Identification of 1S,2R-milnacipran analogs as potent norepinephrine and serotonin transporter inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.04.025

A series of milnacipran analogs were synthesized and studied as monoamine transporter inhibitors, and several potent compounds with moderate lipophilicity were identified from the 1S,2R-isomers. Thus, 15l exhibited IC₅₀ values of 1.7 nM at NET and 25 nM at SERT, which were, respectively, 20- and 13-fold more potent than 1S,2R-milnacipran 1-II.

Full text of DOI:10.1016/j.bmcl.2008.04.025

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3328–3332
Identification of 1S,2R-milnacipran analogs as
potent norepinephrine and serotonin transporter inhibitors
Junko Tamiya,a Brian Dyck,a Mingzhu Zhang,a Kasey Phan,a Beth A. Fleck,b
Anna Aparicio,d Florence Jovic,a Joe A. Tran,a Troy Vickers,a Jonathan Grey,a
Alan C. Fosterc and Chen Chena,*
aDepartment of Medicinal Chemistry, Neurocrine Bioscience, Inc., 12790 El Camino Real, San Diego, CA 92130, USA
bDepartment of Pharmacology, Neurocrine Bioscience, Inc., 12790 El Camino Real, San Diego, CA 92130, USA
cDepartment of Neuroscience, Neurocrine Bioscience, Inc., 12790 El Camino Real, San Diego, CA 92130, USA
dDepartment of Preclinical Development, Neurocrine Bioscience, Inc., 12790 El Camino Real, San Diego, CA 92130, USA
Received 24 March 2008; revised 10 April 2008; accepted 10 April 2008
Available online 15 April 2008
Abstract—A series of milnacipran analogs were synthesized and studied as monoamine transporter inhibitors, and several potent
compounds with moderate lipophilicity were identified from the 1S,2R-isomers. Thus, 15l exhibited IC50 values of 1.7 nM at
NET and 25 nM at SERT, which were, respectively, 20- and 13-fold more potent than 1S,2R-milnacipran 1–II.
Ó 2008 Elsevier Ltd. All rights reserved.
The antidepressant milnacipran (1, Fig. 1),1 marketed as
a racemic mixture, is a hydrophilic molecule (logD
ꢀ 0), and in this respect, differs from many CNS drugs
such as atomoxetine 2 (logD = 0.8).2 Because of its
low molecular weight and low lipophilicity, milnacipran
exhibits almost ideal pharmacokinetics in humans, such
as high oral bioavailability of ꢀ85%, low inter-subject
variability, limited liver enzyme interaction, moderate
tissue distribution, and a reasonably long elimination
half-life of ꢀ8 h.3 Its lack of potential for drug–drug
interaction via cytochrome P450 enzymes is quite attrac-
tive because many CNS drugs are highly lipophilic and
rely heavily on liver enzymes for elimination.4
(Fig. 1), and its ratio at these two transporters is
reported to be about 3:1.7 Milnacipran is currently in
phase III clinical trials for fibromyalgia, and recent
reports have suggested a significant efficacy.8 The SAR
of milnacipran and its analogs based on in vivo efficacy
was reported by Bonnaud and coworkers in 1987.9 We
have described the SAR of a series of N-alkyl and dial-
kyl amides, and potent analogs such as 3a and 3b were
discovered.10 Very recently, Roggen et al. reported the
synthesis and SAR studies of a series of milnacipran
analogs as single stereoisomers with a variation in the
aromatic moiety.11 Here, we report our continued efforts
to discover potent 1S,2R-milnacipran analogs without a
significant change in lipophilicity.
The mechanism of action of milnacipran is believed to
inhibit the monoamine uptake by the norepinephrine
transporter (NET) and the serotonin transporter
(SERT),5 and milnacipran has a negligible activity at
the dopamine transporters (DAT) and many mono-
amine receptors.6 However, milnacipran has only a
moderate potency at both human NET and SERT
The milnacipran derivatives 6–8 were synthesized by a
cyclization of the allyl esters 412 to give the lactones 5,
which were elaborated to the desired products as a pair
of enantiomers using a procedure similar to that for mil-
nacipran (Scheme 1).13
The amide analogs of milnacipran 9–15 were prepared
from phenylacetonitriles 16 and (R)-(ꢁ)-epichlorohy-
drin using a reported stereo-selective synthesis11,13 or a
modified procedure (NaHMDS/THF, Scheme 2).14
Keywords: Milnacipran; Stereoisomer; Monoamine transporter; Nor-
epinephrine; Serotonin; Inhibitor; Structure–activity relationship; Li-
pophilicity; Metabolic stability; Permeability.
