Phenotypical differences in the two families

In Family A (Fig. 1A), no history of neuromuscular disease was found. The couple (I-1 and I-2) had two sons (II-1 and II-2), aged 11 and 8 years, respectively. Both boys were normal in terms of movement, walking, and cognition, and no sign of Gower’s or bilateral pseudohypertrophy of the lower legs was observed. The woman (I-2) was pregnant at a gestational age of 24 weeks when she was referred to the clinic and underwent amniocentesis for routine prenatal genetic diagnosis because of advanced age (36 years). Based on comprehensive analysis and interpretation of the genetic findings, the woman decided to continue the pregnancy and delivered a healthy girl (proband II-3).

In Family B (Fig. 2A), the proband (II-2) was an 8-year-old boy with growing proximal limb weakness since the age of 3. The proband exhibited typical phenotype of DMD with signs of Gower’s and bilateral pseudohypertrophy of the calf muscles. Muscle strength examination revealed that his bilateral lower extremities had diminished muscular weakness (trunk muscles G2, hip flexors G3, hip extensors G3−, knee extensors, and knee flexors G3−). Biochemical tests revealed increased levels of lactate dehydrogenase (LDH), creatine kinase (CK), and creatine enzyme isoenzymes (CK-MB) (LDH, 910 U/L; CK, 8941 U/L; and CK-MB, 264 U/L). No abnormalities were noted on electrocardiography or cardiac echocardiography, and his intelligence was within the normal range. The patient had no family history of muscular dystrophy.

Copy-number gains involving DMD identified using CNV-seq, qPCR and/or MLPA

In Family A, a multi-copy duplication with a segmental size of 0.36 Mb involving exons 64–79 of DMD was identified using CNV-seq. The proband (female, II-3) and her mother (I-2) showed a copy number of seven for the segmental repeat (inferred from the estimated average copy number of 7.14 ± 0.86 and 7.03 ± 0.92, respectively), and her elder brothers (II-1 and II-2) had approximately six copies of the repeat (inferred from the estimated average copy number of 5.51 ± 0.88 and 5.63 ± 0.79, respectively), as shown in Fig. 1B. According to the American College of Medical Genetics and Genomics (ACMG) Guidelines based on the evidences (1 A, 2 J, 3 A, 4E, the final point value: 0.2) [27], the duplication can be classified as a variant of uncertain significance. To validate the copy number and the involved exons of the duplication, qPCR and MLPA were performed. In qPCR, when comparing with the control sample (45,X), the target locus Dup309 of the four calibrator samples (45,X; 46,XX; 47,XXX; 46,XY) presented highly consistent copy number as expected, 1.09 ± 0.24, 2.02 ± 0.05, 2.87 ± 0.06, 1.03 ± 0.11, respectively, after normalizing and adjusting the corresponding copy number of the reference locus XP60. For the test samples, the proband’s father (I-1) and mother (I-2) showed a relative copy number of 0.98 ± 0.08, and 6.89 ± 0.56, respectively, for the target locus, while that for her elder brothers was 5.37 ± 0.47 and 6.05 ± 0.22, respectively, as shown in Fig. 1C. MLPA confirmed that the duplication involved exons 64–79 of DMD, with the FR values of 2.88–3.12 in the proband (II-3) and 2.82–3.48 in her mother (I-2), and 4.47–5.58 and 4.43–5.56 in her elder brothers II-1 and II-2, respectively, all of which were far beyond the reference range provided by the manufacturer (Fig. 1D).

In the proband from Family B (II-2), who exhibited typical DMD neuromuscular manifestations, a duplication involving exons 10–13 of DMD with a copy number of two was identified using MLPA (FR value: 1.95–2.04). This variant was inherited from his unaffected mother (I-2), who had a copy number of three for the segment (FR value: 1.50–1.52). The duplication could be pathogenic and presumably disrupted the function of the dystrophin protein, causing the disease (Fig. 2B).

The cryptic complex rearrangements involving DMD determined using OGM

To further explore the potential patterns underlying the identified duplications, OGM was performed on II-2 from Family A and II-2 from Family B. Genomic mapping for II-2 from Family A enabled the construction of a map exhibiting a complex in-cis tandem repeat with the segment covering exon 64–79 of the DMD gene (chrX: 30889742–31236824) (Fig. 3A), and no other conflicting map was constructed. The inferred upstream breakpoint of the segmental repeat was allocated to the closely upstream region of the TAB3 gene (OMIM #300480) with unknown morbid effects, and the downstream breakpoint was located in intron 63 of the DMD gene. With the limitation in the length of the assembled optical reads form OGM, the rearranged haplotype map only presented four complete copies and partial segment of the repeat; However, CNV analysis using OGM showed an estimated copy number of six for the repeat (Fig. 3A), consistent with CNV-seq and qPCR results. As per the pattern of the rearrangement inferred from OGM, at least one complete copy (reading frame) of the DMD gene were most likely to be retained, regardless of the structural changes (Fig. 3B). Thus, the dosage and function of the gene were presumably maintained.

