Resolution of (+)-abscisic acid using an Arabidopsis glycosyltransferase

Article abstract of DOI:10.1016/j.tetasy.2004.11.062

Abscisic acid (ABA) can exist as two enantiomers, with (+)-ABA as the naturally occurring form. Typically, both enantiomers occur in chemical preparations and both can be modified in the plant to their respective glucose esters. To identify glycosyltransferases capable of discriminating between the different forms of ABA, the Family 1 enzymes of Arabidopsis thaliana were screened for activity towards (±)-ABA. Eight enzymes were found to recognise the plant hormone, with one UGT71B6 showing enantioselective glucosylation towards (+)-ABA. UGT71B6 was used in a whole-cell biocatalysis system as a means of separating (+)- and (-)-ABA, thereby offering an alternative to chemical synthesis for the production of pure (+)-ABA.

Full text of DOI:10.1016/j.tetasy.2004.11.062

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 143–147
Resolution of (+)-abscisic acid using an
Arabidopsis glycosyltransferase
Eng-Kiat Lim,a Charlotte J. Doucet,a Bingkai Hou,a Rosamond G. Jackson,a
Suzanne R. Abramsb and Dianna J. Bowlesa,*
aCNAP, Department of Biology, University of York, York YO10 5DD, UK
bPlant Biotechnology Institute, National Research Council of Canada, 110 Gymnasium Place, Saskatoon,
Saskatchewan, Canada S7N 0W9
Received 27 October 2004; accepted 24 November 2004
Available online 23 December 2004
Abstract—Abscisic acid (ABA) can exist as two enantiomers, with (+)-ABA as the naturally occurring form. Typically, both enan-
tiomers occur in chemical preparations and both can be modified in the plant to their respective glucose esters. To identify glycos-
yltransferases capable of discriminating between the different forms of ABA, the Family 1 enzymes of Arabidopsis thaliana were
screened for activity towards ( )-ABA. Eight enzymes were found to recognise the plant hormone, with one UGT71B6 showing
enantioselective glucosylation towards (+)-ABA. UGT71B6 was used in a whole-cell biocatalysis system as a means of separating
(+)- and (À)-ABA, thereby offering an alternative to chemical synthesis for the production of pure (+)-ABA.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
number of genes encoding enzymes that glucosylate
plant hormones in a regioselective manner.15–17 It was
of interest to determine whether the enzymes from this
family were also capable of stereoselective glucosylation
and whether we would identify glycosyltransferases that
could discriminate between the enantiomers of ABA.
The results described herein show that there are eight
enzymes capable of glucosylating ABA in vitro, but only
one that recognises the naturally occurring enantiomer.
We use this glycosyltransferase in a whole-cell biocatal-
ysis system to resolve (+)-ABA.
Abscisic acid (ABA) is a plant hormone involved in the
regulation of a wide range of developmental and stress-
related events.1–3 ABA is chemically synthesised as
a racemic mixture of (+)- and (À)-stereoisomers,4
although when extracted from plant tissues, the (+)
enantiomer of ABA is recovered, indicating that this is
the natural form of the hormone.5,6 Whilst (À)-ABA is
the ÔunnaturalÕ form, nevertheless when applied to plants
or plant cell culture, the molecule has been reported to
have effects on certain known ABA responses.7–9
In planta, glucosylation of ABA is one recognised means
of controlling hormonal homeostasis,5,6,10 and interest-
ingly, when ( )-ABA racemic mixtures are applied, both
enantiomers are converted to their respective glucose
ester.5 Conjugated ABA is thought to be inactive, but
there is some evidence suggesting it represents a long-
distance transport form of the molecule.11–13
2. Results
2.1. Identification of the Arabidopsis UGTs capable of
glucosylating ( )-ABA in vitro
Members of the Arabidopsis UGT multigene family were
expressed in Escherichia coli, purified as fusion proteins
and screened for activity in vitro against ( )-ABA. The
entire family18,19 was analysed with the exception of two
sequences (UGTs 76E9 and 90A1) which did not form
soluble proteins. The reaction mixture was analysed
individually using reverse-phase HPLC. An authentic
( )-ABA glucose ester, available commercially, was
used as the reference for the chromatography. Of the
As part of our on-going studies into the Family 1 glyco-
syltransferases of Arabidopsis,14 we have identified a
*
Corresponding author. Tel.: +44 1904 328770; fax: +44 1904
328772; e-mail: djb32@york.ac.uk
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.11.062

144
E.-K. Lim et al. / Tetrahedron: Asymmetry 16 (2005) 143–147
these UGTs, UGT84B1 was found to possess the highest
catalytic activity. UGTs 73B3 and 74D1 showed only
trace activity toward ( )-ABA and were therefore not
characterised further.
