Protein kinase FA/glycogen synthase kinase-3 alpha after heparin potentiation phosphorylates tau on sites abnormally phosphorylated in Alzheimer's disease brain

Previously, we identified protein kinase FA/glycogen synthase kinase-3 alpha (GSK-3 alpha) as a brain microtubule-associated tau kinase that phosphorylates Ser235 and Ser404 of tau and causes its electrophoretic mobility shift in gels, a unique property characteristic of paired helical filament-associated pathological tau (PHF-tau) in Alzheimer's disease brains. In this study, we found that the activity of kinase FA/GSK-3 alpha towards phosphorylation of brain tau could be stimulated approximately fourfold by heparin. The phosphorylation molar ratio was increased simultaneously up to 9 mol of phosphates/mol of tau, resulting in a reduced mobility of tau with an apparent molecular mass shift to approximately 68 kDa in sodium dodecyl sulfate gels, which is very similar to that observed in Alzheimer-tau. Tryptic digestion of 32P-labelled tau, followed by HPLC and two-dimensional separation on TLC cellulose plates, revealed eight major phosphopeptides. Phosphoamino acid analysis together with sequential manual Edman degradation and protein sequence analysis further revealed that, in addition to Ser235 and Ser404, heparin generated Thr212, Thr231, Ser262, Ser324, and Ser356, the five extra phosphorylation sites in tau. As Ser235, Ser262, Ser324, Ser356, and Ser404 (particularly the site of Ser262) have been identified as five of the most potent sites in tau responsible for reducing microtubule binding possibly involved in neuronal degeneration, and Thr231, Ser235, Ser262, and Ser404 are four of the most well documented sites abnormally phosphorylated in Alzheimer-tau, the results provide initial evidence that protein kinase FA/GSK-3 alpha after heparin potentiation may represent one of the most potent systems possibly involved in the abnormal phosphorylation of PHF-tau in Alzheimer's disease brains.

Structure, regional and developmental expression of rat MAP2d, a MAP2 splice variant encoding four microtubule-binding domains

MAP2, a major component of microtubule polymers in neurons consists of high molecular weight (HMW) proteins MAP2a, MAP2b and a low molecular weight (LMW) MAP2c, expressed in the developing brain. These isoforms are produced from a single gene by alternative splicing and share identical C-termini encompassing 3 tandem repeats, critical in microtubule binding. We describe the structure, regional and developmental expression of a novel MAP2 splice variant, MAP2d, containing an insertion whose sequence is homologous to the three and four repeats of MAP2 and Tau respectively. This insertion is absent from the mRNAs encoding HMW MAP2. MAP2d mRNAs are expressed at higher levels than MAP2c in all adult nervous tissues of the rat, and are found at low levels in glial cell cultures when compared to primary cultures of cerebellar neurons. Splicing of the fourth repeat in mature Tau precedes that in MAP2d during rat brain development. The tardive expression of a four microtubule-binding domain LMW MAP2 suggests it could play in extended neurites a similar role as mature Tau in axons.

Biopsy-derived adult human brain tau is phosphorylated at many of the same sites as Alzheimer's disease paired helical filament tau

Tau from Alzheimer's disease (AD) paired helical filaments (PHF-tau) is phosphorylated at sites not found in autopsy-derived adult tau from normal human brains, and this suggested that PHF-tau is abnormally phosphorylated. To explore this hypothesis, we examined human adult tau from brain biopsies and demonstrated that biopsy-derived tau is phosphorylated at most sites thought to be abnormally phosphorylated in PHF-tau. These sites also were phosphorylated in autopsy-derived human fetal tau and rapidly processed rat tau. The hypophosphorylation of autopsy-derived adult human tau is due to rapid dephosphorylation postmortem, and protein phosphatases 2A (PP2A) and 2B (PP2B) in human brain biopsies dephosphorylate tau in a site-specific manner. The down-regulation of phosphatases (i.e., PP2A and PP2B) in the AD brain could lead to the generation of maximally phosphorylated PHF-tau that does not bind microtubules and aggregates as PHFs in neurofibrillary tangles and dystrophic neurites.

