Figures
Abstract
Background
Osteoarthritis (OA) is the most common form of arthritis and has become an increasingly important public-health problem. However, the pathogenesis of OA is still unclear. In recent years, its correlation with mtDNA haplogroups attracts much attention. We aimed to perform a meta-analysis to investigate the association between mtDNA haplogroups and OA.
Methods
Published English or Chinese literature from PubMed, Web of Science, SDOS, and CNKI was retrieved up until April 15, 2014. Case-control or cohort studies that detected the frequency of mtDNA haplogroups in OA patients and controls were included. The quality of the included studies was evaluated by the Newcastle-Ottawa Scale (NOS) assessment. A meta-analysis was conducted to calculate pooled odds ratio (OR) with 95% confidence interval (CI) through the random or fixed effect model, which was selected based on the between-study heterogeneity assessed by Q test and I2 test. Subgroup analysis was performed to explore the origin of heterogeneity.
Results
A total of 6 case-control studies (10590 cases and 7161 controls) with an average NOS score of 6.9 were involved. For the analysis between mtDNA haplogroup J and OA, random model was selected due to high heterogeneity. No significant association was found initially (OR = 0.73, 95%CI: 0.52–1.03), however, once any study from UK population was removed the association emerged. Further subgroup analysis demonstrated that there was a significant association in Spain population (OR = 0.57, 95%CI: 0.46–0.71), but not in UK population. Also, subgroup analysis revealed that there was a significant correlation between cluster TJ and OA in Spain population (OR = 0.70, 95%CI: 0.58–0.84), although not in UK population. No significant correlation was found between haplogroup T/cluster HV/cluster KU and OA.
Citation: Shen J-M, Feng L, Feng C (2014) Role of mtDNA Haplogroups in the Prevalence of Osteoarthritis in Different Geographic Populations: A Meta-Analysis. PLoS ONE 9(10): e108896. https://doi.org/10.1371/journal.pone.0108896
Editor: Ramon Trullas, IIBB/CSIC/IDIBAPS, Spain
Received: May 6, 2014; Accepted: August 27, 2014; Published: October 23, 2014
Copyright: © 2014 Shen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by Natural Science Foundation of China, No. 81200446k, CF (http://www.nsfc.gov.cn/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Osteoarthritis (OA), also known as degenerative arthritis and proliferative arthritis, is a chronic progressive disease characterized with degenerative changes and secondary hyperosteogeny of articular cartilage, which leads to pain and dysfunction of joints. OA is the most common form of arthritis, rare before 40 years old, and rising steeply with age. In fact, OA occurs in more than 50% of the population older than 65 years old [1], and is the leading cause of global disability [2]. Due to the aged tendency of population, it is estimated that the population suffering with OA may double in the next three decades [3]. So far, the pathogenesis of OA is still unclear, but it is now accepted that genetic factors play a role in it. Familial aggregation [4] and classical twin [5] studies have both demonstrated the key contribution of genetic susceptibility in OA.
The mitochondrion is a membrane-enclosed organelle that converts dietary intake into ATP by oxidative phosphorylation (OXPHOS) and meanwhile produces reactive oxygen species (ROS) which may lead to oxidative damage of the cell [6]. A great amount of evidence has suggested that mitochondrial degeneration may be associated with some human diseases, including Alzheimer disease, Parkinson disease, Huntington disease, cancer and some aging processes [7]–[9]. Moreover, mitochondrial dysfunction may contribute to the damage in articular cartilage and participate in the occurrence and progression of OA [10].
Mitochondrial DNA (mtDNA) encodes 13 of their mitochondrial proteins [11]. According to the various single nucleotide polymorphisms (SNPs) in mtDNA coding regions, related groups of mtDNA haplogroups are defined [12]. Recently, the association between mtDNA haplogroups and OA has become a hot area of research. Some studies reported that specific mtDNA haplogroups were correlated with the development and progression of OA [13]–[17], while others detected no significant association [18]. The discrepancy may be due to many factors, such as different ethnicity, geological variation and relatively small sample size.
Provided with the studies published in the past year, it is possible to collect sufficient data to explore the authentic interrelation between mtDNA haplogroups and OA. Therefore, in an attempt to ascertain the correlation of mtDNA haplogroups and risk of OA, we performed a meta-analysis to strengthen the statistical power of individual studies.
Materials and Methods
This meta-analysis was developed according to the PRISMA guidelines (Checklist S1). A protocol was registered in PROSPERO, with registration number: CRD42014009640.