2.2. Enantioselectivity of the UGTs towards ( )-ABA
Since the racemic substrate used in the initial screen con-
sisted of two enantiomers, (+)-ABA and (À)-ABA,
the activity of the UGTs were further analysed using
chiral HPLC to determine their enantioselectivity.
As described in Figure 3, the ( )-ABA glucose esters
produced by UGTs in vitro were purified and deglucosyl-
ated by b-glucosidase. The resultant enantiomers were
then analysed by chiral HPLC using authentic (+)-
ABA and (À)-ABA as references. Results are shown in
Figure 4. Analysis of the authentic enantiomers indi-
cated that (À)-ABA had less retention time on the chro-
matogram than (+)-ABA. All of the glucose esters from
the reaction mixtures yielded both (+)-ABA and (À)-
ABA, with the exception of the mix from one UGT.
Thus, UGT71B6 produced a glucose ester from ( )-
ABA that following glucosidase treatment yielded only
the (+)-ABA, indicating the enzyme had only recognised
a single enantiomer. In contrast, both enantiomers were
found in the reaction mixtures from the other UGTs,
showing the enzymes had glucosylated both (+)-ABA
and (À)-ABA. The enantiomeric excess of (+)-ABA
glucose ester in the product fraction of UGT71B6 was
92%.
Figure 1. SDS-PAGE analysis of the recombinant UGTs capable of
glucosylating ( )-ABA in vitro. The UGTs were analysed in a 10%
(w/v) polyacrylamide gel and visualised with Coomassie Brilliant
Blue staining.
105 enzymes analysed, eight UGTs (Fig. 1) were found
to form ABA glucose ester in vitro; the chromatograms
of these reaction mixes are shown in Figure 2. Amongst
Figure 3. The steps involved in the analysis of UGT enantioselectivity.
2.3. Enantioselective synthesis of (+)-ABA glucose ester
in whole-cell biocatalysis
Previous studies have shown that bacterial cells express-
ing plant UGTs can be used as whole-cell biocatalysts to
glucosylate the substrate added to the culture medium in
the absence of supplementary UDP-glucose.20,21 In this
study, E. coli cells expressing UGT71B6 were incubated
with 0.1 mM ( )-ABA. Under the conditions used, 20%
of the aglycone was converted within 24 h and the glu-
cose esters were recovered in the culture medium (Fig.
Figure 2. Reverse-phase HPLC analysis of ( )-ABA glucose ester
formed in the reaction mix of different recombinant UGTs. The
chromatograms (monitored at 270 nm) are shown at the same scale,
except from that of UGT84B1 due to the high level of glucose ester
produced in the in vitro reaction. The product peaks were quantified
using the extinction coefficient of an authentic ( )-ABA glucose ester.
The specific enzyme activity was expressed as nanomoles of ( )-ABA
glucosylated per second (nanokatal, nkat) by 1 mg of protein.

E.-K. Lim et al. / Tetrahedron: Asymmetry 16 (2005) 143–147
145
Figure 5. The use of UGT71B6 as whole-cell biocatalyst to purify the
(+)-ABA enantiomer. (A) The whole-cell biocatalysis process. (B)
Reverse-phase HPLC analysis of the culture medium from the whole-
cell biocatalysis process; insert: chiral HPLC analysis of the ( )-ABA
enantiomeric composition of the product fraction after glucose ester
purification and deglucosylation.
3. Discussion
Our study has revealed that eight Family 1 UGTs from
Arabidopsis thaliana are capable of glucosylating the
plant hormone ABA in vitro. Of these eight, only one
UGT, 71B6, recognises the natural enantiomer (+)-
ABA. There is one previous study of a recombinant en-
zyme able to glucosylate ABA in vitro.22 The gene was
cloned from Adzuki bean and the enzyme was found
to conjugate 2-trans-ABA, a biologically inactive isomer
of the hormone.23,24 The signal from the qualitative
TLC analysis of products was found to be 12-fold higher
towards 2-trans-ABA than that towards (À)-ABA, with
a 2-fold selectivity also towards the (+)-ABA compared
to (À)-ABA.
