GluK1 antagonists from 6-(carboxy)phenyl decahydroisoquinoline derivatives. SAR and evaluation of a prodrug strategy for oral efficacy in pain models

Article abstract of DOI:10.1016/j.bmcl.2013.09.046

The synthesis and structure-activity relationship of decahydroisoquinoline derivatives with various benzoic acid substitutions as GluK1 antagonists are described. Potent and selective antagonists were selected for a tailored prodrug approach in order to f

Full text of DOI:10.1016/j.bmcl.2013.09.046

Bioorganic & Medicinal Chemistry Letters 23 (2013) 6459–6462
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
GluK1 antagonists from 6-(carboxy)phenyl decahydroisoquinoline
derivatives. SAR and evaluation of a prodrug strategy for oral efficacy
in pain models
b
b,
Jose A. Martinez-Perez a, , Smriti Iyengar , Harlan E. Shannon , David Bleakman b, Andrew Alt b,à
,
⇑
Brian M. Arnold b, Michael G. Bell b, Thomas J. Bleisch b, Ana M. Castaño a, Miriam Del Prado a,
Esteban Dominguez a, Ana M. Escribano a, Sandra A. Filla b, Ken H. Ho b, Kevin J. Hudziak b, Carrie K. Jones b,§
Ana Mateo a, Brian M. Mathes b, Edward L. Mattiuz b,–, Ann Marie L. Ogden b, Rosa Maria A. Simmons b,
Douglas R. Stack b, Robert E. Stratford b,||, Mark A. Winter b, Zhipei Wu b, Paul L. Ornstein b,
,
a Centro de Investigación Lilly, Avda. de la Industria, 30, 28108-Alcobendas, Madrid, Spain
b Eli Lilly & Co, Lilly Research Laboratories, Indianapolis, IN 46285, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 July 2013
Revised 13 September 2013
Accepted 13 September 2013
Available online 25 September 2013
The synthesis and structure–activity relationship of decahydroisoquinoline derivatives with various ben-
zoic acid substitutions as GluK1 antagonists are described. Potent and selective antagonists were selected
for a tailored prodrug approach in order to facilitate the evaluation of the new compounds in pain models
after oral administration. Several diester prodrugs allowed for acceptable amino acid exposure and mod-
erate efficacy in vivo.
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Pain
Decahydroisoquinolines
GluK1 antagonists
GluA2
Prodrug
H
N
N
N
H
H
H
Kainate receptors form ligand-gated ion channels which are
activated by the primary excitatory neurotransmitter, glutamate.1
Receptors containing the GluK1 subunit (previously known as
GluR5)2 have been implicated in pain transmission.3 Excessive
stimulation of these receptors can lead to abnormal pain process-
ing, including allodynia and hyperalgesia, and therefore GluK1
receptor antagonists may be useful as analgesic agents.4 Clinical
studies by Sang et al.5 have reported that iv administration of the
non-selective kainate antagonist decahydroisoquinoline amino
acid 1 (LY293558, Chart 1),6 significantly reduced pain intensity
and unpleasantness. Simmons et al. showed that 2 (LY382884,
Chart 1),7 a selective GluK1 receptor antagonist, was active in a
rat model of persistent pain via intraperitoneal (ip) administration
CO H
CO H
2
2
N
NH
NH
HO C
2
H
1
2
Chart 1. GluK1 receptor antagonists with anagelsic efficacy in vivo.
and in a model of peripheral neuropathy in monkeys after admin-
istration into the spinal cord.8 Filla et al. demonstrated oral activity
in animal models of migraine pain with the diethyl ester prodrug
of a selective GluK1 receptor antagonist.9
We sought to examine modifications of 2 (LY382884) in an at-
tempt to discover potent and selective GluK1 antagonists with
activity in preclinical models of persistent pain.10
Affinity in GluK1 and in a representative of the a-amino-3-hy-
⇑
Corresponding author. Tel.: +34 916633419; fax: +34 916633411.
E-mail address: Martinez-Perez_Jose_A@Lilly.com (J. A. Martinez-Perez).
Indiana University School of Medicine, IN 46202, United States.
Bristol-Myers Squibb Company, CT 06492, United States.
Vanderbilt University, TN 37235, United States.
