Coenzyme Q10 quantification in muscle, fibroblasts and cerebrospinal fluid by liquid chromatography/tandem mass spectrometry using a novel deuterated internal standard

Article abstract of DOI:10.1002/rcm.6529

RATIONALE Neurological dysfunction is common in primary coenzyme Q ₁₀ (2,3-dimethoxy, 5-methyl, 6-polyisoprene parabenzoquinone; CoQ₁₀; ubiquinone) deficiencies, the most readily treatable subgroup of mitochondrial disorders. Therapeutic benefit from CoQ₁₀ supplementation has also been noted in other neurodegenerative diseases. CoQ₁₀ can be measured by high-performance liquid chromatography (HPLC) in plasma, muscle or leucocytes; however, there is no reliable method to quantify CoQ₁₀ in cerebrospinal fluid (CSF). Additionally, many methods use CoQ₉, an endogenous ubiquinone in humans, as an internal standard. METHODS Deuterated CoQ₁₀ (d₆-CoQ₁₀) was synthesised by a novel, simple, method. Total CoQ₁₀ was measured by liquid chromatography/tandem mass spectrometry (LC/MS/MS) using d ₆-CoQ₁₀ as internal standard and 5 mM methylamine as an ion-pairing reagent. Chromatography was performed using a Hypsersil GOLD C4 column (150 × 3 mm, 3 μm). RESULTS CoQ₁₀ levels were linear over a concentration range of 0-200 nM (R² = 0.9995). The lower limit of detection was 2 nM. The inter-assay coefficient of variation (CV) was 3.6% (10 nM) and 4.3% (20 nM), and intra-assay CV 3.4% (10 nM) and 3.6% (20 nM). Reference ranges were established for CoQ₁₀ in CSF (5.7-8.7 nM; n = 17), fibroblasts (57.0-121.6 pmol/mg; n = 50) and muscle (187.3-430.1 pmol/mg; n = 15). CONCLUSIONS Use of d₆-CoQ₁₀ internal standard has enabled the development of a sensitive LC/MS/MS method to accurately determine total CoQ₁₀ levels. Clinical applications of CSF CoQ₁₀ determination include identification of patients with cerebral CoQ₁₀ deficiency, and monitoring CSF CoQ₁₀ levels following supplementation. Copyright 2013 John Wiley & Sons, Ltd. Copyright

Full text of DOI:10.1002/rcm.6529

Research Article
Received: 17 October 2012
Revised: 29 January 2013
Accepted: 31 January 2013
Published online in Wiley Online Library
Rapid Commun. Mass Spectrom. 2013, 27, 924–930
(wileyonlinelibrary.com) DOI: 10.1002/rcm.6529
Coenzyme Q10 quantification in muscle, fibroblasts and
cerebrospinal fluid by liquid chromatography/tandem mass
spectrometry using a novel deuterated internal standard
Kate E. C. Duberley1, Iain P. Hargreaves2, Korn-Anong Chaiwatanasirikul3,
Simon J. R. Heales2,4,5, John M. Land2, Shamima Rahman6,7,8, Kevin Mills5
and Simon Eaton3
*
1Department of Molecular Neuroscience, UCL Institute of Neurology, Queen Square, London WC1N 3BG, UK
2Neurometabolic Unit, National Hospital of Neurology and Neurosurgery, Queen Square, London WC1N 3BG, UK
3Unit of Paediatric Surgery, UCL Institute of Child Health, London WC1N 1EH, UK
4Department of Clinical Pathology, Great Ormond Street Hospital for Children, London WC1N 3JH, UK
5Biochemistry Research Group, Clinical and Molecular Genetics Unit, UCL Institute of Child Health, London WC1N 1EH, UK
6Mitochondrial Research Group, Clinical and Molecular Genetics Unit, UCL Institute of Child Health, London WC1N 1EH, UK
7Metabolic Unit, Great Ormond Street Hospital, London WC1N 3JH, UK
8MRC Centre for Neuromuscular Diseases, National Hospital for Neurology and Neurosurgery, Queen Square, London WC1N
3BG, UK
RATIONALE: Neurological dysfunction is common in primary coenzyme Q10 (2,3-dimethoxy, 5-methyl, 6-polyisoprene
parabenzoquinone; CoQ10; ubiquinone) deficiencies, the most readily treatable subgroup of mitochondrial disorders.
Therapeutic benefit from CoQ10 supplementation has also been noted in other neurodegenerative diseases. CoQ10 can
be measured by high-performance liquid chromatography (HPLC) in plasma, muscle or leucocytes; however, there is
no reliable method to quantify CoQ10 in cerebrospinal fluid (CSF). Additionally, many methods use CoQ9, an
endogenous ubiquinone in humans, as an internal standard.
METHODS: Deuterated CoQ10 (d6-CoQ10) was synthesised by a novel, simple, method. Total CoQ10 was measured by
liquid chromatography/tandem mass spectrometry (LC/MS/MS) using d6-CoQ10 as internal standard and 5 mM
methylamine as an ion-pairing reagent. Chromatography was performed using a Hypsersil GOLD C4 column
(150 Â 3 mm, 3 mm).
RESULTS: CoQ10 levels were linear over a concentration range of 0–200 nM (R2 = 0.9995). The lower limit of detection
was 2 nM. The inter-assay coefficient of variation (CV) was 3.6% (10 nM) and 4.3% (20 nM), and intra-assay CV 3.4%
(10 nM) and 3.6% (20 nM). Reference ranges were established for CoQ10 in CSF (5.7–8.7 nM; n = 17), fibroblasts
(57.0–121.6 pmol/mg; n = 50) and muscle (187.3–430.1 pmol/mg; n = 15).
CONCLUSIONS: Use of d6-CoQ10 internal standard has enabled the development of a sensitive LC/MS/MS method
to accurately determine total CoQ10 levels. Clinical applications of CSF CoQ10 determination include identification
of patients with cerebral CoQ10 deficiency, and monitoring CSF CoQ10 levels following supplementation. Copyright
© 2013 John Wiley & Sons, Ltd.
Coenzyme Q10 (2,3-dimethoxy, 5-methyl, 6-polyisoprene
parabenzoquinone; CoQ10; ubiquinone) is a ubiquitous
lipophilic molecule. First proposed by Crane and colleagues
in 1957, the main function of CoQ10 is in the mitochondrial
electron transport chain (ETC).[1] It acts as an electron carrier,
transporting them to complex III (ubiquinol cytochrome c
reductase; EC 1.10.2.2).[2] CoQ10 is also a well-established
antioxidant in its reduced ubiquinol form.[3]
CoQ10 deficiency was first reported by Ogasahara in 1989 in
siblings with progressive muscle weakness, fatigue and central
nervous system dysfunction.[4] Primary CoQ10 deficiency is
usually inherited in an autosomal recessive manner, with
recessive mutations reported in six CoQ10 biosynthetic genes to
date: (A) decaprenyl-diphosphate synthase subunit 1 (PDSS1)[5]
and (B) decaprenyl-diphosphate synthase subunit 2 (PDSS2)[6]
together encode the two subunits of the COQ1 enzyme; (C)
aarF domain containing kinase 3 (ADCK3),[7,8] COQ9[9] and
COQ6[10] are thought to encode enzymes active in the distal
portion of the CoQ10 biosynthetic pathway; and COQ2[5] encodes
4-hydroxybenzoate polyprenyl transferase.[11] Neurological
accepting electrons derived from complex
I (reduced
nicotinamide adenine dinucleotide (NADH) ubiquinone
reductase; EC 1.6.5.3), complex II (succinate ubiquinone
reductase; EC 1.3.5.1) and the electron transfer flavoproteins,
* Correspondence to: S. Eaton, Unit of Paediatric Surgery, UCL
Institute of Child Health, London, WC1N 1EH, UK.
W-mail: s.eaton@ucl.ac.uk
Rapid Commun. Mass Spectrom. 2013, 27, 924–930
Copyright © 2013 John Wiley & Sons, Ltd.

K. E. C. Duberley et al.
involvement is frequently observed in CoQ10 deficiency.
The most widespread presentation of CoQ10 deficiency is the
ataxic phenotype with more than 30 cases having been
reported, typically having childhood onset and variably
associated with seizures.[7,8,12–14] This phenotype is most
commonly due to mutations in the ancestral kinase ADCK3,
which appears to modulate the COQ3 enzyme in the distal
portion of the pathway.[15]
preferred for MS. These compounds share the same chemical
structure as the analyte but differ in mass owing to the
incorporation of stable isotopes. This means the IS will
fragment in an analogous way to the analyte to produce a
characteristic product ion. Although isotopically labelled
CoQ10 has previously been synthesised[32,33] and used in
tandem mass spectrometric analyses of CoQ,[28] the syntheses
are difficult and none are commercially available. In this
paper we describe a novel simple synthesis of deuterated
CoQ10 (d6-CoQ10), and use this as an IS in a LC/MS/MS
assay which is suitable for analysis of total CoQ10 levels in
muscle homogenate, fibroblasts and CSF.
