Mitochondrion 73 (2023) 51–61 Contents lists available at ScienceDirect Mitochondrion journal homepage: www.elsevier.com/locate/mito Complete mitochondrial DNA profile in stroke: A geographical matched case-control study in Spanish population Ana Onieva a, *, Joan Martin b, Daniel R. Cuesta-Aguirre a, Violeta Planells a, Marta Coronado-Zamora c, Katrin Beyer d, Tomás Vega e, José Eugenio Lozano e, Cristina Santos a, Maria Pilar Aluja a, * a Unitat d’Antropologia Biològica, Departament BAVE, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain b Department of Genetics and Microbiology, Faculty of Biosciences, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain c Institut de Biotecnologia i Biomedicina; Department of Genetics and Microbiology, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain d Department of Pathology, Germans Trias i Pujol Research Institute, Badalona 08916 Barcelona, Spain e Dirección General de Salud Pública. Consejería de Sanidad. Junta de Castilla y León, 47007 Valladolid, Spain A R T I C L E I N F O A B S T R A C T Keywords: Introduction: Stroke, the second leading cause of death worldwide, is a complex disease influenced by many risk Stroke factors among which we can find reactive oxygen species (ROS). Since mitochondria are the main producers of Mitochondrial DNA cellular ROS, nowadays studies are trying to elucidate the role of these organelles and its DNA (mtDNA) variation Mitochondrial haplogroups in stroke risk. The aim of the present study was to perform a comprehensive evaluation of the association be­ Mitochondrial DNA copy number tween mtDNA mutations and mtDNA content and stroke risk. Material and methods: Homoplasmic and heteroplasmic mutations of the mtDNA were analysed in a case-controls study using 110 S cases and their corresponding control individuals. Mitochondrial DNA copy number (mtDNA- CN) was analysed in 73 of those case-control pairs. Results: Our results suggest that haplogroup V, specifically variants m.72C > T, m.4580G > A, m.15904C > T and m.16298 T > C have a protective role in relation to stroke risk. On the contrary, variants m.73A > G, m.11719G > A and m.14766C > T appear to be genetic risk factors for stroke. In this study, we found no statistically significant association between stroke risk and mitochondrial DNA copy number. Conclusions: These results demonstrate the possible role of mtDNA genetics on the pathogenesis of stroke, probably through alterations in mitochondrial ROS production. 1. Introduction developing countries (Donkor, 2018; Hankey, 2017). There are stroke risk factors such as hypertension, diabetes mellitus, Nowadays, stroke is the second leading cause of death worldwide, high blood cholesterol, sedentary lifestyle, atrial fibrillation, smoking, accounting for approximately 6.5 million deaths annually, and the first and alcohol consumption that can be modified through a change in cause of acquired disability in adults in some countries (W. Chen et al., lifestyle or medication. Alternatively, there are also non-modifiable risk 2020; Donkor, 2018; Feigin et al., 2022). There are 12.2 million new factors that include age, sex, and genetics (Donkor, 2018). Related to stroke cases per year around the world, which equals one stroke genetics, family studies have shown that a family history of stroke in­ happening every 3 s. Furthermore, 101 million people are living with creases the risk of ischemic stroke by approximately 75% (Floßmann stroke aftermath and over 143 million years of healthy life are lost each et al., 2004), and genome-wide association studies have estimated her­ year due to stroke-related death and disability (Feigin et al., 2022). itability for stroke near 40% (Malik & Dichgans, 2018). From these risk Despite stable incidence rates and declining mortality rates over the past factors, age is the strongest determinant of stroke, followed by hyper­ two decades, the public health burden of stroke is set to rise over future tension (Donkor, 2018). decades due to demographic transitions of populations, particularly in Despite the previously referred risk factors, in the last decade, * Corresponding authors at: Unitat Antropologia Biològica, Dep. Biologia Animal, Biologia Vegetal i Ecologia, Facultat Biociències Edifici C, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain. E-mail addresses: ana.onieva@uab.cat (A. Onieva), mariapilar.aluja@uab.cat (M.P. Aluja). https://doi.org/10.1016/j.mito.2023.10.001 Received 17 January 2023; Received in revised form 28 August 2023; Accepted 1 October 2023 Available online 2 October 2023 1567-7249/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). A. Onieva et al. Mitochondrion 73 (2023) 51–61 different studies have been highlighting the importance of reactive ox­ associated with mitochondrial enzyme activity and adenosine triphos­ ygen species (ROS) in stroke initiation, as ROS are involved in the phate production and can therefore serve as a biomarker of mitochon­ pathogenesis of atherosclerosis (Chang et al., 2010; Peluso et al., 2012). drial function (Ashar et al., 2017; Yue et al., 2018). The most recent studies are focusing on the role of mitochondria in The multiple-copy nature of the mitochondrial genome implies that stroke risk (Yan et al., 2022) since it is widely known that they are when the mtDNA undergoes a new mutational change, the new variant responsible for approximately 90 % of cellular ROS production (Balaban coexists with other mtDNA molecules with the original sequence within et al., 2005). a cell or mitochondria of an individual. The coexistence of different Mitochondria are cytoplasmatic organelles that supply energy for all genomes gives rise to mitochondrial heteroplasmy. cellular metabolic activities via ATP synthesis. Structurally, each mito­ Mitochondrial variants accumulated through time differ between chondrion is bounded by a porous outer membrane and an inner population groups that diverged thousands of years ago. In consequence, membrane that contains multi-enzyme complexes responsible for ATP the related mitochondrial haplotypes (Ht) are grouped into haplogroups synthesis through oxidative phosphorylation (OXPHOS) (Bhagavan & (Hg) which allow the matrilineal inheritance of human beings to be Ha, 2011; Hatefi, 1985). OXPHOS requires the coordinated action of traced from their origins in Africa to the geographical distribution of four respiratory complexes (complexes I-IV), that make up the electron current ethnic groups (Tranah et al., 2011). Mishmar et al. (2003) transport chain (ETC), and an ATP synthase (complex V). However, this proposed that the geographic distribution of human mtDNA hap­ process is not perfectly coupled, since some of the energy present in the logroups had resulted due to selection mainly driven by adaptation to proton gradient (Δp) dissipates as heat (thermogenic role), through climate. According to this hypothesis, in colder climates, natural selec­ pathways independent of ATP synthase (proton leakage) (Cadenas, tion would have favoured mtDNA variants that reduced OXPHOS 2018). coupling efficiency to increase heat production by proton leakage Oxidative Phosphorylation is also the major endogenous source of (Chinnery & Gomez-Duran, 2018; Mishmar et al., 2003). Following this ROS. These reactive oxygen species are generated when electrons leak hypothesis and the idea that OXPHOS coupling also alters ROS pro­ directly to oxygen (O2) to form superoxide (O-), which can be converted duction and thus stroke susceptibility, some studies have been focused to hydrogen peroxide (H2O2). This electron leakage comes mainly from on the association between haplogroups and stroke susceptibility. Some ETC complexes I (NADH dehydrogenase) and III (cytochrome c reduc­ studies based on this idea have obtained significant associations of some tase)(Addabbo et al., 2009; Andalib et al., 2017; Chen & Zweier, 2014; haplogroups, such as K, H1 or U, with stroke risk in different European Krzywanski et al., 2011). The accumulation of high levels of ROS causes populations (review in Andalib et al., 2017) while others did not an intracellular REDOX imbalance, creating oxidative stress that leads to (Umbria et al., 2019). Nonetheless this hypothesis has been refuted by mitochondrial dysfunctions and contributes to the development and/or several groups (Moilanen, J et al., 2003; Sun, C et al., 2007; Jeedigunta, progression of cardiovascular diseases such as atherosclerosis, which is S.P. et al., 2021). involved in stroke initiation. These processes, which may eventually In recent years, the focus of mtDNA studies has expanded, and a new modulate stroke susceptibility, are variable characteristics among the hypothesis has been gaining interest. This complementary approach, population that depend on different individual traits (Andalib et al., known as the mutational load hypothesis, refers to the synergistic effect 2017; Krzywanski et al., 2011). of several variants, not looking for associations between cardiovascular Mitochondrial ROS generation is mainly modulated by OXPHOS disease (CVD) risk and a specific variant but to the accumulative effect of coupling efficiency, which is defined as the percentage of mitochondrial several variants (Piotrowska-Nowak et al., 2019a; Venter et