(1.10.1)--遗传学遗传学Cytoplasmicinheritanceofparent-o.pdf
rspb.royalsocietypublishing.orgResearchCite this article:Shirokawa Y,Shimada M.2016 Cytoplasmic inheritance of parentoffspring cell structure in the clonal diatomCyclotella meneghiniana.Proc.R.Soc.B 283:20161632.http:/dx.doi.org/10.1098/rspb.2016.1632Received:21 July 2016Accepted:12 October 2016Subject Areas:cellular biology,evolution,microbiologyKeywords:centric diatom,cytoplasmic inheritance,non-genetic heritability,quantitative genetics,structural inheritance,rapid evolutionAuthor for correspondence:Yuka Shirokawae-mail:Electronic supplementary material is availableonline at https:/dx.doi.org/10.6084/m9.fig-share.c.3552210.Cytoplasmic inheritance of parentoffspring cell structure in the clonaldiatom Cyclotella meneghinianaYuka Shirokawa and Masakazu ShimadaDepartment of General Systems Studies,The University of Tokyo,3-8-1 Komaba,Meguro-ku,Tokyo 153-8902,JapanYS,0000-0003-4825-7190In cytoplasmic inheritance,structural states of a parent cell could be trans-mitted to offspring cells via two mechanisms.The first is referred to as thehangover of parent structure,where the structure itself remains and faith-fully transmits within offspring cells;the second is structural inheritance,wherein the parent structure functions as a template for development ofnew offspring structure.We estimated to what extent the parent structureaffects the development of offspring structure by structural inheritance,using a clone of the diatom Cyclotella meneghiniana.The cell has two siliceousvalves(a cell wall part at both cell poles):one is inherited from the parentand the other is newly formed.We estimated cytoplasmic heritability bycomparing valve traits(central fultoportulae(CTFP),striae,central area,and cell diameter)of parent and new offspring valves,using single-cellisolation and valve labelling.Parentoffspring valve trait regressionsshowed that all traits,except CTFP,were significantly correlated.We formu-lated a quantitative genetic model considering the diatom inheritance systemand revealed short-term rapid evolution compared with other inheritancesystems.Diatom structural inheritance will have evolved to enable clonalpopulations to rapidly acquire and maintain suitable structures for temporalchanges in environments and life-cycle stages.1.IntroductionInheritance is one of the most important processes of evolution and showsdiverse mechanisms 16.Recently,there has been increasing evidence fornon-genetic inherited information without DNA sequence substitution by var-ious mechanisms such as DNA methylation,non-coding RNA,self-sustainingfeedback loop of gene expression and maternal effects 36.Accordingly,theheritability concept has become expanded to incorporate non-genetic inheri-tance,and has been unified in a general quantitative genetics model thataccounts for the various mechanisms of inheritance 79.To consider non-genetic inheritance for species with single-cell individuals,the states of parent cell structures could be transferred to offspring via twomechanisms.The first mechanism is by faithful transmission of the traits asresidual parent structures in the offspring cells per se(referred to as the hang-over of parent structure).The second mechanism is by parent structureaffecting the development of offspring structure and contributing to producea similar structure,known as structural inheritance.Structural inheritance is acellular transgenerational effect in which existing cellular structures act as tem-plates for the production of similar structures,then the new structures becomethe components of daughter cells 1.Examples of known templates for newstructures include prion in yeast 10,pre-existing cortical structures in ciliates11,12 and the tooth structure of the protozoan Diffugia corona 13.Structuralinheritance has inspired new concepts of evolutionary biology,including thecontribution of the pre-existing cell membrane to eukaryotic cell evolution14 and the evolution of replicators 15.&2016 The Author(s)Published by the Royal Society.All rights reserved.on November 7,2016http:/rspb.royalsocietypublishing.org/Downloaded from However,thereisnogeneralquantitativeframeworktopre-dict evolutionary dynamics of cell structure.The quantitativecell structural traits are expected to have specific evolutionarypatterns compared with evolution of traits that are regulatedby genetic,or other non-genetic,inheritance mechanisms36,because the structure of a single-cell is a mixture of com-ponents with different inheritance patterns.Inheritance by thehangover of parental structure makes the mean trait value ofoffspring populations equal to that of the selected parent,and does not contribute to new trait variation.In comparison,structural inheritance brings about a difference between thepopulation mean of the selected parent and that of the off-spring,and new trait variations could be generated with itsincomplete heritability.Vegetative cell wall formation in the diatom,a unicellulareukaryote,is an example of structural inheritance,in thatthe shape of the parent cell physically affects the shape of itsoffspring 16,17.Each cell consists of two immutable siliceousvalves(cell walls at both cell poles),which can be differentiallylabelled 18:one is inherited from the parent and the other isnewly formed within offspring cells.The new offspringvalves are formed only within a silica deposition vesicle(SDV)that expands inside the solid parent valves 16,19,20.Therefore,parent valves function as the template for the exter-nal shape of offspring valves,and the diameter slightlydecreased by parent valve thickness 20.Indeed,the defor-mation of the valve outline shape and abrupt reduction incell size can be inherited for many generations 16,21,andthe degree of inheritance is thought to be influenced by flexi-bility of the girdle band(bands of silica that support valves)22.In addition,the microscale surface structure patterns onthe offspring valves correlate with the location of parentalorganelles 16,23 and reflect the environmental conditionsthat a parent cell has experienced 2428.Determining whether structural inheritance is affected byenvironmental changes is important for understanding thestability of inheritance 29.Here,we used the euryhalinediatom Cyclotella meneghiniana,which lives in freshwater toestuarine environments 30.The diatom exhibits phenotypicplasticity 31,32(i.e.a trait difference resulting from exposureto different environmental conditions)in response to changingsalinity in the laboratory 26,27.In our system,phenotypicplasticity was detected as the trait mean difference betweenthe parent valves in freshwater and offspring valves in sea-water.To examine the stability of inheritance followingsalinity changes,we investigated parentoffspring valve traitregression between parent valves developed in freshwaterand offspring in seawater.A change in salinity is a ubiquitouscue for the euryhaline diatom,and Cyclotella species arethought to undergo several speciation events according tofreshwater-to-marine migrations 33.The state of vegetative cell structure can influence fitnessand life-cycle transition.We estimated the heritability of fourmicroscale traits(figure 1a:central fultoportulae(CTFP),striae,central area and cell diameter)which could influencemetabolism 20,mechanical strength 34 and whether thecell switches from asexual to sexual reproduction 21.Thenon-genetic inheritance is expected to cause division of cellstates between cell lineages.For instance,sibling cells fromthe same mother cell tend to differentiate into the same sexualfate(egg or sperm)in the centric diatom C.meneghiniana 35.In this study,we estimated the heritable non-genetic com-ponent within cell trait variance(referred to as cytoplasmicheritability)from the regression analysis between the traitson parent and offspring valves using a common euryhalinediatom C.meneghiniana.Furthermore,to characterize theevolutionary dynamics of cell structure,we constructed asimple quantitative genetic model that considered the twoinheritance mechanisms of parent traits i.e.the hangover ofparent structure and structural inheritance.2.Material and methods(a)Culture conditions for the maintenanceof vegetative cell stateWeused asingleclone ofC.meneghinianaspeciescomplex 36thatconstructed from NIES-2364.To investigate cytoplasmic heritabil-ity within vegetative cells,clone construction and maintenancewere performed,considering the following diatom life cycle.The vegetative cell division continues over hundreds of gener-ations 22 with a gradual decrease in cell diameter.When thecellsreachadiameterbelowacertainthreshold,sexualcellABCD(a)(b)(c)offspring valvesparent valvesFWFW+probeoffspring valvesparent valvesFWSW+probeFigure 1.Morphological metrics and generation of the cell wall in the diatom.(a)Variablesevaluatedinthemorphologicalanalysesofvalves(cellwallpartsatboth cellpoles),illustratedfora C.meneghinianaindividual.Each cell consists oftwo siliceous valves.A,the number of central fultoportulae(CTFP);B,thenumber of striae;C,the central area diameter;D,the cell diameter.(b)Sche-matic diagram of offspring valve formation within the parent cell infreshwater medium(FW).A single parent cell was transferred to a well platefilledwith freshwater,afluorescentprobewasadded,and thelabelledoffspringvalves were formed during cell division.The black valves indicate the parentvalves and red valves are newly formed within the offspring cells.(c)Theparent valve formed in freshwater medium and the offspring valve formed inseawater medium(SW).rspb.royalsocietypublishing.orgProc.R.Soc.B 283:201616322 on November 7,2016http:/rspb.royalsocietypublishing.org/Downloaded from reproduction can be induced by species-specific environmentalstimuli.In C.meneghiniana,the stimulus is exposure to salinity37.After sexual reproduction,the vegetative cell with the largercell diameter is formed 21.Sexual reproduction will not be induced in vegetative cells iftheir diameter is above the size threshold,even if the environ-mental stimulus is added.Therefore,to use sufficiently largecells of stable vegetative state,the clone was constructed withthe following steps.(i)Weinducedsexualreproductionbyaddingsalinemedium to a clone of NIES-2364 with a cell diameterbelow the threshold.(ii)Single-cell isolation was performed for cells with a largerdiameter that were produced just after sexual reproduc-tion.In addition,during maintenance of the strain andthe experiment we checked with microscopy that sexualreproduction had not been initiated.To avoid mutation and cell diameter decrease,we madepreservation stocks of the same isolated clone,and the stockwas maintained at low temperature and light intensity to inhibitcell division.For the experiment,cells were cultivated untilreaching the mid-to late-log growth phase using sterile freshwater medium(modified Bolds Basal Medium 38).For adetailed description,see the Material and methods in theelectronic supplementary material.(b)Morphological observations of the parentaland newly formed valvesCyclotella meneghiniana exhibits phenotypic plasticity to changingsalinity 26,27.Therefore,to separate the effect of sharing thecommon external environment from the heritable factor,we inves-tigated whether the inheritancewasmaintained,evenifparent andoffspring valves were formed under different saline environments.Each single cell was isolated using an attenuated Pasteur pipetteand transferred to a well of a 48-well plate(Iwaki,Osaka,Japan).Each well was filled with freshwater or 50%seawater,anda fluorescent probe(LysoTrackerTMHCK-123 yellow;MolecularProbes,Life Technologies,CA,USA)was added to a finalconcentration of 400 nM in order to distinguish parent and off-spring valves.The fluorescent probe is taken up during valveformation 18.Thus,the two parent valves were not labelled,butthetwonewoffspringvalveswithinawellwerelabelled(electronicsupplementary material,figure S1).Fifty per cent seawater solution was prepared by the dissol-ution of half the amount of artificial seawater ingredients usedfor the preparation of a 100%seawater solution in freshwatermedium.The cell trait variables were analysed from phase-contrast images.For a detailed description,see the Materialand Methods in the electronic supplementary material.(c)Statistical analysisValues are reported as the mean and standard error(s.e.).All stat-istical testing was conducted using R software v.3.1.2(RFoundationforStatisticalComputing;http:/www.R-project.org).To estimate the heritability,linear regression analysis wasperformed,with the offspring trait as the dependent variableand the parent trait as the independent variable.A familyconsists of two parent and two offspring valves(figure 1b,c).For the parentoffspring valve trait regression,the mean(forcentral area and cell diameter)or sum(for striae and CTFPnumber)of each of the two valves were used as the traitvalues.For a detailed description of linear regression analysis,see the electronic supplementary material.In the quantitative genetic model for non-genetic inheritance,the phenotypic variance(VP)is decomposed into the sum of thevariance of the genetic component(VG),the variance of thenon-genetic inherited component(VCI:further referred to as thecytoplasmicinheritancecomponent),andthenon-inheritedenvironmental component(VE)2.In this case,VGis assumed tobezerobecauseweusedthediatomcellswithinaclonalpopulation.In the parentoffspring valve trait regression,we defined thecytoplasmic heritability as the regression coefficient 29 betweenthe traits on the parent and offspring valves,as follows:b h2VCIVP,2:1where b is the regression coefficient of the offspring on parenttrait values,and the cytoplasmic heritability is h2.For analysis of phenotypic plasticity,the trait mean differ-ence between the parent valves in a freshwater environmentand the offspring valves in a seawater environment was analysedwith the Students t-test.For the Students t-test,the normality ofthe data was evaluated using the ShapiroWilks test,while thehomogeneity of variance between groups was evaluated usingthe F-test.We also investigated whether the developmental noise ofsimultaneously formed sibling valves would result in valvesthat resemble each other more than other valves within the popu-lation.We then compared the trait difference of actual siblingvalve pairs and that of random sampling pairs from the shuffledvalve population,using the bootstrap method.The followingnull hypothesis was tested against a one-sided alternativehypothesis at the 5%significance level:the null hypothesis isthat the observed mean difference between actual siblingvalves follows the mean distribution of the randomly chosenpair difference between valves.In comparison,the one-sidedalternative hypothesis is that the trait difference between siblingcells tends to be smaller than that of randomly chosen pairs(seethe electronic supplementary material).(d)A model for the evolutionary dynamics of cell traitswith natural selectionWe developed a simple model to i
