Chitosan-Based Nanomedicine to Fight Genital Candida Infections: Chitosomes.

Vaginal infections are associated with high recurrence, which is often due to a lack of efficient treatment of complex vaginal infections comprised of several types of pathogens, especially fungi and bacteria. Chitosan, a mucoadhesive polymer with known antifungal effect, could offer a great improvement in vaginal therapy; the chitosan-based nanosystem could both provide antifungal effects and simultaneously deliver antibacterial drugs. We prepared chitosan-containing liposomes, chitosomes, where chitosan is both embedded in liposomes and surface-available as a coating layer. For antimicrobial activity, we entrapped metronidazole as a model drug. To prove that mucoadhesivness alone is not sufficient for successful delivery, we used Carbopol-containing liposomes as a control. All vesicles were characterized for their size, zeta potential, entrapment efficiency, and in vitro drug release. Chitosan-containing liposomes were able to assure the prolonged release of metronidazole. Their antifungal activity was evaluated in a C. albicans model; chitosan-containing liposomes exhibited a potent ability to inhibit the growth of C. albicans. The presence of chitosan was crucial for the system's antifungal activity. The antifungal efficacy of chitosomes combined with antibacterial potential of the entrapped metronidazole could offer improved efficacy in the treatment of mixed/complex vaginal infections.

In vitro models to estimate drug penetration through the compromised stratum corneum barrier.

OBJECTIVES: The phospholipid vesicle-based permeation assay (PVPA) is a recently established in vitro stratum corneum model to estimate the permeability of intact and healthy skin. The aim here was to further evolve this model to mimic the stratum corneum in a compromised skin barrier by reducing the barrier functions in a controlled manner. METHODS: To mimic compromised skin barriers, PVPA barriers were prepared with explicitly defined reduced barrier function and compared with literature data from both human and animal skin with compromised barrier properties. Caffeine, diclofenac sodium, chloramphenicol and the hydrophilic marker calcein were tested to compare the PVPA models with established models. RESULTS AND DISCUSSIONS: The established PVPA models mimicking the stratum corneum in healthy skin showed good correlation with biological barriers by ranking drugs similar to those ranked by the pig ear skin model and were comparable to literature data on permeation through healthy human skin. The PVPA models provided reproducible and consistent results with a distinction between the barriers mimicking compromised and healthy skin. The trends in increasing drug permeation with an increasing degree of compromised barriers for the model drugs were similar to the literature data from other in vivo and in vitro models. CONCLUSIONS: The PVPA models have the potential to provide permeation predictions when investigating drugs or cosmeceuticals intended for various compromised skin conditions and can thus possibly reduce the time and cost of testing as well as the use of animal testing in the early development of drug candidates, drugs and cosmeceuticals.

Liposomes for (trans)dermal drug delivery: the skin-PVPA as a novel in vitro stratum corneum model in formulation development.

Penetration potential of vesicles destined for trans(dermal) administration remains to be of great interests both in respect to drug therapy and cosmetic treatment. This study investigated the applicability of the phospholipid vesicle-based permeation assay (PVPA) as a novel in vitro skin barrier model for screening purposes in preformulation studies. Various classes of liposomes containing hydrophilic model drug were examined, including conventional liposomes (CLs), deformable liposomes (DLs) and propylene glycol liposomes (PGLs). The size, surface charge, membrane deformability and entrapment efficiency were found to be affected by the vesicle lipid concentration, the presence of the surfactant and propylene glycol. All liposomes exhibited prolonged drug release profiles with an initial burst effect followed by a slower release phase. The permeation of the drug from all of the tested liposomes, as assessed with the mimicked stratum corneum--PVPA model, was significantly enhanced as compared to the permeability of the drug in solution form. Although the DLs and the PGLs exhibited almost the same membrane elasticity, the permeability of the drug delivered by PGLs was higher (6.2 × 10⁻6 cm/s) than DLs (5.5 × 10⁻6 cm/s). Therefore, this study confirmed both the potential of liposomes as vesicles in trans(dermal) delivery and potential of the newly developed skin-PVPA for the screening and optimization of liposomes at the early preformulation stage.

Liposomal formulations of poorly soluble camptothecin: drug retention and biodistribution.

Camptothecin (CPT) represents a potent anticancer drug. However, its therapeutic use is impaired by both drug solubility, hydrolysis, and protein interactions in vivo. Use of liposomes as a drug-formulation approach could overcome some of these challenges. The aim of this study was to perform a mechanistic study of the incorporation and retention of the lipophilic parent CPT compound in different liposome formulations using radiolabeled CPT and thus to be able to identify promising CPT delivery systems. In this context, we also wanted to establish an appropriate mouse tumor model, in vivo scintigraphic imaging, and biodistribution methodology for testing the most promising formulation. CPT retention in various liposome formulations after incubation in buffer and serum was determined. The HT-29 mouse tumor model, (111)In-labeled liposomes, as well as (3)H-labeled CPT were used to investigate the biodistribution of liposomes and drug. The ability of different liposome formulations to retain CPT in buffer was influenced by lipid concentration and drug/lipid ratio, rather than lipid composition. The tested formulations were cleared from the blood in the following order: CPT solution > CPT liposomes > (111)In-labeled liposomes, and liposomes mainly accumulated in the liver. Lipid composition did not influence CPT retention to the same extent as earlier observed from incorporation studies. The set-up for the biodistribution study works well and is suited for future in vivo studies on CPT liposomes. The biodistribution study showed that liposomes circulated longer than free drug, but premature release of drug from liposomes occurred. Further studies to develop formulations with higher retention potential and prolonged circulation are desired.

Bacterial extracellular vesicles: towards realistic models for bacterial membranes in molecular interaction studies by surface plasmon resonance.

One way to mitigate the ongoing antimicrobial resistance crisis is to discover and develop new classes of antibiotics. As all antibiotics at some point need to either cross or just interact with the bacterial membrane, there is a need for representative models of bacterial membranes and efficient methods to characterize the interactions with novel molecules -both to generate new knowledge and to screen compound libraries. Since the bacterial cell envelope is a complex assembly of lipids, lipopolysaccharides, membrane proteins and other components, constructing relevant synthetic liposome-based models of the membrane is both difficult and expensive. We here propose to let the bacteria do the hard work for us. Bacterial extracellular vesicles (bEVs) are naturally secreted by Gram-negative and Gram-positive bacteria, playing a role in communication between bacteria, as virulence factors, molecular transport or being a part of the antimicrobial resistance mechanism. bEVs consist of the bacterial outer membrane and thus inherit many components and properties of the native outer cell envelope. In this work, we have isolated and characterized bEVs from one Escherichia coli mutant and three clinical strains of the ESKAPE pathogens Klebsiella pneumoniae, Acinetobacter baumannii, and Pseudomonas aeruginosa. The bEVs were shown to be representative models for the bacterial membrane in terms of lipid composition with speciesstrain specific variations. The bEVs were further used to probe the interactions between bEV and antimicrobial peptides (AMPs) as model compounds by Surface Plasmon Resonance (SPR) and provide proof-of-principle that bEVs can be used as an easily accessible and highly realistic model for the bacterial surface in interaction studies. This further enables direct monitoring of the effect induced by antibiotics, or the response to host-pathogen interactions.

Tailored anti-biofilm activity - Liposomal delivery for mimic of small antimicrobial peptide.

The eradication of bacteria embedded in biofilms is among the most challenging obstacles in the management of chronic wounds. These biofilms are found in most chronic wounds; moreover, the biofilm-embedded bacteria are considerably less susceptible to conventional antimicrobial treatment than the planktonic bacteria. Antimicrobial peptides and their mimics are considered attractive candidates in the pursuit of novel therapeutic options for the treatment of chronic wounds and general bacterial eradication. However, some limitations linked to these membrane-active antimicrobials are making their clinical use challenging. Novel innovative delivery systems addressing these limitations represent a smart solution. We hypothesized that incorporation of a novel synthetic mimic of an antimicrobial peptide in liposomes could improve its anti-biofilm effect as well as the anti-inflammatory activity. The small synthetic mimic of an antimicrobial peptide, 7e-SMAMP, was incorporated into liposomes (~280 nm) tailored for skin wounds and evaluated for its potential activity against both biofilm formation and eradication of pre-formed biofilms. The 7e-SMAMP-liposomes significantly lowered inflammatory response in murine macrophages (~30 % reduction) without affecting the viability of macrophages or keratinocytes. Importantly, the 7e-SMAMP-liposomes completely eradicated biofilms produced by Staphylococcus aureus and Escherichia coli above concentrations of 6.25 µg/mL, whereas in Pseudomonas aeruginosa the eradication reached 75 % at the same concentration. Incorporation of 7e-SMAMP in liposomes improved both the inhibition of biofilm formation as well as biofilm eradication in vitro, as compared to non-formulated antimicrobial, therefore confirming its potential as a novel therapeutic option for bacteria-infected chronic wounds.

