Synthesis and evaluation of heteroarylalanine diacids as potent and selective neutral endopeptidase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.03.109

Heteroarylalanine derivatives 4 were designed as potential inhibitors of neutral endopeptidase (NEP EC 3.4.24.11). Selectivity over other zinc metalloproteinases was explored through occupation of the S2′ subsite within NEP. Structural optimisation led to

Full text of DOI:10.1016/j.bmcl.2011.03.109

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3404–3406
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of heteroarylalanine diacids as potent and
selective neutral endopeptidase inhibitors
⇑
Melanie S. Glossop , Richard J. Bazin, Kevin N. Dack, David N. A. Fox, Graeme A. MacDonald, Mark Mills ,
Dafydd R. Owen, Chris Phillips, Keith A. Reeves, Tracy J. Ringer, Ross S. Strang , Christine A. L. Watson
Pfizer Global Research and Development, Ramsgate Road, Sandwich CT13 9NJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Heteroarylalanine derivatives 4 were designed as potential inhibitors of neutral endopeptidase (NEP EC
3.4.24.11). Selectivity over other zinc metalloproteinases was explored through occupation of the S20 sub-
site within NEP. Structural optimisation led to the identification of 5-phenyl oxazole 4f, a potent and
selective NEP inhibitor. A crystal structure of the inhibitor bound complex is reported.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 14 March 2011
Revised 25 March 2011
Accepted 29 March 2011
Available online 5 April 2011
Keywords:
Neutral endopeptidase
Angiotensin converting enzyme
Zinc metalloproteinases, Oxazole
Oxadiazole
Neutral endopeptidase (NEP, EC 3.4.24.11) is a zinc metallopro-
teinase involved in the degradation of atrial natriuretic peptide
(ANP), a 28 amino acid vasoactive peptide that exhibits diuretic,
natriuretic and vasodilatory activities.1 Potentiation of ANP
through NEP inhibition may be of clinical benefit for the treatment
of hypertension and congestive heart failure. Candoxatril (previ-
ously disclosed by these laboratories) is the indanyl ester prodrug
of the selective NEP inhibitor Candoxatrilat 1. Candoxatril is lim-
ited by its relatively high dose and short half-life in man.2 The
identification of a potent and selective inhibitor of NEP (over
Angiotensin Converting Enzyme (ACE) and other related zinc
metalloproteinases) with improved pharmacokinetics was sought.
Dual inhibition of ACE and NEP as a strategy for treating hyper-
tension has been extensively investigated, including contributions
from these laboratories through the discovery of Sampatrilat 2.3
Given that both targets are related zinc metalloproteinases, dual
enzyme inhibition can be achieved within a single pharmacophore
possessing a modified peptide backbone linked through to a zinc
ion chelator (carboxylic acid) (Table 1). Selectivity for either ACE
or NEP is highly influenced by the nature of the additional func-
tionality present in the S20 subsite. Selective NEP inhibitors do
oxatrilat 1, Fig. 1), however both ACE and NEP can tolerate biaryl-
alanine substituents as disclosed by Zambon in their dual ACE/NEP
inhibitor 3 (Fig. 1).4 Exploration of the S20 subsite of NEP within the
Candoxatrilat template may enable identification of a selective NEP
inhibitor with enhanced properties. This letter explores the lipo-
philic binding pocket at S20 with conformationally restricted het-
eroarylalanines in place of the tyrosine in Sampatrilat and the
biphenyl group of compound 3. The preferred natural amino acid
stereochemistry was employed, enabling the design and synthesis
of all analogues to be derived from L-aspartic acid.
Oxazoles 7 were synthesised by peptide coupling of the appro-
priately substituted aminoethanol with N-BocAsp-OBn 5, to give
the corresponding amides 6. Oxidation of the alcohol with
Dess–Martin periodinane and cyclisation using either iodine and
triphenylphosphine5 or dibromotetrachloroethane and triphenyl-
phosphine,6 followed by final N-Boc deprotection with
trifluoroacetic acid (TFA) furnished the desired amino esters 7a–c
expressing the heteroaryl alanine (Scheme 1).