*
The target compounds 6–15 were tested in functional
transport assays evaluating the inhibition of human
Corresponding author. Tel.: +1 858 617 7600; fax: +1 858 617 7967;
e-mail: cchen@neurocrine.com
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.04.025

J. Tamiya et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3328–3332
3329
3
O
NH2
O
NH2
O
1
2
O
NH2
N
N
N
H
N
(+/-) 3b
(+/-) 3a
Atomoxetine (2)
Milnacipran (+/- 1)
NET IC50 = 77 nM
SERT IC50 = 420 nM
DAT IC50 = 6,100 nM
NET IC50 = 4.4 nM
SERT IC50 = 13 nM
DAT IC50 = 3,900 nM
NET IC50 = 14 nM
NET IC50 = 5.1 nM
SERT IC50 = 190 nM
DAT IC50 = 3,100 nM
SERT IC50 = 280 nM
DAT IC50 >10,000 nM
Figure 1. Chemical structures of atomoxetine, milnacipran and its potent analogs.
R3'
R3
R2
R3'
R3
R2
N2
R2
R3
R3'
a
b
O
O
NH2
O
N
O
O
4
(+/-) 6: R3 = R3' = H, R2 = Me
(+/-) 7a: R3 = R3' = Me, R2 = H
(+/-) 8a: R2 = R3' = H, R3 = Ph
(+/-) 5
Scheme 1. Reagents and conditions: (a) Rh(OAc)2/CH2Cl2/reflux, 3 h; (b) i—potassium phthalimide/DMF/140 °C, 16 h; ii—SOCl2/CH2Cl2/rt, 2 h;
iii—Et2NH/CH2Cl2/0 °C to rt, 16 h; iv—NH2NH2/EtOH/rt, 16 h.
R1
R1
a
b
CN
R1
O
NH2
O
R5
N
O
R4
16
17 (1S,2R)
9-15 (1S,2R)
Scheme 2. Reagents and conditions: (a) i—NaNH2/toluene or NaHMDS/THF/0 °C; ii—( R)-(ꢁ)-epichlorohydrin/0 °C to rt, 16 h; iii—KOH/EtOH/
4
reflux, 8 h; iv—12N HCl/0 °C to rt, 2.5 h; (b) i—potassium phthalamide/DMF; ii—SOCl2/reflux; iii—R R5NH/CH2Cl2/0 °C to rt, 16 h; iv—NH NH2/
2
EtOH/rt, 16 h.
NET, SERT, and DAT using a procedure similar to that
described by Owens et al.7,15 These results are summa-
rized in Tables 1 and 2.
Table 1. Inhibition (IC50, nM) of monoamine transporters by the
cyclopropane-substituted milnacipran analogs 6–8a
R3'
R3
Since the conformation of milnacipran (1) is an
important part of its pharmacophore, we first exam-
ined the substitution at the cyclopropane. Introducing
a 2-methyl group (compound 6) to milnacipran almost
abolished its potency at both NET and SERT (Table
1). This result was somewhat of a surprise since the
active pharmacophore based on our previous studies
showed that the amino nitrogen is located under the
cyclopropane ring,10 and this additional methyl group
might favor this conformation. One possible explana-
tion is that this cis-2-methyl group limits the rotation
of the 1-phenyl ring to a preferred orientation (Fig. 2).
The 3,3-dimethyl derivative of milnacipran (7a)
also possessed poor potency at the two transporters.
In this case, the 30-methyl group might prevent the
N-alkyl group of 7 from getting close to the 1-phenyl
R2
(+/-) 1: R2 = R3 = R3' = H
(+/-) 6: R2 = Me, R3 = R3' = H
(+/-) 7: R2 = H, R3 = R3' = Me
(+/-) 8: R2 = R3' = H, R3 = Ph
O
NH2
N R5
R4
Compound R4NR5
NET
77
SERT
420
DAT
1
NEt2
NEt2
NEt2
6100
>10,000
6
>10,000 6900
7a
7b
7c
8a
8b
8c
>10,000 >10,000 4500
>10,000 >10,000
>10,000 >10,000 >10,000
EtNCH2CH@CH2 6400
Indolin-1-yl
NEt2
>10,000 6400
>10,000
>10,000
4300
EtNCH2CH@CH2 >10,000 3400
Indolin-1-yl >10,000 560
a Data are the average of two or more independent measurements.

