Evidence linking oxidative stress, mitochondrial dysfunction, and inflammation in the brain of individuals with autism
Daniel A. Rossignol1,* and Richard E. Frye2
Author information ► Article notes ► Copyright and License information ►
This article has been cited by other articles in PMC.
Autism spectrum disorders (ASD) are a group of neurodevelopmental disorders that are defined by behavioral observations including communication and social interaction problems and repetitive behaviors (APA, 1994). ASD affects an estimated 1 out of 88 individuals in the United States (U.S.) (Baio, 2012) with four times more males than females being affected (Rice, 2007). The etiology of ASD is unclear at this time. Although several genetic syndromes, including Fragile X and Rett syndrome, have been associated with ASD; genetic defects account for only a small percentage of ASD cases (Schaefer et al., 2013).
Although many of the cognitive and behavioral features of ASD are thought to arise from dysfunction of the brain, evidence from many fields of medicine has documented physiological abnormalities in organs besides the brain that are associated with ASD, suggesting that, in some individuals, ASD arises from systemic, rather than organ specific abnormalities (Rossignol and Frye, 2012a). Specifically, in recent decades, research and clinical studies have implicated physiological and metabolic systems that transcend specific organ dysfunction, such as immune dysregulation and inflammation, abnormalities in redox regulation and oxidative stress, and dysfunction of energy generation and mitochondrial systems (Ming et al., 2008; Rossignol and Frye, 2012a). In this context, ASD may arise from, or at least involve, systemic physiological abnormalities rather than being a purely central nervous system (CNS) disorder (Herbert, 2005), at least in a subset of individuals with ASD. However, because the CNS is affected in ASD, examining physiological abnormalities in the brain may reveal more about what is abnormal than inspecting abnormalities in blood or urine samples.
A number of studies have reported evidence of oxidative stress in individuals with ASD (Yorbik et al., 2002; James et al., 2004, 2006, 2009a,b; Ming et al., 2005; Chauhan and Chauhan, 2006; Yao et al., 2006; Al-Gadani et al., 2009; Melnyk et al., 2012; Rose et al., 2012a; Rossignol and Frye, 2012a) and their parents (James et al., 2008). Genetic variations in glutathione-related pathways have been observed in ASD (Boris et al., 2004; James et al., 2006; Bowers et al., 2011; Frustaci et al., 2012) and have been correlated in some studies with ASD behaviors (Goin-Kochel et al., 2009; Guo et al., 2012). Several case-control studies have reported lower concentrations of reduced glutathione (GSH), higher levels of oxidized glutathione (GSSG) and a decrease in the GSH/GSSG redox ratio (James et al., 2004, 2006, 2009b), along with a lower mitochondrial GSH reserve (James et al., 2009a) in individuals with ASD compared to controls. In addition, in some studies, lower GSH levels (Adams et al., 2009) and markers of increased oxidative stress (Ghezzo et al., 2013) have been correlated with ASD severity. Markers of oxidative stress have also been correlated with the severity of gastrointestinal problems in ASD individuals (Gorrindo et al., 2013). Notably, these aforementioned studies examined peripheral markers of oxidative stress, including those found in blood and urine. Recently, a number of studies have reported evidence of oxidative stress in post-mortem brain samples from individuals with ASD compared to controls (Evans et al., 2008; López-Hurtado and Prieto, 2008; Sajdel-Sulkowska et al., 2008, 2009; Palmieri et al., 2010; Chauhan et al., 2011; Sajdel-Sulkowska et al., 2011; Chauhan et al., 2012a; Rose et al., 2012b; Gu et al., 2013a,b; Tang et al., 2013).
