BioMed Cen al
Page 1 o 5
(page numbe no o ci a ion pu poses)
BMC Genomics
Open Access
Resea ch a icle
Templa e swi ching can c ea e complex LTR e o ansposon
inse ions in T i iceae genomes
F ançois Sabo 1 and Alan H Schulman*1,2
Add ess: 1MTT/BI Plan Genomics Labo a o y, Ins i u e o Bio echnology, Viikki Biocen e , Uni e si y o Helsinki, P.O. Box 56, FIN-00014
Uni e si y o Helsinki, Finland and 2Plan Genomics, Bio echnology and Food Resea ch, MTT Ag i ood Resea ch Finland, Mylly ie 10, FIN-31600
Jokioinen, Finland
Email: F ançois Sabo - ancois[email p o ec ed]; Alan H Schulman* - [email p o ec ed]
* Co esponding au ho
Abs ac
Backg ound: The LTR (long e minal epea ) e o ansposons o highe plan s a e eplica ed by
a mu agenic li e cycle con aining ansc ip ion and e e se ansc ip ion s eps. The DNA copies a e
o en subjec o ecombina ion once in eg a ed in o he genome. Complex elemen s, whe e wo
elemen s sha e an LTR, a e no uncommon. They a e hough o esul om he e ologous
ecombina ion be ween wo adjacen elemen s ha occu s ollowing hei in eg a ion.
Resul s: He e, we p esen e idence o ano he po en ial mechanism o he c ea ion o complex
elemen s, in ol ing abno mal empla e swi ching du ing e e se ansc ip ion. The empla e
swi ching c ea es a la ge, complex daugh e elemen , o med by he usion o wo pa en
sequences, which is hen inse ed in o he genome.
Conclusion: Those complex elemen s a e pa o he genome s uc u e o plan s in he Poaceae,
especially in he T i iceae, bu no o A abidopsis. Hence, e o ansposon dynamics shaping he
genome a e lineage-speci ic.
Backg ound
Long Te minal Repea (LTR) e o ansposons a e Class I
ansposable elemen s ha eplica e by a "Copy-and-
Pas e" mechanism, called e o ansposi ion, which is
qui e simila o len i i us (such as he HIV) eplica ion.
Highe plan genomes, especially o he g asses (such as
maize, whea and ba ley), ha bo a la ge numbe o hese
elemen s, which o m he as majo i y o he nuclea
DNA. Re o ansposi ion in ol es a e e se ansc ip ion
s ep, whe e cDNA is syn hesized om an RNA empla e.
Re e se ansc ip ion is ca alyzed by e e se ansc ip ase,
which is gene ally encoded by he e o ansposon being
copied, and he cDNA is inse ed in o a new genomic loca-
ion by he in eg ase, which is also sel -encoded [1]. A
canonical e o ansposon inse ion comp ises wo LTRs
and an in e nal domain con aining he coding domain o
in eg ase, e e se ansc ip ase, a p o einase, he s uc u al
p o ein GAG, and he signals o e e se ansc ip ion.
Many composi e s uc u al pa e ns de i ed om canoni-
cal LTR e o ansposon inse ions we e p e iously iden i-
ied in BACs and o he s long genomic sequences om
a ious plan s (Figu e 1; [2-7] and e e ences wi hin).
These appea p ima ily as nes ed inse ions o one e oe-
lemen in o ano he . The nes s can comp ise mo e han
h ee o ou laye s a anged in a "Russian doll" ashion.
In some cases, he nes ed e oelemen s a e solo LTRs
a he han elemen s con aining wo LTRs and a cen al
domain. The solo LTRs a e hough o a ise om non-
Published: 24 July 2007
BMC Genomics 2007, 8:247 doi:10.1186/1471-2164-8-247
Recei ed: 3 Ap il 2007
Accep ed: 24 July 2007
This a icle is a ailable om: h p://www.biomedcen al.com/1471-2164/8/247
© 2007 Sabo and Schulman; licensee BioMed Cen al L d.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/2.0),
which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
BMC Genomics 2007, 8:247 h p://www.biomedcen al.com/1471-2164/8/247
Page 2 o 5
(page numbe no o ci a ion pu poses)
ecip ocal ecombina ion be ween he LTRs o a single ele-
men (Figu e 1, second case; [2]).
In addi ion o he nes s, some complex inse ions a e
cha ac e ized by a hi d LTR sha ed be ween wo po en-
ially comple e elemen s (Figu e 1, bo om, "Complex
inse ion"). Unde s anding he mechanism h ough
which his class o e o ansposon complex is de i ed is
equi ed o a ull ision o genome e olu ion. Au ho s
ha e p e iously desc ibed hem as me ely he esul o
ecombina ion be ween he p oximal LTRs o wo adja-
cen e oelemen s, leading o he elimina ion o he in e -
ening genomic sequence (Figu e 2; om [2,8]). He e we
will desc ibe ano he possible o igin o hese complex
inse ions, an abno mal empla e-swi ching du ing he
e e se ansc ip ion s ep.
Resul s and discussion
In he genome o A abidopsis haliana, a e ho ough anal-
yses, De os e al. iden i ied no complex elemen s o he
han hose o igina ing om ecombina ion be ween wo
e oelemen s [2]. Fo hese, he wo ou e mos LTRs di e
om each o he by no being de i ed om he same
e e se ansc ip ion and in eg a ion. This has wo s uc-
u al consequences. Fi s , a ecombina ion be ween he 3'
LTR om one elemen and he 5' om ano he , closely
ela ed one on he same s and (Figu e 2) gi es ise o a
hi d, in e nal LTR. This LTR is a chime a o he wo LTRs
in ol ed in he ecombina ion. A second consequence,
because he wo elemen s in ol ed come om wo inde-
penden inse ion e en s ha gene a ed wo di e en a -
ge -si e duplica ions (TSDs), is ha he esul ing complex
does no ha bo lanking TSDs. By hese measu es, he
as majo i y o he complex elemen s al eady iden i ied
a ose om unequal and he e ologous ecombina ion
be ween adjacen and independen inse ions [2,8].
Ne e heless, a ca e ul analysis o he complex inse ions
o a ailable la ge genomic sequences om he T i iceae
has e ealed ha he e is ano he g oup o complex ele-
men s. Accession AF497474 (Aegilops auschii, [3]) con-
ains an Angela-like (Copia) complex in posi ion 11808–
29240 ( e e se o ien a ion, nes ed wi h a Sab ina Gypsy in
o wa d o ien a ion). Accession AY368673 (B genome o
T i icum u gidum, [4]) also includes an Angela-like com-
plex (posi ion 218046–233487, in e e se o ien a ion;
Figu e 3). These wo complex inse ions ha bo ea u es
ha a e no consis en wi h an o igin h ough ecombina-
ion. Fi s , hey ha e lanking 5 bp di ec epea s (TATAA
and GCCGG, espec i ely), a leng h cha ac e is ic o
TSDs o Copia elemen s. In addi ion, hei wo ex e nal
LTRs o he se o h ee a e highly homologous (Figu e 4,
do -plo analysis om he AF497474 sequence; he
sequence alignmen s o he LTRs a e p o ided in he addi-
ional ile 1). The p esence o lanking TSDs suppo s hei
o igin om a single in eg a ion e en . The high simila i y
be ween he ou side LTRs is consis en wi h hei o igin,
u he mo e, om a single cycle o e e se ansc ip ion.
Based on hese obse a ions, we checked wi hin o he
Poaceae sequences o he occu ence o such complex
s uc u es. We ca ied ou an ab ini io scan o he ice
pseudomolecules and all he a ailable genomic sequences
om maize, using he LTR_STRUC so wa e [9] o de ec-
ion o comple e LTR e o ansposons. This so wa e
de ec s only comple e elemen s, based on he p esence o
bo h wo LTRs and he TSD mo i s lanking hem. Ou o
4704 iden i ied po en ial LTR e o ansposons, we we e
Fo ma ion o a complex s uc u e by DNA ecombina ionFigu e 2
Fo ma ion o a complex s uc u e by DNA ecombi-
na ion. A he e ologous ecombina ion be ween LTRs om
wo closely inse ed elemen s occu s, elimina ing he in e nal
sequence. Legends a e he same as in Figu e 1 ( om [2]).
Pa e n o inse ions gene ally encoun e ed in he long sequences analyzed om he T i iceae and closely ela ed speciesFigu e 1
Pa e n o inse ions gene ally encoun e ed in he
long sequences analyzed om he T i iceae and
closely ela ed species. Thin black lines ep esen he hos
DNA, hick colo ed lines elemen s' in e nal sequences, ec-
angles he LTRs, small boxes wi hin he ec angle he R
egion, and he iangles he TSD ( a ge -si e duplica ion).
Rec angles and lines om he same colo s de i e om he
same elemen . Dashed ea u es may o may no be p esen .
No mal inse ion, wi h TSD
Solo-LTR, w/ o w/o TSD
Nes ed inse ion in LTR, an i-sense
Nes ed inse ion in LTR, sense
Nes ed inse ion in co e, an i-sense
Nes ed inse ion in co e, sense
Complex inse ion, no TSD ?
Nes ed Solo-LTR in co e, sense
Nes ed Solo-LTR in co e, sense
Nes ed Solo-LTR in LTR, an i-sense
Nes ed Solo-LTR in LTR, sense
Nes ed inse ion in Solo-LTR, sense
Nes ed inse ion in Solo-LTR, an i-sense
BMC Genomics 2007, 8:247 h p://www.biomedcen al.com/1471-2164/8/247
Page 3 o 5
(page numbe no o ci a ion pu poses)
able o clea ly iden i y 2 new complex s uc u es ha bo -
ing he diagnos ic ea u es: an in e nal LTR, 2 comple e
co e sequences, lanking TSDs and simila i y be ween he
ou e mos LTRs. The i s elemen is loca ed on ch omo-
some 5 o ice, in posi ion 14011139–14022766 (TIGR
pseudomolecule), in he o wa d o ien a ion. This ele-
men is a membe o he Squiq sub amily, wi h CAAAC as
he TSD sequence. The second de ec ed complex is a
membe o he Opie amily in he maize BAC AY078063
[10], in posi ion 57992–74088, e e se o ien a ion, wi h
GCATG as he TSDs ( he de ailed alignmen s o LTRs as
he do e images o hose complexes a e p o ided in
addi ional ile 2).
A model ha can explain complex inse ions such as hese
in ol es abno mal empla e swi ching as a pa o e e se
ansc ip ion. Immedia ely p eceding e e se ansc ip-
ion, he RNA ma ix o ms a loop, using he high homol-
ogy be ween he wo R egions (5' and 3') wi hin he LTRs
o buckle he wo ends o a single empla e oge he . This
allows he (-)-s and cDNA, which o he wise canno p o-
ceed once i eaches he 5' end o he RNA empla e, o
jump ac oss o he 3' end and con inue. The jump is called
empla e swi ching. The p ocess leads o pe ec iden i y
be ween he 5' and he 3' LTR o he newly syn hesized
elemen , because he R and U3 segmen s o he 5' LTR and
he U5 segmen o 3' LTR in he RNA a e copied in o bo h
LTRs o he cDNA. The cDNA is ul ima ely inse ed in o a
new genomic loca ion by he in eg ase. The enzyme
liga es he cDNA o one s and o he asymme ic double-
s and b eak in he hos DNA, which is o med concomi-
an ly wi h he liga ion. The epai ing o his b eak leads
o he TSD (Figu e 5A; e iewed in [1]).
(A) No mal in a-s and pai ing o e e se ansc ip ion (middle) and inse ion (bo om) o an LTR e o ansposonFigu e 5
(A) No mal in a-s and pai ing o e e se ansc ip ion
(middle) and inse ion (bo om) o an LTR e o ansposon.
(B) P oposed o ma ion o a empla e-swi ching complex.
In e -s and pai ing (middle) occu s be ween wo di e en
RNAs, and he esul ing inse ion (bo om) ha bo s TSDs as
well as homologies be ween he wo ex e nal LTRs.
AB
A emis [16] iew o Angela complex on he AY368673 sequence ( om [7])Figu e 3
A emis [16] iew o Angela complex on he AY368673 sequence ( om [7]). The LTRs a e shown in da k blue, he
pu a i e polyp o eins in g een, and he whole elemen sequence in ligh blue. The a ows indica e he di ec ion o he inse -
ion.