Advion-Bioanalytical Labs, a Quintiles Company, IN 46241, United States.
Xavier University of Louisiana, LA 70130, United States.
Roosevelt University, IL 60173, United States.
droxy-5-methyl-4-isoxazolepropionic acid (AMPA) family of iono-
tropic receptors (GluA2) were evaluated in ligand binding
screens.11 Potent and selective GluK1 antagonists were then trans-
formed into diethyl ester prodrugs and evaluated in vivo for oral
bioavailability and activity in established models of persistent
pain.12
à
§
–
||
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.09.046

6460
J. A. Martinez-Perez et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6459–6462
Table 2
H
H
H
O
CO Et
2
HO
CO Et
2
GluK1 in vitro affinity of phenyl-substituted acidic amino acids
c
N
N
R
R
CO H
2
H
H
H
R1
R2
X
CO H
2
5, R= CO2tBu
6, R= CO2Me
NH
R4
b
4, R= CO2Me
R3
7, R= CO2tBu
a
d or e
CO2H position
Compda
X
R1 R2
R3
R4
GluK1 Ki (
l
M)b GluA2 Ki ( M)b
l
HO C
H
H
2
ortho-
meta- para-
9
X
CO H
2
8
2
X= C,
X= NH,
X= S,
19
20
21
22⁄
23
24
25
26
27
28
29
30
31
32
O
O
S
NH
O
O
O
O
S
O
O
O
O
O
F
H
H
H
H
H
H
H
H
Cl
Cl
Me
OMe
Ph
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
9.7
1.5
3.7
3.6
9.9
19
93
19
42
3.3
107
87
172
151
22
89
111
8.8
10
13
16
11
14
17
12
15
18
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Cl
Cl
Me
OMe
Ph
H
H
H
H
H
NH
X= O,
Scheme 1. Reagents and conditions: (a) (X = NH) (i) 4, ArNH2, NaBH(OAc)3, rt (20–
90%), (ii) 6 N HCl, 100 °C (50–100%); (b) (i) TMSI, rt, (ii) Boc2O, Et3N, rt (92%, two
steps); (c) CeCl3, NaBH4, EtOH, ꢀ78 °C to rt (60%); (d) (X = S) (i) 6, MsCl, Et3N, rt
(quant.), (ii) ArSH, K2CO3, acetone (40–70%), (iii) 6 N HCl, 100 °C (70–90%); (e)
(X = O) (i) 7, ArOH, DEAD, Ph3P (60–75%), (ii) LiOH (80–100%), (iii) HCl, EtOAc, rt
(90–100%). Ar: ethyl benzoate.
11
2.1
1.3
5.9
6.2
0.5
316
H
H
1470
818
H
Me 228
We first evaluated several heteroatom replacements of the
linker atom between the decahydroisoquinoline nucleus and the
aromatic acid domain, and prepared all three possible regio-iso-
meric benzoic acids for each linker atom. Compounds 8–18 were
a
⁄
All compounds as hydrochloride salts except those with
Values are means of at least two experiments.
.
b
prepared from the known decahydroisoquinoline ketone
4
The data in Table 1 summarizes the in vitro binding affinities for
(Scheme 1).13 The carbon-linked derivatives (2, 8, 9) were previ-
ously obtained via Wittig–Horner reaction and hydrogenation, fol-
lowed by hydrolysis.7 The nitrogen-linked compounds (10–12)
required reductive aminations between appropriate anilines and
the ketone 4 followed by acidic hydrolysis. Two key intermediates
for the preparation of sulfur- and oxygen-linked derivatives were
the known (6R)-alcohols 6 and 7.14 Successful reduction of ketones
4 or 5 was achieved with sodium borohydride in the presence of an
equimolar quantity of cerium trichloride heptahydrate at low tem-
perature. Alcohol 6 was previously obtained as the minor C-6 epi-
mer by reduction with sodium borohydride in the absence of
cerium.14a The nucleophilic substitution of the mesylate from alco-
hol 6 with thiophenols and the deprotection by acid hydrolysis al-
lowed for the preparation of the sulfur-linked derivatives 13–15.
The oxygen-linked compounds (16–18) were synthesized by Mits-
unobu reaction between appropriate phenols and alcohol 7 fol-
lowed by mild ester hydrolysis and amine deprotection.15
GluK1 and GluA2 of the reference compound 2 and of the amino
acids 8–18. The replacement of the carbon linker in 2 (LY382884)
by a heteroatom, as in 12, 15 or 18, reduced the affinity for the
GluK1. No improvement in the in vitro profile was observed with
meta-benzoic acids derivatives (9, 11, 14, and 17), regardless of
the linker atom; on the contrary, affinity for GluA2 was signifi-
cantly increased and compound 14 was more potent in GluA2 than
in GluK1. Surprisingly, while 8 was the least potent GluK1 antago-
nist in the carbon-linked series, ortho-benzoic acids linked by a
heteroatom showed an increased binding for GluK1 while retaining
a moderate selectivity against GluA2.
Next, we evaluated derivatives of these potent ortho-benzoic
acids (10, 13, 16, Table 1) with substitution in the phenyl ring as
a straightforward entry to a broader set of GluK1 antagonists.
The application of the same synthetic strategy from Scheme 1 with
substituted anthranilates, thiosalicylates16 and salicylates allowed
for the preparation of compounds 19-32 (Table 2).17
Table 1
GluK1 and GluA2 in vitro affinity of amino acids derivatives
General structure
Compda
X
GluK1 Kib
GluA2 Kib
H
2
C
NH
S
O
3.6
528
19
267
553
1420
418
913
X
CO2H
NH
12⁄
15⁄
18
HO2C
H
H
9
C
NH
S
O
16
8
46
6
149
238
9
HO2C
X
CO2H
NH
11⁄
14⁄
17
44
H
CO H
8
C
NH
S
O
67
0.9
0.8
2.3
29600
77
2
10⁄
H
X
CO H
2
13⁄
16
15
70
NH
H
a
⁄
.
All compounds as hydrochloride salts except those with
Values (lM) are means of at least two experiments.
b

J. A. Martinez-Perez et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6459–6462
6461
decahydroisoquinoline ring, limiting substitution in these cases. R3
(Cl in Fig. 1) pointed to a large pocket in agreement with SAR results.
A selection of the above antagonists (Tables 1 and 2) was eval-
uated in vivo and compared with the lead compound 2. While
these amino acids showed poor pharmacokinetic properties (data
not shown), we achieved acceptable oral bioavailability of the ami-
no acids in rats with the use of the corresponding diethyl ester
derivatives (Table 3).9,20 Focusing on the AUC and Cmax parameters
in Table 3, the diethyl ester prodrug of 16 allowed for significantly
higher plasma levels of the corresponding amino acid than was the
case for diethyl ester prodrugs of our original compound 2 or the
sulfur-linked 13. Thus, the presence of oxygen as linker showed
better pharmacokinetic parameters than other atoms. Phenyl
ethers substituted at R3 (26 and 30) still showed improved plasma
levels of amino acid over 2, though these substitutions reduced the
bioavailability of the parent amino acid.
Diethyl ester prodrugs were chosen for evaluation in two ani-
mal models of pain via the oral route (Table 4). The diethyl ester
prodrug of compound 2 showed moderate efficacy in the carra-
geenan-induced thermal hyperalgesia model, with a minimal
effective dose of 10 mg/kg. It also showed weak reversal of the pain
caused by formalin, without sedative or ataxic effects as measured
by the rotorod test, at a single oral dose of 30 mg/kg.8a
Figure 1. Docked model of binding of antagonist 26 based on the X-ray structure
2F35 (Glide 5.6 SP docking). Figure made using VIDA.19
Table 3
Rat PK data of GluK1 antagonist amino acidsa
Compdb
AUC0
(ng h/ml)
1
Cmax (ng/ml)
Tmax (h)
%F
to
The prodrugs of the new set of antagonists showed similar
activity to the diethyl ester of 2 in both pain models despite mod-
est improvements in affinity and/or exposure. Though the prodrug
of compound 13 produced a significant larger reversal of pain in
the formalin model than the prodrug of compound 2, this occurred
with a higher dose of 100 mg/kg. Efficacy in vivo of these new com-
pounds may be limited by poor blood-brain barrier penetration of
the parent amino acids. Indeed, experiments with compound 2 and
related GluK1 antagonists, showed that higher in vivo efficacy was
observed after central administration, and furthermore, this was in
the same rank order as the compounds’ in vitro potency.21
In conclusion, we have discovered a new family of GluK1 antag-
onists with a particular linker-acid group relationship. The phenyl
substitution of these benzoic acid derivatives suggested the mini-
mum requirements to maintain acceptable affinity for GluK1 and
may be useful to develop other GluK1 antagonists. In addition,
we have demonstrated that prodrug esters allow for a satisfactory
rat oral bioavailability of the amino acids and for oral evaluation in
pain models.
