Báo cáo khoa học: Effect of monovalent cations and G-quadruplex structures on the outcome of intramolecular homologous recombination doc

Effect of monovalent cations and G-quadruplex structures
on the outcome of intramolecular homologous
recombination
´
´
´
Paula Barros*, Francisco Boan*, Miguel G. Blanco and Jaime Gomez-Marquez
´
´
´
Departamento de Bioquımica e Bioloxıa Molecular, Facultade de Bioloxıa-CIBUS, Universidade de Santiago de Compostela, Spain

Keywords
G-quadruplex; minisatellite MsH43;
monovalent cations; recombination fidelity;
repetitive sequences
Correspondence
´
´
J. Gomez-Marquez, Departamento de
´
´
Bioquımica e Bioloxıa Molecular, Facultade
´
de Bioloxıa-CIBUS, Universidade de
Santiago de Compostela, 15782 Santiago de
Compostela, Spain
Fax: +34 9815969054
Tel: +34 981563100 (ext. 16937)
E-mail:
*These authors contributed equally to this

work
(Received 17 February 2009, revised 18
March 2009, accepted 20 March 2009)
doi:10.1111/j.1742-4658.2009.07013.x

Homologous recombination is a very important cellular process, as it provides a major pathway for the repair of DNA double-strand breaks. This
complex process is affected by many factors within cells. Here, we have
studied the effect of monovalent cations (K+, Na+, and NH4+) on the
outcome of recombination events, as their presence affects the biochemical
activities of the proteins involved in recombination as well as the structure
of DNA. For this purpose, we used an in vitro recombination system that
includes a protein nuclear extract, as a source of recombination machinery,
and two plasmids as substrates for intramolecular homologous recombination, each with two copies of different alleles of the human minisatellite
MsH43. We found that the presence of monovalent cations induced a
decrease in the recombination frequency, accompanied by an increase in
the fidelity of the recombination. Moreover, there is an emerging consensus
that secondary structures of DNA have the potential to induce genomic
instability. Therefore, we analyzed the effect of the sequences capable of
forming G-quadruplex on the production of recombinant molecules, taking
advantage of the capacity of some MsH43 alleles to generate these kinds of
structure in the presence of K+. We observed that the MsH43 recombinants containing duplications, generated in the presence of K+, did not
include the repeats located towards the 5¢-side of the G-quadruplex motif,
suggesting that this structure may be involved in the recombination events
leading to duplications. Our results provide new insights into the molecular
mechanisms underlying the recombination of repetitive sequences.

The integrity of chromosomal material is dependent
upon the efficient repair of DNA double-strand breaks
(DSBs), which arise during DNA replication or are
caused by exogenous agents. Without such systems,

unrepaired breaks can lead to chromosomal translocations, loss of transcriptional control, and promotion of
tumorigenesis [1]. In human cells, the repair of DSBs
can take place through two independent systems:
homologous recombination (HR), and nonhomologous
end-joining. HR is promoted by several enzymes of the
RAD52 epistasis group, which includes RAD51, the
human homolog of Escherichia coli RecA [2,3]. This

protein promotes the key homologous pairing and
strand-exchange reactions leading to the formation of
interlinked recombination intermediates [4].
Changes in ionic strength alter the behavior of some
enzymes involved in the HR process. In this regard,
previous work has shown that high salt concentrations
provoke conformational changes in the RAD51 protein [5] favoring the coaggregation of RAD51–ssDNA
nucleoprotein filaments with duplex DNA, stimulating
the recombination [6]. More recently, a study defining
the effect of salt on human RAD51 activities was
reported [7]. However, as far as we know, none of the

Abbreviations
DSB, double-strand break; HR, homologous recombination.

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

2983


Effect of cations and G-quartets on recombination

P. Barros et al.

studies on the influence of salts on the recombination
process have analyzed how monovalent cations could
affect the fidelity of recombination, defined as the percentages of equal and unequal recombinant molecules,
and the frequency of recombination.
In our laboratory, we have developed an in vitro system with which to analyze HR and nonhomologous
end-joining [8–11]. This system allows us to establish
in vitro the recombinogenic capacity of any DNA
sequence, as well as to determine the nature of the
recombinant molecules generated. In the present study,
we employed this in vitro system to analyze the effect
of the monovalent cations K+, Na+ and NH4+ on
recombination. For this purpose, we used the minisatellite MsH43, a human DNA sequence composed of
pentamers and hexamers organized in a tandem array
[12,13]. The organization in tandem of small repeat
units provides a good substrate with which to study
the frequency of equal and unequal crossovers, as it
facilitates perfect and nonperfect pairings. We found
that the presence of monovalent cations led to a higher
proportion of equal recombinants, and hence an
increase in the fidelity of the recombination events in
the experimental system employed.
On the other hand, it is well known that guaninerich nucleic acids (DNA and RNA) are capable of
forming four-stranded structures named G-quadruplexes (also known as G-tetrads, G4s, or G-quartets)
[14,15]. These structures are further stabilized by the
presence of a monovalent cation (especially K+) in the
center of the tetrad [15,16]. G-quadruplex-forming
sequences have been identified in eukaryotic telomers,
as well as in gene promoters, recombination sites, and

DNA tandem repeats [15]. Whether or not genomic
G-rich structures can form quadruplex-based structures
in vivo remains to be fully demonstrated, although supportive data are starting to emerge [14,15,17,18].
G-quadruplexes have long been hypothesized to play
roles in DNA recombination. Thus, G-quadruplex
DNA might play a role in class switch recombination
in the immunoglobulin genes [19,20], and studies
in yeast suggest possible roles for G-quadruplex DNA
in homologous recombination during meiosis [21]. In
relation to this, Hop1 not only binds to and catalyzes
the formation G-quadruplex DNA in vitro, but also
promotes the pairing of dsDNA molecules via quadruplex structures [22]. However, is not yet clear how the
in vitro activities of this and other proteins on G-quadruplex DNA relate to their in vivo functions.
In the present work, we also analyzed the effect of
the presence of sequences capable of forming G-quadruplex structures on recombination frequency and the
generation of recombinant molecules, taking advantage
2984

of the capacity of the minisatellite MsH43 to form this
kind of structure in the presence of K+ [12]. We found
that the presence of G-quadruplex did not alter the
recombination frequency as compared with the allele
control. Moreover, the great majority of recombinants
containing duplications generated in the presence of
K+ did not include the repeats located at the 5¢-side
of the G-quadruplex motif of MsH43, suggesting that
this structure is involved in the recombination events
leading to duplications. A model to explain this finding, involving replication slippage, is also shown.

Results and discussion

To study the effects of salts on the frequencies of equal
and unequal recombinant products generated in the
recombination experiments, we employed the minisatellite MsH43, as it shows two useful features: (a) an
organization in tandem, which allows the existence of
different types of homologous pairings (in register or
not in register), leading to the formation of equal and
unequal recombinant molecules; and (b) the ability of
allele 80.1 to form a G-quadruplex, as it contains the
motif (TGGGGC)4, which is a G-quadruplex-forming
structure, and the inability of allele 73.1 to form such
a structure, because it contains the motif TGGGGC
repeated only three times instead of four [12]; this differential characteristic allows analysis of the effect of
the presence of G-quadruplex structures on the generation of recombinant molecules in the in vitro system
employed in this work.
To carry out the recombination analyses, we
employed an in vitro system designed to detect intramolecular homologous recombination events [8–11].
We constructed two pBR322-based plasmids, the
recombinant substrates p73.1 and p80.1, bearing two
copies cloned in the same orientation as the corresponding MsH43 allele, 73.1 or 80.1. The map of these
recombinant substrates is shown in Fig. 1A. As the
lacZ gene is situated between the two copies of the
MsH43 inserts, the recombination that takes place
after pairing of the MsH43 homologous sequences
leads to the excision of the lacZ gene from the original
substrate, generating two kind of recombinants, equal
and unequal (Fig. 1B). In the equal recombinants, the
minisatellite remains unaltered, whereas in the unequal
ones, the minisatellite displays size variations caused
by unequal pairings or alterations during the recombination process.
In the recombination experiments, each plasmid

substrate was incubated with the nuclear extract under
standard conditions [8]. After incubation, DNA was
extracted and used to transform bacteria. The recombi-

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS


P. Barros et al.

Effect of cations and G-quartets on recombination

Recombination substrates
A
Sc E P02.1

P02.2 E S

E

P02.1

E

E
MsH43 73.1

MsH43 80.1

p73.1

p80.1

ori

ori
MsH43 73.1
E

B
MsH43

MsH43 80.1

E

Pairing of
homologous
sequences

Rec.
substrate

E

LacZ

Crossover
MsH43
Recombinant
product

(LacZ–)

X

ori
MsH43

Fig. 1. Map of the recombination substrates
and representation of an intramolecular
homologous recombination event. (A) Plasmids p73.1 and p80.1 (not drawn to scale)
were used as substrates in the recombination assays. They contain a replication origin
(ori), an ampicillin resistance gene (Amp),
the lacZ gene, and two identical copies of
the MsH43 sequence (wide black arrows),
cloned in the same orientation. At the top is
shown a scheme of the MsH43 inserts,
minisatellite (open box), and flanking
sequences; thin arrows mark the position of
primers P02.1 and P02.2. E, EcoRI; Sc,
SacII; S, SalI. (B) The intramolecular homologous recombination generates two kinds of
plasmid: equals, in which the MsH43
remains unaltered, and unequals, with alterations in the minisatellite sequence.

P02.2 E

MsH43 80.1

MsH43 73.1

nant plasmids generated lacZ) bacteria (white colonies), and the original plasmids generated lacZ+ bacteria (blue colonies). The recombinant plasmids were

first analyzed by restriction with EcoRI. The digestion
of p73.1 recombinant products yielded two restriction
fragments, one with the minisatellite sequence, and the
digestion of the p73.1 original plasmid generated five
restriction fragments, two of them identical (those containing the minisatellite) (Fig. 2A). If the recombinant
was unequal, then the EcoRI fragment containing the
minisatellite showed size variations (asterisks in
Fig. 2A). In the case of the p80.1, the digestion with
EcoRI of the original plasmid yielded four DNA fragments, whereas the digestion of its recombinant products generated only one fragment (Fig. 2B). To
facilitate the identification of p80.1 recombinants, the
recombinant plasmids were amplified by PCR with the
primers P02.1 and P02.2 [12]. The analysis of amplifications products allowed differentiation of the majority
of the equal and unequal events (Fig. 2C). However,
when the variation affected few repeats, it was difficult
to distinguish length variations of the minisatellite. To
solve this problem, heteroduplex analyses [13] were
carried out by mixing the amplification products of
each recombinant with the PCR product obtained
from the MsH43 sequence present in the original construct. Figure 2D shows the result of a heteroduplex
assay employing the recombinants generated in the
experiments with p80.1. The generation of heteroduplex molecules denotes the presence of an unequal

Amp

ori

Equal
recombinant
ori

Unequal
recombinant

recombinant, whereas the absence of heteroduplex
molecules means that it is an equal recombinant. The
sequencing of 14 equal recombinants corroborated that
the recombinant molecules classified as equal conserved the original minisatellite sequence (data not
shown).
The results of the intramolecular homologous
recombination experiments are summarized in
Table 1. The data were collected from three independent experiments for each assay condition (standard
or supplemented with salts). The LacZ+ colonies
were produced by the transformation with the original plasmids, whereas the LacZ) colonies were the
result of transforming bacteria with the recombinant
plasmids. The lacZ) colonies due to mutations in the
lacZ gene made up less than 1% of the total lacZ)
colonies, and the frequency of the lacZ) colonies in
transformations with the original substrate plasmids
not exposed to the nuclear extract or with heat-inactivated extract (15 min, 100 °C) was about 2 · 10)5.
Under standard conditions, both the recombination
frequencies and the frequency of equal and unequal
recombinants were very similar with plasmids p73.1
and p80.1. Noteworthy, with both plasmids, the
recombination events that generated unequal recombinants were more abundant ( 20%) than those that
maintained the original sequence of MsH43. To verify
the effect of the presence of G-quadruplex-forming
sequences in this in vitro system, we performed several
assays in the presence of 20 mm K+, added to stabi-

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

2985


1

2

3

4 5

6

7

p80.1

2

3

4

5

6

7

3

4

5

6

7

8

9 10 11

**

*

*

D
1

2

3

4

5

6

7

8

M

p80.1

2

*

*

C
1

9 10 11 12 13 14 15

*

*

1

P. Barros et al.

* * *

*

*

B

8

8

p80.1

A

p73.1

Effect of cations and G-quartets on recombination

9

12 13 14 15 16 17 18 19 20 21 22 23

*

*
**

*

*

*

9 10 11 12 13 14 15 16 17 18
Heteroduplex
molecules

lize the G-quadruplex structure. Once again, the
recombination frequencies were very similar with both
plasmids, indicating that the capacity of the MsH43
2986

Fig. 2. Analysis of the recombinant products generated in the in vitro recombination
assays. (A) Analysis in 1.5% agarose gels of
the EcoRI digest of the recombinant plasmids obtained from lacZ) colonies in experiments carried out with p73.1. The EcoRI
restriction pattern of the original plasmid
p73.1 (lane p73.1) and the recombinant
products (lanes 1–15) are shown. The arrow
indicates the DNA fragment that contains
the original MsH43 sequence, and asterisks
mark the DNA fragment containing alterations in the size of MsH43 (unequal recombinants). (B) Analysis in 1.5% agarose gels of
the EcoRI digest of the plasmids obtained
from lacZ) colonies in recombination experiments carried out with p80.1. The EcoRI
restriction patterns corresponding to the
original plasmid p80.1 (lane p80.1) and the
recombinant products (lanes 1–9) are
shown. (C) Analysis in 2% agarose gels of
the amplification products of recombinant

plasmids obtained in experiments carried
out with p80.1. The amplification product of
the original plasmid p80.1 (lane p80.1) and
the recombinant products (lanes 1–23) are
shown. The arrow indicates the DNA fragment that contains the original MsH43
sequence, and asterisks mark unequal
recombinants. (D) Heteroduplex analysis in
a 5% polyacrylamide gel (the presence of
heteroduplex molecules denotes the presence of an unequal recombinant); lane p80.1
is shown as a control of no heteroduplex
formation. The arrow indicates the DNA
fragment that contains the original MsH43
sequence. M, 100 bp ladder (Promega).

80.1 allele to form G-quadruplex does not influence
the recombination frequency, at least in our in vitro
system. However, the presence of K+ caused a strong

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS


P. Barros et al.

Effect of cations and G-quartets on recombination

Table 1. Quantitative analysis of recombination experiments. The recombinant frequencies are given as the ratio between the number of
LacZ) colonies and the total number of colonies obtained in each assay. The equal and unequal recombinant frequencies are given as the
ratio between the total number of each type of recombinant and the total number of recombinants. For columns 4, 6 and 8, mean ± standard deviation of the data is provided.
Equal recombinants
Recombination

substrate

Assay
conditions

Colonies
(LacZ+ ⁄ LacZ))

p73.1

Standard

3863 ⁄ 42
4034 ⁄ 45
4060 ⁄ 50

p80.1

Standard

7832 ⁄ 80
6095 ⁄ 68
6866 ⁄ 74

p73.1

20 mM KCl

8296 ⁄ 39
9183 ⁄ 47

7850 ⁄ 35

p80.1

20 mM KCl

9450 ⁄ 39
12 365 ⁄ 47
9617 ⁄ 44

p80.1

20 mM NaCl

19 559 ⁄ 29
17 360 ⁄ 27
16 261 ⁄ 21

p80.1

20 mM NH4Cl

8114 ⁄ 23
8555 ⁄ 25
7962 ⁄ 22

1.087
1.116
1.232
1.145

1.021
1.116
1.078
1.072
0.470
0.512
0.448
0.476
0.413
0.380
0.458
0.417
0.148
0.156
0.129
0.144
0.283
0.292
0.276
0.284

