L-14 is a divalent, lactosamine-binding lectin expressed in many vertebrate tissues. In the rat nervous system, L-14 expression has been observed previously in restricted neuronal subsets within the dorsal root ganglia and spinal cord. In this study we report that L-14 is expressed by non-neuronal cells in the rat olfactory nerve. We demonstrate that L-14 binds and co-localizes with two ligands in the rat olfactory system: a β-lactosamine-containing glycolipid, and a putative member of the laminin family. The former is expressed on the surfaces of nascent olfactory axons originating from neuron cell bodies in the olfactory epithelium. The latter is present in the extracellular matrix of the axonal path leading to synaptic targets in the olfactory bulb.

In vitro, we find that recombinant L-14 promotes primary olfactory neuron adhesion to two laminin family members, and promotes intercellular adhesion. Both activities are dose-dependent, and are independent of integrinmediated mechanisms. We have thus found that L-14 can serve two distinct adhesive functions in vitro, and propose that L-14 in vivo can promote olfactory axon fasciculation by crosslinking adjacent axons and promote axonal adhesion to the extracellular matrix.

Complex carbohydrates are expressed in the vertebrate nervous system with varying degrees of spatiotemporal specificity, but the functions of these substances on neural cell surfaces remain largely unknown (Yamamoto and Schwarting, 1992). Rat dorsal root ganglion (DRG) neurons and their projections to the dorsal horn of the spinal cord constitute a system in which the co-expression of specific glycoconjugates and complementary binding proteins (lectins) has been characterized (Dodd and Jessell, 1985; Regan et al., 1986; Hynes et al., 1990). One observed pairing of carbohydrate and lectin is that of the lectin, L-14, and a terminal lactosamine epitope recognized by the monoclonal antibodies (mAbs), 1B2 and A5 (Regan et al., 1986). Since L-14 is a homodimer and thus divalent (Leffler and Barondes, 1986), Hynes et al. (1990) suggested that L-14 might promote fasciculation of functional classes of sensory axons.

L-14, also known as L-14-I (Gitt et al., 1992), has particular affinity for polylactosamine chains (Zhou and Cummings, 1990; Cooper et al., 1991), such as those expressed by ‘proto-typical’ laminin (Arumgham et al., 1986; Knibbs et al., 1989). Prototypical laminin purified from the Engelbreth-Holm-Swarm (EHS) mouse sarcoma is a heterotrimer composed of A-, B1-, and B2-chains (reviewed by Martin and Timpl, 1987). Isoforms of the B1-chain known as the S-chain, and of the A-chain known as the M-chain (which forms merosin in conjunction with the B-chains) have also been identified (Hunter et al., 1989; Ehrig et al., 1990). All five chains are observed in nervous tissue (Sanes et al., 1990). Many neuronal receptors for laminin-family members, primarily members of the integrin family, have been described to regulate axon outgrowth (reviewed by Reichardt et al., 1989, and Reichardt and Tomaselli, 1991). L-14, through its potential to crosslink polylactosamine in the extracellular matrix and lactosamine antigens on neuron cell surfaces, might also serve as a receptor that regulates axonal outgrowth. The plausibility of such a mechanism is supported by the report of Zhou and Cummings (1993) that L-14 can promote the adhesion of two cell lines to EHS laminin. A role for L-14 in neuronal axon fasciculation or matrix binding, however, has not been directly tested.

We have demonstrated, by in situ hybridization, that L-14 is produced by non-neuronal cells that populate the olfactory nerve pathway. By immunocytochemistry we have examined the expression of L-14, laminin family members, and the 1B2-epitope in the olfactory system. Within the lamina propria of the olfactory epithelium and the outer nerve layer of the olfactory bulb, we find that L-14 immunoreactivity colocalizes with both laminin-immunoreactivity and the axon-associated 1B2-epitope. Furthermore, in vitro, L-14 binds laminin, merosin, a laminin-related molecule present in olfactory homogenates, and a 1B2-immunoreactive olfactory glycolipid.

Hypothesizing that L-14 may mediate olfactory axon fasciculation by crosslinking adjacent axonal membranes that express the 1B2-epitope, and may serve a guidance function by crosslinking the surface of growing axons to olfactory laminin(s), primary olfactory neurons were cultured on a variety of extracellular matrix molecules and the effects of L-14 on neuronal behavior assayed. We report here that (i) L-14 promotes olfactory neuron adhesion to laminin and merosin, but not fibronectin, in a dose-dependent manner; and (ii) L-14-mediated adhesion appears to be independent of integrinmediated adhesion.

Generation of L-14 RNA probes and in situ hybridization

Poly(A)+ mRNA was isolated from rat olfactory tissue using total RNA separator and mRNA separator kits from Clontech Laboratories. Reverse transcription-polymerase chain reactions (PCR) were performed using a Perkin Elmer Cetus kit. L-14 PCR primers were synthesized that correspond to bases 12-31 (primer A2) and 451-471 (primer B1) of the sequence reported by Clerch et al. (1988) (plus flanking sequences containing PstI and XbaI sites, respectively): A2 = 5′ GGCTGCAGTCGCAGGAATCTCTTCGCTT 3′, and B1 = 5′ GGTCTAGAGGGGCTACGGGCTGGGGCTGG 3′. PCR proceeded through 35 cycles of 1 minute at 95°C and 1 minute at 60°C, and resulted in a single amplified message that was cloned into the pIBI30 vector (International Biotechnologies). The insert was sequenced by the dideoxy chain termination technique using a ‘Sequenase’ kit (United States Biochemical), and the olfactory clone determined to be identical to previously cloned L-14.

In situ hybridizations were carried out by the method of Sassoon et al. (1988) as modified by Ressler et al. (1993) on 8 μm paraffin sections prepared from embryonic day 19 and postnatal day 2 rats. Amplified, purified, and linearized plasmid was transcribed using T3 and T7 RNA-polymerases to generate sense and antisense 35S-RNA probes. Sections were hybridized at 50°C for 16 hours, washed, dehy-drated, and dipped in NTB-2 autoradiography emulsion (Kodak). Slides were developed after exposure for 9 days at 4°C and counterstained with toluidine blue or cresyl violet.

Immunocytochemistry

Immunocytochemistry was performed as detailed in Schwarting et al. (1992a). Staining was performed with polyclonal anti-rat L-14 (Cooper et al., 1991), polyclonal anti-EHS laminin (Sigma), a mAb directed against the rat laminin B1-chain (C21) (Sanes and Chiu, 1983; Green et al., 1992) (gift of A. Y. Chiu, City of Hope), and the 1B2 mAb (Young et al., 1981).

Tissue homogenization for protein analysis

Homogenates of olfactory tissue (ectoturbinates, caudal nasal septum, cribriform plate, and olfactory bulbs) were prepared using a sequential homogenization technique described by Hunter et al. (1992). Briefly, olfactory tissue was dissected from 5 day old rats and stored at −20°C. Tissue was first homogenized using a Polytron tissue homogenizer in 25 mM Tris-HCl, pH 7.5 containing 0.2 mg/ml α-2 macroglobulin as a protease inhibitor. The homogenate was centrifuged at 15,000 g for 20 minutes, the supernatant discarded, and the pellet homogenized again in 20 mM EDTA and 25 mM Tris-HCl, pH 7.4, with 0.2 mM phenylmethyl-sulfonyl fluoride, 0.25% aprotinin, and 60 mM leupeptin (Sigma) as protease inhibitors. The homogenate was centrifuged as above, and the insoluble material was sequentially homogenized and centrifuged in 2 M MgCl2 in 25 mM Tris (pH 7.5); 2% Triton X-100 in 25 mM Tris (pH 7.5); 6 M guanidine-HCl in 25 mM Tris (pH 7.5); and 6 M guanidine-HCl with 10 mM dithiothreitol in 25 mM Tris (pH 7.5). Most of the laminin was extracted from the olfactory tissue in the 6 M guanidine-HCl in 25 mM Tris fraction. This was dialyzed into PBS, precipitated in 40% saturated ammonium sulfate, dissolved in PBS and dialyzed again, then frozen at −80°C.

Purification and radioiodination of recombinant L-14

Recombinant rat L-14 was produced and purified as described by Cooper et al. (1991). Radioiodination was performed as described by Leffler and Barondes (1986); for each reaction approximately 200 μg of L-14 was labeled using 250 μCi of Bolton-Hunter reagent (New England Nuclear). Unreacted and hydrolyzed Bolton-Hunter reagent was removed using a 5 ml ‘Presto’ de-salting column (Pierce Chemical) pre-equilibrated in 1 mg/ml radioimmunoassay-grade BSA (RIA-BSA, Sigma) in buffer A (see below). To eliminate lectin inactivated by radioiodination, labeled protein was repurified by lactose affinity chromatography with 1 ml (packed volume) of lactosyl-agarose beads (Pierce Chemical) with rocking, overnight, at 4°C. Beads were washed 5 times with 10 volumes of buffer A. 125I-L-14 was eluted with 100 mM lactose in buffer A, and concentrated and dialyzed out of lactose by ultrafiltration in a Centricon-10 (Amicon). Buffer A consists of 75 mM KH2PO4/Na2HPO4 (mixed to pH 7.2), 75 mM NaCl, 4 mM β-mercaptoethanol, and 2 mM EDTA (Leffler and Barondes, 1986).

