Synthesis and structure-activity relationship of N-acyl-Gly-, N-acyl-Sar- and N-blocked-boroPro inhibitors of FAP, DPP4, and POP

Article abstract of DOI:10.1016/j.bmcl.2006.11.072

The structure-activity relationship of various N-acyl-Gly-, N-acyl-Sar-, and N-blocked-boroPro derivatives against three prolyl peptidases was explored. Several N-acyl-Gly- and N-blocked-boroPro compounds showed low nanomolar inhibitory activity against fibroblast activation protein (FAP) and prolyl oligopeptidase (POP) and selectivity against dipeptidyl peptidase-4 (DPP4). N-Acyl-Sar-boroPro analogs retained selectivity against DPP4 and potent POP inhibitory activity but displayed decreased FAP inhibitory activity.

Full text of DOI:10.1016/j.bmcl.2006.11.072

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1438–1442
Synthesis and structure–activity relationship of N-acyl-Gly-,
N-acyl-Sar- and N-blocked-boroPro inhibitors
of FAP, DPP4, and POP
Thuy Tran,a Clifford Quan,a Conrad Yap Edosada,b Mark Mayeda,b
Christian Wiesmann,c Dan Sutherlina and Beni B. Wolfb,*
aDepartment of Medicinal Chemistry, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
bDepartment of Molecular Oncology, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
cDepartment of Protein Engineering, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
Received 18 October 2006; revised 17 November 2006; accepted 29 November 2006
Available online 1 December 2006
Abstract—The structure–activity relationship of various N-acyl-Gly-, N-acyl-Sar-, and N-blocked-boroPro derivatives against three
prolyl peptidases was explored. Several N-acyl-Gly- and N-blocked-boroPro compounds showed low nanomolar inhibitory activity
against fibroblast activation protein (FAP) and prolyl oligopeptidase (POP) and selectivity against dipeptidyl peptidase-4 (DPP4).
N-Acyl-Sar-boroPro analogs retained selectivity against DPP4 and potent POP inhibitory activity but displayed decreased FAP
inhibitory activity.
Ó 2006 Elsevier Ltd. All rights reserved.
The prolyl peptidases dipeptidyl peptidase-4 (DPP4),
prolyl oligopeptidase (POP), and fibroblast activation
protein (FAP) cleave bioactive peptides preferentially
after proline residues and represent promising therapeu-
tic targets for diabetes, cognitive disorders, and cancer,
respectively.1 Although these serine proteases share a
preference for proline at the P1 position of substrates,
they display distinct activities. DPP4 displays only
dipeptidyl peptidase (DPP) activity that removes
P2-Pro1-dipeptides from the N-terminus of substrates,
whereas, POP acts solely as a proline-specific endopepti-
dase (Fig. 1). FAP displays both activities; however,
FAP endopeptidase activity is limited to substrates con-
taining a Gly-Pro motif (Fig. 1).2,3
Figure 1. Substrates and boroPro-inhibitors of prolyl peptidases. Note
that FAP has both dipeptidyl peptidase activity and endopeptidase
activity, whereas, DPPs-4, -7, -8, and -9 display only dipeptidyl
peptidase activity and POP only endopeptidase substrates (left panel).
Val-boroPro (PT-100) and Ac-Gly-boroPro (8) are examples of
dipeptidyl peptidase and endopeptidase inhibitors, respectively.
The unique activity of prolyl peptidases has been
exploited for inhibitor development as outlined in Fig-
ure 1. For example, potent inhibition of DPP4 and
FAP has been achieved with aminoacyl-proline boronic
acids (boroPro)4 and inhibitors such as Val-boroPro
(PT-100) stimulate hematopoiesis5 and demonstrate
anti-tumor activity.6,7 Although aminoacyl-boroPros
with free N-termini are poor POP inhibitors,4 they
non-selectively inhibit several proline-specific DPPs
besides DPP4 and FAP such as DPP7, DPP8, and
DPP9. More recently, N-alkyl-Gly-boroPro inhibitors
have been developed and these compounds inhibit
Keywords: Prolyl peptidase; Dipeptidyl peptidase; BoroPro; FAP;
DPP4; POP.
*
Corresponding author. Tel.: +1 6504671954; fax: +1 6502256327;
e-mail: bbwolf@gene.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.11.072

