|
|
|
|||||||
|
|||||||||
(Received for publication, October 6,
1994) From the
High resolution differential scanning calorimetric studies were
performed to investigate the thermotropic phase behavior of 26
molecular species of sn-1 saturated/sn-2
monounsaturated phosphatidylcholines. In parallel with calorimetric
studies, the energy-minimized structures and steric energies of the
diglyceride moieties of these monoenoic lipids were determined using a
molecular mechanics approach. The combined calorimetric and
computational studies led to the following results and conclusions. (i)
When a single cis-carbon-carbon double bond ( Most biological membranes contain a substantial amount of
glycerophospholipids loosely called phospholipids, and a remarkable
feature of these membrane phospholipids is their bewildering diversity.
This diversity originates largely from the various possible
combinations of different fatty acyl chains present in the hydrophobic
moiety of the amphipathic phospholipid molecule. Specifically, the
structures of the two fatty acids hydrolyzed enzymatically from animal
phospholipids display the following general features. (i) With the
possible exceptions of mammalian lung and nerve endings, which contain
large amounts of
dipalmitoylphosphatidylcholine(1, 2) , identical fatty
acids in both acyl positions occur rarely in a naturally occurring
phospholipid molecule. Instead, the two fatty acids have a different
even number of carbon atoms ranging from 14 to
22(3, 4) . (ii) The fatty acids in natural
phospholipids are nonrandomly distributed. The one that is esterified
to the hydroxyl group at glycerol backbone carbon 1, the sn-1
acyl chain, is often a saturated fatty acid, whereas the other, the sn-2 acyl chain, is predominantly an unsaturated
one(3) . (iii) The unsaturated fatty acids derived from the sn-2 acyl chains of major membrane phospholipids in animal
tissues have the structural formula
CH The role played by amphipathic
phosphatidylcholines in biological membranes has, until recently, been
viewed primarily as that of a structural component of biological
membranes in eukaryotic cells. This picture is now changed drastically,
since the breakdown components of phosphatidylcholine molecule, or the
intermediates of the ``phosphatidylcholine cycle,'' in the
cell membrane have turned out to be involved actively in signal
transduction in response to extracellular
stimuli(12, 13) . In addition, there is increasing
evidence suggesting that skeletal muscle phospholipids with
polyunsaturated sn-2 acyl chains may be of great significance
in modulating the insulin receptor activity(14, 15) .
In view of these recent exciting findings, it is important to study
systematically the various natural phosphatidylcholines with well
defined composition and to obtain information regarding the specific
contributions of the characteristic unsaturated acyl chains to the
structure and properties of the lipid bilayer comprised of natural
phosphatidylcholines. As a first step to investigate systematically
the physical properties of naturally occurring phosphatidylcholines
with mixed sn-1 saturated/sn-2 unsaturated acyl
chains, we have semisynthesized 26 molecular species of identical chain
and mixed chain phosphatidylcholines, each containing a single cis-double bond in the sn-2 acyl chain. In this paper
we report the phase transition characteristics of these molecular
species as determined by high resolution differential scanning
calorimetry (DSC). (
Prior to MM calculations, a set of torsion angles for a crude
representation of the molecule under study was first constructed. This
starting set of atomic coordinates was entered, via the MM2 program,
into the computer. An optimized structure corresponding to the minimum
steric energy (E It should
be emphasized that the absolute value of E The initial crude
structure of C(16):C(18:1 In this paper, attempts have been made to compare
the structures of various phosphatidylcholines. Since all of these
lipid molecules have the same head group, being different in their
diglyceride moieties only, we feel that it is justified to simplify our
MM calculations by considering the diglyceride moieties of these lipid
molecules only. Consequently, various sn-1
saturated/sn-2 unsaturated phosphatidylcholines without the
headgroups are used exclusively for our MM calculations. In this
approach, the headgroup-headgroup interactions are assumed to be
identical, and the headgroup-diglyceride interactions are assumed to be
small; hence, they are ignored when a relative comparison among
phosphatidylcholines is made. Furthermore, the water molecules are
considered to be excluded from the hydrophobic interior; hence,
Allinger's MM2 calculations of the diglyceride moiety of the
lipid molecule were performed in vacuo to avoid solvent
interactions. For dimers and tetramers, the MM calculations were
performed according to the same procedure as that used for the monomer,
essentially as described in detail elsewhere(19) . Again, these
aggregates are the aggregated diglyceride moieties of the lipid
molecules.
Figure 1:
Representative DSC heating thermograms
for C(18):C(18:1
It is interesting to recognize from Fig. 1that as the cis-double bond migrates from the
carboxyl end toward the chain methyl terminus, the phase transition
peaks are changed accordingly both in terms of the peak width and the
position of the maximal peak height (T
Figure 2:
T
The least squares fitting
curve in the T
Figure 3:
Representative DSC heating thermograms for
a series of C(X):C(20:1
Aqueous lipid dispersions prepared from four other
series of sn-1 saturated/sn-2 monounsaturated
phosphatidylcholines in which the sn-2 acyl chains are derived
from petroselenic, oleic, erucic, and nervonic acids have also been
studied calorimetrically. The thermodynamic data (T
Figure 4:
T
Upon a closer inspection of Fig. 4, the T The
Figure 5:
Fig. 6illustrates the
energy-minimized graphical structures of the diglyceride moieties of
C(18):C(18:1
Figure 6:
Energy-minimized structures of various
phosphatidylcholine monomers. Panels A and A`) are
two different representations of the same saturated
distearoylphosphatidylcholine. Panels B-G are monoenoic
lipids with the same molecular weight but with different position of
the
Fig. 7illustrates the two types of the energy-minimized
tetramers of the diglyceride moieties of
C(18):C(18:1
Figure 7:
Energy-minimized structures of tetrameric
C(18):C(18:1
To get into more details, the stabilization energy of the tetramer
contributed by the monomer ( The stabilization energy of the tetramer contributed by
the monomer (
Figure 8:
Plots of various calorimetric and
computational data versus the position of the
The first important overall result of our DSC studies is the
quantitative demonstration that a single In
parallel with calorimetric studies, our MM calculations provide
structural evidence indicating that the introduction of a The second important set of results from our
DSC studies indicates that the T Before interpreting the
V-shaped profile shown in Fig. 2, it is important to mention
that all of the DSC heating curves exhibited by aqueous lipid
dispersions of C(18):C(18:1 The lipid chain topologies obtained with MM calculations for a
series of positional isomers of C(18):C(18:1 The third important set of results of
our DSC studies is depicted in Fig. 4, in which the relationship
between the T Before we derive the relationship, it is important to
realize that the longer linear segment of the sn-2 acyl chain
can be either the upper or the lower linear segment separated by the
kink (Fig. 6). The two ends of the upper linear segment are both
covalently linked, whereas one end of the lower segment is not.