Fig. 3

The rearrangement pattern and breakpoints of the complex in-cis tandem repeat in Family A. A. The OGM mapping revealed a complex in-cis tandem repeat involved in DMD. The numbers in the CNV call section indicate the copy number. The blue dashed box represents the partial segment of the repeat that OGM failed to show. The red arrows indicate the breakpoint junctions that will be used for breakpoint analysis. B. The pattern and breakpoint of the rearrangement. The blue line indicates the location of the upstream breakpoint was close to the upstream region of the TAB3 gene. The orange line indicates the location of the downstream breakpoint was in the intron 63 of the DMD gene. “I”: intron, “E”: exon

The OGM rearranged map for II-2 from Family B showed that a genomic segment from a donor chromosome X (chrX: 32595757–37697377) approximately 5.10 Mb in size was inverted and most likely inserted to another chromosome X (accepter). A genomic segment on the accepter chromosome X (chrX: 32660901–37661984) was replaced by the inserted donor segment (Fig. 4A). The reversed insertion generated an unbalanced chromosome X with a segmental repeat involving exons 10–13 of the DMD gene, exons 1–2 of the XK gene (OMIM#314850), and exons 4–5 of the LANCL3 gene, as shown in Fig. 4B. The rearrangement was predicted to break the complete copy and estimated reading frame of DMD, thereby causing dosage insufficiency and gene dysfunction.

Fig. 4

The rearrangement pattern and breakpoints of the unbalanced inversion-insertion rearrangement in Family B. A. The OGM showed an unbalanced inversion-insertion rearrangement involving DMD in Family B. The numbers in the CNV call section indicate the copy numbers. Red and purple arrow indicate the region of the breakpoint verification. B. The pattern of the rearrangement showed that the breakpionts of the complex rearrangement involved exons 10–13 of the DMD gene, exons 1–2 of the XK gene, and exons 4–5 of the LANCL3 gene. “F”: forward primer, “R”: reverse primer

The specific breakpoints of the complex rearrangements validated using NGS and Sanger sequencing

To determine the potential mechanisms underlying the various functional effects of the complex rearrangements, we further validated the specific breakpoints and investigated the characteristics of breakpoint junctions using NGS combined with long-range PCR amplification and Sanger sequencing. The rearrangement involving a multi-copy in-cis tandem repeat in Family A formed an identical breakpoint junction, whose upstream and downstream sequences could be mapped to chrX: 30886535 and chrX: 31238575, respectively, as shown in Fig. 5A and B. Sanger sequencing confirmed the results, but the downstream breakpoint can be mapped to chrX: 31238573 or chrX: 31238575 (Fig. 5C and D) because the dinucleotide repeat “GT” surrounding the breakpoint was homologous between the DMD and TAB3 gene sequences. Based on these findings, further reassessment of the variation was performed according to the evidences (1 A, 2 J, 3 A, 4 K, the final point value: -0.6) from the ACMG guidelines. Although the variation was still reclassified as a variant of uncertain significance, the evidences applicable and the final point values had changed: the evidence (4E) was not appropriate and the evidence (4 K) was adopted, and the corresponding point values changed from 0.2 to -0.6.

Fig. 5

Breakpoint analysis of the rearrangement of Family A. A. Gel analysis of the long-range PCR product; The red arrow indicates the long-range PCR product. Control: healthy control; kb: kilobase. B. The breakpoints were validated using NGS. The integrative genomics viewer (IGV) screenshot showed the locations of breakpoint junctions. C. Gel analysis of the regular PCR product. The red arrow indicates the regular PCR product. Control, healthy control; bp, base pair. D. The result of Sanger sequencing. The precise breakpoint junctions located in intron 63 of the DMD gene and intron 1 of TAB3 gene. The red dashed box represents homologous base

The unbalanced inversion-insertion rearrangement in Family B generated two breakpoint junctions. According to the NGS results, the breakpoint junction 1 can be mapped to chrX: 32664240 and chrX: 37705288 (Fig. 6A and B), respectively, and the breakpoint junction 2 can be mapped to chrX: 32587504 and chrX: 37656582 (Fig. 7A and B), respectively. These were supported by the findings from Sanger sequencing, which further characterized the signatures of the breakpoint junctions. A 19-bp microhomology “AGTAGCTGGGACTACAGGC” was identified in the breakpoint junction 1 (chrX: 32664240 and chrX: 37705288) (Fig. 6C and D), and a nucleotide “T” was inserted in the breakpoint junction 2 (chrX: 32587504 and chrX: 37656582) (Fig. 7C and D).

Fig. 6

Breakpoint analysis in breakpoint junction 1 of rearrangement in Family B. A. Gel analysis of the long-range PCR product. The red arrow indicates the PCR product. Control, healthy control; kb, kilobase. B. The breakpoint was validated using NGS. The IGV screenshot showed the locations of breakpoint junction 1. C. Gel analysis of the long-range PCR product. The red arrow indicates the regular PCR product. C, healthy control; bp, base pair. D. The results of Sanger sequencing showed that the precise locations of breakpoint junctions 1 were in DMD intron 9 and XK intron 1. The red dashed box represents microhomology sequences

Fig. 7

Breakpoint analysis in breakpoint junction 2 of the rearrangement in Family B. A. Gel analysis of the long-range PCR product. The red arrow indicates the long-range PCR product. Control, healthy controls; kb, kilobase. B. The breakpoint was validated using NGS. The IGV screenshot showed that the locations of breakpoint junction 2. C. Gel analysis of the regular PCR product. The red arrow indicates the regular PCR product. C, healthy control; bp, base pair. D. The results of Sanger sequencing showed that the precise locations of breakpoint junctions 2 were in DMD intron 13 and LANCL3 intron 2. The purple dashed box represents the inserted nucleotide