Temporospatial sequence of cellular events associated with etoposide-induced neuronal cell death: role of antiapoptotic protein Bcl-X(L)

Etoposide-induced death comprises such nuclear events as the formation of topoisomerase II-DNA cleavable complex and cytosolic events including caspase activation. By first establishing the temporospatial death sequence triggered by etoposide in a neuronal cell line, MN9D overexpressing Bcl-X(L) (MN9D/Bcl-X(L)) or control vector (MN9D/Neo), we examined whether formation of this complex is primarily responsible for cell death and at which strategic points and how Bcl-X(L) blocks etoposide-induced neuronal death. Etoposide induced death that was dependent on caspase, cycloheximide, and calpain in MN9D/Neo cells. Etoposide also induced death in enucleated MN9D/Neo cells, although this was less severe. The level of topoisomerase II-DNA cleavable complex reached at a maximum of 2 hr after etoposide treatment was identical in MN9D/Neo and MN9D/Bcl-X(L) cells. In MN9D/Neo cells, cytochrome c release into the cytosol and caspase activation occurred as early as 2 hr and 3-6 hr after etoposide treatment, respectively. Etoposide-induced DNA laddering potentially via caspase appeared as early as 12 hr after drug treatment, followed by nuclear swelling in MN9D/Neo cells (>18-20 hr). Subsequently, nuclear condensation started by 24-28 hr and became apparent thereafter. All of these events except for nuclear swelling were substantially blocked in MN9D/Bcl-X(L). At the later stage of cell death (<32-36 hr), a specific cleavage of Bax and fodrin appeared that was completely blocked by calpain inhibitor or by Bcl-X(L). Taken together, our data suggest that Bcl-X(L) prevents etoposide-induced neuronal death by exerting its anticaspase and anticalpain effect on cellular events after the formation of topoisomerase II-DNA cleavable complex that may not be a major contributor to cell death.

Overexpression of calbindin-D28K induces neurite outgrowth in dopaminergic neuronal cells via activation of p38 MAPK

An MN9D dopaminergic neuronal cell line overexpressing calbindin-D28K (MN9D/Calbindin) was established in order to investigate directly the potential role of calcium-binding protein in neuronal differentiation. Overexpression of calbindin-D28K in MN9D cells resulted in significant increases in the number of neurites, the length of primary neurites, and the total extent of neurites. This robust neurite outgrowth occurred without cessation of cell division. Analysis of immunoblots revealed that this morphological differentiation was accompanied by increased expression of such markers of maturation as the synaptosomal protein SNAP-25. During calbindin-D28K-evoked neurite outgrowth in MN9D cells, phosphorylation of p38 mitogen-activated protein kinase (MAPK) dramatically increased while the levels and extent of phosphorylation of such other MAPKs as c-Jun N-terminal kinase (JNK) or extracellular response kinase (ERK) were not altered. Consequently, calbindin-D28K-induced neurite outgrowth was largely abolished by treatment with a p38 inhibitor, PD 169316, while the level of SNAP-25 in MN9D/Calbindin cells was not altered by this treatment. These data support an idea that calbindin-D28K and its associated p38 signaling pathway play a role in dopaminergic neuronal differentiation.

MPP(+) downregulates mitochondrially encoded gene transcripts and their activities in dopaminergic neuronal cells: protective role of Bcl-2

The effects of neurotoxins on levels of mitochondrially encoded gene transcripts in a dopaminergic neuronal cell line, MN9D, were examined following treatment with 200 microM N-methyl-4-phenylpyridinium (MPP(+)) or 6-hydroxydopamine (6-OHDA). As confirmed by a Northern blot analysis, levels of cytochrome c oxidase subunit 3 (COX III) and ATPase subunit 6 (ATPase 6) transcript were decreased in a time-dependent manner following treatment with MPP(+) but not with 6-OHDA. Accordingly, enzymatic activity of cytochrome c oxidase (COX) and the intracellular ATP content were also decreased in MPP(+)-treated cells while these remained unaltered in 6-OHDA-treated cells. In the cell death paradigm induced by MPP(+), overexpression of Bcl-2 in MN9D cells (MN9D/Bcl-2) significantly blocked MPP(+)-induced downregulation of COX III and ATPase 6 transcripts. In MN9D/Bcl-2 cells, MPP(+)-induced downregulation of COX activity and the intracellular level of ATP was also blocked. Treatment with a pan-caspase inhibitor, however, neither prevented MPP(+)-induced downregulation of COX activity nor affected intracellular level of ATP in MN9D cells. Taken together, our present data suggest that Bcl-2 may play a regulatory role in energy metabolism by preventing downregulation of mitochondrially encoded gene(s) at a point distinct from its known anticaspase activity in MPP(+)-induced dopaminergic neuronal death.

Role of tumor necrosis factor-alpha in neuronal and glial apoptosis after spinal cord injury

We investigated the role of tumor necrosis factor (TNF)-alpha in the onset of neuronal and glial apoptosis after traumatic spinal cord crush injury in rats. A few TUNEL-positive cells were first observed within and surrounding the lesion area 4 h after injury, with the largest number observed 24-48 h after injury. Double-labeling of cells using cell type-specific markers revealed that TUNEL-positive cells were either neurons or oligodendrocytes. One hour after injury, an intense immunoreactivity to TNF-alpha was observed in neurons and glial cells in the lesion area, but also seen in cells several mm from the lesion site rostrally and caudally. The level of nitric oxide (NO) also significantly increased in the spinal cord 4 h after injury. The injection of a neutralizing antibody against TNF-alpha into the lesion site several min after injury significantly reduced both the level of NO observed 4 h thereafter as well as the number of apoptotic cells observed 24 h after spinal cord trauma. An inhibitor of nitric oxide synthase (NOS), N(G)-monomethyl-l-arginine acetate (l-NMMA), also reduced the number of apoptotic cells. This reduction of apoptotic cells was associated with a decrease in DNA laddering on agarose gel electrophoresis. These results suggest that: (i) TNF-alpha may function as an external signal initiating apoptosis in neurons and oligodendrocytes after spinal cord injury; and (ii) TNF-alpha-initiated apoptosis may be mediated in part by NO as produced by a NOS expressed in response to TNF-alpha.

Rapid increase in immunoreactivity to GFAP in astrocytes in vitro induced by acidic pH is mediated by calcium influx and calpain I

In higher vertebrates, reactive gliosis resulting from injury to the central nervous system (CNS) is characterized by a rapid increase in immunoreactivity (IR) to glial fibrillary acidic protein (GFAP). Little is known about the extracellular signals that initiate the increase in GFAP-IR following CNS injury. We demonstrated recently [T.H. Oh, G.J. Markelonis, J.R. Von Visger, B. Baik, M.T. Shipley, Acidic pH rapidly increases immunoreactivity of glial fibrillary acidic protein in cultured astrocytes, Glia 13 (1995) 319-322] that a rapid increase in GFAP-IR can be evoked in mature astrocyte cultures by exposing the cells to an acidic medium. We investigated the intracellular pathway(s) involved in initiating increased GFAP-IR, a hallmark of reactive astrocytes. The increase in GFAP-IR produced by exposure to acidic medium was blocked by pretreatment with nickel ions, by such blockers of L-type calcium channels as nifedipine, verapamil and diltiazem, by calpain inhibitor I, or by the intracellular calcium chelator, BAPTA-AM. At physiological pH, treatment with the calcium ionophore, A23187, resulted in increased GFAP-IR which could be blocked by pretreatment with calpain inhibitor I. Astrocytes exposed to low pH exhibited a marked increase in a GFAP fragment with a molecular weight of 48 kDa. In astrocytes exposed to acidic medium, alpha-fodrin, a selective endogenous substrate of calpain, was also found to be hydrolyzed producing fragments with molecular weights of 120-150 kDa. As anticipated, pretreatment with calpain inhibitor I prevented the proteolytic degradation of both GFAP and alpha-fodrin in these samples. These results suggest that the initial increase in GFAP-IR after CNS injury appears to be linked to Ca(++) influx, and is mediated further by a proteolytic process that seemingly involves the activation of the calcium-dependent protease, calpain I.

Cynandione A from Cynanchum wilfordii protects cultured cortical neurons from toxicity induced by H2O2, L-glutamate, and kainate

Oxidative stress has been implicated as a primary cause of neuronal death in certain neurodegenerative disorders and in aging brains. Natural products have been used in Asian societies for centuries for treating such neurodegenerative disorders as senile dementia. In an effort to identify active neuroprotective compounds from these products, we have employed cultures of rat cortical neurons as our screening system. A methanolic extract from dried roots of Cynanchum wilfordii Hemsley (Asclepiadaceae) significantly mitigated the neurotoxicity induced by H2O2 in this screening system. Activity-guided fractionation using several chromatographic techniques resulted in the isolation of the neuroprotective compound, cynandione A, a biacetophenone. At a concentration of 50 microM, cynandione A significantly reduced neurotoxicity induced by H2O2. Cynandione A significantly attenuated decreases in levels of glutathione, superoxide dismutase, and other enzymes that participate in the cellular defense against oxidative stress. Furthermore, cynandione A alleviated neurotoxicity induced by the excitotoxic neurotransmitter, L-glutamate, the neurotoxicity induced by kainate, but not that mediated by N-methyl-D-aspartate. Cynandione A was demonstrated to be a natural antioxidant as it facilitated the breakdown of hydrogen peroxide in vitro; however, no mechanism was uncovered to explain its neuroprotectant effects against glutamate and kainate. Therefore, cynandione A may be efficacious in protecting neurons from oxidative stress mediated via activation of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate/kainate receptors since it exerted significant neuroprotective effects on cultured cortical neurons.

Calcium influx and activation of calpain I mediate acute reactive gliosis in injured spinal cord

Buffering extracellular pH at the site of a spinal cord crush-injury may stimulate axonal regeneration in rats (1; Guth et al., Exp. Neurol. 88: 44-55, 1985). We demonstrated in cultured astrocytes that acidic pH initiates a rapid increase in immunoreactivity for GFAP (GFAP-IR), a hallmark of reactive gliosis (2; Oh et al., Glia 13: 319-322, 1995). We extended these studies by investigating the effects of certain treatments on reactive gliosis developing in situ in a rat spinal cord injury model. A significant reactive gliosis was observed within 2 days of cord lesion in untreated crush or vehicle-treated, crush control animals as evidenced by increased GFAP-IR and hypertrophy of astrocytes. By contrast, infusion of Pipes buffer (pH 7.4) into the lesion site significantly reduced this increase. The increased GFAP-IR appeared to be linked to Ca2+ influx since infusion of a blocker of L-type calcium channels, nifedipine, reduced the ensuing reactive gliosis significantly. While Ca2+ modulates many signaling pathways within cells, its effect on reactive gliosis appeared to result from an activation of calpain I. Calpain inhibitor I, a selective inhibitor of mu-calpain, also significantly reduced reactive gliosis. However, calpain inhibitor II, a close structural analog which blocks m-calpain, had no salutary effect. We suggest, therefore, that the initial reactive gliosis seen in vivo may result from the activation of a neutral, Ca2+-dependent protease, calpain I, through calcium influx.

Protopine from Corydalis ternata has anticholinesterase and antiamnesic activities

While screening extracts of natural products in search of anticholinesterase activity, we found that a total methanolic extract of the tuber of Corydalis ternata (Papaveraceae) showed significant inhibitory effects on acetylcholinesterase. Further fractionation of this extract using acetylcholinesterase inhibition as the parameter screened resulted in the isolation and purification of an alkaloid, protopine. Protopine inhibited acetylcholinesterase activity in a dose-dependent manner. The concentration required for 50% inhibition was 50 microM. The anti-acetylcholinesterase activity of protopine was specific reversible and competitive in manner. Furthermore, when mice were pretreated with protopine, the alkaloid significantly alleviated scopolamine-induced memory impairment. In fact, protopine had an efficacy almost identical to that of velnacrine, a tacrine derivative developed by a major drug manufacturer to treat Alzheimer's disease, at an identical therapeutic concentration. We suggest, therefore, that protopine has both anti-acetylcholinesterase and antiamnesic properties that may ultimately hold significant therapeutic value in alleviating certain memory impairments observed in dementia.

Ginsenosides Rb1 and Rg3 protect cultured rat cortical cells from glutamate-induced neurodegeneration

Certain natural products and Asian herbal remedies have been used in Asia to attenuate neurodegenerative diseases, including senile dementia. We have examined derivatives of several natural products for potential neuroprotective activity in an in vitro test system. In the present study, we assayed a number of compounds that were isolated from Panax ginseng C.A. Meyer (Araliaceae) for an ability to protect rat cortical cell cultures from the deleterious effects of the neurotoxicant, glutamate. We found that ginsenosides Rb1 and Rg3 significantly attenuated glutamate-induced neurotoxicity. Brief exposure of cultures to excess glutamate caused extensive neuronal death. Glutamate-induced neuronal cell damage was reduced significantly by pretreatment with Rb1 and Rg3. Ginsenosides Rb1 and Rg3 inhibited the overproduction of nitric oxide, which routinely follows glutamate neurotoxicity, and preserved the level of superoxide dismutase in glutamate-treated cells. Furthermore, in cultures treated with glutamate, these ginsenosides inhibited the formation of malondialdehyde, a compound that is produced during lipid peroxidation, and diminished the influx of calcium. These results show that ginsenosides Rb1 and Rg3 exerted significant neuroprotective effects on cultured cortical cells. Therefore, these compounds may be efficacious in protecting neurons from oxidative damage that is produced by exposure to excess glutamate.

Acidic pH rapidly increases immunoreactivity of glial fibrillary acidic protein in cultured astrocytes

Neuroepithelial progenitor cells from forebrains of newborn rat pups develop into "mature" astrocytes in an epidermal growth factor-containing medium free of serum (Von Visger et al: Exp Neurol 128:34, 1994). Eight-week-old "mature" astrocyte cultures on poly-L-lysine-coated dishes were exposed to an acidic medium (pH 5.8-6.0) for 2-6 h. Immunoreactivity for glial fibrillary acidic protein (GFAP) dramatically and rapidly increased; this immediate increase was not affected by pretreatment with cycloheximide. In further experiments we found that the increase in GFAP was undiminished for 24-48 h after the acid-treated astrocytes were returned to normal growth medium. The Ca2+ channel antagonists nifedipine and diltiazem attenuated the increase in GFAP immunoreactivity. These results suggest that extracellular acidosis may produce a rapid increase in GFAP immunoreactivity in astrocytes independent of de novo protein synthesis, possibly by increasing intracellular levels of free Ca2+ ions.

Neuroglial responses to the dopaminergic neurotoxicant 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine in mouse striatum

We have studied the reactive responses of both astrocytes and microglia to dopaminergic denervation of the striatum by MPTP. Following MPTP treatment, increased GFAP immunoreactivity reached a peak at 2 days and persisted for at least 6 weeks. Immunoreactivity to vimentin was also markedly increased in astrocytes 48 h after MPTP treatment. Striatal laminin immunoreactivity, however, appeared to be unaffected by drug treatment. GFAP protein levels increased to 196% and 321% of control 24 and 48 hours after MPTP treatment, respectively. Concomitantly, GFAP mRNA levels increased to 560% and 1620% of control, respectively. These reactive changes in striatal astrocytes in response to MPTP treatment were also accompanied by a reactive microglial response as evidenced by increased immunohistochemical visualization of striatal microglia using antibodies to Mac-1. Our results and those reported previously by O'Callaghan et al., strongly suggest that MPTP-induced reactive gliosis in mouse striatum is associated with reactive microglia, albeit without increased interleukin-1 beta.

Differentiation and maturation of astrocytes derived from neuroepithelial progenitor cells in culture

Neuroepithelial progenitor cells from striata of adult mice develop into either astrocytes or neurons when cultured in the presence of epidermal growth factor (B. A. Reynolds, and S. Weiss, Science 255: 1707-1710, 1992). We instituted primary cultures of such progenitor cells from forebrains of newborn rat pups in an epidermal growth factor-containing medium free of serum in order to study the development of astrocytes in culture. At 4-6 days, primary cultures consisted of floating clusters of proliferating cells which expressed nestin, a marker for neuroepithelial progenitor cells, the ganglioside GD3, and vimentin. When clusters were transferred to polylysine-coated dishes, cells attached to the substrate and began to express antigens characteristic of particular differentiated neurons, astrocytes, or oligodendrocytes within 2 weeks. In 4- to 6-week-old secondary cultures, levels of vimentin expression appeared to decrease within maturing astrocytes which had increased levels of glial fibrillary acidic protein. These results suggest that multipotential epidermal growth factor-progenitor cells can give rise to both neurons and macroglia of the adult central nervous system, and that maturation of the astrocytes in vitro may be occurring in a pattern similar to that seen in vivo. Furthermore, no glial fibrillary acidic protein-positive cells expressed the A2B5 antigen in the same cell indicating an absence of type-2 astrocytes.

