Design and optimisation of selective serotonin re-uptake inhibitors with high synthetic accessibility: Part 2

Article abstract of DOI:10.1016/j.bmcl.2009.08.066

A second wave of potential SSRIs with high ease of synthetic accessibility were designed based on the reported selective serotonin re-uptake inhibitor litoxetine and our own previous work in this area. Preparation and subsequent optimisation yielded a ran

Full text of DOI:10.1016/j.bmcl.2009.08.066

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5893–5897
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and optimisation of selective serotonin re-uptake inhibitors with
high synthetic accessibility: Part 2
*
Mark Andrews, Alan Brown , Jean-Yves Chiva, David Fradet, David Gordon, Mark Lansdell,
Malcolm MacKenny
Discovery Chemistry, Pfizer Global Research and Development, Sandwich, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
A second wave of potential SSRIs with high ease of synthetic accessibility were designed based on the
reported selective serotonin re-uptake inhibitor litoxetine and our own previous work in this area. Prep-
aration and subsequent optimisation yielded a range of potent and highly selective SSRIs.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 27 July 2009
Revised 10 August 2009
Accepted 19 August 2009
Available online 21 August 2009
Keywords:
Seretonin re-uptake
SRI
NRI
DRI
SSRI
Over the last 25 years selective serotonin re-uptake inhibitors
(SSRIs) have emerged as a key mechanistic class in the treatment
of depression.1 Sertraline (ZoloftÒ) 1, paroxetine (PaxilÒ) 2 and flu-
oxetine (ProzacÒ) 3 have, amongst others, been utilised extensively
in the clinical treatment of this debilitating condition (Fig. 1).
Our own interest in this area stemmed from the potential ben-
efit of SSRIs in a range of other indications including pre-men-
strual-syndrome and male sexual dysfunction. For these acute
therapy compatible indications we felt that targeting SSRIs which
demonstrated very short pharmacokinetic Tmax’s would potentially
offer significant benefit over currently available SSRIs, which have
relatively long Tmax’s (typically 4–8 h in the clinic).2
To facilitate this approach we set out to identify rapidly a range of
structurallydiverse,potent, selectiveSSRIsforsubsequentpharmaco-
kinetic profiling. A decision was therefore taken to target systems
with a high degree of synthetic accessibility. With this in mind atten-
tion was focused on the previously reported SSRI litoxetine3, 4—a po-
tent, selective inhibitor with a very low level of synthetic complexity.
We have previously reported how, using litoxetine as a starting
point we identified a range of selective SSRIs, illustrated by com-
pounds 5–8 (Fig. 2).4 These systems were designed based on the
simple principles of maintaining the relative orientation of the
naphthyl (benzene) ring to a basic centre (vs litoxetine) and ensur-
ing that targets had high synthetic accessibility. Encouraged by this
success we utilised the same principles to design and prepare a sec-
ond wave of targets, 9–12 (Fig. 3).5
Key data for representative examples of each of these series,
including serotonin re-uptake inhibition (SRI) potency, selectivity
over the related Dopamine and Noradrenalin transporters (Dopa-
mine re-uptake inhibition and Noradrenaline re-uptake inhibition,
6
DRI and NRI, respectively) and binding to the IKr channel7,8 are
detailed in Table 1.
Key SAR points for each of these series are as follows.
For the naphthoic acid amides 13–20 good SRI potency could be
attained provided the amide was tertiary. Molecular modelling
suggested that such systems were significantly twisted out of the
plane around the amide bond/aromatic ring. By way of contrast
secondary amide 13 modelled as essentially flat around the amide
bond/aromatic ring and was found to have no significant SRI activ-
ity. In these naphthoic acid amides it was possible to attain excel-
lent levels of SRI potency as well as NRI, DRI and IKr selectivity in
both 4-piperidine and 3-(R)-pyrrolidine based analogues (see, e.g.,
compounds 15 and 19, respectively). In the 3-(R)-pyrrolidine ser-
ies, it was possible to replace the naphthyl substituent with a suit-
ably substituted phenyl substituent whilst retaining good levels of
potency and selectivity (see, e.g., 23). However, it was found that in
O
N
H
Litoxetine, 4
SRI 5.9nM; DRI 6.7μM, NRI 1.4μΜ
* Corresponding author. Tel./fax: +44 1304648240.
E-mail address: Alan.D.Brown@pfizer.com (A. Brown).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.08.066

