Structural optimization, biological evaluation, and application of peptidomimetic prostate specific antigen inhibitors

Article abstract of DOI:10.1021/jm301718c

Prostate-specific antigen (PSA) is a serine protease produced at high levels by normal and malignant prostate epithelial cells that is used extensively as a biomarker in the clinical management of prostate cancer. To better understand PSA's role in prostate cancer progression, we prepared a library of peptidyl boronic acid-based inhibitors. To enhance selectivity for PSA vs other serine proteases, we modified the P1 site of the inhibitors to incorporate a bromopropylglycine group. This allowed the inhibitors to participate in halogen bond formation with the serine found at the bottom of the specificity pocket. The best of these Ahx-FSQn(boro)Bpg had PSA K_i of 72 nM and chymotrypsin K_i of 580 nM. In vivo studies using PSA-producing xenografts demonstrated that candidate inhibitors had minimal effect on growth but significantly altered serum levels of PSA. Biodistribution of ¹²⁵I labeled peptides showed low levels of uptake into tumors compared to other normal tissues.

Full text of DOI:10.1021/jm301718c

Article
pubs.acs.org/jmc
Structural Optimization, Biological Evaluation, and Application of
Peptidomimetic Prostate Specific Antigen Inhibitors
Maya B. Kostova,† D. Marc Rosen,† Ying Chen,‡ Ronnie C. Mease,‡ and Samuel R. Denmeade*,§,∥
‡
§
∥
†Department of Oncology, Radiology and Radiological Science, Pharmacology and Molecular Science, The Sidney Kimmel
Comprehensive Cancer Center, The Johns Hopkins University, School of Medicine, Baltimore, Maryland 21231, United States
ABSTRACT: Prostate-specific antigen (PSA) is a serine
protease produced at high levels by normal and malignant
prostate epithelial cells that is used extensively as a biomarker in
the clinical management of prostate cancer. To better
understand PSA’s role in prostate cancer progression, we
prepared a library of peptidyl boronic acid-based inhibitors. To
enhance selectivity for PSA vs other serine proteases, we
modified the P1 site of the inhibitors to incorporate a
bromopropylglycine group. This allowed the inhibitors to participate in halogen bond formation with the serine found at the
bottom of the specificity pocket. The best of these Ahx-FSQn(boro)Bpg had PSA Ki of 72 nM and chymotrypsin Ki of 580 nM.
In vivo studies using PSA-producing xenografts demonstrated that candidate inhibitors had minimal effect on growth but
significantly altered serum levels of PSA. Biodistribution of 125I labeled peptides showed low levels of uptake into tumors
compared to other normal tissues.
In particular, PSA demonstrates preference for glutamine in the
P1 position that is a unique characteristic compared to
chymotrypsin,12 which has no ability to cleave after glutamine.
Chymotrypsin is also not inhibited appreciably by peptide
aldehyde inhibitors containing glutamine or glutamine
analogues in the P1 position.11
Previously, using docking studies based on crystal structure
data, we documented important differences in the physical and
electrostatic characteristics of the S1 pocket of these two serine
proteases that explain this differential preference for P1
glutamine on the part of PSA.11−14 In the case of PSA, the
presence of polar chains provides a noticeable polar character
to its shallow bottomed pocket. In contrast, the S1 cavity of
chymotrypsin is less polar and more spacious. A comparative
analysis of the binding modes of the peptide aldehyde inhibitor
SSKLQ-CHO docked in the S1 pocket of PSA and
chymotrypsin respectively revealed drastic differences in
orientation of the glutamine side chain that underlies the
marked difference in Ki values for this inhibitor for these two
INTRODUCTION
■
A hallmark of prostate cancer is the production of the prostate
tissue differentiation protein prostate-specific antigen (PSA).1,2
PSA is aptly named in that its expression is highly restricted to
normal and malignant prostate cells in men.3 Although
controversy exists over the use of PSA as a screening tool for
prostate cancer, it continues to be a useful biomarker for the
detection of cancer recurrence after localized therapy and as a
marker of therapeutic response in the metastatic setting.4 PSA
is found in high μg/mL to mg/mL amounts in the extracellular
fluid where it exists in an enzymatically active state.5 Upon
entering the circulation, PSA is rapidly inhibited by the protease
inhibitors α2-macroglobulin and α1-antichymotrypsin that are
present in molar excess in the serum.6,7 This functional and
quantitative difference in PSA compartmentalization has
resulted in strategies to target PSA’s enzymatic activity for
prodrug/protoxin activity8,9 and to the design of antibodies
directed at PSA’s catalytic site for potential use as imaging
agents.10
Because of the presence of the characteristic His-Asp-Ser
catalytic triad, PSA is classified as a serine protease. For all
serine proteases, substrate and inhibitor recognition is mainly
governed by the binding of the P1 residue to the S1 pocket of
the enzyme. PSA (i.e., KLK3) is also one of 15 members of the
kallikrein family of proteases.3 Most of the other kallikreins
have trypsin-like proteolytic activity due to the presence of an
aspartic acid residue at the bottom of the S1 specificity
pocket.3,4 In contrast, PSA is considered a chymotrypsin-like
protease due to the presence of a serine at the bottom of the S1
pocket (i.e., Ser189).11 However, while PSA has some similarity
to chymotrypsin in its preference for hydrophobic amino acids
at the P1 position, PSA also displays enzymatic properties that
differentiate it from chymotrypsin and other serine proteases.
proteases (PSA Ki = 45
4 μM vs chymotrypsin Ki <1000
μM).11
These modeling studies suggested that a P1 glutamine-based
PSA inhibitor would be the most selective for PSA vs other
chymotrypsin-like proteases. Previously, we generated boronic
acid analogues of leucine and norleucine and determined that
the substitution of boronic acid for aldehyde as the inhibitory
group greatly decreased the Ki for PSA inhibition. Boronic acid
was selected as the inhibitory moiety based on selectivity for
serine protease inhibition and due to the fact that a boronic
acid-based inhibitor bortezomib has achieved FDA approval as
Received: November 21, 2012
Published: May 21, 2013
© 2013 American Chemical Society
4224
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
cancer therapy. However, while boronic acid analogues of these
P1 preferred amino acids can be generated, the synthesis of a
boronic acid analogue of glutamine is not feasible.14 Previous
studies demonstrated that an aldehyde analogue of methionine
in the P1 position produced an inhibitor with a Ki of 3.8 μM for
PSA inhibition but a Ki of >100 μM for chymotrypsin.11 On the
basis of this previous data, in this study we describe our efforts
to produce more potent and potentially more selective PSA
inhibitors by generating a series of peptidyl-based inhibitors
containing a boronic acid analogue of 3-bromopropylglycine
(Bpg) in the P1 position. Bpg resembles Met and Gln as they
have very similar side chain lengths. Like Met, the side chain of
Bpg terminates in a bulky electron-rich atom that can form
hydrogen bonds. We compared the ability of Bpg inhibitors vs
inhibitors containing a boronic acid analogue of phenylalanine
in the P1 position to inhibit both PSA and chymotrypsin.
Selective PSA inhibitors were further evaluated in vitro for
ability to effect the growth of PSA-producing vs nonproducing
prostate cancer cell lines. Further optimization of selected
inhibitors was achieved through modification of the P5 position
of the peptide to enhance circulating half-life. These inhibitors
were labeled with 125I to assess biodistribution in vivo and to
evaluate their potential for imaging of prostate cancers. Finally,
we assessed the effect of the lead inhibitor on the in vivo
growth and PSA production of human prostate cancer
xenografts.
inhibitor sequences synthesized for these studies. All final
products were purified by HPLC and characterized by MS
(Table 1). The first set of peptides have their N-terminus
capped by Cbz and were prepared to determine whether
peptides that included boronic acid analogue of Phe 12 or Bpg
13 maintained the ability to inhibit PSA’s enzymatic activity
(Table 2). In the subsequent peptide inhibitors, aminohexanoic
acid (Ahx) was incorporated at the N-terminus of the peptide
to serve as a linker whereby other groups such as chelating
groups or radiohalogenated prosthetic groups could be
attached.
The binding efficiency of protease substrates and inhibitors is
highly dependent on the amino acid residues binding to the
corresponding pockets within the vicinity of the catalytic site as
illustrated in Scheme 3. The binding specificity is highly
dependent on the amino acid residue present at the P1
position. The structural models of the PSA active site suggest a
hydrophobic character for the sides of the specificity pocket and
a polar character for the bottom of the pocket,13 where the
polar hydroxyl of Ser189 is flanked by the polar hydroxyl side
chains of Ser226 and Thr190.11 PSA modeling studies
suggested that amino acids with medium length hydrophobic
side chains such as leucine, norleucine, or phenylalanine or
those with hydrophobic side chains terminating in a polar
group such as glutamine, methionine, and tyrosine would be
preferred by PSA in the P1 position. However, while aldehyde-
based peptide inhibitors with hydrophobic aliphatic side chains
(Leu, Nle) or aromatic side chains (Phe, Tyr) in the P1 amino
acids could inhibit PSA, these compounds were even more
potent chymotrypsin inibitors. In contrast, aldehyde inhibitors
containing glutamine or methionine in the P1 position were
highly specific for PSA and were not effective at inhibiting
chymotrypsin.
RESULTS
■
For the synthesis of the (boro)phenylalanine containing
peptidomimetics, the building block (R)-borophenylalanine-
(1S,2S,3R,5S)-(+)-2,3-pinanediol ester 5 was synthesized
(Scheme 1) using modified literature precedence originally
Thus, to produce a more potent and specific peptidyl PSA
inhibitor, we synthesized and evaluated (boro)Bpg-based
inhibitors. The bromopropylglycine boronic acid was selected
because it possesses a medium length hydrophobic chain such
as that found in residues like Leu, Nle, Met, and Gln, but the
presence of the bromine also allows for interactions with the
polar groups present within of the bottom of the S1 pocket of
PSA (Scheme 3). A similar strategy has been used to prepare
thrombin17 and proteasome inhibitors.18−21 The presence of
bromine at the end of the side chain of P1 allows for halogen
bond formation (electrostatic interaction between a halogen
and an oxygen), which provides a unique base for inhibitor−
enzyme recognition. Also, halogen−oxygen interactions can
play significant role in intermolecular binding.22 An example in
drug design and halogen bonding is the binding of inhibitor
IDD 594 (5-fluoro-2-(4-bromo-2-fluoro-benzylthiocarbamoyl)-
phenoxyacetic acid) to human aldose reductase.23 In our case of
protease−inhibitor interactions, the polarized bromine can
interact with the carbonyl from the peptide backbone and/or
hydroxyl groups of the Thr and the two Ser residues located at
the bottom of the S1 pocket.
In addition, we synthesized (boro)Phe-based inhibitors to
compare them with the (boro)Bpg-based inhibitors with
respect to the potency and the specificity. The most important
differences in the binding modes of (boro)Bpg and (boro)Phe
inhibitors lie in their interaction with the S1 specificity pocket.
The bromopropyl group is longer and reaches deeper into the
S1 pocket compared to phenylalanine. The additional halogen
interaction with the enzyme backbone facilitates tighter binding
between the enzyme and the inhibitor. Prior modeling studies
Scheme 1. Synthesis of Pinanediol Protected Phenylalanine
a
Boronic Acid,45 (boro-Pin)Phe
a
(a) (+)-pinanediol, Et2O, MgSO4, rt, 85%; (b) LiCHCl2, ZnCl2,
THF, −100 °C, 80%; (c) LiN(SiMe3)2, THF, −78 °C; (d) TFA, Et2O,
0 °C, 73%.
developed by Matteson and Winssinger.15,16 The (boro)-
bromopropylglycine analogues were prepared using commercial
(1R)-1-amino-4-bromobutan-1-ylboronic acid (1S,2S,3R,5S)-
(+)-2,3-pinanediol ester 6 (Scheme 2). All peptides, containing
only L-amino acids, were synthesized using Fmoc solid phase
peptide synthesis. All peptidyl boronic acid inhibitors were
prepared by amide bond formation between the fully protected
peptide and the pinanediol protected boronic acid (Scheme 2).
Consequently, the side chains of the peptide were deprotected,
and in the majority of the sequences, this was followed by
boronic acid deprotection (Scheme 2). For compounds 21 and
23 (Scheme 4), the ester protection of the acid was not
deprotected and is abbreviated as boro-Pin. Table 1 lists the
4225
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
Scheme 2. Schematic Illustration of Solid Phase Peptide Synthesis and Subsequent Conjugation to Boronic Acids ((boro)Phe
and (boro)Bpg Are Used As Examples)
a
Table 1. HPLC (Semi-prep) Retention Times and Observed m/z Ratios by ESI-MS
inhibitor
Z-SSKn(boro)F
Rt, min
exact mass calcd
m/z, [M + H-18]+
[M + H]+
b
12
13
14
15
16
17
18
19
20
21
22
23
24
25
25.6
714.38
744.29
659.40
606.35
636.27
677.39
707.30
753.42
783.33
917.44
1013.26
1147.37
817.45
847.37
697.0
729.0
642.1
589.4
620.9
660.0
691.9
736.6
768.5
nd
nd
b
Z-SSKn(boro)Bpg
24.2
nd
c
Ahx-SSQn(boro)L
21.9
660.0
607.4
638.9
678.0
709.7
754.5
786.5
917.9
1015.8
1147.8
818.0
849.8
b
Ahx-SQn(boro)F
20.9
b
Ahx-SQn(boro)Bpg
19.0
c
Ahx-ASQn(boro)F
23.1
b
Ahx-ASQn(boro)Bpg
Ahx-FSQn(boro)F
20.7
b
24.5
b
Ahx-FSQn(boro)Bpg
Ahx-FSQn(boro-Pin)Bpg
SIB-Ahx- FSQn(boro)Bpg
SIB-Ahx-FSQn(boro- Pin)Bpg
Ahx-NaphSQn(boro)F
Ahx-NaphSQn(boro)Bpg
22.9
b
30.2
b
32.4
nd
b
33.5
nd
b
24.6
800.0
831.9
b
24.9
a
nd: not detected. Eluent A = 0.1% TFA in 5% MeOH/95% water. Eluent B = 0.1% TFA in MeOH. Note: In some cases molecular ion peaks
b
corresponding to [M − (OH)2 + H]+ and/or [M + Na]+ were detected, as well (data not shown). HPLC method: 85% A for 5 min; gradient of
c
15−90% B over 25 min at 5 mL/min; 90% B for 5 min. HPLC method: 90% A for 5 min; gradient of 10−95% B over 25 min at 5 mL/min; 95% B
for 5 min.
with aldehyde inhibitors contrasted the differences in the S1
pockets of PSA and chymotrypsin. As we predicted based on
the results from these studies, (boro)Phe inhibitor 12 was 9-
fold better at inhibiting chymotrypsin and (boro)Bpg inhibitor
13 was ∼8-fold better at inhibiting PSA (Table 2).
Previously, we used an iterative approach that combined
analysis of peptidyl aldehyde inhibitor libraries with molecular
modeling studies. On the basis of these studies, it was
determined that a peptide sequence length of at least four
amino acids was required for PSA inhibition. From a series of
27 natural and unnatural amino acids, we determined that
hydrophobic amino acids were best in the P2 position, with
norleucine being the most preferred. In the P3 position,
glutamine and lysine were equally preferred. Because the
presence of lysine makes the peptide more susceptible to
cleavage by trypsin-like proteases present in the serum, for our
extended studies glutamine was selected as a preferred amino
acid for the P3 position. In addition, having Lys along with Ahx
in the peptide sequence could present more elaborate synthetic
routes when it comes to selective radiolabeling or other N-
terminal modifications. Finally, modeling studies suggested a
preference for serine or alanine in the P4 position. Therefore,
4226
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
Table 2. Comparison of the Inhibition Constant Ki for PSA vs Chymotrypsin
inhibitor
Ki (nM) PSA inhibition
Ki (nM) chymotrypsin inhibition
xspecificity, relative to PSA inhibition
12
13
14
15
16
17
18
19
20
21
22
24
25
Z-SSKn(boro)F
164.503
24.228
43.053
22.01
18.108
196.882
137.837
3.765
0.11
8.13
3.20
0.17
0.83
0.12
3.18
0.002
8.03
6.77
10.14
0.03
4.94
Z-SSKn(boro)Bpg
Ahx-SSQn(boro)L
Ahx-SQn(boro)F
Ahx-SQn(boro)Bpg
Ahx-ASQn(boro)F
302.096
310.286
52.11
250.575
36.22
Ahx-ASQn(boro)Bpg
Ahx-FSQn(boro)F
165.921
0.135
61.484
72.288
73.327
64.409
228.782
102.463
Ahx-FSQn(boro)Bpg
Ahx-FSQn(boro-Pin)Bpg
SIB-Ahx- FSQn(boro)Bpg
Ahx-NaphSQn(boro)F
Ahx-NaphSQn(boro)Bpg
580.724
496.3
652.8
7.214
505.676
a
Scheme 3. PSA Active Site Interactions with Amino Acids in P1−P5 Positions of Peptide Boronic Acid 20
a
Substrate binding pockets on the protease are labeled as S.
on the basis of these studies, we prepared peptide inhibitors
15−25 with the sequence X-SQn(boro)Bpg or X-SQn(boro)F
and characterized their ability to inhibit PSA vs chymotrypsin
(Table 2).
was minimal as in 18 and 20. From this entire group, 20 was
the most selective PSA inhibitor with an 8-fold lower Ki for
PSA vs chymotrypsin. In contrast to the specificity conveyed by
the (boro)Bpg, all of the (boro)Phe inhibitors were much
better (i.e., 19, 450-fold) chymotrypsin inhibitors. In fact, the
inhibitor 19 with a Ki of 135 picomolar, is one of the most
potent chymotrypsin inhibitors ever described (Table 2).
PSA Inhibitors Affect PSA Blood Levels. PSA is secreted
in an enzymatically active form and accumulates to high levels
in the extracellular fluid surrounding prostate cancer cells. A
fraction of this PSA enters the circulation where it is rapidly
inactivated due to the formation of covalent complexes with the
serum protease inhibitors α-2-macroglobulin (A2M) and α-1-
antichymotrypsin (ACT).6,7 To assess whether the formation
of these complexes could be inhibited, PSA was incubated with
either A2M or ACT in aqueous buffer in the presence or
absence of 20 (Figure 2A). Western blot analysis demonstrated
that 20 completely blocked the ability of PSA to bind to both of
these serum protease inhibitors. Subsequently, we evaluated the
effect of the PSA inhibitor 20 on serum PSA levels generated by
PSA-producing human prostate cancer xenografts in nude mice.
First, we determined that the PSA inhibitor 20 had no effect on
the standard ELISA used to measure PSA levels in humans
(Figure 2B). Using different antibodies, this assay can measure
Effects of P5 Modification on PSA Inhibition. Previous
modeling studies indicated that amino acids in P5 position and
beyond in the PSA inhibitor sequence were outside of PSA’s
catalytic site and were solvent exposed. This suggested that
hydrophobic amino acids and bulky adducts such as chelating
groups could be added to the amino terminus of the peptide
without affecting the ability of the compound to act as a PSA
inhibitor (Figure 1). In addition, preliminary in vivo studies
suggested that inhibitors such as 14 had a remarkably short
serum half-life due to rapid renal clearance. Thus, in an attempt
to improve the half-life, inhibitors 19, 20, and 24, 25 were
generated that contained bulky hydrophobic amino acids in the
P5 position. In addition, an aminohexanoic (Ahx) group was
placed at the N-terminus to serve as a linker to chelating groups
(e.g., NOTA, DOTA) or radiolabeled prosthetic groups (SIB,
SFB, etc). The addition of this Ahx group did not affect PSA
inhibition to a significant degree. Analysis of the Ki for the
(boro)Bpg inhibitors demonstrated that in some cases deletion
as in 16 or substitution as in 25 of P5 Ser had a deleterious
effect on PSA inhibition, whereas in other cases the effect on Ki
4227
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
results suggest that the inhibitor may not bind PSA efficiently in
plasma due to nonspecific protein binding.
In Vivo Study. On the basis of these results suggesting 20
can bind to PSA in vivo, we next evaluated whether this
inhibitor could affect the growth of PSA-producing LNCaP
human prostate cancer xenografts growing subcutaneously in
nude mice. On the basis of prior studies with peptidyl boronic
acid inhibitors in this setting, we selected a daily dose of 10 mg/
kg administered intravenously for five consecutive days. This
regimen was well tolerated and induced no significant weight
loss in treated animals. PSA inhibitor treated animals
experienced a delay in tumor growth over the first two cycles
of therapy and had smaller tumors at the end of the treatment
period, although the difference did not achieve statistical
significance (Figure 4).
Addition of 4-Iodobenzoate and 125I Labeling. An
important advantage of our peptide-based inhibitors is that they
can be readily modified to incorporate imaging moieties such as
radionuclide-bearing groups. To assess circulating half-life and
biodistribution, select inhibitors were radiolabeled via the
addition of 4-[125I]iodobenzoate to the Ahx linker. To
introduce this label containing N-hydroxysuccinimidyl-4-[125I]-
iodobenzoate (SIB) was prepared as previously described by
Dekker et al.24 (Scheme 4). This [125I]SIB prosthetic group was
previously developed for radioiodination of proteins and
peptides.25,26 Unlabeled SIB-peptides were prepared as
chromatographic references and also as a comparison with
the noniodinated analogues in the enzyme kinetic studies. The
coupling of SIB to Ahx as in 22, slightly decreased the Ki for
PSA and increased the Ki for chymotrypsin (Table 2).
In initial studies, we generated [125I]SIB-Ahx-SSQn(boro)-
Bpg, which when injected intraperitoneally (ip) into mice had
an extremely short circulating half-life of 12 min (data not
shown). Subsequently, in an effort to improve circulating half-
life, we generated the 125I labeled version of Ahx-NaphSQn-
(boro)F. This compound had a longer circulating half-life of
∼2.7 h following ip injection (Figure 5A). Further, we
performed a biodistribution study in mice bearing PSA-
producing LNCaP tumors. Mice were injected with [125I]SIB-
Ahx-NaphSQn(boro)F, and at 2, 4, and 6 h, individual mice
were sacrificed and their tissues (liver, kidney, heart, brain,
skeletal muscle, and tumor) were collected, weighed, and
analyzed using a gamma counter (Figure 5B). [125I]SIB-Ahx-
FSQn(boro)Bpg 26 and [125I]SIB-Ahx-FSQn(boro-Pin)Bpg 27
were also used for biodistribution studies (Figure 5C). For
these three inhibitors, the majority of activity was detected
within the liver and kidney by 2 h. Tumor levels at each time
point and dose were very similar to those in the heart and
skeletal muscle. Highest amount of activity detected in the
tumors was around 3−4 h postinjection.
Figure 1. Structure of peptide boronic acids with hydrophobic amino
acid substituents in the P5 position.
free PSA, which corresponds to the fraction of PSA in the blood
that is unbound to protease inhibitors because it lacks
enzymatic activity and total PSA, which corresponds to the
sum of the free PSA plus the amount of PSA bound to ACT.
The fraction of PSA bound to A2M cannot be measured due to
lack of antibody that specifically recognizes this complex. In this
experiment, mice received three 5-day courses of 20 at 10 mg/
kg and then blood was obtained for free and total PSA
measurement. Mice treated with 20 had an approximately 40%
lower level of total PSA/g of tumor and a 23% lower level of
free PSA/g of tumor compared to control mice (Figure 2C,D).
These results suggest that the inhibitor is able to block PSA
complex formation with ACT and alter PSA clearance in
treated animals.
To be able to predict the inhibitor’s behavior in vivo, we next
assessed whether 20 maintained the ability to inhibit PSA in
buffer, cell culture media, and plasma (Table 3). Enzyme−
inhibitor interactions are generally measured in buffer, where
we observed the lowest Ki value for Ahx-FSQn(boro)Bpg.
When the assay was performed in RPMI with no additives,
there was no change in the Ki value. However, when the assay
was performed in serum containing media, we saw a linear
decrease in the inhibition potential of the agent (Figure 3). The
study demonstrated that in the presence of serum containing
RPMI the Ki for PSA inhibition increased ∼10-fold. These
DISCUSSION
■
The goal of these studies was to identify a specific and potent
PSA inhibitor that could be used to investigate the role of PSA
in the pathobiology of prostate cancer.4 Such a role is suggested
by the continued production of PSA throughout the course of
the disease, including production by men with poorly
differentiated cancers that have become highly resistant to
multiple forms of androgen ablative therapy. PSA is able to
proteolytically modify growth regulatory proteins that are
important in cancer growth and survival. These include insulin
growth factor binding proteins (IGFBP) 2, 3, and 5,27 PTH-
related protein,28,29 latent TGF-β2,30 and extracellular matrix
4228
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
Figure 2. PSA inhibitor blocks PSA binding to serum protease inhibitors and alters PSA blood levels. (A) Western blot analysis of inhibition of PSA
complex formation with A2M and ACT. FP = control non-PSA containing female plasma; all mixtures were incubated for overnight at room
temperature. (B) PSA measurements using Hybritech assay of 100 ng/mL PSA in RPMI + 1% bovine serum albumin + 1 μM 6 or 1 μM Z-SSKn-
OH peptide + 6 or 1 μM 20. (C) Free and total PSA levels in serum of mice bearing PSA-producing LNCaP prostate cancer xenografts either
untreated or receiving three courses of 20 at dose of 10 mg/kg for five consecutive days/week. (D) Free and total PSA levels per g of tumor and ratio
of free/total PSA in mice treated as in (C).
Table 3. Comparison of the PSA Inhibition Constant of 20
in Different Media
inhibitor Ahx-FSQn(boro)Bpg screened in different media
Ki (nM)
a
PSA buffer (pH 7.8)
72
75
b
RPMI (no additives)
b
c
RPMI (+1% FBS )
142
699
604
b
c
d
e
RPMI (+10% FBS +1% Pen/Strep +1% Gln )
female plasma
a
b
PSA buffer: 50 mM TRIS·HCl, 100 mM NaCl. RPMI: RPMI-1640
medium, cell culture media (Gibco: Invitrogen, Grand Island, NY).
c
d
e
FBS: fetal bovine serum. Pen/Strep: penicillin/streptomycin. Gln:
Figure 3. Evaluation of the activity of inhibitor 20 in the presence of
bovine serum in the media.
L-glutamine.
components fibronectin, and laminin.31 Using siRNA to
knockdown PSA production, Williams et al.32 demonstrated
the importance of PSA production in the in vitro and in vivo
growth of human prostate cancer cell lines. These combined
results suggest that the enzymatic activity of PSA may also
represent a potential therapeutic target in prostate cancer and
further support the development of selective PSA inhibitors to
4229
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
this regard, Ulmert et al.35 recently reported on the
development of 89Zr labeled monoclonal anti-PSA antibody
that selectively binds to the catalytic site of PSA. Positron
emission tomography (PET) imaging demonstrated that this
89Zr antibody accumulated in tumors with a peak activity of
19.6% %ID/g at 24 h postinjection.
In our studies, we have focused on peptidyl boronic acid-
based approach to PSA inhibition. We have pursued this
strategy because peptides provide a platform for easy
introduction of imaging moieties and therapeutic agents.
Other strategies for PSA inhibition have also been described;
the first PSA inhibitors reported in the literature used a
homology model derived from porcine kallikrein to design and
characterize a series of monocyclic β-lactam compounds.36,37
Prior to this study, it had been shown that β-lactams were
efficient and specific inhibitors of several serine proteases,
especially human leukocyte elastase.38,39 The best inhibitor
from the studies of Adlington et al. had an IC50 value of 226
nM for PSA.36,37 However, the specificity and stability of this
compound and others in the class versus chymotrypsin and
other proteases was not reported. Other attempts have been
made to find small-molecule inhibitors for PSA using
compounds other than β-lactams. Azapeptides, which can
inhibit both cysteine and serine proteases, have been developed
and used to effectively inhibit PSA with Ki as low as 500 nM.40
To elucidate the role of PSA in angiogenesis, Koistinen et al.
screened a library of nearly 50000 compounds to discover
inhibitors of PSA.41 Out of the compounds screened, only three
were identified that inhibited PSA at sub μM concentrations.
The active compounds were either benzoxazinone or triazole
derivatives, which inhibited PSA with IC50 values of 300−900
nM and had 10−30-fold specificity vs chymotrypsin.
Figure 4. In vivo evaluation of 20, Ahx-FSQn(boro)Bpg. Mice bearing
subcutaneous PSA-producing LNCaP prostate cancer xenografts were
either untreated (tumor control) or received three courses of 20 at
dose of 10 mg/kg for five consecutive days/week (treatment interval
indicated by gray bars).
determine the significance of PSA’s enzymatic activity in the
progression and growth of prostate cancer.
A further goal of these studies would be to identify a peptidyl
PSA inhibitor as an imaging agent in patients with prostate
cancer. The potential use of a PSA inhibitor for imaging
prostate cancer is particularly attractive because prostate cancer
preferentially metastasizes to the bone, resulting in the
formation of predominantly osteoblastic lesions.33,34 Ther-
apeutic response in the bone is difficult to evaluate with current
bone scan imaging modalities. Thus, there is a compelling need
for better imaging modalities for prostate cancer that can be
used to detect disease recurrence at an earlier time point and to
better evaluate the therapeutic efficacy of new treatments. In
In this work, we have developed a small library of
peptidomimetics containing peptidyl boronic acid, which
shows an improved potency for PSA (<100 nM) and specificity
a
Scheme 4. Peptide Radiolabeling with [125I]SIB and Preparation of Cold Reference
a
(a) N-Hydroxysuccinimidyl-4-iodobenzoate, triethylamine, MeOH, rt, 30 min; (b) [125I]N-hydroxysuccinimidyl-4-iodobenzoate, triethylamine,
DMF, rt, 40 min.
4230
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
Figure 5. Evaluation of 125I labeled PSA inhibitor in vivo. (A) Blood clearance of [125I]SIB-Ahx-NaphSQn(boro)F from mice blood measured over
time. Data shows counts per second normalized to starting counts obtained in plasma as 2 h. Four animals were sacrificed at each time point. (B)
Biodistribution of [125I]SIB-Ahx-NaphSQn(boro)F, shown as percent of injected dose per gram (%ID/gram). Two mice were sacrificed per time
point. Tissues were harvested at 2, 4, and 6 h, weighed, and counted in a gamma counter. Mice had PSA expressing LNCaP xenografts in bilateral
flanks. (C) Biodistribution of [125I]SIB-Ahx-FSQn(boro)Bpg 26 and [125I]SIB-Ahx-FSQn(boro-Pin)Bpg 27, shown as percent of injected dose per
gram (%ID/gram). Each drug was injected into two mice, and one animal was sacrificed per time point. Tissues were harvested at 2 and 4 h,
weighed, and counted in gamma counter. Mice had single PSA expressing LNCaP xenografts.
over chymotrypsin (up to 10-fold). These results suggest a
potential use of these inhibitors as targeted therapeutics or
imaging agents. Systemic administration of inhibitors at doses
of 10 mg/kg for five consecutive days/week ×3 resulted in
tumor growth changes. PSA inhibitor 20 was able to block
complex formation between PSA and serum protease inhibitors
A2M and ACT and did not affect measurement of PSA by
ELISA assay. PSA inhibitor treatment altered PSA blood levels,
reducing the total PSA/g by 40% and the free PSA/g by 23%
while increasing the free/total PSA ratio by ∼25%. These
lowering of the total PSA levels appear consistent with the PSA
inhibitor disrupting PSA binding to ACT, thus decreasing the
amount of PSA-ACT complex that typically makes up about
75% of the total PSA measured in men with prostate cancer.
The reasons for the reduction in free PSA levels in this model is
less clear, as free PSA in the circulation has been demonstrated
to be enzymatically inactive due to a combination of incomplete
processing of the proPSA zymogen and proteolytic process-
ing.42 Potential explanation for this observed decrease in free
PSA may be that the inhibitor changes the kinetics of PSA
clearance upon binding to PSA or may decrease production of
PSA by prostate cancers by some undefined mechanism. While
these changes in serum PSA levels suggest that the inhibitor is
reaching the target intratumorally, in vivo tumor growth assays
demonstrated that the PSA inhibitor has a nonsignificant,
minimal effect on the inhibition of growth of subcutaneously
implanted human prostate cancer xenografts. This lack of effect
may be due to the short biological half-life of the drug as there
is insufficient exposure to completely inhibit PSA, which is
continuously being secreted into the tumor microenvironment
by prostate cancer cells. Alternatively, the lack of effect may be
a contextual one: PSA may play no role in the growth of
subcutaneously implanted prostate cancer xenografts, as
prostate cancer typically forms osteoblastic bone metastases.
The mechanism whereby PSA may stimulate bone adhesion or
osteoblast growth and/or differentiation has not been clearly
defined. However, PSA can release cytokines involved in bone
turnover such as IGF-1 and TGF-β2.27,30 PSA also hydrolyzes
PTHrP, an important mediator of osteolytic bone metastases,
and inactivates PTHrP-stimulated cAMP accumulation in
mouse osteoblasts.28,29 This suggests PSA could alter the
bone metastatic phenotype from osteoclastic to osteoblastic
through PTHrP degradation. Therefore, in future studies, we
plan to assess the effect of PSA inhibitors on growth in
potentially more relevant models of prostate cancer bone
metastases.
CONCLUSION
■
One of our strategies in validating PSA as a “druggable” target is
to inhibit its protease activity by creating PSA specific enzyme
inhibitors. By introducing new structural differences at P5 and
P1, we prepared boronic acid peptidates for targeted inhibition
of PSA. These boronic acid peptidates have Ki values in the
nanomolar range and rank among the best inhibitors of PSA
reported so far. The best of these inhibitors achieved 8-fold
selectivity vs chymotrypsin. In addition, we demonstrated that
the inhibitors could block PSA complex formation with serum
protease inhibitors. In vivo studies showed that inhibition of
PSA resulted in decrease of the total PSA as well as the free
PSA. The inhibitors did not significantly inhibit the growth of
subcutaneous PSA-producing xenografts, and further study is
needed to determine the microenvironmental context in which
PSA activity is most relevant. In addition, these studies suggest
that, while these PSA inhibitors will not be useful as imaging
4231
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
of inhibitors were employed for the assay: 25, 12.5, 6.25, 3.1, 1.56,
0.78, 0.39, 0.19, 0.098, and 0.049 μM.
agents without further optimization, the inhibitors are useful
research tools to help in the elucidation of PSA’s role in the
growth and progression of prostate cancer.
Selectivity of the compounds was determined by screening the
products against chymotrypsin. Chymotrypsin assays were performed
by monitoring the fluorescence of Suc-AAPF-AMC substrate at the
above-described conditions. The Km value of chymotrypsin for the
substrate Suc-AAPF-AMC was determined to be 14 μM. Reactions
were performed on 96-well plate where the final concentration of
chymotrypsin was 2 nM and the substrate was 25 μM. The assay
mixture contained peptidyl boronic acid in DMSO (20 μL), Suc-
AAPF-AMC in PSA buffer (90 μL), and chymotrypsin solution (90
μL). The chymotrypsin substrate solution was added last to each assay
solution, and the kinetic data was collected immediately.
The inhibition constant (Ki) of each compound was determined by
solving the equation: Ki = 1/slope × 1/(1 + [S]/Km), where the slope
is obtained from a linear plot of V0/Vi −1 vs [I]. The reciprocal of the
slope is also equal to the Ki(app), apparent inhibition constant. [S] =
substrate concentration, [I] = inhibitor concentration, Km = Michaelis
constant, V0 = rate of substrate hydrolysis in the absence of inhibitor,
and Vi = rate in the presence of inhibitor. The Ki values are reported in
Table 3.
Biodistribution Studies. Biodistribution studies were conducted
to evaluate the uptake of [125I]SIB-Ahx-NaphSQn(boro)F, [125I]SIB-
Ahx-FSQn(boro)Bpg, and [125I]SIB-Ahx-FSQn(boro-Pin)Bpg in
LNCaP human prostate cancer xenograft models. Mice received
between 230 and 580 μCi of the selected radiolabeled inhibitor in 200
μL of sterile saline through intravenous tail-vein injection. The animals
(n = 2 per group) were euthanized between 2 and 6 h postinjection,
and blood was immediately harvested by cardiac puncture. These time
points were selected based on biodistribution assays previously
performed in our lab. The tumors along with liver, kidney, spleen,
heart, lung, and skeletal muscle were harvested, weighed, and analyzed
using gamma counter for accumulation of 125I radioactivity. Count data
were background and decay corrected, and the tissue uptake for each
sample was calculated as counts/mass.
In Vivo Assay. In vivo studies were performed by using PSA-
producing LNCaP human prostate cancer xenografts as previously
described.8 Nude male mice (n = 10 per group) bearing subcutaneous
tumors were treated once daily for three cycles of five days with doses
of 10 mg/kg. The compound of interest was reconstituted in sterile
saline and given by intravenous injections. Tumors were measured by
caliper, and their volume was determined by the formula: Vol(cc) = L
× W × H × 0.5326. Body weight and tumor sizes were measured twice
per week. Serum levels of free and total PSA were measured 24 h after
the last dose of the inhibitor.
Synthesis. Peptide Synthesis. All peptides were individually
prepared by solid-phase synthesis using AAPPTec Apex 396 40-well
peptide synthesizer. The products were obtained using standard Fmoc
protocol, starting with preloaded Fmoc-Nle-OH onto a 2-ClTrt resin.
The amino acids (0.1 mmol) were coupled for 1 h using double
coupling conditions. HOBt (5 equiv) served as activating reagent; 5
equiv of Fmoc-protected amino acids were used with 0.5 M DIC in
NMP. Once completed, the peptide on the resin was washed
consecutively with NMP, methanol, and DCM. When necessary the
peptides were capped with Cbz (0.4 M) in the presence of DIEA (4
equiv). In other instances, Ahx was added to the N-terminus of the
sequence to serve as a handle containing free amino group. The final
sequences were washed with DCM and cleaved from the resin with 1%
TFA in DCM. On cleavage, the peptides were diluted with ethyl
acetate and after removal of the solvent the peptides were purified by
semiprep RP-HPLC using a Phenomenex Luna 250 mm × 10 mm,
C18 column, followed by lyophilization for 24 h. The purity of all
synthesized peptides was determined by HPLC to be >95%. The
peptides masses were confirmed by mass spectrometry and were used
for further coupling.
EXPERIMENTAL SECTION
■
Chemistry. All amino acids used in this study are of L-form. All
Fmoc-amino acids and 2-chlorotrityl chloride resin used were
purchased from AnaSpec (Fremont, CA). Benzylboronic acid was
purchased from Alfa Aesar (Ward Hill, MA), and (1R)-1-amino-4-
bromobutan-1-ylboronic acid (1S,2S,3R,5S)-(+)-2,3-pinanediol ester
was purchased from BoroChem SAS (Caen, France). Fmoc-(R)-3-
amino-4-(2-naphthyl)-butyric acid was used for the synthesis of Naph-
containing analogues and was purchased from PepTech Corp.
(Burlington, MA). All other reagents and solvents were purchased
from Sigma-Aldrich (USA) and used as received. All reactions were
carried out under anhydrous conditions, nitrogen atmosphere and dry
solvents, unless otherwise noted. Reactions were monitored by thin
layer chromatography (TLC) carried out on 250 μm silica gel plates
(60 F254). All organic extracts were dried over MgSO4, and solvents
were removed with a rotary evaporator, followed by high vacuum oil
pump pressure. The products were characterized by 1H and 13C NMR
on a 400 MHz Bruker Avance NMR spectrometer. Mass spectra were
measured on a Bruker 3000 Esquire mass spectrometer equipped with
ESI. The HPLC used for the analysis and purification was equipped
with photodiode array detector and purchased from Waters (Milford,
MA). The final products were purified on a semiprep HPLC column
from Phenomenex: Luna C18, particle size 10 μm, 250 mm × 10 mm.
The purity of all synthesized peptides was determined by analytical
HPLC to be >95%.
[
125I]NaI was purchased from MP Biomedicals (Solon, OH). All
radiolabeled products were purified on reverse phase radio-HPLC
using Waters Delta 600 pump, Waters 2487 UV/vis dual wavelength
absorbance detector, and Beta-RAM radio HPLC detector, equipped
with sodium iodide cell for soft gammas (I-125) from IN/US Systems
(Brandon, FL). The HPLC column was purchased from Phenomenex:
Luna C18, particle size 5 μm, 150 mm × 4.60 mm. Solid-phase C18
extraction cartridges (Sep-Pak Plus) were purchased from Waters.
Radioactivity was measured in CRC-10R Capintec dose calibrator
(Ramsey, NJ). The CT/SPECT scanning was performed in our small
animal imaging center using X-SPECT imager (Gamma Medica,
Northridge, CA). All animal studies performed were in full compliance
with our institutional guidelines related to conducting experiments
using animals. The radioactivity of the harvested tissue was measured
with automated Gamma 5500 counter (Beckman, Irvine, CA). The
percent injected dose per gram of tissue (%ID/g) was calculated based
on the number of counts measured for 0.01 mL of the dosing solution.
Biology. Chymotrypsin (from bovine pancreas) and PSA (from
human seminal fluid) were purchased from CalBiochem (San Diego,
CA). The fluorescent PSA substrate, Mu-SRKSQQY-AMC (morpho-
linocarbonyl-Ser-Arg-Lys-Gln-Gln-Tyr-aminomethylcoumarin), was
selected as the PSA substrate based on a previous study12 and was
custom synthesized by California Peptides (Napa Valley, CA). The
chymotrypsin substrate Suc-AAPF-AMC was purchased from Sigma-
Aldrich. All kinetic assays were performed on a Beckman DTX-880
multimode detector plate reader.
Enzyme Assays. Peptidyl boronic acids were dissolved in DMSO
at 20 mM. Stock solutions could be stored at −20 °C and were stable
for one year. More diluted solutions were prepared as necessary. The
activity of the inhibitors was determined by monitoring the change of
fluorescence of AMC substrate (ex 370 nm; em 465 nm) due to
hydrolysis by PSA. Product formation was followed over 30 min. The
Km for the substrate Mu-SRKSQQY-AMC was previously determined
to be 140 μM.14 Kinetic measurements were done in triplicate and run
at 37 °C in PSA buffer (50 mM TRIS, 100 mM NaCl, pH 7.8).
Reactions were performed on 96-well plate where the final
concentration of PSA was 2.5 μg/mL (2 nM active PSA) and the
substrate was 300 μM. The assay mixture contained 100 μL of buffer,
peptidyl boronic acid in DMSO (20 μL), PSA solution (20 μL), and
Mu-SRKSQQY-AMC in buffer (60 μL). The following concentrations
Pinanediol Boronic Acid Ester. The synthesis of (+)-pinanediol
protected benzyl boronic acid (boro-Pin)Phe (5) was performed as
described43−46 (Scheme 1). The synthesis of (+)-pinanediol protected
bromopropylglycine boronic acid, (boro-Pin)Bpg, or (1R)-1-amino-4-
bromobutan-1-ylboronic acid (1S,2S,3R,5S)-(+)-2,3-pinanediol ester 6
4232
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
has been previously described as well.16,47 However, both starting
materials are now commercially available and can be purchased from
BoroChem SAS, France. For preparation of (boro)Phe analogues, the
following procedure modifications were performed and the products
were obtained as TFA salts.
Hz, 1H), 3.11 (dd, J = 14.2, 6.4 Hz, 1H), 3.25 (dd, J = 8.0, 6.5 Hz,
1H), 4.50 (d, J = 8.2 Hz, 1H), 7.30−7.37 (m, 5H). 13C NMR
(methanol-d4): δ 24.25, 27.37, 27.49, 28.81, 35.92, 36.46, 39.38, 40.77,
52.40, 80.33, 89.14, 128.58, 130.00, 130.35, 137.88. MS (ESI) m/z
+
calcd for C18H27BNO2 300.21, found 300.1 [M]+.
