Sonic hedgehog-dependent emergence of oligodendrocytes in the telencephalon: evidence for a source of oligodendrocytes in the olfactory bulb that is independent of PDGFRalpha signaling.

Most studies on the origin of oligodendrocyte lineage have been performed in the spinal cord. By contrast, molecular mechanisms that regulate the appearance of the oligodendroglial lineage in the brain have not yet attracted much attention. We provide evidence for three distinct sources of oligodendrocytes in the mouse telencephalon. In addition to two subpallial ventricular foci, the anterior entopeduncular area and the medial ganglionic eminence, the rostral telencephalon also gives rise to oligodendrocytes. We show that oligodendrocytes in the olfactory bulb are generated within the rostral pallium from ventricular progenitors characterized by the expression of PLP: We provide evidence that these Plp oligodendrocyte progenitors do not depend on signal transduction mediated by platelet-derived growth factor receptors (PDGFRs), and therefore propose that they belong to a different lineage than the PDGFRalpha-expressing progenitors. Moreover, induction of oligodendrocytes in the telencephalon is dependent on sonic hedgehog signaling, as in the spinal cord. In all these telencephalic ventricular territories, oligodendrocyte progenitors were detected at about the same developmental stage as in the spinal cord. However, both in vivo and in vitro, the differentiation into O4-positive pre-oligodendrocytes was postponed by 4-5 days in the telencephalon in comparison with the spinal cord. This delay between determination and differentiation appears to be intrinsic to telencephalic oligodendrocytes, as it was not shortened by diffusible or cell-cell contact factors present in the spinal cord.

The early steps of oligodendrogenesis: insights from the study of the plp lineage in the brain of chicks and rodents.

Oligodendrocytes are the myelin-forming cells of the central nervous system. Over the last decade, their development in the embryonic brain and spinal cord has been documented following the discovery of early oligodendroglial markers. These early expressed oligodendroglial genes nevertheless show differences in their spatiotemporal pattern of expression and it is not yet clear if their expression is linked in a linear way. This review highlights the common themes underlying the spatiotemporal aspects of oligodendrogenesis in chick and rodent brain and discusses some recent advances in the knowledge of the cell lineage expressing plp, one of the early oligodendroglial genes. We suggest a model of oligodendroglial commitment whereby definitive oligodendroglial progenitor formation is preceded by a primitive neuroglial progenitor stage and whereby different oligodendrocyte lineages might segregate from either plp-positive or plp-negative primitive progenitor cells.

Spatiotemporal development of oligodendrocytes in the embryonic brain.

In the central nervous system (CNS), oligodendrocytes have long been considered to be the last cell type to be generated during development. In rodents, the progenitor cells that give rise to oligodendrocytes have been reported to originate in the subventricular zone. Here, we review recent data demonstrating the existence of oligodendrocyte precursor cells in the ventricular layer of the neural tube that emerge prior to the progenitor stage. Oligodendrocyte precursors arise in restricted foci that are distributed along the rostrocaudal axis of the neural tube, for the most part ventrally. The generation of oligodendrocyte precursor cells occurs either simultaneously with, or follows closely upon the emergence of the first neurons. Experiments with quail-chick chimeras provide evidence that oligodendrocyte progenitors derived from ventricular precursors migrate either tangentially or radially to colonize extensive or segmentally restricted territories of the brain. The choice depends on their site of origin. Finally, we discuss the possibility that oligodendrocytes could be a mosaic population that originates from at least two types of precursor cells.

Single or multiple oligodendroglial lineages: a controversy.

The text books all say oligodendrocytes are the last cell to arise during development. The analysis of the spatio-temporal pattern of expression of plp/dm-20 during embryonic development in both the chick and the mouse provides evidence that the induction of oligodendrocyte occurs much earlier than we thought. In fact, it seems as though these cells must arise nearly simultaneously with neurons and it is just that they do not mature until later. Furthermore, we review the experimental arguments in favor of the existence of at least two, if not more, oligodendrocyte precursor cells: one is defined by the expression of PDGFRalpha, another characterized by expression of plp/dm-20 is independent from PDGF-AA for its proliferation and survival. We then postulate the existence of a third family of yet unknown origin.

Multiple restricted origin of oligodendrocytes.

