Glycolipids and transmembrane signaling: antibodies to galactocerebroside cause an influx of calcium in oligodendrocytes.

This is the first study to provide evidence that one function for the surface glycolipid galactocerebroside (GalC) is participation in the opening of Ca2+ channels in oligodendroglia in culture. This glycolipid is a unique differentiation marker for myelin-producing cells; antibodies to GalC have been shown to markedly alter oligodendroglial morphology via disruption of microtubules (Dyer, C. A., and J. A. Benjamins. 1988. J. Neurosci. 8:4307-4318). This study demonstrates that extracellular EGTA blocks anti-GalC-induced disassembly of microtubules in oligodendroglial membrane sheets, demonstrating that an influx of extracellular Ca2+ mediates the cytoskeletal changes. The Ca2+ influx was examined directly by loading oligodendroglia with the fluorescent dye Indo-1 in defined medium, and measuring changes in Ca2+ in individual cells with a laser cytometer. Upon addition of anti-GalC IgG, a marked sustained increase in intracellular Ca2+ occurred in 80% of the oligodendroglia observed. EGTA blocked the increase, indicating the increase is due to an influx of extracellular Ca2+, and not due to release from intracellular stores. The effect is specific, since Ca2+ levels remain normal in oligodendroglia treated with nonimmune IgG; astrocytes do not respond to the anti-GalC. The Ca2+ response in oligodendrocytes is dependent on concentration of antibody and GalC on the oligodendroglial membrane surface. The Ca2+ influx is not mediated by voltage-sensitive Ca2+ channels: it is not blocked by cadmium, and depolarization with K+ does not mimic the response. The kinetics of the response suggest that second messenger-mediated opening of Ca2+ channels is involved.

Agonist activity of the 3-hydroxy metabolites of tibolone through the oestrogen receptor in the mouse N20.1 oligodendrocyte cell line and normal human astrocytes.

17beta-oestradiol (E(2)) may have a beneficial impact on the development of age-related diseases, in part through alpha and beta oestrogen receptors (ER) in glia. Tibolone, a synthetic steroid, could influence glial-mediated neuroprotection if agonist oestrogenic activity is demonstrable. We used the N20.1 mouse oligodendrocyte cell line as a glial cell model to evaluate the response of ERalpha and ERbeta through oestrogen-response element (ERE) and AP-1-driven reporters to E(2), 4-hydroxytamoxifen (4OHT) and to two tibolone metabolites, 3alpha-hydroxytibolone (3alpha-OH-Tib) and 3beta-hydroxytibolone (3beta-OH-Tib). In addition, we tested the activity of these same ligands through the endogenous ERalpha in human normal astrocytes. Because endogenous ER was not detected in the N20.1 cells, we tested the ability of exogenous ER to activate transcription in response to ligands (100 nM) using a transient cotransfection assay with an ERalpha expression vector. To test the antagonist activity of 3alpha-OH-Tib and 3beta-OH-Tib, we used them in combination with E(2) (10(-8) M), at concentrations of 10(-7) M and 10(-6) M. The human normal astrocytes were treated similarly, with the exception that no ER-encoding DNA was used. Specific ER ligand mediated activity was shown using the E(2) antagonist ICI 182 780 and the pSG5 empty vector. E(2), 3alpha-OH-Tib, and 3beta-OH-Tib stimulated ERalpha on an ERE-promoter at each concentration (P < 0.001) but not at an AP-1-driven promoter. 4OHT was an effective antagonist, but did not exhibit agonist activity on the ERE-driven promoter. 4OHT was an effective agonist through ERalpha on an AP-1-driven promoter. 3alpha-OH-Tib and 3beta-OH-Tib were not effective antagonists of E(2). Both metabolites acted through the ER because the addition of an E(2) antagonist blocked their activity. These results show that 3alpha-OH-Tib and 3beta-OH-Tib exert agonist activity, yet lack antagonist or additive activity, through the ERalpha and ERbeta on an ERE-driven but not on an AP-1-driven promoter in a glial cell model and in normal human astrocytes. This contrasts with the effects of 4OHT, which exerted little or no agonist activity, but reduced E(2)-stimulated activity through ERalpha on the ERE, in the same cells.

