Substituted piperazines as novel dipeptidyl peptidase IV inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.06.065

The synthesis and SAR of a series of piperazines containing a fluorophenyl β-amino amide moiety is reported. Incorporation of a fluorophenyl β-amino amide moiety into piperazine screening lead 2 has resulted in the discovery of a structurally novel series

Full text of DOI:10.1016/j.bmcl.2004.06.065

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4763–4766
Substituted piperazines as novel dipeptidyl peptidase IV inhibitors
a
a
a
Linda L. Brockunier,a, Jiafang He, Lawrence F. Colwell, Jr., Bahanu Habulihaz,
*
Huaibing He,a Barbara Leiting,c Kathryn A. Lyons,a Frank Marsilio,c Reshma A. Patel,c
Yohannes Teffera,b Joseph K. Wu,c Nancy A. Thornberry,c Ann E. Webera
and Emma R. Parmeea
aDepartment of Medicinal Chemistry, Merck & Co. Inc., PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Drug Metabolism, Merck & Co. Inc., PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Metabolic Disorders, Merck & Co. Inc., PO Box 2000, Rahway, NJ 07065, USA
Received 25 May 2004; revised 17 June 2004; accepted 21 June 2004
Available online 20 July 2004
Abstract—Incorporation of a fluorophenyl b-amino amide moiety into piperazine screening lead 2 has resulted in the discovery of a
structurally novel series of potent and selective DP-IV inhibitors. Simplification of the molecule and incorporation of multiple flu-
orine atoms on the phenyl ring has provided low molecular weight analogs such as compound 32, which is a 19nMDP-IV inhibitor
with >4000-fold selectivity over QPP.
Ó 2004 Elsevier Ltd. All rights reserved.
The preceding paper describes the discovery of a
novel series of (R)-b-homophenylalanine-based dipept-
idyl peptidase IV (DP-IV) inhibitors exemplified by par-
ent compound 1 (Fig. 1).1 In the course of our
investigation of the SAR of a structurally distinct
screening lead 2, we were interested to learn the impact
of a similar substitution. Thus, we have found that a
marked boost in DP-IV inhibitory potency could be
achieved by the incorporation of a (R)-b-amino amide
moiety into the left-hand side of the molecule to give
compounds such as 3. Both enantiomers of the analo-
gous a-amino amide phenylalanine derivatives were
found to be >2000-fold less potent. Removal of the
carbonyl group in either the (R)-a- or (R)-b-amino
amides also resulted in a significant loss of intrinsic
activity2 (data not shown). Due to the clear preference
for b-amino amide substitution, efforts focused on
this series. Further refinement has revealed a number
of structural preferences and these will be discussed
herein.3
Ph
NH2
O
H2N
N
N
O
O
S
N
S
N
N
O
H
H
2 DP-IV IC50 11000 nM
1 DP-IV IC50 3000 nM
Ph
NH2
O
N
O
O
N
S
N
N
O
H
H
3 DP-IV IC50 134 nM
Figure 1. Initial leads.
*
Corresponding author. Tel.: +1-732-5946160; fax: +1-732-5949473; e-mail: linda_brockunier@merck.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.065

4764
L. L. Brockunier et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4763–4766
Ph
Ph
Ph
NH2
O
X
g,b
HN
N
O
O
S
O
HN
O
O
S
*
a-f
N
N
N
BOC
N
H
N
H
N
N
O
H
H
3, 9-13
4
5
Scheme 1. Synthesis of compounds 3, 9–13. Reagents: (a) benzyl chloroformate, Et3N, CH2Cl2; (b) TFA, CH2Cl2; (c) a-Cl-3-NO2-acetanilide, DIEA,
DMF; (d) RaNi, NH2NH2, MeOH, ClCH2CH2Cl; (e) MsCl, pyridine, CH2Cl2; (f) H2, Pd(OH)2, MeOH; (g) (R)- or (S)-ArCH2CH(NHBoc)CH2-
CO2H, EDC, DMAP, CH2Cl2.
