Next generation sequencing in research and diagnostics of ocular birth defects

Abstract

Sequence capture enrichment (SCE) strategies and massively parallel next generation sequencing (NGS) are expected to increase the rate of gene discovery for genetically heterogeneous hereditary diseases, but at present, there are very few examples of successful application of these technologic advances in translational research and clinical testing. Our study assessed whether array based target enrichment followed by re-sequencing on the Roche Genome Sequencer FLX (GS FLX) system could be used for novel mutation identification in more than 1000 exons representing 100 candidate genes for ocular birth defects, and as a control, whether these methods could detect two known mutations in the PAX2 gene. We assayed two samples with heterozygous sequence changes in PAX2 that were previously identified by conventional Sanger sequencing. These changes were a c.527G>C (S176T) substitution and a single basepair deletion c.77delG. The nucleotide substitution c.527G>C was easily identified by NGS. A deletion of one base in a long polyG stretch (c.77delG) was not registered initially by the GS Reference Mapper, but was detected in repeated analysis using two different software packages. Different approaches were evaluated for distinguishing false positives (sequencing errors) and benign polymorphisms from potentially pathogenic sequence changes that require further follow-up. Although improvements will be necessary in accuracy, speed, ease of data analysis and cost, our study confirms that NGS can be used in research and diagnostic settings to screen for mutations in hundreds of loci in genetically heterogeneous human diseases.

Figure 1

Figure 1A and 1B. Average numbers of reads per base (depth of coverage) for each targeted region in Sample 1 (Fig 1A) and Sample 2 (Fig 1B). Although significant variation is noticeable between regions, sufficient coverage for reliable SNP detection was obtained for the majority of targeted exons.

Figure 1

Figure 1A and 1B. Average numbers of reads per base (depth of coverage) for each targeted region in Sample 1 (Fig 1A) and Sample 2 (Fig 1B). Although significant variation is noticeable between regions, sufficient coverage for reliable SNP detection was obtained for the majority of targeted exons.

Figure 2

Figure 2A. Sanger sequencing chromatograms showing the c.527G>C, S176T mutation in two different DNA strands. The mutated base is marked by the red arrow. Figure 2B. A screenshot from the CLCbio analysis software showing the misense mutation c.527G>C, S176T. Reads from different directions are shown in different colors (red and blue). The mutated base is shown between the two vertical red lines.

Figure 2

Figure 2A. Sanger sequencing chromatograms showing the c.527G>C, S176T mutation in two different DNA strands. The mutated base is marked by the red arrow. Figure 2B. A screenshot from the CLCbio analysis software showing the misense mutation c.527G>C, S176T. Reads from different directions are shown in different colors (red and blue). The mutated base is shown between the two vertical red lines.

Figure 3

A screenshot from the CLCbio analysis software showing the frameshift mutation c.del77G. The mutated base is shown between the two vertical grey lines. The reads showing the mutant allele (6 Gs) are grouped at the top, while the reads generated from the wild-type allele (7Gs) are shown at the bottom.

Figure 3

A screenshot from the CLCbio analysis software showing the frameshift mutation c.del77G. The mutated base is shown between the two vertical grey lines. The reads showing the mutant allele (6 Gs) are grouped at the top, while the reads generated from the wild-type allele (7Gs) are shown at the bottom.