Proc. Natl. Acad. Sci. USA Vol. 92, pp. 10590-10594, November 1995

The C-type domain recognizes the adhesion tenascin-R (J1-160/180/janusin) ANDERS ASPBERG, CHRISTOPH BINKERT*, AND ERKKI RUOSLAHTIt Cancer Research Center, La Jolla Cancer Research Foundation, La Jolla, CA 92037 Communicated by Elizabeth D. Hay, Harvard Medical School, Boston, MA, July 25, 1995 (received for review March 31, 1995)

ABSTRACT The core of large chondroitin sul- EXPERIMENTAL PROCEDURES fate contain a C-type lectin domain. The lectin domain of one of these proteoglycans, versican, was expressed Materials. Antiserum against the versican lectin domain was as a recombinant 15-kDa protein and shown to bind to raised against a synthetic peptide (KMFEHDFRWTDG- insolubilized fucose and GIcNAc. The lectin domain showed STLQYEN), corresponding to residues 2244-2262 strong binding in a gel blotting assay to a doublet (10), and affinity purified by using the peptide. The anti- in rat brain extracts. The binding was calcium dependent and versican antiserum used in immunofluorescence staining was abolished by chemical deglycosylation treatment of the ligand raised against recombinant intact versican (1) and affinity glycoprotein. The versican-binding glycoprotein was identi- purified by using a bacterially expressed fragment of versican, fied as the cell adhesion protein tenascin-R, and versican and corresponding to amino acid residues 51-354 (10). Mouse tenascin-R were both found to be localized in the granular monoclonal anti-tenascin-R #596 (18) was a kind gift layer of rat cerebellum. These results show that the versican from Melitta Schachner (Eidgenossiche Technische Hochs- lectin domain is a binding domain with a highly targeted chule, Zurich), and anti-tenascin-C antiserum was from Telios specificity. It may allow versican to assemble complexes Pharmaceuticals (La Jolla, CA). Agarose beads conjugated to containing , an adhesion protein, and hyaluro- various monosaccharides or to streptavidin were from Sigma. nan. Plasmid Construction and Expression of the Versican Lec- tin Domain. The human versican lectin domain (amino acid Versican, , , and constitute a family residues 2177-2305; ref. 10) was amplified from cDNA clone of chondroitin sulfate proteoglycans that form aggregates with 2B (19) by PCR with upstream primer 5'-GTCAACTC- hyaluronan (1-4). The C termini of these proteoglycans con- GAGACAAGATACCGAGACATGTGAC-3' and down- tain a C-type lectin domain, as well as epidermal growth stream primer 5'-TGCTCTAGATCAGGTCGACCCTT- factor-like and complement regulatory-like sequence motifs. TCTTGCACGTATAGGTG-3'. The lectin-coding sequence This domain arrangement is similar to that of the , was then fused to an immunoglobulin signal peptide (20) and which are adhesion receptors involved in leukocyte homing inserted into pcDNA I/Neo (Invitrogen). The sequence of the and extravasation at inflammatory sites (5-8). construct was determined by using Sequenase II (United The C-type animal are defined by the presence of a States Biochemical). CHO/DG44 cells (21) were cotrans- calcium-dependent -recognition domain in which fected with the versican lectin domain construct and pSV2- a pattern of 32 amino acid residues, spaced over a stretch of dhfr (22) by using calcium phosphate transfection (CellPhect; Pharmacia). Transfected cells were selected in nucleoside-free - 120 amino acids, is highly conserved (9). The lectin domains of human and chicken versican show 96% amino acid identity a-modified minimal essential medium (GIBCO/BRL) con- (10, 11), and the human and rat versican lectin domains differ taining 9% (vol/vol) dialyzed fetal calf serum, and lectin in only one amino acid (Thr-2257 in human versican