| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Biophys J, March 2000, p. 1420-1430, Vol. 78, No. 3
and
*Dipartimento di Fisica and Unità INFM, Università
della Calabria, I-87036 Arcavacata di Rende (CS), Italy, and
Abteilung Spektroskopie, Max-Planck-Institut für
biophysikalische Chemie, D-37070 Gottingen, Germany
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Spin-label electron spin resonance (ESR) spectroscopy, together with optical density measurements, has been used to investigate, at both the molecular and supramolecular levels, the interactions of N-poly(ethylene glycol)-phosphatidylethanolamines (PEG-PE) with phosphatidylcholine (PC) in aqueous dispersions. PEG-PEs are micelle-forming hydrophilic polymer-grafted lipids that are used extensively for steric stabilization of PC liposomes to increase their lifetimes in the blood circulation. All lipids had dipalmitoyl (C16:0) chains, and the polymer polar group of the PEG-PE lipids had a mean molecular mass of either 350 or 2000 Da. PC/PEG-PE mixtures were investigated over the entire range of relative compositions. Spin-label ESR was used quantitatively to investigate bilayer-micelle conversion with increasing PEG-PE content by measurements at temperatures for which the bilayer membrane component of the mixture was in the gel phase. Both saturation transfer ESR and optical density measurements were used to obtain information on the dependence of lipid aggregate size on PEG-PE content. It is found that the stable state of lipid aggregation is strongly dependent not only on PEG-PE content but also on the size of the hydrophilic polar group. These biophysical properties may be used for optimized design of sterically stabilized liposomes.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Liposomes containing diacyl lipids with
sterically bulky hydrophilic polar heads are of particular interest
both for basic biomembrane research and for biotechnological
applications (Lasic, 1993
). For example, liposomes obtained by swelling
a variety of bilayer-forming lipids that contain a given proportion of
hydrophilic polymer-lipids (i.e., lipids with water-soluble polymers
covalently attached at the polar head) in water act as very effective
drug encapsulation and delivery systems (Blume and Cevc, 1990; Lasic, 1993; Lasic and Martin, 1995). In particular, it has been established that phosphatidylcholine (PC) host bilayer matrices containing phosphatidylethanolamine that has been derivatized by attachment of
poly(ethylene-glycol) polymers (PEG-PEs, with PEG molecular masses of
2000 and 5000 Da) have a blood circulation time from one to two orders
of magnitude longer (from a few hours to days) than conventional,
unprotected phospholipid liposomes (Blume and Cevc, 1990; Klibanov et
al., 1990; Allen et al., 1991; Papahadjopoulos et al., 1991). The
extended lifetimes in vivo arise from the steric barrier provided by
the grafted polymers that stabilizes the lipid bilayer against attack
by diverse elements of the immune system (Lasic et al., 1991a; Blume
and Cevc, 1993; Torchilin et al., 1994).
From a biophysical perspective, lipid/PEG-lipid/water dispersions have
been studied both experimentally (Lasic et al., 1991b; Hristova et al.,
1995; Kenworthy et al., 1995; Bedu-Addo et al., 1996; Baekmark et al.,
1997; Edwards et al. 1997; Belsito et al., 1998; Szleifer et al., 1998)
and theoretically (Hristova and Needham, 1995; Hristova et al., 1995;
Szleifer et al., 1998). An essential feature for a successful stealth
liposome delivery system is the optimization of the composition of the
liposomes, in particular with regard to the release of their content.
In this light it is useful to study the molecular interactions that
exist, and the dynamics of the aggregates that coassemble, when
micelle-forming lipids, such as PEG-lipids, are dispersed together with
bilayer-forming lipids in an aqueous environment. For the present work,
we have studied fully hydrated binary mixtures of ungrafted and
polymer-lipids with identical acyl chain compositions, namely
dipalmitoylphosphatidylcholine (DPPC) mixed with
dipalmitoylphosphatidylethanolamine (DPPE) bearing poly(ethylene
glycol)s of either low or intermediate average molecular masses
(PEG:350 or PEG:2000, respectively) at the polar head, over the entire
composition range of the two lipids from 0 to 100 mol%. This has been
done by using spectrophotometry at fixed wavelength, and both
conventional and saturation transfer electron spin resonance
spectroscopies (ESR and STESR) of spin-labeled phosphatidylcholine
having the nitroxide moiety at the C-5 or at the C-16 positions in the
sn-2 acyl chain (5- and 16-PCSL).
It is shown here that spin label ESR spectroscopy is particularly
helpful not only for investigating the rotational dynamics of the lipid
chains on different time scales (Thomas et al., 1976; Marsh, 1981,
1989; Hemminga and de Jager, 1989) but also for studying the micelle
formation that inevitably occurs in mixtures of polymer-lipids with
bilayer-forming phospholipids (Hristova and Needham, 1995; Hristova et
al., 1995). Because at low temperature the lamellar bilayer lipid
components are in the gel phase, whereas the micellar components of the
lipid mixture are in a fluid phase, these two environments are readily
resolved and quantified in the conventional spin-label ESR spectra. In
this way, it is found that the stable phases, lamellar or micellar, of
the mixed-lipid systems are affected critically, in a polymer chain
length-dependent manner, by the content of PEG-lipid in the
dispersions. These results have direct relevance both for the stability
of these liposomal formulations in serum, and for the ability of the
liposomes to release their contents in a controlled fashion on
interaction with cells.
| |
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Materials
The synthetic lipid 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) was from Sigma (St. Louis, MO). High-purity PEG-lipids 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-poly(ethylene glycol) with PEG of average molecular mass 350 or 2000 Da (PEG:350-DPPE and PEG:2000-DPPE, respectively), and the spin-labeled lipids 1-palmitoyl-2-(n-(4, 4-dimethyl-oxazolidine-N-oxyl)stearoyl)-sn-glycero-3-phosphocholine (n-PCSL with n = 5, 16) were from Avanti Polar Lipids (Birmingham, AL). The reagent grade salts for the 10 mM phosphate buffer solution (PBS) at pH 7.5 were from Merck (Darmstadt, Germany). All materials were used as purchased with no further purification. Distilled water was used throughout.
Spin-label ESR measurements
Samples for ESR measurement were prepared by dissolving the
required amounts of DPPC and PEG-DPPE, together with 1% by weight of
the spin-labeled lipid (n-PCSL), in chloroform. The solvent was evaporated in a nitrogen gas stream and then kept under vacuum overnight. The dried lipid samples were hydrated fully with PBS at pH
7.5 (final lipid concentration 25 mM), by heating at 60°C and
periodically vortexing for 40 min. The hydrated lipid dispersions were
then sealed in 1 mm (i.d.) 100-µl glass capillaries and then incubated for 24 h at 10°C before ESR measurements. ESR spectra were recorded on a 9-GHz Bruker (Karlsruhe, Germany) spectrometer (model ER 200D-SRC) and digitized with the spectrometer's built-in microcomputer with OS-9 compatible ESP1600 spectral acquisition and
handling software. Sample capillaries were inserted in a standard 4-mm
(i.d.) quartz ESR tube containing light silicone oil for increased
thermal stability and were centered in a TE102
rectangular ESR cavity (ER 4201; Bruker). Measurements were performed
at thermal equilibrium starting from low temperature. Sample
temperature was controlled with a Bruker ER 4111VT variable-temperature
control unit (accuracy ± 0.5°C). Conventional first-harmonic
in-phase absorption ESR spectra were recorded at a microwave power of
10 mW with 1 Gp-p magnetic field
modulation amplitude and a frequency of 100 kHz of the magnetic field
modulation used for phase-sensitive detection. Saturation transfer ESR
spectra were recorded in the second harmonic, 90° out-of-phase
absorption mode with a modulation frequency of 50 kHz and a modulation
amplitude of 5 Gp-p. The microwave
power was set for each sample to give an average microwave field over
the sample of H1 = 0.25 G, according
to the standardized protocol given by Fajer and Marsh (1982) and
Hemminga et al. (1984).
Spectrophotometric measurements
Aqueous dispersions of DPPC/PEG-DPPE for the spectrophotometric measurements were prepared as described above, but with omission of the spin label, at a final lipid concentration of 1 mg/ml. The dried lipid films were first hydrated in PBS at pH 7.5, then transferred to a 3-ml quartz cell with 1-cm optical path, and finally incubated overnight at 10°C before measurement.
Optical density measurements at 400 nm were made with a Jasco 7850 spectrophotometer equipped with a Peltier thermostatted cell holder (model EHC-441) and a temperature programmer (model TPU-436; accuracy ±0.1°C). A heating rate of 1°C/min was used. Data acquisition and manipulation were carried out with the built-in microcomputer accessory of the spectrophotometer.
The data presented in this paper are single measurements, but the reproducibility of the results has been tested by repeating the experiments.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Spectrophotometry
Fig. 1 A shows the
optical density at 400 nm (OD400) for
DPPC/PEG:350-DPPE dispersions, measured as a function of the content of
PEG:350-DPPE, at 10 and 50°C. At 10°C, the optical density first
drops rapidly with increasing PEG:350-DPPE content up to ~7-10
mol%, then decreases gradually to close to zero from 40 mol% onward.
Almost the same dependence on composition is observed at 50°C, except
that the OD400 values are somewhat lower than the
corresponding ones at 10°C. It is interesting to note that the
difference between the optical densities at low and high temperatures (i.e., OD400(T = 10°C) 
OD400(T = 50°C)) decreases with
increasing PEG:350-DPPE content and becomes approximately zero at 50 mol% of PEG:350-DPPE.
|
For hydrated mixtures of DPPC with PEG:2000-DPPE, the dependence of the optical density at 10°C and 50°C on the content of PEG:2000-DPPE shown in Fig. 1 B differs somewhat from that observed with the mixtures containing PEG:350-DPPE. Indeed, at both temperatures, the initial decrease in optical density is steeper and then takes place continually with increasing content of PEG:2000-DPPE, reaching a limiting low value at ~20-30 mol% PEG:2000-DPPE. Moreover, the difference in optical densities between low and high temperatures is reduced, as compared to mixtures containing PEG:350-DPPE, and already attains a value close to zero at 30 mol% PEG:2000-DPPE.