Metabolism of cerebrosides and sulfatides in subcellular fractions of developing rat brain.

Previous studies on myelinating rat brain indicated that microsomes, Golgi-enriched and cytosol fractions may process galactolipids destined for myelin. To extend these findings we labeled brain galactolipids in vivo and determined the specific radioactivity of cerebrosides and sulfatides in several subcellular fractions. 17-day-old rats were treated by intracranial injection with [14C]galactose 60 min prior to and [3H]galactose 15 min prior to killing. Subcellular fractions were prepared from brain stem, and concentrations of cerebrosides and sulfatides were determined, their radioactivity measured and the 3H/14C ratio compared. Our results showed that the heavier Golgi-enriched fraction (designated Fraction 2) is unique in its low galactolipid content and high specific radioactivities of cerebrosides and sulfatides. The low ratio of the specific activity of cerebroside to that of sulfatide in Fraction 2 compared to other fractions indicates that it may be the site of most rapid conversion of newly synthesized cerebrosides to sulfatides. The specific radioactivities of cerebrosides and sulfatides in cytosol are intermediate between those in Golgi-enriched Fraction 2 and microsomes and those in myelin, consistent with the role postulated for cytoplasmic elements in the transport of cerebrosides and sulfatides to myelin.

Characterization of the anti-neural antibodies in the sera of leprosy patients.

Sera from 43 leprosy patients were tested for antibodies that bound to normal human nerve. Thirty-eight percent showed positive staining as demonstrated by indirect immunofluorescence. Only 1 out of 30 control sera tested displayed similar staining. Western blots of myelin and neural intermediate filament (IF) proteins were tested with patient sera. Two of the anti-neural antibody (ANeAb)-positive leprosy sera bound to the P0 protein of PNS myelin. All 17 ANeAb-positive leprosy sera displayed 2 or more bands in the molecular weight range of Mr 45 000-55 000, when tested against IF proteins. One explanation for these findings is that leprosy patients produce antibodies to intermediate filament (IF) proteins released subsequent to the bacterial invasion of the peripheral nerves. The importance of these autoantibodies in the pathogenesis of leprosy is discussed.

Production and characterization of high titer antibodies to galactocerebroside.

High titer antibodies primarily of the IgG class were produced against galactocerebroside (GalC) by including keyhole limpet hemocyanin (KLH) and supplemental M. tuberculosis in the adjuvant mixture used for immunization of rabbits. Antibody titers were determined by an ELISA in which microtiter wells were coated with liposomes containing lecithin, cholesterol and GalC. The antibodies showed reactivity with GalC and psychosine, but not glucocerebroside, sulfatide, mixed gangliosides or asialo GM1. Specificity was further demonstrated by absorption of antibodies with GalC. Binding was inhibited by galactose, but only at high concentrations. Further, the antibodies did not bind to any brain proteins on immunoblots, indicating lack of reactivity with glycoproteins which might contain a terminal galactose. Antibodies to GalC are directed against different determinants than those reacting with peanut agglutinin since the lectin will not react with GalC, and the antibodies will not react with asialo GM1. The antibodies raised to GalC by this method show specific staining for oligodendroglia in culture. Peanut agglutinin binds intensely to process-bearing GalC+ oligodendroglia, but very poorly to the membrane sheets elaborated by oligodendroglia after longer times in culture. Other process-bearing GalC-, GFAP- cells were also stained with peanut agglutinin; these cells may represent glial precursors.

Anti-neural antibodies in leprosy sera: further characterization of the antigens.

