α,α′-Dihydroxyketone formation using aromatic and heteroaromatic aldehydes with evolved transketolase enzymes

Article abstract of DOI:10.1039/c0cc02911d

Transketolase mutants have been identified that accept aromatic acceptors with good stereoselectivities, in particular benzaldehyde for which the wild type enzyme showed no activity.

Full text of DOI:10.1039/c0cc02911d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
a,a0-Dihydroxyketone formation using aromatic and heteroaromatic
aldehydes with evolved transketolase enzymesw
James L. Galman,a David Steadman,a Sarah Bacon,b Phattaraporn Morris,c
Mark E. B. Smith,a John M. Ward,b Paul A. Dalbyc and Helen C. Hailes*a
Received 30th July 2010, Accepted 25th August 2010
DOI: 10.1039/c0cc02911d
Transketolase mutants have been identified that accept aromatic
acceptors with good stereoselectivities, in particular benzalde-
hyde for which the wild type enzyme showed no activity.
WT-TK,13 while a yeast TK crystal structure showed that the
equivalent residue hydrogen bonds to the C-2 hydroxy group
of erythrose-4-phosphate in the active site.14 For these
reasons the D469X library was investigated with a series
of linear aldehydes (C4 to C8) and C3, C5 and C6 cyclic
carboxaldehydes.15 Excellent ees (86–99%) were observed with
the D469E mutant and variable yields (10–58%) which were
generally lower with the cyclic aldehydes.15
The a,a0-dihydroxyketones 3 produced by TK are valuable
building blocks for conversion to other chiral synthons such as
2-amino-1,3-diols.16 For example, (rac)-1,3-dihydroxy-1-
phenylpropane-2-one synthesised non-enzymatically can be
converted using a transaminase enzyme to (1rac,2S)-2-amino-
1-phenyl-1,3-propanediol, a motif present in several antibiotics
such as thiamphenicol and fluoramphenicol.17
The advantages of using biocatalysts as a sustainable resource
in synthesis are well established and include their potential to
achieve high regio- and stereoselectivities.1,2 Transketolase
(TK) (E.C.2.2.1.1) is a thiamine diphosphate (ThDP) dependent
carbon–carbon bond forming enzyme.3 In vivo it reversibly
transfers a two carbon ketol unit to D-erythrose-4-phosphate
and D-ribose-5-phosphate.3,4 To render the reaction irreversible
in vitro, b-hydroxypyruvic acid (HPA 1) has been used
extensively as the ketol donor and E. coli TK shows higher
specific activity towards 1 than yeast and spinach TKs.5,6 TKs
have been used with a range of non-phosphorylated a-hydroxy-
aldehydes, where good conversion rates and stereospecificities
for the (2R)-hydroxyaldehyde acceptor were observed, to give
(S)-a,a0-dihydroxyketones 3 (Scheme 1).4,7 Wild-type (WT) TKs
have been noted to tolerate some non-a-hydroxylated aliphatic
aldehydes, but compared to a-hydroxyaldehydes lower substrate
activities were reported.8 E. coli TK has also been overexpressed,9
and there is interest in industrial applications.10
In the current work the focus was to identify TK variants
able to accept aromatic and heteroaromatic aldehydes. Such
substrates in terms of their reactivity and structure are far
removed from the natural aldosugars used in vivo. Previous
work has highlighted the possible acceptance of aromatic and
heteroaromatic substrates. 2-Furaldehyde has been used with
E. coli TK but very low Vrels were observed,8b as was the case
for benzaldehyde and 2-furaldehyde with yeast TK, although
2-thiophene-carboxaldehyde was more readily accepted.8a
E. coli variants D469E and D469A have been used with
pyridine carboxaldehydes where conversions were observed,
but no reaction occurred with benzaldehyde or 2-furaldehyde.18
However, as previously highlighted, no data corresponding to
the TK products have ever been reported, and reactions were
typically monitored by determining HPA 1 consumption,
which can undergo slow decomposition.6,19 The use of
aldehydes with aromatic moieties have been reported using
WT-TK including phenylacetaldehyde and 2-hydroxy-phenyl-
acetaldehyde, to give products in 26% (isomeric mixture) and
54% yield ((3S,4R)-isomer), respectively.8c,20
To generate TKs with improved properties towards
hydrophobic substrates for synthetic applications, saturation
mutagenesis libraries were created, each targeted to one
TK active-site residue.11 Single point active-site mutants
were identified with improved activity towards propanal
(R = CH2CH3) that were selective for (3S)-3 and (3R)-3.12
Several variants were from the D469X library: D469E with
propanal gave 3S-3 in 90% ee and D469Y 3R-3 in 53% ee.12
The mutant D469E-TK has also been shown to decrease the
acceptance of formaldehyde and glycolaldehyde compared to
Initially, benzaldehyde 2a and Li-1 were used with
WT-TK, but no reaction was observed (Scheme 2, Table 1).