T. Tran et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1438–1442
1439
DPP4, FAP, and DPP7; however, their reactivity with
POP is unknown.8
tert-butyl ether)–water mixture.9 Pinanediol phenylb-
orate was recovered from the organic phase and the
N-acyl-Gly or the N-acyl-Sar-boroPros 6 (Table 1) were
isolated from the aqueous phase by reverse phase
HPLC. Directly N-blocked-boroPros 7 (Table 2) were
prepared by acylation of 1 using the previously de-
scribed coupling conditions followed by removal of the
pinanediol.
Based on FAP’s preference for Gly-Pro-based endopep-
tidase substrates, we recently synthesized Ac-Gly-boro-
Pro (Fig. 1) and tested its reactivity against prolyl
peptidases.2 This compound preferentially inhibited
FAP versus other prolyl peptidases, showing marked
selectivity against DPP8 and DPP9, but only modest
selectivity against DPP4 and POP. To expand the struc-
ture–activity relationship (SAR) and further optimize
inhibitor selectivity for FAP, we created a novel series
Inhibition constants (Ki)10 for FAP, DPP4, and POP
were determined for a series of acyl-Gly-boroPros to
explore the SAR of the N-blocking group (Table 1).
Acyl-Gly-boroPros containing alkyl-R2 groups (com-
pounds 8–10) were nanomolar inhibitors of FAP and
POP, but less effective inhibitors of DPP4. With FAP,
8 was the most potent inhibitor, whereas, 9 was the most
potent POP inhibitor. Although a decrease in FAP
potency was observed for compound 10, selectivity
against DPP4 relative to FAP increased throughout
the series. Cyclopentyl and cyclohexyl analogs (11 and
12) were similar to compound 9, showing slight increases
in potency for FAP and POP, while increasing overall
selectivity against DPP4. Benzoyl-Gly-boroPros (13–
15) also inhibited FAP and POP preferentially, and
the di-chloro-aryl compounds (14 and 15) were among
the most potent FAP/POP inhibitors in this series.
Together, these data show that the acyl-blocking group
of
N-acyl-Gly-,
N-acyl-Sar
(sarcosine)-,
and
N-blocked-boroPros. We report here on their synthesis
and inhibitory activity against FAP, DPP4, and POP.
The amino boronic ester 1 (Scheme 1) was prepared as
previously described.4 Either Boc-glycine-OH or
Boc-sarcosine-OH was coupled to 1 in the presence
of 1-(3-(dimethylamino)propyl)-3-ethylcarbodiimide hy-
drochloride (EDC) to generate the fully protected dipep-
tides 3. The Boc group was removed with HCl to
produce the unprotected amines 4. Acyl or aryl acids
R2 were coupled using the same conditions as for the
Boc amino acids. Deprotection of the boronic ester
was then effected by transesterification of the pinanediol
with phenylboronic acid in a biphasic MTBE (methyl-
Scheme 1. Reagents (i) HOBt, EDC, CH2Cl2, DIPEA; (ii) HCl, dioxane; (iii) HOBt, EDC, CH2Cl2, DIPEA; (iv) H2O, MTBE, PhB(OH)2.

1440
T. Tran et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1438–1442
Table 1. FAP, DPP4, and POP inhibition data for compounds 8–19
Compound
R2
R1
H
Ki (nM)
DPP4
377a
FAP
23a
POP
211a
8
Me
9
10
11
12
13
14
15
16
17
18
H
51
751
20
4300
4.5
191
2.3
2.7
25
H
608,000
9080
H
H
14
7391
H
142
29
38,400
9500
H
2.2
4.4
53
H
12
6871
Me
Me
Me
Me
Me
161
265
191
146
4600
19,500
11,300
68
13
7.4
101
19
a Values from Ref. 2.
consistently confers significant selectivity against DPP4
and that the nature of this group can modulate the
potency of FAP and POP inhibition.
We next synthesized compounds to explore the impor-
tance of the terminal amide carbonyl and NH (Table
2). Compound 20, in which a mesyl-group replaces the
acetyl group of the parental compound (8), shows
decreased potency against FAP, DPP4, and POP. De-
creased potency was also observed when the P2 Gly of
the parental compound was replaced with D-Ala (21).
FAP and POP showed 12- to 15-fold increases in Ki val-
ues with 21 and DPP4 demonstrated an even greater
drop in potency with little inhibition observed at
500 lM inhibitor. Finally, compounds 22–24 were syn-
thesized to examine the contribution of the amide NH
to inhibition. Cyclic amide (22) and ketone (23) com-
pounds from this group showed the most significant
inhibitory activity against FAP and POP, indicating that
the amide NH is not required for these enzymes. By
Sarcosyl analogs (16–18) of selected acyl-Gly-boroPros
were studied to examine the inhibitory activity of dual
N-substitution (Table 1). With FAP, the sarcosyl
analogs showed decreased potency relative to the acyl-
Gly-boroPros; however, these analogs showed equiva-
lent or increased potency with POP. As with the
acyl-Gly-boroPros, the sarcosyl analogs were generally
poor DPP4 inhibitors. One exception was compound
19, which preferentially inhibited DPP4 relative to
FAP and POP. These data indicate that the hydropho-
bic nature of the previously mentioned inhibitors may
be important for FAP and POP selectivity.

