Tau protein induces bundling of microtubules in vitro: comparison of different tau isoforms and a tau protein fragment.

Expression of tau protein in non-neuronal cells can result in a redistribution of the microtubule cytoskeleton into thick bundles of tau-containing microtubules (Lewis et al.: Nature 342:498-505, 1989; Kanai et al.: J Cell Biol 109:1173-1184, 1989). We reconstituted microtubule bundles using purified tubulin and tau in order to study the assembly of these structures. Taxol-stabilized tubulin polymers were incubated with various concentrations of recombinant human tau and examined by electron microscopy. With increasing concentrations of tau 3 (tau isoform containing three microtubule binding domains) or tau 4 (isoform containing four microtubule binding domains) the microtubules changed orientation from a random distribution to loosely and tightly packed parallel arrays and then to thick cables. In contrast, tau 4L, the tau isoform containing four microtubule binding domains plus a 58-amino acid insert near the N-terminus, showed minimal bundling activity. tau 4-induced bundling could be inhibited by the addition of 0.5 M NaCl or 0.4 mM estramustine phosphate, conditions which are known to inhibit tau binding to microtubules. A tau construct that contained only the microtubule binding domains plus 19 amino acids to the C-terminus was fully capable of bundling microtubules. Phosphorylation of tau 3 with cAMP-dependent protein kinase had no effect on its ability to induce microtubule bundling. These results indicate that tau protein is directly capable of bundling microtubules in vitro, and suggests that different tau isoforms differ in their ability to bundle microtubule filaments.     

Site-specific effects of tau phosphorylation on its microtubule assembly activity and self-aggregation

Microtubule-associated protein tau is abnormally hyperphosphorylated and aggregated into neurofibrillary tangles in brains with Alzheimer's disease. The phosphorylation sites of tau are mainly localized in the proline-rich (residues 172–251) and C-terminal tail (residues 368–441) regions, which flank the microtubule-binding repeats. Here, we investigated the effects of tau phosphorylation at these distinct sites/regions on its activity of stimulating microtubule assembly and its self-aggregation. We found that tau phosphorylation at the proline-rich region by dual-specificity tyrosine-phosphorylated and -regulated kinase 1A inhibited its microtubule assembly activity moderately and promoted its self-aggregation slightly. Tau phosphorylation at the C-terminal tail region by glycogen synthase kinase-3β increased its activity and promoted its self-aggregation markedly. Tau phosphorylation at both regions plus the microtubule-binding region by cAMP-dependent protein kinase diminished its activity (∼70% inhibition) and disrupted microtubules. These studies reveal the differential regulation of tau's biological activity and self-aggregation by phosphorylation at various sites/regions.     

A novel tau mutation in exon 9 (1260V) causes a four-repeat tauopathy

A novel mutation in exon 9 of tau, I260V, is associated with a clinical syndrome consistent with frontotemporal dementia with extensive tau pathology; however, neurofibrillary tangles and Pick bodies are absent. Significantly, Sarkosyl-insoluble tau extracted from affected brain tissue consisted almost exclusively of four-repeat isoforms. Consistent with these findings, in vitro biochemical assays demonstrated that the I260V mutation causes a selective increase in tau aggregation and a decrease in tau-induced microtubule assembly with four-repeat isoforms only. The contrasting pathology and biochemical effects of this mutation suggest a different disease mechanism from the other exon 9 mutations and demonstrates the critical role for the first microtubule-binding domain in tau-promoted microtubule assembly and the pathogenic aggregation of tau.     

Developmental regulation of alternatively spliced isoforms of mRNA encoding MAP2 and tau in rat brain oligodendrocytes during culture maturation.

