Dual serotonin transporter inhibitor/histamine H3 antagonists: Development of rigidified H3 pharmacophores

Article abstract of DOI:10.1016/j.bmcl.2007.08.017

A series of tetrahydroisoquinolines acting as dual serotonin transporter inhibitor/histamine H₃ antagonists is described. The introduction of polar aromatic spacers as part of the histamine H₃ pharmacophore was explored. A convergent synthesis of the final products allowing late stage introduction of the aromatic side chain was developed. In vitro and in vivo data are discussed.

Full text of DOI:10.1016/j.bmcl.2007.08.017

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5325–5329
Dual serotonin transporter inhibitor/histamine H3 antagonists:
Development of rigidified H3 pharmacophores
John M. Keith,* Ann J. Barbier, Sandy J. Wilson, Kirstin Miller, Jamin D. Boggs,
Ian C. Fraser, Curt Mazur, Timothy W. Lovenberg and Nicholas I. Carruthers
Johnson & Johnson Pharmaceutical Research and Development L.L.C., 3210 Merryfield Row, San Diego, CA 92121, USA
Received 10 July 2007; revised 6 August 2007; accepted 8 August 2007
Available online 15 August 2007
Abstract—A series of tetrahydroisoquinolines acting as dual serotonin transporter inhibitor/histamine H3 antagonists is described.
The introduction of polar aromatic spacers as part of the histamine H3 pharmacophore was explored. A convergent synthesis of the
final products allowing late stage introduction of the aromatic side chain was developed. In vitro and in vivo data are discussed.
Ó 2007 Elsevier Ltd. All rights reserved.
Fatigue is a frequent symptom experienced by the more
than 340 million people worldwide who are suffering
from depression.1 While antidepressants, particularly
selective serotonin reuptake inhibitors (SSRIs), are fre-
quently able to improve the overall sense of well being
for those who use them, these drugs often fail to im-
prove the symptom of fatigue even as mood improves.2,3
Some SSRIs even induce fatigue and excessive
sleepiness.4,5
explores the effect of replacing the flexible propyloxy lin-
ker common to the earlier series with a more rigid and
polar aromatic spacer in an effort to expand understand-
ing of what is tolerated by the histamine H3 receptor.12
Increasing molecular rigidity has also been associated
with improved oral absorption.13 The target structure
SMe
One possible approach to mitigating the fatigue associ-
ated with depression and/or its treatment is through
the use of a wake promoting agent. Histamine H3 recep-
tor antagonists are known to increase wakefulness6
without showing nonspecific stimulant effects such as in-
creased locomotor activity.7 Thus the case can be made
that H3 antagonists would be useful adjuncts to antide-
pressant therapy.
N
N
O
Me
N
N
O
Me
O
1
2
hSERT Ki = 25.7 nM
hH3 Ki = 0.72 nM
hSERT Ki = 5.2 nM
hH3 Ki = 2.5 nM
As part of our strategy for the development of novel ap-
proaches to the treatment of depression, we have inves-
tigated the possibility of combining histamine H3
antagonism with serotonin transporter (SERT) inhibi-
tion in a single chemical entity. We have recently de-
scribed several chemical series8–11 with high affinities
for both targets (Fig. 1). The research described herein
OMe
OMe
N
N
N
N
N
O
Me
O
Me
3
4
Keywords: Serotonin; Histamine H3; SERT; Tetrahydroisoquinoline.
*
Corresponding author. Tel.: +1 858 784 3275; e-mail: jkeith@
prdus.jnj.com
hSERT Ki = 5.1 nM
hH3 Ki = 2.0 nM
hSERT Ki = 107 nM
hH3 Ki = 12 nM
Address: EnVivo Pharmaceuticals, Inc., 480 Arsenal Street, Building
1, Watertown, MA 02472, USA.
Figure 1. Dual histamine H3/SERT inhibitors.
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.08.017

5326
J. M. Keith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5325–5329
SMe
SMe
H
N
(a)
O
(b,c)
HO
HO
Me
OH
Me
10
N
HO
11
(f)
N
N
OH
H
N
O
Me
Me
O
Figure 2. Target structure type with embedded H3 pharmacophore
shown in bold.
