The discovery of potent, orally bioavailable pyrazolo and triazolopyrimidine CXCR2 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2013.11.074

A hit-to-lead optimisation programme was carried out on the Novartis archive screening hit, pyrazolopyrimidine 2-methyl-5-((phenylthio)methyl) pyrazolo[1,5-a]pyrimidin-7-ol 1, resulting in the discovery of CXCR2 receptor antagonist 2-benzyl-5-(((2,3-difluorophenyl)thio)methyl)-[1,2,4]triazolo[1,5-a] pyrimidin-7-ol 14. The SAR was investigated by systematic variation of the pendant thiol, alkyl and pyrimidinol groups. Replacement of the pyrazolopyrimidine core with a triazolo alternative led to a dual series of antagonists with favourable biological and pharmacokinetic properties.

Full text of DOI:10.1016/j.bmcl.2013.11.074

Bioorganic & Medicinal Chemistry Letters 24 (2014) 72–76
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of potent, orally bioavailable pyrazolo
and triazolopyrimidine CXCR2 receptor antagonists
a
a
b
a
a
David W. Porter a, , Michelle Bradley , Zarin Brown , Riccardo Canova , Steven Charlton , Brian Cox ,
Peter Hunt a, David Kolarik b, Sarah Lewis a, Des O’Connor a, John Reilly a, Carsten Spanka b,
Lauren Tedaldi a, Simon J. Watson a, Roland Wermuth b, Neil J. Press a
⇑
a Novartis Institutes for BioMedical Research, Horsham Research Centre, Wimblehurst Road, Horsham, West Sussex RH12 5AB, United Kingdom
b Novartis Institutes for BioMedical Research, Forum 1, Novartis Campus, CH-4056 Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
A hit-to-lead optimisation programme was carried out on the Novartis archive screening hit, pyrazolo-
pyrimidine 2-methyl-5-((phenylthio)methyl)pyrazolo[1,5-a]pyrimidin-7-ol 1, resulting in the discovery
of CXCR2 receptor antagonist 2-benzyl-5-(((2,3-difluorophenyl)thio)methyl)-[1,2,4]triazolo[1,5-a]pyrim-
idin-7-ol 14. The SAR was investigated by systematic variation of the pendant thiol, alkyl and pyrimidinol
groups. Replacement of the pyrazolopyrimidine core with a triazolo alternative led to a dual series of
antagonists with favourable biological and pharmacokinetic properties.
Received 30 August 2013
Revised 26 November 2013
Accepted 29 November 2013
Available online 4 December 2013
Keywords:
CXCR2
Ó 2013 Elsevier Ltd. All rights reserved.
Chemokines
Pyrazolopyrimidine
Triazolopyrimidine
Neutrophils
COPD
Chemokines are small 8–10 kDa cytokines characterised by four
conserved cysteine residues linked via disulfide bonds.1 According
to the structure and spacing of these cysteine residues they can be
subdivided into four main groups, namely the CXC, CC, C and CX3C
families. Chemokines can be released from a number of inflamma-
tory and structural cell types following stimulation and play an
important role in several inflammatory lung diseases including
chronic obstructive pulmonary disease (COPD),2 severe asthma3
and cystic fibrosis.4
The actions of chemokines are mediated by seven transmem-
brane G protein-coupled receptors (GPCRs) and amongst these is
the receptor designated CXCR2. Neutrophilia is a key component
of the inflammatory lung diseases mentioned above and neutro-
philic inflammation appears to be largely regulated through
CXCR2.5,6 A number of CXC chemokines bind to CXCR2, including
disclosed antagonists and a number of these have progressed
through pre-clinical studies (including in vitro and in vivo) and
are undergoing clinical trials. Noteworthy are navarixin 1,8 dani-
rixin 29 and AZD-5069 (structure not disclosed10) (Fig. 1). Nava-
rixin 1 has been tested in both healthy and moderate to severe
COPD subjects. In healthy volunteers following experimental
ozone challenge, navarixin significantly decreased ozone-induced
airway neutrophilia. Additionally, significant decreases in IL-8
and myeloperoxidase (MPO) levels were noted.11 In patients with
moderate to severe COPD, navarixin was shown to be safe and sig-
nificantly reduced sputum neutrophils and matrix metallopepti-
dase 9 (MMP-9). An improvement in lung function as measured
by an increase in forced expiratory volume in one second (FEV1)
compared to placebo was also noted.12
A HTS was undertaken to identify functional antagonists that
CXCL1 (GRO-
a
), CXCL2, CXCL3, CXCL5, CXCL7 and CXCL8 (IL-8).
