Distinct age-related pattern of mitochondrial somatic mutations across Multiple Sclerosis phenotypes
Description
Multiple sclerosis (MS) is a chronic, autoimmune, inflammatory disease of the central nervous system (CNS) characterized by demyelination and neurodegeneration. Among the various pathogenic mechanisms, oxidative stress plays a critical role in driving both inflammation and neuronal damage, and emerging evidence suggests that mitochondrial dysfunction may significantly contribute to disease progression. Whole mtDNA was sequenced from blood-derived DNA using long-range PCR and Illumina® Nextera XT kit. Somatic mutations were defined based on heteroplasmy levels between 1–5%. Linear regression models were used to assess the association between age and mutation rate. We observed a significant age-dependent increase in low-frequency nonsynonymous mtDNA mutations in MS. Analyses stratified by disease course revealed this effect was driven by PPMS patients, while no association was seen in RRMS, suggesting course-specific mitochondrial trajectories. Furthermore, fast-progressing patients showed a positive linear relationship between age and mtDNA mutations rate, while slow-progressing ones showed a negative trend. These findings support the existence of a differential age-related accumulation of somatic mtDNA mutations across MS courses and underline the importance of mitochondrial genome instability in disease progression.
Files
Steps to reproduce
Genomic DNA extraction was performed with phenol-chloroform standard method or the Maxwell 16 blood DNA purification kit (Promega). MtDNA enrichment was performed with two sets of two long-range PCR reactions with the LA PCR™ Kit Ver.2.1 (TaKaRa). FSequencing libraries were prepared with the Nextera XT DNA Library Prep Kit (Illumina®) and sequenced on the MiSeq® platform (Illumina®) using MiSeq® Reagent Kit v3 (600 cycles) in paired-end mode. Raw sequencing reads were assessed using FastQC v0.11.8 to evaluate quality metrics such as per-base sequence quality, GC content, and duplication levels. Adapters and low-quality bases were removed with Trimmomatic v0.39. To reduce erroneous assignment of reads originating from nuclear mitochondrial DNA segments (NuMTs), all samples were aligned to the human whole-genome reference (hg38) rather than to the mitochondrial genome alone. Read alignment was performed using BWA-MEM v0.7.17 with parameters `-K 100000000 -p -v 3 -t 2 -Y`. Alignments were sorted and indexed using Samtools v1.9, and duplicate reads were marked using the feature Picard MarkDuplicates within GATK v4.1.9.0. Each sample was aligned in parallel to two whole-genome references: (i) the standard hg38 reference containing the canonical mitochondrial chromosome (chrM), and (ii) a modified hg38 reference in which the mitochondrial genome was circularly shifted by 8,000 base pairs. This dual-alignment strategy was implemented to improve variant detection near the mitochondrial control region and to mitigate alignment artifacts introduced by the linearization of the circular mitochondrial genome. Following alignment, reads mapping to chrM were extracted independently from both whole-genome BAM files using Samtools v1.9. Median nuclear coverage was estimated from the whole-genome alignments using Picard CollectHsMetrics and used as a quality control metric. Mitochondrial variant calling was performed separately on chrM reads derived from the standard and shifted references using Mutserve v2.0.0-rc13. Variants were required to pass all internal Mutserve quality filters (“PASS”), specifically calibrated to distinguish true low-frequency variants from sequencing noise. Variant calls generated from the shifted mitochondrial reference were subsequently lifted over to standard mitochondrial genome coordinates and merged with variants identified from the unshifted reference to generate a unified variant set per sample. Final variant annotation was performed using Mutserve’s `annotate` function, and downstream analyses were conducted using custom R scripts. Positions were considered covered if ≥100 reads were present and samples were included in the final analysis only if at least 5,000 sequenced bases were covered at a depth of ≥100×, ensuring sufficient coverage for reliable mitochondrial variant detection.
Institutions
- Ospedale San RaffaeleLombardia, Milano