of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Time Course of Polyglutamine AggregateBody Formation and Cell Death: EnhancedGrowth in Nucleus and an Interval forCell Death I. Toyoshima, 1 * M. Sugawara, 1 K. Kato, 1 C. Wada, 1 T. Shimohata, 2 R. Koide, 2 O. Onodera, 2 and S. Tsuji 2 1 Department of Internal Medicine, Akita University School of Medicine, Akita, Japan 2 Department of Neurology, Brain Research Institute, Niigata University School of Medicine, Niigata, Japan Polyglutamine (polyQ) aggregate bodies are a hallmark ofdentatorubral-pallidoluysian atrophy and related neuro-degenerative disorders, although the relationship be-tween aggregate body formation and cell death is notclear. We analyzed the kinetics of polyQ aggregate for-mation and the time intervals for cell death, trackingindividual cells using fluorescence video microscopy, forthe first time. Expanded polyQ tracts of atrophin-1 withor without nuclear localization signal (NLS) labeled withgreen fluorescent protein (GFP) were constructed,Q57NLS/GFP and Q56/GFP, respectively. All of theQ57NLS/GFP aggregate bodies were in nuclei, and all ofthe Q56/GFP aggregate bodies were in cytoplasm. Ag-gregates of Q56/GFP were larger than those of Q57NLS/ GFP. Surprisingly, a kinetic analysis showed that thelatter grew 5.37 times faster than the former. The timeinterval between transfection and cell death was shorterin Q57NLS/GFP, but the time between the end of therapid growing phase of aggregation and the start ofthe cell death process did not show a significant differ-ence. Aggregate growth was confirmed to correspond tothe accumulated free polyQ by the time of starting ag-gregation. These findings suggest that aggregate bodyformation induced by expanded polyQ stretches is aself-limiting process and is enhanced by factor(s) in nu-clei, whereas it is not tightly bound to the cell deathprocess. ©  2002 Wiley-Liss, Inc. Key words:  DRPLA; green fluorescent protein; polyglu-tamine stretch; neuronal intranuclear inclusion; fluores-cence video microscopy Expanded CAG trinucleotide repeats coding poly-glutamine (polyQ) stretches cause neuronal cell death in atleast nine neuronal degenerative diseases: Huntington dis-ease (HD); spinal and bulbar muscular atrophy (SBMA);spinocerebellar ataxia (SCA)-1, -2, -3/Machado-Josephdisease (MJD), -6, -7, and -17; and dentatorubral-pallidoluysian atrophy (DRPLA; for review see Zoghbiand Orr, 2000; Gusella and MacDonald, 2000; Orr, 2001).Neuronal intranuclear inclusions (NIIs) were found to becritical for neuronal cell death in patients or transgenicmice with HD (Davies et al., 1997; DiFiglia et al., 1997),SBMA (Li et al., 1998), SCA1 (Skinner et al., 1997),SCA2 (Koyano et al., 1999), SCA3/MJD (Paulson et al.,1997), SCA7 (Holmberg et al., 1998), SCA17 (Nakamuraet al., 2001), and DRPLA (Igarashi et al., 1998). Contraryto this, SCA6 has exclusively cytoplasmic inclusions (Ish-ikawa et al., 1999). It is still not clear whether full-lengthproteins or polyQ stretches are responsible for neuronalcell death. Full-length ataxin-1 is reported to be effectivefor NII formation (Skinner et al., 1997), but 3% of theN-terminus region of huntingtin is enough for NIIs inHD (Davies et al., 1997), and only truncated ataxin-3resulted in neurodegeneration in SCA3/MJD (Ikeda et al.,1996).DRPLA is an autosomal dominant neurodegenera-tive disease characterized by myoclonus epilepsy, choreo-athetosis, and dementia (Naito and Oyanagi, 1982). Un-stable expansion of CAG coding polyQ was detected inthe  DRPLA  gene of patients (Koide et al., 1994; Nagafu-chi et al., 1994; Tsuji, 1999). Expanded polyQ stretcheswith shorter flanking regions induced aggregates, whereasfull-length atrophin-1, the gene product of   DRPLA , con-taining expanded polyQ did not result in aggregate for-mation in nuclei or cytoplasm (Igarashi et al., 1998).Green fluorescent protein (GFP)-tagged polyQ stretchesof atrophin-1 developed polyQ length- and time-dependent aggregate formation and cell death in rat cer-ebellar granule cells (Moulder et al., 1999). Transgenic Contract grant sponsor: Ministry of Education, Culture, Sports, Science andTechnology, Japan.*Correspondence to: Itaru Toyoshima, MD, PhD, Department of InternalMedicine, Akita University School of Medicine, 1-1-1 Hondo, Akita010-8543, Japan. E-mail: 8 November 2001; Revised 27 December 2001; Accepted 18 January 2002 Publishedonline16April2002inWileyInterScience .DOI:10.1002/jnr.10233  Journal of Neuroscience Research 68:442–448 (2002) ©  2002 Wiley-Liss, Inc.  mice expressing human atrophin-1 showed massive dep-osition of 120 kDa fragment harboring the N-terminusportion and polyQ tract in nuclei (Schilling et al., 1999).Transgenic mice with 129 repeats of glutamine showedneurological symptoms similar to those in human disease(Sato et al., 1999).Recent experiments using yeast two-hybrid revealedpreferential association of expanded polyQ in atrophin-1with human TATA binding protein (TBP)-associated fac-tor, TAF II 130 (Shimohata et al., 2000). Reports on theassociation of polyQ with nuclear proteins have beenmade for polyQ diseases (Fernandez-Funez et al., 2000;McCampbell et al., 2000; Shibata et al., 2000; Wood et al.,2000; Holbert et al., 2001; Nucifora et al., 2001).In this study, we tested the signi fi cance of the site of aggregate body formation induced by expanded polyQstretches with short  fl anking regions of atrophin-1. Weanalyzed the time course of polyQ aggregate formationand cell death, tracking individual cells using GFP tag and fl uorescence video microscopy, and found a different timecourse and kinetics of aggregate formation and cell deathbetween nuclear and cytoplasmic aggregate bodies. MATERIALS AND METHODSDNA Construct A part of the DRPLA gene containing a polyQ stretchwith short  fl anking regions of both sides was inserted into GFPvector pEGFPN1 (Clontech, Palo Alto, CA) as previously re-ported (Onodera et al., 1997; Igarashi et al., 1998). Q19GFP hada normal length of polyQ, 19, and Q56GFP had 56 repeats of glutamine. Nuclear localization signal (NLS) of SV40 T antigen(PKKKRKV) was added to Q57NLS/GFP with 57 repeats of glutamine (Shimohata et al., 2000). Cell Culture and Lipofection COS-7 cells were plated in Daigo-T medium withoutphenol red (Wako, Tokyo, Japan) supplemented with 10% fetalcalf serum (Hyclone, Logan, UT) 24 hr before transfection at10 5 cells per glass-bottomed Microwell dish (MatTek, Ashland,MA). Lipofection of 0.5   g plasmid DNA with SuperFect(Qiagen, Hilden, Germany) was carried out according to themanufacturer  ’ s instructions. Fluorescence Video Microscopy Most of an inverted microscope (Axiovert; Carl Zeiss,Tokyo, Japan) was covered by a chamber with the temperatureadjusted to 37 ° C    0.2 ° C. Transfected cells in the culture dishwere kept in a small chamber with a continuous  fl ow of 5%carbon dioxide. Emitted light for   fl uorescence microscopy froma mercury lamp was attenuated by AttoArc (Carl Zeiss) to 1%and further with a neutral  fi lter. Final attenuation of the lightsource was to 0.003% of original. This enabled us to trackindividual cells for 48 hr without obvious photobleaching of GFP aggregates or photodamage to cell function. An air-cooled3CCD video camera (Hamamatsu Photonics, Hamamatsu, Ja-pan) captured emitted light and accumulated photons on tip for 45 sec. Each frame was stored in a time-lapse S-VHS videorecorder (Matsushita, Tokyo, Japan) at a rate of 1/240, 1 frameby 4 sec. Dual use of   fl uorescence and phase microscopy wasemployed to visualize cell morphology other than GFP-illuminated structures. Morphometry Volume of aggregate was estimated under the assumptionthat the aggregates form ellipsoids of revolution. Long and shortaxes were calculated with morphometric software, NIH Imageversion 1.62 (free NIH software), after loading images of illu-minated aggregates on S-VHS tapes to a Macintosh computer through a video capture board, CG7 (Scion, Frederick, MD).Three-dimensional reconstruction of aggregates was carried outusing a confocal microscope (Carl Zeiss) to con fi rm that thevolume of aggregates was estimated as ellipsoids of revolution.Nuclei were visualized with Syto16 (Molecular Probes, Eugene,OR), a red dye staining DNA.The time intervals between lipofection, initiation of ag-gregation, end of rapid growing phase of aggregate bodies, andstart of cell death process were measured by reading the timerecords on video frames generated by an SVHS time-lapse videorecorder (Matsushita, Tokyo, Japan). Time Course of Free PolyQ in Nuclei Mean  fl uorescence density in the nuclei other than aggre-gates was measured with NIH Image between the start of aggregation (time 0) and the end of the rapid-growth phase(time 100) in the case of intranuclear Q57NLS/GFP aggregateformation. The highest density was adjusted to 100 and thelowest to 0. Time course of aggregate growing was normalizedfrom time 0 to time 100 and size 0 to 100 and plotted onto thesame format. Mean and standard deviation in each 10% succes-sive time interval were calculated. Statistical Analysis Difference in each parameter was analyzed by two-tailedStudent ’ s  t  -test. RESULTSFluorescence Video Microscopy PolyQ aggregates were well visualized and capturedby a cooled CCD video camera. Illuminated aggregateswith GFP tag (Fig. 1a) were also recognized on phasecontrast microscopy (Fig. 1b). Differential interferencecontrast microscopy revealed fuzzy and spreading fringesfrom aggregates (Fig. 1c). Confocal microscopy showed arevoluted spheroidal structure. Most aggregate bodies bothin nuclei and in cytoplasm were attached to or were closeto the nuclear membrane on three-dimensional analysis(Fig. 1d). Parameters of Aggregate Body Formation Pro fi les of the time course of aggregate body forma-tion were obtained from 24 Q57NLS/GFP aggregates and22 Q56/GFP aggregates. All of the Q57NLS/GFP aggre-gates were in nuclei, and all of the Q56/GFP aggregatebodies were in cytoplasm. Only solitary aggregates wereanalyzed.Figure 2 shows one of the typical records of aggre-gate formation. At time 0, an illuminated aggregate ap- Time Course of PolyQ Aggregate Formation 443  peared in the nucleus and rapidly grew for 30 min. After that, the size of the aggregate was stationary, and abruptdetachment, a sign of cell death, occurred at 152 min.Figure 3 shows a plot of the calculated volume of anaggregate body in sequential video frames against the timeafter start of aggregate formation. Maximal growing ratewas calculated from the plot. End point of aggregategrowing was calculated as the crossing point of the rapid-growth phase and the stationary phase. The rapid-growthphase was  fi tted to a three-degree polynomial equationand the stationary phase to a double reciprocal line withthe assistance of the statistical software StatView version4.0 (Abacus, Berkeley, CA) and KaleidaGraph version 3.0(Synergy, Reading, PA; Fig. 3). Volume of Aggregate Bodies Maximal volumes of aggregate bodies withQ57NLS/GFP and Q56/GFP transfection were 146   122 attoliters (al) and 241    140 al, respectively (Fig. 4a).Mean size of Q56/GFP aggregates was 1.65 times greater than that of Q57NLS/GFP aggregates. Volume of nucleiof COS-7 was estimated as 1,240    667 al. Speed of Aggregate Body Enlargement Maximal speeds of enlargement of Q57NLS/GFPaggregates and Q56/GFP aggregates were 4.93    4.67al/min and 2.31    1.31 al/min, respectively (Fig. 4b).Q57NLS/GFP aggregate bodies grew 2.02 times morerapidly than Q56/GFP aggregates. Only Q57NLS/GFPaggregates showed a speed over 10 al/min.Both Q57NLS/GFP and Q56/GFP aggregatesshowed a correlation between maximal volume and en-largement speed with two different regression lines (Fig.4c). This indicates that Q57NLS/GFP bodies grew 5.37times faster than Q56/GFP aggregates after compensationfor maximal size. Time Course of Aggregate Body Formation Time interval between lipofection and the start of aggregate body formation with Q57NLS/GFP was signif-icantly shorter than that in Q56/GFP aggregation;Q57NLS/GFP aggregation was 2,430    1,300 min, andQ56/GFP aggregation was 3,270    1,480 min (Fig. 5a).The shortest time was 10 hr in both, but some aggregatesstarted had already at that time. Time from lipofection tocell death in Q57NLS/GFP was shorter than that inQ56/GFP; the former was 3,190    1,260, and the latter was 3,880    1,350 min.The duration of the rapid-growth phase inQ57NLS/GFP and Q56/GFP aggregates was 57   64 min and 126  70 min, respectively (Fig. 5b), the latter being longer than the former. The interval between theend of the aggregate growing phase and the start of the celldeath process in Q57NLS/GFP and Q56/GFP was 519  322 min and 400    416 min, respectively (Fig. 5c). Timefrom initiation of aggregation to cell death also did notshow a signi fi cant difference; 578  325 min in Q57NLS/GFP and 529  405 min in Q56/GFP. These intervals didnot correlate with any parameters in this study. Time Course of Free PolyQ in Nuclei To con fi rm that the accumulated polyQ makes ag-gregates, we analyzed the mean  fl uorescence density in thenuclei other than aggregates, in the case of intranuclear Q57NLS/GFP aggregate body formation (Fig. 6). Nor-malized  fl uorescence densities in nuclei clearly showed thecorresponding decrease in aggregate body growth as asigmoid curve. DISCUSSION Here we have shown the time course of polyQaggregate body formation and cell death with or withoutNLS. Both showed sigmoid curves for aggregate bodyvolume against time. After the end of the growing phase,aggregates maintained constant size until the start of thecell death process. The presence of this stationary periodallows the chance to observe aggregate bodies in livingcells.COS-7 cells harboring intranuclear aggregate bodieswith Q57NLS/GFP showed a shorter time for cell deaththan that with cytoplasmic aggregate bodies of Q56/GFP.A major contribution to the shorter period for cell death isa shorter incubation time for the start of aggregation fromlipofection. The time lag between lipofection and theappearance of aggregate bodies might be explained bylocal concentration of free polyQ for the start of aggregateformation. Amyloid  fi bril formation in Alzheimer  ’ s disease Fig. 1. Images of aggregate bodies of Q56/GFP ( a  –  c ) and Q57NLS/GFP ( d ) were obtained by  fl uorescence (a), phase contrast (b), differ-ential interference contrast (DIC; c), and confocal (d) microscopy. The fl uorescent image well re fl ects the shape of an aggregate body visualizedwith a DIC microscope. Confocal image con fi rmed the tangentiallyround shape of an aggregate body in the nucleus. Scale bars    10   min a (for a  –  c); 10   m in d. 444 Toyoshima et al.  (Harper and Lansbury, 1997) and polyQ aggregates in HD(Scherzinger et al., 1999) are reported to be dependent onthe concentration of soluble   -amyloid protein and freepolyQ, respectively. If the production rates of Q57NLS/GFP and Q56/GFP in the cells are similar, the NLSpromotes the transport of Q57NLS/GFP peptides fromcytoplasm to the nucleus to accumulate polyQ in the smallcompartment, whereas Q56/GFP is distributed diffuselyin the whole cell, with a lower concentration of freepolyQ.Nucleation-dependent in vitro  fi bril formation pro-ceeds in a nonlinear sigmoid curve, indicating a self-limiting and thermoequilibrating process (Harper andLansbury, 1997). Aggregate growing in a sigmoid curveand the corresponding decline of accumulated free polyQby the time of start of aggregation was con fi rmed in thisstudy. These  fi ndings suggest for the  fi rst time that aggre-gate body formation induced by expanded polyQ stretchesis a self-limiting process in living cells.Seeding enhances  fi bril formation (Harper and Lans-bury, 1997); there might be an enhancement factor for nucleation and seed formation in nuclei. Aggregate body-associated proteins have the potential for enhancement of aggregation, leucine-rich acidic nuclear protein (LANP) inSCA-1 (Matilla et al., 1997), ubiquitin-proteasome path-way (Cummings et al., 1998), and CREB binding protein(CBP) in HD (Kazantsev et al., 1999), although inhibitionof proteasome function resulted in enhancement of aggre-gation (Chai et al., 1999).A higher growing rate relative to maximal size of intranuclear aggregate bodies was revealed in this study.This may suggest a different acceleration mechanism of aggregate growing between intranuclear and cytoplas-mic inclusion. TAF II 130 is a major binding protein of polyQ in nuclei that preferentially binds expandedpolyQ (Shimohata et al., 2000). All the aggregate bodiesin nuclei and cytoplasm were reported to colocalizewith TAF II 130, although they were exclusively distrib-uted in nuclei in the control. The preferential distribu-tion of TAF II 130 in nuclei may enhance seeding andgrowth of aggregates.The difference in maximal volume of aggregate bod-ies between Q57NLS/GFP and Q56/GFP may be thedifference in the locus of aggregate bodies. Accumulatedfree polyQ by the time of start of aggregation may corre-spond with the maximal size of aggregation, but no cor-relation was observed between maximal size and time lagby the start aggregation. Rigidity of nuclear matrix that isresistant to high-salt treatment (Skinner et al., 1997) maysuppress the growth of aggregates. The limit of transpor- Fig. 2. Time sequence of an aggregate body formation. An aggregate body of Q57NLS/GFPappeared in the nucleus at time 0, rapidly grew by 32 min, and was stationary in the subsequentperiods. al, Attoliters. Scale bar     20   mFig. 3. Parameters of aggregate body formation. Time Course of PolyQ Aggregate Formation 445
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks