Functional MAPT haplotypes: Bridging the gap between genotype and neuropathology

  • Published on

  • View

  • Download


<ul><li><p>Revie</p><p>idg</p><p>sevelt</p><p>Received 23 January 2007; revised 17 April 2007; accepted 27 April 2007Contents</p><p>Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 1Cis-elements and trans-acting factors affect splicing. . . . 2</p><p>FTDP-17 mutations reveal MAPT regulatory splicingsequences . . . . . . . . . . . . . . . . . . . . . . . . 2Alternative splicing regulation at the 5 splicesiteStemloop theory. . . . . . . . . . . . . . . . . 4Alternative splicing regulation at the 5 splicesiteLinear sequence theory . . . . . . . . . . . . . . 4</p><p>Tauopathy mutations . . . . . . . . . . . . . . . . . . . . 4</p><p>It is estimated that 24 million people currently have dementiaworldwide and that this number will double to 42 million by 2020(Ferri et al., 2005). Dementias and movement disorders likeAlzheimers disease (AD) and Parkinsons disease (PD) areexpected to surpass cancer as the second most common cause ofdeath by 2040 (World Health Organization, It is becomingclear that neurodegenerative diseases affecting ageing populationsare an increasingly important public health concern.</p><p>Common to many neurodegenerative diseases is the accumula-tion of abnormal protein in cells affected by neurodegeneration.Characteristic neuropathological aggregations are seen in PD inwhich aggregates of -synuclein form Lewy-bodies and in ADwhere A peptides form neuritic plaques. Intracellular aggregationsThe microtubule-associated protein tau (MAPT) locus has long beenassociated with sporadic neurodegenerative disease, notably progres-sive supranuclear palsy and corticobasal degeneration, and morerecently with Alzheimers disease and Parkinsons disease. However,the functional biological mechanisms behind the genetic associationhave only now started to emerge. The genomic architecture in theregion spanning MAPT is highly complex, and includes a 1.8 Mbblock of linkage disequilibrium (LD). The region is divided into twomajor haplotypes, H1 and H2, defined by numerous single nucleotidepolymorphisms and a 900 kb inversion which suppresses recombina-tion. Fine mapping of the MAPT region has identified sub-clades of theMAPT H1 haplotype which are specifically associated with neurode-generative disease. Here we briefly review the role ofMAPT in sporadicand familial neurodegenerative disease, and then discuss recent workwhich, for the first time, proposes functional mechanisms to linkMAPT haplotypes with the neuropathology seen in patients. 2007 Elsevier Inc. All rights reserved.</p><p>Keywords: MAPT; H1 haplotype; Progressive supranuclear palsy; Tauo-pathy; Splicing; Gene expression; Functional polymorphisms; Suscept-ibility mechanismsAvailable online 5 May 2007Functional MAPT haplotypes: Brand neuropathology</p><p>Tara M. Caffrey and Richard Wade-Martins</p><p>The Wellcome Trust Centre for Human Genetics, University of Oxford, Roo Corresponding author. Current address: Department of Physiology,Anatomy and Genetics, Le Gros Clark Building, University of Oxford,South Parks Road Oxford, OX1 3QX UK. Fax: +44 01865 287501.</p><p>E-mail addresses:, (R. Wade-Martins).</p><p>0969-9961/$ - see front matter 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.nbd.2007.04.006w</p><p>ing the gap between genotype</p><p>Drive, Oxford, OX3 7BN UK</p><p> of Disease 27 (2007) 110</p><p>MAPT haplotype association with sporadic tauopathies . . 4The MAPT H1 and H2 haplotypes . . . . . . . . . . . 4MAPT H1 sub-haplotypes . . . . . . . . . . . . . . . . 6</p><p>Haplotype promoter strength . . . . . . . . . . . . . . . . 6Promoter strength assayed by reporter gene . . . . . . . 6Whole locus analysis of MAPT expression . . . . . . . 