Alterations of Mitochondrial Biology in the Oral Mucosa of Chilean Children with Autism Spectrum Disorder (ASD)

Autistic Spectrum Disorder (ASD) is characterized by the impairment of socio-communicative skills and the presence of restricted and stereotyped behavior patterns. Recent researches have revealed the influence of mitochondrial physiology on the development of ASD. Several research groups have identified defects in respiratory complexes, coenzyme-Q10 deficiency, increased oxidative damage, decreased of superoxide dismutase (SOD2). A study on the influence of mitochondrial physiology on the development of ASD can provide new alternatives and challenges. That is why we set ourselves the general objective to initiate studies of mitochondrial physiology in Chilean children with ASD. A sample of oral mucosa was collected in a group of 12 children diagnosed with ASD and 12 children without ASD. In children with ASD, we found a significant increase in mitochondrial DNA levels. Likewise, in these children, an increase in the protein oxidation was observed. Finally, a downward trend in the expression of the HIGD2A and SOD2 genes was observed, while DRP1, FIS1, MFN1, MFN2, and OPA1 gene expression show an upward trend. The increment of mitochondrial DNA, high oxidative stress, and high expression of the MFN2 gene could help as a scanner of the mitochondrial function in children with ASD.


Introduction
The Autism Spectrum Disorder (ASD) is an alteration of the neurodevelopment with an impact throughout life that present impediments in the social interaction, communication, imagination, and repetitive behaviors, that could affect 1 to 2% of children [1][2][3][4][5][6][7][8][9]. Early diagnosis and intervention allow a better outlook since the progress is significant in linguistic development, social ability, and the possibility of higher behavioral adaptation [10,11]. Therefore, it is essential to advance in the study of molecular mechanism of ASD.
The ASD has a multifactorial etiology in which the oxidative stress is one of the pathology mechanisms widely described [12]. In the last years, many investigations have focused on the study of mitochondrial dysfunction in children with ASD [13][14][15][16][17]. The first studies that associated the alteration of bioenergetics metabolism with the ASD reported high levels of lactic acid, pyruvate, and serotonin [13]. Subsequently, defects were identified in every I, II, III, IV, and V respiratory complex, deficiency of coenzyme Q 10 (CoQ10), defects in mitochondrial genes and nuclear genes [9,[14][15][16]. dynamics, can be critical for understanding how to approach the treatment of children with ASD. The objective for the present work has initiated the study about some aspects of mitochondrial physiology in the oral mucosa of 6-to 13-year-old Chilean children with ASD. From oral mucosa, we isolated DNA, RNA, and proteins, to evaluate the Ala16Val-SOD2 polymorphism, mitochondrial DNA, the HIGD2A, SOD2, DRP1, FIS1, MFN1, MFN2, and OPA1 gene expression. Furthermore, we evaluated the oxidation of total proteins and the expression of proteins of the respiratory complex. Our study reported for the first time a high quantity of mitochondrial DNA in oral mucosa as another biomarker of ASD. Moreover, for the first time, the analyses show higher protein carbonylation and the immunodetection of the respiratory complexes in the buccal mucosa of ASD children. In addition, we showed that mitochondrial dynamics-related genes are more expressed in ASD children.

Experimental Design
The research was a non-experimental, descriptive transactional and non-probabilistic design. The data were collected in a single moment, in a single time. The purpose was to measure and describe variables independently. The focus of the research was quantitative. The data collection to test the hypothesis was based on numerical measurement and statistical analysis to establish patterns of molecular behavior.

Subjects
Cheek swabs were collected from both ASD patients (n = 12, 11 males and 1 female) and healthy controls (n = 12, 9 males and 3 females). The age of the children ranged from 6 to 13 years who achieved a level of understanding of basic instructions by means of pictograms. The selection of children with ASD was carried out at the Magdalena Ávalos Cruz Special School, ASPAUT, in the San Miguel district. All the children who attend this school are diagnosed by the medical professional with Autism Spectrum Disorder. For the selection of healthy control children, the Monte Olivo School in Puente Alto district was chosen. All the children who attend this school are children who access regular school curriculums and do not have a previous medical diagnosis, without any signs or symptoms of neurological disease. In both selections, informed consent for the different samples was elaborated. The documents were authorized by the Institutional Ethics Committee of the Universidad Autónoma de Chile (Act N • 0135-16).

