Journal of Controlled Release 110 (2005) 151 163

www.elsevier.com/locate/jconrel

Lipid nanoparticles for skin penetration enhancementcorrelation

to drug localization within the particle matrix as determined by
fluorescence and parelectric spectroscopy
S. Lombardi Borgia a, M. Regehly b, R. Sivaramakrishnan b, W. Mehnert a,
H.C. Korting c, K. Danker d, B. Roder b, K.D. Kramer e, M. Schafer-Korting a,*
a

Institut f ur Pharmazie, Freie Universitat Berlin, Berlin, Germany

b
Institut f ur Physik, Humboldt-Universitat zu Berlin, Germany
Dermatologische Klinik, Ludwig-Maximilians-Universitat, Munchen, Germany
d
Charite-Universitatsmedizin Berlin, Campus Benjamin Franklin, Germany
e
Institut f ur Physik, Freie Universitat Berlin, Berlin, Germany
Received 19 May 2005; accepted 27 September 2005

Abstract
With topical treatment of skin diseases, the requirement of a high and reproducible drug uptake often still is not
met. Moreover, drug targeting to specific skin strata may improve the use of agents which are prone to cause local
unwanted effects. Recent investigations have indicated that improved uptake and skin targeting may become feasible by
means of nanoparticular systems such as solid lipid nanoparticles (SLN), nanostructured lipid carriers (NLC) and
nanoemulsions (NE). Here we describe techniques to characterize drug loading to carrier systems and skin penetration
profiles by using the lipophilic dye nile red as a model agent. Since the mode of drug association with the particle
matrix may strongly influence the efficiency of skin targeting, parelectric spectroscopy (PS) was used to differentiate
between matrix incorporation and attachment to the particle surface and fluorescence spectroscopy (FS) to solve dye
distribution within NLC particles. Nile red was incorporated into the lipid matrix or the covering tensed shell,
respectively, of SLN and NLC with all the lipids studied (Compritol, Precirol, oleic acid, Miglyol). In NLC, the dye
was enriched in the liquid phase. Next, nile red concentrations were followed by image analysis of vertical sections of
pigskin treated with dye-loaded nanoparticular dispersions and an oil-in-water cream for 4 and 8 h in vitro. Following
the SLN dispersions, dye penetration increased about fourfold over the uptake obtained following the cream. NLC
turned out less potent (b threefold increase) and penetration appeared even reduced when applying a NE. In contrast to
previous studies with glucocorticoids attached to the surface of SLN, a targeting effect was not detected here. Therefore,

* Corresponding author. Institut f ur Pharmazie der Freien Universitat Berlin, Konigin-Luise-Strahe 2-4, D-14195 Berlin, Germany. Tel.: +49
30 838 53283; fax: +49 30 838 54399.
E-mail address: msk@zedat.fu-berlin.de (M. Schafer-Korting).
0168-3659/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconrel.2005.09.045

152

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

drug targeting appears to be more strictly related to the mode of interaction of drug and particle than penetration
enhancement.
D 2005 Elsevier B.V. All rights reserved.
Keywords: Skin absorption; Dye loading; Solid lipid nanoparticles; Nanostructured lipid carriers; Nanoemulsion; Fluorescence spectroscopy;
Parelectric spectroscopy

1. Introduction
Topical therapy of skin disease allows high drug
levels at the site of disease and reduces systemic side
effects. Applied topically, glucocorticoids are first-line
drugs for acute exacerbations and calcineurin inhibitors for long-term suppression of eczema [1].
Retinoids are used for mild to moderate forms of
acne and psoriasis [2,3]. Yet, local side effects such as
skin atrophy following glucocorticoids, lymphomas
feared with immunosuppressants and local irritation
following retinoids are of major concern. Recent
progress in nanotechnology may allow to develop
drug carriers improving the generally low skin
penetration and may also induce targeting to specific
skin strata. This appears to hold true with liposomes
[4] and nanoparticles which can enhance drug uptake
to the epidermis [5,6] or skin appendages [7].
In order to design dermal delivery systems, it is
essential to understand the principles of substance
incorporation into the carrier and the consequences for
penetration into and permeation through the skin.
Therefore, not only specific methods are looked for
which discriminate drug levels within the particle
matrix and bound to its surface but also methods to
quantify the drug at the target site. For a first
approach, dyes as model agents can be used.
Fluorescence microscopy plus image analysis allow
to relate the agent of interest to defined skin strata as
demonstrated by the accumulation of liposomal
carboxyfluorescein in sebaceous glands of Syrian
hamster ear skin [8]. Moreover, confocal laser
scanning microscopy plus integration of pixel brightness values served to investigate the fluorescence
distribution in skin layers following the application of
rhodamine-loaded TransfersomesR to murine skin.
Fluorescence detection was limited to a skin depth of
16 Am [9], while skin penetration of rhodamine red
and calcein contained in ethosomes, liposomes and a
hydroethanolic solution were compared in nude mice

skin until 260 Am depth [10]. Fluorescence measurement allowed to follow the distribution of dyes into
hair follicles of human scalp skin, too [11]. Therefore,
fluorescent dyes of various physico-chemical properties may allow the study of the principles of carrier
effects as well. Pertinent data will allow scientists in
the field of pharmaceutical technology to select the
most suitable system from the start on.
Since drug association with the carrier has been
found to be of great importance in carrier efficiency to
skin targeting, we also aimed for a confirmation of the
validity of results obtained by parelectric spectroscopy
(PS; [6]). Therefore, we used steady state fluorescence
spectroscopy [1214] in parallel, which may be also
useful to characterize the dye distribution within
particles. The application of these two independent
techniques offers the chance of mutual control.
Besides this, the present work aims to establish a
fluorescence method to study the distribution of a
model dye within the skin strata and appendages
following the application of lipid-based nanoparticular
systems of different structure. Nanoparticulate systems investigated are solid lipid nanoparticles (SLN),
nanoemulsions (NE) and nanostructured lipid carriers
(NLC). NLC are mixtures of solid (long chain) and
fluid (short chain) lipids.