A straightforward interpretation of these changes in optical density with increasing content of PEG-lipid in the dispersions is that the initial rapid drop corresponds to a decrease in overall size of the lipid aggregates. At least part of the initial decrease may be attributed to disaggregation of the liposomes. Finally, the closed bilayer vesicles of reduced size convert to micelles, when the OD400 becomes very low. A further criterion for micelle formation is the identity of the optical densities at low and high temperatures. For micelles, there is no longer a change in optical density that results from the change in refractive index of the lipid that is associated with the chain-melting transition in lipid bilayers (see Fig. 1, inset). In terms of the different behavior of mixtures containing PEG:350-DPPE and PEG:2000-DPPE, it appears that the longer polymer-lipid is more effective in reducing the size of the lipid aggregates and achieves micellization more readily than does the shorter polymer-lipid.
Conventional electron spin resonance
Results are reported for three different temperatures, 10°C,
30°C, and 50°C, which correspond to the gel, intermediate, and fluid phases, respectively, of dipalmitoyl phosphatidylcholine bilayers
(Mabrey and Sturtevant, 1976).
Gel phase
Typical conventional ESR spectra of 5- and 16-PCSL spin label positional isomers in mixtures of DPPC with different contents of the short polymer-lipid PEG:350-DPPE, at 10°C, are given in Fig. 2, A and B, respectively. The spectrum of 5-PCSL in dispersions of DPPC is close to the rigid limit of sensitivity to rotational motion on the conventional nitroxide ESR time scale. It is indicative of the slow segmental rotational motion that is typical for phospholipid chains in the gel phase (see, e.g., Bartucci et al., 1993). Addition of up to 5 mol% of
PEG:350-DPPE in the DPPC dispersions does not influence the ESR spectra
of 5-PCSL appreciably, relative to that of DPPC alone. Further
increases in the PEG:350-DPPE content of lipid mixtures up to 60 mol%
produce small changes in the ESR spectra, which nonetheless remain
characteristic of spin labels undergoing slow motion in lamellar gel
phase. At 70 mol% the spectrum has a partially motionally averaged,
axial lineshape, and finally at 100 mol% it shows spectral features of
intermediate mobility characteristic of a fluid environment at low
temperature in dispersions of DPPE-PEG:350 alone.
|
). At the other extreme, the spectrum of 16-PCSL in dispersions of PEG:350-DPPE consists of a near-isotropic triplet, with differential broadening of the three hyperfine manifolds and little residual anisotropy. This type of spectrum for the 16-PCSL
probe is characteristic of a liquid-crystalline environment that, in
this case (cf. the optical density results above), must correspond to
polymer-lipid micelles. Progressive mixing of PEG:350-DPPE with DPPC,
up to 30-40 mol% of the polymer-lipid, leads to a limited motional
restriction of the 16-PCSL, and the spectra are still characteristic of
the lipid gel phase. At 50 and 60 mol% of PEG:350-DPPE the spectra of
16-PCSL clearly consist of two components at 10°C. One component
(designated i in Fig. 2 B) corresponds to the
quasi-isotropic environment that is characteristic of the micelles of
PEG:350-DPPE at this temperature. The second broader component
(designated a in Fig. 2 B) represents the residual gel-phase
environment that is characteristic of the mixtures with a lower content
of the PEG-lipid. Most significantly, the relative proportion of the micelle-like component increases with increasing content of PEG-lipid (compare the spectra for 50 and 60 mol%).
ESR spectra of 5- and 16-PCSL at 10°C were also recorded in
DPPC/PEG:2000-DPPE dispersions (spectra not shown). Increasing the
content of the longer PEG:2000-DPPE polymer-lipid has the same
qualitative effect as that observed with the shorter PEG:350-DPPE. However, the changes in the spectral lineshapes occur at much lower
concentrations of the longer polymer-lipid than is the case for the
shorter one. Specifically, little effect on the typical lamellar
gel-phase lineshape in the ESR spectra of 5-PCSL at 10°C in
DPPC/PEG:2000-DPPE mixtures are seen on adding up to 7 mol% of the
longer polymer-lipid. Further admixture up to 25-30 mol% of
PEG:2000-DPPE monotonically decreases the spectral anisotropy. Beyond
40 mol% of PEG:2000-DPPE, the ESR spectra steadily approach the fast
motional regime of conventional spin label ESR spectroscopy. As
concerns the spectra of 16-PCSL in DPPC/PEG:2000-DPPE mixtures, they
have the same gel-phase lineshape as that in DPPC alone up to 7 mol%
of the polymer-lipid. With a further increase of the content of
PEG:2000-DPPE in the dispersions, progressively increasing proportions
of the quasi-isotropic spectral component are found throughout the
range from 15 to 40 mol% of PEG:2000-DPPE. At 40 mol%, the spectrum
consists almost entirely of this latter component.
The implications of these conventional ESR results are clear. The
longer PEG:2000-DPPE polymer-lipid is much more effective in inducing
micelle formation in DPPC bilayers than is the shorter-chain PEG:350-DPPE analog.
