Synthesis of (E)- and (Z)-fluoro-olefin analogues of potent dipeptidyl peptidase IV inhibitors

Article abstract of DOI:10.1016/S0040-4039(03)01542-9

(E)- and (Z)-fluoro-olefin analogues of potent dipeptidyl peptidase IV inhibitors were synthesized. A Wadsworth-Horner-Emmons reaction, followed by amide formation and reduction of the amide were used for the construction of the α-fluoro-α,β-unsaturated a

Full text of DOI:10.1016/S0040-4039(03)01542-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 6231–6234
Synthesis of (E)- and (Z)-fluoro-olefin analogues of potent
dipeptidyl peptidase IV inhibitors
Pieter Van der Veken, Istva´n Kerte`sz, Kristel Senten, Achiel Haemers and Koen Augustyns*
Department of Medicinal Chemistry, University of Antwerp, Universiteitsplein 1, B-2610 Antwerp, Belgium
Received 15 May 2003; accepted 20 June 2003
Abstract—(E)- and (Z)-fluoro-olefin analogues of potent dipeptidyl peptidase IV inhibitors were synthesized. A Wadsworth–
Horner–Emmons reaction, followed by amide formation and reduction of the amide were used for the construction of the
a-fluoro-a,b-unsaturated amine functionality.
© 2003 Elsevier Ltd. All rights reserved.
1. Introduction
amino terminus, a process analogous to the well-known
diketopiperazine formation in peptide chemistry. This
problem can possibly be overcome by changing the
amide bond in the parent structure by a more rigid
function that stabilizes the ‘trans’ conformation of the
inhibitor. This ‘trans’ conformation is presumed neces-
sary for enzyme binding.2 As early as 1986, the fluoro-
olefin group was proposed as a superior rigid isosteric
and isoelectronic replacement for the amide bond, an
assumption also supported by more recent theoretical
studies.3 Furthermore, the introduction of a fluoro-
olefin function might have pronounced effects on the
pharmacokinetic characteristics of these compounds
compared to their parent amide compounds. This con-
cept was used by the group of Welch in synthesizing
fluoro-olefin analogues of two known inhibitors with
alanyl-c[C(F)ꢀC)]-proline structure (Fig. 1). Although
no general and unambiguous conclusions considering
the effect on inhibitory activity of introducing the
fluoro-olefin group in these compounds can be drawn
from only two examples, Welch could experimentally
verify that there is indeed significant enzyme affinity for
these peptidomimetics.4
Dipeptidyl peptidase IV (DPP IV, EC 3.4.14.5., CD26)
is a serine peptidase cleaving off dipeptides from the
amino terminus of peptides or proteins with proline or
alanine at the penultimate position. Since prolylamides
are known to play a critical role in peptide structure
and function, and because of their high resistance
towards non-specific enzymatic hydrolysis, the few
enzymes capable of cleaving this structural motif have
always attracted considerable scientific attention. As a
result of this research, it was shown that DPP IV-inhi-
bition can be used as a new tool for controlling type II
diabetes.1 Typically, these inhibitors possess a dipeptide
skeleton with a free amino terminus and a pyrrolidine
ring attached to an electrophilic site for Ser-OH scav-
enging such as a carbonitrile or a boronic acid (Fig. 1).
The stability of these compounds is compromised due
to reaction of the electrophilic group with the free
Our research on fluoro-olefin analogues of prolyl-
derived DPP IV inhibitors was focused on N-substi-
tuted glycylprolylanalogues (Fig. 2).
Members of this group of compounds have been
selected for different clinical trials for the management
of type II diabetes.5 Since the (E)-isomers of the exist-
ing fluoro-olefin inhibitors have never been synthesized,
our compounds offer the opportunity to rigorously
examine the cis,trans selectivity of DPP IV. The N-sub-
stituent groups in our compounds (2-phenylethyl-, 1-
Figure 1. Typical structure for DPP IV inhibitors (left) and
existing (Z)-fluoro-olefin mimetics (right).
* Corresponding author. Tel.: +32-(0)3-8202703; fax: +32-(0)3-
8202739; e-mail: koen.augustyns@ua.ac.be
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01542-9

6232
P. Van der Veken et al. / Tetrahedron Letters 44 (2003) 6231–6234
largely undocumented.7 The reaction was optimized
with cyclopentanone and the successful outcome of the
reaction seems to be very much confined to specific
reaction conditions and the use of NaH as base (Table
1).
The yield of the reaction (74%) with compound 1 and
the E/Z isomeric ratio (1.3) were comparable to the
results reported for this reaction when using the Peter-
son reagent tert-butyl a-fluoro-a-trimethylsilylacetate.
Double bond isomers were separated by column chro-
matography; the structural identity of these isomers
was determined by NMR spectroscopy.8
Figure 2. Newer DPP IV inhibitors (left) and synthesized
fluoro-olefin mimetics (right).
adamantyl-, and 4-fluorobenzyl-) were chosen accord-
ing to the activity data for the parent structures
obtained from patent literature.6
The fluoro-olefin products 2 were then subjected to
alkaline hydrolysis in almost quantitative yield and
coupled to the appropriate amine using the standard
peptide coupling reagent TBTU. For the subsequent
reduction, one literature precedent was present of the
successful LiAlH4 mediated transformation of an a-
fluoro-a,b-unsaturated amide into the corresponding
amine in ethereal solution.9 As a test, these conditions
were applied to compound 6, obtained in a similar way
starting from cyclopentanone (Scheme 2). After react-
ing for four days, the reaction mixture which still
contained a considerable amount of starting material
was quenched. Using an extractive work-up, amide and
amine components were separated and the latter were
Boc-protected to allow chromatographic separation.
However, it was found that the yield of the desired
product (3%) was unacceptably low and that the very
slow reaction was accompanied by extensive defluorina-
tion (Table 2, first entry).
2. Results and discussion
Similar to Welch, racemic TBDMS-protected (2-
hydroxymethyl)cyclopentanone was chosen as the start-
ing material for the construction of the pseudodipeptide
skeleton (Scheme 1).
Contrary to Welch’s Peterson olefination methodology
for the introduction of the fluoro-olefin moiety, we
decided to use the much easier accessible Wadsworth–
Horner–Emmons reagent ethyl (diethoxyphospho-
ryl)(fluoro)acetate, obtained in high yield from the
Arbuzov reaction between triethylphosphite and ethyl
bromofluoroacetate. Although this Wadsworth–
Horner–Emmons reagent has been used numerous
times with aldehydes, its reactivity towards ketones is
Contrary to our results, Bartlett and Otake only men-
tion the defluorinated analogue of their target amine in
3–6% yield.
In order to find reaction conditions that suppress this
side reaction, different aluminium hydride-based reduc-
tants were tested on model compound 6.
Table 1. Optimizing the Wadswoth–Horner–Emmons reac-
tion with cyclopentanone
Base
Conditions
Yield (%)
KOt-Bu
LDA
n-BuLi
KHMDS
NaH
DMF/0°C
THF/−78°C
THF/−78°C
THF/−78°C
THF/0°C
12
<5
0
0
70
84
NaH
Et2O/0°C
Scheme 1. Construction of the b-fluoro-b,g-unsaturated
amine group. Synthetic steps for the (E)-isomers are identical.
Reagents and conditions: (a) 1. NaH, Et2O, 0°C, 2. ethyl(di-
ethoxyphosphoryl)(fluoro)acetate, 4 h, 74% (E/Z=1.3); (b)
KOH, MeOH/H2O, 8 h, 98%/97%;a (c) TBTU, DIPEA,
RNH2, CH2Cl2, 4 h, 85–87%/84–91%;b (d) 1. POCl3 (4
equiv.), DIPEA (1 equiv.), CH2Cl2, 0°C, 2 h, 2. LiAlH4 (2
equiv.), Et2O, 0°C.c
aYields separated by ‘/’ are for E- and Z-isomers, respec-
tively.
bYields for E- and Z-isomers fall within the indicated range.
cThe crude reaction products were used in the next steps
without purification.
Scheme 2. Reagents and conditions: (a) LiAlH4, Et2O.

P. Van der Veken et al. / Tetrahedron Letters 44 (2003) 6231–6234
6233
As seen from Table 2, POCl3-preactivation which turns
the amide bond in a more reactive chloro-imidate,
followed by LiAlH4 reduction is a superior method and
was adopted in the synthesis of our target fluoro-olefin
peptidomimetics. To prevent acidolytic cleavage of the
TBDMS group by HCl produced during the reaction
with POCl3, 2 equiv. of DIPEA were added to the
reaction mixture. After reduction, the crude amines 5
were Boc-protected, followed by selective acidic hydroly-
sis of the TBDMS group and chromatographic purifica-
tion of the resulting alcohols (Scheme 3).
The Jones oxidation protocol was then used for the
synthesis of the acids 12, followed by activation with
iso-butylchloroformate and transformation into the
corresponding primary amides 13. Dehydration in high
yield gave the protected carbonitriles 14. Contrary to
the very recently published synthesis of a carbonitrile
fluoroolefin inhibitor by Welch’s group (Fig. 1), the use
of the standard POCl3/imidazole dehydrating agent did
not result in low yields for this step for any of the
products.
Scheme 3. Synthesis of target compounds. The synthetic
methods for the (E)-isomers are identical. (a) (Boc)2O,
DIPEA, CH2Cl2, 6 h;a (b) AcOH/H2O/THF (3:1:1), 12 h,
21–69%/41–72%;b (c) Jones oxidation, 1 h, 78–94%;c (d) 1.
i-ButOC(O)Cl, DIPEA, CH2Cl2, 0°C, 15 min, (2) NH4OH (10
equiv.), 67–91%;c (e) POCl3/imidazole, pyridine, −10°C, 30
min, 85–93%;c (f) TFA/CH2Cl2 (1:1), 30 min, 88–94%.c
aThe crude reaction products were used in he next steps
without purification.
The same holds true for the acidolytic deprotection of
the Boc-group which was done using TFA and pro-
vided final products 15. Although a cyclization reac-
tion, similar to the inactivation reaction described for
amide inhibitors might be expected for the deprotected
(E)-isomers, these compounds were stable for at least
96 h in aqueous or methanolic solution (controlled by
HPLC).
bThis is the combined yield of the reduction, Boc-protection,
and TBDMS-removal steps. Low yields were found for com-
pounds with R=1-adamantyl, probably due to ineffective
Boc-protection caused by steric factors.
3. Conclusion
cYields for E- and Z-isomers fall within the indicated range.
We have shown that fluoro-olefin analogues of glycyl-
prolylcarbonitriles can be prepared from cyclopen-
tanones with a Wadsworth–Horner–Emmons reaction.
Yields and E/Z isomeric ratio are comparable with the
Peterson olefination method. A generally applicable
method for the clean transformation of both (E)- and
(Z)-a-fluoro-a,b-unsaturated amides into the corre-
sponding amines was developed. The synthesized com-
pounds will be of importance as DPP IV inhibitors and
for the study of the selectivity of DPP IV for a ‘trans’
P2ꢁP1 bond.
Table 2. Comparison of reduction methods for model a-
fluoro-a,b-unsaturated amide 6
Acknowledgements
Reducing agent
Conditions
Yield (%)
8
9
P. Van der Veken is a Fellow of the Institute for the
Promotion of Innovation by Science and Technology in
Flanders. I. Kerte`sz thanks the European Community
for his appointment as a Marie Curie Fellow.
a
LiAlH4
Et2O/0°C to rt/96 h
Et2O/0°C to rt/6 h
a. Et2O/0°C to rt/4 h
b. THF/0°C to rt/4 h
Et2O/0°C to rt/8 h
a. 1. DCM/rt/1 h
3
6
LiAlH4/AlCl3 (3:1)b
LiAlH4/ZnCl2 (2:1)c
65
74
72
51
22
6
6
AlH3·Et3N
1. POCl3/2. LiAlH4
19
d
References
2. Et2O/0°C to rt/3 h
b. 1. DCM/rt/1 h
2. THF/0°C to rt/3 h
92
79
3
4
1. Augustyns, K.; Bal, G.; Thonus, G.; Belyaev, A.; Zhang,
X. M.; Bollaert, W.; Lambeir, A. M.; Durinx, C.;
Goossens, F.; Haemers, A. Curr. Med. Chem. 1999, 6,
311–327.
a Reaction was stopped before completion, remainder 22% of 6, 61%
of 7.
2. Fischer, G.; Heins, J.; Barth, A. Biochim. Biophys. Acta
b In situ formation of alane: 3LiAlH4+AlCl3“4AlH3+3LiCl.
c In situ formation of alane and zincane: 2LiAlH4+ZnCl2“2AlH3+
ZnH2+2LiCl.
1983, 742, 452–462.
3. (a) Abraham, M. J.; Ellison, S. L. R.; Schonholzer, P.;
Thomas, W. A. Tetrahedron 1986, 42, 2101–2110; (b)
d POCl3 reacts with amides, forming a reactive chloro-imidate.

6234
P. Van der Veken et al. / Tetrahedron Letters 44 (2003) 6231–6234
Biomedical Frontiers of Fluorine Chemistry; Cieplak, P.;
Scientific Sessions; 62nd Annual Meeting and Scientific
Sessions, San Francisco, 6/14-6/18; American Diabetes
Association: Washington, DC, 2002; 272-OR, A67.
6. Villhauer, E. B. US Patent 6,124,305, 2000.
Kollman, P. A.; Radomski, J. P. ACS Symp. Ser. 639;
American Chemical Society: Washington, DC, 1996, 143–
156.
7. Burton, D. J.; Yang, Z. Y.; Qiu, W. M. Chem. Rev. 1996,
96, 1641–1715.
4. (a) Welch, J. T.; Lin, J. Tetrahedron 1996, 52, 291–304; (b)
Lin, J.; Toscano, P. J.; Welch, J. T. Proc. Natl. Acad. Sci.
USA 1998, 95, 14020–14024; (c) Zhao, K.; Lim, D. S.;
Funaki, T.; Welch, J. T. Bioorg. Med. Chem. 2003, 11,
207–215; (d) Biomedical Frontiers of Fluorine Chemistry;
Welch, J. T.; Lin, J.; Boros, L. G.; DeCorte, B.; Bergmann,
K.; Gimi, R. ACS Symp. Ser. 639; American Chemical
Society: Washinton, DC, 1996, 129–142.
5. (a) Villhauer, E. B.; Brinkman, J. A.; Naderi, G. B.;
Dunning, B. E.; Mangold, B. L.; Mone, M. D.; Russel, M.
E.; Weldon, S. C.; Hughes, T. E. J. Med. Chem. 2002, 45,
2362–2365; (b) Hughes, T. E.; Russel, M. E.; Bolognese, L.;
Li, X.; Burkey, B. F.; Wang, P. R.; Villhauer, E. B.
NVP-LAF237, A Highly Selective and Long-Acting Dipep-
tidyl Peptidase IV Inhibitor, Abstracts from the ADA 62nd
8. 1H–1H NOESY experiments (400 MHz) on compounds 5
were indicative in assigning fluoro-olefin geometries. Char-
acteristic NOEs were seen between the NH-proton and
cyclopentyl-protons. These results were confirmed by com-
paring our 1H and 13C NMR spectra for the aldehyde
derived of compound 1 with data provided by Welch in Ref.
4a. The aldehyde was synthesized by coupling acid 3 to
N-methoxy-N-methylamide using TBTU, followed by
LiAlH4-reduction of the resulting Weinreb-amide. During
the reduction, no defluorination occurred. E- and Z-isomers
were separated by silica gel flash chromatography, using
hexanes/ethylacetate 49:1 as the eluent.
9. Bartlett, P. A.; Otake, A. J. Org. Chem. 1995, 60, 3107–3111.