1,2,4-oxadiazole 9 was obtained by acylating N-hydroxybenz-
amidine with 5 to afford acylamidoxime 8, followed by cyclisation
as a melt. TFA deprotection furnished the desired amino ester
intermediate as a second class of heterocyclic variants (Scheme 1).
Isomeric oxazoles 11a–c were synthesised via N-CbzAspOEt 10,
which was obtained according to literature precedent from
Cbz-aspartic acid anhydride.7 Amide bond formation with the
appropriate aminoketone8 followed by POCl3 cyclisation produced
the desired heterocycles. N-Cbz deprotection with HBr/AcOH
furnished the amino esters (Scheme 2).
not always require an a-amino acid at the C-terminus (e.g., Cand-
⇑
Corresponding author. Tel.: +44 01304 644549.
E-mail address: mel.s.glossop@pfizer.com (M.S. Glossop).
Present address: AstraZeneca R&D Charnwood, Bakewell, Road, Loughborough
LE11 5RH, UK.
Present address: Novartis AG, Basel CH-4002, Switzerland.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.03.109

M. S. Glossop et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3404–3406
3405
S1'
S1
OMe
O
NH2
NHSO2Me
O
OH
S2'
HN
H
H
HO
N
HO
N
O
O
O
O
CO2H
CO2H
Zn
Candoxatrilat
1
Sampatrilat 2
Ph
R
O
Ph
O
Heteroaryl
H
N
HO
H
HO
N
N
H
O
CO2H
O
CO2H
3
4
Figure 1.
R
R
Table 1
NEP and ACE activity of 1–4
HO
b, c or d, e
O
N
O
NH
R
O
a
BocNH
H2N
Het
CO2Bn
CO2Bn
H
N
HO
O
OH
6
7a R = Ph
7b R = Et
BocNH
O
7c R = iBu
O
OH
CO2Bn
5
Ph
Ph
N
Compound
R
Het
NEP
IC50 (nM)
ACE
IC50 (nM)
H2N
N
O
g, h
f
O
N
O
1
2
3
—
—
—
—
—
—
7.8
0.6a
28b
Inactive
3.4
BocNH
H2N
1.5b
CO2Bn
CO2Bn
O
Et
4a
4b
4c
4d
4e
4f
CH3
22
130
75
N
N
N
O
N
O
9
8
O
iBu
Ph
Ph
Ph
Ph
CH3
20
Scheme 1. Reagents and conditions: (a) H2NCH(R)CH2OH, WSCDI, HOBt, NMM,
DCM, 54–85%; (b) Dess–Martin periodinane, DCM, 66–73%; (c) I2, PPh3, NEt3, THF,
25%; (d) Br2C2Cl4, PPh3, DBU, 2,6-di-tbutylpyridine, 30–69%; (e) TFA, DCM, 95–
100%; (f) N-hydroxybenzamidine, WSCDI, HOBt, NMM, DCM, 30%; (g) Melt, 114 °C,
88%; (h) TFA/DCM, quant.
O
N
O
N
CH3
7
25
CH3
0.6a
4.8
0.3a
82
CH3O
CH3O
130
270
A similar protocol was employed to obtain the isomeric 1,3,4-
oxadiazole 12. Benzoyl hydrazine was reacted with the acid chlo-
ride of 10 in good yield, followed by cyclisation in the presence
of chlorodimethylimidazolinium tetrafluoroborate.9 N-Cbz depro-
tection under palladium catalysed hydrogenolysis furnished the
desired amino ester (Scheme 2).
Amide bond formation of the synthesised amino esters with
carboxylic acids 13a and b10 followed by a two step deprotection
strategy (TFA removal of the tBu ester followed by base hydrolysis
of the ethyl ester) completed the synthesis of the novel heterocy-
clic diacids 4 (Scheme 3). These compounds were tested for their
ability to inhibit both ACE and NEP activity in vitro. The experi-
mental details for the pharmacological assays have been previ-
ously described.11 The results are summarised in Table 1.