Effects of interleukin-1 beta and tumor necrosis factor-alpha on the expression of glial fibrillary acidic protein and transferrin in cultured astrocytes

Recent evidence suggests that interleukin (IL)-1 and tumor necrosis factor (TNF) may play a role in astrogliosis following injury to the CNS. The short-term biochemical effects of these immune-related cytokines were determined on cultured rat polygonal and process-bearing astrocytes. Both IL-1 and TNF stimulated the rate of thymidine incorporation in polygonal astrocytes up to 137% and 215%, respectively, over the level observed in untreated controls. By contrast, thymidine incorporation was relatively unaffected by these cytokines in process-bearing astrocytes. The cytokines did not significantly affect the level of glial fibrillary acidic protein (GFAP) within polygonal astrocytes, even though they appeared to downregulate the expression of GFAP mRNA by as much as 62%. Both cytokines increased the intracellular expression of transferrin (Tf) within some polygonal astrocytes. In untreated control cultures, fewer than than 2% of polygonal astrocytes were immunoreactive for Tf. By contrast, approximately 30% of polygonal astrocytes treated with IL-1 or TNF-alpha became strongly immunoreactive for Tf. Neither IL-2 nor a number of other known growth factors appeared to alter the level of immunoreactive Tf in these cells. Process-bearing astrocytes were negative for Tf, regardless of the treatment used. Northern blot analysis demonstrated that the level of Tf mRNA in cultures of polygonal astrocytes increased 148% above the level observed in untreated controls following treatment with either IL-1 or TNF, whereas no change was observed following treatment with IL-2. These results suggest that increased levels of particular cytokines known to be present in injured CNS can produce pronounced biochemical alterations within a subtype of cultured astrocytes.

Interleukin-1-beta and tumor necrosis factor-alpha increase peripheral-type benzodiazepine binding sites in cultured polygonal astrocytes

Peripheral-type benzodiazepine binding sites (PTBBS) are markedly increased in the injured CNS. Astrocytes appear to be the primary cell type which express increased PTBBS. Because certain cytokines within the injured CNS are potent mitogens for astrocytes, we examined the effects of two such cytokines, interleukin (IL)-1 beta and tumor necrosis factor (TNF), on PTBBS in cultured astrocytes using [3H]Ro 5-4864 as the specific ligand. Purified cultures of either polygonal or process-bearing astrocytes were prepared from neonatal rat cerebral hemispheres. At a concentration of 1.8 nM, specific binding of the radioactive ligand to polygonal astrocytes reached equilibrium within 60 min and was half-maximal by 5-10 min. By contrast, specific binding to process-bearing astrocytes barely exceeded background levels. IL-1 and TNF increased PTBBS within polygonal astrocytes in both dose- and time-dependent manners. At 10-50 ng/ml, IL-1 beta and TNF-alpha elevated [3H]Ro 5-4864 binding in polygonal astrocyte cultures 65 and 87%, respectively, above the level in control cultures. However, no changes in PTBBS were seen within polygonal astrocytes after IL-2 treatment. Scatchard analysis of saturation binding experiments suggested that the increase in PTBBS promoted by TNF was due to an increased number of binding sites present in polygonal astrocytes and not due to an increase in receptor affinity. Binding data suggested that PTBBS within cultures of process-bearing astrocytes were virtually absent irrespective of the treatment. These in vitro data suggest that certain cytokines found in the injured brain may be involved in up-regulating PTBBS within a particular subtype of astrocyte.

A neurite-promoting factor from muscle supports the survival of cultured chicken spinal motor neurons

During embryonic development, spinal motor neurons require muscle-derived trophic factors for their survival and growth. We have recently isolated a protein from muscle that is not laminin but that still stimulates neurite outgrowth from embryonic neurons in culture. In the present study, we investigated whether this protein, which we refer to as muscle-derived neurite-promoting factor (NPF), could also promote the survival and growth of motor neurons in culture. Spinal motor neurons were isolated from 6-day-old chicken embryos by a metrizamide step-gradient centrifugation protocol. Most large cells (putative motor neurons) were found in the upper metrizamide fraction (0%-6.8% interface; fraction I). Motor neurons were identified by increased specific activity of choline acetyltransferase (CAT) and by their propensity to transport retrogradely either wheat germ agglutinin-horseradish peroxidase or the fluorescent dye, 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine per chlorate (diI), when those substances were injected into the target field. Labeled motor neurons were 2.6-fold enriched in fraction I and the specific CAT activity was 4.4-fold increased in fraction I as compared to unfractionated cells. When motor neurons were grown on muscle-derived NPF, the protein supported the survival of at least 21% of the neurons for as long as 6 days in culture. The protein showed no significant effect on either CAT specific activity or on high-affinity choline uptake by neurons. There was a substantial increase from 21% to 38% of the survival of motor neurons when a combination of muscle-derived NPF and laminin was used as the substrate. Muscle-derived NPF also promoted the survival of sensory neurons and sympathetic neurons in culture. Our results demonstrate that a neurite-promoting protein derived from muscle promotes both the survival and the outgrowth of neurites from cultured spinal motor neurons as well as from sensory and sympathetic neurons.

Immunocytochemical localization of mitochondrial malate dehydrogenase in primary cultures of rat astrocytes and oligodendrocytes

To assess the oxidative metabolism of glial cells, we visualized mitochondrial malate dehydrogenase (mMDH) in purified cultures of neonatal rat polygonal and process-bearing astrocytes as well as in oligodendrocytes, using indirect immunofluorescence. Double immunofluorescent localization of rabbit anti-mMDH and either mouse monoclonal antiglial fibrillary acidic protein or anti-myelin basic protein demonstrated that both process-bearing astrocytes and oligodendrocytes showed uniformly intense anti-mMDH immunoreactivity in their cell bodies. However, immunoreactivity to mMDH among polygonal astrocytes varied from very weakly positive to intensely positive. Experiments with rhodamine 123, a mitochondrion-specific fluorochrome, indicated that polygonal astrocytes contain relatively similar numbers of mitochondria; this suggested that the variable intensities of anti-mMDH immunoreactivity observed did not result from differences in mitochondrial numbers. In cultures of polygonal astrocytes maintained in a chemically defined medium containing growth factors and hormones, or in complete culture medium containing 1mM N6, O2-dibutyryl adenosine 3',5'-cyclic phosphate, the resultant stellate astrocytes still showed their original variable levels of anti-mMDH immunoreactivity. This suggested that the mMDH distribution pattern did not depend on the degree of morphological differentiation. Furthermore, cultures of polygonal astrocytes isolated from four specific regions of neonatal rat brain showed variable but reproducible profiles of anti-mMDH immunoreactivity. Our results suggest that there may be an appreciable range in the level of oxidative metabolism among individual polygonal astrocytes in culture.

A muscle-derived substrate-bound factor that promotes neurite outgrowth from neurons of the central and peripheral nervous systems

The development and survival of spinal motor neurons depends upon muscle-derived trophic factors. Some circumstantial evidence suggested to us that the regulatory subunit of cyclic adenosine 3':5'-monophosphate-dependent protein kinase (cAMP-dPK)-type II might be involved in neuritic outgrowth from spinal neurons. In the present study, we tested a commercial preparation of cAMP-dPK for neurite-promoting activity. Commercial cAMP-dPK-type II from skeletal and cardiac muscles elicited a significant neurite outgrowth from cultured embryonic chicken neurons when the enzyme preparation was bound to polylysine-coated substrata; type I cAMP-dPK from skeletal muscle was ineffective. Neither cAMP-dPK-type I nor -type II had a significant effect on the survival of spinal neurons in culture. Type II cAMP-dPK also stimulated neurite outgrowth from chicken cerebral hemisphere neurons, dorsal root ganglionic neurons, ciliary ganglionic neurons, and rat sympathetic ganglionic neurons in culture. The neurite-promoting activity appears to reside in a contaminant of the preparation since neither the purified regulatory nor catalytic subunits of cAMP-dPK-type II had an effect on neurite outgrowth per se from cultured neurons and since neurite-promoting activity did not correlate with [3H]cAMP binding or cAMP-dependent kinase activity. The neurite-promoting protein was then partially purified from commercial cAMP-dPK-type II by gel filtration on Sephadex G-200 followed by ion-exchange chromatography on DE-52 cellulose. Sodium dodecyl sulfate gel electrophoresis of the active protein peak revealed a major protein band (MW 50 kDa) and several minor bands (e.g., MW 200 kDa, 52 kDa, 45 kDa). Also, immunoblot analysis and immunoprecipitation revealed that the partially purified neurite-promoting protein was distinct from laminin, heparan sulfate proteoglycan, nerve growth factor, neural cell adhesion molecule, and fibronectin. Furthermore, the neurite-promoting activity was not diminished by treatment with heparinase nor was it bound to heparin conjugated to Sepharose. Our results demonstrate that a protein unrelated to laminin or its associated macromolecules and which copurifies with the type II cAMP-dPK of striated muscle stimulates neurite outgrowth from neurons of the central and peripheral nervous systems.

Immunocytochemical localization of aldolase in normal, denervated, and dystrophic chicken muscles

To investigate whether immunocytochemical localization of muscle-specific aldolase can be used for fiber phenotype determination, we produced specific antibodies against the enzyme and studied its distribution in adult chicken skeletal muscles by indirect immunofluorescence microscopy. Monoclonal antibodies against the myosin heavy chains of fast-twitch (MF-14) and slow-tonic (ALD-58) muscle fibers were also used to correlate aldolase levels with the fiber phenotype. The goat anti-aldolase antibody was found to be specific for the A form of aldolase, as evidenced by sodium dodecyl sulfate gel electrophoresis, immunotitration experiments, and immunoblot analysis. The antibody reacted strongly with the fast-twitch myofibers of normal pectoralis and posterior latissimus dorsi muscles; the phenotype of these muscle fibers was confirmed by a positive immunofluorescent reaction after incubation with MF-14 antibody. By contrast, the slow-tonic myofibers of normal anterior latissimus dorsi, which react positively with ALD-58 antibody, reacted weakly with anti-aldolase antibodies. In denervated chicken muscles, reaction to anti-aldolase antibodies was markedly reduced in fast-twitch fibers, although reaction to MF-14 was not diminished. By contrast, in dystrophic muscle, fast-twitch fibers showed reduced reactivity to anti-aldolase and marked to moderate reduction in MF-14 reactivity. Our results show that: (a) in normal muscles, reactivity to anti-aldolase matches the phenotype obtained by using anti-fast or anti-slow myosin heavy chain antibodies, and therefore can serve to identify mature fibers as fast or slow; and (b) in denervated or dystrophic muscles, the intracellular expressions of aldolase and fast-twitch myosin heavy chains are regulated independently.

Immunocytochemical localization of transferrin and mitochondrial malate dehydrogenase in the developing nervous system of the rat

Transferrin accumulates within neurons of the developing nervous system of humans, sheep, pigs and chickens. To assess the relationship of this accumulation with the ontogeny of oxidative metabolism, we studied the immunocytochemical localization of transferrin (Tf) and the mitochondrial form of malate dehydrogenase (mMDH) in developing neural tissues by the peroxidase-antiperoxidase method. Rabbit anti-rat Tf was obtained commercially and gave a single band of reaction product (MW = 80 kd) on Western blots. Antibodies to porcine heart mMDH were elicited in a rabbit. Western blot analysis showed that this anti-porcine mMDH antibody reacted with the mMDH from porcine, rat or avian tissue but not with the cytosolic MDH from pigs. Tf was first detected in rat brain neurons at about the 18th embryonic day and reached a peak at about the 6th postnatal day. All neurons were immunoreactive with large neurons throughout the brain showing a strong reaction for Tf. From this time onward, the level in brain neurons gradually decreased until adulthood. However, Tf immunoreactivity still remained strongly evident in capillary endothelial cells. The localization of Tf within rat spinal cord neurons peaked as early as the 1st postnatal day and remained elevated to the 6th postnatal day. By contrast, reactivity for Tf within dorsal root ganglia neurons was intense as early as the 18th embryonic day and diminished only gradually. Mitochondrial MDH, a marker for oxidative metabolism, appeared to reach a peak after the crest of intraneuronal Tf had been observed. For example, brain and spinal cord MDH immunoreactivity increased with intense staining in the cell bodies and fibers of neurons from the 6th to the 13th postnatal day; immunoreactivity gradually diminished into adulthood. The gradient of reactivity was low in some areas of the brain but more intense in areas containing large neuronal cell bodies such as the red nucleus. This occurred after the peak of intraneuronal Tf at day 6 and suggested a precursor-product relationship. By contrast, immunoreactivity for neuron-specific enolase, a glycolytic enzyme, showed a developmental pattern that differed from either Tf or MDH in that reactivity appeared later in development and was less intense. These data suggest that as cerebral metabolic rates begin to increase as early as 5-6 days after birth in the rat, an increase in mMDH occurs coincident with the onset of oxidative metabolism. Furthermore, this rise in intraneuronal mMDH follows the peak of intraneuronal Tf and suggests that Tf supplies the iron required for the synthesis of other mitochondrial ferroproteins.

Transferrin: assay of myotrophic effects and method for immunocytochemical localization

1. Primary cultures of dissociated embryonic chicken skeletal muscle cells provide an ideal model for investigating the effects of growth factors such as Tf because these cells undergo a highly integrated pattern of differentiation and maturation. 2. The trophic effects of a growth factor such as Tf can be assessed on muscle cultures by the determination of such parameters as acetylcholinesterase and acetylcholine receptors. These proteins are specific to the cultured myotubes, appear in high levels following fusion of myoblasts into myotubes, and are relatively easy to assay. 3. Tf and other growth factors are internalized by a receptor-mediated mechanism (see Trowbridge et al. and Seligman and Allen, this volume). These growth factors can be localized to specific tissues by immunocytochemistry at the light or electron microscopic level. This information on cellular distribution could be very useful in assessing the pattern of growth and differentiation with regard to the particular growth factor under study.

Immunocytochemical distribution of transferrin and its receptor in the developing chicken nervous system

Transferrin is the plasma protein responsible for iron transport in all vertebrates. While transferrin is known to have growth-promoting activity on a variety of cells in culture, the role of transferrin and its membrane receptor in neuronal development is unknown. Using antibodies to transferrin and transferrin receptors, we studied the immunocytochemical localization of transferrin and its receptor in developing chicken neural tissues by the peroxidase-antiperoxidase method. In 5-day-old embryonic brain, germinal cells of the ventricular zone showed a positive reaction for transferrin receptors but were negative for transferrin. By 6-7 days, transferrin-positive cells were seen in the inner layer of the ventricular zone and a few 'patches' of transferrin-positive cells were also seen in the adjacent area. By 10 days, large neurons throughout the brain were strongly positive for transferrin. By 11-16 days, all neurons in the brain showed a strong positive reaction for the protein. Thereafter, the transferrin-positive reaction became gradually weaker in neurons whereas the walls of blood capillaries showed a positive reaction for transferrin. In the adult brain, neurons showed very weak or negative staining. A similar staining pattern for transferrin was observed in the developing spinal cord and dorsal root ganglia (DRG). By 10-12 days, both spinal cord neurons and DRG neurons showed strong reactions for transferrin. Thereafter, the transferrin-positive reaction gradually diminished in older spinal cord neurons and completely disappeared from DRG neurons. Cultured cerebral hemisphere, spinal cord, and DRG neurons showed positive staining reactions for both transferrin and its receptor. Our results suggest that: transferrin is initially taken up by developing neurons from cerebrospinal fluid via receptor-mediated endocytosis; the accumulation of transferrin ultimately reaches a maximum level within immunoreactive neurons and then declines just prior to hatching; in contrast to other CNS neurons, DRG neurons accumulate transferrin only briefly and then become negative for transferrin by immunocytochemistry; and after closure of the blood-brain barrier, transferrin may reach neurons by transport across capillaries into the 'paravascular' spaces. In view of these results, transferrin may play some important but unrecognized role in early neuronal development in vivo as well as in vitro.

Effects of cell division, cell density, and cyclic nucleotides on choline acetyltransferase activity in a cholinergic neuroblastoma cell line (S-20Y)

We investigated the effects of a number of experimental perturbations on choline acetyltransferase (ChAT) in a cholinergic mouse neuroblastoma cell line (S-20Y). ChAT specific activity increased by 4.5-fold during growth, suggesting that enzyme activity is dependent on increased cell density. This was confirmed by assessing enzyme activity at differential initial seeding densities. ChAT activity was also markedly enhanced by 1 mM dibutyryl cyclic-3',5'-AMP (dBcAMP), an effect that was blocked by cycloheximide. Confirmation of the dBcAMP effect was achieved with forskolin, a compound known to enhance intracellular cyclic AMP; forskolin (100 microM) caused a significant increase in ChAT activity. After a 20-h latent interval ChAT activity was also enhanced significantly by cytosine arabinoside. The common element in these diverse effects on ChAT activity may be cessation of cell division, although cell-cell interactions at the level of the cell membrane may also be important in the control of ChAT in S-20Y.

Synthesis of the transferrin receptor by cultures of embryonic chicken spinal neurons

We have purified a glycoprotein from chicken sciatic nerves, sciatin, which has pronounced trophic effects on avian skeletal muscle cells in culture. Recent studies have shown that sciatin is identical to the iron-transport protein, transferrin, in terms of its physicochemical structure, immunological reactivity, and biological activity. To determine whether transferrin is synthesized and released by neuronal tissue, we incubated cultures of dissociated chicken spinal neurons in a medium free of L-leucine containing either L-3H-amino acids or L-[14C]leucine and immunoprecipitated transferrin with highly specific antibodies. The radiolabeled protein precipitated by rabbit heteroclonal, goat heteroclonal, or mouse monoclonal antitransferrin antibodies increased in specific activity in a linear manner for at least 30 min. Synthesis of this protein was abolished by the presence of puromycin (20 micrograms/ml) or cycloheximide (10(-5) M). The disappearance of the radiolabeled protein from cells was linear with a half-life (t 1/2) of 8-10 h. When immunoprecipitates were separated by SDS gel electrophoresis, a prominent band corresponding to transferrin (Mr 84,000) was visualized by staining with Coomassie Blue. However, when such gels were fluorographed, no radioactivity was apparent in the transferrin region of the gel although a prominent radioactive band was visualized at an Mr of 56,000. The protein of Mr 56,000 was not simply a degradation product of transferrin because this particular protein band was not generated by incubating radiolabeled transferrin with unlabeled neuronal homogenates. The protein of Mr 56,000 was purified from embryonic chicken brain and spinal cord by immunoabsorption chromatography on mouse monoclonal antitransferrin IgG conjugated to Sepharose 4B followed by affinity chromatography on immobilized transferrin. The purified protein bound radioiodinated transferrin and was precipitated by rabbit anti-chicken transferrin-receptor antibodies. Furthermore, this receptor protein was found to be localized on the plasma membrane of dorsal root ganglion neurons by immunocytochemistry using the peroxidase-antiperoxidase technique, and by blocking experiments, which showed that antitransferrin receptor IgG could inhibit the binding of fluorescein-conjugated transferrin at 4 degrees C to cultured neurons in vitro. From these data, we conclude that transferrin is not synthesized by cultures of chicken spinal cord neurons, but that the receptor for transferrin is synthesized by these cultures and is precipitated by antitransferrin antibodies as an antigen-receptor complex.

Sciatin is a transferrin-like polypeptide

Sciatin, an acidic glycoprotein from chicken sciatic nerve, the myotrophic effects on avian skeletal muscle cells in culture. As sciatin was found to have certain structural similarities to transferrin, we further investigated the physiochemical characteristics of sciatin in order to determine the relationship between these two proteins. Sciatin was found to be strikingly similar to ovotransferrin in amino acid composition. In addition, amino acid sequence analysis revealed that sciatin and ovotransferrin and identical amino-terminal sequences for a t least the first 20 amino acid residues. Chicken ovotransferrin, but not human serum transferrin, cross-reacted with rabbit antisciatin antibodies upon rocket immunoelectrophoresis and double immunodiffusion in agar. In addition, in the presence of bicarbonate, sciatin bound approximately 2 mol ferrous iron/mol protein. Using the purification procedure developed for sciatin, we purified a protein from chicken serum that cross-reacted with antisciatin serum, migrated at a position identical to that of sciatin or ovotransferrin on two-dimensional gel electrophoresis, had an amino composition very similar to ovotransferrin and sciatin, and had myotropic effects on cultured muscle cells. From these data, we conclude that sciatin is a growth-promoting polypeptide closely related in structure to transferrin.

Chicken serum transferrin duplicates the myotrophic effects of sciatin on cultured muscle cells

Sciatin, a glycoprotein purified from chicken sciatic nerves, has been shown to have trophic effects on chicken skeletal muscle cells in culture. Since we recently observed pronounced structural similarities between sciatin and chicken serum transferrin [Markelonis et al, 1982a], we decided to investigate the muscle growth-promoting activity of transferrin on cultured muscle cells. Serum transferrin was isolated by the same protocol used to purify sciatin, viz., affinity chromatography on concanavalin A-agarose followed by ion-exchange chromatography on DEAE cellulose. The serum protein recovered by this purification scheme was indistinguishable immunologically from sciatin as evidenced by a positive precipitin reaction against goat anti-sciatin serum on double immunodiffusion in agar. Purified serum transferrin had myotrophic effects identical to those of sciatin when added to skeletal muscle cells in vitro. For example, even when chicken embryo extract--a constituent normally required for chicken muscle cell differentiation in vitro--was omitted from culture medium, either serum transferrin or sciatin promoted myogenesis in culture as measured by a stimulation of the fusion index. Furthermore, both proteins caused a significant increase in the level of protein synthesis, the number of acetylcholine receptors and the activity of acetylcholinesterase in treated muscle cultures. By contrast, commercially obtained ovotransferrin (conalbumin) or FeSO4 (100 microM) were unable to fully support myogenesis of skeletal muscle in vitro if embryo extract was omitted from the culture medium. From these data, we conclude that the neuronal myotrophic protein sciatin is both structurally and biologically related to serum transferrin. Furthermore, we suggest that sciatin may represent a neuronal form of this iron-transport protein.

