New drug delivery nanosystem combining liposomal and dendrimeric technology (liposomal locked-in dendrimers) for cancer therapy

Article abstract of DOI:10.1002/jps.22121

Liposomal locked-in dendrimers (LLDs), the combination of liposomes and dendrimers in one formulation, represents a relatively new term in the drug carrier technology. LLDs undergone appropriate physicochemical investigation can merge the benefits of lipo

Full text of DOI:10.1002/jps.22121

New Drug Delivery Nanosystem Combining Liposomal and
Dendrimeric Technology (Liposomal Locked-In Dendrimers)
for Cancer Therapy
KONSTANTINOS GARDIKIS,1,2 SOPHIA HATZIANTONIOU,2 MADALINA BUCOS,1 DIMITRIOS FESSAS,3
MARCO SIGNORELLI,3 THEODOROS FELEKIS,1 MARIA ZERVOU,1 CONSTANTINOS G. SCRETTAS,1 BARRY R. STEELE,1
MAKSIM IONOV,4 MARIA MICHA-SCRETTAS,1 BARBARA KLAJNERT,4 MARIA BRYSZEWSKA,4 COSTAS DEMETZOS2
1National Hellenic Research Foundation (N.H.R.F.), Institute of Organic and Pharmaceutical Chemistry, Vas. Konstantinou 48,
Athens 11635, Greece
2Department of Pharmaceutical Technology, School of Pharmacy, University of Athens, Panepistimioupolis, Zografou,
Athens 15771, Greece
3
`
Universita di Milano, DISTAM, via Celoria 2, Milano 20133, Italy
4Department of General Biophysics, University of Lodz, Banacha 12/16, Lodz 90-237, Poland
Received 12 October 2009; revised 11 December 2009; accepted 15 January 2010
Published online 29 March 2010 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.22121
ABSTRACT: Liposomal locked-in dendrimers (LLDs), the combination of liposomes and den-
drimers in one formulation, represents a relatively new term in the drug carrier technology.
LLDs undergone appropriate physicochemical investigation can merge the benefits of liposomal
and dendrimeric nanocarriers. In this study generation 1 and 2 hydroxy-terminated dendrimers
were synthesized and were then ‘‘locked’’ in liposomes consisting of DOPC/DPPG. The antic-
ancer drug doxorubicin (Dox) was loaded into pure liposomes or LLDs and the final products
were subjected to lyophilization. The loading of Dox as well as its in vitro release rate from all
systems was determined and the interaction of liposomes with dendrimers was assessed by
thermal analysis and fluorescence spectroscopy. The results were very promising in terms of
drug encapsulation and release rate, factors that can alter the therapeutic profile of a drug with
low therapeutic index such as Dox. Physicochemical methods revealed a strong, generation
dependent, interaction between liposomes and dendrimers that probably is the basis for the
higher loading and slower drug release from the LLDs comparing to pure liposomes. ß 2010
Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 99:3561–3571, 2010
Keywords: liposome; dendrimer; doxorubicin; in vitro release; differential scanning
calorimetry; fluorescence spectroscopy
INTRODUCTION
profile (Absorption, bioDistribution, Metabolism, and
Excretion) of a candidate drug.1
Since the late 1960s, interest in methods of drug
delivery has focused on the creation of new modifica-
tions of established drugs with the objective of getting
a drug into the patient in the simplest possible way.
The proper choice of delivery system can overcome
problems relating to solubility, can regulate bioavail-
ability and can therefore improve the overall ADME
The effectiveness of a drug can generally be
improved in cases where there is need of controlled
release in the bloodstream. This is particularly
important in the case of the treatment of certain
diseases, cancer therapy, for example, in which the
administration of low molecular weight cytostatic
drugs by themselves can cause severe side effects due
to their poor biodistribution, whereas controlled
delivery can greatly improve their therapeutic profile.
In this respect, drug delivery systems based on
nanoscale materials have the potential for minimum
release prior to reaching the target site and selective
accumulation at the desired locations in vivo due to
the enhanced permeation and retention (EPR) effect.2
Polymers and liposomes represent two of the most
thoroughly studied categories of nanoparticles with
Additional Supporting Information may be found in the online
version of this article.
This work is a part of PhD thesis of Konstantinos Gardikis MSc.
Correspondence
to:
Costas
Demetzos
(Telephone:
0302107274596; Fax: 0302107274027;
E-mail: demetzos@pharm.uoa.gr)
Journal of Pharmaceutical Sciences, Vol. 99, 3561–3571 (2010)
ß 2010 Wiley-Liss, Inc. and the American Pharmacists Association
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010
3561

3562
GARDIKIS ET AL.
