Chemotherapeutic deletion of CTG repeats in lymphoblast cells from DM1
http://www.100md.com
《核酸研究医学期刊》
Center for Genome Research, Institute of Biosciences and Technology, Texas A&M University System Health Sciences Center, 2121 West Holcombe Boulevard, Houston, TX 77030-3303, USA 1 Department of Neurology, Baylor College of Medicine, One Baylor Place, and VAMC, Houston, TX 77030, USA and 2 Department of Neurology, The University of Texas Medical Branch, 301 University Boulevard JSA9.128, Galveston, TX 77555-053, USA
* To whom correspondence should be addressed. Tel: +1 713 677 7664; Fax: +1 713 677 7689; Email: Rsinden@ibt.tamushsc.edu
Present addresses: Vera I. Hashem and Merhdad Khajavi, Department of Human and Molecular Genetics, Baylor College of Medicine, Houston, TX 77030, USA
Kuniko Tsuji, Department of Neurology, Osaka Medical College, Takatsuki 568-8686, Japan
ABSTRACT
Myotonic dystrophy type 1 (DM1) is caused by the expansion of a (CTG)·(CAG) repeat in the DMPK gene on chromosome 19q13.3. At least 17 neurological diseases have similar genetic mutations, the expansion of DNA repeats. In most of these disorders, the disease severity is related to the length of the repeat expansion, and in DM1 the expanded repeat undergoes further elongation in somatic and germline tissues. At present, in this class of diseases, no therapeutic approach exists to prevent or slow the repeat expansion and thereby reduce disease severity or delay disease onset. We present initial results testing the hypothesis that repeat deletion may be mediated by various chemotherapeutic agents. Three lymphoblast cell lines derived from two DM1 patients treated with either ethylmethanesulfonate (EMS), mitomycin C, mitoxantrone or doxorubicin, at therapeutic concentrations, accumulated deletions following treatment. Treatment with EMS frequently prevented the repeat expansion observed during growth in culture. A significant reduction of CTG repeat length by 100–350 (CTG)·(CAG) repeats often occurred in the cell population following treatment with these drugs. Potential mechanisms of drug-induced deletion are presented.
INTRODUCTION
The expansion of trinucleotide repeats and other short tandem DNA repeats is associated with an increasing number of inherited neurological diseases. These include myotonic dystrophy type 1 (DM1), DM2, Huntington's disease (HD), HD-Like 2, dentatorubral pallidoluysian atrophy, fragile X syndrome, FRAXE mental retardation, Kennedy's disease, oculopharyngeal muscular dystrophy, Friedreich ataxia and several spinocerebellar ataxias (SCAs) (SCA1, 2, 3, 6, 7, 8, 10, 12 and 17) (1–5).
Several common features are shared by most tandem repeat diseases. First, patients with most of these diseases exhibit an earlier age of onset and increased severity in successive generations, a phenomenon termed anticipation. This phenomenon is generally correlated with increasing repeat length (6). Second, expanded repeats in most of these diseases show both somatic and germline instability usually with a bias toward expansion. In transgenic animal models of these diseases, the longer repeat in the transgene led to increased severity and earlier onset (7–11). Therefore, the repeat length is critical in the pathogenic mechanism and is one of the most important determinants of disease severity.
Repeat expansion leads to the respective disease phenotypes through diverse mechanisms. For example, expanded CAG repeats encoding polyglutamine tracts lead to a gain of function of a protein as with HD, Kennedy's disease and several of the SCAs. A large (CGG)·(CCG) repeat expansion in the 5'-untranslated region (UTR) of the FMR1 and FMR2 genes and an expansion of an intronic (GAA)·(TTC) repeat in the FRDA gene result in a loss of function of the protein products in fragile X syndrome, FRAXE mental retardation and Friedreich ataxia, respectively (1). In SCA12, an expansion of CAG repeat in the 5'-UTR may cause enhanced transcription of the PPP2R2B gene (12). The pathogenic mechanisms of other repeat expansion diseases remain less well defined.
No treatment currently exists for DNA repeat expansion-associated neurological diseases. Because the molecular events downstream of repeat expansion mutations in these diseases are complex and variable, therapeutic strategies to intervene in such mechanisms would be difficult to formulate and must be tailor-made in each disease. Since the underlying etiology is the repeat expansion, treatment with chemicals or drugs that would reduce the repeat size would target the very source of the problem and may be generally applicable to delay onset or reduce the severity of diseases associated with repeat expansion. This may be especially important in diseases where somatic expansion during life may be associated with disease onset and severity.
DM1 is a multisystem disorder, characterized by muscle weakness, wasting and myotonia. DM1 results from an expansion of a CTG repeat in the 3'-UTR of the myotonic dystrophy protein kinase gene (DMPK). The expanded CTG repeat is unstable with a bias toward further expansion in germline and somatic tissues, giving rise to anticipation, and repeat size increases in tissues during life (13,14). The molecular mechanism downstream of the mutation in DM1 is complex and is associated with partial loss of DMPK, decreased transcription of a neighboring gene, Six5, in addition to binding of CUG-binding proteins by the expanded CUG mRNA (15). Thus, DM1 may be considered as a representative disease model to test the feasibility of chemical or drug treatments that decrease the repeat size.
The actual mechanism of tandem repeat expansion remains unclear. Several lines of evidence implicate various processes, including DNA replication, repair and recombination (2,16–18). One common feature of the repeats that may contribute to their instability is their ability to form various DNA secondary structures (2). CTG, CAG, CCG and CGG repeats form stable hairpins (19–21); (CTG)·(CAG) and (CGG)·(CCG) repeats form slipped-strand DNA structures caused by out-of-register misalignment within the repeat tract (22); (GGC)·(GGC) repeats form quadruplex DNA structures (23,24); and (GAA)·(TTC) repeats can form triplex structures and parallel stranded DNA (25–32). (ATTCT)·(AGAAT) repeats can form stable unpaired regions in DNA (33). A reasonable hypothesis remains that the formation of an alternative DNA secondary structure elicits repeat expansion during replication, repair, or recombination (2) or promotes aberrant DNA replication (33).
While mechanisms of repeat expansion are uncertain, mechanisms leading to (CTG)·(CAG) repeat deletion have been generally well accepted. A large body of evidence suggests that the formation of a single-stranded region containing the (CTG)n strand leads to the formation of hairpins that fail to be copied during DNA replication leading to large repeat deletions (17,34–38). Deletions have been thought to result from replication slippage across CTG hairpins formed in the lagging template strand when single-stranded during replication (2,17,34,39,40). Hairpins stabilize misaligned intermediates preferentially in the lagging strand and promote misalignment (41,42). Recent results, however, show that functional RecA and RecB are required for a high rate of repeat deletion, and it is difficult to incorporate these proteins into a simple replication slippage model for repeat deletion (43). An alternative model suggests that repeat deletion results from errors occurring during replication restart following the collapse of the replication fork during synthesis of the repeats via a pathway that involves a double-strand break (43). Unrepaired DNA damage will also stop the replication fork, causing fork collapse. In bacteria and yeast, replication through repeats can cause replication fork blockage (44–46). In yeast, repeats are associated with chromosome ‘fragility’, presumably a double-strand break that initiates recombination (37,45,47–52). Thus, DNA damage can block replication fork progression and lead to double-strand breaks in bacterial, yeast and human cells, where repair occurs by a variety of homologous and non-homologous based mechanisms (53–56). Moreover, many DNA damage repair processes involve the generation of a single-stranded gap in the DNA as a damaged base is removed (57). Thus, we hypothesize that the introduction of DNA damage to cells harboring (CTG)·(CAG) triplet repeats could lead to an increase in the rate of triplet repeat deletion both by creating single-strand regions and by causing replication fork collapse. Our results in a model bacterial system confirmed this hypothesis, showing that the DNA-damaging agents ethylmethanesulfonate (EMS), ethylnitrosourea, and ultraviolet (UV) light, and chemotherapeutic agents mitomycin C, cisplatin and X-rays increased the rate of deletion of (CTG)·(CAG) repeats (58). Moreover, oxidative damage to DNA also increased the rate of deletion of triplet repeats (58). These results provide the molecular basis to suggest that a chemotherapeutic approach to triplet repeat deletion may offer one rationale to slow, stop or reverse the progression of these diseases. In fact, as discussed below, recent experiments in several mammalian cell systems support our basic hypothesis (59–62).
Here, we treat lymphoblast cell lines derived from DM1 patients with DNA-damaging agents, including EMS, a monofunctional alkylating agent, and several chemotherapeutic agents, including mitomycin C, which introduces interstrand crosslinks; mitoxantrone, an inhibitor of topoisomerase II, which intercalates into DNA causing crosslinks and strand breaks; and doxorubicin, which intercalates into DNA and inhibits DNA synthesis. Mitomycin C, mitoxantrone and doxorubicin are currently used in the treatment of cancer and mitoxantrone is also used to treat multiple sclerosis. Our results indicate that drug treatment can prevent expansion commonly observed in lymphoblastoid DM1 cells. In many cases, a significant reduction of (CTG)·(CAG) repeat length occurred in the entire cell population following drug treatment.
MATERIALS AND METHODS
Cell growth
Clonal lymphoblast cell lines derived from DM1 patients have been described previously (63,64). Cell lines 313-1-15, 313-1-20 and 428 initially contained about 770, 740 and 440 (CTG)·(CAG) repeats, respectively (63). DM1 lymphoblast cells were grown between 2 x 105 and 2 x 106 cells/ml at 37°C with 5% CO2 in RPMI (Hi-Clone) media supplemented with 10% calf serum (Hi-Clone) and 1 μM insulin. Cells were grown for a defined number of generations for comparison. Generations is calculated as /log2, where DF is a dilution factor (calculated from the total number of times cells were diluted), NG is the total number of cells at the time of sampling, and N0 is the number of cells at the beginning of the experiment. Previously, growth has been reported as ‘passages’ (63). A passage represents 2–4 generations, when cells are diluted 1:4 or 1:8 for subsequent growth. The calculation of generations in our analysis does not take into account cell death in the population (only population doubling), and thus may represent a minimum or underestimate of the actual number of cell divisions at the time of harvest.
Treatment with drugs
DM1 lymphoblast cell lines were treated with drugs in growth media at 1 x 106 cells/ml for various times. Treatment with 500 or 700 μg/ml EMS or 0.1 or 0.2 μg/ml of mitomycin C was for 14–16 h, treatment with 50 or 100 nM mitoxantrone was for 1 h, while treatment with 0.1–5 μg/ml doxorubicin was for 30 min. Drug exposures were adjusted to reduce survival to 10–30%, except for treatment with mitoxantrone, which reduced survival to 30–50%. Following treatment, the cells were pelleted, washed once with media and resuspended in media at 1 x 106 cells/ml and incubated. Growth was generally continued for up to 10–30 generations. At selected passages aliquots of 1 x 107 cells were harvested by centrifugation for 5 min and frozen at –20°C until the genomic DNA was isolated for small pool PCR. In certain experiments, a second dose of the drug was administered in the same manner as above after recovery of the culture.
DNA purification and PCR analysis
For allele length analyses, genomic DNA was purified from 5–10 x 106 lymphoblast cells using a DNeasy Tissue Kit (Qiagen) followed by one or two ethanol precipitations. For small pool PCR, reactions contained 50–2000 pg of HindIII-digested DNA corresponding to 1–50 diploid genome equivalents per reaction. Two different conditions were used with identical results. First, PCR was performed according to the procedure described by Monckton et al. (13). In addition, PCR was performed as follows: 0.5 U Platinum Taq DNA Polymerase High Fidelity (Invitrogen), 0.2 μM primer DM-DR (dCAGGCCTGCAGTTTGCCCATC) and 0.2 μM primer DM-H (dTCTCCGCCCAGCTCCAGTC), 2 mM MgSO4, 0.2 mM dNTPs, in a total volume of 10 μl. PCR was performed in a Perkin–Elmer 9700 PCR machine with the following conditions: the DNA was denatured at 94°C for 6 min and then subjected to 30–35 cycles of 94°C for 45 s, 67°C for 45 s, 70°C for 2 min 40 s, with a final extension step at 70°C for 35 min.
Southern hybridization
PCR products were separated by electrophoresis through 1.8% agarose gels in 0.5x TBE buffer (1x TBE: 90 mM Tris, 90 mM Borate, 2.5 mM EDTA, pH 8.3), transferred to a Hybond N membrane (Amersham), the DNA fixed by UV irradiation and then the membrane was hybridized with an (AGC)10 oligonucleotide end-labeled with ATP. Hybridization was performed for 16 h at 42°C in 1x SSC buffer. Membranes were then washed twice for 30 min at 49°C in 0.5x SSC, 0.1% SDS, rinsed in H2O and then exposed for 1–2 days to a Molecular Dynamics PhosphorImager plate. Alleles were sized by comparison with molecular weight markers. Lambda/HindIII/pUC19/TaqI/Sau3A1 molecular weight markers (Invitrogen) were labeled with 32P as follows: 1 μg of DNA was added to 10 U Klenow (New England Biolabs), 1 mM dGTP, 1 mM dCTP, 1 mM dTTP and 50 μCi (3000 Ci/mM) dATP, in 20 μl Klenow buffer and the mixture was incubated 15 min at 20°C and then 15 min at 65°C. The number of repeats in individual alleles was calculated from the length of the PCR product and the DNA sequence flanking the (CTG)·(CAG) repeats.
Statistical analysis
For statistical analysis, chi-square tests were calculated for each set of experiments. The numbers of original-length alleles and expansions and deletions were compared between control, non-treated cells and drug-treated cells. Statistical results for a single regiment of drug treatment for both cell lines were then analyzed using a modified Bonferroni comparison to control for type 1 error. The significance following this comparison is reported in the results section.
RESULTS
Repeat instability in lymphoblast cell lines 428 and 313-1-15
Growth of lymphoblast cell lines through many generation is accompanied by expansion of the expanded (CTG)·(CAG) repeat, at a variable magnitude and frequency (63,64). (CTG)·(CAG) repeat lengths were analyzed by small pool PCR using a well-established procedure that can easily detect more than 1000 (CTG)·(CAG) repeats (13,14,63–67). We have analyzed the effect of treatment with a common alkylating agent and three chemotherapeutic drugs on repeat instability. Two of these agents, EMS and mitomycin C, increased the rates of (CTG)·(CAG) deletion in a bacterial model system (17). Cell line 313-1-15 contains about 770 (CTG)·(CAG) repeats that remain relatively stable but do show a slight and gradual increase in the repeat size during growth (data not shown). The initial culture of cell line 428 showed a distribution of allele sizes of about 440 ± 100 repeats (Figure 1A). After growth for 27 generations in one experiment (experiment A), two allele sizes dominated the cell population: 440 and 770 repeats (Figure 1A, experiment A). In a subsequent experiment, which was started from a sub culture of experiment A after five generations growth, the 770 repeat allele dominated the population following an additional 17 generations of growth (experiment B). Figure 1B shows the percentage of alleles plotted as a function of the number of (CTG)·(CAG) repeats for many repetitions of the PCR reactions as shown in Figure 1A. To confirm the rapid expansion in cell line 428, cells from the initial distribution population shown in Figure 1A were sub-cloned by growth of an initial 100–300 cells for 18–20 generations. All populations analyzed showed predominantly the larger 770 repeat allele. Six of nine sub-clones analyzed showed only the 770 repeat allele, while three showed 94–97% the 770 repeat allele (data not shown). Clearly, during these drug experiments cell line 428 was prone to repeat expansion in control, non-treated cultures. Control, of non-treated cultures of cell line 428 was repeated more than 10 times. At later passages, after the repeat had expanded to >700 repeats, spontaneous reductions in allele sizes in the population were never observed without drug treatment. The behavior of this cell line mimics events occurring in certain DM1 tissues, where repeat expansion continues throughout life (13). The rapid change in repeat lengths is presumably due to cells with larger (CTG)·(CAG) repeats having a growth advantage over those with smaller expanded alleles (64). Given the potential for rapid change in cell line 428, it seemed that it may be sensitive to the effects of DNA-damaging agents. On the other hand, cell line 313-1-15 was used for analysis because it is relatively stable and the accumulation of small deletions may be more easily observed in this line. Following growth of this cell line in 17 independent experiments, deletions comprised <2% of the allele population in the absence of drug treatment, and a large deletions of the repeat allele in the entire population were never observed.
