Enzyme Module Systems for the Synthesis of Uridine 5′-Diphospho-α- D -glucuronic Acid and Non-Sulfated Human Natural Killer Cell-1 (HNK-1) Epitope

Article abstract of DOI:10.1002/adsc.201500180

Tailor-made strategies for the stereo- and regioselective multi-step enzymatic synthesis of glycoconjugates require well characterized glycosyltransferases and carbohydrate modifying enzymes. We here report on a novel enzyme cascade for the synthesis of uridine 5′-diphospho-α-D-glucuronic acid (UDP-GlcA) and the non-sulfated human natural killer cell-1 (HNK-1) epitope including in situ regeneration of UDP-GlcA and the cofactor nicotinamide adenine dinucleotide NAD⁺ by the combination of four enzymes in one-pot. In the first enzyme module sucrose synthase 1 (SuSy1) is used to produce uridine 5′-diphospho-α-D-glucose (UDP-Glc) from sucrose and uridine 5′-diphosphate (UDP). The combination with UDP-Glc dehydrogenase in the second enzyme module leads to the synthesis of UDP-GlcA with concomitant in situ regeneration of the cofactor NAD⁺ by nicotinamide adenine dinucleotide hydride (NADH)-oxidase. In the third enzyme module the mammalian glucuronyltransferase GlcAT-P catalyzes the synthesis of the non-sulfated HNK-1 epitope by regioselective transfer of GlcA onto N-acetyllactosamine type 2 (LacNAc type 2). We present a comprehensive study on substrate kinetics, substrate specificities, variation and relation of enzyme activities as well as cross inhibition of intermediate products. With optimized reaction conditions we obtain superior product yields with streamlined synthesis costs for the expensive nucleotide sugar UDP-GlcA and cofactor NAD⁺.

Full text of DOI:10.1002/adsc.201500180

FULL PAPERS
DOI: 10.1002/adsc.201500180
Enzyme Module Systems for the Synthesis of Uridine 5’-Diphos-
pho-a-d-glucuronic Acid and Non-Sulfated Human Natural Killer
Cell-1 (HNK-1) Epitope
a
a
b
a,
Leonie Engels, Manja Henze, Werner Hummel, and Lothar Elling *
a
Laboratory for Biomaterials, Institute of Biotechnology and Helmholtz-Institute for Biomedical Engineering, RWTH
Aachen University, Pauwelsstraße 20, 52074 Aachen, Germany
Fax: (+49)-241-802-387; phone: (+49)-241-802-8350; e-mail: l.elling@biotec.rwth-aachen.de
Institute for Organic Chemistry I, University Bielefeld, 33501 Bielefeld, Germany
b
Received: February 20, 2015; Revised: April 19, 2015; Published online: May 13, 2015
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201500180.
+
Abstract: Tailor-made strategies for the stereo- and NAD by nicotinamide adenine dinucleotide hydride
regioselective multi-step enzymatic synthesis of gly- (NADH)-oxidase. In the third enzyme module the
coconjugates require well characterized glycosyl- mammalian glucuronyltransferase GlcAT-P catalyzes
transferases and carbohydrate modifying enzymes. the synthesis of the non-sulfated HNK-1 epitope by
We here report on a novel enzyme cascade for the regioselective transfer of GlcA onto N-acetyllactosa-
synthesis of uridine 5’-diphospho-a-d-glucuronic acid mine type 2 (LacNAc type 2). We present a compre-
(
UDP-GlcA) and the non-sulfated human natural hensive study on substrate kinetics, substrate specif-
killer cell-1 (HNK-1) epitope including in situ regen- icities, variation and relation of enzyme activities as
eration of UDP-GlcA and the cofactor nicotinamide well as cross inhibition of intermediate products.
+
adenine dinucleotide NAD by the combination of With optimized reaction conditions we obtain superi-
four enzymes in one-pot. In the first enzyme module or product yields with streamlined synthesis costs for
sucrose synthase 1 (SuSy1) is used to produce uri- the expensive nucleotide sugar UDP-GlcA and co-
+
dine 5’-diphospho-a-d-glucose (UDP-Glc) from su- factor NAD .
crose and uridine 5’-diphosphate (UDP). The combi-
nation with UDP-Glc dehydrogenase in the second
enzyme module leads to the synthesis of UDP-GlcA Keywords: carbohydrates; cofactor; enzyme cataly-
with concomitant in situ regeneration of the cofactor sis; glycoconjugates; glycosides
Introduction
and is supposed to regulate their biocatalytic activity.
Laminin 1 of the kidney cortex was reported to bear
The sulfated carbohydrate epitope, human natural non-sulfated HNK-1 showing strong influence on cell-
killer-1 (HNK-1), SO -3GlcA(b1-3)Gal(b1-4)GlcNAc, matrix interaction between laminin 1 and epithelial
3
[
5]
is involved in many physiological processes of mam- cells. In summary, HNK-1 and its non-sulfated pre-
[1]
mals. The sulfated glycan motif is mainly expressed cursor are interesting epitopes for the biofunctionali-
in the nervous system and temporarily regulated zation of biomaterial surfaces to exploit basic molecu-
during the development. The epitope is found on lar mechanisms of cell–cell and cell–matrix interac-
neural cell adhesion molecules (NCAM), myelin pro- tions. Chemical synthesis of HNK-1 derivatives was
tein zero (P0), myelin-associated glycoprotein demonstrated employing laborious and time-consum-
[
6]
(
MAG), extracellular matrix proteins (Tenascin-R, ing protection and deprotection steps. Limited syn-
Phosphacan) and sulfo-glucuronylglycolipids (SGGL- thetic access to these glycan epitopes led to the
[
2]
[7]
1
, SGGL-2) displaying a functional role in cell mi- screening of peptide mimetics, which were used for
[
3]
gration, cell adhesion and neurite outgrowth. Inter- the functionalization of collagen scaffolds to promote
estingly, HNK-1 epitope functionalized biomaterial neurite outgrowth of motor neurons. In contrast, en-
induced neurite outgrowth of motor neurons. The zymatic synthesis is a proved alternative strategy due
non-sulfated HNK-1 epitope is found on metallo-pro- to the high stereo- and regioselectivity of the involved
teases (Meprin, aminopeptidase N) in mice kidneys
[
4]
[
4]
[
4b]
glycosyltransferases and carbohydrate modifying en-
Adv. Synth. Catal. 2015, 357, 1751 – 1762
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1751