Multiple studies have also reported evidence of mitochondrial dysfunction in individuals with ASD (Rossignol and Bradstreet, 2008; Weissman et al., 2008; Giulivi et al., 2010; Guevara-Campos et al., 2010; Shoffner et al., 2010; Zhang et al., 2010; Dhillon et al., 2011; Frye and Rossignol, 2011; Chauhan et al., 2012b; Frye, 2012; Frye and Rossignol, 2012a; Rossignol and Frye, 2012a,b; Frye et al., 2013a,b; Frye and Rossignol, 2013). In some studies, biomarkers of mitochondrial dysfunction have been associated with autistic behaviors or autism severity (Minshew et al., 1993; Mostafa et al., 2005). One systematic review reported that over 30% of children with ASD have biomarkers of abnormal mitochondrial function suggesting that a relatively high percentage of individuals with ASD might have some degree of mitochondrial dysfunction (Rossignol and Frye, 2012b). Another study reported that up to 50% of children with ASD have biomarkers of mitochondrial dysfunction that are valid (that is, they correlate with other biomarkers of mitochondrial dysfunction) and are consistently abnormal (that is, they are repeatedly abnormal) (Frye, 2012). However, like the studies on oxidative stress and ASD, most of the published literature concerning mitochondrial dysfunction has examined blood and urine samples. A number of studies recently have reported evidence of mitochondrial dysfunction in ASD brain samples compared to controls (Palmieri et al., 2010; Chauhan et al., 2011; Anitha et al., 2012, 2013; Ginsberg et al., 2012; Rose et al., 2012b; Tang et al., 2013).
Finally, a number of studies have reported evidence of immune dysregulation and/or inflammation in individuals with ASD (Gupta et al., 2010; Onore et al., 2012; Rossignol and Frye, 2012a; Depino, 2013; Gesundheit et al., 2013; Goines and Ashwood, 2013), including gene changes pertaining to the immune system (Michel et al., 2012; Poultney et al., 2013). In some studies, biomarkers of inflammation or immune dysregulation have been correlated with ASD severity (Mostafa and Kitchener, 2009; Al-Ayadhi and Mostafa, 2011, 2012, 2013; Khakzad et al., 2012; Mostafa and Al-Ayadhi, 2012) and an elevation in TNF-alpha has been reported in ASD lymphocytes (Malik et al., 2011a) and in amniotic fluid in children who develop autism (Abdallah et al., 2013). Particular interest surrounds elevations found in autoantibodies to brain elements and other important molecular targets such as the folate receptor autoantibody (Connolly et al., 1999; Rossignol and Frye, 2012a; Frye et al., 2013c). Although there have been a large number of studies examining immune abnormalities in ASD, almost all of these studies have examined blood and urine samples. However, some studies have recently reported evidence of brain-related immune dysregulation or inflammation in ASD compared to controls (Vargas et al., 2005; Chez et al., 2007; Garbett et al., 2008; Li et al., 2009; Morgan et al., 2010; Wei et al., 2011; Young et al., 2011; Rose et al., 2012b; Suzuki et al., 2013).
Recently, an interrelationship between oxidative stress, mitochondrial dysfunction, and/or inflammation has been reported in some individuals with autism (James et al., 2009a; Mostafa et al., 2010; Zhang et al., 2010; Rose et al., 2012b; Frye et al., 2013a; Napoli et al., 2013; Theoharides et al., 2013). In this manuscript, we concentrate on studies that have documented these physiological abnormalities specifically in the CNS of individuals with ASD. Reviewing the evidence for these physiological abnormalities specifically in the CNS is important for several reasons. Firstly, the CNS is protected from the rest of the body by the blood-brain barrier. Although there is evidence that these physiological abnormalities are present in non-CNS tissue in individuals with ASD, it does not necessarily mean that they are present in the CNS. Demonstrating that these abnormalities also affect the brain would suggest that brain dysfunction in individuals with ASD is not necessarily only secondary to systematic abnormalities, but that the same abnormalities that influence peripheral organs also directly influence brain function. Secondly, there are particular patterns of abnormalities in the CNS that are associated with ASD. Indeed, abnormalities in ASD have been reported in the frontal and temporal cortices, the hippocampus and amygdala as well as the cerebellum. Determining whether these physiological abnormalities are also present in these brain areas would provide insight into whether they could be involved in the pathological mechanisms that result in ASD. Thus, this manuscript reviews the evidence for oxidative stress, mitochondrial dysfunction and immune dysregulation/inflammation in the brains of individuals with ASD compared to controls as well as the evidence linking these abnormalities.
Still, there are larger questions that must be answered beyond confirming the notion that oxidative stress is present in the brain of children with ASD. For example, it is possible that the reduced transportation of folate into the brain as a consequence of the folate receptor alpha autoantibody or mitochondrial dysfunction could reduce the function of methylation and glutathione metabolism specifically within the brain leading to some of the findings described above (Frye et al., 2013c). However, many of these same findings reported for the brain (oxidative damage to lipids, protein and DNA, glutathione abnormalities, reduced function of enzymes essential for regulating oxidative stress) have been found in the blood, immune cells and cell lines derived from individuals with ASD, thereby raising the question of whether these findings are specific for the brain or whether they represent a more general process (Frye et al., 2013c; Frye and James, 2014). Lastly, the etiology of these abnormalities is not clear, as both increases in pro-oxidant influences and reductions in antioxidant defenses have both been associated with ASD (Chauhan and Chauhan, 2006) and it is clear that there are no simple genetic abnormalities that account for these findings (Frustaci et al., 2012; Frye and James, 2014).
Studies of mitochondrial dysfunction in the ASD brain.
Overall these studies provide support for mitochondrial dysfunction in the brain of individuals with ASD. MRS studies using both Phosphorus-31 and 1H techniques have examined energy metabolites in the brain of individuals with ASD, although many more studies have used the latter technique. Phosphorus-31 MRS has found abnormal energy metabolites in the frontal cortex (Minshew et al., 1993; Golomb et al., 2014) while 1H-MRS has found a reduction in NAA in the global white and gray matter and the parietal, anterior cingulate and cerebellum areas (Ipser et al., 2012). ETC function has been reported to be depressed in frontal (Chauhan et al., 2011), temporal (Chauhan et al., 2011; Tang et al., 2013), and cerebellar (Chauhan et al., 2011) brain tissue derived from individuals with ASD, with ETC complex I most commonly reported depressed. Other studies noted decreases in the activity of non-ETC mitochondrial enzymes (aconitase, pyruvate dehydrogenase) in frontal (Gu et al., 2013a), temporal (Rose et al., 2012b), and cerebellar (Rose et al., 2012b) tissue derived from children with ASD. Depressed expression of ETC genes in the occipital and cerebellar areas and of ETC and non-ETC genes in the cingulate, thalamus and frontal areas have been reported (Anitha et al., 2013). In addition, changes in genes that control mitochondrial dynamics have been noted in the temporal lobe (Tang et al., 2013).
As with studies on oxidative stress, studies of mitochondrial function in the brain of individuals with ASD are mostly based on small numbers of samples, involve a wide variety of methods, and study various regions of the brain without consistency across studies. Despite these limitations, these studies demonstrate that mitochondrial dysfunction is consistently found in the brain of individuals with ASD. The studies on ETC function are consistent with studies performed on muscle tissue as both demonstrate ETC complex I deficiency as the most prevalent ETC complex abnormality (Rossignol and Frye, 2012b). Several studies have provided powerful evidence for the correspondence between oxidative stress and mitochondrial dysfunction in the same brain samples (Chauhan et al., 2011; Rose et al., 2012b; Tang et al., 2013). This is one step forward to understanding the interaction between oxidative stress and mitochondrial dysfunction. Future studies could make such evidence more powerful by correlating these abnormalities with peripheral markers of mitochondrial dysfunction and oxidative stress as well as examining clinical characteristics.
Studies of inflammation/immune dysregulation in the ASD brain.
While some studies have reported microglial activation in individuals with autism compared to controls, others have studied differences in the spatial organization of microglial cells in individuals with autism. Microglia are immune cells in the CNS which are activated to eliminate damaged cells or infectious agents through the process of phagocytosis. However, when the microglia are chronically activated, they may increase inflammation through the release of proinflammatory cytokines and free radicals (Dheen et al., 2007). The first study to examine microglia in ASD reported significant activation of microglia and reactive astroglia in the middle frontal gyrus, anterior cingulate gyrus and cerebellum in 11 patients with autism compared to 6 controls (Vargas et al., 2005). The second study examined the dorsolateral prefrontal cortex in 13 males with autism and 9 controls and reported marked microglial activation in 5 out of the 13 autism cases (38.5%) and mild microglial activation in another 4 cases (30.8%) (Morgan et al., 2010). Finally, one study of 20 men with autism and 20 age- and IQ-matched controls reported evidence of microglial activation using positron emission tomography in multiple brain regions (cerebellum, brainstem, corpus callosum, fusiform gyri, superior temporal gyri, anterior cingulate, orbitofrontal, and parietal lobes) in the autism group (Suzuki et al., 2013).
