Hit-to-Lead studies: The discovery of potent, orally bioavailable triazolethiol CXCR2 receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(03)00561-4

A Hit-to-Lead optimisation programme was carried out on the high throughput screening hit, the triazolethiol 1, resulting in the discovery of the potent, orally bioavailable triazolethiol CXCR2 receptor antagonist 45.

Full text of DOI:10.1016/S0960-894X(03)00561-4

Bioorganic & Medicinal Chemistry Letters 13 (2003) 2625–2628
Hit-to-Lead Studies: The Discovery of Potent, Orally Bioavailable
Triazolethiol CXCR2 Receptor Antagonists
Andrew Baxter,* Colin Bennion, Janice Bent, Kerry Boden, Steve Brough, Anne Cooper,
Elizabeth Kinchin, Nicholas Kindon, Tom McInally, Mike Mortimore, Bryan Roberts
and John Unitt
AstraZeneca R&D Charnwood, Bakewell Road, Loughborough LE11 5RH, UK
Received 11 April 2003; revised 4 June 2003; accepted 5 June 2003
Abstract—A Hit-to-Lead optimisation programme was carried out on the high throughput screening hit, the triazolethiol 1,
resulting in the discovery of the potent, orally bioavailable triazolethiol CXCR2 receptor antagonist 45.
# 2003 Elsevier Ltd. All rights reserved.
The use of High Throughput Screening (HTS) is now
widespread in the pharmaceutical industry. There was
an expectation that, once a screen was established for a
particular target, then potent lead compounds or can-
didate drugs would be found. The reality is often far
from this. Bridging the gap between the end of a HTS
and the start of a full Lead Optimisation (LO) project
has been described as Hit-to-Lead (HtL).1 Hits from
HTS are profiled and compared to a generic target lead
criteria. The lead target profile used is shown in Figure 1.
Lead series then have a balance of properties — potency
and SAR an encouraging metabolic and selectivity pro-
file—such that rapid (less than 2 years) further
optimisation should provide Candidate Drugs (CDs).
orders, including asthma and allergic disease, as well as
autoimmune pathologies such as rheumatoid arthritis
and atherosclerosis. These small secreted molecules are
a growing superfamily of 8–14 kDa proteins char-
acterised by a conserved four-cysteine motif. The
chemokine superfamily can be divided into two main
groups exhibiting characteristic structural motifs, the
Cys-X-Cys (CXC) and Cys-Cys (CC) families.2 The
CXC chemokines include several potent chemo-
attractants and activators of neutrophils such as inter-
leukin-8 (IL8), GROa and neutrophil activating peptide
2 (NAP2).
Studies have demonstrated that the actions of the
chemokines are mediated by subfamilies of G protein-
coupled receptors, among which are the receptors
designated CCR1 to CCR10 and CXCR1 to CXCR5.
These receptors represent good targets for drug devel-
opment since agents which modulate these receptors
would be useful in the treatment of disorders and dis-
eases such as those mentioned above. The CXCR2
receptor is one of the receptors that IL8 activates and
antagonists of this receptor should find particular use in
the treatment of a number of diseases.
Chemokines play an important role in immune and
inflammatory responses in various diseases and dis-
A HTS was undertaken to identify compounds that
blocked the binding of [125I]-IL8 to human recombinant
CXCR2 (hrCXCR2) expressed in HEK 293 membranes
using a Scintillation Proximity Assay (SPA). Subse-
quently at the hit evaluation stage compounds with
binding inhibition seen in the HTS SPA assay were
initially validated in a more conventional filter wash
Figure 1. Hit-to-Lead generic lead target profile.
*Corresponding author. Tel.: +44-1509-644772; fax: +44-1509-
645513; e-mail: andrew.jg.baxter@astrazeneca.com
0960-894X/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00561-4

2626
A. Baxter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2625–2628
hrCXCR2 [125I]-IL8 binding assay. Confirmation of
functional antagonism was shown by blockade of
GROa stimulated intracellular calcium mobilisation in
and its size and steric environment are essential for
receptor binding as smaller or sterically demanding
acidic groups (OH, NHAc and NHSO2Ph) are inactive.
All subsequent analogues prepared in HtL preserved the
triazolethiol structure.
isolated human neutrophils using
a Fluorescence
Imaging Plate Reader (FLIPR).3
The triazole 1 emerged inter alia from the CXCR2 HTS
and, while having only modest potency in the binding
assay (IC50 4.6 mM), had demonstrable and comparable
functional activity in the FLIPR assay (IC50 2.4mM).
The profile of this CXCR2 hit is in Table 1 compared
with the generic lead target profile. The DMPK profile
was satisfactory at this stage and HtL focused on
exploring SAR to increase potency and assessing the
importance of the thiol-thione functionality.
Table 2. CXCR2 antagonist binding potencies
R
X
CXCR2 IC50 (mM)a
2
3
4
5
6
7
8
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
SH
OH
NH2
H
SMe
2.4
NA
NA
NA
NA
NA
NA
Table 1. Hit profile of triazolethiol 1
NHAc
NHSO2Ph
aNA,<50% inhibition at 10 mM.
Generic lead criteriaa
Triazolethiol 1
Next variation of the 2-benzyl substituent was investi-
gated. Published routes to these compounds utilise the
reaction of 2-substituted thiosemicarbazides with acid
chlorides,4 or acylisocyanates with hydrazines7 followed
by cyclisation (Scheme 1a and b). The compounds 9–14
in Table 3 were prepared from commercially available
starting materials and offered no improvements in
potency over 2-benzyl. Phenyl, methyl and hydrogen
were all inactive and only the 3-hydroxybenzyl analogue
14 had any activity. The substitution on the 5-phenyl
was investigated next, keeping the 2-benzyl constant,
using the route described above. Table 3 lists the com-
pounds prepared (15–37) and their activity. Various
substitutions were seen as potency enhancing, in partic-
ular 2-chloro (34), 2,4-dichloro (35), 2-bromophenyl
(36) and 2,3-dichloro (37) gave 10-fold increases in
potency compared to the unsubstituted phenyl (2).
Binding IC50<0.1 mM
Ca Flux IC50<0.1 mM
Rat hepatocyte Cl<14
Human microsome Cl<23
Rat iv Cl<35
Rat iv Vss>0.5 L/kg
Rat iv T1/2>0.5 h
Molecular weight<450
ClogP<3.0
4.6 mM
2.4 mM
19
13
12
1.3
3.4
268
3.4
aUnits as Figure 1 where not stated.
Replacement of the pyridinyl ring by phenyl (compound
2)4 gave a slightly more potent analogue which was the
starting point for variation of the thiol (Table 2). Ana-
logues with hydroxy (3), amino (4), and without the
thiol (5) were prepared using literature methods;5ꢀ7 all
were inactive as CXCR2 antagonists. Methylation of
the thiol (6) and acetylation (7) and benzenesulphonyl-
ation (8) of the amino also gave inactive compounds. It
is assumed that the acidic nature of the thiol (pKa ꢁ6.5)
In order to look for further variations of substituents on
the 2-benzyl group, it was necessary to find an alter-
native route that would allow simpler starting materials
Scheme 1. (i) Pyridine, 16 h, rt; (ii) 1 M NaHCO3, 16 h, 100 ꢂC; (iii) Et3N, toluene, 1 h, 80 ꢂC; (iv) aldehyde, MeOH/HCl; (v) Et3SiH, TFA, 0 ꢂC;
(vi) EtOH/HCl, NH4SCN, 18 h, 75 ꢂC.

A. Baxter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2625–2628
2627
to be used. Benzylation of the 3-H compound (11) pro-
ceeds first on sulphur and a di-benzylation followed by
S-debenzylation route did not work in our hands. The
route followed6 is shown in Scheme 1c. Substituted
aldehydes were condensed with benzoylhydrazides to
give the corresponding hydrazones. These hydrazones
were then reduced using triethylsilane in trifluoroacetic
acid to give the substituted benzoylhydrazines. Reaction
with ammonium thiocyanate in hot ethanol containing
hydrogen chloride gave the intermediate semi-
carbazides, which were cyclised in hot sodium bicarbo-
nate solution to give the required triazolethiols (38–45).
A number of analogues were prepared keeping the 5-
substituent constant as 2,4-dichlorophenyl; most potent
of these was the 3-chlorobenzyl having IC50 0.092 mM.
Checking back to other potent 5-substituents led to the
preparation of the lead compound, the 2-(3-chloro-
benzyl)-5-(2-chlorophenyl) analogue (45) (Table 3).
of the two phenyl rings is seen as important. In the 5-
position of the triazole (Table 3), cyclohexyl (15) and
methyl (16) were inactive whilst almost all aromatic
groups tried had some antagonist activity. Phenyl and
thiophene (2 and 26) were more potent than pyridyl and
furan (19 and 20). In general it was found that analo-
gues with a 2-substituted phenyl were the most potent
and that the preferred substituent was chlorine or bro-
mine. 2-Chlorophenyl became the 5-substituent of
choice. 2-Chloro (34) was better than 2-fluoro (30), 2-
methoxy (28) or 2-methyl (27) and 2-hydroxy (18) was
inactive possibly due to an internal hydrogen bond with
the triazole. 2-Chloro was preferred to 2-bromo because
of reduced lipophilicity and better pharmaceutical
acceptability. Some of the potency enhancement could
be lipophilicity driven but the 2-chloro substituent
would be expected to cause a twist out of plane between
the phenyl and triazole. With a single substituent this
could be up to 30 ꢂ without loss of resonance between
the rings. In general, addition of 3- or 4-substituents did
not cause appreciable potency changes; for example.
compare 2-chloro (34) with 2,3-, 2,4-, 2,5-dichloro (37,
35, 33). Optimisation at the 2-position was undertaken
in two separate series (5-phenyl analogues 2, 9–14) and
5-(2,4-dichlorophenyl) analogues 35, 38–44). Initially
only benzyl (2) gave any activity, phenyl, methyl and
hydrogen being inactive (9–11). Activity was removed
by the addition of 2-chloro (13) and 4-trifluoromethyl
(12), only 3-hydroxy (14) maintained activity. The pref-
erence for a 3-substituent was confirmed and extended
in the more potent 5-(2,4-dichlorophenyl) series. 3-
Chloro (44) gave an increase in potency compared to
the unsubstituted compound (35), other 3- and 4-sub-
stitutuents were less active (38–43). One interesting
observation was that the phenethyl analogue (41) had
similar potency to benzyl (35).
Some preliminary SAR conclusions can be drawn on
the basis of the compounds prepared in this paper. The
substitution patterns, orientation in space and presence
Table 3. CXCR2 antagonist binding potencies
R
R0
CXCR2 IC50 (mM)a
2
9
PhCH2
Ph
Me
Ph
Ph
Ph
Ph
Ph
Ph
2.4
NA
NA
NA
NA
NA
4.4
NA
NA
NA
NA
7.7
4.2
3.5
3.5
2.8
2.3
2.0
2.0
1.4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
H
4-CF3PhCH2
2-ClPhCH2
3-OHPhCH2
PhCH2
Ph
Cyclohexyl
Me
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
Investigation of the role of the central triazole ring was
also undertaken. A number of heterocycles substituted
with phenyl and benzyl were obtained from the corpor-
ate compound collection and from external suppliers
but none showed any activity (data not shown). Limited
variation of the 3-position on the triazole was carried
out but only 3-thiol had any activity (2 vs 3–8). Initially
the acidic nature of the thiol (pKa 6.20) was thought to
be important. This trend has been noted by others;
SB2250028 and other urea based CXCR2 antagonists9
are all acidic in nature (Fig. 2). The role of the anionic
group has been investigated further by this group,10
finding that alteration of the phenolic pKa to give
uncharged molecules gave greatly reduced potency. In
the case of the triazoles though more subtle SAR exists.
The hydroxy compound 3 has appreciable ionised form
(pKa 7.80) and the sulphonamide 8 (pKa 6.71) is even
more acidic but both lack activity. With the sulphona-
mide, the presence of a large extra substituent might be
4-MePh
2-OHPh
4-Pyridinyl
2-Furanyl
4-CNPh
3-CF3Ph
4-CF3Ph
4-MeOPh
3,5-DiClPh
2-Thienyl
2-MePh
2-MeOPh
3-ClPh
1.4
1.0
2-FPh
4-ClPh
3,4-DiClPh
2,5-DiClPh
2-ClPh
0.89
0.83
0.80
0.67
0.45
0.41
0.35
0.35
10
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
2,4-DiClPh
2-BrPh
2,3-DiClPh
2,4-DiClPh
2,4-DiClPh
2,4-DiClPh
2,4-DiClPh
2,4-DiClPh
2,4-DiClPh
2,4-DiClPh
2-ClPh
4-MeOPhCH2
3-MeOPhCH2
3-MePhCH2
PhCH2CH2
4-ClPhCH2
3-PhOPhCH2
3-ClPhCH2
3-ClPhCH2
4.2
0.73
0.45
0.30
0.17
0.092
0.028
aNA, <50% inhibition at 10 mM.
Figure 2. Structure of acidic urea CXCR2 antagonists.

2628
A. Baxter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2625–2628
expected to interfere with binding to the receptor but at
present there is no explanation of the lack of activity of
the hydroxy compound 3.
clearly shows scope for further improvement in potency
by looking for additional interactions via further aro-
matic substituents and alteration of linker group
between the triazole and phenyl group in the 2-position.
Selectivity data was acceptable and analogues with sub-
stituents on both aromatic rings are novel. The triazo-
lethiols exemplified by compound 45 formed the basis of
a new LO project.
Table 4. Lead profile of triazolethiol 45
Generic lead criteriaa
Triazolethiol 45
References and Notes
Binding IC50<0.1 mM
Ca flux IC50<0.1 mM
Rat hepatocyte Cl<14
Human microsome Cl<23
Rat iv Cl<35
0.028 mM
0.048 mM
26
14
12
8
1. Michne, W. F. Pharm. News 1996, 3, 19.
2. For reviews, see: (a) Rossi, D.; Zlotnik, A. Annu. Rev.
Immunol. 2000, 18, 217. (b) Horuk, R.; Ng, H. P. Med. Res.
Rev. 2000, 20, 155. (c) Trivedi, B. K.; Low, J. E.; Carson, K.;
LaRosa, G. J. Ann. Rep. Med. Chem. 2000, 35, 191. (d) Onuf-
fer, J. J.; Horuk, R. Trends Pharmacol. Sci. 2002, 23, 459.
3. For full details of the biological assays used, see: Baxter,
A.; Bennion, C.; Cage, P.; Kindon, N.; Mortimore, M.;
Roberts, B. Int. Patent Appl. WO 00/12489.
4. Durant, G. J.; Kline, S. J. Chem. Soc. C 1967, 92.
5. Tsuge, O.; Kanemasa, S.; Tashiro, M. Tetrahedron 1968,
24, 5205.
6. Davidson, J. S.; Dhami, S. S. Chem. Ind. (London) 1978,
92.
Rat iv Vol>0.5 L/kg
Rat iv T1/2>0.5 h
9
Rat po bioavail.>10%
Plasma prot. bind.<99.5%
Molecular weight<450
Solubility>10 mg/mL
clogP<3.0
61%
99.0%
336
20
4.3
3.2
Log D<3.0
aUnits as Figure 1 where not stated.
7. Behringer, H.; Ramert, R. Liebigs Ann. Chim. 1975, 1264.
8. Podolin, P. L.; Bolognese, B. J.; Foley, J. J.; Schmidt, D. B.;
Buckley, P. T.; Widdowson, K. L.; Jin, Q.; White, J. R.; Lee,
J. M.; Goodman, R. B.; Hagen, T. R.; Kajikawa, O.; Mar-
shall, L. A.; Hay, D. W. P.; Sarau, H. M. J. Immunol. 2002,
169, 6435.
9. White, J. R.; Lee, J. M.; Young, P. R.; Hertzberg, R. P.;
Jurewicz, A. J.; Chaikin, M. A.; Widdowson, K.; Foley, J. J.;
Martin, L. D.; Griswold, D. E.; Sarau, H. M. J. Biol. Chem.
1998, 273, 10095.
The profile of the lead compound (45) is shown in Table 4
compared with the lead target profile. Compound (45)
has good lead-like potency, both binding and functional
in the calcium flux assay, and the in vitro rat hepatocyte
and human microsome data was acceptable. It is a weak
acid (pKa 6.4) and has borderline lipophilicity and this is
reflected in the high plasma protein binding. However
the compound has maintained the satisfactory in vivo
DMPK profile having a 9 h intra venous half-life and
over 60% oral bioavailability. The SAR presented
10. Widdowson, K.; Nie, H.; Jurewicz, A. J.; Hertzberg, R. P.;
Sarau, H. M.; Foley, J. J.; Lee, J.; White, J. R.; Veber, D. F.
Lett. Peptide Sci. 1998, 5, 235.