I. R. Greig et al. / Bioorg. Med. Chem. Lett. 23 (2013) 816–820
817
Scheme 1. Synthesis of triarylsulfonamides 4a–4v, 4ah–4ai: Reagents and condi-
tions (a) (3-aminophenyl)methanol, DCM, pyridine (room temp, 18 h); (b) substi-
tuted benzeneboronic acid, (PPh3)4Pd, ethanol, toluene, 2 M Na2CO3, (reflux, 3 h).
Figure 2. Structures of rigidified biphenylsulfonamides.
a more promising starting point, having in vitro potency more than
ten-fold higher than their ester and ketone analogues and better
in vitro stability (30–60 min in human liver microsomes); how-
ever, these compounds showed low oral bioavailability (<5%) and
had poor activity in the collagen-induced arthritis model,13 thus
they were not pursued as developmental candidates.
Scheme 2. Synthesis of triarylsulfonamides 4w–4ag: Reagents and conditions (a)
ethyl 3-aminobenzoate, DCM, pyridine (room temp, 18 h); (b) substituted benzen-
eboronic acid, (PPh3)4Pd, ethanol, toluene, 2 M Na2CO3, (reflux, 3 h); (c) 1 M LiAlH4,
THF, (50 °C, 5 h).
To investigate the potential for improving the properties of the
biphenylsulfonamides, we took the initial approach of incorporat-
ing a phenyl group to rigidify the C4 or C5 alkyl chain, giving tria-
rylsulfonamides of the compound classes 4–7 (Fig. 2), by the
methods shown in Schemes 1 and 2. When the alkyl chain in the
ketone class of compounds 2, was replaced by a phenyl group, to
give a triarylketone, no improvement in potency was obtained;13
however, with the triarylsulfonamides a modest but promising in-
crease in potency was seen in our primary assay (J774 macro-
phages viability, 72 h exposure to the test compound11):
compounds 4b, 5a, 6a and 7a (Fig. 2) had IC50 = 2.5, 3.5, 5.5 and
4.5
lM, respectively. This compares to the 8 and 9 lM of the 4-car-
bon and 5-carbon alkylsulfonamide analogues (3).12 Compounds
Figure 3. Structures of compounds 4al–4ak, 5b, 6b–6c.
8a and 9a (Fig. 2) showed poor activity (IC50 >15 lM). As antici-
pated from previous studies,11 replacement of the sulfonamide
moiety in 4a and 4j with a carboxamide also gave compounds with
poor activity (4al and 4am, IC50 >15 lM, Fig. 3). We conducted a
methyl group proved ideal, giving a further 3 to 4-fold increase
in potency and compound 4t showed 25 nM potency against
J774 macrophages. Representative derivatives based on other
substituents including 4-methyl (4v), 6-chloro (4w), 2-chloro
(4x), 2-hydroxy (4ad), and 5-hydroxymethyl (4ae) gave a decrease
in potency; 4-chloro (4y) and 4-methoxy (4z) had little effect on
potency. 4-Hydroxy (4ac) or 6-hydroxymethyl (4ag) substitution
gave small increases in potency; when combined with the 20,40-
difluorobiphenyl motif these also looked promising candidates
for further development.
structure activity relationship (Table 1) to investigate the potential
for the further development of compounds in classes 4–6. Like
osteoclasts, J774 macrophages are dependent on continued NFB
activation for survival, thereby providing a valuable screen for
compounds with anti-inflammatory activity.15
Depending on the commercial availability of starting materials,
compounds were synthesized by reaction of an appropriately-
substituted 4-bromobenzenesulfonyl chloride with a (3-amino-
phenyl)methanol (compounds 4a–4v and 4ah–4ai), (4-aminophe-
nyl)methanol (5a–5b) or with 2-(4-aminophenyl)ethanol (7a),
and subsequent Suzuki coupling with the required benzeneboronic
acid. 40-Bromobiphenyl derivatives were prepared directly from 40-
bromobiphenylsulfonyl chloride synthesized as described
previously.12,16
Alternatively, 4-bromobenzenesulfonyl chloride was reacted
with the required ethyl 3-aminobenzoate (4w–4ag) or ethyl 2-
(3-aminophenyl)acetate (6a–6c); Suzuki coupling with a benzen-
eboronic acid was then performed as before, and the product re-
duced to the target alcohol in the final step, as shown in
Schemes 1 and 2.
Activity was validated in our secondary assay, mouse osteoclast
formation, in
a
mouse osteoblast–bone marrow co-culture.11
Compounds showed similar potency and rank order to that seen
in the J774 assay, from 4b IC50 = 3000 nM to 4t = 35 nM (osteo-
clasts identified by TRAcP staining11 and subsequent cell counting).
4k and 4s also showed little effect on mouse osteoblast viability
after 72 h, as measured by AlamarBlue.11 As osteoclasts and
osteoblasts are responsible for bone resorption and formation,
respectively, this represents a good therapeutic window and sug-
gests low general toxicity. Further studies in human hepatocytes
demonstrated that there was no reduction in cell viability after
72 h incubation with 50 lM 4j or 4k in an MTT assay. Results are
shown in Table 2. For further assay details, see Supplementary
The SAR of compounds in class 4 showed broad similarities to
those described for compound classes 1–3; thus only a small num-
ber of derivatives were made and the focus was on hydrophobic,
electron-withdrawing substituents. As for class 3, a 20,40-difluoro
(4j; 300 nM) or 20,40-dichloro (4k; 500 nM) substitution on the left
hand ring and 3-methyl group on the middle ring led to the most
potent compounds (4s; 80 nM), giving a combined 25-fold increase
in potency over the unsubstituted parent compound. The addition
of a third ring gave further opportunities for optimisation; a 2-
data.
The SAR of compounds from class 5 was less clear-cut: and the
20,40-difluorobiphenyl derivative (5b, Fig. 3) did not show the same
degree of improvement in potency (3 lM) as was found with com-
pound 4j and these were not further investigated. Compounds
from classes 6 and 7 showed poor metabolic stability, for example,
the 20,40-difluorobiphenyl and 20,40-dichlorobiphenyl derivatives
6b and 6c (Fig. 3) had half lives of just 17 and 13 min, respectively,
in human liver microsomes.
Greig, Iain R.