The evidence reviewed above clearly supports the notion that there are alterations in the immune system upon examination of the brain in individuals with ASD. The strongest evidence for activation of the immune system are the studies which demonstrated histological evidence of microglia cell changes in the frontal (Vargas et al., 2005; Morgan et al., 2010, 2012a; Tetreault et al., 2012), cingulate (Vargas et al., 2005) and cerebellum (Vargas et al., 2005). Neuroimaging supports these histological findings (Suzuki et al., 2013). Evidence for disruption in immune regulation is supported by elevations in proinflammatory cytokines in brain tissue from the frontal (Li et al., 2009), cingulate (Vargas et al., 2005), and cerebellum (Wei et al., 2011) and in CSF (Vargas et al., 2005; Chez et al., 2007) derived from individuals with ASD and elevations in the expression of genes regulating proinflammatory pathways in the temporal (Garbett et al., 2008) and frontal (Young et al., 2011) areas in individuals with ASD.
Although some studies have reported some negative or inconsistent results (Vargas et al., 2005; Zimmerman et al., 2005; Malik et al., 2011b), the majority of studies point to an activation of the innate immune system in the brain of individuals with ASD and some of the findings, particularly the cytokine elevations, parallel abnormal elevations in cytokines reported in non-CNS tissue in children with ASD. Although these studies suffer from small sample sizes and inconsistency in brain areas examined, together they provide support for more comprehensive research into the role of inflammation and immune dysregulation in the brain of children with ASD.
Although ASD is defined by observations of behaviors and is thus classified as a psychiatric disorder, recent evidence has pointed to physiological abnormalities in ASD, suggesting that ASD has a clear biological basis with features of known medical disorders.
Conclusions and perspectives
Overall, the studies reviewed above provide support for the idea that oxidative stress, mitochondrial dysfunction and inflammation/immune dysfunction, which are physiological abnormalities identified in non-CNS tissue in children with ASD, are also found to affect the CNS. A few studies demonstrated the connection between these physiological abnormalities. However, there were several limitations to the studies reviewed, including small sample sizes and inconsistencies in the techniques and biomarkers studied and the brain areas examined. Because of these limitations, at this time, it is difficult to know if the findings are localized to a certain portion of the brain or whether these abnormalities are more diffuse. Another challenge is whether or not these abnormalities can be generalized to all children with ASD, or if they represent a subgroup of children with ASD. However, the consistent positive findings across studies suggest that these effects are not subtle and may be important in the pathological mechanisms that disrupt brain function in ASD.
1. Martha R.Herbert “Autism a brain disorder…., Clinical Neuropsychiatry(2005)2,6,354-379.
2. ValerieW Hu,etal,”Gene expression profiling …..,BMC Genomics 2006,7:118
3. Zimmerman etal, “Neuroglial Activation …., Ann. neurology 2005:57:000-0000
4. Beta Stevens etal, “The classical complement cascade…., Cell 131,1164-1178,December 14,2007 Elsevier
5. Ramachandran VS”Mirror neurons…., http://www.edge.org/3rrdculture/ramachanrdan/ramachandran_p1
6. RizzolattiG.,CraigheroL.,The mirror neuron system,
(http://web.mit.edu/hst.722/www/Topics/Language/RizzolattiReview2004.pdf), annual Review of
7. Down regulation of electroacupressure at ST36 on TNF alpha …., World J Gastroenterol 2003
May;9(5):1028-1033 Li Tian, Yu-Xin Huang,
8. Mostofsky, SH, Powell SK, Simmonds DJ, Goldberg MC, Caffo B, Pekar JJ. (2009). Decreased
connectivity and cerebellar activity in autism during motor task performance. Brain ()132, 2413-25.
9. St Laurent, G., III & Wahlestedt, C. Noncoding RNAs:.…,Trends Neurosci. 30, 612–621 (2007).
10. Georges St Laurent111 etal ‘Expression of non coding RNA…, Nature Medicine,Volume14,Number7, July