Introduction

Coenzyme Q10 (CoQ10) is a lipophilic molecule critical for the transport of electrons from complex I and complex II (and also from the β-oxidation pathway via the electron transfer flavoprotein, ETF) to complex III in the mitochondrial respiratory chain (RC) (Festenstein et al 1955; Crane et al 1957; Frerman 1987). It also participates in extra-mitochondrial electron transport and functions as an antioxidant in cell membranes preventing lipid, protein and DNA oxidation. Moreover, CoQ10 is involved in the regulation of mitochondrial uncoupling proteins and mitochondrial permeability transition pore; it is also required for pyrimidine nucleoside biosynthesis and may modulate apoptosis (Turunen et al 2004).

In humans CoQ10 is synthesized in cells and tissues and no uptake is usually required; 2–4 % of the dietary CoQ10 is recovered in the circulation, but its transfer to the organs seems very limited (Turunen et al 2004).

CoQ10 is composed of a benzoquinone ring derived from tyrosine and a decaprenyl side-chain coming from the mevalonate (MV) pathway after successive additions of isopentenyl-diphosphate (IPP) molecules to farnesyl-diphosphate (FPP) catalyzed by prenyl-diphosphate synthase (COQ1) (Fig. 1) (Dallner and Sindelar 2000). Decaprenyl diphosphate (DPP) and p-hydroxybenzoate (PHB) are condensed by PHB-polyprenyltransferase (COQ2), and further modified by at least six enzymes catalyzing methylation, decarboxylation, and hydroxylation reactions to synthesize the final CoQ10 molecule. The MV pathway comprises the reactions from acetyl-coenzyme A (acetyl-CoA) to FPP, which is precursor for CoQ10, cholesterol, dolichol and isoprenylated proteins (Turunen et al 2004; Dallner and Sindelar 2000).

Fig. 1
figure 1

Biosynthetic pathway of CoQ10. Modified from Dallner and Sindelar (Dallner and Sindelar 2000)

Full size image

Primary CoQ10 deficiencies are described as genetic disorders with good response to supplementation with CoQ10. Early treatment based on early diagnosis is critical to maximize the efficacy of ubiquinone supplementation (López et al 2010). These mitochondrial disorders are rare conditions that have been reported in individuals with various clinical phenotypes showing decreased activities of the RC complexes I+III and II+III, and low levels of CoQ10 (Rahman et al 2012; Ogasahara et al 1989; Rötig et al 2000; Salviati et al 2005; Horvath et al 2006; Quinzii et al 2007; Rustin et al 2004) in muscle or fibroblasts. The diversity of symptoms along with the large number of genes involved in the synthetic pathway and the frequent secondary CoQ10 deficiencies make it difficult to achieve a definitive diagnosis. CoQ10 deficiencies are primary when due to mutations in genes involved in CoQ10 biosynthesis (COQ genes), where even haploinsufficiency for the COQ4 gene has been described to cause CoQ deficiency (Salviati et al 2012). It can also be secondary to genes not directly involved in it, such as APTX (aprataxin) (Quinzii et al 2005), ETFDH (electron-transferring-flavoprotein dehydrogenase) (Gempel et al 2007; Liang et al 2009) or BRAF (Aeby et al 2007). Secondary deficiencies have also been reported in patients with mitochondrial DNA (mtDNA) mutations or deletions (Rahman et al 2012; Sacconi et al 2010; Matsuoka et al 1991), and some specific genetic factors may confer susceptibility to develop secondary CoQ10 deficiency (Sacconi et al 2010).

Skeletal muscle is accepted as the tissue of choice for CoQ10 evaluation, but obtaining a muscle biopsy is invasive. Less invasive procedures such as obtaining lymphoblastoid cell lines, fibroblasts, or lymphocytes have been used for the diagnosis of CoQ10 deficiency (Rahman et al 2012; Montero et al 2008; Arias et al 2012).

For these reasons, our objective was to develop a methodology for the study of the endogenous biosynthesis of CoQ10 in fibroblasts that may allow the identification of primary CoQ10 deficiencies.

Materials and methods

Reagents

13C6-PHB, 2H3-MV, non-labelled PHB, non-labelled MV, cyclodextrine, 5,5′-ditio-bis[−2-nitrobenzoic acid] (DTNB), oxaloacetate, tris(hydroxymethyl)aminomethane (Tris), saccharose, EDTA, coenzyme Q9 (CoQ9), CoQ10, thiazolyl blue tetrazolium bromide (MTT), sodium dodecyl sulphate (SDS), dimethyl sulfoxide, ammonium bicarbonate and cycloheximide (CHX) were provided from Sigma-Aldrich (Madrid, Spain).

Trypsin was from Thermo Scientific (Germany). Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), penicillin (10,000 units/mL) and streptomycin (10,000 μg/mL) were from PAA (Pasching, Austria).

DNA and RNA extraction kits, QIAshredder and RNeasy respectively, were from Qiagen (Germany).

All other solvents and chemicals were of analytical or liquid chromatography grade and were obtained from a variety of sources.

Biosynthesis of labelled CoQ10 in cultured fibroblasts

Our method was based on the previously reported method for radiolabelled substrates (Tekle et al 2008). Skin fibroblasts were grown in DMEM containing 10 % FBS and 1 % penicillin-streptomycin. After culture, cells were rinsed with phosphate buffered saline (PBS), trypsinized, centrifuged for 10 min at 252×g and cultured again in 6 well plates. At 60–70 % confluence the medium was changed for medium containing 13C6-PHB or 2H3-MV, and incubated for 24, 48 or 72 h. 13C6-PHB was tested at 1.65 mmol/L and 3.3 mmol/L and 2H3-MV at 14 μmol/L, 28 μmol/L, 42 μmol/L, 56 μmol/L, 112 μmol/L, 140 μmol/L and 280 μmol/L. After incubation, cells were trypsinized and washed twice with saline. Pelleted-cells were resuspended with 300μL of a buffer solution containing 0.25 mmol/L sucrose, 2 mmol/L EDTA, 10 mmol/L Tris and 100 UI/mL heparin, pH 7,4, and sonicated twice for 5 s. These homogenates were used to determine CoQ10 biosynthesis, total protein and citrate synthase (CS) activity. For CoQ10 determination, 10 μL of CoQ9 (1 μM, as internal standard, IS), and 800_μL of methanol were added to 100 μL of homogenate. The results were expressed in nmol CoQ10/g protein or nmol CoQ10/Unit of citrate syntase (UCS).

Viability test

Viability tests were performed after 24, 48 and 72 h incubation with 13C6-PHB or 2H3-MV at the concentrations above mentioned. Cells were washed with PBS and were incubated 3 h with 100 μL of MTT solution (0.5 mg/mL MTT in PBS). The purple MTT-formazan products were dissolved in dimethyl sulfoxide and optical densities of the solutions were measured by absorbance at 570 nm in an ELISA plate reader. Cells treated with 0.02 % SDS were used as positive control. Untreated cells correspond to the negative control. Cell viability was expressed as the optical density ratio of the treated cells respect to the negative control (% of control). Experiments were performed in triplicate.

HPLC-MS/MS analysis

CoQ10 and 13C6-CoQ10 or 2H3–CoQ10 (the two forms of CoQ10 synthesized depending on whether the substrate is 13C6-PHB or 2H3-MV, respectively) were measured by HPLC-MS/MS, as described in Arias et al (2012). The HPLC (Alliance HT 2795, Waters) was equipped with a 2.1 × 50 mm Symmetry C18 HPLC column (3.5 μm particle size). The mobile phase consisted of 50 % methanol with 5 mM methylamine, 45 % 2-propanol and 5 % water acidified with formic acid (0.5 mL/L), at a flow rate of 0.2 mL/min and isocratic conditions. MS/MS analysis was performed in a Micromass Quattro micro™ (Waters/Micromass, Manchester, UK). The MS/MS was operated in the electrospray positive ion mode with CV and CE of 15 V and 20 eV respectively. The following multiple reaction monitoring (MRM) transitions were selected: m/z 900 > 203 and 897 > 197 for 13C6-CoQ10 or 2H3–CoQ10 respectively, 894 > 197 for the physiological CoQ10 and 826 > 197 for CoQ9 (internal standard). Dwell time for each transition was 200 ms and run-time was 16 min. Nitrogen (at flow rate of 50 L/h) and argon (adjusted to obtain a vacuum of 3 × 10−3 bar) were used as nebulising and collision gas, respectively.

The physiological content of CoQ10 for some fibroblast samples and for muscle tissue was determined by HPLC with electrochemical detection as previously described (Montero et al 2008).

Intra-assay precision (CV) was evaluated in six parallel analyses of the same cell culture. To establish the inter-assay variability, one cell line was independently analysed on six different days.

Citrate synthase and protein determinations

CS activity was measured spectrophotometrically according to the method described by Srere (1969), with 0.1 mM DTNB, 0.2 % Triton X100 and 30–50 μg protein in 500 μL total incubation volume. Proteins were quantified using Protein Assay kit (Bio-Rad Laboratories, EEUU) based on the Lowry method.

Subjects

Thirteen control fibroblast cell lines from the repository bank of our hospital were analysed to establish the reference values. In order to validate the methodology, we studied seven patients with a definite diagnosis and CoQ10 deficiency in fibroblasts, including one patient homozygous for a COQ2 (OMIM*609825) mutation (unpublished results) and six patients with other inborn errors of metabolism: multiple Acyl-CoA dehydrogenase deficiency (MADD; OMIM#231680), very long chain Acyl-CoA dehydrogenase deficiency (VLCADD; OMIM#201475), mitochondrial encephalopathy with complex III deficiency and a mtDNA mutation, and Niemann-Pick type C disease (NPC; OMIM#257220) (Table 1). Then, we investigated three further groups of patients (Table 2) with CoQ10 deficiency in fibroblasts (and in muscle in some cases) but still with no definite genetic diagnosis: Group 1: three patients with a single mutation in one COQ gene; Group 2: five patients responsive to CoQ10 supplementation, without mutations in the genes studied; Group 3: eight patients with CoQ10 deficiency, without further genetic studies or documented response to treatment. All cell lines were analysed in two independent experiments.

Table 1 CoQ10 biosynthesis in patients with definite diagnosis and CoQ10 deficiency in fibroblasts
Full size table
Table 2 CoQ10 biosynthesis in patients with CoQ10 deficiency in muscle or fibroblasts. Group 1: patients with a single detected mutation in a COQ gene. Group 2: patients with good clinical response to CoQ10. Group 3: patients with CoQ10 deficiency in fibroblasts, and no genetic studies or documented response to treatment
Full size table

Patients or parents provided informed consent. The study was approved by the Ethics Committee of the Hospital Clinic-Barcelona, Spain. All samples were obtained in accordance with the current revision of the Helsinki Declaration.

Statistical analysis

Statistical analysis was performed using the SPSS version 18.0.0 software. Kolmogorov-Smirnov test was used to check variables which were under a normal distribution. The reference range was calculated as the mean ± 2 standard deviations. Pearson test was applied to correlate CoQ10 biosynthesis between both substrates.

Genetic studies

Prior to the introduction of the present methodology to evaluate CoQ10 biosynthesis, incomplete studies of some COQ genes had been performed in some of the patients as described beneath. The continuation of those studies was conditioned to the demonstration of an altered biosynthesis in the patients.

Genomic DNA was extracted from blood, skin fibroblasts or tissues using standard protocols, and mutational screening of 13 COQ genes (PDSS1 (OMIM*607429), PDSS2 (OMIM*610564), COQ2 (OMIM*609825), COQ3 (OMIM*605196), COQ4 (OMIM*612898), COQ5, COQ6 (OMIM*614647), COQ8 (OMIM*606980), COQ9 (OMIM:*612837), ADCK1, ADCK2, ADCK4 and ADCK5 was performed using self-designed oligonucleotides.

Total RNA was isolated from cultured fibroblasts using QIAshredder and RNeasy kits and cDNA was synthesized using standard protocols. We also isolated RNA from patients’ fibroblasts that had been treated during 7 h with 500 μg/mL CHX, in order to inhibit possible mRNA degradation by nonsense-mediated decay (NMD). Overlapping segments of the COQ8, PDSS1, PDSS2 and COQ4 cDNAs were PCR amplified and sequenced.

Patients from group 1 (Table 2) were studied for the following genes: PDSS1, PDSS2 and, COQ2-COQ9 in patient 8, and PDSS1, PDSS2, COQ2, COQ4, COQ5, COQ8 and COQ9 in patient 9. In addition, cDNA mutational screening was performed for the mentioned four genes in patients 9 and 10.

Concerning patients of group 2, the 13 mentioned genes were screened in their genomic DNAs.

Results

Method setting up and validation

CoQ10 biosynthesis in control fibroblasts (Fig. 2a) increased linearly with time (24–72 h) using either 1.65 mmol/L or 3.3 mmol/L 13C6-PHB as precursor, and the synthesized 13C6-CoQ10 amounts were alike at both concentrations. When using 2H3-MV as precursor, 2H3-CoQ10 biosynthesis was also linear with time but increased with increasing 2H3-MV from 14 to 56 μmol/L (Fig. 2b). For higher concentrations (112 μmol/L, 140 μmol/L and 280 μmol/L) during 72 h incubation, CoQ10 biosynthesis decreased (Fig. 2c). After those results, the elected conditions were 1.65 mmol/L 13C6-PHB, 28 μmol/L 2H3-MV, and 72 h incubation.

Fig. 2
figure 2

CoQ10 biosynthesis using 13C6-PHB and 2H3-MV as substrates at different concentrations and periods of incubation. a 13C6-CoQ10 biosynthesis when using 13C6-PHB as the precursor, at 1.65 mmol/L or 3.3 mmol/L, it increases linearly with time. b 2H3− CoQ10 biosynthesis when using 2H3-MV as the precursor at concentrations 14 to 56 μmol/L, it increases linearly with time. c 2H3− CoQ10 biosynthesis with higher concentrations of 2H3-MV (28 μmol/L, to 280 μmol/L) and 72 h incubation; it decreases for concentrations greater than 56 μmol/L

Full size image

The quantity of cells grown in the wells and analysed for each experiment was always similar, which is reflected by the protein concentration measured in the preparations (0.63 ± 0.07 mg/mL, n = 102). Therefore, the assay conditions are comparable between cell lines.

Viability tests demonstrated that 2H3-MV at 28 μmol/L does not affect fibroblasts stability at any incubation time tested and neither was there effect on viability for 1.65 mmol/L 13C6-PHB when the incubation time was 24 h or 48 h. And, although the viability decreased slightly (80 % residual) after 72 h incubation, the peak of 13C6-CoQ10 was threefold the LLOQ (S/N>10) (data not shown).

When incubating with 13C6-PHB, inter-assay variability (CV) of the newly synthesized 13C6-CoQ10 was 19 % if results were normalized to protein content and 10 % if normalized to UCS. When incubating with 2H3-MV, inter-assay CV of the newly synthesized 2H3-CoQ10 was 16 % and 13 % related to protein and CS, respectively. Concerning intra-assay CV, when incubating with 13C6-PHB it was 10 % when expressed per g protein and 9 % if expressed per UCS; and, when incubating with 2H3-MV, it was 12 % and 8 % per g protein and UCS, respectively. Due to the lower imprecision when normalizing the results to UCS we decided to use it instead of per protein content.

The mean and reference range for 13C6-CoQ10 and 2H3-CoQ10 biosynthesis were 0.94 (0.84–1.0) nmol/UCS and 0.13 (0.09–0.17) nmol/UCS, respectively (Table 1). The correlation between 13C6-CoQ10 and 2H3- CoQ10 biosynthesis in fibroblasts, in the whole group of controls and patients, showed that the two variables tend to increase together and 63 % of the variance was shared between them (Fig. 3c).

Fig. 3
figure 3

Graphic representation of CoQ10 biosynthesis in controls’ and patients’ fibroblasts using 13C6-PHB and 2H3-MV as substrates. a 13C6-CoQ10 biosynthesis. b 2H3-CoQ10 biosynthesis. c Correlation between 13C6-CoQ10 and 2H3-CoQ10 biosynthesis in fibroblasts (Pearson test, r = 0.63; p < 0.01). Numbers into the figure represent the corresponding patient in the tables

Full size image

Patients studied to validate the methodology are summarized in Table 1. Patient 1 shows significant CoQ10 deficiency in fibroblasts is homozygous for a mutation in COQ2 and presents deficient CoQ10 biosynthesis with both substrates (13C6-PHB and 2H3-MV). In contrast, patients 2–5, with CoQ10 deficiency in fibroblasts and diagnosis of other inborn errors of metabolism, showed normal biosynthesis.

Fibroblasts from patients 6 and 7, affected with NPC, showed deficient CoQ10 and normal biosynthesis with 2H3-MV as precursor, while the rate was decreased using 13C6-PHB as substrate (Table 1).

Patients’ results

The investigations in patients suspected of primary CoQ10 deficiency are summarized in Table 2, Fig. 3a and b. In group 1 only patient 8 (with a mutation in COQ4) showed deficient CoQ10 biosynthesis. Table 2 also shows the results for patient 8’s parents; her father (number 25) gave normal results for both substrates, while her mother’s (number24) biosynthesis was at the lower control range when the precursor was 13C6-PHB (0.84 nmol/UCS; controls 0.84–1.0 nmol/UCS).

Concerning group 2, all but patient 15 showed low CoQ10 biosynthesis, ranging from 0.04 to 0.09 nmol/UCS with 2H3-MV as substrate and from 0.55 to 0.78 nmol/UCS with 13C6-PHB (Table 2).

Finally, CoQ10 biosynthesis was in the normal range in four patients of group 3. In patients 16, 22 and 23, using 2H3-MV the amount of 2H3-CoQ10 synthesized was in the normal range, but with 13C6-PHB the biosynthesis was deficient (0.70, 0.65 and 0.64 nmol/UCS, respectively). The same happened with patient 17 though his CoQ10 biosynthesis with 13C6-PHB was only slightly reduced (0.82 nmol/UCS) (Fig. 3a).

Discussion

Method setting up and validation

We have developed a non-invasive, non-radioactive and sensitive method by HPLC-MS/MS to study the biosynthesis of CoQ10 in fibroblasts. Our results indicated that CoQ10 biosynthesis increased linearly with time for both substrates, but 2H3-MV needed higher concentrations. We finally established 1.65 mmol/L 13C6-PHB, 28 μmol/L 2H3-MV and 72 h of incubation. The concentration of 13C6-PHB used was as described by Tekle et al (2008) and for 2H3-MV it was doubled. Although the conditions for MV were not saturating they were judged adequate to obtain quantifiable peaks of the labelled CoQ10 with limited substrate costs and did not significantly affect cells’ viability. Correlation studies between 13C6-CoQ10 and 2H3-CoQ10 showed that the biosynthesis of both products tend to increase together (Fig. 3c), for that reason and because they measure different steps in the biosynthesis of CoQ10 we maintained the incubation with both substrates.

As expected for primary CoQ10 deficiency (Fig. 1), our results showed clearly decreased rates of CoQ biosynthesis in patient 1 either with 13C6-PHB or 2H3-MV as substrates (Table 1, Fig. 3a, b). Similar rates of CoQ10 biosynthesis were described for a patient harbouring a homozygous mutation in COQ2 with either 14C-PHB or 3H-decaprenyl-diphosphate (Quinzii et al 2006). As expected, the rates of biosynthesis were normal for patients 2–5 with CoQ10 deficiency secondary to other inborn errors of metabolism (MADD, VLCAD and complex III deficiency). MADD is caused by defects in components of flavin metabolism or transport (Frerman and Goodman 2001) that, in mitochondria, mediates the transfer of electrons from flavin to ubiquinone and the RC. Mutations in ETFDH, encoding for electron transfer flavoprotein-ubiquinone oxidoreductase (ETF-QO), have been reported to cause secondary CoQ10 deficiency (Gempel et al 2007; Liang et al 2009). We also studied fibroblasts from a patient with VLCADD. The association between VLCADD and CoQ10 deficiency had previously been noted in one patient (Laforêt et al 2009). Finally, some patients with mtDNA deletions or mt-tRNA point mutations may show secondary CoQ10 deficiencies (Rahman et al 2012; Sznajer et al 2007; Matsuoka et al 1991), as happens with patient 5, with deficient complex III and a homoplasmic mutation in mtDNA. All the mentioned diseases are mitochondrial dysfunctions that may hypothetically increase degradation of CoQ10 or decrease its ATP-dependent transport in some patients, but the actual mechanism of the secondary deficiency in all these conditions remains unknown (Rahman et al 2012).

Additionally, to validate our method, we have studied patients 6 and 7 affected with NPC (Macías-Vidal et al 2011). Previous authors (Schedin et al 1998) had observed that the accumulation of cholesterol in a NPC murine model was paralleled by increased dolichol-P and decreased CoQ10 concentrations. The mevalonate pathway involves condensation of three molecules of acetyl-CoA to 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA). HMG-CoA reductase then converts HMG-CoA to mevalonate (Fig. 1). Cholesterol regulates HMG-CoA reductase by feedback inhibition. In NPC, accumulation of cholesterol may down-regulate HMG-CoA reductase decreasing mevalonate and, consequently, CoQ10 synthesis (Turunen et al 2004). Accordingly, patients 6 and 7 showed normal CoQ10 biosynthesis with 2H3-MV as precursor of the pathway, while it was decreased using 13C6-PHB, as expected if the MV pathway was partially inhibited (Table 1, Fig. 3a, b). In these cases we tried to improve CoQ10 biosynthesis by adding non labelled mevalonate in the 13C6-PHB media but, surprisingly, no amelioration was obtained. Further experiments should investigate this lack of response.

Biosynthesis of CoQ10 in patients’ fibroblasts

As the previous results indicated that our method was suitable for recognizing alterations of CoQ10 biosynthesis, we applied it to different patients with decreased CoQ10 in fibroblasts (Table 2, Fig. 3a and b).

The three patients of group 1 carry a mutation in one allele in one of the COQ genes; we failed to detect a second mutation even after study of cDNA extracted from CHX treated fibroblasts. Only patient 8 (with a mutation in COQ4) showed deficient CoQ10 biosynthesis. Deficient CoQ10 content with diminished biosynthetic rate in cultured fibroblasts has been reported in a patient with haploinsufficiency of COQ4 (Salviati et al 2012). As the father of patient 8 also carries the mutation, we studied the biosynthesis in both parents. Results were normal for her father (number 25), while the mother’s rate (number 24) was at the lower control range with 13C6-PHB as substrate. Therefore, the COQ4 mutation in our patient may not be enough to reduce CoQ10 production. Our observations might be hypothetically explained by an additional maternal mutation in an unknown point of the pathway that lowers the rate of synthesis without clinical effect in heterozygotes. Consequently, patient 8 would carry two genetic alterations (the mutation in COQ4 and that hypothetical mutation) that together cause the deficiency. Whole exome sequencing is on course in this family. The other two patients of group 1 (patients 9 and 10) showed normal RC activities and CoQ biosynthesis. We may conclude that these patients’ CoQ10 deficiency is secondary and that their heterozygous mutations are not the cause of the deficiency.

Therefore, a normal CoQ10 biosynthesis rules out a primary defect but, as exemplified with patients of group 2, the opposite is not always true. In fact, all patients of this group but patient 15 showed deficient CoQ10 biosynthesis. However, screening for COQ genes failed to detect pathological changes. Patient 15 is the father of patient 11; they have previously been described (Pineda et al 2010; Artuch et al 2006). He presented slight clinical alterations including action tremor, mild modification of fluency, transient nystagmus and slight saccadic pursuit that were corrected with treatment. CoQ10 levels were low in his fibroblasts, but muscle CoQ10 and RC activities, as well as fibroblast CoQ10 biosynthesis, were normal. In contrast, his daughter (patient 11), with a more severe ataxia and biochemical alterations, showed clearly deficient CoQ10 biosynthesis (Artuch et al 2006). The disease in this family may be caused by a heterozygous mutation in an unidentified gene that causes mild alterations as seen in the father while the more severe disease in the girl is due to homozygosity (or compound heterozygosity). We cannot exclude that their CoQ10 deficiency is secondary and, in this case, it might only be a modulator of the phenotype, exacerbating the clinical picture in the daughter.

Four patients of group 3 presented CoQ10 biosynthesis in the normal range. Therefore, they are most probably secondary CoQ10 deficiencies. Conversely, patient 23’s biosynthesis was in the lower control range with 2H3-MV and decreased with 13C6-PHB. We could infer that this patient’s deficiency is primary, and mutations in COQ genes should be investigated. Results in patients 16 (Montero et al 2009), 17 and 22 are difficult to conclude because the amount of CoQ10 synthesized using 2H3-MV is in the normal range, but with 13C6-PHB, the biosynthesis is slightly low or deficient. This points to some altered or inhibited step in the biosynthesis of mevalonate, as observed in NPC disease. In fact, patient 22 has recently been diagnosed with neuronal ceroid lipofuscinosis type 2 (OMIM#204500). To our knowledge, the relationship between this disease and CoQ10 deficiency had not been reported previously and should be further investigated.

In conclusion, we have developed a non-invasive non-radioactive method suitable for the detection of defects in CoQ10 biosynthesis, which offers a good tool for the stratification of these treatable mitochondrial diseases. Additionally, our method might be of interest to study unknown aspects about the subcellular turnover of newly synthesized CoQ.