N-formyl hydroxylamine containing dipeptides: Generation of a new class of vasopeptidase inhibitors

Article abstract of DOI:10.1016/S0960-894X(99)00671-X

Four primary zinc-binding pharmacophores (thiols, carboxylates, phosphorus acids, and hydroxamates) have been utilized in generating inhibitors of zinc metalloproteases such as ACE, NEP, the MMPs, and ECE. Although compounds which inhibit the activity of both ACE and NEP (vasopeptidase inhibitors, VPIs) have been reported which incorporate a thiol, carboxylate, or phosphorus acid pharmacophore, the generation of hydroxamate based vasopeptidase inhibitors has remained elusive. Herein we report the first potent vasopeptidase inhibitors which were generated from the incorporation of conformationally restricted dipeptide mimetics to an N-formyl hydroxylamine zinc-binding group. Compounds such as 13c and 13d are among the most potent in this series, exhibiting in vitro activity comparable to other classes of inhibitors. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(99)00671-X

Bioorganic & Medicinal Chemistry Letters 10 (2000) 257±260
N-Formyl Hydroxylamine Containing Dipeptides:
Generation of a New Class of Vasopeptidase Inhibitors
Je€rey A. Robl,* Ligaya M. Simpkins and Magdi M. Asaad
Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400, Princeton, NJ 08543-5400, USA
Received 14 October 1999; accepted 23 November 1999
AbstractÐFour primary zinc-binding pharmacophores (thiols, carboxylates, phosphorus acids, and hydroxamates) have been uti-
lized in generating inhibitors of zinc metalloproteases such as ACE, NEP, the MMPs, and ECE. Although compounds which
inhibit the activity of both ACE and NEP (vasopeptidase inhibitors, VPIs) have been reported which incorporate a thiol, car-
boxylate, or phosphorus acid pharmacophore, the generation of hydroxamate based vasopeptidase inhibitors has remained elusive.
Herein we report the ®rst potent vasopeptidase inhibitors which were generated from the incorporation of conformationally
restricted dipeptide mimetics to an N-formyl hydroxylamine zinc-binding group. Compounds such as 13c and 13d are among the
most potent in this series, exhibiting in vitro activity comparable to other classes of inhibitors. # 2000 Elsevier Science Ltd. All
rights reserved.
The development of vasopeptidase inhibitors (VPIs:
dual inhibitors of angiotensin converting enzyme (ACE)
and neutral endopeptidase (NEP)) is an active ®eld of
study in cardiovascular research, and several excellent
reviews have been published describing this topic in
detail.1±5 Their dual mechanism of action, attenuation
of angiotensin II and potentiation of atrial natriuretic
peptide and bradykinin, a€ords the potential to e€ect
greater reductions in blood pressure in a wider patient
population as compared to other established anti-
hypertensives. Several vasopeptidase inhibitors are cur-
rently in clinical trials for the treatment of hypertension
and congestive heart failure and initial reports have
demonstrated ecacy in man.6±9
described which contain a hydroxamate moiety, the
utilization of this pharmacophore in the design of vaso-
peptidase inhibitors has not yet been reported. In an
attempt to ®nd agents within this class, we were intri-
gued by a report from Roques et al.15 which depicted
the use of N-formyl hydroxylamine amino acids (i.e. 2)
as NEP inhibitors (Fig. 2). Although somewhat less
potent than the related hydroxamic acids, this func-
tional group served as an attractive yet largely unex-
plored16,17 zinc-binding pharmacophore which could be
incorporated with conformationally restricted dipeptide
surrogates to give compounds of the type 3.