WIND-PVPA: Water/Ion NMR Detected PVPA to assess lipid barrier integrity in vitro through quantification of passive water- and ion transport.

Water/Ion NMR Detected - Phospholipid Vesicle Permeability Assay (WIND-PVPA), is presented as a novel, straightforward and automatable method to assess lipid barrier integrity in vitro. The apparent permeability constants of water- and ions across the PVPA barriers are determined in a one-pot experiment under the influence of membrane-active guest molecules. NMR spectroscopy is used to quantify the water directly (D2O) and the ions indirectly (complexed with EDTA) as a function of time. WIND-PVPA is demonstrated using four anti-microbial peptides, to show that membrane active molecules can be differentiated by their disruptive influence on the PVPA system. The results obtained are compared with explicit molecular dynamics simulations of lipid bilayers, AMPs, water and salt, where the motions of all individual water molecules relative to the lipid bilayer are monitored over the course of the simulations, allowing the calculation of theoretical apparent permeability constants of the corresponding single bilayer systems. Proof-of-principle is presented that WIND-PVPA can be used to evaluate the lipid barrier destabilizing effect of active guest molecules by measuring changes in passive water- and ion permeabilities upon exposure. The method is highly flexible in terms of barrier composition, choice of probes and membrane active compounds.

Simultaneous assessment of in vitro lipolysis and permeation in the mucus-PVPA model to predict oral absorption of a poorly water soluble drug in SNEDDSs.

The prediction of the in vivo performance of self-nanoemulsifying drug delivery systems (SNEDDSs) is currently gaining increasing attention. Therefore, the need for reliable in vitro models able to assess the drug solubilization capacity of such formulations upon in vitro lipolysis, as well as to concomitantly evaluate in vitro drug permeation, has become ever so evident. In the current study, the high-throughput in vitro intestinal lipolysis model was combined with the mucus-PVPA in vitro permeation model to study the solubilization capacity of SNEDDSs for the poorly water-soluble drug fenofibrate and to study the consequent drug permeation. Moreover, drug solubilization and permeation were evaluated both in the presence and absence of lipolysis. The results obtained demonstrated that the presence of in vitro lipolysis significantly impacted the solubilization and permeation profiles of fenofibrate compared to its absence. The results were in accordance with already published in vivo data regarding the same fenofibrate-loaded SNEDDSs. Additionally, the correlation between the in vitro permeation data and in vivo plasma concentration in rats was found to be excellent both in the presence and absence of lipolysis (R2 > 0.98), highlighting the ability of the developed combined in vitro model to predict in vivo drug absorption.

Predicting Oral Absorption of fenofibrate in Lipid-Based Drug Delivery Systems by Combining In Vitro Lipolysis with the Mucus-PVPA Permeability Model.

The aim of this work was to develop a new in vitro lipolysis-permeation model to predict the in vivo absorption of fenofibrate in self-nanoemulsifying drug delivery systems (SNEDDSs). More specifically, the in vitro intestinal lipolysis model was combined with the mucus-PVPA (Phospholipid Vesicle-based Permeation Assay) in vitro permeability model. Biosimilar mucus (BM) was added to the surface of the PVPA barriers to closer simulate the intestinal mucosa. SNEDDSs for which pharmacokinetic data after oral dosing to rats was available in the literature were prepared, and the ability of the SNEDDSs to maintain fenofibrate solubilized during in vitro lipolysis was determined, followed by the assessment of drug permeation across the mucus-PVPA barriers. The amount of drug solubilized over time during in vitro lipolysis did not correlate with the AUC (area under the curve) of the plasma drug concentration curve. However, the AUC of the drug permeated after in vitro lipolysis displayed a good correlation with the in vivo AUC (R2 > 0.9). Thus, it was concluded that the in vitro lipolysis-mucus-PVPA permeation model, simulating the physiological digestion and absorption processes, was able to predict in vivo absorption data, exhibiting great potential for further prediction of in vivo performance of SNEDDSs.

Drug loading to mesoporous silica carriers by solvent evaporation: A comparative study of amorphization capacity and release kinetics.

