Báo cáo khóa học: Cloning and expression of murine enzymes involved in the salvage pathway of GDP-L-fucose ppt

Cloning and expression of murine enzymes involvedL-fucokinase and GDP-L-fucose pyrophosphorylase Jaana Niittyma¨ki1, Pirkko Mattila2, Christophe Roos2, Laura Huopaniemi1, Solveig Sjo¨blo

Cloning and expression of murine enzymes involved

L-fucokinase and GDP-L-fucose pyrophosphorylase

Jaana Niittyma¨ki1, Pirkko Mattila2, Christophe Roos2, Laura Huopaniemi1, Solveig Sjo¨blom1*

and Risto Renkonen1,3

1 Department of Bacteriology and Immunology, Haartman Institute and Biomedicum, University of Helsinki; 2 MediCel, Helsinki;

3 HUCH Laboratory Diagnostics, Helsinki University Central Hospital, Finland

In the salvage pathway of GDP-L-fucose, free cytosolic

fucose is phosphorylated by L-fucokinase to form L

-fu-cose-1-phosphate, which is then further converted to

GDP-L-fucose in the reaction catalyzed by GDP-L-fucose

pyrophosphorylase We report here the cloning and

expression of murine L-fucokinase and GDP-L-fucose

pyrophosphorylase Murine L-fucokinase is expressed as

two transcripts of 3057 and 3270 base pairs, encoding

proteins of 1019 and 1090 amino acids with predicted

molecular masses of 111 kDa and 120 kDa respectively

Only the longer splice variant ofL-fucokinase was

enzy-matically active when expressed in COS-7 cells Murine

GDP-L-fucose pyrophosphorylase has an open reading

frame of 1773 base pairs encoding a protein of 591 amino

acids with a predicted molecular mass of 65.5 kDa

GDP-L-fucose, the reaction product of GDP-L -pyrophosphory-lase, was identified by HPLC and MALDI-TOF MS analysis The tissue distribution of murine L-fucokinase and GDP-L-fucose pyrophosphorylase was investigated by quantitative real time PCR, which revealed high expres-sion of L-fucokinase and GDP-L-fucose pyrophosphory-lase in various tissues The wide expression of both enzymes can also be observed from the large amount of data collected froma number of expressed sequence tag libraries, which indicate that not only the de novo pathway alone, but also the salvage pathway, could have a signi-ficant role in the synthesis of GDP-L-fucose in the cytosol Keywords: GDP-L-fucose; L-fucokinase; GDP-L-fucose pyrophosphorylase; salvage pathway; molecular cloning

L-Fucose is an important monosaccharide in the complex

carbohydrates of mammals It decorates N- and O-linked

glycoproteins and glycolipids [1] or is covalently linked to

some serine or threonine residues of proteins [2] Various

functions have been established in biological processes for

fucose residues that are present in the terminal chains of

oligosaccharides of membrane bound or secreted molecules [3] Fucosylated glycans formABO and Lewis blood group antigens in humans [4,5] Glycans that contain a(1,3)-fucosylated modifications, e.g sialyl Lewis x-type glycans, have an important role in inflammation They initiate extravasation of leukocytes by mediating their tethering and rolling on the endotheliumby decorating the leukocyte and endothelial cell counter receptors for selectin family of cell adhesion molecules [6,7] Fucosylation also seems to play an important role in fertilization [8,9], development [10–13], tumor metastasis [14] and programmed cell death [15]

Fucosylation requires GDP-L-fucose as a donor of fucose and as a substrate for fucosyltransferases Two different cytosolic pathways lead to formation of GDP-L-fucose The constitutively active de novo pathway involves conversion of GDP-a-D-mannose to GDP-b-L-fucose by two enzymes, GDP-D-mannose-4,6-dehydratase (GMD) and GDP-4-keto-6-deoxy-D-mannose-3,5-epimerase-4-reductase (FX) [16,17] In the alternative biosynthetic pathway, i.e the

salvage metabolism, L-fucokinase synthesizes L -fucose-1-phosphate from L-fucose and ATP GDP-L-fucose pyrophosphorylase further catalyzes the formation of GDP-L-fucose from L-fucose-1-phosphate and GTP The salvage pathway utilizes fucose obtained fromextracellular sources or fromintracellular degradation of glycoproteins and glycolipids [Fig 1]

Correspondence to: R Renkonen, Department of Bacteriology and

Immunology, Haartman Institute and Biomedicum, PO Box 63,

FIN-00014 University of Helsinki, Helsinki, Finland.

Fax: + 358 9 1912 5155, Tel: + 358 9 1912 5111,

E-mail: Risto.Renkonen@Helsinki.Fi

Abbreviations: CDS, coding sequence; EST, expressed sequence tag;

FX, GDP-4-keto-6-deoxy- D -mannose-3,5-epimerase-4-reductase;

GMD, GDP- D -mannose-4,6-dehydratase; LADII, leukocyte

adhesion deficiency type 2.

Enzymes: GDP-mannose 4,6-dehydratase (EC 4.2.1.47);

GDP-4-keto-6-deoxy- D -mannose 3,5-epimerase/4-reductase (EC 1.1.1.187);

L -fucokinase (EC 2.7.1.52); GDP- L -fucose pyrophosphorylase

(EC 2.7.7.30).

Note: Nucleotide sequence data are available in the DDBJ/EMBL/

GenBank databases under the accession numbers AJ297482,

AJ534942 and AJ276067.

