1、SIGNAL TRANSDUCTIONApoptosis propagates throughthe cytoplasm as trigger wavesXianrui Cheng1and James E.Ferrell Jr.1,2*Apoptosis is an evolutionarily conserved form of programmed cell death critical fordevelopment and tissue homeostasis in animals.The apoptotic control network includesseveralpositive
2、feedbackloopsthat mayallowapoptosistospreadthroughthecytoplasminself-regenerating trigger waves.We tested this possibility in cell-free Xenopus laevis eggextracts and observed apoptotic trigger waves with speeds of 30 micrometers per minute.Fractionation and inhibitor studies implicated multiple fee
3、dback loops in generating thewaves.Apoptoticoocytesandeggsexhibitedsurfacewaveswithspeedsof30micrometersper minute,which were tightly correlated with caspase activation.Thus,apoptosis spreadsthrough trigger waves in both extracts and intact cells.Our findings show howapoptosis canspread over large d
4、istances within a cell and emphasize the general importance of triggerwaves in cell signaling.Xenopus laevis eggsarelargecells,1.2mmin diameter,that are naturally arrested inmetaphase of meiosis II.The eggs ulti-mately adopt one of two fates:Either theybecome fertilized and enter the embryoniccell c
5、ycle,or they remain unfertilized and die,usually through apoptosis.Apoptosis is a rela-tively nonperturbing form of cell death andmay allow the frog to resorb old oocytes and toclean up any eggs retained in the body withoutprovoking an inflammatory response(1).Thepowerful biochemical approaches prov
6、ided bythe Xenopus system have made Xenopus eggsand extracts useful modelsystems for thestudyof apoptosis(2,3).The unusual size of the Xenopus egg raises thequestion of how an all-or-none,global processsuch as apoptosisspreads through the cell.Onepossibility is that apoptosis spreads through theeggb
7、yrandomwalk diffusion,ultimatelytakingoverallofthecytoplasm.Asecondpossibilityissuggested by the existence of multiple positiveand double-negative feedback loops in the reg-ulatorynetworkthatcontrols apoptosis(Fig.1A).These loops may allow apoptosis to propagatethrough self-regenerating trigger wave
8、s.Triggerwavesarepropagatingfronts ofchemicalactivitythat maintain a constant speed and amplitudeover large distances.They can arise when bi-stable biochemical reactions are subject to dif-fusion or,more generally,when bistability orsomething akin to bistability(e.g.,excitabilityor relaxation oscill
9、ation)is combined with aspatial coupling mechanism(e.g.,diffusion orcell-cellcommunication)(46).Familiarexamplesinclude action potentials;calcium waves;andthe spread of a fire through a field,a favorableallele through a population,or a meme throughthe internet.Trigger waves are an importantgeneral m
10、echanism for long-range biologicalcommunication,and apoptotic trigger wavesmay allow death signals to spread rapidly andwithout diminishing in amplitude,even througha cell as large as a frog egg.To distinguish between diffusive spread andtrigger waves in the propagation of apoptoticsignals,we used u
11、ndilutedcell-free extracts fromXenopus eggs(7).Such extracts can be placed inlong tubes(several millimeters)and imaged byvideomicroscopy(8,9).Overmillimeterdistances,the distinction between diffusive spread of apo-ptosis,which would slow down with increasingdistance,and trigger waves,which maintain
12、aconstant speed and amplitude,should be readilyapparent.We incubated one portion of a cycloheximide-treated interphase cytoplasmic extract with horsecytochrome c(2.4 mM)and verified that caspase-3and/or caspase-7,executioner caspases(10)thathave similar peptide sequence specificities,be-came activat
13、ed(fig.S1).A 10-kDa Texas Redconjugated dextran was added to the apoptoticextractasadiffusionmarker,andtheextractwaspipetted into a large(560-mm inner diameter)Teflon reservoir.We then took a second portionof the interphase extract,with added spermchromatin plus a chimeric protein consistingof gluta
14、thione S-transferase,green fluorescentprotein,andanuclearlocalizationsequence(GST-GFP-NLS)but no cytochrome c,and introduceditintoathin(150-mminnerdiameter)Teflontube.The nuclei act as an easily assessed indicator ofapoptosis:In nonapoptotic extracts,GST-GFP-NLS concentrates in the nuclei that form
15、fromthe sperm chromatin,whereas in an apoptoticextract,caspases attack components of the nu-clear pore(11)and allow the GST-GFP-NLS toleak out and disperse.The tube containing thisnave extract was gently inserted a short dis-tance into the reservoir containing the apoptoticextract,and the two tubes
16、were placed undermineraloil(Fig.1B).Weusedtime-lapsefluorescencemicroscopy to determine whether apoptosisspread up the thin tube in a diffusive fashion,withpropagationslowingasapoptosisproceeded,or at a constant velocity as expected for triggerwaves.As shown in movie S1 and Fig.1,B and C,apoptosis p
17、rogressed up the thin tube at a con-stant speed of 27 mm/min over a distance of sev-eralmillimeters.Infiveindependentexperiments,apoptosis always propagated linearly,withoutshowinganysignsofslowingdownordiminishing,and the average speed was 29 2 mm/min(mean SD).In contrast,the 10-kDa dye spreadonly
18、a few hundred micrometers(Fig.1B),im-plying that neither simple diffusion nor any un-intended mixing could account for the spreadof apoptosis.If the apoptotic signals are propagated bytrigger waves,one prediction is that the wavesshould be self-sustaining;that is,once the ac-tivity is established in
19、 the thin tube,continuedcontact with the reservoir of apoptotic extractwould not be required.We tested this possibilityby inserting the thin tube into the apoptoticreservoir for 20 min and then removing it andmonitoring apoptosis.The apoptotic activitypropagated from the induction terminus to thedis
20、tal terminus at a constant speed of32 mm/min,consistent with a self-sustaining process(fig.S2).This procedurewasusedforallsubsequentexperiments because it provided more reliablefocusingfortheimagingandallowedmoretubesto be imaged per experiment.A second way of detecting apoptosis is withthe fluoroge
21、nic caspase substrate carboxybenzylAsp-Glu-Val-Asprhodamine 110(Z-DEVD-R110)(12).This probe is a rhodamine derivative(R110)with twofouraminoacid(DEVD)peptideslinkedto the fluorophore.It is nonfluorescent when theDEVD-fluorophore bonds are intact but be-comes strongly fluorescent once they are hydro-
22、lyzed by caspase-3 or-7.We added the probe(2 mM)to an extract and initiated apoptosisas before.Fluorescence spread up the tube at aconstantspeed(inthisexperiment,33mm/min).Ineightexperiments,theaveragespeedwas303 mm/min(mean SD)(movie S2 and Fig.1D).We also added sperm chromatin and GST-mCherry-NLS
23、as well as Z-DEVD-R110 to com-parethepropagationofthecaspaseactivitywavewith the disappearance of the reconstitutednuclei.In this experiment,the disappearanceofthenucleilagged40minbehindthefrontofZ-DEVD-R110 fluorescence(Fig.1E);in four in-dependent experiments,the lag was 35 5 min(mean SD).The spee
24、ds of the Z-DEVD-R110and GST-mCherry-NLS wavesthe slopes of thedashed lines in Fig.1Ewere indistinguishable.We sought to understand the mechanisticbasis for the trigger waves.The apoptotic controlsystem includes multiple positive feedback loops(Fig.1A).One positive feedback circuit(des-ignated“caspa
25、se loop”in Fig.1A)involves onlycytosolic proteins,including caspases-3,-7,and-9 and XIAP(X-linked inhibitor of apoptosisprotein).Another involves the system that reg-ulates the Bak and Bax proteins(the“BH3protein loop”in Fig.1A),which,when activated,bring about mitochondrial outer membrane per-meabi
26、lization(MOMP).One long loop,in whichRESEARCHCheng and Ferrell,Science 361,607612(2018)10 August 20181 of 61Department of Chemical and Systems Biology,StanfordUniversity School of Medicine,Stanford,CA 94305-5174,USA.2Department of Biochemistry,Stanford UniversitySchool of Medicine,Stanford,CA 94305-
27、5307,USA.*Corresponding author.Email:james.ferrellstanford.eduon October 11,2018 http:/science.sciencemag.org/Downloaded from cytosolic cytochrome c stimulates caspase-3 and-7 activation and caspase activation stimulatescytochrome c release(the“MOMP/caspase/BH3protein loop”),connects the two shorter
28、 loops(Fig.1A).We tested whether the cytosolic com-ponents could support trigger waves in theabsence of mitochondria and whether the mito-chondrial components could support triggerwavesintheabsenceoftheactivationofcaspase-3and caspase-7.We fractionated a crude cytoplasmic extract(2,3)(fig.S3A)and ve
29、rified by immunoblottingthat the resulting cytosol was largely devoidof mitochondria,as indicated by the absenceof a mitochondrial marker protein,the voltage-dependent anion channel(fig.S3B).Horse cyto-chrome c(2.4 mM)was added to this cytosolicfraction,and the activity of caspase-3 and/or-7was asse
30、ssed by the chromogenic caspase assay.In agreement with previous reports(3,13),thecaspases were briskly activated(Fig.2A).Thus,mitochondria are not essential for cytochrome cinduced activation of executioner caspases inXenopus extracts.We tested whether the cytosolic extract couldsupport apoptotic t
31、rigger waves.Because nucleicannot be reconstituted in cytosolic extracts,weusedthe fluorogenicZ-DEVD-R110probe forthisexperiment.The fluorescence spread up the tubein a sublinear fashion(movie S3 and Fig.2E).Tosee if this spread was consistent with simplerandomwalkdiffusion,weidentifiedpointsalongth
32、e propagation front with equal fluorescenceintensities(fig.S4,A and B)and plotted the dis-tance squared(x2)versus time(t)(fig.S4C).Thex2-versus-t relationship was linear for at leastan hour(fig.S4C),consistent with random walkdiffusion.Thus,over this time scale and this dis-tancescale,wefoundnoevide
33、nceforatriggerwave.To further test the role of the mitochondria,we added back purified mitochondria to thecytosolic extractat a 3%volume-to-volume(v/v)ratio,which is approximately physiological esti-mated from the volumes we obtained for thevarious fractions(see fig.S3)and the concentra-tions used b
34、y others in reconstitution studies(3),and again determined whether apoptosis wouldpropagate diffusively or as a trigger wave.As ex-pected,thereconstitutedextractactivatedcaspase-3and/or-7 in response to horse cytochrome c(Fig.2C).Moreover,the reconstitution restoredthe trigger waves(movie S3 and Fig
35、.2D).Thepropagation distance increased linearly with time(Fig.2D),as it did in crude cytoplasmic extractsCheng and Ferrell,Science 361,607612(2018)10 August 20182 of 6ApoptoticextractInterphaseextract+reporternucleiDistance(m)0244872960500100015002000250030003500Time(min)Distance(m)50010001500200025
36、0030003500020406080100Time(min)0Distance(m)0204060Time(min)35002000150010005000BCDABak/BaxCaspase-3 and-7Caspase-9Mitochondrialouter membranepermeabilizationSMACXIAPInputstBidBcl-2Cyt cApoptosisBH3 proteinloopCaspaseloopMOMP/Caspase/BH3 proteinloopE0501001500500100015002000250030003500Time(min)Dista
37、nce(m)Fluorescence(a.u)30004000500060007000800030002500100001500020000250003000035000Fluorescence(a.u)27 m/min33 m/min22 m/minFig.1.Apoptosis propagates in interphase-arrested cytoplasmicX.laevis egg extracts through trigger waves.(A)The control circuit forapoptosis,as conceptualized on the basis of
38、 the present study and otherspreviously published work(2428).Cyt c,cytochrome c.(B)Time-lapsemontage of GST-GFP-NLSfilled nuclei(green)in a cytoplasmic extract in aTeflontubewithitslowerendincontactwithanapoptoticextractreservoir.Theextractinthereservoirismarkedwith10-kDadextranTexasReddye,showninma
39、genta.A time-lapse video of this experiment can be found in movie S1.(C)Correlation between timing and position of nuclear disappearance for theexperiment depicted in(B).The line is a linear least-squares fit to the data.The propagation speed(the slope of the fitted line)is 27 mm/min.(D)Kymograph im
40、age showing the spatial propagation of caspase-3 and/orcaspase-7 activity(indicated by R110 fluorescence)in a crude cytoplasmicextract.The dashed line was manually fitted to the fluorescence front,and ityielded a propagation speed of 33 mm/min.a.u.,arbitrary units.(E)R110fluorescenceandnucleardisapp
41、earance,detectedbyusingGST-mCherry-NLSas a nuclear marker,measured in the same tube.The presence of the nucleimakes the R110 fluorescence less diffuse than it is in(D).One dashed line ismanually fitted to the fluorescence front,and the other is a least-squares fit tothe nuclear data.The propagation
42、speed was 22 mm/min for both waves.RESEARCH|REPORTon October 11,2018 http:/science.sciencemag.org/Downloaded from(Fig.1),and propagation occurred at a constantspeed of 39 mm/min,somewhat faster than thatobservedincytoplasmicextracts.Thesignalprop-agated over a long distance(6000 mm)with littleloss o
43、f amplitude and no loss of speed(Fig.2D).Overa distance scale of a millimeteror so,it iseasy to distinguish a 30-mm/min trigger wavefrom diffusive spread of even a rapidly diffusingsmall molecule such as R110.However,by theend of the experiment depicted in Fig.1B,thespeed of the caspase-3 wave had f
44、allen to only15mm/min;a trigger wave anyslower than thatwould be outpaced by diffusion over the samedistance.Thus,to determine whether triggerwaves were abolished or only slowed in cyto-solicextracts,we made use oflonger tubes(upto3 cm)and longer time courses(up to 24 hours).Display of R110 fluoresc
45、ence on a pseudocolorheat map scale made it easier to distinguish thewave front at both early times and late times(Fig.2E).As was the case in Fig.2B,the speedof the wave front fell during the first 120 min,consistent with diffusive propagation,but oncethe speed reached 14 mm/min it remained con-stan
46、tformanyhours(Fig.2E).Thissuggests thatpurified cytosol is capable of generating apo-ptotic trigger waves,albeit with a substantiallylower speed than that seen in cytoplasm or incytosol supplemented with mitochondria.Trigger wave speeds were measured in eightindependentexperimentswithvariousconcentr
47、a-tions of mitochondria(Fig.2F).From curvefitting,we determined the trigger wave speed tobe half maximal at a mitochondrial concentra-tion of 1.3 0.6%(mean SE),which isestimated to be 40%of the physiologicalmitochondrial concentration in Xenopus eggs.Thus,an average concentration of mitochondriaissu
48、fficienttogenerateapoptotic trigger wavesof near-maximal speed,and the wave speedwouldbeexpectedtodropinmitochondrion-poorregions of the cytoplasm.Apoptosis almost always initiated first at theend of the tube that was dipped in the apoptoticextract(Figs.1 and 2 and fig.S2).However,inexperiments with
49、 either cytoplasmic extracts orreconstitutedextracts,morethanhalfofthetime(in 12 of 22 or 21 of 37 tubes,respectively)a sec-ond spontaneous apoptotic wave emerged else-where in the tube(movie S4 and fig.S5A).Thevelocitiesoftheinducedandspontaneoustriggerwaves were indistinguishable(44 mm/min in thee
50、xample shown in fig.S5A),indicating that theyprobably represent the same basic phenomenon.The entire tube of extract usually became apo-ptotic,as assessed by the Z-DEVD-R110 probe,after 2 hours of incubation(fig.S5)and invari-ablyby4hours.Thisglobalactivationofcaspase-3and-7 did not occur in cytosol