Search strategy
Relevant studies published in English or Chinese were identified by a computerized search via 4 databases, including PubMed, Web of Science, Elsevier ScienceDirect (SDOS), and China National Knowledge Infrastructure (CNKI). The search strategy involved the use of the following key words: Osteoarthritis AND (mtDNA OR “mitochondrial DNA”) AND (haplogroup OR haplotype OR genotype OR “genetic predisposition” OR SNP OR polymorphism* OR variant* OR “genetic susceptibility” OR genetics OR allele), with details in Methods S1. The reference lists of retrieved articles were hand-searched to collect other relevant studies. The non-published studies were searched in the database of OpenSIGLE (System for Information on Grey literature in Europe, http://www.opengrey.eu). The literature in these databases from inception to April 15, 2014 was searched by two separate reviewers. If more than one article was published with the same case series, only the study with the largest sample size was selected.
Inclusion and exclusion criteria
The studies to meet the following inclusion criteria were screened: (1) evaluating the association between haplogroups of mtDNA and the risk of OA; (2) case-control or cohort design; (3) sufficient data provided to calculate odds ratio (OR) with the corresponding 95% confidence interval (95% CI). Studies were excluded if they: (1) explored the correlation between haplogroups of mtDNA and the progression of OA, not the prevalence; (2) focused on other mtDNA mutation except haplogroups. All identified studies were reviewed independently by two investigators to determine if an individual study should be included in this study and the discrepancies were resolved by discussion.
Data extraction
The following data were extracted from each study by two investigators independently: (1) name of the first author; (2) year of publication; (3) country; (4) ethnicity; (5) gender of subjects; (6) age of subjects; (7) diagnosis criteria; (8) types of OA; (9) methods of genotype; (10) conclusion of the study; (11) study type; (12) controlled confounders; and (13) numbers of genotypes in cases and controls. We contacted investigators for additional information when necessary.
Quality assessment of study
The quality of all the included studies was evaluated independently by two authors using the Newcastle-Ottawa Scale (NOS) assessment [19]. Any disagreements were addressed by discussion. The NOS assessment consists of three parameters of quality appraisal: selection of cases and controls, comparability of case and control groups, and ascertainment of exposure.
Statistical analysis
The degree of between-study heterogeneity was determined by Q test and I2 test. A P value (Q test) ≥0.05 or I2<50% was considered to demonstrate no obvious heterogeneity, and then a fixed-effect model (FEM) was applied to calculate the pooled OR, whereas a P value <0.05 or I2≥50% was considered to indicate the presence of heterogeneity, and then a random-effect model (REM) was adopted [20]. The associations of mtDNA haplogroups with OA were evaluated by calculating the pooled ORs and 95% CI. The significance of the pooled OR was assessed by Z test. Subgroup analysis of country was performed to explore the origin of heterogeneity when there is severe between-study heterogeneity. To identify the stability of the results, we conducted a sensitivity analysis by removing an individual study each time. Revman 5 was used for majority of the analysis performed here and STATA version 11.0 was applied to evaluate the publication bias. P<0.05 was considered statistically significant.
Results
Characteristics of the included studies
The database and reference list search identified a total of 874 non-duplicated articles (Fig. 1). No non-published data that met our eligibility criteria were found. Through reading the titles and abstracts, 32 potentially relevant articles were remained for further full-text assessment. Due to the reasons listed in the flow chart, 11 remaining articles were initially included, of which, 5 articles were excluded because of overlapping patients populations [21]–[25]. Finally, 6 studies were included in our meta-analysis.
American College of Rheumatology (ACR) criteria [26] were adopted to make the diagnosis of OA in all the included studies, except the study by Hudson et al. [18], in which the radiographic evidence of disease (Kellgren-Lawrence grade ≥2) or clinical evidence of disease to a level requiring joint replacement was used. The included studies were all case-control studies and generated information for 10590 cases diagnosed with hip or knee OA and 7161 controls. Both the OA and control groups in each study contained female and male cases. According to the reported mtDNA haplogroups that probably correlated with OA risk, we investigated the relationship between haplogroups J, T, G and B and the risk of OA respectively in this meta-analysis. For haplogroups J and T, Caucasian population was recruited; while for haplogroups G and B, East Asian population was examined.
The detailed characteristics of the included studies are shown in the Table 1.
Quality assessment results
The details of the quality evaluation for every study are shown in Table 1. The average NOS score was 6.9. The most common biases were representativeness of the cases and selection of controls.
Meta-analysis results
Since severe between-study heterogeneity for haplogroup J (I2 = 80%, P<0.01) was revealed, a random effect model was selected. Overall, the results showed that there was no significant association between haplogroup J and the risk of OA (OR = 0.73; 95%CI: 0.52–1.03, P>0.05) (Fig. 2). The subgroup analysis stratified by country revealed that there was a significant association in Spain population (OR = 0.57, 95%CI: 0.46–0.71, P<0.01), but no significant association in UK population (OR = 1.19, 95%CI: 0.72–1.95, P>0.05). No significant correlation was found between haplogroup T and risk of OA (OR = 1.01, 95%CI: 0.91–1.12, P>0.05), with moderate between-study heterogeneity (I2 = 49%, P>0.05) (Fig. 3). Further subgroup analysis confirmed the irrelevance between them.
Considering that the mtDNA haplogroup effects may be due to the common features shared by the clusters, we compared the frequency of the clusters HV, KU and TJ between OA and control groups. Since there was severe between-study heterogeneity for cluster TJ (I2 = 76%, P<0.05), subgroup analysis stratified by country was performed. There was a significant correlation between cluster TJ and OA in Spain population (OR = 0.70, 95%CI: 0.58–0.84, P<0.05), although no significant correlation in UK population (OR = 1.02, 95%CI: 0.94–1.11, P>0.05) (Fig. 4). No significant association was found between the cluster HV/KU and OA (OR = 1.04, 95%CI: 0.98–1.11, P>0.05/OR = 0.97, 95%CI: 0.90–1.04, P>0.05) with mild between-study heterogeneity (I2 = 11%, P>0.05/I2 = 0%, P>0.05). Further subgroup analysis confirmed the irrelevance between cluster KU and OA (Fig. 5). A marginal correlation between cluster HV and OA arose in Spain population (OR = 1.17, 95%CI: 1.01–1.35) (Fig. 6).
Sensitivity analysis
Sensitivity analysis was performed by removing any study at one time. For haplogroup J, once a study from UK population [16], [18] was removed, the association between haplogroup J and risk of OA emerged, with ORs and 95%CIs 0.67 (0.46–0.98) [18] and 0.65 (0.45–0.92) [16] respectively; however, removing any study from Spanish population did not impact on the significance of results, with ORs and 95%CIs ranging from 0.75 (0.50–1.11) to 0.80 (0.57–1.12). For cluster TJ/HV, when the study by Hudson et al [18] was removed, there was a significant correlation with ORs and 95%CIs 0.73 (0.62–0.86)/1.15(1.01–1.30). For haplogroup T/cluster KU, the pooled effect was stable and was not influenced significantly by any study, with ORs and 95%CIs ranging from 0.83 (0.66–1.03) to 1.04 (0.94–1.16)/from 0.97(0.90–1.04) to 0.97(0.83–1.13).
Publication bias
Egger's regression test and Begg's funnel plots were applied to evaluate the potential publication bias and suggested no severe publication bias (for haplogroup J: P = 0.202 and P = 0.707 separately; for haplogroup T: P = 0.151 and P = 0.462 separately; for cluster TJ: P = 0.050 and P = 0.462 separately; for cluster HV: P = 0.135 and P = 0.086 separately; for cluster KU: P = 0.752 and P = 0.221 separately) (Fig. 7).
Studies will be distributed symmetrically beside the horizontal line in absence of publication bias.
For haplogroup TJ, the Egger's test showed that the publication bias was suspected with P = 0.05, so the trim and fill analysis was carried out to evaluate the impact of any asymmetry. However, the pooled estimate was not changed.
Discussion
OA is a major cause of pain and impaired mobility in the elderly, and as the aging of population in modern society OA becomes an increasingly important public-health problem. OA is a multi-factor disease with both genetic and environmental causes [27]. It was reported that there was mtDNA deletion mutation and mtDNA damage [28]–[30] in the articular chondrocytes of OA, indicating an association between OA and mtDNA mutation [31]. In 2008 an association between mtDNA haplogroup J and OA was found, which proposed that haplogroup J protected against OA [13]. Later some studies demonstrated the association between mtDNA haplogroup J and risk of OA [14], [17]. However, a large population study in 2003 proposed no evidence of an association between all the haplogroups in Europe and OA. It is quite urgent to ascertain the correlation between haplogroups and OA, and the studies published recently made it possible to carry out a meta-analysis to elucidate this problem. To our best knowledge, this is the first meta-analysis addressing the association between mtDNA haplogroups and the risk of OA.
