Fused tricyclic indoles as S1P1 agonists with robust efficacy in animal models of autoimmune disease

Article abstract of DOI:10.1016/j.bmcl.2012.04.129

Two series of fused tricyclic indoles were identified as potent and selective S1P₁ agonists. In vivo these agonists produced a significant reduction in circulating lymphocytes which translated into robust efficacy in several rodent models of autoimmune disease. Importantly, these agonists were devoid of any activity at the S1P₃ receptor in vitro, and correspondingly did not produce S1P₃ mediated bradycardia in telemeterized rat.

Full text of DOI:10.1016/j.bmcl.2012.04.129

Bioorganic & Medicinal Chemistry Letters 22 (2012) 4404–4409
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Fused tricyclic indoles as S1P1 agonists with robust efficacy in animal models
of autoimmune disease
Daniel J. Buzard , Sangdon Han, Luis Lopez, Andrew Kawasaki , Jeanne Moody à, Lars Thoresen,
⇑
Brett Ullman, Juerg Lehmann, Imelda Calderon, Xiuwen Zhu, Tawfik Gharbaoui §, Dipanjan Sengupta §,
Ashwin Krishnan, Yinghong Gao –, Jeff Edwards k, Jeremy Barden, Michael Morgan, Khawja Usmani,
Chuan Chen, Abu Sadeque, Jayant Thatte , Michelle Solomon, Lixia Fu àà, Kevin Whelan, Ling Liu §§
Hussien Al-Shamma, Joel Gatlin, Minh Le, Charles Xing, Sheryll Espinola ––, Robert M. Jones
,
Arena Pharmaceuticals, 6166 Nancy Ridge Drive, San Diego, CA 92121, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Two series of fused tricyclic indoles were identified as potent and selective S1P1 agonists. In vivo these
agonists produced a significant reduction in circulating lymphocytes which translated into robust efficacy
in several rodent models of autoimmune disease. Importantly, these agonists were devoid of any activity
at the S1P3 receptor in vitro, and correspondingly did not produce S1P3 mediated bradycardia in telem-
eterized rat.
Received 29 March 2012
Revised 26 April 2012
Accepted 27 April 2012
Available online 7 May 2012
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Sphingosine-1 phosphate receptor
S1P1 agonist
Immunomodulators
FTY720
Fingolimod (1) is a potent immunosuppressant that was identi-
fied during a medicinal chemistry campaign aimed at dialing out
the GI side effect profile associated with the secondary fungal
metabolite ISP-1 (2, Fig. 1).1 In vivo, (1) is phosphorylated to the
(S)-isomeric mono-phosphate (3) which is an agonist at four of
the five sphingosine-1-phosphate receptors (S1P1, S1P3–5).2 The
S1P/S1P1 axis regulates the egress of lymphocytes from peripheral
lymphoid tissue and (3) prevents this trafficking via sustained
receptor internalization accompanied by proteasomal degrada-
tion.3 Impaired lymphocyte trafficking reduces the circulation of
auto-reactive immune cells, which has been found to slow the pro-
gression of disease in multiple sclerosis patients.4 Fingolimod also
may be acting directly on astrocytes to reduce inflammation in the
CNS of these patients. S1P1 expression in central MS lesions is dra-
matically increased, and mouse neural S1P1 knockouts phenotypi-
cally show reduced disease severity, astrogliosis, demyelination
and axonal loss when compared to their wild type littermates.5
The precise role of the other S1P subtypes in the amelioration of
disease is less well understood, although (1) may be acting on
the S1P5 and S1P3 receptors expressed centrally on oligodendro-
cytes and astrocytes.3
⇑
Corresponding author.
E-mail address: dbuzard@arenapharm.com (D.J. Buzard).
Present address: Focus Synthesis, 11475 Sorrento Valley Road, Suite 212, San
Diego, CA 92121, USA.
à
Present address: Regulus Therapeutics, 3545 John Hopkins Court, San Diego, CA
92121, USA.
