Patterns and Mechanisms of Ancestral Histone Protein Inheritance in Budding Yeast

19 Pages·2011·4.59 MB·English

Patterns and Mechanisms of Ancestral Histone Protein
Inheritance in Budding Yeast
MartaRadman-Livaja1.,KittyF.Verzijlbergen2.,AssafWeiner3,4.,TiborvanWelsem2,NirFriedman3,4*,
Oliver J. Rando1*, Fred van Leeuwen2*
1DepartmentofBiochemistryandMolecularPharmacology,UniversityofMassachusettsMedicalSchool,Worcester,Massachusetts,UnitedStatesofAmerica,2Division
ofGeneRegulation,NetherlandsCancerInstitute,andNetherlandsProteomicsCenter,Amsterdam,TheNetherlands,3SchoolofComputerScienceandEngineering,The
HebrewUniversity,Jerusalem,Israel,4AlexanderSilbermanInstituteofLifeSciences,TheHebrewUniversity,Jerusalem,Israel
Abstract
Replicatingchromatininvolvesdisruptionofhistone-DNAcontactsandsubsequentreassemblyofmaternalhistonesonthe
newdaughtergenomes.Inbulk,maternalhistonesarerandomlysegregatedtothetwodaughters,butlittleisknownabout
thefinedetailsofthisprocess:domaternalhistonesre-assembleatpreferredlocationsorclosetotheiroriginalloci?Here,
weusearecentlydevelopedmethodforswappingepitopetagstomeasurethedispositionofancestralhistoneH3across
the yeast genome over sixgenerations. We findthat ancestral H3is preferentially retained atthe 59 endsof mostgenes,
withstrongestretentionatlong,poorlytranscribedgenes.Werecapitulatetheseobservationswithaquantitativemodelin
whichthemajorityofmaternalhistonesarereincorporatedwithin400 bpoftheirpre-replicationlocusduringreplication,
with replication-independent replacement and transcription-related retrograde nucleosome movement shaping the
resultingdistributionsofancestralhistones.WefindakeyroleforTopoisomeraseIinretrogradehistonemovementduring
transcription, and we find that loss of Chromatin Assembly Factor-1 affects replication-independent turnover. Together,
theseresultsshowthatspecificlociareenrichedforhistoneproteinsfirstsynthesizedseveralgenerationsbeforehand,and
that maternal histones re-associate close to their original locations on daughter genomes after replication. Our findings
furthersuggest thataccumulation of ancestral histonescould playa roleinshaping histonemodificationpatterns.
Citation:Radman-LivajaM,VerzijlbergenKF,WeinerA,vanWelsemT,FriedmanN,etal.(2011)PatternsandMechanismsofAncestralHistoneProteinInheritance
inBuddingYeast.PLoSBiol9(6):e1001075.doi:10.1371/journal.pbio.1001075
AcademicEditor:PeterB.Becker,AdolfButenandtInstitute,Germany
ReceivedSeptember23,2010;AcceptedApril22,2011;PublishedJune7,2011
Copyright:(cid:2)2011Radman-Livajaetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits
unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited.
Funding:OJRissupportedinpartbyaCareerAwardintheBiomedicalSciencesfromtheBurroughsWellcomeFund.Thisresearchwassupportedbygrantsto
OJRandNFfromtheNIGMS(GM079205)andfromtheUS-IsraelBinationalFoundation,andtoFvLfromtheNetherlandsOrganisationforScientificResearchand
theNetherlandsGenomicsInitiative.Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofthemanuscript.
CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist.
*E-mail:[email protected](NF);[email protected](OJR);[email protected](FVL)
.Theseauthorscontributedequallytothiswork.
