A Kinase-Independent One-Pot Multienzyme Cascade for an Expedient Synthesis of Guanosine 5′-Diphospho-d-mannose

Article abstract of DOI:10.1002/adsc.201600761

Biomimetic synthesis routes towards the important natural d-mannosyl donor guanosine 5′-diphospho-d-mannose (GDP-Man) rely on kinase-catalyzed nucleotide triphosphate (NTP)-dependent phosphorylations of d-mannose (Man), to give d-mannose 6-phosphate or α-d-mannose 1-phosphate (αMan 1-P) as an intermediate product. A GDP-Man synthesis not requiring the kinase/NTP system would be practical and cost-effective. Here, we have developed a multienzyme cascade towards GDP-Man, characterized in that αMan 1-P was obtained by a diastereoselective phosphatase-catalyzed phosphorylation of Man. α-d-Glucose 1-phosphate (αGlc 1-P), prepared in situ through phosphorylase-catalyzed conversion of sucrose in the presence of inorganic phosphate, was used as an expedient phosphoryl donor. The incipient αMan 1-P and guanosine triphosphate (GTP) were converted into GDP-Man by a highly manno compared to gluco selective nucleotidyltransferase. Pyrophosphatase was additionally required to hydrolyze the pyrophosphate released from the GTP, thus driving the reaction towards GDP-Man. The enzymatic cascade was operated with the αMan 1-P and the GDP-Man formation decoupled from one another (sequential mode) or having all steps run concurrently (simultaneous mode). Detailed time course analysis revealed that kinetic pull due to the constant removal of the intermediate αMan 1-P in simultaneous-mode reactions was important to promote phosphorylation of Man from αGlc 1-P in high efficiency, avoiding loss of sugar 1-phosphates by hydrolysis. Under optimized conditions for the one-pot transformation involving four enzymes, 100 mM (67 g L^?1) GDP-Man was prepared from 140 mM sucrose and phosphate, using 400 mM Man as the phosphoryl acceptor. The product was recovered by anion-exchange and size-exclusion chromatography in ≥95% purity in about 50% yield (100 mg). These results demonstrate for the first time the practical use of a phosphorylase-phosphatase combi-catalyst as an alternative to the canonical kinase for the anomeric phosphorylation of the sugar substrate in nucleoside diphospho-sugar synthesis. Phosphorylation from inorganic phosphate via the intermediate αGlc 1-P rather than from NTP, particularly GTP, appears advantageous specifically in cases where the sugar acceptor is a bulk commodity that can be applied in suitable excess to the phosphatase reaction. (Figure presented.).

Full text of DOI:10.1002/adsc.201600761

FULL PAPERS
DOI: 10.1002/adsc.201600761
A Kinase-Independent One-Pot Multienzyme Cascade for an
Expedient Synthesis of Guanosine 5’-Diphospho-d-mannose
Martin Pfeiffer,a Dominik Bulfon,a Hansjoerg Weber,b and Bernd Nidetzkya,c,*
a
Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, NAWI Graz, Petersgasse 12/I,
A-8010 Graz, Austria
b
Institute of Organic Chemistry, Graz University of Technology, NAWI Graz, Stremayrgasse 9/4, A-8010 Graz, Austria
Austrian Center of Industrial Biotechnology, Petersgasse 14, A-8010 Graz, Austria
c
Fax : (+43)-316-873-108400; phone: (+43)-316-8400; e-mail: bernd.nidetzky@tugraz.at
Received: July 15, 2016; Revised: September 30, 2016; Published online: && &&, 0000
Supporting information for this article can be found under: http://dx.doi.org/10.1002/adsc.201600761.
Abstract: Biomimetic synthesis routes towards the kinetic pull due to the constant removal of the inter-
important natural d-mannosyl donor guanosine 5’-di- mediate aMan 1-P in simultaneous-mode reactions
phospho-d-mannose (GDP-Man) rely on kinase-cat- was important to promote phosphorylation of Man
alyzed nucleotide triphosphate (NTP)-dependent from aGlc 1-P in high efficiency, avoiding loss of
phosphorylations of d-mannose (Man), to give d- sugar 1-phosphates by hydrolysis. Under optimized
mannose 6-phosphate or a-d-mannose 1-phosphate conditions for the one-pot transformation involving
(aMan 1-P) as an intermediate product. A GDP- four enzymes, 100 mM (67 gLꢀ1) GDP-Man was pre-
Man synthesis not requiring the kinase/NTP system pared from 140 mM sucrose and phosphate, using
would be practical and cost-effective. Here, we have 400 mM Man as the phosphoryl acceptor. The prod-
developed a multienzyme cascade towards GDP- uct was recovered by anion-exchange and size-exclu-
Man, characterized in that aMan 1-P was obtained sion chromatography in ꢁ95% purity in about 50%
by a diastereoselective phosphatase-catalyzed phos- yield (100 mg). These results demonstrate for the
phorylation of Man. a-d-Glucose 1-phosphate (aGlc first time the practical use of a phosphorylase-phos-
1-P), prepared in situ through phosphorylase-cata- phatase combi-catalyst as an alternative to the can-
lyzed conversion of sucrose in the presence of inor- onical kinase for the anomeric phosphorylation of
ganic phosphate, was used as an expedient phospho- the sugar substrate in nucleoside diphospho-sugar
ryl donor. The incipient aMan 1-P and guanosine tri- synthesis. Phosphorylation from inorganic phosphate
phosphate (GTP) were converted into GDP-Man by via the intermediate aGlc 1-P rather than from NTP,
a highly manno compared to gluco selective nucleoti- particularly GTP, appears advantageous specifically
dyltransferase. Pyrophosphatase was additionally re- in cases where the sugar acceptor is a bulk commodi-
quired to hydrolyze the pyrophosphate released ty that can be applied in suitable excess to the phos-
from the GTP, thus driving the reaction towards phatase reaction.
GDP-Man. The enzymatic cascade was operated
with the aMan 1-P and the GDP-Man formation de- Keywords: carbohydrates; guanosine 5’-diphospho-d-
coupled from one another (sequential mode) or mannose (GDP-Man); nucleotide sugars; one-pot
having all steps run concurrently (simultaneous multienzyme cascade; phosphorylase-phosphatase
mode). Detailed time course analysis revealed that combi-catalyst; phosphorylation
Introduction
various natural GDP-activated sugars, including
GDP-b-l-fucose, GDP-d-rhamnose and GDP-mycos-
Guanosine 5’-diphospho-d-mannose (GDP-Man), the amine, are derived from GDP-Man as the central in-
nucleotide activated form of d-mannose (Man), has termediate.[2] In addition, GDP-Man is required for
many important biological functions. It is the univer- vitamin C biosynthesis in plants.[3]
sal mannosyl donor for mannosyltransferases, a group
GDP-Man is naturally synthesized via two path-
of glycosyltransferases involved in diverse biosynthet- ways. De novo biosynthesis starts with phosphoryla-
ic tasks, including protein glycosylation in eukaryotes tion of d-glucose (Glc) by hexokinase (HK, EC
and O-antigen formation in bacteria.[1] Furthermore, 2.7.1.1). d-Glucose 6-phosphate is converted to a-d-
Adv. Synth. Catal. 0000, 000, 0 – 0
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
mannose 1-phosphate (aMan 1-P) in three steps, cata-
lyzed by glucose 6-phosphate isomerase (EC 5.3.1.9),
phosphomannose isomerase (EC 5.3.1.8) and phos-
phomannomutase (PMM, EC 5.4.2.8) via d-fructose
6-phosphate and d-mannose 6-phosphate. Finally,
aMan 1-P is transformed to GDP-Man by a GDP-
Man pyrophosphorylase (ManC, EC 2.7.7.13). GTP is
the second substrate of the reaction and pyrophos-
phate (PPi) is a co-product. An alternative route to-
wards GDP-Man is the so-called salvage pathway.[4] It
involves the direct conversion of Man into Man 6-P
by a nucleotide triphosphate-dependent kinase, like
HK. The remaining steps are the same as in the de
novo biosynthesis.[4]
Promising chemical routes towards GDP-Man com-
prise either the direct phosphorylation of unprotected
Man by phosphoric acid and tetrahydrofuran[5a] or the
MacDonald reaction,[5b] in which the peracetylated
Man is phosphorylated at the a-anomeric position.
After deprotection the aMan 1-P is reacted with gua-
nosine 5’-monophosphomorpholidate in pyrimidine
using tetrazole as the catalyst. The main disadvantage
of the method is that multiple reactions are per-
Scheme 1. Routes of GDP-Man synthesis involve aMan 1-P
as the central intermediate. Enzymatic methods for aMan 1-
P synthesis (a–c) and its conversion into GDP-Man are
shown. The abbreviations are: AGP (EC 3.1.3.10), a-d-glu-
cose 1-phosphate phosphatase; PEP, phosphoenol pyruvate;
PYR, pyruvate; other abbreviations can be found in the
text.
The anomeric kinase NahK (EC 2.7.1.162) is in-
formed in separate steps and the overall yield is volved in the metabolism of N-acetyl-d-glucosamine
therefore only moderate (phosphorylation 65%[5c] or and N-acetyl-d-galactosamine.[10] Because NahK is
72%,[5a] condensation 76%; overall ~50% based on also active with Man, although less so than on the
the Man used). Furthermore, the reaction times are natural substrates, it was exploited for aMan 1-P syn-
long (1 day for phosphorylation and deprotection,[5c]
2
thesis.[11] By using NahK, the step economy for con-
days for condensation).[5] Conditions for accelerated version of Man into aMan 1-P was further im-
condensation were reported with other sugar 1-phos- proved.[12] Li et al.[12a] demonstrated the synthesis of
phates to form nucleotide sugars,[5i] but to our knowl- GDP-Man (100 mg) in 94% yield based on the initial
edge not with aMan 1-P.
Man (15 mM) used, employing a one-pot three-
To overcome these limitations, conversions with enzyme cascade comprising NahK, ManC and PPase.
cell extracts[6] or biomimetic routes have been consid- Since GTP was applied as substrate for both NahK
ered. Elling and co-workers[7] utilized an enzymatic and ManC, the GDP-Man yield on GTP used was
synthesis according to the salvage pathway, as shown only 40%. Note that GTP is by far the most expensive
in Scheme 1a.
substrate of the overall reaction. Although the more
Typically, HK is about 30% as active with Man as expedient ATP could be used as substrate of NahK in
compared to Glc.[8] ATP was regenerated using phos- principle, its usage in a one-pot reaction is prohibited
phoenol pyruvate (PEP) and pyruvate kinase (PK, by the specificity of ManC which, besides GTP, also
EC 2.7.1.40) (Scheme 1a). Conversion of Man 6-P utilizes ATP for nucleotidyl transfer to aMan 1-P.
into GDP-Man involved the biosynthetic steps cata- Clearly, the resulting ADP-Man is an unwanted side
lyzed by PMM and ManC (Scheme 1a). Pyrophospha- product.[13]
tase (PPase, EC 3.6.1.1) was included to shift the
Therefore, a synthetic route able to combine the se-
equilibrium of the ManC reaction and also to remove lectivity of NahK regarding the anomeric phosphory-
end-product inhibition by PPi. A final product con- lation of Man with the use of an inexpensive phos-
centration of 4 mM was obtained and the yield was phoryl donor would be of high interest for a practical
80% based on the GTP used. The GDP-Man synthe- and also cost-effective GDP-Man production. We
sized within 72 h (581 mg) was isolated in 34.3% show here that a phosphatase-catalyzed selective
yield. Using a similar one-pot approach, however, ap- transphosphorylation, in which aMan 1-P is formed
plying instead of purified HK, PMM and ManC from Man and a-glucose 1-phosphate (aGlc 1-P), rep-
whole cells of E. coli, each expressing one of the resents a possible solution (Scheme 1c). GDP-Man is
three enzymes recombinantly, Honghong et al.[9] dem- thus prepared in an alternative one-pot reaction in-
onstrated formation of GDP-Man (21 mM; 680 mg) volving four enzymes without a kinase (Scheme 2).
from Man (65 mM) and an energy source (d-fructose, The phosphorylase-phosphatase cascade required to
130 mM) within 22 h. The product was not isolated generate the intermediate aMan 1-P was reported by
though.
us in a recent publication.[14] The current study advan-
Adv. Synth. Catal. 0000, 000, 0 – 0
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
Enzymatic Cascade Operated in a Sequential Mode
This approach was used first because it offered the
possibility to analyze separately the two critical trans-
formations of the overall reaction, Man!aMan 1-P
and aMan 1-P!GDP-Man. It was determined that
all enzymes are active and stable at 378C and pH 7.0
(Table S3, Supporting Information), so these condi-
tions were used without further optimization. It was
shown before by Wildberger et al.[14] that transphos-
phorylation from aGlc 1-P, coupled to in situ produc-
Scheme 2. Enzymatic cascade for one-pot synthesis of GDP-
Man. 1 sucrose, 2 aGlc 1-P, 3 aMan 1-P, 4 GDP-Man; Fru,
d-fructose; SPase (EC 2.4.1.7), sucrose phosphorylase (from
Leuconostoc mesenteroides); AGP (from Escherichia coli);
ManC (from E. coli); PPase (from E. coli).
ces the synthetic application of this cascade by dem-
onstrating for the first time its efficient integration
with a nucleotidyltransferase/PPase system. We show
intensification of GDP-Man synthesis to a final prod-
uct concentration of 70 mM at an optimized (>99%)
utilization of GTP in the overall reaction. A compara-
tive evaluation of the pros and cons of anomeric
sugar phosphorylation by phosphorylase-phosphatase
and nucleotide triphosphate-dependent kinase sys-
tems reveals their complementary use for GDP-Man
synthesis. Multistep enzymatic transformations per-
formed in one pot are powerful tools of system bioca-
talysis for nucleotide-activated sugar synthesis.[7b–d,12b]
They are topical in biocatalytic synthesis,[15a–c] includ-
ing carbohydrates in particular.[12b,15d]
Results and Discussion
The enzymatic cascade applied to the synthesis of
GDP-Man is shown in Scheme 2. All enzymes were
recombinantly expressed in E. coli and used as puri-
fied protein preparations. Generally, the enzymes
were expressed reasonably well, with purified yields
ranging from 10 to 50 mgLꢀ1. The Supporting Infor-
mation summarizes the results of enzyme production
and purification, and it also provides some basic bio- from sucrose (100 mM), phosphate (100 mM; ) and Man
chemical properties of the enzymes (Figure S1,
Table S3). Units (U) of enzyme activity refer to stan-
dard assays described in the Supporting Information.
Despite being used in a one-pot reaction, the enzy-
matic cascade can be operated in one of two possible
modes. The sequential mode involves synthesis of the
intermediate aMan 1-P decoupled from the actual
formation of GDP-Man. Alternatively, all individual
reactions are made to run simultaneously. Each mode
was evaluated in detail for GDP-Man production.
Figure 1. GDP-Man synthesis by the enzymatic cascade op-
*
erated in a sequential mode. a) Synthesis of aMan 1-P ( )
^
(400 mM) using SPase (4.5 UmLꢀ1) and AGP (0.3 UmLꢀ1).
The intermediate aGlc 1-P (&) is also shown. Besides aMan
!
1-P, small amounts of Glc 6-P ( ) are formed as side prod-
uct of the transphosphorylation. b) Conversion of aMan 1-P
into GDP-Man, showing also the consumption of GTP (&)
^
and the formation of GDP ( ). Note: the decrease in the
aMan 1-P concentration from synthesis a) to conversion
into GDP-Man b) is due to dilution by adding ManC
(1.7 UmLꢀ1), PPase (3 UmLꢀ1) and GTP (10 mM). (Experi-
ments were performed in triplicates and standard deviations
are shown with error bars.).
Adv. Synth. Catal. 0000, 000, 0 – 0
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
tion of aGlc 1-P, was best performed when the SPase were applied in 1.7-fold molar excess over GTP. Re-
activity exceeded the AGP activity by about 15-fold. sults of the one-pot reaction are summarized in Fig-
It was furthermore known that use of a Man concen- ure 2a. GTP was rapidly consumed and converted
tration exceeding that of sucrose and phosphate by quantitatively into GDP-Man. We note the actual
about 4-fold was useful to prevent loss of the inter- GDP-Man production rate (0.13 mMminꢀ1) to have
mediate aGlc 1-P to the competing enzymatic hydrol- been higher (4-fold) than expected from the limiting
ysis, also catalyzed by AGP. Figure 1 shows typical AGP activity present. The effect is explained by en-
time courses of the aMan 1-P and the GDP-Man hanced turnover of AGP under conditions in which
forming partial reactions of the overall cascade. Syn- an external acceptor (here: Man) replaces water
thesis of aMan 1-P proceeded fast and plateaued during dephosphorylation of the enzyme in the cata-
after about 90 min, giving a final product concentra- lytic cycle. Figure 2a also shows that the product was
tion of 70 mM equivalent to a 70% yield based on the stable over 20 h of incubation. The commercial prepa-
initial phosphate concentration used.
ration of GTP contains GDP (~12%). No additional
The yield based on the Man used was 18%. aGlc 1- GDP was formed in the reaction (Figure 2a). The
P was present in negligible amounts during the reac-
tion (Figure 1a), indicating that its utilization by AGP
was not rate-limiting overall. Glc 6-P was initially
(30 min) formed in low amounts but increased gradu-
ally to about 15 mM at the end of the reaction. This
result is due to the phosphoryl transfer from aGlc 1-P
to the Glc released from aGlc 1-P by hydrolysis. Note
that therefore the time of stopping the reaction is im-
portant. Had we waited longer than 90 min, the rela-
tive abundance of Glc 6-P in the product would have
increased at the expense of aMan 1-P present (data
not shown). A close mass balance for the phosphate
in the reaction was obtained.
In the second part of the overall conversion (Fig-
ure 1b), GDP-Man was synthesized with a 75% con-
version based on the aMan 1-P and GTP present.
Most of the product was formed in the first hour of
reaction, with only a slight increase in GDP-Man con-
centration during a prolonged incubation up to 18 h.
Importantly, the GTP concentration decreased even
after the GDP-Man formation had come to a halt.
GTP consumption was partly reflected by an increase
in GDP, suggesting it to be caused by hydrolysis of
the nucleoside triphosphates, probably catalyzed by
AGP. We therefore assayed AGP with GTP as sub-
strate and determined the enzyme to have a low spe-
cific activity of 0.10ꢂ0.01 Umgꢀ1 for GTP conver-
sion. This was considered to be not prohibitive for
using the enzyme in simultaneous mode.
Enzymatic Cascade Operated in a Simultaneous
Mode
Figure 2. GDP-Man synthesis by the enzymatic cascade op-
erated in a simultaneous mode using GTP as the limiting
Conditions for the simultaneous-mode reaction were
largely based on the evidence from Figure 1. The
substrate concentration, 12 mM in a) and 70 mM in b). Re-
SPase/AGP activity ratio of 15 was maintained, but
action mixtures contained AGP (0.03 UmLꢀ1), SPase
the volumetric AGP activity was decreased to 0.03
(0.45 UmLꢀ1), ManC (1.25 UmLꢀ1) and PPase (6 UmLꢀ1).
UmLꢀ1 for it to become rate-limiting for the overall
conversion. To ensure fast utilization of the in situ
formed aMan 1-P, ManC activity was used in 40-fold
excess over AGP activity. GTP was present in a limit-
The concentrations of sucrose and phosphate were 1.7 times
that of the GTP used. Man was present at 400 mM. Sym-
*
^
bols: ( ) GDP-Man, (&) GTP, ( ) GDP. (Experiments were
performed in triplicates and standard deviations are shown
ing concentration (12 mM). Sucrose and phosphate with error bars.).
Adv. Synth. Catal. 0000, 000, 0 – 0
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
small GTP hydrolase activity of AGP did not, there- on the GTP used therefore strongly improves the cost
fore, impair the synthesis. effectiveness of the process. Minimizing the GTP
Preliminary NMR analysis of the reaction product demand at the expense of the amount of Man used
showed that GDP-Man was formed in high purity. In appears to be a useful strategy. In situ regeneration of
particular, no GDP-glucose was found within limits of the GTP used in substoichiometric amounts in kinase-
detection. Product specificity of the enzymatic cas- dependent cascade transformations provides an alter-
cade can be explained by the high aMan 1-P com- native option for decreasing the GTP input to the re-
pared to aGlc 1-P substrate preference of ManC action.[7,18] However, reaction complexity must be
used.[16] Based on catalytic efficiency (kcat/Km), we de- kept in mind and the sacrificial substrates applied to
termined ManC to be 20 times more specific for uti- nucleotide triphosphate regeneration (e.g., PEP;
lization of aMan 1-P (Table S4, Supporting Informa- Scheme 1a)[7,11a,18] are also relatively expensive when
tion). In addition, the excellent specificity of AGP for compared to sucrose and Pi.
utilization of aGlc 1-P (kcat/Km =1360 mMꢀ1 sꢀ1)[17] en-
sures that the aGlc 1-P concentration is kept low (ꢃ
0.01 mM) during the reaction.
Preparative Synthesis of GDP-Man
These results reveal clear advantages of running the
enzymatic cascade in a simultaneous mode. The re- Using 70 mM GTP under the conditions of Figure 2b,
quirement of meticulous reaction control during the GDP-Man was synthesized in a total volume of 4 mL.
synthesis of aMan 1-P is eliminated. Formation of The product solution (70 mM; 50 gLꢀ1; 180 mg) was
side products of the phosphorylase-phosphatase reac- purified by a combination of anion-exchange (AEX)
tion is suppressed, presumably because aGlc 1-P does and size exclusion chromatography (SEC) (Figure 3).
not accumulate. The yield on the GTP used is higher. GDP-Man (90 mg; sodium salt) was obtained as
The reaction set-up is simpler. However, utilization of
Man was a problem that needed special attention. In
Figure 2a, the GDP-Man yield on the Man added was
only 2.5%. The product titers of 12 mM (8 gLꢀ1) were
also relatively low. Clear targets for enhanced per-
formance of the enzymatic cascade were thus identi-
fied and reactions were therefore performed at ele-
vated substrate concentrations of up to 100 mM GTP.
Biocatalytic Process Intensification
Using the base conditions of Figure 2a regarding the
enzyme activities used, conversions were done at 30,
50, 70, 80 and 100 mM GTP. Sucrose and phosphate
were employed in 1.7-fold excess over GTP in each
case. Figure 2b shows a typical time course from these
experiments.
GTP was completely utilized and GDP-Man was
produced from it in >99% yield. The reaction pro-
ceeded at an almost constant rate up to 75% conver-
sion, only to slow down slightly afterwards to reach
the full conversion (Figure 2b). Typical volumetric
productivities at ꢁ90% GTP conversion were about
9 mMhꢀ1. A final GDP-Man concentration of up to
~100 mM was obtained. This exceeds by a factor of
about 7 the highest product concentration so far re-
ported from the kinase-dependent synthesis.[12a]
The GDP-Man yield on the Man used in the reac-
tion was 40% or lower, depending on the initial GTP
concentration. While this still leaves room for im-
provement, the yield is clearly acceptable for prepara-
mental Section. The solid line shows the absorbance trace,
tive synthesis of GDP-Man. The reagent price of
GTP exceeds that of Man by three orders of magni-
Figure 3. Purification of GDP-Man by AEX a) and SEC b)
is shown. For details of the protocols used, see the Experi-
the dashed line is the conductivity trace due NaOAc in a)
and NaOAc in b). The boxed region in panel b) shows elu-
tude. A synthesis that maximizes the GDP-Man yield tion of pure GDP-Man.
Adv. Synth. Catal. 0000, 000, 0 – 0
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
AGP (0.3 UmLꢀ1) in a total volume of 400 mL. Samples
(taken every 30 min, 30 mL) were quenched by heat treat-
ment at 998C for 5 min and analyzed for Glc 6-P, aGlc 1-P
and Pi (see the Supporting Information for assays used).
After 90 min the reaction was stopped by heat inactivation
(998C, 5 min) and 10 mM GTP, 5 mM MgCl2, PPase
(3 UmLꢀ1), ManC (1.7 UmLꢀ1) and 100 mM MES buffer
(pH 7.0) supplemented with 10 mM MgCl2 were added to
reach a final volume of 2 mL and 10 mM of aMan 1-P. Sam-
ples taken (0, 1, 3, 6 and 18 h; 10 mL each) were analyzed by
HPLC and CE.
a white powder (Figure S2, Supporting Information)
after lyophilization. The purity was at least 95%
(HPLC, capillary electrophoresis, NMR), and the
overall yield was 50% (90 mg). Losses of product
1
occur mainly during the size-exclusion step. H and
13C NMR analysis of the product confirmed its identi-
ty (Figures S3 and S4, Supporting Information).
Conclusions
Reaction mixtures for conversions in simultaneous mode
contained 100 mM MES, 10 mM MgCl2, GTP (12, 30, 50, 70,
80 and 100 mM), sucrose and Pi in a 1.7-fold excess over
GTP and 400 mM Man. Reactions were started by simulta-
neously adding SPase (0.45 UmLꢀ1), AGP (0.03 UmLꢀ1),
ManC (1.25 UmLꢀ1) and PPase (6 UmLꢀ1). Samples taken
(0, 1, 2, 3, 5, 9 and 20 h; 10 mL each) were quenched by ace-
tonitrile addition (1:10, v/v) and analyzed by HPLC and CE.
We show in this study of GDP-Man synthesis that dia-
stereoselective transphosphorylation from aGlc 1-P to
Man presents an interesting and cost-effective alter-
native to direct nucleotide triphosphate-dependent
phosphorylation of Man to generate the central aMan
1-P intermediate of the overall biocatalytic cascade
reaction. We further show that the four-enzyme
system comprising SPase, AGP, ManC and PPase is
operated most effectively by running all enzymatic
steps simultaneously in a one-pot conversion, giving
GDP-Man in the highest so far reported product
titers (100 mM). The yield of GDP-Man on the GTP
used was maximized to its theoretical limit (>99%).
The utilization of Man, by contrast, was not as effi-
cient (ꢃ40% yield), thus suggesting complementary
Preparative Synthesis of GDP-Man
This was performed in a total volume of 4 mL applying the
same conditions as used in simultaneous reactions. The GTP
concentration was 70 mM. The reaction was run for 20 h
and stopped by heat inactivation (958C for 5 min). Precipi-
tated protein was removed by centrifugation (48C, 20,000 g,
synthetic uses for an enzymatic cascade involving the 15 min). To isolate GDP-Man from the reaction, AEX was
used. Per run 1 mL of the cleared reaction mixture was ap-
plied to a FliQ FPLC column (3ꢁ1 mL; 33ꢁ6.2 mm; Gener-
on, Maidenhead, U.K.) packed with SuperQ-650M (Tosh,
Tokyo, Japan) anion-exchange resin. The column was
mounted onto an ꢂKTA prime plus FPLC system (GE
Healthcare, Little Chalfont, U.K.) and operated with a flow
rate of 0.5 mLminꢀ1. The column was pre-equilibrated with
buffer A [20 mM sodium acetate (NaOAc), pH 4.6]. Elution
was done by applying a continuous gradient from 0% to
30% of buffer B (1M NaOAc, pH 4.6) in a volume of
anomeric kinase or the combination of SPase and
AGP. The kinase cascade, possibly expanded by an
enzyme module for nucleotide triphosphate regenera-
tion,[7,11a,18] is clearly the system of choice when effec-
tive utilization of the sugar acceptor is required.
Using inexpensive bulk sugars like Man, however, the
phosphorylase-phosphatase cascade offers the inter-
esting advantage that the input of GTP can be mini-
mized. Since AGP presents a broad range of aldohex-
ose substrates phosphorylated at the anomeric posi- 30 mL. After complete elution of GDP-Man at 30% of
tion,[14] the phosphorylase-phosphatase cascade de-
buffer B, a further step to 100% of buffer B was used to
elute the bound GDP. Elution was monitored by absorbance
scribed here might be applicable, with a case-specific
at 262 nm. Fractions containing GDP-Man were pooled and
adaptation of the nucleotidyltransferase, to the syn-
concentrated to a final volume of 2 mL under reduced pres-
thesis of nucleotide-activated sugars other than GDP-
sure using a rotavapor (Heidolph Laborota 4001, Schwa-
Man. Emerging high throughput-amenable techniques
bach, Germany) at 408C. To remove excess NaOAc the
pooled and concentrated fractions were split into 4ꢁ500 mL
portions and applied to SEC using a XR 16/100 column
(16ꢁ1000 mm; GE Healthcare, Little Chalfont, U.K.)
of reaction analysis are expected to facilitate the de-
velopment and optimization of the multi-enzyme cat-
alyzed transformations.[19]
packed with Sephadex G10 material (exclusion limit
<700 Da) equilibrated in H2O and mounted to an ꢂKTA
prime plus FPLC system (GE Healthcare, Little Chalfont,
U.K.). Compound elution was performed using H2O. It was
Experimental Section
monitored by absorbance at 262 nm and by measuring the
conductivity. Fractions containing salt free GDP-Man were
pooled and stored at 48C. To maximize the yield fractions
Enzymatic Cascade Reaction in Sequential and
Simultaneous Mode of Operation
Reactions were performed at 378C on a Thermomixer Com-
fort (Eppendorf, Hamburg, Germany) in 50 mM 2-(N-mor-
pholino)ethanesulfonic acid (MES) buffer (pH 7.0) and agi-
tation at 650 rpm. Sequential reactions involved conversion
of sucrose and Pi (100 mM each) by SPase (4.5 UmLꢀ1) and
phosphoryl transfer from aGlc 1-P to Man (400 mM) by
that contained GDP-Man and NaOAc were pooled, concen-
trated and reapplied to SEC. Pooled fractions containing
pure GDP-Man were lyophilized and washed with EtOH.
Excess EtOH was removed after centrifugation at 48C and
20,000 g for 10 min. The remaining precipitate was lyophi-
lized, resulting in a dry white powder. The isolated yield
Adv. Synth. Catal. 0000, 000, 0 – 0
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
thus achieved was 50% (90 mg GDP-Man) with a purity of
95%.
[2] a) P. D. Cook, H. M. Holden, Biochemistry 2008, 47,
2833–2840; b) C. J. Thibodeaux, C. E. MelanÅon, H.
Liu, Angew. Chem. 2008, 120, 9960–10007; Angew.
Chem. Int. Ed. 2008, 47, 9814–9859; c) N. Beyer, J.
Alam, T. M. Hallis, Z. Guo, H. Liu, J. Am. Chem. Soc.
2003, 125, 5584–5585; d) S. T. B. Lau, M. E. Tanner, J.
Am. Chem. Soc. 2008, 130, 17593–17602; e) M. Mꢃki,
N. Jꢃrvinen, J. Rꢃbina, H. Maaheimo, P. Mattila, R. Re-
nkonen, Glycobiology 2003, 13, 295–303.
[3] G. L. Wheeler, M. A. Jones, N. Smirnoff, Nature 1998,
393, 365–369.
[4] S. Kuettel, M. C. T. Wadum, M. L. S. Gꢄther, K.
MariÇo, C. Riemer, M. A. J. Ferguson, Mol. Microbiol.
2012, 84, 340–351.
[5] a) G.-L. Huang, X. Liu, H.-C. Zhang, P.-G. Wang, Lett.
Org. Chem. 2006, 3, 668–669; b) D. L. MacDonald, J.
Org. Chem. 1962, 27, 1107–1109; c) D. L. Hill, C. E.
Ballou, J. Biol. Chem. 1966, 241, 895–902; d) G. K.
Wagner, T. Pesnot, R. A. Field, Nat. Prod. Rep. 2009,
26, 1172–1194; e) G. M. Watt, S. L. Flitsch, S. Fey, L.
Elling, U. Kragl, Tetrahedron: Asymmetry 2000, 11,
621–628; f) V. Wittmann, C.-H. Wong, J. Org. Chem.
1997, 62, 2144–2147; g) A. Collier, G. K. Wagner,
Chem. Commun. 2008, 178–180; h) S. C. Timmons,
D. L. Jakeman, Carbohydr. Res. 2008, 343, 865–874.
[6] a) E. S. Simon, S. Grabowski, G. M. Whitesides, J. Org.
Chem. 1990, 55, 1834–1841; b) J. Preiss, E. Greenberg,
Anal. Biochem. 1967, 18, 464–471; c) S. M. Rosen, L. D.
Zeleznick, D. Frankel, I. M. Wiener, M. J. Osborn,
B. L. Horecker, Biochem. Z. 1965, 342, 375–386; d) M.-
M. Sampaio, H. Santos, W. Boos, Appl. Environ. Micro-
biol. 2003, 69, 233–240; e) W. A. Braell, M. A. Tyo, S. S.
Krag, P. W. Robbins, Anal. Biochem. 1976, 74, 484–487.
[7] a) L. Elling, J. E. Ritter, S. Verseck, Glycobiology 1996,
6, 591–597; for reviews see: b) L. Elling, Adv. Biochem.
Eng. Biotechnol. 1997, 58, 89–144; c) T. Bꢄlter, L.
Elling, Glycoconj. J. 1999, 16, 147–159; d) B. Sauer-
zapfe, L. Elling, in: Multi-step Enzyme Catalysis, (Ed.:
E. Garcia-Junceda), Wiley-VCH, Weinheim, 2008,
pp 83–108.
High Pressure Liquid Chromatography (HPLC)
Samples were analyzed on an Agilent 1200 HPLC system
(Agilent Technologies, Santa Clara, USA) equipped with
a 50ꢁ4.6 mm Kinetex C18 reversed Phase HPLC column
(Phenomenex, Torrance, USA) and a UV detector (l=
252 nm). The analytical method comprised an injection
volume of 1–5 mL, an isocratic flow of 87.5% 20 mM potassi-
um phosphate buffer supplemented with 40 mM tetrabuty-
lammonium bromide, at pH 5.9 and 12.5% acetonitrile at
a flow rate of 2 mLminꢀ1 for 4 min. GTP [retention time
(RT): 3.36 min], GDP (RT: 1.01 min), GDP sugars (RT:
0.62 min), GMP (RT: 0.45 min) were analyzed with this
method and referenced against authentic standards.
Nuclear Magnetic Resonance (NMR) Spectroscopy
Ten mg of purified GDP-Man were dissolved in 500 mL of
D2O and analyzed using a Varian (Agilent) INOVA 500-
MHz NMR spectrometer (Agilent Technologies, Santa
Clara, USA). The VNMRJ 2.2D software was used for
measurements. 1H NMR spectra (499.98 MHz) were mea-
sured on a 5 mm indirect detection PFG probe, while
a 5 mm dual direct detection probe with z-gradients was
used for 13C NMR spectra (125.71 MHz). Standard pre-satu-
ration sequence was used: relaxation delay 2 s; 908 proton
pulse; 2.048 s acquisition time; spectral width 8 kHz;
number of points 32 k. HSQC spectra were measured with
128 scans per increment and adiabatic carbon 1808 pulses.
ACD/NMR Processor Academic Edition 12.0 (Advanced
Chemistry Development Inc., Toronto, Canada) was used
for evaluation of spectra.
For more experimental details, see the Supporting Infor-
mation.
[8] L. Berger, M. W. Slein, S. P. Colowick, C. F. Cori, J.
Acknowledgements
Gen. Physiol. 1946, 29, 379–391.
[9] J. Honghong, L. Fuping, L. Yu, L. Xiaoguang, L.
Yihan, W. Hongbin, L. Jing, C. Yueting, Biotechnol.
Lett. 2011, 33, 1145–1150.
Financial support from the Austrian Science Fund FWF (DK
Molecular Enzymology W901, to B.N.) is gratefully acknowl-
edged. Sandra Kulmer (Institute of Biotechnology and Bio-
chemical Engineering, Graz University of Technology) per-
formed cloning of the PPase into the expression vector pET-
STREP3. B.N. acknowledges the EU COST action CM1303
(Systems Biocatalysis SysBiocat).
[10] M. Nishimoto, M. Kitaoka, Appl. Environ. Microbiol.
2007, 73, 6444–6449.
[11] a) Y. Liu, M. Nishimoto, M. Kitaoka, Carbohydr. Res.
2015, 401, 1–4; b) Y. Li, H. Yu, Y. Chen, K. Lau, L.
Cai, H. Cao, V. K. Tiwari, J. Qu, V. Thon, P. G. Wang,
X. Chen, Molecules 2011, 16, 6396–6407.
[12] a) L. Li, Y. Liu, Y. Wan, Y. Li, X. Chen, W. Zhao, P. G.
Wang, Org. Lett. 2013, 15, 5528–5530; b) for reviews,
see: H. Yu, X. Chen, Org. Biomol. Chem. 2016, 14,
2809–2818; c) M. M. Muthana, J. Qu, Y. Li, L. Zhang,
H. Yu, L. Ding, H. Malekan, X. Chen, Chem.
Commun. 2012, 48, 2728–2730.
References
[1] a) G. Samuel, P. Reeves, Carbohydr. Res. 2003, 338,
2503–2519; b) I. Imhof, I. Flury, C. Vionnet, C. Rouba-
ty, D. Egger, A. Conzelmann, J. Biol. Chem. 2004, 279,
19614–19627; c) A. Halim, I. S. B. Larsen, P. Neubert,
H. J. Joshi, B. L. Petersen, S. Y. Vakhrushev, S. Strahl,
H. Clausen, Proc. Natl. Acad. Sci. USA 2015, 112,
15648–15653; d) A. Helenius, M. Aebi, Science 2001,
291, 2364–2369.
[13] R. M. Mizanur, N. L. B. Pohl, Org. Biomol. Chem.
2009, 7, 2135.
[14] P. Wildberger, M. Pfeiffer, L. Brecker, B. Nidetzky,
Angew. Chem. 2015, 127, 16093–16097; Angew. Chem.
Int. Ed. 2015, 54, 15867–15871.
Adv. Synth. Catal. 0000, 000, 0 – 0
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
[15] a) E. Garcia-Junceda, I. Lavandera, D. Rother, J.
Schrittwieser, J. Mol. Catal. B: Enzym. 2015, 114, 1–6;
b) E. Garcia-Junceda, Multi-step Enzyme Catalysis,
Wiley-VCH, Weinheim, 2008; c) E. Ricca, B. Brucher,
J. Schrittwieser, Adv. Synth. Catal. 2011, 353, 2239–
2262; d) M. Nishimoto, M. Kitaoka, Biosci. Biotechnol.
Biochem. 2007, 71, 2101–2104.
[16] Y.-H. Yang, Y.-B. Kang, K.-W. Lee, T.-H. Lee, S.-S.
Park, B.-Y. Hwang, B.-G. Kim, J. Mol. Catal. B:
Enzym. 2005, 37, 1–8.
[18] a) H. Zhao, W. A. van der Donk, Curr. Opin. Biotech-
nol. 2003, 14, 583–589; b) W. Berke, H.-J. Schꢄz, C.
Wandrey, M. Morr, G. Denda, M.-R. Kula, Biotechnol.
Bioeng. 1988, 32, 130–139; c) J. N. Andexer, M. Richter,
ChemBioChem 2015, 16, 380–386; d) Z. Zou, Q. Ding,
L. Ou, B. Yan, Appl. Microbiol. Biotechnol. 2013, 97,
9389–9395; e) A. Kameda, T. Shiba, Y. Kawazoe, Y.
Satoh, Y. Ihara, M. Munekata, K. Ishige, T. Noguchi, J.
Biosci. Bioeng. 2001, 91, 557–563.
[19] C. Wahl, D. Hirtz, L. Elling, Biotechnol. J. 2016, 11,
1298–1308 .
[17] P. Wildberger, M. Pfeiffer, L. Brecker, G. N. Rechberg-
er, R. Birner-Gruenberger, B. Nidetzky, Appl. Environ.
Microbiol. 2015, 81, 1559–1572.
Adv. Synth. Catal. 0000, 000, 0 – 0
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
A Kinase-Independent One-Pot Multienzyme Cascade for
an Expedient Synthesis of Guanosine 5’-Diphospho-d-
mannose
9
Adv. Synth. Catal. 2016, 358, 1 – 9
Martin Pfeiffer, Dominik Bulfon, Hansjoerg Weber,
Bernd Nidetzky*
Adv. Synth. Catal. 0000, 000, 0 – 0
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!