replaced expression in selected colonies was amplified with methotrex- by Ala; Joan Lemire and Thomas N. Wight, personal commu- ate according to standard procedures. Cell supernatants were nication), suggesting that the versican family lectin domains screened for versican lectin secretion by immunoblotting with may have an important conserved function. However, very anti-lectin-domain antiserum. little is known about the function of these lectin domains. The Purification and Biotinylation of Versican Lectin Domain. aggrecan lectin domain has been shown to bind to fucose and Cells expressing the versican lectin were adapted to growth in galactose in a calcium-dependent manner (12, 13). Recently, serum-free medium (Nutri Base + Vita Cyte serum supple- the recombinant C-terminal part (epidermal growth factor- ment; J. Brooks Laboratories, San Diego). Proteins from like, lectin, and complement regulatory-like domains) of conditioned serum-free medium were precipitated in the chicken versican was shown to bind to simple , as presence of 5 mM EDTA with ammonium sulfate (50% well as to heparin and heparan sulfate, in affinity chromatog- saturation) and dissolved in Hepes-buffered saline (HBS; 50 raphy (14). A C-terminal fragment of neurocan has been mM Hepes, pH 7.4/0.88% NaCl) containing 5 mM CaCl2. reported to bind tenascin and the cell-cell adhesion proteins Thereafter, the lectin was enriched through affinity chroma- N-CAM and Ng-CAM (15-17), but it is unclear if the lectin tography on a fucose-agarose column and eluted with HBS domain is responsible for these interactions. containing 20 mM EDTA. After adding CaCl2 to 50 mM and We have studied the potential of the lectin domain of NaOH to pH 8.5, the lectin was biotinylated with NHS-LC- versican to function as a binding domain. Here we report that biotin (Pierce) and purified by C4 reverse-phase HPLC. The recombinant versican lectin domain, produced in mammalian samples were injected onto a 214TP54 column (Vydac Sepa- cells has carbohydrate-binding activity and that it also binds the extracellular matrix glycoprotein tenascin-R (also known Abbreviations: ELC, epidermal growth factor-like, lectin, and com- as J1-160/180 and janusin). plement regulatory-like domains; TFMS, trifluoromethanesulfonic acid; PNGase F, glycopeptide N-glycosidase; Endo Hf, protein fusion of endo-,B-N-acetylglucosaminidase and maltose-binding protein. The publication costs of this article were defrayed in part by page charge *Present address: Samuel Lunenfeld Research Institute, Mount Sinai payment. This article must therefore be hereby marked "advertisement" in Hospital, 600 University Avenue, Toronto ON, Canada M5G 1X5. accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 10590 Downloaded by guest on September 29, 2021 Biochemistry: Aspberg et aL Proc. Natl. Acad. Sci. USA 92 (1995) 10591 rations, Hesperia, CA) equilibrated with 0.06% trifluoroacetic Immunofluorescence Staining. The cerebellum from an acid (TFA). A linear acetonitrile gradient (0-80% acetonitrile adult, male Sprague-Dawley rat was dissected, removed, em- in 0.06% TFA) was applied, and the protein was eluted at 44% bedded in O.C.T. compound (Miles), frozen, and cut into 5-,um acetonitrile. Lectin-containing fractions were lyophilized and cryosections. The sections were air dried for 20 min and resuspended in HBS containing 5 mM CaCl2 and 0.1% bovine rehydrated in TBS. Nonspecific binding was blocked by incu- serum albumin. bating sections in 5% (vol/vol) goat serum/TBS. Primary Construction and Expression of Versican C-Terminal Do- were then applied in 2% (vol/vol) goat serum/TBS. main-IgG Fusion Protein. The 3' part of versican cDNA, The sections were washed, incubated