§
Present address: Dart NeuroScience LLC, 7473 Lusk Boulevard, San Diego, CA
Over the last decade, medicinal chemistry efforts have focused
on identifying alternative agents with improved selectivity. Partic-
ular emphasis has been placed on the cardiac expressed S1P3
receptor that is associated with bradycardia in rodents.6 However,
human trials of the S1P3-sparing agonist BAF312, suggests that the
transient effect on heart rate may not be attributable to S1P3 acti-
vation in humans.7 More recently, researchers have sought ago-
nists with improved pharmacokinetic profiles given the long
half-life of (1) in humans (T1/2 = 8 days).8 These efforts have also
centered around developing non pro-drug agonists that can pene-
trate the CNS and act directly on astrocytes.9 Development of
92121, USA.
–
Present address: EMD Millipore, 10394 Pacific Center Court, San Diego, CA 92121,
USA.
k
Present address: Amylin Pharmaceuticals, 9360 Towne Center Drive, San Diego,
CA 92121, USA.
Present address: PRESCOS LLC, 3550 General Atomics Court, M/S 02-422, San
Diego, CA 92121, USA.
àà
Present address: Synthetic Genomics, 11149 North Torrey Pines Road, La Jolla, CA
92037, USA.
§§
Present address: Pfizer, 10646 Science Center Drive, San Diego, CA 92121-1150,
USA.
––
Present address: Genomics Institute of the Novartis Research Foundation, 10675
John Jay Hopkins Drive, San Diego, CA 92121, USA.
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.04.129

D. J. Buzard et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4404–4409
4405
O
OH NH2
OH
OH
CO2H
NH2
OH
HO
1
2
Fingolimod
ISP-1
[FTY720, Gilenya]
O
P
OH
O
OH
NH2
HO
3
FTY720-P
Figure 1. Fingolimod origin and mechanism.
agonists for other autoimmune indications such as psoriasis
(Ponesimod),10 polymyositis (BAF312),11 and ulcerative colitis
(KRP-203)12 is also underway.
As part of our effort to identify superior S1P1 agonists, we have
expanded upon an early series of 3-(indolin-2-yl)propanoic acids
(4, Fig. 2)13 and herein disclose two fused tricyclic indole S1P1
chemotypes (5, 6).
The cyclopenta[b]indole ring system (8) was prepared via a
palladium mediated cyclization utilizing aniline and ethyl
7
2-(2-oxocyclopentyl)acetate as coupling partners (Scheme 1). Sub-
sequent hydroxy-amidine formation (9), and treatment with an
activated benzoic acid afforded the 5-phenyl-1,2,4-oxadiazole mo-
tif (10). Ester saponification was typically performed using LiOH,
however, compound 17 (Table 1) required a milder procedure
R1
O
N
1
4
2
8
R2
7
N
3
CO2H
R3
N
H
6
R1
5
O
N
5
4
7
R2
3
5
cyclopenta[b]indole
N
2
S1P1 agonists
R3
N
6
CO2H
1
H
R1
O
N
4
8
5
R2
9
3-(indolin-2-yl)propanoic acid
S1P1 agonists
7
N
1
CO2H
R3
N
6
2
3
6
pyrrolo[1,2-a]indole
S1P1 agonists
Figure 2. Cyclopenta[b]indole and Pyrrolo[1,2-a]indole S1P1 agonists.
NH2
NC
I
NC
HO
a
b
N
CO2Et
CO2Et
N
H
N
H
NH2
7
8
9
c
R1
R1
O
N
O
N
N
d
R2
R2
N
CO2H
CO2Et
R3
R3
N
N
H
H
17-22
10
Scheme 1. Cyclopenta[b]indole synthesis. (a) Ethyl 2-(2-oxocyclopentyl)acetate, PPTS, Si(OEt)4 then Pd(OAc)2, DIEA (b) 50% aq hydroxylamine, EtOH (c) ArCOCl, Et3N or
ArCO2H, CDI (d) NaOH or LiBr, Et3N, MeCN, water.

4406
D. J. Buzard et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4404–4409
Table 1
Cyclopenta[b]indole and Pyrrolo[1,2-a]indole SAR
R1
R1
O
O
N
N
R2
R2
N
N
CO2H
CO2H
R3
R3
N
H
N
R1
R2
R3
Cmpd
hS1P1 EC50 (nM)
Cmpd
hS1P1 EC50 (nM)
a
a
OCF3
OCH3
CF3
H
H
H
H
H
H
OCF3
OCH3
O-iPr
CN
CN
CF3
CN
CN
CN
17
18
19
20
21
22
1.1
157
76
3.2
4.6
23
24
25
26
27
28
0.25
4.6
2.1
0.59
2.9
0.61
7.5
a
EC50 values were determined in a human cAMP HTRF assay and are the mean of three or more replicates. All compounds afforded a full response relative to S1P in the
assay.
