Application of Genome Sequencing from Blood to Diagnose Mitochondrial Diseases

Abstract

Mitochondrial diseases can be caused by pathogenic variants in nuclear or mitochondrial DNA-encoded genes that often lead to multisystemic symptoms and can have any mode of inheritance. Using a single test, Genome Sequencing (GS) can effectively identify variants in both genomes, but it has not yet been universally used as a first-line approach to diagnosing mitochondrial diseases due to related costs and challenges in data analysis. In this article, we report three patients with mitochondrial disease molecularly diagnosed through GS performed on DNA extracted from blood to demonstrate different diagnostic advantages of this technology, including the detection of a low-level heteroplasmic pathogenic variant, an intragenic nuclear DNA deletion, and a large mtDNA deletion. Current technical improvements and cost reductions are likely to lead to an expanded routine diagnostic usage of GS and of the complementary "Omic" technologies in mitochondrial diseases.

Keywords: genome sequencing; heteroplasmy; mitochondria; mutation; respiratory chain.

Figure 1

GS detects low levels of heteroplasmy in blood of P1. (a) Integrative Genomics Viewer (IGV) browser shows the identification of m.3243A>T in blood. The black arrow points to the variant (in red) present in 79 of 4235 reads aligned to the m.3243 position. (b) Partial electropherogram showing the heteroplasmic m.3243A>T variant detected in urine. (c) IGV browser shows the identification of m.3243A>T in muscle, the variant (in red) is present in 6108 of 8799 reads aligned to the m.3243 position. (d) Mamit-tRNA [38] schematic of the mt-tRNA^Leu(UUR) cloverleaf secondary structure showing in black positions of reported pathogenic variants. The m.3243A>T variant found in P1 is highlighted with a star and is located in the D-loop of the mt-tRNA^Leu(UUR).

Figure 1

GS detects low levels of heteroplasmy in blood of P1. (a) Integrative Genomics Viewer (IGV) browser shows the identification of m.3243A>T in blood. The black arrow points to the variant (in red) present in 79 of 4235 reads aligned to the m.3243 position. (b) Partial electropherogram showing the heteroplasmic m.3243A>T variant detected in urine. (c) IGV browser shows the identification of m.3243A>T in muscle, the variant (in red) is present in 6108 of 8799 reads aligned to the m.3243 position. (d) Mamit-tRNA [38] schematic of the mt-tRNA^Leu(UUR) cloverleaf secondary structure showing in black positions of reported pathogenic variants. The m.3243A>T variant found in P1 is highlighted with a star and is located in the D-loop of the mt-tRNA^Leu(UUR).

Figure 2

GS detects an intragenic nuclear DNA deletion in AARS2 in P2. IGV coverage plot shows ≈50% fewer reads in the region spanning the deletion. The read pairs that are colored red are flanking the deletion.

Figure 3

OXPHOS protein expression in patient P2 AARS2 and control fibroblasts. (a) Representative Western blot showing decreased protein levels of NDUFB8 (CI) and COXII (CIV) in patient P2 similar to patient A2 (positive control), who harbors likely pathogenic AARS2 variants c.1874G>A (p.Arg625His) and c.665C>T (p.Thr222Ile) in trans [49]; VDAC1 (porin) was used as a mitochondrial loading control. (b) Densitometry analysis suggests 85% lower levels of COXII (CIV) in both P2 and A2, and suggested 85% lower NDUFB8 (CI) in P2 and 65% lower in A2 relative to controls.

Figure 4

Deletion in the mtDNA in P3 identified by GS. (a) IGV coverage plot shows ≈50% fewer reads in the area spanning 9031bp of the mtDNA; (b) MitoBreak figure depicts the single large-scale heteroplasmic deletion spanning 23 mitochondrial genes (black bar).