Consequently, the monoenoic lipid can be divided into two groups: group
I with a longer upper linear segment, and group II with a longer lower
linear segment. For instance, the top two curves in Fig. 4can be identified as lines connecting T Two structural parameters, N and
In Table 1, there are 12 molecular species of group I
C(X):C(Y:1
with Table 1contains 14 T
with With the only exception as C(20):
C(22:1
Volume 270,
Number 5,
Issue of February 3, 1995 pp. 2014-2023
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CALORIMETRIC STUDIES AND MOLECULAR MECHANICS SIMULATIONS (*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
) is
incorporated into a saturated diacylphosphatidylcholine molecule at any
position within the central segment of the long sn-2 acyl
chain, the resulting monoenoic lipid molecules will, in excess water,
exhibit reduced phase transition temperature (T) and transition enthalpy
(
H) as they undergo the gel to liquid-crystalline phase
transition. The T and
H-lowering effects of the bond can be attributed to
a decrease in the chain length of the sn-2 acyl chain, a
change in the chain length difference between the sn-1 and sn-2 acyl chains, and a local perturbation of the chain-chain
van der Waals interaction in the vicinity of the bond. (ii) For a
series of positional isomers of
1-stearoyl-2-cis-octadecenoylphosphatidylcholine,
C(18):C(18:1)PC, with a
bond at
different positions along the sn-2 acyl chain, the T value depends critically on the
position of the
bond. Specifically, the T value is minimal as the
bond is located at the
geometric center of the linear segment of the sn-2 acyl chain,
and the Tvalue is progressively
increased as the
bond migrates toward either end of the sn-2 acyl chain. (iii) The various monoenoic
phosphatidylcholines under study can be divided into two groups. The T values of most lipids in each group can
be correlated in an identical manner with their structural parameters,
yielding a commonT
-structure
relationship.
-(CH
) -
[CH=CH-CH
] -
(CH
) - COOH, where the subscripts
= 1, 4, 5, and 7, 
= 1-6, and
= 2-7, and the cis-double bonds (
=
1-6) are always separated by one methylene group(5) .
Because of the large number of possible ways that permutations of chain
length and unsaturation can occur, a given class of membrane
phospholipids such as phosphatidylcholine isolated from a given cell
type can be expected to give rise to a significant number of specific
molecular species. Indeed, 57 molecular species of phosphatidylcholines
have been detected in mammalian liver(6) . Probably because of
the large number of molecular species of natural phosphatidylcholines,
the physical properties of most membrane phosphatidylcholines with
mixed sn-1 saturated/sn-2 unsaturated acyl chains
have not yet been fully established. The thermotropic phase transitions
of phospholipid bilayers and their relevance to biological membranes,
for instance, have been investigated extensively for nearly 3
decades(7, 8) . Surprisingly, with the notable
exception of 1-palmitoyl-2-oleoylphosphatidylcholine (9, 10, 11) and a few other species (10, 11) , only limited information is available
regarding the phase transition behavior of most natural
phosphatidylcholines. Clearly, a significantly rapid progress in this
field is definitely needed.
)Moreover, molecular mechanics (MM)
calculations are applied to refine the three-dimensional structures and
to compute the steric energies of several series of
bond-containing phosphatidylcholine molecules which, in turn, are used
to account for the thermodynamic data. Finally, the phase transition
temperatures (T) associated with the gel
to liquid-crystalline phase transitions of bilayers prepared from two
groups of a total of 26 monoenoic phosphatidylcholines are analyzed in
terms of their structural parameters by a multiple regression approach,
leading to the derivation of two general equations. It is of the
greatest interest and significance that there is a common T
-structure relationship that holds for
most different monoenoic phosphatidylcholines within each group of the
natural lipids.
Chemicals
Monounsaturated fatty acids with
various chain lengths, each containing a cis-double bond at
different positions, were purchased from Sigma.
Lysophosphatidylcholines with various saturated acyl chain lengths were
provided by Avanti Polar Lipids, Inc. (Alabaster, AL). Silica Gel 60
(mesh 230-400) was obtained from EM Science (Gibbtown, NJ). All
chemicals and organic solvents were of reagent and spectroscopic
grades, respectively.Semisynthesis of sn-1 Saturated/sn-2 Monounsaturated
Phosphatidylcholines
Monounsaturated phosphatidylcholines were
semisynthesized at room temperature by acylation of CdCl adducts of lysophosphatidylcholine, in dry chloroform, with
monounsaturated fatty acid anhydride that was prepared in situ from fatty acid and dicyclohexylcarbodiimide, in the presence of
catalyst 4-pyrrolidinopyridine according to the modified procedure of
Mena and Djerassi (16) as described previously(17) .