Br
Br
Br
a
b
O
O
CO2Et
N
H
N
N
CO2Et
11
12
13
c, d
NH2
HO
NC
g
e, f
N
CO2tBu
CO2tBu
N
N
15
14
R1
R1
O
O
N
N
h
R2
R2
N
N
CO2tBu
CO2H
R3
R3
N
N
16
23-28
Scheme 2. Pyrrolo[1,2-a]indole synthesis. (a) Butyl acrylate, NaH (b) HCl, AcOH (c) BrPh3PCHCO2tBu, THF (d) CuCN, NMP (e) H2, Pd/C (f) 50% aq hydroxylamine, EtOH (g)
ArCOCl, Et3N or ArCO2H, CDI (h) TFA, thioanisole/cysteine.
employing lithium bromide in order to avoid reaction at the cyano
group.14
species as well as the other human subtypes (Table 2). All four
enantiomers maintained a good level of activity for the rodent
receptors, and were selective against the human S1P2, S1P3, and
S1P4 subtypes. In all cases, significant agonism at the S1P5 receptor
was observed. Selectivity data was generated using a DiscoveRX b-
arrestin platform,15 wherein a significant shift in human S1P1 po-
tency was observed. The more potent enantiomer in each chemical
series (17b and 23b) was subsequently selected for further
evaluation.
ADME profiling revealed good stability in mouse, rat, and hu-
man liver microsomes, and 17b and 23b were determined to have
a terminal half-life of approximately 5 h in male Sprague–Dawley
rats after oral administration (Table 3). Both compounds had rela-
tively low systemic clearance and volume of distribution, but 23b
was found to have a higher oral bioavailability achieving plasma
concentrations of >10X that of 17b.
The pyrrolo[1,2-a]indole (12) was prepared upon treatment of
11 with butyl acrylate under basic conditions (Scheme 2). Hydroly-
sis and decarboxylation of the butyl ester gave ketone 13, which
was subjected to a Wittig reaction and subsequent cyanation.
Hydrogenation, followed by treatment with hydroxylamine, affor-
ded intermediate 15, which was used in the oxadiazole forming
step (16). The tert-butyl ester was removed upon treatment with
TFA and a cation scavenger.
Human S1P1 potency for several racemic analogs (17–28) was
determined in a HTRF cAMP assay (Table 1). Several structural per-
mutations were examined, of which the 3-cyano-5-trifluorometh-
oxy-phenyl array appeared to confer the greatest potency across
both chemical series. Compounds 17 and 23 were successfully sep-
arated into their respective enantiomers by chiral chromatography
using a ChiralCelODÒ column (17, IPA/hexanes) and a ChiralPa-
kADHÒ column (23, MeCN). Stereochemical assignments were
not made at this time, but the individual enantiomers were re-
ferred to by their elution order (e.g., 17a first eluting, 17b second
eluting; 23a first eluting, 23b second eluting). The resolved enanti-
omers were subsequently characterized across several pre-clinical
Based on the accepted mechanism of action, a reduction in cir-
culating lymphocytes is an essential requirement for the develop-
ment of new S1P1 agonists. Compounds 17b and 23b were
evaluated in a mouse lymphocyte lowering time-course experi-
ment at a 6 mg/kg oral dose (Fig. 3). Both compounds produced
a
maximal reduction in circulating lymphocytes by the 5 h

D. J. Buzard et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4404–4409
4407
Table 2
Cross species S1P1 activity and human subtype selectivity data
a
a
S1P1 EC50 [cAMP, nM]
hS1P1–5 EC50 [b-arrestin, nM]
Human
Rat
Mouse
S1P1
S1P2
SIP3
S1P4
SIP5
17a
17b
23a
23b
1.73
0.75
0.76
0.08
1.4
1.8
3.4
0.1
8.3
2.6
5.5
0.52
55
36
123
>100,000
>100,000
>100,000
>100,000
>100,000
>100,000
>100,000
>100,000
239
462
>100,000
>100,000
11
15
289
21
4.8
a
EC50 values are the mean of three or more replicates.
