESI pos. ion)

m/z: calc’d for C₁₃H₁₃BrN₂O: 292.0; found: 292.9, 294.9 (Mþ1, Mþ3).

¹H NMR(300 MHz, CDCl₃) δ 3.17ꢀ3.26 (m, 4 H), 3.96ꢀ4.04 (m, 4 H),

6.89 (d, J = 4.97 Hz, 1 H), 7.73 (dd, J = 8.92, 2.19 Hz, 1 H), 7.94 (d,

J = 8.92 Hz, 1 H), 8.16 (d, J = 2.19 Hz, 1 H), 8.76 (d, J = 4.97 Hz, 1 H).

4-(4-(Pyridin-4-ylmethyl)piperazin-1-yl)quinolin-6-ylboronic Acid

(15e). This compound was prepared from 15a (0.360 g, 1.24 mmol) and

1-(pyridin-4-ylmethyl)piperazine (0.500 g, 2.82 mmol) according to the

general S_N2⁰method to afford the title compound (0.370 g, 86%) as a

brown solid. MS (ES, pos.) m/z: calc’d for C₁₉H₂₁BN₄O₂: 348.2; found

349.2 (Mþ1).

(2.01 g, 4.87 mmol) and 15d (1.58 g, 5.38 mmol) according to the

general Suzuki coupling method to afford the title compound (1.71 g,

70%) as an off-white solid. MS (ESI pos. ion) m/z: calc’d for C₂₄H₂₀-

ClFN₄O₃S: 498.1; found 499.0 (Mþ1). HRMS calc’d for C₂₄H₂₀ClF-

1

N₄O₃S (M þ H): 499.0998; found 499.1010. H NMR (400 MHz,

DMSO-d₆) δ 3.27ꢀ3.33 (m, 4 H), 3.87ꢀ3.93 (m, 4 H), 7.09 (d, J = 5.28

Hz, 1 H), 7.42 (t, J = 8.80 Hz, 2 H), 7.79ꢀ7.87 (m, 2 H), 7.99ꢀ8.04 (m,

1 H), 8.06ꢀ8.12 (m, 2 H), 8.21 (d, J = 1.76 Hz, 1 H), 8.65 (d, J = 2.15

Hz, 1 H), 8.75 (d, J = 5.28 Hz, 1 H), 10.87 (br s, 1 H).

N-(2-Chloro-5-(4-(4-(pyridin-4-ylmethyl)piperazin-1-yl)quinolin-

6-yl)pyridin-3-yl)-4-fluorobenzenesulfonamide (16e). This com-

pound was prepared from 4a (0.400 g, 1.1 mmol) and 15e (0.370 g, 1.06

mmol) according to the general Suzuki coupling method to afford the

title compound (0.320 g, 51% yield) as a yellow solid. MS (ES, pos.) m/

z: calc’d for C₃₀H₂₆ClFN₆O₂S 588.2; found 589.1(Mþ1). HRMS calc’d

for C₃₀H₂₆ClFN₆O₂S (M þ H): 589.1579; found 589.1582. ¹H NMR

(400 MHz, DMSO-d₆) δ 2.74 (br s, 4 H), 3.35 (br s, 4 H, overlapped

with water), 3.68 (s, 2 H), 7.08 (d, J = 5.09 Hz, 1 H), 7.35ꢀ7.45 (m, 4

H), 7.83 (dd, J = 8.71, 5.18 Hz, 2 H), 7.97ꢀ8.03 (m, 1 H), 8.05ꢀ8.13

(m, 3 H), 8.17 (s, 1 H), 8.54 (d, J = 4.70 Hz, 2 H), 8.61 (d, J = 1.76 Hz,

1 H), 8.73 (d, J = 5.09 Hz, 1 H), 10.94 (br s, 1H).

N-(2-Chloro-5-(4-(piperidin-1-yl)quinolin-6-yl)pyridin-3-yl)-4-fluoro-

benzenesulfonamide (16f). This compound was prepared from 15a

(0.07 g, 0.2 mmol) and piperidine (0.05 mL, 0.5 mmol) according to the

general S_N2⁰method to afford the title compound (0.04 g, 52%) as a

white solid. MS (ESI pos. ion) m/z: calc’d for C₂₅H₂₂ClFN₄O₂S 496.1;

found 497.5 (Mþ1). HRMS calc’d for C₂₅H₂₂ClFN₄O₂S (M þ H):

497.1205; found 497.1210. ¹H NMR (300 MHz, CDCl₃) δ 1.67ꢀ1.79

(m, 2 H) 1.83ꢀ1.96 (m, 4 H) 3.17ꢀ3.31 (m, 4 H) 6.88 (d, J=4.97Hz, 1H)

7.11ꢀ7.21 (m, 2 H) 7.77ꢀ7.90 (m, 3 H) 8.16 (s, 1 H) 8.21 (d, J = 2.05 Hz,

1 H) 8.34 (d, J = 2.34 Hz, 1 H) 8.49 (s, 1 H) 8.74 (d, J = 5.12 Hz, 1 H).

N-(2-Chloro-5-(4-(4-hydroxypiperidin-1-yl)quinolin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (16g). This compound was pre-

pared from 15a (0.08 g, 0.2 mmol) and 4-hydroxypiperidine (0.2 g,

2 mmol) according to the general S_N2⁰method to afford the title com-

pound (0.035 g, 38%) as a white solid. MS (ESI pos. ion) m/z: calc’d for

C₂₅H₂₂ClFN₄O₃S: 493.1; found 494.3 (Mþ1). HRMS calc’d for C₂₅H₂₂-

ClFN₄O₃S (M þ H): 513.1154; found 513.1130. ¹H NMR (300 MHz,

CDCl₃) δ 1.83ꢀ2.02 (m, 2 H), 2.21 (ddd, J = 6.58, 3.22, 3.07 Hz, 2 H),

3.20ꢀ3.38 (m, 2 H), 3.56ꢀ3.78 (m, 2 H), 4.12 (ddd, J = 8.26, 5.85, 2.85

Hz, 1 H), 6.96 (d, J = 5.55 Hz, 1 H), 7.19 (t, J = 8.55 Hz, 2 H), 7.85 (dd,

J = 9.06, 4.97 Hz, 3 H), 7.88ꢀ7.94 (m, 1 H), 8.19 (s, 1 H), 8.34 (d, J =

2.19 Hz, 1 H), 8.49 (d, J = 2.34 Hz, 1 H), 8.71 (d, J = 5.70 Hz, 1 H).

N-(2-Chloro-5-(4-(4-isopropylpiperazin-1-yl)quinolin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (16h). This compound was pre-

