Enzyme-mediated formation of vesicles from DPPC-dodecyl maltoside mixed micelles

Article abstract of DOI:10.1021/ja980400a

An enzymatic procedure for liposome formation through micelle to vesicle transition is described. Amyloglucosidase hydrolysis of dodecyl-β-D- maltoside (DM) giving dodecyl-β-D-glucoside (DG) leads to dipalmitoylphosphatidylcholine (DPPC)-based vesicle formation from DPPC-DM mixed micelles. Starting from a 1.8 DM/DPPC molar ratio corresponding to mixed micelles, progressive hydrolysis of DM gives DPPC-DG-DM intermediate aggregates ending with DPPC-DG vesicles upon reaction completion. Initial steps of the process corresponding to the exit of the micellar domain were followed by turbidimetry measurements. Next, the reaction progress was investigated by RP-HPLC, HPLC-GEC, and cryofracture electron microscopy. A constant reaction rate is observed in the micellar domain, while the increase of the lamellae proportion considerably decreases the enzyme catalytic activity. Finally, the enzymatic hydrolysis is significantly slowed when closed vesicles are formed. Enzymatic activity is dependent on DM availability in the bulk phase and of the DM/DPPC molar ratio in the aggregates. The presence of mixed micelles or lamellar sheets considerably modulates DM monomer concentration in the aqueous phase. The liposomes formed by the enzymatic process are spherical, unilamellar, and heterogeneous in size with a mean diameter ranging from 10 to 80 nm.

Full text of DOI:10.1021/ja980400a

10588
J. Am. Chem. Soc. 1998, 120, 10588-10595
Enzyme-Mediated Formation of Vesicles from DPPC-Dodecyl
Maltoside Mixed Micelles
Brigitte Carion-Taravella,‡ Sylviane Lesieur,§ Michel Ollivon,§ and Joe1l Chopineau *,‡
Contribution from the Laboratoire de Technologie Enzymatique, CNRS UPRES A 6022,
60205 Compie`gne, France, and Equipe Physicochimie des Syste`mes Polyphase´s, CNRS URA 1218,
92296 Chaˆtenay-Malabry, France
ReceiVed February 4, 1998
Abstract: An enzymatic procedure for liposome formation through micelle to vesicle transition is described.
Amyloglucosidase hydrolysis of dodecyl-â-D-maltoside (DM) giving dodecyl-â-D-glucoside (DG) leads to
dipalmitoylphosphatidylcholine (DPPC)-based vesicle formation from DPPC-DM mixed micelles. Starting
from a 1.8 DM/DPPC molar ratio corresponding to mixed micelles, progressive hydrolysis of DM gives DPPC-
DG-DM intermediate aggregates ending with DPPC-DG vesicles upon reaction completion. Initial steps of
the process corresponding to the exit of the micellar domain were followed by turbidimetry measurements.
Next, the reaction progress was investigated by RP-HPLC, HPLC-GEC, and cryofracture electron microscopy.
A constant reaction rate is observed in the micellar domain, while the increase of the lamellae proportion
considerably decreases the enzyme catalytic activity. Finally, the enzymatic hydrolysis is significally slowed
when closed vesicles are formed. Enzymatic activity is dependent on DM availability in the bulk phase and
of the DM/DPPC molar ratio in the aggregates. The presence of mixed micelles or lamellar sheets considerably
modulates DM monomer concentration in the aqueous phase. The liposomes formed by the enzymatic process
are spherical, unilamellar, and heterogeneous in size with a mean diameter ranging from 10 to 80 nm.
Introduction
However, in these studies, the chemical nature of the
product(s) and the phases formed are far from reproducing cell
conditions. Indeed, the rationale of such work would be greater
if natural molecules and systems of biological relevance such
as phospholipids and liposomes were used instead of organic
solvent-water based mesophases. In this respect, phospho-
lipid(s)-surfactant(s) mixtures form numerous types of ag-
gregates in water depending on their relative proportions.9
Removal of the solubilizing surfactant (detergent) from mixed
micelles results in vesicle formation through the so-called
micelle-vesicle transition.10-18 The latter is involved in many
processes of liposome preparation10-12,14,15,19 and in the extrac-
tion and the reconstitution of biomembrane proteins.20-22
In a preliminary work,23 the possibility of liposome formation
through enzymatic processes was described. Two distinct
The consequences of cell microstructuration on enzyme
functions and the need of geometrically restricted systems to
reproduce the complexity and the diversity of cell organization
led to an approach which used biomimetic environments called
microheterogeneous systems obtained from water/surfactant/
organic solvent mixtures.1,2 The study of relationships between
enzymes and microorganized media was envisaged in micro-
structured systems with the same physicochemical features as
the structures found at the cellular level.3,4 Indeed, ternary
mixtures composed of surfactant, water, and organic solvent
provide a variety of colloidal organizations capable of self-
evolution through the dynamic interactions of their components
with an enzyme.5 Such systems also permit following enzy-
matic reactions in restricted environments,6 wherein the enzyme
is responsible for environmental changes which in turn modulate
enzyme catalytic activity. Self-replicating systems, based on
dynamic interactions between amphiphilic aggregates (micelles,
reversed micelles, vesicles) and enzymes, have been developed
in relationship with autopoiesis.7,8
(8) Schmidli, P. K.; Schurtenberger, P.; Luisi, P. L. J. Am. Chem. Soc.
1991, 113, 8127-8130.
(9) Israelachvili, J. N. In Physics of amphiphiles: micelles, Vesicles,
microemulsions; Degiorgio, V., Corti, M. Eds.; North-Holland Physics
Publishing-Elsevier: Amsterdam, 1985; 24-58.
(10) Lichtenberg, D. CRC Handbook of Chemistry and Physics; CRC
Press: Boca Raton, 1996; pp 199-218.
‡ To whom correspondence should be addressed. E-mail:
joel.chopineau@utc.fr.
(11) Paternostre, M.-T.; Roux, M.; Rigaud, J.-L. Biochemistry 1988, 27,
2668-2677.
§ Laboratoire de Technologie Enzymatique.
(12) Ollivon, M.; Eidelman, O.; Blumenthal, R.; Walter, A. Biochemistry
1988, 27, 1695-1703.
‡ Equipe Physicochimie des Syste`mes Polyphase´s.
(1) Martinek, K.; Levashov, A. V.; Khmelnitsky, Y. L.; Klyachko, N.
L.; Berezin, I. V. Science 1982, 218, 889-891.
(13) Ueno, M. Biochemistry 1989, 28, 5631-5634.
(14) Da Grac¸a Miguel, M.; Eidelman, O.; Ollivon, M.; Walter, A.
Biochemistry 1989, 28, 8921-8928.
(2) Luisi, P. L. Angew. Chem., Int. Ed. Engl. 1985, 24, 439-450.
(3) Klyachko, N. L.; Levashov, A. V.; Pshezhetsky, A. V.; Bogdanova,
N. G.; Berezin, I. V.; Martinek, K. Eur. J. Biochem. 1986, 161, 149-154.
(4) Luisi, P. L.; Giomini, M.; Pileni, M. P.; Robinson, B. H. Biochim.
Biophys. Acta 1988, 947, 209-246.
(15) Vinson, P. K.; Talmon, Y.; Walter, A. Biophys. J. 1989, 56, 669-
681.
(16) Kameyama, K.; Takagi, T. J. Colloid Interface Sci. 1991, 146, 512-
524.
(5) Chopineau, J.; Thomas, D.; Legoy, M. D. Eur. J. Biochem. 1989,
183, 459-463.
(17) Beugin, S.; Grabielle-Madelmont, C.; Paternostre, M.; Ollivon, M.;
Lesieur, S. Prog. Colloid Polym. Sci. 1995, 98, 206-211.
(18) Paternostre, M.; Meyer, O.; Grabielle-Madelmont, C.; Lesieur, S.;
Ghanam, M.; Ollivon, M. Biophys. J. 1995, 69, 2476-2488.
(19) Almog, S.; Litman, B. J.; Wachtel, E. J.; Barenholz, Y.; Ben-Shaul,
A.; Lichtenberg, D. Biochemistry 1990, 29, 4582-4592.
(6) Chopineau, J.; Ollivon, M.; Thomas, D.; Legoy, M. D. Pure Appl.
Chem. 1992, 64, 1757-1763.
(7) Bachmann, P. A.; Walde, P.; Luisi, P. L.; Lang, J. J. Am. Chem.
Soc. 1991, 113, 8204-8209.
10.1021/ja980400a CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/06/1998

Vesicles from DPPC-DM Mixed Micelles
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10589
Spectrophotometric Measurements. Amyloglucosidase activity in
buffer was measured by spectrophotometric analysis of hydrolysis of
p-NPG. In a typical experiment, 1 mL of the aryl-R-D-glucoside
solution (3 mM) in a citrate-phosphate buffer (50 mM, pH 5.2) was
placed inside a spectrophotometer thermostated cuvette and 10 µL of
an enzyme stock solution was added. The p-nitrophenol liberated was
enzymatic preparations of liposomes from micelles or open
structures composed of classical amphiphilic compounds have
been proposed. One example was based on our knowledge
concerning DPPC-alkylglucosides micelle-to-vesicle transition,
while the second involved cholesterol-PEG derivatives as
solubilizing lipids. For the first system, dipalmitoylphosphati-
dylcholine (DPPC)-based vesicles were obtained from DPPC-
dodecyl-â-D-maltoside (DM) mixed micelles. Only the initial
starting mixed micelles and the final liposome suspension were
characterized. The intermediate states of aggregation and the
phases encountered during the process were not approached.
Moreover, the influence of enzyme kinetics on the structural
transformation of the lipid assemblies and vice versa were not
elucidated. The consecutive DM hydrolysis by amyloglucosi-
dase and â-D-glucosidase was initially envisaged to remove the
surfactant entirely, forming two molecules of glucose and one
of dodecanol. In fact, DM hydrolysis by amyloglucosidase
alone leaves an interesting bilayer-compatible surfactant, dode-
cyl-â-D-glucoside (DG), appearing to play the main role in the
micelle-vesicle transition process. This is supported by the
determination of the phase sequences and their boundaries in
the DPPC-DG-DM ternary system upon DPPC and DPPC-
DG vesicle solubilization by DM.24
In this work, DPPC-DG vesicles formation induced by
amyloglucosidase hydrolysis of DM from DPPC-DM mixed-
micelle solutions was considered in the framework of the
DPPC-DG-DM ternary phase diagram. Amyloglucosidase
kinetics toward DM was studied with and without the presence
of lipid. The first steps of the enzyme-mediated vesicle
formation were monitored by turbidimetry. The closure of the
bilayered aggregates was examined by high performance gel
exclusion chromatography (HPLC-GEC) and cryofracture elec-
tron microscopy.
continuously measured by recording optical density at 380 nm (ꢀ380
)
1227 M-1 cm-1 for p-nitrophenol under these experimental conditions).
Reverse Phase High-Performance Liquid Chromatography (RP-
HPLC). Amyloglucosidase activity toward DM was studied following
the surfactant disappearance using a C18 µbondapack column (3.9 ×
300 mm) connected to an HPLC apparatus equipped with a 410
refractometer (Millipore Waters). Prior to use, the column was
equilibrated with a 2.5 mM DM aqueous solution. In a typical
experiment, the reaction was started by the addition of 50 µL of an
enzyme stock solution to the lipid and/or surfactant mixture (6-7 mL)
previously placed inside a 10-mL vial. A Gilson 232-401 automatic
sample processor and injector were used for sampling. For each
measurement, aliquots of (50-100 µL) were withdrawn and either
added to an acetonitrile/water (50% v/v) mixture containing hexanol
(50 mM) as internal standard or not. The column was loaded with 20
µL samples. The eluant (acetonitrile/water 50% v/v) was flushed at 1
mL/min. Under these conditions, the alkylglucosides DM and DG were
eluted in two distinct peaks. Reaction rates were calculated from the
kinetic curves plotting DM concentration as a function of time. One
unit corresponds to the enzymatic activity that will hydrolyze 1 µmol
of substrate per minute under these conditions.
Monitoring of Vesicle Formation by the Enzymatic Process.
Mixed Aggregates Preparation. DPPC-DM mixed aggregates were
prepared from a dried lipid-surfactant film formed from a mixed
surfactant(s)-lipid chloroform solution by removing the organic solvent
under a nitrogen stream followed by 12 h drying under vacuum. Mixed
micelles were obtained by adding buffer and gentle mixing, while DPPC
and DPPC-DG small unilamellar vesicles (SUV) were prepared by
ultrasonic irradiation, according to a procedure already described.28 For
encapsulation experiments, dry DPPC or DPPC-DG films were
hydrated by using 10 mM HEPES, pH 7.4 buffer containing 1 mM
calcein and 140 mM NaCl to maintain the same osmolarity as that of
the 145 mM NaCl buffer used for the gel exclusion chromatography
analysis.
Turbidity Measurements. In a typical experiment, the mixed
micelles solution was placed in a 3 mL optical quartz cell (Hellma,
France) thermostated at 37 °C, equipped with a paddle stirrer that did
not interfere with the light path. To start the reaction, 15 µL of
amyloglucosidase solution was added to 1.5 mL of DPPC-DM mixed
micelles (0.40 µM e final enzyme concentration e 0.90 µM).
Turbidity at 400 nm was recorded using a Perkin-Elmer Lambda 2
double-beam spectrophotometer.
Experimental Section
Chemicals. Amyloglucosidase (1,4-R-D-glucan glucohydrolase EC
3.2.1.3) from Aspergillus niger, DL-R-dipalmitoylphosphatidylcholine
(DPPC, purity 99%), N-dodecyl-â-D-glucopyranoside (DG, purity 98%),
N-dodecyl-â-D-maltoside (DM, purity 98%) and p-nitrophenyl-R-D-
glucopyranoside (p-NPG, purity 99%) were purchased from Sigma.
These products were used without further purification. All of the
experiments were carried out using 145 mM NaCl, 10 mM HEPES
(N-[2-hydroxyethyl]piperazine-N′-[2-ethanesulfonic acid]) buffer, pH
7.4.
Enzymatic Activity Measurements at 37 °C. The amyloglucosi-
dase extract is composed of two isoenzymes, G1 and G2. Their
molecular weights as determined by SDS-Page electrophoresis (Phast-
System, Pharmacia) were equal to 75 000 and 105 000 g mol-1
respectively, in accordance with literature data (MWG1 ) 82 000-
100 000 g mol-1 and MWG2 ) 110 000-112 000 g mol-1).25,26
Enzyme quantities were determined by weight and by measuring
the absorbance at 280 nm using a molar extinction coefficient of 1.53
× 105 M-1 cm-1 (ꢀ280 ) 1.37 × 105 M-1 cm-1 for amyloglucosidase
G1).27
Characterization of the Aggregates. High Performance Gel
Exclusion Chromatography (HPLC-GEC). Separation of nonen-
trapped calcein from calcein encapsulated in the vesicles was performed
by using HPLC TSK-G6000 PW and TSK-G4000 PW 30 × 0.75 cm
columns (Toyo Soda) connected in series.29 The HPLC apparatus was
equipped with a Hitachi pump (model L-6000) and a precision injection
valve (Rheodyne). The eluant was aqueous buffer devoid of calcein.
Prior to analysis, the columns were saturated with DPPC-DG vesicles
prepared without calcein. Sample loading was 200 µL, and eluant flow
rate was 1.0 mL/min. Elution was monitored on line by using a
circulating quartz cell (Hellma) and placed in a spectrofluorimeter SPEX
(FL1T11). The fluorescence emission intensity of calcein at 419 nm
(excitation wavelength, 367 nm) and 90° light-scattering intensity of
particles (incident wavelength, 367 nm) were simultaneously recorded
as a function of elution time. The effective exclusion volume (V0 )
9.7 mL) of the two columns in series was determined according to the
(20) Helenius, A.; Simons, K. Biochim. Biophys. Acta 1975, 415, 29-
79.
(21) Rabilloud, T.; Fath, S.; Baur, E.; Egly, J.-M.; Relevant, O. Biofutur
1993, 126.
(22) Rigaud, J.; Pitard, B.; Levy, D. Biochim. Biophys. Acta 1995, 1231,
223-246.
(23) Chopineau, J.; Lesieur, S.; Ollivon, M. J. Am. Chem. Soc. 1994,
116, 11582-11583.
(27) Clarke, A. J.; Swensson, B. Carlsberg Res. Commun. 1984, 49, 111-
121.
(24) Carion-Taravella, B.; Chopineau, J.; Ollivon, M.; Lesieur, S.
Langmuir, 1988, 14, 3767-3777.
(28) Lesieur, S.; Grabielle-Madelmont, C.; Paternostre, M.-T.; Moreau,
J. M.; Handjani-Villa, R. M.; Ollivon, M. Chem. Phys. Lipids 1990, 56,
109-121.
(29) Andrieux, K.; Lesieur, S.; Ollivon, M.; Grabielle-Madelmont, C. J.
Chromatogr. B 1998, 706, 141-147.
(25) Swensson, B.; Larsen, K.; Svendsen, I.; Boel, E. Carlsberg Res.
Commun. 1983, 48, 529-544.
(26) Pazur, J. H.; Knull, H. R.; Cepure, A. Carbohyd. Res. 1971, 20,
83-96.

10590 J. Am. Chem. Soc., Vol. 120, No. 41, 1998
Carion-TaraVella et al.
â-glucosidase and a â-galactosidase, respectively.31,32 Second,
the enzyme activity was checked on both DM and DG by thin-
layer chromatography according to conditions previously de-
scribed.5 In both cases, no dodecanol was detected, indicating
that DG is not hydrolyzed. This confirms the amyloglucosidase
stereospecificity toward the R-glucosidic bond of DM.
Kinetics of DM hydrolysis by amyloglucosidase were ex-
amined on pure surfactant micelles. The disappearance of DM
was followed by RP-HPLC. The initial reaction rate (Vi) versus
DM concentration is plotted in Figure 1A. For a given enzyme
concentration, Vi is constant in the 1-6 mM DM concentration
range. Moreover, a linear relationship between Vi and amylo-
glucosidase content is found and used to calculate a specific
activity (Figure 1B). The specific activity was found to range
from 0.04 to 0.06 IU mg-1 as a function of the enzyme batch
and storage period. In the following, the specific activity was
normalized to 0.06 IU mg-1 (Figure 1B). These results suggest
that the amyloglucosidase recognizes the DM monomer as
effective substrate. Accordingly, this behavior was demon-
strated for another glucohydrolase, â-D-glucosidase, toward
octyl-â-D-glucopyranoside (OG).6 Unfortunately, due to both
low critical micelle concentration (cmc) of DM (0.13 mM)24
and the limit of refractive index detection (0.1 mM), hydrolysis
experiments below DM cmc cannot be performed under these
RP-HPLC conditions.
Amyloglucosidase activity toward DM in the presence of
DPPC was studied for an initial DM/DPPC molar ratio equal
to 1.8 which imposes the presence of isotropic mixed micelles
solution.24 During enzymatic reaction, DM removal was
monitored by RP-HPLC. The variations of total DM concentra-
tion as a function of time ([DM](t)) are illustrated by Figure
2A at different enzyme and/or lipid contents. The velocity
dependence was obtained by plotting the d([DM](t))/dt as a
function of time (Figure 2B). Three regimes in reaction rate
(corresponding to regions 1, 2, and 3 on parts B and C of Figure
2) can be identified as a function of time. In region 1, a constant
amyloglucosidase initial activity of 0.06 IU mg-1 toward DM
was found at both a given enzyme or DPPC content, this
indicates that the initial rate of substrate hydrolysis only depends
on the DM concentration. This value is identical to that obtained
by hydrolysis from pure DM micelles which suggests that the
presence of lipids does not affect the specific activity of the
enzyme (Figure 1B). In region 2, the rate of the reaction is
Figure 1. (A) Variations of the initial reaction rates versus substrate
concentration for DM (Vi, (b) or p-NPG (Vi p-NPG, (0) hydrolysis at a
constant amyloglucosidase concentration of 0.6 µM or 5.3 µM,
respectively. (B) Variation of initial reaction rate of DM hydrolysis as
a function of amyloglucosidase concentration from (b) pure DM
micelles ([DM]0 ) 4.0 mM) or from (+) mixed DPPC-DM micelles
([DM]0 ) 9.0 mM, [DPPC] ) 4.5 mM).
elution of nonextruded multilamellar egg phosphatidylcholine/egg
phosphatidic acid vesicles. The total volume (Vt ) 21.5 mL) was given
by elution of NaCl by using a 410 differential refractometer as detector
(Millipore Waters).30
Cryofracture Electron Microscopy. Sample suspension (without
cryoprotectant) was placed on a hollow gold-copper cup (Bal-Tec)
with a centered shaft of 1 mm diameter. The preparation was placed
on a sample holder and transferred to a quick-freeze apparatus. Freezing
was obtained by projection of the sample against a highly cleaned
copper block (Reichert and Jung) cooled by liquid helium (4 K) and
maintained under vacuum. The specimen was fractured using a Balzers
freeze-fracture apparatus at 133 K and under high vacuum (2 × 10-10
bar). The fractured sample is replicated by platin/carbon shadowing.
The replica is washed with distilled water overnight and placed on a
gold mesh grill. Electron micrographs were obtained with a transmis-
sion electron microscope (Leo EM912).
significantly affected and decreases rapidly to 0.01 IU mg-1
.
Finally, in region 3, the reaction rate is very slow and close to
null. In Figure 2C, from plotting the enzyme velocity as a
function of the total DM/DPPC molar ratio, each regime change
clearly appears at a constant total DM/DPPC molar ratio and
does not depend on the DPPC or enzyme concentrations. In
particular, the first regime change corresponds to a molar ratio
of 0.7 (Figure 2C, arrow).
Vesicle Formation. The addition of amyloglucosidase to
DPPC-DM mixed micelles leads to progressive production of
DG as a replacement for DM. Consequently, during the
enzymatic process, the system is constituted by aqueous DPPC-
DG-DM evolving ternary mixtures. The technique consisting
of recording the optical density (OD) of the sample as a function
of surfactant concentration is widely used to follow morphologi-
cal changes of the aggregates occurring during vesicle-to-micelle
transition.11,12,17,33-35 In this study, turbidimetry was applied
to identify the first steps of the enzymatic process.
Results
Enzymatic Kinetic Studies. First, amyloglucosidase activity
toward a hydrosoluble substrate, p-NPG, was examined. A non-
Michaelis-Menten mechanism was observed with an inhibition
by excess of substrate obtained above 5 mM of p-NPG (Figure
1A). A specific activity of 6 IU mg-1 was determined using a
Linweaver-Burk plot {1/Vi ) f(1/[p-NPG])} between 0 and 4.6
mM of p-NPG. Similar inhibition effects were found for the
hydrolysis of p-NPG and p-nitrophenyl-â-D-galactoside by a
(31) Gabelsberger, J.; Liebl, W.; Schleifer, K. H. Appl. Microbiol.
Biotechnol 1993, 40, 44-52.
(32) Pulvin, S.; Friboulet, A.; Thomas, D. Biochim. Biophys. Acta 1990,
1041, 97-100.
(30) Beugin-Deroo, S. Ph.D. Thesis, Universite´ Paris VI, 1997.

Vesicles from DPPC-DM Mixed Micelles
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10591
Figure 3. Variation of OD at 400 nm during DM hydrolysis by
amyloglucosidase for different amyloglucosidase concentrations. The
initial medium (t ) 0) is composed of DPPC-DM mixed micelles.
[E] ) 0.40 µM (solid line), [E] ) 0.60 µM (dashed line) and [E] )
0.90 µM (dotted line). [DPPC] ) 1.48mM. Inset: A, B, and C denote
the characteristic break points determined by the intercept of the
tangents to the turbidity curves.
cosidase and DPPC concentrations was studied. Homothetic
curves were obtained in both cases. They are illustrated for
different enzyme concentrations in Figure 3. They present a
similar complex shape depicting an uneven turbidity variation.
The beginning of each curve is characterized by three break
points noted as A, B, and C (Figure 3, inset), corresponding to
drastic OD variations. The initial OD measurements revealed
low values which are attributed to mixed micelles (up to point
A), and the sharp increase which follows (from A to C) depicts
changes in the morphology and/or interactions of the aggregates
formed during the hydrolysis or is the result of phase-separation
phenomena. The higher the enzyme concentration, the faster
the different characteristic break points are reached. According
to Figure 4, this increase in turbidity observed during the earlier
steps of the enzymatic process and characterized by A, B, and
C break points corresponds to regime 1 of the hydrolysis rate,
that is, to a constant hydrolysis rate (Figure 2). Figure 4 easily
allows the determination of the DM/DPPC total molar ratio
corresponding to modifications of the reaction rate regime by
the projection of the breaks of the rate curve d([DM](t))/dt vs
t onto the [DM/DPPC]tot vs t curve. The ends of regimes 1
and 2 correspond to [DM/DPPC]tot equal to 0.7 and 0.3,
respectively.
Figure 2. DM hydrolysis by amyloglucosidase starting from DPPC-
DM mixed micelles. (A) Total DM concentration [DM](t) and (B)
reaction rate d([DM](t))/dt evolutions as a function of time; (C) d([DM]-
(t))/dt variation as a function of total DM/DPPC molar ratio. Respective
enzyme (E) and DPPC concentrations are [E] ) 1.2 µM, [DPPC] )
2.5 mM (b); [E] ) 0.6 µM, [DPPC] ) 2.5 mM (0) and [E] ) 1.2
µM, [DPPC] ) 5.0 mM (+). The d([DM](t))/dt values are related to
enzyme content and expressed in IU mg-1. Numbers 1, 2, and 3 indicate
the different regimes of reaction rate. [DM] or [DM/DPPC]tot values
at the rate changes from regime 1 to regime 2 are either given by the
projection (solid lines) of breaks of the curves in (B) onto the
corresponding curves in (A), or indicated by the arrow in (C).
Two series of turbidity curves by varying DPPC and/or
enzyme concentrations ([E] ) 0.60 µM for 0.3 mM < [DPPC]
< 3 mM; 0.3 < [E] < 1.0 µM for [DPPC] ) 1.5 mM) were
worked out in order to obtain DM concentrations at the
characteristic turbidity points (A to C). The overall DM
concentration ([DM](t)) at any time is calculated during the
reaction using the following equation
[DM](t) ) [DM] - (
V
(t) dt)/Vtot
(1)
∫
0
Reactions were performed starting from DPPC-DM mixed-
micelles solutions with a surfactant-to-lipid molar ratio of 1.8.
In two series of experiments, the influence of both amyloglu-
where [DM](t) equals the total DM concentration at time t (mM),
[DM]0 equals the initial DM concentration (mM), V(t) equals
the instantaneous reaction rate (µmole/min), t equals the time
in minutes, and Vtot equals the total reaction volume (mL).
Throughout the DPPC and DM concentration range correspond-
ing to the turbidity break points, the specific activity of the
(33) Lichtenberg, D.; Robson, R. J.; Dennis, E. A. Biochim. Biophys.
Acta 1983, 737, 285-304.
(34) Seras, M.; Edwards, K.; Almgren, M.; Carlson, G.; Ollivon, M.;
Lesieur, S. Langmuir 1996, 12, 330-336.
(35) De La Maza, A.; Parra, J. L. Biophys. J. 1997, 72, 1668-1675.

10592 J. Am. Chem. Soc., Vol. 120, No. 41, 1998
Carion-TaraVella et al.
Figure 4. Variation of DM/DPPC total molar ratio (b), turbidity (•) and reaction rate d([DM](t))/dt (+) as a function of time (logarithm axis
scaling); [E] ) 0.6 µM, [DPPC] ) 2.5 mM. The d([DM](t))/dt values are related to enzyme content and expressed in IU mg-1. Characteristic
DM/DPPC molar ratio values of 0.7 and 0.3 were determined by the projection of the breaks of the d([DM](t))/dt ) f(t) curve onto ([DM/DPPC]tot
) f(t) curve.
Table 1. Concentration of DM in the Aqueous Phase [DM]bulk
,
enzyme is constant and equal to 0.06 IU mg-1 (Figures 1B and
4). Equation 1 is simplified as follows
Surfactant(s) Mixed Aggregate Compositions [DM/DPPC]agg, and
Total DM/DPPC Molar Ratio [DM/DPPC]tot at the Break Points of
the Turbidity Curves Obtained upon Enzymatic Hydrolysis of DM
from DPPC-DM Mixed Micellesa
[DM](t) ) [DM]0 - 0.06Et/Vtot
(2)
b
break point
[DM]bulkb (mM)
[DM/DPPC]agg
[DM/DPPC]tot
where E corresponds to the enzyme content (mg).
A
B
C
0.05 ( 0.03
0.11 ( 0.06
0.03 ( 0.02
1.66 ( 0.03
1.46 ( 0.07
1.40 ( 0.07
1.71 ( 0.03
1.54 ( 0.08
1.43 ( 0.07
At each break point, A, B, and C, the total DM concentration
([DM](t)) was plotted versus the total DPPC concentration
([DPPC]) (data not shown).18,33,36 The DPPC concentration in
the aggregates is considered equal to the total DPPC content
according to its very low critical aggregation concentration (5
× 10-10M).37 Linear relationships with correlation coefficients
of 0.99 were found for each break point family depicted by the
following equation
a Calculations of total DM concentrations were done according to
eq 2. b Partition parameters were obtained from linear regression
analysis according to eq 3.
fluorescent probe, calcein, (Figure 5). Indeed, this technique
is a useful tool to demonstrate vesicle-like aggregates and to
detect their eventual closure by evaluating their ability to entrap
and retain calcein initially present in the aqueous medium.38
Three initial amyloglucosidase concentrations, 0.3 µM (Figure
5, curve a), 0.6 µM (Figure 5, curve b), and 1.2 µM (Figure 5,
curve c), were examined at a total lipid concentration of 10
mM. Chromatography analysis was performed at 5 days
enzymatic reaction. DM concentrations were determined by
RP-HPLC, leading to calculated total DM/DPPC molar ratios
in the analyzed particles of 1.35 (Figure 5, curve a), 1.04 (Figure
5, curve b), and 0.33 (Figure 5, curve c), respectively. Pure
DPPC vesicles prepared by ultrasonic irradiation in the presence
of calcein were used as a reference (Figure 5, curve d). For
the three enzyme concentrations studied, fluorescence detection
reveals a first peak with a maximum located between 14.5 and
16 mL, which is within the elution region of large particles,
and a second split peak centered at 21 mL very close to the
total elution volume of the columns (Figure 5A). The position
of the first peak is superimposed on those obtained by 90° light-
scattering detection and nearly coincides with the elution profiles
of the DPPC vesicles used as a standard (14 mL maximum
[DM](t) ) [DM]bulk + [DM/DPPC]agg [DPPC]
(3)
The concentration of DM molecules which are not associated
with the lipid [DM]bulk is given by the extrapolation to zero of
the total DPPC concentration, and the surfactant-to-lipid molar
ratio in the aggregates [DM/DPPC]agg is determined from the
slopes of the straight lines. Values of [DM]bulk and [DM/
DPPC]agg at break points A, B, and C are summarized in Table
1, as well as the total DM-to-DPPC molar ratios [DM/DPPC]tot.
The [DM]bulk values, varying from 0.03 to 0.11 mM, agree with
the cmc value of pure DM. Indeed, in the presence of DPPC,
DM (given its amphiphilic character) partitions between lipidic
assemblies and aqueous phase, and its concentration in the water
phase should range from 0 to a maximum equal to the cmc of
DM in the absence of lipid molecules (0.13 mM).24
The aggregates formed during the enzymatic process were
characterized by HPLC-GEC in the presence of a water-soluble
(36) Levy, D.; Gulik, A.; Seigneuret, M.; Rigaud, J. Biochemistry 1990,
29, 9480-9488.
(37) Marsh, D. Handbook of Lipid Bilayers; CRC Press: Boca Raton,
(38) Lesieur, S.; Grabielle-Madelmont, C.; Paternostre, M.; Ollivon, M.
1990; p 276.
Chem. Phys. Lipids 1993, 64, 57-82.

Vesicles from DPPC-DM Mixed Micelles
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10593
Figure 6. Freeze-fracture electron microscopy of (A) DPPC-DG
vesicles obtained at 5 days of enzymatic reaction from DPPC-DM
mixed micelles ([DPPC] ) 5mM, initial DM/DPPC molar ratio equal
to 1.8, [E] ) 2.4 µM, T ) 37 °C) and (B) DPPC-DG-DM (1:1.7:0.1
molar ratio) vesicles obtained by ultrasonic irradiation ([DPPC] ) 10
mM). A constant temperature of 37 °C was maintained until freezing.
Bar ) 130 nm.
Figure 5. Gel exclusion chromatography profile of aggregates formed
during the enzymatic process for a total DPPC concentration of 5 mM
(initial [DM/DPPC] ) 1.8) and in the presence of 1 mM calcein. [E]
) 0.3 µM (a), [E] ) 0.6 µM (b), and [E] ) 1.2 µM (c). Chromatography
analysis was made at 5 days of the enzymatic reaction. Total DM
concentrations of the samples are 1.67 mM (a), 5.2 mM (b), and 6.7
mM (c) from RP-HPLC. Curve d corresponds to sonicated pure DPPC
vesicles ([DPPC] ) 10 mM). (A) Fluorescence (excitation at 367 nm;
emission at 419 nm), (B) 90° light scattering at 367-nm on-line
detection. Flow rate, 1 mL/min. Void (V0) and total (Vt) volumes of
the columns are indicated by arrows. The chromatograms in A are
shifted along the y-axis by adding an arbitrary constant to the base
line.
Discussion
From the results presented in this work, it clearly appears
that the enzyme kinetics in this process are connected to the
lipid-surfactant aggregation state. To clarify the interactions
between amyloglucosidase reactivity and the supramolecular
structures, phase behavior of the system along the enzymatic
pathway from mixed micelles to vesicles was examined. The
pseudoternary DPPC-DG-DM partial phase diagram at 37 °C
was established recently by turbidimetry, performed along the
solubilization process of DPPC and DPPC-DG vesicles by DM,
and the structures of the phases confirmed by small-angle X-ray
scattering.24 A schematic representation of it is reproduced in
Figure 7. With the enrichment in DG (or the decrease of DM),
four main phase domains are crossed, a micellar solution, a
biphasic domain constituted of micelles separated from a water-
rich upper phase (clouding phenomenon), a coexistence domain
including micelles and lamellar assemblies, and a vesicle
dispersion.
elution volume) (Figure 5B). This peak thus corresponds to
closed vesicles containing calcein. The second peak is super-
imposed with the chromatogram of a 1 mM calcein solution
and is ascribable to the nonencapsulated probe molecules. The
intensity of the vesicle elution peaks detected by both fluores-
cence and light scattering increases with increasing amyloglu-
cosidase content, indicating that the higher the enzyme con-
centration, the higher the amount of closed aggregates formed
after a 5-day reaction time.
A cryofracture electron microscopy analysis was made to
visualize the enzymatically formed aggregates. Following
complete reaction, closed vesicles heterogeneous in size with a
diameter ranging from 10 to 80 nm are observed (Figure 6A).
They are comparable with standard vesicles obtained by
ultrasonication of DPPC-DG-DM mixture (1:1.7:0.1 molar
ratio) which are unilamellar and homogeneous in size (mean
diameter of 50 nm) (Figure 6B) and ref 24.
In the diagram of Figure 7, enzymatic pathways are very close
to straight lines joining binary DPPC-DM and DPPC-DG
mixtures of the same molar ratio. This is based on the isomolar
stoichiometry of DM hydrolysis into DG by amyloglucosidase
and on the comparable solubilities of both surfactants in the

10594 J. Am. Chem. Soc., Vol. 120, No. 41, 1998
Carion-TaraVella et al.
turbidity curves fluctuate around high OD values, and no
significant events, which would have been correlated, can be
observed.
To circumvent the above-mentioned limitation, the aggregates
formed upon amyloglugosidase hydrolysis were characterized
by HPLC-GEC and electron microscopy. The velocity of the
enzyme is progressively decreasing from 0.06 to 0.01 IU mg-1
in the 0.7-0.3 DM/DPPC molar ratio range (regime 2, Figures
2 and 4) and from 0.01 to nearly 0 IU mg-1 for DM/DPPC
molar ratios below 0.3 (regimes 3, Figures 2 and 4). This
concentration range coincides with the appearance of lamellar
structures in coexistence with micelles (Figure 7, 9 symbol),
that is, the higher the proportion of lamellar phase, the slower
the reaction rate. Finally, when only lamellar bilayers subsist,
the reaction is significantly slowed. Samples examined at the
same time for different enzyme concentrations by HPLC-GEC
move along the enzymatic pathway in the pure lamellar domain
of the phase diagram. The molar ratios corresponding to the
GEC samples are reported in Figure 7 (+ symbol). As the
enzymatic reaction progresses, the increase in the lamellar
structure proportion is accompanied by an increase in the
number of large aggregates capable of retaining calcein (Figure
5A). These results indicate that the enzymatic kinetic is altered
by the evolution of bilayers toward vesicle formation. This is
supported by a very low reaction rate beyond the DM/DPPC
molar ratio of 0.3 when the system is mainly composed of
vesicles (Figure 5A). The electron micrography of the ag-
gregates obtained at the end of the enzymatic reaction (Figure
6A) confirms the presence of closed vesicles within the region
in which the amyloglucosidase velocity is quite reduced to zero
(Figure 7, O symbol).
A possible explanation for the decrease and finally the
annulment of the amyloglucosidase velocity as the hydrolysis
of DM progresses may be the progressive enzyme encapsulation
inside the vesicles. However, the internal volume of vesicles
formed by this process calculated as a function of mean diameter
(10-80 nm) and total lipid concentration (10 mM) represents
only around 1% of the total dispersion volume, making this
hypothesis unlikely. A second more conceivable suggestion for
the slowing down of the reaction rate is that the DM availability
may be reduced by the occurrence of bilayered structures to
the detriment of micellar ones which may influence the DM
migration from the aggregates toward the aqueous continuum.
In favor of this hypothesis, kinetics studies have shown that
the exchange of surfactant molecules between micelles and bulk
phase and the intermicellar exchange of monomers are indeed
very fast in the case of polyoxyethylene or sugar headgroup-
based surfactants.39-41 In contrast, in the presence of lamellar
structures such as liposomes, the surfactant migration rate is
significantly limited by slow translational diffusion through the
lipid bilayer.42,43 This limitation can be related to the freedom
of movement of the surfactant molecule which is reduced from
a 3-dimensional one in the micelle solution to a 2-dimensional
one in the lamellar assemblies.44
Figure 7. Partial pseudoternary DPPC-DG-DM phase diagram in
excess buffer at 37 °C from ref 24. In such a plot (DM/DPPC as a
function of DG/DPPC molar ratio in the aggregates), the axes
correspond to the binary DPPC-DM and DPPC-DG systems, and
excess water is implied and not represented. The solid lines delimit
the main phase domains and the dotted lines indicate secondary
aggregation state transitions. The arrow indicates the enzymatic reaction
pathway on which are reported the compositions corresponding to the
turbidity break points (4), amyloglucosidase kinetic analysis (9),
samples examined by either HPLC-GEC (+) or cryofracture electron
microscopy (O).
buffer used (cmc ) 0.13 mM for DM and cmc ) 0.16 mM for
DG at 37 °C).24 According to the very low solubility of both
surfactants in the aqueous medium, their bulk concentrations
are negligible. This is partly demonstrated by the [DM]bulk
values in Table 1. Consequently, the total DM-to-DPPC or DG-
to-DPPC molar ratios ([DM/DPPC]tot or [DG/DPPC]tot, respec-
tively) can be considered to be close to the molar ratios in the
mixed aggregates ([DM/DPPC]agg or [DG/DPPC]agg).
In this study, the experiments were performed along the line
corresponding to a 1.8 total surfactant-to-lipid molar ratio and
which is materialized by the arrow in Figure 7. The identifica-
tion of the different aggregation steps involved in the enzymatic
process was investigated by using four techniques, RP-HPLC
(Figure 7, 9 symbol), turbidity measurements (Figure 7, 4
symbol), HPLC-GEC (Figure 7, + symbol), and electron
microscopy (Figure 7, O symbol).
Three regimes of reaction rates are observed as a function of
the enzymatic progress (Figures 2 and 4). In the 1.8-0.7 DM/
DPPC molar ratios range, the reaction rate is constant. The
evolution of the aggregates in this region was studied by
turbidimetry. Three main characteristic events are identified
by break points A, B, and C (Figure 3). The corresponding
compositions (Table 1) are reported on the enzymatic line in
Figure 7 (4 symbol). These events are located in the micellar
domain of the phase diagram. Point A coincides with the
boundary between the micellar solution and the biphasic region
(micelles separated from a water-rich subphase). In the biphasic
region, both points B and C concur with macroscopic phase
separations of mixed micelles from the aqueous phase (clouding
phenomenon). The correspondence of the turbidity events with
the micellar state transitions confort the use of eq 2. As a
consequence, in this region, turbidity measurements can be used
to determine the reaction progress and aggregate composition,
providing enzyme specific activity is known. Beyond point C,
(39) Malliaris, A.; Lang, J.; Sturn, J.; Zana, R. J. Phys. Chem. 1987, 91,
1475-1481.
(40) Frindi, M.; Michels, B.; Zana, R. J. Phys. Chem. 1992, 96, 6095-
6102.
(41) Frindi, M.; Michels, B.; Zana, R. J. Phys. Chem. 1992, 96, 8137-
8141.
(42) Annesini, M. C.; Di Giulio, A.; Di Marzio, L.; Finazzi-Agro, A.;
Mossa, G. J. Liposome Res. 1992, 2, 455-467.
(43) Annesini, M. C.; Di Giorgio, L.; Di Marzio, L.; Finazzi-Agro, A.;
Serafino, A. L.; Mossa, G. J. Liposome Res. 1994, 3, 639-648.
(44) Small, M. S. In The Physical Chemistry of Lipids: From Alkanes
to Phospholipids; Hanahan, D. J., Ed.; Plenum Press: New York, 1986; pp
43-83.

Vesicles from DPPC-DM Mixed Micelles
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10595
Another factor influencing DM avaibility is its partioning
between the lipid aggregates and the aqueous phase upon the
enzymatic reaction. Indeed, it has been shown that the
monomeric detergent concentration in the bulk phase increases
throughout the vesicle-to-mixed micelle transition,13,18,19,45 the
ultimate stage of the solubilization process being characterized
by mixed cmc values very close to the cmc of the pure
surfactant.12,18 The reversibility of the vesicle-to-micelle transi-
tion implies that, conversely, the surfactant content in the bulk
phase progressively decreases with vesicle formation. This was
verified for the PC-DM system.35 Under such circumstances,
DM hydrolysis by amyloglucosidase (the rate of which is
proportional to the DM bulk concentration) would be consider-
ably reduced as the vesicle formation progresses.
The change of the amyloglucosidase reaction rate as the
reaction progresses from mixed micelles to lamellar sheets is
also consistent with two relevant enzymatic studies. The first
one concerns the study of the glucose oxidase catalysis in a
micellar environment.46 It was shown that all of the rate
constants involved in the enzyme kinetics are unaffected by the
presence of surfactant micelles. The second one is the â-D-
glucosidase activity study toward OG micelles (up to 20% OG).5
The authors have shown that in the presence of pure or mixed
micelles, the enzyme activity is limited by the availability of
the detergent in the aqueous continuum and the reaction rate is
constant. Conversely, the formation of an OG/octanol/water
lamellar phase from concentrated OG micelles upon the â-D-
glucosidase reaction shows that the existence of lamellar sheets
considerably decreases the reaction rate.5,6
In conclusion, these results together demonstrate that it is
possible to move inside a ternary phase diagram using a specific
enzyme converting one of the mixture components. Starting
from the region of mixed micelles or similar “open” structures,
the removal of the surface active agent through the enzymatic
polar headgroup hydrolysis allows the system to evolve toward
vesicle formation. Enzyme-induced transformation of non-
vesicle aggregates into liposomes represents an alternative
method for encapsulation of labile substances and, from a
biochemical point of view, provides an insight into how enzymes
may mediate structural transitions in biomembranes.
Acknowledgment. We are grateful to J-P. Lechaire and G.
Frebourg. (Centre Inter-universitaire de Microscopie Electron-
ique CNRS UPR 9042, Paris) for electron microscopy
experiments.
(45) Rigaud, J.-L.; Paternostre, M.-T.; Bluzat, A. Biochemistry 1988, 27,
2677-2688.
(46) Deshaies, C.; Chopineau, J.; Moiroux, J.; Bourdillon, C. J. Phys.
Chem. 1996, 100, 5063-5069.
JA980400A