Association of mitochondrial genome variants with Alzheimer's disease in a Chinese population

Abstract

Background

Research on the mitochondrial genome variants of Alzheimer's disease (AD) in Chinese populations is lacking.

Objective

The study aimed to identify mitochondrial DNA (mtDNA) variants associated with AD risk and explore the relationship between mtDNA variants and plasma biomarkers in AD patients.

Methods

Whole genome sequencing was performed in 1509 AD patients and 2010 controls from the Chinese population. mtDNA variants were called according to GATK's best practice mitochondrial pipeline. We evaluated the association of AD risk with mtDNA variants and mitochondrial haplogroup. Common variant (MAF≥0.01) based association analysis and gene-based tests of rare variants (MAF<0.01) were carried out with PLINK 1.9 and SKAT-O, respectively. Spearman correlation analysis was performed to assess the association between the burden of mtDNA variants and plasma biomarker levels.

Results

The frequency of mitochondrial haplogroup G in AD group was nominally higher than control group (p = 0.019, OR = 1.48). Rare variants of MT-CYB gene were significantly enriched in controls compared to AD patients (p = 2.81 × 10⁻⁴, OR = 0.886). Besides, the control group exhibited considerably lower mRNA expression of MT-CYB in brain regions compared to AD patients in GEO database. Furthermore, the number of mtDNA indel variants per individual correlated positively with plasma Aβ42 levels.

Conclusions

Mitochondrial haplogroup G may serve as a risk factor for AD, while rare variants of MT-CYB gene acted as protective factor against AD in mainland China. Moreover, mtDNA variants were related to AD plasma biomarker levels. Our findings highlighted the role of mitochondrial genome variants in the pathogenesis of AD.

Keywords

Alzheimer's disease mitochondrial DNA variants mitochondrial haplogroup plasma biomarker

Introduction

Alzheimer's disease (AD), the most common form of dementia, is pathologically characterized by extracellular amyloid plaques formed by the aggregation of amyloid-β (Aβ) and neuronal fibrillar tangles formed by abnormally phosphorylated tau protein. The etiology and pathogenesis of AD are exceedingly complicated, involving factors such as inheritance, aging, and the environment.¹ Recent genome-wide association studies (GWAS) have identified more than 80 genetic variants linked to AD,² yet there remains a substantial gap in elucidating the full heritability of the disease. It is estimated that the currently known risk genes identified through GWAS can account for only ∼8% of the heritability.^3–6 Moreover, the majority of GWAS studies have focused on the relationship between AD risk and nuclear genes, neglecting exploration of the mitochondrial genome in AD patients.

Mitochondria are intracellular organelles responsible for producing energy through the oxidative phosphorylation (OXPHOS) pathway, which is essential for cell survival. The nervous system, due to its high energy demands, is particularly vulnerable to mitochondrial abnormalities. Multiple lines of evidence implicated mitochondria play a role in the etiology of AD.⁷ Mitochondrial dysfunction leads to an increase in reactive oxygen species (ROS), causing oxidative stress damage, a crucial hallmark preceding the occurrence of widespread plaque pathology in the brains of AD patients.⁸ Mitochondrial bioenergetic function is impaired in AD patients, as demonstrated by reduced activity of the respiratory chain, enzyme activity of the Krebs cycle, ATP production, and elevated levels of free radicals and ROS.⁹

Several studies have provided evidence supporting the genetic influence of mitochondrial DNA (mtDNA) variants in AD. Individuals with a maternal family history of AD were at a significantly higher risk of developing the disease compared to those with a paternal family history or no family history.¹⁰ Notably, individuals with a maternal history of AD exhibited more obvious endophenotypes, including age at onset,¹¹ cognitive performance,¹² gray matter volume,^13,14 and Aβ deposition load.¹⁵ The mitochondrial cascade hypothesis of AD further highlighted the substantial impact of mtDNA variants on the development of AD.¹⁶

Human mitochondria possess their own circular genome of 16,569 base pairs, known as the mitochondrial genome, with 93% of it being the coding region responsible for encoding a total of 37 genes. Throughout the origin and migration of modern humans, variants accumulate in mtDNA in a chronological order, generating a set of related haplotypes. Genetically, haplotypes are defined based on differences in mtDNA variants, known as mitochondrial haplogroups.¹⁷ Multiple studies have demonstrated a connection between mtDNA variants and the risk of AD. A study comprising 10,831 participants from the Alzheimer's Disease Sequencing Project cohort identified a significant correlation between AD risk and a rare MT-ND4L variant m.10733C > T as well as MT-ND4L gene in a gene-based test.¹⁸ Haplogroup J was much more prevalent in individuals with AD in the Caucasian populations, implying that mitochondrial haplogroup J could be a risk factor for AD.¹⁹ Due to the varying distribution of maternal mitochondrial haplogroups in different regions, a study on the mitochondrial haplogroups in southwestern China found that mitochondrial haplogroup B5 was associated with an increased risk of AD, although this association has not been validated in samples from eastern China.²⁰ These discrepancies indicate that the role of inherited mtDNA variants in AD has yet to be fully established.

In recent years, an increasing number of research has discovered that plasma biomarkers can serve as diagnostic markers for AD and, to some extent, reflect AD pathology.^21,22 They have value as biomarkers to increase understanding on the relationship between mitochondrial variations and AD pathology, thereby enhancing the knowledge of disease pathogenesis. However, no research has been conducted on the relationship between mtDNA variants and plasma biomarkers in AD patients.

Overall, previous research on the relationship of AD risk with mitochondrial genome variants were inconsistent and primarily focused on the Caucasian population. We conducted the first large-scale study of mitochondrial genome among AD patients from mainland China using high-depth whole genome sequencing (WGS), with the aim of identifying mitochondrial variants associated with AD in the Chinese population. In addition, we explored the relationship between mtDNA variants and plasma biomarkers in AD patients for the first time.

Methods

Participants

A total of 1509 Chinese Han patients with AD and 2010 healthy controls were enrolled in this study. AD patients were from the Department of Neurology at Xiangya Hospital, Central South University. These AD patients met the diagnostic criteria of “probable AD” in the 2011 NIA-AA guidelines.²³ The healthy controls were recruited from community cohorts in Hunan Province. There were no neurological diseases found in the healthy controls, and objective examinations such as Mini-Mental State Examination (MMSE) indicated normal cognition. This study followed the Declaration of Helsinki. The study was approved by the Ethics Committee of Xiangya Hospital of Central South University, and all participants or their legal guardians provided written informed consent. The flow chart of the study was shown in Figure 1.

Figure 1.

Flow chart in our cohort. AD: Alzheimer's disease; MAF: minor allele frequency; SNV: single nucleotide variants; Indel: Insertion-deletion; Hom: homoplastic variants; Het: heteroplasmic variants; Aβ: amyloid-β; P-tau: phosphorylated tau; NfL: neurofilament light; GFAP: glial fibrillary acidic protein; αSyn: α-synuclein; *p < 0.05.

Mitochondrial DNA variant calling, annotation, and validation

The genomic DNA was extracted from peripheral blood leukocytes through phenol-chloroform extraction and ethanol precipitation. All DNA samples underwent WGS. In this study, mitochondrial homoplasmic and heteroplasmic single nucleotide variants (SNVs) and small insertion-deletions (indels) (< 50 bp) were identified following the GATK best practices for mitochondrial pipeline (https://github.com/gatk-workflows/gatk4mitochondria-pipeline). The pipeline was executed using GATK v4.2.6.1, with the ChrM sequence from the reference genome GRCh38 (identical to rCRS (GenBank NC_012920.1)) as the mtDNA reference genome. Briefly, the chrM reads were first extracted and restored to unmatched BAM files using RevertSam software, after which the WGS data were aligned to the mtDNA reference genome using Burrow-Wheeler Aligner-Maximum Exact Match. In addition, we realigned the sequences to the mtDNA reference sequence with 8000 bases moved to call variants in the control region across the artificial cleavage of the circular genome. Variants were identified using the GATK Mutect2 variant caller and then parameterized based on a specific “mitochondrial pattern”.

MtDNA variants can affect either all copies of mtDNA within a cell (referred to as homoplasmy) or only a small fraction of the mtDNA molecules (referred to as heteroplasmy).²⁴ The variant allele fraction (VAF) is defined as the ratio of mutant allele reads to total reads for each variant and sample. Given the high number of false positives for variants with VAF < 0.10, we decided to only report variants with VAF ≥ 0.10.²⁵ False positive variants caused by nuclear mitochondrial DNA fragments (NuMTs)^26,27 were also eliminated. Finally, the variants from all samples were combined into a single Variant Call Format (VCF) file. MtDNA variants were annotated using ANNOVAR software,²⁸ and this study utilized various extensive population mitochondrial variant databases, including Mitomap, HelixMTdb, HmtDB, gnomAD, and 1000 Genomes.

Mitochondrial haplogroup classification and association analysis

Mitochondrial haplogroups were assigned utilizing HaploGrep2 software, with predictions based on the rCRS-oriented Phylotree build 17 for global human mtDNA.²⁹ Chi-square test was carried out to compare mitochondrial haplogroup frequencies between the AD group and the control group. Forest plots were generated using the odds ratios (OR) and corresponding 95% confidence intervals (95%CI). Logistic regression analysis of mitochondrial haplogroups was performed to identify AD-associated mitochondrial haplogroups, adjusted by gender and age. Following that, an association analysis was performed between AD-related mitochondrial haplogroups and haplogroup marker (present at frequency of 80% or higher in that specific haplogroup) to explore the possible mechanisms through which mitochondrial haplogroups affect the risk of AD. Statistical significance was determined by a p-value below 0.05.

Mitochondrial DNA variants association analysis

The statistical analyses were conducted using PLINK 1.9, python 3.11, R 4.4 and SPSS 26.0. Characteristics in accordance with normal distribution were analyzed using t test or otherwise Mann–Whitney U test. Stringent quality control filters were applied to exclude variants with minor allele count (MAC) < 10 or call rates < 80%.³⁰ Samples with low genotyping rate (< 80%), discrepancies between recorded sex and genetic sex, genetic relationships, and high or low heterozygosity rates (± 3 standard deviations from the mean) were also removed. Based on the minor allele frequency (MAF) of mitochondrial variants in this cohort, the filtered mitochondrial variants were categorized into common variants (MAF≥0.01) and rare variants (MAF<0.01). Mitochondrial genome association study (MiWAS) was conducted on common variants to explore the relationship between mtDNA variants and phenotypes. PLINK1.9 was used for common variant-based association analysis, with covariates including age, gender, and the first mtDNA principal component (mtPC) taken into account.

The MToolBox³¹ software was utilized to predict the pathogenicity of nonsynonymous variants in the mtDNA coding region, subsequently categorized as either “deleterious variants” or “neutral variants”. The pathogenicity of mitochondrial tRNA variants was assessed through PON-mt-RNA,³² resulting in classification as “pathogenic variants,” “likely pathogenic variants,” “neutral variants,” or “likely neutral variants”. Damaging variants were defined as “deleterious variants” categorized by MToolBox, or “pathogenic variants” and “likely pathogenic variants” classified by PON-mt-RNA. Loss-of-function (LoF) variants encompass nonsense, splicing, and frameshift variants. Rare variants were divided into two categories: rare damaging variants (MAF<0.01, damaging variants or LoF variants) and rare missense variants (MAF<0.01, missense variants). Gene-based associations for rare mitochondrial variants were conducted using the Sequence Kernel Association Test (SKAT) and Sequence Kernel Association Test-Optimal (SKAT-O), with age, gender, and the first mtPC as covariates. The Bonferroni method was employed for p-value correction, with statistically significant differences defined as p * n < 0.05 (where n represents the number of genes or variants analyzed).

To validate the association between the variant m.15758A > G in MT-CYB gene and AD risk, a substitution of the control cohort was conducted for validation purposes. Currently, five specialized mitochondrial databases have reported population frequencies, namely MitoMap,³³ HmtDB,³⁴ MSeqDR,³⁵ HelixMTdb,³⁶ and gnomAD. As the initial three mitochondrial DNA databases exclusively record homozygous variations, our research opted to select the control group from gnomAD.²⁵ Fisher's exact test was utilized to validate the association between the significant variant m.15758 A > G and AD risk.

Mitochondrial gene expression

Based on the Genotype-Tissue Expression (GTEx) project, our study delved into the expression of the MT-CYB gene. Subsequently, the microarray dataset GSE212606 was employed to discern differences in MT-CYB gene expression across specific brain regions, including the hippocampus, middle temporal gyrus, and superior temporal gyrus cortex, among cohorts of healthy individuals (N = 6) and AD patients (N = 6).³⁷ These datasets are accessible via the Gene Expression Omnibus (GEO) Website.

Plasma biomarker testing

A subgroup of 181 AD patients and 114 healthy controls underwent plasma biomarker assays. Participants’ venous blood samples were taken in ethylenediaminetetraacetic acid tubes and centrifuged at 3000 rpm for 15 min at 4°C within two hours of collection. Plasma samples were stored at −80°C and did not undergo a freeze-thaw cycle prior to testing. Plasma biomarker levels were quantified using a fully automated single-molecule detector (AST-Sc-Lite, AstraBio) following the manufacturer's instructions, with detailed parameters and procedures described in the previous study.³⁸ Plasma levels of Aβ40, Aβ42, phosphorylated tau 181 (p-tau181), p-tau217, p-tau231, neurofilament light (NfL), glial fibrillary acidic protein (GFAP), and α-synuclein (αSyn) were detected using kits R64010, R64020, R64030, R64100, R64050, R64040, and R64060 (AstraBioTechnology, Suzhou, China), respectively. The operation process of SMID in the Astra System is included in the Supplementary materials. The assayers were unaware of the grouping of the participants. Spearman partial correlation analysis was performed to assess the association between the number of mtDNA variants and plasma biomarker levels, adjusted by age, sex, and APOE ε4 status.

Results

Clinical demographic description

A total of 1509 AD patients and 2010 healthy controls were involved in this study. After quality control, the statistical analysis included 1448 AD patients (mean age at onset 64.52 ± 10.69 years) and 1832 healthy controls (mean age at enrollment 65.27 ± 8.39 years). There was a statistically significant difference in gender between the two groups (p < 0.001).

Association analysis of mitochondrial DNA common variants

In the analysis of common variants, a total of 318 mitochondrial SNVs were deemed to meet quality control standards and were subsequently utilized in association analyses. In the association analysis of common variants, thirteen mtDNA variants were found to be linked with AD risk following adjustments for gender, age, and the first mtPC (p < 0.05). These associations did not maintain statistical significance after multiple testing correction. It is worth noting that common variants tend to cluster upward at m.16183 variant (Figure 2A). Moderate deviations from anticipated association outcomes, seen as deflation in the QQ plot (Figure 2B, λ = 1.148), indicate that our p-values are higher than expected by random chance. Given that mtDNA is a small genome with closely linked variants, deflation may reflect random variation due to the limited number of independent effects.

Figure 2.

Association analysis of mtDNA common variants. (A) Solar Manhattan plot of the association p-values between mtDNA common variants and AD risk. Each dot represents an association color-coded by mitochondrial gene. The black circle indicates p = 0.05, and the red circle represents the Bonferroni-corrected p-value threshold of 0.05/318 (1.5723 × 10⁻⁴). (B) Quantile-quantile plot of association p-values between mtDNA variants and AD. The x-axis represents the expected -log10(p)-values, and the y-axis represents the observed -log10(p)-values. mtDNA: mitochondrial DNA; AD: Alzheimer's disease.

The study investigated the impact of previously documented AD-related mtDNA variants on Chinese AD patients. Nine of ten previously described AD-related mitochondrial variants were retained in our cohort. Common variant association analysis revealed that two of the variants, m.4833 A > G and m.9861 T > C, were nominally associated with AD (p < 0.05) (Table 1).

Table 1.

Ten previous reported AD related mtDNA variants in our cohort.

Gene	Location	Alleles	Function	OR	Adjusted p
MT-RNR1	MT:856	A > G	ncRNA_exonic	0	1
MT-ND1	MT:3397	A > G	nonsynonymous	0.464	0.361
MT-ND2	MT:4833	A > G	nonsynonymous	1.408	0.039*
MT-CO1	MT:5656	A > G	upstream; downstream	inf	0.441
MT-CO3	MT:9861	T > C	nonsynonymous	4.969	0.043*
MT-ND4	MT:11084	A > G	nonsynonymous	1.089	0.811
MT-ND5	MT:12372	G > A	synonymous	0.855	0.342
MT-ND5	MT:13759	G > A	nonsynonymous	0.860	0.304
MT-ND2	MT:5460	G > A	nonsynonymous	0.963	0.792
MT-ND4L	MT:10733	C > T	nonsynonymous	-	-

AD: Alzheimer's disease; mtDNA: mitochondrial DNA; Adjusted p-value referred to p-value adjusted by gender, age, and the first mtPC.

*Adjusted p-value < 0.05 was considered as statistically significant. “-” represented that the variant didn’t exist in this cohort.

Mitochondrial haplogroup classification and association analysis

We classified the mitochondrial haplogroups of the samples in this cohort and discovered that they were primarily classified into the known East Asian haplogroups, with haplogroup M and D being the most prevalent, followed by haplogroup B and F. Haplogroups with frequencies below 5% are classified as “others”. Chi-square test was utilized to compare the distribution of mitochondrial haplogroups between the AD group and the control group. It was found that the frequency of haplogroup G in the AD group (6.01%) exceeded that in the control group (4.15%) (Figure 3A, B), with a nominally significant difference (OR = 1.48, 95% CI = 1.08–2.03, p = 0.019, Figure 3C). Nevertheless, the significance was not retained after multiple testing corrections. Logistic regression analysis, adjusted by age and gender, revealed that the increased frequency of haplogroup G in the AD group remained nominally significant (p = 0.021). These findings indicated that haplogroup G could confer genetic susceptibility to AD in the Chinese population.

Figure 3.

Mitochondrial haplogroup classification and association analysis. (A) The distribution of haplogroups in the AD group. (B) The distribution of haplogroups in the NC group. (C) Forest plots showing OR and 95% CI for haplogroups. (D) Differential analysis of burden of mtDNA variants per sample between haplogroup G patients and haplogroup non-G patients. ns, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. AD: Alzheimer's disease; NC: normal control; OR: odds ratios; 95% CI: 95th percentile confidence intervals; mtDNA: mitochondrial DNA.

There is no notable difference in the age at onset of haplogroup G and other haplogroup in AD patients (p > 0.05, Supplemental Figure 1). However, the findings showed that there were considerably more females than males in haplogroup G (p = 0.036, Supplemental Figure 2). We performed mitochondrial haplogroup association analysis stratified by sex, and we discovered that the difference in haplogroup G among females was still significant (OR = 1.70, 95% CI = 1.14–2.53, p = 0.011, Supplemental Figure 3A). No significant differences were observed between various haplogroups in male participants (p > 0.05, Supplemental Figure 3B).

To explore the possible mechanism of mitochondrial haplogroup G affecting the risk of AD, we conducted a screening of haplogroup marker SNVs that define haplogroup G, including 489C, 709A, 4833G, 5108C, 10400 T, 14569A, 14783C, 15043A, 15301A, and 16362C (https://www.mitomap.org/foswiki/bin/view/ MITOMAP/HaplogroupMarkers). Analysis revealed a statistically significant difference in the frequency of the m.4833 A > G variant between the two groups (p = 0.039, Table 1), implying that the impact of haplogroup G on AD risk may be mediated by the m.4833 A > G variant. Additionally, the number of mtDNA variants per sample in the whole region and protein-coding region is much higher in AD patients with haplogroup G compared to haplogroup non-G AD patients (Figure 3D).

Gene-based association analysis of mitochondrial DNA rare variants

We conducted gene-based association tests to elucidate the association between mtDNA rare variants and the risk of AD, involving a total of 37 mitochondrial genes in the analysis. There was no significant correlation detected between rare damaging variants of mitochondrial genes and AD risk after Bonferroni correction (Figure 4A). The rare missense variants of MT-CYB gene were substantially associated with the risk of AD, and the missense variants of MT-CYB gene were significantly enriched in controls compared to AD patients (p = 2.811 × 10⁻⁴, OR = 0.886, Figure 4B).

Figure 4.

Gene-based association analysis of mtDNA rare variants. Solar Manhattan plot of the association p-values between mitochondrial genes and AD risk. (A) Gene-based analysis for rare damaging variants of mitochondrial protein-coding genes. (B) Gene-based analysis for rare missense variants of mitochondrial protein-coding genes. Each dot represents an association color-coded by mitochondrial gene. The black circle indicates p = 0.05, and the red circle represents the Bonferroni-corrected p-value threshold of 0.05/37 (3.85 × 10⁻³). mtDNA: mitochondrial DNA; AD: Alzheimer's disease.

Further analysis identified a total of 121 rare missense variants in the MT-CYB gene, three of which were associated with AD (p < 0.05, Table 2). Following corrections for multiple comparisons, the m.15758A > G variant exhibited a significant association with AD susceptibility (p = 6.02 × 10⁻⁴, Table 2). This variant had a frequency of 0.276% in the AD group and 1.419% in the control group. Nuclear DNA coding region variant pathogenicity prediction tools, including CADD, PolyPhen-2, SIFT, Mutpred, and PROVEAN, as well as mtDNA-specific tools like MToolBox, APOGEE2, and Mitoclass.1, all predict the variant to be beneficial or neutral.

Table 2.

The significant variants of MT-CYB gene between AD patients and controls.

Location	Alleles	Gene	Function	Exonic function	Freq in AD	Freq in controls	p	Freq in gnomAD	Validation p
MT:14927	A > G	MT-CYB	exonic	nonsynonymous	3.45 × 10⁻³ (5/1448)	0.011 (20/1832)	0.015	1.22 × 10⁻³ (69/56417)	0.038
MT:15651	C > T	MT-CYB	exonic	nonsynonymous	6.90 × 10⁻⁴ (1/1448)	0.007 (13/1832)	0.005	1.24 × 10⁻⁴ (7/56432)	0.180
MT:15758	A > G	MT-CYB	exonic	nonsynonymous	2.76 × 10⁻³ (4/1448)	0.014 (26/1832)	6.02 × 10⁻⁴	0.011 (639/56401)	5.41 × 10⁻⁴

AD: Alzheimer's disease; Freq: frequency.

In the replication stage, the control group was substituted with control populations sourced from the gnomAD databases to validate the correlation between the MT-CYB m.15758A > G variant and the susceptibility to AD. We found that the association between the m.15758A > G variant and the risk of AD remained significant even after changing the control group (p = 5.41 × 10⁻⁴, Table 2), indicating successful validation of the m.15758A > G variant.

Mitochondrial gene expression

Based on data from the GTEx project database, MT-CYB exhibits notable expression levels in 13 brain regions (Supplemental Figure 4). In order to explore potential alterations in MT-CYB gene mRNA expression within the brain tissue of AD patients, an analysis was conducted using the single-cell transcriptome dataset (GSE212606) sourced from the GEO database. The findings revealed a significant upregulation in the expression of MT-CYB mRNA in the hippocampus and middle superior temporal cortex of AD patients compared to the control group (Supplemental Table 1), which indicating that the mRNA expression of MT-CYB gene was up-regulated in the brain tissue of AD patients.

Association between the burden of mtDNA variants and plasma biomarkers

We calculated the number of various types of mtDNA variants including SNV, indel, homoplasmic, and heteroplasmic variants, and analyzed the relationship between the burden of mtDNA variants and plasma biomarkers after adjusting for age, sex, and APOE ε4 status. The number of mtDNA indel variants per individual was significantly positively correlated with plasma Aβ42 levels in AD patients (R = 0.18, p = 0.019, Figure 5A). Besides, there was a strong positive correlation between the number of mtDNA homoplasmic indel variants per individual and the plasma Aβ42 levels in AD patients (R = 0.19, p = 0.01, Figure 5B). However, no significant correlation was observed between the number of mtDNA indel variants and plasma Aβ42 levels in healthy controls (p > 0.05, Supplemental Figure 5). The relationship between the burden of other type mtDNA variants per sample and plasma biomarkers in AD patients was not significant (Figure 5C).

Figure 5.

Association between the burden of mtDNA variants per sample and plasma biomarkers in AD patients adjusted by age, sex, and APOE ε4 status. (A) Association between the number of mtDNA indel variants per individual and plasma Aβ42 levels. (B) Correlation between the number of mtDNA homoplasmic indel variants per individual and plasma Aβ42 levels. (C) Association between the burden of mtDNA variants per sample and the levels of AD plasma biomarkers. *p < 0.05. AD: Alzheimer's disease; SNV: single nucleotide variants; Indel: Insertion-deletion; Hom: homoplastic variants; Het: heteroplasmic variants; Aβ: amyloid-β; P-tau: phosphorylated tau; NfL: neurofilament light; GFAP: glial fibrillary acidic protein; αSyn: α-synuclein.

Discussion

Mitochondrial dysfunction exerts a critical role in the development of AD. Previous studies on the mitochondrial genome of AD were inconsistent and primarily based on the Caucasian population. This study is the first thorough investigation of mitochondrial genome variants based on high-depth WGS in the AD patients from mainland China. The findings revealed that mitochondrial haplogroup G may serve as a risk factor for AD in the Chinese population, potentially mediated by the m.4833 A > G variant. Rare variants of MT-CYB gene were found to be prominently enriched in the control group. Moreover, the control group exhibited significantly lower mRNA expression of the MT-CYB gene in brain regions compared to AD patients. In addition, we revealed for the first time that the burden of mtDNA indel variants was related to AD plasma biomarker levels.

We conducted a comprehensive analysis of the relationship between mtDNA common variants and the risk of AD, in which thirteen mtDNA common variants were nominally associated with AD risk. However, upon applying the Bonferroni correction, none of these common mtDNA variants exhibited statistical significance. Within our study cohort, ten previously documented mtDNA variants linked to AD risk were retained, with two of them displaying a nominal association with AD, including m.4833A > G and m.9861T > C. The m.4833A > G variant is a nonsynonymous variant in the exon region of the MT-ND2 gene. The MT-ND2 protein is the core subunit of mitochondrial membrane respiratory chain complex I, which activates NADH dehydrogenase (ubiquinone) activity and is involved in mitochondrial respiratory chain electron transfer, NADH to ubiquinone conversion, and assembly of respiratory chain complex I. The m.9861T > C variant is a nonsynonymous variant in the exon region of the MT-CO3 gene. The MT-CO3 protein is a component of cytochrome c oxidase, which is responsible for electron transfer activity, oxidoreduction-driven active transmembrane transport protein activity, and assembly of respiratory chain complex IV. Prior research predominantly focused on Caucasian populations, and the reliance of mtDNA variants on diverse maternal mitochondrial haplogroups suggested that ethnic disparities could impede the complete validation of previous findings.

Numerous studies have shown that mitochondrial haplogroups may contribute to susceptibility to neurodegenerative diseases, including AD, Parkinson's disease.³⁹ Some studies conducted on Caucasian populations have found that mitochondrial haplogroup J is a risk factor for AD patients in Caucasian populations.^19,40 Mitochondrial haplogroups form as a result of environmental changes and human migration, leading to their geographical distribution, with each region exhibiting a unique distribution. In this study, we found that mitochondrial haplogroup G may act as a risk factor for AD patients in southern China, potentially mediated by the m.4833 A > G variant. This finding aligns with a Japanese study comparing 96 AD patients and 96 centenarians found that mitochondrial haplogroup G2a is linked to AD risk, with the m.4833A > G variant contributing to this association.^41,42 Nevertheless, a study conducted by Yao group on 341 AD patients and 435 normal individuals in southwestern China revealed that mitochondrial haplogroup B5 was associated with an increased risk of AD, and these findings were not fully validated in a cohort comprising 371 AD patients and 470 normal individuals from eastern China.²⁰ Besides, a study conducted in Taiwan has found that individuals carrying mitochondrial haplogroup D exhibit poorer cognitive function.⁴³ The discrepancies in findings may be attributed to sample size limitations and different populations, highlighting the necessity for larger cohorts to explore mitochondrial haplogroups associated with AD in the Chinese population. Further validation and functional experimental studies are required to elucidate the role of mitochondrial haplogroup G in the pathogenesis of AD.

Oxidative stress is a contributing factor to the development of AD. Disruption of the respiratory chain complexes leads to an increase in the production of reactive oxygen species (ROS), resulting in oxidative stress. Excessive oxidative stress may play a role in neuronal degeneration and fibrin accumulation within the nervous system.⁴⁴ We found that there were fewer missense variants of MT-CYB gene in the AD group compared to controls, and the expression of MT-CYB was increased in the AD group. Several studies have demonstrated that missense variants of MT-CYB gene could affect the mitochondrial respiratory complex III activity.^45,46 The MT-CYB gene encodes the cytochrome b protein, which is a hydrophobic and conserved protein consisting of eight transmembrane helices. Cytochrome b, a component of the mitochondrial respiratory chain complex III (cytochrome c reductase complex), mediates in the transfer of electrons from coenzyme Q to cytochrome c in the mitochondrial respiratory chain, thereby establishing a proton gradient on the mitochondrial membrane for ATP synthesis. Diseases associated with MT-CYB gene variants include cardiomyopathy,⁴⁷ Leber hereditary optic neuropathy,⁴⁸ and Leigh syndrome⁴⁹ (https://www.mitomap.org/). The m.15758A > G (p.I338 V) variant is the most statistically significant among the rare missense variants in MT-CYB gene. In the Clinvar database,⁵⁰ the m.15758A > G variant was identified as a benign variant for Leigh syndrome based on the modified American College of Medical Genetics and Genomics guidelines. Additionally, research by Guo et al. demonstrated that the m.15885C > T variant in MT-CYB gene had a protective effect on spinal and hip bone mineral density.⁵¹ Moreover, research suggested that the mitochondrial variant m.10398A > G of MT-ND3 gene protected against PD.⁵² All in all, these findings indicated that certain mtDNA variants could be protective against age-related degenerative diseases.

This study reported that rare missense variants of the MT-CYB gene served as protective factors against AD. The expression of MT-CYB protein were elevated in the brains of AD patients. Overexpression of MT-CYB protein causes dysfunction of mitochondrial complex III, potentially leading to the increased electron leakage and subsequent generation of ROS, thereby inducing oxidative stress.⁵³ This increased oxidative stress was believed to facilitate the aggregation of Aβ, thereby heightening the susceptibility to AD.⁵³ Further investigations through functional experiments are necessary to elucidate the precise mechanism by which rare missense variants of the MT-CYB gene confer protective effects against AD.

Furthermore, this was the first study to investigate the relationship between mtDNA variants and AD plasma biomarkers. We observed a significant positive correlation between the number of mtDNA indel variants per sample and plasma Aβ42 levels in AD patients. Research suggests that plasma Aβ42 levels may reflect AD pathology, as they correlate with Aβ42 levels in cerebral fluid and Aβ-PET burden.²¹ Therefore, our finding implied that mtDNA indel variants could influence AD pathophysiology, hence contributing to AD pathogenesis. Recent studies have shown that mitochondrial DNA affects pathological states such as tau phosphorylation in AD.⁵⁴ The copy number of mtDNA in blood and the levels of AD plasma biomarkers, such as plasma NfL levels, were found to be significantly correlated in a recent study based on the Alzheimer's Disease Neuroimaging Initiative database.⁵⁵ These findings strengthened the relationship between mtDNA genetic risk and AD plasma biomarkers. However, due to the limited number of AD patients who have undergone plasma biomarker testing in our study, it is currently not feasible to conduct a comparative analysis of plasma biomarker levels between patients carrying haplogroup G and rare variants of MT-CYB gene and those without these variants. Further larger study is warranted to analyze the differences in plasma biomarkers between variant carriers and non-carriers to explore the impact of mtDNA variants on AD pathology.

This is the first large-scale investigation into the mitochondrial genome using high-depth WGS in the Chinese AD patients. This research still has certain constraints. We utilized blood samples instead of analyzing mitochondrial genomes from a variety of tissue samples due to difficulties in obtaining tissue samples. Comparing our WGS data against the GnomAD database may bring risks of batch effects, including discrepancies in capture and sequencing depth, even if the gnomAD sequences were likewise processed using the same GATK mitochondrial pipeline.²⁵ Furthermore, the results may be impacted by population stratification between our cohort and the general East Asian gnomAD population. Similarly, due to the fact that mitochondrial haplogroups are strongly impacted by regional factors, the observed difference in haplogroup G frequency between AD patients and controls may be influenced by population stratification. Additionally, the study lacked experimental validation of the bioinformatics analysis findings. Future research should incorporate mitochondrial genome analysis in muscle or brain tissue, along with functional studies to explore the mechanism in the AD pathogenesis.

Conclusion

In summary, this was the first study into the comprehensive investigation of mtDNA variants among AD patients in China. Our findings indicated that the mitochondrial haplogroup G may confer genetic susceptibility to AD. Conversely, the rare variants of MT-CYB gene could exert a potential protective role against AD. These findings highlighted the role of mitochondrial genome variants in AD risk.

Supplemental Material

sj-docx-1-alz-10.1177_13872877261442231 - Supplemental material for Association of mitochondrial genome variants with Alzheimer's disease in a Chinese population

Supplemental material, sj-docx-1-alz-10.1177_13872877261442231 for Association of mitochondrial genome variants with Alzheimer's disease in a Chinese population by Xiaoli Hao, Bin Jiao, Yijing Wang, Tianyan Xu, Qijie Yang, Yuan Zhu, Yiliang Liu, Cong Zhang, Xiaoyan Liang, Yafang Zhou, Xinxin Liao, Shilin Luo, Beisha Tang, Jinchen Li, Xuewen Xiao and Lu Shen in Journal of Alzheimer's Disease

Footnotes

Acknowledgements

We sincerely thank all of the participants for their commitment in participating and donating blood samples for this study. We are grateful for technical support from the Center for Computational Biology and Bioinformatics, Furong Laboratory and Bioinformatics Center, Xiangya Hospital, Central South University.

ORCID iD

Lu Shen

Ethical considerations

All human participants provided informed consent. The study was approved by the Ethics Committee of Xiangya Hospital of Central South University (equivalent to an Institutional Review Board), and this study followed the Declaration of Helsinki.

Consent to participate

Written informed consent was obtained from each participant or their legal representatives.

Consent for publication

Not applicable.

Author contribution(s)

Xiaoli Hao: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Software; Writing – original draft; Writing – review & editing.

Bin Jiao: Funding acquisition; Writing – review & editing.

Yijing Wang: Data curation; Investigation; Software.

Tianyan Xu: Data curation; Investigation.

Qijie Yang: Data curation; Investigation.

Yuan Zhu: Data curation; Investigation.

Yiliang Liu: Data curation; Investigation.

Cong Zhang: Data curation; Investigation.

Xiaoyan Liang: Data curation; Investigation.

Yafang Zhou: Data curation; Investigation.

Xinxin Liao: Data curation; Investigation.

Shilin Luo: Resources; Writing – review & editing.

Beisha Tang: Resources; Writing – review & editing.

Jinchen Li: Resources; Writing – review & editing.

Xuewen Xiao: Conceptualization; Funding acquisition; Resources; Writing – review & editing.

Lu Shen: Conceptualization; Funding acquisition; Resources; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the STI2030-Major Projects (2021ZD0201803), the National Natural Science Foundation of China (U22A20300, 82071216, 82371434), the Science and Technology Major Project of Hunan Province (2021SK1020), Outstanding Youth Fund of Hunan Provincial Natural Science Foundation (2024JJ2097), Youth Fund of Hunan Provincial Natural Science Foundation (2025JJ60696), Hunan Health Commission (20232460), Hunan Innovative Province Construction Project (2021SK1010), Hunan Provincial Natural Science Foundation of China (2023JJ40792), the Grant of National Clinical Research Center for Geriatric Disorders, Xiangya Hospital (2022LNJJ16), and the Postdoctoral Fellowship Program of CPSF (GZC20233185).

Outstanding Youth Fund of Hunan Provincial Natural Science Foundation, Hunan Health Commission, Youth Fund of Hunan Provincial Natural Science Foundation, National Natural Science Foundation of China, Grant of National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, STI2030-Major Projects, Hunan Innovative Province Construction Project, Hunan Provincial Natural Science Foundation of China, Science and Technology Major Project of Hunan Province, Postdoctoral Fellowship Program of CPSF, (grant number 2024JJ2097, 20232460, 2025JJ60696, U22A20300, 82071216, 82371434, 2022LNJJ16, 2021ZD0201803, 2021SK1010, 2023JJ40792, 2021SK1020, GZC20233185).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data are available on reasonable request to the corresponding author.

Supplemental material

Supplemental material for this article is available online.

References

Wingo

Lah

Levey

, et al. Autosomal recessive causes likely in early-onset Alzheimer disease. Arch Neurol 2012; 69: 59–64.

Reitz

Pericak-Vance

Foroud

, et al. A global view of the genetic basis of Alzheimer disease. Nat Rev Neurol 2023; 19: 261–277.

Bellenguez

Grenier-Boley

Lambert

. Genetics of Alzheimer's disease: where we are, and where we are going. Curr Opin Neurobiol 2020; 61: 40–48.

Sun

Kurki

Foley

, et al. Genetic associations of protein-coding variants in human disease. Nature 2022; 603: 95–102.

Xiao

Liu

Yao

, et al. LMTK2 And CRB1 are two novel risk genes for Alzheimer's disease in Han Chinese. J Prev Alzheimers Dis 2025; 12: 100087.

Chen

Petty

Sha

, et al. Genetically regulated expression in late-onset Alzheimer's disease implicates risk genes within known and novel loci. Transl Psychiatry 2021; 11: 618.

Wang

Zhao

, et al. Mitochondria dysfunction in the pathogenesis of Alzheimer's disease: recent advances. Mol Neurodegener 2020; 15: 30.

Fleck

Phu

Verschueren

, et al. PTCD1 Is required for mitochondrial oxidative-phosphorylation: possible genetic association with Alzheimer's disease. J Neurosci 2019; 39: 4636–4656.

Ashleigh

Swerdlow

Beal

. The role of mitochondrial dysfunction in Alzheimer's disease pathogenesis. Alzheimers Dement 2023; 19: 333–342.

10.

Edland

Silverman

Peskind

, et al. Increased risk of dementia in mothers of Alzheimer's disease cases: evidence for maternal inheritance. Neurology 1996; 47: 254–256.

11.

Gómez-Tortosa

Barquero

Barón

, et al. Variability of age at onset in siblings with familial Alzheimer disease. Arch Neurol 2007; 64: 1743–1748.

12.

Seshadri

Wolf

Beiser

, et al. Lifetime risk of dementia and Alzheimer's disease. The impact of mortality on risk estimates in the Framingham study. Neurology 1997; 49: 1498–1504.

13.

Honea

Swerdlow

Vidoni

, et al. Reduced gray matter volume in normal adults with a maternal family history of Alzheimer disease. Neurology 2010; 74: 113–120.

14.

Honea

Swerdlow

Vidoni

, et al. Progressive regional atrophy in normal adults with a maternal history of Alzheimer disease. Neurology 2011; 76: 822–829.

15.

Mosconi

Rinne

Tsui

, et al. Increased fibrillar amyloid-{beta} burden in normal individuals with a family history of late-onset Alzheimer's. Proc Natl Acad Sci U S A 2010; 107: 5949–5954.

16.

Swerdlow

Burns

Khan

. The Alzheimer's disease mitochondrial cascade hypothesis: progress and perspectives. Biochim Biophys Acta 2014; 1842: 1219–1231.

17.

Ruiz-Pesini

Mishmar

Brandon

, et al. Effects of purifying and adaptive selection on regional variation in human mtDNA. Science 2004; 303: 223–226.

18.

Zhang

Farrell

Tong

, et al. Association of mitochondrial variants and haplogroups identified by whole exome sequencing with Alzheimer's disease. Alzheimers Dement 2022; 18: 294–306.

19.

Liu

Zhang

Zhao

, et al. The heterogeneity among subgroups of haplogroup J influencing Alzheimer's disease risk. J Adv Res 2021; 33: 117–126.

20.

Zhang

, et al. Mitochondrial DNA haplogroup B5 confers genetic susceptibility to Alzheimer's disease in Han Chinese. Neurobiol Aging 2015; 36: 1604.e1607–1616.

21.

Nakamura

Kaneko

Villemagne

, et al. High performance plasma amyloid-β biomarkers for Alzheimer's disease. Nature 2018; 554: 249–254.

22.

Zettergren

Lord

Ashton

, et al. Association between polygenic risk score of Alzheimer's disease and plasma phosphorylated tau in individuals from the Alzheimer's disease neuroimaging initiative. Alzheimers Res Ther 2021; 13: 17.

23.

Jack

Jr Albert

Knopman

, et al. Introduction to the recommendations from the national institute on aging-Alzheimer's association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011; 7: 257–262.

24.

Klein

Trumpff

Yang

, et al. Characterization of mitochondrial DNA quantity and quality in the human aged and Alzheimer's disease brain. Mol Neurodegener 2021; 16: 75.

25.

Laricchia

Lake

Watts

, et al. Mitochondrial DNA variation across 56,434 individuals in gnomAD. Genome Res 2022; 32: 569–582.

26.

Wei

Schon

Elgar

, et al. Nuclear-embedded mitochondrial DNA sequences in 66,083 human genomes. Nature 2022; 611: 105–114.

27.

Soto-Calderón

Clark

Wildschutte

, et al. Identification of species-specific nuclear insertions of mitochondrial DNA (numts) in gorillas and their potential as population genetic markers. Mol Phylogenet Evol 2014; 81: 61–70.

28.

Wang

Hakonarson

. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 2010; 38: e164.

29.

van Oven

Kayser

. Updated comprehensive phylogenetic tree of global human mitochondrial DNA variation. Hum Mutat 2009; 30: E386–E394.

30.

Yonova-Doing

Calabrese

Gomez-Duran

, et al. An atlas of mitochondrial DNA genotype-phenotype associations in the UK Biobank. Nat Genet 2021; 53: 982–993.

31.

Calabrese

Simone

Diroma

, et al. MToolbox: a highly automated pipeline for heteroplasmy annotation and prioritization analysis of human mitochondrial variants in high-throughput sequencing. Bioinformatics 2014; 30: 3115–3117.

32.

Niroula

Vihinen

. PON-mt-tRNA: a multifactorial probability-based method for classification of mitochondrial tRNA variations. Nucleic Acids Res 2016; 44: 2020–2027.

33.

Ruiz-Pesini

Lott

Procaccio

, et al. An enhanced MITOMAP with a global mtDNA mutational phylogeny. Nucleic Acids Res 2007; 35: D823–D828.

34.

Clima

Preste

Calabrese

, et al. HmtDB 2016: data update, a better performing query system and human mitochondrial DNA haplogroup predictor. Nucleic Acids Res 2017; 45: D698–d706.

35.

Shen

Diroma

Gonzalez

, et al. MSeqDR: a centralized knowledge repository and bioinformatics web resource to facilitate genomic investigations in mitochondrial disease. Hum Mutat 2016; 37: 540–548.

36.

Bayona-Bafaluy

López-Gallardo

Emperador

, et al.

Is population frequency a useful criterion to assign pathogenicity to newly described mitochondrial DNA variants?

Orphanet J Rare Dis 2022; 17: 316.

37.

Sziraki

Lee

, et al. A global view of aging and Alzheimer's pathogenesis-associated cell population dynamics and molecular signatures in human and mouse brains. Nat Genet 2023; 55: 2104–2116.

38.

Weng

Zhang

, et al. Genetic spectrum features and diagnostic accuracy of four plasma biomarkers in 248 Chinese patients with frontotemporal dementia. Alzheimers Dement 2024; 20: 7281–7295.

39.

Hudson

Nalls

Evans

, et al. Two-stage association study and meta-analysis of mitochondrial DNA variants in PARKINSON disease. Neurology 2013; 80: 2042–2048.

40.

Swerdlow

Hui

Chalise

, et al. Exploratory analysis of mtDNA haplogroups in two Alzheimer's longitudinal cohorts. Alzheimers Dement 2020; 16: 1164–1172.

41.

Takasaki

. Mitochondrial haplogroups associated with Japanese Alzheimer's patients. J Bioenerg Biomembr 2009; 41: 407–410.

42.

Takasaki

. Mitochondrial SNPs associated with Japanese centenarians, Alzheimer's patients, and Parkinson's patients. Comput Biol Chem 2008; 32: 332–337.

43.

Yang

Liu

, et al. Mitochondrial DNA haplogroup D and brain microstructure regulate cognitive function among community-dwelling older adults. Arch Gerontol Geriatr 2024; 117: 105197.

44.

Sherer

Betarbet

Stout

, et al. An in vitro model of Parkinson's disease: linking mitochondrial impairment to altered alpha-synuclein metabolism and oxidative damage. J Neurosci 2002; 22: 7006–7015.

45.

Tropeano

La Morgia

Achilli

, et al. Adult leigh syndrome associated with the m.15635T>C mitochondrial DNA variant affecting the cytochrome b (MT-CYB) gene. Int J Mol Sci 2025; 26: 1116.

46.

Petrovic Pajic

Fakin

Jarc-Vidmar

, et al. Atypical Leber hereditary optic neuropathy (LHON) associated with a novel MT-CYB:m.15309T>C(Ile188Thr) variant. Genes (Basel) 2025; 16: 108.

47.

Zarrouk

Finsterer

Mehri

, et al. Dilated cardiomyopathy due to the novel MT-CYB missense mutation m.14757T>C. J Med Cases 2021; 12: 455–459.

48.

Jancic

Rovcanin

Djuric

, et al. Analysis of secondary mtDNA mutations in families with Leber's hereditary optic neuropathy: four novel variants and their association with clinical presentation. Mitochondrion 2020; 50: 132–138.

49.

Ronchi

Cosi

Tonduti

, et al. Clinical and molecular features of an infant patient affected by Leigh disease associated to m.14459G>A mitochondrial DNA mutation: a case report. BMC Neurol 2011; 11: 85.

50.

Landrum

Lee

Benson

, et al. Clinvar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res 2018; 46: D1062–d1067.

51.

Guo

Yang

Liu

, et al. Mitochondria-wide association study of common variants in osteoporosis. Ann Hum Genet 2011; 75: 569–574.

52.

van der Walt

Nicodemus

Martin

, et al. Mitochondrial polymorphisms significantly reduce the risk of Parkinson disease. Am J Hum Genet 2003; 72: 804–811.

53.

Cottrell

Borthwick

Johnson

, et al. The role of cytochrome c oxidase deficient hippocampal neurones in Alzheimer's disease. Neuropathol Appl Neurobiol 2002; 28: 390–396.

54.

Kemna

Kueck

Honea

, et al. Mitochondrial DNA affects tau phosphorylation in aging and Alzheimer's disease. Alzheimers Dement 2025; 21: e70390.

55.

Tong

Zhu

Farrell

, et al. Blood-derived mitochondrial DNA copy number is associated with Alzheimer disease, Alzheimer-related biomarkers and serum metabolites. Alzheimers Res Ther 2024; 16: 234.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

17.81 MB