pared from 15a (0.07 g, 0.2 mmol) and 1-isopropylpiperazine (0.1 mL,

0.8 mmol) according to the general S_N2⁰method to afford the title

compound (0.055 g, 65%) as a white solid. MS (ESI pos. ion) m/z: calc’d

for C₂₇H₂₇ClFN₅O₂S 539.2; found 540.3 (Mþ1). HRMS calc’d for

C₂₇H₂₇ClFN₅O₂S (M þ H): 540.1626; found 540.1650. ¹H NMR

(300 MHz, CDCl₃) δ 1.15 (d, J = 6.58 Hz, 6 H), 2.75ꢀ2.95 (m, 5 H),

3.27ꢀ3.39 (m, 4 H), 6.93 (d, J = 4.97 Hz, 1 H), 7.18 (t, J = 8.55 Hz, 2 H),

7.77ꢀ7.92 (m, 3 H), 8.19 (d, J = 8.62 Hz, 1 H), 8.27 (d, J = 1.90 Hz,

1 H), 8.37 (d, J = 2.34 Hz, 1 H), 8.51 (d, J = 2.34 Hz, 1 H), 8.79 (d, J =

4.97 Hz, 1 H).

N-(2-Chloro-5-(4-chloroquinolin-6-yl)pyridin-3-yl)-4-fluoroben-

zenesulfonamide (16a). This compound was prepared from 4a (2 g,

5 mmol) and 15a (2 g, 7 mmol) according to the general Suzuki coupling

method to afford the title compound (1.5 g, 69%) as a light-yellow solid.

MS (ESI pos. ion) m/z: calc’d for C₂₀H₁₂Cl₂FN₃O₂S: 447.5; found

448.3 (Mþ1). ¹H NMR (300 MHz, CDCl₃) δ 6.99 (br s, 1 H), 7.13ꢀ

7.24 (m, 2 H), 7.60 (d, J = 4.68 Hz, 1 H), 7.88 (dd, J = 8.99, 4.90 Hz,

2 H), 7.97 (dd, J = 8.77, 2.05 Hz, 1 H), 8.28 (d, J = 8.77 Hz, 1 H), 8.33 (d,

J = 2.19 Hz, 1 H), 8.38 (d, J = 1.90 Hz, 1 H), 8.52 (d, J = 2.34 Hz, 1 H),

8.87 (d, J = 4.68 Hz, 1 H).

N-(2-Chloro-5-(4-(dimethylamino)quinolin-6-yl)pyridin-3-yl)-4-

fluorobenzenesulfonamide (16b). This compound was prepared from

4b (0.1 g, 0.3 mmol) and 15b (0.07 g, 0.3 mmol) according to the general

Suzuki coupling method to afford the title compound (0.055 g, 43%) as a

light-yellow solid. MS (ESI pos. ion) m/z: calc’d for C₂₂H₁₈ClFN₄O₂S:

456.1; found 457.0 (Mþ1). HRMS calc’d for C₂₂H₁₈ClFN₄O₂S (M þ H):

457.0893; found 457.0890. ¹H NMR (300 MHz, CDCl₃) δ 3.12 (s, 6 H),

6.85 (d, J = 5.12 Hz, 1 H), 7.18 (t, J = 8.55 Hz, 2 H), 7.84 (td, J = 8.66,

3.58 Hz, 3 H), 8.16 (d, J = 8.77 Hz, 1 H), 8.24 (d, J = 2.05 Hz, 1 H), 8.32 (d,

J = 2.19 Hz, 1 H), 8.50 (d, J = 2.34 Hz, 1 H), 8.72 (d, J = 5.12 Hz, 1 H).

N-(2-Chloro-5-(4-(2-methoxyethylamino)quinolin-6-yl)pyridin-3-yl)-

4-fluorobenzenesulfonamide (16c). This compound was prepared

from 4b (0.07 g, 0.2 mmol) and 15c (0.05 g, 0.2 mmol) according to the

general Suzuki coupling method to afford the title compound (0.015 g,

17%) as a light-yellow solid. MS (ESI pos. ion) m/z: calc’d for C₂₃H₂₀-

ClFN₄O₃S: 486.1; found: 487.0 (Mþ1). HRMS calc’d for C₂₃H₂₀ClF-

N-(5-(4-(4-Benzylpiperazin-1-yl)quinolin-6-yl)-2-chloropyridin-3-

yl)-4-fluorobenzenesulfonamide (16i). This compound was prepared

from 15a (0.05 g, 0.1 mmol) and 1-benzylpiperazine (0.2 g, 1 mmol)

according to the general S_N2⁰method to afford the title compound

(0.035 g, 53%) as a white solid. MS (ESI pos. ion) m/z: calc’d for

C₃₁H₂₇ClFN₅O₂S 587.2; found 588.3 (Mþ1). HRMS calc’d for

C₃₁H₂₇ClFN₅O₂S (M þ H): 588.1626; found 588.1630. ¹H NMR

(300 MHz, CDCl₃) δ 2.81 (br s, 4 H), 3.33 (br s, 4 H), 3.67 (s, 2 H), 6.92

(d, J = 5.12 Hz, 1 H), 7.17 (t, J = 8.48 Hz, 2 H), 7.29ꢀ7.44 (m, 5 H), 7.85

1

N₄O₃S (M þ H): 487.0998; found 487.1000. H NMR (300 MHz,

MeOD) δ 3.41 (s, 3 H), 3.76 (s, 4 H), 6.87 (d, J = 6.87 Hz, 1 H), 7.13ꢀ

7.25 (m, 2 H), 7.83ꢀ7.92 (m, 2 H), 7.94 (s, 1 H), 8.02ꢀ8.10 (m, 1 H),

8.20 (s, 1 H), 8.32ꢀ8.42 (m, 2 H), 8.59 (s, 1 H).

N-(2-Chloro-5-(4-morpholinoquinolin-6-yl)pyridin-3-yl)-4-fluoro-

benzenesulfonamide (16d). This compound was prepared from 4b

4747

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

(dd, J= 8.92, 4.82 Hz, 3 H), 8.18 (d, J=8.77Hz, 1H), 8.25(s, 1H), 8.35(d,

J = 2.34 Hz, 1 H), 8.50 (d, J = 2.34 Hz, 1 H), 8.78 (d, J = 4.97 Hz, 1 H).