with fluorescein- corresponding to the epidermal growth factor-like, lectin, and conjugated secondary antibodies raised in goat, washed again, complement regulatory-like domains (ELC; nucleotides 6530- and mounted in Vectashield mounting medium (Vector Lab- 7493; ref. 10), was amplified by PCR from clone 2B and fused oratories). The slides were inspected in a Zeiss AxioVert in frame with the cDNA coding for the signal peptide and the microscope equipped for fluorescence microscopy, and pic- first eight amino acids of cr5 integrin (residues 1-49; ref. 23). tures were taken with Kodak T-Max 400 film. This fragment was fused to human genomic DNA encoding the Fc region of IgGl (24), inserted into plasmid pcDNA I/Neo, and sequenced. The plasmid was introduced, together with RESULTS pSV2-dhfr, into CHO/DG44 cells by using Lipofectamine Binding of the Versican Lectin Domain to Monosacchar- (GIBCO/BRL). The ELC-IgG fusion protein was purified ides. The versican lectin domain was expressed in CHO cells, from cell culture medium by on and a band corresponding to the expected molecular mass of protein A-Sepharose FF (Pharmacia). the lectin (14.7 kDa) in culture medium was detected by Brain Extracts. Adult, female Sprague-Dawley rats were immunoblotting with anti-versican lectin antibodies. To deter- euthanized with carbon dioxide, and the brains were imme- mine if the lectin-domain protein could bind carbohydrates, diately removed and minced in ice-cold 0.9% NaCl. The tissue the culture medium from the CHO cells was incubated with was homogenized with a Polytron in nine volumes of 4 mM monosaccharides coupled to agarose beads. The lectin domain Hepes, pH 7.5/0.3 M sucrose, containing 0.25 mg of N- bound to beads coupled with fucose or GlcNAc but not to ethylmaleimide per ml, 1 mM EDTA, and 0.2 mM phenyl- beads coupled with GalNAc, , or (Fig. 1). methylsulfonyl fluoride as protease inhibitors. The homoge- Chelation of calcium with EDTA or EGTA eliminated binding nate was centrifuged at 12,000 x g for 30 min. The supernatant (data not shown). Thus, the versican lectin domain appears to was transferred to new tubes and centrifuged at 378,000 x g. be a functional carbohydrate-binding domain. The resulting supernatant was aliquotted, quick frozen in Versican Lectin Domain Binds to a Brain Protein. The liquid nitrogen, and stored at -70°C until used. biotinylated versican lectin domain bound to two proteins of Protein Blotting. Brain extract was boiled 5 min in sample approximate molecular masses 170 and 190 kDa in rat-brain buffer, electrophoresed through precast SDS/4-12% poly- extracts (Fig. 2A). Rat brain was chosen as a primary target acrylamide gels (NOVEX, San Diego) with prestained molec- since versican is prominently expressed in the brain (26, 27), as ular mass markers (GIBCO/BRL), and electroblotted onto are the other members of this proteoglycan family (2, 4, 28). nitrocellulose membranes (Schleicher & Schuell). The mem- The 75- and 130-kDa bands on the blots appear to be non- branes were treated with 3% bovine serum albumin in TTBS specific because they are also detected by the avidin- (25 mM Tris-HCl, pH 7.4/0.88% NaCl/0.1% Tween 20) to horseradish peroxidase conjugate alone (Fig. 2A, control). inhibit nonspecific binding, incubated with biotinylated versi- The binding of the lectin to the 170- and 190-kDa bands was can lectin or antibodies in 0.3% bovine serum albumin in inhibited in the presence of EDTA (Fig. 2A). TTBS, washed with TTBS, incubated with avidin-conjugated The 170- and 190-kDa bands were also detected when blots horseradish peroxidase (Vectastain ABC Elite; Vector Lab- were probed with ELC-IgG recombinant fusion protein, and oratories) or