Leonie Engels et al.
FULL PAPERS
Scheme 1. EMS for the synthesis of non-sulfated HNK-1 (4) epitope with in situ regeneration of UDP-GlcA (2) and cofactor
+
NAD by combining four enzymes in one-pot. Enzyme module A: SuSy module for the synthesis of UDP-Glc (1). Enzyme
+
module B: UGDH module for the synthesis of 2 with regeneration of the cofactor NAD by NOX. Enzyme module C:
GlcAT module for the synthesis of 4. The by-product UDP enters enzyme module A for regeneration of 2 via enzyme
module B.
[
12]
zymes. The biosynthesis of the HNK-1 epitope is on a neo-lactohexaose glycan structure. Subsequent
[
13]
based on the successive transfer of d-galactose (Gal) chemical sulfation yielded sulfated HNK-1.
and d-glucuronic acid (GlcA) onto N-acetyl-d-glucos- The biocatalytic in vitro syntheses of glycoconju-
amine (GlcNAc) of branched N- and O-glycans as gates by cascade reactions in defined enzyme-module
well as glycosphingolipids with subsequent site-specif- systems (EMS) give better control of product forma-
[
8]
[14]
ic sulfation of GlcA (Scheme S1, Supporting Infor- tion under defined reaction conditions. This strat-
mation). egy allows the combination of pathway and non-relat-
The activity of b4Gal-T1 to generate the N-acetyl- ed pathway enzymes from various heterologous sour-
lactosamine (LacNAc) type 2, Gal(b1-4)GlcNAc, ac- ces as well as the variation of substrates. One-pot syn-
ceptor on N-glycans has been shown to be essential thesis facilitates integration of in situ regeneration
for the expression of the HNK-1 epitope in the cycles of nucleotide sugars and avoids laborious isola-
[8c]
brain. Eukaryotic glucuronyltransferase-P (GlcAT- tion of intermediate products resulting in higher prod-
[
15]
P) from rat, mouse and man were cloned and ex- uct yields with minimized process costs.
pressed in different hosts. Recombinant human
[
9]
We here present a novel flexible EMS for the syn-
GlcAT-P produced in E. coli was characterized as thesis of the non-sulfated HNK-1 epitope by the com-
a homo-dimer. GlcAT-P is specific for the transfer of bination of three enzyme modules (Scheme 1). The
GlcA onto LacNAc type 2 as well as the glycoprotein sucrose synthase (SuSy) module generates UDP-a-d-
acceptors neural cell adhesion molecule (NCAM) and glucose (UDP-Glc, 1) from sucrose and UDP with re-
[
16]
asialo-orosomucoid (ASOR), whereas GlcAT-S has combinant SuSy1 (EC 2.4.1.13). from potato. UDP-
a dual specificity for LacNAc type 2 and type Glc is further converted in the second UDP-Glc-de-
[
3,10]
1
(Galb1-3GlcNAc).
This is in accordance with hydrogenase (UGDH) module to UDP-a-d-glucuron-
data for human GlcAT-P expressed in Pichia pasto- ic acid (UDP-GlcA, 2) by recombinant human UDP-
[11]
ris.
GlcAT-P from rat brain expressed in COS- Glc-dehydrogenase (EC 1.1.1.22). NADH oxidase
1
cells (derived from monkey kidney tissue) showed (NOX) from Lactobacillus brevis is integrated for in
+
[17]
also activity toward paragloboside, a glycosphingolipid situ regeneration of NAD
with production of H O
2
[3]
acceptor. Based on these data, murine GlcAT-P was as by-product. In the third glucuronyltransferase
utilized in a metabolic engineered E. coli K12 strain (GlcAT) module recombinant murine glucuronyl-
for the synthesis of the non-sulfated HNK-1 epitope transferase (GlcAT-P, EC 2.4.1.17) transfers GlcA
1752
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1751 – 1762

FULL PAPERS
Module Systems for the Synthesis of Uridine 5’-Diphospho-a-d-glucuronic Acid
[
20b,21,22]
onto linker modified LacNAc type 2 [Gal(b1- with previously published data.
Further
[
15a]
4
)GlcNAcb1-linker-NH -t-Boc, 3]
to yield the non- enzyme kinetic studies showed no inhibition by UDP-
2
sulfated HNK-1 epitope (4). The by-product UDP GlcA and UDP (Figure S6, Supporting Information).
enters the SuSy module for in situ regeneration of Remarkably, inhibition data of human UGDH by the
UDP-GlcA. We performed comprehensive characteri- reduced by-product NADH emphasize the need for
+
[23]
zation of all employed enzymatic reactions to accom- an efficient cofactor regeneration of NAD .
In
plish the synthesis of the non-sulfated HNK-1. The summary, these properties reveal His UGDH as a fa-
6
presented EMS provides a flexible platform for the vorable enzyme for the combination with other
[
18]
combination with other GlcATs involved in the glu- enzyme modules.
[
19]
curonidation of pharmaceutical compounds.
Synthesis of UDP-GlcA 2
Results and Discussion
His UGDH/NOX Enzyme Module System (B)
6
Employing four enzymes in one-pot is an attractive
combinatorial strategy including in situ regeneration In order to provide UDP-GlcA for the characteriza-
+
of UDP-GlcA and the cofactor NAD (Scheme 1). In tion of GlcAT a batch synthesis of 2 was performed
order to optimize the multi-step reaction cascade the (EMS B in Scheme 1). For an economic synthesis we
enzyme properties such as pH optimum, substrate ki- included the in situ regeneration of the cofactor
+
netics and inhibition by intermediate products were NAD . First regeneration experiments with the com-
studied.
mercially available lactate dehydrogenase/pyruvate
system resulted only in 30% conversion of 1 (data not
shown). In contrast, NADH oxidase (NOX) from
Lactobacillus brevis was revealed as a powerful tool
Expression, Purification and Characterization of
+
His UGDH
for efficient NAD cofactor regeneration producing
6
[
17a,b]
water as by-product.
The EMS His UGDH/NOX
6
Optimized expression of the human His UGDH re- (B in Scheme 1) yielded 92% conversion of 5 mM
6
sulted in high level expression in E. coli Rosetta 1 (Figure 1) as analyzed by CE-UV (Figure S7, Sup-
(
DE3)pLysS strain. After induction with 100 mM porting Information). Product identity of isolated
IPTG (final concentration) and a post induction time UDP-GlcA 2 (49% overall yield, 13 mg, 22.4 mmol)
of 18 h at 258C, E. coli Rosetta (DE3)pLysS cells was confirmed by HPLC/ESI-MS (Figure S8, Support-
1
(
21 g) were harvested from 1 L of TB media.. Purifi- ing Information) and H NMR spectroscopy (Fig-
2+
cation of His UGDH from 10 g E. coli cells with Ni - ure S9, Supporting Information).
6
NTA chromatography yielded 51 U of recombinant
enzyme (~52 mg purified enzyme) with a specific ac-
tivity of up to 980 mU/mg protein (Table S1, Support-
ing Information). In comparison to published data
our expression strategy resulted in 7–8-fold higher
[20]
specific activity[ and a 3-times higher amount of pu-
21]
rified enzyme.
Western blot analysis confirmed
His UGDH with an apparent molecular mass of
6
5
8 kDa (Figure S1, Supporting Information). Enzyme
activity assays demonstrate the necessity of reducing
conditions underlining the critical role of the sulfhydr-
yl group of cysteine 276 of human UGDH in the cata-
[21,22]
lytic mechanism.
Further studies showed highest
activity at 308C and ongoing inactivation of Hi-
s6UGDH at temperatures above 308C (Figure S2,
Supporting Information). His UGDH is very sensitive
6
to pH shifts. Highest activity of His UGDH appears
6
at pH 8.7 in Tris-HCl buffer (Figure S3, Supporting
Information). Enzyme stability studies at 308C in
Figure 1. Synthesis of 2 with UGDH/NOX EMS (B in
5
4
0 mM Tris-HCl (pH 8.7) resulted in a half-life of
3.3 h with an inactivation coefficient k of 0.016 h
Scheme 1). The synthesis (10 mL) of 2 was performed with
À1
+
5
0 mM Tris-HCl (pH 8.7), 5 mM 1, 0.5 mM NAD , 10 U
(
Figure S4, Supporting Information). The K values
m
+
His UGDH and 15 U His NOX for 5 h at 308C. Mean
6
6
for UDP-Glc (0.03 mM) and NAD (0.44 mM) (Fig- values with standard deviations for triplicate experiments.
ure S5, Supporting Information) are in accordance Compound 2 was isolated with 49% overall yield.
Adv. Synth. Catal. 2015, 357, 1751 – 1762
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1753

