Oral treatment with Aloe polysaccharide ameliorates ovalbumin-induced atopic dermatitis by restoring tight junctions in skin

2.1 Animals

7-week-old female Balb/c mice were used for all experiments (Orient). The mice were housed in a specific pathogen-free barrier facility at Inha University. Animal care and all experimental procedures were conducted in accordance with the Guide for Animal Experiments published by the Korean Academy of Medical Science. All animal experiments were approved by the Institutional Animal Care Use Committee (IACUC) at Inha University.

2.2 Preparation of processed A vera gel

Processed A vera gel (PAG) was prepared as previously described and then characterized.23-25 Briefly, A vera gel was reacted with cellulase and then passed through a charcoal column to remove anthraquinones and other coloured substances. Finally, it was subjected to ultrafiltration in an Amicon Stirred Cell (MWCO, 5000 Da; Merck). The yield of PAG from A vera gel is ~0.5%. In order to eliminate the possibility of endotoxin contamination, we removed it using a previously reported method.23 It was confirmed that PAG contained <0.1 ng of endotoxin/mg PAG. In addition, we used a TSKgel G4000 PWxl column (7.8 mm × 30 cm; Tosoh Bioscience) to determine the molecular weight distribution of PAG. As described by Im et al,24 the molecular weight distribution of PAG polysaccharides was found to be much smaller than that of native A vera gel. The molecular weight of most polysaccharides contained in A vera gel was >1000 kDa. PAG used in this study consisted of more than 75% 10-500 kDa polysaccharides. PAG was dissolved in 1× phosphate-buffered saline (PBS) for in vivo experiments and in culture media for in vitro experiments.

2.3 AD induction

AD was induced as previously described, with some modifications.26, 27 Briefly, eight mice were used per group. The mice in the group except negative control (NC) were intraperitoneally immunized with 10 μg of chicken OVA (grade V; Sigma-Aldrich) mixed with 4 mg aluminium hydroxide (Imject®Alum; Pierce) in a volume of 150 µL on days 0 and 7. The third immunization was administered in the loose skin over the neck on day 14. On day 20, mice were anaesthetized with a mixture of 10 mg/kg zoletil (Virbac Laboratory) and 10 mg/kg xylazine (Bayer Korea Ltd.). During anaesthesia, dorsal skin was shaved with an electric clipper and hair removal creams. After 24 hours, the mice were epicutaneously sensitized with OVA patches. OVA (10 µg) prepared in 50 µL PBS was contained in a 1 × 1 cm patch of sterile gauze, which was placed on the shaved backs of mice and secured to the skin with a transparent dressing (Tegaderm™). The patch was changed daily, and the skin was kept in contact with OVA for 11 days (ie from days 21 to 31). Then, the indicated concentration of PAG (100 mg/kg for 4 times or 50 mg/kg and 10 mg/kg for 8 times) was administered orally to the mice. After PAG treatment, mouse dorsal skin was shaved again on day 42 and the skin was re-exposed to OVA for 7 days (ie from days 43 to 49). To adjust the AD severity in all mice groups, blood samples were harvested on day 32 before PAG treatment and total IgE levels were analysed using ELISA. Mice were grouped according to their IgE levels. IgE levels were similar in all groups (Figure 2A). On day 49, the mice were sacrificed by CO₂ asphyxiation. After sacrifice, sera and skin biopsy specimens were obtained for ELISA and histological examination, respectively.