12
MeS
Br
(d,e)
+
MeS
OH Me OH
N
type with the embedded H3 pharmacophore14 is shown
in Figure 2.
MeS
SMe
13
The initial linear synthetic route to the desired targets is
shown in Scheme 1. Reductive amination of 6-bromo-2-
pyridinecarboxaldehyde with piperidine gave amino
bromide 5, which was converted to biaryl ether 6 via
an SNAr reaction15 with 3-hydroxylbenzyl alcohol.
Swern oxidation of the alcohol followed by reductive
amination with methylamine gave the benzylic amine 7
in good yield. Alkylation of 7 with 4-thiomethyl-20-
bromoacetophenone followed by reduction16 of the ke-
tone with NaBH4 afforded the secondary alcohol 8.
Ring closure was affected by heating the alcohol in neat
methanesulfonic acid giving the final product 9 in good
yield. Other substrates similar to 8 gave a mixture of 5-
and 7-substituted isomers (see Schemes 1 and 2).
SMe
SMe
OH
(g)
11
+
N
N
HO
Me
Me
14
15
Scheme 2. (a) 2.1 equiv 40%(aq) MeNH2, 1.9 equiv NaBH4, 0.5 M
MeOH, 0 °C, 3.5 h, 97%; (b) 1 equiv 4-thiomethyl-20-bromoacetoph-
enone, 3 equiv DIPEA, 0.1 M CH3CN, 1 h; (c) 2 equiv NaBH4, 0.1 M
EtOH, 16% for two steps; (d) 20 equiv 40%(aq) MeNH2, 0.1 M EtOH,
1 h, then (e) 3 equiv NaBH4, 15 h, 78% for two steps; (f) 1.3 equiv 3-
hydroxybenzaldehyde, 2.5 equiv NaB(OAc)3H,
1 equiv HOAc,
0.25 M, THF, 0 °C rt, 18 h, 94%; (g) 5 mL MSA/g 11, 50 °C, 2.5 h,
67% 7-isomer, 31% 5-isomer, 98% total yield.
While the route shown in Scheme 1 is useful if variation
of the pendant aryl ring at position four is desired, it is
somewhat cumbersome if variation of the H3 side chain
is the objective. From previous work, we found tetrahy-
droisoquinolines with a 4-thiomethylphenyl substituent
in the 4-position yielded potent SERT inhibitors.17
Keeping the 4-thiomethylphenyl substituent constant
may allow for a simplified synthetic route enabling more
rapid evaluation of potential H3 pharmacophores.
SMe
X
X
(a or b)
+
14
N
Ar
N
Br
Ar
N
O
Me
Ar = pyridyl, phenyl or thiazole
X = CH2, CHF, or O
Scheme 3. (a) 1.25 equiv 14, 2 equiv Cs2CO3, 0.5 M NMP, 150 °C, 2 h;
(b) 1.5 equiv 14, 2 equiv Cs2CO3, 0.5 equiv CuI, 0.75 M NMP, 195 °C,
2 h.
(c,d)
(a)
(b)
O
N
N
OH
N
Br
N
Br
N
O
5
6
SMe
As we wished to avoid the use of protecting groups in
the synthesis of the tetrahydroisoquinoline core, we
envisioned 10 as a useful intermediate. Compound
10 was readily preparable via reductive amination,
but was troublesome to work with due to high water
solubility18 and modest solubility in organic solvents.
Nevertheless, it could be converted to 11 via alkyl-
ation with 4-thiomethyl-20-bromoacetophenone and
reduction of the corresponding ketone, albeit in mod-
est yield.