blocked the binding of GRO-
(hrCXCR2) expressed in CHO membranes using a [35S]-GTP
say. In this assay the amount of accumulated [35S]-GTP
S is di-
rectly proportional to the degree of receptor activation. At the hit
evaluation stage, compounds with inhibition >50% at10 M were
re-confirmed as inhibitors in a 10 point IC50 S mem-
35S]-GTP
brane assay and a more conventional filter wash hrCXCR2 [125I]-
GRO-
or [125I]-IL-8 binding assay.13 A good correlation between
a
to human recombinant CXCR2
Given the obvious unmet medical need, alternative approaches
are needed for the treatment of neutrophilic inflammatory diseases
and antagonists targeting this receptor may provide a therapeutic
option.7
To date, there are no CXCR2 receptor antagonists approved for
use in humans. However, several pharmaceutical companies have
cS as-
c
l
[
c
a
the binding and functional assays was noted and binding affinity
was not routinely measured during hit-to-lead. On-target activity
⇑
Corresponding author.
E-mail address: david.porter@novartis.com (D.W. Porter).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.11.074

D. W. Porter et al. / Bioorg. Med. Chem. Lett. 24 (2014) 72–76
73
Me
N
Me
OH
OH
Cl
O
N
O
O
HN
Me
O
HN
Me
N
F
O
O
S
N
H
N
H
F
S
F
O
O
Me
N
H
OH
Me
N
N
H
S
O
O
OH
Navarixin 1
Danirixin 2
AZD-5069 3
Proposed structure
Figure 1. Publicly identified CXCR2 antagonists navarixin and danirixin; compound 3 is believed to be AZD-5069.10
Table 1
Hit profile of pyrazolopyrimidine 1
Table 3
Me
CXCR2 antagonist binding potencies
N
N
N
S
N
HO
N
N
N
S
Measure
Pyrazolopyrimidine 1
2.0
HO
R
hrCXCR2 [125I]-GRO-
35S]-GTP
S IC50
Molecular weight
a
IC50
lM
[
c
10.3
lM
Compound
R
CXCR2 [35S]-GTP M)a
cS IC50 (l
271
HT-eq Solubility pH 6.8
cLogP
0.04 g/l
3.2
2.5
2
3
0.82
>30
Binding LLE (pIC50 À cLogPa)
a
Pr
N
Compound 1 is weakly acidic and for the purposes of this calculation one can
assume LogP = LogD.
4
>30
N
Me
Table 2
Two atom c-5 linkers
O
5
4.1
Me
2
Me
Me
N
N
N
Alternative c-5 linkers
Ph
6
7
11
13
N
7
N
N
N
S
5
S
A
N
HO
Ar
HO
A
Ar
Ar = (un)substituted non-heterocyclic phenyl
NC
OMe
Me
A–A linker group
Effect on potency
Comments
Metabolic instability?
Decreases cLogP by 1.5
Decreases cLogP by 2.0
Increases cLogP by 0.25
Decreases cLogP by 0.5
8
9
11
C–S
—
C–S(O)
C–S(O)2
C–C
30-fold loss
50-fold loss
0?5-fold loss
4?15-fold loss
OMe
C–O
7.8
Cl
Me
was confirmed by showing a lack of functional antagonism in a
SDF-1 stimulated hrCXCR4 [35S]-GTP
S selectivity assay and peri-
OCF3
c
10
11
6.4
4.1
odic spot checking revealed selectivity not to be an issue.
