 |
INTRODUCTION |
Fucose-containing glycans are important in a number of biological
processes, including their roles as a component of the ligands in
mediating leukocyte adhesion to selectins. Mammals possess a number of
fucose-containing structures, including those with Fuc
1-3GlcNAc
-R, Fuc1-6GlcNAc-R, and Fuc1-2Gal-R
(1, 2). These linkages are generated by a variety of
-fucosyltransferases (FTs)1 derived from a
surprisingly high number of genes. Humans have at least six different
1,3FT genes (III, IV, V, VI, VII, and IX), two 1,2FT genes
(FUT1 (H) and FUT2 (Se) and
an 1,2FT pseudogene (SECI)), and at least one 1,6FT
gene (3-5). The functions of all of these genes and their cognate
fucose-containing glycans are not well understood, but the expression
of many of these genes is developmentally regulated and is altered
during embryonic development, differentiation, and tumorigenesis.
Although genetic approaches in mice have revealed many important
insights into the functions of many different genes, particularly in
regard to the 1,3FTs and their role in generating cell adhesion
ligands (5, 6), genetic approaches to study glycoconjugate roles in
development have been frustrated by the complexity of the large
multigene families for most of the glycosyltransferases and consequent
enzyme redundancy. This situation has led to the consideration of model systems that may accelerate our understanding of the overall biological functions of fucose-containing glycoconjugates.
Caenorhabditis elegans is a highly attractive model system
in which to study the potential developmental roles of glycoconjugates, since the worm shares many fundamental biological and biochemical pathways with higher organisms. The potential of using C. elegans to investigate the roles of glycoconjugates in development
requires the identification and characterization of specific
glycosyltransferases and their cognate structures, which may be
important in developmental processes. During the past few years, many
different enzymes involved in glycoconjugate biosynthesis have been
identified by homology in C. elegans (1). In our previous
studies, we identified an 1,3FT and an 1,2FT activity in C. elegans extracts (7). However, further analysis of the acceptor
specificity demonstrates that C. elegans has a novel
-1,2FT activity toward the unusual acceptors Gal1-4Xyl-R and
Gal1-6GlcNAc-R to synthesize the Fuc1-2Gal1-R products; such
acceptor specificity would be unique among known 1,2FTs. However,
recent studies on the O-glycans of C. elegans glycoproteins document the occurrence of such Fuc1-2Gal1-R
linkages (8). Searches of the C. elegans genome revealed
that dozens of putative fucosyltransferase genes are present (3),
and the C. elegans genome contains at least 22 different
genes encoding putative 1,2FTs with some identity to mammalian
1,2FTs.
In this paper, we report our identification of the C. elegans gene (termed CE2FT-1) encoding the 1,2FT
(CE2FT-1) capable of fucosylating the disaccharide acceptors
Gal1-4Xyl-R and Gal1-6GlcNAc-R but unable to fucosylate
lactose. Preliminary studies on other members of this putative 1,2FT
family in C. elegans suggest that they differ from CE2FT-1
in acceptor specificity. The expression pattern of CE2FT-1
is also unusual and is limited strictly to the 20 intestinal cells in
the adult worm. These unexpected findings suggest that each of the FTs
in C. elegans may have a unique enzyme activity and role in development.
| |
EXPERIMENTAL PROCEDURES |
Materials--
Sodium cacodylate, MnCl2, ATP,
L-fucose, phenyl--D-Gal, GlcNAc-pNP,
GalNAc-O-pNP, and Xyl-O-pNP were purchased
from Sigma. 1,2-Fucosidase and 1,3-fucosidase were purchased from
Prozyme, Inc. (San Leandro, CA). 1,6-Fucosidase and GDP-fucose were
purchased from Calbiochem. Gal-O-pNP,
Gal-O-pNP, and Glc-Mu (methylumbelliferon) were
obtained from NBS Biologicals Ltd. (Huntingdon, UK).
Gal1-4Glc-O-pNP, Gal1-4GlcNAc-Bz,
Gal1-4GlcNAc1-2Man-pNP, Gal1-3(Fuc1-4)GlcNAc, and
Gal1-3GlcNAc-O-pNP, and
Gal1-3GalNAc-O-pNP were purchased from Toronto
Research Chemicals Inc. (Downsview, Canada).
Gal1-3GlcNAc1-3Gal1-4Glc was from Oxford GlycoSystems.
Gal1-4GlcNAc1-3Gal1-4GlcNAc was from Dr. A. Veyrieres
(Universite Paris Sud, Orsay, France). Gal1-3Gal1-4Glc was from
Dr. M. Messer (University of Sydney).
Gal1-3Gal1-4Xyl-O-Bzl and
Gal1-4Xyl-O-Bzl were from Dr. T. Norberg (Swedish
University of Agricultural Sciences, Uppsala, Sweden). LNFII/III
(mixture of Gal1-3(Fuc1-4)GlcNAc1-3Gal1-4Glc (LNFII)
and Gal1-4(Fuc1-3)GlcNAc1-3Gal1-4Glc (LNFIII)) and
Gal1-6GlcNAc were from Dr. D. van den Eijnden and C. Koeleman
(Department of Medical Chemistry, VU University Medical Center,
Amsterdam, The Netherlands). GDP-[3H]Fuc and
GDP-[14C]Fuc (PerkinElmer Life Sciences) were
diluted with unlabeled GDP-Fuc to give the desired specific activity.
TaqDNA polymerase and other PCR components were obtained
form Roche Molecular Biochemicals. Oligonucleotide primers were
synthesized by Invitrogen. The eukaryotic TA cloning kit
(bidirectional) and pcDNA4/HisMax vectors were obtained from
Invitrogen. Restriction enzymes were purchased from New England
Biolabs, Inc. (Beverly, MA).
Cloning and Expression of CE2FT-1--
The oligonucleotides
5'-GTATGAGAAATGTGAAAGGACTTTTTAGC-3' and
5'-CATCTATTTTTGAAATTATCGACC-3' were used as forward and reverse primers, respectively, to amplify the CE2FT-1 coding
sequence from a C. elegans cDNA library in the 
ZAP
vector (provided by Dr. Robert Barstead, Oklahoma Medical Research
Foundation, Oklahoma City, OK). Reaction mixtures for amplification by
TaqDNA polymerase contained a 1 µM
concentration of each primer, 200 µM dNTPs, 2.5 mM MgCl2, and 1 µl of 109
plaque-forming units/ml C. elegans adult cDNA library in
a final volume of 100 µl. Amplification was carried out by an initial denaturing step of 94 °C for 5 min, followed by 35 cycles of
94 °C for 1 min, 59 °C for 1 min, and 72 °C for 2 min. These
cycles were then followed by an extension period of 72 °C for 7 min. Following amplification, the PCR product was run on an agarose gel, and
the ~1060-bp DNA band was excised from agarose gel and purified
by a QIAEX II agarose gel extraction kit. The purified DNA
was ligated into the pCR3.1 vector and subsequently transformed into
One Shot Top10F' competent cells according to the procedures provided
with the bidirectional eukaryotic TA cloning kit. Clones containing
inserts were selected and amplified, and minipreps were prepared using
Qiagen kits (Qiagen, Inc., San Clarita, CA) and digested with
ApaI restriction enzyme to check the direction of the
inserts. The clones with right direction insert were sequenced by the
Oklahoma Medical Research Foundation Facility (Oklahoma City, OK).
The clone with full-length cDNA of CE2FT-1 was then
digested with BamHI and XhoI, subcloned in frame
into pcDNA4/HisMax C vector, and subsequently transformed into One
Shot Top10F' competent cells. The clones containing right inserts were
selected and amplified, and maxipreps were prepared using the Qiagen
endotoxin-free plasmid DNA purification kit.
To determine whether the cloned insert encoded an active 1,2FT,
human 293T cells were transfected with 4.5 µg of plasmid DNA
harboring CE2FT-1 cDNA (in the plasmid pcDNA4/HisMax
C) using FuGene 6 transfection reagent, as described by the
manufacturer (Roche Molecular Biochemicals). Mock-transfection
was carried out by transfecting 293T cells with pcDNA4/HisMax C
vector. Seventy-two hours after transfection, cells were harvested,
solubilized in 50 mM sodium cacodylate buffer containing
1% Triton X-100, and assayed for fucosyltransferase activity as
described below.
Construction of an Expression Vector Encoding C-terminal HPC4
Epitope-tagged CE2FT-1--
A mammalian expression vector of
pcDNA4/HisMax (Invitrogen) encoding C-terminal HPC4 epitope-tagged
CE2FT-1 was constructed using PCR for introducing the 12-amino acid
Ca2+-dependent HPC4 epitope (9, 10) into the
cDNA. The forward primer was T7
(5'-TAATACGACTCACTATAGGG-3'). The reverse primer-containing sequence, encoding 12 amino acids of the HPC4 epitope (EDQVDPRLIDGK) immediately following the C- terminal lysine of CE2FT-1, was
5'-CGGAATTCTACTTGCCGTCGATCAGCCTGGGGTCCACCTGG-TCCTCTTTTGAAATTATCGACCCGTTTCGTGAC-3'. The PCR was performed by denaturation at 94 °C for 2 min,
amplification for 30 cycles at 94 °C for 1 min, 56 °C for 1 min
and 72 °C for 2 min using plasmid pcDNA4/HisMaxC/CE2FT-1
cDNA as the template. The expected 1420 bp of PCR product was
purified on 1.2% TAE-agarose gel and digested by EcoRI and
ApaI. The expected 1000-bp DNA fragment was purified and
cloned into ApaI (partially digested)/EcoRI
sites of pcDNA4/HisMaxC/CE2FT-1cDNA plasmid, and its
sequence was confirmed. Human 293T cells were transiently
transfected with 4.5 µg of plasmid DNA harboring CE2FT-1
cDNA with the HPC4 epitope tag fused to the C terminus (in the
plasmid pcDNA4/HisMax C) using FuGene 6 transfection reagent.
Seventy-two hours after transfection, cells were harvested and
solubilized in 50 mM sodium cacodylate buffer containing
1% Triton X-100, and the expressed CE2FT-1 was purified on immobilized
HPC4 monoclonal antibody from the cell extracts. Following
purification, the epitope-tagged CE2FT-1 was analyzed by
Western blotting using the HPC4 antibody and was assayed for fucosyltransferase activity as described below.
Fucosyltransferase Assay--
Standard fucosyltransferase assays
were performed in a 50-µl reaction mixture containing 5 µmol of
sodium cacodylate (pH 7.2), 1 µmol of MnCl2, 0.2 µmol
of ATP, 5 nmol of GDP-[14C]Fuc (3.6 Ci/mol), 0.2% (v/v)
Triton X-100, and 15 µl of cell extract. Acceptor substrate
concentrations were as indicated. After incubation for 120 min at
37 °C, the reaction was stopped, and the product was separated from
unincorporated label by chromatography on a 1-ml column of Dowex 1-X8
(Cl
form) as described (11, 12); alternatively, for
substrates with a hydrophobic aglycon, the products were isolated
by adsorption to Sep-Pak C-18 cartridges (Waters; Milford,
MA), washed with water, and eluted with absolute methanol, as described
(13). Control assays lacking the acceptor substrate were carried out to
correct for incorporation into endogenous acceptors, and assays with
mock-transfected cells were conducted to correct for endogenous fucosyltransferase activity.
C. elegans Culture and Preparation of Extracts--
The standard
laboratory wild type strain N2 worms were grown on NGM
plates seeded with OP50 bacteria to obtain lots of mixed stage worms.
Worms were washed with M9 buffer and were frozen in liquid nitrogen for
2 min, and the worm pellets were stored at 
80 °C. For C. elegans extracts, frozen worms were ground and suspended in 50 mM sodium cacodylate buffer, pH 7.0, containing 1 tablet/10
ml complete protease inhibitor mixture (Roche Molecular Biochemicals),
1% Triton X-100 and then subjected to sonication (three pulses of
10 s each) on a cell disruptor (Branson Sonic Power Co.). The
homogenate was incubated on ice for 30 min to allow solubilization of
proteins and then centrifuged at 14,000 rpm at 4 °C for 10 min. The
supernatant fractions were collected, and they were either used
directly or stored as aliquots at 80 °C. Frozen extracts were
thawed only once for use in enzyme assays.
Characterization of Fucosyltransferase Assay
Products--
Products obtained with the acceptor Gal--phenyl,
GDP-[3H]Fuc, and 293T cell extracts expressing
recombinant CE2FT-1 were isolated by chromatography on Sep-Pak C-18
cartridges (Waters). The isolated, radiolabeled products were incubated
with 0.8 milliunits of 1-2-fucosidase, 1-3,4-fucosidase,
and 1,6-fucosidase in 20 µl of reaction buffer 5 (Prozyme, Inc.)
at 37 °C for 48 h. Following treatment, the samples were
analyzed by descending paper chromatography in a solvent system of
ethyl acetate/pyridine/acetic acid/water (5:5:1:3). Products obtained
upon incubation of C. elegans extracts with the acceptors
Gal1-3Gal1-4Xyl-O-Bzl,
Gal1-4Xyl-O-Bzl, Gal1-3GalNAc-O-pNP, and the donor GDP-[3H]Fuc were isolated by chromatography
on Sep-Pak C-18 cartridges. The isolated, radiolabeled products were
incubated with 0.8 milliunits of 1-2-fucosidase,
1-3,4-fucosidase, 1,6-fucosidase, and water (mock treatment) in
50 µl of reaction buffer 5 (Prozyme) at 37 °C for 48 h,
respectively. After treatment, the samples were isolated by adsorption
to Sep-Pak C-18 cartridges (Waters), washed with water, and eluted with
absolute methanol, and the radioactivity in the eluted material was
determined by scintillation counting. The percentage of
[3H]fucose released was calculated based on the mock treatment.
Preparation of DNA Constructs for Promoter Analysis--
We
prepared fusion constructs of the upstream promoter sequences of
CE2FT-1 fused to the green fluorescent protein to examine the spatial pattern of CE2FT-1 expression during C. elegans
development. A 700-bp fragment of genomic sequence containing the
putative promoter region immediately upstream of the initiation site
for CE2FT-1 translation was obtained by PCR using C. elegans
genomic DNA as template with the primers
5'-AAAAAATTCATTATATAACATTTGTTTTCAG-3', and 5'-ACTTAAAAAAAATACCGGAAC-3'.
The PCR product was run on an agarose gel, and an ~700-bp DNA band
was excised from agarose gel and purified by the Qiagen agarose gel
extraction kit. The purified DNA was ligated into pCR3.1 vector and
subsequently transformed into Top10F' competent cells. Clones
containing inserts were selected and amplified, and minipreps were
prepared using the Qiagen kit and then sequenced. The clone containing
the right insert was digested with HindIII/XbaI
and subcloned into pPD95.67 vector (originally provided by Dr. Andrew
Z. Fire (Carnegie Institute of Washington, Baltimore, MD)), creating
plasmid pPD95.67/CE2FT-1-prom.
Preparation of Transgenic Worms and Promoter Analysis--
DNA
injection into the C. elegans germ line was carried out as
described (14, 15). Transgenic lines were established from F2
descendants of animals injected with 10 ng/µl
CE2FT-1::GFP constructs
(pPD95.67/CE2FT-1-prom), and pBX, which carries the PHA-1 gene, which is involved in the development of the pharynx, served as a transformation marker.
Preparing Synchronous Culture of L1 and
L2-L4 Young Adult Worm--
N2
worms were grown on NGM plates seeded with OP50 bacteria to obtain lots
of gravid adults. These were incubated with 5 ml of alkaline bleach
(0.25 M KOH, 25% commercial Clorox) at room temperature
for 3 min with occasional gentle agitation. This procedure kills all
adults and larvae but leaves the eggs alive. The eggs were washed with
M9 buffer and centrifuged for 1 min in a clinical centrifuge at 1800 rpm. The supernatant was removed, and the washing was repeated two
times. The washed eggs were distributed to one unseeded plate, which
was incubated at 20 °C overnight. This allows the eggs to hatch and
arrests the animals as starved L1, which produces synchrony
of the population. Growth was restarted by distributing these starved
L1 onto plates seeded with an adequate amount of food to
allow the animals to grow without starving. At times during growth,
worms were harvested to generate populations of
L2-L4 and young adult worms.
RT-PCR and Quantitative Real Time RT-PCR Analysis of C. elegans
CE2FT-1 mRNA during Development--
Total RNA was extracted from
staged synchronous populations using a total RNA isolation kit (Ambion,
Austin, TX). Random primed first strand cDNA was synthesized from 2 µg of total C. elegans RNA using the SuperScript
preamplification system for first strand cDNA synthesis kit
(Invitrogen). The first strand cDNA was used as a template in PCRs
to amplify cDNAs encoding CE2FT-1 transcripts using the
oligonucleotides 5'-GTATGAGAAATGTGAAAGGACTTTTTAGC-3' and
5'-CATCATTTTTGAAATTATCGACC-3' in the reactions containing 3 µl of
first strand cDNA in a volume of 50 µl. For controls, a C. elegans actin gene-specific primer pair
5'-GACAATGGATCCGGAATGTGC-3' and 5'-ATGATGGAGTTGTAAGAAGTCTC-3' was used
to amplify the C. elegans actin gene in the identical
conditions described. Following PCR amplification, 20 µl of each PCR
product was run on a 1.5% agarose gel containing 10 µg/ml ethidium
bromide. Quantitative RT-PCR was performed using an ABI Prism 7700 sequence detection (PerkinElmer Life Sciences) using the
double-stranded DNA binding dye, SYBR Green. Random primed first strand
cDNA was synthesized in a 100-µl reaction from 2 µg of
the total C. elegans RNA using TaqMan Reverse Transcription Reagents (PerkinElmer Life Sciences). The first strand
cDNA was used as a template in PCRs to amplify cDNAs encoding CE2FT-1 transcripts using gene-specific primers
5'-GTATGAGAAATGTGAAAGGACTTTTTAGC-3' and
5'-CCCATCCTATTAACAATGAAAATTATGA-3', designed by using Primer Express
software, in the reactions containing 5 µl of first strand cDNA
and 25 µl of 2× SYBR Green Master Mix (PerkinElmer Life
Sciences) in a volume of 50 µl. For controls, a C. elegans actin gene-specific primer pair 5'-ATC GTC CTC GAC TCT GGA
GAT G-3' and 5'-TCA CGT CCA GCC AAG TCA AG-3' was used to amplify the
C. elegans actin gene in the identical conditions
described. Cycling parameters were 95 °C for 10 min, followed
by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. To
confirm the absence of nonspecific amplification, The PCR products were
analyzed by a 1.2% agarose gel containing 10 µg/ml ethidium bromide.
| |
RESULTS |
Identification of a Fucosyltransferase Gene in C. elegans--
Our
previous studies indicated that extracts of C. elegans
contain an unusual 1,3FT activity and in addition contained an unidentified 1,2FT activity (7). To gain a better understanding of
the types of fucosyltransferases present in C. elegans, we used a panel of acceptors to assay extracts of adult worms for possible
fucosyltransferase activities with radioactive GDP-Fuc as the donor
(Table I). The highest activity was found
toward the vertebrate type 1 O-glycan structure
Gal1-3GalNAc-O-pNP (Table I). However, high activity
was also detected toward the unusual disaccharide
Gal1-4Xyl-O-benzyl. This disaccharide represents part
of the core structure of Gal1-4Xyl-O-R found in
vertebrate proteoglycans. We considered that the enzyme capable of
fucosylating the acceptors Gal1-3GalNAc-O-pNP and
Gal1-4Xyl-O-benzyl in C. elegans extracts
might be an 1,2FT acting on the terminal nonreducing Gal residue
rather than the penultimate GalNAc or Xyl residues.
To identify C. elegans sequences with possible homology to
known 1,2FTs, we compared the amino acid sequences of human, rabbit, pig, and mouse 1,2FTs to identify potential conserved sequences among them. We then used these identified conserved domains in a BLAST
search of the C. elegans genome data base. One of the genes identified is harbored in C. elegans cosmid EGAP9
(GenBankTM accession number U80026), derived from
chromosome V, encoding a predicted cDNA of 1068 bp encoding the
hypothetical protein EGAP9.2. A map of the gene structure is shown in
Fig. 1A, and the overall gene
size is 1814 bp, including 700 bp of the 5' upstream sequence. The open
reading frame is encoded within 10 exons predicted to encode a protein
with 355 amino acids (Fig. 1B). The exons are small and
range in size from 66 to 161 bp (average size of ~107 nucleotides).
The introns are also small, ranging in size from 41 to 310 bp, and have
an average size of ~101 bp. The small sizes of the exons and introns
is consistent with previous observations about expressed genes in
C. elegans (16). The exon/intron junctions at all introns
follow the GT/AG rule, although it has been predicted that not all
splicing for short introns in C. elegans requires the AG
(17). There are three predicted N-glycosylation sites at
positions in the predicted luminal domain of the protein at Asn
residues 92, 311, and 349 within the N-glycosylation sequon Asn-X-Ser/Thr (where X does not represent Pro).
There are no predicted sites for O-glycosylation at
Ser and Thr residues, using the NetOGlyc 2.0 Prediction
Server (available on the World Wide Web at
www.cbs.dtu.dk/services/NetOGlyc/) for O-glycosylation
sites, although this program is predictive for mammalian
glycoproteins.
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 1.
Structure of the CE2FT-1
gene and the cDNA encoding CE2FT-1. A, the
exon/intron structure for CE2FT-1 is indicated, with the exons numbered
1-10 and the upper numbers indicating
the sizes in bp of the exons and the lower
numbers indicating the sizes in bp of the introns. Each
splice donor (D) and acceptor (A) site for the
introns is indicated. B, sequence of the cDNA encoding
CE2FT-1. The adenine residue of the putative initiation codon is
assigned as 1, whereas the amino acids encoded by the cDNA are
depicted by single-letter code. The solid line
below the amino acid sequence is the putative transmembrane
region, and the boxes denote N-glycosylation
sequons. The arrows at the 5'- and 3'-ends of the cDNA
indicate the sequences of the forward and reverse primers used for PCR
cloning of the cDNA.
|
|
The Kyte-Doolittle hydrophilicity plot (18) of the protein sequence
suggests the presence of a 16-amino acid transmembrane domain at the N
terminus (Fig. 2). Thus, the encoded
protein displays the typical hallmarks of most Golgi-localized
glycosyltransferases (i.e. type II membrane orientation with
a relatively small cytosolic N terminus and a large, extracellular,
C-terminal region) (1). The most conserved region predicted between the
CE2FT-1 and other 1,2FTs is in the C-terminal domain (Fig.
3), indicating that this region is likely
to contain the catalytic domain, as seen for other Golgi
glycosyltransferases. However, overall the predicted amino acid
sequence from the CE2FT-1 cDNA displays a relatively low identity
(5-10%) to the 1,2FT sequences in humans, rabbits, and mice at the
amino acid level (Fig. 3).
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 2.
Hydropathy plot of the 355-amino acid
polypeptide predicted by the CE2FT-1 cDNA sequence. The
predicted type 2 protein in the membrane with the C-terminal domain
luminal and the reaction catalyzed by CE2FT-1 are also shown.
|
|
View larger version (90K):
[in this window]
[in a new window]
|
Fig. 3.
Comparison of the protein sequences of
CE2FT-1 with vertebrate 1,2FTs.
Comparisons are shown for human, pig, mouse, rabbit, and C. elegans CE2FT-1. Amino acid identities among the mammalian
1,2FTs are indicated by dark shading, and
conserved substitutions are indicated by lighter
shading. Amino acid identities between CE2FT-1 and the
mammalian 1,2FTs are indicated in the CE2FT-1 sequence as
red boxes, whereas the conserved substitutions
between the CE2FT-1 and the mammalian 1,2FTs are indicated in
blue. The three motifs in the C-terminal regions of the
vertebrate 1,2FT gene family termed I, II, and III are indicated by
the solid line (3).
|
|
Cloning of the cDNA from C. elegans Encoding
CE2FT-1--
Gene-specific primers were designed to amplify the entire
coding sequence of the C. elegans CE2FT-1 from a C. elegans cDNA ZAP library, as described under
"Experimental Procedures." Following amplification, the PCR product
was TA-cloned into the vector pCR3.1 and then subcloned in frame into
pcDNA4/HisMax C mammalian expression vector. A clone was isolated,
and DNA sequencing confirmed the exon/intron boundaries predicted by
the data base.
Expression of C. elegans CE2FT-1 cDNA in the 293T
Cell--
Transfection of human 293T cells with cDNA encoding
CE2FT-1 results in significantly higher enzyme activity over
mock-transfected 293T cells by using Gal--p-nitrophenol
as the acceptor (Table I). The mock-transfected 293T cells have no
significant endogenous 1,2FT activity. To verify that the activity
is due exclusively to the expressed cDNA and not to alteration of
an endogenous, unknown enzyme, we used Western blot with Anti-Xpress
monoclonal antibody to probe the Xpress-tagged CE2FT-1 in cell extracts
from transfected 293T cells. Unexpectedly, the results indicated that the N-terminal Xpress tag fused to the N terminus of CE2FT-1 was cleaved, since little antigenic protein was detected. We then constructed an alternative full-length cDNA of CE2FT-1 with the Ca2+-dependent HPC4 epitope (9, 10) tag fused
to the C terminus. Human 293T cells were transiently transfected with
this construct, and the transfected cells were harvested 72 h
after transfection. The presence of the HPC4 epitope-tagged CE2FT-1
protein with a molecular mass of ~50 kDa was identified by Western
blot using the Ca2+-dependent HPC4 monoclonal
antibody (Fig. 4). The recombinant HPC4
epitope-tagged CE2FT-1 was purified by absorption on immobilized HPC4
(9, 10). Using GDP[3H]Fuc as the donor and Gal-phenyl
as the acceptor, the immunoabsorbed recombinant HPC4-tagged CE2FT1
generated 2467 cpm of product, whereas immunoabsorption from
mock-transfected cells generated a background of 58 cpm. These results
confirm that the protein encoded by the gene CE2FT-1 is an
active 1,2FT. To confirm the fucosyl linkages in the reaction
product generated by CE2FT-1, the [3H]Fuc-labeled product
was isolated and treated with either 1,3/4-fucosidase, 1,2-fucosidase, or 1,6-fucosidase, followed by descending paper chromatography. Treatment of the product with 1,2-fucosidase, but
neither the 1,3/4-fucosidase nor 1,6-fucosidase, caused the
quantitative release of [3H]fucose from the radiolabeled
product (Fig. 5A). These
results demonstrate that the product of CE2FT-1 contains Fuc in
1,2-linkage to the Gal residue and demonstrate that
CE2FT-1 encodes a functional 1,2FT.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 4.
Western blot of purified HPC4-tagged CE2FT-1
from crude transfected 293T cell extract with HPC4 monoclonal
antibody. Lane 1, purified HPC4
epitope-tagged CE2FT-1; lane 2, crude cell
extract; lane 3, column flow-through.
|
|
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5.
Characterization of the reaction product
generated by recombinant CE2FT-1 and C. elegans
extracts. A, the radiolabeled product, generated
by recombinant CE2FT-1 with the acceptor Gal--phenyl and
GDP-[3H]fucose donor, was isolated and subsequently
treated with either H2O (mock) or 1,2-, 1,3/4-, or
1,6-fucosidase. The treated material was analyzed by descending
paper chromatography for release of free [3H]fucose, as
indicated by the arrow. B, characterization of
the reaction product generated by C. elegans extract. The
radiolabeled products generated by a C. elegans extract upon
incubation with either Gal1-3Gal1-4Xyl-O-Bzl,
Gal1-4Xyl-O-Bzl, or Gal1-3GalNAc-O-pNP
as acceptor and GDP-[3H]fucose donor were isolated and
treated with either H2O (mock) or 1,2-, 1,3/4-, or
1,6-fucosidase. The treated material was reisolated by
adsorption to Sep-Pak C-18 cartridges (Waters), washed with water, and
eluted with absolute methanol, and the eluted radioactivity was
determined by scintillation counting. The percentage of
[3H]fucose released was calculated based on the mock
treatment.
|
|
Acceptor Specificity of Recombinant CE2FT-1--
To further
characterize the CE2FT-1 activity, a comprehensive analysis of the
acceptor specificity of CE2FT-1 was performed, using a large variety of
acceptors, many of which are known to be acceptors for previously
described 1,2FTs. CE2FT-1 transfers fucose to the monosaccharide
acceptor Gal--p-nitrophenol but not to
Gal--p-nitrophenol, indicating absolute specificity for terminal -linked Gal residues (Table I). However, unexpectedly CE2FT-1 is completely inactive in using
Gal1-3GalNAc-O-pNP as the acceptor and is also
inactive in using either lactose or Gal1-4Glc-O-pNP as acceptors. In addition, the enzyme is unable to transfer fucose to
terminal 1-4-linked galactosyl residues in Man-containing branched
complex-type N-glycans. The enzyme was also inactive with
the acceptors containing 1-3/4-linked fucose, such as LNFII/III and
Gal1-3(Fuc1-4)GlcNAc, indicating that the CE2FT-1 is unable to
transfer Fuc in 1,2 linkage to another fucose to generate Fuc1-2Fuc1-R, as seen for at least one other invertebrate
1,2FT (19). By contrast, the enzyme demonstrated a clear preference for the unusual acceptor Gal1-4Xyl-O-benzyl with
about one-third less activity demonstrated toward Gal1-6GlcNAc.
Importantly, neither of these acceptors has been reported to be an
acceptor for previously described 1,2FTs. CE2FT-1 showed some
activity toward two of the more complex acceptors with the sequences
Gal1-3Gal1-4Xyl-O-benzyl and
Gal1-3Gal1-4Glc (Table I). These two latter acceptors contain a
penultimate Gal residue rather than GlcNAc, as in two of the acceptors
that were inactive, the trisaccharide Gal1-4GlcNAc1-2Man-pNP and the tetrasaccharide Gal1-4GlcNAc1-3Gal1-4GlcNAc.
Overall, CE2FT-1 is inactive toward all acceptors with penultimate
GlcNAc or Glc residues (Table I). Thus, it is possible that the enzyme does not recognize acceptors with penultimate GlcNAc residues but
prefers those with terminal Gal and a penultimate residue that is
neither GlcNAc nor Glc. However, insufficient amounts of product were
generated to allow proof at this time as to which of the Gal residues
was modified by Fuc in reactions with CE2FT-1. Nevertheless, these
results demonstrate that CE2FT-1 is unique compared with known
1,2FTs in its acceptor specificity, since all previously described
1,2FTs are efficient in using lactose or virtually any other
acceptor with terminal 1-4- or 1-3-linked galactose residues
(20, 21). The acceptor specificity of CE2FT-1 suggests that the
presence of an unsubstituted C-6 hydroxyl group on the penultimate
sugar hinders enzyme recognition of the terminal Gal residue.
The activity of CE2FT-1 was compared with other endogenous potential
1,2FTs in extracts of adult C. elegans. These extracts contained a fucosyltransferase activity toward the same acceptors identified for CE2FT-1, but in addition, the extracts contained fucosyltransferase activity toward Gal1-4GlcNAc-benzyl and the O-glycan type acceptor Gal1-3GalNAc1-O-pNP
(Table I). It is likely, based on our previous studies (7), that
Gal1-4GlcNAc-benzyl is fucosylated by the 1,3FT CEFT-1 or
related 1,3FTs rather than an endogenous 1,3FT activity. These
results suggest that adult worms express an active form of both CE2FT-1
and potentially other fucosyltransferases active toward
Gal1-4GlcNAc-benzyl and Gal1-3GalNAc1-O-pNP.
Interestingly, the extracts did not contain significant
fucosyltransferase activities toward Gal1-4Glc-O-pNP. This disaccharide is an excellent acceptor for the known 1,2FTs found in vertebrates, which suggests that of the many potential 1,2FTs encoded by the C. elegans genome (3), none of
these enzymes efficiently utilize the common acceptor
Gal1-4Glc-O-pNP. To confirm that the
fucosyltransferase reaction products using these acceptors were
modified by endogenous -1,2FTs in the C. elegans
extracts, we isolated the products generated by the C. elegans extracts using three of the acceptors utilized by the recombinant CE2FT-1. The acceptors we chose were Gal1,3GalNAc-pNP, Gal1,4Xyl-Bzl, and Gal1,3Gal1,4Xyl-Bzl. These
3H-fucosylated products generated by the
endogenous -fucosyltransferases in C. elegans
extracts were treated with either -1,2-fucosidase, -1,3/4-fucosidase, or -1,6-fucosidase. [3H]Fucose
was quantitatively released from all of the products by
-1,2-fucosidase, and only minimal radioactivity was released by the
other -fucosidases (Fig. 5B). These results indicate that the predominant endogenous -fucosyltransferase in C. elegans extracts is an -1,2FT with the type of acceptor
specificity exhibited by recombinant CE2FT-1.
Expression of CE2FT-1 Promoters at Various Stages in C. elegans Development--
Very little is known about the functions of
glycosyltransferases in C. elegans (7, 22-26), and no
information is available about expression of potential 1,2FT genes.
To examine whether CE2FT-1 is expressed during development
and its pattern of expression, we first analyzed transcript levels by
relative RT-PCR using equivalent amounts of total RNA from different
developmental stages (Fig. 6A). The CE2FT-1
gene is expressed at all developmental stages with little relative
difference in expression compared with -actin. Since there are more
than 22 putative 1,2FTs in C. elegans, and some of them
with more than 85% homology to CE2FT-1, we performed quantitative
RT-PCR to analyze the mRNA expression level at the different
stages. The results indicated that there is little relative difference
in expression at different stages (see Fig. 6, B and C).
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 6.
The relative RT-PCR of CE2FT-1 mRNA from
C. elegans at different stages. A,
total C. elegans RNA from the populations of L1,
L2-L4, and adult stage were prepared and used
as templates in first strand cDNA synthesis. A 500-bp positive
control from the RT-PCR kit was used as a control for the RT-PCR method
(Line Positive Control). The
arrows indicate the expected size of -actin and CE2FT-1.
Size markers in bp are indicated. B, the quantitative
real-time RT-PCR of CE2FT-1 mRNA from C. elegans at
different stages. Total C. elegans RNA from the populations
of L1, L2-L4, and adult stage were
prepared and used as templates in first strand cDNA synthesis. A
106-bp cDNA was amplified. The expression level was indicated by
threshold cycles. C, confirmation of a single product of
amplification in quantitative real time RT-PCR was performed on 1.2%
agarose gel. A 106-bp cDNA fragment was amplified from CE2FT-1, and
a 100-bp cDNA fragment was amplified from actin. Size markers in bp
are indicated.
|
|
To localize the expression of the gene, we used reporter promoter
constructs driving expression of the coelenterate green fluorescent
protein (GFP) in the vector pPD95.67/CE2FT-1-prom. The promoter
construct consisted of a 5' 700-bp fragment containing all the upstream
genomic sequence, which should encompass the putative promoter region
immediately upstream of the initiation site for CE2FT-1 translation
(Fig. 1A). In rats, it has been shown that expression of the
1,2FT gene is regulated by functional promoter elements within
the 5'-flanking region of that gene (27). Likewise, in C. elegans, gene expression is usually regulated by promoter elements
in 5'-untranslated regions of genes (14). Expression of the GFP
reporter gene in the transgenic worms injected with CE2FT-1 promoter
construct pPD95.67/CE2FT-1-prom was examined (Fig.
7, A-H). The results indicate
that the promoter for CE2FT-1 is specifically activated in intestinal
cells beginning at the 2-fold embryonic stage embryo (Fig. 7,
A-D). Promoter function was identified through all later
developmental stages with a highly restricted GFP expression in the 20 cells of the adult intestine (Fig. 7, E-H). In these
studies, the 20 cells aligned to form the entire intestine were clearly
visually identified by transmitted light microscopy compared with the
fluorescent images (Fig. 7, E and F). There was a
surprising restriction of CE2FT-1 expression in the
intestinal cells, and all cells had equivalent expression levels based
on visual observations. Virtually no GFP fluorescence was observable in
nonintestinal cells in adult worms. In control experiments, we used 5'
upstream promoter regions of other potential 1,2FT genes identified
in our Blast search of the C. elegans genome. GFP promoter
analyses with these other constructs gave no restricted GFP expression
in the 20 cells of the intestine, although expression was observed in
other cells (data not shown).
View larger version (70K):
[in this window]
[in a new window]
|
Fig. 7.
Expression of
CE2FT-1::GFP in C. elegans. Shown are fluorescence-based microscopic
images of transgenic C. elegans expressing the
CE2FT-1::GFP reporter construct (A,
C, E, and G) and light microscopic
images of the corresponding C. elegans profile
(B, D, F, and H).
Expression of CE2FT-1::GFP reporter in late embryo
(A and B), egg (C and D),
larva (E and F), and adult (G and
H) is shown.
|
|
| |
DISCUSSION |
Glycoconjugates with the sequence Fuc1-2Gal-R, as in the
ABO(H) blood group antigens, are expressed in a wide variety of tissues
and commonly occur in the digestive mucosae of a large number of
species ranging from amphibians to mammals (1, 28). The synthesis of
the 1,2-fucosylated structures in humans is catalyzed by two
1,2FTs, encoded by FUT1 and FUT2 (also known as the Se gene), which differ slightly in acceptor
specificity (29-33). In addition, humans possess another allele termed
SEC1, which is a pseudogene (33). The orthologues for the
human FUT1, FUT2, and SEC1 genes have
been described in rodents and termed FTA, FTB,
and FTC, respectively (34). Little is known about the
functions of the 1,2-fucosyltransferases and their cognate structure
Fuc1-2Gal-R in animals.
In all species so far examined, FUT2 (Se) is
expressed in the intestine. The ileal epithelium of adult mice in
normal growth cages is characterized by expression of glycoconjugates
with the sequence Fuc1-2Gal-R; however, the intestines of germ-free
mice are deficient in such fucose-containing structures (35).
Inoculation of the germ-free mice with Bacteroides
thetaiotaomicron significantly stimulated expression of
Fuc1-2Gal-R structures and transcripts encoding an 1,2FT (35),
which could be derived from either SEC1 or FUT2
(36). It was recently shown in germ-free mice that administration of
intestinal microbes stimulated expression of MFUT-II, the ortholog of
human FUCT2 (Se) and rat FTB (37, 38). The bacteria-dependent commensal stimulation of host
fucosylation is a complex phenomenon involved in generating metabolic
fucose to sustain bacterial flora (39). We do not yet know whether expression of CE2FT-1 in the intestinal cells of C. elegans
is induced by bacterial factors, but since the worms are constitutively grown using E. coli as a nutrient source, there may be a
role for bacterial products in regulating gene expression.
Other functions of Fuc1-2Gal-R have also been proposed. H-type 1 oligosaccharides with the structure Fuc1-2Gal1-3GlcNAc-R are
differentially expressed during mouse embryogenesis and have been
proposed to be involved in early implantation events (40, 41). The
1,2FT activity varies during the estrous cycle and is elevated
~5-fold during estrous (42). However, implantation appears to be
normal in mice genetically deficient in both murine homologs of
FUT1 and FUT2 (43).
Several other genes are known to be specifically expressed in the
intestinal cells of C. elegans, such as the
N-acetylglucosaminyltransferase I gene Gly14 (24)
and the genes encoding cathepsin L (44) and cysteine protease (45).
There is some information about the regulatory factors driving
gut-specific expression of the cysteine protease gene cpr-1.
The cpr-1 gene has two upstream GATA-like motifs at 51
(GATAA) and 147 (GATAA) to the transcriptional start site (46, 47).
The results showed that the both elements are required for protein
expression (46). Similarly, it was reported that the gut-specific
expression of the Hemonchus contortus gene
AC-2 in C. elegans is dependent on a
reverse GATA element (TTATC) at position 48 to the translation start
site (47). We have inspected the upstream 5' sequence of
CE2FT-1 for the presence of these GATA-like motifs. There
are two such GATA-like motifs at 112 (GATAA) and at 125 (GATAA) 5'
upstream to the translational start site for CE2FT-1. Whether one or
both of these elements is important for expression of the
CE2FT-1 gene is not known at present.
We also inspected the 5'-untranslated sequences of some other C. elegans glycosyltransferase genes for the presence of GATA-like elements that might contribute to intestinal expression. The gene encoding CEFT-1, an 1,3FT that synthesizes Fuc1-3GlcNAc-R
linkages (7), lacks GATA elements within 500 of the translation
initiation site. Our recent expression studies on CEFT-1 expression
show that it is not expressed in intestinal cells of adult C. elegans but is expressed in specific neural
cells.2 However, an open
reading frame within cosmid EGAP9.3, which is in the same
locus as CE2FT-1, also encodes a putative 1,2FT and contains one
GATA-like element at 85 upstream of the translation initiation site,
suggesting that it may also be specifically expressed in intestinal
cells. Interestingly, the Gly14 gene encoding an N-acetylglucosaminyltransferase I (24) was shown to be
expressed specifically in intestinal cells. Our inspection of the
5'-untranslated region of the Gly14 gene reveals that it
contains two GATA-like motifs within 300 of the translation
initiation site. These comparative results suggest that the GATA-like
elements in the 5'-untranslated sequences of some glycosyltransferase
genes, such as CE2FT-1, might be important in regulating
their intestine-specific expression.
The amino acid sequence of CE2FT-1 displays a very low identity
(5-10%) to the sequences of previously described vertebrate 1,2FTs. Previous studies have identified three motifs in the C-terminal regions of the vertebrate 1,2FT gene family termed I, II,
and III (3) and in prokaryotic (Helicobacter pylori) 1,2FT (48), as indicated in Fig. 3. Motifs I and III are conserved in CE2FT-1; however, motif II is not well conserved. The C-terminal or
luminal domain of the protein is predicted to be Golgi-resident and
constitute the catalytic domain based on analogy to mammalian 1,2FTs
(49). The roles of these 1,2FT motifs are not known, but their
presence in the catalytic domain suggests that they may be involved in
interactions with either acceptor glycans or GDP-Fuc. Although purely
speculative, the weak conservation of motif II in CE2FT-1 might be
associated with its inability to transfer to lactose and other simple
Gal1-4Glc(NAc)-R acceptors as do the mammalian 1,2FTs.
Little is currently known about the roles of glycoconjugates in
C. elegans development. Two of the eight sqv
genes regulating vulval epithelial invagination have homologies to
glycosyltransferase genes, including sqv-3, homologous to
members of the 1,4-galactosyltransferase gene family, which may be
involved in elongation of the glycosaminoglycan core linkage to
Xyl-Ser, and sqv-8, homologous to two vertebrate 1,3-glucuronyltransferases (25, 26, 50-52). A third gene, sqv-7, was recently shown to encode a sugar nucleotide
transporter (53). All of these genes have been shown to be functionally involved in glycosaminoglycan biosynthesis in vivo (54).
The C. elegans genome contains at least 22 homologues of
1,2FT genes, at least four genes encoding 1,3/4FT, one 1,6FT
gene, and probably many more fucosyltransferase genes yet to be defined (3, 7, 55) (also see the Web site of the Consortium for Functional
Glycomics at
functionalglycomics.mit.edu/cgi-bin/functional_glycomics/glyt/glyt_index.cgi). For example, a cytosolic 1,2FT named Skp1 was recently identified in
Dictyostelium (56), which lacks significant homology to any previously identified fucosyltransferases. Skp1 is also
significantly different enzymatically and structurally from CE2FT-1
and other known mammalian 1,2FT, in that Skp1 requires divalent
cations and reducing conditions for activity, is cytosolic rather than compartmentalized, lacks a membrane anchor domain, and lacks several sequence motifs that are highly conserved in prokaryotic, microbial, and mammalian 1,2FTs (3).
The CE2FT-1 we have identified may be partly responsible for
synthesizing the unusual Fuc1-2Gal1-6Gal-R linkages recently reported in complex-type O-glycans in adult C. elegans glycoproteins (8). The 1,2-linked fucose occurred
in complex structures, such as
Fuc1-2(Gal1-6)Gal1-6(Gal1-3)(Fuc1-2)Gal1-3(Glc1-6)GalNAc1-4GlcNAc1-Ser/Thr, in which the Fuc residue was 2-O-methylated. CE2FT-1 acts
particularly well on 1-6-branched Gal residues, as in
Gal1-6GlcNAc (Table I), but not Gal1-4GlcNAc residues (Table
I). Interestingly, no Gal1-4GlcNAc residues were observed in
O-glycans from C. elegans (8). In addition, it is
possible that the predicted glycan Gal1-4Xyl1-Ser generated by
enzyme 1,4-galactosyltransferase encoded by the sqv3 gene
(54) could be an endogenous acceptor for CE2FT-1, since we found that
CE2FT-1 shows the highest acceptor activity toward
Gal1-4Xyl-O-benzyl. However, in one study, no 1,2-fucosylated glycosaminoglycan core structures were found in
studies on complex-type O-glycans of C. elegans
(8), but it is possible that such modifications may be restricted to
core structures in only the 20 gut cells of the organism. Further
studies will be required to define the exact role of CE2FT-1 and each of the other members of this large family of 1,2FTs in specific glycoconjugate synthesis and worm development.
1,2-Fucosyltransferase (CE2FT-1) from Caenorhabditis
elegans*
§,
TOP
ABSTRACT
INTRODUCTION
To whom correspondence should be addressed: Dept. of
Biochemistry and Molecular Biology, University of Oklahoma Health
Sciences Center, 975 N.E. 10th St., BRC417, Oklahoma City, OK 73104. Tel.: 405-271-2481; Fax: 405-271-3910; E-mail:
richard-cummings@ouhsc.edu.
