Aminomethylpyrimidines as novel DPP-IV inhibitors: A 105-fold activity increase by optimization of aromatic substituents

Article abstract of DOI:10.1016/j.bmcl.2004.01.019

The influence of aromatic substitution on a newly discovered class of inhibitors of dipeptidyl peptidase IV was investigated. A 10⁵-fold increase in potency was achieved by the optimization of aromatic substituents in a parallel chemistry progr

Full text of DOI:10.1016/j.bmcl.2004.01.019

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1491–1493
Aminomethylpyrimidines as novel DPP-IV inhibitors:
A 105-fold activity increase by optimization of
aromatic substituents
Jens-Uwe Peters,* Silja Weber, Stephane Kritter, Peter Weiss, Angelina Wallier,
Markus Boehringer, Michael Hennig, Bernd Kuhn and Bernd-Michael Loeffler
Pharma Division, Preclinical Research, F. Hoffmann-La Roche Ltd, CH-4070 Basel, Switzerland
Received 29 October 2003; revised 23 December 2003; accepted 9 January 2004
Abstract—The influence of aromatic substitution on a newly discovered class of inhibitors of dipeptidyl peptidase IV was
investigated. A 105-fold increase in potency was achieved by the optimization of aromatic substituents in a parallel chemistry
program. The observed SAR could be explained by an X-ray structure of the protein–ligand complex.
# 2004 Elsevier Ltd. All rights reserved.
Dipeptidyl peptidase IV (DPP-IV) cleaves and inacti-
vates glucagon-like peptide 1 (GLP-1),1 which is an
important stimulator of insulin secretion.2 Inhibition of
DPP-IV increases the level of circulating GLP-1 and
thus increases insulin secretion.3 An improved insulin
secretion could moderate hyperglycaemia in type 2 dia-
betes. Consequently, DPP-IV inhibition has been pro-
posed as a new treatment of type 2 diabetes.4 Several
series of DPP-IV inhibitors have been published as
potential new medicines.5 Clinical proof of concept has
already been established in phase II with the DPP-IV
inhibitor NVP-DPP728.6,7 In a search for novel struc-
tures, we have identified 6-methylenedioxyphenyl-amino-
methylpyrimidine 1a as a relatively weak inhibitor of
DPP-IV in a high-throughput screen.
small substituents (Me, Cl, MeO, F and CF3) in differ-
ent positions at both phenyl rings. 5-Cyanopyrimidines
4 were obtained in a first step by the reaction of benzyl-
amidines 2 and arylidenemalononitriles 3 under basic
conditions (Scheme 1).8 Yields of 4 were greatly
enhanced by treating the reaction mixtures with
KMnO4, e.g. from 13% to 52% for 4c-H. Subsequently,
the nitrile functionality was reduced to give the desired
5-aminomethylpyrimidines 1. The syntheses were
carried out in a parallel fashion using disposable
polypropylene tubes as reaction vessels and automated
preparative, reversed-phase HPLC for purification. The
identities and purities of 4 and 1 were assessed by
HPLC/MS.9
Compounds 1 were evaluated for their ability to inhibit
DPP-IV mediated cleavage of Ala-Pro-7-amido-4-tri-
fluoromethylcoumarin in a fluorogenic assay.10 Table 1
documents the SAR of single substituents at the 6-
phenyl ring of 1. Removal of the methylenedioxy sub-
stituent of 1a gave a reference compound, 1b-H(R ¼H,
Table 1) with a slightly reduced activity. Introduction of
ortho methyl (1b-o-Me), ortho chloro (1b-o-Cl), or ortho
methoxy (1b-o-OMe) substituents increased the inhibi-
tory activity by an order of magnitude. A similar
potency increase was observed for the introduction of
para methyl (1b-p-Me), para chloro (1b-p-Cl), or para
trifluoromethyl (1b-p-CF3) sustituents. The effects of
other ortho and para substituents (1b-o-F, 1b-o-CF3,
1b-p-OMe, 1b-p-F) were generally favorable but less
With the goal of finding analogues of 1a with increased
potency, we prepared 5-aminomethylpyrimidines 1 with
* Corresponding author. Tel.: +41-61-6882636; fax: +41-61-6886459;
e-mail: jens-uwe.peters@roche.com
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.01.019