The plp gene encodes the proteolipid protein and its alternatively spliced product DM-20, major proteins of CNS myelin. In the mouse, plp/dm-20 transcripts are expressed beginning at embryonic day 9.5 (E9.5) by restricted foci of germinative neuroepithelial cells. To determine the identity of the neural precursors expressing plp/dm- 20, a zeomycin resistance gene fused to the lacZ reporter was expressed in transgenic mice under the control of the plp regulatory sequences. In the three different lines generated, the pattern of beta-galactosidase expression was similar and superimposable on the expression pattern of endogenous plp/dm-20. Both in vivo and in vitro, the transgene was expressed by O4(+) pre-oligodendrocytes, and later by RIP+ differentiated oligodendrocytes, but not by neuronal cells, astrocytes, or radial glial cells. After zeomycin selection, a dramatic enrichment in O4(+) pre-oligodendrocytes was observed in cultures derived from E12.5 transgenic embryos. This enrichment indicates the oligodendroglial specification of neural precursors that continuously express plp/dm-20. Early plp/dm-20-expressing precursors, however, appear to be a separate population from previously described PDGFRalpha oligodendrocyte precursors, as shown by the striking differences in their (1) patterns of distribution and (2) responsiveness to PDGF. These data suggest that oligodendrocytes have a plural origin and that early plp/dm-20 defines one of the neural lineages generating oligodendrocytes.

Transcription of myelin basic protein promoted by regulatory elements in the proximal 5' sequence requires myelinogenesis.

Myelination in the central nervous system requires synthesis by oligodendrocytes of enormous amounts of lipids and proteins for incorporation in the developing myelin membranes. To approach the regulatory events coordinating the transcriptional activation of the genes that encode myelin proteins, we examined control of the myelin basic protein (MBP) locus. MBP plays a major role in myelin compaction. During development, MBP is already expressed in mature non-myelinating oligodendrocytes. Here we show that, in transgenic animals in which the E. coli lacZ reporter gene is under the control of increasingly large portions (256, 1900 and 3200 bp) of the MBP promoter, 5' of the initiation of transcription site, reporter gene expression was initiated after myelin formation had started. This delayed expression of the transgene compared to MBP, strongly suggests that premyelinating expression is dependent on regulatory elements located outside of the 3200 bp sequence studied, while expression occurring at the time of myelin formation is dependent on the proximal promoter sequence.

The proximal region of the MBP gene promoter is sufficient to induce oligodendroglial-specific expression in transgenic mice.

To characterize regulatory DNA sequences involved in oligodendroglial expression of myelin basic protein (MBP), transgenic mice carrying a 256 bp fragment of the mouse MBP promoter fused to an Escherichia coli lacZ gene were generated. Of four transgenic families, two (lines 2 and 4) expressed beta-galactosidase activity in the nervous system but not in most other tissues. Histochemical and immunohistochemical analysis of adult brain from these two lines showed oligodendroglial-specific expression of the transgene. In line 2, only a small proportion of oligodendrocytes expressed the transgene, and in labelled cells the product of the enzymatic reaction with beta-galactosidase was confined to a small round vesicle in the vicinity of the nucleus. In contrast, in tissue sections from line 4 adult brain and spinal cord beta-galactosidase activity was much more intense and at least 80-90% of oligodendrocytes expressed the transgene. Detection of the MBP-lacZ transcript by in situ hybridization showed that the transgene mRNA was confined to the oligodendrocyte cell body. These results suggest that cis-acting regulatory elements, specifying oligodendrocytes identity, are located within 256 bp upstream from the MBP gene.

Clonal segregation of oligodendrocytes and astrocytes during in vitro differentiation of glial progenitor cells.

