Syntheses of 2-keto-3-deoxy-D-xylonate and 2-keto-3-deoxy-L-arabinonate as stereochemical probes for demonstrating the metabolic promiscuity of sulfolobus solfataricus towards D-xylose and L-arabinose

Article abstract of DOI:10.1002/chem.201203489

Practical syntheses of 2-keto-3-deoxy-D-xylonate (D-KDX) and 2-keto-3-deoxy-L-arabinonate (L-KDA) that rely on reaction of the anion of ethyl 2-[(tert-butyldimethylsilyl)oxy]-2-(dimethoxy phosphoryl) acetate with enantiopure glyceraldehyde acetonide, followed by global deprotection of the resultant O-silyl-enol esters, have been developed. This has enabled us to confirm that a 2-keto-3-deoxy-D-gluconate aldolase from the archaeon Sulfolobus solfataricus demonstrates good activity for catalysis of the retro-aldol cleavage of both these enantiomers to afford pyruvate and glycolaldehyde. The stereochemical promiscuity of this aldolase towards these enantiomeric aldol substrates confirms that this organism employs a metabolically promiscuous pathway to catabolise the C5-sugars D-xylose and L-arabinose.

Full text of DOI:10.1002/chem.201203489

FULL PAPER
DOI: 10.1002/chem.201203489
Syntheses of 2-Keto-3-deoxy-d-xylonate and 2-Keto-3-deoxy-l-arabinonate as
Stereochemical Probes for Demonstrating the Metabolic Promiscuity of
Sulfolobus solfataricus Towards d-Xylose and l-Arabinose
Robert M. Archer,[a] Sylvain F. Royer,[b] William Mahy,[a] Caroline L. Winn,[c]
Michael J. Danson,[b] and Steven D. Bull*[a]
In memory of Grant Buchanan
Abstract: Practical syntheses of 2-keto-
3-deoxy-d-xylonate (d-KDX) and
developed. This has enabled us to con-
firm that a 2-keto-3-deoxy-d-gluconate
aldolase from the archaeon Sulfolobus
solfataricus demonstrates good activity
for catalysis of the retro-aldol cleavage
of both these enantiomers to afford
pyruvate and glycolaldehyde. The ster-
eochemical promiscuity of this aldolase
towards these enantiomeric aldol sub-
strates confirms that this organism em-
2-keto-3-deoxy-l-arabinonate (l-KDA)
that rely on reaction of the anion of
ethyl 2-[(tert-butyldimethylsilyl)oxy]-2-
(dimethoxy phosphoryl) acetate with
enantiopure glyceraldehyde acetonide,
followed by global deprotection of the
resultant O-silyl-enol esters, have been
Keywords: 3-deoxy-2-ulosonic
acid · aldolase · aldol reaction ·
carbohydrates · metabolism
ploys
a
metabolically promiscuous
pathway to catabolise the C5-sugars
d-xylose and l-arabinose.
Introduction
also capable of catabolising d-galactose to give d-glyceralde-
hyde (3) and pyruvate (4) with similar levels of activity
(Figure 1).[3] This means that these enzymes exhibit broad
substrate-specificity profiles towards diastereomeric C6 sub-
strates that are epimeric at their C4 stereocentres, which en-
ables them to transform d-glucose- and d-glyceraldehyde-
derived substrates. The stereochemically “promiscuous”
nature of this metabolic pathway[4–6] is in stark contrast to
the majority of other bacteria and eukaryotes that normally
catabolise d-galactose through the Delay–Doudoroff path-
way.[7] We have postulated that the presence of this pathway
in S. solfataricus may represent a primitive evolutionary
adaptation that occurs to enable this organism to scavenge
efficiently for different types of sugar feedstock in extreme
environments.[3]
We have recently reported findings on the catabolic path-
way that S. solfataricus employs for metabolism of the C5
sugars d-xylose and l-arabinose (Figure 2),[8] which, along
with d-glucose and d-galactose, represent the four most
commonly occurring sugars in nature. It was shown that the
glucose dehydrogenase responsible for the oxidation of
d-glucose (or d-galactose) also catalyses the oxidation of
d-xylose and l-arabinose to d-xylonate and l-arabinonate,
respectively.[8] However, the gluconate dehydratase responsi-
ble for dehydrating d-gluconate (or d-galactonate) showed
no activity towards d-xylonate or l-arabinonate.[8] Instead, a
KDX-dehydratase is employed to catalyse the dehydration
of both these C5 sugars to afford either 2-keto-3-deoxy-d-
xylonate (d-KDX, (S)-4,5-dihydroxy-2-oxopentanoic acid)
[(S)-5] and 2-keto-3-deoxy-l-arabinonate (l-KDA, (R)-4,5-
dihydroxy-2-oxopentanoic acid) [(R)-5], respectively.[8] Two
competing catabolic pathways were then shown to further
Sulfolobus solfataricus is a hyperthermophilic archaeon that
grows optimally at 80–858C and pH 2–4 and exists as an op-
portunistic heterotroph that can survive on a variety of
carbon sources.[1] S. solfataricus metabolises both d-glucose
and d-galactose through a modified non-phosphorylated var-
iant of the classical Entner–Doudoroff pathway (Figure 1).[2]
In this pathway, a glucose dehydrogenase catalyses the con-
version of d-glucose to d-gluconate, which is then dehydrat-
ed to give 2-keto-3-deoxy-d-gluconate (d-KDGlu) by a gluc-
onate dehydratase.[3] A 2-keto-3-deoxygluconate aldolase
(KDG-aldolase) then catalyses the retro-aldol cleavage of
d-KDGlu (1) to afford d-glyceraldehyde (3) and pyruvate
(4). d-Glyceraldehyde (3) is then converted into a second
molecule of pyruvate by the sequential action of a glyceral-
dehyde oxidoreductase, glycerate kinase, enolase and pyru-
vate kinase.[3] The first three enzymes of this pathway are
[a] R. M. Archer, W. Mahy, Dr. S. D. Bull
Department of Chemistry, University of Bath
Bath BA2 7AY (UK)
Fax : (+44)1225-386231
E-mail: s.d.bull@bath.ac.uk
[b] Dr. S. F. Royer, Prof. M. J. Danson
Centre for Extremophile Research
Department of Biology and Biochemistry, University of Bath,
Bath BA2 7AY (UK)
[c] Dr. C. L. Winn
Syngenta, Jealottꢀs Hill International Research Centre
Bracknell RG42 6EY (UK)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201203489.
Chem. Eur. J. 2013, 19, 2895 – 2902
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2895

Figure 1. Non-phosphorylative Entner–Doudoroff pathway for the catabolism of d-glucose and d-galactose in S. solfataricus (all sugar derivatives are rep-
resented in their open-chain form for clarity).
