Keywords: aerobic exercise, caloric restriction, chronic kidney disease, mitochondrial DNA
Abstract
Circulating cell-free mitochondrial DNA (ccf-mtDNA) may induce systemic inflammation, a common condition in chronic kidney disease (CKD), by acting as a damage-associated molecular pattern. We hypothesized that in patients with moderate to severe CKD, aerobic exercise would reduce ccf-mtDNA levels. We performed a post hoc analysis of a multicenter randomized trial (NCT01150851) measuring plasma concentrations of ccf-mtDNA at baseline and 2 and 4 mo after aerobic exercise and caloric restriction. A total of 99 participants had baseline ccf-mtDNA, and 92 participants completed the study. The median age of the participants was 57 yr, 44% were female and 55% were male, 23% had diabetes, and 92% had hypertension. After adjusting for demographics, blood pressure, body mass index, diabetes, and estimated glomerular filtration rate, median ccf-mtDNA concentrations at baseline, 2 mo, and 4 mo were 3.62, 3.08, and 2.78 pM for the usual activity group and 2.01, 2.20, and 2.67 pM for the aerobic exercise group, respectively. A 16.1% greater increase per month in ccf-mtDNA was seen in aerobic exercise versus usual activity (P = 0.024), which was more pronounced with the combination of aerobic exercise and caloric restriction (29.5% greater increase per month). After 4 mo of intervention, ccf-mtDNA increased in the aerobic exercise group by 81.6% (95% confidence interval: 8.2−204.8, P = 0.024) compared with the usual activity group and was more marked in the aerobic exercise and caloric restriction group (181.7% increase, 95% confidence interval: 41.1−462.2, P = 0.003). There was no statistically significant correlation between markers of oxidative stress and inflammation with ccf-mtDNA. Our data indicate that aerobic exercise increased ccf-mtDNA levels in patients with moderate to severe CKD.
NEW & NOTEWORTHY The effects of prolonged exercise on circulating cell-free mitochondrial DNA (ccf-mtDNA) have not been explored in patients with chronic kidney disease (CKD). We showed that 4-mo aerobic exercise is associated with an increase in plasma ccf-mtDNA levels in patients with stages 3 or 4 CKD. These changes were not associated with markers of systemic inflammation. Future studies should determine the mechanisms by which healthy lifestyle interventions influence biomarkers of inflammation and oxidative stress in patients with CKD.
INTRODUCTION
Complications in patients with chronic kidney disease (CKD) are partially due to risk factors like obesity, oxidative stress, and inflammation (1–3). Exercise and caloric restriction can improve those factors by different mechanisms, including enhancing mitochondrial function (4–6). The role of the mitochondria is essential beyond aerobic ATP generation; for example, circulating cell-free mitochondrial (mt)DNA (ccf-mtDNA) acts as a damage-associated molecular pattern and mediates the innate immune system response to promote inflammation (7, 8).
In patients with certain malignancies, depression, sepsis, and during acute intense exercise, ccf-mtDNA is found to be elevated, and higher ccf-mtDNA levels are associated with worse clinical outcomes (9–14). A recent study showed that elevated ccf-mtDNA correlates with poor clinical status and quality of life in patients on maintenance hemodialysis, likely due to the effects of ccf-mtDNA on inflammation (15). The same study reported an association between higher levels of high-sensitivity C-reactive protein, TNF-α, osteoprotegerin, and chemokine (C-X-C motif) ligand 16 with higher ccf-mtDNA levels (15).
Aerobic exercise decreases systemic inflammation, but the effects on ccf-mtDNA levels are unclear. Shockett et al. (16) found that healthy volunteers after moderate aerobic exercise had lower ccf-mtDNA levels, but other studies reported an elevation in ccf-mtDNA levels after single or repetitive bouts of intense exercise (13, 14). Currently, there are limited studies in patients with CKD evaluating the effects of aerobic exercise on ccf-mtDNA.
In this post hoc analysis of a pilot randomized clinical trial in patients with CKD stages 3 or 4 (NCT01150851), we evaluated the effects of aerobic exercise on plasma ccf-mtDNA. We hypothesized that in patients with moderate to severe CKD, aerobic exercise reduces plasma levels of ccf-mtDNA. We also performed a subanalysis to assess whether the association between ccf-mtDNA and aerobic exercise differed by caloric restriction.
MATERIALS AND METHODS
Study Design
This study was a post hoc analysis of a pilot randomized 2 × 2 factorial design (NCT01150851) (17). The Consolidated Standards of Reporting Trials checklist was used in this report (18). The Institutional Review Boards of all participating sites approved this study. All participants gave written informed consent before study enrollment. This study was unblinded because of the character of the interventions.
We obtained baseline enrollment data and blood work ∼2 wk before the initiation of the 4-mo intervention phase. Patients were assigned to one of the following four groups: 1) aerobic exercise and caloric restriction, 2) aerobic exercise and usual diet, 3) caloric restriction and usual physical activity, and 4) usual physical activity level and usual diet.
Patient Population
We recruited participants from the Vanderbilt University Medical Center, Veterans Affair Tennessee Valley Healthcare System, Providence Medical Research Center/Providence Health Care, University of Washington, and Renal and Transplant Associates of New England. Recruitment started in October 2010, and the study concluded in February 2014. Inclusion criteria consisted of CKD stages 3 and 4 estimated by the creatinine-based modification of diet in renal disease equation with an estimated glomerular filtration rate (eGFR) of 15–60 mL/min/1.73 m2, age of 18–75 yr old, body mass index of ≥25 kg/m2, life expectancy of ≥1 yr, and the capacity to understand and give informed consent for participation in the study. Exclusion criteria covered any acute inflammatory condition (including chronic infection requiring treatment and collagen vascular disease, including active gout); pregnancy; taking high-dose antioxidants (vitamin E or C); chronic use of anti-inflammatory medication, except for low-dose (<10 mg/day) prednisone and aspirin (<100 mg/day); significant cardiac or vascular disease (symptomatic disease or cardiovascular disease event, including congestive heart failure, within 6 mo); significant occlusive atherosclerotic disease or ischemic disease (on noninvasive or invasive diagnostic procedures); significant physical immobility or disabilities (joint replacement or muscular disorders); type 1 diabetes mellitus or type 2 diabetes mellitus requiring insulin therapy; and history of poor adherence to a medical regimen. Patients with a diagnosis of atrial fibrillation or a pacemaker were allowed in the study. We conducted prescreening because direct supervision of exercise was a part of the protocol. We assigned study visits at baseline (before the initiation of interventions) and then at 1, 2, and 4 mo (end of study). We consented to 122 participants; 111 participants were randomized, 104 participants started interventions, and 92 participants completed the study (Fig. 1).
Figure 1.
Consolidated standards of reporting trials flow diagram. ccf-mtDNA, circulating cell-free mitochondrial DNA.
Randomization
We used a permuted block randomization strategy in a 1:1 ratio to allocate subjects to the study arms (Fig. 1). Groups were stratified according to the site (Vanderbilt University Medical Center, Veterans Affair Tennessee Valley Healthcare System, University of Washington, Providence Medical Research Center/Providence Health Care, and Springfield) as well as by diabetes mellitus status.
Standard of Care Dietary Counseling
A registered dietitian with a background in renal nutrition asked all study participants to avoid high-calorie meals rich in processed, quickly digestible, fast absorbable foods and drinks as a part of a CKD diet. All study participants received dietary support that included high numbers of fresh, unprocessed plant intake, with moderate lean protein and fats levels, such as ω-3 and monounsaturated fats.
Caloric Restriction Intervention
A registered dietitian with a background in renal nutrition performed dietary recalls and questionnaires to establish the dietary intake for participants. Data were analyzed using Nutrition Data System for Research software (University of Minnesota, Minneapolis, MN). For subjects randomized to the caloric restriction group, decreasing daily caloric intake by 10–15% (considering an average diet of 3,000 kcal/day) would result in an ∼300–500 kcal decrease per day, leading to 1 lb (restriction of 3,500 kcal/wk) of weight loss per week. We collected dietary recalls before starting the dietary regimen to screen for subjects with low daily caloric intake at baseline (≤2,000 kcal/day) because further dietary restriction in such participants can be potentially harmful.
Aerobic Exercise Intervention
We scheduled subjects randomized to the low-impact aerobic exercise group to complete physical activity for a maximum of 30–45 min for 3 times/wk for 4 mo in duration. Subjects divided up the aerobic exercise time using a treadmill, an elliptical cross-trainer, a NuStep cross-trainer, and a stationary recumbent bicycle to offer variety in the exercise intervention. Since the net caloric expenditure per mile of walking at 3.5 mph is 0.77 kcal/kg per mile during moderate-paced walking, the goal was 200–300 kcal energy expenditure per training session during each exercise session. We acknowledged that previously inactive participants would have a lower level of cardiorespiratory fitness at baseline. Therefore, we customized the exercise prescription considering each subject’s baseline fitness status. The exercise intensity was individualized according to the initial physical fitness assessment and progressively increased throughout the study by certified exercise specialists and clinical exercise physiologists experienced in cardiopulmonary rehabilitation and exercise in participants with chronic diseases. We calculated the oxygen consumption (V̇o2) peak at the beginning of the study. We applied this baseline value to monitor cardiorespiratory fitness improvements over 4 mo with a goal to reach an exercise intensity that represented a 60–80% V̇o2 peak (V̇o2peak). We documented the degree of perceived exertion and metabolic equivalents during the exercise sessions to track exercise progression and tolerance throughout the study.
Compliance and Adverse Events
We monitored the protocol compliance and adverse events at the participating sites by holding monthly conference calls. Agenda details included enrollment, protocol adherence, and adverse events. Study staff at each site registered case report forms into a REDCap database. We reviewed sites for missing or irregular entries and resolved all queries.
We evaluated diet compliance every 2 wk during the collection and analysis of dietary recalls. All participants finished ≥50% of diet recalls in the caloric restriction group, whereas almost one-half of the subjects (42%) achieved all recalls. We monitored and documented exercise compliance and adverse events weekly.
A Data Safety Monitoring Board oversaw the study’s safety profile, and no interim efficacy analyses were planned or conducted.
Outcomes and Measurements
For all groups, we conducted anthropometric measurements (height, weight, and body mass index), blood pressure, medication revision, and standard of care dietary counseling. Data on dual-energy X-ray absorptiometry and bioimpedance analysis have been previously published (17). To monitor changes in cardiorespiratory fitness, we measured V̇o2peak (MedGraphics Ultima, Medical Graphics, St. Paul, MN) at baseline and 4 mo. We used dietary recalls and questionnaires to ascertain participants’ dietary intake (analyzed using Nutrition Data System for Research software, University of Minnesota, Minneapolis, MN).
We obtained routine chemistries from the health records of the participants. We drew blood into Vacutainer (Becton Dickinson, Franklin Lakes, NJ) tubes containing EDTA for plasma separation. Samples were transported on ice and immediately centrifuged at 20°C at 3,000 rpm for 15 min. Supernatants were stored in aliquots at −80°C until further use. We a priori chose to measure kidney function by eGFR by the cystatin C equation (eGFRcysc) to avoid confounding by muscle mass on estimated treatment effects encountered using creatinine-based measures of eGFR. Cystatin C concentrations were measured using a clinical chemistry analyzer (DxC600, Beckman Coulter). We standardized the DxC measurements by reconstituting the cystatin C reference material (ERM-DAY7/IFCC) per its certificate of analysis to yield a cystatin C concentration of 5.48 mg/L (uncertainty of 0.15 mg/L), and eGFRcysC was estimated using the Chronic Kidney Disease Epidemiology Collaboration equation (19).
Cytometric bead arrays (Becton Dickinson, San Jose, CA) were used to determine plasma interleukin (IL) levels. Plasma F2-isoprostane and isofuran concentrations were quantified to determine oxidative stress. The internal standard [2H4]-15-F2T-isoprostane was added to plasma, and sequential C-18 and silica solid-phase extraction purified the sample and then derivatized to pentafluorobenzyl ester, trimethylsilyl ether for gas chromatography/negative ion chemical ionization/mass spectrometry analysis.
We measured ccf-mtDNA at 1, 2, and 4 mo, as previously described (20). We quantified ccf-mtDNA by quantitative RT-PCR directly on patient plasma using plasma-stable DNA polymerase (Omni-KlenTaq-2 DNA Polymerase, DNA Polymerase Technology) and a PCR enhancer cocktail (PEC-2, DNA Polymerase Technology) to avoid inaccuracies associated with DNA purification methods. Primers specific for mitochondrially encoded NADH dehydrogenase 1 (ND1; forward 5′- CCCTAAAACCCGCCACATCT-3′ and reverse 5′- GAGCGATGGTGAGAGCTAAGGT-3′) were used for ccf-mtDNA. Standard curves with parallel quantitative RT-PCRs with oligonucleotide mimic known concentrations were used to determine final DNA concentrations. The ND1 gene and the set of primers are specific for mtDNA detection; no other product was detected by the melting curve analysis.
Statistical Analyses
Primary analyses were performed on an intent-to-treat basis, including all randomized participants with measured plasma ccf-mtDNA (n = 99). For this analysis, the “control” group was defined as the usual activity group, with a total of 48 participants. The usual activity group consisted of 27 subjects who did usual activity with caloric restriction and 21 subjects who did usual activity with a usual diet. The intervention group was defined as the aerobic exercise group and consisted of 51 participants, which included 23 subjects who did aerobic exercise with the usual diet group and 28 subjects who did aerobic exercise with caloric restriction. We tabulated baseline descriptive statistics, including demographics, medical history, clinical, laboratory, and lifestyle characteristics, according to the intervention arm. We reported the median and first and third quartile values of continuous variables and the prevalence within categorical variables.
Our primary analysis examined the effect of the aerobic exercise intervention on the percent change per month in plasma ccf-mtDNA concentrations. We used generalized estimating equations, accounting for within-participant correlations across time, to determine whether the relative change in plasma ccf-mtDNA differed according to exercise intervention. For all analyses, plasma ccf-mtDNA was log transformed. The crude model included a fixed effect term for exercise intervention, an indicator variable for the month, indicator variables for exercise intervention crossed with the month 4 indicator variable, and subject-specific random intercepts. A second model additionally included variables for age, race, and sex. A fully adjusted model additionally included baseline body mass index, diabetes status, baseline systolic blood pressure, and baseline eGFR. We secondarily stratified by diet randomization group (usual diet or caloric restriction) to assess whether the association differed by caloric restriction.
We implemented the same model fitting and hypothesis testing procedures for the intent-to-treat population with the compliant subpopulation.
We also conducted analyses examining the effect of ccf-mtDNA on the percent change per month in inflammation and oxidative stress levels. For all analyses, plasma ccf-mtDNA was log transformed. The slope of plasma ccf-mtDNA was used as the predictor and was calculated using baseline, month 2, and month 4 values. Outcomes included inflammation markers (IL-6, IL-8, IL-10, and TNF) and oxidative stress markers (F2-isoprostanes and isofurans). All outcomes were log transformed. We used generalized estimating equations, accounting for within-participant clustering across time, to determine whether the relative change in inflammation and oxidative stress differed with change in plasma ccf-mtDNA. The models included the slope for ccf-mtDNA, an indicator variable for month, the slope variable cross with the month indicator variable, and subject-specific random.
We performed statistical analyses with R (v. 3.3.0, R Foundation for Statistical Computing) and Stata v. 16.0 (21). We defined the nominal level of significance as P < 0.05 (two-sided).
RESULTS
Participant Characteristics
This post hoc analysis included only participants with a ccf-mtDNA measurement (51 in the aerobic exercise group and 48 in the usual activity group). In terms of the participants, their median age was 57 yr, 44 were women (44.4%) and 55 were men (55.6%), 91 had hypertension (91.9%), and 23 had diabetes (23.2%). Median baseline eGFR was 38.9 (28.6, 50.4) mL/min/1.73 m2. The overall demographics and clinical characteristics were well balanced across the intervention groups (Table 1).
Table 1.
Participant baseline characteristics according to randomization to exercise intervention
| Overall (n = 99) | Usual Activity (n = 48) | Aerobic Exercise (n = 51) | |
|---|---|---|---|
| Age, yr | 57 (49, 63) | 61 (50.5, 65) | 55 (48, 61) |
| Sex | |||
| Female | 44 (44.4) | 20 (41.7) | 24 (47.1) |
| Male | 55 (55.6) | 28 (58.3) | 27 (52.9) |
| Race | |||
| Black | 26 (26.3) | 14 (29.2) | 12 (23.5) |
| White | 68 (68.7) | 32 (66.7) | 36 (70.6) |
| Other | 5 (5.1) | 2 (4.2) | 3 (5.9) |
| Caloric restriction | 55 (55.6) | 27 (56.3) | 28 (54.9) |
| Body mass index, kg/m2 | 32.8 (28.7, 37.2) | 33.2 (28.8, 37.8) | 32.3 (28.5, 35.8) |
| Prevalent disease | |||
| Hypertension | 91 (91.9) | 44 (91.7) | 47 (92.2) |
| Diabetes | 23 (23.2) | 11 (22.9) | 12 (23.5) |
| Congestive heart failure | 3 (3.0) | 2 (4.2) | 1 (2.0) |
| Myocardial infarction | 2 (2.0) | 0 (0.0) | 2 (3.9) |
| Coronary artery disease | 5 (5.1) | 2 (4.2) | 3 (5.9) |
| Estimated glomerular filtration rate, mL/min/1.73 m2 | 38.9 (28.6, 50.4) | 36.9 (28.3, 50.4) | 40.2 (28.6, 51.3) |
| Systolic blood pressure, mmHg | 131 (115, 142) | 133 (118, 141.5) | 129 (114, 142) |
| Diastolic blood pressure, mmHg | 79 (72, 85) | 79 (72.5, 84) | 78 (70, 88) |
Values are presented as numbers of participants with percentages [n (%)] or medians (25th, 75th percentile).
Circulating Cell-Free mtDNA
At baseline, participants in the usual activity group had higher ccf-mtDNA levels than the aerobic exercise group (3.62 vs. 2.01 pM), respectively. After 2 and 4 mo of intervention, plasma ccf-mtDNA gradually decreased in the usual activity group but steadily increased in the aerobic exercise group (Table 2 and Fig. 2).
Table 2.
Plasma circulating cell-free mitochondrial DNA concentrations according to randomization to exercise intervention at all time points
| Overall (n = 99) | Usual Activity (n = 48) | Aerobic Exercise (n = 51) | |
|---|---|---|---|
| Baseline | 2.77 (1.23, 7.50) | 3.62 (1.85, 8.33) | 2.01 (1.08, 7.32) |
| Month 2 | 2.62 (1.18, 8.22) | 3.08 (1.19, 8.70) | 2.20 (1.09, 4.90) |
| Month 4 | 2.62 (1.18, 8.22) | 2.78 (0.79, 8.29) | 2.67 (1.10, 7.23) |
Values are presented as medians (25th, 75th percentile) of plasma circulating cell-free mitochondrial DNA (in pM).
Figure 2.
Boxplot of plasma circulating cell-free mitochondrial DNA (ccf-mtDNA) at baseline, month 2, and month 4 for the control and exercise groups.
Overall results comparing the physical activity intervention without considering a diet regimen showed a 16.1% [95% confidence interval (CI): 2.0−32.1, P = 0.024) higher increase per month of ccf-mtDNA levels in the aerobic exercise group relative to the usual activity group. After the substratification of physical activity by diet, there was a 29.5% (95% CI: 9.0−54.0, P = 0.003) higher increase per month in the group randomized to aerobic exercise with caloric restriction relative to the group randomized to aerobic exercise with usual diet (Table 3). All of the monthly changes in plasma ccf-mtDNA levels persisted after adjusting by age, race, systolic blood pressure, body mass index, diabetes, and eGFR.
Table 3.
Effect of exercise intervention on the percent change (95% CI) in plasma circulating cell-free mitochondrial DNA per month
| Percent Change (95% CI) | P Value | Percent Change (95% CI)* | P Value | Percent Change (95% CI)† | P Value | |
|---|---|---|---|---|---|---|
| Overall (n = 99) | ||||||
| Usual activity | Reference | Reference | Reference | |||
| Aerobic exercise | 15.8 (2.0, 31.4) | 0.024 | 15.7 (1.9, 31.3) | 0.024 | 16.1 (2.0, 32.1) | 0.024 |
| Usual diet (n = 44) | ||||||
| Usual activity | Reference | Reference | Reference | |||
| Aerobic exercise | 1.3 (−15.9, 22.2) | 0.888 | 1.4 (−15.8, 22.1) | 0.885 | 2.0 (−15.7, 23.3) | 0.842 |
| Caloric restriction (n = 55) | ||||||
| Usual activity | Reference | Reference | Reference | |||
| Aerobic exercise | 29.3 (9.2, 53.1) | 0.003 | 29.1 (9.0, 52.8) | 0.003 | 29.5 (9.0, 54.0) | 0.003 |
| P value for interaction: 0.057‡ |
*Adjusted for age, race, and sex.
†Adjusted for age, race, sex, systolic blood pressure, body mass index, diabetes, and estimated glomerular filtration rate.
‡P value for interaction comes from a model with a three-way interaction between aerobic exercise, caloric restriction, and time. CI, confidence interval; n, number of participants.
Overall results comparing the physical activity intervention without considering a diet regimen showed an 81.6% (95% CI: 8.2−204.8, P = 0.024) higher increase of ccf-mtDNA levels in the aerobic exercise group compared with the usual activity group after 4 mo of intervention. After the substratification of physical activity by diet, there was a 181.7% higher increase of ccf-mtDNA levels (95% CI: 41.1−462.2, P = 0.003) in the group of aerobic exercise with caloric restriction relative to controls after 4 mo of intervention (Table 4). Plasma ccf-mtDNA level changes at 4 mo persisted after adjusting by age, race, systolic blood pressure, body mass index, diabetes, and eGFR.
Table 4.
Effect of exercise intervention on the percent change (95% CI) in plasma circulating cell-free mitochondrial DNA over 4 mo
| Percent Change (95% CI) | P Value | Percent Change (95% CI)* | P Value | Percent Change (95% CI)† | P Value | |
|---|---|---|---|---|---|---|
| Overall (n = 99) | ||||||
| Usual activity | Reference | Reference | Reference | |||
| Aerobic exercise | 79.6 (8.2, 198.0) | 0.024 | 79.0 (8.0, 196.9) | 0.024 | 81.6 (8.2, 204.8) | 0.024 |
| Usual diet (n = 44) | ||||||
| Usual activity | Reference | Reference | Reference | |||
| Aerobic exercise | 5.5 (−50.1, 122.9) | 0.888 | 5.7 (−49.8, 122.5) | 0.885 | 8.0 (-49.5, 131.1) | 0.842 |
| Caloric restriction (n = 55) | ||||||
| Usual activity | Reference | Reference | Reference | |||
| Aerobic exercise | 179.7 (42.4, 449.5) | 0.003 | 177.7 (41.4, 445.2) | 0.003 | 181.7 (41.1, 462.2) | 0.003 |
*Adjusted for age, race, and sex.
†Adjusted for age, race, sex, systolic blood pressure, body mass index, diabetes, and estimated glomerular filtration rate. CI, confidence interval; n, number of participants.
Relationship Between ccf-mtDNA and Markers of Inflammation/Oxidative Stress
Overall, there was a correlation between ccf-mtDNA and IL-10 at month 4 of intervention (Table 5). There was no other correlation between ccf-mtDNA with inflammatory and oxidative stress markers in the aerobic exercise group or the usual activity group. A generalized estimating equation analysis evaluated the relative change in markers of inflammation with change in plasma ccf-mtDNA. In the usual activity group, there was a positive association between IL-10 and ccf-mtDNA changes per month during the course of the study (B = 1.001, 95% CI: 1.0−1.1, P = 0.05). There was no other statistically significant association after a generalized estimating equation model was used (Table 6).
Table 5.
Correlations between ccf-mtDNA at month 4 with markers of inflammation and oxidative stress at month 4 [Spearman’s ρ (P value)]
| IL-6 | IL-8 | IL-10 | Tnf-α | F2-Isoprostanes | Isofurans | |
|---|---|---|---|---|---|---|
| Overall | ||||||
| ccf-mtDNA | 0.18 (0.08) |
−0.06 (0.58) |
0.25 (0.02)* |
−0.02 (0.85) |
−0.03 (0.78) |
0.01 (0.95) |
| Aerobic exercise | ||||||
| ccf-mtDNA | 0.25 (0.10) |
−0.13 (0.42) |
0.26 (0.08) |
−0.08 (0.59) |
−0.13 (0.39) |
0.05 (0.77) |
| Usual activity | ||||||
| ccf-mtDNA | 0.15 (0.33) |
0.01 (0.93) |
0.27 (0.07) |
0.05 (0.75) |
0.09 (0.54) |
−0.07 (0.67) |
Correlations were calculated using Spearman’s rank correlation. ccf-mtDNA, circulating cell-free mitochondrial DNA; IL, interleukin; TNF, tumor necrosis factor. *P < 0.05.
Table 6.
Association between circulating cell-free mitochondrial DNA with markers of inflammation and oxidative stress per month
| Estimate (95% Confidence Interval) | P Value | |
|---|---|---|
| Overall | ||
| IL-6 | 0.998 (0.964, 1.034) | 0.924 |
| IL-8 | 1.001 (0.993, 1.008) | 0.860 |
| IL-10 | 1.024 (0.984, 1.065) | 0.243 |
| TNF-α | 0.999 (0.996, 1.003) | 0.682 |
| F2-isoprostanes | 0.998 (0.994, 1.003) | 0.491 |
| Isofurans | 1.002 (0.998, 1.006) | 0.333 |
| Aerobic exercise | ||
| IL-6 | 0.980 (0.934, 1.029) | 0.419 |
| IL-8 | 0.998 (0.991, 1.006) | 0.693 |
| IL-10 | 1.010 (0.958, 1.064) | 0.722 |
| TNF-α | 0.999 (0.994, 1.004) | 0.690 |
| F2-isoprostanes | 0.999 (0.994, 1.004) | 0.632 |
| Isofurans | 1.003 (0.998, 1.009) | 0.268 |
| Usual activity | ||
| IL-6 | 1.039 (0.985, 1.110) | 0.159 |
| IL-8 | 1.007 (0.991, 1.022) | 0.400 |
| IL-10 | 1.064 (1.000, 1.132) | 0.050 |
| TNF-α | 1.001 (0.996, 1.006) | 0.677 |
| F2-isoprostanes | 1.000 (0.993, 1.008) | 0.896 |
| Isofurans | 1.001 (0.995, 1.007) | 0.725 |
Note that all inflammation markers and oxidative stress markers were natural log transformed; circulating cell-free mitochondrial DNA was natural log transformed, and the slope was calculated using baseline, month 2, and month 4 values. IL, interleukin; TNF, tumor necrosis factor.
Adverse Events
The adverse events were as follows: four events of hypotension due to weight loss, one participant had an episode of rapid atrial fibrillation requiring hospitalization, one event of chest pain while exercising, joint pain while exercising, and painful Achilles. There were no deaths during this study. Additional details have been previously published (17).
DISCUSSION
In patients with moderate CKD, aerobic exercise performance for 4 mo increased ccf-mtDNA levels. The increase in ccf-mtDNA is more profoundly observed when aerobic exercise is combined with a 10−15% caloric restriction diet compared with aerobic exercise alone. We did not observe relevant associations between biomarkers of inflammation and oxidative stress and ccf-mtDNA at baseline or in response to study interventions in participants with moderate CKD.
In patients with CKD, there are limited data regarding the impact of ccf-mtDNA levels on systemic inflammation and clinical outcomes. A prospective observational study performed in South Korea evaluated ccf-mtDNA levels in patients on maintenance hemodialysis, showing that higher levels were associated with longer hemodialysis vintage, a more significant comorbidity burden, and higher inflammatory markers (15). Another study from China reported similar results in patients with maintenance hemodialysis compared with healthy controls (22). Chang et al. (23) did not find differences in plasma ccf-mtDNA among CKD stages; however, plasma ccf-mtDNA levels were numerically lower in patients with higher CKD stages.
Based on previous studies showing the potential benefits of aerobic exercise on inflammation and mitochondrial function, we hypothesized that in patients with moderate CKD, aerobic exercise would reduce plasma levels of ccf-mtDNA. Also, caloric restriction has anti-inflammatory and metabolic beneficial effects; some of them may be attributed to mitochondria (e.g., induces mitochondrial biogenesis) (6). However, our results indicate the opposite. mtDNA can appear in plasma after cell necrosis or apoptosis, but it can also be secreted without cell membrane rupture. It has been proposed that mtDNA can escape into the circulation during oxidative stress and dysfunctional mitophagy (24). Interestingly, both phenomena have been described in patients with CKD (1, 25). It is possible that in CKD, exercise and caloric restriction induce an abnormal escape of mtDNA. This may occur under situations like aerobic exercise and caloric restriction that increase mitochondrial biogenesis. On the other hand, it is also possible that we are observing a regression to the mean since there was a notable difference in ccf-mtDNA between the groups at baseline.
The effects of acute and long-term aerobic exercise on plasma levels of ccf-mtDNA are unclear. A study with healthy volunteers after a single bout of strenuous exercise found elevated plasma ccf-mtDNA levels (13). Stawski et al. (14) reported similar results after repeated bouts of strenuous exercise. In contrast, another study in healthy young men found that moderate aerobic exercise decreased plasma ccf-mtDNA over time (16). Further investigation should elucidate the effect of exercise on ccf-mtDNA, according to intensity, duration, and frequency.
We have previously reported that aerobic exercise decreased inflammatory markers in CKD (17); however, the same intervention did not decrease ccf-mtDNA levels. The lack of association could represent the limited sample size evaluated in this study; however, it is possible that in CKD, aerobic exercise affects plasma ccf-mtDNA and inflammatory markers by different pathways. It is possible that in CKD, ccf-mtDNA is not acting as a damage-associated molecular pattern, as occurs in other conditions. Further studies could elucidate the role of ccf-mtDNA on systemic inflammation and its clinic relevance in CKD.
This study’s strengths include clinical exercise physiologists conducting and personalizing the exercise intervention to the precise participant’s physiological capability and a dietary intervention dedicated to limiting overall calories rather than having a particular nutrient intake, perhaps improving participant’s tolerance and adherence. There are also limitations to be acknowledged. This study was a post hoc analysis of a pilot study designed to test feasibility, and it was not created with the biomarker outcome in mind. There is a between-individual variability in the levels of plasma ccf-mtDNA that could confound the results with the possibility of regression to the mean. To help with this possibility, ccf-mtDNA was measured three times to observe the within-individual changes. Also, generalized estimating equations will take into account differences between groups at baseline. A control group without CKD was not evaluated. Due to the limited sample size, we could not evaluate if differences in ccf-mtDNA exist between sex. The exact effect of sex in ccf-mtDNA deserves further evaluation. In addition, our study intervention was relatively short (but longer than previous studies), limiting our capacity to assess longer-term adherence and biological outcomes. Furthermore, other markers that may correlate with ccf-mtDNA were not measured.
Perspective and Significance
In conclusion, long-term aerobic exercise increased ccf-mtDNA levels in patients with moderate to severe CKD and more profoundly when combined with caloric restriction. Future studies should determine relationships of ccf-mtDNA to biomarkers of inflammation and oxidative stress as well as to clinical outcomes of exercise and dietary interventions in CKD.
GRANTS
This work was supported by National Institutes of Health Grants R01HL070938, K24DK62849, P30DK020593, P30DK114809, P30DK035816, P30ES000267, UL1TR000445, and UL1TR000423.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the authors.
AUTHOR CONTRIBUTIONS
B.R., M.G., K.R.T., J.H., and T.A.I., conceived and designed research; S.A.E.H., M.G., G.B., and J.L.G. performed experiments; M.M.P., L.L., T.S., C.R.-C., and J.L.G. analyzed data; J.J.-M., B.K., and J.L.G. interpreted results of experiments; J.J.-M. and M.M.P. prepared figures; J.J.-M. and B.K. drafted manuscript; J.J.-M., B.K., M.M.P., S.A.E.H., G.B., B.R., K.R.T., J.H., C.R.-C., T.A.I., and J.L.G. edited and revised manuscript; J.J.-M., T.A.I., and J.L.G. approved final version of manuscript.
REFERENCES
- 1.Himmelfarb J, Stenvinkel P, Ikizler TA, Hakim RM. The elephant in uremia: oxidant stress as a unifying concept of cardiovascular disease in uremia. Kidney Int 62: 1524–1538, 2002. doi: 10.1046/j.1523-1755.2002.00600.x. [DOI] [PubMed] [Google Scholar]
- 2.Pecoits-Filho R, Heimbürger O, Bárány P, Suliman M, Fehrman-Ekholm I, Lindholm B, Stenvinkel P. Associations between circulating inflammatory markers and residual renal function in CRF patients. Am J Kidney Dis 41: 1212–1218, 2003. doi: 10.1016/s0272-6386(03)00353-6. [DOI] [PubMed] [Google Scholar]
- 3.Yang Y. Metabolically healthy obesity and risk of incident chronic kidney disease in a Korean cohort study. Iran J Public Health 48: 2007–2015, 2019. [PMC free article] [PubMed] [Google Scholar]
- 4.Menshikova EV, Ritov VB, Fairfull L, Ferrell RE, Kelley DE, Goodpaster BH. Effects of exercise on mitochondrial content and function in aging human skeletal muscle. J Gerontol A Biol Sci Med Sci 61: 534–540, 2006. doi: 10.1093/gerona/61.6.534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Tonkonogi M, Sahlin K. Physical exercise and mitochondrial function in human skeletal muscle. Exerc Sport Sci Rev 30: 129–137, 2002. doi: 10.1097/00003677-200207000-00007. [DOI] [PubMed] [Google Scholar]
- 6.Civitarese AE, Carling S, Heilbronn LK, Hulver MH, Ukropcova B, Deutsch WA, Smith SR, Ravussin E; CALERIE Pennington Team. Calorie restriction increases muscle mitochondrial biogenesis in healthy humans. PLoS Med 4: e76, 2007. doi: 10.1371/journal.pmed.0040076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Boyapati RK, Tamborska A, Dorward DA, Ho GT. Advances in the understanding of mitochondrial DNA as a pathogenic factor in inflammatory diseases. F1000Res 6: 169, 2017. doi: 10.12688/f1000research.10397.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhang Q, Raoof M, Chen Y, Sumi Y, Sursal T, Junger W, Brohi K, Itagaki K, Hauser CJ. Circulating mitochondrial DAMPs cause inflammatory responses to injury. Nature 464: 104–107, 2010. doi: 10.1038/nature08780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zachariah RR, Schmid S, Buerki N, Radpour R, Holzgreve W, Zhong X. Levels of circulating cell-free nuclear and mitochondrial DNA in benign and malignant ovarian tumors. Obstet Gynecol 112: 843–850, 2008. doi: 10.1097/AOG.0b013e3181867bc0. [DOI] [PubMed] [Google Scholar]
- 10.Ellinger J, Albers P, Müller SC, von Ruecker A, Bastian PJ. Circulating mitochondrial DNA in the serum of patients with testicular germ cell cancer as a novel noninvasive diagnostic biomarker. BJU Int 104: 48–52, 2009. doi: 10.1111/j.1464-410X.2008.08289.x. [DOI] [PubMed] [Google Scholar]
- 11.Nakahira K, Kyung SY, Rogers AJ, Gazourian L, Youn S, Massaro AF, Quintana C, Osorio JC, Wang Z, Zhao Y, Lawler LA, Christie JD, Meyer NJ, Mc Causland FR, Waikar SS, Waxman AB, Chung RT, Bueno R, Rosas IO, Fredenburgh LE, Baron RM, Christiani DC, Hunninghake GM, Choi AM. Circulating mitochondrial DNA in patients in the ICU as a marker of mortality: derivation and validation. PLoS Med 10: e1001577, 2013. doi: 10.1371/journal.pmed.1001577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lindqvist D, Wolkowitz OM, Picard M, Ohlsson L, Bersani FS, Fernström J, Westrin Å, Hough CM, Lin J, Reus VI, Epel ES, Mellon SH. Circulating cell-free mitochondrial DNA, but not leukocyte mitochondrial DNA copy number, is elevated in major depressive disorder. Neuropsychopharmacology 43: 1557–1564, 2018. doi: 10.1038/s41386-017-0001-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ohlsson L, Hall A, Lindahl H, Danielsson R, Gustafsson A, Lavant E, Ljunggren L. Increased level of circulating cell-free mitochondrial DNA due to a single bout of strenuous physical exercise. Eur J Appl Physiol 120: 897–905, 2020. doi: 10.1007/s00421-020-04330-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Stawski R, Walczak K, Kosielski P, Meissner P, Budlewski T, Padula G, Nowak D. Repeated bouts of exhaustive exercise increase circulating cell free nuclear and mitochondrial DNA without development of tolerance in healthy men. PLoS One 12: e0178216, 2017. doi: 10.1371/journal.pone.0178216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kim K, Jung SW, Cho W-H, Moon H, Jeong KH, Kim JS, Lee SH, Ahn SY, Yang DH, Lee HJ, Lee DY, Moon JY, Kim YG. Associations between cell-free mitochondrial DNA and inflammation, and their clinical implications for patients on hemodialysis: a prospective multicenter cohort study. Blood Purif 50: 214–221, 2021. doi: 10.1159/000510088. [DOI] [PubMed] [Google Scholar]
- 16.Shockett PE, Khanal J, Sitaula A, Oglesby C, Meachum WA, Castracane VD, Kraemer RR. Plasma cell-free mitochondrial DNA declines in response to prolonged moderate aerobic exercise. Physiol Rep 4: e12672, 2016. doi: 10.14814/phy2.12672. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ikizler TA, Robinson-Cohen C, Ellis C, Headley SAE, Tuttle K, Wood RJ, Evans EE, Milch CM, Moody KA, Germain M, Limkunakul C, Bian A, Stewart TG, Himmelfarb J. Metabolic effects of diet and exercise in patients with moderate to severe CKD: a randomized clinical trial. J Am Soc Nephrol 29: 250–259, 2018. doi: 10.1681/ASN.2017010020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ 340: c332, 2010. doi: 10.1136/bmj.c332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Inker LA, Schmid CH, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, Kusek JW, Manzi J, Van Lente F, Zhang YL, Coresh J, Levey AS; CKD-EPI Investigators. Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med 367: 20–29, 2012. doi: 10.1056/NEJMoa1114248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pollara J, Edwards RW, Lin L, Bendersky VA, Brennan TV. Circulating mitochondria in deceased organ donors are associated with immune activation and early allograft dysfunction. JCI Insight 3: e121622, 2018. doi: 10.1172/jci.insight.121622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.StataCorp. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC, 2019 https://www.stata.com/stata16/.
- 22.Cao H, Ye H, Sun Z, Shen X, Song Z, Wu X, He W, Dai C, Yang J. Circulatory mitochondrial DNA is a pro-inflammatory agent in maintenance hemodialysis patients. PLoS One 9: e113179, 2014. doi: 10.1371/journal.pone.0113179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chang C-C, Chiu P-F, Wu C-L, Kuo C-L, Huang C-S, Liu C-S, Huang C-H. Urinary cell-free mitochondrial and nuclear deoxyribonucleic acid correlates with the prognosis of chronic kidney diseases. BMC Nephrol 20: 391, 2019. doi: 10.1186/s12882-019-1549-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Perez-Trevino P, Velasquez M, Garcia N. Mechanisms of mitochondrial DNA escape and its relationship with different metabolic diseases. Biochim Biophys Acta Mol Basis Dis 1866: 165761, 2020. doi: 10.1016/j.bbadis.2020.165761. [DOI] [PubMed] [Google Scholar]
- 25.Forbes JM, Thorburn DR. Mitochondrial dysfunction in diabetic kidney disease. Nat Rev Nephrol 14: 291–312, 2018. doi: 10.1038/nrneph.2018.9. [DOI] [PubMed] [Google Scholar]