Introduction histones carrying epigenetic information do not re-associate with
daughter genomes at the location from which they came, this
In addition to the information encoded in DNA sequence, could lead to ‘‘epimutation,’’ analogous to DNA bases moving
replicatingcellscaninheritepigeneticinformation,whichrefersto relative to one another during genomic replication. Second, it is
variable phenotypes that are heritable without an underlying unknown to what extent newly synthesized histones deposited at
changeinDNAsequence.Itiswidelyacceptedthatchromatin,the differentlocidifferintheircovalentmodificationpatterns.Finally,
nucleoproteinpackagingstateofeukaryoticgenomes,providesone how old histones influence new histones, the basis for positive
potential carrier of epigenetic information. Although definitive feedback, can be considered analogous to asking what the
proofthatchromatinpersecarriesepigeneticinformationduring equivalent of base-pairingisduring chromatin replication.
replicationexistsinveryfewcases[1],geneticstudiesinnumerous Classic radioactive pulse-chase studies demonstrated that, in
organisms have identified key roles for chromatin regulators in bulk,maternalhistonessegregateequallytothetwodaughtercells
multiple epigenetic inheritanceparadigms [2,3]. [4,8–10]. It is unknown, however, whether maternal histones
The idea that chromatin structure carries epigenetic informa- remain close to the locus from which they were evicted by the
tion poses a central mechanistic question—since chromosome replicationfork orwhether maternalhistonesare incorporatedat
replicationinvolvesdramaticperturbationstochromatinstructure preferredgenomiclociinthetwodaughtergenomes[5,7,11].The
ranging from old histone displacement to widespread incorpora- extent of maternal histone dispersal affects the stability of
tion of newly synthesized histones, how can chromatin states be epigenetic states in theoretical models of chromatin inheritance
stably maintained? To understand the mechanism by which [12],making experimentaldetermination ofthisparameter a key
chromatin states could be inherited, it is necessary to understand goal forepigenetics research.
the unique challenges posed by histone protein dynamics during To address these fundamental questions, we carried out a
replication[4–7].First,histonesmustatleasttransientlydissociate geneticpulse-chasewithepitope-taggedhistoneH3[13]tofollow
from the genome during passage of the replication fork—if old ancestralH3forseveralcelldivisionsafterremovaloftheancestral
PLoSBiology | www.plosbiology.org 1 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
generations after releasing yeast into the cell cycle [13]. This
Author Summary
material was hybridized to tiling microarrays covering 4% of the
It is widely believed that chromatin, the nucleoprotein yeastgenome[20],andHA/T7ratiosofnormalizedHAandT7
packaged state of eukaryotic genomes, can carry epige- signalswerecomputedforthe3and6generationdata(Figure1C,
netic information and thus transmit gene expression D).SinceHAiseliminatedviarecombinationleavingnewH3-T7,
patterns to replicating cells. However, the inheritance of highHA/T7ratiosindicatelocienrichedforancestralhistoneH3.
genomic packaging status is subject to mechanistic Surprisingly, many of the highest HA/T7 levels were associated
challenges that do not confront the inheritance of with codingregions (discussed below).
genomic DNA sequence. Most notably, histone proteins
Overall,HA/T7patternsareconsistentat3and6generations,
must at least transiently dissociate from the maternal
butthedynamicrangeofHA/T7enrichmentdiminishedfrom3
genomeduringreplication,anditisunknownwhetheror
to 6 generations (Figure 1D, Figure S2). This is an expected
notmaternalproteinsre-associatewithdaughtergenomes
consequence of thefact that ,1%–2% of cells donot recombine
near the sequence they originally occupied on the
theHAtagaway(FigureS1)—sincetheamountofancestralH3is
maternal genome. Here, we use a novel method for
decreasing by at least 2-fold in each generation, the relative
trackingoldproteinstodeterminewherehistoneproteins
contribution of the ,2% of cells still expressing H3-HA will
accumulateafter1,3,or6generationsofgrowthinyeast.
increase over time, with this genomic background eventually
Tooursurprise,ancestralhistonesaccumulatenearthe59
end of long, relatively inactive genes. Using a mathemat- competingwiththerealsignalfromincreasinglyrareancestralH3
ical model, we show that our results can be explained by (,2%of total H3after 6 doublings).
the combined effects of histone replacement, histone
movement along genes from 39 towards 59 ends, and Ancestral Histones Are Retained Over Long, Poorly
histonespreadingduringreplication.Ourresultsshowthat Transcribed Genes
old histones do move but stay relatively close to their
To extend our analyses to the entire genome, we carried out
original location (within around 400 base-pairs), which
deepsequencingofHAandT7libraries.HA-andT7-taggedH3
places important constraints on how chromatin could
were immunoprecipitated after the tag swap but before release
potentially carry epigeneticinformation. Ourfindings also
from arrest (0 generations), after release into a G2/M cell cycle
suggest that accumulation of the ancestral histones that
block, and at 1, 3, and 6 generations after release. Sequencing
are inherited caninfluencehistone modificationpatterns.
reads were mapped to the yeast genome, normalized for read
count, and HA/T7 ratios were computed genome-wide. These
tag. We find that old histone proteins do not accumulate at data correlated well with our microarray data, and we further
epigenetically regulated loci such as the subtelomeres but instead validatedthesemeasurementsbyq-PCRatSPA2andBUD3,two
accumulate at the 59 ends of long, poorly transcribed genes. As genes which both exhibit high and low HA/T7 ratios at their 59
expected, old histones do not accumulate at loci exhibiting rapid and39 ends,respectively (Figure S3).
histone turnover, but we also find that 39 to 59 movement of old Inpreviouswork,weandothers[13,15–17,21–23]showedthat
histones along coding regions and histone movement during there is a partial correlation between transcription levels and
replicationarerequiredtoexplainthepatternsofancestralhistone replication-independent histone dynamics. To understand how
retention we observe. We estimate that maternal histones stay transcription might affect multigenerational histone retention in
within ,400bp of their original location during replication, our system, we aligned all yeast genes by their transcription start
providing the first measure of this crucial parameter. Finally, we site (TSS) and clustered genes (K-means, K=5) based on the
identify a number of factors that affect old histone localization, pattern of the 3-generation HA/T7 ratios along the gene body
such as topoisomerase I and the H4 N-terminal tail, which both (FiguresS4,S5,TableS1).Weobservedastrikingenrichmentof
affect the 59 bias in localization patterns. In contrast, CAF-1 H3-HAjustdownstreamofthe59endsofgenes(typicallypeaking
mostly affects histone turnover at promoters. Together, these around the +3 nucleosome). One exception to the 59 pattern
results provide a detailed overview of the movement of ancestral describedisfoundinoneclusterofshortgeneswithuniformlylow
histonesacrossmultiplecellgenerationsandidentifyanumberof H3-HA levels (Figure S4, Cluster 1), which is enriched for GO
mechanismsthatplayaroleinshapingthelandscapeofancestral categories (such as protein translation) related to high gene
histone retention. expressionlevels.Incontrast,longgenesweregenerallyassociated
with higherlevels ofancestral H3 (see forexample Cluster 5).
Results Tobettervisualizethesetrends,wesortedgenesbytheextentof
ancestral H3 retention after 3 generations (Figure 2A–B).
Tofollowthemovementofoldhistoneproteinsovermultiplecell Retentionofancestralhistonescorrelatesbothwithlowexpression
generations,weutilizedanovelpulse-chasetechnique[13]tofollow levelsandwithlongergenes(Figure2C–D,FigureS6).Whileitis
ancestral epitope-tagged histone H3 for severalcell divisions after thecasethatlongergenestendtobeexpressedatlowerlevelsthan
swappingepitopetagsfromH3-HAtoH3-T7(Figure1A,B).We short genes (Figure 2E), these factors are partially independent
havepreviouslydescribeduseofthistechniquetoassayreplication- here—even when we focus on genes of 1–2kb length, we still
independent H3 turnover in arrested cells and have shown that observe the correlations between ancestral histone retention and
prior to recombination all cells carry the H3-HA, and that lowexpression(Figure2E–F,andseebelow).Interestingly,inboth
recombination is 98% efficient in cells that are not dividing due microarray and sequencing datasets we found that epigenetically
to nutrient deprivation (Figure S1). Unlike inducible pGAL-based repressed loci such as the silent mating loci and subtelomeres
systems for measuring replication-independent histone dynamics [24,25] did not preferentially accumulate ancestral histone
[14–17],heretheepitope-taggedhistoneisunderthecontrolofits proteins (Figure 1C, Figure S7)—analysis of both unique and
endogenous promoter, avoiding potential artifacts of H3/H4 repetitive subtelomeric genes showed similar H3-HA retention
misexpression[18]onhistonedynamicsthroughoutthecellcycle. patterns to euchromatic genes of similar length and expression.
We used MNase-ChIP [15,19] for the HA and T7 tags after This was not a consequence of silencing defects in our strains, as
recombination but before release into the cell cycle, and 3 and 6 they showedefficientmating (unpublisheddata).
PLoSBiology | www.plosbiology.org 2 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Figure1.Overviewofsystemfortrackingancestralhistoneproteins.(A)Recombination-basedswappingofepitopetagsonhistoneH3.
HistoneH3istaggedatitsendogenouslocuswithaC-terminalHAepitopetagsurroundedbyLoxPsites.UponinductionofCrerecombinasewithb-
estradiol,theHAtagisrecombinedoutandH3isleftwithaC-terminalT7tag.(B)Experimentaloverview.YeastcarryingHA-taggedH3arearrested
bynutrientdepletion,andtheHART7swapisinducedbyovernightincubationwithb-estradiol.Afterthetagswap,yeastarereleasedfromarrest
andHAandT7tagsaremappedacrossthegenomeatvaryingtimespost-release.(C)ChromosomeIIIoverview.HA/T7ratiosareshownasaheatmap
acrosschromosomeIIIat3generationsafterrelease.NotableinthisviewisalackofaccumulationofH3-HAatTEL3Lorthesilentmatingloci.(D)
Close-upviewsoftwogenomicloci.Dataareshownasaheatmapfor3and6generationsafterthetagswap.
doi:10.1371/journal.pbio.1001075.g001
Whatpropertiesofshortorhighlytranscribedgenesmightlead was better correlated with the average turnover rate of several
to loss of ancestral histones? Replication-independent histone surrounding nucleosomes than with the immediate turnover rate
replacementoccursmostrapidlyoverintergenicregionsandover (see, for example, Figure 3B–C). This observation suggests that
the coding regions of highly transcribed genes [15,17,21,23], the maternal histones preferentially re-associate with daughter ge-
converse of the pattern of ancestral H3 retention we observe. nomes near the location from which they originated—if old
Indeed, ancestral histone retention is broadly correlated with histonesscatteredrandomlyatreplication,ancestralH3retention
‘‘cold’’ regions of low H3/H4 turnover (Figure 3A). Importantly, patterns should more precisely anticorrelate with replication-
however, for a given level of H3/H4 turnover, ancestral H3 independent turnover patterns, as is discussed in more detail
retention varied significantly—retention at a given nucleosome below.
PLoSBiology | www.plosbiology.org 3 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
PLoSBiology | www.plosbiology.org 4 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Figure2.AncestralH3moleculesaccumulateatthe59endsoflong,poorlytranscribedgenes.(A–B)HeatmapofsitesofancestralH3
accumulation.GenesarealignedbyTSS(indicated),andLog HA/T7ratiosareindicatedasaheatmap.GenesareorderedbythemedianHA/T7ratio
2
overthe59-most1kbat3generations.Greyovercodingregionsindicatesmissingdata;greydownstreamofgenesindicatessequencedownstream
ofthe39endofthegenetoshowgenelength.Accumulationofancestralhistonesatthe59endsofgenespeaksaroundthe+3nucleosome,as
expectedgiventhatthe+1and+2nucleosomesaregenerallysubjecttohighratesofreplication-independentH3/H4replacement[15,17].(C)An80
geneslidingwindowaverageofPol2ChIPlevels[75]forgenesorderedasin(A–B),showingthatgeneswithlowlevelsofancestralH3retentionare
highlytranscribed.(D)80geneslidingwindowaverageofgenelengths,showingthatgeneswithhighlevelsofancestralH3retentiontendtobe
long.(E)ThemedianHA/T7ratiooverthe59endofgenes(1kb)wascalculatedforallgenes,andmedianvaluesofthisretentionmetricareshown
forgroupsofgenesorderedbytranscriptionrate(x-axis)andgenelength(y-axis).Whilethesearenotindependent—highlyexpressedgenestendto
beshort—foragivengenelengthgenestranscribedathigherlevelsexhibitlowHAretentionlevels.Thisistruemostlyofgenesshorterthan3kb,
whichencompassesthemajorityofyeastgenes.(F)AverageHA/T7ratios(Log2)forgenesbetween1and2kb,brokenintohigh(red),low(green),
andintermediate(blue)transcriptionrates.
doi:10.1371/journal.pbio.1001075.g002
Accumulation of Ancestral Histones at 59 Ends of Genes background of nonswitching cells starts to dominate the profile
Why do old histone proteins accumulate near the 59 ends of at 6 generations (Figure 4). As expected, HA/T7 data at
genes? We considered two alternative possibilities for classes of generation0exhibitedwidespreadHAloss/T7gainatpromoters
mechanisms causing this pattern. In the first mechanism, we and +1 nucleosomes as a result of the rapid replication-
reason that if histone proteins tend to maintain their locations independent turnover at these loci [13,15,17]. Importantly, to
along the genome, the 59 enrichment of old histones implies that rule out the possibility that 59 accumulation of H3-HA was an
old 39 histones are evicted and replaced by new histones during effect of our arrest-release protocol, we also measured HA/T7
some phase of the cell cycle. However, previous measures of distributions6 hafterinducingrecombinationinactivelygrowing
turnover in G1- or G2/M-arrested yeast [15,26] cannot explain midlog cultures of yeast (Figures S3 and S10). Despite heteroge-
the59/39ratiosweobserve.Furthermore,wefoundthatmutations neity in switching times in this protocol (only 65% of yeast have
incandidate59/39-markingcomplexessuchascohesin[27,28]or switched from HA to T7 3 h after switch induction, 85% after
H3K4/K36 methylases [29] did not affect 59-biased retention of 6 h), we nonetheless observed that HA/T7 distributions were
oldhistones at target loci (Figure S8A). remarkably similar in midlog-switching cells to HA/T7 patterns
A second possible explanation for widespread 59 accumulation observed in cells undergoing the arrest/release protocol, with
ofancestralhistoneproteinsisthatthehistoneproteinsmovefrom preferentialancestralhistoneaccumulationatthe59endsoflong,
39 to 59 over genes over time. This could result from RNA poorlytranscribed genes.
polymerasepassage,becausesomeRNApolymerasespasshistone Weaskedwhetherthesedynamicobservationscouldbeusedto
octamers in a retrograde direction during transcription [30,31]. quantitativelyruleoutspecificmodelsconcerningthemechanisms
Although it is debatable whether this is true of Pol2 in vitro for segregation of maternal histones to daughter genomes.
[32,33], in vivo we previously observed that inactivation of Pol2 However, the resolution of this question is complicated by
leads to a modest shift of nucleosomes from 59 to 39 [34], replication-independent processes we discuss above that can
consistent with the idea that Pol2 movement normally shifts remove or shift ancestral histones, and that cannot be fully
nucleosomesina59direction.Totestwhetherthismovementwas removed experimentally (for example, yeast will not proceed
related totranscription,weaskedwhetherthe59peak ofH3-HA through the cell cycle in the absence of RNA polymerase). To
accumulationshiftedfurther59withincreasingtranscriptionrate. understand the relationship between these issues, we designed an
We normalized all gene lengths to one, then plotted the HA/T7 analytical model that accounts for three processes that affect H3
ratio for all genes sorted by transcription rate (Figure S9). moleculesincodingsequences(Figure5A)andthenexaminedthe
Consistent with the prediction of transcription-dependent retro- effectof removinganyofthethree.Briefly, our model includesa
grademovement,wedidobserveasubtlesignalofH3-HApeaks nucleosome-specific term for H3 turnover taken from prior
shiftingfurther59athighertranscriptionrates.Whilethisanalysis experimental results [15], with H3 turnover resulting in loss of
could be confounded by the higher transcription rates seen over HA.Inaddition,itincludesagene-specificparameteraccounting
shorter genes, even when we focus on 1–2kb genes, we observe forlateralmovementofhistones(‘‘passback’’).Further,themodel
that poorly transcribed genes exhibit a much flatter profile than also includes a global parameter that describes the extent of
genes expressed at average levels (Figure 2F), as expected if Pol2 histone ‘‘spreading’’ via dissociation/re-association during repli-
transit were required for H3/H4 ‘‘passback.’’ Finally, we show cation. Finally, the experimentally measured background of 2%
below that per-gene estimates of passback exhibit significant nonswitching cells (Figure S1) was included. The free parameters
correlation with Pol2 levels. Together, these results are most ofthemodel(describingglobalhistonespreadingandgene-specific
consistentwithamodelinwhichhistoneproteinsmovefrom39to lateralmovementpergeneration)wereestimatedtomaximizethe
59 over coding regions over time (further detailed in the likelihood ofexperimental observations (Text S1).
Discussion). To account for any first-pass effects of Pol2 behavior during
initial re-feeding of nutrient-depleted yeast (Figures S3 and S10),
QuantitativeEstimationofNucleosomeDynamicsDuring we examined this model with two starting conditions—the first
Replication startedwithauniformgenomicdistributionofH3-HA,whilethe
A key question we sought to address in this study is whether second started with the experimental distribution of HA/T7
maternal histones re-associate near their original positions after observed after release into G2/M arrest (Figure 4). Both model
passageofthereplicationfork.WereasonedthatchangesinHA/ variants predicted HA/T7 ratios with good correlations to the
T7 patterns over the course of several generations might provide experimentaldata(Figure5Bshowsdatastarting fromauniform
insight into the effects of replication on nucleosome dynamics. distribution, Figure 5E and Figure S11 start from the G2/M
HA/T7 patterns change dramatically between arrest and 1 distribution). Examination of estimated parameters revealed
generation of release (with or without G2/M arrest) and then expectedbehaviors.Forinstance,thedistributionoflateralhistone
are very similar between 1 and 3 generations, before the movement estimates (Figure 5C) was strongly biased towards
PLoSBiology | www.plosbiology.org 5 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Figure3.H3retentionanticorrelateswithreplication-independentturnoverinagenelength-dependentmanner.(A)Scatterplotof
ancestralH3retention(medianLog2HA/T7forthe591kb,y-axis)versusreplication-independentturnover(Dionetal.[15],Zscore,x-axis).(B)HA
retentionisplottedagainst59H3turnoverasabovebutwithshortandlonggenesplottedseparately.ForagivenlevelofH3turnover,ancestral
retentionisgreateratlongergenes.(C)Averagesofthe59HA/T7retentionparameter(medianHA/T7forthe59-most1kb)areshownforgenes
brokenintodifferentlengthand59turnovergroups.Forallturnoverlevels,longergenesretainmoreH3-HAthandoshortergenes.
doi:10.1371/journal.pbio.1001075.g003
PLoSBiology | www.plosbiology.org 6 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Figure4.KineticanalysisofancestralH3retention.HA/T7ratiosweremeasuredgenome-wideafterrecombinationbutbeforerelease(Gen0),
afterreleaseintonocodazole(G2/M),andafter1,3,or6generationsofgrowthpost-release.Dataforallgeneswereaveragedandareplottedas
indicated.
doi:10.1371/journal.pbio.1001075.g004
retrograde 39 to 59 movement of histones, consistent with the nucleosome ina gene) thanwas observed (FiguresS11,S13). We
previously measured effects of rpb1-1 inactivation on nucleosome ascribe these failures to the fact that we considered each gene in
positioning [34]. Passback values were also significantly isolation and therefore did not model shifts of old nucleosomes
(p=9.6439e-19)correlated withtranscription rate (Figure S12). from adjacent genes, which would result in poor fits over short
Ourmodelallowsustoestimatetheextentofhistonemovement genes in particular. Interestingly, the better fit at the +1
during replication. Figure 5D shows the likelihood of the full nucleosome than at the +N nucleosome is consistent with rapid
modelplottedforvariousvaluesforreplication-dependenthistone promoter turnover more effectively isolating genes from one
spreading.Thebestfitmodelallowedhistonestospread,400bp another at their59 ends invivo.
in either direction, or roughly two nucleosome widths, during Overall, the strong correlation between our model and the
replication (more precisely, in this model two-thirds of histones experimental data supports the hypothesis that at least three
stay within 400bp of their original locations, as this value is the dynamicprocessesaffectnucleosomesandshapethelandscapeof
standard deviation of a Gaussian function describing spreading; ancestral histone retention and provide the first quantitative
see Text S1). Results from models with 800 and 1,600bp estimate ofmaternal histonedynamicsduring replication.
spreadingparametersareshowninFigure6forcomparison.Our
estimate of 6400bp spreading is particularly interesting given TopoisomeraseIandtheH4N-TerminalTailPlayRolesin
electron microscopy results demonstrating that nucleosomes are Establishingthe59/39GradientofAncestralH3Molecules
destabilized over 650–1,100bp around the replication fork on To further investigate the mechanism of 59 accumulation, we
replicating SV40 minichromosomes[35,36]. asked whether gene-specific passback parameters were correlated
Elimination of any one or two of the three components of the withspecificgeneannotations(TableS2)[34,37].Interestingly,we
model—spreading,turnover,orpassback—resultedinsignificantly find that the estimated passback distance was much greater at
worse fits between model predictions and experimental data TFIID-dominated (‘‘growth’’) genes than at SAGA-dominated
(Figure5E).Thiscanbeintuitedasfollows.First,intheabsenceof (‘‘stress’’)genes(Figure7A)[38].Asaresult,59accumulationwas
histone spreading, unmitigated histone movement from 39 to 59 much more pronounced at TFIID-dominated than at SAGA-
results in a much tighter 59 ancestral histone peak and results in dominated genes (Figure 7B). Almost every described aspect of
muchmoreextensivechangefromonegenerationtothenextthan chromatin structure and gene expression, from nucleosome
we observe. Second, eliminating histone turnover shifts the 59 positioningtoevolutionarylability(reviewedin[39,40,41]),differs
ancestralpeakclosertothe+1/+2nucleosome.Third,preventing between these two broad types of genes. Mechanistically, one
lateral histone movement results in a 39-shifted, flatter ancestral interestingcorrelateisthatTFIIDrecruitmenthasbeenproposed
histone profile. to be mediated in part by acetylation of the N-terminal tail of
WhileourmodelprovidedgoodquantitativefitsofancestralH3 histoneH4 [38,42].
patterns for many genes, we nonetheless note that many genes To investigate this link experimentally, we examined whether
werenotperfectlyfitbythismodel.Generally,wefoundthatthe mutationsoftheH4tailinfluencedancestralhistoneH3retention.
model poorly fit short genes, and overall the model almost In an H4K5,12R mutant that cannot be acetylated on these two
universallypredictedlowerHA/T7atthe+Nnucleosome(thelast tailresidues,the59-biasedHA/T7waspartiallylost(FigureS8B),
PLoSBiology | www.plosbiology.org 7 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Figure 5. Quantitative modeling reveals three distinct dynamic processes. (A) Outline of quantitative model. From a given starting
distribution, histones are subject to turnover [15], transcription-associated lateral movement (‘‘passback’’), and replication-mediated spreading.
ModelisdescribedindetailinTextS1.(B)Themodelcapturesmajorfeaturesoftheexperimentaldata.HA/T7ratiosforexperimentaldataandmodel
predictionsareshownforallgenesasaheatmap.(C)Distributionoflateralpassbackparameter(pergeneration)forallgenes.Notethatthevast
majority(92%)ofgeneswereassociatedwithretrograde39to59movementalongcodingregions.(D)Estimationofreplication-basedspreadingof
maternalhistones.Modellikelihood(TextS1)isplottedonthey-axisforvariouswidthspreadingdistributions(definedas1standarddeviationofthe
Gaussiandescribinghistonemovementatreplication—seeTextS1formodeldetails).(E)Eliminatinganyofthethreemodelfeaturesworsensfitto
data.Plottedareaveragesat3generationsforgenesover2kbfordataversuspredictionsofvariousmodels(‘‘STP’’refertoreplication-mediated
PLoSBiology | www.plosbiology.org 8 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Spreading,replication-independentTurnover,andPassback).Notethatthemodeleliminatingturnoverunderestimatesturnovereffects,ashistones
thatspreadorarepassedoverthe59endofthegenearestilleliminatedinthismodel(i.e.,inthismodelweeffectivelyonlyeliminateturnoverwithin
CDS,notinintergenicregions),providinganotherbasisforhighlossof59histones.
doi:10.1371/journal.pbio.1001075.g005
Figure 6. Dependency of histone dynamics model on spreading parameter. (A–B) Parameters in the quantitative model described in
Figure5werere-optimizedafterfixingthespreadingtermto400bp(asinFigure5),800,or1,600bp.Dataandsimulationsareshownaveragedfor
genesover2kbformodelsstartingwithauniformH3-HAdistribution(A)orstartingwiththeexperimentallymeasuredG2/MHA/T7distribution(B).
(C–D) Examples of data and three models with different spreading parameters. Genomic coordinates are chromosome 2 490–540kb (C), and
chromosome160–110kb(D).Y-axisshowsmeasured(Data)orpredictedHA/T7values,inLog2.
doi:10.1371/journal.pbio.1001075.g006
PLoSBiology | www.plosbiology.org 9 June2011 | Volume 9 | Issue 6 | e1001075
TransgenerationalHistoneRetentioninYeast
Figure7.Mutantsaffectingancestralhistoneretention.(A)Distributionoflateralnucleosomedistancesfrommodel(Figure5).Shownarethe
passbackparametersforSAGA-dominatedandTFIID-dominatedgenesasdefinedinHuisingaetal.[38].(B)TFIID-dominatedgenespreferentially
accumulate59H3-HA.Averagesof3generationexperimentaldataareshownfortheindicatedgeneclasses.(C)H4taildeletiondramaticallyreshapes
thelandscapeofancestralhistoneretention.YeastcarryinganN-terminalH4taildeletionwereprocessedasinFigure1A–B,andaveragesforall
genesareplottedasindicated.Wenotethatthisstrainhasretainedawild-typeHHT2-HHF2locusforviability,soresultsmustbeinterpretedwith
PLoSBiology | www.plosbiology.org 10 June2011 | Volume 9 | Issue 6 | e1001075