Figure 1. Expansion of (CTG)·(CAG) repeats in cell line 428. (A) Representative small pool PCR reactions with 300 pg DNA are shown for cell line 428 for an initial early passage culture and following growth. In experiment A, the cells were grown for 27 generations. Experiment B was started from sub culture of cells from experiment A after five generations growth, and growth was continued for an additional 17 generations. Size markers are indicated relative to the number of (CTG)·(CAG) repeats in the PCR products. Under the PCR conditions used, 300 pg corresponds to 6–12 genome equivalents per reaction. (B) Allele spectra for 428 cultures at the time of initial growth (panel a) n = 151; experiment A after 27 generations growth (panel b), n = 320; and experiment B after 17 generations growth (panel c), n = 515. n = the number of alleles analyzed.
Treatment with the alkylating agent EMS results in the accumulation of the 440 repeat allele in cell line 428
Lymphoblast cell line 428 was treated with 500 and 700 μg/ml EMS and grown for 10–20 generations. Small pool PCR analysis of alleles shows that the treatment resulted in the accumulation of much smaller allele sizes than present in the non-treated cells grown relatively for the same number of generations (Figure 2). Treatment with 500 μg/ml EMS typically showed the accumulation of two or three repeat sizes (440, 550 and 770 repeats) (Figure 3). When cells were treated with a second dose of EMS, the accumulation of the smaller sized alleles was more pronounced (Figure 3A, panel b and Figure 3B, panel c). Treatment with 700 μg/ml EMS generally resulted in the accumulation of the 440 repeat allele size, a reduction in size of 330 repeats relative to the expansion observed in non-treated cells. In the experiment shown in Figure 3A, the difference in the allele distributions for both drug-treated samples relative to the non-treated distribution was significant (P 0.001) using a modified Bonferroni comparison. In the experiment shown in Figure 3B, while the single treatment of 500 μg/ml EMS resulted in the accumulation of many shorter sized deletions (panel b), the distribution was not significantly different from the control, non-treated distribution comparing the numbers of 440 repeat alleles and other-size alleles (P = 0.072). In this experiment, distributions following the second dose of 500 μg/ml EMS treatment and the single dose of 700 μg/ml EMS treatment (panels c and d) were significantly different (P 0.001). Experiments involving cell growth following treatment with EMS were repeated six times for this cell line with essentially identical results. Occasionally (in <10% of drug-treated samples) an EMS-treated sample with predominantly the expanded (770 repeat) allele was observed. It is important to reiterate that, from independent analyses of the growth of many cell lines and many individual growth analyses, large deletions, especially reductions in the repeat tract in the entire population, were not observed in this, or previous studies (63,64). Thus, the large repeat reductions observed in the population are highly significant and drug dependent.
Figure 2. PCR analysis of the accumulation of deletions in 428 cells following treatment with EMS or mitomycin C. Small pool PCR reactions with genomic DNA are shown for control (non-treated) cells and for cells treated with 0.2 μg/ml mitomycin C or 700 μg/ml EMS. DNA was purified 12–17 generations after drug treatment. PCR reactions contained either 50 or 300 pg DNA, as indicated. Molecular weight markers are indicated as the number of (CTG)·(CAG) repeats.
Figure 3. Accumulation of deletions in 428 cells following EMS treatment. (A) Experiment started with 428 cells using the population of alleles shown in Figure 1B, panel a, at the time of treatment. Panel a, allele distribution following 27 generations growth in control, non-treated cells. Panel b, allele distribution following a second treatment with 500 μg/ml EMS and 16 generations growth (n = 362). Panel c, allele distribution following a single treatment with 700 μg/ml EMS and 18 generations growth (n = 323). In both drug-treated samples, the allele distribution was significantly different from the control, non-treated sample using a chi-square test (P 0.001). (B) Experiment started with 428 cells (as described in text). Panel a, allele distribution following an additional 17 generations growth in control, non-treated cells. Panel b, allele distribution following a single treatment with 500 μg/ml EMS, 18 generations growth (n = 196). Panel c, allele distribution following two treatments with 500 μg/ml EMS, 12 generations growth (n = 269). Panel d, allele distribution following a single treatment with 700 μg/ml EMS, 16 generations growth (n = 603). The allele distribution in panel b was not significantly different from the control (P = 0.071). In panels c and d, the allele distributions were significantly different from the control, non-treated sample using a chi-square test (P 0.001).
Treatment with mitomycin C results in the accumulation of (CTG)·(CAG) deletions
Lymphoblast cell line 428 and 313-1-15 were treated with 0.1 or 0.2 μg/ml mitomycin C and grown following drug exposure. A dramatic effect was observed in both cell lines following treatment with mitomycin C. In the experiment shown in Figures 2 and 4A for cell line 428, a broad range of repeat sizes accumulated following drug treatment. Little of the original 440 repeat size allele was observed in cells treated once or twice with 0.1 μg/ml mitomycin C (Figure 4A, panels b and c). Following the second treatment (Figure 4A, panel c), a broad distribution of allele sizes was observed (from 300 to 675 repeats), corresponding to reductions of 100–470 repeats relative to non-treated cells. Compared with the control non-treated distribution, the allele spectra for cells treated twice with mitomycin C were significantly different (Figure 4A, panel b, P 0.001; panel c, P = 0.002). The allele distribution for cells treated once with 0.2 μg/ml mitomycin C was not significantly different from control considering two populations of alleles (the initial 440 repeat size and all other expansions or deletions). Considering three populations of alleles (440 repeats, 770 repeats, and all other expansions and deletions), the allele spectra were significantly different (P 0.001). This data is representative of five independent experiments with cell line 428.
Figure 4. Accumulation of deletions following treatment with mitomycin C. (A) Mitomycin C treatment of cell line 428. Panel a, allele distribution following 17 generations growth in control, non-treated cells (n = 515). Panel b, allele distribution following treatment with 0.1 μg/ml mitomycin C and 18 generations growth (n = 192). Panel c, allele distribution following a second treatment with 0.1 μg/ml mitomycin C, 16 generations growth (n = 52). Panel d, allele distribution following a single treatment with 0.2 μg/ml mitomycin C, 12 generations growth (n = 231). In panels b and c, the allele distributions were significantly different from the control, non-treated sample using a chi-square test (P 0.001 and P = 0.002, respectively). Results in panel d were not statistically different considering the fraction of starting length (440 repeat) alleles, although the overall distribution is significantly different. (B) Mitomycin C treatment of cell line 313-1-15. Panel a, typical allele distribution for 313-1-15 showing predominantly the full length expanded allele of about 770 repeats after 11 generations growth in control, non-treated cells (n = 503). Panel b, allele distribution following a single treatment with 0.1 μg/ml mitomycin C, 9 generations growth (n = 460). Panel c, allele distribution following two treatments with 0.1 μg/ml mitomycin C, 4–5 generations growth (n = 559). Panel d, allele distribution following a single treatment with 0.2 μg/ml mitomycin C, 4–5 generations growth (n = 401). The allele distribution in panel b was not significantly different from the control (P = 0.096). In panels c and d, the allele distributions were significantly different from the control, non-treated sample using a chi-square test (P 0.001 and P = 0.052).
Similar results were observed for cell line 313-1-15 containing the relatively stable 770 (CTG)·(CAG) repeat allele (Figure 4B). During growth for 5–10 generations following drug treatment, an increase in the frequency of deletions was observed for some samples in this cell line, and alleles with 100 repeats, or fewer, were observed. Following a single exposure to 0.1 μg/ml mitomycin C (Figure 4B, panel b), the frequency of deletions was not significantly different from that in control cells. However, following a second treatment of 0.1 μg/ml mitomycin C (Figure 4B, panel c), the allele spectra were significantly different, after only 4–5 generations growth, with a P-value of <0.001. Following a single treatment with 0.2 μg/ml mitomycin C (Figure 4B, panel d), the difference in allele spectra was not significantly different (P = 0.052). In a separate experiment, a deletion from 770 to 675 repeats was observed in the entire population of 313-1-15 cells following treatment with 0.2 μg/ml mitomycin C (Figure 5). In this experiment, the allele spectra following 17 and 19 generations growth showed a reduction to 675 repeats (Figure 5, panels b and c). Both allele distributions were significantly different from non-treated, control cells (P 0.001). Experiments with cell line 313-1-15 and mitomycin C treatment were repeated 12 times. In another experiment using cell line 313-1-20, a large deletion of about 350 repeats was observed in 10–15% of the culture following treatment with 0.2 μg/ml mitomycin C (data not shown). Experiments with cell line 313-1-20 and mitomycin C treatment were repeated four times.
Figure 5. Accumulation of a 100 repeat deletion during growth in a population of 313-1-15 cells following treatment with mitomycin C. In this experiment, 313-1-15 cells were treated with mitomycin C at 0.2 μg/ml and growth was continued for 17–19 generations. Panel a, allele distribution for 313-1-15 showing predominantly the full-length expanded allele of 770 repeats after 19 generations growth (n = 229). Panel b, allele distribution following 17 generations growth (n = 135). Panel c, allele distribution following 19 generations growth (n = 178). The differences in allele distribution in panels b and c were significantly different from that in the control sample (P 0.001).
Mitoxantrone induces deletions in DM1 lymphoblast cells
A culture of lymphoblast cell line 428, in which the repeat had expanded to 740 (CTG)·(CAG) repeats (>98% of alleles), was treated with 50 and 100 nM mitoxantrone. The allele spectra following treatment with 50 nM mitoxantrone were not significantly different from that for control, non-treated cells after 14 generations growth (Figure 6, panel b). However, after 14 generations a significant shift in the population to about 620 repeats was observed with the sample treated with 100 nM mitoxantrone (P 0.001) (Figure 6, panel c). Eight independent cultures of 428 and 313-1-15 cells have been treated with mitoxantrone. The data shown in Figure 6 represent the most dramatic response.
Figure 6. Accumulation of a 120 repeat deletion during growth in a population of 428 cells following treatment with mitoxantrone. Cell line 428 was treated with mitoxantrone at 50 or 100 nM and growth was continued for 12–14 generations. Panel a, allele distribution in control, non-treated 428 cells. A subculture of cells in which a 770 repeat allele size dominated the population was used and grown for 12 generations (n = 232). Panel b, allele distribution after treatment with 50 nM and subsequent growth for 14 generations (n = 146). The allele distribution was not significantly different from that of control cells. Panel c, allele distribution following treatment with 100 nM and subsequent growth for 14 generations (n = 301). This difference in this allele distribution was statistically significant compared with the distribution in the control, non-treated sample (P 0.001).
Doxorubicin induces deletions in DM1 cells
DM1 lymphoblast cells were treated with several doses of doxorubicin and growth was continued for 10 generations. Figure 7 (top) shows representative data from one of the four independent experiments in cell line 313-1-15. The rate of population doubling was reduced in a dose-dependent fashion following exposure to doxorubicin. Allele distributions, after 10–16 generations growth, showed the accumulation of deletion mutations in 10–20% of the allele population in a dose-dependent fashion that paralleled the initial decrease in survival and growth rate (Figure 7, bottom). In this experiment, the expanded allele measured 840 repeats and most deleted alleles were between 350 and 650 repeats, or deletions of 200–500 repeats. The allele distributions following treatment with 1, 2 and 5 μg/ml doxorubicin were significantly different from that in the control distribution (P 0.001).
Figure 7. Accumulation of deletions following treatment with doxorubicin. Top, growth of 313-1-15 cells following treatment with doxorubicin. Cells were treated with 0.1, 0.5, 1.0, 2.0 or 5.0 μg/ml doxorubicin for 30 min and then growth was monitored, and samples harvested for 10–16 generations. Bottom, allele distributions. Panel a, control, non-treated cells, 10 generations growth (n = 170). Panel b, 0.1 μg/ml doxorubicin, 10 generations growth (n = 401). Panel c, 0.5 μg/ml doxorubicin, 11 generations growth (n = 314). Panel d, 1.0 μg/ml doxorubicin, 13 generations growth (n = 233, P = 0.031). Panel e, 2.0 μg/ml doxorubicin, 16 generations growth (n = 638, P 0.001). Panel f, 5.0 μg/ml doxorubicin, 13 generations growth (n = 485, P 0.001).
DISCUSSION
Our results demonstrate that lymphoblast cells, derived from patients with DM1, which contain expanded (CTG)·(CAG) repeats in the DMPK gene, accumulate smaller (CTG)·(CAG) repeat alleles following treatment with EMS, mitomycin C, mitoxantrone or doxorubicin. Reductions in the length of the repeat from 770 to 100 or fewer (CTG)·(CAG) repeats were frequently observed following drug treatment. In many cases, the length of the expanded repeat was reduced from 770 to 650–440 (CTG)·(CAG) repeats in the entire population of cells. This is a significant result given the propensity of the DM1 (CTG)·(CAG) repeats to expand rather than contract in patients and in culture, as shown here and previously (13,14,64).
Multiple events may be contributing to the reductions in repeat tract length: drug-induced deletion and clonal selection. Events associated with DNA repair or drug-induced replication fork collapse would be expected to result in deletions, and potential mechanisms are discussed below. The mixture of shorter alleles observed following treatment with mitomycin C or doxorubicin likely arise from multiple independent deletions within the population. Clonal selection after drug treatment may also be occurring within the population. As occurs in cancer chemotherapy, drugs may be selectively killing cells in the population with larger numbers of repeats, which have a more rapid rate of cell division than cells with the shorter alleles (64). This would then select for the growth of cells with shorter alleles. Some evidence for clonal selection has been observed following exposure of mouse fibroblasts containing expanded CTG repeats to various chemicals (62). However, alterations in repeat length in mouse and human cells following treatment with various chemicals and replication inhibitors can occur independent of clonal expansion (61,62). In our experiments, clonal expansion is possible when survival was much reduced. However, in many experiments, smaller sized alleles dominated the population prior to the time one would expect population outgrowth of a shorter sized allele from a single clonal cell. From a single cell, 24–25 population doublings are required to generate 107 cells. In the mitoxantrone experiment (Figure 6), in which the starting culture had >98% alleles with 740 repeats, the shorter 620 repeat allele would only generate 8.2 x 103 cells after 14 generations growth (0.1% of the cell population considering analysis of 107 cells) and by this time a complete shift in the population had occurred. The >300 repeat reduction observed in the entire population in cell line 428 in earlier passages when a mixture of allele sizes was present may represent clonal selection. It is important to mention, however, that due to the strong drive toward expansion in the 428 cells, the larger alleles will naturally dominate the population. Mixing experiments using cell lines with different allele sizes have shown that the cell line with the longer allele will dominate the population (64). We also showed that when the ‘initial culture’ of 428 shown in Figure 1 was sub-cloned, all populations analyzed showed predominantly the larger 770 repeat allele by 18–20 generations. Six of nine sub-clones showed only the 770 repeat allele, while three showed 94–97% the 770 repeat allele (data not shown). The drive in the cell culture is toward accumulation of larger alleles. Thus, the accumulation of shorter alleles in the population is highly unusual in the growth of DM1 cell cultures. Therefore, it is unlikely that the domination of the population by the 440 repeat allele following drug treatment is not related to treatment. The important caveat in these experiments is that certain drug treatments can, whether through frequent independent deletions and/or clonal selection, promote large deletions in the population. The outcome is significant and demonstrates a proof of principle in testing our hypothesis.
The data do not allow us, at present, to establish the mechanism(s) responsible for repeat deletion. Several phenomena may be contributing to the observed repeat deletion, and work is in progress to further understand the molecular basis for the drug-induced deletion. The working hypothesis is that DNA damage and/or replication fork collapse at or near the repeats elicits deletion. Deletions may arise through several different pathways. First, results using bacteria suggest that spontaneous deletions occur during a RecA and RecBC pathway of replication restart, in which the introduction of a double-strand break is associated with a high rate of repeat deletion (17,43). As shown in Figure 8A, replication fork collapse resulting from a DNA damage in the leading strand could lead to deletions. Evidence in yeast strongly supports the frequent introduction of double-strand breaks during DNA replication through (CTG)·(CAG) repeats (37,45,47–52). The addition of DNA damage would be expected to increase the frequency of replication fork collapse and the occurrence of double-strand breaks. Second, the formation of CTG hairpins within repair gaps or in single-strand DNA associated with replication restart may also promote repeat deletion, as is believed to occur in bacterial model systems (2,43,58) (Figure 8B). In bacterial cells, the generation of single-stranded DNA greatly increases the probability of deletion of CTG repeats (17,58,68). Single-stranded DNA is generated in different steps of many repair pathways or during replication restart (57,69). Thus, it would be expected that repair of DNA damage by an excision repair pathway could increase the chance of deletion at this DNA repeat hotspot for mutation by allowing the formation of a (CTG)n or (CAG)n hairpin in the single-strand gap formed during repair. Evidence from work with yeast and human cells is consistent with instability occurring as a result of replication encountering a hairpin (37,40,51,52,70–75). The reduction in repeat lengths in 428 and 313-1-15 cells by 100–120 triplet repeats may reflect the size of a preferential hairpin formation. In bacteria, deletion to preferred sizes has been observed (34,68), as has preferential complete deletion of certain sizes repeat tracts (17). The repeated observation of a >300 repeat deletion in 428 cells may also reflect preferential deletion, although, as discussed above, this may also be due to population selection. Third, it is also possible that DNA damage could initiate the formation of a double-strand break, separate from a replication restart pathway (Figure 8C). Following end resection and homologous end joining, DNA repeats would be lost. In fact, repair of a double-strand break within (CTG)·(CAG) repeats in mammalian cells is associated with repeat deletion (75). An additional possibility is that treatment with DNA damage acts in a non-targeted fashion by inducing a generalized state of instability in cells, which results in deletion at the expanded (CTG)·(CAG) repeats. Persistent genomic instability occurring more than 30 generations following exposure to ionizing radiation is a well-documented phenomenon in cultured cells (76–78). In fact, transgenerational genomic instability following exposure to X-rays has been observed in mice (79).
Figure 8. Models for drug-induced repeat deletion. Left panel. (A) Replication restart. DNA damage within a (CAG)n tract in the leading template strand, indicated by the asterisk, may cause DNA synthesis arrest and replication fork collapse. The dark thick region represents a (CTG)n tract, while the lighter thick region represents the complementary (CAG)n strand. Steps a–i are based on the understanding of the pathway for replication fork restart that involves a double-strand break. This pathway appears to be a major source of spontaneous deletion in E.coli (17,43). Following fork collapse (b and c), fork reversal forms a ‘chicken-foot’ structure (d). If cleaved by a Holliday junction resolvase (at the arrows), a double-strand break ensues (e). End resection by a nuclease will lead to a single-strand 3' end (f), which can initiate homologous strand invasion. During this process, misalignment can occur resulting in a (CAG)n hairpin (g). Branch migration will generate a Holliday junction (h), the resolution of which will restore the replication fork (i). Repeat deletion can occur at the point of misalignment at strand invasion or during synthesis of the restored lagging strand, in which the (CTG)n tract may form a stable hairpin. Right panel, following the introduction of a DNA damage within the repeat, two DNA repair pathways could lead to repeat deletion. (B) Repair involving a gap. An excision repair pathway may open up a gap (c) in which a (CTG)n hairpin may form (d). Slipped misalignment with DNA replication would lead to repeat deletion in the nascent strand (e). Excision of the loop or a subsequent round of replication would lead to repeat deletion. (C) Double-strand break repair. A pathway involving a double-strand break within the repeat (c) may also lead to repeat deletion. Following end resection to generate complementary single strands (d), homologous pairing could occur with flaps (e) that, on removal and ligation, would lead to repair with deletion (f).
The results presented here represent a ‘proof of principle’ that a chemotherapeutic approach to reduction in repeat length is feasible, although several issues must be addressed before this approach can be considered for clinical use. The concentrations of chemotherapeutic agents used in these experiments, 50–100 nM for mitoxantrone, 0.1–0.2 μg/ml for mitomycin C and 0.1–5 μg/ml doxorubicin, are within the therapeutic ranges, although EMS, a potent mutagen, cannot be used clinically. Our data are limited to (CTG)·(CAG) repeat expansion at the DM1 locus. The drugs used may have varying effects with repeats associated with other repeat diseases. The drug effect may differ depending on repeat unit motifs, the repeat length, cis elements adjacent to the repeat and trans factors that may influence instability. In fact, in a similar study of lymphoblast cells containing 124 (CTG)·(CAG) repeats, mitomycin C treatment resulted in an increase in repeat length (59). This result may be due to the large differences in repeat length , or other trans-acting factors that may influence the propensity of expansion to occur in this particular cell line. Whether the same chemotherapeutic agents promote deletions in other repeats at different loci remains to be shown. Nevertheless, the ‘proof of principle’ shown in this study raises a possibility that similar approaches might work for repeat expansions of other diseases.
Our results agree favorably with several other studies showing that DNA-damaging agents, various chemicals and replication inhibitors can influence repeat instability (59–62). Two studies have shown that agents that influence DNA replication or repair can influence repeat instability. Gorbunova et al. (60) showed that the DNA replication inhibitor aphidicolin increased the rate of deletions in CHO cells containing 95 but not 61 (CTG)·(CAG) repeats, while Yang et al. (61) showed that replication inhibitors aphidicolin and emetine, but not mimosine, increased the rate of CTG expansion in DM fibroblast cells. Both studies are consistent with hypothesis that DNA replication is involved in repeat expansion. Gomes-Pereia and Monckton (62) showed that long-term exposure of kidney cells from a transgenic mouse carrying 160 CTG repeats to cytosine arabinoside, ethidium bromide, 5-azacytidine and aspirin reduced the rate of repeat expansion, while exposure to caffeine increased the rate of expansion. Our approach is a bit different in that we tested the effect of chemotherapeutic agents known to damage DNA on the instability of CTG repeats in human lymphoblast cells from DM1 patients. While we saw dramatic reductions in repeat tract length and the accumulation of deletions in the cell population, our results with mitomycin C were in contrast with that of Pineiro et al. (59). This difference may result from the short allele size examined, the potential for repeat expansion in their cell line or other differences with experimental protocols. Thus, at present a general consensus on the effects of chemotherapeutic agents, DNA-damaging agents, or replication inhibitors on repeat instability is not yet available from these independent studies with natural DM1 alleles in human lymphoblast or fibroblast cells, or transgenic (CTG)·(CAG) repeats in CHO and mouse cells. However, the basic idea that the length of expanded DNA repeats associated with human genetic disease can be modulated by various treatments seems well established.
While currently used chemotherapeutic regimes may provide a viable rationale for treatment, drugs that react with high specificity to expanding repeats may be designed based on the structure and characteristics of the DNA helix or the unusual properties of these repeats. CTG and CGG tracts exist as an unusually flexible double helix (80–82), and they can form mismatched hairpins (20,21) and slipped strand DNA structures (83–85). GAA repeats can form intramolecular triplex DNA (25–32) and a bimolecular triplex structure (29). Thus, it is possible that drugs that selectively bind to these sequences and mediate deletion in long repeat tracts may eventually be identified, increasing the selectivity of this basic approach. An additional approach would be to couple a DNA-damaging agent to an oligonucleotide or a PNA (peptide nucleic acid) molecule complementary to a repeat since these can bind very tightly to DNA. This approach has proved successful in directing mutations to DNA adjacent to triplex-forming sequences (86–88). A PNA molecule complementary to CTG repeats has been shown to modulate gene expression in cells (89). A targeted approach to repeat damage may also be important in tissue where cell division may be limited, e.g. muscle or brain. Most model systems examined to date have analyzed repeat lengths in dividing cells . However, it is important to note that expansions are observed in the brain of HD patients (90), and one report has suggested that expansion can occur in sperm during maturation (91). Thus, repeat deletion may be expected to occur in non-dividing cells during chromosome repair and maintenance. Moreover, many of the late onset and complex characteristics of DM1 may result from, or be aggravated by, somatic expansion that occurs throughout life (13,14).
In conclusion, DNA repeat expansion diseases form a large class of genetic disorders linked mechanistically by a single type of mutation, the expansion of (CTG)·(CAG), (CGG)·(CCG), (GAA)·(TTC), (CCTG)·(CAGG) or (ATTCT)·(AGAAT) repeats. All known repeat expansion diseases cause progressive neurological manifestations that often lead to total disability and death. At present, no treatment can delay, halt or reverse the progression of these diseases. The application of a chemotherapeutic drug approach may open a gate to a new therapy for the underlying etiology of these diseases.
ACKNOWLEDGEMENTS
We thank William Ondo for helpful comments and suggestions. This work was supported by grants from NIEHS (R.R.S.) and NIH (T.A.).
REFERENCES
Wells,R.D. and Warren,S.T. ( (1998) ) Genetic Instabilities and Hereditary Neurological Diseases. Academic Press, San Diego, CA.
Sinden,R.R. ( (1999) ) Biological implications of the DNA structures associated with disease-causing triplet repeats. Am. J. Hum. Genet., , 64, , 346–353.
Cummings,C.J. and Zoghbi,H.Y. ( (2000) ) Fourteen and counting: unraveling trinucleotide repeat diseases. Hum. Mol. Genet., , 9, , 909–916.
Heintz,N. and Zoghbi,H.Y. ( (2000) ) Insights from mouse models into the molecular basis of neurodegeneration. Annu. Rev. Physiol., , 62, , 779–802.
Cleary,J.D. and Pearson,C.E. ( (2003) ) The contribution of CIS-elements to disease-associated repeat instability: clinical and experimental evidence. Cytogenet. Genome Res., , 100, , 25–55.
Ashizawa,T. and Zoghbi,H.Y. ( (1997) ) Diseases with trinucleotide repeat expansion. In Appel,S.H. (ed.), Current Neurology. IOS Press, Amsterdam, Vol. 17, pp. 79–135.
Sato,T., Oyake,M., Nakamura,K., Nakao,K., Fukusima,Y., Onodera,O., Igarashi,S., Takano,H., Kikugawa,K., Ishida,Y. et al. ( (1999) ) Transgenic mice harboring a full-length human mutant DRPLA gene exhibit age-dependent intergenerational and somatic instabilities of CAG repeats comparable with those in DRPLA patients. Hum. Mol. Genet., , 8, , 99–106.
Monckton,D.G., Coolbaugh,M.I., Ashizawa,K.T., Siciliano,M.J. and Caskey,C.T. ( (1997) ) Hypermutable myotonic dystrophy CTG repeats in transgenic mice. Nature Genet., , 15, , 193–196.
Fortune,M.T., Vassilopoulos,C., Coolbaugh,M.I., Siciliano,M.J. and Monckton,D.G. ( (2000) ) Dramatic, expansion-biased, age-dependent, tissue-specific somatic mosaicism in a transgenic mouse model of triplet repeat instability. Hum. Mol. Genet., , 9, , 439–445.
Mangiarini,L., Sathasivam,K., Mahal,A., Mott,R., Seller,M. and Bates,G.P. ( (1997) ) Instability of highly expanded CAG repeats in mice transgenic for the Huntington's disease mutation. Nature Genet., , 15, , 197–200.
Seznec,H., Lia-Baldini,A.S., Duros,C., Fouquet,C., Lacroix,C., Hofmann-Radvanyi,H., Junien,C. and Gourdon,G. ( (2000) ) Transgenic mice carrying large human genomic sequences with expanded CTG repeat mimic closely the DM CTG repeat intergenerational and somatic instability. Hum. Mol. Genet., , 9, , 1185–1194.
Holmes,S.E., Hearn,E.O., Ross,C.A. and Margolis,R.L. ( (2001) ) SCA12: an unusual mutation leads to an unusual spinocerebellar ataxia. Brain Res. Bull., , 56, , 397–403.
Wong,L.-J.C., Ashizawa,T., Monckton,D.G., Caskey,C.T. and Richards,C.S. ( (1995) ) Somatic heterogeneity of the CTG repeat in myotonic dystrophy is age and size dependent. Am. J. Hum. Genet., , 56, , 114–122.
Martorell,L., Monckton,D.G., Gamez,J., Johnson,K.J., Gich,I., de Munain,A.L. and Baiget,M. ( (1998) ) Progression of somatic CTG repeat length heterogeneity in the blood cells of myotonic dystrophy patients. Hum. Mol. Genet., , 7, , 307–312.
Wansink,D.G. and Wieringa,B. ( (2003) ) Transgenic mouse models for myotonic dystrophy type 1 (DM1). Cytogenet. Genome Res., , 100, , 230–242.
Oussatcheva,E.A., Hashem,V.I., Zou,Y., Sinden,R.R. and Potaman,V.N. ( (2001) ) Involvement of the nucleotide excision repair protein UvrA in instability of (CAG)·(CTG) repeat sequences in E.coli. J. Biol. Chem., , 276, , 30878–30884.
Hashem,V.I., Rosche,W.A. and Sinden,R.R. ( (2002) ) Genetic assays for measuring rates of (CAG)·(CTG) repeat instability in Escherichia coli. Mutat. Res., , 502, , 25–37.
Jakupciak,J.P. and Wells,R.D. ( (1999) ) Genetic instabilities in (CTG)·(CAG) repeats occur by recombination. J. Biol. Chem., , 274, , 23468–23479.
Pearson,C.E. and Sinden,R.R. ( (1998) ) Trinucleotide repeat DNA structures: Dynamic mutations from dynamic DNA. Curr. Opin. Struct. Biol., , 8, , 321–330.
Mitas,M. ( (1997) ) Trinucleotide repeats associated with human disease. Nucleic Acids Res., , 25, , 2245–2253.
Darlow,J.M. and Leach,D.R.F. ( (1998) ) Secondary structures in d(CGG) and d(CCG) repeat tracts. J. Mol. Biol., , 275, , 3–16.
Pearson,C.E. and Sinden,R.R. ( (1998) ) Slipped strand DNA, dynamic mutations, and human disease. In Wells,R.D. and Warren,S.T. (eds.), Genetic Instabilities and Hereditary Neurological Disorders. Academic Press, San Diego, CA, pp. 585–621.
Usdin,K. and Woodford,K.J. ( (1995) ) CGG repeats associated with DNA instability and chromosome fragility form structures that block DNA synthesis in vitro. Nucleic Acids Res., , 23, , 4202–4209.
Fry,M. and Loeb,L.A. ( (1994) ) The fragile X syndrome d(CGG)n nucleotide repeats form a stable tetrahelical structure. Proc. Natl Acad. Sci. USA, , 91, , 4950–4954.
Bidichandani,S.I., Ashizawa,T. and Patel,P.I. ( (1998) ) The GAA triplet-repeat expansion in Friedreich ataxia interferes with transcription and may be associated with an unusual DNA structure. Am. J. Hum. Genet., , 62, , 111–121.
Mariappan,S.V., Catasti,P., Silks,L.A., III, Bradbury,E.M. and Gupta,G. ( (1999) ) The high-resolution structure of the triplex formed by the GAA/TTC triplet repeat associated with Friedreich's ataxia. J. Mol. Biol., , 285, , 2035–2052.
Grabczyk,E. and Usdin,K. ( (2000) ) The GAA·TTC triplet repeat expanded in Friedreich's ataxia impedes transcription elongation by T7 RNA polymerase in a length and supercoil dependent manner. Nucleic Acids Res., , 28, , 2815–2822.
LeProust,E.M., Pearson,C.E., Sinden,R.R. and Gao,X. ( (2000) ) Unexpected formation of parallel duplex in GAA and TTC trinucleotide repeats of Friedreich's ataxia. J. Mol. Biol., , 302, , 1063–1080.
Sakamoto,N., Chastain,P.D., Parniewski,P., Ohshima,K., Pandolfo,M., Griffith,J.D. and Wells,R.D. ( (1999) ) Sticky DNA: self-association properties of long GAA·TTC repeats in R·R·Y triplex structures from Friedreich's ataxia. Mol. Cell, , 3, , 465–475.
Potaman,V.N., Oussatcheva,E.A., Lyubchenko,Y.L., Shlyakhtenko,L.S., Bidichandani,S.I., Ashizawa,T. and Sinden,R.R. ( (2004) ) Length-dependent structure formation in Friedreich ataxia (GAA)n·(TTC)n repeats at neutral pH. Nucleic Acids Res., , 32, , 1224–1231.
Ohshima,K., Montermini,L., Wells,R.D. and Pandolfo,M. ( (1998) ) Inhibitory effects of expanded GAA·TTC triplet repeats from intron I of the Friedreich ataxia gene on transcription and replication in vivo. J. Biol. Chem., , 273, , 14588–14595.
Ohshima,K., Kang,S., Larson,J.E. and Wells,R.D. ( (1996) ) Cloning, characterization, and properties of seven triplet repeat DNA sequences. J. Biol. Chem., , 271, , 16773–16783.
Potaman,V.N., Bissler,J.J., Hashem,V.I., Oussatcheva,E.A., Lu,L., Shlyakhtenko,L.S., Lyubchenko,Y.L., Matsuura,T., Ashizawa,T., Leffak,M. et al. ( (2003) ) Unpaired structures in SCA10 (ATTCT)n·(AGAAT)n repeats. J. Mol Biol., , 326, , 1095–1111.
Kang,S., Jaworski,A., Ohshima,K. and Wells,R.D. ( (1995) ) Expansion and deletion of CTG triplet repeats from human disease genes are determined by the direction of replication in E.coli. Nature Genet., , 10, , 213–218.
Darlow,J.M. and Leach,D.R.F. ( (1995) ) The effects of trinucleotide repeats found in human inherited disorders on palindrome inviability in Escherichia coli suggest hairpin folding preferences in vivo. Genetics, , 141, , 825–832.
Darlow,J.M. and Leach,D.F. ( (1998) ) Evidence for two preferred hairpin folding patterns in d(CGG)·d(CCG) repeat tracts in vivo. J. Mol. Biol., , 275, , 17–23.
Freudenreich,C.H., Stavenhagen,J.B. and Zakian,V.A. ( (1997) ) Stability of a CTG/CAG trinucleotide repeat in yeast is dependent on its orientation in the genome. Mol. Cell. Biol., , 17, , 2090–2098.
Cleary,J.D., Nichol,K., Wang,Y.H. and Pearson,C.E. ( (2002) ) Evidence of cis-acting factors in replication-mediated trinucleotide repeat instability in primate cells. Nature Genet., , 31, , 37–46.
Ireland,M.J., Reinke,S.S. and Livingston,D.M. ( (2000) ) The impact of lagging strand replication mutations on the stability of CAG repeat tracts in yeast. Genetics, , 155, , 1657–1665.
Schweitzer,J.K. and Livingston,D.M. ( (1999) ) The effect of DNA replication mutations on CAG tract stability in yeast. Genetics, , 152, , 953–963.
Trinh,T.Q. and Sinden,R.R. ( (1991) ) Preferential DNA secondary structure mutagenesis in the lagging strand of replication in E.coli. Nature, , 352, , 544–547.
Sinden,R.R., Zheng,G., Brankamp,R.G. and Allen,K.N. ( (1991) ) On the deletion of inverted repeated DNA in Escherichia coli: Effects of length, thermal stability, and cruciform formation in vivo. Genetics, , 129, , 991–1005.
Hashem,V.I., Rosche,W.A. and Sinden,R.R. ( (2004) ) Genetic recombination destabilizes (CTG)n (CAG)n repeats in E.coli. Mutat. Res., , 554, , 95–109.
Pelletier,R., Krasilnikova,M.M., Samadashwily,G.M., Lahue,R. and Mirkin,S.M. ( (2003) ) Replication and expansion of trinucleotide repeats in yeast. Mol. Cell. Biol., , 23, , 1349–1357.
Krasilnikova,M.M. and Mirkin,S.M. ( (2004) ) Replication stalling at Friedreich's ataxia (GAA)n repeats in vivo. Mol. Cell. Biol., , 24, , 2286–2295.
Samadashwily,G.M., Raca,G. and Mirkin,S.M. ( (1997) ) Trinucleotide repeats affect DNA replication in vivo. Nature Genet., , 17, , 298–304.
Callahan,J.L., Andrews,K.J., Zakian,V.A. and Freudenreich,C.H. ( (2003) ) Mutations in yeast replication proteins that increase CAG/CTG expansions also increase repeat fragility. Mol. Cell. Biol., , 23, , 7849–7860.
Freudenreich,C.H., Kantrow,S.M. and Zakian,V.A. ( (1998) ) Expansion and length-dependent fragility of CTG repeats in yeast. Science, , 279, , 853–856.
Freudenreich,C.H. and Lahiri,M. ( (2004) ) Structure-forming CAG/CTG repeat sequences are sensitive to breakage in the absence of Mrc1 checkpoint function and S-phase checkpoint signaling: implications for trinucleotide repeat expansion diseases. Cell Cycle, , 3, .
Lahiri,M., Gustafson,T.L., Majors,E.R. and Freudenreich,C.H. ( (2004) ) Expanded CAG repeats activate the DNA damage checkpoint pathway. Mol. Cell, , 15, , 287–293.
Liu,Y., Zhang,H., Veeraraghavan,J., Bambara,R.A. and Freudenreich,C.H. ( (2004) ) Saccharomyces cerevisiae flap endonuclease 1 uses flap equilibration to maintain triplet repeat stability. Mol. Cell. Biol., , 24, , 4049–4064.
Rolfsmeier,M.L., Dixon,M.J., Pessoa-Brandao,L., Pelletier,R., Miret,J.J. and Lahue,R.S. ( (2001) ) Cis-elements governing trinucleotide repeat instability in Saccharomyces cerevisiae. Genetics, , 157, , 1569–1579.
Dudas,A. and Chovanec,M. ( (2004) ) DNA double-strand break repair by homologous recombination. Mutat. Res., , 566, , 131–167.
van den Bosch,M., Lohman,P.H. and Pastink,A. ( (2002) ) DNA double-strand break repair by homologous recombination. Biol. Chem., , 383, , 873–892.
Symington,L.S. ( (2002) ) Role of RAD52 epistasis group genes in homologous recombination and double-strand break repair. Microbiol. Mol. Biol. Rev., , 66, , 630–70, table.
Lees-Miller,S.P. and Meek,K. ( (2003) ) Repair of DNA double strand breaks by non-homologous end joining. Biochimie, , 85, , 1161–1173.
Friedberg,E.C., Walker,G.C. and Siede,W. ( (1995) ) DNA Repair and Mutagenesis. ASM Press, Washington, DC, pp. 213–214, 288.
Hashem,V.I. and Sinden,R.R. ( (2002) ) Chemotherapeutically induced deletion of expanded triplet repeats. Mutat. Res., , 508, , 107–119.
Pineiro,E., Fernandez-Lopez,L., Gamez,J., Marcos,R., Surralles,J. and Velazquez,A. ( (2003) ) Mutagenic stress modulates the dynamics of CTG repeat instability associated with myotonic dystrophy type 1. Nucleic Acids Res., , 31, , 6733–6740.
Gorbunova,V., Seluanov,A., Dion,V., Sandor,Z., Meservy,J.L. and Wilson,J.H. ( (2003) ) Selectable system for monitoring the instability of CTG/CAG triplet repeats in mammalian cells. Mol. Cell. Biol., , 23, , 4485–4493.
Yang,Z., Lau,R., Marcadier,J.L., Chitayat,D. and Pearson,C.E. ( (2003) ) Replication inhibitors modulate instability of an expanded trinucleotide repeat at the myotonic dystrophy type 1 disease locus in human cells. Am. J. Hum. Genet., , 73, , 1092–1105.
Gomes-Pereira,M. and Monckton,D.G. ( (2004) ) Chemically induced increases and decreases in the rate of expansion of a CAG·CTG triplet repeat. Nucleic Acids Res., , 32, , 2865–2872.
Ashizawa,T., Monckton,D.G., Vaishnav,S., Patel,B.J., Voskova,A. and Caskey,C.T. ( (1996) ) Instability of the expanded (CTG)n repeats in the myotonin protein kinase gene in cultured lymphoblastoid cell lines from patients with myotonic dystrophy. Genomics, , 36, , 47–53.
Khajavi,M., Tari,A.M., Patel,N.B., Tsuji,K., Siwak,D.R., Meistrich,M.L., Terry,N.H. and Ashizawa,T. ( (2001) ) ‘Mitotic drive’ of expanded CTG repeats in myotonic dystrophy type 1 (DM1). Hum. Mol. Genet., , 10, , 855–863.
Monckton,D.G., Wong,L.J., Ashizawa,T. and Caskey,C.T. ( (1995) ) Somatic mosaicism, germline expansions, germline reversions and intergenerational reductions in myotonic dystrophy males: small pool PCR analyses. Hum. Mol. Genet., , 4, , 1–8.
Gomes-Pereira,M., Bidichandani,S.I. and Monckton,D.G. ( (2004) ) Analysis of unstable triplet repeats using small-pool polymerase chain reaction. Methods Mol. Biol., , 277, , 61–76.
Martorell,L., Gamez,J., Cayuela,M.L., Gould,F.K., McAbney,J.P., Ashizawa,T., Monckton,D.G. and Baiget,M. ( (2004) ) Germline mutational dynamics in myotonic dystrophy type 1 males: allele length and age effects. Neurology, , 62, , 269–274.
Rosche,W.A., Jaworski,A., Kang,S., Kramer,S.F., Larson,J.E., Geidroc,D.P., Wells,R.D. and Sinden,R.R. ( (1996) ) Single-stranded DNA-binding protein enhances the stability of CTG triplet repeats in Escherichia coli. J. Bacteriol., , 178, , 5042–5044.
McGlynn,P. and Lloyd,R.G. ( (2002) ) Recombinational repair and restart of damaged replication forks. Nature Rev. Mol. Cell Biol., , 3, , 859–870.
Miret,J.J., Pessoa-Brandao,L. and Lahue,R.S. ( (1998) ) Orientation-dependent and sequence-specific expansions of CTG/CAG trinucleotide repeats in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA, , 95, , 12438–12443.
Bhattacharyya,S. and Lahue,R.S. ( (2004) ) Saccharomyces cerevisiae Srs2 DNA helicase selectively blocks expansions of trinucleotide repeats. Mol. Cell. Biol., , 24, , 7324–7330.
Schweitzer,J.K. and Livingston,D.M. ( (1998) ) Expansions of CAG repeat tracts are frequent in a yeast mutant defective in Okazaki fragment maturation. Hum. Mol. Genet, , 7, , 69–74.
Lee,S. and Park,M.S. ( (2002) ) Human FEN-1 can process the 5'-flap DNA of CTG/CAG triplet repeat derived from human genetic diseases by length and sequence dependent manner. Exp. Mol. Med., , 34, , 313–317.
Panigrahi,G.B., Cleary,J.D. and Pearson,C.E. ( (2002) ) In vitro (CTG)·(CAG) expansions and deletions by human cell extracts. J. Biol. Chem., , 277, , 13926–13934.
Marcadier,J.L. and Pearson,C.E. ( (2003) ) Fidelity of primate cell repair of a double-strand break within a (CTG)·(CAG) tract. Effect of slipped DNA structures. J. Biol. Chem., , 278, , 33848–33856.
Little,J.B., Gorgojo,L. and Vetrovs,H. ( (1990) ) Delayed appearance of lethal and specific gene mutations in irradiated mammalian cells. Int. J. Radiat. Oncol. Biol. Phys., , 19, , 1425–1429.
Little,J.B. ( (1998) ) Radiation-induced genomic instability. Int. J. Radiat. Biol., , 74, , 663–671.
Wright,E.G. ( (1998) ) Radiation-induced genomic instability in haemopoietic cells. Int. J. Radiat. Biol., , 74, , 681–687.
Dubrova,Y.E., Plumb,M., Gutierrez,B., Boulton,E. and Jeffreys,A.J. ( (2000) ) Transgenerational mutation by radiation. Nature, , 405, , 37.
Chastain,P.D., Eichler,E.E., Kang,S., Nelson,D.L., Levene,S.D. and Sinden,R.R. ( (1995) ) Anomalous rapid electrophoretic mobility of DNA containing triplet repeats associated with human disease genes. Biochemistry, , 34, , 16125–16131.
Chastain,P.D. and Sinden,R.R. ( (1998) ) CTG repeats associated with human genetic disease are inherently flexible. J. Mol. Biol., , 275, , 405–411.
Bacolla,A., Gellibolian,R., Shimizu,M., Amirhaeri,S., Kang,S., Ohshima,K., Larson,J.E., Harvey,S.C., Stollar,B.D. and Wells,R.D. ( (1997) ) Flexible DNA - genetically unstable CTG·CAG and CGG·CCG from human hereditary neuromuscular disease genes. J. Biol. Chem., , 272, , 16783–16792.
Pearson,C.E. and Sinden,R.R. ( (1996) ) Alternative structures in duplex DNA formed within the trinucleotide repeats of the myotonic dystrophy and fragile X loci. Biochemistry, , 35, , 5041–5053.
Pearson,C.E. and Sinden,R.R. ( (1998) ) Trinucleotide repeat DNA structures: dynamic mutations from dynamic DNA. Curr. Opin. Struct. Biol., , 8, , 321–330.
Pearson,C.E., Wang,Y.H., Griffith,J.D. and Sinden,R.R. ( (1998) ) Structural analysis of slipped-strand DNA (S-DNA) formed in (CTG)n·(CAG)n repeats from the myotonic dystrophy locus. Nucleic Acids Res., , 26, , 816–823.
Glazer,P.M., Wang,G., Havre,P.A. and Gunther,E.J. ( (1996) ) Targeted mutagenesis mediated by the triple helix formation. Methods Mol. Biol., , 57, , 109–118.
Chan,P.P. and Glazer,P.M. ( (1997) ) Triplex DNA: fundamentals, advances, and potential applications for gene therapy. J. Mol. Med., , 75, , 267–282.
Vasquez,K.M. and Wilson,J.H. ( (2000) ) Triplex-directed site-specific genome modification. Methods Mol. Biol., , 133, , 183–200.
Boffa,L.C., Morris,P.L., Carpaneto,E.M., Louissaint,M. and Allfrey,V.G. ( (1996) ) Invasion of the CAG triplet repeats by a complementary peptide nucleic acid inhibits transcription of the androgen receptor and TATA-binding protein genes and correlates with refolding of an active nucleosome containing a unique AR gene sequence. J. Biol. Chem., , 271, , 13228–13233.
Kennedy,L., Evans,E., Chen,C.M., Craven,L., Detloff,P.J., Ennis,M. and Shelbourne,P.F. ( (2003) ) Dramatic tissue-specific mutation length increases are an early molecular event in Huntington disease pathogenesis. Hum. Mol. Genet., , 12, , 3359–3367.
Kovtun,I.V. and McMurray,C.T. ( (2001) ) Trinucleotide expansion in haploid germ cells by gap repair. Nature Genet., , 27, , 407–411.(Vera I. Hashem, Malgorzata J. Pytlos, El)
* To whom correspondence should be addressed. Tel: +1 713 677 7664; Fax: +1 713 677 7689; Email: Rsinden@ibt.tamushsc.edu
Present addresses: Vera I. Hashem and Merhdad Khajavi, Department of Human and Molecular Genetics, Baylor College of Medicine, Houston, TX 77030, USA
Kuniko Tsuji, Department of Neurology, Osaka Medical College, Takatsuki 568-8686, Japan
ABSTRACT
Myotonic dystrophy type 1 (DM1) is caused by the expansion of a (CTG)·(CAG) repeat in the DMPK gene on chromosome 19q13.3. At least 17 neurological diseases have similar genetic mutations, the expansion of DNA repeats. In most of these disorders, the disease severity is related to the length of the repeat expansion, and in DM1 the expanded repeat undergoes further elongation in somatic and germline tissues. At present, in this class of diseases, no therapeutic approach exists to prevent or slow the repeat expansion and thereby reduce disease severity or delay disease onset. We present initial results testing the hypothesis that repeat deletion may be mediated by various chemotherapeutic agents. Three lymphoblast cell lines derived from two DM1 patients treated with either ethylmethanesulfonate (EMS), mitomycin C, mitoxantrone or doxorubicin, at therapeutic concentrations, accumulated deletions following treatment. Treatment with EMS frequently prevented the repeat expansion observed during growth in culture. A significant reduction of CTG repeat length by 100–350 (CTG)·(CAG) repeats often occurred in the cell population following treatment with these drugs. Potential mechanisms of drug-induced deletion are presented.
INTRODUCTION
The expansion of trinucleotide repeats and other short tandem DNA repeats is associated with an increasing number of inherited neurological diseases. These include myotonic dystrophy type 1 (DM1), DM2, Huntington's disease (HD), HD-Like 2, dentatorubral pallidoluysian atrophy, fragile X syndrome, FRAXE mental retardation, Kennedy's disease, oculopharyngeal muscular dystrophy, Friedreich ataxia and several spinocerebellar ataxias (SCAs) (SCA1, 2, 3, 6, 7, 8, 10, 12 and 17) (1–5).
Several common features are shared by most tandem repeat diseases. First, patients with most of these diseases exhibit an earlier age of onset and increased severity in successive generations, a phenomenon termed anticipation. This phenomenon is generally correlated with increasing repeat length (6). Second, expanded repeats in most of these diseases show both somatic and germline instability usually with a bias toward expansion. In transgenic animal models of these diseases, the longer repeat in the transgene led to increased severity and earlier onset (7–11). Therefore, the repeat length is critical in the pathogenic mechanism and is one of the most important determinants of disease severity.
Repeat expansion leads to the respective disease phenotypes through diverse mechanisms. For example, expanded CAG repeats encoding polyglutamine tracts lead to a gain of function of a protein as with HD, Kennedy's disease and several of the SCAs. A large (CGG)·(CCG) repeat expansion in the 5'-untranslated region (UTR) of the FMR1 and FMR2 genes and an expansion of an intronic (GAA)·(TTC) repeat in the FRDA gene result in a loss of function of the protein products in fragile X syndrome, FRAXE mental retardation and Friedreich ataxia, respectively (1). In SCA12, an expansion of CAG repeat in the 5'-UTR may cause enhanced transcription of the PPP2R2B gene (12). The pathogenic mechanisms of other repeat expansion diseases remain less well defined.
No treatment currently exists for DNA repeat expansion-associated neurological diseases. Because the molecular events downstream of repeat expansion mutations in these diseases are complex and variable, therapeutic strategies to intervene in such mechanisms would be difficult to formulate and must be tailor-made in each disease. Since the underlying etiology is the repeat expansion, treatment with chemicals or drugs that would reduce the repeat size would target the very source of the problem and may be generally applicable to delay onset or reduce the severity of diseases associated with repeat expansion. This may be especially important in diseases where somatic expansion during life may be associated with disease onset and severity.
DM1 is a multisystem disorder, characterized by muscle weakness, wasting and myotonia. DM1 results from an expansion of a CTG repeat in the 3'-UTR of the myotonic dystrophy protein kinase gene (DMPK). The expanded CTG repeat is unstable with a bias toward further expansion in germline and somatic tissues, giving rise to anticipation, and repeat size increases in tissues during life (13,14). The molecular mechanism downstream of the mutation in DM1 is complex and is associated with partial loss of DMPK, decreased transcription of a neighboring gene, Six5, in addition to binding of CUG-binding proteins by the expanded CUG mRNA (15). Thus, DM1 may be considered as a representative disease model to test the feasibility of chemical or drug treatments that decrease the repeat size.
The actual mechanism of tandem repeat expansion remains unclear. Several lines of evidence implicate various processes, including DNA replication, repair and recombination (2,16–18). One common feature of the repeats that may contribute to their instability is their ability to form various DNA secondary structures (2). CTG, CAG, CCG and CGG repeats form stable hairpins (19–21); (CTG)·(CAG) and (CGG)·(CCG) repeats form slipped-strand DNA structures caused by out-of-register misalignment within the repeat tract (22); (GGC)·(GGC) repeats form quadruplex DNA structures (23,24); and (GAA)·(TTC) repeats can form triplex structures and parallel stranded DNA (25–32). (ATTCT)·(AGAAT) repeats can form stable unpaired regions in DNA (33). A reasonable hypothesis remains that the formation of an alternative DNA secondary structure elicits repeat expansion during replication, repair, or recombination (2) or promotes aberrant DNA replication (33).
While mechanisms of repeat expansion are uncertain, mechanisms leading to (CTG)·(CAG) repeat deletion have been generally well accepted. A large body of evidence suggests that the formation of a single-stranded region containing the (CTG)n strand leads to the formation of hairpins that fail to be copied during DNA replication leading to large repeat deletions (17,34–38). Deletions have been thought to result from replication slippage across CTG hairpins formed in the lagging template strand when single-stranded during replication (2,17,34,39,40). Hairpins stabilize misaligned intermediates preferentially in the lagging strand and promote misalignment (41,42). Recent results, however, show that functional RecA and RecB are required for a high rate of repeat deletion, and it is difficult to incorporate these proteins into a simple replication slippage model for repeat deletion (43). An alternative model suggests that repeat deletion results from errors occurring during replication restart following the collapse of the replication fork during synthesis of the repeats via a pathway that involves a double-strand break (43). Unrepaired DNA damage will also stop the replication fork, causing fork collapse. In bacteria and yeast, replication through repeats can cause replication fork blockage (44–46). In yeast, repeats are associated with chromosome ‘fragility’, presumably a double-strand break that initiates recombination (37,45,47–52). Thus, DNA damage can block replication fork progression and lead to double-strand breaks in bacterial, yeast and human cells, where repair occurs by a variety of homologous and non-homologous based mechanisms (53–56). Moreover, many DNA damage repair processes involve the generation of a single-stranded gap in the DNA as a damaged base is removed (57). Thus, we hypothesize that the introduction of DNA damage to cells harboring (CTG)·(CAG) triplet repeats could lead to an increase in the rate of triplet repeat deletion both by creating single-strand regions and by causing replication fork collapse. Our results in a model bacterial system confirmed this hypothesis, showing that the DNA-damaging agents ethylmethanesulfonate (EMS), ethylnitrosourea, and ultraviolet (UV) light, and chemotherapeutic agents mitomycin C, cisplatin and X-rays increased the rate of deletion of (CTG)·(CAG) repeats (58). Moreover, oxidative damage to DNA also increased the rate of deletion of triplet repeats (58). These results provide the molecular basis to suggest that a chemotherapeutic approach to triplet repeat deletion may offer one rationale to slow, stop or reverse the progression of these diseases. In fact, as discussed below, recent experiments in several mammalian cell systems support our basic hypothesis (59–62).
Here, we treat lymphoblast cell lines derived from DM1 patients with DNA-damaging agents, including EMS, a monofunctional alkylating agent, and several chemotherapeutic agents, including mitomycin C, which introduces interstrand crosslinks; mitoxantrone, an inhibitor of topoisomerase II, which intercalates into DNA causing crosslinks and strand breaks; and doxorubicin, which intercalates into DNA and inhibits DNA synthesis. Mitomycin C, mitoxantrone and doxorubicin are currently used in the treatment of cancer and mitoxantrone is also used to treat multiple sclerosis. Our results indicate that drug treatment can prevent expansion commonly observed in lymphoblastoid DM1 cells. In many cases, a significant reduction of (CTG)·(CAG) repeat length occurred in the entire cell population following drug treatment.
MATERIALS AND METHODS
Cell growth
Clonal lymphoblast cell lines derived from DM1 patients have been described previously (63,64). Cell lines 313-1-15, 313-1-20 and 428 initially contained about 770, 740 and 440 (CTG)·(CAG) repeats, respectively (63). DM1 lymphoblast cells were grown between 2 x 105 and 2 x 106 cells/ml at 37°C with 5% CO2 in RPMI (Hi-Clone) media supplemented with 10% calf serum (Hi-Clone) and 1 μM insulin. Cells were grown for a defined number of generations for comparison. Generations is calculated as /log2, where DF is a dilution factor (calculated from the total number of times cells were diluted), NG is the total number of cells at the time of sampling, and N0 is the number of cells at the beginning of the experiment. Previously, growth has been reported as ‘passages’ (63). A passage represents 2–4 generations, when cells are diluted 1:4 or 1:8 for subsequent growth. The calculation of generations in our analysis does not take into account cell death in the population (only population doubling), and thus may represent a minimum or underestimate of the actual number of cell divisions at the time of harvest.
Treatment with drugs
DM1 lymphoblast cell lines were treated with drugs in growth media at 1 x 106 cells/ml for various times. Treatment with 500 or 700 μg/ml EMS or 0.1 or 0.2 μg/ml of mitomycin C was for 14–16 h, treatment with 50 or 100 nM mitoxantrone was for 1 h, while treatment with 0.1–5 μg/ml doxorubicin was for 30 min. Drug exposures were adjusted to reduce survival to 10–30%, except for treatment with mitoxantrone, which reduced survival to 30–50%. Following treatment, the cells were pelleted, washed once with media and resuspended in media at 1 x 106 cells/ml and incubated. Growth was generally continued for up to 10–30 generations. At selected passages aliquots of 1 x 107 cells were harvested by centrifugation for 5 min and frozen at –20°C until the genomic DNA was isolated for small pool PCR. In certain experiments, a second dose of the drug was administered in the same manner as above after recovery of the culture.
DNA purification and PCR analysis
For allele length analyses, genomic DNA was purified from 5–10 x 106 lymphoblast cells using a DNeasy Tissue Kit (Qiagen) followed by one or two ethanol precipitations. For small pool PCR, reactions contained 50–2000 pg of HindIII-digested DNA corresponding to 1–50 diploid genome equivalents per reaction. Two different conditions were used with identical results. First, PCR was performed according to the procedure described by Monckton et al. (13). In addition, PCR was performed as follows: 0.5 U Platinum Taq DNA Polymerase High Fidelity (Invitrogen), 0.2 μM primer DM-DR (dCAGGCCTGCAGTTTGCCCATC) and 0.2 μM primer DM-H (dTCTCCGCCCAGCTCCAGTC), 2 mM MgSO4, 0.2 mM dNTPs, in a total volume of 10 μl. PCR was performed in a Perkin–Elmer 9700 PCR machine with the following conditions: the DNA was denatured at 94°C for 6 min and then subjected to 30–35 cycles of 94°C for 45 s, 67°C for 45 s, 70°C for 2 min 40 s, with a final extension step at 70°C for 35 min.
Southern hybridization
PCR products were separated by electrophoresis through 1.8% agarose gels in 0.5x TBE buffer (1x TBE: 90 mM Tris, 90 mM Borate, 2.5 mM EDTA, pH 8.3), transferred to a Hybond N membrane (Amersham), the DNA fixed by UV irradiation and then the membrane was hybridized with an (AGC)10 oligonucleotide end-labeled with ATP. Hybridization was performed for 16 h at 42°C in 1x SSC buffer. Membranes were then washed twice for 30 min at 49°C in 0.5x SSC, 0.1% SDS, rinsed in H2O and then exposed for 1–2 days to a Molecular Dynamics PhosphorImager plate. Alleles were sized by comparison with molecular weight markers. Lambda/HindIII/pUC19/TaqI/Sau3A1 molecular weight markers (Invitrogen) were labeled with 32P as follows: 1 μg of DNA was added to 10 U Klenow (New England Biolabs), 1 mM dGTP, 1 mM dCTP, 1 mM dTTP and 50 μCi (3000 Ci/mM) dATP, in 20 μl Klenow buffer and the mixture was incubated 15 min at 20°C and then 15 min at 65°C. The number of repeats in individual alleles was calculated from the length of the PCR product and the DNA sequence flanking the (CTG)·(CAG) repeats.
Statistical analysis
For statistical analysis, chi-square tests were calculated for each set of experiments. The numbers of original-length alleles and expansions and deletions were compared between control, non-treated cells and drug-treated cells. Statistical results for a single regiment of drug treatment for both cell lines were then analyzed using a modified Bonferroni comparison to control for type 1 error. The significance following this comparison is reported in the results section.
RESULTS
Repeat instability in lymphoblast cell lines 428 and 313-1-15
Growth of lymphoblast cell lines through many generation is accompanied by expansion of the expanded (CTG)·(CAG) repeat, at a variable magnitude and frequency (63,64). (CTG)·(CAG) repeat lengths were analyzed by small pool PCR using a well-established procedure that can easily detect more than 1000 (CTG)·(CAG) repeats (13,14,63–67). We have analyzed the effect of treatment with a common alkylating agent and three chemotherapeutic drugs on repeat instability. Two of these agents, EMS and mitomycin C, increased the rates of (CTG)·(CAG) deletion in a bacterial model system (17). Cell line 313-1-15 contains about 770 (CTG)·(CAG) repeats that remain relatively stable but do show a slight and gradual increase in the repeat size during growth (data not shown). The initial culture of cell line 428 showed a distribution of allele sizes of about 440 ± 100 repeats (Figure 1A). After growth for 27 generations in one experiment (experiment A), two allele sizes dominated the cell population: 440 and 770 repeats (Figure 1A, experiment A). In a subsequent experiment, which was started from a sub culture of experiment A after five generations growth, the 770 repeat allele dominated the population following an additional 17 generations of growth (experiment B). Figure 1B shows the percentage of alleles plotted as a function of the number of (CTG)·(CAG) repeats for many repetitions of the PCR reactions as shown in Figure 1A. To confirm the rapid expansion in cell line 428, cells from the initial distribution population shown in Figure 1A were sub-cloned by growth of an initial 100–300 cells for 18–20 generations. All populations analyzed showed predominantly the larger 770 repeat allele. Six of nine sub-clones analyzed showed only the 770 repeat allele, while three showed 94–97% the 770 repeat allele (data not shown). Clearly, during these drug experiments cell line 428 was prone to repeat expansion in control, non-treated cultures. Control, of non-treated cultures of cell line 428 was repeated more than 10 times. At later passages, after the repeat had expanded to >700 repeats, spontaneous reductions in allele sizes in the population were never observed without drug treatment. The behavior of this cell line mimics events occurring in certain DM1 tissues, where repeat expansion continues throughout life (13). The rapid change in repeat lengths is presumably due to cells with larger (CTG)·(CAG) repeats having a growth advantage over those with smaller expanded alleles (64). Given the potential for rapid change in cell line 428, it seemed that it may be sensitive to the effects of DNA-damaging agents. On the other hand, cell line 313-1-15 was used for analysis because it is relatively stable and the accumulation of small deletions may be more easily observed in this line. Following growth of this cell line in 17 independent experiments, deletions comprised <2% of the allele population in the absence of drug treatment, and a large deletions of the repeat allele in the entire population were never observed.
Figure 1. Expansion of (CTG)·(CAG) repeats in cell line 428. (A) Representative small pool PCR reactions with 300 pg DNA are shown for cell line 428 for an initial early passage culture and following growth. In experiment A, the cells were grown for 27 generations. Experiment B was started from sub culture of cells from experiment A after five generations growth, and growth was continued for an additional 17 generations. Size markers are indicated relative to the number of (CTG)·(CAG) repeats in the PCR products. Under the PCR conditions used, 300 pg corresponds to 6–12 genome equivalents per reaction. (B) Allele spectra for 428 cultures at the time of initial growth (panel a) n = 151; experiment A after 27 generations growth (panel b), n = 320; and experiment B after 17 generations growth (panel c), n = 515. n = the number of alleles analyzed.
Treatment with the alkylating agent EMS results in the accumulation of the 440 repeat allele in cell line 428
Lymphoblast cell line 428 was treated with 500 and 700 μg/ml EMS and grown for 10–20 generations. Small pool PCR analysis of alleles shows that the treatment resulted in the accumulation of much smaller allele sizes than present in the non-treated cells grown relatively for the same number of generations (Figure 2). Treatment with 500 μg/ml EMS typically showed the accumulation of two or three repeat sizes (440, 550 and 770 repeats) (Figure 3). When cells were treated with a second dose of EMS, the accumulation of the smaller sized alleles was more pronounced (Figure 3A, panel b and Figure 3B, panel c). Treatment with 700 μg/ml EMS generally resulted in the accumulation of the 440 repeat allele size, a reduction in size of 330 repeats relative to the expansion observed in non-treated cells. In the experiment shown in Figure 3A, the difference in the allele distributions for both drug-treated samples relative to the non-treated distribution was significant (P 0.001) using a modified Bonferroni comparison. In the experiment shown in Figure 3B, while the single treatment of 500 μg/ml EMS resulted in the accumulation of many shorter sized deletions (panel b), the distribution was not significantly different from the control, non-treated distribution comparing the numbers of 440 repeat alleles and other-size alleles (P = 0.072). In this experiment, distributions following the second dose of 500 μg/ml EMS treatment and the single dose of 700 μg/ml EMS treatment (panels c and d) were significantly different (P 0.001). Experiments involving cell growth following treatment with EMS were repeated six times for this cell line with essentially identical results. Occasionally (in <10% of drug-treated samples) an EMS-treated sample with predominantly the expanded (770 repeat) allele was observed. It is important to reiterate that, from independent analyses of the growth of many cell lines and many individual growth analyses, large deletions, especially reductions in the repeat tract in the entire population, were not observed in this, or previous studies (63,64). Thus, the large repeat reductions observed in the population are highly significant and drug dependent.
Figure 2. PCR analysis of the accumulation of deletions in 428 cells following treatment with EMS or mitomycin C. Small pool PCR reactions with genomic DNA are shown for control (non-treated) cells and for cells treated with 0.2 μg/ml mitomycin C or 700 μg/ml EMS. DNA was purified 12–17 generations after drug treatment. PCR reactions contained either 50 or 300 pg DNA, as indicated. Molecular weight markers are indicated as the number of (CTG)·(CAG) repeats.
Figure 3. Accumulation of deletions in 428 cells following EMS treatment. (A) Experiment started with 428 cells using the population of alleles shown in Figure 1B, panel a, at the time of treatment. Panel a, allele distribution following 27 generations growth in control, non-treated cells. Panel b, allele distribution following a second treatment with 500 μg/ml EMS and 16 generations growth (n = 362). Panel c, allele distribution following a single treatment with 700 μg/ml EMS and 18 generations growth (n = 323). In both drug-treated samples, the allele distribution was significantly different from the control, non-treated sample using a chi-square test (P 0.001). (B) Experiment started with 428 cells (as described in text). Panel a, allele distribution following an additional 17 generations growth in control, non-treated cells. Panel b, allele distribution following a single treatment with 500 μg/ml EMS, 18 generations growth (n = 196). Panel c, allele distribution following two treatments with 500 μg/ml EMS, 12 generations growth (n = 269). Panel d, allele distribution following a single treatment with 700 μg/ml EMS, 16 generations growth (n = 603). The allele distribution in panel b was not significantly different from the control (P = 0.071). In panels c and d, the allele distributions were significantly different from the control, non-treated sample using a chi-square test (P 0.001).
Treatment with mitomycin C results in the accumulation of (CTG)·(CAG) deletions
Lymphoblast cell line 428 and 313-1-15 were treated with 0.1 or 0.2 μg/ml mitomycin C and grown following drug exposure. A dramatic effect was observed in both cell lines following treatment with mitomycin C. In the experiment shown in Figures 2 and 4A for cell line 428, a broad range of repeat sizes accumulated following drug treatment. Little of the original 440 repeat size allele was observed in cells treated once or twice with 0.1 μg/ml mitomycin C (Figure 4A, panels b and c). Following the second treatment (Figure 4A, panel c), a broad distribution of allele sizes was observed (from 300 to 675 repeats), corresponding to reductions of 100–470 repeats relative to non-treated cells. Compared with the control non-treated distribution, the allele spectra for cells treated twice with mitomycin C were significantly different (Figure 4A, panel b, P 0.001; panel c, P = 0.002). The allele distribution for cells treated once with 0.2 μg/ml mitomycin C was not significantly different from control considering two populations of alleles (the initial 440 repeat size and all other expansions or deletions). Considering three populations of alleles (440 repeats, 770 repeats, and all other expansions and deletions), the allele spectra were significantly different (P 0.001). This data is representative of five independent experiments with cell line 428.
Figure 4. Accumulation of deletions following treatment with mitomycin C. (A) Mitomycin C treatment of cell line 428. Panel a, allele distribution following 17 generations growth in control, non-treated cells (n = 515). Panel b, allele distribution following treatment with 0.1 μg/ml mitomycin C and 18 generations growth (n = 192). Panel c, allele distribution following a second treatment with 0.1 μg/ml mitomycin C, 16 generations growth (n = 52). Panel d, allele distribution following a single treatment with 0.2 μg/ml mitomycin C, 12 generations growth (n = 231). In panels b and c, the allele distributions were significantly different from the control, non-treated sample using a chi-square test (P 0.001 and P = 0.002, respectively). Results in panel d were not statistically different considering the fraction of starting length (440 repeat) alleles, although the overall distribution is significantly different. (B) Mitomycin C treatment of cell line 313-1-15. Panel a, typical allele distribution for 313-1-15 showing predominantly the full length expanded allele of about 770 repeats after 11 generations growth in control, non-treated cells (n = 503). Panel b, allele distribution following a single treatment with 0.1 μg/ml mitomycin C, 9 generations growth (n = 460). Panel c, allele distribution following two treatments with 0.1 μg/ml mitomycin C, 4–5 generations growth (n = 559). Panel d, allele distribution following a single treatment with 0.2 μg/ml mitomycin C, 4–5 generations growth (n = 401). The allele distribution in panel b was not significantly different from the control (P = 0.096). In panels c and d, the allele distributions were significantly different from the control, non-treated sample using a chi-square test (P 0.001 and P = 0.052).
Similar results were observed for cell line 313-1-15 containing the relatively stable 770 (CTG)·(CAG) repeat allele (Figure 4B). During growth for 5–10 generations following drug treatment, an increase in the frequency of deletions was observed for some samples in this cell line, and alleles with 100 repeats, or fewer, were observed. Following a single exposure to 0.1 μg/ml mitomycin C (Figure 4B, panel b), the frequency of deletions was not significantly different from that in control cells. However, following a second treatment of 0.1 μg/ml mitomycin C (Figure 4B, panel c), the allele spectra were significantly different, after only 4–5 generations growth, with a P-value of <0.001. Following a single treatment with 0.2 μg/ml mitomycin C (Figure 4B, panel d), the difference in allele spectra was not significantly different (P = 0.052). In a separate experiment, a deletion from 770 to 675 repeats was observed in the entire population of 313-1-15 cells following treatment with 0.2 μg/ml mitomycin C (Figure 5). In this experiment, the allele spectra following 17 and 19 generations growth showed a reduction to 675 repeats (Figure 5, panels b and c). Both allele distributions were significantly different from non-treated, control cells (P 0.001). Experiments with cell line 313-1-15 and mitomycin C treatment were repeated 12 times. In another experiment using cell line 313-1-20, a large deletion of about 350 repeats was observed in 10–15% of the culture following treatment with 0.2 μg/ml mitomycin C (data not shown). Experiments with cell line 313-1-20 and mitomycin C treatment were repeated four times.
Figure 5. Accumulation of a 100 repeat deletion during growth in a population of 313-1-15 cells following treatment with mitomycin C. In this experiment, 313-1-15 cells were treated with mitomycin C at 0.2 μg/ml and growth was continued for 17–19 generations. Panel a, allele distribution for 313-1-15 showing predominantly the full-length expanded allele of 770 repeats after 19 generations growth (n = 229). Panel b, allele distribution following 17 generations growth (n = 135). Panel c, allele distribution following 19 generations growth (n = 178). The differences in allele distribution in panels b and c were significantly different from that in the control sample (P 0.001).
Mitoxantrone induces deletions in DM1 lymphoblast cells
A culture of lymphoblast cell line 428, in which the repeat had expanded to 740 (CTG)·(CAG) repeats (>98% of alleles), was treated with 50 and 100 nM mitoxantrone. The allele spectra following treatment with 50 nM mitoxantrone were not significantly different from that for control, non-treated cells after 14 generations growth (Figure 6, panel b). However, after 14 generations a significant shift in the population to about 620 repeats was observed with the sample treated with 100 nM mitoxantrone (P 0.001) (Figure 6, panel c). Eight independent cultures of 428 and 313-1-15 cells have been treated with mitoxantrone. The data shown in Figure 6 represent the most dramatic response.
Figure 6. Accumulation of a 120 repeat deletion during growth in a population of 428 cells following treatment with mitoxantrone. Cell line 428 was treated with mitoxantrone at 50 or 100 nM and growth was continued for 12–14 generations. Panel a, allele distribution in control, non-treated 428 cells. A subculture of cells in which a 770 repeat allele size dominated the population was used and grown for 12 generations (n = 232). Panel b, allele distribution after treatment with 50 nM and subsequent growth for 14 generations (n = 146). The allele distribution was not significantly different from that of control cells. Panel c, allele distribution following treatment with 100 nM and subsequent growth for 14 generations (n = 301). This difference in this allele distribution was statistically significant compared with the distribution in the control, non-treated sample (P 0.001).
Doxorubicin induces deletions in DM1 cells
DM1 lymphoblast cells were treated with several doses of doxorubicin and growth was continued for 10 generations. Figure 7 (top) shows representative data from one of the four independent experiments in cell line 313-1-15. The rate of population doubling was reduced in a dose-dependent fashion following exposure to doxorubicin. Allele distributions, after 10–16 generations growth, showed the accumulation of deletion mutations in 10–20% of the allele population in a dose-dependent fashion that paralleled the initial decrease in survival and growth rate (Figure 7, bottom). In this experiment, the expanded allele measured 840 repeats and most deleted alleles were between 350 and 650 repeats, or deletions of 200–500 repeats. The allele distributions following treatment with 1, 2 and 5 μg/ml doxorubicin were significantly different from that in the control distribution (P 0.001).
Figure 7. Accumulation of deletions following treatment with doxorubicin. Top, growth of 313-1-15 cells following treatment with doxorubicin. Cells were treated with 0.1, 0.5, 1.0, 2.0 or 5.0 μg/ml doxorubicin for 30 min and then growth was monitored, and samples harvested for 10–16 generations. Bottom, allele distributions. Panel a, control, non-treated cells, 10 generations growth (n = 170). Panel b, 0.1 μg/ml doxorubicin, 10 generations growth (n = 401). Panel c, 0.5 μg/ml doxorubicin, 11 generations growth (n = 314). Panel d, 1.0 μg/ml doxorubicin, 13 generations growth (n = 233, P = 0.031). Panel e, 2.0 μg/ml doxorubicin, 16 generations growth (n = 638, P 0.001). Panel f, 5.0 μg/ml doxorubicin, 13 generations growth (n = 485, P 0.001).
DISCUSSION
Our results demonstrate that lymphoblast cells, derived from patients with DM1, which contain expanded (CTG)·(CAG) repeats in the DMPK gene, accumulate smaller (CTG)·(CAG) repeat alleles following treatment with EMS, mitomycin C, mitoxantrone or doxorubicin. Reductions in the length of the repeat from 770 to 100 or fewer (CTG)·(CAG) repeats were frequently observed following drug treatment. In many cases, the length of the expanded repeat was reduced from 770 to 650–440 (CTG)·(CAG) repeats in the entire population of cells. This is a significant result given the propensity of the DM1 (CTG)·(CAG) repeats to expand rather than contract in patients and in culture, as shown here and previously (13,14,64).
Multiple events may be contributing to the reductions in repeat tract length: drug-induced deletion and clonal selection. Events associated with DNA repair or drug-induced replication fork collapse would be expected to result in deletions, and potential mechanisms are discussed below. The mixture of shorter alleles observed following treatment with mitomycin C or doxorubicin likely arise from multiple independent deletions within the population. Clonal selection after drug treatment may also be occurring within the population. As occurs in cancer chemotherapy, drugs may be selectively killing cells in the population with larger numbers of repeats, which have a more rapid rate of cell division than cells with the shorter alleles (64). This would then select for the growth of cells with shorter alleles. Some evidence for clonal selection has been observed following exposure of mouse fibroblasts containing expanded CTG repeats to various chemicals (62). However, alterations in repeat length in mouse and human cells following treatment with various chemicals and replication inhibitors can occur independent of clonal expansion (61,62). In our experiments, clonal expansion is possible when survival was much reduced. However, in many experiments, smaller sized alleles dominated the population prior to the time one would expect population outgrowth of a shorter sized allele from a single clonal cell. From a single cell, 24–25 population doublings are required to generate 107 cells. In the mitoxantrone experiment (Figure 6), in which the starting culture had >98% alleles with 740 repeats, the shorter 620 repeat allele would only generate 8.2 x 103 cells after 14 generations growth (0.1% of the cell population considering analysis of 107 cells) and by this time a complete shift in the population had occurred. The >300 repeat reduction observed in the entire population in cell line 428 in earlier passages when a mixture of allele sizes was present may represent clonal selection. It is important to mention, however, that due to the strong drive toward expansion in the 428 cells, the larger alleles will naturally dominate the population. Mixing experiments using cell lines with different allele sizes have shown that the cell line with the longer allele will dominate the population (64). We also showed that when the ‘initial culture’ of 428 shown in Figure 1 was sub-cloned, all populations analyzed showed predominantly the larger 770 repeat allele by 18–20 generations. Six of nine sub-clones showed only the 770 repeat allele, while three showed 94–97% the 770 repeat allele (data not shown). The drive in the cell culture is toward accumulation of larger alleles. Thus, the accumulation of shorter alleles in the population is highly unusual in the growth of DM1 cell cultures. Therefore, it is unlikely that the domination of the population by the 440 repeat allele following drug treatment is not related to treatment. The important caveat in these experiments is that certain drug treatments can, whether through frequent independent deletions and/or clonal selection, promote large deletions in the population. The outcome is significant and demonstrates a proof of principle in testing our hypothesis.
The data do not allow us, at present, to establish the mechanism(s) responsible for repeat deletion. Several phenomena may be contributing to the observed repeat deletion, and work is in progress to further understand the molecular basis for the drug-induced deletion. The working hypothesis is that DNA damage and/or replication fork collapse at or near the repeats elicits deletion. Deletions may arise through several different pathways. First, results using bacteria suggest that spontaneous deletions occur during a RecA and RecBC pathway of replication restart, in which the introduction of a double-strand break is associated with a high rate of repeat deletion (17,43). As shown in Figure 8A, replication fork collapse resulting from a DNA damage in the leading strand could lead to deletions. Evidence in yeast strongly supports the frequent introduction of double-strand breaks during DNA replication through (CTG)·(CAG) repeats (37,45,47–52). The addition of DNA damage would be expected to increase the frequency of replication fork collapse and the occurrence of double-strand breaks. Second, the formation of CTG hairpins within repair gaps or in single-strand DNA associated with replication restart may also promote repeat deletion, as is believed to occur in bacterial model systems (2,43,58) (Figure 8B). In bacterial cells, the generation of single-stranded DNA greatly increases the probability of deletion of CTG repeats (17,58,68). Single-stranded DNA is generated in different steps of many repair pathways or during replication restart (57,69). Thus, it would be expected that repair of DNA damage by an excision repair pathway could increase the chance of deletion at this DNA repeat hotspot for mutation by allowing the formation of a (CTG)n or (CAG)n hairpin in the single-strand gap formed during repair. Evidence from work with yeast and human cells is consistent with instability occurring as a result of replication encountering a hairpin (37,40,51,52,70–75). The reduction in repeat lengths in 428 and 313-1-15 cells by 100–120 triplet repeats may reflect the size of a preferential hairpin formation. In bacteria, deletion to preferred sizes has been observed (34,68), as has preferential complete deletion of certain sizes repeat tracts (17). The repeated observation of a >300 repeat deletion in 428 cells may also reflect preferential deletion, although, as discussed above, this may also be due to population selection. Third, it is also possible that DNA damage could initiate the formation of a double-strand break, separate from a replication restart pathway (Figure 8C). Following end resection and homologous end joining, DNA repeats would be lost. In fact, repair of a double-strand break within (CTG)·(CAG) repeats in mammalian cells is associated with repeat deletion (75). An additional possibility is that treatment with DNA damage acts in a non-targeted fashion by inducing a generalized state of instability in cells, which results in deletion at the expanded (CTG)·(CAG) repeats. Persistent genomic instability occurring more than 30 generations following exposure to ionizing radiation is a well-documented phenomenon in cultured cells (76–78). In fact, transgenerational genomic instability following exposure to X-rays has been observed in mice (79).
Figure 8. Models for drug-induced repeat deletion. Left panel. (A) Replication restart. DNA damage within a (CAG)n tract in the leading template strand, indicated by the asterisk, may cause DNA synthesis arrest and replication fork collapse. The dark thick region represents a (CTG)n tract, while the lighter thick region represents the complementary (CAG)n strand. Steps a–i are based on the understanding of the pathway for replication fork restart that involves a double-strand break. This pathway appears to be a major source of spontaneous deletion in E.coli (17,43). Following fork collapse (b and c), fork reversal forms a ‘chicken-foot’ structure (d). If cleaved by a Holliday junction resolvase (at the arrows), a double-strand break ensues (e). End resection by a nuclease will lead to a single-strand 3' end (f), which can initiate homologous strand invasion. During this process, misalignment can occur resulting in a (CAG)n hairpin (g). Branch migration will generate a Holliday junction (h), the resolution of which will restore the replication fork (i). Repeat deletion can occur at the point of misalignment at strand invasion or during synthesis of the restored lagging strand, in which the (CTG)n tract may form a stable hairpin. Right panel, following the introduction of a DNA damage within the repeat, two DNA repair pathways could lead to repeat deletion. (B) Repair involving a gap. An excision repair pathway may open up a gap (c) in which a (CTG)n hairpin may form (d). Slipped misalignment with DNA replication would lead to repeat deletion in the nascent strand (e). Excision of the loop or a subsequent round of replication would lead to repeat deletion. (C) Double-strand break repair. A pathway involving a double-strand break within the repeat (c) may also lead to repeat deletion. Following end resection to generate complementary single strands (d), homologous pairing could occur with flaps (e) that, on removal and ligation, would lead to repair with deletion (f).
The results presented here represent a ‘proof of principle’ that a chemotherapeutic approach to reduction in repeat length is feasible, although several issues must be addressed before this approach can be considered for clinical use. The concentrations of chemotherapeutic agents used in these experiments, 50–100 nM for mitoxantrone, 0.1–0.2 μg/ml for mitomycin C and 0.1–5 μg/ml doxorubicin, are within the therapeutic ranges, although EMS, a potent mutagen, cannot be used clinically. Our data are limited to (CTG)·(CAG) repeat expansion at the DM1 locus. The drugs used may have varying effects with repeats associated with other repeat diseases. The drug effect may differ depending on repeat unit motifs, the repeat length, cis elements adjacent to the repeat and trans factors that may influence instability. In fact, in a similar study of lymphoblast cells containing 124 (CTG)·(CAG) repeats, mitomycin C treatment resulted in an increase in repeat length (59). This result may be due to the large differences in repeat length , or other trans-acting factors that may influence the propensity of expansion to occur in this particular cell line. Whether the same chemotherapeutic agents promote deletions in other repeats at different loci remains to be shown. Nevertheless, the ‘proof of principle’ shown in this study raises a possibility that similar approaches might work for repeat expansions of other diseases.
Our results agree favorably with several other studies showing that DNA-damaging agents, various chemicals and replication inhibitors can influence repeat instability (59–62). Two studies have shown that agents that influence DNA replication or repair can influence repeat instability. Gorbunova et al. (60) showed that the DNA replication inhibitor aphidicolin increased the rate of deletions in CHO cells containing 95 but not 61 (CTG)·(CAG) repeats, while Yang et al. (61) showed that replication inhibitors aphidicolin and emetine, but not mimosine, increased the rate of CTG expansion in DM fibroblast cells. Both studies are consistent with hypothesis that DNA replication is involved in repeat expansion. Gomes-Pereia and Monckton (62) showed that long-term exposure of kidney cells from a transgenic mouse carrying 160 CTG repeats to cytosine arabinoside, ethidium bromide, 5-azacytidine and aspirin reduced the rate of repeat expansion, while exposure to caffeine increased the rate of expansion. Our approach is a bit different in that we tested the effect of chemotherapeutic agents known to damage DNA on the instability of CTG repeats in human lymphoblast cells from DM1 patients. While we saw dramatic reductions in repeat tract length and the accumulation of deletions in the cell population, our results with mitomycin C were in contrast with that of Pineiro et al. (59). This difference may result from the short allele size examined, the potential for repeat expansion in their cell line or other differences with experimental protocols. Thus, at present a general consensus on the effects of chemotherapeutic agents, DNA-damaging agents, or replication inhibitors on repeat instability is not yet available from these independent studies with natural DM1 alleles in human lymphoblast or fibroblast cells, or transgenic (CTG)·(CAG) repeats in CHO and mouse cells. However, the basic idea that the length of expanded DNA repeats associated with human genetic disease can be modulated by various treatments seems well established.
While currently used chemotherapeutic regimes may provide a viable rationale for treatment, drugs that react with high specificity to expanding repeats may be designed based on the structure and characteristics of the DNA helix or the unusual properties of these repeats. CTG and CGG tracts exist as an unusually flexible double helix (80–82), and they can form mismatched hairpins (20,21) and slipped strand DNA structures (83–85). GAA repeats can form intramolecular triplex DNA (25–32) and a bimolecular triplex structure (29). Thus, it is possible that drugs that selectively bind to these sequences and mediate deletion in long repeat tracts may eventually be identified, increasing the selectivity of this basic approach. An additional approach would be to couple a DNA-damaging agent to an oligonucleotide or a PNA (peptide nucleic acid) molecule complementary to a repeat since these can bind very tightly to DNA. This approach has proved successful in directing mutations to DNA adjacent to triplex-forming sequences (86–88). A PNA molecule complementary to CTG repeats has been shown to modulate gene expression in cells (89). A targeted approach to repeat damage may also be important in tissue where cell division may be limited, e.g. muscle or brain. Most model systems examined to date have analyzed repeat lengths in dividing cells . However, it is important to note that expansions are observed in the brain of HD patients (90), and one report has suggested that expansion can occur in sperm during maturation (91). Thus, repeat deletion may be expected to occur in non-dividing cells during chromosome repair and maintenance. Moreover, many of the late onset and complex characteristics of DM1 may result from, or be aggravated by, somatic expansion that occurs throughout life (13,14).
In conclusion, DNA repeat expansion diseases form a large class of genetic disorders linked mechanistically by a single type of mutation, the expansion of (CTG)·(CAG), (CGG)·(CCG), (GAA)·(TTC), (CCTG)·(CAGG) or (ATTCT)·(AGAAT) repeats. All known repeat expansion diseases cause progressive neurological manifestations that often lead to total disability and death. At present, no treatment can delay, halt or reverse the progression of these diseases. The application of a chemotherapeutic drug approach may open a gate to a new therapy for the underlying etiology of these diseases.
ACKNOWLEDGEMENTS
We thank William Ondo for helpful comments and suggestions. This work was supported by grants from NIEHS (R.R.S.) and NIH (T.A.).
REFERENCES
Wells,R.D. and Warren,S.T. ( (1998) ) Genetic Instabilities and Hereditary Neurological Diseases. Academic Press, San Diego, CA.
Sinden,R.R. ( (1999) ) Biological implications of the DNA structures associated with disease-causing triplet repeats. Am. J. Hum. Genet., , 64, , 346–353.
Cummings,C.J. and Zoghbi,H.Y. ( (2000) ) Fourteen and counting: unraveling trinucleotide repeat diseases. Hum. Mol. Genet., , 9, , 909–916.
Heintz,N. and Zoghbi,H.Y. ( (2000) ) Insights from mouse models into the molecular basis of neurodegeneration. Annu. Rev. Physiol., , 62, , 779–802.
Cleary,J.D. and Pearson,C.E. ( (2003) ) The contribution of CIS-elements to disease-associated repeat instability: clinical and experimental evidence. Cytogenet. Genome Res., , 100, , 25–55.
Ashizawa,T. and Zoghbi,H.Y. ( (1997) ) Diseases with trinucleotide repeat expansion. In Appel,S.H. (ed.), Current Neurology. IOS Press, Amsterdam, Vol. 17, pp. 79–135.
Sato,T., Oyake,M., Nakamura,K., Nakao,K., Fukusima,Y., Onodera,O., Igarashi,S., Takano,H., Kikugawa,K., Ishida,Y. et al. ( (1999) ) Transgenic mice harboring a full-length human mutant DRPLA gene exhibit age-dependent intergenerational and somatic instabilities of CAG repeats comparable with those in DRPLA patients. Hum. Mol. Genet., , 8, , 99–106.
Monckton,D.G., Coolbaugh,M.I., Ashizawa,K.T., Siciliano,M.J. and Caskey,C.T. ( (1997) ) Hypermutable myotonic dystrophy CTG repeats in transgenic mice. Nature Genet., , 15, , 193–196.
Fortune,M.T., Vassilopoulos,C., Coolbaugh,M.I., Siciliano,M.J. and Monckton,D.G. ( (2000) ) Dramatic, expansion-biased, age-dependent, tissue-specific somatic mosaicism in a transgenic mouse model of triplet repeat instability. Hum. Mol. Genet., , 9, , 439–445.
Mangiarini,L., Sathasivam,K., Mahal,A., Mott,R., Seller,M. and Bates,G.P. ( (1997) ) Instability of highly expanded CAG repeats in mice transgenic for the Huntington's disease mutation. Nature Genet., , 15, , 197–200.
Seznec,H., Lia-Baldini,A.S., Duros,C., Fouquet,C., Lacroix,C., Hofmann-Radvanyi,H., Junien,C. and Gourdon,G. ( (2000) ) Transgenic mice carrying large human genomic sequences with expanded CTG repeat mimic closely the DM CTG repeat intergenerational and somatic instability. Hum. Mol. Genet., , 9, , 1185–1194.
Holmes,S.E., Hearn,E.O., Ross,C.A. and Margolis,R.L. ( (2001) ) SCA12: an unusual mutation leads to an unusual spinocerebellar ataxia. Brain Res. Bull., , 56, , 397–403.
Wong,L.-J.C., Ashizawa,T., Monckton,D.G., Caskey,C.T. and Richards,C.S. ( (1995) ) Somatic heterogeneity of the CTG repeat in myotonic dystrophy is age and size dependent. Am. J. Hum. Genet., , 56, , 114–122.
Martorell,L., Monckton,D.G., Gamez,J., Johnson,K.J., Gich,I., de Munain,A.L. and Baiget,M. ( (1998) ) Progression of somatic CTG repeat length heterogeneity in the blood cells of myotonic dystrophy patients. Hum. Mol. Genet., , 7, , 307–312.
Wansink,D.G. and Wieringa,B. ( (2003) ) Transgenic mouse models for myotonic dystrophy type 1 (DM1). Cytogenet. Genome Res., , 100, , 230–242.
Oussatcheva,E.A., Hashem,V.I., Zou,Y., Sinden,R.R. and Potaman,V.N. ( (2001) ) Involvement of the nucleotide excision repair protein UvrA in instability of (CAG)·(CTG) repeat sequences in E.coli. J. Biol. Chem., , 276, , 30878–30884.
Hashem,V.I., Rosche,W.A. and Sinden,R.R. ( (2002) ) Genetic assays for measuring rates of (CAG)·(CTG) repeat instability in Escherichia coli. Mutat. Res., , 502, , 25–37.
Jakupciak,J.P. and Wells,R.D. ( (1999) ) Genetic instabilities in (CTG)·(CAG) repeats occur by recombination. J. Biol. Chem., , 274, , 23468–23479.
Pearson,C.E. and Sinden,R.R. ( (1998) ) Trinucleotide repeat DNA structures: Dynamic mutations from dynamic DNA. Curr. Opin. Struct. Biol., , 8, , 321–330.
Mitas,M. ( (1997) ) Trinucleotide repeats associated with human disease. Nucleic Acids Res., , 25, , 2245–2253.
Darlow,J.M. and Leach,D.R.F. ( (1998) ) Secondary structures in d(CGG) and d(CCG) repeat tracts. J. Mol. Biol., , 275, , 3–16.
Pearson,C.E. and Sinden,R.R. ( (1998) ) Slipped strand DNA, dynamic mutations, and human disease. In Wells,R.D. and Warren,S.T. (eds.), Genetic Instabilities and Hereditary Neurological Disorders. Academic Press, San Diego, CA, pp. 585–621.
Usdin,K. and Woodford,K.J. ( (1995) ) CGG repeats associated with DNA instability and chromosome fragility form structures that block DNA synthesis in vitro. Nucleic Acids Res., , 23, , 4202–4209.
Fry,M. and Loeb,L.A. ( (1994) ) The fragile X syndrome d(CGG)n nucleotide repeats form a stable tetrahelical structure. Proc. Natl Acad. Sci. USA, , 91, , 4950–4954.
Bidichandani,S.I., Ashizawa,T. and Patel,P.I. ( (1998) ) The GAA triplet-repeat expansion in Friedreich ataxia interferes with transcription and may be associated with an unusual DNA structure. Am. J. Hum. Genet., , 62, , 111–121.
Mariappan,S.V., Catasti,P., Silks,L.A., III, Bradbury,E.M. and Gupta,G. ( (1999) ) The high-resolution structure of the triplex formed by the GAA/TTC triplet repeat associated with Friedreich's ataxia. J. Mol. Biol., , 285, , 2035–2052.
Grabczyk,E. and Usdin,K. ( (2000) ) The GAA·TTC triplet repeat expanded in Friedreich's ataxia impedes transcription elongation by T7 RNA polymerase in a length and supercoil dependent manner. Nucleic Acids Res., , 28, , 2815–2822.
LeProust,E.M., Pearson,C.E., Sinden,R.R. and Gao,X. ( (2000) ) Unexpected formation of parallel duplex in GAA and TTC trinucleotide repeats of Friedreich's ataxia. J. Mol. Biol., , 302, , 1063–1080.
Sakamoto,N., Chastain,P.D., Parniewski,P., Ohshima,K., Pandolfo,M., Griffith,J.D. and Wells,R.D. ( (1999) ) Sticky DNA: self-association properties of long GAA·TTC repeats in R·R·Y triplex structures from Friedreich's ataxia. Mol. Cell, , 3, , 465–475.
Potaman,V.N., Oussatcheva,E.A., Lyubchenko,Y.L., Shlyakhtenko,L.S., Bidichandani,S.I., Ashizawa,T. and Sinden,R.R. ( (2004) ) Length-dependent structure formation in Friedreich ataxia (GAA)n·(TTC)n repeats at neutral pH. Nucleic Acids Res., , 32, , 1224–1231.
Ohshima,K., Montermini,L., Wells,R.D. and Pandolfo,M. ( (1998) ) Inhibitory effects of expanded GAA·TTC triplet repeats from intron I of the Friedreich ataxia gene on transcription and replication in vivo. J. Biol. Chem., , 273, , 14588–14595.
Ohshima,K., Kang,S., Larson,J.E. and Wells,R.D. ( (1996) ) Cloning, characterization, and properties of seven triplet repeat DNA sequences. J. Biol. Chem., , 271, , 16773–16783.
Potaman,V.N., Bissler,J.J., Hashem,V.I., Oussatcheva,E.A., Lu,L., Shlyakhtenko,L.S., Lyubchenko,Y.L., Matsuura,T., Ashizawa,T., Leffak,M. et al. ( (2003) ) Unpaired structures in SCA10 (ATTCT)n·(AGAAT)n repeats. J. Mol Biol., , 326, , 1095–1111.
Kang,S., Jaworski,A., Ohshima,K. and Wells,R.D. ( (1995) ) Expansion and deletion of CTG triplet repeats from human disease genes are determined by the direction of replication in E.coli. Nature Genet., , 10, , 213–218.
Darlow,J.M. and Leach,D.R.F. ( (1995) ) The effects of trinucleotide repeats found in human inherited disorders on palindrome inviability in Escherichia coli suggest hairpin folding preferences in vivo. Genetics, , 141, , 825–832.
Darlow,J.M. and Leach,D.F. ( (1998) ) Evidence for two preferred hairpin folding patterns in d(CGG)·d(CCG) repeat tracts in vivo. J. Mol. Biol., , 275, , 17–23.
Freudenreich,C.H., Stavenhagen,J.B. and Zakian,V.A. ( (1997) ) Stability of a CTG/CAG trinucleotide repeat in yeast is dependent on its orientation in the genome. Mol. Cell. Biol., , 17, , 2090–2098.
Cleary,J.D., Nichol,K., Wang,Y.H. and Pearson,C.E. ( (2002) ) Evidence of cis-acting factors in replication-mediated trinucleotide repeat instability in primate cells. Nature Genet., , 31, , 37–46.
Ireland,M.J., Reinke,S.S. and Livingston,D.M. ( (2000) ) The impact of lagging strand replication mutations on the stability of CAG repeat tracts in yeast. Genetics, , 155, , 1657–1665.
Schweitzer,J.K. and Livingston,D.M. ( (1999) ) The effect of DNA replication mutations on CAG tract stability in yeast. Genetics, , 152, , 953–963.
Trinh,T.Q. and Sinden,R.R. ( (1991) ) Preferential DNA secondary structure mutagenesis in the lagging strand of replication in E.coli. Nature, , 352, , 544–547.
Sinden,R.R., Zheng,G., Brankamp,R.G. and Allen,K.N. ( (1991) ) On the deletion of inverted repeated DNA in Escherichia coli: Effects of length, thermal stability, and cruciform formation in vivo. Genetics, , 129, , 991–1005.
Hashem,V.I., Rosche,W.A. and Sinden,R.R. ( (2004) ) Genetic recombination destabilizes (CTG)n (CAG)n repeats in E.coli. Mutat. Res., , 554, , 95–109.
Pelletier,R., Krasilnikova,M.M., Samadashwily,G.M., Lahue,R. and Mirkin,S.M. ( (2003) ) Replication and expansion of trinucleotide repeats in yeast. Mol. Cell. Biol., , 23, , 1349–1357.
Krasilnikova,M.M. and Mirkin,S.M. ( (2004) ) Replication stalling at Friedreich's ataxia (GAA)n repeats in vivo. Mol. Cell. Biol., , 24, , 2286–2295.
Samadashwily,G.M., Raca,G. and Mirkin,S.M. ( (1997) ) Trinucleotide repeats affect DNA replication in vivo. Nature Genet., , 17, , 298–304.
Callahan,J.L., Andrews,K.J., Zakian,V.A. and Freudenreich,C.H. ( (2003) ) Mutations in yeast replication proteins that increase CAG/CTG expansions also increase repeat fragility. Mol. Cell. Biol., , 23, , 7849–7860.
Freudenreich,C.H., Kantrow,S.M. and Zakian,V.A. ( (1998) ) Expansion and length-dependent fragility of CTG repeats in yeast. Science, , 279, , 853–856.
Freudenreich,C.H. and Lahiri,M. ( (2004) ) Structure-forming CAG/CTG repeat sequences are sensitive to breakage in the absence of Mrc1 checkpoint function and S-phase checkpoint signaling: implications for trinucleotide repeat expansion diseases. Cell Cycle, , 3, .
Lahiri,M., Gustafson,T.L., Majors,E.R. and Freudenreich,C.H. ( (2004) ) Expanded CAG repeats activate the DNA damage checkpoint pathway. Mol. Cell, , 15, , 287–293.
Liu,Y., Zhang,H., Veeraraghavan,J., Bambara,R.A. and Freudenreich,C.H. ( (2004) ) Saccharomyces cerevisiae flap endonuclease 1 uses flap equilibration to maintain triplet repeat stability. Mol. Cell. Biol., , 24, , 4049–4064.
Rolfsmeier,M.L., Dixon,M.J., Pessoa-Brandao,L., Pelletier,R., Miret,J.J. and Lahue,R.S. ( (2001) ) Cis-elements governing trinucleotide repeat instability in Saccharomyces cerevisiae. Genetics, , 157, , 1569–1579.
Dudas,A. and Chovanec,M. ( (2004) ) DNA double-strand break repair by homologous recombination. Mutat. Res., , 566, , 131–167.
van den Bosch,M., Lohman,P.H. and Pastink,A. ( (2002) ) DNA double-strand break repair by homologous recombination. Biol. Chem., , 383, , 873–892.
Symington,L.S. ( (2002) ) Role of RAD52 epistasis group genes in homologous recombination and double-strand break repair. Microbiol. Mol. Biol. Rev., , 66, , 630–70, table.
Lees-Miller,S.P. and Meek,K. ( (2003) ) Repair of DNA double strand breaks by non-homologous end joining. Biochimie, , 85, , 1161–1173.
Friedberg,E.C., Walker,G.C. and Siede,W. ( (1995) ) DNA Repair and Mutagenesis. ASM Press, Washington, DC, pp. 213–214, 288.
Hashem,V.I. and Sinden,R.R. ( (2002) ) Chemotherapeutically induced deletion of expanded triplet repeats. Mutat. Res., , 508, , 107–119.
Pineiro,E., Fernandez-Lopez,L., Gamez,J., Marcos,R., Surralles,J. and Velazquez,A. ( (2003) ) Mutagenic stress modulates the dynamics of CTG repeat instability associated with myotonic dystrophy type 1. Nucleic Acids Res., , 31, , 6733–6740.
Gorbunova,V., Seluanov,A., Dion,V., Sandor,Z., Meservy,J.L. and Wilson,J.H. ( (2003) ) Selectable system for monitoring the instability of CTG/CAG triplet repeats in mammalian cells. Mol. Cell. Biol., , 23, , 4485–4493.
Yang,Z., Lau,R., Marcadier,J.L., Chitayat,D. and Pearson,C.E. ( (2003) ) Replication inhibitors modulate instability of an expanded trinucleotide repeat at the myotonic dystrophy type 1 disease locus in human cells. Am. J. Hum. Genet., , 73, , 1092–1105.
Gomes-Pereira,M. and Monckton,D.G. ( (2004) ) Chemically induced increases and decreases in the rate of expansion of a CAG·CTG triplet repeat. Nucleic Acids Res., , 32, , 2865–2872.
Ashizawa,T., Monckton,D.G., Vaishnav,S., Patel,B.J., Voskova,A. and Caskey,C.T. ( (1996) ) Instability of the expanded (CTG)n repeats in the myotonin protein kinase gene in cultured lymphoblastoid cell lines from patients with myotonic dystrophy. Genomics, , 36, , 47–53.
Khajavi,M., Tari,A.M., Patel,N.B., Tsuji,K., Siwak,D.R., Meistrich,M.L., Terry,N.H. and Ashizawa,T. ( (2001) ) ‘Mitotic drive’ of expanded CTG repeats in myotonic dystrophy type 1 (DM1). Hum. Mol. Genet., , 10, , 855–863.
Monckton,D.G., Wong,L.J., Ashizawa,T. and Caskey,C.T. ( (1995) ) Somatic mosaicism, germline expansions, germline reversions and intergenerational reductions in myotonic dystrophy males: small pool PCR analyses. Hum. Mol. Genet., , 4, , 1–8.
Gomes-Pereira,M., Bidichandani,S.I. and Monckton,D.G. ( (2004) ) Analysis of unstable triplet repeats using small-pool polymerase chain reaction. Methods Mol. Biol., , 277, , 61–76.
Martorell,L., Gamez,J., Cayuela,M.L., Gould,F.K., McAbney,J.P., Ashizawa,T., Monckton,D.G. and Baiget,M. ( (2004) ) Germline mutational dynamics in myotonic dystrophy type 1 males: allele length and age effects. Neurology, , 62, , 269–274.
Rosche,W.A., Jaworski,A., Kang,S., Kramer,S.F., Larson,J.E., Geidroc,D.P., Wells,R.D. and Sinden,R.R. ( (1996) ) Single-stranded DNA-binding protein enhances the stability of CTG triplet repeats in Escherichia coli. J. Bacteriol., , 178, , 5042–5044.
McGlynn,P. and Lloyd,R.G. ( (2002) ) Recombinational repair and restart of damaged replication forks. Nature Rev. Mol. Cell Biol., , 3, , 859–870.
Miret,J.J., Pessoa-Brandao,L. and Lahue,R.S. ( (1998) ) Orientation-dependent and sequence-specific expansions of CTG/CAG trinucleotide repeats in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA, , 95, , 12438–12443.
Bhattacharyya,S. and Lahue,R.S. ( (2004) ) Saccharomyces cerevisiae Srs2 DNA helicase selectively blocks expansions of trinucleotide repeats. Mol. Cell. Biol., , 24, , 7324–7330.
Schweitzer,J.K. and Livingston,D.M. ( (1998) ) Expansions of CAG repeat tracts are frequent in a yeast mutant defective in Okazaki fragment maturation. Hum. Mol. Genet, , 7, , 69–74.
Lee,S. and Park,M.S. ( (2002) ) Human FEN-1 can process the 5'-flap DNA of CTG/CAG triplet repeat derived from human genetic diseases by length and sequence dependent manner. Exp. Mol. Med., , 34, , 313–317.
Panigrahi,G.B., Cleary,J.D. and Pearson,C.E. ( (2002) ) In vitro (CTG)·(CAG) expansions and deletions by human cell extracts. J. Biol. Chem., , 277, , 13926–13934.
Marcadier,J.L. and Pearson,C.E. ( (2003) ) Fidelity of primate cell repair of a double-strand break within a (CTG)·(CAG) tract. Effect of slipped DNA structures. J. Biol. Chem., , 278, , 33848–33856.
Little,J.B., Gorgojo,L. and Vetrovs,H. ( (1990) ) Delayed appearance of lethal and specific gene mutations in irradiated mammalian cells. Int. J. Radiat. Oncol. Biol. Phys., , 19, , 1425–1429.
Little,J.B. ( (1998) ) Radiation-induced genomic instability. Int. J. Radiat. Biol., , 74, , 663–671.
Wright,E.G. ( (1998) ) Radiation-induced genomic instability in haemopoietic cells. Int. J. Radiat. Biol., , 74, , 681–687.
Dubrova,Y.E., Plumb,M., Gutierrez,B., Boulton,E. and Jeffreys,A.J. ( (2000) ) Transgenerational mutation by radiation. Nature, , 405, , 37.
Chastain,P.D., Eichler,E.E., Kang,S., Nelson,D.L., Levene,S.D. and Sinden,R.R. ( (1995) ) Anomalous rapid electrophoretic mobility of DNA containing triplet repeats associated with human disease genes. Biochemistry, , 34, , 16125–16131.
Chastain,P.D. and Sinden,R.R. ( (1998) ) CTG repeats associated with human genetic disease are inherently flexible. J. Mol. Biol., , 275, , 405–411.
Bacolla,A., Gellibolian,R., Shimizu,M., Amirhaeri,S., Kang,S., Ohshima,K., Larson,J.E., Harvey,S.C., Stollar,B.D. and Wells,R.D. ( (1997) ) Flexible DNA - genetically unstable CTG·CAG and CGG·CCG from human hereditary neuromuscular disease genes. J. Biol. Chem., , 272, , 16783–16792.
Pearson,C.E. and Sinden,R.R. ( (1996) ) Alternative structures in duplex DNA formed within the trinucleotide repeats of the myotonic dystrophy and fragile X loci. Biochemistry, , 35, , 5041–5053.
Pearson,C.E. and Sinden,R.R. ( (1998) ) Trinucleotide repeat DNA structures: dynamic mutations from dynamic DNA. Curr. Opin. Struct. Biol., , 8, , 321–330.
Pearson,C.E., Wang,Y.H., Griffith,J.D. and Sinden,R.R. ( (1998) ) Structural analysis of slipped-strand DNA (S-DNA) formed in (CTG)n·(CAG)n repeats from the myotonic dystrophy locus. Nucleic Acids Res., , 26, , 816–823.
Glazer,P.M., Wang,G., Havre,P.A. and Gunther,E.J. ( (1996) ) Targeted mutagenesis mediated by the triple helix formation. Methods Mol. Biol., , 57, , 109–118.
Chan,P.P. and Glazer,P.M. ( (1997) ) Triplex DNA: fundamentals, advances, and potential applications for gene therapy. J. Mol. Med., , 75, , 267–282.
Vasquez,K.M. and Wilson,J.H. ( (2000) ) Triplex-directed site-specific genome modification. Methods Mol. Biol., , 133, , 183–200.
Boffa,L.C., Morris,P.L., Carpaneto,E.M., Louissaint,M. and Allfrey,V.G. ( (1996) ) Invasion of the CAG triplet repeats by a complementary peptide nucleic acid inhibits transcription of the androgen receptor and TATA-binding protein genes and correlates with refolding of an active nucleosome containing a unique AR gene sequence. J. Biol. Chem., , 271, , 13228–13233.
Kennedy,L., Evans,E., Chen,C.M., Craven,L., Detloff,P.J., Ennis,M. and Shelbourne,P.F. ( (2003) ) Dramatic tissue-specific mutation length increases are an early molecular event in Huntington disease pathogenesis. Hum. Mol. Genet., , 12, , 3359–3367.
Kovtun,I.V. and McMurray,C.T. ( (2001) ) Trinucleotide expansion in haploid germ cells by gap repair. Nature Genet., , 27, , 407–411.(Vera I. Hashem, Malgorzata J. Pytlos, El)