Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

G-Quadruplexes Involving Both Strands of Genomic DNA Are Highly Abundant and Colocalize with Functional Sites in the Human Genome

  • Andrzej S Kudlicki

    askudlic@utmb.edu

    Affiliations Department of Biochemistry and Molecular Biology, University of Texas Medical Branch, Galveston, Texas, United States of America, Institute for Translational Sciences, University of Texas Medical Branch, Galveston, Texas, United States of America, Sealy Center for Molecular Medicine, University of Texas Medical Branch, Galveston, Texas, United States of America

Abstract

The G-quadruplex is a non-canonical DNA structure biologically significant in DNA replication, transcription and telomere stability. To date, only G4s with all guanines originating from the same strand of DNA have been considered in the context of the human nuclear genome. Here, I discuss interstrand topological configurations of G-quadruplex DNA, consisting of guanines from both strands of genomic DNA; an algorithm is presented for predicting such structures. I have identified over 550,000 non-overlapping interstrand G-quadruplex forming sequences in the human genome—significantly more than intrastrand configurations. Functional analysis of interstrand G-quadruplex sites shows strong association with transcription initiation, the results are consistent with the XPB and XPD transcriptional helicases binding only to G-quadruplex DNA with interstrand topology. Interstrand quadruplexes are also enriched in origin of replication sites. Several topology classes of interstrand quadruplex-forming sequences are possible, and different topologies are enriched in different types of structural elements. The list of interstrand quadruplex forming sequences, and the computer program used for their prediction are available at the web address http://moment.utmb.edu/allquads.

Introduction

The G-quadruplex (G4) is a non-canonical DNA structure consisting of four strands stabilized by Hoogsteen bonds that has received significant attention in the recent years. G4s have been implicated in numerous cellular contexts and functions [1,2], including telomeres [3], cis-acting regulatory elements [4], transcription [5], and replication [6,7,8]. Four runs of guanine must be present in the DNA sequence from which a G4 is created [9,10]. While bimolecular or tetramolecular G-quadruplexes have been discussed in the context of short oligomers, or of interchromosomal interactions of telomeres [11], their significance for nuclear DNA under physiological conditions has been generally dismissed on the grounds of low strand concentration in the cell nucleus [12]. All eukaryotic and prokaryotic genomes are built from double-stranded DNA (dsDNA). In double-stranded genomic DNA, quadruplex structures may be formed using guanines originating either from one strand or from both strands of dsDNA. As first observed by Cao et al [13], the presence of a second strand with a complementary sequence opens the possibility of G-quadruplex configurations in which the four tracts of guanine are distributed between the two strands of DNA. For example, the sequence GGGAGGGACCCACCC is complemented by CCCTCCCTGGGTGGG, and the 12 guanines from both strands may be combined into a single G4 cage structure. The same number of Watson-Crick base pairs need to be broken to create a G4 from this sequence as in the “standard” case of a quadruplex with all guanines coming from the same strand of dsDNA, therefore no major energetic difference between the interstrand and intrastrand configurations should be expected. Nonetheless the definition of a quadruplex-forming sequence used in most genome-wide studies of G-quadruplex DNA is the same as for single-stranded DNA, implicitly assuming that, to form a quadruplex, four tracts of guanine, usually each at least 3nt long, must be positioned in consecutive locations along the same strand of DNA. This assumption has become a nearly unchallenged consensus in the field. As a consequence, the sequence motif G3+N1-7G3+N1-7G3+N1-7G3+ (or C3+N1-7C3+N1-7C3+N1-7C3+ for the complementary strand), see e.g. [14] has been adopted to predict potential sites of G-quadruplex formation in the genome. This motif, or its variants with different limits on the length of the loops, has been used in most algorithms for predicting putative quadruplex sequences (PQS), including Quadparser [12], G4 Calculator [15], QGRS Mapper [16], and others. Likewise, PQS databases, e.g. ([17,18]), and whole-genome analysis studies in higher eukaryotes, e.g. ([5,6,7,10,14,1928]), amounting to several hundred reports published to date, have used this motif or its variant [29]; with notable exceptions in the papers mapping possible quadruplexes in the yeast genome [13] and the human mitochondrial genome [30].

Here, I use a modified version of the approach of Cao et al. [13] to identify sequences potentially forming G-quadruplexes of all types in the human genome, demonstrating their high prevalence. Analysis of enrichment and overlap with functional sites points to association of distinct types of functional loci with the different topologies of G-quadruplex structures, which may suggest that the different G4 topologies are involved in different cellular processes.

Materials and Methods

Prediction of interstrand G-Quadruplexes

Potential quadruplex-forming sequences in the genome have been defined by the regular expression (PCRE type [31]):

  • m/(G{3,}).{1,7}\1.{1,7}\1.{1,7}\1|(C{3,}).{1,7}\2.{1,7}\2.{1,7}\2/g

for single-strand PQS, and by the following regular expressions

  • m/(G{3,}).{1,7}\1.{0,7}(C{3,}).{1,7}\2/g
  • m/(C{3,}).{1,7}\1.{0,7}(G{3,}).{1,7}\2/g
  • m/(G{3,}).{0,7}(C{3,}).{1,7}\2.{0,7}\1|(C{3,}).{0,7}(G{3,}).{1,7}\4.{0,7}\3/g
  • m/(G{3,}).{0,7}(C{3,}).{0,7}\1.{0,7}\2/g
  • m/(C{3,}).{0,7}(G{3,}).{0,7}\1.{0,7}\2/g
  • m/(G{3,}).{0,7}(C{3,}).{1,7}\2.{1,7}\2|(G{3,}).{1,7}\3.{1,7}\3.{0,7}(C{3,})/g m/(C{3,}).{0,7}(G{3,}).{1,7}\2.{1,7}\2|(C{3,}).{1,7}\3.{1,7}\3.{0,7}(G{3,})/g m/(G{3,}).{0,7}(C{3,}).{0,7}\1.{1,7}\1|(C{3,}).{1,7}\3.{0,7}(G{3,}).{0,7}\3/g m/(C{3,}).{0,7}(G{3,}).{0,7}\1.{1,7}\1|(G{3,}).{1,7}\3.{0,7}(C{3,}).{0,7}\3/g

for the different topology classes of cross-strand G-quadruplexes. The first regular expression produces results nearly identical to the Quadparser software [12], with minor differences due to different implicit heuristics applied in situations where alternative or overlapping PQS sequences exist. Similarly, my approach to finding interstrand quadruplex-forming sequences differs from the approach of Cao et al. [13] in that here a separate one-step search is performed for each topology class, while Cao et al. first identify DNA intervals with quadruplex-forming potential and then characterize the topology of the possible quadruplex. As a result, in certain cases of partially overlapping quadruplex-forming sequences some of the alternative topology types may be missed by the two-step approach, although my one-step method requires additional processing of the results if only non-overlapping sequences are desired. A complete Perl program and the results for the hg19 human genome assembly are available as supplementary data, and from the supporting website http://moment.utmb.edu/allquads (the website also provides the results for the hg18 and hg38 assemblies). The program reads a fasta file and outputs PQS’s in text format, one per line, including sequence id, topology class, position and the PQS sequence. The post-processing required to identify overlapping quadruplex-forming sequences is a straightforward task. It can be implemented as an algorithm with O(N log N) computational complexity in the number of quadruplexes when the PQS and DS-PQS sequences are first sorted according to chromosomal coordinate, this function is implemented among others by the intersectBed command of the bedtools package [32].

Analysis of functional associations

Following Gray et al. [5], I also predicted the potential PQS and DS-PQS sequences allowing for loops of up to 12 nt between the guanine tracts (replacing 7 with 12 in the regular expressions above), and used them in the functional analysis of PQS and DS-PQS sites compared to sequencing data on transcriptional initiation and origins of replication. The analysis used the hg19 build of the genome, with the exception of the hf2 antibody pull-down results [19] that have been mapped to hg18. To analyze the prevalence of PQS and DS-PQS sequences in promoter regions, I applied the allquads.pl algorithm directly to the upstream1000.fa fasta file obtained from the UCSC genome database on February 17th, 2015. To infer functions enriched in the DS-PQS loci, I analyzed their overlaps with experimentally identified sites of transcription initiation and origins of replication. The binomial tests for enrichment were performed as in [5], using the gsl_cdf_binomial_Q function included in the Math::GSL::CDF CPAN library [33].

The ChIP-seq, pull-down, G4-seq and nascent DNA sequencing peaks were obtained from GEO accession numbers GSE44849 (GSM1092544, GSM1092545), GSE28911 (GSM716435, GSM716437), GSE63874 and GSE45241 (GSM1099724, GSM1099725, GSM1099726, GSM1099727). Overlaps between peaks and G4’s were defined if there was at least one base pair common to both features.

Results

Prevalence of interstrand quadruplex forming sequences

Nine classes of G4 topologies involving both DNA strands are possible in addition to the previously described case of all guanines located on one strand of dsDNA [13], or five if one ignores the difference between quadruplex starting from the positive and one starting from the negative strand; example topological configurations are shown in Fig 1. Depending on the order of guanine and cytosine tracts on either strand within the sequence, I will denote them as AABB (4), ABAA (8), ABAB (6), ABBA (7), ABBB (2), BAAA (1), BABA (5), BABB (9) and BBAA (3), numbers in parentheses correspond to pattern classes as defined by Cao et al. [13]. AAAA stands for the widely discussed single-strand configuration; generally “A” represents a guanine tract and “B” a cytosine tract, counting from the 5’ end of either strand, reverse complements are not distinguished (e.g. AABA and BABB are the same). Note that each of the ten classes allows several conformations (differing by polarity and arrangement of loops), however they cannot be distinguished based on sequence alone. It is likely that certain topologies will allow formation of a hybrid i-motif [4] in addition to the G4, depending on the lengths of the loops connecting the runs of guanine and cytosine. The i-motif requires a specific range of pH [34,35] and its significance in-vivo is thus limited, therefore although i-motifs in physiological conditions have been reported in certain cases [36,37,38], in this paper I will focus on the G-quadruplex only.

thumbnail
Fig 1. Examples of topology classes of G-quadruplex structures within genomic DNA.

Examples of topological configurations of intrastrand and interstrand quadruplexes are shown schematically. For pairs of topology types that differ only by strand from which the sequence is derived (e.g. ABAB and BABA) only one is shown. Black, green—the Watson and Crick strands. Red: guanines, blue: cytosines, yellow: loops of up to seven nucleotides of any type. AAAA—the intrastrand topology, AABB, ABAB, ABBA, ABBB, BABB: interstrand configurations.

https://doi.org/10.1371/journal.pone.0146174.g001

To identify putative double-strand-derived quadruplex sequences (DS-PQS) in the human genome, I implemented an algorithm representing the sequence search in terms of Perl-compatible regular expressions. The source code of the program allquads.pl is available as a supplementary material (S1 File) as well as from the supporting website http://moment.utmb.edu/allquads. A genome-wide search with loop lengths between 1 and 7 nucleotides (0 to 7 for loops between guanine runs on opposite strands) has revealed 897,935 DS-PQS sequences, of which 196,953 have an overlap of at least one nucleotide with one of the 374,834 single-strand (AAAA) PQS’s. 150,294 DS-PQS are of the BAAA topology class (97,565 without overlap with a single-strand quadruplex), 152,329 (99,299 not overlapping with a single-strand PQS)–ABBB, 69,198 (56,445)–AABB, 142,890 (115,766)–ABAA, 55,735 (50,866)–ABAB, 96,163 (87,940)–ABBA, 49,558 (44,976)–BABA, 128,404 (101,561)–BABB and 53,364 (46,024)–BBAA. Notably, a significant asymmetry exists in the prevalence of some of the ‘mirror image’ configurations: AABB is more abundant than BBAA, ABAB is more abundant than BABA, and ABAA than BABB. Similar differences are present in the yeast genome, see fig. 4a of Cao et al. [13].

Taking into account the overlaps between DS-PQS’s of different topologies, 550,977 independent DS-PQS sites are present in the human genome. Similarly, 832,540 independent groups of overlapping quadruplex-forming sequences of any type (interstrand or intrastrand) are found. The complete list of human DS-PQS sites is available in the online supplement (S2 File) and from the supporting website. The numbers of identified interstrand and intrastrand PQS sites per chromosome are shown in Table 1, and detailed breakdown by topology class is listed in S1 Table. While the average PQS density (per megabase) varies significantly from chromosome to chromosome, the ratios of numbers of intrastrand to interstrand PQS sequences for every chromosome are close to the genome average of 0.68, with the exception of the X and Y chromosomes that are depleted in DS-PQS sites and have intrastrand to interstrand ratios of 0.81 and 0.82 respectively. This statistically significant difference (20.7σ and 9.3σ respectively, Poisson model) may reflect different functions of genes, different regulation, or different chromatin organization in the sex chromosomes compared to autosomes.

thumbnail
Table 1. Intrastrand and interstrand G-quadruplex sequences by human chromosome.

https://doi.org/10.1371/journal.pone.0146174.t001

The presented prevalence of potentially quadruplex-forming sequences have been computed for the standard human genome. While the genomes of human cell lines used in research (such as HeLa or HEK293) differ from the standard genome assembly, most of the differences are translocations or copy number variations that do not have significant impact on the presence or absence of sequences potentially forming G-quadruplexes. Local polymorphisms specific to the cell lines are limited to a relatively small number of sequences, and are not expected to affect the global statistics of PQS sites. For example, the polymorphisms specific to the HEK293T line, as identified by [39], do not overlap with any of the quadruplex-forming sites, either interstrand or intrastrand. Therefore, the results can be readily applied to the genomes of research cell lines.

Functions of sites with potential to form interstrand G4s

The high abundance of DS-PQS sites opens the possibility that interstrand quadruplexes may play a role in a major cellular process. Indirect evidence in favor of such a role may be derived from association between DS-PQS sites and genomic loci with functional properties known to involve G4 structures. Intrastrand G-quadruplexes and G-quadruplex forming sequences have been reported to coincide with promoter regions and play a role in transcriptional regulation [4,28,4046]. Indeed, at least one single-strand (AAAA) PQS is present within 45.0% regions 1-kb upstream of a transcription start site. Searching for DS-PQS sites reveals potential interstrand quadruplexes in 52.5% of these sequences (p < 10−308; binomial); a total of 63.1% of human transcripts have at least one putative G-quadruplex of any type in their 1-kb upstream region.

Additional evidence for the role of G4s in transcription initiation has been provided by a recent ChIP-seq study mapping the binding sites of transcriptional helicases XPB and XPD [5]: approximately 20% of XPB and XPD ChIP-seq peaks overlap with a single-strand PQS (approximately 40% when the PQS definition is relaxed to include loops of up to 12nt connecting the guanine runs). I have analyzed the data in the context of quadruplex—forming sites of all types. The overlaps of XPB and XPD binding sites with all detected G4 sequences—including interstrand—are significantly higher: 45% and 48% respectively for standard loop length of up to 7nt; or 70% and 73% respectively for XPB and XPD when allowing for loops up to 12nt long (see details in Table 2 and S2 Table). This result, along with the enrichment in promoter regions described above, demonstrates a significant association of DS-PQS sites with transcriptional initiation (p < 10−308; binomial test for enrichment of DS-PQS both in XPB and in XPD peaks). Notably, the enrichment of DS-PQS’s in transcriptional helicase binding sites is higher than for interstrand PQS’s, and there is no enrichment at all of peaks containing an intrastrand PQS but no DS-PQS; this observation is consistent with XPB and XPD binding only at interstrand G4’s; the enrichment of intrastrand PQS reported by [5] may be explained by intrastrand PQS overlapping with DS-PQS or present in some of the loci containing a DS-PQS.

thumbnail
Table 2. Intrastrand and interstrand G-quadruplex forming sequences associated with functional sites in the human genome.

https://doi.org/10.1371/journal.pone.0146174.t002

G4 structures have also been associated with origins of replication. To investigate whether this also applies to cross-strand topologies, I considered the overlap between DS-PQS’s and origins of replication that have been mapped by sequencing short nascent DNA in human MCF7 and K562 cells [47]. Again, while single-strand PQS’s are significantly enriched in the replication origins (present respectively in 20% and 26% of the peaks in K562 and MCF7 libraries), DS-PQS’s are even more prevalent (25% and 35%, p < 10−308 in both cases), resulting in 34% and 44% origins respectively overlapping with at least one G4 of any type, see Table 1.

In a recent study, a hf2 antibody that binds to G4 but not to dsDNA was used in a genome-wide pull-down experiment to characterize stable G4 structures in the human genome [19], revealing that 12% of the hf2 binding sites overlap with an intrastrand PQS. In hf2 peaks common to two or more replicate experiments, the ratio is 23%. When both intrastrand and interstrand PQS’s are taken into account, up to 17% of hf2 peaks (33% for two or more libraries) are associated with a potential quadruplex sequence. In this dataset, the ratio of interstrand to intrastrand G4’s is lower than in the functional studies above, and there is no significant enrichment of DS-PQS’s in the sequencing peaks. This result does not however contradict previous findings because the hf2 antibody was designed and tested for specificity only to intramolecular G-quadruplexes, derived from a single strand of DNA.

The resulting low ratio of interstrand to intrastrand quadruplexes is similar in the more sensitive G4-seq study of Chambers et al. [48], who detect quadruplexes by analysing sequencing mismatches between conditions promoting and disfavouring G4 formation. The G4-seq approach to quadruplex detection involves separate analysis for each strand of DNA and thus appears to favour intrastrand quadruplexes. Nonetheless 108,229 quadruplex sites overlap with DS-PQS loci (16-fold enriched in DS-PQS sequences), including 49,377 observed interstrand quadruplexes not overlapping with an intrastrand PQS, corresponding to a very significant 9.1-fold enrichment (this calculation is based on quadruplexes observed by [48] simultaneously in the K+ and the PDS experiments). The enrichment provides evidence that the G4-seq method does detect DS-PQS sites that do not coincide with an intrastrand PQS.

Enrichment of topology classes among DS-PQS associated with different functional loci

While the structures of interstrand G4s with different topologies are yet to be determined crystallographically, structural differences between them may be significant for the specific functions of the quadruplex structures. Specifically, if interstrand quadruplexes are functional in transcription initiation or in replication origin, different ratios of numbers of PQS’s with particular topology classes may be expected in the quadruplex-forming sequences associated with such functional elements. The numbers of potentially quadruplex-forming sequences in each topological category, associated with each type of functional element are listed in S3 Table, along with the fractions of all PQS’s and all DS-PQS’s that they constitute, and a comparison with the ratios computed genome-wide, irrespective of functional site. The enrichment calculation uses all predicted quadruplex-forming structures with a 7nt limit on loop length, including overlapping PQS’s with different topologies, as any of the overlapping PQS’s can be potentially functional. Generally, among the DS-PQS’s coinciding with origin of replication sites, the BABB, ABBB and BAAA topologies are significantly enriched (between 5σ and 24σ; asymptotic estimation for Poisson distribution), compared to the genome-wide prevalence. In the transcriptional helicase binding sites, the ABBA, BABB and BBAA topologies are enriched compared to their genome-wide abundances, while the intrastrand AAAA is consistently very strongly depleted (>18 σ). Interestingly, while the some of the “mirrored” topologies have different abundances genome-wide (e.g. AABB vs. BBAA, or ABAA vs. BABB), their abundances in many functional sites are nearly equal, suggesting different mechanism of selection of quadruplex topologies in functional and non-functional loci. Generally, these results constitute evidence of functional preference of quadruplex-forming sequences of different topology classes, and suggest that the function depends on the topology and structure of the interstrand G-quadruplex formed within the genomic DNA.

Discussion

By integrated sequence-based prediction with results of functional studies, I have shown that sequences potentially forming interstrand G-Quadruplexes, a nucleic acid structure previously not considered in higher eukaryotic nuclear DNA, are highly prevalent in the human genome and colocalize with functionally significant loci. Enrichments of interstrand and intrastrand PQS’s in the functional studies suggest that in DNA replication interstrand G4 conformations may have serve a function similar to intrastrand quadruplexes. In transcription initiation, the role of DS-PQS is, in the light of this analysis, even more prominent than that of intrastrand quadruplexes; possibly only interstrand G-quadruplexes are involved in recruitment of transcriptional helicases. Both single-strand and double-strand PQS’s should be considered in future studies of these and other functions of G4s in the nuclear DNA.

Supporting Information

S1 File. Source code of the AllQuads program for predicting interstrand G4-forming sequences.

https://doi.org/10.1371/journal.pone.0146174.s001

(TAR)

S2 File. The complete list of interstrand and intrastrand quadruplex sites in the human genome (hg19), with at least three guanines per tract and loops not longer than 7nt–(tar archive of compressed text files, one per chromosome).

https://doi.org/10.1371/journal.pone.0146174.s002

(TAR)

S1 Table. Abundance of intrastrand and interstrand G-quadruplex sequences of different topology classes in human chromosomes (separate pdf file).

https://doi.org/10.1371/journal.pone.0146174.s003

(PDF)

S2 Table. Detailed functional analysis of intrastrand and interstrand G-quadruplex sequences in human genome (separate pdf file).

https://doi.org/10.1371/journal.pone.0146174.s004

(PDF)

S3 Table. Relative abundances of PQS’s of different topology classes in the whole genome, and in the functional regions, calculated for all PQS’s and for interstrand PQS only.

XPD, XPB—transcriptional helicase binding sites; ORI–origins of replication; hf2 –antibody to intrastrand PQS, upstream1000 –promoter regions. Enrichments of ratios in functional sites compared to genome-wide proportions of PQS’s with different topology classes suggest that different PQS topologies may be responsible for different functions. Bottom panels—z-transformed enrichments, compared to genome-wide ratios for all PQS’s and for interstrand PQS only; negative numbers denote depletion. (separate pdf file).

https://doi.org/10.1371/journal.pone.0146174.s005

(PDF)

Acknowledgments

This study was conducted with the support of the Institute for Translational Sciences at the University of Texas Medical Branch, supported in part by a Clinical and Translational Science Award (UL1TR000071 and UL1TR001439) from the National Center for Advancing Translational Sciences.

Author Contributions

Conceived and designed the experiments: AK. Performed the experiments: AK. Analyzed the data: AK. Contributed reagents/materials/analysis tools: AK. Wrote the paper: AK.

References

  1. 1. Bochman ML, Paeschke K, Zakian VA. DNA secondary structures: stability and function of G-quadruplex structures. Nat Rev Genet. 13(11):770–80. Epub 2012/10/04. doi: nrg3296 [pii] pmid:23032257; PubMed Central PMCID: PMC3725559.
  2. 2. Rhodes D, Lipps HJ. G-quadruplexes and their regulatory roles in biology. Nucleic Acids Res. 43(18):8627–37. Epub 2015/09/10. doi: gkv862 [pii] pmid:26350216; PubMed Central PMCID: PMC4605312.
  3. 3. Paeschke K, Simonsson T, Postberg J, Rhodes D, Lipps HJ. Telomere end-binding proteins control the formation of G-quadruplex DNA structures in vivo. Nat Struct Mol Biol. 2005;12(10):847–54. Epub 2005/09/06. doi: nsmb982 [pii] pmid:16142245.
  4. 4. Kendrick S, Hurley LH. The role of G-quadruplex/i-motif secondary structures as cis-acting regulatory elements. Pure Appl Chem. 82(8):1609–21. Epub 2010/01/01. pmid:21796223; PubMed Central PMCID: PMC3142959.
  5. 5. Gray LT, Vallur AC, Eddy J, Maizels N. G quadruplexes are genomewide targets of transcriptional helicases XPB and XPD. Nat Chem Biol. 10(4):313–8. Epub 2014/03/13. doi: nchembio.1475 [pii] pmid:24609361; PubMed Central PMCID: PMC4006364.
  6. 6. Besnard E, Babled A, Lapasset L, Milhavet O, Parrinello H, Dantec C, et al. Unraveling cell type-specific and reprogrammable human replication origin signatures associated with G-quadruplex consensus motifs. Nat Struct Mol Biol. 19(8):837–44. Epub 2012/07/04. doi: nsmb.2339 [pii] pmid:22751019.
  7. 7. Valton AL, Hassan-Zadeh V, Lema I, Boggetto N, Alberti P, Saintome C, et al. G4 motifs affect origin positioning and efficiency in two vertebrate replicators. EMBO J. 33(7):732–46. Epub 2014/02/14. doi: embj.201387506 [pii] pmid:24521668; PubMed Central PMCID: PMC4000090.
  8. 8. Comoglio F, Schlumpf T, Schmid V, Rohs R, Beisel C, Paro R. High-Resolution Profiling of Drosophila Replication Start Sites Reveals a DNA Shape and Chromatin Signature of Metazoan Origins. Cell Reports. 2015;11(5):821–34. ISI:000353902900015. pmid:25921534
  9. 9. Sen D, Gilbert W. Formation of parallel four-stranded complexes by guanine-rich motifs in DNA and its implications for meiosis. Nature. 1988;334(6180):364–6. Epub 1988/07/28. pmid:3393228.
  10. 10. Maizels N, Gray LT. The G4 genome. PLoS Genet. 9(4):e1003468. Epub 2013/05/03. PGENETICS-D-13-00197 [pii]. pmid:23637633; PubMed Central PMCID: PMC3630100.
  11. 11. Tarsounas M, Tijsterman M. Genomes and G-Quadruplexes: For Better or for Worse. J Mol Biol. 2013;425(23):4782–9. ISI:000328522600013. pmid:24076189
  12. 12. Huppert JL, Balasubramanian S. Prevalence of quadruplexes in the human genome. Nucleic Acids Research. 2005;33(9):2908–16. Epub 2005/05/26. doi: 33/9/2908 [pii] pmid:15914667; PubMed Central PMCID: PMC1140081.
  13. 13. Cao K, Ryvkin P, Johnson FB. Computational detection and analysis of sequences with duplex-derived interstrand G-quadruplex forming potential. Methods. 57(1):3–10. Epub 2012/06/02. doi: S1046-2023(12)00117-X [pii] pmid:22652626; PubMed Central PMCID: PMC3701776.
  14. 14. Todd AK, Johnston M, Neidle S. Highly prevalent putative quadruplex sequence motifs in human DNA. Nucleic Acids Research. 2005;33(9):2901–7. Epub 2005/05/26. doi: 33/9/2901 [pii] pmid:15914666; PubMed Central PMCID: PMC1140077.
  15. 15. Eddy J, Maizels N. Gene function correlates with potential for G4 DNA formation in the human genome. Nucleic Acids Research. 2006;34(14):3887–96. ISI:000240583800010. pmid:16914419
  16. 16. Kikin O, D'Antonio L, Bagga PS. QGRS Mapper: a web-based server for predicting G-quadruplexes in nucleotide sequences. Nucleic Acids Research. 2006;34(Web Server issue):W676–82. Epub 2006/07/18. doi: 34/suppl_2/W676 [pii] pmid:16845096; PubMed Central PMCID: PMC1538864.
  17. 17. Wong HM, Stegle O, Rodgers S, Huppert JL. A toolbox for predicting g-quadruplex formation and stability. J Nucleic Acids. 2010. Epub 2010/08/21. pmid:20725630; PubMed Central PMCID: PMC2915886.
  18. 18. Zhang R, Lin Y, Zhang CT. Greglist: a database listing potential G-quadruplex regulated genes. Nucleic Acids Research. 2008;36:D372–D6. ISI:000252545400067. pmid:17916572
  19. 19. Lam EY, Beraldi D, Tannahill D, Balasubramanian S. G-quadruplex structures are stable and detectable in human genomic DNA. Nat Commun. 4:1796. Epub 2013/05/09. doi: ncomms2792 [pii] pmid:23653208; PubMed Central PMCID: PMC3736099.
  20. 20. Beaume N, Pathak R, Yadav VK, Kota S, Misra HS, Gautam HK, et al. Genome-wide study predicts promoter-G4 DNA motifs regulate selective functions in bacteria: radioresistance of D. radiodurans involves G4 DNA-mediated regulation. Nucleic Acids Research. 41(1):76–89. Epub 2012/11/20. doi: gks1071 [pii] pmid:23161683; PubMed Central PMCID: PMC3592403.
  21. 21. Rawal P, Kummarasetti VB, Ravindran J, Kumar N, Halder K, Sharma R, et al. Genome-wide prediction of G4 DNA as regulatory motifs: role in Escherichia coli global regulation. Genome Res. 2006;16(5):644–55. Epub 2006/05/03. doi: 16/5/644 [pii] pmid:16651665; PubMed Central PMCID: PMC1457047.
  22. 22. Burge S, Parkinson GN, Hazel P, Todd AK, Neidle S. Quadruplex DNA: sequence, topology and structure. Nucleic Acids Research. 2006;34(19):5402–15. Epub 2006/10/03. doi: gkl655 [pii] pmid:17012276; PubMed Central PMCID: PMC1636468.
  23. 23. Andorf CM, Kopylov M, Dobbs D, Koch KE, Stroupe ME, Lawrence CJ, et al. G-quadruplex (G4) motifs in the maize (Zea mays L.) genome are enriched at specific locations in thousands of genes coupled to energy status, hypoxia, low sugar, and nutrient deprivation. J Genet Genomics. 41(12):627–47. Epub 2014/12/21. doi: S1673-8527(14)00186-6 [pii] pmid:25527104.
  24. 24. Du XJ, Gertz EM, Wojtowicz D, Zhabinskaya D, Levens D, Benham CJ, et al. Potential non-B DNA regions in the human genome are associated with higher rates of nucleotide mutation and expression variation. Nucleic Acids Research. 2014;42(20):12367–79. ISI:000347693200010. pmid:25336616
  25. 25. Nguyen GH, Tang WL, Robles AI, Beyer RP, Gray LT, Welsh JA, et al. Regulation of gene expression by the BLM helicase correlates with the presence of G-quadruplex DNA motifs. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(27):9905–10. ISI:000338514800050. pmid:24958861
  26. 26. Lexa M, Kejnovsky E, Steflova P, Konvalinova H, Vorlickova M, Vyskot B. Quadruplex-forming sequences occupy discrete regions inside plant LTR retrotransposons. Nucleic Acids Research. 2014;42(2):968–78. ISI:000331138100030. pmid:24106085
  27. 27. Nakken S, Rognes T, Hovig E. The disruptive positions in human G-quadruplex motifs are less polymorphic and more conserved than their neutral counterparts. Nucleic Acids Research. 2009;37(17):5749–56. ISI:000271569100015. pmid:19617376
  28. 28. Verma A, Halder K, Halder R, Yadav VK, Rawal P, Thakur RK, et al. Genome-wide computational and expression analyses reveal G-quadruplex DNA motifs as conserved cis-regulatory elements in human and related species. J Med Chem. 2008;51(18):5641–9. ISI:000259342700018. pmid:18767830
  29. 29. Qin MY, Chen ZX, Luo QC, Wen Y, Zhang NX, Jiang HL, et al. Two-Quartet G-Quadruplexes Formed by DNA Sequences Containing Four Contiguous GG Runs. J Phys Chem B. 2015;119(9):3706–13. ISI:000350840600010. pmid:25689673
  30. 30. Dong DW, Pereira F, Barrett SP, Kolesar JE, Cao K, Damas J, et al. Association of G-quadruplex forming sequences with human mtDNA deletion breakpoints. BMC Genomics. 15:677. Epub 2014/08/16. doi: 1471-2164-15-677 [pii] pmid:25124333; PubMed Central PMCID: PMC4153896.
  31. 31. Friedl JEF. Mastering regular expressions: powerful techniques for Perl and other tools. 1st ed. Cambridge; Sebastopol: O'Reilly; 1997. xxiv, 342 p. p.
  32. 32. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26(6):841–2. ISI:000275243500019. pmid:20110278
  33. 33. Comprehensive Perl Archive Network website. Available: http://www.cpan.org.
  34. 34. Phan AT, Mergny JL. Human telomeric DNA: G-quadruplex, i-motif and watson-crick double helix. Nucleic Acids Research. 2002;30(21):4618–25. ISI:000179038100005. pmid:12409451
  35. 35. Singh RP, Blossey R, Cleri F. Structure and Mechanical Characterization of DNA i-Motif Nanowires by Molecular Dynamics Simulation. Biophysical Journal. 2013;105(12):2820–31. ISI:000328597400027. pmid:24359754
  36. 36. Xu Y, Sugiyama H. Formation of the G-quadruplex and i-motif structures in retinoblastoma susceptibility genes (Rb). Nucleic Acids Research. 2006;34(3):949–54. ISI:000235606200017. pmid:16464825
  37. 37. Day HA, Pavlou P, Waller ZAE. i-Motif DNA: Structure, stability and targeting with ligands. Bioorganic & Medicinal Chemistry. 2014;22(16):4407–18. ISI:000340703500009.
  38. 38. Phan AT, Leroy JL. Intramolecular i-motif structures of telomeric DNA. Journal of Biomolecular Structure & Dynamics. 2000:245–51. ISI:000165410200010.
  39. 39. Lin YC, Boone M, Meuris L, Lemmens I, Van Roy N, Soete A, et al. Genome dynamics of the human embryonic kidney 293 lineage in response to cell biology manipulations. Nature Communications. 2014;5 Artn 4767. ISI:000342927700001.
  40. 40. Siddiqui-Jain A, Grand CL, Bearss DJ, Hurley LH. Direct evidence for a G-quadruplex in a promoter region and its targeting with a small molecule to repress c-MYC transcription. Proceedings of the National Academy of Sciences of the United States of America. 2002;99(18):11593–8. ISI:000177843100014. pmid:12195017
  41. 41. Sun DY, Pourpak A, Beetz K, Hurley LH. Direct evidence for the formation of G-quadruplex in the proximal promoter region of the RET protooncogene and its targeting with a small molecule to repress RET protooncogene transcription. Clin Cancer Res. 2003;9(16):6122s–3s. ISI:000187467300218.
  42. 42. Simonsson T, Pecinka P, Kubista M. DNA tetraplex formation in the control region of c-myc. Nucleic Acids Research. 1998;26(5):1167–72. ISI:000072363300005. pmid:9469822
  43. 43. Zhang C, Liu HH, Zheng KW, Hao YH, Tan Z. DNA G-quadruplex formation in response to remote downstream transcription activity: long-range sensing and signal transducing in DNA double helix. Nucleic Acids Research. 2013;41(14):7144–52. ISI:000323050700038. pmid:23716646
  44. 44. Thakur RK, Kumar P, Halder K, Verma A, Kar A, Parent JL, et al. Metastases suppressor NM23-H2 interaction with G-quadruplex DNA within c-MYC promoter nuclease hypersensitive element induces c-MYC expression. Nucleic Acids Research. 2009;37(1):172–83. ISI:000262335700015. pmid:19033359
  45. 45. Sun D, Hurley LH. The Importance of Negative Superhelicity in Inducing the Formation of G-Quadruplex and i-Motif Structures in the c-Myc Promoter: Implications for Drug Targeting and Control of Gene Expression. J Med Chem. 2009;52(9):2863–74. ISI:000265911800025. pmid:19385599
  46. 46. Raiber EA, Kranaster R, Lam E, Nikan M, Balasubramanian S. A non-canonical DNA structure is a binding motif for the transcription factor SP1 in vitro. Nucleic Acids Research. 2012;40(4):1499–508. ISI:000301069400016. pmid:22021377
  47. 47. Martin MM, Ryan M, Kim R, Zakas AL, Fu H, Lin CM, et al. Genome-wide depletion of replication initiation events in highly transcribed regions. Genome Res. 21(11):1822–32. Epub 2011/08/05. doi: gr.124644.111 [pii] pmid:21813623; PubMed Central PMCID: PMC3205567.
  48. 48. Chambers VS, Marsico G, Boutell JM, Di Antonio M, Smith GP, Balasubramanian S. High-throughput sequencing of DNA G-quadruplex structures in the human genome. Nature Biotechnology. 2015;33(8):877-+. ISI:000359274900028. pmid:26192317