Quantitative assay of L-14 binding to proteins

Proteins assayed for binding with 125I-L-14 were BSA (Sigma), bovine serum fibronectin (Sigma), EHS laminin (gift of A. Mercurio, Harvard Medical School), human merosin (initial studies were performed using material that was a gift of E. Engvall, La Jolla Cancer Research Foundation; subsequent studies used material from Chemicon), and olfactory homogenate. All proteins were applied to wells of a 96-well assay plate (Falcon) at 1 μg/well in a volume of 100 μl of PBS, except olfactory homogenate which was applied at 10 μg/well. After incubation at 37°C for 2 hours, wells were aspirated and blocked with 200 μl of 1 mg/ml RIA-BSA at 37°C for 1 hour. Wells were again aspirated, and incubated overnight at 4°C with 100 μl containing 106 cts per minute of 125I-L-14 in buffer A and 1 mg/ml RIA-BSA, with or without 100 mM lactose. The wells were aspirated, washed 3 times with 200 μl buffer A, dried, cut apart, and counted individually in a γ-radiation counter.

Antibody and L-14 binding to blotted proteins

EHS sarcoma laminin, human placental merosin, and olfactory homogenate were subjected to sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions (Laemmli, 1970), transferred to nitrocellulose paper in a wet electroblotter (Bio-Rad Scientific Plastics) in the buffer of Haid and Suissa (1983), and blocked in 1% BSA in phosphate-buffered saline (PBS). Nitrocellulose was then incubated for 1 hour in rabbit anti-EHS laminin (Sigma) diluted 1:1000 in 0.05% Tween-20 in PBS (Tween/PBS), or two different rabbit anti-merosin antisera (see below) diluted 1:500 in Tween/PBS. After 3 washes in Tween/PBS, nitrocellulose was incubated for 1 hour in HRP-coupled, goat antirabbit immunoglobulin G (Boehringer Mannheim) diluted 1:200 in Tween/PBS, washed 3 times, and developed with 4-chloro-naphthol. Anti-merosin antisera were a gift of E. Engvall and were made against whole human placental merosin (thus recognizing M- and B-chains), or against a synthetic peptide corresponding to a unique sequence in the second repeat of the G-like domain of the M-chain (Ehrig et al., 1990).

Alternatively, the nitrocellulose was incubated overnight at 4°C in 20 μg/ml of L-14 in Tween/PBS. In control studies, 100 mM lactose was included in the overnight incubation with L-14. After washing, the nitrocellulose was incubated with affinity purified anti-L-14 (1:200) for 8 hours at 4°C, followed by peroxidase conjugated goat anti-rabbit IgG as above and developed with 4-chloro-naphthol.

Detection of 1B2-, and L-14-binding glycolipids

Neutral glycolipids were extracted from the olfactory tissue of postnatal day 7 Sprague-Dawley rats as previously described (Suchy et al., 1988; Schwarting and Crandall, 1991). Lacto-N-neotetraosyl-ceramide (LNnT) was prepared from human erythrocytes in the same manner. To assay 1B2 binding, the glycolipid preparations were separated by thin layer chromatography (TLC) and prepared as described by Schwarting and Crandall (1991). For binding studies with 125I-L-14 lectin, approximately 2 ml of 2.5×105 cts per minute/ml labeled lectin (1 μg) in 1% RIA-BSA was incubated with the TLC plates at 4°C overnight. The plates were rinsed with PBS, dried, and exposed to X-Omat-AR film (Kodak) for 12-48 hours at −70°C. Control binding studies were performed under the same conditions except that 100 mM lactose was added to the binding buffer.

Cell culture

Primary cultures of fully dissociated olfactory neurons were prepared from postnatal day 1-3 rats as previously described (brain-derived neurotrophic factor was a gift of Amgen) (Mahanthappa and Schwarting, 1993). Substrates were prepared by cleaning 22 mm2 glass coverslips (Corning) as previously described, and coated with 100 μl of 25 μg/ml fibronectin, merosin, or laminin. Coverslips were incubated overnight at 37°C, blocked with 100 μl of 1 mg/ml RIA-grade BSA (Sigma) in Hank’s buffered salt solution (HBSS; Gibco) for 1 hour at 37°C, and rinsed three times with HBSS prior to cell plating. It is important that batches of laminin be tested for their ability to support cell attachment; we previously reported that rat olfactory neurons do not adhere to laminin, but lack of attachment appears to have been due to batch inactivity. Cells were plated in a volume of 100 μl/coverslip, and maintained in a humidified chamber within a 37°C, 5% CO2 incubator. Plating efficiencies were approximately 80%; cell density varied with the amount of starting tissue.

Coated and blocked coverslips were preincubated in the indicated quantities of L-14 in a volume of 100 μl for 2-4 hours at 4°C, and the unbound material rinsed off prior to final plating. By including 125I-L-14 in the preincubation mixture, the amounts of L-14 bound to the substrate after preincubation, and at the end of the culture period could be assessed (as in Table 1). Other additions made to the culture medium at the indicated concentrations were thiodigalactoside (Sigma), Arg-Gly-Asp-Ser peptide (RGDS; Sigma), and a rat antiserum directed against the β1 integrin subunit (Albelda et al., 1989) (gift of Dr C. Buck, Wistar Institute).

Table 1.

Quantification of L-14 bound to laminin-coated substrates before and after the plating of olfactory neurons

Quantification of L-14 bound to laminin-coated substrates before and after the plating of olfactory neurons
Quantification of L-14 bound to laminin-coated substrates before and after the plating of olfactory neurons

Immunostaining and cell counting

We have previously demonstrated that the anti-neural cell adhesion molecule mAb, AG1 (DiFiglia et al., 1989), specifically stains olfactory neurons in culture (Mahanthappa and Schwarting, 1993). Neuron cultures were fixed 14-18 hours after plating, and stained as previously described with the following modifications: 1:100 horse-radish peroxidase-conjugated goat anti-mouse immunoglobulin G (Boehringer Mannheim) was used as the secondary antibody, and the positive cells developed using 0.05% 3,3′-diaminobenzidine with 0.2% nickel ammonium sulfate in Tris-buffered saline.

All cell counting and photography was performed using bright-field microscopy at a total magnification of 100×. Randomly selected fields were scored for number of stained cells, and number of stained aggregates containing greater than 4 cells. Values presented for the number of adherent cells were normalized to the number adhering to laminin under control conditions on the given date. In the dose-response studies, each data point indicates the average value (± s.e.m.) derived from 18 fields (3 fields each from 6 separate cultures established on 2 different dates). Values presented in the studies comparing substrate effects were derived from 12 fields/condition (3 fields each from 4 cultures established on 2 dates), and those presented on the effects of integrin-perturbing reagents were derived from 6 fields/condition (3 fields each from 2 cultures established on the same day). Absolute numbers of adherent AG1+ cells ranged from 50-100 cells/field.

L-14 is synthesized by non-neuronal cells of the olfactory nerve, and co-localizes with laminin-immunoreactivity

In situ hybridization of an antisense RNA probe, corresponding to the entire coding region of rat L-14, to sections of rat olfactory tissue reveals that L-14 message is abundant throughout the olfactory nerve and lamina propria of the olfactory epithelium (Fig. 1A-B). L-14 message is completely absent from the neuroepithelium and thus olfactory neurons are not responsible for L-14 gene expression (Fig. 1E-F). It is thus likely that either glial cells or other mesenchymal cell types within the olfactory nerve are expressing the L-14 gene. Signal specificity is demonstrated by the lack of specific hybridization with a corresponding sense probe (Fig. 1C-D). Furthermore, hybridization patterns in rat DRG and spinal cord using the above anti-sense probe replicate previously observed L-14 hybridization patterns (Hynes et al., 1990; Poirier et al., 1992; Fig. 1G).

Fig. 1.

In situ hybridization of L-14 in the rat. Coronal sections (8 μm) cut through the rat nasal cavity and olfactory bulb were hybridized with 35S-labeled RNA probes transcribed from a cDNA containing the entire coding region of L-14 (see text for details). A and C are counterstained sections of E19 olfactory epithelium (oe) and olfactory bulb (ob) viewed in bright-field illumination. Dark-field illumination (B) of the section shown in A displays specific hybridization of anti-sense L-14 probe over the lamina propria (lp) adjacent to the oe, and the olfactory nerve layer (onl) surrounding the ob. Sense L-14 probe (D) shows no specific hybridization to the section shown in C. Higher magnification of a section from the P2 rat hybridized with anti-sense L-14 probe further demonstrates the restriction of L-14 expression to the lp, and its complete lack of expression within the oe proper (E,F). Hybridization of antisense L-14 probe to the dorsal root ganglia (drg) and motoneurons (mn) of the spinal cord (sc) replicates previously reported L-14 RNA localization and supports the specificity of binding observed in the olfactory system (G). Scale bar, 100 μm (A-D); 50 μm (E-G).

Fig. 1.

In situ hybridization of L-14 in the rat. Coronal sections (8 μm) cut through the rat nasal cavity and olfactory bulb were hybridized with 35S-labeled RNA probes transcribed from a cDNA containing the entire coding region of L-14 (see text for details). A and C are counterstained sections of E19 olfactory epithelium (oe) and olfactory bulb (ob) viewed in bright-field illumination. Dark-field illumination (B) of the section shown in A displays specific hybridization of anti-sense L-14 probe over the lamina propria (lp) adjacent to the oe, and the olfactory nerve layer (onl) surrounding the ob. Sense L-14 probe (D) shows no specific hybridization to the section shown in C. Higher magnification of a section from the P2 rat hybridized with anti-sense L-14 probe further demonstrates the restriction of L-14 expression to the lp, and its complete lack of expression within the oe proper (E,F). Hybridization of antisense L-14 probe to the dorsal root ganglia (drg) and motoneurons (mn) of the spinal cord (sc) replicates previously reported L-14 RNA localization and supports the specificity of binding observed in the olfactory system (G). Scale bar, 100 μm (A-D); 50 μm (E-G).

Since L-14 has been previously determined to bind laminin, and to co-localize with it in muscle basement membranes (Zhou and Cummings, 1990; Cooper et al., 1991), the expression of laminin in olfactory extracellular matrix was examined. As detected with a polyclonal antiserum to the A-B1-B2 form, laminin is observed in the nerve layer of the olfactory bulb from E18 through adulthood (data not shown). Comparison of the localization of L-14 and laminin by immunofluorescence in neonatal rats (Fig. 2) demonstrates the overlapping expression of L-14 and laminin epitopes in the nerve layer of the olfactory bulb and in the lamina propria adjacent to the olfactory epithelium. L-14-immunoreactivity on the luminal surface of the olfactory epithelium is inconsistent and may be due to nonspecific interactions with mucosal oligosaccharides. Olfactory homogenate separated by reducing SDS-PAGE, and immunoblotted with an antiserum directed against recombinant L-14, reveals a single band of Mr 14.5×103 that co-migrates with recombinant L-14 (data not shown). Thus between E18 and P2, when the pattern of olfactory system connections is still developing, the L-14 gene is transcribed and translated by non-neuronal cells of the primary olfactory pathway, and the protein deposited in proximity to laminin immunoreactivity.

Fig. 2.

L-14 and laminin are both concentrated in the lamina propria and olfactory nerve layer of the P2 olfactory system. Immunocytochemistry of a coronal section through the rat nasal cavity and olfactory bulb shows expression of L-14 (A) in the lamina propria (lp) and olfactory nerve layer (onl). Little or no reactivity is seen on cells within the olfactory epithelium (oe) itself or in neuronal regions of the olfactory bulb (ob). Laminin expression (B) is also present throughout the lp and onl, but is absent from the oe and ob. Scale bar, 50 μm.

Fig. 2.

L-14 and laminin are both concentrated in the lamina propria and olfactory nerve layer of the P2 olfactory system. Immunocytochemistry of a coronal section through the rat nasal cavity and olfactory bulb shows expression of L-14 (A) in the lamina propria (lp) and olfactory nerve layer (onl). Little or no reactivity is seen on cells within the olfactory epithelium (oe) itself or in neuronal regions of the olfactory bulb (ob). Laminin expression (B) is also present throughout the lp and onl, but is absent from the oe and ob. Scale bar, 50 μm.

L-14 binds laminin family members and to merosin-immunoreactive peptides derived from olfactory tissue

Co-localization of L-14 and laminin-immunoreactivity in the olfactory nerve raises the possibility that L-14 might functionally interact with olfactory forms of laminin. Thus the binding of 125I-L-14 to various purified proteins and olfactory homogenate was measured in 96-well plates. Binding was measured in the presence and absence of 100 mM lactose in order to distinguish sugar-dependent binding, and sugar-independent binding (i.e. non-specific binding). 125I-L-14 displays significant binding to laminin purified from EHS sarcoma (composed of A-, B1-, and B2-chains), and merosin purified from human placenta (composed of M-, and B-chains) (Fig. 3). The level of 125I-L-14 binding to all proteins tested is reduced to background in the presence of lactose; fibronectin and BSA show no specific binding. The presence of an L-14 ligand in total olfactory homogenate is demonstrated by a significant level of lactose-inhibitable binding (Fig. 3). Thus two members of the laminin family bind significant quantities of L-14, and olfactory protein homogenate contains an L-14 ligand.

Fig. 3.

L-14 binds to laminin, merosin, and olfactory protein in a lactose-inhibitable manner. BSA, bovine serum fibronectin, EHS sarcoma laminin, human placental merosin, and rat olfactory homogenate were applied to a microassay plate (1 μg protein/well, except olfactory homogenate at 10 μg/well) and incubated with 125I-L-14 +/− 100 mM lactose. Identical levels of L-14 bind to all proteins tested in the presence of lactose.

Fig. 3.

L-14 binds to laminin, merosin, and olfactory protein in a lactose-inhibitable manner. BSA, bovine serum fibronectin, EHS sarcoma laminin, human placental merosin, and rat olfactory homogenate were applied to a microassay plate (1 μg protein/well, except olfactory homogenate at 10 μg/well) and incubated with 125I-L-14 +/− 100 mM lactose. Identical levels of L-14 bind to all proteins tested in the presence of lactose.

To examine further the expression of laminin family members in the olfactory system, laminin, merosin, and guanidine-HCl solubilized olfactory proteins were subjected to reducing, 4-20% gradient SDS-PAGE, and transferred to nitrocellulose. Parallel strips of nitrocellulose were exposed to antisera directed against EHS laminin, human merosin, or incubated with L-14 followed by anti-L-14. In EHS laminin, anti-laminin recognizes bands of Mr appropriate for the A-chain (approximately 400×103) and the B-chains (approxi-mately 200×103) (Fig. 4, left). B-chains could not be separated from each other on such gradient gels. In merosin, a het-erotrimeric complex of M- and B-chains (Ehrig et al., 1990), anti-laminin stains solely B-chains; the bands above 400×103 Mr are likely to be unreduced or aggregated merosin. The olfactory homogenate contains peptides that react with antilaminin and co-migrate with the B-chains of merosin. The olfactory homogenate does not contain intact A-chain.

Fig. 4.

Olfactory homogenates contain laminin-related glycoproteins. L, EHS sarcoma laminin; M, human placental merosin; and O, rat olfactory homogenate, were subjected to 4-20% SDS-PAGE, transferred to nitrocellulose, and probed. Anti-laminin antiserum (left) stains A, A-chain of laminin; B’s, B-chains of laminin; the B-chains of merosin; and olfactory peptides that co-migrate with the B-chain of merosin. In addition, the human merosin preparation contains antibody-reactive material migrating above the A-chain that is probably unreduced or aggregated merosin. Anti-merosin antiserum (center) stains M, intact M-chain of merosin, having a Mr of 300×103 and 3 bands between Mr of 180 and 240×103. The merosin aggregate is also labeled. Anti-merosin reacts with a ladder of bands in the olfactory homogenate, most prominently a band of approximately 240×103Mr, which is adjacent to one of the major merosin peptides. L-14 (right) bound to all chains of laminin, all chains of merosin, and 3 bands in the olfactory homogenate (arrows) that migrate near the 200×103Mr band, in the region of B-chains and M-chain peptides. In addition, L-14 binds to 4 smaller olfactory peptides (arrowheads) of between Mr 80 and 150×103.

Fig. 4.

Olfactory homogenates contain laminin-related glycoproteins. L, EHS sarcoma laminin; M, human placental merosin; and O, rat olfactory homogenate, were subjected to 4-20% SDS-PAGE, transferred to nitrocellulose, and probed. Anti-laminin antiserum (left) stains A, A-chain of laminin; B’s, B-chains of laminin; the B-chains of merosin; and olfactory peptides that co-migrate with the B-chain of merosin. In addition, the human merosin preparation contains antibody-reactive material migrating above the A-chain that is probably unreduced or aggregated merosin. Anti-merosin antiserum (center) stains M, intact M-chain of merosin, having a Mr of 300×103 and 3 bands between Mr of 180 and 240×103. The merosin aggregate is also labeled. Anti-merosin reacts with a ladder of bands in the olfactory homogenate, most prominently a band of approximately 240×103Mr, which is adjacent to one of the major merosin peptides. L-14 (right) bound to all chains of laminin, all chains of merosin, and 3 bands in the olfactory homogenate (arrows) that migrate near the 200×103Mr band, in the region of B-chains and M-chain peptides. In addition, L-14 binds to 4 smaller olfactory peptides (arrowheads) of between Mr 80 and 150×103.

The anti-merosin blots were probed with an antiserum directed against a synthetic peptide corresponding to a unique region within the M-chain. As described by Ehrig et al. (1990), this antiserum does not recognize any peptides in EHS laminin. It does, however, bind very strongly to a 300×103 Mr M-chain in placental merosin and to several apparent breakdown products between 180 and 240×103 Mr (Fig. 4, center). Again, the presumed merosin aggregate at the top of the blot is also stained. In olfactory homogenate, this antiserum recognizes multiple bands: the most intensely stained band appears at approximately at the position of 240×103 Mr adjacent to one of the major merosin peptides, while fainter bands appear between 40 and 150×103 Mr (Fig. 4, center). Blots using different anti-merosin antisera generate similar patterns with prominent bands of identical mobility (data not shown). Thus these bands are not likely to be the result of non-specific binding, and probably represent proteolytic products of M-chain.

Nitrocellulose blots probed with L-14 followed by anti-L-14 show lower background than those probed with 125I-L-14. L-14 binds all subunits of purified laminin and merosin (Fig. 4, right). In olfactory homogenate, L-14 binds 3 distinct bands in the area of 200×103 Mr (arrows) that correspond to L-14-reactive bands in the merosin lane. These L-14 binding peptides are likely to be either B-chains or proteolytic products of the M-chain as both migrate in this vicinity. L-14 binds to four additional bands (arrowheads) of between 80 and 150×103 Mr, but does not bind to the 6 anti-merosin reactive bands below Mr of 80×103. These smaller peptides presumably do not contain carbohydrate epitopes recognized by L-14. In the presence of 100 mM lactose, there is no detectable L-14 binding to any of the proteins tested (data not shown). These data suggest that an L-14-binding laminin isoform is present in olfactory tissue, but due to proteolysis similar to that observed for placental M-chain (Leivo and Engvall, 1988), the olfactory proteins remain to be definitively identified.

L-14 binds an axonal glycolipid that is expressed within the olfactory nerve

In previous studies it was demonstrated that the 1B2 mAb reacts with terminal lactosamine structures (Young et al., 1981), and recognizes a glycolipid expressed by nascent olfactory axons in vivo (Schwarting and Crandall, 1991) and in vitro (Mahanthappa and Schwarting, 1993). Double-label immunofluorescence studies in the postnatal day 14 olfactory bulb (Fig. 5A,B) reveal nearly absolute co-localization of L-14 and laminin B1-chain in axon tracks of the nerve layer. There is thus a continuous band of laminin and L-14 immunoreactivity encircling the olfactory bulb, and there are irregular branch points where these proteins are deposited in tracks leading toward the glomerular layer. These laminin-containing tracks never extend into the glomeruli, and often extend only through part of the width of the nerve layer. In an adjacent section, which is double-labeled with 1B2 and antibodies least one form of laminin, are co-expressed within the outer nerve layer of the olfactory bulb.

Fig. 5.

L-14 immunoreactivity co-localizes with laminin deposition and with 1B2 antigens in the nerve layer of the P14 olfactory bulb. L-14 immunoreactivity (A) and laminin B1-chain immunoreactivity (B) are virtually identical (arrows) in the nerve layer of the olfactory bulb, using double-labeled immunofluorescence techniques. L-14 immunoreactivity (C) and 1B2 immunoreactivity (D) also co-localize (arrows) in tracks leading through the nerve layer to 1B2 immunoreactive glomeruli (arrowhead). gl, glomerular layer; onl, olfactory nerve layer. Scale bar, 50 μm.

Fig. 5.

L-14 immunoreactivity co-localizes with laminin deposition and with 1B2 antigens in the nerve layer of the P14 olfactory bulb. L-14 immunoreactivity (A) and laminin B1-chain immunoreactivity (B) are virtually identical (arrows) in the nerve layer of the olfactory bulb, using double-labeled immunofluorescence techniques. L-14 immunoreactivity (C) and 1B2 immunoreactivity (D) also co-localize (arrows) in tracks leading through the nerve layer to 1B2 immunoreactive glomeruli (arrowhead). gl, glomerular layer; onl, olfactory nerve layer. Scale bar, 50 μm.

To test directly whether L-14 can bind 1B2-glycolipid in a manner similar to laminin isoforms, neutral glycolipids extracted from olfactory homogenate were separated by TLC, and compared with a standard glycolipid isolated from human to L-14 (Fig. 5C,D), 1B2-immunoreac-tive axons (arrows) that extend through the nerve layer into 1B2-immunoreactive glomeruli (arrowhead) follow pathways that express L-14 (Fig. 5C) and B1-chain-immunoreactivity (Fig. 5B). Thus L-14, 1B2-glycolipid, and at erythrocytes, LNnT. It has previously been demonstrated that 1B2 antibodies react preferentially with LNnT and other glycolipids that terminate in galactose-(β1→4)-N-acetyl-glucosamine residues (Young et al., 1981). Orcinol staining of TLC plates for total glycolipids reveals a large number of bands in the olfactory preparation (Fig. 6, left). TLC-immunostaining with 1B2, however, reveals one minor immunoreactive glycolipid (Fig. 6, center). Binding studies using 125I-L-14 on TLC plates demonstrate that both LNnT and the 1B2-immunorective olfactory glycolipid bind L-14 (Fig. 6, right). The binding of 125I-L-14 to both glycolipids was inhibited by 100 mM lactose (data not shown). The fact that the olfactory glycolipid does not co-chromatograph exactly with LNnT from human erythrocytes is probably due to differences in ceramide composition.

Fig. 6.

A neutral glycolipid from olfactory tissue is bound by 1B2 and L-14. Total olfactory neutral glycolipids (Olf. Neutral) and lacto-N-neotetraosyl ceramide (LNnT), a positive control, were separated by thin layer chromatography (TLC) and stained with orcinol for total glycolipid (left). Identical chromatograms were reacted with 1B2 antibodies (center) and with 125I-L-14 (right). A minor component of the olfactory glycolipid mixture migrates in the vicinity of standard LNnT on TLC and binds both 1B2 and 125I-L-14.

Fig. 6.

A neutral glycolipid from olfactory tissue is bound by 1B2 and L-14. Total olfactory neutral glycolipids (Olf. Neutral) and lacto-N-neotetraosyl ceramide (LNnT), a positive control, were separated by thin layer chromatography (TLC) and stained with orcinol for total glycolipid (left). Identical chromatograms were reacted with 1B2 antibodies (center) and with 125I-L-14 (right). A minor component of the olfactory glycolipid mixture migrates in the vicinity of standard LNnT on TLC and binds both 1B2 and 125I-L-14.

L-14 in vitro can mediate olfactory neuron adhesion to laminin family members and neuron aggregation

To test the influence of L-14 on adhesion of olfactory neurons to laminin, laminin-coated coverslips were preincubated with quantities of L-14, ranging from 100 ng to 50 μg. Prior to neuron plating, unbound L-14 was rinsed off. Cultures were fixed 14-18 hours after plating, and stained with the AG1 mAb to mark neurons. By including 125I-L-14 in the preincubation solution, it was possible to determine the amount of L-14 bound to the substrate. Table 1 demonstrates that when ≤ 5 μg of L-14 is applied to the substrate, approximately 20% of L-14 present is bound; higher concentrations appear to saturate available laminin binding sites. At the end of the culture period, 20-50% of the L-14 initially bound at the time of plating remained bound; thus a dynamic equilibrium exists between substrate-bound and unbound forms.

At the time of staining, olfactory neurons are observed primarily as single cells, some of which have initiated axon outgrowth. Random fields in each culture condition were scored for number of adherent cells, and number of aggregates containing greater than four AG1+ cell bodies. Due to variability in plating density from date to date, adherent cell numbers were normalized in relation to the average number adhering in the control condition on the given date (control = 100). Aggregation is expressed as the average number of aggregates/100 adherent AG1+ cells. Neurite outgrowth was not observably affected by L-14 and was not scored after the initial experiments (data not shown).

As can be seen in Fig. 7A, the number of adherent cells increases as the amount of substrate-bound L-14 increases. This dose-dependent increase reaches a maximal value of 50% over control levels when the laminin substrate is preincubated in ≥ 5 μg of L-14. As the amount of L-14 present on the substrate increases, the number of aggregates also increases approximately 8-fold when the substrate is preincubated with ≥ 10 μg of L-14 (Fig. 7B). Effects are maximal in the range of L-14 concentrations known to be saturating available laminin binding sites (Table 1). Thus the effects of L-14 on substrate adhesion and aggregation are dose-dependent.

Fig. 7.

The effects on cultured olfactory neurons of preincubating laminin substrates with L-14. Substrates were prepared by preincubating laminin-coated coverslips in the indicated quantities of L-14 for approximately 4 hours and then rinsing away unbound material. Fully dissociated olfactory neurons were cultured on the substrates 14-18 hours, stained for neural cell adhesion molecule, and the labeled cells scored at a total magnification of 100×. For each data point (± s.e.m.), n=18. The number of adherent cells in A was normalized so that control=100 (see text for details). Aggregates in B contained > 4 AG1+ cells.

Fig. 7.

The effects on cultured olfactory neurons of preincubating laminin substrates with L-14. Substrates were prepared by preincubating laminin-coated coverslips in the indicated quantities of L-14 for approximately 4 hours and then rinsing away unbound material. Fully dissociated olfactory neurons were cultured on the substrates 14-18 hours, stained for neural cell adhesion molecule, and the labeled cells scored at a total magnification of 100×. For each data point (± s.e.m.), n=18. The number of adherent cells in A was normalized so that control=100 (see text for details). Aggregates in B contained > 4 AG1+ cells.

To test the substrate specificity of the observed L-14 effects, olfactory neurons were cultured on laminin, merosin, and fibronectin. Neurons were plated on either pure substrate or substrate that had been preincubated with 10 μg of L-14 and rinsed. As can be seen in Fig. 8A, the control levels of neuron adhesion to all three substrates are similar. After substrate preincubation with L-14, however, levels of cell adhesion on merosin increase by approximately 50%, and on laminin increase by approximately 70%. Pre-incubation with L-14 had little effect on cell adhesion to fibronectin. The presence of 10 mM thiodigalactoside, a non-metabolizable competitive inhibitor of L-14 binding (Leffler and Barondes, 1986; Cooper et al., 1991), completely inhibits the ability of the lectin to promote substrate adhesion, but does not affect baseline levels of adhesion.

Fig. 8.

Adhesion (A) and aggregation (B) effects on laminin, merosin, and fibronectin. Three different substrates were treated with and without 10 μg of L-14 as described in Fig. 7, and olfactory neurons plated in the presence or absence of 10 mM thiodigalactoside (TDG). ‘Control’, L-14 and TDG; ‘TDG’, L-14 and TDG+; ‘L-14’, L-14+ and TDG; and ‘L-14/TDG’, L-14+ and TDG+. In A, 100 = control level of adhesion on laminin. Error bars represent the s.e.m. (n=12).

Fig. 8.

Adhesion (A) and aggregation (B) effects on laminin, merosin, and fibronectin. Three different substrates were treated with and without 10 μg of L-14 as described in Fig. 7, and olfactory neurons plated in the presence or absence of 10 mM thiodigalactoside (TDG). ‘Control’, L-14 and TDG; ‘TDG’, L-14 and TDG+; ‘L-14’, L-14+ and TDG; and ‘L-14/TDG’, L-14+ and TDG+. In A, 100 = control level of adhesion on laminin. Error bars represent the s.e.m. (n=12).

Interestingly, as observed in the dose-response study on laminin, increased adhesion to merosin is also accompanied by increased aggregation (Fig. 8B). Control levels of aggregation are higher on merosin than on either fibronectin or laminin. Nevertheless, the number of aggregates on merosin doubles in the presence of L-14, and increases 4-fold on laminin; aggregation induced by L-14 is inhibited by thiodigalactoside. Little aggregation activity is observed on fibronectin unless L-14 is included in the medium with the neurons (data not shown). Thus the ability of L-14 to mediate cell-substrate adhesion appears to be restricted to laminin-family members.

It has been reported that at least one glycolipid forms functional complexes with integrin-type receptors (Cheresh et al., 1987). Since the ligand of L-14 on olfactory neurons is a glycolipid, the possibility that L-14 acts as a modulator of integrin-mediated adhesion was tested. Substrates for this experiment were plain laminin and laminin preincubated with 10 μg of L-14. Olfactory neurons were plated in control medium, 250 μg/ml RGDS, or a 1:100 dilution of rat antiintegrin β1 subunit. All perturbations were performed with and without 10 mM thiodigalactoside. In the absence of substrate-bound L-14, RGDS reduces the number of adherent cells by 20-30%, and the anti-integrin antiserum reduces adhesion by 30-40% (Fig. 9A). When plated on laminin preincubated with L-14, the number of adherent cells increases by 40-50 cells in all conditions, an effect inhibited by thiodigalactoside. Thus, despite varying degrees of inhibition of integrin-mediated adhesion by two completely different reagents, L-14-induced increases hold roughly constant and appear independent of integrin activity. Neither RGDS nor anti-integrin antiserum affect neuron aggregation in a consistent manner (Fig. 9B), thus L-14-mediated aggregation also occurs independent of integrin activity.

Fig. 9.

The effects of integrin-perturbing reagents on L-14 mediated adhesion. Neurons were plated in 10 mM thiodigalactoside (TDG), 250 μg/ml RGDS (RGDS), and/or 1:100 rat anti-integrin β1 subunit (anti-integrin) as indicated. Substrates of untreated laminin (open bars), or laminin that had been preincubated in 10 μg of L-14 and rinsed prior to neuron plating (stippled bars) were used. Adherent cells (A) and aggregates (B) were scored as described in Fig. 7. Error bars represent the s.e.m. (n=6).

Fig. 9.

The effects of integrin-perturbing reagents on L-14 mediated adhesion. Neurons were plated in 10 mM thiodigalactoside (TDG), 250 μg/ml RGDS (RGDS), and/or 1:100 rat anti-integrin β1 subunit (anti-integrin) as indicated. Substrates of untreated laminin (open bars), or laminin that had been preincubated in 10 μg of L-14 and rinsed prior to neuron plating (stippled bars) were used. Adherent cells (A) and aggregates (B) were scored as described in Fig. 7. Error bars represent the s.e.m. (n=6).

Carbohydrates and endogenous lectins in the nervous system

Though a variety of carbohydrates have been characterized in the nervous system of several animal species, few have been directly implicated in functional interactions. The studies of polysialic acid by Landmesser, Rutishauser, and their colleagues are notable for their direct demonstration of the role of this carbohydrate as a negative regulator of interactions requiring intercellular contact (Landmesser et al., 1990; Rutishauser and Landmesser, 1991). Zipser and Cole (1991) have demonstrated that a mannose-containing epitope on the leech axonal membrane protein, Lan3-2, is a required element in defasciculation that precedes arborization of Lan3-2+ axons within the synaptic neuropil.

Their model predicts the presence of a mannose-binding lectin activity in the leech nervous system that is yet to be identified. In the rat cerebellum, Zanetta and his colleagues have characterized two distinct mannose-binding lectins that may play roles in neuronal migration and myelination, but the ligands of these lectins have not been rigorously identified (Zanetta et al., 1992). There is also evidence that L-14 is expressed in a subset of DRG neurons that terminate in laminae I and II of the dorsal horn of the spinal cord. Undefined complementary lactoseries glycoconjugates recognized by 1B2 are specifically expressed within these laminae and the DRG axons innervating them. Thus L-14 has been suggested to play a role in axon guidance (Dodd and Jessell, 1985; Regan et al., 1986; Hynes et al., 1990).

Carbohydrate-dependent interactions involving laminin

Laminin itself has a lectin-like ability to bind sulfated carbohydrates such as sulfoglucuronyl glycolipids (Mohan et al., 1990), sulfatide, and heparin (Taraboletti et al., 1990). In cere-bellar explant cultures, each of these three sulfated glycoconjugates inhibit cell migration and neurite outgrowth, possibly via interactions with laminin (Kunemund et al., 1988).

Laminin is also likely to play important roles in nervous system development as a glycoconjugate ligand for other car-bohydrate-binding proteins. In studies that utilized an ungly-cosylated form of EHS laminin, both a melanoma cell line and the PC12 cell line displayed normal substrate binding, but cell spreading and neurite outgrowth were impaired by the lack of laminin oliogosaccharides (Dean et al., 1990). Cell surface galactosyltransferase has been implicated as an endogenous receptor mediating carbohydrate-dependent interactions with laminin. Galactosyltransferase has been demonstrated to mediate neural crest cell migration on laminin (Runyan et al., 1986; Hathaway and Shur, 1992), and to mediate neurite outgrowth from PC12 cells and DRG neurons cultured on laminin (Begovac and Shur, 1990; Thomas et al., 1990). Thus a variety of cell-matrix interactions depend on recognition of laminin-associated carbohydrates by specific binding proteins.

L-14 as a mediator of laminin interactions

L-14 has been suggested to interact with cell surfaces and extracellular matrix to modulate a variety of cell physiological events in vitro, including adhesion and proliferation (Wells and Mallucci, 1991). However, its abundance and localization in extracellular matrix of both developing muscle (Cooper and Barondes, 1990; Cooper et al., 1991) and the olfactory system suggests that the primary role of L-14 in vivo is to modulate interactions with the extracellular matrix. This hypothesis is supported by the observation that L-14 demonstrates high affinity binding to laminin, a major component of many basement membranes (Zhou and Cummings, 1990; Cooper et al., 1991).

In a direct test of the ability of L-14 to modulate interactions with laminin, Cooper et al. (1991) cultured the mouse myoblast line, C2C12, in the presence of L-14. It was found that L-14 in the culture medium inhibited myoblast adhesion to laminin, but not to fibronectin. This inhibitory effect was blocked by the presence of thiodigalactoside or by endo-β-galactosidase digestion of the laminin-linked polylactosamine. Laminin was the only L-14 ligand identified in this culture paradigm, and it was suggested that L-14 bound to the substrate inhibited myoblast adhesion by sterically hindering interactions between laminin in the presence of L-14. This observation in conjunction with that of Cooper et al. (1991) suggests that the role of L-14 in cell adhesion to laminin is contingent on the relative expression of particular laminin receptors and L-14 ligands on the given cell surface.

L-14 in the rat olfactory system

By in situ hybridization it is clear that L-14 in the olfactory system is synthesized by the non-neuronal cells of the olfactory axon pathway. Furthermore, two L-14 ligands are also present in the olfactory axon pathway. One is a lactosamine-containing glycolipid that is expressed on a subset of olfactory neurons that are likely to be recently postmitotic and are in the process of extending axons into the olfactory bulb (Schwarting and Crandall, 1991; Schwarting et al., 1992b). The other ligand(s) appear to be members of the laminin family. Though prototypical laminin contains A- and B-chains (Timpl et al., 1979), A-chains could not be identified in olfactory homogenates. This observation is similar to the finding that forms of laminin lacking the A-chain are produced in culture by a Schwannoma and cell surface laminin receptors (Cooper et al., 1991). More recently, Zhou and Cummings (1993) have shown that two cell lines, demonstrated to express cell surface polylactosamine, displayed enhanced adhesion to laminin line (Davis et al., 1985; Edgar et al., 1988), and primary astro-cytes (Liesi and Risteli, 1989). However, antisera directed against a homologue of the A-chain, known as the M-chain (Ehrig et al., 1990), stained a series of peptides in olfactory homogenate that may represent breakdown products of M-chain or a related protein (Fig. 4B). Whether olfactory laminin contains M-chain, or other A-chain-like variants (Marinkovich et al., 1992) is currently under investigation.

Both the glycolipid and glycoprotein ligands co-localize with L-14 in the axonal tract leading from the olfactory epithelium to the glomeruli of the olfactory bulb. It is thus hypothesized that L-14, as a divalent lectin, could crosslink 1B2-immunoreactive axons to each other to promote fasciculation, and crosslink axonal surfaces to molecules of the lamininfamily as these axons elongate over extracellular matrix towards targets in olfactory bulb glomeruli. An attractive aspect of this model is the fact that 1B2-immunoreactive gly-colipid is not expressed on the majority of immature olfactory neurons, nor is it present on the majority of mature olfactory axons or terminals. In postnatal rats, the 1B2-epitope is restricted to axons that have recently entered the olfactory bulb along the superficial surface of the nerve layer, where they send branches through the central nerve layer and terminate in dispersed glomeruli. The vast majority (80-90%) of axons and glomeruli in adult olfactory bulbs do not express the 1B2-epitope and we presume that the epitope is lost with maturation (Schwarting and Crandall, 1991; Schwarting et al., 1992b). Thus, in the proposed model, L-14 would only facilitate interactions involving axons that are still growing and/or actively forming synapses in the olfactory bulb.

In order to test such a mechanism, the behavior of primary cultured olfactory neurons in response to recombi-nant L-14 was studied. In the first series of experiments a laminin substrate was preincubated in L-14, the unbound lectin rinsed away, and the neurons plated on the L-14/laminin complex. In such a paradigm, neuron-substrate binding increased in proportion to the concentration of L-14 bound to the laminin substrate. Repeating this experiment on three different protein substrates, it was found that L-14 promotes adhesion to laminin, and merosin to a lesser degree, but not to fibronectin. This result agrees with the previous observation that L-14 binds laminin and merosin, but not fibronectin. That the effect was dependent on carbohydrate binding was demonstrated by its abrogation in the presence of 10 mM thiodigalactoside.

Increased cell attachment to the substrate was accompanied by increased intercellular adhesion. This aggregation did not merely increase in proportion to the number of adherent cells, but was apparent when expressed as a ratio of the number of aggregates to the number of adhering cells. There are two general classes of mechanisms by which the lectin might promote intercellular adhesion. One possibility is that L-14-mediated adhesion to the substrate activates an independent mechanism responsible for aggregation. The second possibility is that laminin-family members act as ‘reservoirs’ of L-14. In such a model, L-14 can mediate substrate adhesion, but by dissociating from the substrate, can also become available to mediate crosslinking of adjacent cells. The second mechanism is made likely by two observations: (i) during the assay, amounts of L-14 that dissociate from the laminin substrate (50-80% of that originally bound) are sufficient to mediate cell aggregation (Table 1); and (ii) if preincubated L-14 is not rinsed away, aggregation is observed regardless of substrate type (data not shown).

Using two distinct perturbations of integrin receptors, the RGDS synthetic peptide (Ruoslahti and Pierschbacher, 1987; Reichardt and Tomaselli, 1991) and an anti-serum directed against the integrin β1 subunit (Albelda et al., 1989), olfactory neuron adhesion to laminin was partially blocked. Pretreatment of the substrate with L-14 gives rise to a consistent increase in cell adhesion however, and thus makes clear that L-14-mediated adhesion is a distinct mechanism from integrinmediated adhesion. That L-14-mediated adhesion works in an additive fashion with integrin-type interactions is shown schematically in Fig. 10.

Fig. 10.

Model of L-14 functions in vitro. In the control case (A), olfactory neurons adhere to the culture substrate through carbohydrate-independent mechanisms (e.g. integrin/RGDS). When the substrate has been preincubated in L-14 (B), both classical and L-14-mediated adhesion take place and adhesion is increased. Furthermore, L-14 that dissociates from the substrate becomes available to crosslink lactosamines on adjacent cells and thus increases aggregation as well.

Fig. 10.

Model of L-14 functions in vitro. In the control case (A), olfactory neurons adhere to the culture substrate through carbohydrate-independent mechanisms (e.g. integrin/RGDS). When the substrate has been preincubated in L-14 (B), both classical and L-14-mediated adhesion take place and adhesion is increased. Furthermore, L-14 that dissociates from the substrate becomes available to crosslink lactosamines on adjacent cells and thus increases aggregation as well.

Conclusions

Adhesive interaction among neurons are mediated by a great variety of molecules (Jessell, 1988). Intercellular adhesive interactions mediated by members of the immunoglobulin gene superfamily (Salzer and Colman, 1989) and the cadherin family (Takeichi, 1988, 1990) are among the best characterized. Studies of neuronal interactions with the extracellular matrix have focused primarily on the integrin family of matrix receptors (Reichardt and Tomaselli, 1991). Despite the discovery of carbohydrate-dependent adhesive interactions in the circulatory system mediated by the selectins (Brandley et al., 1990; Bevilacqua et al., 1991), carbohydrates in the nervous system have been studied primarily as modulators of known neural adhesion systems (Landmesser et al., 1990; Rutishauser and Landmesser, 1991).

In this study we have demonstrated that not only are L-14 and its ligands co-localized in the rat olfactory nerve, but that L-14 can promote both cell-cell and cell-matrix binding by primary rat olfactory neurons cultured on laminin-family members. L-14 thus represents a novel class of adhesion molecule. The adhesion is neither homophilic nor heterophilic in a classic sense, but functions through the crosslinking of related carbohydrate structures on multiple glycoconjugates. These glycoconjugates may be glycolipids or glycoproteins, and may be expressed on the cell surface or the extracellular matrix. Thus the L-14 adhesion mechanism displays a versatility unknown among other adhesion molecules.

We wish to thank E. Engvall for the very generous gift of purified merosin and anti-merosin antibodies, A. Y. Chiu for a panel of antilaminin mAbs, J. Dodd for fresh 1B2 hybridoma cells, C. Buck for a rat antiserum directed against the β1 integrin subunit, J. Miller and R. Rosenfeld of Amgen for BDNF, R. Dahl and F. Smith for their patient tutelage on molecular biological techniques, and K. J. Ressler and L.

B. Buck for advice and instruction on the practice of in situ hybridization. We also thank D. MacCorkle for her excellent technical support. This work was supported by an individual postdoctoral National Service Research Award to N. K. M. (NS09167); funds provided by the Cigarette and Tobacco Surtax Fund of the State of California through the Tobacco-Related Disease Research Program of the University of California (grant no. 2KT0032) to D. N. W. C.; a National Institutes of Health grants to G. S. (DC00953); and the Department of Mental Retardation of the Commonwealth of Massachusetts (contract no. 100220023SC).

Albelda
,
S. M.
,
Daise
,
M.
,
Levine
,
E. M.
and
Buck
,
C. A.
(
1989
).
Identification and characterization of cell-substratum adhesion receptors on cultured human endothelial cells
.
J. Clin. Invest
.
83
,
1992
2002
.
Arumgham
,
R. G.
,
Hsieh
,
T. C.-Y.
,
Tanzer
,
M. L.
and
Laine
,
R. A.
(
1986
).
Structures of the asparagine-linked sugar chains of laminin
.
Biochim. Biophys. Acta
883
,
112
126
.
Begovac
,
P. C.
and
Shur
,
B. D.
(
1990
).
Cell surface galactosyltransferase mediates the initiation of neurite outgrowth from PC12 cells on laminin
.
J. Cell Biol
.
110
,
461
470
.
Bevilacqua
,
M.
,
Butcher
,
E.
,
Furie
,
B.
,
Gallatin
,
M.
,
Gribone
,
M.
,
Harlan
,
J.
,
Kishimoto
,
K.
,
Lasky
,
L.
,
McEver
,
R.
,
Paulson
,
J.
,
Rosen
,
S.
,
Seed
,
B.
,
Siegelman
,
M.
,
Springer
,
T.
,
Stoolman
,
L.
,
Tedder
,
T.
,
Varki
,
A.
,
Wagner
,
D.
,
Weissman
,
I.
and
Zimmerman
,
G.
(
1991
).
Selectins: a family of adhesion receptors
.
Cell
67
,
233
.
Brandley
,
B. K.
,
Swiedler
,
S. J.
and
Robbins
,
P. W.
(
1990
).
Carbohydrate ligands of the LEC cell adhesion molecules
.
Cell
63
,
861
863
.
Cheresh
,
D. A.
,
Pytela
,
R.
,
Pierschbacher
,
M. D.
,
Klier
,
F. G.
,
Ruoslahti
,
E.
and
Reisfeld
,
R. A.
(
1987
).
An Arg-Gly-Asp-directed receptor on the surface of human melanoma cells exists in a divalent cation-dependent functional complex with the disialoganglioside GD2
.
J. Cell Biol
.
105
,
1163
1173
.
Clerch
,
L. B.
,
Whitney
,
P.
,
Hass
,
M.
,
Drew
,
K.
,
Miller
,
T.
,
Werner
,
R.
and
Massaro
,
D.
(
1988
).
Sequence of a full-length cDNA for rat lung β-galactoside-binding protein: primary and secondary structure of the lectin
.
Biochemistry
27
,
692
699
.
Cooper
,
D. N. W.
and
Barondes
,
S. H.
(
1990
).
Evidence for export of a muscle lectin from cytosol to extracellular matrix and for a novel secretory mechanism
.
J. Cell Biol
.
110
,
1681
1691
.
Cooper
,
D. N. W.
,
Massa
,
S. M.
and
Barondes
,
S. H.
(
1991
).
Endogenous muscle lectin inhibits myoblast adhesion to laminin
.
J. Cell Biol
.
115
,
1437
1448
.
Davis
,
G. E.
,
Manthorpe
,
M.
,
Engvall
,
E.
and
Varon
,
S.
(
1985
).
Isolation and characterization of rat Schwannoma neurite-promoting factor: evidence that the factor contains laminin
.
J. Neurosci
.
5
,
2662
2671
.
Dean
,
J. W.
,
Chandrasekaran
,
S.
and
Tanzer
,
M. L.
(
1990
).
A biological role of the carbohydrate moieties of laminin
.
J. Biol. Chem
.
265
,
12553
12562
.
DiFiglia
,
M.
,
Marshall
,
P.
,
Covault
,
J.
and
Yamamoto
,
M.
(
1989
).
Ultrastructural localization of molecular subtypes of immunoreactive neural cell adhesion molecules (NCAM) in the adult rodent striatum
.
J. Neurosci
.
9
,
4158
4168
.
Dodd
,
J.
and
Jessell
,
T. M.
(
1985
).
Lactoseries carbohydrates specify subsets of dorsal root ganglion neurons projecting to the superficial dorsal horn of rat spinal cord
.
J. Neurosci
.
5
,
3278
3294
.
Edgar
,
D.
,
Timpl
,
R.
and
Thoenen
,
H.
(
1988
).
Structural requirements for the stimulation of neurite outgrowth by two variants of laminin and their inhibition by antibodies
.
J. Cell Biol
.
106
,
1299
1306
.
Ehrig
,
K.
,
Leivo
,
I.
,
Argraves
,
W. S.
,
Ruoslahti
,
E.
and
Engvall
,
E.
(
1990
).
Merosin, a tissue-specific basement membrane protein, is a laminin-like protein
.
Proc. Natl. Acad. Sci. USA
87
,
3264
3268
.
Gitt
,
M. A.
,
Massa
,
S. M.
,
Leffler
,
H.
and
Barondes
,
S. H.
(
1992
).
Isolation and expression of a gene encoding L-14-II, a new human soluble lactose-binding lectin
.
J. Biol. Chem
.
267
,
10601
10606
.
Green
,
T. L.
,
Hunter
,
D. D.
,
Chan
,
W.
,
Merlie
,
J. P.
and
Sanes
,
J. R.
(
1992
).
Synthesis and assembly of the synaptic cleft protein s-laminin by cultured cells
.
J. Biol. Chem
.
267
,
2014
2022
.
Haid
,
A.
and
Suissa
,
M.
(
1983
).
Immunochemical identification of membrane proteins after sodium dodecyl sulfate-polyacrylamide gel electrophoresis
.
Meth. Enzymol
.
96
,
192
205
.
Hathaway
,
H. J.
and
Shur
,
B. D.
(
1992
).
Cell surface β1,4-galactosyltransferase functions during neural crest cell migration and neurulation in vivo
.
J. Cell Biol
.
117
,
369
382
.
Hunter
,
D. D.
,
Shah
,
V.
,
Merlie
,
J. P.
and
Sanes
,
J. R.
(
1989
).
A laminin-like adhesive protein concentrated in the synaptic cleft of the neuromuscular junction
.
Nature
338
,
229
234
.
Hunter
,
D.D.
,
Murphy
,
M.D.
,
Olsson
,
C.V.
and
Brunken
,
W.J.
(
1992
).
S-Laminin expression in adult and developing retinae: A potential cue for photoreceptor morphogenesis
.
Neuron
8
,
399
413
.
Hynes
,
M. A.
,
Gitt
,
M.
,
Barondes
,
S. H.
,
Jessell
,
T. M.
and
Buck
,
L. B.
(
1990
).
Selective expression of an endogenous lactose-binding lectin gene in subsets of central and peripheral neurons
.
J. Neurosci
.
10
,
1004
1013
.
Jessell
,
T. M.
(
1988
).
Adhesion molecules and the hierarchy of neural development
.
Neuron
1
,
3
13
.
Knibbs
,
R. N.
,
Perini
,
F.
and
Goldstein
,
I. J.
(
1989
).
Structure of the major conconavalin A reactive oligosaccharides of the extracellular matrix component laminin
.
Biochem
.
28
,
6379
6392
.
Kunemund
,
V.
,
Jungalwala
,
F. B.
,
Fischer
,
G.
,
Chou
,
D. H. K.
,
Keilhauer
,
G.
and
Schachner
,
M.
(
1988
).
The L2/HNK-1 carbohydrate of neural cell adhesion molecules is involved in cell interactions
.
J. Cell Biol
.
106
,
213
223
.
Laemmli
,
U. K.
(
1970
).
Cleavage of structural proteins during assembly of the head of the bacteriophage T4
.
Nature
227
,
680
685
.
Landmesser
,
L.
,
Dahm
,
L.
,
Tang
,
J.
and
Rutishauser
,
U.
(
1990
).
Polysialic acid as a regulator of intramuscular nerve branching during embryonic development
.
Neuron
4
,
655
667
.
Leffler
,
H.
and
Barondes
,
S. H.
(
1986
).
Specificity of binding of three soluble rat lung lectins to substituted and unsubstituted mammalian β-galactosides
.
J. Biol. Chem
.
261
,
10119
10126
.
Leivo
,
I.
and
Engvall
,
E.
(
1988
).
Merosin, a protein specific for basement membranes of Schwann cells, striated muscle, and trophoblast, is expressed late in nerve and muscle development
.
Proc. Natl. Acad. Sci. USA
85
,
1544
1548
.
Liesi
,
P.
and
Risteli
,
L.
(
1989
).
Glial cells of mammalian brain produce a variant form of laminin
.
Exp. Neurol
.
105
,
86
92
.
Mahanthappa
,
N. K.
and
Schwarting
,
G. A.
(
1993
).
Peptide growth factor control of olfactory neurogenesis and neuron survival in vitro: roles of EGF and TGF-βs
.
Neuron
10
,
293
305
.
Marinkovich
,
M. P.
,
Lunstrum
,
G. P.
,
Keene
,
D. R.
and
Burgeson
,
R. E.
(
1992
).
The dermal-epidermal junction of human skin contains a novel laminin variant
.
J. Cell Biol
.
119
,
695
703
.
Martin
,
G. R.
and
Timpl
,
R.
(
1987
).
Laminin and other basement membrane components
.
Ann. Rev. Cell Biol
.
3
,
57
85
.
Mohan
,
P. S.
,
Chou
,
D. K. H.
and
Jungalwala
,
F. B.
(
1990
).
Sulfoglucuronyl glycolipids bind laminin
.
J. Neurochem
.
54
,
2024
2031
.
Poirier
,
F.
,
Timmons
,
P. M.
,
Chan
,
C.-T. J.
,
Guenet
,
J.-L.
and
Rigby
,
P. W. J.
(
1992
).
Expression of the L14 lectin during mouse embryogenesis suggests multiple roles during pre- and post-implantation development
.
Development
115
,
143
155
.
Regan
,
L. J.
,
Dodd
,
J.
,
Barondes
,
S. H.
and
Jessell
,
T. M.
(
1986
).
Selective expression of endogenous lactose-binding lectins and lactoseries glycoconjugates in subsets of rat sensory neurons
.
Proc. Natl. Acad. Sci. USA
83
,
2248
2252
.
Reichardt
,
L. F.
,
Bixby
,
J. L.
,
Hall
,
D. E.
,
Ignatius
,
M. J.
,
Neugebauer
,
K. M.
and
Tomaselli
,
K. J.
(
1989
).
Integrins and cell adhesion molecules: neuronal receptors that regulate axon growth on extracellular matrices and cell surfaces
.
Dev. Neurosci
.
11
,
332
347
.
Reichardt
,
L. F.
and
Tomaselli
,
K. J.
(
1991
).
Extracellular matrix molecules and their receptors: functions in neural development
.
Ann. Rev. Neurosci
.
14
,
531
570
.
Ressler
,
K. J.
,
Sullivan
,
S. L.
and
Buck
,
L. B.
(
1993
).
A zonal organization of odorant receptor gene expression in the olfactory epithelium
.
Cell
73
,
597
609
.
Runyan
,
R. B.
,
Maxwell
,
G. D.
and
Shur
,
B. D.
(
1986
).
Evidence for a novel enzymatic mechanism of neural crest migration on extracellular glycoconjugate matrices
.
J. Cell Biol
.
102
,
432
441
.
Ruoslahti
,
E.
and
Pierschbacher
,
M. D.
(
1987
).
New perspectives in cell adhesion: RGD and integrins
.
Science
238
,
491
497
.
Rutishauser
,
U.
and
Landmesser
,
L.
(
1991
).
Polysialic acid on the surface of axons regulates patterns of normal and activity-dependent innervation
.
Trends Neurosci
.
14
,
528
532
.
Salzer
,
J. L.
and
Colman
,
D. R.
(
1989
).
Mechanisms of cell adhesion in the nervous system: role of the immunoglobulin gene superfamily
.
Dev. Neurosci
.
11
,
377
390
.
Sanes
,
J. R.
and
Chiu
,
A. Y.
(
1983
).
The basal lamina of the neuromuscular junction
.
Cold Spring Harb. Symp. Quant. Biol
.
48
,
667
678
.
Sanes
,
J. R.
,
Engvall
,
E.
,
Butkowski
,
R.
and
Hunter
,
D. D.
(
1990
).
Molecular heterogeneity of basal laminae: isoforms of laminin and collagen IV at the neuromuscular junction and elsewhere
.
J. Cell Biol
.
111
,
1685
1699
.
Sassoon
,
D. A.
,
Garner
,
I.
and
Buckingham
,
M.
(
1988
).
Transcripts of α-cardiac and α-skeletal actins are early markers for myogenesis in the mouse embryo
.
Development
104
,
155
164
.
Schwarting
,
G. A.
and
Crandall
,
J.E.
(
1991
).
Subsets of olfactory and vomeronasal sensory epithelial cells and axons revealed by monoclonal antibodies to carbohydrate antigens
.
Brain Res
.
547
,
239
248
.
Schwarting
,
G. A.
,
Deutsch
,
G.
,
Gattey
,
D. M.
and
Crandall
,
J. E.
(
1992a
).
Glycoconjugates are stage- and position-specific cell surface molecules in the developing olfactory system, 1: the CC1 immunoreactive glycolipid defines a rostrocaudal gradient in the rat vomeronasal system
.
J. Neurobiol
.
23
,
120
129
.
Schwarting
,
G. A.
,
Deutsch
,
G.
,
Gattey
,
D. M.
and
Crandall
,
J. E.
(
1992b
).
Glycoconjugates are stage- and position-specific cell surface molecules in the developing olfactory system, 2: unique carbohydrate antigens are topographic markers for selective projection patterns of olfactory axons
.
J. Neurobiol
.
23
,
130
142
.
Suchy
,
S. F.
,
Yamamoto
,
M.
,
Barbero
,
L.
and
Schwarting
,
G. A.
(
1988
).
A monoclonal antibody WCC4 recognizes a developmentally regulated ganglioside containing α-galactose and α-fucose present in the rat nervous system
.
Brain Res
.
440
,
25
34
.
Takeichi
,
M.
(
1988
).
The cadherins: cell-cell adhesion molecules controlling animal morphogenesis
.
Development
102
,
639
655
.
Takeichi
,
M.
(
1990
).
Cadherins: a molecular family important in selective cell-cell adhesion
.
Ann. Rev. Biochem
.
59
,
237
252
.
Taraboletti
,
G.
,
Rao
,
C. N.
,
Krutzsch
,
H. C.
,
Liotta
,
L. A.
and
Roberts
,
D. D.
(
1990
).
Sulfatide-binding domain of the laminin A chain
.
J. Biol. Chem
.
265
,
12253
12258
.
Thomas
,
W. A.
,
Schaefer
,
A. W.
and
Treadway
,
R. M.
(
1990
).
Galactosyl transferase-dependence of neurite outgrowth on substratum-bound laminin
.
Development
110
,
1101
1114
.
Timpl
,
R.
,
Rohde
,
H.
,
Robey
,
P. G.
,
Rennard
,
S. L.
,
Foidart
,
J.-M.
and
Martin
,
G. R.
(
1979
).
Laminin - a glycoprotein from basement membranes
.
J. Biol. Chem
.
254
,
9933
9937
.
Wells
,
V.
and
Mallucci
,
L.
(
1991
).
Identification of an autocrine negative growth factor: mouse beta-galactoside-binding protein is a cytostatic factor and cell growth regulator
.
Cell
64
,
91
97
.
Yamamoto
,
M.
and
Schwarting
,
G. A.
(
1992
).
Glycoconjugates in the development of the nervous system
. In
Cell Surface Carbohydrates and Cell Development. Boca Raton, FL: CRC Press
.
Young
,
W. W.
,
Portoukalian
,
J.
and
Hakamori
,
S.-I.
(
1981
).
Two monoclonal anticarbohydrate antibodies directed to glycosphingolipids with a lacto-N-glycosyl type II chain
.
J. Biol. Chem
.
256
,
10967
10972
.
Zanetta
,
J.-P.
,
Kuchler
,
S.
,
Lehmann
,
S.
,
Badache
,
A.
,
Maschke
,
S.
,
Marschal
,
P.
,
Dufourcq
,
P.
and
Vincendon
,
G.
(
1992
).
Cerebellar lectins
.
Intl Rev. Cytol
.
135
,
123
154
.
Zhou
,
Q.
and
Cummings
,
R. D.
(
1990
).
The S-type lectin from calf heart tissue binds selectively to the carbohydrate chains of laminin
.
Arch. Biochem. Biophys
.
281
,
27
35
.
Zhou
,
Q.
and
Cummings
,
R. D.
(
1993
).
L-14 lectin recognition of laminin and its promotion of in vitro cell adhesion
.
Arch. Biochem. Biophys
.
300
,
6
17
.
Zipser
,
B.
and
Cole
,
R. N.
(
1991
).
A mannose-specific recognition mediates the defasciculation of axons in the leech CNS
.
J. Neurosci
.
11
,
3471
3480
.

L-14, the β-galactose-binding lectin described in this manuscript, has been renamed, galectin-1, in accordance with suggestions in a February 25, 1994 letter to Cell 76, 597-598.