Intermediate phase
The ESR spectra at 35°C of 5-PCSL in DPPC/PEG:350-DPPE mixtures at different molar ratios are reported in Fig. 3 A. The spectra of 5-PCSL, both in DPPC and in PEG:350-DPPE dispersions, show a considerable reduction of the spectral anisotropy, relative to those recorded for the same systems at 10°C (compare Fig. 2 A). For dispersions of DPPC alone, this temperature corresponds to the intermediate or rippled gel phase. Consequently, the spectra of DPPC alone are asymmetrically broadened. This corresponds to two unresolved components of slightly different mobility that is a characteristic of the intermediate phase (Tsuchida and Hatta, 1988). With increased
content of PEG:350-DPPE in the mixtures to 40 mol% and above, a second
spectral component of spin-labeled lipids with much higher rotational
mobility is very clearly resolved in the spectra of 5-PCSL. Comparison
with the spectrum from dispersions of PEG:350-DPPE alone shows that
this second component corresponds to a population of lipid micelles.
|
). Taken together with the results for 16-PCSL at 10°C, this further confirms that the second spectral component corresponds to a
population of micelles, rather than to lipids with increased rotational
mobility in a lamellar phase.
ESR spectra of the 5-PCSL at 35°C and 16-PCSL at 30°C in
DPPC/PEG:2000-DPPE mixtures containing the longer polymer-lipid are given in Fig. 3, C and D, respectively. A
systematic pattern of micelle-bilayer coexistence is observed that is
similar to that found with the shorter polymer-lipid. For the lipid
with a longer grafted PEG chain, however, the micellar-like component
is already evident at much lower contents of PEG-lipid (from 10 mol%
upward). This is in agreement with the comparative results obtained
with the two polymer-lipids at lower temperatures.
Taking advantage of the good resolution of the two spectral components
in this temperature regime, the relative populations of the
n-PCSL spin labels in the micellar and bilayer environments were quantified by using difference spectroscopy as described by Marsh
(1982). Subtraction of the micellar (lamellar) component from the
two-component spectra of the mixtures, by using the single-component spectrum from dispersions of the PEG-lipid (DPPC) alone, yielded the
fraction f(micelle) (f(lamellar)) of spin labels
in the micellar (lamellar) phase. The fraction of micellar component,
f(micelle), is given as a function of the composition of the
DPPC/PEG-lipid mixtures in Fig. 4. In
cases where reliable subtraction is possible with the spectra of
5-PCSL, the results agree with those obtained with 16-PCSL, even though
the temperatures of measurement differ. Corresponding spectral
subtractions performed on the 16-PCSL spectra at 10°C (cf. Fig. 2
B) were only possible over a much more restricted range of
composition, because of the suboptimal resolution of the two
components. Nevertheless these gave values of f(micelle) in
good agreement with those obtained at 30°C, i.e., the temperature giving the best resolution of the two components. This clearly demonstrates that the mobile spectral component corresponds to micelles, rather than to domains with different chain mobilities in an
extended wholly lamellar phase. It should be noted that micelles
assayed operationally by this method include bilayer fragments that are
sufficiently small so as not to undergo the cooperative chain packing
characteristic of lipid gel phases. Larger fragments that support
cooperative chain immobilization are here classified as lamellar.
|
). This point is considered explicitly in the
Discussion and the Appendix.
Fluid phase
Representative ESR spectra of 5-PCSL and of 16-PCSL at 50°C in hydrated dispersions of DPPC, PEG:350-DPPE, and PEG:2000-DPPE are given in Fig. 5. The spectra of 5-PCSL are all axial, anisotropically averaged powder patterns that are typical of flexible lipid chains in a fluid liquid-crystalline environment. The extent of motional averaging is, however, much larger for the micellar states of the PEG-lipids than for the fluid lamellar state of DPPC. This indicates an increased amplitude of chain segmental mobility in the polymer-lipid that reflects, at least in part, the tendency of the polymer-grafted lipids to form micelles and to decrease the lipid chain packing density on going from bilayers to micelles.
|
).
Note that the ESR spectra of the 5-PCSL and 16-PCSL spin labels in
DPPC/PEG:350-DPPE and DPPC/PEG:2000-DPPE dispersions do not consist of
two resolved components in the fluid phase, at any composition of the
lipid mixtures. This is because, unlike the situation in the gel phase,
the difference in chain motional anisotropy between the lamellar and
micellar fluid phases is insufficient for resolution of the two
overlapping spectral components.
