Synthesis of 3,7-disubstituted imipramines by palladium-catalysed amination/cyclisation and evaluation of their inhibition of monoamine transporters

Article abstract of DOI:10.1002/chem.201100885

We describe a novel approach for the synthesis of a series of 3,7-difunctionalised symmetric and unsymmetrical analogues of the tricyclic antidepressant (TCA) imipramine, which uses a key palladium-catalysed amination/cyclisation of an ester-functionalise

Full text of DOI:10.1002/chem.201100885

DOI: 10.1002/chem.201100885
Synthesis of 3,7-Disubstituted Imipramines by Palladium-Catalysed
Amination/Cyclisation and Evaluation of Their Inhibition of Monoamine
Transporters
Helle Christensen,[a] Christina Schjøth-Eskesen,[a] Marie Jensen,[a] Steffen Sinning,[b] and
Henrik H. Jensen*[a]
Abstract: We describe a novel ap-
proach for the synthesis of a series of
3,7-difunctionalised symmetric and un-
symmetrical analogues of the tricyclic
antidepressant (TCA) imipramine,
which uses a key palladium-catalysed
amination/cyclisation of an ester-func-
tionalised dibromide. Of the ester,
methyl, hydroxymethyl and methoxy-
methyl disubstituted compounds pre-
pared, 3,7-dimethyl-imipramine was
found to be the most potent against the
human serotonin transporter (hSERT).
The inhibitory potency of 3,7-dimethyl
imipramine was found to be at least as
high as the parent imipramine. This
novel TCA also exhibits an increased
selectivity (relative to imipramine) in
binding to hSERT versus the human
norepinephrine transporter (hNET).
Even higher selectivity could be ob-
tained with 3,7-dihydroxymethyl imipr-
amine, which was found to be 167-fold
more selective for hSERT over hNET,
representative of a 120-fold gain in se-
lectivity relative to the parent imipra-
mine. These results further validate our
previous model for the binding of
imipramine and high-affinity analogues
of imipramine to the central binding
site of hSERT.
Keywords: amination · human se-
rotonin transporter (hSERT) · re-
uptake inhibitor · palladium · tricy-
clic antidepressants
Introduction
tonin transporter (hSERT). Imipramine (1) and its metabo-
lites can also be toxic,[2] furthermore, 1 has side effects at-
Depression is a serious disease that results in a significantly
reduced quality of life for the patient. It is expected to
become the second most costly disease to society by 2020
and is amongst the top five leading causes of disability and
disease throughout the world, with treatment options being
relatively new.[1] In the 1950s a revolution in psychopharma-
cology and psychiatry took place with the introduction of
the first types of antidepressants.[2] One of the earliest drugs
was the tricyclic antidepressant (TCA) imipramine (1,
Scheme 1a).[3] To this day, imipramine is considered the
“gold-standard” antidepressant because its ability to lift the
most severe depressive episodes is still unsurpassed.[4] Its
benefits include a high level of inhibition of the human sero-
[a] H. Christensen, C. Schjøth-Eskesen, M. Jensen, Dr. H. H. Jensen
Department of Chemistry, Aarhus University
Langelandsgade 140
8000 Aarhus C (Denmark)
Fax : (+45)8619-6199
E-mail: hhj@chem.au.dk
[b] Dr. S. Sinning
Laboratory of Molecular Neurobiology
Centre for Psychiatric Research
Aarhus University Hospital
Skovagervej 2
Scheme 1. a) Structure of antidepressants imipramine (1) and escitalo-
pram (2). b) One-pot palladium-catalysed synthesis of promazine (3), see
reference [7]. c) Palladium-catalysed cyclisation to 3,7-disubstituted
imipramines (this study).
8240 Risskov (Denmark)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201100885.
10618
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10618 – 10627

FULL PAPER
tributable to inhibition of the human norepinephrine trans-
porter (hNET) or anti-cholinergic effects. Today most anti-
depressants used to treat mild to moderate depressions are
selective serotonin reuptake inhibitors (SSRIs)—for exam-
ple, escitalopram (2, Scheme 1a)—which selectively target
hSERT. Due to the transporter specificity, SSRIs generally
have fewer side effects than TCAs.
Until recently, knowledge about the molecular architec-
ture of the serotonin transporter upon binding of antide-
pressants was limited. We have been successful in construct-
ing and experimentally verifying a homology model of
hSERT bound to 2[5] and 1,[6] which can be used for rational
drug design.
sitions of imipramine (Scheme 1c). Although esters are gen-
erally incompatible with commonly used strong bases, such
as sodium tert-butoxide, they facilitate palladium oxidative
insertion[8] and can be converted to a variety of other func-
tionalities. Very few examples of Pd-catalysed reactions that
result in a seven-membered ring are known.[9] A study by
Buchwald and co-workers found that formation of seven-
membered rings required long reaction times and, in certain
cases, failed to work.[10]
Reaction conditions for the cyclisation process were
screened on two monomeric test systems (4 and 5,
Scheme 2; Table 1). For methyl ester 4 the combination of
JosiPhos[11] and Pd
ACHTNUGTRNE(UNG OAc)2/K3PO4 in dimethoxyethane
In this article, we present the synthesis and inhibition con-
stants of new 3,7-disubstituted imipramine analogues de-
signed to exploit two complementary pockets in hSERT that
were established in our previous study.[6] Mutations in these
pockets near the 3- and 7-positions of bound imipramine
affect the affinity of the 3,7-disubstituted analogues in a
manner consistent with our validated model for TCA bind-
ing.[6] We further demonstrate
(DME) or toluene provided the aniline product 7 in poor
yield (Table 1, entries 1 and 2). Changing the palladium
source to [PdACHTNUTRGNEG(UN dba)2], in combination with K3PO4 and the
XPhos ligand in DME or toluene, gave 7 in improved yields
of 29 and 54%, respectively (Table 1, entries 3 and 4). No
product, however, was formed when [Pd
ACHTUNGRTEN(NUNG dba)2] (5 mol%)
and XPhos (10 mol%) in toluene were used, with either
that it has been possible to
create analogues of 1 that are
more selective for hSERT
versus hNET by the introduc-
tion of substituents in the 3-
and 7-positions. This could po-
tentially eliminate some of the
side effects associated with
hNET inhibition by 1.
Results and Discussion
Synthesis: For the synthesis of
the seven-membered dibenzo-
ACHTUNGTRENNUNG[b,f]azepine ring of the imipra-
mine analogues we were in-
spired by the recent elegant
approach for the synthesis the
anti-psychotic drug promazine
(3, Scheme 1b), established by
Scheme 2. Test system for screening the reaction conditions.
Jørgensen and co-workers[7]
.
The six-membered ring was as-
sembled around the dimethyla-
minopropyl amine in a three-
step one-pot synthesis in the
presence of a palladium cata-
lyst and a strong base. In our
case, a similar final manoeuvre
would require the challenging
task of closing a seven-mem-
bered ring via an eight-mem-
bered intermediate, in the
presence of pre-installed ester
handles (X and Y) in what
would become the 3- and 7-po-
Table 1. Screening of conditions for the reaction shown in Scheme 2 with hexylamine (1.5 equiv).[a]
Entry
Pd cat (mol%)
Ligand (mol%)
Base
Solvent
t [h]
Yield [%][b]
1
2
3
4
5
6
7
8
9
10
Pd
Pd
[Pd
[Pd
N
JosiPhos[c] (5)
JosiPhos[c] (5)
XPhos (10)
XPhos (10)
–
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K2CO3
DME
PhCH3
DME
PhCH3
DME
PhCH3
PhCH3
PhCH3
tBuOH
tBuOH
72
72
48
48
48
48
43
24
72
18
5
11
29
54
16
34
49
65
48
73
ACHTUNGTRENNUNG
A
XPhos precat. (1)
XPhos precat. (1)
XPhos precat. (2)
XPhos precat. (5)
XPhos precat. (5)
XPhos precat. (5)
–
–
XPhos (5)
XPhos (5)
XPhos (5)
[a] Reactions were carried out in a sealed tube in heating block at 1108C; entries 1–7: R=Me; entries 8–10:
R=Et. [b] Isolated product yield after column chromatography. [c] JosiPhos-PPF-tBu.
Chem. Eur. J. 2011, 17, 10618 – 10627
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10619

H. H. Jensen et al.
lithium
bis(trimethylsilyl)-
ACHTUNGTRENNUNGamide (LHMDS) or 1,8-diaza-
bicycloundec-7-ene (DBU) as
the base.
Buchwald and co-workers
have recently developed and
commercialised a stable PdII–
XPhos
precatalyst,
which,
upon treatment with base, pro-
duces indoline and the active
catalyst
Pd0–XPhos
(Scheme 2).[12] Reaction with
this precatalyst (1 mol%) in
the presence of K3PO4 in
either DME or toluene result-
ed in incomplete conversion of
starting material but did pro-
vide the aniline product 7 in
moderate yields, again with the
latter solvent being superior
(Table 1, entries 5 and 6). In-
Scheme 3. Synthetic route to methyl and ethyl ester-substituted imipramines 18 and 19.
creasing the catalyst loading to
2 mol% and performing the
reaction in toluene with K3PO4 (Table 1, entry 7) resulted in
an improved conversion and product yield (49 versus 34%).
The yield was further improved to 65% for reaction of the
ethyl ester 5 in the presence of additional ligand (Table 1,
entry 8) to give aniline 8.
system and reaction completion was not reached when
5 mol% of the precatalyst was used (Table 2, entry 1). Rais-
ing the amount of catalyst to 10 mol% (Table 2, entry 2) re-
sulted in full conversion but the reaction yield was a disap-
Table 2. Cyclisation of ethyl ester 17 in tBuOH.[a]
Reaction of ethyl ester 5 in tert-butanol was investigated
with both K3PO4 and K2CO3 in combination with the XPhos
precatalyst and additional XPhos (Table 1, entries 9 and 10).
In a reasonable reaction time (18 h), K2CO3 was found to
give the best result so far in terms of yield (73%), whereas
K3PO4 reacted more sluggishly under otherwise identical re-
action conditions and produced the same product in 48%
yield.
For ease of reaction monitoring by TLC analysis n-hexyl-
ACHTUNGTRENNUNGamine was used as a model amine. Reaction of dimethylami-
nopropyl amine with 5 under the K2CO3/XPhos precatalyst/
tert-butanol conditions described above gave aniline 9 in
85% isolated yield (Scheme 2).
Entry Me2N
G
XPhos precata-
t
T
Yield
[equiv]
[equiv] lyst [mol%]
[h] [8C] [%][d]
1
2
3
4
5
6
1.5
1.5
1.5
3.0
3.0
1.5
3
3
3
3
6
3
5
10
5
10
10
20
96 110[b] 15[d]
26 110[b] 21
6
140[c] 27[e]
1.5 140[c] 33
1.5 140[c] 34
1.5 140[c] 29
[a] Amount of XPhos precatalyst was varied; XPhos (5 mol%) was
added to every reaction. [b] Sealed tube in heating block. [c] MW irradia-
tion. [d] Isolated yield after column chromatography. [e] Incomplete con-
version of starting material.
Having established successful reaction conditions for the
monomeric test system (Scheme 2), we progressed to the
preparation of the cyclisation precursors (Scheme 1c). Com-
mercially available methyl and ethyl p-iodobenzoate (10 and
11, respectively) were each reacted in a double Sonogashira
coupling with acetylene gas to give alkynes 12 and 13,[13] re-
spectively, in acceptable yields.[14] Palladium/charcoal-cata-
lysed reduction quantitatively resulted in alkanes 14 and 15,
respectively, which each underwent bromination with 1,3-di-
bromo-5,5-dimethyl hydantoin (DBMH)[15] and triflic acid
(TfOH) in dichloromethane to provide the respective dibro-
mides 16 and 17 in good yields (Scheme 3).
pointing 21%. We have previously reported shorter reaction
times and superior yields for reactions heated under micro-
wave (MW) irradiation[16] and, therefore, decided to explore
this in the present case. To our satisfaction this was found to
improve both the yield and reaction time (Table 2, en-
tries 3–6).
With a catalyst loading of only 5 mol%, 27% of the cy-
clised product 19 (Scheme 3) could be isolated after 6 h of
MW irradiation at 1408C (Table 2, entry 3). This was slightly
improved to 34% by raising the catalyst loading to
10 mol% and the amounts of amine and base to three and
six equivalents, respectively. Only small changes in the yield
of 19 resulted from increasing catalyst loading to 20 mol%
or varying the amounts of amine and base (Table 2, en-
tries 4–6).
The cyclisation of ethyl ester 17 was first investigated on
the basis of the reaction conditions established for the test
system (Scheme 2, Table 1, entry 10). The rate of product
formation was greatly diminished compared to the test
10620
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10618 – 10627

3,7-Disubstituted Imipramines
FULL PAPER
The methyl ester-substituted
cyclisation precursor 16 was
cyclised under the conditions
described in Table 2, entry 4 to
give the expected 3,7-disubsti-
tuted imipramine 18 in 21%
yield (Scheme 3).
Several side reactions during
the cyclisation process could
be expected. The only success-
fully isolated and identified by-
products were the products
that originated from monoami-
nation and reduction, rather
than cyclisation of the second
bromide (20 and 21). This was
generally, regardless of reac-
tion conditions, present in half
the amount of the cyclised
product. The formation of
these side products is consis-
tent with the expectation that
formation of
a seven-mem-
bered ring via an eight-mem-
bered intermediate is a slow
reaction. No byproducts origi-
nating from amide formation,
ester hydrolysis by minute
amounts of water, transesterifi-
cation with the solvent or pal-
Scheme 4. Synthesis of imipramine analogues 23 and 25–28.
ladium-catalysed diamination
could be identified.
To further explore the space limitations opposing the
imipramine 3- and 7-positions within hSERT it was decided
to also prepare the tert-butyl ester analogues of 18 and 19.
Because tert-butyl esters are acid labile and, therefore,
cannot be expected to remain stable during the acid-cata-
lysed bromination reaction, a late ester interchange would
be necessary. Dibromo dimethyl ester 16 was hydrolysed in
88% yield and the carboxylic acid reacted with Boc2O
(Boc=tert-butoxycarbonyl) and 4-dimethylaminopyridine
(DMAP) in tert-butanol[17] under gentle heating (Scheme 4).
This produced the di-tert-butyl ester 22 in 73% yield. Palla-
dium-catalysed cyclisation, as described for the correspond-
ing ethyl ester 17, gave imipramine analogue 23 in 49% yield,
together with 20% of the monoaminated and reduced prod-
uct 24. Although this yield seems moderate it has to be kept
in mind that two bonds are formed in the cyclisation step.
The imipramine analogue 23 next underwent either
LiAlH4 reduction to 3,7-dihydroxymethyl imipramine (25)
or ester deprotection with hydrochloric acid to give dicar-
boxylic acid 28. Diol 25 was either methylated by dimethyl
sulfate to give 3,7-dimethoxymethyl imipramine (27) in
34% yield or removed by hydrogenolysis over Pd/C to give
3,7-dimethyl imipramine (26) in 13% yield (Scheme 4).
Until this point, we have prepared six symmetrical 3,7-di-
approach. We next sought to investigate whether this strat-
egy also could be employed to prepare the more challenging
unsymmetrical 3,7-disubstituted imipramines.
First, 2,2,2-trichloroethyl 4-iodobenzoate (30) was syn-
thesised from 4-iodobenzoic acid (29) (Scheme 5). Com-
pound 30 then underwent a Sonogashira coupling with tri-
methylsilyl (TMS) acetylene, followed by removal the TMS
group with tetrabutylammonium fluoride (TBAF) to yield
terminal alkyne 31 in 76% yield. Sonogashira coupling of 31
with ethyl 4-iodobenzoate gave compound 32 in 92% yield.
Pd/C-catalysed hydrogenation of alkyne 32 to alkane 33 pro-
ceeded in 89% yield and was followed by dibromination
with DBMH to give the cyclisation precursor 34 in 73%
yield. However, cyclisation of this unsymmetrical dibromide
34 failed. We speculated whether this negative result was
due to the trichloroethyl ester moiety. Amination was at-
tempted under the conditions outlined in Table 1, entry 10
on the more simple monomeric test system, trichloroethyl
ester bromide 6 (Scheme 2), but no aniline was formed.
As a consequence of the apparent incompatibility of the
trichloroethyl ester with our Pd-catalysed amination/cyclisa-
tion conditions we decided to introduce a tert-butyl ester
group.
Trichloroethyl ester deprotection[18] with zinc dust resulted
in quantitative yield of the carboxylic acid, which was con-
ACHTUNGTRENNUNGsubstituted imipramines through our amination/cyclisation
Chem. Eur. J. 2011, 17, 10618 – 10627
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10621

H. H. Jensen et al.
and 19) being the most potent
compounds of the set (Table 3,
entries 2 and 3). This outcome
was in agreement with the
steric limitations, especially
those of the pocket vicinal to
the 3-position, predicted by
Sinning et al.[6] The fact that
the overall affinity for the
human dopamine transporter
(hDAT) remained low and the
inhibitory effect of bulky 3,7-
diester functionalities was op-
posite, with small increases in
the level of inhibition com-
pared to the parent imipramine
1 (Table 3, entries 4 and 5),
suggests that 1 and its ana-
logues occupy a different site
in hDAT, possibly the non-
competitive low-affinity S2
site,[6,19,20] Unsymmetric imipr-
amine 37 substituted with tert-
butylester and hydroxylmethyl
substituents showed a signifi-
cant decrease in hDAT and
hNET inhibition, but remained
a
micromolar inhibitor of
hSERT (Table 3, entry 6). As
expected, dicarboxylic acid
imipramine 28 exhibited poor
inhibition of all three mono-
Scheme 5. Synthesis of unsymmetric imipramine analogues 36 and 37.
amine transporters (Table 3,
entry 7).
verted into tert-butyl ester 35, in good yield, by treatment
with Boc2O and DMAP in tBuOH. Diester 35 was success-
fully transformed into 3,7-disubstituted imipramine 36 in
37% yield. A second unsymmetric imipramine analogue 37
was attained in 79% yield by LiBH4 reduction of the ethyl
ester group in 36 (Scheme 5).
The most potent compound tested against hSERT was
found to be 3,7-dimethylimipramine (26) with 1.6-fold stron-
ger inhibition than 1 (Table 3, entry 10 versus 1). In addi-
tion, 26 has a 1.8-fold lower affinity for hDAT and a 8.1-fold
lower affinity for hNET than 1, which means 26 has both
higher affinity and selectivity for hSERT versus 1. Again,
this is very interesting for the development of selective anti-
depressants with potentially fewer adverse effects.
3,7-Dihydroxymethylimipramine (25) (Table 3, entry 8)
was found to be a reasonably good inhibitor of hSERT, only
4.3-fold less potent than 1 itself. Whereas hDAT inhibition
by 25 essentially remained unchanged relative to 1, hNET
inhibition was reduced 536-fold relative to 1. As a result,
hNET/hSERT selectivity of the TCA 25 for hSERT over
hNET increased dramatically by 170-fold; a level superior
to the SSRIs fluoxetine (Prozac) and fluvoxamine
(Luvox).[21] Also 3,7-dimethoxymethylimipramine (27)
(Table 3, entry 9) showed a moderate affinity for hSERT
but an even higher hNET/hSERT selectivity than 25. Com-
bined, our findings illustrate that manipulation of the 3- and
7-substituents of imipramines can be successfully utilised to
bestow high selectivity onto otherwise unselective TCAs.
Inhibition of monoamine transporters: We have earlier de-
scribed how the 3- and 7-positions of substituted imipramine
analogues can promote increased affinity for hSERT by in-
teraction with specific residues in the protein.[6] However,
neither the full spectrum of favoured substituents, nor the
steric limitations of the 3- and 7-substituents, have been de-
scribed. Furthermore, 3-substituents that increase affinity
for hSERT do the complete opposite in hNET.
Therefore, the right combination of 3- and 7-substituents
might result in an imipramine analogue with increased selec-
tivity. With nine imipramine analogues in hand, inhibition of
serotonin, norepinephrine and dopamine uptake by their
corresponding transporter proteins was tested (Table 3).
The imipramine analogues with 3,7-diester substitution
were found to have a diminished ability to inhibit both
hSERT and hNET relative to 1, with the smaller esters (18
10622
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10618 – 10627

3,7-Disubstituted Imipramines
FULL PAPER
Table 3. Inhibition constants[a] (Ki) [mM] of imipramine and analogues for radio-tracer uptake inhibition of
Earlier, we have found that
the hSERT mutants A173M
and F335L interfere with 3,7-
disubstituted imipramine ana-
logues, and that the double
mutant A173M_F335L is sensi-
tive to 3-substituted imipra-
mine analogues and even more
sensitive to 3,7-disubstituted
imipramine analogues.[6]
hSERT, hNET and hDAT.
Entry
Structure
K
(hSERT)
K
(hDAT)
K
N
K
K
N
K
K
iA
iA(hSERT)
CHTUNGTRENNUNG
1
0.051
160
0.069
3100
1.4
To confirm that our imipra-
mine analogues adopt the
same overall orientation as we
have shown for 1, clomipra-
mine and cyanoimipramine in
the central binding site of
hSERT, we subjected the mu-
tants A173M, F335L and
A173M_F335L to uptake-in-
hibition studies (Table 4).
2
1.3
34
28
26
21
3
2.5
33
10
13
4.0
Generally, we find a pattern
consistent with our existing
model[6] for inhibitory potential
of imipramines. Whereas no
significant negative effect of
the mutations is observed for
1, compounds 25, 27, 28 and 37
(Table 4) are sensitive to the
A173M_F335L double mutant.
These compounds, which all
carry at least one small un-
charged substituent, adhere to
the expected pattern, in which
the double mutant interferes
considerably with the substitu-
ents and produces a 4.4-fold
(calculated from 37, (p=
0.0219)) or higher (pꢀ0.0112)
loss of affinity relative to wild-
type (WT) hSERT. Fully con-
sistent with our model, the
largest effect is seen with the
hydrophilic imipramine 25,
which experiences a 110-fold
loss of affinity due to the hy-
4
110
15
130
180
1.2
1.7
5
54
40
3.6
2.7
6
6.8
>1000
>1000
190
>1000
>148
>6.6
860
>148
7
150
0.22
3.8
440
2.9
8
37
170
>260
17
drophobic
mutations
in
A173M_F335L, again confirm-
ing the prediction that these
residues control the micro-en-
vironment vicinal to the 3- and
7-positions of imipramine.[6]
In contrast to the imipra-
mine analogues with small sub-
stituents (25, 27, 28 and 37),
imipramines with two bulky
ester substituents (18, 19, 23
and 36) all exhibit micromolar
9
>1000
>1000
>260
10
0.032
280
0.56
8600
[a] Data are from at least three independent experiments.
Chem. Eur. J. 2011, 17, 10618 – 10627 ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10623

H. H. Jensen et al.
Table 4. Inhibition constants[a] (Ki) [mm] of imipramine and analogues for radio-tracer
uptake inhibition of WT hSERT and mutants.
adhere to the same orientation as imipramine
within the central binding site and these positions
could be manipulated to achieve higher hSERT se-
lectivity with no, or only slight, loss of affinity. In
particular, 3,7-dimethylimipramine (26) was found
to be a potent inhibitor of hSERT and dihydroxy-
methyl imipramine (25) demonstrated a remark-
able 121-fold gain in inhibitory selectivity for
hSERT over hNET relative to imipramine (1).
Thus, with only a few modifications of largely un-
selective 1 it was possible to obtain imipramine an-
alogues with selectivity superior to existing SSRIs
on the market. This suggests an unexploited poten-
tial of imipramine analogues that can be accessed
through rational drug design by using our experi-
mentally validated model of TCA binding. Further-
Structure
KiACHTUNGTRENNUNG(WT Ki(A173M Ki(F335L Ki(A173M_F335L Ki(A173M_F335L
hSER- hSERT)
T)
hSERT)
hSERT)
hSERT)/KiACTHUNGETRNNU(G WT
hSERT)
1
0.051
1.3
2.5
110
15
6.8
150
0.22
3.8
0.0073
0.21
0.33
27
3.0
8.7
100
2.9
2.5
0.062
5.6
4.9
190
26
50
870
0.96
33
0.041
1.5
1.8
76
10
62
660
25
44
0.8
1.1
2.0
0.7
0.7
9.1
4.4
110
12
18
19
23
36
37
28
25
27
26
0.032
0.038
0.15
0.45
14
[a] Data are from at least three independent experiments.
affinity for WT hSERT and are insensitive to the
A173M_F335L mutations, with Ki values not statistically dif-
ferent from WT hSERT. Our experimentally validated
model for imipramine binding predicts that substituents of
this size cannot be accommodated in the pocket vicinal to
the 3-position and the present findings that these analogues
have both low affinity and respond atypically to the
A173M_F335L double mutation suggests that they bind in a
different orientation, if not in a different binding site, than
1. It is feasible that the bulky substituents prevent them
from binding in the central binding site (S1) and only allows
them to bind to the promiscuous low-affinity S2 site,[6,20,21]
thus retaining the ability to inhibit uptake, albeit with the
low affinity associated with this vestibular site.
more, it could be imagined that compounds such as 25,
which shows good hSERT selectivity, could act as useful
starting materials for introduction of radioactive labels (for
example, 11C or 18F isotopes) for in vivo positron emission
tomography studies.
Experimental Section
General methods: All reagents, unless otherwise stated, were used as pur-
chased without further purification. Reagents for palladium-catalysed
amination were mixed in a glovebox but reactions were conducted in
sealed tube in an ordinary fumehood or in a microwave vial. Solvents
were dried by standard procedures. Columns for flash chromatography
1
were packed with silica gel (60 ꢁ). H and 13C NMR spectra were record-
ed on a Varian Mercury 400 NMR spectrometer and assigned by using
gHMQC, and gCOSY experiments. Mass spectral analyses were carried
out as electrospray experiments on a Micromass LC-TOF spectrometer.