reduction (near 50%) in the recombination frequencies, as well as an important decrease in the nuclease
activity of the nuclear extract (data not shown). As
the initiation of homologous recombination is mediated by a nuclease activity that introduces DNA
DSBs [23], it is possible that the reduction in recombination frequency was due to inhibition of the nuclease activity by K+.
Remarkably, in the presence of K+, the proportions
of equal and unequal recombinants were inverted with
respect to the results obtained under standard conditions; that is, the equal recombinants were more abundant ( 25%) than the unequal ones (Table 1). Was
this inversion produced specifically by K+? The recombination assays carried out in the presence of Na+ or
NH4+, maintaining the same ionic strength, showed a

marked reduction of the recombination frequency with
respect to the standard conditions, more pronounced
than with K+, and also a predominance of equal recombinants (Table 1). The finding that Na+ and NH4+
caused a greater decrease in the nuclease activity of the
nuclear extract than K+ (data not shown) provides a
coherent explanation of the observation that the

Unequal recombinants

No.

Frequency (%)

No.

Frequency (%)

16
21
19

Recombination
frequency (%)

0.414
0.521
0.468
0.468
0.447
0.459

0.422
0.442
0.289
0.305
0.293
0.296
0.265
0.283
0.322
0.290
0.118
0.121
0.098
0.112
0.222
0.245
0.314
0.260

26
24
31

0.673
0.595
0.764
0.677
0.574
0.657
0.656

0.629
0.181
0.207
0.155
0.181
0.148
0.097
0.136
0.127
0.030
0.035
0.031
0.032
0.061
0.047
0.062
0.057

± 0.077
35
28
29
± 0.048
24
28
23
± 0.033
25
35
31

± 0.039
23
21
16
± 0.014
18
21
17
± 0.008

± 0.054
45
40
45
± 0.018
15
19
12
± 0.008
14
12
13
± 0.029
6
6
5
± 0.013
5
4
5

± 0.048

± 0.085

± 0.048

± 0.026

± 0.027

± 0.003

± 0.008

recombination frequencies obtained with those cations
were lower than with K+.
Cations play essential roles in nucleic acid and protein structure, stability, folding, and catalysis. By
means of their interactions with DNA and proteins,
they could play an important role in recombination.
For instance, changes in K+ concentration could alter
chromatin structure by taking advantage of the unique
sensitivity of quadruplex formation to K+ and other
cations present in the cells [24,25]. On the other hand,
in vitro studies with G-rich telomeric DNA sequences
and the minisatellite MsH43 have shown that they can
form quadruplex structures whose stability is sensitive
to changes in the concentrations of important physiological cations such K+ [12,16]. As mentioned earlier,
RAD51, a key enzyme in HR, is affected by salts. Our
results provide strong evidence that the presence of
monovalent cations causes a strong decrease in recombination frequency, probably due to inhibition of the

nuclease activity that produces DSBs on the plasmid
substrates, and leads to enhancement of the fidelity of
recombination, as the proportion of equal recombinants was higher. The presence of monovalent cations

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

2987


Effect of cations and G-quartets on recombination

P. Barros et al.

in the nucleus of the cells is an important physiological
requirement, and our results suggest that monovalent
cations also influence genomic stability through their
participation in the recombination process. This
increase in fidelity could be related to alterations in the
structure of the minisatellite and to conformational
changes in proteins involved in recombination. In this
regard, it has been reported that high salt induces conformational changes in RAD51, leading to the formation of interlinked recombination intermediates [2] that
are essential for the correct progression of the recombination process. It is worth noting that, with the
in vitro system developed in our laboratory, we did not
observe an influence of the capacity to generate
G-quadruplex DNA on the recombinogenic frequency
of the minisatellite MsH43.
A

The existence of unequal recombinants allowed a
search for the sites where the rearrangement in the

MsH43 occurred. The analysis of 163 recombinant
sequences (Figs 3 and 4) revealed that duplications
occurred less frequently (30%) than deletions, similar
to what was reported for the minisatellite CEB1
[26,27], suggesting that this type of repetitive sequence
is prone to undergoing deletions in the recombination
process. Most of the unequal recombinants involve
simple deletions (Fig. 3), and only four recombinants
derived from p80.1 (S5, S11, K10 and K13 in Fig. 3A)
showed double deletions, suggesting that they did not
arise by a simple recombination event. With regard to
the MsH43 expansions, they seem to be the consequence of simple direct duplications (Fig. 4), except in
one case derived from p73.1, where one repeat was
B

Fig. 3. Organization of MsH43 recombinants containing deletions. (A) Sequence array of the deleted molecules obtained in experiments with
p80.1: standard conditions (S1–S18), with K+ (K1–K21), with Na+ (Na1–Na9), and with NH4+ (N1–N4); asterisks indicate recombinants with
double deletions. The sequence of the MsH43 80.1 allele is shown. (B) Sequence array of the deleted molecules obtained in experiments
with p73.1: standard conditions (S1–S17) and with K+ (K1–K19). The sequence of the MsH43 73.1 allele is shown. The discontinuity in the
sequence indicates the deleted fragment. Minisatellite repeats are depicted by a color code shown at the bottom.

2988

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS


P. Barros et al.

A

MsH43 80.1
allele

S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
S11
S12
S13
K1
K2
K3
K4
K5
K6
K7
K8
K9
K10
K11
K12
K13
K14

Effect of cations and G-quartets on recombination

B
MsH43 73.1
allele

S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
S11
S12
S13
S14
S15
K1
K2
K3
K4
K5
K6
K7
K8

K9

Na1
Na2
Na3
Na4
Na5
Na6
Na7
N1
N2
N3
N4
N5
N6
N7

K10
K11
K12
K13
K14
K15
K16
K17
K18

K19

Fig. 4. Organization of MsH43 recombinants containing duplications. (A) Sequence array of recombinants with duplications obtained in

experiments with p80.1: standard conditions (S1–S13), K+ (K1–K14), Na+ (Na1–Na7), and NH4+ (N1–N7). The sequence of the MsH43 80.1
allele is shown. (B) Sequence array of recombinants presenting duplications obtained in experiments with p73.1 under standard conditions
(S1–S15) and with K+ (K1–K19). The sequence of the MsH43 73.1 allele is shown. The marks in the recombinant S1 denote the mutations
with respect to the original MsH43 allele. The arrowhead indicates the recombinant that has a repeat intercalated between the duplicated
arrays. Arrows indicate the duplicated fragment.

intercalated between the duplicated fragments (S5 in
Fig. 4B). None of the recombinants analyzed showed
truncated repeats. This feature was also observed in
the recombinants generated by the human minisatellite
MsH42 [8,9] and by the human minisatellite CEB1
inserted in yeast [26,27], indicating that the reorganizations produced in the minisatellite MsH43 arose by a
homology-guided mechanism. Interestingly, in the
presence of K+, the p80.1 recombinants displaying
duplications did not include the repeats located at the
5¢-side of the G-quadruplex motif (K1–K14 in
Fig. 4A). In contrast, this limitation was not found in
the assays carried out either under standard conditions
or in the presence of Na+ or NH4+, or with p73.1
(Fig. 4). It is tempting to speculate that the G-quadruplex motif (TGGGGC)4 influences the resolution of

the recombination events, leading to duplications in
the minisatellite sequence.
There is an emerging consensus that secondary
structures of DNA have the potential to induce genomic instability. The role of nonlinear DNA in replication, recombination and transcription has become
evident in recent years. Several studies have predicted
and characterized regulatory elements at the sequence
level. However, little is known about the role of DNA
structures as regulatory motifs. Cells use these structural motifs as signals for processes such as gene regulation or recombination in both prokaryotes and
eukaryotes [28,29]. The coincidence of breakpoints of

gross deletions with non-B DNA conformations has
led to the conclusion that these structures can trigger
genomic rearrangements through recombination ⁄ repair

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

2989


Effect of cations and G-quartets on recombination

P. Barros et al.

activities [30]. Furthermore, G-quadruplex secondary
structures can induce genetic rearrangements and promote RecA-independent homologous recombination
[31]. Genome-wide predictions have shown an abundance of G-quadruplex DNA motifs in the genomes of
Homo sapiens [32] and E. coli [33]. In both species, the
distribution of G-quadruplex structures seems to be
nonrandom and linked to regulatory regions of the
genome. The important finding is that this kind of
structure may play a role in genome dynamics at three
levels: regulation of transcription, recombination and
mutation hotspots in vivo, and blocking the progression of DNA polymerases [34,35].
Our results demonstrate that a minisatellite sequence,
which is not included inside a gene [12], can form Gquadruplex structures that interfere with DNA synthesis
and influence the resolution of recombination. It is
tempting to speculate that this type of repetitive
sequence could be involved in processes related to genome stability. In this regard, although the mechanisms
involved in minisatellite instability are poorly understood, some relevant factors have already been found,
such as the requirement for DSBs [23] and length and

sequence heterozygosity [36]. Furthermore, size alterations of G-rich minisatellites can be caused by the ability
of these sequences to adopt G-quadruplex structures
[37] and by their capacity to undergo slippage during
replication or unequal crossovers [8,9,38]. In the case of
MsH43, we observed that the recombinants containing
duplications, generated in the presence of K+, did not
include the repeats located at the 5¢-side of the G-quadruplex motif of MsH43, suggesting that this structure is
involved in the recombination events leading to duplications. In Fig. 5, we show a hypothetical model to
explain the mechanism involved in the generation of
duplications in the presence of a G-quadruplex. According to this, the generation of duplications is explained
by replication slippage on the strand of new synthesis.
The presence of a G-quadruplex structure stabilized by
K+ in the slippage loop would interfere with the replication at the 5¢-end of the G-quadruplex motif and consequently with the generation of duplications containing
this 5¢-region of MsH43. This effect is not observed if
the slippage occurs either on the 3¢-side of the G-quadruplex DNA structure or in the template strand; in the
latter case, the slippage would produce the observed
deletions. This model explains the generation of all
duplications derived from the experiments with p80.1 in
the presence of K+, with the exception of K1, which
contains the G-quadruplex motif in the duplicated
sequence (Fig. 4A). Possible explanations for
the generation of the recombinant K1 could be that not
all p80.1 plasmid molecules present in the recombina2990

