Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Filters applied. Clear all
. 2016 Oct 28;12(10):e1006391.
doi: 10.1371/journal.pgen.1006391. eCollection 2016 Oct.

Genetic Evidence for Elevated Pathogenicity of Mitochondrial DNA Heteroplasmy in Autism Spectrum Disorder

Free PMC article

Genetic Evidence for Elevated Pathogenicity of Mitochondrial DNA Heteroplasmy in Autism Spectrum Disorder

Yiqin Wang et al. PLoS Genet. .
Free PMC article


Increasing clinical and biochemical evidence implicate mitochondrial dysfunction in the pathophysiology of Autism Spectrum Disorder (ASD), but little is known about the biological basis for this connection. A possible cause of ASD is the genetic variation in the mitochondrial DNA (mtDNA) sequence, which has yet to be thoroughly investigated in large genomic studies of ASD. Here we evaluated mtDNA variation, including the mixture of different mtDNA molecules in the same individual (i.e., heteroplasmy), using whole-exome sequencing data from mother-proband-sibling trios from simplex families (n = 903) where only one child is affected by ASD. We found that heteroplasmic mutations in autistic probands were enriched at non-polymorphic mtDNA sites (P = 0.0015), which were more likely to confer deleterious effects than heteroplasmies at polymorphic mtDNA sites. Accordingly, we observed a ~1.5-fold enrichment of nonsynonymous mutations (P = 0.0028) as well as a ~2.2-fold enrichment of predicted pathogenic mutations (P = 0.0016) in autistic probands compared to their non-autistic siblings. Both nonsynonymous and predicted pathogenic mutations private to probands conferred increased risk of ASD (Odds Ratio, OR[95% CI] = 1.87[1.14-3.11] and 2.55[1.26-5.51], respectively), and their influence on ASD was most pronounced in families with probands showing diminished IQ and/or impaired social behavior compared to their non-autistic siblings. We also showed that the genetic transmission pattern of mtDNA heteroplasmies with high pathogenic potential differed between mother-autistic proband pairs and mother-sibling pairs, implicating developmental and possibly in utero contributions. Taken together, our genetic findings substantiate pathogenic mtDNA mutations as a potential cause for ASD and synergize with recent work calling attention to their unique metabolic phenotypes for diagnosis and treatment of children with ASD.

Conflict of interest statement

The authors have declared that no competing interests exist.


Fig 1
Fig 1. Characteristics of mtDNA variants.
(A) The nonsynonymous-synonymous rate ratio (hN/hS) for all possible nucleotide substitutions, observed heteroplasmies, and homoplasmies on mtDNA. hN is the number of nonsynonymous substitutions divided by the total number of possible nonsynonymous substitutions on mtDNA; similarly, hS is the number of synonymous substitutions divided by the total number of possible synonymous substitutions on mtDNA [34]. The ratio of hN and hS is indicative of purifying selection if it is significantly less than one [34]. (B) The box plot of CADD pathogenicity scores for all possible substitutions, observed heteroplasmies and homoplasmies at nonsynonymous sites on mtDNA. (C, D) hN/hS (C) and CADD pathogenicity scores of nonsynonymous substitutions (D) for heteroplasmies at non-polymorphic sites, heteroplasmies at polymorphic sites and homoplasmies at non-heteroplasmic sites. Each nucleotide substitution was only counted at most once. ***P<2x10-4; **P<2x10-3; P: p values for Chi-squared test (A,C) or Mann-Whitney test (B,D).
Fig 2
Fig 2. Comparison of mtDNA mutation burden and pathogenicity between autistic probands and non-autistic siblings.
Mutation burden was calculated using mutations with DAF ≥5% at mtDNA sites where all members in a family had >40X sequencing depth. The average number of mutations per individual was shown in (A) for all valid sites and non-polymorphic sites, in (B) for sites in the D-loop region and synonymous sites in the coding region, and in (C) for sites in the RNA region, nonsynonymous sites in the coding region, and mutations predicted pathogenic with CADD Phred score >15 or >20. Results in the main text were based on “CADD > 15”. Error bars represent the standard error of the mean. ***P<0.001; **P<0.01; *P<0.05; P: p values for one-tailed paired t-test.
Fig 3
Fig 3. Transmission pattern of pathogenic mtDNA mutations.
The proportion of untransmitted (red), transmitted (purple) and de novo (blue) mutations that were nonsynonymous or predicted pathogenic was shown in (A) for mother-sibling pairs and in (B) for mother-proband pairs. Only mtDNA sites (n = 448) detected with high-confidence heteroplasmies (MAF ≥5%) or de novo homoplasmies were used for calculation. Untransmitted mutations were defined as mutations present in the mother, but were undetectable or had DAF <2% in the child. Likewise, de novo mutations were defined as those present in the child, but were undetectable or had DAF <2% in the mother. Transmitted mutations were mutations shared between the mother and the child with DAF ≥2%. The total number of untransmitted, transmitted and de novo mutations was indicated in parentheses for mother-sibling pairs and mother-proband pairs in the legend. Error bars represent the 95% confidence interval of the proportion estimated based on 10,000 bootstrap samples. CADD: mutations predicted pathogenic with CADD Phred score >15 or >20. Results in the main text were based on “CADD > 15”; P: p values for Fisher’s exact test.

Similar articles

See all similar articles

Cited by 11 articles

See all "Cited by" articles


    1. Jeste SS, Geschwind DH. Disentangling the heterogeneity of autism spectrum disorder through genetic findings. Nat Rev Neurol. Nature Publishing Group; 2014;10: 74–81. 10.1038/nrneurol.2013.278 - DOI - PMC - PubMed
    1. Landa RJ. Diagnosis of autism spectrum disorders in the first 3 years of life. Nat Clin Pract Neurol. 2008;4: 138–47. 10.1038/ncpneuro0731 - DOI - PubMed
    1. Christensen DL, Baio J, Braun KVN, Bilder D, Charles J, Constantino JN, et al. Prevalence and characteristics of Autism Spectrum Disorder among children aged 8 years—Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2012. MMWR Surveill Summ. 2016;65: 1–23. - PubMed
    1. Zablotsky B, Black LI, Maenner MJ, Schieve LA, Blumberg SJ. Estimated prevalence of Autism and other developmental disabilities following questionnaire changes in the 2014 National Health Interview Survey. Natl Health Stat Report. 2015; 1–20. - PubMed
    1. Olsson MB, Westerlund J, Lundström S, Giacobini M, Fernell E, Gillberg C. “Recovery” from the diagnosis of autism—and then? Neuropsychiatr Dis Treat. 2015;11: 999–1005. 10.2147/NDT.S78707 - DOI - PMC - PubMed


Grant support

This work was supported by various funds from Cornell University, NSF MCB-1243588, NIH 1R01AI085286, and a grant for ENN science and technology development to ZG. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.