In this study we found that there was an association between haplogroup J/cluster TJ and risk of OA once a study from UK population was removed. Subgroup analysis also revealed that there was a significant association between haplogroup J/cluster TJ and risk of OA in Spain population, but not in UK population. It seems that there is a marginal correlation between cluster HV and risk of OA in Spain population. The investigation in Asian population indicated that haplogroup G was a risk factor and haplogroup B was protective [15]. All above suggests that mtDNA haplogroups do link to the risk of OA; however, the different ethnic and geological factors have a great influence on the association.
Considering the discrepancies of the association between mtDNA haplogroups and OA in different populations, several probabilities may be taken into account. Firstly, mtDNA haplogroup J consists of two major clades that can be further divided into more clades [32]. The distribution of mtDNA haplogroups were influenced by the past climate events [33], [34] and the constitution of the subhaplogroups of haplogroup J in different regions may be different. Seemingly some specific subhaplogroups function in the protection of OA [14], so the different constitution of subhaplogroups in different countries leads to the inconsistent results. Secondly, mutation of mtDNA may be complicated with epigenetic modification of nuclear genome, which contributes to various phenotypes in the same mtDNA mutation [35], [36]. The crosstalk among mtDNA, nucleus genome, epigenetics, and environment may lead to the different contribution of specific mtDNA haplogroups in OA in different populations.
The exact mechanism underlying the association of mtDNA haplogroups and OA risk remains unclear. Mitochondria produce ROS that play an important role in apoptosis [37]. Chondrocytes in OA exhibit a production increase of ROS and chronic oxidant stress, which may increase chondrocyte death and finally lead to OA [38]. Haplogroup J carriers harbor mtDNA missense mutations that partially uncouple OXPHOS, leading to less ATP production and meanwhile decreasing mitochondrial ROS generation by increasing the oxidation of electron transport chain (ETC). This reduction might decrease the oxidative damage to the cell, protect against apoptosis [39], [40], reduce cartilage degradation, and avoid the risk of OA.
The present study has some limitations. First, all the included studies were case-control studies, which tended to contain selection bias. In addition, the included studies adopted various control groups, including hospital controls with ailments unrelated to OA [14], [15], [17], population with joint fracture [16], university population [17], birth cohort [18], and blood service control groups [18]. Therefore, it is difficult to control the confounding factors and selection bias. Second, the publication bias was suspected for haplogroup TJ. Publication bias may result in an inaccurate estimation of the association between haplogroup TJ and OA. However, the results remained unchanged after “trim and fill” analysis, indicating that the publication bias of haplogroup TJ was not severe. Third, only two studies investigated the mtDNA subhaplogroups [14], [15]. Since we had insufficient data of the contribution of various subhaplogroups, we could not perform the meta-analysis on subhaplogroups data. Fourth, the major geographic cohorts involved in investigating the correlation between OA and mtDNA haplogroups are Spain, England, and China, which are still quite limited. For haplogroups G and B, only one study could be included [15]. More investigations are required to clarify the authentic association and the future studies should pay more attention to geographical difference.
In conclusion, our meta-analysis suggests that mtDNA haplogroup J and cluster TJ correlate with the risk of OA in Spanish population; however, the association in other populations is still not clear. Further investigations in different regions, with satisfying controls, and containing the data of mtDNA subhaplogroups are required to clarify the correlation and the underlying mechanism.
Supporting Information
Methods S1.
Search strategies in PubMed, Web of Science, SDOS and CNKI.
https://doi.org/10.1371/journal.pone.0108896.s002
(DOCX)
Author Contributions
Conceived and designed the experiments: JMS. Performed the experiments: JMS LF. Analyzed the data: CF. Contributed reagents/materials/analysis tools: JMS CF. Wrote the paper: LF CF.
References
- 1. Sangha O (2000) Epidemiology of rheumatic diseases. Rheumatology (Oxford) 39 Suppl 2: 3–12.
- 2. Cross M, Smith E, Hoy D, Nolte S, Ackerman I, et al. (2014) The global burden of hip and knee osteoarthritis: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis 73: 1323–1330.
- 3. Blanco FJ, Rego I, Ruiz-Romero C (2011) The role of mitochondria in osteoarthritis. Nat Rev Rheumatol 7: 161–169.
- 4. McDonnell SM, Sinsheimer J, Price AJ, Carr AJ (2007) Genetic influences in the aetiology of anteromedial osteoarthritis of the knee. J Bone Joint Surg Br 89: 901–903.
- 5. MacGregor AJ, Antoniades L, Matson M, Andrew T, Spector TD (2000) The genetic contribution to radiographic hip osteoarthritis in women: results of a classic twin study. Arthritis Rheum 43: 2410–2416.
- 6. Sahin E, Colla S, Liesa M, Moslehi J, Muller FL, et al. (2011) Telomere dysfunction induces metabolic and mitochondrial compromise. Nature 470: 359–365.
- 7. Yan MH, Wang X, Zhu X (2013) Mitochondrial defects and oxidative stress in Alzheimer disease and Parkinson disease. Free Radic Biol Med 62: 90–101.
- 8. Reddy PH, Shirendeb UP (2012) Mutant huntingtin, abnormal mitochondrial dynamics, defective axonal transport of mitochondria, and selective synaptic degeneration in Huntington's disease. Biochim Biophys Acta 1822: 101–110.
- 9. de Moura MB, dos SLS, Van Houten B (2010) Mitochondrial dysfunction in neurodegenerative diseases and cancer. Environ Mol Mutagen 51: 391–405.
- 10. Kim J, Xu M, Xo R, Mates A, Wilson GL, et al. (2010) Mitochondrial DNA damage is involved in apoptosis caused by pro-inflammatory cytokines in human OA chondrocytes. Osteoarthritis Cartilage 18: 424–432.
- 11. Shigenaga MK, Hagen TM, Ames BN (1994) Oxidative damage and mitochondrial decay in aging. Proc Natl Acad Sci U S A 91: 10771–10778.
- 12. Torroni A, Huoponen K, Francalacci P, Petrozzi M, Morelli L, et al. (1996) Classification of European mtDNAs from an analysis of three European populations. Genetics 144: 1835–1850.
- 13. Rego-Perez I, Fernandez-Moreno M, Fernandez-Lopez C, Arenas J, Blanco FJ (2008) Mitochondrial DNA haplogroups: role in the prevalence and severity of knee osteoarthritis. Arthritis Rheum 58: 2387–2396.
- 14. Rego I, Fernandez-Moreno M, Fernandez-Lopez C, Gomez-Reino JJ, Gonzalez A, et al. (2010) Role of European mitochondrial DNA haplogroups in the prevalence of hip osteoarthritis in Galicia, Northern Spain. Ann Rheum Dis 69: 210–213.
- 15. Fang H, Liu X, Shen L, Li F, Liu Y, et al. (2014) Role of mtDNA haplogroups in the prevalence of knee osteoarthritis in a southern Chinese population. Int J Mol Sci 15: 2646–2659.
- 16. Soto-Hermida A, Fernandez-Moreno M, Oreiro N, Fernandez-Lopez C, Rego-Perez I, et al. (2014) mtDNA haplogroups and osteoarthritis in different geographic populations. Mitochondrion 15: 18–23.
- 17. Fernandez-Moreno M, Tamayo M, Soto-Hermida A, Mosquera A, Oreiro N, et al. (2011) mtDNA haplogroup J modulates telomere length and nitric oxide production. BMC Musculoskelet Disord 12: 283.
- 18. Hudson G, Panoutsopoulou K, Wilson I, Southam L, Rayner NW, et al. (2013) No evidence of an association between mitochondrial DNA variants and osteoarthritis in 7393 cases and 5122 controls. Ann Rheum Dis 72: 136–139.
- 19.
Wells GA, Shea B, O'Connell D, Peterson J, Welch V, et al.. (2000) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta analyses. Available: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed 2014 July 1.
- 20. Higgins JP, Thompson SG (2002) Quantifying heterogeneity in a meta-analysis. Stat Med 21: 1539–1558.
- 21. Fernandez-Moreno M, Soto-Hermida A, Oreiro N, Pertega S, Fenandez-Lopez C, et al. (2012) Mitochondrial haplogroups define two phenotypes of osteoarthritis. Front Physiol 3: 129.
- 22. Rego-Perez I, Fernandez-Moreno M, Deberg M, Pertega S, Fernandez-Lopez C, et al. (2011) Mitochondrial DNA haplogroups and serum levels of proteolytic enzymes in patients with osteoarthritis. Ann Rheum Dis 70: 646–652.
- 23. Fernandez-Moreno M, Soto-Hermida A, Pertega S, Oreiro N, Fernandez-Lopez C, et al. (2011) Mitochondrial DNA (mtDNA) haplogroups and serum levels of anti-oxidant enzymes in patients with osteoarthritis. BMC Musculoskelet Disord 12: 264.
- 24. Rego-Perez I, Fernandez-Moreno M, Deberg M, Pertega S, Fenandez-Lopez C, et al. (2010) Mitochondrial DNA haplogroups modulate the serum levels of biomarkers in patients with osteoarthritis. Ann Rheum Dis 69: 910–917.
- 25. Rego I, Fernández-Moreno M, Fernández-López C, González A, Gómez-Reino JJ, et al. (2008) 370 The mtDNA haplogroup J is a protective factor to hip and knee osteoarthritis. Osteoarthritis and Cartilage 16, Supplement 4: S159.
- 26. Altman R, Asch E, Bloch D, Bole G, Borenstein D, et al. (1986) Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association. Arthritis Rheum 29: 1039–1049.
- 27. Dieppe PA, Lohmander LS (2005) Pathogenesis and management of pain in osteoarthritis. Lancet 365: 965–973.
- 28. Grishko VI, Ho R, Wilson GL, Pearsall AW 4th (2009) Diminished mitochondrial DNA integrity and repair capacity in OA chondrocytes. Osteoarthritis Cartilage 17: 107–113.
- 29. Grishko V, Xu M, Ho R, Mates A, Watson S, et al. (2009) Effects of hyaluronic acid on mitochondrial function and mitochondria-driven apoptosis following oxidative stress in human chondrocytes. J Biol Chem 284: 9132–9139.
- 30. Gavriilidis C, Miwa S, von ZT, Taylor RW, Young DA (2013) Mitochondrial dysfunction in osteoarthritis is associated with down-regulation of superoxide dismutase 2. Arthritis Rheum 65: 378–387.
- 31. Lu HB, Zhou Y, Hu JZ, Lei GH, Zhu M, et al. (2006) Mitochondrial DNA deletion mutations in articular chondrocytes of cartilage affected by osteoarthritis. Zhong Nan Da Xue Xue Bao Yi Xue Ban 31: 640–644.
- 32. Pala M, Olivieri A, Achilli A, Accetturo M, Metspalu E, et al. (2012) Mitochondrial DNA signals of late glacial recolonization of Europe from near eastern refugia. Am J Hum Genet 90: 915–924.
- 33. Vilaça ST, Biosa D, Zachos F, Iacolina L, Kirschning J, et al. (2014) Mitochondrial phylogeography of the European wild boar: the effect of climate on genetic diversity and spatial lineage sorting across Europe. Journal of Biogeography 41: 987–998.
- 34. Mishmar D, Ruiz-Pesini E, Golik P, Macaulay V, Clark AG, et al. (2003) Natural selection shaped regional mtDNA variation in humans. Proc Natl Acad Sci U S A 100: 171–176.
- 35. Carelli V, Giordano C, d'Amati G (2003) Pathogenic expression of homoplasmic mtDNA mutations needs a complex nuclear-mitochondrial interaction. Trends Genet 19: 257–262.
- 36. Chinnery PF, Elliott HR, Hudson G, Samuels DC, Relton CL (2012) Epigenetics, epidemiology and mitochondrial DNA diseases. Int J Epidemiol 41: 177–187.
- 37. Li N, Ragheb K, Lawler G, Sturgis J, Rajwa B, et al. (2003) Mitochondrial complex I inhibitor rotenone induces apoptosis through enhancing mitochondrial reactive oxygen species production. J Biol Chem 278: 8516–8525.
- 38. Terkeltaub R, Johnson K, Murphy A, Ghosh S (2002) Invited review: the mitochondrion in osteoarthritis. Mitochondrion 1: 301–319.
- 39. Ruiz-Pesini E, Mishmar D, Brandon M, Procaccio V, Wallace DC (2004) Effects of purifying and adaptive selection on regional variation in human mtDNA. Science 223–226.
- 40. Wallace DC (2005) A mitochondrial paradigm of metabolic and degenerative diseases, aging, and cancer: a dawn for evolutionary medicine. Annu Rev Genet 39: 359–407.