Discovery of novel quinolinone adenosine A2B antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.10.030

A novel series of quinolinone-based adenosine A_2B receptor antagonists was identified via high throughput screening of an encoded combinatorial compound collection. Synthesis and assay of a series of analogs highlighted essential structural features of the initial hit. Optimization resulted in an A_2B antagonist (2i) which exhibited potent activity in a cAMP accumulation assay (5.1 nM) and an IL-8 release assay (0.4 nM).

Full text of DOI:10.1016/j.bmcl.2010.10.030

Bioorganic & Medicinal Chemistry Letters 20 (2010) 7414–7420
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel quinolinone adenosine A2B antagonists
Brian F. McGuinness a, , Koc-Kan Ho a, Tara M. Stauffer a, Laura L. Rokosz a, Neelima Mannava a,
⇑
Steven G. Kultgen a, Kurt Saionz a, Anthony Klon a, Weiqing Chen a, Hema Desai a, W. Lynn Rogers a,
Maria Webb a, Juxing Yin b, Yan Jiang b, Tailong Li b, Hao Yan b, Konghua Jing b, Shengting Zhang b,
Kanak Kanti Majumdar c, Vikash Srivastava c, Samiran Saha c
a Ligand Pharmaceuticals, 3000 Eastpark Boulevard Cranbury, NJ 08512, USA
b WuXi AppTec Co., Ltd, 288 Fute Zhong Road, Shanghai 200131, China
c TCG Lifesciences Ltd (Chembiotek), Block BN, Plot-7, Sec-V, Salt Lake City, Kolkata 700 091, India
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel series of quinolinone-based adenosine A2B receptor antagonists was identified via high through-
put screening of an encoded combinatorial compound collection. Synthesis and assay of a series of ana-
logs highlighted essential structural features of the initial hit. Optimization resulted in an A2B antagonist
(2i) which exhibited potent activity in a cAMP accumulation assay (5.1 nM) and an IL-8 release assay
(0.4 nM).
Received 4 September 2010
Revised 5 October 2010
Accepted 6 October 2010
Available online 4 November 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
A2B receptor
Adenosine
Antagonist
Combinatorial chemistry
Quinolinone
O
N
O
N
Extracellular adenosine provides regulatory signals through
interaction with a family of G-protein coupled adenosine receptors
(subtypes A1, A2A, A2B, and A3).1 In lung tissue, for example, inter-
action of extracellular adenosine with the A2B receptor has been
shown to cause the release of pro-inflammatory cytokines.2 Hence,
antagonism of the A2B adenosine receptor subtype has been pro-
posed as a treatment for respiratory disease.3 A2B antagonists have
also been suggested as treatments for diabetes,4 diabetic retinopa-
thy,5 colitis,6 and cancer.7 Efforts to develop selective, small mole-
cule A2B antagonists have originally stemmed from xanthine-based
lead structures such as MRS-1754.8 Optimization of this lead led to
CVT-6883, a selective A2B antagonist which entered human clinical
trials.9 Non-xanthine-based A2B antagonists have also been discov-
ered including 2-aminopyrimidines,10 N-(5,6-diarylpyridin-2-yl)
amides,11 2-aminobenzothiazoles,12 and aminothiazoles.13
N
N
N
N
N
N
O
O
N
H
N
H
O
O
F3C
N
NH
MRS-1754
CVT-6883
F
O
O
O
Cl
Cl
O
S
O
O
N
H
N
H
O
F
N
O
N
O
H
1a
1b
Two hit molecules, 1a and 1b, were initially identified as A2B
We have previously described the discovery of various A2A recep-
tor antagonists via the high throughput screening of a large, encoded
combinatorial compound collection.14–16 Herein, we detail the opti-
mization of a novel series of A2B antagonists based on quinolinone
hits also discovered from this encoded combinatorial collection.17
antagonists. While both were active in a hA2B cAMP accumulation
assay (Table 1), neither was stable in a human liver microsome as-
say (0% remaining after 0.5 h treatment with human liver micro-
somes).18 In addition, while these hits proved highly selective
against the A2A receptor subtype, they exhibited only minimal A1
receptor selectivity. Hence, a series of analogs aimed at increasing
both human liver microsome (HLM) stability and A1 receptor selec-
tivity were synthesized.
⇑
Corresponding author. Tel.: +1 609 570 1095 x1527; fax: +1 609 570 1050.
E-mail address: bmcguinness@venenumbiodesign.com (B.F. McGuinness).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.10.030

B. F. McGuinness et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7414–7420
7415
Table 1
SAR of quinolinone analogs
R1
R3
N
R2
O
Compd R1
R2
H
R3
CHO-hA2B cAMP Human A2B binding Human A2A binding A1 binding av
HLM stability (%
Ki (nM)
av Ki (nM)
av Ki (nM)
Ki (nM)
remaining-0.5 h)13
O
O
S
Cl
O
N
1a
315 102
211 36
>10,000
811
6
0
O
H
Cl
O
F
F
O
O
1b
Me
H
633 83
>5000
677 44
>10,000
>10,000
>10,000
3623 1536
6050 3175
0
O
O
N
H
F
F
O
1c
ND
N
H
O
O
S
Cl
Cl
O
N
1d
Me
>5000
>7000
ND
>10,000
ND
O
H
O
O
S
O
1e
1f
H
H
>5000
ND
>10,000
>10,000
10,443 5032
159 34
ND
0
O
N
H
O
O
S
O
O
352 227
70 19
N
H
O
O
F
O
1g
Me
>5000
3655 1332
>10,000
7027 616
ND
N
H
F
O
S
1h
1i
H
H
H
3474 739
1675 671
>5000
11,310 470
>10,000
>10,000
>10,000
ND
>10,000
28
67
ND
MeO
O
O
S
O
O
1334 309
>10,000
N
O
H
O
O
S
O
1j
>10,000
N
O
O
S
H
O
Cl
Cl
O
N
1k
H
1017 37
710 44
>10,000
602 25
0
O
The synthetic route used to generate the majority of the analogs
in this study is outlined in Scheme 1. The hydroxyl moiety of
4-aminophenol was protected as the t-butyldimethylsilyl ether fol-
lowed by protection of the amino group as the t-butoxycarbamate.
The dianion of this protected intermediate was reacted with tribu-
tyltin chloride to generate stannane 4 which was employed in a
Stille coupling with the appropriate acid chloride to yield ketone 5.
Deprotection of the phenol and alkylation with methyl bromoace-
tate produced 6 which was further deprotected by TFA treatment
to yield aniline 7.
Various methods were employed at this stage to generate the
quinolinone ring (8) depending on the targeted R3 group. For the
methanesulfonyl and 2,4-difluorophenyl R3 analogs, a two step
procedure of amide formation followed by base-mediated cycliza-
tion was required. For the 2-thiazolyl R3 analogs, however, cycliza-
tion occurred spontaneously with warming during the EDC-
mediated amide formation conditions. The required 2-thiazolyl-
acetic acid was prepared by the homologation route described in
Scheme 2. Although phosphorus oxychloride (employed for the
synthesis of analogs 2n–2q) was also effective in producing a cy-
clized structure, the 2-chloroquinoline was isolated. Subsequent
treatment with acetic acid under microwave heating was em-
ployed to produce the desired quinolinone 8.
If desired, methylation of the R2 position was completed at this
stage (8) via methyl iodide treatment in the presence of potassium
carbonate. Ultimately, base-mediated deprotection of the methyl
ester followed by amide bond formation yielded the completed
analogs.
A similar route was followed to generate the aminomethyl quin-
olinones 2d–2f (Scheme 3). 4-Chloroaniline was Boc-protected and

7416
B. F. McGuinness et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7414–7420
O
O
R4
R4
a, b, c
d
e, f
HO
TBDMSO
SnBu3
O
O
TBDMSO
MeO
O
O
NH2
NHBoc
NHBoc
NHBoc
4
5
6
O
R4
O
R4
R4
R3
R3
g
h, i, or j
k, l, n
O
O
MeO
R1HN
MeO
O
1a-1k
NH2
N
H
N
O
O
R2
2a-2c
2g-2q
3a-3f
7
8
Scheme 1. Synthetic route for quinolinone-6-oxyacetamide analogs. Reagents and conditions: (a) TBDMSCl, imidazole, THF, rt; (b) Boc anhydride, DCM, rt; (c) n-BuLi
(1 equiv), t-BuLi (1 equiv), (Bu)3SnCl, THF, À78 °C; (d) R4COCl, [Pd(CH3CN)2]Cl2, toluene, 100 °C; (e) TBAF, THF, rt; (f) methyl bromoacetate, K2CO3, MeCN, rt; (g) TFA; DCM, rt;
(h) for R3 = methanesulfonyl and 2,4-difluorobenzyl: (i) R3CH2COOH, EDC, HOBt, DIEA, DMF, rt; (ii) NaOMe, MeOH, 65 °C; (i) for R3 = thiazolyl: 2-(thiazol-2-yl)acetic acid (13),
EDC, HOBt, DMF, 55 °C; (j) For analogs 2n–2q: (i) R3CH2COOH, POCl3 50 °C; (ii) acetic acid, mw, 140 °C, 80 min; (k) (only for analogs 1b, 1d, and 1e) MeI, K2CO3, MeCN, rt; (l)
LiOH, THF, H2O, rt; (n) R1NH2, EDC, HOBt, DIEA, THF, rt.
a
S
N
c
O
H
b
d
S
N
S
N
S
N
S
N
OH
COOH
CN
Cl
9
10
11
12
13
Scheme 2. Synthesis of 2-(thiazol-2-yl)acetic acid (13). Reagents and conditions: (a) NaBH4, MeOH, rt; (b) Ph3P, CCl4, 60 °C; (c) TMSCN, TBAF, DCM, rt; (d) 9% aq KOH, 100 °C.
S
Cl
Cl
SnBu3
Cl
Cl
a, b
c
d, e
N
O
NH2
NHBoc
NHBoc
N
H
O
14
15
16
S
S
g
f
h or i
2d - 2f
NC
N
N
H2N
N
H
O
N
H
O
17
18
Scheme 3. Synthesis route for analogs 2d–2f. Reagents and conditions: (a) Boc anhydride, THF, rt; (b) n-BuLi (1 equiv) t-BuLi (1 equiv), (Bu)3SnCl, THF, À78 °C; (c)
hydrocinnamoyl chloride, [Pd(CH3CN)2]Cl2, toluene, 100 °C; (d) TFA; DCM, rt; (e) 2-(thiazol-2-yl)acetic acid, EDC, HOBt, DMF, 55 °C; (f) Zn(CN)2, Pd(TFA)2, (binaphthyl)P(t-
Bu)2, DMA, mw, 180 °C 1 h; (g) H2, Raney Ni, THF, NH3, MeOH, H2O, rt; (h) for analog 2d: hydrocinnamoyl chloride, Et3N, DCM, rt; (i) benzylamine (for 2e) or cumylamine (for
2f), triphosgene, DIEA, DCM, rt.
converted to the tributyltin derivative (14) as outlined above. Stille
coupling with hydrocinnamoyl chloride produced intermediate 15
which could be cyclized to the quinolinone (16) as before. Palla-
dium-mediated conversion to the cyanoquinolinone (17) followed
by reduction with Raney nickel generated the aminomethyl deriva-
tive 18 which could be converted to the desired urea or amide
derivatives by standard protocols.
N-1 methyl derivative of methylsulfone-containing 1a) lacked po-
tency. Pairing of the 3-methoxypropylamide R1 tail of 1a with the
3-(methylsulfonyl)-quinolinone scaffold of analog 1b also led to
inactivity (1e). Finally, while the presence of a simplified R1
benzylamine was allowed on the 3-(methylsulfonyl)-quinolinone
scaffold (1f), matching of the R1 benzylamine with the 3-(2,4-
difluorophenyl)quinolinone scaffold was not tolerated (1g). Hence,
analogs of 1a and 1b exhibited distinct SAR. Because the c log P of
analog 1f (2.0) was significantly lower than that of analog 1b (4.9),
efforts were focused on optimizing the 3-methylsulfone series.24
The presence of an oxyacetamide in the initial hit structures 1a
and 1b was reminiscent of the 8-phenyl xanthine analog MRS-
1754. It was clear that this portion of the molecule was important
since the truncated methoxy analog 1h was only weakly active.
However, the N-phenyloxyacetamide analog 1i exhibited a loss in
potency, indicating the quinolinone series did not follow the
The synthesized analogs19 were tested for A2B receptor activity
in a CHO-based cAMP accumulation assay,20 an A2B binding as-
say,21 and ultimately a HMC-1 cell IL-8 release assay.22 Compounds
were also tested in A2A and A1 binding assays to access subtype
selectivity.23
A divergent SAR pattern was observed between analogs of the
two initial hit structures 1a and 1b (Table 1). Analog 1c (the
R2-desmethyl analog of 2,4-difluorophenyl-containing 1b) was
not active in the A2B cAMP assay. Additionally, analog 1d (the

B. F. McGuinness et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7414–7420
7417
Table 2
SAR of quinolinone analogs
R1
R3
N
O
H
O
Cl
O
N
N
2a
667 47
330 29
338 187
408 171
>10,000
1085 516
0
H
Cl
O
O
O
2b
2c
2d
2e
2f
1124 440
37 9.5
291 74
19 4.3
ND
407 190
61 18
ND
>10,000
3509 83
>10,000
>10,000
>10,000
2439 184
>10,000
>10,000
436 105
101 57
0
N
H
O
O
O
O
O
O
O
O
S
N
O
21
3
2
N
H
S
N
1926 1621
68 29
3253 752
3
N
H
S
N
ND
68
2
4
N
H
N
H
S
N
17 3.3
ND
68 37
15 12
40
3
N
H
N
H
S
N
O
O
O
O
2g
2h
2i
(S)
(R)
15
8
28 0.7
486 113
28 8.4
1.7 0.4
60 32
N
H
S
N
320 44
5.1 1.1
170 70
19 5.3
ND
53
N
H
S
N
1.3
0
N
H
S
N
2j
4.5 0.8
12 6.2
>10,000
20 14
1.0 0.4
ND
>10,000
ND
13 0.5
1720 178
8.6 2.0
46
ND
22
N
H
O
S
O
O
2k
2l
3413 424
7.4 0.5
N
H
N
S
O
0.4 0.01
1460 572
N
H
N
HO
S
N
O
O
O
O
O
O
2m
2.8 0.5
6.7 2.7
0.44 0.01
1925 807
6.1 0.1
27
N
H
2n
2o
504 14
187 2.6
ND
ND
ND
ND
>10,000
>10,000
1370 219
497 227
ND
83
N
N
H
N
N
H
(continued on next page)

7418
B. F. McGuinness et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7414–7420
Table 2 (continued)
Compd R1
R3
CHO-hA2B
cAMP Ki (nM)
hA2B binding HMC-1 IL-8
Human A2A
binding Ki (nM)
A1 binding
Ki (nM)
HLM stability (%
remaining-0.5 h)
Ki (nM)
release Ki (nM)
O
O
N
O
O
2p
2q
884 12
31 2.0
ND
ND
>10,000
377 29
ND
45
N
H
O
N
ND
ND
2832 27
12
1
N
H
known SAR of the 8-phenyl xanthine analogs (such as MRS-1754).
Interestingly, analog 1i showed an increased stability in the HLM
assay (64% remaining after 0.5 h) suggesting the N-benzyl amide
was a metabolic liability. However, the importance of the N-benzyl
group for potency was highlighted by the lack of activity in the N-
cyclopropylmethyloxyacetamide 1j. Additionally, a threefold drop
in activity was noted in the N-methyl amide derivative 1k. Overall,
this initial study of the R1 SAR did not result in an analog with
higher A2B binding affinity or increased activity in the A2B cAMP
functional assay. Furthermore, as is obvious from the data in Table
1, none of these derivatives addressed the issue of A1 selectivity.
Attention was turned to the R3 portion of the molecule (Table
2). While both a cyano group (2a) and furanyl group (2b) were tol-
erated at this position, inclusion of a 2-thiazolyl group at the R3 po-
sition (2c) increased potency in the cAMP assay 17-fold (when
compared to 1f). Analog 2c maintained good selectivity over the
A2A receptor (95-fold) but was still only threefold selective against
the A1 receptor. The human microsome stability of 2c was also still
poor.
Further investigation of the pendant 6-position side chain (R1)
was initiated while holding the more potent 3-(thiazol-2-yl)quin-
olinone scaffold constant. Derivatives of the 6-aminomethyl
Table 3
SAR of R4 analogs
R4
R1
R3
N
O
H
Compd
R1
R4
R3
CHO-hA2B
cAMP Ki
(nM)
Human A2B
binding Ki
(mM)
HMC-1 IL-8
release Ki
(lM)
Human A2A
binding Ki
(mM)
A1 binding
Ki (nM)
HLM stability
(% remaining-
0.5 h)
O
O
O
O
S
Cl
Cl
Cl
Cl
Cl
Cl
O
O
O
N
H
1a
3a
3b
315 102
6670 40
211 36
5251 1389
ND
172 106
1933 694
ND
>10,000
>10,000
>10,000
811
6
0
O
O
S
Me
N
H
3914 206
37
0
O
O
S
N
H
>10,000
5.1 1.1
8603 1757
O
O
O
O
O
O
S
N
O
2i
28 8.4
ND
1.3
ND
ND
0
>10,000
1080 524
878 131
>10,000
>10,000
19 5.4
5999 1819
1129 984
183 101
53
ND
ND
12
13
N
H
S
N
O
O
O
O
3c
3d
3e
3f
2669 550
5155 682
81 6.7
N
H
S
N
CF3
ND
N
H
S
N
O
70 21
105 78
71 37
10 0.2
N
H
S
N
41 0.2
1153 724
N
H
F

B. F. McGuinness et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7414–7420
7419
5. Grant, M. B.; Davis, M. I.; Caballero, S.; Feoktistov, I.; Biaggioni, I.; Belardinelli, L.
Invest. Ophthalmol. Vis. Sci. 2001, 42, 2068.
6. Eltzschig, H. K.; Jesus, R.-N.; Colgan, S. P. Expert Opin. Ther. Targets 2009, 13,
1267.
7. Fishman, P.; Bar-Yehuda, S.; Synowitz, M.; Powell, J. D.; Klotz, K. N.; Gessi, S.;
Borea, P. A. Handbook of Experimental Pharmacology 2009, 193, 399.
8. Kim, Y.-C.; Ji, X.; Melman, N.; Linden, J.; Jacobson, K. A. J. Med. Chem. 2000, 43,
1165.
9. Elzein, E.; Kalla, R. V.; Li, X.; Perry, T.; Gimbel, A.; Zeng, D.; Lustig, D.; Leung, K.;
Zablocki, J. J. Med. Chem. 2008, 51, 2267.
10. Vidal, B.; Nueda, A.; Esteve, C.; Teresa Domenech, T.; Benito, S.; Reinoso, R. F.;
Pont, M.; Calbet, M.; López, R.; Cadavid, M. I.; Loza, M. I.; Cárdenas, Á.;
Godessart, N.; Beleta, J.; Warrellow, G.; Ryder, H. J. Med. Chem. 2007, 50, 2732.
11. Eastwood, P.; Gonzalez, J.; Paredes, S.; Nueda, A.; Domenech, T.; Alberti, J.;
Vidal, B. Bioorg. Med. Chem. Lett. 2010, 20, 1697.
12. Cheung, A. W.-H.; Brinkman, J.; Firooznia, F.; Flohr, A.; Grimsby, J.; Gubler, M.
L.; Guertin, K.; Hamid, R.; Marcopulos, N.; Norcross, R. D.; Qi, L.; Ramsey, G.;
Tan, J.; Wen, Y.; Sarabu, R. Bioorg. Med. Chem. Lett. 2010, 20, 4140.
13. Press, N. J.; Taylor, R. J.; Fullerton, J. D.; Tranter, P.; McCarthy, C.; Keller, T. H.;
Brown, L.; Cheung, R.; Christie, J.; Haberthuer, S.; Hatto, J. D. I.; Keenan, M.;
Mercer, M. K.; Press, N. E.; Sahri, H.; Tuffnell, A. R.; Tweed, M.; Fozard, J. R.
Bioorg. Med. Chem. 2005, 15, 3081.
intermediate 18 (Scheme 3) were employed to test the require-
ment of the oxyacetamide group. While the 3-phenylpropanoyl
derivative 2d was only weakly active, the N-benzylurea analog
2e was almost as potent as 1f. This boost in potency when compar-
ing analogs 2d and 2e suggests the nitrogen of the N-benzyl group
may be acting as a hydrogen bond donor with the A2B receptor.
(This hypothesis is supported by the drop in potency exhibited
by the N-methylated analog 1k noted earlier.) An increase in the
HLM stability of this urea series was achieved by the dimethylated
benzyl analog 2f (40% remaining after 0.5 h incubation).
The effect of substitution on the R1 benzylic methylene of the
lead structure 2c was similarly investigated. The (S)-methyl-
substituted analog (2g) maintained activity equal to the unsubsti-
tuted 2c but the (R)-methyl-substituted analog 2h was 15-fold less
potent. Gratifyingly, the dimethyl-substituted derivative 2i exhib-
ited increased potency (5.1 nM) in the A2B cAMP assay. As with
analog 2f, a significant increase in the microsomal stability was ob-
served (53% remaining after 0.5 h incubation) for 2i. The cyclopro-
pyl analog 2j was also a potent antagonist of the A2B receptor with
increased microsome stability, further implicating the benzylic
methylene as a metabolic liability. The aromatic benzene ring
proved again to be essential for the potency of these analogs since
t-butyl amide 2k was inactive. Both the ethyl- and hydroxyethyl-
substituted analogs (2l and 2m) were highly active in the A2B re-
lated assays indicating that the boost in activity was not solely re-
lated to an increase in hydrophobicity in this area of the lead
molecule.
14. Cole, A. G.; Stauffer, T. M.; Rokosz, L. L.; Metzger, A.; Dillard, L. W.; Zeng, W.;
Henderson, I. Bioorg. Med. Chem. Lett. 2009, 19, 378.
15. Shao, Y.; Cole, A. G.; Bresia, M.-R.; Qin, L.-Q.; Duo, J.; Stauffer, T. M.; Rokosz, L.
L.; McGuinness, B. F.; Henderson, I. Bioorg. Med. Chem. Lett. 2009, 19, 1399.
16. McGuinness, B. F.; Cole, A. G.; Dong, G.; Brescia, M.-R.; Shao, Y.; Henderson, I.;
Rokosz, L. L.; Stauffer, T. M.; Mannava, N.; Kimble, E. F.; Hicks, C.; White, N.;
Wines, P. G.; Quadros, E. Bioorg. Med. Chem. Lett. 2010, 20, 6845.
17. (a) Ohlmeyer, M. H. J.; Swanson, R. N.; Dillard, L.; Reader, J. C.; Asouline, G.;
Kobayishi, R.; Wigler, M.; Still, W. C. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 10922;
(b) Burbaum, J. J.; Ohlmeyer, M. H. J.; Reader, J. C.; Henderson, I.; Dillard, L. W.;
Li, G.; Randle, T. L.; Sigal, N. H.; Chelsky, D.; Baldwin, J. J. Proc. Natl. Acad. Sci.
1995, 92, 6027.
18. For details of the human liver microsome assay see: Merritt, J. R.; Liu, J.;
Quadros, E.; Morris, M. L.; Liu, R.; Zhang, R.; Jacob, B.; Postelnek, J.; Hicks, C. M.;
Chen, W.; Kimble, E. F.; Rogers, W. L.; O’Brien, L.; White, N.; Desai, H.; Bansal,
S.; King, G.; Ohlmeyer, M. J.; Appell, K. C.; Webb, M. L. J. Med. Chem. 2009, 52,
1295.
The ability of these derivatives to suppress NECA-induced IL-8
release from HMC-1 cells was also examined (Table 2). This assay
confirmed the results from the initial cAMP accumulation assay.
Notably, analogs 2l and 2m inhibited the release of IL-8 at subn-
anomolar concentrations in this assay.
19. Representative spectral data: For 2i: 1H NMR DMSO-d6, 300 MHz: d 8.23 (s, 1H),
8.13 (d, 1H), 7.92 (d, 1H), 7.39 (m, 5H), 7.29 (m, 4H), 7.20 (m, 3H), 4.69 (s, 2H),
3.55 (m, 2H), 2.95 (m, 2H), 1.56 (s, 6H). MS (ES-API): MH+ = 524.2.
20. hA2B cAMP accumulation assay: Chinese Hamster Ovary (CHO) cells
(GlaxoSmithKline) stably transfected with the human A2B receptor were
maintained in 1Â DMEM/F12 medium (Mediatech) supplemented with 10%
FBS (Hyclone), 1% glutamax (Invitrogen), 0.1 U/ml adenosine deaminase
With the R1 position optimized, attention returned to the R3
heterocycle. The pyridine derivatives 2n–2p were significantly less
active than the benchmark compound 2i. The 3-methyllisoxazole
derivative, however, 2q exhibited reasonable potency. Hence, it ap-
pears a five-membered heterocycle may be favored in this position.
The role of the 4-phenethyl group (R4) was also investigated
(Table 3). Replacement of the phenethyl group with a simple
methyl group caused a >20-fold drop in A2B binding affinity and
cAMP activity (compare 1a with 3a). In addition, a 4-phenyl deriv-
ative (3b) was inactive. Furthermore, analogs bearing either a iso-
butyl- or 3,3,3-trifluoropropyl- group in the 4 position of the
quinolinone (3c or 3d) were significantly less potent then 2i.
Hence, the aromatic ring of the 3-phenethylgroup was an impor-
tant contributor to the interaction of 2i with the A2B receptor.
A phenoxymethyl moiety (3e) could replace the phenethyl
group albeit with a 15-fold drop in potency. para-Fluoro substitu-
tion of the phenethyl benzene (3f) also led to a moderate drop in
A2B activity but caused a modest increase in A1 selectivity.
In summary, a series of 4-phenethylquinolinone A2B antagonists
was discovered. Optimization of the high throughput screening
hits generated a lead molecule 2i which was a potent (<10 nM)
A2B antagonist in in vitro functional assays. Significant improve-
ments were made in the human liver microsome stability of these
analogs. Future efforts will need to address the A1 selectivity of this
novel series.
(Roche Applied Science), 1% Penn/Strep (Invitrogen), 200
lg/ml G418
(Mediatech) and 200 g/ml Hygromycin B (Mediatech). cAMP was quantified
l
using the Perkin–Elmer LANCE™ cAMP 384 kit. On the day of the assay, 2500
CHO-hA2B cells were pre-incubated with compound for 15 min and then
stimulated with 75 nM NECA for 15 min in a final volume of 18
conducted in 1Â HBSS (Invitrogen), 5 mM HEPES (Invitrogen), 0.05% BSA,
(Sigma), 100 M papaverine–HCl (Sigma), pH 7.4 buffer at 37 °C/5% CO2. Final
lL. Assay was
l
DMSO concentration was 0.5%. Following stimulation, the cells were lysed with
cAMP detection solution provided with the assay kit. Signal was allowed to
develop for 1 h and detected using the EnVision™ Multiplate Reader or the
Victor™ Multilabel counter (Perkin–Elmer Life and Analytical Sciences). Assays
were performed in duplicate and compounds were tested a minimum of two
times. The data were fit to a one-site competition binding model for IC50
determination using the program GraphPad Prism (GraphPad Software, Inc.,
San Diego, CA) and Ki values were calculated using the Cheng–Prusoff equation.
21. hA2B binding assay: membranes (20
express human A2B (GlaxoSmithKline) were mixed with 30 nM (final) [3H]
ZM241385 in 50 l assay buffer (50 mM Tris–HCl, pH 7.4, 100 mM NaCl,
lg) prepared from CHO cells that stably
l
10 mM MgCl2, 1 mM EDTA, pH 6.5, 0.025 mg/ml adenosine deaminase (Roche
Applied Science) containing 4% DMSO with or without test compounds.
Reactions were carried out for 60 min at room temperature and were
terminated by rapid filtration over GF/B filters (Millipore). Filters were
washed four times with 1 ml cold 50 mM Tris–HCl, 100 mM NaCl, 10 mM
MgCl2, and radioactivity retained on filters was counted in a Microbeta Trilux
microplate scintillation counter (Perkin–Elmer). Nonspecific binding was
determined in the presence of 30 nM ZM241385. Assays were performed in
duplicate and compounds were tested two times. Data were fit to a one-site
competition binding model for IC50 determination using the program
GraphPad Prism (GraphPad Software, Inc) and Ki values were calculated
using the Cheng–Prusoff equation.
22. HMC-1 IL-8 release assay: HMC-1 human mast cells (Mayo Clinic Health
Solutions) were maintained in IMDM (with 2 mM glutamine, Invitrogen), 10%
iron-supplemented calf serum, 1% sodium bicarbonate (Invitrogen), 1%
References and notes
penicillin/streptomycin (Mediatech) and 1.2 mM
the day of the assay, compounds were prepared in growth medium containing
a-thioglycerol (Sigma). On
1. Jacobson, K. A. Handbook of Experimental Pharmacology 2009, 193, 1.
2. (a) Polosa, R. Eur. Respir. J. 2002, 20, 488; (b) Zhong, H.; Wu, Y.; Belardinelli, L.;
Zend, D. Am. J. Respir. Cell Mol. Biol. 2006, 35, 587.
3. (a) Holgate, S. T. Br. J. Pharmacol. 2005, 145, 1009; (b) Kalla, R.; Zablocki, J. Annu.
Rep. Med. Chem. 2009, 44, 265.
1 U/ml adenosine deaminase (Roche Applied Science) and 20 lM NECA
(Sigma). To the compounds, 2.5 Â 104 cells were added in 1:1 ratio to the
prepared compounds/NECA (10 lM final concentration) in a final volume of
250
lL. Final DMSO concentration was 0.5%. Cells and compounds were
4. Ruesing, D.; Mueller, C. E.; Verspohl, E. J. J. Pharm. Pharmacol. 2006, 58,
1639.
incubated for 6 h at 37 °C/5% CO2. Basal IL-8 release was determined in the

7420
B. F. McGuinness et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7414–7420
(GraphPad Software, Inc., San Diego, CA) and Ki values were calculated using
absence of NECA stimulation. Cell supernatant was collected and frozen at
À80 °C until time of ELISA assay, when frozen supernatants were thawed and
the Cheng–Prusoff equation.
IL-8 release was quantified using the QuantiGlo Human IL-8 Chemiluminescent
Immunoassay kit (cat no. Q8000B, R&D Systems). Chemiluminescent signal
was captured using the Microbeta Trilux microplate counter (Perkin–Elmer
Life and Analytical Sciences). Assays were performed in duplicate and
compounds were tested two times. Data were fit to a one-site competition
binding model for IC50 determination using the program GraphPad Prism
23. For details on the A2A and A1 binding assays see the footnotes of Ref. 14. The
values are reported as the mean of at least two determinations.
24. Like many adenosine receptor antagonists, these compounds generally
exhibited poor aqueous solubility. Hence, the series with a lower c log P was
favored.