Portal de Periódicos da CAPES

Mouse Testican-2

2005; Elsevier BV; Volume: 280; Issue: 12 Linguagem: Inglês

10.1074/jbc.m414276200

1083-351X

Anke Schnepp, Patricia Komp Lindgren, Hanni Hülsmann, Stephan Kröger, Mats Paulsson, Ursula Hartmann,

Glycosylation and Glycoproteins Research

Mouse testican-2 was cloned, sequenced, and shown to be a proteoglycan with a multidomain structure closely similar to that of the human ortholog, previously described as a calcium binding extracellular matrix molecule of the BM-40/SPARC/osteonectin family (Vannahme, C., Schübel, S., Herud, M., Gösling, S., Hülsmann, H., Paulsson, M., Hartmann, U., and Maurer, P. (1999). J. Neurochem. 73, 12–20). Recombinant mouse testican-2 was used to prepare specific antibodies that allowed the detection of testican-2 in various brain structures but also in lung, testis, and in several endocrine glands. Although the testican-2 expressed in EBNA-293 cells carried both heparan sulfate and chondroitin/dermatan sulfate glycosaminoglycan chains, the tissue form always contained only heparan sulfate. Both tissue-derived and recombinant testican-2 carried N-linked glycans. Tissue-derived forms of testican-2 were detected as proteoglycans of varying size, whereas a portion of the molecules produced by EBNA-293 cells were core proteins, lacking glycosaminoglycans. Both the proteoglycan and core protein forms of testican-2 inhibited neurite extension from cultured primary cerebellar neurons and may play regulatory roles in the development of the central nervous system. Mouse testican-2 was cloned, sequenced, and shown to be a proteoglycan with a multidomain structure closely similar to that of the human ortholog, previously described as a calcium binding extracellular matrix molecule of the BM-40/SPARC/osteonectin family (Vannahme, C., Schübel, S., Herud, M., Gösling, S., Hülsmann, H., Paulsson, M., Hartmann, U., and Maurer, P. (1999). J. Neurochem. 73, 12–20). Recombinant mouse testican-2 was used to prepare specific antibodies that allowed the detection of testican-2 in various brain structures but also in lung, testis, and in several endocrine glands. Although the testican-2 expressed in EBNA-293 cells carried both heparan sulfate and chondroitin/dermatan sulfate glycosaminoglycan chains, the tissue form always contained only heparan sulfate. Both tissue-derived and recombinant testican-2 carried N-linked glycans. Tissue-derived forms of testican-2 were detected as proteoglycans of varying size, whereas a portion of the molecules produced by EBNA-293 cells were core proteins, lacking glycosaminoglycans. Both the proteoglycan and core protein forms of testican-2 inhibited neurite extension from cultured primary cerebellar neurons and may play regulatory roles in the development of the central nervous system. The testicans form a subgroup within the BM-40/SPARC/osteonectin family of modular extracellular proteins (1Hartmann U. Maurer P. Matrix Biol. 2001; 20: 23-35Crossref PubMed Scopus (108) Google Scholar) and have also been referred to as the SPOCKs (SPARC/osteonectin, CWCV and Kazal-like domains) (2Charbonnier F. Perin J.P. Mattei M.G. Camuzat A. Bonnet F. Gressin L. Alliel P.M. Genomics. 1998; 48: 377-380Crossref PubMed Scopus (29) Google Scholar). Their modular structure is characterized by an N-terminal testican-specific domain followed by the follistatin-like- and extracellular calcium-binding (EC) 1The abbreviations used are: EC, extracellular calcium binding; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight; MMP, matrix metalloproteinase.1The abbreviations used are: EC, extracellular calcium binding; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight; MMP, matrix metalloproteinase. domains that define the BM-40 family. Toward the C terminus, in contrast to most other members of this family, the testicans carry a thyroglobulin-like domain and a novel domain with two potential glycosaminoglycan attachment sites. For testican-1 a substitution with both chondroitin and heparan sulfate was shown (3Bonnet F. Perin J.P. Maillet P. Jolles P. Alliel P.M. Biochem. J. 1992; 288: 565-569Crossref PubMed Scopus (39) Google Scholar, 4Alliel P.M. Perin J.P. Jolles P. Bonnet F.J. Eur. J. Biochem. 1993; 214: 347-350Crossref PubMed Scopus (106) Google Scholar); for testican-1 and -2 it has been demonstrated that the EC domain is functional and can bind calcium (5Kohfeldt E. Maurer P. Vannahme C. Timpl R. FEBS Lett. 1997; 414: 557-561Crossref PubMed Scopus (201) Google Scholar, 6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar). Testican-1 was originally isolated as a proteolytic fragment from human seminal plasma (3Bonnet F. Perin J.P. Maillet P. Jolles P. Alliel P.M. Biochem. J. 1992; 288: 565-569Crossref PubMed Scopus (39) Google Scholar). In the mouse, testican-1 is predominantly expressed in the nervous system during embryonic development, and its expression correlates with periods of neuronal migration and axonal growth (7Charbonnier F. Chanoine C. Cifuentes-Diaz C. Gallien C.L. Rieger F. Alliel P.M. Perin J.P. Mech. Dev. 2000; 90: 317-321Crossref PubMed Scopus (18) Google Scholar). In adult mice, expression is restricted to the brain where it is located at post-synaptic densities (8Bonnet F. Perin J.P. Charbonnier F. Camuzat A. Roussel G. Nussbaum J.L. Alliel P.M. J. Biol. Chem. 1996; 271: 4373-4380Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar). In search for homologous members of the BM-40 family we identified cDNAs that were most similar to testican-1 and were hence termed testican-2 and -3 (6, 1). Testican-2 mRNA was detected in mouse brain, lung, and testis, and within the brain testican-2 was much more widely expressed than testican-1 (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar). Functional information on testicans is still scarce, but there are clear indications that these may serve in the regulation of extracellular protease cascades. When expression cloning was used to identify regulators of pro-matrix metalloproteinase (MMP)-2 processing mediated by membrane-type (MT) 1-MMP, a splice variant of testican-3, called N-Tes, was found (9Nakada M. Yamada A. Takino T. Miyamori H. Takahashi T. Yamashita J. Sato H. Cancer Res. 2001; 61: 8896-8902PubMed Google Scholar). Indeed, it was shown that testican-1 and testican-3 inhibit pro-MMP-2 activation by either MT1-MMP or MT3-MMP and that testican-2 abrogates the inhibition by the other family members (10Nakada M. Miyamori H. Yamashita J. Sato H. Cancer Res. 2003; 63: 3364-3369PubMed Google Scholar). Similarly, testican-1 has been shown to inhibit cathepsin L, but not cathepsin B (11Bocock J.P. Edgell C.J. Marr H.S. Erickson A.H. Eur. J. Biochem. 2003; 270: 4008-4015Crossref PubMed Scopus (43) Google Scholar). Testican-1 was also studied for a potential regulation of neuronal functions and was found to inhibit attachment and neurite outgrowth in cultures of N2a neuroblastoma cells (12Marr H.S. Edgell C.J. Matrix Biol. 2003; 22: 259-266Crossref PubMed Scopus (29) Google Scholar). In the present work we cloned mouse testican-2 and expressed it as a recombinant protein. This allowed us to study its structure, tissue distribution, and interactions with neurons. Isolation of Mouse Testican-2 cDNA Clones—A mouse brain cDNA library (ICR outbred strain, Harlan Sprague-Dawley, newborns; Stratagene) was screened with a 1.4-kb 32P-labeled PstI-BamHI fragment derived from the human testican-2 full-length cDNA (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar). Hybridization was carried out at 42 °C in 50% formamide, 5 × Denhardt′s solution (0.1% bovine serum albumin, 0.1% Ficoll 400, and 0.1% polyvinylpyrrolidone), 5 × SSPE (0.75 m sodium chloride, 50 mm sodium phosphate, pH 7.6, and 5 mm EDTA), 1.5% SDS, and 0.2 mg/ml salmon sperm DNA. Filters were washed twice for 15 min in 0.1 × SSC (15 mm sodium chloride, 1.5 mm sodium citrate, pH 7.5) and 0.1% SDS at 65 °C and exposed to an x-ray film. Positive plaques were excised and rescreened. Several positive plaques from the final rescreen were in vivo excised, yielding cDNA in the pBluescript vector. The plasmids were sequenced on both strands with flanking and internal primers using an ABI Prism Big Dye Terminator Cycle Sequencing Ready Reaction kit, and the products were resolved on an ABI Prism 377 Automated Sequencer (PE Biosystems). Nucleotide sequence analysis and homology searches were performed with the programs of the GCG package. Recombinant Expression of the Full-length Testican-2 Protein—A cDNA fragment spanning the full-length cDNA of murine testican-2 was generated by PCR using the forward primer 5′-GCC CGC TAG CCG AAG GCG ACG CCA AGG-3′ and the reverse primer 5′-CAA TGA CTG CGG CCG CTA TCT ACC AGA TGT AGC C-3′. Primers introduced the new restriction sites NheI and NotI, respectively. PCR-amplified cDNA was cloned in the pCRII vector (Invitrogen) and sequenced on both strands by cycle sequencing. A NheI/NotI restriction fragment of the pCRII/testican-2 plasmid was purified and cloned into the eukaryotic expression vector pCEP-Pu-His6-Myc-factor X (13Wuttke M. Müller S. Nitsche D.P. Paulsson M. Hanisch F.