The in situ reaction and the reacylation were carried out
under an N atmosphere to avoid the possible oxidation of
the unsaturated fatty acid. The synthesized sn-1
saturated/sn-2 monounsaturated phosphatidylcholines were
subsequently purified to about 99% purity by repeated column
chromatography on Silica Gel 60(17) .High Resolution DSC Measurements
Purified
monoenoic phosphatidylcholines were lyophilized from benzene before
weighing. The lyophilized sample was then dispersed in buffered (pH
7.4) aqueous solution containing 50 mM NaCl, 1 mM EDTA, and 5 mM phosphate buffer to give a final lipid
concentration in the range of 3.0-6.0 mM. The exact
lipid concentrations were determined by phosphorus analysis. Prior to
DSC experiments, the lipid sample was kept in the cold room (0 °C)
overnight under an N atmosphere. DSC experiments were
carried out with a high resolution Microcal calorimeter (model MC-2)
equipped with DA-2 digital interface and data acquisition utility for
automatic collection (Microcal, Northampton, MA). After loading the
lipid sample into the sample cell of the calorimeter, it was
preincubated in the calorimeter for a minimum of 60 min at the desired
temperature (usually 15 °C below the estimated T
) followed by the DSC heating scan. The scan was
terminated at about 15 °C above the estimated T. The sample was then allowed to equilibrate
thermally at the elevated temperature for 60 min prior to the cooling
scan. Each sample was scanned at least five times: three heating and
two cooling runs. In all of these experiments, a nominal scan rate of
15 °C/h was used. The T and the transition
enthalpy (H) were determined, as described
in(17) , from the second and third heating curves, and the
average values were reported for each sample.MM Calculations
The MM2 force field (Version 85),
written by Allinger (18) and supplied by Quantum Chemistry
Program Exchange (Department of Chemistry, Indiana University), was
chosen as the software for the MM calculations. All MM computations
were run on an IBM RS/6000 computer workstation. The structural data
from the output of the MM2 calculations were transferred and then
displayed in the form of various two-dimensional representations of the
three-dimensional molecular structure, which was accomplished by the
computer-assisted molecular graphics using the software package
HyperChem (Autodesk Inc., Sausalito, CA) performed on a 486 platform. 
) closest to the starting point
was subsequently determined for the molecule by MM calculations using a
set of potential energy functions (called force fields). These
calculations involved a series of repeated automatic cycles of the
second derivative Newton-Raphson method for energy minimizations. These
cycles of self-adjusted computations came to a halt as the steric
energy difference between two consecutive cycles was less than (0.00008
molecular weight) kcal/mol. For the MM2 program, the steric
energy has the following terms: E = E + E
+ E + E
+ E
+ E
, where E
= bond stretch energy, E
= bond angle bend energy, E = torsional energy, E
= bond
dipole interaction energy, E
= van der
Waals nonbonding interaction energy, E
=
coupling energy between bond stretching and bond bending.
is of
no great significance, since it depends on the software program used.
When the same MM2 program is adopted, however, the E values for a homologous series of positional isomers are quite
useful for making relative comparisons.
)PC was constructed based on
the energy-minimized structure of C(14):C(14)PC (19) , and the
torsion angles of oleate chain of cholesteryl oleate were determined by
x-ray diffraction(20) , as described previously(21) .
Briefly, the torsion angles of the sn-2 acyl chain of
C(16):C(18:1)PC starting from C(3) to the chain methyl
terminus were taken from the crystal data of the cholesteryl oleate,
and all other torsion angles of the lipid molecule were taken from the
energy-minimized structure of C(14):C(14)PC except that the sn-1 acyl chain was extended by two methylene units in the trans form. This assembled structure was taken as the crude
structure of C(16):C(18:1)PC. It was subsequently
subjected to MM calculations in which the internal coordinates were
modified by Allinger's MM2 program until a minimum in the
potential energy surface was reached. The resulting structure was taken
as the optimized or energy-minimized structure of
C(16):C(18:1)PC. The torsion angles of the various
bonds, most importantly those in the vicinity of the bond,
obtained with this optimized structure were used subsequently as the
starting atomic coordinates to search for minimal energy structures of
other monoenoic lipids with the bond at different positions in
the sn-2 acyl chain. For instance, the initial crude structure
of C(18):C(18:1)PC can be readily created by adding
two methylene units in the trans form into the sn-1
acyl chain of C(16):C(18:1
)PC and by moving the
bond to the position in the sn-2 acyl
chain. The refined three-dimensional energy-minimized structure of
C(18):C(18:1
)PC can then be obtained by subjecting
the crude structure to MM calculations using the MM2 program.
Similarly, the same protocol can be used to obtain energy-minimized
structures of other monoenoic phosphatidylcholines. However, when a
bond was introduced into the sn-2 acyl chain at an even
carbon atom such as C(18):C(18:1)PC, the initial
values of the two torsion angles immediately preceding the
bond
( and ) and those (`
and `
) immediately succeeding the bond in
the sequence
`
`
were assigned to be negative, although the absolute magnitudes of
these torsion angles were still taken from those of the
energy-minimized structure of C(16):C(18:1)PC. This
change in sign was simply to reflect the fact that even and odd carbons
in an all trans-acyl chains lie commonly on two opposite and
parallel planes.
Effect of the Position of cis-Double Bond on the Phase
Transition Behavior of
1-Stearoyl-2-octadecenoyl-sn-glycerol-3-phosphocholine
(C(18):C(18:1
The DSC heating thermograms
for aqueous dispersions of six isomers of
C(18):C(18:1
)PC))PC are presented in Fig. 1. These
positional isomers with the same molecular weight have a common sn-1 acyl chain. In addition, each has only one cis-carbon-carbon double bond () in the sn-2
octadecenoyl chain. However, the position of the double bond from the
carboxyl carbon of the acyl chain () is different,
with n = 6, 7, 9, 11, 12, and 13. Each DSC heating
thermogram shown in Fig. 1is characterized by a sharp, single,
nearly symmetrical, endothermic transition that is reproducible upon
subsequent reheatings. The observed transition is assigned as the gel
to liquid-crystalline phase transition or the chain melting transition.
The T and H associated with the
chain melting transition for various isomers of
C(18):C(18:1)PC, in excess water, are summarized in Table 1. The T values obtained with the
cooling exotherms (data not shown) of these lipid samples occur at
temperatures that are either identical to or slightly lower
(0.2-0.4 °C) than the T values obtained
with the heating endotherms.
)PC, C(18):C(18:1
)PC,
C(18):C(18:1)PC, C(18):C(18:1)PC,
C(18):C(18:1
)PC, and
C(18):C(18:1
)PC dispersions. The T
,
H, and
T values associated with the gel to
liquid-crystalline phase transition are given for each lipid sample.
All thermograms were the second DSC heating scans obtained at a
constant scan rate of 15 °C/h.
).