A

5´
´

(TGGGGC)4

´
5´
DSB

B
C

5´
3´
5´

5´
3´

5´
3´

G4 structure

5´

Slippage loop
5´

3´

Slippage loop
5´

Fig. 5. Recombination model involving G-quadruplex (G4) structure

and slippage processes for the generation of recombinant molecules containing duplications and deletions. (A) Generation of a
DSB by the nuclease(s) of the nuclear extract initiates the recombination process; the sequence (TGGGGC)4 represents the G-quadruplex motif in the MsH43 80.1 allele. (B) After the break, there is a
strand invasion of the homologous sequence in the second copy of
the minisatellite included in the recombinant plasmid employed in
the assay. (C) Replication slippage may occur in the strand of new
synthesis (left side). If the slippage loop comprises the region containing the G-quadruplex motif, a G-quadruplex structure could be
generated that would be stabilized by the presence of K+; this secondary structure would interfere with DNA synthesis, avoiding the
formation of duplications involving the 5¢-side of the G-quadruplex
motif. When the slippage loop is located at the 3¢-side of the
G-quadruplex DNA motif [right side of (C)], duplications can be generated without the interference of G-quadruplex structures. According to this model, it is worth noting that the slippage in the
template strand, leading to deletions in the MsH43 sequence,
would not be affected by the generation of G-quartets, as it does
not have the G-quadruplex DNA motif.

tion assay formed G-quadruplex, or that this was
unstable. In both cases, the slippage process would
not be interfered with by the G-quadruplex structure,
making possible the inclusion of the sequence
(TGGGGC)4 in some recombinant molecules. Supporting this reasoning are the results found in dimethyl
sulfate methylation protection assays carried out with
oligonucleotides designed from the sequence of several
MsH43 alleles [12]. In these experiments, even at concentrations of 100 mm K+, there is a residual amount of
oligonucleotides that do not form G-quadruplex.
The present work shows that the presence of monovalent cations increases the fidelity of recombination,
and that this effect is independent of the presence of
G-quadruplex structures in the minisatellite MsH43.
However, the G-quadruplex structure seems to be a
barrier to the events leading to duplications, perhaps
leading to blockage of polymerases at that point.
Therefore, the G-quadruplex would not be a stimulus

for recombination but a source of genomic instability.

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS


P. Barros et al.

It should be noted that the results obtained with
MsH43 cannot be applied to any DNA sequence, as
the repetitive nature of MsH43 favors the existence of
unequal crossovers as well as slippage processes.
Perhaps one of the functions of repetitive DNA in the
genomes is to serve as instability spots that are necessary for genome evolution.
Finally, the results presented here show that the
in vitro system used in this study may be useful for
investigation of the mechanisms involved in recombination and DNA instability, as well for the analysis of
how monovalent cations affect the proteins implicated
in this fundamental biological process.

Experimental procedures
Recombination substrates
The alleles of MsH43 used in this study, 73.1 and 80.1,
were obtained by amplification of human genomic DNA,
with the primers P02.1 and P02.2 [10]. The PCR products
were cloned in the pGEM-T Easy vector (Promega, Madison, WI, USA). The plasmid containing the 73.1 allele was
digested with SacII–SalI, generating a 569 bp fragment,
and the plasmid containing the 80.1 allele was digested with
EcoRI, producing a 586 bp fragment. To generate the
recombination substrates, plasmids p73.1 and p80.1, two
identical copies of each fragment were cloned in pBR322,

in the same orientation, flanking the lacZ gene (Fig. 1A).

In vitro recombination assays
We have previously shown that the recombination products are generated by the nuclear extract and not by the
repair machinery of bacteria [7]. The standard recombination reactions were performed in a final volume of 100 lL
containing 20 mm Tris ⁄ HCl (pH 7.5), 10 mm MgSO4,
1 mm ATP, 0.1 mm each dNTP, 1 lg of plasmid (p73.1 or
p80.1), and 10 lg of rat testis nuclear extract [6]. In the
experiments containing KCl, NaCl, or NH4Cl, the salts
were added to a final concentration of 20 mm. Several
concentrations of salts were assayed (5, 10, 15, 20, and
25 mm), and a concentration of 20 mm was chosen for the
recombination experiments, because it produces a good
number of white colonies, allowing the screening of many
recombinants per assay (data not shown). Concentrations
higher than 20 mm produced few white colonies, probably
because of the decrease caused in the recombination frequency, which could be due to inhibition of nuclease
activity. The inhibition of nuclease activity was observed
by the analysis of plasmid integrity, after incubation with
the nuclear extract in the presence of the different monovalent cation concentrations, by electrophoresis on agarose
gels. After incubation of recombination substrates with the

Effect of cations and G-quartets on recombination

nuclear extracts for 30 min at 37 °C, DNA was phenolextracted, ethanol-precipitated, and used to transform
E. coli DH5a cells. Bacteria were plated onto LB agar
plates containing Blue-O-Gal (BRL, Gaithersburg, MD,
USA) at 0.3 gỈL)1 as lacZ gene indicator. We observed
that lacZ) colonies due to mutations in the lacZ gene
made up less than 1% of the total lacZ) colonies. The frequency of the lacZ) colonies in transformations with the

original substrate plasmids not exposed to the nuclear
extract or with heat-inactivated extract (15 min, 100 °C)
was about 2 · 10)5. The white colonies (recombinant
products) were used for minipreparation of plasmid DNA
and aliquots of  300 ng were digested with EcoRI, and
analyzed in agarose gels.

PCR, DNA sequencing, and heteroduplex analysis
The PCR reactions were performed in 25 lL containing
PCR buffer [67 mm Tris ⁄ HCl, pH 8.8, 16 mm (NH4)2SO4,
0.01% Tween-20], 0.1 ng of plasmid, 0.3 lm each primer
(P02.1 and P02.2), 0.2 mm dNTPs, 1.5 mm MgCl2, and
0.5 U of Taq polymerase. Cycling conditions were 29 cycles
of 95 °C for 1 min, 56 °C for 30 s, and 72 °C for 40 s, and
a final cycle with an extension of 5 min. When the PCR
products were used for direct cycle sequencing employing
the dGTP BigDye Terminator v3.0 Sequencing kit (Applied
Biosystems, Foster City, CA, USA), they were treated with
exonuclease I and alkaline phosphatase (Exo ⁄ Sap-It) (USB,
Cleveland, OH, USA). After this treatment, the PCR products were cycle sequenced by 25 cycles of 96 °C for 10 s
and 68 °C for 2 min in a PTC-200 thermocycler (MJ
Research, Ramsey, MN, USA), and purified by ethanol
precipitation. The sequencing products were analyzed using
the 377 DNA Automated Sequencer (Applied Biosystems).
For the heteroduplex analysis, aliquots of 5 lL of the PCR
products obtained from the recombinants and from the original recombination substrates were mixed at 95 °C for
3 min, and slowly cooled to room temperature. The heteroduplex molecules were detected by electrophoresis in 5%
polyacrylamide gels (29 : 1) at a constant voltage of 140 V
for 6 h using 1· TBE buffer (0.09 m Tris ⁄ borate, 0.002 m
EDTA) and visualized by ethidium bromide staining.

Acknowledgements
This work was supported by the Spanish Ministerio de
´
Educacion y Ciencia (BFU2006-06708) and by the
Xunta de Galicia (PGIDT07PX12001099R).

References
1 van Gent DC, Hoeijmakers JH & Kanaar R (2001)
Chromosomal stability and the DNA double-stranded
break connection. Nat Rev Genet 2, 196–206.

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

2991


Effect of cations and G-quartets on recombination

P. Barros et al.

2 West SC (2003) Molecular views of recombination proteins and their control. Nat Rev Mol Cell Biol 4,
435–445.
3 Benson FE, Stasiak A & West SC (1994) Purification
and characterization of the human RAD51 protein, an
analogue of E. coli RecA. EMBO J 13, 5764–5771.
4 Baumann P, Benson FE & West SC (1996) Human
RAD51 protein promotes ATP-dependent homologous
pairing and strand transfer reactions in vitro. Cell 87,
757–766.

5 Sigurdsson S, Trujillo K, Song B, Stratton S & Sung P
(2001) Basis for avid homologous DNA strand
exchange by human RAD51 and RPA. J Biol Chem
276, 8798–8806.
6 Liu Y, Stasiak AZ, Masson J-Y, Mcllwraith MJ,
Stasiak A & West SC (2004) Conformational changes
modulate the activity of human RAD51 protein. J Mol
Biol 337, 817–827.
7 Shim K-S, Schmutte C, Yoder K & Fishel R (2006)
Defining the salt effect on human RAD51 activities.
DNA Repair 5, 718–730.
´
´
´
8 Boan F, Rodrı´ guez JM & Gomez-Marquez J (1998) A
non-hypervariable human minisatellite strongly stimulates in vitro intramolecular homologous recombination.
J Mol Biol 278, 499–505.
´
9 Boan F, Rodrı´ guez JM, Mourino S, Blanco MG, Vinas
˜
˜
´
´
´
A, Sanchez L & Gomez-Marquez J (2002) Recombination analysis of the human minisatellite MsH42 suggests
the existence of two distinct pathways for initiation and
resolution of recombination at MsH42 in rat testes
nuclear extracts. Biochemistry 41, 2166–2176.
´
´

10 Blanco MG, Boan F, Barros P, Castano JG & Gomez˜
´
Marquez J (2005) Generation of DNA double-strand
breaks by two independent enzymatic activities in
nuclear extracts. J Mol Biol 351, 995–1006.
´
´
´
11 Boan F, Blanco MG, Barros P & Gomez-Marquez J
(2006) DNA end-joining driven by microhomologies
catalyzed by nuclear extracts. Biol Chem 387, 263–267.
´
´
12 Boan F, Blanco MG, Barros P, Gonzalez AI &
´
´
Gomez-Marquez J (2004) Inhibition of DNA synthesis
by K+-stabilised G-quadruplex promotes allelic
preferential amplification. FEBS Lett 571, 112–118.
´
´
´
13 Barros P, Blanco MG, Boan F & Gomez-Marquez J
(2005) Heteroduplex analysis of minisatellite variability.
Electrophoresis 26, 4304–4309.
14 Burge S, Parkinson GN, Hazel P, Todd AK & Neidle S
(2006) Quadruplex DNA: sequence, topology and structure. Nucleic Acids Res 34, 5402–5415.
15 Qin Y & Hurley LH (2008) Structures, folding patterns,
and functions of intramolecular DNA G-quadruplexes
found in eukaryotic promoter regions. Biochimie 90,

1149–1171.
16 Sen D & Gilbert W (1990) A sodium–potassium switch
in the formation of four-stranded G4-DNA. Nature
344, 410–414.

2992

17 Siddiqui-Jain A, Grand CL, Bearss DJ & Hurley LH
(2002) Direct evidence for a G-quadruplex in a promoter region and its targeting with a small molecule to
repress c-MYC transcription. Proc Natl Acad Sci USA
99, 11593–11598.
18 Duquette ML, Handa P, Vincent JA, Taylor AF &
Maizels N (2004) Intracellular transcription of G-rich
DNAs induces formation of G-loops, novel structures
containing G4 DNA. Genes Dev 18, 1618–1629.
19 Dempsey LA, Sun H, Hanakahi A & Maizels N (1999)
G4 DNA binding by LR1 and its subunits, Nucleolin
and hnRNP D, a role for G–G pairing in immunoglobulin switch recombination. J Biol Chem 274, 1066–1071.
20 Maizels N (2005) Immunoglobulin gene diversification.
Annu Rev Genet 39, 23–46.
21 Johnson JE, Smith JS, Kozak ML & Brad Johnson F
(2008) In vivo veritas: using yeast to probe the biological
functions of G-quadruplexes. Biochimie 90, 1250–1263.
22 Muniyappa K, Anuradha S & Byers B (2000) Yeast
meiosis-specific protein Hop1 binds to G4 DNA and
promotes its formation. Mol Cell Biol 20, 1361–1369.
23 Keeney S & Neale MJ (2006) Initiation of meiotic
recombination by formation of DNA double-strand
breaks: mechanism and regulation. Biochem Soc Trans
34, 523–525.

24 Hardin CC, Corregan MJ, Lieberman DV & Brown BA
(1997) Allosteric interactions between DNA strands and
monovalent cations in DNA quadruplex assembly:
thermodynamic evidence for three linked association
pathways. Biochemistry 36, 15428–15450.
25 Schultze P, Hud NV, Smith FW & Feigon J (1999) The
effect of sodium, potassium and ammonium ions on the
conformation of the dimeric quadruplex formed by
the Oxytricha nova telomere repeat oligonucleotide
d(G4T4G4). Nucleic Acids Res 27, 3018–3028.
`
26 Debrauwere H, Buard J, Tessier J, Aubert D, Vergnaud
G & Nicolas A (1999) Meiotic instability of human
minisatellite CEB1 in yeast requires DNA double-strand
breaks. Nat Genet 23, 367–371.
27 Lopes J, Ribeyre C & Nicolas A (2006) Complex minisatellite rearrangements generated in the total or partial
absence of Rad27 ⁄ hFEN1 activity occur in a single generation and are Rad51 and Rad52 dependent. Mol Cell
Biol 26, 6675–6689.
28 Hatfield GW & Benham CJ (2002) DNA topology-mediated control of global gene expression in Escherichia
coli. Annu Rev Genet 36, 175–203.
29 Bacolla A & Wells RD (2004) Non-B DNA conformations, genomic rearrangements, and human disease.
J Biol Chem 279, 47411–47414.
30 Bacolla A, Jaworski A, Larson JE, Jakupciak JP,
Chuzhanova N, Abeysinghe SS, O’Connell CD, Cooper
DN & Wells RD (2004) Breakpoints of gross deletions
coincide with non-B DNA conformations. Proc Natl
Acad Sci USA 101, 14162–14167.

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

P. Barros et al.

31 Shukla AK & Roy KB (2006) RecA-independent
homologous recombination induced by a putative
fold-back tetraplex DNA. Biol Chem 387, 251–256.
32 Huppert JL & Balasubramanian S (2005) Prevalence of
quadruplexes in the human genome. Nucleic Acids Res
33, 2908–2916.
33 Rawal P, Kummarasetti VBR, Ravindran J, Kumar N,
Halder K, Sharma R, Mukerji M, Das SK & Chowdhury
S (2009) Genome-wide prediction of G4 DNA as
regulatory motifs: role in Escherichia coli global
regulation. Genome Res 16, 644–655.
34 Arthanari H & Bolton PH (2001) Functional and
dysfunctional roles of quadruplex DNA in cells. Chem
Biol 8, 221–230.
35 Woodford KJ, Howell RM & Usdin K (1994) A novel
K+ dependent DNA synthesis arrest site in a

Effect of cations and G-quartets on recombination

commonly occurring sequence motif in eukaryotes.
J Biol Chem 269, 27029–27035.
36 Jauert PA & Kirkpatrick DT (2005) Length and
sequence heterozygosity differentially affect HRAS1
minisatellite stability during meiosis in yeast. Genetics
170, 601–612.
37 Nakagama H, Higuchi K, Tanaka E, Tsuchiya N,
Nakashima K, Katahira M & Fukuda H (2006)

Molecular mechanisms for maintenance of G-rich short
tandem repeats capable of adopting G4 DNA
structures. Mutat Res 598, 120–131.
´
´
38 Boan F, Gonzalez AI, Rodrı´ guez JM &
´
´
Gomez-Marquez J (1997) Molecular characterization of
a new human minisatellite that is able to form
single-stranded loops in vitro and is recognized by
nuclear proteins. FEBS Lett 418, 251–257.

FEBS Journal 276 (2009) 2983–2993 ª 2009 The Authors Journal compilation ª 2009 FEBS

2993