Saturation transfer ESR
Because, at 10°C, the rotational rate of 5-PCSL in DPPC/PEG-DPPE
dispersions is close to the limits of motional sensitivity of
conventional spin label ESR spectroscopy, the rotational dynamics at
this temperature can be defined more precisely by use of the STESR
spectroscopy (Thomas et al., 1976; Marsh, 1981; Hemminga and de Jager,
1989). STESR spectra of 5-PCSL at 10°C in dispersions of
DPPC/PEG:350-DPPE mixtures at selected mole ratios are given in the
inset of Fig. 6. The spectral lineshapes
indicate that the rotational motion of the spin-labeled lipid chains
lies in the saturation transfer ESR regime for DPPC/PEG:350-DPPE
mixtures up to 60 mol% of PEG-lipid. Beyond this concentration, the
spectra contain components from motion on the slow conventional
spin-label ESR time scale (see Fig. 2 A) and therefore are
not appropriate for quantitative analysis by STESR.
|
The STESR spectra from 5-PCSL in DPPC/PEG:350-DPPE dispersions at
10°C were analyzed by measuring the diagnostic lineheight ratios,
R (i.e., L"/L,
C'/C, and H"/H in the low-,
central-, and high-field hyperfine manifolds, respectively. The
dependence on lipid composition of the lineheight ratios is given in
Fig. 6. The most striking feature of this dependence is the rapid
initial decrease of all lineheight ratios on addition of the PEG-lipid to gel-phase DPPC. The decrease in the C'/C ratio
may indicate an increase in rate of rotation about the long molecular
axis of the lipids (Marsh, 1980). However, the conventional ESR spectra in Fig. 2 A suggest that there is no increase in chain
segmental rotation rates, which would be reflected in the
L"/L and H"/H ratios.
Therefore the initial decrease in the latter most probably reflects a
decrease in size of the lamellar aggregates with the addition of the
PEG-lipid. In Table 1 are given the
effective rotational correlation times, 
Reff. They
have been deduced from standard calibrations of the lineheight ratios
for isotropic motion taken from Horváth and Marsh (1988), Marsh
(1992), and Marsh and Horváth (1992), using the expression
|
(1) |
. For the short correlation time
regime (for micelles), calibrations were taken from Thomas et al.
(1976) and Fajer and Marsh (1982). As can be seen in Table 1,
comparable values are obtained from the low- and high-field ratios, but
shorter effective correlation times are obtained from the central
ratio. The latter most probably reflects preferential faster rotation
about the long axis of the lipid molecules (Marsh, 1980). The values of
Reff deduced from L"/L and
H"/H decrease up to 5-10 mol% of PEG:350-DPPE. Beyond this, they vary nonsystematically (cf. Fig. 6), which may indicate changing anisotropy in the motion, as possibly suggested by
the varying lineshape in the H" diagnostic region of the
spectrum.
|
The STESR spectra of 5-PCSL in DPPC/PEG:2000-DPPE mixtures at 10°C
(not shown) reveal the absence of submicrosecond motions only over a
much smaller range of polymer-lipid contents. A decrease in
Reff occurs only up to 7 mol% of PEG:2000-DPPE (see
Table 1), again suggesting rapid induction of small lamellar aggregates
with the addition of PEG-lipid.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Optical density
The optical density measurements give qualitative information on
the changes in size of the lipid structures and the formation of
micelles, which are, on the whole, in agreement with previous studies
on related PEG-grafted lipid systems. For the corresponding DSPC/PEG:2000-DSPE mixtures of lipids with matched distearoyl (DS-)
chains, the dependence of the optical density on lipid composition is
very similar to that found here for the DPPC/PEG:2000-DPPE mixtures
with dipalmitoyl chains and the same average PEG polymer length
(Kenworthy et al., 1995). From polarized light microscopy, the latter
authors ascribed the decrease of the absorbance of the lipid
dispersions to a gradual conversion from multilamellar vesicles to
small unilamellar vesicles and to micelles. From the similarity in
behavior of the optical densities, it can be assumed that this is also
the case here for the DPPC/PEG:2000-DPPE lipid mixtures. For
DPPC/PEG:1000-DPPE and DPPC/PEG:3000-DPPE hydrated lipid dispersions,
it was found that the turbidity of the suspensions decreased first
gradually, then more abruptly, and finally the suspensions became
completely transparent at higher content of the PEG-lipid (Bedu-Addo et
al., 1996). Coexisting smaller multilamellar and unilamellar vesicles
were found by optical microscopy for dispersions containing 7.5 mol%
of PEG:2000-DPPE in egg yolk phosphatidylcholine (Lasic et al., 1991b).
In contrast to the system studied here with dipalmitoyl chains, rather
different results have been obtained from optical density measurements
on the shorter PEG:350 polymer-lipid in DSPC/PEG:350-DSPE mixtures with
longer distearoyl lipid chains (Kenworthy et al., 1995). In the latter
case, it was found that the optical density of the mixed lipid
dispersions increased progressively with increasing content of the
PEG-lipid, which was attributed to the existence of multilamellar
vesicles over the entire range of lipid mixtures. This points to a very
sensitive dependence of the optical density behavior on lipid chain
length for PEG-lipids with relatively short polymer polar groups
(compare with Fig. 1 A). The degree of polymerization of
PEG:350 is approximately 7, corresponding to 14 methylene units plus
seven ether oxygen links, as compared with 36 C-units and two acyl
oxygen links in the chains of a distearoyl phospholipid. Because these
sizes are comparable, it is expected that the behavior of the
PEG:350-lipids could be critically modulated by the lipid chain length,
as is observed. This is not expected, however, for the much longer
PEG:2000-lipids, which again is in agreement with experimental
observation (see previous paragraph).
In addition to steric repulsion and hydration forces between the polymer headgroups, longer range electrostatic bilayer-bilayer repulsion likely also contributes to the reduction in size of the lipid aggregates by the negatively charged PEG-lipids. Comparison of results obtained with PEG:350-DPPE and PEG:2000-DPPE, which bear the same charge, indicate the predominance of the headgroup-size-dependent effects. Electrostatic effects are, of course, strongly dependent on the ionic strength of the suspending buffer.
Rotational dynamics of the lipid chains
The conventional ESR spectral lineshapes of the chain-labeled lipids, at a temperature corresponding to the gel phase of DPPC and a temperature corresponding to the fluid phase, contain important information on the segmental mobility of the lipid chains in the various lipid mixtures. There are very marked differences in the effects of the two polymer-lipids on the local chain mobility at 10°C. The PEG-lipid with the longer polymer chain is more effective in disrupting the tight packing of the lipid chains in the gel phase than is the shorter polymer-lipid. The anisotropy of the slow-motion spectra of 5-PCSL is reduced appreciably over the range up to 25 mol% of PEG:2000-DPPE, for which the degree of micellization remains relatively low (cf. Fig. 4). In contrast, the chain mobility remains characteristic of a slightly perturbed gel phase over the range up to 60 mol% of PEG:350-DPPE, despite the fact that the optical density indicates a considerable decrease in particle size (see Fig. 1 A).
In the fluid phase, admixture of the PEG-lipids has a marked effect on the angular amplitude of the lipid-chain motion. The direction of the change is that expected from the difference between DPPC bilayers and PEG-lipid micelles but takes place well before appreciable micelle formation. However, the extent of the effects is not strongly dependent on the size of the polymer headgroup.
Bilayer-micelle conversion
Hristova and Needham (1995) have modeled the transition from
bilayers to micelles, as a function of the total PEG-lipid content, by
calculating the minimum free energies of both structures. Very significant general thermodynamic features emerge from these
calculations that allow fitting of the results given in Fig. 4, in a
manner consistent with only two free parameters, as outlined below.
Over the range from the start of the bilayer to micelle transition at
total mole fraction of PEG-lipid XPEGtr to
the completion at total mole fraction
XPEGcomp, the compositions of the coexisting
micellar and bilayer phases remain constant, at the values specified by
the respective onset and completion points (Hristova and Needham,
1995). This is in agreement with the Gibbs phase rule. Consequently,
the mole ratio of PEG-lipid to phosphatidylcholine is equal to
XPEGtr in the bilayer phase and to
XPEGcomp in the micellar phase, throughout
the transition. The total fraction of lipid (i.e., PEG-lipid + PC) in
the micellar phase is then given by the lever rule (see Appendix):
|
(2) |
and is required by the phase rule.
To interpret the spin label results of Fig. 4, we require the fraction
fPC(micelle) of the
phosphatidylcholine component in the micellar phase. It can be shown
that this is given by (see Appendix)
|
(3) |
; Hristova et al., 1995). The
value of XPEGtr for PEG:2000-DPPE is in
agreement with the explicit calculation by Hristova and Needham (1995),
but that of XPEGcomp is considerably higher
than in the same calculation. The higher completion point is more in
agreement with the experimental measurements by Hristova et al. (1995).
In contrast, experimental estimates for mixtures of phosphatidylcholine
with PEG-lipids of unmatched lipid acyl chain length (DPPC and
PEG:2000-DSPE) yielded degrees of micellization greater than those
predicted by the above theoretical calculations (Baekmark et al.,
1997). These experiments used differential scanning calorimetry and
could conceivably underestimate the bilayer population.
Saturation transfer ESR
In large multilamellar vesicles of DPPC alone, the STESR spectra
are dominated by the molecular rotational mobility of the 5-PCSL
phospholipid probe (Marsh, 1980). As PEG-lipids are mixed with DPPC,
changes in the effective rotational correlation time measured by STESR
may occur because of changes in the local molecular mobility
(characterized by Rmol) and because of a decrease in
size of the lipid aggregates. The effective rotational rate (i.e.,
1/Reff) is then the sum of the rates of molecular
rotation (i.e., 1/Rmol) and of the overall rotation
of the lipid aggregates:
|
(4) |
Ragg is the rotational correlation time
of the lipid aggregates. An estimate of Ragg for the
DPPC/PEG-lipid mixtures can be obtained by assuming that Rmol remains the same as in DPPC alone. This is
likely to be a reasonable approximation for low contents of PEG-lipids,
especially if Reff is taken from the
L"/L or H"/H lineheight
ratios, which are insensitive to rotation around the long molecular axis.
Applying this assumption to the data derived from the
L"/L and H"/H diagnostic
STESR ratios gives the effective values of Ragg that
are listed in Table 1. On this basis, the rotational mobility of the
lipid aggregates increases rather rapidly with increasing content of
PEG-lipid. At PEG:350-DPPE lipid contents in the region of 1-15 mol%,
the effective correlation times for overall rotation of the aggregates
are in the region of 70 and 120 µs, deduced from
L"/L and H"/H,
respectively. Somewhat shorter values are obtained for comparable
contents of PEG:2000-DPPE.