Specifically, the transition peak width at half-height
(T) is seen to increase from 0.3 to 1.1
°C as the position of the
bond () moves from
C(6) to C(13); the T value decreases steadily from
24.8 °C to a minimum value of 3.8 °C as the bond moves
along the chain from C(6) to C(11) and then increases progressively
from 3.8 to 15.9 °C as the bond moves further along the chain
from C(11) to C(13). Most interestingly, the T values in the Tversus plot shown in Fig. 2can be fitted by the least squares
procedure to a smooth parabolic curve with a correlation coefficient of
0.9951. In comparison, the H values are, within
experimental errors, relatively insensitive to the position of the
bond (Table 1), but a general trend similar to that of the T profile is evident.
versus
for
C(18):C(18:1
)PC. The
experimental values of T
are
indicated by open circles, and the solid line connecting the experimental data is the least squares line of a
parabola best fit for the data. For comparison, the dotted line connecting the T
values of
C(18):C(18:1
)PE (PE) obtained
earlier from this laboratory (23) is
included.
versus plot yields = 10 and T = 3.8 °C as the abscissa and ordinate of vertex,
respectively, for monoenoic C(18):C(18:1)PC (Fig. 2). The calorimetric T value for
aqueous dispersion of saturated C(18):C(18)PC is 55.3
°C(22) . The calculated T value for
C(18):C(18:1)PC based on the coordinates of vertex of
the best fit parabola shown in Fig. 2is 3.8 °C,
corresponding to a significant drop in absolute temperature of 15.7%.
However, as the
bond moves toward either end of the sn-2
acyl chain, the T-lowering effect is diminished
considerably. The identification of an inverted bell-shaped
relationship between the T values of monoenoic
phosphatidylcholines and the different positions of the cis-double bond along the sn-2 acyl chain should come
as no surprise, since a similar but not identical relationship with the
lowest T value occurring at the bond
position of C(10) has also been observed for the lipid bilayers of a
different series of chemically distinct
C(18):C(18:1)PE (23) . For comparison, the
statistical least squares line best fit to the calorimetric data
obtained with C(18):C(18:1)PE is presented in Fig. 2as a dotted line.Effect of Increasing sn-1 Acyl Chain Length on the Phase
Transition Behavior of Mixed Chain Monounsaturated
Phosphatidylcholines
We have semisynthesized five additional
series of sn-1 saturated/sn-2 monounsaturated
phosphatidylcholines, and the aqueous dispersions prepared from these
five series of lipids have been subsequently studied calorimetrically.
Within each of the five series, the sn-1 acyl chain length is
increased stepwise by two methylene units, but the chemical composition
of the sn-2 acyl chain is fixed. Fig. 3illustrates the
second DSC heating scans for aqueous dispersions of four lipid samples,
which represent one of the five series of monoenoic lipids under study.
In this series of examples, the sn-1 acyl chains of the four
lipids are derived from stearic, arachidic, behenic, and lignoceric
acids, respectively, whereas the sn-2 acyl chains are derived
commonly from gondoic acids or 11-eicosenoic acids
(20:1). Fig. 3shows clearly a distinct gel to
liquid-crystalline phase transition exhibited by each sample. In this
figure, the T
and H values
associated with the endothermic transition are observed to increase
progressively with increasing sn-1 acyl chain length. Among
the four lipid species shown in Fig. 3,
C(20):C(20:1)PC is the only one that has been studied
previously by DSC(24) . The T
and
H values obtained from Fig. 3for
C(20):C(20:1)PC are 19.8 °C and 7.5 kcal/mol,
respectively, which are comparable to those reported in the literature (24) .
)PC. As the sn-1 acyl chain length is increased successively by two
methylene units, the T
and
H values are also seen to increase progressively. The values of
H, T, and
T are given for each
sample.
, H, S, and
T) associated with the gel to
liquid-crystalline phase transition exhibited by aqueous dispersions
prepared individually from these lipids and lipids of the first series (Fig. 3) are listed in Table 1. In Fig. 4, the T
value is plotted against the total number of
carbon atoms in the saturated sn-1 acyl chain for each of
these monoenoic lipids. Clearly, the T value is
observed to increase with increasing total number of carbon atoms in
the sn-1 acyl chain for all five series of monoenoic lipids.
This general trend reflects that an increase in the sn-1 acyl
chain length is accompanied by an increase in the stability of the
lipid bilayer in the gel state.
values versus the total number of carbon atoms in the sn-1
acyl chain for monoenoic phosphatidylcholines. In this figure,
experimental data from five series of lipids each with a common
monounsaturated sn-2 acyl chain are plotted. The solid
lines connect all experimental T
values within each lipid series.
curves for
C(X):C(18:1)PC,
C(X):C(20:1)PC,
C(X):C(22:1
)PC, and
C(X):C(24:1
)PC show that the larger the
number of carbon atoms that can be found in the sn-2 acyl
chain, the higher the T
curve for a given lipid
series. However, the correlation between the T curve and the total number of carbon atoms in the sn-2
acyl chain, based on the shape of the T curve, is
not perfect. Furthermore, the T curve exhibited by
aqueous dispersions prepared from the series of
C(X):C(18:1)PC, shown in Fig. 4,
appears to be completely out of place among other T curves. Recall the chain melting studies of a homologous series
of lipids with different positions of the bond (Fig. 2);
the misplaced curve shown in Fig. 4is thus most likely the
result of the influence of the position of the bond. In the case
of C(X):C(18:1)PC, the bond is
positioned three C-C bond lengths away from C(10), the geometric center
of the linear segment of sn-2 acyl chain. The bond in
the series of C(X):C(18:1)PC, however, lies
commonly at the geometric center of their sn-2 acyl chains. In
fact, the other two series of lipids also have their bonds
located at or near the geometric center of the sn-2 acyl
chains. In summary, experimental data shown in Fig. 4indicate
that the T value of the lipid bilayer comprised of sn-1 saturated/sn-2 monounsaturated
phosphatidylcholines depends on the total number of carbon atoms in
both acyl chains as well as the position of the bond along the sn-2 acyl chain.H values exhibited by
various sn-1 saturated/sn-2 monounsaturated
phosphatidylcholines can be grouped according to the total number of
carbon atoms in the two acyl chains of the lipid structure. The results
are given in Table 2. They show that the H value
does not, within the large experimental error, vary greatly for each
member of the same group with a common molecular weight (or total
number of carbon atoms). As a result, the weight average transition
enthalpy (H) can be calculated from the
experimental data for each group of the sn-1
saturated/sn-2 monounsaturated phosphatidylcholines as shown
in Table 2. Comparing
H values among
different groups shows, however, that the
H for each group of monoenoic lipids can be best fit by a least
squares line (Fig. 5), yielding
H (kcal/mol) = 0.25
[C(X)+C(Y)] - 2.37, with a
correlation coefficient of 0.9952, where C(X) and
C(Y) are the total number of carbon atoms in the sn-1
and sn-2 acyl chains of the monoenoic lipid, respectively.
