# Rethinking genomics of facioscapulohumeral muscular dystrophy in the telomere-to-telomere era: pitfalls in the hidden landscape of D4Z4 repeats

## TL;DR

This study re-examines the genomic landscape of FSHD, revealing that D4Z4 repeats are more widespread than previously thought, challenging current diagnostic and mechanistic assumptions.

## Contribution

The study identifies D4Z4-like repeats on multiple chromosomes beyond 4q35 and 10q26 using the T2T-CHM13 genome assembly.

## Key findings

- D4Z4-like repeats are found on at least ten additional chromosomes, some with intact DUX4 open reading frames.
- Commonly used PCR primers for DUX4 or DBE-T detection amplify multiple loci beyond 4q/10q.
- The findings suggest that many signals from the 4q locus may originate from paralogous arrays.

## Abstract

Facioscapulohumeral muscular dystrophy (FSHD) is genetically associated with reduction of the D4Z4 macrosatellite array at 4q35 on a permissive 4qA haplotype, a configuration that enables the stable expression of the DUX4 transcription factor. Current diagnostic and mechanistic models, however, rely heavily on the incomplete GRCh38/hg38 reference and assume that D4Z4 repeats are predominantly confined to 4q35 and 10q26 loci. Here we present a systematic re-analysis of the configuration of D4Z4-like repeats in the human genome using the Telomere-to-Telomere human genome assembly (T2T-CHM13 v2.0/hs1) and complementary experimental validation. Using the terminal 4q35 repeat as a query, we annotated the full repertoire of D4Z4-like loci across the genome and characterized their structural completeness, flanking sequences, and coding potential. This survey uncovered clusters and isolated monomers on at least ten additional chromosomes, several of which harbor intact DUX4 open reading frames or polyadenylation signals. In silico PCR and assays on monochromosomal hybrid cell lines demonstrate that primer sets widely employed for DUX4 or DBE-T detection amplify multiple loci beyond 4q/10q. Together, these findings demonstrate that many signals historically attributed to the pathogenic 4q locus may in fact arise from paralogous arrays. Our study establishes the necessity of locus-resolved, repeat-aware approaches, combining long-read sequencing, methylation-aware profiling, and isoform-resolved transcriptomics, for accurate diagnostics and to define the molecular basis of FSHD.