The correlation time for overall rotation of the lipid aggregates about
their symmetry axis is given by a Debye-type expression:
|
(5) |
is the aqueous viscosity,
kB is Boltzmann's constant, and
T is the absolute temperature. This expression is correct to
within an asymmetry factor that depends on the shape of the aggregates.
For the rotational correlation times quoted above, Eq. 5 corresponds to
effective diameters in the region of 100 nm. The volume of the lipid
aggregates can be expressed, quite generally, in terms of the
aggregation number, nagg. The
resulting dependence of the overall correlation time on the mole
fraction, XPEG, of PEG-lipid is given
by
|
(6) |
1.1 nm3) is the core volume of the phospholipid
molecule without polymer attached to the headgroup, and
vPEG is the effective volume of the
PEG-lipid polymer headgroup. For a flexible polymer, the latter depends
on the surface concentration of PEG-lipids. In the low concentration
regime, the PEG polymer forms isolated "mushroom" structures of
volume vPEG (mushroom) = vOEnOE9/5, where
vOE is the volume of an oxyethylene
monomer unit [-(CH2)2O-] and nOE (= 7, 46 for PEG:350, 2000, respectively) is the degree of polymerization of the PEG headgroup (de
Gennes, 1980). In the higher concentration regime, the PEG polymer
forms a continuous brush on the surface with
vPEG (brush) = vOEnOEXPEG
2/3,
which is dependent on the mole fraction,
XPEG, of PEG-lipid (de Gennes, 1980).
This is, of course, an oversimplification of the effects of a flexible
polymer on overall rotational diffusion. However, the different scaling
laws in the two polymer concentration regimes do explain, in part, the
biphasic dependence of the rotational correlation time on the mole
fraction of PEG-lipid that is implicit in Fig. 6. In the
low-concentration regime Ragg is directly
proportional to XPEG, whereas in the
high-concentration regime it increases only as
XPEG1/3 (see Eq. 6), for a fixed aggregation
number. Added to this, the decrease in aggregation number with
progressive micelle formation competes against this slow increase with
the added bulk of the polymer headgroup (see Eq. 6), resulting in the
relatively constant values of Ragg that
are obtained after the initial drop with increasing PEG-lipid content.
The rapid initial drop with the increasing mole fraction XPEG of polymer-lipid suggests that
the rather open "mushroom" configuration of the surface polymer
makes little contribution to the rotational mobility of the lipid
vesicles, which is dominated by the decreasing size, i.e., a rapid
reduction in nagg.
In the case of cylindrical micelles, for which there is direct
structural evidence in PC/PEG-lipid systems (Edwards et al., 1997),
both the aggregation number and rotational correlation time decrease
directly with decreasing length l of the aggregates. For
spherical micelles, on the other hand, the aggregation number is
expected to remain approximately fixed.
At least qualitatively, these results are consistent with direct
structural studies on similar PEG-lipid/PC systems. By using cryotransmission electron microscopy, Edwards et al. (1997) observed the formation of open bilayer discs in PC/PE-lipid mixtures, before the
transition to mixed micelles, which were found to have a thread-like rather than globular shape in the absence of cholesterol. In addition, Lasic et al. (1991b) have reported a reduction in the hydrodynamic radius of lamellar dispersions of egg yolk phosphatidylcholine mixed
with 7.5 mol% of PEG:2000-DPPE. From quasielastic light scattering
measurements on hydrated egg PC/PEG:2000-DPPE mixtures, Szleifer et al.
(1998) observed that the average diameter of the aggregates decreased
with insertion of the polymer-lipid, lying in the range 420-550 nm for
PEG-lipid contents of 0.2-10 mol%. By using measurements of dynamic
light scattering, Bedu-Addo et al. (1996) showed a reduction in
aggregate size of DPPC/PEG:1000,3000-DPPE dispersions from 3000 to 2000 nm on going from 0 to 5 mol% of PEG-lipid, and then further to 40 nm
on increasing the polymer-lipid content to 17 mol%.
| |
CONCLUSIONS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Conventional and saturation transfer ESR spectroscopies, together with spectrophotometry, were used to study the lyotropic phase behavior of DPPC/PEG-DPPE dispersions at full hydration. We have concentrated on PEG-lipid headgroups of two rather different molecular masses, i.e., 350 and 2000 Da.
1. The local segmental mobility of the lipid chains is sensitive to the content of PEG-lipid in the low concentration regime. Chain mobility is increased with increasing PEG-lipid concentration in both gel and fluid PC bilayers. Dependence on polymer headgroup size is marked in the gel phase but smaller in the fluid phase.
2. Conversion from bilayer vesicles to micelles could be quantitated from the resolved two-component ESR spectra and is shown to be consistent with a thermodynamic description of the bilayer-micelle transition. The onset and completion points determined for the transition depend on polymer-lipid headgroup size, in a way that is consistent with shape concepts of lipid polymorphism.
