Image contrast agents activated by prostate specific antigen (PSA)

Article abstract of DOI:10.1016/j.bmcl.2004.04.041

A family of image contrast agent conjugates designed to undergo enzymatic activation has been synthesized. The agents underwent activation both with enzymatically active prostate specific antigen and α-chymotrypsin, releasing free fluorophore via cleavage

Full text of DOI:10.1016/j.bmcl.2004.04.041

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3081–3084
Image contrast agents activated by prostate specific antigen (PSA)
Graham B. Jones,a,* Longfei Xie,a Ahmed El-Shafey,a Curtis F. Crasto,a
Glenn J. Bubleyb and Anthony V. D’Amicoc
aBioorganic and Medicinal Chemistry Laboratories, Department of Chemistry and Chemical Biology,
Northeastern University, Boston, MA 02115, USA
bBeth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, USA
cDana Farber Cancer Institute and Brigham and Womens Hospital, Harvard Medical School, Boston, MA 02215, USA
Received 4 March 2004; accepted 10 April 2004
Abstract—A family of image contrast agent conjugates designed to undergo enzymatic activation has been synthesized. The agents
underwent activation both with enzymatically active prostate specific antigen and a-chymotrypsin, releasing free fluorophore via
cleavage of a three-component system.
Ó 2004 Elsevier Ltd. All rights reserved.
Despite improvements in local therapy and increased
awareness, prostate cancer continues to be second only
to lung cancer as a cause for cancer deaths in men.1
Prior investigations show that the presence of prosta-
tectomy Gleason grade P4 in the radical prostatec-
tomy specimen is the most important predictor of
progression following surgery.2 Unfortunately, the
transrectal ultrasound guided sextant sampling of the
prostate is subject to sampling error, and therefore
biopsy Gleason grade will underestimate prostatectomy
Gleason grade 4 or 5 disease in as many as 40% of men
with clinically localized disease.3 Therefore, an imaging
method capable of identifying Gleason grade P4 dis-
ease within the prostate gland could provide the basis
for patient selection for more aggressive initial thera-
peutic approaches.4 A number of image contrast
enhancing agents have been studied for use in conjunc-
tion with ultrasound methods of detection.5 However,
immunohistochemical studies have also shown that
Gleason grade bears an inverse correlation with the
concentration of enzymatically active prostate specific
antigen (PSA).6 PSA is a serine protease; however, PSA
in serum (but not in the prostatic microenvironment) is
rapidly inactivated by binding to serum proteins.7 An
attractive possibility, therefore, would be the design of
an imaging system, which exploits the enzymatic effi-
ciency of PSA in the prostatic microenvironment. Our
strategy was to conjugate a proteinogenic PSA substrate
to a masked fluorophore via an inert spacer/linker
group, such that the free fluorescent molecule is liber-
ated on proteolysis (Scheme 1).8
Our preferred choice for the inert linker is the p-amino-
benzyl alcohol pioneered by Katzenellenbogen,9 having
previously employed this method for enzyme mediated
cytotoxin release.10 Though a number of high-affinity
peptide substrates for PSA have been identified, we
initially wished to provide proof-of-principle with a
minimal substrate and selected tyrosine conjugates for
examination of appropriate fluorophores (Scheme 2).11
PSA
Cleavage Site
spontaneous
release
LINKER
peptide sequence
(recognized by PSA)
LINKER
fluorophore
LINKER
fluorophore
fluorophore
Scheme 1. Three-component system for PSA activated image contrast agent.
* Corresponding author. Tel.: +1-617-373-8619; fax: +1-617-373-8795; e-mail: gr.jones@neu.edu
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.04.041

3082
G. B. Jones et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3081–3084
Scheme 2. PSA mediated release of Tyr-linker conjugates.
NH
O
O
2
CH
O
3
COOH
H C
3
H N
2
O
H N
2
O
NH
7-AMC
Rhodamine 110
disperse orange 11
For our initial studies three readily available fluoro-
phore dyes were selected––aminomethyl coumarin (7-
AMC), disperse orange 11, and rhodamine 110.
With the alloc route in hand, the coumarin analog 10
was next prepared. This involved coupling of the car-
bonate used in Scheme 4 (7) with isocyanate 8 to give the
masked analog 9, which underwent clean deprotection
to give 10 on workup (Scheme 5). Finally, hoping to
exploit the benefits of intramolecular hydrogen bonding,
the anthraquinone conjugate 14 was assembled. This
necessitated selective removal of the quinone carbonyl
group of aminomethylanthraquinone to allow forma-
tion of the required carbamoyl building block 11
(Scheme 6). Conversion to the p-nitrophenyl carbamate
was inefficient, giving a complex mixture, which allowed
only low recovered yields of 12. However, reaction with
phosgene followed by coupling with 7 gave alloc pro-
tected adduct 13 directly, reoxidation taking place dur-
ing workup. This compares favorably with the
corresponding conversion of 12 to 13 and proved reli-
able on scale-up. Finally, unmasking allowed isolation
Commencing with commercially available Boc-tyrosine,
carbodiimide coupling with p-aminobenzyl alcohol,
followed by reaction with p-nitrophenylchloroformate
gave carbonate 2 without incident (Scheme 3). Coupling
with free rhodamine gave 3 cleanly, albeit in low yield.
However, all attempts to unmask the carbamate group
resulted in decomposition of the molecule, rendering 4
unisolable. Remedy was found using the alternate bis-
alloc substrate 5, which under analogous conditions
gave the intermediate carbonate, and subsequently
underwent rhodamine coupling and unmasking using
the Pd route,12 to allow isolation of the hydrochloride
salt 6 in good yield, and which was soluble in assay
buffer media (Scheme 4).
OH
OH
H N
2
NO
1. EDCI / HOBT
THF, rt, 20h
2
O
OH
O
79%
O
O
O
OH
2.
N
O N
2
BocNH
O
H
BocNH
1
2
O
Cl
72%
Py, THF, rt
rhodamine 110
Pyrdine, 80 C, 48h
o
20%
OH
COOH
OH
COOH
NH
TFA
O
O
O
O
O
N
NH
H
O
O
N
O
H
N
H
4
N
BocNH
3
H
NH
2
Scheme 3. Initial route to tyrosyl rhodamine conjugate via p-aminobenzyl carbamate linker.

G. B. Jones et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3081–3084
3083
EDCI / HOBT THF, rt, 20h
p-aminobenzyl alcohol, 70%
OAlloc
1.
2.
COOH
OH
p-nitrophenylchloroformate
O
Py, THF, rt 99%
O
O
O
N
O
NH
o
3. rhodamine 110, Py, 80 C, 48h, 29%
H
OH
N
H
AllocNH
4. Pd(PPh ) , morpholine
3 4
5
+
-
Cl
NH
6
then HCl, MeOH, 75%
3
Scheme 4. Y-alloc route to tyrosyl rhodamine conjugate.
CH3
O
CH3
O
OAlloc
OAlloc
8
O
OCN
PhCH3
O
O
O
N
O
O
OH
H
∆, 2h, 98%
N
H
N
H
AllocNH
OH
9
AllocNH
7
CH3
O
1. Pd(PPh3)4, morpholine,
DMF-THF
O
O
O
10
N
O
H
2. HCl, MeOH
99%
N
+
Cl-
H
NH3
Scheme 5. Preparation of tyrosyl aminomethylcoumarin conjugate.
O
O
O
NH2
4-NP
NH2
HN
O
4-nitrophenyl
chloroformate
CH3
CH3
Na2S2O4
CH3
DMF, H2O
DMAP, THF 10%
70%
11
12
O
O
1. ClCOCl
DMAP, THF
2. 7, DMAP
7, DMAP
THF
OAlloc
THF 35%
31%
OH
O
Pd(PPh3)4, morpholine
O
O
O
NH O
HCl, MeOH
90%
H3C
O
O
H3C
NH
O
O
N
H
AllocNH
N
+
H
Cl-
NH3
O
14
13
Scheme 6. Preparation of tyrosyl aminomethylanthraquinone conjugate.
of the hydrochloride salt of anthraquinone substrate 14
in good yield.
can be seen (Table 1), though proof-of-concept is estab-
lished, in the present examples, chymotrypsin is more
effective than PSA at cleavage. Though this is unsur-
prising, more complex oligopeptide substrates are
known whose specificity for PSA outranks chymotryp-
sin significantly, the most selective of these (HSSKLQ),
which will now become the target of future synthetic
studies and kinetic analysis.8 Additionally, linker
architecture has been shown to have a marked impact
on substrate half-life in three-component systems,9;12
suggesting that specificity and stability might ultimately
be tailored according to desired application. Though the
With three substrates in hand, spectroscopic and enzy-
matic studies were conducted to establish proof of
concept for use as image contrast agents. Enzymatic
release of the fluorophores was initially probed using
fresh, enzymatically active PSA and chymotrypsin,
using UVdetection to quantitate (and fluorescence in
the case of 10). Release of fluorophore correlated with
release of p-aminobenzyl alcohol and tyrosine, con-
firming the function of the self-immolative linker. As

3084
G. B. Jones et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3081–3084
Table 1. Enzyme mediated release of chromophoresa
Entry
Substrate
UV kmax conjugate
UV kmax free
Chymotrypsinb
PSAb
1
2
3
6
10
14
300
328c
285
497
352d
486
9
24
18
4
10
6
a In duplicate, PSA (Cortex Biochem), 20 lL, 0.5 mg/mL, pH 7.4 0.01 M phosphate buffer saline [PBS]) or a-chymotrypsin (Sigma, TLCK-treated,
20 lL, 0.5 mg/mL, pH 7.4 0.01 M PBS), 160 lL PBS, and 20 lL substrate (1.0 mg/mL; 1:1, EtOH:H2O) were incubated for 24 h at 37 °C. Duplicate
control reactions containing PBS (180 lL) and substrate (20 lL) solutions were incubated under identical conditions. p-Aminobenzyl alcohol was
quantified against authentic standards by HPLC (C18 lBondpak, 1 mL/min, 100% iPrOH, tR ¼ 8:2 min). Specific activity was determined on the
basis of mM released fluorophore per unit time per unit mass of enzyme.
b mM/h/mg fluorophore released.
c Fluorescence emission kmax 397.
d Fluorescence emission kmax 435.
2. Bagshaw, M. A.; Cox, R. S.; Hancock, S. L. J. Urol. 1994,
changes in the kmax range between free and bound con-
152, 1781.
3. Bostwick, D. G. Am. J. Surg. Path. 1994, 18, 796.
4. D’Amico, A. V.; Debruyne, F.; Huland, H.; Richie, J. P.
trast agents prepared are pronounced, differences in
fluorescent characteristics will be of more importance
for imaging purposes where differences in quantum yield
The Prostate 1999, 41, 208.
might be exploited.13
5. Watanabe, M. Nippon Rinsho 1998, 56, 1040; Ragde, H.;
Kenny, G. M.; Murphy, G. P. Prostate 1977, 32, 279;
Moreover, for application with in vitro and in vivo
Bogers, H. A.; Sedelaar, J. P.; Beerlage, H. P. Urology
1999, 54, 97.
6. Nadji, M.; Tabei, S. Z.; Castro, A.; Chu, T. M.; Murphy,
G. P.; Wang, M. C.; Morales, A. R. Cancer 1981, 48, 1229.
7. Lilja, H.; Abrahamsson, P.; Lundwall, P.-A. J. Biol.
Chem. 1989, 264, 1894.
8. Denmeade, S. R.; Nagy, A.; Gao, J.; Lilja, H.; Schally,
A. V.; Isaacs, J. T. Cancer Res. 1998, 58, 2537.
9. Carl, P. L.; Chakravarty, P. K.; Katzenellenbogen, J. A.
J. Med. Chem. 1981, 24, 479.
analysis it will be necessary to employ fluorophores with
spectral characteristics tailored to match imaging de-
vices. Contrast agents in the near IR range (e.g., the Cy
dye family) may prove desirable,14 in that the conjugated
amino function has a profound influence on its quantum
yield.15 The coupling chemistries described herein for
amino substituted fluorophores offer flexibility towards
this goal, providing the potential for in situ CCD based
near-IR imaging of systemic agents that are locally
activated under in vivo conditions.14
10. Jones, G. B.; Mitchell, M. O.; Weinberg, J. S.; D’Amico,
A. V.; Bubley, G. A. Bioorg. Med. Chem. Lett. 2000, 10,
1987.
11. Satisfactory spectroscopic (1H, 13C, MS) and analytical
data was obtained for all new compounds and fluorophore
In summary, a three-component system comprised of
enzyme substrate, inert linker and fluorophore has been
designed and activation by chymotrypsin and PSA
demonstrated. The results support the synthesis and in
vitro evaluation of more complex and selective sub-
strates, which will be reported in due course.
1
homogeneity confirmed by HPLC analysis; 10 H NMR
(CD3OD, 500 MHz) d 7.57 (s, 1H), 7.58–7.50 (m, 3H),
7.39–7.35 (m, 3H), 7.03 (d, J ¼ 8:0 Hz, 2H), 6.72 (d,
J ¼ 8:0 Hz, 2H), 6.15 (br s, 1H), 5.14 (s, 2H), 4.69 (br s,
5H, exch), 3.60 (dd, J ¼ 5:5, 8.0 Hz, 1H), 3.02 (dd,
J ¼ 5:5, 14.0 Hz), 2.75 (dd, J ¼ 5:5, 14.0 Hz, 1H), 2.42
(s, 3H); 13C NMR (CD3OD, 75 MHz) d 169.980, 156.104,
153.860, 153.493, 153.383, 149.638, 142.427, 137.266,
134.171, 132.037, 131.804, 130.632, 130.029, 128.502,
124.925, 120.658, 119.986, 118.898, 117.150, 114.645,
114.519, 111.737, 105.286, 68.824, 56.246, 37.570, 17.925;
RMS calcd for C27H26N3O6 (MHþ) m=z 488.1822, found
488.1250.
Acknowledgements
We thank the Department of Defense (PC-010211) and
the Prostate Cancer Foundation for financial support of
this work.
12. de Groot, F. M. H.; Loos, W. J.; Koekkoek, R.; van
Berkom, L. W. A.; Busscher, G. F.; Seelen, A. E.;
Albrecht, C.; de Brujin, P.; Scheeren, H. W. J. Org.
Chem. 2001, 66, 8815.
13. Hebden, J. C.; Delpy, D. T. Brit. J. Radiol. 1997, 70, S206.
14. Weissleder, R.; Tung, C. H.; Mahmood, U.; Bogdanov,
A., Jr. Nature Biotechnol. 1999, 17, 375.
References and notes
1. Parker, S. L.; Tong, T.; Bolden, S. Ca. Cancer J. Clin.
1996, 46, 5.
15. Mujumdar, R. B.; Ernst, L. A.; Mujumdar, S. R.; Lewis,
C. J.; Waggoner, A. S. Bioconjugate Chem. 1993, 4, 105.