Antibody catalyzed modification of amino acids. Efficient hydrolysis of tyrosine benzoate

Article abstract of DOI:10.1039/b100450f

Esterase antibody 522c2, the first example of a catalytic antibody specifically programmed to control the reactivity of functional groups on the side chain of tyrosine, accelerates the hydrolysis of benzoate esters of L-tyrosine and tyrosine-containing dipeptides by a factor of 10⁴ and is moderately active against other benzoate esters.

Full text of DOI:10.1039/b100450f

View Article Online / Journal Homepage / Table of Contents for this issue
Antibody catalyzed modification of amino acids. Efficient hydrolysis of
tyrosine benzoate†
Fabio Benedetti,*a Federico Berti,*a Alfonso Colombatti,b Massimiliano Flego,a Lucia Gardossi,c Paolo Lindac
and Silvia Peressinia
a Dipartimento di Scienze Chimiche, Università di Trieste, Via Giorgieri 1, I-34127 Trieste, Italy.
E-mail: berti@dsch.univ.trieste.it; Fax: + 39 40 6763903
b Dipartimento di Scienze e Tecnologie Biomediche, Università di Udine, and CRO – IRCSS, Aviano, P.le Kolbe 4,
I-33100, Udine, Italy
c
Dipartimento di Scienze Farmaceutiche, Università di Trieste, P.le Europa 1, I-34127 Trieste, Italy
Received (in Cambridge, UK) 11th January 2001, Accepted 8th March 2001
First published as an Advance Article on the web 29th March 2001
Esterase antibody 522c2, the first example of a catalytic
antibody specifically programmed to control the reactivity of
functional groups on the side chain of tyrosine, accelerates
the hydrolysis of benzoate esters of -tyrosine and tyrosine-
L
containing dipeptides by a factor of 104 and is moderately
active against other benzoate esters.
The control of reactivity of functional groups on the side chains
of amino acids and peptides is a potentially attractive field for
applications of catalytic antibodies.1 The selective modification
of amino acid side chains in peptides and proteins is important
for many areas of chemistry, from organic synthesis to the
irreversible inhibition of enzymes and the improvement of
pharmacokinetic properties of peptide drugs.2 However, the
high level of control over chemo- and regioselectivity which is
required for these transformations is not easily obtained by
chemical or enzymatic approaches.3
While a number of antibody catalyzed reactions taking place
on amino acid substrates have been described, there are no
examples to date dealing specifically with the reactivity of
amino acid side-chains. In this communication we describe the
activity of esterase antibody 522C2, the first catalytic antibody
specifically designed to hydrolyze an ester group on the tyrosine
phenolic side chain. As benzoate is a common modifier and
protecting group in amino acid and peptide chemistry, we chose
the hydrolysis of tyrosine benzoate 1 as the target reaction
(Scheme 1).
Scheme 1
fusion of spleen cells were selected on the basis of the affinity
of the antibodies for an ovalbumin (OVA) conjugate of 2 and
cloned. Only three good binders for the hapten 2 were obtained,
but two of the selected antibodies (517A41 and 522C2)
exhibited catalytic activity for the hydrolysis of benzoate 1 after
purification by protein G and ion exchange chromatography.†
Antibody 517A41 was only a weak catalyst (kcat/k0 = 148),
while 522C2 accelerated the hydrolysis of 1 by a factor of over
104 and was selected for further studies (Table 1). After the
preliminary kinetic assay, carried out on the antibody isolated
from hybridoma supernatants, 522C2 was subcloned and larger
quantities of the antibody were obtained from ascitic fluids. A
fully functional Fab fragment was also obtained by papain
digestion and purification by gel exclusion and ion exchange
chromatography. The observed catalytic activity was always
reproducibile and did not depend on the antibody’s prepara-
tion.
The kinetic parameters for the catalyzed hydrolysis of ester 1
were calculated at pH 8 in TRIS buffer at 25 °C (kcat = 0.063),
and at pH 7.5 in PBS at 30 °C (kcat = 0.017), showing in both
cases a rate enhancement factor in excess of 104 (Table 1).† This
allows the observation of a net enhancement of more than
150-fold in the initial rates of hydrolysis under typical
experimental conditions (5 mM antibody, 50 mM substrate 1), in
spite of the rather high value of KM (370–500 mM), which does
not allow the maximum rate to be reached within the solubility
limit of this substrate.
Hapten 2, a classical phosphonate transition state analog, was
synthesized by a conventional approach,4 via the corresponding
monoester monochloride.† The carboxylate group of tyrosine
The antibody is inhibited by an equimolar amount of the
phosphonate 2. The apparent dissociation constant of the
antibody–hapten complex obtained by a competitive ELISA
assay6 (Kd,app = 75 nM) is in reasonable agreement with the
upper limit for the inhibition constant (Ki ≤ 100 nM) estimated
from kinetic measurements. Inhibition by the hapten demon-
strates that the catalyzed process takes place in the antibody
combining site and is not due to interactions between the
substrate and the protein surface. Nonspecific interactions of
this type can indeed be observed when the hydrolysis of 1 is
carried out in the presence of BSA or unrelated mouse IgGs, but
result in a modest acceleration by a factor of less than 1.5 in the
initial rates (at 5 mM protein and 50 mM substrate 1).
was used to conjugate directly the phosphonate to cationised
bovine serum albumin (cBSA)†. We reasoned that holding the
hapten in close contact with the carrier protein surface would
result in a relatively ‘open’ antigen binding site, able to
recognize and hydrolize the substrate also when tyrosine is part
of a peptide chain. For this reason the use of a linker between the
hapten and the carrier was avoided.
The cBSA conjugate was used to immunize three Balb/c mice
following a standard protocol.5 Hybridomas obtained from the
† Electronic supplementary information (ESI) available: the synthesis and
characterization of compound 2, preparation of the immunogenic conjugate,
preparation and purification of monoclonal antibodies, experimental details
for kinetic studies and Lineweaver-Burk plots. See http://www.rsc.org/
suppdata/cc/b1/b100450f/
For an antibody accelerating a reaction by simple transition
state complementarity, it has been derived that KM/Ki = kcat/k0.7
In the case of 522C2, the values of kcat/k0 (11 000) and KM/
DOI: 10.1039/b100450f
Chem. Commun., 2001, 715–716
This journal is © The Royal Society of Chemistry 2001
715

