Search

Tau proteins are microtubule-associated proteins that promote microtubule polymerization in vitro and in vivo. They are a family of neuronal proteins with apparent molecular weights in the range 50,000-68,000 determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Recently, a new member of this family has been described and its cDNA has been cloned. It has an apparent molecular weight of 116,000 and has been called high-molecular-weight tau (HMW tau). All the tau proteins are encoded by a single gene, which undergoes complex alternative splicing. In the present study, we have cloned into the baculovirus a cDNA fully encoding HMW tau as well as a truncated cDNA encoding a protein beginning 13 amino acids in front of the tau microtubule-binding domain. HMW tau-recombinant-virus-infected Sf9 cells overexpressed HMW tau, which induced the polymerization of microtubules and the formation of long cellular processes similar to those induced by low-molecular-weight tau (LMW tau) overexpression. Process cross sections revealed a larger spacing (approximately 35 nm) between microtubules when induced by HMW tau than when induced by LMW tau (approximately 20 nm). The truncated construct also induces processes, where microtubules were packed far more closely together (approximately 10 nm). Although branching did not occur in processes induced by intact tau S, 10% of the processes induced by the truncated tau protein branched.

beta-Amyloid (A beta) has been strongly implicated in the pathophysiology of Alzheimer's disease (AD), but the means by which the aggregated form of this molecule induces neuronal death have not been fully defined. Here, we examine the role of the c-Jun N-terminal kinases (JNKs) and of their substrate, c-Jun, in the death of cultured neuronal PC12 cells and sympathetic neurons evoked by exposure to aggregated A beta. The activities of JNK family members increased in neuronal PC12 cells within 2 h of A beta treatment and reached 3--4-fold elevation by 6 h. To test the role of these changes in death caused by A beta, we examined the effects of CEP-1347 (KT7515), an indolocarbazole that selectively blocks JNK activation. Inclusion of CEP-1347 (100--300 nM) in the culture medium effectively blocked the increases in cellular JNK activity caused by A beta and, at similar concentrations, protected both PC12 cells and sympathetic neurons from A beta-evoked-death. Effective protection required addition of CEP-1347 within 2 h of A beta treatment, indicating that the JNK pathway acts relatively proximally and as a trigger in the death mechanism. A dominant-negative c-Jun construct also conferred protection from A beta-evoked death, supporting a model in which JNK activation contributes to death via activation of c-Jun. Finally, CEP-1347 blocked A beta-stimulated activation of caspase-2 and -3, placing these downstream of JNK activation. These observations implicate the JNK pathway as a required element in death evoked by A beta and hence identify it as a potential therapeutic target in AD.

We have studied the heterogeneity of the microtubule-associated tau proteins using tau-specific antibodies and two-dimensional electrophoresis. Both monoclonal and polyclonal antibodies to tau proteins recognize five bands in cow brain microtubule proteins run on sodium dodecyl sulfate (SDS)-polyacrylamide gels, with apparent molecular weights between 56,000 and 66,000. Immunoblots of cow brain microtubules separated on two-dimensional gels, using nonequilibrium pH gradient electrophoresis in the first dimension and SDS-gel electrophoresis in the second, reveal that greater than 30 isoforms of tau exist. The tau proteins vary in pI from 6.5 to 8.5, with the higher-molecular-weight forms being more acidic. The microheterogeneity of tau is not induced by cycling of microtubules, because two-dimensional immunoblots of tau from total brain are almost identical to those of tau from cycled tubules. Adult rat brain tau, which appears as three doublet bands on SDS gels, also exhibits considerable isoelectric heterogeneity, as does tau from 7-day-old rats, which appears as only one band on SDS gels. After dephosphorylation of cow brain tau with alkaline phosphatase, the highest-molecular-weight band disappears on SDS gels. On two-dimensional gels, the number of tau variants decreases by more than half after dephosphorylation, and the more basic species increase greatly in intensity. Preliminary experiments with tau labeled in vivo with 32PO4 also indicate that the more acidic tau proteins are the more highly phosphorylated forms. Thus, isoelectric heterogeneity of tau proteins exists at all ages and is due, at least in part to differences in the state of phosphorylation of tau isoforms.

Studies were carried out on the glycoproteins (GPs) released by cultured rat sympathetic neurons and by cultured PC12 rat pheochromocytoma cells with and without nerve growth factor (NGF) treatment. Cultures were prelabeled with [3H]fucose and then incubated for 4-8 h in fresh unlabeled medium. The material released into the medium was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and fluorography. The patterns of labeled material released by all three types of cultures were similar. One of the major components released was of apparent Mr less than or equal to 230,000. Another major component of apparent Mr = 55,000 as well as minor components of apparent Mr less than or equal to 180,000, 140,000, 118,000, and 105,000 were also detected. An additional peptide of apparent Mr less than or equal to 210,000 was released only by the sympathetic neurons. The soluble released Mr less than or equal to 230,000 component appeared to be derived from a previously characterized neuronal integral membrane GP referred to as the NILE (NGF-inducible large external) GP. Evidence for this included recognition of the released component by a monospecific antiserum prepared against membrane-derived NILE GP. At least several of the other released GPs appeared to be derived from membrane-bound components with which they share immuno-crossreactivity. Since the soluble NILE and other released GPs had somewhat faster mobilities on SDS-polyacrylamide gels than their apparent membrane-bound correspondents, release could either be due to, or accompanied by, minor changes in molecular structure.

Exposure of PC12 cells to nerve growth factor (NGF) has been shown to induce an rnRNA that encodes a novel neuronal intermediate filament protein. The findings presented here concern the identity of this filament protein. The major protein in NGF-treated PC12 cell cytoskeletons derived by extraction with 1% Triton X-100 is of apparent Mr= 58,000, focuses by isoelectric focusing as several closely spaced spots of pl 5.6–5.8, and is elevated relative to non-NGF-treated cells. Partial microsequencing of this material reveals 2 internal sequences that are identical to a 14-residue sequence encoded by the NGF-regulated clone 73 mRNA, but not to sequences of other known proteins. An antiserum raised against a 19-residue synthetic peptide corresponding to the deduced C-terminus of the protein encoded by the NGF-regulated clone 73 mRNA specifically recognizes the 58,000-Mr protein. Properties of the 58-kilodalton protein strongly suggest that it corresponds to an intermediate filament protein (peripherin) previously identified in PC12 cells and in peripheral and certain CNS neurons. Identification of the intermediate filament protein encoded by an NGF-induced message should facilitate studies of its regulation and function.

Intact neurofilaments (NF) purified from mammalian brain and spinal cord promote the assembly of microtubules in solutions of pure phosphocellulose (PC)-purified tubulin. This assembly is temperature-dependent and is inhibited by mitotic spindle inhibitors. The ability of NF to induce microtubule formation is 20% of that of purified microtubule-associated proteins (MAPs), whereas MAPs comprise less than 5% of the protein in the NF preparations. The inducing activity of NF is rapidly lost on boiling. When intact NF are incubated with PC-tubulin and then centrifuged, tubulin is sedimented together with the filaments. This association is inhibited by colchicine and podophyllotoxin and is cold-sensitive. NF purified to homogeneity under denaturing conditions and then reassembled completely lack the ability to promote the assembly of PC-tubulin or to bind tubulin on a centrifugation assay. No MAPs are present in these preparations, though these filaments have the ability to bind exogenous MAPs. While these experiments do not rule out an intrinsic microtubule-assembly-promoting activity, they suggest that this activity is due to nontriplet proteins in the preparation, most likely filament-associated MAPs.

The nerve growth factor (NGF)-induced large external (NILE) glycoprotein is an NGF-inducible surface component of PC 12 cells that is also widely distributed in the nervous system. It has recently been shown to be indistinguishable from the high-molecular-weight species of the brain antigens L1 and neuron-glia cell adhesion molecule (Ng-CAM) and may have a function in regulating cell adhesion in the developing nervous system. We have used polyclonal anti-NILE antisera to screen a λgt11 cDNA library made from NGF-treated PC12 cells. Four molecular probes have been isolated that encode parts of the apoprotein, related proteins, or both. These probes are 1,500, 800, 330, and 300 base pairs long, respectively, and in Northern blots they recognize a family of messages having lengths of ∼5.9, 3.4, 2.4, and 1.9 kilobases. The two smaller messages are modestly but reproducibly up-regulated by NGF in PC 12 cells, as is the NILE glycoprotein; however, only the two larger species would appear to be large enough to encode it. These messages are prominent in brain but not in nonneuraltissues, in accordance with the observed levels of the protein. The recombinant phages produce fusion proteins that share specific epitopes with the NILE glycoprotein but not with other proteins. In these experiments, filters coated with recombinant fusion protein were prepared. Antibodies bound to and eluted from these filters specifically immunoprecipitate NILE glycoprotein, but not other proteins, from PC12 cell extracts. Other activities present in the original sera are not specifically retained by recombinant fusion proteins, and Escherichia coll lysates made with nonrecombinant λgt11 do not absorb the anti-NILE activity. The probes described here will be of use in studying the function, biochemistry, and molecular biology of the NILE glycoprotein and related species.