Melting points were measured on a Bꢂchi B-540 apparatus and are cor-
rected. Microwave experiments were carried out in a Biotage Initiator
(Biotage, Sweden). Reaction times listed refer to “hold time” at the
specified temperature.
Conclusion
We have developed a palladium-catalysed amination/intra-
molecular cyclisation process for the synthesis of symmetric
and unsymmetric analogues of the antidepressant imipra-
mine. The challenging seven-membered ring could be
formed in up to 49% isolated yield when tert-butyl ester
groups were present, whereas the cyclisation in the presence
of methyl- and ethyl esters proceeded in lower yields. We
have shown that the established synthetic route is flexible;
the 3,7-di-tert-butyl ester-substituted imipramine could
easily be converted into a series of 3,7-disubstituted imipra-
mine analogues.
By studying the inhibition of hSERT with the new com-
pounds prepared in this study, our previously published
model for the binding of imipramine to the central binding
site of hSERT[6] was further validated. Analogues with large
substituents, predicted to violate the steric restrictions of the
pockets vicinal to the 3- and 7-positions from our experi-
mentally validated model of imipramine binding, were
found to exhibit an atypical response to mutations of A173
and F335. This suggested a different orientation within the
binding site, if not the utilisation of a different binding site
altogether, for these large analogues. Analogues with small-
er substitutents at the 3- and 7-positions were found to
For details on biological assays and the preparation of compounds 7–9,
12 and 14 see the Supporting Information
Compound 16: TfOH (0.95 mL, 10.7 mmol) and DBMH (1.12 g,
3.94 mmol) were added to a solution of 14 (1.07 g, 3.58 mmol) in dry
CH2Cl2 (30 mL) in a flask wrapped in aluminium foil. The reaction mix-
ture was stirred for 2 h at RT then CH2Cl2 and an aqueous solution of
Na2S2O3 were added to the reaction mixture. The aqueous phase was ex-
tracted with CH2Cl2 (ꢃ3) and the combined organic phases were washed
with an aqueous solution of KHSO4, dried over MgSO4, filtered and con-
centrated under reduced pressure. Purification by column chromatogra-
phy (CH2Cl2/pentane 2:3) afforded 16 as a white solid (1.27 g, 78%). Rf =
0.20 (CH2Cl2/pentane 1:1); m.p. 183.1–185.28C (CH2Cl2), lit.[6] 180–
1828C; 1H NMR (400 MHz, CDCl3): d=8.22 (d, J=1.6 Hz, 2H), 7.85
(dd, J=1.6, 8.0 Hz, 2H), 7.17 (d, J=8.0 Hz, 2H), 3.92 (s, 6H), 3.11 ppm
(s, 4H); 13C NMR (100 MHz, CDCl3): d=166.0, 145.5, 134.3, 130.9, 130.4,
128.9, 124.7, 52.7, 36.4 ppm; HRMS (ES): m/z calcd for C18H16O479Br2Na:
476.9313; found: 476.9314. The NMR data were in accordance with those
previously reported.[6]
Compound 18: Compound 16 (54 mg, 0.12 mmol), K2CO3 (52 mg,
0.38 mmol), Buchwaldꢄs XPhos precatalyst (8.7 mg, 0.012 mmol), XPhos
(2.8 mg, 0.0059 mmol), 3-dimethylamino-propylamine (45 mL, 0.36 mmol)
and tert-butanol (1 mL) were added to an oven-dried microwave vial.
The reaction mixture was stirred for 2 h at 1508C under MW irradiation.
The solvent was evaporated and the residue purified by column chroma-
tography (pentane/AcOEt/Et3N 76:19:5) to give a yellow oil (10 mg,
10624
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10618 – 10627

3,7-Disubstituted Imipramines
FULL PAPER
21%). Rf =0.42 (pentane/AcOEt/Et3N 63:32:5); 1H NMR (400 MHz,
CDCl3): d=7.78 (d, J=1.6 Hz, 2H), 7.60 (dd, J=1.6, 8.0 Hz, 2H), 7.15
(d, J=7.6 Hz, 2H), 3.89 (s, 6H), 3.86 (t, J=6.8 Hz, 2H), 3.21 (s, 4H),
2.31 (t, J=7.2 Hz, 2H), 2.15 (s, 6H), 1.72 ppm (quintet, J=7.2, 14.0 Hz,
2H); 13C NMR (100 MHz, CDCl3): d=167.1, 148.1, 139.5, 130.0, 128.7,
124.1, 121.6, 57.6, 52.2, 49.1, 45.6, 32.1, 26.2 ppm; HRMS (ES): m/z calcd
for C23H28O4N2H: 397.2127; found: 397.2137.
7.2 Hz, 2H), 1.38 ppm (t, J=7.2 Hz, 6H; CH3); 13C NMR (100 MHz,
CDCl3): d=166.8, 148.3, 139.6, 130.1, 129.3, 124.2, 121.7, 61.2, 57.8, 49.3,
45.8, 32.3, 26.4, 14.7 ppm; HRMS (ES): m/z calcd for C25H32N2O4Na:
447.2260; found: 447.2276.
Byproduct 21 (29 mg, 16%) was also isolated, as a thick yellow oil. Rf =
0.30 (EtOAc/pentane 2:5+5% Et3N); 1H NMR (400 MHz, CDCl3): d=
7.96 (d, J=8.0 Hz, 2H), 7.30 (dd, J=1.6, 7.6 Hz, 1H), 7.25–7.17 (m, 3H),
6.97 (d, J=7.6 Hz, 1H), 4.36 (q, J=7.2 Hz, 4H), 3.27 (t, J=6.0 Hz, 2H),
2.30 (t, J=8.0 Hz, 2H), 2.80 (t, J=8.0 Hz, 2H), 2.46 (t, J=6.0 Hz, 2H),
2.21 (s, 6H), 1.83 (quintet, J=6.0 Hz, 2H), 1.41–1.36 ppm (m, 6H);
HRMS (ES): m/z calcd for C25H34N2O4Na: 449.2416; found: 449.2419.
Compound 13: [PdCl2ACHTUNGTRENNUNG(PPh3)2] (175 mg, 0.249 mmol) and CuI (91 mg,
0.478 mmol) were added to a solution of 11 (6.69 g, 24.2 mmol) and dis-
tilled iPr2NH (45 mL) in dry THF (35 mL). The reaction vessel was evac-
uated and refilled with acetylene gas (ꢃ3). The mixture quickly changed
colour from yellow to black. The resulting solution was stirred at RT
overnight before it was concentrated under vacuum. The remaining sub-
stance was dissolved in CH2Cl2 and washed with H2O, then the aqueous
phase was extracted with CH2Cl2. The combined organic layers were
dried over MgSO4, filtered and concentrated under reduced pressure.
The crude mixture was purified by recrystallisation from hexane and
column chromatography (CH2Cl2/pentane 1:4) to give 13 as an orange–
brown powder (1.73 g, 44%). Rf =0.23 (pentane/CH2Cl2 1:2); m.p. 148.2–
Compound 22: An aqueous solution of NaOH (2m, 100 mL) was added
to a solution of 16 (1.96 g, 4.29 mmol) in THF (100 mL). The resulting
mixture was stirred at RT overnight and then heated at 458C for 6 h.
THF was removed under reduced pressure and the aqueous phase was
acidified by addition an aqueous solution of HCl (1m) until pH=1,
which resulted in a white precipitate. The precipitate was filtered off and
washed with diethyl ether, then the white powder was dried under
vacuum to afford 4,4’-(ethane-1,2-diyl)bis(3-bromobenzoic acid) as
a
1
150.18C (CH2Cl2), lit.[6] 146–1488C; H NMR (400 MHz, CDCl3): d=8.04
fluffy powder (1.62 g, 88%), which was pure enough to be used without
1
(d, J=8.4 Hz, 4H), 7.60 (d, J=8.4 Hz, 4H), 4.39 (q, J=7.2 Hz, 4H),
1.41 ppm (t, J=7.2 Hz, 6H); 13C NMR (100 MHz, CDCl3): d=166.3,
131.9, 130.6), 129.9, 127.6, 91.9, 61.6, 14.7 ppm; HRMS (ES): m/z calcd
for C20H18O4Na: 345.1103; found: 345.1104. The NMR data are in accord-
ance with those previously reported.[22]
further purification. Rf =0.25 (CH3OH/EtOAc 1:10); H NMR (400 MHz,
[D6]DMSO): d=13.24 (brs, OH), 8.10 (s, 2H), 7.89 (d, J=7.8 Hz, 2H),
7.47 (d, J=7.8 Hz, 2H), 3.08 ppm (s, 4H). The 1H NMR data are in ac-
cordance with those previously reported.[6]
Boc2O (254 mg, 1.16 mmol) and DMAP (47 mg, 0.386 mmol) were added
to a solution of 4,4’-(ethane-1,2-diyl)bis(3-bromobenzoic acid) (191 mg,
0.446 mmol) in dry tBuOH (15 mL). The reaction mixture was stirred
overnight at 458C under an N2 atmosphere. The mixture was diluted with
brine and CH2Cl2, then the aqueous phase was extracted with CH2Cl2 (ꢃ
3). The combined organic layers were dried over MgSO4, filtered and
concentrated under vacuum. The crude product was purified by column
chromatography (CH2Cl2/pentane 1:2) to afford 22 as a white powder
(177 mg, 73%). Rf =0.25 (CH2Cl2/pentane 1:2); m.p. 139.0–140.48C
(CH2Cl2); 1H NMR (400 MHz, CDCl3): d=8.14 (d, J=1.6 Hz, 2H), 7.79
(dd, J=1.6, 8.0 Hz, 2H), 7.13 (d, J=8.0 Hz, 2H), 3.08 (s, 4H), 1.58 ppm
(s, 18H); 13C NMR (100 MHz, CDCl3): d=164.6, 144.9, 134.1, 132.2,
130.7, 128.7, 124.5, 81.8, 36.4, 28.5 ppm; HRMS (ES): m/z calcd for
C24H2879Br2O4Na: 561.0252; found: 561.0254.
Compound 15: Pd/C (10 mol%, 56 mg) was added to a solution of 13
(493 mg, 1.53 mmol) in 1:1 EtOAc/CH2Cl2 (20 mL). The reaction vessel
was evacuated and refilled with H2 gas (ꢃ3) and the reaction mixture
was stirred at RT overnight under H2 atmosphere. The mixture was fil-
tered through Celite and the solvent was removed by evaporation under
vacuum to afford 15 as white crystals (487 mg, 97%). Rf =0.38 (CH2Cl2/
pentane 1:2); m.p. 97.3–99.48C (CH2Cl2), lit[6] 97–988C; 1H NMR
(400 MHz, CDCl3): d=7.94 (d, J= 6.4 Hz, 4H), 7.19 (d, J=6.4 Hz, 4H),
4.36 (q, J=6.8 Hz, 4H), 2.99 (s, 4H), 1.38 ppm (t, J=6.8 Hz, 6H);
13C NMR (100 MHz, CDCl3): d=166.9, 146.7, 130.0, 128.8, 61.2, 37.8,
14.7 ppm; HRMS (ES): m/z calcd for C20H22O4Na: 349.1416; found:
349.1416. The NMR data are in accordance with those previously report-
ed.[23]
Compounds 23 and 24: Compound 22 (217 mg, 0.402 mmol) was added
to an oven-dried microwave vial. The vial was transferred to a glovebox
and K2CO3 (167 mg, 1.21 mmol), Buchwaldꢄs XPhos-precatalyst (30 mg,
0.040 mmol) and XPhos (10 mg, 0.020 mmol) were added, followed by
dry tBuOH (4 mL) and N,N-dimethylpropane-1,3-diamine (152 mL,
1.21 mmol). The vial was sealed and heated at 1408C for 100 min under
MW irradiation. The crude mixture was concentrated under vacuum and
purified by column chromatography (AcOEt/pentane 1:5+3% Et3N) to
afford 23 as a yellow oil (94 mg, 49%). Rf =0.43 (AcOEt/pentane 1:5+
Compound 17: TfOH (1.36 mL, 15.4 mmol) and DBMH (1.61 g,
5.63 mmol) were added to a solution of 15 (1.67 g, 5.12 mmol) in dry
CH2Cl2 (40 mL) in a flask wrapped in aluminium foil. The reaction was
stirred at RT and closely followed by TLC analysis. After 2 h, 15 and
mono-brominated compound were no longer observed. The reaction mix-
ture was then diluted with CH2Cl2 and an aqueous solution of Na2S2O3,
and was stirred for a few minutes until the organic phase turned light
yellow. The aqueous phase was extracted with CH2Cl2 (ꢃ3) and the com-
bined organic layers were washed with an aqueous solution of KHSO4,
dried over MgSO4, filtered and concentrated under vacuum. The product
was purified by column chromatography (pentane/CH2Cl2 3:2) to afford
17 as white powder (2.30 g, 93%). Rf =0.22 (CH2Cl2/pentane 1:1); m.p.
102.3–103.48C (CH2Cl2); 1H NMR (400 MHz, CDCl3): d=8.17 (d, J=
1.6 Hz, 2H), 7.81 (dd, J=1.6, 8.0 Hz, 2H), 7.13 (d, J=8.0 Hz, 2H), 4.34
(q, J=7.6 Hz, 4H), 1.36 ppm (t, J=7.6 Hz, 6H); 13C NMR (100 MHz,
CDCl3): d=165.5, 145.4, 134.2, 130.8, 130.7, 128.9, 124.7, 61.6, 36.4,
14.6 ppm; HRMS (ES): m/z calcd for C20H2081Br2O4Na: 508.9585; found:
508.9586.
1
5% Et3N); H NMR (400 MHz, CDCl3): d=7.71 (d, J=1.6 Hz, 2H), 7.52
(dd, J=1.6, 8.0 Hz, 2H), 7.11 (d, J=8.0 Hz, 2H), 3.84 (t, J=7.4 Hz, 2H),
3.18 (s, 4H), 2.39 (t, J=7.4 Hz, 2H), 2.20 (s, 6H), 1.73 (quintet, J=
7.4 Hz, 2H), 1.57 ppm (s, 18H); 13C NMR (100 MHz, CDCl3): d=166.1,
148.3, 139.1, 130.8, 130.0, 124.0, 121.6, 81.2, 57.8, 49.2, 45.8, 32.3, 28.6,
26.4 ppm; HRMS (ES): m/z calcd for C29H40N2O4Na: 503.2886; found:
503.2885.
Byproduct 24 (39 mg, 20%) was also isolated, as a yellow oil. Rf =0.18
(AcOEt/pentane 1:5+3% Et3N); 1H NMR (400 MHz, CDCl3): d=7.90
(d, J=8.0 Hz, 2H), 7.23–7.17 (m, 4H), 6.96 (d, J=8.0 Hz, 1H), 3.25 (t,
J=6.0 Hz, 2H), 2.98 (t, J=10.0 Hz, 2H), 2.79 (t, J=10.0 Hz, 2H), 2.43
(t, J=6.0 Hz, 2H), 2.19 (s, 6H). 1.81 (quintet, J=6.0 Hz, 2H), 1.59 (s,
6H), 1.58 ppm (s, 6H); HRMS (ES): m/z calcd for C29H42N2O4Na:
505.3042; found: 505.3051.
Compounds 19 and 21: Compound 17 (206 mg, 0.43 mmol) and K2CO3
(180 mg, 1.30 mmol) were added to an oven-dried microwave vial. The
vial was transferred to the glovebox and Buchwaldꢄs XPhos precatalyst
(31.4 mg, 0.043 mmol), XPhos (10.2 mg, 0.021 mmol), 3-dimethylamino-
propylamine (161 mL, 1.28 mmol) and tert-butanol (4 mL) were added.
The reaction mixture was stirred for 1.5 h at 1408C under MW irradia-
tion. The solvent was evaporated and the residue was purified by column
chromatography (pentane/AcOEt/Et3N 76:19:5) to give 19 as a yellow oil
(62 mg, 34%). Rf =0.55 (EtOAc/pentane 2:5+5% Et3N); 1H NMR
(400 MHz, CDCl3): d=7.78 (d, J=1.6 Hz, 2H), 7.60 (dd, J=1.6, 8.0 Hz,
2H), 7.15 (d, J=8.0 Hz, 2H), 4.36 (q, J=7.2 Hz, 4H), 3.57 (t, J=7.2 Hz,
2H), 3.20 (s, 4H), 2.31 (t, J=7.2 Hz, 2H), 2.15 (s, 6H), 1.72 (quintet, J=
Compound 25: LiAlH4 (26 mg, 0.688 mmol) was added to a stirred solu-
tion of 23 (64 mg, 0.134 mmol) in dry diethyl ether (4 mL) at 08C. After
stirring for 2 h at 08C the mixture was diluted with CH2Cl2 and H2O. An
aqueous solution of HCl (1m) was added until pH=2–3, then the aque-
ous layer was extracted with CH2Cl2 (ꢃ3). The combined organic phases
were washed with a saturated aqueous solution of NaHCO3, dried over
MgSO4, filtered and concentrated under vacuum. The product was puri-
Chem. Eur. J. 2011, 17, 10618 – 10627
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10625

H. H. Jensen et al.
fied by column chromatography (AcOEt+1% CH3OH+5% Et3N) to
afford 25 as a colourless oil (39 mg, 85%). Rf =0.34 (AcOEt/Et3N/
CH3OH 90:5:5); 1H NMR (400 MHz, CDCl3): d=7.10–7.06 (m, 4H), 6.89
(d, J=7.6 Hz, 2H), 4.58 (s, 4H), 3.78 (t, J=6.4 Hz, 2H), 3.13 (s, 4H),
2.81–2.77 (m, 2H), 2.39 (s, 6H), 2.07 (s, 2H), 1.98–1.91 ppm (m, 2H);
13C NMR (100 MHz, CDCl3): d=148.1, 139.9, 133.7, 130.4, 121.8, 118.7,
65.3, 62.9, 51.5, 48.5, 32.0, 22.6 ppm; HRMS (ES): m/z calcd for
C21H28O2N2Na: 363.2048; found: 363.2068.
Compound 31: [PdCl2ACHTGUNTERNNU(G PPh3)2] (18 mg, 0.026 mmol), CuI (10 mg,
0.050 mmol) and TMS-acetylene (350 mL, 2.51 mmol) were added se-
quentially to a solution of 30 (866 mg, 2.28 mmol) in Et3N (18 mL). A
precipitate immediately formed. The reaction mixture was stirred at RT
for 4 h, then diluted with H2O and CH2Cl2. The layers were separated
and the aqueous phase was extracted with CH2Cl2 (ꢃ3). The combined
organic layers were washed with an aqueous solution of HCl (0.1m) and
a saturated aqueous solution of NaHCO3, dried over MgSO4, filtered,
concentrated under vacuum and used without further purification in the
next step. Rf =0.39 (CH2Cl2/pentane 1:2).
Compound 26: 10% Pd/C (28 mg) and a concentrated aqueous solution
of HCl (2 drops) were added to a solution of 25 (82 mg, 0.24 mmol) in
1:2 CH3OH/AcOEt (3 mL). The reaction vessel was evacuated and re-
filled with H2 gas (ꢃ3) then the reaction was stirred for 48 h at RT. The
reaction mixture was filtered through Celite with AcOEt. The solvent
was removed under reduced pressure and the remaining oil was treated
with Pd/C, as above, until the reaction was complete (another 96 h). The
mixture was filtered through Celite with AcOEt and purified by chroma-
tography (pentane/AcOEt/Et3N 76:19:5) to give a colourless oil (10 mg,
13%). Rf =0.35 (pentane/Et3N 95:5); 1H NMR (400 MHz, CDCl3): d=
6.96 (d, J=7.6 Hz, 2H), 6.89 (s, 2H), 6.71 (d, J=7.6 Hz, 2H), 3.75 (t, J=
6.8 Hz, 2H), 3.09 (s, 4H), 2.31 (t, J=7.2 Hz, 2H), 2.23 (s, 6H), 2.16 (s,
6H), 1.72 ppm (q, J=7.2 Hz, 2H); 13C NMR (100 MHz, CDCl3): d=
148.8, 136.0, 131.3, 129.8, 123.2, 120.7, 57.9, 48.9, 45.7, 32.1, 26.4,
21.3 ppm; HRMS (ES): calcd for C21H28N2Na: 331.2150; found: 331.2150.
The crude silylated product was dissolved in dry THF (20 mL) and
cooled to ꢁ788C. TBAF (1m in THF, 2.5 mL, 2.51 mmol) was added and
the solution was stirred at ꢁ788C for 1.5 h. The reaction mixture was di-
luted with H2O and extracted with CH2Cl2 (ꢃ3). The combined organic
layers were washed with brine, dried over MgSO4, filtered and concen-
trated under reduced pressure. The product was purified by column chro-
matography (CH2Cl2/pentane 1:10) to afford alkyne 31 as a white powder
(484 mg, 76%). Rf =0.33 (CH2Cl2/pentane 1:2); m.p. 196.4–197.68C
(CH2Cl2); 1H NMR (400 MHz, CDCl3): d=8.07 (d, J=8.6 Hz, 2H), 7.58
(d, J=8.6 Hz, 2H), 4.97 (s, 2H), 3.28 ppm (s, 1H); 13C NMR (100 MHz,
CDCl3): d=164.3, 132.4, 130.0, 128.7, 127.9, 95.1, 82.7, 80.9, 74.7 ppm;
GC-MS (EI): m/z calcd for C11H735Cl3O2: 275.9; found: 276.
Compound 32: [PdCl2ACTHNUGTRNEUNG(PPh3)2] (42 mg, 0.060 mmol), CuI (52 mg,
0.273 mmol) and distilled iPr2NH (1.55 mL, 11.1 mmol) were added to a
solution of ethyl p-iodobenzoic acid (1.15 g, 4.16 mmol) and 31 (770 mg,
2.77 mmol) in dry THF (25 mL). The resulting mixture was stirred at RT
overnight. The mixture was diluted with H2O and CH2Cl2 then the aque-
ous phase was extracted with CH2Cl2 (ꢃ3). The combined organic layers
were washed with an aqueous solution of HCl (0.1m) and then a saturat-
ed aqueous solution of NaHCO3, dried over MgSO4, filtered and concen-
trated under vacuum. The product was purified by column chromatogra-
phy (pentane/CH2Cl2 4:1!CH2Cl2) to afford 32 as a pale brown powder
(1.09 g, 92%). Rf =0.29 (CH2Cl2/pentane 1:1); m.p. 2508C (decomposed);
1H (400 MHz, CDCl3): d=8.13 (d, J=8.4 Hz, 2H), 8.05 (d, J=8.4 Hz,
1H), 7.65 (d, J=8.4 Hz, 2H), 7.62 (d, J=8.4 Hz, 2H), 4.99 (s, 2H), 4.40
(quartet, J=7.3 Hz, 2H), 1.41 ppm (t, J=7.3 Hz, 3H); 13C NMR
(100 MHz, CDCl3): d=166.1, 164.5, 132.1, 131.9, 130.7, 130.3, 129.8,
128.6, 127.3, 95.2, 92.3, 91.3, 74.8, 61.4, 14.5 ppm; MS (ES): m/z calcd for
C40H3035Cl537ClO8Na: 872.99
Compound 27: NaH (60% in mineral oil, 31 mg, 0.775 mmol) was added
to a solution of 25 (37 mg, 0.108 mmol) in dry THF (2 mL) and the mix-
ture was stirred at RT for 1 h. Dimethyl sulfate (31 mL, 0.324 mmol) was
added and the reaction mixture was stirred for 1 h at RT. The mixture
was diluted with H2O and extracted with CH2Cl2 (ꢃ3). The combined or-
ganic phases were dried over MgSO4, filtered and concentrated under re-
duced pressure. The product was purified by column chromatography
(AcOEt/pentane 1:4+1% CH3OH+1% Et3N) to afford 27 as a colour-
less oil (14 mg, 34%). Rf =0.18 (AcOEt/pentane 1:2+2% CH3OH+2%
Et3N); 1H NMR (400 MHz, CDCl3): d=7.07 (d, J=7.6 Hz, 2H), 4.04 (d,
J=1.4 Hz, 2H), 6.89 (dd, J=1.4, 7.6 Hz, 2H), 4.39 (s, 4H), 3.80 (t, J=
6.4 Hz, 2H), 3.36 (s, 6H), 3.14 (s, 4H), 2.84–2.80 (m, 2H), 2.40 (s, 6H),
2.00–1.93 ppm (m, 2H); 13C NMR (100 MHz, CDCl3): d=148.0, 137.0,
133.9, 130.3, 122.6, 119.5, 74.9, 62.9, 58.4, 51.4, 48.5, 32.3, 22.5 ppm;
HRMS (ES): m/z calcd for C23H32N2O2Na: 391.2361; found: 391.2362.
Compound 28: A solution of 23 (49 mg, 0.102 mmol) in dry 1,4-dioxane
(2.5 mL) and a concentrated aqueous solution of HCl (12m, 0.5 mL)
were combined and stirred at 08C for 1.5 h. The mixture was allowed to
warm to RT, stirred for 3 h, then concentrated under reduced pressure to
afford 28 as a bright yellow powder (41 mg, 99%). Rf =0.16 (Et3N/
CH3OH/CH2Cl2 1:3:10); m.p. 199.0–201.78C (CH2Cl2); 1H NMR
(400 MHz, D2O): d=7.38 (brs, 2H), 7.26 (d, J=7.0 Hz, 2H), 6.74 (d, J=
7.0 Hz, 2H), 3.58 (brs, 2H), 2.97 (brs, 2H), 2.68 (s, 10H), 1.78 ppm (brs,
AHCTUNGERTG(NNUN dimer); found: 873.0.