*Present address: Department of Biosciences, Division and Genetics,

University of Helsinki, Finland.

(Received 6 October 2003, accepted 30 October 2003)

L-fucokinase and GDP-L-fucose pyrophosphorylase were

first discovered in pig liver [18,19] To dateL-fucokinase has

been partially purified and characterized fromporcine liver

[19] and thyroid gland [20], and purified to apparent

homogeneity from pig kidney [21] Furthermore, the gene

encoding human fucokinase has been identified [22]

GDP-L-fucose pyrophosphorylase has been purified fromporcine

kidney and the corresponding gene has been cloned from

human [23]

In the present study we have cloned the murine genes

coding for the enzymes involved in the salvage pathway

of GDP-L-fucose L-fucokinase and GDP-L-fucose

pyro-phosphorylase were expressed in COS-7 cells, and the

enzyme activities were determined

Experimental procedures

Cloning of mouseL-fucokinase

Based on the sequence fromthree published pig

fuco-kinase peptides [21], a portion of the mouse fucofuco-kinase

sequence was resolved through tBlastn and

FastA-searches of the EMBL/GenBank/DDBJ sequence

data-base [24,25] Using expressed sequence tags (ESTs),

primers corresponding to putative fucokinase sequence

were designed A region of the mouse fucokinase

sequence was amplified by PCR from the cDNA of

mouse kidney (QuickClone cDNA, Clontech, Palo Alto,

CA, USA), cloned into blunt II-TOPO vector

(Invitro-gen, Carlsbad, CA, USA) and sequenced This sequence

was used as a query tool for further sequence database searches and sequence alignments The IMAGE clone

4190449 (accession number BF538673) was obtained fromMRC geneservice (Cambridge, UK) and sequenced This clone, which was identified as containing a putative fucokinase, contained the full coding sequence (CDS) of

L-fucokinase RT-PCR was performed to confirm the relevance of the IMAGE clone sequence Mouse kidney total RNA (Ambion, Inc., Austin, TX, USA) was used

as a template in the first strand cDNA synthesis (Superscript First Strand synthesis systemfor RT-PCR, Invitrogen) Primers for RT-PCR were designed accord-ing to the sequence data gained fromIMAGE clone

4190449 The gene-specific primer for in vitro reverse transcription, was 5¢-TAGCAGCAGACTTGAAGAGG TA-3¢ PCR was performed by using the forward primer 5¢-GCCAGAATGGAGCAGTCAGAGGGAGTC-3¢ and the reverse primer 5¢-GCAGCTCTAGGTGGTGCCCA CTTCAGAG-3¢ The PCR products were cloned into pCR-XL-TOPO vector (Invitrogen) and sequenced Two clones were identified displaying two putative splice variants [Fig 2]

Expression of fucokinase cDNAs The two splice variants of fucokinase were subcloned into the XbaI site of a pQM vector containing a C-terminal E2-Tag/A (Quattromed Ltd, Tarto, Estonia) The forward

G-3¢ and the reverse primer was 5¢-ATCTCTAGAGGT GGTGCCCACTTC-3¢ All primers contained the XbaI restriction enzyme recognition site (underlined in the oligonucleotide sequences) Long and short splice variants

of fucokinase were transiently transfected into COS-7 cells

by lipofectamine 2000 (Invitrogen) according to the manu-facturer’s instructions After 48 h, the transfected cells were lysed in 50 lL of 50 mMTris/HCl (pH 7.8), 150 mMNaCl, 1% (v/v) Triton X-100, and incubated on ice for 1 h with a protease inhibitor cocktail (BD, Erembodegem, Belgium) Protein concentrations were determined using bicinchoninic acid protein reagent (Pierce Chemical Co., Rockford, IL, USA)

Fig 1 Synthesis of GDP- L -fucose in mammals The constitutively

active de novo pathway converts GDP- D -mannose into GDP- L -fucose

via oxidation, epimerization and reduction catalyzed by two enzymes,

GMD and FX In the alternative salvage pathway, free fucose is

delivered to cytosol fromextracellular sources (shown) or

fromlyso-somal degradation of glycoconjugates (not shown) L -fucose is

phos-phorylated by L -fucokinase to form L -fucose-1-phosphate, which is

converted to GDP- L -fucose in the reaction catalyzed by GDP- L -fucose

pyrophosphorylase GDP- L -fucose is then transported into the Golgi.

Fig 2 Gene structures of the short and the long splice variants of mouse fucokinase (A) and human fucokinase (B) The long splice variant of mouse fucokinase contains exons 1–20, 21a, 22, 23a and 24 The short splice variant contains exons 1–20, 21b, 23b and 24 The human fucokinase has a similar gene structure to the long splice variant of mouse fucokinase at the 3¢ end.

COS-7 cell lysates (30 lg) were detected by Western

blotting using anti-(E2-Tag) primary mAb (Quattromed)

and (mouse IgG) HRP-conjugated secondary

anti-body Detection was performed using enhanced

chemilu-minescence (Amersham Biosciences, Bucks, UK) according

to standard methods

Fucokinase activity assay

Cell lysate (100 lg) was assayed in a 100 lL reaction

mixture containing 50 mM Tris/HCl (pH 8.0), 5 mM

MgSO4, 150 000 c.p.m L-[3H]fucose (specific activity

63.0 CiÆmmol)1, Amersham), 0.1 mM fucose (Sigma,

St Louis, MO, USA), 5 mMATP and 5 mMNaF (Sigma)

in final concentrations The reaction mixture was incubated

at 37C for 30 min and terminated with 100 lL of ethanol

The incubation mixture was applied to two 10 cm

DEA-Bond Elut column (Varian, Palo Alto, CA, USA), which

was then washed with four column volumes of 10 mM

NH4HCO3 to remove the unbound material The [3

H]fu-cose-1-P was eluted with 2 mL of 250 mMNH4HCO3 The

eluate (400 lL) was counted with liquid scintillation and

luminescence counter (Wallac Trilux, Turku, Finland)

Cloning of GDP-L-fucose pyrophosphorylase

The 3¢ end of the pyrophosphorylase gene was cloned from

the mouse kidney UNI-ZAP XR lambda cDNA library

(Stratagene, La Jolla, CA, USA) by screening of

approxi-mately 1· 106 recombinant plasmids The published

human GDP-L-fucose pyrophosphorylase (accession

num-ber AF017445) [23] was used in a BLAST search to locate

mouse ESTs corresponding to the putative

pyrophospho-rylase According to the EST sequence (AA422658), the

forward primer 5¢-GAGTATTCTAGATTGGGGCCT

GA-3¢ and reverse primer 5¢-TGTGGACTGCACGCA

TTTTCC-3¢ were designed PCR was performed using

mouse liver cDNA (QuickClone cDNA, Clontech) as a

template The 330 bp PCR product was labelled with

[32P]dCTP[aP] using the Multiprime DNA labelling kit

(Amersham Biosciences, Buckinghamshire, UK) according

to the manufacturer’s protocol The labelled probe was used

in colony hybridization according to standard methods

The entire 5¢ end was resolved by 5¢ RACE-PCR (Robust

RT-PCR kit, Finnzymes, Espoo, Finland) using mouse

kidney mRNA (Clontech) as a template PCR was

performed using the 5¢ RACE synthesis primer AP1

(Clontech), 5¢-CCATCCTAATACGACTCACTATAGG

GC-3¢ and the gene-specific reverse primer, 5¢-GACTCC

AGGCCTCATGTTTGAGGGGAAATCCACGTAC-3¢

The second round PCR was performed with a nested

adaptor primer AP2 (Clontech), 5¢-ACTCACTATAGG

GCTCGAGCGGC-3¢ together with a nested gene-specific

primer, 5¢-CAAACACTCAAGGGAACAAAG-3¢ All

PCR products were cloned into pCR-Blunt II-TOPO

vector (Invitrogen)

The enzymatic activity of GDP-fucose pyrophosphorylase

The coding sequence of pyrophosphorylase was amplified

by PCR using the forward primer, 5¢-AATGGTACC

ATGGCGTCTCTCCGCGA-3¢ and the reverse

pri-mer, 5¢-CACGGATCCTTAAGATTTCTCTAAATCAG-3¢ creating KpnI and BamHI restriction enzyme recognition sites (underlined), respectively Subcloning of the PCR product into a pCDNA3.1(+) vector (Invitrogen) and the transient transfection into COS-7 cells were per-formed as above The cells were lysed on ice with 50 mM

Tris/HCl (pH 7.5) including protease inhibitor cocktail (Pharmingen), with sonication for 3· 15 s (Branson Sonifier 450, Heinemann, Schwa¨bich Gmund, Austria) Cell lysates (60 lg) were incubated in a 50 lL reaction mixture containing 0.5 M Tris/HCl (pH 7.8), 200 mM

MgCl2, 10 mMb-L-Fuc-1P (Sigma), 100 mMGTP, 0.5 U inorganic pyrophosphorylase (Sigma) at 37C for

30 min

Nucleotide sugars were purified fromthe cell lysates as described by Rabina et al [26] and analyzed by ion-pair reversed-phase HPLC on a Discovery HS C18 column (0.46· 25 cm; Supelco Inc., Pennsylvania, PE, USA) at a flow rate of 1 mLÆmin)1 A linear gradient of 0–1.5% (v/v) acetonitrile in 20 mM triethyammonium acetate buffer (pH 7.0) over 35 min was used and the effluent was monitored at 254 nm The amount of synthesized

GDP-L-fucose was calculated using the peak areas of external nucleotide sugar standards GDP-D-mannose,

GDP-D-rhamnose [27] and GDP-L-fucose (Calbiochem, San Diego, CA, USA) The fraction containing the putative GDP-L-fucose was collected fromthe HPLC assay for further analysis with MALDI-TOF MS

MALDI-TOF-MS MALDI-TOF MS was performed with a Biflex III mass spectrometer (Bruker Daltonics, Bremen, Germany) Nuc-leotide sugars were investigated in a 2,4,6-trihydroxy-acetonephenone–acetonitrile–aqueous ammonium citrate matrix as described previously [26], utilizing the reflector negative-ion mode with delayed extraction External calib-ration was performed with TDP-D-rhamnose (a generous gift fromP Messner, Universita¨t fu¨r Bodenkultur Wien, Wien, Austria) and UDP-GlcNAc (Sigma) GDP-L-fucose (Calbiochem) was used as a positive control

Reverse transcription and quantitative real time PCR Fucokinase and pyrophosphorylase mRNA expression in different tissues was detected by quantitative real time PCR Ambion’s Mouse Total RNA (kidney, liver, brain, ovary, testicle, heart, lung, spleen) was used for the first strand cDNA synthesis For each tissue, 1 lg of total RNA was reverse transcribed with randomhexamers using the Invi-trogen SuperScript cDNA synthesis kit according to the manufacturer’s instructions Parallel reactions in the absence

of SuperScript II (–RT controls) were performed to assess the degree of contaminating genomic DNA The resulting cDNA samples were subjected to real time quantitative PCR assay [28] to detect the expression levels of pyrophosphory-lase and long and short splice variants of fucokinase Primers and probes were designed using thePRIMER EXPRESSprogram (Version 1, PE Applied Biosystems, Foster City, CA, USA),

a software tool provided with the ABI 7000 Sequence Detection System(PE Applied Biosystems) Forward and reverse primers were positioned as close as possible to each

other without overlapping the probe Probes were

synthes-ized incorporating the fluorescent reporter FAM

(6-carboxy-fluorescein) at the 5¢ end and the quencher TAMRA

(6-carboxy-tetramethyl-rhodamine) at the 3¢ end

One microlitre of freshly synthesized cDNA was

ampli-fied in a total volume of 25 lL containing 1· Universal

Master Mix (PE Applied Biosystems) on an ABI Prism 7000

Sequence Detection System Assays for each transcript were

carried out as duplicates on the same plate and real time

PCR amplification was repeated twice Any inefficiencies in

RNA input or reverse transcription were corrected by

normalization to a housekeeping gene (18S rRNA Control

Reagents, PE Applied Biosystems) Primer concentrations

used were 900 nM/300 nM (forward/reverse) for the long

splice variant of fucokinase, 900 nM/900 nM (forward/

reverse) for the short splice variant of fucokinase, and

300 nM/900 nM (forward/reverse) for pyrophosphorylase

The concentration of the double labelled probe was 200 nM

for the long variant of fucokinase and pyrophosphorylase,

and 300 nMfor the short fucokinase splice variant Relative

amounts of the three mRNAs analyzed were based on

standard curves (Applied Biosystems User Bulletin 2)

prepared by a serial dilution of control cDNA

Results

Cloning of putative mouseL-fucokinase and sequence

analysis

Using the three known pig fucokinase peptides [21] as

probes, part of the putative murine fucokinase sequence was

identified from mouse genomic sequence from the EMBL/

GenBank/DDBJ database This sequence was cloned from

mouse kidney cDNA and used as a query in order to find

the com plete sequence from the database The IMAGE

clone 4190449 contained the full CDS of a putative mouse

L-fucokinase, which was utilized in the design of primers for

RT-PCR Two putative cDNAs of different sizes were

cloned representing two splice variants ofL-fucokinase The

long splice variant ofL-fucokinase consisted of 3270 bp,

encoding a protein of 1090 amino acids The sequence of the

shorter cDNA was similar to the sequence of the IMAGE

clone 4190449, consisting of 3057 bp The short splice

variant did not code for amino acids 921–992 present in the

long splice variant, thus the short version consisted of 1019

amino acids [Fig 3]

The long splice variant ofL-fucokinase contains exons

1–20, 21a, 22, 23a and 24 whereas the short one contains

exons 1–20, 21b, 23b and 24 As can be seen in Fig 2, exons

21b and 23b are wholly included in the longer variants of

these exons (21a and 23a respectively) The splice junction

fromexon 20 to exon 21a or 21b is not affected, neither is

the splice junction between exon 23a or 23b, and exon 24 In

conclusion, the alternative splicing maintains the reading

frame along the entire protein, therefore the protein variants

are identical in the amino-terminal end up to the alternative

splice area, in addition to the carboxy-terminal end after the

alternative splice area

There are three methionine codons (ATG) within a

300 bp region at the upstreamend of the longest open

reading frame in the mouse fucokinase mRNA sequence

(accession number AJ534942) The first ATG is estimated

to be the most probable CDS initiation site based on a probabilistic model using multiple parameters, including the Kozak translation initiation signal, as implemented in the

GENSCANanalysis tool [28]

Expression of fucokinase in mammalian cells The two splice variants of the murine fucokinase genes were expressed in COS-7 cells in frame with a 10-amino acid E2-Tag present in the pQM vector The molecular masses

of fucokinase proteins were determined by Western blot analysis; the tagged long splice variant had a mass of

125 kDa and the tagged short splice variant a mass of

115 kDa Both E2-Tagged splice variants had slightly greater molecular masses than the predicted 120 and

111 kDa proteins, respectively [Fig 4]

The production of L-fucose-1-phosphate from L-[3 H]fu-cose and ATP was measured in order to determine whether the expressed splice variants of fucokinase were functionally active The long splice variant showed significant enzyme activity; the specific enzyme activity was determined to be 598.5 pmolÆmg)1Æh)1in transfected COS-7 cells The activity

of the short splice variant was only marginally higher (13.7 pmolÆmg)1Æh)1) than the background in the COS-7 cells (11.4 pmolÆmg)1Æh)1) HumanL-fucokinase, IMAGE clone 4179554 (AJ441184) [22], was also transfected into COS-7 cells and assayed in regard to fucokinase activity The specific enzyme activity of the humanL-fucokinase was the same level, 12.3 pmolÆmg)1Æh)1, as the activity of the shorter mouse splice variant and the vector control (Fig 5A)