Benzaldehyde was then screened against the D469X library
utilising the tetrazolium red colorimetric assay for use with
non-a-hydroxylated aldehydes.19 From this several promising
mutants were selected: D469E, D469K, D469S, and D469T.
Interestingly, D469E, D469K and D469T had also been
identified as suitable variants for use with the cyclic aldehydes
where D469E and D469T gave (3S)-products in >97% ee.15
An engineered TK mutant F434A was also selected for
investigation as this highly conserved first-shell residue is
adjacent to the D469 site and the mutation to alanine was
therefore predicted to improve substrate acceptance for larger
Scheme
ketones 3.
1
TK catalysed reaction to generate a,a0-dihydroxy
a Department of Chemistry, University College London,
20 Gordon Street, London WC1H 0AJ, UK.
E-mail: h.c.hailes@ucl.ac.uk; Fax: +44 (0)20 7679 7463;
Tel: +44 (0)20 7679 4654
b Research Department of Structural & Molecular Biology,
University College London, Gower Street, London WC1E 6BT, UK
c Department of Biochemical Engineering, University College London,
Torrington Place, London WC1E 7JE, UK
w Electronic supplementary information (ESI) available: Character-
isation data for 3a–3f formation. See DOI: 10.1039/c0cc02911d
c
7608 Chem. Commun., 2010, 46, 7608–7610
This journal is The Royal Society of Chemistry 2010

View Article Online
Scheme 2 Formation of dihydroxyketones 3 using TK and side products 4 and 5.
Table 1 Stereoselectivities for WT-TK and TK mutant reactions with aromatic aldehydes
WT-TK ee or
isomer (yield)
D469E ee or
isomer (yield)
D469T ee or
isomer (yield)
D469K ee or
isomer (yield)
F434A ee or
isomer (yield)
Aldehyde
Product
0%
(2%)
[4 (5%)]
70% (3R)
(2%)
[4 (5%)]
82% (3R)
(2%)
[4 (5%)]
82% (3R)
(10%)
no 4
3a
(R = Ph)
No reaction
—
(5%)
[5 (o2%)]
—
(3%)
[5 (o2%)]
—
(3%)
[5 (o2%)]
—
(1%)
[5 (o2%)]
3b
(R = furyl)
No reaction
No reaction
—
(2%)
[5 (o2%)]
—
(3%)
[5 (o2%)]
—
(2%)
[5 (o2%)]
—
(1%)
[5 (o2%)]
3c
(R = thienyl)
3d
(R = CH2Ph)
93% (3S)
(5%)
90% (3S)
(50%)
96% (3S)
(50%)
95% (3S)
(50%)
97% (3S)
(48%)
88% (3S,4R)
12% (3S,4S)
(35%)
95% (3S,4R)
5% (3S,4S)
(30%)
96% (3S,4R)
4% (3S,4S)
(40%)
95% (3S,4R)
5% (3S,4S)
(38%)
85% (3S,4R)
15% (3S,4S)
(35%)
3e
(R = CH(CH3)Ph)
3f
(R = m-(OH)C6H4)
0%
(4%)
53% (3R)
(6%)
No reaction
No reaction
—
aromatic and heteroaromatic aldehydes by removing hydro-
phobic and steric interactions within the active site. From the
colorimetric assay variant D469S gave significant amounts of
product and was explored further. However, product analysis
revealed only a trace of 3a, and instead a double addition product
4 (R = Ph) was isolated, resulting from the aldol addition of 3a to
glycolaldehyde, presumably generated from decarboxylation of
the donor HPA in situ. An additional experiment was performed
using 3a (prepared using the biomimetic reaction21) with Li-1 and
D469S-TK and no 4 was formed, and was repeated with
D469T-TK with the same result. This suggested that the addition
to glycolaldehyde might occur in the active site while the inter-
mediate leading to 3a is still attached to ThDP, or with 3a when it
has been formed. However the formation of 3a, even as an
intermediate product was promising. Benzaldehyde was then used
with the other selected mutants and 3a isolated in yields of 2–10%
(Table 1), with variant F434A giving the highest yield. Care was
taken when isolating 3a due to the ease of rearrangement to 5
(R = Ph). For the first time these experiments established that as
with several other ThDP dependent enzymes, such as pyruvate
decarboxylase (PDC), TK-mutants can accept benzaldehyde even