The sorption of poorly aqueous soluble active pharmaceutical ingredients (API) to mesoporous silica carriers is an increasingly common formulation strategy for dissolution rate enhancement for this challenging group of substances. However, the success of this approach for a particular API depends on an array of factors including the properties of the porous carrier, the loading method, or the attempted mass fraction of the API. At present, there is no established methodology for the rational selection of these parameters. In the present work, we report a systematic comparison of four well-characterised silica carriers and seven APIs loaded by the same solvent evaporation method. In each case, we find the maximum amorphization capacity by x-ray powder diffraction analysis and measure the in vitro drug release kinetics. For a selected case, we also demonstrate the potential for bioavailability enhancement by a permeation essay.

Current challenges and future perspectives in oral absorption research: An opinion of the UNGAP network.

Although oral drug delivery is the preferred administration route and has been used for centuries, modern drug discovery and development pipelines challenge conventional formulation approaches and highlight the insufficient mechanistic understanding of processes critical to oral drug absorption. This review presents the opinion of UNGAP scientists on four key themes across the oral absorption landscape: (1) specific patient populations, (2) regional differences in the gastrointestinal tract, (3) advanced formulations and (4) food-drug interactions. The differences of oral absorption in pediatric and geriatric populations, the specific issues in colonic absorption, the formulation approaches for poorly water-soluble (small molecules) and poorly permeable (peptides, RNA etc.) drugs, as well as the vast realm of food effects, are some of the topics discussed in detail. The identified controversies and gaps in the current understanding of gastrointestinal absorption-related processes are used to create a roadmap for the future of oral drug absorption research.

Impact of Mucin on Drug Diffusion: Development of a Straightforward in Vitro Method for the Determination of Drug Diffusivity in the Presence of Mucin.

Mucosal drug delivery accounts for various administration routes (i.e., oral, vaginal, ocular, pulmonary, etc.) and offers a vast surface for the permeation of drugs. However, the mucus layer which shields and lubricates all mucosal tissues can compromise drugs from reaching the epithelial site, thus affecting their absorption and therapeutic effect. Therefore, the effect of the mucus layer on drug absorption has to be evaluated early in the drug-development phase, prior to in vivo studies. For this reason, we developed a simple, cost-effective and reproducible method employing UV-visible localized spectroscopy for the assessment of the interaction between mucin and drugs with different physicochemical characteristics. The mucin-drug interaction was investigated by measuring the drug relative diffusivity (Drel) in the presence of mucin, and the method was validated by fitting experimental and mathematical data. In vitro permeability studies were also performed using the mucus-covered artificial permeation barrier (mucus-PVPA, Phospholipid Vesicle-based Permeation Assay) for comparison. The obtained results showed that the diffusion of drugs was hampered by the presence of mucin, especially at higher concentrations. This novel method proved to be suitable for the investigation on the extent of mucin-drug interaction and can be successfully used to assess the impact that the mucus layer has on drug absorption.

The Vaginal-PVPA: A Vaginal Mucosa-Mimicking In Vitro Permeation Tool for Evaluation of Mucoadhesive Formulations.

Drug administration to the vaginal site has gained increasing attention in past decades, highlighting the need for reliable in vitro methods to assess the performance of novel formulations. To optimize formulations destined for the vaginal site, it is important to evaluate the drug retention within the vagina as well as its permeation across the mucosa, particularly in the presence of vaginal fluids. Herewith, the vaginal-PVPA (Phospholipid Vesicle-based Permeation Assay) in vitro permeability model was validated as a tool to evaluate the permeation of the anti-inflammatory drug ibuprofen from liposomal formulations (i.e., plain and chitosan-coated liposomes). Drug permeation was assessed in the presence and absence of mucus and simulated vaginal fluid (SVF) at pH conditions mimicking both the healthy vaginal premenopausal conditions and vaginal infection/pre-puberty/post-menopause state. The permeation of ibuprofen proved to depend on the type of formulation (i.e., chitosan-coated liposomes exhibited lower drug permeation), the mucoadhesive formulation properties and pH condition. This study highlights both the importance of mucus and SVF in the vaginal model to better understand and predict the in vivo performance of formulations destined for vaginal administration, and the suitability of the vaginal-PVPA model for such investigations.

Mimicking regional and fasted/fed state conditions in the intestine with the mucus-PVPA in vitro model: The impact of pH and simulated intestinal fluids on drug permeability.

Intestinal drug absorption following oral administration can be influenced by regional conditions (absorbing surface area, bacterial flora, motility, pH, mucus thickness) and food intake, all of which affect drug solubility and permeability. Therefore, it is crucial to assess the impact of these conditions on the drugability of drugs and formulations. In this study, the ability of the liposome-based mucus-PVPA in vitro permeability model to handle relevant intestinal pH conditions was evaluated, together with the investigation on the pH-dependent solubility and permeability profiles of five model drugs. This study additionally evaluated the impact of all commercially available versions of the fasted and fed state simulated intestinal fluids (SIFs) on the integrity of the barriers, and the permeabilities of one hydrophilic and one lipophilic compound were examined under these conditions. The model was found to be well-functioning in all tested pH conditions, and a pH-dependent trend was found for both solubility and permeability profiles for acidic and basic compounds, according to their degree of ionization. Moreover, the mucus layer and its pH-dependent viscosity particularly influenced the permeation of more lipophilic compounds. The PVPA barriers primarily maintained their functionality in the presence of the fed state SIFs, and the permeability of the two tested compounds showed to be influenced by their hydrophilicity/lipophilicity, their degree of interaction with mucus and by the bile salts and phospholipids content in the SIFs. Overall, the obtained results highlight the relevance of studying the effect that pH, mucus and SIFs have on intestinal drug absorption, and suggest the suitability of the mucus-PVPA model for such investigations.

Drug permeability across a phospholipid vesicle-based barrier 4. The effect of tensides, co-solvents and pH changes on barrier integrity and on drug permeability.

In this study the integrity of the recently developed phospholipid vesicle-based permeability barrier in the presence of a variety of co-solvents and tensides has been investigated. Also included are studies of the influence of these additives on drug permeation and the effect of pH changes on the permeability of ionogenic drug compounds. Permeability experiments using the hydrophilic model compound calcein together with polysorbate 80 (Tween 80), polyoxyl 35 castor oil (Cremophor EL), macrogol lauryl ether (Brij 35), sorbitan monolaurate (Span 20), polyethylene glycol 400 (PEG 400), ethanol and dimethylsulphoxide (DMSO) were performed to determine whether the barriers were affected by the presence of these additives in the donor compartment. It was found that the integrity of the phospholipid vesicle-based barriers did not seem to be influenced by Span 20 up to a concentration of 5mg/ml, PEG 400 up to a concentration of 40mg/ml and ethanol and DMSO up to a concentration of 20mg/ml, respectively. Brij 35, Tween 80 and Cremophor EL were however found to be incompatible with the model at all concentrations as the barriers became leaky. Appearance of phospholipid in the donor chamber in presence of these three tensides indicated that the loss of integrity was due to partial dissolution of the phospholipid vesicles from the barrier. The permeability of testosterone was not significantly improved by the presence of the different co-solvents, except for 40 mg/ml PEG 400 and 20 mg/ml DMSO where the permeability was increased. In the pH study the permeability of metoprolol and naproxen was shown to decrease with increasing degree of ionisation according to the pH partition hypothesis. This renders the permeability model suitable for using pH-shift as a factor to influence solubility of drugs as well as to predict segmental absorption in the gastrointestinal tract.

Deformable liposomes for skin therapy with human epidermal growth factor: The effect of liposomal surface charge.

The topical administration of exogenous human epidermal growth factor (hEGF) is a promising approach for improved chronic wound therapy. To develop therapeutically superior hEGF formulation, we prepared hEGF-containing neutral (NDLs), cationic (CDLs) and anionic (ADLs) deformable liposomes (DLs), respectively, since it is expected that the liposomal surface charge can affect both the liposomal physicochemical properties, their skin penetration potential and therapeutic efficacy of liposome-associated drug. All prepared liposomes were of similar size (300-350 nm) with high hEGF load (~80% entrapment efficacy). Among the studied DLs, ADLs were found to be most promising for sustained release of hEGF, as assessed in vitro using the polyamide membrane. Ex vivo studies revealed that all DLs were excellent systems for skin therapy with hEGF and no penetration of hEGF through the full thickness human skin was detected. ADLs provided a depot exhibiting the highest hEGF retention onto the human skin surface. ADLs also revealed enhanced mitogenic activities in human fibroblasts compared to both NDLs and CDLs after 48 hrs treatment. Moreover, hEGF-containing ADLs significantly enhanced mitogenic activity in fibroblast as compared to activity of hEGF solution (positive control). Similar trends were observed in human keratinocytes after 24 hrs of treatment. We proved that the liposomal surface charge affects the therapeutic potential of hEGF-containing liposomes. hEGF-containing ADLs can be a promising nanosystem-based formulation for localized therapy of chronic wounds.

Mucus-PVPA (mucus Phospholipid Vesicle-based Permeation Assay): An artificial permeability tool for drug screening and formulation development.

The mucus layer covering all mucosal surfaces in our body is the first barrier encountered by drugs before their potential absorption through epithelial tissues, and could thus affect the drugs' permeability and their effectiveness. Therefore, it is of key importance to have in vitro permeability models that can mimic this specific environment. For this purpose, the novel mucus phospholipid vesicle-based permeation assay (mucus-PVPA) has been developed and used for permeability screening of drugs and formulations. The model proved to be stable under the chosen conditions and demonstrated the ability to discriminate between compounds with different chemical structures and properties. Overall, a decrease in drug permeability was found in the presence of mucus on top of the PVPA barriers, as expected. Moreover, mucoadhesive (chitosan-coated) and mucopenetrating (PEGylated) liposomes were investigated in the newly developed model. The mucus-PVPA was able to distinguish between the different liposomal formulations, confirming the penetration potential of the tested formulations and the related drug permeability. The mucus-PVPA model appears to be a promising in vitro tool able to mimic the environment of mucosal tissues, and could therefore be used for further drug permeability screening and formulation development.

Drug permeability profiling using cell-free permeation tools: Overview and applications.

Cell-free permeation systems are gaining interest in drug discovery and development as tools to obtain a reliable prediction of passive intestinal absorption without the disadvantages associated with cell- or tissue-based permeability profiling. Depending on the composition of the barrier, cell-free permeation systems are classified into two classes including (i) biomimetic barriers which are constructed from (phospho)lipids and (ii) non-biomimetic barriers containing dialysis membranes. This review provides an overview of the currently available cell-free permeation systems including Parallel Artificial Membrane Permeability Assay (PAMPA), Phospholipid Vesicle-based Permeation Assay (PVPA), Permeapad®, and artificial membrane based systems (e.g. the artificial membrane insert system (AMI-system)) in terms of their barrier composition as well as their predictive capacity in relation to well-characterized intestinal permeation systems. Given the potential loss of integrity of cell-based permeation barriers in the presence of food components or pharmaceutical excipients, the superior robustness of cell-free barriers makes them suitable for the combined dissolution/permeation evaluation of formulations. While cell-free permeation systems are mostly applied for exploring intestinal absorption, they can also be used to evaluate non-oral drug delivery by adjusting the composition of the membrane.

Drug permeability across a phospholipid vesicle based barrier: 3. Characterization of drug-membrane interactions and the effect of agitation on the barrier integrity and on the permeability.

Recently, we reported on the development and structural characterization of a phospholipid vesicle based barrier useful for medium throughput screening of passive drug permeability. Here, we investigate the physical and functional integrity of the phospholipid vesicle based barriers to agitation by stirring or shaking, and whether agitation affects drug permeability of sulpiride, metoprolol and testosterone. In addition, three drugs (caffeine, naproxen and sulphasalazine) which were shown in a previous study to affect the electrical resistance of the barriers, were investigated for their influence on the permeability of a simultaneously applied hydrophilic marker (calcein), and on the thermotropic phase transition of the phospholipid bilayers using differential scanning calorimetry (DSC). Electrical resistance measurements indicated that the barriers should withstand shaking speeds up to 150rpm without losing their integrity, but significant release of phospholipids from the membrane barriers to the donor and acceptor chambers was observed under agitation >or=150rpm. When using agitation up to 100rpm no increase in permeability was observed for sulpiride, metoprolol and testosterone. The phospholipid vesicle-based barrier thus differ from other permeability models in that agitation does not lead to an increase in permeability, not even for highly lipophilic drugs such as testosterone. This is explained by the different morphology of the vesicle-based barrier which is containing a 100microm thick layer of mostly aqueous compartments immobilised within a matrix of phospholipids vesicles. Sulphasalazine and naproxen were shown to decrease the electrical resistance and increase the permeability of the hydrophilic marker calcein. The DSC experiments showed that these two drugs probably interact with the head groups of the phospholipids. In contrast, caffeine gave an increase in electrical resistance and a decrease in permeability of calcein. From the DSC experiments no signs of interaction of caffeine with the phospholipid bilayer could be observed.