Sciatin: immunocytochemical localization of a myotrophic protein in chicken neural tissues

A mytotrophic protein (sciatin) purified from chicken sciatic nerves has "trophic" or "maintenance" effects on cultured muscle. We have elicited a specific antiserum against purified sciatin in rabbits. Using this antiserum, we investigated the distribution of sciatin in embryonic and adult chicken tissues by an unlabeled peroxidase-an-tiperoxidase method at the light microscopic level. The antiserum stained adult chicken neural tissues in situ and cultured embryonic chick neurons. Staining was intense in the cell bodies of spinal cord neurons and the axoplasm of sciatic nerves. These was reaction product seen in the outer margins of myelin sheaths that corresponded to the Schwann cell cytoplasm. Cerebral cortical neurons were weakly stained by the antiserum. No staining was apparent in oligodendrocytes or astrocytes. Nonneural tissues, such as skeletal, smooth and cardiac muscle, kidney, and liver, were also unstained by the antiserum. Cultured spinal cord neurons, cerebral cortical neurons, and sensory neurons were stained immunocytochemically by the antiserum. There was no reaction product seen in the glial cells that are usually present in neuronal cultures or cultured cells from liver, kidney, skeletal muscle, smooth muscle, and cardiac muscle. Our results thus show that the myotrophic protein is localized in neuronal perikarya and their processes in vivo as well as in vitro.

Purification of sciatin using affinity chromatography on concanavalin A-Agarose

A glycoprotein from chicken sciatic nerves, sciatin, has been shown to have trophic effects on the maturation and maintenance of skeletal muscle cells in culture. This protein was purified 24-fold from sciatic nerve extracts by affinity chromatography on concanavalin A-agarose followed by ion-exchange on diethylaminoethyl cellulose. The purity of sciatin obtained by this procedure was greater than 97% as estimated by densitometric integration of sodium dodecyl sulfate gels, and represented 33% of the sciatin present in sciatic nerve extracts as determined by rocket immunoelectrophoresis. Sciatin purified by this technique retained full biological activity since (1) addition of the protein to embryonic chicken skeletal muscle cells in culture enhanced the morphological development of the cells, and (2) the protein increased the number of acetylcholine receptors as measured by binding of 125I-alpha-bungarotoxin to 261% of the control value after 4 days in vitro. The purification procedure described in the present communication provides a more rapid and convenient method for the isolation of this trophic protein.

Dependence of in vitro myogenesis on a trophic protein present in chicken embryo extract

A trophic protein (sciatin) purified from sciatic nerves has been shown to enhance the morphological development and to promote the maintenance of skeletal muscle cells in vitro [Markelonis, G. J. & Oh, T. H. (1979) Proc. Natl. Acad. Sci. USA 76, 2470-2474]. We have elicited a specific antiserum against purified sciatin in rabbits. By using this antiserum, we have examined whether sciatin is also required for the initial differentiation of avian myogenic cells in vitro. Sciatin was found to be a component of chicken embryo extract, a constituent of culture medium required for myogenesis in vitro, by an immunodiffusion assay and by NaDodSO4 gel electrophoresis of immunoprecipitates. The removal of sciatin from chicken embryo extracts by immunoprecipitation with antiserum against sciatin completely inhibited myogenesis. When myogenic cells were grown in culture medium from which sciatin had been removed, the cells failed to differentiate beyond the myoblast stage. However, when sciatin (25 microgram/ml) was added to the sciatin-absorbed culture medium, normal myogenesis ensued. Furthermore, myogenic cells underwent normal myogenesis in the absence of embryo extract if sciatin (25 microgram/ml) was added to the culture medium. These results demonstrate that sciatin is the component of chicken embryo extract required for myogenesis and that the protein influences the initial differentiation of myogenic cells in vitro.

Sciatin: purification and characterization of a myotrophic protein from chicken sciatic nerves

A protein isolated from sciatic nerves of adult chickens promotes the morphological maturation and maintenance of embryonic avian skeletal muscle cells in the absence of innervation and is required for normal myogenesis in vitro. This trophic protein, sciatin, has been purified by ion exchange column chromatography on DEAE-cellulose followed by gel filtration on Sephadex G-100. Sciatin migrated as a single polypeptide chain of molecular weight 84,000 on sodium dodecyl sulfategel electrophoresis. The native molecular weight of sciatin as determined by sedimentation equilibrium centrifugation was 86,400. Amino acid analysis revealed that sciatin is relatively deficient in tryptophan, histidine, glycine, and arginine, but enriched in cysteine, methionine, alanine, and lysine. Carbohydrate determination showed that sciatin in composed of 11% sugar by weight with no detectable N-acetylneuraminic acid residues. Sedimentation velocity centrifugation studies revealed an S20,w0 of 5.11 with a frictional coefficient of 1.31. Sciatin had no detectable protease or acetylcholinesterase activity. The results of the present study provide new biochemical information on a macromolecule with biological activities similar to those expressed by the "maintenance" group of growth factors which includes such proteins as nerve growth factor.

Neurotrophic protein regulates muscle acetylcholinesterase in culture

Skeletal muscles lose acetylcholinesterase in culture as a result of denervation. A protein fraction isolated from peripheral nerves maintained the level of acetylcholinesterase in cultures of aneural embryonic muscle or denervated adult chicken muscle. These results indicate that trophic regulation of muscle acetylcholinesterase might be mediated by a protein produced by nerves.