Discovery of potent and selective phenylalanine derived CCR3 antagonists. Part 1

Article abstract of DOI:10.1016/S0960-894X(01)00248-7

The discovery of a series of phenylalanine derived CCR3 antagonists is reported. Parallel, solution-phase library synthesis has been utilized to delineate the structure-activity relationship leading to the synthesis of highly potent, CCR3-selective antagonists.

Full text of DOI:10.1016/S0960-894X(01)00248-7

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1441–1444
Discovery of Potent and Selective Phenylalanine
Derived CCR3 Antagonists. Part 1
Dashyant Dhanak,* Lisa T. Christmann, Michael G. Darcy, Anthony J. Jurewicz,
Richard M. Keenan, Judithann Lee, Henry M. Sarau, Katherine L. Widdowson
and John R. White
SmithKline Beecham Pharmaceuticals, 1250 South Collegeville Road, PO Box 5089, Collegeville, PA19426-0989, USA
Received 14 December 2000; accepted 6 April 2001
Abstract—The discovery of a series of phenylalanine derived CCR3 antagonists is reported. Parallel, solution-phase library synth-
esis has been utilized to delineate the structure–activity relationship leading to the synthesis of highly potent, CCR3-selective
antagonists. # 2001 Elsevier Science Ltd. All rights reserved.
The chemokine superfamily of secreted proteins exert
their biological activity by binding to and activating
seven-transmembrane domain G-protein coupled cell
surface receptors (GPCRs) present on a variety of cells
including immune and inflammatory cells.1 To date,
over 20 distinct chemokines have been characterized2
and have been shown to be involved in the highly spe-
cific trafficking of pro-inflammatory leukocytes impor-
tant in a number of disease states.3 Eosinophils are
proinflammatory granulocytes that are thought to play
a major role in allergic diseases, such as bronchial
asthma, allergic rhinitis, pruritis and atopic dermatitis.4
Upon activation, eosinophils release lipid mediators,
cytotoxic proteins, oxygen metabolites, and cytokines,
all of which have the potential to produce the clinically
observed pathophysiology. Recently, a CC chemokine,
eotaxin, has been shown to mediate eosinophil infiltra-
tion in vivo by the activation of a novel CC chemokine
receptor, designated CCR3, present on peripheral blood
eosinophils.5 Studies with CCR3 monoclonal antibodies
have demonstrated that receptor expression is primarily
restricted to eosinophils and a subset of Th2 T-cells and
that this restricted expression may be responsible for the
selective recruitment of eosinophils and Th2 T-cells in
allergic inflammation.6 Our interest in this area stems
from the potential therapeutic application of selective
CCR3 antagonists7 as antiinflammatory agents and we
report herein the discovery and initial structure–activity
relationships of a series of highly selective and potent
phenylalanine derived CCR3 antagonists. The cloning
and functional expression of the human CCR3
(hCCR3) has been reported by a number of labora-
tories8 and we have utilized a stable RBL-2H3 cell line
expressing the receptor to identify antagonists that are
able to block the human eotaxin-induced intracellular
calcium mobilization in these cells. High-throughput
screening of our proprietary compound collection using
a fluorescence imaging plate reader (FLIPR) to track
the intracellular calcium changes identified N-benzoyl-
3,5-diiodotyrosine ethyl ester 1 as a modestly effective
antagonist (IC50=2.3 mM). Consistent with its func-
tional activity, in a binding assay using [125I]human
eotaxin as the radioligand and purified human eosino-
phils as the CCR3 source, 1 was determined to have
reasonable CCR3 affinity (IC50=535 nM).9 Although
the presence of the metabolically labile ester precluded
in vivo evaluation, we felt that 1 had sufficient biologi-
cal activity to warrant further chemical optimization
with the aim of improving initially the CCR3 affinity
and ultimately also identifying an acceptable ester sur-
rogate. The structure of 1 suggested that the compound
was amenable to rapid synthesis of analogues, prompt-
ing us to investigate a solution-based parallel synthesis
approach towards the optimization of CCR3 affinity.
*Corresponding author. Fax: +1-610-917-4157; e-mail: dash_
dhanak-1@sbphrd.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00248-7

1442
D. Dhanak et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1441–1444
Chemistry10
compounds of Table 3. The substituted alkyl and benzyl
esters of Table 4 were synthesized using literature pro-
cedures.12 Racemic 24 (Table 3) was obtained in excel-
lent yield via a modified Dakin–West reaction13 and
finally the amides 25 and 26 were prepared using stan-
dard peptide coupling conditions.11
The compounds described in Tables 1 and 2 were either
commercially available or prepared by the benzoylation
of the corresponding amino acid esters.11 Reaction of
tyrosine ethyl ester hydrochloride with a variety of
electrophiles according to Scheme 1 gave access to the
Our initial efforts focused on investigating alternatives
to the diiodotyrosine side chain in 1 while maintaining
both the N-benzoyl group and ethyl ester (Table 1).
Removal of the two iodine atoms either alone (2) or in
combination with the phenolic hydroxyl (3) not only
retained but led to a small increase in CCR3 affinity.
Less conservative changes, for example, removal of the
aromatic system (4), were not tolerated. The CCR3
affinity of 3 could be further improved by substitution,
and of a number of groups surveyed, a 4-nitro group
was found to be associated with particularly high
receptor affinity. A high degree of enantiospecificity in
the CCR3/antagonist interaction was apparent since the
correct amino acid stereochemistry was critical for good
receptor affinity with only the natural l-stereochemistry
having significant activity. For example, the l-enantio-
mer 2 had >200-fold higher CCR3 affinity than its
corresponding d-isomer 6 (Table 1) and a similar trend
was observed for other analogues evaluated. Among
heterocyclic analogues of the phenylalanine ring, the
tryptophan derivative 7 was found to be approximately
equipotent to the carbocyclic parent.
Table 1.
Compound
R
n
CCR3
IC50 (nM)
1
2
3
4
5
6
7
8
9
3,5-I2-4-OH-Ph
4-OH-Ph
Ph
1
1
1
9000
1
1
1
0
2
535
383
190
H1
l-4-NO2-Ph
d-4-OH-Ph
l-3-Indolyl
(Æ)-Ph
49
>50,000
325
>20,000
5950
(Æ)-Ph
Table 2.
The effect of varying the length of the linker between the
a-carbon and the aryl side chain in 3 was equally dra-
matic and either deletion or extension of the methylene
chain to the phenylglycine 8 or to the homophenylalanine
9, respectively, was detrimental for CCR3 affinity.
Compound
R
CCR3
IC50 (nM)
2
COPh
CONHPh
383
>33,000
16,000
>25,000
6250
>25,000
>25,000
80
10
11
12
13
14
15
16
17
18
Having established that a substituted phenylalanine
was the simplest minimal requisite for good CCR3
affinity, we next evaluated potential alternatives to
the amide group in 2. In order to rapidly survey a
number of diverse combinations of linker alternatives
and aryl substituents, a solution-based parallel-synthesis
approach was adopted in which the active tyrosyl and
ethyl ester side chains of the lead were retained and the
SO2(4-Me)Ph
CO-Z-l-Asn
CO-Z-l-Phe
CO-Z-l-Ser
CO-Z-l-Gln
CO-1-napthyl
CO-2-pyridyl
COCH2Ph
1700
2600
Scheme 1. Reagents: (A) 1 equiv RCO2H, 3 equiv NMM, 1.2 equiv HOBT, 1.2 equiv EDCI, DMF; (B) 1 equiv RSO2Cl, 3 equiv Na2CO3, CHCl3/
H2O; (C) 1 equiv RNCO, dioxane.

D. Dhanak et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1441–1444
1443
amino acid nitrogen derivatized. Reaction of tyrosine
ethyl ester under standard conditions with a range of
carboxylic acids, sulfonyl chlorides, and isocyanates
generated a variety of the corresponding amides, sulfo-
namides, and ureas (Scheme 1). Of these, the amide-
linked derivatives were of most interest since the level of
potency displayed was markedly superior to that of the
other linkers surveyed. The urea analogues were found
to be devoid of CCR3 affinity and were not pursued.
Several of the sulfonamides, however, retained some
receptor binding activity although the structure–activity
relationship appeared to be divergent to that of the
amides. Further details of these studies will be the sub-
ject of future publications. Consistent with the presence
of a largely lipophilic binding domain in the vicinity of
the amide moiety, lipophilic aryl amides were generally
associated with the highest CCR3 affinity and the
introduction of either polar or nonaryl moieties was
poorly tolerated, leading to virtually complete abolition
of activity. Extended aromatic systems were highly
favoured with the 1-naphthoyl derivative 16 being among
the most potent analogues identified. Interestingly, amino
acid derived amides were particularly poor receptor
ligands regardless of the nature of the amino acid side
chain, suggesting that the presence of the additional car-
boxylate functionality was disfavored (Table 2).
tion for the alkyl part of the ester moiety suggested that
this part of the molecule was not involved in making
critical receptor interactions and may well be exposed to
solvent when bound to the receptor. Consequently, we
anticipated that replacement of the ester with other
metabolically more stable groups to provide analogues
suitable for in vivo evaluation would not be problematic
providing that any important interactions of the ester
heteroatoms were retained. We selected 19 as the template
to base ester replacement structure–activity relationship on
because the potency of this compound was sufficiently high
that small losses in CCR3 affinity introduced by these
changes should not lead to complete loss of affinity. How-
ever, we quickly realized that the ester group in this series
of antagonists played a key role in CCR3 binding and even
conservative modifications to this group led to huge
decreases in receptor affinity. For example, either removal
of the ester carbonyl to give the corresponding ether or
replacement with a simple alkyl eliminated CCR3 affi-
nity (data not shown). Surprisingly, the ketone 24 was
devoid of all CCR3 affinity and conversion of the ester
to the corresponding secondary (25) or tertiary (26)
amide was similarly not tolerated (Table 3). This exqui-
site sensitivity of the ester moiety for CCR3 binding is
in contrast to that reported for tryptophan ester derived
antagonists of the NK1 receptor where the ketone deri-
vatives retained very high receptor affinity although
interestingly, even in that case, the amide and ether
analogues demonstrated much lower affinity.14
We found that combining the best substituents dis-
covered from the independent optimization of the phe-
nylalanine side chain and the aryl group resulted in the
identification of potent antagonists such as 20 and 27,
whose receptor affinity was comparable to that of the
chemokine agonist, eotaxin (Tables 3 and 4).
Although such reactions are not well precedented with
GPCRs, the apparent absolute requirement for an ester
moiety for CCR3 antagonism prompted us to investi-
gate whether the receptor/antagonist interaction
involved an irreversible transesterification of a critical
residue in the protein. Incubation of CCR3 (primary
human eosinophils) with 5 for 30 min at room tem-
perature followed by extensive washing completely
restored both the binding affinity and calcium mobili-
zation activity of eotaxin, suggesting that the compound
was a simple, freely reversible antagonist. Clearly, the
limited experiments carried out so far would not
preclude a rapid acylation/deacylation of the receptor.
As shown in Table 3, in contrast to the highly specific
requirements shown for the amide side chain, a greater
variety of esters was tolerated without significant loss of
receptor affinity. Linear alkyl derivatives such as methyl
i
20 and ethyl 19 analogues along with the branched Pr
analogue 21 were all found to retain high CCR3 affinity.
t
However, additional steric encumbrance as in the Bu
analogue 22 led to some loss of CCR3 potency, indi-
cating an overall size limitation. A change from an alkyl
to an aromatic ester was well tolerated with the benzyl
ester 23 being equipotent to 21. The lack of discrimina-
Table 4.
Table 3.
Compound
R1
R2
CCR3
IC50 (nM)
Assay
Agonist
IC50 (nM)
CCR3 binding
[
[
125I]-Eotaxin
125I]-MCP-4
Eotaxin
5
7
38
19
20
21
22
23
24
25
26
27
1-Naphthyl
1-Naphthyl
1-Naphthyl
1-Naphthyl
1-Naphthyl
CO2Et
CO2Me
8
5
21
57
18
Ca2+ mobilization
(RBL-2H3-CCR3 cells)
COi2Pr
COt2Bu
Eotaxin-2
MCP-4
Eotaxin
Eotaxin-2
MCP-4
C5a
35
20
32
25
CO2CH2Ph
(Æ)COPr
CONHCH2Ph
CONMe2
CO2Et
1-Naphthyl
1-Naphthyl
>33,000
4000
>33,000
3
Eosinophil chemotaxis
1-Naphthyl
2,4-Dimethylphenyl
55
>330

1444
D. Dhanak et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1441–1444
The functional activity of the CCR3 binding derivatives
was confirmed by assaying their ability to block the
eotaxin-mediated increases in intracellular calcium
mobilization either in HEK293 cells transfected with
human CCR3 or in primary human eosinophils.15 In an
alternative assay, the inhibition of eotaxin induced che-
motaxis of primary human eosinophils from allergic
donors was determined.16 In both assays, the proto-
typical inhibitor 20 was shown to be a potent antagonist
with IC50 values consistent with those determined in the
binding assay (Table 4). In addition, 20 also effectively
inhibited the binding of [125I]MCP-4 to human eosino-
phils, strongly suggesting that 20 mediated its effects by
binding to CCR3 and not the chemokine agonist.15
436. (d) Power, C. A.; Wells, T. N. C. Trends Pharmacol. Sci.
1996, 17, 209. (e) Schwarz, M. K.; Wells, T. N. C. Expert
Opin. Ther. Patents 1999, 9, 1471.
2. Cascieri, M. A.; Springer, M. S. Curr. Opin. Chem. Biol.
2000, 4, 420.
3. (a) Kita, H.; Gleich, G. J. J. Exp. Med. 1996, 183, 2421. (b)
Baggiolini, M. J. Clin. Invest. 1996, 97, 587.
4. (a) Schroder, J.-M. In Chemokines and Skin; Kownatski,
E., Norganer, J., Eds.; Birkhauser: Basel, 1999. (b) Gonzalo,
J.-A.; Lloyd, C. M.; Kremer, L.; Finger, E.; Martinez, A.-C.;
Siegelman, M. H.; Cybrilsky, M.; Gutierrez-Ramos, J.-C. J.
Clin. Invest. 1996, 98, 2332.
5. Teixeira, M. M.; Wells, T. N. C.; Lukacs, N. W.; Proud-
foot, A. E. I.; Kunkel, S. L. J. Clin. Invest. 1997, 100, 1657.
6. Heath, H.; Qin, S.; Rao, P.; Wu, L.; LaRosa, C.; Kassam,
M.; Ponath, P. D.; Mackay, C. R. J. Clin. Invest. 1997, 99,
178.
The chemokine receptors are known to bind and inter-
act with multiple agonists. CCR3 for example has been
shown to bind not only eotaxin but also the chemokines
eotaxin-2, RANTES, MCP-3, and MCP-4.17 Com-
pound 20 effectively blocked the calcium mobilization
and chemotaxis mediated by eotaxin and MCP-4 with
potency identical to inhibition of eotaxin (Table 4),
indicating it will block all CCR3 ligand interactions.
The chemokine receptor selectivity of small molecule
antagonists is therefore critical in obtaining a desired
pharmacological effect whilst minimizing the potential
for undesired side effects. The CCR3 selectivity of 20
against a panel of related chemokine and non-chemokine
7TM GPCRs was determined and the compound found
to be >2500-fold selective for CCR3 over the other 10
receptors in the screen. Compound 20 represents one of
the most potent and selective CCR3 antagonists reported.