(e,f)
(g)
OH
Me
H
N
N
N
N
N
O
Me
N
O
7
8
SMe
5
4
9c
N
N
N
O
Me
7
Recognizing the high polarity of 10 as the source of dif-
ficulty with its manipulation, we desired an alternate
route to 11 with more lipophilic intermediates. Alkyl-
ation of methylamine with 4-thiomethyl-20-bromoace-
tophenone followed by immediate reduction of the
ketone (one pot) gave 12 in reasonable yield, but also
afforded significant amounts of the dialkylated product
13. Increasing the equivalents of methylamine from five
Scheme 1. (a) 1.5 equiv NaB(OAc)3H, 1.0 equiv piperidine, 0.5 M 1,2-
DCE, 71%; (b) 2 equiv 3-hydroxybenzyl alcohol, 2 equiv Cs2CO3,
0.5 M DMSO, 150 °C, 14 h, 59–63%; (c) 1.2 equiv oxalyl chloride, 2.2
equiv DMSO, 5 equiv Et3N, 0.1 M CH2Cl2, ꢀ78 °C, 3 h, 86%; (d) 2.1
equiv 40%(aq) MeNH2, 1.9 equiv NaBH4, 0.5 M MeOH, 0 °C, 3.5 h,
98%; (e) 1 equiv 4-thiomethyl-20-bromoacetophenone, 3 equiv DIPEA,
0.1 M CH3CN, 1 h; (f) 2 equiv NaBH4, 0.1 M EtOH, 80% for two
steps; (g) 1.0 M MSA, 5 h, 60 °C, 64%.

J. M. Keith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5325–5329
5327
7
Table 1. Rat and human SERT19 and human histamine H3 in vitro binding affinities for compounds 9a–m and 16a–c
SMe
SMe
sidechain
N
N
sidechain
Me
Me
9a-m
16a-c
Compound
Side chain
rSERT Ki (nM)
hSERT Ki (nM)
hH3 Ki (nM)
9a
11.7 ( 3)
22.0 ( 0.7)
51 ( 11)
N
N
O
O
O
O
O
9b
16a
9c
11.7 ( 5.8)
403 ( 38)
25.7 ( 7.3)
182 ( 40)
9.7 ( 3.5)
174 ( 19)
8.7 ( 2.7)
4.3 ( 1)
17.7 ( 0.8)
616 ( 123)
10.4 ( 1.0)
157 ( 22)
8.3 ( 2.4)
262 ( 14)
2.4 ( 0.7)
3.8 ( 0.5)
8.0 ( 1.8)
2.3 ( 0.2)
7.9 ( 1.6)
6.5 ( 1.8)
4.3 ( 1.5)
15.5 ( 4.3)
3.2 ( 0.5)
18.0 ( 0.7)
253 ( 36)
323 ( 111)
265 ( 70)
109 ( 18)
511 ( 20)
34.3 ( 5.7)
22.7 ( 2.5)
24.3 ( 8.6)
8.2 ( 0.5)
260 ( 36)
110 ( 10)
20.7 ( 4.7)
201 ( 27)
20.7 ( 2.3)
N
N
N
N
N
N
16b
9d
16c
9e
N
N
N
O
N
N
O
N
O
O
9f
N
N
N
O
O
O
N
9g
9h
9i
7.5 ( 1.8)
4.5 ( 1)
N
N
O
F
11 ( 1.9)
14.3 ( 2.3)
5.2 ( 0.7)
14 ( 3.9)
4.0 ( 0)
N
N
O
9j
N
S
N
O
9k
9l
N
S
O
O
N
N
S
N
N
9m
O
to 20 improved the yield of 12 from 45% to 78%. Reduc-
tive amination of 3-hydroxybenzaldehyde with 12 gave
11 in 94% yield. MSA mediated ring closure of 11 gave
14 and 15 in a 2:1 ratio with a combined yield of 98%.
Compounds 14 and 15 were easily separated
chromatographically.

5328
J. M. Keith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5325–5329
The side chains to be attached to 14 were prepared sim-
ply by reductive amination of the precursor bromo-alde-
hyde with the desired amine (Scheme 3). Attachment of
the side chains onto 14 was accomplished via SNAr or by
coupling in the presence of CuI. The in vitro activity for
the products 9a–m and 16a–c is shown in Table 1.
References and notes
1. Jane-Llopis, E.; Hosman, C.; Jenkins, R.; Anderson, P.
Br. J. Psychiatry; J. Mental Sci. 2003, 183, 384.
2. Nierenberg, A. A.; Keefe, B. R.; Leslie, V. C.; Alpert, J.
E.; Pava, J. A.; Worthington, J. J., 3rd; Rosenbaum, J. F.;
Fava, M. J. Clin. Psychiatry 1999, 60, 221.
3. Fava, G. A.; Fabbri, S.; Sonino, N. Prog. Neuro-Psycho-
pharmacol. Biol. Psychiatry 2002, 26, 1019.