Leonie Engels et al.
FULL PAPERS
[
10a]
native enzyme.
The specific activity of the murine
His catGlcAT-P with the acceptor substrate 8,
6
LacNAc type 2 (Scheme 2) is 60-fold and 47-fold
higher when compared with ProtA-GlcAT-P from rat
[
3]
expressed in COS-1 cells or human GlcAT-P ex-
[
11]
pressed in Pichia pastoris, respectively. SDS-PAGE
analysis confirmed the theoretical molecular mass of
3
6.8 kDa of purified His catGlcAT-P (Figure S12,
6
Supporting Information). His catGlcAT-P depicts
6
a strong requirement of divalent cations (Figure 3).
Highest activity was measured in the presence of
2
+
2+
2
mM Mn followed by Co ions with 83%. The ad-
2+
dition of Mg ions had only a moderate effect (30%)
whereas the addition of Ca ions gave no activity.
With the strong requirement for divalent cations
2
+
Figure 2. Synthesis of 2 with the SuSy/UGDH/NOX EMS
A+B in Scheme 1) monitored by CE-UV analysis. The syn-
(
thesis (1 mL) of 2 was performed in 50 mM Tris-HCl and the typical DxD motif for UDP-sugar-dependent
+
[25]
(
0
3
pH 7.8) with 0.5 mM NAD , 1 mM UDP, 300 mM sucrose, glycosyltransferases His catGlcAT-P belongs to the
6
.5 U SuSy1, 1 U His UGDH and 1.5 U His NOX for 5 h at GT-A fold family displaying a Rossman-like fold with
[26]
6
6
08C. 100% yield of 2 corresponds to 1 mM conversion of
twelve b-strands and seven a-helices
(Scheme S3,
UDP. Mean values with standard deviations for triplicate ex-
periments.
Supporting Information). These results are in accord-
[10a]
ance with the GT-A type human GlcAT-P.
Further
investigations elucidated a pH optimum at pH 6.5
(Figure S13, Supporting Information) and the highest
activity at 458C in 50 mM HEPES-NaOH buffer (Fig-
SuSy/UGDH/NOX Enzyme Module System (A+B)
With respect to our previous work on combinatorial ure S14, Supporting Information). In enzyme stability
strategies for the enzymatic synthesis of nucleotide experiments at 308C in HEPES-NaOH buffer
[
24]
sugars from sucrose
the SuSy module was com- (pH 6.5) a half-life of 61.3 h and an inactivation coef-
À1
bined with the UGDH/NOX EMS (A+B in ficient k of 0.011 h was determined (Figure S15,
Scheme 1). Although the involved enzymes exhibit Supporting Information).
different pH profiles, adjustment of enzyme ratios led
Recombinant murine His catGlcAT-P is highly spe-
6
to complete conversion of UDP at pH 7.8 and 308C cific for LacNAc type 2 (Table 1) which is in accord-
+
[3,27]
(
Figure 2). NAD and the formation of 1 and 2 were ance with human and rat GlcAT-P.
Comparing the
monitored by CE-UV analysis (Figure S10, Support- disaccharide LacNAc type 2 acceptor substrates, the
ing Information). The product UDP-GlcA 2 was fur- type of linker affects the enzyme activity in different
ther confirmed by ESI-MS (Figure S11, Supporting ways. The more hydrophobic t-Boc-protected amino-
Information). In summary, our results demonstrate linker renders LacNAc more favorable compared to
the functionality of the enzyme module system for the azide linker (3 and 9). The favorable influence of
the synthesis of 2 from 1 or from sucrose and UDP. a hydrophobic linker is also seen for lactose and
The latter is an important step towards the in situ re- octyl-lactose (6 and 7). The tetrasaccharide with two
generation of 2. In combination with GlcATs the plat- LacNAc units (10) proves to be a good acceptor sub-
form has the potential for the economic synthesis of strate. Interestingly, lactulose and raffinose (12 and
glucuronides from sucrose and UDP.
13) are accepted as novel substrates of GlcAT-P. The
latter display a-linked galactose, however, with low
enzyme activity pointing out the preference for
b-linked galactose.
Expression, Purification and Characterization of
His catGlcAT-P
Affinity constants for the donor substrate UDP-
6
GlcA (K 0.19 mM) as well as for the acceptor sub-
m
Expression of the catalytic domain of murine strates 8 (K 0.27 mM) and 3 (K 0.24 mM) were de-
m
m
His catGlcAT-P in E. coli BL21(DE3) resulted in termined according to Michaelis–Menten kinetics
6
high level expression yielding 19 g biomass/L of TB (Figure S16, Supporting Information). The slightly
media and 7 U of recombinant enzyme (17 mg puri- better affinity for the linker modified LacNAc reflects
fied enzyme from 10 g cells) with a specific activity of the preference of GlcAT-P for hydrophobic acceptor
2+
0
.44 U/mg after purification by IMAC with Ni -NTA substrates. Most importantly, the addition of alkaline
(
Table S2, Supporting Information). In line with the phosphatase increased the His catGlcAT-P activity
6
human recombinant GlcAT-P our results demonstrate more than two-fold (Figure S17, Supporting Informa-
that the heterologous expression of the murine glucur- tion) indicating a severe inhibition by UDP which is
onyltransferase in E. coli exhibits the activity of the formed as by-product of the transferase reaction. Fur-
1754
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1751 – 1762

FULL PAPERS
Module Systems for the Synthesis of Uridine 5’-Diphospho-a-d-glucuronic Acid
Scheme 2. Tested acceptor substrates for His catGlcAT-P.
6
ports recombinant His catGlcAT-P as a potential en-
6
zymatic tool for the synthesis of the HNK-1 epitope
in combinatorial biocatalysis.
Synthesis of Non-Sulfated HNK-1-linker-t-Boc 4
Based on our substrate screening results, linker modi-
fied LacNAc type (compound 3 in Table 1) was
chosen as acceptor for the synthesis of non-sulfated
HNK-1. We combined stepwise the EMS (Scheme 1)
to reach the best substrate conversion employing min-
imum amounts of enzymes and in situ regeneration of
UDP-GlcA.
Figure 3. Dependency of His catGlcAT-P on divalent metal
ions. The impact of 2 mM divalent metal ions was tested at
6
3
1
08C in 50 mM HEPES-NaOH pH 6.5. Relative activity of
00% corresponds to 0.37 U/mg. Mean values with standard
deviations for triplicate experiments.
UGDH/NOX/GcAT-P Enzyme Module System
(
B+C)
ther studies showed no inhibition by UDP-Glc and In a first approach EMS B and C (Scheme 1) were
sucrose, the product and substrate in the SuSy combined. Focusing on an economic synthesis we car-
enzyme module (Figure S18, Supporting Information). ried out two experiments with an acceptor substrate
In conclusion, the biochemical characterization sup- concentration of 5 mM and 0.5 U/mL and 1.0 U/mL,
Adv. Synth. Catal. 2015, 357, 1751 – 1762
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1755

Leonie Engels et al.
FULL PAPERS
Table 1. Saccharides were tested as acceptors of recombinant His catGlcAT-P (refer to Scheme 2 for glycan structures).
6
2
mM acceptor substrates were tested with 2 mM donor substrate UDP-GlcA in 50 mM HEPES-NaOH buffer pH 6.5. Data
depict mean values and standard deviations for triplicate measurements of each compound. 100% relative activity of
His catGlcAT-P corresponds to 0.38 U/mg.
6
Compound
Acceptor Substrate
Rel. Activity [%]
[
c]
5
6
7
8
d-Gal
Gal(b1-4)Glc
ND
18Æ1.3
33Æ2.4
100Æ6.3
20Æ1.3
116Æ4.5
55Æ3.1
Gal(b1-4)Glcb1-octyl
Gal(b1-4)GlcNAc
Gal(b1-4)GlcNAcAzII
Gal(b1-4)GlcNAc(b1-linker-t-Boc
Gal(b1-4)GlcNAc(b1-3)Gal(b1-4)GlcNAc(b1-linker-t-Boc
Gal(b1-3)GlcNAc
[
[
a]
b]
9
3
1
1
1
1
[
b]
0
1
2
3
[c]
ND
Gal(b1-4)Fru
Gal(a1-6)Glc(a1-2)Fru
35Æ2,9
6.7Æ0.4
[
[
[
a]
[28]
See references Adamiak et al. and Anders et al. for synthesis of this compound.
See reference Rech et al. for synthesis of this compound.
b]
[15a]
c]
ND: no activity detected.
respectively, of all required enzymes. Regarding the
inhibition of the His catGlcAT-P by UDP alkaline
6
phosphatase was added. Complete conversion of
5
2
mM acceptor substrate 3 was accomplished after
4 h and 5 h, respectively (Figure S19 and Figure S20,
+
Supporting Information) with regeneration of NAD .
In contrast, conversion without alkaline phosphatase
yielded only 66% for compound 4 (Figure S21, Sup-
porting Information), which is due to the accumula-
tion of inhibitory UDP confirmed by CE-UV (Fig-
ure S22, Supporting Information).
SuSy/UGDH/NOX/GcAT-P Enzyme Module System
Figure 4. Synthesis of 4 on analytical scale by combination
of SuSy/UGDH/NOX/GlcAT-P enzyme modules with mini-
mum amounts of enzymes (A+B+C in Scheme 1). The
synthesis (1 mL) was performed in buffer A (50 mM Tris-
HCl, pH 7.8, 5 mM DTT) with 0.5 mM NAD , 0.2 mM
UDP, 5 mM of compound 3, 300 mM sucrose, 0.25 U SuSy 1,
(
A+B+C)
The strong inhibition of GlcAT-P by UDP requires
a novel synthesis strategy. Therefore, the enzyme
modules SuSy/UGDH/NOX/GlcAT-P (A+B+C in
Scheme 1) were combined in one-pot to integrate in
situ regeneration of UDP-GlcA from UDP and su-
+
0
.5 U His UGDH, 0.5 U His NOX and 0.5 U His catGlcAT-
6
6
6
P for 24 h at 308C. A conversion of 98% after 24 h corre-
crose. Synthesis of 4 on analytical scale yielded 98% sponds to 4.89 mM product. Mean values with standard de-
conversion of the acceptor substrate 3 (Figure 4). viations for triplicate experiments.
Thus, the in situ regeneration of 2 is accomplished by
starting with catalytic amounts of UDP (0.2 mM) and
+
NAD (0.5 mM) reaching cycle numbers of 24 and 18,
respectively. Surprisingly the scale-up (5 mL batch) of enzyme preparations caused the cleavage of com-
the one-pot synthesis (Figure 5) only yielded 79% for pound 3. Control experiments revealed b-galactosi-
+
4
with similar cycle numbers for UDP and NAD (19 dase side activity in the purified protein pool of re-
and 14 cycles, respectively). The product was con- combinant His UGDH (Figure S25, Supporting Infor-
6
firmed by ESI-MS (Figure S23, Supporting Informa- mation) with a volumetric activity of only 20 mU/mL
1
tion) and H NMR spectrum (Figure S24, Supporting (Figure S26, Supporting Information). Optimizing the
Information).
protein purification method gave no essential im-
HPLC analysis showed the formation of provement. From our results we may postulate aggre-
GlcNAcb1-linker-NH -t-Boc, probably by enzymatic gation of intrinsic b-galactosidase of the expression
2
cleavage of 3 (Figure S25, Supporting Information). strain E. coli Rosetta(DE3)pLysS with the human
We assumed that b-galactosidase in one of our hexamer protein of His UGDH.
6
1756
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1751 – 1762

