A fluorogenic assay for transketolase from Saccharomyces cerevisiae

Article abstract of DOI:10.1016/S0040-4039(02)02634-5

In order to generate transketolase (TK) with new or improved properties by in vitro evolution, an efficient screening system is an absolute prerequisite for identifying the evolved enzyme variants. We report here an assay allowing us to detect wild type TK activity in vitro by fluorescence. We examined the use of the fluorogenic compound 1 as donor substrate of TK, which is itself non fluorescent but releases a fluorescent product: umbelliferone.

Full text of DOI:10.1016/S0040-4039(02)02634-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 827–830
A fluorogenic assay for transketolase from
Saccharomyces cerevisiae
a
a
b
c
a,
Aur e´ lie Sevestre, Virgil H e´ laine, Ghislain Guyot, Christine Martin and Laurence Hecquet *
a
Universit e´ Blaise Pascal, Laboratoire de Synth e` se et Etude de Syst e` mes a` Int e´ r eˆ t Biologique, UMR 6504, 63177 Aubi e` re Cedex,
France
b
Universit e´ Blaise Pascal, Laboratoire de Photochimie Mol e´ culaire et Macromol e´ culaire, UMR 6505, 63177 Aubi e` re Cedex,
France
c
Laboratoire de Microbiologie, INRA, 63200 Theix, France
Received 29 October 2002; accepted 20 November 2002
Abstract—In order to generate transketolase (TK) with new or improved properties by in vitro evolution, an efficient screening
system is an absolute prerequisite for identifying the evolved enzyme variants. We report here an assay allowing us to detect wild
type TK activity in vitro by fluorescence. We examined the use of the fluorogenic compound 1 as donor substrate of TK, which
is itself non fluorescent but releases a fluorescent product: umbelliferone. © 2003 Elsevier Science Ltd. All rights reserved.
TK (EC 2.2.1.1.) is essential in the metabolism of all
living cells. This enzyme is intracellular and is impli-
cated in the pentose phosphate pathway. TK reversibly
transfers the C1ꢀC2 ketol unit from a ketose phosphate
to an aldose phosphate to generate a ketose phosphate
with a new asymmetric center in the (S) configuration.
The enzyme requires two cofactors for activity: thi-
implicated in a latent human genetic disease, Wernicke–
1
Korsakoff syndrome. In the brain of alcohol-fed rats,
2
the enzyme activity and concentration decrease. TK
abnormalities have also been reported to occur in ery-
throcytes and brains of some subjects affected by
3
Alzheimer’s disease.
2
+
amine pyrophosphate (ThPP) and Mg .
TK is also a useful catalyst for ketose synthesis due to
the stereocontrolled formation of the C2ꢀC3 bond.
4
Much work has shown the involvement of TK in
diseases related to the nervous system. TK has been
Transketolase isolated from spinach leaves, baker’s
5a,b
5c,d
yeast,
and Escherichia coli
have been investigated.
Scheme 1. Access to new
D
-threo-aldoses or
Corresponding author. Tel.: +33-3-73-407871; fax: +33-3-73-407717; e-mail: hecquet@chimtp.univ-bpclermont.fr
040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
L-erythro-ketoses by TK engineering.
*
0
PII: S0040-4039(02)02634-5

8
28
A. Se6estre et al. / Tetrahedron Letters 44 (2003) 827–830
Recently, we used Saccharomyces cerevisiae recombi-
available umbelliferone 3 (Scheme 2). It was first ally-
lated in refluxing acetone to give the olefin 4 in 96%
yield. Ozonolysis of the crude material in methylene
chloride with 10% DMF at −30°C and subsequent
reduction of the ozonide intermediate with dimethyl-
sulfide gave the aldehyde 5 in 62% yield. It is notewor-
thy that attempts to carry out this step without DMF
or at lower temperature resulted in precipitation of the
aldehyde. The chiral centers (3S,4R) of the sugar moi-
ety in the fluorogenic substrate 1 were introduced by
aldol condensation using rabbit muscle aldolase
6
,7
nant TK. The TK reaction is reversible except when
the substrate is b-hydroxypyruvic acid (Scheme 1a).
This property is very convenient for synthetic purposes
because the decarboxylation of this donor substrate
makes the overall condensation reaction irreversible.
TK catalyses the transfer of a ketol unit from b-hydroxy-
pyruvic acid to an aldehyde to generate a
D-threo
(
3S,4R) ketose. The enzyme appears highly specific for
ketol donor substrates and for hydroxyaldehyde sub-
strates with the (R) configuration.
(
RAMA, EC 4.1.2.13), dihydroxy acetone phosphate
Our current research interest is to generate TK with
new or improved properties. It has been demonstrated
that in vitro evolution can give new enzymes with
(DHAP) as donor substrate and compound 5 as accep-
tor substrate at pH 7.5 followed by enzymatic dephos-
phorylation of the aldol product using acid
phosphatase (EC 3.1.3.2) at pH 4.8. This strategy has
most often been used to produce ketoses (3S,4R) by
CꢀC bond formation in a highly stereoselective manner.
The utility of RAMA has been demonstrated in the
synthesis of C labeled sugars, heteroatom-substituted
sugars, deoxy sugars, fluoro sugars, long-chain sugars,
and cyclitols. Well over 100 aldehydes have been used
as acceptor substrates. Due to the low solubility of
compound 5 in water, even in the presence of co-sol-
vents such as DMSO or MeOH, addition of a modified
cyclodextrin (mCD) was necessary to obtain a clear
8
9
altered substrate specificity, topology, enantioselectiv-
1
0,11
12
ity,
thermal stability and organic solvent resis-
3
1
tance. By means of this evolution strategy, we are
interested in obtaining variants of TK able to accept
glyoxylate as donor substrate leading to aldoses, or able
to accept (S)-hydroxyaldehydes as acceptor substrates
13
leading to L-erythro ketoses (Scheme 1b).
17
An efficient screening system is an absolute prerequisite
for identifying the evolved enzyme variants that display
improved properties.
18
solution. Thus, substrate 1 was obtained in 35% over-
all yield for the two enzymatic steps.
TK activity can be detected by a spectrophotometric
14
assay involving dependent NADH enzyme. The disad-
vantages of this method as a screening assay is its low
sensitivity, specificity and its non viability with whole
cells.
The fluorogenic compound 1 was used as donor sub-
strate of the enzyme. TK cleaved the C2ꢀC3 bond of 1
and generated 2. A fluorescent signal appeared because
2
is unstable and underwent a rapid b-elimination to
In this paper, we report an assay allowing us to detect
release umbelliferone 3, a highly fluorescent compound
(Scheme 3). According to the literature the b-elimina-
tion reaction of analogs of 2 is accelerated by weak
bases such as ethanolamine, morpholine, or Tris and
wild type TK activity from compound 1 by fluorescence
15,16
according to Reymond’s test.
15
The fluorogenic compound 1 was prepared by a
chemoenzymatic route starting from commercially
catalyzed by bovine serum albumin (BSA). We used
Tris buffer at pH 8.2. At this pH value, the deproto-
Scheme 2. Synthesis of the fluorogenic substrate. (i) CH ꢁCHꢀCH Br, K CO , 96%; (ii) O , Me S, 62%; (iii) RAMA, DHAP,
2
2
2
3
3
2
mCD; (iv) acid phosphatase, 35%.
Scheme 3. Fluorescence assay based on umbelliferone release.

A. Se6estre et al. / Tetrahedron Letters 44 (2003) 827–830
829
nated form in the first excited singlet state of umbellif-
erone 3 (pK =8) gave a fluorescent signal at u =390
a
ex
nm and u =452 nm. We tested the efficiency of the
em
assay towards compound 1 by measuring the fluores-
cence signals of 3 (arbitrary units) during the time.
Fluorescence was correlated with umbelliferone concen-
tration according to a calibration curve. Thus, we cal-
culated the apparent velocity (Vapp) of release of 3.
We checked the stability of 1 using TK without BSA
(
Fig. 1). The results indicated no release of umbellifer-
one. When we added BSA, a strong fluorescent signal
depending on the type of commercial BSA appeared.
Only BSA prepared by cold alcohol precipitation gave
good rates. According to the literature, this might be
attributed to a different folding of the protein. BSA’s
native conformation was necessary for the catalysis but
Figure 1. Influence of the type of BSA on umbelliferone
release in 50 mM aq. Tris buffer, pH 8.2, 2 mM ThPP, 3 mM
−
1
MgCl
, 1 mM D-ribose-5-phosphate, 100 mM 1, 0.01 mg mL
2
−1
TK and 2 mg mL BSA: (a) fractionated by alcohol precipi-
tation (Vapp=1.3×10 mmol min mg ); (b) fractionated by
−
4
−1
−1
1
9
heat shock treatment modified it.
heat shock.
We also checked the stability of 1 in the presence of
BSA without TK (Figs. 2d and 3a%,b%,c%). These control
experiments revealed
umbelliferone.
a
very low release of
In order to show that in this process, the b-elimination
catalysed by BSA was not rate-limiting, we made the
measurements varying the TK quantities. Figure 2
confirmed that the fluorescent signal was proportional
to enzyme quantities. The assay was quite rapid
because a significant fluorescence signal is detectable in
−
1
approximately 15 min in the presence of only 5 mg mL
of TK. Moreover, the velocity of the reaction increased
proportionally with compound 1 (Fig. 3).
In conclusion, we showed the possibility of measuring
wild type TK activity using a highly, sensitive, specific,
throughput fluorogenic assay which should be practical
in cell culture media. This assay will be used in the
evaluation of TK mutants generated by in vitro evolu-
tion, with the fluorogenic substrates containing a sugar
moiety depending on the enzymatic property looked
Figure 2. Influence of TK quantities in 50 mM aq. Tris
buffer, pH 8.2 using 2 mM ThPP, 3 mM MgCl
compound 1, 1 mM -ribose-5-phosphate, 2 mg mL BSA
prepared by cold alcohol precipitation: (a) 0.02 mg mL TK
, 100 mM
2
−1
D
−
1
−
4
−1
−1
−1
(Vapp=2.3×10 mmol min mg ); (b) 0.01 mg mL TK
−
4
−1
−1
−1
(Vapp=1.3×10 mmol min mg ); (c) 0.005 mg mL TK
−
5
−1
−1
(Vapp=6.2×10 mmol min mg ); (d) no TK.
for. To select mutant TK able to produce
D
-threo
aldoses or
L
-erythro ketoses, we will need compounds 6
or 7.
Figure 3. Influence of compound 1 concentration in 50 mM
aq. Tris buffer, pH 8.2 using 2 mM ThPP, 3 mM MgCl , 0.01
2
−
1
mg mL TK, 1 mM
(
D
-ribose-5-phosphate: (a) 200 mM 1
−
4
−1 −1
Vapp=2.7×10 mmol min mg TK); (b) 100 mM 1 (Vapp=
−4 −1 −1 −5
The mutant TK will release umbelliferone that could be
easily detected according to the fluorogenic assay
described with wild type TK herein.
1.3×10 mmol min mg TK); (c) 50 mM 1 (Vapp=6.7×10
−1 −1
mmol min mg TK); (a%) 200 mM 1, no TK; (b%) 100 mM 1,
no TK; (c%) 50 mM 1, no TK.

8
30
A. Se6estre et al. / Tetrahedron Letters 44 (2003) 827–830
11. Fung, S.; Machajewski, T. D.; Mak, C. C.; Wong, C. H.
Acknowledgements
Chem. Biol. 2000, 7, 873–883.
12. Zhao, H.; Arnold, F. H. Protein Eng. 1999, 12, 47–53.
This work was supported by a grant of MENRT. We
would like to thank Professor G. Schneider (Karolinska
Institut, Division of Molecular Biology, S-171 77
Stockholm, Sweden) for the gift of the yeast strain
H420 transformed with pTKL1.
13. Moore, J. C.; Arnold, F. H. Nat. Biotechnol. 1996, 14,
458–467.
14. Hecquet, L.; Bolte, J.; Demuynck, C. Biosc. Biotechnol.
Biochem. 1993, 12, 2174–2176.
15. Klein, G.; Reymond, J. L. Bioorg. Med. Chem. Lett.
1998, 8, 1113–1116.
16. Jourdain, N.; Carlon, R. P.; Reymond, J. L. Tetrahedron
Lett. 1998, 39, 9415–9418.
References
17. Machajewski, T. D.; Wong, C. H. Angew. Chem., Int. Ed.
000, 39, 1352–1374.
2
1
2
. Nixon, P. F.; Price, J.; Norman, M. N.; Williams, G. M.;
Kerr, R. A. Chim. Acta 1990, 192, 89–98.
. Paoletti, F.; Mocali, A.; Marchi, M.; Sorbiand, S.; Pia-
centini, S. Biochem. Biophys. Res. Commun. 1990, 172,
3
18. Experimental procedure for the enzymatic steps:
0.878 g of (2-hydroxypropyl)-b-cyclodextrin (0.63 mmol)
and 100 mg (0.53 mmol, 1.2 equiv.) of 7-(2-
oxoethoxy)coumarin 5 were dissolved in 3.9 mL of
methanol and allowed to stir for 15 min. 2.6 mL of water
were added and the methanol evaporated. 1.3 mL of a
DHAP solution (400 mM, 0.53 mmol, 1 equiv., pH 7.8)
were then poured (to give a 200 mM final substrate
concentration) followed by 150 U of commercially avail-
able RAMA. The mixture was stirred for 48 h at room
temperature. The reaction was followed by TLC using
96–401.
. Jung, E. H.; Itokawa, Y.; Nishino, K. Am. J. Clin. Nutr.
991, 100–105.
3
4
5
1
. Bolte, J.; Demuynck, C.; Hecquet, L.; Samaki, H. Carbo-
hydr. Res. 1990, 206, 79–85.
. (a) Ziegler, T.; Straub, A.; Effenberger, F. Angew. Chem.,
Int. Ed. Engl. 1988, 27, 716–721; (b) Kobori, Y.; Myles,
D. C.; Withesides, G. M. J. Org. Chem. 1991, 22, 5899–
1-propanol/ethyl
acetate/water/ethanol/pyridine/acetic
5
907; (c) Morris, K. G.; Smith, M. E. B.; Turner, N. J.
acid: 35/15/25/15/10/10 as eluent until complete disap-
pearance of the starting aldehyde. The pH was adjusted
to 4.8 and 150 U of acid phosphatase was added. The
Tetrahedron: Asymmetry 1996, 7, 2185–2188; (d) Zimmer-
mann, F. T.; Schneider, A.; Sch o¨ rken, Y.; Sprenger, G.
A.; Fessner, W. D. Tetrahedron: Asymmetry 1999, 10,
mixture was then stirred overnight.
3 volumes of
1
643–1646.
methanol were added to precipitate proteins, the mixture
was centrifuged to 8000 rpm and the subsequent superna-
tant was evaporated to dryness under vacuum. Two flash
chromatographies (methylene chloride/methanol: 9/1)
6
. (a) Wikner, C.; Meshalkina, L.; Nilsson, U.; Nikkola,
M.; Lindqvist, Y.; Schneider, G. J. Biol. Chem. 1994, 269,
3
Gu e´ rard, C.; Schneider, G. FEBS Lett. 1998, 424, 49–52.
. (a) Andr e´ , C.; Gu e´ rard, C.; Hecquet, L.; Demuynck, C.;
Bolte, J. J. Mol. Catal. B: Enzym. 1998, 5, 459–466; (b)
Gu e´ rard, C.; Alphand, V.; Archelas, A.; Demuynck, C.;
Hecquet, L.; Furstoss, R.; Bolte, J. Eur. J. Org. Chem
2144–32150; (b) Nilsson, U.; Hecquet, L.; Gefflaut, T.;
1
gave 56 mg of a white solid (35% yield); H NMR
7
(CD COCD , 400 MHz): l (ppm): 4.15 (dd, J=6.2, 9.8
3 3
Hz, 1H), 4.26 (dd, J=6.2, 9.8 Hz, 1H), 4.40 (td, J=2.4,
6.2 Hz, 1H), 4.45 (d, J=19.6 Hz, 1H), 4.50 (d, J=2.4 Hz,
1H), 4.55 (d, J=19.6 Hz, 1H), 6.20 (d, J=9.2 Hz, 1H),
6.8 (s, 1H), 6.9 (d, J=8.5 Hz, 1H), 7.54 (d, J=8.5 Hz,
1
999, 1, 3399–3402; (c) Crestia, D.; Gu e´ rard, C.;
13
Veschambre, H.; Hecquet, L.; Demuynck, C.; Bolte, J.
1H), 7.87 (d, J=9.2 Hz, 1H); C NMR (CD COCD ,
3 3
Tetrahedron: Asymmetry 2001, 12, 869–876.
. Stemmer, W. P. C. Nature 1994, 370, 389–391.
. MacBeath, G.; Kas, T. P.; Hilvert, D. Science 1998, 279,
100 MHz): l (ppm): 66.8, 68.9, 70.1, 75.7, 101.3, 112.5,
8
9
112.7, 112.8, 129.3, 143.7, 155.8, 160.1, 211.9; HRMS
+
calcd for C14
(LSIMS).
H
O
: [M+H] 295.0818, found 295.0823
14
7
1
958–1961.
0. May, O.; Nguyen, P. T.; Arnold, F. H. Nat. Biotechnol.
000, 18, 317–320.
1
19. Liu, X. C.; Lemma-Gray, P. Biotechnol. Bioeng. 2001, 6,
524–527.
2