Microbiology (2003), 149, 1297–1310
DOI 10.1099/mic.0.26121-0
Detection and analysis of transpositionally active
head-to-tail dimers in three additional Escherichia
coli IS elements
Ildikó Szeverényi,3 Zita Nagy, Tibor Farkas, Ferenc Olasz and János Kiss
Environmental Biosafety Research Institute, Agricultural Biotechnology Centre, Szent-Györgyi
Albert str. 4, H-2101 Gödöllő, Hungary
Correspondence
János Kiss
[email protected]
Ferenc Olasz
[email protected]
Received 12 November 2002
Revised 30 January 2003
Accepted 4 February 2003
This study demonstrates that Escherichia coli insertion elements IS3, IS150 and IS186 are able to
form transpositionally active head-to-tail dimers which show similar structure and transpositional
activity to the dimers of IS2, IS21 and IS30. These structures arise by joining of the left and right
inverted repeats (IRs) of two elements. The resulting junction includes a spacer region (SR) of a few
base pairs derived from the flanking sequence of one of the reacting IRs. Head-to-tail dimers of IS3,
IS150 and IS186 are unstable due to their transpositional activity. They can be resolved in two ways
that seem to form a general rule for those elements reported to form dimers. One way is a sitespecific process (dimer dissolution) which is accompanied by the loss of one IS copy along with the
SR. The other is ‘classical’ transposition where the joined ends integrate into the target DNA. In
intramolecular transposition this often gives rise to deletion formation, whereas in intermolecular
transposition it gives rise to replicon fusion. The results presented for IS3, IS150 and IS186 are in
accordance with the IS dimer model, which is in turn consistent with models based on covalently
closed minicircles.
INTRODUCTION
Mobile DNA elements are widespread in living organisms
and they play an important role in the reorganization and
evolution of the host genome. Transposons and insertion
sequences (ISs) form a diverse group of mobile DNA elements and accordingly, their transposition is explained by
several, sometimes controversial models (for review see
Galas & Chandler, 1989). There are basically three possibilities for transposition and they differ in the reactions
involving the 39 and 59 strand cutting and transfer carried
out by the transposase. Different timing of these processes
can lead to different intermediates which can transpose in a
conservative manner (cut-and-paste model) or a replicative
manner (Mu-type replicative transposition model). The
third way often reported recently is the formation of a
single-strand bridge between the ends of the element. This
results in a ‘figure-of-eight’ intermediate, which can be
resolved into covalently joined inverted repeat (IR) ends
(for review see Chandler & Mahillon, 2002).
Ildikó Szeverényi and Zita Nagy contributed equally to this work.
3Present address: Temasek Life Sciences Laboratory, 1 Research
Link, The NUS, Singapore 117604.
Abbreviations: Ap, ampicillin; Cm, chloramphenicol; DDS, dimer
dissolution; Km, kanamycin; IR, inverted repeat; IRL, left inverted repeat;
IRR, right inverted repeat; IS, insertion sequence; Rif, rifampicin; SR,
spacer region; SSD, site-specific dimerization.
0002-6121 G 2003 SGM
Several recent findings indicate that covalently joined IRs
separated by a short spacer region (SR) are key structures in
the transposition of some ISs and this junction acts as
a transposition intermediate for the element. Such junctions have been identified in IS dimers consisting of two
directly repeated elements and in minicircles formed by
circularization of a single element. IS dimers have been
isolated from IS2 (Szeverényi et al., 1996a), IS21 (Reimmann
& Haas, 1987, 1990; Reimmann et al., 1989), IS30 (Dalrymple,
1987; Olasz et al., 1993, 1997), IS256 (Prudhomme et al.,
2002) and IS911 (Turlan et al., 2000), and covalently closed
circular IS elements have been described for IS1 (Sekine
et al., 1997a, Turlan & Chandler, 1995), IS2 (Lewis &
Grindley, 1997), IS3 (Sekine et al., 1994), IS30 (Kiss & Olasz,
1999), IS117 (Smokvina et al., 1994), IS150 (Welz, 1993;
Haas & Rak, 2002), IS256 (Prudhomme et al., 2002) and
IS911 (Polard et al., 1992, 1996). For some of these elements both circular and dimeric forms have been reported,
suggesting a possible connection between the formation
and transposition of these structures, as described for IS30
(Kiss & Olasz, 1999).
The IS dimer model developed for Escherichia coli element
IS30 (Olasz et al., 1993) is based on reactive IR–IR junctions
and is able to explain all the transpositional reactions:
deletion, inversion, transpositional fusion (cointegration)
and simple insertion (Fig. 1). The key structure of this
model is the IS dimer and is produced by site-specific
dimerization (SSD) from any replicons carrying two IS
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Printed in Great Britain
1297
I. Szeverényi and others
validity might also be extended to additional elements
(e.g. IS3, IS150) for which only minicircles have been
detected so far.
In the present study we investigated the presence or absence
of the dimer in the transposition of three E. coli elements,
IS3, IS150 and IS186. These ISs were selected for the
following reasons. First of all their transpositional activity
was demonstrated in different E. coli strains by using the
pAW1326 trap vector (Szeverényi et al., 1996a). Moreover,
the transposition of two of these elements is fairly well
known. The existence of the IS3 dimer was predicted earlier
(Spielmann-Ryser et al., 1991) and the circular form was
actually detected (Sekine et al., 1994), although a linear
element was suspected as an intermediate in transposition
(Sekine et al., 1996). The linear form of IS3 was recently
shown to be the derivative of the circular one (Sekine et al.,
1999). The circular (Welz, 1993) and linear (Haas & Rak,
2002) form of IS150 was also detected, but without any
indication of the existence of the dimer. Although IS186
seems not to be related to the other two elements, its activity
and the fairly unknown transpositional mechanism give
reasons for its investigation. We report here that IS3,
IS150 and IS186 are also able to produce reactive IS
dimers, thereby extending the class of such IS elements to
eight members.
Fig. 1. A simplified version of the IS dimer model: formation
and transposition of the head-to-tail dimer and the monomeric
circle. Intramolecular transposition of the dimer leading to inversion or deletion formation is not indicated. Note that minicircles
may derive from each replicon containing at least one IS copy,
but are most frequently produced by DDS. IS elements are
shown as open boxes with blank and filled triangles representing
IRL and IRR, respectively. Thin line, donor replicon; thick line, target
replicon. The dashed line arrow indicates non-transpositional
events like homologous recombination or replication/repair, while
other arrows show transposase-dependent processes.
elements. The dimer can promote the transpositional
rearrangements mentioned above, but most frequently
segregates into a replicon carrying only a single IS copy
via a site-specific process, called dimer dissolution (DDS).
This model is consistent with those based on covalently
closed minicircles (Polard et al., 1992; Lewis & Grindley,
1997) and accounts for simple insertion on the basis of
clear analogy between minicircles and IS-dimer-containing
replicons (Kiss & Olasz, 1999). Similar to the dimer structures, the covalently joined IRs are also separated by a few
base pairs in the circular form of the IS elements. These
minicircles can be considered as intermediates of simple
insertion, as reported for IS1 (Sekine et al., 1997a; Turlan &
Chandler, 1995), IS2 (Lewis & Grindley, 1997), IS3 (Sekine
et al., 1994), IS150 (Welz, 1993; Haas & Rak, 2002), IS256
(Prudhomme et al., 2002) and IS911 (Ton-Hoang et al.,
1997, 1998). The synthetic IS dimer model seems to be
applicable for the elements that form both minicircles
and dimers (IS2, IS21, IS30, IS256 and IS911); however, its
1298
METHODS
Bacterial strains and media. Escherichia coli strains JM101
(Messing, 1979) and TG2 (Sambrook et al., 1989) were used for
transformation during cloning, while JM109 (Yanisch-Perron et al.,
1985) or TG2 (both recA2 strains) were used for the transposition
experiments. For the mating-out assay, strain C600 containing the
pOX38Km conjugative plasmid (Chandler & Galas, 1983) was used
as a donor and our rifampicin-resistant (RifR) derivative of the HB101
strain (Boyer & Roulland-Dussoix, 1969) was used as a recipient.
All strains were grown at 37 ˚C in liquid and solid Luria–Bertani
(LB) media supplemented as indicated with antibiotics at the following final concentrations: ampicillin (Ap), 150 mg ml21; chloramphenicol (Cm), 20 mg ml21; kanamycin (Km), 20 mg ml21; rifampicin
(Rif), 60 mg ml21.
Plasmids. Plasmid constructs were cloned into pBluescript KS+
(pKS) (Stratagene), pEMBL19 (Dente et al., 1983) or pJKI88 (Kiss
& Olasz, 1999), respectively. Full IS copies were derived from
pAW1326 : : IS plasmids (Szeverényi et al., 1996b). The cat cassette
encoding the CmR gene (chloramphenicol acetyltransferase gene of
Tn9) was obtained from pAW302 (Stalder & Arber, 1989). The
joined IRs for C1 and C4 constructs were PCR-amplified from
pAW1326 : : IS plasmid templates using the primer pairs oISZ17/
oISZ18 for IS3, oISZ23/oISZ24 for IS150 and oISZ19/oISZ20 for IS186
(see Table 1 for primer sequences). The PCR product for IS3 was cloned
as a 342 bp HindIII fragment containing the right (1054–1258 bp)
and the left (1–138 bp) ends; for IS150, a 560 bp EcoRV–TaqI fragment with joined right (1157–1443 bp) and left (1–274 bp) ends
was cloned; and for IS186, a 416 bp BamHI–SmaI fragment with
joined right (1148–1437 bp) and left (1–211 bp) ends was cloned. For
cloning of C3 constructs, the truncated left ends of the elements
were also amplified from the respective pAW1326 : : IS templates
(Szeverényi et al., 1996b) with the primer pairs IS3in/IS3Lout,
IS150in/IS150Lout and IS186in/IS186Lout or IS186in2/IS186Lout
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149
Head-to-tail dimers from three IS elements
Table 1. Oligonucleotide primers used for PCR and/or sequencing
Upper-case letters refer to the sequence corresponding to the binding sites listed in the third column,
while lower-case letters show additional bases on the 59 end of oligonucleotides containing restriction
sites and/or the bordering sequence of the IRs (in the case of ‘ISin’ oligonucleotides).
Name
Vector-specific primers
T3
T7
pUCfor
pUCrev
IS3-specific primers
oISZ17
oISZ18
IS3Lout
IS3in
IS150-specific primers
oISZ23
oISZ24
IS150Lout
IS150Lout2
IS150Rout
IS150in
IS186-specific primers
oISZ19
oISZ20
IS186Lout
IS186Rout2
IS186in
IS186in2
Sequence (59–39)
ACGCCAAGCGCGCAATTAACCCT
GGCCAGTGAGCGCGCGTAATACGA
CAGGGTTTTCCCAGTCACGAC
TCACACAGGAAACAGCTATGAC
pKS
pKS
pKS
pKS
GTCAAGCGTGGGGCACCGTAAC
AAGCGGCATAATCTGCGTGGAAGTA
gcgtcgaCAAGCTTCAGGGCTTCACTGCGA
gcgtcgacgagTGATCTTACCCAGCAATAGTG
IS3
IS3
IS3
IS3
CTGCACTCTGACCAGGGAACGC
GCAAATTCAGTAACATCGGTAACCCAC
ccgtcgacGAGTACTTTCACTTTTTCGTG
ggaatTCTGGATCAGCACTAACGCCTTTAG
cgggatccGAATACTACAACAGCAGAAGAATTAG
cccaagcttaggTGTACTGCACCCATTTTGTTGG
IS150
IS150
IS150
IS150
IS150
IS150
(1118–1139)
(963–937)
(577–557)
(244–219)
(1328–1353)
(1–22)
TGGCAAATTGAACTGGCTT
GCAACGTCATGGAGCTGAG
cctctagagtcgacGCCGTAAGCCAGCCCCAGAC
gCGGATCCGAAAAGAAGAACTAACTCG
ccggatccggctcgatccccCATAAGCG
ccggatccggctcaatccccCATAAGCG
IS186
IS186
IS186
IS186
IS186
IS186
(982–1000)
(266–248)
(210–191)
(1147–1172)
(1–8)
(1–8)
(for C3 and C39, respectively). The cloned fragments in C3 constructs contained 1–138 bp of IS3, 1–577 bp of IS150 and 1–210 bp
of IS186. For cloning the compound transposons in pKS, the cat
gene from pAW302 and the fragment of pAW1326 : : IS carrying
the full IS copy were used. The whole structures along with flanking
regions derived from pAW1326 : : IS were transferred into pJKI88 to
establish the low-copy-number version of plasmids carrying the
same transposons.
DNA procedures. Commonly used DNA techniques were per-
formed according to Sambrook et al. (1989), unless indicated otherwise. Restriction enzymes and T4 DNA ligase were purchased from
Fermentas, New England Biolabs and Roche, whereas Taq polymerase was obtained from Pharmacia. Gel electrophoresis was performed in 16 TBE buffer at room temperature. DNA fragments
were isolated from gel using the QIAquick Gel Extraction Kit
(Qiagen). DNA hybridization was accomplished using the DIG DNA
Labelling and Detection Kit (Roche) and Biodyne nylon membrane
(Pall). DNA amplifications were performed in a total volume of
25 ml using 2?5 ml 106 PCR buffer supplemented with 15 mM
MgCl2 (Pharmacia), 0?2 mM each dNTP, 1 mM both primers, 0?5
units Taq DNA polymerase and 10–100 ng template DNA. PCR
products were purified for cloning or sequencing with the QIAquick
PCR Purification Kit (Qiagen). DNA sequences were determined
by the chain termination method (Sanger et al., 1977), using the
Sequenase Version 2.0 sequencing kit from USB or the ABI Prism 310
and 3100 Genetic Analyser automated sequencer (Perkin Elmer). All
primers used for PCR or sequencing are listed in Table 1. Computer
analysis was performed with the GCG software package (Wisconsin
Package Version 9.1; Devereux et al., 1984).
http://mic.sgmjournals.org
Binding site (59–39)
(804–782)
(610–633)
(576–596)
(833–812)
(570–549)
(989–1013)
(144–122)
(1–21)
PCR amplification of right IR–left IR (IRR–IRL) junctions. For
detection of the joined IRs from the three IS elements, artificial
compound transposons were constructed from each element and the
plasmids carrying the transposons were introduced into JM109 cells.
Three transformants were grown in 3 ml LB+Km without selection
for the resistance marker of the transposon (CmR). Stationary-phase
cultures (0?1 ml) were diluted 30-fold with fresh media every 12 h
and grown at 37 ˚C. This step was repeated 10 times (each passage
represents about five generations). Plasmid DNA was isolated from
the 1st, 3rd, 5th, 7th and 10th passages. The three parallel samples
were pooled, RNase-treated and used as templates in PCR reactions performed with the following primers: oISZ18/IS3Lout (IS3),
IS150Rout/IS150Lout2 (IS150) and IS186Rout2/IS186Lout (IS186).
The cycling included 1 min 94 ˚C pre-denaturation; 30 cycles of
20 s at 94 ˚C, 1 min at 55 ˚C and 3 min at 72 ˚C; and finally 5 min
at 72 ˚C for terminal elongation. Plasmid samples (pAW1326 : : IS or
pAW1326) and genomic DNA purified from JM101 or JM109 strains
were also used as templates for the amplification of the abutted IRs
using the following primers: oISZ17/oISZ18 (IS3), oISZ23/oISZ24
(IS150) and oISZ19/oISZ20 (IS186). The 30 cycles included 20 s
denaturation at 94 ˚C, 1 min annealing (67 ˚C for IS3, 60 ˚C for
IS150 and 55 ˚C for IS186) and 1 (IS3, IS186) or 1?5 min elongation
(IS150) at 72 ˚C. A final elongation was performed for 5 min at 72 ˚C.
Detection of IS dimerization in C3 structures. C3 constructs
for IS3 and IS186 were introduced into JM109 cells and three transformant colonies were grown in 3 ml LB+Ap. Every 24 h, 30 ml
culture was diluted 100-fold with fresh medium and grown further.
This step was repeated 10 times (each passage represents about
seven generations). Plasmid DNA was purified from the 1st, 3rd,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
1299
I. Szeverényi and others
5th, 7th and 10th passages. The three parallel samples were pooled,
RNase-treated and used as templates in PCR amplifications with
the following primers: oISZ18/pUCrev (template: C3-IS3) and
IS186Rout2/pUCrev (template: C3-IS186). Conditions were 30 cycles
of 20 s at 94 ˚C, 30 s at 55 ˚C and 1?5 min at 72 ˚C.
with 0?9 % NaCl solution to remove antibiotics. Plates were incubated for 6 h at 37 ˚C, then bacteria were resuspended in 5 ml LB
and washed twice with 0?9 % NaCl before plating on LB+Km, Ap,
Rif. The frequency of transposition was calculated as a ratio of the
KmR ApR Rif R and the KmR Rif R transconjugants.
Stability test for C0–C4 structures. The test system used to
determine the activity of different IS-derived structures was composed of five ApR plasmids, each carrying a CmR gene. Plasmids
C0–C4 for IS3, IS150 and IS186, were transformed into the recA2
JM109 strain and three colonies from each transformation were
grown overnight in 3 ml LB+Ap. The cultures were subjected to
seven passages (except for C3 constructs, which were cultured
through 10 passages) under Ap selection as described in the previous
section. Plasmid DNA was isolated after each passage, then the 3rd,
7th and 10th (for C3) were used to transform TG2 cells. The transformants were tested for CmR/CmS phenotype by replica plating.
The frequency of transposition was calculated as a ratio of CmS
ApR/ApR colonies.
Analysis of DDS products. DNA samples of C5 dimers derived
from C4 constructs were digested with an appropriate restriction
endonuclease (IS3, StuI; IS150, BglII; IS186, NdeI), for which the
intact IS element has a unique cleavage site located outside of the
truncated part. JM109 cells were transformed with the digested
plasmid population and DNA was purified from at least 30 of the
resulting ApR colonies that were very likely to contain products of
DDS (C6). The structure of the plasmids was determined by restriction analysis and some of them were also sequenced.
Mating out assay. Plasmids C0–C5 were used to transform the
C600 donor strain containing the pOX38Km (Chandler & Galas,
1983) conjugative plasmid. The transformants were selected on
LB+Km, Ap, Cm (C1–C4) or LB+Km, Ap (C5) plates. Conjugation was carried out on solid LB medium by mixing 0?15 ml O/N
donor and 0?05 ml O/N Rif R HB101 recipient cultures both washed
RESULTS
IRR–IRL junctions are present in DNA
populations containing IS3, IS150 or IS186
As an initial step towards the detection of the dimeric
form of IS3, IS150 and IS186, we first demonstrated the
presence of joined IRs of the elements in bacterial populations. Using active copies entrapped by a trap vector
(Szeverényi et al., 1996b), three compound transposons
were constructed in a low-copy-number vector carrying
two identical IS elements in the same orientation separated
by a CmR gene. Plasmids containing the transposons were
introduced into the recA2 JM109 strain and grown through
10 passages (ca 50 generations) without selection for Cm.
Plasmid DNA was prepared from every second passage
and tested for the occurrence of joined IRs by PCR performed with primers directed outward of the elements.
IRR–IRL junctions could be detected for all the three
elements (Fig. 2).
The presence of IRR–IRL junctions was also tested by PCR
in different DNA samples from JM101 and JM109 cells. The
joined ends were successfully amplified from genomic
DNA and plasmid DNA isolated from cells harbouring the
Fig. 2. PCR-based detection of IRR–IRL junctions in plasmid DNA preparations containing compound transposons. E. coli
strain JM109 was transformed with a low-copy-number plasmid containing a compound transposon, including a CmR gene
flanked by two identical IS sequences in the same orientation. (Such constructs were generated for IS3, IS150 and IS186.)
Cultures were passaged 10 times without Cm selection. Plasmid DNA was prepared from every second passage and used as
template for PCR amplification with primers reading outwards from the IR ends. Panels show PCR products obtained from the
samples collected at the 1st (lanes p1) and 3rd (lanes p3) passages. Lanes M, molecular mass standard (l DNA digested
with PstI). Filled arrowheads point to the PCR products representing the joined IRs (IS3, 416 bp; IS150, 362 bp; IS186,
417 bp), while blank arrowheads indicate the fragments amplified from the original transposons. (Negatives of pictures taken
of gels stained with ethidium bromide are shown to increase visibility of weak fragments.)
1300
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149
Head-to-tail dimers from three IS elements
pAW1326 trap vector with or without an IS element
(pAW1326 : : IS or pAW1326), except in the case of the
pAW1326 : : IS3 template (data not shown). The positive
results obtained with pAW1326 suggested that plasmid
DNA samples contained joined IRs that were situated on
low-molecular-mass DNA species such as minicircles.
However, this experimental set-up did not allow us to
differentiate between dimers and minicircles.
Subsequent Southern hybridization and sequence analysis
of the PCR fragments confirmed the existence of the IRR–
IRL junctions for all three elements. Between the abutted
IRs, the so-called SR was generally identical with the target
duplications that bordered the three elements before
joining of the IRs, but some divergence in spacer length
was detected (see below).
All three IS elements produce head-to-tail
dimers
PCR amplification of the junction fragment of IS3 and
IS186 produced extremely weak products, indicating a
very small amount of dimer in the template DNA, whereas
IS150 resulted in a stronger signal, allowing the isolation
of plasmids carrying head-to-tail dimers (Fig. 2). Plasmid
DNA samples from the 10th passage of the IS150-based
transposon (Fig. 3a) were transformed into strain TG2 and
the transformants were tested by replica plating for the
presence or absence of the CmR marker of the transposon.
The frequency of the CmS segregants was 9?7±3?3 %.
Among 81 plasmid samples prepared from CmS colonies,
64 (79 %) carried a single IS150 copy (monomer), 14 (17 %)
represented three types of deletions and 3 (4 %) were a
mixed population of plasmids containing the head-to-tail
IS150 dimer, monomer and various types of deletions
(Fig. 3b, c, lanes 2). The three isolates containing dimers
were re-transformed into TG2 to separate the different
plasmid species. More than half of the samples (35/61)
extracted from these transformants contained a monomer
element (Fig. 3b, c, lanes 4), five cases were found to be
mixed populations similar to the original isolate (Fig. 3b, c,
lanes 3) and various kinds of deletions seem to have
occurred in the remaining cases. The failure to isolate the
dimer as a unique species suggested that it was highly
active. The fragments corresponding to the head-to-tail
dimer were isolated from the gel and sequenced. The results
confirmed that they carried IRR–IRL junctions with a 3 bp
SR, proving the capacity of IS150 for dimerization.
Our attempts to isolate IS3 and IS186 dimers by using a
similar approach were unsuccessful. The segregation frequency of their transposons was too low (<561024 and
<861024, respectively), when they were located on a
low-copy-number vector. On the other hand, the same
transposons inserted into a high-copy-number vector showed
extremely high instability, even in the case of IS150 (Fig. 4).
