Neu differentiation factor regulates tau protein and mRNA in cultured neonatal oligodendrocytes.

Axonal signals activate myelinogenesis via regulation of the extent to which oligodendrocyte (OLG) processes wrap around the axon. The cytoskeleton in OLG processes is actively involved in myelination and is a putative target for axonal regulation of myelination. The axon-associated neuregulins may regulate the cytoskeleton extensions in OLG processes. Here, we report that the neuregulin neu differentiation factor (NDF) increases the expression of tau mRNA and tau protein in OLGs. Treatment of neonatal OLGs with alpha-NDF or beta-NDF resulted in dramatic increases in the length of OLG processes, which appeared either as singular unbranched extensions or as a network of extensively branched processes. By immunoblot analysis with tau-1 mAb, which recognizes the dephosphorylated form of the tau proteins, neonatal OLGs treated with alpha-NDF or beta-NDF, had an increase in tau protein levels. The increase of tau levels in beta-NDF-treated cells is much greater than the twofold increase present in alpha-NDF-treated cells. By immunoblot analysis with the phosphorylation-insensitive tau-5 mAb, beta-NDF-treated cells had a twofold increase in tau. Immunoblot analysis suggest that alpha-NDF and beta-NDF promote a twofold increase in the tau protein levels in OLG, with the beta-factor also promoting a tau dephosphorylation. Using promoters spanning the amino-terminal region of tau, we found that OLGs treated with alpha-NDF or beta-NDF contained approximately twofold more tau mRNA than untreated cells. However, there was no qualitative difference between control and NDF-treated cells in the pattern of tau mRNA isoforms expressed. A model is proposed in which the axonal NDF-induced regulation of tau expression in OLGs may be part of the mechanism by which the axon regulates myelination.

Pigment epithelium-derived factor (PEDF) in neuroblastoma: a multifunctional mediator of Schwann cell antitumor activity.

Neuroblastoma is notable for its cellular heterogeneity and unpredictable outcome. Tumors are a variable mixture of primitive malignant neuroblasts, more differentiated ganglionic cells, Schwann and endothelial cells. Although often fatal, neuroblastomas can spontaneously regress, possibly due to favorable autocrine and paracrine interactions among these cells. Here, pigment epithelium-derived factor (PEDF), a potent inhibitor of angiogenesis and inducer of neural differentiation, is shown to be produced by ganglionic cells and Schwann cells, but not by more primitive tumor cells. Although undifferentiated neuroblastoma tumor cell secretions were angiogenic primarily due to vascular endothelial growth factor, secretions of Schwann cells were anti-angiogenic due to PEDF. In addition, PEDF was the major factor responsible for Schwann cell's ability to induce tumor cell differentiation in vitro and recombinant PEDF had the same effect in vitro and in vivo. Both the growth and the survival of Schwann cells were enhanced by PEDF. Thus PEDF may serve as a multifunctional antitumor agent in neuroblastomas, inhibiting angiogenesis while promoting the numbers of Schwann cells and differentiated tumor cells that in turn produce PEDF, suggesting that its clinical administration could stimulate a multifaceted antitumor feedback loop with the potential to limit and possibly regress tumor growth.

Myelin basic protein (MBP) and MBP peptides are mitogens for cultured astrocytes.

After CNS demyelination, astrogliosis interferes with axonal regeneration and remyelination. We now provide evidence that myelin basic protein (MBP) can contribute to this observed astrocyte proliferation. We found that astrocytes grown in either serum-containing or serum-free medium proliferate in response to MBP. The mitogenic regions of MBP in both media were MBP(1-44), MBP(88-151) and MBP(152-167). The mitogenic effect of these individual peptides was potentiated by simultaneous treatment with microglia conditioned media (CM). MBP-induced proliferation was inhibited by suramin at concentrations known to block the fibroblast growth factor receptor (FGFR), whereas neither MBP(1-44), MBP(88-151) nor MBP(152-167) were affected. Cholera toxin B, that binds to ganglioside GM(1), inhibited the mitogenicity of MBP(1-44) and had no significant effect on the mitogenicity of MBP, MBP(88-151) or MBP(152-167). Treatment of astrocytes with MBP and MBP(152-167) caused a modest and transitory elevation of intracellular calcium, whereas treatment with MBP(1-44) resulted in a substantial and sustained increase in intracellular calcium. These results suggest that for cultured astrocytes 1) FGFR and extracellular calcium play a major role in MBP mitogenicity; 2) MBP(1-44), MBP(88-151) and MBP(152-167) are the mitogenic regions of MBP; 3) MBP(1-44) and MBP(152-167) interact with ganglioside GM(1) and FGFR, respectively; 4) Component(s) present in microglial CM potentiate the mitogenicity of MBP(1-44), MBP(88-151) and MBP(152-167). These data support the hypothesis that MBP related peptides in conjunction with microglial secreted factors may stimulate astrogliosis after demyelination in vivo.