Structural studies of tau protein and Alzheimer paired helical filaments show no evidence for beta-structure

We have studied the conformation of tau protein and Alzheimer paired helical filaments (PHF) by several spectroscopic, scattering, and imaging methods revealing the overall shape and the conformation of the polypeptide backbone. Tau protein behaves in solution as if it were denatured; no evidence for compact folding was detected. The protein is highly extended, there is no defined radius of gyration, and the scattering is best described by that of a random ("Gaussian") polymer. CD and Fourier transform infrared spectroscopy show only a minimal content of ordered secondary structure (alpha-helix or beta-sheet). Similarly, PHFs from Alzheimer brain tissue show no detectable secondary structure by x-ray diffraction or spectroscopy. It is thus unlikely that the aggregation of tau into Alzheimer PHFs is based on interactions between strands of beta-sheets (a model currently favored for other disease-related polymers such as beta-amyloid fibers of Alzheimer's disease).

Senile dementia of Lewy body type and Alzheimer type are biochemically distinct in terms of paired helical filaments and hyperphosphorylated tau protein

We have used biochemical assays to examine cingulate and occipital cortices from age-matched cases of Alzheimer's disease (AD; n = 12), senile dementia of the Lewy body type (SDLT; n = 13), Parkinson's disease (PD; 5 non-demented cases and 7 cognitively impaired cases) and controls (n = 11) for paired helical filaments (PHFs), phosphorylated and normal tau protein and beta/A4-protein. Whereas cingulate cortex is characterised by relatively high densities of cortical Lewy bodies in the SDLT cases and lower numbers in PD, these inclusion bodies were absent in the cingulate cortex from AD and control cases. Protease-resistant PHFs and hyperphosphorylated tau protein were found in AD and, at low levels, in a minority of SDLT cases. Qualitatively, both of these preparations were indistinguishable in SDLT from those found in AD but levels of both parameters in SDLT were less than 5% of those in AD. SDLT, PD and control groups did not differ from each other in terms of the quantity of protease-resistant PHFs or the level of hyperphosphorylated tau. Furthermore, PHF accumulation did not distinguish between PD cases with or without dementia. The levels of normal tau protein did not differ between the four groups. beta/A4 protein levels did not distinguish between PD and control groups, between AD and SDLT groups, or between SDLT and control groups for either cingulate or occipital cortices. Thus extensive accumulation of PHFs in either neurofibrillary tangles or dystrophic neurites is not a feature of either SDLT or PD. Our findings provide molecular support for the neuropathological and clinical separation of SDLT as a form of dementia that is distinct from AD.

Analysis of microtubule-associated protein tau glycation in paired helical filaments

Alzheimer's disease is typified by the characteristic histopathological lesions of neurofibrillar plaques and tangles. The latter are composed of paired helical filaments (PHFs), the major components of which are modified forms of the microtubule-associated protein tau. The exact nature of these modifications remains unknown, although the presence of hyperphosphorylated tau in PHFs argues strongly that phosphorylation is one of the modifications that result in the polymerization of tau into PHFs. However, hyperphosphorylation alone is insufficient to explain the formation of PHFs. In an attempt to characterize other post-translational modifications of PHF-tau, we have analyzed its glycation. A fraction of PHF-tau seems to be glycated in vivo, whereas soluble tau from either Alzheimer's disease or non-demented human brain is not glycated at all. Purified tau from bovine brain can be efficiently glycated in vitro. Tau glycation is accompanied by a decrease in the tau binding to tubulin. These results support the view that glycation may be one of the modifications hampering the binding of tau to tubulin in Alzheimer's disease, thus facilitating tau aggregation into PHFs.

Complete sequence of 3'-untranslated region of Tau from rat central nervous system. Implications for mRNA heterogeneity

Tau is a family of microtubule associated proteins, heterogeneous in molecular weights, which are expressed specifically in neurons. Tau is encoded by a single gene, while its transcript undergoes a complex and regulated alternative splicing, giving rise to several mRNA species that migrate on Northern blots at approximately 6 and 2 kb. In this report we characterize a full size transcript of tau mRNA from rat brain and demonstrate that it contains 5203 nucleotides (not including exon 2 and 3), which correlates well with the exact size of the transcript as analyzed by Northern blot using RNA standard size markers. The full length of the 3'-untranslated region contains 3848 bp and includes two polyadenylation signals which may yield the two size transcripts in the central nervous system. The first polyadenylation signal is located in the retained intron 13/14 and the second polyadenylation signal is 19 nucleotides before the poly(A) tail. Unspliced intron 13/14 was detected in all RNA preparations tested, including RNA from different ages and different regions of rat brain, RNA from dorsal root ganglia and from undifferentiated and differentiated PC12 cells. In none of the above tissues and cells was a spliced transcript lacking intron 13/14 detected.

