19 April 2006. Lester Binder, Northwestern University, Chicago, Illinois, described the changes in tau conformation that correlate with the stages of neurofibrillary tangle (NFT) formation. Two distinct conformations of tau, Alz50 and Alz66, can exist in the same neurons, but are separated temporally. Carboxy truncation of tau by caspase (tau-C3) facilitates tau polymer formation in vitro, but only occurs after tangles begin to coalesce in situ such that native tau→ tau Alz50 positive→ Alz50 and tau-C3 positive→ tau-C3 and Alz66 positive. Phosphorylation at S422 and tangle formation appears to predate tau truncation in situ. N-terminal truncations and modifications of tau showed that the amino terminus of tau potentiates aggregation. Thus, Binder argued that tau assembly into filamentous aggregates is governed by more than just the tau repeat regions.

Karen Duff, Nathan Kline Institute, Orangeburg, New York, described experiments employing mice overexpressing p25 crossed with mutant tau mice, as well as an inducible p25 cell line. P25/Cdk5 activity was found to influence APP processing in vitro and in vivo. Levels of BACE, C99, and Aβ were elevated in p25 mice. In p25-expressing cells, levels and activity of BACE were increased, and the synthesis rate of BACE was also altered. Several sites in the BACE promoter or 5’UTR are potentially affected by phosphorylation by p25/cdk5, including MEF2 and STAT sites. Effects of Cdk5 activity on the endosomal system may underlie altered substrate/secretase availability or activity. P25/Cdk5 can enhance both Aβ and tangle formation, but neither plaques nor tangles were seen in the p25 mice. Duff also showed that in mice, inhibition of GSK3 using lithium correlated well with reduced insoluble tau and functional decline, but not with pathological (immunohistochemical) markers. The effect of lithium was not mediated through AKT, or other Li-related pathways, since a second GSK3 inhibitor (AR) had a similar effect on insoluble tau in two different mouse models.

Eva-Maria Mandelkow, Max Planck Institute, Hamburg, Germany, described the role of tau in APP and BACE trafficking and presented data showing that tau inhibits anterograde transport of APP into axons and dendrites. However, inhibition of APP transport by tau overexpression could be rescued by MARK kinase. Tau-induced accumulation of APP in the cell body did not lead to increased levels of Aβ, and transfection with tau retarded the secretion of APP and of Aβ. Mandelkow also presented data indicating that vesicles carrying APP or BACE had different movement characteristics, and APP and BACE1 were likely carried by distinct vesicles (in disagreement with recent results reported by Larry Goldstein’s laboratory at UCSD). APP was transported faster than BACE, and APP processing was not observed during the transport of APP down axons. There was no measurable cleavage of APP during vesicle transit: The ratio of N-terminal and C-terminal ends remained constant. Mandelkow also presented data in tau-inducible transgenic mice showing that tau aggregation starts in the entorhinal cortex at 3 months in transgenic mice with a proaggregation mutant. Tau proaggregation mutants aggregated early, while tau anti-aggregation mutants did not aggregate at all. Tau aggregation also progressed with age. Proaggregation mutant mice exhibited evidence for early phosphorylation of S262 tau. Aged transgenic mice revealed neuronal loss in the dentate gyrus. It was also pointed out that Mandelkow’s inducible tau-δK280 mice express tau at only 1.7-fold endogenous levels, but still lead to early aggregates and disease. (In contrast, the tau P301L mutant mice published by Karen Ashe, University of Minnesota, Minneapolis, overexpress tau at >sixfold endogenous levels.)

Michael Hutton, Mayo Clinic, Jacksonville, Florida, reviewed 34 mutations in the tau gene, MAPT, that cause of FTDP-17 in >90 families. FTDP-17 mutations can have a number of effects. Group 1 consists mainly of coding mutations that disrupt microtubule binding and enhance tau aggregation. Group 2 consists of splicing mutations that disrupt alternative splicing of exon 10 and increase four repeat (4R) tau. Mutations that increase 4R tau show clear linkage to disease; absolute increase in 4R tau (not 4R:3R ratio) appears to underlie the rate of aggregation. Hutton also demonstrated that risk for 4R tauopathies (progressive supranuclear palsey [PSP]/corticobasal degeneration [CBD]) and Parkinson disease is associated with the H1/H1 genotype of tau. The SNP rs242557 is also part of the H1 haplotype and associated with risk for PSP. This SNP sits in the LBP-1/LSF/CP2 transcription site in the tau promoter, and the risk allele (as well as the H1 allele) is associated with increased expression of tau. Using a genome-wide association study (Affymetrix 500k SNP chip) for PSP, MAPT was confirmed as a major risk factor genome-wide. In addition, association with PSP (O.R. = 2.1) was observed for SNPs on chromosome 11p11.2 contained within a single block of linkage disequilibrium spanning the genes for DNA damage binding protein (DDB2) and lysosomal acid phosphatase (ACP2).

Also from the Mayo Clinic in Florida, Chad Dickey introduced the hypothesis that depletion of CHIP, a tau-specific ubiquitin ligase, leads to increased abnormal tau accumulation. He showed that in CHIP-/- mice, a large subset demonstrated prenatal morbidity and marked accumulation of cerebral phospho-tau levels. Prenatal CHIP-/- mice developed marked accumulation of soluble, ubiquitin-negative, phospho-tau species. CHIP-/- mice had robust increases in phospho-tau, but no indications of aberrant folding. Caspase-3 activation was increased in CHIP-/- mice and was associated with increased apoptosis and elevated levels of caspase-3 cleavage of tau at Asp421. PAR-1/MARK2 overexpression was able to prevent CHIP from recognizing tau.

In this session's short talk, Kun Ping Lu, Beth Israel Deaconess Medical Center, Boston, demonstrated that knockout of the gene Pin1 on chromosome 19, which catalyzes the conversion of certain phosphorylated Ser/Thr-Pro motifs in polypeptides between cis and trans conformations, led to the accumulation of cis-pAPP, which increased amyloidogenic APP processing and increased Aβ plaque formation. Pin1 knockout also led to cis-p-tau accumulation, which resulted in hyperphosphorylated tau and tangles. Pin1 also changed the conformation of APP and tau from cis to trans. Pin1 knockout, alone or in combination with mutant APP, increased amyloidogenic APP processing in mice and elevated levels of Aβ42. These data suggest that Pin1-catalyzed prolyl isomerization may underlie a common pathogenic pathway leading to both Aβ and tau pathologies in AD.