The coxsackievirus-adenovirus receptor protein as a cell adhesion molecule in the developing mouse brain

Molecular Brain Research 77 (2000) 19–28 www.elsevier.com / locate / bres

Research report

The coxsackievirus-adenovirus receptor protein as a cell adhesion molecule in the developing mouse brain Takao Honda a,b ,1 , Hiroshi Saitoh a,b ,1 , Masayoshi Masuko a , Takako Katagiri-Abe a , Kei Tominaga a,b , Ikuo Kozakai a , Kazuo Kobayashi c , Toshiro Kumanishi c , b b a, Yuichi G. Watanabe , Shoji Odani , Ryozo Kuwano * a

b

Research Laboratory for Molecular Genetics, Niigata University, 1 -Asahimachi, Niigata 951 -8510, Japan Course of Biosystem Science, Graduate School of Science and Technology, Niigata University, Niigata 950 -2181, Japan c Brain Research Institute, Niigata University, Niigata 951 -8585, Japan Accepted 25 January 2000

Abstract In an attempt to elucidate the molecular mechanisms underlying neuro-network formation in the developing brain, we analyzed 130 proteolytic cleavage peptides of membrane proteins purified from newborn mouse brains. We describe here the characterization of a membrane protein with an apparent molecular mass of 46 kDa, a member of the immunoglobulin superfamily of which the cDNA sequence was recently reported, encoding the mouse homologue of the human coxsackievirus and adenovirus receptor (mCAR). Western and Northern blot analyses demonstrated the abundant expression of mCAR in the mouse brain, the highest level being observed in the newborn mouse brain, and its expression was detected in embryos as early as at 10.5 days post-coitus (dpc), but decreased rapidly after birth. On in situ hybridization, mCAR mRNA expression was observed throughout the newborn mouse brain. In primary neurons from the hippocampi of mouse embryos the expression of mCAR was observed throughout the cells including those in growth cones on immunohistochemistry. In order to determine whether or not mCAR is involved in cell adhesion, aggregation assays were carried out. C6 cells transfected with mCAR cDNA aggregated homophilically, which was inhibited by specific antibodies against the extracellular domain of mCAR. In addition to its action as a virus receptor, mCAR may function naturally as an adhesion molecule involved in neuro-network formation in the developing nervous system.  2000 Elsevier Science B.V. All rights reserved. Themes: Cellular and molecular biology Topics: Membrane composition and cell-surface macromolecules Keywords: Coxsackievirus and adenovirus receptor; Nerve growth cone; Adhesion molecule; Subcellular localization; In situ hybridization

1. Introduction The transient expression of an adhesion molecule is a crucial event for neural network formation in the early developmental stages of the brain when neurons are most actively engaged in cell–cell interactions. Cell surface adhesion molecules are thought to play a crucial role in axon guidance and fasciculation in the developing nervous *Corresponding author. Tel.: 181-25-227-2343; fax: 181-25-2270793. E-mail address: [email protected] (R. Kuwano) 1 These authors contributed equally to this work.

system. Membrane proteins on nerve growth cones and growing axons act as recognition molecules involved in neurite extension, pathfinding and targeting of appropriate cells, and consequently the formation of neural connections [7,11]. We purified membrane proteins exhibiting developmental changes, and determined the partial amino acid sequences of more than 130 peptides obtained on proteolytic cleavage of the membrane proteins [1]. Among them we focused on a membrane protein with a molecular mass of 46 kDa, which was found to be expressed extensively in prenatal brains but to be decreased in adult brains on Western blot analysis with antibodies against the purified

0169-328X / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0169-328X( 00 )00036-X

