Measurement of Cell-Monolayer Adhesion in Cells Transfected with N-CAM cDNAs

To measure the adhesion of cells expressing the neural cell adhesion molecule N-CAM, mouse Lmtk fibroblast cells were transfected by a calcium phosphate precipitation technique with eucaryotic expression vectors encoding N-CAM polypeptides. We obtained cell lines expressing the 140-kDa transmembrane isoform of N-CAM at high levels by several rounds of selection by fluorescence-activated cell sorting and compared the adhesion of these cells to that of untransfected cells using a centrifugal removal assay that measures the centrifugal force required to remove radiolabeled probe cells from a cell monolayer. The adhesion of cells prepared from embryonic chicken neural retinas also was examined. Retinal probe cells remained associated with a retinal cell monolayer with an adhesive force of approximately 5 × 10^-6 dyn/cell, and this force was not reduced by treatment with specific anti-N-CAM antibody fragments. Transfected and untransfected mouse L cells each were dislodged from transfected cell monolayers with a removal force of 5 × 10^-5 dyn/cell and thus did not differ in their adhesion. These results support the hypothesis that N-CAM-mediated homophilic adhesion in retinal cells and transfected fibroblasts is relatively, weak and that the major adhesive interaction involved in N-CAM-mediated cell-cell adhesion is heterophilic.     

Olfactory neurons express a unique glycosylated form of the neural cell adhesion molecule (N-CAM)

mAb-based approaches were used to identify cell surface components involved in the development and function of the frog olfactory system. We describe here a 205-kD cell surface glycoprotein on olfactory receptor neurons that was detected with three mAbs: 9-OE, 5-OE, and 13-OE. mAb 9-OE immunoreactivity, unlike mAbs 5-OE and 13-OE, was restricted to only the axons and terminations of the primary sensory olfactory neurons in the frog nervous system. The 9-OE polypeptide(s) were immunoprecipitated and tested for cross-reactivity with known neural cell surface components including HNK-1, the cell adhesion molecule L1, and the neural cell adhesion molecule (N-CAM). These experiments revealed that 9-OE-reactive molecules were not L1 related but were a subset of the 200-kD isoforms of N-CAM. mAb 9-OE recognized epitopes associated with N-linked carbohydrate residues that were distinct from the polysialic acid chains present on the embryonic form of N-CAM. Moreover, 9-OE N-CAM was a heterogeneous population consisting of subsets both with and without the HNK-1 epitope. Thus, combined immunohistochemical and immunoprecipitation experiments have revealed a new glycosylated form of N-CAM unique to the olfactory system. The restricted spatial expression pattern of this N-CAM glycoform suggests a possible role in the unusual regenerative properties of this sensory system.     

Effects of antibodies to neural cell adhesion molecule (N-CAM) on the differentiation of neuromuscular contacts between ciliary ganglion neurons and myotubes in vitro

Previous experiments have suggested that the neural cell adhesion molecule (N-CAM) may have a role in initial nerve-muscle adhesion. To determine whether N-CAM might be involved in synaptic differentiation, we grew ciliary ganglion neurons and embryonic myotubes together in the presence and absence of monovalent antibodies to N-CAM. In normal cultures, undifferentiated neurites contact myotubes, and the nerve at some of these neurite-myotube contacts acquires concentrations of synaptic vesicle antigens. Most of these vesicle antigen-positive contacts become associated with patches of acetylcholine receptor (AChR) on the surface of the underlying myotube. Contacts without concentrations of vesicle antigens do not become associated with AChR patches. In the presence of antibodies to N-CAM, adhesion between neuronal somata and myotubes was reduced, but neurites contacted myotubes with near-normal frequency. The subsequent differentiation of nerve and muscle at these contacts, as assayed by the localization of vesicle antigens and AChR, proceeded normally in the presence of anti-N-CAM antibodies. The results suggest that N-CAM-mediated adhesion between neurite and myotube is not required for synaptic differentiation.     

Specificity of monoclonal antibodies against human thyroglobulin; comparison with autoimmune antibodies.

Ten monoclonal antibodies (mAb) directed against human thyroglobulin (hTgb) were produced, purified and characterized. The mAb avidity for hTgb ranged from 10(-10) to 10(-6) M. The species specificity of the mAb was as follows: eight mAb reacted with monkey Tgb, three with dog Tgb and one with pig Tgb; none with bovine and ovine Tgb. The binding of mAb to hTgb was not significantly inhibited in the presence of Tgb carbohydrate moieties, tyrosine, iodotyrosines and iodothyronines. The topology of the antigenic determinants recognized by the 10 mAb on hTgb was explored by inhibition of Tgb binding of radiolabeled mAb by the other antibodies. Six distinct clusters of reactivity were described. Localization of the antigenic determinants recognized by mAb on hTgb was attempted using tryptic fragments of hTgb to inhibit the binding of mAb to hTgb. The inhibitory effect of hydrolysis products was different for each mAb but exhibited partial analogies between mAb of the same cluster of reactivity. Anti-hTgb autoimmune antibodies (aAb) purified from sera of Graves patients cross-reacted essentially with mAb of one out of the six clusters. These results demonstrate that the large number of antigenic determinants presented by the hTgb are not disseminated on the molecule but are clustered in antigenic regions. Furthermore, from the six antigenic regions evidenced in this paper, only one is involved in autoimmune antibody production in Grave's disease.     

Topography of N-CAM structural and functional determinants. II. Placement of monoclonal antibody epitopes

The accompanying report (Watanabe, M., A. L. Frelinger III, and U. Rutishauser, 1986, J. Cell Biol., 103:1721-1727) describes a set of monoclonal antibodies (mAbs) directed against N-CAM epitopes representing the known major structural and functional domains of the molecule. In this study, we have generated and separated a variety of peptide fragments from N-CAM, and then used their size and reactivity with each antibody to position the antigenic sites along the peptide chain. This epitope map, together with the biological properties of the antibodies and previous studies on N-CAM, have been used to construct a topographical model for the molecule in the cell membrane.     

N-CAM at the vertebrate neuromuscular junction

We have detected the neural cell adhesion molecule, N-CAM, at nerve-muscle contacts in the developing and adult mouse diaphragm. Whereas we found N-CAM staining with fluorescent antibodies consistently to overlap with the pattern of alpha-bungarotoxin staining at nerve-muscle contacts both during development and in the adult, we observed N-CAM staining on the surfaces of developing myofibers and at much lower levels on adult myofibers. Consistent with its function, N-CAM was also detected on axons and axon terminals. Immunoblotting experiments with anti-N-CAM antibodies on detergent extracts of embryonic (E) diaphragm muscle revealed a polydisperse polysialylated N-CAM polypeptide, which in the adult (A) was converted to a discrete form of Mr 140,000; this change, called E-to-A conversion, was previously found to occur in different neural tissues at different rates. The Mr 140,000 component was not recognized by monoclonal antibody anti-N-CAM No. 5, which specifically recognizes antigenic determinants associated with N-linked oligosaccharide determinants on N-CAM from neural tissue. The relative concentration of the Mr 140,000 component prepared from diaphragm muscle increased during fetal development and then decreased sharply to reach adult values. Nevertheless, expression of N-CAM in muscle could be induced after denervation: one week after the sciatic nerve was severed, the relative amount of N-CAM increased dramatically as detected by immunoblots of extracts of whole muscle. Immunofluorescent staining confirmed that there was an increase in N-CAM, both in the cell and at the cell surface; at the same time, however, staining at the motor endplate was diminished. Our findings indicate that, in muscle, in addition to chemical modulation, cell-surface modulation of N-CAM occurs both in amount and distribution during embryogenesis and in response to denervation.     

Myelin-Associated Glycoprotein Shares an Antigenic Determinant with a Glycoprotein of Human Melanoma Cells

A sulfated 100K-dalton glycoprotein has been shown to be released into the culture medium of melanoma cells. Monoclonal antibodies 10C5 and 11B5, which were raised to human melanoma cells, as well as HNK-1 bind to this glycoprotein. It is shown here that mouse anti-myelin-associated glycoprotein (MAG) carbohydrate antibodies raised to human MAG and a human IgM paraprotein associated with neuropathy also bind to the same 100K molecule. However, anti-MAG antibodies recognizing peptide epitopes do not appear to react with this glycoprotein of melanoma cells, a result suggesting that its similarity to MAG is restricted to shared carbohydrate moieties. The anti-melanoma antibodies (10C5 and 11B5) resemble HNK-1 in binding to MAG and to some 19-28K-dalton glycoproteins and sulfated, glucuronic acid-containing sphingoglycolipids of the peripheral nervous system (PNS). In addition, the anti-melanoma antibodies cross-react with neural cell adhesion molecule (N-CAM), an observation emphasizing the shared antigenicity between MAG and other adhesion molecules. The results demonstrate that the anti-melanoma antibodies fall into a class of monoclonal antibodies (including HNK-1, human IgM paraproteins associated with neuropathy, anti-human MAG antibodies, and L2 antibodies) that are characterized by reactivity against related carbohydrate determinants shared by human MAG, N-CAM, and several protein and lipid glycoconjugates of the PNS.     

Human epidermal growth factor (EGF) receptor sequence recognized by EGF competitive monoclonal antibodies. Evidence for the localization of the EGF-binding site

Epitopes recognized by three epidermal growth factor (EGF) competitive monoclonal antibodies, LA22, LA58, and LA90, have been localized to a 14-amino acid region in the extracellular domain of the human EGF receptor. The binding of each of these mutually competitive antibodies to A431 epidermoid carcinoma cells was inhibited up to 87% by EGF. Furthermore, binding to A431 cells was inhibited 100% by the EGF competitive monoclonal antibody 528 IgG. The EGF receptor monoclonal antibody 455 IgG, which recognizes a blood group A-related carbohydrate modification of A431 receptors and does not inhibit EGF binding, did not inhibit the binding of these three antibodies to A431 cells. Antibodies LA22, LA58, and LA90 were unusual in that they bound to recognized denatured and endoglycosidase F-treated antigenic determinants in Western blots. This suggested that the antibodies recognized continuous peptide epitopes. The epitopes for these antibodies were first localized in cyanogen bromide- and V8 protease-generated fragments of a truncated form of the EGF receptor secreted by A431 cells. In experiments with synthetic peptides, all three antibodies were found to bind to the 14 amino acids from Ala-351 to Asp-364 of the mature human EGF receptor. These amino acids are located between the two Cys-rich regions of the extracellular domain of the receptor, and they include an Arg-Gly-Asp-Ser recognition site for adhesion molecule receptors. The homologous sequence in the chicken EGF receptor, which binds mouse EGF with a 100-fold lower affinity than the human EGF receptor, contains four amino acid differences including two in the Arg-Gly-Asp-Ser tetramer. The mutually competitive binding of EGF and antibodies LA22, LA58, and LA90 implied that the amino acids between Ala-351 and Asp-364 participated in the formation of the EGF-binding site of the human EGF receptor.     

Delineation of olfactory pathways in the frog nervous system by unique glycoconjugates and N-CAM glycoforms

The olfactory neuroepithelium, which contains the primary sensory olfactory neurons, continually undergoes neurogenesis and axonal outgrowth throughout life. We describe here several new olfactory system-specific glycoforms of the neural cell adhesion molecule N-CAM in the frog, R. catesbeiana. Using immunochemical methods for in situ localization, we show that the lectin dolichos biflorus agglutinin (DBA) and two monoclonal antibodies, 9OE and 3A6, detect three unique N-CAM forms present on primary sensory olfactory axons. In addition, DBA and monoclonal antibody 9OE recognize glycoconjugates and/or N-CAM glycoforms expressed specifically in discrete central olfactory pathways and regions in frog brain. This is a novel example of unique adhesion molecule forms present in a chain of two neurons within a vertebrate neural pathway. Together these glycoconjugates and N-CAM glycoforms may participate in cellular interactions associated with olfactory system pathway formation and renewal.     

Relationships between epitopes on IgE recognized by defined monoclonal antibodies and by the FC epsilon receptor on basophils

The present study investigated whether the sites on the FC region of the IgE molecule, recognized by different anti-IgE monoclonal antibodies (mAb), are identical to those recognized by the Fc receptor (Fc epsilon R). The anti-IgE mAb recognize different clusters of epitopes on the Fc region of IgE and could interfere to different degrees with the binding of IgE to mast cells and basophils, but still recognized cell-bound IgE. Analysis of the stoichiometry and affinity binding of 125I anti-IgE mAb Fab' to free IgE have revealed that anti-IgE mAb of one group (51.3) recognized three repetitive determinants on the IgE Fc portion, and another group (95.3) recognized only one determinant. When these stoichiometric studies were performed with cell-bound IgE, it was found that only one of the sites recognized by 51.3 mAb was involved in the Fc epsilon R binding site. On the other hand, the site recognized by 95.3 mAb was not the Fc epsilon R binding site. Such findings establish mAb 51.3 as a useful tool for isolating the IgE peptides involved in the binding site to the receptor.     

Isolation of mouse N-CAM-related cDNA: detection and cloning using monoclonal antibodies.

Clones coding for the mouse neural cell adhesion molecule (N-CAM) were isolated from a cDNA library prepared in the expression vector lambda gt 11 from mRNA extracted from a mouse neuroblastoma cell line. This library was screened with two anti-N-CAM monoclonal antibodies directed against different sites on the molecule and with rabbit anti-N-CAM serum. Two clones were identified with the first monoclonal antibody, three with the second one, none reacted with both. The relevance of these cDNA clones to N-CAM was confirmed by several observations. First, cDNA sequences detected with one monoclonal antibody cross-hybridized with those identified by the other antibody. Second, the different fusion proteins all bound the rabbit serum in addition to one monoclonal antibody. Finally, the probes hybridized to discrete mRNA species of sufficient lengths to code for the very large N-CAM polypeptides in RNA preparations from N-CAM-expressing, but not from N-CAM-negative cells. An additional mRNA species not seen in embryonic brain was expressed in adult mouse brain. Genomic blot experiments indicated that sequences corresponding to one of our probes are present only a few times in the mouse genome.     

Mapping of functional epitopes of osteopontin by monoclonal antibodies raised against defined internal sequences.

Osteopontin (OPN) is a secreted protein that has been implicated in diverse physiological and pathological processes. OPN can bind to integrins, via GRGDS or SVVYGLR amino acid sequences, and to other cell surface receptors, and many of OPN's functions are likely mediated via cell adhesion and subsequent signaling. Here we developed and characterized a series of five monoclonal antibodies, raised to distinct internal peptide sequences of human OPN, and have used these sequence-specific reagents, along with the previously described anti-OPN monoclonal antibody mAb53, to map functional epitopes of OPN that are important to cell adhesion and migration. All antibodies were reactive with native as well as recombinant human OPN. One antibody (2K1) raised against the peptide VDTYDGRGDSVVYGLRS could inhibit RGD-dependent cell binding to OPN, with an efficacy comparable to that of mAb53. Furthermore, 2K1 could inhibit alpha9 integrin-dependent cell binding to OPN. The epitope recognized by 2K1 was not destroyed by thrombin digestion, whereas mAb53 has been shown to be unable to react with OPN following thrombin cleavage. The two distinct epitopes defined by 2K1 and mAb53 antibodies are closely related to the SVVYGLR cell-binding domain and the GLRSKS containing thrombin cleavage site, respectively, and are involved in cell binding and cell migration.     

Purification of Neuronal Cell Surface Proteins and Generation of Epitope-Specific Monoclonal Antibodies Against Cell Adhesion Molecules

To establish a procedure for the purification of a broad spectrum of cell surface proteins, three separate methods based on different principles were compared with the aid of four marker proteins. Membrane preparation by sedimentation-flotation centrifugation, temperature-induced phase separation with Triton X-114, and lectin affinity chromatography were used separately as well as in combination. The two-step procedure of membrane preparation and lectin affinity chromatography provided by far the best enrichment of cell surface marker proteins. This result was further substantiated by screening greater than 6,600 hybridoma cultures that originated from mice that had been immunized with protein fractions obtained by different purification protocols. In addition, it was found that solubilized glycoproteins used as immunogens led to many more cell surface-specific monoclonal antibodies than glycoproteins immobilized on lectin-agarose beads. Three monoclonal antibodies that recognize distinct epitopes of cell adhesion molecules (CAMs) were isolated. Monoclonal antibody C4 bound to a detergent-labile epitope of G4 (neuron-glia CAM). Monoclonal antibody D1 recognized specifically nonreduced neural CAM (N-CAM) with intact disulfide bridges, and monoclonal antibody D3 recognized only the 180-kilodalton isoform of N-CAM. Because of these specificities, these monoclonal antibodies promise to be useful tools for the elucidation of the structural organization of adhesion molecules.     

Topographic localization of the heparin-binding domain of the neural cell adhesion molecule N-CAM

Previous studies have reported that the cell-binding region of the neural cell adhesion molecule (N-CAM) resides in a 65,000-D amino-terminal fragment designated Frl (Cunningham, B. A., S. Hoffman, U. Rutishauser, J. J. Hemperly, and G. M. Edelman, 1983, Proc. Natl. Acad. Sci. USA, 80:3116-3120). We have reported the presence of two functional domains in N-CAM, each identified by a specific mAb, that are required for cell-cell or cell-substratum adhesion (Cole, G. J., and L. Glaser, 1986, J. Cell Biol., 102:403-412). One of these domains is a heparin (heparan sulfate)-binding domain. In the present study we have determined the topographic localization of the heparin-binding fragment from N-CAM, which has been identified by our laboratory. The B1A3 mAb recognizes a 25,000-D heparin-binding fragment derived from chicken N-CAM, and also binds to a 65,000-D fragment, presumably Frl, produced by digestion of N-CAM with Staphylococcus aureus V8 protease. Amino-terminal sequence analysis of the isolated 25,000-D heparin-binding domain of N-CAM yielded the sequence: Leu-Gln-Val-Asp-Ile-Val-Pro-Ser-Gln-Gly. This sequence is identical to the previously reported amino-terminal sequence for murine and bovine N-CAM. Thus, the 25,000-D polypeptide fragment is the amino-terminal region of the N-CAM molecule. We have also shown that the B1A3 mAb recognizes not only chicken N-CAM but also rat and mouse N-CAM, indicating that the heparin-binding domain of N-CAM is evolutionarily conserved among different N-CAM forms. Additional peptide-mapping studies indicate that the second cell-binding site of N-CAM is located in a polypeptide region at least 65,000 D from the amino-terminal region. We conclude that the adhesion domains on N-CAM identified by these antibodies are physically distinct, and that the previously identified cell-binding domain on Frl is the heparin-binding domain.     

Polysialic acid and N-CAM localisation in embryonic rat kidney: mesenchymal and epithelial elements show different patterns of expression

The expression of polysialic acid (PSA) and the neural cell adhesion molecule (N-CAM) during the embryonic development of rat kidney was investigated using immunocytochemistry and immunoblotting. A monoclonal antibody (mAb 735), which recognised only long chain PSA, and polyclonal antibodies specific for N-CAM were employed. At the earliest stages of metanephros formation, PSA and N-CAM immunostaining was found in both embryonic anlagen, namely the uretic bud and the metanephrogenic mesenchyme. Reactivity in uretic bud derivatives declined during embryonic development and was generally absent in the collecting duct system by embryonic day 18 (E18). Uninduced metanephrogenic mesenchyme was immunostained throughout development while induced mesenchymal cells showed greatly increased PSA and N-CAM immunoreactivity during their transformation into epithelium. This staining declined rapidly as nephrons differentiated. These processes were preceded by sorting of PSA and N-CAM to the basolateral plasma membrane. Similar N-CAM and PSA patterns were observed in mesonephros development. In adult kidney parenchyma both PSA and N-CAM were undetectable. Immunoblotting of samples of embryonic kidney with mAb 735 revealed a broad band ranging from 140 to greater than 200 x 10(3) Mr. N-CAM antibodies revealed reactivity in a band of 140 x 10(3) Mr after removal of PSA by endoneuraminidase treatment. Expression of N-CAM and PSA in both embryonic anlagen indicates that neither molecular species acts primarily as an inductive signal. These molecules were localised in areas where changes in cell adhesion during organogenesis might be important and thus may be involved in the grouping of developing cells.     

Gonococcal outer-membrane protein PIB: comparative sequence analysis and localization of epitopes which are recognized by type-specific and cross-reacting monoclonal antibodies.

Comparison of the inferred amino acid sequence of outer-membrane protein PIB from gonococcal strain P9 with those from other serovars reveals that sequence variations occur in two discrete regions of the molecule centred on residues 196 (Var1) and 237 (Var2). A series of peptides spanning the amino acid sequence of the protein were synthesized on solid-phase supports and reacted with a panel of monoclonal antibodies (mAbs) which recognize either type-specific or conserved antigenic determinants on PIB. Four type-specific mAbs reacted with overlapping peptides in Var1 between residues 192-198. Analysis of the effect of amino acid substitutions revealed that the mAb specificity is generated by differences in the effect of single amino acid changes on mAb binding, so that antigenic differences between strains are revealed by different patterns of reactivity within a panel of antibodies. The variable epitopes in Var1 recognized by the type-specific mAbs lie in a hydrophilic region of the protein exposed on the gonococcal surface, and are accessible to complement-mediated bactericidal lysis. In contrast, the epitope recognized by mAb SM198 is highly conserved but is not exposed in the native protein and the antibody is non-bactericidal. However, the conserved epitope recognized by mAb SM24 is centred on residues 198-199, close to Var1 , and is exposed for bactericidal killing.     

Adherence epitopes of Mycoplasma genitalium adhesin.

The adherence-mediating sites of the 153 kDa adhesin of Mycoplasma genitalium (MgPa-protein) were characterized at the amino acid sequence level using six monoclonal anti-MgPa antibodies which showed adherence-inhibiting activity. For characterization of the regions to which antibody bound, three segments of the adhesin (N-terminal region, a D1-domain located approximately in the middle of the molecule and a D2-domain located near to the C-terminus) were synthesized as overlapping octapeptides. These regions were chosen in analogy to the three domains of Mycoplasma pneumoniae that are involved in the adhesion process. Whereas two monoclonal antibodies (mAb 5B11 and mAb 6F3) bound exclusively to an epitope in the N-region, mAb 3B7 and mAb 6A2 reacted with two distinct epitopes of the D2-domain only. Binding to short synthetic peptides of different regions was analysed for mAb 3A12 (N-region and D1-region) and mAb 2B6 (N-region and D2-region). Close proximity of the N-region and the D2-region in the native MgPa-protein of M. genitalium was indicated in a competitive ELISA test, using freshly harvested M. genitalium cells. Epitope mapping and competition experiments with monoclonal anti-MgPa antibodies revealed interesting differences in the adherence-mediating sites of MgPa and the adhesin (P1-protein) of M. pneumoniae. Whereas a three-dimensional arrangement of protein loops is suggested for both native adhesins, the MgPa-protein and the P1-protein adherence-mediating epitopes are located in non-homologous regions of these two related proteins.(ABSTRACT TRUNCATED AT 250 WORDS)     

Neoantigens in complement component C3 as detected by monoclonal antibodies. Mapping of the recognized epitopes by synthetic peptides.

The different fragments of the third complement component, C3, generated upon complement activation/inactivation have the ability to bind to several other complement components and receptors as well as to proteins of foreign origin. These multiple reactivities of C3 fragments are associated with a series of conformational changes occurring in the C3 molecule during its degradation. The conformations acquired by the different C3 fragments are also associated with the exposure of neoantigenic epitopes that are specific for (a) particular fragment(s). In order to study these epitopes and thus the conformational changes occurring in C3, monoclonal antibodies (mAbs) recognizing such epitopes were produced in Balb/c mice after immunization with denatured human C3. Two of the three antibodies (7D84.1 and 7D264.6) presented in this study recognized predominantly surface-bound iC3b, and one mAb (7D323.1) recognized both surface-bound and fluid-phase iC3b. Although none of the mAbs recognized any other fluid-phase C3 fragment, all three antibodies detected micro-titre-plate-fixed C3b and iC3b, but not C3c or C3d. In addition to the reaction with human C3, mAb 7D323.1 also bound to micro-titre-plate-fixed rabbit C3. The epitopes recognized by the three mAbs were further localized by using synthetic peptides that were designed on the basis of the differential binding of the mAbs to the C3 fragments. All three antibodies reacted with C3-(924-965)-peptide, which represents the region of C3 between the kallikrein-cleavage site (923-924) and the elastase-cleavage site (965-966). On the basis of the binding of the mAbs to five different overlapping peptides spanning the region between residues 924 and 965 of the human C3 sequence, and the sequence similarity between human C3 and rabbit C3 within this area, the epitopes recognized by these antibodies are mapped. The contribution of the individual amino acid residues in the formation of the epitopes is discussed.     

Alterations in the Xenopus retinotectal projection by antibodies to Xenopus N-CAM

The patterned neural projection from the eye to the optic tectum of lower vertebrates (the retinotectal projection) has been proposed to be ordered by interactions between the optic nerve fibers and their surrounding tissues. To investigate the role of one such defined cell interaction, agarose implants containing antibodies to the neural cell adhesion molecule, N-CAM, were inserted into the tectum of the African clawed frog, Xenopus laevis. Both monoclonal and polyclonal antibodies against N-CAM reversibly and specifically distorted the pattern of the retinotectal projection, decreasing the precision of the projection as determined by electrophysiological techniques as well as decreasing the density of retinal innervation of the tectum and the branching of single axons as determined by horseradish peroxidase tracing. The anatomical effects became maximal at 4 to 6 days after implantation and returned to undetectable levels by 2 weeks, whereas the physiological effects became maximal by 8 to 10 days and a normal physiological map was reestablished within 4 weeks. The results are consistent with the hypothesis that anti-N-CAM antibodies perturb the ongoing growth and retraction of the terminal arbors of the optic nerve fibers, such that a region of the tectum becomes largely denuded of fibers. The physiological defects may then be a consequence both of the initial retraction of optic nerve terminals and of the rapid ingrowth of the perturbed and neighboring optic nerve fibers into the denuded region after the antibodies were cleared from the tectum. These results support the concept of a major role for N-CAM-mediated adhesion during map regeneration and maintenance.