Antibody-catalyzed retro-aldol reaction

Full text of DOI:10.1021/ja954221u

6078
J. Am. Chem. Soc. 1996, 118, 6078-6079
Scheme 1
Antibody-Catalyzed Retro-Aldol Reaction
Mark E. Flanagan, John R. Jacobsen, Elizabeth Sweet, and
Peter G. Schultz*
Howard Hughes Medical Institute and
Department of Chemistry, UniVersity of California
Berkeley, California 94720
ReceiVed December 18, 1995
The aldol reaction plays a central role in cellular metabolism,1
as well as in synthetic chemistry,2 and is therefore an obvious
choice for exploiting the diversity of the immune system3,4 to
find new and selective catalysts. A number of recent examples
of antibody-catalyzed aldol reactions have appeared in which
imine intermediates are involved.5 In our efforts to develop
antibodies to catalyze the bimolecular aldol condensation
between phenylacetone (1) and benzaldehyde (2) (Scheme 1a),
we discovered an antibody that catalyzes the retro-aldol reaction
Henry type6 illustrated in Scheme 1b, which we now report.
Monoclonal antibodies were elicited against the phosphinate
transition state analog 10. This hapten was designed to resemble
the expected transition state for the addition of a phenylacetone-
derived enolate to the carbonyl of benzaldehyde. The negatively-
charged carboxylate group of hapten 10 was intended to
approximate the negatively-charged enolate while the tetrahedral
phosphinate group reflects the developing tetrahedral geometry
and charge at the carbonyl group of the benzaldehyde. Con-
sequently, antibodies generated against 10 were expected to
catalyze the aldol condensation of 1 and 2 by a combination of
both proximity and electrostatic effects.4,7
Hapten 10 was converted to the diazonium derivative by
reaction with NaNO2 and coupled to the carrier proteins keyhole
limpet hemocyanin (KLH) and bovine serum albumin (BSA)
in water (pH 11-12).8 Twenty clonal cell lines exhibiting
binding specificity for BSA-10 were prepared by standard
hybridoma technology.9 Antibodies were purified by chroma-
tography on protein A-coupled Sepharose 4B and determined
to be greater than 95% homogeneous by SDS-polyacrylamide
gel electrophoresis.10
syn
-
Scheme 2a
a Synthesis of hapten 10. Key: (a) neat, 80 °C, 38%; (b) lithium bis
(trimethylsilyl)amide, benzyl cyanoformate, HMPA/THF, 40%; (c) NaI,
acetone (reflux); (d) Pd(OH)2/C, H2 (1 atm), MeOH/H2O, 34% from
9.
formation inhibited by the addition of up to 167 µM hapten
10.12 This lack of hapten inhibition suggested that the observed
catalysis was not active site associated, but rather could be
explained by a catalytic mechanism involving nonspecific imine
formation with surface lysine residues.13
We then examined phenylnitromethane (5) as a substrate for
the antibodies since the electronic structure of the corresponding
nitronate anion more closely matches that of the carboxylate
group in hapten 10. Antibodies generated against 10 might be
expected to stabilize and thereby increase the effective concen-
tration of the anion in the active site. However, the lower pKa
of 5 also significantly shifts the equilibrium of the addition
reaction away from the condensation product. Consequently,
based on the principle of microscopic reversibility,14 the retro-
aldol reaction was studied.
Initially, antibodies were screened for catalysis of the
condensation reaction of phenylacetone (1) and benzaldehyde
(2) (Scheme 1a) under a variety of reaction conditions.11
Product formation, over background, was observed for several
antibodies. However, in no case was the observed product
(1) Marsh, J. J.; Lebherz, H. G. Trends Biochem. Sci. 1992, 17, 110.
(2) (a) Heathcock, C. H. Aldrichimica Acta 1990, 23, 99. (b) Evans, D.
A. Science 1988, 240, 420.
(3) (a) Tramontano, A.; Janda, K. D.; Lerner, R. A. Science 1986, 234,
1566. (b) Pollack, S. J.; Jacobs, J. W.; Schultz, P. G. Science 1986, 234,
1570.