Figure 2. Non-phosphorylative Entner–Doudoroff catabolism of d-xylose and l-arabinose in S. solfataricus (all sugar derivatives are represented in their
open-chain form for clarity).
metabolise d-KDX [(S)-5]. Firstly, a KDX-dehydratase and
2,5-dioxopentanoate dehydrogenase combine to convert ap-
proximately 50% of the d-KDX [(S)-5] to 2-oxoglutarate
(7),[9] whereas the remaining d-KDX [(S)-5] undergoes a
KDG-aldolase mediated retro-aldol cleavage reaction to
afford pyruvate (4) and glycolaldehyde (6).
binonate with semi-purified d-xylonate dehydratase enabled
access to crude samples of d-KDX [(S)-5] and l-KDA [(R)-
5], that were then cleaved by separate incubation with
KDG-aldolase to give pyruvate (4) and glycolaldehyde (6)
in both cases.[8] However, we wished to determine the kinet-
ic parameters of KDG-aldolase towards d-KDX [(S)-5] and
l-KDA [(R)-5] so that their binding efficiencies and turn-
over rates could be compared with those determined previ-
ously for the C6-sugars d-KDGlu (1) and d-KDGal (2).[3]
Consequently, we herein report efficient syntheses of
d-KDX [(S)-5] and l-KDA [(R)-5] that have enabled us to
confirm that KDG-aldolase catalyses the rapid retro-aldol
cleavage of both these enantiomers. This activity profile
contributes further evidence that S. solfataricus employs a
metabolically promiscuous pathway to metabolise the C5
pentoses d-xylose and l-arabinose.
As part of this study,[8] it was shown that KDG-aldolase
catalyses the aldol addition reaction of pyruvate (4) with
glycolaldehyde (6) to afford 4,5-dihydroxy-2-oxo-pentanoic
acid (5) in good yield. However, we were unable to deter-
mine the ratio of d-KDX [(S)-5] to l-KDA [(R)-5] present,
and so the enantioselectivity of this aldol reaction could not
be determined. Neither d-KDX [(S)-5] or l-KDA [(R)-5]
were commercially available, therefore the retro-aldol cleav-
age activity of KDG-aldolase towards [(S)-5] and [(R)-5]
was demonstrated indirectly by using a coupled enzyme
assay.[8] That is, separate incubation of d-xylonate and l-ara-
2896
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2895 – 2902

Syntheses of 2-keto-3-deoxy-d-xylonate and 2-keto-3-deoxy-l-arabinonate
FULL PAPER
Results and Discussion
spectively; however, a final dephosphoylation step to form
l-KDA was not carried out.[27]
Syntheses and characterisation of d-KDX [(S)-5] and
l-KDA [(R)-5]: The enantiomers d-KDX [(S)-5] and
l-KDA [(R)-5] occur widely in nature as part of the lipopo-
lysaccharide capsule of certain Gram-negative bacteria,[10]
and as bacterial and fungal metabolites of d-xylose and
l-arabinose.[11–17] They have been prepared previously by
using biocatalytic protocols that rely on the action of dehy-
dratases on d-xylonate or l-arabinonate,[11–19] or the use of
aldolases to catalyse the aldol reaction of pyruvate with gly-
colaldehyde.[20–21] Syntheses of (rac)-4,5-dihydroxy-2-oxopen-
tanoic acid 5 (50:50 mixture of d-KDX [(S)-5]/L-KDA [(R)-
5]) have also been reported based on the addition of oxaloa-
cetate to glycolaldehyde 6.[13,20,22] A number of these reports
describe that d-KDX [(S)-5] (or l-KDA [(R)-5]) is unstable,
and as a consequence, only limited analytical data has been
reported for this 3-deoxy-ulosonic acid,[23] with structural
characterisation often relying on its chemical conversion to
more stable derivatives.[24] Furthermore, no data on the
enantiopurity of d-KDX [(S)-5] (or l-KDA [(R)-5]) pro-
duced from any of these procedures have been reported,
and the sign and magnitude of its specific rotation are un-
known. Structural analogues of l-KDA [(R)-5] have also
been prepared, with Limberg and Thiem having reported a
five-step b-elimination protocol for the low yielding conver-
sion of l-arabino-1,4-lactone into l-KDA-methyl ester.[25]
Alternatively, Thompson and co-workers prepared l-KDA-
methyl ester through a Wadsworth–Emmons reaction of the
anion of trichloro-tert-butyloxycarbonate-protected phos-
phono-glycolate ester (10b) with l-glyceraldehyde [(R)-9],
followed by zinc mediated O-deprotection of the resultant
adduct 8b.[26] However, in both cases, no conditions were de-
veloped to hydrolyse their methyl ester groups to give
l-KDA. Multi-step synthetic protocols have also been devel-
oped for the preparation of l-KDA-5-phosphate from
d-xylose-3,5-diphosphate or d-glucose-4-6-diphosphate, re-
It was decided to employ an adaptation of the Wads-
worth–Emmons strategy developed by Thompson et al.[26] to
prepare enantiopure d-KDX [(S)-5] and l-KDA [(R)-5] for
enzymatic studies, employing the enantiomers of glyceralde-
hyde acetonide (9) as chiral starting materials. Inspired by
methodology developed previously for the synthesis of
3-deoxy-d-manno-2-octulosonic acid[28] and d-KDGlu (1),[29]
it was proposed that reaction of the anion of ethyl 2-[(tert-
butyldimethylsilyl)oxy]-2-(dimethoxyphosphoryl)
acetate
(10a) with 2,3-O-isopropylidene-d-glyceraldehyde [(S)-9]
would result in formation of the O-silyl-enol ether (E,S)-8a,
which would then be globally deprotected to afford d-KDX
[(S)-5] (Figure 3). This chiral-pool approach would ensure
that the integrity of the stereogenic hydroxyl centre was re-
tained, whereas repeating the synthesis using 2,3-O-isopro-
pylidene-l-glyceraldehyde [(R)-9] would enable stereocom-
plementary access to enantiopure l-KDA [(R)-5].
Figure 3. Retrosynthesis of d-KDX [(S)-5].
Therefore, 2,3-O-isopropylidene-d-glyceraldehyde [(S)-
9][30] was added to a solution of ethyl 2-[(tert-butyldimethyl-
silyl)oxy]-2-(dimethoxyphosphoryl) acetate (10a) and lithi-
um hexamethyldisilazide (LiHMDS) in THF at ꢀ788C, re-
sulting in an inseparable 9:1 mixture of the O-silyl-enol
ether isomers [(E,S)-8a]/
(Scheme 1).[31] The O-silyl-enol ether functionalities of
(E,S)-8a/(Z,S)-11 were deprotected by treatment with
ACHTUNGERTN[NUNG (Z,S)-11] in 72% yield
AHCTUNGTRENNUNG
Scheme 1. Synthesis of d-KDX [(S)-5; represented in its open-chain form for clarity].
