Coenzyme Q10 defects may be associated with a deficiency of Q10-independent mitochondrial respiratory chain complexes

Background Coenzyme Q10 (CoQ10 or ubiquinone) deficiency can be due either to mutations in genes involved in CoQ10 biosynthesis pathway, or to mutations in genes unrelated to CoQ10 biosynthesis. CoQ10 defect is the only oxidative phosphorylation disorder that can be clinically improved after oral CoQ10 supplementation. Thus, early diagnosis, first evoked by mitochondrial respiratory chain (MRC) spectrophotometric analysis, then confirmed by direct measurement of CoQ10 levels, is of critical importance to prevent irreversible damage in organs such as the kidney and the central nervous system. It is widely reported that CoQ10 deficient patients present decreased quinone-dependent activities (segments I + III or G3P + III and II + III) while MRC activities of complexes I, II, III, IV and V are normal. We previously suggested that CoQ10 defect may be associated with a deficiency of CoQ10-independent MRC complexes. The aim of this study was to verify this hypothesis in order to improve the diagnosis of this disease. Results To determine whether CoQ10 defect could be associated with MRC deficiency, we quantified CoQ10 by LC-MSMS in a cohort of 18 patients presenting CoQ10-dependent deficiency associated with MRC defect. We found decreased levels of CoQ10 in eight patients out of 18 (45 %), thus confirming CoQ10 disease. Conclusions Our study shows that CoQ10 defect can be associated with MRC deficiency. This could be of major importance in clinical practice for the diagnosis of a disease that can be improved by CoQ10 supplementation.


Background
Coenzyme Q 10 (CoQ 10 or ubiquinone) is a lipid-soluble component of the mitochondrial inner membrane that plays a central role in mitochondrial respiratory chain (MRC) function, as electrons carrier from complexes I and II to complex III, thus participating in ATP production [1].
CoQ 10 deficiency encompasses several clinical phenotypes such as encephalomyopathy, severe infantile multisystemic disease, cerebellar ataxia, isolated myopathy or nephrotic syndrome [2]. CoQ 10 deficiency can be primary, due to mutations in genes involved in CoQ 10 biosynthesis or secondary, due to mutations in genes unrelated to CoQ 10 biosynthesis [3]. Secondary CoQ 10 deficiency has been described in patients with mitochondrial DNA (mtDNA) mutations or deletions, with mtDNA depletion syndrome (MDS) [4][5][6] and in patients with mutations in APTX [7], ETFDH [8,9], BRAF [10], ACADVL or NPC genes [11]. CoQ 10 defect is the only oxidative phosphorylation (OXPHOS) disorder that can be clinically improved after oral CoQ 10 supplementation with limitation of neurological and renal manifestations, amelioration of muscular symptoms and attenuation of histological alterations. Early treatment is crucial to prevent irreversible damage in organs such as the kidney and the central nervous system [12][13][14]. Reduced activities of CoQ 10 -dependent enzymes by spectrophotometric

Open Access
Biological Research *Correspondence: paquis@hermes.unice.fr 1 School of Medicine, IRCAN, UMR CNRS 7284/INSERM U1081/UNS, Nice Sophia-Antipolis University, 28 av de Valombrose, 06107 Nice Cedex 2, France Full list of author information is available at the end of the article analysis (segments I + III or G3P + III and II + III) are evocative of CoQ 10 deficiency but direct measurement of CoQ 10 levels is the most reliable test for diagnosis [15]. It is widely reported in the literature that, in patients with CoQ 10 deficiency, enzymatic activities of MRC complexes I, II, III, IV, V are normal [16]. In a previous report, we described an 11-year-old boy presenting with a propionic acidemia who succumbed to acute heart failure in the absence of decompensation of his metabolic condition. Spectrophotometric analysis in liver identified CoQ 10dependent activities deficiency that was associated with MRC enzymatic defect. Secondary CoQ 10 deficiency was likely involved in the development of heart complications in this child and we hypothesized that a CoQ 10defect may be associated with MRC deficiency [17]. The aim of this study was to verify this hypothesis in order to improve the diagnosis of this disease.
Over a 6-year period, we analyzed by spectrophotometry 700 tissue samples from 495 patients in whom a mitochondrial disease was suspected. Isolated CoQ 10dependent activity deficiency led to identification of CoQ 10 disease in eight cases. Eighteen patients presented CoQ 10 -dependent enzymatic deficiency associated with MRC defect by spectrophotometry in muscle or in fibroblasts. In order to validate our original observation and to establish if CoQ 10 quantitative defect may be associated with multiple MRC enzymatic deficiency, we measured CoQ 10 in this group of 18 patients. We found decreased CoQ 10 levels by liquid chromatography coupled with tandem mass spectrometry detection (LC-MSMS) in eight patients out of 18 (45 %), thus confirming CoQ 10 disease and its association with MRC enzymatic deficiency. Furthermore, CoQ 10 disease cannot be ruled out in all other patients insofar as the quantitative assay could not always be performed in the affected tissue.

Description of patients involved in the study
We studied 18 patients, including 10 males and eight females, ranging in age from day 1 to 76 years. Clinical presentations were very heterogeneous ( Table 1). The age at onset of the disease was highly variable, ranging from (i) neonatal forms (seven cases with severe phenotypes), (ii) onset before 1 year of age (four cases with either Leigh syndrome or epileptic encephalopathy), (iii) childhoodonset (four cases including two myopathic forms and two complex phenotypes) to (iv) adult-onset (three cases with two myopathic presentations and one cerebellar ataxia). The 18 patients were divided into two different groups according to molecular results.
The first group included 10 patients with identified mutations in responsible genes (Table 1). Patient P01 presented a severe neonatal multisystemic disease secondary to a homozygous missense mutation in the CoQ2 gene [18]. Spectrophotometric analysis in fibroblasts revealed a CoQ 10 -dependent activities defect (segments II + III and G3P + III reduction) associated with a complex IV deficiency ( Table 2). Six patients (P02-P07) presented a mitochondrial disease or dysfunction secondary either to mtDNA abnormalities (P02 and P03) or to mutations in nuclear genes (P04-P07). Patient P02 had a large heteroplasmic mtDNA deletion responsible for Kearns-Sayre syndrome and patient P03 presented with a severe neonatal polyvisceral failure secondary to a heteroplasmic mtDNA mutation in the MT-CYB gene. Patients P04 and P05 presented with sensory ataxia neuropathy dysarthria and ophthalmoplegia (SANDO) phenotype associated with recessive mutations in POLG. Patient P06 had a neonatal encephalopathy with lactic acidosis and mild methylmalonic aciduria linked to mutations in the SUCLG1 gene. P07 had a diagnosis of multiple acyl-CoA dehydrogenation deficiency (MADD) with ETFDH mutation. The last three patients in the first group presented malignant migrating partial seizures with mutations in TBC1D24 (P08), CDG syndrome type Iq with SRD5A3-CDG mutations (P09) and 1p36 deletion syndrome (P10). Patients P02-P10 had a CoQ 10 -dependent activities deficiency (segments I + III or G3P + III and II + III reduction) associated with a multiple MRC defect in muscle or in fibroblasts ( Table 2).
The second group included eight patients suspected of CoQ 10 deficiency with an absence of molecular diagnosis. Except for individual P11, who developed cerebellar ataxia during adulthood, all patients had an early-onset disease ranging from neonatal period to childhood. They presented severe neurological symptoms including two Leigh syndromes (P12 and P16) and one child had an unexplained severe respiratory failure at birth (P17). In the second group, all patients presented a CoQ 10dependent enzymatic deficiency associated with MRC defect in muscle or in fibroblasts ( Table 2).

Confirmation of CoQ 10 disease in eight patients by CoQ 10 quantification
Quantitative analysis of CoQ 10 in muscle or fibroblasts showed that eight patients presented CoQ 10 content below normal values ( Table 2). CoQ 10 defect was found in five patients out of 10 in the first group and in three patients out of eight in the second group. CoQ 10 -deficient individuals were six males and two females, ranging in age from day 1 to 76 years. The age of onset was highly variable, ranging from neonatal forms to diseases appearing after 25 years of age, although six patients had childhood onset. One patient (P01) presented a polyvisceral failure at birth and all the others had neurological symptoms either isolated or combined with muscular and/or other signs. In the first group, the very low CoQ 10 level observed in the fibroblasts of patient P01 confirmed the primary CoQ 10 defect associated with the c.437G > A homozygous missense mutation (p.Ser146Asn) in the CoQ2 gene, involved in CoQ 10 biosynthesis [18]. In the four other patients in the same group, CoQ 10 defect was clearly secondary because the responsible genes were unrelated to CoQ 10 biosynthesis. Three patients had a mitochondrial disease linked to a large mtDNA deletion (patient P02) or to mutations in POLG (patient P04) or SUCLG1 (patient P06). Patient P08 alone did not have a mitochondrial disease, her encephalopathy with refractory malignant migrating partial seizures being linked to mutations in the TBC1D24 gene. In the second group, low CoQ 10 levels were found in three patients with no molecular diagnosis. Two patients were strongly suspected of having a mitochondrial disease: patient P11, who had a cerebellar ataxia with 20-30 % of COX-negative fibers and multiple mtDNA deletions in muscle, and patient P16 who presented with a Leigh syndrome. The last patient (P15) had an encephalopathy with intellectual disability but no histological sign of mitochondrial myopathy.

Discussion
While primary CoQ 10 defects are rare, secondary defects have been observed in various pathologies. In a previous work, we suspected for the first time a secondary CoQ 10 defect in a child with propionic acidemia, who succumbed to acute heart failure in the absence of decompensation of his metabolic condition [17]. CoQ 10 deficiency was not evoked at the outset because CoQ 10dependent activities deficiency was associated with multiple MRC deficiency in the liver of the patient and it had been widely reported that enzymatic activities of MRC complexes are normal in CoQ 10 disease [16]. However, it is likely that a secondary CoQ 10 defect was involved in the development of heart complications leading to the child's death and that oral CoQ 10 supplementation would have been able to prevent cardiac failure if results had been obtained before acute clinical aggravation. This hypothesis is supported by a recent study, which describes a successful reversal of propionic acidemia-associated cardiomyopathy after treatment [19]. The child in this case presented with myocardial CoQ 10 quantitative defect associated with signs of mitochondrial dysfunction such as enlarged mitochondria with atypical cristae and low MRC complex IV activity [19]. Several studies performed on cellular models of CoQ 10 defect suggested a possible association with mitochondrial dysfunction: PDSS2 and COQ9 mutant fibroblasts presented a markedly reduced ATP synthesis and COQ2 mutant fibroblasts presented a partial defect in ATP synthesis, as well as significantly increased ROS production and oxidation of lipids and proteins [20,21]. In 2013, Duberley and colleagues established the first pharmacologically-induced CoQ 10 deficient cellular model in neuroblastoma-derived SH-SY5Y cells by using para-aminobenzoic acid (PABA). They showed that, after PABA treatment, SH-SY5Y cells presented a progressive decrease in the activities of CoQ 10dependent II + III segment but also a deficiency in MRC complexes I and IV. They also reported a concomitant decrease in the level of total cellular ATP with an increase of mitochondrial oxidative stress [22]. Lastly, deficiency of complexes I, II, III and/or IV has also been previously reported in association with CoQ 10 defect in the patient's fibroblasts, muscle or kidney [8,11,18,23]. Today, in most diagnostic laboratories, a spectrophotometric deficiency in one or several MRC enzymes associated with a decrease in CoQ 10 -dependent activities is not considered to be a sign of a CoQ 10 disease, leading to a possible under-estimation of the frequency of this disorder. With the aim of achieving a better diagnostic approach, we quantified CoQ 10 by LC-MSMS in 18 patients presenting a CoQ 10 -dependent enzymatic deficiency associated with a MRC defect by spectrophotometry. CoQ 10 quantitative analysis in muscle or in fibroblast cells confirmed CoQ 10 disease in eight patients (45 %). These data show that a primary CoQ 10 defect can be associated with MRC enzymatic deficiency because patient P01, who carried a deleterious homozygous mutation (c.437G > A; p.Ser146Asn) in the CoQ2 gene, also presented a complex IV deficiency in muscle. Our data also confirm that a secondary CoQ 10 defect can be associated with mitochondrial disease. Indeed, three other patients with a low CoQ 10 level presented a respiratory chain deficiency linked to mtDNA deletion (patient P02) or to mutations in POLG and SUCLG1 genes (patients P04 and P06). Secondary CoQ 10 defect has already been reported in patients with mitochondrial diseases or dysfunctions including Kearns-Sayre syndrome [24], mtDNA depletion and PEO [5] or mutations in ETFDH coding for electrontransferring-flavoprotein dehydrogenase and causing MADD [8,9]. Secondary CoQ 10 defect has also been described in non-mitochondrial disorders linked to genes such as APTX coding for aprataxin and causing ataxia occulomotor-apraxia [7], BRAF coding for serine/threonine-protein kinase B-Raf and causing cardiofaciocutaneous syndrome [10], ACADVL causing very long-chain Acyl-CoA dehydrogenase deficiency or NPC causing Niemann-Pick-type C disease [11]. Here, we report for the first time a secondary CoQ 10 defect associated with mutations in the TBC1D24 gene, leading to malignant migrating partial seizures (Patient P08). The mechanisms linking CoQ 10 defect and decreased activity of MRC complexes are unknown. Studies in patients with metabolic diseases showed an increase in oxidative stress-markers and a decrease in antioxidant defences [25]. More specifically, ubiquinol depletion in patient tissues may lead to increased reactive oxygen species activity [26] and, since all enzymes of the MRC are susceptible to free radical induced oxidative damage [27], we can hypothesize that CoQ 10 -independent MRC dysfunction may result from a high level of mitochondrial oxidative stress creating an imbalance with the CoQ 10 antioxidant capacity, as previously evoked [25]. In parallel, a possible reason for a secondary CoQ 10 defect resulting from a primary MRC deficiency is that the enzymes involved in CoQ 10 biosynthesis are found in a supercomplex in the inner mitochondrial membrane [28]. We hypothesize that the increased oxidative stress resulting from a primary MRC deficiency may inhibit these enzymes resulting in a secondary CoQ 10 defect.

Conclusions
In conclusion, our work highlights the probability that, based on spectrophotometric analysis, the frequency of CoQ 10 disease is underestimated in routine clinical practice. Several studies, which performed a systematic CoQ 10 quantification on muscle biopsies from pediatric and adult populations presenting a wide range of clinical phenotypes, also reported an underestimation of CoQ 10 defects and proposed a systematic evaluation of CoQ 10 content in all muscle biopsies [5,29,30]. However, firstline CoQ 10 quantification seems difficult to set up as a routine analysis in all diagnosis laboratories. Based on our observations, we suggest that CoQ 10 quantification be performed in all tissues presenting a spectrophotometric deficiency of CoQ 10 -dependent enzymes, associated or not with MRC defect, regardless of the patient's age, clinical presentation or molecular diagnosis. This could prove of great value in clinical practice for the diagnosis of a disease that can be improved by CoQ 10 supplementation.

Patients
All patients were explored in the Reference Centre for Mitochondrial Disease (CHU of Nice, France). Selection of the 18 patients was based on the following inclusion criteria: (1) availability of a muscle sample or fibroblast culture and (2) spectrophotometric deficiency of CoQ 10dependent activities (reduction of segments I + III or G3P + III and II + III) associated with MRC defect in muscle or in fibroblasts. The following data were systematically collected: sex, age at biopsy, age of onset, heredity, familial history, clinical presentation, brain MRI, metabolic screening, mitochondrial enzymatic studies, histological and molecular analyses. The age of onset of clinical symptoms ranged from neonatal period to 45 years of age. Blood and tissue samples were obtained after adult patients and parents of affected children had given informed consent.
Patients were divided into two groups (Table 1), according to the results of molecular analysis: (1) individuals with a molecular diagnosis, carrying mutations in mtDNA or in nuclear genes and, (2) individuals with no molecular diagnosis.

Cell culture
Primary fibroblast cultures were obtained from patient skin punches, using standard procedures, in RPMI medium supplemented with 10 % Fetal Bovine Serum, 45 μg/ml uridine and 275 μg/ml sodium pyruvate. Cultures were incubated at 37 ℃ with 5 % CO 2 .

OXPHOS spectrophotometric measurements
Enzymatic spectrophotometric measurements of the OXPHOS respiratory chain complexes and citrate synthase were performed at 37 ℃ on muscle crude homogenates or fibroblasts according to standard procedures [31]. Proteins were measured according to Bradford microassay [32] and results were expressed as nmole/ min/mg of proteins.

Coenzyme Q 10 quantification
Total coenzyme Q 10 was extracted from tissues and analyzed by reverse phase liquid chromatography separation (column C18 symmetry 150 × 2.1 mm, 3.5 µm, Waters, France) as previously described [33]. Detection and quantification were done by mass spectrometry using an API 3000 tandem mass spectrometer (ABSciex, France) equipped with an APCI source. CoQ 10 and CoQ 9 were analyzed in the positive mode using the following m/z 864 → 197 and 796 → 197 transitions. CoQ 9 was used as internal standard for quantification. External calibration was performed using CoQ 10 solutions. A stock solution was prepared by dissolving 10 mg of CoQ 10 in 4 ml of methanol/chloroform (98:2 v/v). This solution was stable for 3 months at −80 °C. The working solutions were prepared daily by diluting the stock solution into methanol to provide a range of 0.05-1 µmol/L. The intra-assay and inter-assay CV's were, respectively, 5.7 and 6.3 % for a CoQ 10 concentration of 0.25 µmol/L.