(4) (a) Schultz, P. G.; Lerner, R. A. Acc. Chem. Res. 1993, 26, 391. (b)
Schultz, P. G.; Lerner R. A. Science 1995, 269, 1835.
The retro-aldol reaction of 415 (Scheme 1b) proceeded with
a measurable background rate to give 2 and 5 in 50 mM NaOAc,
50 mM NaCl, 5% [v:v] acetonitrile (pH 5.0) at 4 °C. Antibodies
were screened by monitoring production of benzaldehyde by
reverse phase high-pressure liquid chromatography (HPLC)
(C18, 50% acetonitrile in 50 mM NaOAc, pH 5.0, 260 nm
detection). A single antibody (29C5.1) was identified that
catalyzed this reaction. The syn diastereomer of 4 (illustrated
in Scheme 1b) was found to be the better substrate for antibody
(5) (a) Reymond, J.-L.; Chen, Y. Tetrahedron Lett. 1995, 36, 2575. (b)
Koch, T.; Reymond, J.-L.; Lerner, R. A. J. Am. Chem. Soc. 1995, 117,
9383. (c) Wagner, J.; Lerner, R. A.; Barbas, C. F., III Science 1995, 270,
1797.
(6) Henry, L. C. R. Acad. Sci. Paris 1895, 120, 1265.
(7) (a) Jencks, W. P. Catalysis in Chemistry and Enzymology; Dover:
Mineola, NY, 1987; p 11501. (b) Fersht, A. Enzyme Structure and
Mechanism; W. H. Freeman: New York, NY, 1985.
(8) Epitope densities of KLH and BSA conjugates were measured by
UV absorbance at 370 nm using ꢀ370 ) 22 900 for the azo linkage and
determined to be 14 for KLH and 8 for BSA.
(9) (a) Harlow, E.; Lane, D. Antibodies: A Laboratory Manual; Cold
Spring Harbor Laboratory: USA, 1988. (b) Shokat, K. M. Ph.D. Thesis,
University of California, Berkeley, CA, 1991.
(10) Fredriksson, U.-B.; Fagerstam, L. G.; Cole, A. W. G.; Lundgren,
T. Protein A-Sepharose Cl-4B Affinity Purification of IgG Monoclonal
Antibodies from Mouse Ascites; Pharmacia AB: Uppsala, Sweden, 1986.
(11) Antibodies were assayed for catalysis of the aldol condensation using
reaction mixtures of benzaldehyde and phenylacetone at pH 8.0 and 9.0.
4-Fluorophenylacetone as well as the ketone containing an ortho azo linkage
to p-cresol was assayed in pH 9.0 buffer.
(12) A derivative of hapten 10 was generated by diazotization and
coupling to p-cresol. This analog was expected to have greater structural
homology to the hapten-protein conjugate used for immunization and
screening. No inhibition of catalysis was observed at concentrations up to
500 µM.
(13) (a) Wittig, G.; Reiff, H. Angew. Chem., Int. Ed. Engl. 1968, 7, 7.
(b) Wittig, G.; Suchanek, P. Tetrahedron Suppl. 1967, 8, 347.
(14) Alberty, W. J. AdV. Phys. Org. Chem. 1993, 28, 139.
(15) Compound 5 was prepared by treatment of benzylbromide with
AgNO2 (ether, room temperature, 24 h, 47%). Compound 4 was prepared
from 2 and 5 by treatment with alumina, basic Brockman activity 1 (neat,
12 h, 42%). The syn diastereomer of 4 was the faster moving on silica (5:1
hexanes/ethylacetate, Rf ) 0.25).
