Palindromes and genomic stress fractures: Bracing and repairing the damage
Introduction
The last several years has seen an increased frequency of publications that implicate palindromes in diverse pathological contexts. Natural palindromes or near-palindromes of about 200–800 bp have been discovered to exist at sites of sporadic and recurrent chromosomal rearrangements in humans [1], [2], [3], [4], [5], [6]. There is evidence that large palindromes arise de novo in tumor cells in disease-specific chromosomal positions [7]. The possibility that palindromes initiate gene amplification has been discussed for a number of years [8] and is closer to being dissected in detail now that different steps have been reproduced in model systems ([9] and cited therein; [10]). An interesting observation made in yeast is that strains compromised for telomere maintenance escape from senescence by replacing their telomeres with long palindromes [11]. Possibly, this “PAL” mechanism of chromosome maintenance has a mammalian correlate in that mouse ES cells in which a telomere is removed infrequently acquire a long apparent palindrome at the site [12]. A provocative link between palindromes in mitochondrial DNA and senescence in fungi also exists [13]. This review will discuss how DNA palindromes are endowed with an ability to disrupt genome integrity, and what we know to date about mechanisms that curtail the potential for damage.
The word “ palindrome” is in common use but has been inconsistently applied to a spectrum of DNA repeat arrangements. Motifs termed palindromes range from small, perfectly or imperfectly matched inverted repeats to extensive stretches of inverted identity separated by kilobase-pair long spacers (diagrammed in Fig. 1). Because these various sequence configurations can have quite different biophysical and biological properties, it is important to stipulate how we define “palindrome” here. “Palindrome” will denote a DNA sequence that is immediately juxtaposed to an exact inverted (that is, reverse complementary) copy of itself. A palindrome has no central spacer and no mismatches between component repeats (Fig. 1).
There are sequences that deviate subtly from our strict definition of a palindrome, and that fall into a gray area, i.e. they are inverted repeats with very small spacers and/or small discrepancies between the two arms (Fig. 1). The biophysical behavior of these “near palindromes” can be similar to real DNA palindromes [14], but this is not easily predicted by sequence gazing. We will use the term “near-palindrome” as a convenient way to refer to these sequences and set them apart from other categories of inverted repeats. Thus (recognizing that this is something of a tautology) near-palindromes are functionally defined as sequences that are like palindromes in terms of biological and biophysical character, with some quantitative differences only.
The reason to take care with semantic distinctions is illustrated in Fig. 1. While there is overlap in the structural and biological possibilities presented by different classes of inverted repeats, non-palindromic and palindromic inverted repeats differ in one major respect. Spaced inverted repeats, though intrinsically self-complementary, are not able to self-pair unless an extensive stretch of single stranded DNA is first exposed. An example of this is diagrammed in Fig. 2, where strand separation is forced by replication. In contrast, palindromes and near-palindromes can buckle under stress into a self-paired “cruciform” structure while still essentially double-stranded. The helix opening that initiates the process can be caused by torsional strain and, in the case of a palindrome, only a limited number of base pairs have to melt before both “Watson” and “Crick” can form hairpins by self-annealing (reviewed in [15]). The extrusion process untwists the two strands, further extending the length of the hairpin arms so that overall, conversion to a four-way branch structure, or cruciform, relieves negative superhelical stress [16]. Hypernegative supercoiling provides both the push needed to initiate cruciform extrusion, and the force necessary to stabilize the four-way branch.
The ability of a DNA palindrome to convert from the lineform to a cruciform structure is a critical link between palindromy and DNA damage. It is known that palindromes can serve positive roles in transcription, replication and specialized developmental processes, and the jobs they do exploit their ability to attain an anomalous DNA structure. Domesticated palindromes are usually (though with exceptions) fairly short, and often work in conjunction with structure- and sequence-specific proteins [17], [18], [19], [20], [21], [22]. However palindromes can also appear at random sites as a result of replication errors, a chance integration event, or some type of illegitimate recombination. Where a palindrome of roughly 100 bp or more occurs, it creates a weak spot in the DNA that, quite literally, cannot take the strain. If a lineform DNA molecule with a palindromic sequence experiences torque, at some threshold level of hypernegative supercoiling, it will extrude into a cruciform structure.
Creation of a cruciform will not itself actually fracture DNA. When a cruciform appears, conditions could change so as to allow it to be resorbed without causing damage. However there is a consistent and growing body of evidence that with cruciform extrusion, there is an increased likelihood of DNA breakage at the site.
Some examples of this are provided by experiments in which palindromes or near-palindromes have been artificially inserted into a yeast genome. One indication of break induction was the creation of a recombination “hotspot” upon introduction a 160 bp palindrome into S. pombe genome [23]. In another study, site-specific DNA breaks were associated with two different 140 bp palindromes introduced at the His4 locus of S. cerevisiae. The breaks could be physically detected on Southern blots [24]. A key observation forged the link between palindromes, cruciforms, DNA breaks and a candidate enzymatic activity with the demonstration that a near-palindrome in yeast chromosome II became the site of a specific chromosomal break in which the resulting DNA ends were hairpin-capped. Accordingly, Lobachev et al. suggested that the breaks were produced by a resolvase acting on the four-way branch of a cruciform [25]. Further, hairpin terminated breaks were suggested to promote chromosomal translocation. It remains to be shown that such palindrome-directed breaks are resolvase-dependent, however this missing piece is understandable, given that identification of resolvase enzyme(s) in eukaryotes is still a work in progress (for a recent discussion see [26]).
The Lobachev results provide a framework for understanding how palindromes might actively undermine genome integrity: breaks are caused by an enzyme designed to act upon an intermediate in homologous recombination when it instead misappropriates a cruciform structure. Resolvases are meant to cleave Holliday junctions, structures comprised of a pair of duplex DNA molecules covalently connected to one another by two shared strands (Fig. 3A-ii). Identified junction-resolving enzymes allow physical separation of the two duplexes by introducing positionally-correlated nicks across the four-way junction in the two shared strands (for a review see [27]). Known resolvase-generated nicks permit recombined duplexes to go their separate ways (Fig. 3A). Known resolvases do not have an inherent ability to discriminate between a Holiday junction and the four-way junction of an extruded cruciform (Fig. 3B [28]). For example T4 DNA resolvase and RuvC both act by making cross-diagonal single strand cleavages at a cruciform base in vitro, breaking apart the four-stranded structure into two linear, hairpin-capped cleavage products [28], [29]. In fact the appearance of a T4 resolvase-sensitive structure is often taken as proof of cruciform extrusion. Thus, while resolution is beneficial in the context of a Holliday junction – two duplex DNAs are successfully separated – “resolution” is definitely undesirable in the context of an extruded cruciform. Resolvases break a once-contiguous duplex into two parts, creating hairpinned ends on either side of the interruption (Fig. 3B) [28].
Moving to humans, the genetic literature supports the notion that pathogenic breaks may be caused by the structure-forming potential of DNA palindromes. Pioneering work, principally by Emanuel and colleagues has demonstrated that the chromosomal exchanges seen for recurrent translocations involving Chr 22, 11 and 17 localize to the centers of large pre-existing near-palindromic regions on each [1], [2], [30], [31]. This is just as might be predicted if the translocations followed DNA breaks created by resolution of a cruciform structure.
Section snippets
A critical review
At present the trail leading from long palindromes to cruciforms to human chromosomal aberrations still involves picking one's way from one clue to the next rather than following a well-beaten evidentiary path. Two key questions are raised by the notion that palindromes are bad because they create cruciform structures: (1) Do extrusion-prone palindromes exist in the human genome? (2) For mammalian cells, are in vivo conditions sufficient to support cruciform extrusion of such palindromes?
How do palindromes arise?
A detailed discussion of the origin of DNA palindromes lies outside the focus of the present review, but possible modes of formation will be touched on briefly. There are four general proposals for palindrome generation, each of which has some experimental support. One is that palindromes are the result of a template-swtiching DNA polymerase. The polymerase might actually traverse the fork junction (changing allegiance from Watson to Crick, as Klenow fragment will do in vitro [60]), similarly a
Stabilizing the weak spots: a mammalian perspective
One productive avenue to study palindromes in higher eukaryotes effectively skirts the cloning barrier: this is to introduce a palindrome directly into the mouse genome and follow its fate by Southern blot analysis. Injection of linear DNA fragments into blastocysts can result in tandem arrays that include head–head and tail–tail insertions [67]. One example, a 15.6 kb palindromic insert in mouse chromosome 17 has been studied in some detail [68], [69], [70], [71]. The transgene in “Line 78” was
Center-Break palindrome revision
A simple scheme that incorporates our current understanding of palindrome revision in eukaryotic cells [70], [71], [75], [76] (A.G. Coté and J. Appleby unpublished) is presented in Fig. 6. Several aspects of the model are speculative, and are currently under investigation.
As diagrammed in Fig. 6, Center-Break palindrome revision is launched once a palindromic sequence has been induced to isomerize from lineform to cruciform. The cruciform structure constitutes a pre-mutagenic lesion at risk of
Next steps
Many issues surrounding the biology of palindromes have not even been touched on here, but invite investigation. To note a few, the physiological conditions and interventions that induce cruciform extrusion in vivo in mammalian cells are not yet experimentally defined. Transcription is one possibility (as raised by studies in E. coli), but neither its cruciform stimulating ability nor that of other processes has been directly tested in higher eukaryotes. Errors in DNA metabolism responsible for
Acknowledgements
This work is supported by an operating grant from National Cancer Institute of Canada (S.M.L.) and an Ontario Graduate Studentship (A.G.C.). The authors regret that many elegant and relevant studies could not be cited directly. We thank H. Lipshitz, A. Rattray, C. Pearson and T. Belsito for comments on the manuscript, K. Lobachev for communicating results prior to publication and numerous other colleagues who kindly responded to requests for reprints and information.
References (96)
- et al.
AT-rich palindromes mediate the constitutional t(11;22) translocation
Am. J. Hum. Genet.
(2001) - et al.
Intramolecular recombination and deletions in mitochondrial DNA of senescent, a nuclear-gene mutant of Neurospora crassa exhibiting “death” phenotype
Fungal Genet. Biol.
(2005) - et al.
Extrusion of an imperfect palindrome to a cruciform in superhelical DNA: complete determination of energetics using a statistical mechanical model
Mol. Cell. Biol.
(2002) - et al.
Helix opening transitions in supercoiled DNA
Biochim. Biophys. Acta
(1992) - et al.
Cruciform structures in palindromic DNA are favored by DNA supercoiling
Mol. Cell. Biol.
(1982) - et al.
The Mre11 complex is required for repair of hairpin-capped double-strand breaks and prevention of chromosome rearrangements
Cell
(2002) - et al.
Recognition and manipulation of branched DNA structure by junction-resolving enzymes
Mol. Cell. Biol.
(1997) - et al.
T4 endonuclease VII cleaves Holliday structures
Cell
(1982) - et al.
The constitutional t(17;22): another translocation mediated by palindromic AT-rich repeats
Am. J. Hum. Genet.
(2003) - et al.
Roles of topoisomerases in maintaining steady-state DNA supercoiling in Escherichia coli
J. Biol. Chem.
(2000)
Cruciform DNA structure underlies the etiology for palindrome-mediated Human chromosomal translocations
J. Biol. Chem.
Nucleotide sequence of the Belgian G gamma+(A gamma delta beta)0− thalassemia deletion breakpoint suggests a common mechanism for a number of such recombination events
Genomics
Cruciform formation in a negatively supercoiled DNA may be kinetically forbidden under physiological conditions
Cell
Torsionally tuned cruciform and Z-DNA probes for measuring unrestrained supercoiling at specific sites in DNA of living cells
J. Mol. Biol.
Large-scale effects of transcriptional DNA supercoiling in vivo
Mol. Cell. Biol.
Torsional tension in the DNA double helix measured with trimethylpsoralen in living E. coli cells: analogous measurements in insect and human cells
Cell
Direct observation of DNA distortion by the RSC Complex
Mol. Cell
The human cruciform-binding protein, CBP, is involved in DNA replication and associates in vivo with mammalian replication origins
J. Biol. Chem.
The mechanism of carcinogen-induced DNA amplification: in vivo and in vitro studies
Mut. Res.
End joining of genomic DNA and transgene DNA in fertilized mouse eggs
Gene
Choreography of the DNA damage response: spatiotemporal relationships among checkpoint and repair proteins
Cell
The 3′ to 5′ exonuclease activity of Mre11 facilitates repair of DNA double-strand breaks
Mol. Cell
Human Rad50/Mre11 is a flexible complex that can tether DNA ends
Mol. Cell
Involvement of poly(ADP-ribose) polymerase-1 and XRCC1/DNA ligase III in an alternative route for DNA double-strand breaks rejoining
J. Biol. Chem.
Exploring rare conformational species and ionic effects in DNA Holliday junctions using single-molecule spectroscopy
Mol. Cell. Biol.
Leading strand specific spontaneous mutation corrects a quasipalindrome by an intermolecular strand switch mechanism
J. Mol. Biol.
Long AT-rich palindromes and the constitutional t(11;22) breakpoint
Hum. Mol. Genet.
A gene deletion ending within a complex array of repeated sequences 3′ to the human b-globin gene cluster
PNAS
A novel sequence-based approach to localize translocation breakpoints identifies the molecular basis of a t(4;22)
Hum. Mol. Genet.
The position of t(11;22)(q23;q11) constitutional translocation breakpoint is conserved among its carriers
Hum. Genet.
A palindrome-mediated mechanism distinguishes translocations involving LCR-B of chromosome 22q11.2
Hum. Mol. Genet.
Widespread and nonrandom distribution of DNA palindromes in cancer cells provides a structural platform for subsequent gene amplification
Nat. Genet.
Vincent A hotspot for novel amplification joints in a mosaic of Alu-like repeats and palindromic A+T-rich DNA
EMBO J.
A mechanism of palindromic gene amplification in Saccharomyces cerevisiae
Genes Dev.
The PAL-mechanism of chromosome maintenance: causes and consequences
Cell Cycle
Chromosome instability as a result of double-strand breaks near telomeres in mouse embryonic stem cells
Mol. Cell. Biol.
Supercoil-induced extrusion of a regulatory DNA hairpin
PNAS
Vaccinia virus nicking-joining enzyme is encoded by K4L (VACWR035)
J. Virol.
Hairpin telomeres and genome plasticity in Borrelia: all mixed up in the end
Mol. Microbiol.
Site-directed mutagenesis of large DNA palindromes: construction and in vitro characterization of herpes simplex virus type 1 mutants containing point mutations that eliminate the oriL or oriS initiation function
J. Virol.
Specific interaction of the nonstructural protein NS1 of minute virus of mice (MVM) with [ACCA](2) motifs in the centre of the right-end MVM DNA palindrome induces hairpin-primed viral DNA replication
J. Gen. Virol.
Cruciform-extruding regulatory element controls cell-specific activity of the tyrosine hydroxylase gene promoter
Nucleic Acids Res.
A 160-bp palindrome is a Rad50.Rad32-dependent mitotic recombination hotspot in Schizosaccharomyces pombe
Genetics
Long palindromic sequences induce double-strand breaks during meiosis in yeast
Mol. Cell. Biol.
Interplay between human DNA repair proteins at a unique double-strand break in vivo
EMBO J.
Escherichia coli RuvC protein is an endonuclease that resolves the Holliday structure
EMBO J.
Unexpectedly high rate of de novo constitutional t(11;22) translocations in sperm from normal males
Nat. Genet.
Cited by (35)
Sak1 kinase interacts with Pso2 nuclease in response to DNA damage induced by interstrand crosslink-inducing agents in Saccharomyces cerevisiae
2014, Journal of Photochemistry and Photobiology B: BiologyA palindrome-mediated recurrent translocation with 3:1 meiotic nondisjunction: The t(8;22)(q24.13;q11.21)
2010, American Journal of Human GeneticsHomologous recombination restarts blocked replication forks at the expense of genome rearrangements by template exchange
2010, Molecular CellCitation Excerpt :Our data indicate that Rad50, Rhp51, and Rad22 act together to promote fork restart at the RTS1 barrier but that a Rad50/Rhp51-independent subpathway can also result in homology-driven template exchange. Chromosomal rearrangements linked to inverted repeats have previously been proposed to result from intramolecular single-strand annealing (SSA) following double-strand breakage in the vicinity of inverted repeats (reviewed in Haber and Debatisse, 2006; Lewis and Coté, 2006). Inversion of sequences enclosed within inverted repeats can also result from break-induced replication (BIR) followed by intermolecular SSA between sister chromatids (Malagón and Aguilera, 2001).
E. coli SbcCD and RecA Control Chromosomal Rearrangement Induced by an Interrupted Palindrome
2010, Molecular CellCitation Excerpt :DNA palindromes are sequences with a rotational symmetry. A perfect palindrome is a DNA sequence that is immediately juxtaposed to an exact reverse complementary copy of itself, whereas an interrupted palindrome (also known as a closely spaced inverted repeat) contains a small unique sequence that separates the two inverted complementary copies (Lewis and Coté, 2006). Due to their symmetry, palindromes have the unique potential to form intrastrand base pairs in hairpin or cruciform DNA structures, and it is this potential that underlies their role in genomic instability (Lewis and Coté, 2006; Tanaka and Yao, 2009).
- 1
Fax: +1 416 813 4931.