Do e analysis o he h ee LTRs om he Angela complex on he AF497474 whea sequenceFigu e 4
Do e analysis o he h ee LTRs om he Angela
complex on he AF497474 whea sequence. Diagonal
lines ep esen he simila i ies be ween he sequences. The
longe and mo e solid he line is, he s onge he simila i y
is. The LTRs a e labeled acco ding o hei posi ion in he
elemen .
LTR-3
LTR-1
LTR-2
LTR-3 LTR-1
LTR-2
BMC Genomics 2007, 8:247 h p://www.biomedcen al.com/1471-2164/8/247
Page 4 o 5
(page numbe no o ci a ion pu poses)
E o s in empla e choice du ing he e e se ansc ip ion
can occu anywhe e along he sequence. The g owing
cDNA can jump o he o he packaged empla e ins ead o
o he o he end o he empla e i is al eady on. Gene ally,
because he wo packaged empla es a e almos iden ical
(de i ed om he same e o ansposon o e o i us
RNA), he phenomenon is unde ec able because he e a e
no majo modi ica ions o he esul ing cDNA. Howe e ,
i wo di e en RNAs a e packaged in he same i us-like
pa icle, a jump o he o he empla e du ing e e se an-
sc ip ion leads o abno mal o new elemen s, opening a
new mode o LTR e o ansposon e olu ion. The VejuL
[11] and BARE2 [8] elemen s appea o ha e been o med
in his way.
I RNAs om wo sligh ly di e en indi idual LTR e o-
ansposons a e co-packaged, he s and swi ch could
occu also be ween he wo R egions. This would lead o
o ma ion o a he e odime (Figu e 5B) a he han a no -
mal monome (Figu e 5A). The esul ing cDNA would
cons i u e a chime ic complex be ween he wo elemen s,
and possess chime ic LTRs. The p ocess o e e se an-
sc ip ion desc ibed abo e ende s he ex e nal LTRs iden-
ical. Thei 3' ends would be he e o e also iden ical and
could se e as subs a es o he same ype o in eg ase.
Thus, a chime ic complex elemen ne e heless would be
in eg a ed ia s anda d in eg ase ca alysis, leading o a
new genomic inse ion ha bo ing TSDs on ei he side
(Figu e 5B). The dime iza ion could occu be ween he
wo packaged RNAs om highly simila elemen s, such as
closely ela ed membe s o he same e o ansposon am-
ily, leading o a complex ha bo ing h ee iden ical LTRs
in e spe sed be ween wo simila in e nal egions. Mo eo-
e , because he LTRs would be comple e and no comp o-
mised by he e oduplex o ma ion, each o hem would be
able o p omo e he exp ession o i s co esponding
downs eam elemen . Thus, he wo o iginal elemen s
could be exp essed as no mal and indi idual copies and
e en p opaga e h ough he genome as sepa a e elemen s.
Conclusion
Only one empla e-swi ching complex could be iden i ied
on he 350 Mb sequence o he ice genome, one on he
a ailable maize sequences, and none in he A abidopsis
genome. On he ~7 Mb o sequences cu en ly a ailable
o he T i iceae (whea , ba ley, and ela ed species), wo
empla e-swi ching complexes we e iden i ied ou o he
20 ecombinan complex elemen s ecognized (Table 1).
Al hough such chime ic complexes, o med by e e se
ansc ip ion, o m a ela i ely mino sha e o he genome
when compa ed o hose o med by pos -inse ional
ecombina ion, hey appea none heless o be mo e abun-
dan in he T i iceae genomes han elsewhe e. The
genome o diploid ba ley is oughly 5 × 109 bp and ha o
hexaploid b ead whea abou 16 × 109 bp. I he obse ed
equency o wo o hese complexes in he a ailable
sequences holds h oughou he ba ley and whea
genomes, he wo ce eals should ha bo ca. 6000 com-
plexes o med by e e se ansc ip ion. Fo ma ion o hese
complexes is ano he mani es a ion, oge he wi h low
eplica ion ideli y and ansduc ion o genomic
sequences, o he luid and lexible na u e o e o anspo-
si ion. Fu he mo e, he complex elemen s epo ed he e
may poin o mechanis ic di e ences be ween plan spe-
cies, in iew o he di e ences in hei abundance be ween
he species we we e able o examine.
The model we p opose is consis en bo h wi h he a aila-
ble da a and wi h he es ablished de ails o he e o ans-
poson li e cycle. A di ec demons a ion o he mechanism
would en ail isola ion o i us-like pa icles con aining
wo pai ed RNAs (Figu e 5B) and demons a ing he RNA
s uc u e. This, howe e , awai s bo h an e icien sys em
o p oduc ion o packaged complexes (pe haps by o e -
exp ession o a e o ansposon wi h a endency o o m
complexes) and a means o dis inguishing he numbe o
mRNAs p esen wi hin he buckle.
Me hods
All cu en ly public a ailable T i iceae (whea and ba ley)
BACs we e e-analyzed as in [7]. The upda ed anno a ions
we e used o analyze he inse ion complexes. The o iginal
analyses o AF497474 om Aegilops auschii, AF368673
om T i icum u gidum and AY078063 om Zea mays we e
pe o med espec i ely by [3,4], and [10]. The sequences
o he ice pseudomolecules (~367 Mb) we e downloaded
om he TIGR websi e [12]. The scanned maize sequences
ep esen he whole la ge sequences a ailable o maize in
he public da abase, i.e., excluding he ace iles and he
gene-only sequences. They we e downloaded om he
NCBI websi e [13] and ep esen ~1 650 Mb.
Table 1: Numbe o inse ions in ~7 Mb o T i iceae la ge-inse
sequences
Type o e en Numbe o e en s
LTR e o ansposon inse ions 400
Copia inse ions 137
Gypsy inse ions 245
LARD inse ions 9
TRIM inse ions 9
Solo-LTR o ma ions 70
O he ecombina ion e en s 220
Recombinan Complexes 20
Templa e Swi ching Complexes 2
LINE inse ions 61
DNA T ansposon inse ions 118
Da a in i alics we e p oduced om he e-anno a ed sequences o [7].
The ecombina ion e en s ep esen all he dele ions, inse ions, and
emodeling e en s de ec able in he elemen s.
Publish wi h BioMed Cen al and e e y
scien is can ead you wo k ee o cha ge
"BioMed Cen al will be he mos signi ican de elopmen o
dissemina ing he esul s o biomedical esea ch in ou li e ime."
Si Paul Nu se, Cance Resea ch UK
You esea ch pape s will be:
a ailable ee o cha ge o he en i e biomedical communi y
pee e iewed and published immedia ely upon accep ance
ci ed in PubMed and a chi ed on PubMed Cen al
you s — you keep he copy igh
Submi you manusc ip he e:
h p://www.biomedcen al.com/in o/publishing_ad .asp
BioMedcen al
BMC Genomics 2007, 8:247 h p://www.biomedcen al.com/1471-2164/8/247
Page 5 o 5
(page numbe no o ci a ion pu poses)
The ab ini io iden i ica ion o LTR e o ansposons wi hin
he ice and maize sequences was pe o med by he
LTR_STRUC so wa e [9] using s anda d speci ica ions.
All o he 4072 po en ial complex elemen s ou pu by his
p og am we e i s sc eened by a home-made Py hon sc ip
acco ding o hei size, and he 1416 candida es mee ing
he c i e ion o >10 kb leng h we e hen manually checked
using Do e [12] o he p esence o he in e nal LTR. The
LTR s. LTR analyses we e pe o med using Do e [12],
and he a ge -si e duplica ion we e manually e i ied.
The LTR alignmen s we e e i ied using Clus alX [15], a e
manual edi ing as necessa y (see Supplemen al da a).
Abb e ia ions
LTR, long e minal epea
TSD, a ge -si e duplica ion.
Au ho s' con ibu ions
FS designed and pe o med he esea ch and w o e he
manusc ip . AHS di ec ed he esea ch as well as edi ed
and con ibu ed o he manusc ip . Bo h au ho s ead and
app o ed he inal manusc ip .
Addi ional ma e ial
Acknowledgemen s
The au ho s hank Jaakko Tanskanen o his help wi h he Py hon sc ip s.
FS was suppo ed by a ellowship om CIMO and by a Uni e si y o Helsinki
Pos doc o al Fellowship. Expe imen s desc ibed he e we e ca ied ou
unde a g an om Academy o Finland, P ojec 106949.
Re e ences
1. Sabo F, Schulman AH: Pa asi ism and he e o ansposon li e
cycle in plan s: A hi chhike 's guide o he genome. He edi y
2006, 97:381-388.
2. De os KM, B own JKM, Benne zen JL: Genome size educ ion
h ough illegi ima e ecombina ion coun e ac s genome
expansion in A abidopsis. Genome Res 2002, 12:1075-1079.
3. Ande son OD, Rausch C, Moulle O, Lagudah ES: The whea D-
genome HMW-glu enin locus: BAC sequencing, gene dis i-
bu ion, and e o ansposon clus e s. Func In Genomics 2003,
3:56-68.
4. Kong X-Y, Gu YQ, You FM, Dubco sky J, Ande son OD: Dynamics
o he e olu ion o o hologous and pa alogous po ions o a
complex locus egion in wo genomes o allopolyploid whea .
Plan Mol Biol 2004, 54:55-69.
5. Chan e N, Cenci A, Sabo F, Ande son OD, Dubco sky J:
Sequencing o he T i icum monococcum Ha dness locus
e eals good mic ocolinea i y wi h ice. Mol Gene Genomics
2004, 271:377-386.
6. Chan e N, Salse J, Sabo F, Rahman S, Bellec A, Laubin B, Dubois I,
Dossa C, Sou dille P, Joud ie P, Gau ie MF, Ca olico L, Becke M,
Aubou g S, Weissenbach J, Caboche M, Be na d M, Le oy P, Chal-
houb B: Molecula basis o e olu iona y e en s ha shaped
he Ha dness, Ha locus in diploid and polyploidy whea spe-
cies, T i icum and Aegilops. The Plan Cell 2005, 17:1033-1045.
7. Sabo F, Guyo R, Wicke T, Chan e N, Salse J, Laubin B, Le oy P,
Sou dille P, Chalhoub B, Be na d M: Upda ing ansposable ele-
men anno a ions om la ge whea genomic sequences
e eals di e se ac i i ies and gene associa ion o elemen s.
Mol Gene Genomics 2005, 274:119-132.
8. Vicien CM, Kalenda R, Schulman AH: Va iabili y, ecombina ion
and mosaic e olu ion o he ba ley BARE-1 e o ansposon.
J Mol E ol 2005, 61:275-291.
9. McCa hy EM, McDonald JF: LTR_STRUC: a no el sea ch and
iden i ica ion p og am o LTR e o ansposons. Bioin o ma -
ics 2003, 19:362-367.
10. S am M, Belele C, Ramak ishna W, Do weile JE, Benne zen JL, Chan-
dle VL: The egula o y egions equi ed o B' pa amu a ion
and exp ession a e loca ed a ups eam o he maize b1
ansc ibed sequences. Gene ics 2002, 162:917-930.
11. Sabo F, Sou dille P, Be na d M: Ad en o a new e o anspo-
son s uc u e: he long o m o he Veju elemen s. Gene ica
2005, 125:325-335.
12. TIGR homepage [h p://www. ig .o g]
13. NCBI homepage [h p://www.ncbi.nlm.nih.go ]
14. Sonnhamme EL, Du bin R: A do -ma ix p og am wi h dynamic
h eshold con ol sui ed o genomic DNA and p o ein
sequence analysis. Gene 1995, 167:GC1-10.
15. Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: imp o ing he
sensi i i y o p og essi e mul iple sequence alignmen
h ough sequence weigh ing, posi ion-speci ic gap penal ies
and weigh ma ix choice. Nucleic Acids Res 1997, 22:4673-4680.
16. A emis a Sange Ins i u e [h p://www.sange .ac.uk/So wa e/
A emis/]
Addi ional ile 1
Clus alX alignmen s o he LTRs om he wo Angela complexes om
T i iceae.
Click he e o ile
[h p://www.biomedcen al.com/con en /supplemen a y/1471-
2164-8-247-S1.pd ]
Addi ional ile 2
Do e and Clus alX alignmen s o he LTRs om he Rice and Maize
complexes.
Click he e o ile
[h p://www.biomedcen al.com/con en /supplemen a y/1471-
2164-8-247-S2.pd ]