Les microtubules sont des éléments clés du cytosquelette impliqué dans de nombreux processus cellulaires. Ce sont des structures dynamiques qui alternent continuellement entre polymérisation et dépolymérisation, un comportement appelé instabilité dynamique. Les microtubules sont particulièrement abondants dans les neurones et sont organisés sous formes de faisceaux dans les axones et les dendrites. Cette organisation particulière leur permet de maintenir la forme de ses cellules hautement spécialisées et d’assurer le transport intracellulaire d’éléments essentiels dans l’ensemble des compartiments neuronaux. De nombreux facteurs participe à la régulation de l’arrangement des microtubules dans les neurones. Parmi ces facteurs, la protéine tau fait partie de la famille des protéines associées aux microtubules (ou MAPs) et est majoritairement neuronale. Tau est un agent pontant majeur des microtubules et est également connue pour stabiliser les microtubules en stimulant leur polymérisation et en inhibant leur dépolymérisation. Malgré de nombreuses études sur l’interaction de tau avec les microtubules, les mécanismes par lesquels cette MAP contrôle leur organisation spatiale restent élusifs. Pour répondre à cette question, nous avons reconstitué in vitro des réseaux de microtubules en présence de divers isoformes, fragments et mutants de tau. La capacité de tau à induire des faisceaux stables de microtubules dépend de deux hexa-peptides localisés dans son domaine de liaison aux microtubules, et est régulée par son domaine de projection N-terminal. Nos résultats montrent que la phosphorylation spécifique de certains sites de tau inhibe soit la formation de faisceaux soit la stabilisation des microtubules, produisant des populations composées de microtubules individuels stable ou de faisceaux dynamiques. De plus, des mutations de tau impliquées dans des démences apparentées à la maladie d’Alzheimer augmentent drastiquement la capacité de tau à former des faisceaux composés de microtubules très dynamiques. Pour finir, des expériences de cryo-microscopie électroniques indiquent que tau génèrent des défauts dans la paroi des microtubules. Ces défauts sont connus pour assouplir les microtubules et pourraient donc constituer un mécanisme structural primaire permettant leur déformation au cours de la formation de faisceaux. En conclusion, nos résultats mettent en évidence un nouveau mécanisme phospho-dépendant par lequel tau régule l’organisation de réseaux de microtubules. De plus, ce travail révèle comment des modifications anormales de tau, telles que des phosphorylations anormales ou des mutations, peuvent altérer l’organisation du cytosquelette dans les maladies neurodégénératives. / Microtubules are key components of the eukaryotic cytoskeleton and are involved in major cellular events. They undergo constant remodeling through alternative cycles of growth and shrinkage of their extremities, a behavior known as dynamic instability. Microtubules are particularly abundant in neurons; they are organized into bundles within axons and dendrites to maintain the polarized shape of these highly specialized cells and to allow cargo transport. Numerous factors regulate the plasticity of the microtubule network in neurons. Among them, tau is a neuro-specific microtubule-associated protein (MAP). Tau is a major microtubule bundler also known to stabilize microtubules by promoting their growth and inhibiting their shrinkage. Although the interaction of tau with microtubules has been widely studied, the mechanisms by which this protein controls the spatial organization of microtubules remain elusive. To address this question, we reconstitute in vitro microtubule self-organization in presence of various tau isoforms, fragments and mutants. We find that the ability of tau to induce stable microtubule bundles depends on two conserved hexapeptides in tau’s microtubule-binding domain and is modulated by tau’s projection domain. Furthermore, our data demonstrate that site-specific phosphorylation of tau inhibits either microtubule bundling or stabilization generating alternative networks composed of stable single or dynamic bundled microtubules. We also show that some disease-related mutations closed to the hexapeptides strikingly enhance the capacity of tau to form bundles of highly dynamic microtubules. Finally, cryo-EM experiments indicate that tau proteins induce microtubule lattice defects known to soften microtubules, a primary structural change allowing microtubule-bending deformation during bundling. Overall, our results highlight novel phospho-dependent mechanisms by which tau regulates microtubule network organization. This work also reveals how abnormal modifications of tau, such as abnormal phosphorylation or mutations found in Alzheimer’s disease and related dementia, might alter cytoskeleton organization during neurodegeneration.

Frontotemporal Dementia with parkinsonism linked to chromosome 17 (FTDP-17) (OMIM # 600274) is a tauopathy caused by mutations in the MAPT gene. This gene encodes for Tau protein and its alternative splicing normally produces 6 different isoforms with three (3R) or four (4R) repeats of microtubule-binding domains, originated from the alternative splicing of exon 10 in the MAPT transcript. In normal adult brain, neurons and glia cells contain both 3R and 4R splicing isoforms in a 1:1 ratio. Several mutations in the MAPT gene impair exon 10 splicing, causing unbalance between 4R and 3R Tau isoforms (4R > 3R), leading to Tau 4R protein accumulation as insoluble neuronal deposits. Therapeutical correction of MAPT splicing isoforms balance is, in principle, possible using either exon-specific siRNAs, which degrade exon-10-containing mRNA in the cytoplasm, or splice-switching antisense RNAs, that induce skipping of exon 10 during the splicing of MAPT pre-mRNA in the nucleus. Both approaches have been explored in the Laboratory of RNA Biology and Biotechnology at CIBIO (University of Trento) using splicing reporters. Subsequently, several siRNAs and antisense RNAs have been demonstrated to efficiently engage their target (pre-)mRNA and restore 4R:3R balance in neuroblastoma cell lines in culture. Aim of the present work is to obtain pre-clinical evidence of the efficiency of the two approaches, in order to move towards clinical studies. To this purpose, I set up a disease model consisting in hiPSCs-derived neurons carrying a mutation in intron 10, where a C is substituted with a T in position 16 (MAPT IVS10+16; EBiSC, depositor Sigma-Aldrich SIGi001-A-12) and compared them with the appropriate isogenic healthy control (EBiSC; depositor Sigma-Aldrich SIGi001-A-1). 3R and 4R Tau mRNA and protein levels were evaluated at various days of differentiation and neuronal maturation. In my hands, IVS10+16 neurons showed increase 4R Tau mRNA expression at 120 days of differentiation, resembling the patient’s phenotype. The unbalance 4R:3R is reflected in the Tau protein, as assessed by Western blotting . I am presently evaluating other outcome measures of disease in this cellular model, such as synaptic impairment and electrophysiology . The Laboratory of RNA Biology and Biotechnology has reported that microRNAs (miRNAs) can be used as biomarkers of Frontotemporal Dementia (FTD). In particular, we recently reported that miR-92a-3p, miR-320a and miR-320b are misregulated in the plasma of FTD patients in comparison to healthy individuals (manuscript under review). I set out to measure these miRNAs in d120 IVS10+16 and isogenic neurons and in their conditioned medium. I found that all three miRNAs of interest were significantly increased in IVS10+16 samples compared to WT neurons, therefore representing a useful measure of therapeutical efficacy in our protocols. With the use of fluorescently labelled siRNAs, I then tackled the problem of delivering siRNA molecules to mature neurons and set up a protocol for their efficient delivery. Consequently, day120 IVS10+16 and WT neurons were transfected with different concentration of scramble and therapeutic siRNAs and the restoration of the 4R:3R Tau balance was assessed. My results suggest a promising potential for the use of isoform-specific siRNAs in FTDP-17 and possibly in other tauopathies. At the same time, I intended to validate in the same hiPSC-derived neuronal disease model, U1 and U7 chimeric splice-switching antisense RNAs that had been previously tested by plasmid transfection in neuroblastoma cell lines. To overcome the limitation represented by poor plasmid transfection efficiency in mature neurons, I encapsulated them into recombinant adeno-associated viruses (rAAVs). After having optimized the production of rAAVs and set the transduction conditions, IVS10+16 mature neurons were transduced with AAV expressing chimeric splice-switching antisense RNAs. Although neurons successfully got transduced and the cassette transcribed, there was no therapeutic effect when viruses were tested in d130 IVS10+16 neurons. I am presently exploring different experimental paradigms, to test the hypothesis that the 4R:3R unbalance can be prevented in mature neurons.

Alzheimer s disease is characterized by synaptic alterations and neurodegeneration. Histopathological hallmarks represent amyloidplaques composed of amyloid-beta (Abeta) and neurofibrillary tangles containing hyperphosphorylated tau. To determine whether synaptic changes and neurodegeneration share common pathways we established an ex vivo model using organotypic hippocampal slicecultures from amyloid precursor protein transgenic mice combined with virus-mediated expression of EGFP-tagged tau constructs. Confocal high-resolution imaging, algorithm-based evaluation of spines and live imaging was employed to determine spine changes and neurodegeneration. We report that Abeta but not tau induces spine loss and shifts spine shape from mushroom to stubby through a mechanism involving NMDA receptor (NMDAR), calcineurin and GSK-3beta activation. In contrast, Abeta alone does not cause neurodegeneration but induces toxicity by phosphorylation of wt tau in a NMDAR-dependent pathway. We show thatGSK-3beta levels are elevated in APP transgenic cultures and that inhibiting GSK-3beta activity or use of phosphorylation-blocking tau mutations prevent Abeta-induced toxicity of tau. FTDP-17 tau mutants are differentially affected by Abeta. While R406W tau shows increased toxicity in the presence of Abeta, no change is observed with P301L tau. While blocking NMDAR activity abolishes toxicity of both wt and R406W tau, the inhibition of GSK-3beta only protects against toxicity of wt tau but not of R406W tau induced by Abeta. Tau aggregation does not correlate with toxicity. We propose that Abeta-induced spine pathology and tau-dependent neurodegeneration are mediated by divergent pathways downstream of NMDA receptor activation and suggest that Abeta affects wt and R406W tau toxicity by different pathways downstream of NMDAR activity.