Elevated mtDNA copy number in older adults is linked to methylation of mitochondrial and nuclear regulatory regions

Abstract

Growing evidence shows that epigenetic modification and mitochondrial dysfunction are hallmarks of aging and are associated with the development of a wide range of age-related diseases. Mitochondrial biogenesis, which is marked by mitochondrial DNA copy number (mtDNAcn), is one of the major regulations of mitochondrial function by a set of transacting elements, including mitochondrial DNA polymerase gamma (POLG), working on the mtDNA control region. In this study, we investigated the mtDNAcn and the methylation status at both mtDNA control and POLGA promoter regions in human blood cells from individuals with a wide range of ages. A total of 119 blood samples were collected, including 24 umbilical cord blood samples from newborns and 95 peripheral blood samples from individuals aged 18 to 96 years. We observed an increase in mtDNAcn, as well as a rise in the methylation levels of the mtDNA control region during aging, particularly in subjects aged ≥ 45. In addition, a positive correlation was also found between the methylation levels of the 4th CpG site in the POLGA promoter region and mtDNAcn during aging. These results suggest epigenetic regulation at mitochondrial and nuclear genes for mitochondrial biogenesis during aging in human blood cells.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s11357-025-02037-2.

Introduction

Aging is a complex biological process that renders organisms more susceptible to diseases and ultimately contributes to their demise. After developmental maturation, it is characterized by a time-dependent decline or loss of function and an accumulation of damage at the molecular, cellular, tissue, and organismal levels. It is driven by a combination of genetic, environmental, and lifestyle factors [1]. The major hallmarks of aging include genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, and altered intercellular communication, among others [2]. These hallmarks are interconnected, and perturbations in one can affect the others [3].

Epigenetics encompasses a broad range of mechanisms that regulate gene expression without altering the DNA sequence. DNA methylation, a key epigenetic modification where a methyl group is added to the DNA molecule, typically at cytosine bases, plays a significant role in regulating gene expression through gene silencing. During aging, DNA methylation patterns undergo notable changes, with certain genes becoming hypermethylated (silenced) or hypomethylated (activated), potentially leading to altered gene expression that could contribute to aging and age-related diseases. These epigenetic changes are believed to influence the function of critical genes involved in inflammation, stress response, and cellular repair [4, 5]. Furthermore, epigenetic clocks, which measure changes in a few hundred specific CpG sites, are proposed to predict chronological age in a variety of species, including humans [6].

Mitochondria harbor their own DNA, mtDNA, which is a double-stranded circular molecule of 16,569 base pairs with 37 genes essential for normal mitochondrial function [7]. One of the key regions of mtDNA is the control region, a non-coding area that contains the origin of replication and plays a vital role in the regulation of mtDNA replication and transcription. The control region is essential for maintaining the copy number of mtDNA, as it controls the replication process. The process of mtDNA replication is regulated to ensure that each cell has an adequate number of mtDNA. In most cells, mtDNA is replicated through a complex process involving the mtDNA polymerase (POLG), transcription factor A (TFAM), and other regulatory proteins. These factors help initiate replication and ensure that the mitochondria maintain sufficient copies of mtDNA for optimal function.

Mitochondria are metabolic and signaling hubs that integrate a plethora of interconnected processes to maintain cell homeostasis [8]. Mitochondrial dysfunction increases with age [9]. Disruptions in respiratory activities, mutations in mtDNA, and changes in mitochondrial biogenesis have been implicated in the progression of aging [10]. mtDNA copy number (mtDNAcn), the levels of mitochondrial genome abundance per cell, is considered a key marker for mitochondrial biogenesis, and it often can vary within a range in a specific tissue [11]. In general, cells with high energy demands, such as muscle and brain cells, contain more copies of mtDNA, suggesting it is closely correlated with mitochondrial function [12]. However, the correlation between aging and mtDNAcn is still a controversial topic. On one hand, mtDNAcn was reported to be decreased in pancreatic islets [13], peripheral blood cells [14, 15], and muscle [16, 17] with aging, and that in turn was suggested to be associated with increased incidence of several aging-related diseases, including cancers [18], cardiovascular disease [19–21], neurodegenerative disorders [22], and diabetes [21], where mitochondrial dysfunction has been recorded. On the other hand, it was also described that mtDNAcn may increase in muscle [23, 24] and lung [25] with age. In addition, no significant changes in mtDNAcn during aging was claimed from studies in muscle [26], fibroblast [27], Cortex cerebis, cortex putamen, muscle iliopsoas, and cardiac muscle [28].

Here, we analyzed mtDNAcn, and DNA methylation levels in the mtDNA D-loop region and promoter region of nuclear DNA encoded POLG in peripheral blood cells from a wide range of people at different ages.

Materials and methods

Samples collection

The research subjects are healthy volunteers of different ages. A total of 24 umbilical cord blood samples from newborns were collected from the Department of Obstetrics at the First Affiliated Hospital of Wenzhou Medical University. Fifty-five peripheral blood samples from healthy young and middle-aged donors were collected from the Physical Examination Center of the same hospital between May 2012 and November 2013. Additionally, 40 peripheral blood samples from older donors were collected from the Wenzhou Senior Citizen Sports Center. All volunteers had no cancer history, neurodegenerative diseases, or other known diseases associated with mitochondrial defects. Informed consent from all donors in this study was obtained under protocols approved by the Wenzhou Medical University Ethics Committee.

DNA extraction

Whole blood samples were digested with Red Blood Cell Lysis buffer (20 mM Tris/HCl, pH 7.6) to collect leukocyte debris. Cell lysis was performed overnight at 55 °C in TES lysis buffer (10 mM Tris/HCl, pH 8.0; 2 mM EDTA, pH 8.0; 0.5% SDS) and 60 µg proteinase K. The following day, DNA was isolated as described previously. In short, lysed cells were incubated on ice for 15 min with 3 M KAc in a 2.5:1 ratio. This mixture was centrifuged for 15 min at 12,000 rpm at 4 °C to precipitate the protein/SDS complex. The supernatant was mixed with an equal volume of isopropanol and incubated for 60 min at −20 °C, followed by centrifugation for 15 min at 12,000 rpm at 4 °C. Total genomic DNA (nuclear and mitochondrial DNA) was washed with 75% ethanol and dissolved with TE buffer. The concentrations of the extracted DNA were quantified using a Nanodrop 2000 spectrophotometer (Thermo Scientific).

Quantification of mtDNA copy number

The mtDNAcn was evaluated based on the ratio of a mitochondrial region to a single-copy nuclear control gene using real-time polymerase chain reaction (QPCR) analysis (ABI Step-One Plus system, Applied Biosystems). Mitochondrial NADH dehydrogenase 1 (MTND1) and GAPDH were selected as the mitochondrial gene and nuclear gene, respectively. The long fragment PCR products containing the two target genes were separately purified and cloned into the pMD-18 T vector to construct recombinant plasmids. A serial of gradient templates in each reaction were diluted 1:100,000, 1:10,000, 1:1000, 1:100, 1:10, and 1:1 to establish standard curves (mtDNA and nDNA R² ≥ 0.99). Genomic DNA (10 ng) was used as a template per 20 µl reaction with Premix Ex Taq™ (TaKaRa) in the Probe QPCR method. The cycling conditions were as follows: 95 °C for 30 s (pre-denaturation), followed by 40 cycles of two steps: 95 °C for 5 s (denaturation) and 58 °C for 30 s (annealing and extension). Each sample was performed in triplicate, and the mean of the threshold cycles values (Ct) was used for analysis. Negative controls (in which ddH₂O instead of DNA was added) were also run in each plate. The relative mtDNAcn was calculated as the following formula: 2^ΔCt (ΔCt = Ct_GAPDH − Ct_MTND1).

Bisulphite treatment

Genomic DNA was bisulfite-converted using the EZ DNA methylation Gold kit (Zymo Research) according to manufacturer’s instructions. Briefly, 400 ng DNA was added with 130 µl of CT Conversion Reagent to a final volume of 150 µl. The mixture was incubated at 98 °C for 10 min and then at 64 °C for 3 h. After adding 600 µl of M-Binding Buffer into a Zymo-Spin IC Column, each sample was loaded into the column and centrifuged at 16,000 g for 30 s. After adding 100 µl of M-Wash Buffer to the columns and centrifuging at 16,000 g for 30 s, 200 µl of M-Desulphonation Buffer was added to the columns and incubated at room temperature for 20 min. The solution was then removed by centrifugation at 16,000 g for 30 s, and the columns were washed twice with 200 µl of M-Wash Buffer. Deaminated DNA was eluted in 10 µl of M-Elution Buffer.

