Tenascin-R Is a Functional Modulator of Sodium Channel β Subunits*

Voltage-gated sodium channels isolated from mammalian brain are composed of α, β1, and β2 subunits. The α subunit forms the ion conducting pore of the channel, whereas the β1 and β2 subunits modulate channel function, as well as channel plasma membrane expression levels. β1 and β2 each contain a single, extracellular Ig-like domain with structural similarity to the neural cell adhesion molecule (CAM), myelin Po. β2 contains strong amino acid homology to the third Ig domain and to the juxtamembrane region of F3/contactin. Many CAMs of the Ig superfamily have been shown to interact with extracellular matrix molecules. We hypothesized that β2 may interact with tenascin-R (TN-R), an extracellular matrix molecule that is secreted by oligodendrocytes during myelination and that binds F3-contactin. We show here that cells expressing sodium channel β1 or β2 subunits are functionally modulated by TN-R. Transfected cells stably expressing β1 or β2 subunits initially recognized and then were repelled from TN-R substrates. The cysteine-rich amino-terminal domain of TN-R expressed as a recombinant peptide, termed EGF-L, appears to be responsible for the repellent effect on β subunit-expressing cells. The epidermal growth factor-like repeats and fibronectin-like repeats 6–8 are most effective in the initial adhesion of β subunit-expressing cells. Application of EGF-L to αIIAβ1β2 channels expressed in Xenopus oocytes potentiated expressed sodium currents without significantly altering current time course or the voltage dependence of current activation or inactivation. Thus, sodium channel β subunits appear to function as CAMs, and TN-R may be an important regulator of sodium channel localization and function in neurons.

Sodium channels from brain are heterotrimeric structures composed of a central, pore-containing ␣ subunit and two auxiliary subunits, ␤1 (or its splice variant ␤1A) and ␤2. The ␤ subunits do not form the pore but play critical roles in channel gating, voltage dependence of activation and inactivation, and expression levels (1-3). 1 ␤1 and ␤2 subunits contain Ig-like extracellular domains and are members of the V-set of the Ig superfamily that includes CAMs 2 (3,4). ␤2 exhibits strong amino acid homology to F3/contactin, a CAM that interacts with the extracellular matrix molecule TN-R as well as with the Ig superfamily adhesion molecules L1 and TAG-1 in rodents and NgCAM and NrCAM in chicken. This homology is most striking in the third Ig-like domain and to the region just proximal to the transmembrane segment of F3/contactin. Two members of the tenascin family bind directly to purified rat brain sodium channels (5). Using immobilized GST fusion proteins, it was shown that purified, heterotrimeric brain sodium channels, as well as a recombinant ␤2 extracellular domain, bind specifically to the fibronectin repeat regions of TN-R and TN-C.
Tenascin molecules play important roles in cellular interactions in the developing nervous system, such as neuronal migration, neuritogenesis, and neuronal regeneration (6 -9). TN-R is expressed predominantly by oligodendrocytes during the onset and early phases of myelin formation and remains expressed by some oligodendrocytes in the adult (10 -14). It is also expressed in some neurons and interneurons in the spinal cord, retina, cerebellum, and hippocampus (10,12,14). TN-R co-localizes with other glial-derived molecules (i.e. myelin-associated glycoprotein and a phosphacan-related molecule) at high density in central nervous system myelinated axons (15). TN-R has been shown to have disparate effects resulting from interaction with one of its neuronal receptors, F3/contactin: promotion of neurite outgrowth when presented as a uniform substrate for some neuronal cell types and inhibition of growth cone advance and axonal outgrowth from retinal ganglion cells when offered as a sharp substrate boundary. TN-R also induces axonal defasciculation in vitro (15)(16)(17)(18)(19)(20)(21).
We postulated that ␤ subunits might function as CAMs in terms of interaction with extracellular matrix molecules. To determine whether sodium channel ␤ subunits are functionally modulated by TN-R, we used stably transfected fibroblasts expressing ␤1 or ␤2 subunits plated on substrates that contained TN-R or recombinant TN-R domains synthesized as GST fusion proteins. We show that cells expressing ␤1 or ␤2 subunits initially recognize TN-R plated on a nitrocellulose substrate. This event is then followed by repulsion. Application of a recombinant peptide domain of TN-R, EGF-L, resulted in the rapid and specific potentiation of sodium currents expressed by coinjection of ␣, ␤1, and ␤2 subunits in Xenopus oocytes. We hypothesize that functional interactions between sodium channel ␤ subunits and TN-R may be important for neuronal defasiculation or growth cone guidance during central nervous system development and may represent a critical communication link between the axon and the node of Ranvier.

EXPERIMENTAL PROCEDURES
Materials-Purification of TN-R from adult mouse brains by immunoaffinity chromatography was performed as described (13). Generation and purification of the recombinant domains of TN-R as fusion proteins with GST were performed as described (21). Chinese hamster lung (CHL) 1610 cells were obtained from the American Type Culture Collection. Nitrocellulose used in the cell repulsion assays was obtained from Schleicher & Schuell (catalog no. 401188, BA85, 0.45 m).
Transfection and Characterization of Cell Lines-1610 cells were transfected with 10 g of cDNA using DOTAP as described previously (22). The following transfections were performed: 1610 ϩ pcDNA3 (mock transfection), 1610 ϩ ␤1, and 1610 ϩ ␤2. Following drug selection with G418 (Life Technologies, Inc.) or Zeocin (Invitrogen), surviving cell clones were expanded and analyzed by Northern and Western blots for ␤ subunit expression. Northern blot analysis was performed as described previously (3) using Trizol reagent-purified total RNA (Life Technologies, Inc.) and digoxigenin-labeled cRNA probes for ␤1 (22) or ␤2 (3). Chemiluminescent labeling and detection were performed using Genius reagents (Roche Molecular Biochemicals). For Western blot analysis, confluent T-225 cell culture flasks of ␤1or ␤2-expressing cells were used to prepare crude membranes, as described previously (22). 50 g of each preparation were separated on 10% acrylamide SDS-polyacrylamide gel electrophoresis gels and blotted to Hybond ECL nitrocellulose membranes (Amersham Pharmacia Biotech) and probed with antibodies to ␤1 or ␤2. Rabbit polyclonal antibodies were generated against multiple-antigenic peptides by Research Genetics (Huntsville, AL). Multiple-antigenic peptides specific to an extracellular domain of ␤1 (KRRSETTAETFTEWTFR) or a cytoplasmic domain of ␤2 (KCVR-RKKEQKLSTD) were synthesized by the Protein and Carbohydrate Structure Core facility at the University of Michigan. Horseradish peroxidase-conjugated anti-rabbit IgG was used as the secondary antibody. Immunoreactive bands were detected using Westdura chemiluminescent reagent (Pierce).
Cell Repulsion Assays-4-well or 24-well tissue culture dishes were coated with methanol-solubilized nitrocellulose as described (23). 2.5-l aliquots of TN-R (15 M) or GST fusion domains of TN-R (25 M) were applied to the nitrocellulose/poly-DL-ornithine (PO)-coated surfaces of the dishes and incubated for 2 h at 37°C in a humidified atmosphere as described previously (21). The dishes were then washed three times with Ca 2ϩ -and Mg 2ϩ -free Hanks' balanced salt solution (CMF-HBSS). The coating efficiency was determined as described previously (21). Substrate boundaries were marked in ink. The source of nitrocellulose (see under "Materials") was critical to obtaining consistent results in this assay. Parental 1610 or transfected cells were plated at a density of 10 5 cells/ml. After 20 h, the cells were fixed with 2.5% glutaraldehyde and stained with Coomassie Blue (Sigma). The number of cells adhering to the extracellular matrix-coated protein areas were counted under a microscope. All experiments were performed at least three times. Statistical analyses were performed using Student's t test.
Cell Adhesion Assays-Tissue culture 4-well or 24-well dishes were coated with methanol-solubilized nitrocellulose (23) and air-dried under a sterile hood. For adhesion assays, 2.5-l spots of different TN-R fragments or GST (each at a concentration of 25 M) were applied to the nitrocellulose-coated surfaces of the dishes and incubated for 2 h at 37°C in a humidified atmosphere. The dried spots were washed with PBS and then flooded with CMF-HBSS containing 2% heat-inactivated fatty acid-free BSA (Sigma) and incubated for 2 h to block residual nonspecific protein binding sites. The dishes were then washed with PBS, and cells from the various cell lines were plated at a density of 10 5 cells/ml in 0.5 ml of growth medium containing 10% BSA. After 20 h of growth (5% CO 2 at 37°C), cultures were fixed with CMF-HBSS containing 2.5% glutaraldehyde. For adhesion blocking assays, a mixture of EGF-L, EGF-S, and FN 6 -8 was added to the culture medium at the concentrations indicated in Table II. After fixation, cultures were stained with 0.5% toluidine blue in 2.5% sodium carbonate. Cells adhering to the various spots of TN-R fragments were photographed and counted. All experiments were performed at least five times.
Electrophysiological Analysis of the Effects of the EGF-L Domains of TN-R on Sodium Channels-Xenopus oocytes were isolated by collagenase treatment of pieces of ovary, as described previously (24). On the day after isolation, oocytes were microinjected with 50 nl of RNA encoding the ␣IIA ␤1, and ␤2 subunits. The concentration of ␣IIA subunit RNA was 20 -50 ng/l, and the concentration of ␤1 and ␤2 subunit RNA was 100 -200 ng/l. Two-electrode voltage clamp recording was performed 2-5 days after injection, using a Turbo TEC 10C amplifier (Adams & List, Westbury, NY) and pCLAMP software (Axon Instruments, Foster City, CA). Electrodes were filled with 3 M KCl and had resistances of Ͻ0.5 M⍀. Data were sampled at 20 kHz and filtered at 2.5 kHz. Capacity transients, as well as leak currents, were subtracted using the P/4 procedure (25). Recordings were performed at room temperature in a 200-l chamber filled with frog Ringer solution (115 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl 2 , 10 mM HEPES, pH 7.2). Fusion proteins were added directly to the bath and subsequently washed out by superfusion. As controls for the electrophysiological effects of the recombinant EGF-L domain of TN-R, the fusion protein was treated with proteinase K (100 g/ml) overnight at 56°C. The enzyme was then inactivated by heating at 90°C for 15 min. This procedure digested the fusion protein such that no silver-stained bands were visible after SDS-polyacrylamide gel electrophoresis on 15% acrylamide gels.

Repulsive Interactions between TN-R and Transfected Cells
Expressing Sodium Channel ␤ Subunits-To determine whether sodium channel ␤ subunits are modulated by TN-R, we examined the growth behavior of transfected 1610 cells (26). We created stable cell lines that express ␤1 alone (1610␤1, clone 4), and ␤2 alone (1610␤2, clone 1). We also created mocktransfected 1610 cells that contained pcDNA3 alone. Characterization of ␤1 or ␤2 expression in these cell lines by Northern and Western blot analyses is shown in Fig. 1.
Because 1610 cells do not express endogenous sodium channel ␣, ␤1, or ␤2 subunits (22) (Fig. 1) or F3/contactin (data not shown), we were able to examine the effects of TN-R on sodium channel ␤ subunits in isolation. We predicted that ␤2 subunits may interact with TN-R because of their homology to F3/contactin. ␤1 also contains an extracellular CAM domain and is structurally homologous to ␤2 (4). We observed that CHL cells transfected with ␤2 (1610␤2) or ␤1 (1610␤1) subunits were strongly repelled by TN-R substrate coated on top of PO on nitrocellulose when allowed to settle in culture for 20 h (Fig. 2). Parental 1610 cells adhered well to TN-R during this time. These results are similar to observations in which cerebellar neurons expressing F3/contactin were repelled from a TN-R substrate (15,18,19,21). One or both of the auxiliary ␤ subunits may elicit signaling events similar to those described for F3/contactin upon interaction with TN-R (15). Parental 1610 cells and cells transfected with ␤1 or ␤2 subunits were not repelled from NCAM or laminin (data not shown). Thus, ␤1 and ␤2 subunits appear to be modulated specifically by TN-R to produce a repellent effect of the transfected cells away from the TN-R substrate. This effect could be the result of other processes (for example, apoptosis or anti-adhesion), although previous studies describing the repulsion of cerebellar neurons from a TN-R substrate excluded cell death as a reason for detachment (18).

Repulsive Effects of Different Domains of TN-R on Sodium
Channel ␤ Subunits Transfected in CHL Cells-We used recombinant peptides of TN-R (summarized in Fig. 3) along with 1610␤1 and 1610␤2 cells to determine which domains are responsible for the observed repellent effects. TN-R has a modular structure containing a cysteine-rich amino-terminal domain followed by EGF-like repeats, fibronectin type III (FN III) domains, and a fibrinogen-like knob at the carboxyl terminus (6 -9, 12, 27-30). The fusion protein EGF-L contains the aminoterminal cysteine-rich domain plus the EGF-like repeats. EGF-S contains only the EGF-like repeats. FN 6 -8 contains the sixth through the eighth fibronectin domains. FG contains the carboxyl-terminal fibrinogen-like knob. CHL cells were plated onto nitrocellulose that had been coated with PO and GST fusion proteins containing various domains of TN-R (21). Adhesion of untransfected CHL cells was the same on the EGF-S, EGF-L, FN 6 -8, and FG domains as GST (data not shown). Both the 1610␤1 and 1610␤2 cell lines were strongly repelled by EGF-L (Figs. 4 and 5) but adhered well to EGF-S, FN 6 -8, FG, and GST (Figs. 4 and 5), suggesting that the cysteine-rich amino-terminal domain of TN-R may be involved in modulating both the ␤1 and ␤2 subunits.
Adhesive Effects of Different Domains of Tenascin-R on Subunits of Sodium Channels Transfected in CHL Cells-A finely tuned balance between opposing principles allows cells to engage in transient contacts leading to long term avoidance reactions (18). Previous studies showed that F3/contactin mediates initial adhesion of neurons and transfected Chinese hamster ovary cells to TN-R. This interaction is followed by repulsion of neuronal cell bodies and neurites. Using cell-substrate binding assays performed under different conditions, it was shown that it is possible to investigate adhesion versus repulsion separately (18). We hypothesized that sodium chan-nel ␤ subunits expressed in CHL cells may initially recognize TN-R as adhesive, prior to repulsion, similar to the situation with F3/contactin. Using a cell-substrate assay described previously to detect cell adhesion, we have found that cells expressing ␤1 subunits adhere to TN-R recombinant domains. As summarized in Table I, ␤1-expressing cells adhere to FN 6 -8 and to EGF-S. Table II shows that a mixture of EGF-L, EGF-S, and FN 6 -8 fusion proteins added to the cell culture medium blocks the adhesion of ␤1-expressing CHL cells to the EGF-like or fibronectin-like domains of TN-R in a concentration-dependent manner. As in the situation for F3-contactin interactions with TN-R, our results suggest that sodium channel ␤1 subunits initially recognize TN-R substrates with an adhesive response. Following that recognition, cells expressing ␤1 subunits are repelled by TN-R, as we show using cell-substrate assays to designed to detect repulsion. Similar experiments were attempted for ␤2-expressing cells. Unfortunately, our results were uninterpretable due to nonspecific interactions with GST alone. Other methods must be explored to answer this question in the future.
The EGF-L Domain of TN-R Potentiates Sodium Currents-Our results from the adhesion and repulsion assays suggested an interaction between TN-R and sodium channel ␤ subunits. This interaction, direct or indirect, might be expected to alter the functional properties of the ion channel as well. To test this hypothesis, we examined the effects of EGF-L and FN 6 -8 fusion proteins on whole cell sodium currents recorded by voltage-clamp. Initially, we assessed sodium currents in SNaIIA cells (26) transfected with ␤1 and ␤2 subunits, using the whole cell configuration of the patch clamp technique. In these experiments, neither EGF-L nor FN 6 -8 fusion proteins significantly altered the amplitude, time course, or voltage dependence of sodium currents (data not shown). Preliminary results FIG. 1. Characterization of transfected cell lines. A, Northern blot analysis. Total RNA was prepared from one 60-mm cell culture dish of each cell line using Trizol reagent. 10 g of each RNA sample were separated on a formaldehyde-agarose gel, blotted to nylon, probed with digoxigenin-labeled cRNA probes specific to ␤1 or ␤2, and detected with CDP-Star (Roche Molecular Biochemicals) as described previously (3). Membranes were prepared from 1610␤1-4 cells, 1610␤2-1 cells, and rat brain. 50 g of each preparation were separated on 10% polyacrylamide SDS gels, transferred to nitrocellulose, and probed with anti-␤1 (left panel, 1:500 dilution) or anti-␤2 (right panel, 1:500 dilution) antibodies followed by horseradish peroxidase-conjugated anti-rabbit IgG. Immunoreactive bands were visualized with Westdura chemiluminescent reagent and Hyperfilm ECL. Molecular mass markers are expressed in kDa. from the cell migration assays suggested that protein phosphorylation may modulate EGF-L-mediated effects on cell migration (data not shown). However, in the whole cell recording configuration used to record sodium currents in SNaIIA ϩ ␤1 ϩ ␤2 cells, the cell cytoplasm was dialyzed into the patch pipette, which may have resulted in loss of intracellular constituents necessary for EGF-L-or FN-dependent modulation of sodium channel function. Therefore, as an alternative approach, we expressed sodium channels in Xenopus oocytes and examined whole cell sodium currents by two-electrode voltage clamp. In two-electrode recordings, EGF-L fusion protein produced a rapid increase in the amplitude of sodium currents (Fig. 6). EGF-L-mediated potentiation was observed in oocytes coexpressing the type IIA ␣ subunit along with ␤1 and ␤2 (Fig. 6A), as well as in oocytes coexpressing ␣IIA alone (Fig. 6B). In contrast, neither FN 6 -8 fusion protein nor GST affected sodium currents in oocytes (Fig. 6E), suggesting that potentiation is a specific effect of the EGF-L domain of TN-R. EGF-L-mediated potentiation was not accompanied by any detectable changes in the voltage dependence of current activation (Fig.  6C) or inactivation (Fig. 6D) or in any obvious effects on current time course (Fig. 6, A and B). The concentration of 50 ng/l was chosen in the experiments shown in Fig. 6 because it elicited a maximal potentiation response. However, clear potentiation was also observed with concentrations as low as 5 ng/l. The potentiating activity was not inactivated by heating EGF samples to 100°C for 10 min, suggesting that a highly stable protein domain is responsible for potentiation. In contrast, potentiation was greatly reduced by pretreatment of the fusion protein with proteinase K (Fig. 6F). DISCUSSION The extracellular matrix protein TN-R modulates cells expressing sodium channel ␤1 or ␤2 subunits. Using a combination of adhesion and repulsion assays, we describe events beginning with an initial recognition of the FN 6 -8 and EGF-like repeats of TN-R followed by a repellent effect on transfected cells away from a TN-R substrate, mediated through the amino-terminal and EGF-like domains. The initial recognition event between TN-R and cells expressing sodium channel ␤ subunits can be blocked by the addition of TN-R recombinant fragments to the cell culture medium, suggesting that TN-R and sodium channel auxiliary subunits interact directly. Our data confirm a recent report showing that purified, heterotrimeric rat brain sodium channels or a recombinant ␤2 subunit extracellular domain synthesized as a GST fusion protein bind with high affinity to recombinant fibronectin-like fusion protein domains of TN-R or TN-C immobilized on microtiter plates and display a low affinity for the EGF-like repeats (5).
It is interesting to consider the developmental time course of ␤1 and ␤2 subunit expression in brain (3,31) compared with the development of oligodendrocytes. Over 90% of sodium channels in the rat brain during early postnatal development are thought to contain a disulfide-linked ␤2 subunit as well as a non-covalently linked ␤1 subunit (32). Most oligodendrocytes develop after postnatal day 7 in rat brain (33). Thus, ␤1 and ␤2 subunits would be expected to be present in the neuronal plasma membrane during the early stages of myelination, when TN-R is secreted. It may be possible that ␤1 and ␤2 subunits function as CAMs apart from ␣ and play roles in neuronal development other than channel modification. It is also possible that ␤1, which is noncovalently bound to ␣, can dissociate and reassociate with the channel complex, such that the channel is dimeric or trimeric, depending on the particular needs of the cell. Thus, the presence or absence of ␤1 in the channel complex may determine the level of responsiveness of the sodium channel to TN-R.
The interaction of TN-R with the CAMs ␤1 and ␤2 in the central nervous system may lead to neuronal defasiculation (20). Central nervous system axons have been shown to form fascicles via homophilic or heterophilic binding of CAMs between axons. Preliminary results from our laboratory show that ␤1 and ␤2 exhibit homophilic binding when expressed individually in Drosophila S2 cells, leading to cell aggregation. 3 Through binding to ␤ subunits or other CAMs on adjacent axons, ␤ subunits may contribute to axonal fasiculation. Interaction of TN-R with ␤1 or ␤2 may then serve to disrupt this interaction, initiating the process of defasciculation.
TN-R has been shown previously to inhibit growth cone advance as well as neurite outgrowth when presented as a sharp substrate boundary (16 -19). The repulsion of ␤ subunitexpressing fibroblasts from TN-R substrates in our hands may occur through a similar mechanism. It is possible that growth cone repulsion, and thus axon guidance, is facilitated through the interaction of TN-R with sodium channel ␤ subunits present at the growth cones, resulting in cell repulsion.
EGF-L-mediated potentiation of sodium currents may represent a novel mechanism for modulation of sodium channel activity. Potentiation could involve an increase in the probability of sodium channel opening, an increase in single channel conductance, or up-regulation of silent channels. Surprisingly, potentiation was observed not only in oocytes coexpressing ␣IIA, ␤1, and ␤2 but also in oocytes expressing ␣IIA alone. This observation suggests the possibility that EGF-L can interact directly with sodium channel ␣ subunits. Alternatively, potentiation of sodium currents in oocytes may be an indirect effect of EGF-L, reflecting a second messenger-mediated signal initiated by EGF-L interaction with some other oocyte membrane protein.
Recent observations on the activity of voltage-dependent sodium channels in TN-R knockout mice are in agreement with the hypothesis that TN-R modulates the activity of sodium channels (34). In these mice, there is no apparent change in expression or distribution of sodium channels, but compound action potential recordings from the optic nerves of the mutant mice show a significant decrease in conduction velocity as compared with wild type controls. Thus, in the absence of TN-R, the observed decrease in the optic nerve conduction velocity may reflect altered channel function. A more thorough analysis of the properties of sodium channels in the TN-R knock-out mu-  Lagenaur and Lemmon (23) and air-dried under a sterile hood. For adhesion assays, 2.5-l spots of different TN-R fragments or GST, each at a concentration of 25 M, were applied to the nitrocellulose-coated surfaces of the dishes and incubated for 2 h at 37°C in a humidified atmosphere. The spots were then washed three times with PBS. The dishes were flooded with CMF-HBSS containing 2% heat-inactivated fatty acid-free BSA and incubated 2 h to block residual nonspecific protein binding sites. Subsequently, the dishes were washed with PBS and the cells of different lines were plated in 0.5 ml medium with 10% BSA at a density of 100,000 cells/ml. After 20 h, the cultures were fixed by flooding with CMF-HBSS containing 2.5% glutaraldehyde. After fixation, cultures were stained with 0.5% toluidine blue in 2.5% sodium carbonate. The number of cells adhering to different protein spots were photographed and counted. Data are presented as the mean and S.D. of five independent experiments. EGF Lagenaur and Lemmon (23) and air-dried under a sterile hood. 2.5-l spots of EGF-L, EGF-S, or FN6 -8 (each at a concentration of 25 M) were applied to the nitrocelluose-coated surfaces of the dishes and incubated for 2 h at 37°C in a humidified atmosphere. The spots were washed with PBS and then flooded with CMF-HBSS containing 2% heat-inactivated fatty acid free BSA and incubated 2 h to block residual non-specific protein binding sites. The dishes were then washed with PBS and cells from the various cell lines were plated at a density of 100,000 cells/ml in 0.5 ml of growth medium containing 10% BSA. A mixture of EGF-L, EGF-S, and FN 6 -8 was added to the culture medium at the concentrations indicated and preincubated with the cells for 1h. After 20 h of growth (5% CO 2 , 37°C) cultures were fixed with CMF-HBSS containing 2.5% glutaraldehyde. Cultures were then stained with 0.5% toluidine blue in 2.5% sodium carbonate. The numbers of cells adhering to the various spots of TN-R fragments were photographed and counted. Data presented are the results of at least five independent experiments. tant and morphometric analysis of nodal area in the optic nerve or other myelinated fiber tracks will be required to investigate these possibilities. The physiological significance of EGF-L-mediated potentiation is, at present, unclear. One intriguing possibility is that it is a means by which neurons or neuronal processes translate contact with TN-R into an intracellular signal. For example, influx of calcium through voltage-gated calcium channels has been shown to be a signal for myelin-evoked growth cone collapse in cultured rat locus coeruleus neurons (35)(36)(37)(38)(39); however, the mechanism by which myelin activates calcium channels is not understood. Our results suggest that calcium channel activation could result, at least in part, from TN-R-mediated potentiation of sodium currents, leading to increased excitability of the growth cone membrane. To our knowledge, no studies have examined the role of sodium channels in growth cone behavior. This will be an interesting area for future investigation.
The potential clinical importance of the role of sodium channel ␤ subunits as CAMs is illustrated by a recent report describing a mutation underlying generalized epilepsy with febrile seizures plus (40). Mutation of a critical cysteine residue (C121W) predicted to be involved in formation of the extracellular Ig fold of ␤1 (41) is postulated to be responsible for the familial epileptic phenotype. These results, combined with the results of our study, suggest that a deficit in the ability of ␤1 subunits to interact with extracellular matrix molecules may cause changes in sodium channel conductance or gating properties to produce excitotoxicity. Alternatively, expression of ␤1 subunits that are incapable of functioning as CAMs may alter sodium channel density and localization in the neuronal plasma membrane or cause changes in axonal guidance events, such as fasiculation or growth cone collapse during brain development.
An interesting parallel may be drawn from the present study to the adhesion molecule on glia (AMOG/␤2), the ␤2 subunit of the murine Na ϩ , K ϩ -ATPase (42). This cell surface glycoprotein is expressed by glial cells during neuronal development, as well as in the adult. It can form a complex with the ␣1 subunit to yield a functionally active Na ϩ ,K ϩ -ATPase enzyme when injected into Xenopus oocytes, indicating that it can act as an integral member of the ion transport complex. In addition, AMOG/␤2 acts as a CAM involved in neuron-glia interactions, promoting neurite outgrowth from cerebellar and hippocampal neurons. It is thought to be unlikely that Na ϩ ,K ϩ -ATPase activity is involved in neurite outgrowth. Instead, AMOG/␤2 may function independently of ␣ subunits to exert this effect. Evidence that AMOG/␤2 may appear at the cell surface independent of an ␣ subunit has been found in the distal colon (43). Like AMOG/␤2, sodium channel ␤ subunits are cell surface glycoproteins expressed during critical stages of neuronal development. ␤1 and ␤2 now appear to have dual functions: as modulators of ion channel activity and as CAMs. We do not yet know whether sodium channel ␤ subunits can appear and function at the cell surface independently of ␣ during brain development. The present study now provides a framework from which the physiological significance of sodium channelextracellular matrix interactions can be investigated.