increased levels of fucosyltransferase ix and carbohydrate lewisx adhesion determinant in human nt2n...

Increased Levels of Fucosyltransferase IXand Carbohydrate Lewisx AdhesionDeterminant in Human NT2N Neurons

Catarina Brito,1 Cristina Escrevente,1 Celso A. Reis,2 Virginia M.-Y. Lee,3

John Q. Trojanowski,3 and Julia Costa1,4*1Instituto de Tecnologia Quımica e Biologica, Oeiras, Portugal2Institute of Molecular Pathology and Immunology of the University of Porto, Portugal3Center for Neurodegenerative Disease Research, Department of Pathology and Laboratory Medicine,University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania4Instituto de Biologia Experimental e Tecnologica, Oeiras, Portugal

The expression of the fucosylated carbohydrate Lewisx

(Lex) determinant (Gal(b1–4)[Fuc(a1–3)]GlcNAc-R) hasbeen found in glycoproteins, proteoglycans, and glyco-lipids from the nervous system. Evidence suggests itsassociation with cell–cell recognition, neurite outgrowth,and neuronal migration during central nervous systemdevelopment. In the present work, we detectedincreased levels of Lex in differentiated human NT2Nneurons cultured in vitro. To identify which fucosyltrans-ferase (FUT) synthesized the Lex in NT2N neurons, RT-PCR, FUTsubstrate specificity and Western blot analysiswere carried out. Strong activity toward acceptorsGalb4GlcNAc-O-R and Fuca2Galb4GlcNAc-O-R [R ¼-(CH2)3NHCO(CH2)5NH-biotin], together with strongFUT9 detection by Western blot and presence of tran-scripts showed that FUT9 was the enzyme associatedwith Lex biosynthesis in NT2N neurons. Lex was detectedat the plasma membrane of NT2N neurons, in lysosomesmarked with lysosomal-associated membrane protein 1(LAMP-1), and it was found for the first time to colocalizewith the tetanus neurotoxin-insensitive vesicle-associ-ated membrane protein (TI-VAMP) that defines the TI-VAMP exocytic compartment that is involved in neuriteoutgrowth. Furthermore, incubation with anti-Lex mono-clonal antibody L5 led to impaired adhesion of NT2Nneurons to the surface matrix and inhibited neurite initia-tion. In conclusion, FUT9 and its product Lex are de-tected specifically in human NT2N neurons and ourresults indicate that they underlie cell differentiation, celladhesion, and initiation of neurite outgrowth in thoseneurons. VVC 2007 Wiley-Liss, Inc.

Key words: NT2N neuron; neuron differentiation; fucosyl-transferase IX; Lewisx; neurite outgrowth

Glycoconjugates containing fucose (Fuc) have beenshown to play important roles in cell adhesion and cell–cell interactions. The fucosylated carbohydrate Lewisx

(Lex) determinant (Gal(b1-4)[Fuc(a1-3)]GlcNAc) fromglycolipids, glycoproteins, or proteoglycans is expressed in

several human tissues, including the brain. It is the mostprobable candidate to play an important role in normalbrain development because its expression was found tem-porally and spatially regulated in the developing centralnervous system (CNS) of diverse species, includinghuman, rat, mouse, chicken, and Xenopus (Dasgupta et al.,1996; Gotz et al., 1996; Mai et al., 1999; Yoshida-Noroet al., 1999). Furthermore, complex-type N-glycans withterminal Lex on one arm are among the most abundantneutral carbohydrates detected in human brain (Chenet al., 1998). At the cellular level, it has been detected inmouse neurons at the surfaces of cell bodies, processes, andgrowth cones (Nishihara et al., 2003). Furthermore, Lex

seems to have a role in neurite outgrowth or adhesionbecause anti-Lex antibodies inhibit neurite outgrowth inexplant brain cultures from Xenopus embryos (Yoshida-Noro et al., 1999). A specific association between Lex andmouse brain cells was also suggested by its role as a neuralstem cell marker (Capela and Temple, 2002).

The final step in the biosynthesis of Lex is catalyzedby a3-fucosyltransferases (FUTs), which add a Fuc residueto Type II-based oligosaccharide chains. To date, six a3-fucosyltransferases have been cloned and characterized(FUT3, FUT4, FUT5, FUT6, FUT7, and FUT9).FUT10 and FUT11 have been identified as fucosyltrans-ferases due to their degree of homology with other fuco-syltransferases (Roos et al., 2002), however, their activity

Contract grant sponsor: European Commission; Contract grant number:

Signalling and Traffic LSHG-CT-2004-503228, CellProm 500039-2;

Contract grant sponsor: Fundacao para a Ciencia e Tecnologia, Portugal;

Contract grant number: POCTI/BCI/38631/2001.

*Correspondence to: Dr. Julia Costa, Instituto de Tecnologia Quımica e

Biologica, Avenida da Republica, Apartado 127, 2781-901 Oeiras, Portugal.

E-mail: [email protected]

Received 3 October 2006; Revised 28 November 2006; Accepted 7

December 2006

Published online 4 March 2007 in Wiley InterScience (www.

interscience.wiley.com). DOI: 10.1002/jnr.21230

Journal of Neuroscience Research 85:1260–1270 (2007)

' 2007 Wiley-Liss, Inc.

has not been validated (Baboval and Smith, 2002). Theenzymes show different specificity patterns: they can allsynthesize the Lex determinant, with the exception ofFUT7 (Becker and Lowe, 2003). FUT4 and FUT9 are themain enzymes responsible for the synthesis of Lex (Gra-benhorst et al., 1998; Baboval et al., 2000). FUT3 has onlyresidual a3-FUT activity and exhibits predominant a4-FUT activity (Costa et al., 1997). Their specificity differ-ences together with their different tissue distributionsunderlie distinct biologic roles. FUT9 is expressed mainlyin the CNS, in both the developing and mature brain ofhuman, rat, and mouse (Kaneko et al., 1999; Babovalet al., 2000; Cailleau-Thomas et al., 2000). a3-FUT hasbeen shown to be responsible for the synthesis of Lex inthe mouse brain (Nishihara et al., 2003), where it is con-trolled by Pax6, a transcription factor involved in brainpatterning and neurogenesis (Shimoda et al., 2002).Recently, the importance of FUT9 for the synthesis androle of Lex in mouse brain was shown unequivocally withthe study of FUT9�/� mouse (Kudo et al., 2006). Thisanimal showed disappearance of Lex in the brain concomi-tantly to increased anxiety-like behaviors.

We have used NTERA-2/cl.D1 (NT2) cells as amodel of the human CNS in this study. The NT2 cell lineis derived from a human embryonic carcinoma (EC), andis characterized by undifferentiated neuronally committedprogenitor cells (NT2�) that after exposure to retinoicacid (RA), differentiate into postmitotic neurons (NT2N)(Pleasure et al., 1992). Differentiation of NT2N neuronsin culture follows a pattern of differential gene expressionsimilar to that of the neuronal precursors during neuro-genesis (Przyborski et al., 2000). NT2N neurons havebeen shown by several laboratories to have characteristicsof primary human neurons (Sandhu et al., 2003).

We have shown that differentiated NT2N neuronsexpressed Lex, which was synthesized by FUT9. The Lex

was detected on the cell surface of both dendrites andaxons and, intracellularly, was detected in lysosomes andfound to colocalize with the exocytic TI-VAMP compart-ment that is associated with neurite outgrowth. Incubationwith anti-Lex antibody caused a decrease in the initiationof neurite outgrowth and cell adhesion.

MATERIALS ANDMETHODS

Neuronal Culture

NTERA-2/cl.D1 cells were cultured and differentiatedinto neurons essentially as described previously (Pleasure et al.,1992). Briefly, undifferentiated NT2� cells were grown inOpti-MEM I (Gibco, Carlsbad, CA) with 5% fetal bovine serum(FBS) (HyClone, Logan, UT), 100 U/ml penicillin (P) and 100lg/ml streptomycin (S) (Gibco). For differentiation, cells werecultured in Dulbecco’s modified Eagle’s medium (DMEM)-High Glucose (HG) (Gibco) with 10% FBS, P, S, and 10 lMretinoic acid (Sigma, St. Louis, MO) for 5 weeks. For neuronenrichment cell culture was in DMEM-HG with 5% FBS, P, Ssupplemented with the mitosis inhibitors 1 lM cytosine arabi-noside (Sigma), 10 lM fluorodeoxyuridine (Sigma), and 10 lM

uridine (Sigma) for further 10 days. NT2N neurons were recov-ered after selective hydrolysis with trypsin (Gibco) and plated on10 lg/ml poly-D-lysine (Sigma) and 0.26 mg/ml Matrigel-coated (BD Biosciences, Bedford, MA) surfaces.

Fucosyltransferase Assays

For FUT assays, cells were solubilized with 50 mMMOPS-NaOH buffer pH 7.5, containing 1% Triton X-100 (TX-100),48C, for 1 hr. The FUT activity in cell extracts was determinedby the incorporation of radioactive Fuc from the GDP-[14C]-Fuc donor onto several acceptors as described previously (Moraiset al., 2001). One unit of enzyme activity was defined as theamount of enzyme catalyzing the transfer of 1 lmol Fuc/minto the acceptor. The synthetic acceptors used were the Type I:Galb3GlcNAc-O-R, Neu5Aca2-3Galb3GlcNAc-O-R, andFuca2Galb3GlcNAc-O-R (H Type I); and the Type II:Galb4GlcNAc-O-R (LacNAc), Neu5Aca2-3Galb4GlcNAc-O-R (sialyl-LacNAc), and Fuca2Galb4GlcNAc-O-R (H Type II)(Lectinity Holding, Moscow, Russia), where R ¼ -(CH2)3NHCO(CH2)5NH-biotin.

SDS-PAGE and Western Blot Analysis

Total cell protein was solubilized with 50 mM Tris-HClbuffer pH 7.4, containing 5 mM EDTA, 1% TX-100, and pro-tease inhibitors (Complete Cocktail; Roche, Mannheim, Ger-many). Proteins were separated by SDS-PAGE and transferredto polyvinylidene difluoride membrane. Membranes wereblocked in 5% nonfat dry milk (Nestle, Oeiras, Portugal) in PBSwith 0.1% Tween-20. Primary antibodies were: L5 anti-Lex

monoclonal antibody (Streit et al., 1996), C-20 anti-FUT4 andC-17 anti-FUT9 polyclonal antibodies (Santa Cruz Biotechnol-ogy, Santa Cruz, CA) at 1:500 dilutions in blocking solution,and H4A3 anti-human LAMP-1 monoclonal antibody (BDBiosciences) at 1:200 in blocking solution. HRP-labeled poly-clonal secondary antibodies anti-mouse IgM, anti-goat IgG, andanti-mouse IgG (Sigma) were diluted in blocking solution.Chemiluminescent detection was by the ECL Plus system(Amersham Biosciences, Piscataway, NJ).

Mucin extract from a gastric mucosa adjacent to a gastriccarcinoma was used as a control in anti-Lex Western blot. Thesample was obtained from a surgical specimen of a patientundergoing surgical resection in the Hospital S. Joao, Porto,Portugal. Sample was collected after informed consent. Mucinextraction was carried out as described previously (Reis et al.,1997).

RNA Extraction and Reverse Transcription

Total cellular RNA was extracted using the RNeasyextraction kit (Qiagen, Valencia, CA) according to the supplier’sprotocol. Reverse transcription was carried out with randomprimers (Invitrogen, Carlsbad, CA) using Moloney murine leu-kemia virus reverse transcriptase (M-MLV) (Invitrogen) asdescribed previously (Escrevente et al., 2006). Amplification ofthe DNA sequences corresponding to the different FUTs andthe primers used were as described previously (Escrevente et al.,

Increased Level of FUT9 in NT2N Neurons 1261

Journal of Neuroscience Research DOI 10.1002/jnr

2006). The PCR products were analyzed by electrophoresis in1.5% agarose gel with 0.5 lg/ml ethidium bromide and theirsize estimated relatively to a gene ruler consisting of 1 kb and100 bp DNA ladders.

Immunofluorescence Microscopy

NT2� cells, grown on coverslips to half confluency, andNT2N neurons, after 8 days in 24-well plates with coverslipscoated with poly-D-lysine and Matrigel, were washed with PBScontaining 0.5 mM MgCl2, fixed with 4% (w/v) paraformalde-hyde and 4% sucrose in PBS for 20 min and, when described,permeabilized with 0.1 or 0.3% (w/v) TX-100 in PBS (NT2�

cells and NT2N neurons, respectively) for 20 min. Blockingwas done with 1% bovine serum albumin (BSA) and 0.1% TX-100 in PBS for 1 hr and primary and secondary antibodies wereincubated at room temperature for 2 and 1 hr, respectively.Antibody solutions were prepared in blocking solution andwashes were carried out with PBS. Coverslips were mounted inAirvol. Primary antibodies were: rat IgM anti-Lex L5 (1:200);mouse IgG anti-Lex SH1 used at 10-fold concentrate containingBSA 1% (Singhal et al., 1990); mouse IgM anti-Lewisy (Ley)AH6 (1:1) (Abe et al., 1983); rabbit IgG anti-NF-L (1:2,000)(Trojanowski et al., 1989); mouse IgG anti-hypophosphorylatedNF-M RMdO.20 (1:1) (Pleasure et al., 1992); rat IgG anti-highly phosphorylated NF-M H014 (1:1) (Pleasure et al., 1992);mouse IgG anti-TI-VAMP clone 158.2 (1:200) (Alberts et al.,2003); mouse IgG anti-EEA-1 (1:200) (BD Biosciences); mouseIgG anti-LAMP1 H4A3 (1:50); mouse IgG anti-GM130 (1:200)(BD Biosciences); and sheep IgG anti-TGN46 (1:200) (Serotec,Oxford, UK). Secondary antibodies were: goat anti-mouse IgGTRITC-conjugate (1:64) (Sigma); goat anti-rat IgM TRITC-conjugated (1:50) (Jackson ImmunoResearch, West Grove,PA); goat anti-mouse IgM TRITC-conjugated (1:50) (JacksonImmunoResearch); goat anti-mouse IgG AlexaFluor488 (1:500)(Molecular Probes. Eugene, OR); donkey anti-mouse IgGAlexaFluor488 (1:500) (Molecular Probes); and donkey anti-sheep IgG AlexaFluor488 (1:500) (Molecular Probes). For dou-ble labeling with anti-LAMP-1 and anti-TI-VAMP mouseIgGs, neurons were permeabilized with 0.1% TX-100 for 10min, and incubated sequentially with the two antibodies. Afterincubation with anti-LAMP-1 followed by goat anti-mouseIgG-AlexaFluor594 conjugated, non-labeled goat anti-mouseIgG Fab fragment (Sigma) at the saturating concentration of 0.2mg/ml was used. Incubations with anti-TI-VAMP followed bygoat anti-mouse IgG-AlexaFluor488 were carried out. As con-trol, the experiment was carried out without anti-LAMP-1 oranti-TI-VAMP, and resulted in staining patterns similar to thoseobtained for each primary antibody independently.

Confocal images were acquired in a confocal microscopeSP2+AOBS (Leica) equipped with a 633 objective usingZoom 2. Each fluorophore used was excited and detected sepa-rately to avoid channel bleed-through. For each channel, photo-multiplier gains and offsets were adjusted to use full imagedynamic range. Each picture consisted of a z-series of 20 imagesof 1,024 3 1,024 pixel resolution with a pinhole of 1.0 Airyunit and typically, 0.6 lm focus intervals. Merge between chan-nels and maximum z-projections were created using the opensource ImageJ software version 1.35 (http://www.rsb.info.nih.gov/ij/). Image editing software (Adobe Photoshop, Adobe

Systems, San Jose, CA) was used to create montages. The onlymodifications made during image editing were linear brightnessand contrast adjustments.

Anti-Lex incubation Assays

NT2N neurons were plated in 24-well plates (5 3 104

cells/cm2) coated with poly-D-lysine and Matrigel. The assayswere carried out in conditioned medium with mitosis inhi-bitors.

Anti-Lex antibody (L5) was added to NT2N neurons atthe time of plating or to neurons already in culture for 4 days.Equivalent volumes of PBS (the solvent of the stock solution ofanti-Lex) and equivalent concentrations of the non-related IgMMOPC-104E (Sigma) that recognizes a1-3 glucan (Klimpel andGoldman, 1988) were used as control. Cells were observed bybright-field microscopy using a Leica DM IRB inverted micro-scope attached to an Olympus DP11 digital camera. Neuritelength was measured using the NeuronJ 1.1.0 plugin (Meijeringet al., 2004) of the open source ImageJ software version 1.35. Aneurite was defined as a process with a length greater than onecell body diameter (approximately 15 lm). At least 60 neuriteswere measured. Experiments were carried out at least twice.Statistical analysis was carried out using the GraphPad Prism 4software.

RESULTS

Level of the Lex Determinant Is Increased DuringNT2N Neurons Differentiation

To monitor Lex during differentiation of the terato-carcinoma NT2� cells into NT2N neurons on retinoicacid induction, we have analyzed the two cell types by im-munofluorescence microscopy with several monoclonalanti-Lex antibodies: L5 (Streit et al., 1996), SH1 (Singhalet al., 1990), and anti-SSEA-1 (Solter and Knowles, 1978).We found that both L5 and SH1 antibodies stained NT2Nneurons but not NT2� cells (Fig. 1). The staining ap-peared in vesicles, which were concentrated in the cellbody, with a perinuclear pattern, and extended into neu-rites. The anti-SSEA-1 antibody presented no reactivity inNT2� nor NT2N neurons (results not shown), asdescribed by other authors (Fenderson et al., 1987). Thelabeling in NT2N neurons did not consist of materialinternalized from FBS nor Matrigel because it was de-tected by immunofluorescence microscopy in NT2N neu-rons cultured in serum-depleted medium and on poly-D-lysine coated surfaces (results not shown). As a control forneuronal differentiation a polyclonal antibody against lowmolecular weight neurofilament protein (NF-L) was usedas neuronal marker, which stained NT2N neurons (Fig. 1)but not NT2� cells (results not shown). Other Lewisdeterminants derived from Type II oligosaccharides werealso monitored. Ley was detected with the AH6 antibodyin NT2� cells but not in NT2N neurons (Fig. 1). Thisdecrease was due probably to a downregulation of a2-FUT during differentiation. Alternatively, reorganizationof glycosyltransferases in the Golgi during differentiationcould occur in a way that a2-FUT would not have accessto the Type II precursor. The sLex epitope was not

1262 Brito et al.


detected in NT2� cells nor in NT2N neurons (results notshown).

Lewis determinants derived from Type I precursors,Lea, Leb, and sLea, were not expressed at detectable levelsin any of the cell lines (results not shown).

Lex From NT2N Neurons Is Synthesized by FUT9

To determine the expression pattern of FUTs duringNT2N differentiation RT-PCR analysis was carried out(Fig. 2). The RT-PCR products of FUT4 (516 bp),FUT5 (555 bp), FUT6 (534 bp), and FUT9 (809 bp) werepresent in NT2� cells, during RA-induced differentiation,and in NT2N neurons. FUT10 transcripts (681 bp) weredetected in all stages of differentiation except in NT2Nneurons. FUT3 (521 bp) and FUT7 (497 bp) were onlydetected, with weak signals, in late differentiating stages,starting after 24 days in the presence of RA, and in NT2Nneurons. FUT11 transcripts (650 bp) were not detected inany of the stages analyzed. These results showed that therewere transcripts of FUT4, FUT5, FUT6, FUT9, and to aminor extent FUT3 and FUT7, in NT2N neurons.

To elucidate the activity profiles of FUTs during thedifferentiation of NT2N cells, we studied the acceptorspecificities of cell homogenates during the differentiationprocess toward a panel of small oligosaccharides of the

Type I and II acceptor substrates coupled to a hydrophobicmoiety. Activity assays showed that Galb4GlcNAc-R(LacNAc-R) was a good substrate, leading to the produc-tion of Lex determinant in every cell homogenate fromeach differentiation step analyzed. A substitution of Galwith a2-linked Fuc (H Type II-R) yielded an increase inactivity. Nevertheless, a substitution of the Gal with a2,3-linked N-acetylneuraminic acid (sLacNAc-R) yieldedvery low radioactivity incorporation (Fig. 3). These resultssuggested that FUT4 or FUT9 would be the enzymesassociated with the synthesis of Lex in differentiatingNT2N neurons, because FUT5 and FUT6 are capable offucosylating sialylated Type II acceptors (Weston et al.,1992a,b), the activity of which was not observed for thecell extracts tested in this study (Fig. 3).

Activity toward Type I acceptors, Galb3GlcNAc-R,NeuAca2,3Galb3GlcNAc-R, and H Type I-R, to gener-ate the Lea, sLea, and Leb epitopes, respectively, was verypoor, from NT2� cells to NT2N neurons, representingresidual or not detectable values (data not shown). Theseresults agreed with the absence or extremely low amountsof FUT3 transcripts detected (Fig. 2).

To distinguish between FUT4 and FUT9 as theenzymes synthesizing Lex during NT2N neuron differen-tiation, we have carried out Western blot analysis withpolyclonal antibodies against carboxy-terminal peptides of

Fig. 1. Immunodetection of Lewis de-terminants in NT2� cells and NT2Nneurons by immunofluorescence confo-cal microscopy. Permeabilized NT2�

cells and NT2N neurons were probedwith the monoclonal antibodies anti-Lex

L5 and SH1; anti-Ley AH6 (shown inred) and a polyclonal antibody anti-NF-L (shown in green). Secondary antibod-ies were: goat anti rat IgM-TRITC, goatanti-mouse IgG-AlexaFluor594, goatanti-mouse IgM-TRITC, and goat anti-rabbit IgG-AlexaFluor488. The experi-ment was done at least three independenttimes. Scale bar ¼ 10 lm.



FUT4 and FUT9. FUT4 and FUT9 proteins weredetected in NT2� cells (Fig. 4). FUT4 had an apparentmolecular mass of 63 kDa, which probably correspondedto the long form of the enzyme, ELFT-L (Goelz et al.,1990). The polypeptide chain has a predicted molecularmass of 59 kDa with two potential N-glycosylation sites(Goelz et al., 1990). Therefore, the difference between theobserved molecular mass and the theoretical predictionindicated that the glycosylation sites were occupied.FUT9 showed an apparent molecular mass of 44 kDa. Thepolypeptide chain has a predicted molecular mass of 42

kDa with three potential N-glycosylation sites (Kanekoet al., 1999). The 2 kDa molecular mass observed wasprobably due to N-glycosylation. Human FUT9 expressedtransiently in HeLa cells was used as a positive control forFUT9 detection (results not shown). After differentiationto NT2N neurons, the levels of FUT9 were increasedmarkedly whereas FUT4 was no longer detected even af-ter long exposure times (Fig. 4). These results showedclearly that FUT9 was the a3-FUT that catalyzed the syn-thesis of Lex in human NT2N neurons. Additionally, wealso found transient higher levels of FUT9 in cells treatedwith RA for 10 and 17 days.

In conclusion, the a3-FUT activity observed inNT2� cells was the result of combined activities of FUT4and FUT9 whereas in NT2N cells only FUT9 participatedin the Lex biosynthesis. The observation that NT2N cellscontained FUT4 transcripts (Fig. 2) but not the corre-

Fig. 2. RT-PCR analysis of fucosyltransferases along NT2N neuronsdifferentiation induced by retinoic acid. cDNA was obtained from un-differentiated NT2� cells (1), cells in retinoic acid for 3 (2), 10 (3), 17(4), 24 (5), and 31 (6) days, cells in mitosis inhibitors for 10 days (7) andmature NT2N neurons plated on poly-D-lysine and Matrigel for 8 days(8). The PCR products were electrophoretically resolved on 1.5% aga-rose gel in the presence of 0.5 lg/ml ethidium bromide. The sizes ofthe amplified fragments are indicated on the left in base pairs (bp). Thegels presented correspond to one representative experiment of the fourexperiments carried out from three independent RNA extractions.

Fig. 3. Fucosyltransferase activity: acceptor specificity profile alongNT2N neurons differentiation induced by retinoic acid. Fucosyltrans-ferase activity was determined using 13 106 cells of each differentiationstep (1–8 as indicated in Fig. 2 legend) per assay with the syntheticacceptors: Galb4GlcNAc-O-R (LacNAc), NeuAca2-3Galb4GlcNAc-O-R (sLacNAc), and Fuca2Galb4GlcNAc-O-R (H type II). R ¼-(CH2)3NHCO(CH2)5NH-biotin.

Fig. 4. Immunodetection by Western blot of FUT4 and FUT9 alongNT2N neurons differentiation induced by retinoic acid. Total proteinfrom 1 3 106 cells of each differentiation step was applied per lane (1–8as indicated in Fig. 2 legend). Detection was carried out using anti-human FUT4 and FUT9 polyclonal antibodies C-17 and C-20, respec-tively. As secondary antibody a polyclonal antibody anti-goat IgGHRP-conjugated was used. Detection was carried out by the ECL+method. Experiments were carried out twice from two independentprotein extractions of each differentiation step. For each experiment,one blot per FUT was done and then reprobed for the other FUT toconfirm that differences did not arise from different protein loads.

1264 Brito et al.


sponding protein (Fig. 4) suggested that FUT4 was post-transcriptionally downregulated in neurons.

Identification of a Lex-Carrier GlycoproteinAssociated Specifically With NT2N Neurons

The L5 antibody has been described to react withLex present on neural cell adhesion molecule L1, neuralchondroitin sulfate proteoglycans, and other neuronal gly-cosylated proteins. Furthermore, it was shown to bind amouse astrocyte proteoglycan of approximately 500 kDa(Streit et al., 1990). Therefore, to identify a possible Lex-carrier glycoprotein from NT2N neurons, cellular extractswere analyzed by Western blot using the monoclonal anti-Lex antibody L5 during differentiation. Most striking wasthe finding of a Lex-containing glycoprotein at a molecularmass higher than 206 kDa, present in NT2N neurons butabsent from NT2� cells (Fig. 5, lanes 1 and 8, respec-tively). This protein was associated endogenously withNT2N neurons and was not internalized from the FBSbecause it was still detected by Western blot in neuronscultured in serum-depleted medium (results not shown).Furthermore, it was not internalized from the Matrigelused to coat the plates because the carrier was alreadydetected in extracts of cells differentiated with RA and inMI for 10 days (Fig. 5, lane 7). Finally, it was absent fromWestern blot analysis of Matrigel extracts (results notshown). A less abundant glycoprotein was also detected inNT2N neurons but not in NT2� cells that had an appa-rent molecular mass of approximately 175 kDa. OtherLex-containing glycoproteins were detected transiently atintermediary differentiation stages at 10–17 days after RAaddition (Fig. 5, lanes 3,4). At these stages, higher levels ofFUT9 were also observed, as shown above (Fig. 4). It isprobable that the higher FUT9 levels would underlie thehigher Lex-carrier glycoprotein detection observed byWestern blot. Lex-containing-mucins extracted from a

gastric mucosa adjacent to a gastric carcinoma were used aspositive control (Fig. 5, lane C).

Lex Determinant Is Localized at the PM and in theTI-VAMP Compartment of NT2N Neurons

We further analyzed Lex subcellular localization inNT2N neurons by immunofluorescence microscopy usingthe L5 antibody. Lex was found to be present at the PM ofnon-permeabilized NT2N neurons at the cell bodies aswell as at the neurites (Fig. 6A). The epitope presented adiscrete distribution, being accumulated in clusters at themembrane. These did not correspond to labeled intracel-lular vesicles because in vivo incubation of NT2N neuronswith L5 antibody at 48C for 1 hr yielded the same labelingpattern (results not shown).

To identify the intracellular vesicles containing Lex

identified above (Fig. 1), we carried out colocalizationanalysis with markers of the exocytic and endocytic path-ways by confocal immunofluorescence microscopy (Fig.6B). Most striking was the finding that Lex colocalizedwith the v-SNARE tetanus neurotoxin insensitive vesicle-associated membrane protein (TI-VAMP) or VAMP7.TI-VAMP is a v-SNARE with a broad neuronal expres-sion, proposed to mediate a specific exocytic pathway(Proux-Gillardeaux et al., 2005). In neuronal cells, the TI-VAMP defines an exocytic pathway involved in neuriteoutgrowth via late endosome/lysosome compartment la-beled by CD63, but not via early recycling endosomescontaining the transferrin receptor or EEA-1 (Coco et al.,1999). It has been shown recently that there was an exten-sive colocalization between TI-VAMP and lysosomal-associated membrane protein 1 (LAMP-1)-positive lateendosomes/lysosomes in developing primary mouse neu-rons from sympathetic cervical ganglia (SCGs) (Arantesand Andrews, 2006). We observed that in human NT2Nneurons there was also colocalization between TI-VAMP

Fig. 5. Immunodetection by Western blot of Lex-carrier proteins alongNT2N neurons differentiation induced by retinoic acid and of LAMP-1 in NT2N neurons. Total protein from 1 3 106 cells of each differen-tiation step was applied by lane (1– 8 as indicated in Fig. 2 legend).Detection was carried out using the anti-Lex monoclonal antibody L5.Mucins containing Lex extracted from a gastric mucosa adjacent to agastric carcinoma were used as positive control (lane C). Total protein

from 1 3 105 NT2N neurons was applied for detection with the anti-LAMP-1 monoclonal antibody H4A3 (lane L). As secondary antibodiespolyclonal anti-mouse IgM and goat anti-mouse IgG HRP-conjugatedwere used. Detection was carried out by the ECL+ method. Experi-ments were carried out three times from at least two independent pro-tein extractions of each differentiation step.



and LAMP-1, both in cell bodies and in neurites (Fig. 6 B),although there was not a complete overlap, contrary tothat found in SGC neurons (Arantes and Andrews, 2006).This may arise from differences between species, type ofneurons, or differences in the development stage analyzed.Lex colocalized extensively with LAMP-1, but not withthe early endosomal marker early endosomal antigen-1(EEA-1) (Fig. 6B). LAMP-1 is a lysosomal highly glyco-sylated protein (Carlsson et al., 1988), therefore, we inves-

tigated if it would be the carrier of the Lex epitope inNT2N neurons. This was not the case, however, becauseLAMP-1 had an apparent molecular weight of approxi-mately 114 kDa as detected in duplicate by SDS-PAGEand Western blot analysis (Fig. 5, lane 8), which is lowerthan the masses of the two bands detected with the L5antibody. Furthermore, immunoprecipitated LAMP-1was not recognized by the L5 antibody, corroborating thatin NT2N human neurons LAMP-1 was not a Lex-carrier

Fig. 6. Localization of Lex in NT2Nneurons by immunofluorescence confo-cal microscopy. A: NT2N neuronswithout permeabilization were probedwith the antibody anti-Lex L5. As sec-ondary antibody a goat anti-rat IgM-TRITC was used. B: Colocalizationanalysis of Lex with markers of the endo-cytic and exocytic pathways and ofLAMP-1 with TI-VAMP. C: Colocali-zation analysis of Lex with axon and den-drite markers. B,C: PermeabilizedNT2N neurons were probed with thefollowing antibodies: monoclonal anti-Lex L5 and monoclonal anti-GM130,polyclonal anti-TGN-46, monoclo-nals anti-EEA-1, anti-LAMP-1, anti-TI-VAMP, anti-hypophosphorylated me-dium molecular weight neurofilamentprotein (NF-M P�), or anti-hyperphos-phorylated medium molecular weightneurofilament protein (NF-M P+). Sec-ondary antibodies were: goat anti-mouseIgG-AlexaFluor594, goat anti-mouseIgG-AlexaFluor488, goat anti-rat IgG-AlexaFluor488, goat anti-rat IgM-TRITC, and goat anti-sheep IgG-AlexaFluor488. Maximum intensityz-projections of 20 optical sections areshown in (A) and (C) and single opticalsections in (B). The experiments werecarried out at least three times and ineach at least five neurons per markerwere analyzed by confocal microscopy.Scale bar ¼ 10 lm.

1266 Brito et al.


(results not shown). The staining observed in the lyso-somes was not due to protein uptake from the serum orfrom the Matrigel coat because the staining with L5 wasalso observed in NT2N neurons cultured in serum-depleted medium for 24 hr and in poly-d-lysine coatedplates.

Lex showed only minor colocalization with theGolgi protein GM130 and the trans-Golgi network(TGN) marker TGN-46 (Fig. 6B).

NT2N neurons possess identifiable dendrites andaxons (Pleasure et al., 1992). Using monoclonal antibodiesthat recognize specifically the hypophosphorylated (RMdO.20)and highly phosphorylated (HO14.4) forms of mediummolecular weight neurofilament protein (NF-M), exclu-sively found on somatodendritic and axonal domainsrespectively, it is possible to distinguish between dendritesand axons. Lex was found on neurites marked with bothantibodies (Fig. 6C). These results suggested that the Lex

Fig. 7. Effect of anti-Lex antibody L5on the adhesion and initiation of neuriteoutgrowth from NT2N neurons. A:NT2N neurons in culture for 4 dayswere incubated with 30 nM anti-Lex

antibody L5; 30 nM of a non-relatedIgM (ctl IgM) was used as control. Thesame areas of the cultures were moni-tored at the time of the antibody addi-tion (0 hr) and 24 hr later (24 hr). Theassay was carried out in one (ctl IgM) orthree independent experiments in dupli-cate. Scale bar ¼ 50 lm. B: Quantifica-tion of NT2N neurons with neuritesin the presence of increasing L5 concen-trations. L5 was added to the cultures atthe time of plating and PBS was used ascontrol. The percentage of neurons withneurites was quantified 24 hr later. Dataare mean 6 SD of at least 150 neuronscounted in 10 equivalent optical fieldsfrom one (30 nM ctl IgM), two (10 and30 nM L5), and three (ctl and 15 nML5) independent experiments in dupli-cate. *Indicate significant difference(P < 0.001) from 30 nM ctl IgM meanby a one-way ANOVA analysis with aTukey’s post-hoc multiple comparisontest.



participated in cell events common to dendrites and axons(e.g., neurite initiation and elongation of outgrowth).

Anti-Lex Antibody Impair Initiation of NeuriteOutgrowth and Adhesion of NT2N Neurons

To investigate the biologic role of Lex in NT2Nneurons, we have studied the effect of anti-Lex monoclo-nal antibody L5 on the neurons. We have used a non-related IgM (ctl IgM) to rule out unspecific effects causedby the IgM size. In cultures that were allowed to adhereand form neurites before the incubation with anti-Lex

antibody, we observed that attachment of neurons andtheir neurites to the surface was significantly impairedwith 30 nM of antibody, when compared to incubationwith 30 nM of control IgM (Fig. 7A, arrows). At 15 nML5, the effect was already visible but was less evident(results not shown). Furthermore, it seemed that traces ofneurite remaining after detachment or retraction wereobserved (Fig. 7A, arrowheads). When incubations wereprolonged for 48 hr, neurons were detached from the sur-face at 10 nM L5 (results not shown). These results indi-cated that the L5 antibody inhibited NT2N cell adhesionto the plates.

To investigate the effect of L5 on neurite outgrowthincubation in the presence of the antibody at the time ofplating was carried out. Twenty-four hours later weobserved that a lower number of NT2N neurons devel-oped neurites on L5 addition. The effect was dose-de-pendent because increasing L5 concentration from 0 nMto 30 nM caused a decrease of the percentage of cells withneurites, from 796 6% to 436 8% (Fig. 7B). There wereno statistically significant differences between controlwithout antibody and incubation with 30 nM of controlIgM (Fig. 7B). Furthermore, differences between incuba-tions with 30 nM of control IgM and with 15 or 30 nML5 were statistically different (P < 0.001) as evaluated by aone-way ANOVA analysis with a Tukey’s post-hoc multi-ple comparison test. In the presence of L5 the cell bodiespresented a round morphology. We have also measuredneurite length in both experiments. In average, 24 hr afterplating, neurites were 51 lm long in the presence (51 636 lm) or absence of L5 (51 6 35 lm). These resultsshowed that L5 prevented the initiation of neurite out-growth in NT2N cells but not its elongation.

DISCUSSION

We have shown that during neuronal differentiationof human NT2N neurons from teratocarcinoma NT2�

cells there was an increase in the levels of the cell adhesioncarbohydrate Lex determinant, concomitantly to FUT9detection. The results obtained from RT-PCR analysis,FUT activities toward several acceptor substrates, andWestern blot analysis indicated clearly that FUT9 was theFUT responsible for the synthesis of Lex in NT2N neu-rons. FUT9 has also been shown to be required for thesynthesis of Lex in mouse brain (Nishihara et al., 2003;Kudo et al., 2006) and to be increasingly expressed duringneuronal RA induced differentiation of PC19 murine EC

cells (Osanai et al., 2001). In addition, FUT9 is regulatedby the transcription factor Pax6 in murine brain (Shimodaet al., 2002). Because Pax6 was increased during RA-induced differentiation of NT2N neurons (Przyborskiet al., 2003), it could be assumed that it also upregulatesFUT9 in NT2N neurons, thus supporting our conclusion.

We found high levels of FUT9 in cells treated withRA for 10 and 17 days. Accordingly, at this stage higherlevels of Lex-containing glycoproteins were also detected(Fig. 3, lanes 3,4). These results suggested that the expres-sion of FUT9 was induced transiently during RA-induceddifferentiation and that it catalyzed the synthesis of Lex invivo, the glycoproteins detected constituting the endoge-nous acceptor substrates. Because, at that stage, NT2 cellsare in a phase of differentiation where committed precur-sor cells began to leave the cell cycle and follow a programof neuronal cell differentiation, as monitored by the tran-scription of neuroD1 gene (Przyborski et al., 2000), it isconceivable that the expression of FUT9 and Lex in thesame phase would have a correlation with such initialevents from neuronal differentiation.

The FUT9 sequence is highly conserved betweenspecies (Kaneko et al., 1999; Baboval et al., 2000), indicat-ing that it has been under strong selective pressure duringevolution. Even if the FUT9�/� mouse develops well andis fertile (Kudo et al., 2004), recent evidence has shownthe disappearance of Lex from its brain concomitant toincreased anxiety-like behaviors. FUT9 seems to play arole in the functional regulation of interneurons from ba-solateral amygdala as detected by a decreased number ofcalbindin-positive neurons in this region of the brain(Kudo et al., 2006).

Regarding cellular localization, we have shown thatLex was predominantly localized on the PM, and intracell-ularly, in the TI-VAMP compartment and in lysosomes,both LAMP-1-containing intracellular compartments.The TI-VAMP compartment has been proposed to definean exocytic pathway involved in neurite outgrowth, basedon the observations that expression of TI-VAMP mutants(Martinez-Arca et al., 2001) and TI-VAMP silencing inneuronal cells (Alberts et al., 2003) impaired neurite out-growth. The exact nature and molecular composition ofthis pathway is not yet clarified. In mouse sympatheticneurons it has been shown that it occurs via late endo-somes/lysosomes immunoreactive for both TI-VAMP andLAMP-1 (Arantes and Andrews, 2006). We have detectedin human NT2N neurons a subset of vesicular structureslabeled with TI-VAMP and LAMP-1, that possibly corre-spond to this pathway. Until now, two proteins are knownto be associated with the TI-VAMP pathway in neurons:L1, a cell–cell adhesion molecule involved in axonalgrowth, whose surface expression and recycling wasshown to be TI-VAMP mediated (Alberts et al., 2003);and synaptotagmin VII, a Ca2+-sensing synaptotagmin(Arantes and Andrews, 2006). A possible candidate ofanother cargo molecule for this pathway could be the highmolecular mass Lex-carrier glycoprotein that we haveidentified (Fig. 5), which was distinct from L1 because itwas not detected by Western blot with monoclonal anti-

1268 Brito et al.


body L1-11A (results not shown) that recognizes L1(Stoeck et al., 2006), and also distinct from synaptotagminVII, which has a molecular mass of 45.6 kDa. We cannot,however, exclude the possibility that by immunofluores-cence microscopy, the L5 antibody detected moleculesdistinct from the Lex-carrier glycoprotein, such as glyco-lipids.

With respect to the possible biologic role of Lex, wehave found that the anti-Lex L5 antibody decreased the ad-hesion of neurons and neurites already formed to the sur-face (Fig. 7A), and also caused a decrease in the initiationof outgrowing neurites (Fig. 7B) in human NT2N neu-rons. It did not, however, inhibit neurite elongation ofoutgrowth contrary to that reported by others for explantXenopus brain cultures (Yoshida-Noro et al., 1999). It ispossible that incubation with the anti-Lex antibody wouldblock interactions between the neurons and matrix com-ponents essential for adhesion and neurite outgrowth.Alternatively, it could be internalized via the TI-VAMPpathway, negatively affecting the intracellular interactionsthat regulate neurite outgrowth. Others also observed thatanti-Lex antibody L5 inhibited process formation of astro-cytes and granule cell migration in mouse cerebellarexplant cultures (Streit et al., 1993). Further studies arerequired to clarify the function at the molecular level ofLex in human NT2N neurons.

In summary, we found that the adhesion carbohy-drate Lex determinant was upregulated during differentia-tion from human NT2� cells to NT2N neurons, where itwas synthesized by FUT9. Lex was localized on the PMand in the exocytic TI-VAMP compartment and it playedan important role in neuronal cell adhesion and neuriteinitiation of outgrowth.

ACKNOWLEDGMENTS

We gratefully acknowledge Prof. A. Streit (King’sCollege, UK) and Prof. T. Feizi (Imperial College, UK) forthe generous gift of L5 antibody and for useful comments,Prof. S. Hakomori (Pacific Northwest Research Institute)for SH1 antibody and Dr. T. Galli (Institut Jacques Monod,France) for the anti-TI-VAMP antibody. The MC-480(anti-SSEA-1) antibody, developed by Dr. D. Solter andDr. B.B. Knowles, was obtained from the DevelopmentalStudies Hybridoma Bank developed under the auspices ofthe NICHD and maintained by the University of Iowa,Department of Biological Sciences. We thank C. Page(CNDR, University of Pennsylvania School of Medicine)for transmitting her expertise on NT2 cells. We thank Dr.H. Conradt (GlycoThera, Braunschweig, Germany) forcritical reading of the manuscript. C.B. was a recipient of aPhD fellowship from FCT.

REFERENCES

Abe K, McKibbin JM, Hakomori S. 1983. The monoclonal antibody

directed to difucosylated type 2 chain (Fuc alpha 1 leads to 2Gal beta 1

leads to 4[Fuc alpha 1 leads to 3]GlcNAc; Y Determinant). J Biol Chem

258:11793–11797.

Alberts P, Rudge R, Hinners I, Muzerelle A, Martinez-Arca S, Irinopoulou

T, Marthiens V, Tooze S, Rathjen F, Gaspar P, Galli T. 2003. Cross talk

between tetanus neurotoxin-insensitive vesicle-associated membrane pro-

tein-mediated transport and L1-mediated adhesion. Mol Biol Cell 14:

4207–4220.

Arantes RM, Andrews NW. 2006. A role for synaptotagmin VII-regulated

exocytosis of lysosomes in neurite outgrowth from primary sympathetic

neurons. J Neurosci 26:4630–4637.

Baboval T, Henion T, Kinnally E, Smith FI. 2000. Molecular cloning of rat

alpha1,3-fucosyltransferase IX (Fuc-TIX) and comparison of the expres-

sion of Fuc-TIV and Fuc-TIX genes during rat postnatal cerebellum de-

velopment. J Neurosci Res 62:206–215.

Baboval T, Smith FI. 2002. Comparison of human and mouse Fuc-TX and

Fuc-TXI genes, and expression studies in the mouse. Mamm Genome

13:538–541.

Becker DJ, Lowe JB. 2003. Fucose: biosynthesis and biological function in

mammals. Glycobiology 13:41R–53R.

Cailleau-Thomas A, Coullin P, Candelier JJ, Balanzino L, Mennesson B,

Oriol R, Mollicone R. 2000. FUT4 and FUT9 genes are expressed early

in human embryogenesis. Glycobiology 10:789–802.

Capela A, Temple S. 2002. LeX/ssea-1 is expressed by adult mouse CNS

stem cells, identifying them as nonependymal. Neuron 35:865–875.

Carlsson SR, Roth J, Piller F, Fukuda M. 1988. Isolation and characteriza-

tion of human lysosomal membrane glycoproteins, h-lamp-1 and h-lamp-

2. Major sialoglycoproteins carrying polylactosaminoglycan. J Biol Chem

263:18911–18919.

Chen YJ, Wing DR, Guile GR, Dwek RA, Harvey DJ, Zamze S. 1998.

Neutral N-glycans in adult rat brain tissue—complete characterization reveals

fucosylated hybrid and complex structures. Eur J Biochem 251:691–703.

Coco S, Raposo G, Martinez S, Fontaine JJ, Takamori S, Zahraoui A, Jahn

R, Matteoli M, Louvard D, Galli T. 1999. Subcellular localization of teta-

nus neurotoxin-insensitive vesicle-associated membrane protein (VAMP)/

VAMP7 in neuronal cells: evidence for a novel membrane compartment.

J Neurosci 19:9803–9812.

Costa J, Grabenhorst E, Nimtz M, Conradt HS. 1997. Stable expression of

the Golgi form and secretory variants of human fucosyltransferase III from

BHK-21 cells. Purification and characterization of an engineered trun-

cated form from the culture medium. J Biol Chem 272:11613–11621.

Dasgupta S, Hogan EL, Spicer SS. 1996. Stage-specific expression of fuco-

neolacto- (Lewis X) and ganglio-series neutral glycosphingolipids during

brain development: characterization of Lewis X and related glycosphingo-

lipids in bovine, human and rat brain. Glycoconj J 13:367–375.

Escrevente C, Machado E, Brito C, Reis CA, Stoeck A, Runz S, Marme

A, Altevogt P, Costa J. 2006. Different expression levels of alpha3/4 fuco-

syltransferases and Lewis determinants in ovarian carcinoma tissues and

cell lines. Int J Oncol 29:557–566.

Fenderson BA, Andrews PW, Nudelman E, Clausen H, Hakomori S.

1987. Glycolipid core structure switching from globo- to lacto- and gan-

glio-series during retinoic acid-induced differentiation of TERA-2-

derived human embryonal carcinoma cells. Dev Biol 122:21–34.

Goelz SE, Hession C, Goff D, Griffiths B, Tizard R, Newman B, Chi-

Rosso G, Lobb R. 1990. ELFT: a gene that directs the expression of an

ELAM-1 ligand. Cell 63:1349–1356.

Gotz M, Wizenmann A, Reinhardt S, Lumsden A, Price J. 1996. Selective

adhesion of cells from different telencephalic regions. Neuron 16:551–

564.

Grabenhorst E, Nimtz M, Costa J, Conradt HS. 1998. In vivo specificity of

human alpha1,3/4-fucosyltransferases III-VII in the biosynthesis of Lew-

isX and Sialyl LewisX motifs on complex-type N-glycans. Coexpression

studies from bhk-21 cells together with human beta-trace protein. J Biol

Chem 273:30985–30994.

Kaneko M, Kudo T, Iwasaki H, Ikehara Y, Nishihara S, Nakagawa S,

Sasaki K, Shiina T, Inoko H, Saitou N, Narimatsu H. 1999. Alpha1,3-

fucosyltransferase IX (Fuc-TIX) is very highly conserved between human

and mouse; molecular cloning, characterization and tissue distribution of

human Fuc-TIX. FEBS Lett 452:237–242.



Klimpel KR, GoldmanWE. 1988. Cell walls from avirulent variants of Histo-

plasma capsulatum lack alpha-(1,3)-glucan. Infect Immun 56:2997–3000.

Kudo T, Fujii T, Ikegami S, Inokuchi K, Takayama Y, Ikehara Y, Nishi-

hara S, Togayachi A, Takahashi S, Tachibana K, Yuasa S, Narimatsu H.

2006. Mice lacking {alpha}1,3-fucosyltransferase IX demonstrate disap-

pearance of Lewis x structure in brain and increased anxiety-like behav-

iors. Glycobiology 5:397–403.

Kudo T, Kaneko M, Iwasaki H, Togayachi A, Nishihara S, Abe K, Nari-

matsu H. 2004. Normal embryonic and germ cell development in mice

lacking alpha 1,3-fucosyltransferase IX (Fut9) which show disappearance

of stage-specific embryonic antigen 1. Mol Cell Biol 24:4221–4228.

Mai JK, Krajewski S, Reifenberger G, Genderski B, Lensing-Hohn S, Ash-

well KW. 1999. Spatiotemporal expression gradients of the carbohydrate

antigen (CD15) (Lewis X) during development of the human basal gan-

glia. Neuroscience 88:847–858.

Martinez-Arca S, Coco S, Mainguy G, Schenk U, Alberts P, Bouille P,

Mezzina M, Prochiantz A, Matteoli M, Louvard D, Galli T. 2001. A

common exocytotic mechanism mediates axonal and dendritic out-

growth. J Neurosci 21:3830–3838.

Meijering E, Jacob M, Sarria JC, Steiner P, Hirling H, Unser M. 2004.

Design and validation of a tool for neurite tracing and analysis in fluores-

cence microscopy images. Cytometry A 58:167–176.

Morais VA, Serpa J, Palma AS, Costa T, Maranga L, Costa J. 2001. Expres-

sion and characterization of recombinant human alpha-3/4-fucosyltrans-

ferase III from Spodoptera frugiperda (Sf9) and Trichoplusia ni (Tn) cells

using the baculovirus expression system. Biochem J 353:719–725.

Nishihara S, Iwasaki H, Nakajima K, Togayachi A, Ikehara Y, Kudo T,

Kushi Y, Furuya A, Shitara K, Narimatsu H. 2003. Alpha1,3-fucosyltrans-

ferase IX (Fut9) determines Lewis X expression in brain. Glycobiology

13:445–455.

Osanai T, Chai W, Tajima Y, Shimoda Y, Sanai Y, Yuen CT. 2001.

Expression of glycoconjugates bearing the Lewis X epitope during neural

differentiation of P19 EC cells. FEBS Lett 488:23–28.

Pleasure SJ, Page C, Lee VM. 1992. Pure, postmitotic, polarized human neu-

rons derived from NTera 2 cells provide a system for expressing exogenous

proteins in terminally differentiated neurons. J Neurosci 12:1802–1815.

Proux-Gillardeaux V, Rudge R, Galli T. 2005. The tetanus neurotoxin-

sensitive and insensitive routes to and from the plasma membrane: fast and

slow pathways? Traffic 6:366–373.

Przyborski SA, Morton IE, Wood A, Andrews PW. 2000. Developmental

regulation of neurogenesis in the pluripotent human embryonal carci-

noma cell line NTERA-2. Eur J Neurosci 12:3521–3528.

Przyborski SA, Smith S, Wood A. 2003. Transcriptional profiling of neuro-

nal differentiation by human embryonal carcinoma stem cells in vitro.

Stem Cells 21:459–471.

Reis CA, David L, Nielsen PA, Clausen H, Mirgorodskaya K, Roepstorff

P, Sobrinho-Simoes M. 1997. Immunohistochemical study of MUC5AC

expression in human gastric carcinomas using a novel monoclonal anti-

body. Int J Cancer 74:112–121.

Roos C, Kolmer M, Mattila P, Renkonen R. 2002. Composition of Dro-

sophila melanogaster proteome involved in fucosylated glycan metabo-

lism. J Biol Chem 277:3168–3175.

Sandhu JK, Pandey S, Ribecco-Lutkiewicz M, Monette R, Borowy-

Borowski H, Walker PR, Sikorska M. 2003. Molecular mechanisms of

glutamate neurotoxicity in mixed cultures of NT2-derived neurons and

astrocytes: protective effects of coenzyme Q10. J Neurosci Res 72:691–

703.

Shimoda Y, Tajima Y, Osanai T, Katsume A, Kohara M, Kudo T, Nari-

matsu H, Takashima N, Ishii Y, Nakamura S, Osumi N, Sanai Y. 2002.

Pax6 controls the expression of Lewis x epitope in the embryonic fore-

brain by regulating alpha 1,3-fucosyltransferase IX expression. J Biol

Chem 277:2033–2039.

Singhal AK, Orntoft TF, Nudelman E, Nance S, Schibig L, Stroud MR,

Clausen H, Hakomori S. 1990. Profiles of Lewisx-containing glycopro-

teins and glycolipids in sera of patients with adenocarcinoma. Cancer Res

50:1375–1380.

Solter D, Knowles BB. 1978. Monoclonal antibody defining a stage-specific

mouse embryonic antigen (SSEA-1). Proc Natl Acad Sci U S A 75:5565–

5569.

Stoeck A, Keller S, Riedle S, Sanderson MP, Runz S, Le Naour F, Gut-

wein P, Ludwig A, Rubinstein E, Altevogt P. 2006. A role for exosomes

in the constitutive and stimulus-induced ectodomain cleavage of L1 and

CD44. Biochem J 393:609–618.

Streit A, Nolte C, Rasony T, Schachner M. 1993. Interaction of astrochon-

drin with extracellular matrix components and its involvement in astro-

cyte process formation and cerebellar granule cell migration. J Cell Biol

120:799–814.

Streit A, Yuen CT, Loveless RW, Lawson AM, Finne J, Schmitz B, Feizi

T, Stern CD. 1996. The Le(x) carbohydrate sequence is recognized by

antibody to L5, a functional antigen in early neural development. J Neu-

rochem 66:834–844.

Trojanowski JQ, Kelsten ML, Lee VM. 1989. Phosphate-dependent and

independent neurofilament protein epitopes are expressed throughout the

cell cycle in human medulloblastoma (D283 MED) cells. Am J Pathol

135:747–758.

Weston BW, Nair RP, Larsen RD, Lowe JB. 1992a. Isolation of a novel

human alpha (1,3)fucosyltransferase gene and molecular comparison to

the human Lewis blood group alpha (1,3/1,4)fucosyltransferase gene.

Syntenic, homologous, nonallelic genes encoding enzymes with distinct

acceptor substrate specificities. J Biol Chem 267:4152–4160.

Weston BW, Smith PL, Kelly RJ, Lowe JB. 1992b. Molecular cloning of a

fourth member of a human alpha (1,3)fucosyltransferase gene family. Mul-

tiple homologous sequences that determine expression of the Lewis x,

sialyl Lewis x, and difucosyl sialyl Lewis x epitopes. J Biol Chem 267:

24575–24584.

Yoshida-Noro C, Heasman J, Goldstone K, Vickers L, Wylie C. 1999.

Expression of the Lewis group carbohydrate antigens during Xenopus

development. Glycobiology 9:1323–1330.

1270 Brito et al.


increased levels of fucosyltransferase ix and carbohydrate lewisx adhesion determinant in human nt2n...

Documents