General Method for Synthesis of Peptidyl Boronic Acids
(12−25). The desired N-terminus protected peptide Boc-Ahx- or Z-
(1 mmol) was dissolved in 50 mL of CH2Cl2, and the solution was
cooled in an ice bath. HOBt (1.5 mmol, 0.229 g) and EDC (1.5 mmol,
0.288 g) were added to the mixture and stirred for 15 min. Boronic
acid pinanediol ester (1 equiv) was added, and after 10 min of stirring
at 0 °C, N-methylmorpholine (2 mmol, 0.23 mL) was added. The
solution was stirred overnight, and the temperature was allowed to
warm slowly to room temperature. The reaction mixture was washed
with water ×2 and brine ×2. The organic layer was dried with MgSO4
and evaporated to dryness to yield crude protected peptide in 95−97%
yield. The resulting white powder was then stirred with 95% TFA in
water to remove all of the peptide side chain protecting groups. After 1
h of stirring at room temperature, the reaction was concentrated in
vacuo and resulting residue was resuspended in a 1:1 mixture of ethyl
ether/water (40 mL). Phenyl boronic acid (0.305 g, 2.5 mmol) was
then added to remove the (+)-pinanediol protecting group. The
deprotection was allowed to go overnight at room temperature. The
aqueous layer containing the fully deprotected peptidyl boronic acid
was washed with ethyl ether (30 mL, 3×) and lyophilized. The product
was purified by HPLC, and the resulting white powder was
characterized by MS. The purity of the product was determined by
HPLC to be >95%.
(4-Iodo-benzoylamino)-Ahx-FSQn(boro)Bpg (22). To a solution of
20 (0.0047 g, 0.006 mmol) in 0.3 mL of dry methanol was added 1.2
equiv of N-hydroxysuccinimidyl-4-iodobenzoate24 (0.0025g, 0.007
mmol), followed by 3 equiv of triethylamine (0.0025 mL, 0.018
mmol). The mixture was stirred for 30 min at room temperature and
purified on an analytical RP-HPLC column (Phenomenex: Luna C18,
particle size 5 μm, 150 mm × 4.60 mm), using a gradient of 30−100%
eluent B over 16 min at 1.25 mL/min, then gradient of 0−70% eluent
A over 12 min (buffer A, 0.1% TFA in 5% acetonitrile; eluent B, 0.1%
TFA in acetonitrile), product eluted at 16 min. The purity of the
product was determined by HPLC to be >95%. MS (ESI) m/z calcd
for C40H58BBrIN7O10 1013.26, found 1037.8 [M + Na]+.
(+)-Pinanediol Phenylmethyl-1-boronate (2). To a solution of 1.0
g (7.36 mmol) of benzyl boronic acid 1 in 80 mL of diethyl ether were
added 1 equiv of (+)-pinanediol (1.25 g) and 3.0 g of MgSO4. The
solution was stirred for 2−3 days, and the product formation was
followed by NMR. The reaction mixture was filtered out, and the
organic layer was washed with saturated solution of NaHCO3 (20 mL
× 2) and brine (20 mL × 2). After drying, the organic layer was
concentrated under low pressure and kept under high vacuum
overnight to yield 85% colorless oil. The product was used for the next
step without purification. 1H NMR (CDCl3): δ 0.85 (s, 3H), 1.07 (d, J
= 11.0 Hz, 1H), 1.28 (s, 3H), 1.39 (s, 3H), 1.9 (m, 2H), 2.05 (m, 1H),
2.2 (m, 1H), 2.28 (m, 1H), 2.35 (s, 2H), 4.29 (dd, J = 8.9, 2.1 Hz,
1H), 7.16−7.34 (m, 5H). 13C NMR (CDCl3): δ 23.95, 26.50, 28.31,
28.68, 36.22, 38.36, 39.48, 51.31, 78.34, 85.89, 110.64, 124.84, 128.47,
129.13, 138.86. MS (ESI) m/z calcd for C17H23BO2 270.18, found
288.0 [M + H + H2O]+, 293.1 [M + Na]+.
(+)-Pinanediol (S)-1-Chloro-2-phenylethaneboronate (3). In a
three-neck flask equipped with thermometer, a solution of 10 mL of
freshly distilled THF and 1.3 mL of DCM was cooled to −100 °C (in
ethanol/liquid nitrogen bath) under argon atmosphere and chilled at
−78 °C. n-BuLi (8.6 mmol; 2.5 M in hexane) was added via syringe by
running it down the cold wall of the reaction flask over a period of 15
min. White precipitate of LiCHCl2 was formed, and after 10 min of
stirring, cold solution of 2 (0.38 g, 1.4 mmol) in 2 mL of dry THF was
added. After 10 min of stirring at −100 °C, 8 mL of ZnCl2 solution
(0.5 M in THF) were added. The solution was stirred overnight, and
the temperature was allowed to rise to room temperature. The
reaction mixture was evaporated and redissolved in hexane. The
organic phase was washed with satd NH4Cl, water, and brine, then
dried and concentrated to give about 80% crude product as colorless
1
oil. The product was used for the next step without purification. H
NMR (CDCl3): δ 0.84 (s, 3H), 1.08 (d, J = 11.2 Hz, 1H), 1.28 (s,
3H), 1.36 (s, 3H), 1.90 (m, 2H), 2.08 (m, 1H), 2.19 (m, 1H), 2.35 (m,
1H), 3.11 (dd, J = 13.7, 8.6 Hz, 1H), 3.23 (dd, J = 13.9, 7.9 Hz, 1H),
3.68 (dd, J = 8.5, 7.8 Hz, 1H), 4.46 (dd, J = 11.3, 2.0 Hz, 1H), 7.20−
7.35 (m, 5H). 13C NMR (CDCl3): δ 23.95, 26.22, 27.02, 28.33, 29.71,
35.15, 38.24, 39.35, 40.45, 51.22, 78.56, 86.79, 126.72, 126.75, 128.35,
129.20, 129.24, 138.42. MS (ESI) m/z calcd for C18H24BClO2 318.16,
found 283.0 [M − Cl + H]+.
(4-Iodo-benzoylamino)-Ahx-FSQn(boro-Pin)Bpg (23). 23 was
prepared from 21 using the same procedure as in 22. The product
eluted at 18 min. The purity of the product was determined by HPLC
to be >95%. MS (ESI) m/z calcd for C50H72BBrIN7O10 1147.37,
found 1147.8 [M + H]+, 1171.7 [M + Na]+.
(+)-Pinanediol (R)-1-(N-Bis(trimethylsilyl)amino)-2-phenyl-
ethane-1-boronate (4). An oven-dried flask containing 5 mL of
freshly distilled THF was cooled to −78 °C and flushed with argon.
After addition of 7.2 mL (1.25 equiv) of lithium bis(trimethylsilyl)-
amide and LiN(SiMe3)2 (1 M solution in THF), the contents of the
flask were stirred for 10 min. Crude boronic ester 3 (1.8 g, 5.6 mmol)
was dissolved in 10 mL of THF and added dropwise. The reaction
mixture was stirred overnight and allowed to warm to room
temperature. The resulting solution was concentrated under reduced
pressure, reconstituted in 25 mL of DCM, and filtered through Celite.
The organic layer was dried with MgSO4 and evaporated to give 2.5 g
General Method for Preparation of Radiolabeled Peptidyl
Boronic Acids Such As 26 and 27. [125I]SIB was prepared
according to a procedure suggested by Dekker et al. and purified by
radio-HPLC (retention time: iodobenzoic acid, 10.57 min; SIB, 12.54
min; Sn-precursor, 24.40 min). Radiolabeling of the peptidyl boronic
acids using [125I]SIB was performed as suggested by Chen et al.48 To a
solution of 0.005 g (or about 0.007 mmol) of peptidyl boronic acid
(20) or peptidyl boronic pinanediol ester (21) in 0.1 mL of DMF was
added 0.002 mL of triethylamine, followed by [125I]N-hydroxysucci-
nimidyl-4-iodobenzoate (0.100 mL, 0.5−5 μCi). After stirring at room
temperature for 40 min, 0.005 mL of TFA was added, the mixture was
subjected to HPLC purification (Phenomenex: Luna C18, 5 μm, 150
mm × 4.60 mm), using gradient of 30−100% eluent B over 16 min at
1.25 mL/min, then gradient of 0−70% eluent A over 12 min (buffer A,
0.1% TFA in 5% acetonitrile; eluent B: 0.1% TFA in acetonitrile). The
purity of the product was determined by HPLC to be >95%. The
1
of crude product as light-brown oil. H NMR (CDCl3): δ 0.09 (s,
18H), 0.83 (s, 3H), 0.97 (d, J = 10.6 Hz, 1H), 1.27 (s, 3H), 1.38 (s,
3H), 1.82 (m, 2H), 2.01 (m, 1H), 2.12 (m, 1H), 2.30 (m, 1H), 2.67
(m,1H), 2.74 (dd, J = 13.2, 7.2 Hz, 1H), 3.05 (m, 1H), 4.30 (dd, J =
8.6, 2.1 Hz, 1H), 7.20−7.32 (m, 5H). MS (ESI) m/z calcd for
C24H43BNO2Si2 444.29, found 444.1 [M + H]+.
(+)-Pinanediol (R)-1-Amino-2-phenylethane-1-boronate Trifluor-
oacetate Salt (5). The crude 4 (1.1 mmol, 0.5 g) was dissolved in 10
mL of ethyl ether and cooled down to 0 °C. To the mixture was added
TFA (0.25 mL, 3.3 mmol), and the reaction was allowed to stir for 3 h.