When C(X) = C(Y), this linear function can be
subtracted from another linear function derived from the experimental
H values obtained with lipid bilayers comprised of
saturated identical chain phosphatidylcholines (shown as a dashed
line in Fig. 5), leading to a new linear function (Fig. 5, inset) with the expression H (kcal/mol) = 0.23 [C(X) +
C(X)] - 5.26 = 0.46 C(X) -
5.26. Here, the term H specifies the decrease in
H as a cis-double bond is introduced into the
identical chain phosphatidylcholine's sn-2 acyl chain.
For instance, the experimental H value associated with
the gel to liquid-crystalline phase transition of C(18):C(18)PC
bilayers is 9.8 kcal/mol(25) ; the H value
can be calculated as 0.46 (18) - 5.26 = 3.0
kcal/mol. The H value associated with the
gel to liquid-crystalline phase transitions of
C(18):C(18:1
)PC bilayers (n = 6, 7, .
. 13) can thus be estimated as 9.8 - 3.0 = 6.8 kcal/mol.
It should be emphasized that this equation, H = 0.46 C(X) - 5.26, can be used not only to
estimate the weight average H value for the positional
isomers of monoenoic phosphatidylcholines, it also indicates that
H = 0 when the total number of carbon atoms
in the sn-1 or the sn-2 acyl chain of the identical
chain phosphatidylcholine approaches 11. This implies that short chain
C(11):C(11)PC bilayers are already quite disordered at T < T, and as a consequence the introduction of a cis-double bond into the sn-2 acyl chain will not
further disrupt and destabilize the gel state bilayer of C(11):C(11)PC.
Hversus the total number of carbon atoms in both acyl chains of monoenoic
lipids. The
H value is the weight-averaged
H value obtained with aqueous dispersions of monoenoic
lipids, C(X):C(Y:1)PC, with
the same molecular weight as shown in Table 2. The dashed
line connects the
H values obtained with aqueous
dispersions of saturated identical chain phosphatidylcholines,
C(X):C(X)PC, reported by Lewis et al.(25) . The difference in H values between
the dashed and solid lines (H) is
plotted against the total number of carbon atoms in both acyl chains of
phospholipids in the upper left
inset.
MM Computations of Monoenoic Series of
C(18):C(18:1
In a recent study, the energy-minimized structures of
C(16):C(18:1)PC with n = 6, 7, 9, 11, 12, and
13)PC with various motifs of the
crankshaft-like kink in the sn-2 acyl chain have been
identified(21) . Specifically, four general groups of the
crankshaft-like kinks have been derived from a common protocol for the sn-2 acyl chain of C(16):C(18:1)PC, at T < T, by molecular mechanics calculations
using Allinger's MM2 program. Within each group, four different
kink-forming sequences are known; only minor variations in E are detected among these four kink models within
each group. The torsion angles of the various bonds in the immediate
neighborhood of the bond, or the kink-forming sequence, for a
representative rotomer of C(16):C(18:1)PC from these
four groups are presented in Table 3. Most interestingly, the
energy-minimized structure of the C(16):C(18:1)PC
rotomer with the lowest steric energy of 22.78 kcal/mol, shown in Table 3, is identical to the one derived independently using the
torsion angles of the oleate chain of crystalline cholesteryl oleate
followed by energy minimization using Allinger's MM2 program as
described under ``Experimental Procedures.'' The atomic
coordinates of this lowest-energy structure have thus been used as the
basic set of structural data for constructing the initial
three-dimensional structure of all other related members of monoenoic
lipids as described under ``Experimental Procedures.'' Since
the basic set of atomic coordinates is derived from crystal data, the
computed energy-minimized structure presented in this paper should
represent the lipid's simulated equilibrium structure at T < T.
)PC with n = 6, 7, 9, 11,
12, and 13. For comparison, the energy-minimized graphical structure of
the diglyceride moiety of C(18):C(18)PC with all-trans-acyl
chains is also included in Fig. 6. In each of the graphical
structures, the long axis joining all even carbons in the
all-trans-sn-1 acyl chain is assigned as the x axis
with the origin of the Cartesian coordinates fixed at C(4) and the
positive direction of the x axis oriented toward the chain
methyl terminus. The y axis is oriented from C(4) toward the
midpoint of an imaginary line connecting C(3) and C(5) of the sn-1 acyl chain; the z axis is, then, perpendicular
to the plane of the paper pointing toward the viewer. Several distinct
features are revealed by the energy-minimized structures of monoenoic
rotomers. (i) The sn-2 acyl chains of these monoenoic lipids
are seen to have a crankshaft-like kink with average torsion angles for
carbon-carbon bonds around the bond (,
, , `, `) being
(66.0° ± 2.3, 134.7° ± 6.4, 0.3° ±
0.4, 155.4° ± 3.0, 176.5° ± 0.3). The values of
, , `, and
` are positive for lipids with the bond
(negative value) positioned at the odd carbon, and negative for those
with the bond (positive value) at the even carbon (Table 4). (ii) In each structural model, the zigzag planes of
the two segments of the sn-2 acyl chain separated by the kink
are nearly parallel to the x-z plane, which is virtually
perpendicular to the zigzag plane of the sn-1 acyl chain.
However, the long chain axis of the all-trans-sn-1 acyl chain
is nearly parallel to the chain axes of the two segments of the sn-2 acyl chain. The average spatial distance separating the
two terminal methyl groups along the x axis direction (C)
is 5.67 Å ± 0.3 or 4.46 ± 0.24 C-C bond lengths for
this series of C(18):C(18:1)PC. The corresponding
C value for saturated C(18):C(18)PC can be calculated from the MM2
program to be 4.64 Å or 3.66 C-C bond lengths. Hence, the
presence of a -containing kink shortens the effective chain length
of the sn-2 acyl chain by 1.03 Å or 0.81 C-C bond
lengths. (iii) The number of trans-C-C bonds in the upper
segment of the sn-2 acyl chain, C(18:1), is (n - 5). In fact, the total number of C-C bonds in the
upper segment is (n - 1). The first two C-C bonds are gauche, leading to a 90° bend in the sn-2 acyl
chain. The two C-C bonds ( and )
immediately preceding the bond are gauche and skew (Fig. 6). Consequently, the total number of trans-C-C bonds (180° ± 10) is (n -
1) - 2 - 2 = (n - 5). Similarly, it
can be shown that the total number of trans-C-C bonds in the
lower segment is (16 - n). These numbers,
(C-C) and
(C-C)
, are given in Fig. 6for each molecular species of the series of
C(18):C(18:1
)PC. (iv) The E for
C(18):C(18:1)PC, 23.75 kcal/mol, is the minimal value
among all E values calculated for this lipid
series, and the E of 24.55 kcal/mol for
C(18):C(18:1)PC is the maximal. The average value of E
is 24.19 ± 0.31 kcal/mol.
bond. The local sequence around the bond in the sn-2 acyl chain is expressed as gsst and the torsion angles (in degrees) of the various bonds in the
sequence are given under each energy-minimized structure after gsst, where g is the gauche bond or conformation, t is the trans conformation, and s is the skew conformation of
a C-C single bond. E denotes the steric
energy and
C the effective chain length difference between the two
acyl chains. (C-C) and
(C-C)
denote the trans-carbon-carbon bond lengths in the upper and lower
segments of the sn-2 acyl chain,
respectively.
MM Computations of the Aggregates of Tetrameric
C(18):C(18:1
Once the energy-minimized
structures of the diglyceride moieties of monoenoic
C(18):C(18:1)PC)PC were constructed as shown in Fig. 6, the Cartesian coordinates of these structures were then
chosen as the new starting points to yield the energy-minimized
tetramers with the front to back and the up and down
motifs(19) . The aim was to combine these two types of
energy-minimized tetramers in creating a simplified bilayer model in
which the lipid molecules were packed in an orthorhombic lattice in the
two-dimensional plane of the bilayer(19) . )PC. The F-B packing motif is
characterized by the superposition of two trans-bilayer dimers
stacked in an eclipsed position, allowing van der Waals contact
interactions to occur optimally between the two contiguous faces (Fig. 7A). The energy-minimized U-D tetramer is
illustrated in Fig. 7B, in which two trans-bilayer dimers, lying on a common plane, are aligned
side by side. In this packing motif, the dimer-dimer contact surface is
limited to the lateral chain-chain interaction. The steric energies
associated with the two types of tetrameric packings, after energy
minimization by the MM2 program are also given in Fig. 7as E
and E
, where the subscript t denotes the
tetramer and the superscripts F-B and U-D denote the F-B and U-D
packing motifs, respectively. It is apparent from Fig. 7that
the general structural features of monoenoic lipids seen in Fig. 6are basically preserved in the two tetrameric motifs.
)PC. Panel A, the tetramer with
an F-B motif. Panel B, the tetramer with a U-D motif. The
stabilization energy of the tetramer contributed by the monomer is
expressed as E. These two energy terms are
related to each other as follows: E = (E - 4E)/4, where E is the steric energy of the monomer,
corresponding to the E term seen in Fig. 6. Both E
and E values are obtained by Allinger's MM2
program.
E) has also been
determined by MM calculations using the MM2 program. In general, the
intermolecular interactions among four monomers within a tetramer are
intimately associated with the stability of the tetramer. If the
geometry of the monomer is such that extensive attractive interactions
can occur among them, then this tetramer can be expected to be stable
at equilibrium. The affinity of the four monomers for each other in the
aggregated tetramer is a measure of the stabilization energy of the
tetramer contributed by its constituent monomer, which can be
calculated from the relationship E = (E - 4E)/4. The magnitudes of
E and
E, given in Fig. 7, for
the F-B and U-D motifs, are -19.32 and -8.75 kcal/mol,
respectively. Clearly, the tetrameric C(18):C(18:1
)PC
with an F-B packing motif is a more stable structure. If two tetramers,
one with an F-B motif and the other with a U-D motif, are closely
packed to mimic an orthorhombic two-dimensional lattice, then the
overall averaged stabilization energy,
E, is
(
E +
E) = -14.04
kcal/mol.
E +
E) and the overall averaged
stabilization energy (
E) for
C(18):C(18:1
)PC with n = 6, 7, 9, 11,
12, and 13 are listed in Table 5. Computational results indicate
that the calculated -E values for the simplified bilayer models of
C(18):C(18:1
)PC are in the range of 14.03-14.41
kcal/mol with a mean of 14.18 ± 0.22; the largest difference
between the mean and the calculated values amounts to only a few tenths
of a kcal/mol. It is important to realize that various approximations
have been made in our MM calculations in the determination of
-E values. Nevertheless,
with the exception of C(18):C(18:1
)PC, the
-
E values for all other
C(18):C(18:1
)PC shown in Table 5appear to
exhibit a V-shaped profile in the
-Eversus
plot, with the minimum at (Fig. 8C). Recall from Table 2that the
experimental
H values associated with the gel to
liquid-crystalline phase transitions of
C(18):C(18:1)PC bilayers are in the range of 6.0
± 0.7 to 7.1 ± 0.5 kcal/mol with a
H value of 6.6 ± 0.7 kcal/mol and
that the minimal
H value occurs at .
The calculated -
E values
resulting from MM computations thus harmonize with the experimental
H values obtained calorimetrically (Fig. 8, B and C), indicating that the energy contents of
C(18):C(18:1)PC bilayers in the gel state are not
greatly different and that the stability of the
C(18):C(18:1)PC bilayer appears to be the weakest.
bond in
C(18):C(18:1)PC. The various notations used
in the y axis of different plots are explained in the legends
of Fig. 1and Fig. 6and in the legend of Table 4.
Data in plots A and B are derived from DSC
experiments; plots C and D are from MM
calculations.
bond has the profound
effect of altering the gel to liquid-crystalline phase transition
behavior of fully hydrated phosphatidylcholines. In particular, when a
single bond is incorporated into a long sn-2 acyl chain
containing 18-24 carbons and is positioned at the carbon atom
numbering from C(6) to C(15), the T and H values associated with the phase transition are shown
calorimetrically to decrease significantly. For example, the T and H values for the C(18):C(18)PC
bilayer are, respectively, 55.3 °C and 9.8 kcal/mol(25) .
In contrast, the corresponding values for the
C(18):C(18:1)PC bilayer are 5.6 °C and 6.5
kcal/mol (Table 1). A reduction in the transition entropy
(S) of 6.6 cal/mol
K can thus be calculated.
These fairly pronounced T- and
H-lowering effects of the bond imply a decrease in
the overall stability of the monoenoic gel state bilayer. bond
into the sn-2 acyl chain allows the chain to kink in the shape
of a crankshaft at T < T ( Fig. 6and Fig. 7). This kink shortens the effective chain
length of the sn-2 acyl chain by about 1.03 Å, which
will undoubtedly diminish somewhat the stability of the gel state
bilayer. Furthermore, a shortening of the sn-2 acyl chain may
lead to an increase in the effective chain length difference between
the sn-1 and sn-2 acyl chains (C), which has
long been known to perturb the stability of the gel state
bilayer(17) . In addition, the nearly uniform chain-chain
separation distance is distorted by the kink; hence, favorable van der
Waals interactions between acyl chains must be reduced to some extent.
Our MM study thus indicates that a shortening of the sn-2 acyl
chain, a change in the C value, and a perturbation in the overall
chain-chain separation distance as induced by the bond are three
important factors that must be involved in the reduction in the overall
average stabilization energy of the lipid aggregates
(E) and also in the depression
of the
H value associated with the lipid phase
transition. In conclusion, it may be worthwhile to emphasize that when
a single bond is introduced locally into the sn-2 acyl
chain, there is a decrease in the overall chain length in the sn-2 acyl chain within the lipid molecule. Associated with
this chain length shortening, there are rotational isomerizations about
the C-C bonds in the vicinity of the bond which affect both the
intra- and intermolecular van der Waals interactions in the gel state
bilayer, thus producing a decrease in the stability of the entire lipid
molecule in the gel state bilayer at T < T. value exhibited
by bilayers prepared from C(18):C(18:1)PC depends
critically on the position of the bond in the sn-2 acyl
chain (). Notably, a V-shaped curve in the Tversus plot with the
minimal T value occurring at = 10 is observed (Fig. 2). This V-shaped profile is
in excellent agreement with the recent calorimetric data obtained with
aqueous lipid dispersions prepared from a series of sn-1
saturated/sn-2 monounsaturated
C(18):C(18:1)PE(23) . A similar V-shaped
profile has also been reported earlier by Barton and Gunstone (26) for a series of sn-1
monounsaturated/sn-2 monounsaturated
C(18:1):C(18:1)PC. However, for this
series of dioctadecenoylphosphatidylcholine, the minimal T value is observed when the bond in both
acyl chains lies in between C(9) and C(10).)PC arecharacterized by a
single, sharp, endothermic peak (Fig. 1). This common feature
can be reasonably taken as evidence to indicate that each endothermic
process is a two-state transition; hence, a simple equilibrium exists
between the two phases (gel and liquid-crystalline) at T. The T associated with the
gel to liquid-crystalline phase transition can thus be expressed by the
Clausius equality: T =
H/S. The variations in calorimetric
H values for aqueous dispersions of
C(18):C(18:1)PC are small with considerably large
errors ( Table 2and Fig. 8B). Nevertheless, these
experimental H values do exhibit a discernible trend
characterized by the V-shaped profile over a range of from n = 6 to n = 13 (Fig. 8B). On the other hand, the S values for the same series of C(18):C(18:1)PC are
essentially constant (Table 1). From this information and the
Clausius equality, the V-shaped T profile seen in Fig. 2and Fig. 8A for
C(18):C(18:1)PC can be simply interpreted by the
observed small variations in H as a function of
, which is also characterized by a V-shaped profile (Fig. 8B). This H profile is, by and
large, consistent with the
(-E) values calculated
for C(18):C(18:1
)PC (Fig. 8C). Based
on the calorimetric and computational data (Fig. 8, A-C), we can conclude that the V-shaped T profile of C(18):C(18:1)PC can be explained by a
systematic variation in the stability of the
C(18):C(18:1)PC bilayer, at T < T, as a function of . One
question can immediately be raised: What are the molecular origins of
the stability variations for the various gel state bilayers of
C(18):C(18:1)PC? A plausible answer to this question
seems to lie in the energy-minimized structures of
C(18):C(18:1)PC shown in Fig. 6and Fig. 7, which will be discussed in the following paragraph. )PC are
illustrated graphically in Fig. 6and Fig. 7. These
molecular structures show that the sn-2 acyl chain of each of
C(18):C(18:1)PC molecules is kinked. Comparisons of
kinked chains reveal that the two linear segments of the sn-2
acyl chain separated by the -containing kink have different
(C-C) values. The values of the longer segments,
(C-C)
, for
C(18):C(18:1
)PC are 10, 9, 7, 6, 7, and 8,
respectively, as n = 6, 7, 9, 11, 12, and 13 (Fig. 6). The relative percentage differences among these trans-C-C bond lengths are significant; moreover, the absolute
(C-C) value varies in an inverted
bell-shaped manner as a function of
, with the
smallest (C-C) value of 6 trans-C-C bond lengths identified for
C(18):C(18:1
)PC (Fig. 8D). If the
longer linear segment is assumed to pack more favorably with its
neighboring chains in the two-dimensional plane of the gel state
bilayer than does the shorter linear segment, it is then possible to
rationalize that the stability of the gel state bilayer will increase
with increasing length of the longer segment. Based on the preferential
interaction, the
H and hence the T values would be expected to display V-shaped profiles, each with
the minimum at . These expectations are of course
borne out by calorimetric data ( Table 1and Fig. 2). Thus,
we propose that the longer linear segment of the sn-2 acyl
chain separated by the
-containing kink will tend to pack more
effectively with its neighboring chains in the bilayer interior,
thereby changing the thermal stability of the bilayer according to the
length of the longer segment. values of lipid bilayers prepared
from different series of monoenoic phosphatidylcholines and the total
number of carbon atoms in the sn-1 acyl chains of these lipids
is presented. As mentioned under ``Results,'' the
calorimetric data shown in Fig. 4can be taken to indicate that
the T values of these lipids depend critically on
the total number of carbon atoms in both acyl chains as well as the
position of the bond along the sn-2 acyl chain. The
total number of carbon atoms in the two acyl chains of a phospholipid
molecule is related to the thickness of the hydrophobic core of the
lipid bilayer in the gel state (N). The effect of the cis-double bond position on the T has
just been proposed in the above paragraph to be related to the
(C-C) value. Earlier, we mentioned
that the chain length difference between the two acyl chains (
C)
can perturb the stability of the gel state bilayer. This is an
important term for sn-1 saturated/sn-2
monounsaturated phospholipids, since C remains unchanged for sn-1 monounsaturated/sn-2 monounsaturated
phospholipids or when both chains are unsaturated. Next, we want to
show that the T values of sn-1
saturated/sn-2 monounsaturated phosphatidylcholines can indeed
be correlated quantitatively with their structural characteristics in
terms of N, (C-C), and
C. values obtained with group I lipids
(C(X):C(24:1)PC and
C(X):C(22:1
)PC), the other three curves
connecting T
values derived from group II lipids
(C(X):C(18:1)PC,
C(X):C(20:1)PC, and
C(X):C(18:1
)PC). A structural parameter, N, can be introduced to specify the importance of
the (C-C) values as follows: N
= (X - 1) +
(C-C) for group I lipids and N
= (X - 1) +
(C-C) for group II lipids. The
values of (C-C)
and
(C-C)
for
C(X):C(Y:1
)PC can be calculated as
follows: (C-C) = n - 5 and (C-C)
= Y - n - 2. Hence, the structural
parameter N
specifies the sum of the trans-C-C bond lengths in the sn-1 acyl chain and the
longer segment of the sn-2 acyl chain for
C(X):C(Y:1)PC at T < T.C, have been used to correlate with the T value of saturated mixed chain phosphatidylcholines(22) .
The parameter N specifies the thickness of the hydrocarbon
core of the gel state bilayer, and the other, C, is the chain
length difference between the two acyl chains. For both parameters, the
unit is the C-C bond length along the long molecular axis running
perpendicular to the bilayer surface. For a saturated mixed chain
C(X):C(Y)PC, N and C are related to X and Y as follows: N = X + Y - 0.5, and C = X - Y + 1.5
(22) . In the case of
a monoenoic lipid C(X):C(Y:1
)PC, the
structural parameters N and C have slightly different expressions: N = X + Y - 1.5, and
C = X - Y + 2.5(23) .
Here, the shortening of the sn-2 acyl chain by one C-C bond
length caused by the bond has been taken into account in these
two expressions. The shortening of one C-C bond length is, in fact, an
average value, which can be estimated based on the C values and
their conformational weights (the P value given in Table 3) for all 16 possible kink models of
1-palmitoyl-2-oleoylphosphatidylcholine(21) .)PC with longer upper
segments in the sn-2 acyl chains. The T values of these monoenoic lipids are subjected to multiple
regression analysis in an attempt to establish a quantitative T-structure relationship. The statistical analysis
is conceptually similar to that established for saturated mixed chain
C(X):C(Y)PC(22) , except that an additional
structuralparameter, N, is introduced for
C(X):C(Y:1)PC. This analysis yields
= 0.9973 and the root mean square error
= 0.8269, where
is the correlation coefficient. The
calculated T
values based on for all
12 molecular species of group I lipids are given in Table 1. With
the only exception as C(20):C(22:1)PC, all of the
calculated T
values are within ± 0.8 °C
of the experiment alones. The calculated T value
for C(20):C(22:1)PC, however, is 2.4 °C smaller
than the experimentally observed value.
values determined calorimetrically for group II
monoenoic lipids. These values have also been subjected to multiple
regression analyses using various combinations of the three structural
parameters. gives the best
correlation
= 0.9914 and the root mean square error
= 0.9094. The calculated values for the 14 group II lipids are
presented in Table 1. With the exception of
C(24):C(18:1
)PC, the differences between the
experimental and calculated T values are all
within ±1.5 °C. The calculated T value
for C(24):C(18:1)PC is 1.8 °C smaller than the
experimentally observed T value. This largest
deviation amounts to only a relative error of 0.62% in absolute
temperature.)PC, the agreement between the calculated and
observed T
values for sn-1
saturated/sn-2 monounsaturated phosphatidylcholines shown in Table 1is quite reasonable. In the case of
C(20):C(22:1)PC, additional DSC experiments with
highly purified C(20):C(20:1
)PC will be required in
the future to clarify the disagreement. Despite this disagreement, it
is nevertheless of great significance that there is a
commonT
-structure relationship that holds for most
different monoenoic phosphatidylcholines within each group of sn-1 saturated/sn-2 mono unsaturated
phosphatidylcholines. This common relationship suggests strongly that
the various parameters characterizing the fundamental structures of
monoenoic lipids at T <T also contribute
collectively to the unique T values of these
lipids packed in the bilayer.
)H, transition enthalpy; E, the steric energy, in
kcal/mol, of an energy-minimized lipid species;
C(X):C(Y)PC, saturated phosphatidylcholine or
1,2-diacyl-sn-glycero-3-phosphocholine with X carbons
in the sn-1 and Y carbons in the sn-2 acyl
chain; C(X):C(Y:1)PC, monounsaturated
phosphatidylcholine with a saturated sn-1 acyl chain
containing X carbons and a monounsaturated sn-2 acyl
chain containing Y carbons with a cis-carbon-carbon
double bond () at the nth carbon atom from the carboxyl
end; T, transition peak width at
half-height; PE, phosphatidylethanolamine;
S, phase
transition entropy change; H, weight
average transition enthalpy; F-B, front to back; U-D, up to down.
We thank Professor T. E. Thompson for critical
comments on the manuscript.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C.-h. Huang, H.-n. Lin, S. Li, and G. Wang Influence of the Positions of cis Double Bonds in the sn-2-Acyl Chain of Phosphatidylethanolamine on the Bilayer's Melting Behavior J. Biol. Chem., August 29, 1997; 272(35): 21917 - 21926. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Wang, H.-n. Lin, S. Li, and C.-h. Huang Phosphatidylcholines with sn-1 Saturated and sn-2 cis-Monounsaturated Acyl Chains J. Biol. Chem., September 29, 1995; 270(39): 22738 - 22746. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |