Yao Y, O’Sullivan B, Mondal TK, Uddin MN, Manley K, et al. (2025) Mitochondrion-Mediated Metabolism and Microbiome Biodiversity Influence Autism-alike Behaviors. J Neurol Transl Neurosci 10(1): 1100

Background: The BTBR T+Itpr3tf/J (BTBR) strain has autism spectrum disorder (ASD)-like behaviors, which has been associated with mitochondrial dysfunction. Therefore, a new mouse strain was developed. The BTBR-mtB6 strain has the nuclear genome of BTBR but mitochondria from C57BL/6J (B6) mice,hich have normal behaviors and immunity. The BTBR-mtB6 strain had more normal behavios and immunity. Therefore, the mechanisnisms associated with the improvements were investigated.

Aim: Since replacement of mitochondria in BTBR mice improved behavior and some immune differences, the associated mechanisms were researched.

Methods: Since mitochondria functions affect gut microbiota and metabolomics, the bacteria in fecal samples and metabolites in the blood and organs including the brain were investigated. Microbiome sequences were retrieved from Illumina BaseSpace. Genetic and molecular changes incuding metabolomics were assayed.

Results: The fecal microbiomes of BTBR, B6 and BTBR-mtB6 mice were different from each other. The serum and brain cholesterol levels were intermediate between males of the BTBR and B6 strain. The liver PPARg level also was intermediate between the BTBR and B6 strain which may relate to the BTBR-mtB6 intermediate amount of lipid in the liver. The BTBR-mtB6 mice also had an intermediate number of T cells in the white adipose tissue compared to the BTBR and B6 mice. Complex IV of the ETC in the liver was slightly lower in the BTBR-mtB6 mice than B6 mice. The BTBR-mtB6 strain lost production of IgG to brain antigen that is observed in BTBR mice.

Conclusion: The mitochondrial shift is shown to affect fecal microbiota, mitochondrion-dependent metabolism affecting lipid accumulation, the levels of cholesterol in the brain and serum, and brain expression of myelin basic protein (MBP) and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase), which improves myelination in BTBR-mtB6 brains. The metabolite and microbiome differences likely relate to mitochondrial/nuclear differences affecting metabolism, immunity, and behavior.

Autism spectrum disorders (ASD) are characterized by behavioral deficits in sociability and communication and almost one-third of children with ASD experience metabolic abnormalities [1]. Mitochondrial disease and aberrant redox metabolism are associated with metabolic changes, which have been reported with ASD [2-6]. BTBR T+ Itpr3tf/J (BTBR) mice with ASD-like behaviors have metabolic disturbances affecting behavior and immunity [7-9]. Since mitochondria are “metabolic hubs” having a major influence on the metabolism of all organs [10], BTBR mice had their mitochondria replaced with mitochondria from C57BL/6J (B6) mice, which have normal behavior and immunity. The BTBR-mtB6 strain was derived through the initial breeding of a BTBR male with a B6 female; after that, female offspring were bred with a BTBR male for twelve generations. Although the new BTBR-mtB6 strain was confirmed to have the nuclear genome of BTBR and B6 mitochondria, there may be mitochondrial effects on epigenetics and nuclear gene effects on the mitochondria [10-12]. The BTBR-mtB6 strain had some improved behaviors, higher levels of glutamate and astrocytes in the brain, and less neuroinflammation than the BTBR strain [13].

BTBR T+Itpr3tf/J (BTBR) mice have larger bodies with more fat than C57BL/6J (B6) mice, which have normal behaviors. The BTBR strain also has immunity that differs from the B6 mouse strain, which has normal immunity. Since mitochondria utilize lipids and sugars for their energy support to provide ATP for all cells, molecular differences of mitochondria and nuclei of BTBR, B6, and BTBR-mtB6 mice have been evaluated. BTBR-mtB6 mice have the nuclear genome of BTBR and mitochondria of B6 mice.

In the current study, we researched lipid metabolism, including absorption, transport, and metabolism along with the fecal microbiota of B6, BTBR, and BTBR-mtB6 mice. The gut microbiome has been implicated in ASD [14-16], and the gut microbiome has been linked to mitochondrial function and reactive oxygen species (ROS) [17-19]. Lessening oxidative stress (OS) has been reported to attenuate the autism-like behavior of BTBR mice [20]. BTBR mice have an elevated level of OS, which affects immunity and brain functions [21,22]. Since the BTBR- mtB6 mice had less ASD-like behavior than the BTBR mice [13], we investigated whether replacing the mitochondria of BTBR with mitochondria from B6 mice also affected their gut microbiome, genes, and proteins known to affect metabolites and brain functions. Gender differences were considered since sex influences metabolic processes and diseases [23-25], as well as metabolic influences on adipose tissue [26], and neuropsychiatric disorder [27]. The metabolic differences of males and females of the three mouse strains were assayed. Although mitochondrial influences on metabolism differentially affect the brain development tempo of mice compared to humans [28], mitochondria influence brain development and functions throughout life for all species. The mitochondrial shift is shown to affect fecal microbiota, mitochondrion- dependent metabolism affecting lipid accumulation, the levels of cholesterol in the brain and serum, and brain expression of myelin basic protein (MBP) and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase), which improves myelination in BTBR-mtB6 brains. The metabolite and microbiome differences likely relate to mitochondrial/ nuclear differences affecting metabolism, immunity, and behavior. The results and conclusions of this ASD mouse model add to our knowledge of human studies researching ASD.

Mice

BTBR and B6 male and female mice were purchased from Jackson Laboratories (Bar Harbor, ME, USA) and maintained in the Wadsworth Center Animal Facility (room temperature: 22 ± 2°C; humidity: 50 ± 10%; lights on 7 AM-7 PM). The BTBR-mtB6 strain was developed as previously described [13]; the conplastic BTBR-mtB6 strain was developed by serial backcrossing to restore nuclear DNA to that of the BTBR strain with the mitochondrial parent always being the female B6 mouse during the process. After the twelfth generation, the male and female BTBR-mtB6 minimum of 10 backcross generations to develop a new conplastic strain. Adult mice at 10 ± 2 weeks old were used in this study. Tissues were collected from mice perfused as previously described [8,21]. All mice were provided with food and water ad libitum. The three mouse strains were housed in the same room, fed the same food (Prolab®IsoPro®RMH 3000 LabDiet, from ScottPharm, Marlborough, MA), and had cages changed with the same bedding at the same frequency. All animal experiments complied with and followed the National Research Council’s Guide for the Care and Use of Laboratory Animals. All procedures described involving live vertebrate animals have been reviewed and approved by the Wadsworth Center’s IACUC; IACUC approval for the study (protocol # 21-278) was received on 07/19/2021. The Wadsworth Center maintains a PHS-assured and AAALAC-accredited animal use program (Animal Welfare Assurance Number D16-00115), which is largely based on the guidelines outlined in the 8th edition of the Guide for the Care and Use of Laboratory Animals.

Fecal Microbiome DNA Isolation and PCR Amplification

Bacterial DNA was extracted from frozen fecal samples using the Shoreline Complete Lyse and Purify Kit (Shoreline Biome, Farmington, CT) according to manufacturer protocols [29]. For each mouse, approximately 5 mg of feces were mixed with 50 µL of nuclease-free water (Invitrogen, Waltham, MA) and 50 µL of 0.4 M KOH (Fisher Scientific, Waltham, MA) in wells containing dried Lysis-1. Samples were heated at 95°C for 5 min to lyse cells and afterward spun to pellet fecal debris; 50 µL of supernatant was added to 50 µL of purification buffer containing magnetic beads. Pelleted beads were washed with 70% ethanol, and DNA was eluted into 200 µL of TE Buffer (Thermo Scientific, Waltham, MA).