S0002-7863(95)04221-1 CCC: $12.00 © 1996 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 118, No. 25, 1996 6079
I - Ki′) + [(Et + I + Ki′)2 - 4EtI]1/2}, where Vi is the inhibited
rate, Vo is the uninhibited rate, Et is the total enzyme concentra-
tion, and I is the total inhibitor concentration.18,19
Chemical modification20 of 29C5.1 with diethyl pyrocarbonate
resulted in an ∼4-fold reduction in catalytic activity (no
inactivation was observed when the modification was carried
out in the presence of 1 mM 10). These results are consistent
with the presence of an active site histidine. The rate of the
antibody-catalyzed reaction can be compared with the rate of
the reaction catalyzed by imidazole (kimid ) 2.5 × 10-4 M-1
min-1). This affords a ratio of second-order rate constants (kcat/
KM/kimid) of 5.0 × 105 indicating the antibody to be an effective
catalyst.21
An effort was made to alter the equilibrium of the reaction
to favor formation of the condensation product (4) by reacting
2 and 5 in isooctane in the presence of purified Fab (29C5.1).22
Unfortunately, these conditions also failed to produce a measur-
able rate of formation of 4 for both the background and antibody
(Fab)-catalyzed reactions. In conclusion, these results demon-
strate that antibodies can stabilize the aldol transition state but
point to the need for improved strategies for enolate formation
under aqueous conditions.
Figure 1. Velocity (V) vs substrate concentration [syn-4] plot for
antibody-catalyzed retro-aldol reaction (Scheme 1b). Initial velocities
were measured by monitoring the rate of formation of 2 by HPLC (UV,
260 nm).
catalysis by 2:1 over the anti diastereomer (250 µM substrate
and 7 µM 29C5.1).16,17
Initial reaction rates were determined for the syn diastereomer
of 4 at concentrations ranging from 50 µM to 1.0 mM in the
presence of 3.0 µM IgG (6.0 µM combining site).18 Analysis
of the reaction rate (V) for antibody catalysis over background
as a function of substrate concentration (Figure 1) afforded an
apparent second-order rate constant (kcat/KM) of 125 M-1 min-1
(the limited solubility of syn-4 under the reaction conditions
precluded determination of kcat and KM). The inhibition constant
Ki,app for hapten 10 was determined to be 2.6 µM by curve-
fitting data for Vi/Vo as a function of inhibitor (10) concentration
to the equation for competitive inhibition: Vi/Vo ) 1/2Et{(Et -
Acknowledgment. We are grateful for financial support for this
work from the National Institutes of Health (Grant No. R01 AI24695).
P.G.S. is a Howard Hughes Medical Institute Investigator. M.E.F.
acknowledges a fellowship from the National Institutes of Health (Grant
No. 1F32AI09093-01A1), and J.R.J. acknowledges an NIH Molecular
Biophysics Training Grant.
JA954221U
(19) Reactions were carried out in 50 mM NaOAc, 50 mM NaCl, 5%
acetonitrile, pH 5.0 with 50 µM syn-4 and 3 µM 29C5.1, 4 °C.
(20) Holbrook, J. J.; Ingram, V. A. Biochem. 1973, 131, 729.
(21) A study of pH effects on antibody catalysis revealed that the rates
of the background and antibody-catalyzed reactions increased linearly as a
function of hydroxide ion concentration from pH 4.5 to 6.5. Above pH 6.5
the fast background rate of the retro-aldol reaction complicated analysis
(kuncat ) 7.4 × 10-4 min-1 at pH 5.0).
(16) The relative stereochemical configuration of syn-4 was determined
by single-crystal X-ray analysis.
(17) No evidence of enantioselectivity was observed for catalysis by
monitoring the relative consumption of the enantiomers of syn-4 by chiral
HPLC (Chiralcel OD-R column, 30-70% acetonitrile in 50 mM NaOAc,
pH 5.0).
(22) Enzyme digests for Fab preparation used immobilized Papain
(Pierce). Isooctane reactions were conducted in the presence of 1.0% Aerosol
OT (Sigma) as described by Paradkar, V. M.; Dordick, J. S. J. Am. Chem.
Soc. 1994, 116, 5009.
(18) (a) Cha, S. Biochem. Pharmacol. 1975, 24, 2177. (b) Angeles, T.
S. et al. Biochemistry 1993, 32, 12128.