results were examined. Long repeats (full mutation length alleles), which typically have clear clinical phenotypes, were not included in this study. DNA obtained from both previously tested clinical specimens as well as cell lines obtained from the Coriell Cell Repository were analyzed after PCR amplification. These samples were used as in-house controls by two clinical laboratories. Our initial measurements focused on measurement variability: (1) between slab-PAGE and capillary (CE) separation systems, (2) interlane variability, (3) intergel variability (slab-PAGE), and (4) variability during PCR amplification. We performed statistical analyses on system reproducibility and interlane and intergel variability. Samples were run in triplicate for all measurements and the analysis was performed using GeneScan(TM) analysis software. DNA sequencing was performed to verify repeat lengths.
2.1.1. Slab-PAGE Analysis As expected, the shorter alle-les were more easily amplified and sequenced than longer alleles. The standard deviations for interlane measurements in slab-gels ranged from 0.05 to 0.35. The variation in size measurements performed on different gels and PCR amplifications ranged from 0.06 to 0.30 (7). This suggests that these measurements varied by up to a single nucleotide (0.33 of a three-nucleotide repeat).
2.1.2. CE Analysis The CGG repeat measurements performed by capillary electrophoresis were slightly more precise, with standard deviations ranging from 0.02 to 0.29. However, allele sizes observed after CE separations were significantly smaller than those obtained after slab-gel elec-trophoresis (Table 1, samples 6910 and 6968). DNA sequence analysis confirmed that the size measurements were correct for the slab-gel data and inaccurate for the CE results. It was hypothesized that the proprietary gel matrix used for capillary electrophoresis (POP-4(TM)) leads to anomalous rapid elec-trophoretic mobility of CG-rich sequences (10).
2.1.3. Allele Size Analysis The peak ratio of each allele was compared within each female sample (Table 1). The detection method with peak ratios closest to 1.00 is more accurate but not necessarily more precise. The results indicate a gradual decrease in peak ratio with increase of allele size. As shown, premutation alleles contain the highest peak ratio discrepancy (comparison of samples 6910 and 6968 with 7541 and 13664). This bias would be even more pronounced in the detection of larger, premutation and full mutation length alleles (7).
Table 1 also shows that CE analysis has greater error detecting the presence of long alleles than slab-PAGE, as shown by the difference between peak ratios (i.e., shaded areas samples 6968; 0.59 for slab-PAGE vs 0.25 for CE). This suggests that the electrokinetic injection used in CE results in a bias toward capillary loading of smaller alleles obtained from female specimens. Therefore, both amplification and amplicon loading appear to contribute to peak ratio discrepancy.
2.2. SUMMARY OF ACCURATE SIZING METHODS FOR FRAGILE X SYNDROME The accuracy of our sizing data for Fragile X measurements within the normal, gray zone, and pre-mutation allele sizes was within one repeat length for slab-PAGE measurements. The precision was equally high in lane-to-lane comparisons, comparisons of PCR results between gels, and in multiple PCR amplifications. There is no statistically significant evidence for heterogeneity of size determination after separation by either slab-PAGE or CE measurements. Hence, the gel matrix and running conditions for slab-PAGE were suitable for accurate size determinations (as confirmed by DNA sequencing) despite known migration anomalies (10). A GC-rich sizing standard would improve the accuracy of sizing by capillary electrophoresis, as this separation method results in high-precision measurements but incorrect size determinations.
Our data reveal several important considerations in the performance of Fragile X testing by PCR. First, size measurements were not directly comparable between the two separation systems for the larger, premutation, and presumably full mutation length alleles. Second, the POP4 polymer used in this study resulted in premutation size measurements that varied from actual size (as measured by DNA sequencing) by four to eight repeat elements. The greatest error (three repeats) detected in slab-PAGE measurements was found in the sample with the longest allele size—112 repeats as determined by DNA sequencing. All other normal, gray zone, and premutation measurements agreed with DNA sequencing. In addition, elec-trokinetic injection by CE resulted in allele-biased loading of premutation alleles. Because PCR methods are developed that robustly amplify full mutation alleles, the bias for loading smaller alleles could impact the ability to detect these alleles. High-precision measurements for samples containing long pre-mutation and full mutation length alleles are currently under validation. Sizing standards that effectively allow cross-platform and interlaboratory comparisons are under development as a NIST SRM.
Was this article helpful?
Are You Prepared For Your First Baby? Endlessly Searching For Advice and Tips On What To Expect? Then You've Landed At The Right Place With All The Answers! Are you expecting? Is the time getting closer to giving birth to your first baby? So many mothers to be are completely unprepared for motherhood and the arrival of a little one, but stress not, we have all the answers you need!