Download Expression and identification of the RfbE protein from Vibrio

Glycobiology vol. 11 no. 8 pp. 655–661, 2001
Expression and identification of the RfbE protein from Vibrio cholerae O1 and its use for
the enzymatic synthesis of GDP-D-perosamine
Christoph Albermann and Wolfgang Piepersberg1
Chemische Mikrobiologie, Bergische Universität GH Wuppertal,
Gauss-Str. 20, D-42097, Germany
Received on January 15, 2001; revised on March 26, 2001; accepted on
March 28, 2001
The 4-amino-6-deoxy-monosaccharide D-perosamine is an
important element in the glycosylation of interesting cell
products, such as antibiotics and lipopolysaccharides (LPS) of
Gram-positive and Gram-negative bacteria. The biosynthetic
pathway of the precursor molecule, GDP-D-perosamine, in
Vibrio cholerae O1 starts with an isomerisation of fructose6-phosphate catalyzed by the bifunctional enzyme phosphomannose isomerase–guanosine diphosphomannose
pyrophosphorylase (RfbA; E.C. 2.7.7.22) creating the
intermediate mannose-6-phosphate, which is subsequently
converted by the phosphomanno-mutase (RfbB; E.C.
5.4.2.8) and further by RfbA to GDP-D-mannose, to GDP-4
keto-6-deoxymannose by a 4,6-dehydratase (RfbD; E.C.
4.2.1.47) and finally to GDP-D-perosamine by an
aminotransferase (RfbE; E.C. not yet classified).
We cloned the rfbD and the rfbE genes of V. cholerae O1
in Escherichia coli expression vectors. Both biosynthetic
enzymes were overproduced in E. coli BL21 (DE3) and
their activities were analyzed. The enzymatic conversion
from GDP-D-mannose to GDP-D-perosamine was optimized
and the final product, GDP-D-perosamine, was purified
and identified by nuclear magnetic resonance, mass
spectrometry, and chromatography. The catalytically
active form of the GDP-4-keto-6-deoxy-D-mannose-4-aminotransferase seems to be a tetramer of 170 kDa. The His-tag
RfbE fusion protein has a Km of 0.06 mM and a Vmax value
of 38 nkat/mg protein for the substrate GDP-4-keto-6deoxy-D-mannose. The Km and Vmax values for the cosubstrate
L-glutamate were 0.1 mM and 42 nkat/mg protein, respectively. The intention of this work is to establish a basis for
both the in vitro production of GDP-D-perosamine and for
an in vivo perosaminylation system in a suitable bacterial
host, preferably E. coli.
Key words: GDP-α-D-perosamine/6-deoxyhexose/enzymatic
synthesis/GDP-4-keto-6-deoxy-D-mannose 4-aminotransferase
Introduction
The 6-deoxyhexose D-perosamine is an important and characteristic element in glycoconjugates of bacteria. In Gram-positive
1To
whom correspondence should be addressed
© 2001 Oxford University Press
bacteria, for example, the D-perosamine was first detected in the
antibiotic perimycin of Streptomyces coelicolor var. aminophilus
(Lee and Schaffner, 1966, 1969). In Gram-negative bacteria it
is also found as a main constituent of lipopolysaccharides
(LPS) in Brucella spp., Eschericha coli O157, and Vibrio
cholerae O1 (Wu and Mackenzie, 1987; Bilge et al., 1996;
Villeneuve et al., 2000).
The O-antigen of V. cholerae O1, the causative agent of
cholera, is a homopolymer of the dideoxy amino sugar
perosamine, which is substituted via the amino group with a
4-carbon aliphatic acid, 3-deoxy-L-glycero-tetronic acid. This
structure is repeated on the average 18 times and is joined to a
lipid A molecule by a linker containing 6-, 7-, and 8-carbon
sugars (core unit) (Hisatsune et al., 1993). The introduction of
the perosamine monosaccharidic unit into the glycosylated
endproducts is dependent on nucleotide activation as a GDPhexose and subsequent modification. The rfb region of V. cholerae O1 encodes the gene cluster that is responsible for the
perosamine biosynthesis, the 3-deoxy-L-glycero-tetronate
biosynthesis, the O-antigen transport, and of the O-antigen
modification (Hisatsune et al., 1996; Stroeher et al., 1998).
The perosamine biosynthesis cluster consists of four genes that
are predicted to encode the enzymes for the conversion of fructose-6-phosphate to the activated sugar metabolite GDP-Dperosamine in five steps (Stroeher et al., 1995). The first step is
the conversion of fructose-6-phosphate to mannose-6-phosphate catalyzed by the phosphomannose isomerase–guanosine
diphosphomannose pyrophosphorylase (RfbA; E.C. 2.7.7.22);
subsequently, the mannose-6-phosphate is converted to
mannose-1-phosphate by the phosphomanno-mutase (RfbB;
E.C. 5.4.2.8). The third step, the nucleotide activation, is catalyzed once again by the bifunctional phosphomannose
isomerase–guanosine diphosphomannose pyrophosphorylase
(RfbA) to GDP-D-mannose. The two further biosynthetic steps
characteristic for the formation GDP-D-perosamine basically
follow that of other 6-deoxyhexoses, which are made via the
dTDP- or CDP-activated sugars. It starts with a dehydratation
reaction catalyszed by an NAD+-dependent NDP-hexose 4,6dehydratase, which forms NDP-6-deoxy-D-4-hexulose
(Piepersberg, 1994; Liu and Thorson, 1994; Piepersberg and
Distler, 1997). By further reactions this intermediate can be
modified by a large variety of following steps, for example, the
4-carbonyl group can be enantioselectively reduced, transaminated, or fully reduced. In other cases the NDP-6-deoxy-D-4hexuloses are converted first by an 3,5-epimerase reaction to
NDP-6-deoxy-L-4-hexulose. In the case of GDP-D-perosamine,
it is supposed that the activated sugar metabolite GDP-Dmannose is converted by the GDP-D-mannose 4,6-dehydratase
(RfbD; E.C. 4.2.1.47) to GDP-4-keto-6-deoxy-D-mannose
655
C. Albermann and W. Piepersberg
followed by a transamination catalyzed by the perosamine
synthetase (RfbE; E.C. not yet classified) (Figure 1).
To produce novel glycoconjugates or oligosaccharides by
the use of glycosyltransferases, it is necessary to synthesize the
required sugar donors in a large scale. Other nucleotideactivated sugars, for instance, GDP-D-mannose, GDP-Lfucose, or dTDP-L-rhamnose, are already available fully on the
basis of enzymatic reactions (Elling et al., 1996; Pastuszak et
al., 1998; Elling and Bütler, 1999; Albermann et al., 2000)
with the respective enzymes overproduced in a suitable expression system such as E. coli. In this article we describe the
cloning and the overproduction of the RfbE (GDP-D-perosamine synthetase) from V. cholerae O1. This work presents
the evidence that the final product of the enzymatic reaction
catalyzed by RfbE-protein is GDP-α-D-perosamine.
Results
Cloning and expression of GDP-D-mannose 4,6-dehydratase
(RfbD) and GDP-D-perosamine synthetase (RfbE)
analysis. Additionally the genes were cloned into the pET16b
expression vector to obtain their products as N-terminal His-tag
fusion proteins. The plasmids pCAW 13.1 (rfbD), pCAW 13.2
(His-tag rfbD), pCAW 14.1 (rfbE), and pCAW 14.2 (His-tag
rfbE) constructed this way were transformed into E. coli BL21
(DE3) pLysS, where their expression under induced conditions
resulted in the overproduction of large amounts of RfbD and
RfbE proteins. The apparant molecular masses in analyses by
sodium dodecyl sulfate–polyacrylamide gel electrophoresis
(SDS–PAGE) agreed with the calculated masses of the proteins
(RfbD 42.0 kDa; His-tag RfbD 44.5 kDa; RfbE 40.9 kDa; His-tag
RfbE 43.4 kDa). Crude ribosome-free extracts from the
recombinant E. coli strains were preliminarily tested for the
activities of GDP-D-mannose 4,6-dehydratase and of GDP-4keto-6-deoxy-D-mannose 4-aminotransferase (GDP-D-perosamine
synthetase). The recombinant native RfbD protein and the
His-tag fusion protein variant of RfbD did not show any
activity. Therefore the recombinant GDP-D-mannose 4,6dehydratase (Gmd; plasmid pCAW21.1) of E. coli K12 was
The two open reading frames rfbD and rfbE of the LPS biosynthesis cluster of V. cholerae strain O1 (Stroeher et al., 1992;
Manning et al., 1995) are predicted to encode a GDP-Dmannose 4,6-dehydratase and GDP-D-perosamine synthetase.
The RfbD protein has high similarity with other 4,6-dehydratases from E. coli, Salmonella enterica, Yersinia enterocolitica, or Helicobacter pylori. The RfbE protein possesses a
pyridoxal-phosphate-binding domain, which is supposed to be
a group VI (secondary metabolic) aminotransferase and is in
the cluster of orthologous proteins COG0399 (Tatusov et al.,
2001) (Figure 2). Both genes were amplified and modified by
polymerase chain reaction (PCR), respectively, to introduce each
an N-terminal NdeI, overlapping the ATG start codon, and a
BamHI restriction site downstream from the stop codon. These
were used to clone the respective PCR products into the
expression vector pET11a and were verified by sequence
Fig. 1. Biosynthetic pathway of GDP-α-D-perosamine. The steps are catalysed
by the following enzymes: (1) Phosphomannose isomerase–guanosine
diphosphomannose pyrophosphorylase (RfbA), (2) phosphomannomutase
(RfbB), (3) phosphomannose isomerase–guanosine diphosphomannose
pyrophosphorylase (RfbA), (4) GDP-α-D-mannose 4,6-dehydratase (RfbD,
Gmd), (5) GDP-4-keto-6-deoxy-α-D-mannose 3,5-epimerase-4-reductase
(RfbE).
656
Fig. 2. Alignment of the V. cholerae O1 RfbE protein (acc. no. X59554 )
(1) with amino acid sequences of (2) the GDP-perosamine-synthetase PerA
from Brucella melitensis 16M (acc. no. AF047478), (3) LmbS from
Streptomyces lincolnensis 78–11 (acc. no. X79146) and (4) StrS from
Streptomyces griseus N2311 (acc. no. Y00459). All four proteins are in cluster
COG0339 (Tatusov et al., 2001).
RfbE protein from V. cholerae
used; it had previously already been shown to catalyze the
synthesis of GDP-4-keto-6-deoxy-D-mannose (Sturla et al.,
1997; Albermann et al., 2000). This finding was surprising,
because the amino acid sequence of the RfbD protein (V. cholerae
O1) and the Gmd protein (E. coli K12) have 79% identity and
90% homology. In contrast, both the native RfbE protein and its
His-tag fusion variant were catalytically active in the conversion
of GDP-4-keto-6-deoxy-D-mannose to GDP-D-perosamine.