Pyrazolopyrimidine 1 was discovered from the CXCR2 HTS dis-
playing a modest and comparable level of potency in both the
binding and functional assays. This activity translates into accept-
able ligand-lipophilicity efficiency (LLE)14 and the profile of this
CXCR2 hit is shown in Table 1.
The plan was to examine in sequence the three pyrazolopyrim-
idine substituents looking for SAR whilst aiming to improve po-
tency and solubility and lower cLogP. Alternative linkers to the
c-5 pendant phenyl ring, pyrazolo versus triazolo head groups
and isosteres for the pyrimidinol functionality were explored.
Me
Cl
12
4.8
Cl
(continued on next page)

74
D. W. Porter et al. / Bioorg. Med. Chem. Lett. 24 (2014) 72–76
were preferred and bis-halogenation gave rise to an increase in
Table 3 (continued)
potency. 2,4- and 2,3-dichloro 12–13 were comparable in activity
and swapping of the chlorine substituents in 13 for the
fluorine substituents in 14 gave an increase in activity.
2,3-Difluoro 14 also has reduced lipophilicity relative to 13. The
3,4-difluoro analogue 15 was found to be less active and prefer-
ence for 2,3-difluoro has been reported in a thiazolopyrimidine
series of antagonists.16
The synthetic route to this series of triazolopyrimidines is
shown in Scheme 1. Condensation of 3-benzyl-1H-1,2,4-triazol-5-
amine17 with methyl 4-chloro-3-oxobutanoate gave 5-(chloro-
methyl)triazolopyrimidine 16. DBU mediated SN2 displacement
of chlorine in MeCN gave the desired phenyl and alkylthio triazol-
opyrimidines 2–15. For the 2,3-difluorophenyl analogue 14, the
precursor thiol was synthesised from 2,3-difluorophenol as shown
in Scheme 2. Treatment of 2,3-difluorophenol with N,N-dim-
ethylthiocarbamoyl chloride in DMF in the presence of DABCO
gave O-thiocarbamate intermediate 17. Miyazaki–Newman–Kwart
type rearrangement18 under high temperature in Dowtherm A pro-
ceeded smoothly to give S-aryl thiocarbamate 18. Hydrolytic
deprotection with NaOMe in MeOH yielded 2,3-difluorobenze-
nethiol in 45% yield over the three steps.
Having established the optimal nature of the 2,3-(difluoro-
phenyl)thiomethyl moiety, variation of the pyrazolo/triazolo
head group was investigated. Routes to these compounds are
shown in Scheme 3 and utilised the reaction of methyl 4-((2,3-
difluorophenyl)thio)-3-oxobutanoate, prepared from methyl
4-chloro-3-oxobutanoate, with a series of commercial and inter-
nal sources of triazoles and pyrazoles—a selection of these is
shown in Table 4. In the triazole series (19–23), the R1 position
was tolerant of lipophilicity and increasing chain length. Efficacy
was comparable between methyl 19, isopropyl 20, cyclohexyl 21,
phenyl 22 and phenethyl 23. A larger set of compounds were
made in the pyrazole series exploring both mono- and fused
substitution (24–31). Heteroaromatics at the R1 position were
reasonably well tolerated (analogues 24, 25 and 26) and the
tricyclic phenyl analogue 28 was found to be more potent than
saturated cyclohexyl 27. Small groups at the R2 position
(nitrile 30 and ethyl 31) were weakly active in comparison to
Compound
R
CXCR2 [35S]-GTP
1.7
cS IC50 (l
M)a
Cl
F
Cl
F
13
14
15
0.35
1.7
F
F
a
Mean of n P 2.
