Structural Heterogeneity and Genomic Distribution of Drosophila melanogaster
Lucı ´a Alonso-Gonza ´lez, Ana Domı ´nguez, and Jesu ´s Albornoz
Area de Gene ´tica, Departamento de Biologı ´a Funcional, Universidad de Oviedo, Oviedo, Spain
Structural heterogeneity of five long terminal repeat (LTR) retrotransposon families (297, mdg 1, 412, copia, and 1731)
was investigated in Drosophila melanogaster. The genomic distribution of canonical and rearranged elements was
studied by comparing hybridization patterns of Southern blots on salivary glands from adult females and males with in
situ hybridization on polytene chromosomes. The proportion and genomic distribution of noncanonical copies is
distinctive to each family and presents constant features in the four different D. melanogaster strains studied. Most
elements of families 297 and mdg 1 were noncanonical and presented large interstock and intrastock polymorphism.
Noncanonical elements of these two families were mostly located in euchromatin, although not restricted to it. The
elements of families 412 and copia were better conserved. The proportion of noncanonical elements was lower. The 1731
family is mainly composed of noncanonical, b-heterochromatic elements that are highly conserved among stocks. The
relation of structural polymorphism to phylogeny, transpositional activity and the role of natural selection in the
maintenance of transposable elements are discussed.
A large fraction of the eukaryotic DNA is composed
of transposable elements that can cause mutations when
they transpose to novel sites. Due to their significant
influence on genomic diversity, the nature of forces
affecting the transpositional spread of transposons and
their distribution throughout the genome is a major
problem of evolutionary genetics. The forces controlling
their abundance within their host genomes have been the
subject of a great deal of empirical and theoretical research
(e.g., Langley, Brookfield, and Kaplan 1983; Capy 1997),
although their nature and relative significance is still
controversial. Such forces include selection against the
deleterious effects of transposable elements in the host
genome, self-regulation, and interaction with the host
genome (Nuzhdin 1999).
In previous studies on a set of mutation-accumulation
lines, we found that the main source of variability affecting
transposable elements was the generation of internal re-
arrangements. The rate of rearrangement was 8.5 3 10–6
(Domı ´nguez and Albornoz 1996; Albornoz and
Domı ´nguez 1999; Domı ´nguez and Albornoz 1999). This
applied to class I (297, 412, and copia), class II (P and
hobo), and Foldback (FB) elements. The existence of
this kind of mutation with a nonnegligible rate allows us to
presume structural heterogeneity among transposable
elements of the Drosophila genome. Structural heteroge-
neity in the elements pertaining to the same family is
common for class II elements, retroposons, and the FB
element of Drosophila (Potter et al. 1980; O’Hare and
Rubin 1983; Vaury, Bucheton, and Pelisson 1989;
Bucheton et al. 1992; Capy, David, and Hartl 1992). In
contrast, the sequences of Drosophila retrotransposons
(long terminal repeat [LTR] elements) have been shown to
be closely conserved in the genome, present mainly as full-
sized elements with only a few rearranged copies
(Finnegan 1985). The existence of noncanonical elements
in the heterochromatin of D. melanogaster has been
reported for some retrotransposon families (Shevelyov,
Balakireva, and Gvozdev 1989; Vaury, Bucheton, and
Pellison 1989; Montchamp-Moreau et al. 1993). In plants,
however, the structural heterogeneity of retrotransposon
families is the rule (Marillonnet and Wessler 1998).
Rearranged copies of transposable elements are often
considered as old genomic components, usually embedded
in heterochromatin (Bucheton et al. 1992; Vaury et al.
1994). There are different nonexclusive hypotheses to ex-
plain the accumulation of transposable elements in het-
erochromatin. This region may either contain insertional
hotspots or may function like a trap in which trans-
posable elements become immobilized by position-effect
inactivation. Another possibility is that once inserted there,
elements are not so easily eliminated by natural selection
and recombination as are euchromatic insertions (reviewed
in Charlesworth, Sniegowosky, and Stephan 1994). Recent
evidence suggests that the relationship between transpos-
able elements and heterochromatin will not be quite so
straightforward. Constitutive heterochromatin contains
genetically active domains, and there is evidence for
elements that underwent fixation, probably under selective
pressure for regulatory roles (reviewed in Dimitri 1997;
Dimitri and Junakovic 1999).
Recent studies have shown the existence of de-
generate euchromatic elements pertaining to the 297 and
412 families (Cizeron and Bie ´mont 1999; Domı ´nguez and
Albornoz 1999) in D. melanogaster and D. simulans.
Studies based on the released sequence of the D.
melanogaster and Caenorhabditis elegans genomes have
also reported structural heterogeneity of euchromatic LTR-
retrotransposons (Bowen and McDonald 2001; Frame,
Cutfield, and Pulter 2001; Ganko, Fielman, and McDonald
2001). In this paper, we address the question of the
structural heterogeneity and genomic distribution of five
families of D. melanogaser LTR elements (297, 412,
1731, copia, and mdg1). Southern blot analysis of
individuals allows us to reveal low-frequency bands and
evaluate intrastock polymorphism for rearranged elements.
The comparison of laboratory strains and a wild popula-
tion show the existence of characteristic features of each
Key words: Transposable elements, retrotransposons, rearrange-
ments, heterochromatin, Drosophila melanogaster.
Mol. Biol. Evol. 20(3):401–409. 2003
? 2003 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038
family both in the extent of heterogeneity and in the
proportion of rearranged elements per individual. The
genomic distribution of canonical and noncanonical
elements was investigated in one fully homozygous line
by comparing the results of Southern blot analysis and
fluorescent in situ hybridization.
Materials and Methods
Three laboratory strains and a natural population were
analyzed in this work. The laboratory lines were Oregon
R, Canton S (both from the Bowling Green Stock Center)
and the fully homozygous line C66 from Domı ´nguez and
Albornoz (1996). Wild flies were collected from a natural
population recently caught in Llanera (in the north of
Genomic DNA was extracted from the salivary
glands of third-instar larvae, single flies, and pools of
flies as described in Di Franco et al. (1989), Domı ´nguez
and Albornoz (1996), and Albornoz and Domı ´nguez
(1999), respectively. Probes for 297, 1731, and copia
were described in Di Franco, Galuppi, and Junakovic
(1992). Probes for 412 and mdg 1 were obtained from
clones pOR 708 and 5B257 bearing, respectively, a full-
size 412 element (Shepherd and Finnegan 1984) and the
2.8-kb Hin dIII fragment from Dm 58 (Ilyin et al. 1980).
These clones were digested with the appropriate combi-
nation of restriction enzymes, and, after electrophoresis,
the fragment to be used as a probe was extracted from the
agarose gel. Southern blots were performed as previously
described (Albornoz and Domı ´nguez 1999). To check for
the completeness of enzyme digestion, filters were retested
with the 1.9-kb Eco RI-Bam HI fragment from Dm FB1,
which is a piece of single-copy genomic DNA (Potter et al.
1980). The autoradiographs were scanned and analyzed
with the GelReader software from the National Center for
Supercomputing Applications (1991). In situ hybridization
was carried out following the protocol of Gatti, Bonaccorsi
and Pimpinelli (1994) and fluorescein-labeled probes.
Deductions on the Proportion of Noncanonical Elements
and Their Genomic Distribution
To categorize elements, we used the sequences
described in GenBank as representatives of each family
as the canonical element (X03431 for 297, X59545 for
mdg 1, X04132 for 412, X04456 for copia, and X07656
for 1731). Classification of elements as canonical or
noncanonical was based on obtaining of the expected
internal restriction fragments in Southern blots. Genomic
DNA was digested with the restriction enzyme or the
combination of two that rend the largest internal fragment
for each family. Combinations of enzymes and the
expected size of the fragment detected for the canonical
element of each retrotransposon family are shown in
figure 1. Noncanonical elements will generate fragments
of a different size. A larger size than expected might be
a consequence of the loss of a restriction site or insertion
of a new sequence, and a smaller size might be due to
either internal deletions or the presence of additional
restriction sites. This approach is limited by the availability
of restriction enzymes and adequate probes, it being
possible that some noncanonical elements were not
The total number of elements of each family was
scored from Southern blots. Digestions with Hin dIII were
used for 297, copia, and mdg 1 and with Eco RI for 1731
and 412. The 1731 element has no Eco RI internal sites,
thus, the number of bands will fit with the number of
copies. Elements 412, copia, and mdg 1 have one or more
internal sites for the restriction enzyme. Consequently, the
digestion of genomic DNA produces two or more
fragments, but only one of the ends of the element
hybridizes with the probe, and hence the number of bands
equals the number of copies. Element 297 has two internal
Hin dIII sites, and so each element produces three frag-
ments after digestion. The probe reveals one of the ends
of the element and the central fragment. Consequently,
the number of bands exceeds the number of copies by
one, which corresponds to all the central fragments (see
The genomic distribution of canonical and non-
canonical elements was approached by comparing in situ
hybridization on polytene chromosomes with Southern
blots. In the salivary glands of Drosophila larvae,
FIG. 1.—Sites for restriction enzymes that were used in Southern
blots and probes. Size of the expected fragments for canonical copies of
each retrotransposon family.
402Alonso-Gonza ´lez et al.
euchromatin, and b-heterochromatin undergo polyteniza-
tion (John 1988; Dimitri 1997). b-heterochromatin does
not present the banding pattern characteristic of euchro-
matin and can only be seen as an amorphous region at the
base of each chromosome arm. a-heterochromatin under-
goes few, if any, rounds of replication during polyteniza-
tion and is found as a dense region centrally located at
the chromocenter. It corresponds to the bulk of constitu-
tive heterochromatic segments so evident in the mitotic
chromosomes. From these considerations, we have
estimated the number of euchromatic elements in the fully
homozygous line C66 as the number of signals detected
in chromosome arms by in situ hybridization. The number
of elements in b-heterochromatin was estimated as the
difference between the number of bands in Southern blots
from salivary glands and the number of in situ signals in
chromosome arms. The comparison of the autoradiographs
from adults and salivary glands was used to identify the
elements located in underreplicated regions in a-hetero-
chromatin. Finally, the number of elements on chromo-
some Y was deduced from the comparison of blots from
adult males and females.
