Diagnostics of short tandem repeat expansion variants using massively parallel sequencing and componential tools

Abstract

Short tandem repeats (STRs) are scattered throughout the human genome. Some STRs, like trinucleotide repeat expansion (TRE) variants, cause hereditable disorders. Unambiguous molecular diagnostics of TRE disorders is hampered by current technical limitations imposed by traditional PCR and DNA sequencing methods. Here we report a novel pipeline for TRE variant diagnosis employing the massively parallel sequencing (MPS) combined with an opensource software package (FDSTools), which together are designed to distinguish true STR sequences from STR sequencing artifacts. We show that this approach can improve TRE diagnosis, such as Oculopharyngeal muscular dystrophy (OPMD). OPMD is caused by a trinucleotide expansion in the PABPN1 gene. A short GCN expansion, (GCN[10]), coding for a 10 alanine repeat is not pathogenic, but an alanine expansion is pathogenic. Applying this novel procedure in a Dutch OPMD patient cohort, we found expansion variants from GCN[11] to GCN[16], with the GCN[16] as the most abundant variant. The repeat expansion length did not correlate with clinical features. However, symptom severity was found to correlate with age and with the initial affected muscles, suggesting that aging and muscle-specific factors can play a role in modulating OPMD.

Fig. 1

A flowchart of the DNAseq procedure

Fig. 2

Examples of bioanalyses and FDSTools sequence results in OPMD and controls. Each row shows the result from a single subject. The left panel shows DNA histograms, which were generated by the DNA Bioanalyzer application. The right panel shows the graphical output of FDStools sequence analysis for an anonymous control and OPMD patients with five different GCN expansion lengths. The genuine alleles are highlighted in bold. The wild-type allele is depicted in black and the expanded allele in red. Non-genuine reads (PCR stutters, PCR-generated point mutations and sequencing errors) are shown in gray

Fig. 3

Stutters in trinucleotide GCN expansion. a Analysis of the wild-type allele (GCN[10]) in control and OPMD patients. The number of reads (left panel) or the percentage of reads with respect to the highest allele in the sample (right panel) are plotted against GCN length results. In the control group, results are from one PCR experiment. In the OPMD group, results are from two experiments: 20 amplification cycles (PCR1) or 40 amplification cycles (PCR2). The stutter percentage was significantly higher in the second experiment. Genuine allele length is depicted in bold and stutter allele length is in plain text. P-value was calculated by the Student’s T-test. b Analysis of sequencing results in OPMD subjects from two experiments. Plots show percentage of reads with respect to the highest allele in the sample for subjects with the same genotype (heterozygous: GCN[12], GCN[13], GCN[14], GCN[15], GCN[16], and homozygous GCN[11]). In each plot, the left side shows the wild-type GCN[10] and corresponding stutter GCN[9], the right side shows the expanded allele and its corresponding stutters. Genuine GCN length is depicted in bold and stutters are plain. Each dot represents an individual. Median is marked with a black line

Fig. 4

The frequency of GCN expansion length in the Dutch OPMD patients. Bar chart shows the frequency of GCN expansion variants in familial and sporadic OPMD. The GCN[16] expansion is enriched in the Dutch familial OPMD. The GCN[15] expansion is found only in sporadic OPMD, the GCN[11] expansion is only familial