Oligodendrocytes are responsible for the formation and maintenance of the myelin sheaths in the central nervous system (CNS), and microtubules essentially participate in the elaboration and stabilization of myelin-containing cellular processes. We have shown before that the two major groups of neuronal microtubule-associated proteins (MAPs), MAP2 and tau, are expressed in the myelin forming cells of the CNS (Mueller et al. [1997] Cell Tissue Res. 288:239-249). Here we demonstrate for the first time that during culture maturation, changes in mRNA splicing and a shift from immature to mature MAP2 and tau mRNAs occur in oligodendrocytes. Similarly to neurons, a developmental shift from MAP2 isoforms with 3 microtubule (MT)-binding domains (3R) to the isoforms with 4 MT-binding domains (4R) is observable. MAP2c constitutes the major MAP2 isoform in oligodendrocytes. They contain tau mRNA splice products with both 3 and 4 MT-binding repeats (3R, 4R) with no amino terminal insert or with exon 2, and do not express isoforms containing exon 3. The shortest form tau 1 (3R; no inserts) representing the immature tau isoform is most prominently expressed in early progenitor cells and gradually decreases during culture maturation, while tau 5 (4R; with exon 2) appears later during in vitro differentiation. The product corresponding to tau 2 (3R; with exon 2) and tau 4 (4R; no inserts) remains approximately at the same level. Hence, the occurrence of MAPs in oligodendrocytes is developmentally regulated. While in progenitor cells, 3R- and 4R-MAP2c are expressed at approximately the same level, in mature oligodendrocytes after 12 days in vitro, the ratio of 4R- to 3R-MAP2c is nearly 2. In contrast, the ratio of 4R- to 3R-tau in progenitor cells is 1:3 and shifts to 1:1 after 12 days in culture.     

Pick's disease is associated with mutations in the tau gene

Recently, mutations within the tau gene have been associated with some familial forms of frontotemporal dementia. To investigate whether tau gene mutations are also associated with Pick's disease, we analyzed the tau gene in 30 cases of pathologically confirmed Pick's disease. Two coding mutations were identified in separate cases of Pick's disease. A glycine-to-arginine mutation at codon 389 was detected in 1 case and a lysine-to-threonine mutation at codon 257 was identified in another. Analysis of dephosphorylated tau from the brain of the patient with the codon 389 mutation revealed a prominent band representing tau, with four microtubule-binding domains and no amino terminal inserts. This is in contrast to Pick's disease without any tau gene mutations, which consist of tau with mainly three microtubule-binding domains and only a trace of tau, with four microtubule-binding domains. Functional analysis of tau with these two mutations demonstrated a reduced ability of tau to promote microtubule assembly. Surprisingly, these mutations increased tau's susceptibility to calpain I digestion, suggesting that this feature may be related to the formation of a Pick type of histology. Moreover, these data suggest that Pick's disease is not a separate entity but part of the frontotemporal dementia disease spectrum.     

An R5L tau mutation in a subject with a progressive supranuclear palsy phenotype

MAPT, the gene encoding tau, was screened for mutations in 96 progressive supranuclear palsy subjects. A point mutation (R5L) was identified in a single progressive supranuclear palsy subject that was not in the other progressive supranuclear palsy subjects or in 96 controls. Functionally, this mutation alters the ability of tau to promote microtubule assembly. Analysis of soluble tau from different brain regions indicates that the mutation does not affect the ratio of tau isoforms synthesized. Aggregated insoluble tau from subcortical regions was predominantly four-repeat tau with no or one amino terminal insert (0N4R and 1N4R). Insoluble tau from cortical regions also contained 1N3R tau. Thus, the R5L mutation causes a progressive supranuclear palsy phenotype, presumably by a gain-of-function mechanism.     

Pathological inclusion bodies in tauopathies contain distinct complements of tau with three or four microtubule-binding repeat domains as demonstrated by new specific monoclonal antibodies

Pathological inclusions containing fibrillar aggregates of hyperphosphorylated tau protein are a characteristic feature in the tauopathies, which include Alzheimer's disease, frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP-17), progressive supranuclear palsy, corticobasal degeneration and Pick's disease. Tau isoform composition and cellular and regional distribution as well as morphology of these inclusions vary in each disorder. Recently, several pathological missense and exon 10 splice-donor site mutations of the tau gene were identified in FTDP-17. Exon 10 codes for the second of four microtubule-binding repeat domains. The splice-site mutations result in increased inclusion of exon 10 which causes a relative increase in tau isoforms containing four microtubule-binding repeat domains over those containing three repeat domains. This could be a central aetiological mechanism in FTDP-17 and, perhaps, other related tauopathies. We have investigated changes in the ratio and distribution of three-repeat and four-repeat tau in the different tauopathies as a basis of the phenotypic range of these disorders and the selective vulnerability of different subsets of neurones. In this study, we have developed two monoclonal antibodies, RD3 and RD4 that effectively distinguish these closely related tau isoforms. These new isoform-specific antibodies are useful tools for analysing tau isoform expression and distribution as well as pathological changes in the human brain.     