al., 2018). respiration linked to ATP synthesis (Gnaiger et al., 2015). When Based on the evidence that ROS are mainly generated when electrons OXPHOS is highly coupled, Δp is higher since only a low percentage of it leak from ETC complexes, we suspect that ROS formation could also be dissipates as heat. In those cases, respiration slows down increasing modified by individual variations in ETC complexes. Therefore, varia­ oxygen availability and the steady-state concentration of reduced-state tion in genes that encode these complexes may play an important role in ETC complexes (higher electron availability), which favours superox­ ROS regulation, and thus in stroke susceptibility. Furthermore, based on ide formation by electron leakage. However, mild uncoupling of the evidence that mitochondrial dysfunction conditions the cells’ energy OXPHOS dissipates Δp attenuating ROS production (Brand, 2000; homeostasis, variation in mtDNA-CN and genetic variation in the NCR, Cadenas, 2018; Krzywanski et al., 2011; Munro & Treberg, 2017). which could alter both replication and transcription of the mtDNA Apart from OXPHOS coupling efficiency, Chen and Zweier (2014) genome, may also play a role in CVD risk. proposed that ROS generation could also be modulated by individual Given the aforementioned premises, the main goal of this study is to variations in OXPHOS rate and that a decrease in the rate of oxidative perform a comprehensive evaluation of the association between stroke phosphorylation could increase ROS production. Some OXPHOS com­ risk and mitochondrial genetics in a Spanish population from Castilla y plexes are partially encoded by the mitochondrial DNA (mtDNA). Thus, León. The more specific objectives are: i) to evaluate the relation be­ mtDNA variation could be involved in ROS regulation and, therefore, in tween mtDNA-CN and stroke risk, ii) to determine the mitochondrial stroke susceptibility. haplotype and haplogroup of individuals with stroke history and The mitochondrial genome is a circular double-stranded and investigate their role concerning stroke risk and iii), to analyse possible maternally inherited DNA molecule of 16.6 kb present in numerous differences in the distribution of mitochondrial mutations between copies within the mitochondrial matrix. Concretely, mtDNA coding re­ stroke cases and controls, and examine the possible accumulative effect gion (577–16,023 bp) encodes 13 essential polypeptides of the OXPHOS of common variants in stroke risk. system (complexes I, III, IV and V) and the necessary RNA machinery for their translation within the mitochondria (Giles et al., 1980; Robin & 2. Material and methods Wong, 1988). The mtDNA also has a non-coding region (NCR) spanning approximately 1.1 kb (16,024–16,569; 1–576 bp) that comprises three 2.1. Sample collection and DNA extraction hypervariable regions (namely HVR1, HVR2, and HVR3), the point of origin of heavy strand replication (OH) and the transcription promoters A case-control study was conducted using 110 S patients and 110 of both strand (HSP and LSP respectively) (Nicholls & Minczuk, 2014). corresponding controls taken from the Cardiovascular Disease Risk The number of molecules of mitochondrial DNA, also known as Study of Castilla y León (Vega Alonso et al., 2007). Cases and controls mtDNA copy number (mtDNA-CN), is variable in cells and tissues. Since were matched according to nonmodifiable risk factors, namely, age, sex mtDNA expression is a prerequisite for OXPHOS functionality, mtDNA- and geographic origin. CN must be adjusted according to metabolic needs (Filograna et al., The study of Vega Alonso et al. (2007) was a cross-sectional study 2021). The mtDNA-CN, while not a direct measure of mtDNA damage, is initiated in 2004, and followed up in 2009 and 2014, in which 4013 52 A. Onieva et al. Mitochondrion 73 (2023) 51–61 subjects (older than 15 years of age) were randomly selected from follows: HN: F 5′- AGT ACC TAA CAA ACC CAC AGG − 3′, R 5′ - Castilla y León population and underwent health examinations in rela­ TTGGATCAATTGAGTATAGTAGT − 3′; PSEN2: F 5′ - CTG TGC ACG CCT tion to cardiovascular risk factors. From these subjects, blood samples CTT CAG T − 3′, R 5′ - CTG TGA GCC TTG GTC TCA A − 3′. Primers were were also obtained, after informed consent, and following the Helsinki designed to assure HN1-specidicity, carrying out primer blast (NCBI; protocol (World Medical Association, 2013). Biological samples https://www.ncbi.nlm.nih.gov/tools/primer-blast/) and also alignment included in this study were centralized, processed and provided by the using clustal Omega - Multiple Sequence Alignement. EMBL-EBI (htt Spanish National DNA Bank Carlos III of the University of Salamanca ps://www.ebi.ac.uk/Tools/msa/clustalo/) to check specificity against (biobank ID B.0000716 and ISCII-EU grant number PT20/00085), in­ the MT-RNR2 like pseudogenes 1–13 and thus ruling out any possibility tegrated in the “ISCIII Platform Biobanks and Biomodels”. The DNA of of cross amplification with nuclear genes. qPCR conditions consisted of these samples was extracted using an automated process on an Autopure an initial denaturing step of 15 min at 95 ◦ C followed by 35–40 cycles of LS (Gentra System, Minneapolis, MN) and stored at − 20 ◦ C. 15 sec of denaturation at 95 ◦ C, 25 sec of annealing at 57–60 ◦ C and 25 For each subject, information about age (categories ≤ 44, 45–49, sec of extension at 72 ◦ C. The Cycle Threshold (Ct) difference between 50–54, 55–59, 60–64 and ≥ 65 years), sex, geographic origin (North, replicates was<0.5. Any sample crossing this threshold was repeated to Central and South regions of Castile and Leon), history of hypertension minimize the error in the assay and discarded if the difference persisted. (≥140/90 mmHg), history of diabetes, history of hypercholesterolemia The mtDNA-CN results were calculated using the comparative-cycle (greater than200 mg/dl), cigarette consumption (smokers, former threshold (ΔΔCt) (Livak & Schmittgen, 2001) and Fold Change (Fold smokers and non-smokers), presence of overweight or obesity (body change = 2− ΔΔCt ). mass index ≥ 25 kg/m2), presence of abdominal perimeter in risk range (risk: ≥ 80 cm for women and ≥ 94 cm for men) and presence of high 2.4. Data analysis levels of triglycerides (≥170 mg/dl), was available from the databases of the original study (Vega Alonso et al., 2007). 2.4.1. Analysis of mtDNA sequencing data FASTQ files from each sample were processed with FASTQC program 2.2. Mitochondrial DNA amplification and sequencing to perform quality control analyses (Babraham Institute Enterprise, BIE). After FASTQC quality control, data for each sample was imported into For each sample, the whole mitochondrial genome was amplified in mtDNA-Server 1.0.7 (Weissensteiner et al., 2016), which is a free two overlapping fragments of approximately 9 kb and 11 Kb using two mtDNA NGS analysis pipeline, to identify the sample’s haplogroup, distinct sets of primers (SET1 and SET2) previously described (Mikkel­ homoplasmic and heteroplasmic variants in the sequence and possible sen et al., 2009). This primer design allows to rule out the possibility of contamination. Finally, mtDNA-Server output data was processed with NUMT co-amplification due to their location and the target amplicon an ad-hoc R application based on Shiny to facilitate downstream ana­ size, as 77,8% of NUMTs are<500bps (Wei, W. et al., 2022). PCR mix for lyses. The script is available on the GitHub repository https://github.co each reaction consisted of 1 μL (10 pmol/μL) of each primer (Forward m/marta-coronado/mtDNA-analysis. and Reverse), 1 μL (10 μM) of dNTPs, 5 μL of 10 X reaction buffer, 2 μL (50 mM) of MgSO4, 0.2 μL (2 U/μL) of Platinum® Taq High Fidelity 2.4.2. Mutational load calculations and adjustments using MutPred DNA Polymerase and 1 μL (approximately 50 ng) of DNA in a final pathogenicity scores volume of 50 μL. The PCR reactions were performed in a Bio-Rad MutPred pathogenicity scores were used to derive “mutational S1000™ Thermal Cycler (Bio-Rad, Hercules, CA) and the amplifica­ loads”, which are useful to test the possibility of a cumulative effect of tion program consisted of an initial denaturation step of 2 min at 94 ◦ C, mildly deleterious (rare) variants in cases and control groups (Elson followed by 35 cycles of PCR (30 s at 94 ◦ C, 30 s at annealing temper­ et al., 2006). The MutPred algorithm assigns a pathogenicity score be­ ature – 60 ◦ C for SET1 and 59 ◦ C for SET2 – and 10 s at 68 ◦ C) and a final tween 0 and 1, with 0 being a benign substitution. Pathogenic scores extension step of 7 min at 72 ◦ C. The PCR amplification results were above 0.5 can be considered “actionable hypotheses”, while scores visualized through 1% agarose gel electrophoresis to further validate the above 0.75 can be considered “confident hypotheses”. specificity of the PCR. For mutational load calculations, the MutPred scores above 0.5 PCR products were purified using DNA Clean & Concentrator™-5 Kit generated for each of the non-synonymous homoplasmic variants on an (Zymo Research, Irvine, CA) and quantified using Qubit® dsDNA HS individual’s mtDNA were summed. Variants with a pathogenicity score Assay Kit (ThermoFisher Scientific, Waltham, MA), according to the below 0.5, which tend to be the most common population variants, were manufacturer’s specifications. Once quantified, for each sample, SET1 excluded to highlight the impact of rare variants while reducing the and SET2 PCR products were diluted to 0.4 ng/μL and equally mixed to effect of population stratification (Venter et al., 2017). Then, the avoid over- representation of any of the fragments. Finally, to prepare mutational load was adjusted by dividing it by the number of non- the Next Generation Sequencing (NGS) library, 2.5 μL of each individual synonymous variants presented in each individual. In this way, the ef­ equimolar mix was placed in a single well of a 96-well plate. fect of the position of an individual in the phylogeny was reduced Dual indexed libraries were generated using Nextera® XT DNA (Venter et al., 2017). Sample Preparation Kit, according to the manufacturer’s protocol (Illumina, San Diego, CA). The library was sequenced with paired-end, 2.4.3. Statistical comparison between cases and controls dual-indexing 2x250 bp cycles per read in the Servei de Genòmica McNemar’s test (dichotomous variables) a marginal mean test (var­ (Universitat Autònoma de Barcelona, http://sct.uab.cat/genomic iables with more than two categories) and a conditional logistic a-bioinformatica/es). Two FASTQ files were generated per sample. regression were used to compare the frequency of biochemical and clinical variables (cigarette consumption, hypertension, diabetes, hy­ 2.3. Mitochondrial DNA copy number (mtDNA-CN) quantification percholesterolemia, overweight or obesity, abdominal perimeter and triglycerides) between cases and controls to determine potential con­ To a subset of 146 samples, 73 cases and 73 controls, the mtDNA-CN founding variables (stroke risk factors). was also evaluated. The relative mtDNA-CN, defined as the copy number A t-test for paired samples was used to determine if there were dif­ ratio of a mitochondrial gene (Humanin, HN1) to a single-copy nuclear ferences in mtDNA Copy number between stroke cases and controls. gene (Presenilin 2, PSEN2), was assayed in duplicate by a quantitative Haplogroups frequencies were compared using McNemar’s test and real-time PCR (qPCR) method using the Rotor-gene 6000 System (Qia­ Odds Ratios (and 95% confidence interval, CI) were calculated. gen) and the Rotor-Gene SYBR Green PCR kit (Qiagen). Specific primers To compare the mtDNA profile between cases and controls, homo­ used to amplify the mt-HN gene and the nuclear PSEN2 gene were as plasmic and heteroplasmic mtDNA variants were compiled into different 53 A. Onieva et al. Mitochondrion 73 (2023) 51–61 matrixes. Concerning heteroplasmies, only reliable point heteroplasmies Table 1 with an individual frequency of the minor variant of at least 3% and Sociodemographic, biochemical and clinical characteristics of stroke cases and 1000x depth were compiled into the matrix (González et al., 2020). controls. Region 309 – 315 was excluded from the analyses due to difficulties in Cases Controls p- CLR p- the alignment of repetitive regions. n (%) n (%) value value To assess a possible imbalance in the distribution of homoplasmic Cigarette consumption mutations throughout the mitochondrial genome, the number of Non-smoker 60 62 0.431b 0.786 homoplasmic variants was compared between cases and controls with a (54.5) (56.4) Wilcoxon test. Comparisons include the whole mtDNA molecule, Former smoker 36 38 dividing the molecule into coding (CR) and non-coding regions (NCR), (32.7) (34.5) and stratifying the coding region in groups of genes that configure the Smoker 14 10 (9.1) (12.7) ETC complexes and RNAs and the NCR in the three hypervariable re­ Hypertension (≥140/90 mmHg) 89 88 (80) 1a 0.738 gions. Identical analyses were made in the case of point heteroplasmic (80.9) variants. Adjusted mutational loads (AML) were compared between Diabetes 19 15 0.556a 0.884 cases and controls using the one-tailed Wilcoxon test to determine if (17.3) (13.6) a Hypercholesterolemia (>200 mg/ 58 44 (40) 0.082 0.109 differences in the number of homoplasmic variants were related to a dL) (52.7) pathogenic effect. This comparison was performed for both, the whole Overweight or obesity (≥25 Kg/m2) 76 81 0.568a 0.842 coding region and complex of the ETC. P-values were corrected for (69.1) (73.6) multiple testing (False Discovery Rate, FDR < 5%). High abdominal perimeter (>80 cm 79 84 0.542a 0.539 Finally, each mtDNA individual variant - homoplasmic or in heter­ or > 94 cm) (71.8) (76.4) a Triglycerides (≥170 mg/dL) 13 7 (6.4) 0.238 0.206 oplasmy - detected, in at least 5% of cases and/or controls, was indi­ (11.8) vidually tested with a McNemar’s test to determine if associations with a stroke were present. Odds Ratios (and 95% confidence interval, CI) were p-value of McNemar (used for dichotomous variables); b p-value of marginal homogeneity test (used for variables with more than two also calculated. categories). Statistical analyses were performed using IBM SPSS Statistics 23 (IBM Corp., 2015) and JAMOVI 2.3.21 (Jamovi, 2022). Differences were considered significant at P-values equal or below 0.05. showed no statistically significant differences for any of the risk factors. Given that these biochemical and clinical characteristics could be 2.4.4. Possible significance of variants associated with stroke influencing each other a conditional logistic regression was performed including those variables that weren’t matched between cases and controls. No risk factor presented a statistically significant association 2.4.4.1. Hits in the phylogeny, population database and nucleotide con­ with stroke risk (Table 1). servation index (NCI). For variants that appear to be associate with Since these variables do not present differences between cases and stroke, the number of hits or the number of occurrences in the phylogeny controls, they will not be included in further analyses. was compiled from the updated mtDNA phylogeny - mtDNA tree Build 17 (van Oven & Kayser, 2009) and from Soares et al., (2009). To calculate the frequency of each variant in the population an online 3.2. Mitochondrial DNA copy number population database with 54,594 human mitochondrial DNA sequences (with size greater than 15.4 kb) was used (Lott et al., 2013). The mitochondrial DNA copy number was similar in stroke cases The nucleotide conservation index (NCI) was estimated for variants (mean ΔCt = -5.86) in relation to the controls (mean ΔCt = -5.78) (Fold associated with stroke across reference sequences of different primate change = 1.35) and the difference was not statistically significant (t-test species. A total of 48 sequences were used (for list of species and p-value = 0.430). accession numbers see Supplementary Table S1). Sequences were aligned using Clustal W (Thompson et al., 1994). 3.3. Haplogroup frequencies A summary of haplogroup frequencies for stroke cases and controls is 2.4.4.2. RNA structure prediction. Secondary structures were performed to understand the structural impact of the different variants. The Sec­ ondary structure was generated using the RNAfold web server Table 2 MtDNA macrohaplogroup and haplogroup frequencies in stroke cases and (http://rnna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi, last controls. accession November 2022) with default parameters established (Gruber et al., 2008). The sequence introduced was either the complete gene or Cases Controls McNemar p-value Odd Ratio (CI 95%) the hypervariable region containing the variant. The minimum free n (%) n (%) energy, base-pair probabilities and positional entropy of predicted R0 55 (50) 71 (64.55) 0.027 0.5493 (0.312 – 0.943) structures were used to estimate the implications in the RNA molecule. H 48 (43.64) 57 (51.82) 0.253 0.720 (0.423 – 1.224) HV 4 (3.64) 3 (2.73) 1 1.346 (0.294 – 6.159) 3. Results V 2 (1.82) 10 (9.09) 0.039 0.185 (0.040 – 0.866) R0 1 (0.91) 1 (0.91) 1 1 (0.0618 – 16.192) JT 20 (18.18) 14 (12.73) 0.327 1.524 (0.726 – 3.197) 3.1. Classical risk factors J 12 (10.91) 7 (6.36) 0.332 1.802 (0.681 – 4.764) T 8 (7.27) 7 (6.36) 1 1.154 (0.404 – 3.300) The percentage of individuals (cases – controles) included in risk U 28 (25.45) 18 (16.36) 0.144 1.745 (0.900 – 3.386) categories for different non-genetic risk factors is shown in Table 1. U 19 (17.27) 12 (10.91) 0.248 1.705 (0.784 – 3.709) K 9 (8.18) 6 (5.45) 0.581 1.545 (0.531 – 4.497) Sociodemographic variables (age, sex and geographic origin) used to No-R 7 (6.36) 7 (6.36) 1 1 (0.339 – 2.953) match stroke cases with their corresponding controls can be found in I 2 (1.82) 1 (0.91) 1 2.019 (0.180 – 22.591) Supplementary Table S2. The most notorious difference is found in hy­ L 0 (0) 2 (1.82) 0.5 0.196 (0.009 – 4.138) percholesterolemia, being more frequent in stroke cases (McNemar p- M 1 (0.91) 0 (0) 1 3.027 (0.122 – 75.133) W 3 (2.73) 2 (1.82) 1 1.514 (0.248 – 9.243) value 0.082). However, the comparison between cases and controls X 1 (0.91) 2 (1.82) 1 0.495 (0.044 – 5.545) 54 A. Onieva et al. Mitochondrion 73 (2023) 51–61 shown in Table 2. Macrohaplogroup R0 (including haplogroups H, HV, V a decrease in homoplasmies in stroke cases. Complex III (cytochrome b) and R0) was significantly overrepresented in stroke controls (64.55% of also presented a significant increase in homoplasmies in stroke cases in controls vs 50% of cases; McNemar’s test: P = 0.027). When macro­ relation to controls (C = 287 vs CL = 230, P = 0.002). Homoplasmic haplogroup R0 was closely examined, haplogroup V appeared over­ variants in Complex IV (Cytochrome c oxidase) were similar between represented in stroke controls (9.09% of controls vs 1.82% of cases, cases and controls (C = 176 vs CL = 171, P = 0.771), a characteristic McNemar’s: P = 0.039). In this sense, our results showed that in­ observed in all 3 genes (CO1, CO2 and CO3). Similar results were ob­ dividuals from macrohaplogroup R0 and hg V were 1.82 and 5.41 tained for Complex V (ATP synthase, C = 155 vs CL = 160, P = 0.554) (inversed odds ratios) times less likely to have a stroke, respectively. and ATP6, ATP8 and the overlapping region between these genes. This association is not statistically significant after multiple testing Regarding RNA genes, our results showed an increase in the number correction (FDR < 5%). of homoplasmies in cases in relation to controls in both rRNAs (C = 476 vs CL = 429, P = 0.022) and tRNAs (C = 81 vs CL = 66, P = 0.21). Differences in rRNA genes were due to differences in RNR1 and RNR2 as 3.4. Mutation profile and mutation load no mutation was found in RNR3. The number of heteroplasmies did not show significant differences Individual mtDNA profile is reported in Supplementary Table 3 and between stroke cases and controls in any of the analyses except for the the number of homoplasmic variants and heteroplasmic positions in gene CO2, in which heteroplasmic positions were significantly more cases and controls stratified by mtDNA regions is shown in Table 3. A prevalent in cases than in controls (C = 4 vs CL = 0; Wilcoxon test P = comparison between the number homoplasmies and heteroplasmies 0.046). between cases and controls was performed to assess a possible imbal­ In addition, mutational loads of cases and controls were compared to ance in their distribution throughout the mtDNA.The analysis of the determine if differences in the number of homoplasmic variants were whole mtDNA molecule (Table 3) showed that the number of homo­ related to a pathogenic effect. Table 4 reports adjusted mutation loads plasmic mutations was significantly higher in cases (cases (C) = 2525 vs by mtDNA complexes and genes. A higher adjusted mutational load was controls (CL) = 2176; Wilcoxon test: P = 0.02). Similar results were observed in stroke cases (x‾ = 0.364) in comparison to controls (x‾ = obtained when the non-coding region was studied separately (Wilcoxon 0.296), although this difference was not statistically significant (Wil­ test: P = 0.006). Within the NCR, a significant increase in homoplasmic coxon test: P = 0.109). Analysing by complex and genes, a higher mutations in the HVR1 (C = 256 vs CL = 179; Wilcoxon test P = 0.002) adjusted mutational load in cases was seen in Complex IV and ATP6 with was found. This tendency is also observed in HVR2 and HVR3, although values at the limit of statistical significance (Wilcoxon test: P = 0.05). differences were not significant in these cases. Variants inside the coding region were analysed complex by complex and gene-by-gene. Complex I (NADH dehydrogenase) showed a signif­ 3.5. Individual variation association icant increase in homoplasmic variants in stroke cases in relation to controls (C = 658 vs CL = 556, P = 0.041). This can also be observed for Each mtDNA variant – in homoplasmy or in heteroplasmy - was genes ND1, ND3, ND4L, ND4 and ND5, with differences in genes ND4 individually analyzed to determine if they were associated with stroke and ND5 being statistically significant. On contrary, ND2 and ND6 show risk. The comparison of each between cases and controls allowed to Table 3 Number of homoplasmic variants and heteroplasmic positions in stroke cases and controls stratified by mtDNA regions. Number of Homoplasmic Variants Number of Heteroplasmic Positions Region Cases Controls Wilcoxon p-value Cases Controls Wilcoxon p-value Σ (%) Σ (%) Σ (%) Σ (%) NC Region 685 (27.13) 557 (25.60) 0.006b 45 (34.09) 51 (32.69) 0.665 HVR1 256 (10.14) 179 (8.23) 0.002b 17 (12.88) 15 (9.62) 0.713 HVR2 300 (11.88) 258 (11.86) 0.076 19 (14.39) 22 (14.10) 0.729 HVR3 44 (1.74) 36 (1.65) 0.393 9 (6.82) 14 (8.97) 0.445 Outsidea 85 (3.37) 84 (3.86) 0.912 0 (0) 0 (0) 1 Coding Region 1840 (72.87) 1619 (74.40) 0.056 87 (65.91) 105 (67.31) 0.296 Complex I 658 (26.06) 556 (25.55) 0.041 35 (26.52) 41 (26.28) 0.792 ND1 61 (2.42) 55 (2.53) 0.469 6 (4.55) 4 (2.56) 0.564 ND2 166 (6.57) 173 (7.95) 0.665 3 (2.27) 4 (2.56) 0.705 ND3 33 (1.31) 23 (1.06) 0.210 0 (0) 2 (1.28) 0.157 ND4L 26 (1.03) 16 (0.74) 0.085 1 (0.76) 1 (0.64) 1 ND4 172 (6.81) 125 (5.74) 0.022 10 (7.58) 15 (9.62) 0.614 ND5 167 (6.61) 120 (5.51) 0.017 12 (9.09) 11 (7.05) 0.835 ND6 33 (1.31) 44 (2.02) 0.209 3 (2.27) 4 (2.56) 0.705 Complex III (CYB) 287 (11.37) 230 (10.57) 0.002b 8 (6.06) 6 (3.85) 0.593 Complex IV 176 (6.97) 171 (7.86) 0.771 15 (11.36) 14 (8.97) 0.847 CO1 101 (4) 109 (5.01) 0.480 6 (4.55) 5 (3.21) 0.705 CO2 26 (1.03) 25 (1.15) 0.894 4 (3.03) 0 (0) 0.046 CO3 49 (1.94) 37 (1.70) 0.188 5 (3.79) 9 (5.77) 0.285 Complex V 155 (6.14) 160 (7.35) 0.554 7 (5.30) 12 (7.69) 0.251 ATP8 12 (0.48) 10 (0.46) 0.655 0 (0) 1 (0.64) 0.317 ATP8/6 3 (0.12) 1 (0.05) 0.317 0 (0) 0 (0) 1 ATP6 140 (5.54) 149 (6.85) 0.232 7 (5.30) 11 (7.05) 0.346 rRNA 476 (18.85) 429 (19.72) 0.022 19 (14.39) 27 (17.31) 0.179 RNR1 254 (10.06) 233 (10.71) 0.020 7 (5.30) 11 (7.05) 0.346 RNR2 222 (8.79) 196 (9.01) 0.130 12 (9.09) 16 (10.26) 0.375 RNR3 0 (0) 0 (0) 1 0 (0) 0 (0) 1 tRNA 81 (3.21) 66 (3.03) 0.210 3 (2.27) 5 (3.21) 0.480 Whole molecule 2525 (1 0 0) 2176 (1 0 0) 0.028b 132 (1 0 0) 156 (1 0 0) 0.239 a b Outside Hypervariable regions; Results that remain statistically significant after FDR < 5% correction. 55 A. Onieva et al. Mitochondrion 73 (2023) 51–61 Table 4 significance (Supplementary Table 4). Mean adjusted mutational load in homoplasmic variants in stroke cases and Stability analyses were performed to predict the impact of these controls. mutations. Several measures as the number of hits in the mtDNA phy­ Cases Controls Wilcoxon p- logeny, the frequency in the population database and the conservation value index (NCI) at the nucleotide level were calculated, and results are x‾ (CI) x‾ (CI) presented in Table 6. Positions 4580, 11719, 14,766 and 15,904 appear to be fairly stable Complex I 0.250 0.168 0.098 (-0.049–––0.550) (-0.104–––0.440) as they present 4 or fewer hits in the phylogeny while positions 72, 73 ND1 0.15 0.085 (-0.13–––0.3) 0.091 and 16,298 present a higher number and could be considered non-stable (-0.121–––0.421) positions. Variants in positions 73, 11,719 and 14,766 are highly prev­ ND2 0.117 0.06 0.086 alent in the database population (frequency greater than 75%), whereas (-0.113–––0.347) (-0.112–––0.232) ND3 0 (0–––0) 0.005 (-0.05–––0.06) 0.317 variants in positions 72, 4580, 14,766 and 15,904 have a low frequency ND4 0.017 0.01 0.225 (<2%). (-0.084–––0.118) (-0.064–––0.084) Using the previously proposed method (RNAFold web server) to ND4L 0.007 0 (0–––0) 0.317 predict the impact of these mutations on the stability of the secondary (-0.062–––0.076) structure of the mtDNA, we studied the 7 positions that had a statisti­ ND5 0.038 0.039 0.850 (-0.109–––0.185) (-0.113–––0.191) cally significant association with stroke risk. A summary of these results ND6 0.007 0.034 0.176 is presented in Table 7. Variants m.73A > G, m.4580G > A, m.11719G (-0.065–––0.079) (-0.111–––0.179) > A imply a conformational rearrangement of the fragments (HVR2, Complex III 0.120 (-0.123 – 0.129 0.927 ND2 and CYB respectively). While the minimum free energy (MFE) of (CYB) 0.364) (-0.118–––0.376) Complex IV 0.064 (-0.133 – 0.026 0.050 variants m.72 T > C and m.73A > G decreases compared to the wild-type 0.261) (-0.094–––0.146) molecule, the MFE of variants m.4580G > A and m.11719G > A in­ CO1 0.022 0.005 0.343 creases. Variants m.14766C > T, m.15904C > T and m.16298 T > C led (-0.09–––0.134) (-0.051–––0.061) to a folded structure with the same MFE as the wild-type structure. CO2 0.016 0.01 0.480 Although variant m.15904 T does no change the MFE of the folded (-0.084–––0.116) (-0.061–––0.081) CO3 0.031 0.011 0.128 structure, changes in the entropy of nucleotides surrownding this posi­ (-0.116–––0.178) (-0.068–––0.09) tion can be seen in Fig 1. Similar graphical representations for each Complex V 0.027 0.057 0.112 variant can be foun in Supplementary Table 5. (-0.097–––0.151) (-0.126–––0.239) ATP6 0.017 0.048 0.050 (-0.083–––0.117) (-0.123–––0.219) 4. Discussion ATP6/8 0.006 0 (0–––0) 0.317 (-0.053–––0.065) Stroke, the second leading cause of death worldwide, is a complex ATP8 0.01 0.005 0.414 disease influenced by different modifiable and nonmodifiable risk fac­ (-0.062–––0.082) (-0.044–––0.054) tors (Donkor, 2018). Among these, reactive oxygen species (ROS) are Global 0.364 0.296 0.109 (0.067–––0.661) (-0.004–––0.596) gaining importance and since mitochondria are their main producers, the latest stroke-associated research is trying to elucidate the role of these organelles in stroke risk (Balaban et al., 2005; Chang et al., 2010). detect three variants (m.73G, m.11719A and m.14766G)- all of them in In this line, some studies are focusing on the mitochondrial genome, homoplasmy - that were significantly overrepresented and four variants which encodes for OXPHOS complexes related to ROS production, to (m.72C, m.4580A, m.15904 T and m.16298C) – also homoplasmic - that find significant associations with stroke risk (Andalib et al., 2017). To were underrepresented in stroke cases (Table 5). date, most of these studies have been working with mitochondrial Variants m.72C, m.4580A, m.15904 T and m.16298C were more haplogroups; however, few of them have evaluated the importance of prevalent in controls and individuals with these variants were between the mtDNA variation of the whole molecule. In the present study, we 3.06 and 5.99 times (inversed odd ratios) less likely to have a stroke. aimed to determine the role of whole mtDNA variation in stroke. These variants may be related to the protective role of haplogroup V, as The characterization of biochemical and clinical variables showed no they are descriptive of the Hg (Supplementary Fig. 1). In contrast, var­ statistically significant differences between cases (C) and controls (CL) iants m.73G, m.11719A and m.14766G are extremely common across (Table 1). This result was unexpected as available literature considers linages and are not generally associated with a particular haplogroup. hypertension and hypercholesterolemia as the strongest determinants of Individuals with either one of these three risk variants were approxi­ stroke risk after age (Donkor, 2018; Umbria et al., 2017). On one hand, mately 2 times more likely to have a stroke. As variants m.72C, the fact that hypertension was not significantly more frequent in cases m.4580A, m.15904 T and m.16298C define Hg V, their association with could be due to the high number of control subjects that suffered from stroke risk was evaluated altogether using a conditional logistic hypertension (80%) in our sample. This could be explained because of regression (Cox regression). None of the variants reached statistical the advanced age of the subjects since more than 85% of them were Table 5 Possible stroke-associated variants. Variant Location Cases Controls McNemar p-value Odds Ratio (95% CI) Putative role n (%) n (%) m.72 T > C HVR2 2 (1.82) 11 (10.00) 0.012 0.167 (0.036 – 0.771) Protective m.73A > G HVR2 57 (51.82) 38 (34.55) 0.012 2.038 (1.185 – 3.506) Risk m.4580G > A ND2 2 (1.82) 11 (10.00) 0.022 0.167 (0.036 – 0.771) Protective m.11719G > A ND4 55 (50.00) 37 (33.64) 0.014 1.973 (1.145 – 3.340) Risk m.14766C > T CYB 56 (50.91) 38 (34.55) 0.014 1.965 (1.142 – 3.380) Risk m.15904C > T TT 2 (1.82) 11 (10.00) 0.022 0.167 (0.036 – 0.771) Protective m.16298 T > C HVR1 5 (4.55) 14 (12.73) 0.049 0.327 (0.113 – 0.941) Protective 56 A. Onieva et al. Mitochondrion 73 (2023) 51–61 Table 6 Complete results of stroke homoplasmic mutations analysed.rCRS = revised Cambridge reference sequence. Stability analysis Variant Cases: n (%)/ McNemar’s p- OR (95% CI) rCRS Distribution in population N. Hits phylogeny N. Hits Nucleotide Conservation Controls: n (%) value database (%) (PhyloTree.org) Soares et al. Index (%) m.72 T > C 2 (1.82) / 11 (10) 0.012 0.167 (0.036 T T: 98.12; C: 1.79; A: 0.005; 9 6 GAP: 87.5; T: 12.5 – 0.771) G: 0.08; Del: 0.002 m.73A > G 57 (51.82) / 38 0.012 2.038 (1.185 C G: 75.91; A: 24.08; C: 12 11 GAP: 87.5; A: 2.1; C: 6.3; (34.55) – 3.506) 0.005; Del: 0.002 G: 4.2 m.4580G > 2 (1.82) / 11 (10) 0.022 0.167 (0.036 G G: 98.31; A: 1.69 2 2 A:85.4; G:14.6 A – 0.771) m.11719G 55 (50) / 37 0.014 1.973 (1.145 G A: 77.23; G: 22.77 4 7 A: 60.4; T: 4.2; C: 25; >A (33.64) – 3.340) G:10.4 m.14766C 56 (50.91) / 38 0.014 1.965 (1.142 C T: 76.71; C: 23.29 4 10 A: 25; T: 8.3; C: 66.7 >T (34.55) – 3.380) m.15904C 2 (1.82) / 11 (10) 0.022 0.167 (0.036 C C: 98.40; T: 1.59; A: 0.002 1 1 GAP: 4.2; A: 2.1; T: 25; >T – 0.771) C: 66.7; G:2.1 m.16298 T 5 (4.55) / 14 0.049 0.327 (0.113 T T: 93.37; C: 6.61; A: 0.002; 18 13 A: 29.2; C: 8.3; T:37.5; >C (12.73) – 0.941) G: 0.007; Del: 0.012 G:2.1; GAP: 22.9 above 60 years old, and it is known that hypertension prevalence et al., 2010) and H1 (Rosa et al., 2008) have been reported as protective significantly increases with age, especially from 60 years old (Fryar factors for stroke in European population and haplogroup J with a lower et al., 2017). On the other hand, hypercholesterolemia differences presence of ischemic cardiomyopathy (Fernández-Caggiano et al., among groups (C: 52.7% and CL:40%) were higher than with hyper­ 2012). tension, triglycerides and the percentage of smokers, and the absence of Concerning the mtDNA profile, we observed a significant increase of significant differences could be due to sample size. However, finding no homoplasmic variants in stroke cases in comparison to controls differences in these possible risk factors in our sample strengthens the (Table 3), this increase was especially important in the genes corre­ reliability of the results produced in this study given that there was not a sponding to complexes I and III, and in rRNAs genes. Moreover, our possible bias induced by these variables. results showed a similar distribution of point heteroplasmies between The association between mtDNA-CN and cardiovascular disease risk, stroke cases and controls. In general, it is accepted that heteroplasmy is although commonly accepted, has yet to be thoroughly investigated. associated with many clinical presentations (Wallace & Chalkia, 2013). Lower mitochondrial DNA content has been associated with an Nevertheless, our results can be explained by the proven fact that het­ increased risk of cardiovascular disease (Wei et al., 2021; Yue et al., eroplasmy also occurs with appreciable frequency in the general popu­ 2018). Nonetheless, this association has been observed to vary in lation (Floros et al., 2018; Ramos et al., 2013) and may also been magnitude depending on the specific CVD studied, and even between acquiered tue to aging and be unrelated to cardiovascular health status. studies. In our study, we found no association between mitochondrial Regarding homoplasmic variant results, together with the fact that DNA copy number and stroke risk. MtDNA-CN varies depending on mitochondrial ROS is mainly generated when electrons leak from ETC many factors, many of which are matched between cases and controls in complexes I and III (Chen & Zweier, 2014), supports the idea that in­ our sample (age, sex and geographical origin) or present no differences dividual variations in these complexes can modify electron leakage (cigarette consumption, hypertension, hypercholesterolemia, etc). This predisposition, and thus ROS formation and stroke risk (Chang et al., fact increases the value of our results, as we have controlled many fac­ 2010). The increase in mutations in the control region observed in our tors that could contribute to mtDNA-CN changes, testing the association results could be influencing mitochondrial DNA regulation (replication between mtDNA-CN and stroke risk with more reliability. Even so, some and transcription). However, it is risky to obtain a definitive conclusion remarks should be considered: a) the sample size, as it could be as we do not know the character - neutral or pathogenic - of the different considered small; b) the fact that we perform a relative quantification variants (Sun & Yu, 2019) and we did not observe differences in mtDNA- instead of an absolute one; and c) the non-differentiation between CN that could indicate that the increase in mutations in this region is cellular and cell free mtDNA in our analyses. affecting the mitochondrial genome replication. Regarding haplogroup distribution between cases and controls, our The mutational load analysis presented in this study focuses on analyses showed that haplogroup V was the only haplogroup signifi­ homoplasmic population variants with a score above 0.5 that are pre­ cantly overrepresented in controls (Table 2). This result, which marks V dicted to be mildly deleterious to protein function, and thus tend to be as a potential genetic protective factor for stroke has not been reported rare variants that are less likely to be impacted by population stratifi­ before in any stroke-associated study. However, this association loses cation (Pereira et al., 2011; Venter et al., 2017). We observed a higher statistic significance after multiple testing correction. Hg V has been global mutational load (adjusted) in stroke cases, although the differ­ linked with Diabetes mellitus. While Piotrowska-Nowak et al. (2019b) ence is not statistically significant (Table 4). An association between a found a lower than expected prevalence of Hg V in Type 2 Diabetes higher mutational load and atherosclerosis was found by Piotrowska- mellitus when compared with the control group (in a Polish population), Nowak et al. (2019a), in contrast Venter et al. (2017) could not identify Soini et al. (2012) found a 3-fold higher frequency of mtDNA Hg V in a relationship between the mutational load and CVD phenotypes. patients with matrilineal diabetes (in Finnish population). Haplogroup V Interestingly, the adjusted mutational load of Complex IV is significantly is most frequent in the Sami population (33–68%), inhabiting the higher