T. Tran et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1438–1442
Table 2. FAP, DPP4, and POP inhibition data for compounds 20–24
1441
Compound
R
Ki (nM)
DPP4
FAP
POP
20
21
246
350
1430
1359
2705
17% (500 lM)a
22,700
22
23
7.5
94
2.8
1.7
451,000
24
2900
182,000
353
a Percent inhibition at the concentration indicated.
contrast, compound 24, which maintains the planarity
of compound 22, but is more compact and has a less
electrophilic oxygen, was a relatively poor inhibitor of
FAP and POP. Compounds 22–24 showed marked
selectivity against DPP4 as observed with the previous
series of compounds.
tor’s P1-pyrrolidine ring. As well, each protease contains
a conserved arginine (e.g., FAP R123) and tyrosine (e.g.,
FAP Y541) that may hydrogen bond with the carbonyl
oxygens of the inhibitor’s P2 residue and indole ring,
respectively. Therefore, structural differences distal to
these interactions must underlie the observed SAR. In
this context, POP Cys 255, Trp 595, and other hydro-
phobic residues provide a non-polar environment for
the inhibitor’s indole ring, whereas, the di-Glu repeat
at the base of the FAP and DPP4 binding pockets pro-
vides a relatively more polar environment. This poten-
tially explains the higher affinity binding observed with
POP and compound 22, but does not account for its
inhibition of FAP and selectivity against DPP4. FAP
inhibition may involve hydrogen bonding of Glu 203
To identify potential structural differences that underlie
the observed SAR, we modeled binding of compound 22
to FAP, DPP4, and POP. As shown in Figure 2, the ac-
tive site residues of FAP and DPP4 are nearly identical
and highly similar to POP, particularly with regard to
the position of the catalytic serine, which likely forms
a covalent adduct with the inhibitor’s boronic acid moi-
ety and the S1 subsite, which accommodates the inhibi-
Figure 2. Inhibitor binding models. Compound 22 was docked into the active site of FAP (pdb 1z68)11, DPP4 (pdb 2ajd),12 and POP (pdb 1e8m)13
using Pymol (http://www.pymol.org). Inhibitor carbon atoms are colored yellow, whilst those of FAP are cyan, DPP4 gray, and POP green. Oxygen
atoms are red, nitrogens blue, borons pink, and sulfurs yellow.