The requisite acids 4 and 5 were generated following the
procedure described by Roques.15 In the case of the b-
benzyloxylamino acid 5, the compound could be for-
mylated to a€ord 6 or was resolved via its (1R,2S)-( )-
ephedrine salt to give 7 as a single enantiomer.18
Both ACE and NEP are zinc metalloproteases and a
historical survey of the ®eld has demonstrated that four
primary zinc-binding pharmacophores (thiols, carboxy-
lates, phosphorus acids, and hydroxamates) have been
incorporated into the design of potent and selective
inhibitors of each enzyme. Additionally, a variety of
vasopeptidase inhibitors have been described which
contain a thiol, carboxylate, or phosphorus acid zinc-
binding pharmacophore. Representative examples
include omapatrilat,10,11 sampatrilat,12,13 and com-
pound 114 respectively (Fig. 1). Although potent and
selective inhibitors of zinc metalloproteases have been
In order to compare the e€ect of chain length (n), acyl
substituent (Y), and the incorporation of confor-
mationally restricted dipeptide surrogates in this frame-
work, a sub-set of compounds based on Leu-Phe as
the dipeptide was initially explored. Leu-Phe was chosen
as a representative dipeptide since the related mercap-
toacetyl derivative exhibited IC50s of 73 and 54 nM
versus ACE and NEP, respectively.19 Thus, EDAC/
HOBT mediated coupling of 4, 5 or 6 with Leu-
Phe(OBn) (8) a€orded the adducts 9a±c in good yield
(Scheme 2). In the case of 9a and 9b, simultaneous
removal of the benzyl ether and ester via hydrogenolysis
*Corresponding author. Tel.: +1-609-818-5048; fax: +1-609-818-
3550; e-mail: roblj@bms.com
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(99)00671-X

258
J. A. Robl et al. / Bioorg. Med. Chem. Lett. 10 (2000) 257±260
a€orded the ®nal products 11a and 11b, respectively.
The related N-acetyl 11c and hydroxy urea 11d ana-
logues were generated by treatment of 9c with acetic
anhydride or trimethylsilylisocyanate to give 10c and
10d respectively, followed by hydrogenolysis.
potency against both enzymes. The shortened derivative
11a was substantially less active versus both enzymes.
Paralleling results obtained by Roques in the N-acyl
hydroxyl amino acid series, the N-acetyl 11c and hydroxy
urea 11d derivatives were poorly active in this series,
highlighting the absolute requirement for the N-formyl
substituent. In previous studies, we demonstrated that
the incorporation of suitable conformationally restricted
dipeptide surrogates in mercaptoacyl dipeptides leads to
a substantial enhancement in in vitro inhibitory active
In vitro inhibition of ACE and NEP for compounds
11a±d (diastereomeric mixture of isomers at *) are
compared in Table 1. Although ꢀ12-fold more selective
for ACE than NEP, compound 11b exhibited reasonable
Figure 1.
Figure 2.
Scheme 1.
Scheme 2. (a) compound 4, 5 or 6, EDAC, HOBT, CH2Cl2, rt (61±73%); (b) for compound 9d to 10d: xs TMSNCO, THF, re¯ux, (68%); for
compound 9c to 10c: Ac2(O), THF, rt, (80%); (c) H2, 10% Pd/C, MeOH, rt (94±100%).

J. A. Robl et al. / Bioorg. Med. Chem. Lett. 10 (2000) 257±260
259
Scheme 3. Conditions: For 12a: (a) 4, EDAS, HOBT, CH2Cl2 (84%); (b) H2, 10% Pd/C, MeOH, rt, then separate diastereomers by HPLC (51%).
For 12b±d: (a) 5, EDAC, HOBT, CH2Cl2 60±72%); (b) HCO2H, A2O, THF, 0 ꢁC (77±97%); (c) H2, 10% Pd/C, MeOH, rt (90±97%); (d) NaOH,
H2O, MeOH, then H3O+ (85±94%).
Table 1.
vasopeptidase inhibitors disclosed to date. The generally
soft and di€use zinc chelating ability of the N-formyl
hydroxylamine phamacophore closely parallels that of
the related thiols. Unlike carboxylates and phosphorus
acids, the N-formyl hydroxylamines (like the thiols)
exhibit higher pKa values and their incorporation in
charged dipeptides may represent and advantage in the
design of VPIs with greater propensity for oral absorp-
tion. The evaluation of these compounds in vivo as well
as the generation of other hydroxamate based VPIs will
be the topic of future disclosures.
Compound
Stereochemistry
at *
ACE
(IC50, nM)
NEP
(IC50, nM)
omapatrilat
11a
11b
11c
11d
Ð
R,S
R,S
R,S
R,S
R
R
R
R
R
6
9
93,000
1270
214,000
705,000
290
1.7
18
1.8
1.3
228,000
106
461,000
529,000
210,000
93
5.4
3.4
5.3
13a
13a0
13b
13c
13d
Acknowledgement
against both enzymes.20,21 In analogy, a representative
sample of mono- and bicyclic dipeptide surrogates
(compounds 12a±d)22 were incorporated as Leu-Phe
replacements in 11b (Scheme 3). In the case of 12a,
coupling with 6 followed by hydrogenolysis a€orded a
diastereomeric mixture of 13a and 13a0 which were
separated by preparative HPLC and individually
assayed. In the case of 12b±d, coupling of the enatio-
merically pure 7 a€orded single diastereomers which
were subsequently formylated, debenzylated, and saponi-
®ed to provide 13b±d.
Appreciation is expressed to Mary Malley for perform-
ing the crystallographic analysis of the ephedrine salt of
compound 7.
References and Notes
1. Robl, J. A.; Trippodo, N. C.; Petrillo, E. W. In Anti-
hypertensive Drugs; Van Zwieten, P. A., Greenlee, W. J., Eds.;
Harwood Academic Publishers: Amsterdam, 1997; p. 113.
2. Bhagwat, S. In XIVth Internat. Sympos. Med. Chem., Proc.
1997, p. 15.
3. De Lombaert, S.; Chatelain, R. E.; Fink, C. E.; Trapani, A.
J. Curr. Pharmaceut. Des. 1996, 2, 443.
Taking into consideration diastereomeric purity, incor-
poration of the benzofused oxazepine in 13a0 resulted in
a 300-fold increase in potency as compared to 11b
against NEP while activity versus ACE was retained.
Comparison of 13a0 with its stereoisomer 13a clearly
indicates a preference for the R isomer, paralleling
observations made in the related mercaptoacyl dipep-
tide series of VPIs. The most active in this series, com-
pounds 13c and 13d, exhibit high potency against both
enzymes and are comparable in activity to the most
potent VPIs reported to date.
4. Fink, C. A. Exp. Opin. Ther. Patents 1996, 6, 1147.
5. Flynn, G. A.; French, J. F.; Dage, R. C. In Hypertension:
Pathophysiology, Diagnosis, and Management, 2nd ed.; Laragh,
J. H., Brenner, B. M., Eds.; Raven Press, Ltd.: New York,
1995; p. 3099.
6. Graul, A.; Leeson, J.; Castaner, J. Drugs of the Future 1999,
24, 269.
7. Norton, G. R.; Woodiwiss, A. J. et al. Am. J. Hypertens.
1999, 12, 563.
8. Rousso, P.; Buclin, T.; Nussberger, J.; Brunner-Ferber, F.;
Brunner, H. R.; Biollaz, J. J. Cardiovasc. Pharmacol. 1998, 31,
408.
9. Laurent, S.; Boutouyrie, P. et al. J. Hypertens. 1997, 15
(suppl 4), Abs 170.
10. Robl, J. A.; Sun, C.-Q.; Stevenson, J.; Ryono, D. E.;
Simpkins, L. M.; Cimarusti, M. P.; Dejneka, T.; Slusarchyk,
W. A.; Chao, S.; Stratton, L.; Misra, R. M.; Bednarz, M. S.;
Conclusion
The activity demonstrated by this series of compounds
is signi®cant, representing the fourth class of potent

260
J. A. Robl et al. / Bioorg. Med. Chem. Lett. 10 (2000) 257±260
Asaad, M. M.; Cheung, H. S.; Abboa-O€ei, B. E.; Smith, P.
L.; Mathers, P. D.; Fox, M.; Schae€er, T. R.; Seymour, A. A.;
Trippodo, N. C. J. Med. Chem. 1997, 40, 1570.
11. Trippodo, N. C.; Robl, J. A.; Asaad, M. M.; Fox, M.;
Panchal, B.; Schae€er, T. R. Am. J. Hypertens. 1998, 11, 363.
12. Brealy, C. J.; Allen, M. J.; Grimwood, V. C.; MacMahon,
M. Eur. Meet. Hypertens. 1995, 7th Milan Abs 107.
13. Williams Lawson, J. S.; Barclay, P. L.; Barnes, P. E.;
Shepperson, N. B. Eur. Meet. Hypertens. 1995, 7th Milan Abs
914.
14. De Lombaert, S.; Tan, J.; Stamford, L.; Sakane, Y.; Berry,
C.; Ghai, R. D. Bioorg. Med. Chem. Lett. 1994, 4, 2715.
15. Fournie-Zaluski, M.-C.; Coulaud, A.; Bouboutou, R.;
Chaillet, P.; Devin, J.; Waksman, G.; Costentin, J.; Roques, B.
P. J. Med. Chem. 1985, 28, 1158.
16. Jin, Y.; Kim, D. H. Bioorg. Med Chem. Lett. 1998, 8, 3515.
17. Handa, B. K.; Johnson, W. H.; Machin, P. J. Eur. Pat.
Appl. EP 236872 A2, 1987.
18. For the resolution of compound 5 to compound 7: a solu-
tion of acid 5 (2.56 g, 8.9 mmol) in CH3CN (ꢀ20 mL) was
treated with (1R,2S)-( )-ephedrine (1.52 g) and swirled until
homogeneous. Most of the CH3CN was removed by rotary
evaporation and the residue was redissolved in Et2O (ꢀ25 mL)
followed by hexane (ꢀ16 mL) until the solution became tur-
bid. The solution was seeded and let stand overnight. The
resulting precipitate was collected by ®ltration and washed
with 1:1-Et2O:hexane to a€ord a white solid (2.10 g, [a]d
16.4ꢁ (c=0.6, CH2Cl2). Repeated crystallization from hot
CH3CN a€orded enatiomerically pure material (mp 118 ꢁC,
[a]d 19.7ꢁ (c=0.4, CH2Cl2). Absolute con®guration of 7 (as
shown) was determined by single crystal X-ray analysis of the
puri®ed ephedrine salt. The free acid 7 (an oil) was obtained
by partitioning the salt between EtOAc and 5% aqueous
KHSO4 (pH adjusted to 2.5).
19. The mercaptoacetyl derivative of 11, HSCH(Bn)CO-Leu-
Phe-CO2H, is readily generated by coupling of (S)-a-(acetyl-
thio)-2-benzenepropionic acid with 8 followed by saponi®ca-
tion and acidi®cation. Representative experimentals can be
found in Robl, J. A. et al. J. Med. Chem. 1996, 39, 494.
20. Robl, J. A.; Simpkins, L. M.; Stevenson, J.; Sun, C. Q.;
Murugesan, N.; Barrish, J. C.; Asaad, M. M.; Bird, J. E.;
Schae€er, T. R.; Trippodo, N. R.; Petrillo, E. W.; Kar-
anewsky, D. S. Bioorg. Med. Chem. Lett. 1994, 4, 1789.
21. Robl, J. A.; Simpkins, L. M.; Sulsky, R.; Sieber-McMaster,
E.; Stevenson, J.; Kelly, Y. F.; Sun, C. Q.; Misra, R. N.;
Ryono, D. E.; Asaad, M. M.; Bird, J. E.; Trippodo, N. C.;
Karanewsky, D. S. Bioorg. Med. Chem. Lett. 1994, 4, 1795.
22. Compounds 12a and 12d were described in references 10a
and 10b respectively. Compound 12b was reported in Robl, J.
A.; Sulsky, R.; Sieber-McMaster, E.; Ryono, D. E.; Cimarusti,
M. P.; Simpkins, L. M.; Karanewsky, D. S.; Chao, S.; Asaad,
M. M.; Seymour, A. A.; Fox, M.; Smith, P. L.; Trippodo, N.
C. J. Med. Chem. 1999, 42, 305. Compound 12c was described
in Flynn, G. A.; Beight, D. W.; Mehdi, S.; Koehl, J. R.; Gir-
oux, E. L.; French, J. F.; Hake, P. W.; Dage, R. C. J. Med.
Chem. 1993, 36, 2420.