Bioscience Reports, Vol. 6, No. 9, 1986
Solvation Properties of Ubiquinone-lO in
Solvents of Different Polarity
Monica Ondarroa, Santosh K. Sharma and Peter J. Quinn
Received August 27, 1986
KEY WORDS: ubiquinone-lO;coenzymeQ; solubility.
The solvation properties of ubiquinone-10 and ubiquinol-10 in a wide variety of
solvents of polarity varying from alkanes to water are reported. Greatest solubility is
observed in solvents of intermediate polarity and particularly where low polarity is
combined with a pronounced tendency to interact with the benzoquinone substituent
of the ubiquinone molecule. This includes solvents like chloroform and benzene.
Ubiquinone-10 is somewhat less polar than ubiquinol-10 as judged by comparative
solubilities of the two molecules. Proton-NMR chemical shift measurements and
aggregation studies in selected solvents indicate that in ubiquinone-10 in the liquid
phase and in solution in hydrocarbons like dodecane the molecules have a preferred
association possibly involving stacking of the benzoquinone rings. Surface balance
studies indicated that the surface-active character of ubiquinone-10 is relatively weak
and only in a comparatively polar and highly structured solvent, formamide, was there
evidence of an effect on surface tension of the solvent. The critical micelle
concentratiom in this solvent was estimated to be about 5 #M on the basis of surface
tension measurements. Ubiquinone-10 is well known to form virtually insoluble
monolayers at the air/water interface. Studies of the partition of ubiquinone-10 in
binary mixtures of solvents suggest that the interaction of the benzoquinone ring
substituent with structured polar solvents is considerably weaker than the internal
cohesion between molecules of the solvent. No evidence on the basis of wide-angle Xray diffraction measurements was obtained to indicate that solvent molecules were a
component of the crystal lattice of ubiquinone-10 that had precipitated from solvent
mixtures.
Department of Biochemistry,King'sCollegeLondon,CampdenHill, LondonW8 7AH, UK.
783
0144-8463/86/0900-0783505.00/0 9 1986 Plenum Publishing Corporation
784
Ondarroa, Sharmaand Quinn
INTRODUCTION
Ubiquinones and related plastoquinones are the only lipid components of energytransducing electron transport chains. It is often assumed that these molecules,
because of their lipophilic character, are confined to a hydrophobic domain within the
mitochondrial or chloroplast membrane respectively. Transfer of electrons requires
contact between the components of the redox couple. The redox couples in which
ubiquinone is involved are Complexes I and II on the one hand and Complex III on the
other and interaction between these complexes and ubiquinone is required for electron
(and proton) translocation to take place. Although a special class of proteins has been'
proposed to interact specifically with ubiquinones and assist in their arrangement and
function in the mitochondrial membrane [1-3], identification of such Q-binding
proteins apart from components of the different Complexes themselves has not so far
been successful.
Information about the solvation properties of ubiquinones can provide a useful
insight into the partition of these molecules into phospholipid bilayer structures.
Lenaz and coworkers [4, 5] have used ultraviolet absorbance measurements in
solvents of ethanol, isooctane and aqueous buffers to characterise the polarity of the
environment about the chromophore. They observed that the polarity of the medium
has a pronounced effect on the absorbance spectrum [6]. Incorporation and partition
studies of ubiquinones into phospholipid bilayers and biomembranes indicated an
environment of the chromophore that was similar to isooctane [7-] and removed from
solvation by polar molecules.
The present study of the solvation properties of ubiquinone-10 and ubiquinol-10
is focused on the effects on chemical shifts of proton magnetic resonances at different
locations within the molecules and correlation with previous studies undertaken with
binary mixtures of ubiquinone-10 and phospholipids dispersed in aqueous systems
[8-]. Physical measurements were also performed to investigate the surface-active and
colligative properties of ubiquinone-10 in solvents of different polarity to determine
whether micelle-type structures could be formed under certain conditions. Finally
solvation effects and partition coefficients of ubiquinone-10 in an ethanol/dodecane
two-phase system has been examined as a model of the behaviour of more polarized
phospholipid/water systems.
MATERIALS AND METHODS
Ubiquinone-10 was a gift from Eisai Co. Ltd., Tokyo, Japan and was used as
supplied in its oxidized form. Ubiquinone-10 concentrations were determined in
ethanolic solution using an extinction coefficient at 275nm of 1.4 x 104 [9].
