Discovery and optimization of 2-phenyloxazole derivatives as diacylglycerol acyltransferase-1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.09.039

In a discovery effort to find safe and effective DGAT-1 inhibitors, we have identified 2-phenyloxazole 4-carboxamide 1 as a conformationally constrained analog of a hydrazide hit, which was previously identified from high-throughput screening. Further optimization of this series has led to chemically more stable 2-phenyloxazole-based DGAT-1 inhibitor 25 with improved solubility, cell-based activity, and pharmacokinetic properties. Compound 25 also demonstrated in vivo efficacy in a diet-induced obesity (DIO) rat model.

Full text of DOI:10.1016/j.bmcl.2011.09.039

Bioorganic & Medicinal Chemistry Letters 21 (2011) 7205–7209
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and optimization of 2-phenyloxazole derivatives as diacylglycerol
acyltransferase-1 inhibitors
Weiya Yun a, , Mushtaq Ahmad a, Yingsi Chen a, Paul Gillespie a, Karin Conde-Knape a, Sonja Kazmer a,
⇑
Shiming Li a, Yimin Qian a, Rebecca Taub b, Stanley J. Wertheimer a, Toni Whittard a, David Bolin a
a Roche Research Center, Hoffmann-La Roche, Inc., Nutley, NJ 07110, USA
b Via Pharmaceuticals, Inc., 101 College Road East, Princeton, NJ 08540, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
In a discovery effort to find safe and effective DGAT-1 inhibitors, we have identified 2-phenyloxazole 4-
carboxamide 1 as a conformationally constrained analog of a hydrazide hit, which was previously iden-
tified from high-throughput screening. Further optimization of this series has led to chemically more sta-
ble 2-phenyloxazole-based DGAT-1 inhibitor 25 with improved solubility, cell-based activity, and
pharmacokinetic properties. Compound 25 also demonstrated in vivo efficacy in a diet-induced obesity
(DIO) rat model.
Received 8 August 2011
Revised 7 September 2011
Accepted 12 September 2011
Available online 22 September 2011
Keywords:
DGAT-1 inhibitor
Obesity
Ó 2011 Elsevier Ltd. All rights reserved.
Phenyloxazole
F
F
Excessive accumulation of triacylglycerides is often associated
with diseases such as obesity, diabetes and steatohepatitis. Diacyl-
glycerol acyltransferase (DGAT) plays an essential role in obesity as
it catalyzes the rate-limiting step in triacylglyceride synthesis.1–3 It
is known that DGAT-1 deficient mice are lean, have increased en-
ergy expenditure, increased insulin sensitivity and are resistant
to diet-induced obesity.4 Consequently, DGAT-1 inhibition is ex-
pected to be a feasible therapeutic strategy for human obesity
and obesity-associated complications, such as type 2 diabetes.
In a continued discovery effort to find safe and effective DGAT-1
inhibitors, we have identified 2-phenyloxazole 4-carboxamide 15–7
as a conformationally constrained analog of a hydrazide hit
F
O
O
N
N
N
O
N
1
The first round of our SAR campaign involved replacing the indazo-
lone piece in compound 1 with a library of 210 structurally diverse
amines via amide coupling to 2-phenyl-5-(trifluoromethyl)oxazole-
4-carboxylic acid. Inhibitory activities of these compounds against
human DGAT-1 were measured at two concentrations, 80 and
(IC50 = 0.23
throughput screening.8 While 1 showed good initial activity
toward human DGAT-1 in both enzyme (IC50 = 0.06 M) and
DGAT-1/CHOK1 cell-based assays (EC50 = 2.5 M), overall physico-
lM), which was previously identified from high-
2 l
M, in a human DGAT-1 enzyme assay.6–8 Compounds with inhi-
l
l
bition percentage greater than 50% at 80 lM were further titrated
chemical property concerns related to its poor solubility prompted
us to continue to optimize this series. Herein we report our optimi-
zation of the series, which led to chemically more stable 2-phenyl-
oxazole-based DGAT-1 inhibitors with improved solubility,
cell-based activity, and pharmacokinetic properties, and led to a
compound that demonstrated in vivo efficacy in a diet-induced
obesity (DIO) rat model.
down to obtain IC50 values. Selected test results from the library
are highlighted in Table 1.