1492
J.-U. Peters et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1491–1493
Scheme 1. Reagents and conditions: (a) K2CO3, MeOH, 50 ꢀC, 3 h, then, after evaporation of solvent, KMnO4, acetone, rt, 3 h (yield for 4c-H
(R1=H; R2=2,4-Cl2): 52%); (b) LiAlH4, THF, 40 ꢀC, 3 h (yield for 1c-H(R 1=H; R2=2,4-Cl2): 40%).
Table 1. Structure–activity relationship of 6-phenyl substituents on
1b: DPP-IV—IC50 data (mM)
these and other (1c-m-Me, 1c-m-Cl, 1c-m-CF3, 1c-p-Me,
1c-p-Cl, 1c-p-OMe, 1c-p-CF3) meta- and para-
substituted derivatives.
R=H: 42
R=
Me
Cl
OMe
F
CF3
ortho
meta
para
1.5
20
1.0
2.5
31
1.4
1.5
80
47
14
40
18
14
170
1.1
X-ray crystal structure determination (Fig. 1)11,12
revealed that 1e binds to the active site of DPP-IV. The
2,4-dichlorophenyl motif of the inhibitor occupies very
well the hydrophobic S1 pocket of the enzyme explain-
ing the beneficial effect of ortho and para substituents in
this part of the molecule. The aminomethyl group of the
ligand forms a symmetrical hydrogen bonding network
with a tyrosine (Y662) and two glutamate (E205, E206)
residues of the protein. This interaction substitutes the
binding of the N-terminus of the substrates to this key
recognition motif of the active site. Other interactions
include an additional hydrogen bond from the aromatic
amino function of 1e to a backbone amide carbonyl
pronounced. In contrast, meta substituents decreased
activity (1b-m-OMe, 1b-m-CF3) or had little influence
(1b-m-Me, 1b-m-Cl, 1b-m-F).
The obvious assumption that a combination of ortho
and para substituents should further increase activity
led to the 2,4-dichloro substituted compound 1c-H
(R¼H, Table 2). The activity of 1c-Hsurpassed our
anticipations with a 1000-fold increase in activity as
compared to screening hit 1a! Taking 1c-Has a basis for
further improvement, we investigated the effects of sub-
stitution of the 2-phenyl ring. ortho Substitution gave
less active compounds (1c-o-Me, 1c-o-OMe, 1c-o-F,
Table 2). A single meta fluoro (1c-m-F) or a single para
fluoro (1c-p-F) substituent increased the activity into the
picomolar range. A combination of these substituents in
one molecule (1d) was less efficient. Single meta meth-
oxy substitution (1c-m-OMe) led to a drop of activity.
Surprisingly, the introduction of a second meta methoxy
substituent gave a compound with an outstanding inhi-
bitory activity, 1e. No clear-cut SAR emerged from
Table 2. Structure–activity relationship of 2-phenyl-substituents on
1c: DPP-IV—IC50 data (mM)
R=H: 0.01
R=
Me
Cl
OMe
F
CF3
ortho
meta
para
1.75
0.0009
0.090
—
0.24
0.053
0.35
0.34
0.10
0.047
0.0002
0.0002
—
0.13
0.18
Figure 1. X-ray structure by co-crystallization of 1e (bonds are
colored cyan) and DPP-IV (magenta). Dashed lines indicate hydrogen
bonds between protein and ligand.

