Isolation and characterization of glial filaments from human brain.

Intermediate (8--9 nm) filaments of human central nervous system astrocytes were isolated from the gliosed white matter of cases of adrenoleukodystrophy (ALD). This hereditary lipidosis is characterized pathologically by demyelination, loss of axons, and replacement of the white matter of the caudal cerebrum by a glial scar. Glial filaments were composed largely of a single protein component with a mol wt of about 49,000 daltons. Smaller components (44,000--39,000 daltons) were detected in some samples, and appear to represent degradation products of the filament protein. Human neurofilaments were isolated from the normal frontal white matter of ALD cases by the standard myelin-free axon technique. Isolated glial and neurofilament proteins comigrated during acrylamide gel electrophoresis in SDS. Polypeptides resulting from cyanogen bromide cleavage of the two filament proteins were the same. Both proteins reacted with rabbit antisera raised against isolated bovine neurofilament protein and human glial fibrillary acidic protein.

Neuronal soma and whole neuroglia of rat brain: a new isolation technique.

Minced rat brain softened by treatment with trypsin is disrupted by filtration through nylon and steel meshes to produce a suspension of free-floating cells and debris. The cells are separated and purified by centrifugation on discontinuous sucrose gradients. Preparations of neuronal perikarya, retaining stumps of processes, so obtained are 90 percent pure and yield 33.6 x 10(6) cells per brain (3 milligrams, dry weight). The glial cells, apparently intact with extensive branched processess, are about 70 percent pure by weight and are obtained in a yield of 6.6 x 10(6) cells per brain (2 milligrams dry weight). The neurons are smaller and have less lipid than the glial cells.

Neuronal soma and whole neuroglia of rat brain: a new isolation technique.

In the report "Neuronal Soma and Whole Neuroglia of Rat Brain: A New Isolation Technique" by W. T. Norton and S. E. Podulso [167, 1144 (1970)], sentence 2, paragraph 3, column 1, p. 1144, should read: "The brains are trimmed of cerebellum and chopped fine (approximately 1 mm(3)) in an ice-cold medium consisting of 5 percent glucose, 5 percent fructose, and 1 percent bovine serum albumin (14) in 10 mM KH(2)PO(4)-NaOH buffer (pH 6.0)."

Axons: isolation from mammalian central nervous system.

Centrifugation of a homogenate of white matter, in a solution of buffered sucrose containing salt, produces a floating layer of myelinated axons. When these are suspended in hypotonic buffer, the mnyelin swells and strips away from the axon. Axons are then separated from the myelin by centrifugation. The resulting preparation consists of a variable population of processes with lengths up to 200 micrometers and diameters between 0.3 and 5.0 micrometers. The axons contain neurofilaments and mitochondria, although no axolemma or neurotubules are evident. The preparation contains cerebroside and sulfatide, yet is essentially free of myelin.

Isolation of filaments from brain.

A method is presented for the isolation of filaments of 90-angstrom diameter from the white matter of bovine brain by first floating the myelinated axons in a centrifugal field and then fractionating the axons on a series of density gradients. This results in a fraction that contains two types of bundles of filaments but few other constituents. The filaments are stable over a wide range of temperatures and at both low and high ionic strength. Their density and their resistance to digestion by ribonuclease and deoxyribonuclease indicate that they are primarily protein. The molecular weight of the subunit is approximately 60,000. The protein does not comigrate with microtubule protein and does not bind cholcicine or nucleotides.

Peroxisomal and mitochondrial defects in the cerebro-hepato-renal syndrome.

The cerebro-hepato-renal syndrome is a rare familial malady with cerebral, renal, and skeletal abnormalities, severe hypotonia, cirrhosis, iron and lipid storage, and death within 6 months. Correlated electron microscopic, histochemical, and biochemical studies demonstrate defects in two oxidative organelles. Peroxisomes cannot be found in hepatocytes and renal proximal tubules. In hepatocytes and cortical astrocytes, mitochondria are distorted in their appearance and glycogen stores are increased. Oxygen consumnption of brain and liver mitochondrial preparations with succinate and with substrates reducing nicotinamide adenine dinucleotide is markedly diminished, but the consumption is normal with ascorbate and tetramethylphenylenediamine, which suggests a defect in electron transport prior to the cytochromes. Histochemical studies of mitochondrial oxidation point to a defect between the succinate dehydrogenase flavoprotein and coenzyme Q, possibly in the region of nonheme iron protein.

Multiple sclerosis: altered expression of 70- and 27-kDa heat shock proteins in lesions and myelin.

Recent studies have implicated heat shock proteins (HSP) in the pathogenesis of the multiple sclerosis (MS) lesion. Expression of the 73 kDa constitutive HSP (HSC70), the 72 kDa stress-inducible HSP (HSP70), and the 27 kDa small HSP (HSP27) was analyzed in white matter and myelin from central nervous system (CNS) tissue of MS and normal subjects using a combination of immunocytochemistry and quantitative immunoblotting. Plaques of all types were sharply defined by reduced immunostaining for HSC70, and shown by immunoblotting to contain 30 to 50% less HSC70 than surrounding white matter or normal tissue. In contrast, HSP27 was markedly enhanced 2.5- to 4-fold in plaque regions, especially in fibrous astrocytes and in hyperplastic interfascicular oligodendrocytes at the lesion edge. HSP70 was less abundant than HSC70, and no significant differences in HSP70 levels were noted between MS and normal white matter. Myelin isolated from active plaques contained 3- to 4-fold more HSC70 than normal myelin. Pronounced expression of HSP70 and HSP27 was also found in MS myelin, although neither protein was detected in normal myelin. Thus, white matter undergoing immune-mediated destruction in MS was associated with altered distribution and expression of HSC70 and HSP27. These changes may initially serve to protect myelin from further destruction and facilitate repair; however, enhanced expression of HSC70, HSP70, and HSP27 in myelin may subsequently present as additional immune targets involved in the progression of disease.

Biochemistry of demyelination.

The myelin sheath, a lipid-rich multilamellar membrane of relative stability, both insulates and enhances conduction in nerve axons. A notable feature of myelin-specific proteins, in particular myelin basic protein, is their susceptibility to proteolytic activity and their encephalitogenicity, which induces inflammatory demyelination in the CNS. The final common pathway of myelin breakdown in vivo is well documented and there is evidence that myelin disruption can be mediated directly by soluble (circulating) factors and for following receptor-driven phagocytosis by macrophages. However the exact mechanism(s) of demyelination in multiple sclerosis is still unresolved, both antigen-specific and--non-specific events having the potential to generate the myelinolytic process.

Mucolipidosis IV. Clinical, ultrastructural, histochemical, and chemical studies of a case, including a brain biopsy.

A 7-year-old Ashkenazi Jewish boy with normal early development started to regress at 8 months of age and made no further developmental progress. Corneal clouding was noted at age 10 months. Corneal and conjunctival biopsy at 14 months, cerebral biopsy at 24 months, and fibroblast cultures at 32 months showed lysosomal inclusions, suggesting the storage of lipid-like and mucopolysaccharide-like material. In the brain, dense fluorescent inclusions resembled those in ceroid-lipofuscinosis. Total ganglioside content of white matter was raised, but the pattern was normal. The level of nonlipid hexosamine in the brain was normal. The cornea and conjunctiva contained electronlucent vacuoles resembling those in the mucopolysaccharidoses. Cornea, brain, and lymphocytes contained concentric membranous lamellar structures reminiscent of those in the gangliosidoses. The clinical picture and ultrastructural findings support the impression that this case belongs to a new variant of the mucolipidoses, mucolipidosis IV.

Huntington disease: normal lipid composition of purified neuronal perikarya and whole cortex.

This is the first report of the lipid composition of human neurons. Neuronoal perikarya were isolated from frozen samples of the cerebral cortex of persons with Huntington disease and two normal controls. These were analyzed for total lipid, individual lipids, and gangliosides. No differences were detected between diseased and normal cells. In addition, gray matter samples from the same patients, and one additional patient and control sample, were analyzed and found not to differ. Thus the ultrastructural abnormalities seen in cortical biopsies are not reflected in the concentration of the major lipid classes.

Biochemical and pathological studies of myelin in hexachlorophene intoxication.

Adult female rats were fed a diet containing 500 ppm hexachlorophene (HCP). Morphological study of brains from these animals showed vacuolation of the myelin sheaths due to separation of myelin lamellae at the minor dense line. However, myelin could be isolated from the brains of these animals in normal yield. The myelin isolated from HCP-fed animals had normal lipid and protein compositions as shown by analyses of the individual lipids and by disc gel electrophoresis of the proteins. Assay of the myelin-specific enzyme, 2',3'-cyclic nucleotide-3'-phosphohydrolase, showed normal specific activity in myelin obtained from HCP-fed rats. Brains of HCP-fed rats showed an increase in wet weight and a decrease in dry weight, with the chloroform-methanol insoluble fraction showing the greatest weight loss. During isolation of myelin from HCP-fed rats material was found floating over 0.32 M sucrose. This "floating fraction" contained a higher ratio of lipid to protein but the same relative proportions of the individual lipids as are found in myelin. The yield of "floating fraction" from each HCP-fed rat was less than 10 percent of the yield of myelin. Disc gel electrophoresis demonstrated the presence of the usual myelin proteins in this fraction, but with a slight increase in the relative amount of the low molecular weight basic protein. The data were compared to reports on the biochemistry of triethyltin poisoning, and it was concluded that vacuolation of myelin in HCP poisoning is probably due to increased permeability of myelin lamellae to water and electrolytes.

