Introduction

Mitochondria are responsible for a diversity of essential functions in cardiomyocytes including the production of ATP by oxidative phosphorylation (OXPHOS), several signaling pathways, control of apoptosis, as well as control of the cytosolic calcium concentration [1]. Mitochondria contain 3-10 16,570 bp circular DNA molecules, mitochondrial DNA (mtDNA), comprising a non-coding control region and a coding region containing 37 genes. The genes encode 13 polypeptide components, constituting a small fraction of the many proteins involved in the OXPHOS system, consisting of complex I-IV and complex V (ATP synthase). The rest of the genes code for two species of ribosomal RNA (rRNAs) and 22 species of transfer RNAs (tRNA) required for intra-mitochondrial translation.

During evolution, mutations have accumulated sequentially along maternal mtDNA lineages forming groups of related mtDNA genotypes, haplogroups which tend to be regionally and ethnically specific [2] (Figure 1). The haplogroup defining variants encompass transitions and transversions in both protein- and RNA-coding genes (Tables 1 and 2), hence, mtDNA haplogroups are likely to confer different functional characteristics. Indeed, mtDNA haplogroup H has been shown to exhibit differences in VO_2max and reactive oxygen species (ROS) damage [3], and H, J and K to confer differences in penetrance or expression of Leber Hereditary Optic Neuropathy [4]. Likewise, haplogroup J, U and H has been associated with reduced and increased risk of Alzheimer and Parkinson disease [5,6] and haplogroup K with reduced risk of in transient ischaemic attack and ischaemic stroke [7]. In addition, haplogroup H and J are associated with increased and reduced of developing ischemic cardiomyopathy, respectively [8].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Simplified mtDNA tree based on PhyloTree mtDNA tree Build 15 (30 Sep 2012) (www.phylotree.org).

https://doi.org/10.1371/journal.pone.0071904.g001

The heart has the highest density of mitochondria and the highest oxygen uptake rate of human organs [9-11], consequently, functional differences in mitochondria must be expected to play a more prominent role in defining susceptibility to cardiac disease than to diseases in other organs. This is clearly supported by the many mitochondrial diseases, where cardiomyopathy is part of the clinical syndrome [1] and the finding of mitochondrial dysfunction in HCM [12]. HCM is the most common inherited cardiac disease with a prevalence of 1 in 500 individuals in the general population [13]. In adult patients, 50-60% of all cases are associated with mutations in genes coding for proteins involved in sarcomere function [14]. The disease exhibits a considerable intra- and inter-familial variability with respect to the clinical presentation, some affected present with syncope, dyspnea, angina and palpitations, whilst others are asymptomatic. HCM is a common cause of sudden cardiac death in young adults and athletes [15]. Recently, a study found that haplogroup T might be a risk factor for developing HCM in the Spanish population [16]. Mutations in mtDNA have been found to be frequently occurring (>1:200) in unselected neonates [17] and some of the mutations found have previously been found associated with HCM [18]. These findings indicate that mtDNA haplogroups, in particular J, K, U and H, might play a role as susceptibility factors or modifiers for HCM.

In the present study we examined whether the distribution of mtDNA haplogroups and haplogroup clusters (U+K, J+T, I+W+X and H+V) differed between a consecutively collected Danish HCM cohort and the normal population. We found that haplogroup H and haplogroup cluster HV were susceptibility factors with respect to developing HCM, whereas haplogroup J and K as well as haplogroup cluster UK were protective factors.

Materials and Methods

Results

The mtDNA haplogroup distribution of the 91 HCM patients is shown in Table 4. Also shown are the reported haplogroup distributions in a large Danish population (n= 9254) study by Benn et al. [24], performed in the same geographic location of Copenhagen as the hospital where the HCM patients, participating in this study, were recruited, and in a smaller Danish study (n=201) performed as a control study for forensic examinations by Mikkelsen et al. [25], Not all haplogroups were identified in the control population studies. MtDNA haplotyping performed by Mikkelsen et al. were based on variants in the hyper-variable regions 1 and 2 according to [26] and the mtDNA haplotying performed by Benn et al. were based on haplogroup specific markers in the coding region of the mtDNA (Table 2).

All major haplogroups were identified in this study and compared with the haplogroup frequencies obtained from two studies by Benn et al. and Mikkelsen et al.

Haplogroup H was found to have a higher frequency in HCM probands, 60% versus 41%(p = 0.006), and 46%, (p = 0.003) as compared to controls (Table 4). This difference translates into odds-ratios (OR) for having HCM of 1.8 (CI = [1.16-2.83]) and 2.17 (CI = [1.27-3.72]) (Tables 5 & 6), respectively. On the contrary, the prevalence of haplogroup J was significantly lower in HCM probands, 3.3% vs. 12.4% (p = 0.017) (Table 5) when compared to the forensic control population, yielding an OR=0.24, CI= [0.05-0.82] (Table 6). The difference in prevalence of haplogroup J was only borderline significant when the HCM group was compared with the population controls, 3.3% vs. 9.1% (p = 0.06) (Table 5) yielding an, OR = 0.34, CI = [0.07-1.03]) (Table 6). These results indicate that haplogroup H acts as a susceptibility factor and haplogroup J as a protective factor for HCM.

Comparing the prevalence of haplogroup K in the HCM group with that of the two control populations disclosed a significant difference (Table 4) with the forensic control group, p = 0.018 and an OR for HCM of 0.20, CI = [0.02-0.87] (Table 6). However, the difference was not significant when compared with the population based control population, p =0.181 and an OR= 0.34, CI= [0.04-1.28] (Table 5). None of the other haplogroup frequencies differed significantly between the HCM and control groups (Tables 5 and 6).

When comparing the prevalence of the evolutionarily related haplogroups (HV, JT, KU and IWX) (Figure 1), we found that the prevalence of the JT cluster was not significantly different from that of the control populations (Tables 5 and 6), indicating that the haplogroup defining variants distinguishing haplogroup J and T could explain the protective effect of haplogroup J. The prevalence of the UK cluster was found to be significantly reduced compared to both population controls (p = 0.02, OR = 0.49, CI= [0.23-0.92]) and forensic controls (p = 0.038, OR = 0.46, CI = [0.21-0.97]). This suggests that the UK cluster is a protective factor against the development of HCM.

The prevalence of the HV cluster was found to be significantly increased in HCM probands when compared to the population based control group, p = 0.003, OR = 1.91, CI = [1.22-3.07]. Haplogroup V was not identified in the forensic control material precluding a comparison with that group.

There was no appreciable difference with respect to age, gender distribution, frequency of sarcomere mutations identified and left ventricle dimensions as a function of haplogroup when the HCM group was broken down with respect to mtDNA haplogroups (Table 7).

To test for bias as a result of population stratification all HCM probands and 90 controls were analyzed for 15 short tandem repeats (STR). All STR loci investigated were in Hardy-Weinberg equilibrium, P=0.99 and P=0.83 for HCM and control populations respectively; a pairwise F_ST value of 0.003 indicates that there is no difference between the HCM population and the control population. Similarly, an F_ST of 0.001 indicates that there are no differences between the H-haplotype carrying HCM population and the non-H-haplotype carrying HCM population, indicating that the H-haplotype carrying population of HCM probands investigated here is not stratified. The HCM population is as mixed as the background population.

Discussion

Here we have shown that mtDNA haplogroup H and haplogroup cluster HV constitute susceptibility factors and haplogroup J and the UK cluster constitute protective factors for HCM in a Danish population. Our finding, that haplogroup H is a risk factor and J a protective factor is consistent with the recent finding that haplogroup H is a risk factor and J a protective factor for ischemic cardiomyopathy where ROS has the same putative etiological role [8].

The STR analysis showed that the association between haplogroup H and HCM is not a result of population stratification.

Haplogroup H is defined by the non-synonymous variant, m.14766C, p. MT-CYB: Thr7Ile (also defining HV and V), and the synonymous variant m.7028C in MT-COI and the rRNA variant m.2706G in MT-RNR2. p. MT-CYB: Thr7Ile is not conserved, CI=1.4% [27], but has been predicted, based on structural analysis, to have impact on the efficiency on ETS [28]. When comparing the haplogroup defining variants of H with the ones of haplogroup J, the differences in non-synonymous variants are, apart from m.C14766C, m.A10398G in MT-ND3, m. G13708A in MT-ND5, and m. C15452A in MT-CYB. These variants result in amino acid substitutions in subunits of Complex I and Complex III, both organized in supercomplexes with Complex IV in the constellation CI₁CIII₂CIV₁ [9,29,30] which are the primary sites of ROS generation. These mtDNA variants may influence the stability and/or activity of the supercomplexes and thereby induce variation in ATP production, electron leakage and/or ROS production [31,32]. As the control populations were not haplotyped to the level of sub-haplotypes, we cannot examine if there are significant differences in the sub-haplotype frequencies. Hence, it was not possible to identify single SNPs that could be responsible for the difference in association with HCM.

Haplogroup H has previously been reported to be physiologically different from other common European haplogroups; Martínez-Redondo et al. [3] found that mitochondria belonging to haplogroup H have higher VO_2max and is associated with increased mitochondrial oxidative damage compared to mitochondria of haplogroup J in skeletal muscle. However, the difference disappears with steady exercise [33]. The group suggests that the high VO_2max reflects that the electron transport system (ETS) is more tightly coupled in haplogroup H than in haplogroup J mitochondria. Consequently, haplogroup H confers higher ROS production than haplogroup J [3,34].

Altered mitochondrial function is evident in pathological cardiac hypertrophy and ROS is suggested to contribute to this alteration [12]. ROS act as signaling molecules in several pathways, which participate in the control of cell proliferation, differentiation, apoptosis and senescence [10,35,36]. ROS are cytotoxic if not controlled, as the reactive species can modify DNA, oxidate and inactivate iron–sulfur proteins and stimulate lipid peroxidation, causing altered function or degradation of these [1]. Cells have an antioxidant defense compromising several enzymes, such as mitochondrial superoxide dismutase (mtSOD), which maintain balance (redox homeostasis), establishing protection from and control of ROS generation [29]. Oxidative stress is known to cause contractile failure in the myocardium and can induce myocyte hypertrophy, apoptosis, interstitial fibrosis [10,37] and cardiomyopathy [38]. Increased ROS production may lead to myocardial remodeling through activation of matrix metalloproteinases [10,37]. In accordance with the tentative role of increased ROS production as the cause of the haplogroup associated differences in HCM susceptibility, the myocardial ROS level has been found to correlate with the functional reduction of the failing myocardium [39,40].

Interestingly, Benn et al (2008) found no association with over-all morbidity, longevity or ischemic cardiovascular disease and mtDNA haplogroups in the Danish population-based control population [24]. This suggests that the effect of mtDNA haplogroups is a modifier effect requiring the presence of a specific pathological process. This is in agreement with studies on transgenic mouse and rat models that have shown that there is an increased tension-dependent ATP consumption associated with mutations in sacomere genes causing HCM [41]. In thirty-two of the 91 HCM probands analysed here, HCM causing mutations in sarcomere genes had been identified. Such mutations may result in impaired ATP utilization and thereby an increase in the workload of the mitochondria and - concomitantly - increased ROS production. We suggest that the increased baseline ROS production in haplogroup H could lower the threshold of the cardiomyocytes and thereby exacerbate or precipitate HCM. Ascribing ROS as a modifying factor is in concordance with the finding that haplogroup J has a lower OXPHOS coupling, i.e. lower ATP and ROS production, than haplogroup H [3].

Clustering haplogroup UK and comparing with the two control populations indicated that cluster UK is a protective factor for HCM. Recently a study investigating differences in OXPHOS of haplogroup H and K (H1, H5, H13, K1 and K2), using osteosarcoma cybrids have reported that K cybrids had significantly lower mitochondrial inner membrane potential than H, which is in agreement with the significantly lower endogenous leaking and uncoupled respiration found in K per ETS unit compared to H [27]. They found no differences in ROS levels. However, as mentioned by the authors, methods used for ROS measurements are not reliable and this poses a challenge for precise quantitation of ROS [27,42,43]. The authors also mention that the only haplogroup defining non-synonymous SNP (m.14798T>C, p. MT-CYB: Phe18Leu) present in the cybrid cell K is also a marker for haplogroup J1c, and suggests that it is of functional importance because of its conservation and location in cytochrome b [27]. Furthermore, both J1c and K are underrepresented in patients with Parkinson’s disease and over-represented in centenarians and in patients with LHON and multiple sclerosis, indicating a similar physiological characteristic of these haplogroups [5,27,44].

Haplogroup T, defined by m.13368A, has previously been associated with an increased risk of developing HCM in a Spanish population [16]. Our study could not confirm this finding, however a trend is present (Tables 5 and 6). The discrepancy could be due to the differences in haplogroup/sub-clade frequencies in the Danish and Spanish population or specific differences that may modify the significance of the modifiers, through variation of other factors, e.g. ethnic background, life style and environmental influences [45,46].

We found no significant differences when comparing the mean age of diagnosis or the number of genetic diagnosed patient between the haplogroups. This supports that the mtDNA haplogroup is not causative of the disease. However, the small number of probands in most of the haplogroups precludes any firm conclusions as to a possible gender-bias or differential clinical characteristics, as previously described for the U haplogroup in Alzheimer’s disease [6]. Even when corrected conservatively for multiple testing using Bonferroni, the association with susceptibility for HCM was significant (p = 0.036 for Benn et al. and p = 0.021 for Mikkelsen et al.) for haplogroup H, whereas the association with other haplogorups where not (data not shown). However the findings are not just a result of multiple testing, but rather based on a hypothesis of experimentally founded functional and clinical significance of mtDNA haplogroups.

In conclusion, our findings indicate that mtDNA haplogroup H and haplogroup cluster HV are susceptibility factors for developing HCM. The opposite is seen for haplogroup J and the UK cluster. As ORs are considerable, i.e. ~ 2, ~ 0.3 and 0.5, respectively, we suggest that mtDNA haplotyping is assessed with respect to its potential as a risk marker for development of HCM in families with sarcomere mutations and in conditions predisposing to HCM of non-genetic etiology, i.e. hypertension, excessive exercise or obesity among others.