Interpreting the clinical significance of multiple large-scale mitochondrial DNA deletions (MLSMD) in skeletal muscle tissue in the diagnostic evaluation of primary mitochondrial disease

Abstract

Background and objectives: Improved detection sensitivity from combined Long-Range PCR (LR-PCR), Next-Generation Sequencing (NGS), and droplet digital PCR (ddPCR) to identify multiple large-scale mtDNA deletions (MLSMD) and quantify deletion heteroplasmy have introduced clinical interpretation challenges. We sought to evaluate clinical, biochemical, and histopathological phenotypes of a large clinical cohort harboring MLSMD in muscle to better understand their significance across a range of clinical phenotypes.

Methods: A single-site retrospective study was performed of 212 diagnostic muscle biopsies obtained from patients referred for Primary Mitochondrial Disease (PMD) evaluation with muscle mitochondrial (mt)DNA sequencing performed at our institution, including electronic medical record (EMR) review of symptoms, biochemical results, and Mitochondrial Myopathy Composite Assessment Tool (MM-COAST) scores.

Results: MLSMD were identified in 50 of 212 (24%) diagnostic tissue biopsies, and were universally present. in subjects ≥50 years (n = 18/18). In 45 of 50 (90%) subjects with MLSMD, no definitive genetic etiology was identified, despite clinical whole exome sequencing (WES) and/or whole genome sequencing (WGS). MLSMD heteroplasmy levels quantified by ddPCR ranged from 0% to 33%, exceeding 10% heteroplasmy in 5/45 (11%). Subjects with MLSMD (n = 45) were more likely to demonstrate mitochondrial abnormalities on histopathology, upregulation (≥150% of control mean) of one or more electron transport chain (ETC) complex enzyme activities, and reduced citrate synthase indicative of mitochondrial depletion (<60% of control mean) relative to subjects without MLSMD (n = 155). As clinical phenotypes varied across the MLSMD cohort, Bernier diagnostic criteria major/minor symptoms were used to discriminate 13 of 45 subjects with "suspected" PMD having unrevealing WES/WGS results and 32 of 45 subjects scored as "less likely" to have PMD. Relative to the "less likely" cohort, a significantly higher frequency of biochemical and muscle histopathological abnormalities (ragged red and COX negative fibers) were observed in the "suspected" cohort, further supporting a higher index of suspicion for PMD, p < 0.05.

Discussion: MLSMD in skeletal muscle tissue were a common molecular finding (24%) in our cohort and consistently present in subjects ≥50 years. Among those with genetically undiagnosed MLSMD (n = 45), the "suspected" PMD subset (n = 13/45) represent a promising cohort for novel gene discoveries.

Keywords: electron transport chain (ETC) enzymatic activity; mitochondrial DNA (mtDNA); multiple large-scale mitochondrial DNA deletions (MLSMD); primary mitochondrial disease (PMD); ragged blue fibers (RBF); ragged red fibers (RRF).

FIGURE 1

Study Overview. Flowchart outlines the study cohort and analyses conducted. * Includes 1 cardiac biopsy case. ** resulted in m.3243A>G considered diagnostic (14% in blood, 42% in muscle). MLSMD = multiple large-scale mitochondrial DNA deletions. PMD = Primary mitochondrial disease. DM2 = myotonic dystrophy type 2.

FIGURE 2

MLSMD detected by long-range PCR and NGS. (A–D). Subjects with MLSMD heteroplasmy ≥10%. (E–G). Three representative subjects with MLSMD heteroplasmy <10%. (H) Deletion negative control. Left panels: NGS coverage profile to indicate the presence of large-scale mtDNA deletions. Right panels: long range PCR gel pictures. The presence of multiple amplification bands supports the detection of MLSMD by NGS. M: molecular weight marker; C: controls; P: Subject. X-axis: mtDNA position; Y-axis: NGS coverage depth. The deletion heteroplasmy was evaluated by ddPCR. Heteroplasmy <10% was considered as low level and may not be accurately quantified by ddPCR assay.

FIGURE 3

Detection rate of MLSMD across different age groups. (A). Subjects with and without MLSMD were stratified into 10-year age groups. The presence (orange) and lack of (blue) MLSMD subjects were demonstrated in each age group. All subjects in the ≥50 years age group demonstrated MLSMD in their diagnostic muscle biopsies (n = 18). (B). Mean age of the MLSMD cohort (n = 45) was significantly higher as compared to subjects without MLSMD (n = 155), even after comparison of subjects below 50 years of age only (n = 28). ****: p < 0.0001.

