Chemistry Reference
In-Depth Information
with maintaining properties that did not compromise the excellent CNS
penetration of the known agents. The majority of CNS drugs have a low
number of hydrogen bond donors (HBD) (all of the top 25 selling CNS drugs
from 2004 had
o
3 HBD), lower polar surface area (PSA) (96% of the top 25
CNS drugs had TPSA
o
90A
˚
2
) and relatively high clog P (68% of the top 25
CNS drugs had clog P values between 2 and 5).
17
As a result, we focused on
compounds that had the following properties: MW
o
400, log D
¼
2-3, HBD
count
o
3 and TPSA
o
80A
˚
2
.
While, in principle, this strategy is equally applicable to all lipophilic SSRIs,
several factors led us to select compound 3 as our starting point: (1) 3 has
proven ecacy in PE clinical trials; (2) published
18
and in-house structure-
activity relationships (SAR) in this series gave us a clear understanding of the
structural features essential for potency; (3) the template is amenable to syn-
thetic modifications, allowing rapid SAR development; (4) 3 has low CYP
inhibition and low CYP2D6 anity,
19
minimising the risk of drug-drug
interactions and variable inter-subject metabolism.
Based on the known SAR around 3, we did not try to vary the dichlorophenyl
or cis-methylamino substituents, but concentrated on substituting on the aro-
matic part of the tetrahydronaphthalene ring.
20,21
We quickly found that polar
substitution is well tolerated, with substitution at C8 (6-8), C7 (9-18) and C6
(19-23) leading to potent SSRIs (selected examples shown in Table 7.1).
Interestingly, the acid intermediates (6, 9, 19) used to make amides at the three
positions all retain significant SRI activity. As these compounds are zwitter-
ionic, they would be expected to have an inherently low V
D
,
22
but unfortunately
they did not meet our potency criteria and, ultimately, were not investigated
further. The acidic substituents also raised concerns as to how CNS penetrant
they would be. Substitution at C8 resulted in a loss in selectivity over DRI (e.g.
36-fold for 7), so focus was given to investigating substitution at C7 and C6. In
general, substitution at C7 gave compounds with the best SRI activity, while
substitution at C6 gave compounds with the best selectivity over DRI and NRI
(compare 14 with 23, 11 with 21, 12 with 22). As the C7 position was synthe-
tically the most accessible and many analogues were found with acceptable
selectivity versus DRI and NRI, this became the focus of the project's efforts and
ultimately led to the compounds that were progressed for further studies.
Both the primary and secondary amides (10, 11) are potent SSRIs, but the
tertiary amide lost some SRI potency and as a result lost selectivity over DRI
and NRI. Nitrile substitution (13) resulted in an SSRI which is more lipophilic
than our starting point 3 (log D 3.4 vs. 3.1) and so did not meet our goal of
reducing log D. This is presumably a result of the electron-withdrawing nitrile
group reducing the basicity of the amine and therefore increasing the log D.
Primary and secondary sulfonamides (14, 15) are both potent SSRIs, with the
primary sulfonamide being one of the most potent compounds we identified
(1 nM). As with the amides, the tertiary sulfonamide 16 is a weaker SSRI than
the primary and secondary sulfonamides, but in this case the drop-off was
greater than 10-fold. The N-linked sulfonamide 17 is also potent and polar, but
has unacceptably high levels of DRI and NRI activity (less than 20-fold
Search WWH ::
Custom Search