Biochemical and immunocytochemical changes in glial fibrillary acidic protein after stab wounds.

Changes of glial fibrillary acidic protein (GFAP) in the forebrain of rats with stab wounds were determined by quantitative immunoblots and by immunohistochemistry. Bilateral stab wounds were made stereotaxically in the cortex and hippocampus. In control rats, the scalp was retracted and depressions were etched on the intact skull. At various times up to 21 days postoperation, one cerebral hemisphere was homogenized, proteins were separated by polyacrylamide gel electrophoresis and immunoblots were quantitated by densitometry. The contralateral hemisphere was immunostained for GFAP. Three hours postoperation GFAP+ cells were detected around the wound but there was no increase of total GFAP. At 6 h postoperation total GFAP in the forebrain decreased to 80% of the sham-operated control value and the number of GFAP+ cells was lower, compared to the controls, in layer 1 of the cortex, corpus callosum, cingulum, external capsule, internal capsule, hippocampus, optic tracts and around blood vessels. This early relative decrease in GFAP levels was actually due to an increase in GFAP in the sham-operated controls, which mounted a stronger gliotic response during the first 24 h. In neither group of animals did the GFAP levels drops below those of intact unoperated animals. At 24 h total GFAP began to increase. The number and intensity of reactive glia in the vicinity of the wound increased steadily, appearing to reach a maximum at about 7 days, then declining significantly by 21 days. The glial reaction was most pronounced in the hippocampus. Total GFAP reached 180% of the control value by 7 days and then declined to 117% by 21 days.(ABSTRACT TRUNCATED AT 250 WORDS)

GFAP mRNA levels following stab wounds in rat brain.

We previously reported that glial fibrillary acidic protein (GFAP) levels increased significantly at 3 days after stab wounds, relative to sham-operated controls, reaching a maximum of 200% of control value at 5-7 days. They then fell to near-normal values by 21 days. To determine whether these protein changes correlated with changes in GFAP mRNA we performed Northern blot analyses. Total RNA, isolated from lesioned, sham-operated and intact rat forebrains, was hybridized with 32P-labeled mouse GFAP cDNA and quantified by densitometry. The maximum increase in total RNA content in lesioned animals was only 20% over controls at 12 h. GFAP mRNA levels increased to 2-fold control values at 6 h and reached 5-fold at 12 h. Thereafter they remained at 3.5- to 6-fold until 5 days and then declined to 1.5-fold by 21 days. The rapid increase of GFAP message at 12 h preceded a significant increase in GFAP by 2 days and the decrease of message after 5 days was more precipitate than the slow decrease in GFAP content. Sham-operated animals showed no significant changes in GFAP mRNA, compared to intact controls, during the period 3 h to 14 days postoperation. GFAP mRNA and GFAP in the stab-wound model reached levels similar to those found in the experimental autoimmune encephalomyelitis (EAE) model, but returned to normal much more rapidly.

The long term culture of bulk-isolated bovine oligodendroglia from adult brain.

Oligodendroglia isolated from adult bovine brain by the method of Farooq et al. could be plated on polylysine-coated plastic dishes with an efficiency of 55-80%, and maintained in culture for as long as 4 months. The addition of cytosine arabinoside to the nutrient medium resulted in cultures that were approximately 90% oligodendroglia and 10% large fibroblasts. From 50 g of white matter 100-160 X 10(6) oligodendroglia, containing approximately 6-10 mg protein, could be obtained in culture. These small round cells started to send out processes at 5 days in vitro and by 2 weeks they formed an extensive network of processes. By immunofluorescence, all cells of this morphology were positive for galactocerebroside (GC) and myelin basic protein (MBP), and negative for glial filament protein and fibronectin. Most of the large flat cells were positive for fibronectin and negative for GC, MBP and glial filament protein. As the cultures aged the oligodendroglia tended to clump and blebs formed on the surface of both perikarya and processes. By 4 months they showed evidence of degeneration and detached from the substrate. Electron microscope examination showed that the cells had the appearance typical of oligodendroglia in situ. The somata were round to elliptical, with eccentrically placed nuclei, and were larger than freshly isolated cells. They grew directly on the substrate or on the surface of the fibroblasts. In older cultures the cells formed tight nests. The somata were enveloped by sheets of oligodendrocyte cytoplasm, sometimes having a myelin-like appearance. Gap junctions and small desmosomes were seen between oligodendroglial processes and between oligodendroglia and fibroblasts. The cytoplasm was characterized by a prominent Golgi apparatus, many mitochondria and lysosomes, scattered rough endoplasmic reticulum, free ribosomes, frequent centrioles and an abundance of microtubules. In cells from older cultures large vacuoles were common, and rarely they had multilamellar walls with alternating major and minor dense lines resembling myelin.

Quantitation of myelin carbonic anhydrase-development and subfractionation of rat brain myelin and comparison with myelin from other species.

A number of related studies have been performed to characterize further the carbonic anhydrase activity of myelin. Recent assertions that carbonic anhydrase activity is intrinsic to the myelin sheath were subjected to the additional test of isolation of rat brain myelin in the presence of purified carbonic anhydrase. This procedure did not increase the carbonic anhydrase activity in myelin above the endogenous level, indicating that this enzyme does not stick to myelin membranes. A developmental study of rat brain carbonic anhydrase showed that the enzyme activity increased in whole brain homogenates and in myelin, with the greatest increments in enzyme activity occurring before the animals were 60 days old. When myelin from adult rat brains was fractionated on a density gradient, carbonic anhydrase activity was relatively enriched in the heavy subfraction but was present in all three layers. This finding suggested that the activity in myelin preparations was not due to contamination with a carbonic anhydrase-rich membrane fragment. Carbonic anhydrase in myelin was not confined to the rat. Beef brain homogenates and myelin had low activities of the enzyme, but myelin from rabbit, cat, monkey and mouse had carbonic anhydrase activities comparable to that of the rat, accounting for 6.3--13.6% of the respective homogenate activities.

The isolation of cerebral neurons with partial retention of processes.

A novel tissue disaggregation technique has been devised which permits the isolation of neurons with fairly extensive processes attached. Cortex is dissociated by aspiration through nozzles of decreasing size followed by agitation on a vortex mixer, rather than by the usual technique of forcing tissue through sieves. After each aspiration step, dissociated cells are separated from undisrupted tissue by coarse filtration and the latter is subjected to repeated treatment. This prevents unnecessary trauma to the free cells. After disruption is complete, small pieces of undisrupted tissue are removed from the cell suspension by floating on the foam created by degassing the suspension under vacuum. Cells are purified by conventional velocity-gradient centrifugation. This procedure has been applied successfully to fresh rat brain, with or without a preincubation with trypsin, frozen human brain and frozen bovine brain. The cell yields from rat brain were comparable to or better than, those obtained by other procedures (37 X 10(6) cells/g brain) while the purity was comparable. Cell yields from human brain were similar to those from rat brain but the purity was lower. The lowered particle purity of human and bovine cells can probably be attributed to the conditions of storage of the tissue and to trapping of free nuclei in the meshwork of dendritic processes. Values are given for the amount of protein, RNA and DNA per cell.

Astrocytic reactivity and intermediate filament metabolism in experimental autoimmune encephalomyelitis: the effect of suppression with prazosin.

In either actively or passively transferred experimental autoimmune encephalomyelitis (EAE), increased immunocytochemical staining of glial fibrillary acidic protein (GFAP) in astrocytes was detected early in the disease process in both the gray and white matter of the spinal cord. Staining was not restricted to areas of perivascular mononuclear infiltration, and was observed at all levels of the cord. This enhanced staining pattern was delayed in rats in which clinical signs of EAE had been suppressed by treatment with the alpha 1-adrenoceptor antagonist prazosin. This glial reaction in EAE was not accompanied by increased GFAP synthesis, as measured by in vitro labeling of spinal cord slices, nor an increase in GFAP content, as measured by densitometry of intermediate filament fractions separated by polyacrylamide gel electrophoresis. Total protein synthesis was increased, with vimentin being labeled especially heavily; in prazosin-treated EAE animals, the increase in total protein synthesis was reduced and delayed.

Complement potentiates the degradation of myelin proteins by plasmin: implications for a mechanism of inflammatory demyelination.

A previous finding, that the basic protein in lyophilized bovine myelin was degraded by macrophage-conditioned media in the presence of plasminogen, suggested that the macrophage-secreted plasminogen activator, along with plasminogen, might have a role in destruction of myelin during inflammatory demyelination. To approximate more closely the conditions expected in vivo, plasmin, or macrophage supernatants plus plasminogen, were incubated with freshly homogenized bovine white matter or freshly isolated myelin, as distinguished from lyophilized myelin. Under these conditions basic protein was not degraded. Phospholipase or lysolecithin potentiated the degradation of basic protein in fresh bovine myelin by plasmin; however, the cultured macrophages did not secrete significant amounts of phospholipase and plasminogen activator simultaneously into the culture media after activation with any of several different agents. Recently myelin was shown to activate complement. After preincubation of fresh myelin with guinea pig serum, as a source of complement, the basic and proteolipid proteins were vulnerable to plasmin or to macrophage-conditioned media plus plasminogen. C3-depleted and C4-deficient sera were not effective, suggesting that these complement components were required for the serum effect. Hypothetically, then, degradation of myelin proteins in the CNS could be initiated by plasminogen activator, secreted by infiltrating macrophages, plus complement and plasminogen, which could enter the CNS through lesions in the blood-brain barrier.

Ganglioside content of astroglia and neurons isolated from maturing rat brain: consideration of the source of astroglial gangliosides.

Previous biochemical and histochemical studies have failed to clarify the nature or quantity of gangliosides in CNS astrocytes. Using improved methodologies for bulk isolation of both neurons and astrocytes as well as for ganglioside purification, we find significantly higher ganglioside concentration in astrocytes and very similar thin-layer chromatography (TLC) patterns for the two cell types. However, in vivo labeling of glycoconjugates via intracerebral injection of [3H]glucosamine prior to cell isolation revealed a different picture: whereas glycoproteins were well-labeled in both cell types after labeling periods of 1-2 h, gangliosides were appreciably labeled only in neurons. With longer time periods (8-48 h) between injection and sacrifice, there was convergence of specific radioactivity of gangliosides from the two isolated cell preparations. These changes are compared to those observed in synaptosomes and microsomes that were isolated simultaneously. The results suggest limited ganglioside synthetic ability in astrocytes as compared to neurons, a conclusion supported by assay of UDP-galNAc:GM3 N-acetylgalactosaminyltransferase in the isolated cells. Nevertheless, the presence of ganglioside GM1 in a substantial portion of bulk-isolated astrocytes was demonstrated by indirect immunofluorescent detection of cholera toxin binding. Ideas on the reconciliation of these apparently contradictory phenomena, including the possibility of intercellular transfer and/or phagocytosis are discussed.

Studies on the encephalitogenic effects of purified preparations of human and bovine oligodendrocytes.

Bulk-isolated human and bovine oligodendroglia, practically free from myelin, have been used in attempts to elicit an autoimmune response which has been compared with acute experimental allergic encephalomyelitis (EAE). For these experiments, a total of 20 Hartley guinea pigs, 33 Lewis rats and 16 rabbits have been studied. Animals were inoculated with a range of doses of purified preparations of both human and bovine oligodendroglial cells in complete Freund's adjuvant (CFA) and compared with others challenged with whole white matter in CFA. The latter animals all developed clinical and histological signs of experimental allergic encephalomyelitis (EAE) 2-3 weeks post-inoculation. In general, oligodendroglial cells were encephalitogenically less potent than white matter. Guinea pigs were the most susceptible to inoculations of oligodendroglia. In several given human oligodendroglia 14 days earlier, a paraparesis indistinguishable from conventional EAE was seen. Animals receiving bovine cells showed no clinical signs. Histologically, the CNS of afflicted guinea pigs displayed severe inflammation but, in contrast to conventional EAE in the same species, demyelination was rare in the small group of animals tested. After sensitization with oligodendroglia, rats displayed no clinical disease. Histologically, some given human cells had positive evidence of disease while bovine cells in others gave a mild response. Rabbits showed no clinical and very little histological disease. Although more extensive studies are needed to confirm the findings, from the animals studied it appears that (1) variation in response to inocula containing oligodendroglia exists among the species tested, (2) that human oligodendroglia are more potent immunologically than bovine cells, (3) that CNS lesions produced by these cells in guinea pigs, lack a strong demyelinative component and (4) a specific antigen might exist in oligodendrocytes which is distinct from myelin basic protein. The possible reasons underlying our findings are discussed.