Interestingly, of the compounds within this library, significant
activity was observed only for aryl amines with a para substituent
(see compounds 4, 6, 8 and 10 in Table 1). Compared with their
corresponding ortho/meta-substituted counterparts, the para-
substituted compounds showed better activities. The IC50 value
of compound 10 even reached sub-micromolar activity. Results
from a follow up library further showed that compounds with ami-
no-substituents on the phenyl ring are favored over ether or alkyl
derivatives. These results encouraged us to further investigate the
⇑
Corresponding author. Tel.: +1 973 235 3126; fax: 1 973 235 6084.
E-mail address: weiya.yun@roche.com (W. Yun).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.039

7206
W. Yun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7205–7209
Table 1
substitution with morpholine or an aliphatic amine followed by
reduction of the nitro group afforded an arylamine, which was then
coupled with a biarylcarboxylic acid to give the final compounds
(Scheme 1a). The thiazole analog 11 was prepared in a similar fash-
ion as shown in Scheme 1b.
To study the SAR of the R group on the B ring, the oxazole ring
was synthesized in two different ways.6 Scheme 2 outlines the
preparation of analogs with a simple alkyl chain at position 5 of
the oxazole ring.9 Doubly protected glycine was first acylated on
Assay results of representative compounds from the initial librarya
F
F
F
O
NHR
N
O
a
Compound number
R
Inhibition% at 80
l
M
IC50 (lM)
the
a-carbon with an acid chloride under basic conditions. After
2
3
4
5
6
7
8
9
2-i-Pr-Ph
3-i-Pr-Ph
4-i-Pr-Ph
3-O-i-Pr-Ph
4-O-i-Pr-Ph
3-n-Pr-Ph
4-n-Pr-Ph
2-Morpholine-Ph
42.1
15.3
83.3
17.7
83.6
5.7
ND
ND
9.9
ND
7.4
ND
13.9
ND
0.33
deprotection, the amino group was acylated again with benzoyl
chloride. The oxazole ring was then formed via dehydration of
the 1,4-dicarbonyl compounds.
For R substituents other than alkyl, such as chlorine and bro-
mine, an alternative synthetic route was used as outlined in
Scheme 3. Starting from commercially available 2-chloro-oxa-
zole-4-carboxylic acid ethyl ester, Suzuki coupling with phenyl
boronic acid followed by chlorination with N-chlorosuccinimide
(NCS) or bromination with N-bromosuccinimide (NBS) provided
the desired oxazole intermediate.
67.6
5.3
10
4-Morpholine-Ph 82.1
ND: not determined.
a
Mean value of at least three experiments.
Scheme 4 outlines a representative synthesis of the regio-iso-
meric oxazole.6,10 2-Amino-propionic acid was first acylated with
benzoyl chloride. The amide 12 formed was subjected to oxalyl
chloride to yield the 2-phenyl-4-methyl-oxazole-5-carbonyl chlo-
ride 13, presumably via a lactonization–isonitrile rearrangement
sequence.10 Subsequent treatment of 13 with N-hydroxy-
succinimide provided an activated ester. Stirring the desired ani-
line with this activated ester at 85 °C overnight afforded the
target analogs.
SAR of this class. The following discussion focuses on analogs with
an amino substituent (especially morpholine) and also on modifi-
cation of the A and B rings, as shown in Fig. 1 below.
The majority of the compounds for our SAR study were synthe-
sized from para-substituted nitro-halo-benzenes or heteroaro-
matic analogs as shown in Scheme 1.6 Nucleophilic aromatic
SAR of representative examples of A ring modifications is
shown in Table 2. Compounds were assayed in a human DGAT-1
enzyme assay and in a cell-based assay of human DGAT-1 transfec-
ted into CHOK1 cells. The observed activities were expressed as
IC50 for enzyme assay and EC50 for cellular assay, respectively.8
As shown in Table 2, replacement of the central A ring of com-
pound 10 by a pyridine ring (14) improved the activity by 2-fold
in the human DGAT-1 enzyme assay (n > 3) and 3-fold in the
R
O
B
N
N
A
R2
O
N
R1
Figure 1. General structure of phenyloxazole based DGAT-1 inhibitors.
