Antibody-catalyzed activation of a model tripartate prodrug by a tandem hydrolysis-1,6-elimination reaction

Article abstract of DOI:10.1039/b103971g

Activation of a model tripartate prodrug was achieved by abzyme-catalyzed hydrolysis of an N-methylcarbamate moiety followed by a spontaneous 1,6-elimination reaction.

Full text of DOI:10.1039/b103971g

Antibody-catalyzed activation of a model tripartate prodrug by a
tandem hydrolysis–1,6-elimination reaction†
A. Nicole Dinaut and Scott D. Taylor*
Department of Chemistry, University of Waterloo, 200 University Avenue West, Waterloo, Ontario,
Canada, N2L 3G1. E-mail: s5taylor@sciborg.uwaterloo.ca
Received (in Corvallis, OR, USA) 26th April 2001, Accepted 7th June 2001
First published as an Advance Article on the web 11th July 2001
Activation of a model tripartate prodrug was achieved by
abzyme-catalyzed hydrolysis of an N-methylcarbamate moi-
ety followed by a spontaneous 1,6-elimination reaction.
Researchers have long been searching for ways of delivering
drugs specifically to cancer cells. One such tactic is known as
antibody-directed enzyme prodrug therapy (ADEPT).1 In this
approach, an enzyme–antibody conjugate is used to target an
enzyme to the surface of tumour cells. An inactive prodrug is
administered which is converted into the drug only by the
tumour-bound enzyme. This results in a high concentration of
the drug in the vicinity of the cancer cell while minimizing its
presence at healthy cells. One of the key requirements for a
clinically useful ADEPT system is the absence of an equivalent
endogenous enzyme in humans. Consequently, the enzyme
component has commonly been of non-human origin. However,
the immunogenicity of the non-human enzymes has severely
limited the clinical potential of ADEPT. To overcome this
problem, researchers have suggested replacing the enzyme
component with a catalytic antibody (antibody-directed abzyme
prodrug therapy—ADAPT).2a–d Abzymes can be designed to
act upon substrates that are not readily acted upon by human
Scheme 1
enzymes. In addition, antibodies can be humanized thus
reducing the serious problem of immunogenicity.3
A number of abzymes have been developed that convert
to TSA’s in which the drug was not incorporated into the TSA.5
prodrugs into drugs.2a–d In most instances, the transition state
analogues (TSA’s) to which these abzymes were raised
contained the drug or a structural analogue of the drug.2a–c Such
abzymes were designed to activate bipartate prodrugs in which
the moiety acted upon by the abzyme (known as the specifier4)
is directly attached to the drug. An alternative to using bipartate
prodrugs is to use tripartate prodrugs. In tripartate prodrugs, the
specifier is separated from the drug by a linker. After cleavage
of the specifier from the linker, the linker–drug product is
designed to spontaneously fragment to release the drug. An
example of a tripartate prodrug system is outlined in Scheme 1.
Here, specific hydrolysis of the moiety separating the specifier
and linker in 1 releases unstable intermediate 2 which
undergoes a spontaneous 1,6-elimination reaction to release the
drug. This prodrug design was first described by Katzenellenbo-
gen and coworkers.4 Prodrugs based upon this general design
have been used extensively in ADEPT and for other applica-
tions.
Several years ago, we suggested that abzymes that are
designed to activate tripartate prodrugs might be more versatile
for ADAPT than those that act upon bipartate prodrugs.5 If an
abzyme could be designed to recognize only the specifier and
linker of a tripartate prodrug, then a single abzyme could be
used to activate a variety of different prodrugs.5 This would be
advantageous both in terms of economy and for dealing with
tumours that have developed resistance to a certain drug. We
reasoned that this could be accomplished by raising antibodies
Instead, the carrier protein (necessary for antibody production)
could be attached to a position on the TSA that is equivalent to
the point of attachment of the drug to the prodrug. Since
abzymes are usually insensitive to changes in the region of the
substrate that corresponds to the part of the TSA that is attached
to the carrier protein,6 then abzymes raised to such TSA’s might
be capable of acting as generic activators of tripartate
prodrugs.
To test this approach we constructed the model tripartate
prodrug 3† (Scheme 2). In this system, drug release is triggered
by abzyme-catalyzed hydrolysis of the N-methylcarbamate
moiety to release p-nitrophenol (PNP, the specifier), CO2 and
intermediate 4. Intermediate 4 then undergoes the spontaneous
1,6-elimination reaction to produce the amino acid -tryptophan
L
(Trp) as the model ‘drug’.7 Abzymes were raised to TSA-hapten
6,8 which mimics the rate-determining transition state for the
hydrolysis of the N-methylcarbamate moiety in 3. We recently
reported that an antibody raised to 6, called ST51, is capable of
catalyzing the hydrolysis of N-methylcarbamate 7.8 Since the
carrier protein in 6 was attached to a position on the TSA that
corresponds to the point of attachment of tryptophan to 3, we
reasoned that ST51 should be capable of activating model
prodrug 3.
Model prodrug 3 was examined as a substrate for ST51 by
removing aliquots from solutions containing ST51 (5 mM) and
varying amounts of 3, in buffer at pH 9.0 (100 mM bicine,
100 mM NaCl, 5% DMSO, 25 °C), at various time intervals,
and then examining the aliquots for production of both PNP and
Trp by HPLC. Under these conditions, we were unable to detect
any PNP or Trp in the absence of ST51 after 40 hours. However,
in the presence of ST51, both of these products were readily
detectable. The antibody-catalyzed reaction obeys saturation
† Electronic supplementary information (ESI) available: the synthesis and
characterization of 3, and 9–11, and experimental details for kinetic studies
and Lineweaver–Burk plots. See http://www.rsc.org/suppdata/cc/b1/
b103971g/
1386
Chem. Commun., 2001, 1386–1387
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b10371g