PHF-tau from Alzheimer's brain comprises four species on SDS-PAGE which can be mimicked by in vitro phosphorylation of human brain tau by glycogen synthase kinase-3 beta

Extensive in vitro phosphorylation of a purified preparation of control human brain tau consistently produces four rather than, as previously believed, three tau species on SDS-PAGE. The species thus generated are shifted on SDS-PAGE to positions that match those of PHF-tau isolated from Alzheimer's disease brain. A mixture of recombinant human tau isoforms phosphorylated by GSK-3 beta gave similar results to those obtained with control human brain tau. In vitro phosphorylation of the individual recombinant isoforms by GSK-3 beta showed that the four bands of PHF-tau are likely to consist of isoforms 3R,0 alone; 4R,0 with 3R,29; 4R,29 with 3R,58 and 4R,58 alone.

Epitope mapping of monoclonal antibodies to the paired helical filaments of Alzheimer's disease: identification of phosphorylation sites in tau protein

Tau is a neuronal phosphoprotein the expression of which is developmentally regulated. A single tau isoform is expressed in fetal human brain but six isoforms are expressed in adult human brain, with the fetal isoform corresponding to the shortest adult isoform. Phosphorylation is also developmentally regulated, as fetal tau is phosphorylated at more sites than adult tau. In Alzheimer's disease, the six adult tau isoforms become hyperphosphorylated and form the paired helical filament (PHF), the major fibrous component of the neurofibrillary lesions. One way to identify phosphorylated sites in tau is to use antibodies that recognize phosphorylated residues within a specific amino acid sequence. We here characterize the two novel phosphorylation-dependent anti-tau antibodies AT270 and AT180 and identify their epitopes as containing phosphorylated Thr-181 and Thr-231 respectively. With these antibodies we show that these two threonine residues are partially phosphorylated in fetal and adult tau and almost fully phosphorylated in PHF tau. This result contrasts with previous studies of Ser-202 and Ser-396 which are partially phosphorylated in fetal tau, unphosphorylated in adult tau but almost fully phosphorylated in PHF tau.

Characterization of a shared epitope in cortical Lewy body fibrils and Alzheimer paired helical filaments

The straight fibrils of the Lewy body contain an epitope related to phosphorylation of the KSPV motif common to the C termini of the 200- and 170-kDa neurofilament subunits and tau. To further characterize this phosphorylated neurofilament/tau epitope in Lewy bodies and to analyze the constituents of isolated Lewy bodies we used a combined biochemical and immunochemical approach. In formalin-fixed paraffin-embedded tissue cortical Lewy bodies were labelled by monoclonal antibodies directed to phosphorylation-dependent KSPV epitopes in the sequences of neurofilament and phosphorylation-independent epitopes. Immunoblotting of solubilized Lewy body fibrils with the same antibodies which stained Lewy bodies in tissue sections revealed that the immunoreactive Lewy body proteins were phosphorylated neurofilament subunits. An antibody to the 68-kDa neurofilament subunit labelled Lewy bodies and Lewy body protein at 50-68 kDa. We conclude that the shared phosphorylated epitope in Lewy body fibrils and paired helical filaments is related to the common KSPV sequence in neurofilament and tau, and that all three neurofilament subunits are present in the Lewy body. This result indicates that although Lewy bodies and neurofibrillary tangles share epitopes they are comprised of distinct structural subunits.

Truncated tau protein as a new marker for Alzheimer's disease

Molecular analysis of histological hallmarks (neurofibrillary tangles, neuritic plaques, neuropil threads and dystrophic neurites) of Alzheimer diseased brain tissues showed that these lesions contain paired helical filaments. Their major constituent is microtubule associated protein tau that is in an abnormally hyperphosphorylated and truncated state. These diseased forms of tau protein are unable to promote full microtubule assembly. Understanding of the molecular basis of the processes leading to the modifications of tau proteins and paired helical filament formation will form a firm step toward rational drug development and the cure of the Alzheimer's disease.