The V4 region of the 16S rRNA gene was amplified with the Shoreline Complete V4 Amplify Set A Kit (Shoreline Biome, Farmington, CT). The extracted DNA (10 µL) and 10 µL of 2X PCR Premix were added to wells containing barcoded primers. PCR was performed according to the manufacturer’s protocol. The PCR results were analyzed and pooled using the previously described (Qiagen, Hilden, Germany) methodology [30]. Samples were screened using the Fast Analysis kit (Qiagen, Hilden, Germany) and were analyzed with the ScreenGel software (Qiagen, Hilden, Germany); 5 µL of each sample was pooled together and purified with the GeneRead Size Selection Kit (Qiagen, Hilden, Germany); 80 µL of the library was added to 320 µL of Buffer SB1 (Qiagen, Hilden, Germany) and applied to a MinElute Spin Column (Qiagen, Hilden, Germany). The sample was centrifuged at 20,000 x g at 4°C for 1 minute. The bound DNA was rinsed with 700 µL of 80% ethanol and centrifuged at 20,000 x g at 4°C for 1 minute twice. After repeating the centrifugation step to remove any remaining ethanol, the DNA was eluted into a fresh tube with 90 µL of Buffer TE (Qiagen, Hilden, Germany) by centrifuging at 20,000 x g at 4°C for 1 min. Buffer SB1 (360 µL) was added to the eluted DNA and was bound to the spin column by centrifuging at 20,000 x g at 4°C for 1 min. The ethanol rinse was repeated, and the purified library was eluted into 50 µL of Buffer EB (v) by centrifuging at 20,000 x g at 4°C for 1 minute. Sequencing was performed with a MiSeq Reagent Kit V2 (500-cycles, Illumina, San Diego, CA) on an Illumina MiSeq sequencer by the Microbial Analysis, Resources, and Services facility at the University of Connecticut.

Bioinformatic Analysis of Microbiome Samples

Demultiplexed sequences were retrieved from Illumina BaseSpace and imported into QIIME2 ver. 2021.8 [31]. DADA2 was used to call amplicon sequence variants (ASVs) [32]. Taxonomy was assigned using the Silva database version 138. QIIME2 was used to create a phylogenetic tree with use of MAFFT to align sequences and FastTree to construct the tree. Data was then imported into R, and the Decontam package was used to remove contaminants found in the negative controls [33]. Phyloseq was used to calculate alpha and beta diversity metrics [34]. ANCOM-BC was used to identify taxa that were differentially abundant between different mouse strains [35]. Samples were rarefied to a depth of 9,127 reads.

Isolation of small intestinal epithelial cells and hepatocytes

Small intestinal epithelial cells (SIECs) were collected from B6, BTBR, and BTBR-mtB6 male and female mice after intracardial perfusion with warm phosphate-buffered saline (PBS) of 50 ml/mouse following the method previously reported [21]. The small intestines were excised, placed in ice-cold PBS (pH 7.2), and washed with ice-cold buffered saline to remove intestinal constituents. Then the emptied small intestine was cut longitudinally, and immersed in PBS containing 1.5 mM EDTA, 3 U/ml heparin, and 0.5 mM dithiothreitol for 20 min to loosen the small intestine mucosa. SIECs were scraped, placed in PBS, and pelleted by spinning at 2,000 g for 10 min at 4°C. The pelleted cells were washed twice with PBS and immediately used for gene and protein detections. Hepatocytes were collected and purified separately after inferior vena cava perfusion as described [36]. Briefly, the livers were perfused with PBS, cut into small pieces, placed in 5 ml complete medium containing collagenase 150 µl (stock 10 mg/ml) and DNase I 100 µl (stock 10 mg/ ml), incubated for 1 h at 37°C with shaker, and then poured the mixture through 100 µM nylon mesh and collected the cells, and spun down to remove the tissue debris at 30 g for 5 min at 4°C. The supernatant spun down again at 350 g for 10 min at 4°C. The cell pellet was mixed with 1 ml 33% Percoll, and centrifuged at 4000 rpm, 25 min at 20°C. The supernatant contained hepatocytes while the pelleted cells were liver immune cells.

Real-time PCR for gene detections

Qiagen miRNeasy Mini Kit (Cat. No. / ID: 217004) was used to isolate and purify total RNA including small RNAs from tissues. SIECs and Hepatocytes (no more than 1 x 107 cells) from B6, BTBR and BTBR-mtB6 mice were collected as described earlier and were added to 700 µl QIAzol Lysis Reagent (QIAGEN), samples were homogenized using a syringe and needle then vortexing for at least 1 min, incubated for 5 min at room temperature (RT), 140 µl chloroform was added and the mixture was centrifuged for 15 min at 12,000g at 4°C. The upper aqueous phase was transferred to a new collection tube and washed with 100% ethanol, 700µl RWT buffer and 500 µl RPE buffer twice. OD260/OD280 and OD260/ OD230 using a NanoDrop Spectrophotometer, together with RNA electrophoresis were utilized to detect the RNA purity. Collected RNA (1µg) was used for cDNA reverse transcription then cDNA as template with 5 µl PowerUp SYBR Green Master Mix (Thermo Fisher Scientific) for 10 µl Real-time PCR system and 1 µl primer for each gene detection assay (Applied Biosystems 7500 Fast Real-time PCR system). Relative expression data were calculated using 2-??Ct, and β-actin was used as reference. Primers for each gene: nuclear receptor subfamily 1 group H member 3 (Nr1h3); NPC1 like intracellular cholesterol transporter 1 (NPC1L1), acetyl-CoA acetyltransferase 2 (Acat2), ATP binding cassette subfamily G member 5 (Abcg5) and member (8Abcg8), and actin beta (Actb) are listed (Table 1).

Table 1: Primers for detected genes

Mitochondrial DNA/Nuclear DNA (mtDNA/nDNA) Ratio Detection

An experiment for the detection of mitochondrial numbers SIECs and livers was performed as previously reported [8]. NADH dehydrogenase 1 (ND1) and hexokinase isoform (HK2) were set as the mtDNA and nDNA markers, respectively. The ratio of mtDNA/nDNA was calculated by comparison of Ct values of mtDNA and nDNA PCR products.

Western Blot detection

Perfused tissues from B6, BTBR and BTBR-mtB6 mice were cut into small pieces in ice-cold lysis RIPA buffer (60µl/mg tissue). Western blot detection was performed as previous described [8]. Briefly, 20 µg total protein mixed with NuPAGE™ LDS Sample Buffer (4X) plus 2-mercaptoethanol was heated at 95°C for 10 min and added onto 12.5% SDS-PAGE gels. After electrophoresis, proteins in the gel were transferred to PVDF membrane and blocked with 5% fat free milk for 1h at RT. The membrane was incubated with mouse monoclonal anti-aquaporin9 (AQP9) (sc-74409, 1:1000, Santa Cruz Biotechnology, Inc), rabbit anti-fatty acid binding protein 1 (FABP1) pAb (A5311, 1:1000, ABclonal), anti-3-hydroxy- 3-methylglutarlyl-CoA reductase (HMGCR) (MA5-32521, 1:800, Invitrogen), mouse monoclonal anti-myelin basic protein (MBP) (F-6) (sc-271524, 1:1000, Santa Cruz Biotechnology, Inc), or mouse monoclonal anti-2′,3′-cyclic nucleotide-3′-phosphodiesterase (CNPase) (H-2) (sc- 166558, 1:800, Santa Cruz Biotechnology, Inc) in TBST buffer at 4°C overnight. After washing in TBST buffer three times, the membrane was incubated with goat anti-mouse or rabbit IgG H&L (HRP) (ab205719 or ab205718, 1:5000; Abcam) in TBST buffer at RT for 1h on an orbital shaker, and Pierce™ ECL Western Blotting Substrate was added for 1 min and exposed accordingly. Images were obtained with LAS-3000 Imaging System (Fuji-Film). Anti-β-actin (MA5- 15739-HRP, 1:5000, Invitrogen) was used as reference.

