J Biol Chem, Vol. 275, Issue 11, 8007-8015, March 17, 2000

Molecular Cloning of Mouse Ganglioside Sialidase and Its Increased Expression in Neuro2a Cell Differentiation^*

Takafumi Hasegawa^§, Kazunori Yamaguchi, Tadashi Wada, Atsushi Takeda^§, Yasuto Itoyama^§, and Taeko Miyagi^¶

From the  Division of Biochemistry, Research Institute, Miyagi Prefectural Cancer Center, 47-1 Nodayama, Medeshima-shiode, Natori, Miyagi 981-1293 and the ^§ Department of Neurology, Tohoku University School of Medicine, Sendai, Miyagi 980-8574, Japan

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Ganglioside sialidases have been implicated in neuronal differentiation processes, including neurite outgrowth. To understand further the roles and regulation mechanisms of the sialidase in neuronal systems, we have cloned mouse ganglioside sialidase cDNA and observed its expression in Neuro2a cell differentiation. A 3339-base pair cDNA, cloned based on the sequence information of previously cloned enzymes, encodes 418 amino acids containing three Asp boxes characteristic of sialidases. Northern blot analysis revealed a 3.4-kilobase transcript expressed highly in heart but also in several other tissues including brain. In situ hybridization of mouse brain demonstrated the mRNA to be present in the cerebral cortex, as well as in the granule cell layer, Purkinje cells, and deep cerebellar nucleus of the cerebellum. Transient expression of the cDNA in COS-1 cells resulted in over 300-fold increase in sialidase activity toward gangliosides compared with the control level, with a preference for ganglioside substrate. During 5-bromodeoxyuridine-induced Neuro2a cell differentiation, the expression of the sialidase was increased as assessed by activity assays and quantitative reverse transcription-polymerase chain reaction analyses. Stable transfection of the sialidase in Neuro2a cells resulted in accelerated neurite arborization following 5-bromodeoxyuridine treatment, indicating the direct participation of this ganglioside sialidase in neuronal cell differentiation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In recent years, great interest has focused on the elucidation of the biological function of glycosphingolipids, in particular gangliosides, which are characteristic constituents at the surface of mammalian cells and abundant in neuronal membranes (1, 2). In the nervous system, gangliosides have been observed to undergo metabolic alterations during development and have been implicated in various biological phenomena, including neuritogenesis (3, 4), synaptic function (5), neuronal death (6), neural repair (7), cell-cell recognition, and cell adhesion (8). Ganglioside accretion is largest during dendritic arborization and synaptogenesis and is accompanied by marked changes in the ganglioside composition, which then persist during adult life (9, 10). Neuroblastoma cell lines can be triggered to undergo neuronal differentiation with a marked alteration of ganglioside patterns by several pharmacological agents (11-13). On the other hand, a number of exogenous gangliosides are known to cause neuritogenesis, including ganglioside GM1 (14), enhances the effects of neurotrophic factors, such as nerve growth factor (15, 16), in several cell lines. Tsuji et al. (17) reported that very low amounts of exogenous GQ1b specifically stimulated neurite outgrowth in a human neuroblastoma cell line, GOTO cells. Although the general mechanism by which gangliosides promote neuritogenesis has not been elucidated, their desialylation by endogenous as well as exogenous sialidases has been suggested as a critical event for neuronal differentiation (18-21), myelination (22), synaptogenesis, and synaptic function (23). A potent ganglioside sialidase activity, in fact, is located in the neuronal membranes (24), with a relative enrichment in the synaptosomal fraction as compared with the neuronal perikarya (23, 25). In this context, it is necessary to clarify the expression mechanisms of endogenous sialidase responsible for ganglioside degradation in the nervous system. In the present study, we obtained new insights into the role of the ganglioside sialidase during neuronal differentiation by studies with a newly cloned cDNA.

Mammalian sialidases, catalyzing the first step in glycoconjugate degradation, have been classified based on their subcellular distribution into at least four types: cytosolic, lysosomal matrix, lysosomal membrane, and plasma membrane (26-28). They differ from each other not only in subcellular location but also catalytic and immunological properties. Membrane-associated sialidases hydrolyze gangliosides preferentially, and those in plasma membrane are distinct from lysosomal membrane sialidase in acting specifically on gangliosides residing in the same membrane (28, 29).We previously cloned a cytosolic sialidase from rat skeletal muscle (30), and recently membrane ganglioside sialidases of bovine (31) and human (32) brains. Lysosomal sialidases of human (33-35) and mouse (36-38) were also cloned by another three groups. We have now isolated a cDNA encoding mouse ganglioside sialidase capitalizing on the high degree of conservation in the primary structures of bovine plasma membrane (31), rat cytosolic (30), and human lysosomal (33-35) sialidases. The spatial and temporal expression of the sialidase mRNA was investigated in adult mouse brain and during differentiation of Neuro2a cells induced by 5-bromodeoxyuridine (BrdUrd).¹ The results revealed the ganglioside sialidase to be widely expressed in the central nervous system and to play a positive regulatory role in differentiation of neuronal cells.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

