Metabolic Dysfunction in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome Not Due to Anti-mitochondrial Antibodies

Metabolic profiling studies have recently indicated dysfunctional mitochondria in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). This includes an impaired function of pyruvate dehydrogenase complex (PDC), possibly driven by serum factor(s), which leads to inadequate adenosine triphosphate generation and excessive lactate accumulation. A reminiscent energy blockade is likely to occur in primary biliary cholangitis (PBC), caused by anti-PDC autoantibodies, as recently proposed. PBC is associated with fatigue and post-exertional malaise, also signifying ME/CFS. We herein have investigated whether ME/CFS patients have autoreactive antibodies that could interfere with mitochondrial function. We found that only 1 of 161 examined ME/CFS patients was positive for anti-PDC, while all PBC patients (15/15) presented significant IgM, IgG, and IgA anti-PDC reactivity, as previously shown. None of fibromyalgia patients (0/14), multiple sclerosis patients (0/29), and healthy blood donors (0/44) controls showed reactivities. Anti-mitochondrial autoantibodies (inner and outer membrane) were negative in ME/CFS cohort. Anti-cardiolipin antibody levels in patients did not differ significantly from healthy blood donors. In conclusion, the impaired mitochondrial/metabolic dysfunction, observed in ME/CFS, cannot be explained by presence of circulating autoantibodies against the tested mitochondrial epitopes.

Accumulating evidence are pointing toward an autoimmune phenotype for ME/CFS. The presence of self-reacting antibodies in the circulation of patients include nuclear and membrane structures, neurotransmitters and their receptors, neoautoantigens formed by oxidative or nitrosative damage, and autoantibodies targeted to mitochondrial components (Table 1). However, both the frequency and the titers of autoantibodies and their correlation to disease severity or symptoms has had limited reproducibility between different studies and patient cohorts. Still, a subset of ME/CFS patients presented amelioration of symptoms following antibody removal treatment (18). Specific changes in the proteome of CSF of ME/CFS patients involved the accumulation of complement components, which signify antibody activity (19). In a recent publication from our group, the serological profile of the same ME/CFS patient cohorts demonstrated evidence of minor alterations of antibody reactivities against the ubiquitous herpesviruses when compared to healthy controls (20). These alterations may indicate shortcomings in humoral responses in ME/CFS which are hallmarks of autoimmune diseases.
Recent reports point toward a central metabolic defect in ME/CFS, which affects aerobic energy production via the tricarboxylic acid (TCA) cycle in mitochondria, leading to a diminished production of adenosine triphosphate (ATP) and excessive lactate generation upon exertion, possibly explaining PEM (21,22). The transition between anaerobic and aerobic energy production is catalyzed by the pyruvate dehydrogenase complex (PDC). Autoantibodies specific for PDC is a hallmark of primary biliary cholangitis (PBC), a potential disease model of autoantibody-mediated energy blockade (23,24). In analogy with PBC, in which energy production is inhibited by antibodies (25), circulating energy inhibitors have also been detected in ME/CFS (21), however, their molecular nature is unknown. It would be reasonable if these circulating inhibitors turned out to be immunoglobulins, presumably directed against mitochondrial antigens. We have therefore investigated the presence of antimitochondrial antibodies and anti-PDC reactive autoantibodies, in ME/CFS patients.

Participants
All ME/CFS patients included in this study were diagnosed according to the Canadian consensus criteria (3). ME/CFS patients reported impairment was assessed by the Fibro-fatigue scale (26). Blood samples were acquired from three ME/CFS cohorts. Cohort 1 (n = 74): 46 ME/CFS patients, 17 ME/CFS + fibromyalgia (FM) patients, and 11 FM patients. This cohort also included 29 multiple sclerosis (MS) patients. Cohort 2 (n = 61): 61 ME/CFS patients; Cohort 3 (n = 40): 18 ME/CFS patients, 19 ME/CFS/FM patients, 3 FM patients, and 15 age-matched healthy donors in cohort 3 (HD3). Samples from cohorts 1-3 originated from the Gottfries Clinic, Mölndal, Sweden. The characteristics of the patients are summarized in Table 2. Plasma samples from 15 PBC patients were collected at the blood bank of The Medical School in The University of Newcastle upon Tyne, UK. Additional controls included serum samples from 46 anonymous healthy blood donors from Uppsala Academic Hospital University, Sweden.

Ethics Statement
Informed consent for blood collection was obtained from all patients according to the Declaration of Helsinki, and ethical approval was granted by the ethical review committee at Medical Faculty, University of Gothenburg Dnr 029-13, Dnr 867-13, T 091-16, and Dnr 852-12, T 092-16 (CGG). Blood donor samples were collected anonymously according to ethical consent from the Uppsala Institutional Review Board (IRB) UPS_01_367, and Linköping Regional Office of Swedish Ethical Review Authority. The 15 PBC patient samples from Newcastle upon Tyne, were under REC number 12/NE/0095.

Validation of Anti-PDC Antibodies
Extracted human PDC was resolved with 10% SDS PAGE and immunoblotting was performed as previously described (24,27). The membrane was blocked with 5% (w/v) skimmed milk powder and then probed with patient plasma from three ME/CFS and one ME/CFS/FM, all diluted 1:500 in 0.5% w/v of BSA in PBS-T. Bound antibodies were detected using goat anti-human IgG peroxidase-conjugated antibodies (Sigma, Poole, UK) and enhanced chemiluminescence (ECL; Amersham, Aylesbury, UK) according to the manufacturer's protocol.

Statistical Analysis
For the comparison between multiple patient groups, we used one-way analysis of variance (ANOVA) using GraphPad Prism version 6.0 software (GraphPad Software, San Diego, CA, USA). Only statistically significant differences p < 0.05 were reported.

Autoantibodies Against Mitochondrial Pyruvate Dehydrogenase Complex (PDC) Not Detected in ME/CFS
In this study, we analyzed whether dysfunctional energy generation from mitochondria in ME/CFS could be explained by the presence of reactive autoantibodies directed against the PDC enzyme, in analogy to what has been observed in PBC. ME/CFS, PBC, FM, MS and healthy blood donor controls were analyzed (Figure 1). PBC patient plasma samples were all positive for IgG, IgM, and IgA anti-PDC antibodies and hence presented with statistically significant differences (p < 0.0001). In repeated analyses, these samples remained positive down to dilutions of 1:10 5− 1:10 6 . In contrast to our hypothesis, ME/CFS samples were negative for autoantibodies against PDC, with the exception of three ME/CFS patients: #43 (A = 1.217 at 1:500); ME/CFS patient #69 (A = 0.406 at 1:500); ME/CFS patient #166 (A = 0.418 at 1:500); and #21, a ME/CFS patient with FM comorbidity (A =1.658 at 1:500) that was found weakly/intermediately positive. These plasma samples were tested additionally by immunoblot analysis for validation. Only sample from patient #21 (ME/CFS+FM) was positive, whereas the other samples were negative (Figure 2). Autoreactivity was identified against the major autoreactive epitopes within the PDC complex, dihydrolipoamide acetyltransferase (E2) and the E3-binding protein (E3BP) as previously described (24).

AMA and ACA Not Detected in ME/CFS Patients
The ME/CFS patient group did not exhibit AMAs, against inner and outer mitochondrial membrane, nor antibodies against other mitochondrial structures, nor other structures (nuclear, smooth muscle, gastric parietal cell). Anti-cardiolipin antibody (IgG, IgM, IgA) levels were not significantly different in plasma of ME/CFS or FM patients compared with blood donors (Figure 2).

DISCUSSION
We conclude from the results presented in this study, that in contrast to our initial expectation, ME/CFS patients do not have autoantibodies to epitopes of mitochondrial components. This does not rule out that other important indirect or secondary mechanisms exist in blocking the complex mitochondrial pyruvate metabolism in ME/CFS. An example of autoimmunity with autoreactive antibodies against a key metabolic factor is PBC. It was previously shown that anti-PDC antibodies in these patients are mainly directed against the inner lipoyl domain of the PDC-E2 component, which has an alpha-lipoic acid covalently bound to a specific lysine residue, which is an absolute requirement for its enzymatic activity. Lipoylation is a posttranslational modification, which also occurs in bacteria like E. coli or Novosphingobium sp. (28). Autoreactive responses in PBC have been suggested to arise after infection in the gut with these bacteria (29,30) or beta-retroviruses (31), or exposure to environmental xenobiotics that mimic the native lipoic acid moiety (32). PBC patients have a PEM reminiscent of PEM in ME/CFS and FM patients (33). FM is a common comorbidity in both ME/CFS, occurring in ∼30-70% of the patients, and autoimmunity (34)(35)(36). Metabolic disturbances in ME/CFS with reduced energy supply and increased lactate may account for most of the malaise in PEM as well as cognitive and physical disturbances (37). Notably, accumulation of lactate in blood and muscles after exercise, along with increased concentrations of lactate in cerebrospinal fluid (CSF) have been found in ME/CSF (38,39). However, our present comparative analysis for AMA, anti-PDC, and anti-cardiolipin autoantibodies in the 3 cohorts did not yield a positive or diseasespecific result. A link between mitochondrial dysfunction and innate immune dysregulation is suggested which demonstrate that the energy producing organelles (mitochondria and peroxisomes) are coupled via mitochondrial antiviral signaling proteins (MAVS) to the inflammasome (40). Hypothetically, dangerassociated molecular patterns (DAMPs), such as mitochondrial DNA (mtDNA), heat shock proteins (Hsps), and cardiolipins could be released from the mitochondria (41)(42)(43) in ME/CFS and act as autoantigens. We previously found that a subset of ME/CFS patients had higher levels of IgM antibodies against epitopes of both mitochondrial and bacterial Hsp-60 (despite the absence of an infectious pathogen) which potentially may lead to dysfunctional mitochondria (11). Our hypothesis on the presence of autoantibodies against cardiolipin in ME/CFS patients was based on previous publications (ref in Table 1), could not be confirmed here, possible explained by different inclusion and exclusion criteria of patients in previous studies that did not used Canadian criteria, but included leukemia/lymphoma and type 1 diabetes patients. Further studies are required to analyse unique epitope alterations such as oxidation epitopes of cardiolipin (42), length and 3D-structure, which each may signal differently in this complex immunemetabolic scenario.
The hypothesis of a direct blocking of PDC due to autoreactive antibodies is excluded based on our present findings, however, blocking could be secondary and a sign of compensatory efforts to balance some other abnormalities in ME/CFS, as suggested by Fluge et al. (21). Their data indicate a functional impairment of PDC leading to increased consumption of specifically ketogenic and anaplerotic amino acids that fuel alternative pathways for ATP production independently of PDC. The results were supported by increased mRNA expression of pyruvate dehydrogenase kinases that are inhibitory to PDC (21). Recent studies also show that many fatty acids may generate acetyl-CoA that fuel the TCA-cycle independently of PDC (44). Similar observations on altered energy metabolism may occur in different types of cellular stress such as starvation (45). Blocking of PDC could also be generated by other mechanisms including reduction-oxidation (redox) reactions known to modulate lipoic acid moiety in the PDC (46). Therefore, it would be of interest to investigate in future studies the PDC redox status in ME/CFS patients. Finally, multiple enzymes such as mitochondria pyruvate carrier, pyruvate carboxylases, PDkinases, PD-phosphatases, in addition to PDC, modulate overall pyruvate carbon flux (47).

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the ethical review committee at Medical Faculty, University of Gothenburg; Uppsala Institutional Review Board; Linköping Regional Office of Swedish Ethical Review Authority, Newcastle upon Tyne, under REC number 12/NE/0095; Regional ethic committee of Stockholm. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
IN, JP, and EA designed and performed experiments, analysis, and authored the manuscript. MR and C-GG provided critical materials. CD performed experiments and analysis. AR conceived the design and execution of all experimental procedures and authored the manuscript. All authors critically reviewed the final manuscript.

FUNDING
This work was financed with grants from the Swedish ME Association (AR), Solve ME/CFS (AR), the Swedish Cancer Association (AR), and the Open Medicine Foundation (OMF) (AR).