Figure 4. Chiral HPLC analysis of the ( )-ABA enantiomers released
from the glucose esters generated by different UGTs. Authentic (+)-
ABA and (À)-ABA enantiomers were used as references.
5Aand B). The glucose ester peak ( Fig. 5B) was purified
and deglucosylated as described before and the products
analysed by chiral HPLC, as shown in Figure 5B (in-
sert). As expected, (+)-ABA was the major enantiomer
present, indicating that the whole-cell biocatalysis sys-
tem could be used to separate ( )-ABA enantiomers.
The enantiomeric excess of (+)-ABA glucose ester was
slightly lower (84%) than that observed for the in vitro
system, probably reflecting a trace activity of UGT71B6
towards the (À)-ABA, when the enzyme was used in
bacterial cells.
There has been interest in the glucosylation of ABA for
many years, since the glucose ester is readily recoverable
from plant extracts and whilst known to be biologically
inactive, it has been implicated in the long-distance
transport of the hormone.11–13 Rapid progress has been
made in the identification of genes encoding enzymes

146
E.-K. Lim et al. / Tetrahedron: Asymmetry 16 (2005) 143–147
involved in the hydroxylation of ABA,25–27 the principle
means of catabolism,6 but until now little information
has been available of the nature and number of enzymes
involved in glucosylation. The UGTs described in this
report are clearly capable of recognising ( )-ABA as a
substrate, but the data provide no insight into whether
the native enzymes function in ABA homeostasis in
planta.
glucose ester (retention time, Rt = 13.4 min) was sepa-
rated from ( )-ABA (Rt = 16.5 min) by a linear gradient
of 10%–90% methanol in H2O (all solutions contained
2.5 mL/L of acetic acid and 0.4 mL/L of triethylamine)
at 1 mL/min over 20 min. The chromatogram was mon-
itored at 270 nm. Authentic ( )-ABA glucose ester pur-
chased from OlChemIm Ltd (Czech Republic) was used
as reference to confirm the ( )-ABA glucose ester pro-
duced in the in vitro reactions. An extinction coefficient
for quantification of reaction products was obtained
using authentic ( )-ABA glucose ester.
Interestingly, UGT84B1 showing the highest activity in
vitro towards ( )-ABA also glucosylates IAA and can
disturb IAA homeostasis when the gene is constitutively
over-expressed in planta.15,28 The biological significance
of this dual activity is as yet unknown, but it will be of
interest to determine whether glucosylation plays a role
in balancing the relative cellular levels of ABA and IAA.
5.3. Sample preparation for chiral HPLC
The ( )-ABA glucose ester produced by recombinant
UGTs was purified using reverse-phase HPLC equipped
with a Gilson FC 240 fraction collector. The sample was
freeze-dried in vacuo and dissolved in 75 mM sodium
phosphate, pH 5.0. The glucose ester was then treated
with b-glucosidase (Sigma) at 37 ꢁC for 1 h to release
the aglycone ( )-ABA. The completion of deglucosyla-
tion was confirmed using reverse-phase HPLC. Water–
saturated butanol was used to extract the aglycone.
The organic extract was then dried in vacuo, dissolved
in 75 mM sodium phosphate, pH 5.0, and subjected to
chiral HPLC analysis.
4. Conclusion
To our knowledge, this is the first demonstration of a
glycosyltransferase from this family recognising a sub-
strate in an enantioselective manner. The glucosylation
of (+)-ABA by UGT71B6 offers a novel means of sepa-
ration of ( )-ABA enantiomers and we have used a
whole-cell biocatalysis system to produce the (+)-ABA
glucose ester and go on to recover the single enantiomer.
This method provides an alternative to chemical synthe-
sis and chromatographic resolution for the large-scale
preparation of the natural form, (+)-ABA,29–34 and
establishes a new foundation for using glycosyltransfe-
rases for chiral resolution.
5.4. Chiral HPLC analysis
Chiral HPLC was performed using a Chiral-AGP col-
umn (100 · 4.0 mm, ChromTech). ( )-ABA enantio-
mers were separated by an isocratic mobile phase,
75 mM sodium phosphate, pH 5.0, at a flow rate of
0.9 mL/min. The chromatogram was monitored at
270 nm. (À)-ABA (Rt = 23.1 min) and (+)-ABA
(Rt = 32.2 min) was confirmed by comparing the chro-
matogram with that of the authentic enantiomers pur-
chased from Sigma.
5. Experimental
5.1. Glycosyltransferase activity assay
The construction of the UGT expression plasmids has
been described previously.35 The plasmids were trans-
formed into E. coli XL-1 Blue individually for recombi-
nant protein expression. Recombinant UGTs were
purified as fusion proteins with glutathione-S-transferase
attached to the N-terminus of the UGTs.34 105 UGTs
out of the total 107 UGTs were subjected to the initial
screen. UGTs 76E9 and 90A1 were not included due to
the inability to recover soluble recombinant proteins.
The glucosyltransferase assay was carried out following
the conditions described by Schwarzkopf and Miersch
with modifications.36 Each assay mix (200 lL) contained
1 lg of recombinant protein, 50 mM Tris–HCl, pH 7.0,
2.5 mM UDP-glucose and 0.5 mM ( )-ABA. The reac-
tion was carried out at 30 ꢁC for 1 h, and was stopped
by the addition of 20 lL of trichloroacetic acid
(240 mg/mL), quick-frozen and stored at À20 ꢁC prior
to the reverse-phase HPLC analysis.
The enantiomeric excess (eep) of the product fraction
(glucose esters) was calculated from the equation
eep = (P À Q)/(P + Q) where P and Q are the (+)-ABA
glucose ester and (À)-ABA glucose ester produced in
the process.37
5.5. Whole-cell biocatalysis with UGT71B6
The E. coli BL21 culture expressing UGT71B6 was
grown in 100 mL 2 · YT medium at 37 ꢁC until the
A600 reading reached 1.0. The bacterial cells were then
exchanged to 50 mM sodium phosphate (pH 7.4) con-
taining 10 g/L glucose, 0.1 mM isopropyl-1-thio-b-D-
galactopyranoside and 1 mM ( )-ABA. After 24 h
incubation at 28 ꢁC, the culture was centrifuged at
5000g for 5 min. The supernatant fraction (culture med-
ium) was collected for HPLC analysis described in the
Section 5.4.
5.2. Reverse-phase HPLC analysis
Reverse-phase HPLC (SpectraSYSTEM HPLC systems
and UV6000LP Photodiode Array Detector, Thermo-
Quest) analysis was carried out using a Columbus 5 l
C18 column (250 · 4.60 mm, Phenomenex). ( )-ABA
Acknowledgements
The research was supported by the Biotechnology and
Biological Sciences Research Council (BBSRC) grant

E.-K. Lim et al. / Tetrahedron: Asymmetry 16 (2005) 143–147
147
19. Paquette, S.; Møller, B. L.; Bak, S. Phytochemistry 2003,
62, 399–413.
20. Arend, J.; Waszecha, H.; Hefner, T.; Sto¨ckigt, J. Biotech-
nol. Bioeng. 2001, 76, 126–131.
87/EGA16205 and funding from The Garfield Weston
Foundation.
21. Lim, E.-K.; Ashford, D. A.; Hou, B.; Jackson, R. G.;
Bowles, D. J. Biotechnol. Bioeng. 2004, 87, 623–631.
22. Xu, Z.-J.; Nakajima, M.; Suzuki, Y.; Yamaguchi, I. Plant
Physiol. 2002, 129, 1285–1295.
23. Watanabe, H.; Takahashi, K. Plant GrowthRegul. 1999,
28, 5–8.
24. Finkelstein, R. R.; Rock, C. D. In The Arabidopsis Book;
Somerville, C., Meyerowitz, E., Eds.; American Society of
Plant Biologists, Rockville, MD, 2002.
25. Zhou, R.; Cutler, A. J.; Ambrose, S. J.; Galka, M. M.;
Nelson, K. M.; Squires, T. M.; Loewen, M. K.; Jadhav, A.
S.; Ross, A. R. S.; Taylor, D. C.; Abrams, S. R. Plant
Physiol. 2004, 134, 361–369.
References
1. Finkelstein, R. R.; Gampala, S. S. L.; Rock, C. D. Plant
Cell 2002, 14, S15–S45.