In addition, secondary CoQ10 deficiency may be observed
in oculomotor apraxia type
I
(AOA1),
a
form of
spinocerebellar ataxia caused by APTX mutations.[16,17]
CoQ10 supplementation has proved successful in the
treatment of Friedreich’s ataxia, where serum CoQ10 status
was found to be ‘the best predictor of a positive clinical
outcome.[18] The prevalence of neurological involvement in
CoQ10 deficiency highlights the need for a method suitable
for the assessment of cerebral CoQ10 status. Patients with
cerebral CoQ10 deficiency often present with normal
peripheral CoQ10 levels. In order to detect patients with a
defect it is necessary to examine cerebral CoQ10 levels,
as has been reported for other small molecules such as
5-methyltetrahydrofolate.[19] In view of the reported low
levels of CoQ10 detected in cerebrospinal fluid (CSF;
<20 nmol/L), a highly sensitive analytical method would
be required for this analysis.[20–22]
EXPERIMENTAL
Synthesis of deuterated CoQ10 (d6-CoQ10)
The two methoxy groups of CoQ10 can be exchanged with other
alcohols in alkaline solution. This has been exploited to
synthesise diethoxy-CoQ10[34] or dipropoxy-CoQ10[23] as internal
standards for HPLC. We used the same approach to replace the
two CH3O- groups of CoQ10 with two CD3O- groups, yielding
d6-CoQ10. Preliminary experiments attempting exchange of
CH3O- groups directly with a large excess of deuterated
methanol in alkaline solution yielded a mixture of d6-CoQ10 with
a major contaminant of unexchanged CoQ10 (results not shown),
so instead, we opted for a two-step synthesis, first synthesising
and purifying diethoxy-CoQ10, then exchanging the two
ethoxy groups with d3-methoxy groups and repurifying to yield
d6-CoQ10 which was substantially free from unlabelled CoQ10
and therefore suitable as an internal standard.
Initially, a solution of diethoxy-CoQ10 was synthesised, as
described by Edlund:[34] 100 mg CoQ10 (Sigma-Aldrich) was
dissolved in 1 mL hexane and diluted in 4 mL dry ethanol.
Sodium hydroxide (100 mL) in ethanol (40 g/L) was added
to the CoQ10 solution and was left for 30 min. The reaction
was then stopped with 100 mL glacial acetic acid. Hexane
(10 mL) was added and the solution was centrifuged at
1000 g for 5 min. The top (organic) phase was removed
and washed by adding 10 mL H2O and centrifuging at
1000 g for 5 min. This was repeated twice. The organic
phase was then dried down under nitrogen gas at 60 ꢀC
and reconstituted in 1 mL methanol. Semi-preparative
reversed-phase HPLC was then used to purify the
diethoxy-CoQ10 from unexchanged CoQ10 and the partially
exchanged monoethoxy, monomethoxy-CoQ10. A Hypersil
250 Â 10 mm HyperPrep HS C18 column (Thermo-Hypersil,
Runcorn, UK) plus a C18 guard column (Phꢀenomenex, UK)
were used at a flow rate of 3.2 mL/min (25 C). The mobile
phase was isocratic and was composed of methanol and
ethanol, both containing 50 mM ammonium acetate at a
ratio of 25:75 (v/v), respectively. The effluent was monitored
at 275 nm and fractions were collected using glass tubes.
We then synthesised d6-CoQ10 by utilising the backward
version of this reaction, converting diethoxy-CoQ10 into d6-
CoQ10. Purified diethoxy-CoQ10 (200 mL) was diluted in 800
mL hexane and d4-methanol (4 mL, CD3OD, 99.8% isotopic
purity, Cambridge Isotope Laboratories) was added to the
mixture. Sodium hydroxide (100 mL) dissolved in d4-methanol
(40 g/L) was added to the solution and left for 5 h at
room temperature. The reaction was stopped with 100 mL
Current analytical techniques employ high-performance
liquid chromatography (HPLC) with either ultraviolet (UV)
or electrochemical (EC) detection. UV detection utilises the
ability of the benzoquinone core to absorb UV light at
275 nm. This method has been widely used and is highly
reliable. However, it is time consuming, takes over 15 min
to analyse one sample, and is relatively insensitive.[23] EC
methods are also widely described, have previously been
utilised for measurement of CoQ10 in CSF, and can also be
used for the measurement of CoQ10/CoQ10H2 redox
state.[20–22,24] However, both UV and EC detection can
potentially be subject to interference, due to either electrical
interference or mobile phase contamination when measurements
are made close to the detection limit. A robust and sensitive
tandem mass spectrometry method (LC/MS/MS) would
ensure a fast, selective and highly sensitive method of total
CoQ10 quantification.
Various liquid chromatography/tandem mass spectrometry
(LC/MS/MS) methods for measurement of CoQ10 have been
described.[25–28] However, CoQ10 forms Na+ and K+ adducts
which can be difficult to fragment, thus limiting the sensitivity
and utility of these methods. This has been overcome by use of
methylamine to produce a dominant ion ([M + CH3NH3]+),
which fragments easily, and this significantly increases
sensitivity.[29] Other methods use atmospheric pressure
chemical ionisation (APCI), which is less prone to adduct
formation (reviewed by Barshop and Gangoiti and Hansen
et al. [30,31]) but may be less widely available than electrospray
ionisation (ESI).
Another problem with current mass spectrometric methods
has been selection of a suitable internal standard (IS). The ISs
currently used for CoQ10 analysis include ubiquinone
analogues such as CoQ9, which will be influenced by
endogenous CoQ9 in human tissue, and chemically
synthesised analogues such as CoQ11, diethoxy-CoQ10 and
dipropoxy-CoQ10. These are type 2 ISs, structurally related
to CoQ10 with similar chemical properties. Type 1 ISs are
Rapid Commun. Mass Spectrom. 2013, 27, 924–930
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

Coenzyme Q10 analysis by LC/MS/MS using d6-CoQ10 as internal standard
deuterium chloride. Hexane (10 mL) was added and the
solution was centrifuged at 1000 g for 5 min. The top (organic)
phase was removed and washed by adding 10 mL deuterium
oxide and centrifuging at 1000 g for 5 min. This was repeated
twice. The organic phase was then dried down under
nitrogen gas at 60 ꢀC and reconstituted in 1 mL methanol.
The d6-CoQ10 was then purified by semi-preparative HPLC
as described above for diethoxy-CoQ10.
Electrospray ionisation-MS/MS
ESI-MS/MS of CoQ10 and d6-CoQ10 was carried out using a
Quattro micro triple quadrupole mass spectrometer with a
2975 HPLC system (Waters, Manchester, UK). To optimise
analytical conditions, initially standards were directly infused
into the electrospray source via a 25 mm (i.d.) fused silica
transfer line by means of a syringe pump at a flow rate
of 40 mL/min. The instrument was operated in positive
ionisation mode. The capillary voltage was maintained at
3.38 kV with a cone voltage of 30 V. The sourceꢀtemperature
Preparation of standard solutions
ꢀ
was held constant at 150 C, desolvation at 350 C. Nitrogen
Stock solutions for CoQ10 and d6-CoQ10 were prepared to a final
concentration of 2 mM in ethanol/methanol (1:1). Samples were
quantified spectrophotometrically using the molar extinction
coefficient of 14.6 Â 103 M–1cm–1 (at 275 nm) for CoQ10.[35] The
CoQ10 stock solution was serially diluted in 50:50 (v/v)
methanol/ethanol to construct a calibration curve. Samples
were diluted to concentrations of 200, 100, 50, 20, 10, 8, 6, 4, 2
and 0 nM. Each sample then had 20 nM d6-CoQ10 added to it.
was used as the nebulising gas at a flow rate of 950 L/h
(desolvation) and 50 L/h (cone). The masses of all CoQ10
homologues were determined in full scan mode (m/z 500–
1000) of MS1. Product ions were determined in MS3 following
collision-induced dissociation using argon as the collision gas.
The optimum collision energy was determined to be 27 eV.
Data were acquired in multiple reaction monitoring (MRM)
mode, with a dwell time for each ion species of 50 ms.
For quantitative analyses samples were separated using
HPLC coupled to MS. The mobile phases consisted of (A)
methanol, (B) 4 mM ammonium acetate 0.1% formic acid and
(C) methanol/isopropanol/formic acid (45:55:0.5, v/v/v)
containing 5 mM methylamine as an ion-pairing reagent.[29]
The column was a 3 mM Hypersil GOLD C4 (150 mm 3 mm,
3 mm) with a GOLD C4 guard column (3 mm, 10 mm length,
Subjects
The method was evaluated using patient frozen skeletal
muscle, cultured skin fibroblast and CSF samples. Random
samples were obtained from patients attending the National
Hospital, Queen Square, and metabolic clinic at Great
Ormond Street Hospital, London, as part of diagnostic
investigations after informed consent. Anonymised control
skeletal muscle and cultured skin fibroblast samples had no
evidence of ETC deficiency and control CSF samples had no
evidence of CSF monoamine metabolite disorder.
ꢀ
3 mm) operated at 40 C at a flow rate of 0.4 mL/min. The
initial conditions were 50% A, 50% B for 2 min, followed
by a switch to 50% B, 50% C at 4 min and a gradient
to 100% C for a further 8 min, after which the column was
re-equilibrated under the starting conditions for a further
5 min before the next injection.
Fifty microlitres of sample was injected and CoQ10 and
d6-CoQ10 eluted at approximately 8 min (Fig. 1). MS/MS detection
was performed by selected reaction monitoring (SRM) from the
methylammonium adduct molecule ([M + CH3NH3]+).
Preparation of patient samples
Skeletal muscle samples (vastus lateralis) were prepared
according to the method of Duncan et al.[23] Fibroblasts
were grown to confluency in 25 cm2 flasks and harvested in
100 mL Hank’s balanced salt solution ready for analysis.
Protein content was analysed in skeletal muscle and
fibroblast samples according to the method by Lowry et al.[36]
Unfiltered CSF was used. The IS was added to the samples
(Table 1), which were then extracted twice with 700 mL
hexane/methanol (5:2, v/v). Following centrifugation at
14 000 rpm for 3 min the upper organic layer was retained.
Due to the low CoQ10 concentration in CSF, samples were
extracted one further time to increase recovery (Table 1).
Samples were then evaporated to dryness using a rotary
concentrator and resuspended in 50:50 methanol/ethanol
(final volume shown in Table 1).
RESULTS AND DISCUSSION
Deuterated internal standard
As described by Teshima and Kondo,[29] use of a methylamine
adduct yields a precursor ion of m/z 895 for CoQ10 (Fig. 2(a))
and a product benzoquinone ring ion of m/z 197 (Fig. 2(a)).
d6-CoQ10 gave similar ions with the expected 6 Da difference
(precursor m/z 901 and product m/z 203; Fig. 2(b)), indicating
that d6-CoQ10 is an appropriate internal standard for the
measurement of CoQ10. d6-CoQ10 was 99.5% isotopically pure,
with only 0.5% contamination with CoQ10, demonstrating the
efficiency of the conversion and suitability for use as an internal
standard. Although we synthesised d6-CoQ10 using diethoxy-
CoQ10, substitution of the unlabelled methoxy groups with
other aliphatic groups, such as dipropoxy-CoQ10,[23] may give
an increased yield. We also used d4-methanol (CD3OD) in our
conversion, as this was available in our laboratory, although
d3-methanol (CD3OH) will also yield d6-CoQ10.
Table 1. Extraction procedure and internal standard (IS)
concentration
Number of hexane/
methanol extractions
(5:2, v/v, 700 mL)
Final
IS volume
(nM) (mL)
Linearity, precision, recovery
Muscle
Fibroblasts
CSF
2
2
3
200
50
10
300
300
70
A concentration curve over the range 0–200 nM CoQ10 (with a
d6-CoQ10 concentration of 20 nM) gave excellent linearity
(R2 = 0.9995; Fig. 3(A)). At lower concentrations (0–10 nM),
wileyonlinelibrary.com/journal/rcm
Copyright © 2013 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 924–930

K. E. C. Duberley et al.
Figure 1. Chromatography of a control fibroblast sample. (a) CoQ10: product ion 197 m/z;
precursor ion 865m/z. (b) Deuterated CoQ10: product ion 203 m/z; precursor ion 901 m/z.
Figure 2. (a) CoQ10 and (b) d6-CoQ10 mass spectra: 895 m/z, methylamine CoQ10 adduct precursor
ion; 197 m/z CoQ10 product ion; 901 m/z methylamine d6-CoQ10 adduct precursor ion; 203 m/z d6-
CoQ10 product ion).
linearity was less (R2 = 0.9964; Fig. 3(B)), giving a detection
limit of 2 nM. The reproducibility of the method was very
good (Table 2). The precision of the method was assessed
by examining the inter- (between runs) and intra-assay
(within run) coefficient of variation (CV). The CSF was
pooled from 64 disease controls. The inter-assay CV was
calculated by comparing 10 pooled CSF repeats (one run)
over five separate days. The intra-assay CV was calculated
by comparing 10 repeats in the same run; an average was
then taken of each run’s intra-assay CV. The inter-assay CV
was 3.6% when 10 nM CoQ10 was added to pooled CSF
and 4.3% with addition of 20 nM CoQ10. Intra-assay CV
was 3.4% (10 nM) and 3.6% (20 nM). Accuracy of the method
was tested by measuring the concentration of a pooled CSF
sample before and after addition of 10 nM CoQ10. Pre-
addition concentration was 4.09 nM Æ 0.03 (standard
deviation; SD), post-addition was 14.04 nM Æ 0.61 (SD),
giving a recovery of 0.995.
Rapid Commun. Mass Spectrom. 2013, 27, 924–930
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

Coenzyme Q10 analysis by LC/MS/MS using d6-CoQ10 as internal standard
spectrophotometric enzyme assay analysis (Table 3; Fig. 4(a);
mean: 307.7 pmol/mg; 187.3–430.1 pmol/mg; mean (SE)
age, 24.5 (3.9) years; range, 0.5–59 years; ratio of males to
females, 7:6). These values are comparable with those reported
in the literature using an EC detection method (110–480 pmol/
mg protein; mean 231 pmol/mg).[38]
Analysis of CoQ10 in muscle homogenates by the MS/MS
method was compared with analysis of the same samples
by the UV detection method currently used for clinical
samples in our laboratory.[23] We compared five muscle
biopsy samples previously analysed with the UV detection
method with the described MS/MS method. Correlation
analysis revealed an R2 value of 0.9914, indicating a strong
positive correlation between methods.
Fibroblast and CSF analysis
We have also established a reference range for fibroblasts
using 50 samples (Table 3; Fig. 4(b); 57.0–121.6 pmol/mg;
mean (SE) age, 20.75 (1.4) years; range, 0.03–55 years; ratio
of males to females, 2:3) with a mean concentration of
89.3 pmol/mg. We have confirmed the potential application
of our method to detect primary CoQ10 deficiency in
fibroblasts by measurement of CoQ10 in fibroblasts from a
patient with established CoQ10 deficiency due to a mutation
in COQ9.[9,39] The CoQ10 level determined for this patient
was 13.17 pmol/mg; 14.7% of the mean control CoQ10
concentration for fibroblasts.
Figure 3. Concentration curves for CoQ10 standard: (A) 0–200
nM (R2 = 0.9995); (B) 0–10 nM (R2 = 0.9964).
The observed reference range is comparable to many observed
reference ranges reported in the literature (48–112 nmol/g and
56–184 nmol/mg);[5,38] however, reference ranges for fibroblasts
appear to be variable. In particular some reference ranges appear
to be lower (39–75 pmol/mg and 34–70.4 ng/mg).[4,23] This
difference may be due to differences in tissue culture media,
fibroblast passage number and/or the protein assay used, which
underlines the importance of using standard operating
procedures when harvesting fibroblast samples.
The limit of detection for the UV detection method is 6
nM,[23] making it unsuitable for CSF analysis. The higher
sensitivity seen with the MS/MS method has enabled
the detection of the low levels of CoQ10 in CSF, allowing
us to establish a reference range (Table 3; Fig. 4(c); 5.7–8.7
nM; mean 7.38 nM; mean (SE) age, 6.15 (0.5) years; range,
0.1–22 years; ratio of males to females, 10:7). There were
no age (R2 = 0.03; Spearman correlation: p = À0.17) or gender
(t-test: p = 0.98) effects on CSF CoQ10 concentration. Although
Isobe et al. measured CoQ10 (as oxidised and reduced CoQ10)
in CSF (filtered with a 10 000 MW cut-off) of adults using
EC detection, a reference range was not reported; from the
data reported in their papers the total CoQ10 concentration
Table 2. Inter- and intra-assay coefficient of variation (CV)
and spiking accuracy of the CoQ10 MS/MS method on
pooled control cerebrospinal fluid samples
Average intra-
assay CV (%)
Inter-assay
CV (%)
Spike
recovery
Low spike
(10 nM)
High spike
(20 nM)
3.43
3.55
3.55
4.26
0.995
—
Oxidation of CoQ10H2
Although the majority of CoQ10 is present in biological fluids in
the reduced form, CoQ10H2, the reduced form is completely
oxidised during storage and sample preparation unless
rigorous steps are taken to preserve the redox state.[24] We have
shown previously that the extraction method used leads to
complete oxidation of CoQ10.[36,37] In addition, we divided a
fibroblast pellet into two aliquots. One aliquot was extracted
using the normal extraction procedure, whereas the other was
treated with 0.1 mL of 6 mM ferric chloride in 0.24 M
hydrochloric acid[34] before extraction. There was no increase
in measured CoQ10 after FeCl3 treatment, verifying complete
oxidation of CoQ10H2 during the extraction procedure.
Table 3. Reference intervals for skeletal muscle, fibroblasts
and cerebrospinal fluid (CSF)
Reference
range
Mean
concentration
Muscle analysis
Muscle (pmol/mg; n = 15) 187.3–430.1
307.7
89.3
Fibroblasts (pmol/mg;
n = 50)
57.0–121.6
We obtained a reference range by analysis of 15 samples
taken from patients who were subsequently found to
have no biochemical evidence of ETC disorder following
CSF (nM; n = 17)
5.697–8.681
7.39
wileyonlinelibrary.com/journal/rcm
Copyright © 2013 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 924–930

K. E. C. Duberley et al.
Figure 4. CoQ10 levels in (a) skeletal muscle homogenate (n = 15), (b) fibroblasts
(n = 50), and (c) cerebrospinal fluid (CSF; n = 17).
of CSF appears to be approximately 2.8 nM.[20,21] It is possible
that the difference between their data and our measured CSF
CoQ10 concentration is due to their use of a 10 000 MW cut-off
filter. In a preliminary experiment we found that filtered CSF
had undetectable levels of CoQ10, suggesting that CSF CoQ10
may be associated with lipoproteins or other CSF proteins, as
in plasma.[37] In addition, the average age of the controls used
by Isobe et al.[20,21] was 67, so there may also be differences in
CoQ10 concentration between childhood/adolescence and the
more aged controls studied by Isobe et al.
[2] L. Ernster, G. Dallner. Biochemical, physiological and
medical aspects of ubiquinone function. Biochim. Biophys.
Acta 1995, 1271, 195.
[3] M. Turunen, J. Olsson, G. Dallner. Metabolism and function
of coenzyme Q. Biochim. Biophys. Acta Biomembranes 2004,
1660, 171.
[4] S. Ogasahara, AG Engel, D. Frens, D. Mack. Muscle coenzyme
Q deficiency in familial mitochondrial encephalomyopathy.
Proc. Natl. Acad. Sci. USA 1989, 86, 2379.
[5] J. Mollet, I. Giurgea, D. Schlemmer, et al. Prenyldiphosphate
synthase, subunit
1
(PDSS1) and
OH-benzoate
polyprenyltransferase (COQ2) mutations in ubiquinone
deficiency and oxidative phosphorylation disorders. J. Clin.
Invest. 2007, 117, 765.
CONCLUSIONS
[6] L.C. López, M. Schuelke, C.M. Quinzii, et al. Leigh
syndrome with nephropathy and CoQ10 deficiency due to
decaprenyl diphosphate synthase subunit
mutations. Am. J. Hum. Genet. 2006, 79, 1125.
2 (PDSS2)
We have devised a simple synthesis of d6-CoQ10, enabling us
to establish a method suitable for total CoQ10 quantification
by tandem mas spectrometry (MS/MS). The method is
applicable to muscle, fibroblasts and CSF and is advantageous
over several published CoQ10 methods in which CoQ9, an
endogenous ubiquinone to human tissue, is used as an IS. As
CoQ10H2 is oxidised to CoQ10 during sample preparation/
analysis, the method is not currently suitable to measure the
CoQ10/CoQ10H2 redox state. However, the method could
potentially be adapted to measure the CoQ10/CoQ10H2
redox state. The ability to assess CoQ10 levels in CSF will
potentially be of considerable clinical utility, both for diagnosis
(e.g., to identify patients with primary neurological CoQ10
deficiency), and for monitoring therapeutic CSF CoQ10
levels following supplementation in mitochondrial and
other neurodegenerative diseases.
[7] C. Lagier-Tourenne, M. Tazir, L. López, et al. ADCK3 , an
ancestral kinase, is mutated in a form of recessive ataxia
associated with coenzyme Q10 deficiency. J. Hum. Genet.
2008, March, 661.
[8] J. Mollet, A. Delahodde, V. Serre, et al. CABC1 gene
mutations cause ubiquinone deficiency with cerebellar
ataxia and seizures. J. Hum. Genet. 2008, 82, 623.
[9] A. Duncan, M. Bitner-Glindzicz, B. Meunier, et al. A
nonsense mutation in COQ9 causes autosomal-recessive
neonatal-onset primary coenzyme Q10 deficiency:
potentially treatable form of mitochondrial disease. Am. J.
Hum. Genet. 2009, 84, 558.
a
[10] S.F. Heeringa, G. Chernin, M. Chaki, et al. COQ6 mutations
in human patients produce nephrotic syndrome with
sensorineural deafness. J. Clin. Invest. 2011, 121, 2013.
[11] S. Rahman, C.F. Clarke, M. Hirano. 176th ENMC International
Workshop: Diagnosis and treatment of coenzyme Q(10)
deficiency. Neuromuscular Disord. 2012, 22, 76.
[12] O. Musumeci, A. Naini, A.E. Slonim, et al. Familial
cerebellar ataxia with muscle coenzyme Q10 deficiency.
Neurology 2001, 56, 849.
[13] C. Lamperti, A. Naini, M. Hirano, D. C. de Vivo, E. Bertini,
S. Servidei. Cerebellar ataxia and coenzyme Q10 deficiency.
Neurology 2003, 60, 1206.
REFERENCES
[1] F. Crane, Y. Hatefi, R. Lester, C. Widmer. Isolation of a
quinone from beef heart mitochondria. Biochim. Biophys.
Acta 1957, 25, 220.
Rapid Commun. Mass Spectrom. 2013, 27, 924–930
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

Coenzyme Q10 analysis by LC/MS/MS using d6-CoQ10 as internal standard
[14] R. Horvath, B. Czermin, S. Gulati, et al. Adult-onset
cerebellar ataxia due to mutations in CABC1/ADCK3.
J. Neurol. Neurosurg. Psych. 2012, 83, 174.
[15] A. Tauche, U. Krause-Buchholz, G. Rödel. Ubiquinone
biosynthesis in Saccharomyces cerevisiae: the molecular
organization of O-methylase Coq3p depends on Abc1p/
Coq8p. FEMS Yeast Res. 2008, 8, 1263.
[26] S. Takamiya, T. Matsui, H. Taka, K. Murayama, M. Matsuda,
T. Aoki. Free-living nematodes Caenorhabditis elegans possess
in their mitochondria an additional rhodoquinone, an
essential component of the eukaryotic fumarate reductase
system. Arch. Biochem. Biophys. 1999, 371, 284.
[27] L. Li, D. Pabbisetty, P. Carvalho, M. Avery, B. Avery.
Analysis of CoQ10 in rat serum by ultra-performance liquid
[16] I. Le Ber, O. Dubourg, J. Benoist, et al. Muscle coenzyme Q10
deficiencies in ataxia with oculomotor apraxia 1. Neurology
2007, 68, 295.
[17] C. Quinzii, G. Kattah, A. Naini, et al. Coenzyme Q
deficiency and cerebellar ataxia associated with an
aprataxin mutation. Neurology 2005, 64, 539.
chromatography
mass
spectrometry
after
oral
administration. J. Pharmaceut. Biomed. Anal. 2008, 46, 137.
[28] W.H. Schaefer, J.W. Lawrence, A.F. Loughlin, et al.
Evaluation of ubiquinone concentration and mitochondrial
function relative to cerivastatin-induced skeletal myopathy
in rats. Toxicol. Appl. Pharmacol. 2004, 194, 10.
[18] J.M. Cooper, L.V.P. Korlipara, P.E. Hart, J.L. Bradley,
H.V. Schapira. Coenzyme Q10 and vitamin E deficiency in
Friedreich’s ataxia: predictor of efficacy of vitamin E and
coenzyme Q10 therapy. Eur. J. Neurol. 2008, 15, 1371.
[19] K. Hyland, J. Shoffner, S. J. Heales. Cerebral folate
deficiency. J. Inherit. Metab. Dis. 2010, 33, 563.
[20] C. Isobe, T. Abe, Y. Terayama. Levels of reduced and
oxidized coenzyme Q-10 and 8-hydroxy-2’-deoxyguanosine
in the CSF of patients with Alzheimer’s disease demonstrate
that mitochondrial oxidative damage and/or oxidative
DNA damage contributes to the neurodegenerative process.
J. Neurol. 2010, 257, 399.
[29] K. Teshima, T. Kondo. Analytical method for ubiquinone-9
and ubiquinone-10 in rat tissues by liquid chromatography/
turbo ion spray tandem mass spectrometry with 1-alkylamine
as an additive to the mobile phase. Anal. Biochem. 2005,
338, 12.
[30] B. Barshop, J. Gangoiti. Analysis of coenzyme Q in human
blood and tissues. Mitochondrion 2007, 7(Suppl.), S89.
[31] G. Hansen, P. Christensen, E. Tüchsen, T. Lund. Sensitive
and selective analysis of coenzyme Q10 in human serum
by negative APCI LC-MS. The Analyst 2004, 129, 45.
[32] K. Hamamura, Y. Yamagishi, S. Kijima, T. Suzuki, Y. Mori,
T. Nakamura. Synthesis of deuterium-labelled coenzyme
Q10. J. Labelled Compd. Rad. 2002, 45, 831.
[21] C. Isobe, T. Abe, Y. Terayama. Increase in the oxidized/total
coenzyme Q-10 ratio in the cerebrospinal fluid of
Alzheimer’s disease patients. Dement. Geriatr. Cogn. Disord.
2009, 28, 449.
[33] K. Hamamura, I. Yamatsu, N. Minami, et al. Synthesis
of [3’-14C] coenzyme Q10. J. Labelled Compd. Rad. 2002,
45, 823.
[22] T. Murata, C. Ohtsuka, Y. Terayama. Increased mitochondrial
oxidative damage and oxidative DNA damage contributes to
the neurodegenerative process in sporadic amyotrophic
lateral sclerosis. Free Radical Res. 2008, 42, 221.
[34] P.O. Edlund. Determination of coenzyme Q10, a-tocopherol
and cholesterol in biological samples by coupled-column
liquid chromatography with coulometric and ultraviolet
detection. J. Chromatography 1988, 425, 87.
[23] A. Duncan, S. Heales, K. Mills, S. Eaton, J. Land, I.
Hargreaves. Determination of coenzyme Q10 status in
blood mononuclear cells, skeletal muscle, and plasma by
HPLC with di-propoxy-coenzyme Q10 as an internal
standard. Clin. Chem. 2005, 51, 2380.
[35] F.L. Crane, R. Barr. Determination of ubiquinones. Methods
Enzymol. 1971, 18, 137.
[36] O. Lowry, N. Rosebrough, A. Farr, R. Randall. Protein
measurement with the Folin phenol reagent. J. Biol. Chem.
1951, 193, 265.
[24] P.H. Tang, M.V. Miles, A. DeGrauw, A. Hershey, A. Pesce.
HPLC analysis of reduced and oxidized coenzyme Q10 in
human plasma. Clin. Chem. 2001, 47, 256.
[37] S. Eaton, R. Skinner, J.P. Hale, et al. Plasma coenzyme Q10 in
children and adolescents undergoing doxorubicin therapy.
Clin. Chim. Acta 2000, 302, 1.
[25] J. Ruiz-Jiménez, F. Priego-Capote, J.M. Mata-Granados,
J.M. Quesada, M.D. Luque de Castro. Determination of
the ubiquinol-10 and ubiquinone-10 (coenzyme Q10) in
human serum by liquid chromatography tandem mass
spectrometry to evaluate the oxidative stress. J. Chromatogr. A
2007, 1175, 242.
[38] R. Montero, M. Pineda, A. Aracil, et al. Clinical, biochemical
and molecular aspects of cerebellar ataxia and coenzyme
Q10 deficiency. Cerebellum 2007, 6, 118.
[39] S. Rahman, I. Hargreaves, P. Clayton, S. Heales. Neonatal
presentation of coenzyme Q10 deficiency. J. Pediatrics 2001,
139, 456.
wileyonlinelibrary.com/journal/rcm
Copyright © 2013 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 924–930