Sera or plasmas from 129 leprosy patients were tested by immunoblotting for antibodies that bound to proteins in a Triton-insoluble fraction enriched in neural intermediate filaments (IF fraction) from human or bovine spinal cord. Sixty samples (47%) showed positive staining of proteins at 35 kDa, 42 kDa or both. The presence of these antibodies appeared to be evenly distributed across the spectrum of disease. The frequency of these antibodies in samples from 12 healthy Ethiopians was similar to that in the leprosy group. Similar antibodies were found in only three of 28 samples from U.S. patients with neurologic diseases and in seven of 35 normal U.S. sera. Sera from U.S. tuberculosis patients stained multiple bands in the 50-30 kDa region of the blots; 11 of 16 stained bands corresponding to the 35 kDa or 42 kDa bands along with a number of other bands in this region. The 35 kDa and 42 kDa antigens do not appear to be breakdown products of neural filaments or glial fibrillary acidic protein, since antibodies to these proteins do not react with the 35 kDa or 42 kDa antigen. Further, the staining pattern with the leprosy sera is unchanged following Ca2+-mediated proteolysis of the IF-enriched fraction. The two antigens differ from one another in isoelectric point: the pI of the 35 kDa antigen is 5.9, and the pI of the 42 kDa antigen is 4.8. Staining of the immunoblots with antibodies against a number of known neural antigens failed to identify the 35 kDa and 42 kDa antigens. The 42 kDa antigen appears to be a component of axolemma, since 42 kDa-positive leprosy sera stained a protein with identical migration in preparations of bovine peripheral nervous system and human central nervous system axolemma. In some sera, antibodies reacting with the 35 kDa antigen were adsorbed by D-O bovine serum albumin, a synthetic analogue of the terminal disaccharide portion of the phenolic glycolipid 1 of Mycobacterium leprae. Antibodies to the 42 kDa antigen were not removed by this treatment.

Chronic Theiler's virus infection in mice: appearance of myelin basic protein in the cerebrospinal fluid and serum antibody directed against MBP.

Myelin basic protein (MBP) appears frequently in the cerebrospinal fluid (CSF) of mice with chronic demyelination following intracerebral infection with Theiler's murine encephalomyelitis virus (TMEV); antibody to MBP can frequently be found in the sera. The peaks of the immune responses to both MBP and TMEV coincide with the time course of the clinical signs of disease. Adsorption of mouse sera with TMEV or MBP indicate the non-identity of the antigens and the specificity of the antisera as measured by ELISA. Immunoblot analysis of sera confirmed the ELISA findings. The mechanism of induction of antibody directed against MBP and its role in TMEV-associated demyelination remain to be determined.

Opsonization of normal myelin by anti-myelin antibodies and normal serum.

Fc receptor-dependent myelin phagocytosis has been proposed as a possible important effector mechanism in several immune-mediated demyelinating diseases. The present study was designed to determine whether myelin is opsonizable by anti-myelin antibodies. Thioglycolate-elicited mouse peritoneal macrophages were cultured with 125I-labelled bovine central myelin pretreated with normal or immune serum. Serum opsonic activity was determined by a kinetic study comparing macrophage uptake of opsonized and untreated 125I-myelin. Heat-stable and heat-labile myelin opsonins were detected in normal rabbit serum. Myelin was also opsonized by normal rabbit gamma globulin and by heat-inactivated normal mouse, human, and guinea pig serum. Increased opsonic activity was detected in rabbit anti-myelin antiserum and the gamma globulin fraction prepared from this serum, in anti-myelin basic protein and anti-galactocerebroside antiserum but not in anti-myelin-associated glycoprotein antiserum or in serum from rabbits injected with Freund's adjuvant alone. One out of three anti-sheep red blood cell antisera tested also showed increased myelin opsonic activity. It is concluded that anti-myelin antibodies can promote opsonic phagocytosis, and that normal serum and normal serum gamma globulin also opsonize myelin.

DNA, ganglioside and sulfatide in brains of rats given corticosterone in infancy, with an estimate of cell loss during development.

Evidence of a postnatal loss of cerebral cells has been presented, based on counts of the activity of cerebral DNA following labeling with [3H]thymidine on embryonic day 14. In otherwise untreated rats, the loss was about 15 percent of the labeled cells between postnatal days 3 and 15. There was no further loss of activity thereafter. Littermates given corticosterone at day 3 showed little additional loss of labeled cells. Ganglioside NANA (N-acetylneuraminic acid) was reduced in proportion to the reduction in cerebral weight in the corticosterone treated rats. Sulfatide was reduced more, so that the concentration was 11 percent below that of the controls; It is concluded that the deficit in DNA after postnatal corticosterone treatment must be due primarily to an irreversible suppression of DNA synthesis, involving mainly glia. The reduction in gangliosides may represent a deficit in the growth of neuronal processes, leading to a reduction in the amount of neuropil, and perhaps contributing to a decrease in the number or size of myelinated axons.