7. Bertrand, C. P.; Ponath, P. D. Expert Opin. Invest. Drugs
2000, 9, 43.
8. (a) Kitaura, M.; Nakajima, T.; Imai, T.; Harada, S.;
Combadiere, C.; Tiffany, H. L.; Murphy, P. M.; Yoshie, O.
J. Biol. Chem. 1996, 271, 7725. (b) Daugherty, B. L.; Sici-
liano, S. J.; DeMartino, J. A.; Malkowits, L.; Sirotina, A.;
Springer, M. S. J. Exp. Med. 1996, 183, 2349. (c) Ponath,
P. D.; Qin, S.; Post, T. W.; Wang, J.; Wu, L.; Gerard, N. P.;
Newman, W.; Gerard, C.; Mackay, C. R. J. Exp. Med. 1996,
183, 2437.
9. The IC50 values reported are the mean of duplicate samples
from at least two determinations.15
10. All compounds gave spectroscopic and analytical data
consistent with their assigned structure.
11. Bodanszky, M.; Bodanszky, A. In The Practice of Peptide
Synthesis; Springer: Berlin, 1994.
12. (a) Widmer, U. Synthesis 1983, 135. (b) Wang, S. S.;
Gisin, B. F.; Winter, D. P.; Makofske, R.; Kulesha, I. D.;
Tzougraki, C.; Meienhofer, J. J. Org. Chem. 1997, 42, 1286.
13. Steglich, W.; Hofle, G. Angew. Chem. 1969, 81, 1001.
14. MacLeod, A. M.; Cascieri, M. A.; Merchant, K. J.;
Sadowski, S.; Hardwicke, S.; Lewis, R. T.; MacIntyre, D. E.;
Metzger, J. M.; Fong, T. M.; Shepheard, S.; Tattersall, F. D.;
Hargreaves, R.; Baker, R. J. Med. Chem. 1995, 38, 934.
15. White, J. R.; Lee, J. M.; Dede, K.; Imburgia, C. S.; Jur-
ewicz, A. J.; Dhanak, D.; Christmann, L. T.; Darcy, M. G.;
Widdowson, K.; Foley, J. J.; Schmidt, D. B.; Sarau, H. M. J.
Biol. Chem. 2000, 275, 36626.
16. Berkout, T. A.; Sarau, H. M.; Moores, K.; White, J. R.;
Elshourbagy, N.; Appelbaum, E.; Reape, R.-J.; Brawner, M.;
Makwana, J.; Foley, M. S.; Schmidt, D. B.; Imburgia, C.;
McNulty, D.; Matthews, J.; O’Donnell, K.; O’Shannessy, D.;
Scott, M.; Croot, P. H. E.; Macphee, C. J. Biol. Chem. 1997,
272, 16404.
A phenylalanine-based antagonist of the human CCR3
has been discovered and initial structure–activity rela-
tionship studies revealed a key role for both the N-acyl
and ester moieties. Parallel, solution-phase library synth-
esis was utilized to delineate the structure–activity rela-
tionship and highly potent, CCR3-selective antagonists
have been synthesized. In contrast to other small molecule
CCR3 antagonists reported,18 the presence of a basic
nitrogen functionality within the molecule is not a
requirement for good CCR3 binding affinity in this series,
suggesting a unique binding site.
References and Notes
17. Pease, J. E.; Wang, J.; Ponath, P. D.; Murphy, P. M. J.
Biol. Chem. 1998, 273, 19972.
1. (a) Ward, S. G.; Westwick, J. Biochem. J. 1998, 3333, 457.
(b) Baggiolini, M.; Dewald, B.; Moser, B. Annu. Rev. Immunol.
1997, 15, 675. (c) Epstein, F. H. N. Engl. J. Med. 1998, 338,
18. Sabroe, I.; Peck, M. J.; Kenlen, J. V. B.; Jorritsma, A.;
Simmons, G.; Clapham, P. R.; William, T. J.; Pease, J. E. J.
Biol. Chem. 2000, 275, 25985.