2c
1506
7531
24743
9193
539
2981
9205
6449
2695
1
62.5
38.8
75.3
32.9
22.2
13
16
26
30
0.5
0.5
0.5
1
7226
a
Pharmacokinetics (PK) parameters of amino acids in Fischer rats after 30 mg/kg
oral dose. Three rats per time point were used. AUC0 to 1: area under the amino acid
plasma concentration curve. Cmax: maximum observed amino acid plasma con-
centration. Tmax: time at Cmax. %F: the bioavailability of the amino acid following its
oral administration as diethyl ester relative to the amino acid administered by the
intravenous (iv) route.
b
Amino acid dosed p.o. as diethyl ester hydrochloride salt.
Data from a 10 mg/kg oral dose.
c
Table 4
In vivo effect of diethyl esters of GluK1 antagonists amino acidsa
Compdb Carr. Therm. Hyperalg.
MEDc (mg/kg)
Formalin% Reversal @
30 mg/kg p.o
Rotorod MEDc
(mg/kg)
2
13
16
26
10 (N = 6)
10 (N = 6)
10 (N = 6)
>10 (N = 6)
13.4 11.6 (N = 9)
49.4 4.6 (N = 8)d
21.8 13.6 (N = 9)
14.9 8.0 (N = 8)
>30 (N = 8)
>100 (N = 8)
>30 (N = 8)
>30 (N = 8)
Acknowledgments
a
The total number of animals used for in vivo characterization was 255,
The authors would like to thank Howard Broughton for the
preparation of the GluK1 model. For the in vivo work, all experi-
ment protocols were approved by the Eli Lilly & Co. Institutional
Animal Care and Use Committee. Principles of Replacement,
Reduction, and Refinement were implemented.
including vehicle controls. Vehicle controls for carrageenan, N = 6, formalin, N = 8–
12, rotorod, N = 8. Experimental methods based on Ref. 14b.
b
Amino acid dosed p. o. as diethyl ester hydrochloride salt.
Several doses were evaluated. MED = minimal effective dose.
c
d
Result obtained with 100 mg/kg p. o.
A small group like fluorine ortho- to the acid group (19) pro-
duced a modest weakening of GluK1 binding. Substitution at R4,
ortho- to the linker atom X, (31, 32) was deleterious to the affinity
for GluK1, which may indicate conformational preferences around
the linker, disturbed by crowded environments. The substitutions
at R2 or R3 were well tolerated when the linker atom was sulfur
or oxygen, while a large reduction of the selectivity against GluA2
was noticed with the amino as linker (22). The most important po-
sition appears to be R3 where a variety of substitution was allowed
and offered the possibility of obtaining a set of potent and selective
antagonists, such as 26–30.
References and notes
1. (a) Watkins, J. C.; Jane, D. E. Br. J. Pharmacol. 2006, 147, S100; (b) Mellor, J. R.
Biochem. Soc. Trans. 2006, 34, 949.
2. Collingridge, G. L.; Olsen, R.; Peters, J. A.; Spedding, M. Neuropharmacology
2009, 56, 2.
3. (a) Li, P.; Wilding, T. J.; Kim, S. J.; Calejesan, A. A.; Huettner, J. E.; Zhuo, M.
Nature 1999, 397, 161; (b) Procter, M. J.; Houghton, A. K.; Faber, E. S. L.; Chizh, B.
A.; Ornstein, P. L.; Lodge, D.; Headley, P. M. Neuropharmacology 1998, 37, 1287;
(c) Kerchner, G. A.; Wilding, T. J.; Huettner, J. E.; Zhuo, M. J. Neurosci. 2002, 22,
8010; (d) Qiu, C.-S.; Wyhe, L. L.-V.; Sasaki, M.; Sakai, R.; Swanson, G. T.; Gereau,
R. W., IV Pain 2011, 152, 1052; (e) Jane, D. E.; Lodge, D.; Collingridge, G. L.
Neuropharmacology 2009, 56, 90.
