SAR studies on thiazolo[4,5-d]pyrimidine based CXCR2 antagonists involving a novel tandem displacement reaction

Article abstract of DOI:10.1016/j.bmcl.2007.02.080

As part of a Lead Optimisation programme to identify small molecule antagonists of the human CXCR2 receptor, a series of substituted thiazolo[4,5-d]pyrimidines was prepared via the application of a novel tandem displacement reaction.

Full text of DOI:10.1016/j.bmcl.2007.02.080

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2731–2734
SAR studies on thiazolo[4,5-d]pyrimidine based CXCR2
antagonists involving a novel tandem displacement reaction
Fraser Hunt,* Caroline Austin, Rupert Austin, Roger Bonnert, Peter Cage,
Jadeen Christie, Mark Christie, Clare Dixon, Steven Hill, Robert Jewell, Ian Martin,à
David Robinson and Paul Willis
AstraZeneca R&D Charnwood, Bakewell Road, Loughborough LE11 5RH, UK
Received 6 February 2007; revised 28 February 2007; accepted 28 February 2007
Available online 3 March 2007
Abstract—As part of a Lead Optimisation programme to identify small molecule antagonists of the human CXCR2 receptor, a
series of substituted thiazolo[4,5-d]pyrimidines was prepared via the application of a novel tandem displacement reaction.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Chemotatic cytokines (Chemokines) are a class of po-
tent inflammatory mediators. Interaction of chemokines
with specific cell populations is mediated by G-protein
coupled receptors. These receptors are divided into four
subtypes (CC, C, CX3C and CXC), depending upon the
position of the N-terminal cysteine residues within the
protein. Interleukin 8 (IL-8, CXCL8) and growth-re-
lated oncogene a (GROa) are members of the CXC che-
mokine subfamily and play a role in the activation and
recruitment of neutrophils to sites of inflammation med-
iated through the CXCR2 receptor.1,2 Elevated levels of
CXCL8 have been observed in diseases such as arthri-
tis,3 chronic obstructive pulmonary disease (COPD),
asthma,4 ulcerative colitis and psoriasis.5 In the light
of this finding small molecule antagonists of the CXCR2
receptor are attractive biological targets.
by compound 1, Table 1, which formed the basis of a
new Lead Optimisation (LO) project. This paper will
disclose a further programme of synthetic work towards
primarily increasing potency and highlighting further
structure activity relationship (SAR) within this series.
Simple variation of the 7-substituent of 1 used the same
methodology applied during the Lead Identification (LI)
Table 1. Lead profile of thiazolopyrimidine 1
OH
HN
7
S
N
H2N
5
S
2
N
N
F
F
As part of a Hit-to-Lead study, our laboratory has pre-
viously disclosed the identification of a series of novel
thiazolopyrimidines as potent CXCR2 receptor antago-
nists.6 Early evaluation had indicated that S-benzyl sub-
stitution at the 5-position and hydroxy-alkylamino
substitution at the 7-position are preferred, exemplified
Assay
Thiazolopyrimidine 1
Binding CXCR2 pIC50
Functional pIC50
7.6
7.8a
98.4
2.5b
25
Human plasma protein binding %
Solubility (pH 7.4) lg/ml
Rat iv Cl ml/min/Kg
Rat iv Vss 0.5 L/Kg
Rat iv T1/2
1.9
1.2
h
Keywords: CXCR2; Thiazolopyrimidine.
*
Hu CaCo2 cm/s · 106
Rat po bioavailability %
3.6
Corresponding author. E-mail: FRASER.HUNT@ASTRAZENECA.
COM
15
a Functional antagonism was shown by blockade of GROa stimulated
intracellular calcium mobilisation in isolated human neutrophils
using a fluorescence imaging plate reader (FLIPR).7
Present address: UCB Granta Park, Great Abington, Cambridge
CB21 6GS, UK.
à
Present address: Organon Research Scotland, Newhouses, Lanark-
shire ML1 5SH, UK.
b Solubility corrected from value given in reference 1, Table 5.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.02.080