peroxidase-conjugated secondary antibody, this binding was also calcium dependent (Fig. 2B). washed with TTBS, and visualized by chemiluminescence The 170- and 190-kDa React with Anti- (ECL; Amersham). For the lectin blots, all solutions contained Tenascin-R Antibodies. The sizes of the 170- and 190-kDa either 5 mM CaCl2 or 5 mM EDTA. glycoproteins we detected with the biotinylated lectin sug- Deglycosylation of Brain Extracts. Aliquots of brain extract gested to us that they might be tenascin-R (29). Furthermore, were precipitated in acetone, dissolved in appropriate buffers its presence in the brain and the reduction in apparent for enzyme digestions, and incubated with or without recom- molecular mass of "20 kDa by PNGase F treatment (see Fig. binant glycopeptide N-glycosidase (PNGase F; EC 3.5.1.52; CD New England Biolabs), recombinant protein fusion of Endo- 0 o c) En 0 ,B-N-acetylglucosaminidase (EC 3.2.1.96) and maltose-binding (0 protein (Endo Hf; New England Z61 Zc cC 0 Biolabs), or Vibrio cholerae CA O 2 a L c neuraminidase (EC 3.2.1.18; Boehringer Mannheim). The oD CD~ 0 buffers for these incubations were provided by the manufac- S B S B S B S B SB 0 turer or prepared according to the manufacturer's instructions. After a 12-h incubation, the samples were boiled in SDS/ PAGE sample buffer. Deglycosylation with trifluoromethane- 67 sulfonic acid (TFMS) was performed as described (25). 43 Precipitation of Tenascin-R with Versican Lectin. Brain 30 extract was incubated for 1 h on ice with biotinylated versican 20.1 lectin in the presence of 5 mM CaCl2 or 20 mM EDTA. 14.4 Streptavidin-conjugated agarose beads were then added, and the incubation was continued for 1 h at 4°C with gentle FIG. 1. Versican lectin domain binds specifically to fucose and rotation. The beads were washed four times in Tris-buffered GlcNAc. Culture supernatants from versican-lectin-domain- expressing CHO cells were incubated with monosaccharide- saline (TBS; 25 mM Tris-HCl, pH 7.4/0.88% NaCl/5 mM conjugated beads, the beads were pelleted by centrifugation, and beads CaCl2 or 20 mM EDTA) and boiled in SDS/PAGE sample (B) and supernatants (S) were analyzed for versican lectin by SDS/ buffer containing 2.5% 2-mercaptoethanol. The samples were PAGE and immunoblotting with anti-versican-lectin-domain-specific electrophoresed and electroblotted, and tenascin-R was de- antibodies. The positions of molecular mass markers (in kDa) are on tected as described above. the left. The control lane contains untreated culture supernatant. Downloaded by guest on September 29, 2021 10592 Biochemistry: Aspberg et al. Proc. Natl. Acad. Sci. USA 92 (1995)

± < bands. Immunoblotting with anti-tenascin-C antiserum gave +\ + < only weakly staining bands in the adult rat-brain extracts (data A: not shown); a strongly staining 230-kDa band and a weakly staining triplet between 210 and 220 kDa were seen in extract A >u >u °u B { cm prepared from brains of 11-day-old rats (Fig. 3A). The versican lectin did not bind tenascin-C in brain extracts prepared from either adult or 11-day-old rats (data not shown). To confirm the identity of the gel bands detected with the 203 versican lectin domain as tenascin-R, the ligand was precipi- 215- tated from brain extracts by employing biotinylated versican 105- lectin domain and streptavidin-conjugated agarose beads, and analyzed by immunoblotting. As shown in Fig. 4, the lectin 71 0651 - 43 domain-bound material reacted with anti-tenascin-R. No bands were detected in the lectin domain-bound fraction when 29 EDTA had been added to the brain extract prior to the 28 18- incubation with the lectin (Fig. 4, lane 3) or when the extract 14 was incubated with beads without the lectin domain (Fig. 4, FIG. 2. Calcium-dependent binding of versican lectin domain to lane 4). 