2.4 Histology

Skin tissues from each mouse were filled with 10% neutral-buffered formalin and immersion fixed for 16 h at 4°C. Tissues were dehydrated, embedded in paraffin and sectioned at 5 μm. Mast cells, T cells and eosinophils in the skin were assessed using toluidine blue (Sigma–Aldrich), anti-CD3 antibodies (BD Bioscience), and haematoxylin and eosin (H & E) staining, respectively. Cells were counted from three randomly selected fields per slide using a microscope. Two slides per mouse were examined. The three fields were first identified by scanning the entire slide at ×100 magnification. Positively stained cells were counted in a high-powered field (HPF, ×400). Skin thickness analysis was performed on H & E staining photographs from individual mice skin at ×100 magnification. Dermal thickness was defined as the distance from the epidermal-dermal junction down to the adipose tissue of hypodermis. Epidermal thickness was defined as the distance from the epidermal-dermal junction up to the stratum corneum. Using Image J software (Image Processing and Analysis in Java, NIH), thickness measurements were obtained on light microscopy images by drawing a line to the average orientation. Four thickness measurements were taken per slide. The pixel-based thickness measurement was converted to micrometres. Antigens were detected with anti–Ki-67, claudin-1 (Abcam) and IL-4 (Abbiotec). Antigen-antibody complexes were detected with Alexa594-conjugated secondary antibodies (Molecular Probes). We performed control staining with secondary antibody alone (data not shown). Slides were counterstained with 1 μg/mL 4′,6-diamidino-2-phenylindole (DAPI). Images were acquired using a confocal microscopy (FluoView FV1000). For analysis of claudin-1, fluorescence area analysis was performed on immunofluorescence staining photographs from four mice skins at ×400 magnification. Epidermal area of the skin and positive fluorescence area were selected using the selection tool in ImageJ. Then, percentages of positive area per total epidermal area were calculated. The calculation is following; % of positive staining intensity = expression area/total epidermal area ×100. Ki-67-positive cells were counted at ×400 magnification of the immunofluorescence staining photographs in the basal layer using the selection tool in ImageJ.

2.5 Cytokine assay

To evaluate the effect of PAG on naïve T cells, lymphocytes were isolated from lymph nodes (inguinal, brachial, superficial cervical and axillary) in naïve Balb/c mice. To detect IL-4 cytokines, purified CD4⁺ T cells (1 × 10⁶) were stimulated with 1 μg/mL of coated anti-CD3 (145-2C11) and soluble anti-CD28 (37.51) antibodies in the presence of PAG in a 24-well plate. To detect IFN-γ and IL-17A, total lymphocytes from lymph nodes were stimulated with 1 μg/mL soluble anti-CD3 and anti-CD28 antibodies. After 48 h of stimulation, cytokine levels were measured by ELISA according to the manufacturer's protocol (BD Bioscience). To evaluate the effect of PAG on antigen-specific T cell response, we first immunized Balb/c mice twice with OVA/alum at days 1 and 7. At day 15, after the 1st immunization, lymphocytes were isolated from the spleen and draining lymph nodes and then cultured with different concentrations of OVA and PAG. After 48 h, IL-4 levels were measured in culture media by ELISA.

2.6 Isolation of RNA and RT-PCR analysis

To evaluate the mRNA levels of tight junction genes, we separated only the epidermis from AD skin lesions. To measure the mRNA levels of cytokines, skin lesions with epidermis and dermis were used. Total RNA was extracted from biopsied skin using easy-BLUE^TM (iNtRON), and cDNA was prepared using a QuantiTect Reverse Transcription Kit (Qiagen), according to the manufacturer's instructions. cDNA products were amplified using SYBR Green Quantitative PCR Master Mix (Applied Biosystems). The primers (QuantiTect Primer Assays) are all commercially available and validated by QIAGEN. Reactions were carried out in the StepOnePlus™ Real-Time PCR System (Applied Biosystems); the PCR running conditions were as follows: for activating the DNA polymerase, hot start was performed for 10 min at 95°C and then cycling at 95°C for 15 s and 60°C for 1 min, for a total of 40 cycles.

2.7 Statistical analysis

Statistical significance was determined by Student's unpaired t tests. Error bars represent mean ± SEM. Asterisks indicate statistical significance, where P < .05 was used as significance cut-off. Levels of statistical significance are indicated in the figure legends.