J.-U. Peters et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1491–1493
1493
(E205), and a cation-p-interaction between an arginine
residue (R125) and the inhibitor’s pyrimidine core. The
3,5-dimethoxyphenyl residue points mainly into the sol-
vent beside some interaction of one methoxy substituent
with the side chain of a tyrosin (Y547). No interaction
with the catalytic serine (S630) is observed.
1324. (g) Mitani, H.; Takimoto, M.; Hughes, T. E.;
Kimura, M. Japanese Journal of Pharmacology 2002, 88,
442. (h) Mitani, H.; Takimoto, K. M. Japanese Journal of
Pharmacology 2002, 88, 451. (i) Sudre, B.; Broqua, P.;
Ashworth, D.; Evans, E. M.; Haigh, R.; Junien, J. L.;
Aubert, M. L. Diabetes 2002, 51, 1461.
4. (a) Holst, J. J.; Deacon, C. F. Diabetes 1998, 47, 1663. (b)
Zander, M.; Madsbad, S.; Madsen, J. L.; Holst, J. J.
Lancet 2002, 359, 824. (c) For a review, see Drucker, D. J.
Expert Opinion on Investigational Drugs 2003, 12, 87.
5. For a review, see: Augustyns, K.; Van der Veken, P.; Senten,
K.; Haerners, A. Expert Opin. Ther. Patents 2003, 13, 499.
6. Villhauer, E. B.; Brinkman, J. A.; Naderi, G. B.; Dun-
ning, B. E.; Mangold, B. L.; Mone, M. D.; Russell, M. E.;
Weldon, S. C.; Hughes, T. J. Med. Chem. 2002, 45, 2362.
7. Ahren, B.; Simonsson, E.; Larsson, H.; Landin-Olsson,
M.; Torgeirsson, H.; Jansson, P. A.; Sandqvist, M.;
Bavenholm, P.; Efendic, S.; Eriksson, J. W.; Dickinson,
S.; Holmes, D. Diabetes Care 2002, 25, 869.
In summary, a series of novel DPP-IV inhibitors 1 was
derived from screening hit 1a. We focussed on the deri-
vatization of the 2-phenyl ring and the 6-phenyl ring of
1. A 2,4-disubstitution pattern at the 6-phenyl ring
appears to be highly favored. This is exemplified by 1c-
Hwith a 1000-fold increase in activity as compared to
the original screening hit 1a. Picomolar compounds
could be obtained by the introduction of fluoro and
methoxy substituents into the meta and para positions
of the 2-phenyl ring. By optimizing aromatic sub-
stituents at both phenyl rings, a 100,000-fold increase in
activity was achieved (1e vs 1a).
8. Abd-Elfattah, A. M.; Hussain, S. M.; El-Reedy, A. M.;
Yousif, N. M. Tetrahedron 1983, 39, 3197.
9. For detailed procedures, see: Boehringer, M.; Loeffler, B.-
M.; Peters, J.-U.; Steger, M.; Weiss, P. International
Patent Application WO 03/68757, 2003.
References and notes
10. DPP-IV inhibitors were measured in triplicate at 5 to 7
concentrations in the range of 100 mM to 100 pM. IC50
values were calculated with a non-linear best fit regression
model. All assays were calibrated with NVP-DPP728
as internal standard inhibitor. NVP-DPP728 under the
conditions of the assay showed an IC50 of 15ꢁ4 nM
(MꢁSD, n=12) at 50 mM substrate concentration and a
Ki of 11ꢁ3 nM determined at substrate concentration
range of 10 mM to 600 mM. IC50 values of unkonwn
compounds were accepted when the IC50 (ꢂ) measured
for NVP-DPP728 in the assay was 11<ꢂ<19 nM.
11. Thoma, R.; Loffler, B.; Stihle, M.; Huber, W.; Ruf, A.;
Hennig, M. Structure 2003, 11, 947.
12. Data collection, processing and refinement statistics:
resolution range: 15.0–2.2 A; measured reflections:
216,345; unique reflections: 79,169; completeness: 84.3%;
Rsym: 8.7%; no. of protein atoms: 11 962; no. of waters
98; no. of ligand atoms: 54; r.m.s. distances: 0.010 A;
r.m.s. bond angles: 1.8; Rcryst: 23.8%; Rfree (5% of data):
28.9%; mean B-factor, protein: 43.4 A2; mean B-factor,
ligand: 43.6 A2; mean B-factor, waters: 40.4 A2. The
coordinates of this structure have been deposited in the
Protein Data Bank (accession code 1RWQ).
1. (a) Mentlein, R.; Gallwitz, B.; Schmidt, W. E. Eur. J.
Biochem. 1993, 214, 829. (b) Kieffer, T. J.; McIntosh,
C. H. S.; Pederson, R. A. Endocrinology 1995, 136, 3585.
(c) Lene, H.; Deacon, C. F.; Orskov, C.; Holst, J. J.
Endocrinology 1999, 140, 5356. (d) For a review, see:
Rosenblum, J. S.; Kozarich, J. W. Current Opinion in
Chemical Biology 2003, 7, 496.
2. (a) Holst, J. J. Gastroenterology 1994, 107, 1048. (b)
Orsakov, C. Diabetologia 1992, 35, 701. (c) Drucker, D. J.
Diabetes 1998, 47, 159.
3. (a) Ahren, B.; Holst, J. J.; Martensson, H.; Balkan, B.
Eur. J. Pharmacol. 2000, 404, 239. (b) Deacon, C. F.;
Hughes, T. E.; Holst, J. J. Diabetes 1998, 47, 764. (c)
Pederson, R. P.; White, H. A.; Schlenzig, D.; Pauly, R. P.;
McIntosh, C. R. P.; Demuth, H. U. Diabetes 1998, 47,
1253. (d) Pauly, R. P.; Rosche, R.; Schmidt, J.; White,
H. A.; Lynn, F.; McIntosh, C. H. S.; Pederson, R. A.
Metabolism 1999, 3, 385. (e) Pospisilik, J. A.; Stafford,
S. G.; Demuth, H. U.; Brownsey, R.; Parkhouse, W.;
Finegood, D. T.; McIntosh, C. H. S.; Pederson, R. A.
Diabetes 2002, 51, 943. (f) Balkan, B.; Kwasnik, L.;
Miserendino, R.; Holst, J. J.; Li, X. Diabetologia 1999, 42,