2. Materials and methods

2.1. Materials
CompritolR 888 ATO (glyceryl behenate) and
PrecirolR ATO 5 (glyceryl palmitostearate) were gifts
from Gattefosse (Weil a. Rh., Germany), MiglyolR
812 (medium chain triglycerides) and base cream (oilin-water cream prepared as described in Deutscher
Arzneimittel Codex 2004 containing: glycerol monostearat 60 4.0 g, cetylalcohol 6.0 g, medium chain
triglycerids 7.5 g, white vaselin 25.5 g, macrogol 20-

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

glycerolmonostearate 7.0 g, propylenglycol 10.0 g,

water 40.0 g) were obtained from Caelo (Hilden,
Germany), the emulsifier PluronicR F68 was supplied
by BASF (Ludwigshafen, Germany). Nile red (log
P = 3.10; octanol/water; [15]) and oleic acid were
purchased from Sigma-Aldrich (Munchen, Germany).
All other reagents were obtained from Merck (Darmstadt, Germany) and were of the highest quality
available.
2.2. Methods
2.2.1. Preparation of the test formulations
SLN, NLC and NE were prepared as described
elsewhere [5,1618]. Briefly, solid lipids were melted
at 95 8C. Following the addition of nile red solution,
the fluid lipid phase was dispersed in a PluronicR F68
2.5% solution in water using an ultra turrax to form a
premix which was then passed through a Lab 60 high
pressure homogenizer (APV Gaulin, Lubeck, Germany). Three cycles at 500 bar and 85 8C were
performed. As lipid phase, SLN dispersions contained
a pure solid lipid (Compritol, C; Precirol, P) and
nanoemulsions a pure fluid lipid (oleic acid, O;
Miglyol, M). NLC lipid phase was a mixture of a
solid and a fluid lipid. The composition of lipid
nanoparticle dispersions is given in Table 1. For
physical characterization and penetration studies, dye
free as well as nile red loaded dispersions (dye content
from 0.001% to 0.1%) were prepared. Nile red 0.04 %
cream served as reference in skin penetration studies.
Nile red dispersed in some droplets of paraffin oil was
incorporated into the base cream. All preparations
were stored at 8 8C until used.
2.2.2. Characterization of particles: crystallinity, size
and distribution
The degree of crystallinity was determined by
differential scanning calorimetry (DSC) measurements [19]. Particle size analysis was performed by
photon correlation spectroscopy (PCS, Malvern Zetasizer IV, Malvern Instruments, Malvern, UK) which
yields the mean particle size (z-average) and the
polydispersity index (PI) as a measure of the width of
the distribution. In addition, laser scattering (LD,
Coulter LS 230, Miami, FL) served to detect larger
particles. pH values of the dispersions were determined and dispersions repeatedly inspected by light

153

microscopy to detect nile red crystallisation during

storage [20].
2.2.3. Characterization of nile red entrapment by
parelectric spectroscopy
It has been demonstrated that the drug association
with the particles is of highest relevance for epidermal
targeting [5]. Parelectric spectroscopy, successfully
applied previously, allows to investigate the interaction between drug and SLN [6]. This technique
applied for glucocorticoids, now served to study nile
red labelling of SLN and NLC. The respective dyefree dispersions served for control.
The basic ideas underlying this relatively novel
method have been expressed as early as 1930 [21]
describing the frequency dependence of the complex
dielectric constant e = eV( f )
i d eW( f ) and giving the
phenomenologically introduced relaxation time s a
microscopic meaning. This is now known as the
EinsteinDebye relation stating that sm with m as
mass of the dipole-carrying molecule. The essential
parameters in Debyes description are the density De
and the mobility f 0 = 1/2Po of the permanent electric
dipole moments. To measure the complex dielectric
constant, a transmitter feeds an electromagnetic wave
into an open-ended coaxial probe, the cut inner and
outer diameters forming a condenser in contact with
the samples under test [22]. The reflected wave is
analysed by comparison with the wave sent into the
probe, thus yielding the desired parameters De(c) and
f 0(c) in dependence on the concentration c of the dye
molecules incorporated into or attached to the carrier
particles.
Within the narrow region of mass changes as
caused by the added dye molecules, the abovementioned curves have to be described by a linear
dependence on the mass if there is pure dye
incorporation. We can state the validity of the EinsteinDebye relation to explain the relatively small
changes in De(c) and f 0(c). As soon as we observe
deviations from linearity, we have to assume a
changing surface of the carrierdye system. These
effects can be explained on the basis of theory of rate
processes [23]. As given in a previous paper [6], the
model describes the thermal fluctuations of adjacent
particles by jump heights h 0 to be overcome by an
activation energy DE 0, both quantities being large in
zero attachment or fully clad carriers. In between these

