April 2016, VOLUME137 /ISSUE 4

Mitochondrial Copy Number as a Biomarker for Autism?

  1. Christelle Golzio, PhD and
  2. Nicholas Katsanis, PhD
  1. Center for Human Disease Modeling, Duke University School of Medicine, Durham, North Carolina
  • Abbreviations:
    autism spectrum disorder
    mitochondrial, mtCN, mitochondrial copy number
    mitochondrial DNA
  • Autism (AS) and autism spectrum disorders (ASD) represent a group of neurodevelopmental disorders of acute clinical and pathomechanistic heterogeneity. The serious impact of AS/ASD in families1,2 and the alarming growth of AS/ASD in the developed world3,4 have highlighted the pressing need both to develop new early diagnostic tools and to understand causality as a means of developing new therapeutic paradigms. In this issue of Pediatrics, Wong et al5 studied mitochondrial (mt) copy number (mtCN) and the incidence of mt deletions in parents and children with autism (refs 68 and review in ref 9). They report an increased mtCN in patients with AS compared with matched control children, as well as an increase in the number of mt deletions, although there was no evidence of excessive mt gene loss between these 2 groups.

    In addition, they found that fathers of AS children had a higher incidence of mitochondrial DNA (mtDNA) deletions (but not mtCN) compared with fathers of control children, with the converse being observed in mothers of AS children. Of note, increased maternal or paternal age did not seem to have an effect on the magnitude of this difference. Finally, because of the contributory role of p53 in mediating mtCN and possibly mtDNA integrity, the authors studied the p53 locus, where they found an increased incidence of microdeletions in both the children diagnosed with AS and their parents.

    These studies touch on 2 important areas of autism research. First, it is known that several genes causing AS/ASD, such as PTEN, CHD8, and CUL3, are regulators of p53.1013 However, given that each of these proteins has multiple functions, the implication of p53 directly might help us focus on a subset of their activities with regard to AS-relevant pathways. Second, these studies, performed in peripheral blood monocytes, have the potential to contribute to the keenly needed search for reliable biomarkers. As such, it is tantalizing to speculate that counts of quantitative analyses, such as mtCN, together with genomics and early-infantile behavioral characteristics,1416 might generate matrices of diagnostic and prognostic value.

    At the same time, several key questions remain. First, observational studies of heterogeneous neurodevelopmental traits, such as autism, have been notorious for generating false positives. Thus, replication of these data in a significantly larger cohort is mandated, not only for confirmation but also for a more accurate assessment of relative risk. This is particularly important in order to enable the community to assess the possibility that the mtCN/deletion phenotype might be independent of paternal age, in sharp contrast to the current thought about the mechanisms of de novo mutations (novel genetic lesion due to an error of copying of the genetic material in the germ cell of 1 of the parents or in the fertilized egg itself) in the nuclear genome.13,17,18

    Second, the question of cause and effect remains. The observation that mt deletions did not appear to affect gene numbers argues that the observed mt phenotypes might be outcomes or other processes, not drivers. This does not belittle the value of these data, but focuses their utility as a marker as opposed to an AS driver. The authors highlight, appropriately, the known roles of p53 in regulating both mtCN and mtDNA integrity.1921 At the same time, p53 negatively regulates the proliferation and survival of adult neuronal stem cells, without affecting their differentiation potential.22 In its absence (mouse p53-null, in which p53 is no longer functional), neuronal stem cell self-renewal and the number of differentiated neurons increases. Furthermore, in the absence of both p53 and the autism-associated gene Pten (mouse mutant model lacking both genes), neuronal stem cells have a higher renewal and neuronal differentiation potential.23

    As such, it is possible that the observed mtDNA defects are a mechanistically unrelated surrogate of p53-mediated defects in the central nervous system that are potent drivers of neurodevelopmental circuitry anomalies. This is a testable hypothesis; the identification of numerous AS-associated genes from exome studies (high-throughput sequencing of all the exons of known coding genes in the human genome) is potentiating the generation of a host of new mouse models. It will be interesting to ask whether such models exhibit mtDNA number and content pathologies and whether these pathologies extend to the developing nervous system. At the same time, studies of mt function in both animal models and in AS patients might provide further illumination.


      • Accepted January 6, 2016.
    • Address correspondence to Nicholas Katsanis, PhD, 300 N Duke St, Suite 47–101, Durham, NC 27701. E-mail: katsanis{at}
    • Opinions expressed in these commentaries are those of the author and not necessarily those of the American Academy of Pediatrics or its Committees.

    • FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

    • FUNDING: Dr Golzio was supported by National Institutes of Health Grant 1R01-MH106826-01. Funded by the National Institutes of Health (NIH).

    • POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

    • COMPANION PAPER: A companion to this article can be found on online at