Compound 33: 10% Pd/C (149 mg) was added to a solution of 32 (1.09 g,
2.55 mmol) in 1:1 EtOAc/CH2Cl2 (30 mL). The reaction vessel was evacu-
ated and refilled with hydrogen gas (ꢃ3). The reaction mixture was
stirred for 48 h at RT, filtered through a bed of Celite and concentrated
under reduced pressure. Alkane 33 was obtained without further purifica-
tion as an orange–brown solid (971 mg, 89%). Rf =0.44 (CH2Cl2/toluene
1:1); m.p. 136.2–138.88C (CH2Cl2); 1H NMR (400 MHz, CDCl3) d=8.04
(d, J=8.2 Hz, 2H), 7.97 (d, J=8.2 Hz, 2H), 7.26 (d, J=8.2 Hz, 2H), 7.21
(d, J=8.2 Hz, 2H), 4.97 (s, 2H), 4.37 (q, J=7.2 Hz, 2H), 3.02 (s, 4H),
1.40 ppm (t, J=7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=166.6,
164.9, 147.8, 146.3, 130.4, 129.9, 129.0, 128.7, 128.6, 126.8, 95.4, 74.5, 61.0,
37.5, 37.4, 14.5 ppm; HRMS (ES): m/z calcd for C20H1935Cl3O4Na:
451.0247; found: 451.0243.
1
2H); H NMR (400 MHz, CD3OD): d=7.91 (d, J=1.6 Hz, 2H), 7.73 (dd,
J=1.6, 7.8 Hz, 2H), 7.36 (d, J=7.8 Hz, 2H), 4.06 (t, J=6.4 Hz, 2H), 3.34
(s, 4H), 3.32–3.28 (m, 2H), 2.90 (s, 6H), 2.13–2.09 ppm (m, 2H);
13C NMR (100 MHz, D2O): d=169.7, 148.9, 141.1, 131.4, 130.7, 125.8,
122.1, 57.2, 43.6, 32.9, 24.2 ppm; HRMS (ES): m/z calcd for
C21H24N2O4H: 369.1814; found: 369.1820.
Compound 30: DMF (one drop) was added to a solution of 29 (400 mg,
1.61 mmol) and oxalyl chloride (211 mL, 2.42 mmol) in dry CH2Cl2
(8 mL). The reaction mixture was stirred at RT for 1.5 h and then con-
centrated under vacuum. The residue was dissolved in dry CH2Cl2 (8 mL)
then Et3N (0.34 mL, 2.45 mmol), trichloroethanol (0.31 mL, 3.23 mmol)
and a catalytic amount of DMAP (20 mg) were added. The reaction mix-
ture was stirred at RT for 4 h, then diluted with CH2Cl2 and H2O. The or-
ganic phase was washed with an aqueous solution of HCl (0.1m), then a
saturated aqueous solution of NaHCO3. The organic phase was dried
over MgSO4, filtered and concentrated under vacuum. The product was
purified by column chromatography (heptane/CH2Cl2 7:1) to afford 30 as
an off-white powder (549 mg, 90%). Rf =0.55 (CH2Cl2/heptanes 1:1);
m.p. 63.5–65.28C (CH2Cl2); 1H NMR (400 MHz, CDCl3): d=7.84–7.79
(m, 4H), 4.95 ppm (s, 2H); 13C NMR (100 MHz, CDCl3): d=164.6, 138.3,
131.6, 128.4, 102.2, 95.1, 74.7 ppm; GC-MS (EI): m/z calcd for
C9H635Cl3IO2: 377.8; found: 378.
Compound 34: DBMH (704 mg, 2.46 mmol) and CF3SO3H (475 mL,
5.37 mmol) were added to a solution of 33 (962 mg, 2.24 mmol) in dry
CH2Cl2 (60 mL) in a flask wrapped in aluminium foil. The reaction was
stirred at RT and closely monitored by TLC analysis. After 4.25 h, the re-
action mixture was diluted with CH2Cl2 and an aqueous solution of
Na2S2O3, then vigorously stirred for 5 min. The aqueous phase was ex-
tracted with CH2Cl2 (ꢃ3) and the combined organic layers were washed
with an aqueous solution of KHSO4, dried over MgSO4, filtered and con-
centrated under vacuum. The product was purified by column chroma-
tography (pentane/CH2Cl2 3:1) to afford 34 as a white powder (964 mg,
73%). Rf =0.36 (pentane/CH2Cl2 1:2); m.p. 129.3–131.28C (CH2Cl2);
1H NMR (400 MHz, CDCl3): d=8.27 (d, J=1.6 Hz, 1H), 8.21 (d, J=
1.6 Hz, 1H), 7.93 (dd, J=1.6, 8.0 Hz, 1H), 7.85 (dd, J=1.6, 8.0 Hz, 1H),
7.23 (d, J=7.6 Hz, 1H), 7.18 (d, J=8.0 Hz, 1H), 4.96 (s, 2H), 4.36 (q, J=
7.2 Hz, 2H), 3.11 (s, 4H), 1.39 ppm (t, J=7.2 Hz, 3H); 13C NMR
10626
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10618 – 10627

3,7-Disubstituted Imipramines
FULL PAPER
(100 MHz, CDCl3): d=165.4, 163.8, 146.6, 145.2, 134.7, 134.2, 131.1,
130.8, 130.7, 129.3, 128.9, 128.8, 124.9, 124.6, 95.2, 74.8, 61.6, 36.5, 36.3,
14.6 ppm; HRMS (ES): m/z calcd for C20H1779Br235Cl3O4Na: 606.8457;
found: 606.8412.
(EtOAc/pentane 1:2+5% Et3N+5% CH3OH); 1H NMR (400 MHz,
CDCl3): d=7.69 (d, J=1.6 Hz, 1H), 7.53 (dd, J=1.6, 8.0 Hz, 1H), 7.14–
7.09 (m, 3H), 6.94 (dd, J=1.6, 7.6 Hz, 1H), 4.64 (s, 2H), 3.83 (t, J=
6.0 Hz, 2H), 3.20–3.14 (m, 4H), 2.85–2.81 (m, 2H), 2.42 (s, 6H), 1.99–
1.95 (m, 2H), 1.58 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d=166.0,
148.0, 147.8, 139.9, 139.1, 133.9, 130.7, 130.4, 130.3, 124.1, 122.2, 120.9,
119.0, 81.3, 65.5, 62.9, 51.4, 48.6, 32.7, 31.7, 28.5, 22.4 ppm.
Compound 35: Freshly activated zinc powder (4.51 g) was added to a so-
lution of 34 (707 mg, 1.20 mmol) in 1:1 dry CH2Cl2/99% acetic acid
(50 mL). The mixture was stirred at RT overnight under a N2 atmosphere.
The reaction mixture was filtered through a pad of glass wool and con-
centrated under vacuum. The residue was dissolved in H2O and EtOAc,
then the aqueous phase was extracted with EtOAc (ꢃ3). The combined
organic phases were dried over MgSO4, filtered and concentrated under
vacuum to afford 3-bromo-4-(2-bromo-4-(ethoxycarbonyl)phenethyl)ben-
zoic acid as a white powder (545 mg, 99%). Rf =0.46 (CH2Cl2/CH3OH
98:2); m.p. 176.4–178.68C (CH2Cl2); 1H NMR (400 MHz, CDCl3): d=
8.29 (d, J=1.6 Hz, 1H), 8.23 (d, J=2.0 Hz, 1H), 7.92 (dd, J=1.6, 8.0 Hz,
1H), 7.87 (dd, J=2.0, 8.0 Hz, 1H), 7.22 (d, J=8.0 Hz, 1H), 7.19 (d, J=
8.0 Hz, 1H), 4.38 (q, J=6.8 Hz, 2H), 3.13 (s, 4H), 1.40 ppm (t, J=6.8 Hz,
3H); 13C NMR (100 MHz, CDCl3): d=171.1, 165.6, 146.6, 145.3, 134.9,
134.3, 131.0, 130.8, 130.7, 129.5, 128.9, 124.8, 124.7, 61.7, 36.5, 36.3,
14.6 ppm; HRMS (ES): m/z calcd for C18H1679Br2O4Na: 476.9313; found:
476.9301.
Acknowledgements
We thank The Carlsberg Research Foundation for funding to H.H.J. and
OChem Graduate School, Aarhus University for support. We are also
grateful to the Skrydstrup research group, Aarhus University, for fruitful
scientific discussions and technical assistance.
[1] The World Health Report 2001: Mental Health: New Understanding,
New Hope, WHO, Geneva, 2001.
[2] F. Lꢅpez-MuÇoz, C. Alamo, Curr. Pharm. Des. 2009, 15, 1563–1586.
