Relationships between human vitality and mitochondrial respiratory parameters, reactive oxygen species production and dNTP levels in peripheral blood mononuclear cells

Introduction

The Medical Outcomes Study Short Form 36 (SF-36) vitality scale used in this study is a frequently cited and validated subscale or component of the multi-component SF-36 fatigue scale. The SF-36 vitality scale focuses on energy level to capture differences in subjective well-being and disease burden [1]. SF-36 vitality is strongly associated with other components of the SF-36 fatigue scale, such as SF-36 physical functioning, mental health, bodily pain scores, and decreased sleep quality (all P < 0.001) [2, 3]. Of the eight SF-36 scales, the SF-36 vitality scale is the best at discriminating between levels of health, even in the healthiest subgroups [2]. In addition, SF-36 vitality shows significant association with all three components of another fatigue scale, the Fatigue Symptom Inventory (FSI) [4]. Therefore, SF-36 vitality is often used as a surrogate for fatigue. Fatigue/low vitality is a common complaint among older adults and one of the most frequent reasons for physician visits in general practice [1, 2, 4, 5]. Fatigue/low vitality is often a symptom of underlying psychiatric or medical illness and is increasingly seen as an early indicator of frailty, caused by increased vulnerability in multiple biologic and physiologic systems [5]. Previous studies find that the SF-36 vitality level is associated with several clinical conditions including anemia, congestive heart failure, chronic obstructive pulmonary disease, chronic fatigue syndrome, mortality risk, and with negative outcomes including inability to work, job loss, and hospitalization [1, 6]. It is likely that several cellular factors contribute to development of low vitality, and mitochondrial dysfunction is an obvious culprit due to its function in ATP production. Surprisingly, there have been no studies to assess a potential role of defects in mitochondrial oxidative phosphorylation, or other mitochondrial-related defects, in human vitality. Towards this end, we measured several mitochondrial-related cellular parameters in peripheral blood mononuclear cells (PBMCs) to assess potential associations of mitochondrial activities with vitality score. Such analysis may give insight into mechanisms underlying low vitality or lead to biomarkers for diseases that have low vitality as a core symptom. PBMCs are commonly used in peripheral biomarker discovery research since they are easily isolated from the blood. In addition, accumulating evidence suggests that alterations in markers in peripheral serum or cells are identifiable footprints that reflect dysfunctions occurring elsewhere, such as in the brain [7]. For example, changes in psychiatric disorders are not only found in the brain, but also in peripheral markers, such as cytokines [8, 9]. Moreover, PBMCs have been shown to be a suitable cell model for mitochondrial dysfunction and oxidative stress in biomarker discovery for Alzheimer's disease, fibromyalgia and chronic fatigue syndrome [10-12].

Since mitochondria produce the majority of cellular energy (in the form of ATP), and mitochondrial dysfunction can result in metabolic and degenerative diseases [13], it is possible that mitochondrial dysfunction could contribute to fatigue, low vitality and the decline in functional status with aging. Cells generate ATP and biosynthetic precursors through a combination of oxidative and glycolytic metabolism. ATP production capacity can be measured conveniently by the Seahorse Bioscience's extracellular flux analyzer through detection of the oxygen consumption rate (OCR), to quantify mitochondrial respiration, and the extracellular acidification rate (ECAR), an indicator of glycolysis. Bioenergetics analysis suggests that measurement of mitochondrial respiratory fluxes give more information about the ability to make ATP than do measurements of membrane potential or ATP levels [14-16]. In this report, our interest lies with mitochondrial respiration capacity and plasticity under mitochondrial respiration stress or changing ATP demands, not with the cellular level of ATP at time of blood withdrawal. Therefore, our focus in this respect was to look for correlations of vitality score with mitochondrial respiration and glycolysis parameters, by using well-defined modulators of mitochondrial respiration (oligomycin, FCCP and antimycin A; see below and Figure 1).

We also measured cellular reactive oxygen species (ROS) production, in light of their relationship with mitochondrial function. Mitochondria are a significant source of ROS, however, there are also significant non-mitochondrial sources of endogenous ROS, including endoplasmic reticulum and peroxisomes [17]. Oxidative stress reflects the balance between the production rate of ROS and the removal rate of ROS by the antioxidant system. If oxidative stress is increased, it can damage DNA, RNA, proteins, and lipids and can lead to cell death and senescence and can have deleterious effects on all major organ systems [18, 19]. ROS are associated with the aging process and many age-associated diseases. ROS are also important signaling molecules, having crucial roles in normal physiological functioning. ROS-related diseases can stem from lack of ROS (such as certain autoimmune disorders) or surplus of ROS (such as cardiovascular and neurodegenerative diseases) [20].

In addition, we examined cellular deoxyribonucleotides (dNTPs) levels. dNTPs are the precursors used by polymerases for replication and repair of nuclear and mitochondrial DNA. The accuracy of these events requires sufficient amounts and correct balance of the four dNTPs. Imbalance of dNTP pools leads to genomic instability and is harmful to the cell and can lead to disease [21, 22]. In fact, recent studies have led to the hypothesis that there is an interconnected relationship between dNTP metabolism, genome stability, aging and cancer [23]. Several enzymes involved in biosynthesis of dNTPs are located in the mitochondria membrane. Studies suggest that mitochondrial dysfunctions affect these metabolic pathways and limit the available levels of cytosolic dNTPs [24]. Our rational for measuring dNTP ratios in this study is this connection to mitochondrial dysfunction, and that, unlike OCRs, ECARs and ROS, dNTPs can be measured on frozen samples and thus would serve well as biomarkers of fatiguing illnesses.

We performed association analysis between participant vitality scores and several OCR and ECAR parameters, as well as cellular ROS production and cellular dNTP levels. The objective of our study was to provide association-based evidence for or against underlying molecular defects that contribute to subjective feelings of low vitality, as well as search for potential biomarkers towards ailments that have low vitality as a core symptom.

Results

Discussion

Our study is the first to examine mitochondrial respiration in PBMCs with respect to vitality. Our bioenergetics analysis showed that none of the OCR parameters was associated with vitality score, indicating that the capacity of mitochondrial ATP production may not be related to vitality. In addition, there was no association of vitality score with basal ECAR or with the oligomycin-induced ECAR shift (glycolytic reserve), suggesting that alternative ATP generation by glycolysis (as needed when mitochondrial ATP generation is reduced) is also not linked to vitality. Although no studies before us have investigated potential links of mitochondrial dysfunction to vitality, several studies have reported mitochondrial dysfunctions in fatiguing illness [27-30], by examining levels of electron transport chain components and by estimating mitochondrial oxidative phosphorylation by way of ATP profiles. However, two recent studies in which the activities of the electron transport chain components were measured (in PBMCs in one of the studies) concluded that chronic fatigue syndrome patients have normal oxidative phosphorylation capacity [31, 32].

Our study is the first to report an association between ROS production in PBMCs and SF-36 vitality. This result is consistent with findings from illnesses that have low vitality and fatigue as core symptoms. Reports indicate that chronic fatigue syndrome and fibromyalgia are accompanied by higher oxidized low density lipoproteins and elevated protein carbonyl levels in the blood [10, 33-36]. In addition, elevated levels of 8-hydroxy-deoxyguanosine, a marker of oxidative damage to DNA, had been found in the urine of chronic fatigue patients [37]. These findings suggest that increased ROS levels may be involved in the development of these fatiguing disorders [38]. At moderate concentrations, ROS participate in the regulation of many cellular processes, including differentiation, proliferation, apoptosis, and cytoskeletal regulation [20, 39-44]. ROS initiates and regulates these pathways by directly interacting with critical signaling molecules, such as ATM, PI3 kinase, iron regulatory protein and redox factor 1 (for complete current review of this so called “oxidative interface” pathway activation process see Ray et al. [45]). Many of the ROS-mediated responses actually protect the cell against oxidative stress and re-establish redox homeostasis and regulate mitochondrial energy production [43]. We speculate that our reported association of ROS with vitality score could involve signaling pathway imbalance. Along these lines, studies in humans have shown that oxidative changes in plasma thiol/disulphide redox status are correlated with aging-related pathophysiological processes, such as immune response, thyroid function, and cognitive function [20, 40, 46]. However, many correlation studies, including our current study, cannot imply cause and effect.

The dTTP level was inversely associated with vitality score, yet the dCTP level showed a significant association with vitality score in the opposite direction. Therefore, our data suggests that low vitality may be linked to both a lower dCTP and higher dTTP level. Thus, going forward, a potentially useful predictor of low vitality/fatigue or related illnesses could be low dCTP/dTTP ratio. However this would have to be verified in future studies. We did not specifically measure this ratio in this study because, due to lack of sufficient material, dCTP measurements were done on a different set of participants (N of 34; see Table 1) than for the other three dNTPs (N of 79; see Table 1); we had to set aside 34 frozen samples that had not been measured yet, specifically for dCTP-only measurements to allow us to obtain data from all four dNTPs. One could also look for alterations in the level or activity of deoxycytidylate (dCMP) deaminase. This enzyme catalyzes the conversion of dCMP to dUTP in the de novo synthesis of thymidine nucleotides and, via this activity, is believed to play a key role in regulating the dCTP/dTTP balance in the cell [22]. The physiological consequences of dCMP deaminase defects or imbalance in dCTP/dTTP are not well understood. However there is evidence for a role of dCMP deaminase in genomic integrity and tumorigenicity [23]. Speculation based on the above associations would be that dCMP deaminase is overactive in low vitality (high fatigue) patients (since this enzyme catalyzes a step in dTTP synthesis and the dTTP levels we measured in this study were inversely associated with vitality). Other important enzymes important in dNTP balance that could be examined, with respect to vitality, are dihydroorotate dehydrogenase (DHODH) and ribonucleotide reductase (RNR) [22, 24]. DHODH is important in de novo synthesis of pyrimidines and is located in the inner mitochondrial membrane and its activity is influenced by the activity of the electron transport chain. RNR catalyzes the formation of deoxyribonucleotides from ribonucleotides and can be inhibited by nitric oxide that is created by mitochondrial nitric oxide synthase (mtNOS) located in the inner mitochondrial membrane. The activity of mtNOS is linked to the mitochondrial membrane potential. We note that our measurement of total cellular dNTP levels and total ROS levels are not designed to differentiate the source of these molecules (mitochondrial versus cytosolic); this would require more elaborate work (involving isolation of mitochondria), not amenable to a cohort study. However, typically mitochondrial dNTP pools represent at most 10% of the total cellular pools [21]. This suggests that the dNTP imbalance that we observe, as a function of vitality, could stem from a defect in cytosolic, as opposed to mitochondrial, dNTP pathways. From the above data, we cannot say whether the ROS and dNTP associations with vitality have any mitochondrial-defect link.

We found no association of the mitochondrial OCR parameters with ROS production or dNTP levels (Table 3). Data from previous studies suggest that mitochondrial dysfunction may be linked to oxidant levels [47-50] and dNTP imbalance [24, 51]. However, those studies did not measure mitochondrial oxygen flux respiratory parameters, but rather focused on other measures of mitochondrial dysfunction, that are linked to respiration, such as oxidized protein carbonyls [52], peroxidation of membrane phospholipids [47-49], and mitochondrial DNA mutations [51]. An exception is a recent report by Furda et al. [53] in which treatment of mouse embryonic fibroblasts with H₂O₂ resulted in a reduction in the OCR parameters ATP turnover and reserve capacity. However, sustained H₂O₂ treatment was necessary to see this effect. To get a more complete picture of bioenergetics, with respect to vitality, further activities could be examined, such as electron delivery to the respiratory chain, substrate transport into the cell, and ATP export to the cytoplasm. We did not measure ATP concentration as our focus was on mitochondrial ATP generation dysfunction as measured via oxygen flux respiratory parameters. However, Myhill et al. [29] recently reported a strong correlation between an “ATP profile” test measure (which includes ATP concentration) in neutrophils and the severity of chronic fatigue syndrome. However, ATP-related experiments would also have to assess whether observed changes in cellular ATP reflects dysfunctional mitochondria or an independent change in the metabolism of adenine nucleotides [15].

It is known that several physiological/bioenergetic conditions regulate the OCR levels, and the OCR parameters are interrelated: for example, basal OCR level is controlled to a large degree by ATP turnover (as evident by our strong correlation between these two parameters; Table 2), and partly by proton leak (both of which consume the proton-motive force generated by the substrate oxidation) [15]. Therefore, differences in ATP demand between individuals (for example, due to differences in protein synthesis or plasma membrane ion cycling [15, 26, 54]) strongly influence basal OCR levels. OCRs can also be altered by difference in levels of pyruvate or glucose; this was controlled by careful preparation of incubation medium containing defined levels of pyruvate and glucose. The maximum OCR reports the maximum activity of electron transport that is achievable by the cells under the assay conditions (artificial decoupling of electron transport from ATP synthesis by FCCP). A decrease in maximum OCR is a strong indicator of mitochondrial dysfunction in level or efficiency of electron transport (hence, our strong correlation of maximum capacity with basal OCR, ATP turnover and reserve capacity). Reserve capacity (also called spare-respiratory capacity) is an important diagnostic of the ability of the cell to response to changes in ATP demand. As is obvious in Figure 1A, it represents the maximum OCR above the basal OCR and so indicates how close to its bioenergetic limit the cell is operating. It was correlated with basal OCR, ATP turnover, and maximum capacity, which indicated that all these parameters are linked to the cells ability to respond to changes in ATP demand. The glycolysis parameter, basal ECAR (Figure 1B), did not correlate with any mitochondrial respiratory parameters; however it did correlate with the other measured glycolysis (ECAR) parameter, glycolytic reserve. The glycolytic reserve correlated with all the OCR parameters except for proton leak. Previous studies have validated, in cultured cells, the concept of switching to glycolysis for generation of ATP when the mitochondrial ATP generation is dysfunctional [16, 55]; our OCR-glycolytic reserve correlations validates this concept in PBMCs and thus reiterate that in a population a person with high mitochondrial consumption of oxygen (in their PBMCs) would also tend to have a high ability to switch to glycolytic generation of ATP when the mitochondrial ATP generation is dysfunctional. A thorough discussion of the links between the mitochondrial and glycolytic bioenergetic measures can be found in other articles [15, 16, 56]. Interestingly the non-mitochondrial OCR showed significant associations with several mitochondrial OCR parameters. Future in-vitro experiments will be needed to extent this observation, by examining specific non-mitochondrial oxygen consuming reactions in PBMC or other cells types that may alter with induced changes in respiratory parameters. These include cyclooxygenases, lipoxygenases and potentially NADPH oxidases and nitric oxide synthases, all of which generate either reactive lipids or oxidants such as hydrogen peroxide, which can directly modify mitochondrial function. In should be noted that the term non-mitochondrial OCR may not be completely accurate, as there may be a component of this OCR parameter that stems from mitochondrial sources other than oxidative phosphorylation; therefore “non-oxidative phosphorylation OCR” may be a more accurate term, however, “non-mitochondrial” is almost exclusively used in the literature in this context [15, 16, 57], and therefore we use this term. Moreover, in immune cells, such as in PBMCs, the predominant antimycin A-insensitive OCR is due to oxygen consuming immune cell activities such as NADPH oxidases [15]. Due to the above relationships of the respiratory parameters one would expect that if one respiratory parameter was associated with vitality, then other respiratory parameters would also be associated with vitality. We in fact saw that none of the OCR or ECAR parameters was associated with vitality score. We conclude that mitochondria respiratory dysfunction at the level of the above respiratory parameters are not factors in human vitality.

Our data suggests that low vitality is not linked to defects in mitochondrial respiration or glycolysis, but may have other underlying molecular defects that impact ROS levels and dNTP balance. We speculate that high ROS production and dNTP pool imbalance may interact with development of low vitality. These measures could be pursued as components of early biomarker strategies for fatiguing illness. The ROS association data supports the practice of antioxidant intake as a method to lift or maintain energy levels in healthy individuals and suggests that antioxidant intake could be examined as part of an intervention strategy in patients that need to maintain high vitality for quick recovery.

Methods

Supplementary Materials

Supplemental Data