2732
F. Hunt et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2731–2734
R3
R4
N
R
R
Cl
N
N
i
S
N
S
N
i
S
N
N
N
N
H2N
H2N
S
N
H2N
N
S
H2N
N
SH
S
N
S
21-22
2
F
3-9
F
F
OH
OH
HN
HN
F
19
20
R
ii
R=
ii
iii
Cl
N
iii
Cl
N
R1
S
N
S
N
N
N
R2
Br
N
S
S
R
N
S
10
F
N
11
S
N
F
H2N
N
iv
F
H2N
N
F
N
S
R'
S
O
O
R3
R4
N
iv
23-24
25-34
N
N
R1
S
N
Scheme 2. Reagents and conditions: (i) Na, NH3,NH4Cl (70%); (ii)
R0Br, NaH (45–85%); (iii) OxoneTM (96%); (iv) R0SH, NaBH4, DMSO,
60 ꢁC (20–65%).
N
R2
S
12-18
F
F
21–22 with sodium hydride, followed by reaction with
a range of activated halides, gave an array of sulfides
25–34. The second approach involved reaction of the
benzylsulfones 23–24 prepared by oxidation of 19–20
with Oxoneꢂ, with a range of thiols. Due to the propen-
sity of thiols to form disulfides, the optimised conditions
involved treatment of sulfone with a thiol under an inert
atmosphere, using degassed solvent, and in the presence
of a stoichiometric quantity of sodium borohydride.
Scheme 1. Reagents and conditions: (i) R3R4NH 100 ꢁC; (ii) CHBr3,
iso-amylnitrite, CH3CN, 50 ꢁC (82%); (iii) R1R2NH (rt), Et3N; (iv)
R3R4NH 100 ꢁC.
phase, which involved treatment of the chloro com-
pound 2 with a range of primary and secondary amines
at elevated temperatures8 to give compounds 3–9,
Scheme 1. The development of a new tandem displace-
ment reaction permitted the independent exploration
of the SAR at both the 2- and 7-positions via a ‘‘one
pot’’ procedure. This route exploits the difference in
reactivity of positions 2 and 7 in compound 10 towards
nucleophiles.
Previous SAR studies around the 7-position had shown
that a hydroxylated alkylamino substituent is beneficial
for potency. In our follow up programme based on this
structural knowledge we treated compound 2 in a com-
binatorial manner with a designed set of amines in order
to explore the favoured steric and electronic parameters
of this motif, Scheme 1. From over 290 compounds syn-
thesised, only compounds that had a hydroxyethyl ami-
no core had binding affinity (pIC50) of >7.0 Table 2.
Further hydroxyl substitution of the alkyl side chain is
also tolerated as exemplified by the diols 3–5. Explora-
tion of substitution upon the hydroxyethyl amino group
showed that small alkyl groups are preferred, compare,
for example, the (R)-alaninol derivative 6 with 8, giving
the alaninol series a 6-fold advantage in potency over 1.
Moreover comparison of 6 with 7 shows a 100-fold pref-
erence for (R) enantiomer over the (S). Introduction of a
secondary cyclic amine at the 7-position was less well
tolerated as shown by 9, a cyclic analogue of 6.
Diazotisation of 2 with iso-amyl nitrite in the presence
of bromoform at 50 ꢁC9 afforded 10. This intermediate
was treated with an array of amines at room tempera-
ture affording the mono-substituted intermediates 11
efficiently. After an hour, a second set of amines was
added and the temperature increased to 100 ꢁC to give
the disubstituted compounds 12–18. Purification of the
final compounds was performed using automated high
performance liquid chromatography. The isolated prod-
ucts were quantified using Chemi-luminescent Nitrogen
Detection (CLND)10 and the purified samples were sol-
ubilised to a pre-determined concentration for in vitro
biological screening (affinity data and in vitro DMPK
data obtained). The biological activity of key com-
pounds was confirmed using re-synthesised solid, fully
characterised samples.
Turning our attention to varying both the 2- and 7-sub-
stituents using a 2-dimensional matrix via the tandem
displacement protocol, the results further reinforced
the importance of substitution in the 7-position by
hydroxyalkyl amines. Investigation of the substituent
at the 2-position showed that amino was most preferred,
small secondary alkyl amines, for example, methyl 16,
isopropyl 14–15 and cyclopropyl 12–13, led to a 2- to
5-fold drop in potency and small tertiary amines 17–18
a 100-fold drop when compared to 6. The (R)-alaninol
substituent was found to confer most potency regardless
Two synthetic approaches were employed to explore the
SAR at the 5-position. Both routes started with the ben-
zylthio ethers 19–20, Scheme 2. This strategy allowed
the opportunity to increase the diversity of substitution
of the products 25–34 by not limiting the synthesis to a
single reagent class.
Hydrogenolysis of 19–20 with sodium in liquid ammo-
nia11 gave the thiols 21–22. Subsequent treatment of

F. Hunt et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2731–2734
2733
Table 2. Substitution at the 2- and 7-positions
ortho and/or meta positions, of such groups lipophilic
and electron withdrawing substituents were the most po-
tent Table 3, fluorine and chlorine being most preferred.
Y
S
N
X
N
N
S
There was an increase in potency observed for 2,3-disub-
stitution with fluorine over both ortho and meta substi-
tution, compare 6–25 and 26, however this did not
transfer to the 2,3-dichloro analogue 30. Combining
the substitution of an ortho fluoro with a meta chloro
29 gave a compound equipotent with 6. Replacement
of the benzyl moiety with heteroaromatic rings is gener-
ally less well tolerated. Again best potency is observed
when the 7-position is substituted by the (R)-alaninol
motif as shown by 31–34.
F
F
Y
X
CXCR2 Bindinga pIC50
OH
OH
3
4
NH2
NH2
NH2
NH2
NH2
NH2
NH2
HN
7.8
7.1
7.5
8.4
7.2
7.6
<6.0
7.7
7.0
7.5
7.0
8.1
6.4
5.8
HN
OH
OH
HN
OH
In summary, our SAR investigation designed to opti-
mise all three substitution positions on compound 1
5
OH
HN
HN
Table 3. Substitution at the 5-position
OH
6
R
S
N
H2N
R'
N
N
S
OH
7
HN
R
R0
CXCR2 Bindinga pIC50
OH
OH
8
OH
HN
N
HN
1
6
7.6
F
F
F
F
9
OH
OH
OH
OH
OH
OH
8.4
7.9
7.8
7.5
7.9
8.4
7.8
7.0
7.5
6
HN
HN
HN
HN
HN
HN
OH
12
13
14
15
16
17
18
HN
25
26
27
28
29
30
31
32
33
34
F
OH
OH
HN
HN
HN
F
OH
HN
Cl
OH
OH
HN
HN
HN
Cl
H
N
F
Cl
OH
HN
OH
OH
N
N
HN
HN
Cl
Cl
OH
OH
HN
O
a Binding affinity was determined via a scintillation proximity assay
(SPA) using [125I]IL-8 and human recombinant CXCR2 (hrCXCR2)
receptor expressed in HEK 293 membranes.7
OH
OH
HN
HN
O
N
of the 2-substituent. Variation to the 5-position was
undertaken by one of two synthetic routes, reaction of
the thiols 21–22 with a range of benzyl and heteroaro-
matic halides, or through the treatment of the sulfones
23–24 with thiolates Scheme 2. Screening results gave
good correlation between the measured binding affinity
of the compound with substituent size and substitution
pattern, with a preference for small substituents at the
S
N
OH
6.9
HN
S
a Binding affinity was determined via a scintillation proximity assay
(SPA) using [125]IL-8 and human recombinant CXCR2 (hrCXCR2)
receptor expressed in HEK 293 membranes.7

2734
F. Hunt et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2731–2734
Table 4. Profile of thiazolopyrimidine 6
ments for the thiazolopyrimidine heterocycle with a view
to addressing the factors influencing the low bioavail-
ability will form the basis of a new programme of work.
OH
HN
N
S
N
N
H2N
S
References and notes
F
F
1. Holmes, W. E.; Lee, J.; Kuang, W. J.; Rice, G. C.; Wood,
W. I. Science 1991, 253, 1278.
Assay
Thiazolopyrimidine 6
2. Murphy, P. M.; Tiffany, H. L. Science 1991, 253, 1280.
3. Matsukawa, A.; Yoshimura, T.; Fujiwara, K.; Maeda, T.;
Ohkawara, S.; Yoshinaga, M. Lab. Invest. J. Tech. Meth.
Pathol. 1999, 79, 591.
4. Hay, D. W. P.; Sarau, H. M. Curr. Opin. Pharmacol. 2001,
1, 242.
5. Bizzarri, C.; Allegretti, M.; Di Bitondo, R.; Cervellera,
MN.; Colotta, F.; Bertini, R. Curr. Med. Chem. Anti-
Inflammatory Anti Allergy Agents 2003, 2, 67.
6. Baxter, A.; Cooper, A.; Kinchin, E.; Moakes, K.; Unitt, J.;
Wallace, A. Bioorg. Med. Chem. Lett. 2006, 16, 960.
7. For full details of the biological assays used, see: Austen,
R.; Baxter, A.; Bonnert, R.; Hunt, F.; Kinchin, E.; Willis,
P. Int. Patent Appl WO 00/09511.
Binding CXCR2 pIC50
Functional pA2
8.4
9.0
Human plasma protein binding %
Solubility (pH 7.4) lg/ml
Rat iv Cl ml/min/Kg
97.4
1.5
17
Rat iv Vss L/Kg
1.03
1.2
Rat iv T1/2
h
Hu CaCo2 mean A to B cm/s · 106
Hu CaCo2 mean B to A cm/s · 106
Mean efflux ratio
2.47
13.68
5.5
Rat po bioavailability %
9
8. Synthesis of compound 6: A mixture of 2 (0.1 g) and (2R)-2-
amino-propan-1-ol (0.5 g) in a solution in tetrahydrofuran
(10 ml) was heated in a sealed vessel at 100 ꢁC for 18 h. The
mixture was evaporated to dryness and purified (HPLC,
Novapakꢂ C18, 0.1% aqueous ammonium acetate:aceto-
nitrile, gradient elution 70:30 to 0–100 over 15 min) to
afford 6 0.051 g (46%) as a colourless solid. MS (APCI) 384
led to the identification of 6, a more potent CXCR2
receptor antagonist. This was achieved in part through
the development of a novel tandem displacement reac-
tion allowing the opportunity to rapidly explore two
points of diversity on the thiazolopyrimidine hetero-
cycle. A key finding from this programme of work high-
lighted that the (R)-alaninol substituent is the best
7-substituent regardless of the variations at the 2- and
5- positions. The in vitro metabolic stability and
in vivo pharmacokinetics were determined for 6 and
the results are shown in Table 4. Compound 6 is more
stable in vivo and less plasma protein bound than 1,
however the rat oral bioavailability of 6 is lower, which
could be attributed to a poorer absorption due to the
decrease in its measured aqueous solubility, and/or the
possibility of being a substrate for the efflux transporter
P-glycoprotein (pGp). CaCo2 permeability measure-
ments showed a modest but significant efflux component
with a mean efflux ratio of ꢀ6, which was not com-
pletely reduced in the presence of verapamil a known
competitive inhibitor of pGp. Investigation of replace-
1
(M+H, 100%); H NMR (399.98 MHz, DMSO) d 8.03 (s,
2 H), 7.40–7.26 (m, 2 H), 7.17–7.10 (m, 1 H), 7.01 (d,
J = 7.9 Hz, 1H), 4.70 (s, 1H), 4.39 (dd, J = 18.5, 14.4 Hz,
2H), 4.19 (quintet, J = 6.5 Hz, 1H), 3.44 (q, J = 5.0 Hz, 1H),
1.10 (d, J = 6.7 Hz, 3H); 13C NMR (100.585 MHz, DMSO)
d 170.88 (s), 168.70 (s), 164.85 (s), 155.19 (s), 149.61 (dd
JCF = 245, 12 Hz), 148.04 (dd, JCF = 246, 12 Hz, 128.63 (d,
JCF = 11.5 Hz), 126.06 (t, JCF = 3 Hz), 124.26 (dd,
JCF = 5,7 Hz),115.97 (d, JCF = 17 Hz), 98.09 (s), 64.27 (s),
47.97 (s), 27.01 (s), 17.26 (s).
9. Nair, V.; Richardson, G. S. Synthesis 1982, 8, 670.
10. Peake, D. A.; Duckworth, D. C.; Perun, T. J.; Scott, W.
L.; Kulanthaivel, P. Combinatorial Chemistry and High
Throughput Screening 2005, 8, 477.
11. Chandrakanta, G.; Schmidt, D. G.; Bimal, C. P. J. Org.
Chem. 1984, 49, 5256.