[3] R. Kuhn, Am. J. Psychiatry 1958, 115, 459–464.
[4] I. M. Anderson, J. Affective Disord. 2000, 58, 19–36.
[5] H. Koldsø, K. Severinsen, T. T. Tran, L. Celik, H. H. Jensen, O.
Wiborg, B. Schiøtt, S. Sinning, J. Am. Chem. Soc. 2010, 132, 1311–
1322.
[6] S. Sinning, M. Musgaard, M. Jensen, K. Severinsen, L. Celik, H.
Koldsø, T. Meyer, M. Bols, H. H. Jensen, B. Schiøtt, O. Wiborg, J.
Biol. Chem. 2010, 285, 8363–8374.
Boc2O (124 mg, 0.568 mmol) and DMAP (14 mg, 0.111 mmol) were
added to a heterogeneous solution of 3-bromo-4-(2-bromo-4-(ethoxycar-
bonyl)phenethyl)benzoic acid (145 mg, 0.338 mmol) in 4:1 dry tBuOH/
1,4-dioxane (12 mL). The mixture was stirred overnight at 458C under an
N2 atmosphere. TLC analysis showed unreacted starting material, so ad-
ditional Boc2O (52 mg, 0.121 mmol) and dry 1,4-dioxane (3 mL) were
added, which resulted in a homogenous mixture. The mixture was stirred
for additional 4 h, then diluted with H2O and extracted with CH2Cl2 (ꢃ
3). The combined organic layers were washed with brine, dried over
MgSO4, filtered and concentrated under vacuum. Column chromatogra-
phy (CH2Cl2/pentane 1:7) afforded 35 as a white powder (142 mg, 82%).
Rf =0.37 (CH2Cl2/pentane 1:2); m.p. 89.7–91.58C (CH2Cl2); 1H NMR
(400 MHz, CDCl3): d=8.30 (d, J=1.6 Hz, 1H), 8.23 (d, J=1.6 Hz, 1H),
7.94 (dd, J=1.6, 7.7 Hz, 1H), 7.88 (dd, J=1.6, 8.0 Hz, 1H), 7.25 (d, J=
8.0 Hz, 1H), 7.23 (d, J=7.7 Hz, 1H), 4.46 (q, J=7.0 Hz, 2H), 3.18 (s,
4H), 1.68 (s, 9H), 1.48 ppm (t, J=7.0 Hz, 3H); 13C NMR (100 MHz,
CDCl3): d=165.6, 164.7, 145.4, 144.9, 134.2, 134.1, 132.2, 130.9, 130.7,
130.6, 128.9, 128.8, 124.7, 124.5, 81.9, 61.6, 36.5, 36.4, 28.5, 14.6 ppm;
HRMS (ES): m/z calcd for C22H2479Br2O4Na: 532.9939; found: 532.9922.
[7] T. Dahl, C. W. Tornøe, B. Bang-Andersen, P. Nielsen, M. Jørgensen,
Angew. Chem. 2008, 120, 1750–1752; Angew. Chem. Int. Ed. 2008,
47, 1726–1728.
[8] A. Jutand, A. Mosleh, Organometallics 1995, 14, 1810–1817.
ˇ
[9] a) M. Prashad, Y. Liu, X. Y. Mak, D. Har, O. Repic, T. J. Blacklock,
Tetrahedron Lett. 2002, 43, 8559–8562; b) C. Xue, S.-H. Kung, J.-W.
Wu. F.-T. Luo, Tetrahedron 2008, 64, 248–254; c) D. Tsvelikhovsky,
S. L. Buchwald, J. Am. Chem. Soc. 2010, 132, 14048–14051.
[10] J. P. Wolfe, R. A. Rennels, S. L. Buchwald, Tetrahedron 1996, 52,
7525–7546.
[11] a) Q. Shen, J. F. Hartwig, Org. Lett. 2008, 10, 4109–4112; b) Q.
Shen, T. Ogata, J. F. Hartwig, J. Am. Chem. Soc. 2008, 130, 6586–
6596.
[12] M. R. Biscoe, B. R. Fors, S. L. Buchwald, J. Am. Chem. Soc. 2008,
130, 6686–6687.
[13] A. P. Rudenko, A. V. Vasil’ev, Russ. J. Org. Chem. 1995, 31, 1360–
1379.
[14] Recently, a 1,2-dianion equivalent other than acetylene was elegant-
ly developed by Molander and Sandrock, see: G. A. Molander, D. L.
Sandrock, Org. Lett. 2009, 11, 2369–2372.
[15] H. Eguchi, H. Kawaguchi, S. Yoshinaga, A. Nishida, T. Nishiguchi,
S. Fujisaki, Bull. Chem. Soc. Jpn. 1994, 67, 1918–1921.
[16] P. Nilsson, K. Olofsson, M. Larhed, Top. Curr. Chem. 2006, 266,
103–144.
[17] K. Takeda, A. Akiyama, H. Nakamura, S.-I. Takizawa, Y. Mizuno,
H. Takayanagi, Y. Harigaya, Synthesis 1994, 1063–1066.
[18] R. B. Woodward, K. Heusler, J. Gosteli, P. Naegeli, W. Oppolzer, R.
Ramage, S. Ranganathan, H. Vorbrꢂggen, J. Am. Chem. Soc. 1966,
88, 852–853.
Compound 36: Compound 35 (92 mg, 0.179 mmol) and a magnetic stirrer
bar were added to an oven-dried microwave vial. The vial was trans-
ferred to a glovebox and K2CO3 (74 mg, 0.537 mmol), Buchwaldꢄs XPhos
precatalyst (13 mg, 0.018 mmol), XPhos (4 mg, 0.009 mmol), dry tBuOH
(2 mL) and N,N-dimethylpropane-1,3-diamine (68 mL, 0.537 mmol) were
added. The vial was sealed and heated under MW irradiation (1408C,
100 min). The crude mixture was concentrated under vacuum and puri-
fied by column chromatography (AcOEt/pentane 1:7+3% Et3N) to
afford 36 as a colourless oil (30 mg, 37%). Rf =0.26 (AcOEt/pentane
1:5+5% Et3N); 1H NMR (400 MHz, CDCl3): d=7.77 (d, J=1.6 Hz,
1H), 7.73 (d, J=1.6 Hz, 1H), 7.59 (dd, J=1.6, 7.6 Hz, 1H), 7.54 (dd, J=
1.6, 8.0 Hz, 1H), 7.14 (d, J=7.6 Hz, 1H), 7.12 (d, J=8.0 Hz, 1H), 4.36 (q,
J=7.2 Hz, 2H), 3.85 (t, J=7.0 Hz, 2H), 3.20 (s, 4H), 2.31 (t, J=7.0 Hz,
2H), 2.15 (s, 6H), 1.72 (quintet, J=7.0 Hz, 2H), 1.58 (s, 9H), 1.38 ppm
(t, J=7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=166.9, 166.0, 148.4,
148.2, 139.5, 139.2, 130.8, 130.2, 129.9, 129.2, 124.1, 124.0, 121.7, 121.6,
81.2, 61.2, 57.9, 49.2, 45.9, 32.4, 32.2, 28.6, 26.4, 14.7 ppm; HRMS (ES):
m/z calcd for C27H36N2O4Na: 475.2573; found: 475.2564.
Compound 37: LiBH4 (1 mg, 0.060 mmol) was added to a solution of 36
(20 mg, 0.043 mmol) in dry THF (2 mL). The mixture was stirred at 508C
under a N2 atmosphere. After 2.5 h, the mixture was diluted with CH2Cl2
and H2O. An aqueous solution of HCl (1m) was added until pH=2–3,
then the aqueous layer was extracted with CH2Cl2 (ꢃ2). The combined
[19] Z. Zhou, J. Zhen, N. K. Karpowich, R. M. Goetz, C. J. Law, M. E. A.
Reith, D. N. Wang, Science 2007, 317, 1390–1393.
[20] S. K. Singh, A. Yamashita, E. Gouaux, Nature 2007, 448, 952–956.
[21] J. Hyttel, Int. Clin. Psychopharmacol. 1994, 9, 19–26.
[22] N. G. Pschirer, U. H. F. Bunz, Tetrahedron Lett. 1999, 40, 2481–2484.
[23] P. Kaszynski, D. A. Dougherty, J. Org. Chem. 1993, 58, 5209–5220.
organic layers were washed with
a saturated aqueous solution of
NaHCO3, dried over MgSO4, filtered and concentrated under vacuum.
The product was purified by column chromatography (AcOEt/pentane
2:5+5% Et3N) to afford 37 as a colourless oil (14 mg, 79%). Rf =0.23
Received: March 22, 2011
Published online: August 18, 2011
Chem. Eur. J. 2011, 17, 10618 – 10627
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10627