154

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

limiting cases, neighbouring particles find empty sites

on the dyecarrier system connected to jump heights
h b h 0 and activation energies DE b DE 0. Following
from Arrhenius law, the mobility is augmented in the
latter situation. Thus, the curves f 0(c) pass a maximum
and De(c) pass a minimum, both deviations from linear
fits by far exceed error bars.
Measurements were performed as described [6]
using a commercial frequency analyser (type ZVRE;
Rohde and Schwarz, Mu nchen, Germany). The
dependence of the parameters mobility f 0 and density
De of the dipole-carrying lipids on the dye concentration gives insight into the type of dye loading, in
our findings an incorporation of the dye molecules
into carrier particles. The PS method cannot give
insight into specific location inside the carrier
particles. It is, however, a quick method to measure
large numbers of samples. Moreover, any deviation of
the answers of the dispersion eV( f) and absorption
eW( f) from the pure Debye behaviour allows to
calculate the polydispersity via a broadening of the
absorption curve.
2.2.4. Steady state fluorescence measurements
To study the dye localization within the particles,
SLN and NLC dispersions containing nile red
(4 F 0.1) d 10
2% were subjected to steady state
fluorescence measurements. The dye dissolved in
pure lipids of identical concentrations served as
source of reference spectra. For measurement of the
fluorescence spectra, the SLN and NLC dispersions
were diluted with water in order to minimize
scattering of the fluorescence by nanoparticles and
to avoid re-absorption effects. Measurements were
carried out using a diode-pumped Yt:YAG cw-laser
(VersaDisk, Darmstadt, Germany). Dispersions stirred
by a magnetic bar were excited by a second harmonic
generation laser (SHG, Princeton, New Jersey) at 548
nm with an output power of 100 mW. The fluorescence was measured using a polychromator (ORIEL,
Darmstadt, Germany) equipped with a CCD detector
system (Instaspec, Darmstadt, Germany).
2.2.5. Penetration experiments
2.2.5.1. Method establishment and validation
Fluorescence pictures were taken (20 magnification)
using an inverted fluorescence microscope (Axiovert

200, ZEISS, Jena, Germany) equipped with a monochrome camera (AxioCam HR, version 5.05.10,
ZEISS). Nile red fluorescence was recovered in the
red band (beam splitter wavelength 585 nm) exciting
the samples at 546 nm. Pictures were recorded setting
the camera integration time to 10 ms. Fluorescence
yield was quantified by an image treatment software
(Axiovision 3.1.3.1, ZEISS); the integration of pixel
brightness values (arbitrary brightness values, ABU)
gives the relative dye contents.
First, the fluorescence intensity had to be studied
for linearity in relation to the dye concentration in
solution. 500 ll of nile red solutions (0.0020 to 8.0
lg/ml in methanol) contained in polystyrene holders
was subjected to fluorescence microscopy. Four
samples for each dye concentration were tested and
measurements were repeated twice on different days.
The entire area of the recorded picture was analysed
for pixel brightness. Camera light intensity scale
ranged from 0 (black) to 16383 (white) ABU. A
linear relation between ABU and dye concentration
exists for ABU values ranging from 130 F 25%
(0.015 lg/ml) to 14 000 F 3% (1.3 lg/ml;
R 2 = 0.9998). The coefficients of variation for intraassay precision were less than 9.2% and inter-assay
precision was 2.4% to 7.1%.
Next, we investigated the relation of nile red
fluorescence in vertical skin slices treated with
increasing dye concentrations. These slices (20 Am
thickness) obtained by a freeze microtome (Frigocut
2800 N, Leica, Bensheim, Germany) from untreated
skin were kept for 5 min in dye solutions (water/
ethanol 70/30 v/v) with increasing nile red concentration (three independent experiments each with 4
slices per concentration). Slices were then withdrawn,
dried for 1 h at 25 8C, and then kept at 4 8C for 24
h until analysed for fluorescence. Nile red bleaching
was taken into consideration: After 1 s of exposition
to the beam, the dye showed a bleaching of 416%,
which is related to the initial fluorescence. Therefore,
fluorescence pictures had to be recorded in the first 10
ms of exposition. Fluorescence ABU reading was
performed within user-defined areas as shown in Fig.
1A. Pixel intensities (I) were integrated in 5 rectangular shaped areas located within the stratum corneum
(SC, I sc), the viable epidermis (VE, I ve), the superficial dermis (SD, I sd 100 Am from the basal
membrane) and in the deep dermis (DD, I dd400

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

155

Fig. 1. (A) Scheme of the fluorescence intensity data taken from skin pictures. (B) Fluorescence ABU values measured in skin slices soaked into
nile red solutions at different concentrations in water/ethanol 70/30. The relation existing between the fluorescence emission and the dye
concentration is described by a linear regression fit, the R 2 values ranging between 0.9562 and 0.9947.

Am from the basal membrane). A further reading was

performed outside the skin surface.
Fluorescence intensity units (ABU) obtained from
untreated skin were defined as background level (B).
These are: horny layer 105 F 16%, viable epidermis
52 F 12%, superficial dermis 51 F 8%, and deep
dermis 49 F 8% as mean values of 10 skin samples.
Corrected ABU values were calculated by subtracting
the above given B values from the corresponding pixel
intensities of dye-treated skin. Linear relations existed
between the dye concentration and corrected ABU
values measured in the different skin strata (Fig. 1B)
up to about 12 000 ABU as was obtained from stratum
corneum soaked in 0.31 Ag/ml nile red solution and
viable epidermis treated with 0.50 Ag/ml dye solution.
With both superficial and deep dermis, however,
response was always clearly below the camera plateau.
Skin strata response to identical nile red concentrations
varied. As can be seen from Fig. 1B, the horny layer
showed the highest fluorescence intensity, followed by
viable epidermis, superficial dermis, and deep dermis
which reflects differences in the binding capability of
this lipophilic dye.
2.2.6. Skin penetration tests
Cutaneous uptake was studied in vitro using pigskin
[2426]. Fresh ears of domestic pigs were obtained

