PL 407 hydrogels at 30% w/w were dispersed in deionized water at 4°C by magnetic stirring (150 rpm) for 12 h until complete dissolution. 15d-PGJ₂ was then solubilized in dimethyl sulfoxide (DMSO) and dispersed into the hydrogel at 15 ng/µl. The final DMSO concentration into the hydrogels was 0.015%, which is sufficiently low to avoid skin toxicity.

The 15d-PGJ₂-micelle interaction and micellar self-assembly were investigated using dynamic light scattering [(DLS; Nanoseries Zetasizer ZS-Malvern^® particle analyzer (Malvern Instruments, Ltd., Malvern, UK)] for determining the micellar hydrodynamic diameter and mean distribution size. For samples preparation, PL or PL-PGJ₂ systems (3% w/v) were filtered across a polycarbonate membrane (pore 0.22 µm) and measurements acquired at least three times for sample at a fixed 173° angle, at 25°C to 37°C.

Drug loading (DL, %) and entrapment efficiency (EE, %) parameters were determined for 3% PL micellar formulation. Aliquots (100 µl) were diluted in 0.02 M monobasic sodium phosphate pH 3.5/acetonitrile (60/40% v/v) solution and analyzed by HPLC method. DL, % (Eq. 1) and EE, % (Eq. 2) were determined as follow:

where C_PGJ2 is 15d-PGJ₂ concentration, C_PL is PL concentration, and C_total is the total PGJ₂ concentration into the samples.

Differential Scanning Calorimetry (DSC) was performed to determine temperature and enthalpy relative to micellization. The hydrogels (30 mg) were placed in sealed aluminum receptacles and underwent three heating-cooling cycles (0 to 50°C) at 5°C/min in a DSC equipment (Q-200; TA Instruments, New Castle, DE, USA). An empty receptacle was used as negative control. All thermograms were described as heat flux (cal/g) against temperature (°C).

The sol-gel transition temperature (Tsol-gel) and gelation kinetics were determined by an oscillatory rheometer (Kinexus Lab., Malvern Instruments, Ltd.) with a cone-plate geometry, under a temperature range from 10 to 50°C and frequency at 1 Hz. From the results, parameters related to the elastic (G′), viscous modulus (G″) and viscosity (η) were obtained and data analyzed by rSpace for Kinexus^® software.

For investigating the 15d-PGJ₂ release mechanisms from PL hydrogel, in vitro assays were carried out using a vertical two-compartment diffusion model Franz-type cells (1.76 cm² area, Microette Plus^®, Hanson Research, Chatsworth, CA, USA). An artificial membrane (cellulose acetate sheets, MWCO 1000 Da, Spectrum Lab) was used as a barrier for separating the two compartments. The donor compartment was filled with 250 µl of 15d-PGJ₂ (in ultrapure water) or PL404-PGJ₂. 15d-PGJ₂ final concentration of 3.75 µg/250 µl for both formulations. Receptor compartment was filled with 7.0 ml of 5 mM Hepes, 154 mM NaCl buffer (pH 7.4, at 37°C) and maintained under magnetic stirring (350 rpm). Aliquots of 1.0 ml were withdrawn from the receptor compartment at intervals from 0.5 to 24 h. Samples were analyzed by HPLC. Data were expressed as 15d-PGJ₂ released percentage against time (h).

Release profiles were then analyzed according to Zero-order (Eq. 3), Higuchi (Eq. 4) and Hixson-Crowell (Eq. 5) models, as described below:

where Qt is the cumulative amount of drug released at time t, Q0 is the initial amount of drug, K₀ is the zero-order release constant, and t is time.

where the rate of drug release is linear as a function of square root of time and the drug is the only component that diffuses through the medium, which the release mechanism is a diffusion process dependent on Fick law. K_H is the release coefficient, and Qt is the drug released amount.

Q₀ is the initial amount of drug, Q_t is the cumulative amount of drug released, K_HC is the release constant and t is time.

The 15d-PGJ₂ quantification was performed by High Performance Liquid Chromatography (Ultimate 3000 with Chromeleon 7.2 software; Dionex Corporation, Sunnyvale, CA, USA) system composed of quaternary pump, DAD detector and C18 column (150×4.6 mm, 5 µm-Phenomenex). Samples were detected at 216 nm, 0.6 ml/min flow rate (25°C) and mobile phase composed of 0.02 M monobasic sodium phosphate pH 3.5/acetonitrile (60/40 v/v). Drug retention time was 2.8 min. A calibration curve was obtained from standard solutions (2.5, 5, 50, 250 and 300 ng/ml). The detection (LOD) and quantification (LOQ) limits values were 0.063 and 0.189 ng/ml, respectively, obtained from the previously determined equation (y=0.8378 × + 3.339, R²=0.989).