20

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

peptides. We cloned cDNAs by RT-PCR with degenerate oligonucleotides corresponding to the sequences of the purified peptides and found that this membrane protein consists of two immunoglobulin domains, one V-like and one C2-like domain, followed by a transmembrane domain and a cytoplasmic domain. This structure is found in a number of proteins in the immune system [25] rather than in proteins expressed in the central nervous system. In the course of our study, Bergelson et al. [2] and Tomko et al. [22] isolated a coxsackievirus and adenovirus receptor (CAR) cDNA. The 46-kDa membrane protein purified from newborn mouse brain was identified as the mouse homologue (mCAR) [2,3] of the human CAR on prediction of the amino acid sequence from the cDNA. Coxsackievirus is known to exhibit affinity for newborn tissues [6,13], and was first isolated on intracerebral inoculation into newborn mice. Suckling mice, 3–7 days of age, became paralyzed, while mice of 10–12 g in weight did not [6]. This critical feature of coxsackievirus pathogenicity coincided with our observation of abundant expression up to birth followed by sharp decreases in both the protein and mRNA, and no detection in adult brains. However, the endogenous native function of mCAR during the development of the mouse brain has not been elucidated. We report here the structural features of mCAR, a member of the immunoglobulin superfamily in the brain, and show its subcellular localization in cultured hippocampal neurons, and the developmental changes in the levels of mCAR and its mRNA. We further demonstrate the cell adhesion activity of mCAR toward cultured cells transfected with cDNA in a eukaryotic expression vector.

2. Materials and methods

2.1. Purification of membrane proteins Growth cone-enriched and non-enriched fractions were prepared from newborn mouse brains by the discontinuous sucrose density gradient centrifugation method [18]. Membrane proteins were isolated from each fraction and digested with lysylendopeptidase, and then the purified proteins were subjected to amino acid microsequencing [1].

2.2. Western blot analysis Two peptides (a, KIYDNYYPDLKC; and b, KTQYNQVPSEDFERAPQC) were synthesized and used to immunize white rabbits as described previously [1]. Mouse brains were homogenized in 20 mM Tris–HCl (pH 7.4) containing 0.32 M sucrose and then centrifuged for 10 min at 4000 rpm. The supernatants were centrifuged at 55 000 rpm for 1 h (Optima TL, Beckman). The precipitated crude membranes were washed with 20 mM Tris–

HCl (pH 7.4) containing 1 mM phenylmethylsulfonyl fluoride. The washed membrane proteins (5 mg) were subjected to SDS–PAGE (12.5% acrylamide) and then electro-transferred to a nitrocellulose membrane. The blot was incubated with antibodies against peptide-a or -b, followed by reaction with HRP-labeled swine anti-rabbit IgG (Dako) for 60 min and development with an ECL detection system (Amersham).

2.3. Primary hippocampal culture and immunocytochemistry Primary hippocampal cultures were prepared from mouse brains on embryonic day 16 [8,17]. Briefly, dissected hippocampi were suspended in a 2.5% trypsin solution for 15 min at 378C, washed three times with calcium- and magnesium-free Hank’s balanced salt solution, and then triturated with a glass pipette to dissociate the cells. The cells were then plated on acid-washed, poly-L-lysine treated glass coverslips (Matsunami, Japan) in 6-cm Petri dishes containing minimum essential medium (MEM) supplemented with 10% horse serum. After the neurons had become attached to the substrate, the coverslips were inverted and transferred to dishes containing a confluent monolayer of astroglia, and then maintained in serum-free medium (MEM containing the N2 supplements of Bottenstein and Sato [4] together with 0.1 mM sodium pyruvate and 0.1% ovalbumin). Small dots of paraffin on the coverslips supported them just above the glial monolayers. For immunocytochemistry, cells were fixed with 4% paraformaldehyde and 4% sucrose in phosphate-buffered saline (PBS, pH 7.4) for 30 min, and then permeabilized with PBS containing 0.1% Trixon X-100 and 10% goat serum. Cells were incubated with primary antibodies against peptide-b (11.8 mg / ml) for 60 min, biotinylated goat anti-rabbit IgG (Nichirei, Japan) for 15 min, and then FITC-conjugated streptavidin (1:50; Vector Laboratories, CA) for 15 min. The same cells were double-stained with monoclonal antibodies against class III beta-tubulin (TuJ1, 1.7 mg / ml; Babco, CA) or microtubule-associated protein (MAP2, 1.67 mg / ml; Boehringer Mannheim, Germany) for 60 min, followed by reaction with TRITC-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, PA). To visualize F-actin, cells labeled with antipeptide-b antibodies were incubated with Texas Red-phalloidin (1:30, Molecular Probes, OR) in PBS for 60 min. Fluorescent images of the neurons were obtained using a Zeiss LSM 310 laser-scanning confocal microscope.