Flow Cytometric analysis of peroxisome proliferator-activated receptor gamma (PPARγ)

Livers from perfused mice were removed and treated as decribed earlier to obtain dispersed cells. Cells were then counted and adjusted to 10 x 106 cells/ml; 100 µl of cells were used for staining. Cells were stained for surface antigens with antibody (CD45) for leukocytes, washed and then fixed and permeabilized with fixation/ permeabilization buffer followed by intracellular staining with anti-PPARγ antibody (E-8) PE (sc-7273 PE, Santa Cruz Biotechnology, Inc) conducted in the presence of permeabilization buffer. Data were attained by immunophenotyping with a FACSCanto flow cytometer (BD Biosciences). Cytometric results were analyzed by Flow Jo-V10 software.

Oil red O staining of liver sections

Livers were collected, embedded in Fisher Healthcare™ Tissue-Plus™ O.C.T. Compound and frozen on dry ice. Frozen sections were cut at 20 µm and collected in PBS w/o azide. Oil Red O (ORO; Sigma-Aldrich O0625-25G) at 0.5% was prepared in isopropanol (250 mg into 50 ml) Slides were fixed in formalin, briefly washed with running tap water 5 min then rinsed with 60% isopropanol. Slides were stained with freshly prepared Oil Red O working solution for 15 min at RT, rinsed with 60% isopropanol then distilled water. Slides were mounted in glycerine jelly. Photos were taken using Nikon 50i w/Color camera after being mounted in aqueous mountant. For the quantification of images, we used ImageJ software as described (https:// www.youtube.com/watch?v=gxnWYhZ0z4k) with the aid of our Advanced Light Microscopy & Image Analysis Core.

Serum and brain cholesterol detection

Serum and brain cholesterol levels in three groups of male and female mice were detected using Amplex™ Red Cholesterol Assay Kit (Invitrogen™, Catalog number: A12216). Briefly, for serum, 2.5 μL of the serum samples were diluted (20 times) in 48 μL of the 1X Reaction Buffer. For the brain, after perfusion, mice brain was rinsed with 1X PBS (pH7.4), homogenized using cell lysis buffer (2mg/100μL, 150 mM sodium chloride, 1.0% Triton X-100,

0.1 M NaOH, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8.0) with a PX-SR-50-E homogenizer at 1,200 rpm for 30 min then centrifuged at 10,000 g for 20 min. After that, the supernatant was collected, diluted, and added into 1X Reaction Buffer provided in the kit for a total 50 μL volume at room temperature. 1X Reaction Buffer without cholesterol as a negative control while diluting the 20 mM H2O2 working solution to 10 μM in 1X Reaction Buffer as a positive control. 50 μL of the Amplex® Red reagent/ HRP/cholesterol oxidase/cholesterol esterase working solution was added to each microplate well containing 50 μL samples and controls. The mixed 100 μL solution was incubated for 30 min at 37°C, protected from light. The fluorescence data was measured in a fluorescence microplate reader using excitation in the range of 530–560 nm and emission detection at ~590 nm. Serum cholesterol concentration was calculated with a cholesterol standard curve from 0 to 8 μg/mL.

Statistical analysis

Data were expressed as means ± SEM. Statistical data analysis was determined with two-way ANOVA to compare different groups (strain and sex) and by t-test for comparison of two groups; p < 0.05 was considered as significant changes. All statistics were performed with GraphPad Prism 9.4.0.

Fecal microbiome of B6, BTBR, and BTBR-mtB6 mice

The fecal samples from all mice yielded V4 reads that predominately belonged to amplicon sequence variants (ASVs) of taxa within the phyla Bacteroidetes and Firmicutes, especially by members of the families Lactobacilleae, Bacteroidaceae, and Muribaculaceae (Figure 1A).

Figure 1: Fecal microbiome analysis. Phylum and Family of bfetal bacteria in females (F) and males (M) of the three strains. (A) Relative abundance of the microbial communities of mice at the phylum level and of the 20 most abundant ASVs at the family leve. (B) Plots of alpha diversity metrics and the number of taxa observed at phylum, family, and genus levels. Horizontal bars denote the median. (C) NMDS plots of beta diversity in the fecal microbiome by weighted and unweighted UniFrac, Jaccard index, and Bray-Curtis dissimilarity. Ellipses represent a 95% confidence interval. (D) Plots of relative abundances of genera that were identified as differentially abundant by ANCOM. Boxes show interquartile range, and error bars represent a 95% confidence interval. * = P ≤ 0.05, ** = P ≤ 0.005, *** = P ≤ 0.0005, **** = P ≤ 0.00005. Of the genera different in all mouse strains, Lachnoclostridium, Muribaculum, and Turicibacter were elevated in B6 mice, while Bacteroides, Blautia, and Marvinbryantia were most abundant in BTBR mice; BTBR-mtB6 microbiome was most diversified.

Overall, the BTBR-mtB6 mice had more diverse microbial communities than the B6 and BTBR mice. The number of genera, families, and phyla observed in BTBR-mtB6 mice was higher than in either BTBR or B6 mice (p<0.038). These different taxa belonged to diverse organisms as indicated by the greater Faith’s Phylogenetic Distance in BTBR-mtB6 mice than in BTBR and B6 mice (p<0.0051) (Figure 1B). This is reflected by the significant differences in the overall microbial community in feces from mice with different genetic backgrounds using distance metrics that are presence-absence (Jaccard and unweighted UniFrac) or relative-abundance based (Bray- Curtis and weighted UniFrac) and include phylogenetic distance (UniFrac) or not (Jaccard and Bray-Curtis) (Figure 1C, PERMANOVA p≤0.030). Out of 54 genera detected in the mice, an ANCOM-BC analysis found that 49 were differentially abundant in at least one mouse strain, and 11 were different in all three strains. Of the genera different in all mouse strains, Lachnoclostridium, Muribaculum, and Turicibacter were elevated in B6 mice, while Bacteroides, Blautia, and Marvinbryantia were most abundant in BTBR mice (Figure 1D).

Mitochondrial Number and Activity in Small Intestine Epithelial Cells (SIECs) and Livers

Mitochondrial numbers were detected as previously described [8]. The numbers of mitochondria in both SIECs and livers were only slightly but not significantly decreased in BTBR mice. On average, it was highest in SIECs in BTBR- mtB6 mice, but also with no significance, compared with B6 and BTBR mice. There was no significant change in liver mitochondria number between BTBR and BTBR-mtB6 mice, both were slightly lower than B6 on average. The numbers of mitochondria in all three groups individually varied greatly. For the mitochondrial activity, the five complexes (complex I to complex V) of the electron transport chain (ETC) were detected as previously described [13]. There were no obvious mitochondrial activity changes in SIECs among the B6, BTBR, and BTBR-mtB6 mice (Figure 2).

Figure 2: Electron Transport Chain (Complex I-V) of SIEC and Liver.

However, the mitochondrial activities in BTBR livers were lower than those in B6 livers. Although the mitochondria in BTBR-mtB6 mice originally came from B6 mice, there were some differences between BTBR-mtB6 and B6 mice, indicating mitochondria regulation was not completely independent but partly regulated by the host cells.

Genes of small intestine epithelial cells (SIECs) and liver affecting lipid absorption

Some lipid absorption-related factors were detected in SIECs at both gene and protein levels in females and males of BTBR, B6, and BTBR-mtB6 mice. NPC1-like intracellular cholesterol transporter (NPC1L1) and acetyl-CoA acetyltransferase 2 (Acat2) were significantly increased in SIECs of BTBR and BTBR-mtB6 mice (Figure 3A),

Figure 3: Small intestine epithelial cell (SIEC) (A) and liver (B) gene expression affecting lipid absorption. Nr1h3 inhibits the expression of NPC1L1 and promotes the expression of ABCG5/8, thereby preventing cholesterol accumulation in SIECs. Less Nr1h3 in BTBR and BTBR-mtB6 mice SIECs and livers allows higher expression of NPC1L1 suggesting enhanced cholesterol absorption in BTBR and BTBR-mtB6 mice. Each strain (6 females + 6 males) is shown as box and whiskers with Min to Max and medians from Prism analyses.

whereas nuclear receptor subfamily 1 group H member 3 (Nr1h3), which is also named as Liver x receptor α (LXRα) and members of the superfamily of ATP-binding cassette (ABC) transporters Abcg5 and Abcg8 were decreased compared with B6 mice. No significant change was found between BTBR and BTBR-mtB6 mice, and there were no significant female and male differences within a strain (Table 2). Similar strain differences were also observed for livers (Figure 3B), and there also were no sex differences within each strain (Table 3).

Table 2: Gene expression in SIECs from females and males of each strain.

Table 3: Gene expression in livers from females and males of each strain.

Glycerol transport receptor aquaporin-9 (AQP9) and Fatty Acid Binding Protein 1 (FABP1) involved in lipid transportation from SCIEs to livers also were assayed (Figure 4). AQP9 of the family of water-selective membrane channels was increased in SIECs from female and male BTBR mice compared with cells from B6 mice (Table 4).

Figure 4: Western blot analysis of aquaporin-9 (AQP9) and Fatty Acid Binding Protein 1 (FABP1) in SIECs. Compared with B6 mice, AQP9 was increased in BTBR mice and BTBR-mtB6 mice, and FABP1 was increased in both BTBR and BTBR-mtB6 mice (Table 4). There were no sex differences for AQP9 or FABP1 expression within each strain

Although FABP1 primarily serves as a fatty acid binding protein in the liver, it was also easily detected and found to be increased in SIECs from BTBR mice compared with the B6 mice regardless of sex (Table 4). No significant changes were detected between females and males of BTBR mice and BTBR-mtB6 mice (Table 4).

Table 4: Protein expression of AQP1 and FABP1 in SIECs

Enhanced hepatic cholesterol synthesis and steatosis

The enzyme 3-hydroxy-methylglutaryl-CoA reductase (HMG-CoA-R or HMGCR) is a key factor for hepatic de novo cholesterol synthesis since it is the rate-limiting enzyme. HMGCR was significantly increased in BTBR mice in both male and female mice compared with B6 mice (Figure 5; Table 5).

Figure 5: Western blot analysis of liver 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR) which is a rate-limiting enzyme for de novo cholesterol synthesis. Compared with B6 mice, HMGCR was increased in BTBR male and female mice and BTBR-mtB6 male mice; BTBR males also had higher expression than BTBR-mtB6 males and there were sex differences within the BTBR and BTBR-mtB6 strains (Table 5).

Table 5: Liver HMGCR protein expression

HMGCR protein expression was decreased in BTBR-mtB6 male and female mice compared with the same gender BTBR mice (Table 5). interestingly, BTBR-mtB6 females had lower levels than the males. Similar results were obtained with the analysis of HMGCR gene expression except BTBR-mtB6 males and females have equivalent levels (Table 6).

Table 6: Liver HMGCR gene expression

PPARγ protein liver expression (Figure 6A),

Figure 6: Hepatocyte (CD45- liver cells) expression of PPARγ. Flow cytometric gating on CD45- cells (A) for analysis of intracellular expression of PPARγ (B). Compared with B6 male mice, PPARγ was significantly increased in BTBR male mice (P < 0.05) but not BTBR-mtB6 male mice (P > 0.05). Compared with B6 female mice, PPARγ was significantly increased in BTBR female mice (P < 0.001)

and BTBR-mtB6 female mice (P < 0.01). It was improved after mitochondria replacement in BTBR female mice (P < 0.05). which affects the regulation of fatty acid storage and glucose metabolism, was measured in CD45- cells (hepatocytes) from the liver, and it also had higher expression in BTBR females and males compared to same-sex B6 mice (Figure 6B). The level of PPARγ was more at an intermediate level in female and male BTBR-mtB6 mice, and the sex differences of B6 and BTBR-mtB6 did not significantly differ (Figure 6B).

Fat deposition in liver

The oil red O staining of liver sections indicated there was more lipid in BTBR liver than in B6 liver (Figure 7),

Figure 7: Oil red O staining of liver sections from B6, BTBR, and BTBR-mtB6 mice. Frozen sections were cut at 20 µm. Photos were taken as 40X using Nikon 50i w/ Color camera. Scale bar = 100 µm. Sections were imaged using ImageJ software.

which is in agreement with the HMGCR and PPARγ results. Also similar to the HMGCR and PPAR results, the BTBR- mtB6 livers had an intermediate level of lipid.

CD3+ T Cells and CD19+ B Cells in White Adipose Tissue

Since BTBR mice have more fat and OS than B6 mice and immune cells in adipose tissue are known to increase OS, we assayed the presence of T cells in adipose tissue. As shown in Figure 8,

Figure 8: CD45+ lymphoid cells were gated on in white adipose tissue to quantify the presence of CD3+ T cells and CD19+ B cells in males of the three strains.

the mean percentage of CD3+ T cells in white adipose tissue in BTBR male mice (14.67%), was significantly increased (p=0.002) compared with B6 male mice (5.08%). Interestingly, it was reduced (p=0.0037) in BTBR-mtB6 males (7.23%). Although the average of CD3+ T cells in white adipose tissue in BTBR-mtB6 male mice was higher than in B6 male mice, the difference was not significant (P>0.05).

The mean percentages of B cells in white adipose tissue of B6 (3.39%), BTBR (6.40%), and BTBR-mtB6 (5.83%) males did not significantly differ. These results are considered preliminary because only a few males were assayed. We did not assay females due to limited funds for analysis of estrogen, and it is known that white adipose tissue homeostasis and expansion are influenced by estrogen during female ovulation.

Elevated serum cholesterol level in BTBR mice

Both B6 and BTBR mice had higher serum cholesterol levels in males compared to females; this sex difference was not significant for the BTBR-mtB6 males and females. However, increased serum cholesterol levels were detected in both males and females of the BTBR and BTBR- mtB6 mice compared with the same gender B6 mice (Figure 9).

Figure 9 Serum cholesterol levels. Compared with B6 male mice, serum cholesterol levels were greatly increased in BTBR male mice, while improved in BTBR-mtB6 male mice. The serum cholesterol levels were increased in both BTBR and BTBR-mtB6 female mice compared with B6 female mice. No obviouschanges existed after mitochondria replacement in BTBR female mice. Two-way ANOVA (sex x strain) was F(5,25)=26.39, p<0.0001.There were sex differences in serum cholesterol levels within the B6 and BTBR strains, which were higher for males.

The serum cholesterol levels of BTBR-mtB6 males, but not females, were significantly decreased compared with the same gender BTBR mice. However, compared with B6 male mice, it was still higher in BTBR-mtB6 male mice. Two- way ANOVA (sex x strain) was F(5,25)=26.39 p<0.0001.

Reduced cholesterol level in BTBR mice brain

The cholesterol levels in the brains of BTBR males and females were significantly lower compared with those of B6 males and females, respectively (Figure 10).

Figure 10: Brain cholesterol levels. Compared with B6 male mice, brain cholesterol levels were decreased in BTBR male mice, while improved in BTBR- mtB6 male mice. Compared with B6 female mice, the brain cholesterol levels were also significantly decreased in BTBR and BTBR-mtB6 females. Two-way ANOVA (sex x strain) was F(5,35)=7.228, p<0.0001. There was no sex difference within each strain

Although the brain cholesterol levels of BTBR-mtB6 mice were significantly increased in males compared with BTBR males, it was not significantly lower than the level of B6 males. BTBR females also had significantly less brain cholesterol than B6 females, but unlike the males, the brain cholesterol level of BTBR-mtB6 females was still significantly lower than that of B6 females. Two-way ANOVA (sex x strain) was F(5,35)=7.228 p<0.0001.

IgG Antibodies to Brain Antigens

Antibodies to brain antigens of B6 brain lysate were assayed with serum from the B6, BTBR, and BTBR-mtB6 males (Figure 11).

 Figure 11: Western blot analyses of IgG to B6 brain homogenate with sera from B6, BTBR, and BTBR-mtB6 mice.

The binding of IgG autoantibodies to an approximate 45KDa protein from a B6 brain was only observed with BTBR sera. Mass spectrometry analysis of  the protein(s) in the band with bound IgG will be needed to identify the antigen(s); this will be under further investigation in the future.

Reduced myelination in BTBR mice cerebral cortex

Myelination-related factors myelin basic protein (MBP) and 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNPase) were detected in the cerebral cortex of the three mouse strains (Figure 12).

 Figure 12: Western blot analyses brain MBP and CNPase. MBP was significantly decreased in BTBR male and female mice, compared with B6 mice (Tables 7 and 8).

The male BTBR mice had significantly less MBP than the B6 males; the MBP levels in the B6 and BTBR females also were significantly different (Table 7). The BTBR-mtB6 males and females had intermediate levels of MBP that did not differ from the B6 or BTBR males and females, respectively (Table 7). The CNPase brain levels were significantly different only between B6 and BTBR- mtB6 males and between male and female BTBR-mtB6 mice (Table 8).

Table 7: Brain MBP protein expression

Table 8: Brain CNPase protein expressison

The BTBR mouse strain has been a useful animal model for studying ASD, and it has been widely accepted for neuroscience and behavioral studies of ASD [38,39]. The new BTBR-mtB6 strain had improved social behaviors even though its nuclear genome is mainly that of BTBR mice [13]. The BTBR-mtB6 strain has mainly the nuclear genome of BTBR, although there may be some mitochondrial epigenetic influences. Besides producing ATP, mitochondria affect metabolic precursors and metabolic regulation of lipids, proteins, glucose, and nuclear acids, as well as metabolic by-products like ROS and ammonia that may affect genetic expression.

Mitochondrial disease (MD) and abnormal redox metabolism [40], has been associated with ASD. The BTBR strain was reported to have more oxidative products and fewer antioxidants than the B6 strain [22]. Many significantly different levels of metabolites have been found between mothers of ASD children and the mothers of normal children at 2–5 years after birth [41]. Abnormalities in lipid metabolism during early development, such as changes in adipokine levels, are associated with ASD [42-44]. The gut microbiota is known to influence host metabolism with the modification of lipids, bile acids, short-chain fatty acids, and other dietary nutrients, which mediate effects through the gut-brain axis [45,46]. Microbial metabolism can make nutrients available to the host that otherwise would not be used. Regulation of lipid availability is critical for many essential biological functions. Considering that many ASD patients have disturbed intestinal functions and altered gut microbiota, which play a critical role in physiological homeostasis, immunological development, and nutrition metabolism, fecal microbiomes of the three strains were assessed. The composition of the fecal microbiome differed significantly between BTBR, B6, and BTBR-mtB6 mice. Using  high-tech DNA analysis, Kang et al. [47], reported that children with ASD had significantly decreased intestinal microbial diversity, but there was no link between this lower biodiversity and the severity of GI symptoms and no diet differences between ASD and normal children. Some gut microbes were increased in the BTBR mice. BTBR mice had different microbiota from B6 mice, and when treated with palmitoylethanolamide, BTBR behavior and microbiota were more like B6 mice [48].

Children with ASD were reported to have a higher gut abundance of Bacteroides, Parabacteroides, Clostridium, Faecalibacterium, and Phascolarctobacterium [49]. Gut microbes such as Bacteroides, Prevotella, and Porphyromonas, have been reported to produce sphingolipids (SPs) and other groups of lipids. Brown etal.ceramide phosphoinositol and deoxy-sphingolipids, modulate intestinal inflammation and regulate host ceramide pools in animals. Interestingly, the BTBR- mtB6 microbiome has an intermediate level of microbial phenotypes between the BTBR and B6 mice, and it is also more diverse than that of either B6 or BTBR mice. Although significantly different, the fecal microbiome of BTBR-mtB6 mice slightly resembled the BTBR microbiome more than it did the B6 microbiome by both alpha and beta diversity metrics.

Bacterial symbionts can interact with and even produce neurotransmitters and their precursors in mammals. Given the bidirectional flow of information as part of the gut- brain axis and the frequency of GI distress in ASD, microbes  have often been hypothesized to play a role in this system. Neurotransmitters are a potential mechanism of this interaction. The gut microbiome can affect host serotonin levels [51,52], and serotonin levels influence autism [53]. Turicibacter, which has been observed to import and benefit from serotonin, was present in 9 out of 10 of the B6 mice but was completely absent in both the BTBR and BTBR-mtB6 mice. Another microbial genus, Blautia, can metabolize tryptophan to tryptamine. Tryptophan is the precursor to serotonin, and tryptamine can bind to serotonin receptors. Blautia was highly elevated in the 

BTBR mice and was more abundant in BTBR-mtB6 mice than in B6 mice. The gut-brain axis has been expanded to the microbiome making the triparitite microbiota-gut-brain “connectome” [54], and the gut microbiome influences neuroimmunology affecting different neuropathologies [55], including ASD [56].

On average, the BTBR mice have slightly fewer mitochondria in SIECs and liver than the B6 mice (P>0.05). No obvious changes in mitochondrial numbers were found after the mitochondria replacement in BTBR-mtB6 mice in the liver. The activity of the ETC complexes of BTBR livers, which are the main indicators for mitochondrial capacity, were lower in the BTBR mice and might promote subsequent progression of non-alcoholic fatty liver diseases from steatosis to steatohepatitis and fibrosis/ cirrhosis. Mitochondria replacement could attenuate this pathological progression. However, due to the dependent regulation of mitochondria on host cells, this improvement was limited. Due to variances in the removal of mesenteric adipose tissue, the data greatly varied. The numbers of mitochondria were increased in SIECs in BTBR-mtB6 mice on average; however, the activity assay showed no obvious changes among the three groups, neither improvement nor decline.

The gut microbiota can affect lipid metabolism and lipid levels in sera and tissues [57]. There are two main sources of cholesterol generation. One is self-generation in the liver and peripheral tissues as endogenous sources while another is absorbed in the intestine from food as dietary sources. Most lipids (~95%) are absorbed through the SIECs. Absorption of dietary lipids is a major determinant of animal energy balance and metabolic health. The regulation of intestinal cholesterol absorption and secretion is mainly mediated by the liver X receptor (LXR). In mice, the LXR homolog is NR1H3. As a cholesterol sensor, NR1H3 was decreased in BTBR mice compared with B6 mice. No significant change was found between BTBR and BTBR-mtB6 mice. Dietary cholesterol is absorbed into SIECs through NPC1L1, esterified by ACAT2 in the endoplasmic reticulum, and then incorporated into chylomicrons (CMs). CMs are transported to the Golgi apparatus, secreted into the lymphatic system, and finally into the bloodstream [58]. NR1H3 inhibits the expression of NPC1L1 and promotes the expression of ABCG5/8, thereby preventing cholesterol accumulation in SIECs [59]. When NR1H3 expression in SIECs is low as in BTBR mice, the expression of NPC1L1 is enhanced, which results in enhanced cholesterol absorption. We found both NPC1L1 and ACAT2 were increased in BTBR mice compared with B6 mice. The glycerol transport receptor AQP9 and FABP1 of SIECs of BTBR mice were also both significantly increased compared with B6 mice. There is a small amount of excess free cholesterol that can be excreted into the intestinal lumen via ABCG5 and ABCG8. Inhibition of NR1H3 in SIECs decreases ABCG5/8, which would result in less cholesterol excretion. Both ABCG5 and ABCG8 were significantly decreased in BTBR mice compared with B6 mice. All these results suggest more dietary cholesterol is absorbed into SIECs and esterified, incorporated into CMs, while less cholesterol is excreted. No sex difference was found within each of the three strains. A significant difference between BTBR and BTBR-mtB6 mice was not obtained, although the BTBR-mtB6 mice did tend to have expression levels more intermediate between BTBR and B6 mice, which suggests potential differences in lipid metabolite levels.

The liver is the main site for de novo cholesterol synthesis in mammals with approximately 80% of total cholesterol synthesis. Cholesterol synthesis is a complex multi-step process in which HMGCR is the critical rate- limiting enzyme [60]. HMGCR expression in the liver was increased in BTBR male and female mice compared with B6 same-gender mice. It was decreased in BTBR-mtB6 male and female mice, especially male mice, compared with BTBR same-gender mice. All these levels suggest increased cholesterol synthesis in BTBR mice, further promoting increased serum cholesterol. Inhibition of cholesterol synthesis critical genes such as HMGCR could lower serum cholesterol in BTBR-mtB6 mice, especially male mice. Like in SIECs, Nr1h3 expression was lower in the livers of BTBR mice compared with B6 mice. As mentioned above, the most important target genes of Nr1h3 include ABCG5/G8,

NPC1L1, and ACAT-2. Decreased liver expression of Nr1h3 in BTBR mice can reduce hepatic cholesterol excretion via decreased ABCG5/8, promote hepatic cholesterol absorption via increased NPC1L1, and promote cholesterol storage via increased ACAT-2. No obvious difference was found between BTBR and BTBR-mtB6 mice on these genes. To confirm the lipid deposition differences between the strains, lipid deposition was assessed using Oil red O staining. As anticipated based on adipogenic expression, increased lipid deposition was observed in the liver of BTBR mice, which can cause hepatic steatosis. There was less lipid deposition in the livers of BTBR-mtB6 mice compared to the livers of BTBR mice, but it was still higher than that observed in B6 livers. Molecular evidence for hepatic steatosis in these three strains was further investigated with an assessment of PPARγ. Peroxisome proliferator-activated receptors (PPARs) are members of the superfamily of ligand-activated nuclear transcription factors, and PPARγ is a classical marker of hepatic steatosis; PPARγ is preferentially expressed in adipose for fatty acid synthesis and storage [61]. PPARγ was increased in both female and male BTBR livers compared with B6 livers, suggesting increased adipose synthesis and storage in BTBR livers. It was significantly decreased in BTBR-mtB6 mice compared with BTBR mice.

Adipose tissue affects the metabolic state of the immune system via adipokines, which control immune cells such as T cell activity. It performs immunoregulatory functions besides lipid storage. Various T cells in different adipose tissues play a critical role in metabolic homeostasis maintaining immunometabolism [62], and sex hormones affect adipose tissue [25,26]. Once irregulated, some pathological conditions such as obesity, diabetes, cardiovascular diseases, and brain disorders occur. The adipose tissue of the BTBR mice did have the presence of more T cells than that of B6 or BTBR-mtB6 mice. More T cells in adipose tissue have related to more inflammation and the production of ROS [63], and BTBR mice do have more inflammation and oxidative stress than B6 mice [22].

There is a balance between cholesterol absorption and synthesis in a healthy body [64,65], which means if the cholesterol synthesis increases, then the cholesterol absorption decreases, and vice versa. However, we found both the cholesterol absorption and synthesis were increased in BTBR male and female mice, compared with B6 same-gender mice. After mitochondria replacement, hepatic cholesterol synthesis was significantly decreased in BTBR-mtB6 mice with no obvious effects on cholesterol absorption compared with BTBR mice. Increased cholesterol absorption and cholesterol synthesis will be reflected in the blood lipid panel. The serum cholesterol

level was elevated in BTBR mice, and it was lower in the BTBR-mtB6 mice. Abnormalities of cholesterol are highly associated with ASD [66,67]. The level of cholesterol in the brain is suggested to play a critical role in brain development [68,69]. The brain cholesterol level was decreased in the brains of the BTBR mice compared with B6 mice; it was significantly increased in BTBR-mtB6 mice compared with BTBR mice.

Cholesterol is critical for many brain functions. Although the brain is the most cholesterol-rich organ (approximately 25% of the whole body), serum cholesterol can’t be directly delivered to the brain due to blockage at the blood-brain barrier. Astrocytes are important cells for in situ and de novo cholesterol synthesis in the brain, although evidence indicates a role for different neurons [68]. We previously reported a decreased number of astrocytes in the brain of BTBR mice compared to an increased number of astrocytes in the brain of BTBR- mtB6 mice [13]. These results help to explain the reduced astrocyte influence in BTBR brains, which synthesize a lower amount of cholesterol that affects the brain’s normal functions, including social behaviors.

In the mammalian brain, the highest cholesterol content is found in myelin as the integral myelin component. A high cholesterol level is required for myelin membrane growth. It is the rate-limiting factor for developing myelin in the brain, while lower myelin may cause neurons to be vulnerable to attack. Myelin abnormalities are highly associated with ASD, which will affect signal processing and cognition. Myelination is the generation of layers of myelin that wrap around the neuronal axons. Our study found that MBP was significantly decreased in BTBR mice compared with B6 mice of the same sex. This myelination inhibition might be partly due to the decreased brain cortex cholesterol level. BTBR mice do have a major absence of corpus callosum and a severely reduced hippocampal commissure [70]. The corpus callosum is a large bundle of myelinated nerve fibers connecting the two brain hemispheres, thus permitting right and left hemispheric communication; it is highly associated with ASD [71,72]. The slightly improved levels of MBP and CNPase in the BTBR-mtB6 mice may explain their improved behavior; additionally, the BTBR and BTBR-mtB6 females also had slightly higher levels. However, in general, unlike humans with higher ASD prevalence in males, BTBR females and males have equivalent behaviors [73,74].

Antibodies directed against brain antigens play a direct role in brain disease pathogenesis. Antibodies to brain antigens were assayed with serum from three strains of mice. The binding of autoantibodies (mostly IgG) was only 

observed with BTBR sera, indicating a potential mechanism of autoimmune neural modulation. These results suggest the mitochondrion replacement affects the neuroimmune network and the microbiome-gut-brain axis [75,76], which lessens the autoimmune [21], and neuroinflammatory [22], phenotype of the BTBR strain. Although most anti-brain antibodies target epitopes expressed on neurons, anti- brain antibodies can also be directed to antigens expressed on glia, thereby causing different brain pathology. Our next step is to target serum antibodies directly against different types of neural cells and glia in three groups of mice brains, to find out the potential autoimmune neuronal/glia injury target.

In conclusion, compared with B6 mice, BTBR mice have more microbiota related to lipid secretion. The fecal microbiome of BTBR-mtB6 mice was distinct and more diverse than the BTBR or B6 strain. Lipid absorption was enhanced in SIECs of BTBR mice, along with the liver’s enhanced de novo cholesterol synthesis, as shown with elevated serum cholesterol. There was no obvious improvement in cholesterol absorption but more controlled cholesterol synthesis in BTBR-mtB6 mice compared to BTBR mice. The brain cholesterol level was significantly lower in BTBR brains compared to B6 brains and significantly increased in BTBR-mtB6 brains compared with BTBR brains. Myelination was also inhibited in BTBR brains. The improved myelination in the BTBR-mtB6 strain may be related to lower levels of neuroinflammation and serum IL-17 levels, together with higher levels of glutamate and astrocytes in the brain [13]. Although further research is needed, the results indicate that lipid metabolism and cholesterol levels are involved in the ASD-like behaviors of the BTBR mice. Mitochondria and microbiomes are implicated in biochemistry and behavior since there are improvements with B6 mitochondria in the BTBR strain. This animal model study adds to our knowledge of ASD and further aids our understanding of the complexities associated with ASD, which may help to address future interventions. Mitochondrial dysfunction has been associated with ASD, but there’s no definitive evidence on whether fewer mitochondria/cell also has an influence. Future studies are needed to assess if nutraceuticals to increase mitochondrial numbers and functions can be effective in reducing ASD.

We appreciate the kind help from Richard Cole in Light Microscopy Core, Wadsworth Center for the microscopy of liver sections, and the animal facility staff for the maintenance of the mouse strains.

Funding Information

The work reported in this manuscript was supported by an NIH grant (R01 ES025584) to DAL.

Data Availability

All data used for this study are available upon request.

Conceptualization, D.A.L. and Y.Y.; methodology, D.A.L., J.G. and Y.Y.; software, K.M. and J.G.; validation, Y.Y.,

J.G., and D.A.L.; formal analysis, Y.Y., K.M., M.N.U, T.K.M.

and B.O.; investigation, Y.Y., K.M., M.N.U, T.K.M. and B.O.; resources, D.A.L. and J.G.; data curation, Y.Y. and B.O.; writing—original draft preparation, Y.Y., B.O. and D.A.L.; writing—review and editing, all authors; visualization, Y.Y.; supervision, D.A.L. and J.G.; project administration, D.A.L.; funding acquisition, D.A.L. All authors have read and agreed to the published version of the manuscript.

Author Contributions

1. Cheng N, Rho JM, Masino SA. Metabolic Dysfunction Underlying Autism Spectrum Disorder and Potential Treatment Approaches. Front Molec Neurosci. 2017; 10: 34.
2. Kim Y, Vadodaria KC, Lenkei Z, Kato T, Gage FH, Marchetto MC, et al. Mitochondria Metabolism and Redox Mechanisms in Psychiatric Disorders. Antioxid Redox Signal. 2019; 31: 275-317.
3. Frye RE. Mitochondrial Dysfunction in Autism Spectrum Disorder: Unique Abnormalities and Targeted Treatments. Semin Pediatr Neurol. 2020; 35: 100829.
4. Thorsen M. Oxidative stress metabolic and mitochondrial abnormalities associated with autism spectrum disorder. Prog Mol Biol Transl Sci. 2020; 173: 331-354.
5. Bjørklund G, Tinkov AA, Hosnedlová B, Kizek R, Ajsuvakova OP, Chirumbolo S, et al. The role of glutathione redox imbalance in autism spectrum disorder: A review. Free Radic Biol Med. 2020; 160: 149- 162.
6. Gevezova M, Minchev D, Pacheva I, Sbirkov Y, Yordanova R, Timova E, et al. Cellular Bioenergetic and Metabolic Changes in Patients with Autism Spectrum Disorder. Curr Top Med Chem. 2021; 21: 985-994.
7. Queen NJ, Boardman AA, Patel RS, Siu JJ, Mo X, Cao L. Environmental enrichment improves metabolic and behavioral health in the BTBR mouse model of autism. Psychoneuroendocrinology. 2020; 111: 104476.
8. Yao Y, Uddin MN, Manley K, Lawrence DA. Constitutive activation of Notch signalling and T cell activation characterize a mouse model of autism. Cell Biochem Funct. 2022; 40: 150-162.
9. Cao C, Li Q, Chen Y, Zou M, Sun C, Li X, et al. Untargeted Metabolomic Analysis Reveals the Metabolic Disturbances and Exacerbation of Oxidative Stress in the Cerebral Cortex of a BTBR Mouse Model of Autism. J Mol Neurosci. 2023; 73: 15-27.
10. Suomalainen A, Nunnari J. Mitochondria at the crossroads of health and disease. Cell. 2024; 187: 2601-2627.
11. Healy TM, Burton RS. Strong selective effects of mitochondrial DNA on the nuclear genome. Proc Natl Acad Sci U S A. 2020; 117: 6616- 6621.
12. Heuer B. The effects of nuclear DNA mutations on mitochondrial function. J Am Assoc Nurse Pract. 2023; 35: 2-4.
13. Yao Y, Uddin MN, Manley K, Lawrence DA. Improvements of autism- like behaviors but limited effects on immune cell metabolism after mitochondrial replacement in BTBR T+Itpr3tf/J mice. J Neuroimmunol. 2022; 368: 577893.
14. Mangiola F, Ianiro G, Franceschi F, Fagiuoli S, Gasbarrini G, Gasbarrini A. Gut microbiota in autism and mood disorders. World J Gastroenterol. 2016; 22: 361-368.
15. Li Q, Zhou JM. The microbiota-gut-brain axis and its potential therapeutic role in autism spectrum disorder. Neurosci. 2016; 324: 131-139.
16. Saurman V, Margolis KG, Luna RA. Autism Spectrum Disorder as a Brain-Gut-Microbiome Axis Disorder. Dig Dis Sci. 2020; 65: 818-828.
17. Rose S, Niyazov DM, Rossignol DA, Goldenthal M, Kahler SG, Frye RE. Clinical and Molecular Characteristics of Mitochondrial Dysfunction in Autism Spectrum Disorder. Mol Diagn Ther. 2018; 22: 571-593.
18. Yardeni T, Tanes CE, Bittinger K, Mattei LM, Schaefer PM, Singh LN, et al. Host mitochondria influence gut microbiome diversity: A role for ROS. Sci Signal. 2019; 12: 3159.
19. Ballard JWO, Towarnicki SG. Mitochondria the gut microbiome and ROS. Cell Signal. 2020; 75: 109737.
20. Wu H, Zhao G, Liu S, Zhang Q, Wang P, Cao Y, Wu L. Supplementation with selenium attenuates autism-like behaviors and improves oxidative stress inflammation and related gene expression in an autism disease model. J Nutr Biochem. 2022; 107: 109034.
21. Uddin MN, Yao Y, Mondal T, Matala R, Manley K, Lin Q, et al. Immunity and autoantibodies of a mouse strain with autistic-like behavior. Brain Behav Immun Health. 2020; 4: 100069.
22. Uddin MN, Mondal T, Yao Y, Manley K, Lawrence DA. Oxidative stress and neuroimmune proteins in a mouse model of autism. Cell Stress Chaperones. 2023; 28: 201-217.
23. Gaignard P, Fréchou M, Liere P, Thérond P, Schumacher M, Slama A, Guennoun R. Sex differences in brain mitochondrial metabolism: influence of endogenous steroids and stroke. J Neuroendocrinol. 2018; 30
24. Govender P, Ghai M, Okpeku M. Sex-specific DNA methylation: impact on human health and development. Mol Genet Genomics. 2022; 297: 1451-1466.
25. Palmisano BT, Zhu L, Eckel RH, Stafford JM. Sex differences in lipid and lipoprotein metabolism. Mol Metab. 2018; 15: 45-55.
26. Palmer BF, Clegg DJ. The sexual dimorphism of obesity. Mol Cell Endocrinol. 2015; 402: 113-119
27. Ceasrine AM, Devlin BA, Bolton JL, Green LA, Jo YC, Huynh C, et al. Maternal diet disrupts the placenta-brain axis in a sex-specific manner. Nat Metab. 2022; 4: 1732-1745.
28. Iwata R, Casimir P, Erkol E, Boubakar L, Planque M, Gallego López IM, et al. Mitochondria metabolism sets the species-specific tempo of neuronal development. Science. 2023; 379: eabn4705.
29. Graf J, Ledala N, Caimano MJ, Jackson E, Gratalo D, Fasulo D, et al. High-Resolution Differentiation of Enteric Bacteria in Premature Infant Fecal Microbiomes Using a Novel rRNA Amplicon. mBio. 2021; 12: e03656-20.
30. Benjamino J, Beka L, Graf J. Microbiome Analyses for Toxicological Studies. Curr Protoc Toxicol. 2018; 77: e53.
31. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible interactive scalable and extensible microbiome
32. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016; 13: 581-583.
33. Davis NM, Proctor DM, Holmes SP, Relman DA, Callahan BJ. Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome. 2018; 6: 226.
34. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013; 8: e61217.
35. Lin H, Peddada SD. Analysis of compositions of microbiomes with bias correction. Nat Commun. 2020; 11: 3514.
36. Jung Y, Zhao M, Svensson KJ. Isolation culture and functional analysis of hepatocytes from mice with fatty liver disease. STAR Protoc. 2020; 1: 100222.
37. Mondal TK, Emeny RT, Gao D, Ault JG, Kasten-Jolly J, Lawrence DA. A physical/psychological and biological stress combine to enhance endoplasmic reticulum stress. Toxicol Appl Pharmacol. 2015; 289: 313-322.
38. McFarlane HG, Kusek GK, Yang M, Phoenix JL, Bolivar VJ, Crawley JN. Autism-like behavioral phenotypes in BTBR T+tf/J mice. Genes Brain Behav. 2008; 7: 152-163.
39. Meyza KZ, Blanchard DC. The BTBR mouse model of idiopathic autism - Current view on mechanisms. Neurosci Biobehav Rev. 2017; 76: 99-110.
40. Frye RE, Delatorre R, Taylor H, Slattery J, Melnyk S, Chowdhury N, et al. Redox metabolism abnormalities in autistic children associated with mitochondrial disease. Transl Psychiatry. 2013; 3: e273.
41. Hollowood-Jones K, Adams JB, Coleman DM, Ramamoorthy S, Melnyk S, James SJ, et al. Altered metabolism of mothers of young children with autism spectrum disorder: a case control study. BMC Pediatr. 2020; 20: 557.
42. Rodrigues DH, Rocha NP, Sousa LF, Barbosa IG, Kummer A, Teixeira AL. Changes in adipokine levels in autism spectrum disorders. Neuropsychobiology. 2014; 69: 6-10.
43. Chen J, Chen J, Xu Y, Cheng P, Yu S, Fu Y, et al. Retinol-binding protein 4 in combination with lipids to predict the regression phenomenon of autism spectrum disorders. Lipids Health Dis. 2021; 20: 93.
44. Esposito CM, Buoli M, Ciappolino V, Agostoni C, Brambilla P. The Role of Cholesterol and Fatty Acids in the Etiology and Diagnosis of Autism Spectrum Disorders. Int J Mol Sci. 2021; 22: 3550.
45. Jia X, Xu W, Zhang L, Li X, Wang R, Wu S. Impact of Gut Microbiota and Microbiota-Related Metabolites on Hyperlipidemia. Front Cell Infect Microbiol. 2021; 11: 634780.
46. Xiao W, Su J, Gao X, Yang H, Weng R, Ni W, et al. The microbiota- gut-brain axis participates in chronic cerebral hypoperfusion by disrupting the metabolism of short-chain fatty acids. Microbiome. 2022; 10: 62.
47. Kang DW, Park JG, Ilhan ZE, Wallstrom G, Labaer J, Adams JB, et al. Reduced incidence of Prevotella and other fermenters in intestinal microflora of autistic children. PLoS One. 2013; 8: e68322.
48. Cristiano C, Pirozzi C, Coretti L, Cavaliere G, Lama A, Russo R, et al. Palmitoylethanolamide counteracts autistic-like behaviours in BTBR T+tf/J mice: Contribution of central and peripheral mechanisms. Brain Behav Immun. 2018; 74: 166-175.
49. Iglesias-Vázquez L, Van Ginkel Riba G, Arija V, Canals J. Composition of Gut Microbiota in Children with Autism Spectrum Disorder: A Systematic Review and Meta-Analysis. Nutrients. 2020; 12: 792.
50. Brown EM, Ke X, Hitchcock D, Jeanfavre S, Avila-Pacheco J, Nakata T, et al. Bacteroides-Derived Sphingolipids Are Critical for Maintaining Intestinal Homeostasis and Symbiosis. Cell Host Microbe. 2019; 25: 668-680.e7.
51. Agus A, Planchais J, Sokol H. Gut Microbiota Regulation of Tryptophan Metabolism in Health and Disease. Cell Host Microbe. 2018; 23: 716- 724.
52. Legan TB, Lavoie B, Mawe GM. Direct and indirect mechanisms by which the gut microbiota influence host serotonin systems. Neurogastroenterol Motil. 2022; 34: e14346.
53. Takumi T, Tamada K, Hatanaka F, Nakai N, Bolton PF. Behavioral neuroscience of autism. Neurosci Biobehav Rev. 2020; 110: 60-76.
54. Gershon MD, Margolis KG. The gut its microbiome and the brain: connections and communications. J Clin Invest. 2021; 131: e143768.
55. Yu LW, Agirman G, Hsiao EY. The Gut Microbiome as a Regulator of the Neuroimmune Landscape. Annu Rev Immunol. 2022; 40: 143-

56. 167.Srikantha P, Mohajeri MH. The Possible Role of the Microbiota-Gut- Brain-Axis in Autism Spectrum Disorder. Int J Mol Sci. 2019; 20: 2115.

57. Schoeler M, Caesar R. Dietary lipids gut microbiota and lipid metabolism. Rev Endocr Metab Disord. 2019; 20: 461-472.
58. Turley SD. The role of Niemann-Pick C1 - Like 1 (NPC1L1) in intestinal sterol absorption. J Clin Lipidol. 2008; 2: S20-S28.
59. Williams K, Segard A, Graf GA. Sitosterolemia: Twenty Years of Discovery of the Function of ABCG5ABCG8. Int J Mol Sci. 2021; 22: 2641.
60. Brown AJ, Coates HW, Sharpe LJ. Cholesterol Synthesis (Chapter 10) in Biochemistry of Lipids Lipoproteins and Membranes 7th Edition 2021 (eds: Ridway N McLeodR)ElsevierScienceISBN:9780323983280
61. Smith SA. Peroxisome proliferator-activated receptors and the regulation of mammalian lipid metabolism. Biochem Soc Trans. 2002; 30: 1086-1090.
62. Wang Q, Wu H. T Cells in Adipose Tissue: Critical Players in Immunometabolism. Front Immunol. 2018; 9: 2509.
63. Wang Q, Wang Y, Xu D. The roles of T cells in obese adipose tissue inflammation. Adipocyte. 2021; 10: 435-445.
64. Cohen DE. Balancing cholesterol synthesis and absorption in the gastrointestinal tract. J Clin Lipidol. 2008; 2: S1-s3.
65. Afonso MS, Machado RM, Lavrador MS, Quintao ECR, Moore KJ, Lottenberg AM. Molecular Pathways Underlying Cholesterol Homeostasis. Nutrients. 2018; 10: 760.
66. Tierney E, Bukelis I, Thompson RE, Ahmed K, Aneja A, Kratz L, et al. Abnormalities of cholesterol metabolism in autism spectrum disorders. Am J Med Genet B Neuropsychiatr Genet. 2006; 141B: 666-668.
67. Franco R, Rivas-Santisteban R, Navarro G, Reyes-Resina I. Cholesterol in autism spectrum disorders. Exploration Neuroprotective Therapy 2021; 1: 10-18.
68. Genaro-Mattos TC, Anderson A, Allen LB, Korade Z, Mirnics K. Cholesterol Biosynthesis and Uptake in Developing Neurons. ACS Chem Neurosci. 2019; 10: 3671-3681.
69. Lu F, Ferriero DM, Jiang X. Cholesterol in Brain Development and Perinatal Brain Injury: More than a Building Block. Curr Neuropharmacol. 2022; 20: 1400-1412.
70. Wahlsten D, Metten P, Crabbe JC. Survey of 21 inbred mouse strains in two laboratories reveals that BTBR T/+ tf/tf has severely reduced hippocampal commissure and absent corpus callosum. Brain Res. 2003; 971: 47-54.
71. Loomba N, Beckerson ME, Ammons CJ, Maximo JO, Kana RK. Corpus callosum size and homotopic connectivity in Autism spectrum disorder. Psychiatry Res Neuroimaging. 2021; 313: 111301.
72. Yao S, Zhou M, Zhang Y, Zhou F, Zhang Q, Zhao Z, et al. Decreased homotopic interhemispheric functional connectivity in children with autism spectrum disorder. Autism Res. 2021; 14: 1609-1620.
73. Heo Y, Zhang Y, Gao D, Miller VM, Lawrence DA. Aberrant immune responses in a mouse with behavioral disorders. PLoS One. 2011; 6: e20912.
74. Zhang Y, Gao D, Kluetzman K, Mendoza A, Bolivar VJ, Reilly A, et al. The maternal autoimmune environment affects the social behavior of offspring. J Neuroimmunol. 2013; 258: 51-60.
75. Khalil M, Zhang Z, Engel MA. Neuro-Immune Networks in Gastrointestinal Disorders. Visc Med. 2019; 35: 52-60.
76. Jacobson A, Yang D, Vella M, Chiu I.M. The intestinal neuro-immune axis: crosstalk between neurons immune cells and microbes. Mucosal Immunol. 2021; 14: 555-565.