1442
T. Tran et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1438–1442
5. Jones, B.; Adams, S.; Miller, G. T.; Jesson, M. I.;
Watanabe, T.; Wallner, B. P. Blood 2003, 102, 1641.
6. Adams, S.; Miller, G. T.; Jesson, M. I.; Watanabe, T.;
Jones, B.; Wallner, B. P. Cancer Res. 2004, 64, 5471.
7. Cheng, J. D.; Valianou, M.; Canutescu, A. A.; Jaffe, E. K.;
Lee, H. O.; Wang, H.; Lai, J. H.; Bachovchin, W. W.;
Weiner, L. M. Mol. Cancer Ther. 2005, 4, 351.
8. Hu, Y.; Ma, L.; Wu, M.; Wong, M. S.; Li, B.; Corral, S.;
Yu, Z.; Nomanbhoy, T.; Alemayehu, S.; Fuller, S. R.;
Rosenblum, J. S.; Rozenkrants, N.; Minimo, L. C.; Ripka,
W. C.; Szardenings, A. K.; Kozarich, J. W.; Shreder, K. R.
Bioorg. Med. Chem. Lett. 2005, 15, 4239.
and Glu 204 to the amide nitrogen of the inhibitor’s in-
dole ring, which may also interact with Phe 350 and Phe
351. Although DPP4 Glu 205 and Glu 206 could simi-
larly bind the inhibitor, the polar nature of Arg 358
and Arg 669 may disfavor binding of the indole ring.
Additionally, an amino acid difference more removed
from the active site (eg. FAP Ala657 = DPP4 Asp663;
not shown) imparts significant differences in FAP and
DPP4 substrate specificity11 and thus may influence
inhibitor selectivity as well. Future mutagenesis and
crystallography studies will test these possibilities.
9. Coutts, S. J.; Adams, J.; Krolikowski, D. A.; Snow, R. J.
Tetrahedron Lett. 1994, 35.
In conclusion, various N-acyl-Gly-, N-acyl-Sar-, and
N-blocked-boroPros were synthesized and tested as
inhibitors of FAP, DPP4, and POP. Several of the N-ac-
yl-Gly- and N-blocked-boroPros showed low nanomo-
lar inhibitory activity against FAP and POP, and
marked selectivity against DPP4, suggesting that they
will be useful tools for the study of FAP and POP biol-
ogy. N-Acyl-Sar-boroPro analogs retained selectivity
against DPP4 and potent POP inhibitory activity but
displayed decreased FAP inhibitory activity. The results
presented here provide a framework for future studies
aimed at developing selective inhibitors for FAP and
POP.
10. Recombinant FAP, DPP4, and POP were expressed and
purified as described previously (Ref. 2). Inhibitor stocks
were made in DMSO and assayed at various concentra-
tions up to 0.5 mM. Final DMSO levels in reactions were
less than 1% v/v to avoid inhibition of proteases by excess
DMSO. Ki values for protease inhibition were determined
by the method of progress curves (Ref. 2). Briefly,
proteases were added to a mixture of inhibitor and
substrate (Gly-Pro-AFC for FAP and DPP4; zGP-AMC
for POP) in 50 mM Tris–HCl, 100 mM NaCl, pH 7.4, at
23 °C. Dithiothreitol was included in POP reactions at
final concentration of 1 mM. Protease activity was mon-
itored with time by following release of AFC (FAP,
DPP4) or AMC (POP) fluorophores. Data were plotted as
v0/vi À 1 versus [I], where v0 is the rate of substrate
hydrolysis in the absence of inhibitor, vi is the steady-state
rate of substrate hydrolysis in the presence of inhibitor,
and [I] is the concentration of inhibitor. Plots of v0/vi À 1
versus [I] were linear and the apparent inhibition constant
(Kapp) was determined from the reciprocal of the slope.
The true equilibrium inhibition constant (Ki) was calcu-
lated from the relationship: Ki = Kapp/(1 + [S]/Km), where
[S] is the concentration of substrate used in the assay and
Km is the Michaelis constant for substrate hydrolysis. The
presented values represent the average of at least three
independent experiments with standard errors less than
10%.
Acknowledgment
We thank Janie Pena for graphics help.
References and notes
1. Rosenblum, J. S.; Kozarich, J. W. Curr. Opin. Chem. Biol.
2003, 7, 496.
2. Edosada, C. Y.; Quan, C.; Wiesmann, C.; Tran, T.;
Sutherlin, D.; Reynolds, M.; Elliott, J. M.; Raab, H.;
Fairbrother, W.; Wolf, B. B. J. Biol. Chem. 2006, 281, 7437.
3. Edosada, C. Y.; Quan, C.; Tran, T.; Pham, V.; Wiesmann,
C.; Fairbrother, W.; Wolf, B. B. FEBS Lett. 2006, 580, 1581.
4. Coutts, S. J.; Kelly, T. A.; Snow, R. J.; Kennedy, C. A.;
Barton, R. W.; Adams, J.; Krolikowski, D. A.; Freeman,
D. M.; Campbell, S. J.; Ksiazek, J. F.; Bachovchin, W. W.
J. Med. Chem. 1996, 39, 2087.
11. Aertgeerts, K.; Levin, A.; Shi, L.; Snell, G. P.; Jennings,
A.; Prasad, G. S.; Zhang, Y.; Kraus, M. L.; Salakian, S.;
Sridhar, V.; Wijnands, R.; Tennant, M. G. J. Biol. Chem.
2005, 280, 19441.
12. Engel, M.; Hoffmann, T.; Manhart, S.; Heiser, U.;
Chambre, S.; Huber, R.; Demuth, H.; Bode, W. J. Mol.
Biol. 2006, 355, 768.
13. Fulop, V.; Bocskei, Z.; Polgar, L. Cell 1998, 94, 161.