We assume that hybridization signals are due to
single elements. Hence the number of euchromatic ele-
ments may be underestimated if in some instances there
is more than one element per hybridization signal.
Southern blot pictures of digestions performed to
reveal noncanonical elements in individuals from the four
strains are shown in figure 2 and the data are summarized
in table 1. The proportion of noncanonical elements
was estimated from the densitometry analysis of
autoradiographs as the relative intensity of bands other
than that expected for the canonical element.
The number of noncanonical elements per individual
was computed as the number of restriction fragments other
than that expected for the canonical elements (see fig. 2).
This will be an estimation of the minimum number of
structurally variant elements per individual, since the
actual number depends on the heterozygosity of the lines
and on the existence of multiple loci with noncanonical
elements coincident in size. Bands present in all males
from a strain, and only in the males, were classified as
noncanonical elements located on the Y chromosome.
The number of noncanonical size classes of each
family of elements, as well as the percentage that they
account for in the genome, were similar for the four
strains. Families 297 and mdg 1 shared similar features:
they have 17 to 20 classes of noncanonical elements per
individual that account for about 80% of the hybridization
signal. Rearranged elements were highly heterogeneous in
size. Only two 297 (6.7 and 4.8 kb) and three mdg 1 (5.3,
4.3, and 3.5 kb) noncanonical classes were common to the
four strains. In the wild population, noncanonical elements
were also very heterogeneous, most of these segregating at
low frequencies (fig. 3a and b). The total number of size
classes in the sample from Llanera for element 297 was 64
with a mean frequency of 0.30, and for mdg1, the number
of classes was 79 with a mean frequency of 0.24.
FIG. 2.—Banding patterns in Southern blots made to detect
noncanonical elements in individuals from the four strains. Bands
generated by potentially canonical elements (see fig. 1) are indicated by
Noncanonical Retrotransposons and Heterochromatin403
Noncanonical classes of both families that appeared in the
four strains were at frequency 1 in the Llanera population.
For the 412 and copia families, the proportion of
noncanonical elements per individual was lower, as was the
number of size classes (table 1). Noncanonical elements
were heterogeneous in size. As can be seen from figure 2,
the autoradiographic bands corresponding to noncanonical
elements are faint and quite close to the canonical element,
particularly for copia. This observation contrasts with the
results obtained for the other three elements included in the
study, where prominent noncanonical bands can be
observed. Only two 412 noncanonical classes (4.8 and
4.6 kb) were common to the four strains; these were also
present in every individual from the wild population. In the
sample from the Llanera population, the number of size
classes of noncanonical elements for families 412 and
copia were 32 and 23 with mean frequencies of 0.27 and
0.21, respectively (fig. 3c and d).
The 1731 family was mainly composed of rearranged
elements, the proportion of putative canonical copies being
between 6% and 11%. Southern blots showed between 13
and 20 noncanonical size classes per individual in the
different lines studied. The number of size classes in the
sample of the wild population was 24 with a mean
frequency of 0.55 (fig. 3e). Ten noncanonical classes were
common to the four strains, but only four noncanonical
classes were present in every individual; two of the latter
were insertions on chromosome Y. The common non-
canonical class of 3.6 kb was the most abundant in the four
strains (fig. 2d) and accounted for 20% to 25% of the
hybridization signal in every strain.
The genomic distribution of structurally variant
elements was analyzed in the line C66 (table 2). The
complete homozygosity of this line allows us to test the
number of hybridization signals obtained by different
methods and to compare them so as to deduce the
distribution of elements in this genotype (see Materials
and Methods). First, the number of elements per haploid
genome was deduced from Southern blots on adult flies.
The number of noncanonical size classes is the minimum
number of noncanonical elements. With the exception of
copia, these numbers agree quite well with those that can
be estimated by the product of their relative intensity in
autoradiographs by the number of elements: 26, 26, 8, 11,
and 17 for families 297, mdg1, 412, copia, and 1731,
respectively. This implies that most size classes of re-
arranged elements are present in one or a few loci per
haploid genome. This estimation might contain a sub-
stantial error, since if some bands corresponding to non-
canonical elements were close to those generated by the
canonical elements, their absorption profiles would over-
lap and hence the proportion of noncanonical elements
would be overestimated. If, on the other hand, the full-
size element is very abundant, saturation of the photo-
graphic sheet may be another cause of underestimation
of the proportion of canonical elements. This source of
error is of particular importance for copia (see fig. 2)
and could explain the discrepancy between the two estima-
tions of the number of noncanonical elements.
The number of in situ hybridization signals is the
minimum number of euchromatic elements, given that
there might be more than one element per chromosome
The 297 family had 32 elements per haploid genome.
Five elements were inserted in a-heterochromatin and the
remaining 27 were euchromatic. One of them is located in
band 20A, at the border between euchromatin and b-
heterochromatin. At least 20 elements were noncanonical
Number of Size Classes per Individual and Percentage Intensity of Noncanonical Elements in Southern Blots of the Four
Number of IndividualsNoncanonical Classes per Individual 6 SD
Mean % of Noncanonical
Copies per IndividualElementStrain Females Males FemalesMales
20.33 6 0.37
18.50 6 1.50
17 6 0
20 6 0
16.89 6 0.42
17.00 6 1.00
17 6 0
20 6 0
79.91 6 1.87
80.35 6 4.03
76.25 6 2.87
80.30 6 1.92
17.29 6 0.68
16 6 0
18 6 0
18 6 0
18.71 6 1.23 (1)
21.5 6 1.50 (5)
20 6 0 (2)
20 6 0 (2)
81.10 6 0.76
78.30 6 1.37
76.26 6 0.66
81.46 6 1.14
6.20 6 0.76
5.33. 6 0.33
5 6 0
7 6 0
8.90 6 0.64 (2)
8.20 6 0.58 (1)
6.50 6 0.50
7 6 0
46.82 6 3.06
39.99 6 3.51
42.40 6 2.14
38.04 6 3.84
3.33 6 0.47
6.20 6 0.49
3 6 0
3 6 0
4.62 6 0.56
9.80 6 0.77
4 6 0 (1)
3 6 0
26.73 6 3.50
55.37 6 3.45
53.07 6 8.34
44.62 6 2.26
13.13 6 0.67
13.25 6 0.48
14 6 0
13 6 0
13.63 6 0.68 (2)
20.20 6 0.37 (6)
18 6 0 (4)
16 6 0 (3)
91.00 6 1.32
93.18 6 0.50
93.06 6 0,32
88.72 6 0.48
NOTE.—Numbers in parentheses indicates the number of size classes in chromosome Y.
404 Alonso-Gonza ´lez et al.
and only one of them was located in the a-heterochroma-
tin. Consequently, there is a maximum of eight complete
elements in the euchromatin and four in the a-heterochro-
The mdg 1 family was represented by 32 elements per
haploid genome. Seven elements were inserted in a-
heterochromatin (two on the Y chromosome) and 19 were
euchromatic; the remaining six elements not detected by in
situ hybridization were classified as b-heterochromatic.
Twenty elements were found to be noncanonical; of these,
seven were a-heterochromatic (two on the Y chromo-
some). The remaining 13 noncanonical elements would be
distributed between b-heterochromatin and euchromatin.
Consequently, there is a maximum of 12 complete
elements in euchromatin.
The 412 family had 20 elements per haploid genome.
Six were in a-heterochromatin (one on the Y chromo-
some) and 14 were euchromatic. Seven noncanonical
FIG. 3.—Size distributions of autoradiographic bands in the samples from the Llanera population.
Noncanonical Retrotransposons and Heterochromatin405
elements were observed; three of these being a-hetero-
chromatic. Consequently, there is a maximum of 10
canonical elements in euchromatin and three in a-
Copia was represented by 27 elements. Six elements
were a-heterochromatic and 18 euchromatic. The other
three elements not detected by in situ hybridization might
be b-heterochromatic. Moreover, two elements were
mapped to the polytene band 41F, at the border between
b-heterochromatin and euchromatin. Only three non-
canonical elements were detected, and these were
not a-heterochromatic. Therefore, six potentially full-size
elements were inserted in the a-heterochromatin.
The element 1731 presented 19 copies per haploid
genome, and the number of noncanonical size classes was
16. The noncanonical size class of 3.6 kb (see fig. 2) ac-
counted for 25% of the Southern blot hybridization sig-
Therefore, the number of canonical elements would
probably be one, whereas there would be two or three
copies of the element giving the restriction fragment of 3.6
kb. Three noncanonical elements were a-heterochromatic
and were located on chromosome Y. Only two signals were
detected by in situ hybridization, one of which was in band
20A, at the b-heterochromatin–euchromatin boundary.
Thus, the number of b-heterochromatic elements would
be 14. Accordingly, an intense signal can be seen at the
chromocenter region corresponding to b-heterochromatin.
Studies carried out by in situ hybridization to mitotic
chromosomes have revealed prominent stable clusters of
transposable elements located in heterochromatic regions
(Carmena and Gonza ´lez 1995; Pimpinelli et al. 1995).
Studies of reassociation kinetics suggest that transposable
1981), that is, 16 Mb out of the 180 Mb that constitute the
genome of D. melanogaster (Adams et al. 2000). The
consideration that there are 50 families with an average size
of 5.4 kb and an average copy number in the euchromatic
arms of 32 (Maside et al. 2001) leads to the conclusion that
approximately 8 Mb out of the 120 Mb that constitute
euchromatin corresponds to transposable elements. This
implies that half of the sequences are heterochromatic, and
hence the density of transposable elements in heterochro-
matin must be twofold higher than in euchromatin.
With our technical approach, only 38% (49 out of
a total of 130 elements for the five families considered) of
the bands were heterochromatic in line C66. This value is
somewhat lower than expected, bearing in mind the above
considerations. One possible source of this discrepancy is
that heterochromatic elements are linked to a fraction of
DNA that remains a high-molecular-weight fraction in
restriction digests and is not transferred to the membrane.
Another possibility is that this fraction of DNA might
contribute to a Southern blot picture as a smear of DNA
fragments resulting from degradation during extraction
(Terrinoni et al. 1997). Our study is hence restricted to the
fraction of transposable elements that are detectable with
the regular Southern blot method, a fraction that is biased
towards euchromatic elements. It is worth noting that the
comparison of DNA from adults and salivary glands (a
method devised by Di Franco et al. 1989) allows a-
heterochromatic copies to be properly distinguished, as
can be noticed by the fact that the elements on
chromosome Y were independently classified as hetero-
chromatic. On the other hand, Southern blot estimates of
family copy numbers were always higher than in situ
family copy numbers, as was expected, denoting a good
level of resolution of this technique.
Some specific characteristic of each family of
transposable elements can be observed with regards to
the intragenomic distribution of elements. Family 1731
was the only one preferentially inserted in heterochro-
matin, only two of the 19 copies detected in line C66
were viewable in situ and one of these was in band
20A, at the junction between euchromatin and b-
heterochromatin. The other four elements showed a small
proportion of heterochromatic copies. On the other hand,
families 297, 412, and copia had potentially active
elements in a-heterochromatin. This is in line with some
reports that showed that heterochromatic transposable ele-
ments may be active (Hochstenbach et al. 1996; Chalvet
et al. 1998).
The proportion of noncanonical elements is a partic-
ular characteristic of each family. Elements copia and 412
Genomic Distribution of Elements in Line C 66
Distribution in Chromatin
Euchromatin Element Total
Noncanonical Elements in
b-Heterochromatin 1 Euchromatin
NOTE.—The numbers were obtained by comparing results of in situ hybridization to polytenic chromosomes and Southern blots with different combinations of enzymes
on DNA extracted from adult males and females and larval salivary glands (see text). Numbers in parentheses indicate the number of noncanonical elements in
406Alonso-Gonza ´lez et al.
were mainly represented by putative full-size elements,
mdg 1 and 297 had around 75% to 80% of rearranged
elements that were not conserved between stocks, and
most copies of 1731 were noncanonical elements with
a remarkable conservation within and between stocks.
These results can be compared with the study of LTR-
retrotransposons on the release of the complete euchro-
matic genome sequence of D. melanogaster by Bowen and
McDonald (2001) that showed many of the characterized
elements to contain sequence deletions.
In previous studies (Domı ´nguez and Albornoz 1996),
we estimated a rate of transposable element structural
mutation of 8.5 3 10–6. A similar value was recently
reported (6.8 3 10–6) by Maside et al. (2001). The
existence of this kind of mutation is consistent with the
structural heterogeneity observed and its rate is congruent
with that reported by Petrov, Lozovskaya, and Hartl
(1996) for small deletions in Drosophila, as dicussed in
Albornoz and Domı ´nguez (1999) and Domı ´nguez and
Albornoz (1999). However, several other processes, such
as insertions, unequal crossing-over, ectopic recombina-
tion, and abortive gap repair may be responsible for the
observed structural mutations (Brunet et al. 2002).
It is rational to think that structurally variant elements
are functionally different, and this must be taken into
account in models to explain their population dynamics.
Kaplan, Darden, and Langley (1983) proposed a model
that allows mutation to functionally distinct mutant
elements to describe the evolution of transposable
elements. The model suggests that when an element enters
a population, it can become extinct within the first few
generations, and if not, then the average copy number per
host genome of the wild type increases quickly in the early
stages of the element’s evolution to then decrease slowly
as the average copy number of the mutant starts to grow.
The average copy number in the population of the wild
type and mutant then appears to stabilize, but the wild type
ultimately disappears from the population, leaving only the
mutant. At this point, transposition stops and the mutant
eventually becomes extinct due to the forces of deletion
Following this model, the composition of the 1731
family is that expected for an ancient element that has
become inactive. The abundance of noncanonical ele-
ments, in addition to the wild type that was observed for
the 297 and mdg 1 families, will be associated with
a reduction in transpositional activity. Finally, the
composition of the copia and 412 families is that expected
for young, actively transposing elements. The phyloge-
netic distribution of the 1731 family, which is widespread
throughout the Drosophila genus as well as being present
in the Scaptomyza and Zaprionus genera (Montchamp-
Moreau et al. 1993), is consistent with this element being
an old component of the Drosophila genome, leaving the
defective elements that were inserted in the pericentro-
meric regions only as vestiges. In addition, 1731 was
inactive in the three experiments on transposition rate that
included this element (reviewed in Junakovic, Di Franco,
and Terrinoni 1997). It has been proposed that during
genome evolution, transposable elements that lose their
transposition activity may conversely acquire new func-
tions (von Sternberg et al. 1992; McDonald 1993). In
relation to this topic, Kalmykova, Maisonhaute, and
Gvozdev (1999) proposed that the fused gag-pol poly-
peptide encoded by a fraction of 1731 element might serve
for the normal development of host testes. In that case, the
1731 element would be subject to positive selection. This
is congruent with the conservation observed within and
between stocks of 1731 noncanonical elements and with
the reported conservation between stocks and species of
the Bam HI Sal I fragment (Vaury, Bucheton, and Pelisson
1989; Montchamp-Moreau et al. 1993) that includes the
gag gene and most of the pol ORF (Champion et al. 1992).