3. Saturation transfer ESR measurements of the effective rotational diffusion rates of the lipid aggregates demonstrate the sensitive dependence in the "mushroom" regime at low concentrations of PEG-lipid, and a much weaker dependence in the "brush" regime with increasing content of PEG-lipid.
| |
APPENDIX: BILAYER-MICELLE TRANSITION THERMODYNAMICS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
For a mixture of polymer-lipid and phosphatidylcholine (i.e.,
C = 2, with water in excess), in the bilayer-micelle
coexistence region (i.e., P = 2), the number of degrees
of freedom given by the Gibbs phase rule is F = C p + 1 = 1, at constant pressure. Therefore, at a fixed temperature, the compositions of the coexisting bilayer and micellar phases are fixed, independent of the total composition (see, e.g., Cevc and Marsh, 1987). Varying the total composition results in bilayer-micelle conversion. The composition of
the bilayer phase is specified by the end of the tie line at the
beginning of the transition (viz., mole fraction of PEG-lipid = XPEGtr). The composition of the micellar
phase is specified by the end of the tie line at the completion of the
transition (viz., mole fraction of PEG-lipid = XPEGcomp).
If ftot(micelle) is the fraction of
total lipid in the micellar phase, the total mole fraction of
PEG-lipid, XPEG, in the mixture is
given in terms of the compositions of the individual phases by
conservation of mass:
![X<SUB><UP>PEG</UP></SUB>=[1−f<SUB><UP>tot</UP></SUB>(<UP>micelle</UP>)]X<SUP><UP>tr</UP></SUP><SUB><UP>PEG</UP></SUB>+f<SUB><UP>tot</UP></SUB>(<UP>micelle</UP>)X<SUP><UP>com p</UP></SUP><SUB><UP>PEG</UP></SUB>](/content/vol78/issue3/fulltext/1420/img007.gif)
|
(A1) |
|
(A2) |
Within the micellar phase, the mole fraction of phosphatidylcholine is
simply 1 XPEGcomp. For the whole
sample, the mole fraction of phosphatidylcholine present in the
micellar phase is therefore (1 XPEGcomp).ftot(micelle),
viz., when referred to both phases, where the total mole fraction of
phosphatidylcholine is 1 XPEG). Hence, the fraction of
phosphatidylcholine that is in the micellar phase is given by
|
(A3) |
| |
ACKNOWLEDGMENTS |
|---|
This work, financially supported by the Ministero dell'Università e della Ricerca Scientifica e Tecnologica (MURST) and by the Istituto Nazionale per la Fisica della Materia (INFM), is part of the Ph.D. thesis of SB. SB and GM thank MURST and European Community, respectively, for the award of Ph.D. fellowships.
| |
FOOTNOTES |
|---|
Received for publication 4 August 1999 and in final form 11 November 1999.
Address reprint requests to Dr. L. Sportelli, Dipartimento di Fisica, Università della Calabria, I-87036 Arcavacata di Rende (CS), Italy. Tel.: +39-0984-493131; Fax: +39-0984-493187; E-mail: sportelli{at}fis.unical.it.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSIONS
APPENDIX
REFERENCES
|
|---|
Biophys J, March 2000, p. 1420-1430, Vol. 78, No. 3
© 2000 by the Biophysical Society 0006-3495/00/03/1420/11 $2.00
This article has been cited by other articles:
 |
|
 |
|
  M. Johnsson and K. Edwards Liposomes, Disks, and Spherical Micelles: Aggregate Structure in Mixtures of Gel Phase Phosphatidylcholines and Poly(Ethylene Glycol)-Phospholipids Biophys. J., December 1, 2003; 85(6): 3839 - 3847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. L. Hansen, J. A. Cohen, R. Podgornik, and V. A. Parsegian Osmotic Properties of Poly(Ethylene Glycols): Quantitative Features of Brush and Bulk Scaling Laws Biophys. J., January 1, 2003; 84(1): 350 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Dubertret, P. Skourides, D. J. Norris, V. Noireaux, A. H. Brivanlou, and A. Libchaber In Vivo Imaging of Quantum Dots Encapsulated in Phospholipid Micelles Science, November 29, 2002; 298(5599): 1759 - 1762. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Rovira-Bru, D. H. Thompson, and I. Szleifer Size and Structure of Spontaneously Forming Liposomes in Lipid/PEG-Lipid Mixtures Biophys. J., November 1, 2002; 83(5): 2419 - 2439. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||
) and
50°C (
). The errors are smaller than the symbols. The inset shows
the temperature dependence of the optical density for DPPC
(upper curve) and PEG:350-DPPE (line)
aqueous dispersions. The arrows indicate the temperatures
Tp and Tm at
which the pretransition and main transition, respectively, are detected
in DPPC dispersions.
), in the low-, central-, and
high-field regions, respectively, of the STESR spectra of 5-PCSL at
10°C in aqueous dispersions of DPPC/PEG:350-DPPE mixtures. The errors
range from ±4 to ±7% for L"/L, from
±3 to ±6% for C'/C, and from ±15 to
±25% for H"/H. Inset:
Selected STESR spectra of 5-PCSL at 10°C in aqueous dispersions of
DPPC/PEG:350-DPPE mixtures. The lineheights L",
L; C', C; and
H", H in the low-, central-, and
high-field regions of the spectrum are indicated. Total scan width = 100 gauss.