Oral Sample Preparation
The sample collection process was carried out by a non-invasive procedure, where a sample of oral mucosa was taken. Each child was given 10 mL of sterile saline (0.9% NaCl), indicating that mouthwash should be done by sucking the cheeks and not swallowing the saline solution. The indications were accompanied by pictograms. After the mouthwash, the saline was returned to the container and centrifuged at 6000 rpm for 10 min at 4 • C. The cellular precipitate containing the oral mucosa was resuspended in 1 mL of TRIzol™ Reagent (Invitrogen) and stored at −20 • C until the isolation of RNA, DNA, and proteins.

Isolation of RNA, DNA, and Proteins
From oral mucosa RNA, DNA, and proteins were extracted with TRIzol™ Reagent (Invitrogen), according to the manufacturer´s protocol. The quantification of the quantity and quality of the RNA, DNA, and proteins isolated in each sample was carried out by measuring the absorbance (A) using the Infinite M200 Pro spectrophotometer (TECAN). The quality and integrity of the DNA and RNA were evaluated on 1.5% agarose gels.

Reverse Transcriptase and Quantitative Real-Time PCR (qRT-PCR)
Complementary DNA synthesis (cDNA) was performed from 2 µg of purified RNA using the RevertAid First Strand cDNA Synthesis kit (Thermo Scientific, Waltham, MA, USA) which in addition to the reverse transcriptase (RevertAid M-MulV RT), includes a mixture of oligo (dT) and random hexamer starters optimizing the synthesis of cDNA from RNA. For a final volume of 40 µL per reaction, 2 µg of RNA contained in 20 µL of nuclease-free water, 2 µL random hexamer starter, 2 µL of oligo (dT) was prepared. It was incubated for 5 min at 65 • C, and the following mixture was added; 8 µL Buffer 5×, 2 µL RiboLock RNase Inhibitor, 4 µL mixture of deoxyribonucleotide triphosphates (dNTPs), 2 µL reverse transcriptase. The mixture was homogenized gently and centrifuged briefly. Later in the thermocycler, a 5 min incubation was scheduled at 25 • C followed by an incubation for 60 min at 42 • C and 5 min at 72 • C. The resulting cDNA was stored at −20 • C until used. For the relative quantification of cDNA levels by qPCR, the LightCycler ® 96-Real-time PCR system (Roche, Risch-Rotkreuz" Switzerland) with capacity for 96-well plates was available and the FastStart Essential DNA Green Master (Roche) kit based on the fluorophore SyBR Green. Data are presented as relative gene expression normalized to PPIA gene expression.

Tetra-Primer Amplification-Refractory Mutation System (ARMS)-PCR
Genotyping of the Ala16Val-SOD2 SNP polymorphism was performed by PCR using the DNA from the oral mucosa sample and a "Tetra-Primer ARMS-PCR" assay [36]. For this, the following primers were used: ARMS-SOD2-

Western Blot
For the immunodetection of respiratory complexes, 20 µg of total protein samples from oral mucosa were fractionated in 12% polyacrylamide gel with running buffer: twenty-five millimolar Tris, 190 mM glycine, 0.1% SDS, pH 8.3 was transferred onto a polyvinyl difluoride (PVDF) membrane using a wet transfer tank with the transfer buffer: twenty-five millimolar Tris, 190 mM glycine, 20% methanol, pH 8.3. The PVDF membrane was blocked for 1 h with 3% bovine serum albumin (BSA) in Tris-buffered saline with Tween 20 (TBST) buffer; 20 mM Tris pH 7.5, 150 mM NaCl, 0.1% Tween 20, at room temperature. Then, the PVDF membrane was incubated overnight at 4 • C with MitoProfile ® Total Oxidative Phosphorylation (OXPHOS) Human Western Blot (WB) Antibody Cocktail (ab110411 Abcam) used at 1/200 dilution in 1% BSA in TBST. The blot was rinsed three times for 5 min with TBST buffer. Primary antibodies were detected with the secondary antibodies: Horseradish peroxidase (HRP)-conjugated anti-mouse Immunoglobulin G (IgG) used at 1/300 in 1% BSA in TBST for 1 h at room temperature and the Chemiluminescent Western Blot Detection (ThermoFisher Scientific).

Protein Oxidation Detection
Total protein oxidation was measured with the Oxyblot Detection Kit (Millipore) according to the manufacturer's instructions. This measurement is based on the detection of carbonyl groups that are introduced into protein side chains when proteins are exposed to oxidative stress [37]. Briefly, total protein levels were quantified using the Bradford assay and using bovine serum albumin as a standard. The carbonyl groups in the protein side chains are derivatized to 2,4-dinitrophenylhydrazone (DNP-hydrazone) by reaction with 2,4-dinitrophenylhydrazine (DNPH). The protein derivatization reaction was performed with 15 µg of total protein. The DNP-derivatized proteins samples were fractionated in 12% polyacrylamide gel electrophoresis (PAGE) and then transferred onto a polyvinyl difluoride membrane using a wet transfer tank.

Statistical Analysis
Statistical analysis was performed with GraphPad Prism 6 software (GraphPAd Software, Inc., La Jolla, CA, USA). To compare the results of two variables that participated in the same experiment, it was analyzed by means of an unpaired student T-test. For the statistical analysis of the polymorphism the Chi-square test was used to evaluate the differential hypothesis between the two categorical variables ASD and HC, a contingency table was used where each variable was divided into different categories (A/A, A/V, and V/V alleles). Subsequently, the expected frequencies were calculated.

Subjects Clinical Data and Demographic Information
The age of the 12 ASD children, one female, and 11 males ranged from 7 to 12 years, with an average of 9.6 ± 1.9 years. In Table 1 presents the clinical data and demographic information of the children with ASD. The different therapies mentioned in Table 1, behavioral, communication, and sensory, are focused on Behavioral Programs that allow, through cognitive/behavioral strategies, the modification of maladaptive and disruptive behaviors with sensory and communicative nuances, giving greater comfort to the child and the family in activities of daily life.

Oral Mucosa mtDNA Levels in Chilean Children with ASD
Mitochondrial DNA copies are a very common marker in mitochondrial dysfunction, and alterations in mtDNA levels have been reported in autism [14,21]. However, there is no consensus on the different findings obtained by various research groups. Therefore, we set out to explore the levels of mtDNA in Chilean children with ASD. For this, extraction of total DNA from the oral mucosa was performed, and by qPCR, relative quantification of the mtDNA was carried out with a nuclear gene as a gene reference. Our results show that there is a significant increase in the mtDNA levels in the oral mucosa of Chilean children with ASD ( Figure 1). These results are consistent with those obtained by the research from other groups that indicate an increase in mtDNA copies and mtDNA deletions in children with ASD, which is correlated with mitochondrial dysfunction [14,21]. However, other groups have not been able to identify differences in mtDNA levels in ASD [18,23]. The increase in mtDNA levels could account for a compensatory mechanism to strengthen the different mitochondrial functions that govern cellular bioenergetics [14]. In addition, this condition is suggestive of alterations in the mitochondrial DNA replisome, where different proteins are involved than those used for nuclear DNA replication; the T7-Like Mitochondrial DNA helicase (TWINKLE) and the DNA Polymerase Gamma (POLγ), among others [38,39]. For mitochondrial DNA quantification, the mitochondrial gene tRNA-Leu (UUU) was amplified by qPCR, and the nuclear gene B2microglobulim amplification was used for the normalization. Each bar graph represents the mean ± SEM, n = 12 ASD and 12 HC children, analyzed by unpaired t-test (P < 0.05) and producing a P value = 0.0006 (***).

Oral Mucosa Total Protein Oxidation Levels and Western Blotting Analysis of Respiratory Complexes in Chilean Children with ASD
ASD has enhanced reactive oxygen species (ROS) generation due to alterations in the respiratory chain activities. Total protein oxidation in the oral mucosa of children with ASD was higher than healthy control children ( Figure 2). In addition, the protein expression of respiratory complexes was assessed by immunoblot. However, the protein samples were depleted, so only five samples from the control children and six samples from the ASD children were analyzed. The manufacturers of MitoProfile ® Total OXPHOS Human WB Antibody Cocktail (ab110411 Abcam) recommended to perform the immunodetection test of the respiratory complexes with 50 μg of protein, but we could For mitochondrial DNA quantification, the mitochondrial gene tRNA-Leu (UUU) was amplified by qPCR, and the nuclear gene B2-microglobulim amplification was used for the normalization. Each bar graph represents the mean ± SEM, n = 12 ASD and 12 HC children, analyzed by unpaired t-test (P < 0.05) and producing a P value = 0.0006 (***).

Oral Mucosa Total Protein Oxidation Levels and Western Blotting Analysis of Respiratory Complexes in Chilean Children with ASD
ASD has enhanced reactive oxygen species (ROS) generation due to alterations in the respiratory chain activities. Total protein oxidation in the oral mucosa of children with ASD was higher than healthy control children ( Figure 2). In addition, the protein expression of respiratory complexes was assessed by immunoblot. However, the protein samples were depleted, so only five samples from the control children and six samples from the ASD children were analyzed. The manufacturers of MitoProfile ® Total OXPHOS Human WB Antibody Cocktail (ab110411 Abcam) recommended to perform the immunodetection test of the respiratory complexes with 50 µg of protein, but we could only load 20 µg, for this reason, the quality of the immunoblot is low. Although, this result confirms the possibility of detecting proteins from respiratory complexes from samples of the oral mucosa, it is necessary to perform more analyzes that to obtain the levels of respiratory complexes in the oral mucosa, and, thus, consider them in the study profile of children with ASD ( Figure 3). only load 20 μg, for this reason, the quality of the immunoblot is low. Although, this result confirms the possibility of detecting proteins from respiratory complexes from samples of the oral mucosa, it is necessary to perform more analyzes that to obtain the levels of respiratory complexes in the oral mucosa, and, thus, consider them in the study profile of children with ASD ( Figure 3).  PAGE Coomassie blue staining of SDS-PAGE of derivatized proteins. (c) Densitometry quantification of protein carbonylation was made with the ImageJ software. Carbonylation of proteins was normalized by Coomassie blue staining. Each bar represents the mean ± SEM, analyzed by unpaired t-test. Healthy controls (HC), n = 11 and ASD children, n = 11.

The Presence of Ala16val-SOD2 Polymorphism in Chilean Children with ASD
The single nucleotide polymorphism C47T at the 6q25 locus of the gene encoding for SOD2 results in the substitution of the Ala16Val amino acid and in the partial retention of SOD2 in the mitochondrial intermembrane space, reducing enzymatic activity and increasing oxidative stress [19]. However, the variability of the results obtained by the different research groups who have studied how this polymorphism may affect SOD2 activity indicates that it may be due to differences in diet, environmental exposure, and additional genetic factors [40][41][42][43][44][45]. In this study, we analyzed the presence of Ala16Val-SOD2 polymorphism in Chilean children (Figure 4a,b). Seventy-five percent of the population of children with ASD (9 of 12) and 92% of the population of children with HC (11 of 12), could be identified as being heterozygous for the Ala16Val-SOD2 (A/V) polymorphism. No HC or ASD children were homozygous for the Ala16 (A/A) allele.
Regarding the Val16 allele (V/V), only one HC child was homozygous for it, and three children with ASD presented the polymorphism in homozygosis. When performing a Chi-square test, we found that these differences were not significant (Figure 4b,c). However, the number of children tested with ASD and HC is not enough to infer results. We were interested in analyzing the