Table 3
Microsomal stability and rat pharmacokinetics
Cmpd
LMa (T1/2
)
Male Sprague–Dawley rat pharmacokineticsb
Cl (L/h/kg)
Vss (L/kg)
T1/2 (h)
Cmax (ng/mL)
AUClast (hr⁄lg/mL)
%F
17b
23b
>60 min (m, r, h)
>60 min (r, h)
0.0834
0.0253
0.734
0.234
5.46
4.68
1.40
13.3
11.0
124
32.5
100
a
Microsomal half-life.
2.0 mg/kg IV (100% PEG400), 3.0 mg/kg PO (0.5% methyl cellulose).
b
veh: 0.5% MC
veh: 0.5% MC
23b 6mg/kg
10
8
8
17b 6mg/kg
6
4
2
6
4
2
0
0
0
0
5
10
15
Time (hr)
20
25
5
10
15
20
25
Time (hr)
Figure 3. BALB/c mice lymphocyte lowering. Compounds 17b and 23b were dosed orally (6 mg/kg) as a suspension in 0.5% methyl cellulose.
time-point, with 23b maintaining this level of suppression out
to 24 h. Lowering the dose of 23b to 1.0 mg/kg produced an
equivalent response at 5 h, but normalization of lymphocyte levels
occurred by 24 h.
Both 17b and 23b were efficacious in several rodent models of
autoimmune disease; however, a much higher dose (30 mg/kg) of
17b was required to achieve complete suppression of lymphocytes
over 24 h and maximal benefit in these chronic studies. Consistent
with the 24 h lymphocyte data, 23b was efficacious at lower doses
and showed a clear dose response (Fig. 4 and 5). Compound 23b
was found to be effective in delaying the onset and reducing the
severity of disease in a mouse experimental autoimmune enceph-
alomyelitis (EAE) model, a myelin/oligodendrocyte glycoprotein
(MOG) 35–55 induced demyelinating autoimmune response simi-
lar to MS (Fig. 4).16 At the highest dose, 23b showed comparable
efficacy to 1 in terms of reduced disease score and incidence.
Similarly, this compound was effective at preventing disease in a
cyclophosphamide-induced Type I diabetes model performed in
non-obese diabetic (NOD) mice (Fig. 5a).17 Treatment with 23b
maintained normal glucose levels over the duration of the experi-
ment, indicating a reduction in T-cell mediated b-cell destruction
and delayed disease progression. Prophylactically, in a collagen in-
duced arthritis (CIA) model performed in female Lewis rats
(Fig. 5b),18 23b reduced joint inflammation (as measured by mean
ankle diameter) at all doses evaluated. Compound 23b appeared to
exceed the efficacy of the 1 mg/kg dose of (1), and approached the
5
100
0.5% MC
0.5% MC
1 1 mg/kg
23b 1mg/kg
23b 3mg/kg
1 1mg/kg
4
3
2
1
0
80
60
40
20
0
23b 1mg/kg
23b 3mg/kg
23b 10mg/kg
23b 10mg/kg
0
7
14 21 28 35 42
Days
8
10 12 14 16 18 20 22 24
Days
Figure 4. Mouse experimental autoimmune encephalomyelitis (EAE) model. Female C57Bl/6 mice were immunized with MOG35-55 peptide and Freund’s complete adjuvant.
On day 3 mice were dosed orally (q.d.) with test compound in 0.5% methylcellulose. The animals were monitored daily for signs of disease. Disease severity was graded on a
scale of 1–5, with one being limp tail or hind limb weakness and five being death.

4408
D. J. Buzard et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4404–4409
a. NOD Mouse T1D
600
b. Rat CIA
0.5% MC
1 1mg/kg
23b 3mg/kg
0.5% MC
500
400
300
200
100
0
23b 1mg/kg
23b 3mg/kg
23b 10mg/kg
1 1 mg/kg
0.33
23b 10mg/kg
0.31
0.29
0.27
0.25
MTX 0.075 mg/kg
0
10 20 30 40 50 60 70
Days
9
8
10111213141516 1718
Days
Figure 5. NOD mouse cyclophosphamide-induced Type I diabetes model and rat collagen induced arthritis model. (a) Non–obese diabetic (NOD) mice were dosed with
cyclophosphamide (i.p.) on day 0 and 14 to accelerate the onset of diabetes. Oral dosing (q.d.) of test compound in 0.5% methyl cellulose was performed from day 0 to day 25.
Animals were considered diabetic if they had a non-fasting blood glucose value >250 mg/dL. (b) Bovine type II collagen emulsified in Freund’s complete adjuvant was
administered to female Lewis rats on days 0 and 6. Starting on day 0, test compound was delivered orally as a suspension in 0.5% methyl cellulose. Caliper measurements of
ankle diameter were taken on days 9–17. MTX = methotrexate.
VEH
450
400
350
300
250
23b 3mg/kg
23b 30mg/kg
1 3mg/kg
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Time (hr)
Figure 6. Effects of 23b on heart rate in telemeterized Charles–River Sprague–Dawley rats. Test compound was dosed orally as a suspension in 0.5% methyl cellulose.
efficacy threshold achieved by methotrexate (MTX) which is a first
line therapy for the treatment of rheumatoid arthritis.
dents. Additional studies for these compounds and other second
generation S1P1 agonists will be reported elsewhere in due course.
In vitro safety profiling included screening for unwanted cross-
reactivity at the CYP P450 isozymes and the hERG channel, as well
as evaluation of the cytotoxicity and mutagenicity potential. Com-
References and notes
1. (a) Adachi, K.; Kohara, T.; Nakao, N.; Arita, M.; Chiba, K.; Mishina, T.; Sasaki, S.;
Fujita, T. Bioorg. Med. Chem. Lett. 1995, 5, 853; (b) Kiuchi, M.; Adachi, K.; Kohara,
T.; Minoguchi, M.; Hanano, T.; Aoki, Y.; Mishina, T.; Arita, M.; Nakao, N.;
Ohtsuki, M.; Hoshino, Y.; Teshima, K.; Chiba, K.; Sasaki, S.; Fujita, T. J. Med.
Chem. 2000, 43, 2946.
2. Hale, J.; Yan, L.; Neway, W.; Hajdu, R.; Bergstrom, J.; Milligan, J.; Shei, G.;
Chrebet, G.; Thornton, R.; Card, D.; Rosenbach, M.; Rosen, H.; Mandala, S.
Bioorg. Med. Chem. Lett. 2004, 12, 4803.
3. Oo, M.; Thangada, S.; Wu, M.; Liu, C.; Macdonald, T.; Lynch, K.; Lin, C.; Hla, T. J.
Biol. Chem. 2007, 282, 9082.
4. Kappos, L.; Radue, E.; O’Conner, P.; Polman, C.; Hohlfeld, R.; Calabresi, P.;
Salmaj, K.; Agoropoulou, C.; Layk, M.; Auberson, L.; Burtin, P. N. Eng. J. Med.
2010, 362, 387.
5. Choi, J.; Gardell, S.; Herr, D.; Rivera, R.; Lee, C.; Noguchi, K.; Teo, S.; Yung, Y.; Lu,
M.; Kennedy, G.; Chun, J. Proc. Natl. Acad. Sci. 2011, 108, 751.
6. Hamada, M.; Nakamura, M.; Kiuchi, M.; Marukawa, K.; Tomatsu, A.; Shimano,
K.; Sato, N.; Sugahara, K.; Asayama, M.; Takagi, K.; Adachi, K. J. Med. Chem. 2010,
53, 3154.
7. Gergely, P.; Wallström, E.; Nuesslein-Hildesheim, B.; Bruns, C.; Zécri, F.; Cooke,
N.; Traebert, M.; Tuntland, T.; Rosenberg, M.; Saltzman, M. Multiple Sclerosis
2009, 15, S125.
8. http://www.receptos.com/clinical-development-RPC1063.php.
9. Demont, E.; Arpino, S.; Bit, R.; Campbell, C.; Deeks, N.; Desai, S.; Dowell, S.;
Gaskin, P.; Gray, J.; Harrison, L.; Haynes, A.; Heightman, T.; Holmes, D.;
Humphreys, P.; Kumar, U.; Morse, M.; Osborne, G.; Panshal, T.; Philpott, K.;
Taylor, S.; Watson, R.; Willis, R.; Witherington, J. J. Med. Chem. 2011, 54, 6724.
10. (a) Piali, L.; Froidevaux, S.; Hess, P.; Nayler, O.; Bolli, M.; Schlosser, C.; Steiner,
B.; Clozel, M. J. Pharmacol. Exp. Ther. 2011, 337, 547; (b) Bolli, M.; Abele, S.;
Binkert, C.; Bravo, R.; Buchmann, S.; Bur, D.; Gatfield, J.; Hess, P.; Kohl, C.;
Mangold, C.; Mathys, B.; Menyhart, K.; Müller, C.; Nayler, O.; Scherz, M.;
Schmidt, G.; Sippel, V.; Steiner, B.; Strasser, D.; Treiber, A.; Weller, T. J. Med.
Chem. 2010, 53, 4198; (c) http://www.clinicaltrials.gov/ct2/show/
NCT00852670?term=ACT-128800&rank=3.
pound 23b was not an inhibitor (IC50 > 30
P450 isoforms (3A4, 2D6, 2C9, 2C19, 1A2), and no interaction with
the hERG channel was observed in astemizole binding (IC50
30 M) or in patch clamp (IC50 > 400 M). Little activity was ob-
served in an in vitro cellular toxicity assay measuring changes in
membrane integrity (IC50 = 25 M), nuclear size (IC50 > 1000 M),
mitochondrial membrane potential (IC50 > 1000 M), and intracel-
lular calcium release (IC50 = 346 M). Ames testing showed no
lM) of the major CYP
>
l
l
l
l
l
l
frame shift or base pair mutations in the five strains tested.
Given the bradycardia observed in human trials of 1, 23b was
examined in telemeterized Sprague–Dawley rats to detect any
changes in heart rate. Compound 23b was dosed orally at 3 and
30 mg/kg and 1 was included as a positive control. Heart rate
was measured over a 20 h period and the hourly readings are
shown in Figure 6. Consistent with literature reports,19 1 produced
a pronounced reduction in heart rate in this experiment. Overall,
compound 23b was very similar to the vehicle treated animals,
suggesting limited interaction with the rodent S1P3 receptor, and
supporting the in vitro findings.
In conclusion, we have identified two new S1P1 chemotypes
comprised of a fused tricyclic indole motif. Optimization efforts
led to the identification of two lead molecules which possess excel-
lent selectivity against the cardiac expressed S1P3 subtype. In vivo,
our most potent compound 23b demonstrated robust lymphocyte
lowering in mouse, which translated into impressive efficacy in
three different rodent models of autoimmune disease. Further-
more, this compound was found to have an acceptable in vitro
safety profile and did not induce S1P3 mediated bradycardia in ro-
11. http://clinicaltrials.gov/show/NCT01148810.
12. (a) http://clinicaltrials.gov/show/NCT01375179.; (b) Song, J.; Matsuda, C.; Kai,
Y.; Nishida, T.; Nakajima, K.; Mizushima, T.; Kinoshita, M.; Yasue, T.; Sawa, Y.;
Ito, T. J. Pharmacol. Exp. Ther. 2008, 324, 276.

D. J. Buzard et al. / Bioorg. Med. Chem. Lett. 22 (2012) 4404–4409
4409
13. Unpublished work.
14. Mattsson, S.; Dahlströrm, M.; Karlsson, S. Tetrahedron Lett. 2007, 48, 2497.
15. http://www.discoverx.com/gpcrs/arrestin.php.
16. Kataoka, H.; Sugahara, K.; Shimano, K.; Teshima, K.; Koyama, M.; Fukunari, A.;
Chiba, K. Cell. Mol. Immunol. 2005, 2, 439.
17. Brode, S.; Raine, T.; Zaccone, P.; Cooke, A. J. Immunol. 2006, 177, 6603.
18. (a) Bendele, A. J. Musculoskelet. Neuron. Interact. 2001, 1, 377; (b) Matsuura, M.;
Imayoshi, T.; Okumoto, T. Int. J. Immunopharmacol. 2000, 22, 323.
19. Forrest, M.; Sun, S.; Hajdu, R.; Bergstrom, J.; Card, D.; Doherty, G.; Hale, J.;
Keohane, C.; Meyers, C.; Milligan, J.; Mills, S.; Nomura, N.; Rosen, H.;
Rosenbach, M.; Shei, G.; Singer, I.; Tian, M.; West, S.; White, V.; Xie, J.; Proia,
R.; Mandala, S. J. Pharmacol. Exp. Ther. 2004, 309, 758.