Chem. Eur. J. 2013, 19, 2895 – 2902
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2897

S. D. Bull et al.
cesium fluoride in CH3CO2H/MeCN (1:100) to afford the
a-keto-ester (S)-12 in 85% yield.[32] However, numerous at-
tempts to deprotect the acetonide and ester groups of
a-keto-ester (S)-12 under a variety of aqueous acidic condi-
tions (pH 1–5) proved unsuccessful, affording intractable
mixtures of compounds that were difficult to characterise. A
review of the literature revealed previous reports describing
the instability of d-KDX [(S)-5] (or l-KDA [(R)-5]) under
acidic conditions,[24] and as a consequence, a two-step enzy-
matic deprotection procedure for a-keto-ester (S)-12 that
would avoid exposure of d-KDX [(S)-5] to strong acid was
devised. Therefore, a-keto-ester (S)-12 was stirred in water
at pH 1.9 for 24 h, which resulted in the clean removal of
the acetonide fragment to afford diol-ester (S)-13. The resul-
tant solution of diol-ester (S)-13 was then neutralised to
pH 7.0, followed by the addition of Novozyme 435 (recombi-
nant Candida antarctica lipase B[33] on macroporous acrylic
resin). The resultant reaction mixture was stirred for 5 h,
whilst ensuring that the pH remained above 6.0, to afford
d-KDX [(S)-5] in 94% yield.[34] The resultant d-KDX [(S)-
5] could be purified further by chromatography on silica gel
using a highly polar eluent (CH2Cl2/MeOH/H2O 5:5:1);[35]
however, this purification step led to partial decomposition,
resulting in an approximately 30% loss in mass recovery.
l-KDA [(R)-5] was then prepared in an analogous manner
by using 2,3-O-isopropylidene-l-glyceraldehyde [(R)-9] as
the starting material[36–37] for this Wadsworth–Emmons/de-
protection methodology, which resulted in a 52% overall
yield over 3 steps.
assigned by comparison of the chemical shifts of their re-
spective C2-HA and C2-HB protons with those previously re-
ported for the a- and b- furanosyl anomers of the sodium
salt of d-KDGlu (1).[25] Therefore, resonances centred at d=
2.40 and 1.98 ppm were assigned to the C2-HA/C2-HB pro-
tons of the a-furanosyl anomer 5B (cf. d=2.45 and
1.95 ppm for the C2-HA/C2-HB protons of the a-furanosyl
anomer
of
the
sodium
salt
of
d-KDGlu (1)).[25] An overlapping multiplet centred at d=
2.21–2.26 was then assigned to the C2-HA/C2-HB protons of
the b-furanosyl anomer 5C (cf. d=2.24 and 2.19 ppm for
the C2-HA/C2-HB protons of the b-furanosyl anomer of
sodium salt of d-KDGlu (1)).[25] The pair of resonances at
d=103.8 and 103.9 ppm in the 13C NMR spectrum were also
diagnostic for the presence of the hemi-acetal C1 atoms of
the a- and b- furanose anomers 5B/5C, with no evidence of
any hydrate being present.
The large percentage of acyclic a-keto acid form 5A (ap-
proximately 30%) observed for l-KDA [(R)-5] (and
d-KDX [(S)-5]) is in contrast to what is observed for
d-KDGlu (1) and d-KDGal (2), which exist as rapidly inter-
converting mixtures of their a-/b-furanose and a-/b-pyra-
nose forms in aqueous solution, with very little of their re-
spective acyclic a-keto acid forms being present (<1% for
d-KDGlu (1); ꢁ3% for d-KDGal (2)).[25] It also contrasts
with the results reported from 13C NMR spectroscopic anal-
ysis of calcium-d-threo-pentulosonate (14; Figure 5a), which
exists as a mixture of its a-/b- furanose anomers, with essen-
Analysis of the 1H and
13C NMR spectroscopic data
obtained for l-KDA [(R)-5] (or
d-KDA [(S)-5]) in D2O at room
temperature revealed that it
exists as a rapidly interconvert-
Figure 5. a) The calcium salt of d-threo-pentulosonate (14) exists as a mixture of a-and b-anomers.[35] b) The
circular dichroism spectrum of the methyl a-furanose anomer 15 displays a positive Cotton effect at 210 nm.[38]
c) The circular dichroism spectrum of the methyl b-furanose anomer 16 exhibits a mixed Cotton effect with a
negative absorption at 203 nm and positive absorption at 226 nm.[38]
ing 30:33:37 mixture of its acy-
clic form 5A, and a- and b-fur-
anose forms 5B and 5C
(Figure 4). Diagnostic resonan-
tially no open-chain form being present.[37] Furthermore, it
is clear that the free carboxyl group of l-KDA [(R)-5] helps
stabilise the open-chain form 5A because the l-KDA ethyl
ester 13 exists as an essentially 50:50 mixture of its a-/b- fur-
anose epimers.
The specific rotations of d-KDX [(S)-5] ([a]2D6 = +13.5
(c=4.75 in H2O/MeOH, 1:1) and l-KDA [(R)-5] ([a]2D6 =
ꢀ13.3 (c=4.95 in H2O/MeOH, 1:1) were found to be of sim-
ilar magnitude and opposite sign, implying that these enan-
tiomers had been prepared without racemisation occurring.
The circular dichroism spectra of d-KDX [(S)-5] and
l-KDA [(R)-5] were also recorded, revealing equal and op-
posite mirrored optical refractory dispersion (ORD) curves,
as expected from their enantiomeric relationship (Figure 6).
The negative absorption at 215 nm in the CD spectrum of
l-KDA [(R)-5] is similar to CD spectra observed for other
2-keto-3-deoxy-d-ulosonic acids, which exhibit similar ab-
Figure 4. l-KDA [(R)-5] in H2O at pH 7.0 at room temperature exists as
a rapidly interconverting mixture of its open-chain keto-form 5A and its
a- and b-furanose forms 5B and 5C, respectively.
ces for the diastereotopic C3-HA and C3-HB protons of the
1
a-keto acid form 5A were observed in the H NMR spec-
trum at d=2.92 and 2.84 ppm, whilst a resonance at d=
203.8 ppm in the 13C NMR spectrum and a carbonyl absorp-
tion at 1756 cmꢀ1 in the IR spectrum confirmed the presence
of its ketone functionality. Resonances for the a- and b-
1
anomers 5B and 5C in the H NMR spectrum of (R)-5 were
2898
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2895 – 2902

Syntheses of 2-keto-3-deoxy-d-xylonate and 2-keto-3-deoxy-l-arabinonate
FULL PAPER
Figure 7. Acid-catalysed dehydration of d-KDX [(S)-5; only the a-epimer
is shown].
l-KDA [(R)-5]) is significantly more prone to dehydration
under acidic conditions than the corresponding C6-ana-
logues d-KDGlu (1) and d-KDGal (2) that only underwent
partial dehydration (<10%) under these conditions.
Kinetic assay of KDG-aldolase towards d-KDX and l-KDA
and its use as a biocatalyst to catalyse the aldol reaction of
pyruvate with glycolaldehyde: Recombinant KDG-aldolase
was expressed in Escherichia coli using our previously de-
scribed procedure and was employed directly for kinetic
assays and for carrying out a preparative aldol reaction
using pyruvate and glycolaldehyde.[3,42] The kinetic parame-
ters of KDG-aldolase towards d-KDX [(S)-5] and l-KDA
[(R)-5] were determined using our previously described,
modified thiobarbituric acid assay to determine the initial
rate of reaction.[3,42] Previous assays of KDG-aldolase activi-
ty towards d-KDGlu and d-KDGal were carried out at
708C;[3,42] however, d-KDX [(S)-5] (and l-KDA [(R)-5]) are
less thermally stable than these C6 analogues, and so their
kinetic assays were carried out at 508C. Both d-KDX [(S)-5]
and l-KDA [(R)-5] were shown to be excellent substrates
for KDG-aldolase at 508C, with the enzyme demonstrating
higher KM and kcat values for d-KDX [(S)-5] than for
l-KDA [(R)-5; Table 1]. The efficiency of KDG-aldolase in
Figure 6. The circular dichroism spectra of d-KDX [(S)-5; c], l-KDA
[(R)-5; b] and 4,5-dihydroxypentanoic acid (g; <5% ee for
d-KDX) produced from the KDG-aldolase-mediated reaction of pyru-
5
vate (4) with glycolaldehyde (6).
sorptions associated with the n!p* Cotton transitions of
their carboxyl groups.[38] The presence of a smaller absorp-
tion band of opposite sign at 235 nm is also consistent with
the CD spectra of other chiral a-hydroxy acid derivatives,
which have been assigned to n!p* Cotton transitions of
carboxyl groups of different rotational conformers present
in solution.[39] However, analysis of the CD spectrum of
l-KDA [(R)-5] (or d-KDX [(S)-5]) is further complicated by
the presence of a rapidly interconverting mixture of a-keto-
acid 5A and the a-/b-anomers 5B/5C that are present in ap-
proximately equal amounts. Redmond and co-workers have
previously described that the stable a-anomer of methyl
Table 1. Kinetic data were analysed using direct linear plots[43] (values in
parentheses are standard errors).
Substrate
KM
[mM]
kcat
[sꢀ1
kcat/KM
[sꢀ1 mmꢀ1
]
3-deoxy-a-d-manno-oct-2-ulo-furanosid)onate
(15;
Fig-
G
]
U
ure 5b) exhibits a positive Cotton effect at 210 nm, whereas
the stable b-anomer (16; Figure 5c) exhibits a mixed Cotton
effect with a minimum at 203 nm and a maximum at
226 nm.[40] Therefore, it can be speculated that the signs of
the ORD curves of the a-/b-anomeric species 5B/5C of
l-KDA [(R)-5] in the 215 nm region should contribute equal
and opposite absorptions that combine to cancel each other
out, leaving a small positive contribution to the CD spec-
trum from the b-isomer 5C at 226 nm. This combination of
absorptions would then be superimposed on to the contribu-
tion of the a-keto-acid form, 5A, to the CD spectrum,
which would contribute to the strong negative absorption
observed at 215 nm.
d-KDX [(S)-5][a]
l-KDA [(R)-5][a]
d-KDGlu (1)[b]
d-KDGal (2)[b]
33.3 (ꢂ3.3)
17.6 (ꢂ1.1)
25.7 (ꢂ1.2)
9.9 (ꢂ0.4)
15.0 (ꢂ1.12)
9.3 (ꢂ0.36)
0.45 (ꢂ0.06)
0.53 (ꢂ0.04)
1.1 (ꢂ0.2)
28.2 (ꢂ1.4)
6.8 (ꢂ0.2)
0.7 (ꢂ0.04)
[a] Kinetic parameters were determined at 508C by using a modified thi-
obarbituric acid assay.[3] [b] Kinetic parameters were determined at 708C
by using a modified thiobarbituric acid assay
catalysing the retro-aldol cleavage reaction of d-KDX
(kcat/KM=0.45sꢀ1 mmꢀ1) and l-KDA (kcat/KM =0.53sꢀ1 mmꢀ1)
at 508C are similar to the levels of efficiency previously de-
termined for the C6-homologue d-KDGal (2; kcat/KM =
0.7sꢀ1 mmꢀ1) at 708C.[3] Consequently, these results demon-
strate that KDG-aldolase exhibits good retro-aldol activity
towards both d-KDX [(S)-5] and l-KDA [(R)-5], thus pro-
viding further confirmation of the metabolic promiscuity of
Sulfolobus solfataricus towards the C5 sugars d-xylose and
l-arabinose.[8]
The general instability of d-KDX [(S)-5] under acidic con-
1
ditions was confirmed by H NMR spectroscopic analysis in
D2O at pH 1.0 over time. This analysis revealed that dehy-
dration occurs to afford dihydrofuran 17 and furan 18 as
major products after 24 h,[41] with further uncharacterised
minor products appearing after longer exposure to acid
(Figure 7). Therefore, it appears that d-KDX [(S)-5] (or
Recombinant KDG-aldolase has previously been shown
to be a versatile biocatalyst for catalysing preparative aldol
Chem. Eur. J. 2013, 19, 2895 – 2902
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2899

S. D. Bull et al.
reactions of pyruvate with a wide range of non-phosphory-
lated C3 and C4 aldoses with varying levels of stereocon-
trol.[3,44–46] For example, it catalyses the aldol reaction of pyr-
uvate and d-glyceraldehyde with essentially no stereocontrol
to give a 55:45 diastereomeric ratio of d-KDGlu (1) and
d-KDGal (2).[3] The kinetic parameters determined for
KDG-aldolase for d-KDX [(S)-5] and l-KDA [(R)-5;
Table 1) revealed that both these enantiomers were excel-
lent substrates for this aldolase. Therefore, it followed from
the principle of microscopic reversibility that KDG-aldolase
would be expected to catalyse the preparative aldol reaction
of pyruvate with glycolaldehyde with essentially no stereo-
control.
nide (9), followed by global deprotection of the resultant
O-silyl-enol esters 8a/11. This enabled us to confirm that
KDG-aldolase from the archaeon Sulfolobus solfataricus
catalyses the retro-aldol cleavage of both enantiomers of
4,5-dihydroxy-2-oxo-pentanoic acid (5) to afford pyruvate
and glycolaldehyde. These results confirm that this organism
employs the same series of enzymes to catabolize the C5
sugars d-xylose and l-arabinose as it uses to catabolise the
C6 sugars d-glucose and d-galactose. However, the thermal
and acidic lability of d-KDX and l-KDA implies that this
hyperthermophilic organism is likely to generate these spe-
cies as transient intermediates that are rapidly metabolised
in vivo. We propose that the existence of this metabolically
promiscuous pathway in Sulfolobus solfataricus represents a
primitive evolutionary feature that this hyperthermophilic
archaeon has evolved, which enables it to survive in extreme
environments. Its presence allows the organism to scavenge
efficiently for both pentose and
To confirm this hypothesis, recombinant KDG-aldolase
was used to catalyse the preparative aldol addition reaction
by using a 2-fold excess of sodium pyruvate (4) with glyco-
laldehyde (6) at 508C and pH 6.5 (Scheme 2).[47] Yeast cells
hexose sugars by using a flexi-
ble metabolic strategy that may
be indicative of the primitive
state of the organism.
Scheme 2. KDG-aldolase-catalysed aldol reaction of glycolaldehyde (6) with pyruvate (4) to afford essentially
racemic 4,5-dihydroxy-2-keto-pentanoic acid (5; <5% ee for d-KDX; represented in its open-chain form for
clarity).
Experimental Section
All reactions were performed with
starting materials and solvents ob-
tained from commercial sources with-
expressing pyruvate decarboxylase were used to remove
excess pyruvate,[48] which enabled 4,5-dihydroxy-2-oxo-pen-
tanoic acid (5) to be obtained in 54% yield after purifica-
tion by chromatography on silica gel. This enzymatically
produced aldol product (5) was shown to be essentially race-
mic by comparison of its specific rotation ([a]2D6 = +0.7 (c=
4.95 in H2O/MeOH, 1:1) with that determined for enantio-
pure d-KDX [(S)-5; [a]2D6 = +13.5 (c=4.75 in H2O/MeOH,
1:1; Scheme 2).[49] The lack of enantioselectivity in this aldol
reaction was also consistent with the CD spectrum of the
enzymatically produced product, which displayed much-re-
duced absorptions at 215 and 235 nm, consistent with the
presence of an essentially racemic aldol product containing
a small excess of d-KDX [(S)-5; <5% ee; Figure 6). There-
fore, although KDG-aldolase catalyses the aldol addition of
pyruvate (4) with glycolaldehyde (6), it proceeds with very
little stereocontrol to provide essentially racemic 4,5-dihy-
droxy-2-oxo-pentanoic acid (5).
out further purification. All dry organic solvents were used under an at-
mosphere of nitrogen. 1H NMR spectra were recorded at 300 or
500 MHz, with 13C{1H} NMR spectra recorded at 75 MHz. Chemical
shifts are quoted in parts per million and are referenced to the residual
solvent peak, with coupling constants reported in Hertz (Hz). NMR peak
assignments were confirmed by using 2D 1H COSY NMR analysis when
necessary. Infrared spectra were recorded as thin films with internal
background calibration in the range 600–4000 cmꢀ1. High resolution mass
spectra were recorded in either positive or negative mode using electro-
spray (ES) ionization. Specific rotations were recorded with a path
length of 1 dm; concentrations (c) are quoted in g/100 mL. CD spectra
were recorded by using a Jasco J-600 spectrometer with a cell length of
0.2 cm at a concentration of 6.66 mm of 4,5-dihydroxy-2-oxopentanoic
acid (5) in H2O/MeOH (1:1). Details of the synthetic protocols used to
prepare (S)-9, (R)-9 and (S)-10a are reported in the supplementary infor-
mation to this paper.
AHCTUNGTRENNUNG
A
5.87 mmol) in THF (6 mL) was added dropwise to a solution of 1m
LiHMDS in THF (6 mL) at ꢀ788C under a nitrogen atmosphere. The re-
action mixture was stirred for 10 min before a solution of aldehyde (S)-9
(0.762 g, 5.87 mmol) in THF (6 mL) was added dropwise. The reaction
mixture was then stirred for 5 min at ꢀ788C before being warmed to
08C, extracted with Et2O (90 mL), washed sequentially with HCl(aq) (1m)
and a saturated solution of NaHCO3(aq), dried (MgSO4), filtered and con-
centrated in vacuo. The residue was purified by flash chromatography
over silica gel (EtOAc/hexanes, 1:20) to give the title compound as a col-
Conclusion
Practical syntheses of 2-keto-3-deoxy-d-xylonate (d-KDX;
(S)-5) and 2-keto-3-deoxy-l-arabinonate (l-KDA; (R)-5)
have been developed and full spectroscopic data for these
biologically important 2-deoxy-ulosonic acids reported for
the first time. Their syntheses involve reaction of the anion
of ethyl 2-[(tert-butyldimethyl-silyl)oxy]-2-dimethoxyphos-
phoryl acetate (10a) with enantiopure glyceraldehyde aceto-
ourless oil (1.404 g, inseparable 9:1 mixture of (E,S)-8a/ACTHNUTRGNE(NUG Z,S)-11 isomers,
1
72%). Major (E,S)-8a isomer: H NMR (300 MHz, CDCl3) d=5.39 (1H,
d, J=7.5 Hz; CH=C), 5.17 (1H, apparent q, J=6.9 Hz; CHO), 4.12 (1H,
dd, J=8.1, 6.5 Hz; CHAHBO), 4.07 (2H, q, J=7.1 Hz; OCH2CH3), 3.42
A
(1H, dd, J=8.1, 6.8 Hz; CHAHBO), 1.29 (3H, s; CCH 3), 1.23 (3H, s;
CCHB3), 1.16 (3H, t, J=7.2 Hz; CH2CH3), 0.79 (9H, s;
CACHTUNGTRENNUNG
0.00 ppm (6H, s; SiACHTUNGTRENNUNG
2900
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2895 – 2902

Syntheses of 2-keto-3-deoxy-d-xylonate and 2-keto-3-deoxy-l-arabinonate
FULL PAPER
123.8, 109.3, 72.5, 70.0, 61.2, 26.7, 25.6, 18.2, 14.1, ꢀ4.8, ꢀ4.9 ppm. I. (thin
film) n˜max =1720 cmꢀ1 (m, C=O); HRMS (ESI): m/z calcd for
C16H30O5SiNa: 353.1760 [M+Na]+; found: 353.1744. Minor isomer (S,Z)-
11: 1H NMR (300 MHz, CDCl3) d=5.81 (1H, d, J=8.6 Hz; CH=C),
4.92–4.82 (1H, m; CHO), 4.06 (2H, q, J=7.0 Hz, OCH2CH3), 3.97 (1H,
dd, J=8.1, 6.2 Hz; CHAHBO) 3.42 (1H, dd, J=8.1, 6.7 Hz; CHAHBO),
1.29 (3H, s; CCHA3), 1.24 (3H, s; CCHB3), 1.15 (3H, t, J=7.2 Hz;
(R)-4,5-Dihydroxy-2-oxopentanoic acid (l-KDA) (5): (R)-5 was prepared
in the same way described for (S)-5 by using (R)-12 as a substrate to give
the title compound as a yellow oil (80 mg, 87%) whose spectroscopic
data were identical to those reported for (S)-5. [a]2D6 =ꢀ13.3 (c=4.95 in
H2O/MeOH, 1:1).
Expression and purification of recombinant KDG-aldolase[3]: KDG-aldo-
lase was expressed by using the pET-3a expression vector (Novagen, Not-
tingham, UK) with the gene cloned into the NdeI and BamHI restriction
CH2CH3), 0.80 (9H, s; C
(S)-Ethyl 3-(2,2-dimethyl-1,3-dioxolan-4-yl)-2-oxopropanoate
Cesium fluoride (0.46 g, 3 mmol) and acetic acid (0.43 mL, 7.5 mmol)
were added to a solution of O-silyl-enol ether (E,S)-8a/(S,Z)-11 (500 mg,
ACHTUNGTRENNU(G CH3)3), 0.03 ppm (6H, s; SiAHCTUNGTREN(NUGN CH3)2).
sites. One litre cultures of E. coli BL21ACTHUNRGTNE(NUG DE3) (Novagen) containing the
(12):
vector were grown overnight at 378C without induction. Cells were col-
lected by centrifugation at 12000ꢂg for 20 minutes and re-suspended at
0.2 gmLꢀ1 in water. These cells were lysed by 5ꢂ30 bursts of sonication
at room temperature before heat precipitation at 708C for 30 min.
Debris was removed by centrifugation at 12,000ꢂg for 20 min, and the
resulting KDG-aldolase sample was lyophilized.
AHCTUNGTRENNUNG
1.51 mmol) in dry MeCN (50 mL) at 08C. The resulting mixture was stir-
red at 08C for 30 min before being allowed to warm to room temperature
and then stirred for 18 h. The reaction mixture was then diluted with
EtOAc (150 mL) and hexane (150 mL), washed with a saturated solution
of NaHCO3(aq), dried (MgSO4) and filtered. The combined organic layers
were concentrated in vacuo to afford the title compound as a yellow oil
(286 mg, 85%). [a]2D6 = +12.0 (c=1.25 in CH2Cl2); 1H NMR (300 MHz,
CDCl3) d=4.45 (1H, quin, J=6.4 Hz; CHO), 4.24 (2H, q, J=7.2 Hz;
OCH2CH3), 4.10 (1H, dd, J=8.2, 6.1 Hz; CHAHBO), 3.52 (1H, dd, J=
8.5, 6.3 Hz; CHAHBO), 3.21 (1H, dd, J=17.6, 6.1 Hz; CHAHB), 2.92 (1H,
dd, J=17.6, 6.8 Hz; CHAHB), 1.32 (3H, s; CCHA3), 1.28 (3H, t, J=
7.2 Hz; CH3), 1.26 ppm (3H, s; CCHB3); 13C NMR (75 MHz, CDCl3) d=
192.0, 160.4, 109.3, 71.1, 69.1, 62.7, 43.8, 26.8, 25.5, 14.0 ppm; IR (thin
film) n˜max =1729 cmꢀ1 (m; C=O); HRMS (ESI): m/z cald for C10H16O5Na:
239.0895 [M+Na]+; found: 239.0881.
Kinetic study with a modified thiobarbituric acid assay[3]: Different con-
centrations of d-KDX [(S)-5] (or l-KDA [(R)-5]) (2.5–80 mm) in 50 mm
sodium phosphate buffer, pH 6.0) were prepared with a total volume of
100 mL. These samples were then incubated with purified KDG-aldolase
(5 mL; 2.1 mg in 1 mL of 20 mm Tris/HCl, pH 8.5) at 508C for 10 min
before each kinetic assay was terminated by the addition of trichloroace-
tic acid (10 mL, 12% w/v). Precipitated proteins were removed by centri-
fugation (13,000 rpm, 5 min) and 50 mL of the supernatant was oxidized
by the addition of periodic acid (25 mm) in 0.25m H2SO4 solution
(125 mL) followed by incubation at room temperature for 20 min. This
oxidation step was terminated by the addition of sodium arsenite (2%
w/v) in HCl (0.5m, 250 mL). Thiobarbituric acid (0.3% w/v, 1 mL) was
then added and the chromophore was developed by heating at 1008C for
10 min. A 25 mL sample of this solution was then removed, diluted with
water (1400 mL), and the absorbance of the solution was determined at
549 nm. The absorption coefficient of the resultant chromophore was de-
(S)-4,5-Dihydroxy-2-oxopentanoic acid (d-KDX; 5): A solution of the
a-keto-ester 12 (200 mg, 0.92 mmol) in H2O/toluene (5.0:0.25 mL) was
carefully acidified to pH 1.9 by the dropwise addition of HCl(aq) (1m).
After stirring at room temperature for 24 h, the reaction mixture was
neutralised by the dropwise addition of NaOH(aq) (1m), and then Novo-
zyme 435 (200 mg) was added. The reaction mixture was then heated to
408C and the pH of the solution was maintained between 6.0–7.5 by the
dropwise addition of NaOH(aq) (1m). After 5 h, the reaction was filtered
and concentrated in vacuo to give the title compound as a yellow oil
(0.128 g, 94% crude yield; no other major impurities were present). This
product could be purified by flash chromatography using silica gel
(CH2Cl2/MeOH:H2O, 5:5:1) to afford the title compound as a yellow oil
(65 mg, 48%). NMR spectroscopic analysis in D2O revealed that d-KDX
[(S)-5] exists as a rapidly interconverting mixture of its open chain a-
keto-acid 5A and a-/b-furanose acid anomers 5B/5C. [a]2D0 = +13.5 (c=
4.75 in H2O/MeOH, 1:1). a-Keto acid isomer 5A: 1H NMR (300 MHz,
D2O) d=4.18–4.11 (1H, m; CHOH), 3.55 (1H, dd, J=11.8, 4.2 Hz;
CHAHBOH), 3.48 (1H, dd, J=11.8, 6.4 Hz; CHAHBOH), 2.92 (1H, dd,
J=17.1, 4.4 Hz; CHAHB), 2.84 ppm (1H, dd, J=17.1, 8.3 Hz; CHAHB);
13C NMR (75 MHz, D2O) d=203.8, 169.5, 67.5, 65.0, 42.8. a-Furanose
acid anomer 5B: 1H NMR (300 MHz, D2O) d=4.55–4.49 (1H, m;
CHOH), 4.05 (1H, dd, J=9.8, 4.7 Hz; CHAHBO), 3.95 (1H, dd, J=9.8,
2.5 Hz; CHAHBO), 2.40 (1H, dd, J=14.3, 6.3 Hz; CHAHB), 1.98 ppm
(1H, m; CHAHB); 13C NMR (75 MHz, D2O) d=177.0, 103.8, 75.7, 71.4,
43.9 ppm. b-Furanose acid anomer 5C: 1H NMR (300 MHz, D2O) d=
4.51–4.47 (1H, m, CHOH), 4.09 (1H, dd, J=9.7, 4.3 Hz; CHAHBO), 3.84
(1H, dd, J=9.7, 2.1 Hz; CHAHBCO), 2.26–2.21 ppm (2H, m; CHAHB);
13C NMR (75 MHz, D2O) d=176.9, 103.9, 75.0, 70.6, 44.2 ppm. IR (thin
film) n˜max =3359 (br., OH); 2476 (br., CO2H), 1756 (m, C=O), 1710 (m,
termined to be 14519mꢀ1 cmꢀ1
.
Preparation of bakers yeast expressing pyruvate decarboxylase[47]: Type II
bakers yeast (7.0 g) was suspended in cold water (100 mL) and stirred for
18 h at <48C. The cells were harvested by centrifugation (8,500 rpm,
30 min), washed twice with cold water (2ꢂ100 mL, <48C), before re-sus-
pension of the harvested cells in cold water (20 mL) and storage at <48C
for future use.
KDG-aldolase catalysed aldol reaction between glycolaldehyde and pyru-
vate: Glycolaldehyde (6; 270 mg, 4.54 mmol) and sodium pyruvate (4;
800 mg, 9.08 mmol) were added to a solution of lyophilized KDG-aldo-
lase (5 mg) in water (23 mL, pH 6.5). The reaction was heated at 508C by
using a shaking incubator and reaction progress was followed by in-proc-
ess analysis using an Agilent 1200 HPLC fitted with a Bio-Rad Aminex
HPX-87H column (300 mmx7.8 mm; 0.6 mLminꢀ1
, 1m formic acid,
608C). Peak detection was carried out by using a refractive-index detec-
tor. Once all the glycolaldehyde had been consumed after 48 h, the reac-
tion mixture was quenched by acidification to pH 2 with HCl(aq) (1m) and
then neutralised to pH 7.0 with NaOH(aq) (2m) after 2 h. An air bubbler
was attached to the reaction vessel and a suspension of yeast cells
(20 mL) expressing pyruvate decarboxylase added. The resulting solution
was then stirred for 3 h, with the pH maintained above 6.5 by careful ad-
dition of NaOH(aq) (1m), until HPLC analysis revealed that that all the
excess pyruvate had been consumed. The residual yeast cells were re-
moved by centrifugation (3ꢂ8500 rpm, 30 min) and the mother liquors
were concentrated to give a crude reaction product that was purified by
flash chromatography on silica gel (CH2Cl2/MeOH/H2O, 5:5:1) to furnish
4,5-dihydroxy-2-oxo-pentanoic acid as a yellow oil (5; 363 mg, 54%,
<5% ee for d-KDX [(S)-5]). [a]2D6 = +0.7 (c=4.95 in H2O/MeOH, 1:1).
C=O), 1602 cmꢀ1 (s, C=O); HRMS (ESI): m/z calcd for [C5H7O5]:
ꢀ
ꢀ
147.0294 [M H] ; found: 147.0301.
ACHTUNGTRENNUNG(E,R)-Ethyl 2-[(tert-butyldimethylsilyl)oxy]-3-(2,2-dimethyl-1,3-dioxolan-
4-yl)acrylate (8a): (R)-9 was employed as a substrate under the same
protocol employed for the synthesis of (E,S)-8a to give the title com-
pound as a colourless oil (500 mg, inseparable 9:1 mixture of (E,R)-8a:-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Acknowledgements
(R)-Ethyl 3-(2,2-dimethyl-1,3-dioxolan-4-yl)-2-oxopropanoate (12): (R)-
12 was prepared in the same way described for (S)-12 by using a 9:1 mix-
ture of (E,R)-8a/ACHTUNGTRENNUNG(Z,R)-11 as substrates to give the title compound as a
We would like to thank the Biological and Biophysical Sciences Research
Council (BBSRC), the Engineering and Physical Sciences Research
Council (EPSRC) and Syngenta for funding.
yellow oil (281 mg, 86%), whose spectroscopic data were identical to
those reported for (S)-12. [a]2D6 =ꢀ8.9 (c=0.45 in CH2Cl2).
Chem. Eur. J. 2013, 19, 2895 – 2902
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2901

S. D. Bull et al.
[32] G. Bꢃlanger, F. T. Hong, L. E. Overman, B. N. Rogers, J. E. Tellew,
W. C. Trenkle, J. Org. Chem. 2002, 67, 7880.
[33] E. M. Anderson, M. Karin, O. Kirk, Biocatal. Biotransform. 1998,
16, 181.
[1] D. W. Grogan, J. Bacteriol. 1989, 171, 6710.
[2] M. De Rosa, A. Gambacorta, B. Nicolaus, P. Giardina, E. Poerio, V.
Buonocore, Biochem. J. 1984, 224, 407.
[3] H. J. Lamble, N. I. Heyer, S. D. Bull, D. W. Hough, M. J. Danson, J.
Biol. Chem. 2003, 278, 34066.
[34] For an enzymatic hydrolysis of an a-keto-ester by using Candida
rugosa lipase, see: A. Sutherland, C. L. Willis, J. Org. Chem. 1998,
63, 7764.
[35] R. Winzar, J. Philips, M. Kiefel, Synlett 2010, 583.
[36] C. Hubschwerlen, Synthesis 1986, 962.
[37] T. C. Crawford, G. C. Andrews, H. Faubl, G. N. Chmurny, J. Am.
Chem. Soc. 1980, 102, 2220.
[38] R. D. Anand, M. K. Hargreaves, Chem. Commun. 1967, 421.