The Presence of Ala16val-SOD2 Polymorphism in Chilean Children with ASD
The single nucleotide polymorphism C47T at the 6q25 locus of the gene encoding for SOD2 results in the substitution of the Ala16Val amino acid and in the partial retention of SOD2 in the mitochondrial intermembrane space, reducing enzymatic activity and increasing oxidative stress [19]. However, the variability of the results obtained by the different research groups who have studied how this polymorphism may affect SOD2 activity indicates that it may be due to differences in diet, environmental exposure, and additional genetic factors [40][41][42][43][44][45]. In this study, we analyzed the presence of Ala16Val-SOD2 polymorphism in Chilean children (Figure 4a,b). Seventy-five percent of the population of children with ASD (9 of 12) and 92% of the population of children with HC (11 of 12), could be identified as being heterozygous for the Ala16Val-SOD2 (A/V) polymorphism. No HC or ASD children were homozygous for the Ala16 (A/A) allele.
Regarding the Val16 allele (V/V), only one HC child was homozygous for it, and three children with ASD presented the polymorphism in homozygosis. When performing a Chi-square test, we found that these differences were not significant (Figure 4b,c). However, the number of children tested with ASD and HC is not enough to infer results. We were interested in analyzing the correlation of the A16VSOD2 polymorphism with the ASD severity levels. We observed that six of 12 (50%) of children with genotype A/V have a level II of severity, 17% (2 of 12) with genotype A/V have a level III severity, and 8% (1 of 12) present a level I of severity. On the other hand, 17% (2 of 12) with the V/V genotype presented a level II of severity and 8% (1 of 12) with a level III of severity. When performing a Chi-square test, we found that these differences were not significant (Figure 4d). correlation of the A16VSOD2 polymorphism with the ASD severity levels. We observed that six of 12 (50%) of children with genotype A/V have a level II of severity, 17% (2 of 12) with genotype A/V have a level III severity, and 8% (1 of 12) present a level I of severity. On the other hand, 17% (2 of 12) with the V/V genotype presented a level II of severity and 8% (1 of 12) with a level III of severity. When performing a Chi-square test, we found that these differences were not significant (Figure 4d). The observed genotypes were analyzed with a Chisquared (X 2 ) test with an alpha <0.05 statistical significance, obtaining a P value = 0.2231 n/s (nonsignificant). (d) The distribution of Ala16val-SOD2 polymorphism in the ASD severity level was analyzed with a Chi-squared (X 2 ) test with an alpha <0.05 statistical significance, obtaining a P value = 0.1897 n/s (non-significant).

Expression of the Genes Encoding for HIGD2A, SOD2, DRP1, FIS1, MFN1, MFN2 and OPA1 from the Oral Mucosa of Chilean Children with ASD
The SOD2 enzyme is a metalloenzyme containing manganese. It is the only mitochondrial enzyme that dismutate the superoxide anion to H2O2 and is very important in the antioxidant defense of the eukaryotic cells [12,18]. Alterations in SOD2 enzyme function have been associated with different neurobiological disorders as well as with autism [12,18].
The deficit of SOD2 results in a higher production of free radicals, which cause damage to mitochondrial activity [12]. Likewise, as oxidative stress progresses, energy production weakens, and there is an increase in excitotoxicity in autistic children, which damages different macromolecules and genetic factors. Among them, an increase in the peroxidation of lipids, higher protein, and DNA oxidation can be noted [12,46,47]. The observed genotypes were analyzed with a Chi-squared (X 2 ) test with an alpha <0.05 statistical significance, obtaining a P value = 0.2231 n/s (non-significant). (d) The distribution of Ala16val-SOD2 polymorphism in the ASD severity level was analyzed with a Chi-squared (X 2 ) test with an alpha <0.05 statistical significance, obtaining a P value = 0.1897 n/s (non-significant). The SOD2 enzyme is a metalloenzyme containing manganese. It is the only mitochondrial enzyme that dismutate the superoxide anion to H 2 O 2 and is very important in the antioxidant defense of the eukaryotic cells [12,18]. Alterations in SOD2 enzyme function have been associated with different neurobiological disorders as well as with autism [12,18].

Expression of the Genes
The deficit of SOD2 results in a higher production of free radicals, which cause damage to mitochondrial activity [12]. Likewise, as oxidative stress progresses, energy production weakens, and there is an increase in excitotoxicity in autistic children, which damages different macromolecules and genetic factors. Among them, an increase in the peroxidation of lipids, higher protein, and DNA oxidation can be noted [12,46,47].
As an initial overview of the status of the SOD2 enzyme in Chilean children with ASD, we decided to analyze the expression of the gene that encodes for SOD2 in the oral mucosa through RT-qPCR.
Our results indicate, a downward trend in the expression of the SOD2 gene in Chilean children with ASD (Figure 5a), which suggests that there is no alteration in the transcription of SOD2 in ASD. We analyzed the correlation between the expression of the SOD2 gene and the Ala16Val-SOD2 polymorphism; however, no correlation was observed (HC: R 2 = 0.0473; ASD: R 2 = 0.0889), indicating that the polymorphism does not affect the transcription of the SOD2 gene. The PPIA gene was taken as a reference during the qPCR procedure. Each bar graph represents the mean ± SEM, n = 10 ASD and 11 HC children, analyzed by unpaired t-test (** P < 0.05).
The mitochondrial protein HIG2A promotes cell survival during hypoxic conditions [48]. Furthermore, HIG2A mediates the assembly of the complexes of the electron transport chain in respiratory supercomplexes or respirasomes [31,49]. These respirasomes enhance the transfer of electrons, channel substrates, stabilize protein complexes, enhance enzymatic activity, and reduce oxidative stress [32].
Our research group has focused on understanding the role of HIG2A as it relates to mitochondrial physiology. So far it has been established that the HIGD2A gene-encoding for the HIG2A protein-is upregulated by hypoxia and by glucose availability. The HIGD2A gene exhibits a differential expression based on tissue and age [33]. The absence of HIGD2A alters mitochondrial dynamics [33]. Within the framework of this study, we evaluated the expression of HIGD2A in the oral mucosa of Chilean children with ASD. It is shown that there is a downward trend in the expression of the HIGD2A gene (Figure 5b), suggesting that the transcription of HIGD2A gene does not have a direct impact on the development of ASD.
The assessment of the genes involved in mitochondrial dynamics, such as MFN1, MFN2, OPA1 (mitochondrial fusion) and DRP1 and FIS1 (mitochondrial fission) was performed in the oral mucosa of children with ASD. It is shown that there is an upward trend in the expression of the MFN1, OPA1, DRP1 and FIS1 genes in the oral mucosa in children with ASD (Figure 5c-e,g) while a significant increase in the expression of the MFN2 gene (Figure 5f) was shown.

Prediction of the Association of the Analyzed Genes with the Genetic Bases of ASD
With the purpose of understanding the results of the gene expression reported above, we review the genetic associations of these genes with the computational tool (http://asd.princeton.edu) from the Lewis-Sigler Institute of Integrative Genomics at Princeton University (USA) that can predict the association of each of the genes of the human genome with the genetic basis of ASD. Among 25,825 genes, SOD2 is located at position 12001, suggesting that the probability of association with ASD is low (Figure 6a) [50]. Upon analyzing the HIGD2A gene with the computational tool, we found that HIGD2A is ranked at position #21384 out of 25825, which suggests that the likelihood of association with ASD is very low (Figure 6b) [50].
The computational analysis of the genes involved in mitochondrial dynamics, such as MFN1, MFN2, OPA1 (mitochondrial fusion) and DRP1 and FIS1 (mitochondrial fission) suggests a higher likelihood of association with the development of ASD. In particular, OPA1 which is located at position #3271 of the ranking of genes associated with the development of ASD (Figure 6g), while DRP1 is at position #2558 and FIS1 at position #2822 (Figure 6c,d). In turn, these genes are related in the network with other genes that have a higher likelihood of association with the development of ASD [50]. MFN1 is at position #22827 and MFN2 is at position #11607. However, this last MFN2 gene was the only one that showed a significantly higher expression in children with autism. When we review the networks of genetic associations of the MFN2 gene (Table 2), the integration with genes that are more likely to be associated with the development of autism is evident.
The use of the computational tool from the Lewis-Sigler Institute for Integrative Genomics at the University of Princeton (USA) will be beneficial for future studies of different genes that might be involved in the development of ASD.

Discussion
The Chilean Association of Parents and Friends of the Autistic, ASPAUT, was one of the first centers created in Chile (1983) specialized in the treatment of persons with ASD, delivering pedagogical and therapeutic attention to achieve an integral development of persons with autism. Its Integral Diagnostic Team mentions that the diagnosis of ASD is delayed because families have the first complex experiences with children when they face the educational system and it is the pedagogues who refer them to suitable specialists (Table 1). Currently, only medical professionals (Neurologists and Child Psychiatrists) are responsible for this prescription and depend on their skills on the behavioral characteristics of Autism Spectrum Disorder. The generation of public policies on early detection is necessary for the timely identification of each child with ASD, and it is essential that the treatment of children begin at an early age to achieve a significant impact on their evolution. A study conducted in Chile by Hartmann et al. showed associations between ASD traits, physician support, community engagement, stigma, and family stress, highlighting "the importance of future research to better understand and treat Latin American children with ASD and their families" [51]. This study was an exploratory study carried out by a non-invasive sampling of the oral mucosa, which is a multi-stratified flat epithelial tissue that originates in the ectoderm, as does the nervous system [24].
In this study, we show that in the oral mucosa of Chilean children with ASD have a significant increase in mtDNA levels ( Figure 1). These results agree with those obtained by the investigation of other groups that indicate an increase in mtDNA copies, and mtDNA deletions in children with ASD, which correlates with mitochondrial dysfunction [14,21]. However, other groups have not been able to identify differences in mtDNA levels in ASD [18,23]. The increase in mtDNA levels could account for a compensatory mechanism to strengthen the different mitochondrial functions that control cellular bioenergetics [14,21]. Furthermore, this condition is suggestive of alterations in the mitochondrial DNA replisome where different proteins are involved than those used for the replication of nuclear DNA, such as the TWINKLE helicase, the DNA polymerase γ, POLγ, and TFAM (Transcription Factor A, Mitochondrial), among others [38,39]. TFAM is an activator of mitochondrial transcription, it also functions as a histone-like in the packaging of mtDNA and as mtDNA levels follow TFAM levels [52,53].
An increase in the number of copies of mitochondrial DNA has been associated with good prognosis in glioma patients [54]. In addition, an improvement in survival in experimental conditions where human TFAM is overexpressed in an animal model of myocardial infarction [55]. The overexpression of human TFAM in mice diminishes the memory deterioration associated with aging [56]. The consequences of a high number of mtDNA copies in vivo are nucleoids enlargement, defects in transcription, accumulation of deletions in mtDNA, and deficiencies in the respiratory chain [57].
Several studies have reported oxidative stress in postmortem brain samples from patients with ASD, demonstrating oxidative damage to DNA, lipids, and proteins, and also reported alterations in the enzymatic activity of redox metabolism and lower concentrations of reduced glutathione (GSH) [12,58]. In this study, we evidenced higher protein carbonylation in the total protein of oral mucosa of ASD children ( Figure 2). Furthermore, we show the presence of the respiratory complexes in the total protein of oral mucosa (Figure 3). Remarkably, Tang et al. (2013) in ASD postmortem brains in the BA21 temporal cortex, revealed significantly lower levels of respiratory complex I and decreased complex I and IV activities [18].
SOD2 is a metalloenzyme that contains manganese; it is the only enzyme in the mitochondria that converts superoxide to H 2 O 2 [46,47], therefore, it has a fundamental role in dealing with the free radicals generated by the electron transport chain. In the gene coding for SOD2 (locus 6q25), the C47T single nucleotide polymorphism that results in the substitution of the amino acid Ala16Val has been studied extensively. Where the change from nucleotide cytosine (C) to thymidine (T) generates change in the amino acid sequence of SOD2 at the position of residue 16 of alanine (Ala) changes to valine (Val) occur. The Ala16Val-SOD2 polymorphism has been suggested to affect the transport of the SOD2 enzyme to the mitochondrial matrix as a result of modification to the mitochondrial targeting sequence. The 16Ala variant reaches the mitochondrial matrix, while the 16 Val variant is partially retained in the intermembrane space of the mitochondria. The presence of valine in position 16 reduces enzymatic activity and increases oxidative stress [19]. The altered cellular localization of SOD2 could be part of the neurodegeneration process that occurs in ASD. An exploratory study of the nutritional and metabolic biomarkers in ASD carried out by the Department of Integrative Medicine of the University of Kansas (USA), analyzed nine patients with ASD, of which four presented the Ala16Val polymorphism, three of them were homozygous (V/V) [20]. In our study three of twelve ASD children were homozygous (V/V) for the Ala16Val polymorphism while one of twelve HC children were homozygous ( Figure 4). However, the number of children tested with ASD and HC is not enough to infer results. In addition to the exploration of Ala16Val-SOD2 polymorphism, we also studied the SOD2 gene expression in the oral mucosa of ASD children, which shows a downward trend (Figure 5a). In autism postmortem brains, the Thalamus region-specific present alteration with SOD2 gene shows upregulated expression [59]. While, in ASD postmortem brains in the BA21 temporal cortex presented a decreased mitochondrial antioxidant enzyme SOD2 [18].
Despite the known poor overall correlation between mRNAs and their protein products, differential mRNA expression studies implicitly assume that changes in mRNA expression have biological meaning, most likely mediated by corresponding changes in protein levels [60]. Transcript levels of MFN1, MFN2, OPA1, DRP1, and FIS1 by themselves are not enough to predict MFN1, MFN2, OPA1, DRP1, and FIS1 protein levels in many scenarios and to, thus, explain genotype-phenotype relationships, however, high-quality data quantifying different levels of gene expression are necessary for the comprehensive understanding of MFN1, MFN2, OPA1, DRP1, and FIS1 biological processes in the development of ASD. In this study, there is an upward trend in the expression of the MFN1, OPA1, DRP1, and FIS1 genes in the oral mucosa in children with ASD (Figure 5c-e,g) while a significant increase in the expression of the MFN2 gene (Figure 5f) was shown. In autism postmortem brains, the Anterior Cingulate Gyrus region-specific MFN2 gene shows downregulated expression [59]. The expression of DRP1 (DNML1) was reduced in the Anterior Cingulate Gyrus and Motor Cortex region-specific alteration of autism patients [59]. Tang et al. (2013) show altered mitochondrial dynamics proteins in the brain of ASD patients, with decreased levels of mitochondrial fusion proteins; MFN1, MFN2, OPA1, and an increased level of mitochondrial fission proteins; DRP1 and FIS1 [18]. The treatment with butyrate (1 mM) of lymphoblastoid cell lines (LCL) derived from ASD patients increased the expression of genes involved in mitochondrial fission (DRP1, FIS1) [30]. Butyrate is a ubiquitous short-chain fatty acid principally derived from the enteric microbiome, and it is possible that the oral microbiome butyrate conditions the expression of the genes of the oral mucosal cells and this could explain the differences concerning that described in the brain.

Conclusions
There is a significant increase in mtDNA levels in ASD children in relation to HC children (healthy controls).
The oral mucosa sample is a reliable tissue for assessing mtDNA levels in children with ASD. Furthermore, from this sample, it was possible to extract RNA for the evaluation of gene expression and proteins, which opens the door for new research possibilities in ASD or other diseases that are difficult to have access to or which make for risky blood samples.
Since the oral mucosa has an embryonic origin shared with that of the nervous system, it is a good alternative for exploratory studies on gene expression analysis associated with the development of ASD in humans.
ASD is a complex neurodevelopmental disorder with a strong genetic basis. Despite this, only a small fraction of the genes that can be the cause of it are known and supported by strong genetic evidence from sequencing studies. This study constitutes a small contribution towards the understanding of the incidence of mitochondrial biology in the development of ASD.