The TFA salt crashed out of solution and was filtered. The final
product was washed three times with cold ethyl ether, resulting in
white powder with 73% yield (0.34 g). 1H NMR (methanol-d4): δ 0.89
(s, 3H), 1.11 (d, J = 10.8 Hz, 1H), 1.34 (s, 3H), 1.42 (s, 3H), 1.93 (m,
2H), 2.06 (m, 1H), 2.26 (m, 1H), 2.43 (m, 1H), 3.03 (dd, J = 14.0, 8.1
[
125I]-peptidyl boronic acid containing fraction was neutralized with 1
M NaHCO3, concentrated under vacuum, and reconstituted in PBS
(pH 7.4). The solution was passed through a 0.22 μm syringe filter
into a sterile vial.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: (410) 955-8875. E-mail: denmesa@jhmi.edu.
4233
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
(8) Denmeade, S. R.; Jakobsen, C.; Janssen, S.; Khan, S. R.; Lilja, H.;
Christensen, S. B.; Isaacs, J. T. Prostate-specific antigen (PSA)
activated thapsigargin prodrug as targeted therapy for prostate cancer.
J. Natl. Cancer Inst. 2003, 95, 990−1000.
(9) Williams, S. A.; Merchant, R. F.; Garrett-Mayer, E.; Isaacs, J. T.;
Buckley, J. T.; Denmeade, S. R. A prostate-specific antigen-activated
channel-forming toxin as therapy for prostatic disease. J. Natl. Cancer
Inst. 2007, 99, 376−385.
(10) Ulmert, D.; Evans, M. J.; Holland, J. P.; Rice, S. L.; Wongvipat,
J.; Pettersson, K.; Abrahamsson, P. A.; Scardino, P. T.; Larson, S. M.;
Lilja, H.; Lewis, J. S.; Sawyers, C. L. Imaging androgen receptor
signaling with a radiotracer targeting free prostate-specific antigen.
Cancer Discovery 2012, 2, 320−327.
(11) LeBeau, A. M.; Singh, P.; Isaacs, J. T.; Denmeade, S. R. Prostate-
specific antigen is a ‘chymotrypsin-like’ serine protease with unique P1
substrate specificity. Biochemistry 2009, 48, 3490−3496.
(12) Denmeade, S. R.; Lou, W.; Malm, J.; Lovgren, J.; Lilja, H.;
Isaacs, J. T. Specific and efficient peptide substrates for assaying the
proteolytic activity of prostate specific antigen. Cancer Res. 1997, 57,
4924−4930.
(13) Singh, P.; LeBeau, A. M.; Lilja, H.; Denmeade, S. R.; Isaacs, J. I.
Molecular insights into substrate specificity of prostate specific antigen
through structural modeling. Proteins 2009, 77, 984−993.
(14) LeBeau, A. M.; Singh, P.; Isaacs, J. T.; Denmeade, S. R. Potent
and selective peptidyl boronic acid inhibitors of the serine protease
prostate-specific antigen. Chem. Biol. 2008, 15, 665−674.
(15) Matteson, D. S.; Michnick, T. J.; Willett, R. D.; Patterson, C. D.
[(1R)-1-Acetamido-3-(methylthuio)propyl]boronic acid and the X-ray
structure of its ethylene glycol ester. Organometallics 1989, 8, 726−
729.
(16) Debaene, F.; Da Silva, J. A.; Pianowski, Z.; Duran, F. J.;
Winssinger, N. Expanding the scope of PNA-encoded libraries:
divergent synthesis of libraries targeting cysteine, serine and
metalloproteases as well as tyrosine phosphatases. Tetrahedron 2007,
63, 6577−6586.
(17) Elgendy, S.; Patel.; Green, D.; Goodwin, C. A.; Scully, M. F.;
Husman, W.; Skordalakes, E.; Kakkar, V. V.; Deadman, J. J. Design of a
novel class of bifunctional thrombin inhibitors, synthesised by the first
application of peptide boronates of solid phase chemistry. Tetrahedron
Lett. 1997, 38, 3305−3308.
(18) Zhu, Y.; Zhao, X.; Zhu, X.; Wu, G.; Li, Y.; Ma, Y.; Yuan, Y.;
Yang, J.; Hu, Y.; Ai, L.; Gao, Q. Design, synthesis, biological
evaluation, and structure−activity relationship (SAR) discussion of
dipeptidyl boronate proteasome inhibitors, part I: comprehensive
understanding of the SAR of alpha-amino acid boronates. J. Med.
Chem. 2009, 52, 4192−4199.
(19) Zhu, Y.; Yao, S.; Xu, B.; Ge, Z.; Cui, J.; Cheng, T.; Li, R. Design,
synthesis and biological evaluation of tripeptide boronic acid
proteasome inhibitors. Bioorg. Med. Chem. 2009, 17, 6851−6861.
(20) Watanabe, T.; Abe, H.; Momose, I.; Takahashi, Y.; Ikeda, D.;
Akamatsu, Y. Structure−activity relationship of boronic acid
derivatives of tyropeptin: proteasome inhibitors. Bioorg. Med. Chem.
Lett. 2010, 20, 5839−5842.
(21) Watanabe, T.; Momose, I.; Abe, M.; Abe, H.; Sawa, R.;
Umezawa, Y.; Ikeda, D.; Takahashi, Y.; Akamatsu, Y. Synthesis of
boronic acid derivatives of tyropeptin: proteasome inhibitors. Bioorg.
Med. Chem. Lett. 2009, 19, 2343−2345.
(22) Auffinger, P.; Hays, F. A.; Westhof, E.; Ho, P. S. Halogen bonds
in biological molecules. Proc. Natl. Acad. Sci. U. S. A. 2004, 101,
16789−16794.
(23) Howard, E. I.; Sanishvili, R.; Cachau, R. E.; Mitschler, A.;
Chevrier, B.; Barth, P.; Lamour, V.; Van Zandt, M.; Sibley, E.; Bon, C.;
Moras, T. R.; Joachimiak, A.; Podjarny, A. D. Ultrahigh resolution drug
design I: Details of interactions in human aldose reductase−inhibitor
complex at 0.66 Å. Proteins: Struct., Funct., Genet. 2004, 55, 792−804.
(24) Dekker, B.; Keen, H.; Shaw, D.; Disley, L.; Hastings, D.;
Hadfield, J.; Reader, A.; Allan, D.; Julyan, P.; Watson, A.; Zweit, J.
Functional comparison of annexin V analogues labeled indirectly and
directly with iodine-124. Nucl. Med. Biol. 2005, 32, 403−413.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
Abbreviations for common amino acids are in accordance with
the recommendations of IUPAC.
ACKNOWLEDGMENTS
■
Financial assistance for this work provided by Department of
Defense Prostate Cancer Training Award (W81XWH-10-1-
0420) to M.B.K. and NCI Prostate SPORE Award
(2P50CA58236) to S.R.D. We are grateful to W. N. Brennen
for his help during the course of this work. We thank G. Green
and J. Fox for technical assistance.
ABBREVIATIONS USED
■
Ahx, 6-aminohexanoic acid; Boc, tert-butoxycarbonyl; DIC,
diisopropylcarbodiimide; DIEA, N,N-diisopropylethylamine;
DMAP, 4-(dimethylamino)pyridine; EDC, N-ethyl-N′-(3-
dimethylaminopropyl)carbodiimide; Fmoc, fluorenylmethoxy-
carbonyl; Z or Cbz, benzyloxycarbonyl protecting group for N-
terminus; HBTU, O-benzotriazole, N′,N′,N′,N′-tetramethy-
luronium hexafluorophosphate; HOBt, N-hydroxybenzotria-
zole; Mu, morpholinocarbonyl; NMP, N-methylpyrrolidone;
Bpg, 3-bromopropylglycine; Naph, (2-naphthyl)-β-homoala-
nine; aminoboronic acid or (boro)aa, CO2H of the equivalent
amino acid is replaced by B(OH)2, H is used to designate an
unprotected amino function; Pin, pinanediol protecting group;
PSA, prostate-specific antigen; KLK3, kallikrein-related pepti-
dase 3; IGFBP, insulin-like growth factor binding protein;
TGF-β2, transforming growth factor-beta 2; PTHrP, para-
thyroid hormone-related protein; A2M, α-2-macroglobulin;
ACT, α-1-antichymotrypsin; cAMP, cyclic adenosine mono-
phosphate; IDD, Institute for Diabetes Discovery
REFERENCES
■
(1) Watt, K. W.; Lee, P. J.; M’Timkulu, T.; Chan, W. P.; Loor, R.
Human prostate-specific antigen: structural and functional similarity
with serine proteases. Proc. Natl. Acad. Sci. U. S. A. 1986, 83, 3166−
3170.
(2) Lilja, H. A kallikrein-like serine protease in prostatic fluid cleaves
the predominant seminal vesicle protein. J. Clin. Invest. 1985, 76,
1899−1903.
(3) Yousef, G. M.; Diamandis, E. P. An overview of the kallikrein
gene families in humans and other species: emerging candidate tumor
markers. Clin. Biochem. 2003, 36, 443−452.
(4) Williams, S. A.; Singh, P.; Isaacs, J. T.; Denmeade, S. R. Does PSA
play a role as a promoting agent during the initiation and/or
progression of prostate cancer? Prostate 2007, 67, 312−329.
