Investigating Splice Defects in USH2A Using Targeted Long-Read Sequencing

Abstract

Biallelic variants in USH2A are associated with retinitis pigmentosa (RP) and Type 2 Usher Syndrome (USH2), leading to impaired vision and, additionally, hearing loss in the latter. Although the introduction of next-generation sequencing into clinical diagnostics has led to a significant uplift in molecular diagnostic rates, many patients remain molecularly unsolved. It is thought that non-coding variants or variants of uncertain significance contribute significantly to this diagnostic gap. This study aims to demonstrate the clinical utility of the reverse transcription-polymerase chain reaction (RT-PCR)-Oxford Nanopore Technology (ONT) sequencing of USH2A mRNA transcripts from nasal epithelial cells to determine the splice-altering effect of candidate variants. Five affected individuals with USH2 or non-syndromic RP who had undergone whole genome sequencing were recruited for further investigation. All individuals had uncertain genotypes in USH2A, including deep intronic rare variants, c.8682-654C>G, c.9055+389G>A, and c.9959-2971C>T; a synonymous variant of uncertain significance, c.2139C>T; p.(Gly713=); and a predicted loss of function duplication spanning an intron/exon boundary, c.3812-3_3837dup p.(Met1280Ter). In silico assessment using SpliceAI provided splice-altering predictions for all candidate variants which were investigated using ONT sequencing. All predictions were found to be accurate; however, in the case of c.3812-3_3837dup, the outcome was a complex cryptic splicing pattern with predominant in-frame exon 18 skipping and a low level of exon 18 inclusion leading to the predicted stop gain. This study detected and functionally characterised simple and complex mis-splicing patterns in USH2A arising from previously unknown deep intronic variants and previously reported variants of uncertain significance, confirming the pathogenicity of the variants.

Keywords: Oxford nanopore sequencing; USH2A; cryptic splice-altering variant; deep intronic variant; inherited retinal disease; long-read sequencing; nasal epithelial cells.

Figure 1

(A) Sashimi plot from IGV showing the splicing junctions of Patient 1 (top, red) and a control (bottom, blue). This fragment includes exons 48–52 of USH2A from right to left (reverse strand, 3′ to 5′). Each bar represents an exon, and the height of the bars represents the read depth. The control (bottom, blue) shows the expected canonical splicing. Patient 1 (top, red) shows mis-splicing, resulting in the retention of a pseudoexon comprising 113 bp derived from intron 50. The reference USH2A transcript (NM_206933.4) is represented at the bottom of the track. (B) Enlarged view of sashimi plot showing exons 50–51 from (A). (C) Graphical representation of the mis-splicing in Patient 1 relative to the variant location and showing the mutant donor site motif.

Figure 2

(A) Sashimi plot from IGV showing splice junctions for USH2A exons 12 and 13 of Patient 2 (top, red) and a control (bottom, blue). Exons are depicted from right to left (reverse strand, 3′ to 5′) with each bar depicting an exon and the height of the bars indicating the number of reads. The control (bottom, blue) exhibits the expected normal splicing. In contrast, Patient 2 (top, red) shows mis-splicing (partial exon 12 skipping) due to the variant. The reference USH2A transcript (NM_206933.4) is represented by the blue line at the bottom. (B) Graphical representation of the mis-splicing pattern in Patient 2 relative to the variant location and the mutant sequence.

Figure 3

(A) Sashimi plot showing the splice junctions of Patient 3 (red, top) and a control (bottom, blue). USH2A exons 17–20 are depicted from right to left (reverse strand, 3′ to 5′). The patient sample (top, red) shows multiple mis-splicing events—the inclusion of the duplicated nucleotides (mapping as partial intron 17 inclusion) and exon 18 skipping with or without partial exon 19 skipping. The control sample (bottom, blue) shows canonical splicing. (B) Enlarged view of exons 17–19 from (A). (C) Graphical representation of the multiple mis-splicing events identified in Patient 3.

Figure 4

(A) Sashimi plot showing the splice junctions of Patient 4 (red, top) and a control (bottom, blue). USH2A exons 43–44 are depicted from right to left (reverse strand, 3′ to 5′), with each bar representing an exon and the height of the bars indicating the number of reads. The control (bottom, blue) shows canonical splicing only. The patient (top, red) shows mis-splicing—a pseudoexon inclusion of 130 bp. The reference USH2A transcript (NM_206933.4) is represented by the blue line at the bottom. (B) Graphical representation of the mis-splicing patterns in Patient 4, showing the mutant donor site motif.

Figure 5

(A) Sashimi plot showing the splice junctions of Patient 5 (red, top) and a control (bottom, blue). USH2A exons 45–47 are depicted from right to left (reverse strand, 3′ to 5′) with each bar representing an exon and the height of the bars indicating the number of reads. The control sample (bottom, blue) shows only canonical splicing. The patient (top, red) shows mis-splicing—a pseudoexon inclusion of 78 bp in intron 45. The blue line at the bottom represents the reference USH2A transcript (NM_206933.4). The overall read depth was low due to short flow cell run time and low template in the library preparation. (B) Graphical representation of the mis-splicing pattern in Patient 5 showing the mutant donor site motif.