Ubiquinol-10 was obtained by reducing the ubiquinone with sodium dithionite as
described previously [10]. The reduced ubiquinone solution was divided into small
aliquots, the solvent removed under a stream of dry nitrogen and aliquots stored in
liquid nitrogen. Purity of the reduced compound was determined
spectrophotometrically [9]. Even with precautions, the ubiquinol was not stable and
SoIvation of Ubiquinone-I0
785
gradually oxidised to the quinone; this precluded a more comprehensive comparison
of the physical properties of the reduced and oxidised forms of the coenzyme. All
solvents were of the highest quality grade available. Dipalmitoylphosphatidylcholine
was purchased from Sigma, London. Ubiquinone-10 and phospholipid were mixed in
chloroform solution and the solvent evaporated under nitrogen prior to freeze-drying
for at least 10 hrs to remove residual traces of solvent. The lipids were dispersed in
D 2 0 at 60~ using a vortex mixer; at least 20 min mixing was required to achieve a
stable dispersion, The hydrated lipids were stored at 4~ for at least 1 hr before
measurement of N M R spectral parameters.
Solubility Measurements
Small volumes of solvent were added to a weighed amount of ubiquinone-10 in a
stoppered vial. Complete incorporation of solute into the solution was achieved by
brief bath sonication or moderate warming of the samples which were then
equilibrated for at least 24 hr prior to solubility measurements. Heating was avoided
with ubiquinol-10 and the determinations carried out immediately as these procedures
accelerate reoxidation of the unstable reduced state, The point at which precipitation
occurred was determined by visual observation of the turbidity in a light beam; this
concentration was recorded as the solubility limit. In some cases spectrophotometric
determinations of the concentration were required, The values quoted are averages of
several independent measurements,
Vapour Pressure Osmometry
The instrument used was the Mechrolab Vapour Pressure Osmometer Model
301A. The instrument provides a value of electrical resistance corresponding to a
temperature difference between two thermistor beads and has a significance equivalent
to boiling point elevation or freezing point depression, except that with vapour
pressure osmometry it corresponds to the thermostat temperature of measurement
(37~ in these experiments). Benzil used for calibrating the osmometer was twice
crystatlised from anhydrous ethanol and stored in a vacuum desiccator over P205
until required, Linde molecular sieve was used in the solvent chamber to ensure that
solvents were completely dry.
aH-NMR Experiments
1H-NMR spectra were recorded on a Nicolet N T 200 (200 MHz) spectrometer
equipped with Fourier transform facilities. 200 transients were accumulated using
pulse angles of 90 and 15~. Chemical shifts of proton resonances were referred to
external tetramethylsilane dissolved in deuterated benzene and placed in a sealed tube
mounted coaxially with the sample tube. Small aliquots of solvent ~1) were added
stepwise to an N M R tabe containing a weighed amount of ubiquinone-10 in the liquid
phase (above 50~ and mixed thoroughly to make the desired solvent/ubiquinone
molar ratio. In this way, two series of spectra were recorded at 51~ starting with the
786
Ondarroa, Sharma and Quinn
melt and continuing with the progressively diluted sample, one in chloroform and
another in n-dodecane.