7</p><p>Specific tau isoform expression. . . . . . . . . . . . . . . 7Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 8References . . . . . . . . . . . . . . . . . . . . . . . . . . . 8</p><p>Introduction</p><p>Future demographic projections are predicting the number ofpeople aged 60 years or greater to reach nearly 1.2 billion by 2025( Correspondingly, this will in-crease the prevalence of diseases affecting ageing populationssuch as the neurodegenerative dementias and movement disorders.of abnormally hyperphosphorylated microtubule-associated proteintau (tau), known as neurofibrillary tangles (NFTs), are the majorpathological feature of tauopathies, a diverse group of neurode-generative dementias and movement disorders. Tauopathies includediseases such as progressive supranuclear palsy (PSP), AD, fronto-</p></li><li><p>temporal dementia with parkinsonism linked to chromosome 17(FTDP-17), corticobasal degeneration (CBD), argyrophilic graindisease (AGD) and Picks disease (PiD). Identification of tau as themajor component in neurofibrillary tangles positioned the MAPTlocus as a leading causal candidate gene in these neurodegenerativediseases.</p><p>The MAPT gene locus is located on chromosome 17q21 andconsists of 16 exons (Andreadis et al., 1992) (Fig. 1). Tau protein isexpressed predominantly in the neurons of the peripheral andcentral nervous systems where it has a role in building and sta-bilizing microtubules, neuronal polarity and signal transduction (forreview see Shahani and Brandt, 2002). In the adult human centralnervous system, six proteins isoforms are generated by alternativesplicing of exons 2, 3 and 10 (Goedert et al., 1988; Goedert et al.,1989a,b; Andreadis et al., 1992) (Fig. 1). Exons 2 and 3 code forshort amino terminal inserts that which part of the acidic projectiondomain that may interact with the plasma membrane (Brandt et al.,1995) and regulate spacing between microtubules (Chen et al.,1992). The interaction of tau with microtubules is mediated by themicrotubule binding domains formed by imperfect repeats of 31 or32 amino acids at the carboxyl terminus, encoded by exons 912(Goedert et al., 1988; Gustke et al., 1994; Trinczek et al., 1995). Thealternative splicing of exon 10 generates proteins with either three(exon 10; 3R tau) or four (exon 10+; 4R tau) microtubule bindingdomains (Fig. 1).</p><p>Neurofibrillary tangles are found in AD, PSP, FTDP-17, CBD,AGD and PiD. The filamentous tau that aggregates to form NFTs inAD consists mainly of paired helical filaments 1020 mm in dia-</p><p>composition differs between tauopathies, being comprised of both3R tau and 4R tau in AD (Sergeant et al., 1997) and predominately3R tau in PiD (Delacourte et al., 1996). The major component of theinsoluble tangles in PSP, CBD, AGD and FTDP-17 is 4R tau (BueeScherrer et al., 1996; Arai et al., 2001; Togo et al., 2002). Conse-quently, these diseases are referred to as 4R tauopathies.</p><p>In 1998 the discovery of multiple MAPT mutations in FTDP-17provided the first evidence that changes in tau alone could causeneurodegenerative disease. In addition, the FTDP-17 splice sitemutations within MAPT that increase the inclusion of exon 10 intranscripts show an imbalance in the ratio of 3R and 4R tau isoformsis sufficient to cause disease (Hutton et al., 1998; Spillantini et al.,1998; DSouza et al., 1999). In-depth analysis of the region aroundexon 10 containing these rare pathogenic mutations has uncoveredseveral functional cis-elements in the MAPT sequence which maygive insight into how intronic and non-coding mutations orpolymorphisms may cause sporadic tauopathies.</p><p>Cis-elements and trans-acting factors affect splicing</p><p>FTDP-17 mutations reveal MAPT regulatory splicing sequencesThe alternative splicing of MAPT exon 10 is moderated in part</p><p>by the nature of the intronic sequences upstream and downstream ofthe exon (Gao et al., 2000). Neither the 3 nor the 5 exon 10 splicesites conform to the splicing consensus sequence, differing from thecanonical sites by one and four nucleotides, respectively, (An-dreadis et al., 1992) resulting in weak binding of the U1 snRNP(Hutton et al., 1998; Spillantini et al., 1998; Jiang et al., 2000). In</p><p>tivelyats eaus sys</p><p>2 T.M. Caffrey, R. Wade-Martins / Neurobiology of Disease 27 (2007) 110meter (Kidd, 1963; Crowther, 1990) whereas the PSP tangles arecomposed of 15- to 18-mm-diameter straight filaments (Tellez-Nagel and Wisniewski, 1973; Powell et al., 1974). Tangle isoform</p><p>Fig. 1.MAPT locus consists of 16 exons of which exons 2, 3 and 10 are alternathe N-terminal projection domain (black stripes). Exons 912 imperfect repeare absent from the CNS, but exon 4A is expressed in the peripheral nervo</p><p>splicing of exons 2 and 3 results in proteins with 0, 1 or 2 N-terminal inserts (ON, 1Nbinding repeats (3R or 4R tau protein).vivo, the 5 splice site is strengthened by mutations S305N and exon10 +3, resulting in increased exon 10 inclusion in transcripts(Hutton et al., 1998; Spillantini et al., 1998) (Fig. 2A). Weak splice</p><p>spliced in the adult CNS (grey or black stripes). Top: Exons 2 and 3 code forch coding for a microtubule-binding domain (grey). Exons 4A and 8 (white)tem. Bottom: Six tau isoforms are expressed in the adult CNS. Alternative</p><p>, 2N). Splicing of exon 10 generates proteins with either 3 or 4 microtubule-</p></li><li><p>eurobT.M. Caffrey, R. Wade-Martins / Nsites around exon 10 may allow subtle spatial and temporal regula-tion of 3R and 4R isoforms, by allowing splicing regulation througha collection of local splicing enhancer and silencer elements.</p><p>Systematic deletions across exon 10 have revealed that nearlyall the exon is involved in splicing regulation (DSouza andSchellenberg, 2000) (Fig. 2A). At the 5 of exon 10, three exonicsplicing enhancers have been identified: a SC35-like enhancer, apolypurine enhancer (PPE) and an AC-rich element (ACE)(DSouza et al., 1999; DSouza and Schellenberg, 2000).</p><p>Fig. 2. Cis-elements affecting splicing ofMAPT exon 10. Silencer (red) and enhancsequence is designated by capital letters and intronic sequence is indicated by small c3: a SC35-like enhancer, a polypurine enhancer (PPE), an A/C-rich enhancer (ACintronic splicing silencer (ISS) and an intronic splicingmodulator (ISM). Select FTDthe tau exon 10 5 splice site. Two of the variant stemloop structures proposed to real., 1998). (C) The linear model of tau exon 10 5 splicing regulation postulates that balternative splicing. Shown here is a trans-acting factor bound to the ISM. This ISMfrom binding the silencing cis-element, thereby allowing access of the U1 snRNP3iology of Disease 27 (2007) 110The polypurine enhancer has a high purine content and is theregion affected by the mutations N279K and K280del. N279Kenhances exon 10 splicing by adding an extra AAG-repeat thatstrengthens the PPE (DSouza et al., 1999; Hasegawa et al., 1999).In contrast, the K280del abolishes exon 10 splicing by removing anAAG-repeat (DSouza et al., 1999; Rizzu et al., 1999). Trans-actingfactors that bind the PPE include SF2/ASF, Tra2, SRp30c andSRp54 (Jiang et al., 2003; Kondo et al., 2004; DSouza andSchellenberg, 2006).</p><p>er (green) sequences within and surrounding exon 10 are indicated. Exon 10ase letters. (A) Several cis-elements affectMAPTexon 10 splicing, listed 5 toE), an exonic splicing silencer (ESS), an exonic splicing enhancer (ESE), anP-17 and PSPmutations are also shown. (B) Alternative splicing regulation atgulate splicing at the tau exon 5 splice site (Hutton et al., 1998; Spillantini etinding of trans-acting factors to the cis-regulatory elements mediates exon 10binding protein sterically hinders a trans-acting silencer (ISS binding protein)to the 5 splice site (D'Souza and Schellenberg, 2002, 2005).</p></li><li><p>eurobDownstream of the PPE is a putative A/C rich enhancer. Thedeletion of nucleotides in this region results mainly in decreasedinclusion of exon 10 in transcripts. This region is affected by thesilent mutation CTTNCTC (L284L) that is thought to exert itseffect by increasing the content of AC nucleotides (DSouza et al.,1999; DSouza and Schellenberg, 2000).</p><p>Further downstream of these enhancers, three different patho-genic mutations have been found at the same codon. The missenseN296H and silent AATNAAC (N296N) mutations enhance exon 10inclusion, while a deletion at the same codon (N296del) is eitherneutral or enhances exon 10 splicing (Spillantini et al., 2000;Grover et al., 2002; Yoshida et al., 2002). These mutations havebeen proposed to either disrupt an 18-nucleotide silencer or changethe silencer into an enhancer (DSouza and Schellenberg, 2000).</p><p>Alternative splicing regulation at the 5 splicesiteStemloop theory</p><p>Mutations in the MAPT exon 10 5 splice site have highlightedthe silencing function of this region on exon 10 inclusion. Onetheory proposes that these mutations disrupt a stemloop structurethat blocks U1 snRNP or another factor from binding, thus pre-venting inclusion of exon 10 in transcripts (Hutton et al., 1998;Spillantini et al., 1998) (Fig. 2B). Mutations within the predictedstem increase inclusion of exon 10 in transcripts, while mutationsgenerated in the putative loop have no effect on splicing (Grover etal., 1999). In addition, stabilizing the stemloop by either insertingextra bases to elongate the stem or by generating more stable pairingalong the stem, leads to a decrease in exon 10 inclusion (Grover etal., 1999; Donahue et al., 2006). Stemloop secondary structureshave been indicated through in vitro gel migration assays (Grover etal., 1999), by exposure of transcripts to RNase H (Jiang et al., 2000)and by UV melting experiments (Donahue et al., 2006).</p><p>While secondary structures can be observed in in vitro ex-periments (Grover et al., 1999; Varani et al., 1999), it is not clear ifthe same structure would form in vivo because in vivo there aremultiple splicing regulatory proteins that coat the pre-mRNApossibly preventing secondary structures from forming (DSouzaand Schellenberg, 2000). Additionally, the stemloop structuresproposed are often ambiguous and do not agree between groups(Fig. 2B). Importantly, regulation through the stemloop structuredoes not necessarily allow for regulation by trans-acting factors thatmay affect differential splicing seen throughout development.</p><p>Alternative splicing regulation at the 5 splicesiteLinear sequence theory</p><p>An alternative hypothesis is that the inhibitory sequence inintron 10 is a linear cis-acting element that binds trans-actingsplicing factors (DSouza et al., 1999; DSouza and Schellenberg,2000) (Fig. 2C). Some of the mutations that destabilize the stem-loop structure may also function to increase exon 10 inclusion byincreasing affinity for the U1 and U6 snRNPs which have both beenimplicated in exon 10 splicing inclusion. A proposed intronicsilencer at position exon 10+11 to exon 10+18 retains its functioneven after it has been translocated to different positions, including aposition within exon 10 as well as a heterologous setting, where thecomplementary bases are not present to form a stemloop structure(DSouza and Schellenberg, 2002). An intronic splicing modulatorlocated at position exon 10+19 to exon 10+26, acts in conjunctionwith the intron...</p></li></ul>


View more >