Highly sensitive and rapid detection of antibody catalysis by luminescent bacteria

Article abstract of DOI:10.1016/S0960-894X(00)00455-8

A highly sensitive, inexpensive, and facile bioluminescent assay for the detection of catalytic antibodies has been developed. This assay may be used for the early detection of antibody catalysis. The efficiency of this technique was exemplified by the use of the luminescent bacterium VhM42 for monitoring an antibody-catalyzed retroaldol fragmentation reaction with aldolase antibodies 38C2 and 24H6. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00455-8

Bioorganic & Medicinal Chemistry Letters 10 (2000) 2353±2356
Highly Sensitive and Rapid Detection of Antibody Catalysis by
Luminescent Bacteria
a
a,y
b,{
Hagit Shulman, Anatol Eberhard, Carolyn Eberhard,
b
a,c, ,x
Shimon Ulitzur and Ehud Keinan *
aDepartment of Chemistry and Institute of Catalysis Science and Technology, Technion-Israel Institute of Technology, Technion City,
Haifa 32000, Israel
b
Department of Food Engineering and Biotechnology, Technion-Israel Institute of Technology, Technion City, Haifa 32000, Israel
cDepartment of Molecular Biology and The Skaggs Institute for Chemical Biology, The Scripps Research Institute,
1
0550 North Torrey Pines Road, La Jolla, CA 92037, USA
Received 13 June 2000; accepted 8 August 2000
AbstractÐA highly sensitive, inexpensive, and facile bioluminescent assay for the detection of catalytic antibodies has been devel-
oped. This assay may be used for the early detection of antibody catalysis. The eciency of this technique was exempli®ed by the
use of the luminescent bacterium VhM42 for monitoring an antibody-catalyzed retroaldol fragmentation reaction with aldolase
antibodies 38C2 and 24H6. # 2000 Elsevier Science Ltd. All rights reserved.
The most common practice in the design and production
of catalytic antibodies for a given reaction involves
immunization against a transition state analogue (TSA)
and isolation of the best binding antibodies to this ana-
0.1 mL.2 Previous e€orts to detect catalytic activity
involved detection of the reaction products using either UV
spectroscopy3 or anti-product antibodies (cat-ELISA).
Alternatively, the reaction products were selectively
tagged with easily detectable groups. Thus, tagging with
4
1
logue. While this approach was proven to be a powerful
5
strategy to elicit ecient catalysts, not every antibody
binds the TSA in such a way that will a€ect catalysis.
Furthermore, many reactions involve more than one
transition state that requires stabilization. Considering
that only a small fraction of the immune response is
typically sampled by the hybridoma technology, the
commonly applied screening for best binders may not
necessarily uncover the best catalyst in a given popula-
tion of monoclonal antibodies. A better approach
would involve the screening for actual catalytic activity
rather than for TSA binding. Such a screening technique,
however, should be suciently sensitive to allow detection
of catalytic activity when the concentration of catalyst in
the hybridoma supernatant solution does not exceed
biotin allowed for detection with an avidin reagent,
6
tagging with DNA allowed for PCR detection, and
labeling with acridone allowed for ¯uorescence detection
on TLC plates.7
Intact marine bioluminescent bacteria provide interesting
opportunities for sensitive detection of speci®c com-
8
pounds. For example, Vibrio harveyi produces light as a
result of the luciferase-catalyzed reaction of reduced ¯avin
9
mononucleotide, with a long chain aldehyde and oxygen.
An aldehyde-negative mutant (VhM42) produces low or
no light except in the presence of exogenous aldehyde.
The addition of aldehyde to these bacteria induces a
sharp increase in light emission within less than a second.
The use of this mutant with a simple luminometer can
therefore rapidly detect very small amounts of long
chain aldehydes.
�
7
10
M and the volume of the entire reaction mixture is
*Corresponding author. Tel.: +1-858-784-8511; fax: +1-858-784-8732;
e-mail: keinan@scripps.edu
We reasoned that VhM42 could be used to detect the
catalytic activity of an antibody in a reaction that pro-
duces a long chain aldehyde. The commercially avail-
able aldolase antibody 38C2, which was obtained via
reactive immunization,10 has already been shown to
catalyze the aldol addition and retroaldol fragmentation
y
Permanent address: Department of Chemistry, Ithaca College, Ithaca,
NY 14850, USA.
{
Permanent address: Department of Plant Biology, Cornell University,
Ithaca, NY 14853, USA.
x
Incumbent of Benno Gitter & Ilana Ben-Ami chair of Biotechnology,
Technion.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00455-8

2
354
H. Shulman et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2353±2356
reactions with a broad variety of substrates.11 Therefore,
we anticipated that this antibody would also catalyze the
retroaldol fragmentation of 4-hydroxy-2-dodecanone, 1,
to produce nonanal, 2, and acetone (Scheme 1).12
from Michaelis±Menten analysis of the experimental
data which were obtained via the bioassay (Fig. 2).
Samples of several nonrelevant antibodies15 (2 mg/mL)
as well as 38C2 at various concentrations (0.27 to
2.0 mg/mL) were tested for aldolase activity according
to this procedure (Fig. 3). We have also tested the
amount of aldehyde formed solely by the serum-based
medium which is normally used to grow hybridomas.
These results clearly show that the catalytic antibody,
38C2, can readily be distinguished from the other, non-
catalytic antibodies as well as from serum proteins.
Here we show that 38C2 indeed catalyzes the fragmen-
tation of 1 to produce 2 and that this product is easily
detected using VhM42. This assay, which is sensitive,
inexpensive, and easy to perform, could be suitable for
the early detection of catalytic antibodies in hybridoma
cultures.
VhM4213,14 was grown at 25 C for approximately 24 h
on Petri plates containing ASW (0.3 M NaCl, 0.01M
KCl, 0.05 M MgSO , and 0.01 M CaCl ) with 5 g/L of
ꢀ
We examined our recently discovered aldolase antibody,
24H6,15 under the assay conditions. Our previous
attempts to detect catalysis of the retroaldol fragmenta-
tion with aliphatic substrates, such as 1, using this anti-
body were unsuccessful, probably due to the high
4
2
peptone (Difco), 3 g/L of yeast extract (Difco) and 15 g/L
of agar (Difco). Bacterial colonies were dispersed in
ASW to an OD600 of approximately 1.5 and kept on ice
for the assays.
A calibration plot for nonanal was obtained using
� 6
�
3
standard aqueous solutions (10 ±10 M, containing
0% acetonitrile). These standard solutions were pre-
pared from stock solutions (0.1 M in CH CN, made
1
3
fresh daily) for which nonanal that had been distilled
under reduced pressure was used. For each measure-
ꢀ
ment, 1 mL of ASW was placed in a vial at 23 C, 10 mL
of the bacterial suspension was added and the light
output was measured using a luminometer (Deltatox
model PSI, Azur Environmental Co.). The standard
solution of nonanal (10 mL) was added and the light
output was measured immediately. The di€erence
between these two readings was taken as the net light
output caused by the aldehyde. The resultant calibra-
tion line (Fig. 1) indicates that nonanal can be easily
�
11
detected at levels as low as 0.01 mM (10 mol).
Figure 1. Calibration line for nonanal. Light emission is given in
arbitrary units as a function of nonanal concentration in the assay
In the absence of antibody no aldehyde was detectable
�
vial. In this experiment, the practical detection limit was found to be 3
� 12
mols aldehyde in the assay vial).
3
in an aqueous solution of 1 (10 M containing 10%
CH CN) by this bioassay even after several weeks at
light units (10
3
room temperature. All antibody-catalyzed reactions
�
3
were carried out with 1 (10 M) in phosphate bu€ered
saline (PBS, 0.05 M sodium phosphate, 0.1 M NaCl, pH
ꢀ
7
.4) containing 5% acetonitrile at 25 C, unless other-
wise indicated. Aliquots were removed and assayed for
nonanal as described above. On the basis of these mea-
surements we can conclude that the detection of as little
�
8
as 5 mg/mL (ꢁ3Â10 M) of this antibody is feasible
under optimal conditions. The 38C2-catalyzed retro-
aldol fragmentation reaction of 1 was found to follow
Michaelis±Menten kinetics. The kinetic parameters
�
1
(kcat=0.001 min
and K =190 mM) were obtained
M
Figure 2. Michaelis±Menten kinetic analysis for the antibody 38C2-
ꢀ
catalyzed retroaldol fragmentation of 1 at 25 C. No background
reaction could be detected.
Scheme 1.

H. Shulman et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2353±2356
2355
In conclusion, we have provided evidence showing that
the bioluminescent assay, which is sensitive, inexpensive,
and easy to perform, could be used for the early detection
of catalytic antibodies. The eciency of this technique was
exempli®ed by monitoring the antibodies 38C2- and
2
4H6-catalyzed retroaldol fragmentation reaction with
the luminescent baterium VhM42. Luminescent bacteria
with other speci®cities can be developed and used to
assay other types of catalytic activities.
Acknowledgements
We thank the Israel Science Foundation and the Skaggs
Institute for Chemical Biology for ®nancial support.
A.E. thanks Ithaca College for a sabbatical leave. C.E.
thanks Cornell University for a leave of absence.
Figure 3. Nonanal formation from 4-hydroxy-2-dodecanone incu-
ꢀ
bated with various proteins at 25 C for 6 days. Concentration of all
proteins was 2 mg/mL unless otherwise indicated. Serum: 10% FCS,
References and Notes
5
% HCF, HAT.
1
(
. (a) Schultz, P. G.; Lerner, R. A. Science 1995, 269, 1835.
b) Jacobsen, J. R.; Schultz, P. G. Curr. Opin. Struct. Biol.
preference of 24H6 for aromatic substrates. The
increased sensitivity of our bioluminescence assay,
however, allowed for detection of the 24H6-catalyzed
fragmentation of 1 to produce 2 (2.3 mM of product
with 24H6 versus 0.12 mM in the bu€er-catalyzed reac-
tion after an identical incubation time).
1
3
1
2
995, 5, 818. (c) MacBeath, G.; Hilvert, D. Chem. Biol. 1996,
, 433. (d) Keinan, E.; Lerner, R. A. Israel J. Chem. 1996, 36,
13. (e) Thomas, N. R. Nat. Prod. Rep. 1996, 479.
. (a) Goding, J. W. Monoclonal Antibodies: Principles and Prac-
tice, 2nd Ed.; Academic: San Diego, 1986. (b) Tramontano,
A.; Schloeder, D. Methods in Enzymol. 1989, 178, 531.
3
D. Nature 1995, 373, 228.
4. (a) Taw®k, D. S.; Green, B. S.; Chap, R.; Sela, M.; Eshhar,
Z. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 373. (b) MacBeath,
G.; Hilvert, D. J. Am. Chem. Soc. 1994, 116, 6101.
. Thorn, S. N.; Daniels, R. Q.; Auditor, M. T. M.; Hilvert,
The usefulness of the bioluminescence detection technique
is further exempli®ed by the fact that attempts to follow
this antibody-catalyzed reaction by GC analysis (using a
DB-5 column and an FID detector) gave irreproducible
results, probably due to thermal decomposition of 1
within the GC injector port.
5
. Lane, J. W.; Hong, X.; Schwabacher, A. W. J. Am. Chem.
Soc. 1993, 115, 2078.
. Fenniri, H.; Janda, K. D.; Lerner, R. A. Proc. Natl. Acad.
Sci. U.S.A. 1995, 92, 2278.
. Reymond, J.-L.; Koch, T.; Schro
6
An important feature of the bioluminescence analysis is
that it is a very facile and inexpensive bioassay. Detection
is performed at room temperature, using simple instru-
mentation. Even when operated manually, approxi-
mately 60 samples can be measured per hour. This rate
would allow for complete screening of an entire set of
antibodies from a single fusion protocol (which typically
yields 2000±3000 hybridoma clones).
7
È
er, J.; Tierney, E. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 4251.
8. Ulitzur, S. J. Biolumin. Chemilumin. 1997, 12, 179.
9. Hastings, J. W. In Methods in Enzymology; DeLuca, M. A.
Ed.; Academic: New York, 1978; Vol. LVII, pp 125±135.
1
2
1
0. Wagner, J.; Lerner, R. A.; Barbas, C. F., III Science 1995,
70, 1797; Aldrich #47,995-0.
1. (a) Barbas, C. F., III; Heine, A.; Zhong, G.; Ho€mann,
T.; Gramatikiva, S.; Bjo
Stura, E.; Wilson, I. A.; Lerner, R. A. Science 1997, 278, 2085.
b) Zhong, G.; Ho€mann, T.; Lerner, R. A.; Danishefsky, S.;
È
rnestedt, R.; List, B.; Anderson, J.;
In addition to the above-described procedure, the bio-
luminescent assay was examined with samples of 2 in a
(
Barbas, C. F., III J. Am. Chem. Soc. 1997, 119, 8131. (c) List,
B.; Shabat, D.; Barbas, C. F., III; Lerner, R. A. Chem. Eur. J.
1998, 4, 881. (d) Ho€mann, T.; Zhong, G.; List, B.; Shabat,
D.; Anderson, J.; Gramatikova, S.; Lerner, R. A.; Barbas,
C. F., III J. Am. Chem. Soc. 1998, 120, 2768.
96-well microtiter plate using concentrations that could
occur in the early screening for catalytic activity in
hybridoma clones. Solutions of 2 (10 mL each, at a con-
centration range between 0 and 0.001 M) were added to
the wells of a microtiter plate. Each well contained
1
2. The aldol adduct 1 was synthesized in one step from
freshly distilled nonanal and the lithium enolate of acetone.
Puri®cation by ¯ash chromatography (silica gel, hexane:ethyl
acetate 85:15) a€orded 1 in the form of a colorless oil. 1
NMR: 3.9 (m, 1H), 3.1 (d, J=3.6 Hz, 1H), 2.47 (m, 2H), 2.07
1
(
00 mL of bu€er solution. The bacterial suspension
10 mL) was added rapidly to every well and light produc-
tion was determined immediately with a luminometer
H
�
5
(Lucy I, Anthos Co.). Samples containing 10 M nonanal
(
(
s, 3H), 1.3 (m, 14H), 0.77 (t, J=6.2 Hz, 3H). CI-MS: m/z 201
+
were readily detected (data not shown). The sensitivity
of this bioassay could be increased by using an aldehyde
knock-out mutant, which is absolutely dark in the
absence of 2, in place of the somewhat leaky strain
VhM42 used here.
MH ). Its purity was not readily veri®able by GC due to its
tendency to undergo a partial retroaldol reaction to give the
starting materials. However, the bioassay as described above
showed it to contain no detectable nonanal.
13. Ulitzur, S.; Hastings, J. W. In Methods in Enzymology;

2
356
H. Shulman et al. / Bioorg. Med. Chem. Lett. 10 (2000) 2353±2356
DeLuca, M. A., Ed.; Academic: New York, 1978; Vol. LVII,
pp 189±194.
J. Am. Chem. Soc. These antibodies were induced against a hapten
with a diketone chemical trigger using the reactive immuniza-
tion technique. The antibodies chosen in this case are ones that
were previously identi®ed as non-catalytic clones using both the
usual HPLC- or UV-based assays for aldolase activity.
1
1
1
4. Ulitzur, S.; Hastings, J. W. Proc. Natl. Acad. Sci. U.S.A.
979, 76, 265.
5. Shulman, H.; Keinan, E.; submitted for publication in