3.1 PAG ameliorates ovalbumin (OVA)-induced AD in mice

To investigate the effect of PAG on OVA-induced AD, mice were administered 10 μg OVA with aluminium hydroxide on days 0, 7 and 14 (Figure 1A). Epicutaneous sensitization was carried out using patches with 10 μg OVA in 50 μL PBS and comprised two 1-week exposures to OVA. PAG was administered orally between epicutaneous sensitizations. Mice received three different doses of PAG: 100 mg/kg every two days for 4 times or 50 or 10 mg/kg every day for 8 times. Two groups received 400 mg/kg PAG and one group received 80 mg/kg PAG in total. Mice were killed on day 49 after the second epicutaneous sensitization procedure (Figure 1A). Inflammation in skin lesions and immunoglobulin levels in sera were analysed. Histopathological examination revealed an amelioration of epidermal hypertrophy and hyperkeratosis in PAG-treated mice. H & E staining indicated decreased cell infiltration in the skin of PAG-treated mice compared with positive control (PC) mice (Figure 1B, C). Treatment with 100 or 50 mg/kg PAG resulted in decreased epidermal thickness, whereas treatment with 10 mg/kg PAG did not. Furthermore, there were no significant effects of PAG on dermal thickness or number of mast cells, T cells or eosinophils (Figure 1C, D). Serum immunoglobulin levels were measured by ELISA. Total and OVA-specific IgE levels significantly decreased only in the mice treated with 100 mg/kg PAG (Figure 2B, C). IgG1 and IgG2a levels were not affected by PAG treatment.

3.2 PAG inhibits keratinocyte proliferation and restores tight junction expression

As we observed that treatment with PAG decreased epidermal thickness, we stained the skin tissues with Ki-67, a cell proliferation marker. Consistent with the H & E staining assessments, increased Ki-67 staining was inhibited by PAG in a dose-dependent manner. The least staining was observed in the 100 mg/kg PAG-treated skin, whereas the most staining was observed both in the PC and 10 mg/kg PAG-treated skin. There was more Ki-67 staining in the epidermal field than in the dermal field (Figure 3A). We quantified Ki-67-positive cells and observed decreased cell number in 100 and 50 mg/kg PAG-treated skin compared with the PC mice (Figure 3B).

CLDN1, a major component of TJs in the epidermis, which plays a key role in barrier function, is decreased in the skin of patients with AD.9, 28, 29 Therefore, we stained AD skin with CLDN1 antibodies. Consistent with the results of previous studies, we observed decreased CLDN1 expression in AD skin lesions. CLDN1 expression increased in 100 and 50 mg/kg PAG-treated skin lesions, whereas it decreased in the outer layer compared to the inner layer of 10 mg/kg PAG-treated skin lesions (Figure 4A). We quantified the percent of claudin-1 positive area per total epidermal area. As shown in Figure 4B, treatment of 100 and 50 mg/kg PAG can restore the claudin-1 expression. In addition, we also measured other tight junction gene mRNA levels. Expression of the tight junction components ZO-1 (TJP1), claudin-1 (CLDN1) and claudin-8 (CLDN8) was decreased in AD epidermis compared with normal epidermis. These decreases were partially rescued by treatment with PAG (Figure 4C), suggesting that PAG ameliorates AD through restoration of tight junction gene expression.

3.3 PAG suppresses the production of pro-inflammatory cytokines in AD skin lesions

Inflammatory cytokines are known to be involved in AD. We measured the expression of IFN-γ, IL-4 and IL-17A in AD skin lesions. The mRNA expression of IFN-γ does not significantly change in any group. Consistent with AD symptoms, mRNA levels of IL-4 and IL-17A were increased in AD, and PAG treatment decreased these cytokine levels (Figure 5A). We could also detect lower number of IL-4-positive cells in skin lesions of PAG-treated mice compared with PC mice in a dose-dependent manner (Figure 5B, C). In addition, we investigated whether PAG treatment affected the OVA-specific response. We immunized mice twice with OVA/alum at days 1 and 7. At day 15 after the 1st immunization, lymphocytes were isolated from the spleen and draining lymph nodes and then cultured with different concentration of OVA and PAG. After 48 h, IL-4 levels in the culture media were measured by ELISA. As shown in Figure 5D, we could observe decreased IL-4 expression by PAG treatment in a dose-dependent manner.

Furthermore, we isolated lymph nodes from naïve mice and total lymphocytes were stimulated with anti-CD3 and anti-CD28 antibodies in the presence or absence of different concentrations of PAG. Cytokines in the culture media were measured by ELISA. IL-4 and IFN-γ levels were significantly suppressed in a dose-dependent manner (Figure 5E). IL-17A significantly decreased after treatment with 2 mg/mL PAG. These results indicated that PAG suppresses the production of AD-related pro-inflammatory cytokines, not only in total T cells but also in ova-specific T cells.