Removal of the projection domain of microtubule-associated protein 2 alters its interaction with tubulin

Microtubule-associated proteins (MAPs) can promote microtubule assembly in vitro. One of these MAPs (MAP2) consists of a short promoter domain which binds to the microtubule and promotes assembly and a long projection domain which projects out from the microtubule and may interact with other cytoskeletal elements. We have previously shown that MAP2 and another MAP, tau, differ in their interactions with tubulin in that tau, but not MAP2, promotes extensive aggregation of tubulin into spiral clusters in the presence of vinblastine and that microtuubles formed with MAP2 are more resistant than those formed with tau to the antimitotic drug maytansine [Luduena, R. F., et al. (1984), J. Biol. Chem. 259, 12890-12898; Fellous, A., et al. (1985), Cancer Res. 45, 5004-5010]. Here we have used chymotryptic digestion to remove the projection domain of MAP2 and examined the interaction of the digested MAP2 (ctMAP2) with tubulin in the presence of vinblastine and maytansine. We have found that ctMAP2 behaves very much like tau, but not like undigested MAP2, in the presence of vinblastine, in that ctMAP2 causes tubulin to polymerize into large clusters of spirals. In contrast, microtubule assembly in the presence of ctMAP2 is much more resistant to maytansine inhibition than is assembly in the presence of tau or undigested MAP2. Our results suggest that the projection domain of MAP2 may play a role in the interaction of tubulin with MAP2 during microtubule assembly.

Differential effect of phosphorylation and substrate modulation on tau's ability to promote microtubule growth and nucleation

The neuronal microtubule-associated protein tau promotes microtubule assembly and has been implicated in the development of axonal morphology. To study the effect of phosphorylation and substrate modulation on tau's distinct activities to promote growth of existing microtubules and nucleation of new ones, we phosphorylated bacterially expressed human tau by cAMP-dependent protein kinase in the absence or presence of heparin, an acidic substrate modulator. We found that heparin increased phosphorylation of tau by a factor of more than 2 and produced tau bands with decreased electrophoretic mobility. We demonstrate that phosphorylation of tau in the absence or presence of heparin similarly reduced tau's activity to promote microtubule growth, whereas tau's activity to promote microtubules was suppressed much more after phosphorylation in the presence of heparin. Using recombinant tau fragments we showed that heparin-induced phosphorylation caused a specific shift in electrophoretic mobility indicative of a change in tau's conformation. By aminoterminal sequencing of a tau fragment starting at residue 154 we provide evidence that phosphorylation of serine 156 is responsible for this mobility shift and for the effect on tau's nucleation activity. We conclude that tau's activities to promote growth of existing microtubules and nucleation of new ones are differentially affected by the phosphorylation of specific tau residues. Regulation of the phosphorylation state by substrate modulation may play an important role in regulating tau's function.

Microtubule-associated protein tau is hyperphosphorylated during mitosis in the human neuroblastoma cell line SH-SY5Y

A phosphorylated tau epitope specific for paired helical filaments in Alzheimer's disease is recognized by monoclonal antibody PHF-1. Healthy adult brains lack the PHF-1 epitope (PHF-1 tau), but it is transiently expressed by immature neurons during development. We have found that proliferating SH-SY5Y human neuroblastoma cells also express PHF-1 tau. Consistent with the recent finding that cell-cycle-dependent kinases can phosphorylate tau in vitro, flow cytometry showed that mitotic SH-SY5Y cells were up to 18-fold more PHF-1 immunoreactive than nonmitotic cells. On immunoblots, PHF-1 tau in mitotic and nonmitotic cells also was strikingly different. First, mitosis induced a prominent PHF-1 reactive band at 120 kDa, which likely accounted for the large increase in PHF-1 signal seen at mitosis. Although the size of the 120-kDa band is consistent with it being the high-molecular-weight form of tau, other antibodies to tau did not recognize it. Second, mitosis caused a hyperphosphorylation of the PHF-1 immunoreactive tau band normally seen at 50 kDa. In mitotic cells this band had an increased intensity and molecular weight. Alkaline phosphatase treatment abolished tau M(r) heterogeneity, verifying that the variations in mobility were due to phosphorylation. These data show that cell-cycle-dependent hyperphosphorylation of tau occurs in intact cells, and they support the hypothesis that aberrant activity of cell-cycle-dependent kinases may contribute to tau phosphorylation and PHF formation in Alzheimer's disease.

Amino acid residues 226-240 of tau, which encompass the first Lys-Ser-Pro site of tau, are partially phosphorylated in Alzheimer paired helical filament-tau

A synthetic peptide corresponding to residues 226-240 (E9 peptide) of human tau, which contains an Lys-Ser-Pro motif, was used to raise a polyclonal antibody. The antibody, E9, was 10-fold less reactive with phospho-E9 peptide than with native E9 peptide. E9 antibody was used to study the extent of phosphorylation in a modified form of tau (PHF-tau) that is found in Alzheimer's disease brain and is incorporated into paired helical filaments (PHFs). E9 immunolabeled Alzheimer's disease neurofibrillary tangles and abnormal neurites in brain sections intensely, with increased immunoreactivity detected after pretreatment of sections with phosphatase. On immunoblots and ELISA, E9 reacted with PHF-tau and recombinant human tau but not with the high and middle molecular weight neurofilament proteins. Phosphatase treatment of PHF-tau improved the E9 immunoreactivity by 30-50%. Dephosphorylated high but not middle molecular weight neurofilament protein became reactive with E9. These results indicate that < 50% of the PHF-tau is phosphorylated in the subregion corresponding to residues 226-240 of tau and suggest that the phosphorylation of this region may not be essential for PHF formation.

Pathological Tau proteins of Alzheimer's disease as a biochemical marker of neurofibrillary degeneration

Paired Helical Filaments (PHF) accumulate in the degenerating neurons from the associative cortical brain areas during Alzheimer's disease. They are composed of a triplet of hyperphosphorylated microtubule-associated protein Tau, called Tau 55, 64, 69 or PHF-Tau. The distribution of PHF-Tau in the different brain areas corroborates neuropathological observations and specifies that: the entorhinal cortex and hippocampus are vulnerable regions specifically affected by Alzheimer-type neurofibrillary degeneration during aging, the temporal cortex is already affected at the very first stage of clinical manifestations, almost the whole brain is concerned by neurofibrillary degeneration at the end-stages of the disease. Tau-PHF are also observed in the cortical areas from Parkinson patients with dementia, and more especially in the prefrontal cortex. Tau pathology for Progressive Supranuclear Palsy is significantly different, with a doublet of pathological Tau, namely Tau 64 and 69, in almost all cortical and subcortical areas. Therefore, the presence of pathological Tau proteins in several associative cortical areas is always associated with severe intellectual impairment. Finally, PHF-Tau are powerful biochemical markers of the degenerating process which could be used for setting up an early biological diagnosis test of Alzheimer's disease based upon the immunodetection of PHF antigens in the CSF, as well as for developing experimental models of neurodegeneration.

Analysis of magnetic resonance spectra by mole percent: comparison to absolute units

A variety of metabolites present in perchloric acid extracts of brain tissue were measured by 1H and 31P magnetic resonance spectroscopy (MRS) and HPLC in the same tissue sample and the MRS results were expressed both in terms of mole % and mumole/g based on an internal standard. The levels of 16 metabolites were compared by linear regression analysis and the mole % results were found to correlate very well with the results expressed as mumole/g. To compare the two units under typical experimental conditions, the percent change in metabolites in a group of Alzheimer's disease brains was compared to a control group using both units. The results were essentially identical for the mole % and mumole/g methods. We conclude that the use of the mole % method of expressing MRS data yields results which are equivalent to those expressed in absolute units and suggest that, for in vivo MRS studies, use of the mole % method is preferable because fewer artifacts, such as partial volume effects, are introduced.

In vivo phosphorylation sites in fetal and adult rat tau

Fetal tau and tau in paired helical filaments show similar immunoreactivities with several phosphorylation-dependent paired helical filament-polyclonals and monoclonals, suggesting that the two molecules share several distinct phosphorylated epitopes. To make clear the similarities and differences between the two, we have undertaken work to identify the in vivo phosphorylation sites in fetal rat tau. We have approached this problem by identifying phosphopeptides by means of mass spectrometry and sequencing of those phosphopeptides after modification with ethanethiol. Although remarkable heterogeneity was present, fetal tau was found to bear at most 10 phosphates at Ser-189, Ser-190, Ser-193, Ser-226, Ser-387, Ser-395, Thr-172, Thr-222, and, presumably, Ser-391 and Thr-208 (numbering is according to the longest form of rat tau; Kosik, K. S., Orecchio, L. D., Bakalis, S., and Neve, R. L. (1989) Neuron 2, 1389-1397). In contrast, adult rat tau was much less phosphorylated; only Thr-172, Ser-190, Ser-193, Thr-222, and Ser-395 were phosphorylated to a slight-to-moderate extent. All these sites except for Ser-189 and Ser-391 were followed by Pro residues. Thus, tau is an in vivo substrate for proline-directed protein kinase(s), and its phosphorylation state is developmentally regulated.

Characterization of a PC12 cell sub-clone (PC12-C41) with enhanced neurite outgrowth capacity: implications for a modulatory role of high molecular weight tau in neuritogenesis

To address the means by which diversity of neuronal morphology is generated, we have isolated and characterized naturally occurring variants of rat PC12 pheochromocytoma cells that exhibit altered neurite outgrowth properties in response to nerve growth factor (NGF). We describe here a PC12 cell sub-clone, designated PC12-clone 41 (PC12-C41), that displays significant increases in neurite abundance and stability when compared with the parental line. This difference does not appear to be due to an altered sensitivity or responsiveness to NGF or to a more rapid rate of neurite extension. Because of the role of the cytoskeleton in neuritogenesis, we examined a panel of the major cytoskeletal proteins (MAP 1.2/1B, beta-tubulin, chartins, peripherin, and high and low molecular weight (HMW and LMW) taus) whose levels and/or extent of phosphorylation are regulated by NGF in PC12 cultures. Although most cytoskeletal proteins showed little difference between PC12 and PC12-C41 cells (+/- NGF treatment), there was a significant contrast between the two lines with respect to tau expression. In particular, while NGF increases the total specific levels of tau in both cell types to similar extents (by about twofold), the proportion comprising HMW tau is threefold higher in the PC12-C41 clone than in PC12 cells. A comparable difference was observed under substratum conditions that were non-permissive for neurite outgrowth and so this effect was not merely a consequence of the differential neuritogenic capacities of the two lines. The distinction between the expression of HMW and LMW taus in PC12 and PC12-C41 cells (+/- NGF) was also observed at the level of the messages encoding these proteins. Such findings indicate that initiation of neurite outgrowth in PC12 cultures does not require a massive induction of tau expression and raise the possibility that HMW and LMW taus may have differential capacities for modulating neuronal morphology.

Analysis of paired helical filaments (PHFs) found in Alzheimer's disease using freeze-drying/rotary shadowing

Paired helical filaments (PHFs) isolated from Alzheimer's disease (AD) brains were analyzed using freeze-drying/rotary shadowing and immunoelectron microscopy. These filaments are the major contributors to the formation of neurofibrillary tangles (NFTs) found in AD, and are composed primarily of the microtubule-associated protein tau. We have focused on the identification of PHF-tau protein within isolated PHFs using anti-tau antibodies (tau 14, 46, 60). These PHFs consisted of both helically twisted and "straight" paired filaments. With freeze-drying/rotary shadowing, we are able to demonstrate in 2-D and 3-D subtle twists within the straight filaments as well as immunolabeling of the individual filamentous strands composing the PHFs. Additionally, projections emanating from numerous filaments and bridges between PHFs often were immunolabeled with anti-tau antibodies. Our results suggest that tau proteins are present in discrete, nonconfluent patterns within the PHFs and are components of both strands composing the PHFs. Tau proteins are present in the bridges between individual PHFs and may contribute to interconnecting PHFs into a complex macromolecular network such as the NFTs found in AD.

Heterogeneity of the high molecular weight tau proteins in N115 neuroblastoma cells

The sequence of a high molecular weight (HMW) tau cDNA cloned from a neuroblastoma N115 library contains, in addition to the C- and N-terminal and middle regions present in the low molecular weight mouse brain tau proteins, a 711-bp nonhomologous domain (exon 4a) and a region of 198 bp corresponding to exon 6 of the tau gene. Protein immunoblot analysis, performed with antibodies specific either for a sequence present in the N-terminal region of all the tau variants or for exon 4a revealed several bands suggesting that more than one tau form is expressed in this cell line. Northern blot experiments performed with a number of cDNA probes spanning domains common and uncommon to low molecular weight and HMW tau allowed the identification of four tau transcripts differing in the size of their coding and noncoding regions. All these transcripts contain the sequence encoded by exon 6, but two of them lack exon 4a. As shown by RNase protection assays, the N-terminal region of these transcripts is also variable and contains either exon 1, or exons 1 and 2, or exons 1-3. Yet all these HMW tau forms contain four homologous repeats in their C-terminal domain both in the differentiated and nondifferentiated cells, i.e., have adult characteristics. In conclusion, the data reported in this article demonstrate that several HMW tau variants are expressed in neuroblastoma N115 cells and that the transition between immature to mature tau forms occurring during brain development is not required for neurite outgrowth during morphological differentiation of this cell line.

Phosphorylation of tau by cyclic-AMP-dependent protein kinase

Alzheimer's disease paired helical filaments contain abnormally phosphorylated tau (PHF-tau) which has reduced electrophoretic mobility on sodium dodecyl sulphate polyacrylamide electrophoresis. We have investigated the effects of cyclic-AMP-dependent protein kinase (PKA) on recombinant human tau isoforms and two recombinant tau fragments. PKA phosphorylated tau and reduced its electrophoretic mobility, phosphorylation towards the C-terminus of tau having a major influence on this property. Substitution of serine396 (phosphorylated in PHF-tau) or serine416 (phosphorylated by calcium/calmodulin kinase II) by alanine demonstrated that these are not major sites for PKA phosphorylation. Although the phosphorylated forms of tau generated by PKA are not identical to those of PHF-tau, PKA may be involved in the generation of PHF-tau in Alzheimer's disease via phosphorylation of additional, as yet unidentified, sites on tau.

The balance between tau protein's microtubule growth and nucleation activities: implications for the formation of axonal microtubules

The microtubule-associated protein tau is found primarily in neuronal tissues and is highly enriched in the axon. It promotes microtubule assembly in vitro and stabilizes microtubules in cells. To study how tau protein might be involved in the unique features of axonal microtubules, we have analyzed the effect of E. coli-synthesized tau protein using an in vitro centrosome-mediated microtubule regrowth assay over a wide range of tau/tubulin ratios. We report that microtubule assembly promoted by tau protein exhibits characteristic changes dependent on the tau/tubulin ratio. Above a threshold level, nucleation of new microtubules is favored over growth of existing ones. tau isoform variation does not change this phase transition in microtubule assembly. We discuss how tau might participate in the elaboration of axonal morphology based on our results and present evidence that the phase transition from microtubule growth to nucleation is critical for axonal development.

High and low molecular weight tau proteins are differentially expressed from a single gene

Both high and low molecular weight (HMW and LMW) tau proteins are expressed in the immature and adult mouse spinal cord. Northern blot analysis, performed with probes complementary to domains common and uncommon to the LMW and HMW entities, suggested that HMW tau proteins found in the immature mouse spinal cord are not translated from the single transcript of 6 kb expressed at these stages, but are transported within this nervous structure by axons arising in the periphery. In contrast, another minor transcript of 8 kb was detected in the adult mouse spinal cord by a HMW tau specific probe, suggesting that a small fraction of the HMW tau forms present in adulthood are translated within mouse spinal cord neurons. LMW spinal cord tau forms are encoded by mRNAs of 6 kb that contain three and four homologous repeats at immature and mature stages, respectively, whereas adult HMW entities contain four repeats. PCR analysis performed with mouse genomic DNA also showed that the nonhomologous region specific for HMW tau is a single exon. Southern blot and gene mapping showed that the same gene, located on the murine chromosome 11, encodes all the LMW and HMW tau variants. All these tau forms, therefore, are produced by an alternative splicing mechanism that is neuron-specific and developmentally regulated.