Genetic diagnosis of facioscapulohumeral muscular dystrophy type 1 using rare-variant linkage analysis and long-read genome sequencing

Kun Li, Daniel Quiat, Fei She, Yuanwei Liu, Rong He, Alireza Haghighi, Fang Liu, Rui Zhang, Steven R. DePalma, Ying Yang, Wen Wang, Christine E. Seidman, Ping Zhang, Jonathan G. Seidman

TL;DR

This paper presents a novel, combined approach using rare-variant linkage analysis and ultra-long-read genome sequencing to diagnose Facioscapulohumeral Muscular Dystrophy type 1 (FSHD1). This method successfully maps the disease locus and precisely genotypes the pathogenic allele, offering a more efficient and comprehensive diagnostic tool than current standards.

What Problem Does This Paper Address?

Facioscapulohumeral Muscular Dystrophy type 1 (FSHD1) is a progressive skeletal muscle disorder with variable penetrance. The underlying genetic cause is an epigenetic de-repression of the DUX4 gene, which is located within a complex, repetitive region (D4Z4 locus) on chromosome 4q. Diagnosing FSHD1 has traditionally relied on Southern blotting, a method that is labor-intensive, requires significant DNA input, and can yield inconclusive results in a substantial percentage of cases. More recent techniques like optical genome mapping and molecular combing have improved D4Z4 repeat length assessment, but direct sequencing of the pathogenic allele remains challenging due to the repetitive and homologous nature of the region. This paper aims to overcome these diagnostic hurdles by leveraging advanced sequencing technologies and analytical methods.

Methods

The researchers employed a two-pronged genomic approach:

  1. Short-Read Genome Sequencing and Rare-Variant Linkage Analysis: They first performed high-throughput, short-read genome sequencing on affected and unaffected individuals from a large kindred. This data was then used for parametric linkage analysis. By filtering for rare heterozygous biallelic single nucleotide variants (SNVs) with low population frequency, they identified markers that could be used to map the disease locus. This analysis was performed assuming an autosomal dominant inheritance model with a 90% penetrance.
  2. Ultra-Long-Read Genome Sequencing (Oxford Nanopore): To directly genotype the pathogenic FSHD1 allele, they utilized ultra-long-read sequencing technology (Oxford Nanopore). This method is capable of generating reads that span kilobases, allowing for the sequencing of entire repetitive regions. The long reads were aligned to a highly contiguous reference genome (chm13) to accurately assess the D4Z4 repeat array length and identify the presence of the permissive 4qA haplotype, which is crucial for FSHD1 pathogenesis.

Datasets

  • Human Genome Reference: The study utilized the hg38 reference genome for initial alignment and the chm13 reference genome for ultra-long-read analysis, noting its improved representation of repetitive regions.
  • 1000 Genomes Project Phase 3: Used for phasing SNVs.
  • gnomAD (v3): Used to filter for rare heterozygous biallelic SNVs with a population allele frequency less than 1x10⁻⁴ in East Asian subjects.
  • Genomic DNA: Extracted from peripheral blood samples of 8 affected and 2 unaffected individuals from a 4-generation kindred.

Key Insights and Ideas

  • Linkage Analysis with Rare Variants is Powerful: The study demonstrates that genome-wide rare variant-based linkage analysis can effectively map disease loci in multiplex kindreds, even when traditional diagnostic methods are challenging.
  • Ultra-Long Reads Revolutionize Repetitive Region Analysis: Ultra-long-read sequencing, specifically Oxford Nanopore, is shown to be a powerful tool for directly genotyping complex repeat alleles like those found in the D4Z4 locus, which are problematic for short-read technologies.
  • A Single Assay Potential: The combined approach suggests the possibility of a single molecular test that can assess D4Z4 repeat number, 4qA/B haplotype status, and potentially variants in other related genes (like SMCHD1 for FSHD2), offering a more comprehensive diagnostic solution.
  • Improved Reference Genomes are Crucial: The use of the chm13 reference genome, which has a more complete assembly of repetitive regions, was critical for accurate alignment and analysis of the D4Z4 locus.

Key Results

  • Disease Locus Mapping: Parametric linkage analysis identified a single 1.7 Mb interval on chromosome 4q35.2 with a logarithm of the odds (LOD) score of 3.2, strongly suggesting linkage to the disease.
  • Pathogenic Allele Genotyping: Ultra-long-read sequencing successfully identified reads containing 5 KpnI repeat units and the 4qA permissive haplotype, consistent with a pathogenic FSHD1 allele in an affected individual.
  • Non-pathogenic Allele Characterization: A read with 11 KpnI repeats and lacking the 4qA polyadenylation sequence was identified, potentially representing the non-pathogenic allele, though mosaicism or misalignment were also considered.
  • Efficiency: Despite the complexity of the region, the study obtained 3 high-confidence reads from a pathogenic allele among over 750,000 total reads.

Limitations and Future Work

  • Limited Genotyping of Non-pathogenic Allele: The study was unable to obtain high-confidence reads fully spanning the non-pathogenic allele.
  • Potential for Recombination Events: The authors acknowledge the inability to exclude additional recombination events at the D4Z4 locus in other affected family members.
  • Southern Blotting Limitations: Clinical genotyping in affected individuals by Southern blotting was unavailable, limiting the completeness of the molecular analysis.
  • Future Directions: The authors suggest increasing sequencing depth, using adaptive sampling to selectively sequence target regions, and employing target capture before sequencing to improve the yield of ultra-long reads. They also anticipate that continued advances in long-read sequencing technologies, reduced costs, and improved bioinformatic tools will enhance the ability to evaluate 4q-D4Z4 genotypes using genome-wide approaches. They also highlight the potential for this approach to evaluate loci for other repeat-mediated myopathies.

Summary updated

Abstract

Facioscapulohumeral dystrophy type 1 (FSHD1) is a progressive, debilitating skeletal myopathy that requires a multimodal approach for complete molecular characterization of pathogenic genotypes. Here, we report genomic analyses of a family with suspected FSHD1. We first performed short-read genome sequencing, followed by parametric linkage analysis using rare variants to map the disease locus to a single 1.7 Mb interval on chromosome 4q35.2 with a logarithm of the odds score of 3.2. We then used ultra-long-read genome sequencing as a single molecular test to genotype a pathogenic FSHD allele containing a 4qA permissive haplotype and 5 KpnI repeat units at the D4Z4 locus. These results demonstrate that genome-wide rare variant-based linkage analysis is a powerful tool for mapping disease loci in families, and ultra-long-read genome sequencing is capable of genotyping pathogenic FSHD1 alleles.
BibTeX
@inproceedings{li2024genetic,
  author = {Kun Li and Daniel Quiat and Fei She and Yuanwei Liu and Rong He and Alireza Haghighi and Fang Liu and Rui Zhang and Steven R. DePalma and Ying Yang and Wen Wang and Christine E. Seidman and Ping Zhang and Jonathan G. Seidman},
  title = {{Genetic diagnosis of facioscapulohumeral muscular dystrophy type 1 using rare-variant linkage analysis and long-read genome sequencing}},
  booktitle = {Genetics},
  year = {2024},
  doi = {10.1016/j.gimo.2024.101817},
}
Figure