R
a
NH2
B
NO2
Ph
R
O
N
X
R2
R1
B
a, b
c
Ph
A
O
A
A
X
N
N
N
N
X
N
O
N
Hal
Hal: F or Cl; X: CH or N
R1
R2
b
CF3
O
N
S
O2N
d
S
N
a, b
S
N
N
O
Br
O
N
O
N
N
11
Scheme 1. General synthetic routes for A and B ring modifications. Reagents and conditions: (a) NHR1R2, Et3N, CH2Cl2, 23 °C, 3 h; (b) 10% Pd/C, H2, 50 psi, EtOAc, 23 °C, 3 h; (c)
1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC), 1-hydroxy-7-azabenzotriazole (HOAT), CH2Cl2, 23 °C, 18 h; (d) 2-phenyl-5-(trifluoromethyl)oxazole-
4-carboxylic acid, EDC, HOAT, CH2Cl2, 23 °C, 18 h.
O
N
R
R
O
N
R
O
O
Ph
Ph
Ph
Ph
c, d
a
O
b
N
CO2Bn
Ph
CO2Bn
CO2Bn
Ph
N
O
R = Et, n-Pr or i-Pr
Scheme 2. Synthesis of 5-alkyl substituted 2-phenyloxazole-4-carboxylic acid. Reagents and conditions: (a) LiN(Me3Si)2, RCOCl, THF, À78 °C to rt, 3 h; (b) (i) 2 M HCl, rt, 1 h;
(ii) PhCOCl, pyridine, CH2Cl2, rt, 1 h; (c) Ph3P, I2, Et3N, THF, À78 °C to rt, 1 h; (d) LiOH–H2O, THF/MeOH/H2O (3:1:1), rt, 1 h.

W. Yun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7205–7209
7207
X
X
O
O
O
O
c
a
b
Ph
Cl
Ph
CO2Et
Ph
CO2Et
CO2Et
N
N
N
N
CO2H
X = Cl or Br
Scheme 3. Synthesis of 5-halogen substituted 2-phenyloxazole-4-carboxylic acid. Reagents and conditions: (a) PhB(OH)2, Pd(Ph3P)4, dimethoxyethane/2 M aqueous Na2CO3
(5:1), 90 °C, overnight; (b) N-chlorosuccinimide (NCS) or N-bromosuccinimide (NBS), sulfuric acid, CCl4, 90 °C, overnight; (c) LiOH–H2O, THF/MeOH/H2O (3:1:1), rt, 1 h.
O
N
O
a
c
b
d
N
Ph
20
O
N
O
N
COOH
Ph
N
COOH
Ph
COCl
O
13
O
O
12
Scheme 4. Synthesis of regio-isomeric oxazole derivatives. Reagents and conditions: (a) Me3SiCl, PhCOCl, DIPEA, CH2Cl2, rt to reflux, 1.5 h; (b) Oxalyl chloride, THF, 50 °C,
overnight; (c) N-hydroxysuccinimide, Et3N, CH2Cl2, 50 °C, overnight; (d) 6-morpholin-4-yl-pyridin-3-ylamine, Acetonitrile, 85 °C, 18 h.
Table 3
Table 2
B ring modifications and assay resultsa
A ring modifications and assay resultsa
F
R
F
B
F
O
N
N
N
O
A
N
X
O
N
O
O
Compound number
B ring
R
X
IC50
(l
M)
EC50 (lM)
Compound number
A ring
IC50
(lM)
EC50
2.7
(lM)
14
16
17
CF3
Cl
Et
N
N
N
0.16
0.61
0.70
0.84
2.99
3.42
O
10
0.33
N
O
18
CF3
N
0.14
0.49
14
0.16
0.84
N
N
N
N
19
20
21
Me
Me
Me
N
0.66
0.15
43
11.8
7.46
NDb
15
11
0.38
0.1
1.5
3.7
S
N
S
N
O
N
N
a
Mean value of at least three experiments.
CH
N
a
Mean value of at least three experiments.
ND: not determined.
b
DGAT-1/CHOK1 cell-based assay (n > 5). Although it is moderate,
two to 4-fold improvement was observed in all similar analogs pre-
pared and tested. However, no significant improvements were ob-
served with other A ring modification in terms of cell-based
activity. Therefore, we retained this morpholine–pyridine–amine
motif during the optimization of the B ring and R group.
As shown in Table 3, when the oxazole ring is replaced with fur-
an (18), thiazole (19) or the regio-isomeric oxazole (20), no signif-
icant loss of potency in the human DGAT-1 enzyme assay was
observed. However, replacement of the oxazole with a six-mem-
bered pyrimidine ring (21) is detrimental to the compound’s activ-
ity. These results revealed that the five-membered oxazole ring
likely serves as a scaffold and that the hetero atoms in the ring
may not be involved in any hydrogen bonding interactions with
the enzyme. In contrast, the R groups on the oxazole are critical
for enzyme inhibition. Replacement of the trifluoromethyl group
by a slightly smaller substituent, such as chlorine (16) resulted in
a slight loss of activity. As the R group size increases from trifluo-
romethyl to ethyl, (17), activity decreases by 4-fold. Removal of the
trifluoromethyl or phenyl group from the oxazole results in
complete loss of activity (data not shown). Nevertheless, of the
compounds in Table 3, the parent compound 14 showed the best
cell-based activity among all other analogs except for the furan
analog 18. Considering the potential liabilities associated with
the furan ring, we retained the 2-phenyl-5-(trifluoromethyl)oxa-
zole ring for subsequent SAR studies.
In order to find better alternatives for the morpholine side
chain, several acyclic amines were systematically examined and
the results are given in Table 4. While diethyl amine analog 22 is
equally potent to the parent compound 14, compounds with an
NH group (23) or an OH group (24) reduced the cell-based activity
significantly, likely due to reduced cell permeability caused by the
extra H-bond donors associated with the NH or OH groups. On the
other hand, the unsymmetrical acyclic analog 25 showed 4-fold
improvement in the enzyme inhibition assay and slightly
improved cell-based activity (n > 9). Increasing the size of the
N-substituted group resulted in significant loss of activity (26).

7208
W. Yun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7205–7209
Table 4
Table 5
Optimization of morpholine motifa
Mean pharmacokinetic parameters of compound 25 following an intravenous dose of
5 mg/kg with a solution formulation or an oral dose of 50 mg/kg with an aqueous
suspension in male Wistar rats (n = 3)
CF3
O
O
IV dose (5 mg/kg)
Vd (L/kg)
AUC (ng h/mL)
CL (mL/min/kg)
T1/2 (h)
PO dose (50 mg/kg)
Cmax (ng/mL)
AUC (ng h/mL)
Tmax (h)
F%
N
N
1.6
10200
47158
1
6156
13.5
1.95
N
R
77
Compound number
R
IC50
(
l
M)
EC50 (lM)
N
N
14
0.160
0.84
As shown in Table 5, when dosed at 5 mg/kg iv and 50 mg/kg po
in rats, compound 25 exhibited low clearance and good oral bio-
availability. The maximum exposure level in plasma (Cmax) is
O
22
0.253
0.91
24 lM at the 50 mg/kg oral dose, about 36-fold higher than its
EC50 value. The exposure levels of 25 in target tissues, such as liver
and adipose, are at least 2-fold higher than that in plasma in rats.
These data suggested that compound 25 should be a good candi-
date for in vivo proof of concept efficacy studies in DIO rats.
Three-month-old male Sprague Dawley DIO rats (fed with a
high-fat diet for 5 weeks prior to the study and during the study)
were dosed with compound 25 orally and once daily for three
weeks. The treated groups demonstrated dose-dependent reduc-
tion of weight gain compared to the vehicle group. As shown in
Fig. 2a , a 4% reduction in cumulative body weight gain compared
to vehicle was observed for the 25 mg/kg dose group on day 21 (p
<0.0001) and 8% for the 50 mg/kg dose group (p <0.0001), respec-
tively. The dose-dependent efficacy correlates well with measured
O
O
23
24
N
H
0.407
0.554
11.06
8.09
N
N
N
O
O
25
26
0.038
0.653
0.66
8.90
O
O
27
0.699
>25
N
drug levels (Cmax of 2.6, 7.8 and 12.2 lM at doses of 5, 25 and
50 mg/kg, respectively, on day 21). The high dose (50 mg/kg) group
also showed significant reduction in plasma triglycerides (64%, p
<0.0009 in Fig. 2b) and an improved lipid profile (decreased plasma
free fatty acid and total cholesterol, data not shown). Dexa scan re-
sults also showed a trend of decrease in fat pad mass in the high
dose treated rats relative to the vehicle group, but no change in
lean muscle. These findings are consistent with the observations
from DGAT-1 knock out mouse studies.4 Compound 25 was well
tolerated in rats. After 22 days of treatment, no change in organ
weight was observed in liver, heart, brain and kidney; and liver en-
zymes were not affected. This compound is also clean in a number
of in vitro toxicity tests such as MNT, Ames test and Cerep in vitro
screening. The major liability identified for 25 was its inhibitory
a
Mean value of at least three experiments.
Similar loss of activity was observed when the tertiary amine of
compound 25 was changed to an amide (27). Taken together, these
SAR results suggest that tertiary amine is preferred at this position,
and morpholine may not be optimal.
Further studies of these analogs revealed that compound 25 is a
better candidate for advancement to pharmacokinetic and in vivo
efficacy studies based on its superior potency (IC50 = 38 nM for hu-
man DGAT-1 and 120 nM for rat DGAT-1), good selectivity (>100-
fold over DGAT-2 and ACAT), and other pharmaceutical profiling
activity on the hERG channel (IC20 = 0.2 lM) and a strong QT pro-
data (kinetic solubility: 1
ulated gastric fluid: 770
tein binding: 1.5%; human and rat hepatocyte clearance of 6.5 and
lg/mL in PBS at pH 6.5; solubility in sim-
longation was noted at low concentrations (P0.47 lM) in the gui-
lg/mL; free fraction in human serum pro-
nea pig Langendorff heart assay.8
In summary, our lead optimization studies have led to the dis-
covery of 2-phenyl-5-(trifluoromethyl)oxazole amide (compound
6.5 mL/min/kg).
(a) Reduction of cumulative body weight
gain
(b) Plasma triglyceride levels
300
18%
250
14%
10%
6%
Vehicle
Vehicle
200
5 mg/kg
25 mg/kg
50 mg/kg
5 mg/kg
150
100
50
25 mg/kg
50 mg/kg
2%
-2%
0
1
6
11
16
21
day
Figure 2. (a) Dose dependent effects of compound 25 on cumulative body weight gain in DIO rats (n = 10 per group); 4% reduction compared to vehicle on day 21 for the
25 mg/kg dose group (p <0.0001); 8% for the 50 mg/kg dose group (p <0.0001); (b) Dose dependent reduction in plasma triacylglycerides measured on day 21 (n = 10 per
group); 64% for the 50 mg/kg dose group, (p <0.0009). Values are the average for each group. Anova contrast test was used for the statistic analysis.

W. Yun et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7205–7209
7209
3. Yen, C.-L.; Stone, S. J.; Koliwad, S.; Harris, C.; Farese, R. V., Jr. J. Lipid Res. 2008,
49, 2283.
25) as a novel human DGAT-1 inhibitor with excellent in vitro po-
tency, improved solubility, good pharmacokinetics, and demon-
strated in vivo efficacy. It paved the way to the discovery of a
new sub-class of DGAT-1 inhibitors with better in vivo efficacy
and reduced hERG liability, which are detailed in separate
accounts.8,11
4. (a) Smith, S. J.; Cases, S.; Jensen, D. R.; Chen, H. C.; Sande, E.; Tow, B.; Sanan, D.
A.; Raber, J.; Eckel, R. H.; Farese, R. V., Jr. Nat. Genet. 2000, 25, 87; (b) Buhman, K.
K.; Smith, S. J.; Stone, S. J.; Repa, J. J.; Wong, J. S.; Knapp, F. F., Jr; Burri, B. J.;
Hamilton, R. L.; Abumrad, N. A.; Farese, R. V., Jr. J. Biol. Chem. 2002, 277, 25474;
(c) Chen, H. C.; Smith, S. J.; Ladha, Z.; Jensen, D. R.; Ferreira, L. D.; Pulawa, L. K.;
McGuire, J. G.; Pitas, R. E.; Eckel, R. H.; Farese, R. V., Jr. J. Clin. Invest. 2002, 109,
1049; (d) Liu, L.; Yu, S.; Khan, R. S.; Ables, G. P.; Bharadwaj, K. G.; Hu, Y.;
Huggins, L. A.; Eriksson, J. W.; Buckett, L. K.; Turnbull, A. V.; Ginsberg, H. N.;
Blaner, W.; Huang, L.-S.; Goldberg, I. J. J. Lipid Res. 2011, 52, 732.
