Human Molecular Genetics, 2002, Vol. 11, No. 7 791-798
© 2002 Oxford University Press
Age and insertion site dependence of repeat number instability of a human DM1 transgene in individual mouse sperm
Department of Experimental Radiation Oncology and 1Department of Molecular Genetics, The University of Texas M. D. Anderson Cancer Center, 2Department of Environmental Sciences, The University of Texas School of Public Health, Houston, TX 77030, USA and 3Institute of Biomedical and Life Sciences, University of Glasgow, Anderson College, Glasgow G11 6NU, UK
Received November 28, 2001; Revised and Accepted January 28, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Precise measurement of germline repeat number mutations is important for understanding the molecular etiology of expanded trinucleotide repeat diseases. We used single genome-equivalent PCR of sperm DNA to measure the mutation frequencies in two lines of Dmt transgenic mice containing an expanded CTG·CAG tract on an identical genetic background. Single genome-equivalent PCR indicated that apparent mutational spectra derived in other investigations from PCR of bulk sperm DNA were largely the consequence of PCR stutter and not mutations. Here we show that sperm from 8-week-old Dmt-D mice had a significantly higher mutation frequency (change of >1 repeat) (14.2%) than those of Dmt-E mice of the same age (5.5%), in agreement with pedigree analysis. Furthermore, the mutation frequency in sperm of Dmt-D mice increased significantly with age (28.0% at 17 weeks). The age dependence of the degree of expansion implies that mutations accumulate with time in spermatogenic stem cells. Similar rates of expansion per spermatogenic cycle in man would yield the large expansions observed in human diseases such as myotonic dystrophy type 1. Pedigree data showed a significant age-dependent bias toward repeat contraction in female transmissions and a trend towards expansion with age in male transmissions. Thus, direct single genome-equivalent PCR of the sperm DNA of an individual male appears to predict the distribution of mutant allele sizes that might be inherited by its offspring. In further contrast to a recent report, the sex of the offspring had no detectable effect on the direction of the mutational length change.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Several trinucleotide repeat loci in the human genome are associated with genetic diseases such as myotonic dystrophy type 1 (DM1), Huntington’s disease (HD) and dentatorubral-pallidoluysian atrophy (DRPLA). The size of the repeat expansion is directly related to the occurrence and severity of the disease. Increasing repeat length of expanded disease-associated alleles from one generation to the next underlies the commonly observed anticipation in these disorders. Although the molecular mechanisms involved in the repeat number instability in DM1 and other human trinucleotide repeat diseases are as yet unclear (1,2), large repeats are subject to greater additional instability both in somatic and germinal cells (3) and both cis- and trans-acting factors are involved (4,5).
Pedigree analyses have attempted to measure germline instability from the relationship of the allele sizes in the parent and their offspring by direct comparison of the repeat numbers in their somatic cell DNA. There is only one report confirming that a CAG·CTG repeat expands with increasing age of the father (6). One explanation for the paucity of such reports is that in many trinucleotide repeat diseases, including DM1, somatic instability is a major factor, and often the large differences in age between the parent and their offspring at the time of blood sample collection confounds the analysis (4). Only by studying the DM1 instability in the sperm from a father may one predict the distribution of allele lengths that could be inherited by his offspring (7). However, attempts to describe the effects of age on germline repeat number mutations in human sperm have had only marginal success (8).
Transgenic mouse models of human diseases caused by trinucleotide repeat expansion provide a useful tool to elucidate the molecular mechanisms underlying trinucleotide repeat instability. Monckton et al. (9) generated transgenic mice with the same genetic background using only the 3'-untranslated region (3'-UTR) of the DMPK gene containing 162 CTG·CAG repeats from a DM1 patient. Two of these lines, Dmt-D and Dmt-E, carrying only a single copy of the expanded trinucleotide repeat, had high and low mutation frequencies, respectively, suggesting that the sites of transgene integration may contribute to the mutation frequency. The Dmt-D line showed a dramatic age-dependent, expansion-biased and tissue-specific somatic instability, which is similar to the somatic mutations observed in DM1 patients (10). The expanded DM1 CTG repeat was also unstable upon intergenerational transmission with a bias toward further expansion (9), but the age dependence of this effect was not analyzed. In some other transgenic mouse models, pedigree analyses did show an age-dependent increase in paternally transmitted expansions of trinucleotide repeats (6,11,12), but such an increase could not be demonstrated in all cases (13).
Most recently, a study using PCR amplification of bulk sperm DNA reported that there was no age-dependent alteration in repeat length of an expanded HD transgene in male mice (14). However, it seemed unlikely to us that measurements on bulk DNA (15) could detect small changes in repeat number consisting of both expansions and contractions that differed from the progenitor by one to three repeats, because they would be hidden by the already broadened distribution of repeat sizes produced by PCR stutter when this large trinucleotide repeat is amplified. Small-pool PCR (SP-PCR) involving up to 40 genome-equivalents (g.e.) of input DNA has already been applied to measure somatic mutation frequencies at expanded trinucleotide repeat loci (4). However, when the tract is long, the artifactual stutter due to DNA replication slippage during the PCR amplification obscures mutant bands involving small shifts. There are two approaches to obtaining single DNA molecules to refine this analysis. One involves sorting of individual sperm cells (16,17). In this study we took the second approach, diluting the DNA to the single g.e. level in each PCR tube, and showed that this could resolve most mutations from the progenitor. Employing such techniques, we were able to directly demonstrate differences in germ-line mutation frequencies between two lines of Dmt mice and an age-dependent increase in mutation frequency in one of the lines.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Determination of allele sizes in sperm and relationship to progenitor
When performing PCR of expanded trinucleotide repeats with many g.e. of input DNA, it is not possible to determine which changes are due to PCR stutter and which are due to mutant molecules present within the sample. However, with single g.e. PCR, the stutter bands, which are 3 bp apart, observed in the PCR products of the expanded DM1 CTG repeat (Fig. 1, lanes 2 and 4) must be caused by slippage during PCR amplification in vitro because most of the reaction mixtures would have contained only a single molecule with the DM1 repeat. Since the breadth of the distribution of the stutter bands limits the detection of small repeat number changes, we sought to minimize this stutter. We chose a single round of PCR with 35 cycles, since that was the fewest number of cycles with which we could reliably detect the PCR product using fluorescent detection with a DNA sequencer.
|
The distribution of bands with different numbers of repeats in the 100 g.e. PCR (Fig. 1, lane 1) was only slightly broader than the distributions from the single g.e. PCR bands (lanes 2 and 4). The former represented the combination of in vivo mutation and PCR stutter, whereas the latter represented only PCR stutter. Therefore, the multiple peaks shown by Kovtun et al. (15) must largely have represented PCR stutter, not just mutations in sperm as claimed by the authors. It is only by using single g.e. PCR that mutations falling within the range of the stutter can be distinguished.
Although most workers choose the highest peak on the trace as the progenitor band (6,11,18), we do not believe that this choice is correct. PCR stutter with small trinucleotide repeats always generates additional bands smaller than the progenitor in that reaction (19). Furthermore, the distribution of stutter bands from single g.e. PCR amplifications of a large trinucleotide repeat was asymmetric (Fig. 1, lanes 2 and 4); there appeared to be a better defined first band on the right (large repeat number) side of the distribution and a broader tail on the left side. In an attempt to fit these observations, we developed a model of PCR stutter in which the progenitor is the band with the largest repeat size, and during each PCR cycle there is the same probability, S, of loss of one repeat. As the number of cycles increases, the location of the highest peak moves down, away from the progenitor, and the distribution remains skewed to the left. This model fits the observed peak heights for 35 cycles of amplification of single g.e. sperm DNA when S is chosen to be equal to 0.0775 per PCR cycle (Fig. 2).
|
The model supports the concept that the mode bears a fixed relationship to the progenitor that depends only on the number of PCR cycles (fixed at 35) and S. In practice, since it is often difficult to identify the largest band on the right side of the distribution, measuring the maximum point (mode) of the distribution is more practical. To account for variations in S between PCR amplifications, the modes for individual sperm samples were compared with those for 100 g.e. PCR amplifications on mouse-tail DNA run simultaneously. The modes were determined by fitting a cubic equation (Fig. 1, lanes 1, 2 and 4), since this is more accurate than simply choosing the highest peak.
To determine the reproducibility of the method, we loaded identical single g.e. PCR products from the same reaction in different lanes of the same gel or on different gels. The variations within a gel and between pairs of gels run on different days were 0.06 ± 0.02 repeats and 0.13 ± 0.04 repeats, respectively. The variation between the PCR amplifications, determined using the products from different 300 g.e. PCR amplifications performed on the same DNA sample, which should have similar mixtures of in vivo repeat number mutations, and loaded on the same gel, was 0.5 ± 0.2 repeats. Thus, the overall variation of the assay was a little more than 0.5 CTG repeats. A result in which a sperm has one repeat change could sometimes be due to variation in the assay, so a cut-off point of greater than one repeat shift was used for scoring mutations.
Because of the significant somatic and germline mutation frequencies, in addition to the PCR stutter, we were not able to precisely determine the progenitor allele size for each mouse. Therefore, a consistent method for measuring differences of individual sperm repeat sizes from the progenitor for that mouse was developed. Tail DNA obtained at 3 weeks of age was used as a standard for that mouse. The modal size of the DM1 CTG repeat from the 3-week tail, using PCR with 1 ng DNA input, was not significantly different from the sizes obtained with newborn tail or 3-week blood, lung, heart or testes from the same Dmt-D mouse (data not shown). The absence of a shift with age in this range or tissue type suggests that the average size of the DM1 CTG repeat in 3-week tail is representative of that of the progenitor for the mouse. Using these methods, we identified mutations in the samples in Figure 1 as follows. The tail DNA distribution had a mode of 154.4 repeats (data not shown), similar to that of the 100 g.e. PCR on sperm (lane 1). Whereas the sperm in lane 2 was normal (mode of 154.4 repeats), the one in lane 4 (modal repeat number of 157.5) showed a gain of almost exactly three repeats from the progenitor represented by the tail DNA. Empty lanes (e.g. lane 3) were expected because single g.e. PCR is based on dilution rather than single sperm sorting.
Spontaneous mutation frequencies in sperm from different mice
Multiple Dmt-D mice at 8 and 17 weeks of age were used to determine the effect of age on the spontaneous germline mutation frequency. Eight-week-old Dmt-E mice were used to compare the spontaneous germline mutation frequencies with Dmt-D mice of the same age. Because expanded trinucleotide repeats have high spontaneous mutation frequencies, some mice, as observed with the Dmt-D line (9), have two major populations of cells with different repeat numbers resulting from the generation of mutant alleles in the early post-zygotic cell divisions. To identify such mice, we performed single g.e. PCR on tail DNA from nine 3-week-old Dmt-D mice. The apparent mutation frequencies, assuming a single progenitor with a modal value from the 100 g.e. PCR on the same DNA, were between 2.8 and 12.2% for eight of the mice (7.0 ± 2.9% SD) but 23.8% for one mouse (no. 747). This last frequency was significantly (P < 0.05) higher than those in all other mice, but there were no significant differences between the frequencies in any other mice. The single g.e. PCR on both 3-week-old tail and 8-week-old sperm DNA indicated that there might be two subpopulations with 152 and 155 repeats, a feature that was not apparent from PCR on 100 g.e. of DNA. Since mouse no. 747 was a mosaic arising from a repeat number mutation in the DM1 CTG repeat in early embryogenesis, it was excluded from the study. Mutation frequencies in sperm DNA from the remaining 12 mice (four in each group) were measured by single g.e. PCR (Fig. 3). Because large expansions could be missed with this technique (10), sperm were also analyzed by single g.e. PCR and SP-PCR with 20 g.e. of DNA, using agarose gel electrophoresis and Southern blotting. The mutations observed on the DNA fragment analyzer were confirmed, and no large expansions were detected (data not shown).
|
The measured mutation frequencies in these groups of mice are shown in Table 1. The average numbers of molecules per PCR amplification (
) were below 1.0 and were similar in each group except for the Dmt-E embryo, which was even lower than the others. Within each group of animals tested the mutation frequencies measured were not statistically significantly different from each other, indicating a high degree of determinism in the mutational process in animals with a common genetic background and shared environment. As expected, embryonic DNA from Dmt-E mice, with a 3.2% mutation frequency, was the most homogeneous at the DM1 locus. The average spontaneous mutation frequency in the sperm of 8-week-old Dmt-E mice, 5.5%, was not significantly different from that in the embryo. In contrast, the average spontaneous mutation frequency in the sperm of 8-week-old Dmt-D mice was 14.2%, which was significantly higher than that in the Dmt-E mice by two methods of statistical analyses (P < 0.01), showing the dependence of mutation frequency on the site of insertion of the transgene. The average spontaneous mutation frequencies in the sperm of 17-week-old Dmt-D mice, 28.0%, was significantly greater than that at 8 weeks of age (P < 0.01), indicating an age dependence.
|
The distributions of the DM1 allele sizes in individual sperm from the three groups of mice (Fig. 3) showed that gains and losses of CTG repeats were not equally distributed. Significant biases to deletion were observed in the sperm of the Dmt-E mice (Fig. 3A, two gains versus 16 losses, P = 0.02) and Dmt-D mice (Fig. 3C, six gains versus 38 losses, P < 0.001) at 8 weeks of age. A significant bias towards small expansions was observed in sperm DNA from the 17-week-old Dmt-D mice (Fig. 3D, 64 gains versus 23 losses, P = 0.001). The overall distribution of sizes of repeat number mutations also differed between the sperm from the two lines at 8 weeks of age and between Dmt-D mice at 8 and 17 weeks of age (P < 0.001).
Further analysis of pedigree data
Having observed an age-of-parent effect in the sperm data, we analyzed our previously collected transmission data (9) to see if this effect was also reflected in mouse pedigrees. These pedigree data had previously revealed a bias toward deletions during female transmissions and a bias toward expansions during male transmissions. Regression analysis revealed significant correlations between the age of the mother and the size of the deletion in maternal transmissions for all four lines studied: Dmt-B (r = 0.46, P < 0.005); Dmt-C (r = 0.33, P < 0.005); Dmt-D (r = 0.63, P < 0.05); and Dmt-E (r = 0.71, P < 0.0001) (Fig. 4A). Analysis of male transmissions revealed a significant correlation between the age of the father and the size of the expansion only for the Dmt-B line (r = 0.46, P < 0.05). Similar trends were observed for Dmt-C and Dmt-D (r = 0.42, P = 0.09), but they were not significant, perhaps because the numbers of offspring were small (n = 20 for Dmt-C and n = 17 for Dmt-D). No trend was apparent for Dmt-E, but the male germline mutation frequency for this line was very low (only three mutants observed in 35 transmissions). However, combining the male transmission data for all four lines revealed a highly significant correlation between the size of the allele transmitted and the age of the father (r = 0.37, P < 0.0005) (Fig. 4A).
|
Recently, it has been reported that the direction of repeat number change mutations observed in intergenerational transmissions in another transgenic model of expanded CAG·CTG repeats was dependent on the sex of the offspring (15). We have analyzed our pedigree data from all four lines for such an effect, but there were no significant deviations, nor indeed any apparent trends, from the overall distribution of mutant alleles in male or female offspring from male or female transmissions for each individual line (data sets small, not shown) or from the combined data from all four lines (P > 0.1, Fisher’s exact test) (Fig. 4B).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In order to understand the anticipatory process in inherited human disorders it is essential to accumulate accurate data regarding the germline stability of expanded triplet repeat sequences. This can be achieved by pedigree analysis, but the numbers of transmissions from individuals is usually small and such studies are confounded by the effects of somatic mosaicism. Therefore, one of our goals was to develop a sensitive method to detect small length-change mutants at expanded CAG·CTG repeat loci in individual sperm without the requirement for physical separation of cells. Because of the problem with PCR stutter when the progenitor is large, dilution down to a single g.e. of DNA was essential for the detection of such mutants. The data presented here highlight the problem of PCR stutter and the relative insensitivity of bulk DNA analysis in detecting small length-change mutants from highly expanded repeats and caution against the use of bulk sperm DNA analysis in trying to determine the frequencies or spectra of this class of germline mutations in males.
The mutation frequencies determined by sperm PCR were comparable with those obtained by pedigree analysis. Mutation frequencies involving changes of more than one repeat upon transmission of the DM1 CTG repeat by male mice were 5.7% in the Dmt-E line (two of 35 offspring) and 50% in the Dmt-D line (14 of 28 offspring) (9). Our finding of a frequency of 5.5% in the Dmt-E mice by sperm PCR was very close to the value from pedigree data. In contrast, the mutation frequencies of sperm from the Dmt-D mice at 8 and 17 weeks of age were 14 and 28%, respectively, which were much lower than the values from pedigree analysis. However, the higher frequency measured by pedigree analysis is likely to be a result of the older age of the males (mean age of 20 weeks) at conception of their offspring, at which time they would have accumulated more mutations. Nonetheless, the single g.e. PCR of sperm DNA reflected the much higher frequency of repeat number mutations in the multiple Dmt-D mice than in the Dmt-E mice, as was observed by pedigree analysis. These differences in mutational stability observed between two lines on identical inbred genetic backgrounds indicate that cis-acting factors also play an important role in trinucleotide repeat instability in the male germline with the same line-dependent effects as observed in the soma (10). The cis effects observed in this model system are consistent with the genomic position effects observed from both pedigree and sperm studies in humans (5).
These results indicate that the distribution of DM1 CTG·CAG repeats in the progeny represents alleles present in the sperm of the transmitting father. Similarly, data from DM1 patients revealed that expansion of the DM1 CTG repeat detected in the offspring by pedigree analysis had already accumulated in the paternal germline (7). These results are in contrast to conclusions reported with PCR amplification of bulk DNA from a 117-CAG·CTG repeat from the HD locus in transgenic mice (15); however, we believe the supposedly mutant bands observed in bulk sperm DNA were largely an artifact of PCR stutter. The study of Kovtun et al. (15) also indicated an effect of the gender of the embryo on the spectrum of mutations observed. Our data revealed no such effect with the Dmt lines, suggesting that the Kovtun et al. observation may be peculiar to the R6/1 transgenic line, the mixed genetic background of the strain, or the breeding scheme used.
Single g.e. PCR demonstrated that the significant increase in mutation frequencies in Dmt-D sperm with age was due to a highly significant increase in the number of expansions. A similar trend towards expansion was observed in the pedigree data with the Dmt-D mice, although the effect was not statistically significant, probably due to the relatively small number of offspring analyzed. A significant trend towards increased allele size was observed though when pedigree data from all four Dmt lines of mice were pooled, indicating that this effect is probably a general one. Likewise, three previous transgenic studies have shown an increase in expansions in male transmissions with increasing age (6,11,12). One previous attempt to measure age-dependent changes in repeat number directly in sperm using PCR on bulk DNA failed to detect a change in the distribution or modal size of a large CAG repeat between 7 and 16 weeks of age (14), but as indicated by the present analysis, the sensitivity of that technique was insufficient to detect small changes. In another study (12), PCR of bulk DNA was able to show overall trinucleotide expansions in sperm from 8–10-month-old mice carrying >300 DM1 CTG repeats because the length changes observed were relatively large (mean gain +30 repeats). However, the age dependence was not studied in detail. Using pedigree analysis, three studies have shown an increase in contraction during female transmissions with age (6,13,20), in agreement with the present results, whereas one failed to show an increase (12). These data confirm that as in humans (5), the mutational pathway in the male and female germline is different.
There is also evidence from pedigree analysis that there is an effect of age on length mutations of trinucleotide repeats in humans. The mean sizes of small expansions of a CAG·CTG repeat in DRPLA patients were significantly increased with paternal age (6), and there is an inverse relationship between age at onset of HD and paternal age (21). In an attempt to confirm this by direct analysis of sperm, Leeflang et al. (8) modeled the distribution of repeat numbers in individual sperm of HD patients of different ages, an analysis which suggested a paternal age effect on expansion of the CAG·CTG repeat. However, there was neither a sufficient age range between multiple sperm samples donated from the same donor, nor a simple correlation between age and mutation frequency among the various donors to demonstrate this effect directly.
To our knowledge, the present study is the first to show an age dependence in the accumulation of expansions of trinucleotide repeats, or even any mutations in the male germline, by direct analysis of sperm. These results highlight the utility of a controlled experimental model with a homogeneous genetic background, allowing simple comparisons between animals, and the large sample sizes afforded by single g.e. PCR sperm analysis (>300 DM1 alleles per group of animals), facilitating the detection of even relatively subtle effects.
We observed that the allele distribution in the sperm of the Dmt-D mice shifted towards expansion with increasing repeat size heterogeneity over a period of 9 weeks. There are reports that the mutations at an expanded trinucleotide repeat in an HD transgene take place in the spermatids (14) and that human minisatellite mutations occur at meiosis (22). It is theoretically possible that the increase in repeat length we have observed over time could be explained by age dependence in the error rate during meiotic or spermatid development. However, we consider this is unlikely, particularly since spermatogenesis in the Dmt-D mice seemed unaffected by age. Thus, the majority of DM1 repeat mutations in the Dmt-D mice seem to occur in the renewing spermatogonial stem cell population, likely by a replication-dependent process since the gene is hemizygous. The expansion bias of mutations is most likely fundamental to the mutational pathway, although it could be due to the selective proliferation of spermatogonial stem cells carrying larger alleles by ‘mitotic drive’ (23). However, such an effect would not be expected in our system since the Dmt transgene does not contain any coding sequences and is not associated with any obvious phenotype in hemi- or homozygous Dmt-D mice.
The very high mutation frequency and length changes heavily biased toward expansions in the male germline of the Dmt-D transgene are both features that are observed at expanded repeat loci in humans. However, the mutation frequencies do not appear to be as high, nor are the length changes as large as observed at the expanded DM1 locus in humans (4,7), but are more similar to some of the more stable loci, such as ERDA1, in humans (5). This may simply reflect differences in the activity of cis-acting genomic modifiers. However, in numerous mouse models, the expanded trinucleotide repeat transgenes are more stable relative to their equivalent-sized human counterparts, and none have yet replicated the very large length changes observed at human loci such as the DM1 repeat. This may just reflect the position of the repeat, which in humans may be in a particularly unstable site, but is random in the case of the transgenes, or a fundamental difference in the male germinal pathways for repeat metabolism between men and mice. However, it is possible to speculate that the rate of expansion in spermatogonial stem cells in mice is the same as that in man, and that differences in reproductive age account for the observed length change discrepancy. In the Dmt-D mice the mean repeat length change in the sperm increased from –1.2 repeats to +1.5 repeats over 9 weeks. It is estimated that stem cells divide once during each cycle of the seminiferous epithelium, which lasts 8.6 days in mouse (24); therefore the rate of expansion can be calculated as [1.5 – (–1.2)]/(9 x 7/8.6) = 0.37 repeats per spermatogenic cycle. Spermatogenesis in human males initiates when they reach puberty at 
13 years of age and each seminiferous epithelial cycle lasts 16 days (25). Therefore, assuming that spermatogenic stem cells also divide once per cycle, in men of reproductive ages 20, 30, 40 and 50 years, the cells will have undergone 160, 388, 616 and 844 cycles, respectively. Thus, if the rate of expansion per division cycle were the same, then the mean expansion in these men’s sperm would be 59, 143, 228 and 312 repeats at ages 20, 30, 40 and 50 years, respectively. These estimates fall within the observed ranges for sperm analysis in DM1 patients with similar progenitor allele lengths (4,7) and indicate that small expansions produced by replication slippage could accumulate in the spermatogenic stem cells resulting in the large expansions observed in pedigrees. This provides experimental support for one aspect of the model proposed for age-dependent expansion in the HD repeat, in which both replication slippage contributes to addition of a single repeat at a time and ligation of displaced Okazaki fragments contributes to larger incremental increases (8). Nevertheless, calculations assuming that time per se, rather than cell division, were the critical factor would also be consistent with large expansions accumulating in the spermatogenic stem cells of humans by a cell division-independent mechanism.
Overall, these data reveal that intergenerational differences in repeat length do reflect differences present in the fathers’ sperm. They also further highlight the dependence of the germinal mutational spectrum on both cis-acting modifiers and age. Since we can provide a quantitatively plausible explanation for the absence of the large expansions observed in humans in the male germline of mice, these experiments reinforce the utility of transgenic models in unraveling some aspects of the complex biology of expanded trinucleotide repeats associated with human disease.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Mice
Male mice carrying a single copy of an expanded human Dmt transgene (Dmt-D and Dmt-E lines) were generated on an inbred FVB genetic background and maintained on this background by intercrossing or backcrossing to wild-type FVB mice. Mice were maintained in the facilities of the Department of Veterinary Medicine and Surgery, M. D. Anderson Cancer Center, Houston, TX, and all procedures were approved by the Institutional Animal Care and Use Committee. Only hemizygous males were used for the experiments. Unless otherwise stated, four mice were used for each experimental point. Pedigree analyses in lines Dmt-B, -C, -D and -E were calculated from raw data based on Southern blot analysis of PCR products of tail DNA separated on agarose gels as summarized previously (9).
Tissue DNA preparation
DNA samples from the tail, sperm and somatic tissues were prepared as described previously (9,26). Briefly, tail DNA was prepared from biopsies of 3–4-week mice, and sperm DNA was isolated from both cauda epididymides collected at the time of killing. To minimize the risk of contamination by other DNA sources, the operations were performed in a laminar flow hood, and all solutions and equipment were autoclaved and treated with UV light. One cell-free control was processed in parallel by the same procedure during each DNA extraction to rule out contamination from other DNA sources.
PCR and detection method
Single g.e. PCR and detection were performed as described previously (26). In brief, sperm DNA digested with HindIII (Gibco BRL Life Technologies, Gaithersburg, MD) was diluted to obtain (within a factor of two) a single g.e. amount in each PCR amplification (26). DM-C (5'-AAC GGG GCT CGA AGG GTC CTC-3') and DM-ER (5'-AAA TGG TCT GTG ATC CCC CC-3') primers were used for PCR amplifications. DM-C was fluorescently labeled with FAM-6 (PE Applied Biosystems, Foster City, CA). Amplification conditions employed 96°C for 6 min for one cycle, then 96°C for 45 s, 65°C for 45 s and 70°C for 2 min 45 s for 35 cycles, followed by a chase at 65°C for 1 min and 70°C for 30 min.
The PCR products and an internal standard, GeneScan 1000-ROX, were sized by an ABI 377 sequencer/fragment analyzer (Perkin-Elmer) using GeneScan software for data collection and analysis. For each gel the following were loaded: one negative control (tail DNA from a wild-type mouse) and one positive control (PCR product from 1 ng DNA of a Dmt-E hemizygous embryo) for gel quality, four replicates of PCR products from 1 ng 3-week tail DNA from the mouse whose sperm were studied to mark the progenitor, two replicates of 100 g.e. PCR on sperm DNA to check the quality of the DNA, and 28 replicates of single g.e. PCR on sperm DNA.
A cubic equation, y = a + bx + cx2 + dx3, was fit to the four or five highest peaks to determine the modal repeat size of the products of that PCR amplification, which was not necessarily an integer since the mode shifts continuously with increasing cycle number. The average size of four replicates of 3-week tail DNA was assumed to represent the size of the DM1 progenitor allele for the individual mouse on that particular gel. For analyses of sperm DNA from each mouse, data were pooled from at least three gels (26).
The Poisson distribution was used to calculate the average number of amplifiable DM1 DNA molecules () for each PCR amplification. Gels with > 2 were excluded from the analysis, because stutter bands from additional progenitor molecules obscure mutants with small shifts in repeat number.
Statistics
Data derived from single g.e. PCR were presented as the mean mutation frequency (µ) for each group with standard error of the mean (SEM). The mutation frequency was calculated as m/T, where m is the total number of mutants and T is the total number of amplifiable DNA molecules screened. Differences in the mutation frequencies between different groups were tested in two ways.
Since Dmt mice are on an inbred background, variations between individual mice can reasonably be assumed to be small, so that the data within groups may be pooled for analysis. The error on m was modeled by a Poisson distribution and with the variance V(m) = m (27). Since the single g.e. samples were based on dilution and not sorting individual sperm, T is proportional to the value of , the error on which can be computed by maximum likelihood estimation (27). The variance of µ, V(µ), was then calculated from µ2·[(V()/2) + (V(m)/m2)]. The mutation frequencies between two groups was considered significant (
< 0.05) if 
µa – µb > 1.96 x [V(µa) + V(µb)]1/2.
Alternatively, the mean and SEM of the mutation frequencies for each group were calculated from the values for the four animals in the group. A two-tailed Student’s t-test was used to determine the significance of any differences.
Determination of whether there were biases towards gains or losses was performed using two-tailed 
2 or Fisher’s exact tests. The difference of the overall distribution of sizes of DM1 repeat mutations between groups was tested by the two-sample Kolmogorov-Smirnov test (28).
Analysis of the data from single g.e. PCR on 3-week tail DNA samples was performed using two-tailed 2 or Fisher’s exact tests. Although these tests do not take into account uncertainty of the number of amplifiable DNA molecules, this component of the error is smaller than that from m. SPSS 10.0 statistical software (SPSS Inc., Chicago, IL) was used for all tests.
|
ACKNOWLEDGEMENTS |
|---|
We thank Rashmi Pershad and her staff in DNA Core Laboratory, the University of Texas M. D. Anderson Cancer Center for DNA fragment size analyses, and Mary Coolbaugh-Murphy for helpful discussions and assistance with SP-PCR technique. Walter Pagel’s editorial advice is appreciated. D.G.M. is a Lister Institute Research Fellow. This work was supported by Grant CA-78973 (to M.L.M.), Core Grant CA-16672 and Training Grant CA-77050.
|
FOOTNOTES |
|---|
+ To whom correspondence should be addressed at present address: CIIT Centers for Health Research, 6 Davis Drive, Research Triangle Park, NC 27709, USA. Tel: +1 919 558 1264; Fax: +1 919 558 1300; Email: yzhang@ciit.org
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Lia,A.-S., Seznec,H., Hofmann-Radvanyi,H., Radvanyi,F., Duros,C., Saquet,C., Blanche,M., Junien,C. and Gourdon,G. (1998) Somatic instability of the CTG repeat in mice transgenic for the myotonic dystrophy region is age dependent but not correlated to the relative intertissue transcription levels and proliferative capacities. Hum. Mol. Genet., 7, 1285–1291.
2 Timchenko,L.T. (1999) Myotonic dystrophy: the role of RNA CUG triplet repeats. Am. J. Hum. Genet., 64, 360–364.[Web of Science][Medline]
3
Cummings,C.J. and Zoghbi,H.Y. (2000) Fourteen and counting: unraveling trinucleotide repeat diseases. Hum. Mol. Genet., 9, 909–916.
4 Monckton,D.G., Wong,L.-J.C., 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.
5
Brock,G.J.R., Anderson,N.H. and Monckton,D.G. (1999) Cis-acting modifiers of expanded CAG/CTG triplet repeat expandability: associations with flanking GC content and proximity to CpG islands. Hum. Mol. Genet., 8, 1061–1067.
6
Sato,T., Oyake,M., Nakamura,K., Nakao,K., Fukusima,Y., Onodera,O., Iqarashi,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.
7 Martorell,L., Monckton,D.G., Gamez,J. and Baiget,M. (2000) Complex patterns of male germline instability and somatic mosaicism in myotonic dystrophy type 1. Eur. J. Hum. Genet., 8, 423–430.[Web of Science][Medline]
8
Leeflang,E.P., Tavaré,S., Marjoram,P., Neal,C.O.S., Srinidhi,J. MacFarlane,H., MacDonald,M.E, Gusella,J.F., de Young,M., Wexler,N.S. and Arnheim,N. (1999) Analysis of germline mutation spectra at the Huntington’s disease locus supports a mitotic mutation mechanism. Hum. Mol. Genet., 8, 173–183.
9 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. Nat. Genet., 15, 193–196.[Web of Science][Medline]
10
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.
11 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. Nat. Genet., 15, 197–200.[Web of Science][Medline]
12
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.
13
La Spada,A.R., Peterson,K.R., Meadows,S.A., McClain,M.E., Jeng,G., Chmelar,R.S., Haugen,H.A., Chen,K., Singer,M.J., Moore,D. et al. (1998) Androgen receptor YAC transgenic mice carrying CAG 45 alleles show trinucleotide repeat instability. Hum. Mol. Genet., 7, 959–967.
14 Kovtun,I.V. and McMurray,C.T. (2001) Trinucleotide expansion in haploid germ cells by gap repair. Nat. Genet., 27, 407–411.[Web of Science][Medline]
15
Kovtun,I.V., Therneau,T.M. and McMurray,C.T. (2000) Gender of the embryo contributes to CAG instability in transgenic mice containing a Huntington’s disease gene. Hum. Mol. Genet., 9, 2767–2775.
16 Zhang,L., Leeflang,E.P. and Arnheim,N. (1994) Studying human mutations by sperm typing: Instability of CAG trinucleotide repeats in the human androgen receptor gene. Nat. Genet., 7, 531–535.[Web of Science][Medline]
17
Leeflang,E.P., Zhang,L., Tavare,S., Hubert,R., Srinidhi,J., MacDonald,M.E., Myers,R.H., de Young,M., Wexler,N.S., Gusella,J.F. et al. (1995) Single sperm analysis of the trinucleotide repeats in the Huntington’s disease gene: quantification of the mutation frequency spectrum. Hum. Mol. Genet., 4, 1519–1526.
18
Wheeler,V.C., Auerbach,W., White,J.K., Srinidhi,J., Auerbach,A., Ryan,A., Duyao,M.P., Vrbanac,V., Weaver,M., Gusella,J.F. et al. (1999) Length-dependent gametic CAG repeat instability in the Huntington’s disease knock-out mouse. Hum. Mol. Genet., 8, 115–122.
19 Leeflang,E.P., Tavare,S., Marjoram,P., Grewal,R., Neal,C.O.S. and Arnheim,N. (1998) Human germline mutation analysis by single genome PCR: application to dynamic mutations. In Wells,R.D., Warren,S.T. and Sarmiento,M. (eds), Genetic Instabilities and Hereditary Neurological Diseases. Academic Press, San Diego, CA, pp. 543–558.
20
Kaytor,M.D., Burright,E.N., Duvick,L.A., Zoghbi,H.Y. and Orr,H.T. (1997) Increased trinucleotide repeat instability with advanced maternal age. Hum. Mol. Genet., 6, 2135–2139.
21 Farrer,L.A., Cupples,L.A., Kiely,D.K., Conneally,P.M. and Myers,R.H. (1992) Inverse relationship between age at onset of Huntington disease and paternal age suggests involvement of genetic imprinting. Am. J. Hum. Genet., 50, 528–535.[Web of Science][Medline]
22 Jeffreys,A.J., Neil,D.L. and Neumann,R. (1998) Repeat instability at human minisatellites arising from meiotic recombination. EMBO J., 17, 4147–4157.[Web of Science][Medline]
23
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.
24 Clermont,Y. and Trott,M. (1969) Duration of the cycle of the seminiferous epithelium in the mouse and hamster determined by means of 3H-thymidine. Fertil. Steril., 20, 805–817.[Web of Science][Medline]
25 Heller,C.G. and Clermont,Y. (1964) Kinetics of the germinal epithelium in man. Recent Prog. Horm. Res., 20, 545–575.
26 Zhang,Y., Monckton,D.G., Siciliano,M.J., Connor,T.H. and Meistrich,M.L. (2002) Detection of radiation and cyclophosphamide-induced mutations in individual mouse sperm at a human expanded trinucleotide repeat locus transgene. Mutat. Res., in press.
27 Zheng,N., Monckton,D.G., Wilson,G., Hagemeister,F., Chakraborty,R., Connor,T.H., Siciliano,M.J. and Meistrich,M.L. (2000) Frequency of minisatellite repeat number changes at the MS205 locus in human sperm before and after cancer chemotherapy. Environ. Mol. Mutagen., 36, 134–145.[Web of Science][Medline]
28 Fisher,L.D. and van Belle,G. (1993) Biostatistics: A Methodology for the Health Sciences. Wiley, New York, pp. 304–344.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Jung and N. Bonini CREB-Binding Protein Modulates Repeat Instability in a Drosophila Model for PolyQ Disease Science, March 30, 2007; 315(5820): 1857 - 1859. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Lloret, E. Dragileva, A. Teed, J. Espinola, E. Fossale, T. Gillis, E. Lopez, R. H. Myers, M. E. MacDonald, and V. C. Wheeler Genetic background modifies nuclear mutant huntingtin accumulation and HD CAG repeat instability in Huntington's disease knock-in mice Hum. Mol. Genet., June 15, 2006; 15(12): 2015 - 2024. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. V. Kovtun, G. Welch, H. D. Guthrie, K. L. Hafner, and C. T. McMurray CAG Repeat Lengths in X- and Y-bearing Sperm Indicate That Gender Bias during Transmission of Huntington's Disease Gene Is Determined in the Embryo J. Biol. Chem., March 5, 2004; 279(10): 9389 - 9391. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Savouret, C. Garcia-Cordier, J. Megret, H. te Riele, C. Junien, and G. Gourdon MSH2-Dependent Germinal CTG Repeat Expansions Are Produced Continuously in Spermatogonia from DM1 Transgenic Mice Mol. Cell. Biol., January 15, 2004; 24(2): 629 - 637. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||