mtDNA variability determines spontaneous joint aging damage in a conplastic mouse model

Introduction

Aging involves the progressive loss of physiological integrity, with consequent time-dependent functional decline and increased vulnerability to mortality. The aging phenotype is determined by several cellular and molecular hallmarks, such as genomic instability, cellular senescence, telomere attrition, and mitochondrial dysfunction [1].

Aging-related changes affect all of the joint tissues, including cartilage, menisci, subchondral bone, synovium, and ligaments. Among these tissues, the articular cartilage, with alterations in chondrocytes and the disruption of extracellular matrix homeostasis, presents the most intense aging-related changes. In association with other risk factors, such as obesity, mechanical injury, and genetics, these aging-related changes may lead to an earlier onset or the development of a more severe form of osteoarthritis (OA) [2]. However, it is important to note that OA is not an inevitable consequence of aging. Joint aging is a process that is distinct from the changes caused by the initiation of OA [3].

Many in vitro and in vivo studies have shown the contribution of mitochondria to various processes associated with aging, including inflammation, senescence, and age-dependent decline in tissue and organ function [4]. Interestingly, studies using transmitochondrial cybrids have demonstrated that mitochondria and mtDNA variation can regulate different nuclear target genes [5].

Increased production of ROS with consequent mtDNA damage is a result of dysfunction of the mitochondrial respiratory complex, which characterizes the aging phenotype of OA [6]. Apoptosis, decreased autophagy, and progressive telomere shortening accompany this gradual decline in mitochondrial function, which results in an increase of cartilage matrix destruction [7, 8]. Regarding the direct relationship between mitochondrial haplogroups and aging, several studies found that certain mtDNA variants, including the mtDNA haplogroup J, have a protective role against aging with a decrease in ROS production and a consequent decline in mtDNA damage [9, 10]. Moreover, many studies have reported the association of these mtDNA variants with increased human longevity [11–13]. Interestingly, in OA, haplogroup J was found to be associated with less apoptosis and the production of nitric oxide by articular chondrocytes [14]. However, such association studies are affected by various confounding factors making the assessment of patients difficult, such as nuclear genetic variability, lifestyle differences, and environmental factors.

An interesting tool that can be used to control for confounding effects is conplastic mice [13]. These mice have the same nuclear background, but different wild-type mtDNA variants, and are maintained under identical conditions. This allows study of the functions of the mtDNA variants in an in vivo context. Various strains of conplastic mice have been developed [14]. In a recent study, we showed that the extension of a healthy life was highly dependent on the mtDNA variant [15]. In the current study, the conplastic mice employed were generated with the C57BL/6JOlaHsd nuclear genome and NZB/OlaHsd mtDNA. The difference between the mtDNA of the NZB/B1NJ strain and BL/6J mtDNA resembles the sequence difference between human mtDNAs of distant populations, at around 90 single-nucleotide polymorphisms (SNPs). The mtDNAs of C57BL/6 and NZB/OlaHsd mice differ by 12 missense mutations, 4 transfer RNA (tRNA) mutations, 8 ribosomal RNA (rRNA) mutations, and 10 non-coding-region mutations. These correspond to a level of divergence comparable to that between human Eurasian and African mtDNAs [15].

In this study, we evaluated whether the substitution of the mtDNA variant from C57BL/6JOlaHsd with the mtDNA from NZB/OlaHsd would change the aging-related joint damage. For this purpose, we used an aging model to analyze the joint alterations in both strains. Using histological and immunohistochemical assays, we found that the mtDNA replacement reduced aging-related joint damage.

Results

Discussion

Aging is a biological process characterized by the disruption of cellular homeostasis, progressive tissue deterioration, and an increased risk of cell death [1]. As individuals age, their joints show some changes in mobility due in part to alterations in the connective tissues.

The cellular changes that occur during aging contribute to the development of OA, but OA is not a simple consequence of joint aging. In fact, not all older adults develop this pathology. Thus, OA is a multifactorial disease, of which aging is one of the major risk factors but not the only one. Indeed, there are important differences between an aged joint and one with OA. The link between aging and OA was believed to be only a result of continuous mechanical “wear and tear” of articular cartilage. However, the age-related changes to cartilage matrix and the progressive accumulation of damaged organelles and macromolecules are now known to contribute to the initiation and progression of cartilage degradation [19]. Several hallmarks of aging, including genomic instability, cellular senescence, epigenetic alterations, and mitochondrial dysfunction, contribute to OA [20].

The aging phenotype of OA is characterized by a change in the activity of mitochondrial respiratory complexes, which causes an increase of ROS production and consequent mtDNA damage [6]. Many in vitro studies have found an association between aging and mtDNA variants [9, 21–23]. For in vivo analysis of the functional role of mtDNA variants in aging-related joint damage, we generated for the first time to our knowledge a mouse model of spontaneous aging in conplastic mouse strains.

Conplastic mice have emerged as a potent tool to study the function of mtDNA variants in an in vivo setting. These mice present the same nuclear background, but have different mtDNA variants. Here, we used two wild-type mtDNA variants derived from C57BL/6JOlaHsd (BL/6^C57) and NZB/OlaHsd (BL/6^NZB), which correspond to a level of divergence comparable to that between human Eurasian and African mtDNAs. The proof-of-concept study demonstrated that the mtDNA haplotype profoundly influences mitochondrial proteostasis and reactive oxygen species generation, insulin signaling, and aging parameters, including telomere shortening and mitochondrial dysfunction, resulting in profound differences in health longevity between conplastic strains [15]. Notably, the abnormal accumulation of mtDNA mutations (mutator mice) induces premature aging and affects musculoskeletal tissues [24, 25].

The aim of this project was to evaluate whether the substitution of mtDNA would also result in changes in the joints during spontaneous aging. Our results showed that conplastic (BL/6^NZB) mice developed fewer signs of aging than the original strain BL/6^C57, demonstrating that the mtDNA haplotype can alter the aging response in joints, but without OA alterations.

In line with our results, very recently, Geurts’ group found that the accumulation of mtDNA mutations in a DNA mutator mouse model conferred a predisposition to elevated subchondral bone turnover and hypertrophy in calcified cartilage, but not to the accelerated development of osteoarthritis [26].

In this study, we first analyzed the influence of aging on the joint cartilage of both strains. To achieve this, we evaluated histopathological changes in the cartilage by a modified Mankin scoring system [16]. Three criteria, namely, structure, cellularity, and matrix staining of the cartilage, were analyzed and scored. In terms of the total Mankin scores, there were significant increases with the aging process among mice at 25, 75, and 90 weeks of age in BL/6^C57 and conplastic (BL/6^NZB) strains (Figure 1). Specifically, these differences at 25 and 75 weeks were significant for all three of the criteria analyzed. However, at 90 weeks of age, we observed a significant difference only in terms of the structure (data not shown).

Next, we continued comparing the Mankin scores of BL/6^C57 and conplastic (BL/6^NZB) mice. Interestingly, we found that conplastic (BL/6^NZB) mice developed fewer signs of aging with a consequently lower Mankin score than the original strain BL/6^C57 (Figure 1). We observed significant differences in all age groups analyzed. Specifically, the medial compartment and the medial femoral condyle were the joint regions where we found significant differences between conplastic (BL/6^NZB) mice and the original strain BL/6^C57 (Table 1), in accordance with the development of aging damage in human knee. Notably, we observed localized fibrillation and decreased chondrocyte density, but we did not find complete loss of cartilage typical of the OA process, which we observed in a model of surgically induced OA [27]. These findings on Mankin score levels reveal that mtDNA manipulation can attenuate the aging-related changes that affect the cartilage at joints, reducing their potential contribution to the development of OA. Besides, we analyzed the presence of histopathological changes in synovial tissue and subchondral bone (Supplementary Figure 2). Although we previously detected that conplastic (BL/6^NZB) mice under a surgically-induced model of OA presented lower damage in these tissues than observed in an original strain BL/6^C57 [27], in this study we failed to observed any significant difference between both strains in term of pathological changes in these articular tissues. It could be due to the fact that age-related spontaneous model of OA performed in this study is less aggressive and generally causes a lower articular damage. Nevertheless, conplastic (BL/6^NZB) mice showed an attenuation in age-associated reduction of growth plate of tibial epiphysis in comparison with detected in BL/6^C57 mice (Supplementary Figure 2).

Autophagy declines with aging and during OA [2]. Here, we measured two of the most important factors that participate in the autophagy process: LC3 and Beclin-1. The cartilage expression of LC3 and Beclin-1 was significantly increased in joints from conplastic (BL/6^NZB) mice at 75 (Supplementary Figures 3) and 90 weeks of age (Figure 2) compared with those of BL/6^C57 mice at the same ages. No significant differences were found in mice at 25 weeks of age (Supplementary Figures 3). Thus, our results confirmed that there were no basal differences between the two strains and that these differences appeared only in old age. Moreover, in our previous study using a model of surgically induced OA, we observed that conplastic (BL/6^NZB) mice also presented more cartilage expression of LC3 and Beclin-1 than the original strain [27]. In order to confirm that increment of LC3 and Beclin-1 proteins reflected an augmentation of autophagy flux, we analyzed p62 levels, a classical receptor and indicator of autophagy activity as P62 accumulates when autophagy is inhibited, and decreased levels can be observed when autophagy is elicited [2]. Interestingly, BL/6^NZB mice showed lower levels of P62 than BL/6^C57 mice, suggesting its higher degradation in the autolysosome and thus increased activation of autophagy flux in the conplastic mice. These findings imply that mtDNA variability affects the autophagic pathway in both aging and OA, suggesting a potential mechanism mediating the effects of mtDNA variants in the process of joint damage.

To study the senescence activity, we measured the cartilage expression of MMP13, β-Galactosidase, and p16, markers involved in the senescence process, and Ki67, an indicator of proliferation activity. Increased production of MMPs, in particular MMP13 and MMP3, was detected in cartilage with aging [28], while β-Galactosidase and p16 are widely used markers for cellular senescence [29–31]. In mice at 90 weeks of age, we found less cartilage expression of MMP13, β-Galactosidase, and p16 in joints from conplastic (BL/6^NZB) mice than in the original strain, whereas proliferation activity measured as Ki67 expression was increased in BL/6^NZB strain (Figure 3). We did not detect significant differences in mice at 25 and 75 weeks of age (Supplementary Figure 4). These results are in accordance with our data on the surgically induced OA in the conplastic (BL/6^NZB) mice and the original strain, where we did not observe differences between sham groups at 25 weeks of age [27]. Thus, at the basal time (25 weeks), there were no differences between the strains and only when aging progressed did we observe how different mtDNA variants influence the expression of senescence markers predisposing the mice to a worse disease outcome.

In chondrocytes, age-related mitochondrial dysfunction together with reduced activity of mitochondrial superoxide dismutase (SOD2) results in an increase in mitochondrially derived ROS with consequent DNA damage. Moreover, the aging-associated increase of MMP13 production in joint also contributes to the increase of ROS production [32]. To analyze oxidative stress, we measured the cartilage expression of 8-oxo-dG, an excellent biomarker, to determine the extent of oxidative damage. We found significantly less cartilage expression of 8-oxo-dG in joints from conplastic (BL/6^NZB) mice at 90 weeks of age compared with that in BL/6^C57 mice (Figure 4). Interestingly, we also found significant differences in the marker 8-oxo-dG in mice at 25 and 75 weeks of age (Supplementary Figure 5). Notably, in our recent study where we analyzed the effect of surgically induced OA in conplastic (BL/6^NZB) mice compared with that in the original strain BL/6^C57, we also observed a significant basal (sham group) difference in 8-oxo-dG between the two strains at 25 weeks of age. Thus, this marker of oxidative stress was less expressed in cartilage from sham conplastic (BL/6^NZB) mice than in the sham original strain [27]. These findings suggest that mitochondrially derived oxidative stress could be the first potential mechanism for initiating the aging process, characterizing the different responses between the two strains, and contributing to the development of OA. Aging-associated decreases in autophagy and the accumulation of ROS and MMPs induce chondrocyte apoptosis with consequent cartilage damage [2]. Here, we measured the cartilage expression of cleaved-caspase 3 in both studied strains. BL/6^C57 mice at 90 weeks of age showed more CC3 expression in cartilage than conplastic (BL/6^NZB) mice (Figure 4). We did not find significant differences between the strains in mice at 25 and 75 weeks of age (Supplementary Figure 5).

Notably, we also observed differences in the autophagic and apoptotic responses, at the joint level, between conplastic (BL/6^NZB) mice and the BL/6^C57 strain, in a model after the surgical induction of OA. Thus, these mechanisms seem to mediate the effects of the mtDNA variants and be key factors in both processes. These findings further shed light on the pivotal role of harboring specific mtDNA haplogroup and cluster such as J or JT [12, 23] in the OA pathogenesis; mitochondrial polymorphisms that have also been associated, like conplastic (BL/6^NZB) mice [15], to higher longevity and lower mtDNA damage and production of age-related mitochondrial ROS/RNS. Likewise, they could represent a potential novel approach to target aging-related changes in the joints in order to prevent or slow the development and progression of OA.

Nonetheless, the study has some limitations that must be highlighted. We here employed conplastic mice with NZB/OlaHsd mitochondrial genome and C57BL/6JOlaHsd nuclear DNA in order to evaluate the role of mitochondrial variants in OA development. Reciprocal conplastic animals, that is, animals with the NZB/OlaHsd nuclear genome and C57BL/6JOlaHsd mitochondrial DNA were not assayed as having lesser therapeutic significance, although it would be interesting to test with the aim of further confirming our current findings. In addition, activation of pathological pathways in the cartilage was evaluated by histological and immunohistochemical analysis by different markers for each process, however future studies should be warranted to reinforce our findings.

In conclusion, our findings demonstrated that, in response to aging, conplastic (BL/6^NZB) mice showed reduced joint deterioration compared with the original strain BL/6^C57. Moreover, we found that mtDNA variants differentially modulate the aging process at the joint level through modulating autophagy, senescence, oxidative stress, and apoptosis (Figure 5). Taken together, our results support the hypothesis that the mtDNA background plays a role in the process of joint damage, confirming that mtDNA variability is a prognostic factor for OA in aging and suggesting that modulation of mitochondrial oxidative phosphorylation (OXPHOS) is a novel therapeutic target in OA associated with aging.

Materials and Methods

Supplementary Materials

Supplementary Figures