Isolation and serum-free culture of primary Schwann cells from human fetal peripheral nerve.

We have developed a method for isolating Schwann cells (SC) from human fetal peripheral nerve and maintaining these SC in vitro under serum-free conditions. This method yields essentially pure SC which have a bipolar, spindle-shaped morphology; align in fascicles; and express typical glial cell markers. Human fetal SC can be maintained for months under serum-free conditions with the neuregulin NDF beta. These human fetal SC can mimic axonal contact in vivo by retaining the functional capacity to strongly associate with neurites of cultured human fetal dorsal root ganglia. These isolation, culture, and coculture techniques provide a method for investigating SC-neuron interactions as well as development and function of human fetal SC.

Neurofibrosarcoma-derived Schwann cells overexpress platelet-derived growth factor (PDGF) receptors and are induced to proliferate by PDGF BB.

Neurofibromatosis type 1 (NF1) is characterized by the formation of neurofibromas, benign tumors of the peripheral nerve consisting essentially of Schwann cells, which can sometimes turn malignant to form neurofibrosarcomas. The mechanism of progression toward a malignant phenotype remains largely unknown. In this report, we show that platelet-derived growth factor (PDGF) BB, and to a lesser extent fibroblast growth factor 2, are mitogenic for two neurofibrosarcoma-derived Schwann cell lines, but not for a Schwann cell line derived from a schwannoma (from a non-NF1 patient) or for transformed rat Schwann cells. Levels of expression of both PDGF receptor alpha and beta are significantly increased in the two neurofibrosarcoma-derived cell lines compared to the non-NF1 Schwann cell lines. The level of tyrosyl-phosphorylated PDGF receptor beta is strongly increased upon stimulation by PDGF BB. In comparison, only modest levels of tyrosyl-phosphorylated PDGF receptor alpha are observed, upon stimulation by PDGF AA or PDGF BB. Accordingly, PDGF AA is only a weak mitogen for the neurofibrosarcoma-derived cells by comparison to PDGF BB. These results indicate that the mitogenic effect of PDGF BB for the neurofibrosarcoma-derived Schwann cell lines is primarily transduced by PDGF receptor beta. Neu differentiation factor beta, a potent mitogen for normal Schwann cells, was unable to stimulate proliferation of the transformed Schwann cell lines, due to a dramatic down-regulation of the erbB3 receptor. Therefore, aberrant expression of growth factor receptors by Schwann cells, such as the PDGF receptors, could represent an important step in the process leading to Schwann cell hyperplasia in NF1.

Expression of Kit in neurofibromin-deficient human Schwann cells: role in Schwann cell hyperplasia associated with type 1 neurofibromatosis.

Type 1 Neurofibromatosis (NF1) is characterized by the formation of neurofibromas, benign tumors composed mainly of Schwann cells, which can turn malignant to form neurofibrosarcomas. Neurofibromin, the protein product of the Nf1 gene, is believed to act as a tumor suppressor, accelerating the conversion of the oncogene Ras to its inactive form. The absence of neurofibromin could therefore lead to higher Ras activity in Schwann cells, resulting in uncontrolled growth through a cascade of events not yet elucidated. We describe the abnormal expression of high levels of the Kit tyrosine kinase receptor in both NF1-derived Schwann cell lines and tissue, as compared to primary Schwann cells or schwannoma-derived cells. High levels of Kit expression in the neurofibrosarcoma-derived Schwann cells correlate with a decrease in neurofibromin expression. Using inhibitors of tyrosine kinase receptors, we found that proliferation of the neurofibrosarcoma-derived cells is dependent upon activation of a subclass of tyrosine-kinase receptors. The proliferation of these cells is not dependent upon an autocrine loop involving typical Schwann cell mitogens. Finally, the proliferation of the neurofibrosarcoma-derived Schwann cells can be increased by stimulation with Kit ligand. These data implicate Kit as one of the components leading to the Schwann cell hyperplasia observed in NF1.