170- and 190-kDa proteins in rat brain extracts. (A) Rat-brain extracts Chemical Deglycosylation of Tenascin-R Eliminates Versi- were electrophoresed, transferred to nitrocellulose, and probed with can Lectin Domain Binding. When all carbohydrates were biotinylated versican lectin domain (VcL) in the presence of 5 mM removed from brain-extract proteins through chemical degly- CaCl2 or EDTA. Bound lectin was detected with avidin-horseradish peroxidase, chemiluminiscent substrate, and autoradiography. In the cosylation with TFMS, versican lectin domain binding to control lane, the lectin was omitted to show nonspecific binding by tenascin-R was abolished (Fig. 5). Incubation with PNGase F, avidin-horseradish peroxidase. (B) As in A, but probed with versican which removes N-linked carbohydrates, reduced the apparent ELC-IgG fusion protein and horseradish peroxidase-conjugated anti- molecular mass of tenascin-R but did not decrease lectin human IgG antibodies. The control lane (IgG) was probed with human domain binding (Fig. 5). Incubation with Endo Hf had no IgG instead of ELC-IgG. effect on versican lectin binding, although concanavalin A staining of a parallel blot was decreased after the Endo Hf 5) are in agreement with the reported expression pattern of treatment (data not shown). V cholerae neuraminidase caused tenascin-R and the reduction of its molecular mass after a slight reduction in the apparent molecular mass without removal of N-linked carbohydrates (18). affecting lectin domain binding (Fig. 5). The effect of TFMS Immunoblotting of brain extracts fractionated by SDS/ did not seem to have been caused by degradation of the protein PAGE with either an anti-tenascin-R monoclonal antibody or backbones in the extracts, as judged from Ponceau S staining the biotinylated lectin domain resulted in virtually identical of the blots (data not shown) and the continued presence of the staining patterns with reduced samples (Fig. 3A). When un- 75- and 130-kDa avidin-binding bands in the TFMS-treated reduced samples were probed, the lectin and the anti-tenascin- lane (Fig. 5). The size of the tenascin-R core protein in the R antibody stained the same bands, but an additional band was TFMS treatment was evaluated by stopping the reaction after detected with the antibody (Fig. 3B). This band may represent different incubation times. As shown in Fig. 6B, the versican an alternative glycoform or splicing product (29) that is not reactive with the versican lectin domain and that is not seen in : the reducing gel because of comigration with one of the other M LUL o U E A u cr: + B n, + <: VP z Z m - U LU 0 4J 1 23 4 JJ1 4J1 -i 4-X C C C u 0 CI X > m > > u

203 - tV 203 -

105 - 105 - w 71 - _ w w 71 - 44 - - 28 - 44 Reducing Non red ucing 28 - FIG. 3. Versican lectin domain ligand comigrates with tenascin-R. FIG. 4. Precipitation of tenascin-R by biotinylated versican lectin Rat-brain extract was boiled in SDS/PAGE sample buffer with (A; domain and streptavidin beads. Brain extract was incubated with Reducing) or without (B; Nonreducing) 2-mercaptoethanol, electro- biotinylated versican lectin in the presence of 5 mM CaCl2 or 20 mM phoresed, transferred, and probed with biotinylated versican lectin EDTA or without versican lectin (beads only). The lectin was then domain (VcL), as in Fig. 2, with mouse monoclonal anti-tenascin-R collected by addition of streptavidin-conjugated agarose beads, which antibody #596 (anti-TN-R), or with polyclonal anti-tenascin C (anti- were then washed and boiled in SDS/PAGE sample buffer containing TN-C). Tenascin-C staining in adult rat-brain extract was very weak; reducing agent, and the solubilized sample was analyzed by immuno- thus, the anti-tenascin-C lane in A shows staining in postnatal day 11 blotting with the anti-tenascin-R monoclonal antibody. Lane 1, brain rat-brain extract, while the sample in all other lanes is adult rat-brain extract only; lane 2, brain extract, lectin, CaCl2, and beads; lane 3, brain extract. The control lane shows the nonspecific staining by avidin- extract, lectin, EDTA, and beads; and lane 4, brain extract, CaCl2, and horseradish peroxidase in the absence of versican lectin domain. beads. Downloaded by guest on September 29, 2021 Biochemistry: Aspberg et aL Proc. Natl. Acad. Sci. USA 92 (1995) 10593

0 cn A anti-versican B anti-tenascin R o C-Co ML U) co IO ECr13

F- w Z +_ +- +- PCL

GL

203- WM 105- 71 - ^s '- i ML 44- PCL 28 - FIG. 5. Effect of deglycosylation on versican lectin domain binding to brain extract proteins. Rat brain extract was chemically deglycosy- GL lated with TFMS or enzymatically deglycosylated with different glycosidases, electrophoresed under reducing conditions, and analyzed by lectin blotting as in Fig. 2. Aliquots of brain extract were incubated in the same buffer with (+) or without (-) enzyme or TFMS. The arrow indicates the position of the unaltered tenascin R. Endo H, Endo Hf. WM lectin domain binding to tenascin-R was lost after a 10-sec incubation in the acid, whereas anti-tenascin-R antibodies C Rabbit lgG D Mouse IgG detected a band with the size of the intact core polypeptide even after 120 min FIG. 7. Both versican and tenascin R are localized in the granular (Fig. 6A). layer of rat cerebellum. Adjacent cryosections of rat cerebellum were Versican and Tenascin-R Immunolocalization in Rat Cer- incubated with affinity-purified anti-versican antiserum (A) or anti- ebellum. Immunofluorescence staining of rat cerebellum tenascin-R monoclonal antibody (B) followed by fluorescein- showed versican immunoreactivity predominantly in the gran- conjugated secondary antibodies. Control sections, where the primary ular layer and to some extent also in the white matter tracts antibodies were replaced with rabbit (C) or mouse (D) IgG at the same (Fig. 7A), in agreement with previous reports (27, 30). Tena- concentrations, were completely devoid of staining. ML, molecular scin-R staining was mainly found in the granular layer, but layer; PCL, Purkinje cell layer; GL, granular layer; and WM, white staining could also be found in the molecular layer and to a matter. lesser extent in the white matter (Fig. 7B), also in agreement with previous findings (18). calcium-dependent manner to tenascin-R. The carbohydrate binding of versican lectin domain was demonstrated in two DISCUSSION ways. First, the lectin domain (as well as the ELC-IgG chimera; data not shown) bound to some, but not all, simple We show that the lectin domain of versican interacts with carbohydrates. This binding was dependent on divalent cat- fucose and GlcNAc and that it binds specifically and in a ions. Similar binding to simple has been observed for =A<,° -c S B ° '-cUS~~~~~~~~~~~~c the lectin domain of aggrecan, another member of the same A uln E E E proteoglycan family (12, 13), and recently also for the C- o o o o 0 OQOQQc~~~~~~~~~~~i o0 u~~~~~~~~~4aOc O O0o r terminal part of the chicken homolog of versican (14). In the latter study, only a fragment containing the ELC was found to be active in carbohydrate binding, whereas a smaller fragment composed of the epidermal growth factor-like and lectin ~~~~~~~~~~~~~~~~~...... fK domains was inactive. The reason for the inactivity of the 2 - 250 smaller fragment may be that, as a bacterially produced protein, it may not have folded correctly. We have _.,i ,,'.. .s! attempted :.,:...... *.-...... bacterial expression of the versican lectin domain in two 98 -...... - 98 different bacterial expression systems, but in both cases the has-been inactive In 64 - 64 product (unpublished results). contrast, 50 the eukaryotic expression system used here yielded versican lectin domain that has carbohydrate-binding activity. 36 - - 36 The second line of evidence supporting the function of the anti-TN-R VcL-overlay (Ca2+) versican lectin domain as a lectin is the loss of binding to tenascin-R in chemically deglycosylated brain extract. Thus, FIG. 6. Effect of chemical deglycosylation on tenascin-R versican the versican lectin domain would appear to be a lectin, the binding and core protein size. Aliquots of rat brain extract were of which to tenascin-R is calcium and deglycosylated with TFMS for the indicated times, electrophoresed binding dependent under reducing conditions, and analyzed by immunoblotting and lectin carbohydrate directed, as is characteristic of C-type lectins (9). overlay blot assay as in Fig. 2. (A) Anti-tenascin-R (anti-TN-R) However, these data regarding the possible carbohydrate immunoblot. (B) Versican lectin domain (VcL) binding to an identical nature of the versican lectin binding site in tenascin R have to filter to the one in A. be interpreted cautiously. At least one C-type lectin, the IgE

Downloaded by guest on September 29, 2021 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.S...... 10594 Biochemistry: Aspberg et al. Proc. Natl. Acad. Sci. USA 92 (1995) receptor FceR2, binds its ligand in a carbohydrate-independent 4. Yamada, H., Watanabe, K., Shimonaka, M. & Yamaguchi, Y. manner (31). Moreover, the chemically deglycosylated tenas- (1994) J. Bio. Chem. 269, 10119-10126. cin-R could have lost its ability to bind versican lectin domain as 5. Bevilacqua, M. P. & Nelson, R. M. (1993) J. Clin. Invest. 91, a result of a modification in the protein backbone. Thus, whether 379-387. the tenascin-R binding is purely carbohydrate mediated or might 6. Springer, T. A. (1994) Cell 76, 301-314. also depend on protein to 7. Varki, A. (1994) Proc. Natl. Acad. Sci. USA 91, 7390-7397. recognition remains be determined. 8. Lasky, L. A. (1992) Science 258, 964-969. The highly selective binding of the versican lectin to tena- 9. Drickamer, K. (1988) J. Biol. Chem. 263, 9557-9560. scin-R in brain extracts suggests that this interaction is phys- 10. Zimmermann, D. R. & Ruoslahti, E. (1989) EMBO J. 8, 2975- iologically significant. The similarity of the tissue localizations 2981. of versican and tenascin-R in cerebellum provides an addi- 11. Shinomura, T., Nishida, Y., Ito, K. & Kimata, K. (1993) J. Bio. tional indication of such significance. We found strong staining Chem. 268, 14461-14469. for both proteins in the granular layer of cerebellum and also 12. Saleque, S., Ruiz, N. & Drickamer, K. (1993) Glycobiology 3, detected both of them in the white matter. In addition, 185-190. tenascin-R, but not versican, staining was observed in the 13. Halberg, D. F., Proulx, G., Doege, K., Yamada, Y. & Drickamer, molecular layer with the anti-tenascin-R used in this study. K. (1988) J. Biol. Chem. 263, 9486-9490. This antibody reacts with all of the isoforms of tenascin-R (18). 14. Ujita, M., Shinomura, T., Ito, K., Kitagawa, Y. & Kimata, K. As one of the isoforms of tenascin-R detected in nonreducing (1994) J. Biol. Chem. 269, 27603-27609. 15. Grumet, M., Milev, P., Sakurai, T., Karthikeyan, L., Bourdon, M., SDS/PAGE did not react with versican lectin domain, it is Margolis, R. K. & Margolis, R. U. (1994) J. Biol. Chem. 269, possible that one or more of the tenascin-R isoforms are not 12142-12146. present in the molecular layer and that versican-binding tena- 16. Grumet, M., Flaccus, A. & Margolis, R. U. (1993) J. Cell Biol. scin-R codistributes with versican more completely than is 120, 815-824. apparent from our study. 17. Friedlander, D. R., Milev, P., Margolis, R. K., Margolis, R. U. & The various proteoglycan lectin domains may have different Grumet, M. (1994) J. Cell Bio. 125, 669-680. ligand specificities and functions; the versican lectin domain 18. Pesheva, P., Spiess, E. & Schachner, M. (1989) J. Cell Bio. 109, bound to both fucose and GlcNAc, whereas aggrecan lectin 1765-1778. domain binds to fucose and galactose but only weakly to 19. Krusius, T., Gehlsen, K. R. & Ruoslahti, E. (1987)J. Bio. Chem. GlcNAc (12, 13). Thus, the saccharide-binding specificities of 262, 13120-13125. the proteoglycan lectins may be related but distinct. As neu- 20. Rogelj, S., Weinberg, R. A., Fanning, P. & Klagsbrun, M. (1988) rocan may bind tenascin-C (15, 32, 33), versican and neurocan Nature (London) 331, 173-175. 21. Urlaub, G. & Chasin, L. A. (1980) Proc. Natl. Acad. Sci. USA 77, may each bind to a different member of the tenascin protein 4216-4220. family. Both tenascins and proteoglycans have been classified 22. Subramani, S., Mulligan, R. & Berg, P. (1981) Mol. Cell. Biol. 1, as antiadhesive and have been found to be expressed in barriers 854-864. of neuronal outgrowth in vivo (for review see ref. 34). Such 23. Argraves, W. S., Suzuki, S., Arai, H., Thompson, K., Piersch- specialized matrixes consisting of a proteoglycan, tenascin- bacher, M. D. & Ruoslahti, E. (1987) J. Cell Biol. 105, 1183-1190. family glycoprotein, and hyaluronan may be formed as a result 24. Aruffo, A., Stamenkovic, I., Melnick, M., Underhill, C. B. & of the lectin interactions described in this paper. Seed, B. (1990) Cell 61, 1303-1313. 25. Horvath, E., Edwards, A. M., Bell, J. C. & Braun, P. E. (1989) J. We thank Dr. Melitta Schachner for providing the anti-tenascin-R Neurosci. Res. 24, 398-401. antibody, Dr. Brian Seed for the IgG-fusion plasmid, and Drs. Hudson 26. Perides, G., Rahemtulla, F., Lane, W. S., Asher, R. A. & Big- Freeze and Minoru Fukuda for helpful discussions. We are also nami, A. (1992) J. Biol. Chem. 267, 23883-23887. grateful to our colleagues at the La Jolla Cancer Research Foundation 27. Bignami, A., Perides, G. & Rahemtulla, F. (1993)J. Neurosci. Res. for comments on the manuscript. This work was supported by Grants 34, 97-106. CA42507, CA28896, and Cancer Center Support Grant CA30199 28. Yamagata, M., Shinomura, T. & Kimata, K. (1993) Anat. Em- (E.R.) from the National Cancer Institute. A.A. was supported by bryol. 187, 433-444. fellowships from the Swedish Cancer Research Fund and the Swedish 29. Fuss, B., Wintergeist, E.-S., Bartsch, U. & Schachner, M. (1993) Institute. C.B. was supported by fellowships from the European J. Cell Bio. 120, 1237-1249. Molecular Organization (ALTF 129-1990) and the Swiss 30. Bignami, A. & Dahl, D. (1986) Proc. Natl. Acad. Sci. USA 83, National Foundation (823A-033322.92). 3518-3522. 31. Vercelli, D., Helm, B., Marsh, P., Padlan, E., Geha, R. S. & 1. LeBaron, R. G., Zimmermann, D. R. & Ruoslahti, E. (1992) J. Gould, H. (1989) Nature (London) 338, 649-651. Biol. Chem. 267, 10003-10010. 32. Hoffman, S. & Edelman, G. M. (1987) Proc. Natl. Acad. Sci. USA 2. Rauch, U., Karthikeyan, L., Mauel, P., Margolis, R. U. & Mar- 84, 2523-2527. golis, R. K. (1993) J. Biol. Chem. 267, 19536-19547. 33. Hoffman, S., Crossin, K. L. & Edelman, G. M. (1988)J. Cell Biol. 3. Doege, K. J., Sasaki, M., Kimura, T. & Yamada, Y. (1991)J. Biol. 106, 519-532. Chem. 266, 894-902. 34. Faissner, A. & Steindler, D. (1995) Glia 13, 233-254. Downloaded by guest on September 29, 2021