Purification and properties of the His-tag fusion protein RfbE
After harvesting of E. coli BL21 (DE3) pLysS cells containing
the plasmid pCAW14.2, a crude extract was prepared and
chromatographed on a Ni-NTA-agarose column, which bound
the His-tag RfbE fusion protein tightly. The elution of the
recombinant protein from the affinity column was accomplished by increasing the imidazole concentration up to
250 mM. The imidazole was removed from the protein fraction
by gel filtration. The eluted fractions each contained one major
band in an SDS–PAGE analysis (Figure 3) consistent with the
calculated molecular masses of His-tag Gmd (44.5 kDa) and of
His-tag RfbE (43.4 kDa), respectively.
A gel permeation analysis was used to investigate the
molecular mass of the His-tag RfbE fusion protein under
native conditions. The apparent mass of 170 kDa, measured by
comparison of the elution volume of the His-tag RfbE peak
with those of known molecular weight standards, suggested
that the native protein is a tetramer.
Determination of enzyme specificity and specific activity
The conversion of GDP-α-D-mannose to GDP-4-keto-6-deoxyamnnose and futher to GDP-D-perosamine was monitored by
high-performance liquid chromatography (HPLC) (Figure 4).
In coupled reaction assays, with recombinant Gmd and RfbE
or His-tag RfbE, the complete conversion of GDP-α-Dmannose to GDP-D-perosamine could be obtained. As an
amino-donor out of the tested alternatives the amino acid Lglutamate was used as substrate of the RfbE protein. From the
other tested aminodonors (L-alanine, L-arginine, L-aspartate,
L-glycine, L-glutamine, L-leucine, L-lysine, L-ornithine, L-phenylalanine, L-serine) only L-glutamine showed activity, for which
the conversion rate was only 10% of the activity compared
with L-glutamate. It was also shown that the addition of the
cosubstrate pyridoxal-5-phosphate was essential for catalysis
of the aminotransfer reaction by RfbE. The specific activities
of the crude extract of E. coli BL21(DE3) pLysS pCAW 14.2
and of the purified His-tag fusion protein under the given
reaction conditions were 8 and 44 nkat/mg protein, respectively.
This indicated approximately a 6.5-fold enrichment of the
purified protein. In the negative controls, that is, crude extracts
from E. coli BL21 (DE3) pLysS with either vector plasmid,
pET11a, or pET16b, no activity of either GDP-D-mannose 4,6dehydratase or GDP-D-perosamine synthetase could be
detected under the same conditions.
For the determination of the kinetic properties the concentration
of GDP-4-keto-6-deoxy- D-mannose was measured by the UV
absorption in alkaline medium at 320 nm. The Km and Vmax
values for the substrate GDP-4-keto-6-deoxy-D-mannose were
0.06 mM and 38 nkat/mg protein, respectively. The Km and
Vmax values for the cosubstrate L-glutamate were 0.1 mM and
42 nkat/mg protein, respectively. The reaction catalyzed by the
His-tag RfbE protein had its pH optimum at 8.0, and the
optimum concentration of the divalent cation Mg2+ was
20 mM.
Preparative synthesis and isolation of GDP-α-D-perosamine
To identify the product of the RfbE reaction, we used the overproduced enzyme GDP-D-mannose 4,6-dehydratase (Gmd)
and the purified GDP-D-perosamine synthetase (His-tag RfbE)
for a preparative conversion of GDP-α-D-mannose. No
product inhibition of the GDP-D-mannose 4,6-dehydratase by
Fig. 3. Purification of the His-tag RfbE protein by Ni-NTA-agarose. The
SDS–PAGE analysis of the following protein fractions is shown in the lanes:
(1) Low molecular weight marker; (2) S30 of E. coli BL21 pLysS pCAW14.2
(His tag RfbE); (3) wash step 1 with 20 mM imidazole; (4) wash step 2 with
50 mM imidazole; (5) His-tag RfbE after purification on Ni-NTA-agarose and
gel filtration; (6) elution step with 250 mM imidazole; (7) S30 of E. coli BL21
pLysS pCAW14.1 (native RfbE).
Fig. 4. HPLC analysis of the enzymatic reactions catalyzed by Gmd and RfbE.
(A) Enzyme assay after 1 min of incubation, GDP-α-D-perosamine (7 min),
GDP-α-D-mannose (24 min), GDP-4-keto-6-deoxy-D-mannose (37 min).
(B) Enzyme assay after complete conversion, GDP-α-D-perosamine (7 min).
657
C. Albermann and W. Piepersberg
GDP-D-perosamine could be observed. Such a one-pot method
was applied to the complete synthesis of GDP-D-perosamine. The
synthesis was started from GDP-α-D-mannose and the formation
of the product GDP-D-perosamine and the consumption of the
precursors, GDP-α-D-mannose and GDP-4-keto-6-deoxymannose, was monitored by HPLC and of L-glutamate by thinlayer chromatography (TLC). After removal of proteins and
separation of the side product, α-ketoglutarate, by ion
exchange chromatography and size exclusion chromatography
the final product GDP-D-perosamine was obtained in 70%
overall yield. The result of 1H-nuclear magnetic resonancy
(NMR); 1H1H-COSY; 31P-NMR spectroscopy (Table I), of the
mass spectrometry (MS) (GDP-α-D-perosamine sodium salt
[(M + Na)+] 611.13 amu) and the TLC analysis of the D-perosamine component proved the structure of the product to be in
accord with GDP-α-D-perosamine.
Discussion
The intention of this work was to identify the function of the
RfbE protein by an overexpression of the rfbE gene and an
in vitro production of GDP-α-D-perosamine. As yet, the
specific function of a GDP-D-perosamine-synthetase has not
been directly proven by in vitro assays. Rather it could been
shown by mutational studies that the encoded proteins are
involved in the biosynthesis of the LPS antigens (Hisatsune
Table I. 1H-, 31P-NMR, chemical shifts, and coupling for GDP-D-perosamine
Residue
1H/31P
α-D-Perosamine
H-1″
H-2″
H-3″
H-4″
H-5″
Ribose
Chemical
shift (p.p.m.)
5.41
3.97
3.94
3.01
4.08
Coupling
constants (Hz)
3J
1″2″
3J
PH1″
3J
2″3″ 3.5
3J
1″2″
1.8
3J
3″4″
10.1
3J
2″3″
3.5
3J
3″4″
10.1
3J
4″5″
9.8
3J
5″6″
6.2
3J
4″5″
9,8
3J
5″6″
6.2
1.8
7.1
Materials and methods
H-6″ (CH3)
1.21
H-5′ a,b
4.20
H-4′
4.31
H-3′
4.45
3J
2′3′
3.7
H-2′
4.70
3J
1′2′
6.1
3J
2′3′
3.7
Bacterial strains, growth conditions, and media
3J
1′2′
6.1
The bacterial strains used in this study were E. coli DH5α and
E. coli BL21 (DE3) pLysS. Strains of E. coli were grown at
37°C in Luria-Bertani broth (LB) (Miller, 1972); for solid
media agar-agar in a concentration of 18 g/L was added.
Antibiotics were used at the following final concentrations:
ampicillin 100 µg/ml, chloramphenicol 25 µg/ml. For induc-
H-1′
5.90
Base
H-1′
8.10
Phosphate
P-1
–10.1
3J
PP
19.6
–12.6
3J
PP
19.6
3J
PH-1″
P-2
658
et al., 1996; Godfroid et al., 1998; Cloeckaert et al., 2000). The
RfbE protein shows homology to pyridoxal-phosphate-binding
proteins (COG0399; Tatusov et al., 2001), which in part catalyze the aminogroup transfer onto deoxysugars, like NysDII
(3-amino-3,6-dideoxy-D-mannose, mycosamine) in Streptomyces noursei (Brautaset et al., 2000). Therefore, as the
required tools we herein evaluate the overexpressed D-perosamine biosynthetic enzymes RfbD and RfbE from V. cholerae
O1, as well as the purification of GDP-D-perosaminesynthetase RfbE to homogeneity and its characterization.
Surprisingly, the native and the His-tag RfbD protein were
enzymatically inactive, even though they could be overexpressed in E. coli BL21 pLysS as soluble proteins. This is
even though the amino acid sequence of the RfbD protein
(V. cholerae) and the Gmd protein (E. coli) are 79% identical,
indicating a high degree of structural congruence in the 3D
structure. However, also in case of the Gmd protein of E. coli
it is known that the enzyme is unstable (Sturla et al., 1997) and
that the N-terminal His-tag fusion variant of the Gmd protein is
also inactive (Albermann et al., 2000). Nevertheless, the recombinant Gmd protein from E. coli K12 catalyzed the formation of
GDP-D-mannose to GDP-4-keto-6-deoxymannose (Sullivan et
al., 1998; Albermann et al., 2000). The GDP-4-keto-6-deoxymannose, which can be synthesized this way, is used as the
substrate of the GDP-4-keto-6-deoxy-D-mannose 4-aminotransferase (perosamine synthetase; RfbE). The recombinant
RfbE proteins, the native and the His-tag fusion proteins,
respectively, converted GDP-4-keto-6-deoxymannose in presence of L-glutamate to a new product. The NMR and MS analyses and the results of the chromatography supported that this
product of the Gmd and RfbE catalyzed reactions is GDP-αD-perosamine. The in vitro preparation could be carried out
efficiently in a one-step (“one-pot”) reaction system. In
contrast to GDP-β-L-fucose (Sturla et al., 1997), the GDP-αD-perosamine did not show any effect of product inhibition on
the activity of the GDP-D-mannose 4,6-dehydratase. Whether
the GDP-α-D-perosamine exerts feedback inhibition of other
biosynthetic enzymes, for instance, RfbA, RfbB, and RfbD, of
V. cholerae O1 will be the subject of further investigations.
In the present work, we have established a preparative
enzymatic route to the synthesis of GDP-α-D-perosamine by
use of the two recombinant enzymes, Gmd and RfbE, which
can easily be used now for the production of GDP-α-D-perosamine, at gram scale.
7.1
Biochemicals
All protein standards, antibiotics, isopropyl-β-thiogalactoside
(IPTG), and nucleotides were obtained from Sigma (Deisenhofen,
Germany); GDP-α-D-mannose was obtained from Calbiochem
(Schwalbach, Germany).
RfbE protein from V. cholerae
tion of the lac operator IPTG was used at a final concentration
of 24 µg/ml.
Cloning of rfbD and rfbE
The DNA-sequence of the rfb-gene cluster for lipopolysaccharide
biosynthesis in V. cholerae O1 has been described by others
(Manning et al., 1995). The genes rfbD and rfbE were amplified
by PCR with Vent-polymerase (New England BioLabs,
Schwalbach/Taunus, Germany). The following primers, used
to amplify the genes rfbD and rfbE from the chromosomal
DNA of V. cholerae O1 were designed on the basis of the
published sequences and were obtained from Interactiva (Ulm,
Germany): for the rfbD gene 5′-GGATATTTACATATGAATAAAAAAGTTG-3′ (forward direction); 5′-CAGGAATCATTTAAAAGATCTCACTCTAC-3′ (reverse direction); for
rfbE 5′-GTGAGGTCCTTCATATGATTCCTGTAT-3′ (forward
direction); 5′-GGAGGTAAGGGATCCCAAACTACTA-3′
(reverse direction). The products of the PCR amplification
were further treated with NdeI and BglII restriction enzymes
(Gibco BRL, Eggenstein, Germany) in the case of the rfbD
fragment and with NdeI and BamHI for the rfbE PCR product,
and were ligated into the vectors pET11a and pET16b (Novagen,
Madison, WI), which were hydrolyzed with NdeI and BamHI,
respectively (Sambrook et al., 1989). After ligation with T4DNA-ligase (BioLabs, Schwalbach/Taunus), the recombinant
plasmids pET11a/rfbD (pCAW 13.1), pET11a/rfbE (pCAW
14.1), pET16b/rfbD (pCAW 13.2), and pET16b/rfbE (pCAW
14.2) were obtained and transformed into competent cells of
E. coli DH5α (Hanahan, 1983). The respective cloning of the
gmd gene from E. coli K12 in plasmid pCAW21.1 had been
published previously (Albermann et al., 2000).
DNA sequencing
The insert structure of each recombinant derivative was
verified by restriction analysis and DNA sequencing. The
sequence of the amplified DNA was determined by the dideoxynucleotide chain-termination method (Sanger et al., 1977)
using the Thermo Sequenase fluorescent labeled primer cycle
sequencing kit with 7-deaza-dGTP (Amersham, Braunschweig,
Germany) and an A.L.F.-Express DNA sequencer (Pharmacia,
Freiburg, Germany). The sequences of both strands each were
determined from the DNA of double stranded plasmids
prepared by the QIAprep spin miniprep kit (Qiagen, Hilden,
Germany).
Overproduction of GDP-D-mannose 4,6-dehydratase (RfbD,
Gmd) and GDP-4-keto-6-deoxy-D-mannose-4aminotransferase (RfbE)
The recombinant plasmids (pCAW 13.1, pCAW 13.2, pCAW
14.1, pCAW 14.2, pCAW 21.1) were retransformed into
competent E. coli BL21 (DE3) pLysS (Noreagen). The E. coli
BL21 (DE3) pLysS cells with the recombinant plasmids were
grown in LB media to an optical density of 0.6 at 540 nm. Then
the cells were induced by IPTG for 90 min. Subsequently, the
cells were harvested by centrifugation, washed twice in icecold 50 mM Tris–HCl buffer, pH 8, and suspended in extraction
buffer (50 mM Tris–HCl buffer, pH 8, 150 mM NaCl, 20 mM
MgCl2, 5 mM β-mercaptoethanol, 5 mM pyridoxal-5-phosphate). After disruption by using a 5-ml French press cell
(Aminco, Silver Spring, MD) the crude extract was clarified by
centrifugation at 30,000 × g for 30 min (S30).
Purification of RfbE-His-tag fusion protein
The RfbE His-tag fusion-protein was purified as soluble
protein by affinity chromatography, using Ni-NTA-agarose
(Qiagen) (Le Grice and Grueninger-Leitch, 1990; Schmitt et
al., 1993). The crude extract (containing His-tag RfbE) was
loaded on a Ni-NTA-agarose-column (15 mm × 60 mm) which
was equilibrated with 50 mM Tris–HCl buffer, pH 8, with 300
mM NaCl, 10 mM MgCl2, and 5 mM β-mercaptoethanol. The
loaded column was washed with 50 mM Tris–HCl buffer, pH 8,
containing 300 mM NaCl, 10 mM MgCl2, 5 mM β-mercaptoethanol, and 20 mM imidazole. After the elution with 50 mM
Tris–HCl buffer, pH 8, with 300 mM NaCl, 10 mM MgCl2,
5 mM β-mercaptoethanol, 250 mM imidazole, the proteincontaining fractions were analysed by SDS–PAGE (Laemmli,
1970) to check their purity and molecular masses. The finial
purification step was carried out by gel filtration with a
Sepharyl S-200 HR column (2.5 × 90 cm) (Pharmacia). The
column was equilibrated with 50 mM Tris–HCl buffer, pH 8,
with 10 mM MgCl2, 5 mM β-mercaptoethanol, and 300 mM
NaCl.
Characterization of recombinant His-tag RfbE protein
The size determination of the native His-tag RfbE was
performed by size exclusion chromatography on an HPLC
system (Pharmacia) with a Superose 12 gel fitration column
(Pharmacia). A 50 mM Tris–HCl buffer, pH 8, containing
150 mM NaCl was used as mobile phase, at a flow rate 0.5 ml/min.
For the calibration curve the following proteins were applied:
ovalbumin 43 kDa; bovine serum albumin 67 kDa; alcohol
dehydrogenase of yeast 150 kDa; β-amylase of potato
200 kDa; apoferritin from horse spleen 443 kDa; bovine
thyroglobulin 669 kDa.
HPLC
GMP, GDP, GTP, GDP-α-D-mannose, GDP-4-keto-6-deoxyand GDP-α-D-perosamine were separated by
HPLC and detected by UV photometry (Beckman, Munich,
Germany) at 260 nm. As stationary phase a reversed phase
column, Eurospher ODS18, 5 µm, 250 × 4.6 mm (Knauer,
Berlin, Germany) was used. As mobile phase a phosphate
buffer (30 mM potassium phosphate, pH 6.0; 5 mM tetrabutylammonium hydrogensulfate, 2% acetonitrile) and acetonitrile
was used (Payne and Ames, 1982).
D-mannose
TLC
To analyze the sugar components, the nucleotide moieties were
cleaved from the sugar by incubating in 1 M HCl for 10 min at
100°C. The resulting mixture and the standards of 6-deoxy-Lmannose, 6-deoxy-L-galactose, D-glucose, 2-amino-2-deoxyD-glucose, 2-amino-2-deoxy-D-galactose, 2-amino-2-deoxy-Dmannose, and D-mannose were chromatographed on TLC
sheets, Silica gel F254 (Merck, Darmstadt, Germany), using a
solvent system consisting of acetic acid:chloroform:water
(7:6:1). All the sugar components were localized by naphthoresorcinol staining (0.2 g naphthoresorcinol in 96 ml ethanol
and 4 ml concentrated sulfuric acid), aminosugars by MorganElson reagent (Chaplin and Kennedy, 1986). L-glutamate was
chromatographied by the same mobile and stationary phase
(see above) and detected by staining with ninhydrine (0.3 g
ninhydrine in 100 ml butanol).
659
C. Albermann and W. Piepersberg
Enzyme assays
For the measurement of the activity of the GDP-D-mannose
4,6-dehydratase (Gmd, RfbD) each a spectroscopic (Okazaki
et al., 1962; Kornfeld and Ginsburg, 1966) and chromatographic test system was used to estimate the increase of GDP4-keto-6-deoxy-D-mannose. The standard assay contained
50 mM Tris–HCl buffer, pH 8, 10 mM MgCl 2, 4 mM GDPD-mannose, 50 µM NADP+, and different amounts of GDP-Dmannose 4,6-dehydratase containing crude extract in a final
volume of 100 µl. Protein concentrations were determined
according to Bradford (1976). The reactions were preformed at
37°C and measured at different times between 0 and 30 min.
For the photometric analysis, the reactions were stopped by
adding 950 µl 100 mM NaOH to 50 µl of the reaction mixtures;
the solutions were further incubated for 20 min at 37°C. After
this, the absorption was measured at 320 nm (ε320nm = 2200 L
mol–1 cm–1). For the HPLC analysis the samples were boiled
for 1 min, and the proteins were removed by centrifugation.
The enzyme activity of the GDP-4-keto-6-deoxy-D-mannose
4-aminotransferase RfbE was determined by measuring the
decrease of GDP-4-keto-6-deoxy-D-mannose (see above).
GDP-4-keto-6-deoxy-D-mannose was obtained by a previously published procedure (Albermann et al., 2000). The
standard assay contained 50 mM Tris–HCl buffer, pH 8, 2 mM
GDP-4-keto-6-deoxy-D-mannose, 20 mM MgCl2, 50 µM pyridoxal-5-phosphate, and different amounts of RfbE or His-tag
RfbE and 4 mM sodium-L-glutamate. Other tested aminodonor
cosubstrates (L-alanine, L-arginine, L-aspartate, L-glycine, Lglutamine, L-leucine, L-lysine, L-ornithine, L-phenylalanine,
L-serine) were also added at 4 mM final concentration. The
reaction was started by the addition of different concentrations of GDP-4-keto-6-deoxy-D-mannose (final volume 50 µl).
Also, the increase of GDP-α-D-perosamine was observed by
HPLC analysis.
Enzymatic synthesis of GDP-α-D-perosamine
The preparative synthesis of GDP-α-D-perosamine was carried
out in a coupled enzymatic reaction by recombinant Gmd from
E. coli K12 (plasmid pCAW21.1; Albermann et al., 2000) and
the purified His-tag RfbE protein. The assay contained
33 µmol (20 mg) GDP-D-mannose, sodium-L-glutamate 38 µmol
(7.1 mg), 50 mM Tris–HCl buffer, pH 7.5, 20 mM MgCl2,
50 µM pyridoxal phosphate, 50 µM NADP+, crude extract of
E. coli BL21 pLysS pCAW21.1 and His-tag RfbE, with a
GDP-D-mannose 4,6-dehydratase and GDP-4-keto-6-deoxy-Dmannose 4-aminotransferase activity each of 25 nkat in a final
volume of 3 ml. This mixture was incubated for 60 min at
37°C. The proteins were removed by boiling for 1 min and
subsequent centrifugation at 10,000 × g for 30 min. The time
course of the reaction to GDP-α-D-perosamine was monitored
by assaying small samples with HPLC analysis.
Isolation of GDP-α-D-perosamine
The separation of GDP-α-D-perosamine form the side products
α-ketoglutarate and L-glutamate was done by an anion-exchange
chromatography. This was performed with a Dowex 1 × 8 resin
(mesh 200–400; formate form) (Serva, Heidelberg, Germany),
elution with a linear gradient of ammonium-formate solution
from 0.05 M up to 1 M in 2 h at a flow rate of 5 ml/min. GDPα-D-perosamine-containing fractions were pooled and evapo660
rated to a volume of 8 ml under reduced pressure (20 mbar) at
20–25°C. For the desalting of the preparation a gel filtration on
a Sephadex G-10 column (SR 25/100; Pharmacia), with a total
volume of 420 ml, was used and eluted at a flow rate of 1 ml/min.
The GDP-α-D-perosamine-containing fractions were detected
by UV absorption, pooled, and applied to a membrane anion
exchanger Q15 (Sartorius, Göttingen, Germany) by which the
GDP-α-D-perosamine preparation was equilibrated against
150 mM NaCl to yield the sodium form. After another volume
reduction by evaporation to a volume of about 10 ml and a
further gel filtration (Sephadex G-10 column), the sodium
GDP-α-D-perosamine was lyophilized (Cryograph LCD-1,
Christ, Osterrode, Germany) (yield 14 mg).
The enzymatically synthesized GDP-α-D-perosamine was
analyzed by NMR spectroscopy: 1H-NMR (400 MHz, D2O) δ
(p.p.m.) 1.20 (d, H-6″ 3J5″6″ = 6.2 Hz); 3.01 (dd, H-4″ 3J3″4″ =
10.1 Hz, 3J4″5″ = 9.8 Hz); 3.94 (dd, H-3″ 3J3″4″ = 10.1 Hz, 3J2″3″
= 3.5 Hz); 3.97 (dd, H-2″ 3J2″3″ = 3.5 Hz, 3J1″2″ = 1.8 Hz); 4.08
(m, H-5″ 3J5″6″ = 6.2 Hz, 3J4″5″ = 9.8 Hz); 5.41 (H-1″ 3J1″2″ =
1.8 Hz, 3JPH1″ = 7.1 Hz); 4.20 (2H-5′); 4.31 (m, H-4′); 4.45
(dd, H-3′ 3J2′3′ = 3.7 Hz); 4.70 (dd, H-2′ 3J1′2′ = 6.1 Hz, 3J2′3′ =
3.7 Hz); 5.90 (dd, H-1′ 3J1′2′ = 6,1 Hz); 8.10 (s, H-1).
31P-NMR (160 MHz, D O) δ (p.p.m.) –10.1 (P-1 2J = 19.6 Hz);
2
PP
–12.6 (P-2 2JPP = 19.6 Hz, 3JP-2H-1″ = 7.1 Hz).
Acknowledgments
We thank S. Kuberski for expert technical assistance. We are
grateful to J. Hacker for providing DNA of V. cholerae O1 and
F. Wedekind for MS analysis. This work was supported by a
grant from Roche Diagnostics, Penzberg (Germany).
Abbreviations
HPLC, high performance liquid chromatography; IPTG,
isopropyl-β-thiogalactoside; LB, Luria-Bertani; LPS, lipopolysaccharide; MS, mass spectrometry; NMR, nuclear magnetic
resonance; PCR, polymerase chain reaction; SDS–PAGE,
sodium dodecyl sulfate–polyacrylamide gel electrophoresis;
TLC, thin-layer chromatography.
References
Albermann, C., Distler, J., and Piepersberg, W. (2000) Preparative synthesis of
GDP-beta-L-fucose by recombinant enzymes from enterobacterial sources.
Glycobiology, 10, 875–881.
Bilge, S.S., Vary, J.C., Dowell, S.F., and Tarr, P.I. (1996) Role of the
Escherichia coli O157:H7 O side chain in adherence and analysis of an rfb
locus. Infect. Immun., 64, 4795–4801.
Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye
binding. Anal. Biochem., 72, 248–254.
Brautaset, T., Sekurova, O.N., Sletta, H., Ellingsen, T.E., Strm, A.R., Valla, S.,
and Zotchev, S.B. (2000) Biosynthesis of the polyene antifungal antibiotic
nystatin in Streptomyces noursei ATCC 11455: analysis of the gene cluster
and deduction of the biosynthetic pathway. Chem. Biol., 7, 395–403.
Chaplin, M.F., and Kennedy, J.F. (1986) Carbohydrate analysis: a practial
approach. IRL Press, Oxford.
Cloeckaert, A., Grayon, M., Verger, J.M., Letesson, J.J., and Groidfroid, F.
(2000) Conservation of seven genes involved in the biosynthesis of the
lipopolysaccharide O-side chain in Brucella spp. Res. Microbiol., 151,
209–216.
RfbE protein from V. cholerae
Elling, L., and Bütler, T. (1999) Enzymatic synthesis of nucleotide sugars.
Glycoconj. J., 16, 147–159.
Elling, L., Ritter, J.E., and Verseck, S. (1996) Expression, purification and
characterization of recombinant phosphomannomutase and GDP-alpha-Dmannose pyrophosphorylase from Salmonella enterica, group B, for the
synthesis of GDP-alpha-D-mannose from D-mannose. Glycobiology, 6,
591–597.
Godfroid, F., Taminiau, B., Danese, I., Denoel, P., Tibor, A., Weynants, V.,
Cloeckaert, A., Godfroid, J., and Letesson, J.J. (1998) Identification of the
perosamine synthetase gene of Brucella melitensis 16M and involvement
of lipopolysaccharide O side chain in Brucella survival in mice and in
macrophages. Infect. Immun., 66, 5485–5493.
Hanahan, D. (1983) Studies on transformation of Escherichia coli plasmids.
J. Mol. Biol., 127, 100–104.
Hisatsune, K., Kondo, S., Isshiki, Y., Iguchi, T., and Haishima, Y. (1993)
Occurrence of 2-O-methyl-N-(3-deoxy-L-glycero-tetronyl)-D-perosamine
(4-amino-4, 6-dideoxy-D-manno-pyranose) in lipopolysaccharide from
Ogawa but not from Inaba O forms of O1 Vibrio cholerae. Biochem.
Biophys. Res. Commun., 190, 302–307.
Hisatsune, K., Kondo, S., Iguchi, T., Ito, T., and Hiramatsu, K. (1996)
Lipopolysaccharides of Escherichia coli K12 strains that express cloned
genes for the Ogawa and Inaba antigens of Vibrio cholerae O1: identification
of O-antigenic factors. Microbiol. Immunol., 40, 621–626.
Kornfeld, R.H., and Ginsburg, V. (1966) Control of synthesis of guanosine 5′diphosphate D-mannose and guanosine 5′-diphosphate L-fucose in bacteria.
Biochim. Biophys. Acta, 117, 79–87.
Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature, 277, 680–685.
Le Grice, S.F., and Grueninger-Leitch, F. (1990) Rapid purification of
homodimer and heterodimer HIV-1 reverse transcriptase by metal chelate
affinity chromatography. Eur. J. Biochem., 187, 307–314.
Lee, C.H., and Schaffner, C.P. (1966) Perimycin: chemistry of perosamine.
Tetrahedron Lett., 47, 5837–5840.
Lee, C.H., and Schaffner, C.P. (1969) Perimycin: the structure of some degradation products. Tetrahedron, 25, 2229–2232.
Liu, H.-W., and Thorson, J.S. (1994) Pathways and mechanisms in the biogenesis
of novel deoxysugars by bacteria. Ann. Rev. Microbiol., 48, 223–256.
Manning, P.A., Stroeher, U.H., Karageorgos, L.E., and Morona, R. (1995)
Putative O-antigen transport genes within the rfb region of Vibrio cholerae
O1 are homologous to those for capsule transport. Gene, 158, 1–7.
Miller, J.H. (1972) Experiments in molecular genetics. Cold Spring Harbor
Laboratory, New York.
Okazaki, R., Okazaki, T., Strominger, J.T., and Michelson, A.M. (1962)
Thymidine diphosphate 4-keto-6-deoxy-D-glucose, an intermediate in
thymidine diphosphate L-rhamnose synthesis in Escherichia coli strains.
J. Biol. Chem., 237, 3014–3026.
Pastuszak, I., Ketchum, C., Hermanson, G., Sjoberg, E.J., Drake, R., and
Elbein, A.D. (1998) GDP-L-fucose pyrophosphorylase. Purification, cDNA
cloning, and properties of the enzyme. J. Biol. Chem., 273, 30165–30174.
Payne, S.H., and Ames, B.N. (1982) A procedure for rapid extraction and
high-pressure liquid chromatographic separation of the nucleotides and
other small molecules from bacterial cells. Anal. Biochem., 123, 151–161.
Piepersberg, W. (1994) Pathway engineering in secondary metaboliteproducing actinomycetes. Crit. Rev. Biotechnol., 14, 251–285.
Piepersberg, W., and Distler, J. (1997) Aminoglycosides and sugar components in other secondary metabolites. In Rehm, H.-J., Reed, G., Pühler, A.,
and Stadler, P., eds., Biotechnology, 2nd ed., vol. 7. VCH-Verlagsgesellschaft, Weinheim, pp. 397–488.
Sambrook, J., Fritsch, E.F., and Maniatias, T. (1989) Molecular cloning: a
laboratory manual. Cold Spring Harbor Laboratory, New York.
Sanger, F., Nicklen, S., and Coulson, A.R. (1977) DNA sequencing with
chain-terminating inhibitors. Proc. Natl Acad. Sci. USA, 74, 5463–5467.
Schmitt, J., Hess, H., and Stunneberg, H.G. (1993) Affinity purification of
histidine-tagged proteins. Mol. Biol. Rep., 18, 223–230.
Stroeher, U.H., Karageorgos, L.E., Morona, R., and Manning, P.A. (1992)
Serotype conversion in Vibrio cholerae O1. Proc. Natl Acad. Sci. USA, 89,
2566–2570.
Stroeher, U.H., Karageorgos, L.E., Brown, M.H., Morona, R., and
Manning, P.A. (1995) A putative pathway for perosamine biosynthesis is
the first function encoded within the rfb region of Vibrio cholerae O1.
Gene, 166, 33–42.
Stroeher, U.H., Jedani, K.E., and Manning, P.A. (1998) Genetic organization
of the regions associated with surface polysaccharide synthesis in Vibrio
cholerae O1, O139 and Vibrio anguillarum O1 and O2: a review. Gene,
223, 269–282.
Sturla, L., Bisso, A., Zanaridi, D., Benatti, U., De Flora, A., and Tonetti, M.
(1997) Expression, purification and characterization of GDP-D-mannose 4,
6-dehydratase from Escherichia coli. FEBS Lett., 412, 126–130.
Sullivan, F.X., Kumar, R., Kriz, R., Stahl, M., Xu, GY., Rouse, J., Chang, X.J.,
Boodhoo, A., Potvin, B., and Cumming, D.A. (1998) Molecular cloning of
human GDP-mannose 4, 6-dehydratase and reconstitution of GDP-fucose
biosynthesis in vitro. J. Biol. Chem., 273, 8193–8202.
Tatusov, R.L., Natale, D.A., Garkavtsev, I.V., Tatusova, T.A., Shankavaram, U.T.,
Rao, B.S., Kiryutin, B., Galperin, M.Y., Fedorova, N.D., and Koonin, E.V.
(2001) The COG database: new developments in phylogenetic classification
of proteins from complete genomes. Nucleic Acids Res., 29, 22–28.
Villeneuve, S., Souchon, H., Riottot, M.-M., Mazie, J.C., Lei P., Glaudemans, C.P.,
Kovac, P., Fournier, J.M., and Alzari, P.M. (2000) Crystal structure of an
anti-carbohydrate antibody directed against Vibrio cholerae O1 in
complex with antigen: molecular basis for serotype specificity. Proc. Natl
Acad. Sci. USA, 97, 8433–8438.
Wu, A.M., and Mackenzie, N.E. (1987) Structural and immunochemical
characterization of the O-haptens of Brucella abortus lipopolysaccharides
from strains 19 and 2308. Mol. Cell. Biochem., 75, 103–111.
661