[39] I. Listowsky, G. Avigad, S. Englard, J. Org. Chem. 1970, 35, 1080.
[40] P. A. Mcnicholas, M. Batley, J. W. Redmond, Carbohydr. Res. 1986,
146, 219.
[41] For similar acid-catalysed dehydrative reactions of 3-deoxy-d-
manno-oct-2-ulosonic acid (d-KDO), see: P. A. Mcnicholas, M.
Batley, J. W. Redmond, Carbohydr. Res. 1987, 165, 17.
[42] C. L. Buchanan, H. Connaris, M. J. Danson, C. D. Reeve, D. W.
Hough, Biochem. J. 1999, 343, 563.
[43] R. Eisenthal, A. Cornish-Bowden, Biochem. J. 1974, 139, 715.
[44] H. J. Lamble, S. F. Royer, D. W. Hough, M. J. Danson, G. L. Taylor,
S. D. Bull, Adv. Synth. Catal. 2007, 349, 817.
[45] H. J. Lamble, M. J. Danson, D. W. Hough, S. D. Bull, Chem.
Commun. 2005, 124.
[46] S. F. Royer, L. Haslett, S. J. Crennell, D. W. Hough, M. J. Danson,
S. D. Bull, J. Am. Chem. Soc. 2010, 132, 11753.
[47] C. H. Lin, T. Sugai, R. L. Halcomb, Y. Ichikawa, C. H. Wong, J. Am.
Chem. Soc. 1992, 114, 10138.
[48] In unpublished work, we have shown that the presence of a two-fold
excess of pyruvate is sufficient to ensure that the KDG-aldolase-cat-
alysed aldol addition reaction with d-glyceraldehyde (3) proceeds
under kinetic control, affording an unchanged 55:45 ratio of d-
KDGlu (1) and d-KDGal (2) throughout the course of the aldol re-
action. This precedent, as well as the good activity demonstrated by
KDG-aldolase for the retro-aldol cleavage of both enantiomers of
4,5-dihydroxy-2-oxopentanoic acid (5), suggests that this essentially
racemic aldol product is formed under kinetic control and is not a
consequence of thermodynamic equilibration of its (R)- and (S)
enantiomers through a reversible retro-aldol/aldol pathway.
[49] Reverse-phase HPLC analysis revealed that treatment of the crude
reaction product with bakers yeast does not lead to significant con-
sumption of aldol 5, and the optical rotation of a crude sample of an
aldol product containing excess pyruvate (prior to bakers yeast
treatment) was essentially zero. Therefore, we are confident that the
KDG-aldolase is responsible for producing the essentially racemic
aldol product 5, and that its enantiomeric ratio does not change sig-
nificantly during the bakers yeast mediated pyruvate removal step.
[4] C. C. Milburn, H. J. Lamble, A. Theodossis, S. D. Bull, D. W. Hough,
M. J. Danson, G. L. Taylor, J. Biol. Chem. 2006, 281, 14796.
[5] H. J. Lamble, A. Theodossis, C. C. Milburn, G. L. Taylor, S. D. Bull,
D. W. Hough, M. J. Danson, FEBS Lett. 2005, 579, 6865.
[6] A. Theodossis, H. Walden, E. J. Westwick, H. Connaris, H. J.
Lamble, D. W. Hough, M. J. Danson, G. L. Taylor, J. Biol. Chem.
2004, 279, 43886.
[7] J. De Ley, M. Doudoroff, J. Biol. Chem. 1957, 227, 745.
[8] C. E. M. Nunn, U. Johnsen, P. Schoenheit, T. Fuhrer, U. Sauer, D. W.
Hough, M. J. Danson, J. Biol. Chem. 2010, 285, 33701.
[9] S. J. J. Brouns, J. Walther, A. P. L. Snijders, H. J. G. V. van de Werk-
en, H. L. D. M. Willemen, P. Worm, M. G. J. de Vos, A. Andersson,
M. Lundgren, H. F. M. Mazon, R. H. H. van den Heuvel, P. Nilsson,
L. Salmon, W. M. de Vos, P. C. Wright, R. Bernander, J. van der
Oost, J. Biol. Chem. 2006, 281, 27378.
[10] B. Lindberg, K. Samuelssson, W. Nimmich, Carbohydr. Res. 1973,
30, 63.
[11] N. J. Palleroni, M. Doudoroff, J. Bacteriol. 1957, 74, 180.
[12] R. Weimberg, J. Biol. Chem. 1959, 234, 727.
[13] A. C. Stoolmiller, R. H. Abeles, J. Biol. Chem. 1966, 241, 5764.
[14] A. S. Dahms, Biochem. Biophys. Res. Commun. 1974, 60, 1433.
[15] N. J. Novick, M. E. Tyler, J. Bacteriol. 1982, 149, 364.
[16] A. M. Elshafei, Acta. Biotechnol. 1989, 9, 479.
[17] A. C. Ghosh, N. C. Mandal, Ind. J. Exp. Biol. 1998, 36, 1056.
[18] S. Watanabe, N. Shimada, K. Tajima, T. Kodaki, K. Makino, J. Biol.
Chem. 2006, 281, 33521.
[19] W. Niu, M. N. Molefe, J. W. Frost, J. Am. Chem. Soc. 2003, 125,
12998.
[20] A. C. Stoolmiller, Meth. Enzym. 1975, 41B, 101.
[21] A. M. Elshafei, O. M. Abdel-Fatah, Enz. Microb. Technol. 1989, 11,
367.
[22] C. J. Vlahos, M. A. Ghalambor, E. E. Dekker, J. Biol. Chem. 1985,
260, 5480.
[23] D. Portsmouth, Carbohydr. Res. 1968, 8, 193.
[24] N. J. Palleroni, M. Doudoroff, J. Biol. Chem. 1956, 223, 449.
[25] G. Limberg, J. Thiem, Carbohydr. Res. 1995, 275, 107.
[26] D. Horne, J. Gaudino, W. J. Thompson, Tetrahedron Lett. 1984, 25,
3529.
[27] F. Trigalo, L. Szabo, J. Chem. Soc. Perkin Trans. 1 1975, 604.
[28] H. Itoh, T. Kaneko, K. Tanami, K. Yoda, Bull. Chem. Soc. Jpn.
1988, 61, 3356.
[29] R. Plantier-Royon, D. Anker, J. Robert-Baudouy, J. Carbohydr.
Chem. 1991, 10, 239.
[30] J. Kuszmann, E. Tomori, I. Meerwald, Carbohydr. Res. 1984, 128, 87.
[31] E. Nakamura, Tetrahedron Lett. 1981, 22, 663.
Received: September 29, 2012
Revised: November 12, 2012
Published online: January 11, 2013
2902
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2895 – 2902