potential application as carriers of bioactive mole-
cules. Liposomes, which constitute the earliest used
category of nanocarriers, are able to encapsulate
either lipophilic or hydrophilic drugs in their lipidic
chains or aqueous interior, respectively. Research
has proven that they are able to ameliorate the
pharmacokinetic and pharmacological profile of
many drugs3–5 giving way to the appearance of
several liposomal formulations in the market. The
application of liposomes, though, is limited because
of their thermodynamic instability giving rise to
phenomena such as aggregation, fusion, or drug
leakage upon storage. However, these problems have
been overcome to a large extent due to freeze
drying.6,7
Dendrimers, a so-called 4th new architectural class
of polymers, represent a much newer category of drug
delivery vehicles.8–10 The dendritic macromolecular
structure is well-defined and consists of a central core,
branching units, and terminal functional groups
which can be further chemically modified. Due to
their precise architecture, dendrimers possess an
advantage over other generally polydisperse nano-
particles and this allows for greater control over their
pharmacodynamic profile, while as vehicles for drug
delivery they can be used either for encapsulation of
bioactive compounds or for their covalent or non-
covalent attachment at the periphery. They also offer
other potential advantages such as prolongation of
drug circulation time, protection of a drug from its
surroundings, increase in drug stability (and possibly
effectiveness), and the ability to target diseased
tissue.11–14
efficiency for the acidic anticancer drug methotrex-
ate. The loading of the drug indeed increased
proportionally to dendrimer generation while the
leakage of the drug from the system decreased. The
results from this work were impressive, though little
data on the physicochemical interactions between the
components were given. An analogous study was
made by Papagiannaros et al.20 in which a PAMAM
G4–doxorubicin (Dox) complex was formed prior to
encapsulation in liposomes and the results seem
promising in terms of drug release and cytotoxic
activity against cancer cell lines. The immobilization
of anionic liposomes with PAMAM G4 has been
explored using FT-IR, X-ray diffraction, and SPR21
and the results revealed that PAMAM G4 dendrimers
may be used to fabricate porous carrier films on the
liposome surface in which ions or small molecules can
be released from liposomes and can diffuse through
the PAMAM layers. Another study by 31P NMR and
AFM, using neutral liposomes and lipid bilayers
interacting with PAMAM G7 dendrimers gave
similar results which were attributed to the forma-
tion of lipid–dendrimer aggregates.22,23
A very important method for studying interactions
in complex systems such as LLDs is differential
scanning calorimetry (DSC) and in recent years our
groups have published several reports of thermal
analysis data for the interaction of dendrimers with
model lipid membranes or liposomes.24–26 Despite the
numerous methods applied for dendrimer–lipid
interactions there is certainly a gap in the literature
concerning the physicochemical characterization of
LLDs.1
Although the first attempt to incorporate a drug
into dendrimers was done in 198915, it was only in
2001 that a combination between dendrimers and
liposomes and the study of the interactions of the
components took place for the first time.16 Liposomal
locked-in dendrimers (LLDs) technology—liposomes
incorporating dendrimers—is a relatively new term
in the drug delivery literature. Locked-in dendrimers
may be viewed as a dendrimer-based class of
modulatory liposomal controlled release systems
(MLCRS) leading to high entrapment and modifica-
tion of the release profile of bioactive molecules from
liposomal vesicles.1 It has been established that
liposomal formulations of certain anticancer agents
are extremely sensitive to the drug release rates, with
the slowest releasing systems exhibiting the best
efficacy profiles.17,18 Therefore, for liposomal formu-
lations, it is very important to control the drug release
rate.
In this study new synthetic generation 1 and 2
polyether–polyester dendrimers (PG1&PG2) were
incorporated into liposomes consisting of DOPC/
DPPG. The anticancer drug Dox was loaded into
pure liposomes or LLDs and the final products were
subjected to lyophilization. The loading of Dox as well
as its in vitro release rate from all systems was
determined and the interaction of liposomes with
dendrimers was assessed by thermal analysis and
fluorescence spectroscopy.
MATERIALS AND METHODS
PG1&PG2 Dendrimer Synthesis
We have synthesized two new hydroxy-terminated
dendrimers (Scheme 1) containing an aliphatic
polyether–polyester backbone. Both bear branches
composed of glycerol and acetic acid monomers. The
choice of building blocks was based on a requirement
for biocompatible, biodegradable, and water soluble
compounds.
The milestone for the creation of the liposomal
locked in dendrimer concept was the work by
Khopade et al.19 In that study, cationic PAMAM
dendrimers were incorporated in the aqueous interior
of liposomes in order to increase the encapsulation
Following a divergent strategy we prepared den-
drimers G1 and G2. Using the pentaerythritol
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010
DOI 10.1002/jps

NEW DRUG DELIVERY NANOSYSTEM
3563
Scheme 1. Structure of G1 and G2 PG dendrimer.
molecule as the core, esterification with 2-[1,3–
bis(benzyloxy)propan-2-yloxy]acetic acid gave G1-
Bn. Removal of the benzyl groups by hydrogenolysis
led to G1-OH. Following the same procedure, we also
prepared G2-OH (Scheme 2).
We chose to use the benzyl ether moiety as
protective group for its additional ability to serve
as an NMR ‘‘tag’’ for the characterization of the
growing molecule.
tal setup. Elemental analyses were performed at the
National Hellenic Research Foundation in Athens
using a Perkin Elmer (Massachusetts, USA) PE2400
II analyzer. MS analyses were performed using a TSQ
7000 Finnigan MAT instrument operating in ESI
mode.
Synthesis of 1,3-Bis(Benzyloxy)-2-Propanol (1)
1,3-bis(benzyloxy)-2-propanol was prepared by the
reaction of 2 equiv. of benzyl alcohol with epichlor-
ohydrin according to a literature procedure.28
General Procedures
All reactions requiring dry or inert conditions were
carried out in flame-dried glassware under an atmo-
sphere of argon. Pd/C (10%, w/w) was purchased
from Sigma Aldrich (St Louis, MO, USA), N,N0-
dicyclohexylcarbodiimide (DCC) from Merck (Darm-
Synthesis of 2-{2-(Benzyloxy)-1-[(Benzyloxy)Methyl]
Ethoxy}Acetic Acid (2)
To a suspension of NaH (5.2 g, 130 mmol, washed
with toluene) in 120 mL dry THF, were added slowly
27.2 g of compound 1,3-bis(benzyloxy)-2-propanol
(100 mmol). Dry sodium a-chloroacetate (12.5 g,
108 mmol) was added and the solution then refluxed
for 48 h. After removal of organic solvent, the solid
residue was dissolved in hot water. The aqueous
solution was extracted with toluene and hexane. The
aqueous phase was acidified with conc. aq. HCl and
then extracted with CH2Cl2. After removal of organic
solvent 25.7 g of the product were obtained as a yellow
liquid (yield 78%).29 1H-NMR (CDCl3): d 3.56 (d,
J ¼ 5.4 Hz, 4H, CH–CH2), 3.79 (q, J ¼ 5.4 Hz, 1H, CH),
stadt,
Germany).
4-(dimethylamino)pyridinium
p-toluenesulfonate (DPTS) was prepared according
to the literature.27 Solvents were dried under argon
by conventional methods. (THF distilled over sodium
benzophenone and DMF over molecular sieve 40A.)
Reactions were monitored by TLC (Merck Kieselgel
60 F254). After aqueous work-up of reactions
mixtures, organic solutions were routinely dried over
anhydrous sodium sulfate. Column chromatography
was carried out on Kieselgel 60 (particle size 40–
63 mm) as supplied by Merck. Size exclusion chroma-
tography (SEC) was performed using Bio-Beads SX1
Beads 200-400Mesh from Bio-Rad (California, USA).
HR NMR experiments were acquired to Varian 600
and 300 MHz spectrometers at 258C. Compounds
were dissolved in CDCl3 or D2O. The 2D 1H-1H DQF-
COSY, 1H-13C edited-HSQC, and 1H-13C HMBC
experiments assisted structure characterization. Ex-
perimental data were processed using VNMR rou-
tines. Chemical shifts (d) are reported in ppm while
spectra were referenced by the standard experimen-
–
4.31 (s, 2H, CH –C O), 4.57 (s, 4H, benzyl-CH ), 7.37
–
2
2
(m, 10H, arom.), 13C NMR (CDCl3): d 68.45, 69.67,
73.46, 79.21, 127.68, 127.88, 128.39, 136.92, 172.8
–
(C O).
–
Synthesis of G1-Bn (3)
9.9 g (30 mmol) of 2-{2-(benzyloxy)-1-[(benzyloxy)-
methyl]ethoxy}acetic acid (2), 5.1 g (5.1 mmol) of
pentaerythritol, and 1.76 g (6 mmol) DPTS were
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010

3564
GARDIKIS ET AL.
Scheme 2. Divergent synthesis of G1 and G2 PG dendrimer.
dissolved in 50 mL of dry CH2Cl2. The mixture was
flushed with argon and 6.2 g (30 mmol) DCC was
added. The reaction mixture was stirred for 72 h at
room temperature under an argon atmosphere.
After filtration of urea, extraction with CH2Cl2,
washing with water and drying over Na2SO4, an
oily residue was obtained which after purification
by silica gel column chromatography (hexane/ethyl
acetate: 70/30) gave the product as an oil in a yield
of 4.6 g (60%). 1H-NMR (CDCl3): d 3.64 (d, J ¼ 9.6 Hz,
16H, CH2-6), 3.85 (q, J ¼ 9.6 Hz, 4H, CH-5), 4.08
(s, 8H, CH2-2), 4.39 (s, 8H, CH2-4), 4.54 (s, 16 H,
benzyl-CH2), 7.34 (m, 40H, arom.), 13C NMR (CDCl3):
d 42.0 (C1), 62.1(C2), 67.7(C4), 70.6 (C6), 73.4
(C7benzyl), 78.7(C5), 127.7, 128.4, 138.1(aþrom.),170.2
Synthesis of G1
To a solution of 0.2 g (0.144 mmol) of G1-Bn in a
mixture of 3 mL CH2Cl2 and 8 mL EtOH was added
Pd/C (10%, w/w). The flask was first evacuated and
then filled with H2 and the reaction mixture stirred
for 4 h at RT. After completion of the reaction, the
catalyst was removed by filtration through Celite and
washed with EtOH. Evaporation of the filtrate gave
0.07 g (73%) of the desired product as a colorless
viscous oil.
1H NMR (D2O) d: 3.42 (s, 8H, CH2-2), 3.45 (m, 4H,
CH-5), 3.48 ((dd J ¼ 11.8, 5.6 Hz), 8H, CH2-6), 3.55
((dd J ¼ 11.8, 3.8 Hz), 8H, CH2-6), 4.07 (s, 8H, CH2-4);
13C NMR (D2O) d: 45.3 (C1), 60.6 (C6), 60.9 (C2), 67.6
(C4), 81.5 (C5), 176.2 (C3) ESI m/z: 687.5 (M þ Naþ),
(theory: 664.2 M). Anal. Calcd for C25H44O20: C, 45.18;
H, 6.67. Found: C, 45.00; H 6.45.
–
(C3, C O), ESI m/z: 1408.9 (MH þ Na ), (theory:
–
1385.6 MH). Anal. Calcd for C81H92O20: C, 70.21; H,
6.69. Found: C, 70.18; H 6.57.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010
DOI 10.1002/jps

NEW DRUG DELIVERY NANOSYSTEM
3565
Synthesis of G2-Bn (5)
using the reverse phase evaporation method (REV).30
Sonication was applied to afford SUVs with reduced
P.I. and the extraliposomal pH was changed to 7.5
through gel permeation chromatography using a
Sephadex G75 column with PBS 10 mM pH 7.5/
150 mM sucrose as a mobile phase.
Dox was loaded to pure liposomes or LLDs by
incubation in room temperature for 1 h. Unentrapped
Dox was removed by gel permeation chromatography
(Sephadex G75).
G2-Bn was obtained following a procedure similar to
that described for compound G1-Bn. 0.390 g
(0.58 mmol) of G1, 2.29 g (5.1 mmol) of 2-{2-(benzy-
loxy)-1-[(benzyloxy)methyl]ethoxy}acetic acid (2) and
0.733 g (6 mmol) DPTS were dissolved in 22 mL of dry
THF and (1.74 mL) dry DMF. The mixture was
flushed with argon and 1.44 g (30 mmol) DCC was
added. The reaction mixture was stirred for 10 days at
room temperature under an argon atmosphere. After
filtration of urea and concentration to remove solvent,
an oily residue was obtained which was purified by
SEC (CH2Cl2) in a yield of 40% (0.73 g).
Freeze Drying of Liposomal Suspensions
Free or Dox-loaded liposomes were frozen at À808C
overnight and were subjected to lyophilization in
order to overcome stability issues concerning liposo-
mal suspensions.31 The lyophilization was achieved
using a freeze drier (TELSTAR Cryodos-50, Terrassa,
Spain) under the following conditions: condenser
temperature from À508C, vacuum 8.2 Â 10À2 mb).
Reconstitution was made by adding the appropriate
amount of HPLC-grade water.
1H NMR (CDCl3) d: 3.61 ((dd, J ¼ 14 Hz, 4.5 Hz),
32H, CH2-10), 3.7 ((q, J ¼ 4.8 Hz), 4H, CH-5), 3.82 (m,
8H, CH-9), 4.04 (s, 8H, CH2-2), (4.04-4.23) (m, 16H,
CH2-6), 4.29 (s, 8H, CH2-4), 4.33 (s, 16H, CH2-8), 4.5
(br.s, 32H, CH2-benzyl), (7.25–7.31) (m, 80H, ar-H);
13C NMR (CDCl3) d: 42.1 (C1), 62.0 (C2), 62.8 (C6),
67.7 (C4), 67.8 (C8), 70.6 (C10), 73.4 (C11), 75.8 (C5),
78.7 (C9), 127.5, 128.4 (phenyl C13-C15), 138 (phenyl
C12), 170.1 (C3), 170.4 (C7) ESI m/z: 3.200.4
(M þ Kþ), 3.161.4, (theory: 3.161.3 M).
Characterization of Free and Doxorubicin-Loaded
Liposomes
Anal. Calcd for C177H204O52: C, 67.20; H, 6.50.
Found: C, 66.9; H 6.35.
The hydrodynamic diameter of empty and Dox-loaded
liposomes was measured by light scattering. Size and
z-potential of liposomes are the parameters that
indicate their physical stability. 100 mL of the
liposomal suspension was 30-fold diluted in HPLC-
grade water (pH 5.6–5.7) immediately after prepara-
tion or after reconstitution and z-average mean and z-
potential of the empty and Dox-loaded liposomes were
measured. Samples were scattered (633 nm) at 908,
and measurements were made at 258C in a photon
correlation spectrometer (Zetasizer 3000 HS, Mal-
vern Instruments, Malvern, UK) and analyzed by the
CONTIN method (MALVERN software).
Dendrimer and lipid quantification was done by
HPTLC-FID (Iatroscan)32 with chloroform/methanol/
water 60:20:3.2 (v:v) as a mobile phase.
The incorporation of Dox into liposomes was
determined by UV spectrometry (UV-1700, UV–
Visible Spectrophotometer, SHIMADZU, Pharmas-
pec, Kyoto, Japan) at wavelength 480 nm after the
addition of methanol to the liposomal suspension,
with the aid of a Dox calibration curve in methanol.
Pure methanol was used as blank.
Synthesis of G2
G2-OH was obtained following a similar procedure to
that used for compound 4. To a solution of 0.5 g
(0.158 mmol) of G2-Bn in a mixture of 5 mL THF and
5 mL MeOH was added Pd/C (10%, w/w). The flask
was first evacuated and then filled with 50 psi of H2
and the reaction mixture stirred for 20 h at room
temperature. After completion of the reaction, the
catalyst was removed by filtration through Celite and
washed with MeOH and THF. Evaporation of the
filtrate gave 0.245 g (yield 90%) of the desired product
1
as a colorless viscous oil. H-NMR (D2O) d: 3.47 (s,
8H, CH2-2), 3.62-3.51 (m, 60H, CH-5,9, CH2-6,10),
4.21 (s, 24H, CH2-4,8); 13C NMR (D2O) d: 42.1 (C1),
60.5 (C6, C10), 60.9 (C2), 66.8 (C4, C8), 81.5 (C5, C9),
174.7(C3, C7). ESI m/z: 1744.5 (M þ Naþ), (theory:
1721.52 M). Anal. Calcd for C65H108O52: C, 45.35; H,
6.32. Found: C, 45.55; H 6.45.
Pure Liposome and Liposomal Locked in Dendrimer
Preparation
In Vitro Release Studies
The liposomes prepared in this study consisted of
DOPC and DPPG at a 10:0.6 ratio (the lipid system
from now on will be referred to as just DOPC). In the
case of LLDs, dendrimeric solutions in methanol were
mixed with the lipid solutions. The initial dendrimer/
lipid molar ratio was 0.1. (NH4)2SO4 150 mM, pH 5.5,
with 150 mM sucrose as a cryoprotectant was added
and the mixture was vortexed until the induction of a
homogenous emulsion. MLV preparation was made
Dox-loaded Liposomal suspensions were placed in
12000 MWCO 25 mm width dialysis sacks (Sigma
Aldrich, St Louis, MO, USA). Dialysis sacks were
inserted in RPMI 5% medium in shaking water bath
(Selecta, Barcelona, Spain) set at 378C. Aliquots of
samples (1 mL) were taken from the external solution
at specific time points and that volume was replaced
with RPMI incubated at 378C. Dox concentrations
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010

3566
GARDIKIS ET AL.
were measured with UV spectrometry after the
addition of 2 mL HPLC-grade water. As reference
sample RPMI at 378C, diluted threefold with HPLC-
grade water was used. The cumulative percentage of
drug release was calculated and plotted versus time
using the equation:
dendrimers. The measurements were performed with
a TA Instruments DSC 2920. The lipid bilayer
samples where prepared by hydration of dry lipid/
dendrimer mixtures at appropriate ratios (0%, 3%,
10%, 20% of dendrimer content) in excess of HPLC-
grade water and were placed in stainless steel
pressure-resistant 60 mL pans which were sealed.
Empty pan was used as the reference. Two cooling-
heating cycles were performed from 25 to À508C at
28C/min scanning rate. The second heating run was
taken into account.
½DoxŠ
released
% Released Dox ¼
½DoxŠ
initial
Membrane Fluidity Measurements
The raw data were worked out with the software
‘‘THESEUS’’33 dedicated for handling raw calori-
metric data. Briefly, the output signal in mW units
was divided by the sample lipid mass and by the
heating rate to be converted into apparent heat
capacity, CaPpp in kJ KÀ1 molÀ1 units. The trace of CaPpp
was finally scaled with respect to the baseline to
obtain the excess (with respect to the low temperature
lipids state) specific heat, CePxðTÞ. The area underlying
the recorded peaks, so treated, directly corresponds to
the lipid phase transitions relevant enthalpy in
kJ molÀ1 units. Errors were evaluated on the basis
of at least three replicas.
The change of fluidity of the lipid bilayer due to
increasing dendrimer incorporation in the membrane
was measured using a steady-state fluorescence pol-
arization technique. Two different fluorescent probes
were used: 1,6-diphenyl-1,3,5-hexatriene (DPH), an
apolar molecule which is incorporated into the hydro-
phobic region of the liposome bilayer with its long axis
parallel to the acyl chains, and 1-(4-trimethylammo-
niumphenyl)-6-phenyl-1,3,5-hexatriene (TMA-DPH),
which is anchored at the surface of the liposome
bilayer in contact with the water due to its positively
charged amino groups. Since DPH probe is incorpo-
rated deeper into the lipid bilayer than TMA-DPH,
the use of both probes in the same lipid membrane
allows for the comparison of membrane order at
different depths of the bilayer. Measurements were
made using a Perkin Elmer luminescence spectro-
meter LS-50B equipped for fluorescence polarization
measurement. Three hundred micromolar liposomal
suspension was added to the cuvette followed by the
addition of 1 mM DPH (in tetrahydrofuran) or TMA-
DPH (in methanol). The sample was stirred well and
incubated in the dark at room temperature for 20 min.
The cuvette holder was temperature controlled by
water thermostat (MLW-U) with 378C. The readings
were taken at intervals of 2 s. The fluorescence aniso-
tropy values (r) of the samples were calculated by the
fluorescence data manager program FL WinLab
The Perkin-Elmer corporation, Version 3.00 using
the following equation:
Statistical Analysis
Results are shown as mean ꢀ SD of n ¼ 3 independent
experiments. To analyze differences in variables
before and after freeze drying the paired Student’s
t-test was used. Comparison between different groups
was done using One-way ANOVA followed by Tukey
multiple comparison test when equal variances were
assumed and Dunnett’s C multiple comparison test
when equal variances were not assumed. In order to
assess the correlations between the pairs of the
variables parameters of the regression line were
estimated together with the regression coefficient r.
The no correlation hypothesis was rejected on the
p ¼ 0.05 significance level. p-Values <0.05 were
considered statistically significant. Statistical analy-
sis was done using SPSS 14.0.
ðIVV À GIVH
Þ
RESULTS AND DISCUSSION
r ¼
ðIVV þ 2GIVH
Þ
Characterization of Free and Dox-Loaded Liposomes
where IVV and IVH are the vertical and horizontal
fluorescence intensities, respectively, to the vertical
polarization of the excitation light beam. The factor
G ¼ IHV/IHH (grating correction factor) corrects the
polarizing effects of the monochromator. The excita-
tion wavelengths were 348 and 340 nm the fluores-
cence emission was measured at 426 and 430 nm nm
for DPH and TMA-DPH, respectively.
The physicochemical characteristics of free and Dox-
loaded liposomes and LLDs before freeze drying or
after reconstitution can be seen in Table 1. PG1
incorporation in DOPC liposomes did not affect their
size, z-potential or P.I. Respective data were obtained
for PG2 LLDs, except the case of vesicular size that
was augmented from 70 to 96.7 nm. This fact could be
due to bigger dendrimer size and higher encapsulated
dendrimer/lipid ratio in the case of PG2 leading to
higher level pressure of the aqueous interior of the
liposome and consequently to larger vesicles.34 Dox
Thermal Analysis Measurements
The DSC method was used for the stability char-
acterization of the DOPC lipid bilayers incorporating
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010
DOI 10.1002/jps

NEW DRUG DELIVERY NANOSYSTEM
3567
Table 1. Physicochemical Characteristics and Component Ratios of Final Nanosystems
Size
z-Potential
Nanosystem
DOPC
(nm)
SD
PI
SD
(mV)
SD
Dox/Lipid
0.19
SD
PG/Lipid
SD
Lyophilization
70
4.4
1.6
0.4
10.7
2.6
1.1
0.273
0.272
0.308
0.386
0.271
0.246
0.274
0.369
0.298
0.285
0.316
0.382
0.012
0.021
0.029
0.028
0.01
0.006
0.016
0.038
0.129
0.054
0.043
0.037
À24.4
À30
5.3
1.8
4.6
6.9
0.1
0.7
1.3
3.2
18.9
5.7
1.6
5.2
Before
After
Before
After
Before
After
Before
After
Before
After
Before
After
71.2
73.2
106.5
62.3
67.9
63.9
112.6
96.7
96
DOPC/DOX
À22.8
À26.1
À19
0.02
DOPC/PG1
À22
DOPC/PG1/DOX
DOPC/PG2
2.8
À19.1
À17.8
À32.5
À29.4
À19.6
À19.1
0.24
0.28
0.02
0.03
0.21
0.31
0.03
0.12
15.6
14.2
4.2
16.6
20.1
DOPC/PG2/DOX
93.8
109.6
incorporation in pure liposomes or LLDs did not affect
any of the physicochemical characteristics of the
empty systems. Reconstitution after freeze drying
was successful in all cases, with Dox-loaded vesicles
presenting size increase and higher P.I. index, as seen
elsewhere in the literature.35 This could be explained
by a slight leakage of Dox during freeze drying due to
thermomechanical stress during freezing stage on the
membrane affecting its permeability.36 As reported in
the literature37 extraliposomal Dox may induce
aggregation of fusion of negatively charged vesicles
leading to higher mean diameter and P.I. values. In
our case the slight increase of the parameters above
should be due to partial vesicular fusion. In the case of
aggregated particles, according to Fonseca et al., the
extraliposomal Dox can be removed by chromatogra-
phy or cation exchange resins and the process
becomes reversible. In our case there was no
reversibility when removing the extraliposomal
Dox, which furthermore accounted to <5%. From
these remarks it is concluded that the leakage of Dox
during freeze drying leads to fusion of the lipidic
vesicles while probably, an amount of free Dox gets
reencapsulated inside the fused vesicles during the
process.
Dendrimer locking into liposomes was almost
quantitative, as measured by HPTLC-FID. Although
there was significant phospholipid loss during pro-
duction steps, more than 90% of initial dendrimer was
entrapped in the liposomal vesicle leading to a
dendrimer/lipid ratio of 0.21 ꢀ 0.03 and 0.31 ꢀ 0.12
for PG1 and PG2 LLDs, respectively (see Tab. 1).
Freeze drying of LLDs did not affect dendrimer
entrapment, as more than 97% of dendrimers
remained in the vesicle as measured by HPTLC-
FID after gel permeation chromatography.
0.19 ꢀ 0.02) and even higher in the case of PG2 LLDs
(0.28 ꢀ 0.03). These results can be explained by the
size difference between PG1 and PG2, with PG2
bearing more chemical groups and steric space able to
interact with bioactives like Dox in a stronger
manner. These results confirm the findings of
Khopade et al.19 In that work, methotrexate/lipid
ratio increased by increasing PAMAM dendrimer
generation, reaching a plateau at generation 5. This
fact was explained either by the basicity caused by
PAMAM dendrimers establishing a pH gradient, or
by the interaction of final –NH2 groups of PAMAM
with –COOH of methotrexate. In our case, since PG
dendrimers are uncharged, interactions with Dox
should involve hydrogen bonding and hydrophobic
interactions.
Freeze drying did not affect Dox entrapment, as is
reported in the literature.39 More than 95% of Dox
remained in the vesicle, as measured by UV–Vis after
gel permeation chromatography following reconstitu-
tion.
In Vitro Release Studies
The entrapment of dendrimers in the liposome
affected significantly the in vitro release rate of
encapsulated Dox. Although there was fast leakage
during the first hour of the experiment for both pure
liposomes and LLDs (32.6 ꢀ 1.5% for DOPC and
19.7 ꢀ 5.4% and 18.7 ꢀ 1.9% for DOPC/PG1 and
DOPC/PG2, respectively) the release of Dox from
LLDs after that time point was significantly lowered
comparing to pure liposomes. At 96 h the cumulative
release for pure liposomes was 74.6 ꢀ 7.8% while for
DOPC/PG1 it was 32.2 ꢀ 1% and for DOPC/PG2 it was
27.9 ꢀ 2.8% (see Fig. 1). From the data obtained it is
obvious that the interaction of dendrimers with Dox
created the appropriate force for the latter to be
maintained in the interior of the liposome. These
findings are in accordance with Khopade et al.19 who
found significant lowering of methotrexate release
from LLDs compared with pure liposomes. Khopade
Dox loading to either pure liposomes or LLDs using
ammonium sulphate gradient was in the range of 95%
as reported in the literature.38 Interestingly, Dox/
lipid ratio was found to be higher when comparing
PG1 LLDs to pure liposomes (0.24 ꢀ 0.02 against
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010

3568
GARDIKIS ET AL.
Figure 2. Fluorescence anisotropy of DPH probe incor-
porated in DOPC liposomes with increasing concentrations
of PG1 dendrimer.
Figure 1. In vitro release of Dox-loaded pure liposomes or
LLDs in RPMI 5%.
et al. also noticed dependence of release rate on
dendrimer generation, comparing generations 2, 3,
and 4. In our case the difference between generations
1 and 2 was not significant and this fact should be
attributed to the open structure of small generation
dendrimers (below generation 3).40,41 Such open
structure could not provide the appropriate dendri-
mer conformation for encapsulation of relatively big
molecules such as Dox. We may thus assume that
there is no internalization of Dox in the dendrimer
cavity but more likely the formation of Dox–dendri-
mer network, meaning Dox molecules surrounded by
and interacting with dendrimer molecules. In this
kind of network the size of dendrimer and the
dendrimer to lipid molar ratio would be less
significant than the conformation of generation 1
and 2 dendrimers that should be similar for both
dendrimers due to their similar chemistry and is
affected by the greater mobility of dendrimer
molecules at 378C compared to 258C that the Dox
loading process takes place. The fast release of Dox
from LLDs during the first hour should be due to the
liquid crystalline structure of DOPC membranes at
378C. Liposomes in liquid crystalline state bear lipid
chains in gauche conformation and in a state of great
mobility, permitting Dox to cross the membrane
easily.42 Thus, probably, during the first hour of the
experiment the Dox fraction in the interior of the
liposome that does not interact with dendrimers leaks
out of the vesicle leading to high release rate. From
then on the release rate diminishes as the inter-
liposomal Dox inside the liposome is in dendrimer
‘‘bound’’ form.
dendrimer incorporation (data not shown). PG1
exhibited a significant interaction with the apolar
part of the membrane, inducing a concentration-
dependent fluidization of the membrane (Fig. 2). The
DPH anisotropy decreased until a PG1/DOPC ratio of
0.09 and then remained constant, meaning that
higher dendrimer concentration did not induce more
fluid membrane. In the case of PG2 (that is
significantly bigger compared to PG1) the data were
very scattered probably due to the induction of
membrane phase separation (see thermal analysis
chapter). Thus clear anisotropy conclusions cannot be
obtained for the DOPC/PG2 system (Fig. 3).
Thermal Analysis Measurements
The incorporation of PG1 and PG2 dendrimers into
the DOPC lipid bilayers affected significantly the
thermotropic behavior of the membrane lowering the
Tm, as can be seen in Figures 4 and 5. On the other
hand, the DH of the transition was not affected by the
incorporation of the dendrimers (all enthalpies were
in the order of 36 ꢀ 2 kJ molÀ1) leading to the
conclusion that the destabilization effect induced by
dendrimer incorporation was predominantly of entro-
pic nature. This is in line with the lack of interactions
of the lipid polar groups with the dendrimers
Membrane Fluidity Measurements
Fluorescence spectroscopy was applied in order to
estimate the interaction of PG1 and PG2 dendrimers
with DOPC liposomes. The polar head groups of
phosphatidyl choline did not seem to interact
significantly with either dendrimer as TMA-DPH
anisotropy values did not change significantly upon
Figure 3. Fluorescence anisotropy of DPH probe incor-
porated in DOPC liposomes with increasing concentrations
of PG2 dendrimer.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010
DOI 10.1002/jps

NEW DRUG DELIVERY NANOSYSTEM
3569
hypothesis that the high flexibility (without compro-
mise the enthalpic interactions) allow more phases to
be hosted within the outer envelope of the liposomes
in the presence of dendrimers. Addition of Dox could
imply further possibilities of molecular arrangements
pushing the system to a more stable thermodynamic
state (including the migration of some Dox þ
dendrimer complexes toward the liposomal aqueous
core). This is maybe is on the basis of the higher Dox
encapsulation efficiency of DOPC/PG2 system.
CONCLUSION
Figure 4. DSC thermograms of DOPC lipid bilayers
in the presence of increasing concentrations of PG1
dendrimer.
Despite the numerous advantages of dendrimers and
liposomes, a combination of them has appeared very
few times in the literature. LLDs could be considered
as an efficient class of MLCRS with a great potential
for carrying a high drug load and modifying the drug
release rate from the system. The modification of the
above parameters may increase the therapeutic index
profile of the carried bioactive substance.
mentioned above (see fluorescence spectroscopy).
These destabilizing effects are much more pro-
nounced in the case of PG2 which permits the system
to form more than one phase (see Fig. 5) with different
stability and order parameters. Similar destabilizing
effects concerning the thermal interaction between
dendrimers and liposomes have been documented
in the literature.24,25 This tendency, in fact, may
prevent a clear measurement with the fluorescence
anisotropy method (see Fig. 3). The lowering of the Tm
of the membrane due to incorporated dendrimers
leads to the conclusion that PG1 and, especially, PG2
induce a more fluid lipid bilayer, that reflects a higher
mixing affinity of the dendrimers in the liquid
crystalline phase.43,44 This is favorable in the case
of DOPC liposomes, since vesicular incorporation of
dendrimers and Dox as well as in vivo administration
of the final product take place with vesicles in the
liquid crystalline phase.
In this work liposomes and new polyether–polye-
ster dendrimers were combined to afford a novel
lyophilized LLD system incorporating the potent
anticancer drug Dox. The LLDs exhibited high drug
loads and slower in vitro release rate compared to
pure liposomes. In order to study the physicochemical
interactions between lipidic components and dendri-
mers, fluorescence spectroscopy and thermal analysis
were applied. The combination of those two techni-
ques revealed a strong, concentration dependent,
interaction between the lipidic chains of DOPC and
PG1 or PG2 dendrimers. On the other hand, the
interaction of the polar head groups of the lipids with
both dendrimers was found to be not significant.
Furthermore the interaction between lipids and
dendrimers is of entropic nature, as derived from
thermal analysis results, leading to the conclusion
that probably there is no bond formation between
lipids and dendrimers but the interactions are of
steric nature and are favored in the interior of the acyl
chains. PG2 dendrimer at high concentrations
induces a clear phase separation of the lipid bilayer
due to its larger size, probably leading to higher
encapsulation percentage into the more thermody-
namically favorable aqueous interior of a unilamme-
lar liposomal vesicle.
Although lipid bilayers and liposomes are formed in
different regions on the DOPC/DPPG/water/dendri-
mer phase diagram, one may nonetheless advance the
The loading of Dox to the liposomal locked in
dendrimers through pH gradient method was highly
successful. Apart from the protonation of Dox shifting
the equilibrium of free to encapsulated Dox to the
encapsulated form, in the systems we studied there
was a second force, incorporated dendrimers, which
increased the antracycline’s loading. Probably the
complex formed by dendrimers and Dox while cross-
ing the membrane finds a more thermodynamically
Figure 5. DSC thermograms of DOPC lipid bilayers
in the presence of increasing concentrations of PG2
dendrimer.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010

3570
GARDIKIS ET AL.
stable location at the interior of the liposome leading
to high drug loading and slower release compared to
pure liposomes. The interaction between Dox and
LLDs is the subject of a forthcoming publication by
our group.
To summarize, LLDs seem like a promising system
for the delivery of bioactive molecules. In the case of a
drug with low therapeutic index such as Dox, high
drug to carrier ratios and minimum release in the
bloodstream prior to reaching the target tissue are of
utmost importance and LLDs seem to be able to help
in both directions. Care has to be taken, though, to the
greatest consideration of physicochemical interac-
tions of such complicated systems in order to
rationally design stable, safe, and efficient drugs.
16. Purohit G, Sakthivel T, Florence AT. 2001. Interaction of
cationic partial dendrimers with charged and neutral lipo-
somes. Int J Pharm 214:71–76.
17. Charrois GJR, Allen TM. 2004. Drug release rate influences
the pharmacokinetics, biodistribution, therapeutic activity,
and toxicity of pegylated liposomal doxorubicin formulations
in murine breast cancer. Biochim Biophys Acta Biomembr
1663:167–177.
18. Drummond DC, Meyer O, Hong KL, Kirpotin DB, Papahadjo-
poulos D. 1999. Optimizing liposomes for delivery of chemo-
therapeutic agents to solid tumors. Pharmacol Rev 51:691–743.
19. Khopade AJ, Caruso F, Tripathi P, Nagaich S, Jain NK. 2002.
Effect of dendrimer on entrapment and release of bioactive
from liposomes. Int J Pharm 232:157–162.
20. Papagiannaros A, Dimas K, Papaioannou GT, Demetzos C.
2005. Doxorubicin-PAMAM dendrimer complex attached to
liposomes: Cytotoxic studies against human cancer cell lines.
Int J Pharm 302:29–38.
21. Moraes ML, Baptista MS, Itri R, Zucolotto V, Oliveira ON, Jr.
2008. Immobilization of liposomes in nanostructured layer-by-
layer films containing dendrimers. Mater Sci Eng C28:467–
471.
REFERENCES
1. Tekade RK, Kumar PV, Jain NK. 2009. Dendrimers in oncol-
ogy: An expanding Horizon. Chem Rev 109:49–87.
2. Gullotti E, Yeo Y. 2009. Extracellularly activated nanocarriers:
A new paradigm of tumor targeted drug delivery. Mol Pharm
6:1041–1051.
3. Abraham SA, Waterhouse DN, Mayer LD, Cullis PR, Madden
TD, Bally MB. 2005. The liposomal formulation of doxorubicin.
In: Nejat D, editor. Methods in enzymology, edition. San Diego,
CA, USA: Elsevier Academic Press. pp. 71–97.
22. Mecke A, Uppuluri S, Sassanella TM, Lee D-K, Ramamoorthy
A, Baker JJR, Orr BG, Banaszak Holl MM. 2004. Direct
observation of lipid bilayer disruption by poly(amidoamine)
dendrimers. Chem Phys Lipids 132:3–14.
23. Hong S, Hessler JA, Banaszak Holl MM, Leroueil P, Mecke A,
Orr BG. 2006. Physical interactions of nanoparticles with
biological membranes: The observation of nanoscale hole for-
mation. J Chem Health Saf 13:16–20.
24. Gardikis K, Hatziantoniou S, Viras K, Wagner M, Demetzos C.
2006. A DSC and Raman spectroscopy study on the effect of
PAMAM dendrimer on DPPC model lipid membranes. Int J
Pharm 318:118–123.
4. Massing U, Fuxius S. 2000. Liposomal formulations of antic-
ancer drugs: Selectivity and effectiveness. Drug Resist Updates
3:171–177.
5. Schiffelers RM, Storm G. 2008. Liposomal nanomedicines as
anticancer therapeutics: Beyond targeting tumor cells. Int J
Pharm 364:258–264.
25. Klajnert B, Epand RM. 2005. PAMAM dendrimers and model
membranes: Differential scanning calorimetry studies. Int J
Pharm 305:154–166.
6. Hatziantoniou S, Dimas K, Georgopoulos A, Sotiriadou N,
Demetzos C. 2006. Cytotoxic and antitumor activity of lipo-
some-incorporated sclareol against cancer cell lines and human
colon cancer xenografts. Pharmacol Res 53:80–87.
7. Miyajima K. 1997. Role of saccharides for the freeze-thawing
and freeze drying of liposome. Adv Drug Deliv Rev 24:151–159.
8. Buhleier E, Wehner W, Vogtle F. 1978. Cascade-chain-like and
nonskid-chain-like syntheses of molecular cavity topologies.
Synthesis (Stuttgart) 2:155–158.
9. Tomalia DA, Naylor AM, Goddard WA. 1990. Starburst den-
drimers—Molecular-level control of size, shape, surface-chem-
istry, topology, and flexibility from atoms to macroscopic
matter. Angew Chem Int Ed Engl 29:138–175.
26. Klajnert B, Janiszewska J, Urbanczyk-Lipkowska Z, Brys-
zewska M, Epand RM. 2006. DSC studies on interactions
between low molecular mass peptide dendrimers and model
lipid membranes. Int J Pharm 327:145–152.
27. Moore JS, Stupp SI. 1990. Room temperature polyesterifica-
tion. Macromolecules 23:65–70.
28. Nemoto H, Wilson JG, Nakamura H, Yamamoto Y. 1992.
Polyols of a cascade type as a water-solubilizing element of
carborane derivatives for boron neutron capture therapy. J Org
Chem 57:435–435.
29. Rao TS, Revankar GR. 1995. Synthesis of Certain Acyclic
Nucleoside Analogs of 1,2,4-Triazolo[3,4-f][1,2,4]triazine and
Pyrimido[5,4-d]pyrimidine. Nucleosides Nucleotides Nucleic
Acids 14:1601–1612.
´
10. Tomalia DA, Frechet JMJ. 2002. Discovery of dendrimers and
dendritic polymers: A brief historical perspective. J Polym Sci
[A] 40:2719–2728.
11. Esfand R, Tomalia DA. 2001. Poly(amidoamine) (PAMAM)
dendrimers: From biomimicry to drug delivery and biomedical
applications. Drug Discov Today 6:427–436.
12. Gillies ER, Frechet JMJ. 2002. Designing macromolecules for
therapeutic applications: Polyester dendrimer-poly(ethylene
oxide) ‘‘bow-tie’’ hybrids with tunable molecular weight and
architecture. J Am Chem Soc 124:14137–14146.
13. Gillies ER, Frechet JMJ. 2005. Dendrimers and dendritic
polymers in drug delivery. Drug Discov Today 10:35–43.
14. Svenson S, Tomalia DA. 2005. Dendrimers in biomedical appli-
cations–reflections on the field. Adv Drug Deliv Rev 57:2106–
2129.
30. Szoka F, Papahadjopoulos D. 1978. Procedure for preparation
of liposomes with large internal aqueous space and high cap-
ture by reverse-phase evaporation. Proc Natl Acad Sci USA
75:4194–4198.
31. Wang N, Wang T, Li TF, Deng YJ. 2009. Modulation of the
physicochemical state of interior agents to prepare controlled
release liposomes. Colloids Surf
238.
B Biointerfaces 69:232–
32. Hatziantoniou S, Demetzos C. 2006. Qualitative and quanti-
tative one-step analysis of lipids and encapsulated bioactive
molecules in liposome preparations by HPTLC/FID (IATROS-
CAN). J Liposome Res 16:321–330.
33. Barone G, Del Vecchio P, Fessas D, Giancola C, Graziano G,
1992. Theseus: A new software package for the handling and
analysis of thermal denaturation data of boilogical macromo-
lecules. Journal of Thermal Analysis and Calorimetry 38:2279–
2790.
15. Naylor AM, Goddard WA, Kiefer GE, Tomalia DA. 1989. Star-
burst dendrimers. 5. Molecular shape control. J Am Chem Soc
111:2339–2341.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010
DOI 10.1002/jps

NEW DRUG DELIVERY NANOSYSTEM
3571
34. Hupfeld S, Moen HH, Ausbacher D, Haas H, Brandl M. Lipo-
some fractionation and size analysis by asymmetrical flow field-
flow fractionation/multi-angle light scattering: Influence of
ionic strength and osmotic pressure of the carrier liquid. Chem
Phys Lipids 163:141–147.
35. Papagiannaros A, Hatziantoniou S, Dimas K, Papaioannou GT,
Demetzos C. 2006. A liposomal formulation of doxorubicin,
composed of hexadecylphosphocholine (HePC): Physicochem-
ical characterization and cytotoxic activity against human
cancer cell lines. Biomed Pharmacother 60:36–42.
36. Van Bommel EMG, Crommelin DJA. 1984. Stability of doxor-
ubicin-liposomes on storage: As an aqueous dispersion, frozen
or freeze-dried. Int J Pharm 22:299–310.
37. Fonseca M, vanWinden ECA, Crommelin DJA. 1997. Doxor-
ubicin induces aggregation of small negatively charged lipo-
somes. Eur J Pharm Biopharm 43:9–17.
38. Fritze A, Hens F, Kimpfler A, Schubert R, Peschka-Suss R.
2006. Remote loading of doxorubicin into liposomes driven by a
transmembrane phosphate gradient. Biochim Biophys Acta
Biomembr 1758:1633–1640.
39. vanWinden ECA, Crommelin DJA. 1997. Long term stability of
freeze-dried, lyoprotected doxorubicin liposomes. Eur J Pharm
Biopharm 43:295–307.
40. Tomalia DA. 2005. Birth of a new macromolecular architec-
ture: Dendrimers as quantized building blocks for nanoscale
synthetic polymer chemistry. Prog Polym Sci 30:294–
324.
41. Tomalia DA, Naylor AM, Goddard WA. 1990. Starburst
dendrimers: Molecular-level control of size, shape, surface
chemistry, topology, and flexibility from atoms to macro-
scopic matter. Angew Chem Int Ed Engl 29:138–
175.
42. Charrois GJR, Allen TM. 2004. Drug release rate influences
the pharmacokinetics, biodistribution, therapeutic activity,
and toxicity of pegylated liposomal doxorubicin formulations
in murine breast cancer. Biochim Biophys Acta Biomembr
1663:167–177.
43. Ivanova VP, Makarov IM, Schaffer TE, Heimburg T. 2003.
Analyzing heat capacity profiles of peptide-containing mem-
branes: Cluster formation of gramicidin A. Biophys J 84:2427–
2439.
44. Freire E, Markello T, Rigell C, Holloway PW. 1983. Calori-
metric and fluorescence characterization of interactions
between cytochrome b5 and phosphatidylcholine bilayers. Bio-
chemistry 22:1675–1680.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 8, AUGUST 2010