cDNA Cloning of a Mouse Ganglioside Sialidase-- Possible mouse ganglioside sialidase cDNA fragments were first amplified by PCR using oligonucleotide primers, based on the amino acid sequences of previously cloned sialidases. One pair of the primers, 1S (5'-GGACACCGGACCATGAACCCCTGTCCT-3') (sense) and 2A (5'-CCTGGCCCCACAGCAAAAGTGGCCCA-3') (antisense), and degenerate primers 3S (5'-AC(C/T)TT(C/T)CT(G/C)GCCTT(C/T)GC(A/T)GAG-3') and 4A (5' CTG(A/G)AT(A/G) CCAI I(A/G)CC(C/T)GG(G/T)CCIII(A/G)GCAAA-3') were designed from sequences in regions in which the amino acids of the bovine membrane sialidase are identical to those of rat cytosolic and of human lysosomal sialidases, respectively. The first strand cDNAs were synthesized from poly(A)⁺ RNA from BALB/c mouse brain, prepared by the acid guanidium-phenol-chloroform extraction procedure (39), followed by oligo(dT) column chromatography, using random primers and murine leukemia virus reverse transcriptase (Superscript RNase H, Life Technologies, Inc.). The cDNA was then used as a template for PCR under the following conditions: initial denaturation at 95 °C for 5 min followed by 40 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 1 min, and extension at 72 °C for 2 min, with final elongation at 72 °C for 10 min. Appropriate PCR products were gel purified and blunt end-ligated to the SmaI site of a Bluescript vector. At this step, three positive fragments (MBI, MBIIA, and MBIIB) were obtained.

To obtain a complete cDNA, the 3' and 5' ends of the mouse membrane sialidase cDNA were extended according to the rapid amplification of cDNA ends (RACE) methods described by Frohman et al. (40) with slight modifications. The primer 5A (5'-ATCAGCCCTCGTCTGAGCACCA-3'), derived from the 5' end sequence obtained from the degenerate PCR product MBIIA, was used in the reverse transcription of the 5' end of the mRNA from mouse brain poly(A)⁺ RNA. After reverse transcription, the RNA was degraded by incubation for 15 min at 37 °C with 2 units of RNase H. The purified cDNA was polyadenylated at its 3' end by incubation with 10 units of terminal deoxynucleotidyl transferase in the presence of 6 mM ATP at 37 °C for 10 min. The mixture was then heated to 65 °C to denature the enzyme, and the polyadenylated single-strand cDNA (one-tenth of the aliquot of the cDNA pool synthesized above) was applied directly in the first PCR amplification. The poly(A) tail generated was used to initially amplify the cDNA with the (dT)₁₇ adaptor primer (5'-GACTCGAGTCGACATCGATTTTTTTTTTTTTTTTT-3'), followed by further amplification with the adaptor primer (5'-GACTCGAGTCGACATCG-3') and the specific primers 6A (5'-AGGCAGCATCCTCATCTCTGAC-3') and 7A (5'-TCATCTCTGACTGAGGTCCGCT-3'), deduced from the MBIIA fragment. In the first amplification, the cDNA was amplified using 37.5 pmol of (dT)₁₇ adaptor, 12.5 pmol of adaptor, and 25 pmol of 6A primer, using the following conditions: 95 °C for 5 min, 3 cycles of 94 °C for 30 s, 50 °C for 1 min, and 72 °C for 2 min followed by 37 cycles of 94 °C for 30 s, 55 °C for 1 min, and 72 °C for 1 min with final elongation at 72 °C for 10 min. A 2-µl aliquot of the PCR product was reamplified using 25 pmol of each of the nested primers, adaptor, and 7A primer, under the same conditions except for an annealing temperature of 55 °C.

For derivation of the 3' end, a (dT)₁₇ adaptor, an adaptor, and a set of oligonucleotides, 8S (5'-TTGACCGAGGAGGTCATTGG-3') and 9S (5'-TCATTGGCTCAGAGGTGAAG-3') primers specific to the 3' end sequence of the MBIIA fragment, were used as primers. First strand cDNA was synthesized by reverse transcription of mouse brain poly(A)⁺ RNA (1 µg) using the (dT)₁₇ adaptor primer. The single-stranded cDNA generated by reverse transcription was purified from primers using a NICK column (Amersham Pharmacia Biotech). The cDNA (one-tenth of the aliquot synthesized above) as template, adaptor (12.5 pmol), and primer 8S (25 pmol) were used for amplification as described above. A 2-µl sample of the product was reamplified in 50 µl with adaptor primer (25 pmol) and a nested specific primer (9S, 25 pmol).

cDNA Cloning of a Mouse Cytosolic Sialidase-- Mouse cytosolic sialidase was cloned based on the primary structure of rat cytosolic sialidase by RT-PCR and RACE methods in a manner similar to that described above. Degenerate primers (sense, 5'-GAAGTCTATGCTTAC(A/C)GIA(T/C)CC-3'; antisense, 5'-AGCTTCCGGTAGGCATAAGC-3') were used in combination with template cDNA reverse-transcribed from mouse skeletal muscle poly(A)⁺ RNA. A 513-bp PCR fragment was obtained, and its 5' and 3' sequences were extended by the RACE methods using specific primers derived from the fragment sequence: for 5'-RACE, mcSD1A (5'-TGCTGGCTCGCTTTTCCGCAAA-3'), mcSD2A (AAAGGCCAGCAGGGTCTTCTGCTT-3'), and mcSD3A (5'-TGCTTCTTCAGGTAGAGCAGAGCA-3'); for 3'-RACE, mcSD4S (5'-GTCTGCAGCTGC GGAACCCAGCT-3') and mcSD5S (5'-AGCTGGGAGCCTGCTGGTACCT-3').