5894
M. Andrews et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5893–5897
NHMe
NH
O
NHMe
O
CF3
O
O
F
Cl
Cl
Fluoxetine
3
Sertraline
1
Paroxetine
2
Figure 1. Representative approved SSRIs.
O
N
H
Litoxetine, 4
H
Me
O
O
O
OMe
N
F
NC
F
N
H
N
H
N
H
N
H
5
6
7
8
Figure 2. Litoxetine and analogues with high synthetic accessibility previously reported by our group.
Table 1
Potency and selectivity of new SRIs
the 3-(S)-pyrrolidine naphthoic acid amides larger right hand side
substituents often resulted in somewhat reduce NRI selectivity (as
illustrated by compound 25).
Structure
SRI IC50
5.9 nM
NRI IC50 DRI IC50 IKr Ki
O
The corresponding 3-(S)-pyrrolidine naphthoic acid amides
(e.g., 20) typically carried significant NRI activity as compared to
the corresponding 3-(R) enantiomers (compare 20 with 19). As a
result these compounds were not pursued further for our SRI
programme.
N-Naphthylmethylsulphonamides 26–29 showed excellent SRI
activity. Once more the 4-piperidine and 3-(R)-pyrrolidine
analogues typically carried good to excellent selectivity (see, e.g.,
compounds 27 and 28, respectively). In contrast the corresponding
3-(S)-pyrrolidine systems (as in compound 29) typically carried in-
creased NRI activity.
N-Naphthylmethylamides also demonstrated excellent levels of
SRI potency and selectivity, both in 4-piperidine based systems
(30, 31) and in the corresponding 3-(R)-pyrrolidine analogues (33
and 34). Interestingly, increasing the right hand side substituent
from acetyl to propionyl (30 to 31 and 33 to 34) gave significantly
increased DRI activity. Indeed 31 and 34 displayed amongst the
highest degrees of DRI activity observed throughout our work on
this litoxetine based programme.
Finally, replacement of the right hand side acetyl substituent
with a pyrimidine gave 32 which also demonstrated good levels
of SRI activity, albeit with a somewhat reduce NRI selectivity
window.
1.4
l
M
6.7
l
M
n.t.a
n.t.a
12.7
N
H
Litoxetine, 4
O
NH
>400 nM n.t.a
n.t.a
N
H
13
O
O
N
28.3 nM
>40
l
M
>40
l
M
lM
N
H
14
N
The preparation of a representative analogue, compound 25 is
described in Scheme 1.9 Commercially available 2,3-dichlorophe-
nol was regioselectively carboxylated using the Kolbe–Schmitt
reaction10 to give 36. Conversion to the corresponding ester/triflate
gave 37 which smoothly underwent Suzuki coupling with trim-
ethylboroxine. Hydrolysis then gave key 1,2,3,4-substituted
10 nM
>40
lM
>40
lM
>10 lM
N
H
15

M. Andrews et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5893–5897
5895
Table 1 (continued)
Table 1 (continued)
Structure
SRI IC50
15 nM
NRI IC50 DRI IC50 IKr Ki
Structure
SRI IC50
NRI IC50 DRI IC50 IKr Ki
O
O
Me
Cl
Cl
N
N
7.5
l
M
>14
l
M
n.t.a
n.t.a
n.t.a
7 nM
5.7
l
M
>40
>28
>68
>28
l
l
l
l
M
M
M
M
n.t.a
N
N
H
H
24
16
O
Me
O
Cl
Cl
N
N
15 nM
17 nM
4.6 nM
357 nM
n.t.a
9 nM
12 nM
5 nM
3 nM
6
l
M
5.8
7.4
9.5
lM
N
H
N
H
25
17
SO2Me
N
O
N
7
l
M
4.2 lM
N
677 nM
lM
H
N
H
26
18
Me
SO2Me
N
O
735 nM
1.6
l
l
M
F
N
N
1.2
lM
l
M
7 lM
H
27
N
H
19
SO2Me
N
O
3.5 nM
3.5 nM
4.5 nM
1.0 nM
10 nM
2.5
l
M
4.9
2.8
1.5
l
M
3.2
M
N
N
H
24 nM
744 nM
n.t.a
n.t.a
n.t.a
n.t.a
28
29
N
H
SO2Me
20
N
370 nM
l
M
11 lM
O
N
H
Cl
Me
N
>400 nM n.t.a
n.t.a
Cl
O
N
H
N
21
8.9
1.3
1.2
l
l
l
M
M
M
l
M
>20 lM
O
O
Me
N
H
Cl
Cl
Me
N
30
58 nM
n.t.a
n.t.a
O
N
N
H
N
22
Me
339 nV
12 lM
N
H
31
32
Cl
Cl
N
10 nM
17.4
lM
>40 lM
N
N
N
H
1.3
l
M
n.t.a
23
N
H
(continued on next page)

5896
M. Andrews et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5893–5897
benzoic acid 38 which was coupled (via the acid chloride) to BOC
Table 1 (continued)
diamine 39. Acid deprotection then furnished benzamide 25.
In summary, using litoxetine as a starting point we have identi-
fied a range of potent, highly selective SSRIs all of which have a
high degree of synthetic accessibility, compared to existing SSRIs
such as Sertraline, Paroxetine and Fluoxetine. Compounds 15, 19,
28, 31 and 33 in particular, represent attractive potential leads in
the search for a short Tmax SSRI.11
Structure
SRI IC50
5.2 nM
NRI IC50 DRI IC50 IKr Ki
O
N
726 nM
>12
l
M
8 lM
N
H
33
Acknowledgements
O
We would like to acknowledge the contributions of the follow-
ing co-workers: James Gosset, Hugh Verrier and Robin Ward.
N
2.4 nM
224 nM
266 nM
n.t.a
N
H
References and notes
34
1. Preskorn, S. H.; Ross, R.; Stanga, C. Y. Handbook of Experimental Pharmacology;
Springer, 2004. 157 (Antidepressants), 241.
a
Not tested.
2. For more background on this short Tmax based approach see: Middleton, D. S.;
Andrews, M. D.; Glossop, P.; Gymer, G.; Hepworth, D.; Jessiman, A.; Johnson, P.
O
O
R
N
N
R
( )1,2
( )1,2
N
N
H
9
H
11
O
N
H
O
Litoxetine
R'
R'''
SO2R
( )1,2
N
N
( )1,2
R''
N
N
H
12
H
10
Figure 3. A second wave of targets with high synthetic accessibility based on the SSRI litoxetine.
OH
OH
OTf
Cl
Cl
CO2H
Cl
Cl
CO2Me
Cl
Cl
(i)
(ii), (iii)
37
36
35
(iv), (v)
H
N
Me
O
Me
38
(vi), (Vii)
(viii)
N
Cl
Cl
39
CO2H
Cl
Cl
N
boc
N
H
25
Scheme 1. Synthesis of compound 25. Reagents and conditions: (i) K2CO3/CO2, 100 atm (71%); (ii) TMSCHN2, toluene/MeOH (quant); (iii) Tf2O, Et3N, DCM; (iv)
trimethylboroxine, Pd(Ph3)4, K2CO3, aq Dioxan, (quant, two steps); (v) 1 N NaOH (quant); (vi) ClCOCOCl, Et3N, DCM, DMF; (vii) 39, Et3N, DCM (85%, two steps); (viii) HCl/DCM
(quant).