View Article Online
Table 1 Kinetic parameters for the 522C2 catalyzed hydrolysis of tyrosine benzoate and other esters
KM
(mmol dm23
kcat/KM
(dm3 s21 mol21
k
cat/KMk0
Substrate
k0/s21
kcat/s21
)
kcat/k0
)
(dm3 mol21
)
(S)-1a
(S)-1b
(S)-1b,c
3ab
5.06 3 1026
1.53 3 1026
1.53 3 1026
8.16 3 1026
1.04 3 1024
4.49 3 1025
1.94 3 1024
2.90 3 1026
5.70 3 1028
8.21 3 1028
0.063
0.017
0.020
500
370
410
12 400
11 100
13 100
126
45.9
48.8
44.9
35.4
8.5
—
0.39
1.2
2.5 3 107
3.0 3 107
3.2 3 107
5.5 3 106
3.4 3 105
1.9 3 105
—
1.4 3 106
2.1 3 107
1.4 3 107
3bb
3cb
3db
5b
6d
5.5 3 1024
7.5 3 1024
460
650
9 700
9 100
7d
1.2
a TRIS 10 mM, pH 8, dioxane 10%, 25 °C. b PBS, pH 7.5, 10% dioxane, 30 °C. c Fab fragment. d PBS, pH 7.5, 30% DMSO, 30 °C.
Kd,app (4900) indicate that selective recognition of the transition
state significantly contributes to the catalytic activity. Thus
522C2 behaves similarly to other ester hydrolyzing antibodies
for which oxyanion stabilization has been recognized as the
main source of catalytic activity.8
It is noteworthy that the substrate selectivity exhibited by
522C2 is reversed with respect to that displayed by a hydrolytic
enzyme such as a-chymotrypsin. This protease is known to
hydrolyze esters with a broad specificity. However, under
similar conditions, the acceleration in the hydrolysis of p-
nitrophenyl ester 3a is six times higher than the acceleration in
the hydrolysis of tyrosine ester 1. Another major difference
between a-chymotrypsin and 522C2 is representend by the pre-
steady state burst which is observed with the enzyme but not
with 522C2, in agreement with the mechanism proposed for the
antibody.
Finally, the activity of 522C2 was also tested on the
dipeptides 6 and 7 (Table 1). 522C2 retains most of its catalytic
activity on both the dipeptide benzoates even in the presence of
30% DMSO, which is necessary for solubility. Replacement in
7 of the N-Cbz protecting group with the somewhat isosteric
residue of phenylalanine does not lead to a loss of activity.
In conclusion, 522C2 is an efficient and selective catalyst for
the hydrolysis of (S)-tyrosine benzoate and retains its activity
when the substrate is part of a simple dipeptide. This
preliminary result is promising in view of a possible extension
of this approach to the selective deprotection of tyrosine
benzoate in more complex peptide structures, which would be a
valuable application and is currently being investigated.
We are grateful to CNR (Progetto Finalizzato Biotecnologie)
and to MURST (Programma PRIN) for financial support.
Antibody 522C2 is highly specific for the S-enantiomer of
tyrosine benzoate, mirroring the configuration of the hapten,
and no acceleration over background is observed in the
hydrolysis of the corresponding R-ester. While displaying such
a high enantiospecificity, antibody 522C2 is able to hydrolyze a
number of simplified esters 3 in which tyrosine is replaced by p-
nitrophenol. Solubility, and the high values of KM for these
substrates, restrict the accessible substrate concentrations to a
range in which [S] < < KM. Therefore, second order rate
constants, corresponding to kcat/KM ratios assuming Michaelis–
Menten behaviour, were obtained for this set of esters and the
ratio kcat/KM·k0 (catalytic proficiency) was chosen to compare
the efficiency of the antibody on these substrates (Table 1).9
Data in Table 1 show that one order of magnitude in the
catalytic proficiency is lost on going from N-Cbz-tyrosine
benzoate 1 to p-nitrophenyl benzoate 3a. This decrease in the
catalytic activity of the antibody is paralleled by its affinity for
the corresponding phosphonates. In fact, while the tyrosine
phosphonate 2, as we have seen, is a strong binder, the simple
phenyl phenylphosphonate 4 does not inhibit 522C2. This
Notes and references
1 D. Hilvert, Annu. Rev. Biochem., 2000, 69, 751; G. M. Blackburn and A.
Garçon, in Biotechnology, ed. D. R. Kelly, Wiley-VCH, Weinheim, 2nd
edn., vol. 8b, 2000, p. 403; J. D. Stevenson and N. R. Thomas, Nat. Prod.
Rep., 2000, 17, 535; J. L. Reymond, Top. Curr. Chem., 1999, 200, 59;
G. M. Blackburn, A. Datta, H. Denham and P. Wentworth Jr., Adv. Phys.
Org. Chem., 1998, 31, 249.
2 S. P. Vyas, P. Venugopalan, A. Sood and N. Mysore, Pharmazie, 1997,
52, 339; D. K. Clodfelter, A. H. Pekar, D. M. Rebhun, K. A. Destrampe,
H. A. Havel, S. R. Myers and M. L. Brader, Pharmacol. Res., 1998, 15,
254; M. Baudys, D. Letourneur, F. Liu, D. Mix, J. Jozefonvicz and S. W.
Kim, Bioconjugate Chem., 1998, 9, 176.
3 B. T. Miller, Biochem. Biophys. Res. Commun., 1996, 218, 377; R. D.
Lundbald, Techniques in Protein Modification, CRC Press, Boca Raton,
1995; A. W. Yem, H. A. Zurcher-Neely, K. A. Richard, N. D. Staite, R. L.
Heinrikson and M. R. Deibel, J. Biol. Chem., 1989, 264, 17691; L.
Gardossi, D. Bianchi and A. Klibanov, J. Am. Chem. Soc., 1991, 113,
6328.
4 M. F. Hersman and L. F. Audrieth, J. Org. Chem., 1958, 23, 1889.
5 E. Harlow and D. Lane, Antibodies, a Laboratory Manual, Cold Spring
Harbor Laboratory, 1988, p. 139.
6 B. Friguet, A. F. Chaffotte, L. Djavadi-Ohaniance and M. E. Goldberg,
J. Immunol. Methods, 1985, 77, 305.
7 J. D. Stewart and S. J. Benkovic, Nature, 1995, 375, 388.
8 J. L. Buchbinder, R. C. Stephenson, T. S. Scanlan and R. J. Fletterick,
J. Mol. Biol., 1998, 282, 1033; J. B. Charbonnier, B. Gigant, B. Golinelli-
Pimpaneau and M. Knossow, Biochimie, 1997, 79, 653; G. MacBeath and
D. Hilvert, Chemistry & Biology, 1996, 3, 433.
9 A. R. Radzicka and R. A. Wolfenden, Science, 1995, 267, 90; C. L. Lo,
P. Wentworth, K. W. Jung, J. Yoon, J. A. Ashley and K. D. Janda, J. Am.
Chem. Soc., 1997, 119, 10 251.
indicates that in the substrate 1 and hapten 2 the region of the
molecule corresponding to the protected a-amino acid group
plays a significant role in recognition by the antibody.
Replacing the benzoyl group by aliphatic residues, as in the
hydrolysis of esters 3b and 3c, results in a further decrease of
catalytic proficiency by one order of magnitude (Table 1). The
catalytic activity is completely lost when branching is in-
troduced in the ester 3d. Surprisingly however, antibody 522C2
shows good efficiency in the hydrolysis of benzyl p-nitro-
benzoate 5.
716
Chem. Commun., 2001, 715–716