2.4. Deglycosylation Aliquots of crude membrane proteins (5 mg) were solubilized in 10 ml of 0.2 M sodium phosphate buffer (pH 8.6), 0.5% SDS, and 3% 2-mercaptoethanol, and then heated in a boiling water bath for 3 min. The denatured protein solution was mixed with 12 ml of 0.2 M sodium

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

21

phosphate buffer (pH 8.6), and then 5 ml of 7.5% NP40. N-glycopeptidase F (1 m unit; Takara, Japan) was added to the protein solution, followed by incubation for 16 h at 378C [21].

cells by electroporation and cells were selected with 500 mg (titer) / ml of G418. Recombinant C6 cells transfected with pfGMP46 and prfGMP46 are referred to as the fC6 and rfC6 cell line, respectively.

2.5. cDNA cloning

2.8. Aggregation assay

Degenerate oligonucleotides corresponding to both the forward and reverse directions of peptides-a and -b were synthesized. A SalI site, CTCGAG, was added at the 59 end of each oligonucleotide. The oligonucleotide sequences were as follows: aF, 59-CGCTCGAGAT(A / C / T)TA(C / T)GA(C / T)AA(C / T)TA(C / T)TA(C / T)CC, as a forward primer, and aR, 59-CGCTCGAGGG(A / G)TA(A / G)TA(A / G)TT(A / G)TC(A / G)TA(A / G / T)AT, as a reverse one corresponding to the peptide-a sequence, IYDNYYP, and bF, 59-CGCTCGAGAA(A / G)AC(A / C / G / T)CA(A / G)TA(C / T)AA(C / T)CA(A / G)GT, as a forward primer, and bR, 59-CGCTCGAGAC(C / T)TG(A / G)TT(A / G)TA(C / T)TG(A / C / G / T)GT(C / T)TT, as a reverse one of the peptide-b sequence, KTQYNQV. A plasmid clone, pGMP729, was obtained by PCR with the aF and bR primers and cDNA from total brain RNA as a template. Unique oligonucleotides, 59-AATGTCCGACAGCTG CAG and 59-GAAGGAAGTTCATCATGATATC, in pGMP729 were synthesized, and used as primers for 59 RACE and 39 RACE, respectively, using a Marathon cDNA amplification kit (Clontech). A plasmid pGMP46 encoding the 46-kDa protein was constructed by ligating three overlapping cDNA clones, i.e. pGMP729, 59 RACE and 39 RACE cDNAs.

Cell adhesion activity was determined by means of an aggregation assay [19,20,26]. Stable transfectants fC6 and rfC6, and parental C6 cells were washed three times with PBS, and then treated with 0.001% trypsin and 0.01 mM EDTA for 15 min at 378C. The cells were collected by brief centrifugation and then dispersed as single-cell suspensions at 2310 6 / ml in 2 ml of Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum. These suspensions were incubated at 378C without agitation. In the cell aggregation inhibition assay, cells were suspended in 0.5 ml of the same medium and allowed to stand for 30 min on ice in the presence or absence of antibodies (0.1 mg / ml) against the mCAR-a peptide in the extracellular domain before incubation at 378C. To measure the progress of aggregation, isolated cells and aggregated-cell masses, each being counted as one particle, were counted at 15-min intervals with a haemocytometer. The degree of aggregation was represented as Nt /N0 , where Nt and N0 are the total numbers of particles at incubation times t and 0 min, respectively [20]. In the mixed cell aggregation assay, parental cells were labeled with 40 mg / ml of DiI (Molecular Probes, OR) for 1 h [10]. The labeled parental C6 cells and non-labeled fC6 cells were dispersed in the culture medium in a 1:1 ratio and then incubated for 60 min at 378C. The mixed culture was observed by laser scanning confocal microscopy.

2.6. RNA analysis Total RNA (10 mg) was separated on a 1.5% agarose gel in MOPS buffer (pH 7.0) containing formaldehyde and then transferred to a nitrocellulose membrane filter. The filter was hybridized with a [ 32 P]-labeled cDNA fragment derived from pGMP729 in a solution comprising 50% formamide, 53 SSC, 50 mM sodium phosphate (pH 6.5), 53 Denhardt’s solution, and 250 mg / ml heat-denatured salmon sperm DNA at 428C. In situ hybridization was carried out by means of a reported method [23] using the [ 35 S]-labeled cDNA fragment from pGMP729 as a probe.