Table 1 shows the structure–activity relationship for the con-
formationally restricted heteroarylalanine analogues 4. Within
the 4-substituted oxazoles 4a–c, we found that increasing lipophil-
N
N
4g
4h
CH3O
CH3O
5
210
195
Ph
O
Cl
0.3a
O
Ph
4i
4j
CH3O
CH3O
6
>3
58
lM
O
N
N
N
Ph
1.9
N
O
a
Where IC50 was <1.5 nM, lower enzyme concentrations were used to determine
the true IC50, below the tight binding limit of the standard assay.
b
Literature values.

3406
M. S. Glossop et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3404–3406
R
Ar
A
a, b, c
11a R = H, Ar = Ph
11b R = Me, Ar = Ph
11c R = H, Ar = 4-ClPh
N
O
CO2H
H2N
CbzNH
CO2Et
CO2Et
10
Ph
d, e, f
N
O
N
12
H2N
N542
R110
B
R102
H587
CO2Et
Y697
Scheme 2. Reagents and conditions: (a) SOCl2, NEt3, DMF, H2NCH(R)C@OAr, 50–
77%; (b) POCl3, 100 °C, toluene, 50–73%; (c) HBr AcOH 65–100%; (d) (i) (COCl)2,
DMF, DCM, (ii) Py, DCM, benzoyl hydrazine, 62%; (e) (i) 2-chloro-1,3dimethyl
imidazolinium tetrafluoroborate, NEt3, DCM; (ii) Toluene, 80 °C, 33%; (f) 10% Pd/C,
EtOH, 50%.
W693
R717
T696
M579
Figure 2. (A) Crystal. Structure of 4c complex with human NEP.
A surface
Het
representation illustrating the 5-phenyl oxazole packing within the S20 pocket.
The surface is coloured according to hydrophobicity with brown indicating the
most hydrophobic areas of the binding site. Zn is represented as a grey sphere. (B)
The binding site with key residues highlighted.
H2N
O
OBn/OEt
R
a, b, c
+
Het
OH
for NEP. The i.v. pharmacokinetics of compound 4f compare
favourably with those of candoxatrilat.2 Compound 4f has an i.v.
clearance of 4.8 ml/min/kg and a half-life of 2.1 h in rat. Oral bioav-
ailablity for 4f was found to be low (1.7%) in a seperate rat study.
In conclusion, this letter demonstrates that selective NEP inhibi-
tion can be achieved with heteroarylalanine derivatives 4. Natural
amino acid stereochemistry for the novel heterocyclic substituents
at P20 in combination with a methoxyethyl S1 substituent enables
phenyl oxazoles to exhibit potent and selective NEP inhibition.
Compound 4f has a useful pharmacokinetic half-life in rat.
H
N
R
HO2C
O
O
OH
tBuO2C
O
4
13a R = CH3
13b R = CH3O
Scheme 3. Reagents and conditions: (a) WSCDI, HOBt, NMM, DCM, 50–92%; (b)
TFA, DCM, 85–98%; (c) NaOH, dioxan, H2O, 53–96%.
Acknowledgments
icity through to the 4-phenyloxazole 4c increased NEP inhibition;
however selectivity over ACE was modest. Co-crystallisation of 4c
with human NEP12 shows the binding mode of the P20 phenyl oxa-
zole in the deep, lipophilic S20 pocket (Fig. 2). Key interactions in-
clude the left hand acid (as drawn) zinc coordination, and
hydrogen bonding between the inhibitor amide group and Arg
717 and Asn 542. The natural amino acid stereochemistry enables
the right-hand side acid to interact with key Arg 110 and 102 res-
idues in the active site, orientating the phenyl oxazole deep into
the S20 cavity. The subsite is fully occupied and replacement of
the pendant phenyl in P20 for a less lipophilic alkyl substituent
such as ethyl and isobutyl (4a, 4b) only served to lower NEP
activity.