5. Bolin, D. R.; Cai, J.; Cheung, A. W.-H.; Goodnow, R. A. Jr.; Hamilton, M. M.; Li, S.;
McDermott, L. A.; Yun, W. U.S. Patent 2009/0076275.
6. Bolin, D. R.; Cheung, A. W.-H.; Firooznia, F.; Hamilton. M. M.; Li, S.; McDermott,
L. A.; Qian, Y.; Yun, W. U.S. Patent 7714126, B2, 2010.
7. Yun, W.; Bolin, D. R.; Li, S.; Ahmad, M.; Wertheimer, S. J.; Conde-Knape, K.;
Chen, Y.; Kazmer, S. Abstract of papers, The 237th National Meeting of the
American Chemical Society, Salt Lake City, UT, March, 2009, Abstract MEDI-
256.
8. Qian, Y.; Wertheimer, S. J.; Ahmad, M.; Cheung, A. W.-H.; Firooznia, F.;
Hamilton, M. M.; Hayden, S.; Li, S.; Marcopulos, N.; McDermott, L.; Tan, J.; Yun,
W.; Guo, L.; Pamidimukkala, A.; Chen, Y.; Huang, K.-S.; Ramsey, G. B.; Whittard,
T.; Conde-Knape, K.; Taub, R.; Rondinone, C. M.; Tilley, J.; Bolin, D. J. Med. Chem.
2011, 54, 2433.
9. Grabowska, U.; Rizzo, A.; Farnell, K.; Quibell, M. J. Comb. Chem. 2000, 2, 475.
10. (a) Bolin, D. R.; Sytwu, I.-I.; Humiec, F.; Meienhofer, J. Int. J. Pept. Protein Res.
1989, 33, 353; (b) Cynkowski, T.; Cynkowska, G.; Ashton, P.; Crooks, P. A. J.
Chem. Soc., Chem. Commun. 1995, 2335; (c) Gilchrist, T. L. In Heterocycl. Chem.,
2nd ed. 1992; pp 75–77.
Acknowledgments
We thank Dr. Wayne Levin and Ms. Linda Reik for DGAT-con-
taining membrane preparations; Drs. Robert Goodnow, Jr. and Jian-
ping Cai for the early DGAT-1 hit identification work; Mr. Jia_Kui Li
and Ms. Anjula Pamidimukkala for the formulation and PK study;
Mr. Liang Guo for hERG data and Langendorff study; Dr. Kuo-Sen
Huang and Ms. Xiaolei Zhang for developing DGAT enzyme assay;
Mr. Theodore Lambros for compound purification; Mrs. Gino Sasso,
Vance Bell and Ms. Theresa Burchfield for 1H NMR and LC/MS anal-
ysis; Dr. Adrian W.-H. Cheung for helpful discussions and reading
the manuscript.
References and notes
1. Cases, S.; Smith, S. J.; Zheng, Y.-W.; Myers, H. M.; Lear, S. R.; Sande, E.; Novak, S.;
Collins, C.; Welch, C. B.; Lusis, A. J.; Erickson, S. K.; Farese, R. V., Jr. Proc. Natl.
Acad. Sci. U.S.A. 1998, 95, 13018.
2. Cases, S.; Stone, S. J.; Zhou, P.; Yen, E.; Tow, B.; Lardizabal, K. D.; Voelker, T.;
Farese, R. V., Jr. J. Biol. Chem. 2001, 276, 38870.
11. Cheung, A. W.-H.; Ahmad, M.; Conde-Knape, K.; Firooznia, F.; Guo, L.; Hayden,
S.; Li, S.; Marcopulos, N.; Pamidimukkala, A.; Qian, Y.; Rondinone, C. M.; Tan, J.;
Taub, R.; Wertheimer S. J.; Yun, W.; Bolin, D. Manuscript in preparation.