3330
J. Tamiya et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3328–3332
ring, which is a pharmacophore feature of this series
of compounds.10
disrupted in 7. Similarly, the 2-phenyl analogs 8a–c were
also poorly active.
The N-allyl-N-ethyl amide 3a is about 6-fold more po-
tent at NET than milnacipran, and indoline 3b has over
30-fold improvement at SERT. However, these func-
tionalities as the amide side-chain of 7 had little effect
on potency (compounds 7b and 7c), indicating that the
conformation required by the pharmacophore of 3 is
Roggen et al. have shown that the NET activity of mil-
nacipran resides in the 1S,2R-isomer,11 and substitu-
tion at the phenyl group of the active 1S,2R-isomer
has a negative impact in NET potency, although
replacement of the phenyl group of the much less ac-
tive 1R,2S-isomer with a lipophilic 3,4-dichlorophenyl
Table 2. Inhibition (IC50, nM) of monoamine transport by the amides 9–15a
S
O
Ar
Ar =
2R
O
NH2
F
1S
1R
2S
9
10
14
11
15
O
NH2
R5
N
R4
1-II, 9-15
O
O
O
O
O
N
1-I
12
13
Compound
R4NR5
NET
SERT
DAT
clogPc
1–I
1–II
9
NEt2
NEt2
NEt2
NEt2
NEt2
NEt2
5500
40
3500
320
370
1100
280
2100
430
1700
270
130
145
210
460
480
480
150
58
>10,000
3200
1.2
1.2d
1.3
3.4
1.7
1.2
1.5
2.0
1.9
1.4
0.9
1.9
2.0
2.4
1.1e
1.4
1.8
0.8
1.8
2.5
2.5
1.9
2.1
2.4
0.9
1.2
1.9
1.3
0.8
0.6
1.1
1.4
2.5
2.1
1.7
0.7
3.3f
73
47
>10,000
1300
10
11b
12ab
12b
12c
12d
12e
12f
12g
12h
13
47
>10,000
>10,000
>10,000
>10,000
>10,000
1>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
4300
160
18
EtNCH2CH@CH2
PrNCH2CH@CH2
N(CH2CH@CH2)2
CH2@CHCH2NCH2C„CH
N(CH2C„CH)2
CH2@CHCH2NCH2C„CMe
PrNCH2C„CMe
NEt2
21
4.5
14
17
40
80
160
12
14ab
14b
14c
14d
14e
14f
14g
14h
14i
NEt2
EtNCH2CH@CH2
N(CH2CH@CH2)2
N(CH2C„CH)2
CH2@CHCH2NCH2C„CMe
1-Indolinyl
7.1
7.0
7.6
49
64
15
>10,000
640
5100
4.2
5.8
28
4-(1,4-Benzoxazinyl), 3,4-dihydro-2H
Isoindolin-1-yl
320
77
20
3600
>10,000
3600
5-Thieno[3,2-c]pyridinyl, 4,5,6,7-tetrahydro
2-Isoquinolinyl, 1,2,3,4-tetrahydro
NEt2
9.4
6.5
6.3
2.9
5.2
7.1
5.9
4.4
3.9
8.1
3.2
4.0
3.5
1.7
5.1
14
26
14j
15ab
15bb
15c
15d
15e
15f
15g
15h
15i
140
65
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
2,900
EtNCH2CH@CH2
EtNCH2CF@CH2
EtNCH2Pr-c
112
540
125
160
210
100
39
EtNCH2C„CH
cPrNCH2C„CH
cPrCH2NCH2CH„CH
MeNCH2CF@CH2
CH2@CHCH2NCH2CF@CH2
CH„CCH2NCH2CF@CH2
N(CH2CH@CH2)2
15j
54
32
>10,000
6300
15k
15l
N(CH2C„CH)2
25
190
>10,000
3100
2
a Data are the average of two or more independent measurements.
b The ee value was determined using a chiral HPLC method: 11 (91.6%), 12a (88.7%), 14a (92.4%), 15a (95.6%), and 15b (96.4%).
c Calculated using ACD software.
d Measured logP was 1.6.
e Measured logP was 1.7.
f Measured logP was 3.5.