FULL PAPERS
Module Systems for the Synthesis of Uridine 5’-Diphospho-a-d-glucuronic Acid
Figure 7. Batch synthesis of 4 by combination of SuSy/
UGDH/NOX/GlcAT-P enzyme modules with optimized
enzyme amounts (A+B+C in Scheme 1). The synthesis
(5 mL) was performed in buffer A (50 mM Tris-HCl, pH 7.8,
5 mM DTT) with 0.5 mM NAD , 0.2 mM UDP, 5 mM of
compound 3, 300 mM sucrose, 1.25 U SuSy 1, 2.5 U
Figure 5. Batch synthesis (5 mL) of 4 by combining SuSy/
UGDH/NOX/GlcAT-P with minimum amounts of enzymes
+
(
A+B+C in Scheme 1). The synthesis was performed in
buffer A (50 mM Tris-HCl, pH 7.8, 5 mM DTT) with
+
0
.5 mM NAD , 0.2 mM UDP, 5 mM of compound 3,
His UGDH, 5.0 U His NOX and 7.5 U His catGlcAT-P for
6
6
6
3
00 mM sucrose and 1.25 U SuSy1, 2.5 U His UGDH, 2.5 U
6 h at 308C. A conversion of 98.6% of 3 after 6 h corre-
sponds to 4.93 mM of product 4. Mean values with standard
deviations for triplicate experiments. Compound 4 was iso-
lated with 48% overall yield.
6
His NOX and 2.5 U His catGlcAT-P for 24 h at 308C. A
6
6
conversion of 79% 3 after 24 h corresponds to 3.9 mM of 4.
Formation of 5 as cleavage product of 3 is indicated. Mean
values with standard deviations for triplicate experiments.
[12]
We concluded that increasing the enzyme amount HNK-1 with metabolic engineered E. coli cells,.
of His UGDH by 5-fold in a batch synthesis would The optimization of synthesis strategies on analytical
6
result in cleavage of the acceptor substrate 3. Howev- and batch scales highlights the complexness and ra-
er, increasing NOX and GlcAT-P activity in excess to tionale of enzymatic cascade reactions in combinatori-
SuSy1 and UGDH (0.25 U SuSy/0.5 U UGDH/1.0 U al biocatalysis.
NOX/1.5 U GlcAT-P) resulted in a complete conver-
sion of compound 3 (100%, 5 mM) within 4 h on an synthesis (5 mL) yielded 98.6% (4.93 mM) conversion
analytical scale (Figure 6). of compound 3 within 6 h without cleavage of the ac-
Transforming these reaction conditions to a batch
In conclusion, the combinatorial strategy presented ceptor substrate 3 (Figure 7). The integrated in situ
+
herein, is a conclusive alternative strategy to the al- regeneration of 2 and NAD is accomplished with
ready reported bacterial synthesis of non-sulfated nearly maximum cycle numbers of 24 and 18, respec-
tively.
Conclusions
Here we present a novel tailor made EMS for the
synthesis of UDP-GlcA and non-sulfated HNK-1 epi-
tope with in situ regeneration of UDP-GlcA and the
+
cofactor NAD .
A detailed characterization of
His UGDH and His catGlcAT-P allowed the step by
6
6
step optimization of the enzyme cascades. The one-
pot synthesis avoids tedious isolation and purification
of the intermediate products leading to high overall
product yields. Thus, this strategy rationalizes synthe-
sis costs with the expensive nucleotide UDP and nu-
cleotide sugar UDP-GlcA. Furthermore, the demon-
strated EMS has the potential to be used for the syn-
thesis of non-sulfated HNK-1 epitope modified by
various linkers as well as glucuronides by exchange of
the acceptor substrates and glucuronyltransferases, re-
Figure 6. Synthesis of 4 on analytical scale by combination
of SuSy/UGDH/NOX/GlcAT-P enzyme modules with opti-
mized enzyme amounts (A+B+C in Scheme 1). The syn-
thesis (1 mL) was performed in buffer A (50 mM Tris-HCl,
+
pH 7.8, 5 mM DTT) with 0.5 mM NAD , 0.2 mM UDP,
5
mM of compound 3, 300 mM sucrose, 0.25 U SuSy 1, 0.5 U
His UGDH, 1.0 U His NOX and 1.5 U His catGlcAT-P for
6
6
6
4
h at 308C. A conversion of 100% 3 after 4 h corresponds
to 5.0 mM of product 4. Mean values with standard devia- spectively.
tions for triplicate experiments.
Adv. Synth. Catal. 2015, 357, 1751 – 1762
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1757

Leonie Engels et al.
FULL PAPERS
Experimental Section
primers
Fw-catglcatp-5’ GACGACGACAAGATGCTCG
CACCTCTGCTTG 3’ and Rev-catglcatp-5’ GAGGAGAAGCCCGGT
CAGATTCACTGAGGGG 3’ and later ligated into the expression
vector pET46Ek/LIC according to the instructions of Nova-
gen (Novagen/Merck, Darmstadt, Germany). Ligation re-
sulted in pET46Ek/LIC_his catglcAT-P (Scheme S2B, Sup-
porting
General Methods
Amplification of all genes was performed with KOD Hot
Start DNA Polymerase-Kit from Novagen (Novagen/Merck,
Darmstadt, Germany). PCR purification and agarose gel ex-
traction were carried out with “Nucleo Spin Extract II”
from Macherey–Nagel (Düren, Germany). Plasmids were
purified with “peqGold Plasmid Miniprep KitI” from
PEQLAB Biotechnologie GmbH (Erlangen, Germany).
The standard 1 kb DNA ladder, restriction enzymes, T4-
6
Information).
Heterologous
expression
of
His catglcAT-P was performed in E. coli BL21(DE3) (Nova-
6
gen/Merck, Darmstadt, Germany).
Expression, Purification of His NOX, His catGlcAT-P
and His UGDH
6
6
TM
6
ligase and alkaline phosphatase “Fast AP ” were delivered
from MBI Fermentas (Vilnius, Lithuania). For transforma-
Expression of pET28a_his nox was carried out in E. coli
6
À
+
tion the strain E. coli NovaBlue endA1 hsdR17(r
m
)
BL21(DE3), pETDuet-1_his ugdh in E. coli Rosetta(DE3)-
K12
K12
6
+
+
q
supE44 thi-1 recA1 gyrA96 relA1 lac Fꢁ[proA B lacI ZDM
pLysS and pET46Ek/LIC_his catglcAT-P in E. coli
6
R
1
5:Tn10 (Tc )] (Novagen/Merck, Darmstadt, Germany) was
BL21(DE3) in overnight cultures. Recombinant cells were
grown in 100-mL flasks containing 25 mL LB medium with
appropriate amounts of antibiotic at 378C and 120 rpm. 5-L
Erlenmeyer flasks containing 1 L TB medium were inoculat-
ed with 25 mL pre-cultures. Main cultures were shaken at
+
used. UDP, UDP-GlcA, UDP-Glc, cofactor NAD and
NADH were purchased from Sigma Aldrich (Deisenhofen,
Germany). If not otherwise indicated all other reagents
were purchased from Carl Roth (Karlsruhe, Germany),
Merck (Darmstadt, Germany), Roche (Basel, Switzerland)
and Sigma Aldrich GmbH (Deisenhofen, Germany). Integ-
rity of the gene constructs was confirmed by sequencing (Se-
quiserve, Vaterstetten, Germany).
378C and 80 rpm until induction of the promoter (OD600
=
0.6–0.8). LacZ promoter was induced with a final concentra-
tion of 0.1 mM IPTG. After an incubation time of 18–20 h
at 258C cells were harvested by centrifugation (30 min at
4
8C, 7000 rpm). For protein purification all three histidine-
tagged enzymes were purified by IMAC. A cell suspension
of 40% (w/w) was lysed by ultrasound (4”15 sec). Cellular
debris was removed after centrifugation (15000 rpm, 30 min)
and later analyzed for the production of inclusion bodies by
SDS-PAGE or Western blot. The crude extract (15 mL) was
Cloning and Expression of Recombinant Enzymes
Ugdh from Homo sapiens
Human cDNA of mesangial cells was kindly provided by
Dr. Claudia van Roeyen, (Medical Clinic II, University Hos-
pital of RWTH Aachen, Germany) for amplification of the
gene ugdh (UniProtKB/Swiss-Prot/accession number:
O60701) with PCR using the primers: Fw-ugdh 5’ ATGTTT
GAAATTAAGAAGATCTGTTGCATCGG 3’ and Rev-ugdh-5’ CTA
CACTTTAGGTTTCTTGTTAGGTGGATC 3’. In a second step the
introduction of the restriction sites was carried out with oli-
gonucleotides Fw-PstI-ugdh-5’ AAAACTGCAGCCAATGCATTG
GAAAATGTTTGAAATTAAGAAGATCTGT 3’ and Rev-NotI-
ugdh-5’ ATAAGAATGCGGCCGCTAAACTATCTACACTT
À1
loaded onto an IMAC column (1.6”20 cm, 1.5 mLmin ) of
2+
Ni -NTA Sepharose (Qiagen, Hilden, Germany), which
was equilibrated with lysis buffer. Removal of unspecific
proteins was carried out using washing buffers with 0.02M
imidazole. The final elution step with buffers containing
3
00 mM imidazole allowed the detection of the target pro-
tein, respectively. All fractions containing the recombinant
enzyme were analyzed for protein concentration with Brad-
ford assay and for enzyme activity. For purification of
TAGGTTTCTTGT 3’. After ligation of the amplified product
into pETDuet-1 sequencing results confirmed the construct
His NOX the recombinant protein was activated with 10 mm
6
FAD and purified with 50 mM KP buffer (100 mMK PO
i
2
4
pETDuet-1 his ugdh (Scheme S2A, Supporting Informa-
6
and 100 mM KH PO adjusted to pH 6.5). Lysis buffer addi-
2 4
tionally contained 1 mM DTT, washing buffer 300 mM
NaCl, 20 mM imidazole and elution buffer 300 mM NaCl,
tion). For heterologous expression E. coli Rosetta(DE3)-
pLysS (Novagen/Merck, Darmstadt, Germany) was chosen.
3
00 mM imidazole. For purification of recombinant
His UGDH and His catGlcAT-P lysis buffer: 50 mM
NaH PO pH 8; 1 mM DTT; washing buffer: 50 mM
Glcatp from Mus musculus
6
6
2
4
The gene glcatp (UniProtKB/Swiss-Prot/accession number:
Q8R531) was cloned starting from mRNA of mouse brain
NaH PO with 0.3M NaCl, 1 mM DTT, 0.02M imidazole,
2
4
pH 8.0; elution buffer: 50 mM NaH PO with 0.3M NaCl,
2
4
(
kindly provided by Prof. W. Jahnen-Dechent, Institute of
1
mM DTT, 0.3M imidazole, pH 8.0, 20% (v/v) glycerol was
Cell and Molecular Biology at Interfaces, Medical Faculty,
RWTH Aachen University). The first strand cDNA synthe-
sis performed with the anchored oligo (dT)18 primer follow-
ing the manual of the Transcriptor High Fidelity cDNA Syn-
thesis Sample Kit from Roche (Basel, Switzerland) was fur-
ther used for amplification of the entire length of glcatp
with the following primers Fw-glcAT-P-5’ ATGGGTAATGAG
used.
Expression of pET28a_sus1 in E. coli BL21(DE3)
The expression of pET28a_sus1 in E. coli BL21(DE3) was
performed as described in Sauerzapfe et al.
was carried out according to the method of Rçmer et al.
[16]
Purification
[29]
GAGCTGTG 3’
and
Rev-glcAT-P-5’ TCAGATCTCCACT
GAGGGGT 3’. Eukaryotic glycosyltransferases belong to type
II membrane proteins, anchored in the Golgi apparatus
membrane. Therefore, only the C-terminal catalytic domain
of GlcAT-P (amino acids 27–334) was amplified with
SDS-PAGE and Western Blot Analysis
The expression level of all recombinant enzymes was moni-
tored by SDS-PAGE or Western blot analysis. Acrylamide
1
758
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1751 – 1762