L-fucokinase activity is present in many different tissues, and exhibits high activity in kidney [21] The COS-7 cell line

is derived from monkey kidney cells and thus has some intrinsic fucokinase activity In order to discriminate the possible fucokinase activity of a short splice variant fromthe kidney cell backround, the short splice variant was also transfected into epithelial HeLa S3-cells The relatively weak enzymatic activity of the short splice variant could be detected in HeLa cells; the specific enzyme activity was 30.6 pmolÆmg)1Æh)1whereas the specific activity of the mock control was 8.4 pmolÆmg)1Æh)1(Fig 5b)

Cloning of murine GDP-L-fucose pyrophosphorylase The cloned human pyrophosphorylase (accession number AF017445) [23] was used as a query in BLAST searches to find a mouse EST corresponding to the putative pyro-phosphorylase Using this mouse EST as a probe, the 3¢ end

of the GDP-L-fucose pyrophosphorylase was cloned froma mouse kidney cDNA library by screening 1 · 106 recom-binant plasmids The 5¢ end of the sequence was resolved by the RACE-PCR method, using mouse kidney mRNA as the template as described in Experimental procedures The isolated cDNA consisted of 3480 bp, and the predicted CDS encoded a protein of 591 amino acids [Fig 6]

Pyrophosphorylase activity assay and the identification

of GDP-L-fucose Because we could detect only a faint protein band in SDS/ PAGE from cell lysate with the estimated molecular mass of 65.5 kDa that relates to GDP- -fucose pyrophosphorylase

(data not shown), we decided to identify accurately the

product of a GDP-L-pyrophosphorylase assay The cell

lysate expressing the pyrophosphorylase gene was incubated

with L-fucose-1-phosphate and GTP, and the resulting

product of the reaction was analyzed by ion-pair

reversed-phase HPLC The analysis revealed a peak with the same

retention time as the GDP-L-fucose standard (F) at

29.6 min in a sample containing the pyrophosphorylase,

whereas the vector control gave only a faint peak at

29.7 min [Fig 7] The peak was purified and subjected to

further analysis by MALDI-TOF MS, which gave a single

peak at 588.08 m/z, thus being identical to the GDP-L

-fucose control

Quantitative PCR and tissue distribution levels

ofL-fucokinase and pyrophosphorylase The primer and probe sequences and their positions in the mRNA sequence, for GDP-L-fucose pyrophosphorylase and the short and long splice variants ofL-fucokinase, are listed in Table 1

Various mouse tissues were analyzed for the expression of the three enzymes (GDP-L-fucose pyrophosphorylase, and short and long splice variants ofL-fucokinase) involved in the salvage pathway of GDP-L-fucose, to elucidate the possible differences between the various tissues Moreover, the ratio of long to short splice variants of -fucokinase in

Fig 3 Nucleotide sequence and deduced amino acid sequence of mouse L -fucokinase The predicted amino acid sequence for the coding area of the long splice variant of fucokinase consists of 1090 amino acids Due to alternative splicing, the amino acids 921–992 (bold letters) are not coded in the short splice variant of fucokinase The amino acids corresponding to the published peptide sequences of pig fucokinase [21] are underlined The sequence data of the short splice variant is available in the EMBL/GenBank/DDBJ Nucleotide Sequence Databases under Accession No AJ297482 and the long splice variant under the Accession No AJ534942.

different tissues was also determined (Table 2) Relative

expression levels, shown in Fig 8, were calculated following

normalization to 18S RNA In the subsequent calculations,

expression levels of those enzymes found in mouse liver were

assigned a relative expression value of one The expression

of both splice variants of L-fucokinase was found to be

relatively high in brain, ovary, testis and kidney In spleen,

heart and lung the expression was lower When calculating

the ratio between the long and short splice variants of

fucokinase it could be seen that the long splice variant was

more abundantly expressed in liver, kidney, ovary, testis,

spleen and heart In the lung the expression levels were

equal, whereas in brain the expression of the short splice

Fig 5 Fucokinase activities of the cell lysates of COS-7 cells (A) and

HeLa cells (B) transfected with the fucokinase cDNAs Enzyme activity

is expressed as pmol of L -[ 3 H]fucose incorporated onto ATP per hour

devided by the total protein content (A) Enzyme activities of COS-7

cells, transfected with the short and long splice variants of mouse

fucokinase (mFK) and human fucokinase (hFK, AJ441184) (B)

Fucokinase activities of HeLa cells transfected with vector or the short

splice variant of mouse fucokinase.

Fig 4 Western blot analysis of the expressed murine L -fucokinase in

COS-7 cells detected with E2-Tag antibodies Lane 1, negative COS-7

cell control; lane 2, short splice variant of mouse fucokinase and lane 3,

long splice variant of mouse fucokinase.

Fig 6 Nucleotide sequence and deduced amino acid sequences of murine

GDP- L -fucose pyrophosphorylase The 3.5 kb nucleotide sequence

predicts an amino acid sequence of 590 residues for the coding region

of GDP- L -fucose pyrophosphorylase The sequence data is available in

the EMBL/GenBank/DDBJ Nucleotide Sequence Databases under

Accession No AJ276067.

variant was higher than that of the longL-fucokinase splice

variant The expression pattern of GDP-L-fucose

pyro-phosphorylase resembles the pattern of L-fucokinase, i.e

expression was high in brain, ovary, testis and kidney

Again, the expression levels were lower in liver, spleen, heart and lung (Fig 8C)

Discussion The de novo synthesis of GDP-L-fucose, that converts

GDP-D-mannose to GDP-L-fucose, is evolutionary conserved and the enzymes involved in this pathway have been cloned from several bacteria [17], plants [29] and mammals [30] In addition, the de novo synthesis of GDP-L-fucose has been characterized in silico fromthe fruit fly [31] The alternative pathway of GDP-L-fucose synthesis, the salvage pathway, allows cells to activate monosaccharides that come from nutrition or fromlysosomal degradation of glycoproteins and glycolipids The sugars are phosphorylated by kinases and activated by pyrophosphorylases To date the salvage pathway of GDP-L-fucose has been identified only in mammals [21,23] The specific salvage pathway is also found for UDP-galactose, UDP-glucuronic acid and UDP-N-acetylgalactosamine [32]

The salvage pathway of GDP-L-fucose involvesL -fuco-kinase which catalyzes the transfer of phosphate fromATP

to free L-fucose, forming L-fucose-1-phosphate GDP-L -fucose pyrophophorylase then condensates L -fucose-1-phosphate with GTP to formGDP-L-fucose In the present study we have cloned the murine enzymes involved

in the salvage pathway of GDP-L-fucose and expressed themas functionally active enzymes Two splice variants of

L-fucokinase were cloned, but only the long splice variant was enzymatically active when expressed in mammalian cells The short splice variant did not show significant

Table 1 Probe and primer sequences in quantitative PCR FK short, short splice variant of L -fucokinase; FK long, long splice variant of

L -fucokinase; PP, GDP-l-fucose pyrophosphorylase; F, forward primer; R, reverse primer; P, probe.

Target gene Primer/Probe sequence Starting position in mRNA Length of amplicon

P a

P 5¢-AGTGTCTCTCCAAGTGTTCCTGAGCGCT-3¢ 1144

a Probe is antisense strand.

Table 2 Ratio of long splice variant to short splice variant of

L -fucokinase.

Fig 7 Ion-pair reversed-phase HPLC analysis of the product of the

enzymatic reaction catalyzed by GDP- L -fucose pyrophosphorylase (A)

Vector control in COS-7 cell lysate; (B) putative mouse

pyrophos-phorylase in COS-7 cell lysate; (C) GDP-sugar standards, 500 pmol of

each M, GDP- D -mannose, 18.6 min; R, GDP- D -rhamnose, 24.4 min;

F, GDP- L -fucose, 29.6 min.

enzymatic activity, but was expressed abundantly in many

tissues, especially in brain, which may indicate an

uniden-tified role for this variant

When comparing both splice variants of mouseL

-fuco-kinase cDNA sequence with the previously published human

L-fucokinase cDNA sequence (accession number AJ441184)

[22], it can be observed that the human fucokinase cDNA is

similar to the long splice variant of mouse fucokinase cDNA

at the 3¢ splice region The first and third methionines in the

murine sequence, in the upstream end of the CDS, are also

found in the human sequence (e.g BC032542) while the

second one has evolved into a leucine Although the

beginning of the human CDS has been proposed to start

fromthe position that corresponds to the third ATG in the

murine sequence [22], we suggest that the first ATG would be

a better starting codon than the third one; indeed, it is

predicted to be the first triplet in the CDS by several gene

prediction tools, e.g.GENSCAN analysis tool [28]

Further-more, a high degree of sequence similarity exists between the

mouse and the human cDNA sequences upstream of the

third ATG, suggesting that this segment is part of the CDS

In conclusion, we propose that the CDS starts not at the third

but at the first ATG in the murine sequence, and that the

human CDS starts at the corresponding position Thus, we suggest that the human CDS ofL-fucokinase becomes 94 amino acids longer than the corresponding CDS in the previous study [22]

L-fucose is a fundamental component of many mamma-lian glycoproteins and glycolipids Fucosylation requires GDP-L-fucose as a donor of L-fucose, and a specific fucosyltransferase to catalyze the transfer of L-fucose to the acceptor molecules The synthesis of GDP-L-fucose and its import into the Golgi lumen for a specific fucosyltrans-ferase is essential for selectin-dependent leukocyte traffick-ing and for normal human development Leukocyte adhesion deficiency type 2 (LADII), also known as a congenital disorder of glycosylation IIc, is a rare human disorder of fucose metabolism in which the patient suffers fromrecurrent infection, persistent leukocytosis and severe mental and growth retardation [33,34] Missense mutations

in a Golgi-localized GDP-fucose transporter lead to parti-ally defective function and are responsible for the defective fucosylation in LADII patients [35,36] Studies with LADII patients show that oral supplementation of fucose can restore selectin ligands and correct the immunodeficiency [37,38] In this scenario, GDP-L-fucose is synthesized from oral fucose through the salvage pathway, which elevates the

am ount of GDP-L-fucose in the cytosol, leading to enhanced GDP-fucose uptake into the Golgi [35] In a study by Smith et al [39], the targeted disruption of the FX locus in the mouse ablates the de novo pathway for GDP-fucose synthesis fromGDP-mannose causing adult animals

to lack almost completely the fucosylated glycans in multiple tissues, leading to symptoms similar to those of LADII The FX-deficient mice are completely dependent on dietary fucose, which restores the synthesis of GDP-fucose through the salvage pathway

The salvage metabolism accounts for approximately only 10% of the intracellular pool of GDP-L-fucose [40] However, the enzymes of the salvage pathway are expressed with relatively high intensities in various animal tissues, e.g brain, ovary, testis, kidney and liver, as shown by the quantitative real time PCR analysis in the present study and also in previous studies [21,23] The wide expression of the enzymes involved in the salvage pathway of GDP-L-fucose can also be deduced fromthe large amount of data available fromdifferent EST libraries (e.g http://www.ncbi.nlm.nih gov/UniGene) Our analysis of the expression of the enzymes involved in the salvage pathway of GDP-L-fucose indicates that not only the de novo pathway alone, but also the salvage pathway could have an essential role in the synthesis of GDP-L-fucose in the cytosol The importance and the regulatory mechanisms of the enzymes in the salvage pathway of GDP-L-fucose have not been elucidated, thus futher studies are needed

Acknowledgements

We thank Tuula Kallioinen and Sirkka-Liisa Kauranen for skilled technical assistance in molecular biology, and Kati Vena¨la¨inen and Leena Penttila¨ for assistance in HPLC and MALDI-TOF MS analysis The work was supported in part by Research Grants of the Academy of Finland, the Technology Development Center (TEKES), Helsinki, the Sigrid Juselius Foundation, and the Helsinki University Central Hospital Fund (EVO).

Fig 8 Tissue expression patterns of murine L -fucokinase short and long

splice variants and GDP- L -fucose pyrophosphorylase The expression

levels of the long splice variant of L -fucokinase (A), the short splice

variant of L -fucokinase (B) and GDP- L -fucose pyrophosphorylase (C)

were detected by quantitative real time PCR The mRNA expression

levels in each tissue were expressed relative to expression in the liver.

1 Walz, G., Aruffo, A., Kolanus, W., Bevilacqua, M & Seed, B.

(1990) Recognition by ELAM-1 of the sialyl-Lex determinant on

myeloid and tumor cells Science 250, 1132–1135.

2 Harris, R.J & Spellman, M.W (1993) O-linked fucose and other

post-translational modifications unique to EGF modules

Glyco-biology 3, 219–224.

3 Staudacher, E., Altmann, F., Wilson, I.B & Marz, L (1999)

Fucose in N-glycans: fromplant to man Biochim Biophys Acta

1473, 216–236.

4 Greenwell, P (1997) Blood group antigens: molecules seeking a

function? Glycoconjugate J 14, 159–173.

5 Lloyd, K.O (2000) The chemistry and immunochemistry of blood

group A, B, H, and Lewis antigens: past, present and future.

Glycoconjugate J 17, 531–541.

6 Lasky, L.A (1992) Selectins: interpreters of cell-specific

carbo-hydrate information during inflammation Science 258, 964–969.

7 Lowe, J.B (1997) Selectin ligands, leukocyte trafficking, and

fuc-osyltransferase genes Kidney Int 51, 1418–1426.

8 Mori, E., Hedrick, J.L., Wardrip, N.J., Mori, T & Takasaki, S.

(1998) Occurrence of reducing terminal N-acetylglucosamine

3-ulfate and fucosylated outer chains in acidic N-glycans of porcine

zona pellucida glycoproteins Glycoconjugate J 15, 447–456.

9 Johnston, D.S., Wright, W.W., Shaper, J.H., Hokke, C.H., Van den

Eijnden, D.H & Joziasse, D.H (1998) Murine sperm-zona

bind-ing, a fucosyl residue is required for a high affinity sperm-binding

ligand A second site on spermbinds a nonfucosylated,

beta-ga-lactosyl-capped oligosaccharide J Biol Chem 273, 1888–1895.

10 Xiang, J & Bernstein, I.A (1992) Differentiative changes in

fucosyltransferase activity in newborn rat epidermal cells.

Biochem Biophys Res Commun 189, 27–32.

11 Ruggiero-Lopez, D., Biol, M.C., Louisot, P & Martin, A (1991)

Participation of an endogenous inhibitor of fucosyltransferase

activities in the developmental regulation of intestinal fucosylation

processes Biochem J 279, 801–806.

12 Moloney, D.J., Panin, V.M., Johnston, S.H., Chen, J., Shao, L.,

Wilson, R., Wang, Y., Stanley, P., Irvine, K.D., Haltiwanger, R.S.

& Vogt, T.F (2000) Fringe is a glycosyltransferase that modifies

Notch Nature 406, 369–375.

13 Bruckner, K., Perez, L., Clausen, H & Cohen, S (2000)

Glyco-syltransferase activity of Fringe modulates Notch–Delta

inter-actions Nature 406, 411–415.

14 Fukuda, M (1996) Possible roles of tumor-associated

carbo-hydrate antigens Cancer Res 56, 2237–2244.

15 Hiraishi, K., Suzuki, K., Hakomori, S & Adachi, M (1993) Le(y)

antigen expression is correlated with apoptosis (programmed cell

death) Glycobiology 3, 381–390.

16 Tonetti, M., Sturla, L., Bisso, A., Benatti, U & De Flora, A.

(1996) Synthesis of GDP- L -fucose by the human FX protein.

J Biol Chem 271, 27274–27279.

17 Tonetti, M., Sturla, L., Bisso, A., Zanardi, D., Benatti, U &

De Flora, A (1998) The metabolism of 6-deoxyhexoses in

bacterial and animal cells Biochimie 80, 923–931.

18 Ishihara, H., Massaro, D.J & Heath, E.C (1968) The metabolism

of L -fucose III The enzymatic synthesis of beta- L -fucose

1-phosphate J Biol Chem 243, 1103–1109.

19 Ishihara, H & Heath, E.C (1968) The metabolism of L -fucose IV.

The biosynthesis of guanosine diphosphate L -fucose in porcine

liver J Biol Chem 243, 1110–1115.

20 Kilker, R.D., Shuey, D.K & Serif, G.S (1979) Isolation and

properties of porcine thyroid fucokinase Biochim Biophys Acta

570, 271–283.

21 Park, S.H., Pastuszak, I., Drake, R & Elbein, A.D (1998)

Puri-fication to apparent homogeneity and properties of pig kidney

L -fucose kinase J Biol Chem 273, 5685–5691.

22 Hinderlich, S., Berger, M., Blum e, A., Chen, H., Ghaderi, D & Bauer, C (2002) Identification of human L -fucose kinase am i-noacid sequence Biochem Biophys Res Commun 294, 650–654.

23 Pastuszak, I., Ketchum, C., Hermanson, G., Sjoberg, E.J., Drake,

R & Elbein, A.D (1998) GDP- L -fucose pyrophosphorylase Purification, cDNA cloning, and properties of the enzyme J Biol Chem 273, 30165–30174.

24 Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W & Lipman, D.J (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs Nucleic Acids Res 25, 3389–3402.

25 Pearson, W.R (1990) Rapid and sensitive sequence comparison with FASTP and FASTA Methods Enzymol 183, 63–98.

26 Rabina, J., Maki, M., Savilahti, E.M., Jarvinen, N., Penttila, L & Renkonen, R (2001) Analysis of nucleotide sugars fromcell lysates by ion-pair solid-phase extraction and reversed-phase high-performance liquid chromatography Glycoconjugate J 18, 799–805.

27 Maki, M., Jarvinen, N., Rabina, J., Roos, C., Maaheim o, H., Mattila, P & Renkonen, R (2002) Functional expression of Pseudomonas aeruginosa GDP-4-keto-6-deoxy- D -mannose reduct-ase which synthesizes GDP-rhamnose Eur J Biochem 269, 593–601.

28 Heid, C.A., Stevens, J., Livak, K.J & Williams, P.M (1996) Real time quantitative PCR Genome Res 6, 986–994.

29 Bonin, C.P., Potter, I., Vanzin, G.F & Reiter, W.D (1997) The MUR1 gene of Arabidopsis thaliana encodes an isoformof

GDP-D -mannose-4,6-dehydratase, catalyzing the first step in the de novo synthesis of GDP- L -fucose Proc Natl Acad Sci USA 94, 2085–2090.

30 Reitman, M.L., Trowbridge, I.S & Kornfeld, S (1980) Mouse lymphoma cell lines resistant to pea lectin are defective in fucose metabolism J Biol Chem 255, 9900–9906.

31 Roos, C., Kolm er, M., Mattila, P & Renkonen, R (2002) Com -position of Drosophila melanogaster proteome involved in fuco-sylated glycan metabolism J Biol Chem 277, 3168–3175.

32 Bulter, T & Elling, L (1999) Enzymatic synthesis of nucleotide sugars Glycoconjugate J 16, 147–159.

33 Etzioni, A., Frydman, M., Pollack, S., Avidor, I., Phillips, M.L., Paulson, J.C & Gershoni-Baruch, R (1992) Brief report: recurrent severe infections caused by a novel leukocyte adhesion deficiency N Engl J Med 327, 1789–1792.

34 Becker, D.J & Lowe, J.B (1999) Leukocyte adhesion deficiency type II Biochim Biophys Acta 1455, 193–204.

35 Luhn, K., Wild, M.K., Eckhardt, M., Gerardy-Schahn, R & Vestweber, D (2001) The gene defective in leukocyte adhesion deficiency II encodes a putative GDP-fucose transporter Nat Genet 28, 69–72.

36 Lubke, T., Marquardt, T., Etzioni, A., Hartmann, E., von Figura,

K & Korner, C (2001) Complementation cloning identifies CDG-IIc, a new type of congenital disorders of glycosylation, as a GDP-fucose transporter deficiency Nat Genet 28, 73–76.

37 Marquardt, T., Luhn, K., Srikrishna, G., Freeze, H.H., Harms, E.

& Vestweber, D (1999) Correction of leukocyte adhesion defi-ciency type II with oral fucose Blood 94, 3976–3985.

38 Luhn, K., Marquardt, T., Harms, E & Vestweber, D (2001) Discontinuation of fucose therapy in LADII causes rapid loss

of selectin ligands and rise of leukocyte counts Blood 97, 330–332.

39 Smith, P.L., Myers, J.T., Rogers, C.E., Zhou, L., Petryniak, B., Becker, D.J., Homeister, J.W & Lowe, J.B (2002) Conditional control of selectin ligand expression and global fucosylation events

in mice with a targeted mutation at the FX locus J Cell Biol 158, 801–815.

40 Yurchenco, P.D & Atkinson, P.H (1977) Equilibration of fucosyl glycoprotein pools in HeLa cells Biochemistry 16, 944–953.