Clinical and molecular aspects of the association of skewed X chromosome inactivation and recurrent spontaneous abortion

• Main Author: Beever, Christy L.
• Format: Others
• Language: English
• Published: 2009
• Online Access: http://hdl.handle.net/2429/12063

Recurrent spontaneous abortion (RSA), defined as three or more pregnancy losses at less than 20 weeks of gestation, is a devastating problem that affects 1-2% of couples trying to have children. Despite extensive medical testing and experimental treatments couples with RSA may go through, a clear cause for their pregnancy losses is often not found. The finding that skewed X chromosome inactivation (XCI), the non-random inactivation of one of two X chromosomes in female cells, is increased in women with RSA has led to the exciting prospect that a cause for these pregnancy losses can be identified in a subset of women with RSA. The objective of this thesis was to explore the etiology of the association of RSA and skewed XCI and the potential use of skewed XCI in counseling of couples with RSA. The reliability of the assay for skewed XCI was also examined by comparison of different methodologies. Skewed XCI is currently used in both clinical and research settings as a biological marker for clonality and carrier status of certain X-linked diseases. DNA methylation-based XCI assays generally involve amplification by PCR of an X-linked locus which is both differentially methylated, allowing distinction of the inactive and active X via digestion with a methylation sensitive enzyme, and polymorphic, allowing discrimination of alleles of the locus. To test the reproducibility of such assays, the following aspects of the XCI assay were altered: 1) method of quantification (densitometry of silver-stained polyacrylamide gels vs. automated fluorescent analysis using an ABI 310 Genetic Analyzer); 2) the use of a secondary cutter (a restriction enzyme which is not methylation sensitive) in conjunction with the methylation sensitive enzyme; 3) locus (androgen receptor (AR) vs. FMR-1). In 60 samples tested by both quantification methods, XCI results were well correlated (r=0.839) with a mean difference in percent skewing between replicates of 6.4%±0.6% with a range of 0.0 to 21.2. By analyzing replicate measurements of XCI skewing in which the same quantification method was used, the ABI 310 was observed to be more reproducible than with densitometry. The mean absolute difference in XCI skewing between 77 samples assayed with and without a secondary cutter was 6.4%±0.7% with a range of 0 to 35 and a correlation of 0.795. Within sample reproducibility was higher with the secondary cutter than without the secondary cutter. A relatively, poor correlation (r=0.673). was observed between 37 samples analyzed at both the AR and FMR-1 loci. The mean difference was 7.8%±1.6% with a range of 0.5 to 44.9. Further analyses of 3 discrepant samples with a third locus, DXS6673E, were more consistent with the AR results, suggesting additional influences upon FMR-1 methylation may cause assays using the FMR-1 locus to be less reliable. In conclusion, we recommend using the AR locus with a secondary cutter and quantifying alleles with the ABI 310. With this methodology we observed a mean difference in 45 replicated samples of only 2.9%±0.7%. Although this methodology is the most reproducible, it is unknown if it is the best reflection of XCI status. This study showed DNA methylation XCI assays can be prone to bias and caution must be taken when interpreting results from this assay. Skewed XCI (defined as >90% inactivation of one allele) was estimated with the DNA methylation sensitive assay and was observed in 12% of women with RSA (n=207) compared to 9% of controls (n=100), demonstrating an increase, although not statistically significant, of skewed XCI in women with RSA. Furthermore, 54% of karyotyped losses (n=26) from women with skewed XCI were trisomic compared to 31% (n=145) from women without skewed XCI (p=0.02, Fisher's exact test), suggesting skewed XCI in women with RSA reflects a predisposition to trisomic losses. These trisomic losses from women with RSA and skewed XCI were found to be associated with advanced maternal age. An increase of male losses or an increase of female livebirths was not observed in women with skewed XCI, as would be expected if male lethal X-linked mutations were the cause of the association of skewed XCI and RSA. Skewed XCI was not more prevalent in women with RSA who tested positive for an autoimmune, structural chromosomal, endocrine, anatomical, or infectious factor; nor was skewed XCI more prevalent in women with unexplained RSA. Telomere length was assessed in women with skewed XCI and RSA to investigate a possible association of skewed XCI and chromosome structure; no association between skewed XCI and telomere length was found. Although skewed XCI has been linked to RSA in several studies, the cause of this association and the extent to which skewed XCI reflects an etiology for losses in couples with RSA remains unclear. For this reason, the value of determining skewed XCI in women with RSA is extremely limited for the purposes of counseling. However, this study provides evidence that factors associated with trisomy rather than male lethal X-linked mutations, are the main cause of the association of RSA and skewed XCI. Further research of skewed XCI and maternal age-related mechanisms of trisomy may provide further insight into the cause of the association of skewed XCI and RSA. === Medicine, Faculty of === Medical Genetics, Department of === Graduate