A model based on the public crystal structure of an antagonist
(UBP302; pdb 2F35) bound to the ligand binding domain of GluK1
allowed rationalization of some of these observations (Fig. 1).18 R1
and R2 were close to the protein surface and R4 was close to the
4. (a) Williams, M.; Kowaluk, E. A.; Arneric, S. P. J. Med. Chem. 1999, 42, 1481; (b)
Bräuner-Osborne, H.; Egebjer, J.; Nielsen, E. Ø.; Madsen, U.; Krosgaard-Larsen,
P. J. Med. Chem. 2000, 43, 2609.
5. Gilron, I.; Max, M. B.; Lee, G.; Booher, S. L.; Sang, C. N.; Chappell, A. S.; Dionne, R.
A. Clin. Pharmacol. Ther. 2000, 68, 320.

6462
J. A. Martinez-Perez et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6459–6462
6. Ornstein, P. L.; Arnold, M. B.; Allen, N. K.; Bleisch, T.; Borromeo, P. S.; Lugar, C.
W.; Leander, J. D.; Lodge, D.; Schoepp, D. D. J. Med. Chem. 1996, 39, 2232.
7. Bleisch, T.; Ornstein, P. L.; Allen, N. K.; Wright, R. A.; Lodge, D.; Schoepp, D. D.
Bioorg. Med. Chem. Lett. 1997, 7, 1161.
Stack, D. R.; Stratford, R. E.; Winter, M. A.; Johnson, K. W.; Wu, Z.; Ornstein, P. L.
J. Med. Chem. 2005, 48, 4200; (c) The assignment of the stereochemistry is
based on the axial versus equatorial coupling constants observed for each of
the epimeric alcohols obtained.
8. (a) Simmons, R. M. A.; Li, D. L.; Ho, K. H.; Deverill, M.; Ornstein, P. L.; Iyengar, S.
Neuropharmacology 1998, 37, 25; (b) Palecek, J.; Neugebauer, V.; Carlton, S. M.;
Iyengar, S.; Willis, W. D. Pain 2004, 111, 151.
9. Filla, S. A.; Winter, M. A.; Johnson, K. W.; Bleakman, D.; Bell, M. G.; Bleisch, T. J.;
Castaño, A. M.; Clemens-Smith, A.; Del Prado, M.; Dieckman, D. K.; Dominguez,
E.; Escribano, A.; Ho, K. H.; Hudziak, K. J.; Katofiasc, M. A.; Martinez-Perez, J. A.;
Mateo, A.; Mathes, B. M.; Mattiuz, E. L.; Ogden, A. M. L.; Phebus, L. A.; Stack, D.
R.; Stratford, R. E.; Ornstein, P. L. J. Med. Chem. 2002, 45, 4383.
15. The use of a more labile nitrogen protecting group, as in 7, is recommended for
the preparation of the oxygen-linked derivatives.
16. Novel substituted phenyl thiols were prepared from salicylates: Arnould, J. C.;
Didelot, M.; Cadilhac, C.; Pasquet, M. J. Tetrahedron Lett. 1996, 37, 4523.
17. (a) Arnold, M. B.; Bleisch, T. J.; Castaño, A. M.; Dominguez-Manzanares, E.;
Escribano, A. M.; Martinez-Perez, J. A.; Mateo, A. I.; Ornstein, P. L.; Winter, M.
A.; May, S. A.; Ratz, A. M; Wilson, T. M. WO2003024453, 2003; PCT Int. Appl.
2003.; (b) For synthetic details of N-linked compounds see Bleisch, T. J.;
Castaño, A. M.; Dominguez-Manzanares, E.; Escribano, A. M.; Ornstein, P. L.
WO2003024934, 2003; PCT Int. Appl, 2003.