Human immunoglobulins and Fc fragments promote microtubule assembly via tau proteins and induce conformational changes of neuronal microtubules in vitro

The influence of human immunoglobulins (Ig) in neuronal cytoskeleton stability was studied in vitro. Here we show that human Ig and Fc fragments stimulate animal and human microtubule assembly by binding to microtubules via tau isoforms. In presence of Ig, microtubules show increased aggregation, twisting and rigidity. Non-immune Ig and Fc fragments promote microtubule assembly in temperature-dependent manner and stabilize microtubules at a molecular ratio of 1 Ig per 4 tubulin dimers. These in vitro data provide an experimental support for an immuno-mediated modulation of the cytoskeleton. In conjunction with previous neuropathological data, they suggest that Ig could participate in early stages of neurodegeneration by affecting the microtubule stability in vivo.     

The novel <Emphasis Type="Italic">Tau</Emphasis> mutation G335S: clinical,neuropathological and molecular characterization

Mutations in Tau cause the inherited neurodegenerative disease, frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17). Known coding region mutations cluster in the microtubule-binding region, where they alter the ability of tau to promote microtubule assembly. Depending on the tau isoforms, this region consists of three or four imperfect repeats of 31 or 32 amino acids, each of which contains a characteristic and invariant PGGG motif. Here, we report the novel G335S mutation, which changes the PGGG motif of the third tau repeat to PGGS, in an individual who developed social withdrawal, emotional bluntness and stereotypic behavior at age 22, followed by disinhibition, hyperorality and ideomotor apraxia. Abundant tau-positive inclusions were present in neurons and glia in the frontotemporal cortex, hippocampus and brainstem. Sarkosyl-insoluble tau showed paired helical and straight filaments, as well as more irregular rope-like filaments. The pattern of pathological tau bands was like that of Alzheimer disease. Experimentally, the G335S mutation resulted in a greatly reduced ability of tau to promote microtubule assembly, while having no significant effect on heparin-induced assembly of recombinant tau into filaments.     

Missense point mutations of tau to segregate with FTDP-17 exhibit site-specific effects on microtubule structure in COS cells: a novel action of R406W mutation

Missense and splicing point mutations have been found in the tau gene in families with frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP-17). Of these mutations, we examined four exonic missense point mutations (G272V, P301L, V337M and R406W) in 3-repeat or 4-repeat tau isoform on the transfection experiment. The effects of two mutations (G272V or P301L) on microtubules were subtle whereas those of two other mutations (V337M or R406W) were dramatically significant when these two mutations were constructed into 3-repeat tau but not into 4-repeat tau. The R406W mutation induced an alternation of microtubules to form dotted or fragmented forms retaining colocalization of tau with tubulin whereas the V337M mutation predominantly disrupted microtubule networks and diminished colocalization of tau and tubulin. The effect of the mutations on microtubules were thus site-dependent and isoform-dependent. Tau with R406W mutation was found to be colocalized with tubulin without filamentous structures on confocal views, suggesting that the carboxyl region of tau played a different role from tubulin-binding domain on microtubule assemble. Another abnormal property was identified in tau with R406W mutation that failed to suffer phosphorylation. Thus, diverse effects of tau mutations on microtubules may explain the various clinicopathologies of FTDP-17 and related tauopathies.     