The authors would like to thank Ed Hawkeswood and Justin
Kewney for ACE and NEP enzyme assays. We also thank Dr. Dennis
Smith for his assistance.
References and notes
1. Bevan, E. G.; Connell, J. M. C.; Doyle, J.; Carmichael, H. A.; Davies, D. L.; Lorimer,
A. R.; McInnes, G. T. J. Hypertension 1992, 10, 607.
2. (i) Kaye, B.; Brearley, C. J.; Cussans, N. J.; Herron, M.; Humphreys, M. J.; Mollatt,
A. R. Xenobiotica 1997, 27, 1091; (ii) McDowell, G.; Nicholls, D. P. Exp. Opin.
Invest. Drugs 1999, 8, 79.
3. Bralet, J.; Schwartz, J.-C. Trends Pharmacol. Sci. 2001, 22, 106.
4. Pellacini, F.; Norcini, G.; Botta, D.; Romagnano, S.; Santangelo, F. WO 95/35307,
1995.
The isomeric 5-phenyloxazole 4d not only exhibited increased
potency against NEP, but also a decreased ACE inhibition thus pro-
viding encouraging selectivity. Changing the P1 group from an n-
propyl to a more polar methoxyethyl at the left hand acid (as seen
in compound pairs 4c–4e and 4d–4f) decreased ACE inhibition fur-
ther such that 4f showed around 1000-fold selectivity over ACE.
Introduction of a second substituent onto the oxazole, such as with
4-methyl-5-phenyl-oxazole 4g, gave a slight reduction in activity,
whilst simple para- chlorination 4h of the phenyl group gave sim-
ilar levels of activity to the parent 4f. The two isomeric oxadiazoles
4i and 4j offered no pharmacology advantage over the oxazole
systems.
5. Wipf, P.; Venkatraman, S. J. Org. Chem. 1995, 60, 7224.
6. Pattenden, G.; Plowright, A. T. Tetrahedron Lett. 2000, 41, 983.
7. Voelter, W.; Zech, K.; Grimminger, W.; Breitmaier, E.; Jung, G. Chem. Ber. 1972,
105, 3650.
8. Han, Y.; Hu, H. Synthesis 1990, 2, 122. Han, H.; Hu, H. T.L. 1989, 39, 5285.
9. Isobe, T.; Ishikawa, T. J. Org. Chem. 1999, 64, 6989.
10. (i) Challenger, S.; Cook, A. S.; Gillmore, A. T.; Middleton, D. S.; Pryde, D. C.;
Stobie, A. WO0279143, 2002;(ii)Barber, C.G.; Cook, A. S.;Maw, G. N.;Pryde, D. C.;
Stobie, A. WO 0202513, 2002.
11. Maw, G. N.; Wayman, C. P. EP 1097719, 2001.
12. (i) Oefner, C.; D’Arcy, A.; Hennig, M.; Winkler, F. K.; Dale, G. E. J. Mol. Biol. 2000,
296, 341. Data were collected using a Raxis IV image plate mounted on an F-RD
rotating anode source. The CCP4 suite of programs was used for data
processing. (ii) The CCP4 Suite: Programs for Protein Crystallography. Acta
Cryst. 1994 D50, 760. The crystals diffracted to 2.4 Å resolution, the structure
was solved by difference Fourier analysis. The BUSTER refinement package was
used. (iii) Blanc, E, Roversi, P., Vonrhein, C., Flensburg, C., Lea, S. M. And
Bricogne, G. Acta Cryst. 2004 D60, 2210. Coordinates and Structure Factors are
deposited with the PDB with accession code s2yb9 and r2yb9 respectively.
Compound 4f is 25-fold more potent against NEP than Candox-
atrilat 1, and significantly more selective over ACE than 2 or 3.
Interactions within the S20 lipophilic pocket are well optimised