10. For other medicinal chemistry approaches see: (a) Larsen, A. M.; Bunch, L. ACS
Chem. Neurosci. 2011, 2, 60; (b) Venskutonyte, R.; Butini, S.; Coccone, S. S.;
˙
Gemma, S.; Brindisi, M.; Kumar, V.; Guarino, E.; Maramai, S.; Valenti, S.; Amir,
A.; Valadés, E. A.; Frydenvang, K.; Kastrup, J. S.; Novelino, E.; Campiani, G.;
18. (a) Mayer, M. L.; Ghosal, A.; Dolman, N. P.; Jane, D. E. J. Neurosci. 2006, 26, 2852;
(b) Standard procedures in the Schrodinger Molecular Modeling suit, 2006,
were used; An antagonist bound rather than an agonist bound x-ray was used
due to known changes in domain closure - see (c) Unno, M.; Shinohara, M.;
Takayama, K.; Tanaka, H.; Teruya, K.; Doh-ura, K.; Sakai, R.; Sasaki, M.; Ikeda-
Saito, M. J. Mol. Biol. 2011, 413, 667; The GluK1 X-ray crystal structure of a
decahydroisoquinoline antagonist subsequently disclosed (PDB entry 2QS4)
shows a heavy-atom RMSD of 0.75 Å from the 2F35-based model for residues
within 5 Å of the ligand - see (d) Alushin, G. M.; Jane, D.; Mayer, M. L.
Neuropharmacology 2011, 60, 126.
19. VIDA 4.0.0. OpenEye Scientific Software: Santa Fe, NM, USA, 2009.
20. (a) Dialkyl ester compounds were largely devoid of affinity for GluK1 receptor;
(b) Ester compounds were usually prepared from the amino acids by treatment
with the corresponding alcohol saturated with hydrogen chloride; For a review
on prodrugs see (c) Beaumont, K.; Webster, R.; Gardner, I.; Dack, K. Curr. Drug
Metab. 2003, 4, 461.
Pickering, D. S. J. Med. Chem. 2011, 54, 4793; (c) Venskutonyte, R.; Frydenvang,
˙
K.; Valadés, E. A.; Szyman´ ska, E.; Johansen, T. N.; Kastrup, J. S.; Pickering, D. S.
ChemMedChem 2012, 7, 1793.
11. GluK1 binding was measured using [3H]ATPA - see (a) Ho, K. H.; Legutko, B.;
Rizkalla, G.; Deverill, M.; Hawes, C. R.; Ellis, G. J.; Stensbol, T. B.; Krogsgaard-
Larsen, P.; Skolnick, P.; Bleakman, D. Neuropharmacology 1999, 38, 1811; GluA2
binding was measured using [3H]AMPA - see (b) Small, B.; Thomas, J.; Kemp,
M.; Hoo, K.; Ballick, B.; Deverill, M.; Ogden, A. M.; Rubio, A.; Pedregal, C.;
Bleakman, D. Neuropharmacology 1998, 37, 1261.
12. Ettmayer, P.; Amidon, G. L.; Clement, B.; Testa, B. J. Med. Chem. 2004, 47, 2393.
13. Ornstein, P. L.; Arnold, M. B.; Augenstein, N. K.; Paschal, J. W. J. Org. Chem. 1991,
56, 4388.
14. (a) Ornstein, P. L.; Arnold, M. B.; Allen, N. K.; Bleisch, T.; Borromeo, P. S.; Lugar,
C. W.; Leander, J. D.; Lodge, D.; Schoepp, D. D. J. Med. Chem. 1996, 39, 2219; (b)
Dominguez, E.; Iyengar, S.; Shannon, H. E.; Bleakman, D.; Alt, A.; Arnold, B. M.;
Bell, M. G.; Bleisch, T. J.; Buckmaster, J. L.; Castaño, A. M.; DelPrado, M.;
Escribano, A.; Filla, S. A.; Ho, K. H.; Hudziak, K. J.; Jones, C. K.; Martinez-Perez, J.
A.; Mateo, A.; Mathes, B. M.; Mattiuz, E. L.; Ogden, A. M. L.; Simmons, R. M. A.;
21. Jones, C. K.; Alt, A.; Ogden, A. M.; Bleakman, D.; Simmons, R. M. A.; Iyengar, S.;
Dominguez, E.; Ornstein, P. L.; Shannon, H. E. J. Pharmacol. Exp. Ther. 2006, 319,
396.