Schwann cells stimulated by axolemma-enriched fractions express cyclic AMP responsive element binding protein.

Both axolemma-enriched fractions (AEF) and cyclic AMP have been shown to regulate the proliferation and differentiation of cultured primary Schwann cells (SC). We have evaluated the role of CREB, a transcription factor that binds to the cAMP-responsive element, in mediating the AEF-stimulated SC proliferation and differentiation. We detected CREB in nuclear extracts derived from SC stimulated with 40 micrograms/ml of AEF for 16, 24, 48, 72, and 96 hr, using a DNA-electrophoretic mobility shift assay. Unstimulated quiescent SC contained low levels of CREB which increased to a maximal level after 48 hr of AEF treatment. Using anti-CREB antibodies and Western blot analysis, after 24 hr of AEF treatment we first detected CREB as a 45 kDa protein which reached a maximal level of expression after 72 hr. Double labeled immunocytochemistry using anti-CREB and anti-5-bromo-2'-deoxy-uridine antibodies demonstrated maximal CREB expression after 72 hr of AEF treatment, closely coinciding with the temporal expression of SC proliferation. At all times examined, all AEF-treated SC labeled by anti-CREB antibodies were also labeled with anti-BrdU antibodies. These observations are consistent with the view that CREB could play an important role in the induction of SC proliferation by AEF.

Glucocorticoids enhance the potency of Schwann cell mitogens.

Previous studies have documented that cultured Schwann cells require serum-containing medium to respond maximally to mitogens. We now report that Schwann cells are able to proliferate to a mitogenic response in a serum-free defined medium termed oligodendrocyte defined media (ODM). Glucocorticoids are the essential component of ODM which allow Schwann cell proliferation in the serum-free medium. Charcoal treatment of the fetal calf serum decreases the mitogenic potency of the axolemma-enriched fraction (AEF) by 50%. The addition of 2 microM hydrocortisone to charcoal-treated fetal calf serum restores 75% of the lost mitogenicity. These observations are consistent with the view that glucocorticoids present in fetal calf serum are potent co-mitogens essential for AEF-induced Schwann cell proliferation. The synthetic glucocorticoid, dexamethasone, is a more potent co-mitogen than hydrocortisone, with a maximal effect at concentrations less than 10 nM. In contrast, other steroids including aldosterone, progesterone, testosterone, and 17 beta-estradiol have no effect on enhancing the mitogenic response of Schwann cells to the AEF. The glucocorticoid antagonists RU 486 and dehydroepiandrosterone (DHEA), but not the antiestrogenic compound tamoxifen, block AEF-induced Schwann cell proliferation. These results suggest that glucocorticoid-induced Schwann cell proliferation is mediated through a glucocorticoid receptor (GR) mechanism. We detected immunoreactivity to the GR in the cytoplasm, but not in the nuclei of Schwann cells grown in ODM lacking dexamethasone. The addition of 100 nM dexamethasone to these cultures resulted in immunoreactivity in the nucleus. This data suggests that glucocorticoids working through the GR are potent co-mitogens for Schwann cell proliferation.

cAMP-induced morphological changes in an immortalized Schwann cell line: a prelude to differentiation?

Schwann cells (SC), the myelinating cells of the peripheral nervous system, show a remarkable capacity to switch from a differentiated state to a proliferative state both during development and peripheral nerve regeneration. In order to better understand the regulatory mechanisms involved with this change we are studying a Schwann cell line transfected with the SV-40 large T gene (TSC). Serum-free medium combined with elevating intra-cellular cAMP levels produced a slower proliferating TSC whose morphology changed from pleiomorphic to process bearing, reminiscent of primary SC in culture. This change was abrogated by colcemid but was unaltered by cytochalasin D, indicating a major role for microtubules. Ultrastructural studies demonstrated numerous microtubules in the cellular extensions which correlated with strong immunocytochemical staining for tubulin in the processes. Analysis of cytoskeletal fractions from the treated cells revealed a greater proportion of tubulin in the polymerized state compared with untreated cells which closely resembled the distribution in primary SC. The cytoskeletal changes observed in the TSC as a result of elevating the intra-cellular cAMP levels may reflect the earliest cellular changes in the induction of myelination.

Distribution of fibroblast growth factor in cultured dorsal root ganglion neurons and Schwann cells. I. Localization during maturation in vitro.

The distribution of acidic and basic fibroblast growth factor in co-cultures of dorsal root ganglion neurons and Schwann cells was examined as a function of time in culture. After two days in vitro, the cytoplasm of the neuronal cell bodies demonstrated both acidic and basic FGF immunoreactivity, whereas the cytoplasm of the neurites was not immunoreactive. Schwann cells, in contrast, exhibited both acidic and basic fibroblast growth factor cytoplasmic immunoreactivity. After two days in culture, immunoreactivity was not detected on the plasma membrane surface of either the neurons or the Schwann cells. By 10 days in vitro, fibroblast growth factor immunoreactivity was observed in the cytoplasm of the most proximal portion of some, but not all, neurites but was unchanged in Schwann cells. At 20 days in vitro, immunoreactivity was still restricted to the intracellular compartment of both Schwann cells and neurons. Acidic fibroblast growth factor was primarily localized to the cytoplasm of Schwann cells, neuron cell bodies and along the entire length of the neurites. In contrast, basic fibroblast growth factor was predominantly localized to the nuclei of Schwann cells and small to medium size neurons. In many cases, the nucleolar region demonstrated the most intense basic fibroblast growth factor. The cytoplasm of the neurites was also immunoreactive for basic fibroblast growth factor. At 30 days in vitro the intracellular distribution of fibroblast growth factor immunoreactivity was similar to that observed at 20 days. However, both acidic and basic fibroblast growth factor were detected on the surface of the neurites. In contrast, no fibroblast growth factor immunoreactivity was detected at the Schwann cell surface at any time point examined. The distribution of fibroblast growth factor in Schwann cells cultured by themselves was similar to that of Schwann cells co-cultured with neurons after 20 days in vitro. Both Schwann cells and dorsal root ganglia exhibited increased fibroblast growth factor immunoreactivity with increased time in culture and an increased expression of basic fibroblast growth factor in the nucleus. Of particular interest was the appearance of fibroblast growth factor on the surface of neurites after 30 days in vitro where it could function to modulate neuron-glial cell interactions.

Distribution of fibroblast growth factor in cultured dorsal root ganglion neurons and Schwann cells. II. Redistribution after neural injury.

The localization of fibroblast growth factor was examined in both immature (< 20 days in vitro) and mature (> 30 days in vitro) dorsal root ganglion neuron-glial cell co-cultures as a function of time after in vitro crush injury of the neurites. In the 20 day cultures, neuritic membrane vesicles were seen adhering to Schwann cells following neurite injury. Fibroblast growth factor was not detected on the surface of these membrane vesicles when they were associated with either the degenerating neurites or the surface of Schwann cells. However, the cytoplasm of the Schwann cells demonstrated fibroblast growth factor immunoreactivity at all times. In contrast, injury to neurites after 30 days in vitro resulted in demonstrable fibroblast growth factor immunoreactivity on the surfaces of the neuritic membrane vesicles both before and after their association with the Schwann cells. Furthermore, there was a change in the pattern of fibroblast growth factor immunoreactivity on the surface of Schwann cells after injury: initially the staining was patchy but with increasing time it became more uniform and more intense. A similar pattern of staining was noted on the surface of oligodendrocytes co-cultured with dorsal root ganglion neurons. However, astrocytes which were co-cultured with dorsal root ganglion neurons did not show any fibroblast growth factor immunoreactivity. Also, after injury at 30 days in vitro, the neuronal cell bodies began to express fibroblast growth factor immunoreactivity on their extracellular surfaces and the regenerating neurites exhibited fibroblast growth factor immunoreactive material on the surface of their plasma membranes. This redistribution of fibroblast growth factor via degenerating neuritic membrane vesicles to the plasma membrane of Schwann cells may be involved in neuronal signalling to glial cells after neuronal injury.