though the yields are low. For D469E, D469K and D469T 5% of
4 (R = Ph) was also generated. By comparison, the mutant
F434A gave no double addition product 4, but 3a in 10% yield,
perhaps reflecting that with increased accessibility to the active
site, with the substitution of Phe to Ala, the product 3a can
more readily exit the active site region avoiding glycolaldehyde
addition. For HPLC analysis of the optical purities of 3a, it was
dibenzoylated and determination of the absolute stereochemistry
was achieved by the formation of the Mosher’s ester at the
primary hydroxyl with (S)- and (R)-MTPACl.22 Mutant D469E
gave 3a as a racemate, in contrast to the high stereoselectivities
observed with this mutant for the aliphatic linear and cyclic
aldehydes.15 Variants D469K, D469T and F434A gave 3a in
82% ee, 70% ee and 82% ee, respectively, and in all cases the
3R-isomer was formed predominantly, comparable to the (R)-
hydroxyketones formed using ThDP dependent PDC and
benzaldehyde lyase. By analysis of an alignment of 382
TPP-dependent enzyme sequences described previously,23 the
position equivalent to F434 is always Phe or Tyr, except in
phospho- or sulfo-pyruvate decarboxylases (Asn), and most
interestingly is Ala only in benzoyl formate decarboxylase and
benzaldehyde lyase enzymes, indicating the potential role of the
F434A mutation for acceptance of the benzene ring. Position
D469 is typically Asp or Asn and occurs naturally as Lys of Thr
only in a few PDC and PDC-related enzymes.
With several TK mutants able to accept benzaldehyde, the
use of aromatic and heteroaromatic aldehydes 2b–2f was also
investigated. Products 3b–3f were prepared for chiral assay
development from aldehydes 2b–2f using the biomimetic
reaction.21 For the determination of ees by chiral HPLC 3d
and 3e were monobenzoylated at the primary alcohol and 3f
dibenzoylated, and Moshers’ ester derivatisation of 3d, 3e, and
3f performed.22 For 2b and 2c no conversions to 3b and 3c
were observed with WT-TK. When using the D469 mutants
and F434A products 3b and 3c were formed in low yields
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 7608–7610 7609

View Article Online
o5%, with o2% of the rearranged dihydroxyketone 5 and no
4. Attempts to derivatise 3b and 3c as esters for absolute
stereochemistry and ee determination were not successful:
rearrangement to 5 occurred. Phenylacetaldehyde 2d and
(rac)-2-phenylpropionaldehyde 2e were initially used with
WT-TK and 3d and 3e formed in 5% and 35% yield, where
the more bulky a-methylated aldehyde was accepted more
readily. Phenylacetaldehyde has been used with E. coli
WT-TK to give 3d and 5 (R = CH2Ph) as a mixture of
isomers, although here only 3d was formed, possibly due to
maintenance of the pH during the reaction. The D469 mutants
and F434A led to the formation of 3d in approximately 50%
yield, and 3e in 30–40% yield. For 3d ees in the range 90–97%
(3S-isomer) were determined with the highest stereoselectivity
observed with F434A. 1H NMR analysis of 3e formed by
D469T indicated the presence of two diastereoisomers, one
major and one minor. Monobenzoylation and chiral HPLC
analysis revealed two products in a ratio of 96 : 4. Use of
(2R)-2e in the biomimetic reaction and chiral HPLC peak
correlation, together with the Mosher’s method indicated that
the major product formed using (rac)-2e with the D469
mutants and F434A was (3S,4R)-3e, and the minor isomer
(3S,4S)-3e. The D469 mutants were therefore enantioselective
for (2R)-2e, and stereoselectively formed the (3S)-isomer.
When (2R)-2e was used with D469T only (3S,4R)-3e was
formed. 2-Hydroxy-phenylacetaldehyde has been used with
E. coli WT-TK and generates the (3S,4R)-isomer.20 Despite
removal of the key hydrogen-bonding interaction between the
aldehyde C-2 hydroxyl and D469, H100, H26, indicated from
yeast TK studies, and replacement with a methyl group this
aldehyde enantioselectivity was maintained.11,14 To probe
whether a hydroxylated benzaldehyde might influence active
site hydrogen bonding interactions, the aldehyde 2f was used.
No reaction was observed with WT-TK or D469E, but with
D469T racemic 3f and F434A (3R)-3f formed (53% ee).
These results are extremely interesting, for the first time
it has been shown that selected TK mutants can accept
benzaldehyde, however the stereoselectivity observed is the
opposite to that reported for the aliphatic series. In addition,
phenylacetaldehyde and 2-methyl phenylacetaldehyde gave
products in good yields and high ees when using the single
point mutants. Higher reactivities with phenylacetaldehydes
compared to benzaldehyde may reflect increased conforma-
tional flexibilities, less steric interactions, and higher
reactivities. The yields for the formation of 3a–3c and 3f
probably reflect low rates due to poor access to the active site,
since the use of F434A clearly enhanced product formation.
However, for more productive reactions, the yield may also be
influenced by product inhibition or enzyme deactivation:
aldehyde solubility is not limiting with these substrates. This
work also highlights the importance of product isolation: TK
assays based on HPA consumption or colorimetric detection
of hydroxyketones cannot distinguish between products 3 and
4. With identification of the first TK mutants to definitively
accept benzaldehyde, further studies are now underway to
produce improved combination mutants.
and Engineering (BiCE) programme (GR/S62505/01). The
UCL Department of Chemistry are thanked for funding
D.S., the Thai government for support to P.M. and the
BBSRC for a studentship to S.B.
Notes and references
1 A. Schmid, J. S. Dordick, B. Hauer, A. Kiener, M. Wubbolts and
B. Witholt, Nature, 2001, 409, 258.
2 K. M. Koeller and C. H. Wong, Nature, 2001, 409, 232.
3 E. Racker, in The Enzymes, ed. P. D. Boyer, H. Lardy and
K. Myrzback, Academic Press, New York, 1961, vol. 5, p. 397.
4 N. J. Turner, Curr. Opin. Biotechnol., 2000, 11, 527.
5 P. Srere, J. R. Cooper, M. Tabachnick and E. Racker, Arch.
Biochem. Biophys., 1958, 74, 295.
6 G. A. Sprenger and M. Pohl, J. Mol. Catal. B: Enzym., 1999, 6, 145.
7 (a) J. Bolte, C. Demuynck and H. Samaki, Tetrahedron Lett., 1987,
28, 5525; (b) F. Effenberger, V. Null and T. Ziegler, Tetrahedron
Lett., 1992, 33, 5157; (c) L. Hecquet, J. Bolte and C. Demuynck,
Tetrahedron, 1994, 50, 8677; (d) Y. Kobori, D. C. Myles and
G. M. Whitesides, J. Org. Chem., 1992, 57, 5899;
(e) R. K. Mitra, J. M. Woodley and M. D. Lilly, Enzyme Microb.
Technol., 1998, 22, 64.
8 (a) C. Demuynck, J. Bolte, L. Hecquet and V. Dalmas, Tetra-
hedron Lett., 1991, 32, 5085; (b) G. R. Hobbs, M. D. Lilly,
N. J. Turner, J. M. Ward, A. J. Willets and J. M. Woodley,
J. Chem. Soc., Perkin Trans. 1, 1993, 165; (c) K. G. Morris, M. E.
B. Smith, N. J. Turner, M. D. Lilly, R. K. Mitra and
J. M. Woodley, Tetrahedron: Asymmetry, 1996, 7, 2185.
9 M. D. Lilly, R. Chauhan, C. French, M. Gyamerah, G. R. Hobbs,
A. Humphrey, M. Isupov, J. A. Littlechild, R. K. Mitra, K. G. Morris,
M. Rupprecht, N. J. Turner, J. M. Ward, A. J. Willetts and
J. M. Woodley, Ann. N. Y. Acad. Sci., 1996, 782, 513.
10 (a) J. Bongs, D. Hahn, U. Schorken, G. A. Sprenger and C. Wandrey,
Biotechnol. Lett., 1997, 19, 213; (b) J. Shaeri, R. Wohlgemuth and
J. M. Woodley, Org. Process Res. Dev., 2006, 10, 605; (c) J. Shaeri,
I. Wright, E. B. Rathbone, R. Wohlgemuth and J. M. Woodley,
Biotechnol. Bioeng., 2008, 101, 761.
11 (a) E. G. Hibbert, T. Senussi, S. J. Costelloe, W. Lei, M. E.
B. Smith, J. M. Ward, H. C. Hailes and P. A. Dalby,
J. Biotechnol., 2007, 131, 425; (b) E. G. Hibbert, T. Senussi, M.
E. B. Smith, S. J. Costelloe, J. M. Ward, H. C. Hailes and
P. A. Dalby, J. Biotechnol., 2008, 134, 240.
12 M. E. B. Smith, E. G. Hibbert, A. B. Jones, P. A. Dalby and
H. C. Hailes, Adv. Synth. Catal., 2008, 350, 2631.
13 U. Schorken, H. Sahm and G. A. Sprenger, in Biochemistry and
¨
Physiologyof Thiamine Diphosphate Enzymes, ed. H. Bisswanger
and A. Schellenberger, Intemann, Prien, 1996, ch. 6, pp. 543–554.
14 U. Nilsson, L. Meshalkina, Y. Lindqvist and G. Schneider, J. Biol.
Chem., 1997, 272, 1864.
15 A. Ca
´
zares, J. L. Galman, L. G. Crago, M. E. B. Smith,
os-Solıs, G. L. Lye, P. A. Dalby and
J. Strafford, L. Rı
´
H. C. Hailes, Org. Biomol. Chem., 2010, 8, 1301.
´
16 (a) C. U. Ingram, M. Bommer, M. E. B. Smith, P. A. Dalby,
J. M. Ward, H. C. Hailes and G. J. Lye, Biotechnol. Bioeng., 2007,
96, 559; (b) U. Kaulmann, K. Smithies, M. E. B. Smith, H. C. Hailes
and J. M. Ward, Enzyme Microb. Technol., 2007, 41, 628; (c) M. E.
B. Smith, B. H. Chen, E. G. Hibbert, U. Kaulmann, K. Smithies,
J. L. Galman, F. Baganz, P. A. Dalby, H. C. Hailes, G. J. Lye,
J. M. Ward, J. M. Woodley and M. Micheletti, Org. Process Res.
Dev., 2010, 14, 99.
17 K. Smithies, M. E. B. Smith, U. Kaulmann, J. L. Galman,
J. M. Ward and H. C. Hailes, Tetrahedron: Asymmetry, 2009, 20, 570.
18 U. Schorken, PhD thesis, University of Dusseldorf, 1997.
¨
¨
19 M. E. B. Smith, K. Smithies, U. Kaulmann, J. M. Ward and
H. C. Hailes, Bioorg. Med. Chem., 2006, 14, 7062.
20 A. J. Humphrey, N. J. Turner, R. McCague and S. J. C. Taylor,
J. Chem. Soc., Chem. Commun., 1995, 2475.
21 M. E. B. Smith, K. Smithies, T. Senussi, P. A. Dalby and
H. C. Hailes, Eur. J. Org. Chem., 2006, 1121.
22 J. Galman and H. C. Hailes, Tetrahedron: Asymmetry, 2009, 20, 1828.
23 S. J. Costelloe, J. M. Ward and P. A. Dalby, J. Mol. Evol., 2008,
66, 36.
The EPSRC are thanked for DTA studentships to J.L.G.
and support of the Bioconversion Integrated with Chemistry
c
7610 Chem. Commun., 2010, 46, 7608–7610
This journal is The Royal Society of Chemistry 2010