Expression of embryonic tau protein isoforms persist during adult neurogenesis in the hippocampus

Tau is a microtubule-associated protein with a developmentally regulated expression of multiple isoforms. The neonatal isoform is devoid of two amino terminal inserts and contains only three instead of four microtubule-binding repeats (0N/3R-tau). We investigated the temporal expression pattern of 0N-tau and 3R-tau in the rat hippocampus. After the decline of 0N- and 3R-tau immunoreactivity during the postnatal development both isoforms remain highly expressed in a few cells residing beneath the granule cell layer. Coexpression of the polysialylated neuronal cell adhesion molecule, doublecortin, and incorporated bromodeoxyuridine showed that these cells are proliferating progenitor cells. In contrast mature granule cells express the adult tau protein isoform containing one aminoterminal insert domain (1N-tau). Therefore a shift in tau isoform expression takes place during adult neurogenesis, which might be related to migration, differentiation, and integration in the granule cell layer. A model for studying shifts in tau isoform expression in a defined subset of neurons might help to understand the etiology of tauopathies, when isoform composition is crucial for neurodegeneration, as in Pick's disease or FTDP-17.     

Molecular characterization of microtubule-associated proteins tau and MAP2.

Tau and MAP2 are two of the major microtubule-associated proteins in the vertebrate nervous system. They promote microtubule assembly and stability, and might be involved in the establishment and maintenance of neuronal polarity. In nerve cells immunohistochemistry shows complementary distributions, with tau being concentrated in axons and high molecular mass MAP2 being confined to dendrites. Each protein consists of multiple isoforms that contain three or four homologous tandem repeats near the carboxy-terminus, which constitute microtubule-binding domains. In humans, tau consists of at least six isoforms of related amino acid sequences that are produced from a single gene by alternative mRNA splicing and that are expressed in a stage- and cell type-specific manner. Tau is also a component of the paired helical filaments associated with Alzheimer's disease and other disorders of the CNS. Rat MAP2 consists of at least three isoforms produced from a single gene: high molecular mass MAP2a and MAP2b, and low molecular mass MAP2c. MAP2c is expressed only during early development and has so far been seen only in axons; MAP2a appears to replace MAP2c, whereas MAP2b is expressed throughout life. Messenger RNAs for MAP2 of high molecular mass are expressed both in cell bodies and in dendrites, consistent with the dendritic localization of the corresponding protein isoforms.     

Mutated tau binds less avidly to microtubules than wildtype tau in living cells

Some forms of genetically inherited dementia, including frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP-17), are caused by mutations in tau. We have examined several mutations in the microtubule-binding portion of tau for their effect on microtubule binding, cellular distribution and cytoskeletal structure in mammalian cells. Using constructs coding for mutant (P301L and V337M) and wildtype human tau fused to a green fluorescent protein analog (EGFP) we followed the disposition of tau in live cells after transient transfection using confocal microscopy. Most of the tau protein localized to structures that resembled microtubules or microtubule bundles and co-localized with tubulin. At 3 days post-transfection mutant tau proteins showed a higher abundance of free tau in the cytoplasm than did wildtype tau. Cells expressing the P301L mutation showed proportionally more cytoplasmic localization of tau. Confirming these results, fractionated cells with mutant tau had a higher percentage of tau in the cytoplasmic compartment as compared to the cytoskeletal compartment. Cells with wildtype tau had most tau in the cytoskeletal fraction. Because the mutations (V337M, P301L) are associated with genetic tauopathies, these results suggest that a factor in disease etiology of genetic tauopathies and other dementias with altered tau is a greater abundance of tau in the cytoplasm due to decreased binding to microtubules. This increased cytoplasmic presence may be a significant factor in promoting tau aggregation.     

The L266V tau mutation is associated with frontotemporal dementia and Pick-like 3R and 4R tauopathy

We report a case of rapidly progressive frontotemporal dementia presenting at age 33 years. At autopsy there was severe atrophy of the frontal and temporal lobes. Tau-positive Pick bodies, which ultrastructurally were composed of straight filaments, were present, accompanied by severe neuronal loss and gliosis. RD3, a tau antibody specific for the three-repeat (3R) isoforms, labeled the Pick bodies. ET3, a four-repeat (4R) isoform-specific tau antibody, did not label Pick bodies, but highlighted rare astrocytes, and threads in white matter bundles in the corpus striatum. Analysis of the tau gene revealed an L266V mutation in exon 9. Analysis of brain tissue from this case revealed elevated levels of exon 10+ tau RNA and soluble 4R tau. However, both 3R and 4R isoforms were present in sarkosyl-insoluble tau fractions with a predominance of the shortest 3R isoform. The L266V mutation is associated with decreased rate and extent of tau-induced microtubule assembly, and a 3R isoform-specific increase in tau self assembly as measured by an in vitro assay. Combined, these data indicate that L266V is a pathogenic tau mutation that is associated with Pick-like pathology. In addition, the results of the RD3 and ET3 immunostains clearly explain for the first time the presence of both 3R and 4R tau isoforms in preparations of insoluble tau from some Pick's disease cases.     

Transient expression of fluorescent tau proteins promotes process formation in PC12 cells: contributions of the tau C-terminus to this process

The neuronal microtubule-associated protein tau promotes microtubule assembly and has been implicated in the development of axonal morphology. In this study, PC12 cells were transiently transfected with constructs coding fusion proteins of human tau with green fluorescent protein (GFP). Expression of tau constructs actively stabilized microtubules. Expression of the C-terminus of tau can mimic this effect in living cells, though to a lesser extent because of the absence of the tau N-terminus. However, tau colocalization with microtubules did not require the presence of the tau N-terminus. Transient expression of tau (including tau24, a four-repeat human tau isoform encoded in 383 residues, and tau23, human fetal tau isoform encoded in 352 residues) stimulated process formation in PC12 cells, and this occurred faster with tau24 than with tau23. The residues (residues 154-172 in tau23) that confer microtubule nucleation activity of tau in vitro are not required for tau-directed process formation. However, when tau induces the formation of cellular processes in response to cortical breakdown by cytochalasin B, residues 154-172 must be present. Thus, it appears that tau may serve to promote cellular process outgrowth in cultured neuronal cells and that C-terminus of tau is essential to this process.     

Sequence of a human MAP-2 region sharing epitopes with Alzheimer neurofibrillary tangles

Microtubule-associated protein 2 (MAP-2), an abundant neuronal protein, consists of a short microtubule-binding domain and a long projection arm. MAP-2 shares epitopes with Alzheimer neurofibrillary tangles (NFT). However, most anti-MAP-2 antibodies do not stain detergent-extracted NFT, and the role of MAP-2 in NFT formation has therefore been unclear. We have determined the sequence of a 1.7 kb partial MAP-2 cDNA encoding at least three NFT epitopes. The epitopes are not removed by detergent extraction of tangle preparations, suggesting that they are integral components of NFT. Expression vectors containing restriction fragments of the cDNA were used to assign the epitopes to a 51-amino-acid region near the end of the MAP-2 projection arm, distal to the microtubule.     

Sporadic Pick's disease: a tauopathy characterized by a spectrum of pathological tau isoforms in gray and white matter

Pick's disease is characterized neuropathologically by distinct tau-immunoreactive intraneuronal inclusions known as Pick bodies and by insoluble tau proteins with predominantly three microtubule-binding repeat tau isoforms. However, recent immunohistochemical studies showed that the antibody specific for exon 10, which encodes the fourth microtubule-binding repeat, detected other tau lesions in Pick's disease. To better define the spectrum of tau pathology in Pick's disease, we used biochemical, immunohistochemical, and ultrastructural techniques to analyze the tau isoform composition in 14 Pick's disease brains. Western blot analysis showed that both three and four microtubule-binding repeat pathological tau isoforms are present in gray and white matter of various brain regions. Using phosphorylation-dependent anti-tau antibodies, we show that major tau phosphoepitopes are present in sarcosyl-insoluble gray and white matter regions of Pick's disease brains. Also, for the first time to our knowledge, we demonstrated that isoforms with four microtubule-binding repeat tau isoforms are present in Pick bodies from selected brains. Isolated tau filaments were straight or twisted and formed by three microtubule-binding repeat or four microtubule-binding repeat tau isoforms. Major tau phosphorylation-dependent and exon 10-specific epitopes were present in filaments. Therefore, Pick's disease is characterized by an accumulations of Pick bodies in the hippocampal region and cortex as well as the presence of three and four microtubule-binding repeat tau pathology in both cortical gray and white matter that distinguish this tauopathy from other neurodegenerative disorders.     

Regulation and differential expression of tau mRNA isoforms as oligodendrocytes mature in vivo: implications for myelination

Oligodendrocytes and neurons derive from the same cell type but develop distinct morphologic and functional properties as they mature in vivo. Both cells express tau protein, a developmentally regulated protein in the central nervous system. The regulation of tau has been investigated extensively in neurons but not in oligodendrocytes, so we studied regulation of tau in oligodendrocytes in vivo. The amino-derived tau isoforms consist of isoforms with zero (A0), one (A1), or two (A2) inserts. We examined the developmental regulation of tau mRNA isoforms at the amino domain by comparing tau expression in oligodendrocytes (OLGs) isolated from 1- and 20-day-old rat brain and in age-matched cortex, which abounds in neurons. In the rat brain, myelination peaks at 20 days. By using semiquantitative RT-PCR, we found that OLGs and cortex from 1-day-old rat brain largely had amino-derived tau isoforms with no insert, whereas OLGs from 20-day-old rat brain had similar levels of amino-derived tau isoforms with no insert or with one insert. We also found that 20-day-old OLGs had twofold more tau mRNA levels than younger OLGs. In contrast to OLGs from 20-day-old rat brain, age-matched cortex had comparable levels of A0, A1, and A2 tau amino-derived isoforms. Further, younger and older OLGs had a reciprocal pattern of expression of both carboxy-derived tau mRNA isoforms with either three (3R) or four (4R) repeats. In contrast, younger and older cortex expressed either 3R or 4R tau. This study showed an upregulation of tau mRNA and cell-specific tau mRNA isoform expression in OLGs forming myelin.     

The interrelationship between selective tau phosphorylation and microtubule association

The purpose of this study was to examine the modulation of tau phosphorylation mediated by protein kinase A, a kinase with low intrinsic activity, and by the constitutively active glycogen synthase kinase, as well as to examine the subsequent effects on tau-microtubule association in differentiated human SH-SY5Y neuroblastoma cells. Activation of protein kinase A with forskolin and rolipram significantly increased tau phosphorylation at Ser262/356 only in the presence of okadaic acid, indicating that phosphates at these sites are normally turned over rapidly. In contrast, glycogen synthase kinase appears to maintain tau phosphorylation at Thr181 and Ser396/404 since inhibition of glycogen synthase kinase with lithium reduced phosphorylation at these sites. Lithium treatment also significantly decreased tau and tyrosinated α-tubulin levels. Perturbation of microtubules with nocodazole or taxol induced tau dephosphorylation at Tau-1 sites, Thr181 and Ser396/404, indicating that both constitutive kinase activity and microtubule state modulate tau phosphorylation at these sites. Nocodazole- or taxol-induced tau dephosphorylation was blocked by the protein phosphatase 2A/1 inhibitor okadaic acid, but not by the protein phosphatase 2B inhibitor cyclosporin A. In addition, osmotic stress, such as treatment with 20 mM NaCl, selectively increased tau phosphorylation at the Tau-1 epitope. To investigate the effect of phosphorylation on tau association with microtubules and microtubule stability in situ, a Triton X-100 extraction assay was utilized to separate the detergent-soluble cytosolic components from the detergent-insoluble cytoskeletal components. In control cells or cells treated with lithium very little tau was detected in the cytosolic fraction. Activation of protein kinase A in the presence of okadaic acid elevated tau levels in the detergent-soluble fraction, which contained all the tau phosphorylated at Ser262/356, and also decreased microtubule stability, as indicated by decreased acetylated α-tubulin levels. In conclusion, the phosphorylation state of tau in differentiated SH-SY5Y cells is regulated by glycogen synthase kinase, microtubule dynamics and osmotic stress at overlapping sites which apparently have little influence on tau-microtubule association. In contrast, phosphorylation of tau at Ser262/356 within the microtubule-binding, which was mediated in part by protein kinase A, prevented the association of tau with microtubules in situ.