from a local abattoir on the day of the experiment and

kept on ice for the transport to the laboratory. Within 2
h, post-mortem skin was removed from adhering
subcutis avoiding contamination of the skin surface
by subcutaneous lipids. Full thickness skin was then
punched to obtain specimens of 15 mm diameter and
mounted to Franz flow-through diffusion cells 9 mm in
diameter (Crown Scientific, Somerville, NJ) the
stratum corneum facing the air and the dermis being
in contact with the acceptor medium (phosphatebuffered saline, PBS; skin surface temperature 32
8C). 300 Al of SLN, NLC, NE and a cream containing
(4 F 0.1) d 10
2% nile red were applied non-occlusively to the skin surface for 4 and for 8 h. Then surplus
material was carefully removed by a cotton swab and
the skin was removed from the Franz cell and cut into
vertical slices (bottom to surface) of 20 Am thickness
each; three skin slices were taken per treated area.
Slices were stored on slides at + 4 8C and analysed for
fluorescence within 24 h. All preparations were tested
in parallel using the skin of four pigs.
Skin slices were subjected to both normal light- and
fluorescence-microscopy, the integration of pixel
intensity of the recording areas (Fig. 1A) was
performed as described above. Since fluorescence
intensity of the system slide and freezing medium did
not exceed 4% of the horny layer response, only minor

156

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

3. Results

sizes were followed for up to 82 days of storage at

8 8C. PCS results are shown in Table 1: Compritoland Precirol-based SLN were of similar size (mean
diameter 184 and 211 nm), while Precirol-based
NLC (NLC-P-O, 153 nm; NLC-P-M, 162 nm) were
significantly smaller than Compritol-based NLC
(NLC-C-O, 236 nm; NLC-C-M, 199 nm). The
nanoemulsions contained droplets of a mean size of
227 (NE-O) and 179 nm (NE-M). Polydispersity
index was always less than 0.25 except for NLC-CM (0.30). LD results confirmed that particles and
nanoemulsions are constituted as a homogeneous
population except for NLC-C-O where a micrometric
sub-population is present (LD95%: 8 Am). Even
storage for almost 3 months did not change particle
sizes and their distribution. This does not hold true
for the NLC-C-O dispersion which after an increase
of size and size distribution gelled. Therefore,
parelectrical and fluorescence spectroscopies as well
as skin penetration were studied within the first
month after preparation.
pH values of oleic acid-free dispersions ranged
from 5.0 to 5.9. Oleic acid addition resulted in a
decline to 4.4 to 5.0.

3.1. Particle characterization

3.2. Parelectric spectroscopy results

At day 3 after preparation, particle crystallinity,

size and size distribution were characterized by
conventional techniques [19,20,27]. Crystallinity
(8997%) and size of SLN were well in accordance
with previous results [6,19] while the addition of
oleic acid and Miglyol reduced crystallinity of the
NLCs to 8084% and 6386%, respectively. Particle

Six different types of nile red dispersions have

been studied, the single step of the dispersion recordings to the level of water excludes the presence of
particular systems of different types such as micelles
which form from free Miglyol [28] or aggregates from
oleic acid at the pH values bpK a (9.85; [29].
Depending on the dye concentration c, the results of

amounts of fluorescence label may have diffused from

the skin into the surrounding medium.
2.3. Statistics
To compare the influence of nanoparticulate
systems upon dye uptake, four independent experiments were performed using the skin of four donator
pigs, each treated separately with test preparations and
cut into four slices each. For all experiments, the
corrected ABU values obtained from dye uptake
following the nanoparticular dispersions were then
related to ABUs obtained from the skin treated with
the nile red loaded cream. The resulting parameter
called penetration enhancing effect (PEE) was calculated for all skin layers of interest. Obviously, the PEE
value for the cream is 1 by definition.
All data are presented as arithmetic mean values
including the standard deviations. Because of the
small sample size, differences were analysed using Utest where p V 0.05 is considered as significant.

Table 1
Composition, size (diameter, /nm) and size distribution (polydispersity index, PI) of nanoparticle dispersions at days 3 and 82 of storage at 8 8C
Components

SLN-C
SLN-P
NLC-C-O
NLC-C-M
NLC-P-O
NLC-P-O
NE-O
NE-M

Day 3

Day 82

Solid lipid

Fluid lipid

/nm

PI

/nm

PI

Compritol 10%
Precirol 10%
Compritol 8%
Compritol 8%
Precirol 8%
Precirol 8%

Oleic acid 2%
Miglyol 2%
Oleic acid 2%
Miglyol 2%
Oleic acid 10%
Miglyol 10%

184
211
236
199
153
162
227
179

0.246
0.151
0.303
0.215
0.157
0.127
0.207
0.150

186
195
Gelled
201
154
199
232
191

0.210
0.165
0.185
0.118
0.095
0.169
0.121

Composition is reported as % lipid in the water phase, the tensid is Pluronic F68, 2.5%. Average values of 10 measurements; coefficients of
variation range between 0.9% and 2.6% for diameters and between 17% and 50% for PI.

