Evolutionary dynamics of an ancient retrotransposon family provides insights into evolution of genome size in the genus Oryza.

Abstract

Long terminal repeat (LTR) retrotransposons constitute a significant portion of most eukaryote genomes and can dramatically change genome size and organization. Although LTR retrotransposon content variation is well documented, the dynamics of genomic flux caused by their activity are poorly understood on an evolutionary time scale. This is primarily because of the lack of an experimental system composed of closely related species whose divergence times are within the limits of the ability to detect ancestrally related retrotransposons. The genus Oryza, with 24 species, ten genome types, different ploidy levels and over threefold genome size variation, constitutes an ideal experimental system to explore genus-level transposon dynamics. Here we present data on the discovery and characterization of an LTR retrotransposon family named RWG in the genus Oryza. Comparative analysis of transposon content (approximately 20 to 27,000 copies) and transpositional history of this family across the genus revealed a broad spectrum of independent and lineage-specific changes that have implications for the evolution of genome size and organization. In particular, we provide evidence that the basal GG genome of Oryza (O. granulata) has expanded by nearly 25% by a burst of the RWG lineage Gran3 subsequent to speciation. Finally we describe the recent evolutionary origin of Dasheng, a large retrotransposon derivative of the RWG family, specifically found in the A, B and C genome lineages of Oryza.

Full text

Page 1

Evolutionary dynamics of an ancient retrotransposon
family provides insights into evolution of genome size
in the genus Oryza
Jetty S. S. Ammiraju1,†, Andrea Zuccolo1,†, Yeisoo Yu1, Xiang Song1, Benoit Piegu2, Frederic Chevalier2, Jason G. Walling3,‡,
Jianxin Ma3, Jayson Talag1, Darshan S. Brar4, Phillip J. SanMiguel5, Ning Jiang6, Scott A. Jackson3, Olivier Panaud2and
Rod A. Wing1,*
1Arizona Genomics Institute, Department of Plant Sciences, BIO5 Institute, University of Arizona, Tucson, AZ 85721, USA,
2Laboratoire Ge ´nome et De ´veloppement des Plantes, Universite ´ de Perpignan, Perpignan, France,
3Department of Agronomy, Purdue University, West Lafayette, IN 47907-2054, USA,
4Department of Plant Breeding and Genetics, International Rice Research Institute, Los Ban ˜os, The Philippines,
5Genomics Core Facility, Purdue University, West Lafayette, IN 47907-2054, USA
6Plant and Soil Sciences, Michigan State University, East Lansing, MI 48823, USA
Received 30 March 2007; revised 11 June 2007; accepted 21 June 2007.
*For correspondence (fax +1 520 621 1259; e-mail rwing@ag.arizona.edu).
†These authors contributed equally to this work.
‡Present address: Department of Horticulture, University of Wisconsin, Madison, WI, USA.
Summary
Long terminal repeat (LTR) retrotransposons constitute a significant portion of most eukaryote genomes
and can dramatically change genome size and organization. Although LTR retrotransposon content
variation is well documented, the dynamics of genomic flux caused by their activity are poorly understood
on an evolutionary time scale. This is primarily because of the lack of an experimental system composed of
closely related species whose divergence times are within the limits of the ability to detect ancestrally
related retrotransposons. The genus Oryza, with 24 species, ten genome types, different ploidy levels and
over threefold genome size variation, constitutes an ideal experimental system to explore genus-level
transposon dynamics. Here we present data on the discovery and characterization of an LTR retrotrans-
poson family named RWG in the genus Oryza. Comparative analysis of transposon content (approximately
20 to 27 000 copies) and transpositional history of this family across the genus revealed a broad spectrum
of independent and lineage-specific changes that have implications for the evolution of genome size and
organization. In particular, we provide evidence that the basal GG genome of Oryza (O. granulata) has
expanded by nearly 25% by a burst of the RWG lineage Gran3 subsequent to speciation. Finally we describe
the recent evolutionary origin of Dasheng, a large retrotransposon derivative of the RWG family, specifically
found in the A, B and C genome lineages of Oryza.
Keywords: Oryza, long terminal repeat (LTR) retrotransposon, RWG family, genome size, large retrotrans-
poson derivative.
Introduction
Plant genomes are extraordinarily dynamic. One promi-
nent force behind this is the flux of transposable elements,
especially long terminal repeat (LTR) retrotransposons,
that can result in dramatic increases or decreases in gen-
ome size (Bennetzen et al., 2005; Devos et al., 2002;
Feschotte et al., 2002; Hawkins et al., 2006; Ma and
Bennetzen, 2004; Ma et al., 2004; Meyers et al., 2001;
Petrov and Wendel, 2006; Piegu et al., 2006; SanMiguel
et al., 1996, 1998; Schulman and Kalendar, 2005; Shirasu
et al., 2000; Uozu et al., 1997; Vicient et al., 1999; Vitte and
Panaud, 2005; Wendel and Wessler, 2000; Wendel et al.,
2002; Zhang and Wessler, 2004). However, little is known
about how LTR retrotransposons evolve and reshape
genomes within a family structure. For example, previous
342
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd
The Plant Journal (2007) 52, 342–351doi: 10.1111/j.1365-313X.2007.03242.x

Page 2

comparative studies of orthologous regions between
maize, sorghum and wheat were unable to identify con-
served LTR retrotransposons that pre-date their divergence
(Ramakrishna et al., 2002; SanMiguel et al., 2002; Tikhonov
et al., 1999). This is mainly because of long divergence
times between these species relative to the half-life of LTR
retrotransposons, which is estimated to be <6 million years
(MY) (Ma and Bennetzen, 2004), thus confounding the
study of patterns of ancestral LTR retrotransposon evolu-
tion and their impact on genome biology.
The Oryza genus represents a powerful model system to
test and validate the mechanisms and timing of genome size
evolution within a genus (Piegu et al., 2006; Wing et al.,
2005), having ten distinct genome types (six diploid and four
polyploid), a threefold genome size variation (Ammiraju
et al., 2006), a robust phylogenetic tree (Ge et al., 1999), a
comprehensive collection of BAC libraries (Ammiraju et al.,
2006), BAC end sequences (BES) and physical maps (http://
www.omap.org; Kim et al., unpublished results), and a
reference genome sequence for the cultivated species
O. sativa (International Rice Genome Sequencing Project,
2005; Yu et al., 2005). Previously, using this system, we
demonstrated that there is a positive correlation between
LTR retrotransposon content and genome size in Oryza
(Ammiraju et al., 2006), and that the O. australiensis EE
genomedoubled insize by three LTR retrotransposonbursts
(RIRE1, Wallabi and Kangaroo) approximately 3 million
years ago (MYA) (Piegu et al., 2006).
Here we present the discovery and characterization of a
second case of genomic obesity in Oryza caused by a Ty-3-
gypsy retrotransposon family named Gran3 that populates
nearly one quarter of the O. granulata [GG] genome. Most
compellingly, the 22% increase in genome size of this most
ancestral lineage of the genus (Ge et al., 1999) occurred
subsequent to its speciation. These results are significant in
light of previous predictions of unidirectional genome size
reduction in the genus Oryza, resulting from super-imposi-
tion of genome sizes onto known phylogenies (Kellogg,
1998). Gran3 is closely related to two other LTR retrotrans-
posonfamiliespreviouslycharacterizedinO.sativa[AA]and
O. australiensis [EE], i.e. RIRE2 and Wallabi, respectively.
Comparative analysis of the transpositional history of these
three elements (Gran3, RIRE2 and Wallabi) across the genus
Oryza allowed, for the first time, the study of evolutionary
dynamics of the same LTR retrotransposon family across a
plant genus covering a time span of 25 MY (Piegu et al.,
2006). This analysis revealed a broad spectrum of indepen-
dent and lineage-specific changes that have implications on
the evolution of genome size in Oryza as a whole. Finally we
analyzed theevolutionary dynamics of
Dasheng, a large retrotransposon derivative (LARD) (Kalen-
dar et al., 2004) of the retrotransposon RIRE2 family, thus
providing data and insights on the evolution of a non-
autonomous LTR retrotransposon.
theelement
Results and discussion
Isolation and characterization of the Gran3 LTR retrotrans-
poson family from the O. granulata [GG] genome
As part of a large-scale comparative sequencing project
within the genus Oryza (http://www.omap.org), a BAC clone
(149.7 kb) containing the Adh1–Adh2 region from the
O. granulata [GG] genome was isolated, sequenced and
annotated. O. granulata is the most basal lineage of the
genus Oryza (Ge et al., 1999), and possess the second larg-
est genome(882Mbp) among thediploid species(Ammiraju
et al., 2006). Four Ty-3-gypsy retrotransposons and a relic in
the form of a solitary LTR were identified from this single
BAC and found to belong to the same family, henceforth
named Gran3. They displayed typical LTR retrotransposon
consensus 5¢-TG…CA-3¢ sequences at the start and end of
the LTRs, and intact target site duplications. Two of the
elements, Gran1 and Gran2, contained deletions in their
internal coding domains relative to the structurally complete
element Gran3 (Supplementary Figure S1). These two ele-
ments are therefore considered defective, and were either
trans-mobilized by an autonomous Gran3 element or have
degraded or been rearranged after transposition. Another
member of the same family, Gran4, was considerably rear-
ranged (Figure 1a). Gran1 and Gran2 are 5003 and 4620 bp
long, respectively, whereas Gran3 element is 12 098 bp after
removal of two unrelated solo LTRs located within this ele-
ment (Figure 1a). In addition to sharing very similar and
short LTR lengths of 425–431 bp, Gran1–3 possessed highly
conserved primer binding sites (PBS) and polypurine tract
(PPT) sites adjacent to their 5¢ and 3¢ LTRs (Figure 1b).
Nearly one quarter of the O. granulata genome
is composed of the Gran3 family
The presence of four members on a single BAC strongly
suggested that Gran3 family is highly abundant in the
O. granulata genome. To evaluate this hypothesis, the copy
number of Gran3 family in the O. granulata genome was
estimated using three approaches. We first calculated the
copy number mathematically [LTR copy number = matches
in BACends· 882 Mb(O.granulatagenomesize)/thesizeof
the BAC end database (93 Mb)] (Jiang et al. (2002a) and
obtained an estimate of 59 556 LTRs. Secondly, probes
specific to the LTR domain and an internal domain common
to both complete and defective Gran3 elements (Figure 1b)
were hybridized to dot blots containing stochiometrically
estimated concentrations of the O. granulata genome. This
analysis identified nearly 25 700 Gran3 copies with both
LTRs (based on the internal domain) and 8600 ? 2000
apparent single LTRs (Table 1). Lastly, we validated the copy
number estimates by screening the O. granulata BAC library
(Ammiraju et al., 2006), and showed that about 60–80% of
Evolutionary dynamics of ancient retrotransposon family 343
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 3

the tenfold redundancy BAC library contained sequences
homologous to the Gran3 family (data not shown).
These results indicated that the Gran3 LTR retrotranspo-
son family is highly abundant in the O. granulata genome,
but could not distinguish the relative proportion of intact
versus defective elements. Evidence was gathered from two
different analyses to understand such dynamics. Firstly, an
additional probe specific to the reverse transcriptase (RT)
domain of Gran3 element that distinguished the defective
forms from complete forms was designed (Figure 1b) and
hybridized to dot blots. Copy number estimates indicated
the presence of approximately 8900 ? 1000 complete Gran3
elements. Subtraction of this number from the entire
number of Gran3 family members indicated that nearly
16 800 ? 3000 defective forms with internal deletions exist
in the GG genome, accounting for up to 22% of the genome
size (Table 1). Secondly, an in silico survey of the BES
database of O. granulata was also conducted using the RT
domain of Gran3, and the results were compared to those
obtained using LTR sequences. The results indicated a
striking imbalance of RT to LTR domains, with a ratio 0.08,
which is far from that expected (0.5) if all the Gran3 elements
are complete (Supplementary Table S1). These results point
to an overwhelming presence of incomplete/non-autono-
mous elements, and are consistent with the results obtained
from hybridization approaches and the annotated O. granu-
lata BAC. Similar results have been found for other LTR
retrotransposon families, such as Cinful-1 and Zeon in maize
(Meyers et al., 2001). Further, a considerable portion of the
total number of LTR sequences appears to be solo LTRs
Figure 1. Organization and structural features of
Gran3 retrotransposon family in O. granulata
BAC.
(a) Distribution of Gran3 elements in the Oryza
granulata BAC. Arrows indicate the 5¢ fi 3¢ ori-
entation. The numbers indicate the position of
theelementsintheBACOG-Aba077F15(149.7 kb
long).
(b) Detailed structures of the Gran3 elements
isolated. TSD, target site duplication; PBS, pri-
mer binding site; PPT,polypurinetract.Proposed
PBS and PPT sequences are shown and their
shared nucleotide positions are underlined;
sequences belonging to LTRs are shown in red.
The light-green boxes in Gran3 represent two
solo LTRs insertions flanked by TSD.
(c) Isolated full-length RWG family members
from various Oryza lineages. The arrows indicate
tandem repeats.
Table 1 Copy numbers and relative contribution of the Gran3 family
in the Oryza granulata genome
Gran3 elements
Number
of copies
Size
(bp)
Size in
genome
(Mb)
Genome
fraction
(%)
Intact elements
Defective forms
Apparent solo LTRs
8900 ? 1000 12 000
16 800 ? 3000
8600 ? 2000
107 ? 12
84 ? 15
3.6 ? 0.9
194.6 ? 24.9 22.0 ? 3.2
12.1 ? 1.4
9.5 ? 1.7
0.4 ? 0.1
5000
430
Total
344 Jetty S. S. Ammiraju et al.
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 4

(25%). However, the contribution of these solo LTRs is only
0.4 ? 0.1% of the genome (Table 1).
Genomic organization of the Gran3 family in the
O. granulata genome
The copy number estimates above indicate that the Gran3
retrotransposon family is highly abundant in the GG gen-
ome. To better understand the genomic organization of the
Gran3 family, we hybridized a fluorescently labeled Gran1
probe (see Experimental procedures) to O. granulata
metaphase chromosomes. As shown in Supplementary
Figure S4, a significant clustering of Gran3 in the hetero-
chromatic peri-centromeric regions is particularly striking
for most chromosomes. Additionally, we also mapped
Gran3 family members using BES to a phase I BAC-based
physical map of the O. granulata genome. The results are
presented in Supplementary Table S2 and Supplementary
Figure S5, and are congruent with the results of fluorescent
in situ hybridization. The abundance of Gran3 appeared to
be non-uniform, with the highest accumulation frequency
(portion/presence; LTRs/megabase) found on chromo-
some 8 and the lowest on chromosome 10 in the peri-
centromeric regions. Ratios of accumulation frequencies in
the peri-centromeric regions versus non-pericentromeric
regions were derivedfor each chromosome(Supplementary
Table S2). A ratio above 1 is expected in the case of peri-
centromeric clustering. The results indicated that there was
significant peri-centromeric clustering on all chromosomes
except chromosome 10. Interestingly, the frequency of
Gran3 accumulation (LTR/Mb) is approximately five times
higher in peri-centromeric regions than in chromosome
arms.
Our data indicate that the Gran3 family is a major
component of peri-centromeric regions in the O. granulata
genome. This type of organization is similar to that previ-
ously found for the LTR retrotransposons Dasheng and
RIRE2 in O. sativa (Jiang et al., 2002a,b).
Genomic paleontology of the Gran3 family in O. granulata
A major retrotranspositional burst underlies Gran3 abun-
dance in O. granulata. We estimated the timing of inser-
tional activity of Gran3 using an approach described by
Piegu et al. (2006). Although, such global simulations may
tend to over-estimate the timing, they should nevertheless
provide insights into the approximate time frame for such
activity.Inbrief,wefirstdeterminedthepopulationstructure
of the Gran3 retrotransposon family within the O. granulata
genome using a phylogenetic approach. The resulting
neighbor-joining tree, generated using all 150 paralogous
RT sequences (Supplementary Table S1) retrieved from the
BES database of O. granulata (methods) has three
major monophyletic clades, namely M1 (which includes 92
paralogous RT), M2 (which includes 22 paralagous RT) and
M3 (which includes 13 paralagous RT), plus a fourth group
(which includes 23 paralogous RT) that is not clearly
monophyletic and is very basal to the first three groups
(Figure 2).
Secondly, pairwise distances (Kimura, 1980) were calcu-
lated by comparing each paralog to all other paralogs
belonging to the same monophyletic group. Distances were
then translated into insertion dates as described by San-
Miguel et al. (1996) but using a mutation rate of 1.95 · 10)8
(Vitte et al., 2004) to be consistentwith our previousanalysis
(Piegu et al., 2006). This enabled the identification of three
retrotranspositional waves, mostly overlapping in terms of
time (Figure 3). From a quantitative perspective, a signifi-
cant part of the amplification seems to have occurred at
around 3–6.2 MYA. Further, a 4.4 MY insertion time was
estimated for complete and intact Gran3 element, which is
consistent with the dating of the proliferation waves.
The Gran3 LTR retrotransposon burst is more recent than
host species radiation. Two lines of evidence suggest that
the Gran3 burst in the GG genome occurred subsequent to
the species radiation. Firstly, the estimated time of diver-
gence between O. granulata and other Oryza species is
approximately 9–25 MY (Guo and Ge, 2005; Piegu et al.,
2006). On the other hand, the proliferation waves were dated
at <6.2 MY (Figure 3), suggesting that the genome size
increase in O. granulata is recent. Secondly, all Gran3 RT
paralogs from O. granulata form a distinct cluster relative to
the other Oryza species at high bootstrap values (Figure 4).
A strikingly similar scenariowasobserved in O. australiensis
[EE], echoing our previous observations (Piegu et al., 2006).
Combined, these results provide an intriguing picture of
recent and independent increases in genome sizes for the
GG and EE genome types of Oryza because of the
same ancestral retrotransposon family. As in the case of
O. australiensis, the increase in the genome size was sub-
sequent to its speciation. Further, phenetic analysis sug-
gests that the episodic burst dated at around the same
evolutionary time in O. australiensis (Piegu et al., 2006) was
independent of the bursts in O. granulata.
Indirect evidence for the ongoing activity of Gran3 was
detected in the O. granulata genome. Gran3 LTR sequences
identified from the BES database were compared to each
other in all pairwise combinations and then classified into
groups according to the extent of sequence divergence
(Supplementary Figure S3). Identical LTRs that could poten-
tially result from cloning bias, i.e derived from the same
restriction site cloned during the BAC library construction,
were removed by careful manual comparison. A very
conservative estimate indicated that more than 9% of the
LTRs (142 out of 1559 LTRs analyzed) have at least one
identical counterpart within the same group. As LTRs of an
intact element are identical at the time of insertion, these
Evolutionary dynamics of ancient retrotransposon family 345
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 5

observations may imply ongoing Gran3 activity or such
activity in the recent past. However, the overall majority of
LTRs showed only partial similarity with others, reflecting
the heterogeneity of the LTR set as expected in the case of
an ancient retrotranspositional process. Solo LTRs with
terminal site duplications (TSDs) (Supplementary Figure S3)
were present in most of these classes, suggesting that the
process leading to their formation has been active since the
Figure 2. Neighbor-joining phylogenetic tree of the Gran3 retrotransposon in
Oryza granulata.
Reverse transcriptase coding domain nucleotide sequences were used.
Bootstrap values are calculated for 1000 replicates, and are shown only if
greater than 50. Three monophyletic clades are indicated (M1, M2 and M3).
Figure 3. Timing of the Gran3 retrotranspositional burst in Oryza granulata.
Groups of paralogs were isolated on the basis of the monophyletic clades
indicated in Figure 2. Each paralog was compared to all others included in the
same clade; Kimura distances were then converted to million of years ago
using a substitution rate of 1.95 · 10)8. The y axis shows the percentage of
the total pairwise comparisons.
Figure 4. Neighbor-joining phylogenetic tree of the Gran3 retrotransposon
from 12 Oryza species.
Gran3 reverse transcriptase (RT) coding domain nucleotidic sequences were
used. Of all the available Gran3 RT (Supplementary Table S1), 20% were
randomly chosen for the analysis for each Oryza species. Bootstrap values
werecalculatedfor1000replicates.CirclesindicateO. australiensissequences
and triangles indicate O. granulata sequences.
346 Jetty S. S. Ammiraju et al.
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 6

very beginning of the Gran3 colonization of the O. granulata
genome.
RIRE2, Wallabi and Gran3 are all members of the
same LTR retrotransposon family
To determine whether the Gran3 family isolated from
O. granulata was related to other LTR retrotransposons in
Oryza, similarity searches were conducted. This analysis
revealed that Gran3 family members were similar to two
previously characterized LTR retrotransposons, RIRE2 from
O.sativa(Ohtsubo et al.,
O. australiensis [EE] (Piegu et al., 2006). Gran3 was 55%
similarovertheentirelengthofRIRE2,55.9%intheLTRs and
71% in the RT domain at the nucleotide level. In addition,
they shared remarkable conservation in sequence length
(11 365 bp for RIRE2 versus 12 098 bp for Gran3), and a PBS
sequence that is homologous to tRNAArg, thereby suggest-
ing that Gran3 is a distant member of the RIRE2 clade of LTR
retrotransposons.For Wallabi, Gran3was53.8%similarover
the entire length, 57.4% in the LTRs and 72.1% in the RT
regions.
Phylogenetic analysisof
(O. granulata) and Wallabi (O. australiensis) with those
belonging to the most abundant Ty-3-gypsy elements from
O. sativa resulted in a distinct grouping of these two
elements with that of RIRE2 at high bootstrap value
(Supplementary Figure S2).
structural conservation as well as phylogenetic grouping,
we propose that RIRE2, Wallabi and Gran3 belong to the
same LTR retrotransposon family, henceforth named RWG
(RIRE2, Wallabi, Gran3).
Several lines of evidence also suggested that the RWG
LTR retrotransposon family was ancient in the genus
Oryza. Using Southern blots, we previously demonstrated
the presence and differential abundance of the Wallabi
element (now a member of the RWG family) across all 10
genome types of Oryza (Piegu et al., 2006). These results
were further confirmed by the presence of RT homologs
(see Experimental procedures) of this family in BES
derived from all 10 known Oryza genome types (http://
www.omap.org) and by isolation of complete RWG family
elements from several sequenced BACs of O. punctata
[BB], O. minuta [BBCC] and O. ridleyi [HHJJ] (Figure 1c;
GenBankaccessionnumbers
addition, a detailed phylogenetic analysis of RT domains
derivedfrom theOryza BES
coherence between element phylogeny and organismal
phylogeny, suggesting primary colonization by the RWG
family in the ancestor of Oryza (Figures 4 and 5) and
subsequent diversification during the course of evolution.
The closest homologs for the RWG family outside Oryza
are the Athila and Grande-1 elements in Arabidopsis and
Zea mays, respectively (Ohtsubo et al., 1999).
1999)and Wallabifrom
RT sequencesof Gran3
Based onsequence and
EF193783–EF193785).In
databasesrevealeda
Discontinuous phylogenetic distribution of the RWG LTR
retrotransposon family in the genus Oryza
Dot-blotanalysiswasconductedusingprobesspecifictoLTR
and internal domains to estimate the relative abundance of
the RWG family in Oryza and its contribution to genome size
variation. Different steps were taken to minimize biases
inherent in the use of dot-blot hybridization approaches to
estimatecopynumbers.Firstly,wedesignedspecificprimers
from the RWG family sequences (LTR and RT) mined from
BES of each Oryza genome type (see Experimental proce-
dures), and polymerase chain reaction (PCR)-amplified
probeswereusedtohybridizedotblotsofthesamespeciesto
minimize biases from sequence divergence. Secondly,
hybridization conditions were calibrated by first estimating
RWGcopynumberfromO.sativagenomicDNAdotblotsand
then comparing the results to those obtained from in silico
analysis of the completed reference sequence (International
Rice Genome Sequencing Project, 2005). Thirdly, the results
obtained from hybridization analysis were largely supported
by copy number estimates from a set of an unbiased sam-
pling of random sheared shotgun sequences generated for
each Oryza species (Supplementary Table S3) and the num-
ber of significant hits from BES analysis (Supplementary
Table S1).TheresultsarepresentedinTable 2.Copynumber
variation of an unprecedented magnitude was found across
Oryza, from approximately 20 in O. brachyantha [FF] to
27 000inO.australiensis[EE](Pieguet al.,2006).Twodiploid
species, O. brachyantha [FF] and O. glaberrima [AA], and the
tetraploid O. coarctata [HHKK] contained the lowest number
of Gran3 members. The copy number in the case of O. coa-
rctata should be considered a rough estimate as the genome
size of this species is unknown (Ammiraju et al., 2006). Two
diploid genome types (GG and EE) contained the largest
number of RWG family members among the ten known
genome types (Table 2). Other genome types contained
moderately increased numbers of RWG family elements in
comparison to the International Rice Genome Sequencing
Figure 5. Evolutionary scenario of the RWG family and its impact on the
evolution of genome size. (Phylogenetic tree is from Ge et al., 1999.)
Evolutionary dynamics of ancient retrotransposon family 347
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 7

Project reference sequence, suggesting differential lineage-
specific proliferation and/or elimination or regulation. We
estimated the ratio of the number of LTRs versus the internal
region (LTR/in) for the RWG family for each Oryza genome
type. A ratio of 2 is expected in the case of an intact element,
and deviations from this could result from deletion or
recombination. The LTR/in ratios for these species suggest a
differential excess of LTRs in all genome types except
O. coarctata and O. brachyantha (Table 2). The contributions
of the RWG family varied widely across the genus, ranging
from 0.02% in O. coarctata to 25% in O. australiensis.
Recent evolution of a non-autonomous element of the
RWG family
Previous studies (Jiang et al., 2002a,b) showed that the
O. sativa [AA] genome harbors a high copy number of a
large non-autonomous retrotransposon derivative (LARD)
(Kalendar et al., 2004) called Dasheng. Its autonomous
counterpart is the LTR retrotransposon RIRE2, based on
sequence similarity of the LTRs of both elements, target site
preference and genomic distribution. Two alternative
hypotheses have been proposed to account for the origin of
LARDs (Jiang et al., 2002a). One is that they are the product
of transduction of a genomic sequence from the host gen-
ome, flanked by two solo LTRs. The other is that LARDs
originate in the virus-like particle by co-encapsulation of an
mRNA of the autonomous element with an mRNA of any
host gene, followed by strand exchange between the two
during the reverse transcription step. So far, there is
no experimental evidence to support either of these
hypotheses. In this study, we have provided evidence that
the RWG family is ancient in the genus Oryza. We thus
looked for the presence of Dasheng-related sequences in the
genus to tentatively trace its origin by probing Southern
blots containing genomic DNA from representatives of all
ten Oryza genometypesusingDasheng-specificprobes. The
results clearly indicate the presence of strong hybridization
signals in Oryza lineages belonging to the AA, BB and CC
genomes, and their absence from other distantly related
genomes (Supplementary Figure S6). The Southern blot
results were further confirmed by PCR using a primer in the
5¢ LTR of Dasheng and another in its internal region, for
which PCR amplification bands were only detected in the
AA, BB and CC genome species (data not shown). These
results indicate that retrotransposition of Dasheng has
occurred recently and specifically in the A/B/C genome
lineages, i.e. around 7 MYA (Figure 5) (Panaud et al.,
unpublished results).
Dynamics of the RWG family and its impact on the evolution
of genome size in Oryza
Here we have presented a comparative evolutionary
dynamics study of an ancestral LTR retrotransposon fam-
ily (RWG) in terms of its recent history, timing of trans-
position and differential fates of proliferation among
lineages of the genus Oryza. Recent activity of Dasheng
and RIRE2 has already been demonstrated in domesti-
cated rice (O. sativa [AA]) by estimating their age as well
by identifying transcripts in cDNA libraries (Jiang et al.,
2002a,b). Nearly 25% of all full-length Dasheng and RIRE2
elements isolated from O. sativa were amplified within the
last 150 000 years (Jiang et al., 2002a,b). A global estimate
Table 2 Copy number and genomic contribution of the RWG family across Oryza
Species Genome type
Genome size
(Mb)
Copy number Percentage genome share
LTR/in LTR Internala
By intact
elements
By apparent
solo LTR Total
O. nivara
O. rufipogon
O. glaberrima
O. punctata
O. officinalis
O. minuta
O. alta
O. australiensisb
O. brachyantha
O. granulatab
O. ridleyi
O. coarctata
AA
AA
AA
BB
CC
BBCC
CCDD
EE
FF
GG
HHJJ
HHKK
448
439
357
425
651
1124
1008
965
362
882
1283
1283
3240 ? 645
1620 ? 321
207 ? 35
2291 ? 469
5946 ? 573
2850 ? 533
1327 ? 200
66 000 ? 1000
< 20
60 000 ? 3000
1033 ? 257
< 20
512 ? 130
209 ? 37
30 ? 7
188 ? 32
762 ? 113
407 ? 113
759 ? 222
27 000 ? 3000
< 20
25 700 ? 2000
469 ? 155
< 20
1.37
0.57
0.10
0.53
1.40
0.43
0.90
24.9
0.03*
21.6
0.44
0.01*
0.2
0.1
0.17
0.2
0.3
0.1
0.0
0.6
0.0
0.4
0.004
0.0
1.6
0.7
0.3
0.7
1.7
0.5
0.9
25.5
0.03*
22.0
0.45
0.01*
4–10.2
5.3–11.3
4.6–10.8
8.3–17.7
6.1–10
4.5–11.5
1.1–2.8
2.2–2.8
NA
2.1–2.7
1.2–4.1
NA
aInternal values represent the copy numbers obtained from hybridization of the reverse transcriptase domain of RWG elements isolated from BAC
end sequences of each Oryza species, except in the case of O. granulata and O. australiensis.
bFor O. granulata, internal values represent copy numbers obtained using probes that are common for both complete and defective Gran3
elements (Table 1). O. australinesis values were obtained from Piegu et al. (2006).
NA, not available; asterisks indicate that the value is a rough estimate
348 Jetty S. S. Ammiraju et al.
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 8

of their insertion time derived by dating phenetically
structuredgroups ofDasheng-
sequences indicated Dasheng to be one of the youngest
elements (7.2 MY) and RIRE2 to be relatively ancestral
(17.2 MY) (Jiang et al., 2002b). However, the substitution
rate used for molecular clock estimation is at least three
times slower than that used in our analysis. Extrapolation
of the previous results of Jiang et al. (2002b) to the
mutation rate we used suggests that most RIRE2- and
Dasheng-like elements were inserted within the past 5–6
MY in the O. sativa genome. These results indicate that, as
in the case of the distant ancestors O. australiensis and
O. granulata, the O. sativa genome also experienced
recent amplification of the RWG family (Figure 5). The
magnitude of such amplification, however, is several
thousand-fold different among these lineages, leading to
massive increases in genome size for both the EE and GG
genomes. The contribution of this single LTR retrotrans-
poson family in these two genomes parallels the total
contribution of all classes of retrotransposons in the
O. sativa genome (20%) (International Rice Genome
Sequencing Project, 2005). These results suggest that a
large proportion of the more than twofold genome size
difference in the domesticated species versus the most
ancestral lineage is mainly accounted for by differential
amplification of the RWG family.
We also demonstrated that the RWG family propagates to
different levels, resulting in a number of lineages with
complex phylogenetic patterns. The findings raise more
questions regarding the biological causes for massive RWG
family proliferation in some lineages (i.e. O. granulata and
O. australiensis) and almost sudden death or attenuation in
others (i.e. O. brachyantha and O. coarctata) (Figure 5). Such
differential fates of the RWG family could mechanistically
resultfromlineage-specificdifferencesintheregulationand/
or suppression of transposition. Alternatively, they could
alsoresultfromdifferencesintheratesandefficienciesofthe
removal processes. One indirect line of evidence, a lack of
correlation between transposition and transcription of this
familyinO.sativa,O.australiensisandO.granulata,suggests
that differential amplification of the RWG family is mainly
regulatory, at transcriptional or post-transcriptional level.
Beyond this mechanistic information, surprisingly little is
known about the global evolutionary forces underlying
genome size variation. Several models for the evolution of
genome size have been proposed, ranging from the role of
adaptive changes to a particular ecological niche (Kalendar
et al., 2000) to selection at the organismal level (Kidwell and
Lisch, 2000). With large-scale sequences from several Oryza
species expected in the near future, the stage is set for
exploring such exciting avenues. The results presented here
bring clarity to the bi-directional evolution of genome size
and show that the RWG family has played a major role in
restructuring and evolution of genome size in Oryza.
andRIRE2-like LTR
Experimental procedures
Data mining and phylogenetic analysis
Similarity searches were carried out using the BLASTN algorithm
(Altschulet al.,1997);onlyhitswithanEvalue£ 1e-10andspanning
at least 80 bp of the Gran3 elements were taken into account. All
alignmentswere performedusing
(Thompson et al., 1994). Sequences homologous to the RT domain
of Gran3 elements were identified in the BES databases through
BLASTN searches (Altschul et al., 1997) run under relaxed settings
(-q -4 –r 5) in order to isolate the most diverged elements of the
family; only sequences spanning at least 90% of the length of the RT
domain were considered. The specificity and the origin of these
sequences were checked by comparison with RTs from the most
abundant LTR retrotransposons of O. sativa. Only those having the
best scores of similarity with Gran3 elements were retained and
used to build neighbor-joining trees using the program MEGA
version 3 (Kumar et al., 2004).
A similar approach was used to identify Gran3 LTRs from
O. granulata BES. Alignments of LTR pairs were performed using
the program
STRETCHER (EMBOSS package; Rice et al., 2000).
Kimura distances (Kimura, 1980) were calculated for each pair
using the DISTMAT program (EMBOSS package).
theprogram
CLUSTAL W
Hybridization analysis
Total genomic DNA from young leaf tissue of the Oryza
species O. brachyantha ([FF]; accession 101232), O. alta ([CCDD];
accession 105143), O. officinalis ([CC]; accession 100896), O. ridleyi
([HHJJ]; accession 100821), O. punctata ([BB]; accession 105690),
O. coarctata ([HHKK]; accession 104502), O. minuta ([BBCC];
accession101141), O. granulata
O. glaberrima variety CG14 ([AA]; accession 96717), O. rufipogon
([AA]; accession 105491), O. nivara ([AA]; accession 100897) and
O. australiensis ([EE]; accession 100882) and O. sativa ssp. japonica
(cv. Nipponbare; [AA]) was isolated as described by Doyle and
Doyle (1987). Large-scale genomic resources in the form of BAC
libraries, BES and physical maps (Ammiraju et al., 2006; Wing et al.,
2005; Kim et al., unpublished results) are available for these Oryza
accessions as part of the OMAP consortium (http://www.omap.org).
LTR and the internal domain (1817–2128 bp) of Gran2 (Genbank
accession number EF187427) and the RT domain (3993–4430 bp) of
Gran3 (Genbank accession number EF187429) were used as BLAST
probes to retrieve paralogs from BES databases of all Oryza species.
Retrieved paralogs were aligned, and specific PCR primers were
designed to amplify these domains for each species to reduce the
copy number biases inherent with heterologous probes (Supple-
mentary Table S4). All PCR primers were designed using PRIMER3
software (Rozen and Skaletsky, 2000). PCR reactions were carried
out in a 25 ll reaction volume, using 10 ng of genomic DNA, 0.2 mM
of each dNTP, and 10 pmol of each primer. The following thermal
cycling conditions were used; 94?C for 4 min, then 35 cycles of 94?C
for 30 sec, 65?C for 30 sec and 72?C for 90 sec, followed by 72?C
for 5 min.
Polymerase chain reaction products were purified from the
agarose gels. Genomic DNA from Oryza species as well as the
isolated PCR products was quantified using a Nano Drop?ND-1000
spectrophotometer(Nano Drop;
Dot-blot preparation, hybridization, washing conditions and copy
number estimations were essentially as described by Piegu et al.
(2006). At least three replications were carried out for each species.
High-density colony filters of the O. granulata library were screened
([GG]; accession102118),
http://www.nanodrop.com).
Evolutionary dynamics of ancient retrotransposon family 349
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 9

using LTR and internal domain probes. Hybridizations and imaging
were carried out as described by Ammiraju et al. (2006).
Two probes corresponding to the internal region (IR) of Dasheng
were generated. Primer pair 1 (5¢-TATTATGCCACCGTTGTTGG-3¢/
5¢-AACAACGACGGCATAATAGG-3¢) was designed to span the tan-
demely repeated sequence within the IR. Primer pair 2 (5¢-TGG-
GAAACTTCGTCATCATAGTC-3¢/5-GTGCTTAAGTCCCTAGAAGACA-
C-3¢) was designed to amplify a portion of the IR that did not contain
any repeat motif. PCR amplification with primer pair 1 resulted in a
ladder-like pattern on agarose gel. The PCR products were cloned,
and the inserts obtained from the recombinant clones were of
various sizes, as expected. The largest clone was sequenced to
confirm the presence of five tandem repeats, and was then used in
Southern hybridization experiments. PCR amplification with primer
pair 2 yielded a single band, which was cloned, sequenced and then
used as a hybridization probe.
Fluorescent in situ hybridization analysis
The Gran1elementwassub-cloned fromtheO.granulataBACclone
(OG_ABa077 F15) using a PCR strategy. PCR amplification was
performed in a final volume of 50 ll containing 0.4 lM of primers
(CTGCGACAGGCAGGGTAA and
CATTATCGT), 2.5 U of Klen Taq DNA polymerase (Sigma, http://
www.sigmaaldrich.com/),5 ll
0.12 5 mM of each dNTP, 1.5 mM MgSO4, 5 ll of PCRx enhancer
solution (Invitrogen, http://www.invitrogen.com/), 3 ll of DMSO
and 50 ng of BAC DNA. PCR amplification was performed in a
Tetrad thermocycler (Bio-Rad, http://www.bio-rad.com/) with 35
cycles of 96?C for 15 sec and 60?C for 15 min, and checked by 1%
agarose gel electrophoresis with 1 · TAE buffer. The gel-purified
insert was cloned into pGEM-T Easy vector (Promega, http://
www.promega.com/) according to the manufacturer’s recommen-
dation, and the recombinant clones were validated by NotI
restriction enzyme digestion and sequencing using T7 (TA
ATACGACTCACTATAGGG)and
AGAA) primers. Chromosomes of O. granulata were prepared
according to the method described by Walling et al. (2005). Probe
labeling, fluorescent in situ hybridization and image analysis were
carried out as described by Walling et al. (2006).
TGGGTTCTCTAGACCTA-
of 10 ·
PCR buffer (Sigma),
Sp6(ATTTAGGTGACACTAT
Physical mapping
Finger printed contigs (FPC) for the BAC clones that span Gran3
LTRs in their ends were identified in the phase I physical map of
O. granulata (http://www.omap.org). The distance between two
ends of each BAC clone was measured in CB units. The CB distances
were then translated into physical distances based on the total
lengthofthephysicalcontigandalsofromanchoringinformationof
these contigs to the O. sativa reference genome sequence using
SYMAP software (Soderlund et al., 2006). The relative positions of
Gran3 LTRs in the physical map were inferred from these data.
Acknowledgements
This work was supported by grants from the National Science
Foundation Plant Genome Program DBI-0321678 to R.A.W., S.A.J.
and P.J.S.M., and the Bud Antle Endowed Chair to R.A.W. We thank
members of the Arizona Genomics Institute’s Resource Center,
Sequencing, Finishing and Bioinformatics groups for generating
and processing the high-quality sequence data used for this analy-
sis. We are grateful to Dr Paul Sanchez for providing genomic DNA
of O. granulata for initial dot-blot analysis. We also thank other
members of the OMAP research team and advisory committee for
fruitful discussions.
Supplementary Material
The following supplementary material is available for this article
online:
Figure S1. Dot plot comparisons of complete (horizontal) and
incomplete (vertical) forms of Gran3 elements.
Figure S2. Neighbor joining tree generated from reverse transcrip-
tase domains of most abundant Ty-3-gypsy elements from Oryza
sativa with those from Gran3 (O. granulata) and Wallabi
(O. australiensis).
Figure S3. Distribution of Gran3 LTR sequence in O. granulata
according to their reciprocal similarity.
Figure S4. Genomic distribution of Gran3 family in the O. granulata
genome as revealed by Fluorescent in situ hybridization (FISH).
Figure S5. Distribution of Gran3 LTRs in the physical map of
O. granulata.
Figure S6. Southern hybridization of Dasheng on Oryza species.
Table S1. Dynamics of RWG family across the genus Oryza.
Table S2. Distribution of Gran3 LTRs in the comparative physical
map of O. granulata genome.
Table S3. Copy number estimates of RWG element LTRs in different
Oryza species obtained using random sheared libraries.
Table S4. List of specific primers used for amplification of specific
domains from RWG family for each Oryza species.
This material is available as part of the online article from http://
www.blackwell-synergy.com.
References
Altschul, S.F., Madden, T.L., Scha ¨ffer, A., Zhang, J., Zhang, Z.,
Miller, W. and Lipman, D.J. (1997) Gapped BLAST and PSI-
BLAST: a new generation of protein database search programs.
Nucleic Acids Res. 25, 3389–3402.
Ammiraju, J.S., Luo, M., Goicoechea, J.L. et al. (2006) The Oryza
bacterial artificial chromosome library resource: construction and
analysis of 12 deep-coverage large-insert BAC libraries that rep-
resent the 10 genome types of the genus Oryza. Genome Res. 16,
140–147.
Bennetzen,J.L.,Ma,J.andDevos,K.M.(2005)Mechanismsofrecent
genome size variation in flowering plants. Ann. Bot. 95, 127–132.
Devos, K.M., Brown, J.K. and Bennetzen, J.L. (2002) Genome size
reduction through illegitimate recombination counteracts gen-
ome expansion in Arabidopsis. Genome Res. 12, 1075–1079.
Doyle, J.J. and Doyle, J.L. (1987) Genomic Plant DNA preparation
from fresh tissue – CTAB method. Phytochem. Bull. 19, 11–15.
Feschotte, C., Jiang, N. and Wessler, S.R. (2002) Plant transposable
elements: where genetics meets genomics. Nature Rev. Genet. 3,
329–341.
Ge, S., Sang, T., Lu, B.R. and Hong, D.Y. (1999) Phylogeny of rice
genomes with emphasis on origins of allotetraploid species.
Proc. Natl Acad. Sci. USA, 96, 14400–14405.
Guo, Y.L. and Ge, S. (2005) Molecular phylogeny of Oryzeae
(Poacea) based on DNA sequences from chloroplast, mitochon-
drial, and nuclear genomes. Am. J. Bot. 92, 1548–1558.
Hawkins, J.S., Kim, H., Nason, J.D., Wing, R.A. and Wendel, J.F.
(2006) Differential lineage-specific amplification of transposable
elements is responsible for genome size variation in Gossypium.
Genome Res. 16, 1252–1261.
International Rice Genome Sequencing Project (2005) The map-
based sequence of the rice genome. Nature, 7052, 793–800.
350 Jetty S. S. Ammiraju et al.
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351

Page 10

Jiang, N., Bao, Z., Temnykh, S., Cheng, Z., Jiang, J., Wing, R.A.,
McCouch, S.R. and Wessler, S.R. (2002a) Dasheng: a recently
amplified nonautonomous long terminal repeat element that is a
major component of pericentromeric regions in rice. Genetics,
161, 1293–1305.
Jiang, N.,Jordan,I.K. andWessler,S.R.(2002b) Dashengand RIRE2.
A nonautonomous long terminal repeat element and its putative
autonomous partner in the rice genome. Plant Physiol. 130,
1697–1705.
Kalendar, R., Tanskanen, J., Immonen, S., Nevo, E. and Schulman,
A.H. (2000) Genome evolution of wild barley (Hordeum sponta-
neum) by BARE-1 retrotransposon dynamics in response to sharp
microclimatic divergence. Proc. Natl Acad. Sci USA, 97,
6603–6607.
Kalendar, R., Vicient, C.M., Peleg, O., Anamthawat-Jonsson, K.,
Bolshoy, A. and Schulman, A.H. (2004) Large retrotransposon
derivatives: abundant, conserved but nonautonomous retroele-
ments of barley and related genomes. Genetics, 166, 1437–1450.
Kellogg, E.A. (1998) Relationships of cereal crops and other grasses.
Proc. Natl Acad. Sci. USA, 95, 2005–2010.
Kidwell, M.G. and Lisch, D.R. (2000) Transposable elements and
host genome evolution. Trends Ecol. Evol. 15, 95–99.
Kimura, M. (1980) A simple method for estimating evolutionary rate
of base substitutions through comparative studies of nucleotide
sequences. J. Mol. Evol. 16, 111–120.
Kumar, S., Tamura, K. and Nei, M. (2004) MEGA3: integrated soft-
ware for molecular evolutionary genetics analysis and sequence
alignment. Brief. Bioinform. 5, 150–163.
Ma, J. and Bennetzen, J.L. (2004) Rapid recent growth and diver-
gence of rice nuclear genomes. Proc. Natl Acad. Sci. USA, 101,
12404–12410.
Ma, J., Devos, K.M. and Bennetzen, J.L. (2004) Analyses of LTR-
retrotransposon structures reveal recent and rapid genomic DNA
loss in rice. Genome Res. 14, 860–869.
McCarthy, E.M., Liu, J., Lizhi, G. and McDonald, J.F. (2002) Long
terminal repeat retrotransposons of Oryza sativa. Genome Biol.
3(10): research 0053.1-0053.11.
Meyers, B.C., Tingey, S.V. and Morgante, M. (2001) Abundance,
distribution, and transcriptional activity of repetitive elements in
the maize genome. Genome Res. 11, 1660–1676.
Ohtsubo, H., Kumekawa, N. and Ohtsubo, E. (1999) RIRE2, a novel
gypsy-type retrotransposon from rice. Genes Genet. Syst. 74,
83–91.
Petrov, D.A. and Wendel, J.F. (2006) Evolution of eukaryotic gen-
ome structure. In Evolutionary Genetics: Concepts and Case
Studies (Fox, C.W. and Wolf, J.B., eds). Oxford: Oxford University
Press, pp. 144–156.
Piegu, B., Guyot, R., Picault, N. et al. (2006) Doubling genome size
without polyploidization: dynamics of retrotransposition-driven
genomic expansions in Oryza australiensis, a wild relative of rice.
Genome Res. 16, 1262–1269.
Ramakrishna, W., Dubcovsky, J., Park, Y.J., Busso, C., Emberton, J.,
SanMiguel,P.andBennetzen,J.L.(2002)Different typesand rates
of genome evolution detected by comparative sequence analysis
of orthologous segments from four cereal genomes. Genetics,
162, 1389–1400.
Rice, P., Longden, I. and Bleasby, A. (2000) EMBOSS: the European
molecular biology open software suite. Trends Genet. 16, 276–
277.
Rozen, S. and Skaletsky, H.J. (2000) Primer3 on the WWW for
general users and for biologist programmers. In Bioinformatics
Methodsand Protocols:Methods
(Krawetz, S. and Misener, S., eds). Totowa, NJ: Humana Press,
pp. 365–386.
inMolecularBiology
SanMiguel, P., Tikhonov, A., Jin, Y.K. et al. (1996) Nested
retrotransposons in the intergenic regions of the maize genome.
Science, 274, 765–768.
SanMiguel, P., Gaut, B.S., Tikhonov, A., Nakajima, Y. and
Bennetzen, J.L. (1998) The paleontology of intergene retrotrans-
posons of maize. Nature Genet. 20, 43–45.
SanMiguel, P.J., Ramakrishna, W., Bennetzen, J.L., Busso, C.S. and
Dubcovsky, J. (2002) Transposable elements, genes and recom-
bination in a 215-kb contig from wheat chromosome 5A(m).
Funct. Integr. Genomics, 2, 70–80.
Schulman, A.H. and Kalendar, R. (2005) A movable feast: diverse
retrotransposons and their contribution to barley genome
dynamics. Cytogenet. Genome Res. 110, 598–605.
Shirasu, K., Schulman, A.H., Lahaye, T. and Schulze-Lefert, P. (2000)
A contiguous 66-kb barley DNA sequence provides evidence for
reversible genome expansion. Genome Res. 10, 893–894.
Soderlund, C., Nelson, W., Shoemaker, A. and Paterson, A.H. (2006)
SyMAP: a system for discovering and viewing syntenic regions of
FPC maps. Genome Res. 16, 1159–1168.
Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W:
improving the sensitivity of progressive multiple sequence align-
mentthroughsequenceweighting,position-specificgappenalties
and weight matrix choice. Nucleic Acids Res. 22, 4673–4680.
Tikhonov, A.P., SanMiguel, P.J., Nakajima, Y., Gorenstein, N.M.,
Bennetzen, J.L. and Avramova, Z. (1999) Colinearity and its
exceptions in orthologous adh regions of maize and sorghum.
Proc. Natl Acad. Sci. USA, 96, 7409–7414.
Uozu, S., Ikehashi, H., Ohmido, N., Ohtsubo, H., Ohtsubo, E. and
Fukui, K. (1997) Repetitive sequences: cause for variation in
genome size and chromosome morphology in the genus Oryza.
Plant Mol. Biol. 35, 791–799.
Vicient, C.M, Suoniemi, A., Anamthawat-Jonsson, K., Tanskanen,
J., Beharav, A., Nevo, E. and Schulman, A.H. (1999) Retrotrans-
poson BARE-1 and its role in genome evolution in the genus
Hordeum. Plant Cell, 11, 1769–1784.
Vitte, C. and Panaud, O. (2005) LTR-retrotransposons and flowering
plant genome size: emergence of the increase/decrease model.
Cytogenet. Genome Res. 110, 91–107.
Vitte, C., Ishii, T., Lamy, F., Brar, D.S. and Panaud, O. (2004)
Genomic paleontology provides evidence for two distinct ori-
gins of Asian rice (Oryza sativa, L.). Mol. Genet. Genomics,
272, 504–511.
Walling, J.G., Pires, J.C. and Jackson, S.A. (2005) Preparation
of samples for comparative studies of plant chromosomes
using in situ hybridization methods. Methods Enzymol. 395,
443–460.
Walling, J.G., Shoemaker, R., Young, N., Mudge, J. and Jackson, S.
(2006) Chromosome-level homeology in paleopolyploid soybean
(Glycine max) revealed through integration of genetic and chro-
mosome maps. Genetics, 172, 1893–1900.
Wendel, J.F. and Wessler, S.R. (2000) Retrotransposon-mediated
genome evolution on a local ecological scale. Proc. Natl Acad. Sci
USA, 97, 6250–6252.
Wendel, J.F., Cronn, R.C., Johnston, J.S. and Price, H.J. (2002) Feast
and famine in plant genomes. Genetica, 115, 37–47.
Wing, R.A., Ammiraju, J.S.S., Luo, M. et al. (2005) The Oryza map
alignment project: the golden path to unlocking the genetic
potential of wild rice species. Plant Mol. Biol. 59, 53–62.
Yu, J., Wang, J., Lin, W. et al. (2005) The genomes of Oryza sativa:
a history of duplications. PloS Biol. 3(2), e38.
Zhang, X. and Wessler, S.R. (2004) Genome-wide comparative
analysis of the transposable elements in the related species
Arabidopsis thaliana and Brassica oleracea. Proc. Natl Acad. Sci.
USA, 101, 5589–5594.
Evolutionary dynamics of ancient retrotransposon family 351
ª 2007 The Authors
Journal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 52, 342–351