FULL PAPERS
Module Systems for the Synthesis of Uridine 5’-Diphospho-a-d-glucuronic Acid
gels (10%) were casted according to the manual of Invitro-
gen (Invitrogen, Paisley, UK). For each single enzyme crude
ware GmbH, Erkrath, Germany). Product inhibition and
cross inhibition by UDP-GlcA and UDP were tested using
the photometric assay by adding concentrations of 0.02 mM
to 5 mM to the assay solution. Enzyme stability, pH and
temperature optimum were also investigated with the con-
tinuous assay described above.
extract, cell debris and purified His -tagged protein samples
6
were investigated after staining with Coomassie Blue. Deter-
mination of the molecular mass was carried out with a pre-
TM
stained Protein Ladder PageRuler
(Fermentas, Vilnius,
Lithuania). The His -tag of recombinant His UGDH was
6
6
recognized by immunoblot using a monoclonal anti-His -an-
6
tibody conjugated to horse radish peroxidase (HRP) (Roche Activity Assays and Biochemical Characterization of
Diagnostics, Mannheim, Germany). Therefore, the instruc- His catGlcAT-P
6
tions of the NuPAGE Western transfer protocol (BioRad,
Two different assays were performed to determine the
München, Germany) were followed.
enzyme activity of His catGlcAT-P in the crude extract and
6
IMAC fractions. A continuous photometric assay in a micro-
titer plate, was performed using the pyruvate kinase/lactate
dehydrogenase system for the detection of released UDP
Activity Assays and Biochemical Characterization of
His UGDH
6
(
method C). The assay solution consisted of 2 mM UDP-
Enzyme activity of His UGDH in the crude extract and
GlcA, 2 mM LacNAc type 2, Gal(b1-4)GlcNAc
8
6
IMAC fractions were determined by a continuous photo-
metric assay (method A) and capillary electrophoresis with
UV-detection (CE-UV analysis, method B). The reaction
mixture for the microtiter plate assay contained 2 mM
UDP-Glc, 3 mM NAD ; 5 mM DTT, 50 mM Tris-HCl,
pH 8.7 in a total volume of 200 mL (method A) or 400 mL
(Scheme 2), 1 mM phosphoenolpyruvate, 20 mM KCl, 2 mM
K HPO , 0.25 mM NADH, 2 mM MnCl , 4 mM MgCl ,
2
4
2
2
5 mM DTT, 5 U pyruvate kinase and 5 U lactate dehydro-
genase in 50 mM HEPES-NaOH, pH 6.5 in a total volume
of 250 mL. The reaction was started by addition of 100 mL
enzyme solution and followed at 340 nm at 378C. Control
experiments for the detection of side activities were done
without the acceptor substrate. Enzyme activity for the sub-
strates Gal(b1-4)GlcNAc(b1-linker-t-Boc 3 and 8 was also
determined by CE-UV analysis for detection of released
UDP at 254 nm (method D). The assay solution (500 mL)
consisting of 2 mM UDP-GlcA, 2 mM LacNAc-linker-t-Boc
or 2 mM type 2 LacNAc, 2 mM MnCl2, 5 mM DTT and
200 mL purified enzyme was incubated at 378C for 15 min.
The reaction was stopped by heating at different time points
as described above. In both analyzing methods 1 U of
+
(
method B). The reaction was started by addition of 50 mL
enzyme solution at 308C. The formation of NADH was de-
tected at 340 nm (method A). With method B formation of
UDP-GlcA was followed over time by CE-UV at 254 nm.
Samples were taken at different time points where the con-
version rate was linear and heated for 5 min at 958C. CE-
UV analysis was performed on a P/ACE MDQ apparatus
from Beckman Coulter (Krefeld, Germany), equipped with
a UV detector. The separation of nucleotide sugars and co-
+
factors NAD and NADH was accomplished in a fused-
silica capillary (i.d. 75 mm, 57 cm total capillary length,
His catGlcAT-P is determined as the amount of enzyme
6
5
0 cm to the detector) with 50 mM Na B O ·10H O buffer,
which produces 1 mmol UDP per min under standard assay
conditions.
Further experiments for biochemical characterization
were performed with method C under standard assay condi-
tions. For investigation of the temperature optimum reaction
temperatures were varied from 27–458C. To study the opti-
2
4
7
2
pH 8.9. Conditions for migration and detection were 25 kV
23 mA) at 258C and UV detection at 254 nm, respectively.
(
Samples were injected by pressure (5.0 sec at 0.5 psi in the
forward direction). UDP-Glc and UDP-GlcA were quanti-
fied by a standard calibration curve with commercially avail-
able nucleotide sugars. For both assays 1 U of His UGDH is
mal pH value for the activity of recombinant His catGlcAT-
6
6
the amount of enzyme which produces 1 mmol product
P samples were measured in different buffer systems includ-
ing HEPES-NaOH buffer (pH 5.9–7) and 50 mM Tris-HCl
buffer (pH 7–9) at 378C. Enzyme stability was monitored by
incubating the purified protein solution in HEPES-NaOH
buffer (pH 6.5) at 308C over a period of 7 days with subse-
quent determination of the residual enzyme activity. Addi-
tion of metal ions 2 mM NiCl2, MnCl , CaCl , MgSO and
(
UDP-GlcA) per min under standard assay conditions. Tem-
perature optimum, pH optimum and enzyme stability were
investigated with the continuous photometric assay (method
A). Studying the optimal pH value for highest enzyme activ-
ity of recombinant His UGDH the assay was performed at
6
different pH values in 50 mM Tris-HCl buffer ranging from
2
2
4
7
.2 to 9.0 at 308C. The influence of temperature on the
CoCl2 to the standard assay allowed investigation of the
metal ion dependency. The impact of alkaline phosphatase
enzyme activity was determined under standard assay condi-
tions at different temperatures in a range of 258C to 378C.
Enzyme stability was tested over a period of 7 days at 308C
on the transferase activity of His catGlcAT-P was deter-
6
mined by method D with 5 mM UDP-GlcA, 5 mM LacNAc-
in 50 mM Tris-HCl (pH 8.7).
linker-t-Boc 3 or type 2 LacNAc 8, respectively, and 5 mM
+
Kinetic data of His UGDH for UDP-Glc and NAD were
MnCl 5 mM DTT and 200 mL purified enzyme.
6
2,
analyzed with method A. Kinetic constants were derived
from initial rate analysis at variable substrate concentra-
tions. UDP-Glc was varied from 0.02 mM to 5 mM at a con-
stant concentration of 6 mM NAD and NAD was altered
from 0.02 to 5 mM at a constant concentration of 6 mM
UDP-Glc. Kinetic constants were determined by non-linear
regression analysis according to the Michaelis–Menten equa-
tion using the Sigma Plot 10 software (SPSS Science Soft-
Kinetic analysis was done for the acceptor substrates 8
and 3 in a range of 0.03 to 5 mM and 0.03 to 2.5 mM, re-
spectively. For the donor substrate UDP-GlcA the concen-
tration range was 0.03 to 5 mM. Determination of kinetic
constants was accomplished by non-linear-regression analy-
sis according to the Michaelis–Menten equation using the
Sigma Plot 10 software (SPSS Science Software GmbH,
Erkrath, Germany). Cross inhibition studies were carried
+
+
Adv. Synth. Catal. 2015, 357, 1751 – 1762
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1759