-G. Maurer P. J. Biol. Chem. 2001; 276: 36839-36848Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar). Correct insertion of the fragment in the pCEP-Pu/testican-2 construct was verified by sequencing. Plasmids were transfected into the human embryonic kidney cell line EBNA-293 (Invitrogen) using an electroporator (Bio-Rad) according to the instructions of the manufacturer. Growth and selection of transfected cells were carried out as described (5Kohfeldt E. Maurer P. Vannahme C. Timpl R. FEBS Lett. 1997; 414: 557-561Crossref PubMed Scopus (201) Google Scholar). Conditioned serum-free medium of EBNA-293 cells (2.6 liters) was collected and passed over a nickel-nitrilotriacetic acid column (Qiagen). Testican-2 was eluted with a linear gradient of 0.005–0.25 m imidazole in 50 mm sodium phosphate, pH 8.0, containing 0.3 m sodium chloride. The testican-2-containing fractions were pooled and, after dialysis against 50 mm Tris-HCl, pH 8.6, submitted to ion exchange chromatography on a Resource Q column (Amersham Biosciences). The column was eluted with a linear NaCl gradient (0–1 m) and testican-2 detected in two peaks at 0.4 and 0.6 m NaCl. Preparation of Specific Antibodies against Mouse Testican-2—A mixture of equal amounts of the proteoglycan and core protein form of recombinant mouse testican-2, obtained after chromatography on a Resource Q column, was used to immunize a rabbit. The resulting antiserum was adsorbed by repeated passage through a column of CNBr-Sepharose (Amersham Biosciences) to which proteins extracted from EBNA-293 cells with 0.15 m NaCl, 50 mm Tris-HCl, 10 mm EDTA, pH 7.4, had been coupled. The adsorbed antiserum was then passed through a second affinity column with recombinant testican-2 coupled to CNBr-Sepharose and the bound antibodies to testican-2 eluted with 3 m KSCN, 50 mm Tris-HCl, pH 7.4, followed by rapid dialysis against 0.15 m NaCl, 50 mm Tris-HCl, pH 7.4. To block a remaining cross-reactivity with testican-1, 1 μg of purified recombinant testican-1/10 μg of antibody was added before use. Indirect Immunofluorescence Microscopy—Tissues from adult (older than 8 weeks) C57/BL6 mice were embedded in Tissue Tek, and 7-μm cryosections were prepared and stored at –20 °C. Sections were air dried for 1 h and fixed with 2% paraformaldehyde/phosphate-buffered saline for 20 min. After washing with 0.1% Triton X-100 in Tris-buffered saline and 0.1 m glycine in Triton/Tris-buffered saline, sections were blocked for 1 h with 10% bovine serum albumin in Triton/Tris-buffered saline. The affinity-purified antibodies against testican-2 were diluted in Triton/Tris-buffered saline containing 1% fetal calf serum and 1% bovine serum albumin and incubated on the sections at 4 °C overnight. Detection was done with Cy3-labeled goat anti-rabbit IgG (Jackson ImmunoResearch), and the slides were examined under a Zeiss Axiophot fluorescence microscope. Hematoxylin/eosin staining of parallel sections was done by standard methods. Preparation of Tissue Extracts—Tissues from adult (more than 8 weeks old) C57/BL6 mice were briefly homogenized in 5 vol (vol/weight) 0.15 m NaCl, 50 mm Tris-HCl, pH 7.4, containing 10 mm EDTA and a protease inhibitor mixture (Complete; Roche Applied Science) and stirred for 2–3 h at 4 °C. For extraction of lung, 4 m urea was included in the buffer. Supernatants were prepared by centrifugation at 21,500 × g for 30 min at 4 °C. Glycan Analysis, SDS-Polyacrylamide Gel Electrophoresis, and Immunoblotting—Tissue extracts or recombinant forms of testican-2 were digested with specific enzymes degrading glycosaminoglycans or N-glycans, respectively, to determine the nature of the glycosylation. In each case digestions were performed on aliquots corresponding to 50–200 ng of recombinant protein or extracts from 50–200 mg of tissue. Heparan sulfate was removed by simultaneous digestion with 0.1 mIU heparinase I (Sigma) and 0.5 mIU heparinase III (Sigma) in 50 mm NaCl, 50 mm Tris-HCl, 4 mm CaCl2, pH 7.5, at 37 °C overnight. Dermatan sulfate and chondroitin sulfate were degraded by digestion with 30 mIU chondroitinase ABC (Seikagaku) in 50 mm NaCl, 50 mm sodium acetate, 50 mm Tris-HCl, pH 7.5, at 37 °C for 4 h with an equal amount of enzyme added again after 2 h. Combined digestions were performed sequentially in the buffer given above for the heparinase treatment. For analysis of N-glycan content the samples were first treated with glycosaminoglycan-degrading enzymes as described above. Detergents were added to a final concentration of 0.1% SDS and 0.5% Nonidet P40, and the samples were digested with 1 (recombinant testican-2) or 2 (tissue extracts) IU of peptide N-glycosidase F (Roche Applied Science) overnight at 37 °C. SDS-polyacrylamide gel electrophoresis was done under reducing conditions in 4–15% polyacrylamide gels, and proteins were electrophoretically transferred to nitrocellulose. Testican-2 was detected with the affinity-purified primary antibodies followed by a swine anti-rabbit IgG-horseradish peroxidase conjugate (Dako). MALDI-TOF Mass Spectrometry—The samples were dissolved in 5 μl of 0.1% aqueous trifluoroacetic acid. MALDI-mass spectrometry was carried out in linear mode on a Bruker Reflex IV equipped with a video system, a nitrogen UV laser (λmax = 337 nm), and a HiMass detector. 1 μl of the sample solution was placed on the target, and 1 μl of a freshly prepared saturated solution of sinapinic acid in acetonitrile/H2O (2:1) with 0.1% trifluoroacetic acid was added. The spot was then recrystallized by addition of another 1 μl of acetonitrile/H2O (2:1) that resulted in a fine crystalline matrix. For recording of the spectra an acceleration voltage of 20 kV was used, and the detector voltage was adjusted to 1.9 kV. Approximately 500 single laser shots were summed into an accumulated spectrum. Calibration was carried out using the single and double protonated ion signal of bovine serum albumin for external calibration. Neurite Outgrowth Assays—Primary cerebellum cells were prepared from 6–7-day-postnatal C57/BL6 mice as previously described (14Trenkner E. Sidman R.L. J. Cell Biol. 1977; 75: 915-940Crossref PubMed Scopus (117) Google Scholar) and suspended in Dulbecco's modified Eagle's medium/F12 1:1, 0.5% glucose, 10% horse serum, 2 mm l-glutamine, 10 IU/ml penicillin, 100 IU/ml streptomycin (Invitrogen). Lab-Tek chamber slides (Nunc) were coated with 15 μl/ml poly-l-ornithine (0.01% sterile solution; Sigma) and incubated at 37 °C for 1 h. The wells were washed twice with distilled H2O, and laminin-1 from the mouse Engelbreth-Holm-Swarm tumor (15Paulsson M. Aumailley M. Deutzmann R. Timpl R. Beck K. Engel J. Eur. J. Biochem. 1987; 166: 11-19Crossref PubMed Scopus (329) Google Scholar) was added at 80 μg/ml and incubated at 37 °C overnight. Cells were seeded at a density of 1–5 × 104 cells/well either in pure medium or in medium supplemented with 20 μg/ml recombinant testican-2 core protein or proteoglycan and kept in culture at 37 °C for 48 h. After fixation in 4% paraformaldehyde for 10 min at room temperature, the neuronal cells were depicted by staining with a rabbit antiserum to mouse NCAM (kindly provided by Dr. C. Goridis, Paris) and a Cy3-labeled secondary antibody and examined by fluorescence microscopy. Cloning and Sequencing of Mouse Testican-2—A mouse brain cDNA library was screened with a 1.4-kb probe derived from the human full-length cDNA (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar), and three individual clones were isolated (Fig. 1A). One of these covered an open reading frame of 1522 bp, whereas the other two were shorter but overlapping and together covered the same sequence with differences to the first clone only in the length of the 5′ and 3′-untranslated regions. A conserved translational start site context (16Kozak M. Mamm. Genome. 1996; 7: 563-574Crossref PubMed Scopus (755) Google Scholar) surrounds the putative ATG initiation codon (Fig. 1B). No consensus polyadenylation sequence is found within the isolated 3′-untranslated region, suggesting an mRNA of >1.6 kb. The cDNA encodes a putative protein sequence of 423 amino acids. Sequence comparison showed that this protein is 93.8% identical to human testican-2 (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar) over its complete length, indicating that the mouse ortholog had been cloned. The overall domain structure is nearly identical to human testican-2. The N-terminal portion is predicted to function as a signal peptide. It contains a hydrophobic stretch of 22 amino acids ending with a consensus signal peptidase cleavage site (17Nielsen H. Engelbrecht J. Brunak S. von Heijne G. Protein Eng. 1997; 10: 1-6Crossref PubMed Scopus (4911) Google Scholar). The mature protein of 401 amino acids, with a calculated molecular mass of 44678 dalton, can be further subdivided into five domains. The N-terminal region (residues 23–88) does not show any homology to known proteins except testican-1 and -3 (4, 1). It is followed by a cysteine-rich domain (residues 89–192) homologous to follistatin. A third domain (residues 193–310) is homologous to the EC domain of BM-40 (18Hohenester E. Maurer P. Timpl R. Jansonius J.N. Engel J. Nat. Struct. Biol. 1996; 3: 67-73Crossref PubMed Scopus (132) Google Scholar), characterized by two Ca2+-binding EF-hand motifs that were shown to be functional in human testican-2 (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar). Following the EC domain, a thyroglobulin-like domain (residues 311–376) and a C-terminal domain (residues 377–423), unique to testicans, is found. The latter contains two putative glycosaminoglycan attachment sites at Ser-382 and Ser-387. Recombinant Expression and Purification of Mouse Testican-2—A cDNA fragment spanning the full-length cDNA was generated by PCR, cloned in the eukaryotic expression vector pCEP-Pu-His6-Myc-factor X (13Wuttke M. Müller S. Nitsche D.P. Paulsson M. Hanisch F.-G. Maurer P. J. Biol. Chem. 2001; 276: 36839-36848Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar), and expressed in the human embryonic kidney cell line EBNA-293. The recombinant testican-2 carrying an N-terminal His6-Myc tag was harvested from serum-free cell culture medium and purified by affinity chromatography on a column with chelated nickel. When the testican-2 secreted to the medium and the material purified by nickel affinity was analyzed by SDS-PAGE and immunoblot, two components were seen, a distinct band at ∼65 kDa and a diffuse smear in the range from 70 to 130 kDa (Fig. 2). As the polydispersity of the latter band indicated a substitution with glycosaminoglycan side chains, the affinity-purified material was further fractionated on a Resource Q ion exchanger. Here the 65-kDa protein, presumably representing the testican-2 core protein, was eluted at 0.4 m NaCl and the 70–130-kDa presumed proteoglycan form at 0.6 m (not shown). In a typical purification from 2.6 liters of conditioned medium, a yield of 2.6 mg of testican-2 core protein and 3.7 mg of testican-2 proteoglycan was obtained, but the proportion of the two forms of testican-2 varied between different medium batches. Expression of Testican-2 in Mouse Tissues—A mixture of equal amounts of testican-2 core protein and proteoglycan was used to immunize a rabbit. The antiserum obtained had a good titer, but initial experiments showed cross-reactions both with other members of the BM-40 protein family and with unrelated antigens (not shown). The cross-reacting antibodies could be removed only by use of a rather elaborate protocol for affinity purification, including an initial adsorption of the serum by repeated passage through a Sepharose column to which proteins from a crude extract of EBNA-293 cells had been coupled, followed by binding of the specific testican-2 antibodies to recombinant testican-2 coupled to Sepharose. The eluate from the testican-2-Sepharose still showed a moderate cross-reactivity with testican-1, but this activity could be blocked by the addition of recombinant testican-1 to the antibody dilution prior to use. All immunoblots and immunohistochemical experiments were performed with such highly testican-2-specific antibody preparations. The antibodies were used to assess the tissue distribution of testican-2 in adult mice. In previous work, we have shown that testican-2 mRNA is present in mouse brain, lung, and testis, with most parts of the brain being positive in in situ hybridization (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar). These results could now be confirmed by immunohistochemistry, showing that testican-2 is actively translated and deposited (Fig. 3). In brain sections testican-2 was seen associated with the cell bodies, axons, and sometimes also dendrites of large neurons such as the mitral cells in the olfactory bulb (Fig. 3A), the pyramidal cells in cerebral cortex (Fig. 3B), and hippocampus (Fig. 3C), and the Purkinje cells in the cerebellum (Fig. 3D). In addition, blood vessels were stained in all regions of the brain, and there was a widespread, diffuse staining that could not be found in other tissues. Interestingly, a particularly strong signal for testican-2 was detected in the pituitary gland (Fig. 3, E and F). All three anatomical regions were positive, with the strongest staining seen in pars intermedia, separating the adenohypophysis from the neurohypophysis. This finding, together with the known expression in testis, made us look specifically for the presence of testican-2 in other endocrine glands. Indeed, testican-2 is strongly expressed in the adrenal medulla, whereas the cortex is negative (Fig. 3, G and H). In the pancreas, the Langerhans islets were stained, but not the exocrine portions (Fig. 3, I and K). Within the islets testican-2 is ubiquitous, and it appears that both A- and B-cells are positive. In testis, testican-2 was deposited in basement membrane structures, showing a colocalization with nidogen-1 (Fig. 3, L–N). A weak punctate staining may be associated with Sertoli cells, spermatides, and sperms. Immunohistochemical analysis of ovaries showed only a weak signal in the theca layer (not shown). Heterogeneity in Glycosylation of Testican-2 Expressed by Established Cell Lines and in Different Mouse Tissues—Several established cell lines were tested for their expression of testican-2 by immunoblot of medium samples. Neuroblastoma cell lines N2a and N18TG2, a pheochromocytoma cell line derived from adrenal medulla (PC12), and a myoblast cell line (C2C12) showed clear positive reactions (Fig. 4). NG108 cells, hybrids between N18TG2 and C6 astrocytoma cells, showed a weak signal, whereas C6 cells alone were negative, indicating that testican-2 expression was due to the neuronal component. Cell lines derived from embryonal kidney cells (EBNA-293), fibroblasts (Wi26), and keratinocytes (HaCaT) were negative, indicating clear restrictions in testican-2 expression. The band patterns observed were heterogenous and varied between cell lines. In each case a polydisperse population of large molecules was observed, presumably representing testican-2 proteoglycans, but the differences in the character of these bands between different cellular sources indicated a cell-specific glycosylation. Sharper bands were also seen between 35 and 50 kDa representing either non-glycosylated core proteins or proteolytic cleavage products. The proportion of these bands appeared to increase with the age of the conditioned medium, an observation that would favor the latter alternative. Immunoblot analysis of mouse tissue extracts showed mainly the proteoglycan form of testican-2, with strong signals obtained in extracts of whole brain, pituitary, adrenal gland, and lung (Fig. 5). Again, the position and characteristics of the bands varied between tissues and were also not identical with that of recombinant testican-2 proteoglycan. It appears that the cell-specific glycosylation observed in established cell lines reflects differences present also in vivo. To verify that the observed variability in testican-2 structure is due to differential glycosylation, the nature of the glycans bound to recombinant and tissue-derived testican-2 was determined by enzymatic digestion. First the larger, more acidic form of recombinant testican-2 was digested with chondroitinase ABC or a mixture of heparinase I and III (Fig. 6A). Treatment with heparinase I/III gave a shift of the preparation to a smaller size without a sharp band resulting, whereas chondroitinase ABC digestion shifted a portion of the material to the position of the presumed core protein. When both enzymes were used, almost all testican-2 ran as a sharp 65-kDa core protein band. Apparently, recombinant testican-2 is a hybrid proteoglycan carrying both chondroitin/dermatan sulfate and heparan sulfate chains. A sample of the 65-kDa testican-2 core protein was digested with peptide N-glycosidase F, an endoglycosidase cleaving N-glycosidically linked glycan. A clear shift in molecular mass was seen both in SDS-polyacrylamide gel electrophoresis (Fig. 6B) and in MALDI-TOF mass spectrometry (Fig. 6C) with the latter method giving the masses 57545 Da before and 53538 Da after digestion. N-glycans are clearly present, and the difference of ∼4 kDa is rather large considering that only one N-glycan attachment site, located in the EC domain, was predicted from the sequence by the NetNGlyc software. The value of 53538 daltons for the core protein devoid of glycosaminoglycans and N-glycans is still considerably higher than the 49837 daltons calculated from the amino acid sequence of the recombinant protein. Four potential binding sites for O-glycans were identified by NetOGlyc software, all being present in the follistatin-like domain. Their substitution with O-glycans could explain this remaining discrepancy in molecular mass. Similar experiments were performed on testican-2 extracted from mouse brain, pituitary, and adrenal gland (Fig. 7). Here digestion with heparinase I/III in each case yielded a sharp core protein band, whereas chondroitinase ABC had no effect, indicating that the tissue forms of testican-2 are pure heparan sulfate proteoglycans (Fig. 7A). Sequential treatment of testican-2 from pituitary gland with heparinase I/III and peptide N-glycosidase F gave a further reduction in mass (Fig. 7B), showing the presence of N-glycans. Effects of Testican-2 on Laminin-supported Neurite Outgrowth—Several proteoglycans have been shown to regulate neuronal migration, neurite outgrowth, and axonal guidance, thus contributing to the patterning of the nervous system (1Hartmann U. Maurer P. Matrix Biol. 2001; 20: 23-35Crossref PubMed Scopus (108) Google Scholar, 19Oohira A. Katoh-Semba R. Watanabe E. Matsui F. Neurosci. Res. 1994; 20: 195-207Crossref PubMed Scopus (75) Google Scholar, 20Ruoslahti E. Glycobiology. 1996; 6: 489-492Crossref PubMed Scopus (359) Google Scholar, 21Margolis R.U. Margolis R.K. Cell Tissue Res. 1997; 290: 343-348Crossref PubMed Scopus (164) Google Scholar). To test for effects of testican-2 on neurite outgrowth, mouse primary cerebellum cells were seeded on slides coated with mouse laminin-1, a substrate known to induce neurite outgrowth (22Edgar D. Timpl R. Thoenen H. EMBO J. 1984; 3: 1463-1468Crossref PubMed Scopus (439) Google Scholar), in the presence or absence of testican-2. To determine whether the testican glycosaminoglycan chains play a role, either the core protein or the proteoglycan fraction was added. After a 48-h incubation the neuronal cells were depicted by using antibodies to the neuronal marker NCAM in indirect immunofluorescence microscopy (Fig. 8). In each case a similar number of cerebellum cells attached to the laminin-1 substrate, but significant differences were seen with regard to neurite extension. Control cultures grown on laminin-1 alone showed substantial neurite outgrowth and extension (Fig. 8A), whereas both the core protein (Fig. 8B) and proteoglycan form (Fig. 8C) of testican-2 caused a significant decrease in neurite growth when added in solution, indicating that the core protein is the active component. Mouse testican-2 cDNA was cloned and the full-length protein expressed in human embryonic kidney EBNA-293 cells. This allowed the production of the first antiserum against testican-2 and determination of the tissue distribution of testican-2 protein to complement information on the expression of the corresponding mRNA that we previously obtained by Northern blot and in situ hybridization (6Vannahme C. Schübel S. Herud M. Gösling S. Hülsmann H. Paulsson M. Hartmann U. Maurer P. J. Neurochem. 1999; 73: 12-20Crossref PubMed Scopus (54) Google Scholar). This is an important step in determining the nature of the extracellular structures in which testican-2 functions. In brain we could detect a widespread distribution with testican-2 antibodies labeling the cell surface of many types of neurons. In addition, immunohistochemistry showed a strong and unexpected expression in the pituitary gland and in other endocrine organs such as the pancreas, the adrenal gland, and the testis. Thus, with the exception of its presence in the lung, testican-2 is only expressed in neural or endocrine tissues in adult mice. At present the functional implications of this restricted expression are unclear, but the localization to the central nervous system is shared with testican-1, which may be a component of synaptic structures (8Bonnet F. Perin J.P. Charbonnier F. Camuzat A. Roussel G. Nussbaum J.L. Alliel P.M. J. Biol. Chem. 1996; 271: 4373-4380Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar). In most organs testican-2 appears to be present diffusely throughout the extracellular space or localized at the cell surface, indicating that the proteoglycan may interact with membrane-bound molecules. Interactions of proteoglycans with neuronal cell surface receptors were already shown for neurocan and phosphacan, which both bind to members of the immunoglobulin superfamily (23Milev P. Friedlander D.R. Sakurai T. Karthikeyan L. Flad M. Margolis R.K. Grumet M. Margolis R.U. J. Cell Biol. 1994; 127: 1703-1715Crossref PubMed Scopus (284) Google Scholar, 24Friedlander D.R. Milev P. Karthikeyan L. Margolis R.K. Margolis R.U. Grumet M. J. Cell Biol. 1994; 125: 669-680Crossref PubMed Scopus (389) Google Scholar). This binding is thought to modulate cell adhesion, neurite growth, and signal transduction during the development of the nervous system. We could demonstrate an interaction between testican-2 and neurons that resulted in an inhibition of neurite growth when testican-2 was added to primary cerebellar neurons cultured on laminin-1 as a permissive substrate. Similar results were obtained with both proteoglycan and core protein forms, a result pointing to the protein part carrying the biological activity. The mechanism by which neurons sense the presence of testican-2 has not yet been studied, but we aim to determine the nature of the putative testican-2 receptor as well as the physiological consequences of the in vivo testican-2-cell interactions. The inhibition of neurite growth may be a more general function of the testican family, because this effect can also be observed when testican-1 is added to cultured N2a neuroblastoma cells (12Marr H.S. Edgell C.J. Matrix Biol. 2003; 22: 259-266Crossref PubMed Scopus (29) Google Scholar). A different localization of testican-2 could be observed in testis. Here colocalization with nidogen-1 demonstrated that the proteoglycan, as previously shown for other members of the BM-40 family (25Dziadek M. Paulsson M. Aumailley M. Timpl R. Eur. J. Biochem. 1986; 161: 455-464Crossref PubMed Scopus (110) Google Scholar, 26Vannahme C. Smyth N. Miosge N. Gösling S. Frie C. Paulsson M. Maurer P. Hartmann U. J. Biol. Chem. 2002; 277: 37977-37986Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar), can also be associated with basement membranes. Whether the different localization also reflects a different function remains to be elucidated. The presence of a stretch of hydrophobic amino acids at the N terminus of testican-2, serving as a signal peptide, and the lack of a second hydrophobic domain already indicate that the protein is secreted from the cell. Here, we have shown that the signal peptide is functional and testican-2 is indeed present in the supernatants of different neuroblastoma cell lines (N2a, N18TG2, NG108) as well as cell lines derived from adrenal gland (PC12) or skeletal muscle satellite cells (C2C12). Recombinantly expressed testican-2 occurred in two forms, with one of these carrying glycosaminoglycan side chains and the other not. This may indicate that testican-2 is a "part-time" proteoglycan, but the analysis of endogenous testican-2 secreted by established cell lines or extracted from mouse tissues failed to convincingly show this for the in vivo form of testican-2. The occurrence of a glycosaminoglycan-free form of testican-2 produced in EBNA-293 cells may be due to a massive overexpression of the protein and a concomitant saturation of the biosynthetic enzyme machinery, because this was not observed when testican-2 was recombinantly expressed in other cell lines that produce lower amounts of protein (data not shown). Smaller forms of endogenous testican-2 were also detected in addition to the polydisperse proteoglycan forms, especially in cell culture supernatants where they may be products of proteolytic cleavage. Similarly, the proteoglycan form of testican-1 is converted to smaller stable forms in samples of human blood plasma (27BaSalamah M.A. Marr H.S. Duncan A.W. Edgell C.J. Biochem. Biophys. Res. Commun. 2001; 283: 1083-1090Crossref PubMed Scopus (18) Google Scholar). At present the physiological significance of testican cleavage is not clear, but for testican-1 it was proposed that size conversion may be important for activation or decay. Testican-2 from EBNA-293 cells and from mouse tissues differed in that the former contained both chondroitin/dermatan sulfate and heparan sulfate side chains whereas the latter are pure heparan sulfate proteoglycans. This does, however, not influence the effects seen for the EBNA-293 cell-derived testican-2 on neurite outgrowth as the activity resides in the core protein. Our future work will address the interaction of testican-2 with cells and the role of this proteoglycan in the development of the nervous system and of endocrine glands.