Table 3. Effect of truncation or replacement of the piperazine core on
DP-IV inhibitory properties
Inhibitors were initially synthesized as mixtures of dia-
stereomers from compound 4 (Scheme 1). Protection
as the benzyl carbamate (Cbz) and removal of the tert-
butylcarbamate (Boc) group was followed by alkylation
with alpha-chloro-3-nitroacetanilide. Selective reduction
of the nitro group in the presence of the Cbz-protecting
group, sulfonamide formation, and removal of the ben-
zylcarbamate provided intermediate 5. Standard 1-[3-
Ph
NH2
O
N
F
X
R
Compd
X
R
DP-IV IC50, nMQPP IC 50, nM
(dimethylamino)propyl]-3-ethylcarbodiimide
(EDC)
18
19
20
21
22
23a
N
–CH2Ph
–CH2CONH2
405
114
15,000
66,000
mediated peptide coupling of 5 with appropriately sub-
stituted optically active Boc-protected b-amino acids1
followed by deprotection gave the final compounds4 3,
and 9–13 (structures defined in Table 1). The stereoselec-
tive synthesis of compounds 14–32 (structures defined in
Tables 2–5) from enantiomerically pure intermediates5 6
is shown in Scheme 2. Peptide coupling of 6 with appro-
priately substituted b-amino acids, and Boc deprotec-
N
N
–CH3
H
––
225
139
>100,000
>100,000
>100,000
>100,000
N
O
CH
108
1040
H
a Mixture of diastereomers at piperidine 2-position.
tion of the resultant intermediate
7
provided
compound 18. Alternatively, the benzyl group in 7 was
Table 4. Effect of C-2 substitution of truncated analogs on DP-IV
inhibitory properties
Table 1. Inhibitory properties of (R) and (S) b-amino amides
NH2
O
R
Ph
NH2
*
O
N
N
O
O
S
F
NH
X
N
N
N
O
H
H
Compd
R
DP-IV IC50, nMQPP IC 50, nM
21
24
25
26
27
28
29
30
Benzyl
H
139
3700
1600
922
>100,000
>100,000
>100,000
50,000
Compd
X
H
Amine
stereochemistry
DP-IV
IC50, nM
QPP
IC50, nM
Methyl
Isopropyl
3
R
R
R
S
S
S
134
44
7700
450
4-Fluorophenyl
4-Thiazolylmethyl
3-Pyridylmethyl
4-Fluorobenzyl
610
143
23,000
9
3,4-DiF
2-F
>100,000
>100,000
>100,000
10
11
12
13
34
13,000
19,000
15,000
26,000
155
166
H
65,000
36,000
19,000
3,4-DiF
2-F
Table 2. Effect of C-2 stereochemistry on DP-IV inhibitory properties
Ph
Table 5. Effect of incorporation of multiple fluorine atoms on DP-IV
inhibitory properties
NH2
O
*
N
O
O
S
X
F
N
N
N
R
Ph
O
H
H
NH2
O
N
Compd
X
Benzyl group
stereochemistry
DP-IV
IC50, nM
QPP
IC50, nM
F
NH
14
15
16
17
3,4-DiF
2-F
R
R
S
S
30
14
640
19,000
280
Compd
R
DP-IV IC50, nMQPP IC
50, nM
3,4-DiF
2-F
1100
690
31
32
H
F
51
19
65,000
78,000
12,000

L. L. Brockunier et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4763–4766
4765
BOC
d-f, b
g, b
h, b
b
BOC
14-17
19
NH
O
R
NH
O
R
R
N
N
a
c
HN
N
Ph
NH
N
Ph
20
X
X
7
6
8
b
21, 24-32
18
Scheme 2. Diastereoselective synthesis of compounds 14–32. Reagents: (a) (R)-ArCH2CH(NHBoc)CH2CO2H, EDC, HOBt, DIEA, DMF; (b) TFA,
CH2Cl2; (c) H2, Pd(OH)2, MeOH [for compound 28 ACE–Cl, ClCH2CH2Cl]; (d) a-Cl-3-NO2-acetanilide, DIEA, DMF; (e) H2, Pd(OH)2, MeOH; (f)
˚
MsCl, pyridine, CH2Cl2; (g) 2-chloroacetamide, DIEA, DMF; (h) NaBH(OAc)3, (CH2O)n, 4A mol. sieves, ClCH2CH2Cl.
removed by hydrogenolysis (or by treatment with 1-
chloroethyl chloroformate when R=4-thiazolylmethyl)
and the resultant intermediate 8 converted to com-
pounds 14–17 by elaboration of the right-hand side as
discussed in Scheme 1. Alkylation of 8 with 2-chloro-
acetamide and Boc removal provided compound 19.
Reductive amination of 8 with paraformaldehyde fol-
lowed by deprotection of the Boc group gave compound
20. Compounds 21 and 24–32 were prepared by Boc de-
protection of 8. The requisite precursor of morpholine
analog 22 was synthesized according to literature prece-
dent.6 Compound 23 was synthesized from racemic 2-
benzylpiperidine.
Having demonstrated the dramatic influence of amine
stereochemistry on DP-IV inhibitory potency, we then
wished to define the stereochemical preference of the
substituent at the 2-position of the piperazine ring.
The individual diastereomers were prepared as described
in Scheme 2 and the compounds evaluated in vitro (Ta-
ble 2). Once again there was a strong preference for one
isomer. For the 2-fluoro analog, the (R,R) diastereomer
15 was 50-fold more potent (DP-IV IC50 =14nM) than
the corresponding (R,S) isomer 17. Importantly, com-
pound 15 also had an excellent selectivity profile as it
was >1000-fold selective over QPP.
With potent and selective DP-IV inhibitors now in hand,
we wanted to evaluate the pharmacokinetic profiles of
compounds in this series. The initial results were not
encouraging: compound 15 had an unusually high clear-
ance (130mL/min/kg) and low (<1%) oral bioavailabi-
lity when dosed in rats.9 To address this issue we first
began simplifying the molecule by truncation of the
right-hand side at various points. Gratifyingly, the low
molecular weight unsubstituted piperazine 21 and
morpholine 22 retained good in vitro potency (DP-IV
IC50 =139 and 108nM, respectively) and excellent selec-
tivity (>700- and >900-fold, respectively) over QPP (Ta-
ble 3). Interestingly, a significant loss in potency (DP-IV
IC50 =1040nM) was observed with piperidine com-
pound 23.
All compounds synthesized were tested for DP-IV activ-
ity and selectivity over DP-IV homologs and other
proline specific enzymes with DP-IV-like activity.7,8
Without exception, the piperazine analogs described
herein are inactive (IC50Õs >100lM) at PEP (prolyl
endopeptidase), prolidase, and APP (aminopeptidase
P) and as such these data will not be presented.
However, some inhibition of QPP (quiescent cell pro-
line peptidase, also known as DPP-II or DPP7), which
has been implicated to play a role in immune function,
has been observed in this series. Thus inhibitory poten-
tial of test compounds at DP-IV and QPP will be
reported.
The work described in the previous paper focused on the
optimization of a b-homophenylalanine thiazolidide ser-
ies and we hoped to utilize this information in the design
of analogs in the piperazine series. A key modification
was the introduction of one or more fluorine atoms on
the phenyl ring. In particular, the 3,4-difluoro and 2-
fluoro-substituted analogs were quite promising, thus
the corresponding isomers were prepared in the pipera-
zine series. While the (R) stereochemistry was shown to
be essential in the thiazolidide series, we wanted to con-
firm this in the structurally unique piperazine series. As
shown in Table 1, the same general trend for increased
DP-IV inhibitory potency when going from the unsub-
stituted phenyl analog 3 to the 3,4-difluorophenyl ana-
log 9 (3-fold increase) and the 2-fluorophenyl analog
10 (4-fold increase) was also observed in the piperazine
series. Preference for the (R) stereochemistry of the
b-amino amide moiety is clearly maintained. Most inter-
estingly, there was quite a pronounced increase in selec-
tivity (>350-fold) over QPP seen with the 2-fluorophenyl
analog and, therefore, the remaining SAR studies con-
centrated primarily on this substitution pattern.
We next began a survey of replacements of the benzyl
group at the 2-position of the piperazine ring, which
would not have the potential metabolic liabilities of an
unsubstitued phenyl ring. As shown in Table 4, removal
of the benzyl group altogether, or replacement by an
alkyl group resulted in a significant loss of potency.
The 4-fluorophenyl compound 27, lacking a benzylic
methylene group, suffered a 4-fold loss in DP-IV inhib-
itory potency whereas all of the other benzylic aromatic
groups were of comparable potency (DP-IV IC50s of
143–166nM) and selectivity (>600-fold over QPP) to
that of the unsubstituted benzylic analog 21.
