| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center for Genome Research, Institute of Biosciences and Technology, Texas A&M University System Health Science Center, The Texas Medical Center, Houston, Texas, USA
1Correspondence: Center for Genome Research, Institute of Biosciences and Technology, Texas A&M University System Health Science Center, The Texas Medical Center, 2121 W. Holcombe Blvd., Houston, TX 77030-3303, USA. E-mail: rwells{at}ibt.tamhsc.edu
 |
ABSTRACT |
|---|
Key Words: sticky DNA • GAA·TTC • transcription silencing • non-B DNA conformations • genetic instability • polyamides
|
INTRODUCTION |
|---|
, 2)
. The gene has been mapped to chromosome 9q13-q21.1 by linkage studies. The FRDA gene, FXN or X25, contains 7 exons and encodes a 210 amino acid protein called frataxin, a mitochondrial protein involved in iron-sulfur cluster biosynthesis. Of the patients with FRDA, 98% have an expanded GAA·TTC trinucleotide repeat sequence (TRS) in the first intron of the FXN gene and 
2% have point mutations. Normal alleles contain 6–34 repeats, but the expanded alleles of patients range from 66 to 1700 repeats (3
, 4)
. The age of onset and the severity of the disease are inversely correlated with the length of the GAA·TTC repeats (1
2
3
4
5
6)
, in many cases. Patients carrying expanded GAA·TTC repeats in both alleles have reduced levels of frataxin. The somatic instability of the expanded repeats progresses throughout life (7)
.
The etiology of FRDA involves the loss of transcriptional activity of the FXN gene, which leads to reduced levels of functional frataxin protein. This inhibition may be due to the repeat expansion (1
2
3
4
5
, 8
9
10
11
12)
and the capacity of the expanded GAA·TTC repeats to form non-B DNA structures, especially triplexes (13
14
15
16
17
18
19
20
21
22)
and sticky DNA (23
24
25
26)
. Other types of non-B DNA structures that have been characterized for the polymorphic GAA·TTC sequence include hairpins (22
, 27
, 28)
and parallel DNA (28)
. The expansion of the GAA·TTC sequence is performed by replication, recombination, and repair in model systems (4
, 29
30
31
32
33
34
35)
. After the TRS are expanded, the repeat may influence these processes since long repeats are known to stall replication forks in vivo and in vitro (36
37
38
39)
. The pausing of replication is attributed to the expanded TRS and its ability to form these types of non-B DNA structures (9
, 14
, 21
, 36)
.
Excellent recent reviews exist on clinical and pathophysiological issues (6
, 40)
, experimental therapeutics (41)
, evolution and instability of the TRS (30
, 40
, 42)
, mouse models (43
, 44)
, the involvement of triplexes and sticky DNA (45
, 46)
, iron dysregulation (47
, 48)
, oxidative stress (49)
, and frataxin as a protein (50)
. Furthermore, a brief overview has been published (51)
of a very recent meeting of researchers in the field covering all of these topics.
Interestingly, this laboratory published a review in FASEB Journal on the chemistry and biology of triplexes in 1988 (52)
. We stated that "it is likely that these unorthodox structures play an important role in the function of the eukaryotic genome"; obviously, we now recognize that this prediction was correct.
|
DNA CONFORMATIONS ASSOCIATED WITH LONG GAA·TTC TRACTS |
|---|
, 6
) discovered that the expansion of a GAA·TTC repeat sequence in intron 1 of the FXN gene was responsible for the disease etiology in most cases. Predictably, the attention of many investigators immediately turned toward this interesting repeat sequence at both the DNA level as well as its mRNA. Numerous investigations prior to 1996 (53
and references cited therein) revealed the polymorphic nature of this sequence since its capacity to form three-stranded DNA structures (triplexes) was well established due to its mirror repeat R·Y sequence with the concomitant possibility of forming a number of base-paired structures. The types of triplexes formed as well as the conditions for their stabilization and their biochemical and genetic functions have been reviewed (9
, 13
, 23
24
25
26
, 35
, 46
, 54)
. Inter- as well as intramolecular DNA triplexes are well known (13)
as are DNA·RNA triplexes where the RNA strand fills the major groove in the DNA duplex (55
56
57)
. As stated in the introduction, several other types of non-B DNA structures may be adopted by this repeating sequence.
However, in 1999 when Sakamoto et al. (23)
undertook investigations directly related to FRDA on long tracts of (GAA·TTC)n (where n=up to 270), they had substantial reservations due to the relatively imprecise types of triplexes that were formed by much shorter lengths (n=20–58) of GAA·TTC in plasmids (16)
. Indeed, these authors were astonished to realize that extremely well-defined types of a certain triplex were formed by tracts of GAA·TTC, which were longer than 59 repeats in length (75, 90, 115, 150, and 270 repeats). Thus, whereas we thought that this investigation might not yield interpretable and quantitative data, exactly the opposite occurred. Extremely unusual behavior was observed since segments of the recombinant plasmids were tenaciously adhering to each other, which gave rise to the concept of sticky DNA (Fig. 1
). Sticky DNA has been the subject of intense study in this and other laboratories over the past 8 years (9
, 14
, 23
24
25
26
, 35
, 45
, 46
, 54)
. Investigations have focused on the types of triplexes that are adopted by the long GAA·TTC repeats, both in vitro and in vivo, the role of the structure as a mutagen [see Effects of Triplexes (Sticky DNA) on Other DNA Metabolic Events], its involvement in replication, repair, and recombination-mediated genetic instabilities, the influence of small sequence interruptions on the structure, and the capacity of DNA sequence-specific ligands to alter the conformation to alleviate transcriptional repression (see Therapeutic Strategies).
|
Sticky DNA (Fig. 1)
is a long GAA·GAA·TTC triplex that is formed intramolecularly in vivo and in vitro (reviewed in ref. 45
) by a pair of repeating GAA·TTC tracts. Two tracts of GAA·TTC repeats within a given DNA interact to form a complex that can only be broken with high temperatures (80°C) and with the addition of EDTA to remove divalent metal ions. The net effect of these intramolecular interactions is to lock the DNA molecule into a dumbbell-shaped structure (Fig. 1)
, which impedes virtually all biological processes in vitro including transcription, replication, repair, and recombination (9
, 14
, 23
24
25
26
, 35
, 45
, 46
, 54)
and excludes nucleosome assembly (H. Ruan, M. Napierala, R. D. Wells, and Y.-H. Wang, unpublished results). Because of the extreme stability of sticky DNA, the two ends of the circular plasmid behave independently in terms of their supercoiled domains (26)
.
Since this sequence is important in the etiology of FRDA, substantial attention has been devoted to the in vivo properties of sticky DNA (26)
. Clearly, sticky DNA exists and functions in bacterial systems (26
, 35
, 46)
. Likewise, long tracts (n=60, 150) of (GAA·TTC)n are mutagenic in African green monkey kidney (COS-7) cells as well as CV-1 and human embryonic kidney HEK-293 cells (58)
, which implies behaviors similar to those observed in bacteria. However, direct determinations in eukaryotic cells remain to be conducted to ascertain the precise type of non-B structure responsible.
Extensive investigations have been conducted in living Escherichia coli to evaluate the role of the lengths and orientations of GAA·TTC repeat tracts (either one or two in number) on the supercoil relaxation (26)
. This procedure has been used extensively for triplexes, cruciforms, and Z-DNA, because when DNA structural transitions occur in small regions of plasmids as a result of DNA underwinding, the global negative supercoil value of the plasmid decreases (becomes less negative). To maintain the physiological negative supercoil density of the plasmid (–0.025) in the cell, DNA gyrase adds negative supercoils into the DNA to reestablish the normal homeostatic supercoil levels. Hence, when the plasmid is isolated and analyzed in vitro, the result of this negative supercoil change on agarose gels containing chloroquine is an overly negative supercoiled DNA with a greater negative linking number difference (26)
. When a family of plasmids containing either one or two tracts of GAA·TTC of varying lengths were subjected to these analyses, a unique behavior was observed since plasmids that have the capacity to form sticky DNA displayed an overwound, not an underwound, primary helix. This behavior was unexpected and indicates an apparent, but theoretically intractable, compression of the primary helices, an unprecedented result with any other non-B structure forming sequence (26)
. Since these investigations involve measurements of the gel mobility of intact plasmids in vitro to enable conclusions about in vivo behaviors, this apparent compression of the bp may be due to other interactions, which remain to be identified. These data clearly demonstrate that sticky DNA exists in vivo and has important functional consequences (interactions with proteins or roles in metabolic processes), which have not been observed previously. Further work will be required to elucidate the details of these properties.
Sticky DNA requires GAA·TTC tracts, which are at least 59 repeats (and up to 300) in length and a pair of repeat tracts in the direct repeat orientation. If the inserts are in the indirect (inverted) repeat orientation, no sticky DNA is observed. Divalent metal ions (magnesium) as well as negative supercoiling are also required for the sticky DNA structure formation (26)
.