Leonie Engels et al.
FULL PAPERS
out by adding UDP-Glc (0.02 mM to 5 mM) and sucrose
300 mM) to the reaction mixture using the photometric
assay (method C).
final volume in a 15-mL centrifuge tube under gentle stir-
ring at 20 rpm). Samples of 50 mL were taken at different
time points, enzyme reaction was stopped at 958C for 5 min,
centrifuged and analyzed by CE-UV. Enzymes were re-
moved from the reaction mixture by heating for 5 min at
958C and subsequent centrifugation. The crude product so-
lution was purified by preparative hydrophilic interaction
chromatography (HILIC) HPLC (Knauer, Germany, Soft-
ware: EZChrom Elite) on a Multospher APS-HP-5m 250”
3 mm column (CS Chromatographie, Langerwehe, Germa-
ny) using a gradient ranging from 100% solvent A (MilliQ
H O) to 100% solvent B (0.4 mM NH CH COO pH 6.7) at
(
Activities of His catGlcAT-P with 2 mM acceptor sub-
6
strates (Scheme 2) were investigated under standard assay
conditions of method C. The relative activity was deter-
mined by comparison with Gal(b1-4)GlcNAc. Compounds
[15a,28]
3, 9, and 10 were synthesized as described previously.
The relative activity of compound 5 was determined using
the assay for CE-UV analysis as described above.
Activity Assay of SuSy1
2
4
3
À1
a flow rate of 5 mLmin . Fractions were monitored for the
target product with CE-UV analysis and finally concentrated
by speedvac. Product 2 was obtained with a final yield of
49% (13 mg, 22.4 mmol) and confirmed by ESI-MS and
Enzyme activity was measured using a continuous photo-
[16]
metric assay in a microtiter plate as described previously.
Furthermore, enzyme activity was determined at linear con-
version rate with CE-UV analysis of UDP-Glc at 254 nm as
described above. The assay solution of 500 mL contained
1
H NMR spectrometry.
250 mM sucrose, 50 mM HEPES-NaOH pH 7.6 and 1 mM
UDP. After adding 250 mL enzyme solution the reaction was
started at 308C. Samples at different time points were
stopped by heating for 5 min at 958C and centrifuged for
Synthesis and Purification of Compound 3, LacNAc-
linker-t-Boc
1
0 min at 15,000 rpm. The concentration of the reaction
Expression, protein purification and activity assays of
human b1,4-galactosyltransferase-1 and UDP-Glc-4’-epimer-
ase from Campylobacter jejuni were performed as previously
product UDP-Glc was quantified by a standard calibration
curve of UDP-Glc. For both assays 1 U of SuSy1 is the
amount of enzyme which produces 1 mmol product per min
under standard assay conditions.
[15a]
described by Rech et al.
was carried out as described before
0 mL in a 100-mL flask containing 5 mM GlcNAc-linker-t-
The preparative synthesis of 3
[15a]
in a total volume of
2
Boc and 10 mM UDP-glucose. The isolation of product 3
was performed by preparative reversed-phase (RP) HPLC
(Knauer, Germany, Software: EZChrom Elite) on a Euros-
Activity Assay of His NOX
6
The activity of His NOX was accomplished with CE-UV
6
[15a]
analysis at 308C in 50 mM KP buffer. 10 mL of His NOX
were added to 290 mL 50 mM KP buffer (pH 6.5) including
i
6
pher 100–10 C 18 column with 15% (v/v) acetonitrile.
i
1
mM NADH. Samples were taken at different time points
by heating at 958C for 5 min. After centrifugation (10 min,
HPLC Analysis for Detection of 2 UDP-GlcA
1
2
5000 rpm) the supernatant was analyzed by CE-UV at
+
54 nm for NAD formation as described above. The con-
Separation and detection of product 2 (synthesized with
SuSy/UGDH/NOX enzyme module system) was accom-
plished with hydrophilic interaction chromatography
(HILIC) using Thermo Quest Hypersil APS2–3m column
(100”3 mm). A 10-mL sample of the crude reaction solution
was separated with 75% acetonitrile and 8% 500 mM
+
centration of NAD was quantified by a standard calibra-
tion curve. The enzyme activity was determined under
linear conversion rate; 1 U of His NOX is the amount of
6
+
enzyme which produces 1 mmol NAD per min under stan-
dard assay conditions.
À1
NH CH COO pH 4.5 at a flow rate of 0.5 mLmin . The re-
4
3
tention time of 2 was compared with a sample of 2 mM
commercially available UDP-GlcA.
Synthesis of UDP-GlcA 2 with the SuSy/UGDH/
NOX Enzyme Module System
Synthesis of 2 on an analytical scale (1 mL) was performed
by combination of both enzyme modules in one pot. The re-
action solution consisting of 50 mM Tris-HCl buffer
Synthesis of Compound 4, GlcA(b1-3)Gal(b1-
4
)GlcNAc(b1-linker-t-Boc, by the UGDH/GlcAT-
Enzyme Module System
+
(
pH 7.8), 0.5 mM NAD , 5 mM DTT, 1 mM UDP, 300 mM
sucrose, 0.5 U SuSy1, 1 U His UGDH and 1.5 U His NOX
6
6
was incubated for 5 h at 308C. Samples at different time
points were stopped by heating for 5 min at 958C, centri-
fuged for 10 min at 15,000 rpm, and the supernatant ana-
lyzed by CE-UV analysis.
The synthesis of 4 was first performed on analytical scale
(1 mL) with 5 mM acceptor substrate 3 in buffer A (50 mM
Tris-HCl, pH 7.8, 5 mM DTT) with 6 mM MnCl
, 0.5 mM
NAD , 6 mM of compound 1, and 5 mM acceptor substrate
. The enzyme amount was set to 1 U or 0.5 U (for
His UGDH, His NOX, and His catGlcAT-P) with (2 U) and
2
+
3
6
6
6
Batch Synthesis of UDP-GlcA 2 with the UGDH/
NOX Enzyme Module System
without addition of alkaline phosphatase. The reaction solu-
tions were then incubated for 24 h at 308C. Aliquots of
20 mL (diluted 1:5) were taken during incubation for RP-
Batch synthesis of 2 was carried out in 50 mM Tris-HCl
buffer (pH 8.7) containing 5 mM UDP-Glc 1, 0.5 mM
ꢂ
HPLC analysis using LiChrospher 100RP 18, LiChro-Cart,
+
NAD , 5 mM DTT, 10 U His UGDH and 15 U His NOX.
column (Merck, Germany) with a mobile phase of 15% (v/
6
6
[15a]
Reaction solution was incubated for 5 h at 308C (10 mL
v) acetonitrile containing 0.1% formic acid.
1760
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1751 – 1762

FULL PAPERS
Module Systems for the Synthesis of Uridine 5’-Diphospho-a-d-glucuronic Acid
Synthesis of Compound 4, GlcA(b1-3)Gal(b1-
)GlcNAc(b1-linker-t-Boc, by the SuSy/UGDH/
GlcAT-Enzyme Module System
Acknowledgements
4
The authors acknowledge financial support by the Federal
Ministery for Education and Research (BMBF) within the
program “Basistechnologien” for the project “The Golgi
Glycan Factory” (FKZ031A162). The authors acknowledge
support by the DWI Leibniz-Institut für Interaktive Materia-
lien e.V. , RWTH Aachen University, Germany, for carrying
The synthesis of product 4 (1 mL/5 mL batch) was per-
+
formed in buffer A with 0.5 mM NAD , 0.2 mM UDP,
3
00 mM sucrose, 5 mM acceptor substrate 3, 5 mM MnCl2,
0
.25 U/mL SuSy1, 0.5 U/mL His UGDH, 0.5/1.0 U/mL
6
His NOX and 0.5/1.5 U/mL His catGlcAT-P for 24 h at
1
6
6
out the H NMR measurement and thank M.Sc. Ruben R.
308C. The 5 mL batch was performed in a 25-mL flask
Rosencrantz for performing NMR analysis. Furthermore the
authors thank M.Sc. Vera Werthmann for characterization
work on GlcAT-P and the DECHEMA Max Buchner Foun-
dation for financial support (L. Engels). We thank Prof. Dr.
Vladimír K rˇ en (Institute of Microbiology, Academy of Scien-
ces of the Czech Republic Center of Biocatalysis and Bio-
transformation) for providing the starting substrate GlcNAc-
linker-t-Boc.
under gentle (20 rpm) stirring. At different time points ali-
quots were analyzed with RP-HPLC. Experimental condi-
[15a]
tions for product isolation are described by Rech et al.
Product 4 was obtained with an overall yield of 48% (9 mg,
1
1
2 mmol) and characterized by ESI-MS and H NMR spec-
trometry.
Control Experiment of the Cleavage of the Acceptor
Substrate Gal(b1-4)GlcNAc(b1-linker-t-Boc with RP-
HPLC Analysis
References
The experiment was carried out by incubating 0.5 U
His UGDH, 0.5 U His NOX, 0.5 U His catGlcAT-P and
[1] R. Kleene, M. Schachner, Nat. Rev. Neurosci. 2004, 5,
195–208.
[2] I. Morita, Y. Kizuka, S. Kakuda, S. Oka, J. Biochem.
2008, 143, 719–724.
6
6
6
0
5
.5 U SuSy1 with 2 mM Gal(b1-4)GlcNAc(b1-linker-t-Boc in
0 mM Tris-HCl (pH 7.8) for 24 h at 308C, respectively. A
sample without the addition of the enzyme was used as neg-
ative control.
[3] S. Kakuda, Y. Sato, Y. Tonoyama, S. Oka, T. Kawasaki,
Glycobiology 2005, 15, 203–210.
[
4] a) S. N. Masand, I. J. Perron, M. Schachner, D. I.
Shreiber, Biomaterials 2012, 33, 790–797; b) H. Tagawa,
Y. Kizuka, T. Ikeda, S. Itoh, N. Kawasaki, H. Kurihara,
M. L. Onozato, A. Tojo, T. Sakai, T. Kawasaki, S. Oka,
J. Biol. Chem. 2005, 280, 23876–23883.
Determination of b-Galactosidase Side Activity in the
Purified Protein Pool of Recombinant His UGDH
6
For quantification of the b-galactosidase side activity in the
purified protein pool of recombinant His UGDH 10 mL
6
His UGDH were added to 90 mL of 4 mM pNPG in 50 mM
[5] Y. Kizuka, K. Kobayashi, S. Kakuda, Y. Nakajima, S.
Itoh, N. Kawasaki, S. Oka, Glycobiology 2008, 18, 331–
338.
6
citrate-Na HPO buffer (pH 6.0) for 24 h at 308C. Samples
2
4
were taken at different time points and reaction was stopped
by the addition of 200 mL 0.2M Na CO . The released pNP
[6] a) F. Belot, A. Otter, M. Fukuda, O. Hindsgaul, Synlett
2003, 1315–1318; b) Y. E. Tsvetkov, M. Burg-Roder-
feld, G. Loers, A. Arda, E. V. Sukhova, E. A. Khatunt-
seva, A. A. Grachev, A. O. Chizhov, H.-C. Siebert, M.
Schachner, J. Jimenez-Barbero, N. E. Nifantievt, J. Am.
Chem. Soc. 2012, 134, 426–435.
2
3
was determined with a spectrophotometer at 405 nm using
a calibration of pNP. 1 U of b-galactosidase is the amount of
enzyme which produces 1 mmol pNP per min under standard
assay conditions. As negative control 10 mL of Milli Q H O
2
instead of the enzyme was added to the assay solution.
[
7] O. Simova, A. Irintchev, A. Mehanna, J. Liu, M. Dihne,
D. Bachle, N. Sewald, G. Loers, M. Schachner, Ann.
Neurol. 2006, 60, 430–437.
Mass Spectrometry
To confirm the integrity of products 2 and 4 they were ana-
lyzed by HPLC/ESI mass spectrometry (Thermo Finnigan
Surveyorꢂ MSQꢃ) using the following settings: negative
ESI ionization, probe temperature: 4008C; needle voltage:
[
8] a) A. Togayachi, T. Akashima, R. Ookubo, T. Kudo, S.
Nishihara, H. Iwasaki, A. Natsume, H. Mio, J. Inoku-
chi, T. Irimura, K. Sasaki, H. Narimatsu, J. Biol. Chem.
2
001, 276, 22032–22040; b) E. Ong, M. Suzuki, F. Belot,
4.5 kV; cone voltage: 100 V. 10 mL of a 0.1 mM sample solu-
J.-C. Yeh, I. Franceschini, K. Angata, O. Hindsgaul, M.
Fukuda, J. Biol. Chem. 2002, 277, 18182–18190; c) T.
Kouno, Y. Kizuka, N. Nakagawa, T. Yoshihara, M.
Asano, S. Oka, J. Biol. Chem. 2011, 286, 31337–31346.
9] a) S. Yamamoto, S. Oka, Trends Glycosci. Glycotech-
nol. 2001, 13, 193–208; b) S. Yamamoto, S. Oka, F.
Saito-Ohara, J. Inazawa, T. Kawasaki, J. Biochem.
tion were concentrated on a Multospher 120 RP 18 HP-3 m
À1
(
60”2 mm), and applied at a flow rate of 0.2 mLmin using
50% acetonitrile in water as a mobile phase.
[
NMR Spectroscopy
1
H NMR spectra were measured on a Bruker AV 400 Ultra
2
002, 131, 337–347.
10] a) S. Kakuda, S. Oka, T. Kawasaki, Protein Expr. Purif.
004, 35, 111–119; b) S. Kakuda, T. Shiba, M. Ishiguro,
H. Tagawa, S. Oka, Y. Kajihara, T. Kawasaki, S. Wakat-
suki, R. Kato, J. Biol. Chem. 2004, 279, 22693–22703.
11] M. Fondeur-Gelinotte, V. Lattard, S. Gulberti, R. Oriol,
G. Mulliert, M. W. Coughtrie, J. Magdalou, P. Netter,
Shield spectrometer in D O (400 MHz; 308C). Chemical
2
[
[
shifts are given in the d scale and coupling constants in Hz.
2
Adv. Synth. Catal. 2015, 357, 1751 – 1762
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1761