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

PS measurements are presented in Fig. 2. The

parameters of interest, i.e. dipole density De(c) and
dipole mobility f 0(c), have been extracted from
experimental curves with a signal/noise ratio of 10,
approximately. As each of the four PS parameters [6]
is derived from a fit of 200 measured frequencies, the
error bars will be F 3% as an average of all resulting
points. Compared to this, errors in drug concentration
and temperature can be neglected. As these curves
obviously follow a linear dependence on c, we have to
assume an almost complete incorporation of the agent
into SLN and NLC following the model of Einstein
and Debye.
3.3. Steady state fluorescence results

their accessibility to water molecules results in a

lower hydrophobicity. The increased re-orientation
probability of surrounding water molecules leads to
the pronounced red shift of the corresponding
emission maximum.
First, SLN-C dispersions containing different
amounts of nile red were measured in order to verify
a linear relationship of fluorescence intensity and dye
concentration. The verification of this linearity
excludes a remarkable self-quenching by aggregation.
Next, reference spectra of nile red in Miglyol and
oleic acid were recorded while the lack of light
transparency in the visible range impeded to obtain
calibration spectra in solid lipids. The use of solid
lipid may change the dye fluorescence emission.
Therefore we dispersed SLN in aqueous nile redloaded SLN in water, assuming that the dye amount is
preserved in the lipid. This assumption is based on the
much higher solubility of nile red in lipid as compared
to an aqueous surrounding [12]. Dye spectra were
recorded from the four NLC preparations, too,
containing combinations of the solid and liquid lipids
of interest. Fig. 3A depicts the nile red fluorescence in
NLC-C-O particles and the dye spectra as measured in
a liquid and a solid lipid surrounding. The fluorescence from NLC particles can be modelled as the sum
of the emission spectra from nile red in the subcompartments. Dye spectra in pure solid lipid F(k) at
the known concentration c N and nile red spectra in
pure fluid lipid O(k) at c N were measured separately.
The verified linear relationship between concentration

0.5

500

0.4

400

Dipole density (
)

Dipole mobility (f0 /MHz)

Because of a solvent-dependent fluorescence

emission, nile red is the first choice probe for
steady state fluorescence investigation. In fact, nile
red is very soluble in organic solvents and strongly
fluorescent in a hydrophobic lipid environment. The
emission maximum is shifted towards the blue
region with increasing hydrophobicity of the
solvent [1214]. In the case of nile red molecules
embedded between the relatively long hydrocarbon
chain as to be found in the solid lipids [13], this
location is likely to prevent any contact with
adjacent water molecules. Thus, corresponding to
a high hydrophobicity, the emission maximum is
found near 600 nm. Due to the shorter chain length
of an oily environment of the nile red molecules,

0.3
0.2
0.1

157

300
200
100
0

0.0
0.00

0.02

0.04

0.06

0.08

Nile red concentration (%)

0.10

0.00

0.02

0.04

0.06

0.08

0.10

Nile red concentration (%)

Fig. 2. Influence of nile red incorporation on f 0 and De. Particles were loaded with increasing dye concentrations as described in Methods and
then subjected to parelectric spectroscopy varying the frequency from 0.01 to 100 MHz at 305 K. Results are presented for f 0(c) and De(c) for
two SLN and four NLC dispersions. SLN-C (5), SLN-P (+), NLC-C-O (*), NLC-C-M (w), NLC-P-O (D), NLC-P-M (o).

158

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

Fig. 3. (A) Fluorescence spectra of NLC-C-O loaded with nile red and reference spectra of the dye in a Compritol-based SLN dispersion and
dissolved in oleic acid. (B) Comparison of the measured NLC-C-O spectrum with the fitted model spectra constructed by a superposition of the
nile red fluorescence in the pure lipids.

and fluorescence intensity allows to express the

theoretical spectrum N theor.(k) as
Ntheor: k cF d F k=cN cO d Ok=cN :
The quantities c F and c O refer to the unknown
concentrations of nile red in the solid and liquid lipid
sub-compartments of the NLC particles. The different
fluorescence quantum yield of nile red in the lipids
does not need to be considered due to the previously
measured function F(k) and O(k). Starting from a
measured spectrum N meas.(k) of nile red, the dye
concentrations c F and c O in the sub-compartments can
be extracted by minimization of the integrated least
square error function MIN(c F d c O).
MINcF ; cO
Z

cF d F k=cN cO d Ok=cN
Nmeas: k2 dk:

The integration has to cover the whole fluorescence emission band of nile red. This least square
minimization leads to the dye concentrations c F and
c O in the two sub-compartments (Fig. 3B). The
adjusted reference spectra, namely the products
c F d F(k)/c N and c O d O(k)/c N are shown in Fig. 3B,
too. In the NLC dispersions, solid lipid exceeded the
fluid one by a factor of 4.
The amount of nile red in the solid and liquid
lipid compartments of the NLC particles (Table 2)
were determined with an error of + 6%. The
consistency between the measured NLC spectrum

and the fitted theoretical spectrum provides evidence

for the reliability of the proposed model. Besides the
quality of the fit in the wings of the fluorescence
band, there is still a discrepancy left in the peak
region. This misfit can be expressed by the area
between the two curves. Near this peak, the model
function gives somewhat smaller values than the
measured spectrum, indicating that dye molecules are
located in interfacial regions between the different
lipid droplets. The above-defined misfit area amounts
to 5% of the total area for all NLC formulations.
Dye distribution within the nanoparticles follows the
solubility of the lipophilic agent in the sub-compartments. The good solubility of nile red in Miglyol
explains the strong dye enrichment in the oily
component of these NLC dispersions. Since Compritol and Precirol differ only slightly in their
hydrocarbon chain length, nile red distribution is
almost equal for NLC made up from both solid
lipids. Varying the types of lipid for NLC formulation, one can adjust the distribution of lipophilic
agents between the different sub-compartments.

Table 2
Nile red content (%) in both phases of NLC dispersions
Formulations

Solid lipid phase (%)

Fluid lipid phase (%)

NLC-C-O
NLC-C-M
NLC-P-O
NLC-P-M

54 F 6
35 F 4
47 F 6
27 F 3

46 F 6
65 F 8
53 F 6
73 F 9

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

159

Fig. 4. Penetration of nile red from cream and Precirol-based nano-particles into pig skin after 8 h. (A) Cream. (B) SLN. (C) NLC, oleic acid.
(D) NLC, Miglyol. The vertical section depicts the stratum corneum (a), the epidermis (b), and the dermis (c). Pictures are obtained by
superposing normal light and fluorescence images of the same area; scale (black bar) is 100 Am.

Fig. 5. Penetration enhancing effect (PEE) of nano-particular formulations with respect to a oil in water cream after application for 4 and 8 h
(x F S.D., n = 4, *p b 0.05). Cream PEE value (=1) is indicated by dashed lines. SC: stratum corneum; VE: viable epidermis; SD: superficial
dermis.

160

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

3.4. Cutaneous uptake

Fig. 4 depicts representative examples of fluorescence microscopy images of vertical pigskin sections
following the topical application of the nile red
formulations for 8 h. With all the preparations, dye
staining is fading from the horny layer towards deeper
skin. The visualisation qualitatively indicates the most
pronounced trends of fluorescent staining in the various
skin strata: SLN clearly promoted the dye penetration
as compared to the cream. Detailed information is
obtained from pixel intensities derived from fluorescence measurements of the skin sections. Image
analysis was possible with stratum corneum, viable
epidermis and superficial dermis (Fig. 5). Penetration
of deep dermis was not calculated as fluorescence did
not significantly differ from the background level.
Since PEE values are given as relative to the cream and
do not give information on the absolute fluorescence
measured, the corrected ABU values obtained with
cream are presented as reference in Table 3.
Dye distribution was greatly influenced by the
particulate carriers: When applied using SLN, nile red
concentrations in the horny layer, viable epidermis
and superficial dermis exceeded those found with the
cream about fourfold. PEE values ranged from 1.51 to
5.91 ( p V 0.05) with a clear increase from 4 h to 8 h. A
minor, yet still often significantly improved penetration was seen with NLC dispersions, too, while there
was no improvement over cream with the nanoemulsion. In fact, at 8 h nile red levels in the horny
layer following the NE were even lower than with the
cream ( p V 0.05). With the NLC dispersions, PEE
values were 1.05 to 2.18 in SC, NLC-P-M induced a
significant increase of 2.18 F 0.47 at 8 h and NLC-PO resulted in an increase, significant for viable
epidermis and superficial dermis (p V 0.05). Varying
the solid lipids in the SLN and NLC dispersions as
well as varying the liquid component of NLC does not
Table 3
Nile red fluorescence (corrected ABU values) in pig skin 4 h and 8 h
after the application of an oil in water cream (xF CV, n = 12)
Time (h)

Stratum
corneum
(103 ABU)

Viable
epidermis
(103 ABU)

Superficial
dermis
(103 ABU)

Deep dermis
(103 ABU)

4h
8h

2.2 F 16%
2.5 F 23%

0.98 F 24%
0.15 F 25%

0.16 F 30%
0.23 F 29%

0.05 F 13%
0.06 F 40%

strikingly change the results. Therefore, the different

chain lengths of the solid lipids Compritol and
Precirol do not influence dye uptake to a relevant
extent and a penetration enhancement by oleic acid
[30] was not detectable either. A new aspect is
obtained from the comparison of the time dependence
of dye penetration following SLN and NLC application (Fig. 5): With the former, an increase of PEE over
the cream is seen as soon as 4 h whereas with NLC it
takes 8 h to reach some increase in PEE.
Since nile red fluorescence strongly depends on the
solvent we had to exclude that the enhanced fluorescence response is an artifact resulting from a modified
lipid environment in the horny layer following SLN
application. Although intact lipid particles do not
penetrate into stratum corneum [27,31], uptake of
their components (lipids and tensids) is to be
expected. Therefore, staining was also performed by
soaking skin previously treated for 8 h with the
respective dye-free vehicles into nile red solution. A
significant change of the fluorescence intensity was
not detected (data not shown). Therefore, the increased fluorescence intensities (ABU) and PEE
values exceeding 1 reflect an improved capacity of
carriers to deliver nile red to the skin.

4. Discussion
Topical drug therapy of skin diseases is confounded by poor and erratic drug absorption. High interand intra-individual absorption also limits transdermal
therapy to rather small and highly lipophilic agents.
To improve uptake, well-tolerated penetration enhancers and carrier systems are still looked for.
Therefore, using a lipophilic model dye and applying
image analysis, we investigated the potential of lipid
particles to enhance skin uptake and to possibly
induce targeting to specific skin strata. As compared
to conventional oil-in-water cream, nanoemulsions
appear to reduce dye penetration into pigskin while
SLN strongly improve nile red uptake which results in
increased concentrations from the horny layer at least
down to superficial dermis at 4 h and 8 h (Figs. 4 and
5). A minor increase of uptake was also found
following the application of NLC dispersions at 8 h.
Although dye penetration into the skin is favoured by
the lower lipid content of the particulate systems

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

(10%) as compared to the cream (25% vaseline, 7.5%

Miglyol), the striking lack of an improvement by the
nanoemulsion excludes the lipid content to be of
preponderate importance. In fact, the cream contains
the penetration enhancer propylene glycol (10%).
All formulations were applied without occlusion
and water evaporation was allowed. Thus, with SLN a
semisolid film covering the skin surface was formed,
while a more creamy texture was obtained with NLC.
With both dispersions, the loss of water increased the
superficial dye concentration in contact with skin
surface as compared to that observed with the cream
and nanoemulsion. This may explain the increased
skin penetration. Moreover, also the loss of water
inducing a change in crystal modification of the SLN
matrix may induce drug expulsion and penetration
[17]. This can also explain the more rapid dye
penetration following SLN as compared to NLC
(Figs. 4 and 5). The more rapid penetration from
SLN is in accordance with 100% of the dye being in
the solid phase capable for expulsion while b54% are
to be found in the solid phase of NLC (Table 2).
Finally, an occlusive effect [17,32] cannot be excluded to be of relevance for this improved uptake.
Independent of the mechanism, the results obtained
here are well in accordance with previous findings of
an about fourfold increased penetration of two
glucocorticoids, prednicarbate and betamethasone
17-valerate, not only when applied loaded to SLN
but also when drug-free SLN were present [5,6].
Therefore, SLN use should be a reliable basis for
more constant drug levels in the skin due to their
enhancement capability for skin penetration.
Yet SLN also have the potential to induce
epidermal targeting which is also derived from the
previous studies with topical glucocorticoids: Applying PS we already showed that glucocorticoid
attachment to the particle surface is associated with
an epidermal targeting. This method is generally
applicable for dipole-carrying lipids. Drug adsorption
to the particle surface was complete with prednicarbate but clearly far from complete with betamethasone
17-valerate [5,6]. As demonstrated in this study, also
nile red penetration into the skin was raised about
fourfold by SLN, yet there was no targeting effect
(Fig. 5). Applying parelectric spectroscopy, we found
that the dye is neither attached to the particle/tenside
surface nor located within micelles nor present in the

161

free form yet it is incorporated into the particles or the

covering tenside shell (Fig. 2). This was verified using
a second and independent physical technique, namely
fluorescence spectroscopy, which in principle gives a
better insight into type of interaction of the incorporated agent with its carrier yet limited to a fluorescence of the agent of interest. Thus, the latter method
allowed to differentiate between nile red incorporation
into the liquid and into the solid lipid phase of NLC
(Fig. 3) which are described to consist of oil droplets
dispersed in a solid lipid matrix [17]. Alternatively, an
attachment of the liquid lipid to the particle surface
has been suggested [33]. As given in Table 2, the dye
accumulated in the fluid lipid can explain the
penetration behaviour described above.

5. Conclusion
Combining PS and FS, we obtained a clearer
picture of dye particle interactions as it is possible
with one single method. The discriminatory potential
of PS does not depend of the specific physicochemical
behaviour of the loaded agent and relies only on
changes in dipole mobility of lipid particles if drugs
are loaded to lipid phase. Therefore, if future studies
will verify the relevance of the type of drug
interaction with lipid matrices for skin uptake and
targeting, the tool of parelectric spectroscopy may
become a standard method in pharmaceutical technology comparable to the well-established methods
such as photon correlation spectroscopy and laser
diffractometry.

Acknowledgement
Financial support of the Deutsche Forschungsgemeinschaft, FG 463 (SCHA 382, KR 1696), is
gratefully acknowledged.

References
[1] T. Ruzicka, T. Assmann, M. Lebwohl, Potential future
dermatological indications for tacrolimus ointment, Eur. J.
Dermatol. 13 (2003) 331 342.
[2] A. Haider, J.C. Shaw, Treatment of acne vulgaris, Jama 292
(2004) 726 735.

162

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

[3] A. Roeder, M. Schaller, M. Schafer-Korting, H.C. Korting,

Tazarotene: therapeutic strategies in the treatment of psoriasis,
acne and photoaging, Skin Pharmacol. Physiol. 17 (2004)
111 118.
[4] P.L. Honeywell-Nguyen, A.M. de Graaff, H.W. Groenink, J.A.
Bouwstra, The in vivo and in vitro interactions of elastic and
rigid vesicles with human skin, Biochim. Biophys. Acta 1573
(2002) 130 140.
[5] C. Santos Maia, W. Mehnert, M. Schaller, H.C. Korting, A.
Gysler, A. Haberland, M. Schafer-Korting, Drug targeting by
solid lipid nanoparticles for dermal use, J. Drug Target. 10
(2002) 489 495.
[6] R. Sivaramakrishnan, C. Nakamura, W. Mehnert, H.C.
Korting, K.D. Kramer, M. Schafer-Korting, Glucocorticoid
entrapment into lipid carrierscharacterisation by parelectric
spectroscopy and influence on dermal uptake, J. Control.
Release 97 (2004) 493 502.
[7] N. Otberg, H. Richter, H. Schaefer, U. Blume-Peytavi, W.
Sterry, J. Lademann, Variations of hair follicle size and
distribution in different body sites, J. Invest. Dermatol. 122
(2004) 14 19.
[8] L.M. Lieb, C. Ramachandran, K. Egbaria, N. Weiner, Topical
delivery enhancement with multilamellar liposomes into
pilosebaceous units: I. In vitro evaluation using fluorescent
techniques with the hamster ear model, J. Invest. Dermatol. 99
(1992) 108 113.
[9] A. Schatzlein, G. Cevc, Non-uniform cellular packing of the
stratum corneum and permeability barrier function of intact
skin: a high-resolution confocal laser scanning microscopy
study using highly deformable vesicles (Transfersomes), Br. J.
Dermatol. 138 (1998) 583 592.
[10] E. Touitou, B. Godin, N. Dayan, C. Weiss, A. Piliponsky, F.
Levi-Schaffer, Intracellular delivery mediated by an ethosomal
carrier, Biomaterials 22 (2001) 3053 3059.
[11] Y.Y. Grams, J.A. Bouwstra, Penetration and distribution of
three lipophilic probes in vitro in human skin focusing on the
hair follicle, J. Control. Release 83 (2002) 253 262.
[12] P. Greenspan, S.D. Fowler, Spectrofluorometric studies of the
lipid probe, nile red, J. Lipid Res. 26 (1985) 781 789.
[13] P. Greenspan, E.P. Mayer, S.D. Fowler, Nile red: a selective
fluorescent stain for intracellular lipid droplets, J. Cell Biol.
100 (1985) 965 973.
[14] W. Hen, T. Liu, F. Himo, A. Toutchkine, D. Bashford, K.
Hahn, L.A. Noodleman, Theoretical study of the UV/VIS
absorption and emission solvatochromic properties of solventsensitive dyes, ChemPhysChem 4 (2003) 1084 1094.
[15] CAS; http://www.cas.org/SCIFINDER/scicover2.html.
[16] V. Jenning, M. Schafer-Korting, S. Gohla, Vitamin A-loaded
solid lipid nanoparticles for topical use: drug release properties, J. Control. Release 66 (2000) 115 126.
[17] R.H. Muller, M. Radtke, S.A. Wissing, Solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC) in
cosmetic and dermatological preparations, Adv. Drug Deliv.
Rev. 54 (Suppl 1) (2002) S131 S155.
[18] A. zur Muhlen, C. Schwarz, W. Mehnert, Solid lipid nanoparticles
(SLN) for controlled drug delivery-drug release and release
mechanism, Eur. J. Pharm. Biopharm. 45 (1998) 149 155.

[19] C. Schwarz, W. Mehnert, Solid lipid nanoparticles (SLN) for

controlled drug delivery: II. Drug incorporation and physicochemical characterization, J. Microencapsul. 16 (1999)
205 213.
[20] W. Mehnert, K. Mader, Solid lipid nanoparticles. Production,
characterization and applications, Adv. Drug. Deliv. ReV. 47
(2001) 165 196.
[21] P. Debye, Polare Molekeln, Hirzel, S., Leipzig, 1930.
[22] F. Kremer, A. Schonhals, Broadband Dielectric Spectroscopy,
Springer, Berlin, 2003.
[23] S. Glasstone, K.J. Laidler, H. Eyring, The Theory of Rate
Processes, 1941, New York.
[24] OECD, Guidance document for the conduct of skin absorption
studies, Series on Testing Assessment, vol. 28, 2003.
[25] OECD, Skin absorption: in vitro method, Guideline 428, 2004.
[26] S. Schreiber, A, Mahmoud, A. Vuia, M.K. Rubbelke, E.
Schmidt, M. Schaller, H. Kandarova, A. Haberland, U.F.
Schafer, U. Bock, H.C. Korting, M. Liebsch, M. SchaferKorting, Reconstructed epidermis versus human and animal
skin in skin absorption studies., Toxicology in Vitro, in press.
[27] M. Schafer-Korting, W. Mehnert, in: C. Nastruzzi (Ed.),
Lipospheres in Drug Targets and Delivery, CRC Press, Boca
Raton, 2004, pp. 127 142.
[28] J.A. Hamilton, Fast flip-flop of cholesterol and fatty acids in
membranes: implications for membrane transport proteins,
Curr. Opin. Lipidol. 14 (2003) 263 271.
[29] J.R. Kanicky, D.O. Shah, Effect of degree, type, and position
of unsaturation on the pKa of long-chain fatty acids, J. Colloid
Interface Sci. 256 (2002) 201 207.
[30] A.C. Williams, B.W. Barry, Penetration enhancers, Adv. Drug
Deliv. Rev. 56 (2004) 603 618.
[31] J.A. Bouwstra, P.L. Honeywell-Nguyen, G.S. Gooris, M.
Ponec, Structure of the skin barrier and its modulation by
vesicular formulations, Prog. Lipid Res. 42 (2003) 1 36.
[32] V. Jenning, A. Gysler, M. Schafer-Korting, S.H. Gohla,
Vitamin A loaded solid lipid nanoparticles for topical use:
occlusive properties and drug targeting to the upper skin, Eur.
J. Pharm. Biopharm. 49 (2000) 211 218.
[33] K. Jores, A. Haberland, S. Wartewig, K. Mader, W. Mehnert,
Solid lipid nanoparticles (SLN) and oil-loaded SLN studied
by spectrofluorometry and Raman spectroscopy, Pharm. Res.,
in press.

Glossary of abbreviations
C: CompritolR 888 ATO
P: PrecirolR ATO 5
O: oleic acid
M: MiglyolR 812
SLN: solid lipid nanoparticle(s); C, P added indicate the lipid
component
NLC: nanostructured lipid carrier(s); C, M, P, O added indicate the
lipid component
NE: nanoemulsion(s); M, O added indicate the lipid component
PS: parelectric spectroscopy

S. Lombardi Borgia et al. / Journal of Controlled Release 110 (2005) 151163

FS: fluorescence spectroscopy
DSC: differential scanning calorimetry
PCS: photon correlation spectroscopy
PI: polydispersity index
LD: laser scattering
LD95%: maximum particles size of 95% of the particles
f 0: dipole mobility
De: dipole density
ABU: arbitrary brightness unit
R 2: coefficient of correlation
SC: stratum corneum
VE: viable epidermis
SD: superficial dermis
DD: deep dermis
I sc, I ve, I sd, I dd: pixel intensity (ABU) recorded on the SC, VE, SD,
DD

163

B: background pixel intensity (ABU)

PBS: phosphate-buffered saline
PEE: penetration enhancing effect
F(k): dye spectrum in pure solid lipid
O(k): dye spectrum in pure fluid lipid
N theor.(k): dye theoretical spectrum in dispersion
N meas.(k): dye measured spectrum in dispersion
c N: dye concentration
c F, c O: dye concentration in solid or liquid NLC sub compartment
MIN(c F d c O): minimization of the integrated least square error
function
: particle diameter (nm)
x : arithmetic mean value
S.D.: standard deviation
CCD: charge coupled device