Abstract
Although knowledge of the functions of neurotrophins has advanced rapidly in recent years, studies concerning the involvement of neurotrophins in glial–neuronal interactions rarely extend further than their roles in supporting the survival and differentiation of neuronal cells. In this study endogenous brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT3) were identified in Schwann cell/dorsal root ganglia neuronal cocultures and shown to modulate the myelination program of the peripheral nervous system. The differential expression of BDNF and NT3 were examined and compared with the expression profiles of myelin proteins in the cocultures throughout the myelination process. BDNF levels correlated with active myelin formation, whereas NT3 expression was initially high and then down regulated throughout the proliferation and premyelination periods. Addition of exogenous BDNF enhanced myelination, whereas the removal of the endogenous BDNF by using the BDNF receptor TrkB-Fc fusion protein inhibited the formation of mature myelin internodes. Interestingly, exogenous NT3 significantly inhibited myelination, whereas the removal of the endogenous NT3 by using the NT3 receptor TrkC-Fc fusion protein resulted in an enhancement similar to that obtained with the addition of BDNF. In addition, in vivo studies were performed during the development of the mouse sciatic nerve. Subcutaneous injections of BDNF resulted in an enhancement of myelin formation in the sciatic nerve, whereas the removal of the endogenous BDNF dramatically inhibited myelination. Injections of NT3 inhibited myelin formation, and the removal of the endogenous NT3 enhanced myelination. These results demonstrate that BDNF and NT3 possess different modulatory roles in the myelination program of the peripheral nervous system and that their mechanisms of action are specific and highly regulated.
The myelin sheath is a unique component of the nervous system that functions to maximize the efficiency and velocity of action potentials transmitted through nerve cell axons. The composition of myelin differs in the peripheral and central nervous systems principally in the nature of the proteins that are embedded in the lipid bilayers of the multiple myelin wraps. The proteins in peripheral myelin have received considerable attention because alterations in genes encoding these proteins are responsible for the demyelinating peripheral neuropathies, especially the Charcot-Marie-Tooth diseases (1). The formation of peripheral myelin is a complex, dynamic process involving two different cell types, the myelin forming glia (Schwann cells) and the neurons, that involve a series of neuronal–glial interactions controlling the various stages of myelination (2–4). Much has been learned about the interactions that define the three major phases of Schwann cell growth and differentiation in myelination. These include the proliferation and migration of Schwann cells on axons in the proliferative stage, the elongation and ensheathment of the axon by the Schwann cell in the premyelination stage, and the initiation, rate, and extent of growth of the myelin sheath in the final myelination stage (5–7). Accompanying Schwann cell differentiation are reciprocal interactions that are responsible for proper axon development, including the survival and differentiation of the neurons (2, 3).
The potential role for neurotrophins in peripheral myelination is best illustrated when the Schwann cell–axonal interactions are disrupted by nerve injury and then allowed to reestablish themselves. For example, on transection or focal crush injury a rapid induction of the expression of nerve growth factor (NGF) (8, 9) and the p75 neurotrophin receptor (NTR) (10) occurs in the segment distal from the site of injury (11, 12). In contrast, brain-derived neurotrophic factor (BDNF) (13) and neurotrophin-4 (NT4) (14, 15) are only induced in the distal segment 2 weeks after injury, whereas neurotrophin-3 (NT3) (16–19) levels decrease initially after transection but recover soon after (20, 21). For a review on neurotrophins and their corresponding receptors, see refs. 22–25. It is reasonable to assume that the well known effects of these neurotrophins and their receptors in promoting neuronal survival and differentiation come into play after peripheral nerve injury. However, the functions of neurotrophins are not restricted exclusively to neurons. There are numerous descriptions of their role in the reproductive, endocrine, cardiovascular, and immune systems, as well as with glial cells (23, 24, 26–28). Previous studies have documented the effects of NT3 on oligodendrocyte proliferation, survival, and differentiation (29–31). It has also been reported that neurotrophins influence Schwann cell migration through the p75NTR (32, 33) and that BDNF affects myelin formation during nerve regeneration (34, 35). However, it has been difficult to determine the influence of neurotrophins, including BDNF, on the various stages preceding myelin synthesis in complex systems. For this reason we have initially used myelination in Schwann cell/neuronal cocultures where the separation of these stages is possible all the way through to an examination of mature myelin internodes. Our results demonstrate that endogenous BDNF and NT3 have different expression profiles during the development of myelin in these cultures and that they exert different modulatory actions on the myelination program in the peripheral nervous system (PNS). We have confirmed these effects in vivo during PNS myelin development, providing one more example of the complex and dynamic regulation of Schwann cell–neuronal interactions, this time by neurotrophins.
Materials and Methods
Materials.
NGF was obtained from Serotec. BDNF and NT3 were gifts from Regeneron Pharmaceuticals (Tarrytown, NY). TrkB-Fc and TrkC-Fc were also gifts from Regeneron Pharmaceuticals and are chimeric proteins of the Fc fraction of the human Ig fused to the extracellular domain of TrkB or TrkC receptors, respectively (36).
Dorsal Root Ganglia (DRG) Neuronal/Schwann Cell Cocultures.
Purified neuronal and Schwann cell cultures were prepared by using methods described in ref. 37. Neuronal cultures were established from DRG neurons obtained from Sprague–Dawley rat embryos at 15 days gestation (Simonsen Laboratories, Gilroy, CA). DRG neurons were dissociated and plated onto collagen coated coverslips. Nonneuronal cells were eliminated by cycling (three 2-day cycles) with a fluorodeoxyuridine-containing medium (10 μM). NGF-dependent neurons were then maintained for 1 week in a medium consisting of 10% FBS in MEM and 100 ng/ml of NGF.