For each bisulfite treatment procedure, unconverted primers randomly covering the entire mtDNA molecule, including those used for control analysis, were used in polymerase chain reaction (PCR) reactions as negative controls.

Primer design for PCR reactions

We designed primers specifically to amplify mtDNA while avoiding the contamination from nuclear mitochondrial sequences using MethPrimer software and Primer Premier 5.0. Several considerations were taken into account for primer design: (a) cytosines within CpG sites were avoided; (b) cytosines in forward primers and guanines in reverse primers were replaced with thymines and adenines, respectively; (c) the amplicon size was kept short (range: 150–350 bp); (d) when possible, DNA regions characterized by low polymorphism were preferred. Three pairs bisulfite-specific primers for the mitochondrial control and MT-RNR1 genes, as well as nuclear POLGA gene, were designed, respectively (Supplementary Table 1).

Bisulfite sequencing

Bisulfite sequencing PCR is considered the gold standard method for methylation detection. The PCR mixture (25 µl) contained 3 µl of bisulphite-treated DNA, 2.5 µl 10 × Ex Taq Buffer, 1.75 mM MgCl₂, 0.3 mM of each primer, 0.2 mM dNTP mix, and 0.75 U of Ex Taq enzyme (TaKaRa). The thermal profile for the reaction included a 4-min heat activation of the enzyme at 95 °C, followed by 45 cycles of denaturation at 94 °C for 20 s, annealing at different temperatures (Supplementary Table 1) for 30 s, and extension at 72 °C for 1 min, with one final cycle at 72 °C for 3 min.

The obtained PCR products were purified with a Gel/PCR Extraction Kit (BioMiga) and cloned into the pMD18-T vector (Thermo Scientific) according to the manufacturer’s protocol. The recombinant plasmids were transformed into E. coli DH5α, followed by blue-white screening and restriction enzyme identification. Positive individual clones purified with MiniBEST Plasmid Purification Kit (TaKaRa) were sent for sequencing to BGI Genomics (Shenzhen, China). Bisulfite sequencing results were analyzed using the online tool QUMA (www.quma.cdb.riken.jp/) [55] and CodonCode software (https://www.codoncode.com/aligner/download.htm).

Statistical analysis

All statistical analyses were performed using SPSS software (version 25.0) (SPSS Inc.,Chicago, IL, USA) and Microsoft Excel. The mtDNA copy number (log values) and methylation levels in all samples were determined using one-way or two-way ANOVA in combination with nonparametric tests. To adjust for the contribution of age, regression analysis were also performed. Figures were generated using GraphPad PRISM version 8.0.2. The p values were two-tailed, and statistical significance was defined as *p < 0.05, **p < 0.01, ***p < 0.001.

Results

The alterations of mtDNA copy numbers during aging

To determine if mtDNAcn in human blood samples can be a biomarker for aging, we measured mtDNAcn in DNA samples extracted from blood cells of donors with wide range of ages using the QPCR method as described in the “Methods” section. As shown in Fig. 1, we found that an increase in mtDNA copy number occurred with aging, especially in populations aged > 80 years (R² = 0.474, p < 0.0001, Fig. 1A). Since sex is a significant biological variance, we further analyzed the mtDNAcn changes in both female and male populations, and we found that it increased in both female (Fig. 1B) and male populations (Fig. 1C).

Fig. 1.

The variations of mtDNA copy number (log values) with age in total sample (A), in female (B) and male population (C). (*p < 0.05, **p < 0.01, ***p < 0.001)

The methylation of the mtDNA D-loop region in aging

The mtDNA displacement loop (D-loop) is a DNA structure where a short DNA strand displaces one of the two strands of a double-stranded DNA molecule, creating a three-stranded region. It is located in the mtDNA control region and contains the heavy strand origin of replication (OH), the light strand origin of replication (OL), and several conserved sequence blocks necessary for the proper regulation of mtDNA replication. DNA methylation is a major epigenetic modification that can regulate gene expression. This modification typically occurs at CpG dinucleotides, and it typically leads to blocking or recruiting the binding of regulatory proteins to the promoter region. To explore the underlying mechanism of age-related change of mtDNAcn and the potential role of this epigenetic alteration, we first investigated the DNA methylation status at the mtDNA D-loop region in the blood samples from different age groups. Interestingly, we detected the presence of methylated cytosines not only in CpG nucleotides but also in residues located outside of CpG nucleotides (non-CpG pattern, CpH) across age groups (Supplementary Fig. 1). As shown in Fig. 2A, a variable median percentage of methylation (the percentage of methylated cytosines to the overall cytosine content of the analyzed sequence) was 24.29% (range 15.27–43.18%) observed in DNA samples from old-aged donors, which is significantly higher than that in middle-aged (0.75% (range 0.39–25.68%)), young (0.49% (range 0.26–2.06%)), and newborn donors (0.37% (range 0.22–5.79%)). Similarly, significantly higher methylation levels at CpH sites (22.86% (range 14.60–38.28%)) and CpG sites (20.19% range (12.50–39.15%)) were detected in old-aged donors (Fig. 2B, C).

Fig. 2.

DNA methylation levels of mtDNA control region in four age groups: newborn (< 1 month), young (18 ≤ age < 45), middle-aged (45 ≤ age < 60), and old (60 ≤ age < 96). The total methylation percentage (percentage median (third percentile–first percentile) (A), methylated CpGs (B), and methylated CpHs (C). (*p < 0.05, **p < 0.01, ***p < 0.001)

Further, regression analysis also demonstrated a positive correlation between CpG methylation and age when we pooled all samples together (R² = 0.248, p = 0.003, Fig. 3A). However, when we divided the samples into two age groups, we found that the correlation is largely driven by subjects’ age ≥ 45 (R² = 0.349, p = 0.013, Fig. 3B), while no significant correlation was observed in subjects aged < 45 (R² = 0.010, p = 0.703, Fig. 3C). A similar relationship was observed for the CpH methylation in total sample (R² = 0.245, p = 0.003, Fig. 3D), subjects aged ≥ 45 (R² = 0.382, p = 0.008, Fig. 3E) and subjects aged < 45 (R² = 0.001, p = 0.885, Fig. 3F).

Fig. 3.

The relationship between the methylation levels of mtDNA control region and age. Methylation of CpGs in total sample (A), in age ≥ 45 subjects (B), and in age < 45 subjects (C). Methylation of CpHs in total sample (D), in age ≥ 45 subjects (E), and in age < 45 subjects (F)

The methylation of the mtDNA D-loop region and mtDNAcn during aging

To explore the potential role of methylation of the mtDNA D-loop region on mtDNAcn during aging, we plot the mtDNAcn against the methylation level. Interestingly, we observed that only in subjects aged ≥ 45, both the CpG and CpH methylations showed positive correlations with mtDNA copy number (R² = 0.600, p = 0.005, Fig. 4A; R² = 0.613, p = 0.004, Fig. 4B, respectively), while no significant correlation was found in subjects aged < 45 (R² = 0.000, p = 0.968, Fig. 4C; R² = 0.001, p = 0.972, Fig. 4D, respectively).

Fig. 4.

The relationship between the methylation levels of mtDNA control region and mtDNA copy number (log values). Methylation of CpGs (A) and methylation of CpHs (B) in age ≥ 45 subjects. Methylation of CpGs (C) and methylation of CpHs (D) in age < 45 subjects

The changes in methylation of the promoter region of POLG gene during aging

While the D-loop region contains the cis-acting element for mtDNA replication and transcription, the mtDNA polymerase gamma (POLG) is a major transacting regulator in mtDNA replication. POLG is encoded by the POLG gene, which is located in the nuclear genome and produces two primary subunits: POLG and POLG2 [29]. POLG is the catalytic subunit, and POLG2 functions as a regulatory subunit. The promoter region of the POLG gene is the key regulatory element responsible for controlling the transcription of the POLG gene. The POLG promoter is suggested to interact with a variety of transcription factors, such as NRF-1 (Nuclear Respiratory Factor-1) and CAAT-Box Binding Factors, involved in cellular stress response or energy metabolism. In parallel to mtDNA, we analyzed the methylation status of the nuclear DNA encoded POLG promoter region. We identified a 222 bp fragment enriched with 15 CpG sites (Supplementary Fig. 2) within the POLG promoter region and analyzed the methylation using the bisulfite sequencing PCR method. As we did with the mtDNA D-loop region, we compared the methylation level of the POLG promoter regions from different age groups. As shown in Fig. 5 and Supplementary Fig. 3, we found the lowest levels of DNA methylation were detected in umbilical cord blood from newborns (12.56% ± 2.19), which is significantly lower than that observed in middle-aged donors (22.63% ± 1.74) and older donors (22.81% ± 1.86) (Fig. 5A). However, we did not detect a significant difference among the young, middle-aged, and older groups. Nevertheless, linear regression analysis showed a positive correlation between the methylation of the promoter region of the POLG gene and age (Fig. 5B). Specifically, among these CpG sites, we found the age-dependent methylations more likely occurred at two areas: 2nd, 4th, 5th; and 11th, 13th, 14th, 15th CpG sites increased their methylation levels to different extents with age (Supplementary Fig. 4, A–O). Further multiple linear regression results showed that the regression equation was significant (F = 2.258, p = 0.034). Among them, only the 4th CG methylation was correlated with age (β = 0.330, p = 0.031), as shown in Table 1.

Fig. 5.

DNA methylation levels of the promoter region of POLG gene among different age groups (A) and the relationship between the total methylation levels of the promoter region of POLG gene and age (B). (*p < 0.05, **p < 0.01, ***p < 0.001)

Table 1.

The results of multiple linear regression analysis of the relationship between age and the methylation levels of different CG sites in the promoter of POLG gene

CG site	B	SE	β	t	p	Lower 95% CI	Upper 95% CI	F	Adjusted R2	p
1st CG	− 0.212	0.545	− 0.064	− 0.389	0.700	− 1.312	0.888	2.058	0.221	0.034
2nd CG	0.616	0.450	0.229	1.368	0.179	− 0.293	1.525
3rd CG	− 0.179	0.241	− 0.107	− 0.742	0.462	− 0.667	0.309
4th CG	1.297	0.582	0.330	2.231	0.031	0.123	2.472
5th CG	0.321	0.418	0.138	0.769	0.446	− 0.522	1.165
6th CG	− 0.250	0.531	− 0.084	− 0.471	0.640	− 1.322	0.823
7th CG	− 0.018	0.305	− 0.010	− 0.060	0.953	− 0.634	0.598
8th CG	− 0.293	0.515	− 0.110	− 0.568	0.573	− 1.334	0.748
9th CG	− 0.120	0.323	− 0.067	− 0.371	0.713	− 0.773	0.533
10th CG	0.029	0.527	0.011	0.054	0.957	− 1.035	1.093
11th CG	0.073	0.415	0.034	0.176	0.861	− 0.766	0.912
12th CG	− 0.045	0.575	− 0.014	− 0.078	0.938	− 1.206	1.116
13th CG	0.585	0.493	0.208	1.187	0.242	− 0.410	1.581
14th CG	0.200	0.369	0.098	0.542	0.591	− 0.545	0.945
15th CG	0.451	0.476	0.154	0.949	0.348	− 0.509	1.412

R² = 0.430; adjusted R² = 0.221; F = 2.058. p = 0.034

B unstandardized coefficient, SE standard error, β standardized coefficient, t t-test, CI confidence interval, R² coefficient determination

The methylation of the POLG promoter region and mtDNAcn during aging

To explore the potential role of methylation at the POLG promoter region in the mtDNAcn control during aging, we conducted a linear regression analysis. We found a correlation between the total methylation levels of POLG promoter region and mtDNAcn in the total sample (R² = 0.105, p = 0.019, Fig. 6A). In contrast to what we found with the mtDNA D-loop region, no significant correlation was found when we divided the samples into two age groups: aged < 45 (R² = 0.017, p = 0.501, Fig. 6B) or aged ≥ 45 (R² = 0.031, p = 0.420, Fig. 6C).

Fig. 6.

The relationship between mtDNA copy number (log values) and the methylation levels of the promoter region of the POLG gene. Total POLG methylation in total sample (A), in age < 45 subjects (B), and in age ≥ 45 subjects (C)

We then analyzed the correlation of the methylation of each CpG site to the copy number. Interestingly, among these CpG sites, we found that the methylation at 2nd, 4th, 14th, and 15th was correlated to mtDNAcn in the total sample separately according to the single regression analysis (Supplementary Fig. 5, A–O). Further, multiple linear regression results showed that the regression equation was not significant (F = 1.550, p = 0.135). Among these CpG sites, however, the 4th CG methylation exhibited a positive correlation with mtDNAcn (β = 0.416, p = 0.013), as shown in Table 2.

Table 2.

The results of multiple linear regression analysis of the relationship between mtDNA copy number (log values) and the methylation levels of different CG sites in the promoter of the POLG gene

CG site	B	SE	β	t	p	Lower 95% CI	Upper 95% CI	F	Adjusted R2	p
1st CG	0.000	0.011	0.007	0.042	0.967	− 0.022	0.023	1.550	0.133	0.135
2nd CG	0.008	0.009	0.151	0.816	0.419	− 0.011	0.027
3rd CG	− 0.003	0.005	− 0.114	− 0.734	0.467	− 0.013	0.006
4th CG	0.030	0.012	0.416	2.616	0.013	0.007	0.054
5th CG	0.005	0.008	0.109	0.562	0.578	− 0.012	0.022
6th CG	− 0.010	0.011	− 0.172	− 0.902	0.373	− 0.032	0.012
7th CG	− 0.008	0.006	− 0.242	− 1.279	0.209	− 0.020	0.005
8th CG	0.000	0.010	− 0.002	− 0.011	0.991	− 0.021	0.023
9th CG	0.004	0.007	0.127	0.620	0.539	− 0.010	0.018
10th CG	0.003	0.011	0.060	0.269	0.790	− 0.019	0.025
11th CG	0.000	0.009	− 0.012	− 0.055	0.956	− 0.018	0.017
12th CG	− 0.001	0.013	− 0.011	− 0.054	0.957	− 0.027	0.025
13th CG	0.006	0.010	0.107	0.558	0.580	− 0.015	0.026
14th CG	0.002	0.008	0.063	0.321	0.750	− 0.013	0.018
15th CG	0.010	0.012	0.159	0.827	0.413	− 0.014	0.034

R² = 0.373; adjusted R² = 0.133; F = 1.550; p = 0.135

B unstandardized coefficient, SE standard error, β standardized coefficient, t t-test, CI confidence interval, R² coefficient determination

Discussion

Mitochondrial alteration and epigenetic changes have been suggested as primary hallmarks for aging. In the present study, we investigated mtDNAcn, DNA methylation levels in the mtDNA D-loop region, and the promoter region of the nuclear POLG gene at different age groups: newborns, young, middle-aged, and older populations. In the population studied in this work, an increase in mtDNAcn with aging was observed. In addition, a positive correlation between mtDNA methylation during aging was found with the methylation levels of the control region in subjects aged ≥ 45, while a positive correlation between the methylation levels of the POLG promoter region and mtDNAcn is also recorded.

mtDNAcn serves as a surrogate marker for mitochondrial biogenesis. Changes in mtDNAcn are often interpreted as indicative of alterations in mitochondrial biogenesis. While higher mtDNAcn generally reflects greater mitochondrial content or a response to increased energy demand, lower mtDNA copy number is often associated with impaired biogenesis and mitochondrial dysfunction. Interestingly, environmental insults or cellular stressors, including oxidative stress, hypoxia, toxins and pollutants, inflammation, and infection, can stimulate or impair mitochondrial biogenesis depending on the nature, intensity, and duration of the stress. In this study, we observed a dramatic increase in mtDNAcn with age from young to old, especially in those aged older than 80 years. Previous studies that measured mtDNAcn in various tissues across different ages showed conflicting results. Multiple reports recorded reductions in mtDNAcn in elderly individuals, including in the lymphocytes of Sardinian participants [14] and in peripheral blood from a Danish population [15]. However, mtDNAcn in nonagenarians and centenarians showed a significant increase compared to elderly individuals aged between 50 and 70 years [30]. While multiple accumulated stresses could contribute to the observed rise in mtDNAcn, it is worth noting that the asymmetry of non-CpG methylation, i.e., mCpC > mCpA/mCpT > mCpG, which was observed in the control region in this study, was reported to be acutely regulated in response to environmental stimuli/stressors [31]. Furthermore, a strong positive correlation between mtDNAcn and the CpH methylation of the control region was observed in subjects aged ≥ 45.

The vast majority of mitochondrial proteins are encoded in the nuclear genome, including all the regulatory factors for mtDNA replication and transcription through binding at the control region [32]. Among them, POLG encodes the catalytic subunit of mtDNA polymerase, with an established role for mtDNA replication and repair. A previous study suggested a reverse association between methylation of a CpG island within exon 2 of the POLG gene with decreased mtDNAcn in highly proliferative versus terminally differentiated cells and several cancer cell lines [33, 34]. One explanation was underscored by a recent study that evaluated tissue-specific changes in the methylation status of CpG islands within POLG, POLG2, TFAM, FIS1, and OPA1 genes in differently aged rats on standard versus low-calorie diet [35]. Another possibility could be that the site specificity of methylation on POLG determined the recruitment of either enhancer or suppressor for POLG expression, as we found the methylation levels at the 4th CpG site of the POLG promoter region is positively correlated with mtDNAcn.

While the nuclear genome has well-characterized CpG methylation, mtDNA methylation is understudied and more controversial [36–38], with conflicting data about its existence and biological significance. Despite some studies suggesting the low CpG methylation compared with nuclear DNA, others demonstrated mtDNA methylation in both CpG and non-CpG contexts using various techniques [39–42]. The identification of DNA methyltransferase 1 (DNMT1) and other DNA methylation enzymes, such as DNMT3a, DNMT3b, TET1, and TET2, in mitochondria in various systems also supports the existence of mtDNA methylation [31, 43–45]. In addition, emerging evidence from animal models and human tissues suggests that epigenetic modifications in mtDNA could contribute to the aging process. Hypomethylations of the mtDNA non-coding region [46] and MT-CO1 [47] alongside with an increase in mtDNAcn and MT-CO1 expression were reported in senescent cells compared to proliferative endothelial cells. However, in cells treated with 5-aza-2′-deoxycytidine, MT-CO2 methylation was inhibited, but MT-CO2 expression was downregulated [48]. A study involving subjects aged 38 to 107 years claimed that the methylation levels of the MT-RNR1 gene were positively associated with age, particularly among women older than 85 years [49]. A subsequent study demonstrated that the methylation of two CpG sites (M1215 and M1313) located within the MT-RNR1 gene showed a strong inverse correlation with the subject’ age [50]. Perhaps the tissue-specific and site-specific methylation plays a role in these diverse regulations. Lastly, it is also interesting to note that compared with the nuclear DNA-encoded POLG promoter region, mtDNA D-loop region methylation exhibited a sharp increase in old-aged subjects (graphic abstract), indicating an aging-dependent activation of the mechanism of mtDNA methylation.

Limitations and conclusion

While our studies provided evidence to support the correlation between the epigenetic changes in both the nuclear and mitochondrial DNA control regions for mtDNA replication and mtDNAcn in the context of aging, we want to acknowledge several limitations of the present study. First, it is a study with a relatively small sample size across the age groups, although our case-control of 30 samples per group offers sufficient power to detect differences in control methylation levels of approximately 1.2% between groups, higher than a previous similar study with a difference of about 1% [51]. Second, our investigation on mtDNA methylation represents only a small portion of the mitochondrial genome. Third, our samples are from volunteers recruited randomly without comprehensive background information. Further studies with expanding samples and more sophisticated technology to determine the role of epigenetic control of mitochondrial biogenesis are warranted. Mechanistic study targeting methylation in cultured cell models with CRISPR technology would also be an ideal next step in the future directions.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

HC designed, performed, and interpreted the data. HS analyzed the data and wrote the manuscript. YY performed the experiments. YI, JW, and MA interpreted the data and wrote the manuscript; ZW and HZ analyzed the data, and YB conceived the project, directed the research, and wrote the manuscript. All authors read and approved the final manuscript, and the authors declare that they have no competing interests.

Funding

The authors are thankful for the support provided by the College of Laboratory Medicine and Life Sciences, Wenzhou Medical University. Hongying Shu and Yidong Bai are supported by the Baptist Health Foundation of San Antonio and the San Antonio Medical Foundation.

Data availability

All detailed information about the primers used in this study is listed in Supplementary Table 1. Supplementary Tables and Figures are available Online.

Declarations

Ethics approval and consent to participate

Informed consent from all donors in this study was obtained under protocols approved by the Wenzhou Medical University Ethics Committee.

Conflict of interest

The authors declare no competing interests.