Leonie Engels et al.
FULL PAPERS
M. Ouzzine, S. Fournel-Gigleux, Glycobiology 2007, 17,
57–867.
12] E. Yavuz, S. Drouillard, E. Samain, I. Roberts, B.
[20] a) J. W. Huh, H. Y. Yoon, H. J. Lee, W. B. Choi, S. J.
Yang, S. W. Cho, J. Biol. Chem. 2004, 279, 37491–
37498; b) B. J. Sommer, J. J. Barycki, M. A. Simpson, J.
Biol. Chem. 2004, 279, 23590–23596.
8
[
[
[
Priem, Glycobiology 2008, 18, 152–157.
13] L. Bastide, B. Priem, S. Fort, Carbohydr. Res. 2011, 346,
[21] K. E. Easley, B. J. Sommer, G. Boanca, J. J. Barycki,
M. A. Simpson, Biochemistry 2007, 46, 369–378.
[22] S. Egger, A. Chaikuad, K. L. Kavanagh, U. Opper-
mann, B. Nidetzky, J. Biol. Chem. 2011, 286, 23877–
23887.
3
48–351.
14] L. Engels, L. Elling, in: Carbohydrate-Modifying Bio-
catalysts, (Ed.: P. Grünwald), Standford Publishing,
Standford, 2011, pp 237–251.
[
15] a) C. Rech, R. R. Rosencrantz, K. K rˇ enek, H. Pelan-
tovµ, P. Bojarovµ, C. E. Roemer, F.-G. Hanisch, V.
K rˇ en, L. Elling, Adv. Synth. Catal. 2011, 353, 2492–
[23] T. Eixelsberger, B. Nidetzky, Adv. Synth. Catal. 2014,
356, 3575–3584.
[24] a) A. Zervosen, A. Stein, H. Adrian, L. Elling, Tetrahe-
dron 1996, 52, 2395–2404; b) T. Bülter, C. Wandrey, L.
Elling, Carbohydr. Res. 1997, 305, 469–473; c) T. Bülter,
L. Elling, J. Mol. Catal. B: Enzymatic 2000, 8, 281–284;
d) L. Elling, C. Rupprath, N. Günther, U. Rçmer, S.
Verseck, P. Weingarten, G. Dräger, A. Kirschning, W.
Piepersberg, ChemBioChem 2005, 6, 1423–1430.
[25] M. Negishi, J. Dong, T. A. Darden, L. G. Pedersen,
L. C. Pedersen, Biochem. Biophys. Res. Commun. 2003,
303, 393–398.
2
500; b) C. Rupprath, M. Kopp, D. Hirtz, R. Müller, L.
Elling, Adv. Synth. Catal. 2007, 349, 1489–1496; c) B.
Sauerzapfe, L. Elling, in Multi-Step Enzyme Catalysis,
Wiley-VCH, 2008, pp 83–107.
[
[
16] B. Sauerzapfe, L. Engels, L. Elling, Enzyme Microb.
Technol. 2008, 43, 289–296.
17] a) W. Hummel, B. Riebel, Biotechnol. Lett. 2003, 25,
5
1–54; b) B. Geueke, B. Riebel, W. Hummel, Enzyme
Microb. Technol. 2003, 32, 205–211; c) M. Kuzu, K. Nie-
find, W. Hummel, D. Schomburg, Acta Crystallogr.
Sect. F 2005, 61, 528–530.
[26] C. Breton, L. Snajdrova, C. Jeanneau, J. Koca, A. Im-
berty, Glycobiology 2006, 16, 29R–37R.
[
18] a) A. Radominska-Pandya, S. M. Bratton, M. R. Redin-
bo, M. J. Miley, Drug Metab. Rev. 2010, 42, 133–144;
b) A. Alonen, O. Aitio, K. Hakala, L. Luukkanen, M.
Finel, R. Kostiainen, Drug Metab. Dispos. 2005, 33,
[27] M. Fondeur-Gelinotte, V. Lattard, R. Oriol, R. Molli-
cone, J. C. Jacquinet, G. Mulliert, S. Gulberti, P. Netter,
J. Magdalou, M. Ouzzine, S. Fournel-Gigleux, Protein
Sci. 2006, 15, 1667–1678.
6
57–663.
[28] a) K. Adamiak, T. Anders, M. Henze, H. Keul, M.
Mçller, L. Elling, J. Mol. Catal. B: Enzymatic 2012, 84,
108–114; b) T. Anders, K. Adamiak, H. Keul, L. Elling,
M. Mçller, Macromol. Biosci. 2011, 11, 1201–1210.
[29] U. Rçmer, H. Schrader, N. Günther, N. Nettelstroth,
W. B. Frommer, L. Elling, J. Biotechnol. 2004, 107, 135–
149.
[
19] a) M. Kittelmann, U. Rheinegger, A. Espigat, L.
Oberer, R. Aichholz, E. Francotte, O. Ghisalba, Adv.
Synth. Catal. 2003, 345, 825–829; b) V. Kren, J. Ulricho-
va, P. Kosina, D. Stevenson, P. Sedmera, V. Prikrylova,
P. Halada, V. Simanek, Drug Metab. Dispos. 2000, 28,
1
513–1517.
1762
asc.wiley-vch.de
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1751 – 1762