Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs: A Comparison of Important Physicochemical Characteristics of Different Vesicle Types

Copyright © 2012 American Scientific Publishers All rights reserved Printed in the United States of America

Journal of Biomedical Nanotechnology Vol. 8, 1–11, 2012

Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs: A Comparison of Important Physicochemical Characteristics of Different Vesicle Types VassilikiNtimenou1 , Alfred Fahr2 , and Sophia G. Antimisiaris1 3 ∗ 1

2

Laboratory of Pharmaceutical Technology, Department of Pharmacy, University of Patras, Rio 26510, Greece Department of Pharmaceutical Technology, Friedrich-Schiller-University Jena, Lessing-str. 8, D-07743 Jena, Germany 3 Institute of Chemical Engineering and High Temperatures, FORTH/ICE-HT, Rio 26504, Patras, Greece

Keywords: Elastic Liposomes, Elasticity Of Vesicles, Transferosomes, Invasomes, Edge Activators, Integrity, Stability, Dermal Transdermal Delivery.

1. INTRODUCTION During the past decades there has been wide interest in exploring new techniques for increasing the absorption of drugs through the skin.1–4 Since the first paper to report the effectiveness of liposomes for skin delivery was published by Mezei and Gulasekharam in 1980,5 it has become evident that in most cases classic liposomes are of little or no value as carriers for transdermal drug delivery, as they cannot penetrate deeply into the skin. While conventional liposomes [CLs] were reported to demonstrate local and rarely transdermal effects,6 deformable liposomes were found to penetrate into deeper skin layers, carrying therapeutic concentrations of drugs, when applied under non-occluded conditions.7 Deformable liposomes or transfersomes [TRs] were the first generation of elastic ∗

Author to whom correspondence should be addressed.

J. Biomed. Nanotechnol. 2012, Vol. 8, No. 4

vesicles.6 They consist of phospholipids and edge activators. An edge activator is a surfactant (usually single chain with a high-curvature radius) that destabilizes lipid bilayers and increases vesicle deformability,8 e.g., sodium cholate, sodium deoxycholate, span (60, 65, 80), tween (20, 60, 80) or di-potassiumglycyrrhizinate. Other types of elastic vesicle compositions studied include the so-called invasomes (INVs), second generation elastic vesiclesthat consist of mixtures of phospholipids and terpenes(hydrocarbonsemployed as percutaneous penetration enhancers) in ethanol.9 10 In this case, the skin penetration enhancement is a result of the increased drug solubility in the terpene-treated stratum corneum (SC). Furthermore, terpenes interact with intercellular lipids perturbing their lamellar packing. In general, exploitation of the mechanisms by which the various types of deformable vesicles improve drug delivery in skin, has been the main objective of

1550-7033/2012/8/001/011

doi:10.1166/jbn.2012.1426

1

RESEARCH ARTICLE

The aim of this study is to evaluate the influence of different lipid vesicular systems on the skin permeation abilityof hydrophilic molecules, and understand if and which vesicle physicochemical properties may be used as predictive tools. Calcein and carboxyfluorescein were used as hydrophilic drug models. All vesicles (conventional liposomes [CLs], transfersomes [TRs] and invasomes [INVs], were characterized for particle size distribution, -potential, vesicular shape and morphology, encapsulation efficiency, integrity, colloidal stability, elasticity and finally in vitro human skin permeation. Dynamic light scattering (DLS) and cryo-transmission electron microscopy (cryo-TEM) defined that almost all vesicles had spherical structure, low polydispersity (PI < 02) and nanosize. Elasticity values (measured by extrusion through membranes) were in the order INVs > TRs > CLs. Three vesicle types were selected (having different elasticity) and in vitro skin permeation experiments demonstrated thatcalcein permeation was minimal from an aqueous solution, slightly enhanced from CLs, and enhanced by 1.8 and 7.2 times from TRs and INVs, respectively. Permeation and elasticity values were correlated by rank order but not linearly, indicating that elasticity can be used as a crude predictive tool for enhancement of skin transport. Drug encapsulation efficiency was not found to be an important factor in the current study.

RESEARCH ARTICLE

Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs

several investigations, and two main theorieshave been proposed.11 12 According to the first, vesicles act as drug carrier systems and intact vesicles enter into the SC carrying vesicle-bound drug molecules. The driving force for vesicles to enter into the skin is xerophobia (tendency to avoid dry surroundings), although it has been evidenced that the water gradient across the skin may not be linear and there may be a relatively ’dry’ region within the SC.13 According to the secondtheory, vesicles act as penetration enhancers; vesicle bilayers enter the SC and subsequently modify the intercellular lipid lamellae facilitating the penetration of free drug molecules into and across the SC. Regardless of the exact mechanismby which each elastic vesicle-type works, it would be interesting to know which vesicle physicochemical properties should be considered in order to select the best formulation for each specific application. One important difference between deformable and CLs is the high and stress-dependent adaptability of deformable vesicles, which enables them to squeeze alone between the cells of the SC as intact entities (at least to a certain depth), despite their relatively large average size (compared to the size of the intercellular channels between keratinocytes). This is usually evaluated by measuring the elasticity of vesicles.14 Other important physicochemical factors may be vesicle size, colloidal stability, and providing that the drug is transported in encapsulated form, also the vesicle encapsulation efficiency and integrity. The aim of this study is to compare various physicochemical characteristics of two different types of elastic liposomes which are widely used for enhancement of the transdermal delivery of hydrophilic drugs, TRs15 16 and INVs.9 10 For each elastic liposome type, appropriate control formulations were evaluated under the same conditions in order to understand the importance of each specific property for determination of the vesicle transdermal delivery potential. In both cases (TR and INV), vesicles with different compositions and increasing edge activator concentrations were constructed and evaluated. Small hydrophilic dyes (5,6-carboxyfluorescein (CF) orcalcein) were used as models of low-molecular weight hydrophilic drugs, and the permeation of calcein through human skin from selected (calcein-encapsulating) vesicle types, was evaluated.

2. MATERIALS AND METHODS 2.1. Materials Purified soybean phosphatidyl-choline (Lipoid S-75) was purchased from Lipoid, Germany. Mixtures of soybean lecithin in ethanol 75:25 w/w (NAT 8539) was from Nattermann Phospholipid GmbH, Germany. 5,6-Carboxyfluorescein (CF), calcein and cholesterol (Chol) were obtained by Sigma-Aldrich, Greece. Sodium cholate(Schol) was from Serva (Germany) and tween 80 was from ICI Chemicals, UK. Chloroform and methanol 2

VassilikiNtimenou et al.

were of analytical grade and were obtained from Merck, Germany. R-limonene (LIM) was from Aldrich, SigmaAldrich, Greece. Citral, cineol and all salts used for preparation of buffers were from Merck, Germany. Female human abdominal skin from plastic surgery was used. The subcutaneous fatty tissue was removed from the skin by using a scalpel and surgical scissors and then frozen at −20  C with aluminium foil packed for later use in 3 months at most. 2.2. Preparation of Liposomes 2.2.1. Transfersomes and Conventional Liposomes The thin film method was used.17 In brief, lipid (Lipoid S75), Chol and/or surfactant are dissolved in a chloroform/methanol mixture (1:1 v/v)and placed in a round bottom flask. Solvents are evaporated under low pressure at 40  C (BuchiRota-Vapor). The lipid thin film is then flashed with nitrogen, for removal of possible solvent traces, and finally hydrated at room temperature with buffer (PBS) pH 7.40 (in the case of empty liposomes) or Tris buffer pH 7.40 containing CF or calcein (3 or 100 mM, respectively; adjusted to be isotonic). For size reduction, the MLV (multilamellar) vesicles are extruded through polycarbonate membranes with pore size of 100 nm (initially) and 50 nm (finally) at room temperature with the aid of a Lipo-so-Fast extruder (Avestin, Canada). Non-entrapped CF or calceinis separated from encapsulated solute, by the Quix Sep micro dialyzer device (Carl Roth GmbH), or alternatively, by size- exclusion chromatography (Sephadex G-50 columns). All formulations are finally adjusted to a concentration of 10% (w/w) of lipid. 2.2.2. Invasomes INVs were prepared using anethanolic phospholipid solution (NAT8539), together with a mixture of terpenes (LIM/citral/cineole at 10:45:45 v/v/v) [as penetration enhancers (PE)] or individual terpenes (LIM, citral or cineole) at 0–2% w/w, and an aqueous solution.9 10 For this, the terpene-mixture (PE) or individual terpeneis initially dissolved in NAT8539 and the mixture is vortexed for 5 min followed by sonication to produce a clear transparent solution. PBS buffer pH 7.40 (for empty liposomes) or a solution of CF or calcein (3 or 100 mM, for CF- or calcein-encapsulating liposomes) is then added drop-wise, with a syringe, under constant vortexing. Vortex is applied for 5 additional min. The MLV dispersions produced are subsequently extruded as described above. The preparation is performed at room temperature. In case of CF- or calcein-encapsulating INVs, non-entrapped soluteis separated from encapsulated, as described above. Final dispersions were adjusted to contain 10% w/w lipid. J. Biomed. Nanotechnol. 8, 1–11, 2012

VassilikiNtimenou et al.

Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs

2.3. Particle Size Distribution, -Potential and Size Stability Study The mean diameter, polydispersity and zeta potential of the vesicles are determined by DLS (dynamic light scattering) measurements using a Nano-Zetasizer (Malvern Instruments). For the measurements, 5 l from each vesicle suspension are mixed with buffer PBS pH 7.40. Each sample is measured four times over 5 min. All measurements are conducted at 25  C. The polydispersity index (PI) is a measure of the homogeneity of the vesicle suspensions. A value of PDI < 02 indicates a homogenous vesicle population. For the size stability study, INVs with increasing amounts of PE, encapsulating a hydrophilic dye (CF, 3 mM), and TRs containing tween 80 or Scholare prepared (as described above) and stored at different temperatures (4  C, room temperature (RT) and 37  C). At specified time periods (usually 15 d and 30 d, and in some cases up to 60 or 120 d) samples are taken and vesicle size distribution and PI are determined as described above. 2.4. Extrusion Measurements—Elasticity of Vesicles

The elasticity (E) of the bilayers is directly proportional to j (flux) which is the rate of penetration through a permeability barrier, rv is the size of vesicles (after the extrusion) and rp is the pore size of the barrier. The average vesicle size is measured before extrusion and after extrusion, by DLS as described above. For each vesicle composition, three suspensions are measured 5 times each. The compositions of the different types of formulations are summarized in Table I. 2.5. Vesicle Encapsulation Efficiency and Membrane Integrity Calcein-encapsulating vesicles are prepared as described above, by using a calcein solution (100 mM in buffer pH 7.40, adjusted with NaCl to be isotonic) as hydrating solution of the lipidic thin film. The encapsulation efficiency of each vesicle type is calculated by measuring the fluorescence intensity (FI) of a given amount of vesicles (lipid), after they are disrupted with TRITON X-100 J. Biomed. Nanotechnol. 8, 1–11, 2012

Latency % =

F Ifinal − F Iinitial × 100 F Ifinal

Where FIinitial and FIfinal are the FIs of the dispersions before and after disruption of the liposomes with Triton X-100, respectively (FIfinal values were corrected for dilution). 2.6. Morphology of Vesicles by Cryo-TEM The morphology of the elastic vesicles was examined by cryo-transmission electron microscopy (cryo-TEM), at the Institute for Electron Microscopy of the FSU Jena, Germany. The diluted liposome formulations (1:1 v/v) are applied to a carbon coated microscopy grid in such a way that a very thin aqueous film is formed, which is then rapidly plunged into a cooling medium (ethane) just above its freezing point20 (where the film vitrifies rapidly without crystallization). The grid with the vitrified film is then transferred to the microscope by a cryo-transfer unit and examined at liquid nitrogen temperature in transmission mode. A dimension of about 4–5 nm is the smallest that can be resolved and therefore ordinary surfactant micelles are usually seen as dots.14 2.7. Calcein permeation Experiments in Human Skin Liposomes (purified by Sephadex G-50 column) were measured in terms of lipid concentration and adjusted to 1 mg/ml. A calcein solution, diluted properly to havethe same concentration as that in the vesicle dispersions was used as control. The experiment was conducted using human skin (from plastic surgery) from three different donors in triplicate each time. The formulations tested were INVs with 1% PE; TRs with 1.1% Schol and Liposomes S75:Chol (2:1 mol/mol) as control formulation of non-elastic CLs. Skin permeation was studied using Franz diffusion cells (PermeGear, Inc. USA) with an effective permeation area and receptor cell volume of 3.14 cm2 and 17 ml, respectively. The temperature was maintained at 37 ± 1  C. The receptor compartment contained 17 ml PBS (pH 7.4) and was constantly stirred by magnetic stirrer at 100 rpm. Human female skin from abdominal areas was obtained from cosmetic surgery, isolated from fat tissue and stored at −20  C. The full skin was mounted 3

RESEARCH ARTICLE

In order to measure vesicle elasticity (E), 1 ml of each formulation is extruded 10 times (Lipo-so-Fast Extruder, Avestin) through polycarbonate filters with a pore size of 50 nm (rp  at a constant pressure of 0.5 MPa (Lipo-soFast Pneumatic Actuator, Avestin). The temperature was approximately 20  C throughout the whole experiment. The time (s) required to extrude the formulations for 10 times is measured and used for the calculation of Jflux .18 E is finally calculated according the equation:  2 r E=j· v rp

(1% w/w final concentration). The amount of calceinis calculated by use of a calibration curve. The membrane integrity of the various elastic vesiclestypes is evaluated, as previously described19 by measuring the retention of calcein in the vesicles during incubation of the vesicle dispersions for a 48 h period, at 37  C. Liposome samples are drawn from the incubated dispersions at specified time points and the retention of calcein in the vesicles is calculated after measuring its latency by use of the following equation:

Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs

VassilikiNtimenou et al.

Table I. Size of conventional and elastic liposomes, before extrusion through polycarbonate membrane with a pore size of 50 nm and size alteration caused by extrusion. Encapsulation efficiency of calcein (D/L) in all types of vesicles, and -potential (of CF encapsulating vesicles), in mV. Values given are mean values of 3 different batches ± standard deviation of the mean. Formulations Conventional liposomes S-75/Chol Transfersomes Schol 0.5% 1.1% 1.8% 4.3% Tween 80 0.5% 1.1% 1.8% 4.3% Invasomes Control 1% cineole 1% citral 1% LIM 0.5% PE 1% PE 1.5% PE

Size before extrusion

Size after extrusion

Size alteration (%)

D/L (mole/mole)

-potential (mV)

108 ± 1.9

98 ± 1.8

−9.26%

−0.045 ± 0.0028

—

— 82 ± 3.5 85 ± 1.9 110 ± 22.3

— 83 ± 7.4 86 ± 7.3 104 ± 24

— +1.71% +1.18% −5.45%

0.030 ± 0.0015 0.021 ± 0.0014 0.014 ± 0.00026 —

— −11.3 ± 2.2 −13.7 ± 4.1 −8.5 ± 3.5

— 88 ± 3.4 87 ± 2.2 78 ± 1.7

— 83 ± 5.8 79 ± 3.2 73 ± 0.2

— −5.68% −9.20% −6.41%

0.068 ± 0.0016 0.062 ± 0.0032 0.045 ± 0.0044 —

— −12.1 ± 1.9 −9.6 ± 3.3 −11.1 ± 2.1

110 ± 1.0 108 ± 7.1 105 ± 0.9 143 ± 2.8 — 112 ± 0.5 298 ± 33.8

100 ± 1.3 95 ± 2.6 100 ± 2.8 122 ± 2.1 — 108 ± 4.6 132 ± 12.4

−9.09% −12.04% −4.76% −14.69% — −3.57% −55.70%

0.021 ± 0.00016 — — — 0.013 ± 0.00039 0.013 ± 0.0010 0.019 ± 0.0026

−10.8 ± 1.9 — — — −13.1 ± 2.5 −11.9 ± 3.4 −15.7 ± 2.8

RESEARCH ARTICLE

Notes: Not measured.

on a receptor compartment with the stratum corneum side towards the donor compartmentand the dermal side bathed in the receptor fluid. The liposomal formulations or plain calcein solution (1 ml) were applied on the skin in the donor compartment, which was then covered with parafilm to avoid sample evaporation. Samples (10 l) were withdrawn through the sampling port of the diffusion cell at predetermined time intervals over 48 h and analyzed for drug content by a fluorophotometer (Shimadzu RF-1501). The receptor phase was immediately replenished with equal volume of fresh PBS buffer. Sink conditions were maintained throughout all the experiment. The cumulative amount of drug permeated per unit area was plotted as a function of time; the steady-state permeation rate (Jss) and lag time (LT, h) were calculated from the slope and X-intercept of the linear portion, respectively. The enhancement ratio (ER) was calculated as: transdermal flux from vesicular formulation/transdermal flux from plain drug solution.

3. RESULTS AND DISCUSSION 3.1. Physicochemical Characteristics of the “Elastic” Liposome Formulations The mean diameters and PIs of the various compositions of CLs, TRs and INVs, as measured after their preparation, are presented in Tables I and II, respectively. Meandiameters ofTRs range between 82–110 nm. The vesicles with the highest amount of Schol (4.3% w/w) are significantly larger (mean diameter 110 nm) compared to the 4

other two formulations, and also have significantly higher PI (Table II). INVs are generally slightly larger than TRs, with diameters ranging between 105 and 298 nm, nevertheless most INVs have diameters similar to that of the control INVs(with noterpenes), with the exception of LIM-containing-INVs (143 nm) and 1.5% PE-containing INVs (298 nm) in which a considerable (and significant p < 001) increase in the vesicle diameter (compared to

Table II. PI values of conventional and elastic liposomes, before and after extrusion through polycarbonate membrane with a pore size of 50 nm. Values given are mean values of 3 different batches ± standard deviation of the mean. Formulations Liposomes S-75/Chol(2:1) TRs Schol 1.1% 1.8% 4.3% Tween 80 1.1% 1.8% 4.3% INVs Without PE Cineole 1% Citral 1% LIM 1% PE 1% PE 1.5%

PI before extrusion

PI after extrusion

0.073 ± 0.0090

0.100 ± 0.0075

0.078 ± 0.020 0.057 ± 0.017 0.307 ± 0.0382

0.096 ± 0.020 0.061 ± 0.033 0.255 ± 0.0111

0.068 ± 0.014 0.089 ± 0.0042 0.066 ± 0.0143

0.072 ± 0.0013 0.084 ± 0.0272 0.065 ± 0.0034

0.118 ± 0.0086 0.097 ± 0.0087 0.107 ± 0.017 0.108 ± 0.0085 0.086 ± 0.0092 0.295 ± 0.0536

0.085 ± 0.0065 0.075 ± 0.0008 0.078 ± 0.0052 0.082 ± 0.0014 0.077 ± 0.0064 0.091 ± 0.0133

J. Biomed. Nanotechnol. 8, 1–11, 2012

VassilikiNtimenou et al.

Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs

control) is observed (30% and 170% increase, respectively). The last formulation also had a high PI. Similar results were also obtained in separate experiments performed with CF- encapsulating CLs and elastic vesicles, as seen in Figures 1(A) and (B), for Schol and tween-80 containing-TRs respectively, and in Figure 1(C) for PE-containing INVs. As seen, the mean diameter of Schol (or Tween) containing TRs was almost not affected when surfactant content was increased up to 1.8%;but the PI of the

(A)

transfersomes with sodium cholate 0.4

150

0.3

PI z-average

100

0.2

50

0.1

0

PI

z-average (nm)

200

0.0 0.0

0.5

1.0

1.5

2.0

%Schol

0.4

150

0.3

100

0.2

50

0.1

0

0.0 0.0

0.5

1.0

1.5

PI

transfersomes with tween 80 200

Schol 1.1 PE-0

2.0

160

%tween 80

Schol 1.8 Tween 1.1 PE-0.5 PE-1 PE-1.5

Tween 1.8 PE-2.0

RT

120

Invasomes 0.4

150

0.3

100

0.2

50

0.1

0

0.0

80

Mean Diameter (nm)

200

PI

z-average (nm)

(C)

40 0

Day 1 160

Day 15

Day 60

Day 120

37 ºC

120 80 40

0.0

0.5

1.0

1.5

2.0

% PE Fig. 1. Mean diameter (black solid squares) and Polydispersity Index (PI) (blue open squares) of CF-encapsulating TRs that contain increasing amounts of Schol (A) or Tween-80 (B), and INVs with various amounts of PE (terpene mixture) (C).

J. Biomed. Nanotechnol. 8, 1–11, 2012

0

Day 1

Day 30

Fig. 2. Size stability of various types of TRs (containing Schol or Tween-80), and various types of INVs during storage at room temperature (RT) or 37  C. Each value is the mean from at least 3 different measurements and the bars represent standard deviation of each mean.

5

RESEARCH ARTICLE

z-average (nm)

(B)

Schol-containing dispersions gradually increased with the increase of Schol amount, as also observed for Tweencontaining dispersions when surfactant content was 1.8%. Both, vesicle mean diameter and PI of INV dispersions were significantly increased when PE concentration was higher than 1% (w/w) (Fig. 1(C)), in line with the measurements of the empty INVs (Table I). The fact that the actual mean diameter values as well as the PI values are not the same in the two sets of formulations (empty elastic vesicles vs. CF-encapsulating ones) (Table I and Fig. 1) may be attributed to the effect of sample color on the DLS measurements, since encapsulation of a hydrophilic molecule in the aqueous phase of vesicles is not expected to have an effect on the vesicle size. As seen in Table I, the zeta potential values of all the elastic liposome types measured, was negative (ranging from −8.5 to −17.7) and quite similar for all cases, suggesting that the addition of different edge activator types in the concentration range used do not influence the surface charge of the vesicles produced. In regards to the physical stability of the various types of TRs and INVs evaluated, during storage at RT (approx. 23  C) or 37  C (Fig. 2), all types of elastic vesicles studied retain their mean size value during 15 or 30 d of storage under all study conditions (stability was also followed at 4  C [results not shown]), and in some cases even after 60 or 120 d. The only significant increase of vesicle mean diameter observed was in the case of 1.8% Schol-containing TRs after 30 d at 37  C (Fig. 2, lower graph). In general, for most of the elastic vesicle types only slight, non-significant, size increases were observed after 60 or 120 d, indicating that they have very high colloidal stability.

Elastic Vesicles for Transdermal Drug Delivery of Hydrophilic Drugs

Elasticity is an important feature of ultra-deformable vesicles which differentiates them from CLs (with more rigid membranes).14 21 Size measurements (Table I) showed that the original vesicle size did not change substantially after extrusionin most cases (size alterations were always