DNA Sequence Analysis-- The appropriate PCR products in Bluescript were sequenced with a Model 4200L-2 DNA autosequencer (LI-COR) using a Thermo Sequenase Cycle Sequencing Kit (Amersham Pharmacia Biotech) with fluorescence-labeled M13 forward and reverse primers. Nucleotide and amino acid sequences thus obtained were compared with the nonredundant sequence data bases present at the National Center for Biotechnology Information using the BLAST network service. Multiple amino acid sequence alignment was performed using the Parallel Protein Information Analysis system (41) (Parallel Application TRC Laboratory, Real World Computing Partnership, Tsukuba, Japan). Protein sorting signals and subcellular localization were predicted using the PSORT World Wide Web server (42).

Expression of Mouse Ganglioside Sialidase cDNA in COS-1 Cells-- To construct the sialidase expression plasmid, a set of primers, 10S (5'-CCCGAATTCGCCATGGAGGAAGTCCCA-3') and 11A (5'-CCCGAATTCTTAGTCGCTACTAGGGCT-3'), both containing EcoRI restriction sites at the 5' end, was used to amplify the sequence encoding the entire open reading frame, including the Kozak sequence (43), with the template cDNA synthesized from mouse brain poly(A)⁺ RNA. After digestion with EcoRI, PCR products were ligated into the EcoRI site of the pME18S eukaryotic expression vector. The presence, orientation, and fidelity of the cDNA in the vector were confirmed by digestion of restriction enzyme and DNA sequencing. To verify the ganglioside-hydrolyzing sialidase activity, COS-1 cells (Riken Cell Bank) were transfected by electroporation with the empty expression vector or the expression vector containing the mouse ganglioside sialidase cDNA by the following procedures. Cells washed twice with phosphate-buffered saline (PBS) were resuspended in HEPES-PBS at a final concentration of 1.0 × 10⁷ cells/ml, and an aliquot of the suspension (750 µl) was added to sterile cuvettes with 40 µg of plasmid DNA. The cuvette was placed on ice for 10 min and electroporated at 240 V, 975 µF using a Bio-Rad Genepulser. The cuvette was placed on ice for an additional 10 min, and cells were cultured in 40 ml of Dulbecco's modified Eagle's medium (DMEM) supplemented with penicillin (100 IU/ml), streptomycin (100 µg/ml), and 10% heat-inactivated fetal bovine serum. After 48 h, cells were pelleted, lysed in 9 volumes of ice-cold PBS by sonication, and centrifuged at 1000 × g at 4 °C for 10 min. The supernatant (crude homogenate) was further centrifuged at 100,000 × g at 4 °C for 1 h. The resulting supernatant and pellet were used as cytosolic and membrane fractions, respectively, and were assayed for sialidase activity with various substrates that included gangliosides. Stable transfectants of the sialidase were obtained using mouse neuroblastoma Neuro2a cells (American Type Culture Collection, CCL-131), cultured in DMEM containing high glucose supplemented with penicillin, streptomycin, and 10% fetal bovine serum. After overnight culture at 2 × 10⁵ cells per 35-mm culture dish, cells at 50-60% confluence were washed with serum-free medium and then incubated with a DNA (2 µg)-PLUS (6 µl)-LipofectAMINE (8 µl) (Life Technologies, Inc.) reagent complex for 5 h. At 48 h after lipofection, selection started in growth medium containing 500 µg/ml G418 (Life Technologies, Inc.). Following 5 weeks of selection, discrete colonies were subcloned and expanded.

Sialidase Assay-- Sialidase activity was assayed using various substrates as described elsewhere (27). Briefly, the reaction mixture contained 30 nmol of substrate as bound sialic acid, 0.2 mg of bovine serum albumin, 10 µmol of sodium acetate (pH 4.6), and 0.2 mg of Triton X-100. After incubation at 37 °C for 10-30 min, released sialic acid was determined by the thiobarbituric acid method of Aminoff (44). Sialidase activity toward 4-methylumbelliferyl-neuraminic acid was assayed by spectrofluorometrical measurement of 4-methylumbelliferone released. To determine the pH profile of the activity, HEPES, 2-morpholinoethanesulfonic acid, and citrate-phosphate buffers also were used. One unit of sialidase was defined as the amount of enzyme that catalyzed the release of 1 nmol of sialic acid/h.