2.7. cDNA transfection Plasmid pCAGGSneo was constructed by insertion of the XhoI and BamHI fragment derived from pMC1neoPolyA (Stratagene) into the blunt-ended PstI site of a eukaryotic expression vector, pCAGGS [15]. The blunt-ended NotI and XhoI fragment of pGMP46 was subcloned in the sense (pfGMP46) or reverse orientation (prfGMP46) into the unique blunt-ended XhoI site of pCAGGSneo. These plasmids were introduced into C6

3. Results

3.1. Isolation of a membrane protein and cDNA cloning A membrane protein exhibiting a molecular mass of 46 kDa on SDS–PAGE was purified from the growth coneenriched fraction. To determine whether peptide-a and -b were generated from a single protein, RT-PCR analysis were carried out with two different combinations of oligonucleotide primers, aF and bR, and bF and aR. The primer combination of sense strand aF for peptide-a and anti-sense strand bR for peptide-b produced a 729-bp long DNA fragment. But the other set of primers, bF and aR, failed to amplify a specific PCR product. These results indicate that the two peptides are the products of a single protein generated through lysylendopeptidase digestion. We obtained plasmid pGMP46, which contains a ligated 1242 bp long cDNA encompassing the entire coding sequence. Judging from the predicted amino acid sequence, the first hydrophobic 19 amino acids seem to be an N-terminal signal sequence followed by one V-like and one

22

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

C2-like immunoglobulin domain, a hydrophobic transmembrane domain and an intracellular domain. Searches through databases revealed that the membrane protein containing peptide-a and -b was identical to a mouse homologue of human CAR (GenBank accession numbers Y07593 for human cDNA and Y10320 for the mouse one). We refer to this 46-kDa membrane protein hereinafter as mCAR.

3.2. Tissue distribution and developmental changes in the levels of mCAR and its mRNA Western blotting showed that anti-mCAR-b (peptide-b) antibodies reacted intensely with a protein with an apparent molecular mass of 46 kDa from the brain, with only faint signals for the heart, and none for the lung, liver and kidney, of newborn mice. Expression of mCAR was observed as early as embryonic 10.5 dpc, and its level increased until the perinatal period, followed by a rapid decrease after birth, and mCAR was not detected in the adult brain (Fig. 1). We examined the tissue distribution and developmental changes of mCAR mRNA by Northern blotting (Fig. 2). Fig. 2A shows the most abundant expression of an approximately 6 kb mCAR was observed in the mouse brain, with slight expression in the kidney, heart and liver, but none in the lung. On Northern blotting, mCAR mRNA was detected in the kidney while the mCAR protein was

not observed. We did not attempt to determine whether or not mCAR expression in the kidney was under posttranscriptional regulation. mCAR mRNA was detected as early as embryonic 10.5 dpc and underwent progressive increases with age, reaching a maximum level in the perinatal period and then rapidly decreasing after birth (Fig. 2B). In the olfactory bulb, relatively abundant expression was observed at postnatal 7 days and the highest expression occurred 1 week later than that in the cerebral cortex and brainstem. In the cerebellum the expression of mCAR mRNA was significantly lower throughout the postnatal period in comparison with in the cerebral cortex, brainstem and olfactory bulb (Fig. 2E).

3.3. Distribution of mCAR mRNA in the brain The cDNA fragment of pGMP729, which was used as a probe and gave a single band on Northern blot analysis, was used again for in situ hybridization. Strong expression was observed in the cerebral cortex, midbrain, hippocampus, various thalamic nuclei and the meninges, whereas in the cerebellum and olfactory bulb the expression was relatively low in newborn mice (Fig. 3A). In a coronal section we observed hybridization signals of similar intensity in the amygdala (Fig. 3B). There were no signals in the corpus callosum, anterior commissure or cerebellar white matter, where cells are sparse or absent. These findings indicate that the mCAR gene is transcribed widely

Fig. 1. Expression of mCAR revealed by Western analyses. (A) Tissue distribution of mCAR. Crude membrane proteins (5 mg) of kidney, heart, brain, lung, liver and testis from newborn mice were separated by SDS–PAGE. Proteins were transferred to a nitrocellulose membrane and then reacted with anti-mCAR-b antibodies. (B) Western blot analysis of mCAR in developing brains. Crude membrane proteins were prepared from brains at the embryonic 10.5 dpc (E10.5) and 16.5 dpc (E16.5), newborn (P0), and postnatal 7 days (P7), 21 days (P21) and 60 days (P60), stages. The positions of molecular mass markers are indicated on the left of the gel.

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

23

Fig. 2. Tissue distribution and developmental changes of mCAR mRNA. (A) Equal amounts of total RNA (10 mg) from kidney, heart, brain, lung and liver of newborn mice were separated and hybridized with the [ 32 P]-labeled fragment from pGMP729 as a probe. (B) Total RNAs were prepared from whole brains at the embryonic (E10.5, E12.5, E14.5 and E16.5 dpc), newborn (P0), and postnatal 7 days (P7), stages, and from cerebral cortex (Ctx) and cerebellum (Cb) at postnatal 40 days. (E) Total RNAs from postnatal (P0, P7, P14 and P21) cerebral cortex, cerebellum, brainstem and olfactory bulb were examined. (C, D) Ethidium bromide staining of 28S and 18S was performed to assess the integrity and loading of total RNAs from the tissues examined.

in the newborn mouse brain. In the adult brain, positive signals were barely detectable even when the film was exposed for 36 s, as compared with 5 s for newborn samples. The bright signals in the cerebellum and at the

margin of the brain are considered to be nonspecific since these signals were outside cells and did not appear as silver grains in a bright field (Fig. 3C and D). These distribution and developmental changes of mCAR mRNA coincide

24

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

Fig. 3. In situ hybridization of mCAR mRNA in mouse brains. Cryostat sections were hybridized with the [ 35 S]-labeled mCAR cDNA probe and then exposed for 8 weeks at 48C with a desiccant. The sections were then dipped in an emulsion and hybridization signals were visualized by dark field microscopy (A–C). (A) Sagittal section of a newborn mouse head. Intense signals can be observed in the cerebral cortex (Ctx), midbrain (MB), hippocampus (Hi), thalamic nuclei (Th), and meninges (Me), and moderate reaction in the cerebellum (Cb) and olfactory bulb (OB), but no signal in the corpus callosum (CC) or anterior commissure (AC). (B) Coronal section of a newborn mouse brain. Strong signals can be seen in the amygdala (Am) as well as the areas described above. (C) Sagittal section of an adult brain. (D) The same section of the adult brain was counterstained with methylgreen pyronin and observed by bright field microscopy. The exposure times for newborn and adult samples in a dark field were 5 and 36 s, respectively. Scale bars: 1 mm in A and B; 2 mm in C and D.

with the results obtained on Western and Northern bolt analyses (Figs. 1 and 2).

3.4. Immunocytochemistry To confirm the subcellular localization of mCAR we examined cultured hippocampal cells by means of immunocytochemistry with polyclonal antibodies against mCAR-b (Fig. 4). In dissociated hippocampal neurons, mCAR was distributed throughout the cell bodies, neurites and growth cones. mCAR as well as F-actin (Fig. 4D–F) was detected on the filopodia of the growth cones. mCAR was detected in the areas which reacted with antibodies against MAP2, which was used as a dendrite marker protein.

3.5. Deglycosylation assay To determine whether or not the broad band material was generated through posttranslational modifications, we performed a deglycosylation assay (Fig. 5A). Two canonical N-glycosylation sites (Asn–Val–Thr at amino acid position 106 and Asn–Ala–Ser at 201) are located in the extracellular domain. Crude membrane proteins were digested with N-glycopeptidase F and then analyzed by Western blotting with antibodies against mCAR-b, which is located in the intracellular domain. Treatment with N-glycopeptidase F generates a single band corresponding to an apparent molecular mass of 39 kDa, which is similar to the size calculated from the deduced amino acid sequence. These results indicate that the broad band material comprised different carbohydrate units linked to

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

25

Fig. 4. Distribution of mCAR and TuJ1, MAP2 or F-actin in hippocampal neurons. Hippocampal cells were plated on poly-L-lysine coated coverslips. After 2 days, cells were fixed, permeabilized, and then double stained with anti-mCAR-b antibodies (B, E, H; green) and anti-neuron-specific class III beta-tubulin isotype (TuJ1) antibodies (A; red), Texas Red phalloidin (D; red), or anti-MAP2 antibodies (G; red). The right panels (C, F, and I) show merging of the two labelings. Scale bars: 20 mm.

two possible N-glycosylation sites on the extracellular domain. It remains to be determined, however, whether glycosylation occurs at both sites or whether different carbohydrate chains are attached to the same site, or both.

3.6. mCAR as a homophilic adhesion molecule C6 cells were transfected with mCAR cDNAs in a eukaryotic expression vector. Expression of mCAR in the

transfectant line was verified by Western blot analysis. As shown in Fig. 5B, fC6 cells transfected with pfGMP46 produced a protein of the same size as the native molecule from the brain, suggesting an N-glycosylated protein. Both the parental cells and the rfC6 cells transfected with cDNA in the reverse direction as a control did not express mCAR. To determine whether or not mCAR functions as an adhesion molecule like many other proteins of the immunoglobulin superfamily, we analyzed cell–cell interaction by means of an aggregation assay (Fig. 6). mCAR-

26

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

4. Discussion

Fig. 5. Western blot analyses of deglycosylation and transfection. (A) Crude membrane proteins were treated with (1) or without (2) Nglycopeptidase F for 16 h at 378C. (B) C6 cells were transfected with mCAR cDNA in the sense (fC6) and reverse orientation (rfC6). Crude membrane proteins from brain (brain), and total proteins from stable transfectants (fC6, rfC6) and parental cells (C6) were separated by SDS–PAGE and then reacted with anti-mCAR-b antibodies. The positions of molecular mass markers are indicated on the left of the gels.

expressing fC6 cells formed large aggregates (Fig. 6F), whereas rfC6 and parental C6 cells did not (Fig. 6G and H). To examine the specificity of the cell aggregation mediated through the mCAR molecule we used specific antibodies against the mCAR-a peptide, which is located in the extracellular domain. The aggregation of mCAR-expressing fC6 cells was inhibited in the presence of antimCAR-a antibodies (Fig. 6B). Non-immunized immunoglobulin did not impede aggregation. In a series of cell aggregation inhibition assays, cells were treated with antibodies on ice, by which aggregation may be delayed (Fig. 6A and B). We investigated whether or not the aggregation via the mCAR molecule depends on homophilic or heterophilic adhesion by means of a mixed cell aggregation assay (Fig. 6I–K). In this assay the parent C6 cells labeled with DiI were mixed in a 1:1 ratio with non-labeled mCAR-expressing fC6 cells, and then were allowed to aggregate during the incubation. The labeled C6 cells, which were identified as fluorescent cells, were scattered around aggregates composed of non-fluorescent cells. These results indicate that the aggregation of transfectant fC6 cells occurs at least in part through the mCAR molecule in a homophilic manner.

We describe here the characterization of a membrane protein purified from a growth cone-enriched fraction which is identical with mCAR [2,3]. As judged on amino acid sequence comparison, the mCAR is a member of the immunoglobulin superfamily, of which the highest identity of the primary structure is with CTX [5] and human A33-antigen [9]. CTX and A33-antigens are composed of a combination of one V- and one C-like immunoglobulin domain, a hydrophobic transmembrane domain and an intracellular domain, and are expressed in the thymocytes of Xenopus, and human small and large intestines, respectively. This structure is distinct from those of other immunoglobulin superfamily members, such as telencephalin, N-CAM, L1, TAG-1, contactin, DM-GRASP, P0 and MAG, shown to exist in the central nervous system [24,25]. mCAR consists of seven exons (unpublished observation by K. Tominaga) and the insertion sites of introns are almost the same as in CTX [5], indicating that mCAR is evolutionally close to CTX. Recent studies have shown that viruses use a variety of adhesion molecules on host cells as receptors. Cell surface molecules that belong to the immunoglobulin superfamily are also used as receptors, such as CD4 for HIV [12], ICAM for rhinovirus [16], and poliovirus receptor [14]. Coxsackievirus and adenovirus use a neuronal adhesion molecule as a receptor, although there is no direct evidence of virus infection via this adhesion molecule in vivo. In vitro, receptor-negative NIH3T3 cells [22] and nonpermissive CHO cells [3] transfected with mCAR cDNA become susceptible to infection by coxsackievirus B. In the newborn mouse brain, expression of mCAR mRNA was clearly demonstrated on the meninges by in situ hybridization. These findings might be not inconsistent with infantile meningitis, which is caused by coxsackievirus via the mCAR molecule as a receptor. Adhesion molecules are involved in a variety of cell– cell interactions through homophilic or heterophilic ligands such as integrins and extracellular matrices during the development of the brain. To determine whether or not mCAR mediates homophilic binding, we performed an aggregation assay on C6 cells transfected with plasmid pGMP46. Like many other members of the immunoglobulin superfamily, mCAR on cells binds at least in a homophilic manner in vitro. An immunocytochemical study clearly demonstrated that mCAR is present throughout the cells not only in nerve growth cones but also dendrites and axons of cultured hippocampal neurons. The abundant mCAR expression seen in the embryo decreases rapidly in the postnatal brain. These findings suggest that mCAR on the cell surface is likely to mediate tissue organization in the developing brain as a native function. Further experiments are necessary to elucidate the molecular mechanisms of mCAR’s involvement in neurite exten-

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

27

Fig. 6. Aggregation assay. (A) Parental C6 cells (s), and stable transfectants fC6 (h) and rfC6 (d) were cultured in a CO 2 incubator at 378C without agitation and allowed to aggregate. (B) Transfectant fC6 cells were incubated in the absence (h) or presence (m) of antibodies against mCAR-a, which is located in the extracellular domain. Normal rabbit non-immune serum was used as a control (n). Each point is the average for four independent experiments. Error bars indicate S.E.M. (C–H) Transfectants fC6 (C, F) and rfC6 (E, H), and parental C6 cells (D, G) were dispersed as single-cell suspensions (C–E) and then incubated for 60 min at 378C (F–H). (I–K) Laser scanning confocal micrographs of a mixed cell culture; phase-contrast (I), fluorescent (J), and combined (K) images. Parental C6 cells labeled with DiI and unlabeled transfectant fC6 cells were mixed in a 1:1 ratio and then allowed to aggregate for 60 min at 378C. DiI was visualized as fluorescence. Untransfected C6 cells showing fluorescence are scattered around the aggregates composed of non-fluorescent fC6 cells. Scale bars: 200 mm.

28

T. Honda et al. / Molecular Brain Research 77 (2000) 19 – 28

sion, axon guidance, fasciculation and synapse formation as an adhesion molecule.

Acknowledgements We thank Drs Tadahiro Hamada and Akiko Nishiyama for the helpful discussions, and Dr Hiroaki Asou for the valuable comments on the aggregation assay. This work was supported in part by Grants-in-Aid from the Ministry of Education, Science, Sports, and Culture, and the Ministry of Health and Welfare of Japan (to RK).

References [1] T.K. Abe, H. Tanaka, T. Iwanaga, S. Odani, R. Kuwano, The presence of the 50-kDa subunit of dynactin complex in the nerve growth cone, Biochem. Biophys. Res. Commun. 233 (1997) 295– 299. [2] J.M. Bergelson, J.A. Cunningham, G. Droguett, E.A. Kurt-Jones, A. Krithivas, J.S. Hong, M.S. Horwitz, R.L. Crowell, R.W. Finberg, Isolation of a common receptor for Coxsackie B viruses and Adenoviruses 2 and 5, Science 275 (1997) 1320–1323. [3] J.M. Bergelson, A. Krithivas, L. Celi, G. Droguett, M.S. Horwitz, T. Wickham, R.L. Crowell, R.W. Finberg, The murine CAR homolog is a receptor for coxsackie B viruses and adenoviruses, J. Virol. 72 (1998) 415–419. [4] J.E. Bottenstein, G.H. Sato, Growth of a rat neuroblastoma cell line in serum-free supplemented medium, Proc. Natl. Acad. Sci. USA 76 (1979) 514–517. [5] I. Chretien, J. Robert, A. Marcuz, J.A. Garcia-Sanz, M. Courtet, L. Du Pasquier, CTX, a novel molecule specifically expressed on the surface of cortical thymocytes in Xenopus, Eur. J. Immunol. 26 (1996) 780–791. [6] G. Dalldorf, G.M. Sickles, An unidentified, filtrable agent isolated from the feces of children with paralysis, Science 108 (1948) 61–62. [7] C.S. Goodman, C.J. Shatz, Developmental mechanisms that generate precise patterns of neuronal connectivity, Cell Suppl. 72 (1993) 77–98. [8] K. Goslin, G. Banker, Rat hippocampal neurons in low density culture, in: G. Banker, K. Goslin (Eds.), Culturing Nerve Cells, MIT Press, Cambridge, MA, 1991, pp. 251–281. [9] J.K. Heath, S.J. White, C.N. Johnstone, B. Catimel, R.J. Simpson, R.L. Moritz, G.F. Tu, H. Ji, R.H. Whitehead, L.C. Groenen, A.M. Scott, G. Ritter, L. Cohen, S. Welt, L.J. Old, E.C. Nice, A.W. Burgess, The human A33 antigen is a transmembrane glycoprotein and a novel member of the immunoglobulin superfamily, Proc. Natl. Acad. Sci. USA 94 (1997) 469–474. [10] M.G. Honig, R.I. Hume, Fluorescent carbocyanine dyes allow living neurons of identified origin to be studied in long-term cultures, J. Cell Biol. 103 (1986) 171–187.

[11] R.O. Hynes, A.D. Lander, Contact and adhesive specificities in the associations, migrations, and targeting of cells and axons, Cell 68 (1992) 303–322. [12] P.J. Maddon, A.G. Dalgleish, J.S. McDougal, P.R. Clapham, R.A. Weiss, R. Axel, The T4 gene encodes the AIDS virus receptor and is expressed in the immune system and the brain, Cell 47 (1986) 333–348. [13] J.L. Melnick, Advances in the study of the enteroviruses, in: E. Berger, J.L. Melnick (Eds.), Progress in Medical Virology, Karger, Basel, 1958, pp. 59–105. [14] C.L. Mendelsohn, E. Wimmer, V.R. Racaniello, Cellular receptor for poliovirus: molecular cloning, nucleotide sequence, and expression of a new member of the immunoglobulin superfamily, Cell 56 (1989) 855–865. [15] H. Niwa, K. Yamamura, J. Miyazaki, Efficient selection for highexpression transfectants with a novel eukaryotic vector, Gene 108 (1991) 193–199. [16] N.H. Olson, P.R. Kolatkar, M.A. Oliveira, R.H. Cheng, J.M. Greve, A. McClelland, T.S. Baker, M.G. Rossmann, Structure of a human rhinovirus complexed with its receptor molecule, Proc. Natl. Acad. Sci. USA 90 (1993) 507–511. [17] R.G. Perez, H. Zheng, L.H. Van der Ploeg, E.H. Koo, The betaamyloid precursor protein of Alzheimer’s disease enhances neuron viability and modulates neuronal polarity, J. Neurosci. 17 (1997) 9407–9414. [18] K.H. Pfenninger, L. Ellis, M.P. Johnson, L.B. Friedman, S. Somlo, Nerve growth cones isolated from fetal rat brain: subcellular fractionation and characterization, Cell 35 (1983) 573–584. [19] E. Taira, N. Takaha, H. Taniura, C.H. Kim, N. Miki, Molecular cloning and functional expression of gicerin, a novel cell adhesion molecule that binds to neurite outgrowth factor, Neuron 12 (1994) 861–872. [20] M. Takeichi, Functional correlation between cell adhesive properties and some cell surface proteins, J. Cell Biol. 75 (1977) 464–474. [21] A.L. Tarentino, T.H. Plummer Jr., Oligosaccharide accessibility to peptide: N-glycosidase as promoted by protein-unfolding reagents, J. Biol. Chem. 257 (1982) 10776–10780. [22] R.P. Tomko, R. Xu, L. Philipson, HCAR and MCAR: The human and mouse cellular receptors for subgroup C adenoviruses and group B coxsackieviruses, Proc. Natl. Acad. Sci. USA 94 (1997) 3352– 3356. [23] H. Usui, T. Katagiri, Y. Yoshida, A. Nishiyama, T. Ichikawa, R. Kuwano, Y. Takahashi, T. Kumanishi, In situ hybridization histochemistry of Spot 35 protein, a calcium-binding protein, in the rat brain, Mol. Chem. Neuropathol. 15 (1991) 207–216. [24] D.E. Vaughn, P.J. Bjorkman, The (Greek) key to structures of neural adhesion molecules, Neuron 16 (1996) 261–273. [25] A.F. Williams, A.N. Barclay, The immunoglobulin superfamilydomains for cell surface recognition, Annu. Rev. Immunol. 6 (1988) 381–405. [26] T. Yazaki, M. Miura, H. Asou, K. Kitamura, S. Toya, K. Uyemura, Glycopeptide of P0 protein inhibits homophilic cell adhesion: competition assay with transformants and peptides, FEBS Lett. 307 (1992) 361–366.