a tandem hydrolysis–1,6-elimination reaction.9 We have dem-
onstrated that it is possible to obtain an abzyme that can activate
a tripartate system without incorporating the ‘drug’ into the
TSA by attaching the carrier protein to a position on the TSA
that corresponded to the point of attachment of the ‘drug’ to the
tripartate prodrug. Although ST51 can catalyze the hydrolysis
of N-methylcarbamates at physiological pH, the rate of the
reaction is too slow for an accurate kinetic analysis and effective
prodrug activation. It has been estimated that for ADEPT
systems, a kcat value of about 1.0 s21 is required.2c Nevertheless,
the approach outlined here should be readily applicable to
tripartate prodrugs of type 1 bearing moieties that are more
amenable to antibody–catalyzed hydrolysis, under physio-
logical conditions, than N-methylcarbamates, which are highly
challenging substrates for antibody catalysis.8 An N-H carba-
mate linkage between the specifier and linker, of the type
recently exploited by Blackburn and coworkers for abzyme
catalysis,2c should be very suitable to the approach reported
here.11 However, tripartate prodrugs of type 1 bearing such a
moiety cannot be used as substrates for ST51 since ST51 will
not hydrolyze N-H carbamate 8, amide 9, or even ester 10.
Studies to elucidate the mechanism of ST51 are in progress and
we are continuing our work to develop abzymes that can trigger
tripartate prodrugs of type 1 under physiological conditions.
We thank the Canadian Institutes of Health Research (CHIR)
for financial support of this work. We would also like to thank
Katherine Majewska and Parham Daneshvar for assistance in
synthesizing compounds 8–10.
Scheme 2
Notes and references
1 For recent reviews see: K. N. Syrigos and A. A. Epenetos, Anticancer
Res., 1999, 19, 605; W. A. Denny and W. R. Wilson, J. Pharm.
Pharmacol., 1998, 50, 387.
2 (a) H. Miyashita, Y. Karaki, M. Kikuchi and I. Fujii, Proc. Natl. Acad.
Sci, USA, 1993, 90, 5337; (b) D. A. Campbell, B. Gong, L. M.
Kochersperger, S. Yonkovich, M. A. Gallop and P. G. Schultz, J. Am.
Chem. Soc., 1994, 116, 2165; (c) P. Wentworth, A. Datta, D. Blakey, T.
Boyle, L. J. Partridge and G. M. Blackburn, Proc. Natl. Acad. Sci USA,
1996, 93, 799; (d) D. Shabat, C. Rader, B. List, R. A. Lerner and C. F.
Barbas III, Proc. Natl. Acad. Sci. USA, 1999, 96, 6925.
3 G. Winter and C. Milstein, Nature, 1991, 349, 293.
4 P. L. Carl, P. K. Chakavarty and J. A. Katzenellenbogen, J. Med. Chem.,
1981, 24, 479.
5 S. D. Taylor, M. J. Chen, A. N. Dinaut and R. A. Batey, Tetrahedron,
1998, 54, 4223.
6 L. J. Liotta, P. A. Benkovic, G. P. Miller and S. J. Benkovic, J. Am.
Chem. Soc., 1993, 115, 350; F. Tanaka, K. Kinoshita, R. Tanimura and
I. Fujii, J. Am. Chem. Soc., 1996, 118, 2332.
7 Since the primary objective of these studies was to determine if ST51
could trigger the tripartate system, it was not necessary to use an actual
drug to make the model ‘prodrug’. Indeed, for these studies any amine-
bearing molecule would have sufficed for the ‘drug’ so long as it (1) had
no resemblance to the specifier (PNP) portion of the prodrug, (2) was
relatively easy to detect and quantitate, (3) had a charged moiety that
would enhance the solubility of the prodrug and, (4) was readily
available and allowed for economic and facile synthesis of the model
prodrug. Tryptophan was chosen since it satisfied all of these
requirements.
kinetics, and approximately equimolar quantities of PNP and
Trp were detected for the duration of the time the reactions were
monitored. When following the production of PNP, ST51
exhibited a kcat = 0.075 h21 and a Km = 137 mM. Similar
values were obtained from data following the production of Trp.
We also found that ST51 was unable to catalyze the hydrolysis
of Z-Trp. Taken together, these results indicate that the
production of Trp is not a result of hydrolysis of the N-H
carbamate moiety, but rather initial hydrolysis of the N-
methylcarbamate followed by the spontaneous fragmentation
reaction and that the abzyme-catalyzed step is slow compared to
the fragmentation reaction. We also found that ST51 was
capable of catalyzing the activation of 3 with multiple turnover.
This indicates that 5, which is produced as an intermediate
during the reaction, does not inactivate the abzyme after a single
turnover by reacting with crucial residues in the active site.
Assuming that the spontaneous rate of hydrolysis of the N-
8 A. N. Dinaut, M.-J. Chen, A. Marks, R. A. Batey and S. D. Taylor,
Chem. Commun., 2000, 385.
9 While our work was in progress, Shabat et al. reported abzyme-
catalyzed activation of tripartate prodrugs by a tandem retro-aldol–
retro-Michael reaction (see ref. 2d). Although the abzyme was not
originally designed to act as a prodrug activator (see also ref. 10),
generic tripartate prodrug activation was achievable because of the
broad substrate specificity of the abzyme.
10 J. Wagner, R. A. Lerner and C. F. Barbas, Science, 1995, 270, 1797.
11 For tripartate prodrugs of type 1 based upon N-H carbamates of the kind
examined by Blackburn and coworkers (see ref. 2c), the specifier would
be an amino acid and atom X in Scheme 1 would be oxygen and so the
self-immolative intermediate 2, would be a phenol derivative. Phenols
also undergo the spontaneous 1,6-elimination reaction. For a recent
example of this see: I. Niculescu-Duvaz, D. Niculescu-Duvaz, F.
Friedlos, R. Spooner, J. Martin, R. Marais and C. J. Springer, J. Med.
Chem., 1999, 42, 2485.
1
methylcarbamate moiety in 3 is similar to that of 7 (t = 5.7
2
years in 100 mM bicine, 100 mM NaCl, 5% DMSO, pH 9.0)
then the rate enhancement (kcat/kuncat) with 3 is about 5000-fold.
This is only slightly less than the rate enhancement found for
ST51 using substrate 7 (6500-fold).8 It is also important to note
that 3 and 7 exhibit similar Km values (266 mM for 7 and 137 mM
for 3). These results indicate that the ‘drug’ portion of the
prodrug is not an important recognition site for the antibody-
catalyzed reaction and this is consistent with our hapten
design.
In summary, we have reported the first example of antibody-
catalyzed activation of a model tripartate prodrug of type 1 by
Chem. Commun., 2001, 1386–1387
1387