The preceding paper1 describes the dramatic effect on
potency when additional fluorine atoms are incorpo-
rated into the fluorophenyl b-amino amide moiety of a
thiazolidide series of DP-IV inhibitors. Thus, the 2,5-di-
fluorophenyl and 2,4,5-trifluorophenyl analogs are 3-
and 7-fold more potent, respectively, than the 2-fluoro-
phenyl analog. We were also interested in incorporating
these same modifications into the current piperazine

4766
L. L. Brockunier et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4763–4766
lead. When the corresponding 2,5-difluorophenyl analog
31 and 2,4,5-trifluorophenyl analog 32 were prepared in
the truncated series, similar increases in potency relative
to the corresponding 2-fluorophenyl analog 21 were
achieved. Excellent selectivity of >1000-fold over QPP
was maintained.
References and notes
1. Xu, J.; Ok, H. O.; Gonzalez, E. J.; Colwell, L. F., Jr.;
Habulihaz, B.; He, H.; Leiting, B.; Lyons, K. A.; Marsilio,
F.; Patel, R. A.; Wu, J. K.; Thornberry, N. A.; Weber, A.
E.; Parmee, E. R. Bioorg. Med. Chem. Lett. 2004, 14,
preceding paper. doi:10.1016/j.bmcl.2004.06.099.
2. The a- and b-amino amines were prepared by reductive
amination of piperazine intermediate 5 with the requisite
aldehydes. The aldehydes were available by lithium alumi-
num hydride reduction of the corresponding Weinreb amides.
3. This work was reported in part; Brockunier, L. L. 227th
National Meeting of the ACS, Anaheim, CA, March 2004.
Abstract 100.
To determine whether we had achieved any improve-
ment in oral bioavailability, a subset of potent and selec-
tive compounds were chosen from within these series for
pharmacokinetic analysis in rats.9 Unfortunately, com-
pounds 21, 22, 29, and 30 also showed uniformly poor
pharmacokinetics typified by an exceptionally high
clearance, short half-life (<2h), and low overall oral bio-
availability (<4%). With the hope of providing some
guidance for future synthetic efforts within this class of
compounds, we initiated in vitro metabolism studies,
choosing compound 21 as a representative example.
Incubation with rat liver microsomes resulted in exten-
sive metabolism of this compound, primarily by pipera-
zine ring oxidation.
4. Final compounds were characterized by 1H NMR, mass
spectrometry and HPLC analysis.
5. Requisite enatiomerically pure piperazines 6 were prepared
according to a literature procedure outlined for the
synthesis of (R) and (S)-2-methylpiperazine. Kiely, J. S.;
Priebe, S. R. Org. Prep. Proced. Int. 1990, 22, 761.
6. Shawe, T. T.; Koenig, G. J.; Ross, A. A. Synth. Commun.
1997, 27, 1777. Chromatographic separation provided the
individual diastereomers.
7. (a) All IC50 determinations for DP-IV and other proline
peptidases were carried out as described in: Leiting, B.;
Pryor, K. D.; Wu, J. K.; Marsilio, F.; Patel, R. A.; Craik,
C. S.; Ellman, J. A.; Cummings, R. T.; Thornberry, N. A.
Biochem. J. 2003, 371, 525. (b) With the exception of
compound 31, all multiple determinations of the DP-IV
IC50 values were within 1.6-fold of the reported average;
compound 31 was within 2.8-fold of the average. All QPP
values were within 1.3-fold of the reported average.
8. For references on proline specific peptidases see: (a)
Rosenblum, J. S.; Kozarich, J. W. Curr. Opin. Chem. Biol.
2003, 7, 1; (b) Cunningham, D. F.; OÕConnor, B. Biochim.
Biophys. Acta 1997, 1343, 160.
In summary, we have discovered a series of structurally
novel compounds, which are potent and selective
DP-IV inhibitors. For example, compound 15 has an
IC50 of 14nMat DP-IV with >1000-fold selectivity
over QPP. Significantly, truncation of the right-hand
side of the molecule provided low molecular weight
analog 32, which maintained good in vitro potency
(DP-IV IC50 =19nM) and selectivity (>4000-fold) over
QPP. Having ascertained the metabolic liabilities of
compounds in this series, we have modified the pipera-
zine ring of compound 32, resulting in major improve-
ments in the pharmacokinetic profile. These
modifications will be the subject of future
communications.
9. Male Sprague-Dawley rats were dosed intravenously at
1mg/kg and orally at 2mg/kg. Plasma drug levels were
determined by LC/MS/MS.