The age of the other four elements estimated on the
basis of their phylogenetic distribution does not conform
to the predictions of the model depicted above. The
elements copia and 412 are ancient components of the
Drosophila genome, as they have been found in
representatives of all the major Drosophila radiations
(Stacey et al. 1983; Cizeron et al. 1998; Bie ´mont and
Cizeron 1999). Data on mdg 1 is scarce. This family must
have been acquired before the Sophophoran radiation
because it is present in the Obscura group (de Frutos,
Peterson, and Kidwell 1992). Element 297 is restricted to
D. melanogaster, its sibling species D. simulans and D.
mauritana, and the closely related D. yakuba; thus this
family is probably of recent origin (Martin, Wiernasz, and
Schedl 1983). The four elements have shown reduced
transpositional activity in one or another stock (data
reviewed in Junakovic, Di Franco, and Terrinoni 1997).
The element copia has been shown to transpose actively,
at a rate of 10–3to 10–2, in some lines with permissive
alleles (Pasyukova, Nuzhdin, and Filatov 1998), which is
congruent with its structural conservation.
It has been shown that the full-length elements from
the D. melanogaster genome are very young (Bowen and
McDonald 2001). These authors posed the question of the
immediate source of the full-length LTR-retrotransposons
and pointed to three possibilities. The full-length elements
are descendants from older elements that have been
actively eliminated from the D. melanogaster genome or
from older elements sequestered within the heterochroma-
tin. A third possibility is that the LTR-retrotransposons
currently present in the D. melanogaster genome have
been subject to vertical and horizontal transmission during
their evolution, as has been suggested for copia (Jordan
and McDonald 1998). This possibility could explain the
lack of correspondence between the age of the elements as
inferred on a phylogenetic basis and from the proportion of
noncanonical elements in the genome.
A final consideration concerns the close coincidence
of the proportion and number of noncanonical elements in
unrelated stocks within each of the different families
studied. This observation suggests the possibility that non-
canonical elements could play some role in the regula-
tion of the transpositional activity of wild-type elements.
Posttranscriptional inhibition at greater copy numbers
has been demonstrated for I elements and is attributed
to cosuppression (Jensen, Gassama, and Heidmann 1999).
Cosuppression does not require any translatable sequence,
and the severity of repression correlates with copy number.
There is considerable evidence from a number of trans-
Noncanonical Retrotransposons and Heterochromatin 407
posable elements and hosts that this kind of interaction can
be important to activity (Birchler, Pal-Bhadra, and Bhadra
We are grateful to D. J. Finnegan for providing clones
pOR 708 and 5B257, to R. Pin ˜eiro for providing samples
of the Llanera population, to Hero S.A. for providing
culture vials, to S. Pimpinelli for training L.A-G. in FISH
and to an anonymous referee for their helpful comments.
L.A-G. was supported by an FPI fellowship from the
Ministerio de Ciencia y Tecnologı ´a. This work was
supported by grant PB-97–1268 (to J.A.) from the
Ministerio de Educacio ´n y Cultura.
Adams, M. D., S. E. Celniker, R. A. Holt et al. (190 co-authors).
2000. The genome sequence of Drosophila melanogaster.
Albornoz, J., and A. Domı ´nguez. 1999. Spontaneous changes in
Drosophila melanogaster transposable elements and their
effects on fitness. Heredity 83:663–670.
Bie ´mont, C., and G. Cizeron. 1999. Distribution of transposable
elements in Drosophila species. Genetica 105:43–62.
Birchler, J. A., M. Pal-Bhadra, and U. Bhadra. 1999. Less from
more: cosuppression of transposable elements. Nat. Genet.
Bowen, N. J., and J. F. McDonald. 2001. Drosophila euchromatic
LTR retrotransposons are much younger than the host species
in which they reside. Genome Res. 11:1527–1540.
Brunet, F., T. Giraud, F. Godin, and P. Capy. 2002. Do deletions
of Mos I-like elements occur randomly in the Drosophilidae
family? J. Mol. Evol. 54:227–234.
Bucheton, A., C. Vaury, M. C. Chaboissier, P. Abad, A. Pe ´lisson,
and M. Simonelig. 1992. I elements and the Drosophila
genome. Genetica 86:175–190.
Capy, P. ed. 1997. Evolution and impact of transposable
elements. Kluwer, Dordretch, The Netherlands.
Capy, P., J. R. David, and D. L. Hartl. 1992. Evolution of the
transposable element mariner in the Drosophila melanogaster
species group. Genetica 86:37–46.
Carmena, M., and C. Gonza ´lez. 1995. Transposable elements
map in a conserved pattern of distribution extending from b-
heterochromatin to centromeres in Drosophila melanogaster.
Chalvet, F., C. Di Franco, A. Terrinoni, A. Pelisson, N.
Junakovic, and A. Bucheton. 1998. Potentially active copies
of the gypsy retroelement are confined to the Y chromosome
of some strains of Drosophila melanogaster possibly as the
result of the female-specific effect of the flamenco gene. J.
Mol. Evol. 46:437–441.
Champion, S., C. Maisonhaute, M. H. Kim, and M. Best-
Belpomme. 1992. Characterization of the reverse transcriptase
of 1731, a Drosophila melanogaster retrotransposon. Eur. J.
Charlesworth, B., P. D. Sniegowski, and W. Stephan. 1994. The
evolutionary dynamics of repetitive DNA in eukaryotes.
Cizeron, G., and C. Bie ´mont. 1999. Polymorphism in structure of
the retrotransposable element 412 in Drosopila simulans and
D. melanogaster populations. Gene 232:183–190.
Cizeron, G., F. Lemeunier, C. Lœvenbruck, A. Brehm, and C.
Bie ´mont. 1998. Distribution of the retrotransposable
element 412 in Drosophila species. Mol. Biol. Evol.
de Frutos, R., K. R. Peterson, and M. G. Kidwell. 1992.
Distribution of Drosophila melanogaster transposable ele-
ment sequences in species of the obscura group. Chromosoma
Di Franco C., D. Galuppi, and N. Junakovic. 1992. Genomic
distribution of transposable elements among individuals of an
inbred Drosophila line. Genetica 86:1–11.
Di Franco, C., C. Pisano, P. Dimitri, S. Gigliotti, and N.
Junakovic. 1989. Genomic distribution of copia-like trans-
posable elements in somatic tissues and during development
of Drosophila melanogaster. Chromosoma 98:402–410.
Dimitri, P. 1997. Constitutive hetrochromatin and trans-
posable elements in Drosophila melanogaster. Genetica 100:
Dimitri, P., and N. Junakovic. 1999. Revising the selfish DNA
hypothesis. New evidence on accumulation of transposable
elements in heterochromatin. Trends Genet. 15:123–124.
Domı ´nguez, A., and J. Albornoz. 1996. Rates of movement of
transposable elements in Drosophila melanogaster. Mol. Gen.
———. 1999. Structural instability of 297 element in Drosophila
melanogaster. Genetica 105:239–248.
Finnegan, D. J. 1985. Transposable elements in eukaryotes. Int.
Rev. Cytol. 93:281–326.
Frame, I. G., J. F. Cutfield, and R. T. M. Poulter. 2001. New
BEL-like LTR-retrotransposons in Fugu rubipes, Caenorhab-
ditis elegans, and Drosophila melanogaster. Gene 263:219–
Ganko, E. W., K. T. Fielman, and J. F. McDonald. 2001.
Evolutionary history of Cer elements and their impact on the
C. elegans genome. Genome Res. 11:2066–2074.
Gatti, M., S. Bonaccorsi, and S. Pimpinelli. 1994. Looking at
Drosophila mitotic chromosomes. Methods Cell. Biol. 44:
Hochstenbach, R., H. Harhangi, K. Schouren, P. Bindels, R.
Suijkerbuijk, and W. Hennig. 1996. Transcription of gypsy
elements in a Y-chromosome male fertility gene of Drosoph-
ila hydei. Genetics 142:437–466.
Ilyin, Y. V., V. G. Chmeliauskaite, E. V. Ananiev, N. V.
Lyubormirskaya, V. V. Kulguskin, A. A. Bayev, and G. P.
Georgiev. 1980. Mobile dispersed genetic element mdg1 of
Drosophila melanogaster: structural organization. Nucleic
Acids Res. 22:5333–5346.
Jensen, S., M. P. Gassama, and T. Heidmann. 1999. Taming of
transposable elements by homology-dependent gene silenc-
ing. Nat. Genet. 21:209–212.
John, B. 1988. The biology of heterochromatin. Pp. 1–128 in
R.S. Verma, ed. Heterochromatin: Molecular and structural
aspects. Cambridge University Press, Cambridge.
Jordan, I. K., and J. F. McDonald. 1998. Evolution of the copia
retrotransposon in the Drosophila melanogaster species
subgroup. Mol. Biol. Evol. 15:1160–1171.
Junakovic, N., C. Di Franco, and A. Terrinoni. 1997. Evidence
for a host role in regulating the activity of transposable
elements in Drosophila melanogaster: the case of the
persistent instability of Bari 1 elements in Charolles stock.
Kalmykova, A., C. Maisonhaute, and V. Gvozdev. 1999.
Retrotransposon 1731 in Drosophila melanogaster changes
retrovirus-like expression strategy in host genome. Genetica
Kaplan, N., T. Darden, and C. H. Langley. 1985. Evolution and
extinction of transposable elements in mendelian populations.
Langley, C. H., J. F. Y. Brookfield, and N. L. Kaplan. 1983.
408 Alonso-Gonza ´lez et al.
Transposable elements in mendelian populations. I. A theory. Download full-text
Marillonnet, S., and S. R. Wessler. 1998. Extreme structural
heterogeneity among the members of a maize retrotransposon
family. Genetics 150:1245–1256.
Martin, G., D. Wiernasz, and P. Schedl. 1983. Evolution of
repetitive-disersed DNA. J. Mol. Evol. 19:203–213.
Maside, X., C. Bartolome ´, S. Assimacopoulos, and B. Charles-
worth. 2001. Rates of movement and distribution of trans-
posable elements in Drosophila melanogaster: in situ
hybridization vs. Southern blotting data. Genet. Res. 78:
McDonald, J. F. 1993. Evolution and consequences of transpos-
able elements. Curr. Opin. Genet. Develop. 3:855–864.
Montchamp-Moreau, C., S. Ronsseray, M. Jacques, M. Leh-
mann, and D. Anxolabe ´here. 1993. Distribution and conser-
vation of sequences homologous to the 1731 retrotransposon
in Drosophila. Mol. Biol. Evol. 10:791–803.
National Center for Supercomputing Applications, 1991. NCSA
GelReader for the Macintosh. Version 2.0.5. University of
Illinois at Urbana-Champaign, Urbana-Champaign.
Nuzhdin, S. V. 1999. Sure facts, speculations, and open questions
about the evolution of transposable element copy number.
O’Hare, K., and G. M. Rubin. 1983. Structures of P transposable
elements and their sites of insertion and excision in the
Drosophila melanogaster genome. Cell 34:25–35.
Pasyukova, E. G., S. V. Nuzhdin, and D. A. Filatov. 1998. The
relationship between rate of transposition and transposable
element copy number for copia and Doc transposable
elements of Drosophila melanogaster. Genet. Res. 72:1–11.
Petrov, D. A., E. R. Lozovskaya, and D. L. Hartl. 1996. High
intrinsic rate of DNA loss in Drosophila. Nature 384:346–
Pimpinelli, S., M. Berlocco, L. Fanti, P. Dimitri, S. Bonaccorsi,
E. Marchetti, R. Caizzi, C. Caggese, and M. Gatti. 1995.
Transposable elements are stable structural components of
Drosophila melanogaster heterochromatin. Proc. Natl. Acad.
Sci. USA 92:3804–3808.
Potter, S., M. Truett, N. Phillips, and A. Maher. 1980. Eukaryotic
transposable genetic elements with inverted terminal repeats.
Shepherd, B. M., and D. J. Finnegan. 1984. Structure of circular
copies of the 412 transposable element present in Drosophila
melanogaster tissue culture cells and isolation of a free 412
long terminal repeat. J. Mol. Biol. 180:21–40.
Shevelyov, Y. Y., M. D. Balakireva, and V. A. Gvozdev. 1989.
Heterochromatic regions in different Drosophila melano-
gaster stocks contain similar arrangements of moderate re-
peats with inserted copia-like elements (MDG1). Chromosoma
Spradling, A. C., and G. M. Rubin. 1981. Drosophila genome
organization: conserved and dynamic aspects. Annu. Rev.
Stacey, S. N., R. A. Lansman, H. W. Brock, and T. A. Grigliatti.
1986. Distribution and conservation of mobile elements in the
genus Drosophila. Mol. Biol. Evol. 3:522–534.
Terrinoni, A., C. Di Franco, P. Dimitri, and N. Junakovic. 1997.
Intragenomic distribution and stability of transposable ele-
ments in euchromatin and heterochromatin of Drosophila
melanogaster: non-LTR retrotransposon. J. Mol. Evol. 45:
Vaury, C., A. Bucheton, and A. Pellison. 1989. The b-
heterochromatin sequences flanking the I elements are
themselves defective transposable elements. Chromosoma
Vaury, C., M. C. Chaboissier, M. E. Drake, O. Lajoinie, B.
Dastugue, and A. Pe ´lisson. 1994. The doc transposable
element in D. melanogaster and D. simulans: genomic
distribution and transcription. Genetica 93:117–124.
von Sternberg, R. M., G. E. Novick, G. P. Gao, and R. J. Herrera.
1992. Genome canalization: the coevolution of transposable
and interspersed repetitive elements with single copy DNA.
Diethard Tautz, Associate Editor
Accepted November 6, 2002
Noncanonical Retrotransposons and Heterochromatin409