in stroke cases than controls (p = 0.05). The enzyme cytochrome Northern Shield - north of finland, Norway, Sweden and the Koala c oxidase (CO, Complex IV) is the last enzyme in the OXPHOS located in Peninsula of Russia- (Ingman & Gyllensten, 2007). As previously stated, the membrane and although CO is encoded by both the mitochondrial in colder climates, OXPHOS coupling efficiency decreases to increase and the nuclear genomes, its catalytic centres are encoded exclusively by heat production (Chinnery & Gomez-Duran, 2018) and, although there two mitochondrial genes, CO1 and CO2 (encoding CO subunits I and II, are no studies about hg V coupling efficiency, we could hipotetize that a respectively). Mutations in CO genes have been linked to Leigh Syn­ sligh uncoupling might exists reducing ROS production, and mtDNA drome (Bruno et al., 1999; Campos et al., 2001; Clark et al., 1999) and damage in these individuals. Our study did not replicate other associa­ MELAS syndrome (Wang et al., 2021). And the majority of poly­ tions previously found in other studies. Mitochondrial Hgs K (Chinnery morphism associated with stroke risk reported by Rosa et al. (2008) 57 A. Onieva et al. Mitochondrion 73 (2023) 51–61 were localized in complexes I and IV, whose deficiencies are the most − 54.30 − 60.50 − 44.40 frequently observed abnormalities of the OXPHOS system. Considering 70.84 0.00 Mut this information, our results are consistent with an increase of muta­ m.16298 T > C HVR1 tional load in patients with stroke, as an accumulation of mildly dele­ terious mutations in these genes that may alter CO function and in turn − 54.30 − 60.42 − 47.10 71.28 0.00 WT mitochondrial function and ROS generation, both known risk modifiers No of stroke. In the individual analysis of mitochondrial mutations, variants − 14,8 − 13,8 27,12 m.72C, m.4580A, m.15904 T and m.16298C were found to be over­ 5,24 − 14 Mut m.15904C > T represented in controls and have a protective role against stroke risk No (or slight) TT (Table 6). Variant m.72C > T is located in the control region, more − 15,09 precisely in the HVR2. Umbria et al. (2021) described it as a protective − 13,8 16,97 9,14 − 14 WT variant against myocardial infarction, which is aligned with our results, as seems to also be a protective factor against stroke. M.4580G > A is a synonymous SNP in the ND2 gene and has not been associated with any − 199,15 − 132,85 − 180,7 215,66 Mut condition. The ND2 gene shares a similar structure to sodium and po­ 0 tassium antiporters and may be involved in proton pumping. Variant m.14766C > T No (or slight) CYB m.15904C > T is located in the mitochondrial threonine tRNA gene, in Summary of the results obtained with the secondary structure prediction of the previously described variants. WT = wild type variant, Mut = mutated variant. the D-loop of the secondary structure, not modifying the codon- − 198,84 − 132,85 − 180,7 210,67 WT anticodon pairing. This variant has been described as benign in cases 0 of juvenile myopathy, encephalopathy, lactic acidosis and stroke (Wong et al., 2020) and has been classified as likely benign (MitoTIP score = − 180,54 − 198,93 1.7). Although the functional impact of the MFE changes in D-loop − 145,6 213,82 Mut 0 variants and even non-tRNA gene variants is unclear, we could hypos­ m.11719G > A tatize that the decrease in entropy observed in this tRNA gene variant ND4 results in an increased stability and therefore increased availability and − 181,64 − 200,52 219,45 better protein translation. M.16298C is located in HVR1 and has not − 145 WT 0 been described before as a variant that modifies risk for disease. None­ Yes theless, these four variants may not play a role in stroke risk indepen­ dently and could be reflecting the protective role of haplogroup V. − 114,88 − 128,37 − 100,7 109,85 In contrast, variants m.73G, m.11719A and m.14766G were found to Mut 0 be overrepresented in cases and increase the risk of stroke. Variant m.4580G > A ND2 m.73A > G, also located in HVR2, has been linked to an increased risk of obesity (Eaaswarkhanth et al., 2019), a known risk factor for CVD. In − 117,68 − 131,12 − 103,5 105,5 WT contrast, Umbria et al., (2020), found this polymorphism to play a 0 Yes protective role against myocardial infarction. It had been hypothesized that this variant could alter the replication and/or transcription of the − 71,2 − 75,7 − 67,5 32,79 mtDNA (Umbria et al., 2020). With the results of the mtDNA-CN analysis 0,07 Mut performed in this study, we can infer that these variants do not alter HVR2 Yes (slight) m.73A > G significantly the replication of the mitochondrial genome. M.11719G > − 72,97 A is a synonymous SNP in the ND4 gene. ND4 gene product is a subunit − 67,4 43,93 0,01 48,2 WT of the respiratory complex I which accepts electrons from NADH, transfers them to ubiquinone and uses the energy released to pump protons across the mitochondrial inner membrane. A mutation in ND4 − 74,31 − 69,4 − 62,9 37,99 0,03 Mut (m.11778G > A) has been related to over 50% and 90% of all LHON cases among Caucasians and Asians, respectively. In concordance with HVR2 m.72 T > C our results, Rosa et al.(2008) described m.11719G as a protective − 72,97 − 67,4 variant for ischemic stroke and proposed that the proximity of this 43,93 0,01 48,2 WT variant to the known pathogenic polymorphism m.11778G > A suggests No that it is located close or in an important functional domain and could potentially alter ND4′s function. Furthermore, 11719G > A has been Free energy of the thermodynamic ensemble (kcal/mol) found to be more prevalent in multiple sclerosis patients (Al-Kafaji et al., Freq of the MFE structure in the ensemble (%) 2022). Finally, variant m.14766C > T is a non-synonymous variant Centroid Minimum free energy (kcal/mol) located in the CYB gene. This mutation results in the amino-acid sub­ Minimum free energy (kcal/mol) Secondary structure differences stitution Thr7Ile. This change is involved in electron flow and proton pumping and could lead to a decrease in the efficiency of the complex III Ensamble diversity Region of analysis Q cycle and an increase in the production of ROS (Beckstead et al., 2009). Our stroke-associated case-control study based on mtDNA next- generation sequencing has allowed us to suggest several genetic risk modifiers for stroke. Several studies have analysed changes in the distribution of mtDNA variants mostly in relation to cancer (Kazdal et al., 2017; Pérez-Amado et al., 2020), Down Syndrome (Hefti et al., 2017) and Leber Hereditary Table 7 Optic Neuropathy (LOHN) (Piotrowska-Nowak et al., 2019b). However, to date, no study of this kind related to stroke has been published, being 58 A. Onieva et al. Mitochondrion 73 (2023) 51–61 Fig. 1. Entropy changes in tRNA Thr with variant m.15904C > T. A decrease in entropy can be observed in the nucleotides surrounding the variant m.15904 T. the present the first study of this type. In this regard, we observed an Balaban, R.S., Nemoto, S., Finkel, T., 2005. Mitochondria, oxidants, and aging. Cell 120 (4), 483–495. https://doi.org/10.1016/J.CELL.2005.02.001. increase in homoplasmic variants in complex III further supporting the Beckstead, W. A., Ebbert, M. T. W., Rowe, M. J., & McClellan, D. A. (2009). Evolutionary implication of this complex in ROS production and thus in stroke risk. Pressure on Mitochondrial Cytochrome b Is Consistent with a Role of CytbI7T The study of the mutational load proved the cumulative effect of Affecting Longevity during Caloric Restriction. PLoS ONE, 4(6). https://doi.org/ homoplasmic mtDNA complex IV variants with pathogenicity scores 10.1371/JOURNAL.PONE.0005836. Bhagavan, N. v., & Ha, C.-Eun. (2011). Essentials of medical biochemistry : with clinical above 0.5 in stroke risk and finally, the analyses of the different mtDNA cases. 581. variants exposed variants m.73A > G, m.11719G > A and m.14766C > T Brand, M.D., 2000. Uncoupling to survive? The role of mitochondrial inefficiency in as genetic risk factors. ageing. Exp. Gerontol. 35 (6–7), 811–820. https://doi.org/10.1016/S0531-5565(00) 00135-2. In our study, we did not find an association between mtDNA-CN and Bruno, C., Martinuzzi, A., Tang, Y., Andreu, A.L., Pallotti, F., Bonilla, E., Shanske, S., stroke risk. This lack of association is quite robust as it is not affected by Fu, J., Sue, C.M., Angelini, C., DiMauro, S., Manfredi, G., 1999. A stop-codon the impact of other variables also associated with changes in mtDNA- mutation in the human mtDNA cytochrome c oxidase I gene disrupts the functional structure of complex IV. Am. J. Hum. Genet. 65 (3), 611–620. https://doi.org/ CN. 10.1086/302546. In conclusion, our findings support the hypothesis that mtDNA Cadenas, S., 2018. Mitochondrial uncoupling, ROS generation and cardioprotection. variation involved in OXPHOS coupling efficiency and in other mecha­ Biochim. Biophys. Acta Bioenerg. 1859 (9), 940–950. https://doi.org/10.1016/j. bbabio.2018.05.019. nisms that modulate ROS production, influences stroke risk but these Campos, Y., García-Redondo, A., Fernández-Moreno, M.A., Martínez-Pardo, M., risk does not appear to be dependent on mtDNA copy number. Goda, G., Rubio, J.C., Martín, M.A., del Hoyo, P., Cabello, A., Bornstein, B., Garesse, R., Arenas, J., 2001. Early-onset multisystem mitochondrial disorder caused by a nonsense mutation in the mitochondrial DNA cytochrome C oxidase II gene. Declaration of Competing Interest Ann. Neurol. 