(5) Denmeade, S. R.; Sokoll, L. J.; Chan, D. W.; Khan, S. R.; Isaacs, J.
T. Concentration of enzymatically active prostate-specific antigen
(PSA) in the extracellular fluid of primary human prostate cancers and
human prostatic cancer xenograft models. Prostate 2001, 48, 1−6.
(6) Lilja, H.; Christensson, A.; Dahlen, U.; Matikainen, M. T.;
Nilsson, O.; Pettersson, K.; Lovgren, T. Prostate-specific antigen in
serum occurs predominantly in complex with a1-antichymotrypsin.
Clin. Chem. 1991, 37, 1618−1625.
(7) Stenman, U. H.; Leinonen, J.; Alfthan, H.; Rannikko, S.;
Tuhkanen, K.; Alfthan, O. A complex between prostate specific antigen
and a1-antichymotrypsin is the major form of prostate-specific antigen
in serum of patients with prostatic cancer: assay of the complex
improves clinical sensitivity for cancer. Cancer Res. 1991, 51, 222−226.
4234
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235

Journal of Medicinal Chemistry
Article
(25) Nanus, D. M.; Milowsky, M. I.; Kostakoglu, L.; Smith-Jones, P.
M.; Vallabahajosula, S.; Goldsmith, S. J.; Bander, N. H. Clinical use of
monoclonal antibody HuJ591 therapy: targeting prostate specific
membrane antigen. J. Urol. 2003, 170, S84−S88 discussion S88−S89..
(26) Smith-Jones, P. M.; Vallabahajosula, S.; Navarro, V.; Bastidas,
D.; Goldsmith, S. J.; Bander, N. H. Radiolabeled monoclonal
antibodies specific to the extracellular domain of prostate-specific
membrane antigen: preclinical studies in nude mice bearing LNCaP
human prostate tumor. J. Nucl. Med. 2003, 44, 610−617.
(27) Cohen, P.; Graves, H. C.; Peehl, D. M.; Kamarei, M.; Giudice, L.
C.; Rosenfeld, R. G. Prostate-specific antigen (PSA) is an insulin-like
growth factor binding protein-3 protease found in seminal plasma. J.
Clin. Endocrinol. Metab. 1992, 75, 1046−1053.
(28) Iwamura, M.; Hellman, J.; Cockett, A. T.; Lilja, H.; Gershagen,
S. Alteration of the hormonal bioactivity of parathyroid hormone-
related protein (PTHrP) as a result of limited proteolysis by prostate-
specific antigen. Urology 1996, 48, 317−325.
(29) Cramer, S. D.; Chen, Z.; Peehl, D, M. Prostate specific antigen
cleaves parathyroid hormone-related protein in the PTH-like domain:
inactivation of PTHrP-stimulated cAMP accumulation in mouse
osteoblasts. J. Urol. 1996, 156, 526−531.
(41) Koistinen, H.; Wohlfart, G.; Mattsson, J. M.; Wu, P.;
Lahdenpera, J.; Stenman, U. H. Novel small molecule inhibitors for
prostate-specific antigen. Prostate 2008, 68, 1143−1151.
(42) Mikolajczyk, A. D.; Marks, L. S.; Partin, A. W.; Rittenhouse, H.
G. Free prostate-specific antigen is becoming more complex. Urology
2002, 59, 797−802.
(43) Matteson, D. S.; Sadhu, K. M.; Lienhard, G. E. R-1-Acetamido-
2-phenylethaneboronic acid. A specific transition-state analog for
chymotrypsin. J. Am. Chem. Soc. 1981, 103, 5241−5242.
(44) Matteson, D. S.; Sadhu, K. M. Synthesis of 1-amino-2-
phenylethane-1-boronic acid derivatives. Organometallics 1984, 3,
614−618.
(45) Matteson, D. S.; Jesthi, P. K.; Sadhu, K. M. Synthesis and
properties of pinanediol α-amido boronic esters. Organometallics 1984,
3, 1284−1288.
(46) Hansen, K. K.; Grosch, B.; Greiveldinger-Poenaru, S.; Bartlett,
P. A. Synthesis and evaluation of macrocyclic transition state analog
inhibitors for α-chymotrypsin. J. Org. Chem. 2003, 68, 8465−8470.
(47) Wityak, J.; Earl, R., A.; Abelman, M. M.; Bethel, Y. B.; Fisher, B.
N.; Kauffman, G. S.; Kettner, C. A.; Ma, P.; McMillan, J. L.; Mersinger,
L. J.; Pesti, J.; Pierce, M. E.; Rankin, F. W.; Chorvat, R. J.; Confalone,
P. N. J. Org. Chem. 1995, 60, 3717−3722.
(48) Chen, Y.; Foss, C. A.; Byun, Y.; Nimmagadda, S.; Pullambhatla,
M.; Fox, J. J.; Castanares, M.; Lupold, S. E .; Babich, J. W.; Mease, R.
C.; Pomper, M. G. Radiohalogenated prostate-specific membrane
antigen (PSMA)-based ureas as imaging agents for prostate cancer. J.
Med. Chem. 2008, 51, 7933−7943.
(30) Dallas, S. L.; Zhao, S.; Cramer, S. D.; Chen, Z.; Peehl, D. M;
Bonewald, L. F. Preferential production of latent transforming growth
factor-2 by primary prostatic epithelial cells and its activation by
prostate-specific antigen. J. Cell. Physiol. 2005, 202, 361−370.
(31) Lilja, H.; Piironen, T. P.; Rittenhouse, H. G.; Mikolajczyk, S. D.;
Slawin, K. M. Comprehensive Textbook of Genitourinary Oncology;
Lippincott Williams and Wilkins Publishers: Philadelphia, 2000; pp
638−650.
(32) Williams, S. A.; Jelinek, C. A.; Litvinov, I.; Cotter, R. J.; Isaacs, J.
T.; Denmeade, S. R. Enzymatically active prostate-specific antigen
promotes growth of human prostate cancers. Prostate 2011, 71, 1595−
1607.
(33) O’Keefe, R. J.; Guise, T. A. Molecular mechanisms of bone
metastasis and therapeutic implications. Clin. Orthoped. Relat. Res.
2003, 415, S100−S104.
(34) Roodman, G. D. Mechanisms of bone metastasis. N. Engl. J.
Med. 2004, 350, 1655−1664.
(35) Ulmert, D.; Evans, M. J.; Holland, J. P.; Rice, S. L.; Wongvipat,
J.; Pettersson, K.; Abrahamsson, P. A.; Scardino, P. T.; Larson, S. M.;
Lilja, H.; Lewis, J. S.; Sawyers, C. L. Imaging androgen receptor
signaling with a radiotracer targeting free prostate-specific antigen.
Cancer Discovery 2012, 2, 320−327.
(36) Adlington, R. M.; Baldwin, J. E.; Becker, G. W.; Chen, B.;
Cheng, L.; Cooper, S. L.; Hermann, R. B.; Howe, T. J.; McCoull, W.;
McNulty, A. M.; Neubauer, B. L.; Pritchard, G. J. Design, synthesis,
and proposed active site binding analysis of monocyclic 2-azetidinone
inhibitors of prostate specific antigen. J. Med. Chem. 2001, 44, 1491−
1508.
(37) Adlington, R. M.; Baldwin, J. E.; Chen, B.; Cooper, S. L.;
McCoull, W.; Pritchard, G. J.; Howe, T. J.; Becker, G. W.; Hermann,
R. B.; McNulty, A. M.; Neubauer, B L. Design and synthesis of novel
monocyclic β-lactam inhibitors of prostate-specific antigen. Bioorg.
Med. Chem. Lett. 1997, 7, 1689−1694.
(38) Knight, W. B.; Green, B. G.; Chabin, R. M.; Gale, P.; Maycock,
A. L.; Weston, H.; Kuo, D. W.; Westler, W. M.; Dorn, C. P.; Finke, P.
E. Specificity, stability, and potency of monocyclic β-lactam inhibitors
of human leukocyte elastase. Biochemistry 1992, 35, 8160−8170.
(39) Green, B. G.; Chabin, R.; Mills, S.; Underwood, D. J.; Shah, S.
K.; Kuo, D.; Gale, P.; Maycock, A. L.; Liesch, J.; Burgey, C. S.;
Doherty, J. B.; Dorn, C. P.; Finke, P. E.; Hagmann, W. K.; Hale, J. J.;
MacCoss, M.; Westler, W. M.; Knight, W. B. Mechanism of inhibition
of human leucocyte elastase by β-lactams. Stability, reactivation
kinetics, and products of β-lactam-derived E−I complexes. Biochemistry
1995, 44, 14331−14243.
(40) Huang, A.; Knoell, C.; Fey, G.; Hazegh-Azam, M.; Tashjian, A.;
Hedstrom, L.; Abeles, R. H. Modulation of recombinant human
prostate-specific antigen: activation by hofmeister salts and inhibition
by azapeptides. Biochemistry 2001, 40, 11734−11741.
4235
dx.doi.org/10.1021/jm301718c | J. Med. Chem. 2013, 56, 4224−4235