Analysis of the human MBP promoter in primary cultures of oligodendrocytes: positive and negative cis-acting elements in the proximal MBP promoter mediate oligodendrocyte-specific expression of MBP.

Since the regulation of myelin basic protein expression depends primarily on the initiation of transcription, we analyzed the 5' flanking region of the human myelin basic protein gene in transient transfection studies in primary cultures of developing oligodendrocytes. We demonstrated that 149 base pairs 5' of the initiation of transcription was sufficient to direct oligodendrocyte-specific expression of myelin basic protein. The capsite of the fusion transcript was identical with that of the endogenous myelin basic protein transcript, and chloramphenicol acetyl transferase reporter gene expression was restricted to oligodendrocytes in these cultures. Within this 149 base pair region, one distal, negative cis-acting segment, containing a consensus nuclear factor I site, and one proximal, positive cis-acting segment were identified. The distal segment behaved more negatively in Cos-7 cells than in oligodendrocytes, reducing expression to background levels. Furthermore, these functionally important cis-acting segments bound oligodendrocyte nuclear proteins in a pattern differing from other cells, including Cos-7 cells. Interestingly, the distal segment increased heterologous SV40 promoter activity in oligodendrocytes but had no effect on the SV40 promoter in Cos-7 cells. We conclude that the functionally negative distal segment may mediate oligodendrocyte-specific expression of MBP by restricting its expression in other cells. These experiments strongly support using primary cultures of oligodendrocytes for analyzing the myelin-specific promoters.

Integrin on developing and adult skeletal muscle.

Avian integrin is a complex of integral membrane glycoproteins that appears to function as a dual receptors for both intracellular cytoskeletal and extracellular matrix components. Antibodies were raised against this complex and used to (1) immunolocalize integrin on cryosections of developing and adult muscle tissue and on developing myotube cultures in vitro and (2) immunoaffinity purify integrin from various fiber-type specific muscles. Integrin localization was compared with that of its putative cytoskeletal-associated and extracellular matrix ligands, talin and vinculin and fibronectin and laminin, respectively. The goal was to identify putative sites of interaction between the muscle sarcolemma and the cytoskeleton and the extracellular matrix and to reveal any differences in the molecular composition at these sites. Integrin's distribution on the sarcolemma of early (Day 12) embryonic limb muscle was random and punctate. On late embryonic (Days 17-19) limb muscle tissue its distribution was generally uniform but with occasional increased densities at specific sites along the sarcolemma. Posthatch (greater than 3 weeks) fast twitch muscle showed a highly regionalized distribution. These regions of integrin concentration coincided with densities of acetylcholine receptors, revealed by TRITC alpha-bungarotoxin labeling, and regions of muscle-tendon interaction, identified by morphological criteria. Tissue culture studies also demonstrated integrin densities at analogous sites in vitro, e.g., acetylcholine receptor clusters and sites at which myofibrils terminate at the sarcolemma. These integrin-rich sites were also shown to be Triton X-100 insoluble and therefore presumably are linked to the cytoskeleton or extracellular matrix. The localization of integrin on developing and adult muscle tissue was compared with that of fibronectin, laminin, vinculin, and talin using double, immunofluorescently labeled cryosections. In general, integrin did not colocalize exclusively with any one of its putative ligands. In the embryo, discrete densities of both talin and vinculin were observed at the myotendinous junction, whereas integrin immunoreactivity was widely distributed on muscle, vasculature, nerve, and connective tissue with no discernible sites of increased density. Laminin was primarily associated with muscle and nerve whereas fibronectin was prominent on connective tissue. On posthatch tissue, the distributions of talin, vinculin, laminin, and fibronectin were similar to those in the embryo, whereas the distribution of integrin was restricted to specific sites. The distribution of integrin was also examined for fiber-type specific differences on adu

The participation of a putative cell surface receptor for laminin and fibronectin in peripheral neurite extension.

We have used the CSAT (cell substrate attachment) monoclonal antibody (Mab), which is directed against a putative laminin and fibronectin receptor, to examine its role in the adhesive phenomena of neurons. This antibody was previously found to disturb the adhesion of several classes of fibroblasts and muscle. Here we report its effects upon neuronal-substrate adhesion. Two sources of neurons were investigated--the dorsal root and ciliary ganglia. Both responded similarly. Neurons plated in the presence of the CSAT Mab did not adhere to the substratum and process formation was inhibited completely for at least 24-48 hr. In explant cultures, when neurons were first allowed to extend processes prior to addition of the CSAT Mab, the results depended on the particular substrate. With some substrates, the neurites bundled and detached from the substratum; with others, they retracted and regrew to form large fascicles or bundles of processes. In dissociated cultures that already had extended processes, neurites fasciculated and cell bodies aggregated in response to the presence of the CSAT Mab. The magnitude of this response varied, depending upon the substrate. The antigen was localized, using immunofluorescence, on neuronal cell bodies, axons, and growth cones. This distribution correlated with its biological effects on all parts of the neuron. The antigen was isolated from neuronal cultures by immunoaffinity purification. It migrated in the molecular weight range of 140 kDa on reducing SDS-PAGE. This antigen is very similar to that isolated from fibroblasts, which is an integral membrane glycoprotein complex. The data presented implicate the participation of the CSAT antigen in neurite extension and fasciculation.