This study was performed on male Wistar rats weighing 200 to 300 g (n=5/per group) and kept in cages (5 per cage) in a temperature-controlled room (23±1°C), 12:12 light cycle, with water and food ad libitum. All animals were obtained from the Multidisciplinary Center for Biological Investigation on Laboratory Animal Science (CEMIB-UNICAMP) and the experimentation was approved by the Committee on Animal Research of the University of Campinas (approval no. 4088-1), which followed the guidelines by the Brazilian National Council for Control of Animal Experimentation (CONCEA).

Induction of AD-like lesions was adapted from previously published guidelines (12). The DNCB is an aromatic hydrocarbon that when directly apply in the skin induce an inflammation. The skin from the dorsum of the rats was shaved to an area of 1×1 cm and painted once with 200 µl of 1% DNCB. Two weeks after sensitization, the target area on the skin was challenged with 200 µl of 0.2% DNCB solution twice weekly for 2 weeks. Subsequently, one of the following treatments was topically applied once daily over 14 days: i) No treatment; ii) vehicle (PL-407, 3 µl); iii) 15d-PGJ₂ hydrogel (75 ng/3 µl) or iv) Tacrolimus 0.1% (Tarfic^® 0.1%, Tacrolimus Monohydrate; Libbs Pharmaceutics, São Paulo, Brazil). The treatments were maintained at the site of lesion induction. When the experiment was complete, the animals were sacrificed by CO₂ inhalation and skin biopsies were harvested. The AD-protocol is summarized in Fig. 1.

A portion of the skin biopsies were fixed in neutral formalin and paraffin-embedded. Seven-micrometer tissue sections were taken and stained with toluidine blue for mast cell count.

Initial analysis of the toluidine blue sections were performed by four examiners (ABS, NM, MJ, PA) using a multi-headed microscope. Toluidine blue staining was evaluated both qualitatively and quantitatively. Qualitative analysis was performed via cell positivity in all areas of the section. Quantitative analysis of mast cells was performed by positive cell counting within the subepithelial connective tissue over 10 fields per case at magnification, ×400 (×40 objective lens, field diameter of 0.44 mm) using a CCD camera on a Nikon Eclipse Ci microscope. The individual who performed the counting was blind to the experimental groups.

Five-micrometer sections were immune-stained for ROR-γt and TNF-α. following endogenous peroxidase activity quenching in 3% hydrogen peroxide (Dinâmica, Diadema, SP, Brazil). Antigen retrieval (AR) was performed in boiling citrate buffer (pH 6.0). The primary antibody was incubated overnight at 4°C, followed by EnVision HRP and Envision+ (K1491; Dako; Agilent Technologies, Inc., Santa Clara, CA, USA) at 37°C for one hour. The sections were then stained with 3,3′-diaminobenzidine tetrahydrochloride (DAB, Dako; Agilent Technologies, Inc.) for five min at 37°C and counter-stained with hematoxylin.

ROR-γt and TNF-α expression was evaluated by inflammatory positive cell counting within the subepithelial connective tissue over 10 fields per case at magnification, ×400 (×40 objective lens, field diameter of 0.44 mm) using a CCD camera on a Nikon Eclipse Ci microscope. Epithelial cells were not taken into account for ROR-γt and TNF-α expression. The individual who performed the counting was blind to the experimental groups.

Whole blood was collected by cardiac puncture to quantify IgE levels following 15d-PGJ₂ hydrogel administration. The blood samples were stored in EDTA Vacutainer tubes containing EDTA (BD Biosciences, Franklin Lakes, NJ, USA) and blood plasma was then isolated. IgE measurements were obtained using ELISA following the manufacturer's instructions (BD Biosciences) via optical density (O.D.) measured at 450 nm and the readings were expressed as pg/ml, according to the standard.

To determine if there were significant differences (P<0.05) among groups, the data were analyzed using one-way analysis of variance (ANOVA) with post hoc contrasts using the Tukey's test. Data are presented in figures as mean ± standard deviation (SD). All statistical calculations were performed on GraphPad Prism 6^® (GraphPad Software, Inc., La Jolla, CA, USA).

The hydrodynamic diameter was the parameter used for evaluating the PL 407 micelles formation in the presence or absence of 15d-PGJ₂. In general, there were no overall significant changes on micellar hydrodynamic diameter for PL407 systems after 15d-PGJ₂ incorporation. Micellar diameters of ~60 nm (average distribution of 88.1±0.7%) and ~5 nm (12.1±0.2%) were observed at 25°C, while at 37°C, micellar dimensions were reduced to ~30 nm with 99.6±0.7% and polydispersion values of ~0.25, showing the influence of temperature variation on micellar self-assembly, even in the presence of 15d-PGJ₂. For DL % and EE % parameters were obtained values of 44.1±0.2 and 98.0±0.3%, respectively, indicating that PL407 micelles are able to carry high amounts of 15d-PGJ₂.

Calorimetric analysis (Table I) showed that micelles formation is an endothermic process (enthalpy values greater than zero), with micellization temperature (Tm) at 17.8 and 15.2°C, before and after 15d-PGJ₂ incorporation. Even slightly different Tm was observed for 15d-PGJ₂-PL407, this system presented high enthalpy variation (ΔH°=0.31 cal/g) than that observed for PL407 isolated system (ΔH°=0.21 cal/g).

For PL-based formulations, the rheological behavior provides essential information to study the hydrogel formation and the influence of incorporated molecules into its structure. For this reason, rheological parameters such as elastic (G′) and viscous (G″) moduli, viscosity (η) and Tsol-gel (when the most pronounced viscosity variation is observed) were determined before and after 15d-PGJ₂ insertion (Table I). Fig. 2 presents the rheograms for PL407 and PL407-15d-PGJ₂ under temperature variation. The incorporation of 15d-PGJ₂ did not, significantly, shift the Tsol-gel, but evoked pronounced changes on elastic modulus (G′) reaching values ~10 times higher than that observed for G″.

As demonstrated in Fig. 3 and Table II, the 15d-PGJ₂ release profiles and their mathematical models. The 15d-PGJ₂ release from aqueous solution reached a maximum release percentage after 4 h. On the other hand, the 15d-PGJ₂ released from PL407 hydrogels was sustained and lower drug release percentages were observed until 24 h (62.3%), when compared to 15d-PGJ₂ in solution (100%). In general, low release constant (Krel) values were obtained for PL407-PGJ₂ (1.4%.h⁻¹; 10.4%.h^−1/2; 0.15%.h^−1/3 for Zero Order, Higuchi and Hixson-Crowell models, respectively) in relation to 15d-PGJ₂. However, the Hixson-Crowell mathematical model showed the highest correlation coefficient value (R²=0.97) compared to Zero Order (R²=0.95) and Higuchi (R²=0.91).

To establish whether 15d-PGJ₂ hydrogel reduces mast cell infiltration into AD-like skin lesions, toluidine blue staining was performed on the skin biopsies following topical administration of 15d-PGJ₂ or vehicle (Fig. 4). Mast cell infiltration was detected in the AD-like group and the AD-like group treated with vehicle (PL-407), where the 15d-PGJ₂ hydrogel significantly decreased (P<0.05) such infiltration of mast cells into the skin when compared with AD-like and AD-like + PL-407 (Fig. 5). Moreover, the group treated with Tacrolimus 0.1% also decreased mast cell infiltration, although no statistically significant difference was detected when compared to both untreated groups. The data on mast cell counts is shown in Fig. 5.

High IgE levels are a major feature of AD and those diagnosed with AD often exhibit high levels of total IgE and also allergen-specific IgE. Serum levels of IgE in the AD-like group and AD-like treated with vehicle were significantly higher than that in the disease-free group. The administration of topical 15d-PGJ₂ hydrogel or Tacrolimus 0.1% significantly reduced the serum levels of IgE (P<0.05) compared to AD-like groups (Fig. 6).

To determine whether 15d-PGJ₂ hydrogel decreases Th17 type lymphocyte, we performed immunohistochemistry to quantify the transcription factor ROR-γt (Fig. 4). We found that topical administration of 15d-PGJ₂ hydrogel significantly decreased the number of immunostained cells compared to the AD-like group (P<0.05). No significant difference was observed between the AD-like group and the Tacrolimus-treated group (P>0.05), nor was it observed between the 15d-PGJ₂ hydrogel and the Tacrolimus groups (P>0.05). The data are shown in Fig. 7A.

Immunohistochemistry was performed to verify whether 15d-PGJ₂ hydrogel would reduce TNF-α expression (Fig. 4). The number of immunostained cells in the AD-like group and the AD-like group treated with vehicle significantly increased in comparison to the disease-free specimens (P<0.05). Moreover, topical administration of 15d-PGJ₂ hydrogel significantly reduced the TNF-α-positive cell count from the AD-like groups (P<0.05). It is important to highlight that no significant difference was detected between the AD-like group and the Tacrolimus-treated group (P>0.05). The data are described in Fig. 7B.