J. Tamiya et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3328–3332
3331
= 3.3). Tetrahydrothienopyridine 14i exhibited almost
a 1:1 ratio between NET and SERT, although it was
slightly more lipophilic (clogP = 2.1) than milnacipran.
None of the compounds showed significant activity at
DAT.
Compounds 14a, 15a, 15b, 15k, and 15l were also tested
in an in vitro human liver microsomal assay17 to com-
pare their metabolic stability with that of milnacipran
and atomoxetine. The scaled systemic clearance (CLsys
)
of the 3,4-methylenedioxyphenyl compound 14a was
7.9 mL/min kg which was slightly higher than 1–II
(CLsys = 5.0 mL/min kg). In comparison, the 3,4-ethy-
lenedioxyphenyl analog 15a (CLsys = 4.4 mL/min kg)
was comparable to 1–II. In contrast, atomoxetine 2,
which has high plasma protein binding18 and is mainly
eliminated as metabolites in humans,19 exhibited high
clearance (14.3 mL/min kg) in this assay. The potent
analogs 15 k and 15l exhibited only slightly higher met-
abolic clearance than 1–II, suggesting these compounds
might still be, at least in part, eliminated by renal clear-
ance in humans20 since their lipophilic profiles were sim-
ilar to milnacipran.
Figure 2. A low-energy conformation of 6. If the phenyl ring rotates
about 90°, its ortho-proton may clash with the 2-methyl group,
preventing its free rotation.
or 2-naphthyl moiety increases its potency at all the
three transporters up to 20-fold. We explored a set of
oxygen-containing substituted phenyl groups of the ac-
tive 1S,2R-milnacipran (1–II) in an effort to improve
the potency while still maintaining the low lipophilicity
of these analogs. In our assay, the 1R,2S-isomer (1–I)
had a negligible activity on all the three transporters,
and 1S,2R-milnacipran (1–II) exhibited IC50 values of
40 and 320 nM, respectively, at NET and SERT (Table
2). Compound 1–II was only weakly active at DAT
(IC50 = 3200 nM).16 In comparison, the 3-fluorophenyl
analog 9 slightly reduced the potency at both NET
and SERT. The more lipophilic benzofuran 11 (clogP
= 1.7) exhibited a similar pharmacological profile com-
pared to 1–II (clogP = 1.2), while benzothiophene 10
was less active at SERT, and showed some activity at
DAT. Dihydrobenzofuran 12a had a clogP value sim-
ilar to 1–II, but was much less potent at both NET and
SERT. Benzocyclopentane 13 showed a moderate po-
tency, while the 3,4-methylenedioxyphenyl analog
(14a, NET IC50 = 12 nM) had improved NET activity,
indicating an important role of the 3-oxygen in 14a for
NET. Finally, the 3,4-ethylenedioxyphenyl 15a exhib-
ited about 6-fold higher NET potency than 1–II. More
importantly, compounds 14a and 15a were not much
more lipophilic than milnacipran. For example, 14a
and 1–II had measured logP values of 1.7 and 1.6,
respectively.
The renal clearance of milnacipran (CLR ꢀ 350 mL/
min)20 in healthy human subjects is about 3-times higher
than the glomerular filtration rate, indicative of an ac-
tive secretion process possibly caused by P-glycoprotein
(P-gp) activity.21 In an in vitro Caco-2 assay, 1–II dis-
played a moderate permeability (Papp = 51 nm/s) and a
possible efflux mechanism reflected by its higher perme-
ability from the basolateral to apical direction compared
to the reverse direction [(b to a)/(a to b) = 7.4]. In com-
parison, the methylenedioxy and ethylenedioxy analogs
14 and 15 also showed high efflux ratios, possibly due
to their high polar surface area (PSA). For example,
2
2
˚
˚
14a (PSA = 65 A ) and 15l (PSA = 65 A ) showed Papp
values of 47 and 52 nm/s, respectively, and efflux ratios
of 6 and 15, which were similar to that of 1–II
2
˚
(PSA = 46 A ). The strong activity of these com-
pounds as P-gp substrates should facilitate active
tubular secretion for drug elimination, although the
efflux mechanism might also limit brain penetration of
its substrates22.
In summary, a series of milnacipran analogs were stud-
ied for their structure–activity relationships as inhibitors
of NET and SERT. Many potent compounds were iden-
Table 3. Metabolic stability of 14a, 15a, 15b, 15k and 15l in vitroa
Compound
CLsys (mL/min kg)
1–IIb
14ac
15a
15b
15k
15l
5.0
7.9
Previously we have found that replacing one of the ethyl
groups of milnacipran with an allyl moiety increases
its NET potency by about 5-fold (3a, NET
IC50 = 14 nM).10 We utilized this information to explore
the SAR of three hydrophilic aryl analogs (12, 14, and
15), and the results are summarized in Table 3. Many
potent NET inhibitors were discovered and their ratios
to the SERT activity varies from 1- (14i) to almost 80-
fold (15d). Their lipophilicities were not much higher
than 1–II but lower than that of atomoxetine 2 (clogP
4.4
8.8
8.1
7.9
2d
14.3
a See Ref. 16 for assay conditions.
b Measured logD=ꢁ0.2 and pKa = 9.6.
c Measured logD=0.4
d Measured logD=0.8 and pKa = 10.1.

3332
J. Tamiya et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3328–3332
12. (a) Doyle, M. P.; Davies, S. B.; Hu, W. Org. Lett. 2000, 2,
1145; (b) Doyle, M. P.; Wu, W. Adv. Synth. Catal. 2001,
343, 299.
tified from the 1S,2R-isomers with a range of NET/
SERT ratios and they possessed moderate lipophilicity
similar to milnacipran. For example, 15l exhibited an
IC50 value of 1.7 nM at NET, which was 20-fold more
potent than 1S,2R-milnacipran (1–II), and only slightly
lower metabolic stability in human liver microsomes.
Compound 15a had a similar in vitro pharmacological
profile to atomoxetine 2 but was more metabolically sta-
ble. These results indicate that compounds such as 15a
may have an elimination profile not much different than
milnacipran in humans. Further studies on the pharma-
cokinetic characterization including brain penetration of
this series of compounds will be published in due course.
13. Shuto, S.; Ono, S.; Hase, Y.; Kamiyama, N.; Takada, H.;
Yamasihita, K.; Matsuda, A. J. Org. Chem. 1996, 61, 915.
14. A typical procedure for the synthesis of lactone 17. To a
stirred solution of NaHMDS (1 M in THF, 12.6 mL,
12.6 mmol) at 0 °C was added dropwise a solution of 3,4-
(ethylenedioxy)phenylacetonitrile (16 h, 1.0 g, 5.7 mmol)
in THF (4 mL) over 30 min. R-(ꢁ)-Epichlorohydrin (97%
ee, 0.4 mL, 5.7 mmol) in THF (4 mL) was added dropwise.
The mixture was stirred at 0 °C for 2 h, then at rt
overnight. KOH (1 N, 1 mL) and EtOH (3 mL) were
added and the solvents were evaporated in vacuo. The
residue was diluted with KOH (1 N, 3 mL) and EtOH
(8 mL) and the resulting mixture was refluxed for 8 h.
After cooled to 0 °C, the reaction mixture was treated with
concentrated HCl to pH ꢀ 1, and stirred at rt for 2 h. The
mixture was concentrated in vacuo and the residue was
partitioned between water and ethyl acetate. The aqueous
layer was extracted with ethyl acetate and the combined
extracts were dried over MgSO4, filtered, and concentrated
in vacuo. The crude product was purified by chromatog-
raphy on silica gel using 3:1 hexane/EtOAc to afford
lactone 17h (500 mg, 38% yield). For a recently published
procedure, see: Xu, F.; Murry, J. A.; Simmons, B.; Corley,
E.; Fitch, K.; Karady, S.; Tschaen, D. Org. Lett. 2006, 8,
3885.
15. Owens, M. J.; Morgan, W. N.; Plott, S. J.; Nemeroff, C. B.
J. Pharmacol. Exp. Ther. 1997, 283, 1305, (for experimen-
tal detail, see WO 07098356).
16. Following IC50 values for the synaptosomal uptake
inhibition of this isomer are reported by Roggen: 7 nM
(NET), 120 nM (SERT) and 10,000 nM (DAT), see Ref.
11.
17. For assay conditions, see: Guo, Z.; Zhu, Y. F.; Gross, T.
D.; Tucci, F. C.; Gao, Y.; Moorjani, M.; Connors, P. J.,
Jr.; Rowbottom, M. W.; Chen, Y.; Struthers, R. S.; Xie,
Q.; Saunders, J.; Reinhart, G.; Chen, T. K.; Bonneville, A.
L.; Chen, C. J. Med. Chem. 2004, 47, 1259.
18. Chalon, S. A.; Desager, J. P.; Desante, K. A.; Frye, R. F.;
Witcher, J.; Long, A. J.; Sauer, J. M.; Golnez, J. L.; Smith,
B. P.; Thomasson, H. R.; Horsmans, Y. Clin. Pharmacol.
Ther. 2003, 73, 178.
Acknowledgment
The authors thank Mr. Rajesh Huntley for his technical
assistance in this study.
References and notes
1. (a) Preskorn, S. H. J. Psychiatr. Pract. 2004, 10, 119; (b)
Spencer, C. M.; Wilde, M. I. Drugs 1998, 56, 405.
2. Calculated logP values were 1.2 for milnacipran and 3.3
for atomoxetine using ACD software.
3. (a) Puozzo, C.; Leonard, B. E. Int.. Clin. Psychopharma-
col. 1996, 11, 15; (b) Puozzo, C.; Panconi, E.; Deprez, D.
Int. Clin. Psychopharmacol. 2002, 17, S25.
4. (a) Puozzo, C.; Lens, S.; Reh, C.; Michaelis, K.; Rosillon,
D.; Deroubaix, X.; Deprez, D. Clin. Pharmacokinet. 2005,
44, 977; (b) Puozzo, C.; Hermann, P.; Chassard, D. Int.
Clin. Psychopharmacol. 2006, 21, 153.
5. Briley, M. CNS Drugs Rev. 1998, 4, 137.
6. Mochizuki, D.; Tsujita, R.; Yamada, S.; Kawasaki, K.;
Otsuka, Y.; Hashimoto, S.; Hattori, T.; Kitamura, Y.;
Miki, N. Psychopharmacology 2002, 162, 323.
7. Vaishnavi, S. N.; Nemeroff, C. B.; Plott, S. J.; Rao, S.
G.; Kranzler, J.; Owens, M. J. Biol. Psychiatry 2004, 55,
320.
8. For recent review articles, see (a) Rooks, D. S. Curr. Opin.
Rheumatol. 2007, 19, 111; (b) Leo, R. J.; Brooks, V. L.
Curr. Opin. Investig. Drugs 2006, 7, 637.
19. Sauer, J. M.; Ponsler, G. D.; Mattiuz, E. L.; Long, A. J.;
Witcher, J. W.; Thomasson, H. R.; Desante, K. A. Drug
Metab. Dispos. 2003, 31, 98.
9. Bonnaud, B.; Cousse, H.; Mouzin, G.; Briley, M.; Stenger,
A.; Fauran, F.; Couzinier, J. P. J. Med. Chem. 1987, 30,
318.
20. Puozzo, C.; Pozet, N.; Deprez, F.; Baille, P.; Ung, H. L.;
Zech, P. Eur. J. Drug Metab. Pharmacokinet. 1998, 23,
280.
10. Chen, C.; Dyck, B.; Fleck, B. A.; Foster, A. C.; Grey, J.;
Jovic, F.; Mesleh, M.; Phan, K.; Tamiya, J.; Vickers, T.;
Zhang, M. Bioorg. Med. Chem. Lett. 2008, 18, 1346.
11. Roggen, H.; Kehler, J.; Stensbol, T. B.; Hansen, T. Bioorg.
Med. Chem. Lett. 2007, 17, 2834.
21. Shitara, Y.; Horie, T.; Sugiyama, Y. Eur. J. Pharm. Sci.
2006, 27, 425.
22. Linnet, K.; Ejsing, T. B. Eur. Neuropsychopharmacol.
2008, 18, 157.