Schwann cells were isolated from the sciatic nerve of 4-day-old rat pups as described (37). Schwann cells were purified by using cytosine arabinoside and Thy-1.1 antibody mediated lysis of the fibroblasts (the anti-Thy-1.1 antibody was obtained from the American Type Culture Collection). Approximately 100,000 purified Schwann cells were than seeded onto purified neuronal cultures of ≈50,000 cells and allowed to proliferate and ensheath the axons (≈1 week). Myelination was then initiated with the addition of ascorbic acid (50 μg/ml), which was replenished with feeding every 2 to 3 days.
Western Blot Analysis.
Samples from Schwann cell/neuronal cocultures and sciatic nerves were prepared for Western blot analysis by homogenization in radioimmunoprecipitation assay (RIPA) buffer [PBS with 1% Nonidet P-40/0.5% deoxycholate/0.1% SDS/1 mM PMSF/Complete protease inhibitor tablets (Roche Molecular Biochemicals)] followed by high-speed centrifugation. Protein determination was made by using the Bicinchoninic Acid Kit (Sigma). Equivalent amounts of total protein extract from each sample were mixed with sample buffer, boiled, and loaded onto SDS polyacrylamide gels (38). Electrophoretic separation of the extracts was typically performed on 10–15% (depending on the molecular weight of the protein of interest) discontinuous acrylamide gels under denaturing conditions. The proteins were then transferred to pure nitrocellulose membranes (PROTRAN BA85, Schleicher and Schuell, 0.45 μm) and probed with specific antibodies. The mouse monoclonal anti-myelin-associated glycoprotein (MAG; Chemicon) was used at a concentration of 2.5 μg/ml (under nonreducing conditions); the mouse monoclonal anti-P0 antibody (gift from J. J. Archelos, Karl-Franzen-Universität, Graz, Austria) was used at a dilution of 1:5,000. All primary antibodies were incubated overnight at 4°C. Secondary HRP-conjugated anti-mouse IgG and anti-rabbit IgG antibodies were used at a dilution of 1:10,000 (Jackson ImmunoResearch). The blots were developed by chemiluminescence (Renaissance, DuPont/NEN) as described by the manufacturer. All of the blots were imaged and quantitated in the linear range for the corresponding antibodies. Protein concentrations of the samples were serial diluted until linear intensities were achieved. In many instances, the blots were stripped and reprobed with different antibodies.
Immunocytochemistry.
Immunocytochemistry was performed as described (37, 39). Cocultures were fixed in 4% paraformaldehyde before dehydration through a graded ethanol series (50, 70, 90, and 100%). Samples were then permeabilized and blocked by incubation with 20% normal goat serum or 10% FCS. Primary antibodies included the mouse monoclonal anti-P0 antibody used at a dilution of 1:500 and the mouse monoclonal anti-MAG at a concentration of 2.5 μg/ml. The Texas red-conjugated anti-mouse IgG (Jackson ImmunoResearch) was used as a secondary antibody at a dilution of 1:1,000. Cellular nuclei were examined by using the Hoechst dye. Samples were mounted and fluorescence microscopy was accomplished by using a Nikon Microphot FXA.
ELISA Analysis.
ELISAs were performed by using the TMB (tetramethylbenzidine) Peroxidase Substrate System (Kirkegaard & Perry Laboratories), as described (40). Briefly, the BDNF and NT3 ELISAs were accomplished by using Immobilon plates (Nunc) coated with the TrkB-Fc fusion protein and the #4704 anti-NT3 antibody (Regeneron Pharmaceuticals), respectively, followed by incubation of the samples. The BDNF ELISA was developed by using a polyclonal chicken anti-BDNF antiserum and an anti-chicken-HRP antibody (Promega) for detection. The NT3 ELISA was developed by using a biotinylated monoclonal anti-NT3 antibody (Regeneron Pharmaceuticals) and streptavidin-HRP (Sigma) for detection. The substrate was incubated in the ELISA reaction for ≈2–5 min or until adequate signal was detected. ELISA reactions were stopped with the addition of 1 M phosphoric acid. Using a Bio-Rad Model 550 Microplate Reader, optical densities were measured at a wavelength of 450 nm.
Injections in Mouse Sciatic Nerve.
BDNF, NT3, TrkB-Fc, and TrkC-Fc (3 μg each; Regeneron Pharmaceuticals) were injected s.c., starting from the caudal portion of the greater trochanter region and running parallel along the sciatic nerve (total volume of 5 μl). The contralateral leg served as a control for each factor with the injection of saline. Injections were performed on 1-day-old mouse pups (C57BL/6, Simonsen Laboratories) and the sciatic nerves were extracted and processed 48 h later. A second set of mice were reinjected with the factors and then examined after an additional 48 h (4 days of total treatment). Nerves used for electron microscopy were trimmed and incisions were made at the flexure of the greater trochanter. In total, 11 animals were analyzed after injection with BDNF, 12 with TrkB-Fc, 11 with NT3, and 7 with TrkC-Fc.
Electron Microscopy.
Electron microscopy was performed by the Electron Microscopy Facility in the Department of Microbiology and Immunology (Stanford University, CA). Processing of the sciatic nerve was accomplished by fixation in 2% glutaraldehyde and 4% paraformaldehyde solution in PBS, followed by postfixation in 1% OsO4. Staining was achieved with 1% aqueous uranyl acetate, followed by dehydration with an ethanol gradient and treatment with propylene oxide. Finally, samples were infiltrated and embedded in pure epoxy.