2. Zhu, J.-K. Annu. Rev. Plant Biol. 2002, 53, 247–273.
3. Himmelbach, A.; Yang, Y.; Grill, E. Curr. Opin. Plant
Biol. 2003, 6, 470–479.
4. Roberts, D. L.; Heckman, R. A.; Hege, B. P.; Bellin, S. A.
J. Org. Chem. 1968, 33, 3566–3569.
5. Oritani, T.; Kiyota, H. Nat. Prod. Rep. 2003, 20, 414–425.
6. Cutler, A. J.; Krochko, J. E. Trends Plant Sci. 1999, 4,
472–478.
26. Saito, S.; Hirai, N.; Matsumoto, C.; Ohigashi, H.; Ohta,
D.; Sakata, K.; Mizutani, M. Plant Physiol. 2004, 134,
1439–1449.
27. Kushiro, T.; Okamoto, M.; Nakabayashi, K.; Yamagishi,
K.; Kitamura, S.; Asami, T.; Hirai, N.; Koshiba, T.;
Kamiya, Y.; Nambara, E. EMBO J. 2004, 23, 1647–1656.
28. Jackson, R. G.; Kowalczyk, M.; Li, Y.; Higgins, G.; Ross,
J.; Sandberg, G.; Bowles, D. J. Plant J. 2002, 32, 573–583.
29. Kienzle, F.; Mayer, H.; Minder, R. E.; Thommen, H.
Helv. Chim. Acta 1978, 61, 2616–2627.
7. Sondheimer, E.; Galson, E. C.; Chang, Y. P.; Walton, D.
C. Science 1971, 174, 829–831.
8. Walton, D. C. In Abscisic Acid; Addicott, F. T., Ed.;
Praeger: New York, 1983; pp 113–146.
9. Balsevich, J. J.; Cutler, A. J.; Lamb, N.; Friesen, L. J.;
Kurz, E. U.; Perras, M. R.; Abrams, S. R. Plant Physiol.
1994, 106, 135–142.
10. Kleczkowski, K.; Schell, F. Crit. Rev. Plant Sci. 1995, 14,
283–298.
11. Sauter, A.; Hartung, W. J. Exp. Bot. 2000, 51, 929–935.
12. Hartung, W.; Sauter, A.; Hose, E. J. Exp. Bot. 2002, 53,
27–32.
´ˇ
30. Soukup, M.; Lukac, T.; Lohri, B.; Widmer, E. Helv. Chim.
Acta 1989, 72, 361–364.
13. Sauter, A.; Dietz, K.-J.; Hartung, W. Plant Cell Environ.
2002, 25, 223–228.
14. Lim, E.-K.; Bowles, D. J. EMBO J. 2004, 23, 2915–2922.
15. Jackson, R. G.; Lim, E.-K.; Li, Y.; Kowalczyk, M.;
Sandberg, G.; Hoggett, J.; Ashford, D. A.; Bowles, D. J.
J. Biol. Chem. 2001, 276, 4350–4356.
31. Welch, C. J. Chirality 1993, 5, 569–572.
32. Kiyota, H.; Nakabayashi, M.; Oritani, T. Tetrahedron:
Asymmetry 1999, 10, 3811–3817.
33. Hanson, J. R.; Uyanik, C. J. Chem. Res. 2003, 426–427.
34. Dunstan, D. I.; Bock, C. A.; Abrams, G. D.; Abrams, S.
R. Phytochemistry 1992, 31, 1451–1454.
35. Lim, E.-K.; Li, Y.; Parr, A.; Jackson, R.; Ashford, D. A.;
Bowles, D. J. J. Biol. Chem. 2001, 276, 4344–4349.
36. Schwarzkopf, E.; Miersch, O. Biochem. Physiol. Pflanzen
1992, 188, 57–65.
16. Lim, E.-K.; Doucet, C. J.; Elias, L.; Worrall, D.; Li, Y.;
Ross, J.; Bowles, D. J. J. Biol. Chem. 2002, 277, 586–592.
17. Hou, B.; Lim, E.-K.; Higgins, G. S.; Bowles, D. J. J. Biol.
Chem. 2004, 276, 47822–47832.
37. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J.
J. Am. Chem. Soc. 1982, 104, 7294–7299.
18. Ross, J.; Li, Y.; Lim, E.-K.; Bowles, D. J. Genome Biol.
2001, 2, 3004.1–3004.6.