Appearance of myelin proteins in rat sciatic nerve during development.

Rats between 5 and 45 days of age were sacrificed and their sciatic nerves dissected. Myelin was prepared from these sciatic nerves by a procedure involving purification on discontinuous sucrose gradients. The proteins of whole sciatic nerves at different ages and the proteins derived from the myelin isolated from these sciatic nerves were examined by discontinuous polyacrylamide gel electrophoresis in buffers containing sodium dodecyl sulfate. Over half of the proteins of sciatic nerve myelin migrated in a single band on the gel (P0). There were only minor changes in the protein distribution of sciatic nerve mylein during development. In contrast, the polyacrylamide gel patterns of whole sciatic nerve homogenate changed markedly during development between 5 and 15 days of age. The amount of P0 protein as a proportion of the total sciatic nerve protein increased from 3% at 5 days of age to 13% at 15 days of age after which it remained constant. Several other proteins which were also characteristic of the isolated myelin increased in relative importance during this time period. Parallel experiments dealing with a metabolic parameter of myelinogenesis, incorporation of intraperitoneally injected [35S]sulfate into sulfatide, were conducted. The maximum synthesis of sulfatide occurred between 6 and 16 days of age, coincident with the marked accumulation of myelin proteins in sciatic nerve.

Effects of ceramide analogs on myelinating organ cultures.

Analogs of ceramide which inhibit galactocerebrosidase also demyelinate or inhibit myelination in organ cultures of rat cerebellum. The potency of the analogs in culture correlated with their effectiveness as inhibitors of cerebrosidase, but not with their effectiveness as inhibitors of galactosyl transferase. The most effective compound was the decanoyl amide of 3-phenyl-2-amino-1,3-propanediol with erythroconformation. Stimulators of cerebrosidase also demyelinated cultures. With both groups of compounds, myelin sheaths became distorted, then broke into lipid droplets. Axons were preserved, but neurons showed some nuclear changes and granularity. Metabolic studies with the most effective inhibitor showed that glucose incorporation into cerebroside and other alkali-stable lipids was initially depressed compared to proteins and total lipids.

Myelin synthesis during postnatal nutritional deprivation and subsequent rehabilitation.

Newborn Long-Evans rats were undernourished by maternal deprivation so that by 20 days of age their body and brain weights were about 45 and 80%, respectively, of the values obtained for control (well-nourished) values. Proteins from myelin of undernourished and control rats were separated by polyacrylamide gel electrophoresis in buffers containing sodium dodecyl sulfate. At 15 and 20 days of age the proportion of basic and proteolipid protein was reduced in the starved animals relative to controls, indicative of a delay in maturation. However, by 30 days of age the composition of myelin from starved and control animals appeared similar. At all ages the yield of myelin from brains of starved rats was less than 25% of that obtained from control animals. A series of isotope labeling experiments, using a double label design, was carried out to compare relative rates of incorporation of radioactive amino acids into individual proteins of various brain subcellular fractions. In 20-day-old rats the incorporation of [3H] OR [14C] leucine or glycine into myelin proteins, relative to incorporation into proteins of other subcellular fractions, is preferentially depressed (about 60%) in starved animals. Synthesis of all the myelin proteins was depressed, supporting the hypothesis that the high molecular weight proteins isolated with myelin are true myelin constituents. Similar experiments were conducted using [3H]-and [14C] acetate, choline, or glycerol as precursors of lipids. Incorporation of isotope into lipids of myelin, relative to lipids of other subcellular fractions, was also depressed by about 60% in starved animals. In several experiments we studied synthesis during rehabilitation (ad libitum feeding) following 20 days of postnatal starvation. After 6 days of rehabilitation, incorporation of radioactive precursors into myelin, relative to other subcellular fractions, was still depressed. This result was true for both proteins and lipids, and was interpreted as evicence against the initiation of a process leading to a net recovery of myelin (i.e., an irreversible deficit of myelin synthesis is induced by this regime of nutritional deprivation).

Differential effects of Th1, monocyte/macrophage and Th2 cytokine mixtures on early gene expression for immune-related molecules by central nervous system mixed glial cell cultures.

Cytokines secreted within the central nervous system (CNS) are important in the development of multiple sclerosis (MS) lesions. The balance between Th1, monocyte/macrophage (M/M) and Th2 cytokines in the CNS may be pivotal in determining the outcome of lesion development. We examined the effects of mixtures of cytokines on gene expression by CNS glial cells, as mixtures of cytokines are present in MS lesions, which in turn contain mixtures of glial cells. In this initial analysis by gene array, we examined changes at 6 hours to identify early changes in gene expression that represent primary responses to the cytokines. Rat glial cells were incubated with mixtures of Th1, M/M and Th2 cytokines for 6 hours and examined for changes in early gene expression employing microarray gene chip technology. A minimum of 814 genes were differentially regulated by one or more of the cytokine mixtures in comparison to controls, including changes in expression in a large number of genes for immune system-related proteins. Expression of the proteins for these genes likely influences development and inhibition of MS lesions as well as protective and regenerative processes. Analysing gene expression for the effects of various combinations of exogenous cytokines on glial cells in the absence of the confounding effects of inflammatory cells themselves should increase our understanding of cytokine-induced pathways in the CNS.

Maintenance of membrane sheets by cultured oligodendrocytes requires continuous microtubule turnover and Golgi transport.

Oligodendrocytes in murine shakeoff cultures elaborate extensive membrane sheets containing networks of microtubules. Several membrane components, including proteolipid protein (PLP) and sulfatide, are transported through the Golgi en route to the plasma membrane or myelin (1,2). This transport is essential for membrane assembly, but its role in continuing maintenance of the sheets is not known. We examined the stability of the membrane sheets following microtubule stabilization with taxol or block of transport into the Golgi with brefeldin A. Within one to three hours, both agents had marked effects on the membrane sheets. While some oligodendrocytes maintained regions of normal membrane sheets, many showed retraction of the sheets, with the majority now exhibiting multiple processes rather than sheets. The distribution of sulfatide, PLP and tubulin in cell bodies, processes and sheets was altered in treated cells, as analyzed by immunocytochemical staining with antibodies to these components. The Golgi apparatus also showed reorganization in the presence of taxol, as visualized by binding of wheat germ agglutinin, a lectin with high affinity for distal Golgi vesicles. All of these effects were reversible when the agents were removed after 3 hours. Thus, maintenance of membrane sheets by oligodendrocytes in culture is a dynamic process, requiring ongoing microtubule turnover and transport of molecules through the Golgi.

Antibodies to glycolipids and cholera toxin B subunit do not initiate Ca++ signaling in rat Schwann cells.

Antibodies to glycolipids have been implicated in the pathogenesis of several immune-mediated PNS demyelinating diseases. This study focuses on antibodies to galactocerebroside (GalC) and sulfatide and on the B subunit of cholera toxin (CTB), which reacts with GM1 ganglioside, to examine whether these agents have any direct effects on Schwann cells (SC) as measured by Ca++ responses. While surface levels of GalC and sulfatide were markedly upregulated by 8 Br-cAMP treatment, as reported by others, very little expression of surface GM1 ganglioside was detected with or without 8 Br-cAMP treatment. Schwann cells, under either condition, showed no changes in intracellular Ca++ levels when exposed to purified monoclonal antibodies reacting with GalC or sulfatide. Thus upregulation of surface levels of GalC or sulfatide does not lead to antibody-induced Ca++ influx, in contrast to previous findings in mature oligodendrocytes (OLs) exposed to antibodies to GalC. Further, cross-linking with one of the antibodies (R-mAb) did not produce Ca++ responses. No Ca++ responses were elicited by CTB in Schwann cells either with or without 8 Br-cAMP treatment. Since surface binding of CTB was very low and sparsely punctate in Schwann cells +/- 8 Br-cAMP, we tested whether increasing levels of GM1 ganglioside on the surface would lead to induction of a Ca++ signaling pathway, as reported for fibroblasts. GM1 ganglioside on the surface of SC was markedly increased by exposing cells to exogenous GM1 ganglioside, but no Ca++ responses were observed in the treated cells. Thus undifferentiated or partially differentiated SC lack the glycoconjugate-mediated Ca++ signaling pathways found in mature OLs or fibroblasts.

Galactocerebroside and sulfatide independently mediate Ca2+ responses in oligodendrocytes.

Galactocerebroside (GalC) and sulfated galactocerebroside (sulfatide) are sphingolipids highly enriched in myelin. The binding of antibodies reactive with either sulfatide or GalC to cultured oligodendrocytes causes a Ca2+ influx, followed by microtubule depolymerization; however, antisulfatide is less effective than anti-GalC in altering cytoskeleton. Typical Ca2+ responses are delayed for both antibodies but are transient for sulfatide-reactive antibodies in contrast to the sustained responses previously reported for anti-GalC (Dyer and Benjamins, J Cell Biol 111: 625-633, 1990). Approximately one-half as many oligodendrocytes respond to sulfatide-reactive antibodies (about 39%) as to anti-GalC (about 75%). Subpopulations of oligodendrocytes were identified that responded to neither antibody, only one antibody, or both antibodies, indicating that sulfatide and GalC independently mediate Ca2+ responses. These results suggest that sulfatide and GalC have different physiologic roles in regulating elaboration of myelin membrane by oligodendrocytes in vivo and support the possibility that viral or immune attack via GalC or sulfatide on oligodendrocytes may mimic normal signals in a manner that disrupts the sequence of events that coordinates myelination or maintenance of myelin in vivo.

Release of intracellular calcium stores leads to retraction of membrane sheets and cell death in mature mouse oligodendrocytes.

The ability of mature oligodendrocytes (OLs) to recover from insult is important in repair of damage following demyelination. Since regulation of Ca2+ levels within cells plays a critical role in function and survival, this study investigates the effects of changes in cytoplasmic Ca2+ on the viability of cultured mouse OLs and their ability to maintain membrane sheets. Mature OLs in culture respond rapidly to the calcium ionophore A23187 and promptly return to resting Ca2+ levels when the ionophore is removed. Longer exposure to 0.1-1.0 microM A23187 leads to microtubule disruption, membrane sheet retraction and eventual cell death; nuclear lysis occurs in many of the OLs, as reported by Scolding, et al. (1) for rat OLs. In our cultures, mature OLs were more susceptible to nuclear lysis than were immature OLs or astroglia. Release of intracellular Ca2+ stores with thapsigargin at 5-10 microM also leads to retraction of membrane sheets. Following 6 hours of continuous exposure to thapsigargin, the effects on membrane sheets are reversed over the next 12 hours. After 18 hours of continuous exposure to thapsigargin, only occasional nuclear lysis is observed, but a number of the mature OLs show signs of DNA fragmentation, indicating that apoptotic death is occurring. Our results suggest that mature OLs cannot survive a prolonged influx of extracellular calcium as readily as immature OLs and astroglia, but have mechanism to withstand similar increases in cytoplasmic Ca2+ following sustained release of intracellular stores.

Kinetics of entry P0 protein into peripheral nerve myelin.

Sciatic nerves from 9-day-old rat pups were removed, sliced into 0.4-mm sections, and incubated with [3H]fucose or [14C]glycine precursors. The nerve slices system gave nearly linear incorporation of [3H]fucose as a function of time for 3 h, after an initial lag of approximately 30 min for homogenate and approximately 60 min for myelin. Incorporation of [3H]fucose at constant specific radioactivity was directly proportional to exogenous fucose levels over the range 3.0 x 10(-8) M to 1.5 x 10(-6) M. Analysis of labeled proteins by sodium dodecyl sulfate polyacrylamide gel electrophoresis showed that greater than 50% of labeled glycoprotein was P0, with no other major constituents. This system was used in fucose-chase experiments to determine that a period of approximately 20 min elapses between fucosylation and assembly of P0 into myelin. Cycloheximide inhibition of protein synthesis was used to determine that a period of approximately 33 min elapses between protein synthesis and appearance of P0 myelin.