Analyses of Cellular Sialic Acids and Glycolipids-- Cells (10⁷) harvested at subconfluency were washed with PBS and lyophilized. The glycolipids were extracted in sequence with 5 ml of chloroform/methanol (C/M) (1:1, v/v), 2.5 ml of C/M (2:1, v/v), and 2.5 ml of C/M (1:2, v/v), and then evaporated to dryness. After desalting by dialysis, the glycolipids were again lyophilized and dissolved in a small volume of C/M (2:1, v/v), and one-tenth of the amount was chromatographed on high performance thin layer chromatographic plates (Baker, Phillipsburg, NJ) in C/M/0.5% CaCl₂ (60:40:9, v/v/v). Glycolipids were visualized with the orcinol-H₂SO₄. To determine the cellular sialic acids, the glycolipid extracts (lipid-bound sialic acid) and the residues after extraction (protein-bound sialic acid) were treated in 0.1 N H₂SO₄ at 80 °C for 1 h and assayed for sialic acids. To estimate small amounts of sialic acids, fluorometric high performance liquid chromatography with 1,2-diamino-4,5-methylene dioxybenzene was employed (45).

BrdUrd-induced Neuritogenesis in Neuro2a Cells-- Neuro2a cells were initially grown in DMEM with penicillin, streptomycin, and 5% fetal bovine serum, with cell passage every 2-3 days. They were seeded in 150-cm² sterile tissue culture flasks at a density of 2 × 10⁴/cm². After overnight incubation in the above medium, cells were exposed for variable periods (0-5 days) to 0.25 mM BrdUrd (46). The morphology of the cells was examined, and photographs were taken at × 100 magnification. To quantify neuritogenesis, five randomly chosen fields in each flask were counted, and cells bearing neurites at least 1.5-fold the length of the soma diameter were considered as differentiated. All of the cells collected were assayed for sialidase activity using bovine brain mixed gangliosides as substrates and for acetylcholinesterase activity using the spectrophotometric procedure described by Ellman et al. (47).Total RNAs were extracted for quantitative PCR by the acid guanidium-phenol-chloroform extraction procedure (39).

Preparation of Deleted cDNA and Quantitative RT-PCR-- To evaluate the mRNA level of the sialidase in variously differentiated Neuro2a cells, we adopted competitive PCR methods (48) using the deleted mouse sialidase cDNA clone as a competitor. The competitor cDNA was prepared by double-digestion of the mouse sialidase cDNA encoding the entire open reading frame in Bluescript vector with MunI and BbsI. From 5 µg of total RNA, cDNA was synthesized using 200 units of MoMuLV-RT and oligo(dT) primers in a final volume of 20 µl. 10 µl of the 25-fold-diluted solution was subjected to quantitative PCR analysis. PCR amplifications were performed with a set of primer pairs 12S (5'-GCCATGGAGGAAGTCCCA-3') and 13A (5'-TTAGTCGCTACTAGGGCT-3') under the following conditions: 1 min at 94 °C, 1 min at 54 °C, and 2 min at 72 °C for 40 cycles, preceded by 10 min at 95 °C and followed by 10 min at 72 °C, generating a 1254-bp product from the target cDNA and a 937-bp product from the competitor cDNA. The quantitative PCR analysis consisted of two steps: the first PCR for rough measurement using 10 fg of competitor DNA, and the second PCR for precise quantification based on the amount of target DNA estimated from the first PCR. To check for sample-to-sample variation, the expression level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (49) was measured for each cDNA sample using a GAPDH competitor, prepared by digestion of the 0.5-kilobase cDNA (nucleotides 566-1017) with StyI and BspMI. After amplification, 10-µl aliquots were electrophoresed in 1% agarose gels, followed by photographic recording of the gels stained with ethidium bromide. Gel photos were scanned and densitometric analyses of PCR products were performed using the image analysis program NIH Image, version 1.61 (Research Services Branch, NIMH, National Institutes of Health). Under the PCR conditions, a linear correlation between the densitometry intensity and amount of template was also confirmed using standard cDNAs.

Northern Blot Analysis-- Mouse multiple tissue Northern blots (CLONTECH) were hybridized with an [-³²P]dCTP-labeled MBIIA fragment amplified by the polymerase chain reaction, according to the manufacturer's specifications. After overnight incubation at 42 °C, membranes were washed twice in 2× SSC, 0.1% SDS at 42 °C for 30 min and used to expose imaging plates. The radioactivity was visualized with an Image Analyzer BAS2000 system (Fuji Photo Film). The tissue mRNA level was further estimated by quantitative RT-PCR methods as described above.