Release of membrane-associated growth factors during neural injury.

The release of membrane-associated growth factors after neural injury may influence the outcome of the recovery. For example, for remyelination to occur after neural injury it is critical for the glial cell to proliferate prior to remyelination in both the PNS and CNS. In the CNS, the relative response of the oligodendrocytes and astroglia to growth factors mobilized during neural injury may play a role in the cellular dynamics of repair of neural injury or scarring and subsequent failure to repair neural injury. In support of this view, we have studied the mitotic potential and cell cycle kinetics of cultured adult oligodendrocytes and found that these adult cells respond only weakly to factors such as FGF which are known to be potent mitogens for neonatal cells. However, given the same dose of FGF, adult astrocytes are mitotically stimulated to a much greater degree than are the adult oligodendrocytes (Vick and De Vries, unpublished observations). Given the pathways which may be operative in the release of growth factors after injury, it has not escaped our attention that, provided the released factors are in equilibrium with easily accessible and peripheral body fluids, these released factors may serve as new markers for neural injury. Further experiments are in progress to explore this possibility.

Surface behavior of axolemma monolayers: physico-chemical characterization and use as supported planar membranes for cultured Schwann cells.

The axolemma membrane forms a stable and reproducible monomolecular layer at the air-aqueous interface. The major lipids and proteins are present in this monolayer in molar ratios similar to the original membrane. Acetylcholinesterase and Na-K-ATPase activities are preserved in the monolayer to levels of 64% and 25%, respectively. The total lipid fraction forms a homogeneously mixed phase. The presence of proteins in the monolayer introduces surface inhomogeneties. Among other features, this is revealed by the presence of two values of lateral pressure at which the monolayer shows partial or total collapse: a broad partial collapse at surface pressures between 13 to 30 mN/m and a sharp collapse point at 46 mN/m. The average molecular areas, the broad collapse point, and the variation of the surface potential per molecule suggest the relocation of protein components at surface pressures between 13 to 30 mN/m. The behavior is consistent with the extrusion and exposure of proteins toward the aqueous medium that depends on the lateral pressure. Schwann cells grown on coverslips coated with axolemma monolayers at 13 mN/m (beginning of the broad collapse) and 34 mN/m (above the broad collapse) recognize the difference in the surface organization of axolemma caused by the lateral pressure which affects their proliferation, morphology, and spatial pattern of organization. Our results show for the first time that response of Schwann cells depends on the intermolecular organization of the axolemma surface with which they interact. These results suggest that the local expression of putative surface molecules of axolemma that may mediate membrane recognition and the signalling of morphological and proliferative changes can be modulated by long range supramolecular properties.

Evidence for the colocalization of the axonal mitogen for Schwann cells and oligodendrocytes.

Axons that normally will encounter either CNS or PNS glia have been shown to contain a powerful mitogen for both Schwann cells and oligodendrocytes. The normally nonmyelinated, nonglial ensheathed cerebellar granule cells have been shown to possess a proliferative signal for Schwann cells, suggesting that a glial mitogen is common to all axons. To determine if a glial mitogen capable of stimulating both Schwann cells and oligodendrocytes is colocalized on all types of axons we have (1) cocultured granule cells with oligodendrocytes, (2) incubated oligodendrocytes with granule cell membranes, and (3) evaluated the ability of heparin extracts of granule cell membranes, splenic nerve microsomes, and axolemma-enriched fractions isolated from rat and bovine CNS to stimulate mitosis of cultured oligodendrocytes. Neither the intact granule cells nor the granule cell membrane fraction stimulated cultured oligodendrocytes to divide. However, heparin extracts of the granule cell membranes were significantly mitogenic to the cultured oligodendrocytes. Heparin extracts of splenic nerve microsomes were more mitogenic than the comparable extract obtained from bovine CNS axolemma-enriched fractions. These results suggest that the neuronal mitogen for oligodendroglia is colocalized with the neuronal mitogen for Schwann cells.

Monoclonal antibody production to human and bovine 2':3'-cyclic nucleotide 3'-phosphodiesterase (CNPase): high-specificity recognition in whole brain acetone powders and conservation of sequence between CNP1 and CNP2.

Monoclonal antibodies against human and bovine 2':3'-cyclic nucleotide 3'-phosphodiesterase (CNPase) were generated by fusing FOX-NY myeloma cells with spleen cells from RBF/Dn mice previously immunized with the purified brain antigens. The enzyme isolated from bovine brain was quite basic, with an isoelectric point of 9.71 and both the bovine and human enzymes consisted of a closely spaced doublet at approximately 44 and 46 kDa on SDS-PAGE. Six monoclonals were were identified as strongly recognizing the enzyme on both ELISA plates and on immunoblots of whole brain protein. Four monoclonals very weakly cross-reacted with guinea pig myelin basic protein. In contrast with two previous reports, some of our monoclonal antibodies did immunostain 2 or 3 protein bands in peripheral nerve, two bands closely corresponding to those immunostained in central nervous system (CNS) myelin, the Wolfgram protein fraction and in acetone powders of whole brain. Each of the 6 monoclonals reacting strongly on immunoblots recognized the enzyme in from 2 to 5 of the species examined (human, bovine, rat, mouse and rabbit). In addition, all 6 monoclonals that immunostained the enzyme in whole brain, myelin and Wolfgram protein immunoblots recognized both CNP1 (44 kDa) and CNP2 (46 kDa). The two closely spaced protein bands observed on SDS-PAGE and previously stained on immunoblots of CNS CNPase using polyvalent rabbit anti-bovine CNPase antisera, and now different monoclonal antibodies, appear to be immunologically related and to contain highly conserved sequences.

Freeze-fracture characterization of isolated myelin and axolemma membrane fractions.

The macromolecular organization of membranes isolated from the rabbit optic nerve and tract was analyzed using the freeze-fracture technique. A myelin fraction and two axolemma-enriched fractions were prepared from a preparation of myelinated axons isolated by flotation in a buffered salt-sucrose medium. In the myelinated axon preparation, axolemma and myelin membranes were easily identified. Larger areas of the axon membrane and myelin membrane totally lacked intramembranous particles. The particles remaining on the myelin membrane formed patches of evenly distributed elongated and globular particles. In contrast, the particles remaining on the axolemma were globular in shape and tightly clustered. Particle clustering and particle-free areas were not characteristic of either the axolemma or myelin membrane of whole nerves fixed in situ and processed for freeze-fracture. The isolated myelin membrane fraction contained a large number of vesicles completely lacking intramembranous particles. Of the remaining membrane vesicles, profiles with dispersed elongated and globular particles predominated. A small percentage of vesicles displayed intramembranous particles of the same size, shape and clustering pattern as that seen on the axolemma of the myelinated axon preparation. The two axolemma fractions were enriched in membrane containing tightly clustered globular particles. Particle-free vesicles as well as some myelin membrane vesicles were also seen in the axolemma fractions.

The protein composition of bovine myelin-free axons.

The proteins of axons prepared from myelinated axons and isolated as myelin-free entities were separated by sodium dodecyl sulfate polyacrylamide electrophoresis and found to consist of more than 10 different molecular weight species. The molecular weights range from 13 000 to over 200 000 with a prominent protein of molecular weight 47 000. The amino acid composition of the seven major proteins showed that the protein with a molecular weigt of 47 000 is distinct from all the other proteins analyzed. A group of three low molecular weight proteins have amino acid compositions which are similar to each other as do a group of three high molecular weight proteins although the two groups are distinctly different from each other and the major axonal protein. Histones, DNA, myelin basic protein and glycoprotein were absent from the proteins but neurotubule protein was present as indicated by cochicine binding activity in the axonal preparations. The cellular origin of these proteins and their relationship to other central nervous system proteins are discussed.