Results
Expression Profiles of Endogenous Neurotrophins During Myelination in Schwann Cell/Neuronal Cocultures.
The synthesis of neurotrophins and the expression of neurotrophin receptors have previously been documented in Schwann cells and DRG neurons (11, 21, 41–43). To investigate the role of neurotrophins in the myelination process, BDNF and NT3 levels were examined in Schwann cell/neuronal cocultures established as described in Materials and Methods. The cells were grown to maturity separately and contaminating cells were removed. Approximately 1 week after removal of the antimitotic agent, neuronal cultures were seeded with Schwann cells. On contact with the axons, Schwann cells proliferated rapidly (proliferation stage). Approximately 4 days after seeding, the axons were fully populated and proliferation had ceased. The Schwann cells then began to elongate and ensheath the axons (premyelination stage). At this time (7 days after seeding the Schwann cells) the cocultures were induced to myelinate by the addition of ascorbic acid (myelination stage). Active myelin formation occurred between ≈4 and 7 days after induction, as characterized by lipid analyses, internode diameter measurements, and electron microscopy (37, 44). By examining the expression profiles of MAG and P0 in the Schwann cell/neuronal cocultures (Fig. 1A), the initial induction of the myelin protein synthesis was observed ≈2 days after the addition of ascorbic acid. The expression of MAG and P0 leveled off by 6 days after induction, suggesting that active myelin synthesis was complete at this time. This extensive and reproducible characterization of the cocultures allowed for further detailed investigations and/or distinctions to be made between the proliferation, premyelination, and myelination phases of the Schwann cells as noted above.
Figure 1.
Expression profiles for MAG, P0, BDNF, and NT3 during the myelination process in Schwann cell/neuronal cocultures. (A) Western blot analysis of MAG and P0 in Schwann cell/neuronal cocultures throughout the myelination process. (B) ELISA for BDNF and NT3 in cultures of Schwann cells and DRG neurons alone, and in cocultures. Schwann cells were induced to myelinate at ≈7 days after seeding onto DRG neurons (day 0) with the addition of ascorbic acid. The results are shown as the mean value ± SD.
Because the DRG neurons were consistently cultured in the presence of NGF, only the expression of BDNF and NT3 were examined. Conditioned media from neurons, Schwann cells, or cocultures were collected every two days and assayed for BDNF and NT3 by ELISA (Fig. 1B). Whereas DRG neurons secreted both BDNF and NT3 at high concentrations (1–2 ng/ml), Schwann cells only secreted NT3, at concentrations 10-fold lower than from the DRG neurons (0.2 ng/ml). On seeding the Schwann cells onto the DRG neurons, NT3 levels immediately decreased. In addition, during the proliferation and premyelination periods, NT3 levels gradually diminished until near undetectable amounts were observed at the day of induction. Interestingly, BDNF levels remained relatively constant throughout the entire proliferation and premyelination periods, and only began to decrease at 2 days after induction of the onset of active myelin synthesis (Fig. 1B). Both BDNF and NT3 were undetectable at about 6 days after induction with ascorbic acid. These results demonstrate different expression profiles for BDNF and NT3 throughout the myelination process in Schwann cell/neuronal cocultures, which could imply a potential difference in function.
Endogenous BDNF and NT3 Exert Different Modulatory Actions on Myelination in Schwann Cell/Neuronal Cocultures.
To investigate the potential role of BDNF and NT3 on the myelination process, exogenous BDNF (100 ng/ml) or NT3 (100 ng/ml) were added at the day of induction in Schwann cell/neuronal cocultures. BDNF significantly enhanced the expression of MAG by ≈2-fold and P0 by ≈1.5-fold over control cultures, whereas NT3 diminished the expression of these myelin proteins by ≈2- and 3-fold, respectively (Fig. 2). To further examine the role of endogenous BDNF and NT3, the TrkB-Fc and TrkC-Fc fusion proteins were used to diminish the endogenous neurotrophin levels. The addition of TrkB-Fc (1 μg/ml) at the day of induction inhibited the expression of MAG by 3-fold and P0 by 5-fold, whereas the addition of TrkC-Fc (1 μg/ml) resulted in an enhancement greater than that obtained with the addition of BDNF (≈3-fold; Fig. 2). These results indicate that endogenous BDNF plays an important role in signaling the initiation of myelin formation, whereas endogenous NT3 acts as an inhibitor of the myelination process.
Figure 2.
Western blot analysis of the effects of endogenous and exogenous neurotrophins on the expression of myelin proteins in Schwann cell/neuronal cocultures. Exogenous BDNF (100 ng/ml), TrkB-Fc (1 μg/ml), NT3 (100 ng/ml), and TrkC-Fc (1 μg/ml) were added to cocultures on the day of induction of myelin formation. After 6 days of induction, cocultures were extracted and probed for MAG and P0 proteins. The results are shown as the mean value ± SD.
To examine the effects of BDNF and NT3 on the formation of mature myelin internodes, immunocytochemical analyses for P0 were performed in the presence of exogenous neurotrophins or the BDNF and NT3 scavengers TrkB-Fc and TrkC-Fc. Six days after the addition of ascorbic acid to the Schwann cell/neuronal cocultures, the formation of mature myelin internodes was detected through P0 immunostaining (Fig. 3B). The addition of BDNF produced an enhancement in myelin formation, especially at earlier time points (data not shown), although this effect was not easily distinguishable through fluorescence microscopy at 6 days after induction because of the high degree of myelination in control conditions (Fig. 3C). In contrast, the removal of endogenous BDNF by addition of TrkB-Fc greatly inhibited the formation of mature myelin internodes (Fig. 3D). Conversely, NT3 had the opposite effect on myelin formation. Exogenous NT3 inhibited the appearance of myelin internodes, whereas TrkC-Fc, by removing NT3, enhanced myelin formation (Fig. 3 E and F). As was noted with the addition of BDNF, the enhancement by TrkC-Fc was significantly greater at earlier time points (data not shown), but was not easily distinguishable at 6 days after induction. In addition, the few internodes that were detected after addition of TrkB-Fc or NT3 were significantly thinner and shorter in length than internodes obtained from the control cultures. Neither BDNF nor NT3 caused any changes in Schwann cell proliferation or in the morphology of the ensheathed Schwann cells (data not shown). There were no observable changes in the axonal processes of the DRG neurons and there was no sign of increased cell death, as determined by neurofilament and nuclear staining (data not shown). These results suggest that under our in vitro culture conditions, endogenous BDNF and NT3 modulate the myelination process in opposite ways. BDNF is required for proper myelin internode formation, whereas an excess of NT3 inhibits the myelination process and the formation of normal myelin internodes.
Figure 3.
The effects of endogenous and exogenous neurotrophins in the formation of myelin internodes in Schwann cell/neuronal cocultures determined by immunocytochemical analysis of P0. (A) Control cultures without the addition of primary antibody. (B) Control cultures with the addition of the anti-P0 antibody. (C) Addition of exogenous BDNF (100 ng/ml). (D) Addition of TrkB-Fc (1 μg/ml). (E) Addition of exogenous NT3 (100 ng/ml). (F) Addition of TrkC-Fc (1 μg/ml). All factors were added to cocultures on the day of induction of myelin formation. Immunocytochemistry was performed on cocultures after 6 days of induction.
Endogenous BDNF and NT3 Exert Different Modulatory Actions on Myelination in the Developing Sciatic Nerve.
The effects of BDNF and NT3 on myelin formation in vivo were analyzed during the development of the sciatic nerve in newborn mice. Subcutaneous injections of the neurotrophins or the neurotrophin scavengers were made starting at the caudal portion of the greater trochanter region and running parallel along the sciatic nerve. Injections of BDNF, NT3, TrkB-Fc, and TrkC-Fc (3 μg each) were performed on 1-day-old mouse pups, followed by the extraction and processing of the sciatic nerves for Western blot analyses 48 h later. A second set of mice were reinjected with the same factors and then examined after an additional 48 h (4 days of total treatment). Contralateral legs were injected with an equivalent volume of the saline buffer used as vehicle and provided the specific controls for the injections of all of the factors reported. BDNF was found to significantly enhance the expression of MAG and P0 in the sciatic nerves by more than 50% (Fig. 4). On the contrary, injections of NT3 inhibited MAG and P0 expression in the sciatic nerves of mice treated for 2 days by ≈25%. Interestingly, the inhibition by NT3 was time-dependent and was not detected in mice treated for a total of 4 days. In agreement with the results from the coculture experiments, the physiological effects of BDNF and NT3 in mature myelin formation were clearly demonstrated after injection with the neurotrophin scavengers. TrkB-Fc reduced both P0 and MAG levels, whereas TrkC-Fc significantly enhanced the expression of the myelin proteins in both the 2- and 4-day-treated animals. The effects obtained with both of the receptor-fusion proteins demonstrate once again that endogenous BDNF and NT3 possess different modulatory actions in vivo, during the development of the sciatic nerve.
Figure 4.
Western blot analysis of the effect of neurotrophins and neurotrophin scavengers during the development of the sciatic nerve. Newborn mice (P1) were s.c. injected with 3 μg each of BDNF, NT3, TrkB-Fc, or TrkC-Fc as indicated under Materials and Methods. The contralateral leg was injected with vehicle alone as a control for each one of the individually treated mice. Two days later (P3, 2 days treatment), the sciatic nerves were isolated and processed for Western blot analysis. In some instances, the animals were reinjected at this stage and allowed to proceed for 2 more days (P5, 4 days treatment). The sciatic nerves were then processed and analyzed in the same manner. (A) Quantification of MAG protein content after treatment with the different factors for 2 or 4 days. Representative Western blots are shown (Lower). (B) Quantification of P0 protein content and representative Western blots. The results are shown as the mean value ± SD of the percentage of the levels expressed in the contralateral leg (injected with saline vehicle alone).
To analyze the effect of BDNF on the formation of the myelin sheath in the sciatic nerves, electron microscopy was used (Fig. 5 A and B). Sciatic nerves treated with BDNF displayed a 2-fold decrease in ensheathed axons accompanied by a significant correlative increase in the number of myelinated axons (Fig. 5C). BDNF not only produced an increase in the number of myelinated axons, but also an enlargement of the myelin sheath itself. By examining the distribution of the number of lamellae in the myelinated axons of the control nerves and the BDNF-treated nerves, it became evident that the myelin sheath of the BDNF nerves were significantly thicker. This increase in size was due to an increase in the number of lamellae in the myelin internodes. Fig. 5D shows the distribution of the thickness (number of wraps) of the myelinated axons from control and BDNF-treated nerves. Whereas the control nerves had a higher percentage of axons with lower number of wraps of myelin, the BDNF-treated nerves had a profile that was shifted, because of a greater percentage of axons with a higher number of wraps. Less than 10% of the myelinated axons in the control nerve had more than 25 wraps of myelin, whereas more than 30% of all of the myelinated axons in the BDNF-treated nerves were in this same population. On average, BDNF-treated nerves had 25% more wraps of myelin. Similar analyses were also performed after injection with the chimeric protein TrkB-Fc (Fig. 6 A and B). By removing endogenous BDNF with the TrkB-Fc scavenger, a significant increase in the number of ensheathed axons (≈50%) was detected as compared with controls, and a corresponding decrease in the number of myelinated axons was also observed (Fig. 6C). In similar fashion, but with an opposite distribution as seen with the BDNF-treated nerves, TrkB-Fc had an effect not only in the proportion of axons that were myelinated, but also on the thickness of the remaining myelinated axons. In examining the distribution of the number of lamellae, the TrkB-Fc-treated nerves displayed thinner myelin sheaths or a greater population of axons with fewer wraps of myelin than the controls (Fig. 6D). These results were consistent with the Western blot analyses and with the effects observed in the cocultures, suggesting that BDNF plays a fundamental role in the initiation and progression of the myelination program of the peripheral nervous system.
Figure 5.
Effects of BDNF treatment on the myelin ultrastructure during sciatic nerve development. Newborn mice (P1) were injected with 3 μg of BDNF as indicated under Materials and Methods and reinjected again 2 days later. The contralateral leg served as a control with the injection of saline vehicle alone. At P5 the sciatic nerves from both legs were removed and processed for the electron microscopy study. Low magnification electron micrographs from (A) control nerve treated with saline alone and (B) BDNF-treated nerve. Axons ensheathed by Schwann cell cytoplasm without the formation of myelin are indicated with an asterisk. The scale bar represents 1 μm. (C) Ensheathed and myelinated axons from control and BDNF-treated nerves were counted and the proportions shown as a percentage of the sum of both. (D) The thickness of the myelin sheath was determined by counting the number of lamellae of individual myelinated axons. The distribution is shown as the mean value ± SD of the percentage of myelinated axons that falls within a certain range in the number of lamellae.
Figure 6.
Effects of TrkB-Fc treatment on the myelin ultrastructure during sciatic nerve development. Newborn mice were injected with 3 μg of TrkB-Fc and their sciatic nerves analyzed as in Fig. 5. Electron microphotographs from (A) control nerve (saline alone) and (B) TrkB-Fc-treated nerve. (C) Distribution of myelinated axons against ensheathed axons as a percentage of the sum of both. (D) The thickness of the myelin sheath was determined by counting the number of lamellae. Myelinated axons were distributed as a function of the thickness of the myelin sheath. The distribution is shown as the mean value ± SD of the percentage of myelinated axons that falls within the given range in the number of lamellae.
Discussion
During development and/or after nerve injury, the complex interactions between glial cells and neurons are responsible for the reciprocal regulation and dramatic modulation of gene expression in both cell types. It is the action of multiple axonal/glial factors involved in an intricate neuron/Schwann cell cross-talk that allows for a fundamental relationship conducive to the formation of the myelin sheath (for review, see refs. 2–4). Schwann cell/neuronal cocultures provide a powerful tool to dissect the complex interactions necessary for the formation of peripheral myelin by permitting the characterization of three major stages in the process, proliferation, premyelination, and myelination stages (45). A schematic diagram of these stages including the expression of the neurotrophins is shown in Fig. 7. The results described here show that the neurotrophins BDNF and NT3 are essential components in Schwann cell/neuronal interactions that prepare axons and/or Schwann cells for myelination. As determined by MAG and P0 synthesis, the active myelination stage begins in these cocultures by 2 days after induction with ascorbic acid. At this stage BDNF levels are significant but fall to undetectable levels 4 days after induction. That BDNF is indispensable for normal myelination is shown by the enhancement in the expression of the two myelin proteins both in the cocultures and in the sciatic nerve in vivo on addition of BDNF. This effect is further illustrated by the inhibition of myelin protein synthesis and the formation of myelin in cocultures when endogenous BDNF levels are reduced by addition of the receptor-based scavenger TrkB-Fc. The findings that BDNF increases the number of myelinating axons and the thickness of the myelin sheath in vivo add weight to the conclusion that BDNF is a key regulator of myelination and validate the coculture system as a reliable model for myelination. In keeping with this are the further in vivo observations that TrkB-Fc increases the number of axons that remain in an ensheathed, premyelinating stage and reduces the number of myelin wraps of the remaining axons. BDNF is initially secreted by the DRG neurons and inhibition of its synthesis only occurs after induction of myelination. The short-term (acute) modulation of neurotrophin levels was obtained by localized infusion of the factors or the scavengers in the area surrounding the sciatic nerve during development. This was done at a period in which the survival and innervation of the neurons had already been established. Therefore, the results obtained are the direct effect of the neurotrophins on the myelination process, and not an indirect effect due to the selection of a particular neuronal population. In agreement with this hypothesis, the ultrastructural studies did not reveal any differences in the total number of axons present in the BDNF- or TrkB-Fc-treated nerves compared with the contralateral control nerves (data not shown). NT3 on the other hand inhibits myelination and in keeping with this the levels of NT3 secreted by cocultures decrease steadily over 5 days after seeding the Schwann cells until they are undetectable after induction. When NT3 is added at the day of induction, myelin protein synthesis in cocultures is decreased significantly and very few myelin internodes are observed. Addition of TrkC-Fc to remove endogenous NT3 increases myelin protein synthesis and restores normal myelination, again indicative of an inhibitory effect of NT3 on myelination. Of interest is the observation that NT3 applied in vivo does not inhibit myelin protein expression as much as it does in coculture. This difference may be due, in part, to a timing effect. Whereas NT3 was added to cocultures during the premyelination stage, in vivo injection took place postnatally and myelination may have already been initiated. This result also implies that NT3 exerts its inhibitory action during premyelination, a hypothesis consistent with the slight inhibition of protein synthesis observed 2 days after NT3 injection in vivo and not after 4 days. The limitations of in vivo studies might also be considered. Differences from animal to animal were always larger than between cocultures and although the contralateral leg was used as a control there is no simple way of discerning whether the lack of effect is due to the loss of NT3 responsiveness at center stage of myelin formation or to a decrease bio-availability of the factor. Nevertheless, BDNF and NT3 clearly have different modulatory effects at different stages in the myelination program.
Figure 7.
Schematic model representing the modulation of the endogenous levels of neurotrophins and their possible roles during myelin formation. BDNF and NT3 are expressed at high levels during the initial phases of myelin development. Concomitant with the proliferation and premyelination phases, there is a marked decrease in NT3 levels, whereas BDNF remains constant. High levels of NT3 do not allow the myelination program to proceed further and keep the Schwann cell/axonal unit in an ensheathed premyelinated stage. When NT3 levels are diminished, the Schwann cell initiates the formation of a myelin internode surrounding the axon. On the contrary, high levels of BDNF are required for the myelination process to proceed and BDNF levels will decrease only after the myelination program has already been initiated. Elevated levels of BDNF during the early stages of myelination increase the speed and extent of the final process. An illustration of the timeline for the proliferation, premyelination, and myelination stages of Schwann cells in the coculture system appears below the model.
Previous studies on the role of BDNF in nerve injury experiments have documented an increase in the amount of PNS myelin (34, 35). Although these results confirm our own findings, it is difficult to exclude the influence of BDNF on the various stages preceding myelination, as well as its direct influence on the neuronal cells. In addition, a study from Pruginin-Bluger et al. (46) reported an increase in the expression of a Schwann cell myelin protein in cultures of quail nonneuronal cells in the presence of BDNF. Although the results of this study are quite similar to our data, the examination of mature myelin internodes was not explored. It seems evident from our results and that of others that BDNF is an essential component of the myelination program in the PNS.
Although the role of neurotrophins in central nervous system myelination remains unclear, studies with BDNF knockout mice (47) document hypomyelination of the optic nerve as compared with wild-type mice. However, no changes in myelin formation of the facial and sciatic nerves were observed in these mice. This study was conducted in mice between postnatal days 10–20. From the expression profile of myelin proteins during the development of the mouse sciatic nerve, ≈80% of the final MAG and P0 protein levels had already been reached by 10 days postnatal (data not shown). Therefore, studies performed on younger pups at the start of myelination may result in quite different conclusions. In addition, the same study reported abnormalities in the Schwann cell cytoplasm surrounding axons of the facial nerve (47), but did not examine the ultrastructure of the myelin sheath. By analyzing the average number of lamellae in the myelinated axons, differences in the myelin sheath may be distinguished from effects on the axonal caliber. Finally, whereas our results were obtained by using relatively short-term exposure with neurotrophins, the long-term deprivation of BDNF in vivo may result in compensatory mechanisms that result again in different conclusions.
Data concerning the effects of NT3 on myelin formation are even more scarce than that of BDNF and are always obscured by the predominant function of this neurotrophin in neuronal survival and differentiation. Newborn transgenic mice overexpressing NT3 under the control of the nestin promoter show an increase in the number of axons projecting to the spinal cord that parallels an increase in the number of DRG neurons (48). Over time, this effect is reversed so that NT3 overexpression produces a reduction in the number of DRG neurons and myelinated axons in adult animals. Although the reduction in the number of neuronal cells is only 20%, there is 37% depletion in the number of myelinated axons in the dorsal roots. Although the number of surviving axons in the overexpressing animals and the ultrastructure of the myelin sheath were not analyzed in this study, the effect of NT3 overexpression on the number of myelinated axons compared with the number of surviving neurons is in agreement with a negative role of NT3 during myelin formation.
Our results demonstrate that endogenous neurotrophins are key mediators of the myelination program in the PNS. The obvious therapeutic implications of these findings relate specifically to the demyelinating peripheral neuropathies and to nerve injury. It is hoped that this previously uncharacterized role for neurotrophins on myelination will aid in separating and distinguishing the factors involved in the complex process of axonal regeneration from the process of remyelination. Neurons and Schwann cells share a mutual dependence in establishing or reestablishing a functional relationship through multiple axonal/glial signals. The mechanism of neurotrophin signaling is complex and depends on numerous factors. These signaling events rely heavily on the accessibility of the full-length and truncated Trk receptors and on the p75NTR, the relative binding affinities and specificities of the ligands, and the relative amounts of the receptors and ligands present in the particular system (for a review, see refs. 49 and 50). By elucidating the mechanism of neurotrophin action on the myelination process, and characterizing this previously uncharacterized neuronal–glial interaction, new therapeutic strategies into myelin repair and the functional recovery of demyelinating peripheral neuropathies may be possible.
Acknowledgments
We thank Drs. B. Barres and A. Kruttgen for insightful discussions and for reviewing the manuscript. We also thank Dr. R. Tolwani, Dr. N. Liu, and S. Varma for their technical advice. We are grateful to Regeneron Pharmaceuticals for the generous gifts of BDNF, NT3, TrkB-Fc, and TrkC-Fc. In addition, we thank Nafisa Ghori for her expertise and work with the electron microscopy. This work has been supported by grants from the National Institute of Neurological Disorders and Stroke (National Institutes of Health), the Muscular Dystrophy Association, the Christopher Reeve Paralysis Association, the McGowen Charitable Trust (to E.M.S.), the National Institutes of Health Neurobiology Postdoctoral Training Fellowship (to J.R.C.), and a postdoctoral fellowship from the Ministerio de Educación y Cultura, Spain (to J.M.C.).
Abbreviations
- BDNF
brain-derived neurotrophic factor
- NT3
neurotrophin-3
- MAG
myelin-associated glycoprotein
- NGF
nerve growth factor
- PNS
peripheral nervous system
- DRG
dorsal root ganglia
Footnotes
This contribution is part of the special series of Inaugural Articles by members of the National Academy of Sciences elected on May 2, 2000.
References
- 1.Lupski J R, Garcia C A. Brain Pathol. 1992;2:337–349. doi: 10.1111/j.1750-3639.1992.tb00710.x. [DOI] [PubMed] [Google Scholar]
- 2.Lemke G. Annu Rev Neurosci. 2001;24:87–105. doi: 10.1146/annurev.neuro.24.1.87. [DOI] [PubMed] [Google Scholar]
- 3.Martini R. Muscle Nerve. 2001;24:456–466. doi: 10.1002/mus.1027. [DOI] [PubMed] [Google Scholar]
- 4.Reynolds M L, Woolf C J. Curr Opin Neurobiol. 1993;3:683–693. doi: 10.1016/0959-4388(93)90139-p. [DOI] [PubMed] [Google Scholar]
- 5.Bunge M B, Williams A K, Wood P M. Dev Biol. 1982;92:449–460. doi: 10.1016/0012-1606(82)90190-7. [DOI] [PubMed] [Google Scholar]
- 6.Bunge R P. Curr Opin Neurobiol. 1993;3:805–809. doi: 10.1016/0959-4388(93)90157-t. [DOI] [PubMed] [Google Scholar]
- 7.Mirsky R, Jessen K R. Curr Opin Neurobiol. 1996;6:89–96. doi: 10.1016/s0959-4388(96)80013-4. [DOI] [PubMed] [Google Scholar]
- 8.Levi-Montalcini R. Science. 1987;237:1154–1162. doi: 10.1126/science.3306916. [DOI] [PubMed] [Google Scholar]
- 9.Shooter E M. Annu Rev Neurosci. 2001;24:601–629. doi: 10.1146/annurev.neuro.24.1.601. [DOI] [PubMed] [Google Scholar]
- 10.Radeke M J, Misko T P, Hsu C, Herzenberg L A, Shooter E M. Nature (London) 1987;325:593–597. doi: 10.1038/325593a0. [DOI] [PubMed] [Google Scholar]
- 11.Heumann R, Lindholm D, Bandtlow C, Meyer M, Radeke M J, Misko T P, Shooter E, Thoenen H. Proc Natl Acad Sci USA. 1987;84:8735–8739. doi: 10.1073/pnas.84.23.8735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Johnson E M, Jr, Taniuchi M, DiStefano P S. Trends Neurosci. 1988;11:299–304. doi: 10.1016/0166-2236(88)90090-2. [DOI] [PubMed] [Google Scholar]
- 13.Barde Y A, Edgar D, Thoenen H. EMBO J. 1982;1:549–553. doi: 10.1002/j.1460-2075.1982.tb01207.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Berkemeier L R, Winslow J W, Kaplan D R, Nikolics K, Goeddel D V, Rosenthal A. Neuron. 1991;7:857–866. doi: 10.1016/0896-6273(91)90287-a. [DOI] [PubMed] [Google Scholar]
- 15.Hallbook F, Ibanez C F, Persson H. Neuron. 1991;6:845–858. doi: 10.1016/0896-6273(91)90180-8. [DOI] [PubMed] [Google Scholar]
- 16.Ernfors P, Ibanez C F, Ebendal T, Olson L, Persson H. Proc Natl Acad Sci USA. 1990;87:5454–5458. doi: 10.1073/pnas.87.14.5454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hohn A, Leibrock J, Bailey K, Barde Y A. Nature (London) 1990;344:339–341. doi: 10.1038/344339a0. [DOI] [PubMed] [Google Scholar]
- 18.Maisonpierre P C, Belluscio L, Squinto S, Ip N Y, Furth M E, Lindsay R M, Yancopoulos G D. Science. 1990;247:1446–1451. doi: 10.1126/science.247.4949.1446. [DOI] [PubMed] [Google Scholar]
- 19.Rosenthal A, Goeddel D V, Nguyen T, Lewis M, Shih A, Laramee G R, Nikolics K, Winslow J W. Neuron. 1990;4:767–773. doi: 10.1016/0896-6273(90)90203-r. [DOI] [PubMed] [Google Scholar]
- 20.Sebert M E, Shooter E M. J Neurosci Res. 1993;36:357–367. doi: 10.1002/jnr.490360402. [DOI] [PubMed] [Google Scholar]
- 21.Funakoshi H, Frisen J, Barbany G, Timmusk T, Zachrisson O, Verge V M, Persson H. J Cell Biol. 1993;123:455–465. doi: 10.1083/jcb.123.2.455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Barde Y A. Prog Growth Factor Res. 1990;2:237–248. doi: 10.1016/0955-2235(90)90021-b. [DOI] [PubMed] [Google Scholar]
- 23.McAllister A K, Katz L C, Lo D C. Annu Rev Neurosci. 1999;22:295–318. doi: 10.1146/annurev.neuro.22.1.295. [DOI] [PubMed] [Google Scholar]
- 24.Snider W D. Cell. 1994;77:627–638. doi: 10.1016/0092-8674(94)90048-5. [DOI] [PubMed] [Google Scholar]
- 25.Thoenen H. Science. 1995;270:593–598. doi: 10.1126/science.270.5236.593. [DOI] [PubMed] [Google Scholar]
- 26.Snider W D, Lichtman J W. Mol Cell Neurosci. 1996;7:433–442. doi: 10.1006/mcne.1996.0031. [DOI] [PubMed] [Google Scholar]
- 27.Tessarollo L. Cytokine Growth Factor Rev. 1998;9:125–137. doi: 10.1016/s1359-6101(98)00003-3. [DOI] [PubMed] [Google Scholar]
- 28.Bibel M, Barde Y A. Genes Dev. 2000;14:2919–2937. doi: 10.1101/gad.841400. [DOI] [PubMed] [Google Scholar]
- 29.Barres B A, Schmid R, Sendnter M, Raff M C. Development (Cambridge, UK) 1993;118:283–295. doi: 10.1242/dev.118.1.283. [DOI] [PubMed] [Google Scholar]
- 30.Barres B A, Raff M C, Gaese F, Bartke I, Dechant G, Barde Y A. Nature (London) 1994;367:371–375. doi: 10.1038/367371a0. [DOI] [PubMed] [Google Scholar]
- 31.McTigue D M, Horner P J, Stokes B T, Gage F H. J Neurosci. 1998;18:5354–5365. doi: 10.1523/JNEUROSCI.18-14-05354.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Anton E S, Weskamp G, Reichardt L F, Matthew W D. Proc Natl Acad Sci USA. 1994;91:2795–2799. doi: 10.1073/pnas.91.7.2795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bentley C A, Lee K F. J Neurosci. 2000;20:7706–7715. doi: 10.1523/JNEUROSCI.20-20-07706.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Zhang J Y, Luo X G, Xian C J, Liu Z H, Zhou X F. Eur J Neurosci. 2000;12:4171–4180. [PubMed] [Google Scholar]
- 35.Namiki J, Kojima A, Tator C H. J Neurotrauma. 2000;17:1219–1231. doi: 10.1089/neu.2000.17.1219. [DOI] [PubMed] [Google Scholar]
- 36.Shelton D L, Sutherland J, Gripp J, Camerato T, Armanini M P, Phillips H S, Carroll K, Spencer S D, Levinson A D. J Neurosci. 1995;15:477–491. doi: 10.1523/JNEUROSCI.15-01-00477.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Chan J R, Rodriguez-Waitkus P M, Ng B K, Liang P, Glaser M. Mol Biol Cell. 2000;11:2283–2295. doi: 10.1091/mbc.11.7.2283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Laemmli U K. Nature (London) 1970;227:680–685. doi: 10.1038/227680a0. [DOI] [PubMed] [Google Scholar]
- 39.Notterpek L, Snipes G J, Shooter E M. Glia. 1999;25:358–369. [PubMed] [Google Scholar]
- 40.Kruttgen A, Moller J C, Heymach J V, Jr, Shooter E M. Proc Natl Acad Sci USA. 1998;95:9614–9619. doi: 10.1073/pnas.95.16.9614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Ernfors P, Wetmore C, Olson L, Persson H. Neuron. 1990;5:511–526. doi: 10.1016/0896-6273(90)90090-3. [DOI] [PubMed] [Google Scholar]
- 42.Mu X, Silos-Santiago I, Carroll S L, Snider W D. J Neurosci. 1993;13:4029–4041. doi: 10.1523/JNEUROSCI.13-09-04029.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Acheson A, Barker P A, Alderson R F, Miller F D, Murphy R A. Neuron. 1991;7:265–275. doi: 10.1016/0896-6273(91)90265-2. [DOI] [PubMed] [Google Scholar]
- 44.Chan J R, Phillips L J, II, Glaser M. Proc Natl Acad Sci USA. 1998;95:10459–10464. doi: 10.1073/pnas.95.18.10459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Kleitman N, Wood P M, Bunge R P. In: Culturing Nerve Cells. Banker G, Goslin K, editors. Cambridge, MA: MIT Press; 1991. pp. 337–377. [Google Scholar]
- 46.Pruginin-Bluger M, Shelton D L, Kalcheim C. Mech Dev. 1997;61:99–111. doi: 10.1016/s0925-4773(96)00623-5. [DOI] [PubMed] [Google Scholar]
- 47.Cellerino A, Carroll P, Thoenen H, Barde Y A. Mol Cell Neurosci. 1997;5:397–408. doi: 10.1006/mcne.1997.0641. [DOI] [PubMed] [Google Scholar]
- 48.Ringstedt T, Kucera J, Lendahl U, Ernfors P, Ibanez C F. Development (Cambridge, UK) 1997;124:2603–2613. doi: 10.1242/dev.124.13.2603. [DOI] [PubMed] [Google Scholar]
- 49.Patapoutian A, Reichardt L F. Curr Opin Neurobiol. 2001;11:272–280. doi: 10.1016/s0959-4388(00)00208-7. [DOI] [PubMed] [Google Scholar]
- 50.Lee F S, Kim A H, Khursigara G, Chao M V. Curr Opin Neurobiol. 2001;11:281–286. doi: 10.1016/s0959-4388(00)00209-9. [DOI] [PubMed] [Google Scholar]