Nuclear DNA amounts in angiosperms

124 Pages·2011·1.15 MB·English

Annalsof BotanyPage1 of124
doi:10.1093/aob/mcq258,availableonlineat www.aob.oxfordjournals.org
Nuclear DNA amounts in angiosperms: targets, trends and tomorrow
M. D.Bennett* and I. J.Leitch
Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AB, UK
*For correspondence. E-mail: [email protected]
Received:25August2010 Returnedforrevision:18October2010 Accepted:24November2010
CONTENTS
INTRODUCTION 2
Extendingthe range of genome sizes encounteredin angiosperms 3
The needforreference lists 4
TARGETSIN GENOMESIZE RESEARCH 4
Meetingtargetsforspeciesrepresentation 4 D
o
Progresstowardstargetsforfamilialrepresentation 5 w
n
Improved systematicrepresentation forgenera 6 lo
a
d
Improved representation ofother groups 6 e
d
TRENDSIN TECHNIQUESUSED TO ESTIMATE GENOMESIZE 7 fro
The risein flowcytometryasthe techniqueof choiceforgenome size estimations 7 m
Developmentofdifferentisolation buffersforflow cytometry 7 http
TRheceeanptpdleicvaetlioopnmoefntflsowincthyteomapeptrliycatotiopnlaonftflsyoswtecmyatotimcsetryto genome size studies 88 ://ao
b
(i)The useof seeds 8 .o
x
(ii)Ease of accessto methodologicaldata 8 fo
(iii)Newequipment 8 rdjo
Are there anynew techniques forestimating genome size onthe horizon? 9 urn
(i)Canreal timePCR beusedforestimating plantgenome sizes? 9 als
(ii)Will ‘complete’ genome sequencinggive useable genome size estimates? 9 .o
rg
TOMORROW 13 a/
Uncoveringandcollatinggenome size datafromdiversepublishedsources 14 t N
Screeningex situ andinsitu collectionsas sources of target taxa 15 IH
DEDICATION 15 L
ib
LITERATURECITED 16 ra
APPENDIX 19 ry o
Notesto the Appendix 19 n D
Original references forDNAvalues 121 e
c
e
m
b
†BackgroundandAimsTheamountofDNAinanunreplicatedgameticchromosomecomplementisknownasthe e
C-valueandisakeybiodiversitycharacteroffundamentalsignificancewithmanypracticalandpredictiveuses. r 30
Since1976,BennettandcolleagueshaveassembledeightcompilationsofangiospermC-valuesforreferencepur- , 2
0
posesandsubsequentlythesehavebeenpooledintotheAngiospermDNAC-valuesDatabase(http://data.kew.org/ 15
cvalues/).Sincethelastcompilationwaspublishedin2005,alargeamountofdataonangiospermgenomesizehas
beenpublished.ItisthereforetimelytobringthesedatatogetherintoaninthcompilationofDNAamounts.
†ScopeThepresentworklistsDNAC-valuesfor2221speciesfrom151originalsources(includingfirstvalues
for 1860 species not listed in previous compilations). Combining these data with those published previously
showsthat C-values are nowavailable for6287 angiospermspecies.
†KeyFindingsAnalysisofthedataset,whichisbyfarthelargestoftheninecompilationspublishedsince1976,
showsthatangiospermC-valuesarenowbeinggeneratedatthehighestratesincethefirstgenomesizeswereesti-
matedinthe1950s.Thecompilationincludesnewrecordholdersforthesmallest(1C¼0.0648pginGenliseamar-
garetae) and largest (1C¼152.23 pg in Paris japonica) genome sizes so far reported, extending the range
encountered in angiosperms to nearly 2400-fold. A review of progress in meeting targets set at the Plant
GenomeSizemeetingsshowsthatalthoughrepresentationforgenera,geographicalregionsandsomeplantlife
forms (e.g. island floras and parasitic plants) has improved, progress to increase familial representation is still
slow.Intermsoftechniqueitisnowclearthatflowcytometryissoonlikelytobecometheonlymethodavailable
forplantgenomesizeestimations.Fortunately,thishasbeenaccompaniedbynumerouscarefulstudiestoimprove
thequalityofdatageneratedusingthistechnique(e.g.designofnewbuffers,increasedawarenessandunderstanding
ofproblemscausedbycytosolicinhibitors).ItisalsoclearthatalthoughthespeedofDNAsequencingcontinuesto
rise dramatically with the advent of next-generation and third-generation sequencing technologies, ‘complete
genomesequencing’projectsarestillunabletogenerateaccurateplantgenomesizeestimates.
Keywords: DNAC-value,nucleargenome size, PlantDNAC-values Database, flow cytometry.
#The Author2011.Publishedby OxfordUniversity Presson behalfofthe AnnalsofBotanyCompany.Allrights reserved.
For Permissions,please email:[email protected]
Page 2 of 124 Bennett & Leitch – Nuclear DNA amounts in angiosperms
INTRODUCTION amount in the unreplicated gametic nuclear chromosome
complement [defined by Swift (1950) as the 1C-value) of
It has been possible to estimate the amount of DNA in plant
different taxa. [Recently this was defined as the holoploid
and animal nuclei using various methods for over 60 years.
genome size by Greilhuber et al. (2005).]
Following the discovery of the key role of DNA in biology
Easy access to data for these key characters is clearly an
in1953(WatsonandCrick,1953)suchresearchhasincreased
important facility. Yet in practice it has often been difficult
ineachsuccessivedecade,andthistrendhasfurtherintensified
to know whether a C-value exists for a particular taxon, and
in the 21st century. These decades saw the loss of many who
if so, where to find it, especially as estimates are widely scat-
played key roles in founding and developing this interest
teredintheliteratureorevenunpublished.Althoughsmalllists
[including Hewson Swift (1920–2004), John McLeish
of nuclear DNA amounts were published in reviews and
(1929–1971), Arnold Sparrow (1914–1976) and Hugh Rees
research papers as genome size data started to accumulate in
(1923–2009)]. Yet until now the remarkable birth and
the 1960s and 1970s, the first large compilation of DNA
explosion of genome size data in ‘the DNA age’ occurred in
amounts for angiosperms, compiled primarily as a reference
the scientific lifetime of some individuals who witnessed or
source, was published in 1976 with data for 753 species
experienced the entire development of this field. However,
from 54 sources (Bennett and Smith, 1976). This noted an
the death of Jim Price (1944–2005 – Arnold Sparrow’s post-
intention to publish supplementary compilations for reference
doctoralstudent)in2004,andtheretirementofthefirstauthor
purposes at intervals and seven followed (see Table 1). D
(Mike Bennett – Hugh Rees’s postdoctoral student) in 2006 o
Increasing accessibility and ease of access to nuclear DNA w
ahnads bJoehgaunnnaGsrtehielhubbaetorninp2as0s1e0ssthoowgesntehraattioannshiusntobroircnchatantghee amounts was made possible through pooling of data in the nload
sstiazretroesfeathrcehDreNflAectasgwe,riwttehnosaecckonuonwtsleadlgoeneo,fweitahrlynogfeunrothmeer pavuabilliasbhleedorenfetrheenceintceormnepti.laTtihoensAinntgoioosnpeeremlecDtroNnAicCda-vtaabluaeses ed from
recourse to supplementary first-hand sources. Happily, recent Dfirasttabfiavseep(urbelliesahseed1c)o,mwphiliacthioncsol(lTataebdlea1ll),twheendtaltiavefrionmApthriel http
trends noted in this review show that tomorrow’s studies of 1997. After several updates, a major new release (3.1) of ://a
nuclearDNAamountsinangiospermsrestwithanenthusiastic o
new cohort of practitioners focused on carefully improving angiosperm data was combined with databases for gymnos- b.ox
techniques and data representation to better understand the pCe-rvmalsu,esptDeraidtaobpahsyete(srelaenadseb1ry.0o)phinyte2s0,01in.toThtihsedaPtlaabnatseDwNaAs ford
origin, extent and effects of variation in plant genome size. jo
As noted by Bennett and Leitch (2005a), ‘Nuclear DNA further expanded in 2004 (release 3.0) to incorporate not urn
acmhaoraucntterasn,dwgheonsoemsteusdiyzepr(oCv-ivdaelsuea)satrroenigmupnoirftyainntgbeioledmiveenrstiitny oAnnlgyiofisrpsetrvmaluDeNsAforC-6v2a8lusepsecDieastabnaostep,rbeuvtioaulssloy,lfiostredtheinfitrhset als.org
time, DNA amounts for 253 algal species. An additional a/
biology with practical and predictive uses’. Of all the major update of the Plant DNA C-values Database (release 4.0), t N
taxonomic groups studied, Bennett and Leitch (2005a) noted IH
that angiosperms were probably the most intensively studied. adding data for over 300 angiosperm species (from L
Zonneveld et al., 2005), went live in 2005. Release 5.0 of ib
With published prime C-values for over 6280 species now ra
thePlantDNAC-values Database(December2010)incorpor- ry
available (Table 1), they remain so today. Indeed, work on o
atesallthedatapresentedintheAppendixhere,andbringsthe n
plants has played a leading part in the research on the DNA D
e
c
e
m
b
e
TABLE 1. Total numbers of first estimates for angiosperm species and of original references from which they come in nine r 30
compilationsofnucleargenomesizeestimatescollatedforreferencepurposessince1976 , 20
1
5
No..offirstestimatesfor No.ofprimeestimatesforspecies No.oforiginal No.offirstestimatesper
‘species’inabstractof inthePlantDNAC-valuesDatabase referencesin originalreference(column
Originalcompilation compilation (release5.0) compilation 2/4)
1.BennettandSmith(1976) 753 596† 54 13.9
2.Bennettetal.(1982) 240 195† 53 4.5
3.BennettandSmith(1991) 588 552† 56 10.5
4.BennettandLeitch(1995) 899 868† 106 8.5
5.BennettandLeitch(1997) 471 481† 37 12.7
6.Bennettetal.(2000) 691 686† 71 9.7
7.BennettandLeitch(2005a) 628 636† 88 7.7
8.Zonneveldetal.(2005) 308 308 1 308.0
9.Presentwork 1860 1974† 151 13.1
Total 6438* 6287* 617 –
*ThetotalnumberoftaxainthePlantDNAC-valuesDatabase(i.e.6287)islowerby151thanthetotalforpapercompilations(i.e.total¼6438)because
severaltypesofentryincludedinthelatterareomittedfromtheformer.Theseinclude:87DNApercellvalues,35genussp.valueswhereothervaluesfor
namedspeciesinthesamegenusareincluded,and29valuesforsomevarietiesofspecieswherevaluesforothervarietiesofthesamespeciesareincluded.
†ThenumberofprimeestimatesforspeciesinthePlantDNAC-valuesDatabase(release5.0)(BennettandLeitch,2010)maydifferfromthenumberof
firstestimatesforspecieslistedintheabstractofpapercompilations(column2)owingtothereallocationofprimestatusfromthefirstestimatepublishedfor
ataxontoanestimatepublishedlater.
Bennett & Leitch – Nuclear DNA amounts in angiosperms Page 3 of 124
totalnumberofangiospermswithC-valuedataavailableinthe relationship between genome size and pollen size (Knight
database to 6287 (http://data.kew.org/cvalues/, Bennett and et al., 2010).
Leitch, 2010). The database is in a format which can be Given their ongoing uses as reference sources, the publi-
queried through a variety of search options, thus enabling cation of a further supplementary compilation of angiosperm
users to analyse, sort and filter the data, while at the end of C-values is needed, and indeed, it is overdue. Whereas the
each query, the number of records returned together with first three compilations were published in journals of the
their summary statistics (i.e. minimum, maximum, mean and Royal Society (London), the five later ones, from 1995
standard deviation) are given. In addition taxonomic (i.e. onwards, have all been published in Annals of Botany, a
family, higher group, voucher status), cytological (i.e. ploidy decision based, in part, on its wider availability to intended
level, chromosome number), technical (i.e. method used to users, especially those in institutions in the USA. Annals of
estimate genome size) and bibliographic information can be Botany (AoB) has played a major role in plant genome size
displayed where such data are available. research through its sponsorship of discussion meetings and
Compiling the reference lists and, more recently, the elec- workshops on plant genomes [at the Royal Botanic Gardens,
tronic databases has clearly been of value, given the scope Kew (RBG, Kew) in 1997, 2003 and 2007, and at the XVII
and range of uses made of them by the science community. International Botanical Congress held in Vienna, Austria, in
An indication of their value can be gleaned from noting that 2005]. These led to the publication of three special issues
the published compendia listing angiosperm C-values have ‘Genome size in Plants’ (December 1998), ‘Plant Genome D
o
been cited over 2100 times and over 85 citations for just the Size’ (January 2005) and ‘Plant Genome Horizons – Vistas w
n
seventh compilation (Bennett and Leitch, 2005a). In addition, and Visions’ (April 2008). Partly as a consequence of these lo
a
theelectronicdatabaseshavebeencitedover230timesandthe developments, AoB has played a growing role as a vehicle de
d
total number of hits on the Plant DNA C-values Database is for research papers on plant genome size and has rapidly fro
now well over 250000 (most of which are enquiries for becomethejournalofchoiceforfirstpublicationofsignificant m
h
angiosperms). work. Indeed, Bennett and Leitch (2005a) noted that the ttp
Particular uses of DNA C-values for individual taxa have highest proportion of papers containing C-values were pub- ://a
o
been reviewed previously (Bennett et al., 2000; Bennett and lished in AoB – a role maintained until now. Thus, 17% b
.o
Leitch, 2005a; Leitch and Bennett, 2007). Together, these (15 of the 88) of the original sources cited in Bennett and x
fo
resources have clearly continued to provide large samples of Leitch (2005a) were originally published in AoB, which is rd
data essential for many diverse comparative studies. Recent similar to 16% (24 of 151) of the original sources cited in jou
examples of large surveys include studies of the relationship the present compendium. It is also worth noting that in both rna
between genome size and B-chromosomes (Levin et al., cases the proportion of first estimates for species (26.7 and ls.o
2005), duration of cell cycle (Francis et al., 2008), seed size 34.7%, respectively) published in AoB was much higher arg/
and mass (Beaulieu et al., 2007b), photosynthetic rate than the proportion of original sources. t N
(Beaulieu et al., 2007a), plant growth form and distribution Over theyears, the citation impact factor for AoB has risen IH
(Ohri, 2005), leaf cell size and stomatal density (Beaulieu from 1.127 in 1995 to 3.501 in 2010 and plant genome size Lib
et al., 2008; Hodgson et al., 2010), and patterns of genome papers have contributed strongly to this increase. For rary
size evolution (Leitch et al., 2005, 2009, 2010; Beaulieu example, compilations published in 1995–2000 together o
n
et al., 2010). havereceivedover600citations,whiletwopapers,oneonter- D
e
Variation in genome size has interesting relationships with minology (Greilhuber, 2005) and the other on technique ce
m
characters at many levels ranging from the nucleus and cell (Dolezˇel and Bartos, 2005), each with over 100 citations, b
e
to tissues and whole organisms (Bennett, 1987; Bennett and were bothamong the five most cited papers in AoB published r 3
0
Leitch, 2005c), and recent work continues to confirm this. in 2005. Clearly AoB is an important platform to showcase , 2
0
For example, relationships between genome size and a plant genome size research and reference data, and prac- 15
range of nuclear characters have been detected in surveys titioners would be wise to maintain a critical mass of such
using the Plant DNA C-values Database since 2003. work to benefit their field into the future.
Houben et al. (2003) showed that the extent of methylation
of histone H3 in euchromatin of plant chromosomes
Extendingtherangeofgenomesizesencounteredinangiosperms
depends on basic nuclear DNA content. Prokopowich et al.
(2003) found a positive correlation between rDNA copy The minimum and maximum DNAC-values listed in 1972
number and genome size in eukaryotes (Gregory, 2005), forArabidopsisthalianaandFritillariaassyriaca,respectively
whilst Ross-Ibarra (2007) found that genome size is signifi- (Bennett, 1972), remained unchanged for over 30 years,
cantly correlated with recombination rate, with changes in although the number of species listed increased more than
genome size explaining a meaningful proportion (approx. 15-fold from 273 (Bennett, 1972) to more than 4100 (Plant
20%) of variation in this character. In contrast, Barow and DNA C-values Database, release 4.0 October 2005). Lower
Meister (2003) reported that genome size had only a values were claimed for several taxa (including A. thaliana),
minor, albeit significant, effect on endopolyploidization in but most of these were discounted or withdrawn later,
seed plants (Barow and Jovtchev, 2007), with other factors mainly on technical grounds (Bennett and Leitch, 2005a).
such as phylogeny and life strategy having a stronger For example, the very low value for Cardamine amara of
impact. Similarly, a large-scale analysis of a phylogenetically 1C¼0.051 pg reported by S. R. Band in 1984 as a personal
diverse sample of species failed to find a strong significant communication to the first author has subsequently been
shown to be an underestimate. Three independent
Page 4 of 124 Bennett & Leitch – Nuclear DNA amounts in angiosperms
measurementsusingflowcytometryobtainedavalueof1C¼ 24.3%. [Actually, Dolezˇel et al. (1998) reported a mean
0.24 pg (Bennett and Leitch, 2005b). 2C-valueof1.22and1.26pgforRaphanussativusdepending
Nevertheless, data listed in the present work have recently on which species was used as the primary standard.] Had
extended the range of values for angiosperms at both ends of Schmuthsetal.(2004)usedthevalueof1.11pgtheirestimate
the scale. At the lower end, Greilhuber et al. (2006) reported for A. thaliana ‘Columbia’ would have been 0.166 pg, and
convincing C-values much lower than for A. thaliana for within 2% of the prime C-value estimate for A. thaliana
several Lentibulariaceae species, including 1C¼0.09 pg for ‘Columbia’ of 0.163 pg (Bennett et al., 2003). The electronic
Utricularia gibba, 0.065 pg for Genlisea aurea and a new databaselackssuchdetailedcommentsbutitcitesthecompen-
record minimum angiosperm C-value estimate of 0.0648 pg diumreviewtowhichreferencecanbemadetocheckifuseful
for Genlisea margaretae. This drop in the minimum known extra information is available and analysis provided.
C-value was not small, but constituted a remarkable 60% Theexampleaboveconcernsonetaxon.However,compen-
reduction to well under half the value for A. thaliana (0.16 dia reviews also provide opportunities to publish results of
pg). Indeed, Bennett and Leitch (2005a) previously discussed usefulwidercomparisonstoidentifytrends,ortestperceptions
what was the lowest C-value fora free-living angiosperm and (and correct misconceptions). For example, the compendium
suggested that the theoretical minimum monoploid genome papers give overviews of trends in the latest sample of esti-
size (1Cx value) may approach 0.05 pg for a true diploid. mates and papers, and report on progress towards reaching
The results for Genlisea (Greilhuber et al., 2006) confirm milestones andtargets in gapfilling and inthe take upof gui- D
o
this prediction, and should provide a practical rationale and dance regarding best practice issued by the GEnome Size w
n
spur for new surveys to see if other taxa have yet lower Initiative (GESI – Bennett and Leitch, 2005b) by the plant lo
a
C-values and approach the suggested theoretical minimum genome size community (http://data.kew.org/cvalues/pgsm/ de
d
even more closely. #9%20Key%20recommendations). In the past, none of these fro
At the other end of the scale, the maximum angiosperm was routinely included with updates or new releases of the m
h
C-value has also increased recently. Zonneveld (2010) electronic database, but instead may be mentioned there, ttp
reported a 1C-value of 132.5 pg for a hexaploid hybrid citing a compendium as the reference for full details. Unlike ://a
Trillium×hagae (2n¼6x¼30; Melanthiaceae), just less the Plant DNA C-values Database, compendia also often ob
.o
than the reported maximum genome size for an animal (133 draw attention to some of the novel and key uses to which x
fo
pg in the marbled lungfish Protopterus aethiopicus, DNAC-value metadata have been put, and review the results rd
Pedersen, 1971). Yet even this has been exceeded in the and conclusions of such studies. Having the collected new jou
genus Paris (also Melanthiaceae) in which a 1C-value of data together with such analyses and reviews is a valuable rna
152.23 pg for the octoploid Paris japonica (2n¼8x¼40) resource providing overviews inthe broadestpossiblecontext. ls.o
rg
was recently estimated by Pellicer et al. (2010). This value However, given the retirement of the first author and the a/
is 15% larger than all previous plant genome size estimates need to update the Plant DNA C-values Database on a more t N
and extends the range encountered in angiosperms to nearly regular basis than in the past, future releases of new data IH
L
2400-fold. will be uploaded with or without publication of any compen- ib
Suchstudiesillustrateagainthatwedonotyetknowthefull dium paper such as this. Thus, the database will continue to rary
ranges of basic genome parameters (cf. Bennett, 1998) and be updated and made available on the internet through the o
n
confirm the continued need for new surveys of plant DNA website of the RBG, Kew, who are committed to maintaining D
e
C-values, chromosome numbers and ploidy levelsto establish and managing this important biodiversity resource for the ce
m
their full ranges experimentally. genome size community. b
e
r 3
0
, 2
The need for reference lists TARGETS IN GENOME SIZE RESEARCH 0
1
5
Asthedatagatheredinthiscompendiumhavebeen entered The present work marks several major milestones in plant
into the Plant DNA C-values Database, it might be asked if genome size research which concern both quantitative and
publishing them in this form has added value to their elec- qualitativeaspectsoftheDNAC-valuedata.Thecompendium
tronic listing. Indeed it does, and at various levels ranging lists DNA C-value estimates for 2221 angiosperm species
from individual values to overviews and reviews of the meta- taken from 151 original references (including first values for
data.Forexample,givenonlythenumberspublisheditispuz- 1860 species from 132 sources). Thus, it is by far the largest
zling why Schmuths et al. (2004) obtained a 1Cx value of number of new and novel species’ values, and the largest
0.206 pg for A. thaliana ‘Columbia’, which was 26% higher number of new sources, published together, among the nine
than the value of 0.163 pg for the same species measured compendia published since 1976 (see Table 1). So how do
using the completely sequenced genome of Caenorhabditis the new data contribute to the targets set for plant genome
elegans as the calibration standard (Bennett et al., 2003). size research?
However,thisseemscleargiventheextrainformationinfoot-
note‘ac’(fororiginal reference number 494).This showsthat
Meeting targets for species representation
Schmuths etal.(2004)choseRaphanussativusastheinternal
calibration standard using a 2C-value of 1.38 pg, which they Previous compilations have tracked the annual output of
reported had been taken from Dolezˇel et al. (1998), rather DNA C-value estimates and shown a strong rising trend in
than the 2C-value estimate for Raphanus sativus ‘Saxa’ of the number of total and first estimates since 1960 (Bennett
1.11 pg given by Dolezˇel et al. (1992) – a difference of+ and Leitch, 2005a). Updating this analysis to include values
Bennett & Leitch – Nuclear DNA amounts in angiosperms Page 5 of 124
publishedhere(comprisingdataincludedincompilations1–8 analysis shows better progress as the addition of C-values
and the present Appendix – see Table 1) shows that these for a further 2155 species compiled over this time frame
trends were strongly maintained. Thus, the mean total shows that 86% of the target was met.
number of C-values for angiosperm species per year rose 69 However, recently the total number of angiosperm species
% from 311 in 2000–2004 to 527 in 2005–2008, whilst the has been revised upward to a widely accepted 352000
mean total number of first values for angiosperm species (Paton et al., 2008). Using this higher value, the percentage
rose 72% from 219 in 2000–2004 to 378 in 2005–2008 of species with C-values still stands at only 1.8%.
(Fig. 1). Clearly, encouraging progress towards increasing Nevertheless, if present rates of data acquisition are main-
the representation of angiosperm species with genome size tained, the sample of species for which C-values are listed
data continues to be made. Indeed the slope of the graph in should reach 2% (i.e. 7040) and 3% (i.e. 10560) of the
Fig. 1 shows that the last 3-year period (2005–2008) had the higher number within about 3 and 12 years, respectively.
highestrateofgenomesizedatagenerationforanyperiodana- Clearly the setting of targets has been invaluable for focus-
lysed since the first plant genome sizewas estimated in 1950. ingandimprovingtherepresentationofspecieswithC-values.
Itisthereforetimelytoreviewandassesshowfarprogresshas It therefore seems reasonable to maintain a target of
been made towards meeting targets set for increasing species first C-values for angiosperm species at 2500 for the next
representation for angiosperms. 5 years. If rates of data acquisition continue at the same rate
Bennett and Smith (1976) listed C-values for 753 angios- then such a goal seems reasonable, but attaining 10% D
o
perm species, and it was not until 1995 that data for 2500 species representation (i.e. approx. 35200, the ultimate goal w
n
species were available. It was then conservatively estimated which seemed sensible to Bennett and Leitch, 2005a) would lo
a
that the total number of angiosperm species was still be estimated to take 60–80 years. de
d
about 250000, thus C-value data were already available for fro
1% of species. At the 1997 Angiosperm Genome Size m
Progress towardstargets for familial representation h
Workshop, with C-values for 2802 species compiled into the ttp
Angiosperm DNA C-values Database, a target was set to In1997,C-valueswerestillunavailablefor68%ofangios- ://a
o
measure a further 1% of species (i.e. approx. 2500 species) perm families recognized by the first Angiosperm Phylogeny b
.o
by 2003. Analysis shows that this target was aspirational, as Group (APG) classification published in 1998 (APG, 1998). x
fo
a review at the Plant Genome Size meeting held in 2003 Thus, at the first Angiosperm Genome Size workshop (1997) rd
showed that C-values for only a further 1032 species had a goal was set to complete familial representation by 2002. jou
been collated, corresponding to just 41% of the target, This proved over-optimistic. Progress was initially very slow, rna
although knowledge of further data as yet unpublished and/ as only 12 (1.7%) of the 691 species’ values in the fifth ls.o
rg
or collated suggested that approx. 66% of the target may DNA compilation (Bennett et al., 2000) were also first a/
have been met (Bennett and Leitch, 2005a). Nevertheless, values for families. Nevertheless, new work focused to t N
the 2003 meeting reset the target to measure a further 2500 addressthe shortfall produced first C-values forover 80 unre- IH
L
species in the next quinquennium (i.e. 2003–2008) and presentedfamiliesby2003(e.g.Hansonetal.,2001a,b,2003; ib
Leitch and Hanson, 2002). This increased familial represen- rary
tation to 47.5% (Bennett and Leitch, 2005a), based on o
n
C-value data for 217 families out of the 457 families recog- D
500 Total ec
First nized in the revised angiosperm phylogeny (APG II, 2003) em
(Fig. 2). As this still fell well short of the original target, it b
e
mates 400 w20a0s3retvoisaedndewowgnoaatltohfe7P5la%nt GfaemniolimalerSeipzreesMeneteattiinogn.hBeladseidn r 30, 2
esti on the number of families recognized by APG II (2003), this 015
of 300 would require the measurement of an additional 114 families.
er Analysis of the 1860 first C-values for species listed in the
b
m current Appendix shows that only 24 (0.13%) are also first
n nu 200 estimates for families, representing just 21% of the target of
ea 114 families. Combining these with data in the Plant DNA
M
C-values Database shows that in total there are currently 240
100
angiosperm families with genome size data. Given that a
further revision by the APG (APG III, 2009) now recognizes
0 only 415 families (10% down on the number recognized
54 59 64 69 74 79 84 89 94 99 04 08 in APG II), the current familial representation now stands at
9 9 9 9 9 9 9 9 9 9 0 0
1 1 1 1 1 1 1 1 1 1 2 2 58%, so progress towards achieving even the reduced target
– – – – – – – – – – – –
50 55 60 65 70 75 80 85 90 95 00 05 set of 75% familial representation in the last quinquennium
9 9 9 9 9 9 9 9 9 9 0 0
1 1 1 1 1 1 1 1 1 1 2 2 (i.e. 2003–2008) has remained poor. Major factors limiting
Period progress have been discussed previously (Hanson et al.,
2003)ashastheexpectationthatsuchdifficultieswillintensify
FIG. 1. Mean number per year of total and ‘first’ DNAC-value estimates as familial representation approaches 100%. Nevertheless,
communicatedin11successive 5-yearperiodsandthe4-yearperiod2005–
2008, between 1950 and 2008. Data taken from the Plant DNA C-values progress since 2003 has been disappointing in relation to the
Database(release5.0,December2010). size and distribution of the shortfall and has again depended
Page 6 of 124 Bennett & Leitch – Nuclear DNA amounts in angiosperms
60 representation would require a first C-value for an additional
approx. 400 genera and so this target was set. Subsequently,
the total number of genera was revised down to 12962
50
s (Mabberley, 2008), which would make achieving 10% rep-
e
mili resentation easier, requiring a first C-value for only approx.
G fa 40 2ta5i0nsgCen-evraal.uAesnafolyrs5is5o4fgtehneedraataofinwthhiechAp3p2e9n(d5ix9.s4h%ow)sarietcfiorns-t
P
A C-values for a genus. Overall the 6287 species with genome
% of 30 size data include 1444 genera, an increase of 402 genera
e since 2003. Thus, work in the last 5 years has made consider-
v
ati 20 able progress towards meeting the generic target, increasing
ul representation to 11.2%. Indeed, progress towards this target
m
u was outstanding, greatly exceeding that at both specific and
C
10 familial levels noted above. Given the current focus on
genera it seems reasonable to keep a similar target of first
C-values for a further 450–500 genera over the next 5 years,
0
1950 1960 1970 1980 1990 2000 to enable representation to approach approx. 15%. D
Year ow
n
lo
FIG. 2. Cumulative percentage of angiosperm families recognized by the Improved representation of other groups ad
AngiospermPhylogenyGroup(APG)(APGIII,2009)withafirstC-valuerep- ed
resentedinthepresentPlantDNAC-valuesDatabase(release5.0,December Although the Plant Genome Size Workshops have noted fro
2010). other gaps in general discussions, they have not set any m
h
targets to fill them, choosing to focus first on improved sys- ttp
mainlyonworkbyjustoneinstitution – RBG,Kew[i.e.RBG, tematic representation in angiosperms, and on widening rep- ://a
o
Kew, estimated 65 of the 74 (87%) first values for families resentation of other plant groups, such as gymnosperms, b.o
added in 1997–2002, and 18 of the 25 (75%) first values pteridophytes (i.e. monilophytes and lycophytes), bryophytes xfo
for families added in the present Appendix]. Bennett and and algae. Nevertheless, previous papers in this series have rd
Leitch (2005a) previously noted that ‘...international collab- identified particular major gaps in our knowledge of angios- jou
rn
oration to locate materials and estimate genome sizes for still perm C-values (e.g. Bennett and Leitch, 1995, 2005a), als
unsampled families will be essential if the long term goals which merit research (but without recommending targets and .org
of 75% and then complete familial representation are to be time frames). Clearly this was influential as many of these a/
achieved’. Such progress will require champions to focus on gaps have subsequently received significant attention. For t N
this target, collaboration to share the load more broadly and example, Bennett and Leitch (1995) noted that ‘estimates IH L
funds for collecting exceptional and esoteric material of the wererareformanytypesofangiospermssuchas:palms,para- ib
approx.70or175familiesstillneededfor75or100%familial sitic plants, noxious weeds and succulents’. Studies soon rary
representation, respectively. addressed some, but not all, of these groups listing first on
C-values for 74 palms (Ro¨ser et al., 1997) and 39 weeds D
e
(Bennett et al., 1998). ce
m
Althoughsomeprogresshadbeenmade,BennettandLeitch b
Improved systematic representation for genera (2005a) noted several ongoing gaps in geographical and plant er 3
0
Within the keyrecommendations forangiosperms set at the life form, including ‘island floras rich in endemics’ and ‘taxa , 2
0
Plant Genome Size meeting in 2003 (http://data.kew.org/ from bog, fen, tundra, alpine and desert environments, and 15
cvalues/pgsm/#9%20Key%20recommendations) was a target halophytic, insectivorous, parasitic, saprophytic and epiphytic
of achieving 10% generic representation. Although targets species andtheirassociated taxa’. Examining the 151original
for species and familial representation have been set and references contributing to the Appendix shows they include
tracked since 1997, none was set for genera prior to 2003, so significant contributions to filling some of the above gaps.
2003–2008 is the first quinquennium for which this target Several provide data for island floras. Murray et al. (2003,
has been monitored. The reason for adding this new level 2005) give genome sizes for 158 endemic and indigenous
was to reflect developments in the field of angiosperm phylo- grasses of New Zealand, whilst Suda et al. (2003, 2005) list
genetics based on molecular studies. By 2003 the APG had values for species representing 40% of the Macaronesian
transformed the view of higher order angiosperm phylogeny, angiosperms. Data for 13 Oncidiinae orchids (Chase et al.,
bringing new certainty and stability at the familial level and 2005) and 18 Cuscuta vines (McNeal et al., 2007) have use-
replacing several different conflicting non-molecular systems. fully increased representation for epiphytes. The latter,
Consequently, the focus of new molecular-based work was together with estimates for approx. 40 taxa of holoparasitic
shifted to the generic level. Orobanche (Weiss-Schneeweiss et al., 2005) and for several
Analysis of the 4119 species listed by 2003 showed DNA taxa in two genera of Cistaceae (Halimium and Xolantha;
C-values for 1042 genera, out of the then approx. 14000 Boscaiu et al., 2008) have also significantly improved rep-
genera recognized (M. Chase, Rotal Botanic Garden, Kew, resentation for parasitic plants with first values for approx.
London, pers. comm.), equivalent to approx. 7.4% represen- 70 species. Finally, estimates for four Adriatic seagrasses
tation overall. Thus, it was calculated that achieving 10% (Dolenc Koce et al., 2003) and 13 marine halophytes
Bennett & Leitch – Nuclear DNA amounts in angiosperms Page 7 of 124
(DolencKoceetal.,2008)havemadeapositivestartindirect 300
A
response to address a gap identified in Bennett and Leitch Fe (Fe and CIA)
(2005a).
ar 250 FC (all fluorochromes)
e
y
s/
e
TRENDS IN TECHNIQUES USED TO ESTIMATE mat 200
GENOME SIZE sti
e
Therise in flowcytometryasthe technique ofchoiceforgenome st
size estimations of fir 150
A continued interest in genome size research has focused on ber
m 100
the techniques used to obtain such data. Previous analyses u
n
haveshownatrendtowardstheincreasinguseofflowcytome- n
trysince 1983 (Galbraith et al., 1983) to reach 58.4% of first ea 50
M
estimates listed in the seventh compilation (i.e. Bennett and
Leitch, 2005a). Updating this analysis shows a further strong
0
rise in the use of flow cytometry in recent years, with 4 4 4 4 4 4 D
Feulgen microscopy methods continuing to fall (Fig. 3A). 95 96 97 98 99 00 ow
1 1 1 1 1 2 n
Indeed, of the 1860 first estimates listed in the present – – – – – – lo
0 0 0 0 0 0 a
1A5p.p4e%ndixu,se8d4.5F%eulwgeenre mmiacdreosucsoipnyg flmoewthcoydtso.mDetirvyidbiuntgonthlye 195 196 197 Yea198r 199 200 ded fro
latter into traditional Feulgen microdensitometry and m
h
computer-basedimageanalysis(CIA)ofFeulgen-stainedprep- 100 ttp
arations(Fig.3B)showsthatalthoughtheuseofCIAgenerally B Traditional Fe ://a
o
rose between 2000 and 2006, more recently its use has b
CIA .o
declined (contributing just 31 C-value estimates in 2007 and 80 xfo
2008, Fig. 3B) along with that of the more traditional s rd
Feulgen microdensitometry methods. This trend is not unex- ate jou
m rn
psietcotmedetberust arenfldecthtsetfhaeilfiurnealtolodssevoeflotrpadait‘iostnaanldparhdotionmexicpreondseivne- st esti 60 als.org
CafecIrcA’e,scskotinot,dlieatainodniansggwrelahebdicohrbaetBosertinepnsreaftoctrtaictnreadinCLinIeAgitcahtnedc(2hte0nc0iqh5unaeo)losaugngydgtereasatnessyd- er of fir 40 at NIH/
mb L
would be essential for CIA to become the realistic alternative u ib
to flow cytometry for estimating C-values. Indeed, CIA has N 20 rary
advantages for some materials in providing cytological sights o
n
and insights into the material being studied (such as chromo- D
e
c
some number, aneuploidy, B-chromosomes and nuclear stab- e
0 m
ility) unavailable to users of flow cytometry without also b
making separate cytological preparations. Clearly, the use of 2000 2001 2002 2003 2004 2005 2006 2007 2008 er 3
Year 0
Feulgen microscopy with CIA has not been favoured by most , 2
0
practitioners in the plant genome size community. FIG. 3. (A) Mean number per yearof ‘first’ DNAC-value estimates made 15
Analysis of data in the Animal Genome Size Database usingFeulgenmicrodensitometry(Fe)andflowcytometry(FC)communicated
in 11 successive 5-year periods and the 4-year period 2005–2008, between
(Gregory, 2006) shows a different picture. Here, 1151 esti-
1950 and 2008. (B) Number of first estimates made using traditional
mates were made using CIA for the period 1991–2009 (com- Feulgen methods (Traditional Fe) and Feulgen staining measured using
pared with 190 in plants). Nevertheless, of the 29 original computer-aidedimageanalysis(CIA)since2000.DatatakenfromthePlant
sources giving such data, 13 (which list data for 549 taxa) DNAC-valuesDatabase(release5.0,December2010).
were published by T. R. Gregory, so (as with many other
aspects of genome size research) this trend has been strongly
of some equipment made for a different prime use in a
influenced by the output of just one productive practitioner
larger (e.g. medical) market.
and his colleagues. In plants, all recent estimates made using
CIA were co-authored by Vilhar in Slovenia or come from
the University of Vienna where Greilhuber and colleagues
Development of different isolation buffers for flow cytometry
use the CIRES system, which is now unavailable to purchase.
Inviewoftheabove,theuseofCIAtoestimateplantgenome Since 1983,flowcytometry hasbeen increasinglyusedasa
sizesisunlikelytoincreaseagainunlessmanufacturerssupply method to estimate genome size (Galbraith et al., 1983), and
and support complete and inexpensive CIA packages which the composition and number of different types of isolation
compete with the options provided by flow cytometry buffers has grown considerably, reflecting the complex diver-
makers. This now seems unlikely given the small size of the sity of biochemistry in plant cells. Indeed, in the FLOWer
genome size market, but could still occur as a marginal use database for plant flow cytometry (Loureiro et al., 2007b,
Page 8 of 124 Bennett & Leitch – Nuclear DNA amounts in angiosperms
2008) there are 28 different buffers which have been used Recent developments in the application of flow cytometry to
based on 826 research articles included in the database. genome size studies
Recently, the performances of four of the most commonly
(i)Theuseofseeds. Recent reports suggest that dryseeds may
used flow cytometry buffers (Galbraith’s, LB01, Otto’s and
be suitable material for genome size estimations made using
Tris.MgCl ) on seven plant species were compared by
2 flow cytometry (Sliwinska et al., 2005, 2009; Jedrzejczyk
Loureiro et al. (2006a). By analysing various parameters of
and Sliwinska, 2010). This approach, if shown to be widely
the flow histogram including forward scatter, side scatter and
applicable, has several advantages: (1) it enables the direct
relative fluorescence intensity of propidium iodide-stained
analysis of seed stored in seed banks (provided the owners
nuclei, the performance of each buffer was assessed. They
of the genetic resources approve), (2) it overcomes the pro-
concluded that no single buffer worked well for all species.
blems associated with analysing fresh material collected in
Instead, the optimal buffer for a particular species depended
the field within a reasonable time frame before material has
on a number of factors, although Otto’s and LB01 buffers
degraded, and (3) it provides an alternative source of tissue
wereconsideredtobethebestforthemajorityofspeciesana-
for analysis if the accumulation of staining inhibitors in
lysed. Subsequently, Loureiro and colleagues have developed
other tissues has been shown to be problematic. Indeed, dry
twonewbuffers,ageneral-purposebufferandonespecifically
seed may contain a lower level of cytosolic inhibitors than
developed for use with woody material (Loureiro et al.,
other parts of the plant and hence may be easier to analyse
2007a). D
(Greilhuber et al., 2007). However, there is the potential that o
Tworecentstudieshaveextendedourunderstandingonhow w
the seeds are intra- or inter-specific hybrids rather than n
cytosolic compounds, which are released into the isolation lo
representing the genomic make-up of the parent. In addition, a
bufferduring nucleiisolation,can interferewith fluorochrome de
using whole seeds for genome size studies means that all the d
stainingandleadtoerroneousresultsandpseudo-genomesize material is destroyed, thereby preventing further genomic fro
plasticity (Loureiro et al., 2006b; Bennett et al., 2008). m
analysis (chromosome counts etc.). An alternative approach h
Bennett et al. investigated the effects of anthocyanin, while is to germinate the collected seed and analyse the growing ttp
Loureiro et al. examined tannic acid. Both groups of com- plant material. In a large survey of the Macronesian flora by ://ao
pounds, which are widespread across angiosperms, were b
Suda et al. (2005) such an approach was taken. Seeds for .o
shown tohavepotenteffects onDNA stainingleadingto stoi- x
over 100 taxa were collected from the Canary Islands during fo
chiometric errors. Such studies emphasize the need to use rd
internal standardization and the importance of testing for fieldwork and then brought back to the Czech Republic for jou
inhibitory compounds as recommended by Dolezˇel (1991) germination and subsequent flow cytometric and cytological rna
analysis. ls
and as more recently discussed in detail by Greilhuber et al. .o
rg
(2007). (ii) Ease of access to methodological data. The launch of the a/
FLOWer database (Loureiro et al., 2007b, 2008) in 2007 has t N
IH
The application of flow cytometry to plant systematics providedavaluableresourceenablingusersofflowcytometry L
toaccessandassessdetailsoftheprotocolsusedbyotherprac- ib
Withtheabilityofflowcytometrytoanalyselargenumbers titionerstoassistthem inchoosing themostlikelymethod for rary
of individuals within and between populations, genome size analysing the plant species of interest. on
data are now increasingly being used to help resolve taxo- D
e
c
nomic issues. These have included examples where genome (iii) New equipment. Another development which may assist e
m
size has been used to provide new insights into species in extending the ease of genome size analysis includes b
e
relationships (e.g. Agapanthus, Zonneveld and Duncan, the launch of a new type of flow cytometer, the Accuri C6, r 3
0
2003; Galanthus, Zonneveld et al., 2003; Nerine, Zonneveld by Accuri Cytometers in December 2006 (www. , 2
0
and Duncan, 2006; Curcuma, Leong-Sˇkornickova` et al., AccuriCytometers.com). The company promised a revolution 15
2007; Hieracium, Suda et al., 2007; Narcissus, Zonneveld, in flow cytometry with ‘reliable, high performance fluidics,
2008) as well as for recognizing new taxonomic entities (e.g. inexpensive, robust optics, and sophisticated and simplified
Cochlearia borzaeana, Kochjarova´ et al., 2006; Pandanus electronics’ – allforapprox.£30000.Oneofthemaindiffer-
fascicularis, Panda et al., 2009), supporting existing ones ences between such a machine and conventional flow cyt-
(e.g. Lilium bosniacum, Muratovic et al., 2005) or rejecting ometers such as those of Partec, Beckman Coulter or
them (e.g. Lamium endtmannii, Rosenbaumova´ et al., 2004). Beckton Dickinson is the greater dynamic range of operation
Genome size has also been used as a reliable guide in the (as it uses a 24-bit analog-to-digital converter for signal pro-
separation of closely related species and their hybrid such as cessing), enabling species with widely different genome
shown in studies of Elytrigia repens, E. intermedia and their sizes to be analysed together in a single sample. Indeed,
hybrid (Mahelka et al., 2005) (original reference 593 in the tests on the machine carried out by Galbraith (2009) have
Appendix). Mahelka et al. (2005) analysed 238 individuals demonstrated that species with 1C-values ranging from 0.16
from 55 different populations and found that there was so to40pgcanbevisualizedandanalysedonthesameflowhis-
little intraspecific variation in DNA amount for each of the togram. Galbraith demonstrated that the machine had the
two species and the hybrid that genome size could be used potential to measure genome sizes in species with C-values
as a reliable taxonomic indicator. Indeed, genome size was ranging from 0.032 to 185 pg and hence was unlikely to be
showntobemorerobustandlesssubjectivethanmorphologi- limited by this character. Galbraith concluded that ‘for
cal characters that varied more or less continuously between routine analysis of plant nuclear DNA contents, ploidy or
the species and hybrid. investigations of other issues requiring C-value
Bennett & Leitch – Nuclear DNA amounts in angiosperms Page 9 of 124
determinations, the Accuri C6 flow cytometer provides an interest and (2) an accurate method to estimate concentration
excellent platform’. He also suggested that given the reprodu- of genomic DNA. qPCR was then performed and genome
cibility and simplicity of use it might now be feasible to rea- size calculated by dividing the amount of DNA present in
listically consider estimating genome sizes in the remaining the sample by the number of copies of the single copy gene
98% of angiosperm species (Galbraith, 2009). Nevertheless, detected. The method has since been used to estimate
althoughGalbraithconsideredthatitmightbepossibletoesti- genome sizes in several eukaryotic species with genomes up
mate genome sizes in 12 species per hour, in reality he also to 2900 Mb. However, comparisons with estimates made
recognized that potential problems arising from the need to using other methods have not always produced similar
optimize buffers and test for secondary products would inevi- results (Table 2) and reasons for the discrepancies remain
tablyslowthingsdown.Inaddition,theneedtomakechromo- unclear. Recent modifications to the technique may overcome
some counts will further impede output – a factor often some of the initial problems identified (e.g. Jeyaprakash and
omitted by flow cytometer practitioners when extolling the Hoy, 2009), but given that (1) considerable time is required
greater speed of flow cytometry compared with Feulgen to optimize conditions for qPCR and (2) prior molecular
microscopy. Nevertheless, the availability of such a compact, knowledge is needed to identify suitable, specific primers for
inexpensive, yet easy to use machine is likely to make an amplifying a unique sequence, whether the method is widely
impactinthefieldofgenomesizeresearchasolder,moretra- applicable remains unknown even for species with small
ditional flow cytometers are replaced and updated. genomes. Moreover, there are as yet no data to demonstrate D
o
howaccuratethemethodisforgenomeslargerthan3000Mb. w
n
Indeed, the general suitability of the method for angios- lo
a
Are there any new techniques forestimating genome size on the perms remains unclear as there are no reports of plant de
d
horizon? genome size estimations using qPCR. Nevertheless, given fro
(1) the propensityof plant genomesto undergo both localized m
Sincethe1950s,nuclearDNAamountshavebeenestimated h
by at least eight different techniques, including: chemical andwholegenomeduplicationsand(2)theneedtoempirically ttp
extraction, Feulgen microdensitometry, flow cytometry, CIA, select a suitable single copy gene and develop appropriate ://ao
primers for each species investigated, it seems unlikely the b
reassociation kinetics, pulse field gel electrophoresis and .o
method will be widely used. It is too time consuming and x
‘complete’ genome sequencing. From an analysis of all the fo
expensiveevenifitcanbeshowntoworkacrossthespectrum rd
genome size data in the Plant DNA C-values Database and of genome sizes encountered in plants. jou
as noted above, just two (Feulgen microdensitometry and rn
a
flow cytometry) have contributed over 96% of first estimates ls
.o
forangiosperm species and it seems clear that flow cytometry (ii)Will‘complete’genomesequencinggiveuseablegenomesize rg
will remain the dominant method of choice for estimating estimates? It is now 10 years since the first complete genome a/
t N
DNA C-values in plants in the near future. Indeed, it may sequence for an angiosperm was published (Arabidopsis IH
become almost the sole method unless CIA or some new Genome Initiative, 2000), and to date similar data for a L
method takes off. So are there any realistic alternatives on further eight angiosperm species has followed (Table 3). It ibra
the horizon? istherefore timely to assessthe actual contribution that com- ry
o
pletegenomesequencinghasmadetoourknowledgeofplant n
D
(i)Canreal-timePCRbeusedforestimatingplantgenomesizes? DNA C-values, and to ask if this molecular approach has e
c
e
In 2003, a molecular-based method to estimate genome size yielded more accurate genome size measurements than were m
b
was published which used the quantitative real-time polymer- previously available, and thereby provided the new ‘gold er 3
asechainreaction(qPCR)(Wilhelmetal.,2003).Themethod standard’ calibration standards against which other species 0
required (1) a cloned single copy gene from the species of can be compared. Bennett and Leitch (2005a) discussed , 20
1
5
TABLE 2. Comparison of genome sizes estimated by quantitative real-time PCR (qRT-PCR) with other methods for various
eukaryotic species (PFGE, pulse field gel electrophoresis; FC, flow cytometry; Fe, Feulgen microdensitometry; RK, reassociation
kinetics;Ch,chemicaldetermination).
Species Group 1CvalueestimatedusingqRTPCR Reference Rangeof1Cvaluesestimatedbyothermethods*
Saccharomycescerevisiae(yeast) Fungus 12.1Mb Wilhelmetal.(2003) 12.06–12.2Mb(completegenomesequencing)1
Piriformosporaindica Fungus 15.6–24Mb Zuccaroetal.(2009) 15.7Mb(PFGE)2
(dependingongeneused)
Thalassiosirapseudonana Diatom 35.9Mb vonDassowetal. 31.3Mb(completegenomesequencing)3
(2008)
Muscadomestica(housefly) Insect 295Mb GaoandScott(2006) 950Mb(spectrophotometry)4,900Mb(FC)5
Drosophilamelanogaster(fruitfly) Insect 184Mb GaoandScott(2006) 157–206Mb(FC)4
Xiphophorusmaculatus(platyfish) Fish 550Mb Wilhelmetal.(2003) 410–550Mb(Fe,RK)6,760–990Mb(FC)4
Homosapiens(human) Mammal 2900Mb Wilhelmetal.(2003) 2940Mb(Ch)7to3355Mb(Ch)8
*1FungalGenomeSizeDatabase(Kullmanetal.,2005);2Zuccaroetal.(2009);3Armbrustetal(2004);4AnimalGenomeSizeDatabase(Gregory,2006);
5J.S.Johnston(TexasA&MUniversity,pers.comm.,2008);6Wilhelmetal.(2003);7VendrelyandVendrely(1948);8Mandeletal.(1950).
TABLE 3. ComparisonbetweentheamountofDNAsequencedandtotalgenomesizein‘complete’plantgenomesequencingprojects Pa
g
e
Actualamountof 1
0
Totalgenomesize assembledsequence o
reportedbythe generatedbythe Percentageofgenome f
1
‘complete’genome ‘complete’genome estimatedtohavebeen 2
sequencingproject sequencingproject sequenced(column3/ 1C-valueestimatedbyother 4
Species‡ (1C-valueinMbp) (Mbp) column2) methods Comments
Arabidopsisthaliana 125† 115.4 92%assuminga 157MbusingFC:PIand A1C-valueof125MbgivenbytheArabidopsisGenome
‘Columbia’1 1C-valueof125Mbbut Caenorhabditiselegansas Initiativewasbasedonadding115.4Mbofassembled
(Arabidopsis) actuallyonly73.5%if standard(1C¼100Mb) sequencetoanestimateof10Mbfortheunsequenced
FamilyBrassicaceae thehigher1C-valueof (Bennettetal.,2003) centromeresandribosomalDNAregions(Arabidopsis
157Mbisused GenomeInitiative,2000).
Oryzasativasubsp. 466† 362 77.6% 510MbusingFeandVigna Yuetal.(2002)estimatedthattherepetitivesequenceswhich Be
indica‘93-11’2 radiata(1C¼0.53pg) weredifficulttoalignaccountedfor104MbofDNA.When nn
(Rice) (BennettandSmith,1991) addedtothetotalamountofalignedsequencedatathisgave e
tt
FamilyPoaceae anestimatedgenomesizeof466Mbasstatedincolumn2.
&
Oryzasativasubsp. 420* 389.8 n/a 510MbusingFeandVigna Noattemptwasmadetoestimateatotalgenomesize.Instead, L
j‘aNpiopnpiocnabare’3 r(Badeinantaett(1aCnd¼S0m.5it3h,p1g9)91) aSy1nCg-evnatlauegroofup4)20alMthobupgwhatshelisstoeudrcbeyoGfothffisetvaallu.e(,2a0n0d2h–ence eitch
(Rice) howitwasdetermined,wasnotstated. –
Oryzasativasubsp. 389† 370 95% 510MbusingFeandVigna ThisworkbuiltonthatofSyngenta(seeabove,Goffetal., Nu
japonica radiata(1C¼0.53pg) 2002)toproduceahigherqualitysequenceestimatedto c
‘Nipponbare’4 (BennettandSmith,1991) comprise95%ofthewholegenome,includingallthe lea
(Rice) euchromaticregions.The1C-valueof389Mbpwasbasedon r
addingtheestimatedlengthof62gaps(¼18.1Mb),which DN
includedninecentromeresand17telomerestothetotallength A
ofalignedsequence(¼370Mb). a
m
Populustrichocarpa 485†+10 410 85%oftotalgenomebut 686MbusingFC:DAPIand Theauthorsconsideredthatthe410Mbofassembled o
u
genotype‘Nisqually approx.90%of chickenredbloodcellsas sequencedatarepresentedtheeuchromaticportionofthe n
1’5 euchromaticregion standard(1C¼1.21pg) genome.Thetotalgenomesizewasbasedonaddingthisto ts
(blackcottonwood) (BradshawandStettler, 75Mbofgenomicsequencethatcouldnotbeassembledinto in
FamilySalicaceae 1993) anyofthemajorscaffolds.Theauthorssuggestedthiswas a
n
consistentwithcytogeneticevidenceshowingapprox.30%of g
thegenomewasheterochromatic.ThiswasbasedonDAPI io
s
stainingofsevenprophaseandmetaphasecellsfollowedby p
e
physicalmeasurementsofheterochromaticbands(NB: rm
adjustmentsweremadeforthedifferentialcontractionof s
heterochromatinversuseuchromatin).
Vitisvinifera 475* 487 Seecomments 415MbusingFC:PIand Noattemptwasmadetoestimateatotalgenomesize.Instead,
genotypePN400246 chickenredbloodcellsas the1C-valueof475MbgivenbytheFrench-ItalianPublic
(grapevine) standard(LodhiandReisch, ConsortiumforGrapevineGenomeCharacterization(2007)
FamilyVitaceae 1995) wastakenfromLodhiandReisch(1995).Actually,thisfigure
comesfromtheabstractofLodhiandReischwhostatedthat
‘theVitisgenomesizeis475Mb’basedonmeasurementsfor
19speciesincludingV.vinifera.ThemeasurementsofLodhi
andReischweremadeusingflowcytometryandchickenred
bloodcellsasstandard(assuming2C¼2.33pg)andshowed
therewasintraspecificvariationbetweendifferentvarietiesof
Vitisviniferarangingfrom1C¼415to511Mb.Thecultivar
‘PinotNoir’,whichisclosesttothevarietyusedforgenome
sequencing,wasestimatedtohaveagenomesizeof415Mb.
5102 ,03 rebmeceD no yrarbiL HIN ta /gro.slanruojdrofxo.boa//:ptth morf dedaolnwoD