Faculty Bibliography

Neurofilament proteins are continuously modified during their lifetime by a succession of protein kinases and phosphatases. Site-specific phosphorylation or dephosphorylation within different polypeptide domains of each neurofilament subunit is now believed to regulate such behaviors of neurofilaments as subunit polymerization and exchange, axonal transport, interactions with other cytoskeletal proteins and degradation. Local regulation of phosphorylation events could account for variations in the size, morphology and dynamics of the neurofilament network in different regions of the neuron. The apparent greater plasticity of the neurofilament network in regions like the perikaryon, initial segment and nodes along the axon may provide some insight into the vulnerability of these regions in neurofibrillary disease

Disturbed lysosomal function may be implicated at several stages of Alzheimer's pathogenesis. Lysosomes and acid hydrolases accumulate in the majority of neocortical pyramidal neurons before typical degenerative changes can be detected, indicating that altered lysosome function is among the earliest markers of metabolic dysfunction in Alzheimer's disease. These early alterations could reflect accelerated membrane and protein turnover, defective lysosome or hydrolase function, abnormal lysosomal trafficking or any combination of these possibilities. Because APP is partly metabolized in lysosomes, early disturbances in lysosomal function could promote the production of abnormal and/or neurotoxic APP fragments within intact cells. Lysosomal abnormalities progressively worsen as neurons begin to degenerate. Based on existing literature on cell death, increased perturbation and instability of the lysosomal system may be expected to contribute to the atrophy and eventual lysis of the neuron. Finally, the release of hydrolase-filled lysosomes and lipofuscin aggregates from dying neurons accounts for the abundant deposition of enzymatically active acid hydrolases of all classes in the extracellular space--a phenomenon that may be unique to Alzheimer's disease. Acting on APP present in surrounding dystrophic neurites, cellular debris and astrocyte processes, dysregulated hydrolases may cleave APP in atypical sequential patterns, thereby generating self-aggregating protease-resistant APP fragments that can be only processed to beta-amyloid. Genetic mutations or posttranslational factors of APP should further enhance the generation of amyloidogenic fragments by a dysregulated lysosomal system. Given that very little, if any, beta-amyloid is detected intracellularly, yet extracellular beta-amyloid is very abundant, our data suggest that the final steps of APP processing and the generation of most beta-amyloid in the brain parenchyma occur extracellularly and may involve one or more lysosomal proteases

The respective roles of cAMP-dependent protein kinase (protein kinase A [PKA]) and protein kinase C (PKC) in the early stages of neurite outgrowth were examined in SH-SY-5Y human neuroblastoma cells. Forskolin or dbcAMP, agents that increase intracellular cAMP levels, and intracellular delivery of PKA catalytic subunit induced neurite outgrowth. The PKA inhibitor, N-(2-guanidinoethyl)-5-isoquinolinesulfonamide (HA 1004), prevented the increases, and decreased further the percentage of cells possessing short, filopodia-like neurites in the absence of inducers. In contrast to effects on PKA activation, PKC activation by 12-0-tetradecanoylphorbol-13-acetate (TPA) reduced the percentage of filopodia-like neurites elaborated by otherwise untreated cells, and prevented neurite outgrowth induced by PKA activators. PKC inhibitors 1-(5-isoquinolinesulfonyl)-2-methylpiperazine dihydrochloride (H7), staurosporine, and sphingosine induced neurite outgrowth. Neurites induced by PKA activation contained higher levels of tubulin immunoreactivity than those induced by PKC inhibition. Furthermore, PKA-induced neurites rapidly retracted in the presence of colchicine, while those elaborated following PKC inhibition were more resistant. These data suggest that neurites elaborated in response to PKA activation are dependent upon microtubule polymerization, and that neurite induction following PKC inhibition is mediated by a different mechanism. PKA activators and PKC inhibitors exerted additive effects on neurite outgrowth, suggesting that the distinct pathways regulated by these two kinases function cooperatively during neuritogenesis

NB2a/d1 neuroblastoma cells constitutively express multiple isoforms of the microtubule-associated protein tau and incorporate this protein into the axonal neurites elaborated during serum deprivation. To examine whether or not tau played an essential role in axonal outgrowth, cells cultured in serum-free medium were treated at 24 h intervals with antisense- and sense-oriented cDNA oligonucleotides (25 or 36 mers that span or are upstream of tau initiation codon) and were simultaneously serum deprived. Oligonucleotide uptake was confirmed by determination of intracellular levels of radiolabeled oligonucleotides. Treatment for 48 h with tau antisense oligonucleotides reversibly inhibited the expression of tau and the number of neurite-bearing cells compared with treatment with sense oligonucleotides. By contrast, tubulin expression was not affected. When cells were treated with antisense oligonucleotide simultaneously with serum deprivation, the initial outgrowth of neurites was unaffected, but continued neurite elongation was prevented. By contrast, neurite outgrowth at 4 h was inhibited when cells were pretreated with tau antisense 24 h before serum deprivation. Furthermore, intracellular delivery of anti-tau antiserum prevented neurite outgrowth and, in cells that had previously been deprived of serum for 24 h, induced retraction of existing neurites. These findings indicate that both the initiation and the continued outgrowth of neurites are dependent on tau and that pre-existing cytoplasmic pools of tau can mediate initial neuritogenesis

The millimolar form of calcium-activated neutral proteinase (mCANP) is generally regarded as a cytosolic enzyme in nonneuronal systems, although its subcellular localization in brain is less well established. To resolve conflicting reports on the localization of mCANP based on activity measurements, we developed an immunoassay for CANP and compared the content and activity of the molecule in soluble and membrane fractions of mouse and human brain. Western blot immunoassays, using two different antibodies specific for mCANP, demonstrated that mCANP content is 4.5 ng/g in human or mouse brain, about 0.0005% of the total protein. More than 95% of the total immunoreactive mCANP remained in the soluble fraction after 15,000 g centrifugation of the whole homogenate. mCANP activity was determined with [14C]azocasein as substrate after removing endogenous CANP inhibitor(s) by ion-exchange chromatography on DEAE-cellulose. Caseinolytic activity was detected only in fractions derived from the supernatant extract. The distribution of mCANP content and enzyme activity were unchanged when tissues were extracted with different concentrations of Triton X-100. These findings establish the usefulness and validity of the CANP immunoassay and demonstrate that mCANP in mouse and human brain is localized predominantly within the cytosol

Certain biological actions of phorbol esters cannot be duplicated by diacylglycerol (DAG). Thus, the human neuroblastoma cell line SH-SY5Y differentiates when exposed to 12-tetradecanoyl-13-acetyl-beta-phorbol (TPA) and protein kinase C (PKC) inhibitors, but not when exposed to DAG. To investigate the specific features of the phorbol diester molecule that might be responsible for these effects, we examined the extension of neurites, expression of neuron-specific enolase, and appearance and localization of phosphorylated high molecular weight neurofilament subunits (NF-H). TPA, 12-deoxy-13-tetradecanoyl-beta-phorbol, and staurosporine, but not DAG or 4-O-methyl-TPA, caused neurite outgrowth. Neuron-specific enolase was expressed in cells treated with TPA and 12-deoxy-13-tetradecanoyl-beta-phorbol but not with DAG, staurosporine, or 4-O-methyl-TPA. NF-H increased in the perikarya of cells treated with DAG and 4-O-methyl-TPA, in processes and to varying degrees in perikarya of TPA- and 12-deoxy-13-tetradecanoyl-beta-phorbol-treated cells, but much more in the processes than in the perikarya of staurosporine-differentiated cells. These findings and additional differences between the differentiation induced by TPA (a PKC activator) and staurosporine (a PKC inhibitor), including distinct morphology of the cell body and processes and time of appearance of the morphological phenotype, suggest that activators and inhibitors of PKC induce differentiation of SH-SY5Y cells by different mechanisms, and that the five-membered/seven-membered terpene ring region present in TPA must be intact for the induction of morphological differentiation

Exposure of each of the three neurofilament proteins (NFPs) to AlCl3 resulted in their failure to migrate into sodium dodecyl sulfate (SDS)-containing gels. This effect was dependent on length of incubation (minimum, 2 h) and AlCl3 concentrations (minimum, 50 microM) and was not reversed by 20% SDS, 6 M urea, freeze-thawing, boiling, or extensive dialysis. The migration of vimentin and glial fibrillary acidic protein was not affected by AlCl3. The high-molecular-weight neurofilament subunit (NF-H) entered SDS-containing gels after exposure to aluminum lactate but migrated aberrantly as a long high-molecular-weight streak. Migration of the 160-kDa alpha-chymotryptic cleavage product of NF-H, which contains the higher phosphorylated tail domain, was also prevented from migrating into SDS-containing gels by AlCl3. Dephosphorylation of NF-H and the middle-molecular-weight neurofilament subunit (NF-M) eliminated these effects on gel migration. EDTA, EGTA, MgCl2, CaCl2, or FeCl3 had no effect on NF-H or NF-M migration; furthermore, preincubation with, or simultaneous exposure to, CaCl2 or FeCl3 did not alter the effect of AlCl3. One interpretation of these results is that Al3+ interacts with phosphate groups on extensively phosphorylated C-terminal sidearms of NFPs, resulting in intermolecular cross-linking. These findings demonstrate a direct effect of aluminum on NFPs and provide a possible mechanism for neurofilament accumulation in perikarya during aluminum intoxication

Astroglial-conditioned medium (GCM) induced two distinct, but intimately related, phases of neuritogenesis in NB2a/d1 neuroblastoma cells--a 'rapid-outgrowth,' unstable phase, and a delayed, relatively stable phase, which are apparently regulated by glial-derived protease inhibitors and laminin, respectively. The initial rapid outgrowth (less than 4 hr) may be mediated by inhibition of a thrombin-like protease, present as a serum component and/or adsorbed to the outer neuronal surface, since (1) a similar effect was obtained by serum removal or by adding the specific thrombin inhibitor, hirudin; (2) exogenous thrombin inhibited the rapid outgrowth of neurites by GCM; and (3) cell-free enzyme assays confirmed the presence of thrombin-inhibitory activity in GCM. Although neurites induced by removal of serum removal or hirudin addition are rapidly resorbed following serum replenishment or hirudin depletion, GCM-induced neurites continued to elongate after GCM removal, indicating that GCM contained additional neurite-promoting factors. Anti-laminin antiserum did not inhibit the initial elaboration of neurites by GCM but prevented their continued elongation. Anti-laminin antiserum had no affect on neurite outgrowth induced by serum deprivation. The more protracted, second phase of neurite outgrowth could also be achieved by the addition of soluble purified laminin to undifferentiated cells. Unlike neurites at 4 hr, neurites at 24 hr were no longer dependent on the protease inhibitors in GCM, since exogenous thrombin no longer caused them to retract. Simultaneous addition of thrombin and anti-laminin antiserum with GCM had identical inhibitory effects on continued neurite elaboration at 24 hr as did anti-laminin antiserum without thrombin.(ABSTRACT TRUNCATED AT 250 WORDS)

New studies provide further evidence that the neuronal cytoskeleton is the product of a dynamic interplay between axonal transport processes and locally regulated assembly mechanisms. These data confirm that the axonal cytoskeleton in mammalian systems is largely stationary and is maintained by a smaller pool of moving subunits or polymers. Slow axonal transport in certain lower species, however, may exhibit quite different features