To study the clonal lineage of the glial progenitor population, isolated from newborn rat brain (Lubetzki et al. J Neurochem 56:671, 1991), we combined somatic transgenesis using a retroviral vector encoding a modified bacterial beta-galactosidase with nuclear localization, and triple immunofluorescence labeling with A2B5, anti-galactosylceramide, and anti-glial acidic fibrillary protein antibodies. This allowed clonal analysis of the postnatal glial lineage with precise phenotypic identification of each cell within the lacZ-positive clones. When infected cells were cultivated under constant conditions, in the presence of either 1% or 10% fetal calf serum (FCS)-containing medium, all the 250 lacZ-positive clusters examined were homogeneous, i.e., either oligodendroglial or astroglial. Mixed astrocyte-oligodendroglial clones were observed when cells cultivated in the presence of 1% FCS were switched to a 10% FCS-containing medium, confirming the bipotentiality of glial progenitor cells (Temple and Raff Nature 313:223, 1985). However, even under the switch culture conditions, segregation into homogeneous clones of either oligodendrocytes or astrocytes still predominated, and the percentage of mixed clones dropped from 25 to 8 or to 3, when the switch took place at 8, 16, or 22 days in vitro, respectively. Two additional observations lead us to suggest that microenvironmental factors are responsible for the clonal segregation of glial progenitor cells: 1) the uneven distribution of oligodendrocyte and astrocyte clusters, the latter being seen mostly on the edge of the coverslips; and 2) the presence, in the vicinity of an homogeneous lacZ-positive clone, of some lacZ-negative cells expressing the same phenotype.

Morphological, biochemical, and functional characterization of bulk isolated glial progenitor cells.

We describe a simple, rapid, and efficient method, based on separation on a Percoll centrifugation gradient, to purify glial progenitor cells from newborn rat brains. Cytofluorimetry analysis of the isolated cell population showed that 75 +/- 8 and 86 +/- 7% of the cells were A2B5- and R24-positive, respectively. Transmission electron microscopy examination of the purified cell population confirmed their homogeneity and illustrated their typical morphology, as previously described in situ. Assay of UDP-galactose-ceramide galactosyltransferase, 3'-phosphoadenosine 5'-phosphosulfate galactosylceramide sulfotransferase, and 2',3'-cyclic nucleotide 3'-phosphohydrolase activities showed that the levels of these enzymes were 446, 76, and 11 times lower, respectively, than the levels measured in mature oligodendrocytes. Low levels of mRNA coding for 2',3'-cyclic nucleotide 3'-phosphohydrolase and myelin proteolipid protein, but not for myelin basic protein, were present in the glial progenitor cells. At the time of isolation, 40% of the cells in the population were dividing, and the cells could easily be expanded in culture. After 3 weeks of culture in the presence of 1% fetal calf serum, 75% of the cells had differentiated into galactosylceramide-positive oligodendrocytes. When the culture took place in the presence of 10% fetal calf serum, only 2% of the cells expressed galactosylceramide, and 60% were glial fibrillary acidic protein-positive astrocytes; half of them were also A2B5 positive.

Developmental expression of myelin proteolipid, basic protein, and 2',3'-cyclic nucleotide 3'-phosphodiesterase transcripts in different rat brain regions.

RNA was extracted from five different rat brain regions during development, starting from embryonic day 15 (E15) until postnatal day 60 (P60). These RNA preparations were analyzed by both Northern and dot blot for their content of 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNPase), myelin proteolipid protein (PLP), and myelin basic protein (MBP) -specific transcripts. CNPase mRNA was readily detectable at E15 and PLP mRNA at P1 in all brain regions examined. In contrast, expression of MBP mRNA followed a caudorostral gradient. It was first observed at P1 in the mesencephalon and at P9-P11 in the olfactory bulb. Expression of these three transcripts displayed two types of developmental profiles. One was termed biphasic because the specific mRNA level increased regularly and then reached a plateau level. The other developmental profile was termed triphasic, because there was a gradual increase in the level of specific transcripts with a sudden appearance of a sharp peak followed by a decline to a plateau level. When the triphasic pattern was observed, the date of the peak appearance was probe-, but not region-, dependent. It was P15 for CNPase, P18 for MBP, and P21 for PLP. As these peaks occurred at a time during development when myelination was the most active, we postulate the existence of a transient external signal, perhaps neuronal, which would be responsible for this increased amount of myelin-related transcripts.

Immunological determination of galactosylceramide level in blood as a serum index of active demyelination.

An enzyme-linked immunosorbent assay (ELISA) to determine the level of galactosylceramide (GalC) in biological fluids is described. The assay uses GalC-coated plastic microtiter plates, with binding of an antibody to GalC detected by a peroxidase-labeled second antibody. The GalC level was directly estimated in the biological samples, without prior extraction, by competition with the coated hapten. This method allows the detection of 62 pmol of GalC (1.2 nmol/ml). Results using this procedure revealed positive sera only among patients suffering a myelin-destructive process: either primary, as in multiple sclerosis, or secondary to brain damage, as during ischemic strokes.

Production and characterization of a mouse monoclonal antibody to the glycolipid asialo-GM1.

The glycosphingolipid asialo-GM1 (aGM1) is a true differentiation antigen of murine lymphoid cells. This glycolipid is highly immunogenic in the rabbit, but the antisera produced shows some cross reactivity with GM1, the naturally occurring sialylated derivative of aGM1. In the present study we examined the ability to raise anti-aGM1 antisera in the mouse. We compared the efficiency of several immunization methods in various strains of mice. The most effective procedure involved repeated intraperitoneal injections of aGM1-cholesterol rich particles in the NZB mouse. Hybrid B cell lines were generated by fusion of mouse myeloma cells with the splenocytes of an NZB mouse immunized with aGM1. The specificity of the antisera produced and of the monoclonal antibody secreted by one of these hybridomas (103HT30) was defined by ELISA and by immunostaining on thin layer chromatograms. The monoclonal antibody 103HT30 is an IgM. It reacted with aGM1 but not with any of the structurally-related ganglioside or neutral glycolipids tested. In particular, 103HT30 monoclonal antibody did not present any detectable cross-reactivity with GM1.

Marmoset red blood cell receptor for membrane-associated complement components is not related to human CR1: partial characterization of the C3-binding proteins responsible for the spontaneous rosette formation between marmoset red blood cells and human leukocytes.

Cells from all the human B-lymphoblastoid cell lines tested and most human monocytes form rosettes with marmoset red blood cells (MaRBC). Because previous reports suggested the involvement of complement components in this phenomenon, the mechanism of rosette formation and the eventual similarities between the MaRBC receptor and the CR1 receptor present on human erythrocytes have been analyzed herein. The binding of MaRBC to human leukocytes strongly differs from the immune adherence phenomenon: rabbit anti-human CR1 did not react with MaRBC and the MaRBC receptor-binding activity is Ca2+-dependent. Rosette formation required intact energy metabolism and cytoskeleton integrity of leukocytes. Our attempts to purify the receptor from MaRBC membranes revealed the absence of CR1. Nevertheless, C3-binding proteins were isolated by selective desorption by Sepharose iC3 column chromatography. A three-band pattern was observed under reduced conditions with 74,000, 70,000, and 53,000 molecular weights. It was not possible to further separate these components. This protein complex inhibited the rosette phenomenon between MaRBC and both Raji and U-937 cells, exhibited a very poor cofactor activity, and had no decay-accelerating activity toward the classical C3 convertase. This material did not cross-react with antibodies directed to human proteins. These results showed that erythrocytes from new world monkeys do not express a receptor analogous to the human CR1, but expressed C3-binding protein with low cofactor activity that could recognize membrane-associated complement components.

Cell-free synthesis of opiate binding sites.

A procedure is described for the detection of opiate binding sites synthesized during in vitro translation of various mRNA preparations. RNA were isolated from membrane bound polysomes which were prepared from NG 108-15 hybridoma, C6BU1 glioma cells, as well as from N18TG2, NB2aAg and NB41A3 neuroblastoma cells. Polyadenylated [poly(A)(+)] RNA were purified, translated in vitro in a rabbit reticulocyte lysate and the translation products assayed for their ability to bind [(3)H] bremazocine. Bound and free ligands were separated by column chromatography. After translation of poly(A)(+) RNA obtained from NG 108-15 cells we demonstrated a stereospecific, saturable binding of [(3)H]bremazocine (displaced by levorphanol and not by dextrorphan) with a K(d) of 2.4 +/- 1.0 nM. The total amount of opiate binding sites synthesized was 6.2 +/- 0.5 fmol per ?g of poly(A)(+) RNA. Opiate binding sites were undetectable at zero time and a plateau was reached after translation had proceeded for 20 min. Five time less opiate binding sites were synthesized when the poly(A)(+) RNA purified from N18TG2 neuroblastoma cells were used under the same experimental conditions. There was no detectable binding of opiate ligands with poly(A)(+) RNA obtained from C6BU1 glioma cells, NB2aAg or NB41A3 neuroblastoma cells.

Schwann cell marker defined by a monoclonal antibody (224-58) with species cross-reactivity. I. Cellular localization.

We have demonstrated by indirect immunofluorescence the cellular localization of a monoclonal antibody (mAb 224-58), produced after immunization of a mouse with human central nervous system (CNS) myelin. Serologically, mAb 224-58 was found to be specific for 3'-sulfomonogalactosylglycolipids, namely 3'-sulfogalactosylceramide (SGC) and 3'-sulfogalactosyl 1-O-alkyl ether 2-O-acylglycerol (seminolipid). This mAb did not bind to SGC-containing tissues such as kidney, liver, spleen, or brain, nor to muscle. However mAb 224-58 did stain positively mouse, rat, and human peripheral nerve sections. In these latter sections, mAb 224-58 was bound to Schwann cell bodies and processes. The specificity of mAb 224-58 for Schwann cells was ascertained on teased rat sciatic nerves and rat Schwann cell cultures. Cells positive for mAb 224-58 were also positive for laminin, and negative for Thy 1-1 antigens both in teased fibers and Schwann cell cultures. In addition, in teased nerve preparations, mAb 224-58-positive cells were also galactosylceramide (GalC)- and SGC-positive. Isolated Schwann cells also expressed 224-58 antigen, even after prolonged time in culture. On testis sections, which contain both SGC and seminolipid, the SGC-positive cells, i.e., the spermatogonia, were always 224-58-negative. But the other germinal cells were 224-58-positive. This suggests that although 224-58 does not discriminate between SGC and seminolipid in serological tests, these lipids in their naturally occurring membrane acquire a spatial configuration that renders them distinguishable to their respective antibody.

Schwann cell marker defined by a monoclonal antibody (224-58) with species cross-reactivity. II. Molecular characterization of the epitope.

A monoclonal antibody (mAb) designated 224-58 (IgM-kappa) has been raised by fusion of Sp2/0-Ag14 mouse hybridoma cell line with spleen cells of a mouse immunized with human brain myelin. This mAb binds specifically to mouse, rat, and human Schwann cells membrane. Serological tests showed that mAb 224-58 reacted with total lipid extract, but not with total protein extract of human myelin. Combination of ELISA, complement fixation, and immunoautoradiographic detection on silica gel TLC allowed us to determine that mAb 224-58 reacted with sulfomonogalactolipids, namely 3'-sulfogalactosylceramide (SGC) and 3'-sulfogalactosyl 1-O-alkyl-ether 2-O-acylglycerol (seminolipid). The fine molecular structure of the epitope recognized by mAb 224-58 was established by studying the cross-reactivity of this mAb toward closely related sulfolipids and by comparing its reactivity after submitting either purified sulfolipids or total lipid extracts to various chemical and enzymatic treatments. The lipidic hapten-binding site to mAb 224-58 is dependent on (1) the sulfoester on carbon 3 of the galactose molecule, (2) the osidic bond, and (3) the carbonyl group of the fatty acid. Interestingly enough, neither the amide bond and the long-chain base nor the OH group of the fatty acid belongs to the antigenic determinant recognized by mAb 224-58.

Analysis of DR-like molecules on a marmoset Epstein-Barr virus-induced cell line using a monomorphic anti-human HLA-DR monoclonal antibody.

Mouse anti-HLA D region-related (DR) monoclonal antibodies have been found to cross-react with peripheral blood leukocytes from one primate species, the common marmoset (Callithrix jacchus). Immunoprecipitates of radioactively labeled cells extracted from a marmoset Epstein-Barr virus-induced cell line were analyzed by one- and two-dimensional gel electrophoresis and compared with the DR antigens of a human lymphoblastoid B cell line. Two chains of estimated molecular weights of 34 000 (alpha) and 28 000 (beta), similar to the human alpha and beta chains, have been observed in marmoset immunoprecipitates. Additionally, a set of spots located in the same area as the set of invariant spots found in human HLA-DR antigens is shown by two-dimensional gel electrophoresis. Thus, cross-reacting anti-human HLA-DR monoclonal antibodies could be used to analyze the expression and the structure of marmoset DR-like antigens.