N-(2-Chloro-5-(4-phenylquinolin-6-yl)pyridin-3-yl)-4-fluoroben-

zenesulfonamide (16j). This compound was prepared from 15a (0.07 g,

0.2 mmol) and phenylboronic acid (0.06 g, 0.5 mmol) according to the

general Suzuki coupling method to afford the title compound (0.030 g,

39%) as a light-yellow solid. MS (ESI pos. ion) m/z: calc’d for C₂₆H₁₇-

ClFN₃O₂S: 489.1; found 490.0 (Mþ1). HRMS calc’d for C₂₆H₁₇ClF-

off-white solid. The combined yield of this reaction was 11.11 g, 72%. To

this 7-bromoquinoxalin-2-ol (8.01 g, 35.6 mmol) was added toluene

(70 mL) and POCl₃(14.9 mL, 160 mmol) and the mixture was stirred at

room temperature for overnight. The reaction mixture was poured into

iceꢀwater (200 mL). The aqueous phase was extracted with EtOAc

(3 ꢁ 150 mL). The combined organic phases were washed with satu-

rated aqueous NaCl (400 mL), dried over sodium sulfate, filtered, and

concentrated. Upon addition of DCM to the residue, some of the

precipitated product was collected via filtration to afford the title

compound as a light-brown powder (3.07 g). The filtrate containing

the rest of the product was purified by silica gel column chromatography

(0ꢀ8% EtOAc in hexanes) to afford more of the title compound

(5.20 g) as a pink solid. Total yield was 8.27 g, 95%. MS (ESI pos. ion)

m/z: calc’d for C₈H₄BrClN₂: 241.9; found 242.9, 244.9 (Mþ1, Mþ3).

¹H NMR (300 MHz, CDCl₃) δ 7.83ꢀ7.90 (m, 1 H), 7.95ꢀ8.02 (m,

1 H), 8.21 (d, J = 2.05 Hz, 1 H), 8.78 (s, 1 H).

7-Bromo-2-fluoroquinoxaline (17b). To a 150 mL round-bottomed

flask, 7a (1.73 g, 7.12 mmol) was suspended in DMSO (20 mL).

Tetrabutylammonium fluoride (1.0 M in THF, 8.55 mL, 8.55 mmol)

was added. The reaction mixture was stirred at room temperature for 1.5 h.

Water (100 mL) was added and the resulting solid was collected, washed

with water, and air-dried to obtain the title compound (1.54 g, 95%) as a

light-yellow solid. This compound was not detectable in mass spectros-

copy. ¹H NMR (300 MHz, DMSO-d₆) δ 7.92ꢀ8.41 (m, 3 H), 8.90ꢀ

9.14 (m, 1 H).

7-Bromo-N,N-dimethylquinoxalin-2-amine (17c). This compound

was prepared from 17b (0.113 g, 0.499 mmol) and dimethylamine (2 M in

THF, 1.0 mL, 2.00 mmol) according to the general S_N2⁰method to afford

the title compound (0.124 g, 99%) as a pale-yellow solid. MS (ESI, pos. ion)

m/z: calc’d for C₁₀H₁₀BrN₃: 251.0; found 251.9, 253.9 (Mþ1, Mþ3). ¹H

NMR (400 MHz, CDCl₃) δ 3.27 (s, 6 H), 7.42 (dd, J = 8.71, 2.05 Hz, 1 H),

7.71 (d, J = 8.61 Hz, 1 H), 7.86 (d, J = 2.15 Hz, 1 H), 8.49 (s, 1 H).

7-Bromo-2-morpholinoquinoxaline (17d). This compound was

prepared from 17c (0.201 g, 0.826 mmol) and morpholine (0.200 mL,

2.29 mmol) according to the general S_N2⁰method to afford the title

compound (0.124 g, 51%) as light-yellow crystals. MS (ESI, pos. ion) m/z:

calc’d for C₁₂H₁₂BrN₃O 293.0; found 294.0, 295.9 (Mþ1, Mþ3). ¹H

NMR (300 MHz, CDCl₃) δ 3.72ꢀ3.83 (m, 4 H), 3.83ꢀ3.93 (m, 4 H),

7.49 (dd, J = 8.8, 2.2 Hz, 1 H), 7.75 (d, J = 8.6 Hz, 1 H), 7.88 (d, J = 2.0 Hz,

1 H), 8.55 (s, 1 H).

1

N₃O₂S (M þ H): 490.0784; found 490.0770. H NMR (300 MHz,

CDCl₃) δ 7.01ꢀ7.12 (m, 2 H), 7.44 (d, J = 4.53 Hz, 1 H), 7.54ꢀ7.64 (m,

5 H), 7.78 (dd, J= 9.06, 4.97 Hz, 2 H), 7.92 (dd, J= 8.77, 2.19 Hz, 1 H), 8.08

(d, J = 1.75 Hz, 1 H), 8.20 (d, J = 2.34 Hz, 1 H), 8.32 (d, J = 8.62 Hz, 1 H),

8.37 (d, J = 2.19 Hz, 1 H), 9.02 (d, J = 4.53 Hz, 1 H).

N-(2-Chloro-5-(4-(pyridin-3-yl)quinolin-6-yl)pyridin-3-yl)-4-fluoro-

benzenesulfonamide (16k). To a 5 mL vial was added 15a (0.100 g,

0.2 mmol), 3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine (0.05 g,

0.3 mmol), potassium acetate (0.05 g, 0.6 mmol), dichlorobis(di-tert-butyl-

phenylphosphine)palladium(II) (0.007 g, 0.01 mmol), water (0.3 mL),

and 1-butanol (3 mL). The reaction vial was heated at 100 ꢀC for 2 h.

The solvent was removed and the residue was partitioned between water

and CHCl₃. The organic layer was dried over magnesium sulfate and

concentrated in vacuo. The crude product was purified by silica gel

chromatography (DCM/EtOAc/MeOH = 75:22:3) to afford the title

compound (0.070 g, 64%) as a white solid. MS (ESI pos. ion) m/z: calc’d

for C₂₅H₁₆ClFN₄O₂S 490.1; found 490.9 (Mþ1). HRMS calc’d for

C₂₅H₁₆ClFN₄O₂S (M þ H): 491.0734; found 491.0730. ¹H NMR (300

MHz, CDCl₃) δ 7.01ꢀ7.07 (m, 1 H), 7.11ꢀ7.21 (m, 2 H), 7.47 (d, J =

4.53 Hz, 1 H), 7.58 (d, J = 3.07 Hz, 1 H), 7.74ꢀ7.84 (m, 2 H), 7.89ꢀ8.00

(m, 3 H), 8.20 (d, J = 2.34 Hz, 1 H), 8.34ꢀ8.42 (m, 2 H), 8.79ꢀ8.88 (m,

2 H), 9.07 (d, J = 4.38 Hz, 1 H).

N-(2-Chloro-5-(4-(pyridin-4-yl)quinolin-6-yl)pyridin-3-yl)-4-fluoro-

benzenesulfonamide (16l). This compound was prepared from 15a

(0.1 g, 0.2 mmol) and 4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-

yl)pyridine (0.05 g, 0.3 mmol) in a similar manner as described for

16k to afford the title compound (0.07 g, 64%) as a white solid. MS (ESI

pos. ion) m/z: calc’d for C₂₅H₁₆ClFN₄O₂S 490.1; found 490.9 (Mþ1).

HRMS calc’d for C₂₅H₁₆ClFN₄O₂S (M þ H): 491.0734; found

1

491.0740. H NMR (300 MHz, CDCl₃) δ 7.04ꢀ7.19 (m, 3 H), 7.44

(d, J = 4.38 Hz, 1 H), 7.51 (dd, J = 4.46, 1.53 Hz, 2 H), 7.72ꢀ7.84 (m, 2

H) 7.92ꢀ8.02 (m, 2 H), 8.22 (d, J = 2.34 Hz, 1 H), 8.31ꢀ8.40 (m, 2 H),

8.87 (dd, J = 4.38, 1.61 Hz, 2 H), 9.07 (d, J = 4.38 Hz, 1 H).

N-(2-Chloro-5-(3-chloroquinoxalin-6-yl)pyridin-3-yl)-4-fluoro-

benzenesulfonamide (18a). This compound was prepared from 4b

(0.189 g, 0.457 mmol) and 17a (0.100 g, 0.410 mmol) according to the

general Suzuki coupling method, except bis(di-tert-butylphosphine)-

dichloropalladium(II) was used as a catalyst, to afford the title com-

pound (0.126 g, 68%) as an off-white solid. MS (ESI pos. ion) m/z: calc’d

N-(2-Chloro-5-(4-(4-(dimethylamino)phenyl)quinolin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (16m). This compound was

prepared from 15a (0.07 g, 0.2 mmol) and N,N-dimethyl-4-(4,4,5,

5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzenamine (0.05 g, 0.2 mmol)

in a similar manner to prepare 16k to afford the title compound (0.035 g,

42%) as a red solid. MS (ESI pos. ion) m/z: calc’d for C₂₈H₂₂ClFN₄O₂S

532.1; found 532.9 (Mþ1). HRMS calc’d for C₂₈H₂₂ClFN₄O₂S (M þ

1

for C₁₉H₁₁Cl₂FN₄O₂S 448.0; found 448.8, 450.8 (Mþ1, Mþ3). H

NMR (300 MHz, DMSO-d₆) δ 7.38ꢀ7.50 (m, 2 H), 7.79ꢀ7.89 (m,

2 H), 8.18ꢀ8.35 (m, 3 H), 8.40 (d, J = 1.75 Hz, 1 H), 8.80 (d, J = 2.34 Hz,

1 H), 9.05 (s, 1 H), 10.56 (br s, 1 H).

1

H): 533.1205; found 533.1190. H NMR (300 MHz, CDCl₃) δ 3.07

(s, 6 H), 6.88ꢀ6.97 (m, 2 H), 7.04ꢀ7.13 (m, 3 H), 7.42 (d, J = 4.53 Hz,

1 H), 7.47ꢀ7.54 (m, 2 H), 7.77ꢀ7.84 (m, 2 H), 7.90 (dd, J = 8.77, 2.05 Hz,

1 H), 8.22ꢀ8.26 (m, 2 H), 8.28 (d, J = 8.77 Hz, 1 H), 8.41 (d, J = 2.34 Hz, 1

H), 8.96 (d, J = 4.53 Hz, 1 H).

N-(2-Chloro-5-(3-(dimethylamino)quinoxalin-6-yl)pyridin-3-yl)-

4-fluorobenzenesulfonamide (18c). This compound was prepared

from 4b (0.161 g, 0.389 mmol) and 17c (0.0952 g, 0.378 mmol)

according to the general Suzuki coupling method to afford the title

compound (0.136 g, 79%) as a yellow solid. MS (ESI, pos. ion) m/z:

calc’d for C₂₁H₁₇ClFN₅O₂S 457.1; found 457.9 (Mþ1). HRMS calc’d

for C₂₁H₁₇ClFN₅O₂S (M þ H): 458.0846; found 458.0870. ¹H NMR

(400 MHz, DMSO-d₆) δ 3.27 (s, 6 H), 7.45 (t, J = 8.80 Hz, 2 H), 7.64

(dd, J = 8.61, 1.76 Hz, 1 H), 7.78ꢀ7.88 (m, 3 H), 7.93 (d, J = 8.41 Hz, 1

H), 8.05 (d, J = 2.15 Hz, 1 H), 8.74 (s, 2 H), 10.52 (br s, 1 H).

N-(2-Chloro-5-(3-morpholinoquinoxalin-6-yl)pyridin-3-yl)-4-fluoro-

benzenesulfonamide (18d). This compound was prepared from 4b

(2.00 g, 4.85 mmol) and 17d (1.58 g, 5.36 mmol) according to the

general Suzuki coupling method to afford the title compound (1.07 g,

7-Bromo-2-chloroquinoxaline (17a). To a stirred solution of

quinoxalin-2-ol (10.0 g, 68.4 mmol) in 500 mL of AcOH in 1-L

round-bottomed flask, bromine (3.60 mL, 70.3 mmol) was added

slowly. The mixture was stirred at room temperature for 1.5 h. The

resulting precipitate was collected via filtration and was washed with

water and MeOH. This solid was dissolved into DMSO and water was

added to create a precipitate, which was collected via filtration and

washed with water and MeOH to afford 7-bromoquinoxalin-2-ol (3.85 g) as

a light-pink solid. To the first filtrate was added water (1 L) and the

resulting precipitate was collected and washed with copious amounts of

water to afford another batch of 7-bromoquinoxalin-2-ol (7.26 g) as an

4748

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

44%) as a pale-yellow solid. MS (ESI, pos. ion) m/z: calc’d for C₂₃H₁₉-

ClFN₅O₃S 499.1; found 500.0 (Mþ1). HRMS calc’d for C₂₃H₁₉ClF-

H), 8.01 (d, J = 8.48 Hz, 1 H), 8.24 (s, 1 H), 8.29 (d, J = 2.19 Hz, 1 H),

8.46 (d, J = 2.19 Hz, 1 H).

1

N₅O₂S (M þ H): 500.0951; found 500.0960. H NMR (300 MHz,

N-(2-Chloro-5-(3-(2-(piperidin-1-yl)ethylamino)quinoxalin-6-yl)-

pyridin-3-yl)-4-fluorobenzenesulfonamide (18k). 38.5 mg. Yield 63%.

MS (ESI pos. ion) m/z: calc’d for C₂₆H₂₆ClFN₆O₂S 540.2; found

540.8 (Mþ1). HRMS calc’d for C₂₆H₂₆ClFN₆O₂S (M þ H): 541.1579;

found 541.1570. ¹H NMR (300 MHz, CDCl₃) δ 1.85ꢀ2.23 (m, 7 H),

2.76 (br s, 2 H), 3.43 (br s, 2 H), 3.66 (br s, 1 H), 4.09 (br s, 2 H), 7.05 (s,

1 H), 7.14ꢀ7.22 (m, 2 H), 7.58 (dd, J = 8.62, 2.05 Hz, 1 H), 7.81ꢀ7.87

(m, 2 H), 7.89 (d, J = 1.75 Hz, 1 H), 8.02 (d, J = 8.48 Hz, 1 H), 8.34 (d,

J = 2.34 Hz, 1 H), 8.40 (s, 1 H), 8.50 (d, J = 2.34 Hz, 1 H).

N-(2-Chloro-5-(3-(2-morpholinoethylamino)quinoxalin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (18l). This compound was pre-

pared from 18b (0.045 g, 0.10 mmol) and 4-(2-aminoethyl)morpholine

(0.040 mL, 0.31 mmol) according to the general S_N2⁰method to afford

the title compound (0.043 g, 76%) as an off-white solid. MS (ESI pos. ion)

m/z: calc’d for C₂₅H₂₄ClFN₆O₃S 542.1; found 543.0 (Mþ1). HRMS

calc’d for C₂₅H₂₄ClFN₆O₃S (M þ H): 543.1372; found 543.1381. ¹H

NMR (300 MHz, DMSO-d₆) δ 2.55ꢀ2.88 (m, 6 H), 3.54ꢀ3.77 (m, 6

H), 7.39 (t, J = 8.84 Hz, 2 H), 7.56 (dd, J = 8.48, 2.05 Hz, 1 H), 7.71 (d,

J = 1.75 Hz, 2 H), 7.77ꢀ7.91 (m, 3 H) 7.97 (d, J = 2.19 Hz, 1 H), 8.37 (s,

1 H), 8.51 (br s, 1 H), 10.28 (br s, 1 H).

DMSO-d₆) δ 3.67ꢀ3.91 (m, 8 H), 7.37ꢀ7.52 (m, 2 H), 7.70 (dd, J = 8.6,

2.0 Hz, 1 H), 7.76ꢀ7.89 (m, 3 H), 7.96 (d, J = 8.5 Hz, 1 H), 8.06 (d, J =

2.3 Hz, 1 H), 8.72 (d, J = 2.2 Hz, 1 H), 8.87 (s, 1 H), 10.53 (s, 1 H).

N-(2-Chloro-5-(3-(pyridin-4-yl)quinoxalin-6-yl)pyridin-3-yl)-4-

fluorobenzenesulfonamide (18e). This compound was prepared from

18a (0.106 g, 0.236 mmol) and pyridine-4-ylboronic acid (0.043 g, 0.35

mmol) according to the general Suzuki coupling method, except bis(di-

t-butylphosphine)dichloropalladium(II) was used as the catalyst and

potassium acetate was used for the base, to afford the title compound

(0.066 g, 57%) as a light-yellow solid. MS (ESI pos. ion) m/z: calc’d for

C₂₄H₁₅ClFN₅O₂S 491.1; found 491.9 (Mþ1). HRMS calc’d for

C₂₄H₁₅ClFN₅O₂S (M þ H): 492.0690; found 492.0700. ¹H NMR

(300 MHz, CDCl₃) δ 7.06ꢀ7.23 (m, 3 H), 7.82ꢀ7.92 (m, 2 H), 8.01

(dd, J = 8.77, 2.05 Hz, 1 H), 8.14 (dd, J = 4.53, 1.61 Hz, 2 H), 8.31 (d, J =

8.77 Hz, 1 H), 8.40 (t, J = 1.83 Hz, 2 H), 8.57 (d, J = 2.34 Hz, 1 H), 8.89

(dd, J = 4.68, 1.46 Hz, 2 H), 9.42 (s, 1 H).

General Method for Library Synthesis of 18fꢀ18k, 18mꢀ18p, and

18r. To a solution of 18b (40 mg, 92 μmol) in DMSO (1 mL) was added

the amine (1 M in DMF, 231 μL, 231 μmol) and DMSO (0.57 mL). The

reaction was heated at 60 ꢀC for 24 h. The reaction was then cooled to

23 ꢀC and submitted to mass-directed purification to give the product as

a yellow solid. The yield and spectroscopic data are shown below.

N-(2-Chloro-5-(3-(2-methoxyethylamino)quinoxalin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (18f). 6.4 mg. Yield 11%. MS (ESI

pos. ion) m/z: calc’d for C₂₂H₁₉ClFN₅O₃S 487.1; found 488.0 (Mþ1).

HRMS calc’d for C₂₂H₁₉ClFN₅O₃S (M þ H) 488.0951, found

N-(2-Chloro-5-(3-(piperidin-1-yl)quinoxalin-6-yl)pyridin-3-yl)-4-

fluorobenzenesulfonamide (18m). 16.2 mg. Yield 32%. MS (ESI pos.

ion) m/z: calc’d for C₂₄H₂₁ClFN₅O₂S 497.1; found 498.0 (Mþ1). HRMS

calc’d for C₂₄H₂₁ClFN₅O₂S (M þ H): 498.1158; found 498.1170. ¹H

NMR (300 MHz, CDCl₃) δ 1.70ꢀ1.86 (m, 6 H), 3.80ꢀ3.87 (m, 4 H),

7.04 (s, 1 H), 7.13ꢀ7.20 (m, 2 H), 7.53 (dd, J = 8.48, 2.05 Hz, 1 H), 7.81ꢀ

7.89 (m, 3 H), 7.97 (d, J = 8.48 Hz, 1 H), 8.33 (d, J = 2.34 Hz, 1 H), 8.50

(d, J = 2.34 Hz, 1 H), 8.63 (s, 1 H).

1

488.0950. H NMR (300 MHz, CDCl₃) δ 3.44 (s, 3 H), 3.70 (d, J =

4.82 Hz, 2 H), 3.80 (d, J = 5.55 Hz, 2 H), 7.16 (t, J = 8.62 Hz, 2 H), 7.51ꢀ

7.62 (m, 2 H), 7.81ꢀ7.89 (m, 3 H), 7.98 (d, J = 8.62 Hz, 1 H), 8.25 (s,

1 H), 8.33 (d, J = 2.19 Hz, 1 H), 8.50 (d, J = 2.19 Hz, 1 H).

N-(2-Chloro-5-(3-((2-methoxyethyl)(methyl)amino)quinoxalin-6-yl)-

pyridin-3-yl)-4-fluorobenzenesulfonamide (18g). 14.1 mg. Yield

31%. MS (ESI pos. ion) m/z: calc’d for C₂₃H₂₁ClFN₅O₃S 501.1; found

502.0 (Mþ1). HRMS calc’d for C₂₃H₂₁ClFN₅O₃S (M þ H): 502.1107;

found 502.1120. ¹H NMR (300 MHz, CDCl₃) δ 3.36 (s, 3 H), 3.39 (s,

3 H), 3.69ꢀ3.75 (m, 2 H), 3.94 (t, J = 5.33 Hz, 2 H), 7.01 (s, 1 H), 7.18

(t, J = 8.55 Hz, 2 H), 7.53 (dd, J = 8.48, 2.05 Hz, 1 H), 7.81ꢀ7.89 (m,

3 H), 7.99 (d, J = 8.48 Hz, 1 H), 8.33 (d, J = 2.19 Hz, 1 H), 8.51 (d, J =

2.34 Hz, 1 H), 8.61 (s, 1 H).

N-(2-Chloro-5-(3-(cyclohexylmethylamino)quinoxalin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (18h). 16.9 mg. Yield 29%. MS

(ESI pos. ion) m/z: calc’d for C₂₆H₂₅ClFN₅O₂S 525.1; found 526.0

(Mþ1). HRMS calc’d for C₂₆H₂₅ClFN₅O₂S (M þ H): 526.1470; found

526.1480. ¹H NMR (300 MHz, CDCl₃) δ 1.00ꢀ1.38 (m, 5 H),

1.66ꢀ1.93 (m, 6 H), 3.45 (t, J = 6.21 Hz, 2 H), 7.05 (br s, 1 H), 7.18

(t, J = 8.55 Hz, 2 H), 7.55 (dd, J = 8.40, 1.97 Hz, 1 H), 7.82ꢀ7.90 (m, 3

H), 7.98 (d, J = 8.48 Hz, 1 H), 8.31 (d, J = 2.19 Hz, 2 H), 8.48 (d, J = 2.19

Hz, 1 H).

N-(2-Chloro-5-(3-(4,4-difluoropiperidin-1-yl)quinoxalin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (18n). 5.2 mg. Yield 9%. MS (ESI

pos. ion) m/z: calc’d for C₂₄H₁₉ClF₃N₅O₂S 533.1; found 534.0. HRMS

calc’d for C₂₄H₁₉ClF₃N₅O₂S (M þ H): 534.0970; found 534.0970. ¹H

NMR (300 MHz, CDCl₃) δ 2.07ꢀ2.24 (m, 4 H), 3.97ꢀ4.04 (m, 4 H),

7.01 (s, 1 H), 7.12ꢀ7.21 (m, 2 H), 7.60 (dd, J = 8.48, 2.05 Hz, 1 H), 7.81ꢀ

7.91 (m, 3 H), 8.02 (d, J = 8.62 Hz, 1 H), 8.34 (d, J = 2.19 Hz, 1 H), 8.50

(d, J = 2.34 Hz, 1 H), 8.67 (s, 1 H).

N-(5-(3-((1s,5s)-3-Aza-bicyclo[3.2.2]nonan-3-yl)quinoxalin-6-yl)-

2-chloropyridin-3-yl)-4-fluorobenzenesulfonamide (18o). 3.9 mg.

Yield 7%. MS (ESI pos. ion) m/z: calc’d for C₂₇H₂₅ClFN₅O₂S 537.1;

found 538.0 (Mþ1). HRMS calc’d for C₂₇H₂₅ClFN₅O₂S (M þ H):

538.1470; found 538.1480. ¹H NMR (300 MHz, CDCl₃) δ 1.76 (t, J =

1.75 Hz, 8 H), 2.29 (br s, 2 H), 4.02 (d, J = 4.09 Hz, 4 H), 7.00 (s, 1 H),

7.12ꢀ7.25 (m, 2 H), 7.51 (dd, J = 8.48, 2.05 Hz, 1 H), 7.81ꢀ7.91 (m, 3

H), 7.97 (d, J = 8.48 Hz, 1 H), 8.33 (d, J = 2.34 Hz, 1 H), 8.50 (d, J = 2.34

Hz, 1 H), 8.69 (s, 1 H).

N-(2-Chloro-5-(3-(4-methoxypiperidin-1-yl)quinoxalin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (18p). 1.8 mg. Yield 3%. MS (ESI

pos. ion) m/z: calc’d for C₂₅H₂₃ClFN₅O₃S 527.1; found 528.0 (Mþ1).

HRMS calc’d for C₂₅H₂₃ClFN₅O₃S (M þ H): 528.1263; found

528.1271. ¹H NMR (300 MHz, CDCl₃) δ 1.43 (s, 3 H), 1.77 (br s, 2

H), 1.98ꢀ2.10 (m, 2 H), 3.51ꢀ3.67 (m, 3 H), 4.10ꢀ4.22 (m, 2 H), 7.01

(s, 1 H), 7.12ꢀ7.20 (m, 2 H), 7.55 (dd, J = 8.55, 1.97 Hz, 1 H), 7.80ꢀ

7.89 (m, 3 H), 7.98 (d, J = 8.62 Hz, 1 H), 8.33 (d, J = 2.34 Hz, 1 H), 8.50

(d, J = 2.19 Hz, 1 H), 8.65 (s, 1 H).

N-(2-Chloro-5-(3-(4-isopropylpiperazin-1-yl)quinoxalin-6-yl)pyridin-

3-yl)-4-fluorobenzenesulfonamide (18q). This compound was pre-

pared from 18b (0.031 g, 0.072 mmol) and 1-isopropylpiperazine

(0.026 mL, 0.18 mmol) according to the general S_N2⁰method to afford

the title compound (0.021 g, 54%) as a pale-yellow solid. MS (ESI pos.

ion) m/z: calc’d for C₂₆H₂₆ClFN₆O₂S 540.2; found 541.1 (Mþ1).

HRMS calc’d for C₂₆H₂₆ClFN₆O₂S (M þ H): 541.1579; found

541.1570. ¹H NMR (300 MHz, DMSO-d₆) δ 1.18 (d, J = 6.6 Hz, 6 H),

N-(5-(3-(Benzyl(methyl)amino)quinoxalin-6-yl)-2-chloropyridin-

3-yl)-4-fluorobenzenesulfonamide (18i). 9.3 mg. Yield 31%. MS (ESI

pos. ion) m/z: calc’d for C₂₇H₂₁ClFN₅O₂S 533.1; found 534.0 (Mþ1).

¹H NMR (300 MHz, CDCl₃) δ 3.33 (s, 3 H), 4.99 (s, 2 H), 7.01 (s, 1 H),

7.11ꢀ7.25 (m, 2 H), 7.27ꢀ7.38 (m, 5 H), 7.54 (dd, J = 8.48, 2.05 Hz, 1

H), 7.81ꢀ7.92 (m, 3 H), 7.99 (d, J = 8.48 Hz, 1 H), 8.32 (d, J = 2.19 Hz,

1 H), 8.48ꢀ8.57 (m, 2 H).

N-(2-Chloro-5-(3-(phenethylamino)quinoxalin-6-yl)pyridin-3-yl)-

4-fluorobenzenesulfonamide (18j). 20.8 mg. Yield 35%. MS (ESI pos.

ion) m/z:calc’dforC₂₇H₂₁ClFN₅O₂S 533.1; found 534.0(Mþ1). ¹HNMR

(300 MHz, CDCl₃) δ 3.11 (t, J = 6.87 Hz, 2 H), 3.90 (d, J = 4.97 Hz,

2 H), 7.11 (br s, 1 H), 7.18ꢀ7.25 (m, 3 H), 7.28 - 7.38 (m, 4 H), 7.62

(dd, J = 8.48, 1.90 Hz, 1 H), 7.86 (d, J = 1.90 Hz, 1 H), 7.88ꢀ7.95 (m, 2

4749

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

3.04 (s, 4 H), 3.14ꢀ3.23 (m, 1 H), 4.00 (s, 4 H), 7.32 (t, J = 8.8 Hz, 2 H),

7.61 (dd, J = 8.5, 1.9 Hz, 1 H), 7.71 (d, J = 1.8 Hz, 1 H), 7.75ꢀ7.86

(m, 2 H), 7.85ꢀ7.97 (m, 2 H), 8.26 (s, 1 H), 8.88 (s, 1 H).

NBS, N-bromosuccinimide; PD, pharmacodynamic; PI3K, phos-

phoinositide 3-kinase; PO, orally; PTEN, phosphate and tensin

homologue gene; QD, once daily; TBAF, tetrabutylammonium

ﬂuoride; TFA, triﬂuoroacetic acid; Vd,ss, volume of distribution at

steady state

tert-Butyl 3-((7-(6-chloro-5-(4-fluorophenylsulfonamido)pyridin-

3-yl)quinoxalin-2-ylamino)methyl)piperidine-1-carboxylate (18r).

21.0 mg. Yield 37%. MS (ESI pos. ion) m/z: calc’d for C₃₀H₃₂ClFN₆O₄S

626.2; found 626.8. ¹H NMR (300 MHz, CDCl₃) δ 1.38ꢀ1.60 (m, 11

H), 1.71 (d, J = 9.21 Hz, 1 H), 1.89ꢀ2.10 (m, 3 H), 2.97ꢀ3.23 (m, 2 H),

3.56 (br s, 2 H), 3.83 (d, J = 12.86 Hz, 1 H), 7.10ꢀ7.24 (m, 3 H), 7.62 (d,

J = 8.18 Hz, 1 H), 7.85ꢀ7.93 (m, 3 H), 8.04 (d, J = 8.48 Hz, 1 H), 8.27

(d, J = 2.19 Hz, 1 H), 8.45 (d, J = 2.34 Hz, 1 H), 8.50 (br s, 1 H).

N-(2-Chloro-5-(3-(piperidin-3-ylmethylamino)quinoxalin-6-yl)-

pyridin-3-yl)-4-fluorobenzenesulfonamide (18s). A solution of 18r

(5 mg, 8 μmol) in DCM (1 mL) was treated with TFA (1 mL) and stirred

at 23 ꢀC for 15 min. The reaction was then concentrated in vacuo. DCM

(1 mL) was added and the solvent was evaporated. This procedure was

repeated 2 additional times to give the TFA salt of the title compound

(5 mg, 100%) as a yellow film. MS (ESI pos. ion) m/z: calc’d for C₂₅H₂₄-

ClFN₆O₂S 526.1; found 527.0. HRMS calc’d for C₂₅H₂₄ClFN₆O₂S

(M þ H): 527.1423; found 527.1410. ¹H NMR (300 MHz, CDCl₃) δ

1.61ꢀ1.99 (m, 4 H), 2.17 (d, J = 14.62 Hz, 1 H), 2.64 (t, J = 11.84 Hz,

1 H), 2.71ꢀ2.83 (m, 1 H), 3.22 (br s, 1 H), 3.32ꢀ3.47 (m, 3 H),

7.02ꢀ7.09 (m, 2 H), 7.45 (dd, J = 8.48, 2.05 Hz, 1 H), 7.69ꢀ7.77 (m,

3 H), 7.83 (d, J = 8.48 Hz, 1 H), 8.17ꢀ8.21 (m, 2 H), 8.37 (d, J = 2.19 Hz,

1 H) (no exchangeable protons were observed because a small amount

of MeOD was added for solubility).

’ REFERENCES

(1) Cantley, L. C. The phosphoinositide 3-kinase pathway. Science

2002, 296, 1655–1657.

(2) Engelman, J. A.; Luo, J; Cantley, L. C. The evolution of

phosphatidylinositol 3 kinases as regulators of growth and metabolism.

Nat. Rev. Genetics 2006, 7, 606–619.

(3) Liu, P.; Cheng, H.; Roberts, T. M.; Zhao, J. Targeting the

phophoinositide 3-kinase pathway in cancer. Nat. Rev. Drug Discovery

2009, 8, 627–644.

(4) Yap, T. A.; Garrett, M. D.; Walton, M. I.; Raynaud, F.; De Bono,

J. S.; Workman, P. Targeting the PI3K-AKT-mTOR Pathway: Progress,

Pitfalls, and Promises. Curr. Opin. Pharm. 2008, 8, 393–412.

(5) Ihle, N. T.; Powis, G. Take your PIK: phosphotidylinositol-3-

kinase inhibitors race through the clinic and toward cancer therapy. Mol.

Cancer Ther. 2009, 8, 1–9.

(6) D’Angelo, N.; Kim, T.-S.; Andrews, K.; Booker, S. K.; Caenepeel,

S.; Freeman, D.; Jiang, J.; McCarter, J.; San Miguel, T.; Mullady, E.;

Schrag, M.; Subramanian, R.; Tang, J.; Wang, L.; Whittington, D. A.; Wu,

T.; Xi, N.; Xu, Y.; Yakowec, P.; Zhang, N.; Hughes, P.; Norman, M. H.

The discovery and optimization of a series of benzothiazole PI3K/

mTOR dual inhibitors. J. Med. Chem. 2011, 54, 1789–1811.

(7) Booker, S. K.; D’Angelo, N. D.; D’Amico, D. C.; Kim, T. S.; Liu,

L.; Meagher, K.; Norman, M. H.; Panter, K.; Schenkel, L. B.; Smith, A. L.;

Tamayo, N. A.; Whittington, D. A.; Xi, N.; Yang, K. Preparation of

2-aminobenzothiazole derivatives as phosphoinositide 3-kinase (PI3

kinase) modulators. PCT Int. Appl. WO2009017822A2, 2009.

(8) Stec, M. M.; Tamayo, N.; Booker, S. K.; Liao, L.; Yang, K.;

Caenepeel, S.; Zhang, N.; Jiang, J.; Subramanian, R.; Andrews, K.;

Hughes, P.; Norman, M. H. Structure-activity relationships of PI3KR

inhibitors: alternatives to a benzothiazole core and their eﬀect on

metabolism. Unpublished results.

(9) Booker, S.; Kim, T. S.; Liao, H.; Liu, L.; Norman, M. H.; Peterson,

E. A.;Stec, M.; Tamayo, N. A. Benzoimidazole derivatives as PI3 kinase

inhibitors and their preparation, pharmaceutical compositions and use in

the treatment of cancers. PCT Int. Appl. WO2009085230A1, 2009.

(10) Bo, Y. Y.; Liu, L.; Nishimura, N.; Norman, M. H.; Siegmund,

A. C.; Tamayo, N. A.; Yang, K. Inhibitors of PI3 kinase. PCT Int. Appl.

WO2010108074A2, 2010.

’ ASSOCIATED CONTENT

S

Supporting Information. Biological assays, in vivo study

b

protocols, and X-ray crystallographic data. This material is

available free of charge via the Internet at http://pubs.acs.org.

Accession Codes

^£The crystal structure of PI3K-gamma in complex with com-

pound 16d has been deposited in the Protein Data Bank with

accession code 3S2A.

’ AUTHOR INFORMATION

Corresponding Author

*Nobuko Nishimura: Tel: 805-447-0339. Fax: 805-480-3016.

E-mail: nobukon@amgen.com.

(11) Knight, Z. A.; Gonzalez, B.; Feldman, M. E.; Zunder, E. R.;

Goldenberg, D. D.; Williams, O.; Loewith, R.; Stokoe, D.; Balla, A.;

Toth, B.; Balla, T.; Weiss, W. A.; Williams, R. L.; Shokat, K. M. A

Pharmacological map of the PI3-K family deﬁnes a role of p110R in

insulin signaling. Cell 2006, 125, 733–747.

’ ACKNOWLEDGMENT

The authors thank Randy Hungate, Terry Rosen, Rick

Kendall, and Glenn Begley for their support of this research

program. Thanks also go to Paul Andrews for his assistance with

the in vitro cellular assays and Paul Schnier for his high resolution

mass spectroscopy analysis. In addition, we are grateful to Jin

Tang and Peter Yakowec for expression and puriﬁcation of the

PI3Kγ enzyme.

(12) The importance of CH-carbonyl hydrogen bond has been estab-

lished in protein-ligand binding for kinases. SeePierce, A. C.; Sandretto,

K. L.; Bemis, G. W. Kinase Inhibitors and the Case for CH O Hydrogen

3 3 3

Bonds in Protein-Ligand Binding. Proteins 2002, 49, 567–576.

(13) Suzuki, A. Recent advances in the cross-coupling reactions of

organoboron derivatives with organic electrophiles, 1995ꢀ1998.

J. Organomet. Chem. 1999, 576, 147-168, and references therein.

(14) Dhanak, D.; Knight, S. D. Preparation of thiazolones for use as

PI3 kinase inhibitors. WO2007103756, Sep 13, 2007.

’ ABBREVIATIONS

(15) Sun, H.; DiMagno, S. G. Room-temperature nucleophilic

aromatic ﬂuorination: experimental and theoretical studies. Angew.

Chem. Internat. Edit. 2006, 45, 2720–2725.

(16) Albert, A. Hydration of CdN bonds in heteroaromatic sub-

stances. Angew. Chem. Internat. Edit. 1967, 6, 919-928, and the reference

therein.

(17) Cho, M. J.; Pitman, I. H. Linear free energy relationships govern-

ing the covalent addition of nucleophilic reagents to a nitrogen-containing

heteroaromatic molecule. J. Am. Chem. Soc. 1974, 96, 1843–1849.

Ac, acetyl; BID, twice daily; CL, clearance; DCM, dichloromethane;

DMF, N,N-dimethylformamide; DMSO, dimethylsulfoxide; DNA-

PK, DNA-dependent serine/threonine protein kinase; dppf, 1,

1⁰-bis(diphenylphosphino)ferrocene; EC, eﬀective concentration;

ED, eﬀective dose; EtOAc, ethyl acetate; HGF, hepatocyte growth

factor; HPMC, hydroxypropyl methylcellulose; IV, intravenous;

MRT, mean residence time; mTOR, mammalian target of rapamy-

cin; mTORC1, mTOR containing complex 1; MeOH, methanol;

4750

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Journal of Medicinal Chemistry

ARTICLE

(18) During the course of this study, a structurally similar PI3K/

mTOR dual inhibitor, GSK2126458, was disclosed: Knight, S. D.; et al.

Discovery of GSK2126458, a Highly Potent Inhibitor of PI3K and the

Mammalian Target of Rapamycin. ACS Med. Chem. Lett. 2010, 1, 39–43.

(19) 16d, 16k and 18d were also tested in the Ambit kinase panel at

1 μM. (see Supporting Information).

(20) Experimental details of the PD study can be found in the

Supporting Information.

(21) Experimental details of the xenograft study can be found in the

Supporting Information.

(22) The doses used in the xenograft study were selected based on

the pharmacokinetic and pharmacodynamic proﬁles of compound 16d

and were estimated to give g50% target coverage for up to 24 h.

4751

dx.doi.org/10.1021/jm200386s |J. Med. Chem. 2011, 54, 4735–4751

Article Doi

Phospshoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) dual inhibitors: Discovery and structure-activity relationships of a series of quinoline and quinoxaline derivatives

DOI: 10.1021/jm200386s

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm200386s

Full text of DOI:10.1021/jm200386s

Products guided by the article

R&D Labs maybe for 1112982-79-2

Relevant to this article

Hot Product