Although the PCR test showed the presence of the joined
IRs for all the three elements in the high-copy-number
plasmid DNAs (data not shown), after transformation and
http://mic.sgmjournals.org
replica plating the CmS segregants proved to be plasmids
with a single element (ca 95–100 %) or various deletions
(<5 %). This observation suggests that the dimer is
extremely unstable and quickly segregates into monomers
by excision of one IS element preventing the isolation of
plasmids still carrying the dimer, as predicted for (IS3)2
(Spielmann-Ryser et al., 1991). To decrease this activity,
truncated transposons (Fig. 5, C3 structure) were constructed using a full and an incomplete copy of IS3 and
IS186. These constructs were suitable for detecting
dimerization by PCR amplification using a primer pair
annealing to the full IS copy and to the vector sequence,
respectively (Fig. 6a). This experimental set-up is able to
distinguish between IS dimers and minicircles, as only
dimers and the original C3 constructs can serve as templates for the amplification and the products can easily
be distinguished according to their different sizes.
C3 constructs for IS3 and IS186 were introduced into
JM109 cells, respectively, and transformants were grown
through 10 passages (ca 70 generations) without selection
for the CmR marker of the transposons. Isolated plasmid
DNA from every second passage was tested by PCR for the
presence of the junction between the full and truncated
elements. The amplification confirmed the formation of
dimers that had arisen by site-specific deletion of the
intervening DNA in both C3 constructs (Fig. 6b). The
PCR products characteristic for the dimers were isolated,
cloned and sequenced. The segregation of C3 constructs
for IS3 and IS186 showed some differences. Although the
quantity of the dimers did not change significantly between the
passages, the amount of the original C3 construct clearly
decreased in the case of IS186, but not in that of IS3 (Fig. 6b,
open arrowheads). The same phenomenon was observed
when the plasmid preparations were transformed into TG2
and the transformants were tested for Cm resistance by
replica plating: the frequency of CmS transformants in the
10th passage was <8?661022 % for C3-IS3, but reached
almost 100 % for C3-IS186. This phenomenon may hint at
a replicative manner of dimer formation for IS3, while a
conservative manner seems more likely for IS186. It is worth
noting that both pathways are possible, as suggested for
IS911 (Polard et al., 1996) and IS30 (Kiss & Olasz, 1999).
Both dimers and monomers (C5 and C6, Fig. 5) were found
among CmS segregants of C3-IS3. The DNA isolates of
C5-IS3 dimers seemed to be uniform by restriction analysis,
but when they were reintroduced to TG2 cells, the resulting
transformants contained C5 dimers or C6 monomers, which
supported the idea of C5 being an active intermediate.
The isolation of the C5-IS186 dimer was somewhat more
problematic, because hundreds of CmS segregants turned
out to be C6-type monomers. Pre-screening of 192 CmS
transformants by colony PCR that amplifies dimers
identified a colony still containing the C5-IS186 structure
(demonstrated by sequencing). The plasmid DNA isolated
from this colony was a mixed population of the dimer
and numerous different segregation products, indicating
that C5-IS186 was indeed very unstable.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
1301
I. Szeverényi and others
Fig. 3. Analysis of segregation products of the IS150-based compound transposon. The transposon was introduced into
recA2 E. coli strain JM109 and cultured for 10 passages as described in the legend to Fig. 2. Plasmid DNA was isolated after
the 10th passage step and transformed into TG2 cells. Transformants were tested for the presence/absence of the CmR gene
by replica plating. (a) Schematic structures of the compound transposon (top), the dimer (middle) and the monomer (bottom),
with the characteristic restriction sites indicated. Abbreviations: B, BamHI; Bg, BglII; E, EcoRI; Xh, XhoI. IS150 copies are
indicated as open boxes with blank and filled triangles representing IRL and IRR, respectively. The CmR gene of the
transposon is indicated. (b and c) Restriction analysis of the compound transposon and its segregation products doubledigested with XhoI/BglII (b) or with EcoRI/BamHI (c). Lanes 1, original plasmid containing the compound transposon prior to
the passages; lanes 2, mixed plasmid population (monomer, dimer and deletion product) from a single CmS colony obtained
by transforming TG2 cells with the DNA isolated after the 10th passage (3 out of 81 colonies contained dimer; one example is
shown here); lanes 3 and 4, two DNA samples obtained from re-transformation of DNA shown in lanes 2; lanes 3, mixed
population containing dimer and its derivatives; lanes 4, stable and uniform monomer; lanes 5, molecular mass standard
(l DNA digested with PstI). Abbreviations: D, deletion product; V, vector; T, transposon; D, dimer; M, monomer; ML and
MR, left and right part of the monomer, respectively (the 0?46 kb BamHI–EcoRI fragment derived from the CmR gene is not
shown in c).
The IRR–IRL junction actively participates in
transposition
We developed a test system for comparing the stability
of IS-derived constructs containing different parts of IS
elements (Fig. 5, C0–C4). The system is suitable for the
1302
detection of deletions leading to the loss of the CmR
marker gene. The constructs carried the following structures: C0, CmR gene without any IS-derived sequence (control to detect spontaneous mutations); C1, CmR gene
and a PCR-amplified IRR–IRL junction (truncated dimer)
without functional transposase production; C2, CmR gene
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149
Head-to-tail dimers from three IS elements
and a single intact IS element (monomer); C3, an intact
and a truncated IS copy flanking the CmR gene on both
sides; and C4, the full IS element and the truncated dimer
flanking the CmR gene on both sides. These plasmids
were constructed for all three IS elements, introduced
into JM109 cells and the ratio of the plasmid population
that had lost the CmR phenotype was determined after the
3rd and 7th passages (Table 2).
Fig. 4. Segregation frequency of compound transposons
inserted into a high-copy-number vector. pBluescript KS plasmids containing compound transposons of IS3 (open bars),
IS150 (hatched bars) and IS186 (filled bars) were introduced
into JM109 cells and transformant colonies were grown without
selection for CmR. One-day-old cultures were 30-fold diluted
with fresh medium and grown for an additional 24 h. This passage step was repeated five times for IS150 and IS186, and
10 times for IS3. Plasmid DNA was isolated from each passage and the segregation frequency was determined by transforming into TG2 cells followed by replica plating of the
transformants. The frequency was calculated as the ratio of
CmS and ApR transformants. Each column represents data from
three parallel experiments.
Fig. 5. Schematic structure of the original plasmids (C1–C4)
and their main segregation products (C5–C6) in the test
system used for comparing the stability of IS-derived constructs. C0, control plasmid with CmR gene to determine the
rate of spontaneous mutation. The structure of these plasmids
was the same for IS3, IS150 and IS186. Symbols are defined
in the legend to Fig. 1.
http://mic.sgmjournals.org
The C0 control was stable and the frequency of CmS
segregants was undetectable. The truncated dimer (C1) was
stable for IS150 or nearly stable for IS186, but C1-IS3
produced CmS segregants in considerable proportion at
the 7th passage. The truncated dimer structures were presumably activated by complementation of genomic IS186
and IS3 copies. The C2 constructs carrying single elements
showed no or only low levels of segregation. C3-IS3 and
C3-IS150 were nearly stable during the first seven passages
and the CmS derivatives became detectable after three
further passages. C3-IS186 behaved quite differently, since
its segregation exceeded 95 % by the 7th passage. The C4
constructs showed a high level of activity in all three cases.
IS186 was again found to be the most active, 98 % of the
population was already CmS after the 3rd passage. These
results indicate the increased activity of the truncated dimer,
suggesting that the joined IR structure plays an important
role in transposition of these elements.
Detection of transposition activities
characteristic for IS dimers
While analysing the structure of plasmids obtained from
the CmS segregants of C4 constructs, a C5-type dimer,
containing an intact and a truncated IS element, was isolated (IS3, in 11 samples out of 45 analysed; IS150, 22/75;
IS186, 49/100). This product presumably arose by SSD or
by transposition of the IRR–IRL junction to the IR end of
the full IS copy, as shown for IS911 (Turlan et al., 2000)
and IS30 (Kiss & Olasz, 1999; Olasz et al., 1997). The C5
dimers were similar to those that arose from C3 and the
SR was identical with the sequence originally separating
the truncated IRs in C4 (Fig. 7). The truncated dimers
(C5) derived from C4 constructs (and also from C3) were
not the end products of the rearrangements. They could
be stabilized by the loss of the full IS element, giving rise
to plasmids containing the single truncated element (C6,
Fig. 5). To prove that C6 originated from the dimer, the
DNA of C5 isolates was reintroduced into strain TG2 and
the plasmid DNA of transformants was analysed. C6-type
structures were found among plasmid samples of all three
elements, supporting the idea that C6 arose from C5. These
results suggested a two-step transition C4RC5RC6, albeit
a C4RC6 reaction can also occur, since C6 structures
were already present in a high proportion among the
CmS segregants of passaged C4 constructs (IS3, 43–53 %;
IS150, 46–71 %; IS186, 10–83 %).
After demonstrating the intramolecular activity of these
dimers, the intermolecular transposition was investigated
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
1303
I. Szeverényi and others
Fig. 6. PCR-based detection of IS3 and IS186 dimers by using a truncated transposon as a template. PCR was carried out
using the primer annealing to the vector outside of the truncated IS copy (pUCrev) combined with that annealing to the intact
IS copies (oISZ18 for IS3 and IS186Rout2 for IS186, see Table 1). (a) Schematic representation of the templates. The
primers are indicated with arrows, with blank and filled triangles representing IRL and IRR, respectively. (b) Amplification
products for C3-IS3 (left) and C3-IS186 (right). Lanes p1–p10 refer to the passage steps from which the template plasmid
population was isolated. Filled arrowheads point to the PCR product representing the joined IRs (IS3, 470 bp; IS186,
464 bp), while blank arrowheads show the fragment amplified from the original C3 templates. Other bands may represent
different deletion products present in the template plasmid populations.
Table 2. Comparison of the stability of IS-derived constructs
Data were obtained by replica plating of colonies derived from transformation with plasmid DNA isolated from the 3rd and 7th (and 10th
for C3–IS3) passages of C0–C4 constructs. The frequency of segregation was calculated as the CmS/ApR colony ratio.
Proportion of CmS segregants (%)
Construct
IS3
IS150
IS186
p3
p7
p10
p3
p7
p10
p3
p7
C0
<0?05
<0?05
NE
<0?05
<0?05
NE
<0?05
<0?05
C1
0?7
24?0
NE
<0?05
<0?05
NE
0?04
0?08
C2
<0?05
0?2
NE
0?2
0?1
NE
<0?05
0?8
C3
<0?05
<0?05
0?09
<0?05
<0?05
1?0
5?0
98?0
C4
41?0
74?0
NE
49?0
87?0
NE
98?0
NE,
100
Not examined.
1304
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149
Head-to-tail dimers from three IS elements
Fig. 7. SSD of C3 and C4 structures for IS3, IS150 and IS186. (a and b) Schematic representation of the C3RC5 and
C4RC5 rearrangements. Sequences flanking the ends of the single and truncated IS element (a) or those flanking the IRR
end of the single element and sequences of the dimer SRs (b) are shown for the three IS elements. The asterisk in (a)
indicates that sequence data for IS150 originated from the full transposon instead of C3-IS150. C3-IS186 was constructed
in two slightly different versions: C39 contained A instead of G in the 27 position of the flanking region of the truncated
element. (c) The SRs identified in C5 structures derived from C3 and C4, respectively.
by a mating out assay with the conjugative plasmid
pOX38Km (Chandler & Galas, 1983). The frequency of
KmR ApR transconjugants gives us information about the
transposition activity of C0–C5 structures (Fig. 8). Both C4
and C5 for the three elements resulted in at least one order
of magnitude higher frequency of transconjugants than the
single IS (C2). Conjugal transmission of each C3 structure
also exceeded that of C2, indicating the ability of C3 to
produce a more active form, potentially a C5-like structure. The truncated IRR–IRL junction in C1 structures
showed considerable activity for all three ISs, indicating
possible complementation by the genomic copies. In the
case of IS186, the higher activity of C3 and C4 structures
suggest that C5-IS186 isolates may not be the most active
form of the intermediate. It is noteworthy that the relatively lower activity of C5-IS186 was at least one order of
magnitude higher than that of C2-IS186, suggesting that
the dimer represents a more active form than the single
element. Sequencing of the full IS186 copy of C5 excluded
http://mic.sgmjournals.org
the possibility that the reduced activity was caused by
mutations, although the length and sequence of the spacer
might play a role in this phenomenon.
Origin of the SR in IRR–IRL junctions
SRs separating the abutted IRs were analysed by comparing the sequence data of the IRR–IRL junctions collected
in the experiments described above. The SRs identified
seem to have been derived from the bordering sequence
of one of the two reacting IRs. They were generally 3 bp for
IS3 and IS150 and varied between 10 and 20 bp for IS186.
In many cases the length of SR was identical to that of the
target duplication generated by the IS element investigated.
In the case of IS3, the SR in C4-derived C5 structures was
GAG (Fig. 7c), corresponding to the triplet in the spacer of
the truncated dimer in C4. On the other hand, C3-derived
C5-IS3 structures had a CCA spacer (with one exception,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
1305
I. Szeverényi and others
have derived from the junction in C4, while the spacer of
dimers originating from the low-copy-number compound
transposon showed variability: one of them had an AAG
spacer, which flanked the IS elements in the transposon,
while the AGG spacer may represent an unusual dimerization process. Additional samples were sequenced from the
pAW1326 : : IS150 isolates. The majority of them contained
AGG, but we found AAG, GTA and ACC triplets, as well.
The AAG spacer corresponds to the target duplication
bordering the element in pAW1326 : : IS150 and GTA flanks
the IRL of the genomic IS150 copy (EMBL/GenBank
PRO02 : ECAE433), while AGG and ACC do not match
the bases bordering IS150 in pAW1326 : : IS150 or those
that delimit the chromosomal copy.
Three C5-IS186 isolates derived from the C4 structure all
contained the same 10 bp spacer, GGCTCGATCC, that
corresponded to the SR in the IRR–IRL junction and the
flanking sequences of the full IS186 copy on C4. To identify the origin of SR in C5 structures we constructed two
almost identical C3-IS186 plasmids, which differed only
in a single GRA change that was introduced to the
sequence adjacent to the truncated element (C39-IS186,
Fig. 7a). This enabled us to trace the origin of SR in the
junction. The amplified IRR–IRL junctions in the C3 and
C39 plasmid populations showed considerable diversity
(Fig. 7c). However, it was clear that the SRs were derived
from the sequences flanking either the left or the right IR
in C39 structures. Further junctions amplified from one
pAW1326 : : IS186 DNA template comprised the 10 bp
spacer GGCTCGATCC, which was identical with the target
duplication flanking the element in pAW1326 : : IS186.
SR is lost during DDS
Fig. 8. Analysis of the intermolecular transposition of C0–C5
constructs by mating them out with a conjugative plasmid. The
six types of constructs for all three IS elements (all with the
ApR gene) were transformed into bacteria containing the KmR
plasmid pOX38Km, then the fusion products of the two plasmids were conjugated to a recipient. The frequency of fusion
was calculated as the ratio of KmR ApR/KmR colonies. The C5
structures used were derived from C4 constructs. For the
structures, see Figs 5 and 7.
where it was a single C), which originally flanked the intact
IS element in C3. Further IRR–IRL junction sequences
were investigated which we found in JM101 or JM109 cells
containing the pAW1326 plasmid. We found two samples
with an AGT spacer, probably derived from the genomic
element IS3A (Umeda & Ohtsubo, 1989; EMBL/GenBank
PRO02 : ECAE137). Four other isolates had a GAG triplet
in between the IRs and one sample contained ATG. The
origin of these latter spacers remains unknown.
The SRs in three C5-IS150 isolates were AGG, which may
1306
The bordering sequence in C6 structures may provide
information on the mechanism of dissolution of C5 dimers,
therefore C6 was isolated and sequenced. The flanking
sequences in C6-IS3 (Fig. 9a) differed from the SR of
C5-IS3, indicating the loss of the spacer during DDS. This
corresponds to the results obtained for dissolution of the
(IS30)2 dimer (Olasz et al., 1993). However, the bordering
sequences in different C6-IS3 samples were not identical.
Only in one case (out of 14) was the single IR end flanked
by the original CCA target duplication. In the remaining
cases the IR was bordered by sequences generated from
the original flanking region of C5 by a short (1–5 bp)
deletion. For IS150, the bordering sequence in C6 was
identical to the flanking region of the intact IS element
in C5 (AAG) in two out of six isolates. In the remaining
cases an additional base (G), presumably originating
from the SR, was found right beside the IR end (Fig. 9b).
The sequence bordering the left end of the IS186 copy
in six isolates of the C5 structure was identical with that
of the SR; therefore, the origin of the sequence flanking the truncated IS copy in C6 derivatives could not
be determined.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149
Head-to-tail dimers from three IS elements
Fig. 9. Dissolution of two C5 dimers generates a set of C6 products. The sequence and occurrence of C6 products isolated
from C5-IS3 (a) and C5-IS150 (b) are shown. The three positions flanking the IRL are indicated in bold type; base G, which
could be derived from the SR, is underlined in (b). IS39 and IS1509 refer to the truncated elements.
DISCUSSION
In the present study we demonstrate that head-to-tail
dimers are formed by IS elements IS3, IS150 and IS186.
Both the reactive structure (the IR–IR junction) and the
specific reactions mediated by the joined ends resemble
those of the previously described dimers (IS2)2 (Szeverényi
et al., 1996a), (IS21)2 (Reimmann & Haas, 1987, 1990),
(IS30)2 (Dalrymple, 1987; Farkas et al., 1996; Olasz et al.,
1993) and (IS911)2 (Turlan et al., 2000). The characteristic
reaction of two directly oriented elements is the SSD
leading to the formation of a dimer structure (Figs 3 and
7). This intermediate can disintegrate by DDS, resulting
in the loss of one IS element (Fig. 9) or by promoting
transposition (deletion or fusion of replicons – Fig. 8).
These transposition reactions accompany the disassembly of the IRR–IRL junction and lead to more stable rearranged DNA species. We have demonstrated that plasmid
constructs carrying two directly oriented IS elements can
undergo SSD, then DDS reactions, which are illustrated by
the C3RC5RC6 transition. It has been shown that constructs carrying the joined IRs are at least one order of
magnitude more active in intermolecular transposition
than the single element, which suggests that the abutted
ends act as an intermediate for transposition. All of these
data support the notion that IS3, IS150 and IS186 have the
ability to translocate via an intermediate that is formed by
joining the IRs.
Although the C5 structure for both IS3 and IS150 was
the most active species in intermolecular transposition,
the C4-borne C5 structure of IS186 showed only as much
activity as the truncated dimer in C1 (Fig. 8), whereas
both C3-IS186 and C4-IS186 were more active than C5.
The fact that C1, C4 and C5 structures, all of which contain
http://mic.sgmjournals.org
an IRR–IRL junction, and C3, which has a potential to
form such junction, were more active than C2 containing
an IS186 monomer, clearly indicates that the IRR–IRL
junction represents an active transpositional intermediate.
The observations that (i) both C3 and C4 segregated into
C6 with high frequency, (ii) the C5-type molecules could
hardly be isolated among segregants of C3, (iii) the only
C3-borne C5 isolate was not uniform, and (iv) C5 structure
could not be recovered by re-transformation of this isolate
all suggest that IS186 may behave similarly to the other
two elements. The formation of C5-IS186 dimers results
in the appearance of a variety of structures differing in
their SR (Fig. 7c) and possibly in their transposition
activity. However, the only structure we could recover by
re-transformation was a C5 derivative carrying a 10 bp SR,
presumably because it was the least active.
Although formation and dissolution of the joined IRs
share similar basic characteristics, the different ISs show
important differences. One of these features is the difference in the formation of the spacer between the joined
IRs during SSD. In the dimers of IS3 or IS150 the length
of the spacer was generally the same as that of the target
duplication generated by the respective element, while IS186
produced IRR–IRL junctions with divergent spacer length
(Fig. 7c). The sequence of the SRs, where it was traceable,
always derived from the flanking regions of one or the
other IR taking part in dimerization, as has been reported
for other elements [IS1 (Sekine et al., 1997a, Turlan &
Chandler, 1995); IS3 (Sekine et al., 1994); IS30 (Olasz et al.,
1993); IS911 (Polard et al., 1992)]. The C4-derived C5
dimers of all three ISs contained the same SRs that separated the truncated IRs rather than the bases flanking the
intact element in C4, whereas C3-borne IRR–IRL junctions
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
1307
I. Szeverényi and others
behaved differently. In IS3 junctions the spacer always
derived from the target duplication bordering the full IS
copy in C3 and never from the bordering region of the
truncated one, while in IS186 junctions the SRs derived
either from the flanking sequence of the full or the truncated copies. The different origin of SRs in C3- or C4-borne
dimers may reflect the differences in the molecular processes that can occur when two IRs (C3) or an IRR–IRL
junction and a third IR (C4) take part in the dimerization.
It can be assumed that the length of the SR depends on
the site of the staggered cuts adjacent to the reacting IRs.
This cleavage site in the cases of IS30 (Olasz et al., 1997,
1998) and IS911 (Ton-Hoang et al., 1998; Turlan et al.,
2000) proved to be at the same distance from the end of
IRs both in dimerization or integration into the target site,
resulting in the same length of the spacer and of the target
duplication. On the other hand, the spacer in (IS2)2
(Szeverényi et al., 1996a) and in (IS21)2 (Reimmann & Haas,
1987) does not correspond to the target duplication
characteristic for these elements. In these cases the cleavage
site is different in SSD and integration into the target
DNA. These deviations can play a role in the regulation of
transposase gene expression, especially in the formation of
new hybrid promoters, as summarized for IS911 (TonHoang et al., 1997) and proposed for IS30 (Dalrymple,
1987). Although there are no obvious hybrid promoters
(235 promoter box in the IRR and correctly positioned
210 box in the IRL) in the joined IRs of IS3 (DuvalValentin et al., 2001), IS150 and IS186, all of these elements
can generate promoters by insertion, as shown by their
entrapment in pAW1326 (Szeverényi et al., 1996b).
Furthermore, the target specificity may also play a role in
the formation of spacers in IS186 junctions. The flanking
DNA of the truncated element in C3-IS186 contains
hotspot-like sequences for IS186 that differ only in one
position from the consensus G3N6C3 target site of the
element (Keller, 1996; SaiSree et al., 2001), which may
influence the dimerization reaction and consequently the
length of SR. The fact that different flanking sequences
of IS186 influenced the length of the resulting SR (Fig. 7c)
supports the role of flanking sequences in spacer formation via dimerization. It is worth noting that IS186 generates a variable length of target duplication via insertion as
well (Sengstag et al., 1986).
The ability of joining the left and right IRs of an element
accompanies the potential to form both dimers and minicircles. The primary isolates of IRR–IRL junctions (i.e. those
generated by PCR from the compound transposons and
pAW1326 or pAW1326 : : IS templates) could have originated
either from tandemly repeated or circular IS elements. The
fact that joined ends for IS3, IS150 and IS186 were also
detectable in bacteria containing pAW1326 without IS
elements suggests that several amplified IRR–IRL junctions
might have derived from minicircles of genomic IS copies.
Since the SR derives from the flanking sequences of the
original IS copy, this allows us to determine the ancestry
1308
of some IRR–IRL junctions isolated. When the junctions
were amplified from pAW1326 : : IS186 DNA templates,
the abutted IRs were separated by the GGCTCGATCC
sequence, which was identical with the direct repeat
bordering the element in pAW1326 : : IS186. This suggested
that the junction was formed by the inserted element via
dimerization or circularization. The same origin could be
assumed for the IRR–IRL junction of IS150 with the AAG
spacer, while GTA could derive from the genomic IS150
copy. Similarly, the abutted IRs of IS3 separated by GAG
must have derived from the IS3A element of E. coli.
Numerous SRs of unidentified origin were found in the
IRR–IRL junctions of IS3 and IS150, as was also reported
for IS1 and IS3 circles (Sekine et al., 1994, 1997b). This
indicates that at any time the bacterial population carries
transpositional intermediates deriving from new insertions, which may be present only in subclones of the cells.
The other explanation for this phenomenon may be the
ability of these IS elements to undergo multiple cycles of
rearrangements, e.g. successive cycles of circularization and
dimerization, which has been reported for IS30 (Kiss
& Olasz, 1999).
Our experiments provide evidence that DDS takes place
in IS3, IS150 and IS186 dimers similarly to those described
by Olasz et al. (1993) and Turlan et al. (2000). However, this
process also has some unique features in IS3 and IS150
(Fig. 9). In the dissolution of the C5-IS3 dimer, the
bases that directly flanked the IRL of the full element (the
original target duplication) were lost in many cases and
the bases positioned originally 2–5 bp upstream from
the IRL bordered the remaining IR end in the C6 product
(Fig. 9a). This phenomenon may provide an explanation
for the interesting fact that only 4 out of the 13 complete
IS3 copies available in the GenBank database are delimited
by the common 3 bp direct repeat. The others flanked by
different sequences presumably underwent a dimerization–
dissolution cycle, which could alter either of the bordering
sequences, similar to those presented in Fig. 9a. In C6-IS150
isolates the flanking bases were preserved, but an additional base – probably originating from the SR in C5 –
was detected. One may conclude that for some ISs the
bordering sequence of the remaining element is not affected
during DDS [IS2 (Szeverényi et al., 1996a); IS30 (Olasz et al.,
1993); IS911 (Turlan et al., 2000); IS186, this work], while
other elements, like IS3 and IS150, generate a few base
pair deletions or insertions, respectively, in the flanking
DNA (Fig. 9).
Our findings reported here extend the number of IS
elements, in which an active dimer-type intermediate was
detected, to eight (IS2, IS3, IS21, IS30, IS150, IS186, IS256
and IS911). These elements are dispersed in five IS families
[IS3, IS4, IS21, IS30 and IS256 (Mahillon & Chandler, 1998)].
However, additional experimental evidence is required to
determine the overall occurrence of this intermediate in
other IS elements, and further potential members may be
found among the elements for which minicircle formation
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149
Head-to-tail dimers from three IS elements
has already been reported [e.g. IS1 (Sekine et al., 1997a;
Turlan & Chandler, 1995)]. It has been suggested that
covalently closed IS circles and dimers could be alternative representations of the same transpositional pathway
(Lewis & Grindley, 1997; Kiss & Olasz, 1999). Recent studies
(Berger & Haas, 2001; Prudhomme et al., 2002; Turlan
et al., 2000) seem to further support this possibility.
ISs that have been reported to produce dimers and/or
minicircles are scattered over at least seven IS families:
IS1, IS3, IS4, IS21, IS30, IS110 and IS256 (Mahillon &
Chandler, 1998). However, the presence of covalently
attached transposon ends and their participation in the
integration of mobile DNA is not restricted to bacteria,
they can also be found in eukaryotic organisms, for example:
Tc1 (Rose & Snutch, 1984; Ruan & Emmons, 1984), copia
(Flavell & Ish-Horowicz, 1983) and the P element (Roiha
et al., 1988) of Drosophila; Mu1 transposon (Sundaresan &
Freeling, 1987) and Ac/Ds of maize (Gorbunova & Levy,
1997). The formation and transposition of abutted ends
of mobile elements may reflect another common hallmark
of these entities. It should be mentioned, however, that
several verified models exist which explain the transposition/integration processes differently and further investigations in this field will contribute to a deeper insight into
transposition processes.
Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive
set of sequence analysis programs for the VAX. Nucleic Acids Res 12,
387–395.
Duval-Valentin, G., Normand, C., Khemici, V., Marty, B. &
Chandler, M. (2001). Transient promoter formation: a new feed-
back mechanism for regulation of IS911 transposition. EMBO J 20,
5802–5811.
Farkas, T., Kiss, J. & Olasz, F. (1996). The construction and
characterization of an effective transpositional system based on IS30.
FEBS Lett 390, 53–58.
Flavell, A. J. & Ish-Horowicz, D. (1983). The origin of extra-
chromosomal circular copia elements. Cell 34, 415–419.
Galas, D. J. & Chandler, M. (1989). Bacterial insertion sequences.
In Mobile DNA, pp. 109–162. Edited by D. E. Berg & M. M. Howe.
Washington, DC: American Society for Microbiology.
Gorbunova, V. & Levy, A. A. (1997). Circularized Ac/Ds transposons:
formation, structure and fate. Genetics 145, 1161–1169.
Haas, M. & Rak, B. (2002). Escherichia coli insertion sequence IS150:
transposition via circular and linear intermediates. J Bacteriol 184,
5833–5841.
Keller, M. P. (1996). Regulation of transposition activity in the
prokaryotic insertion sequences IS30 and IS186. PhD thesis,
Philosophisch-Naturwissenschaftlichen Fakultät der Universität
Basel, Biozentrum der Universität Basel, Switzerland.
Kiss, J. & Olasz, F. (1999). Formation and transposition of the
covalently closed IS30 circle: the relation between tandem dimers
and monomeric circles. Mol Microbiol 34, 37–52.
Lewis, L. A. & Grindley, N. D. (1997). Two abundant intramolecular
transposition products, resulting from reactions initiated at a single
end, suggest that IS2 transposes by an unconventional pathway. Mol
Microbiol 25, 517–529.
ACKNOWLEDGEMENTS
This study was supported by grants T019365, T025729 and F038204
from the Hungarian National Scientific Research Fund (OTKA) and
NKFP4/006 from the Ministry of Education. We wish to thank László
Orbán for useful suggestions with the PCR amplifications and for
comments on the manuscript, and Mick Chandler for critical reading
of the manuscript. We are also grateful to Ágota Bakos and Ilona
Keresztúri for their skilful help with basic bacterial procedures as well
as Erika Keresztúri, Zsuzsanna Lengyel and Mária Turai for technical
assistance.
Mahillon, J. & Chandler, M. (1998). Insertion sequences. Microbiol
Mol Biol Rev 62, 725–774.
Messing, J. (1979). Multipurpose cloning system based on the
single-stranded DNA bacteriophage M13. Recomb DNA Tech Bull
NIH 2, 43–48.
Olasz, F., Stalder, R. & Arber, W. (1993). Formation of the tandem
repeat (IS30)2 and its role in IS30-mediated transpositional DNA
rearrangements. Mol Gen Genet 239, 177–187.
Olasz, F., Farkas, T., Kiss, J., Arini, A. & Arber, W. (1997). Terminal
inverted repeats of insertion sequence IS30 serve as targets for
transposition. J Bacteriol 179, 7551–7558.
REFERENCES
Berger, B. & Haas, D. (2001). Transposase and cointegrase:
Olasz, F., Kiss, J., König, P., Buzás, Z., Stalder, R. & Arber, W.
(1998). Target specificity of insertion element IS30. Mol Microbiol
specialized transposition proteins of bacterial insertion sequence
IS21 and related elements. Cell Mol Life Sci 58, 403–419.
Polard, P., Prere, M. F., Fayet, O. & Chandler, M. (1992).
Boyer, H. & Roulland-Dussoix, J. (1969). A complementation
analysis of the restriction and modification of DNA in E. coli. J Mol
Biol 41, 459–472.
28, 691–704.
Transposase-induced excision and circularization of the bacterial
insertion sequence IS911. EMBO J 11, 5079–5090.
Chandler, M. & Galas, D. J. (1983). Cointegrate formation mediated
Polard, P., Ton-Hoang, B., Haren, L., Betermier, M., Walczak, R. &
Chandler, M. (1996). IS911-mediated transpositional recombination
by Tn9. II. Activity of IS1 is modulated by external DNA sequences.
J Mol Biol 170, 61–91.
Prudhomme, M., Turlan, C., Claverys, J. P. & Chandler, M. (2002).
Chandler, M. & Mahillon, J. (2002). Insertion sequences revisited. In
Mobile DNA II, pp. 305–366. Edited by N. L. Craig, R. Craigie,
M. Gellert & A. Lambowitz. Washington, DC: American Society for
Microbiology.
Dalrymple, B. (1987). Novel rearrangements of IS30 carrying
plasmids leading to the reactivation of gene expression. Mol Gen
Genet 207, 413–420.
Dente, L., Cesareni, G. & Cortese, R. (1983). pEMBL: a new family
of single stranded plasmids. Nucleic Acids Res 11, 1645–1655.
http://mic.sgmjournals.org
in vitro. J Mol Biol 264, 68–81.
Diversity of Tn4001 transposition products: the flanking IS256
elements can form tandem dimers and IS circles. J Bacteriol 184,
433–443.
Reimmann, C. & Haas, D. (1987). Mode of replicon fusion mediated
by the duplicated insertion sequence IS21 in Escherichia coli. Genetics
115, 619–625.
Reimmann, C. & Haas, D. (1990). The istA gene of insertion
sequence IS21 is essential for cleavage at the inner 39 ends of
tandemly repeated IS21 elements in vitro. EMBO J 9, 4055–4063.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
1309
I. Szeverényi and others
Reimmann, C., Moore, R., Little, S., Savioz, A., Willetts, N. S. & Haas,
D. (1989). Genetic structure, function and regulation of the
Smokvina, T., Henderson, D. J., Melton, R. E., Brolle, D. F., Kieser, T.
& Hopwood, D. A. (1994). Transposition of IS117, the 2?5 kb
transposable element IS21. Mol Gen Genet 215, 416–424.
Streptomyces coelicolor A3(2) ‘minicircle’: roles of open reading
frames and origin of tandem insertions. Mol Microbiol 12, 459–468.
Roiha, H., Rubin, G. M. & O’Hare, K. (1988). P element insertions and
rearrangements at the singed locus of Drosophila melanogaster.
Genetics 119, 75–83.
Rose, A. M. & Snutch, T. P. (1984). Isolation of the closed circular
form of the transposable element Tc1 in Caenorhabditis elegans.
Nature 311, 485–486.
Spielmann-Ryser, J., Moser, M., Kast, P. & Weber, H. (1991). Factors
determining the frequency of plasmid cointegrate formation
mediated by insertion sequence IS3 from Escherichia coli. Mol Gen
Genet 226, 441–448.
Stalder, R. & Arber, W. (1989). Characterization of in vitro
Ruan, K. & Emmons, S. W. (1984). Extrachromosomal copies of
constructed IS30-flanked transposons. Gene 76, 187–193.
transposon Tc1 in the nematode Caenorhabditis elegans. Proc Natl
Acad Sci U S A 81, 4018–4022.
Sundaresan, V. & Freeling, M. (1987). An extrachromosomal form
SaiSree, L., Reddy, M. & Gowrishankar, J. (2001). IS186 insertion at
of the Mu transposons of maize. Proc Natl Acad Sci U S A 84,
4924–4928.
a hot spot in the lon promoter as a basis for lon protease deficiency
of Escherichia coli B: identification of a consensus target sequence for
IS186 transposition. J Bacteriol 183, 6943–6946.
Szeverényi, I., Bodoky, T. & Olasz, F. (1996a). Isolation, character-
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning:
Szeverényi, I., Hodel, A., Arber, W. & Olasz, F. (1996b). Vector for IS
a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold
Spring Harbor Laboratory.
element entrapment and functional characterization based on turning
on expression of distal promoterless genes. Gene 174, 103–110.
Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing
Ton-Hoang, B., Betermier, M., Polard, P. & Chandler, M. (1997).
with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74,
5463–5467.
Assembly of a strong promoter following IS911 circularization and
the role of circles in transposition. EMBO J 16, 3357–3371.
Sekine, Y., Eisaki, N. & Ohtsubo, E. (1994). Translational control in
Ton-Hoang, B., Polard, P. & Chandler, M. (1998). Efficient trans-
production of transposase and in transposition of insertion sequence
IS3. J Mol Biol 235, 1406–1420.
Turlan, C. & Chandler, M. (1995). IS1-mediated intramolecular
Sekine, Y., Eisaki, N. & Ohtsubo, E. (1996). Identification and
characterization of the linear IS3 molecules generated by staggered
breaks. J Biol Chem 271, 197–202.
ization and transposition of an (IS2)2 intermediate. Mol Gen Genet
251, 281–289.
position of IS911 circles in vitro. EMBO J 17, 1169–1181.
rearrangements: formation of excised transposon circles and replicative deletions. EMBO J 14, 5410–5421.
Turlan, C., Ton-Hoang, B. & Chandler, M. (2000). The role of
Isolation and characterization of IS1 circles. Gene 191, 183–190.
tandem IS dimers in IS911 transposition. Mol Microbiol 35,
1312–1325.
Sekine, Y., Izumi, K., Mizuno, T. & Ohtsubo, E. (1997b). Inhibition
Umeda, M. & Ohtsubo, E. (1989). Mapping of insertion elements
of transpositional recombination by OrfA and OrfB proteins
encoded by insertion sequence IS3. Genes Cells 2, 547–557.
IS1, IS2 and IS3 on the Escherichia coli K-12 chromosome. Role of
the insertion elements in formation of Hfrs and F9 factors and in
rearrangement of bacterial chromosomes. J Mol Biol 208, 601–614.
Sekine, Y., Eisaki, N., Kobayashi, K. & Ohtsubo, E. (1997a).
Sekine, Y., Aihara, K. & Ohtsubo, E. (1999). Linearization and
transposition of circular molecules of insertion sequence IS3. J Mol
Biol 294, 21–34.
Sengstag, C., Iida, S., Hiestand-Nauer, R. & Arber, W. (1986).
Terminal inverted repeats of prokaryotic transposable element IS186
which can generate duplications of variable length at an identical
target sequence. Gene 49, 153–156.
1310
Welz, C. (1993). Funktionelle Analyse des Bakteriellen Insertion-
selements IS150. PhD thesis, Fakultät für Biologie der AlbertLudwigs-Universität Freiburg, Germany.
Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13
phage cloning vectors and host strains: nucleotide sequences of the
M13mp18 and pUC19 vectors. Gene 33, 103–119.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 17:08:29
Microbiology 149