To elucidate the effect of sequence on sticky DNA formation, various extents of GGA·TCC interruptions were introduced into long repeats of GAA·TTC tracts in a family of homologous plasmids. The plasmids were prepared in vitro but then were investigated in vivo. For plasmids that harbored more than 20% of GGA·TCC interruptions in their TRS, the formation of sticky DNA was abolished. However, the DNAs with GAA·TTC repeats with less than 11% of the GGA·TCC interruptions formed triplexes and/or sticky DNA similar to the uninterrupted sequences (46
, 54)
. Thus, when considering therapeutic strategies in the future for FRDA, site-directed mutagenesis could be an effective way to destabilize the sticky DNA conformation (45)
. Interestingly, an individual with late onset ataxia was found to be heterozygous for an unusual (GAAGGA·TCCTTC)65 sequence and a normal GAA·TTC repeat in the frataxin gene. This repeating hexanucleotide is nonpathogenic (59)
. Furthermore, this sequence alone will not adopt the sticky DNA structure or a triple helical conformation, as expected from its base pairing capabilities. However, if a tract of the repeating hexanucleotide sequence is present in the same plasmid as a pure long GAA·TTC tract, sticky DNA is formed (25
, 26)
as expected on the basis of base pairing possibilities. The kinds of base-paired structures that may be formed (24
25
26)
are highly predictable with respect to the formation of sticky DNA.
Sticky DNA apparently occurs and functions only in the FXN gene (in its first intron), which is one of the 20–25,000 genes in humans. Whereas the GAA·TTC sequence is not uncommon (60)
, the extremely long tracts that are found to be associated with FXN are apparently unique. Closely related R·Y sequences can also adopt this structure (25)
, as expected, and thus future investigations will be required to evaluate the ubiquity of this non-B DNA conformation. Note that no direct evidence exists for triplexes or sticky DNA having an obligatory role in the etiology of FRDA.
|
TRANSCRIPTION INHIBITION BY LONG TRACTS OF GAA·TTC |
|---|
DNA conformational effects
All molecular biological investigators in the Friedreich ataxia field concur that the inhibition of FXN transcription is the principal and critical step affected in this autosomal recessive disease (4)
. Long tracts of GAA·TTC suppressed FXN gene expression (1
2
3
4
5
6
, 61)
. This pathogenic loss of function mechanism agrees with the well-known recessive nature of disease and, thus, the deficiency of the protein frataxin is explicable. Furthermore, virtually all molecular biological investigators have implicated non-B DNA structures, either triplexes or sticky DNA, in the mechanism of this inhibition (6
, 8
9
10
11
12
, 14
, 21
, 33
, 45
, 46
, 54
, 62
, 63)
.
After the discovery of sticky DNA (23)
, in vitro investigations were conducted (9)
to directly evaluate the transcription behaviors of this unorthodox conformation. As expected, sticky DNA was an effective inhibitor of the synthesis of RNA, consistent with the well-known inhibitory effect of triplexes on transcription (55
56
57
, 64
65
66)
. Surprisingly, transcriptional inhibition was observed not only for the sticky DNA template but also another DNA molecule used as an internal control in an orientation-independent manner. Thus, the molecular mechanism for the transcription inhibition by sticky DNA was a sequestration of the RNA polymerases by direct binding to the complex DNA structure. Only sticky DNA showed this behavior (9)
. Furthermore, a family of GGA·TCC-interrupted long GAA·TTC repeat tracts have been cloned and characterized (see above). As expected, the greater the extent of the interruptions of the GAA·TTC repeats, the less inhibition of in vitro transcription was observed, consistent with the capacity of the interruptions to inhibit the formation of sticky DNA. The interruptions introduced base mismatches into the R·R·Y triplex, which explains its observed chemical and biological properties (54)
.
Grabczyk et al. (66)
have recently reported the formation of a persistent RNA·DNA hybrid by transcription of the FXN GAA·TTC repeat sequence in E. coli and by T7 RNA polymerase in vitro. RNA·DNA hybrids were observed with repeats numbering 44 or 88 in length. During in vitro transcription of the longer repeats, T7 RNA polymerase arrested in the promoter distal end of the GAA·TTC tracts and an extensive RNA·DNA hybrid was tightly linked to this arrest. Thus, RNA·DNA hybrid formation appears to be an intrinsic property of transcription through long GAA·TTC repeats. An attractive model (Fig. 2
) was proposed to explain the transcription-coupled RNA·DNA hybrid formation in the GAA·TTC repeats. Initially, the repeating DNA d(TTC)n strand serves as the template for synthesis of r(GAA)n to form a moderate length of DNA·RNA hybrid. Due to the stability of this hybrid, the DNA triplex is dislodged behind the growing transcription complex to give an even longer RNA·DNA hybrid. The waves of negative supercoiling behind the translocating RNA polymerase facilitates these processes from a topological standpoint. As stated above, the poly R·Y sequence of the FXN repeat sequence enables these interactions from a nucleic acid structural standpoint. The formation of the quasi-stable DNA triplex and DNA·RNA hybrid generates a pause site at the TRS. However, the possible involvement of a DNA·RNA triplex (two strands of DNA and one strand of RNA) (56
, 57)
has not been eliminated (62
, 66
; E. Grabczyk, personal communication).
|
In 1998, the inhibitory effects of the expanded GAA·TTC repeats (up to 270 U in length) were investigated in vivo by transient transfection into COS-7. A length and orientation-dependent inhibition of a reporter gene was observed (8)
. Very low levels of mature mRNA were found when plasmids with longer TRS were investigated with no accumulation of the primary transcript. Inhibitory effects were also found in DNA replication studies with long tracts of the TRS (8)
.
Influences of chromatin structure
A rather different mechanism of transcriptional silencing has been investigated by Saveliev et al. (67)
based on the packaging of DNA within chromatin. DNA embodied in chromatin and heterochromatin has been implicated in gene silencing. DNA wraps around histones, creating nucleosomes in chromatin. Repetitive DNA sequences have the propensity to package genomic regions into inaccessible heterochromatin structures that can lead to gene silencing (67)
. Prior in vitro investigations (68)
revealed that long tracts of repeating CTG·CAG bind histones tightly and "position" nucleosomes. Alternatively, long tracts of the fragile X CCG·CGG repeats displayed strong nucleosome exclusion (69
, 70)
. Hence, a precedent was established in vitro for a profound influence of TRS on chromatin structure. The FXN TRS confer variegation of expression on a linked transgene in mice (67)
. Silencing was correlated with a decrease in promoter accessibility and was enhanced by the classical position effect variegation modified heterochromatin protein 1. Interestingly, the TRS-associated variegation was not restricted to classical heterochromatic regions but occurred irrespective of chromosomal location. Hence, the mechanisms responsible for the heterochromatin-mediated silencing might have a role in gene regulation throughout the mammalian genome and, thus, modulate the extent of gene silencing and possibly the severity of several TRS diseases (67)
.
In contrast, direct binding measurements have been conducted in vitro on plasmids containing long tracts of GAA·TTC duplex as well as GAA·GAA·TTC triplexes (H. Ruan and Y.-H. Wang, unpublished results). The duplex excludes nucleosome assembly, and the triplex further lowers the nucleosome assembly efficiency. The difference in assembly efficiency is amplified when hypoacetylated histones were used, compared to assembly with hyperacetylated histones. Also, the triplex structure destabilizes the extent of adjacent sequences to assemble nucleosome arrays. Thus, these in vitro investigations on the GAA·TTC duplex and triplex structures reveal profound influences on local chromatin structure.
Relationship between in vitro and in vivo chromatin structure determinations
This apparent discrepancy in FRDA systems between in vitro binding measurements, where the GAA·TTC tracts and/or triplexes inhibit nucleosome assembly, and the in vivo genetic determinations (67)
, which indicate an enhancement of heterochromatin formation by the long GAA·TTC repeats, may be resolved by further biochemical experiments in the future. Other studies with histone deacetylase inhibitors indicate that a heterochromatin structure is present in the region surrounding the expanded GAA·TTC repeats and inhibits the transcription of the FXN gene in cells from patients (71)
. Heterochromatin is enriched with hypoacetylated histones, which our results (see above) suggest could distinguish the GAA·TTC duplex and the triplex from adjacent (flanking) sequences more than with hyperacetylated histones. Hence, the presence of the hypoacetylated histones at the expanded GAA·TTC site might increase accessibility to other chromatin-remodeling proteins, which result in heterochromatin structure and/or promoting the formation of the GAA·GAA·TTC triplex.
Also, the observation of a difference between heterochromatin-like structure present in vivo adjacent to the expanded GAA·TTC repeat and in vitro nucleosome exclusion of the GAA·TTC repeats is reminiscent of the results with the CGG·CCG repeats in the FMR1 gene of fragile X syndrome patients (69
, 72)
. For the fragile X syndrome, the discrepancy between in vitro and in vivo studies suggests that epigenetic events, such as DNA CpG methylation, play a role in the formation of unusual chromatin structure over the expanded CGG·CCG repeats. Although the GAA·TTC repeats in FRDA do not contain intrinsic CpG dinucleotides, a CpG island is located 1 kb upstream of the GAA·TTC repeats at the boundary of exon/intron 1 of the FXN gene (H. Ruan and Y.-H. Wang, unpublished results). In fact, three CpG dinucleotides located at the 5' end of the GAA·TTC repeat have been shown to posses preferential methylation in FRDA patients (73)
. The nucleosome exclusion of the GAA·TTC repeats could increase the accessibility of these regions to various protein machineries, such as DNA methyltransferase. In summary, further investigations must be conducted in FRDA systems in order to resolve the apparent discrepancy between the conclusions derived from in vivo genetic investigations on heterochromatin formation and the in vitro direct binding studies.
A different epigenetic study has focused on DNA methylation in segments of the FXN gene, which flank the GAA·TTC repeat tract in intron 1. Greene et al. (73)
showed that a region adjacent to the TRS was methylated in both the unaffected and affected individuals. However, methylation was more extensive in patients. Furthermore, three residues were almost completely methylation-free in unaffected individuals but almost always methylated in patients with FRDA. One of these residues is located within an E-box, whose deletion caused a significant drop in promoter activity in reporter assays. Elevated levels of histone H3 dimethylated on lysine 9 were found in FRDA cells, consistent with a more repressive chromatin organization. This type of chromatin is known to reduce transcription elongation. Hence, this could be a mechanism where the expanded repeats contribute to the frataxin deficit in FRDA.
In summary, in vitro and in vivo investigations on the role of long tracts of GAA·TTC on transcription are consistent with the recessive nature of the disease and its loss of function characteristics. Considering all of the 30 hereditary neurological diseases (74)
caused by expansions of triplet repeats (4)
, FRDA may be the clearest case of a well-defined biochemical step (transcription) that is inhibited, thus causing the disease etiology. Accordingly, this realization elevates the hope of being able to develop effective therapeutics (see Therapeutic Strategies) to ameliorate the disease manifestations.
|
EFFECTS OF TRIPLEXES (STICKY DNA) ON OTHER DNA METABOLIC EVENTS |
|---|
, 76
and references cited therein) have clearly demonstrated the mutagenic capacity of both inter- and intramolecular triplexes. A wide range of triplexes is effective in inducing recombination and DNA repair (75)
. Indeed, specific mutations have even been induced in mice by triplex-forming oligonucleotides (77)
. Furthermore, naturally occurring triplex-forming sequences are mutagenic in mammalian cells and generate substantial deletions, translocations, and other mutations (78
, 79)
. Thus, it is not surprising that triplexes (sticky DNA) associated with the FXN gene show mutagenic behaviors.
Whereas ample evidence demonstrates the involvement of triplexes in genetic instabilities, DNA repair and recombination (75
, 76
and references cited therein), a new paradigm in disease etiology has been discovered with the finding of triplexes coinciding with breakpoints of gross deletions (reviewed in refs. 80
, 81
). Triplexes as well as other non-B DNA conformations are found at the breakpoints of a wide range of genetic rearrangements (58
, 60
, 80
81
82
83
84
85)
. The types of genomic rearrangements include translocations, deletions, insertions, inversions, and duplications and include the TRS as well as regions of flanking DNA sequences. The non-B DNA conformations are directly involved in the repair-recombination functions, since remnants of the triplexes and other non-B conformations are found in the genetically rearranged products. Thus, the residual portions of the non-B structures cleaved at their labile contorted residues leave telltale tracks of the roles of these unorthodox conformations (80)
. These rearrangements are the genetic basis for approximately 50 human diseases, including adrenoleukodystrophy, follicular lymphomas, and spermatogenic failure (reviewed in refs. 17
, 80
). Hence, it would not be surprising if other physiological roles are discovered in the future for the long and stable sticky DNA triplex in FRDA.
|
Genetic instabilities |
|---|
, 14
, 39
, 74
, 86)
occur by DNA replication, recombination, and repair (especially methyl-directed mismatch repair and nucleotide excision repair) as studied in pro- and eukaryotic model systems. These biochemical-genetic processes probably act in concert due to slippage of the complementary strands relative to each other. The biophysical properties of the folded-back repeating sequence strands have a critical role in these instabilities. The non-B DNA structural elements include hairpins, slipped structures, DNA unwinding elements, tetraplexes, triplexes, and sticky DNA. The replication mechanisms are influenced by pausing of the replication fork, orientation of the triplet repeat strands, location of the repeat sequences relative to the replication origins, and flap endonucleases. The long GAA·TTC tracts associated with FRDA are expanded and deleted by mechanisms similar to the other known TRS (4
, 14
, 39
, 74)
. However, in contrast to the other TRS involved with myotonic dystrophy, Huntington’s disease, fragile X syndrome, etc., it seems that sticky DNA occurs and functions only in the FXN gene, which is one of the 20–25,000 genes in humans. Whereas other related sequences will form sticky DNA (25)
, none of these sequences has been identified yet with disease syndromes.
Genetic instability investigations on TRS responsible for FRDA, myotonic dystrophy, fragile X syndrome, and Huntington’s disease from 1990 to 2003 indicated that the TRS underwent either expansions or deletions with few or no base pair changes to the flanking sequences. However, more recent investigations (14
, 17
, 58
, 80
, 82
, 87
88
89)
showed that substantial deletions, rearrangements, duplications, translocations, and inversions occurred in flanking sequences, in addition to the repeating triplet or tetranucleotide repeats. Rearrangements of several kbp were observed and were promoted by long tracts of GAA·TTC, CTG·CAG, CCAG·CTGG, or CGG·CCG. Thus, the presence of the simple repeating sequence has a profound influence on the repair-recombination properties of flanking DNA sequences.
An insightful family of investigations (58)
revealed that the non-B DNA conformations (triplexes or sticky DNA and slipped structures) formed by a portion of the long repeating tracts of FRDA, myotonic dystrophy type 1, and myotonic dystrophy type 2 genes, not the sequences per se, promoted mutagenesis in the flanking DNA regions. Studies in E. coli and three types of mammalian (COS-7, CV-1, and HEK293) fibroblast-like cells revealed that conditions that promoted the formation of the non-B structures enhanced the genetic instabilities, both in the repeat tracts and in the flanking sequences of up to 4 kbp. The three strategies utilized included: the in vivo modulation of global negative supercoil density using topA and gyrB mutant E. coli strains; the in vivo cleavage of hairpin loops, which are an obligate consequence of slipped-strand structures, cruciforms, and intramolecular triplexes, by inactivation of the SbcC protein; and by genetic instability studies with plasmids containing long repeating sequence inserts that do, and do not, adopt non-B DNA structures in vitro (58)
. Thus, it was possible to tease apart the effects of the sequences vs. the conformations. Clearly, the conformations are the culprits.
Even earlier investigations (87)
demonstrated that the slipped structures, cruciforms, and sticky DNA adopted by long tracts of GAA·TTC, CGG·CCG, and CTG·CAG repeats were responsible for the genetic instabilities of these sequences. The influence of negative superhelical density on the genetic instabilities of these repeat tracts was studied in vivo in topologically constrained plasmids in E. coli. The capacity of these DNA tracts to undergo deletions-expansions was explored with three genetic-biochemical approaches, including first, the utilization of topoisomerase I and/or DNA gyrase mutants; second, the specific inhibition of DNA gyrase by novobiocin; and third, the genetic removal of the HU protein, thus lowering the negative supercoil density. All three strategies revealed that higher negative supercoil densities in vivo enhanced the formation of deleted repeat sequences. Higher levels of negative supercoiling stabilize the formation of triplexes, sticky DNA, and slipped structures at appropriate repeat tracts. Also, numerous prior genetic instability investigations invoke a role for these structures in promoting the slippage of the DNA complementary strands. Thus, the authors concluded that the in vivo modulation of the DNA structure, localized to the repeat tracts, is responsible for these behaviors. These data agree with the conclusions (58)
derived from alternate types of mutagenic investigations.
Recombination
The triplex (sticky DNA) structures adopted by long GAA·TTC sequences from intron 1 of the FXN gene exert profound structure-dependent influences on the recombination behaviors of these repeating sequences (35)
. Intramolecular and intermolecular recombination studies showed that the frequency of recombination between the GAA·TTC tracts was as much as 15 times higher than the nonrepeating control sequences. Homologous, intramolecular recombination between GAA·TTC tracts and GAAGGA·TCCTTC repeats also occurred with a very high frequency (0.8%). Biochemical analyses of the recombination products demonstrated the expansions and deletions of the GAA·TTC repeats. These results, together with our previous studies (90
, 91)
on the CTG·CAG sequences, suggest that the recombinational hot spot characteristics may be a common feature of all triplet repeat sequences. Unexpectedly, we found that the recombination properties of the GAA·TTC tracts were unique, compared with CTG·CAG repeats, because they depended on the DNA secondary structure polymorphism. Specifically, increasing the length of the GAA·TTC repeats decreased the intramolecular recombination frequency between these tracts. This result was unexpected. Also, a correlation was found between the propensity of the GAA·TTC tracts to adopt the sticky DNA conformation and the inhibition of intramolecular recombination. Investigations with novobiocin to modulate the intracellular DNA topology (i.e., the lowering of the negative superhelical density, repressing the formation of the sticky DNA structure) restored the expected positive correlation between the length of the GAA·TTC tracts and the frequency of intramolecular recombination. Thus, these studies support other investigations (58
, 86)
demonstrating the role of the DNA structure, not the sequence per se, in the biological behaviors.
|
THERAPEUTIC STRATEGIES |
|---|
, 93)
.
Polyamides
Napierala et al. (35)
stated that "the use of pharmacological agents, capable of interfering with the formation of these unusual (DNA) structures, may up-regulate the expression of the FRDA gene, thereby providing an effective therapeutic approach." Gottesfeld et al. (94)
have shown that sequence-specific polyamides (95)
alleviate transcription inhibition associated with long GAA·TTC repeats. Since excellent reviews (95
, 96)
on this topic are available, only the FRDA salient issues will be discussed briefly herein. The pyrrole-imidazole polyamides synthesized by P. Dervan and associates are the only available class of small molecules that are designed to recognize virtually any predetermined DNA sequence (97
and papers cited therein). These molecules have affinities and specificities that equal or exceed natural eukaryotic transcriptional regulatory proteins. A major strength of polyamides from a therapeutic strategy is their cell permeability and their capability of localizing in the nucleus in various cultured cell lines; also, they down-regulate target genes in these cells. Studies with model gene systems (95)
and a variety of eukaryotic and viral transcription factors showed that these small ligands are potent inhibitors of protein-DNA interactions.
Burnett et al. (94)
reasoned that appropriate polyamides might bind to long GAA·TTC tracts with high affinity and, thereby, disrupt the intramolecular DNA·DNA-associated region of the sticky DNA conformation formed by these long TRS. Fluorescent polyamide-Bodipy conjugates localize in the nucleus of a lymphoid cell line derived from a FRDA patient. The synthetic ligands increase transcription of the frataxin gene in cell culture, resulting in increased levels of frataxin protein. DNA microarray analyses indicated that a limited number of genes were significantly affected in FRDA cells. Thus, the polyamides may increase transcription by altering (abolishing) the sticky DNA conformation of the FXN gene harboring the long GAA·TTC repeats; alternatively, a chromatin opening mechanism was discussed (94)
.
HDAC inhibitors
Second, a quite different therapeutic strategy has been broached by attempting to modulate gene regulation at the epigenetic level. Gottesfeld et al. (71)
have explored the chromatin structure of the FXN gene in normal and FRDA cell lines using antibodies to the various modification states of the core histones and chromatin immunoprecipitation methods. Gene silencing at expanded FXN alleles is accompanied by hypoacetylation of histones H3 and H4, and methylation of histone H3 at lysine 9, consistent with a heterochromatin-mediated repression mechanism. Hence, these authors conclude that histone deacetylase (HDAC) inhibitors, compounds that reverse heterochromatin, might activate the FXN gene. One commercial HDAC inhibitor, BML-210, which partially reverses silencing in the FRDA cell line, was identified. Based on the structure of this compound, Gottesfeld et al. synthesized and assayed a series of derivatives of BML-210 and identified HDAC inhibitors that reverse FXN silencing in primary lymphocytes from Friedreich’s patients. These molecules act directly on the histones associated with the FXN gene, increasing acetylation at particular lysine residues on histones H3 and H4 (H3-K14, H4-K5, and H4-K12). Since the expanded GAA·TTC repeats do not alter the coding potential of the FXN gene, gene activation should be beneficial from a therapeutic standpoint (98)
.
|
UNANSWERED QUESTIONS |
|---|
What is the mechanism of transcription silencing? Is it one of the three mechanisms described in Transcription Inhibition by Long Tracts of GAA·TTC or another process?
Does sticky or triplex DNA exist and function in humans? What is the relationship to FRDA?
What is the detailed molecular structure of sticky DNA, a unique non-B DNA conformation, both in vivo and in vitro?
What is the equilibrium in vivo between the triplex (sticky) DNA structure and the orthodox B-DNA form of the GAA·TTC repeats? Since sticky DNA inhibits replication, transcription, recombination, and repair, cells could not survive or divide if this conformation was dominant and very long lived in vivo. Thus, what is the lifetime of this transient structure?
What is the role, if any, of heterochromatin in FRDA gene silencing?
What is the role of DNA methylation in FRDA?
What is the molecular biological mechanism in humans of the genetic instability of GAA·TTC? Is this process conducted by replication, recombination, or repair or a combination of all three mechanisms acting in concert?
What is the role in FRDA etiology, if any, of mutagenesis promoted by GAA·TTC in DNA tracts flanking the TRS?
Could genomic translocations (copy number variations, deletions, insertions, inversions, etc.) play a role, at least for some patients, in the mutation spectra? Many reviews and papers repeat the declaration that "98% of all FRDA cases are due to the expansion of the GAA·TTC repeats and 2% are due to point mutations" (1)
. Is this the entire range of mutations in the FXN gene or might future detailed analyses reveal the existence of other types of mutations that could be caused by the presence of long GAA·TTC repeats (see Mutagenesis)? Note that the fragile X mutation spectra is broad (reviewed in refs. 82
, 84
), as probably promoted by its highly mutagenic TRS.
What is the optimum therapeutic strategy and ideal agent for treating FRDA patients?
What are the functions of other moderate-length GAA·TTC repeats, which are shorter than in the FXN intron 1, in genomes? What are the roles of triplexes (sticky DNA) in these biological processes?
Obviously, substantial additional work remains to be conducted to unravel these critical questions in the future.
|
PROSPECTS FOR THE FUTURE |
|---|
. One conclusion from this review article is an agreement with this sentiment and an acknowledgment that substantial additional work needs to be concentrated on fundamental mechanisms responsible for the initial steps in the disease etiology. FRDA is a conceptually tractable disease for the reasons that are stated above (Therapeutic Strategies), since molecular events at the DNA level are primary, hence, the research focus should be localized on this initial step. Accordingly, any normalization of the TRS mutation should be focused at this primary step and not targeted to downstream events. It is critical to determine the fundamental molecular biological cause for transcription silencing (non-B DNA structures or heterochromatin or methylation or other mechanisms). Therapeutics for FRDA may be more advanced than for any other of the 30 hereditary neurological diseases caused by repeat expansion mechanisms.
Since the number of investigators conducting innovative research in this area has expanded greatly over the past decade (51)
, hope is raised for viable treatment and therapeutic strategies for the future. The development of high-throughput assays to evaluate the efficacy of a large number of ligands that may be therapeutically successful is currently underway.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 2, 2007. Accepted for publication December 27, 2007.
|
REFERENCES |
|---|
- and 
-diaminobutyric acid-linked hairpin polyamide-alkylator conjugates. Nucleic Acids Res 35,307-316This article has been cited by other articles:
|
|
 |
|
  L. E.L.M. Vissers, S. S. Bhatt, I. M. Janssen, Z. Xia, S. R. Lalani, R. Pfundt, K. Derwinska, B. B.A. de Vries, C. Gilissen, A. Hoischen, et al. Rare pathogenic microdeletions and tandem duplications are microhomology-mediated and stimulated by local genomic architecture Hum. Mol. Genet., October 1, 2009; 18(19): 3579 - 3593. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Torigoe, K. Sasaki, and T. Katayama Thermodynamic and Kinetic Effects of Morpholino Modification on Pyrimidine Motif Triplex Nucleic Acid Formation under Physiological Condition J. Biochem., August 1, 2009; 146(2): 173 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Wells Discovery of the Role of Non-B DNA Structures in Mutagenesis and Human Genomic Disorders J. Biol. Chem., April 3, 2009; 284(14): 8997 - 9009. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sureshkumar, M. Todesco, K. Schneeberger, R. Harilal, S. Balasubramanian, and D. Weigel A Genetic Defect Caused by a Triplet Repeat Expansion in Arabidopsis thaliana Science, February 20, 2009; 323(5917): 1060 - 1063. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Soragni, D. Herman, S. Y. R. Dent, J. M. Gottesfeld, R. D. Wells, and M. Napierala Long intronic GAA*TTC repeats induce epigenetic changes and reporter gene silencing in a molecular model of Friedreich ataxia Nucleic Acids Res., November 1, 2008; 36(19): 6056 - 6065. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
TOP
ABSTRACT
INTRODUCTION