The lipophilic S-phenyl was seen as a potential site of metabolic
instability and a set of alternative linkers to substituted and unsub-
stituted phenyl rings were assessed early on (Table 2). Methylene
O-phenyl and sulfoxide/sulfone analogues had improved cLogP
values but were generally less potent antagonists than the corre-
sponding S-phenyl. A series of ethyl linked phenyl compounds dis-
played comparable potency and optimisation of the phenyl ring
was undertaken with C–S and C–C linkers. Data from the C–S series
will be discussed in this letter.
Replacement of the 2-methyl pyrazolo head group with a 2-
benzyl triazolo gave a slightly more potent analogue 2 and this
was used as the starting point for variation of the thioether
(Table 3). CXCR2 affinity was lost changing the S-phenyl substitu-
ent to an S-alkyl as exemplified by acyclic pentyl 3. In general, it
was found that changing the S-phenyl to an S-heteroaromatic led
to a loss in activity. Of the five ring heteroaromatics synthesised,
imidazole
4 was inactive, whilst 2,5-dimethylfuran 5 and
4-phenylthiazole 6 had some activity. A decrease in activity was
observed for six ring heteroaromatics including 3-substituted nic-
otinonitrile 7 relative to 2. Almost all S-phenyl groups had some
antagonistic activity and the nature and position of substituents
was important. Mono-, bis- and tri-substituted aromatics were
tolerated but SAR was flat in the range 1–10
lM (8–11). As noted
unsubstituted 29 and larger substituents tended to cause
a
previously elsewhere,15 analogues with a 2-substituted phenyl
complete loss of activity.
N
N
O
O
i
ii
N
N
Cl
OMe
N
N
N
N
53%
(20-90)%
Cl
S
HO
HO
R
16
2-15
Scheme 1. Reagents and conditions: (i) 3-benzyl-1H-1,2,4-triazol-5-amine, AcOH, 100 °C, 16 h; (ii) thiol, DBU, MeCN.
F
F
F
HO
F
(Me)2N
O
F
(Me)2N
S
F
i
ii
S
O
87%
67%
17
18
F
HS
F
iii
78%
Scheme 2. Reagents and conditions: (i) ClCSN(Me)2, DABCO, DMF; (ii) Dowtherm A, K2CO3, 250 °C, 3 h; (iii) NaOMe, MeOH, 100 °C, 3 h.

D. W. Porter et al. / Bioorg. Med. Chem. Lett. 24 (2014) 72–76
75
R1
N
R2
N
X
F
F
O
O
i
ii
HS
F
F
S
N
F
OMe
S
F
79%
(13-80)%
HO
19-31
Scheme 3. Reagents and conditions: (i) ClCH2COCH2CO2Me, NaOMe, MeOH; (ii) pyrazole or triazole, AcOH, 100 °C, 16 h.
Table 4
CXCR2 antagonist binding potencies
Table 5
Lead profile of triazolopyrimidine 14
R1
R2
X
N
N
N
N
F
N
N
S
F
N
F
HO
F
S
HO
Compound
X
R1
R2
CXCR2 [35S]-GTP
1.5
cS IC50 (l
M)a
Measure
Triazolopyrimidine 14
Me
19
N
—
—
hrCXCR2 [125I]-GRO-
35S]-GTP
S IC50
Rat microsome Cl (ml/min/kg)
a
IC50
0.33
0.35
31
l
l
M
M
[
c
20
21
N
N
0.91
0.51
Human microsome Cl (ml/min/kg)
Rat iv Cl (ml/min/kg)
Rat iv Vss (l/kg)
16
1.4
0.9
9
—
—
Rat iv t (h)
½
Rat po bioavailability (%)
Rat plasma protein binding (%)
Molecular weight
52
22
23
N
N
0.85
0.47
>99.5
384
0.09
4.1
HT-eq Solubility pH 6.8 (g/l)
cLogP
Binding LLE (pIC50 À cLogPa)
2.4
—
a
Compound 14 is weakly acidic and for the purposes of this calculation one can
assume LogP = LogD.
O
N
24
25
C
C
H
H
0.15
0.28
at the 4-position of the pyrimidine ring. We were concerned that
the compounds in our chemical series (C–S and C–C) would have
high in vivo clearance following conjugation of the phenol with
glucuronic acid, and in order to address this, a limited investigation
into variation of the phenol at the 7-position was therefore carried
out. Representative examples of this are shown in Figure 2. Meth-
ylation of the phenol to give 33 led to a loss in activity compared to
the parent phenol 32—consistent with the findings of others, that
acidity in this region is important. However, methanesulfonation
of 14 to give 34 also yielded a less potent compound. The methan-
sulfonamide group should be partially ionised at physiological pH
(pKa 5.0) and this makes it difficult to draw direct parallels be-
tween acidic isosteres in this and other published series.
O
26
C
H
0.87
Me
27
28
C
C
1.7
0.26
29
30
C
C
H
H
H
0.81
4.1
CN
Me
The profile of lead triazole compound 1421 is shown in Table 5.
This compound has a reasonable lead like profile with acceptable
potency in the binding and functional assays. The in vitro rat and
human microsomal clearance is low to moderate and the com-
pound is a weak acid (pKa 6.5) with moderate to high lipophilicity.
Plasma protein binding as a consequence is high and the com-
pound has a suitable rat in vivo profile with low clearance and
good oral bioavailability. The PK profile suggests that clearance
via glucuronidation is not an issue. Metabolite identification stud-
ies indicate the formation of sulfone and sulfoxide metabolites
in vitro and in vivo but further studies are required to determine
the levels of each. LLE is unchanged relative to the HTS hit 1 but
SAR within the series demonstrated room for manoeuvre, in partic-
ular via variation of substituents on the pyrazole/triazole head
31
C
H
20
a
Mean of n P 2.
Although many chemotypes have emerged since the first disclo-
sure of phenol-containing diarylurea antagonists,19 this compound
class in various isosteric guises has proven to be one of the most
successful to date. The ionisation state of the phenol has been
shown to be critical for binding affinity. A pH-range binding study
with a compound from the danirixin series demonstrated that the
ionised form was >10 fold more potent than the non-ionised
form20 and removal of the phenol led to a complete loss of activity.
In addition, several thiazolopyrimidine based AZD-5069 analogues
have been disclosed containing a hydroxyl or reverse sulfonamide

76
D. W. Porter et al. / Bioorg. Med. Chem. Lett. 24 (2014) 72–76
HO
N
HO
N
N
N
F
N
7
N
F
Methylation
Me
F
F
O
HO
33
IC50 = 10.5 M
32
µ
µ
IC50 = 1.1 M
N
Methanesulfonation
N
N
N
F
N
N
F
O O
S
S
F
HO
S
F
Me
N
H
14
34
µ
µ
IC50 = 0.35 M
IC50 = 13.9 M
Figure 2. Alternative c-7 groups.
12. Magnussen, H.; Holz, O.; Watz, H.; Sauer, M.; Khanskaya, I.; Gann, L.; Stryszak,
P.; Sadeh, J. Presented at the European Respiratory Society, Brussels, September
2010; Session 65.
13. For full details of the biological assays used see: Porter, D.; Press, N. J.; Spanka,
C. U.S. Pat. Appl. US 20100069407, 2010—(i) receptor binding assay: [125I]-IL-8
hrCXCR2 was performed in a 96-well micro plate format with each reaction
mixture containing 0.05 mg/ml CXCR2 membrane protein. Compounds of
interest were pre-dissolved in DMSO so as to reach a final concentration of
group. Compound 14 was inactive (IC50 >10 lM) against a broader
panel of 49 GPCRs and triazoles and pyrazoles as illustrated by 14
form the basis for on-going studies with this series of compounds.
In conclusion, HTS identified novel pyrazolopyrimidine chemo-
types as selective CXCR2 antagonists. The SAR of the scaffold was
explored, resulting in identification of triazolopyrimidine 14 which
is a functional antagonist of the human CXCR2 receptor and shows
good oral bioavailability in the rat.
between 10
was initiated by addition of 0.02 nM [125I]-IL-8; (ii) [35S]-GTP
for hrCXCR2 receptor using SPA technology: assay was performed in duplicate
in 96 well OptiplateTM microplates in a final volume of 250 l per well.
lM and 0.0005
l
M [final concentration of DMSO 2% (v/v)]. Binding
c
S binding assay
l
References and notes
Compounds were diluted in DMSO (0.5% final concentration). Data were
expressed as the% response to 100 nM IL-8 minus basal.
14. Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881.
15. Baxter, A.; Bennion, C.; Bent, J.; Boden, K.; Brough, S.; Cooper, A.; Kinchin, E.;
Kindon, N.; McInally, T.; Mortimore, M.; Roberts, B.; Unitt, J. Bioorg. Med. Chem.
Lett. 2003, 13, 2625.
16. Baxter, A.; Cooper, A.; Kinchin, E.; Moakes, K.; Unitt, J.; Wallace, A. Bioorg. Med.
Chem. Lett. 2006, 16, 960.
17. Chembridge Corporation, 11199 Sorrento Valley Road, San Diego, CA 92121,
USA.
18. Zonta, C.; De Lucchi, O.; Volpicelli, R.; Cotarca, L. Thione–Thiol Rearrangement
In Miyazaki–Newman–Kwart Rearrangement and Others. Topics in Current
Chemistry; Schaumann, E., Ed.; Springer: Berlin/Heidelberg, 2007; Vol. 275,
pp 131–161.
19. Widdowson, K. L. Int. Pat. Appl. WO 1997/49286, 1997.
20. 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. Pept. Sci. 1998, 5, 235.
21. Selected data for 14: dH (500 MHz, DMSO-d6) 13.41 (1H, br), 7.45–7.15 (8H, m),
5.79 (1H, s), 4.13 (2H, s), 4.07 (2H, s); C19H14F2N4OS requires C, 59.37; H, 3.67;
N, 14.57; S, 8.34. Found: C, 58.98; H, 3.94; N, 14.86; S, 8.10.
1. Uguccioni, M.; D’Apuzzo, M.; Loetscher, M.; Dewald, B.; Baggioloini, M. Eur. J.
Immunol. 1995, 25, 64.
2. Pauwels, R. A.; Rabe, K. F. Lancet 2004, 364, 613.
3. Bel, E. H.; Sousa, A.; Fleming, L.; Bush, A.; Chung, K. F.; Versnel, J.; Wagener, A.
H.; Wagers, S. S.; Sterk, P. J.; Compton, C. H. Thorax 2011, 66, 910.
4. Anderson, P. Ther. Adv. Respir. Dis. 2010, 4, 177.
5. Murphy, P. M. Semin. Hematol. 1997, 34, 311.
6. Chapman, R. W.; Phillips, J. E.; Hipkin, R. W.; Curran, A. K.; Lundell, D.; Fine, J. S.
Pharmacol. Ther. 2009, 121, 55.
7. Barnes, P. J. J. Allergy Clin. Immunol. 2007, 119, 1055.
8. Nair, P.; Gaga, M.; Zervas, E.; Alagha, K.; Hargreave, F. E.; O’Byrne, P. M.;
Stryszak, P.; Gann, L.; Sadeh, J.; Chanez, P. Clin. Exp. Allergy 2012, 42, 1097.
9. Busch-Petersen, J. U.S. Pat. Appl. US 20070249672, 2007.
10. (a) Gullberg, B. A. I.; Larsson, P. T.; Stonehouse, J. P. Int. Pat. Appl.
WO2012007748, 2012.; (b) Norman, P. Expert Opin. Ther. Patents; 2013, 23,
113.
11. Holz, O.; Khalilieh, S.; Ludwig-Segpiel, A.; Watz, H.; Stryszak, P.; Soni, P.; Tsal,
M.; Sadeh, J.; Magnussen, H. Eur. Respir. J. 2010, 35, 564.