In Situ Hybridization-- The mouse sialidase cDNA fragment stretching from base 374 to 769 was cloned into the Bluescript vector in the correct orientation. To generate sense and antisense probes, the plasmids were then linearized by HindIII and XbaI and transcribed by T3 and T7 RNA polymerase, respectively, in the presence of digoxigenin-labeled UTP (Roche Molecular Biochemicals). Probes were fragmented to approximately 150 bp in length by alkaline treatment before use. Sagittal sections (15-µm) of brain from a 7-week-old BALB/c mouse were cut on a Micron cryostat at 20 °C and thaw-mounted onto silane coated slides. The sections were fixed in 4% paraformaldehyde plus 0.5% glutaraldehyde/PBS and incubated with 10 µg/ml proteinase K for 30 min at 37 °C. After acetylation with 0.25% acetic anhydride for 10 min in room temperature, the sections were dehydrated and prehybridized for 1 h at 48 °C in a hybridization buffer consisting of 50% deionized formamide, 4× SSC, 20% dextran sulfate, 1× Denhardt's solution, and 0.25 mg/ml yeast tRNA. They were then incubated with an antisense riboprobe (or a sense probe as a negative control) at 48 °C overnight in a humidified box followed by successive incubations in 2× SSC at 48 °C for 15 min twice, in 2× SSC with 100 µg/ml RNaseA at 37 °C for 30 min, and in 0.1× SSC at room temperature for 10 min. Positive signals were detected by immunoassay with anti-digoxigenin-alkaline phosphatase conjugate and the color substrates NBT/X-phosphate (Roche Molecular Biochemicals).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The PCR with primers designed on the basis of conserved amino acid sequences between bovine membrane and rat cytosolic sialidases yielded a specific product MBI of 240 bp showing 80% amino acid sequence identity with the bovine sialidase. PCR with degenerate primers from the sequence conserved between human membrane and human lysosomal sialidases gave 380- and 420-bp products, the deduced amino acid sequences of which were highly homologous to the lysosomal and membrane sialidases, respectively. Sequencing analyses of the 420-bp products revealed the presence of two distinct amplicon sequences, MBIIA and MBIIB. The deduced amino acid sequence of MBIIA showed 80% identity to the corresponding region of the bovine membrane sialidase and included the MBI sequence obtained as described above. On the other hand, the sequence of MBIIB was completely identical to the human membrane sialidase. Northern blot analysis by MBIIA probe revealed the presence of a hybridizing transcriptant of approximately 3.4 kilobases in mouse brain as described below, but the MBIIB probe showed a 2.4-kilobase faint band that disappeared under relatively high stringency conditions. Genomic Southern blot experiments using MBIIA and MBIIB probes demonstrated two genomic patterns indistinguishable from each other (data not shown). Although these results cannot exclude the possible presence of two types of putative membrane sialidases in mouse brain, we decided to obtain a mouse major membrane sialidase cDNA by the RACE methods based on the nucleotide sequence of MBIIA. 5'-RACE generated six positive products (312-433 bp) after two rounds of PCR amplification, the sequences of which were found to overlap with each other. A single 2646-bp product obtained by 3'-RACE contained the 3' end of MBIIA and two Asp boxes, known to be the consensus sequences for sialidases (50). These two cDNAs were used to construct a cDNA encoding the entire sequence (3339 bp) of the mouse sialidase, as shown in Fig. 1. An in-frame stop codon (TGA, nucleotides 11-13) was found located upstream of the first methionine codon (nucleotides 266-268), and the sequence surrounding the translation start site fit reasonably well with the Kozak consensus (43). The 3'-untranslated region contained the polyadenylation signal (AATAAA), 18 bases upstream from the poly(A) tail. The predicted open reading frame encodes a protein of 418 amino acids with a calculated molecular mass of 46.8 kDa. Protein and nucleotide data base homology searches revealed that the mouse sialidase has significant homology with mammalian, bacterial, and viral sialidases. Among mammalian sialidases so far cloned, the highest homology was detected with the bovine and human plasma membrane-associated sialidase (31, 32) (67.1 and 67.8%, respectively), and lower but still significant similarity was found with mouse (51), rat (30), and human (52) cytosolic sialidases (40.4, 39.4, and 39.8%, respectively). There were less significant sequence identities with human (33-35) and mouse (36-38) lysosomal sialidases (29.0 and 26.4%, respectively). A multiple alignment of the amino acid sequences of the mouse sialidase and other mammalian sialidases is shown in Fig. 2. The sequence motif (F/Y)RIP (amino acids 24-27), which is highly conserved near the N terminus of all of the sialidases described thus far, and three copies (SEDAGCSW, amino acids 131-138; SDDFGVTW, amino acids 203-210; and STDSGGCF, amino acids 254-261) of Asp boxes are present in the mouse sialidase. There is a putative hydrophobic segment at 174-194 split by an internal arginine but neither a potential site of N-glycosylation nor an affirmative sorting signal. Northern blot analysis revealed a transcript of approximately 3.4 kilobases (Fig. 3A) expressed most highly in heart but also in brain, spleen, lung, kidney, and testis. Further estimation of the mRNA level by competitive RT-PCR methods provided evidence that brain, especially cerebellum, is rich in this sialidase, as well as heart (Fig. 3B). Determination of the spatial localization of the sialidase mRNA in adult mouse brain by in situ hybridization analysis revealed positive signals mainly in the layers II-IV of cerebral cortex and the granule cell layer, Purkinje cells, and deep cerebellar nucleus of the cerebellum (Fig. 4). The signal intensity with the antisense probe in the hippocampal formation was slightly higher than with the sense probe (data not shown).

View larger version (38111K): [in this window] [in a new window]   Fig. 1.   Nucleotide and deduced amino acid sequences of the mouse ganglioside sialidase. The predicted amino acid sequence is shown by the single letter amino acid code under the nucleotide sequence. The polyadenylation signal is underlined, the Asp boxes are boxed, and the YRIP motif is indicated by a dotted line.

View larger version (77K): [in this window] [in a new window]   Fig. 2.   Amino acid sequence alignment of mammalian sialidases. Alignment of the entire amino acid sequences of the mouse ganglioside sialidase (Mus GangSD) with the sequences of the bovine plasma membrane (28) (Bov PlasSD), mouse cytosolic (Mus CytoSD), rat cytosolic (26) (Rat CytoSD), and mouse lysosomal (32) (Mus LysoSD) sialidases. Conserved amino acid residues are indicated by filled boxes. The Asp boxes are solid underlined and F/YRIP motif is double underlined. Dashes indicate gaps introduced by the Parallel Protein Information Analysis algorithm to maximize alignment (42).

View larger version (37K): [in this window] [in a new window]   Fig. 3.   Analysis of the ganglioside sialidase mRNA expression in various mouse tissues. A, a mouse multiple tissue Northern blot (CLONTECH) was hybridized with an [-32P]dCTP-labeled sialidase cDNA (nucleotides 374-769). RNA sizes in kilobases (kb) were determined relative to a RNA ladder. B, the RNA level in mouse brain, heart, and skeletal muscle was determined by competitive PCR as described under "Experimental Procedures." Expression of GAPDH was also measured to check for sample-to-sample variation. Gel photos were scanned and densitometric analysis was performed using the image analysis program NIH Image, version 1.61.

View larger version (77K): [in this window] [in a new window]   Fig. 4.   In situ hybridization analysis of the ganglioside sialidase in adult mouse brain. Frozen parasaggital sections of a BALB/c mouse brain were hybridized with digoxigenin-labeled ganglioside sialidase antisense (A and C) and sense (B and D) RNAs. A and B, cerebral cortex; C and D, cerebellum. All of the photographs were taken at × 100 magnification.

To confirm that the isolated cDNA clone encodes a functionally active sialidase, expression plasmids (pMEmmSD) under the SR promoter were transiently transfected into COS-1 cells. The pMEmmSD-transfected cells showed an increase of 1500-fold in sialidase activity toward gangliosides in the presence of 0.1% Triton X-100 as compared with untransfected cells or with vector-transfected cells (Fig. 5). Over 85% of the expressed sialidase activity in the crude homogenates was recovered in the particulate fraction, suggesting a membrane-bound form. Table I summarizes the substrate specificity of the expressed sialidase. The sialidase hydrolyzed gangliosides GD3, GM3, GD1a preferentially and also oligosaccharides, such as  23 sialyllactose, to some degree, with optimum pH at 4.1-4.5. When the kinetic properties of the sialidase were compared using GM3, GD1a, and sialyllactose as substrates, the K_m for sialyllactose (584 µM) was found to be 10 times higher than the values for GM3 (47 µM) and GD1a (57 µM), indicating that gangliosides are able to be hydrolyzed much more efficiently than sialyllactose by the sialidase. Computer sequence analysis by PSORT World Wide Web server predicted that the polypeptide might be localized in microsomes.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Expression of the ganglioside sialidase in COS-1 cells. Sialidase activity in cells 2 days after pMEmmSD transfection by electroporation. Cells were collected and assayed using mixed gangliosides as substrates.

To investigate an involvement of the ganglioside sialidase in the differentiation of neuronal cells, we measured levels of sialidase activity (Fig. 6A) and mRNA (Fig. 7) in Neuro2a cells differentiated by BrdUrd. The total cell numbers increased only slightly at 2-3 days of BrdUrd treatment, and there were 10-15% dead cells in the culture medium at 4-5 days. As the proportion of cells bearing neurites rose to 65% by 4 days, elevation of ganglioside sialidase activity was observed, with a concomitant increase of acetylcholinesterase (Fig. 6B), a neuronal differentiation marker, indicating that BrdUrd causes functional as well as morphological differentiation of murine neuroblastoma cells. Successive quantitative PCR analyses showed an increase of over 10-fold in the amount of transcripts observed after 3 days of treatment and a 21-fold increase after another 2 days, as compared with the untreated cell case (Fig. 7), equal amounts of template cDNA being confirmed by means of GAPDH.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Ganglioside sialidase activity during BrdUrd-induced differentiation of Neuro2a cells. Neuro2a cells were seeded at a density of 2 × 104 cells/cm2 in DMEM supplemented with 0.25 mM BrdUrd. At the indicated day, the number of neurite-bearing cells was counted (A, ), cells were harvested by trypsinization, and the homogenates were assayed for ganglioside sialidase (A, black bars) and acetylcholinesterase activities (B), respectively.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Quantitative measurement of the ganglioside sialidase mRNA in BrdUrd-stimulated Neuro2a cells. Neuro2a cells were seeded at a density of 2 × 104/cm2 in medium containing 0.25 mM BrdUrd. At the indicated day, cells were harvested for total RNA preparation. The mRNA expression of each samples was quantified by competitive PCR using deleted cDNA competitors, as described in Fig. 3.

To further elucidate the functional role of ganglioside sialidase in the process of neuronal differentiation, Neuro2a cells were stably transfected with the expression plasmid. Three independent positive clones showed markedly increased levels of ganglioside sialidase activity (Fig. 8A). Two of the three clones appeared morphologically indistinguishable from parental Neuro2a cells, but the stable transfectant Neuro-mSD10 with the highest activity (about 18-fold versus control cells) had more neurites compared with the parental cells. All three clones showed accelerated neurite outgrowth on BrdUrd treatment: 45, 40, and 36% of the transfectants Neuro-mSD10, -2, and -16, in the order of activity level, respectively, bore neurites more than 1.5-fold longer than the soma diameter at day 2 (Fig. 8B). Consequently there seemed to be a correlation between ganglioside-hydrolyzing activity and acceleration of neurite outgrowth. Unlike the sialidase-overexpressing cells, less than 15% of the parental Neuro2a cells contained neurites at day 1 and 31% at day 2. At day 3, more than 44% of both parental and sialidase-transfected Neuro2a cells had neurites. It should be noted that the length and complexity of neurites in the sialidase-transfected Neuro2a cells appeared longer and greater than those of parental Neuro2a cells.

View larger version (51K): [in this window] [in a new window]   Fig. 8.   Effect of overexpression of the ganglioside sialidase cDNA on neuro2a cell differentiation induced by BrdUrd. A, Neuro2a cells were stably transfected with the expression construct pMEmmSD as described under "Experimental Procedures." Three independent positive clones (Neuro-mSD2, Neuro-mSD10, and Neuro-mSD16) showed high levels of ganglioside sialidase activity compared with parental (Neuro2a) and empty vector-transfected (Neuro-neo) cells. B, cells were cultured in medium containing 0.25 mM BrdUrd, and photographs were taken at × 100 magnification on the day indicated. To determine the effect of sialidase gene expression, cells were counted and scored for the expression of neurites at least 1.5-fold the diameter of the soma.

In confirmation of an direct involvement of the sialidase in neurite formation through desialylation of gangliosides, we observed changes in cellular sialic acid contents and ganglioside patterns that the sialidase would bring about. Lipid-bound sialic acids of the cells (10⁶) were determined to be 0.83, 1.0, and 0.23 nmol, whereas protein-bound sialic acids were 2.81, 2.74, and 2.76 nmol in parental Neuro2a cells, Neuro-neo, and Neuro-mSD10, respectively. These data indicate that the overexpressing sialidase desialylates specifically gangliosides, consistent with the results on its substrate specificity. Ganglioside patterns by thin layer chromatography revealed that GM3 (55-60%), GM2 (23-27%), and GD1a (10-14%) were found to be major gangliosides in the cells, and a decrease in these gangliosides was observed in the sialidase-overexpressing cells. After BrdUrd treatment, a small amount of GM1(1.5-2% of the total gangliosides) became apparent in these cells, and the percentage of GM1 was two times higher in the sialidase-transfected cells than in the control cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study, we report cDNA cloning and expression for a mouse ganglioside sialidase with a YRIP sequence and Asp boxes playing a major role in ganglioside degradation in the neuronal system. Thus transfection of the cDNA into COS-1 cells led to a marked increase of sialidase activity with ganglioside substrate preference. When neurite outgrowth of Neuro2a cells was stimulated by BrdUrd, the enzymatic and transcriptional activities of the sialidase were strikingly enhanced in proportion to neurite arborization, and transfection of the sialidase cDNA into Neuro2a cells resulted in accelerated neurite outgrowth.

The observed similarity of primary structure and enzymatic properties to bovine and human membrane sialidases argues that the ganglioside sialidase represents the mouse homolog of these sialidases (31, 32). Degenerate PCR amplification suggested a possible presence of another ganglioside sialidase with very low expression, at least partially identical to the human membrane sialidase. In addition, a cytosolic sialidase could be considered as an another sialidase responsible for ganglioside degradation (27, 30).We thus identified a cDNA for a mouse cytosolic sialidase from mouse skeletal muscle by the homology-based RT-PCR and RACE methods as shown in the multiple alignment (Fig. 2). The primary structure was found to be 88% identical to that of rat cDNA. Compared with the nucleotide sequence of the mouse cDNA recently reported (51), our cDNA sequence differs in several nucleotides, leading to amino acid differences in both the N and C termini. Measurement of expression levels by RT-PCR gave values over 10 times higher for the membrane than the cytosolic sialidase in mouse brain (data not shown). These data strongly suggest that the membrane sialidase cloned here may be the major ganglioside sialidase in mouse brain and Neuro2a cells. The substrate specificity of this sialidase for ganglioside preference and the fact that the transfection of the membrane sialidase resulted only in a decrease of glycolipid-bound sialic acids also support the idea that the sialidase expression in neuronal cells is associated with specific desialylation of gangliosides compared with other sialoglycoconjugates. In situ hybridization analysis revealed interesting results consistent with ganglioside expression patterns previously reported by other investigators. Regional differences in ganglioside composition have been described in mammalian brain (53, 54), and immunological analysis of rat brain has suggested a cell- and layer-specific expression of major and minor gangliosides. GM3 is expressed intensely in the white matter and only slightly in the granular layer of the rat cerebellar cortex; GD3 in both the granular layer and the white matter, as well as in the cerebral cortex; and GD1a exclusively in layers I, II/III, and Va, as well as the upper part of layer VI. Our in situ hybridization analysis showed an mRNA pattern strikingly similar to the ganglioside expression pattern of rat brain. The physiological significance of this sialidase in heart, its site of highest expression on Northern blotting, is uncertain. It might play a role in the regulation of myocardial Ca²⁺ channel function, because it has been reported that bacterial sialidase selectively enhances transient Ca²⁺ currents in cardiac myocytes (55).

Modulation of endogenous gangliosides has been shown to influence the process of neuronal differentiation. For example, treatment of Neuro2a cells with cholera toxin B-subunit and anti-GM1 antibody inhibited neurite outgrowth (19), whereas anti-GM3 stimulated differentiation (56). Transfection of Neuro2a cells with ganglioside GD3 synthase cDNA increases the amount, not of only GD3, but also of b-series gangliosides, resulting in changes in cell morphology, along with cholinergic differentiation (57). Neurite outgrowth from murine neuroblastoma cells has been found to be significantly inhibited by threo-1-phenyl-2-hexadecanoylamino-3-pyrrolidino-1-propanol, an effective inhibitor of glycosphingolipid synthesis (58). Sialidases and specific inhibitors have proven to be useful tools for alteration of ganglioside patterns on the neuronal surface, and modification of de novo degradation of gangliosides may affect cellular differentiation. This sialidase inhibitor, 2,3-dehydro-2-deoxy-N-acetylneuraminic acid, abolished increase of neuronal differentiation marker acetylcholinesterase in a human neuroblastoma cell line (20, 21). In granule cells from rat cerebellum, a marked increase of sialidase activity is observed during differentiation with formation of functional synapses (18). Exogeneous bacterial sialidase from Clostridium perfringens stimulates neurite outgrowth in murine neuroblastoma cells (19). Bacterial sialidase-induced neuritogenesis is closely related to the alteration of intracellular calcium homeostasis secondary to the up-regulation of GM1 on the cell surface and nuclear envelope (59).

Together with these previous reports, our experiments on Neuro2a cells provide direct evidence of crucial role of the ganglioside sialidase in neurite formation. The sialidase expression may affect the rate of which cell express neurites rather than the overall extent because the transfection effects are observed at earlier times. As GD1a, GM3, and GM2 were found to be major gangliosides in the cells, which is consistent with previous reports (46, 60), and as they were good substrates for the sialidase except for GM2 (as shown in Table I), quantitative and qualitative changes of these gangliosides may occur with increased expression of the sialidase during neurite formation. Our present study confirmed that one of the characteristic features is an appearance of GM1, which is likely to occur by increased expression of the sialidase. A GM2-hydrolyzing sialidase activity has been reported (28, 61, 62) in rodents and canine cells, but the enzyme(s) seems to be distinct from our sialidase. Our preliminary data suggest that sialidase activity toward GM2 gradually rises, but only a 1.7-fold increase was noted even at day 3 of BrdUrd treatment, suggesting only a limited involvement in BrdUrd-induced neuritogenesis.

In conclusion, the results of the present study indicate that a membrane ganglioside sialidase is important for neuronal cell differentiation. Although the regulation mechanisms are not fully understood at present, it is feasible that the sialidase is modulated by various kinases, including protein kinase C (63), as well as controlled at the gene transcription level. This sialidase, on the other hand, would give an influence on signal transduction processes by alteration of gangliosides affecting various kinase activities (64-66). Taken together, our experiments provide strong evidence for an essential participation of the sialidase in ganglioside modulation in the central nervous system and in neuroblastoma cells.

	ACKNOWLEDGEMENTS

We are grateful to Dr. Kazuo Maruyama, Tokyo Medical and Dental University, for kindly providing the pME18S vector and to Dr. Masashi Sawada, Tohoku University School of Medicine, for helpful discussion. We also thank Hiroyuki Kato and Setsuko Moriya, Divisions of Pathology and Biochemistry, respectively, Miyagi Prefectural Cancer Center, for expert technical assistance.

	FOOTNOTES

^* This work was supported in part by a grant-in-aid for scientific research on priority areas from the Ministry of Education, Science, Sports and Culture of Japan.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AB026842 and AB008184 for mouse membrane-associated ganglioside sialidase and mouse cytosolic sialidase, respectively.

^¶ To whom correspondence should be addressed. Tel.: 81-22-384-3151; Fax: 81-22-381-1195; E-mail: tmiyagi@mcc.pref.miyagi.jp.

	ABBREVIATIONS

The abbreviations used are: BrdUrd, 5-bromodeoxyuridine; RACE, rapid amplification of cDNA ends; PCR, polymerase chain reaction; PBS, phosphate-buffered saline; DMEM, Dulbecco's modified Eagle's medium; RT, reverse transcription; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; bp, base pair(s); C/M, chloroform/methanol.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES