S ge

12. S. Sen, P. Brahma, S. K. Roy, D. K. Mukherjee, S. B. 18. P. Y. Yan and I.-C. Khoo, IEEEJ. Quantum Electron. the 34 mW/cm2 as the correct value of /1 for the
Roy, Mol. Cryst. Liq. Cryst. 100, 327 (1983). 25, 520 (1989). beam coupling calculations. Subsequently, with both
13. G. P. Wiederrecht, W. A. Svec, M. P. Niemczyk, M. 19. The value for /1 in this experiment is given by the beams incident on the sample, beam 1 was found to
R. Wasielewski, J. Phys. Chem. 99, 8918 (1995). relation increase in intensity to 64 mW/cm2.
14. M. R. Wasielewski, R. L. Smith, A. G. Kostka, J. Am. /1 = /00 - IT') (2) 20. A. G. Chen and D. J. Brady, Opt. Lett. 17, 441
Chem. Soc. 102, 6923 (1980). where /l is the intensity in beam 1 after the sample in (1992).
15. J. C. Scott, L. Th. Pautmeier, W. E. Moerner, J. Opt. the absence of diffraction and IT is the total diffrac- 21. We gratefully acknowledge support from the Office
Soc. Am. B 9, 2059 (1992). tion efficiency of the grating. For the case of an ap- of Computational and Technological Research, Divi-
16. H. J. Eichler, D. Gunter, D. W. Pohl, Laser-Induced plied electric field of 0.4 kV/cm, the value for /1 was sion of Advanced Energy Projects, U.S. Department
Dynamic Gratings (Springer-Verlag, Berlin, 1986). 68% (34 mW/cm2) of the value of /l without the of Energy, under contract W-31-109-ENG-38.
17. Q. Wang, R. M. Brubaker, D. D. Nolte, M. R. Melloch, electric field. In other words, 32% of the total power
J. Opt. Soc. Am. B 9,1626 (1992). lies in the diffracted spots. It is reasonable to consider 18 July 1995; accepted 18 October 1995

Efficient Aldolase Catalytic AntibodiesUseThat activation energy (E) for proton abstrac-
tion from the Cot atom and subsequent
the Enamine Mechanism of Natural Enzymes enamine formation. The enamine acts as
the carbon nucleophile, or aldol donor,
JOrgen Wagner, Richard A. Lerner,* Carlos F. Barbas 111* which reacts with an aldehyde electrophile,
the aldol acceptor, to form a new C-C
Antibodies that catalyze the aldol reaction, a basic carbon-carbon bond-forming reaction, bond. The Schiff base is then hydrolyzed
have been generated. The mechanism for antibody catalysis of this reaction mimics that and the product is released (in this case, a

Downloaded from http://science.sciencemag.org/ on March 9, 2017

used by natural class I aldolase enzymes. Immunization with a reactive compound co- 3-hydroxy ketone). The essence of the
valently trapped a Lys residue in the binding pocket of the antibody by formation of a mechanism is the formation of the enam-
stable vinylogous amide. The reaction mechanism for the formation of the covalent ine, which is the nascent carbon nucleo-
antibody-hapten complex was recruited to catalyze the aldol reaction. The antibodies use phile. Although transition state models
the s-amino group of Lys to form an enamine with ketone substrates and use this enamine have been proposed for aldol reactions in-
as a nascent carbon nucleophile to attack the second substrate, an aldehyde, to form a volving metals (8), models for the enamine
new carbon-carbon bond. The antibodies control the diastereofacial selectivity of the case remain to be studied.
reaction in both Cram-Felkin and anti-Cram-Felkin directions. We designed haptens that could both trap
the requisite Lys residue in the active site to
then form the essential enamine intermediate
and induce the appropriate binding sites for
One of the major goals of organic chemis- the antigen used to induce it. To test these the two substrates to overcome the entropic
try is to use the understanding of reaction ideas, we have studied the aldol condensa- barrier intrinsic to this bimolecular reaction.
mechanisms to design new catalysts. This is tion which is, arguably, the most basic C-C The simple 1,3-diketone hapten 1 provides
often not easy because one must address bond forming reaction in chemistry and elements of both a chemical and entropic trap
intermediates that are of high energy and biology. A variety of effective reagents have (Fig. IB). In water, the keto-form of the hap-
complex structure. Antibody catalysts offer been developed to control the stereochem- ten shown predominates over the enol-form
an interesting solution to the problem in istry of the aldol, but these reagents are at a 3:1 ratio (9). The reaction coordinates of
that they can be programmed by the exper- stoichiometric and require preformed eno- the aldol addition and the reaction mecha-
imenter to interact with the rate-limiting lates and extensive protecting-group chem- nism expected when the hapten interacts
transition state of a chemical reaction to istry (3). Recently, catalytic aldol reactions with some antibodies share several common
lower its energy and increase the reaction that use preformed enolates have been de- intermediates. In both cases, a tetrahedral
rate (1), but even here catalyst design is veloped, including the Mukaiyama cross- carbinolamine intermediate forms that dehy-
usually limited to the more global aspects of coupling aldol (4). A number of enzymes drates to afford the cationic iminium that
the transition state rather than the detailed catalyze the aldol condensation, and al- tautomerizes to the enamine. It was expected
reaction mechanism. Thus, while one can though much is understood about their that antibodies induced according to the hap-
deal with high-energy charges, stereoelec- mechanism (5) they accept a limited range tenic reaction mechanism would stabilize the
tronic, and geometrical features that appear of substrates (6). Thus, our goal was to analogous transition states and cationic inter-
along the reaction coordinate, the organi- induce antibodies that use the reaction mediates along the reaction coordinate of the
zation of multiple complex reaction inter- mechanisms that give aldolases their effi- aldol reaction. The driving force for the reac-
mediates remains difficult. ciency but that take advantage of the range tion of the 1,3-diketone hapten with the an-
For some reactions, the problem of com- of substrates and stereochemical specifici- tibody is the formation of a stable covalent
plex intermediates may be solved by using ties available with antibodies. vinylogous amide or conjugated enamine be-
relatively reactive compounds rather than Two mechanistic classes of aldolase en- tween the hapten and the e-amino group of
the more usual inert antigens to immunize zymes have evolved (7). Class I aldolases Lys. Calculations that make use of the Wood-
animals or select antibodies from libraries utilize the £-amino group of a Lys in the ward rules for enones indicated that the vinyl-
such that the process of antibody induction active site to form a Schiff base with one of ogous amide would have an absorption max-
involves an actual chemical reaction in the the substrates, which activates the substrate imum in the ultraviolet (UV) that would
binding site (2). This same reaction then as an aldol donor. Class II aldolases are allow for its identification, Xn.x = 318 nm
becomes part of the catalytic mechanism metalloenzymes that facilitate enolate for- (10). The stability and spectral characteristics
when the antibody interacts with a sub- mation by coordination to the substrate's ofthis type ofcompound were previously not-
strate that shares chemical reactivity with carbonyl oxygen. We chose class I aldolases ed in the studies of acetoacetate decarboxylase
as our model (Fig. lA). The reaction is by Westheimer and co-workers (11). We ex-
Departments of Chemistry and Molecular Biology, bimolecular and proceeds through covalent pected an entropic advantage by incorpora-
Scripps Research Institute, 10666 North Torrey Pines catalysis through multiple intermediates. tion of the second substrate (aldol acceptor)
Road, La Jolla, CA 92037, USA. An iminium ion or Schiff base forms that in the diketone chemical trap. Entropic ef-
*To whom correspondence should be addressed. acts as an electron sink, which lowers the fects could provide as much as 10t3 to 1011 to
SCIENCE * VOL. 270 * 15 DECEMBER 1995 1797
the rate acceleration of natural enzymes (12). because rotation of both enamine and alde- chiometry of the titration should corre-
The reaction we studied is the aldol hyde faces should provide a reasonable spond to the concentration of antibody ac-
addition of acetone to 3-phenylpropion- Heathcock angle (Fig. 2). tive sites. The titration gives a ratio of
aldehyde derivatives. The second substrate The diketone hapten 1 was synthesized acetylacetone to antibody of 1.9, indicating
is represented by the 3-phenylpropiononyl (Fig. 3) and coupled to keyhole limpet he- that each of the two identical antigen bind-
portion of the hapten. The tethering of the mocyanin. After immunization of 129 GIx± ing sites of the antibody form the vinylogous
two substrates in the diketone hapten mice, 20 hybridomas producing antibodies amide adduct (Fig. 5B). Catalysis of the
would present a substrate complex wherein to 1 were obtained with standard methods formation of the vinylogous amide was es-
the Heathcock angle (13, 14) for attack of (15). Antibodies from each cell-line were sentially complete upon mixing of the anti-
the enamine on the aldehyde would be purified by ammonium sulfate precipitation, body with hapten and sufficiently rapid that
distorted to the extreme of 900 in the rate- anion exchange, and protein G affinity determination of the rate of this reaction
determining transition state of C-C bond chromatography (16). would require stopped-flow kinetic studies.
formation. We did not expect this to be a All 20 antibodies were screened in a Antibodies 38C2 and 33F12 were as-
major impediment in the catalytic reaction microtiter plate assay for their ability to sayed for their ability to catalyze the addi-
form the proposed stable vinylogous amide tion of acetone to aldehyde 3 (Table 1).
2 by incubation of 20 ,uM antibody with Consumption of 3 and production of the
A HB 100 p.M of the diketone hapten 1 (Fig. 4). ,B-hydroxy ketone 4 were monitored by re-
(V N Two antibodies, 38C2 and 33F12, demon- versed-phase high-performance liquid chro-
)JZt7N:NH2-LYSyEs Enzz strated a strong absorption maximum at 316 matography (HPLC) (17). Aldol addition

Downloaded from http://science.sciencemag.org/ on March 9, 2017

nm characteristic of the proposed vinylo- catalyzed by both antibodies followed
H2O gous amide 2, approximating the calculated Michaelis-Menten kinetics. Generation of
H , Lys- Enz absorption maximum in the absence of pro- acetone and aldehyde 3 from the ,-hydroxy
H = _ 8' tein of 318 nm (Fig. 4). Incubation of 1 ketone 4 in the retro-aldol reaction by the
QN-Lyss s___H with Lys under identical conditions resulted antibodies was monitored by following the
/A E¶nzs /nz ( i
in no increase in absorbance. The extinc-
tion coefficient of the antibody-enamine
decrease in UV absorbance at 340 nm in a
coupled enzymatic assay with alcohol dehy-
H
Ly?N HH H complex was determined to be 15,000 cm- drogenase and P-nicotinamide adenine
M-' after subtraction of the absorbance of dinucleotide, reduced form (NADH) (18).
/K RK. ~-> the antibody (Fig. 5A), approximating that Production of aldehyde 3 (19) was moni-
observed in the reaction of acetopyruvate tored by its subsequent reduction by alcohol
+H20 with the enzyme acetoacetate decarboxyl- dehydrogenase and consumption of NADH.
Enz ase, 19,000 cm-,1 M- (11). The retro-aldol reaction was also studied by
\Lys Because the antibodies are expected to HPLC and the same results were obtained
*AtHBHH-H0 H H0 OH form an enamine with acetone in the syn-
thetic reaction, observation of the vinylo-
(17). The Michaelis constant KM and cat-
alytic rate constant kcat values were 54 p.M
+ H gous amide chromophore should not be de- and 4.4 x 1i-0 min-1, respectively, for
:NH2-Lys - Enz pendent on the aldol acceptor (benzyl) por- antibody 38C2. The remaining 18 antibod-
B 0 o tion of the hapten. We tested acetylacetone ies were unable to catalyze the synthetic
R :NH2-Lys-Ab as the minimal diketone expected to gener- and retrosynthetic aldol reactions, indicat-
1 ate the chromophore. Both antibodies re- ing that only those that formed the critical
acted with this compound and produced the intermediate were active.
expected absorbance spectrum. To deter- We studied the ability of the hapten 1
HBLysBAb
H mine the stoichiometry of the antibody- and acetylacetone to inhibit the aldol reac-
0 HA, 'NL A enamine complex, a titration of the anti- tion. When three equivalents of either hap-
Rl_~~~ body with acetylacetone was carried out ten 1 or acetylacetone were provided prior
-H2O 1 ( 1). If we assume that the reaction that to the aldol addition or retro-aldol assays,
forms the enamine is irreversible, the stoi- catalytic activity was completely inhibited,
H @ Lys - Ab

H H 0Q

° CH R R H

H, Lys-Ab H .Lys-Ab
13,~~~~~~~ ~0 0
N02
O' N: __2 0 ~N: NO

R R X= NH-A
Fig. 1. (A) General mechanism of a class aldo- o o
lase-catalyzed aldol-addition reaction (5, 7) (Enz, Fig. 2. The 1,3-diketone hapten structure con-J O
enzyme; B, base). (B) Mechanism of trapping the tains the elements of a chemical and entropic trap.nn
essential e-amino group of a Lys residue in the The binding pocket induced with the hapten 1 0 0
antibody (Ab) binding pocket by using the 1,3- does not preclude attainment of a reasonable Fig. 3. Synthesis of hapten 1. (a) LDA 12 equiva-
diketone hapten 1. The formation of the stable Heathcock angle for attack of the aldol donor on lents (eq)], THF, 4000, 1 hour. (b) 4-nitrobenzyl-
covalent vinylogous amide 2 can be detected at A the acceptor. A proper attack geometry is at- bromide, hexamethylphosphoramide, -78° to
= 316 nm (£ = 15,000). R = p(HOOC(0H9)3- tamned by simple rotation of both enamine and -40°C, 48% yield. (c) (i) Pd/C, H2, ethanol; (ii)
CONH)C6H4-. aldehyde faces. glutaric anhydride, 0H2012, 74% yield.
1798 SCIENCE * VOL. 270 * 15 DECEMBER 1995
showing that trapping of the enamine in- zymes in synthesis is their rather poor catalysts that direct the aldol addition in
termediate with the 1,3-diketones precludes acceptance of a range of substrates (6). either direction (Table 1, entry 1 or 7).
catalysis involving the Lys £--amino group. Although natural enzymes display broad Antibodies 38C2 and 33F12 are restricted to
To establish that enamine formation with specificity with respect to the aldol accep- direct the aldol addition with acetone or
the hapten in the trapping assay and ace- tor, the aldol donor is usually limited to aliphatic ketones as donors and 3-phenylpro-
tone in the catalytic assay are dependent on the natural substrate. For example, among
the same Lys residue, antibodies incubated the ketones studied for antibody catalysis Table 1. Substrate specificity of antibody 3802; R
with acetone were treated with NaBH4. (Table 1), only acetone is a substrate for a = 0H3C0NHC6H4CH2-. The kinetic parameters
Reduction with NaBH4 of the imine inter- natural enzyme. In contrast, antibody al- ka and Km of each reaction were calculated with
mediate formed in the reaction of acetone dolases can use various aldol donors and respect to the aldehyde. The aldol donors (ace-
with the Lys E--amino group in the antibod- acceptors. The antibodies accept acetone, tone, 2-butanone, 3-pentanone, 2-pentanone,
ies would result in irreversible isopropyla- fluoroacetone, chloroacetone, 2-butanone, and acetaldehyde) were fixed at a constant con-
nion of the essential amine (5). After treat- 3-pentanone, 2-pentanone, and dihydroxy- centration of 5% v/v in each experiment (1 7, 19).
ment, the antibodies could not form the acetone as aldol donor substrates. In reac- Products 7/8 and 9/1 0 were formed at ratios 94 to
vinylogous amide with the diketones. These tions with 2-butanone and 2-pentanone, the 4 and 73 to 27, respectively.
experiments provide evidence that the re- antibodies exhibit some control of the regi- (1O-t Km
action mechanism and residues induced oselectivity of the aldol addition by prefer- Reaction kca3mi-1 (jiM)
with the 1 ,3-diketone hapten are the same ential formation of the most substituted 0 OH
as those recruited in the catalytic reactions. enamine. The relative efficiency of catalysis R~JL. RJA.)
F1 6.7 17

Downloaded from http://science.sciencemag.org/ on March 9, 2017

The antibody aldolases showed a broad pH with these substrates decreases 42-fold as 3 4
optimum between 6 and 10 approximating reflected by k/IK in the acetone to pen- 0 0 OH 0 83* 125*
catM R~) . R
that observed with natural class I aldolases tanone series (Table 1, entries 1 and 3 to 5).TY'6-- 40 48
(5, 7). Thus, antibody aldolases recapitulate The failure of the antibodies to accept acet- 0
{ OH
the mechanisms of natural class I aldolases aldehyde as a donor shows that the aldol R OH_ 8 5
R 7
in the catalysis of this complex, multistep addition is directed with a ketone as the O

reaction (Fig. IA). aldol donor (Table 1, entry 6). In principle, 8

We probed the substrate specificity of the diketone hapten should induce antibod- RQ0 0
RO5. 5.
11
15
these catalysts because the most limiting ies that react at either of the two keto posi-
aspect of the application of natural en- tions of the hapten 1, thereby generating OH 0
0 0 R~OH

Fig. 4. (A) Screening of anti-

bodies for the formation of the
A
0 0
RO 1
H)L
vinylogous amide intermediate _H__ _No reaction
2. Hapten 1, 5 eq, was added 0
to 20 p.M solutions of each an- OJ~H, orato
tibody in PBS buffer (pH 7.5) in *Reaction using antibody 33F12.
a microtiter plate format. (B)
Antibodies with aldolase cata- B
lytic activity presented the typi- 1.0 ,II r r -
cal absorption maximum of the I- - .... .tC H3 OH3
vinylogous amide at 316 nm 0.8 J1_... ._ cH I # I II
H...Antib/dy
~~~~~~~0
33F12
0
~OH0
(example shown in red), where- - L .~-. --.--1--'A.v L~-..
_.. 1-..-.- pH=7.5
as none ofthe inactive antibod- c 0.6 -- -s 4- -.4,- ----I
ies did (example shown in blue). e 1 4 f---- -i f- . 00 100
Two antibodies out of 20 0 * -90
formed the intermediate 2and ~ - ~ i-8
were subsequently determined IT 1
I225 780
to be catalytic. 0. i T- T f~~ -60 ts

250 290 330 370 410 450 2 4

Wavelength (nm) 40
-30
75
-20
Fig. 5. (A) Determination of 10
the extinction coefficient of 0.4 A
A 20 *
2. A fixed concentration 4 0 5 1 1'5 20 25 30 35
(1 00 j.tM) of hapten 1 was 0315 Time (hours)
added to the indicated con- .0 1~~~~0
3 .Fig. 6. The aldol addition reaction of aldehyde 5
centrations (micromolar) of 02and acetone was monitored over a 36-hour period
33F1 2. The antibody enam- .o2 in the presence of 1.5% catalyst (20) (Ac, acyl). The
mne complex01*catalyst
could easilybe showed multipDle turnovers (-2 turnovers
pionaldehyde derivatives as acceptors. also reflected in differences in Km for 5 in 11 W. Tagaki, P. Guthrie, F. H. Westheimer, Biochem-
The reaction of the branched 3-phenyl- the antibodies (Table 1). The degree of 1.istry 905 (1968).
7,Page
h reaciM.
1 andPa W. P. JencksJen Proc.roNati.tAcad.d.Sci.
propionalehyde accPtoacce'torTable
propionaldehde Tbe11, nr enry
trohmclcnrloof the
stereocheical eciniis
contro
U.S.A. 68,1678 (1971).
2) with acetone was the most efficient and exceptional as no stereochemical informa- 13. C. H. Heathcock and L. A. Flippin, J. Am. Chem.
showed little product inhibition (Fig. 6) tion was introduced into the hapten. The Soc. 105,1667 (1983).
(20). In fact, less than 1 mole % catalyst is E. P. Lodge and C. H. Heathcock, ibid. 109, 3353
chemical reaction of the lithium enolate 14. (1987).
sufficient to achieve high conversion of of acetone with (S-2-methyl-3-phenyl- 15. G. Kohier and C. Milstein, Nature 256, 495 (1975).
substrate in a relatively short time. The propionaldehyde yields the analogous 16. V. E. Gouverneur et al., Science 262,'204 (1993).
reaction produces only the desired aldol (4S,5S) product with a diastereomeric ex- 17. A RP-C18 column (MICROSORB-MV, 0.45 cm by
product as each mole of aldehyde consumed cess of 5% in favor of this Cram-Felkin 22 cm) was used with an isocratic program of 75/25;
H20 (0. 1 % trifluoroacetic acid)/1-20:acetonitrile 1: 1,
is converted to the P3-hydroxy ketone prod- product (13). The generation of the at 1.5 mI/min, monitored at 254 nm. T-he retention
uct (Fig. 6). For this reaction, the rate of the (4S,5R) and (4R,5R) products at an 11:1I times of the aldehyde 3 and the aldol product 4 are
uncatalyzed background reaction at pH 7.5 ratio under antibody 38C2 catalysis shows ketonee 6.35 and 8.78 min, respectively. For kinetic studies,
under
underidential uused
condition
identialcondition sed in th an-
an- a reveral of the tpical
a Cram-Flthe
tyical Cram-Flkincste-rcocentrationoo
concentrationra
3 ora5
wasnfixedixat 5% v/v/vanddthe
wasvariedrfrom00 to 200 pM
tibody assays has been determined, kL ca reoselectivity of the aldol, in this case in the study of the aldol addition reaction. Antibodies
2.28 X 10-7 M '1 min-' (21). For 38C2 following the disfavored and energetically- were also assayed after an additional purification
and 33F12, (kcat/IKM)/kuna is - i O'. The more demanding anti-Cram-Felkin mode step over an anion-exchange column with identical
efficiency of catalysis is due in a large part of attack. i 8. A solution, 1 00 R1, pH 7.5, containing 0.2 mM EDTA,
to an entropic advantage in'the antibody- Catalysts that proceed by defined reac- 1 00 mM tris, 0.045 mg of yeast alcohol dehydroge-

Downloaded from http://science.sciencemag.org/ on March 9, 2017

catalyzed reaction, which is reflected as a nion mechanisms can be induced by im- nase, 0.43 mg of NADH, and 4 mM f3-hydroxy ke-
tone 4 was introduced into four wells of a microtiter
high effective molarity, kcat/kuncat >i~M munization with reactive compounds. plate. Antibodies 3802 and 33F1 2 (1 00 R.I, 34.6 p.M,
The catalytic efficiency (kcat/KM) of anti- This approach is not limited to Schiff base tris buffer, pH 7.5) were added to two different wells;
body aldolases is only -4000-fold lower or enamine mechanisms and may be used 1 00 RI buffer only was added to the remaining wells,
which were the blanks. T-he UV absorbance was
than that
than ofthe
tht of tudiedenzyme
te stdied
most
most enzyme whenevr the
whenever hemistr to
th chemistry o beaccom-
be accom- measured at 340 nm every 15 min for 24 hours. The
fructose-i1,6-bisphosphate aldolase (5). plished is beyonid that easily achieved by absorbance of the blanks was subtracted from the
The catalytic efficiency of antibody 38C2 even a concert of noncovalent interac- catalyzed reactions, and the rate was determined by
for the reaction given in entry 2, Tal Table 1,
tos tons. Thes
hs results
eut rovide
rvd futher
ute evi-
v- using s = 6220 cm -1 M'- Antibodies 3802 and
,

is 64-fold greater than that obtained with- dence of the utility of antibody catalysts to
33F1 2 had the same kcat =4.53 x 10-3i
which correlates well with the HPLC measurements.
m- 1,
catalysis by the enzyme 2-deoxyribose-5- perform disfavored chemical transforma- 19. The aldehydes 3 and 5 were prepared by using the
tions (16, 24). Heck reaction [T. Jeffrey, J. Chem. Soc. Chem. Corn-
popaealdolasephosphate
(22). ~~~~~~~~~~~~~~~~~~~~~~~~~~mun.
1984, 1287 (1984)]. The products of the aldol
We studied the product distribution of additions of entries 1 to 5 have been prepared by
the antibody-catalyzed reaction of 5 with REFERENCES AND NOTES independent synthesis. Typically, 50 to 1 00 mg aIde-
acetone after 30% conversion with a nor- 1R.ALenrS.JBnkvcP.GShut,cice
A Lerer,S. Bekovc, P G.Schltz,Scince hyde, 1 ml ketone, 4 ml H20, and 1 0 p.1 of saturated
1 NaOH solution were shaken for 1 houriThe products
mal-phase HPLC chiral support column 252, 659 (1991). were separated and purified by preparative reversed-
(Scheme 1) (22). Both antibodies catalyze 2. C. F. Barbas Ill, A. S. Kang, R. A. Lemer, S. J. Ben- phase HPLC. All new compounds were characterized
kovic, Proc. Natl. Acad. Sci. U.S.A. 88, 7978 (1991); by 'H nuclear magnetic resonance (NMR), 13C-NMR,
K. D. Janda et al., ibid. 91, 2532 (1994); P. Wirsching and infrared and mass spectrometry.
Racemic mixture ~ b 0 et al., Science, in press. 20. The reaction was monitored by HPLC as in (1 7) but
0 0 Ab 3. ~~~~~~~~~C.
H.Heathcock, Aldrichim. Acta 23, 99 (1990); C. at 77/23 ratio instead of 75/25. The retention times
for the aldehyde 5 and the 3-hydroxy ketone 6 are
RyLKH +
cH3
R..,,.J["
,
~~~~~~pH7.5
H. Heathcock, Science 214,395 (1981); D. A. Evans,
ibid. 240, 420 (1988); S. Masamune, W. Choy, J. 1. n 19 m epciey
CH3 CH3
~~~~~~~~~~~Peterson,
L. R. Sita, Angew. Chem. Int. Ed. Engl. 24, 219.2
a.L emnd and9Yin Chen,petrahedrnyef..6
>95% de OH 0W OH U 83%de 1 (1985); 0. A. Evans, J. V. Nelson, T. R. Taber, Top. 2575(195);JL-
ReymondanY.Ce, petrshdonalcomuiction.
(38C2) Rl(38C2) I I Stereoc(3802) 3, 1(1982); C. H.Heathcock eta. n 22. The four diastereoisomers have been separated on a
I I ~~~~~~~~~~Comprehensive Organic Synthesis, B. M. Trost, Ed. DIE hrla Jclm iha ortcpo
>9%de CH H3 ..u65% de (Pergamon, Oxford, 1991), vol. 2, pp. 133-319; grAmCE7/Chexane/ethancol,ummin 254h
a , nmiorai The-
(33F12) (48,54S,S) (33F712) Paterson, Pure Appi. Chem. 64, 1821 (1992). gretention tiesaforethenofour mVisoer were 19.74h
OH 0 OH 0 4. S. Kobayashi, H. Uchiro, I. Shiina, T. Mukaiyama, 23.3 t4R5Se 25.15i4omers ande 297.9
4retetintme
R R.
Tetrahedron 49, 1761 (1993); K. Furuta, T. Ma- 4P5)232 R5S,5154SR,an279
CH, ruyama, H. Yamamoto, J. Am. Chem. Soc. 113 min (4S,5S). The relative configuration had been de-
CH (4, S H(R 5Ruy1041 (19)H. Bachot,Angew. Chem. St.oE. Engi termined previously [see (20)]. The absolute config-
(4R, 5S (4R, R) 1041(1991) T. Bac, Ange. Chem.Int. E. EngL uration was deduced from an experiment wherein
33, 417 (1994), and references therein; E. M. Car- tectls a -exrbs--hsht loa
Scheme 1. reira, W. Lee, R. A. Singer, J. Am. Chem. Soc. 1 17, the cDEAtays wEAs2-doxribxclsiely-phopaealdola-odc
5.6C.9Y1Lai5
5. C.Y. Li, NNaa,D
.
hn,Sine13
aki, D Chan, 24
Scence183,1204 possessing the (5) configuration at 0-4 [see (6)]. The
aldol product generated by DERA consists of a 1: 4.5
the diastereoselective addition of acetone (1 974); A. J. Morris and D. R. Tolan, Biochemistry 33, mixture of (4S5,R)-6 (92% diasteriomeric excess, de)
tothe re-face
tO of 5 regardless of the stereo- 06.12291
~ ~ ~ ~ ~ ~ ~.-H. (1994).and G. M. Whitesides, Enzymes in Syn-
~ 6 Wong ad(S5)6(9%d) h kinetic
tipandti4ular- (>95%frmde)oThre iei parameterso
aaeeso
chemistry at C-2 of this substrate. The aldol thetic Organic Chemistry (Pergamon, Oxford, 1994); this-partcua transformationwrM.45x1-
reactions follow the Cram-Felkin (23)
mode of attack on (S-5 to generate the
M. 0. Bednarski, in Comprehensive Organic Synthe-
2isppM.455473t, Fd. BaerbasaIll OxfodWang,0.-H. minD1JCa andK A.3400 p.M.z .Am hm.Sc
74, 5828 (1952); M. Ch6rest, H. Felkin, N. Prudent,
(4S,5S )-6 product adteatCrmFl and
the
anti-ram-Eel- Won,
WogJ.A.Ce.Sc Hag. C.-M.
455m73Che. So.Babs112,2013(990)
pp. 1221 90)H..M. Tetrahedron Lett. 1968, 2199 (1 968); N. T. Ahn, Top.
Currn Chem. 88,145 (1980).
kin mode of attack on (R)-5 to generate the Gijsen and 0.-H. Wong, ibid. 11 7, 2947 (1995); C.- 4 .D ad,C .Seln .A enr cec
(4S,5R-6 product. The products formed H. Wonget et al., ibid., p. 3333; L. Chen, D. P. Du- 24.2KD04 Janda, C.GF.Sheavain,R. A.Lerner,Scienc
mas, 0. -H. Wong, ibid. 1 14, 741 (1992).26,24(93;BF.rvatJ.AAslyK.0
w7ith similar y7ields-,so thre)n was. no k-ine-tic 7.AWIJ.Ruter Fed. Proc Am.CSc.- ExpBiol 231248 Janda, D. L. Boger, R. A. Lerner, J. Am. Chem. Soc.
 Efficient Aldolase Catalytic Antibodies That Use the Enamine
Mechanism of Natural Enzymes
Jürgen Wagner, Richard A. Lerner and Carlos F. Barbas III (December
15, 1995)
Science 270 (5243), 1797-1800. [doi: 10.1126/science.270.5243.1797]

Editor's Summary

This copy is for your personal, non-commercial use only.

Downloaded from http://science.sciencemag.org/ on March 9, 2017

Article Tools Visit the online version of this article to access the personalization and article
tools:
http://science.sciencemag.org/content/270/5243/1797

Permissions Obtain information about reproducing this article:

http://www.sciencemag.org/about/permissions.dtl

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in
December, by the American Association for the Advancement of Science, 1200 New York Avenue NW,
Washington, DC 20005. Copyright 2016 by the American Association for the Advancement of Science;
all rights reserved. The title Science is a registered trademark of AAAS.