50 (3), 409–413. https://doi.org/10.1002/ANA.1141. Chang, J.C., Kou, S.J., Lin, W.T., Liu, C.-S., 2010. Regulatory role of mitochondria in The authors declare that they have no known competing financial oxidative stress and atherosclerosis. World J. Cardiol. 2 (6), 150–159. https://doi. org/10.4330/wjc.v2.i6.150. interests or personal relationships that could have appeared to influence Chen, W., Huang, J., Hu, Y., Khoshnam, S.E., Sarkaki, A., 2020. Mitochondrial Transfer the work reported in this paper. as a Therapeutic Strategy Against Ischemic Stroke. Transl. Stroke Res. 11 (6), 1214–1228. https://doi.org/10.1007/S12975-020-00828-7. Chen, Y.R., Zweier, J.L., 2014. Cardiac mitochondria and reactive oxygen species Acknowledgements generation. Circ. Res. 114 (3), 524–537. https://doi.org/10.1161/ CIRCRESAHA.114.300559. This work is supported by Ministerio de Economía y Competitividad Chinnery, P.F., Elliott, H.R., Syed, A., Rothwell, P.M., 2010. Mitochondrial DNA haplogroups and risk of transient ischaemic attack and ischaemic stroke: a genetic (project: CGL2014-53781-R) and by Generalitat de Catalunya (Ref. 2017 association study. The Lancet. Neurology 9 (5), 498–503. https://doi.org/10.1016/ SGR 1630). We thank Dr. Pere Puig Casado for his advice regarding the S1474-4422(10)70083-1. statistical analyses. Chinnery, P.F., Gomez-Duran, A., 2018. Oldies but Goldies mtDNA Population Variants and Neurodegenerative Diseases. Front. Neurosci. 12 (OCT) https://doi.org/ 10.3389/FNINS.2018.00682. Appendix A. Supplementary material Clark, K.M., Taylor, R.W., Johnson, M.A., Chinnery, P.F., Chrzanowska-Lightowlers, Z.M. A., Andrews, R.M., Nelson, I.P., Wood, N.W., Lamont, P.J., Hanna, M.G., Lightowlers, R.N., Turnbull, D.M., 1999. An mtDNA mutation in the initiation codon Supplementary data to this article can be found online at https://doi. of the cytochrome C oxidase subunit II gene results in lower levels of the protein and org/10.1016/j.mito.2023.10.001. a mitochondrial encephalomyopathy. Am. J. Hum. Genet. 64 (5), 1330–1339. https://doi.org/10.1086/302361. Donkor, E.S., 2018. Stroke in the 21st Century: A Snapshot of the Burden, Epidemiology, References and Quality of Life. Stroke Research and Treatment 2018. https://doi.org/10.1155/ 2018/3238165. Addabbo, F., Montagnani, M., & Goligorsky, M. S. (2009). Mitochondria and reactive Eaaswarkhanth, M., Melhem, M., Sharma, P., Nizam, R., al Madhoun, A., Chaubey, G., oxygen species. Hypertension (Dallas, Tex. : 1979), 53(6), 885–892. https://doi.org/ Alsmadi, O., Al-Ozairi, E., & Al-Mulla, F. (2019). Mitochondrial DNA D-loop 10.1161/HYPERTENSIONAHA.109.130054. sequencing reveals obesity variants in an Arab population. The Application of Clinical Al-Kafaji, G., Bakheit, H.F., AlAli, F., Fattah, M., Alhajeri, S., Alharbi, M.A., Daif, A., Genetics, 12, 63. https://doi.org/10.2147/TACG.S198593. Alsabbagh, M.M., Alwehaidah, M.S., Bakhiet, M., 2022. Next-generation sequencing Elson, J.L., Herrnstadt, C., Preston, G., Thal, L., Morris, C.M., Edwardson, J.A., Beal, M. of the whole mitochondrial genome identifies functionally deleterious mutations in F., Turnbull, D.M., Howell, N., 2006. Does the mitochondrial genome play a role in patients with multiple sclerosis. PLoS One 17 (2). https://doi.org/10.1371/ the etiology of Alzheimer’s disease? Hum. Genet. 119 (3), 241–254. https://doi.org/ JOURNAL.PONE.0263606. 10.1007/S00439-005-0123-8. Andalib, S., Divani, A.A., Michel, T.M., Høilund-Carlsen, P.F., Vafaee, M.S., Gjedde, A., Feigin, V.L., Brainin, M., Norrving, B., Martins, S., Sacco, R.L., Hacke, W., Fisher, M., 2017. Pandora’s Box: mitochondrial defects in ischaemic heart disease and stroke. Pandian, J., Lindsay, P., 2022. World Stroke Organization (WSO): Global Stroke Fact Expert Rev. Mol. Med. 19 https://doi.org/10.1017/ERM.2017.5. Sheet 2022. Int. J. Stroke 17 (1), 18–29. https://doi.org/10.1177/ Ashar, F.N., Zhang, Y., Longchamps, R.J., Lane, J., Moes, A., Grove, M.L., Mychaleckyj, J. 17474930211065917. C., Taylor, K.D., Coresh, J., Rotter, J.I., Boerwinkle, E., Pankratz, N., Guallar, E., Fernández-Caggiano, M., Barallobre-Barreiro, J., Rego-Pérez, I., Crespo-Leiro, M.G., Arking, D.E., 2017. Association of mitochondrial DNA copy number with Paniagua, M.J., Grillé, Z., Blanco, F.J., Doménech, N., 2012. Mitochondrial cardiovascular disease. JAMA Cardiol. 2 (11), 1247–1255. https://doi.org/10.1001/ jamacardio.2017.3683. 59 A. Onieva et al. Mitochondrion 73 (2023) 51–61 Haplogroups H and J: Risk and Protective Factors for Ischemic Cardiomyopathy. Pereira, L., Soares, P., Radivojac, P., Li, B., Samuels, D.C., 2011. Comparing phylogeny PLoS One 7 (8), e44128. and the predicted pathogenicity of protein variations reveals equal purifying Filograna, R., Mennuni, M., Alsina, D., Larsson, N.G., 2021. Mitochondrial DNA copy selection across the global human mtDNA diversity. Am. J. Hum. Genet. 88 (4), number in human disease: the more the better? FEBS Lett. 595 (8), 976–1002. 433–439. https://doi.org/10.1016/J.AJHG.2011.03.006. https://doi.org/10.1002/1873-3468.14021. Pérez-Amado, C.J., Tovar, H., Gómez-Romero, L., Beltrán-Anaya, F.O., Bautista-Piña, V., Floros, V. I., Pyle, A., DIetmann, S., Wei, W., Tang, W. W. C., Irie, N., Payne, B., Capalbo, Dominguez-Reyes, C., Villegas-Carlos, F., Tenorio-Torres, A., Alfaro-Ruíz, L.A., A., Noli, L., Coxhead, J., Hudson, G., Crosier, M., Strahl, H., Khalaf, Y., Saitou, M., Hidalgo-Miranda, A., Jiménez-Morales, S., 2020. Mitochondrial DNA Mutation Ilic, D., Surani, M. A., & Chinnery, P. F. (2018). Segregation of mitochondrial DNA Analysis in Breast Cancer: Shifting From Germline Heteroplasmy Toward heteroplasmy through a developmental genetic bottleneck in human embryos. Homoplasmy in Tumors. Front. Oncol. 10, 2208. https://doi.org/10.3389/ Nature Cell Biology, 20(2), 144–151. https://doi.org/10.1038/s41556-017-0017-8. FONC.2020.572954/BIBTEX. Floßmann, E., Schulz, U.G.R., Rothwell, P.M., 2004. Systematic review of methods and Piotrowska-Nowak, A., Elson, J.L., Sobczyk-Kopciol, A., Piwonska, A., Puch-Walczak, A., results of studies of the genetic epidemiology of ischemic stroke. Stroke 35 (1), Drygas, W., Ploski, R., Bartnik, E., Tonska, K., 2019a. New mtDNA Association 212–227. https://doi.org/10.1161/01.STR.0000107187.84390.AA. Model, MutPred Variant Load, Suggests Individuals With Multiple Mildly Deleterious Fryar, C.D., Ostchega, Y., Hales, C.M., Zhang, G., Kruszon-Moran, D., 2017. Hypertension mtDNA Variants Are More Likely to Suffer From Atherosclerosis. Front. Genet. 9 Prevalence and Control Among Adults: United States, 2015–2016. NCHS Data Brief (JAN) https://doi.org/10.3389/FGENE.2018.00702. 289, 1–8. Piotrowska-Nowak, A., Kosior-Jarecka, E., Schab, A., Wrobel-Dudzinska, D., Bartnik, E., Giles, R. E., Blanc, H., Cann, H. M., & Wallace aD., C. (1980). Maternal inheritance of Zarnowski, T., Tonska, K., 2019b. Investigation of whole mitochondrial genome human mitochondrial DNA. Proceedings of the National Academy of Sciences of the variation in normal tension glaucoma. Exp. Eye Res. 178, 186–197. https://doi.org/ United States of America, 77(11), 6715–6719. https://doi.org/10.1073/ 10.1016/J.EXER.2018.10.004. PNAS.77.11.6715. Ramos, A., Santos, C., Mateiu, L., del Gonzalez, M., Alvarez, M., Azevedo, L., Amorim, L. Gnaiger, E., Boushel, R., Søndergaard, H., Munch-Andersen, T., Damsgaard, R., A., Aluja, M.P., 2013. Frequency and Pattern of Heteroplasmy in the Complete Hagen, C., Díez-Sánchez, C., Ara, I., Wright-Paradis, C., Schrauwen, P., Human Mitochondrial Genome. PLoS One 8 (10), 19–21. https://doi.org/10.1371/ Hesselink, M., Calbet, J.A.L., Christiansen, M., Helge, J.W., Saltin, B., 2015. journal.pone.0074636. Mitochondrial coupling and capacity of oxidative phosphorylation in skeletal muscle Robin, E.D., Wong, R., 1988. Mitochondrial DNA molecules and virtual number of of Inuit and Caucasians in the arctic winter. Scand. J. Med. Sci. Sports 25 (Suppl 4), mitochondria per cell in mammalian cells. J. Cell. Physiol. 136 (3), 507–513. 126–134. https://doi.org/10.1111/SMS.12612. https://doi.org/10.1002/JCP.1041360316. González, M.D.M., Ramos, A., Aluja, M.P., Santos, C., 2020. Sensitivity of mitochondrial Rosa, A., Fonseca, B. v., Krug, T., Manso, H., Gouveia, L., Albergaria, I., Gaspar, G., DNA heteroplasmy detection using Next Generation Sequencing. Mitochondrion 50, Correia, M., Viana-Baptista, M., Simões, R. M., Pinto, A. N., Taipa, R., Ferreira, C., 88–93. https://doi.org/10.1016/j.mito.2019.10.006. Fontes, J. R., Silva, M. R., Gabriel, J. P., Matos, I., Lopes, G., Ferro, J. M., … Oliveira, Hankey, G.J., 2017. Stroke. Lancet (London, England) 389 (10069), 641–654. https:// S. A. (2008). Mitochondrial haplogroup H1 is protective for ischemic stroke in doi.org/10.1016/S0140-6736(16)30962-X. Portuguese patients. BMC Med. Genet., 9(1), 1–10. https://doi.org/10.1186/1471- Hatefi, Y., 1985. The mitochondrial electron transport and oxidative phosphorylation 2350-9-57. system. Annu. Rev. Biochem 54 (1), 1015–1069. https://doi.org/10.1146/ Soares, P., Ermini, L., Thomson, N., Mormina, M., Rito, T., Röhl, A., Salas, A., ANNUREV.BI.54.070185.005055. Oppenheimer, S., Macaulay, V., Richards, M.B., 2009. Correcting for Purifying Hefti, E., Bard, J., Blanco, J.G., 2017. Analysis of Heteroplasmic Variants in the Cardiac Selection: An Improved Human Mitochondrial Molecular Clock. Am. J. Hum. Genet. Mitochondrial Genome of Individuals with Down Syndrome. Hum. Mutat. 38 (1), 48. 84 (6), 740–759. https://doi.org/10.1016/j.ajhg.2009.05.001. https://doi.org/10.1002/HUMU.23071. Soini, H.K., Moilanen, J.S., Finnila, S., Majamaa, K., 2012. Mitochondrial DNA sequence Ingman, M., Gyllensten, U., 2007. A recent genetic link between Sami and the Volga-Ural variation in Finnish patients with matrilineal diabetes mellitus. BMC. Res. Notes 5, region of Russia. European Journal of Human Genetics : EJHG 15 (1), 115–120. 350. https://doi.org/10.1186/1756-0500-5-350. https://doi.org/10.1038/SJ.EJHG.5201712. Sun, C., Kong, Q.P., Zhang, Y.P., 2007. The role of climate in human mitochondrial DNA Jeedigunta, S.P., Minenkova, A.V., Palozzi, J.M., Hurd, T.R., 2021. Avoiding Extinction: evolution: a reappraisal. Genomics 89 (3), 338–342. https://doi.org/10.1016/j. Recent Advances in Understanding Mechanisms of Mitochondrial DNA Purifying ygeno.2006.11.005. Selection in the Germline. Annu. Rev. Genomics Hum. Genet. 22, 55–80. https://doi. Sun, H., Yu, G., 2019. New insights into the pathogenicity of non-synonymous variants org/10.1146/annurev-genom-121420-081805. through multi-level analysis. Sci. Rep. 9 (1) https://doi.org/10.1038/S41598-018- Kazdal, D., Harms, A., Endris, V., Penzel, R., Kriegsmann, M., Eichhorn, F., Muley, T., 38189-9. Stenzinger, A., Pfarr, N., Weichert, W., & Warth, A. (2017). Prevalence of somatic Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving the sensitivity mitochondrial mutations and spatial distribution of mitochondria in non-small cell of progressive multiple sequence alignment through sequence weighting, position- lung cancer. British Journal of Cancer 2017 117:2, 117(2), 220–226. https://doi.org/ specific gap penalties and weight matrix choice. Nucleic Acids Res. 22 (22), 10.1038/bjc.2017.155. 4673–4680. https://doi.org/10.1093/NAR/22.22.4673. Krzywanski, D.M., Moellering, D.R., Fetterman, J.L., Dunham-Snary, K.J., Sammy, M.J., Tranah, G.J., Manini, T.M., Lohman, K.K., Nalls, M.A., Kritchevsky, S., Newman, A.B., Ballinger, S.W., 2011. The mitochondrial paradigm for cardiovascular disease Harris, T.B., Miljkovic, I., Biffi, A., Cummings, S.R., Liu, Y., 2011. Mitochondrial susceptibility and cellular function: a complementary concept to Mendelian genetics. DNA variation in human metabolic rate and energy expenditure. Mitochondrion 11 Laboratory Investigation; a Journal of Technical Methods and Pathology 91 (8), (6), 855–861. https://doi.org/10.1016/J.MITO.2011.04.005. 1122–1135. https://doi.org/10.1038/LABINVEST.2011.95. Umbria, M., Ramos, A., Santos, C., Aluja, M.P., 2017. The role of control region Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real- mitochondrial DNA mutations in cardiovascular disease: stroke and myocardial time quantitative PCR and the 2(-Delta Delta C(T)). Method. Methods (San Diego, infarction. En Revisión. Calif.) 25 (4), 402–408. https://doi.org/10.1006/METH.2001.1262. Umbria, M., Ramos, A., Caner, J., Vega, T., Lozano, J.E., Santos, C., Aluja, M.P., 2019. Lott, M.T., Leipzig, J.N., Derbeneva, O., Michael Xie, H., Chalkia, D., Sarmady, M., Involvement of mitochondrial haplogroups in myocardial infarction and stroke: A Procaccio, V., Wallace, D.C., 2013. mtDNA Variation and Analysis Using Mitomap case-control study in Castile and Leon (Spain) population. Mitochondrion 44, 1–6. and Mitomaster. Curr. Protoc. Bioinformatics 44 (123). https://doi.org/10.1002/ https://doi.org/10.1016/j.mito.2017.12.004. 0471250953.BI0123S44. Umbria, M., Ramos, A., Aluja, M.P., Santos, C., 2020. The role of control region Malik, R., Dichgans, M., 2018. Challenges and opportunities in stroke genetics. mitochondrial DNA mutations in cardiovascular disease: stroke and myocardial Cardiovasc. Res. 114 (9), 1226–1240. https://doi.org/10.1093/CVR/CVY068. infarction. Sci. Rep. 10 (1) https://doi.org/10.1038/S41598-020-59631-X. Mikkelsen, M., Rockenbauer, E., Wächter, A., Fendt, L., Zimmermann, B., Parson, W., van Oven, M., Kayser, M., 2009. Updated comprehensive phylogenetic tree of global Nielsen, S.A., Gilbert, T., Willerslev, E., Morling, N., 2009. Application of full human mitochondrial DNA variation. Hum. Mutat. 30 (2), 386–394. https://doi.org/ mitochondrial genome sequencing using 454 GS FLX pyrosequencing. Forensic 10.1002/humu.20921. Science International: Genetics Supplement Series 2 (1), 518–519. https://doi.org/ Vega Alonso, A.T., Lozano Alonso, J.E., Álamo Sanz, R., Llera Muñoz, S., Escribano 10.1016/J.FSIGSS.2009.09.028. Hernández, A., de la Iglesia Rodríguez, P., 2007. Design of a population-based study Mishmar, D., Ruiz-Pesini, E., Golik, P., Macaulay, V., Clark, A. G., Hosseini, S., Brandon, of cardiovascular risk in Castile and Leon [Spain] through primary care teams. Gac. M., Easleyf, K., Chen, E., Brown, M. D., Sukernik, R. I., Olckers, A., & Wallace, D. C. Sanit. 21 (1), 84–87. https://doi.org/10.1016/S0213-9111(07)71976-3. (2003). Natural selection shaped regional mtDNA variation in humans. Proceedings of Venter, M., Malan, L., van Dyk, E., Elson, J.L., van der Westhuizen, F.H., 2017. Using the National Academy of Sciences of the United States of America, 100(1), 171–176. MutPred derived mtDNA load scores to evaluate mtDNA variation in hypertension https://doi.org/10.1073/PNAS.0136972100. and diabetes in a two-population cohort: The SABPA study. Journal of Genetics and Moilanen, J.S., Finnila, S., Majamaa, K., 2003. Lineage-specific selection in human Genomics =. Yi Chuan Xue Bao 44 (3), 139–149. https://doi.org/10.1016/J. mtDNA: lack of polymorphisms in a segment of MTND5 gene in haplogroup. JGG.2016.12.003. J. Molecular biology and evolution 20 (12), 2132–2142. https://doi.org/10.1093/ Venter, M., van der Westhuizen, F.H., Elson, J.L., 2018. The aetiology of cardiovascular molbev/msg230. disease: A role for mitochondrial DNA? Cardiovasc. J. Afr. 29 (2), 122–132. https:// Munro, D., Treberg, J.R., 2017. A radical shift in perspective: mitochondria as regulators doi.org/10.5830/CVJA-2017-037. of reactive oxygen species. J. Exp. Biol. 220 (Pt 7), 1170–1180. https://doi.org/ Wallace, D.C., Chalkia, D., 2013. Mitochondrial DNA genetics and the heteroplasmy 10.1242/JEB.132142. conundrum in evolution and disease. Cold Spring Harb. Perspect. Biol. 5 (11) Nicholls, T.J., Minczuk, M., 2014. In D-loop: 40 years of mitochondrial 7S DNA. Exp. https://doi.org/10.1101/CSHPERSPECT.A021220. Gerontol. 56, 175–181. https://doi.org/10.1016/j.exger.2014.03.027. Wang, W., Sun, Y., Lin, Y., Xu, X., Zhao, D., Ji, K., Li, W., Zhao, Y., Yan, C., 2021. A novel Peluso, I., Morabito, G., Urban, L., Ioannone, F., Serafi, M., 2012. Oxidative Stress in nonsense variant in MT-CO3 causes MELAS syndrome. Neuromuscular Disorders : Atherosclerosis Development: The Central Role of LDL and Oxidative Burst. Endocr NMD 31 (6), 558–565. https://doi.org/10.1016/J.NMD.2021.02.020. Metab Immune Disord Drug Targets 12 (4), 351–360. https://doi.org/10.2174/ Wei, R., Ni, Y., Bazeley, P., Grandhi, S., Wang, J., Li, S.T., Hazen, S.L., Wilson Tang, W. 187153012803832602. H., LaFramboise, T., 2021. Mitochondrial DNA Content Is Linked to Cardiovascular 60 A. Onieva et al. Mitochondrion 73 (2023) 51–61 Disease Patient Phenotypes. Journal of the American Heart Association: variants. Genetics in Medicine : Official Journal of the American College of Medical Cardiovascular and Cerebrovascular Disease 10 (4), 1–720. https://doi.org/ Genetics 22 (5), 917–926. https://doi.org/10.1038/S41436-019-0746-0. 10.1161/JAHA.120.018776. World Medical Association. (2013). World Medical Association Declaration of Helsinki: Wei, W., Schon, K.R., Elgar, G., Orioli, A., Tanguy, M., Giess, A., Tischkowitz, M., ethical principles for medical research involving human subjects. JAMA, 310(20), Caulfield, M.J., Chinnery, P.F., 2022. Nuclear-embedded mitochondrial DNA 2191–2194. https://doi.org/10.1001/JAMA.2013.281053. sequences in 66,083 human genomes. Nature 611 (7934), 105–114. https://doi.org/ Yan, F., Tang, H., Wang, L., Huang, L., Zhang, J., 2022. Editorial: Mitochondrial 10.1038/s41586-022-05288-7. Dysfunction in Stroke. Front. Aging Neurosci. 14, 888952 https://doi.org/10.3389/ Weissensteiner, H., Forer, L., Fuchsberger, C., Schöpf, B., Kloss-Brandstätter, A., FNAGI.2022.888952. Specht, G., Kronenberg, F., Schönherr, S., 2016. mtDNA-Server: next-generation Yue, P., Jing, S., Liu, L., Ma, F., Zhang, Y., Wang, C., Duan, H., Zhou, K., Hua, Y., Wu, G., sequencing data analysis of human mitochondrial DNA in the cloud. Nucleic Acids Li, Y., 2018. Association between mitochondrial DNA copy number and Res. 44 (April), gkw247. https://doi.org/10.1093/nar/gkw247. cardiovascular disease: Current evidence based on a systematic review and meta- Wong, L.J.C., Chen, T., Wang, J., Tang, S., Schmitt, E.S., Landsverk, M., Li, F., Wang, Y., analysis. PLoS One 13 (11). https://doi.org/10.1371/JOURNAL.PONE.0206003. Zhang, S., Zhang, V.W., Craigen, W.J., 2020. Interpretation of mitochondrial tRNA 61