4. Beasley, C. M., Jr.; Koke, S. C.; Nilsson, M. E.; Gonzales,
J. S. Clin. Ther. 2000, 22, 1319.
From Table 1 it is apparent 5-substituted compounds
16a–c have much lower affinity for the rat and human
SERTs and for the H3 receptor than the corresponding
7-isomers 9a–m. Compounds having 1,3-disubstituted
pyridyl spacers 9a, c–e (relative spacing) were all less po-
tent than the corresponding 1,3-disubstituted phenylene
derivative 9b. Pyridine ring nitrogen position had a large
impact on H3 affinity, but a more modest impact on
hSERT binding. Greater H3 affinity was generally found
in the corresponding 1,4-disubstituted arylene deriva-
tives, but again, the presence and location of a ring
nitrogen had a significant effect on potency. Compounds
possessing either 2,4 or 2,5 substituted thiazole spacers
(9k and 9m respectively) in the side chain retained excel-
lent hSERT and reasonable H3 affinity. None of the
compounds prepared were as potent as 2 at the hista-
mine H3 receptor.
5. Zajecka, J. M. J. Clin. Psychiatry 2000, 61(Suppl. 2), 20.
6. Monti, J. M.; Jantos, H.; Boussard, M.; Altier, H.;
Orellana, C.; Olivera, S. Eur. J. Pharmacol. 1991, 205, 283.
7. Barbier, A. J.; Berridge, C.; Dugovic, C.; Laposky, A. D.;
Wilson, S. J.; Boggs, J.; Aluisio, L.; Lord, B.; Mazur, C.;
Pudiak, C. M.; Langlois, X.; Xiao, W.; Apodaca, R.;
Carruthers, N. I.; Lovenberg, T. W. Br. J. Pharmacol.
2004, 143, 649.
8. Stocking, E. M.; Miller, J. M.; Barbier, A. J.; Wilson, S. J.;
Boggs, J. D.; McAllister, H. M.; Wu, J.; Lovenberg, T. W.;
Carruthers, N. I.; Wolin, R. L. Bioorg. Med. Chem. Lett.
2007, 17, 3130.
9. (a) Keith, J. M.; Gomez, L. A.; Wolin, R. L.; Barbier,
A. J.; Wilson, S. J.; Boggs, J. D.; Mazur, C.; Fraser, I.
C.; Lord, B.; Aluisio, L.; Lovenberg, T. W.; Carruthers,
N. I. Bioorg. Med. Chem. Lett. 2007, 17, 2603; (b)
Keith, J. M.; Gomez, L. A.; Barbier, A. J.; Wilson, S.
J.; Boggs, J. D.; Lord, B.; Mazur, C.; Aluisio, L.;
Lovenberg, T. W.; Carruthers, N. I. Bioorg. Med. Chem.
Lett. 2007, 17, 4374.
We examined two of the more promising compounds, 9h
and 9k, in our H3 functional assay.7 Both compounds
were antagonists with pA2 values in the range of 7.7–7.9.
10. (a) Keith, J. M.; Gomez, L. A.; Letavic, M. A.; Ly, K. S.;
Jablonowski, J. A.; Seierstad, M.; Barbier, A. J.; Wilson,
S. J.; Boggs, J. D.; Fraser, I. C.; Mazur, C.; Lovenberg, T.
W.; Carruthers, N. I. Bioorg. Med. Chem. Lett. 2007, 17,
702; (b) Letavic, M. A.; Keith, J. M.; Jablonowski, J. A.;
Stocking, E. M.; Gomez, L. A.; Ly, K. S.; Miller, J. M.;
Barbier, A. J.; Bonaventure, P.; Boggs, J. D.; Wilson, S. J.;
Miller, K. L.; Lord, B.; McAllister, H. M.; Tognarelli, D.
J.; Wu, J.; Abad, M. C.; Schubert, C.; Lovenberg, T. W.;
Carruthers, N. I. Bioorg. Med. Chem. Lett. 2007, 17, 1047.
11. Letavic, M. A.; Keith, J. M.; Ly, K. S.; Barbier, A. J.;
Boggs, J. D.; Wilson, S. J.; Lord, B.; Lovenberg, T. W.;
Carruthers, N. I. Bioorg. Med. Chem. Lett. 2007, 17,
2566.
12. For other work involving benzyl amine H3 pharmaco-
phores, see: Miko, T.; Ligneau, X.; Pertz, H. H.; Arrang,
J.-M.; Ganellin, C. R.; Schwartz, J.-C.; Schunack, W.;
Starck, H. Bioorg. Med. Chem. 2004, 12, 2727, Earlier
work performed by our laboratories involved chemotypes
possessing benzyl amines, but these substituents were not
the dominant pharmacophoric functionality from an
activity standpoint: Apodaca, R.; Carruthers, N. I.;
Dvorak, C. A.; Rudolph, D. A.; Shah, C. R.; Xiao, W.
PCT Int. Appl. 2002, 155 pp, WO 2002012214.
Throughout our H3/SERT program, we have been able
to qualitatively correlate slow absorption into the brain
(or slow onset of pharmacological response) with brain
and tissue accumulation along with concomitantly high
volumes of distribution.20 We therefore elected to use
the 5-hydroxytryptophan (5-HTP) induced head twitch
model of SERT blockade21 to triage molecules on the
basis of desirable physical properties. Only those mole-
cules with a robust response in the head twitch model
after 1 h and little to no response after 24 h would be
profiled further. One hour post ip injection (10 mg/kg)
of 9h elicited no change in head twitch response
(HTR) as compared with control animals. After 24 h,
an HTR of 272% over control animals was observed.
These data suggest 9h inhibits the serotonin transporter,
but gets into the brain slowly either due to slow absorp-
tion through the peritoneum or slow accumulation into
the brain due to a very high volume of distribution.
Compound 9k did not show significant differences in
the HTR over controls at either time point. In an effort
to improve the physical properties of 9h, the piperidine
ring was replaced with morpholine and 4-fluoropiperi-
dine.22 When such substitutions were made in this series
(9i and 9j) an unacceptable decrease in H3 affinity was
observed.
13. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P.
J. Adv. Drug Delivery Rev. 1997, 23, 3.
14. Apodaca, R.; Dvorak, C. A.; Xiao, W.; Barbier, A. J.;
Boggs, J. D.; Wilson, S. J.; Lovenberg, T. W.; Carruthers,
N. I. J. Med. Chem. 2003, 46, 3938.
15. An SNAr reaction was not possible with some of our
examples due to electronically unfavorable substitution of
the aromatic ring. In those instances CuI was used as the
coupling agent.
16. It is important to carry out the reduction immediately due
to rapid oxidation of the amino ketone. For similar
examples: (a) Babad, E.; Carruthers, N. I.; Jaret, R. S.;
Steinman, M. Synthesis 1988, 966; (b) Lipton, S. H.;
Dutky, R. C. Chem. Ind. (London) 1972, 33.
In conclusion, we have prepared a series of novel dual
SERT inhibitor/H3 antagonists possessing rigidified H3
pharmacophores. The compounds described generally
were potent SERT inhibitors and some 9b, 9f, 9h, 9k,
and 9m were potent H3 antgonists. One compound,
9h, was modestly effective in the 5-HTP model of SERT
inhibition after 24 h, but lacked the desired physical
properties to warrant further development.

J. M. Keith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5325–5329
5329
17. See Refs. 9a and 10b.
compounds possessing histamine H3 and SERT
activity.
21. Darmani, N. A.; Reeves, S. L. Pharm. Biochem. Behav.
1996, 55, 387.
18. Compound 10 was not extracted efficiently from water
with methylene chloride or ethyl acetate, but was readily
extracted with n-butanol.
19. Carruthers, N. I.; Gomez, L. A.; Jablonowski, J. A.;
Keith, J. M.; Letavic, M. A.; Ly, K. S.; Miller, J. M. B.;
Stocking, E. M.; Wolin, R. L. PCT Int. Appl. 2006, WO
2006066197 A1.
20. While the two phenomena could not be expected to
always correlate, the trend held for quite a few dibasic
22. Previous studies by us (Ref. 9a) found replacement of the
piperidine in the H3 pharmacophore with morpholine or
4-fluoropiperidine resulted in more rapid brain penetra-
tion and subsequent elimination thus resulting in a strong
physiological response after 1 h but a return to near
baseline levels after 24 h.