M. Andrews et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5893–5897
5897
J.; MacKenny, M.; Stobie, A.; Tang, K.; Morgan, P.; Jones, B. Bioorg. Med. Chem.
Lett. 2008, 18, 5303. and the references cited therein.
3. Angel, I.; Schoemaker, H.; Prouteau, M.; Garreau, M.; Langer, S. Z. Eur. J.
Pharmacol. 1993, 232, 139.
measuring the amount of radioactivity taken up into the cells by scintillation
counting. Potency of test compounds was quantified as IC50 values, that is,
concentration required to inhibit the specific uptake of radiolabelled substrate
into the cells by 50% relative to maximum (vehicle only) over a 10-point dose
response range. A minimum of three experiments were made. Monoamine
reuptake IC50 values are geometric means of at least three experiments. Values
within twofold of each other should be considered equivalent. (b) IKr data
reflects the inhibition of binding of 3H-radiolabelled dofetilide to human hERG
Potassium channels expressed in HEK-293 cells. Ki values are geometric means
of at least two experiments. Values within twofold of each other should be
considered equivalent.
4. (a) Andrews, M.; Brown, A.; Chiva, J-Y.; Fradet, D.; Gordon, D.; Lansdell, M.;
MacKenny, M. Bioorg. Med. Chem. Lett. 2009, 19, 2329; For a recent review of
new chemical entities see: (b) Walter, M. W. Drug Dev. Res. 2005, 65, 97.
5. The perceived high degree of synthetic accessibility of pyrrolidine containing
targets was based upon the (wide) range of commercially available, suitably
protected, homochiral 3-substituted pyrrolidines. For example, (S) and (R)-3-
(BOC-amino)pyrrolidine (the homochiral starting materials utilised in the
synthesis of all of the chiral compounds described in this Letter) are both
available from Sigma–Aldrich in P98% purity (P96% ee).
6. (a) SRI, NRI and DRI data reflects inhibition of the relevant (radiolabelled)
monoamine uptake into HEK-293 cells transfected with the appropriate human
monoamine transporter. Specifically, these assays were a modification of those
described by Blakey et al. Anal. Biochem. 1991, 194, 302. HEK293 cells
7. For a review of IKr (hERG) affinity and the issues that it presents in small
molecule drug discovery, see: Jamieson, C.; Moir, E. M.; Rankovic, Z.; Wishart,
G. J. Med. Chem. 2006, 49, 5029.
8. In-house experience suggested that, for certain SRI series, IKr affinity can
represent a significant issue. We were therefore keen to identify at an early
stage whether or not this issue was likely to arise in any newly identified lead
series.
expressing
a single human amine transporter protein (7500 cells/well in
Milliporte 96-well filter bottom plates) were pre-incubated at 25 °C for 5 min
with assay buffer containing vehicle (DMSO in water) or test compound.
Uptake of neurotransmitter into the cells was initiated by the addition of
tritiated 5-HT (50 nM), Noradrenaline (200 nM) or Dopamine (200 nM)
substrates. The samples were shaken in an incubator at 25 °C for 5 min (5-
HT, Dopamine) or 15 min (Noradrenaline). The assays were stopped by an ice-
cold buffer wash followed by filtration. The filters were then dried before
9. In fact, the synthesis of 25, straightforward as it was, represents, by some way,
the most complex synthesis (in terms of total steps) carried out during the
course of the work described in this Letter, where our focus was very much on
accessing new templates which had a high degree of synthetic accessibility.
10. Lindsey, A. S.; Jesley, H. Chem. Rev. 1957, 57, 583.
11. Our subsequent efforts in this area will be disclosed in due course.