Surface Tension Measurements
Surface tension measurements were made with a Du N o u y tensiometer
(Cambridge Instruments) using the ring detachment method at a temperature of 25~
Particle-Sizing Techniques
Two methods were employed to characterise the shape and size of particles in
suspension. A Malvern Autosizer (Malvern Instruments Ltd.) based on the principle of
light scattering, and a Coulter Counter (Model TA 11), a conductivity method, were
used. Solution samples were prepared by sonicating appropriate amounts of
ubiquinone in a Kerry sonicating bath in water phases. All measurements were carried
out at 20~
Wide Angle X-Ray Diffraction
Pure ubiquinone-10 and precipitates from ethanol/water solvent systems were
sealed in glass capillary tubes of 1 mm outside diameter and 0.01 mm wall thickness
(Glas, Berlin-West, W. Germany) under partial vacuum. Samples were exposed for
1 hr to X-rays produced in a Phillips Generator with a fixed copper anode and a finefocus system in a standard Debye-Scherrer powder camera. Measurements were all
performed at 20~
RESULTS
The solubility properties of ubiquinone-10 in solvents of different types is one way
of obtaining useful information relating to the phase behaviour of the coenzyme in the
lipid domain of biological membranes. The solubility limits of ubiquinone-10 and its
reduced form, ubiquinol-10, have been determined and the results are presented in
Table 1. The data have been arranged such that the solvents are listed in order of a
polarity parameter [11, 12]. It is recognised that other parameters are involved in
solute solvation and different parameters including dielectric constant and refractive
index are also given as these relate to particular solvation features. It can be seen from
Table 1 that solubility increases with increasing solvent polarity from alkanes to reach
a maximum solubility in solvents of intermediate polarity. The solubility in more polar
solvents decreases abruptly and is particularly low in solvents with a combined Hdonor and acceptor character. In terms of solvent classification the low solubility in
class 2 solvents might indicate that the polarity of the benzoquinone substituent is
sufficiently high to cause aggregation of the ubiquinone and shielding of the polar
groups away from the solvent molecules. Accordingly, solvation of the benzoquinone
would be necessary to achieve high solubility and it appears that diethyl ether and
Solvation of Ubiquinone-10
787
Table 1. Solubility limits of ubiquinone-10 and ubiquinol-10 in solvents listed according to increasing
order of polarity
Solvent
n-Dodecane
n-Undecane
n-Decane
n-Hexane
DiethyI ether
Benzene
Chloroform
Acetone
Ethanol
N,N-Dimethyl formamide
N-Methyl formamide
Formic acid
1,4-Butanediol
1,3-Butanediol
1,2-Propanediol
1,3-Propanediol
Formamide
2-Amino ethanol
Water
Dielectric P' (Polarity
constant
parameter
(25~
Ref. 11, 12)
2.0
2.0
2.0
1.9
4.3z
2.32
4.8Z
20.7
24.3
36.7
182.4
58.5
-28.8
32.0
35.0
109.5
-78.5
--0.1
2.8
2.7
4.1
5.1
4.3
6.4
6.0
--9.6
-10.2
R.I.
Solvent
(refractive group
index, 25~
(Ref. 13)
---1.375e
1.350
1.498
1.443
1.356
1.359
1.428
1.447
1.3712
-1.4402
-1.4402
1.447
1.454
1.333
2
2
2
2
2
2
3
3
3
3
3
1
1
1
1
1
1
1
1
Solubility
(20~
( m m o ldm-3) 1
Qox
653
653
653
1053
2303
> 270
> 390
40
5.8
40
2.3
Insoluble
Insoluble
Insohible
Insoluble
Insoluble
0.2
Insoluble
Insoluble
Qred
75
55
85
105
240
320
100
20
190
4.0
Insoluble
Insoluble
Insoluble
Insoluble
Insoluble
4.6
Insoluble
Insoluble
1 The molar concentration at which a saturated solution is formed at 20~ is calculated from the amount of
ubiquinone added to a given volume of solvent taking density of ubiquinone-10 as 1.
2 At 20~
3 These solubility values are approaching the saturation point.
benzene a m o n g s t this solvent class possess sufficient polarity to achieve this condition.
It is interesting to note that polarity is due to a hydrogen acceptor property in the case
of diethyl ether, a n d to its polarizability in the case of benzene. Greatest solubility,
however, is observed in chloroform, which is a class 3 solvent, a n d in which the
aggregation tendency of amphiphiles is low. The d o m i n a n t character of chloroform is
its h y d r o g e n d o n o r property. T a k i n g the three solvents together it appears that
solvation of the b e n z o q u i n o n e residue can be achieved by either h y d r o g e n acceptors or
d o n o r s or solvents with a dipole character b u t in addition, to solvate the whole
molecule, it is essential that the overall polarity of the solvent is low. As polarity of the
solvent increases the ability to solvate the extensive isoprenoid constituent decreases
a n d solubility is c o n s e q u e n t l y reduced. It is also n o t e w o r t h y that a m o n g s t the class 3
solvents the solubility of the reduced u b i q u i n o l - 1 0 relative to the oxidised u b i q u i n o n e 10 tends to increase as the polarity of the solvent increases. This reflects the m o r e polar
character of the reduced form of the q u i n o n e . This m a y also be related to the fact that
the more polar solvents of the class 3 type are strong hydrogen acceptors like N , N dimethylformamide, while the less polar like chloroform are strong hydrogen d o n o r s .
We examined the solvation effects of representative class 2 a n d 3 solvents using
p r o t o n - N M R methods. The effect of solvating u b i q u i n o n e - 1 0 by chloroform a n d
dodecane o n the different p r o t o n resonances referenced to the pure c o m p o u n d are
788
Ondarroa, Sharma and Quinn
"~ 250 I
(a)
2oor
r
~ 150g
g
Mean +_ S.D. of at[
other 1H-resonances
lOO-4,--
-EI'H 2_
13..
~
o
-0CH 3
I
I
I IH
/~
6
8 190
SoLvent/Ubiquinone-lO molar ratio
-50
-100 ~- (b)
Fig. 1. Variationin proton resonanceof ubiquinone-10melt with dilution in (a)
C2HC13 (up field shifts) and (b) dodecane (down field shift). Chemical shifts are
determined relative to pure ubiquinone-10. The ordinate shows the differencein
chemical shift between the resonancein the melt and the resonancein the diluted
sample. Positive differencesindicate an upfield shift and negative differences are
downfield shifts.
presented in Figure 1. Addition of dodecane to ubiquinone-10 causes a downfield shift
of the entire proton resonance spectrum such that there is no selectivity with regard to
particular chemical groups within the molecule. Solvation by chloroform, however,
indicates that the protons in the proximity of the benzoquinone substituent are affected
differently from the remainder of the molecule. One of the remarkable features of the
solvation by chloroform is the dramatic chemical shift in proton resonances associated
with the polyisoprene chain in the presence of 1 mole solvent per 10 moles of
ubiquinone. Since the dipole component of chloroform is relatively low it is unlikely
that this is due to a direct interaction of solvent with the non-polar chain. One
interpretation of this effect could be that the melt is not an isotropic liquid and that
Solvation of Ubiquinone-10
789
Table 2. Relative I chemical shift of proton resonances of ubiquinone-10 in different solvents
Resonance assignment
"Trans" terminal --CH 3
C3'--CH 3
Isoprenoid side chain
--CH z C2--CH 3
C3--CH 2 --C5,6--OCH 3 U
D
CI'--CH=
Isoprenoid side chain
--CH=
Melt Phospho- Dodecane Benzene
1.16 M 5 lipid
0.66 M 5 0.02 M
0.17 M 4
CC14
dil.2
CHC13 Ethanol
0.02 M
dil, 3
-0.14
-0.16
0.07
0.14
0.07
0.07
0.09
0.15
0.08
0.14
---
0.43
0,43
1.55
2.29
0.41
0.41
1.53
2,29
0.42
0.42
1.54
2.28
0.39
0.42
1.51
2.353
3.38
--
3,36
0.52
0,19
1.46
1.97
1.98
3.46
--
0.41
0.41
1.59
2.38
2.39
3.35
---2.35
2.36
--
3.53
35t
3,53
3,69
3.45
3.52
--
~H-NMR chemical shifts are determined relative to the isoprenoid side chain CH a. Spectra recorded at
25~
2 Ref. 16.
3 2.32 according to ref. 15.
4 Ubiquinone-10 in a proportion of 15 mole ~ocodispersed with dipalmitoylphosphatidylcholine at 10 ~ by
weight of phospholipid in D20. Spectrum recorded at 30~
5 Spectrum recorded at 50~
preferred associations between u b i q u i n o n e molecules are preserved in the liquid phase.
The s o l v a t i o n of the p o l a r g r o u p m a y cause d i s r u p t i o n of these associations a n d
different regions of u b i q u i n o n e are then able to interact. A n o t h e r feature t h a t is
a p p a r e n t from the p r o t o n N M R spectra of u b i q u i n o n e - 1 0 solubilized by c h l o r o f o r m is
that the - - O C H 3 resonances are split when the m o l a r p r o p o r t i o n of solvent to
u b i q u i n o n e - 1 0 reaches 5:1. This does n o t occur with d o d e c a n e . T h e s o l v a t i o n of
u b i q u i n o n e - 1 0 in m o n o m e r i c form by c h l o r o f o r m can be i n t e r p r e t e d on the basis of
s o l v a t i o n studies o f o t h e r lipids [-14]. It is also n o t e w o r t h y t h a t in c h l o r o f o r m the
- - O C H 3 resonances of u b i q u i n o l are n o t split [15].
W e e x a m i n e d the effects of o t h e r solvents on p r o t o n - N M R shifts a n d the d a t a are
s u m m a r i s e d in T a b l e 2. This shows t h a t solvents of the class 3 t y p e a n d also class 1
which are likely to interact with the p o l a r g r o u p of u b i q u i n o n e - 1 0 all cause a splitting
of the - - O C H 3 p r o t o n resonances while m e m b e r s of class 2 solvents like d o d e c a n e a n d
c a r b o n - t e t r a c h l o r i d e do not. A l t h o u g h benzene is a class 2 solvent its large d i p o l e
m o m e n t allows i n t e r a c t i o n with the b e n z o q u i n o n e substituent. This is also consistent
with a preferred a s s o c i a t i o n of the b e n z o q u i n o n e g r o u p s in the melt which is preserved
in the presence of class 2 solvents.
To d e t e r m i n e w h e t h e r u b i q u i n o n e - 1 0 forms aggregates or micelles in solvents of
differing p o l a r i t y three types of m e a s u r e m e n t were m a d e ; v a p o u r pressure o s m o m e t r y ,
surface tension m e a s u r e m e n t s a n d particle size analysis. V a p o u r pressure
m e a s u r e m e n t s were p e r f o r m e d on u b i q u i n o n e - 1 0 in solvents of increasing p o l a r i t y :
hexane < e t h a n o l < f o r m a m i d e . The d a t a o b t a i n e d were e v a l u a t e d using c o n c u r r e n t
c a l i b r a t i o n d a t a for benzil solutions in the a p p r o p r i a t e solvents. T h e d a t a for hexane
a n d e t h a n o l are presented in T a b l e 3 a n d show t h a t u b i q u i n o n e - 1 0 in hexane is a l m o s t
entirely in the form of m o n o m e r s in s o l u t i o n with very little t e n d e n c y for the molecules
790
Ondarroa, Sharma and Quinn
Table 3. Weight of particles in hexane and ethanol formed by ubiquinone-10
when present in varying concentrations calculated from vapour pressure
measurements obtained at 37~
n-hexane
ethanol
Cone. (raM)
M. Wt (g)
Cone. (mM)
M. Wt (g)
7,2
14.4
29.0
40.3
43.0
52.0
57.5
752
786
841
878
875
892
902
0.7
1.5
2.2
2.9
482
549
599
592
to aggregate even at relatively high concentrations. The 4 ~ increase in calculated
molecular weight in hexane is consistent with a weak interaction between ubiquinone
molecules in this solvent. The values obtained for ethanol are less than the calculated
molecular weight of ubiquinone-10 (863.4) and could be due to the presence of nonvolatile impurities in the soIvent. Despite strenuous efforts to obtain solvent free of
possible impurities, values less than the molecular weight of ubiquinone-10 were
invariably obtained. Attempts at measuring molecular weight of ubiquinone-10 in
formamide were also frustrated by a low solubility of, the lipid in the solvent such that
measurable vapour pressures could not be obtained. The surface-active properties of
ubiquinone-10 were examined in solvents of increasing polarity; dodecane < hexane <
ethanol < N,N-dimethyl formamide < N-methyl formamide < formamide. The only
solvent in which ubiquinone exhibited any surface activity was formamide and the
results are presented in Figure 2, This shows that with increasing ubiquinone-10
concentration there is a corresponding decrease in surface tension at the solvent-air
interface reaching a limiting value of about 46.5 m N . m - 1 . The concentration of
ubiquinone-10 at this point is about 5 ~tmoles. d m - a. This could be interpreted as the
X
60-
Z
E
tO
t-
t~
t~5
0
I
I
I
L
I
I
I
2
/+
6
8
10
12
1/+
Ubiquinone-lO concentrafion x 105 (mole x dm-3)
Fig. 2. Variation of surface tension with increasing ubiquinone-10
concentration in formamide.
Solvation of Ubiquinone-10
791
critical micelle concentration of ubiquinone-10 in formamide. Finally, the size
distribution of aggregates of ubiquinone-10 in water or dilute salt solutions were
investigated to determine whether a uniform aggregation of micellar-type structures
were present in highly polar solvents. Preliminary solubility experiments showed that
ubiquinone-10 was dispersed in water to an extent less than 5#g.m1-1. Results
obtained with highly sonicated suspensions of ubiquinone-10 in aqueous systems using
a Coulter Counter are shown in Figure 5. These show that the lipid is dispersed with a
range of particle sizes varying from about 7 to 30 #m in diameter with a skewed
distribution towards larger particle sizes and a median in the order of 20 #m. There was
a tendency to form particles of larger sizes with increasing concentrations of
ubiquinone-10. Filtration through a 0.22 #m filter and examination of the filtrate with
a Malvern particle Autosizer again showed a range of particle sizes in the 1 to 2 #m size
range which did not indicate the existence of a homogeneous population of micellarlike aggregates. Furthermore, it is clear that aggregation of particles had taken place
after filtration, since the particle diameter exceeded the exclusion limit of the filter.
More precise solubility properties were examined in ethanol-water solutions in
which the solvent polarity could be increased by increasing the proportion of water in
the system. The results shown in Figure 3 indicate a very marked decrease in solubility
of ubiquinone-10 when the proportion of water in the solvent mixture exceeds about
A275
---8
o
0
0
1.5
0
0
0.5
I
I
I
5
10
15
Volume % of H20 in efhano[
20
Fig. 3. Solubility of ubiquinone-lO in ethanol-water mixtures as
judged by absorbance measurements of the supernatant after
centrifugation.
792
Ondarroa, Sharma and Quinn
Fig. 4. Wide-angle X-ray diffraction patterns obtained from crystalline
ubiquinone-10 (a) and the crystalline material precipitating from ethanol-water
mixtures (b).
10~o by volume. The solutions were centrifuged after equilibration for 24 hr to
determine the concentration of ubiquinone in solution. Centrifugation again after
standing for a total of 48 hr did not result in further sedimentation of ubiquinone-10.
Light scattering measurements (data not shown) essentially confirmed the data
obtained by spectrophotometric determination of aggregated ubiquinone-10.
To determine whether the ubiquinone-10 in the material precipitating from
ethanol/water mixtures contained solvent molecules, samples of the precipitate were
examined by wide-angle X-ray methods. A typical diffraction pattern of the precipitate
is presented in Figure 4. The wide-angle diffraction maxima are presented in Table 4. It
can be seen that these spacings are almost identical to those obtained from the pure
crystalline compound. Other X-ray measurements of ubiquinone-10 precipitating
from aqueous systems also indicated no change in the spacings of the molecules in the
precipitating material to suggest the presence of solvating molecules in the crystal
lattice.
Table 4. Principal wide-angle X-ray diffraction spacings
of ubiquinone-10 crystals and aggregates formed on
precipitation from a 20~o aqueous ethanol solvent
Aggregates*
Crystals
0.480
0.396
0.303
0.283
0.240
0.467
0.383
0.293
0.273
0.234
* Means of two values in nm.
Solvation of Ubiquinone-10
793
O
15-
.-e
-,r - 10
tD"I
5
I
0
I
10
I
I
20
:30
/,0
Mean size of parfic[es (IJrn)
Fig. 5. Particle size distribution by weight ofubiquinone-10 aggregates dispersed
by ultrasonication in an aqueous medium. The mean diameter was 15.5 #m, s.d. 1.6
and surface volume mean diameter was 13.83 #m. The measurements were
performed by Coulter counter.
F u r t h e r a t t e m p t s to d e t e r m i n e w h e t h e r u b i q u i n o n e - 1 0 was s o l v a t e d b y p o l a r
molecules were u n d e r t a k e n using a t w o - p h a s e solvent system consisting of d o d e c a n e /
e t h a n o l , 45:55 (by vol.). The results of the presence of 10 a n d 30 m g of u b i q u i n o n e - 1 0
p e r ml of solvent m i x t u r e on the p a r t i t i o n of u b i q u i n o n e - 1 0 between the two phases
that form at equilibrium, on the v o l u m e r a t i o of the t w o o r g a n i c layers or the
c o n c e n t r a t i o n of e t h a n o l in t he d o d e c a n e layer are presented in Table 5. There were no
significant differences in the phase b e h a v i o u r of the solvent system due to the presence
of ubiquinone-10. The absence of increased e t h a n o l in the d o d e c a n e layer suggests that
e t h a n o l does n o t solvate the b e n z o q u i n o n e substituent in the h y d r o c a r b o n phase.
Table 5. Partition coefficients of ubiquinone-10 in an ethanol-dodecane two phase system and solvation
properties in the respective phases
Concentration
of ubiquinone-10
mg/ml solvent mixture
Partition
coefficient
Volume
ratio of
organic layers
Concentration of ethanol
in dodecane layer
( ~ by volume)
0
10
30
-3.70 ___0.48
3.42 • 0.19"
0.27 _+ 0.04
0.31 • 0.03
0.35 • 0.03*
0.84 _+0.11
1.83 _+0.06
1.85 • 0.11
Mean values + S.D. of 4 experiments.
* 3 experiments.
794
Ondarroa, Sharmaand Quinn
DISCUSSION
Examination of the chemistry of ubiquinone-10 indicates that this molecule is a
lipid and by operational definition is soluble in solvents of relatively low polarity. The
most efficient solvents, however, are not the least polar types like the alkanes but
solvents of intermediate polarity such as benzene and chloroform. Solvents of this type
are able to solvate both the benzoquinone substituent via polar interactions, but at the
same time can solvate the more hydrophobic polyisoprenoid chain. The importance of
solvent polarity is exemplified in particular by comparing ubiquinol-10 with
ubiquinone-10 in the relatively polar solvent, N,N-dimethyl formamide; the more
polar reduced form of the molecule is considerably more soluble in this solvent. In very
polar solvents like short-chain alcohols and water, solubility of the lipid decreases to
very low values as expected from the lipophilic nature of the molecule.
The factors that determine solubility at either end of the polarity scale are complex
and depend on a combination of three characteristics, namely, hydrogen donor or
acceptor strength and dipolar character. The low solubility in solvents of combined
hydrogen donor and acceptor character may be due to the fact that neither character is
dominant resulting in the absence of any strong interactions between ubiquinone-10
and solvent, a generally high polarity of the solvent, which does not solvate the
polyisoprenoid substituent, or a high degree of structure within the solvent.
A detailed picture appears to emerge from the proton NMR resonances of
ubiquinone-10 in different solvents which may be relevant to the type of interactions of
ubiquinone-10 in phospholipid bilayers. The chemical shift of the met hoxy protons are
equivalent in solvents of low polarity and they are chemically shifted with dilution to
the same extent as the remaining resonances in the molecule. In solvents of increasing
polarity the methoxy proton resonance is first shifted relative to the remaining proton
resonances as a single resonance to a position of lower field strength (carbon
tetrachloride). In solvents of even greater polarity the two methoxy groups become
non-equivalent and split into two resonances which occupy a similar position to the
single resonance observed in carbon tetrachloride. The fact that the --OCH 3
resonances are not split in ubiquinol-10 dissolved in chloroform [15] could be
explained because the solvent and reduced ubiquinol are both proton donating
compounds. Comparison of the methoxy proton resonances of pure ubiquinone-10 in
a melt indicates that the benzoquinone substituent is in a relatively non-polar
environment similar to that of dodecane. There are also no differences in chemical shift
of these protons in the melt compared with ubiquinone-10 codispersed with
phospholipid in water suggesting that if the trend seen in Table 2 can be extrapolated
to water the benzoquinone residues are not in contact with water or more polar
domains in the mixed dispersion. Again this is consistent with an interpretation of
highly sonicated dispersions of ubiquinone-10 and phospholipid in which two types of
methoxy proton resonances are observed [ 15]. The peak at lower field strength, on the
basis of the present solvent study, would represent a component of the molecule with
the benzoquinone substituent in a relatively polar domain of the system and conversely
with the other fraction with methoxy proton resonances at higher field strength. In
contrast to this fraction, the former component has been found to undergo reduction
Solvation of Ubiquinone-10
795
by water soluble reductants [-17], and is consistent with this argument.
The polarity of ubiquinone-10, as judged by surface-active properties in different
solvents, tends to be relatively weak and this may explain the reason why the molecule
is apparently not oriented in a phospholipid bilayer with the benzoquinone ring
substituent extending between the phospholipid molecules to the aqueous interface
of the bilayer and the polyisoprene chain extending into the hydrophobic interior. This
is believed to be how the shorter chain homologues (< ubiquinone-6) are arranged in
phospholipid bilayers [,18], and is readily explained in terms of the amphipathic
balance within the molecule. The more hydrophobic longer-chain homologues would
be expected to partition into a more hydrophobic domain.
There is accumulating evidence to suggest that the extent to which ubiquinone-10
can partition into phospholipid bilayers is limited [,19] and the amount incorporated
per phosphatidylcholine molecule depends markedly on the character of the: fatty acyl
chains. Thus amounts that can be incorporated appear to increase in the order
dipalmitoyl- > dimyristoyl- > egg phosphatidylcholine [20], but the reason for this is
presently obscure. The form in which ubiquinone-10 is incorporated is also unknown
but several possibilities have been considered [-21-23]. One is that ubiquinone
molecules are sandwiched between the bilayers of phospholipid and another in the
form of aggregates either fixed in position relative to the bilayer matrix or rotating
isotropically within the bilayer and extending across the structure.
Our studies to determine whether micellar forms of ubiquinone-10 form in
different solvents failed to provide evidence to support this arrangement. In general,
when solvation was low, crystalline ubiquinone-10 tended to precipitate out and
expulsion of solvent molecules from the crystal lattice was commonly observed. Even
solvation of the molecule in binary solvent systems like ethanol/dodecane did not
appear to take place probably due to weak interactions between the ubiquinone and
the respective solvent. It should, however, be emphasized that the environment created
by a phospholipid bilayer is quite exceptional in that a hydrocarbon domain of
thickness approaching the length of the ubiquinone-10 molecule is oriented between
aqueous interfaces. How this structural arrangement might influence the organization
of ubiquinone-10 has yet to be determined.
ACKNOWLEDGEMENTS
The work was aided by grants by the SERC and the British Heart Foundation.
Equipment used was purchased with funds provided by the Central Research Fund of
London University. MO was supported by a Spanish Ministry of Education and
Science scholarship.
REFERENCES
1. Yu, C. A., Nagoaka,S., Yu, L. and King,T. E. (1980). Arch. Biochem. Biophys. 204:59-70.
2. Yu, C. A. and Yu, L. (1980). Biochemistry 19:3579-3585.
3. Nagoaka,S., Yu, L. and King,T. E. (1981). Arch. Biochem. Biophys. 208:334-343.
796
Ondarroa, Sharma and Quinn
4. Lenaz, G., Degli-Esposti, M., Bertoli, E., Parenti-Castelli, G., Mascarello, S., Fato, R. and Casali, C.
(1982). In: Functions of Quinones in Energy Conservin 9 Systems (B. L. Trumpower, Ed.), Academic
Press, New York, pp. 111-124.
5. Lenaz, G. and Degli-Esposti, M. (1985). In: Coenzyme Q (G. Lenaz, Ed.), John Wiley, Chichester,
pp. 83-105.
6. Degli-Esposti, M., Ferri, E. and Lenaz, G. (1981). Ital. J. Biochem. 30:437-452.
7. Degli-Esposti, M., Bertoli, E., Parenti-Castelli, G., Fato, R., Mascarello, S. and Lenaz, G. (1981). Arch.
Biochem. Biophys. 210:21-32.
8. Ondarroa, M. and Quinn, P. J. (1986). Eur. J. Biochem. 155:353-361.
9. Lester, R. L., Hatefi, Y., Widmer, C. and Crane, F. L. (1959). Biochim. Biophys. Acta 33:169-185.
10. Rich, P. R. (1981). Biochim. Biophys. Acta 637:28-33.
11. Snyder, L. R. (1974). J. Chromatoor. 92:223-230.
12. Snyder, L. R. (1978). J. Chromatogr. Sei. 16:223-234.
13. Ray, A. (1971). Nature (London) 231:313-315.
14. Zahler, P. and Niggli, V. (1977). Meth. Membr. Biol. 8:1-50.
15. Kingsley, P. B. and Feigenson, G. W. (1981). Biochim. Biophys. Acta 635:602-618.
16. Gale, P. H., Arison, B. H., Trenner, N. R., Page, A. C. and Folkers, K. (1963). Biochemistry 2 : 196-200.
17. Ulrich, E. L., Girrin, M. E., Cramer, W. A. and Markley, J. L. (1985). Biochemistry 24:2501-2508.
18. Katsikas, H. and Quinn, P. J. (1982). Biochim. Biophys. Acta 689:363-369.
19. Stidham, M. A., Mclntosh, J. L. and Siedow, J. N. (1984). Biochim. Biophys. Acta 767:423-431.
20. Katsikas, H. and Quinn, P. J. (1982). Eur. J. Biochem. 124:165-169.
21. Quinn, P. J. (1980). Biochem. Int. 1:77-83.
22. Katsikas, H. and Quinn, P. J. (1981). FEBS Lett. 133:230-234.
23. Alonso, A., Gomez-Fernandez, J. C., Aranda, F. J., Belda, F. J. F. and Goni, F. M. (1981). FEBS Lett.
132 : 19-22.