FIGURE 4

Comparison of muscle biopsy histopathology results, ETC enzyme activity, citrate synthase (CS) activity, and mtDNA content in MLSMD and non-MLSMD groups. Subjects with confirmed PMD were excluded. (A) Comparison of the proportion of subjects with ragged red fibers (RRFs), ragged blue fibers (RBFs), cytochrome oxidase negative fibers (COX-), and increased subsarcolemmal mitochondrial (SSM) staining in diagnostic muscle biopsies between MLSMD and the non-MLSMD groups by Fischer’s exact test. Results showed significantly increased subjects with RRFs (p = 0.038), COX- fibers (p = 0.015) and SSM (p = 0.037) in the MLSMD groups as compared to the non-MLSMD groups (<50 years). No statistically significant differences in RBFs were observed between MLSMD and non-MLSMD groups most likely related to the small cohort size. (B) Comparison of the proportion of subjects with ETC enzyme activity abnormalities between the MLSMD and non-MLSMD groups. ‘Deficiency’ denotes one or more ETC complex activities meeting the modified Walker diagnostic criteria (<30% of the control value); ‘Unremarkable’ indicates ETC enzyme activities between 30%–149% of the control value; ‘Upregulated’ signifies one or more complex activities ≥150% of the control value. Results showed significantly increased subjects with ‘Upregulated’ ETC enzyme activity in the MLSMD group as compared to the non-MLSMD group (p = 0.015). (C) Comparison of the proportion of subjects with abnormal CS activity results between the MLSMD and non-MLSMD groups. ‘Reduced CS’ indicates CS activity <60% of the normal control value; ‘Unremarkable CS’ indicates CS activity between 60%–149% of normal control value; iii) ‘Upregulated CS’ indicates CS activity ≥150% of normal control value. Results showed that subjects with MLSMD had a significantly higher likelihood of having reduced CS activity (p = 0.0006), whereas non-MLSMD subjects had significantly higher likelihood of unremarkable CS activity (p = 0.005). There were no significant differences observed in the ‘Upregulated CS’ category between MLSMD and non-MLSMD groups. (D) Comparison of mtDNA content between MLSMD and non-MLSMD groups. ‘Depleted’ indicates mtDNA content <50% of age and tissue matched control values; ‘Reduced’ includes mtDNA content between 50%–59%; ‘Unremarkable’ denotes mtDNA content between 60%–139%; ‘Proliferated’ refers to mtDNA content ≥140% of the age-matched control value. Results showed that the majority of subjects in both the MLSMD and non-MLSMD groups displayed ‘unremarkable’ mtDNA content, which was seen in 67% and 78% of the respective cohorts. No significant differences were observed between mtDNA content results between groups. The bar charts present the rates of abnormal results in the MLSMD and non-MLSMD groups. The corresponding tables show the number of subjects with abnormal results (numerator) and the total number of subjects with available results in each test category (denominator). Fisher’s exact test was used to calculate p-values, determining the significance of the observed differences. *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 5

Clinical phenotype across MLSMD Groups I-III. Group I comprised of subjects with a high suspicion of PMD. Group II included subjects age ≥50 years with a lower index of suspicion, classified as ‘Less likely PMD’. The remaining subjects in Group III were <50 years of age, also considered as ‘Less likely PMD’. The percent (%) in each colored box represents the number of subjects with the stated phenotype/total subjects in Groups I (dark blue), II (green) or III (light blue, see key). Muscle weakness and exercise intolerance was consistently the most prevalent symptom across all three groups (85%–100%). Other common features observed across all three groups included myalgias/muscle pain, gastrointestinal (GI) symptoms, ptosis, and memory difficulties/cognitive decline. PEO: Progressive external ophthalmoplegia; POTS: Postural orthostatic tachycardia syndrome; FTT: Failure to thrive; GDD: Global developmental delay.

FIGURE 6

Comprehensive analysis of subjects with MLSMD (n = 45) and correlation with heteroplasmy levels (A). Results overview for 45 subjects with MLSMD lacking a genetic diagnosis. Results display individual subject demographics, histopathologic, and biochemical characteristics across the MLSMD cohort. Subject # corresponds to the subject list in Supplementary Table S1. Five subjects with known PMD genetic diagnosis (or DM2) were excluded from these analyses (Supplementary Table S1). MLSMD heteroplasmy levels were detected and quantified by combined NGS, LR-PCR, and ddPCR analysis. Please note, heteroplasmy levels <10% are below ddPCR detection limit and therefore may not be accurately quantified by ddPCR, but can be detected by NGS and LR-PCR. M: Missing report or no result available. (B). Abnormal results identified in MLSMD cohort by mtDNA heteroplasmy levels (≥10% and <10%). (B) illustrates the rates of abnormal results in histopathology (RRFs: ragged red fibers; RBFs: ragged blue fibers; COX-: cytochrome oxidase negative fibers; Increased SSM: Increased subsarcolemmal mitochondrial staining/accumulation), muscle electron transport chain (ETC) assay, mtDNA content, Creatine kinase (CK), and growth differentiation factor 15 (GDF15) in each test category, distinguishing between MLSMD heteroplasmy levels of 10% and above and those below 10%. The number of subjects with abnormal results, the total number of subjects with available results in each diagnostic test category, and the ’abnormal percentages’ are listed in the table. Histological abnormalities including RRFs, RBFs, and COX– fibers occurred significantly more frequently in the ≥10% heteroplasmy groups when compared to the <10% heteroplasmy groups (p = 0.018, p = 0.0081, p = 0.0094, respectively by Fisher’s exact test). *p-value <0.05, **p-value <0.01.